Uploaded by debbytu0428


Shirley Tilghman, President
Julie Theriot, Program Chair
Karen Oegema, Local Organizer
Search Your Smartphone
App Store for ASCB2015
Oral Presentations‐Sunday, December 13
Symposium 1: Pushing the Limits: Visualization of Hidden Biological Processes
In vivo imaging of cellular dynamics from the nanoscale to the macroscale.
E. Betzig1; 1Janelia Research Campus, Ashburn, VA
The hallmark of life is that it is animate. To gain a better understanding of how inanimate molecules
assemble to create animate life, it is necessary to image living organisms noninvasively at high resolution
in both space and time. However, the imaging of biological specimens involves inevitable tradeoffs of
spatial resolution, speed, non‐invasiveness, and imaging depth. I will describe three methods that
balance these tradeoffs in different ways: structured illumination microscopy at 50‐80 nm resolution,
which we apply to study endocytic and cytoskeletal dynamics at the plasma membrane; lattice light
sheet microscopy, which we use to image the rapid three‐dimensional dynamics of single molecules,
cells and embryos at hundreds of image planes per second; and adaptive optics, which we use to
recover optimal resolution of fine neural processes deep in the brains of zebrafish and mice.
Illuminating biology at the nanoscale with single‐molecule and super‐resolution fluorescence
X. Zhuang1; 1Department of Chemistry and Chemical Biology, HHMI/Harvard University, Cambridge, MA
Dissecting the inner workings of a cell requires imaging methods with molecular specificity, molecular‐
scale resolution, and dynamic imaging capability such that molecular interactions inside the cell can be
directly visualized. Fluorescence microscopy is a powerful imaging modality for investigating cells largely
owning to its molecular specificity and dynamic imaging capability. However, the spatial resolution of
light microscopy, classically limited by diffraction to a few hundred nanometers, is substantially larger
than molecular length scales in cells, making many sub‐cellular structures difficult to resolve. We
developed a super‐resolution fluorescence microscopy method, stochastic optical reconstruction
microscopy (STORM), which overcomes the diffraction limit by using photo‐switchable fluorescent
probes to temporally separate the spatially overlapping images of individual molecules. This approach
has allowed multicolor and three‐dimensional imaging of living cells with nanometer‐scale resolution
and enabled discoveries of novel sub‐cellular structures. In this talk, I will discuss the technological
advances of STORM and biological discoveries enabled by STORM.
I will also describe a single‐cell transcriptome imaging method that we recently developed. System‐wide
analyses of the abundance and spatial organization of RNAs in single cells promise to transform our
understanding in many areas of cell and developmental biology, such as the mechanism of gene
regulation, the heterogeneous behavior of cells, and the development and maintenance of cell fate.
Single‐molecule imaging approaches are powerful tools for counting and mapping RNA; however, the
number of RNA species that can be simultaneously imaged in individual cells has been limited, making it
challenging to perform transcriptome‐scale analysis of single cells in a spatially resolved manner. To
overcome this challenge, we developed a transcriptome imaging approach, multiplexed error‐robust
fluorescent in situ hybridization (MERFISH), which allows numerous RNA species to be localized and
quantified in single cells in situ. In this talk, I will also discuss the technology development and
applications of MERFISH.
The story of single molecules, from early spectroscopy in solids, to super‐resolution microscopy,
to 3D dynamics of biomolecules in cells.
W.E. Moerner1; 1Chemistry, Stanford University, Stanford, CA
More than 25 years ago, low temperature experiments aimed at establishing the ultimate limits to
optical storage in solids led to the first optical detection and spectroscopy of a single molecule in the
condensed phase. At this unexplored ultimate limit, many surprises occurred where single molecules
showed both spontaneous changes (blinking) and light‐driven control of emission, properties that were
also observed in 1997 at room temperature with single green fluorescent protein variants. In 2006,
PALM and subsequent approaches showed that the optical diffraction limit of ~200 nm can be
circumvented with single molecules to achieve super‐resolution fluorescence microscopy with relatively
nonperturbative visible light. Essential to this is the combination of single‐molecule fluorescence
imaging with active control of the emitting concentration and sequential localization of single
fluorophores decorating a structure. Super‐resolution microscopy has opened up a new frontier in
which biological structures and behavior can be observed in fixed and live cells with resolutions down to
20‐40 nm and below. Examples range from protein superstructures in bacteria to details of the shapes
of amyloid fibrils and much more. Current methods development research addresses ways to extract
more information from each single molecule such as 3D position and orientation, and both of these can
be obtained by proper point‐spread function engineering of a wide‐field microscope. It is worth noting
that in spite of all the current focus on super‐resolution, even in the “conventional” low concentration,
single‐molecule tracking regime where the motions of individual biomolecules are recorded rather than
the shapes of extended structures, much can still be learned about biological processes. For example,
my laboratory has explored the motions of single Smoothened proteins in the primary cilium, where the
molecular motions show clear evidence of binding sites at the ciliary base whose affinity is modulated by
Hedgehog pathway activation. Using 3D precision tracking at high speed, correlations in the motions of
pairs of DNA loci in the yeast nucleus highlight the complexity of the nuclear environment through the
appearance of subdiffusive motion. I warmly thank all members of the Moerner Lab and all
collaborators for their contributions to this work.
Symposium 2: Wisdom of Crowds: Collective Decision‐Making by Cells and
Collective cell migration: the power of many.
R. Mayor1; 1Cell and Developmental Biology, University College London, London, United Kingdom
Collective behaviour is commonly found in nature and it has been widely studied in animal groups such
as bird flocks, fish schools and insect swarms. From these studies a picture has developed in which
collective movement of animals emerges as consequence of local interaction based on the ability of
each animal to sense their neighbours and respond accordingly. This collective migration has important
consequences for the group as it allows a more efficient response to external stimulus as depredators.
An apparently equivalent collective behaviour has been described in recent years for the migration of
cells. However most of these studies have focused on the migrations of epithelial tissues with strong
cell‐cell adhesion, where individual cell behaviour is unlikely and therefore different to the collective
movement based on the behaviour of individual animals. We have discovered that individual cells, such
as mesenchymal neural crest cells, can also exhibit collective cell migration. Here we will discuss the
cellular and molecular basis of this collective cell migration. Neural crest is an embryonic cell population
whose invasive behaviour has been likened to cancer metastasis. We showed that each neural crest cell
sense their neighbours by a combination of transient cell‐cell adhesion and short range chemotaxis. The
interaction with its neighbours leads to cell repolarization, which is at the basis of collective cell
migration. Similar to the more efficient response of animals to depredators, we show that collective
cell migration allows a more efficient response to external signals such as chemoattractants and
chemorepressors. Finally, we have developed a mathematical model that integrates the behaviour of
individual cells to generate an efficient directional collective cell migration of neural crest.
The ecology of collective behavior.
D.M. Gordon1; 1Biology, Stanford University, Stanford, CA
Like many biological systems, an ant colony operates without central control using networks of local
interactions. Ant species differ in collective behavior, reflecting diversity in ecology, including the
distribution of resources and the rates at which resources are obtained and used. In ants as in all natural
systems, the network motifs and feedback loops that regulate collective behavior evolve in relation to
patterns of environmental change. For example, desert harvester ants and tropical arboreal turtle ants
differ in the algorithms that use local interactions to regulate foraging activity. The relevant interactions
are olfactory, including antennal contact in which one ant assesses the cuticular hydrocarbons of the
other, and detection by one ant of a pheromone recently deposited by another. Using feedback from
simple interactions, colonies search for and collect food, create and prune trail networks, respond to
competing species, and adjust foraging activity to food availability. How interactions provide feedback
for each ant species corresponds to the challenges of that species' environment. Similar ecological
constraints, in many natural systems from cells to ants, may correspond to similar algorithms that
regulate collective outcomes, such as search, group movement, and the exclusion of enemies.
Microsymposium 1: Cell Motility and Migration
Cancer‐associated fibroblasts promote directional migration of cancer cells via parallel
organization of the fibronectin matrix.
B. Erdogan1, M. Ao1, B.M. Brewer2, O.E. Franco3,4,5, S.W. Hayward3,4,5, D. Li2, D.J. Webb1,5; 1Biological
Sciences, Vanderbilt University, Nashville, TN, 2Mechanical Engineering, Vanderbilt University, Nashville,
TN, 3Urologic Surgery, Vanderbilt University, Nashville, TN, 4Surgery, NorthShore University
HealthSystem, Evanston, IL, 5Cancer Biology, Vanderbilt University, Nashville, TN
Cancer cell migration is a critical step in cancer progression, and is required for metastasis. An emerging
concept is that cell movement is influenced by cells in the surrounding tumor stroma, pointing to the
importance of the tumor microenvironment in disease progression. Fibroblasts are prominent
components of the tumor stroma; however, the mechanisms by which they regulate cancer cell
migration are not well understood. To study this, we used microfluidic devices to co‐culture cancer cells
with primary fibroblasts isolated from human prostate cancer tissues (cancer‐associated fibroblasts,
CAF) or normal tissue fibroblasts (NAF) and performed time‐lapse microscopy. We observed that CAF
promote directionally persistent migration of cancer cells, which is not observed in co‐cultures with NAF.
An important function of fibroblasts is secretion and organization of the extracellular matrix (ECM);
therefore, we hypothesized that directional migration in co‐cultures with CAF could be due to changes in
the ECM organization. Fibronectin (Fn) is a major ECM protein secreted by fibroblasts. We investigated
Fn matrix organization in CAF and NAF. We observed that CAF organize Fn into parallel fibers, whereas
NAF assemble Fn into a fibrillar meshwork. To determine whether CAF‐mediated matrix organization
plays a role in the directional migration of cancer cells, cell‐derived matrices were generated. DU‐145
prostate cancer cells, and SCC61 and JHU012 head and neck cancer cells were plated onto CAF/NAF‐
derived matrices, and their migration directionality was assessed. Cancer cells exhibited directionally
persistent migration along the aligned Fn fibers of the CAF‐generated matrix, suggesting that the
directional migration of cancer cells is due to the topography of the CAF‐derived matrix. We next
hypothesized that the changes in the ECM organization by CAF could be due to altered traction forces
generated by CAF. Analysis of cellular traction forces using traction force microscopy revealed that CAF
exert higher traction stresses on the substratum compared to NAF. Traction forces are generated by
actin‐myosin‐II contractions, which are regulated by Rho GTPases such as Rac. Integrins act as
transducers of these forces to the Fn matrix. Our results show that CAF have elevated levels of active
α5β1 integrin (a major Fn integrin) and active Rac compared to NAF. Furthermore, inhibition of myosin‐II
activity, using blebbistatin, changed the Fn matrix generated by CAF from an aligned into a meshwork‐
like organization. Collectively, our results suggest that CAF promote directional migration of cancer cells
via parallel organization of the Fn matrix, in an α5β1 integrin, Rac and myosin‐II‐dependent mechanism.
Centrosomes define the rear of migrating cells by modulating the distribution of inhibitory
J. Zhang1, Y. Wang1; 1Biomedical Engineering, Carnegie Mellon University, Pittsburgh, PA
The centrosome is conventionally believed to localize in front of the nucleus during directional migration
to define the front of the cell. This view, however, has been challenged by observations that centrosome
position relative to the nucleus and cell polarity is variable among cell types and environmental
conditions, and even for the same cell over time. To clarify the role of the centrosome in directional
migration, we studied the effect of centrosome localization on the polarity of RPE‐1 cells migrating on
2D surfaces and along micropatterned tracks, taking advantage of the well‐defined polarity, ease of the
control of migration, and amenability to microamputation for cells migrating along 1D. Centrosomes
were visualized in living cells by expressing GFP‐centrin. We found that while the relationship between
the centrosome and the nucleus was variable relative to the direction of migration, centrosomes were
always localized behind the centroids of cells in both 1D and 2D. This relationship also held for
enucleated cells, which migrated as persistently as control cells as long as the centrosome was present.
However, cells without a centrosome showed primarily oscillatory movements, suggesting that the
centrosome is necessary for defining migration polarity and that the nucleus is dispensable. Moreover,
during the initial establishment of directional migration, known as symmetry breaking, the end closer to
the centrosome turned into the tail of most cells plated on 1D tracks. The same relationship held when
the position of the centrosome relative to the two ends was changed microsurgically. Similarly, as a cell
on a 1D track reached the end of the track, migration stopped for a variable period of time and
retraction of the stalled leading edge and reversal of migration direction happened only after the
centrosome had relocated across the cell centroid from the opposite end. Longer cells took more time
to reverse the direction, likely as a result of the increased time required in relocating the centrosome. To
validate the intuitive interpretation that the centrosome defines the rear of a migrating cell, we
developed a computational model that allowed the centrosome to modulate the distribution of
inhibitory signals under the general framework of a local‐enhancement‐global‐inhibition mechanism.
The simulation was able to validate all the experimental phenomena, such that the edge closer to the
centrosome has an increased probability of retraction. Together, our results demonstrate that the
localization of centrosomes defines the rear rather than the front of cells undergoing directional
A novel actin‐adhesion structure requiring the formin FMN2 positions the nucleus and protects
it from DNA damage during confined migration.
C.T. Skau1, H. Racine Thiam1, G.M. Alushin1, P. Gurel1, A. Tubbs2, M. Piel3, A. Nussenzweig2, C.M.
Waterman1; 1National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD,
National Cancer Institute, National Institutes of Health, Bethesda, MD, 3Systems Cell Biology of Cell
Polarity and Cell Division, Institut Curie, Paris, France
Previous studies have shown that the nucleus is mechanically coupled to the extracellular matrix (ECM)
via actin, actin‐binding proteins, and integrins during cells migration. However, it has so far been unclear
what the precise role of this coupling is. Force transmission to the nucleus has been shown to
mechanically initiate signaling events and control chromatin organization, but the contribution of the
actin cytoskeleton versus nuclear lamins in this process is poorly understood. Similarly, coupling the
nucleus to the cytoskeleton can control the position of the nucleus within the cell, although it is not
known if this is mediated by linking the nucleus to the previously‐characterized adhesion/actin system
used for migration, or if nucleus‐ECM coupling is accomplished via a specialized cytoskeletal system. We
examined the interplay between actin, adhesions, and the nucleus in fibroblasts using fluorescence
microscopy. We identified novel adhesion structures located underneath the nucleus termed subnuclear
adhesions that are compositionally and dynamically distinct from the canonical focal adhesions at the
leading edge. Subnuclear adhesions are nucleated along an existing actin fiber, in contrast to leading
edge adhesions. We show that the actin fibers connecting two subnuclear adhesions can control nuclear
shape by physically impinging on the nucleus. These subnuclear fibers have elevated levels of the IIB
isoform of myosin but reduced levels of α‐actinin as compared with dorsal stress fibers, and are less
dependent on the contractile activity of myosin than dorsal stress fibers. Furthermore, subnuclear fibers
are independent of the Arp2/3 complex but dependent on activity of the formin FMN2. We find that
FMN2 localizes as thin dynamic fibrils underneath the nucleus in cells plated on 2 dimensional
substrates, partially co‐localizes with subnuclear actin bundles, and in a cup‐like organization around the
rear of the nucleus in cells in 3D collagen gels. Critically, FMN2 is essential for both subnuclear actin and
adhesions; loss of FMN2 eliminates both structures. These cells exhibit defects in nuclear positioning
and show increased double‐stranded breaks in DNA upon mechanical compression while migrating. We
therefore propose a unique role for the novel actin‐adhesion system generated by FMN2: The nucleus‐
associated actin‐adhesion system functions to protect the nucleus from mechanical damage during
confined migration as well as control the position of the nucleus in migrating cells. This is the first
demonstration of an actin‐adhesion system responsible for protecting the nucleus from physical damage
in migrating fibroblasts.
Reconstitution of tumor microenvironment‐associated high‐speed breast cancer cell motility on
aligned nanofibers.
V.P. Sharma1,2, J. Williams3, E. Leung1, J. Sanders3, R.J. Eddy1, J. Castracane3, J.S. Condeelis1,2; 1Anatomy
and Structural Biology, Albert Einstein College of Medicine, Bronx, NY, 2Gruss Lipper Biophotonics
Center, Albert Einstein College of Medicine, Bronx, NY, 3Colleges of Nanoscale Science and Engineering,
SUNY Polytechnic Institute, Albany, NY
One of the key features of primary breast tumor progression is the extensive ECM remodeling and the
presence of aligned collagen fibers oriented perpendicular to blood vessels. These aligned fibers provide
topography for the rapid migration of single tumor cells (streaming migration) to invade the surrounding
stroma, intravasate into blood vessels and form distant metastases. Here, inspired by the in vivo tumor
microenvironment ECM architecture (i.e. alignment, topography and composition) and intravital movies
of tumor cell streaming migration on collagen fibers in vivo, we developed a high fidelity reconstitution
in vitro system to study of tumor cell streaming migration. To faithfully capture the ECM fiber
topography in our in vitro system, we investigated the in vivo tumor ECM architecture in two mice
models of breast cancer ‐ PyMT and rat mammary carcinoma MTLn3, using second harmonic intravital
imaging. We found a wide distribution of collagen I fiber diameters with a peak falling in the range, 2‐3
µm. Moreover, we found that tumor cells in vivo prefer to exhibit streaming migration on 2‐3 µm fibers.
Based on these in vivo observations, we engineered cylindrical poly(lactic‐co‐glycolic acid) (PLGA)
nanofibers matching in vivo dimensions and coated them with fluorescently‐labeled ECM molecules
(collagen I and/or fibronectin). Carcinoma cells plated on ECM‐coated nanofibers showed enhanced
motility compared to motility of these cells on 2D flat surfaces. The average tumor cell speed on 2 µm
thick fibers was 1.2 µm per min, consistent with previous in vivo observations. Employing both
nanofibers and micro‐patterned 1D stripes, we evaluated the effect of varying 1D diameter (from 0.7 µm
to 20 µm) on tumor cell migration characteristics and found that tumor cells move the fastest (~1.2 µm
per min) with highest persistence on smaller 1D thickness (0.7‐3 µm) range, with numbers approaching
the slow 2D migration values as the 1D thickness is increased to 20 µm. Similarly, F‐actin fibers were
highly orientated along 1D dimension at smaller 1D diameters and approached random orientation with
increasing 1D diameters. Interestingly, we also observed nuclear deformation during carcinoma cell
migration on fibers, similar to the nuclear deformation observed in cells in vivo, suggesting that
carcinoma cell nucleus is inherently plastic and ECM space constraint is not a requisite for nuclear
deformation. In summary, our 1D nanofiber bioassay, matching key in vivo ECM characteristics, could
provide a diagnostic platform for high‐throughput screening of experimental drugs targeting fast‐
moving cancer cells during metastasis.
CAMSAP2 and CAMSAP3 Block Trailing Edge MT Disassembly and Nucleate Leading Edge MT
growth during Endothelial Cell Polarization and Migration.
P. Jones1, K.A. Myers1; 1Biological Sciences, University of the Sciences, Philadelphia, PA
Angiogenesis is the process by which endothelial cells alter their morphology, polarize and migrate to
form new blood vessels. This process requires the coordinated reorganization of the microtubule (MT)
cytoskeleton. MTs are intrinsically polar cytoskeletal structures, consisting of a free, peripherally
roaming plus end and a minus end that is free or anchored. Calmodulin‐regulated Spectrin Associated
Proteins (CAMSAPs) 1, 2, and 3 have been shown to bind to and stabilize the free minus ends of MTs, as
well as to nucleate new MT growth away from the centrosome; yet, the contribution of this activity to
MT organization, cell polarity, and directional migration has not been investigated. We hypothesized
that CAMSAP association with free minus ends of MTs functions to promote polarized organization of
the MT array and thereby contribute to cell polarity and directed migration. To test this hypothesis, we
performed live‐cell imaging of CAMSAP localization and MT growth dynamics during wound‐edge
polarization of Human Umbilical Vein Endothelial Cells (HUVECs) expressing fluorescently‐labeled
CAMSAP2 or 3. Our results show that both CAMSAP2 and CAMSAP3 are localized to the peri‐
centrosomal regions of the cell and specifically to the trailing edge of wound‐edge HUVECs.
Pharmacologic disassembly of MTs with nocodazole resulted in CAMSAP dissociation from trailing edge
MTs followed by the formation of new CAMSAP‐nucleated MT growth at the leading edge of polarized
HUVECs. Unlike nocodazole‐induced MT disassembly, co‐expression of CAMSAP2 or CAMSAP3 with
MCAK, a MT depolymerizing protein, resulted in longer stretches of CAMSAP2 and CAMSAP 3 along the
MT lattice, and inhibited MCAK‐induced MT disassembly specifically at the plus‐end tip of the CAMSAP
stretch. This result supports in‐vitro investigations and suggests that the in‐vivo function of CAMSAP2
and 3 is to protect MT minus ends from MCAK‐induced disassembly. Measurements of MT growth
dynamics revealed that both CAMSAP2 and 3 promoted fast and short‐lived (aka “dynamic”) MT growth.
Regional analysis of MT dynamics revealed that CAMSAP2 specifically promoted fast MT growth within
the leading edge of wound‐edge HUVECs, while CAMSAP2 and CAMSAP3 significantly reduced MT
growth lifetimes without regional specificity. Migration studies revealed overexpression of either
CAMSAP2 or CAMSAP3 increased cellular migration velocity (36% and 174%, respectively), and increased
directional migration distance to origin (41% and 72%, respectively). Together, these results suggest that
during HUVEC polarization and wound‐healing, CAMSAP2 and 3 function to protect trailing edge MTs
against MCAK‐mediated disassembly and also function to nucleate new, dynamic MTs at the leading
Leukocyte integrin LFA‐1 is aligned and oriented by actin flow during cell migration.
T.I. Moore1,2,3, P. Nordenfelt1,2,3,4, S. Mehta5, T. Lambert6, V. Swaminathan3,4,7, J.K. Mathew3,4,8, N.
Koga9,10, D. Baker9,10, T. Tani5, S. Mayor3,4,8, C.M. Waterman3,4,7, T.A. Springer1,2,3; 1Program in Cellular and
Molecular Medicine, Boston Childrens Hospital, Boston, MA, 2Biological Chemistry and Molecular
Pharmacology, Harvard Medical School, Boston, MA, 3Whitman Center, Marine Biological Laboratory,
Woods Hole, MA, 4Physiology Course, Marine Biological Laboratory, Woods Hole, MA, 5Eugene Bell
Center, Marine Biological Laboratory, Woods Hole, MA, 6Nikon Imaging Center, Harvard Medical School,
Boston, MA, 7Cell Biology and Physiology Center, National Heart, Lung, and Blood Institute, NIH,
Bethesda , MD, 8 TIFR, National Centre for Biological Sciences, Bangalore, India, 9Biochemistry,
University of Washington, Seattle, WA, 10Howard Hughes Medical Institute, Seattle, WA
Integrins are transmembrane, heterodimeric, surface receptors that mediate numerous cell‐cell and
cell–matrix interactions. Signaling bidirectionally, integrins bind ligand to their large ectodomains and
couple the actin cytoskeleton to their short β‐subunit cytoplasmic tails. In this work we sought to test
the traction force model of integrin activation, which proposes that force generated by the actin
cytoskeleton acts as an allosteric effector, stabilizing integrin conformational transition to a high affinity
state. During cell migration, actin typically undergoes “retrograde flow” and would be predicted to
cause integrins to align and adopt a specific orientation relative to actin flow direction. To test this
prediction engineered fluorescent protein integrin fusions and two independent polarization‐imaging
techniques were used. We show that the leukocyte integrin LFA‐1 (αL/β2, CD11a/CD18) becomes
aligned and oriented relative to the leading edge of migrating T‐cells. This orientation is aligned with the
direction of actin flow in the lamellipodia and its coupling to the cytoplasmic tail of β2. LFA‐1 orientation
is dependent on activation via inside‐out signaling, talin mediated coupling to actin, and integrins
binding their specific ligand, ICAM‐1, while being in the fully activated extended‐open conformation.
Engineering of the GFP‐integrin chimeras with increased linker length results in a significant decrease in
alignment of the fluorescence emission dipole of GFP and allow for validation of integrin orientation at
the leading edge of migrating cells from computer simulations using Rosetta. Together, these results
lend strong support for the cytoskeletal force model of integrin activation and place the molecular
understanding of integrin organization and function in the context of leukocyte migration.
LSP‐1 is a myosin‐IIA binding regulator of podosome dynamics and macrophage migration.
P. Cervero1, A. Bouiossou2, I. Maridonneau‐Parini2, S. Linder1; 1Institute for Medical Microbiology,
University Medical Center Eppendorf, Hamburg, Germany, 2CNRS UMR 5089, 2Institut de Pharmacologie
et de Biologie Structurale , Toulouse, France
Subcellular fine‐tuning of the actin cytoskeleton is a prerequisite for regulated cell migration and
adhesion structure turnover and function. We now identify LSP (lymphocyte‐specific protein)‐1 as a
critical regulator of the cortical actin cytoskeleton of primary macrophages and also as a novel
component of the podosome cap, a recently identified substructure on top of the F‐actin‐rich podosome
core. LSP‐1 is important for podosome dynamics and mechanosensing, as LSP‐1 overexpression leads to
increased oscillations and protrusive force of podosomes on pliant substrates, whereas siRNA‐based
depletion shows reverse effects. Moreover, LSP‐1 depleted macrophages show increased mobility and
turnover of podosome clusters, coupled with increased speed of cell migration. LSP‐1 function is
apparently based on its binding to myosin IIA, as shown by immunoprecipitation and proximity ligation
assays. Accordingly, LSP‐1 knockdown leads to strongly decreased myosin levels at podosomes.
Importantly, LSP‐1 localization to the detergent‐resistant actin cytoskeleton is regulated by its
phosphorylation on Ser252 downstream of p38 kinase and MK2, as shown by respective
phosphomutants. Collectively, we identify LSP‐1 as a novel regulator of podosome dynamics and
function, as well as of macrophage migration, and demonstrate that LSP‐1 activity is based on its binding
to myosin IIA and its phosphorylation on Ser252.
Microsymposium 2: Signaling in Health and Disease
Spatial control of Shoc2 scaffold ‐mediated ERK1/2 signaling requires remodeling activity of the
E. Galperin1, E. Jang1, H. JANG1, P. Shi1, G. Popa1, M. jeoung1; 1Mol and Cellular Biochemistry, University
Of kentucky, Lexington, KY
Enzyme‐binding scaffolds organizing the macro‐molecular signaling assemblies represent a considerable
portion of protein in cells. These scaffolding proteins guide the spatial organization of the signaling
enzymes and the flow of molecular information. However, mechanisms that control assembly and
dynamics within scaffolding complexes remain largely unknown. We unravel a novel, multi‐level
paradigm in which allosteric modifications alter the ability of the scaffold protein Shoc2 actively
accelerate transmission of ERK1/2 signals. Shoc2 is the scaffold protein that accelerates the ERK1/2
signaling pathway in response to growth factors. Initially identified in C. elegans as SUR‐8/SOC2, Shoc2 is
a critical positive regulator of the ERK1/2 signaling pathway that integrates the Ras and RAF‐1
components of the ERK1/2 pathway into a multi‐protein complex. Mutations in Shoc2 result in Noonan‐
like RASopathy, a developmental disorder with a wide spectrum of symptoms. The amplitude of the
ERK1/2 signals transduced through the complex is fine‐tuned by the HUWE1‐mediated ubiquitination of
Shoc2 and its signaling partner RAF‐1. However, the mechanistic basis of how ubiquitination of the
Shoc2 scaffold and RAF‐1 is controlled is unknown. We demonstrate that the (AAA+) ATPase PSMC5 is a
binding partner of Shoc2 that triggers translocation of Shoc2 to endosomes. PSMC5 then displaces the
E3‐ligase HUWE1 from the scaffolding complex to attenuate ubiquitination of Shoc2 and RAF‐1. We
show that Noonan‐like Rasopathy mutation that changes the subcellular distribution of Shoc2 lead to
alterations in Shoc2 ubiquitination due to the loss of accessibility to PSMC5. In summary, our results
demonstrate that PSMC5 is a novel critical player involved in regulating ERK1/2 signal transmission
through the remodeling of Shoc2 scaffold complex in a spatially‐defined manner. This study makes a
significant advance in our understanding of how the ERK1/2 signaling pathway is governed by the critical
scaffold Shoc2. Furthermore, this study describes a novel mechanism of how scaffolds can regulate
specificity and dynamics of cellular networks through remodeling mechanisms.
Activation of the proteinase‐activated receptor‐2‐β‐arrestin‐2 signaling axis by household
allergens in the lung.
M.C. Yee1, H.L. Nichols1, K. Pal1, D. Polley2, K.J. Lee1, M. Ming1, M.D. Seigler1, M.D. Hollenberg2, S.
Boitano3, K.A. DeFea1; 1•
Division of Biomedical Sciences, University of California, Riverside,
Department of Physiology Pharmacology, University of Calgary, Calgary,
Riverside, CA, 2•
Arizona Respiratory Center and Department of Physiology, University of Arizona,
Canada, 3•
Tucson, AZ
Exposure to household allergens such as the fungus Alternaria alternata is associated with increased
incidence of morbidity and risk of fatal asthma attacks for both children and elderly patients. Alternaria
allergens contain serine proteases capable of activating proteinase‐activated receptor‐2 (PAR2) and
promote inflammatory responses in the airway that are dependent upon protease activity. We have
previously shown that PAR2‐dependent airway inflammation is mediated through β‐arrestin‐2 signaling.
Thus, we hypothesized that proteases from Alternaria filtrates will promote activation of the β‐arrestin‐
2 (βarr2) dependent cellular signaling pathways that mediate airway inflammation in murine models of
Alternaria‐induced asthma. We show that both Alternaria allergenic filtrates and isolated proteases
promote recruitment of β‐arrestin‐2 to PAR2 and activation of PAR2‐βarr2 signaling in infiltrating
immune cells from Alternaria‐treated mice was reduced in βarr2‐/‐ compared to wild‐type mice. Using
histological and flow cytometric analyses, we demonstrate that Alternaria‐induced airway inflammation
requires both PAR2 and βarr2, as demonstrated by the reduced recruitment of leukocytes (particularly
eosinophils and CD4+T‐cells) into the lung, goblet cell hyperplasia, and thickening of the lung epithelium
in PAR2‐/‐ or βarr2‐/‐, compared to wild‐type, mice. These in vivo findings are consistent with previous
findings in the laboratory that PAR2 promotes leukocyte chemotaxis through a β‐arrestin‐dependent, G‐
protein‐independent mechanism involving scaffolding of proteins involved in actin assembly at the
leading edge of the cell. These are the first experiments demonstrating that the PAR2‐β‐arrestin‐
dependent signaling axis is important for asthma‐induced by an allergen associated with human asthma.
Thus, targeting this pathway via biased antagonism of PAR2 could be a new avenue of therapeutic
intervention for asthma.
Optogenetic spatial control of TrkA‐mediated pathways reveals a potential role for Raf/ERK
pathway in inducing polarity in PC‐12 cell differentiation model.
Q. Ong1, K. Zhang2, A. McGuire1, S. Guo1, F. Santoro1, C. Zeng1, A.Y. Sarro‐Schwartz1, R. Zhang1, B. Cui1;
Department of Chemistry, Stanford University, Stanford, CA, 2Department of Biochemistry, University of
Illinois at Urbana‐Champaign, Urbana, IL
Neurite outgrowth is an important process in the formation of neuronal networks. It is widely accepted
that exogenous stimuli such as nerve growth factor (NGF) exert their effects on neurite outgrowth via
Trk neurotrophin receptors. In the event of TrkA/NGF binding and activation, the downstream pathways
such as PI3K/Rac, Raf/MEK/ERK and PLC/PKC pathways are subsequently upregulated and these
contribute to neurite outgrowth in the cells. However, the contribution of each pathway to the exact
neurite morphology and the nature of these signaling pathways inducing polarity were not well
characterized. In our study, we utilized the light protein interaction pair, CRY2PHR and CIBN, to elicit
recruitment of key signaling proteins and partners and result in selected pathway activation. For
example, the recruitment of Raf to the plasma membrane had been demonstrated by our group to
activate the Raf/ERK pathway. [1] We characterized the growth morphology of neurites, where cells
with upregulated Raf/ERK and PI3K/Rac had lower (1.8) and higher (3.7) average number of primary
neurite per cell respectively compared to cells induced with NGF (2.9). We also plated PC12 cells
expressing the opto‐genes onto devices which exhibit global planarity on its growing surface but were
micropatterned to have well‐defined light‐blocking regions and transparent regions. Cells plated at the
edges of the light‐blocking regions were thus exposed to light partially and we examined the polarity of
neurite growth of these cells. Notably, while the PI3K/Rac pathway was known to induce polarity in
neurite growth and thus prefer neurite growth towards the light region in our experiments, we also
identified the Rac/ERK pathway to induce high selectivity towards the lit area as compared to the dark
area. Together, these results demonstrate the individual contributions of various pathways in
determining the neurite morphology and a potential role for Raf/ERK pathway to induce and sustain
polarity in neurite growth in PC12 cells.
Zhang K, Duan L, Ong Q , Lin Z, Verman PM, Sung K, Cui B (2014). “Light‐mediated kinetic control reveals
the temporal effect of the Raf/MEK/ERK pathway in PC12 cell neurite outgrowth”, PLoS ONE , 9(3),
Inhibition of one substrate phosphorylation of a protein kinase out of many substrates by a
selective peptide inhibitor of kinase‐substrate interaction.
N. Qvit1, D. Mochly Rosen1, M. Disatnik1; 1Chemical and Systems Biology, Stanford, Stanford, CA
Many kinases phosphorylate multiple substrates. To assess the role of an individual phosphorylation
event without using laborious mutagenesis studies, selective pharmacological tools that inhibit
phosphorylation of specific one substrate are needed. We focused on delta protein kinase C (delta‐PKC)
because delta‐PKC activation after heart attack leads to cardiac damage. We reasoned that since
pyruvate dehydrogenase kinase (PDK) phosphorylation by delta‐PKC leads to inhibition of ATP
production, it might be critical in the injurious effect delta‐PKC after heart attack. To determine if of the
many substrates of delta‐PKC, PDK is the substrate that mediates this injurious effect in the heart, we
designed a short protein‐protein interaction inhibitory peptide (PPIIP) to selectively inhibit delta‐PKC
phosphorylation of PDK. PDK/delta‐PKC interaction was inhibited by PPIIP in vitro and in vivo. PDK/delta‐
PKC PPIIP selectively inhibited phosphorylation of PDK without affecting phosphorylation of several
other delta‐PKC substrates. Treatment with PDK/delta‐PKC PPIIP inhibited myocardial infarction‐induced
PDK phosphorylation without inhibiting the phosphorylation of other delta‐PKC substrates. Further,
PDK/delta‐PKC PPIIP treatment led to a 50% reduction in infarct size, in release of cardiac enzyme and in
JNK phosphorylation, all markers of cardiac injury. PDK/delta‐PKC PPIIP is a selective inhibitor of PDK
phosphorylation by delta‐PKC. Of the several delta‐PKC‐mediated phosphorylation events following
heart attack, delta‐PKC‐induced PDK phosphorylation is sufficient to cause cardiac injury as a selective
inhibitor of this phosphorylation event alone was sufficient to provide great reduction in delta‐PKC‐
mediated cardiac injury after heart attack.
The Tetraspanin CD82 Regulates Hematopoietic Stem Cell Fitness.
C.A. Saito Reis1, K.D. Marjon1, K.L. Karlen1, R.J. Dodd1, C.M. Termini1, J.M. Gillette1; 1Pathology,
University of New Mexico Health Science Center, Albuquerque, NM
Hematopoietic stem and progenitor cells (HSPCs) are responsible for the continued production of new
blood and immune cells. Cell signaling and adhesion within the bone marrow microenvironment tightly
regulate the self‐renewal and differentiation of HSPCs. The tetraspanin CD82 is a membrane scaffold
protein shown to be highly expressed on HSPCs. Previous work in our lab showed that human CD34+
HSPCs in the G0 phase of the cell cycle displayed an enrichment of CD82 in polarized domains, essential
for HSPC homing and engraftment to the bone marrow. Moreover, when cells were pre‐treated with a
CD82 blocking antibody, homing to the bone marrow was impaired. Therefore, we hypothesize that
CD82 scaffold regulates HSPC signaling and adhesion within the bone marrow. To test this hypothesis,
we are utilizing a global CD82 knock out (CD82KO) mouse to evaluate changes in the HSPC population.
First, we assessed the HSPC populations within the bone marrow and found a significant decrease in the
number of long term HSCs (LT‐HSC) in the CD82KO mice. Next, we assessed the cycling status of the LT‐
HSC populations, since the decrease in LT‐HSCs could be due to an increase in cycling. Using a
combination of flow cytometry and BrdU assays, we identified a significant decrease in the number of
quiescent HSCs as well as an increase in the HSPC proliferation status of the CD82KO mice. We also
measured the transcript levels of cell cycle inhibitors p21 and p27 finding that both inhibitors were
significantly decreased in the CD82KO lineage negative population compared to WT. In addition, we
analyzed the fitness of the CD82KO HSPCs by assessing the engraftment potential of WT and KO HSPCs.
Our data indicate that there is no change in the HSPC engraftment potential following primary and
secondary engraftment. However, we detect a significant reduction of HSPC engraftment potential
following a quaternary engraftment. Furthermore, in a competitive repopulation assay, HSPCs from WT
mice significantly outcompete HSPCs from CD82KO mice within the bone marrow. These data suggest
that CD82 expression regulates stem cell fitness in a competitive environment and that CD82KO HSCs
have the potential to exhaust following multiple transplantations. Finally, competitive homing studies
suggests a significant decrease in the bone marrow homing potential of CD82KO cells when compared to
WT cells. Moreover, we detected an increase of CD82KO cells in the blood, suggesting a homing delay or
an inability for CD82KO cells to be maintained in the bone marrow. Together, these data demonstrate
that the CD82 scaffold plays a critical role regulating the self‐renewal and repopulation potential of LT‐
HSCs. Future studies will analyze how CD82 regulates specific signaling within the bone marrow niche.
Pathological lymphangiogenesis is regulated by galectin‐8‐dependent crosstalk among VEGF‐C,
podoplanin and integrin pathways.
W. Chen1, H. Leffler2, U.J. Nilsson3, L. Xia4, N. Panjwani1,5; 1Ophthalmology, Tufts University, Boston, MA,
Microbiology Immunology and Glycobiology, Lund University, Lund, Sweden, 3Center for Analysis and
Synthesis, Lund University, Lund, Sweden, 4Cardiovascular Biology Research Program, Oklahoma Medical
Research Foundation, Oklahoma City, OK, 5New England Eye Center, Boston, MA
Background: Lymphangiogenesis (LA) plays a vital role in diverse pathological conditions including
corneal graft rejection, dry eye and glaucoma. The goal of the current study was to characterize the role
galectin‐mediated carbohydrate recognition system in the modulation pathological LA.
Methods: Lymphatic endothelial cell (LEC) sprouting assays, corneal micropocket assays, gene
knockdown and antibody blocking assays, and galectin‐8 and podoplanin knockout mice were used to
assess the role and the mechanism of galectin‐8‐mediated LA.
Results: The study revealed that galectin‐8 is a potent lymphangiogenic factor. Galectin‐8 was markedly
upregulated in inflamed human and mouse corneas, and inhibitors of galectin‐8 reduced inflammatory
LA. In corneal micropocket assays and 3D sprouting assays, galectin‐8 promoted LA in a carbohydrate‐
dependent manner. Galectin‐8 was identified as a key mediator of integrin‐dependent crosstalk
between VEGF‐C and podoplanin lymphangiogenic pathways. Galectin‐8 inhibitors reduced VEGF‐C‐
induced LA. Conversely, exogenous galectin‐8 markedly enhanced VEGF‐C‐induced lymphangiogenesis in
a carbohydrate‐dependent manner. Knockdown of podoplanin reduced not only galectin‐8 but also
VEGF‐C‐mediated LEC sprouting. Also, in corneal micropocket assays, VEGF‐C‐induced LA was
significantly reduced in the galectin‐8‐/‐ and podoplanin‐/‐ mice; likewise, galectin‐8‐induced
lymphangiogenesis was reduced in podoplanin‐/‐ mice. Interestingly, knockdown of VEGFR‐3 did not
affect galectin‐8‐mediated LEC sprouting. Instead, inhibiting integrins α1β1 and α5β1 curtailed both
galectin‐8‐ and VEGF‐C‐mediated LEC sprouting. Additionally, podoplanin knockdown in LECs interfered
with integrin activation. Immunoprecipitation assays further confirmed galectin‐8‐dependent
interactions between podoplanin and integrins α5 and β1.
Conclusions: This study has uncovered a unique lymphangiogenic pathway in which galectin‐8‐mediated
interactions between PDPN and integrins α1β1/α5β1 play a key role.
Activation of HuR in a Gq‐p38 MAPK‐dependent manner promotes cardiac fibrosis and
pathological remodeling.
M. Tranter1, S.R. Anthony1, S. Slone1; 1Internal Medicine, Division of Cardiovascular Health and Disease,
University of Cincinnati, Cincinnati, OH
The initial development of cardiac hypertrophy is a compensatory response to increased wall stress and
allows the heart to maintain systolic function in the face of increased hemodynamic load. However, the
development of hypertrophy in response to pathological etiologies is also associated with increased risk
of arrhythmias and the progression to a decompensated pathological state followed by the
development of heart failure. Thus, increasing our understanding of the pathways that mediate cardiac
hypertrophy is critical to developing therapeutic strategies to treat heart failure, which remains a
leading cause of death in the United States. Activation of Gq proteins downstream of Gq‐coupled GPCRs
such as angiotensin receptors (AT1R) or α‐adrenergic receptors (α1‐AR) is a known mediator of
pathological hypertrophy, but the responsible signaling pathways downstream of Gq have yet to be fully
Our results demonstrate activation of the RNA binding protein HuR (Human antigen R) downstream of
Gq‐coupled AT1R and α1‐AR in cardiac myocytes. HuR binding regulates stability and expression of target
mRNA, and despite being highly expressed in the heart, nothing is currently known about the role of
HuR in pathological hypertrophy. To address the functional role of HuR in pathological hypertrophy, we
generated inducible cardiomyocyte‐specific HuR‐deficient (iCM‐HuR‐/‐) mice. Preliminary
characterization of these mice indicates that, while there is no overt functional phenotype at baseline,
deletion of HuR in the adult heart preserves cardiac function while decreasing pathological ventricular
remodeling following transverse aortic constriction (TAC), a model of hypertension‐induced pathological
cardiac hypertrophy. Furthermore, HuR activation in the hypertrophic heart is strongly co‐localized with
regions of fibrosis, and we show that the development of cardiac fibrosis following TAC is completely
inhibited in iCM‐HuR‐/‐ mice. In vitro experiments using primary neonatal rat ventricular cardiac
myocytes (NRVMs) demonstrate that HuR activation downstream of Gq‐coupled GPCRs is dependent on
p38 MAPK, but not the canonical Gq‐PLC or Gq‐PKC pathways. HuR knockdown in NRVMs also reduced
hypertrophy in response to the α1‐AR‐Gq agonist phenylephrine and inhibited pro‐fibrotic TGF‐β gene
Thus, in conclusion, our results show that HuR activation downstream of a Gq‐p38 MAPK signaling axis
promotes the initiation of cardiac fibrosis via myocyte expression of TGF‐β and subsequent pathological
left ventricular remodeling. These findings are significant as we have identified HuR as a novel Gq‐
dependent mediator of pathological hypertrophy and suggest a pro‐fibrotic role as a potential
mechanism for this effect.
Microsymposium 3: Membrane Dynamics and Visualization
S. Upadhyayula1, S. Arumugam2, R. Gaudin1, F. Aguet1, C. Wunder2, E. Betzig3, L. Johannes††2, T.
Kirchhausen1; 1Cell Biology, Harvard Medical School, Boston, MA, 2PSL Research University, Institut
Curie, Paris, France, 3Howard Hughes Medical Institute, Janelia Research Campus, Ashburn, VA
So far, attempts to follow in real time endocytic events in cultured cells have been limited to events
occurring on the ventral (attached) surface or relatively small regions of the dorsal (free) surface. The
advent of the powerful lattice light‐sheet microscope (LLSM) based on ultrathin light sheets for real time
3D imaging with single molecule sensitivity (Chen et al. , 2014) provided us with the unique opportunity
to simultaneously explore the dynamics of cargo uptake mediated by the clathrin‐dependent and
clathrin‐independent routes within the same cell. Here, we used LLSM for direct visualization and
molecular counting during endocytosis of fluorescently tagged transferrin (canonical cargo for the
clathrin endocytic route) and Shiga toxin (a cargo for the clathrin‐independent route) in genome‐edited
cells expressing fluorescently tagged chimeras of clathrin light‐chain A, the endocytic adaptor AP‐2, and
endophilin (constituent of the clathrin‐dependent and independent routes). As expected, we found that
most clathrin‐coated vesicles captured transferrin regardless of the cell surface from which they formed.
Shiga toxin was also captured by a very large number of clathrin‐coated structures. Formation of
endophilin‐positive carriers associated with the clathrin‐independent route was prominent at the
periphery of the attached surfaces of the cells, and as previously described, they also captured Shiga
toxin while sorting of transferrin to these carriers was minimal. The current goal is to use the LLSM data
calibrated for single‐molecule sensitivity to determine the global capture dynamics of both cargoes and
to correlate the 2D clustering potential of Shiga toxin to the specificity of its uptake.
Chen, B.‐C. et al. (2014). Lattice light‐sheet microscopy: imaging molecules to embryos at high
spatiotemporal resolution. Science 346 , 1257998.
Rapid exocytosis of an endolysosome‐derived membrane domain forms a polarized invasive
protrusion that clears basement membrane during cell invasion.
K.M. Naegeli1, Q. Chi2, D.R. Sherwood2; 1Department of Pharmacology and Cancer Biology, Duke
University, Durahm, NC, 2Department of Biology, Duke University, Durham, NC
Cell invasion through basement membrane barriers is required for many developmental, physiological,
and pathogenic processes including neural crest cell migration, leukocyte trafficking, and cancer
metastasis. How invasive cells regulate their plasma membrane during invasion, though, is not well
characterized, particularly in vivo. During the larval development of C. elegans, a specialized gonadal
cell, the anchor cell, invades through two juxtaposed basement membranes to initiate direct uterine‐
vulval contact. After an invadopodium breaches the underlying basement membrane, netrin receptor
UNC‐40/DCC traffics to the breach and promotes formation of a large protrusion through this breach
towards the underlying vulval precursor cells. Formation of this protrusion displaces basement
membrane, thereby opening a hole to allow for contact between the two tissues. We have found that
generation of the invasive protrusion requires a rapid 20% increase in cell volume and 12% expansion of
cell surface area over the course of thirty minutes. Animals lacking the netrin receptor UNC‐40/DCC fail
to generate an invasive protrusion, fail to efficiently clear underlying basement membrane, and fail to
exhibit the same marked increase in cell volume and surface area. Animals lacking UNC‐40/DCC
effectors, including MADD‐2/TRIM‐9, also fail to efficiently clear basement membrane under the anchor
cell. This indicates that UNC‐40/DCC and its effectors coordinate membrane addition at the anchor cell’s
invasive membrane to clear the underlying basement membrane. Intriguingly, LMP‐1::GFP‐labeled
endolysosomes localize to the site of basement membrane breach and are heavily polarized to the
invasive protrusion during its extension, suggesting that endolysosomes may be the source of
membrane for the invasive protrusion. Perturbation of the endolysosome through mutation of
lysosomal membrane protein cup‐5 yields animals with anchor cells that also fail to increase in size,
indicating that integrity of the endolysosomal membrane is necessary to facilitate protrusion formation
and cell expansion during invasion. Targeted photobleaching of the anchor cell’s invasive protrusion
leads to significant loss of fluorophore signal in the invasive protrusion that is not mirrored in the apical
domain of the cell, thus indicating a barrier to free diffusion of cell membrane between the protrusion
and body of the cell. Together, these studies identify a new subcellular structure, the invasive
protrusion, a specialized membrane domain rapidly formed from exocytosis of the endolysosome that
clears basement membrane barriers during cell invasion.
Measuring GLUT4 vesicle exocytosis using intracellular intravital microscopy.
A.J. Kee1, A. Masedunskas1, C.A. Lucas1, W. Han2, P.W. Gunning1, E.C. Hardeman1; 1School of Medical
Sciences, University of New South Wales Australia, Sydney, Australia, 2Singapore Bioimaging
Consortium, Agency for Science, Technology and Research (A*STAR), Singapore, Singapore
Defects in insulin‐stimulated glucose uptake are central events in the development of Type II diabetes
and metabolic syndrome. Insulin‐stimulated glucose uptake in muscle and adipose tissue requires the
movement of the glucose transporter GLUT4 from intracellular storage vesicles (GSVs) to the cell
membrane via an exocytic pathway. Our understanding of GLUT4 trafficking comes almost exclusively
from studies using cultured adipocytes and myoblasts that are morphologically quite distinct from
adipocytes and muscle fibres in adult tissue. They do not have complex 3‐dimensional tissue
architecture or the cellular and systemic interactions of tissues in the living organism. Therefore, we
have developed an intracellular intravital microscopy approach to measure GLUT4 trafficking in vivo in
skeletal muscle. One of the limitations of previous efforts to study GSV trafficking in vivo was that
individual GSV fusion events could not be unambiguously detected due to the absence of a clear
fluorescent signature. We have overcome this limitation by using a dual colour GLUT4 probe
(rGLUTpHluor) comprised of pHluorin inserted into the first exofacial loop of GLUT4 and tdTomato at the
C‐terminus. This allows the simultaneous tracking of GLUT4 fusion events (visualized as a rapid increase
in GFP/Tomato signal ratio upon exposure to the neutral pH of the extracellular environment) and
GLUT4 vesicle movement. We electroporated mouse skeletal muscle in vivo with rGLUTpHluor and
detected for the first time GLUT4 fusion events that increased at cell membranes after insulin
stimulation. GLUT4 fusion events were observed both at the T‐tubules and at sarcolemma indicating
that the T‐tubule membrane is a significant site of insulin‐stimulated GLUT4 fusion and glucose uptake in
skeletal muscle. This represents the first demonstration of sub‐diffraction‐sized vesicle fusion events in
High‐Resolution Imaging of Living Cells by Atomic Force Microscopy.
A.L. Slade1, I. Medalsy1, S. Hu1, J.E. Shaw1, H. Schillers2; 1Bruker Nano Surfaces, Santa Barbara, CA,
Institute of Physiology II, University of Muenster, Muenster, Germany
Atomic force microscopy (AFM) provides the unique capability to obtain three‐dimensional images of
the morphology of individual living cells under physiological conditions without the need for labeling or
staining. However, due to their flexible and dynamic nature, visualization of cellular structures by AFM
with nanoscale resolution has remained challenging. Microvilli are common structures found on
epithelial cells that increase the apical surface thus enhancing the transmembrane transport capacity
and also serve as mechanosensors. Changes in both the density and structure of microvilli are associated
with various disease states. We have used a newly developed AFM probe, having a unique 17 micron tall
tip and 65nm fixed end radius, together with the low piconewton imaging forces enabled by PeakForce
Tapping mode to resolve individual microvilli structures on the surface of living cells for the first time by
AFM. Our studies also revealed a direct relationship of the observed structure of the microvilli with the
interaction force of the AFM probe. With regards to studying the dynamic behavior of cell structures,
traditional AFM imaging has been restricted in this area of research due to the typically longer
acquisition times required to obtain a single AFM image. With recent advances in high‐speed AFM
imaging, where images are now obtained in a matter of seconds, we have successfully applied AFM
imaging to investigate cell dynamics. Protrusion formation is one of the essential first steps in the
process of cell migration. High‐resolution imaging of this step has often been limited using typical optical
microscopy techniques. The unique combination of high‐resolution and high‐speed AFM imaging has
now allowed us to directly observe the formation and advancement of individual lamellipodia and
fillipodia at the leading edge of living stem cells during migration. In other HS‐AFM studies we were able
to obtain high‐resolution time sequence images of the native structure of a bacterial outer membrane,
obtained directly on the surface of living Escherichia coli cells. The increased time resolution of HS‐AFM
allowed us to observe changes in the nanoscale structure of the outer membrane in direct response to
an antimicrobial peptide (AmP), at timescales relevant to the mechanism of AmP‐induced cell death. The
results of these studies have provided the first opportunity to resolve the dynamics of AmP activity in a
native cell membrane environment in real‐time and with nanoscale resolution.
ω‐3 polyunsaturated fatty acids direct differentiation of the membrane phenotype in
mesenchymal stem cells to potentiate osteogenesis.
K.R. Levental1, M.A. Surma2, J.H. Lorent1, A. Skinkle3, C. Klose2, I. Levental1; 1Integrative Biology and
Pharmacology, The University of Texas Health Science Center at Houston, Houston, TX, 2Lipotype,
Dresden, Germany, 3Rice University, Houston, TX
Membranes are involved in nearly aspect of cellular physiology. The plasma membrane is the interface
between a cell and its environment and is therefore responsible for a myriad of parallel processing tasks
that must be tightly regulated to avoid aberrant signaling. To achieve this functional complexity,
mammalian cells produce thousands of lipid species that are actively turned over and trafficked to
produce spatial and temporal gradients between cellular compartments. In addition to the plethora of
regulatory roles performed by individual lipid molecules, membrane physiology is strictly dependent on
the biophysical phenotypes – including membrane fluidity, rigidity, lipid packing, and lateral organization
– arising from the collective behaviors of lipids. Dietary perturbations of lipid profiles can have severe
deleterious consequences (e.g. hypercholesterolemia) or be broadly therapeutic, as with ω‐3
polyunsaturated fatty acids (PUFAs). The molecular mechanisms underlying these effects are not clear;
however, an intriguing possibility is that dietary fats integrate into membrane lipids, change membrane
properties, and thereby affect cell signal transduction. Up to now, neither the lipid repertoires of
different cellular membranes nor their modulation by dietary fats have been widely explored. Here, we
show extensive, cell‐autonomous remodeling of both cellular and plasma membrane (PM) lipidomes
during human mesenchymal stem cell (MSC) differentiation into adipocytes and osteoblasts. Lipidomic
differentiation results in unique, cell‐specific membrane phenotypes, with osteoblasts containing longer
and more polyunsaturated lipids which result in more ordered membranes and more stable raft
microdomains. The unique features of osteoblast membranes enabled rational remodeling of
membrane phenotypes to direct lineage specification in MSCs. Supplementation with a lipid component
characteristic to osteoblasts (the ω‐3 PUFA docasohexaenoic acid, DHA) induces broad lipidomic
remodeling in MSCs, producing an osteoblastic membrane phenotype. This membrane remodeling
alters membrane signaling through Akt to augment RUNX2 expression and ultimately potentiate
osteogenic differentiation of MSCs. These results demonstrate that membrane phenotypes are central
drivers of cell function and are susceptible to remodeling by exogenous fatty acids, suggesting a novel
mechanism by which dietary fats ultimately affect cellular physiology.
Nanoscale spatiotemporal organization of Fas receptor (CD95) during early stages of signaling
revealed by quantitative superresolution microscopy.
P. Sengupta1, A. Cruz2, R. Siegel2, J. Lippincott‐Schwartz1; 1CBMP, NICHD, Bethesda, MD, 2Autoimmunity
branch, NIAMS, Bethesda, MD
The Fas receptor (FasR) mediates programmed cell death via its cytoplasmic death‐domain (DD) and
plays a critical role in physiological processes like tumor surveillance and immune tolerance. The initial
model postulating the initiation of signaling by trimeric Fas ligand mediated oligomerization of
monomeric FasR has been challenged by recent biochemical and in vitro structural studies, which
suggest that the FasR can exist as preassembled oligomers even before ligand binding. Additionally, a
membrane proximal palmitoyl‐anchored cysteine residue was shown to be critical for Fas‐signaling,
leading to the hypothesis that the unstimulated FasR organization is governed by clustering of FasR into
specialized lipid domains mediated by saturated palmitoyl lipid‐anchor. However, since the receptor
spatial organization and ligand‐induced remodeling happens at spatial scales inaccessible by diffraction‐
limited microscopy, the actual oligomerization state of FasR in living cells and mechanism of transition to
signaling state has remained unclear. We have used single‐molecule based superresolution‐microscopy
to characterize the oligomerization state of FasR and its remodeling following binding of ligand,
uncovering novel conceptual insights about the mechanism of early steps of FasR‐mediated signaling.
Quantitative cluster analysis of superresolution‐data reveal that unstimulated FasR is organized into
clusters of three molecules. This pre‐ligand organization is independent of palmitoyl‐anchors but is
disrupted by removal of the exoplasmic N‐terminal domain of FasR. Following ligand binding, the FasR
undergoes reorganization into clusters of tens of molecules and the palmitoyl‐anchor is critical for this
ligand‐triggered remodeling. Thus, in contrast to the prevailing notion, we find that the palmitoyl
anchors do not mediate spatial configuration of unstimulated receptors, but participate in the transition
of the FasR to the signaling state. Palmitoyl‐anchors might enable the ligand‐bound receptors to access
specific membrane environment where their interaction with other molecules facilitates the formation
of signal‐competent receptor platforms. Additionally, experiments with mutant FasR lacking the DD
showed that the DD is also crucial for the transition to larger receptor clusters, indicating the role of
cytoskeleton and downstream signaling molecules in stabilization of the signaling clusters. Thus, using
quantitative superresolution microscopy, we have unraveled details of FasR nanoscale organization and
previously unappreciated role of the various structural features of FasR in mediating the early steps of
signaling. We anticipate that this approach will be useful for unraveling the mechanistic underpinnings
of other receptor mediated signaling pathways.
A new ER structure revealed by live custom STED microscopy.
L.K. Schroeder1, M. Deline2, J. Bewersdorf1, S. Bahmanyar2; 1Dept. of Cell Biology, Yale University, New
Haven, CT, 2MCDB, Yale University, New Haven, CT
The ER is a highly dynamic network of membrane sheets and tubules. ER tubules grow and retract by
interacting with the microtubule cytoskeleton and ER tubules and sheets undergo interconversions. The
interconversion of tubules and sheets is also regulated by microtubules. Depolymerization of
microtubules by nocodazole treatment transforms the ER network into a uniform sheet‐like
morphology. During cell division, a similar ER structural change occurs ‐ microtubules are reorganized to
form the mitotic spindle and the ER transforms from a mixed network of sheets and tubules to a
uniform membrane morphology. While mitotic‐ER morphology is important for mitotic progression in
many systems, ER‐membrane structure in intact mitotic cells has remained elusive because of difficulties
in imaging rounded mitotic cells with high optical resolution. Here, we visualized ER membrane
structure live using a custom built Stimulated Emission Depletion (STED) microscope in cells treated with
nocodazole to mimic mitotic microtubule reorganization. We found that regions that appear as smooth
ER sheets by conventional confocal microscopy are instead lace‐like in structure when imaged using live
STED microscopy. This membrane structure, which we call “ER lace,” is comprised of flat membrane
regions with holes of approximately 100‐200 nm in diameter spaced evenly apart. ER‐lace is not an
outcome of nocodazole treatment, because we also observed peripheral ER‐lace like structures in
untreated cells. To track events that lead to the formation of ER‐lace, we imaged ER dynamics live using
confocal microscopy following microtubule depolymerization. Convergence of several ER rings or
cinching of ER nodes preceded sheet formation illustrating unique ER dynamics that might contribute to
formation of ER‐lace. We are currently customizing our approach to determine whether mitotic‐ER
morphology is composed of ER‐lace as would be predicted from our results. Taken together, the
discovery of a new ER structure using cutting edge imaging technology leads to previously unexplored
questions on how the ER is shaped.
Microsymposium 4: Cell Division and Cytokinesis
Symmetry and scale orient Min protein patterns in shaped bacterial sculptures.
F. Wu1, C. Dekker1, J.E. Keymer1; 1Bionanoscience, Delft University of Technology, Delft, Netherlands
The boundary of a cell defines the shape and scale of its subcellular organization. However, it has been
difficult to systematically and quantitatively probe the effect of cell boundary on the intracellular
networks and macromolecular machineries in bacteria. Here, we show that, nanofabricated structures
can be used to ‘sculpt’ living E. coli cells into defined shapes, e.g. squares and rectangles, with sizes
ranging from 2x1x1 to 11x6x1 μm3. We use these bacteria to study the spatial adaptation of Min
proteins, which oscillate pole‐to‐pole in rod‐shape cells to assist division machineries to position at cell
middle. In a broad geometric parameter regime, we found that they exhibit versatile oscillation
patterns, sustaining rotational, longitudinal, diagonal, stripe, and even transversal modes. These
patterns are found to directly capture the symmetry and scale of the cell boundary, and the Min
concentration gradients scale in adaptation to the cell size within a characteristic length range of 3–6
μm. Numerical simulations reveal that local microscopic Turing kinetics of Min proteins can yield global
symmetry selection, gradient scaling, and an adaptive range, when and only when facilitated by the
three‐dimensional confinement of cell boundary. Our data thus show that spatial boundaries can
facilitate simple molecular interactions to result in far more versatile functions than previously
understood. Furthermore, we demonstrate how the ‘cell‐sculpting’ approach can be extended to
understand pattern multistability phenomena in the context of cell growth and perturbations, and to
uncover the role of cell boundary in facilitating chromosome compaction.
References: 1. Wu F et al, Symmetry and scale orient Min protein patterns in shaped bacterial
sculptures, Nature Nanotechnology 2015 (DOI: 10.1038/nnano.2015.126) 2. Wu F et al, Multicolor
imaging of bacterial nucleoid and division proteins with blue, orange and near infrared fluorescent
proteins, Frontiers in Microbiology 6 (607) 2015 3. Mannik J et al, Robustness and accuracy of cell
division in Escherichia coli in diverse cell shapes. PNAS 109 (18), 2012 4. Wu F & Dekker C,
Nano/microfabricated structures for bacteria: from techniques to biology. Chemical Society Reviews
(under review 2015)
Cdc42EP1 is a novel regulator of Septin organization during cytokinesis.
A.L. Wilson1, S.J. Terry1, U.S. Eggert1; 1Randall Division of Cell and Molecular Biophysics, Kings College
London, London, United Kingdom
Cytokinesis involves the spatially and temporally co‐ordinated actions of cell cycle regulation,
cytoskeleton rearrangements and membrane reorganization to ensure each of the two daughter cells
created has a complete set of chromosomes and organelles. During a functional RNAi screen to identify
novel proteins involved in cytokinesis, several surprising hits were discovered; these included Cdc42EP1.
Cdc42EP1 is a member of a family of Cdc42 effector proteins, whose activity is believed to be negatively
regulated by the small GTPase Cdc42.
Depletion of Cdc42EP1 leads to a late stage cytokinesis failure, with cells able to pass unaffected
through cytokinesis until the terminal abscission stage, where a relapse of the intercellular bridge is
observed. This phenotype can be rescued by expressing an RNAi resistant epitope tagged version of
Cdc42EP1. Immunofluorescence analysis of epitope tagged versions of Cdc42EP1 revealed changes in its
localization throughout the cell cycle, and an enrichment of the protein at the ingressing furrow in
dividing cells. Immunoprecipitation of Cdc42EP1 revealed an interaction with Septin 9, but not Septins 2,
6 or 7. In the absence of Cdc42EP1, aberrant Septin 9 localization is observed during cytokinesis,
suggesting Cdc42EP1 is essential for maintaining the correct spatial and temporal localization of Septin 9
throughout the cell cycle. Mutation of a predicted Septin binding domain in Cdc42EP1 abolishes the
interaction with Septin 9, and overexpression of this mutant version is unable to rescue the cytokinesis
failure in the absence of endogenous protein. These results indicate the interaction between Cdc42EP1
and Septin 9 is essential for the function of Cdc42EP1 in cytokinesis.
This work has identified Cdc42EP1 as a regulator of Septin localization during cell division and thereby
provides insights in to the signaling mechanisms that control the organization of individual septins
during cytokinesis. Furthermore, this work elucidates a previously unidentified role for the Cdc42EP
family in cytokinesis.
Understanding cellular variation in the molecular regulation of cytokinesis.
T. Davies1, N. Romano Spica1, Y. Zhuravlev1, M. Shirasu‐Hiza2, J. Dumont3, J.C. Canman1; 1Department of
Pathology and Cell Biology, Columbia University, New York, NY, 2Department of Genetics and
Development, Columbia University, New York, NY, 3CNRS, Institut Jacques Monod, Paris, France
Accurate cytokinesis, the physical division of one cell into two, is essential for the development of all
multicellular organisms. Cytokinesis is driven by constriction of an actomyosin contractile ring that is
positioned and controlled by the mitotic spindle. Conventionally, contractile ring assembly and
constriction is thought to be controlled by the same molecular mechanisms in all dividing animal cells.
However, human genetics have revealed that some cell types are exquisitely sensitive to mutational
disruption of cytokinesis genes, while other cell types are naturally programed to fail in cytokinesis (e.g.
cells in the blood, heart). Thus to fully understand the molecular regulation of cytokinesis, we need to
understand how cytokinesis is differentially regulated in different cell types. To address this question,
we used a collection of fast‐acting, tunable, temperature sensitive (ts) cytokinesis‐defective mutants
from the worm C. elegans to study differences in the regulation of cytokinesis in each cell within the
dividing 4‐to‐8 cell embryo. At this early stage of development, each cell already has a distinct and
reproducible cell fate that is controlled by well understood, conserved, cell fate‐signaling molecules (e.g.
Notch, Wnt). We have probed the differential requirement for six ts proteins in each individual cell of
the dividing 4‐to‐8 cell embryo in two ways: First, we determined the level of protein function required
in each cell by systematically monitoring the success or failure of cytokinesis across a thermal range.
Second, we determined the functional temporal window of activity required during cytokinesis by
shifting (and maintaining) individual embryos to restrictive temperatures at distinct times throughout
the process with high temporal resolution (~60 sec). Using both methods, we found repeatable, cell type
specific differences in the functional and temporal requirements for individual proteins. For example,
we found specific cells have a lower functional requirement for diaphanous forminCYK‐1‐mediated actin
nucleation than other cells within the same embryo. Further, we found that the functional temporal
requirement during cytokinesis for forminCYK‐1 and other cytokinesis proteins reproducibly varies
between individual cell types. Together, our results show that cytokinesis is not regulated via the same
molecular mechanisms in all cells, even in genetically identical sister cells within the same embryo. We
are currently investigating how cell fate signaling and other pathways unique to a multicellular organism
(e.g. cell contacts, adhesions, and polarity) affect the molecular requirements for cytokinesis.
Force‐dependent inhibition of formin Cdc12 by myosin Myo2 during in vitro reconstituted
cytokinesis Search, Capture and Pull.
D. Zimmermann1, G.M. Hocky2, L.W. Pollard3, M.J. Lord3, D.R. Kovar1; 1Molecular Genetics and Cell
Biology, The University of Chicago, Chicago, IL, 2Chemistry, The University of Chicago, Chicago, IL,
Molecular Physiology and Biophysics, The University of Vermont, Burlington, VT
Cytokinesis, the process by which a dividing cell is physically separated into two daughter cells, presents
a hallmark of all eukaryotic cells. From fungi to animal cells, class II myosins along with specific sets of
actin‐binding proteins (ABPs) assemble a tension‐generating contractile ring comprised of F‐actin. Basic
mechanistic insights about cytokinesis have been deduced from studies on the model organism fission
yeast Schizosaccharomyces pombe. In dividing S. pombe cells the contractile ring assembles around the
cell equator from ~65 cytokinetic nodes, constituting ensembles of defined sets of ABPs along with the
class II myosin Myo2. Due to a joint‐action of these ABPs and Myo2, nodes condense in an actin‐
dependent manner into a compact ring. According to the Search‐Capture‐Pull‐Release model (Vavylonis
et al., Science 2008), an actin filament that is nucleated from one node by the formin Cdc12 randomly
'searches' the cortex and is ultimately 'captured' by the essential Myo2 myosin from a neighboring node.
Subsequently, Myo2 motor 'pulling' on the captured actin filament brings the nodes closer together
before that interaction is 'released'. Here, by using biomimetics along with multi‐color TIRF microscopy,
we reconstituted Search‐Capture‐Pull between beads containing Cdc12 and Myo2 in vitro. With this
system at hand, we are able to dissect the underlying molecular principles and specific roles for each
ABP involved during contractile ring assembly. We discovered a new phenomenon, where Myo2‐
mediated pulling by one ‘node’ on Cdc12‐elongated filaments from a neighboring ‘node’ results in the
immediate arrest of formin‐mediated elongation, suggesting a novel mechanism of formin regulation
that may be crucial for the proper formation of the cytokinetic ring as well as other actin‐dependent
Protein phosphatase 1 regulates ZYG‐1 levels to limit centriole duplication.
J. Iyer1, N. Peel2, A. Naik2, A.A. Hyman3, K.F. O'Connell1; 1NIDDK, National Institutes of Health, Bethesda,
MD, 2Department of Biology, The College of New Jersey, Ewing, NJ, 3Molecular Biology and Genetics,
Max Planck Institute of Molecular Biology and Genetics, Dresden, Germany
Centrioles are cylindrical microtubule‐based structures that are required for the formation of cilia,
flagella and the mitotic spindle. Centrioles are duplicated only once during each cell cycle and this
involves the assembly of a single new (daughter) centriole next to each existing (mother) centriole.
Dysregulation of this process yields an abnormal centriole number and may lead to spindle
abnormalities, aneuploidy, cell signaling defects, and disease. The nematode C. elegans has proven
particularly useful in unraveling the molecular details of centriole duplication. So far six conserved core
factors have been identified in the worm including the master regulator ZYG‐1, a member of the Plk4
family of kinases. Several studies have demonstrated that controlling ZYG‐1/Plk4 levels is critical for
ensuring a precise doubling of centrioles but a thorough understanding of the mechanisms involved is
lacking. We have utilized a genetic approach to identify several novel regulators of centriole
duplication. Specifically, we find that partial depletion of protein phosphatase 1 β (PP1β) or either of its
regulators I‐2 or SDS‐22 can rescue the centriole duplication defect of a zyg‐1 mutant. Furthermore,
strong inhibition of PP1β activity results in embryonic lethality and the simultaneous formation of
multiple daughter centrioles around a single mother. Significantly, we find PP1β functions post‐
translationally by down regulating the level of ZYG‐1. Thus, we have identified a new pathway that
limits expression of this key regulator. To further define this pathway we have conducted a series of
pull‐down experiments and have identified an RNA‐binding protein named CSTL‐2 that interacts with
PP1β, I‐2, SDS‐22, and ZYG‐1. Interestingly, depletion of CSTL‐2 also rescues the centriole duplication
defect of a zyg‐1 mutant. Our data are consistent with a model in which PP1β down regulates ZYG‐1
protein levels via CSTL‐2. In summary, we have identified a novel pathway that plays a critical role in
ensuring that the correct number of centrioles are formed each cell cycle.
A Regulatory Switch Alters Chromosome Motions at the Metaphase to Anaphase Transition.
K. Su1, Z. Barry2, N. Schweizer3, A.J. Pereira3, H.J. Maiato3,4, M. Bathe2, I.M. Cheeseman1,5; 1Whitehead
Institute, Cambridge, MA, 2Department of Biological Engineering, Massachusetts Institute of Technology,
Cambridge, MA, 3Institute for Molecular and Cell Biology, University of Porto, Porto, Portugal,
Department of Experimental Biology, Faculda de de Medicina, University of Porto, Porto, Portugal,
Department of Biology, Massachusetts Institute of Technology, Cambridge, MA
During mitosis in vertebrate cells, replicated sister chromatids are first aligned at the metaphase plate
and subsequently segregated to opposing spindle poles at anaphase. To define the parameters of
anaphase chromosome movement with high temporal resolution in human cells, we developed a live
cell imaging‐based assay, which revealed multiple discrete phases of anaphase chromosome motions.
Although prior work has largely assumed that the primary change at the metaphase to anaphase
transition is the removal of the physical connections between sister chromatids by the cleavage of
cohesin molecules, we observed an initial delay in poleward movement at anaphase onset and
discontinuous poleward movements that are inconsistent with a simple physical transition. Our analysis
demonstrates that these behaviors are not due to persistent physical connections between sister
chromatids, but instead reflect changes in the cyclin‐dependent kinase (CDK)‐based regulatory
environment at metaphase vs. anaphase. Perturbing dephosphorylation during anaphase dramatically
increases the oscillatory behavior of separated anaphase chromatids in a manner that is at least partially
dependent on chromokinesin‐based polar ejection forces. Together, our work suggests that a switch in
cellular phosphorylation status alters chromosome movement to drive poleward chromosome
segregation during anaphase.
Microsymposium 5: Mechanics in Cellular Maintenance and Disease
Directed transport of a kinesin/cargo pair along newly polymerized microtubules into dendritic
spines undergoing synaptic plasticity.
D.P. McVicker1, A.M. Awe1, K. Richters1, D.A. Cowdrey1, X. Hu1, E.R. Chapman1, E.W. Dent1;
Neuroscience, University of Wisconsin‐Madison, Madison, WI
Many forms of synaptic plasticity involve changes in the structure and composition of dendritic spines.
Here, we uncover a novel mechanism by which spines are altered by synaptic plasticity which involves
the trafficking of specific material into spines. Three prevailing synaptic trafficking models include: (1)
diffusion of cytoplasmic material into the spine from the dendrite shaft, (2) fusion of vesicles containing
transmembrane proteins in the dendritic shaft followed by diffusion in the plane of the plasma
membrane into spines, or (3) a kinesin/microtubule (MT) to myosin/actin motor hand‐off mechanism at
the spine neck, followed by myosin‐based transport into the spine. Using live cell total internal reflection
fluorescent (TIRF) microscopy, we present evidence for targeting specific cargo into dendritic spines
undergoing remodeling via transport along MTs that directly polymerize into spines from the dendrite
shaft. We demonstrate that during transient MT invasions of spines, the kinesin KIF1A, and an
associated cargo, synaptotagmin‐IV (syt‐IV), translocate in unison into spines along MTs. Additionally,
we show that this “direct‐deposit” model is activity‐dependent, specific for a particular kinesin/cargo
pair and is followed by exocytosis of the syt‐IV‐containing vesicles in the spine head. Thus, we define a
new mechanism for targeting dendritic cargo directly into spines during synaptic plasticity.
The role of N‐cadherin signaling on endothelial barrier integrity.
K.J. Kruse1, F. Huang1, Y. Sun1, S.M. Vogel1, Y.A. Komarova1, A.B. Malik1; 1Pharmacology, University of
Illinois at Chicago, Chicago, IL
Vascular Endothelial (VE)‐and Neuronal (N)‐cadherins expressed in endothelial cells (EC) exert their
respective biological function via assembly of homophilic adhesions. The role of VE‐cadherin adhesion in
regulating endothelial barrier function has been well established, however the role of N‐cadherin in ECs
remains mainly unknown. VE‐cadherin is recruited to sites of cell‐cell adhesion between ECs, whereas N‐
cadherin forms homophilic adhesion between ECs and mural cells (pericytes, smooth muscle cells) at the
abluminal surface of the endothelium. Using an EC specific, inducible knockout mouse model developed
in our lab, we have shown that deletion of Cdh2 gene (N‐cadherin) results in increased permeability of
pulmonary endothelial microvessels to fluid and solutes, suggesting that N‐cadherin adhesion restricts
permeability in the endothelial barrier. To study N‐cadherin adhesion‐mediated signaling in endothelial
monolayers, we have utilized surface chemistry techniques to mimic N‐cadherin homophilic interaction
in vitro . Covalent attachment of N‐cadherin extracellular domain to an Ni‐NTA surface in an oriented
manner induced formation of N‐cadherin adhesions proximal to tricellular junctions and promoted
recruitment of VE‐cadherin to adherens junctions (AJs). Depletion of N‐cadherin in ECs or denaturation
of N‐cadherin extracellular domain abolished this effect, suggesting that N‐cadherin “outside‐in”
signaling promotes assembly of VE‐cadherin adhesion. Using nanoscale biomimetic platforms combined
with mass spectroscopy, we isolated and analyzed N‐cadherin complexes. We found that a dual Rac1
and RhoA activator, TRIO, is recruited to N‐cadherin adhesion complexes to activate both RhoA and
Rac1 signaling at the sites of AJs. This was consistent with increased acto‐myosin tension across VE‐
cadherin analyzed with FRET‐based biosensor as well as formation of lamellipodia protrusions above AJs.
Furthermore, inhibition of ROCK signaling or depletion of Neuronal navigator 1 (Nav1) abolished
recruitment of VE‐cadherin to AJs specifically in cells plated into N‐cadherin but not gelatin‐coated
surfaces. These data cumulatively demonstrate that N‐cadherin promotes assembly of VE‐cadherin
adhesion by providing spatial control of RhoA and Rac1 activities.
Live multiplexed imaging of stem cell mechanotransduction and mechanoadaptation.
I. Jalilian1, R. Oldfield1, P.W. Gunning2, M.L. Knothe Tate1; 1Graduate School of Biomedical Engineering,
UNSW, Sydney, Australia, 2School of Medical Sciences, UNSW, Sydney, Australia
The mechanical milieux of stem cells modulate their lineage commitment. Given the role of cytoskeleton
in sensing the physical forces and in mechanotransduction, it is crucial to understand the exact effect of
mechanical stress on cytoskeletal behaviour in stem cells and subsequent gene regulation, culminating
in differentiation. During the process of lineage commitment, live imaging methods allow for precise
measurement of changes in volume and shape as well as spatiotemporal tracking of cytoskeletal
dynamics in stem cells. However, until recently, it has not been possible to measure these processes in
concert. Here we developed a novel method for live cell multiplexed fluorescent imaging for the
spatiotemporal changes of the cell nucleus, membrane and cytoskeleton concomitant to measurement
of the subcellular strains under controlled mechanical stress.
To image the cell nucleus, membrane and cytoskeleton at the same time and to measure their
spatiotemporal distribution, these cellular compartments were stained simultaneously with specific
fluorescent dyes during exposure to controlled mechanical stress and/or strains in embryonic murine
mesenchymal stem cells (C3H1T1/2). The nucleus and membrane of the cells were stained using
Hoechst and wheat germ agglutinin, respectively. In addition, to track cell membrane displacement for
strain mapping using digital image correlation, live cells were tagged at time zero with Concavalin A
conjugated microbeads, which bind noncovalently to the glycoproteins of the cell glycocalyx. The
cytoskeleton structures (actin filaments and tubulin) were tagged using BacMam fluorescent probes
during transcription. In this way, cytoskeleton compartments exhibit unique excitation and emission
wavelength. Through this multiplexed approach, the displacement and spatiotemporal distribution of
each individual cellular compartment was tracked independently under the controlled mechanical
stress. Finally, by pairing a multiphysics computational model with the live cell experimental mechanics
approach, stem cell mechanoadaptation was mapped as a function of specific mechanical cue libraries
(referred to as 'mapping the mechanome'). This novel imaging technique enabled us to image changes in
the cell membrane, nucleus and cytoskeleton at the same time. Furthermore, by using the multiplexed
imaging fluorescence technique and different fluorescent dyes, it is now possible to separate and study
each cellular compartment individually and simultaneously at different time points.
Understanding the mechanisms underlying stem cell mechanotransduction and mechanoadaptation will
enable prospective guidance of targeted tissue genesis and healing by stem cells.
Regulation of Collateral Branch Formation in Axons by MAP7‐mediated Microtubule Bundle
L. Ma1, S. Tymanskyj1; 1Neuroscience, Thomas Jefferson University, Philadelphia, PA
Microtubules play important roles in cell morphogenesis, but their functions in axonal branch formation
have not been fully understood. Our recent study of microtubule‐associated protein 7 (MAP7, also
known as ensconsin) suggests that regulation of microtubule bundle formation plays a critical role here.
First, MAP7 is expressed at a time when sensory axons in the dorsal root ganglion (DRG) just start to
sprout interstitial collaterals into the spinal cord. Overexpression of MAP7 precociously in young
neurons promotes branch formation, whereas knockdown of MAP7 in old neurons reduces branching.
Structure‐function studies have revealed that the branching activity resides in the amino region that
includes the microtubule binding domain and a dimerization domain rich in phosphorylation sites. A
truncated protein including these two domains binds and bundles microtubules in vitro. When
expressed in non‐neuronal cells, it induced the formation of long, curly, and stable microtubule bundles.
Interestingly, when express in DRG neurons, it also induces axonal branches. Time‐lapse imaging of live
DRG neurons shows that this activity enters nascent branches and accumulates in the base, revealing its
role in branch maturation. In addition, both microtubule bundling and axon branching can be regulated
by phosphorylation at a specific site. To test the role of MAP7 in vivo, we analyzed a spontaneous
mutant mouse (mshi) that expresses this truncated form. Genetic labeling of proprioceptive axons in
the spinal cord reveals more lateral entry of these collaterals, a phenotype that is consistent with the
branch‐promoting activity. These results thus provide strong evidence to support a role of microtubule
bundle regulation in axonal branch formation.
Controlling Cell Shape Affects the Spatial Distribution of Load Across Vinculin.
K.E. Rothenberg1, S.S. Neibart1, A.S. LaCroix1, B.D. Hoffman1; 1Biomedical Engineering, Duke University,
Durham , NC
The shape of an adherent cell is a key determinant of many cellular processes, including gene
expression, proliferation, apoptosis, and stem cell differentiation. Thus, an increased understanding of
the mechanisms mediating cell shape sensing will likely aid endeavors to manipulate cell behavior in
pathophysiological and technological contexts. Shape sensing is at least partially due to mechanically‐
sensitive signaling within focal adhesions. Previous work has shown that the ability of the mechanical
linker protein vinculin to bear tensile loads is a critical determinant of force‐induced focal adhesion
dynamics. Therefore, in this work we evaluate the effects of alterations of cell shape on vinculin load
using a Forster Resonance Energy Transfer (FRET)‐based biosensor. The sensor is comprised of two
fluorescent proteins capable of undergoing FRET connected by a deformable linker. In response to
tensile forces, the linker extends, and the fluorophores separate, leading to a decrease in FRET.
Conversely, compressive forces decrease the separation distance of the fluorophores leading to
enhanced FRET. To control cell shape while measuring the loads supported by vinculin, vinculin deficient
MEFs stably expressing the force‐sensitive vinculin sensor were confined to shapes of varying area or
geometry created through photopatterning techniques. We find that the tension across vinculin
increases at the edges of the cells. However, surprisingly, vinculin supporting compressive loads exists in
the center of cells. Both tensile and compressive loads increase as the aspect ratio and area of the
patterns are increased. Interestingly, pharmacological perturbation of the actomyosin cytoskeleton
ablates tensile loads, but has a limited effect on compressive loads. Overall, these data are consistent
with previous observations of high forces at the edges of cells and compressive forces directly under the
nucleus in traction force microscopy. In these previous studies, the compressive forces were attributed
to apical stress fibers compressing the nucleus. Based on an observed correlation between increased
nuclear deformation, enhanced apical actin organization, and increased compressive loads supported by
vinculin, we suggest a similar mechanism occurs in the patterned fibroblasts used in the current study.
The results from this study show that both tensile and compressive loads across vinculin are able to
maintain focal adhesion assembly. This suggests the existence of a new paradigm for mechanosensitivity
in adhesion mechanobiology, where the compressive and tensile loads across particular proteins must
be considered. Current work addresses the possibility that differential signaling occurs in focal adhesions
supporting distinct types of mechanical load.
Formin‐mediated cortex mechanics coordinate invasion by cell collectives.
T. Fessenden1, Y. Beckham1, G.R. Ramirez‐SanJuan1, M. Manning2, M.L. Gardel1,3,4; 1Institute for
Biophysical Dynamics, University of Chicago, Chicago, IL, 2Physics, Syracuse University, Syracuse, NY,
Physics, University of Chicago, Chicago, IL, 4James Franck Institute, University of Chicago, Chicago, IL
The onset of local invasion defines the transition from benign to malignant tumors, typified by egress of
tumor cells singly or collectively from the primary lesion. While this pathological definition has clear cell
biological implications, a unifying mechanism by which intact cell collectives initiate invasion remains
elusive. We investigated roles of actin nucleators from the formin family and the Arp2/3 complex during
invasion by two cell types: primary mammary tumor explants from mice, and branching morphogenesis
by Manin‐Darby Canine Kidney (MDCK) cells, which undergo collective invasion in response to HGF.
Surprisingly, inhibition of Arp2/3 did not prevent invasion in either cell type. Formin inhibition, however,
drastically reduced invasive fronts in both types. This phenotype was specific to multicellular collectives
in 3D, as formins were dispensable for 2D motility by single cells or cell collectives. We identified Dia1
and FHOD1 as formin family members required for invasion by MDCK acini. MDCK acini in which these
isoforms were stably depleted could develop from single cells and polarize normally, but were unable to
initiate invasion in response to HGF.
We reasoned that formins could act through focal adhesion signaling or though cortical actin stability.
Increasing focal adhesion signaling by plating acini in bundled collagen gels rescued invasion defects
caused by loss of FHOD1, but not by Dia1. To assess roles for Dia1 in actin cortex stability, we used a
physical model describing cell shape in uniformly packed epithelia. We found a softening or fluidization
of tissues deficient for Dia1, but not FHOD1. Together with direct measurements of cell cortical stiffness,
these results show that impaired cortical actin mechanics are required to launch stable invasive fronts
from epithelial tissues. Scattering assays of Dia1‐deficient MDCK cells in 2D showed impaired cell‐cell
rupturing and unstable leading edge dynamics.
Thus we find that, independent of focal adhesion stability, actin cortex mechanics determine invasion
only in multicellular contexts, where dynamic cell‐cell and cell‐ECM interactions make specific local
requirements of cortical stability to promote protrusion and initiate invasion. Further, this work
provides a bridge between cell motility in 3D and canonical models of 2D migration. Our work suggests a
novel biophysical approach for predicting and treating the transition from benign to malignant tumors.
Mena INV localizes to invadopodium precursors in breast carcinoma cells and dysregulates
cortactin phosphorylation to promote matrix degradation by invadopodia.
M.D. Weidmann1, V.P. Sharma1,2, R.J. Eddy1, J.S. Condeelis1,2; 1Anatomy and Structural Biology, Albert
Einstein College of Medicine, Bronx, NY, 2Gruss Lipper Biophotonics Center, Albert Einstein College of
Medicine, Bronx, NY
Invadopodia, actin‐polymerization driven protrusions of invasive carcinoma cells that focally secrete
ECM‐degrading proteases, are essential for tumor cell migration and intravasation during tumor cell
dissemination from the primary tumor. Invadopodium formation involves the recruitment of a core
complex of proteins including cortactin, cofilin, actin, N‐WASP, and Tks5/FISH at the cell membrane. We
have shown that cortactin phosphorylation at tyrosine residues, especially Y421, regulates the stability
of this complex and promotes actin polymerization at invadopodium precursors. However, the
mechanism by which cells regulate the cortactin phosphorylation‐dephosphorylation cycle at
invadopodia is unknown. Mena, an actin barbed‐end capping protein antagonist, is expressed in various
splice‐isoforms. One isoform of Mena, termed MenaINV, is upregulated in invasive sub‐populations of
breast carcinoma cells, and is involved in tumor cell invasion and intravasation. Here we show that
expression of MenaINV sensitizes breast cancer cells to epidermal growth factor (EGF)‐induced
invadopodium precursor formation and EGF‐induced phosphorylation of Y421‐cortactin at invadopodia.
Forced MenaINV expression also increases invadopodium degradation to a far greater extent than
equivalent forced expression of other Mena isoforms, and MenaINV forced expression accelerates the
invadopodium maturation process. Specific knock down of MenaINV, but not Mena11a, with siRNA
significantly reduces invadopodium maturation while sparing the formation of invadopodium
precursors. We show that MenaINV is recruited to invadopodia in the late precursor stage and
dysregulates cortactin Y421 phosphorylation by inhibiting normal dephosphorylation of this residue. The
resulting hyperphosphorylated pY421 cortactin reduces inhibition of cofilin activity, resulting in
enhanced actin polymerization, invadopodium assembly and invadopodium function, leading to ECM
degradation and invasion. These results are consistent with the requirement for MenaINV in
invadopodium‐dependent transendothelial migration during intravasation.
Microsymposium 6: Studying Organelle Function: New Trends and Technologies
S. Guo1,2, R. Veneziano2, S. Gordonov2, D. Park3, T. Kulesa2,4, P. Blainey2,4, E. Boyden2,3, M. Bathe2;
Chemistry, MIT, Cambridge, MA, 2Biological Engineering, MIT, Cambridge, MA, 3Media Lab, MIT,
Cambridge, MA, 4Broad Institute of Harvard and MIT, Cambridge, MA
Neuronal synapses form critical junctions of communication in neuronal networks, mediating neuronal
signal transmission and circuit function. Synapses consist of thousands of proteins organized on the sub‐
micron scale, and their dysregulation via genetic aberrations including copy number variations and site‐
specific mutations is associated with a large number of neurological and psychiatric diseases.
Understanding how these genetic aberrations affect the localization and structural organization of
synapse proteins at the single synapse level is crucial for understanding neuronal function and related
pathogenesis. Super‐resolution fluorescence imaging is a powerful approach to resolving nanometer‐
scale organization of synapse molecules. However, conventional super‐resolution imaging is limited to
simultaneous interrogation of only 2‐4 proteins in a single synapse. As an alternative, here we apply
DNA‐PAINT (Points Accumulation for Imaging in Nanoscale Topography) that enables highly multiplexed
super‐resolution imaging of synaptic proteins. PAINT generally employs transiently binding imaging
probes to molecular targets in order to generate target blinking while simultaneously allowing probe
wash‐out or exchange, thereby in principle enabling sequential imaging of arbitrary numbers of
molecular targets using a single dye and laser source. To test the affinity of the synaptic antibody is not
altered by DNA conjugation, we compare co‐localization scores of the antibody staining with the
synaptic marker. We employ DNA‐PAINT to resolve the localization and organization of ~10 targets
simultaneously including synaptic proteins and cytoskeletal markers. Distributions of synaptic proteins
within individual synapses are consistent with previous study using three‐color STORM.
A novel video bioinformatics toolbox to study mitochondrial morphology, dynamics, and
mitophagy in stressed stem cells.
A. Zahedi1, R. Phandthong1, V. On2, P. Talbot1; 1Cell Biology Neuroscience, University of California
Riverside, Riverside, CA, 2Electrical Engineering, University of California Riverside, Riverside, CA
Video bioinformatics is a powerful technology for studying changes in mitochondrial morphology,
dynamics, and health. The objective of this project was to test, using video bioinformatics tools, the
hypothesis that electronic cigarette (EC) liquids and aerosols damage mitochondria. Mouse neural stem
(mNSC), selected for their well‐defined mitochondria, were nucleofected to stably express a MitoTimer
reporter with the colorimetric Timer protein tagged to the cytochrome c oxidase subunit VIII gene.
MitoTimer‐mNSC report on mitochondrial protein oxidation levels by an irreversible fluorescent shift
from green to red. Time‐lapse images were collected at millisecond resolution, and bioinformatics
software was developed to quantify protein oxidation based on the relative red/green fluorescence. To
induce stress, MitoTimer‐mNSCs were treated for 24 hours with EC liquids or aerosols from a major
tobacco company. The liquid/aerosol treated cells showed significant elevation in mitochondrial protein
oxidation (p<0.01 versus untreated controls). Oxidation was preventable by a reactive oxygen species
scavenger, N‐acetyl‐L‐cysteine (NAC). The mitochondria were then segmented using CellProfiler and
classified as round, networked, or swollen. EC aerosol/fluid‐treated mNSC exhibited a dose‐dependent
shift from the round (control) to networked (low dose) to swollen (high dose) phenotypes.
Mitochondrial dynamics were quantified using a pixel‐based motion magnification algorithm written in
Matlab. The control group exhibited slight movement (mode = 0.02 Δ intensity motion vector),
correlating with basal levels of mitochondrial motion. Low dose treatment with menthol‐flavored E‐
liquids resulted in a significant increase in movement (mode = 0.04 Δ intensity motion vector), possibly
due to increase of mitochondrial fusion resulting in the networked phenotype. High doses with menthol
E‐liquids resulted in little or no movement (mode = 0 Δ intensity motion vector). TMRM fluorescence,
which measures mitochondrial membrane potential, was lost at high doses, indicating leaky, damaged
mitochondria. To determine if EC liquid/aerosol can induce mitophagy, a three‐channel mitophagy‐
reporter cell line was used consisting of a mCerulean‐tagged mitochondria and a GFP‐RFP‐tagged‐LC3
(marker for autophagosomes). EC aerosol‐treated cells exhibited an increase in the number and size of
autophagosomes compared to controls. Video bioinformatics software showed that exposure to EC
fluid/aerosol increased oxidation of mitochondrial proteins, disrupted mitochondrial morphology and
membrane potential, and increased mitophagy. Disruption of normal mitochondrial dynamics and the
mitophagy cycles can cause deleterious effects, which have been linked to neurodegenerative disorders.
Analyzing the spatial organization of synaptic molecules using SIM and object‐based statistics.
T. Lagache1, A. Grassart1, N. Sauvonnet1, O. Faklaris2, L.A. Danglot2, J. Olivo‐Marin1; 1Cell Biology and
Infection, Institut Pasteur, Paris, France, 2Institut Jacques Monod, Université Paris Diderot, Paris, France
Determining the precise localization of molecules using image based analysis is a standard and powerful
tool to probe molecular assembly in cellular compartments. While standard colocalization techniques
have been widely used in standard fluorescence microscopy, the emergence of super resolution
microscopy made obsolete the classical overlap approaches and calls for new methods of analysis. We
developed a method and software (SODA: Statistical Object Distance Analysis) that statistically
quantifies the relative spatial positioning of several molecules' populations. The method computes the
coupling distance and specifically quantifies the morphology (size, shape, intensity) of coupled spots. We
studied the apposition of three consensual synaptic molecules (synapsin, PSD95 and homer) in primary
hippocampal neurons with structured‐illumination microscopy. We found that a large pool of small
PSD95 spots, that were previously unobservable in widefield or confocal microscopy, were neither
synaptic nor associated with homer and that only half of the synaptic PSD95 were effectively coupled to
homer. Moreover, we found that the distance between post‐synaptic anchoring molecules was
significantly smaller (2‐3 fold) than the distance between pre‐ and post‐apposed molecules. SODA thus
provides a general automatic method to statistically map the diversity and geometry of molecular
Global membrane geometry rather than membrane curvature underlies Min oscillations in
Escherichia coli.
J. Shen1,2,3, Y. Chang4, C. Chou1; 1Institute of Physics, Academia Sinica, Taipei, Taiwan, 2Department of
Engineering and System Science, National Tsing Hua University, Hsinchu, Taiwan, 3Nano Science and
Technology Program, Taiwan International Graduate Program, Academia Sinica, Taipei, Taiwan,
Department of Physics, National Taiwan Normal University, Taipei, Taiwan
Septum positioning at mid‐cell in many Proteobacteria is ascribed to the sensing capability of the Min
proteins, which undergo dynamic pole‐to‐pole oscillations; it is currently uncertain whether these
oscillations are driven by curvature‐mediated localization of MinD around membranes with high
negative curvature (diffusion and capture) or whether geometric boundaries imposed by membrane
patches determine the reaction‐diffusion propagation axis of the Min proteins. Here, the use of
microfluidic confinement to gently reshape round mutants of the rod‐shaped bacterium Escherichia coli
revealed a dominant distribution of MinD around positively curved regions of the cell periphery, directly
challenging the diffusion‐and‐capture hypothesis. Our phenomenological formalism based on the
principle of wave spreading and the permissive modes of chemical waves indicated that global
membrane geometry is the primary cue for the localization of membrane‐bound MinD. The
experimental results analyzed by the phenomenological formalism also suggest that while Min dynamics
could be stochastic in time in certain cellular geometries, the spatial pattern of Min localization over the
cell periphery is deterministic. Taken together, our investigations inspires a framework for quantitative
analysis of other wave‐like biopatterning systems without solving reaction‐diffusion equations, such as
(cancer) cell migration driven by cytoskeletal waves and deformation‐induced morphogen patterning in
early embryonic stem cells.
Protein disorder and protein‐RNA interactions drive phase separation into liquid droplets with
tunable viscoelasticity and dynamics.
S. Elbaum‐Garfinkle1, N. Vaidya1, N. Taylor1, C.P. Brangwynne1; 1Chemical Bioengineering, Princeton
University, Princeton, NJ
P granules, nucleoli and stress granules are membrane‐less RNA/Protein organelles that display liquid‐
like properties and may assemble by intracellular phase separation, similar to the condensation of water
vapor into droplets. However, the molecular driving forces and the mechanisms underlying these
dynamic condensed phases remain poorly understood. Recently we have shown that an intrinsically
disordered protein domain of the P granule RNA helicase LAF‐1, drives phase separation into liquid
droplets with properties similar to P granules in vivo. Through a combination of microrheology,
fluorescence recovery after photobleaching (FRAP) and confocal imaging, we precisely measure changes
in droplet material properties and dynamics as a function of droplet component or buffer condition. We
find that RNA and salt concentration can tune the viscoelastic properties and molecular dynamics within
LAF‐1 droplets. Specifically, we find that droplet properties are sensitive to length and sequence of RNA
substrates. We further examine the contribution of protein disorder and protein‐RNA interactions
towards the assembly of additional RNA/protein assemblies such as the nucleolus. This work provides
new insight into the molecular mechanism by which disordered proteins and protein‐nucleic acid
interactions give rise to the tunable material properties of dynamic liquid‐phase organelles.
Surface to Volume Relationships in the Phenotype of Vascular Smooth Muscle Cells.
R. Calizo1, P. Rangamani2, R. Iyengar1; 1Pharmacology and Systems Therapeutics, Icahn School of
Medicine, New York, NY, 2Mechanical and Aerospace Engineering, University of California, San Diego,
San Diego, CA
The causal relationships between the plasma membrane shape and downstream signaling events have
not been explored systematically. The local surface to volume (StV) ratios of the plasma membrane, and
their effect on downstream signaling events have been studied recently (1, 2), however, the relationship
between cell shape and their StV relationships between the subcellular organelles, such as the ER and
the nucleus, have not been characterized or explored. Here, we show that global cell shape and plasma
membrane curvature also impacts intracellular StV ratios between the cytosol and subcellular
organelles, such as the ER and the nucleus. Local and global StV relationships between the plasma
membrane and the cytoplasm alter signal propagation through IP3 and calcium. We modeled the
signaling pathway from the M3R/Gαq/PLCβ/IP3 to calcium and its downstream effectors using reaction‐
diffusion formulation. We ran simulations using different cellular geometries with spatial specification
of the different organelles. We used COMSOL Multiphysics to run simulations using the finite element
method. We tested the experimental predictions of this model with vascular smooth muscle cells, which
has 2 biologically relevant shape dependent phenotypes, circular and fusiform, reflecting their
differentiation states. We measured IP3 and calcium signal propagation as well as the activation of the
transcription factor CREB in cells with these 2 mathematically tractable shapes in 2D chips. We further
enhanced the spatial model using averaged topography of plasma membrane and ER using datasets
from super‐resolution microscopy of circular and elongated cells. Using a combination of simulations
and experiments, we were able to show that spatial distribution of plasma membrane receptors and
subcellular organelles are important determinants of a differentiated phenotype of VSMC.
Rangamani P, et al. (2013) Decoding Information in Cell Shape. Cell 154(6):1356‐1369.
Neves SR, et al. (2008) Cell shape and negative links in regulatory motifs together control spatial
information flow in signaling networks. Cell 133(4):666‐680.
Mutant KRAS‐dependent Argonaute 2 (Ago2) sorting regulates miRNA secretion into exosomes.
A.J. McKenzie1, D. Hoshino2, D.J. Cha3, J.G. Patton3, R.J. Coffey4, A.M. Weaver1,5,6; 1Cancer Biology,
Vanderbilt University, Nashville, TN, 2Cancer Cell Research, Kanagawa Cancer Center, Yokohama, Japan,
Biological Sciences, Vanderbilt University, Nashville, TN, 4Medicine, Vanderbilt University, Nashville, TN,
Cell and Developmental Biology, Vanderbilt University, Nashville, TN, 6Pathology, Microbiology, and
Immunology, Vanderbilt University, Nashville, TN
Secretion of miRNAs in extracellular vesicles (EVs) is a novel form of intercellular communication. The
presence of miRNA binding and processing proteins in EVs can affect the efficacy of secreted miRNAs.
However, the mechanisms governing miRNA and miRNA‐binding protein secretion into exosomes
remain largely unknown. Recently, mutant KRAS was shown to regulate both the protein and miRNA
content in exosomes. Here we use an isogenic cell line model to show that mutant KRAS controls the
secretion of the critical RNA‐induced silencing (RISC) machinery component, Argonaute 2 (Ago2) into
exosomes. Ago2 phosphorylation downstream of KRAS leads to decreased association of Ago2 with
multivesicular endosomes (MVE) and a concomitant decrease in Ago2 secretion in exosomes.
Mechanistically, inhibition of MEKI/II, but not Akt downstream of active KRAS reverses the KRAS‐
dependent phosphorylation of Ago2 and leads to increased Ago2‐MVE association and Ago2 secretion
into exosomes. Moreover, inhibition of Ago2 phosphorylation rescues the secretion of KRAS‐dependent
miRNAs in exosomes. These data establish a mechanism for regulating miRNA secretion into exosomes
downstream of KRAS and demonstrate that oncogenic signaling pathways can regulate the secretion of
specific miRNAs.
Minisymposium 01: Cell Migration in Tissues
Mechanical force dynamics during 3D collective migration.
A.S. Piotrowski1, V.D. Varner1, C.M. Nelson1,2; 1Chemical Biological Engineering, Princeton University,
Princeton, NJ, 2Molecular Biology, Princeton University, Princeton, NJ
Collective cell migration drives tissue remodeling during morphogenesis, wound repair, and metastasis.
The physical mechanisms that enable cells to move cohesively during three‐dimensional (3D) collective
cell migration remain unclear. Here, we quantified the physical forces exerted by multicellular cohorts as
they migrated collectively through 3D collagenous extracellular matrix (ECM), and found that this form
of migration occurs via a dynamic pulling mechanism. Tensile forces accumulated at the invasive front of
the multicellular cohorts, and served both a physical, propelling role as well as a regulatory one by
conditioning the cells and matrix for further extension. These forces induced mechanosensitive signaling
within the cells located at the leading edge and caused them to align the ECM, creating microtracks
conducive to further migration. Surprisingly, the tensile forces and collective movements were not
continuous: cells periodically released their grip on the surrounding ECM, causing the collective to
retract in phase with ECM deformations. Our data suggest that migrating cohorts use spatially localized,
long‐range forces and to align and navigate through the ECM.
A novel image‐guided genomics approach to dissect the mechanisms of collective cancer cell
J.M. Konen1, A. Marcus2; 1Graduate Program in Cancer Biology, Emory University, Atlanta, GA,
Hematology and Medical Oncology, Emory University, Atlanta, GA
Collective migration is a normal mechanism used by cells during embryonic development, wound
healing, and vessel development, where cells migrate in a pack towards a target. During tumor
progression, cancer cells can aberrantly activate this collective invasion program to invade into the
tumor microenvironment. The collective cellular pack is phenotypically heterogenous and contains
leader cells which pioneer invasion into the microenvironment along with follower cells that
immediately attach to and follow the leaders. To dissect the genomic and molecular basis of collective
cancer invasion, we developed a technique termed spatiotemporal genomic analysis (SAGA), which uses
image‐guided genomics to precisely select living, rare cell populations that are maintained within a
physiologically relevant environment for downstream genomic and molecular analyses. Using 3‐D lung
cancer cell spheroids, we precisely selected as few as 10 leader cells using the SAGA technique,
extracted them from the bulk of a multicellular cancer spheroid embedded in a 3‐D matrix, and
compared leader cell gene expression patterns to follower cells. This analysis revealed significant
enrichment in alterations of genes within the cell adhesion and vascular endothelial growth factor
(VEGF) signaling pathways in leader cells as compared to follower cells. The SAGA technique also
allowed purified cells of interest to be cultured post‐selection. Therefore, we created the first ever
leader and follower purified cell cultures. The purified leader cell cultures show highly dynamic invasive
patterns, thus maintaining the initial invasive potential they possessed in the parental line, while
follower cells have limited invasive capabilities. Reintroducing limited numbers of leader cells, or even
leader cell conditioned media, into follower cell cultures promoted invasion and motility in the follower
cells through upregulation of VEGF secretion. Overall, our data show that SAGA can precisely select
living cells based upon dynamic behaviors for genomic analysis and can amplify rare cell populations for
subsequent molecular, cellular, and proteomic analyses. Therefore, this image‐guided method has the
potential to impact the field of tumor heterogeneity by uncovering genomic signatures of rare yet
dynamic subpopulations within a heterogeneous cancer population.
Twist1‐induced epithelial dissemination is regulated by cell adhesion and heterotypic cell‐cell
E.R. Shamir1, K. Sirka1, K. Coutinho1,2, M. Auer2, A.J. Ewald1; 1Cell Biology, Johns Hopkins School of
Medicine, Baltimore, MD, 2Life Sciences, Lawrence Berkeley National Laboratory, Berkeley, CA
Cancer metastasis involves the dissemination of cancer cells away from the primary tumor, a process
that is challenging to visualize in vivo. We recently described a novel 3D organoid assay in which acute,
inducible expression of the transcription factor Twist1 is sufficient to induce rapid, robust dissemination
of otherwise normal cells out of primary mouse mammary epithelium. A core concept in metastasis is
that cancer cells must lose cell‐cell adhesion to detach from the tumor. Contrary to expectation, Twist1+
disseminating cells had membrane‐localized E‐cadherin and β‐catenin, and complete knockdown of E‐
cadherin strongly inhibited Twist1‐induced dissemination.
We are now leveraging this model to study how Twist1+ cells initially detach from a multicellular,
adherent tissue and migrate through the extracellular matrix as single cells. By time‐lapse microscopy,
we observe that Twist1+ cells disseminate and migrate by filopodial, amoeboid motility, and their initial
migration path shows high directional persistence. By electron microscopy, protrusive Twist1+ cells at
the basal surface are connected to the main organoid by multiple cell‐cell junctions, consistent with our
previous data. However, these junctions are distinct from classic zonula adherens and tight junctions
and appear to represent desmosomes. We hypothesize that desmosomes limit dissemination and are
now testing the requirement for desmosomal adhesion by shRNA knockdown of desmosomal cadherins.
A major question is why some Twist1+ cells disseminate while many remain in the epithelium. We
previously showed that constitutive Twist1 expression induced dissemination of both mammary cell
types: inner luminal and basal myoepithelial cells. To distinguish whether these cell types have different
dissemination potential, we have developed mouse models to restrict Twist1 to distinct mammary
lineages. Myoepithelial‐specific Twist1 expression induces cell autonomous myoepithelial dissemination.
In contrast, luminal‐specific Twist1 expression rarely results in dissemination. Most breast tumors arise
from luminal cells, and the integrity of the myoepithelium is the major clinical diagnostic criteria used to
distinguish in situ from invasive ductal carcinoma. In turn, we hypothesized that normal myoepithelial
cells block Twist1+ luminal cell escape. Knockdown of the myoepithelial contractility gene smooth muscle
actin (SMA) in organoids with luminal‐specific Twist1 expression significantly increases dissemination,
supporting a role for the myoepithelium in restraining dissemination and for SMA in maintaining
myoepithelial integrity. Taken together, our data demonstrate that intercellular adhesion and
heterotypic cellular interactions critically regulate the efficiency of epithelial dissemination.
PDGF signaling directs the medial movement of cardiomyocytes during the assembly of the
heart tube.
J. Bloomekatz1, A. Dunn1, M. Vaughan1, D. Yelon1; 1Division of Biology, University of California, San Diego
(UCSD), San Diego, CA
Heart tube assembly begins with the movement of bilateral epithelial sheets of cardiomyocytes toward
the midline, where they merge together to form a tube in a process called cardiac fusion. Although, cell
behaviors underlying cardiac fusion have been previously characterized, the central questions of which
signals and forces direct this medial collective movement remain unanswered. Using a combination of
forward genetics and high‐resolution cell tracking in zebrafish, we have revealed a novel role for
platelet‐derived growth factor (PDGF) signaling in directing cardiomyocyte movement. We identified an
ENU‐induced mutation named refuse‐to‐fuse (ref) that inhibits cardiac fusion, resulting in a bifurcated
ventricle. Positional cloning revealed that the ref mutation disrupts the splicing of pdgfra, causing
premature truncation of the receptor PDGFRa. Although Pdgfra has been shown to be expressed in
cardiac progenitors and to be important for later aspects of cardiac morphogenesis, it has not been
previously implicated in cardiac fusion. To identify the mechanism through which PDGF signaling
regulates cardiac fusion, we first examined the anterior endoderm, since it is known to facilitate
cardiomyocyte movement. However, there are no evident anterior endoderm defects in ref mutants.
Epithelial cells are known to create medial movements through the regulated exchange of neighbors,
but cardiomyocytes seem to maintain their neighbors during cardiac fusion. Instead, our cell tracking
studies reveal that cardiomyocytes in ref mutants are not directed medially and can even move laterally,
suggesting a laterally directed force opposing the medial movement of the cells. Additionally, ref mutant
cardiomyocytes fail to transform from the symmetrical hexagonal shapes found prior to cardiac fusion
into the more elongated shapes found at the end of the process. pdgfra is expressed in cardiomyocytes
prior to their movement toward the midline, but is then downregulated during cardiac fusion.
Interestingly, we found that the PDGF ligands pdgfaa and pdgfab are expressed in tissues medially
adjacent to cardiomyocytes as they begin their migration. Global overexpression of pdgfaa causes
cardiac fusion defects, suggesting an instructive role for PDGF signaling in regulation of cardiomyocyte
movement. Together, our data indicate that PDGF signaling is one of the previously unknown pathways
that acts within developing cardiomyocytes to direct their collective movement towards the midline.
Role of exosomes in promoting directional migration of cancer cells.
B. Sung1, T. Ketova2, D. Hoshino3, A. Zijlstra1,2, A.M. Weaver1,2,4; 1Cancer Biology, Vanderbilt University
School of Medicine, Nashville, TN, 2Pathology, Microbiology and Immunology, Vanderbilt University
School of Medicine, Nashville, TN, 3Division of Cancer Cell Research, Kanagawa Cancer Center,
Yokohama, Japan, 4Cell and Developmental Biology, Vanderbilt University School of Medicine, Nashville,
Exosomes are late endosome‐derived extracellular vesicles that carry multiple motility‐promoting
cargos. The exact role of exosomes in promoting motility under diverse circumstances is unclear.
Likewise, whether the process of exosome secretion is essential for cell migration is an open question.
Here, we show that exosome secretion promotes directionally persistent migration of cancer cells
through tissues. Intravital live imaging reveals that exosome secretion promotes a polarized cell
morphology and stabilization of lamellipodial protrusions. Mechanistic in vitro experiments demonstrate
that extracellular matrix (ECM) is an important exosome cargo that enhances motility speed. Consistent
with a major role for ECM as a critical motility‐promoting exosome cargo, live imaging experiments
reveal that exosomes are secreted at the sites of nascent adhesions and promote their assembly.
Additional experiments in diverse in vitro environments suggest a role for additional exosome cargoes in
directional sensing. Altogether, our results indicate that autocrine secretion of exosomes is a critical
feedback mechanism that promotes cancer cell motility.
Differentiation of the invasive phenotype requires G1 cell cycle arrest and HDAC‐mediated
regulation of gene expression.
A.Q. Kohrman1, M. Chandhok1, W. Zhang1, D.Q. Matus1; 1Biochemistry and Cell Biology, Stony Brook
University, Stony Brook, NY
Cell invasion serves as a key mechanism underlying cell dispersal and organ formation during normal
development but is misregulated during many pathogenic processes, including metastatic cancer. Due
to difficulties of examining this dynamic behavior in vivo, we have a poor understanding of the genetic
and epigenetic controls of cell invasion. We are utilizing a powerful in vivo model to examine cell
invasive behavior by combining functional genomic and genetic tools with single‐cell visual analyses. We
examine anchor cell (AC) invasion into the vulval epithelium during C. elegans larval development. From
an RNA interference (RNAi) screen we have identified a transcriptional program that directs the invasive
AC into a cell cycle arrested state. Using lineage tracing, single‐cell transcript analysis, cell cycle
manipulation and live imaging we show that the nuclear hormone receptor nhr‐67 (NR2E1/TLX) is
required in the AC to prevent cell division, in part through upregulation of the cyclin‐dependent kinase
inhibitor cki‐1 (p21Cip1/p27Kip1). Loss of nhr‐67 results in non‐invasive mitotic ACs that fail to express
matrix metalloproteinases (MMPs) and actin regulators or form invadopodia, F‐actin rich membrane
protrusions that facilitate invasion. AC invasion can be rescued through induction of G1 arrest,
preventing cell division and promoting differentiation. Additionally, we show that G1 arrest is necessary
for the activity of the histone deacetylase, HDA‐1, a key regulator of cell differentiation, to regulate the
expression of pro‐invasive genes and invadopodia formation. Together these results suggest that
invasive cell fate is a post‐mitotic differentiated state, which may explain the paradoxical observation
that many metastatic cancers are non‐proliferative.
Intravital Imaging Reveals Ghost Fibers as Architectural Units Guiding Muscle Progenitors
During Skeletal Muscle Regeneration.
M.T. Webster1, U. Manor2, J. Lippincott‐Schwartz2, C. Fan1; 1Embryology, Carnegie Institution for Science,
Baltimore, MD, 2NICHD, National Institutes of Health, Bethesda, MD
Many adult tissues regenerate following injury by means of resident stem cells and their immediate
progenitors. Understanding how these cells’ behaviors are orchestrated to rebuild tissues of pre‐injury
size and organization, i.e. proportional regeneration, has been limited by inability to directly visualize
the regenerative process. Using intravital two‐photon imaging, we present the first direct visualization
of skeletal muscle stem/progenitor cell mediated regeneration in live mice. By combining fluorescent
stem cell lineage markers with second harmonic generation that reveals the extracellular matrix
remnant of damaged muscle fibers, termed “ghost fibers,” we have determined how ghost fibers
influence stem/progenitor cell behavior. Consistent with prevailing models, stem cells are immobile and
quiescent without injury whereas their activated progenitors migrate and divide after injury.
Unexpectedly, progenitors are contained within individual ghost fibers, where cell divisions and
migration are bi‐directionally oriented along the ghost fiber’s long axis. This bi‐directional, uniaxial
movement leads to spreading of myogenic progenitors throughout ghost fibers prior to fusion and
muscle formation. Re‐orienting ghost fibers impacts myogenic progenitors’ migratory paths and division
planes, causing disorganization of regenerated muscle fibers. Thus we have established the architectural
role of ghost fibers for proportional regeneration after tissue injury by directly observing cellular
mechanisms underlying muscle regeneration.
Decoding Embryonic Developmental Pathways Using 4D‐High Content Imaging of C. elegans
R.A. Green1,2, S.D. Ochoa1,2, R. Khaliullin1,2, S. Wang1,2, Z. Zhao1,2, R.J. Biggs1,2, A. Gerson1,2, A.B. Desai1,2, K.
Oegema1,2; 1Ludwig Cancer Research, San Diego, CA, 2Cellular and Molecular Medicine, University of
California, San Diego, San Diego, CA
An important current challenge is to systematically define the genetic pathways that drive
embryogenesis‐ a complex process requiring coordination of cell division, signaling, migration,
differentiation, and death. To this end, we have developed a 4D‐high‐content screening based approach
to functionally classify the ~2600 essential developmental genes in the model metazoan, C. elegans. We
imaged two specially engineered marker strains for the duration of embryogenesis (~10hrs), following
RNAi of target genes; these strains read out defects in (1) germ layer specification and positioning and
(2) cell shape changes and cell migration, during morphogenesis. Imaging was performed on a high
throughput spinning disk confocal imaging system (CV1000), which enabled collection of high resolution
developmental data for 80‐100 embryos in a single experiment. Using the first 500 genes as a pilot set,
we validated that our experimental approach recovered the expected phenotypes for well described
developmental genes, and revealed as yet unreported phenotypes for many uncharacterized genes.
Further, this pilot data set has been used to develop custom analysis algorithms, which streamline data
processing and perform automated scoring of specific phenotypic features. Using this approach, each
individual embryogenesis movie was scored for a variety of parameters that describe phenotypic
defects, phenotypic profiles were compared, and genes were clustered into functional groups. To date,
we have collected 4D developmental data on >800 genes. When complete, this will be the first systems‐
level view of embryonic development in a complex multicellular organism. We anticipate this effort will
help reveal the genetic basis for human congenital defects, such as neural tube, craniofacial, and ventral
body wall closure abnormalities.
Macrophage delivery service – a migrating source of extracellular matrix components is
necessary for Drosophila embryogenesis.
Y. Matsubayashi1, B.M. Stramer1; 1Randall Division of Cell and Molecular Biophysics, King's College
London, London, United Kingdom
During development, Drosophila macrophages (hemocytes) are born in the head of the embryo and
subsequently disperse throughout the body cavity (hemocoel) where they adopt an even dispersal
pattern. We recently revealed that these cells require precisely regulated intercellular interactions for
this even patterning to emerge (Davis et al., Cell 2015), yet we still do not understand whether
hemocyte spreading is functionally important for morphogenesis. Here we demonstrate that rapid and
efficient hemocyte dispersal is essential for proper formation of the basement membrane (BM)
throughout the embryo. During development, hemocytes secrete virtually all known components of the
BM, such as Laminin, Collagen IV, and Perlecan. By live imaging endogenously labelled BM during
hemocyte dispersal, we show that while some components such as Laminin are freely diffusing in the
hemocoel, others such as Collagen IV appear to require local deposition. These non‐diffusible BM
components are deposited by hemocytes as they undergo their developmental migrations, and the end
result is an even spreading of matrix. To examine whether even spreading of hemocytes is indeed
essential for embryogenesis and BM formation, we genetically perturbed hemocyte migration.
Concentrating hemocytes in the head of the embryo leads to a gradient of BM from head to tail. This
uneven deposition of matrix leaves organs that span the length of the embryo, such as the gut and
nerve cord, devoid of BM in the posterior of the embryo. While over time the spreading of BM appears
to catch up in these defective embryos, it is too late for embryogenesis: 50% fail to hatch with the
remainder dying as 1st instar larvae due to a number of morphogenetic defects. This work unveils a
surprising role for hemocyte dispersal in the distribution logistics of extracellular matrix, which allows
for an efficient and timely delivery of BM components during embryogenesis.
Davis J, Luchici A, Mosis F, Thackery J, Salazar J, Mao M, Dunn G, Betz T, Miodownik M, Stramer B.(2015)
Inter‐cellular Forces Orchestrate Contact Inhibition of Locomotion. Cell. 161:361‐373.
Minisymposium 02: Cellular Decision‐Making
Combining whole cell modeling and optically reversible spatial mutations to dissect the 3D
circuitry regulating the Caulobacter asymmetric developmental program.
K. Lasker1, L. Shapiro1; 1Developmental Biology, Stanford University, Stanford, CA
How a cell executes different developmental programs from the same genome is a fundamental
question in systems biology. The bacterium Caulobacter crescentus is a valuable model system for cell
cycle control in which a temporally‐controlled genetic circuit is interwoven with the 3D deployment of
regulatory and morphological proteins. This transcriptional circuit is robustly controlled by dynamically
localized signaling networks and targeted cell‐type‐specific proteolysis. Many individual events
comprising the regulatory circuit have been well characterized both genetically and biochemically in
exquisite detail. This unique depth of biochemical and biophysical data brings within reach a systems‐
level computational model of a complex, highly integrated regulatory network operating dynamically in
both space and time.
We have constructed a whole cell spatiotemporal model of the cell cycle regulatory circuit by integrating
all available kinetic and spatial information encompassing the master transcription factors and their
control by subcellular localization of regulatory proteins, phosphosignaling, and proteolysis. The model
is designed to be dynamic such that it can be continually updated as new experimental results are
reported. We provide a web‐based interface to the model to be used by the entire community to test
hypothesis and communicate ideas. One of the key predictions from the model is the critical role of
dynamic protein localization in determining cell fate. To test these predictions we have adapted a light‐
inducible dimerization system in Caulobacter introducing light controlled "spatial‐mutations" for driving
a diffuse protein to specific cellular addresses. Using our optogenetic tools we demonstrated in vivo that
asymmetry is broken by abrogating the pattern of differential localization of key signaling proteins.
Role of the microtubule cytoskeleton in the control of Cdc42 GTPase and fission yeast cell shape
M. Rodriguez1, F. Verde1; 1Molecular and Cellular Pharmacology, University of Miami Miller School of
Medicine, Miami, FL
NDR (nuclear dbf2‐related) kinases, a subfamily of the AGC group of protein kinases, are highly
conserved from humans to yeast. Several studies suggest NDR kinases play an essential role in
fundamental cellular processes, including cell morphogenesis, neuronal differentiation and
tumorigenesis. In Schizosaccharomyces pombe, the NDR kinase Orb6 plays a role in the establishment
of cell polarity and the control of polarized cell growth. S. pombe cells are a great model system to
study cell morphogenesis and polarity. S. pombe cells grow from the cells ends, maintaining a cylindrical
cell shape and a constant cell width. Polarized cell growth is regulated by the spatial and temporal
activation of the small GTPase Cdc42. The levels of active Cdc42 oscillate at the cell tips in an anti‐
correlated manner as a result of coordinated function of Cdc42 regulators, including the Cdc42 GEF
(Guanine Exchange Factor) proteins Gef1 and Scd1, the Cdc42 GAP (GTPase Activating Protein) Rga4,
and Cdc42‐dependent Pak1 kinase, in an interplay of positive and negative feedback mechanisms (Das,
M. et al. 2012). Additionally, Cdc42 activation at the cell tips is negatively regulated by the NDR Orb6,
via direct phosphorylation of Cdc42 GEF Gef1, an orthologue of mammalian TUBA/DNMBP (Das, M. et
al. 2009; Das, M. et al. 2015). Gef1 phosphorylation promotes Gef1 interaction with the 14‐3‐3 protein
Rad24, resulting in the reduced availability of Gef1 and decreased Cdc42 activation (Das, M. et al. 2015).
Here, we report that the microtubule cytoskeleton modulates Orb6‐mediated phosphorylation of Gef1.
Cells treated with microtubule depolymerizing agent, MBC, exhibit asymmetric distribution of active
Cdc42 and reduced bipolar growth activation. Similarly, cells expressing a mutated form of the
microtubule‐dependent protein Tea4‐V223A, unable to bind to phosphatase Dis2, show active Cdc42
localization at only one cell tip. Loss of Orb6‐mediated Gef1 phosphorylation suppresses these
phenotypes in MBC treated or Tea4‐V223A cells and restores the normal symmetric distribution of
active Cdc42 (Das, M. et al. 2015). Furthermore, the levels of phosphorylation of Gef1 are reduced upon
over‐expression of Tea4. Thus, the microtubule cytoskeleton opposes Orb6‐mediated Gef1
phosphorylation, either by protecting Gef1 from interaction with Orb6 kinase, or through the
recruitment of a phosphatase needed to dephosphorylate Gef1. These results suggest the presence of a
microtubule‐mediated mechanism regulating Gef1 phosphorylation, and define a novel mechanism for
the microtubule cytoskeleton to control Cdc42 activity and promote cell shape emergence.
A Handoff Model for How Asymmetric Cell Division Triggers Cell‐Specific Gene Expression in
Bacillus subtilis.
N. Bradshaw1, R. Losick1; 1Molecular and Cellular Biology, Harvard University, Cambridge, MA
How genetically identical daughter cells adopt dissimilar programs of gene expression following cell
division is a fundamental problem for multicellular organisms and for unicellular organisms with multiple
cell types. To form dormant spores, B. subtilis cells assemble a division septum near a randomly chosen
cell pole, generating asymmetry, which must be passed on to the daughter cells to differentiate. An
unanswered question is how polar septation activates a transcription factor (σF) selectively in the small
cell. We present evidence that the upstream regulator of σF, the phosphatase SpoIIE, is
compartmentalized in the small cell by handoff from the polar septum to the adjacent cell pole where
SpoIIE is protected from proteolysis and activated as a phosphatase. Polar recognition, protection from
proteolysis, and stimulation of phosphatase activity are linked to oligomerization of SpoIIE. This
mechanism for initiating cell‐specific gene expression is independent of additional sporulation factors;
vegetative cells engineered to divide near a pole sequester SpoIIE and activate σF in small cells. Thus, a
simple handoff model explains how SpoIIE responds to a stochastically‐generated cue to activate
transcription at the right time and in the right place.
Studying Chemoattractant Signal Transduction Dynamics in Dictyostelium by BRET.
A.T. Islam1, P.G. Charest1; 1Chemistry and Biochemistry, University of Arizona, Tucson, AZ
Chemotaxis, the directed migration of cells in response to chemical cues (chemoattractants), is central
to normal physiology and is often implicated in the onset and progression of disease, including cancer
metastasis. Therefore, the ability to therapeutically target the chemotaxis machinery would be highly
beneficial. However, the signaling networks and molecular mechanisms underlying chemoattractant
gradient sensing and directional migration of cells are not understood. Key to our understanding of the
mechanisms underlying the directed migration of cells is a better knowledge of chemoattractant
signaling dynamics. The established chemotaxis model Dictyostelium discoideum has proven to be a
great tool to decipher chemoattractant signal transduction, but the methods available to quantitatively
measure signaling dynamics in this organism are limited. We have developed a protocol to quantitatively
study chemoattractant signal transduction in Dictyostelium by monitoring protein‐protein interactions
and conformational changes using Bioluminescence Resonance Energy Transfer (BRET). We have used
BRET to analyze the kinetics and dose‐dependency of heterotrimeric G protein subunit (G 2 and G
dissociation in response to chemoattractant stimulation in Dictyostelium. We show that BRET allows
obtaining quantitative data with high temporal resolution, well suited for the study of chemoattractant
signal transduction in Dictyostelium and the quantitative modeling of chemotactic responses.
Kinome analysis in the giant ciliate Stentor coeruleus.
S.B. Reiff1, P. Sood1, G. Ruby1, M. Slabodnick1, J. DeRisi1, W.F. Marshall1; 1Biochemistry and Biophysics,
University of California San Francisco, San Francisco, CA
The giant unicellular ciliate Stentor coeruleus has the ability to fully regenerate after being cut in half, in
a way that perfectly preserves cell polarity and structure. This regenerative ability has made it a classical
model system for studying regeneration at the cellular level. So far, however, the molecular details
behind this incredible phenomenon have remained largely unstudied. Recently, our laboratory has
developed a system for RNAi knockdown of Stentor genes, and additionally sequenced the Stentor
coeruleus genome. Interestingly, not only do Stentor’s introns appear to possess the smallest average
intron length of any organism described to date at 15 bp, but Stentor also seems to use the standard
genetic code, unlike other ciliates. We wish to understand how the regeneration process is coordinated
at the molecular level. Some of the details of the regeneration process parallel the events of cell
division, and thus we additionally wish to understand whether the cell co‐opts certain cell division
signaling pathways for regeneration.To identify candidates for RNAi knockdown we analyzed the kinome
of Stentor by looking for protein kinase domains among the predicted protein coding genes. Stentor was
found to encode more than 2000 kinases, making up 6% of the total protein coding genes. Many of
these consist of expansions in mitotic kinase families such as PLKs, NDRs, and NEKs. There are also
expansions of families absent in animals and yeast; over 12% of the kinome consists of the calcium‐
dependent CDPK family, originally identified in plants. We also analyzed additional protein domains
found on kinase genes in Stentor, revealing a few novel domain architectures. The most notable
example is an adenylate kinase fused to a calcium‐dependent protein kinase, with a large region in
between containing a AAA+ ATPase and other protein domains. RNAi screening of kinase genes is
ongoing, and will ultimately reveal which of these kinases help to coordinate the many different
precisely timed cellular events required for successful regeneration. In the future, a better
understanding of the mechanisms behind single cell regeneration will have important implications for
basic biology as a whole, and will reveal how these single cells can establish and maintain their polarity
and cortical organization with such a high degree of precision.
Sensing and fusing: how fission yeast use pheromone signaling to achieve fusion.
O. Dudin1, S.G. Martin1; 1Fundamental Microbiology, University of Lausanne, Lausanne, Switzerland
Cell‐cell fusion is a vital process that occurs in many organisms during development and sexual
reproduction. Yeasts cells undergo cell fusion in response to pheromone signalling in order to produce a
diploid zygote. For fusion of walled cells, such as yeasts, the cell wall must be degraded at a precise
location but maintained in surrounding regions to avoid lysis. In this work, we show that a novel actin
structure, which we name the actin fusion focus, is required to achieve cell‐cell fusion. Structured
illumination microscopy and live‐cell imaging revealed that the formin Fus1 nucleates actin filaments
organized as an aster with barbed ends focalized near the plasma membrane. We further show that the
enzymes necessary for cell wall digestion are delivered and focalized at the site of fusion by type V
myosins, while cell wall synthases remain broadly distributed (Dudin et al, The Journal of Cell Biology
2015). The pheromone signaling machinery, including pheromone receptors and transporters, also
focalize at the fusion site. Interestingly, a hyperactive endocytosis‐deficient receptor allele that fails to
focalize, exhibits fusion focus instability and is fusion defective, suggesting that pheromone signaling
focusing is necessary for cell‐cell fusion. Consistently, an M‐cell autocrine strain, expressing the receptor
for its own pheromone, not only readily grows in all directions, but also assembles a fusion focus and
lyses. As cell lysis is suppressed by deleting fus1, this suggests an attempted cell‐fusion event without
partner cell. Furthermore, closely apposed autocrine M‐cells are even occasionally able to fuse together.
These results indicate that cell fusion relies on a positive entrainment between actin fusion focus
formation and focalization of the signaling machinery, which in turn stabilizes the fusion focus. This
allows both partner cells to align their fusion machineries and precisely digest their cell wall at the
future fusion site while avoiding lysis.
Reference: Dudin, O., F.O. Bendezú, R. Groux, T. Laroche, A. Seitz, and S.G. Martin. 2015. A formin‐
nucleated actin aster concentrates cell wall hydrolases for cell fusion in fission yeast. J. Cell Biol.
From plants to yeast and back again: synthetic biology and plant development.
J. Nemhauser1; 1Biology, University of Washington, Seattle, WA
The fundamental building blocks of plant development—cell‐cell communication, cellular
differentiation, cell growth and proliferation—are characterized by complex gene regulatory networks,
many of which include plant hormones. By recapitulating the Arabidopsis thaliana forward auxin signal
transduction pathway in Saccharomyces cerevisiae, we were able to identify and analyze the parameters
of auxin response without interference from other networks. This work revealed that members of the
large Aux/IAA family of auxin repressors exhibit a range of degradation rates and that Aux/IAA
degradation rates drive transcriptional dynamics. These synthetic experiments and subsequent
experiments in transgenic plants demonstrate that Aux/IAA degradation rate can set the pace for critical
developmental events. Our findings lead us to conclude that Aux/IAAs act as auxin‐initiated timers to
facilitate coordinated cell behaviors during developmental transitions.
Minisymposium 03: Chromosome Segregation
Aneuploidy confers a selective advantage to cancer cells by promoting karyotypic
S.D. Rutledge1,2, T.A. Douglas3, C.L. Kantzler1,2, D. Wangsa4, S.D. Kale2, E. Logarinho5, D. Cimini1,2;
Biological Sciences, Virginia Tech, Blacksburg, VA, 2Virginia Bioinformatics Institute, Virginia Tech,
Blacksburg, VA, 3School of Biomedical Engineering and Sciences, Virginia Tech, Blacksburg, VA, 4Genetics
Branch, NIH ‐ National Cancer Institute, Bethesda, MD, 5Instituto de Biologia Molecular e Celular,
Universidade do Porto, Porto, Portugal
An abnormal chromosome number, a condition known as aneuploidy, is a ubiquitous feature of cancer
cells. A number of studies have shown that aneuploidy impairs cellular fitness. However, there is also
evidence that aneuploidy can arise in response to specific challenges and can confer a selective
advantage under certain environmental stresses in unicellular eukaryotes. Cancer cells are likely
exposed to a number of challenging conditions arising within the tumor microenvironment. To
investigate whether aneuploidy may confer a selective advantage to cancer cells, we employed a
controlled experimental system. We used the diploid, colorectal cancer (CRC) cell line DLD1 and two
DLD1‐derived CRC cell lines carrying single‐chromosome aneuploidies, namely trisomy 7 (DLD1+7) and
trisomy 13 (DLD1+13), to assess a number of cancer cell properties. Such properties, which included
rates of proliferation and apoptosis, anchorage‐independent growth, and invasiveness, were assessed
both under normal culture conditions and under conditions of stress (i.e., serum starvation, drug
treatment, hypoxia). Overall, our data show that aneuploidy can confer selective advantage to cancer
cells under environmental stress conditions. These findings indicate that aneuploidy can increase the
adaptability of cells, even those, such as cancer cells, that are already characterized by increased
proliferative capacity and aggressive tumorigenic phenotypes. Interestingly, we did not find any
consistently significant difference between cells with trisomy 7 and cells with trisomy 13. This
observation suggests that the selective advantage conferred by aneuploidy, at least under the
conditions assessed in this study, is not due to mis‐expression of specific aneuploidy‐associated genes.
Instead, our data suggest that aneuploidy may confer a selective advantage to cancer cells by inducing
karyotypic heterogeneity, which we recently found to arise in these aneuploid cell lines due to high rates
of chromosome mis‐segregation.
Quantitative Assessment of Chromosome Instability Induced through chemical disruption of
mitotic progression.
S. Markossian1, A. Arnaoutov1, N.S. Saba2, V. Larionov3, M. Dasso1; 1Laboratory of Gene Regulation and
Development, National Institute of Child Health and Human Development, Bethesda, MD, 2Section of
Hematology and Medical Oncology, Department of Medicine, Tulane University, New Orleans, LA,
Developmental Therapeutics Branch, National Cancer Institute, Bethesda, MD
Most solid tumors are aneuploid, carrying an abnormal number of chromosomes, and they missegregate
whole chromosomes in a phenomenon termed chromosome instability (CIN). CIN is associated with
poor prognosis in many cancer types, and targeting of CIN is an attractive strategy for anti‐cancer
therapeutics. The mechanisms causing CIN and its contributions to tumor initiation and growth are not
well defined, partly because there is no straightforward, quantitative assays for CIN in human cells. To
address this problem, we have developed the first Human Artificial Chromosome (HAC)‐based
quantitative live‐cell assay for mitotic chromosome segregation in mammalian cells, with which we can
easily score the rates of CIN within one cell division under different experimental conditions. We have
constructed a HAC encoding copies of enhanced green fluorescent protein (eGFP) fused to the
destruction box (DB) of hSecurin, a substrate of the anaphase promoting complex/cyclosome (APC/C)
ubiquitin ligase, which becomes active during anaphase to catalyze the proteolysis of critical mitotic
target proteins. This HAC also contains tet operator (tetO) arrays and sequences encoding the
tetracycline repressor fused to monomeric cherry fluorescent protein (tetR‐mCherry). We have
produced human U2OS cells (U2OS‐Phoenix) carrying this HAC, in which we monitor HAC segregation in
two ways: First, APC/C degrades the DB‐eGFP fusion expressed from the HAC at anaphase onset, and
DB‐eGFP re‐accumulates in the daughter cells after G1 phase, when APC/C becomes inactive. Daughter
cells that do not obtain a copy of the HAC will thus be GFP negative in the subsequent interphase.
Second, because tetR‐mCherry binds to the tetO arrays, the HAC itself could be followed by live imaging.
Following the HAC by live cell imaging experiments, we show that U2OS‐Phoenix cells have low inherent
levels of CIN, but HAC mis‐segregation is markedly increased by treatment with Reversine, an inhibitor
of Mps1, and microtubule agents Nocodazole and Taxol. In summary, we have developed new assays to
score CIN levels in human cells and have shown that CIN levels increase upon chemical disruption of
mitotic progression, demonstrating the utility of this assay for chemical screens of CIN‐inducing
Aurora Kinase Phosphorylation of the Ndc80 Tail Antagonizes Ska Complex‐Dependent
Lockdown of Microtubule Attachments.
D.K. Cheerambathur1, K. Oegema1, A.B. Desai1; 1Dept. of Cellular and Molecular Medicine, University of
California, San Diego, La Jolla, CA
Aurora kinases ensure accurate chromosome segregation by controlling the stability of kinetochore‐
microtubule attachments. A major conserved Aurora target at kinetochores is the 4‐subunit Ndc80
complex, which is the primary mediator of load‐bearing attachments that ensure faithful chromosome
segregation. The basic, unstructured N‐terminal tail (N‐tail) of the Ndc80 subunit harbors multiple
Aurora target sites. Phosphorylation of the Ndc80 N‐tail is proposed to regulate kinetochore‐
microtubule attachments by altering the electrostatic affinity of the basic N‐tail for the acidic
microtubule lattice. However, using an in vivo assay that reports on the timing and strength of
kinetochore‐microtubule attachments, we find that deletion of the N‐tail does not affect either
kinetochore‐microtubule attachment kinetics or strength. In contrast, mutations in the adjacent
calponin homology (CH) domain of Ndc80 that docks onto inter‐monomer interface on the microtubule
lattice, cause significant defects in attachment formation. Surprisingly, despite the lack of an effect of
removing the N‐tail, mutation of Aurora target sites in the N‐tail to alanine resulted in significantly
accelerated stabilization (‘lockdown’) of kinetochore‐microtubule attachments, which was entirely
dependent on lattice docking by the CH domain. Attachment lockdown triggered by the
unphosphorylatable Ndc80 N‐tail was accompanied by premature (late prometaphase) and enhanced
recruitment of the Ska complex to kinetochores; in contrast, in wildtype and the N‐tail deletion, Ska was
only abruptly recruited in a short period prior to anaphase onset. Consistent with a role for the Ska
complex as the effector responsible for attachment lockdown, Ska depletion rescued the premature
attachment lockdown triggered by the unphosphorylatable Ndc80 N‐tail. Based on these results, we
propose that the Ndc80 N‐tail is phosphorylated early in mitosis to limit Ska recruitment and keep
attachments dynamic. Following chromosome biorientation and N‐tail dephosphorylation, Ska is
recruited by an unknown Ndc80‐dependent mechanism to lockdown the attachments. Taken together,
our findings argue against the widely accepted model that N‐tail phosphorylation controls electrostatic
affinity of the Ndc80 complex for the microtubule lattice and instead implicates N‐tail phosphorylation
in controlling Ska‐mediated attachment lockdown.
The Ska complex recruits Protein Phosphatase 1 to the kinetochore and promotes the
metaphase‐anaphase transition.
S. Sivakumar1, P. Janczyk2, Q. Qu1, P. Stukenberg2, H. Yu1, G.J. Gorbsky3; 1Pharmacology, University of
Texas‐Southwestern, Dallas, TX, 2Biochemistry and Molecular Genetics, University of Virginia ,
Charlottesville, VA, 3Cell Cycle and Cancer Biology, Oklahoma Medical Research Foundation, Oklahoma
City, OK
The Spindle‐ and kinetochore‐associated (Ska) protein complex is a hetero‐trimeric complex that
localizes to the kinetochores and spindle microtubules and is required for anaphase onset. The Ska1
protein c‐terminus was shown to bind microtubules. This binding has been proposed to directly stabilize
kinetochore fiber microtubules and to increase the processivity of the Ndc80 complex in binding
depolymerizing microtubules at kinetochores. We now show that the Ska complex is required for
accumulation of Protein Phosphatase 1 (PP1) at kinetochores and that this recruitment requires the C
terminus of Ska1 that contains the microtubule‐binding domain (MTBD). The C terminus of Ska1 is
required for Ska complex interaction with PP1 in vitro. Replacing the MTBD of Ska1 with a direct fusion
to PP1 results in a nearly complete rescue of the Ska depletion phenotype. A point mutant encoding a
phosphatease dead Ska1‐PP1 fusion does not rescue. Indeed, simply expressing the phosphatase‐dead
but not the phosphatase‐active Ska1‐PP1 fusion in otherwise wild type cells phenocopies Ska depletion.
These experiments indicate that a major role of the Ska complex, specifically the C terminus of Ska1, is
the recruitment of PP1 to kinetochores to promote the metaphase‐anaphase transition. We propose
that the microtubule binding properties of the Ska complex serve primarily to regulate the kinase‐
phosphatase balance of kinetochores in metaphase cells rather than directly mediate kinetochore
movement on microtubules.
The Number of Satellite Repeats Dictates Centromere Strength in Mammals.
A. Iwata‐Otsubo1, S.J. Falk2, L. Chmatal1, M.A. Lampson1, B.E. Black3; 1Biology, University of Pennsylvania,
Philadelphia, PA, 2Graduate Group in Cell and Molecular Biology, University of Pennsylvania,
Philadelphia, PA, 3Biochemistry and Biophysics, University of Pennsylvania, Philadelphia, PA
Biased chromosome segregation in female meiosis I, in violation of Mendel’s First Law, is a form of
meiotic drive leading to preferential retention of a chromosome in the egg at the expense of its
homolog, which is lost in the polar body. This phenomenon has significant implications for chromosome
evolution, but the underlying mechanisms remain unclear. Using mouse as a model system, we
previously showed that differences in centromere strength between homologous chromosomes predict
the direction of drive. Stronger centromeres, manifested by increased kinetochore protein levels and
altered interactions with spindle microtubules, are preferentially retained in the egg. What determines
centromere strength is unknown, but likely depends on something intrinsic to the centromere because
stronger and weaker centromeres can exist simultaneously on a single meiotic bivalent (Chmatal et al.,
2014, Curr. Biol. 24:2295‐2300). Here, we analyze centromeric DNA using two parallel approaches,
whole genome sequencing and fluorescence in situ hybridization (FISH). We find that the amount of
minor satellite DNA repeats, at the kinetochore‐forming portion of the centromere, is reduced tenfold at
weaker centromeres relative to stronger centromeres. In contrast, the amount of major satellite repeats
at the adjacent constitutive heterochromatin, where kinetochores never form, is unchanged. In
addition, we find no substantial changes in the DNA sequences of minor satellite monomers between
weaker and stronger centromeres.
We previously showed that a signature 100 bp fragment of DNA is protected from nuclease digestion at
nucleosomes formed from the histone H3 variant, CENP‐A, which epigenetically defines the centromere
(Hasson, et al., 2013, NSMB 20:687‐695). We now find that this signature is a small component of total
minor satellite at stronger centromeres but a large component at weaker centromeres. This result
suggests that minor satellites are primarily occupied by CENP‐A nucleosomes at weaker centromeres, as
opposed to interspersed CENP‐A and histone H3 nucleosomes at stronger centromeres. We propose
that limiting the amount of minor satellite repeats leads to different chromatin organization and
reduced centromere strength. Conversely, expansion of the repeat sequences at the site of kinetochore
assembly leads to stronger centromeres and preferential retention in the egg in female meiosis,
providing a molecular basis for meiotic drive.
Centromere maintenance through error correction of CENP‐A deposition during DNA
Y. Nechemia‐Arbely1, K.H. Miga2, M. McMahon1, D. Fachinetti1, A. Lee1, B. Ren1, D.W. Cleveland1;
Cellular and Molecular Medicine, Ludwig Institute for Cancer Research, UCSD, LA JOLLA, CA, 2Center for
Biomolecular Science Engineering, University of California Santa Cruz, Santa Cruz, CA
The human centromere is defined epigenetically by chromatin assembled with the histone H3 variant
CENP‐A. Although CENP‐A previously bound at each centromere is quantitatively re‐deposited onto each
daughter centromere during DNA replication, assembly of new CENP‐A into centromeric chromatin
occurs only after exit from mitosis, suggesting different forms of centromeric chromatin before and
after S phase. By combining isolation of CENP‐A chromatin, high throughput sequencing, and mapping
onto novel sequence reference models for human chromosome satellite arrays, we now determine that
at least 90% of human CENP‐A bound to centromeric α‐satellite DNA at all cell cycles phases (G1, S/G2
and M) is assembled into homotypic CENP‐A containing octameric nucleosomes. In contrast with
centromeric histone H3‐containing nucleosomes whose crossed DNA termini yield 147 bp of protected
α‐satellite DNA, CENP‐A‐containing centromeric nucleosomes are structurally divergent, with DNA
unwrapping at entry/exit producing 133 bp of α‐satellite DNA protected at all cell cycle points.
Furthermore, we show that during DNA replication CENP‐A previously bound at centromeres is re‐
deposited at centromere sites with the same sequence preferences as seen in its initial loading at
mitotic exit. In contrast, DNA replication removes ectopic deposition of CENP‐A outside of the
centromere that occurred at exit from mitosis. Thus, we have identified a DNA replication‐dependent
error correction mechanism for maintaining CENP‐A loading at centromeres.
Mitotic synthesis and processing of centromere‐derived RNAs promotes kinetochore and
spindle assembly in Xenopus.
A.W. Grenfell1, M. Strzelecka1, R. Heald1; 1Molecular and Cell Biology, University of California, Berkeley,
Berkeley, CA
Noncoding RNAs (ncRNAs) transcribed from centromeric DNA have been identified in many model
systems. These centromeric transcripts are reported to be involved in the deposition of the centromeric
histone CENP‐A, as well as the mitotic retention of CENP‐C, a constitutive centromere protein essential
for kinetochore assembly and function. Binding of centromere‐derived ncRNAs is also essential for full
Aurora B kinase activation, and RNA‐dependent Aurora B activity is important for the regulation of the
microtubule depolymerase MCAK during cell division. Despite a growing number of biological functions
attributed to centromeric ncRNAs, little is known about their biogenesis pathway. Using metaphase‐
arrested Xenopus egg extract, we show that the spliceosome, which removes introns from RNA
Polymerase II (RNAPII) transcripts, is involved in centromeric ncRNA biogenesis. Small molecule
inhibition of spliceosome assembly led to the accumulation of high molecular‐weight centromeric
transcripts. Automated image analysis showed that this treatment also resulted in decreased CENP‐A
and CENP‐C on mitotic chromosomes, as well as spindle defects that correlate with increased
accumulation of MCAK on spindle microtubules. Similar effects were apparent upon mitotic inhibition of
RNAPII transcription initiation, as well as after knockdown of individual snRNAs that make up the
catalytic core of the spliceosome. Our work suggests that co‐transcriptional recruitment of the RNA
processing machinery to nascent centromeric transcripts contributes to kinetochore and spindle
assembly. Furthermore, these results challenge the idea that RNA splicing is globally and completely
repressed during mitosis.
Purification of the chromosome passenger complex from mitotic chromosomes reveals
functions in centromere transcription and cohesion.
L. Liu1, M.C. Barnhart1, E. Zasadzinska1, C.A. Kestner1, J. Yates iii2, D. Daniel Foltz 1, P. Stukenberg1;
Biochemistry and Molecular Genetics, University of Virginia, School of Medicine, Charlottesville, VA,
Chemical Physiology, The Scripps Research Institute, La Jolla , CA
The chromosome passenger complex, which contains the Aurora B kinase as a catalytic subunit, has
established functions as a regulator of kinetochores. However, the bulk of protein localizes to inner
centromeres from late prophase to metaphase. We purified the CPC from micrococcal nuclease
digested mitotic chromosomes to identify the proteins associated with inner centromere CPC. A set of
RNA binding proteins including RBMX was associated with the CPC chromatin. RBMX localizes to
centromeres where it interacts with the CPC. Depletion of RBMX mislocalized CPC, Bub1, and two
histone marks (H2ApT120 and H3pT3) that are required to localize CPC from centromeres. Each of these
events was rescued by targeting Aurora‐B to centromeres. The premature sister chromatid separation
caused by depletion of RBMX was rescued by targeting Aurora‐B to centromeres, arguing that the
cohesion function of RBMX is to localize Aurora B. Since Aurora B has an established role to remove
cohesion from chromosome arms, we also confirmed that Aurora B has a role in protection of
centromeric cohesion. Indeed, Aurora‐B activity is required to maintain centromeric cohesion in cells
where arm cohesion has been removed. In addition cellular accumulation of alpha‐satellite transcripts in
mitosis requires Aurora B activity. Thus our unbiased visualization of the proteins interacting with the
CPC bound to mitotic chromosomes has identified interactions with RNA processing proteins and
functions in centromere transcription and sister chromosome cohesion.
The Genome Proximity Sensor: A novel failsafe that triggers p53 accumulation in aneuploid cells
in response to chromosome missegregation.
C. Day1, Z. Dong1, K.T. Vaughan2, E.H. Hinchcliffe1; 1Hormel Institute, University of Minnesota, Austin,
MN, 2Biological Sciences, University of Notre Dame, Notre Dame, IN
Inadvertent chromosome missegregation in anaphase generates aneuploid cells, but the proliferation of
these cells is normally blocked, because chromosome missegregation also triggers a p53‐dependant
failsafe that triggers cell cycle arrest in the ensuing G1. The molecular mechanisms underlying this
trigger are not known. Here we identify a conserved feedback mechanism that monitors the relative
position of lagging chromosomes during anaphase via the differential phosphorylation of the histone
variant H3.3 at Ser31. During normal mitosis H3.3 Ser31 is phosphorylated exclusively at peri‐
centromeres, which are rapidly dephosphorylated in anaphase. We induced non‐transformed cells to
missegregate chromosomes by transiently depolymerizing spindle microtubules with cold. These cells
undergo the metaphase‐anaphase transition in the presence of one or more misaligned chromosomes
that lack BubR1 labelling. These cells transit mitosis with relatively normal timing and lack DNA damage.
After re‐warming, correlative same cell live and fixed imaging revealed that isolated chromosomes (e.g.
lagging in anaphase) have hyper‐phosphorylated H3.3 Ser31 (pS31) along their arms that persists into
G1 as these chromosomes assemble into a micronucleus. Surprisingly, during telophase Ser31
phosphorylation along individual chromosomes initiates global phosphorylation of H3.3 Ser31 in both
reforming nuclei, suggesting both an amplification step of the aneuploid failsafe, and an explanation for
why both daughter cells trigger p53 activation in response to a single chromosome missegregation
event. pS31 is mimicked by the hyperlocalization of ATRX to isolated chromosome arms. ATRX – a
member of the SWI/SNF family of chromatin binding protein – is known to load histone H3.3 into
nucleosomes. Unlike H3.3 S31 phosphorylation during anaphase, the association of ATRX with isolated
chromosomes is transient; by nuclear envelope reformation ATRX is absent from the resulting
micronucleus. Finally, we demonstrate that post‐anaphase H3.3 pS31 and ATRX are required to trigger
p53 stabilization in the subsequent G1. Microinjection of monospecific antibodies against either pS31 or
ATRX into anaphase cells containing lagging chromosomes blocks p53 accumulation in G1 nuclei¬. Here
we show that p53 cell cycle arrest – triggered by chromosome missegregation – is mediated via a novel
signaling mechanism dependent upon H3.3 S31 phosphorylation and ATRX recruitment to lagging
chromosomes. This work provides insight into how aneuploidy is normally monitored and suppressed.
Furthermore, driver mutations in H3.3 (flanking Ser31) and null mutations in ATRX are both found in
pediatric glioblastomas, suggesting that disrupting the aneuploidy failsafe contributes to neoplastic
Minisymposium 04: Genome Organization and Stability
Early replication stress leads to abnormal mitosis and genome rearrangement.
S.A. Sabatinos1,2, N.S. Ranatunga1, J. Yuan1, M. Green1, S.L. Forsburg1; 1Molecular Computational
Biology, University of Southern California, Los Angeles, CA, 2Chemistry Biology, Ryerson University ,
Toronto, ON
Yeast cell cycle genetics suggests that cells under replication stress generate damage signals that are
sufficient to activate checkpoints and restrain mitosis. Yet in mammalian cancer cells, replication stress
is not necessarily sufficient to block division. This can lead to abnormal mitotic structures including
chromosome bridges, lagging chromosomes, ultrafine anaphase bridges, and micronuclei. These
abnormal mitotic structures are a source of increased mutations including dramatic chromosome
rearrangements, resulting in characteristic genome instability.
We have examined the response to different forms of replication stress in the fission yeast, S. pombe,
employing visual methods and single‐cell live‐cell analysis. First, we observe distinct patterns of single‐
strand binding protein RPA and homologous recombination protein Rad52 depending on the timing and
type of stress we induce. This creates a fingerprint of different S phase‐specific stress responses which
we can correlate to distinct outcomes. Second, we identify a novel stress‐associated phenotype in a
temperature sensitive mutation of the Mcm4 helicase subunit. Despite under‐replicated DNA, the cells
fail to arrest during temperature shift, and following release. Our dynamic pedigrees reveal surprising
examples of division under stress including all the mitotic structures described above, including
apparent micronuclei. These abnormal mitoses are associated with checkpoint evasion, genome
instability, mutations, and chromosome rearrangement in the surviving cells. Our study shows that a
subset of cells within a whole population can continue to divide, despite significant negative effects on
genome integrity, and establishes a genetic model to study mechanisms that allow division to occur
despite stress.
DNA damage and chromothripsis from chromosome segregation errors.
A. Spektor1,2,3, Z. Cheng‐Zhong2,3,4,5, N.T. Umbreit2,3, H. Cornils2,3, J.M. Francis4,5, E.K. Jackson2,3,6, S. Liu2,3,
M.L. Meyerson4,5,7,8, D.S. Pellman2,3,5,6; 1Radiation Oncology, Dana‐Farber Cancer Institute, Boston, MA,
Pediatric Oncology, Dana‐Farber Cancer Institute, Boston, MA, 3Cell Biology, Harvard Medical School,
Boston, MA, 4Medical Oncology, Dana‐Farber Cancer Institute, Boston, MA, 5Broad Institute of Harvard
and MIT, Boston, MA, 6Howard Hughes Medical Institute, Chevy Chase, MD, 7Center for Cancer Genome
Discovery, Dana‐Farber Cancer Institute, Boston, MA, 8Pathology, Harvard Medical School, Boston, MA
Recent cancer genome sequencing suggests that many cancers accumulate large numbers of mutations
rapidly, in some cases even within the course of a single cell cycle. The most dramatic example of such
“punctuated equilibrium‐type” rapid genome evolution is chromothripsis, a new mutational process
characterized by massive chromosome rearrangements and a unique pattern of DNA copy number
oscillations, all curiously restricted to one or a few chromosomes. The mechanism(s) leading to
chromothripsis have been unclear, but our group previously suggested that it could result from the
physical isolation of chromosomes in abnormal nuclear structures called micronuclei. Micronuclei are
deficient in key nuclear functions such as nuclear envelope stability, DNA replication, and transcription,
and because of these abnormalities develop DNA damage upon entry into S‐phase. By combining live‐
cell imaging with single‐cell whole genome sequencing, we show that micronucleation can lead to a
large number of rearrangements on the missegregated chromosome within one cell cycle, in some cases
recapitulating all the hallmarks of chromothripsis. Here, we present our latest efforts to describe in
detail the mechanisms of DNA breakage and chromosome reassembly that ultimately lead to
A chemical proteomics approach reveals direct binders of DNA‐damage‐associated histone
variant gammaH2AX.
R.E. Kleiner1, P. Verma1, K.R. Molloy2, B.T. Chait2, T.M. Kapoor1; 1Laboratory of Chemistry and Cell
Biology, Rockefeller University, New York, NY, 2Laboratory of Mass Spectrometry and Gaseous Ion
Chemistry, Rockefeller University, New York, NY
Post‐translational modifications on histone tails can regulate the localization of chromatin‐associated
proteins implicated in essential cellular processes. These modifications are often sub‐stoichiometric and
mediators of low‐affinity interactions, which makes identifying their ‘readers’ challenging. In particular,
the efficient repair of DNA double‐strand breaks involves the phosphorylation of the histone variant
H2AX (‘γH2AX’), which accumulates in foci at the site of damage. In current models, the recruitment of
multiple DNA repair proteins to γH2AX foci depends mainly upon direct recognition of this mark by a
single protein, MDC1 (mediator of damage checkpoint 1). However, DNA repair proteins can accumulate
at γH2AX foci without MDC1, suggesting that other γH2AX ‘readers’ exist. Here, we use a chemical
proteomics approach to profile direct and phospho‐selective binders of γH2AX in native proteomes. We
identify proteins that ‘read’ γH2AX, including the DNA damage response mediator, 53BP1, which we
show interacts with this ‘mark’ through its BRCT domains. Furthermore, we investigate targeting of wild‐
type 53BP1 or a mutant form deficient in γH2AX binding to chromosomal breaks resulting from
endogenous and exogenous DNA damage. Our results show how direct recognition of γH2AX can
modulate protein localization at DNA damage sites, and suggest how specific chromatin ‘mark’‐‘reader’
interactions contribute to essential mechanisms that ensure genome stability.
Dynamic phosphoregulation of axis proteins underlies chromosome remodeling during meiosis.
Y. Kim1,2,3, S.C. Rosenberg4,5, N. Kostow1,2,3, O. Rog1,2,3, S. Köhler1,2,3, K.D. Corbett4,5, A.F. Dernburg1,2,3,6;
Molecular and Cell Biology, University of California, Berkeley, Berkeley, CA, 2California Institute for
Quantitative Biosciences, Berkeley, CA, 3Howard Hughes Medical Institute, Chevy Chase, MD, 4San Diego
Branch, Ludwig Institute for Cancer Research, La Jolla, CA, 5Department of Cellular and Molecular
Medicine, University of California, San Diego, La Jolla, CA, 6Life Sciences Division, Lawrence Berkeley
National Laboratory, Berkeley, CA
Segregation of chromosomes in meiosis requires their dramatic, stepwise reorganization during meiotic
prophase. Each chromosome must pair, synapse, and recombine with its homolog to give rise to a stable
bivalent structure that can biorient and then divide during Meiosis I. A fundamental mystery is how
these events are coordinated during the meiotic cell cycle. Pairing, synapsis, and recombination depend
on the formation of linear “axes” along each chromosome at meiotic entry, followed by assembly of the
synaptonemal complex (SC) between paired axes. Chromosome axes in C. elegans are comprised of
cohesins and four related HORMA domain proteins: HIM‐3, HTP‐1, HTP‐2, and HTP‐3. We recently
reported that the largest of these proteins, HTP‐3, recruits the other three paralogs through short
peptide sequences (closure motifs) in its C‐terminal tail, which are bound by the HORMA domains of
HTP‐1, HTP‐2, and HIM‐3 (Kim, Rosenberg, et al., 2014). Here we show that these interactions are
dynamically regulated by phosphorylation of the closure motifs in HTP‐3 by two meiotic kinases, CHK‐2
and PLK‐2. Phosphorylation of the four central motifs reduces their binding affinity for HIM‐3, and
occurs in two temporally distinct waves. In early meiosis, phosphorylation along the entire axis by CHK‐2
is required for efficient synapsis. A second wave of phosphorylation by PLK‐2 occurs after crossover
formation, and specifies the “short arm” of the bivalent where cohesion will be released during the first
meiotic division. This second wave of HTP‐3 phosphorylation requires both crossover formation and
recruitment of PLK‐2 to the chromosomes through a binding site in an SC component, SYP‐1. Thus,
phosphorylation‐dependent regulation of HORMA domain protein assembly promotes dynamic
remodeling of chromosome axes during meiotic progression, and is essential for proper segregation of
holocentric chromosomes in C. elegans meiosis.
Pol I transcription and nucleolar structure play key roles in nuclear organization.
C. Wang1, S.B. Sondalle2, S.J. Baserga2, S. Huang1; 1Cell and Molecular Biology, Northwestern University,
Feinberg School of Medicine, Chicago, IL, 2Department of Genetics, Yale University School of Medicine,
New Haven, CT
The nucleolus is the site of ribosome synthesis including the transcription of ribosomal DNA (rDNA) by
RNA polymerase I (Pol I), processing of pre‐rRNA, and assembly of pre‐ribosomal particles. In addition
to ribosomal synthesis, increasing numbers of functions have been implicated for the nucleolus,
including cell cycle regulation, apoptosis, viral infection, DNA replication and DNA damage. Although the
nucleolus is among the first cellular organelles described, how and why it forms and the functional
relationship between nucleolus and the rest of the nucleus remain open questions. Here we ask what
would happen to the nucleus if we selectively disrupt the nucleolus. Although stresses and genotoxicity
have been shown to impact nucleolar structure, these processes often influence more than just the
nucleolus. To selectively tease out the role of nucleoli in nuclear organization, we specifically disrupted
rDNA transcription, the first step in the synthesis of ribosomes by siRNA depletion of the Pol I large
subunit, RPA194. Depletion of RPA194 significantly reduced pre‐rRNA synthesis and induced structural
distortions of the nucleolus similar to those found in ActD treated cells. Nucleoli were segregated to
caps labeled with antibodies against transcription factors, and to another fraction that contains other
components. When cells were treated with an siRNA against Cirhin, while the nucleolar number was
reduced, the subcellular distribution of Pol I transcription, pre‐ribosomal RNA processing factors, and
ribosomal proteins within the nucleolus remained similar to those in the control cells, demonstrating
that disruption of Pol I transcription, but not ribosome synthesis in general, is responsible for nucleolar
segregation. The influence of selective inhibition of Pol I transcription upon nuclear organization will be
"TSA‐Seq" Reveals 3D Organization of the Human Genome.
Y. Chen1, Y. Zhang2, L. Zhang1, E. Brinkman3, Y. Wang4, B. van Steensel3, J. Ma2,4, A.S. Belmont1,4; 1Cell and
Developmental Biology, University of Illinois, Urbana‐Champaign, Urbana, IL, 2Bioengineering, University
of Illinois, Urbana‐Champaign, Urbana, IL, 3Division of Gene Regulation, Netherlands Cancer Institute,
Amsterdam, Netherlands, 4Program in Biophysics and Computational Biology,, University of Illinois,
Urbana‐Champaign, Urbana, IL
Current technologies to study three‐dimensional organization of the genome include microscopy‐based
approaches (such as FISH) and genomic approaches (such as Hi‐C). However, it is difficult to directly link
what you see under the microscope and the results generated by genomic approaches. Here we present
a new immunochemistry‐based, genomic approach “TSA‐Seq” that measures cytological distance of the
whole genome from a particular target, such as a certain nuclear compartment in the nucleus. It is a
“What‐You‐See‐is‐What‐You‐Get” approach that directly connects microscopic images with genome‐
wide maps. Also, compared to Chromatin Immunoprecipitation (ChIP) and DNA Adenine
Methyltransferase Identification (DamID) methods, which report a binary state of on/off molecular
interaction, TSA‐Seq provides a quantitative signal proportional to cytological distance.
Using TSA‐Seq, we have generated complete genome maps reporting average distance of any genomic
region to nuclear speckles and nuclear lamina in human erythroleukemia K562 cells. Our LaminA/C TSA‐
Seq result agrees remarkably well with Lamina‐Associated Domains (LADs) identified by LaminB1 DamID,
but also reveals variations in distance to the nuclear periphery. Computational analysis of our results
shows that speckle association is highly correlated with high gene expression. We found that more than
half of the top 5th‐percentile expressed genes have mean speckle distance less than 0.4 microns,
suggesting nuclear speckles constitute a major nuclear compartment for RNA pol 2 transcription. We
also found a striking correlation of the density of various epigenetic marks as well as replication timing
with distance to nuclear speckles. In addition, our TSA‐Seq map stays the same genome‐wide after heat
shock treatment of the cells, revealing that such genome organization is largely maintained despite of
global transcription inhibition.
The human genome is dynamically polarized during epidermal differentiation.
A.M. Wood1, B. Poll1, S.T. Kosak1; 1Cell Molecular Biology, Northwestern University, Feinberg School of
Medicine, Chicago, IL
Aberrant regulation of gene expression often leads to cell death or disease. The organization of
chromatin within the nucleus is one factor that is involved in regulation of gene expression. Dynamic
changes to this organization are thought to be necessary to achieve the precise and coordinated
modifications in gene expression that occur during normal cell growth. However, an aspect of nuclear
organization that has been largely ignored in mammalian systems thus far is how extra‐nuclear cues
relate to nuclear organization. Such functions often result in asymmetric signaling to the nucleus that
could potentially create a polarized nuclear structure. This study tests the hypothesis that nuclear
polarization is established during epidermal differentiation. Formation of the epidermis involves the
establishment of a stratified epithelium through proliferation and differentiation of progenitor cells
attached to the basement membrane. Therefore, the epidermis is a polarized tissue containing
differentiating cells that exhibit polarized interactions with the extracellular matrix creating an attractive
system to examine nuclear polarity. Using this system, we found that genes involved in the attachment
of progenitor cells to the basement membrane show a dramatic polarization in the basal cell population
where the genes are expressed. However, this polarized gene localization is lost in differentiating cells
that have detached from the basement membrane and silenced the genes of interest. Additionally, we
have shown that artificial tissue dissociation from the basement membrane as well as disruption of the
linker of nucleoskeleton and cytoskeleton (LINC) complex result in loss of the identified nuclear polarity.
Our findings suggest that genome organization is polarized in relation to tissue polarity in the epidermis
providing evidence for nuclear polarity, a novel form of nuclear organization that may play a role in the
regulation of gene expression during cell differentiation.
Minisymposium 05: Mechanisms for Shaping Membranes
Determining the molecular basis of cristae structure by electron cryo‐tomography.
K.M. Davies1,2, A.W. Mühleip1,2, T. Blum1, B. Daum1, C. Anselmi3, J. Faraldo‐Gomez3, W. Kühlbrandt1,2;
Structural Biology, Max Planck Institute of Biophysics, Frankfurt am Main, Germany, 2Cluster of
Excellence 'Macromolecular Complexes', Goethe University, Frankfurt am main, Germany, 3Theoretical
Molecular Biophysics Section, National Heart, Blood and Lung Institute, Bethesda, MD
Mitochondria are an essential organelle of eukaryotic cells. Deeply integrated into the metabolic
pathways of cells, mitochondria not only generate vast quantities of ATP through aerobic respiration but
are also involved in lipid, heme and amino acid biosynthesis, Ca2+ storage, cellular signalling and
programmed cell death. The basic architecture of a mitochondrion consists of a smooth outer
membrane and an inner membrane, which invaginates forming protrusions called cristae. The structure
of the cristae varies both between species and between different types of tissue. Using the technique of
electron cryo‐tomography together with subtomogram averaging and cell biology we are investigating
how proteins influence cristae structure. So far we have shown that the mitochondrial ATP synthase
from all species studied to date form rows of dimers along the highly curved edges of cristae. The
structure of the dimers and the architecture of the rows they form, vary between species and are
probably responsible for the species‐specific variation in cristae shape. Unlike the ATP synthase, the
respiratory chain complexes are randomly distributed in the flat regions of the cristae either side of the
rows of ATP synthase dimers causing the proton sources (the respiratory chain complexes) to be
spatially separated from the proton sinks (ATP synthases). The importance of this organisation is
currently unknown but disrupting the formation of the ATP synthase dimers leads to an increase in
generation time. In addition, we have also been investigating how cristae junctions are shaped. In
metazoan species, cristae are connected to the inner boundary membrane by multiple circular cristae
junctions whereas in fungi, lamellar cristae are connected to the inner boundary membrane by a single
slot‐like junction. Deletion of single proteins of the MICOS complex (‘Mitochondrial Contact Site and
Cristae Organizing System’) causes the cristae junctions in yeast to shorten. The degree of shortening
correlates exceptionally well with the mutant phenotype. Mutants which struggle to grow on non‐
fermentable media have either no cristae junctions or junctions <10nm in diameter, whereas those with
only mild growth defects have mainly wild type cristae junctions.
Ultrafast endocytosis of synaptic vesicles.
S. Watanabe1, T. Trimbuch2, M. Camacho‐Perez2, B. Rost2, C. Rosenmund2, E. Jorgensen1; 1Department of
Biology, University of Utah, Salt Lake City, UT, 2Neurocure, Charite Universitatsmedizin, Berlin, Germany
In the early 70s, two models for synaptic vesicle endocytosis were put forward based on the
ultrastructural studies of frog neuromuscular junctions. Heuser and Reese found that vesicles fuse and
collapse into the membrane and synaptic vesicles are recovered via slow, clathrin‐mediated
endodcytosis. Ceccarelli and his colleagues observed a fusing vesicle with a narrow neck and deduced
that vesicles do not fully collapse into the membrane and that vesicles are recovered thorough a rapid
reversal of the fusion pore. Since then, conflicting evidence has accumulated: the molecular studies
suggest that clathrin and clathrin‐associated proteins are essential, but the kinetics studies have
indicated the existence of both fast and slow mechanisms. The major criticism of the original
morphological studies was that they used intense stimulation to ensure the capture of endocytic events.
To determine how endocytosis takes place after a single action potential, we coupled channelrhodopsin‐
induced synaptic transmission with rapid high‐pressure freezing and captured membrane dynamics
following fusion of synaptic vesicles. We found that vesicle membrane is retrieved within 100 ms after
fusion in mouse hippocampal neurons. This ultrafast endocytic pathway is compensatory – the amount
of membrane retrieved equals the amount of membrane exocytosed. Following ultrafast endocytosis,
large endocytic vesicles are delivered to endosomes. Synaptic vesicles are then regenerated from the
endosomes ~5 s after stimulation. The newly formed vesicles return to active zone, suggesting that
these vesicles are functional synaptic vesicles. To test the requirement for clathrin, we generated
shRNA against clathrin heavy chain. We found that ultrafast endocytosis is intact in the clathrin knock‐
down cells. However, synaptic endosomes were not resolved into synaptic vesicles, suggesting that
clathrin is required for regenerating vesicles from endosomes. These results suggest that regeneration
of synaptic vesicles is a two‐step process: rapid, clathrin‐independent internalization of vesicle
membrane followed by slower, clathrin‐dependent reconstitution of vesicles.
Superresolution imaging of clathrin‐mediated endocytosis in yeast.
M. Mund1, A. Picco1,2, M. Kaksonen1,2, J. Ries1; 1Cell Biology and Biophysics, EMBL, Heidelberg, Germany,
Biochemistry, University of Geneva, Geneva, Switzerland
Clathrin‐mediated endocytosis is a highly intricate cellular process, which involves the ordered
recruitment and disassembly of around 60 proteins. Diffraction‐limited live‐cell microscopy has led to
tremendous insight into composition and dynamics of the endocytic machinery. Electron microscopy on
the other hand offers nanometer resolution, but lacks molecular specificity. Thus, the structural
organization of endocytic proteins in situ is largely unknown.
We employ single‐molecule localization microscopy (PALM/STORM) to study endocytic structures in
Saccharomyces cerevisiae. In this method, achieving the highest possible resolution requires static
structures, which is why cells are typically fixed during sample preparation and temporal resolution is
lost. This is mirrored in considerable heterogeneity among the imaged endocytic sites, which
complicates their interpretation. To address this issue, we simultaneously image a second reference
structure and use their known spatial relation to estimate the time point of single endocytic sites.
We focused on visualizing the coat and actin assembly preceding scission. We found that the endocytic
coat protein Sla1 assembles in a ring shape independent of membrane curvature [1]. Furthermore, we
imaged endocytic proteins, which are involved in the regulation of actin polymerization at endocytic
sites, including actin nucleators, nucleation promoting factors and crosslinkers. Here, we employed dual‐
color superresolution microscopy to directly visualize the structural relation between polymerized actin
and actin interacting proteins at individual endocytic sites with highest spatial resolution. By imaging a
high number of sites and determining their respective time points, we attempt to describe the dynamic
organization of the endocytic actin structures and ultimately mechanistically understand how actin
polymerization is regulated in situ.
Extending the approach by dual‐color imaging of more components, we are pursuing to obtain a
comprehensive structural picture of the endocytic machinery in yeast.
[1] Picco, A. et al. Visualizing the functional architecture of the endocytic machinery. Elife 4, (2015).
Direct probing of dynamic phosphoinositide switches during clathrin‐mediated endocytosis.
K. He1,2, E. Song1,2, M. Ma1,2, R. Gaudin1,2, T. Kirchhausen1,2; 1Program in Cellular and Molecular Medicine,
Boston Childrens Hospital, Boston, MA, 2Department of cell biology, Harvard Medical School, Boston,
Regulatory roles have been postulated for the phosphoinositides PI(4,5)P2 and PI(3,4)P2 during the
formation phase of endocytic clathrin coated pits. These lipids are believed to function as effectors to
mediate capture of components of the clathrin coat. A role during the uncoating phase of coated
vesicles has also been suggested for PI(4)P, in this case acting as the effector mediating the acute
recruitment of auxilin. These roles have been inferred either from in vitro reconstitution studies or by
analyzing the consequence of long‐term genetic perturbations on phosphoinositides kinases or
phosphatases associated with the endocytic pathway.
In an attempt to study more directly the phosphoinositide composition of coated pits and coated
vesicles, we describe here a new generation of phosphoinositide‐specific probes that selectively read in
real time the lipid composition and their dynamic changes during all formation stages of clathrin coated
pit and coated vesicles.
The eGFP‐fluorescently tagged probes are based on protein chimeras containing different domains with
binding specificity for defined phosphoinositides fused to the clathrin‐binding region of auxilin. By
themselves, each of these domains were unable to associate with clathrin coated pits; in contrast,
defined chimeras were recruited to forming pits, with recruitment specificity and dynamics uniquely
linked to the phosphoinositide‐specificity of the lipid binding domain. From the binding patterns we
uncovered that (1) PI(4,5)P2 and PI(3,5)P2 are present during all stages of coated pit formation; (2) a
small amount of PI(4)P is also present during the formation stage of the pits followed by a significant
burst in coated vesicles immediately after coated pit budding; (3) appearance of a PI(3)P burst in coated
vesicles was also detected after coated pit budding and distinct from the PI(3)P pool found in the early
endosomal compartment; (4) steady‐state presence of PI(3,4)P2, first appearing in coated vesicles and
remaining until arrival of the uncoated vesicle to the Rab5‐positive early endocytic compartment; (5) the
PI(4)P burst in the newly formed coated vesicle reflected conversion of PI(4,5)P2 into PI(4)P whereas
appearance of PI(3,4)P2 resulted from the conversion of PI(4)P into PI(3,4)P2. And finally, (5) we were
unable to detect association of PI(5)P or PI(3,4,5)P2 with clathrin‐coated pits and coated vesicles.
Membrane curvature regulates the biogenesis of COPII coated transport carriers.
M. Hanna1, I. Mela2, J. Edwardson2, A. Audhya1; 1Biomolecular Chemistry, University of Wisconsin‐
Madison, Madison, WI, 2Department of Pharmacology, University of Cambridge, Cambridge, United
The majority of biosynthetic secretory proteins initiate their journey through the endomembrane
system from specific subdomains of the endoplasmic reticulum (ER). At these locations, coated
transport carriers are generated, with the Sar1 GTPase playing a critical role in membrane bending,
recruitment of coat components, and ultimately membrane scission. How these events are
appropriately coordinated remains poorly understood. Here, we demonstrate that Sar1 acts as the
curvature sensing component of the COPII coat complex and highlight the ability of Sar1 to bind more
avidly to membranes of high curvature. Additionally, using an atomic force microscopy‐based approach,
we further show that the intrinsic GTPase activity of Sar1 is necessary for remodeling individual lipid
bilayers. Our results also indicate that Sar1 GTPase activity is stimulated by membranes that exhibit
elevated curvature, potentially enabling Sar1 membrane scission activity to be spatially restricted to
highly bent membranes that are characteristic of a bud neck. Taken together, our data support a
stepwise model in which the amino‐terminal amphipathic helix of GTP‐bound Sar1 stably penetrates the
ER membrane, promoting local membrane deformation. As membrane bending increases, Sar1
membrane binding is elevated, ultimately culminating in GTP hydrolysis, which may destabilize the
bilayer sufficiently to facilitate membrane fission.
Intrinsically Disordered Proteins Drive Membrane Curvature and Modulate the Cargo Content
of Coated Vesicles.
D.J. Busch1, J.R. Houser1, C.C. Hayden1, M.B. Sherman2, E.M. Lafer3, J.C. Stachowiak1,4; 1Department of
Biomedical Engineering, University of Texas at Austin, Austin, TX, 2Department of Biochemistry and
Molecular Biology, University of Texas Medical Branch, Galveston, TX, 3Department of Biochemistry and
Center for Biomedical Neuroscience, University of Texas Health Science Center at San Antonio, San
Antonio, TX, 4Institute for Cellular and Molecular Biology, University of Texas at Austin, Austin, TX
The assembly of highly curved membrane structures, such as coated vesicles, is essential to cellular
physiology. The prevailing view has been that structured protein motifs such as wedge‐like amphipathic
helices and crescent‐shaped BAR domains drive membrane curvature. However, many proteins that
contain these structural motifs also contain large intrinsically disordered domains of 300‐1500 amino
acids, including most clathrin adaptor proteins and COPII coat components. While these domains have
long been regarded simply as binding motifs, we have recently reported the surprising result that
disordered domains, including the C‐terminal domains of the endocytic adaptor proteins Epsin1 and
AP180 are highly potent drivers of membrane curvature. Specifically, disordered domains bend
membranes with greater efficiency than the ENTH domain of Epsin1, which contains a well‐
characterized amphipathic helix domain (Busch et al., Nat Comms 2015). How can molecules without a
defined structure drive membrane bending? We utilize in vitro measurements of membrane curvature
and protein diffusivity to demonstrate that disordered domains occupy a significantly larger area on
membrane surfaces compared to structured domains, making them efficient drivers of membrane
curvature through the recently discovered mechanism of protein crowding (Stachowiak et al., NCB 2012,
Copic et al., Science 2012). From this perspective, if disordered domains drive curvature on the coat side
of a membrane, then presenting them on the cargo side would resist internalization through a balance
of pressures on the two surfaces of the membrane. Consistent with this model, when we expressed
disordered domains as cargo molecules on the surface of mammalian cells, steric pressure from the
crowding of bulky cargo excluded them from clathrin‐coated pits, resulting in their accumulation at the
plasma membrane. Interestingly, these findings imply that any receptor could be retained at the plasma
membrane through extracellular binding of ligands or antibodies conjugated to large, disordered
polymers such as PEG chains. The ability to control the accumulation of key receptors at the plasma
membrane could provide a tool for addressing specific clinical disorders. For example, insulin receptors
in type 2 diabetics, and mutant variants of Cystic Fibrosis Transmembrane Conductance Regulator
(D565G) in CF patients are functional at the molecular level, yet do not accumulate at the plasma
membrane in sufficient densities to support healthy levels of downstream signaling. Our latest work
uses bulky engineered ligands to precisely regulate the plasma membrane levels of therapeutically
relevant receptors, creating a new, physical approach to address diverse diseases by controlling
membrane traffic.
Subcellular recruitment of clathrin‐mediated endocytosis machinery in genome‐edited cells to
sites of nanostructure‐induced membrane curvature.
W. Zhao1, L.A. Hanson2, P. Chowdary2, J.R. Marks3, S. Hong3, D.G. Drubin3, Y. Cui1,4, B. Cui2; 1Department
of Materials Science and Engineering, Stanford University, Stanford , CA, 2Department of Chemistry,
Stanford University, Stanford , CA, 3Department of Molecular Cell Biology, University of California,
Berkeley, Berkeley, CA, 4Stanford Institute for Materials and Energy Sciences, SLAC National Accelerator
Laboratory, Menlo Park, CA
Endocytosis is a nanoscale plasma membrane (PM) invagination process, which requires temporally
coordinated recruitment and function of numerous proteins to mediate membrane curvature
progression. Protein‐based curvature generation and/or stabilization have been proposed to be coupled
to biochemical reactions during clathrin‐mediated endocytosis. However, whether or how pre‐existing
membrane curvature affects protein recruitment and dynamics in live cells is difficult to study, due to a
lack of suitable technologies. Here, we used substrate‐attached nanostructures to induce stable
membrane curvature in live cells, and studied how nanoscale curvature affected the recruitment and
dynamics of endocytic machinery on the PM. Patterned nanopillar arrays with controllable diameter
from 200nm to 1000nm were fabricated by electron‐beam lithography. They were found to induce local
PM curvature in living cells with membrane diffusion of CAAX‐tagged green fluorescent proteins
unchanged. Clathrin‐mediated endocytosis proteins, i.e., clathrin and dynamin, fluorescently tagged and
expressed at endogenous levels by genome editing, accumulated at the sites of nanopillar‐induced
membrane curvature in a spatially and temporally specified fashion, favoring high PM curvature on
diameter below 500nm. Additionally, the impact of membrane curvature on endocytic protein
subcellular distribution was successfully demonstrated by varying the shape of the nanostructures, thus
modulating membrane curvature. This work introduces a new tool for the study of cell processes at a
nano‐bio interface, and demonstrates the use of nanostructures for manipulating membrane curvature
in live cells to decipher the influence of PM curvature on endocytic dynamics.
Autoregulation and membrane composition coordinate the membrane remodeling and actin
assembly activities of the F‐BAR/SH3 protein Nervous Wreck.
C.F. Kelley1, E. Messelaar1, T. Stanishneva‐Konovolova2, T.L. Eskin1, S.A. Wasserman1, O. Sokolova2, A.A.
Rodal1; 1Molecular and Cell Biology, Brandeis University, Waltham, MA, 2Bioengineering, Moscow State
University, Moscow, Russia
Eukaryotic membranes are shaped and remodeled by a host of membrane‐binding proteins, which must
be highly regulated to enable the rapid dynamics of intracellular traffic. Nervous Wreck (NWK) encodes
an F‐BAR/SH3 protein that regulates the traffic and signaling output of synaptic growth receptors at the
Drosophila neuromuscular junction (NMJ) and induces membrane deformation in vitro. The dogma in
the membrane‐remodeling field has been that SH3 domain‐mediated autoregulation of crescent‐shaped
F‐BAR proteins serves as an on‐off switch for membrane binding and deformation. Surprisingly, we
found that membrane binding and deformation are not directly coupled. Instead, the isolated Nwk F‐
BAR domain only efficiently assembles into higher‐order structures and remodels membranes within a
limited range of negative membrane charge, and autoregulation serves to elevate this range. Using
single particle electron microscopy, we confirmed that the orientation of the F‐BAR domain on the
membrane is dependent on membrane composition. We also show that the SH3 domains of Nervous
Wreck undergo major conformational changes between membrane‐bound and non‐membrane‐bound
structures. Finally, in addition to controlling membrane binding, SH3‐mediated autoregulation inhibits
Nwk activation of WASp/Arp2/3‐dependent actin filament assembly. We demonstrate that
autoregulation is critical to constrain and regulate membrane binding and WASp activation by Nwk in its
normal vivo context. Taken together, our results suggest a novel mechanism by which autoregulation
and lipid composition may act in tandem to target actin polymerization to sites of membrane
deformation in order to coordinate membrane trafficking events at the synapse.
Regulation of the ESCRT‐III membrane scission machine by a ubiquitin hydrolase.
N.K. Johnson1, M. West1, G. Odorizzi1; 1Molecular, Cellular, and Developmental Biology, University of
Colorado, Boulder, CO
Dynamic assembly and disassembly of Endosomal Sorting Complex Required for Transport‐III (ESCRT‐III)
drives membrane scission reactions at endosomes and the plasma membrane, but regulation of this
activity is poorly understood. The objective of this study was to identify ESCRT‐III interactions that
influence both the polymerization state of the complex and membrane scission activity by the complex.
To measure ESCRT‐III assembly/disassembly in vivo, we used rate‐zonal density gradients of detergent
solubilized membranes to obtain the distribution of monomeric to polymerized ESCRT‐III present in the
cell at steady state. To assay membrane scission activity, endosome morphology was visualized by
electron microscopy and 3D tomography. We show that ESCRT‐III in yeast is regulated by Doa4, the
ubiquitin hydrolase that deubiquitinates transmembrane proteins being sorted by ESCRT‐III into
intralumenal vesicles (ILVs) at endosomes. Deletion of Doa4 destabilizes the ESCRT‐III complex and
results in smaller ILVs. Conversely, Doa4 overexpression increases ESCRT‐III stability, and this activity is
inhibited through its interaction with Vps20, the subunit of ESCRT‐III that initiates complex assembly.
These results suggest that the timing of ILV membrane scission by ESCRT‐III is coordinated with the
removal of ubiquitin from transmembrane protein cargoes sorted into nascent ILVs.
Minisymposium 06: Molecular Motors and the Cytoskeleton: Measurement,
Manipulation, and Mechanics
Engineering cytoskeletal motors.
Z. Bryant1,2; 1Bioengineering, Stanford University, Stanford, CA, 2Structural Biology, Stanford University
School of Medicine, Stanford, CA
Engineering molecular motors can provide direct tests of structure‐function relationships and potential
tools for controlling cellular processes. I will describe the design and characterization of a panel of
cytoskeletal motors that reversibly change gears — speed up, slow down, or switch directions — when
exposed to exogenous signals including blue light. Genetically encoded light‐responsive motors will
expand the optogenetics toolkit, complementing precise perturbations of ion channels and intracellular
signaling with spatiotemporal control of cytoskeletal transport and contractility.
Force Generation by Membrane Associated Myosin‐IC.
S. Pyrpassopoulos1, G. Arpağ2, E. Feeser1, H. Shuman1, E. Tuzel2, E.M. Ostap1; 1The Pennsylvania Muscle
Institute and Department of Physiology, Perelman School of Medicine at the University of Pennsylvania,
Philadelphia, PA, 2Department of Physics, Worcester Polytechnic Institute, Worcester, MA
Vertebrate myosin‐IC (Myo1c) is a class‐1 myosin that links cell membranes to the underlying actin
cytoskeleton. Actin binding occurs via the motor domain, while the tail domain interacts with
phosphatidylinositol 4,5‐bisphosphate (PtdIns(4,5)P2 ) through a pleckstrin homology domain. We have
previously shown that Myo1c is able to propel actin filaments while attached to a fluid supported‐lipid‐
bilayer containing PtdIns(4,5)P2 in an in vitro gliding assay. However, Myo1c in this ensemble assay
undergoes its working stroke under very low‐loads, and it is of interest to measure the ability of the
motor to develop and sustain force while bound to a fluid bilayer. Therefore we used optical tweezers in
the three‐bead assay configuration to measure at the single molecule level the dissipation of load for
membrane bound Myo1c. We engineered spherical supported bilayers as pedestal beads with
membranes consisting of 98% DOPC and 2% PtdIns(4,5)P2 . Membrane fluidity of the bilayers was
confirmed by fluorescence recovery after photobleaching of labeled lipids. From force dissipation
measurements, we found that the diffusion coefficient of single membrane bound Myo1c molecules is D
= 0.17 ± 0.11 um2 /s. Similar values were obtained by TIRF microscopy assays in the absence of load.
Interestingly at high concentrations of Myo1c we found that ensembles of motors dynamically
interacting with lipid membranes and 1 um of actin filament at saturating [ATP] (1 mM) can develop and
sustain forces of ~ 1 pN. A computational approach using a simple 1D model and the kinetic parameters
for membrane attachment and actin mediated ATP hydrolysis from previous studies recapitulated the
basic features of our ensemble data. More importantly our computational model predicts that if
diffusion barriers are introduced to only 5% of the motors that simultaneously interact with the actin
and membrane, the force developed almost quadruples. We also showed that the minimal physiological
and dynamic interaction between Myo1c and PtdIns(4,5)P2 is sufficient for the transport of lipid cargoes
along actin filaments immobilized on a glass coverslip.
Shape Remodelling and Blebbing of Active Cytoskeletal Vesicles.
E. Loiseau1, J.A. Schneider2, F.C. Keber1, G. Salbreux2, A.R. Bausch1; 1Physics Departement, Technische
Universität München, Munich, Germany, 2Lincoln's Inn Fields Laboratories, The Francis Crick Institute,
London, UK
Morphological transformations of lipid membranes such as shape adaptation to external stimuli,
blebbing, invagination or tethering result from an intricate interplay of an active tension generating
shear elastic cytoskeleton with a fluid lipid membrane. Cellular complexity defies a clear identification of
the competing processes which lead to such a rich phenomenology, therefore a model system starting
from few building blocks is of avail. Here we present a minimal in vitro model system which
demonstrates that tension generation of an encapsulated active acto‐myosin network suffices for a
remodeling and global shape transformation of cell‐sized lipid vesicles. Our in vitro model system uses
purified cytoskeletal elements inside vesicles, in which coupling to the membrane, elasticity of the
cytoskeletal network, and contractile activity can all be precisely tuned. The observed polymorphism of
membrane deformation can be rationalized by an analytical model taking into account the membrane
tension, the anchoring between the membrane and the actin network, and the forces exerted by
molecular motors. The identification of the physical mechanisms for shape transformations, sets a
conceptual and quantitative benchmark for the further exploration of the adaptation mechanisms of
On the force‐generating capacity of disassembling microtubules.
C.L. Asbury1, J.W. Driver1, E. Geyer2, L.M. Rice2; 1Physiology Biophysics, University of Washington,
Seattle, WA, 2Biophysics, University of Texas Southwestern, Dallas, TX
Microtubules can generate force independently of motor enzymes, especially at kinetochores where
disassembling microtubules drive movement of mitotic chromosomes. A popular explanation for this
microtubule‐powered motility is the conformational wave model, where individual rows of tubulin
subunits, the protofilaments, push or pull on a kinetochore as they curl outward from a disassembling
microtubule tip. One pioneering study has demonstrated that curling protofilaments are capable of
generating force, but the measured forces were far weaker (~0.24 pN) than forces generated at
kinetochores in vivo (4 – 8 pN) or in vitro (~9 pN). Thus it has remained unclear whether the wave
mechanism can contribute significantly to kinetochore force production. Here, by developing a wave
assay using recombinant tubulin and a feedback‐controlled laser trap, we show that the conformational
wave can generate forces 50‐fold higher than previously recorded. Measuring its full capacity for work
output reveals that the wave carries at least three times the energy harnessed by kinetochores in vitro,
and enables an estimate of the mechanical strain energy trapped per tubulin dimer in the microtubule
lattice. Surprisingly, a β‐tubulin mutation that allosterically enhances microtubule stability has little
effect on wave energy. Our work indicates that the conformational wave mechanism can make a major
contribution to kinetochore motility and it provides a new, direct way to examine tubulin
Microtubules self‐repair in response to mechanical stress.
L. Schaedel1, K. John2, J. Gaillard1, M.V. Nachury3, L. Blanchoin1, M. Théry1,4; 1iRTSV, CEA, Grenoble,
France, 2Laboratoire Interdisciplinaire de Physique, CNRS, Grenoble, France, 3Department of Molecular
and Cellular Physiology, Stanford University School of Medicine, San Francisco, CA, 4IUH, INSERM, Paris,
Microtubules ‐ which define the shape of axons, cilia and flagella, and provide tracks for intracellular
transport ‐ can be highly bent by intracellular forces, and microtubule structure and stiffness are
thought to be affected by physical constraints. Yet how microtubules tolerate the vast forces exerted on
them remains unknown. Here, by using a microfluidic device, we show that microtubule stiffness
decreases incrementally with each cycle of bending and release. Similar to other cases of material
fatigue, the concentration of mechanical stresses on pre‐existing defects in the microtubule lattice is
responsible for the generation of larger damages, which further decrease microtubule stiffness.
Strikingly, damaged microtubules were able to incorporate new tubulin dimers into their lattice and
recover their initial stiffness. Our findings demonstrate that microtubules are ductile materials with self‐
healing properties, that their dynamics does not exclusively occur at their ends, and that their lattice
plasticity enables the microtubules' adaptation to mechanical stresses.
Developing a Method for Mathematical Computation of Three‐Dimensional EB1‐GFP Motion
Visualized by Lattice Light‐Sheet Microscopy.
N. Yamashita1, M. Morita1, W.R. Legant2, B. Chen2,3, E. Betzig2, H. Yokota1, Y. Mimori‐Kiyosue4; 1Image
Processing Research Team, RIKEN Center for Advanced Photonics, RIKEN, Wako, Saitama, Japan, 2Janelia
Research Campus, Howard Hughes Medical Institute, Ashburn, VA, 3Research Center for Applied
Sciences, Academia Sinica, Taipei, Taiwan, 4Cellular Dynamics Analysis Unit, RIKEN Center for Life Science
Technologies, RIKEN, Kobe, Hyogo, Japan
In this study, we adapted mathematical computing and geometric representation technique to analyze
spatial variations in microtubule growth dynamics within the mitotic spindle apparatus observed by 3D
live imaging using lattice light‐sheet microscopy (LLSM). Imaging whole cells with high resolution at sub‐
second intervals is a key technique for observing the microtubule growth dynamics. Recently, we
succeeded in the 3D tracking of microtubule growth dynamics throughout the cellular space using LLSM
in conjunction with microtubule growth marker protein end‐binding 1 fused to GFP (EB1‐GFP). This
microscopic technology enables 3D live imaging with high spatiotemporal resolution. The lattice light
sheet is generated from a massive parallel array of nondiffracting light beams that mutually interfere to
create an ultrathin light sheet extending over cellular dimensions. This strategy allowed us to image
whole cells for hundreds of volumes at subsecond intervals. Automated computational strategies for the
observed data are also required to extract biologically significant information in an efficient and
reproducible manner because LLSM generates thousands of numerical data sets containing far more
information than can be practically analyzed using conventional approaches. In this study, we developed
geometric representation using mathematical computing to analyze the microtubule growth dynamics.
EB1‐GFP comets were automatically tracked and their dynamic and spatial properties were further
analyzed by developing a pipeline implemented in MATLAB software. Grouping and separately
displaying the trajectories by their mean travel speeds highlighted the trend for trajectories with fast or
slow migration speeds to be associated with particular regions of the cell. Fast trajectories often
originated near the centrosomes, while slowly migrating ones tended to accumulate at inter‐
centrosome region, considering the slowly moving EB1‐GFP comet traveling along microtubules
associating with kinetochores. We also arrange the trajectories according to their subcellular location
and their angle against the centrosomal axis. In anaphase and telophase cells, a sub set of trajectories
started growing at an area between the centrosome and the midzone and these have growth angle that
is unlikely to have initiated from that centrosome, suggesting that these are noncentrosomal
microtubules. Our computational analysis framework combined with LLSM will enable spatiotemporal
information regarding the movements of fluorescent proteins within living specimens to be analyzed
with previously unobtainable precision. We therefore believe that this method of advanced imaging
analysis will prove to be an invaluable research tool for a wide range of scientific fields in the future.
Direct measurement of the binding rate constant of kinesin to microtubules in living cells.
T. Kambara1, Y. Okada1; 1Quantitative Biology Center, Riken, Suita, Japan
It has been established that conventional kinesin (kinesin‐1, KIF5 in mammalian cells) selectively moves
along a specific subset of microtubules in living cells. For example, KIF5 is specifically recruited to the
microtubules in the axon initial segment in neurons, which would enable efficient transport into the
axon. It has been proposed that kinesin recognizes cues on microtubules. However, the mechanism of
this selective binding is still controversial. Some groups have proposed that acetylation or other post
translational modifications of tubulin serve as the cue for selective binding, while we are proposing that
conformational differences between the GTP‐form and GDP‐form of microtubules provide the cue. To
understand the mechanism of the selective binding, it would be important to examine whether kinesin
binding to specific subsets of microtubules is enhanced, inhibited or both. Here, we directly measured
the binding rate constant of kinesin to microtubules in living cells using single molecule fluorescence
microscopy. To our surprise, there are three populations of microtubules. Some microtubules bound to
KIF5 with the binding rate constant similar to the value in vitro. Some other microtubules showed about
3 times higher binding rate constant, some minor population of microtubules showed nearly ten times
higher rate constant than in vitro. The velocity and the run length were same among these microtubules
and with the in vitro values. These data suggest the existence of at least two different mechanisms exist
as the guidance cues for KIF5 by accelerating its binding to some specific subset of microtubules.
O‐Myo! A ring‐shaped myosin gliding assay for characterizing the lifetime of myosin motors.
R.F. Hariadi1, A. Appukutty2, S. Sivaramakrishnan3; 1Wyss Institute for Biologically Inspired Engineering,
Harvard University, Boston, MA, 2Department of Biomedical Engineering, University of Michigan, Ann
Arbor, MI, 3Department of Genetics, Cell Biology, and Development, University of Minnesota,
Minneapolis, MN
Nature has evolved molecular machines that are critical in cellular processes occurring over broad
timescales, ranging from seconds to days. As an example, myosin in the human heart take ~10 steps per
second and are collectively responsible for ~105 heart beats per day. Analogous to the engineering of
macro‐scale machines, the evolution of molecular machines is likely to be constrained by the
requirements of high performance and a long lifetime. Here, we developed a ring‐shaped gliding assay
(O‐Myo) that utilizes engineered micron‐scale DNA nanotube rings with precise arrangements of dimeric
myosin VI, a model system for myosin motor. The O‐Myo gliding assay platform allows the same
individual actin filament to glide over the same myosin ensemble (50–1000 motors per ring) multiple
times, once per revolution. Actin filaments glide along the nanotube rings with high processivity for up
to 5 revolutions (40 µm total run length; 11 minutes), consistent with interactions between a single actin
filament and multiple myosin motors. To benchmark our assay, we show actin gliding speed is robust to
the variation in motor number, which is similar to the observations from classic gliding assays on non‐
processive motor‐coated surfaces and from experiments utilizing non‐ring myosin VI‐labeled DNA
nanotubes. Surprisingly, actin gliding speed is also independent of ring curvature within our sample
space (ring diameter of 1‐5 µm). This result can be explained by the interplay between the force
required to bend an actin filament and the mechanical load experienced by myosin motors. Our circular
gliding assay may provide a closed‐loop platform for characterizing the robustness of broad classes of
molecular motors, all the while allowing for a more in depth study of friction and wear in a biological
Structural insights into cytokinesis: super‐resolution imaging of cytokinesis nodes and
contractile rings in live fission yeast.
C. Laplante1, F. Huang2, I.R. Tebbs1, J. Bewersdorf2,3, T.D. Pollard1,2,4; 1MCDB, Yale University, New Haven,
CT, 2Cell Biology, Yale University, New Haven, CT, 3Biomedical Engineering, Yale University, New Haven,
CT, 4Molecular Biophysics and Biochemistry, Yale University, New Haven, CT
Studies of fission yeast have contributed much to our understanding of cytokinesis, but progress has
been limited by a lack of information about nodes, small organizing centers for the contractile ring. We
used ultra‐fast, super‐resolution fluorescence nanoscopy of live fission yeast cells to localize the
distributions of photoconvertible mEos3.2 fused to the N‐ or C‐ termini of six constituent node proteins
at ~35 nm resolution (σloc ~ 14 nm). We used the stochastic blinking nature of the fluorescent marker
mEos3.2 to calculate the positions and relative numbers of each protein. Assuming radial symmetry,
statistical analysis of the localizations of mEos3.2 established that each node marker occupies a distinct
zone within the node. The C‐terminus of anillin Mid1p is located centrally in nodes during contractile
ring formation with 75% of localizations at radii <35 nm. The C‐terminus of formin Cdc12p and the N‐
terminus of F‐BAR Cdc15p are at radii <40 nm, while the C‐terminus of Cdc15p is at 55 nm. The N‐
terminus of IQGAP Rng2p and the C‐terminus of myosin‐II Myo2p are at radii <50 nm, while the N‐
terminus of Myo2p and the C‐terminus of its regulatory light chain Rlc1p are at radii <67 nm. Expression
of pairs of proteins labeled with mEos3.2 showed that Mid1p, Cdc15p, Myo2p and Rlc1p colocalize to
the same punctate structures. The ratio of localizations in nodes of forming contractile rings was 1
Mid1p to 2 Cdc15p, Rng2p, Myo2p and Rlc1p and <1 Cdc12p. This data together with prior studies of
molecular interactions shows that Mid1p forms the base of the node that together with Rng2p anchors
the end of myosin‐II tail to the plasma membrane with myosin heads extending into the cytoplasm. Cells
expressing a node marker along with the calponin homology domain mEos3.2‐CHD to mark actin
filaments have bright puncta interconnected by linear structures consistent with a network of actin
filaments connecting the nodes. The organization of Cdc15p, Rng2p, Myo2p, Rlc1p and Cdc12p in
punctate structures persists in contractile rings and these ring nodes move bidirectionally in constricting
rings. Tethering a bouquet of myosins to the plasma membrane explains how their interactions with
actin filaments produce forces to assemble and constrict the contractile ring. The application of ultra‐
fast FPALM to live yeast cells not only provides the first insight into the substructure of nodes and the
cytokinetic ring but also lays the ground work for the broad application of live cell super‐resolution
microscopy relevant to all biological questions.
Education Minisymposium: Teaching How to Teach and Learn
The development and validation of tools to help biology departments navigate from Vision to
S.E. Brownell1, S. Freeman2, M. Wenderoth2, A. Crowe2; 1School of Life Sciences, Arizona State
University, Tempe, AZ, 2Biology, University of Washington, Seattle, WA
Biology as a discipline has expanded dramatically, and as a community, we struggle to choose what is
most important to teach. The Vision and Change report outlined broad national consensus for five core
concepts that should be emphasized in undergraduate biology (Brewer and Smith 2011). Using a
grassroots approach of soliciting input from faculty at a diverse range of institutions nationally, we have
incorporated feedback of 244 faculty members to create the BioCore Guide. The BioCore Guide lays out
a set of general principles and specific concept statements that elaborate on each of the five core
concepts for the three major sub‐disciplines of biology: molecular/cellular biology, physiology, and
ecology/evolution. Through this process we achieved strong national agreement with over 91% of
responders being in agreement with the scientific accuracy of the statements and over 93% of
responders agreeing that the concept statements were relevant for a general biology curriculum. We
have now designed a general biology program‐level assessment called BioMaps that is aligned with the
BioCore Guide. Biology departments can use this test to monitor the progression of students through a
general biology curriculum. The test consists of a series of multiple true/false questions focused on 35
scenarios. The questions have been face‐validated through over 100 student think aloud interviews,
feedback from 20 expert biologists, and a large‐scale pilot involving over 3000 students at 10 diverse
institutions. We will present on the overall validation process and results of our pilot study. Overall, we
recommend that the BioCore Guide and the BioMaps assessment be implemented as tools for biology
departments to better align their curriculum with the goals of Vision and Change.
Converting a lecture‐based introductory biology class to an active learning studio environment.
J.N. Schoonmaker1, C.J. Ramey1; 1Chemical and Biological Engineering, Colorado School of Mines,
Golden, CO
We recently transformed a traditional introductory biology course into an active learning experience
that resonates with engineering students. Our course re‐design involved: (1) a renovated classroom to
create a studio environment with wet‐lab capability; (2) a stream‐lined curriculum with measurable
learning outcomes designed to explore the fundamental concepts of biology through active learning
strategies and open inquiry labs; and (3) a training program to develop graduate teaching assistants who
support a student‐centered learning experience. Our classroom accommodates twenty‐one groups of
three. Students are seated around a cantilevered island, maximizing interactions. Workstation
computers allow internet access and connectivity to data acquisition systems. Dual monitors stacked
vertically give students a view of their local computer as well as the signal from the instructor podium.
To support whole‐class instruction, audio‐visual technology transmits a variety of signals from an
instructor podium to all twenty‐one workstations. The classroom design meets the traditional needs of a
biology lab, including access to sinks, use of compound microscopes, data acquisition, gel
electrophoresis and thermal cyclers. The flipped course structure (reading assignment plus online
homework) is supported by group discussions. Misconceptions are exposed during a short quiz and
clarified during follow‐up peer instruction. The remainder of class time is spent in explorations or inquiry
labs. Whiteboards facilitate discussion as students solve problems, create concept maps, plan
experiments and interpret experimental data. Initial impacts of our studio environment indicate
increased student success as measured by a decrease in students withdrawing or not passing the course
(Lecture: 10.93% +/‐ 0.49% (n= 484 over 3 semesters) compared to Studio: 4.41% +/‐ 1.5% (n= 338 over
3 semesters) and increased normalized learning gains as measured by a pre/post basic biology concept
inventory (Lecture: 0.42+/‐ 0.02, n=398 (Fall 2012); Studio: 0.70 +/‐ 0.06, n=248 (Fall 2013) and 0.82 +/‐
0.03, n=47 (Spring 2014)). Student survey responses indicated an overall positive attitude toward
learning in the new studio environment. Students felt they had: (1) learned about the fundamentals of
biology (83% positive, 11% neutral, 4% negative); (2) learned a lot about experimental design (85%
positive, 13% neutral, 1% negative) and (3) improved their analytical thinking skills (70% positive, 21%
neutral, 10% negative). This new learning space supports a constructivist approach to learning, moving
conversations past rote repetition of textbook material to evaluation and synthesis of ideas.
Teaching Computational Approaches for Life Scientists.
A. Rubinstein1; 1Computer Science, Tel Aviv University, Tel Aviv, Israel
Modern biology has been undergoing a revolution in recent decades. One consequence of this
revolution is that computational methods are increasingly being used to help solve problems in
molecular biology and genetics. Still, life sciences curricula have been only slightly affected by this
revolution. In some universities, life sciences students are required to take an introductory programming
course, possibly followed by a course on bioinformatics tools. While these courses usually promote
some computational thinking, they often focus on programming skills and tool handling, rather than on
abstract, algorithmic thinking, and basic notions in computer science relevant for biology. Numerous
articles have been published in recent years, calling for the incorporation of "computational thinking"
into the life sciences curricula and labs ([1], [2])
We designed a new course – "Computational Approaches for Life Scientists" (website:
http://ca4ls.wikidot.com), which aims to promote the computational "way of thinking", beyond the level
of programming and tool handling. The course is a non‐introductory course designed specifically for life
scientists. It is taught in a less technical manner than in "pure" computer science courses, and avoids
unnecessary formalism. It assumes prior basic programming skills (e.g. from high school). The focus is on
abstract and algorithmic thinking, fundamental computing concepts and ideas, and basic computational
approaches to biological problem solving. An additional goal is to expose students to discrete
mathematics notions, a branch of mathematics underrepresented in life‐sciences curricula. The course
was taught 3 times so far. At the first 2 weeks students are acquainted with the programming language
Python, later used for demonstrating course topics and for developing hands‐on skills. Course topics
include biological image processing, network notions and algorithms, basic sequence analysis concepts,
and discrete approaches to simulation of dynamic biological systems.
In this talk I will present the challenges of computational education for life scientists, overview some
recent initiative (e.g. [3], [4]), and describe our own course and its evaluation ([5]).
References: 1. Bialek W, Botstein D (2004) Introductory science and mathematics education for 21st‐
century biologists. Science Signaling 2. Pevzner P, Shamir R (2009) Computing has changed biology‐‐
biology education must catch up. Science 3.
Libeskind‐Hadas R, Bush E (2013) A first course in
computing with applications to biology. Briefings in bioinformatics. 4. Pevzner P, Shamir R (2011)
Bioinformatics for biologists. Cambridge University Press. 5.
Rubinstein A, Chor B (2014)
Computational Thinking in Life Science Education. PLoS Comput Biol
MACH: A Model for Explaining Molecular and Cellular Mechanisms Across the Life Science
C.M. Trujillo1,2, T. Anderson3, N. Pelaez2; 1Plant Biology, Michigan State University, East Lansing, MI,
Biological Sciences, Purdue University, West Lafayette, IN, 3Chemistry, Purdue University, West
Lafayette, IN
Life scientists use mechanistic explanations to understand behaviors of the immense complexity of
molecular and cellular systems. In undergraduate biology courses, students face many challenges when
learning about molecular and cellular mechanisms. To identify and model the key components, we
reviewed the literature to inform an initial model, followed by thematic analysis of interviews collected
from biologists of different sub‐disciplines to validate the model. The results revealed four themes
present in the explanations: research Methods (M), Analogies (A), Contexts (C), and How (H) the
mechanism works. These themes formed the components of the MACH model. Once developed, MACH
was transitioned into the undergraduate classroom to explore its usefulness as a lens to understand
students’ written explanations and as an instructional tool for a range of courses. Analysis of essays and
interviews suggests that MACH may act as a metacognitive tool by helping students to monitor their
understanding, communicate explanations, and identify explanatory gaps. The presented MACH model
provides a promising framework for researching, teaching, and explaining molecular and cellular
Barriers to engagement with the primary literature and the journey from a novice to a
competent reader.
C. Abdullah1, R. Lie2, E. Tour3; 1Biomedical Sciences Program, UCSD, La Jolla, CA, 2Department of
Neurosciences, UCSD, La Jolla, CA, 3Cell and Developmental Biology, UCSD, La Jolla, CA
Primary literature offers rich opportunities to teach students how to “think like a scientist.” Calls for
increased incorporation of original scientific literature into biology education have been issued by a
variety of educational organizations (e.g., “Vision and Change”, AAAS, 2011). Understanding the barriers
that students face when they approach research articles is needed to create a strong basis for
developing effective teaching methods that utilize primary literature. Here we present the first analysis,
to our knowledge, of what recent biology undergraduates perceive as the most challenging aspects of
engaging with the primary literature. This study was conducted in a Master’s‐level course that offered a
structured analysis of four recent papers from diverse fields of biology, including one flawed paper, one
exemplary paper, and two conflicting papers. We analyzed 69 pairs of pre‐ and post‐course free
responses to the question: “What aspects of reading and analyzing primary literature do you find most
challenging?” The analysis was conducted by three raters who were blind to both the identity of the
students and the pre‐/post‐ status of the response. We describe six over‐arching categories of
difficulties. Before instruction, “Unfamiliar techniques” was the most frequently identified challenge,
while after instruction “Conclusions” was identified as such. We also detected a shift in the cognitive
level of the identified difficulties: while the challenges aligned the Lower Order Cognitive Skills (LOCS,
Bloom et al., 1956) dominated the responses prior to the instruction, after instruction their frequency
declined. At the same time, the frequency of challenges aligned with Evaluation (a Higher Order
Cognitive Skill) increased two‐fold. We place these changes within the context of the Model of Domain
Learning that describes the journey toward acquisition of expertise in an academic domain (Alexander,
2003) and suggest that these changes are consistent with the transition from a novice to a competent
reader. We discuss immediate implications of our findings and recommendations for instruction that
utilizes scientific papers in graduate and undergraduate classes.
Spreading Vision and Change Through Faculty Mentorship: The ASCB Mentoring in Active
Learning and Teaching (MALT) program.
M.J. Wolyniak1, A.J. Prunuske2, J.J. Adler3, A.J. Crowe4, L.C. Keller5, B.J. Kolber6, B.A. Leland7, S.
Murugesan8, S.M. Schreiner7, Z. Whatley9, S.M. Wick10; 1Biology, Hampden‐Sydney College, Hampden‐
Sydney, VA, 2Biomedical Sciences, University of Minnesota Medical School, Duluth, MN, 3Biology, Brescia
University, Owensboro, KY, 4Biology, University of Washington, Seattle, WA, 5Biological Sciences,
Quinnipiac University, Hamden, CT, 6Biological Sciences, Duquesne University, Pittsburgh, PA, 7Cell
Biology, Yale University School of Medicine, New Haven, CT, 8Cell Biology and Physiology, National
Heart, Lung, and Blood Institute, Bethesda, MD, 9Biology, Gettysburg College, Gettysburg, PA, 10Plant
Biology, University of Minnesota, St. Paul, MN
The life science education community has responded to the recommendations of the American
Association for the Advancement of Science (AAAS) Vision and Change document with several initiatives
designed to improve the way in which undergraduates learn science. These initiatives have often taken
the form of one‐time workshops and institutes that have succeeded in generating awareness of and
interest in active learning techniques among participants. However, they have been less successful with
respect to generating the sustainable change necessary to bring real reform to undergraduate science
education. To create sustainable change, long‐term faculty development initiatives focused on
mentorship are needed so that instructors seasoned in active learning can successfully convey their
experiences to mentees who are interested in using these pedagogical techniques as the centerpiece of
their own teaching. In this spirit, the American Society for Cell Biology (ASCB) Education Committee
created the Mentoring in Active Learning and Teaching (MALT) program to provide a means for
members with an interest in developing improved and sustainable active learning techniques to gain
experience in this style of teaching through close, long‐term interaction with a veteran teaching mentor.
Since the program’s inception in late 2013, MALT has organized over 25 mentor/mentee pairings that
are individually tailored to meet the particular needs of each mentee. While initially conceived strictly
from the perspective of spreading Vision and Change reforms to a larger audience, MALT has also served
as a valuable means for graduate student and postdoctoral level mentees with strong interests in
teaching‐based careers to gain the best practice teaching experiences they need to be competitive in a
tight academic job market. The majority of pairings have been organized with respect to geography;
however, virtual mentorships have also been established in which the mentee meets regularly with his
or her mentor via Skype or phone conversation to discuss teaching or career strategies. To assess
MALT’s effectiveness in spreading active learning techniques, mentees receiving financial support from
the program videotape their classroom interactions before and after mentoring and submit them for
coding via the Classroom Observation Protocol for Undergraduate STEM (COPUS) scoring (Smith et al.,
2013). Initiatives like MALT provide an excellent means to equip current and future faculty with the
proven pedagogical techniques needed to effectively train the next generation of scientists.
Developing an understanding of scientific research helps prepare STEM undergraduates to
E.M. Stone1, A. Baranger2; 1Berkeley Science Math Initiative, University of California, Berkeley, Berkeley,
CA, 2Department of Chemistry, University of California, Berkeley, Berkeley, CA
Laboratory research experiences provide undergraduate students with a deeper understanding of
science practices, including how to test hypotheses, collect and analyze data, create models to explain
results, and use evidence to support conclusions [1]. For science, engineering and mathematics (STEM)
majors who are considering teaching as a career, these experiences equip them with a foundation
needed to guide their own future students to develop an understanding of research practices and
problem‐solving. Cal Teach Berkeley is an interdisciplinary teacher education program that allows STEM
undergraduates to complete their disciplinary degrees while simultaneously earning a single‐subject
teaching credential with a focus on teaching in urban schools. We hypothesize that the integration of
science and math disciplinary coursework with education coursework and apprentice teaching through
multiple field placements in local K‐12 classrooms will produce highly qualified science and math
teachers [2]. As part of their coursework, Cal Teach undergraduates are required to participate in a
scientific research experience and concurrently enroll in a research methods course. We present data
that address the following research questions: How does participation in undergraduate research
influence their understanding of science and math, beliefs about science and math teaching and
learning, and the development of pedagogical content knowledge (PCK) for teaching? How are Cal Teach
Berkeley graduates able to translate their understandings, beliefs and PCK into classroom practices that
promote K‐12 student achievement? We have modified a Knowledge Integration scoring rubric [3] that
effectively measures conceptual understanding of undergraduate research by analyzing open‐ended
responses from pre/post surveys. Additionally, we have developed an instrument for evaluating inquiry‐
based K‐12 lesson plans created and taught by participants. We are also in the process of designing and
testing approaches for evaluating weekly journal reflections about research projects and a culminating
scientific research poster, as well as initiating interviews to assess participants’ perceptions of their
research experiences and ways in which the experience has influenced their K‐12 teaching. We show
that the opportunity to engage in scientific research experiences provides participants with a deeper
understanding of the process of research, experimentation and science practices, as well as strategies
for teaching these concepts and skills to K‐12 students.
Linn, Palmer, Baranger, Gerard & Stone (2015). Science 347, 627
Newton, Poon, Nunes & Stone (2012). Evaluation and Program Planning 36, 88–96.
Stone (2014). CBE‐ Life Sci Educ, 13, 90–101.
Stop Telling Me to do Active Learning and Show Me How: Biology Examples of Active Learning
K. McCoy‐Simandle1, L.B. Jones2, D. Cox1; 1Anatomy and Structural Biology, Albert Einstein College of
Medicine, Bronx, NY, 2Biological Sciences, Lehman College, Bronx, NY
Recent reports in education research consistently demonstrate that incorporating active learning
strategies in a traditional classroom setting leads to numerous benefits including enhanced
understanding of material, fewer failing students as well as higher order thinking skills improvement
(Freeman et al., PNAS 2014, Deslauriers et al., Science 2011). As such, discussion of science education
emphasizes the use of active learning techniques in place of traditional lectures. The problem lies in the
fact that the majority of scientists were taught by a traditional lecture style. While the idea of
modernizing the classroom into a more student‐centric environment enhancing student understanding
and retention appeals to many, most do not know where or how to begin. Here we will discuss the goal
of active learning and strategies for designing and implementing alternatives or supplements to lecture.
Moreover, we discuss and give examples of different active learning techniques (not just clickers)
implemented in several different biology classes such as think‐pair‐share, Jigsaw activities,
brainstorming, team‐based learning, peer‐led team‐based learning, concept mapping etc. These
examples will help the uninitiated or old‐school lecturers by providing ideas on how to transform a
traditional lecture class with a focus on blending active learning into traditional lecture classrooms.
StarCellBio: a new cell and molecular biology experiment simulator.
A.L. Brauneis1, L.M. Aleman1,2, I. Ceraj2, S. Kini2, W. Lisa1, A. Gavrilman2, P. Pinch2, C.A. Kaiser1, G.
Walker1; 1Biology, Massachusetts Institute of Technology, Cambridge, MA, 2Office of Digital Learning,
Massachusetts Institute of Technology, Cambridge, MA
Acquisition of scientific reasoning is crucial for students’ ability to accomplish real‐world scientific tasks,
including designing and conducting scientific investigation. Students struggle with learning and
understanding the experimental design process, which requires authentic research activities. To provide
students with real experimentation opportunities, faculty, research scientists, and software developers
at MIT developed a freely available, web‐based cell and molecular biology experiment simulator called
StarCellBio (http://starcellbio.mit.edu/). StarCellBio enhances student learning of core cell biology
concepts and experimentation by providing students with opportunities to design, perform, and analyze
their own simulated experiments using three experimental techniques: western blotting, flow
cytometry, and microscopy. To develop the simulator, we used an iterative, multi‐faceted design
process incorporating focus groups, prototype testing, user interface and graphic design, and usability
testing. The result is a user‐friendly, educational, and inquiry‐based simulator that introduces research
experiences into cell and molecular biology courses, compensating for the dearth of laboratory
components in upper‐level undergraduate cell and molecular biology courses at MIT and other
institutions. Affective interview and survey data following StarCellBio implementation in MIT’s Cell
Biology course indicate that StarCellBio helps students develop a deeper understanding of experimental
design and analysis, but that students struggle with proper experimental control design, a finding that
we are currently probing in more detail. Once fully disseminated, StarCellBio will support students’
learning of cell biology concepts and experimentation in both residential and online courses around the
E.E. Just Award Lecture
Dissecting the molecular mechanisms of vocal learning and spoken language: a personal
E.D. Jarvis1; 1Department of Neurobiology, Duke University Medical Center HHMI, Durham, NC
In the spirit of being the recipient of the American Society for Cell Biology’s 2015 Ernest Everett Just
Awardee, I will present a lecture on both my science and personal journey. I have long been interested
in how the brain generates complex behaviors, particularly from molecular and cellular perspectives.
The behavior I have chosen to study is spoken‐language, one of the most advanced behaviors humans
have evolved. The most specialized and rare component of spoken‐language is vocal learning, the ability
to imitate sounds and pass on vocal repertoires culturally from one generation to the next. Only several
groups of mammals and several groups of birds, such as songbirds and parrots, have vocal learning. It is
not found in our closest living non‐human primate relatives, such as chimpanzees, nor in the closest
relatives of each of the vocal learning bird lineages. We have found that the convergent behavior and
brain circuitry for vocal learning in birds and humans are associated with convergent specialized
expression of multiple genes in the unique brain regions that control song and speech. Many of these
genes function in brain development and formation of neural connections; mutations in some that lead
to spoken‐language deficits. Our findings lead us to propose a motor theory of vocal learning origin,
where brain pathways that control vocal learning in different species independently evolved from
surrounding brain pathways that control motor learning, due to repeated evolutionary changes in genes
that control neural circuitry of specific cell types. These pre‐existing anatomical, cellular, and molecular
constraints indicate that there is a limited way in which brain pathways for complex traits can evolve.
These discoveries and the manner in how I approached science were heavily influenced by my
upbringing as an underrepresented minority in the Americas. Diversity and education were heavily
valued, and the challenges faced were well recognized. I have learned approaches to overcome these
challenges, which I believe can help science broadly.
Bruce Alberts Award for Excellence in Science Education
D. Harmon Hines1; 1School Services, University of Massachusetts Medical School, Worcester, MA
Since the 1970s, the University of Massachusetts Medical School (UMMS) has provided programs to
increase access to professions requiring STEM preparation, particularly the health professions,
biotechnology and biomedical research. Initially there were two programs: the High School Health
Careers Program and the Summer Enrichment Program for undergraduates interested in the health
professions. Tracking data are available for both programs.
In the early 1990s, the Regional Science Center (RSRC), became part of UMMS and provided
professional development for K‐12 science and math teachers in over 100 Massachusetts school districts
annually, a home for the Central Massachusetts STEM network, managed the Massachusetts State
Middle School Science Fair and supported a science laboratory and K‐12 STEM curriculum library for
area teachers and students.
In 1993, a fourth program was created, the Summer Undergraduate Research Program (SURP). Funded
continuously since then by the NHL&BI, the SURP initially targeted students underrepresented in
biomedical research. In 2008 it was expanded to target students from disadvantaged backgrounds or
with disabilities. Simultaneously in 2008, the SURP was combined with an internally‐funded Summer
Undergraduate Research Experience (SURE) for undergraduates who did not meet NHL&BI eligibility
requirements. Both programs operated as the Combined Summer Undergraduate Research
Opportunity. Since 1993, the NHL&BI program has hosted over 400 undergraduates and SURE has
hosted approximately 100 undergraduates. As of 2012, of 353 NHL&BI trainees: 27 participated for two
years; 251 graduated from four‐year colleges and universities; 21 completed doctorate programs, while
8 were still enrolled; 21 completed MD programs, while 24 were still enrolled; 38 completed master’s
degrees; 2 completed DVM; 4 completed DDM/DDS; 32 entered or completed MD/PhD; 265 were listed
as authors, coauthors or acknowledged in peer articles in peer reviewed journals. These data are being
The Worcester Pipeline Collaborative (WPC) was initiated in 1996. It is comprised of several partner
institutions, including Worcester Public Schools, Quinsigamond Community College, Worcester State
University, Plumley Village (low income housing project), AbbVie, and UMass Memorial Health Care.
Initially funded by the Robert Wood Johnson Foundation with matching funds from the partners, since
2001 the WPC has been solely supported by the partners. Serving over 6000 students annually with
linked K‐16+ programs, the WPC goal is to increase the numbers of underrepresented and
disadvantaged students entering the health professions and biomedical research. Tracking data are
available for this program was well.
Keith R. Porter Lecture
Monitoring translation in space and time with ribosome profiling.
J.S. Weissman1,2; 1Cellular Molecular Pharmacology, University of California San Francisco, San
Francisco, CA, 2Howard Hughes Medical Institute, Chevy Chase, MD
The ability to sequence genomes has far outstripped approaches for deciphering the information they
encode. We have developed a suite of techniques based on ribosome profiling (deep sequencing of
ribosome protected fragments) that dramatically expand our ability to follow translation in vivo. I will
present recent applications of our ribosome profiling approach including the following: (1) Development
of ribosome profiling protocols for a wide variety of eukaryotic and prokaryotic organisms. (2) Uses of
ribosome profiling to globally monitor when chaperones, targeting factors or processing enzymes
engage nascent chains. (3) Deciphering the driving force and biological consequences underlying the
choice of synonymous codons.
Oral Presentations‐Monday, December 14
Symposium 3: Embraces across the Species Barrier: Complex Cell Interactions
Wolbachia, Microtubules and Big Sur.
W.T. Sullivan1, L.R. Serbus2, P. White1; 1MCD Biology, University of California, Santa Cruz, Santa Cruz , CA,
Biology, Florida International University, Miami, FL
Wolbachia are obligate intracellular bacteria carried by millions of insect and nematode host species.
The success of Wolbachia is largely a result of its combined ability to be efficiently transmitted through
the female lineage and to profoundly alter host development and reproduction to favor females. This
includes male‐killing, induction of parthenogenesis, sex reversal (Wolbachia induces males to develop as
fertile females) and a conditional form of sterility favoring infected over uninfected females. Like
mitochrondria, Wolbachia are exclusively transmitted through the female germline. This requires that
the bacteria navigate the ever‐changing cytoskeletal environment of the developing oocyte to
concentrate at the posterior pole, the future site of the germline in the next generation. To accomplish
this, Wolbachia must undergo precisely timed engagements with host minus and plus‐end directed
microtubule motor proteins followed by an association with specific posterior determinants. The
regulation of these interactions will be discussed. In addition, the recovery of informative Wolbachia
variants in oocyte localization and transmission isolated from wild‐populations will be described.
Interactions across biological scales within the cheese rind ecosystem.
R.J. Dutton1; 1Division of Biological Sciences, Section of Molecular Biology, University of California, San
Diego, La Jolla, CA
Experimentally tractable microbial communities are needed to bridge the gap between observing
patterns of microbial diversity and defining the mechanisms that give rise to these patterns. We
developed cheese rinds as model microbial communities by characterizing in situ patterns of diversity,
culturing all dominant species, and developing a simple in vitro system for community reconstruction.
One key characteristic is the widespread positive and negative interactions amongst bacterial and fungal
community members. These interactions span many biological scales, from modification of the
environment to horizontal transfer of genes within the community. We are now developing genetic, cell
biological, and chemical approaches to studying species interactions in this model microbial community
with the goal of defining molecular mechanisms and principles of community formation.
Bacteriophage Adherence to Mucus (BAM) Immunity: How Phage and the Microbiome form
Innate and Acquired Immune Systems on Mucosal Surfaces.
F. Rohwer1; 1Biology , San Diego State University, San Diego, CA
Bacteriophage Adherence to Mucus (BAM) Immunity is a novel, microbiome‐derived, phage‐mediated
immunity active at mucosal surfaces. Phage stick to mucin glycoproteins via hypermutable,
immunoglobulin‐like domains displayed on their capsids and protect the underlying epithelial cells from
invading bacteria. The complexity of these interactions between phage, microbes and the macrobial
host are just starting to be de‐convoluted. Using an organ‐on‐a‐chip approach, we show that mucus‐
binding phage have a subdiffusive hunting pattern that resulted in more frequent host interactions than
non‐mucus‐adherent phages moving by normal Brownian diffusion. A supporting theory and
experiments show that subdiffusion as a more efficient phage search strategy at low bacterial
concentrations found deeper within mucus layers. These findings extend previous observations that
phage form an acquired and adaptive immune system on mucosal surfaces and for phage manipulation
of microbiomes.
Symposium 4: Like Oil and Water: New Principles Governing Cell Organization
Physical Mechanisms of Cell Organization on Micron Length Scales.
M.K. Rosen1; 1Biophysics, UT Southwestern Medical Center/HHMI, Dallas, TX
Cells are organized on length scales from Angstroms to microns. However, the mechanisms by which
Angstrom‐scale molecular properties are translated to micron‐scale macroscopic properties are not well
understood. We have shown that interactions between multivalent proteins and multivalent ligands can
cause liquid‐liquid demixing phase transitions, resulting in formation of micron‐sized liquid droplets in
aqueous solution and micron‐sized puncta on membranes. These transitions appear to occur
concomitantly with sol‐gel transitions to form large, dynamic polymers within the droplets/puncta. I will
discuss how such transitions may control the spatial organization and biochemical activity of actin
regulatory signaling pathways, and contribute to formation of PML nuclear bodies in the mammalian
nucleus. Our data suggest a general mechanism by which cells may achieve micron‐scale organization
based on interactions between multivalent macromolecules.
Phase separation in cytoplasm: Implications for polarity and neurodegeneration.
A.A. Hyman1; 1Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany
Many proteins contain disordered regions of low sequence complexity, which cause aging‐associated
diseases because they are prone to aggregate. The molecular properties of these proteins, and how they
relate to disease, are not well understood. Here, we study FUS, a prion‐like protein containing
intrinsically disordered domains associated with neurodegenerative disease. We show that in cells, FUS
forms liquid compartments at sites of DNA damage and in the cytoplasm upon stress. We confirm this
by reconstituting liquid FUS compartments in vitro. Using an in vitro “aging” experiment, we
demonstrate that liquid droplets of FUS protein convert with time from a liquid to an aggregated state,
and this conversion is accelerated by patient‐derived mutations. We conclude that the physiological role
of FUS is to form dynamic liquid‐like compartments. We propose that aberrant phase transitions of a
liquid‐like compartment lie at the heart of ALS and presumably many other age‐related diseases.
Microsymposium 7: Microtubule Dynamics: From +TIPs to Membrane
Active Contraction of Microtubule Networks.
P.J. Foster1,2, S. Fürthauer3,4, M.J. Shelley4, D.J. Needleman1,2,3; 1FAS Center For Systems Biology, Harvard
University, Cambridge, MA, 2School of Engineering and Applied Sciences, Harvard University, Cambridge,
MA, 3Department of Molecular and Cellular Biology, Harvard University, Cambridge, MA, 4Courant
Institute of Mathematical Science, New York University, New York, NY
Many cellular processes including cell motility, cell shape control, and cell division are driven by
cytoskeletal assemblies. It remains unclear how cytoskeletal filaments and motor proteins organize into
cellular scale structures and how the molecular properties of cytoskeletal components affect the large
scale behaviors of these systems. Here we investigate the self‐organization of stabilized microtubules in
Xenopus oocyte extracts and find that they can form macroscopic networks that spontaneously
contract. We propose that these contractions are driven by the clustering of microtubule minus ends by
dynein. To test this idea, we construct a mesoscopic model of motor induced microtubule end
clustering, which we coarse grain to develop a theory of network contractions. This theory predicts a
dependence of the timescale of contraction on initial network geometry, a development of density
inhomogeneities during contraction, a final network density that is independent of initial geometry and
dynein inhibition, and a strong influence of dynein inhibition on the rate of contraction, all in
quantitative agreement with experiments. These results demonstrate that the motor driven clustering
of filament ends is a generic mechanism leading to contraction, and argue that dynein generated
contractile stresses could play a role in spindle assembly.
The kinesin‐4 KIF21B is specialized to independently regulate trafficking and microtubule
dynamics in dendrites.
A.E. Ghiretti1, E.L. Holzbaur1; 1Department of Physiology, University of Pennsylvania Perelman School of
Medicine, Philadelphia, PA
Neurons are highly polarized cells, extending distinct processes specialized to send (axons) and receive
(dendrites) information. The accurate trafficking of the correct protein cargoes to either the axonal or
dendritic compartments is required to maintain this polarity and therefore proper neuronal function.
While axonal trafficking is well studied due to the uniform polarity of axonal microtubules, dendritic
microtubules are of mixed polarity; as a result, little is known about how dendritic trafficking is achieved
and only a handful of dendrite‐specific kinesins have been identified to date. Through a combination of
single molecule assays and live imaging in rodent hippocampal neurons, we have identified a dual role
for the kinesin‐4 KIF21B in dendrites as a molecular motor as well as a positive regulator of microtubule
dynamicity. Intriguingly, KIF21B motor activity is determined by an N‐terminal, nucleotide‐dependent
microtubule binding region, while dynamic remodeling of microtubules is entirely separable and
regulated through nucleotide‐independent activity of the C‐terminal region of the protein. KIF21B is
highly processive, with run lengths of nearly 10 μm in vitro, and exhibits a strong bias towards
retrograde motility in dendrites, where it mediates the trafficking of TrkB‐positive signaling endosomes
in the proximal dendrite. KIF21B also directly affects microtubule dynamics, increasing both the growth
rate and catastrophe frequency of dynamic microtubules in vitro. In neurons, KIF21B knockdown
enhances dendritic EB3 comet length, and overexpression of a C‐terminal KIF21B fragment is sufficient
to rescue this defect. This demonstrates that the motor activity of this kinesin is completely dispensable
for its role in regulating microtubule dynamics. Overall, our studies of KIF21B show that, distinct among
the kinesins, it is uniquely tuned to navigate the complex cytoskeletal environment of dendrites by
compartmentalizing functions to disparate protein regions. This provides new insight into how the
processive motility of cargoes and proper microtubule arrangement is maintained in the face of
dendritic arbor dynamics and synaptic plasticity changes that occur in response to neuronal activity.
Cytoskeletal processes in vitro in 3D: new perspective on the motor tug‐of‐war.
M. Vershinin1, O. Osunbayo1, J. Bergman1; 1Physics and Astronomy, University of Utah, Salt Lake City, UT
Intracellular cargo routing along microtubules often features cargos navigating 3D filament
intersections. Such a process is therefore not reducible to the simplified picture of a single cargo moving
along a single microtubule, and it cannot be fully modeled in a surface‐bound in vitro assay. We have
developed a novel 3D microtubule motility assay which allows multiple microtubules to be
independently held and positioned in 3D. In addition, forces exerted by cargos on microtubules can be
accurately quantified. We will present relevant details of the approach and its capabilities. With this
approach we demonstrate that cargos driven by kinesin‐1 motors experience significant tug‐of‐war at
intersections even in the low motor limit, i.e. under conditions where a single motor is likely engaged on
each intersecting microtubule. We further show that geometry of the intersection affects cargo
navigation across the intersection.
Kinesin‐5 is a microtubule polymerase.
W.O. Hancock1, Y. Chen1; 1Biomedical Engineering, Penn State University, University Park, PA
Kinesin‐5 slides antiparallel microtubules during spindle assembly and regulates the branching of
growing axons. Besides the mechanical activities enabled by its tetrameric configuration, the specific
motor properties of kinesin‐5 that underlie its cellular function remain unclear. By engineering a stable
kinesin‐5 dimer and reconstituting microtubule dynamics in vitro, we demonstrate that kinesin‐5
promotes microtubule polymerization by increasing the growth rate and decreasing the catastrophe
frequency. Strikingly, microtubules growing in the presence of kinesin‐5 have curved plus‐ends,
suggesting that the motor stabilizes growing protofilaments. Single‐molecule fluorescence experiments
reveal that kinesin‐5 remains bound to the plus‐ends of static microtubules for 7 seconds, and tracks
growing microtubule plus‐ends in a manner dependent on its processivity. We propose that kinesin‐5
pauses at microtubule plus‐ends and enhances polymerization by stabilizing longitudinal tubulin‐tubulin
interactions, and that these activities underlie the ability kinesin‐5 to slide and stabilize microtubule
bundles in cells.
Direction specific microtubules are the rails for interflagellar transport trains.
G. Pigino1, L. Stepanek1; 1Molecular Cell Biology end Genetics, Max Planck Institute, Dresden, Germany
The cilium is a large and complex microtubule‐based molecular machine that is performs vital motility,
signaling, and sensing functions in most eukaryotic cells. The assembly of the cilium requires the
intraflagellar transport (IFT), a multifactorial machinery that organizes in large IFT trains and moves
ciliary precursors between the cell body and the distal tip of the cilium. Numerous IFT trains and their
cargoes are moved along the microtubules of the cilium by IFT specific molecular motors. Plus‐end‐
directed kinesins drive anterograde IFT from the cell body to the tip and minus‐end‐directed cytoplasmic
dynein 2 moves retrograde IFT back to the cell body. Although we know that IFT is necessary for cilia
assembly, the logistics that ensures fast and efficient cargos transport in the extremely crowded ciliary
environment in not understood. By TIRF imaging of IFT trains in Chlamydomonas reinhardtii cilia, we
show that oppositely directed trains pass by each other without obvious interactions. We hypothesize
the presence of a cellular mechanism that prevents collisions between anterograde and retrograde IFT,
contributes to the control of motors activity, and ensures fast reliable cargo transport. To test this
hypothesis, we developed a novel time‐resolved correlative fluorescence and 3D‐electron microscopy
approach that allows the investigation of IFT dynamics with TEM spatial resolution. We show that
anterograde and retrograde IFT trains avoid collisions by using distinct sets of axonemal microtubules.
We provide direct evidence for a new revised ultrastructural classification of anterograde and
retrograde IFT trains, and we also identify a third class of static IFT trains with distinguished morphology.
Spatiotemporal control of intracellular microtubule dynamics by light.
J. Van Haren1, A.W. Ettinger1, H. Wang2, K.M. Hahn2, T. Wittmann1; 1Cell and Tissue Biology, University of
California, San Francisco, CA, 2Pharmacology, University of North Carolina, Chapel Hill, NC
Dynamic remodeling of the microtubule (MT) cytoskeleton through stochastic switching between phases
of growth and shortening is critical for all MT functions in cells. This dynamic non‐equilibrium behavior
underlies the ‘search‐and‐capture’ hypothesis first proposed by Mitchison and Kirschner in which
dynamic MTs explore the cytoplasm to establish and maintain polarized cell dynamics. However, the
molecular underpinnings of spatial and temporal control of intracellular MT dynamics and interactions
are incompletely understood. In particular, how MT search‐and‐capture is spatially and temporally
controlled at rapid timescales relevant to MT‐driven intracellular processes remains unclear, and most
importantly the causality of local MT regulation and associated cell dynamics has not been rigorously
established. Control of MT polymerization dynamics as well as dynamic interactions of growing MTs with
other intracellular components are mediated to a large extent by a heterogeneous class of proteins
commonly referred to as +TIPs. Interactions of these +TIPs with growing MT plus ends require central
adapters of the EB family, in particular EB1 and EB3 in most mammalian cells. It is not known how the
function of this complex +TIP network at MT ends is dynamically regulated, which represents a
significant gap in our understanding of local MT function, especially since different +TIPs can have
opposing activities. For example, interactions with EBs are involved in recruiting both the MT
polymerase XMAP215/ch‐TOG as well as the MT depolymerase MCAK to growing MT ends, and it is not
known how these opposing polymerizing and depolymerizing activities are spatially and temporally
balanced in cells. To decipher +TIP activities at high spatial and temporal resolution in cells, we therefore
constructed a light‐inactivated EB1 molecule by utilizing a recently developed phototropin‐based
protein interaction pair that rapidly dissociates upon blue light exposure. We show that this “opto‐EB1”
construct can replace endogenous EB function, is rapidly inactivated by blue light and recovers to the
dark state within seconds allowing highly spatially controlled dynamic manipulation of the MT
cytoskeleton in cells. We further demonstrate that opto‐EB1 inactivation induces catastrophes of a
population of cell edge associated MTs and reduces persistent growth of MTs in the cell body consistent
with +TIP functions as capture factors and a predominant role of XMAP215/ch‐TOG in promoting
interphase MT growth. We are currently testing how local control of MT dynamics by patterned
illumination affects cellular behaviors such as migration and division.
A complex relationship between motor activity, cytoplasmic flows and the organisation of the
actin and microtubule cytoskeletons.
I.M. Palacios1, M. Drechsler1, S. Ganguly2; 1Zoology, University of Cambridge, Cambridge, United
Kingdom, 2Yale University, New Haven, CT
Motor driven transport along cytoskeletons constitutes a fundamental mechanism involved in many
cellular processes. We use the Drosophila oocyte as a versatile model to study the complex relationship
between motor proteins, the cytoskeletons and the viscoelastic properties of the cytoplasm. Kinesin‐1
(Kin) driven transport induces bulk movements of the oocyte cytoplasm – called flows. The topology of
flows correlate with the architecture of the underlying microtubules, while the velocity is directly linked
to Kin speed1. Flows also depend on a novel cytoplasmic actin structure known as the “actin mesh”2,
which negatively regulates flows by a yet unknown mechanism.
Following our biophysical characterisation of the micro‐rheology of the ooplasm1, we are carrying on
the analysis of this complex relationship between flows and the cytoskeletons. We have found that
altered flow velocities strongly alter microtubule conformation. In the absence of flows microtubules
appear immobile and with a lower degree of bundling. Conversely, faster flows, induced by abolishing
the actin mesh both genetically as well as chemically, increase microtubule bulk movement, order and
bundling. Our first results suggest that the mesh regulates Kin binding to microtubules, potentially
explaining the observed bundling phenotypes.
Furthermore, Kin activity itself not only affects microtubule and flows, but also the organisation of the
actin mesh, indicating a mutual crosstalk between both cytoskeletons. Preliminary data obtained by
Digital Fourier Microscopy (in collaboration with R. Cerbino, Milan) suggest that the dynamics of the
actin mesh correlate with the features of flows, and therefore with Kin activity. More strikingly,
abolishing Kin causes a progressive breakdown of the mesh, which illustrates an unexpected and novel
requirement for a microtubule motor in actin organisation.
To further understand the coupling between motor‐induced forces, fluid dynamics and cytoskeletal
organisation, we are currently extending our analysis to super‐resolution microscopy and advanced
image analysis, as well as to the mouse oocyte.
Ganguly, S., et al. Proc Natl Acad Sci U S A 109, 15109‐15114 (2012)
Dahlgaard, K., et al. Dev Cell 13, 539‐553 (2007)
Microsymposium 8: The Role of the Cytoskeleton in Disease and Repair
Dynamic force patterns promote collective cell migration and rapid wound repair.
T. Zulueta‐Coarasa1, R. Fernandez‐Gonzalez1,2,3; 1Institute of Biomaterials and Biomedical Engineering,
University of Toronto, Toronto, ON, 2Department of Cell and Systems Biology, University of Toronto,
Toronto, ON, 3Developmental and Stem Cell Biology Program, The Hospital for Sick Children, Toronto,
An outstanding question in cell biology is how cells coordinate their behavior during tissue assembly and
repair. During embryonic wound healing, the contraction of a supracellular cable formed by filamentous
actin and the motor protein non‐muscle myosin II coordinates the cell movements that drive wound
closure. We showed that, in Drosophila embryos, the actin cable displays a heterogeneous pattern, with
regions of high and low actin density. Mutants in which the heterogeneous distribution of actin at the
wound margin is lost display significantly slower wound closure. However, the mechanisms by which a
non‐uniform distribution of actin favors rapid wound repair are unknown. We used in vivo microscopy,
quantitative image analysis and in silico modeling to investigate how the distribution of cytoskeletal
molecules affects collective cell migration during wound closure in Drosophila embryos. We found that
myosin displays a heterogeneous distribution at the wound margin that matches the actin pattern.
Using laser ablation and particle image velocimetry, we demonstrated that segments of the cable with
high actomyosin levels sustain higher contractile forces and contract faster than areas of the cable with
low actomyosin density, indicating that contraction of the cable around the wound is non‐uniform. A
simple force balance showed that the closure forces that result at the interface between regions of high
and low actomyosin contractility are greater than in regions of uniform contractility, suggesting that a
non‐uniform distribution of tension at the wound margin favors rapid wound repair. Moreover, we
developed an in silico vertex model of wound repair and we found that, for similar levels of total force
generated, wound healing was faster when the pattern of tension along the wound margin was
heterogeneous ‐corresponding to the in vivo distribution of myosin‐ than when we made it
homogeneous. Our model predicted that, to maintain a heterogeneous distribution of tension around
the wound over time, the pattern of cytoskeletal molecules must change as cell boundaries deform.
Consistent with this, we found that myosin rapidly accumulates in segments of the wound margin that
are stretched as a consequence of the contraction of adjacent, myosin‐rich segments. Together, our
results suggest that a non‐uniform distribution of contractile forces along multicellular actomyosin
networks generates mechanical signals that facilitate rapid network contraction and remodeling, and
efficient collective cell movements.
Abnormal cortical neuron migration by perturbed nucleus‐centrosome coupling underlies the
pathophysiology of autism with abnormality in RBFOX1/A2BP1 gene.
K. Nagata1, N. Hamada1, H. Ito1, I. Iwamoto1, R. Morishita1, H. Tabata1; 1Molecular Neurobiology,
Institute for Developmental Research, Aichi Human Service Center, Kasugai, Japan
Background: RBFOX1/A2BP1 is an RNA‐binding protein necessary for proper pre‐mRNA splicing events in
genes crucial for neuronal functions. Critical functions of RBFOX1 in brain development have been
approved by the gene abnormalities that cause neurodevelopmental disorders including autism
spectrum disorder (ASD), attention deficit hyperactivity disorder (ADHD) and schizophrenia. Methods:
To elucidate the pathophysiological relevance of RBFOX1, we performed a battery of in vivo and in vitro
analyses of RBFOX1 during mouse corticogenesis. In vivo analyses were based on in utero
electroporation, and we examined the role of RBFOX1 in cortical neuron migration, axon elongation,
dendritic arbor formation and proliferation. Migration profiles were examined carefully with confocal
microscope‐assisted live‐imaging. In vitro analyses were done with primary cultured neurons and
focused on various morphological aspects including spine dynamics. Results: Knockdown of RBFOX1 in
utero caused abnormal cortical neuron migration during brain development. Intensive time‐lapse
imaging analyses revealed that RBFOX1‐silencing caused defects in 1) the transformation from
multipolar to bipolar shape, an essential event occurring in the ventricular zone at the initiation of radial
migration, 2) the radial migration in the cortical plate, and 3) the terminal translocation near the pial
surface. These abnormal phenotypes during the cortical migration process were due to the impairment
of nucleokinesis (intracellular translocation of the nucleus during cell migration). These results indicate
that the microtubule cytoskeleton and centrosome play essential roles in the cortical neuron migration.
In addition, RBFOX1 was found to regulate neuronal network formation in vivo since axon growth and
dendritic arborization were suppressed in RBFOX1‐deficient cortical neurons. Aberrant morphology was
further confirmed in vitro; RBFOX1‐silencing in primary cultured hippocampal neurons resulted in the
reduction of primary axon length, total length of dendrites, spine density and mature spine number.
Conclusions: Functional abnormalities of RBFOX1 induced abnormal neuronal migration during the
cortical development. Centrosome‐mediated microtubule regulation was found to be essential for the
neuronal migration and morphology during corticogenesis. Disturbance of RBFOX1 function may lead to
structural and functional defects in the cerebral cortex, and consequently contribute to ASD and other
neurodevelopmental and psychiatric disorders.
The physical interaction between tumor and endothelial cells is a key inducer for liver cancer
angiogenesis in a HepG2‐HUVEC co‐culture system.
G. Chiew1, A. Fu1, K.Q. Luo1; 1School of Chemical and Biomedical Engineering, Nanyang Technological
University, Singapore, Singapore
Blood vessel remodeling is crucial in tumor growth. Growth factors released by tumor cells and
endothelium‐extracellular matrix interactions are highlighted in tumor angiogenesis, however the
physical tumor‐endothelium interactions are highly neglected. Here, we report that the physical
supports from hepatocellular carcinoma, HepG2 cells, are essential for the differentiation and
remodeling of endothelial cells (HUVEC).
We developed a novel HepG2‐HUVEC co‐culture system for studying tumor‐endothelium interactions.
When HUVEC cells were introduced to a monolayer of HepG2 cells, the endothelial cells undergo
differentiation and reorganization to form tubular networks similar to those plated on matrigel, with
protrusions and sprouts structures formed above this sheet of HepG2 cells. Cytoskeleton staining of
actin and intermediate filament revealed phenotypic activation of HUVEC cells in response to HepG2
cells, and the corresponding changes HepG2 cells undertook to accommodate the differentiated HUVEC
cells. This differentiation was found to be induced by physical interaction rather than growth factors
secreted by HepG2 cells. Only HUVEC cells in direct contact with HepG2 could form angiogenic sprouts.
Furthermore, in separate experiments, fixed HepG2 cells could stimulate endothelial cells differentiation
while the conditioned media could not, indicating that the physical interactions between tumor and
endothelial cells were indispensable. To further investigate the endothelium‐remodeling mechanisms,
we treated the co‐culture system with different inhibitors, including inhibitors of PI3K/mTOR pathway,
Ras/Raf/MEK/ERK pathway, p38 and JNK pathway, focal adhesion kinase (FAK) and Rho family. Inhibitors
targeting focal adhesions and Ras/Raf/MEK pathway effectively inhibited the differentiation of HUVEC,
while the inhibitor targeting vascular endothelial growth factor receptor (VEGFR) displayed little effects.
In conclusion, the novel co‐culture system of HepG2‐HUVEC induced tubular network formations
representative of angiogenesis. We were able to eliminate secretory factors released by tumor cells as
the potential inducers of HUVEC differentiation under this co‐culture system. The physical interaction
between HepG2 and HUVEC is the key factor in tilting the angiogenic balance. Our co‐culture system can
function as a platform to study the cellular signaling pathways in tumor angiogenesis. Furthermore, a
hepatocellular carcinoma (HCC) tumor microenvironment was generated for investigating tumor‐
endothelium interactions. Understanding the physical interactions of the tumor‐endothelium provides
valuable insights on tumor angiogenesis and offers new targets in the search of therapeutic agents to
inhibit HCC and tumor angiogenesis.
Abnormal Expression of Laminin‐332 Promotes Cell Proliferation and Cyst Growth in Autosomal
Recessive Polycystic Kidney Disease.
S. Vijayakumar1, S. Dang2, D.P. Wallace3, B.K. Yoder4, P. Marinkovich5, Z. Lazarova6, V.E. Torres7; 1Natural
Sciences Mathematics Department, SUNY Cobleskill, Cobleskill, NY, 2Department of Medicine, University
of New England College of Osteopathic Medicine, Biddeford,, ME, 3Department of Medicine, University
of Kansas Medical Center, Kansas City, KS, 4Department of Cell Biology, University of Alabama,
Birmingham, AL, 5Program in Epithelial Biology, Stanford University School of Medicine, Stanford, CA,
Department of Dermatolgy, Medical College of Wisconsin, Milwaukee, WI, 7Division of Nephrology
Hypertension, Mayo Clinic, Rochester, MN
Polycystic kidney diseases (PKD) are a common indication for renal transplantation and dialysis and a
leading cause of end‐stage renal disease (ESRD). Although the autosomal recessive polycystic kidney
disease (ARPKD), caused by mutations in PKHD1 (encoding fibrocystin) is not as common as the
autosomal dominant polycystic kidney disease (ADPKD) and affects only 1 in 20,000 live births, it is an
important cause of ESRD and mortality in infants and children. Basement membrane abnormalities have
often been observed in the kidney cysts of polycystic kidney disease (PKD) patients and animal models
but their role in cystogenesis is yet unclear. Knockdown of PKD1 (an ADPKD gene) paralogs in zebrafish
leads to dysregulated synthesis of the extracellular matrix, suggesting that altered basement membrane
assembly may be a primary defect in ADPKD. Previous studies have shown that in squamous cell
carcinoma, laminin‐332 promotes tumor invasion and cell survival through interactions with several cell‐
surface receptors (including α6β4 and α3β1 integrins). In this study, for the first time, we demonstrate
that laminin‐332 is aberrantly expressed in cysts and precystic tubules of human autosomal recessive
(ARPKD) kidneys as well as in the kidneys of PCK rats, an orthologous ARPKD model. Aberrant laminin‐
332 expression was observed as early as postnatal day 30 (PN30) coinciding with the appearance of
distinctly visible renal cysts. We also show that a kidney cell line derived from orpk mice, another model
of ARPKD, exhibited abnormal lumen‐deficient and multi‐lumen structures in matrigel culture. These
cells had increased proliferation rates and altered expression levels of laminin‐332 compared to their
rescued counterparts. A function‐blocking polyclonal antibody to laminin‐332 significantly inhibited their
abnormal proliferation rates and rescued their aberrant phenotype in matrigel culture. Furthermore,
our results show that in human end‐stage ADPKD kidneys abnormal laminin‐332 expression is observed
not only in the cysts originating from collecting ducts as well as proximal tubules, but also in precystic
tubules. Also our results show that integrin β4 expression was upregulated in PCK rat kidneys compared
to the control SD (sprague dawley) kidneys. Taken together, our results suggest that the abnormal
laminin‐332 expression may be an underlying mechanism for cystogenesis in various forms of PKD and
that it may contribute to cystogenesis by altering the proliferation rates of tubular epithelia via the
integrin α6β4 ‐ ERK pathway.
Sentinel cells of the social amoeba Dictyostelium traps and kills bacteria by casting DNA nets.
X. Zhang1, O. Zhuchenko2, K. Adam2, T. Soldati1; 1Department of Biochemistry, University of Geneva,
Geneva, Switzerland, 2Verna and Marrs McLean Department of Biochemistry and Molecular Biology,
Baylor College of Medicine, Houston, TX
As part of human innate immune defense, activated neutrophils are able to generate reticulated nets of
DNA decorated with antimicrobial granules and proteins, known as extracellular traps (ETs), to capture
and kill extracellular pathogens. Recent findings indicate that ET generation is not a unique feature of
the human innate immune system, but represents an ancient host‐defense mechanism shared by
several specific phagocytic cell types across vertebrates and invertebrates. However, when and how this
antimicrobial mechanism originated, and whether it is a feature unique to the animal kingdom are still
unclear. As the only cell type with phagocytic capability, Sentinel (S) cells in the multicellular slug of the
amoebal model organism, Dictyostelium discoideum, function as a primitive innate immune system for
clearing invading bacteria. Therefore, the S cells represent a unique pre‐metazoan model system to
study the evolution and conserved mechanisms of innate immunity. In this study we observed that,
upon stimulation with bacteria or lipopolysaccharide (LPS), only the S cells are able to produce ETs for
trapping and killing invading bacteria. Contrary to human neutrophils, S cells produce mitochondrial
DNA‐based ETs instead of chromosomal DNA based ETs, without immediate cell death. In addition, we
showed that S cells are the major producer of reactive oxygen species (ROS) in the slug, and inhibition of
ROS production suppressed ET generation in a dose‐dependent manner. Similar to human neutrophils,
production of ETs by S cells requires a conserved Toll/Interleukin 1 receptor (TIR) domain containing
protein and ROS‐generating NADPH oxidases (NOXs), and knocking out of these genes results in
decreased clearance of bacterial infections during development. Our results demonstrate that the
multicellular stage of D. discoideum is an excellent model to study the early evolution of innate
immunity, and further suggests that the origin of DNA‐based ETs as an innate immune defense predates
the appearance of animals.
Modeling Polycystic Kidney Disease Cystogenesis with Genome‐Modified Human Pluripotent
Stem Cells.
B.S. Freedman1,2, T.I. Steinman3, J. Zhou2, J.V. Bonventre2; 1Department of Medicine, Division of
Nephrology, University of Washington, Seattle, WA, 2Department of Medicine, Renal Division, Brigham
and Womens Hospital/Harvard Medical School, Boston, MA, 3Department of Medicine, Division of
Nephrology, Beth Israel Deaconess Medical Center/Harvard Medical School, Boston, MA
Human pluripotent stem cells (hPSCs) can self‐renew extensively and differentiate into diverse tissues,
including recently tubular organoids expressing kidney markers. To date, however, no study has
demonstrated a phenotype in these tubules relevant to kidney disease. Polycystic kidney disease (PKD)
is an extremely common, life‐threatening Mendelian disorder in which balloon‐like cysts arise from
kidney tubules. Using the CRISPR/Cas9 genome editing system, we tested the ability of hPSCs and
derived kidney organoids to model PKD cystogenesis in vitro.
To generate PKD hPSCs, Cas9 nuclease and guide RNAs targeting the disease‐causative genes, PKD1 or
PKD2, were transfected into hPSCs. Chromatogram analysis and immunoblotting indicated biallelic,
frame‐shift mutations at the target site and the absence of the corresponding full‐length proteins. hPSCs
in 3D cultures were treated with specific growth factors to direct stepwise differentiation into kidney
progenitor cells (SIX2+PAX2+) and subsequently proximal tubules (LTL+LRP2+). PKD cultures were
inspected microscopically for cystogenesis phenotypes, compared to unmodified controls of identical
genetic background.
Phenotypically, PKD hPSCs exhibited self‐renewal and pluripotency characteristics comparable to
isogenic controls, and differentiated into tubular organoids with similar efficiencies. Interestingly, in PKD
hPSC cultures, we observed formation of large, translucent, cyst‐like structures alongside tubules.
Confocal microscopy revealed that cyst‐lining epithelia reacted strongly with LTL, were highly
proliferative, and surrounded hollow interior compartments devoid of cells. Time‐lapse imaging
revealed that cysts arose from tubular structures. Importantly, isogenic control hPSCs plated and
differentiated side‐by‐side did not form cysts under these conditions. PKD mutations did not augment
the size of cavitated spheroids of undifferentiated (epiblast‐stage) hPSCs, a tissue not known to be
affected in PKD patients.
Our findings suggest that PKD‐specific cyst formation from tubules can be reproducibly modeled in a
minimal system in vitro. Cysts arise from both PKD1 and PKD2 mutants, but not in healthy controls of
identical genetic background. They are furthermore specific to the kidney lineage. These findings
support the identification of hPSC‐derived tubules as kidney nephron structures. Genome‐modified
hPSCs represent a new, pre‐clinical model in which to investigate PKD pathophysiology and evaluate
candidate therapeutics.
Macrophage‐dependent activation of Notch1 signaling regulates breast tumor cell
J. Pignatelli1, J. Bravo Cordero1,2,3, M. Roh1, S. Gandhi1, Y. Wang1, R.H. Singer1,2, L. Hodgson1,2, M. Oktay1,4,
J.S. Condeelis1,2; 1Anatomy and Structural Biology, Albert Einstein College of Medicine, Bronx, NY, 2Gruss
Lipper Biophotonics Center, Albert Einstein College of Medicine, Bronc, NY, 3Medicine, Mount Sinai ,
New York, NY, 4Pathology, Albert Einstein Collge of Medicine, Bronx, NY
The process of intravasation, a key component to the metastatic cascade, remains poorly understood.
How the multi‐cell type tumor microenvironment facilitates tumor cell intravasation is largely unknown.
Intravital imaging of rodent mammary tumors has shown that direct contact between a Mena
expressing tumor cell, a perivascular macrophage and an endothelial cell forms a microanatomical
structure named TMEM (Tumor Microenvironment of Metastasis) which is the site where intravasation
occurs in mammary tumors. Clinical studies have shown that the number of TMEM is correlated with
increased risk of developing distant metastasis in breast cancer patients. While TMEM is an excellent
prognostic marker for predicting metastasis, the mechanisms of TMEM assembly and function are not
well understood. Recently, we showed that heterotypic cell contact between tumor cells and
macrophages induces the formation of invadopodia in tumor cells, invasive structures necessary for
matrix degradation and required for tumor cell intravasation. In further work we found that Notch1, a
known receptor involved in heterotypic cell contact signaling, is involved. In the absence of Notch1
signaling, macrophage‐induced invadopodium formation and tumor cell intravasation are abolished.
This heterotypic tumor cell – macrophage interaction regulates the expression profile of Mena in the
tumor cell. Upon touching, the transcription of Mena shows a rapid kinetic response, with detectable
changes in single transcriptome activity within 1 hour. Inhibition of Notch1 in vivo results in decreased
intravasation of mammary tumor cells. Our findings indicate that Notch1 signaling regulates heterotypic
cell contact mediated invadopodium formation, transendothelial migration and intravasation and
reveals a novel Notch1/Mena pathway as a molecular target to prevent TMEM function and therefore
Microsymposium 9: Membrane Trafficking
Human Sar1 paralogs differ biochemically in the assembly of the COPII coat.
D.B. Melville1,2, S. Studer1,2, R.W. Schekman1,2; 1Department of Molecular and Cell Biology, University of
California Berkeley, Berkeley, CA, 2Howard Hughes Medical Institute, University of California, Berkeley,
Berkeley, CA
COPII coated vesicles are the primary mediators of ER‐to‐Golgi trafficking. Sar1, one of the five core
COPII components, is a highly conserved small GTPase, which, upon GTP binding, recruits the other
COPII proteins to the ER membrane. Mammals have two paralogs of Sar1, SAR1A and SAR1B, which
share 90% sequence identity and can both promote vesicle budding in vitro. It has been found that
mutations in SAR1B, but not SAR1A, are associated with Chylomicron retention disease (CMRD) in
humans, suggesting that, despite their strong sequence similarity, they have distinct physiological roles.
We have previously identified a functional difference between the paralogs, namely that when
combined with a disease‐associated allele of SEC23A (F382L), SAR1A is more efficient than SAR1B at
recruiting the Sec13‐Sec31 complex that comprises the COPII outer coat. Because the primary defect of
the F382L‐SEC23A allele appears to be a diminished capacity to recruit SEC31A, we therefore
hypothesized that the differences observed between the Sar1 paralogs may be due to their specific
interactions with SEC31A and possible changes to the kinetics of SAR1 GTPase activity related to SEC31A
interaction. Using in vitro biochemical assays we have identified amino acid residues sufficient to
account for the paralog‐specific differences seen in SAR1, as well as the site where these residues
interact with SEC31A. We have also identified novel factors that may play a role in Sar1 GTPase kinetics.
These data illuminate the relationship between Sar1 paralogs and the COPII outer coat, suggest a
possible mechanism for the paralog‐specific defects seen in CMRD, and provide insights into the role of
GTPase kinetics during large cargo trafficking.
Calcyon Forms a Complex with Cytoplasmic Dynein and Regulates Cargoes Transport in Mature
L. Shi1, N. Muthusamy2, Z. Roth1, C. Bergson3, D. Smith1; 1Department of Biological Sciences, University of
South Carolina, Columbia, SC, 2Department of Molecular Biomedical Sciences, North Carolina State
University, Raleigh, NC, 3Department of Pharmacology and Toxicology, Georgia Regents University,
Augusta, GA
The endosome‐enriched neuronal Calcyon (Caly) is implicated in various neuropsychiatric disorders,
including schizophrenia and attention deficit hyperactivity disorder. Caly stimulates clathrin‐mediated
endocytosis, and is important for axonal targeting of AP‐3 cargoes, but the extent to which this involves
molecular motors is not known. In the current study, mass spectrometric (MS) investigation of the Caly
associated complex in brain revealed the presence of the motor protein, cytoplasmic dynein 1 heavy
chain (DYNC1H) in Caly pulldowns. Further biochemical studies confirmed the strong association of the
Caly‐C terminus with dynein. Caly distribution in COS7 cells was altered by the expression of either the
dynein inhibitor CC1, or the cargo binding intermediate chain, DYNC1I2 (isoform IC‐2C), indicating that
the minus‐end directed transport of Caly associated cargoes was depended on dynein motors. In axons
of cultured adult rat sensory neurons (dorsal root ganglia – DRG neurons) Caly‐labeled organelles co‐
localized with DYNC1I1 (isoform IC‐1B) and a dynein regulatory protein, lissencephaly 1 (LIS1).
Manipulation of dynein function changed the motile properties of Caly‐labeled organelles in axons,
indicating that Caly vesicles utilize dynein motors in mature neurons. More interestingly, manipulation
of Caly expression levels also altered the movement of acidic organelles, which are known to utilize
dynein. Furthermore, over‐expression of wild type Caly, but not a mutant defective in binding AP‐3,
increased the motility of phosphoinositol 4‐kinase IIα (PI4KIIa), which is targeted to axon terminals by
AP‐3. Altogether, these findings suggest that Caly may function as a cargo adaptor, to stimulate dynein
motility of synaptic vesicles and lysosome‐related organelles.
Peroxisomes move by hitchhiking on early endosomes using the novel adaptor protein PxdA.
J. Salogiannis1, S.L. Reck‐Peterson1,2; 1Cell Biology, Harvard Medical School, Boston, MA, 2Cellular and
Molecular Medicine, University of California San Diego, La Jolla, CA
Eukaryotic cells use microtubule (MT)‐based intracellular transport for the delivery of organelles such as
peroxisomes, endosomes, and mitochondria, and this transport is required for cell growth, metabolism
and maintenance. Perturbations in MT‐based transport are a hallmark of neurodegenerative,
neurodevelopmental, and ciliary diseases. The canonical view of MT‐based transport is that cargo
adaptors link molecular motors to organelles. We report here in the filamentous fungi, Aspergillus
nidulans, a mechanism of peroxisome movement that does not require direct recruitment of a
molecular motor. Instead, long‐range motility of peroxisomes requires hitchhiking onto endosomes via a
novel endosome‐linked tethering protein PxdA. Specifically, PxdA is required for normal distribution and
long‐range motility of peroxisomes, but not other organelles such as endosomes, nuclei, and
autophagosomes. Using simultaneous time‐lapse imaging, we find that endosome‐localized PxdA
colocalizes with motile peroxisomes, but not immotile ones. We also determine that a ~600 amino acid
coiled‐coil region within PxdA located on its C‐terminus is required for its endosome‐localization and its
interaction with peroxisomes, while its N‐terminus is dispensable for both functions. These results
demonstrate that hitchhiking is a novel mechanism used for MT‐based transport of membrane‐bound
Akt stabilizes a Rab11‐WDR44 interaction to regulate pre‐ciliary vesicle trafficking and
ciliogenesis initiation.
V. Walia1, C. Insinna1, Q. Lu1, S. Specht1, Z. Meng1, M. Zhou1, D. Ritt1, D. Morrison1, C.J. Westlake1;
National Cancer Institute, Frederick, MD
Serum starvation is widely used to induce primary cilia assembly in mammalian cells, yet the mechanism
by which serum factor(s) block ciliogenesis is not known beyond the obvious links to cell cycle control.
Our previous work showed that Rabin8, a Rab8 GEF, can be recruited to vesicular Rab11 membranes
within minutes of serum removal to activate Rab8 for ciliary membrane growth. By one hour, the
majority of cells show a shift in Rabin8 recruitment from the cytoplasm to Rab11 membranes marking
one of the earliest steps in ciliogenesis. To investigate how serum withdrawal activates this pre‐ciliary
trafficking step and ciliogenesis, we screened a panel of growth factors and discovered that
lysophosphatidic acid (LPA) inhibits Rabin8 trafficking and ciliogenesis in RPE‐1 cells. LPA blocks
ciliogenesis via signaling through the LPAR1 receptor. A chemical inhibitor screen of the known LPAR1
downstream pathways indentified PI3K/Akt as the most likely signaling pathway regulating these events.
Bioinformatics and proteomics analysis suggested that Akt phosphorylation may regulate binding of
Rab11 associated proteins and thus prevent Rab11‐Rabin8 association in the presence of serum. To test
this theory, we performed a siRNA screen of candidate Rab11 effectors/binding proteins analyzing
Rabin8 pre‐ciliary trafficking in the presence of serum. We observed a strong effect on Rabin8 trafficking
in serum fed cells following depletion of WDR44/Rabphillin‐11 similar to serum starvation. WDR44 was
the first reported Rab11 interacting protein but its Rab11 associated function remains poorly
understood. Interestingly, we show serum‐dependent Akt phosphorylation of Serine‐342 in the Rab11
binding domain of WDR44. Importantly, phospho‐inactive WDR44 showed reduced binding to Rab11.
Expression of the phospho‐active WDR44 S342D mutant, but not S342A strongly blocked ciliogenesis in
RPE‐1 cells and zebrafish embryos. Together these findings uncover a growth factor signaling pathway
mediated by Akt kinase and its novel downstream substrate, WDR44, in the regulation of ciliogenesis
initiation via modulating Rab11‐Rabin8 interaction and pre‐ciliary vesicle trafficking to the centrosome.
A Ras‐like domain in the Light Intermediate Chain bridges the Dynein motor to a cargo‐binding
C.M. Schroeder1, J.M. Ostrem1, N.T. Hertz2, R.D. Vale1; 1Cellular and Molecular Pharmacology , University
of California, San Francisco, San Francisco, CA, 2Laboratory of Brain Development and Repair, The
Rockefeller University, New York, NY
Dynein is a multi‐subunit microtubule‐based motor that is implicated in the transport of many
intracellular cargos, and the dynein light intermediate chain (LIC) is an accessory protein in the dynein
complex that is a critical subunit for binding cargo, particularly membranous organelles. However, little
is known about the mechanism by which the LIC carries out its role in membrane traffic. We have
determined the crystal structure of the conserved N‐terminal domain of a fungal LIC, which revealed a
structure similar to small GTPases (or G proteins). However, despite having a Ras‐like fold, the fungal LIC
did not co‐crystallize with guanosine nucleotide and revealed an unusual binding pocket that excludes
nucleotide. Unlike fungal LIC, the amino acid sequence of human LIC1 contains nucleotide‐binding
motifs found in GTPases. We discovered human LIC1 can bind nucleotide but has a strong preference for
GDP over GTP, which is unusual for G proteins. Collectively, our findings suggest that the LIC evolved
from the GTPase superfamily but no longer acts as a GTPase switch. The LIC may be a member of a new
class of “pseudo‐GTPases,” which have a structure typical of G proteins but cannot hydrolyze GTP. We
further discovered that the Ras‐like domain of the LIC alone binds the dynein heavy chain, and a patch
of aromatic residues, which are universally conserved among all species of LICs, is critical for the LIC
binding to the dynein heavy chain. While the N‐terminal domain of the LIC binds dynein, the divergent
C‐terminal portion of the protein directly binds several Rab effectors involved in membrane transport.
Importantly these Rab effectors were recently shown to also connect dynein to the complex dynactin to
form an ultraprocessive motor. Therefore the LIC may play an important role in selecting only cargo‐
bound dyneins for processive motility along microtubule tracks. Our work provides insight into how the
LIC binds both the dynein complex and cargo, and it also opens questions for the GTPase field and
membrane traffic.
Components of ESCRT‐III and the AAA‐ATPase Vps4 complex are involved in the release of
preperoxisomal vesicles from the ER.
F.D. Mast1,2, R.A. Saleem1,2, L.R. Miller1,2, R.A. Rachubinski3, J.D. Aitchison1,2; 1Center for Infectious
Disease Research, Seattle, WA, 2Institute for Systems Biology, Seattle, WA, 3Cell Biology, University of
Alberta, Edmonton, AB
Peroxisomes are formed by two separate, and possibly complementary, biogenesis pathways: de novo
budding from the ER and growth and division of existing peroxisomes. In cells that lack peroxisomes, de
novo synthesis starts with the recruitment of Pex19, a cytosolic chaperone, to the ER to aid in the
formation of preperoxisomal vesicles (PPVs) that contain peroxisomal membrane proteins. At least two
distinct PPV populations have been characterized and their subsequent maturation into mature
peroxisomes is proposed to be mediated by heterotypic vesicle fusion. However, other than the
requirement for Pex19, and ATP hydrolysis, the mechanisms of PPV formation are unknown.
Furthermore, additional machinery regulating ER‐mediated PPV formation has eluded identification.
Using systems‐level proteomics and multi‐parameter quantitative phenotypic analyses of an arrayed
mutant collection of yeast cells to study peroxisome dynamics, we have identified several protein
complexes with putative roles in the budding of PPVs from the ER. The first, components of ESCRT‐III
and the AAA‐ATPase Vps4 complex (III‐associated), is a cytosolic complex with diverse functions
including, notably, membrane scission during cytokinesis and multivesicular body formation.
Characterization of yeast mutants of ESCRT‐III, and III‐associated, revealed defects in peroxisome
biogenesis and morphology. To investigate ESCRT‐III, and III‐associated, function biochemically, we used
an in vitro PPV budding assay. Cells expressing Pex3‐GFP, a peroxisomal membrane protein, and lacking
Pex19, to trap Pex3 in the ER, were converted to spheroplasts, gently lysed and the ER membranes were
recovered and mixed with cytosol, from wild‐type (WT) or ESCRT deletion mutants, and an ATP‐
regenerating system. Cytosols lacking ESCRT‐III and III‐associated components did not release PPVs from
the ER. We propose that ESCRT‐III, and III‐associated, function as positive regulators of PPV egress. This
represents a novel role for ESCRT‐III, and III‐associated, in regulating the formation of vesicles not only
away from, but also into, the cytosol.
Endolysosomal Deficits Augment Mitochondria Pathology in Spinal Motor Neurons of
Asymptomatic fALS Mice.
Y. Xie1, B. Zhou1, M. Lin1, S. Wang1, Z. Sheng1; 1NINDS, NIH, Rockville, MD
One pathological hallmark in ALS‐linked motor neurons (MNs) is axonal accumulation of damaged
mitochondria, which produce energy and buffer Ca2+ less efficiently, and initiate apoptotic cascades. It
has long been thought that the accumulation of those mitochondria in axons is due to impaired
mitochondrial transport. We previously tested this by genetically crossing fALS‐linked hSOD1G93A mice
and syntaphilin (snph) knockout mice. SNPH acts as an axonal mitochondrial docking receptor; deleting
snphrobustly increases axonal mitochondrial motility (Kang et al., Cell 2008).We found that the two‐fold
increase in axonal mitochondrial motility does not slow ALS‐like disease progression (Zhu and Sheng, JBC
2011), thus challenging the hypothesis that defective mitochondrial transport contributes to rapid‐onset
MN degeneration. These observations raise a fundamental question:does impaired degradation of
damaged mitochondria by the autophagy‐lysosome system play a more pathological role during the
early asymptomatic stage of fALS‐linked mice?
We recently reveal for the first time spinal MN‐targeted progressive lysosomal deficits starting at
asymptomatic stages in fALS‐linked hSOD1G93A mice (Xie et al. Neuron 2015). These deficits impair
autophagic degradation, resulting in aberrant accumulation of autophagic vacuoles engulfing damaged
mitochondria along MN axons. These phenotypes were captured in cultured adult (P40) spinal MNs
from the hSOD1G93A mice. Such early deficits are due to reduced late endosome (LE) retrograde
transport via binding of mutant hSOD1G93A to dynein, and can be reversed by introducing dynein adaptor
snapin transgene. Snapin competes with hSOD1G93A for binding to dynein, thereby recruiting more
dynein to LEs for transport. Thus, snapin and hSOD1G93A play opposite roles in LE retrograde
transport.Expressing snapin efficiently reverseslysosome deficits and facilitates removal of damaged
mitochondria. Injecting AAV9‐snapin into the diseased mice rescues lysosome deficits in vivo and slows
MN degeneration and disease progression.Thus, our study advances our knowledge of early pathological
mechanisms underlying MN degeneration.Elucidating this early pathological mechanism is broadly
relevant, because defective transport, lysosomal deficits, and mitochondrial pathology are all associated
with ALS, Huntington’s, Parkinson’s and Alzheimer’s diseases. Therefore, enhancing clearance of
damaged mitochondria by regulating endolysosomal trafficking may be a potential therapeutic strategy
for ALS and perhaps other neurodegenerative diseases. (Supported by the Intramural Research Program
Xie et al., Endolysosomal Deficits Augment Mitochondria Pathology in Spinal Motor Neurons of
Asymptomatic fALS Mice. Neuron 87, 355–370 (2015).
Microsymposium 10: Applications of Cell Biology in the Real World
Inhibiting endothelium directed tumor cell streaming by targeting the HGF/C‐Met signaling
E. Leung1, A. Xue2, Y. Wang1, P. Rougerie1, V.P. Sharma1,3, R.J. Eddy1, D. Cox1,4, J.S. Condeelis1,3; 1Anatomy
and Structural Biology, Albert Einstein College of Medicine, Bronx, NY, 2Pelham Memorial High School,
Pelham, NY, 3Gruss Lipper Biophotonics Center, Albert Einstein College of Medicine, Bronx, NY,
Department of Developmental Molecular Biology, Albert Einstein College of Medicine, Bronx, NY
In order to metastasize, tumor cells need to migrate towards blood vessels. Streaming, a specialized
mode of migration involving chemotactic signaling between macrophages and tumor cells, allows linear
streams of tumor cells to migrate directionally towards blood vessels on collagen I‐containing ECM fibers
in vivo. We have successfully reconstructed tumor cell streaming in vitro by co‐plating tumors cells,
macrophages, and endothelial cells on 2.5 um thick ECM‐coated micro‐patterned substrates. We found
that tumor cells and macrophages, when plated together on the micro‐patterned substrates, do not
demonstrate directional migration in only one direction but show random walk. Directional streaming
migration of tumor cells in only one direction was established when beads coated with human umbilical
vein endothelial cells (HUVEC) were placed at one end of the micro‐patterned “ECM fibers” within the
assay. Using this in vitro reconstituted streaming system, we found that directional streaming can be
blocked by inhibiting the HGF/C‐Met signaling pathway between endothelial cells and tumor cells, which
led to a marked decrease in endothelium‐directed streaming of the tumor cells. Key observations made
with the in vitro reconstituted system were reproduced in mammary tumors using the in vivo invasion
assay and in intravital observations of tumor cell streaming with C‐Met inhibition. These results highlight
a promising role of C‐met inhibitors in blocking tumor cell streaming and metastasis in vivo and for use
in human trials.
Age‐ and lineage‐dependent gene expression is maintained by microenvironment imposed
epigenetic states in human mammary epithelial cells.
M. MIYANO1, M. Stoiber1, M. Stampfer1, J.B. Brown1, M.A. LaBarge1; 1Life Sciences, Lawrence Berkeley
National Laboratory, Berkeley, CA
Normal healthy tissues show changing patterns of gene expression as a consequence of aging, and there
is a functional cost to these changes that can be relevant to disease pathology. The most obvious age‐
related disease in breast is cancer, with the large majority of newly diagnosed breast cancer occurring in
women over 50. We have shown that breast gene expression changes that occur with age have
functional consequences in the epithelial progenitor and differentiated cells, i.e, a decline of the
myoepithelial lineage, loss of luminal cell specificity, and accumulation of differentiation defective
multipotent progenitor cells. We have hypothesized that these tissue‐level changes make older epithelia
more susceptible to transformation. Because gene expression patterns reflect the wiring and response
diagrams of cells, it is of central importance to understand the origins of age‐related transcriptomes. In
our studies, early passage normal human mammary epithelial cells (HMEC) show age‐dependent
functional and molecular hallmarks consistent with in vivo, suggesting that the underlying gene
expression patterns are metastable. DNA methylation is a stable, but malleable, form of epigenetic
regulation that may underpin these biologically metastable states. Genome‐wide analysis of primary
epithelia was used to identify a set of genes that exhibit age‐ and lineage‐specific expression that was
inversely correlated with promoter CpG methylation. Chemical perturbation of DNA methylation in pre‐
menopausal HMEC resulted in a biochemical phenocopy of more advanced age. Tissue‐mimetic cultures
were used to demonstrate that lineage‐specific gene expression and methylation in luminal cells were
imposed by distinct microenvironments. Optimal maintenance of the luminal phenotype required direct
contact with the apical surface of myoepithelial cells. Mimetic tissues built with HMEC that differed in
chronological donor age revealed that age‐dependent gene expression and methylation patterns are
communicated between the two different lineages, as exposure of pre‐menopausal luminal cells to a
post‐menopausal microenvironment imposed transcriptional patterns in luminal cells consistent with
post‐menopause. These data demonstrate that lineage‐ and age‐dependent phenotypes in HMEC are
maintained by microenvironment‐imposed metastable epigenetic states.
Actin facilitates chromosome capture by microtubules in meiosis of starfish oocytes.
M. Burdyniuk1, M. Mori2, N. Monnier3, P. Lénárt1; 1Cell Biology and Biophysics, European Molecular
Biology Laboratory, Heidelberg, Germany, 2Department of Experimental Genome Research, Osaka
University, Osaka, Japan, 3Department of Biological Engineering, Massachusetts Institute of Technology,
Cambridge, MA
Chromosome capture by microtubules is an early step of cell division essential for alignment and
subsequent separation of sister chromatids. The initially proposed model of ‘search and capture’ by
dynamic astral microtubules (Kirschner and Mitchison, 1986) has been validated by recent studies, and
additionally revealed mechanisms that facilitate capture by biased microtubule nucleation and
stabilization in the proximity of chromatin and kinetochores (reviewed in Pavin and Tolic, 2014). These
mechanisms ensure rapid and efficient capture of chromosomes in somatic cells (Wollman et al, 2005).
However, in specialized cell types such as oocytes with large nucleus chromosomes are located much
further from microtubule asters. In these cells, the models that work in small somatic cells are
insufficient to explain chromosome capture. Recently, our group has shown that in starfish oocyte an
actin‐driven mechanism facilitates chromosome congression and is required to prevent chromosome
loss: contractile actin meshwork transports chromosomes to within the capture range of microtubule
asters of approx. 30 µm (Lenart et al, 2005; Mori et al, 2011).
Here, we investigated whether this actin‐driven mechanism merely functions to transport chromosomes
closer to microtubule asters, or whether there is additional synergy between the actin meshwork and
microtubules. To this end, we performed high spatio‐temporal resolution tracking of chromosome
motion in 3D together with drug‐perturbation experiments. This assay allowed us to identify when and
where individual chromosomes are captured in the presence or absence of the functional actin
meshwork. Our data show that the actin meshwork not only transports chromosomes to the vicinity of
microtubule asters, but also greatly facilitates microtubule ‘search and capture’ to the extent that
chromosome capture appears instantaneous and distance independent. If the actin meshwork is
depolymerized by cytochalasin D, chromosome capture is slower and occurs in the distance‐dependent
manner, as predicted by the ‘search and capture’ model. These observations point to a tight
coordination between actin‐driven transport and microtubule capture of chromosomes in starfish
oocytes. Currently, we are focusing on understanding the mechanisms underlying this coordination
between the cytoskeletal systems and identifying the molecular players involved.
Orthogonal and modular gene regulation using engineered CRISPR/Cas9.
A. Didovyk1, B. Borek1, J. Hasty1, L. Tsimring1; 1BioCircuits Institute, University of California, San Diego, La
Jolla, CA
The promise of synthetic biology lies in predictable and modular design of genetic circuits. Similarly to
electrical engineering the use of modular independent components greatly simplifies circuit design
allowing increased complexity of synthetic circuits necessary to implement complex behavior. Until
recently, the complexity of synthetic genetic circuits has been limited by the number and inflexibility of
available orthogonal transcription factors. Recently CRISPR/Cas9 system has been used to control gene
expression in various organisms. However, this system is limited by the its imperfect DNA sequence
specificity, leading to cross‐talk with the host genome or other circuit components. Furthermore,
CRISPR/Cas9 mediated gene regulation is context dependent, limiting the modularity of CRISPR/Cas9
based transcription factors.
In this work we address the problems of specificity and modularity by developing an approach for
selecting Cas9/gRNA transcription factor/promoter pairs that are maximally orthogonal to each other as
well as to the host genome and synthetic circuit components. We validate the method by designing and
experimentally testing four orthogonal promoter/repressor pairs in the context of a strong promoter PL
from phage lambda. We demonstrate that these promoters can be interfaced by constructing a double
and a triple inverter circuits. Finally, to further address the problem of modularity we propose a scheme
to predictably incorporate the designed orthogonal CRISPR/Cas9 regulation into an important large class
of natural inducible promoters. We validate this approach on three ubiquitously used inducible
promoters PLlacO‐1, PBAD, and PLuxI.
Defective endosome maturation and trafficking in models of macular degeneration.
K.A. Toops1,2, L. Tan1,3, A. Lakkaraju1,2,3; 1Ophthalmology Visual Sciences, University of Wisconsin‐
Madison, Madison, WI, 2McPherson Eye Research Institute, University of Wisconsin‐Madison, Madison,
WI, 3Division of Pharmaceutical Sciences, University of Wisconsin‐Madison, Madison, WI
Abnormalities in the endo‐lysosomal system are emerging as a common theme in neurodegenerative
conditions such as Alzheimer’s disease, Parkinson’s disease and amyotrophic lateral sclerosis. A direct
consequence of endo‐lysosomal defects is the accumulation of cellular debris due to autophagic stress.
We recently reported that autophagy is impaired in the Abca4‐/‐ mouse model of Stargardt macular
degeneration, raising the possibility that endo‐lysosomal defects could worsen autophagic stress and
contribute to vision loss in this disease. Here, we used high‐speed live imaging and 4‐D image analysis to
investigate endosome maturation and trafficking in polarized monolayers of the retinal pigment
epithelium (RPE). The RPE, which sits beneath the light‐sensing photoreceptors in the eye, performs
several essential functions indispensable for healthy vision. The post‐mitotic RPE is also the initial site of
damage in both inherited and age‐related macular degenerations. A primary function of the RPE is the
daily phagocytosis and digestion of shed photoreceptor outer segment tips to counteract the effects of
light damage and oxidative stress. With age, undegraded outer segment components including vitamin
A metabolites accumulate within RPE lysosomes in the form of lipofuscin bisretinoids. Bisretinoids are
wedge‐shaped molecules that trap cholesterol within RPE lysosomes. Our data show that in primary RPE
monolayers and in RPE from Abca4‐/‐ mice, excess cholesterol, which accumulates secondary to
bisretinoids, interferes with the acquisition of active Rab7‐GTP by maturing early endosomes. This leads
to abnormally enlarged Rab5‐positive early endosomes and fewer Rab7‐positive late endosomes.
Paradoxically, bisretinoids paralyze late endosome trafficking but increase long‐range lysosome
transport in the RPE. Excess lysosomal cholesterol reroutes the p150glued subunit of the dynein‐dynactin
complex from late endosomes to lysosomes via interactions mediated by the Rab7 effector RILP and the
cholesterol sensor ORP1L. As a result, lysosomes associate more, and late endosomes associate less,
with microtubules, leading to opposing effects of bisretinoids on organelle traffic. Intra‐peritoneal
administration of the liver X receptor (LXR) agonist TO901317 decreases cholesterol and restores
endosome maturation in the RPE of Abca4‐/‐mice. Our studies suggest that endo‐lysosomal defects occur
early in the cascade that eventually causes RPE dysfunction. Therefore, regulating endosome maturation
and trafficking could be a potential therapeutic approach for preventing vision loss in retinal
Drug metabolism and clearance system in normal and malignant plasma cells: Relationship to
clinical outcome.
W. HASSEN1,2, A. KASSAMBARA1,3, B. KLEIN1,3, J. Moreaux1,3; 1Suivi des Therapies Nouvelles, CNRS
UPR1142, Institute of Human Genetics; Montpellier, 34090, France, Montpellier, France, 2Cellular
Biology and Physiology, High Institute of Biotechnology of Monastir, Monastir, Tunisia, 3Department of
Biological Haematology, Laboratory for Monitoring Innovative Therapies, Montpellier, France
Chemical‐host interaction plays an important role in the etiology and pathogenesis of many diseases
including cancer. Drug metabolism and clearance (DMC) system is a defense system that sensors,
inactivates and excretes chemicals from a variety of sources including dietary components and drugs
that can profoundly impair the structure and function of cells and tissues. Drug sensing is mediated by
members of the superfamily of nuclear receptors (PXR, CAR, LXR and FXR) and some cytosolic
transcription factors. After xenobiotic binding, xenobiotic receptors translocate to the nucleus and
govern the tandem expression of genes encoding for phase I and II Drug Metabolizing Enzymes (DMEs)
and transporters. DMC system proceeds then through enzymatic conversion of xenobiotics into more
water‐soluble metabolites that are better efluxed from the cell through membrane transporters (ATP
binding cassette (ABC) transporters or Solute Carriers (SLC)) and discharged into urinary and biliary
systems. The DMC is a major feature of the hepatocyte function and at lesser extent other cell types.
Unsurprisingly, DMC system has been demonstrated to compromise the efficacy of cancer
chemotherapy and lead to treatment failure through promoting the metabolism and the elimination of
chemotherapeutic agents.
In the present study, we have interrogated normal plasma cells and Multiple Myeloma cells (MMC) of
newly‐diagnosed patients for the expression of 350 genes encoding for uptake carriers, xenobiotic
receptors, phase I and II Drug DMEs and efflux transporters in relation to their clinical outcome (relapse
and survival). Interestingly, both normal plasma cells and malignant plasma cells with good prognosis
(Event free survival and overall survival) exhibits high expression of large number of influx transporters
and phase I/II DMEs driven by high expression of xenobiotic receptors (RXRα LXR, CAR and FXR) with
110 out 350 genes upregulated with a fold Change>2. On the contrary, MMC of patients with
unfavourable outcome displayed a global down regulation of DMC genes with only 14 genes
upregulated out of 350 including several members of ABC transporters. This study shed light for the first
time on the existence of high drug metabolism capacities within immunological cells from B lineage.
More importantly, our results pointed out a significant correlation between DMC and cancer drug
resistance and treatment failure. Further investigations are needed to unveil the role of possible high
drug metabolism capacities in normal plasma cells and to better understand their unexpected
association to a better prognosis.
Altered trafficking of connexin 43 participates to the development of ventricular arrhythmias in
cardiomyopathy caused by mutations in A‐type lamins gene.
C. Macquart1, C. Le Dour2, M. Chatzifrangkeskou1, H.J. Worman2,3, G. Bonne1, A. Muchir1; 1Center of
Research in Myology, UPMC ‐ Inserm UMRS 974, CNRS FRE3617, Paris, France, 2Department of
Medicine, Columbia University, New York, NY, 3Department of Pathology and Cell Biology, Columbia
University, New York, NY
Compared with other forms of dilated cardiomyopathy, mutations in LMNA encoding nuclear A‐type
lamins are responsible for a more aggressive clinical course due to a high rate of malignant ventricular
arrhythmias. A better understanding of factors and mechanisms that drive ventricular arrhythmias is
crucial to generation of potential therapies. Inter‐cellular communication is essential for proper cardiac
function. Mechanical and electrical activities must synchronize so that the work of individual
cardiomyocytes transforms into the pumping function of the heart. Gap junctions are specialized cell‐
cell junctions that mediate inter‐cellular communication. They are composed of connexin proteins,
which form transmembrane channels for small molecules. Here, we showed that altered distribution of
connexin 43 occurs prior to any electrical disturbances in a mouse model of dilated cardiomyopathy due
to LMNA mutations. We next assessed in vitro the molecular mechanisms of connexin 43 re‐localization
in pathological context. We showed that the presence of LMNA mutations triggers an abnormal
trafficking of connexin 43 along both microtubules and actin networks leading to a loss of cell‐cell
communication. Going further, we demonstrated that modulating this process could restore the correct
localization and function of connexin 43 at the cell‐cell junction in cardiomyocytes carrying LMNA
mutation. Our work could break new ground for future work towards developing novel treatment for
malignant arrhythmias.
Microsymposium 11: Nucleus Biology and Disease
A role for the nucleus in sarcomere assembly.
A.L. Auld1, E.S. Folker1; 1Department of Biology , Boston College, Chestnut Hill, MA
The syncytial nature of the myofiber makes it a powerful system for the elucidation of both the
mechanisms and functions of nuclear movement. During normal muscle development, nuclei undergo a
series of complex movements to achieve a precise distribution pattern in which most nuclei are spaced
to maximize the distance between neighbors. Although the mechanisms of these movements are
emerging, it is not clear why nuclei are being moved and precisely positioned. To address this question,
we have used a combination of genetics and in vivo time‐lapse microscopy to investigate what aspects
of muscle development rely on the nuclei. These experiments demonstrated a role for the nucleus in
sarcomere assembly. We found that muscle nuclei are fully positioned prior to sarcomere formation
and that Zasp‐GFP, a Z‐line component, localizes to nuclei prior to myofibril assembly with the earliest
punctate accumulations of Zasp‐GFP colocalizing with nuclei. Higher resolution imaging of the puncta
showed that they were linear extensions emanating from the nucleus. These extensions were also found
to contain F‐actin suggesting that muscle nuclei may act as a nucleating center for latent sarcomeric
complexes. The relationship between latent sarcomeric complexes and nuclei was transient with muscle
nuclei nucleating Zasp assemblies prior to moving on and nucleating new Zasp assemblies. These latent
sarcomeric structures then developed into the repetitive Z‐line structure that is characteristic of the
sarcomere. Genetic manipulation of embryos that result in mispositioned nuclei did not affect the ability
of the nucleus to recruit or assemble Zasp‐GFP, but did therefore alter the position of the earliest
sarcomeres. Conversely, two nuclear envelope proteins, Klarsicht (Nesprin) and Klaroid (Koi) were
necessary for the recruitment and assembly of Zasp‐GFP. This indicates that there is a molecular
recruitment of Zasp‐GFP to the nucleus and implicates the nuclear envelope as a critical regulator of
early sarcomere assembly. Finally, both blocking the ability of the nucleus to recruit and assemble
sarcomere proteins and causing the nuclei to assemble sarcomeres in the wrong location result in
weaker muscles that often tear or completely lack a contractile network. Thus, the nucleus is crucial for
the assembly of a stable myofibril network.
Myonuclear position is regulated by different mechanisms during muscle development and
muscle growth.
M.A. Collins1, T. Mandigo1, G. Vazquez1, E.S. Folker1; 1Department of Biology, Boston College, Chestnut
Hill, MA
Muscle cells are a syncytial cell type conserved from Drosophila to humans. In these cells, nuclei are
positioned at the cell periphery to maximize the distance between adjacent nuclei. The importance of
this feature of muscle cells is highlighted by the correlation between mispositioned nuclei and muscle
disease. However, it remains unclear why nuclei are mispositioned, and whether nuclear positioning
contributes to muscle weakness and wasting. As a first step toward understanding the relationship
between the position of the nucleus and muscle disease, we examined whether mispositioned nuclei in
two different diseases, Centronuclear Myopathy (CNM) and Emery‐Dreifuss Muscular Dystrophy
(EDMD), result from a common mechanism. Although both diseases presented an effect in Drosophila
embryos and larvae, the resulting disease phenotypes differ drastically. Two CNM‐linked genes,
Amphiphysin and Myotubularin which have known roles in membrane organization, presented a subtle
effect in embryos and larvae. Nuclei were positioned too close to the MTJ with some nuclei found in the
center of the muscle. In contrast, two of the gene linked to EDMD, all of which encode for nuclear
envelope proteins, presented a distinct phenotype. Nuclei failed to separate into two distinct clusters in
embryos with mutant Nesprin (Klarsicht) and Emerin (Bocksbeutel) genes. Additionally, nuclei in larvae
were found either misaligned or clumped together. These data suggest that although the feature of
mispositioned nuclei is common to both diseases, the underlying mechanism that drives this phenotype
is not similar.
Turnover of proteins at the nuclear periphery.
A.L. Buchwalter1, M.W. Hetzer1; 1Molecular and Cell Biology Laboratory, The Salk Institute for Biological
Studies, La Jolla, CA
The nuclear periphery is defined by the nuclear lamina and the proteins of the inner nuclear membrane
(INM). In cooperation with INM proteins, the lamina connects the nucleus to the cytoskeleton. The
lamina and INM proteins also scaffold organization of the genome by recruiting and repressing specific
genomic loci at the nuclear periphery. Proliferating cells have ample opportunity to remodel the nuclear
periphery through the processes of nuclear envelope breakdown and reassembly. It is unknown
whether or how the nuclear periphery is remodeled while maintaining nuclear integrity in non‐dividing
cells. The importance of this question is underscored by the fact that disrupting protein homeostasis at
the nuclear periphery is associated with at least 15 separate human diseases, referred to as
“laminopathies”. To understand protein homeostasis at the nuclear envelope, we used stable isotope
labeling in cell culture (SILAC) and quantitative mass spectrometry to define protein turnover rates in
non‐dividing cells. C2C12 mouse myotubes were used as an in vitro model of differentiation. Our data
indicate that unlike nuclear pore complexes (NPCs) and histones, lamins and INM proteins exchange
within intact nuclei. This implies that lamin‐templated processes may be dynamic in long‐lived cell types.
A‐type lamins turn over just as efficiently as their cytoplasmic relatives vimentin and desmin, while B‐
type lamins exchange more slowly. INM proteins that cooperate in cytoskeleton‐associated complexes
turn over at similar rates to each other. INM proteins generally turn over at similar rates to
transmembrane proteins of the endoplasmic reticulum (ER), indicating that removal from the INM is not
a rate‐limiting step in degradation of these proteins. Further, there is no correlation between INM
protein size and turnover rate, which suggests that transport of nucleoplasmic domains through the NPC
also does not limit INM protein turnover. To determine the mechanism of INM protein turnover, we
screened disease‐linked mutations of INM proteins with the goal of identifying a mutant INM protein
that is rapidly degraded. We identified a small in‐frame deletion in emerin that is linked to Emery‐
Dreifuss muscular dystrophy (EDMD) and is normally targeted to the INM, but is characteristically
expressed at low levels. We found that this EDMD‐linked mutant is rapidly degraded, and that
degradation is inhibited by blocking activity of the proteasome or the ER‐associated degradation (ERAD)
enzyme p97. Altogether, these data suggest that INM proteins are subject to ERAD and that a disease‐
linked INM protein mutant is a viable model substrate for dissecting the INM‐ERAD pathway.
Structural Organization of the Nuclear Lamin Isoforms in the Nuclear Lamina Revealed by Super
Resolution Microscopy.
M. Kittisopikul1, T. Shimi2,3, J. Tran4, A.E. Goldman3, S.A. Adam3, Y. Zheng4, R.D. Goldman3, K. Jaqaman1;
Biophysics, UT Southwestern, Dallas, TX, 2Human Genetics, University of Chicago, Chicago, IL, 3Cell and
Molecular Biology, Northwestern University, Feinberg School of Medicine, Chicago, IL, 4Embryology,
Carnegie Institution for Science, Baltimore, MD
The nuclear lamina is a key structural element of the metazoan nucleus. However, the structural
organization of the major proteins composing the lamina remains poorly defined. Using three‐
dimensional Structured Illumination Microscopy and computer vision algorithms, we have characterized
the structures of the four lamin isoforms in mouse embryo fibroblast nuclei. Each isoform forms a
distinct fiber meshwork, having comparable physical characteristics with respect to mesh edge length,
mesh face area and shape, and numbers of edges per face. In this study, we quantitatively compare
each isoform according to the distribution of these meshwork properties. In fibroblasts null for the
expression of either lamins A/C or lamin B1, we show that the meshwork enlarges in both cases but
does so in a different manner. Those lacking lamins A/C expand the number of edges per face while
those lacking lamin B1 extend the edge lengths. In contrast, the fibroblasts lacking lamin B2, show little
enlargement. These studies demonstrate that individual lamin isoforms assemble complex
interdependent networks within the nuclear lamina and that each has a distinct role in the organization
of the nuclear lamina that we can quantitatively measure.
Brightness Characterization Of Nuclear Envelope Proteins By Z‐Scan Fluorescence Fluctuation
J. Hennen1, C.A. Saunders2, E.M. Smith1, J.D. Mueller1, G. Luxton2; 1Physics and Astronomy, University of
Minnesota, Minneapolis, MN, 2Genetics, Cell Biology, and Development, University of Minnesota,
Minneapolis, MN
Traditionally, fluorescence fluctuation spectroscopy (FFS) has been used to quantify the stoichiometry of
soluble proteins in the nucleus and the cytoplasm of mammalian cells by brightness analysis. The
development of z‐scan FFS broadened the capability of brightness analysis to include proteins
distributed across stratified layers, such as the cytoplasm and the plasma membrane. In this work, we
further extend z‐scan FFS to study proteins that reside on or within endomembranes, including the
endoplasmic reticulum (ER) membrane/lumen and the nuclear envelope/perinuclear space.
Experimentally, we place the 20 amino acid ER signal sequence (SS) from the lumenal AAA+ ATPase
torsinA in front of EGFP (SS‐EGFP); the complex is translated into the lumen of the ER where of the
signal sequence is cleaved, leaving EGFP to diffuse within the ER and nuclear envelope. The brightness of
SS‐EGFP is determined by performing z‐scan FFS measurements where corrections are applied for both
the thin layer geometry and coexcitation of adjacent layers. We then generated a tandem dimeric EGFP
protein (SS‐EGPF2) to establish a model for calibrating brightness and stoichiometry. Finally, we tested
the limits of our technique and applied z‐scan FFS to study the assembly of the inner nuclear membrane
proteins SUN1 and SUN2 within the perinuclear space. SUN proteins are type II membrane proteins that
interact with A‐type nuclear lamins and chromatin within the nucleoplasm and with the cytoskeletal‐
interacting outer nuclear membrane nesprin proteins within the perinuclear space. Together, SUN and
nesprin proteins form the linker of nucleoskeleton and cytoskeleton (LINC) complex, which is critical for
nuclear positioning and movement as well as chromosome dynamics. The crystal structure of SUN2
reveals a mushroom‐like trimer with a cap composed of SUN domains and a stalk of three coiled‐coils.
Each SUN trimer binds three nesprin proteins in grooves that are formed between adjacent SUN
domains. Currently, direct evidence for this structure in living cells does not exist nor does structural
information for SUN1. Here, we show that the lumenal regions of both SUN1 and SUN2 oligomerize
within the perinuclear space. SUN2 oligomerizes into trimers while SUN1 oligomerizes up to a tetramer,
potentially through an interaction with the nuclear pore complex. Surprisingly, when we express the
lumenal regions of SUN1 and SUN2 within the cytoplasm, they oligomerize at lower concentrations than
those required in the perinuclear space suggesting that LINC complex assembly is tightly regulated
within this subcellular compartment. Taken together, our work establishes z‐scan FFS as a powerful
method for the biochemical and biophysical analysis of internal endomembrane proteins within living
The nuclear import receptor KPNA7 is critical for neuronal development and function.
L.T. Oostdyk1,2, C. Snow2, K. Geffken2, C. Yang2, A.R. Paciorkowski3, M.J. McConnell1, B.M. Paschal1,2;
Department of Biochemistry and Molecular Genetics, University of Virginia, Charlottesville, VA, 2Center
for Cell Signaling, University of Virginia, Charlottesville, VA, 3Departments of Neurology, Pediatrics and
Biomedical Genetics, University of Rochester Medical Center, Rochester, NY
The karyopherin alpha (KPNA) family of transport receptors facilitates translocation of signal‐containing
protein cargos from the cytoplasm to the nucleus. In humans the KPNA family is encoded by seven
genes, all generating import receptors with highly conserved structures that include 10 armadillo (ARM)
repeats arranged in tandem. Despite the high degree of similarity (45‐85% identity), including a concave
surface that contains both a major and a minor contact surface for nuclear localization signals (NLSs),
there is clear biochemical evidence for protein cargo selectivity amongst the different KPNA receptors.
Moreover, the expression level of individual KPNA genes in different cells and tissues might be relevant
for nuclear pathways that are context‐specific. We recently showed that autosomal recessive mutations
in KPNA7 are associated with infantile spasms and cerebellar malformation, with additional clinical
manifestations that include severe developmental disability and intractable epilepsy (Paciorkowski et al.
2014). Protein modeling with KPNA2 as a template indicates the missense mutations map to the NLS
binding region of KPNA7, and suggests that defective import of protein cargoes in neurons might
underlie the disease phenotypes. To understand the function of KPNA7 in neurons, we have examined
its expression, localization and NLS binding properties. KPNA7 mRNA can be detected in the cerebral
cortex, cerebellum and mid‐brain of adult mouse brain. In dissociated cultures of rat brain, KPNA7
protein is expressed in both cortical and hippocampal neurons based on co‐staining with MAP2. To
determine if KPNA7 expression is related to neuronal development, we used neural progenitor cells
(NPCs) and neurons derived from human induced pluripotent stem cells (hiPSC). Within six weeks of
neuronal differentiation, NPCs develop extensive branching, express neuronal markers including MAP2
and b‐III‐tubulin and show a 4‐fold increase in KPNA7 mRNA levels. Within neurons, KPNA7 was localized
predominantly to the nucleus. To study the biochemical properties of KPNA7, we expressed the protein
in insect cells. By ELISA we determined that KPNA7 binds to the monopartite NLS (GRKRKKKRTS; amino
acids 351‐360) in Brn2/Pou2f, a transcription factor with established roles in neuronal differentiation.
The KPNA7 interaction is highly specific since altering a single amino acid (K356E) in the Brn2 NLS
eliminates binding. Taken together, our data suggest the working hypothesis that mutations in KPNA7
result in disease phenotypes by virtue of defective import of transcription factors required for proper
development and function of neurons.
Microsymposium 12: Signaling in Differentiation and Cancer
Histone deacetylase 1 and 3 inhibition blocks TGFβ‐mediated conversion of tumor endothelial
cells into myofibroblasts and “re‐educates” carcinoma‐activated fibroblas.
D. Kim1, L. Xiao1, C.A. Otey1, M. Troeste2,3, A.C. Dudley1,3,4; 1Department of Cell Biology Physiology,, The
University of North Carolina at Chapel Hill, Chapel Hill, NC, 2Gillings School of Global Public Health, The
University of North Carolina at Chapel Hill, Chapel Hill, NC, 3Lineberger Comprehensive Cancer Center,
The University of North Carolina at Chapel Hill, Chapel Hill, NC, 4McAllister Heart Institute, The University
of North Carolina at Chapel Hill, Chapel Hill, NC
Smooth muscle actin positive (SMA+) myofibroblasts (MFs) are prominent in the tumor
microenvironment (TME) and they promote tumor progression and metastasis. Multiple cell types in the
TME, including epithelial cells, fibroblasts, and tumor endothelial cells (TECs) are converted to MFs by
TGFβ. Because TGFβ deprograms TECs by globally modifying chromatin conformation which promotes
endothelial‐to‐mesenchymal transition (EndMT), we screened a library of epigenetic‐modifying drugs
that might regulate TGFβ‐driven MF differentiation from TEC precursors. From our screen, we narrowed
our focus on histone deacetylase (HDAC) inhibitors that block conversion of TECs into SMA+ MFs and
“re‐educate” tumor‐associated MFs (also called carcinoma activated fibroblasts or CAFs). We show that
Scriptaid, a potent inhibitor of HDACs 1, 3, and 8, represses MF marker genes in TECs and CAFs derived
from human and mouse melanoma and breast carcinoma. Scriptaid strikingly suppresses TGFβ‐mediated
EndMT, blocks collagen contraction, and reduces expression of the MF markers collagen I, fibronectin,
SMA, and palladin. Scriptaid also delays tumor growth in murine orthotopic mammary tumor models
and diminishes the number of collagen I+/SMA+ myofibroblasts in the TME in vivo. Thus, inhibiting
specific HDACs “re‐educates” tumor‐associated MFs into SMA negative fibroblasts, blocks EndMT‐
mediated conversion of endothelial precursors into MFs, and impairs tumor growth in mice. Because
MFs support tumor progression, new therapies that impair MF function or block conversion of cellular
precursors into MFs could arrest cancer growth and spread.
Mediator kinase module as a transducer of oncogenic Wnt/β‐catenin signaling.
A.D. Clark1, M. Oldenbroek1, J.M. Spaeth1, T.G. Boyer1; 1Molecular Medicine, UT Health Sci Center‐San
Antonio, San Antonio, TX
Aberrant Wnt/β‐catenin signaling promotes colorectal cancer (CRC) and other types of malignancies
through unprogrammed changes in gene transcription that drive tumorigenesis. CDK8 is a CRC
oncoprotein whose amplification‐dependent overexpression identifies a significant subset of CRC
patients with poor prognosis. CDK8 kinase activity, along with Cyclin C (CycC), drives tumorigenesis by
stimulating β‐catenin transcriptional activity. Therefore, inhibition of CDK8 kinase function offers a
promising therapeutic approach for CDK8‐overexpressing CRCs. CycC‐CDK8, along with MED12 and
MED13, compose a discrete 4‐subunit “kinase” module within Mediator, a conserved multiprotein
interface between gene‐specific transcription factors and RNA Polymerase II. We have previously
identified a network of physical and functional interactions within the Mediator kinase module critical
for oncogenic Wnt/β‐catenin signaling. Mechanistically, β‐catenin binds directly to MED12 in Mediator,
thus activating CycC‐dependent CDK8 kinase activity and β‐catenin transcriptional activity. More
specifically, MED12‐dependent CDK8 activation occurs through a direct interaction involving the MED12
N‐terminus and a phylogenetically conserved surface groove on CycC. Here, we demonstrate that
mutagenic disruption of the MED12/CycC interface in CRC cell lines leads to uncoupling of CycC‐CDK8 to
MED12 and core Mediator, and concomitant loss of Mediator‐associated CDK8 activity. In addition, gene
expression and functional analyses of cells lacking CDK8 kinase function revealed downregulation of
Wnt/β‐catenin signaling, which consequently impairs CRC cell proliferation and clonogenicity. Our
studies therefore identify the MED12/CycC interface as a critical transducer of oncogenic Wnt/β‐catenin
signaling. By validating MED12/CycC as a potential therapeutic target, our findings have significant
implications in current methods of treating CRC as they pave a way for development of novel, targeted
inhibitors of Wnt/β‐catenin signaling to treat colorectal and other CDK8‐driven malignancies.
Sperm TRP‐3 channel mediates the timely onset of the fertilization calcium wave in the
nematode C. elegans.
J. Takayama1, S. Onami1,2; 1Lab. for Developmental Dynamics, RIKEN QBiC, Kobe, Japan, 2NBDC, JST,
Tokyo, Japan
Fertilization calcium waves are a trigger for embryonic development. The calcium waves in most animal
species are generated by egg calcium channels, which are activated by sperm factors introduced upon
sperm–egg fusion. However, the involvement of sperm calcium channels has been unclear. Here, by
using high‐speed in vivo calcium imaging, we show that a sperm‐specific calcium channel TRP‐3
mediates the onset of the fertilization calcium wave in the nematode C. elegans. We visualized the
fertilization calcium response by using spinning disk confocal microscopy and fluorescent chemical
calcium indicator. Subsequent image processing on the time‐lapse microscopy images revealed that the
fertilization calcium response was composed of a rapid local calcium rise upon sperm entry near the
sperm entry point and a following global traveling wave. Mutants defective in sperm entry such as spe‐9
and spe‐42 showed no calcium responses. Mutants for trp‐3 gene, which encodes a sperm‐specific
plasma membrane calcium‐permeable channel, showed no local calcium rise and delay in the onset of
the global calcium wave after sperm entry. Rescue experiments using several sperm‐specific
promoter/drivers demonstrated that the stronger the driver activity for TRP‐3 expression, the larger the
local calcium rise became. Moreover, the larger the local calcium rise was, the earlier the onset of the
global calcium wave became. This correlation was explained by a simple numerical simulation on the
basis of Nagumo equation, which assumes calcium‐induced calcium release machinery in the fertilized
oocyte. To visualize the localization dynamics of TRP‐3 protein during fertilization, we generated a
transgenic strain expressing TRP‐3 fused C‐terminally with TagRFP‐T (TRP‐3::TagRFP‐T) in sperm. By
visualizing the fertilization between TRP‐3::TagRFP‐T‐expressing sperm and the oocyte whose plasma
membrane was labeled with GFP::PH, we found that C. elegans fertilization took place as a direct plasma
membrane fusion and that TRP‐3 was transferred from sperm to the oocyte upon fusion. The calcium
concentration in the cytoplasm of mature trp‐3 mutant sperm did not differ from that of the wild type.
Furthermore, high‐speed calcium imaging suggested that the calcium concentration in the fused sperm
cytoplasm increased after the sperm–oocyte fusion. Taken together, these results suggest that the
sperm plasma membrane channel TRP‐3 induces a local calcium rise and mediates the timely onset of
the global calcium wave.
Discrete and Continuous Cell States Revealed by Single Cell Sequencing.
G. Stanley1, O. Gokce2, B. Treutlein3, S. Quake3,4, T. Sudhof2,4; 1Biophysics, Stanford, Stanford, CA,
Molecular and Cellular Physiology, Stanford, Stanford, CA, 3Bioengineering, Stanford, Stanford, CA,
Howard Hughes Medical Institute, Stanford, CA
The hallmark of multicellular biology is the division of labor: different cell types perform different jobs.
The textbook picture of these cell types is that they are discrete and that a given cell has a clearly‐
defined role. Cell types are typically distinguished by their function and gene expression pattern.
However, cell identification techniques like FISH or FACS measure the expression of a small number of
genes, and it has remained unclear just how distinct different cell types are. Single‐cell RNA sequencing
measures the expression of thousands of genes and provides a much more complete picture of cell
state. In this study we have sequenced the RNA of around one thousand single cells from the adult
mouse striatum using microfluidic capture and whole‐transcriptome amplification. We performed tSNE
dimensionality reduction, which revealed 10 discrete cell types, including one cluster of 361 D1 and D2
neurons. We discovered that striatal neurons can be separated into both discrete and continuous states,
defined by two orthogonal sets of genes respectively. The continuous state previously has been
undescribed in this system, and may represent a novel functional organization of striatal neurons. These
results also raise a fundamental question about the division of labor in animal tissues: when is cell
function discrete and when is cell function continuous?
Identification of Ran Binding Protein 6 (RanBP6) as a Novel EGFR Regulator that is Frequently
Silenced in Cancer.
W. Hsieh1,2, B. Oldrini1,3, H. Erdument‐Bromage4, M. Squatrito3, I.K. Mellinghoff1,5,6; 1Human Oncology
and Pathogenesis Program, Memorial Sloan Kettering Cancer Center, New York, NY, 2Pharmacology,
Weill Cornell Graduate School of Medical Sciences, New York, NY, 3Seve Ballesteros Foundation Brain
Tumor Group, F‐BBVA Cancer Cell Biology Program, National Cancer Research Centre , Madrid, Spain,
Molecular Biology Program, Memorial Sloan Kettering Cancer Center, New York, NY, 5Pharmacology,
Weill Cornell Medical College, New York, NY, 6Neurology, Memorial Sloan Kettering Cancer Center, New
York, NY
Resistance to EGFR kinase inhibitors in glioblastoma (GBM) has been associated with loss of PTEN
expression. We recently found that PTEN loss raises EGFR levels, in part by interfering with CBL‐
mediated EGFR degradation. To better characterize the effects of PTEN on EGFR, we characterized the
EGFR interactome by EGFR immunoaffinity purification and LC‐MS/MS in cancer cells with and without
PTEN knockdown. We identified RanBP6 as a previously not described EGFR‐interacting protein that only
bound EGFR in PTEN positive cells. Activation of PI3K by EGF impaired the interaction between RanBP6
and EGFR in PTEN expressing cells whereas the PI3K inhibitor BKM120 restored the interaction in PTEN
deficient cells. Further studies of the effect of RanBP6 on EGFR revealed that RanBP6 depletion by
shRNA or CRISPR/Cas9‐mediated gene silencing resulted in increased EGFR mRNA levels and
upregulation of EGFR promoter activity. Consistent with a model of negative EGFR regulation by RanBP6,
we observed an inverse correlation between RanBP6 and EGFR mRNA levels in a panel of human cancer
cell lines (Cancer Cell Line Encyclopedia). To determine the molecular basis of EGFR transcriptional
regulation by RanBP6, we examined RanBP6 interacting proteins using affinity purified V5‐tagged
RanBP6. We observed interactions of RanBP6 with multiple proteins involved in nucleo‐cytoplasmic
transport, including importin alpha/beta, RanGAP, Ran, RCC1, and Nup 93. Based on these findings, we
examined the effects of RanBP6 depletion on the nuclear localization of several transcription factors and
found a decrease in nuclear STAT3 in RanBP6 knockout cells. We observed that STAT3 acts as repressor
of EGFR transcription and that pharmacological blockade of STAT3 using the JAK inhibitor, ruxolitinib,
was sufficient to antagonize the effects of RanBP6 depletion on EGFR. Since RanBP6 is frequently
silenced in cancer, typically by gene copy loss together with its genomic neighbor CDKN2A/B, we also
examined its potential role as tumor suppressor. We found that reconstitution of RanBP6 in RanBP6‐low
GBM tumor sphere lines reduced colony formation in soft agar and that RanBP6 knockdown reduced
survival in an orthotopic glioma model driven by PDGF‐B‐RCAS/TVA. In summary, we have identified
RanBP6 as novel EGFR regulator and candidate tumor suppressor in multiple human cancers.
A novel GSK3‐regulated APC:Axin interaction regulates Wnt signaling by driving a catalytic cycle
of efficient βcatenin destruction.
M. Pronobis1, M. Peifer1,2, N.M. Rusan3; 1Curriculum in Genetics and Molecular Biology, University of
North Carolina at Chapel Hill, Chape Hill, NC, 2Biology , University of North Carolina at Chapel Hill, Chapel
Hill, NC, 3National Heart Blood and Lung Institute, National Institute of Health, Bethesda, MD
Adenomatous polyposis coli (APC) is a key negative regulator of Wnt signaling, a powerful pathway
regulating both development and oncogenesis. APC acts in the destruction complex with the scaffold
Axin and the kinases GSK3 and CK1 to target the Wnt effector and transcriptional co‐activator βcatenin
for ubiquitination and subsequent proteasomal destruction. Mutations in APC lead to constitutive active
Wnt signaling and thus contribute to cancer initiation and progression. Despite 20 years of research,
APC’s mechanistic function as negative regulator of Wnt signaling remains mysterious. Most models of
the destruction complex portray it as a static entity that binds, phosphorylates and hands off βcat to the
E3‐ligase. Our work prompted us to consider an alternative hypothesis, in which dynamic assembly and
conformational change within the destruction complex are key to its function in maintaining low βcat
levels. We used FRAP, super‐resolution microscopy, functional tests in mammalian cells and flies, and
other approaches to define APC’s mechanistic role in the active destruction complex. Our study provides
the first glimpses of the internal structure of the APC:Axin complex, revealing that Axin alone forms
strands and sheets, while addition of APC stimulates Axin multimerization, with APC strands intertwined
with those of Axin in a larger destruction complex. Our FRAP and biochemical data reveal that Axin and
APC interact in a complex and regulated way. Based on our data we found that APC plays two roles
inside the destruction complex: (1) APC promotes efficient Axin multimerization through one known and
one novel APC:Axin interaction site, and (2) GSK3 acts through putative phosphorylation sites in APC
motifs R2 and B to regulate the second of these APC:Axin interactions, thus triggering a conformational
change that promotes high‐throughput of βcatenin to destruction. These data support a new dynamic
model of how the destruction complex regulates Wnt signaling and how this goes wrong in cancer, and
more broadly may provide insights into the assembly and dynamics of other large multiprotein
complexes that assemble via dynamic multivalent interactions involving motifs within intrinsically
disordered regions.
Deciphering Noncanonical Fzd2 Signaling in Cancer Metastasis.
T. Gujral1, M.W. Kirschner1; 1Systems Biology, Harvard Medical School, Boston, MA
Metastasis is responsible for as much as 90% of cancer‐associated mortality; yet progress has remained
slow in developing effective drugs either specifically targeting metastasis or targeting cells with
metastatic potential. To date, surgical resection is considered the best treatment for most cancers;
however, it provides limited benefit in metastatic cancers. Although there are numerous studies
showing the migratory potential of metastatic cells and relating metastasis to the biology of the
epithelial‐mesenchymal transition (EMT), little is known about how tumor cells engage this fundamental
cellular program. We have recently discovered a new non‐canonical Wnt pathway through Wnt5–Fzd2
that drives EMT, cellular migration, invasion in vitro and tumor growth and metastasis in vivo. Using
focused proteomics approaches, we have identified several novel signaling proteins downstream of
Fzd2, including Fyn, Stat3, Mek‐Erk‐S6k, and Pyk2 kinase. Our new findings highlight critical nodes in the
Fzd2 pathway that are linked to metastatic phenotypes such as increased motility, and EMT. Overall,
these findings establish the driver role of Fzd2 signaling in diverse solid tumors and make Fzd2 a
potential therapeutic target for metastatic cancers.
Minisymposium 07: Centrosomes and Spindles
Proper organization of the interphase centrosome structure through the coordinated activities
of Centrosomin and Pericentrin‐like protein is essential for viability.
D.A. Lerit1, H.A. Jordan1, J.S. Poulton2, C.J. Fagerstrom1, G.J. Brian1, M. Peifer2, N.M. Rusan1; 1Cell Biology
and Physiology Center, National Heart, Lung, and Blood Institute, National Institutes of Health,
Bethesda, MD, 2Biology Department, University of North Carolina at Chapel Hill, Chapel Hill, NC
Centrosomes are the microtubule‐organizing centers (MTOCs) of most eukaryotic cells and serve critical
functions in maintaining cellular organization and genome stability. Dynamic changes to pericentriolar
material (PCM) levels and organization throughout the cell cycle define centrosome function and
activity. Much of our understanding of centrosome activity comes from studies that focus on
centrosome maturation, the rapid recruitment of PCM to the centrosome as cells enter mitosis. Far less
is known about how centrosome function is regulated during interphase. Numerous studies show
centrosomes are essential during the abridged mitotic cycles that mark the early development of many
animals. In syncytial Drosophila embryos, the centrosome undergoes massive structural rearrangements
with each cell cycle. However, mechanisms regulating this centrosome reorganization – and its
physiological importance – remain unclear. In this work, we show Centrosomin (Cnn) and Pericentrin‐
like protein (PLP) form interphase‐specific extensions, or flares, that define the outer edge of the
interphase centrosome structure. Dual‐color live imaging reveals Cnn and PLP are packaged together at
the tips of interphase flares and released into dynamic cytoplasmic particles. Intriguingly, mutations in
either of the human orthologs of Cnn and PLP result in microcephalic disorders. Our genetic analysis in
Drosophila demonstrates a functional relationship between Cnn and PLP in interphase centrosome
organization and allows us to examine the physiological relevance of interphase centrosomes in
development. Live imaging and super‐resolution microscopy reveal PLP is required to scaffold both Cnn
and γ‐Tubulin at interphase centrosomes, as loss of PLP results in massive PCM fragmentation and
dispersal. Further, we demonstrate that PLP is required for normal centrosome segregation, MT
organization, and embryo compartmentalization. Similar to Cnn, PLP is required for viability, as plp
mutant embryos fail to develop into adults. Moreover, our data indicate Cnn and PLP interact directly at
two defined domains, indicating Cnn and PLP form a complex. Systematically disrupting the Cnn‐PLP
complex prevents the cell cycle‐dependent centrosome reorganization and impairs PCM structure.
Collectively, our data support a model whereby proper organization of the interphase centrosome is
mediated by a Cnn‐PLP scaffold that is required for centrosome segregation, MT organization, genome
stability, and normal development.
A splice variant of Centrosomin converts mitochondria to MTOCs to facilitate sperm
morphogenesis in Drosophila.
J.V. Chen1, T.L. Megraw1; 1Biomedical Sciences, Florida State University, Tallahassee, FL
Mitochondria are energy centers in cells. In Drosophila spermatids, however, they also play an
important role in sperm morphogenesis by providing a structural platform for microtubule (MT)
organization to support the elongating tail. Centrosomin (Cnn) is a core centrosomal protein whose gene
expresses several variants falling into two major forms: the centrosomal form (CnnC) and a non‐
centrosomal form in testes (CnnT). CnnC is essential for functional centrosomes, the major
Microtubule‐Organizing Centers (MTOCs) in animal cells. We found that CnnT is expressed exclusively in
testes, and unlike CnnC that resides at centrosomes, it localizes to nebenkerns (giant mitochondria) in
spermatids. In cell culture, CnnT targets to the mitochondrial surface and recruits the MT‐nucleating
complex γ‐TuRC to assemble MTs on mitochondria, converting mitochondria to MTOCs. We have
mapped two separate domains on CnnT that are necessary and sufficient to target it to mitochondria,
and to recruit the γ‐TuRC and nucleate MTs, respectively. Disrupting the conserved CM1 domain in CnnT
(which is also shared with CnnC) abolishes the MT‐nucleating function but does not block γ‐TuRC
recruitment, indicating that CM1 is essential for the activation but not recruitment of γ‐TuRCs. In vivo,
CnnT forms speckles on the surface of nebenkerns in spermatids, where it is required to recruit γ‐
Tubulin. CnnT mutant males have significantly smaller seminal vesicles, shorter mature sperm tails, and
reduced fertility. We propose that CnnT assembles unique MTOCs on nebenkerns to facilitate the
morphogenesis of the extremely long sperm that are found in Drosophila.
A microscope adaptation that allows high‐speed 3D imaging from a single plane of focus.
C.J. Cogswell1, J. Yu1, R.N. Zahreddine1, S. Chen1, J. Xing1, R.H. Cormack1, J.S. Tyler2, M. Winey2;
Electrical, Computer and Energy Engineering, University of Colorado, Boulder, CO, 2Molecular, Cellular
and Developmental Biology, University of Colorado, Boulder, CO
The goal of our work is to demonstrate how the mainstream microscopes found in nearly every cell or
developmental biology lab can be easily adapted to perform high resolution 3D imaging that greatly
exceeds the acquisition speed and 3D localization accuracy of confocal and other 3D fluorescence
microscopes. By engineering a small number of add‐on optical components, we have produced a new
prototype system that overcomes the limited depth of field of high NA microscope objectives and
collects all photons from an extended sample volume into a single image plane, doing away with the
need to change the microscope focus (or build up an image stack). As our recent results will
demonstrate, this makes it possible to acquire high‐speed images of yeast centrosome segregation with
minimal photobleaching and phototoxicity. It also makes possible the precise localization and tracking of
moving particles and cell components throughout a cell volume so that 3D videos can be produced for
further analysis.
The proposed new system is based on our novel approach to optical design, called expanded point
information content (EPIC) and is built on the premise that it is possible to encode much more
information about the original three‐dimensional sample into standard microscope images and then
easily retrieve this information using custom signal processing algorithms. Implementation of the new
design requires only a few modifications to a standard fluorescence microscope (inserting a phase plate
and installing a software package). The result is the new EPIC system is able to record high resolution 3D
images of dynamic cell structures at speeds an order of magnitude faster than existing microscopes and
simultaneously locate features to an accuracy of 20nm or better, without the need to change the original
In addition to imaging point‐like fluorescent structures, the microscope can be adapted to acquire
volumetric information from extended objects and present it as a maximum intensity projection all from
a single recorded plane of focus. Alternatively, the microscope can be configured to generate high‐speed
3D reconstructions of live‐cell dynamics such as mitotic spindles, all from a single recorded image plane.
Identification of a mitotic SKAP isoform reveals roles in astral microtubule behavior and spindle
D.M. Kern1,2, P.K. Nicholls2, D.C. Page1,2, I.M. Cheeseman1,2; 1MIT, Cambridge, MA, 2Whitehead Institute,
Cambridge, MA
Accurate mitotic spindle positioning is critical for cell division and organismal development. To generate
the force for spindle positioning, astral microtubules form transient, end‐on attachments with dynein
motor localized to the cell cortex. However, little is known about the factors mediating these astral
microtubule‐dynein interactions. The Astrin/SKAP complex plays multiple roles during mitosis, including
in chromosome segregation and centrosome stability, but previous work found conflicting results
regarding SKAP localization and function. Here, we demonstrate that SKAP has two distinct
transcriptional isoforms in mammals; a testis‐specific isoform that has been used ectopically in previous
studies in mitotic cells, and a novel mitotic isoform. The mitotic SKAP isoform displays robust
microtubule plus‐end tracking and rescues the depletion of endogenous SKAP. A SKAP mutant defective
for plus‐end tracking facilitates proper chromosome segregation, but it displays dramatic spindle
positioning defects due to imbalanced dynein‐dependent cortical pulling forces. In mutant cells, spindles
often closely contact one side of the cell with an accumulation of lateral interactions between astral
microtubules and cell cortex. We propose that the Astrin/SKAP complex helps mediate the interaction
between cortical dynein and astral microtubules. Together, our work reveals unappreciated roles for
SKAP and activities governing mitotic spindle positioning.
Microcephaly protein Asp focuses the spindle microtubule minus ends independent of Ncd
motor protein.
A. Ito1, G. Goshima1; 1Division of Biological Science, Nagoya University, Nagoya, Japan
Microcephaly protein Asp focuses the spindle microtubule minus ends independent of Ncd motor
Ami Ito and Gohta Goshima Division of Biological Science, Graduate School of Science, Nagoya University
(Div. of Biol. Sci., Grad. Sch. of Sci., Nagoya Univ.)
The bipolar spindle with two focused poles helps ensure equal segregation of sister chromatids to the
two daughter cells. In Drosophila, abnormal spindle protein (Asp), the orthologue of microcephaly
protein ASPM, has been identified as critical factors for spindle microtubule focusing at the pole.
However, it remains unclear how Asp contributes to pole focusing, a process that also requires Ncd
(kinesin‐14), a minus‐end‐directed motor. In this study, Drosophila S2 cells were utilised as a model
system to determine the role of the Asp protein in spindle pole focusing. We observed that in the
absence of Asp, the spindle pole is persistently unfocused from nuclear envelope breakdown (NEBD) to
metaphase. Surprisingly, the phenotype observed after Ncd motor depletion was distinct, wherein
focusing occurred transiently after NEBD. Double RNAi and reciprocal rescue experiments suggested
that Asp‐induced spindle microtubule focusing is independent of Ncd. Along with our re‐evaluation of
the phenotype, we concluded that Ncd is a global spindle coalescence factor, and not limited to
microtubule ends, as previously proposed. We identified a domain required for Asp function, which has
shown a strong microtubule crosslinking activity in vivo and in vitro. Furthermore, we observed
poleward movement of Asp‐GFP along spindle microtubules in addition to enrichment at the pole. We
obtained evidence that Asp is also localised to the minus ends of intra‐spindle microtubules that are
nucleated in an augmin‐dependent manner, and translocated towards the poles by spindle microtubule
flux involving kinesin‐13 and kinesin‐5. Based on these results, we propose a revised molecular model
for spindle pole focusing, in which Asp crosslinks the minus ends of microtubules at the pole and within
the spindle. Additionally, this study provides new insight into the dynamics of intra‐spindle
The Golgi matrix protein GM130 participates in spindle assembly by activating TPX2 and
capturing microtubules.
J. Wei1,2, Z. Zhang3,4, R. Wynn5, J. Seemann2; 1Biochemistry and Biophysics, University of California San
Francisco, San Francisco, United States, 2Cell Biology, UT Southwestern, Dallas, TX, 3Institute of Life
Sciences, Southeast University, Nanjing , China, 4Pharmacology, UT Southwestern, Dallas, TX, 5Internal
Medicine, UT Southwestern, Dallas, TX
Successful cell division requires concerted efforts to assemble a microtubule‐based spindle that
partitions genomic information and intracellular contents. The function of centrosomes and
chromosomes in spindle formation is well established, but the role of membrane‐bound organelles
remains largely unknown. Here we present a novel microtubule assembly pathway in mitosis that is
initiated from the Golgi and triggered by the Golgi matrix protein GM130. Upon mitotic entry when Cdk1
phosphorylates GM130, importin alpha is recruited to the Golgi membranes via a classical NLS in
GM130. Binding and sequestration of importin alpha by GM130 liberates the spindle assembly factor
TPX2, which activates Aurora‐A kinase and stimulates microtubule nucleation at the Golgi. Once
filaments assemble, the microtubules are further captured and bundled by GM130, thereby linking Golgi
membranes to the spindle. This non‐canonical, Golgi‐derived spindle assembly activity thus empowers
the Golgi to actively shape the spindle to ensure faithful organelle inheritance.
EML3 participates in mitotic spindle assembly by regulating the acentrosomal microtubule
J. Luo1, Z. Deng1, B. Yang1, Q. Jiang1, C. Zhang1; 1The MOE Key Laboratory of Cell Proliferation and
Differentiation and the State Key Laboratory of Membrane Biology, College of Life Sciences, Peking
University, Beijing, China
The proper mitotic spindle is essential for accurate segregation of the duplicated chromosomes into two
daughter cells during the cell division. In vertebrate somatic cells, the mitotic spindle assembly is
dependent on the centrosomal and acentrosomal microtubule nucleation. During the mitotic entry, the
duplicated and separated centrosomes, serving as the main microtubule organization centers (MTOCs)
and the spindle poles, play a dominant role in the microtubule nucleation for the bipolar mitotic spindle
assembly. More recently, the microtubule‐based microtubule nucleation was found to be one of the
acentrosomal microtubule nucleation pathways important for the mitotic spindle assembly. In this
pathway, the Augmin protein complex has been proposed to play a crucial role by recruiting the γ‐
tubulin ring protein complex (γ‐TuRC) to the spindle microtubules, yet the underlying mechanism
remains largely unknown. Here, we show that EML3, a microtubule‐associated protein, participates in
the regulation of the microtubule‐based microtubule nucleation. We find that RNAi knockdown of EML3
in HeLa cells leads to abnormal spindle assembly, severe misalignment of chromosomes, decreased
numbers of the kinetochore‐fibers, inter‐kinetochore tension loss and microtubule intensity decrease.
Furthermore, EML3 interacts with the Augmin complex and promotes its localization on the spindle
microtubules. Moreover, EML3 is phosphorylated at T881 by CDK1, and this regulates the binding of
EML3 with the Augmin complex. Together, our results demonstrate that EML3 participates in the mitotic
spindle assembly by regulating the localization of γ‐TuRC on the spindle microtubules through stabilizing
the interaction of the Augmin complex with the spindle microtubules.
Centrosome age regulates kinetochore microtubule stability and biases chromosome mis‐
I. Gasic1, P. Meraldi1; 1Cellular Physiology and Metabolism Department, University of Geneva, Geneva,
The two poles of the mitotic spindle contain centrosomes of different ages, one old and one young. In
asymmetric stem cell divisions, the age of centrosomes affects their behaviour and their probability to
remain in the stem cell. In contrast, in symmetric divisions old and young centrosomes are thought to
behave the same. This hypothesis has, however, never been addressed directly. Here, we show in
symmetrically dividing human cells, that chromosomes bound to kinetochore‐microtubules from old
centrosomes align onto the metaphase plate less efficiently than those connected to young
centrosomes. Our data indicate that kinetochore‐microtubules associated to old centrosomes are more
calcium‐resistant and thus more stable, favoring unproductive end‐on attachments on unaligned
chromosomes that delay alignment. The bias in alignment caused by the differential kinetochore‐
microtubule stability depends on cenexin, a protein enriched on old centrosomes. Finally, we
demonstrate that this asymmetry biases chromosome segregation in anaphase, causing daughter cells
with old centrosomes to have a 8‐fold higher probability to retain non‐disjoint mis‐segregating
chromosomes. We conclude that centrosome age imposes via cenexin a functional asymmetry on all
mitotic spindles.
Temporal and Spatial Dynamics of Spindle Midzone Assembly Revealed by Lattice Light Sheet
S. Forth1, P. Verma1, M. Sen1, M.C. Pamula1, W.R. Legant2, E. Betzig2, T.M. Kapoor1; 1Laboratory of
Chemistry and Cell Biology, Rockefeller University, New York, NY, 2Janelia Research Campus, HHMI,
Ashburn, VA
Successful cell division in eukaryotes requires the proper assembly of the microtubule‐based spindle.
During anaphase, the spindle elongates and a subset of microtubules undergoes a dramatic
reorganization, particularly near the midzone of the spindle. In order to better understand how this
rapid micron‐scale rearrangement of the central spindle occurs, we have employed lattice light sheet
microscopy to follow the three‐dimensional real‐time dynamics of two microtubule‐associated proteins,
PRC1 and EB1, in dividing human cells. PRC1 has been shown to preferentially crosslink antiparallel
microtubules in vitro, while EB1 tracks the growing plus‐ends of microtubules.
Lattice light sheet microscopy permits us to follow the dynamics of these two proteins in dividing RPE1
cells with diffraction‐limited spatial resolutions in all three dimensions at data acquisition rates
approaching one full cell volume per second. We observe persistent ‘bundles’ decorated with PRC1 that
form prior to anaphase, and we quantify the time trajectories of PRC1’s central spindle accumulation
along each of the hundreds of individual filaments per cell through cytokinesis. Surprisingly, these
‘bundles’ appear to persist for many minutes and are nearly immobile when viewed along a cross‐
sectional slice at the spindle midzone perpendicular to the spindle’s long axis. By performing
quantitative analyses using custom‐written 3D fiber tracking software, we characterize the lateral
coalescence of PRC1 bundles near the midzone, and propose a novel mechanism for organizing the
microtubules of the central spindle. In contrast, highly dynamic EB1 tracks are observed to pass through
this same central ‘slice’, yet appear to be occluded from regions containing PRC1‐decorated bundles.
Together these data, along with chemical perturbations to the actomyosin network, reveal how the time
evolution of lateral interactions between stabilized microtubule bundles contribute to the assembly of
the central spindle.
Minisymposium 08: Lipid Organization, Transport,Composition and
Local control of membrane composition by Integrin receptors.
J.K. Mathew1, S. Mayor1; 1National Centre for Biological Sciences (TIFR), Bangalore , India
Cells sense and react to their extracellular environment via parsing vital information in both directions
via membrane receptors embedded in the plasma membrane. In a living cell, the membrane exhibits
lateral heterogeneities in the organisation of membrane lipids and proteins, not explainable by simple
equilibrium mechanisms. These lateral heterogeneities facilitate the establishment of very specific local
composition in the vicinity of the membrane receptors, allowing a cell to specifically respond to and
interpret its physical and chemical environment. Here we explore how a membrane receptor may exert
control on its local composition. Using the technique of homo‐FRET microscopy based on imaging
fluorescence emission anisotropy, we and others (1) find that the engagement of integrin receptors with
fibronectin, leads to the nanoclustering of fluorescently‐tagged GPI‐anchored proteins in the vicinity of
the activated receptor. GPI‐anchored proteins in turn are linked to cytoplasmic actin filaments via trans‐
bilayer linkages to inner‐leaflet phosphatidylserine in small lo like lipidic environments (2). These lo‐like
nanodomains are stabilised upon immobilisation of inner leaflet‐lipids by associated actin‐filaments,
resulting in the generation of a local lo environment in the vicinity of the receptor. Upon ligation of the
integrin receptor with its ligand, the ensuing signal transduction leads to the generation of dynamic
cortical actin filaments, which in turn create local contractile actin platforms (asters) in conjunction with
myosin activity (3). These asters lead to the active focusing of GPI‐anchored proteins connected to
dynamic actin filaments, into small scale nanoclusters with unusual spatial and temporal statistics (3).
Integrin receptors thus fine‐tune their immediate membrane organisation by exploiting an active actin‐
based mechanism, thereby facilitating the creation of localised lo domains. The subsequent recruitment
of molecules that have an affinity for these domains is likely to play an important role in the control of
integrin function.
References: 1. Van Zanten TS, Cambi A, Koopman M, Joosten B, Figdor CG, Garcia‐Parajo MF. Hotspots
of GPI‐anchored proteins and integrin nanoclusters function as nucleation sites for cell adhesion. Proc
Natl Acad Sci U S A. 2009;106(44):18557–62. 2. Raghupathy R, Anilkumar AA, Polley A, Singh PP, Yadav
M, Johnson C, et al. Transbilayer lipid interactions mediate nanoclustering of lipid‐anchored proteins.
Cell; 2015;161(3):581–94. 3. Gowrishankar K, Ghosh S, Saha S, Rumamol C, Mayor S, Rao M. Active
remodeling of cortical actin regulates spatiotemporal organization of cell surface molecules. Cell.
Interplay between Membrane Traffic and Sphingolipid Organization in the Plasma Membrane.
M.L. Kraft1, H.A. Klitzing1, R. Kim1, P.K. Weber2, J. Zimmerberg3; 1School of Chemical Sciences, University
of Illinois, Urbana Champaign, Urbana, IL, 2Chemical Biology and Nuclear Science, Lawrence Livermore
National Laboratory, Livermore, CA, 3Eunice Kennedy Shriver National Institute of Child Health and
Human Development, National Institutes of Health, Bethesda, MD
Lateral variations in the distributions of the cholesterol, various lipids, and proteins in the plasma
membrane are required to coordinate mammalian cell function. Though various lipid species are
thought to be compartmentalized into domains within cellular membranes, the precise distributions of
most lipid species in the plasma membrane are not well‐characterized. To address this issue, we have
imaged the distributions of stable isotope‐labeled lipids in the plasma membranes of intact cells by
combining metabolic isotope incorporation with high‐resolution secondary ion mass spectrometry
(SIMS) performed on a Cameca NanoSIMS 50. In this approach, we first metabolically incorporate
distinct stable isotopes into the lipid species of interest. Then the cells are chemically fixed, and the
lipid‐specific rare isotope enrichments on their surfaces are imaged with high‐resolution SIMS. Using this
approach, we previously imaged the sphingolipids in the plasma membranes of fibroblast cells that had
been metabolically labeled with 15N‐sphingolipid precursors (15N‐sphingosine and 15N‐sphinganine) for 6
days so that the resulting 15N‐sphingolipids reached a steady‐state distribution in their membranes.
High‐resolution SIMS revealed micrometer‐scale 15N‐sphingolipid patches in their plasma membranes.
By imaging the effects of drugs on the sphingolipid distribution, we determined the sphingolipids were
confined to distinct plasma membrane domains by the cytoskeleton and its associated proteins.
Here we use this approach to investigate the role of intracellular membrane trafficking on sphingolipid
organization within the plasma membrane. By imaging the sphingolipid organization at the surfaces of
cells that were metabolically labeled for 30 min to 6 days, we found that the newly synthesized 15N‐
sphingolipids were rapidly incorporated into sphingolipid‐rich plasma membrane domains. The 15N‐
sphingolipid distributions that we observed at different metabolic labeling times, and complementary
studies of sphingolipid metabolism will be presented. The implications of these findings on membrane
organization and function will be discussed.
PI4P/phosphatidylserine countertransport at ORP5‐ and ORP8‐mediated ER‐plasma membrane
J. Chung1, F. Torta2, K. Masai1, L. Lucast1, H. Czapla1, L.B. Tanner2, P. Narayanaswamy2, M.R. Wenk2, F.
Nakatsu1, P. De Camilli1; 1Department of Cell Biology, Yale School of Medicine/HHMI, New Haven, CT,
Department of Biochemistry, Yong Loo Lin School of Medicine, National University of Singapore,
Singapore, Singapore
Lipid transfer between cell membrane bilayers at contacts between the endoplasmic reticulum (ER) and
other membranes help to maintain membrane lipid homeostasis. We found that two similar ER integral
membrane proteins, oxysterol‐binding protein (OSBP)–related protein 5 (ORP5) and ORP8, tethered the
ER to the plasma membrane (PM) via the interaction of their pleckstrin homology domains with
phosphatidylinositol 4‐phosphate (PI4P) in this membrane. Their OSBP‐related domains (ORDs)
harbored either PI4P or phosphatidylserine (PS) and exchanged these lipids between bilayers. Gain‐ and
loss‐of‐function experiments showed that ORP5 and ORP8 could mediate PI4P/PS countertransport
between the ER and the PM, thus delivering PI4P to the ER‐localized PI4P phosphatase Sac1 for
degradation and PS from the ER to the PM. This exchange helps to control plasma membrane PI4P levels
and to selectively enrich PS in the PM.
M. Sohn1, P.T. Ivanova2, A.H. Brown2, Y. Kim1, T. Balla1; 1NICHD, National Institutes of Health, Bethesda,
MD, 2Department of Pharmacology and Biochemistry, Vanderbilt‐Ingram Cancer Center, Vanderbilt
University, School of Medicine, Nashville, TN
Phosphatidylinositol 4‐phosphate (PI4P) is a quantitatively minor but metabolically very active lipid
component of eukaryotic membranes. PI4P has been recognized for its role in vesicular trafficking from
the Golgi and in transmembrane signaling at the plasma membrane (PM). Phosphatidylserine (PS), on
the other hand, is an important structural lipid primarily found in the inner leaflet of the PM and in
endosomes. PS is also converted to phosphatidylethanolamine in the mitochondria. In the present
study, we report on the close relationship between PI4P turnover and PS metabolism that plays a critical
role in the metabolic control of PS synthesis and distribution to different membranes. Prolonged
pharmacological inhibition of phosphatidylinositol 4‐kinase IIIa (PI4KA), an enzyme that phosphorylates
PI to generate PI4P in the PM, led to a 50% reduction in PS levels based on lipidomic analysis. Metabolic
labeling experiments showed a dramatic decrease in PS synthesis after acute PI4KA inhibition, whereas
phosphatidylinositol‐4‐kinase IIIb (PI4KB) inhibitors had no significant effect. Interestingly,
overexpression of mutant PS synthase 1 (PSS1) enzymes that are resistant to PS‐induced product
inhibition alleviated the inhibition of PS synthesis by PI4KA inhibitors. Notably, overexpression of mutant
PSS1 caused over‐production and accumulation of PS in the ER, observed in live cells using the
Lactadherin C2 domain as a specific PS probe. There was no PS detectable in ER of naive cells or cells
expressing the wild‐type PSS1 enzyme. Overexpression of the mutant PSS1 enzyme also caused an
increase in Sac1 phosphatase activity of ER membranes as assessed by a direct phosphatase assay.
Furthermore, mutant PSS1‐overexpressing cells exhibited a decrease in the PI4P pools both in the Golgi
and PM evaluated by live cell imaging and also by PI4P immunostaining. Taken together, these data
indicated that PI4KA activity is required to maintain PS synthesis, most likely by preventing PS
accumulation in the ER. Conversely, increased PS levels in the ER negatively regulate PI4P levels by
activating the Sac1 phosphatase. Our hypothesis is that PS removal from the ER is aided by PI
phosphorylation by PI4KA and inhibition of this enzyme causes accumulation of PS in the ER, which
results in inhibition of PS synthesis. Current studies are aimed at determining whether the contact sites
between the ER and the PM serve as the structural basis for the metabolic interplay between PI4P and
PS metabolism and testing our hypothesis that PI 4‐phosphorylation is important to drive PS removal
from the ER.
The acyltransferase LYCAT regulates phosphoinositides and specific stages of endomembrane
L.N. Bone1, R.M. Dayam1, R.J. Botelho1, C.N. Antonescu1; 1Department of Chemistry and Biology, Ryerson
University, Toronto, ON
Phosphoinositides (PIs) are signaling lipids that control organelle dynamics, proliferation, nutrient
uptake, autophagy and apoptosis. There are seven species of PIs defined by the phosphorylation of their
inositol headgroups, a process controlled by specific lipid kinases and phosphatases. While much has
been learned about PI inositol headgroup phosphorylation, much less is known about the regulation and
function of the incorporation of specific fatty acyl chains within PIs. Importantly, PIs exhibit unique
specificity of composition of fatty acyl chains, such that the majority of PIs are 1‐steroyl (18:0), 2‐
arachidonyl (20:4). This specific composition of fatty acyl groups within PIs is controlled in part by the PI
acyltransferase LYCAT. How LYCAT, and in turn incorporation of specific fatty acids within PIs, controls
the function of PIs is poorly understood, which we examined here. LYCAT gene silencing altered the
cellular localization and levels of phosphatidylinositol‐(4,5)‐bisphosphate (PIP2) and
phosphatidylinositol‐3‐phosphate (PIP3), which control membrane traffic from the plasma membrane to
early endosomes, but was without effect on other PI species examined. This suggests that LYCAT may
function subsequently to PI synthesis in the ER to control the acylation of specific pools of PIs.
Consistent with this interpretation, while LYCAT has some ER localization, substantial LYCAT signal did
not co‐localize with canonical ER markers. The control of specific PIs by LYCAT impacted endomembrane
traffic, as evinced by the fact that silencing of LYCAT reduced cell surface levels of transferrin receptor
(TfR), a membrane protein that undergoes constitutive clathrin‐mediated endocytosis (CME) from the
plasma membrane followed by recycling to the cell surface. CME occurs by the assembly of proteins
including clathrin to a small, invaginating region of the plasma membrane termed a clathrin‐coated pit,
some of which eventually undergo scission to produce internalized vesicles. Live cell imaging of cells
expressing GFP fused to clathrin light chain (GFP‐CLC) coupled to custom image analysis revealed that
LYCAT silenced cells exhibited alterations in the assembly and dynamics of clathrin‐coated pits (CCPs),
consistent with the role of PIP2 in the regulation of CCPs. LYCAT silencing also delayed TfR transit to
early endosomes as well as TfR recycling, consistent with the role of PI3P in this process. Collectively,
these results show that the PI acyltransferase LYCAT controls the function of specific species of PIs,
which in turn selectively impacts specific stages of endomembrane traffic. Hence, the regulation of fatty
acyl content of PIs is an important new dimension for the control of PI function.
Lipid engineering approaches reveal cellular roles for membrane viscosity.
I. Budin1, J.D. Keasling2; 1Miller Institute, University of California Berkeley, Berkeley, CA, 2Chemical
Engineering, University of California Berkeley, Berkeley, CA
Biological membranes feature characteristic lipid compositions that can vary dramatically between
different organisms, tissues, and organelles. Addressing the functional basis for this diversity requires
better tools for manipulating lipid composition in vivo. I will present our efforts to apply metabolic
engineering approaches to systematically modulate lipid composition in model organisms. By taking
advantage of titratable promoters and knowledge of lipid synthesis pathways, we have generated a set
of bacterial (E. coli) and yeast (S. cerevisiae) strains in which parameters of native membrane
stoichiometry is under experimental control. Using a combination of mass spectrometry and biophysical
assays, we then correlate measurements of membrane composition and physical properties to the
physiology of cells and the activity of specific molecular processes. I will present of set of experiments
that use this approach to characterize cellular roles for phospholipid unsaturation, which regulates
membrane viscosity in both bacteria and yeast. These experiments have elucidated two potentially
universal roles for viscosity in cell membranes. Genetic titration of the stoichiometry of unsaturated
phospholipids controls flux through cellular respiration, which we hypothesize is a result of diffusional
regulation of electron carriers (quinones) in the membrane. Low levels of lipid unsaturation also result in
defects in protein folding, which we propose is a result of frictional resistance of membrane lipids to
polypeptide translocation. I will discuss physical models for these observations and their implications for
the evolution of lipid composition in energy transducing and protein folding membranes.
Lipid Binding by Osh4p, an OSBP homologue, is Required for a Discrete Step in Polarized
R.J. Smindak1, D.M. Hyatt1, K.G. Kozminski1,2, R.C. Deutscher1; 1Biology, University of Virginia,
Charlottesville, VA, 2Cell Biology, University of Virginia, Charlottesville, VA
Polarized exocytosis is required for diverse events such as polarized cell growth and cell migration. The
seven‐member Osh protein family, the S. cerevisiae homologues of the mammalian oxysterol‐binding
protein (OSBP) family, is required for polarized exocytosis in yeast (Alfaro et al., Traffic, 2011). Within
the Osh protein family, Osh4p (Kes1p) binds and transfers specific lipid species between membranes in
vitro. However, in vivo, it is unclear whether lipid binding and transfer by Osh4p is an essential activity
(de St. Jean et al., JCB, 2012; Georgiev et al., Traffic, 2011). Three questions regarding Osh4p function
are i) does Osh4p regulate non‐polarized exocytosis in addition to polarized exocytosis; ii) does lipid
binding by Osh4p regulate its role in polarized exocytosis; and, iii) what steps within the process of
polarized exocytosis requires Osh protein function. In answer to the first question, we show data,
obtained by assaying invertase secretion, that excludes a role for Osh4p in non‐polarized exocytosis.
Secondly, we show, by assaying a panel of lipid binding deficient mutants for Bgl2p secretion, that
Osh4p must be able to bind both sterol and phosphatidylinositol‐4‐phosphate (PI4P) for polarized
exocytosis to occur normally. Thirdly, we show that Osh protein family activity, and specifically lipid
binding by Osh4p, are required for vesicle docking at the plasma membrane. In strains lacking Osh
protein family activity or in strains expressing only lipid binding deficient Osh4p and no other Osh family
members, the amount of assembled SNARE complexes (Sso1/2p and Snc2p) decreases, indicating a
vesicle docking defect, answering a long‐standing question as to where Osh family proteins function in
the process of polarized exocytosis. Together, these data show that the docking of vesicles, which
support polarized cell growth, at the plasma membrane, requires lipid binding by Osh4p. These results
suggest that other Osh proteins in yeast or OSBP family proteins in other organisms may regulate
polarized exocytosis or the formation of other membrane‐membrane contact sites, in a lipid dependent
Sac1 regulates microtubule stability and trafficking of cell surface adhesion molecules in the
developing Drosophila eye.
L.M. Del Bel1,2, R. Wilk2, J. Burgess1,2, G. Polevoy2, H. Wei2, J.A. Brill1,2; 1Molecular Genetics, The
University of Toronto, Toronto, ON, 2Cell Biology, The Hospital for Sick Children, Toronto, ON
Phosphatidylinositol phosphates (PIPs) play key roles in membrane trafficking, cytoskeletal organization
and signaling during cell morphogenesis. The enzymes that control the production and destruction of
PIPs have been well studied in budding yeast and cultured mammalian cells. However, the in vivo
requirements for particular PIP pathway enzymes have been difficult to predict from results in single
cells. In experiments examining regulation of phosphatidylinositol 4‐phosphate (PI4P) in the fruit fly
Drosophila melanogaster, we discovered that the PIP phosphatase Sac1 is required for normal
patterning of the retinal epithelium. Using a temperature‐sensitive allele of sac1, we showed that Sac1
is needed at early stages of eye development for plasma membrane delivery of the Neph1 homolog
Roughest (Rst), an immunoglobulin family cell surface adhesion molecule required for retinal patterning
and pigment cell specification. Notably, the Rst paralog Kirre and other cell surface proteins (DE‐cad,
Dlg, Arm) localize normally in Sac1 mutant eyes, indicating that trafficking of Rst is exquisitely sensitive
to elevated levels of PI4P. Impaired trafficking of Rst was associated with microtubule defects and
altered distribution of the exocyst complex subunit Sec8, which colocalized with Rst to enlarged
structures in sac1 mutants. Our data indicate that restriction of PI4P by Sac1 is needed for microtubule
stability and proper trafficking of cell adhesion molecules within the developing eye. Because Sac1 is
highly conserved, our results have important implications for the role of Sac1 in cell‐cell interactions in
human development and disease.
Dietary fats remodel the plasma membrane lipidome to regulate the stability of membrane
K.R. Levental1, J.H. Lorent1, X. Lin1, A. Gorfe1, I. Levental1; 1Integrative Biology and Pharmacology, The
University of Texas Health Science Center at Houston, Houston, TX
The spatial organization of biological membranes is crucial for the their extensive functionality. One
example of such organization is the partitioning of metazoan plasma membranes into lipid‐driven
ordered domains termed membrane rafts. While the physicochemical interactions responsible for raft
domain formation are clearly established, the specific properties and biological consequences of these
domains in living cells remain unclear. In particular, little is known about how the properties of raft
domains are regulated, either by endogenous mechanisms or exogenous perturbations. In this context,
an intriguing question concerns the effect of dietary fatty acids on the lipid composition, biophysical
properties, and ultimate function of membrane domains. Here, we investigate the consequences of the
w‐3 polyunsaturated fatty acid docosohexaenoic acid (DHA) on the biochemical and biophysical
properties of the mammalian plasma membrane (PM). By detailed lipidomic analysis, we observe that
DHA supplemented into cell culture media is robustly incorporated into specific plasma membrane
lipids, and that this incorporation leads to wholesale remodeling of the plasma membrane lipidome.
Specifically, both saturated lipids and cholesterol are upregulated as a compensatory response to the
fluidizing effect of the PUFA‐containing lipids. The combined effect of DHA incorporation and
compensatory response result in biophysical remodeling of the PM, which we address in membrane
model systems ranging from coarse‐grained molecular dynamics simulations to intact PMs isolated from
live cells. In all systems, DHA–containing lipids promote the formation of raft domains by disordering the
non‐raft domains, thereby increasing the differences between coexisting raft and non‐raft domains. The
correlation between interdomain disparity and stability of phase separation holds for a variety of
perturbations, including manipulation of cholesterol levels and treatment by exogenous amphiphiles,
suggesting it as a general design feature of the organization of biological membranes.
Minisymposium 09: Microtubule‐Based Motility and Dynamics
A proteomics survey reveals the human cytoplasmic dynein transportome.
W.B. Redwine1, I. Hollyer2, M. DeSantis2, S.L. Reck‐Peterson1,2; 1Cell Biology, Harvard Medical School,
Boston, MA, 2Cellular and Molecular Medicine, University of California San Diego, La Jolla, CA
The extensive microtubule network, in combination with a diverse set of microtubule‐based molecular
motors, provides an efficient way to distribute intracellular cargos within eukaryotic cells. The proper
organization of sub‐cellular components is crucial to human health, especially within neuronal cells,
where mutations within microtubule‐based motors and their regulators are linked to many neurological
diseases. Microtubules are polar structures with plus ends often located at the cell periphery and minus
ends embedded in MT organizing centers that are typically close to the cell center. This polarity offers
tracks along which the molecular motors dynein and kinesin move unidirectionally. While there are
numerous members of the plus‐end‐directed kinesin family, a single gene encodes cytoplasmic dynein‐1,
the primary minus‐end‐directed motor in the cytoplasm of eukaryotic cells. Thus, a single dynein must
coordinate a wide‐range of processes, including transporting organelles, vesicles, RNAs, proteins and
viruses. In addition, dynein has critical functions in both mitosis and meiosis. Given these diverse
functions, dynein must be highly regulated, both at the level of its motor activity and in how it interacts
with distinct cargos. To determine how dynein is regulated we have employed a proteomics approach
that combines in vivo biotinylation with tandem mass spectrometry to identify novel interacting
proteins of the dynein holoenzyme complex. Using this approach in human cells we have discovered
novel cargo adaptor proteins and candidate cellular cargos, providing the first glimpse of the dynein
Dynarrestin is a novel small molecule inhibitor of cytoplasmic dynein.
T. Yeh1, S. Höing2,3, M. Baumann3, H.C. Drexler4, S.A. Ketcham1, B. Klebl3, H.R. Schöler2,5, H. Waldmann6,
J.L. Sterneckert2,7, T.A. Schroer1; 1Biology, Johns Hopkins University, Baltimore, MD, 2Cell and
Developmental Biology, Max Planck Institute for Molecular Biomedicine, Münster, Germany, 3Lead
Discovery Center GmbH, Dortmund, Germany, 4Bioanalytical Mass Spectrometry, Max Planck Institute
for Molecular Biomedicine, Münster, Germany, 5Medical Faculty, University of Münster, Münster,
Germany, 6Chemical Biology, Max Planck Institute for Molecular Physiology, Dortmund, Germany, 7DFG‐
Research Center for Regenerative Therapies Dresden, Cluster of Excellence / Technische Universität
Dresden, Dresden, Germany
Hedgehog (Hh) signaling involves multiple steps, some of which require the activity of cytoplasmic
dyneins. Because aberrant Hh signaling contributes to the pathogenesis of multiple cancers, small
molecule inhibitors that block the Hh pathway at different steps are in great demand. Currently
available inhibitors target the Smoothened receptor (Smo) which is trafficked within the primary cilium
and cell body. Because Smo can acquire mutations that lead to drug resistance, a major challenge is to
identify compounds that inhibit Hh signaling downstream of Smo. Dynarrestin is a novel small molecule
inhibitor that blocks Hh signaling and the proliferation of Hh‐dependent tumor cells. Chemical
proteomics identified the target of dynarrestin as cytoplasmic dynein. We find that dynarrestin
reversibly inhibits endosome movement, mitotic spindle positioning and pole focusing, mitotic
progression, and cytoplasmic dynein‐based microtubule translocation in vitro without blocking ATP
hydrolysis or changing dynein abundance, composition or S‐value. Dynarrestin is thus a valuable new
tool for studying dynein‐dependent cellular processes and has great potential for development into
novel anti‐cancer drugs aimed at controlling Hh signaling downstream of Smo.
IFT‐dynein dynamics in vivo: an ensemble and single‐molecule quantification.
J. Mijalkovic1, B. Prevo1, P.J. Mangeol1, F. Oswald1, E.J. Peterman1; 1Physics and Astronomy, VU
Amsterdam, Amsterdam, Netherlands
Cytoplasmic dyneins are the main drivers of microtubule‐based retrograde transport in eukaryotic cells.
Cytoplasmic dynein 1 plays a role in retrograde intracellular transport and cell division, whereas
cytoplasmic dynein 2, also known as IFT‐dynein, co‐operates with kinesin‐2 motors to assemble and
maintain cilia in a process called intraflagellar transport (IFT). While cytoplasmic dynein 1 has been the
subject of many recent studies, relatively little is known about IFT‐dynein. Here, we focus on the
mechanism and dynamics of IFT‐dynein at both the ensemble and single‐molecule level. In other words,
how does IFT‐dynein behave in vivo and how does it work together with the kinesins to drive transport?
To this end, we use fluorescence microscopy to visualize labeled IFT‐dynein motors in the chemosensory
cilia of living C. elegans. Transgene worms were generated using the Mos1‐mediated single copy
insertion (MosSCI) method to ensure endogenous motor expression levels. The movies were processed
to kymographs, from which location‐dependent velocities and motor numbers were obtained using in‐
house developed kymograph‐analysis software. To obtain insight into the behavior of individual motors,
we employed photoactivation of PA‐GFP‐labeled IFT‐dynein, which allowed, for the first time, the
tracking of individual IFT‐dynein motors in vivo.
The ensemble analysis revealed that IFT‐dynein moves in trains consisting of tens of motor proteins.
Retrograde trains are smaller but more frequent than anterograde (kinesin‐driven) trains. We find that
anterograde and retrograde IFT‐dynein flux are equal along cilia, indicating that the cilium is a closed
system for dynein. While the kinesin composition of a motor train varies along the track, the amount of
dynein per train remains relatively constant. Remarkably, this does not result in directionality changes
along the track, such as has been reported for other opposite polarity, 'tug‐of‐war' motility systems in
vivo and in vitro, thus suggesting that the activity of retrograde and anterograde motors is carefully
orchestrated in IFT.
Single‐motor observations of IFT‐dynein are in the line with the ensemble findings. Trajectories showed
distinct features of IFT‐dynein motility: diffusive behavior at the ciliary base, pauses, turns and directed
motion. Pauses in retrograde or anterograde trajectories are never followed by a directional switch.
Moreover, retrograde‐to‐anterograde turn events are rare.
This combined ensemble and single‐molecule approach has provided novel quantitative insight into IFT‐
dynein transport dynamics in living organisms, shedding light on the complex functioning of dynein
motors in general.
The C. elegans Ninein‐Related Protein NOCA‐1 Functions Coordinately with γ‐tubulin and in
Parallel to Patronin to Assemble Non‐Centrosomal Microtubule Arrays.
S. Wang1,2, S. Quintin3,4, R.A. Green1, D.K. Cheerambathur1, S.D. Ochoa1, D. Wu1, B. Prevo1, A.B. Desai1,5,
K. Oegema1,5; 1Ludwig Institute for Cancer Research, La Jolla, CA, 2Biomedical Sciences Graduate
Program, University of California, San Diego, La Jolla, CA, 3CNRS, Université de Strasbourg, Illkirch,
France, 4CNRS, Institut de Génétique et de Biologie Moléculaire et Cellulaire, Illkirch, France, 5Cellular
and Molecular Medicine, University of California, San Diego, La Jolla, CA
In metazoans, non‐centrosomal microtubule arrays assemble in differentiated tissues to perform
mechanical and transport‐based functions. How these arrays form remains poorly understood. We
identify C. elegans NOCA‐1 as a protein with homology to vertebrate ninein that is specifically required
to assemble non‐centrosomal, but not centrosomal, microtubule arrays. In vitro analysis revealed that
the region of NOCA‐1 with homology to ninein directly binds microtubules. In C. elegans tissues, NOCA‐1
co‐localizes with the microtubule nucleator, γ‐tubulin, to non‐centrosomal sites, but is absent from
centrosomes. In the germline, knockdown of NOCA‐1 or γ‐tubulin results in an essentially identical
defect in microtubule assembly. NOCA‐1 and γ‐tubulin co‐localize to the plasma membrane, and NOCA‐1
targeting is γ‐tubulin‐dependent when a non‐essential putatively palmitoylated cysteine at its N‐
terminus is mutated to alanine. In the larval epidermis, NOCA‐1 acts in parallel with the minus end
protection factor Patronin/PTRN‐1 to direct assembly of a circumferential microtubule array required for
worm growth and morphogenesis. A tissue‐specific protein knockout method further revealed that γ‐
tubulin functions coordinately with NOCA‐1 and in parallel to Patronin/PTRN‐1. These results show that
NOCA‐1 functions with γ‐tubulin to direct the assembly of non‐centrosomal arrays in multiple tissues
and highlight functional overlap between the ninein and Patronin families of microtubule cytoskeleton‐
controlling proteins.
Towards an understanding of the molecular mechanism of the unique biomechanical properties
of the yeast kinesin‐8 Kip3.
H. Arellano‐Santoyo1,2,3, E. Stokasimov1,2,3, X. Su2,4, D.S. Pellman1,2,3; 1Cellular Biology, Harvard Medical
School, Boston, MA, 2Howard Hughes Medical Institute, Boston, MA, 3Pediatric Oncology, Dana Farber
Cancer Institute, boston, MA, 4School of Medicine, UCSF, San Francisco, CA
The kinesin‐8 family of motors is best known for its conserved role in microtubule length control and in
the regulation of spindle size. Kinesin‐8s are highly processive plus‐end directed motors that are able to
dwell at the plus‐ends of microtubules and alter their dynamics. Kip3, the yeast kinesin‐8, can actively
depolymerize GMPCPP‐stabilized microtubules and has been shown to be a catastrophe factor on
dynamic microtubules in vitro. Although the importance of spindle length regulation by kinesin‐8s in a
variety of cell types and organisms is well established, the structural elements that uniquely allow for
the dual‐functions of Kip3 are unknown. Through domain swapping analysis and biophysical
characterization of modular elements in the motor domain we have dissected the minimal elements
necessary and sufficient for Kip3’s depolymerase activity, plus‐end dwelling and its high processivity.
Furthermore we will report progress in testing the two current models of kinesin‐8 function in the field,
namely the bump‐off model and the plus‐end capping model. Future structural work will be required to
understand the interaction of Kip3 with the tubulin dimer and the microtubule and how it diverges from
other motile kinesins.
A small GTPase ARL‐8 regulates synapse formation by unlocking the autoinhibition of the axonal
kinesin UNC‐104/KIF1A.
K. Shen1,2, S. Niwa1,3; 1Biology, Stanford University, Stanford, CA, 2Howard Hughes Medical Institute,
Stanford, CA, 3Frontier Research Institute for Interdisciplinary Sciences, Tohoku University, Sendai, Japan
The kinesin motor proteins hydrolyze ATP and generate movements along microtubules. Intramolecular
autoinhibitory interactions have been shown for several kinesins, however, it remains unknown how
such autoinhibition controls motor function and cargo transport in vivo. Here, we characterize the
autoinhibitory mechanism for the synaptic vesicle kinesins, UNC‐104/KIF1A. We identified three
mutations in the stalk region and motor domain of unc‐104 that specifically disrupt the autoinhibition.
These mutations augment microtubule and cargo vesicle binding in vitro. In vivo, these mutations cause
excessive activation of UNC‐104, leading to defects in cargo stopping and ultimately a decrease in
synapse density and smaller synapses. We also show that the synaptic‐vesicle bound small GTPase ARL‐8
activates UNC‐104 by unlocking the autoinhibition. These results demonstrate that the autoinhibitory
mechanism is utilized to regulate the distribution of transport cargoes and is important for
synaptogenesis in vivo.
Interplay between kinesin‐1 and cortical dynein during axonal outgrowth and microtubule
organization in Drosophila neurons.
U. del Castillo1, M. Winding1, W. Lu1, V.I. Gelfand1; 1Dept. of Cell and Molecular Biology, Feinberg School
of Medicine, Northwestern University, Chicago, IL
One of the key differences between axons and dendrites is that axons have uniformly polarized
microtubules, with plus‐ends out, while dendrites contain microtubules of mixed polarity (in mammals)
or with mostly minus‐ends‐out (in Drosophila and C. elegans). This difference in microtubule polarity
affects the selectivity of cargo sorting. Here, we used live imaging of plus‐ and minus‐end microtubule
markers in cultured Drosophila neurons to understand how this polarity is established. Surprisingly, we
found that axonal microtubules underwent dramatic rearrangements during development. The initial
protrusion of the plasma membrane in lengthening axons was driven by the minus‐ends of microtubules
pushing against the plasma membrane, and nascent neurites contained microtubule arrays of mixed
polarity. This configuration was established by kinesin‐1 driven microtubule‐microtubule sliding. During
axonal outgrowth, microtubule orientation shifted from mixed to uniform polarity with plus‐ends out.
This transition depended on the activity of cortically attached cytoplasmic dynein. Here, we present and
discuss a unified model to describe the role of kinesin‐1 and cytoplasmic dynein in the organization of
cytoplasmic microtubules and formation of Drosophila axons.
Purified Recombinant Human Tubulin Isotypes Show Distinct Polymerization Properties In Vitro.
M.C. Pamula1, S. Ti1, T.M. Kapoor1; 1Laboratory of Chemistry and Cell Biology, The Rockefeller University,
New York, NY
Microtubules are dynamic cytoskeletal polymers that are essential for a diverse array of cellular
processes including intracellular transport and cell division. The composition of cellular microtubules
varies as multiple tubulin isotypes differing in amino acid sequence have been identified in all
eukaryotes and often have complex expression patterns in multicellular organisms. In vivo observations
have suggested the non‐interchangeable roles of β‐tubulin isotypes in neural development, perhaps
involving different protein partners or distinct intrinsic microtubule polymerization dynamics. However,
evidence for clear functional roles of tubulin isotypes has remained elusive as direct testing of this cell
biological hypothesis is lacking. This knowledge gap is in large part due to the immense challenge in
isolating specific and homogeneous human tubulin. Here, we develop a general method for purifying
specific forms of untagged recombinant human tubulin. This protocol typically yields 1.5 mg tubulin per
liter of cultured insect cells with >95% purity. We purify two forms of beta tubulin, isotype 2B (TUBB2B)
and isotype 3 (TUBB3), and use total internal reflection fluorescence (TIRF) microscopy to characterize
the polymerization dynamics of single filaments in vitro. We quantify parameters of dynamic instability
for each isotype and show that TUBB2B and TUBB3 have distinct intrinsic dynamic properties. Together,
our findings provide insight into the biochemical properties and polymerization dynamics of specific
human tubulin isotypes. This approach can be applied further to dissect other aspects of tubulin
function and its effect on diverse cellular processes.
Mechanism of Microtubule Lumen Entry for the α‐Tubulin Acetyltransferase Enzyme αTAT1.
C.E. Coombes1, A. Yamamoto1, M. McClellan1, M. Plooster1, J. Alper2,3, G. Luxton1, J. Howard2, M.
Gardner1; 1Genetics, Cell Biology, and Development, University of Minnesota, Minneapolis, MN,
Molecular Biophysics and Biochemistry, Yale University, New Haven, CT, 3Physics and Astronomy,
Clemson University, Clemson, SC
Microtubules are structural polymers inside of cells that are subject to post‐translational modifications.
These post‐translational modifications create functionally distinct subsets of microtubule networks in
the cell, and acetylation is the only modification that takes place in the hollow lumen of the
microtubule. While it is known that the α‐tubulin acetyltransferase αTAT1 is the primary enzyme
responsible for microtubule acetylation, the mechanism for how αTAT1 accesses the acetylation site
inside of the microtubule lumen is not well understood. By performing biochemical assays, fluorescence
and electron microscopy experiments, and computational simulations, we found that αTAT1 enters into
the microtubule lumen predominantly through the microtubule ends, and is preferentially targeted to
tapered microtubule ends with exposed acetylation sites. In addition, once αTAT1 enters the
microtubule lumen, it has limited mobility to travel down the microtubule lumen due to rapid rebinding
of αTAT1 on highly concentrated α‐tubulin acetylation sites, leading to short bursts of acetylation at the
microtubule tips. These results provide important insights into the mechanism for αTAT1 microtubule
acetylation and suggest a mechanism for how αTAT1 may recognize and specifically acetylate older
Minisymposium 10: New Technologies and Immuno‐Signaling
LOV and Zdark, a protein pair that dimerize selectively in the dark, provide a versatile
optogenetic toolbox.
H. Wang1, K. Bryant1, C. Der1, K.M. Hahn1; 1Pharmacology, university of North Carolina Chapel Hill,
Chapel Hill, NC
Using mRNA display we developed Zdark (Zdk), a small protein based on the Z domain (1), that binds
only to the dark state of the LOV2 domain from Avena sativa phototropin. The LOV2 domain undergoes
a large, reversible conformational change induced by light between 400 and 500 nm. Zdk binds to the
dark state of LOV2 with a Kd of ~100nM, but shows no detectable binding to the lit state. So far, we
have developed four optogenetic approaches based on different combination of Zdk and LOV2. The first
method is LOVTRAP, in which the LOV domain is anchored at the mitochondrion and different proteins
of interest are fused to Zdk. In the dark, the Zdk‐protein fusion is sequestered at the mitochondrion, but
upon irradiation it is reversibly released within < 0.5 secs. LOVTRAP has been successfully applied to
regulate RhoA, Rac1 and Vav2. The second method we developed is called Z‐bridge, where Zdk and
LOV2 molecules are joined using a flexible linker. In the dark, Zdk is blocked by the intramolecular LOV2;
while in the light, Zdk is released from the LOV2 and able to bind intermolecularly to a LOV2 dark‐state
mutant. This technique has been applied to regulate KRas localization, thus inducing different cell fates:
proliferation by recruitment of KRas to the plasma membrane, and apoptosis by recruitment of KRas to
mitochondria or ER. The third method, Zlock, controls short peptides using the Zdk‐LOV2 interaction. In
the Zlock system, Zdk and LOV2 are fused to the N‐ and C‐ termini of the peptide. In the dark the
construct forms a loop because Zdk binds to LOV2, thus preventing the peptide from binding to its
target. When Zdk is released from LOV2 by light, the peptide becomes flexible and can bind to its target.
This approach has been used to generate photoactivatable inhibitors of endogenous Cdc42, PP1, JNK
and Gqα. The last approach is to decrease unwanted background activity of our previously published
LANS system. LANS is a LOV2‐based optogentic approach that controls protein nuclear localization with
light (2). To reduce nuclear localization in the dark, Zdk molecules were anchored to mitochondria to
sequester the LANS molecules. Upon irradiation, LANS was released from mitochondria and the NLS
fragment in LANS was also unblocked, resulting in translocation into the nucleus. Zdk is a versatile
optogenetic tool for cellular signal transduction research.
Using subcellular optogenetics to uncover signaling dynamics that control immune cell
P.R. O'Neill1, V. Kalyanaraman1, N. Gautam1,2; 1Anesthesiology, Washington University School of
Medicine, St. Louis, MO, 2Genetics, Washington University School of Medicine, St. Louis, MO
We have developed optogenetic tools that act downstream of native G protein coupled receptors
(GPCRs) and provide direct control over the activity of endogenous heterotrimeric G protein subunits.
These optogenetic tools allowed us to create reversible gradients of intracellular signaling activity in
immune cells. Migratory responses generated by this approach show that a gradient of active G protein
subunits is sufficient to generate directed cell migration. They also provide the most direct evidence so
far for a global inhibition pathway triggered by Gi signaling in directional sensing and adaptation. These
optogenetic tools may additionally be useful for dissecting the functions of G protein signaling at
different intracellular membranes.
Engineered regulation of protein tyrosine phosphatases in living cells.
J.B. Klomp1, V. Huyot1, A.V. Karginov1, A. Ray1; 1Pharmacology, University of Illinois at Chicago, Chicago,
We describe here a new broadly applicable approach to activate the catalytic function of specific protein
phosphatases in living cells. Insertion of an engineered allosteric switch, the iFKBP domain, at a
structurally conserved position within the catalytic domain makes the modified phosphatase inactive.
Treatment with rapamycin triggers interaction with a small FKBP‐rapamycin‐binding domain (FRB) and
restores the activity of the phosphatase. The reagents used in this method are genetically encoded and
membrane permeable, enabling ready application in many systems. Based on the structural similarity of
catalytic domains this method should be applicable to many tyrosine phosphatases. We report here
development of rapamycin‐regulated (RapR) analogs of Shp2, PTP‐PEST and PTP1B. Analysis of RapR‐
Shp2 activity in living cells reveals that it can stimulate endogenous Erk1/2 kinases demonstrating that it
functions similarly to wild type Shp2. Through conjugation of FRB to a selected protein, we are able to
restrict phosphatase activation to a complex with a specific downstream target, and/or to a specific
subcellular location. Using this strategy we demonstrate that targeted activation of RapR‐Shp2 in
complex with focal adhesion kinase downregulates its signaling. This novel approach opens new
opportunities for interrogation of phosphatase‐mediated signaling pathways.
Mapping the human calcineurin phosphatase signaling network through global identification of
short linear motifs that mediate substrate recognition.
M.S. Cyert1, J. Roy1, N. Damle1, C.P. Wigington1, P.M. Kim2,3, N.E. Davey4, Y. Ivarsson5; 1Biology, Stanford
University, Stanford, CA, 2Department of Molecular Genetics, University of Toronto, Donnelly Centre,
Toronto, Canada, 3Department of Computer Science, University of Toronto, Donnelly Centre, Toronto,
Canada, 4Conway Institute of Biomolecular and Biomedical Research, University College Dublin , Dublin,
Ireland, 5Department of Chemistry, Uppsala University, Uppsala, Sweden
Systems‐level analyses of phosphorylation‐based signaling networks has transformed our understanding
of kinase function, but knowledge of phosphatase signaling has lagged behind, primarily because global
approaches to identify phosphatase substrates are lacking. Calcineurin, the conserved Ca2+/calmodulin‐
dependent protein phosphatase and target of immunosuppressants, FK506 and Cyclosporin A, is
ubiquitously expressed, and critically regulates Ca2+‐dependent processes in the immune system, heart,
and brain. However, in the literature only 27 substrates are attributed to calcineurin. Systematic
identification of calcineurin targets is now feasible due to insights into its conserved mechanism of
substrate recognition. Calcineurin acts on phosphosites with little primary sequence similarity; thus
specificity is not encoded within regions contiguous to the phosphosite. Rather, the enzyme binds to
short linear motifs (SLiMs),“PxIxIT” and “LxVP”, which can occur hundreds of residues away from
dephosphorylation sites. CsA , FK506 and the viral A238L protein inhibit calcineurin by blocking SLiM
binding to conserved surfaces on the enzyme. SLiMs are a growing class of sequences that localize
within intrinsically disordered regions, i.e. flexible protein domains that lack a defined structure. SLiMs
mediate most protein‐protein interactions in cells and evolve rapidly to mediate rewiring of signaling
networks, including that of calcineurin. However, degenerate sequences and low affinities for their
target domains make SLiMs challenging to identify. We have used novel experimental and
computational approaches to identify calcineurin‐binding SLiMs systematically in the human proteome.
Proteome peptide Phage Display (ProP‐PD) was used to directly select calcineurin‐binding sequences of
the PxIxIT and LxVP types from predicted disordered regions of the human proteome. We also
developed a novel computational tool to predict PxIxIT sequences, which makes use of their
characteristic structural features (i.e. intrinsic disorder and beta strand formation), and predicts binding
to the conserved PxIxIT‐docking surface on CNA, the calcineurin catalytic subunit. Sequences identified
either experimentally or computationally were confirmed to bind calcineurin in vitro, and their parent
proteins tested for co‐immunoprecipitation with calcineurin in HEK‐293 cells. These studies have
identified new PxIxIT sites in calcineurin substrates, KSR2 and amphiphysin, and have identified more
than 50 new targets for calcineurin, including ion channels, kinases, and receptors, that reveal points of
cross‐talk between calcineurin and other signaling pathways in human cells. Furthermore, our methods
can be applied to systematic characterization of any SLiM‐based signaling network.
Image‐driven Analysis of Molecular Inter‐regulation in Live Cells.
s. lu1, P. Wang1; 1Bioengineering, UCSD, La Jolla, CA
Genetically encoded biosensors based on fluorescence resonance energy transfer (FRET) have been
widely applied to visualize molecular activities in live cells with high spatiotemporal resolution.
However, it remains a challenge to quantitatively evaluate the subcellular coordination among different
molecules based on the dynamic time courses visualize by FRET and fluorescent proteins via live‐cell
imaging. For this purpose, we developed a correlative FRET imaging microscopy (CFIM) approach to
quantitatively analyze the subcellular coordination between the enzymatic Src activation and the
structural FA disassembly. CFIM revealed that only the Src kinase activity within the microdomain of
lipid rafts at the plasma membrane is coupled with FA dynamics in a linear fashion. CFIM further showed
that the level of Src‐FA coupling, as well as the time delay, was regulated by cell‐matrix interactions, as a
tight enzyme‐structure coupling occurred in FA populations mediated by integrin αvβ3, but not in those
by integrin α5β1. Therefore, different FA subpopulations have distinctive regulation mechanisms
between their local kinase activity and structural FA dynamics. Therefore, our work highlights the
importance of dynamic single live‐cell imaging and its integration in‐depth computational analysis.
TFE3 and TFEB regulate autophagy induction, lysosomal biogenesis, and cytokine production in
activated macrophages.
O.A. Brady1, H.I. Diab1, J.A. Martina1, L. Sun1, J. Lim2, N. Raben2, R. Puertollano1; 1Laboratory of Cell
Biology, National Heart, Lung, and Blood Institute, Bethesda, MD, 2Laboratory of Muscle Stem Cells and
Gene Regulation, National Institute of Arthritis and Musculoskeletal and Skin Diseases, Bethesda, MD
Lysosomal gene expression is dynamically regulated in response to nutrient status indicating that cells
monitor lysosomal function and respond to degradation requirements and environmental conditions.
Transcription factor EB (TFEB) was the first reported transcription factor capable of promoting lysosomal
biogenesis and autophagy in response to nutrient levels. Recently, we identified transcription factor E3
(TFE3) as another regulator of lysosomal function that plays a crucial role in nutrient sensing and energy
metabolism. We sought to address whether TFEB and TFE3 contribute to cellular adaptation to stressors
other than starvation. In particular, we studied the role of these transcription factors in immune
response. Innate immune responses are crucial for host defense against invading pathogens. This
defense requires a number of mechanisms including autophagy up‐regulation and cytokine production.
We found that TFE3 translocates from the cytosol to the nucleus in response to toll like receptor
activators and live bacteria in RAW 264.7 cells, microglia, and bone marrow derived primary
macrophages. After 24 hours of lipopolysaccharide (LPS) stimulation, nuclear localization of TFE3 was
comparable to that of mTORC1 inhibition by Torin or starvation. Interestingly, this localization is mTOR
independent, as cells stimulated with LPS retained mTOR activity. In cultured macrophages, LPS induced
autophagy and lysosomal biogenesis by a mechanism dependent upon TFE3, but possibly independent
of TFEB since endogenous levels of TFEB protein were greatly diminished after 24 hours of LPS
stimulation. Further, CRISPR‐TFEB/TFE3 double knockout macrophages were unable to induce
autophagy and lysosomal biogenesis upon LPS stimulation. To understand the transcriptional behavior
of TFE3 upon LPS stimulation we assessed TFE3 gene targets using ChIP‐seq analysis. We found that not
only were the promoters of vacuolar ATPase, lysosome, and autophagic genes occupied by TFE3 in LPS
stimulated cells, but also those involved in host immune response. Relative quantitative RT‐PCR analysis
demonstrated that transcriptional induction of genes involved in the inflammatory response was greatly
hampered in cells deficient in TFE3 or TFEB; and, even more in the TFE3/TFEB double knockout cell lines.
In addition, proteome cytokine arrays revealed that after 24 hours of stimulation with LPS, cells deficient
in TFE3/TFEB expression were unable to induce secretion of key mediators of the inflammatory
response (GM‐CSF, IL‐1β, IL‐2, and IL‐27), macrophage differentiation (M‐CSF), and macrophage
infiltration and migration to sites of inflammation (CCL2). Altogether, our data reveal novel and crucial
roles of TFEB and TFE3 in regulating host immune response to infection.
A novel master adaptor for toll‐like receptors tails pro‐inflammatory responses.
L. Luo1,2, N.J. Bokil1,2, A.A. Wall1,2, M.J. Sweet1,2, J.L. Stow1,2; 1Institute for Molecular Bioscience, The
University of Queensland, Brisbane, Australia, 2IMB Centre for Inflammation Research and Disease, The
University of Queensland, Brisbane, Australia
In innate immunity, danger signals such as Gram‐negative bacterial lipopolysaccharide (LPS) initiate host
inflammatory responses by engaging members of the Toll‐like Receptor (TLR) family. TLR4 activates
inflammatory responses through recruitment of a small number of TIR‐domain containing adaptor
proteins that shape signaling and transcriptional responses. We hypothesized that additional adaptors
may be necessary to diversify cytokine outputs. Here, using pull‐downs and mass spectrometry, we
identified SCIMP as a novel binding partner for TLR4. SCIMP is a transmembrane adaptor and belongs to
a member of the pTRAP family; it is a non‐TIR domain containing protein but we identified that a specific
region at C terminus of SCIMP binds to the TIR domain of TLR4 in an LPS‐induced and novel manner.
SCIMP and TLR4 are co‐located in lipid raft domains on filopodia and surface ruffles of activated
macrophages. SCIMP scaffolds Lyn kinase and is responsible for Lyn‐dependent tyrosine phosphorylation
of TLR4. SCIMP silencing and functional studies show that SCIMP induces pro‐inflammatory signaling and
a surprisingly specific output of pro‐inflammatory cytokines, confined to IL‐6 and IL‐12p40.
Phosphorylation deficient mutations in SCIMP that abrogate TLR4 binding also prevent LPS‐inducible
production of SCIMP‐dependent cytokines. Our studies implicate SCIMP as the adaptor for TLR4 tyrosine
phosphorylation and reveal SCIMP as a wholly new proximal adaptor that imparts remarkable specificity
in downstream inflammatory cytokine responses.
Minisymposium 11: Nuclear Mechanics and Transport
Nuclear Mechanics and Genome Regulation.
G. Shivashankar1; 1Mechanobiology Institute, National University of Singapore, Singapore, Singapore
Extracellular matrix signals sculpt cellular geometry which result in altered functional nuclear landscape
and gene expression. While these alterations regulate diverse biological processes including stem‐cell
differentiation, developmental genetic programs and cellular homeostasis; the mechanisms underlying
such control systems are unclear. I will describe our ongoing work that provides modular links between
cell geometry and nuclear mechanics and its impact on transcription dependent 3D chromosome
organization. In these studies we combine high‐resolution imaging of single cells adhered onto micro
patterned substrates and functional genomics. In particular, I will highlight the mechanisms by which
chromatin dynamics is regulated by cell shape and the accompanying mechanical reorganization of
relative chromosome intermingling to facilitate modular gene expression programs.
Chromatin and lamin A dominate two different regimes of nuclear mechanical force response.
A.D. Stephens1, E. Banigan1,2, L. Almassalha3, Y. Stypula‐Cyrus3, V. Backman3, S.A. Adam4, R.D. Goldman4,
J.F. Marko1,2; 1Department of Molecular Biosciences, Northwestern Universtiy, Evanston, IL,
Department of Physics and Astronomy, Northwestern University, Evanston, IL, 3Department of
Biomedical Engineering, Northwestern University, Evanston, IL, 4Feinberg School of Medicine,
Northwestern University, Chicago, IL
The nucleus must resist constant mechanical inter‐ and intracellular forces to maintain spatial
positioning of genes, as well as transduce forces to dictate proper gene expression. In many human
diseases, including aging, heart disease and cancer, the nucleus exhibits severe shape change along with
alterations in both lamins and chromatin. The intermediate filament lamin A is a well‐known mechanical
component of the nucleus that when altered is proven to perturb wild type nuclear mechanics.
However, the mechanical contribution of chromatin in both healthy and diseased nuclei is relatively
unknown. We developed a novel technique to dissect a single nucleus from a live mammalian cell and
perform micromanipulation force measurements via stretching the whole nucleus with micropipettes.
Using a third “spray” micropipette, we can dynamically treat the isolated nucleus with biochemicals to
compact or digest the chromatin, enabling us to separate the relative contributions of chromatin and
lamin A to nuclear mechanical response. We also utilize drug treatments and genetic approaches to
increase or decrease the amount of heterochromatin or lamin A in the nucleus. Here we show that
chromatin, along with its compaction/histone modification state, is a major mechanical component of
the nucleus. Chromatin is necessary for and dominates force response at small extensions while lamin A
induces strain stiffening and determines the majority of force response at longer extensions. We have
constructed a simple physical simulation model of the nucleus that recapitulates the qualitative
behavior of our experimental results and provides a framework to interpret experimental data.
Altogether, our research suggests that alterations of chromatin compaction and quantity of lamin A that
occur in many human diseases differentially disrupt proper nuclear force response.
Formin mDia2 mediated nuclear actin regulates CENP‐A deposition at centromeres.
C. Liu1, Y. Mao1; 1Department of Pathology and Cell Biology, Columbia University Medical Center, New
York, NY
Epigenetic landscape of the chromosome is well inherited over cellular and organismal generations,
largely independent of underlying DNA sequences. In higher eukaryotes, centromeres are epigenetically
defined by nucleosomes containing the histone H3 variant CENP‐A. In order to maintain centromere
identity against CENP‐A dilution due to S‐phase genome replication, newly synthesized CENP‐A proteins
are deposited at preexisting centromeres specifically during early G1 phase of each cell cycle. However,
how CENP‐A deposition is physically controlled remains unclear. Here, we identified the mammalian
diaphanous formin mDia2 to be essential for CENP‐A loading. Quantitative imaging, pulse‐chase analysis
and high resolution ratiometric live cell profiling demonstrated that mDia2 is required for newly
synthesized CENP‐A to be stably deposited at G1 centromeres. This novel function of mDia2 depends on
its nuclear localization and its ability to polymerize actin. Dynamic, short nuclear actin filaments were
observed in live G1 cells, and the percentage of cells containing these nuclear actin filaments is
significantly reduced upon mDia2 depletion. Moreover, ectopic expression of polymerizable actin inside
the nucleus enhances CENP‐A loading at centromeres. To further test the physical role of mDia2
mediated nuclear actin during G1 CENP‐A deposition, single particle tracking of centromere movement
was performed in early G1 cells over the time scale of initial CENP‐A loading. Quantitative analysis
revealed subdiffusive behaviors where normal G1 centromere movements are relatively confined. The
confinement of centromere motion, however, is significantly impaired upon mDia2 knockdown. Finally,
knocking down mDia2 results in prolonged centromere association of HJURP, a chaperone required to
undergo timely turnover to allow for new CENP‐A loading at centromeres. Our findings suggest that the
physical confinement of early G1 centromere is pivotal to the chemical reactivity of CENP‐A deposition,
possibly through timely turnover of HJURP. Nuclear actin polymerized by formin mDia2, in turn,
contributes directly to the confined, viscoelastic environment and thereby regulates the assembly of
new CENP‐A containing nucleosomes to mark centromere’s epigenetic identity.
In vivo single particle imaging of nuclear mRNA export in budding yeast.
A. Lari1, C. Smith2, C. Derrer3, A. Ouwehand2,4, A. Rossouw2,4, M. Huisman2,4, T. Dange4, M. Hopman4, A.
Joseph2, D. Zenklusen5, K. Weis3,6, D. Grunwald2,4, B. Montpetit1; 1Department of Cell Biology, University
of Alberta, Edmonton, AB, 2Department of Biochemistry and Molecular Pharmacology, RNA
Therapeutics Institute, University of Massachusetts Medical School , Worcester, MA, 3Department of
Biology, Institute of Biochemistry, ETH Zurich, Zurich, Switzerland, 4Bionanoscience, Kavli Institute of
NanoScience, Delft University of Technology, Delft, Netherlands, 5Départment de Biochimie et médecine
moléculaire, Université de Montréal, Montréal, QC, 6Department of Cell and Developmental Biology,
University of California, Berkeley, Berkeley, CA
Dozens of mRNA export proteins have been identified; yet the spatial and temporal activities of these
proteins and how they determine directionality of messenger ribonucleoprotein complex (mRNP) export
from the nucleus remain largely undefined. Here we use the bacteriophage PP7 RNA‐labeling system in
live Saccharomyces cerevisiae cells to follow single‐molecule mRNA export events with high spatial
precision and temporal resolution. Using this approach, we tracked hundreds of mRNPs , analyzed the
kinetics of mRNA export events, and characterized the role of the critical mRNA export factor Mex67p.
We show that mRNA export is fast (~200 ms), but release from the nuclear pore is slow. Upon traversing
the nuclear envelope, mRNPs persist at nuclear pores before being released into the cytoplasm,
suggesting that cytoplasmic release of an mRNP is regulated. These interactions might function to
remodel mRNPs for cytoplasmic release or subsequent events in the mRNA life cycle (e.g., translation).
Strikingly, we observe prolonged cytoplasmic docking at nuclear pore complexes (NPCs) and retrograde
transport of mRNPs into the nucleus in a mex67‐5 mutant. Thus proving an essential role for Mex67p in
cytoplasmic mRNP release and directionality of transport. Importantly, this approach provides a
platform upon which to address fundamental questions related to nuclear mRNA export, including
kinetics, regulation, and mechanism(s) of transport through NPCs.
Influenza Virus mRNA Splicing and Export Through Nuclear Speckles.
A. Mor1, A. White1, R. Munoz‐Moreno2, A. Garcia‐Sastre2, B.M. Fontoura1; 1Department of Cell Biology,
University of Texas Southwestern Medical Center, Dallas, TX, 2Department of Microbiology, Department
of Medicine, Division of Infectious Diseases, Global Health and Emerging Pathogens Institute, Mount
Sinai School of Medicine, New York, NY
The M1 and M2 proteins of influenza A viruses are essential for viral trafficking and budding, and are
encoded in the M1 mRNA. While the M1 matrix protein is generated by the unspliced M1 mRNA, the M2
ion channel is encoded when an internal intron is removed. The intranuclear pathway that mediates
influenza mRNA alternative splicing and export has not been elucidated. Using single molecule RNA‐FISH
(smFISH) combined with genetic approaches we show that influenza virus utilizes cellular nuclear
speckles to promote post‐transcription splicing of its M1 mRNA. The template for generation of M1
mRNA, M vRNA, is not found at nuclear speckles indicating that M1 to M2 mRNA splicing occurs post‐
transcriptionally at nuclear speckles. We found that the viral NS1 protein and its cellular interacting
binding protein NS1‐BP promote nuclear speckles entry of viral M1 mRNA. The RNA binding protein
hnRNP K, which interacts with NS1‐BP and with M1 mRNA, then mediates M1 mRNA splicing into M2
mRNA at nuclear speckles. We also show that the SON protein, which is important for nuclear speckle
formation, is part of the pathway that promotes splicing of M1 mRNA. M1 mRNA that fails to localize at
nuclear speckles is not spliced and accumulates in the nucleoplasm. Additionally, we found that the
mRNA export factors Aly/REF and UAP56 are required for both spliced M2 and unspliced M1 mRNA
export to the cytoplasm. M1 mRNA export inhibition appears to enhance M1 mRNA splicing, a process
that is dependent on NS1‐mediated M1 mRNA accumulation at speckles. In sum, we reveal here an
intranuclear trafficking mechanism that is important for splicing of a subset of viral and likely cellular
Nucleic Acid‐Programmed RNA Tracking in Living Cells with CRISPR/Cas9.
D.A. Nelles1,2, M. Fang2, M.R. O'Connell3, S.J. Markmiller2, J.A. Doudna3, G. Yeo1,2,4; 1Materials Science
and Engineering Graduate Program, University of California, San DIego, San Diego, CA, 2Cellular and
Molecular Medicine, University of California, San Diego, San Diego, CA, 3Departments of Molecular and
Cell Biology and Chemistry, University of California, Berkeley and Howard Hughes Medical Institute,
Berkeley, CA, 4Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore
RNA‐programmed genome editing using CRISPR/Cas9 from Streptococcus pyogenes (spCas9) has
enabled rapid and accessible alteration of any genomic locus in a variety of organisms. The spCas9 target
search mechanism begins with recognition of a short DNA motif known as the protospacer adjacent
motif (PAM) with subsequent interrogation of the adjacent DNA sequence by the Cas9‐bound single
guide RNA (sgRNA). An analogous means to target RNA also determined by nucleic acid specificity would
allow alteration and imaging of endogenous RNA transcripts, but most RNA tracking methods rely on
incorporation of exogenous tags. Recent work has demonstrated the ability of spCas9 to bind RNA in
purified cell extracts while avoiding encoding DNA by supplying the PAM as part of an oligonucleotide
(PAMmer) hybridized to the target RNA. We have recently achieved recognition of RNA in living cells
programmed by a PAMmer and sgRNA with nuclease‐inactive spCas9 fused to fluorescent proteins. This
approach, termed RNA‐targeted Cas9 (RCas9), supported tracking of endogenous RNAs with signal
distributions that correlate highly with fluorescence in situ hybridization. We have also tracked RNA
translocation in living cells to stress granules. Our results establish RCas9 as a means to bind and track
RNA in living cells in a programmable manner without the requirement of genetically encoded tags and
indicates the potential of spCas9 as a means to manipulate the transcriptome in addition to the genome.
Regulation of intracellular transport and diffusion.
K. Weis1, R.P. Joyner1, J.H. Tang1, E.M. Dultz1; 1Department of Biology, ETH Zurich, Zurich, Switzerland
The organization and biophysical properties of the cytosol govern molecular transport and interactions
within the cell. Little is known whether cells have mechanisms by which they globally regulate cytosolic
properties and rates of intracellular diffusion and transport. We provide evidence that the cytosol of
budding yeast can undertake a dramatic transition upon starvation, which severely confines the mobility
of large macromolecules. This affects the transport of macromolecules though the nuclear pore complex
and reduces the movement of both chromatin in the nucleus and mRNPs in the cytoplasm. The global
macromolecular confinement can be explained by a reduction of cytoplasmic volume and subsequent
molecular crowding. These results reveal a novel mechanism by which cells globally alter their
biophysical properties to establish a unique homeostasis during starvation.
Minisymposium 12: Organelle Dynamics, Structure, and Function
Biogenesis of the eukaryotic carbon‐concentrating organelle.
L.C. Mackinder1, M.T. Meyer2, T. Mettler‐Altmann3, V.K. Chen1,4, M.C. Mitchell2, O.D. Caspari2, E.S.
Freeman Rosenzweig1,4, L. Pallesen1, A. Itakura1,4, G. Reeves1, R. Roth5, F. Sommer3, S. Geimer6, T.
Mühlhaus 3, M. Schroda3, U. Goodenough5, M. Stitt3, H. Griffiths2, M.C. Jonikas1; 1Plant Biology, Carnegie
Institution for Science, Stanford, CA, 2Department of Plant Sciences, University of Cambridge,
Cambridge, United Kingdom, 3Max Planck Institute of Molecular Plant Physiology, Potsdam‐Golm,
Germany, 4Department of Biology, Stanford University, Stanford, CA, 5Department of Biology,
Washington University, St. Louis, St. Louis, MO, 6Cell Biology Electron Microscopy, University of
Bayreuth, Bayreuth, Germany
One of the most remarkable feats of photosynthesis is that it has completely changed the composition
of our planet's atmosphere, causing a precipitous drop in CO2 and a dramatic rise in O2 concentrations.
These new conditions have promoted the evolution of carbon concentrating mechanisms (CCMs) that
enhance the ability of organisms to assimilate CO2 from their environment to feed the carbon‐fixing
enzyme Rubisco. Approximately one‐third of the carbon fixation on the planet is mediated by a
eukaryotic algal‐type CCM. The heart of the algal CCM is a poorly‐characterized organelle called the
pyrenoid, which contains tightly packed Rubisco. Little is known about the pyrenoid's protein
composition or biogenesis. We have discovered a novel pyrenoid protein critical for pyrenoid biogenesis.
We present evidence for this protein's central function in the carbon concentrating mechanism and will
propose a model for its mechanism of action.
A sol‐gel transition of the cytoplasm driven by adaptive intracellular pH changes promotes entry
into dormancy.
M.C. Munder1,2, D. Midtvedt1, T.M. Franzmann2, E. Nüske2, L. Malinovska2, O. Otto3, E. Ulbricht3, J.
Guck3, V. Zaburdaev1, S. Alberti2; 1Max Planck Institute for the Physics of Complex Systems, Dresden,
Germany, 2Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany,
Biotechnology Center, Technische Universität Dresden, Dresden, Germany
Cells can enter into a dormant state when faced with unfavorable conditions. However, how cells enter
into and recover from this state is still poorly understood. Here, we study dormancy in different
eukaryotic organisms and find it to be associated with a significant decrease in the mobility of organelles
and foreign tracer particles—a phenomenon we term cytoplasmic freezing. We show that cytoplasmic
freezing is caused by a marked acidification of the cytoplasm, which leads to widespread
macromolecular assembly of proteins and triggers a transition of the cytoplasm from a sol to a gel state
with increased mechanical stability. We further demonstrate that gel formation is required for cellular
survival under conditions of starvation. Our findings have broad implications for understanding
alternative physiological states, such as quiescence and dormancy, and create a new view of the
cytoplasm as an adaptable fluid that can reversibly transition into a protective gel state.
Mitochondrial fission and fusion dynamics are required to generate network topologies that are
robust, efficient and well‐distributed in budding yeast cells.
M.P. Viana1,2, I.A. Mueller1,2, C. Goul1,2, S.M. Rafelski1,2; 1Developmental and Cell Biology, University of
California Irvine, Irvine, CA, 2Center for Complex Biological Systems, University of California Irvine, Irvine,
Mitochondria form dynamic interconnected networks that undergo balanced fission and fusion
dynamics, constantly remodeling/reshaping network topology. The acts of dividing and fusing
mitochondria are implicated in the segregation of unhealthy mitochondria away from the rest of the
network. However, how the network topology itself, a consequence of fission and fusion dynamics,
contributes to mitochondrial function is unclear. We performed the first quantitative analysis of 3D
mitochondrial network topology in budding yeast using our MitoGraph software. We analyzed
mitochondrial networks in the presence and absence of fission and fusion (Δdnm1Δfzo1 mutant) in cells
that are either rapidly undergoing polarized growth or maintained in an unbudded ‘steady state’. We
compared experimental networks to computational simulations. We found that mitochondrial networks
obey the same scaling laws as other physical networks (e.g. street networks) independent of fission and
fusion dynamics. However, in the absence of fission and fusion, mitochondrial networks display altered,
tree‐like topologies and are composed of much longer tubules compared to wild‐type networks. This
altered topology reduces the efficiency and speed of diffusive processes spreading through Δdnm1Δfzo1
networks and makes the networks more sensitive to random connection failures. Fission and fusion are
also required to generate mitochondrial networks with connective properties that are insensitive to the
amount of mitochondria within the cell. Finally, fission and fusion are necessary to distribute
mitochondria uniformly throughout the cell. Together these results reveal the importance of the
underlying network topology, generated by fission and fusion dynamics, for creating the robust, efficient
and well‐distributed mitochondrial networks required for proper mitochondrial function.
Dynamic actin cycling through mitochondrial subpopulations regulates mitochondrial
A.S. Moore1, Y.C. Wong1, E.L. Holzbaur1; 1Physiology, University of Pennsylvania, Philadelphia, PA
To maintain cellular homeostasis, mitochondria undergo dynamic fission/fusion events. We used live cell
imaging to investigate the dynamics of mitochondria in the cell, and to probe the role of the
cytoskeleton in regulating the fission/fusion balance. We monitored actin dynamics in interphase HeLa
cells using either GFP‐LifeAct or GFP‐actin, and identified a robust but transient assembly of actin
filaments on the outer membrane of a subpopulation of cellular mitochondria, labeled with the marker
sBFP‐mito. Actin dynamically assembled around ~10% of the cellular pool of mitochondria; this assembly
was blocked by either latrunculin B or CK‐666, indicating that the active assembly of actin around
mitochondrial subpopulations is mediated by Arp2/3. Actin disassembly from these mitochondria
occurred within 3‐5 min, followed by reassembly of the labeled actin around a neighboring population
of mitochondria. Over time, actin cycled through the entire mitochondrial population, at a rate of 14
min/cycle. Cycling through mitochondrial subpopulations followed a stochastic one‐dimensional random
walk model. Actin polymerization around mitochondrial subpopulations led to the localized
fragmentation of the normal long tubular morphology of these organelles, with significant decreases in
both length and area of actin‐associated mitochondria. Live cell imaging indicated that this
fragmentation was due to a decrease in the extent of fusion, likely due to the decreased motility
observed for mitochondria enmeshed in an actin filament network. This decrease in fusion was
paralleled by a significant increase in the likelihood of fission events. Once actin disassembled from the
affected subpopulation, fusion rates were rapidly restored, leading to a regional recovery of the tubular
mitochondrial network within 5 min. Thus, cycles of actin assembly/disassembly promote localized and
transient fragmentation of mitochondrial networks that spatially regulates mitochondrial morphology
and may facilitate mitochondrial quality control mechanisms. Actin recruitment did not appear to result
from the transient depolarization of individual mitochondria. However, whole‐cell depolarization with
CCCP rapidly blocked actin cycling, inducing more stabilized actin filament assembly around all cellular
mitochondria within 2 min. This rapid response led to overall fragmentation of the tubular
mitochondrial network followed by eventual mitophagy of the mitochondrial fragments. Together, these
studies highlight actin cycling as a dynamic regulator of mitochondrial morphology and identify a novel
mechanism involved in the maintenance of mitochondrial homeostasis in the cell.
MICOS coordinates with respiratory complexes and lipids to establish mitochondrial inner
membrane architecture.
J.R. Friedman1, A. Mourier2, J. Yamada1, J. McCaffery3, J. Nunnari1; 1MCB, University of California, Davis,
CA, 2Department of Mitochondrial Biology, Max Planck Institute for Biology of Ageing, Cologne,
Germany, 3Integrated Imaging Center, Johns Hopkins University, Baltimore, MD
The conserved MICOS complex functions as a primary determinant of mitochondrial inner membrane
structure. We address the organization and functional roles of MICOS and identify two independent
MICOS subcomplexes: Mic27/Mic10/Mic12, whose assembly is dependent on respiratory complexes and
the mitochondrial lipid cardiolipin, and Mic60/Mic19, which assembles independent of these factors.
Our data suggest that MICOS subcomplexes independently localize to cristae junctions and are
connected via Mic19, which functions to regulate subcomplex distribution, and thus, potentially also
cristae junction copy number. MICOS subunits have non‐redundant functions as the absence of both
MICOS subcomplexes results in more severe morphological and respiratory growth defects than
deletion of single MICOS subunits or subcomplexes. Mitochondrial defects resulting from MICOS loss are
caused by misdistribution of respiratory complexes in the inner membrane. Together, our data are
consistent with a model where MICOS, mitochondrial lipids and respiratory complexes coordinately
build a functional and correctly shaped mitochondrial inner membrane.
Multiple dynamin family members collaborate to drive mitochondrial division.
J.E. Lee1, G. Voeltz1; 1Molecular, Cellular, Developmental Biology, University of Colorado Boulder,
Boulder, CO
The current model of mitochondrial division represents dynamin‐related protein 1 (Drp1) as the sole
dynamin family member responsible for driving membrane fission. However, in vitro studies identified
that Drp1‐mediated constriction was unable to induce spontaneous fission of liposomes. The biophysical
limitations of Drp1 suggest that other constriction mechanisms are necessary to complete mitochondrial
division. Thus, we tested whether another dynamin, which is capable of driving fission in vitro, might
play a role at mitochondria. We discovered that classical dynamin‐2 (Dyn2) depletion phenocopies Drp1
depletion resulting in increased mitochondrial length in multiple cell types. This phenotype is rescued by
re‐expression of wild‐type Dyn2, but not by dominant‐negative K44A or the CMT‐associated G358R
mutant. Dyn2 depletion does not disrupt Drp1 recruitment to mitochondrial constrictions, which
suggests that Dyn2 is acting downstream of Drp1. Dyn2‐depletion also prevents mitochondrial fission in
the presence of BAPTA or Staurosporine, which would normally stimulate Drp1‐dependent division.
Finally, we show that Dyn2 is recruited to constrictions with the timing of Drp1‐dependent division.
Therefore, our model is that Dyn2 and Drp1 work in concert to orchestrate sequential constriction
events leading to mitochondrial division.
Changing the paradigm for Drp1 oligomerization during mitochondrial fission: roles for actin
and myosin II at specific stages in the process.
W. Ji1, A. Hatch1, H. Higgs1; 1Biochemistry, Dartmouth Medical School, Hanover, NH
Mitochondrial fission is a fundamental cellular process, and is required for mitochondrial inheritance
and for maintenance of mitochondrial and cellular homeostasis through mitophagy and apoptosis.
Defects in mitochondrial fission are linked to neurodegenerative diseases such as Alzheimer’s,
Parkinson’s, Huntington’s, and Charcot‐Marie‐Tooth disease. The dynamin GTPase Drp1 plays a critical
role during mitochondrial fission, during which Drp1 oligomerizes as a ring around the mitochondrion,
then hydrolyzes GTP to constrict its ring. Drp1 ring constriction constricts the underlying mitochondrion.
The current paradigm for Drp1 recruitment to the fission site is that Drp1 comes directly from the
cytosol. Using live‐cell microscopy, we find a very different sequence of events during Drp1
oligomerization. A broad range of mitochondrially‐associated Drp1 oligomers exists on the
mitochondrial outer membrane, varying in size, motility, and morphology. Large Drp1 assemblies arise
through incorporation of smaller Drp1 oligomers that are already mitochondrially‐bound, through a
process we call “maturation”. Surprisingly, most Drp1 oligomers are non‐productive for fission, and a
substantial lag time occurs between stable Drp1 oligomer assembly and fission. Previously, we have
shown that actin filaments (assembled by the endoplasmic reticulum‐bound formin INF2) and myosin II
promoted mitochondrial Drp1 accumulation and mitochondrial fission (Korobova, Ramabhadran &
Higgs, 2013 Science). To investigate the sequence of events in fission more precisely, we developed a
method to stimulate mitochondrial fission rapidly. We find that stimulation results in rapid
accumulation of actin filaments (28.1 ± 19.8 sec), and subsequent accumulation of mitochondrial Drp1
oligomers (99.9 ± 19.0 sec) at fission sites. Inhibiting actin polymerization, myosin II, or INF2 reduces
both stimulation‐induced Drp1 accumulation and mitochondrial fission. An additional finding, revealed
by super‐resolution microscopy, is that Drp1 oligomers can translocate along the mitochondrial outer
membrane at up to 46 nm/sec, with a characteristic motility pattern. Based on these findings, we
propose a three‐step sequence for Drp1 oligomerization on the mitochondrial outer membrane: 1)
RECRUITMENT of small oligomers to the outer membrane, 2) MATURATION of a stable Drp1 oligomer
by incorporation of smaller, mitochondrially‐bound Drp1 oligomers; and 3) CONVERSION of the Drp1 to
a fission‐competent state. Actin filaments promote specific steps in this process through direct binding
to Drp1 at the fission site, which we show in biochemical experiments.
SPD‐2/CEP192 and CDK are limiting for microtubule organizing center function at the
R. Yang1, J.L. Feldman1; 1Biology, Stanford University, Stanford, CA
The centrosome acts as the microtubule organizing center (MTOC) during mitosis in animal cells.
Microtubules are nucleated and anchored by γ‐tubulin ring complexes (γ‐TuRCs) embedded within the
centrosome’s pericentriolar material (PCM). The PCM is required for the localization of γ‐TuRCs and
both are steadily recruited to the centrosome, culminating in a peak in MTOC function in metaphase. In
differentiated cells, the centrosome is often attenuated as an MTOC and MTOC function is reassigned to
non‐centrosomal sites such as the apical membrane in epithelial cells, the nuclear envelope in skeletal
muscle, and down the lengths of axons and dendrites in neurons. Hyperactive MTOC function at the
centrosome is associated with epithelial cancers and with invasive behavior in tumor cells. Little is
known about the mechanisms that limit MTOC activation at the centrosome. Here, we find that MTOC
function at the centrosome is completely inactivated during cell differentiation in C. elegans embryonic
intestinal cells and MTOC function is reassigned to the apical membrane. In cells that divide after
differentiation, the cellular MTOC state switches between the membrane and the centrosome. Using
cell fusion experiments in live embryos, we find that the centrosome MTOC state is dominant and that
the inactive MTOC state of the centrosome is malleable; fusion of a mitotic cell to a differentiated or
interphase cell results in rapid reactivation of the centrosome MTOC. We show that conversion of MTOC
state involves the conserved centrosome protein SPD‐2/CEP192 and CDK activity from the mitotic cell.
Temporal And Spatial Dynamics Of Centrosome Assembly In C. elegans.
T. Laos1, G. Cabral1, A. Dammermann1; 1Max F. Perutz Laboratories, University of Vienna, Vienna, Austria
Centrosomes are cytoplasmic organelles whose primary function is to organize the microtubule
cytoskeleton. As such, they play important roles in cell division, directing chromosome segregation as
well as the partitioning of cytoplasmic contents including cell fate determinants. Centrosomes are
formed by the recruitment of microtubule‐organizing pericentriolar material (PCM) to a pair of
centrioles whose duplication once per cell cycle ensures the formation of a bipolar spindle in mitosis.
While the molecular mechanisms underlying centriole duplication are increasingly well characterized,
how centrioles recruit PCM and how this material is organized remains poorly understood.
In this study we examine the dynamics of PCM assembly in the one‐cell C. elegans embryo, taking
advantage of the substantial size of the centrosome in this experimental model (~60x larger in volume
than that of vertebrate somatic cells or Drosophila embryos), which affords much better resolution for
light microscopy studies. We focus our analysis on SPD‐5, a protein which is essential for PCM assembly
in C. elegans and one of the best candidates for a building block or scaffold component of the PCM. We
show that, in contrast to Cnn in Drosophila syncytial embryos, SPD‐5 is recruited throughout the volume
of the PCM rather than specifically at centrioles. Consistent with it forming part of a stable scaffold
structure, SPD‐5 displays no cytoplasmic exchange subsequent to its incorporation into the PCM.
Further, we find no evidence for rearrangements, or ‘flux’, within the PCM.
These findings have a number of important implications for the nature of the PCM and the role of
centrioles in its assembly. First and foremost, there is no privileged role for centrioles in PCM
recruitment. Rather, the PCM expands isotropically by incorporation of additional SPD‐5 throughout the
PCM volume. Unlike the growth of crystals or typical polymers, the PCM lattice must be able to stretch
to accommodate additional subunits, implying hitherto unsuspected internal flexibility. Centrioles
clearly do play an important role in controlling the rate and location of PCM assembly. However, they
must be able to perform their role at a distance, potentially via a diffusible signal such as Plk1. We are
currently examining the contribution of centrioles and centriole‐localized regulators using the
quantitative live microscopy assays we have developed.
Kaluza Minisymposium
Target the adaptability of heterogeneous aneuploidy.
G. Chen1,2, W.A. Mulla1,3, A. Kucharavy1,3, H. Tsai1,3, B. Rubinstein1, J. Conkright1, S. McCroskey1, D.
Bradford1, L. Weems1, J. Haug1, C. Seidel1, J. Berman4, R. Li1,2,3; 1Stowers Institute for Medical Research,
Kansas City, MO, 2Molecular and Integrative Physiology, University of Kansas Medical Center, Kansas
City, MO, 3Department of Cell Biology, Johns Hopkins University, Baltimore, MD, 4Department of
Molecular Microbiology and Biotechnology, Tel Aviv University, Ramat Aviv, Israel
A mutation of overwhelming size, aneuploidy (chromosomal copy number alteration), was the first type
of genetic abnormality identified in human cancer, discovered by Theodore Boveri in 1914, a pre‐
double‐helix era then. In mammalians, studies from pioneers ascertained the tie between tumor
development and chromosomal instability (CIN) which produces aneuploidy. However, in cellular level,
evidence was missing to prove the advantage of CIN and the resulted karyotype heterogeneity in
adaptation. We demonstrated, in yeast, that the cell population with high karyotype diversity left many
folds more survivors than the normal population under various unrelated stress conditions, such as
fluconazole, benomyl, etc. Furhter, canalization for specific karyotype was noticed for different selective
conditions, a direct evidence supporting the stress‐adaptive role of aneuploidy.
By altering dosage of large number of genes, karyotype diversity produces phenotype diversity fueling
cellular adaptation. By analyzing the phenotype heterogeneity (growth standard deviation) of a
heterogeneous aneuploidy population under various conditions, I found a peculiar scaling: the more
heavily the overall growth is suppressed by stress, the more heterogeneous the phenotype is expressed
by aneuploidy population.
Even though superior adaptability caused by karyotype diversity can fuel cellular adaptation, the
karyotype diversity shrunk after selection, leaving a targetable population. For example, under near‐
lethal dose of Hsp90 inhibitor radicicol (Stress X), all yeast survivors shared gain of Chr XV, despite varied
dosage of other chromosomes. The channeled population offered a more approachable target than the
random aneuploidy: when I applied a compound (Stress Y: hygromycin) selecting against the gain of Chr
XV simultaneously with radicicol, the heterogeneous aneuploidy population was eradicated by the
evolutionary trap.
Through computational analysis of pharmaceutical profiling datasets for aneuploidy human cancer cell
lines, I found that: First, stress (drug treatment) also exaggerates phenotype heterogeneity in
aneuploidy cancer cells, as in yeast; secondly, some types of cancer exhibit convergent dosage change of
a specific chromosome under tumor evolution pressure (Stress X), yielding a new class of potential
target for drug (Stress Y). For example, the analysis suggests that Chr 7p gain, identified in 80% of brain
tumors, associated with hypersensitivity towards topotecan, a FDA‐approved cancer drug.
These works from yeast model to human cells, outline a potentially applicable strategy to eradicate
heterogeneous aneuploidy with highly challenging adaptability.
Disentangling the Gordian knot: From understanding to preventing the tumorigenicity of
human pluripotent stem cells.
U. Ben‐David1,2; 1Department of Genetics, The Hebrew University of Jerusalem, Jerusalem, Israel,
Cancer Program, Broad Institute of Harvard and MIT, Cambridge, MA
The promise of human pluripotent stem cells (hPSCs) for the future of regenerative medicine has been
long cast by major safety concerns. My Ph.D. thesis explored the genomic instability and tumorigenicity
of hPSCs, and introduced novel strategies to eliminate undifferentiated tumorigenic cells.
One troubling aspect for the translational applications of hPSCs is their genomic instability. To address
this issue, we developed a new methodology to analyze chromosomal integrity using global gene
expression data. We used this technology to dissect the chromosomal integrity of hPSCs, and revealed
three distinct mechanisms that promote their genomic instability. Cross‐species comparison of
chromosomal aberrations in PSCs revealed both evolutionarily‐conserved and species‐specific
aberrations. By functionally analyzing aneuploid hPSCs, we found that aneuploidy induced profound
changes in the gene expression signature, the proliferation rate and the in vivo tumorigenicity of the
cells. These findings led us to assess the genomic integrity of human adult stem cells (hASCs), which are
already in clinical use. We revealed that hASCs could rapidly acquire chromosomal aberrations in
culture, and that the same aberrations recur in tumors of the respective cell lineage. Our novel
methodology is now freely available, and can be implemented by every lab to detect aneuploid cell
The tumorigenic risk of hPSCs, however, is not limited to aneuploid cells; when differentiated cells are
injected into patients, residual undifferentiated cells may form tumors, even in the absence of
chromosomal changes. It is therefore essential to get rid of residual hPSCs prior to transplanting the final
cell product into patients. We therefore sought to devise robust chemical ways to selectively ablate
tumorigenic hPSCs. We identified a tight‐junction protein, CLDN6, which is highly‐specific for
undifferentiated hPSCs, and targeted it to selectively eradicate hPSCs. We also performed an unbiased
high‐throughput screen of >50,000 small molecules and identified 15 pluripotent stem cell‐specific
inhibitors (PluriSIns). These compounds selectively eliminated hPSCs without harming a large array of
progenitor and differentiated cells. Cellular and molecular analyses identified the target of the most
selective molecule, PluriSIn#1, as the metabolic enzyme SCD1, revealing a unique role for lipid
metabolism in hPSCs. Exposure of cell cultures to PluriSIn#1 completely prevented teratoma formation.
PluriSIn#1 is now commercially available, and is commonly used to generate hPSC‐free cell cultures.
This work has therefore expanded our understanding of the tumorigenic potential of hPSCs, and
presented novel technologies to selectively eliminate undifferentiated cells from culture.
The kinetochore encodes a mechanical toggle‐switch to control the spindle assembly
P. Aravamudhan1, A.P. Joglekar2, A.A. Goldfarb2; 1Department of Pediatrics, Vanderbilt University, Ann
Arbor, TN, 2Cell and Developmental Biology, University of Michigan, Ann Arbor, MI
During cell division, kinetochore attaches duplicated sister chromatids with spindle microtubules to
enable their equal segregation. In the absence of microtubule attachment, kinetochore activates the
Spindle Assembly Checkpoint (SAC), which stalls cell division. An unattached kinetochore recruits an
array of proteins that generate the biochemical SAC signal, and these SAC proteins are removed
following microtubule attachment leading to SAC silencing. Although the biochemistry of SAC signaling
has been well established, how the kinetochore communicates its attachment status to the SAC
machinery remained a mystery. Attachment‐induced mechanical changes within the multi‐protein
architecture of the kinetochore have been proposed to provide the cue to the SAC machinery. However,
identifying whether and how such architectural changes control the SAC biochemistry has been difficult
to address due to the molecular complexity of the kinetochore. To address this, we utilized budding
yeast as the model system, where the architecture is well understood. Within the ~ 80 nm length of the
yeast kinetochore‐microtubule attachment, we localized key SAC proteins at distinct positions and
studied the effect on SAC signaling. Among the localized proteins, only the master regulator of SAC,
Mps1 kinase activated the SAC from attached kinetochores. Crucially, this SAC activity was position‐
specific: Mps1 activated the SAC only when it was localized proximal to its essential kinetochore
substrate in SAC, Spc105 (KNL1), and it did not activate the SAC when localized away from Spc105 in the
outer kinetochore. Forced localization of the phosphodomain in Spc105 to the same kinetochore
positions yielded reciprocal results. The phosphodomain localized to the outer kinetochore activated the
SAC from attached kinetochores, and not when localized to inner kinetochore. These results
demonstrated that active Mps1 localizes to the outer kinetochore even in the presence of attachment,
likely mediated by its interaction with the Calponin Homology domains in the Ndc80 complex.
Furthermore, this spatial restriction of Mps1 at the outer kinetochore and Spc105 at the inner
kinetochore is essential to prevent phosphorylation of Spc105 in attached kinetochores. These insights
demonstrate that the kinetochore operates the SAC like a mechanical toggle‐switch, with Ndc80 and the
phosphodomain of Spc105 acting as two terminals of the switch. Proximity between these protein
terminals in the absence of attachment allows phosphorylation of Spc105 and triggers SAC signaling.
Attachment‐induced changes in the kinetochore physically separates the two terminals and thereby
prevents phosphorylation to silenc
Cell biology of cheating ‐ mechanism of selfish element transmission through asymmetrical
L. Chmatal1, K. Yang1, R.M. Schultz1, M.A. Lampson1; 1Biology, University of Pennsylvania, Philadelphia,
Mendel’s First Law of genetics states that a pair of alleles or chromosomes segregates randomly during
meiosis so that one copy of each is represented equally in gametes. Whereas male meiosis produces
four equal sperm, in females only one cell survives and the other three degenerate. During meiotic drive
a selfish DNA element exploits such female meiotic asymmetry and segregates preferentially to the egg,
increasing its transmission to the offspring and frequency in the population. Although meiotic drive
affects karyotype evolution in mammals, the cellular mechanism is not known. Rapid and massive
karyotype change is represented by the western house mouse. Karyotype conversion from mostly
telocentric (centromere terminal) to mostly metacentric (centromere internal) typically reflects fixation
of Robertsonian (Rb) fusions, a common chromosomal rearrangement that joins two telocentric
chromosomes at their centromeres to create one metacentric. Although preferential transmission of Rb
fusions during female meiosis can explain their fixation and karyotype change, there was no mechanistic
explanation of why fusions preferentially segregate to the egg in some mouse populations, leading to
karyotype change, while other populations preferentially eliminate the fusions and maintain a
telocentric karyotype. Using laboratory mouse models, we established a system where biased
segregation of Rb fusions can be measured and studied. Our findings suggest a model of how the Rb
fusion can drive in either direction in different mouse populations. To test that model, we captured and
studied wild mice with Rb fusions in their natural habitats in Europe. Based on these results, we show
that differences in centromere strength predict the direction of drive. Stronger centromeres, with higher
kinetochore protein levels and altered interactions with spindle microtubules (MTs), are preferentially
retained in the egg. Rb fusions preferentially segregate to the polar body in laboratory mouse strains
when the fusion centromeres are weaker than those of telocentrics. Conversely, fusion centromeres are
stronger relative to telocentrics in natural house mouse populations that have changed karyotype by
accumulating metacentric fusions. Our data represent the first direct evidence for theoretical
predictions of centromere drive: increased MT binding leads to a transmission advantage during
meiosis. Exploiting the system with Rb fusions allowed us to study general mechanisms of recognizing
and destabilizing erroneous kinetochore‐MT attachments. Improper attachments typically lack tension
between kinetochores and are positioned off‐center on the spindle. Low tension is a widely accepted
mechanism for recognizing errors, but whether chromosome position also regulates MT attachments
was u
Mechanisms of target specificity in eukaryotic gene silencing pathways.
P.A. Dumesic1, H.D. Madhani1; 1Biochemistry and Biophysics, University of California, San Francisco, CA
Eukaryotic genome defense depends on specific silencing of transposons, despite the fact that these
foreign genetic elements are diverse and utilize the same gene expression machinery as do host genes.
Our work explores mechanisms by which two gene silencing pathways—RNA interference (RNAi) and
heterochromatin—achieve target specificity.
Eukaryotes employ RNAi to silence transposons, but how this pathway recognizes its targets is unclear.
Using the yeast Cryptococcus neoformans, we observed that endogenous small interfering RNAs
(siRNAs) are synthesized using unspliced transposon pre‐mRNAs as templates. We also characterized a
nuclear protein complex, SCANR, that mediates siRNA biogenesis and physically associates with the
spliceosome. These findings suggest a model for RNAi in which rapidly spliced transcripts are not
targeted for siRNA synthesis, whereas transposon transcripts, by virtue of their poor splicing kinetics,
accumulate in spliceosomes and are targeted for RNAi by SCANR. Indeed, mutations that stall a
transcript’s splicing increase the level of its corresponding siRNA.
Thus, the suboptimal gene expression signals encoded by foreign genetic elements betray them to the
host RNAi pathway. Our work establishes a new function for introns and the spliceosome in genome
defense, and raises the possibility that the ubiquity of introns in eukaryotic genomes is driven by their
contribution to self/non‐self recognition.
Repetitive elements are also silenced by heterochromatin. To examine how these chromosomal
domains are specified, we established C. neoformans as a model for chromatin biology. We identified
Polycomb group proteins in this organism and showed that they methylate histone H3K27 to establish
facultative heterochromatin. We characterized the Polycomb repressive complex (PRC2) responsible for
this activity and found that it directly binds its own product: H3K27me. Remarkably, disruption of this
product recognition activity causes a genome‐wide redistribution of facultative heterochromatin such
that it inappropriately overlaps with a different chromatin domain: constitutive heterochromatin. This
PRC2 redistribution is suppressed by elimination of constitutive heterochromatin. Therefore, PRC2 can
respond to signals from constitutive heterochromatin, but this latent promiscuity is normally masked by
PRC2’s product recognition activity.
Since product recognition is a widespread feature of chromatin‐modifying complexes, it may represent a
general mechanism to insulate these complexes from incorrect signals and prevent commingling of
distinct chromatin domains. Our work helps explain how chromatin‐modifying enzymes achieve
specificity despite high nuclear concentrations of a largely identical substrate: the nucleosome.
Epigenetic Transmission of a Gene Repression Memory Across Generations and During
L.J. Gaydos1, A. Rechtsteiner2, T. Egelhofer2, S. Strome2; 1Basic Sciences, Fred Hutchinson Cancer
Research Center, Seattle, WA, 2Molecular, Cell and Developmental Biology, University of California,
Santa Cruz, Santa Cruz, CA
To retain cell identity during development, cells must remember patterns of gene expression through
cell division. The MES proteins in Caenorhabditis elegans are key regulators of gene expression in the
germline and are necessary in parents for fertility in the next generation. MES‐2, MES‐3, and MES‐6 form
the C. elegans Polycomb Repressive Complex 2 (PRC2) and generate a repressive histone modification,
methylation on Lysine 27 of Histone H3 (H3K27me), on genes repressed in the germline. MES‐4
generates a different histone modification, H3K36me, on genes expressed in the germline. To further
define the targets of MES protein regulation, determine if they work together to regulate gene
expression, and investigate when they are important during germline development, I used genomics,
genetics, and microscopy approaches. Transcript profiling of dissected mutant germlines revealed that
MES‐2/3/6 and MES‐4 cooperate to promote expression of germline genes and repress somatic genes
and genes on the X chromosome. Loss of MES‐4 from germline genes causes H3K27me to spread to
germline genes, resulting in reduced H3K27me elsewhere on the autosomes and especially on the X.
This finding supports the model that methylation of H3K36 antagonizes methylation of H3K27 on the
same histone tails. My finding that loss of MES‐2, 3, or 6 results in expression from the X chromosomes
and sterility, unless the X chromosome is repressed by other means, showed that the essential role of
MES‐2/3/6 in worms is repression of the X chromosomes in germ cells. I determined that repressive
H3K27me is transmitted to embryos by both sperm and oocytes. By generating embryos containing
some chromosomes with and some lacking H3K27me, I showed that in embryos lacking MES‐2/3/6
enzyme, H3K27me is transmitted to daughter chromatids through several rounds of cell division. In
embryos with MES‐2/3/6 enzyme, the mosaic pattern of H3K27me is perpetuated through
embryogenesis. Subsequently, during germline proliferation in larval stages, H3K27me accumulates on
all chromosomes. During germline proliferation in larvae is also when I found MES‐3 to be most
important for germline development in the next generation. These latter two findings suggest that
germline memory is reset during germline proliferation in the larval stages. Taken together, my findings
support a “germline memory model” in which MES‐4 and H3K36me act as a memory of germline gene
expression and help concentrate MES‐2/3/6 repression on the X chromosomes. MES‐2/3/6 and
H3K27me act as a memory of germline gene repression, most importantly on the X chromosomes. The
germline memory generated by the MES proteins is epigenetically transmitted across generations and is
critical for the proper development of nascent germ cells.
Investigation of coordinated stem cell behaviors in the skin by live imaging.
K.R. Mesa1, V. Greco1; 1Genetics, Yale University, New Haven, CT
Tissue homeostasis is a fundamental metazoan process that requires continuous replacement of cells
lost by either terminal differentiation or cell death. Previous work has described the importance of
tissue resident stem cells in fueling cellular turnover due to their ability to both self‐renew and replace
other cell types. Work from our lab and others has described that stem cells can be directed to perform
either of these behaviors by different environmental cues. Therefore, given that stem cells can choose
to either proliferate or commit toward differentiated cell fates, we hypothesized that the stem cell pool
has a counterbalancing mechanism to prevent stem cell accumulation or depletion. To investigate stem
cell behavior in vivo, we have developed an intravital two‐photon microscopy system to non‐invasively
image the skin of live mice. Here we have utilized the mouse epidermis, which is comprised of a highly
proliferative basal epithelial layer that feeds upward into the terminal differentiated supra basal layers
forming the watertight barrier of our skin. We have found using both genetic and unbiased photo‐
activatable tools that environmental differences such as local cell density and tissue organization can
predict basal cell behavior. Furthermore by in vivo revisits of the same cells, we observe that as basal
cells move into the supra basal layers neighboring basal cells extend laterally to maintain cell‐cell
contacts in the basal layer. Strikingly, by time‐lapse recordings we find laterally extended basal cells
undergo cell divisions oriented along their long‐axis. Through genetic and mechanical manipulation of
cell density we find that loss of terminal differentiated layers promote upward departure and
differentiation of basal cells, which leaves remaining basal cells to expand laterally and divide to balance
any loss in tissue cellularity. This study identifies a regulatory relationship of cell differentiation and
division in the skin epidermis and provides a model for how cells achieve coordinated tissue
Molecular mechanisms of vinculin activation and nanoscale organization at focal adhesions.
L.B. Case1, M.A. Baird1,2, G. Shtengel3, S.L. Campbell4, M.W. Davidson2, H.F. Hess3, C.M. Waterman1; 1Cell
Biology and Physiology Center, NHLBI, National Institutes of Health, Bethesda, MD, 2National High
Magnetic Field Laboratory, The Florida State University, Tallahassee, FL, 3 , Janelia Farm Research
Campus, HHMI, Ashburn, VA, 4Biochemistry and Biophysics, University of North Carolina at Chapel Hill,
Chapel Hill, NC
Focal adhesions (FAs) are multi‐protein complexes that link the extracellular matrix (ECM) to the actin
cytoskeleton to mediate cell adhesion, migration, mechanosensing and signaling. Previously,
interfermoteric photoactivatable localization microscopy (“iPALM”), which uses single molecule
localization to measure lateral position with ~20nm accuracy and interferometry to measure the axial
position with ~10nm accuracy, revealed that FAs have a conserved laminar structure, with an integrin
signaling layer (ISL) ~10‐20 nm from the plasma membrane, an actin regulatory layer (ARL) ~100 nm
from the membrane that extends into the stress fiber, and a force transduction layer (FTL) that spans
the intervening space. However, it was previously unknown how the layered organization of proteins
within FAs contributes to the regulation of protein activity and function. Vinculin (Vcl) is a protein that
undergoes a conformational change at FAs to interact with multiple binding partners and regulate
diverse cellular functions including protrusion, migration, mechanosensing and survival. Vcl has over 10
binding partners distributed throughout the FA including paxillin in the ISL, talin in the FTL, and actin in
the ARL. We hypothesized that the spatial compartmentalization of different binding partners into
distinct FA layers contributes to Vcl activation, regulation and functional specificity at FAs. To test this,
we used point mutants to perturb specific protein interactions and assayed Vcl nanoscale localization
and Vcl activation state at FAs. We used iPALM to image the 3D localization of vinculin molecules with
nanometer accuracy as well as a FRET biosensor to assay vinculin conformational changes. By
systematically perturbing specific protein interactions, we found that inactive Vcl localizes in the lower
ISL by binding to phospho‐paxillin. Talin binding activates Vcl and targets active Vcl >10nm higher in FAs
where Vcl can engage retrograde actin flow. Furthermore, the upwards repositioning of Vcl is
dependent on Myosin‐II activity. By performing complimentary experiments with iPALM and FRET, we
have been able to study the mechanism of Vcl activation at a structural level in a cellular setting. We
have found that specific protein interactions are spatially segregated within FAs at the nanoscale to
regulate Vcl activation and function.
Oral Presentations‐Tuesday, December 15
Symposium 5: Bending Nature to Our Purposes: Engineering of Cells and Tissues
CRISPR‐Cas Genome Surveillance: From Basic Biology to Transformative Technology.
J.A. Doudna1; 1Molecular Cell Biology and Chemistry, Howard Hughes Medical Institute, Berkeley, CA
The advent of facile genome engineering using the bacterial RNA‐guided CRISPR‐Cas9 system in animals
and plants is transforming biology. I will present a brief history of CRISPR biology from its initial
discovery through the elucidation of the CRISPR‐Cas9 enzyme mechanism, providing the foundation for
remarkable developments using this technology to modify, regulate or mark genomic loci in a wide
variety of cells and organisms. These results highlight a new era in which genomic manipulation is no
longer a bottleneck to experiments, paving the way to both fundamental discoveries in biology, with
applications in all branches of biotechnology, and strategies for human therapeutics. Recent results
regarding the molecular mechanism of Cas9 and its use for targeted cell‐based therapies will be
Giving New Life to Materials for Energy, the Environment and Medicine.
A. Belcher1; 1Biological Engineering, MIT, Cambridge, MA
Organisms have been making exquisite inorganic materials for over 500 million years. Although these
materials have many desired physical properties such as strength, regularity, and environmental benign
processing, the types of materials that organisms have evolved to work with are limited. However, there
are many properties of living systems that could be potentially harnessed by researchers to make
advanced technologies that are smarter, more adaptable, and that are synthesized to be compatible
with the environment. One approach to designing future technologies which have some of the
properties that living organisms use so well, is to evolve organisms to work with a more diverse set of
building blocks. The goal is to have a DNA sequence that codes for the synthesis and assembly of any
inorganic material or device. We have been successful in using evolutionarily selected peptides to
control physical properties of nanocrystals and subsequently use molecular recognition and self‐
assembly to design biological hybrid multidimensional materials. These materials could be designed to
address many scientific and technological problems in electronics, military, medicine, and energy
applications. Currently we are using this technology to design new methods for building batteries, fuel
cells, solar cells, carbon sequestration and storage, enhanced oil recovery, catalysis, and medical
diagnostics and imaging. This talk will address conditions under which organisms first evolved to make
materials and scientific approaches to move beyond naturally evolved materials to genetically imprint
advanced technologies with examples in lithium ion batteries, lithium‐air batteries, dye‐sensitized solar
cells, and ovarian cancer imaging.
Hydrogels as synthetic extracellular matrices: from tissue engineering to 4‐D cell biology.
K. Anseth1; 1Chemical and Biological Engineering, University of Colorado and HHMI, Boulder, CO
Methods for culturing mammalian cells in a biologically relevant context are increasingly needed to
study cell and tissue physiology, expand and differentiate progenitor cells, and to grow replacement
tissues for regenerative medicine. Two‐dimensional culture has been the paradigm for in vitro cell
culture; however, evidence and intuition suggest that cells behave differently when they are isolated
from the complex architecture of their native tissues and constrained to petri dishes or material surfaces
with unnaturally high stiffness, polarity, and surface to volume ratio. As a result, the field is often faced
with the need for more physiologically relevant culture environments, and many researchers are
realizing the advantages of hydrogels as a means of creating custom 3D microenvironments with highly
controlled chemical, biological and physical cues. Further, the native ECM is far from static, so ECM
mimics must also be dynamic to direct complex cellular behavior, so called 4D biology. In general, there
is an un‐met need for materials that allow user‐defined control over the spatio‐temporal presentation of
important signals, such as integrin‐binding ligands, growth factor release, and biomechanical signals.
Developing such hydrogel mimics of the ECM for 4D cell culture is an archetypal engineering problem,
requiring control of numerous properties on multiple time and length scales important for cellular
functions. New materials systems have the potential to significantly improve our understanding of how
cells receive information from their microenvironment and the role that these dynamic processes may
play in controlling the stem cell niche to cancer metastasis. This talk will illustrate recent efforts to
advance hydrogel chemistries for 3D cell culture, as well as dynamic control of biochemical and
biophysical properties through orthogonal, photochemical click reaction mechanisms.
Symposium 6: Going the Distance: Determining Size and Spacing of Biological
Mechanisms of mitosis and size control in Xenopus.
M. Strzelecka1, R. Heald1; 1Molecular and Cell Biology, University of California, Berkeley, Berkeley, CA
The goal of my laboratory is to elucidate the molecular mechanisms of cell division and size control. Our
interdisciplinary approaches take advantage of cytoplasmic extracts prepared from eggs of the frog
Xenopus laevis that reconstitute mitotic chromosome condensation and spindle assembly and function
in vitro. To study mechanisms of spindle and organelle size control, we utilize a smaller, related frog,
Xenopus tropicalis, to investigate interspecies scaling, and extracts prepared from fertilized eggs at
different stages of embryogenesis to study developmental scaling. We seek to provide new insight into
the underlying principles of spindle assembly and biological size control, as well as the molecular basis of
variation that contributes to genomic instability and evolution.
Turing pattern formation without diffusion.
S. Kondo1; 1Frontier Bioscience, Osaka University, Osaka, Japan
The reaction‐diffusion mechanism, presented by AM Turing more than 60 years ago, is currently the
most popular theoretical model explaining the biological pattern formation including the skin pattern.
This theory suggested an unexpected possibility that the skin pattern is a kind of stationary wave (Turing
pattern or reaction‐diffusion pattern) made by the combination of reaction and diffusion. At first,
biologists were quite skeptical to this unusual idea. However, the accumulated simulation studies have
proved that this mechanism can not only produce various 2D skin patterns very similar to the real ones,
but also predict dynamic pattern change of skin pattern on the growing fish. Now the Turing’s theory is
accepted as a hopeful hypothesis, and experimental verification of it is awaited. Using the pigmentation
pattern of zebrafish as the experimental system, our group in Osaka University has been studying the
molecular basis of Turing pattern formation. We have identified the genes related to the pigmentation,
and visualized the interactions among the pigment cells. With these experimental data, it is possible to
answer the crucial question, “How is the Turing pattern formed in the real organism?” The
pigmentation pattern of zebrafish is mainly made by the mutual interactions between the two types of
pigment cells, melanophores and xanthophores. All of the interactions are transferred at the tip of the
dendrites of pigment cells. In spite of the expectation of many theoretical biologists, there is no
diffusion of the chemicals involved. However, we also found that the lengths of the dendrites are
different among the interactions, which can substitute the difference of diffusion constant in the RD
model. Therefore the real mechanism we found in the zebrafish skin is not the classic RD mechanism,
but is mathematically equivalent to the original Turing mechanism.
Microsymposium 13: Morphology of the Cytoskeleton Leading to Morphology in
Association of the RNAi machinery with the zonula adherens regulates growth‐related signaling.
A. Kourtidis1, J.M. Carr1, I.K. Yan1, T. Patel1, E.A. Thompson1, P.Z. Anastasiadis1; 1Department of Cancer
Biology, Mayo Clinic, Jacksonville, FL
Cell‐cell junction formation at the apical zonula adherens (ZA) is critical for epithelial tissue development
or maintenance and is perturbed during cancer progression. E‐cadherin (Ecad) and p120 catenin (p120)
are core components of the ZA in epithelial cells. Recently, a novel Ecad‐p120 partner called PLEKHA7
was identified to support the integrity of the ZA by tethering the microtubules to the junctions. Here, we
reveal that PLEKHA7 associates with two major complexes of the RNA interference (RNAi) machinery at
the ZA of non‐transformed epithelial cells, the microprocessor and the RISC complex, to suppress pro‐
tumorigenic signaling via miRNAs. By using confocal microscopy of polarized epithelial monolayers, co‐
immunoprecipitation (IP), and proximity ligation assays, we show that PLEKHA7 co‐localizes and co‐IPs
with a non‐nuclear subset of the core microprocessor proteins DROSHA and DGCR8. In addition, using
the same assays and by performing proteomics we found that PLEKHA7 associates with all the major
RISC components, including Ago2, GW182 and PABPC1, at the ZA. Recruitment of DROSHA, DGCR8 and
Ago2 to the ZA is PLEKHA7‐ and p120/cadherin‐dependent, opposed by active Src, and independent of
microtubules. PLEKHA7 also co‐precipitates with primary miRNAs (pri‐miRNAs), which are the
microprocessor substrates, and possesses pri‐miRNA processing activity, as indicated by RNA‐IP and
processing activity assays. Depletion of PLEKHA7 results in mislocalization and decreased processing
activity of the junctional microprocessor. Examination of miRNA expression by Nanostring revealed that
PLEKHA7 regulates the levels of a specific set of miRNAs, including miR‐24, miR‐30a, miR‐30b and let‐7g.
We show that this regulation occurs at the pri‐miRNA processing level, at least for miR‐30b. As a result,
PLEKHA7 suppresses anchorage‐independent growth (AIG) and expression of growth‐related markers,
including SNAI1, MYC, and CCND1, via these miRNAs and particularly miR‐30b. Indeed, both the mature
miR‐30b and the SNAI1 mRNA, which are substrates of the RISC complex, co‐precipitate with PLEKHA7
and localize at the ZA by in situ hybridization assays, in addition to phosphorylated S387‐Ago2 that
promotes miRNA‐mediated, mRNA silencing. By interrogating the RNAi machinery for the first time in
the context of normal or non‐transformed epithelial cells, the present work a) describes a new
mechanism through which the ZA regulates cellular behavior via its surprising association with RNAi
complexes, and b) reveals the presence of the otherwise nuclear microprocessor outside the nucleus, as
well as of the cytoplasmic RISC, at the ZA.
Cell shape changes required for brain morphogenesis are mediated by calcium signaling and
non‐muscle myosin II.
S.U. Sahu1, C. Kwas1, M.R. Visetsouk1, R.J. Garde1, J.H. Gutzman1; 1Biological Sciences, University of
Wisconsin‐Milwaukee, Milwaukee, WI
Elucidating the molecular mechanisms that regulate the cell shape changes required for shaping brain
tissue during development is critical because normal brain function requires proper formation of brain
shape. Cell shape changes are mediated by mechanical forces generated within individual cells, leading
to cytoskeletal rearrangements that are integrated to effect whole tissue shape changes. The generation
of force within a cell often depends on motor proteins, particularly non‐muscle myosins. We are using
zebrafish to study the molecular mechanisms that regulate the cell shape changes that lead to the first
fold in the vertebrate brain, the highly conserved midbrain‐hindbrain boundary. The contractile state of
the neuroepithelium is tightly regulated by non‐muscle myosin II activity; therefore, we tested the role
of non‐muscle myosin IIA (NMIIA) (myh9a and myh9b) and non‐muscle myosin IIB (NMIIB) (myh10) in
regulating cell shape changes that occur during midbrain‐hindbrain boundary morphogenesis.
Knockdown of NMIIA and NMIIB indicated that both proteins are required for normal midbrain‐
hindbrain boundary tissue angle formation. However, quantification of cell shapes revealed that NMIIA
and NMIIB have a differential role in regulating midbrain‐hindbrain boundary formation. NMIIA is
required for the shortening of cells specifically at the midbrain‐hindbrain boundary constriction, while
NMIIB is required for the proper width of cells throughout the midbrain‐hindbrain boundary region.
Non‐muscle myosin II proteins are tightly regulated via the phosphorylation of their associated myosin
regulatory light chains; however, the upstream signaling pathways that initiate differential regulation of
cell shape during midbrain‐hindbrain boundary morphogenesis are not known. Our current studies have
revealed that calcium signaling is critical for the regulation of cell length, but not cell width, during
midbrain‐hindbrain boundary formation. In particular, manipulation of cytosolic calcium levels resulted
in abnormal midbrain‐hindbrain boundary cell length and inhibition of cytosolic calcium rescued the cell
length phenotype observed in embryos with over activation of non‐muscle myosin II. In addition, we
found that calcium signals mediate phosphorylation of myosin light chain in the midbrain‐hindbrain
boundary region. Together these data suggest that modulation of myosin activity by calcium signals may
be critical for proper regulation of cell length to determine embryonic brain shape and possibly for
shaping other epithelial cells and tissues throughout development.
Epithelial cell migration in the intestinal villi depends on actin‐driven cell protrusions and
mitotic pressure in the crypts.
D. Krndija1, E. Hannezo2, S. Richon1, A. Simon1, D.M. Vignjevic1; 1UMR 144, Institut Curie, Paris, France,
Cavendish Laboratory, University of Cambridge, Cambridge, United Kingdom
Small intestinal epithelium is one of the most proliferative tissues in the body, and the entire epithelium
is renewed every week. Epithelial renewal in gut homeostasis is driven by cell division in the intestinal
crypts, coupled with cell migration along the villi, and cell shedding at the tip of the villi. Mechanisms
responsible for the migration of intestinal epithelial cells from the crypt to the villus tip remain largely
unknown. Dynamic analysis of cell migration in the gut has never been performed due to limitations in
imaging tools.
The mechanisms proposed to explain cell migration in gut homeostasis include mitotic pressure and
active cell movement, among others. Our objective is to experimentally test these hypotheses, using
adult murine small intestine as a model. To this end, we have developed long‐term imaging of living
intestinal tissue ex vivo in 3D, using bi‐photon confocal microscopy. Double‐fluorescent reporter mice
were generated, allowing us to label individual epithelial cells and observe cell migration and membrane
protrusions in high contrast. These unprecedented imaging experiments revealed that enterocytes
migrate collectively and directionally, with a variable velocity – moving slower at the base and faster
towards the villus tip. We also demonstrate for the first time that epithelial cells’ basal protrusions are
dynamic and enriched in actin. Treatment with CK666, an Arp2/3 inhibitor, significantly slowed down
cell migration along the villus, implying an important role for actin and dynamic actin protrusions.
To determine the role of mitotic pressure, we specifically ablated dividing cells in the crypts using
hydroxyurea, a mitotic inhibitor. This short‐term treatment did not abolish cell migration – the cells still
migrated, but with a slower velocity. We also determined that cell density in untreated tissue was higher
at the base and tip of the villi, compared to the middle. Interestingly, the cell density changed after the
hydroxyurea treatment, with a significant decrease in the basal and middle part of the villus.
In addition, we are developing a computational model for gut cell migration that accounts for cell
adhesions, division and differentiation. The model predicts that the observed differences in cell density
along the villus could not be the result of mitotic pressure acting alone.
In summary, our data suggest that epithelial cells migrate both actively, using actin‐rich basal
protrusions, as well as passively, pushed by dividing cells in the crypt. The observed differences in cell
densities and velocities along the villus suggest that mitotic pressure is acting at the villus base, whereas
actin‐driven migration is predominant in the rest of the villus.
Muscle‐specific ribosome synthesis coordinates overall body growth and development in
Drosophila by regulating systemic insulin signaling.
S.S. Grewal1, A. Ghosh1; 1Biochemistry and Molecular Biology, University of Calgary, Calgary, AB
As they develop, multicellular organisms need to coordinate growth among tissues and organs in order
to achieve proper body size and proportions. This coordination relies on organ‐to‐organ communication
and endocrine signaling. Drosophila larvae have emerged as an excellent model system to study how
inter‐organ signaling controls body growth. During the four‐day larval period, animals increase in mass
almost 200‐fold. This dramatic growth is nutrition‐dependent and involves activation of the conserved
TOR kinase signaling. While TOR is required for cell‐autonomous growth in almost all tissues, TOR
activity in specific tissues can also coordinate overall body size through endocrine or systemic effects.
Here we describe our work identifying how muscle‐specific ribosome synthesis controls body growth
and development in larvae in response to nutrient/TOR signaling. One important growth‐regulatory
target of TOR signaling is ribosome biogenesis. Studies in yeast and mammalian cell culture have
described how TOR controls rRNA synthesis ‐ a limiting step in ribosome biogenesis ‐ via the Pol I factor
TIF‐IA. We find that nutrient starvation and TOR inhibition lead to reduced levels of TIF‐IA, and
decreased rRNA synthesis in larval muscle. When we mimic this decrease in muscle ribosome synthesis
using RNAi‐mediated knockdown of TIF‐IA, we observe delayed development and reduced body size.
This reduction in growth is caused by lowered systemic insulin signaling – the major endocrine regulator
of body growth in larvae. This reduced insulin signaling occurs via two endocrine responses: reduced
expression of Drosophila insulin‐like peptides (ILPs) from the brain and increased expression of Imp‐L2 ‐
a secreted factor that binds and inhibits ILP activity ‐ from muscle. We also find that maintaining TIF‐IA
levels in muscle partially restores the starvation‐mediated suppression of insulin signaling. Finally, we
show that activation of TOR specifically in muscle increases body size and that this requires TIF‐IA. Our
data suggest that muscle ribosome synthesis functions as a nutrient‐dependent checkpoint for overall
body growth: in nutrient rich conditions, TOR maintains TIF‐IA levels and ribosome synthesis to promote
normal systemic insulin signaling. But upon starvation, reduced muscle ribosome synthesis triggers an
endocrine response that limits systemic insulin signaling to restrict growth and maintain homeostasis.
This work emphasizes how tissue‐specific changes in protein synthesis can exert non‐autonomous
effects on overall body growth.
Mechanics serves as an instructional cue driving heart progenitor cells to undergo a
mesenchymal‐to‐epithelial transition during early heart morphogenesis.
T.R. Jackson1, H. Kim1, L.A. Davidson1,2,3; 1Department of Bioengineering, University of Pittsburgh,
Pittsburgh, PA, 2Department of Developmental Biology, University of Pittsburgh, Pittsburgh, PA,
Department of Computational and Systems Biology, University of Pittsburgh, Pittsburgh, PA
Congenital heart defects, often resulting from errors in early heart development, occur in 40,000
newborns per year in the U.S. Although recent studies have highlighted the importance of how
mechanics drive organogenesis, the physical mechanisms that underlie early heart formation remain
poorly understood. Bilateral fields of heart progenitor cells (HPCs) originate in the anterior lateral plate
mesoderm at the end of gastrulation, undergo a mesenchymal‐to‐epithelial transition (MET), then
merge into a single population on the ventral midline to initiate heart formation. Previous work with the
amphibian model Xenopus laevis has shown that mechanical cues can drive MET in embryonic
mesenchymal cells. In this study we test the hypothesis that mechanical cues instruct MET in HPCs and
perturbation of that MET, through changes in tissue mechanics, causes heart defects due to altered cell
behaviors. Using live intravital long‐term confocal time lapse imaging we assessed cooperative cell
movements with a novel nuclei‐tracking algorithm and calculated directional traction forces with ex vivo
traction force microscopy (TFM). Trajectory analysis combined with TFM reveals that early HPCs are first
mechanically passive in their movements then after MET, begin to migrate collectively and actively exert
traction. To determine the role of cell and tissue contractility in the onset of MET we treated embryos
with small molecules to perturb bulk tissue stiffness. MET progression was monitored by localization of
ZO‐1 and aPKC, and extracellular matrix proteins, including fibronectin and fibrillin. HPCs in embryos
treated with 100uM blebbistatin or 50uM Y27632 to decrease tissue stiffness showed nearly two‐fold
reduced ZO‐1 and aPKC levels and localization while those treated with 40nM calyculin A to increase
tissue stiffness had slightly increased localization of epithelial markers. All small molecule treatments
resulted in heart defects and the failure of cardiomyocytes to incorporate into the adult heart. Our
findings expose the role played by biomechanics in driving MET during early heart organogenesis and
provide deeper insights into the cellular etiology of congenital heart defects.
Mechanical strain guides the formation of global planar axis in ciliated epithelia.
Y. Chien1, R. Keller2, C. Kintner1, D. Shook2; 1Molecular Neurobiology Laboratory, Salk Institute for
Biological Studies, La Jolla, CA, 2Department of Biology, University of Virginia, Charlottesville, VA
Cilia on multiciliated epithelia orient and beat along a common planar axis. The core planar cell polarity
(PCP) pathway patterns the planar axis through local cell‐cell interactions. However, the cues that guide
the PCP across an entire tissue are largely unknown. Here we use the Xenopus embryonic epidermis,
which produces directed ciliary flow at the tailbud stage, to study the mechanism of global planar axis
formation. The planar axis of Xenopus embryo is fixed after gastrulation. We found that two early
features of planar axis establishment, microtubule (MT) polarization and the formation of stabilized PCP
protein complexes, aligned with the anterior‐posterior axis during gastrulation. When we eliminated
anterior‐posterior and dorsal‐ventral patterning signals with UV irradiation after fertilization, the
resulting ventralized embryos still gastrulated and produced directed cilia flow. MT alignment and stable
PCP complexes were also detected in these ventralized embryos along the animal‐vegetal axis, but only
in the region where mechanical strain on the tissue builds up as a result of gastrulation movement.
Explants of embryo skin prepared before gastrulation that did not undergo mechanical strain were thus
unable to align microtubules or form PCP stable complexes. By applying exogenous strain on these
explants, we restored MT alignment and PCP stable complexes along the direction of strain.
Furthermore, exogenous strain applied on embryos during gastrulation changed the planar axis and
ciliary flow direction of ciliated epithelia. We thus propose that the mechanical strain from gastrulation
is sufficient to guide the global planar axis for ciliated epithelia.
Gut on the chip: Extracellular matrix composition and crypt‐villus topography dictate the
growth and maintenance of intestinal epithelium.
M. Verhulsel1, A. Simon2, D. Ferraro1, C. Bureau1, J. Viovy1, S. Descroix1, D.M. Vignjevic2; 1UMR 168,
Institut Curie, Paris, France, 2UMR 144, Institut Curie, Paris, France
The epithelium of the small intestine is composed of a single layer of epithelial cells lining the villi that
project into the lumen of the gut, and the crypts that descend into the underlying connective tissue.
Dividing stem cells are contained within the crypts and give rise to five types of specialized epithelial
cells. Most of those cells travel upwards from the crypt towards the villus tip where they shed into the
lumen. The basement membrane underlines the basal surface of epithelium and separates it from the
stroma mostly composed of collagen I and fibroblasts.
The whole intestinal epithelium is renewed every week. Many biochemical pathways that control
intestinal homeostasis are discovered using mouse models. In contrast, in vitro models systems, such as
organoids, provide a mean to investigate questions hard to be addressed in vivo. Despite their obvious
interest, organoids do not fully recapitulate intestinal features: the total number of cells does not
remain constant, villi‐like structures are missing as well as cells and matrix constitutive of the stroma.
To address this lack, we developed an innovative device that recapitulates both the composition and
topography of the intestinal lining. We microstructured collagen I scaffolds that respect the anatomic
dimensions of mice intestine (3D sinusoid with 400µm period and 400µm height) by adapting methods
from soft lithography field. Primary fibroblasts were embedded in collagen scaffolds coated with thin
and homogeneous layer of laminin that replicated basement membrane. Early‐stage cyst‐like organoids
were seeded on the scaffolds. They opened up, spread and, after 7 days, successfully colonized the
whole structure forming a confluent epithelial monolayer. Thus, our microfabrication protocol
permitted to reproduce the anatomical compartmentalization observed in vivo where basement
membrane separates fibroblasts residing in the stroma from epithelial cells. In vivo‐like intestinal
features were successfully obtained as proliferative cells were spatially restricted to the crypts. The
presence of fully differentiated cells such as Goblet and Paneth cells attested of the functionality of the
3D intestinal tissue engineered. Our preliminary results suggest that fibroblasts promote epithelial cell
migration and are actively involved in the maintenance of stem cell proliferation. Thanks to the flexibility
of our platform, we are currently investigating the influence of biophysical parameters of the matrix
(geometry and stiffness) on intestinal homeostasis.
Altogether, our device provides the first in vitro model of functional intestinal tissue with physiologically
relevant geometries and microenvironment paving the way to both biological and biophysical
Microsymposium 14: Actin Cytoskeleton Dynamics
Sharpin is a novel activator of the Arp2/3 complex.
M.H. Khan1,2, S. Salomaa1, A. Augenlicht1, T. Deguchi3, g. jacquemet4, e. kremneva5, a. byron4, M.J.
Humphries4, P. Hänninen3, P. Lappalainen5, J. Pouwels1; 1Centre for Biotechnology, University of Turku,
Turku, Finland, 2Turku Doctoral Programme of Molecular Medicine, University of Turku, Turku, Finland,
Laboratory of Biophysics, University of Turku, Turku, Finland, 4Faculty of Life Sciences, University of
Manchester, Manchester, United Kingdom, 5Institute of Biotechnology, University of Helsinki, Helsinki,
Sharpin is an important integrin inhibitor (Rantala, Pouwels, et al., Nature Cell Biol. 2011), which
promotes lymphocyte detachment during transmigration (Pouwels et al., Cell Rep. 2013). In addition,
Sharpin has oncogenic properties, Sharpin knockout mice display a psoriasis‐like phenotype with chronic
inflammation and, in addition to its integrin‐inhibitory function, Sharpin is a component of the linear
ubiquitin chain assembly complex (LUBAC), which enhances signal‐induced Nuclear factor‐kappaB (NF‐
κB) activity. We performed a mass spectroscopy screen that identified many novel Sharpin interactors
involved in a wide array of cellular processes. One of these novel Sharpin interactors is the Arp2/3
complex, which catalyzes branching of actin filaments and promotes formation of lamellipodia,
filopodia, and peripheral and dorsal ruffles. Super‐resolution STED microscopy showed that Sharpin and
Arp2/3 colocalize in distinct spots along actin filaments, which are enriched in the cell periphery. The
interaction between Arp2/3 and Sharpin was confirmed using proximity ligation assays, fluorescence
resonance energy transfer (FRET/FLIM) and co‐immunoprecipitation. Interestingly, inhibitors of actin
polymerization (Cytochalasin D) or Arp2/3 activity (CK666) demonstrated that the Arp2/3‐Sharpin
interaction depends on an intact cytoskeleton and Arp2/3 activity. We mapped the Arp2/3 interaction
site to the conserved ubiquitin like domain of Sharpin and, importantly, identified a Sharpin mutation
(V240A/L242A) that disrupts Arp2/3 binding but does not affect the ability of Sharpin to inhibit integrins
or support LUBAC function. Functionally, Arp2/3‐dependent formation of peripheral and dorsal ruffles is
inhibited in the absence of Sharpin. Importantly, expression of WT GFP‐Sharpin, but not GFP‐
Sharpin(V240A/L242A), rescued peripheral ruffle formation in Sharpin‐silenced cells, showing that
Sharpin‐mediated peripheral ruffle formation depends on the Sharpin‐Arp2/3 interaction. In addition,
Sharpin deficient cells spread slower after release from CK666‐mediated Arp2/3 inhibition, suggesting
that Sharpin also promotes Arp2/3 activity in lamellipodia. All in all, we identify Sharpin as a novel
interactor of the Arp2/3 complex and demonstrate that Sharpin promotes Arp2/3‐dependent cell
spreading and formation of peripheral and dorsal ruffles.
Nuclear actin interactome reveals new functions for actin in the nucleus.
T. Viita1, G. Huet1, H. Asan‐Liski1, A. Hyrskyluoto1, J. Virtanen1, M. Varjosalo1, M.K. Vartiainen1; 1Institute
of Biotechnology, University of Helsinki, Helsinki, Finland
Even though actin is still best known from its actions in the cytoplasm, recent studies have strengthened
our understanding about the role of actin inside the nucleus, where it has been linked to many different
nuclear processes from transcription to chromatin remodeling. However, molecular mechanisms behind
these events are still unclear. To tackle this problem, we have analyzed the nuclear actin interactome
both at the level of the genome (ChIP‐seq) and proteins. We have used two mass spectrometry (MS)
based techniques: Strep‐HA affinity purification (AP‐MS) and BioID to identify proteins that interact with
nuclear actin. These two methods complemented each other’s weaknesses and gave us broad and less
bias view about nuclear actin interactome. We have also utilized different actin constructs to
discriminate nuclear vs. cytoplasmic interactions and also to assess the requirement for actin
polymerization for the putative interactions. Initial validation of the hits was done with light microscopy
technique Bimolecular Fluorescent Complementation (BiFC).
The two MS screens had significant number of shared hits, but the BioID approach gave much broader
interactome than the AP‐MS, likely because the BioID allowed us to catch transient interactions as well
as the more stable ones. Shared hits from both screens clarified the role of actin in the different
chromatin remodeling complexes, including a novel actin‐containing complex hATAC, and support the
notion that actin acts as a monomer in these complexes. Moreover, the data as a whole indicates that
actin prefers to interact with most of the nuclear complexes as a monomer. Broader interactome from
BioID data linked actin with novel functions in DNA replication, DNA damage response and in RNA
processing. Indeed, change in nuclear actin levels has an effect on mRNA processing. Interestingly, our
ChIP‐seq data on genome‐wide actin distribution on chromatin revealed that actin peaks seemed to
correlate more with introns than exons. Taken together, this nuclear actin interactome analysis will be
the first step towards understanding the molecular mechanism by which actin operates in the nucleus,
and has suggested novel roles for nuclear actin beyond gene expression.
Palladin promotes actin polymerization at pointed ends.
R. Gurung1, R. Yadav1, M.R. Beck1; 1Chemistry, Wichita State University, Wichita, KS
Palladin is a recently discovered actin binding protein that plays a key role in both normal cell migration
and invasive cell motility, yet its precise function in organizing the actin cytoskeleton is unknown. The
majority of this previous research has focused on the scaffolding activity of palladin, whereas our results
here add a new dimension to the relationship between palladin and actin by highlighting the direct role
in actin assembly. Here we show that the C‐terminal immunoglobulin‐like domain of palladin, which is
directly responsible for actin binding and bundling, also stimulates actin polymerization in vitro. Palladin
eliminates the lag phase that is characteristic of the slow nucleation step of actin polymerization and
dramatically reduced depolymerization. Actin growth was not inhibited by barbed‐end blockers, thus
palladin appears to stimulate filament growth from the pointed‐end. Microscopy and in vitro
crosslinking assays reveal differences in actin bundle architecture when palladin is incubated with actin
before as opposed to after polymerization. Similar to metal ions, palladin also appears to stimulate a
polymerization‐competent form of G‐actin, either through charge neutralization or conformational
changes. We also explore palladin’s role in actin‐based motility by examining the involvement of
palladin during the formation of actin comet tails by Listeria monocystogenes, a model system for
studying dynamic actin polymerization. I will discuss our current progress in utilizing NMR spectroscopy,
microscopy and kinetic assays that integrate protein structure with cytoskeletal dynamics. The affects of
mutations, lipid binding, and phosphorylation on palladin structure and function will also be presented,
where the goal is to understand how palladin regulates actin polymerization. Together with the
localization of palladin in Z‐disks, podosomes, and stress fibers, our results demonstrate that palladin
stimulates actin polymerization and indicate that palladin is part of an actin organization nucleation
Identification of the possible states of organisation of the actin cytoskeleton using high content
image screening of a high diversity chemical library.
N.S. Bryce1, A. De Laurentiis1,2, T. Failes3, G.M. Arndt3, J.R. Stehn1, E.C. Hardeman1, P.W. Gunning1;
School of Medical Sciences, UNSW Australia, Sydney, Australia, 2Department of Clinical and
Experimental Medicine, University Magna Graecia of Catanzaro, Cantanzaro, Italy, 3ACRF Drug Discovery
Centre for Childhood Cancer, Children’s Cancer Institute Australia, Lowy Cancer Research Centre, UNSW
Australia, Sydney, Australia
The local organisation of the actin cytoskeleton is responsible for the regulation of many cellular
processes such as cell motility, adhesion, endocytosis and invasion. Dysregulation of actin organisation
is seen in many tumour types and disease processes, often through the involvement of actin binding
proteins. Due to the high sequence similarity between the 6 actin isoforms expressed in humans, actin
itself is not a highly sought after drug target due to the cardiotoxic side effects of current actin‐targeting
drugs. Actin binding proteins are a diverse group of proteins that have distinct functions and tissue and
cellular expression profiles, and as such present as more attractive therapeutic targets. We have
screened a chemical library to identify compounds that reorganise the actin cytoskeleton in ways
different from known anti‐actin drugs. Using a high content image‐based screening approach, we have
screened over 70,000 chemicals from the 115,000 compound WECC diversity library. Changes in
filamentous actin structures within the cells were easily visualised by staining with fluorescently labelled
phalloidin. Through the use of custom designed image analysis algorithms and hierarchical cluster based
analysis in 95‐parameter space, we have identified >800 compounds that impact actin filament
organisation. Interestingly, the compounds identified fall into discrete clusters of 25 different
phenotypes. This suggests that there are a discrete number of ways that polymeric actin can be
reorganised in a mammalian cell. This screen has identified new compounds that can break down or
enhance stress fibres, induce lamellipodia or filopodia formation, induce nuclear filamentous actin
structures, alter filaments associated with organelles such as the Golgi and alter cortical actin. These
compounds will be a great resource to the actin community to understand the molecular mechanisms
underlying the organisation of the actin cytoskeleton.
EB1, CLIP‐170, and mDia1 trigger ultrafast actin filament polymerization from microtubule plus
J.L. Henty‐Ridilla1, A. Rankova1, J. Eskin1, B.L. Goode1; 1Biology, Brandeis University, Waltham, MA
Dynamic microtubule plus ends are thought to govern the remodeling of actin networks in diverse
biological settings, yet the mechanisms underlying such regulation have remained elusive. Here, we
show that the microtubule plus end‐associated protein CLIP‐170 links the two polymer systems by
forming a high affinity complex with Diaphanous‐family formins that promotes ultrafast actin
polymerization from growing microtubule plus ends in the presence of EB1. Using single‐molecule
fluorescence microscopy, we directly observe mDia1 dimers and CLIP‐170 dimers forming complexes
that track the growing ends of actin filaments for minutes. Together, they promote actin polymerization
at the fastest rates observed to date, 187 subunits s‐1 µM‐1, and simultaneously enhance protection from
capping protein by almost 4‐fold. This ‘ultrafast’ actin filament elongation depends on profilin, the FH1
and FH2 domains of mDia1, and specific sequences in CLIP‐170 located outside of its microtubule‐ and
EB1‐binding regions. Using TIRF microscopy co‐reconstitution assays that simultaneously monitor actin
polymerization and microtubule dynamic instability, we observed mDia1‐CLIP‐170 complexes being
recruited to microtubule plus ends by EB1. This triggers almost immediate and ‘ultrafast’ polymerization
of new actin filaments from growing microtubule plus ends, with the mDia1‐CLIP‐170 complex shifting
from the microtubule surface to the barbed ends of the actin filaments and with EB1 maintaining
attachment of the actin filament pointed ends to the microtubule plus end. This work establishes both a
novel mode of formin regulation and one of the first molecular mechanisms for coordinating and
physically linking dynamic actin filaments with growing microtubules.
Myosin VIII Links Actin to Microtubules During Polarized Growth.
S. Wu1, M. Bezanilla1; 1Biology, University of Massachusetts Amherst, Amherst, MA
Plants are an excellent system to study the interaction between the actin and microtubule
cytoskeletons. Plants only have two families of actin‐based molecular motors: class VIII and class XI
myosins. In our lab we have generated a plant that lacks the entire family of class VIII myosins and
discovered that myosin VIII links microtubules to the actin cytoskeleton. Using a functional myosin VIII‐
GFP fusion, we recently demonstrated that myosin VIII localizes to the ends of microtubules where it
interacts with actin to guide cell division. We also observed that myosin VIII null plants grow slower
than wild type plants. During polarized growth, myosin VIII is found at the ends of cytoplasmic
microtubules that are focused behind the growing tip, where there is an accumulation of actin filaments.
In wild type cells, the accumulation of actin at the tip appears as a motile actin spot, which is persistent
and stays close to the apex. Movement of the actin spot correlates with the direction of growth. In the
absence of myosin VIII, the actin spot is rarely present at the tip. Instead, actin filaments often
accumulate in random spots further back in the cell, which quickly disappear after formation. We also
observed abnormal cortical actin in these cells, consisting of huge arrays of actin bundles that move
along the shank of the cell. The abnormal actin array found in myosin VIII null plants can be
recapitulated in wild type cells by treating with the microtubule‐depolymerizing drug oryzalin,
suggesting microtubules are required for proper actin organization and myosin VIII is the link between
actin and microtubule in this process. Additionally, when myosin VIII null cells encounter an obstacle
during growth, selection of a new site of polarized growth is impaired leading to branched cells.
Together, these results suggest that myosin VIII coordinates the interaction between the microtubule
and actin cytoskeletons. Our data suggest that microtubules may deliver polarity‐establishing factors
that guide subsequent stabilization of actin filaments responsible for delivering secretory vesicles to the
growth site thereby optimizing rates and directionality of growth.
Molecular mechanisms of force transmission through linkers of the nucleoskeleton and
Z. Jahed1, M. Mofrad1, H. Shams1; 1Bioengineering, University of California, Berkeley, CA
Several cell functions including polarization, differentiation, division and migration rely on the ability of
cells to endure mechanical forces generated by the cytoskeleton on the nucleus. The only known
physical linkage between the nucleoskeleton and cytoskeleton is through LINC complexes, which are
formed by the interaction of SUN (Sad‐1 and Unc), and KASH (Klarsicht, ANC‐1, Syne Homology) domain
containing proteins in the perinuclear space. SUN‐KASH complexes are ideally structured
mechanosensors, essential for the transmission of forces between the cytoskeleton and nucleoskeleton
and thereby involved in cellular mechanotrandsuction. In this study, we explore the response of the
human SUN2‐KASH2 complex to mechanical forces, and show that it is remarkably stable under
physiologically relevant tensile forces and large strains. Our studies also suggest that a covalent disulfide
bond between two highly conserved cysteine residues of SUN2 and KASH2 is crucial for the stability of
this interaction and the transmission of forces through the complex.
Microsymposium 15: Membrane Regulation and Signaling
A tale of deception told in 3D: Golgi modifications by an alphavirus to engineer its exit‐pod.
R. Sengupta1, E. Mihelc1, S. Angel1, R.J. Kuhn1, J.K. Lanman1; 1Biological Sciences, Purdue University, West
Lafayette, IN
Enveloped viruses are known to exploit cellular trafficking pathways in their morphogenesis and egress
from the host cell. Many of these viral pathways have either led to the discovery of new cellular
trafficking pathways or have revealed novel and fundamental aspects of previously known pathways.
Alphaviruses are one of the most geographically widespread positive strand RNA viruses which include
significant human and animal pathogens. Early EM studies on alphavirus infected cells revealed the
presence of virus induced cytopathic vesicles (CPVs), namely CPV‐I and CP‐VII that was later found to be
endo‐lysosomal and trans‐Golgi in origin respectively. While CPV‐I has been better characterized and
directly linked with replication of the viral genome, CPV‐II is the predominant vacuolar structure
observed late in infection hinting at its involvement in virus maturation and egress. However the exact
mechanism behind formation of CPV‐II and its trafficking is largely unknown. We reconstructed the
ultrastructure of a large volume (5 x 5 x 2.5 μm) of an alphavirus infected cell in 3D using EM
tomography to understand the biogenesis of CPV‐II from the trans‐Golgi and its role in virus egress. Our
findings reveal that the viral nucleocapsid cores comes in contact with the trans face of the Golgi and
converts it into CPV‐II structures via various intermediate stages of membrane modifications. Gradual
accumulation of viral cores on the Golgi initially results in curling of the cisternal ends of the exposed
face of the trans‐Golgi away from the stack. Further accumulation of cores results in large scale peeling
of the exposed cisterna leading ultimately to cisternal fragmentation and its conversion into vesicular
structures. This early stage CPV‐II that derives its internal and external membranes from the two
cisternal faces of Golgi is decorated with viral nucleocapsid cores on both membranes. These vesicles
then mature gradually to single membrane structures as they move away from the Golgi on its journey
to the plasma membrane, the observed site for alphavirus budding and egress. We reconstructed and
analyzed six independent Golgi stacks with numerous vesicles in different stages of maturation
emanating from them. Additionally we were able to assign all the nucleocapsid cores either to the Golgi
membrane, the CPV‐II or the plasma membrane. Thus by using large volume electron tomography, we
were able to reconstruct piece‐by‐piece a pathway of pathogenic modification of a cell organelle by a
virus for its morphogenesis and egress.
Identification of a novel negative regulator of lipophagy and its role in macrophage foam cell
T.Y. Nazarko1, S.H. Choi2, A. Glieder3, Y.I. Miller2, S. Subramani1; 1Section of Molecular Biology, UC San
Diego, La Jolla, CA, 2Department of Medicine, UC San Diego, La Jolla, CA, 3Institute of Molecular
Biotechnology, Graz University of Technology, Graz, Austria
Heart attack and stroke are the leading causes of death in the United States. Both are the manifestations
of atherosclerosis, a slow, life‐long buildup of lipid (cholesterol and fat) in the plaques of the blood
vessel wall. Eventually, the lipid‐overloaded and inflamed plaques rupture, and clog the blood vessels in
the heart or the brain, causing heart attack or stroke. Macrophages are the major type of lipid‐loaded
cells in the plaques. Proper removal of lipid from macrophages is critical for the prevention and
treatment of atherosclerosis. Most eukaryotic cells accumulate lipid in the organelles called lipid
droplets (LDs). The abundance of LDs is controlled by lipolysis via cytosolic lipases and by lipophagy, the
autophagic trafficking pathway that delivers LDs to the lysosomal lipases. The lack of lipolysis and/or
lipophagy creates an excess of LDs that leads to the lipid accumulation diseases, such as atherosclerosis.
The recent discovery of lipophagy opened a new opportunity for the prevention and treatment of these
diseases, but our knowledge of this autophagic pathway is still very limited. The goal of this study was to
shed light on the regulation of lipophagy in eukaryotic cells in relation to atherosclerosis.
We developed the yeast, Pichia pastoris, as a simple lipophagy model and found that acb1 cells exhibit
de‐repressed lipophagy in the rich growth medium. However, the increased lipophagy in acb1 cells could
not be rescued by the ACB1 gene indicating that a secondary mutation, named nrl1 for the negative
regulator of lipophagy 1, is responsible for the phenotype. Disruption of the ATG5 gene in nrl1 cells
abrogated lipophagy proving that a secondary mutation affected an autophagic process. At the same
time, nrl1 cells had normal non‐selective autophagy pointing to the lipophagy‐specific nature of the
mutated regulator. We sequenced the genome of nrl1 and parental cells and identified the mutated
NRL1 gene. The gene was disrupted by a non‐sense mutation close to the start codon. Reintroduction of
the NRL1 gene in nrl1 cells rescued the phenotype proving that NRL1 is indeed a negative regulator of
lipophagy. Moreover, the overexpression of NRL1 blocked lipophagy in wild‐type cells even under
nitrogen starvation conditions that are known to induce lipophagy. Importantly, our preliminary results
indicate that the knockdown of mouse homologue of NRL1 in macrophages reduces the accumulation of
LDs, which was induced by the pro‐atherogenic minimally oxidized low‐density lipoprotein.
We conclude that NRL1 is a novel negative regulator of lipophagy conserved from yeast to mammalian
cells and that it promotes the accumulation of lipid in macrophages under pro‐atherogenic conditions.
Activated GPCRs exit cilia through ectosome release or BBSome‐mediated retrieval.
A.R. Nager1, F. Ye1, V. Herranz‐Pérez2, J.S. Lee1, J. Manuel Garcia‐Verdugo2, M.V. Nachury1; 1Department
of Molecular and Cellular Physiology, Stanford University School of Medicine, Stanford, CA, 2Laboratorio
de Neurobiología Comparada, Instituto Cavanilles, Universitat de València, Valencia, Spain
Loss of function of the BBSome coat complex results in Bardet‐Biedl Syndrome, a multi‐organ disease
involving dysfunction of the Hedgehog and Planar Cell Polarity pathways, as well as numerous GPCR
signaling pathways. The 8‐protein BBSome coat has been proposed to traffic GPCRs and lipidated cargos
both to and from the cilium. In this study, we document the agonist‐dependent exit of the GPCR
Somatostatin Receptor 3 (SSTR3) from neuronal cilia and, using low‐expression promoters to mimic
neuronal expression levels, we reconstitute agonist‐dependent exit of the GPCRs SSTR3, MCHR1,
GPR161, and NPY2R in immortalized kidney epithelial cells. Surprisingly, we do not observe any overt
GPCR entry defect upon CRISPR‐mediated knockout or siRNA‐mediated depletion of BBSome subunits or
the BBS‐associated GTPase ARL6. Instead, SSTR3, MCHR1, and GPR161 fail to be exported from cilia back
into the cell. In further support of a role for the BBSome in ciliary export, upon GPCR activation the
BBSome and its cargo GPCRs accumulate at the tip where they assemble into cargo‐laden retrograde
BBSome trains. Additionally, we identify β‐Arrestin 2 as a candidate BBS gene as β‐Arrestin 2 is recruited
to cilia upon GPCR activation, interacts with both BBSome and activated GPCRs, and loss of β‐Arrestin 2
results in ciliary export defects. Most unexpectedly, live‐cell fluorescence imaging and electron
microscopy reveal that in the absence of ARL6, BBSome or β‐Arrestin 2 function, activated GPCRs
hyperaccumulate at the ciliary tip, and are frequently released as ectosomes. Based on prior and
present work, we propose the BBSome routes activated GPCRs away from the ectocytosis pathway and
into a retrieval pathway by directly recognizing a motif located in the third intracellular loop of most
ciliary GPCRs, and indirectly reading the state of GPCR activation through β‐Arrestin 2. To test the
biological relevance of this two‐part code, we examined NPY2R, a ciliary GPCR lacking both parts of the
BBSome export code, and found that the default pathway for ciliary exit of NPY2R is through ectocytosis.
Our findings show that activated ciliary GPCRs are redistributed to the tip where they can either be
packaged into ectosomes or loaded onto retrograde BBSome trains for retrieval into the cell. Past work
observing lower abundance of SSTR3, MCHR1, and other ciliary proteins in fixed imaging of BBS‐deficient
cilia is likely due to loss of material following excessive ectosome release. Release of activated GPCRs in
ectosomes may function as an ancient desensitization pathway or to generate bioactive packages for
paracrine signaling.
Intercellular communication pathways are hijacked by bacterial pathogens during cell‐to‐cell
R. Lamason1, M.D. Welch1; 1MCB, UC Berkeley, Berkeley, CA
Eukaryotic cells exchange materials and information via pathways such as trans‐endocytosis or cell‐cell
adhesion. These pathways are also likely hijacked by intracellular bacterial pathogens during the process
of cell‐to‐cell spread, allowing the pathogens to move throughout host tissues while remaining within
host cells. To better understand intercellular communication pathways in infected and uninfected cells,
we investigated the cell‐to‐cell spread process of two bacterial pathogens: Rickettsia parkeri and Listeria
monocytogenes. These bacteria hijack the host actin cytoskeleton to drive actin‐based motility, which
propels them to the plasma membrane, where they form membrane protrusions from the donor cell
that are engulfed into vesicles by a recipient cell. It was previously assumed that the mechanisms of
spread were similar for both pathogens; however, using live cell microscopy, we discovered surprising
differences. In particular, R. parkeri generates shorter protrusions that resolve more quickly into vesicles
when compared with L. monocytogenes, suggesting that each pathogen employs a distinct molecular
mechanism of spread. To uncover these mechanisms, separate screens were done to reveal important
host or bacterial factors. First, we used siRNA screening to test the role of 115 host genes normally
involved in cell‐cell communication (e.g. cell adhesion, membrane remodeling, and endocytosis). This
uncovered distinct sets of host factors, including different cadherin and caveolin isoforms that are
selectively required by each pathogen. Second, bacterial factors that promote spread were identified
using transposon mutagenesis of R. parkeri. One such factor was surface cell antigen 4 (Sca4), which is a
secreted bacterial protein. We found that Sca4 promotes spread by increasing the efficiency of
protrusion engulfment. At the molecular level, Sca4 interacts with the host cell adhesion protein vinculin
and blocks association with its binding partner, α‐catenin. At maturing cell‐cell junctions, α‐catenin
normally recruits vinculin to reinforce reciprocal, actin‐mediated pulling forces and increase membrane
tension. In ongoing experiments, we are testing if Sca4 specifically disrupts the α‐catenin:vinculin
association in the donor cell to relieve opposing forces in the donor cell and shift the force balance
towards the recipient cell, thus improving protrusion engulfment efficiency. Overall, this work has
uncovered significant differences in the features and molecular mechanisms of spread by different
pathogens, and has highlighted that targeting host pathways of cell‐cell communication are key to
promoting bacterial spread. Ultimately, this work will also reveal important regulatory mechanisms of
intercellular communication in uninfected cells.
Endocytic membrane‐associated septins are required for macropinosome maturation and
fusion with lysosomes.
L. Dolat1, E.T. Spiliotis1; 1Biology, Drexel University, Philadelphia, PA
Macropinocytosis is a form of clathrin‐independent endocytosis (CIE). At the plasma membrane, actin
polymerization triggers membrane ruffles that fold back and form large fluid‐filled vesicles termed
macropinosomes. Following ruffle closure, macropinosomes undergo a maturation process that leads to
fusion with lysosomes. Macropinocytosis is critical for signaling, immune cell surveillance and pathogen
invasion, but little is known about the mechanisms that underlie the progression of macropinosomes
through the endocytic pathway. Septins are a family of oligomeric GTP‐binding proteins that assemble
into higher‐order structures and associate directly with actin, microtubules and cell membranes. Here,
we have identified a novel population of membrane‐associated septins that localize to macropinosomes
and tubulovesicular membranes in the peripheral lamellae of Madin‐Darby canine kidney (MDCK)
epithelia. Two‐color, 3D live‐cell imaging of GFP‐tagged septins and membranes labeled with Lyn‐
mCherry revealed that septins are recruited to nascent macropinosomes as they tubulate and begin to
shrink and disappear. Moreover, septins localize to a dynamic network of endotubular membranes and
are particularly enriched at their contact sites with macropinosomes. Using dextran uptake assays, we
show that septin depletion does not impede macropinosome formation, but increases significantly the
number and size of dextran‐positive macropinosomes. Live‐cell imaging of Lyn‐mCherry showed that
septin depletion results in macropinocytic pileups that consist of large clusters of long‐lived
macropinosomes, some of which form elongated undynamic tubules. Because this phenotype was
indicative of a defect in macropinosome maturation and fusion with late endosomes or lysosomes we
tested whether septins affect the recruitment of the Rab5 and Rab7 GTPases, which regulate membrane
traffic and fusion. Interestingly, septin depletion did not impede the recruitment of Rab5 or Rab7 to
macropinosomes. However, septin knock‐down impeded the delivery of the fluid phase marker dextran
to lysosomes; dextran was largely accumulated in macropinocytic membranes devoid of LAMP1. To test
whether septins are required for macropinosome membrane fusion, we performed in vivo and in vitro
two‐color dextran fusion assays, which showed a reduction in the fusion of green‐ and red‐labeled
macropinosomes upon SEPT2 depletion or incubation with a function‐blocking antibody against SEPT2.
Taken together, these data indicate that septins are a novel component of the molecular machinery that
drives the maturation of macropinosomes in the endocytic pathway, and suggest that septins play a
critical role in the homotypic and heterotypic fusion of macropinosomes with lysosomes.
β‐arrestin drives MAP kinase signaling after dissociating from its activating GPCR through
kinetic arrest of clathrin‐mediated endocytosis.
K. Eichel1, D. Jullié1, M.E. Von Zastrow1,2; 1Psychiatry , University of California, San Francisco, San
Francisco, CA, 2Cellular Molecular Pharmacology, University of California, San Francisco, San Francisco,
A fundamental question in cell biology is how endocytosis and cellular signaling are interconnected.
While there are many examples of endocytosis regulating signaling, there are relatively few examples of
signaling regulating endocytosis. Here, we show here that (1) signaling receptors control their clathrin‐
coated pit (CCP) dynamics and (2) CCP dynamics regulate MAP kinase signaling. β‐arrestins are key
regulators of G protein‐coupled receptors (GPCRs), which not only 'arrest' G protein signaling but also
modulate endocytosis and initiate a wave of G protein‐independent signaling via MAP kinase. Our
understanding of how these events occur in intact cells remains rudimentary and based largely on
inference from structural and biochemical study of isolated proteins. The present dogma is based on
two key assumptions: (1) β‐arrestin traffics to clathrin‐coated pits (CCPs) in obligate physical complex
with its activating GPCR, and (2) the MAP kinase signal emanates from the GPCR / β‐arrestin complex
after endocytosis of the formed complex. Here we explicitly test both assumptions using real‐time
imaging combined with direct manipulation of surface protein dynamics in living mammalian cells.
Refuting the first assumption, we show that ligand‐activated GPCRs can drive robust trafficking of β‐
arrestin to CCPs without themselves moving there. Refuting the second assumption, we show that β‐
arrestin promotes downstream MAP kinase activation not through endocytosis, but by capturing CCPs to
prolong their surface lifetime before endocytosis. This endocytic delay, and not endocytosis itself,
determines the magnitude of downstream MAP kinase signaling. These results redefine the cellular basis
of β‐arrestin function and demonstrate a discrete, non‐endocytic role of CCPs as dynamic signaling
stations on the plasma membrane.
Autophagy preserves the functional capacity of hematopoietic stem cells during aging.
T.T. Ho1, M.R. Warr1, J. Debnath2, E. Passegue1; 1Department of Medicine, University of California, San
Francisco, San Francisco, CA, 2Department of Pathology, University of California, San Francisco, San
Francisco, CA
Aging is the greatest risk factor for a wide range of diseases, and one emerging hallmark of aging is that
reduction in tissue repair usually correlates with a reduction in stem cell activity. With age,
hematopoietic stem cells (HSCs) lose their regenerative capacity and ability to produce all blood cells,
resulting in a decline in immune responses and increased rate of blood diseases in the elderly.
Autophagy is a fundamental proteostasis mechanism important for cellular health and healthspan and
lifespan in general. We previously showed that autophagy, and its longevity‐associated transcription
factor FoxO3a regulator, plays a critical role in protecting young HSCs from metabolic stress. We also
found that a subset of old HSCs isolated from aged mice have constitutive activation of autophagy.
However, it remains largely unknown how autophagy controls HSC function and blood production in
both young and old animals. We generated mice with conditional deletion of the essential autophagy
gene Atg12 in HSCs (Atg12cKO). We found that autophagy inactivation in young HSCs resulted in pre‐
mature aging phenotypes resembling those of physiologically old HSCs, including enhanced
differentiation towards the myeloid lineage, rapid HSC exhaustion with loss of self‐renewal and
regenerative potential, and ultimately HSC depletion under regenerative challenge such as serial
transplantation. Analyses of Atg12cKO HSCs by electron microscopy revealed expanded endoplasmic
reticulum and Golgi compartments, and an excess of small vesicles and mitochondria, similar to what we
observed in the subset of physiologically old HSCs that do not show constitutive autophagy induction.
Microarray gene expression analyses confirmed significant changes in mitochondria and metabolic
pathways in both young Atg12cKO and physiologically old HSCs. Inhibition of autophagy in
physiologically old HSCs dramatically reduced function and multipotentiality, while having little effect on
young wild type HSCs. We also aged GFP‐LC3 autophagy‐reporter mice and found that old HSCs that are
unable to activate autophagy lose their regenerative capacity and rapidly exhaust after transplantation,
hence resembling young autophagy‐deficient Atg12cKO HSCs. Our results demonstrate that autophagy
is essential for the regenerative capability of young HSCs. They also show that with age, HSCs become
increasingly dependent on their ability to mount a proper autophagy response for their survival and
functional competency, and that old HSCs unable to undergo autophagy are severely impaired, likely
causing the age‐associated declines in regenerative potential and blood production. They suggest that
increasing autophagy in old HSCs could help promote HSC fitness and rejuvenate the aging blood
Microsymposium 16: Cell Biology of Genetic Information
Cell‐to‐cell transfer of mRNA via membrane nanotubes.
G. Haimovich1, J.E. Gerst1, R.H. Singer2,3; 1Molecular Genetics, Weizmann Institute of Science, Rehovot,
Israel, 2Anatomy Structural Biology, Albert Einstein College of Medicine , Bronx, New‐York, NY,
Dominick P. Purpura Department of Neuroscience, Albert Einstein College of Medicine , Bronx, New‐
York, NY
Communication between cells is an essential function in multicellular organisms. The possible transfer of
RNA molecules (particularly miRNA and mRNA) between cells was recently suggested as a form of
communication that could have a regulatory role in the acceptor cells. The current model suggests that
such transfer occurs by secreting the RNA to extracellular fluids, either as a free RNP particle or packed
in nanovesicles (e.g. exosomes). However, most work in the field has been done on whole cell
populations, using mostly biochemical methods. Many questions still remain. In particular, can we show
a quantitative analysis of RNA transfer and what are the exact mechanism and kinetics of this process.
Here, we used single molecule fluorescent in situ hybridization (smFISH) and live imaging of single mRNA
molecules to show that mRNAs are transferred in cultures of mouse embryonic fibroblasts (MEFs),
human cell lines and mouse/human co‐cultures. We found that the transfer of either endogenous or
tagged mRNAs occurs in both immortalized and primary cells. Among the mRNAs we followed are those
encoding mouse β‐actin, human Cyclin D1, human BRCA1, human GAPDH, SV40 Large T antigen, and
others. We were unable to detect transfer of human HER2 mRNA, suggesting this process has specificity.
co‐FISH experiments revealed that the transferred β‐actin mRNA makes up ~2‐5% of the total β‐actin
mRNA in the acceptor cell. We further show that this process can be modulated by various treatments,
such as heat‐shock or oxidative stress. mRNA transfer is quick (detected within 30 minutes after co‐
culture) and is independent of de‐novo protein synthesis.
Contrary to the current model, we found that mRNA transfer requires close proximity between cells and
is not mediated by diffusion. Rather, we suggest that mRNA is transferred through membrane
nanotubes. These are very long (>100µm) yet thin (50‐300nm) cytoplasmic projections that were
recently shown to be involved in direct, contact‐dependent, intercellular communication. We present
both biological and imaging data to support this hypothesis.
The biological significance of intercellular transfer hinges on whether these mRNAs are translated and
effect cell physiology. We currently are developing methods to assess this.
Association of heterochromatin with the nuclear envelope drives nuclear stiffness.
S.M. Schreiner1, Y. Zhao2, P. Koo2, S. Mochrie2,3, M.C. King1; 1Cell Biology, Yale School of Medicine, New
Haven, CT, 2Physics, Yale University, New Haven, CT, 3Applied Physics, Yale University, New Haven, CT
Tissues (and the cells that comprise them) possess variable degrees of stiffness according to their
specific function(s). There is an increasing appreciation that the nucleus is integrated into this
mechanical environment, and its mechanical properties must be coordinately modulated to support
cellular (and tissue level) function. In stiff tissues, high degrees of nuclear stiffness correlate with the
level of lamin A expression. However, it is not yet clear if the lamin A polymer itself imparts nuclear
stiffness, or if the ability of lamin A to tether DNA, particularly heterochromatin, to the nuclear envelope
plays a role. Further, many cell types express little (or no) A‐type lamins; what factors define nuclear
stiffness in these contexts remains poorly defined. Yeast, which lack a nuclear lamina, provide a model
system in which to study how tethering of chromatin to the nuclear envelope influences nuclear
stiffness; in this model integral inner nuclear membrane proteins act as the sole nuclear envelope
tethers. Using a quantitative imaging platform capable of measuring 3D nuclear contours in live cells and
an in vitro optical tweezers assay to probe the mechanical properties of S. pombe nuclei, we found that
integral inner nuclear membrane protein tethers promote nuclear stiffness but also restrict chromatin
flow to prevent long‐term changes in nuclear shape in response to cytoskeletal forces. This suggests that
modulating the extent of chromatin tethering to the inner nuclear membrane can modulate both
nuclear stiffness and nuclear viscosity; both parameters may impact nuclear deformation during cell
migration in mammalian cells. Here, we extend this work to investigate how the epigenetic state of
chromatin influences nuclear mechanics. Interestingly, we find that cells lacking the HP1 homologue
have compromised nuclear stiffness in vivo and in vitro, while having little change in nuclear viscosity.
Similar results were observed in cells lacking the histone H3 K9 methylase. These results suggest that the
ability of chromatin tethering to restrain chromatin flow is not influenced by the chromatin state.
However, tethering of euchromatin does not provide the same degree of nuclear stiffness as
heterochromatin. This work provides a new framework for thinking about the age‐old observation that
heterochromatin is constitutively associated with the nuclear envelope in most eukaryotes. Further, it
suggests that physical softness of nuclei and euchromatic state may be tied during development;
indeed, both these attributes apply to stem cell nuclei. We are currently testing this model by applying
these same techniques to other eukaryotic models.
Constricted cell migration drives lamin segregation, repair factor demixing, and DNA damage.
J. Irianto1, C.R. Pfeifer1, A. Athirasala1, I.L. Ivanovska1, R.A. Greenberg2, D.E. Discher1; 1Biophysical
Engineering Labs, University of Pennsylvania, Philadelphia, PA, 2Department of Cancer Biology,
Abramson Family Cancer Research Institute, University of Pennsylvania School of Medicine, Philadelphia,
Normal and diseased cells in vivo sometimes squeeze their large nucleus through a stiff, fibrotic tissue or
a basement membrane barrier, but any effect on genome integrity is unknown. Here, migration of
human lung and bone cancer derived cells through rigid micro‐pores causes nuclear blebs with
segregated lamins and DNA tethers as well as an increase in DNA double‐strand breaks (DSBs).
Mechanical enhancement of DSBs by 50% due to migration through small pores (but not large pores) is
further enhanced by lamin‐A suppression. Nuclear blebs seen in the majority of cells after migration are
enriched in lamin‐A but deficient in DNA and lamin‐B as well as a progeria mutant of lamin‐A. Extension
of a specific DSB site on chromosome‐1 is directly visualized within nuclei pulled into micropipettes,
showing that extension modulates the density of DSB repair factors. Numerous DSB repair proteins –
when mobile rather than bound to chromatin – demix from the chromatin and can even be ruptured
from the nucleus as it squeezes into an aspirating micropipette and also through a small micro‐pore
during migration. DSB repair tends to be impeded, favoring accumulation of DNA damage and genome
instability as assessed by copy number variation (CNV) and expression profiling. Migration‐induced
genome instability, with subsequent mutation or cell death, might thus couple to microenvironment
rigidification as occurs in processes ranging from cancer to aging.
Structural Insights into the Recognition of the Histone Tails by Karyopherins.
M. Soniat1,2, Y. Chook1,2; 1Biophysics, University of Texas Southwestern, Dallas, TX, 2Pharmacology,
University of Texas Southwestern, Dallas, TX
Import‐Karyopherins or Importins bind nuclear localization signals (NLSs) to transport proteins into the
nucleus. Nuclear import of the core histones is one of the major nuclear import activities during S‐phase
of the cell cycle. Multiple Importins were previously reported to bind histone tails and import histones
into the nucleus. However, nothing is known about how these different Importins recognize histone
tails. It is unclear if there are common nuclear localization signals (NLSs) in the tails that are recognized
by all Importins or if each Importin recognizes distinct NLSs within the tails. Furthermore prior to nuclear
import, acetylation of core histones by acetyltransferases (HATs) at lysine residues of their N‐terminal
tails have been shown to occur in the cytoplasm prior to their import into the nucleus, but it is not
understood how acetylation controls nuclear import of histones. Structural and biochemical analysis
show that the histone H3 and H4 tails are recognized by six Importins (Impβ, Kapβ2, Imp4, Imp5, Imp7,
and Imp9) through a ‘universal’ basic epitope in the histone H3 Tail that binds to all six Importins. Within
the ‘universal’ basic epitope, we show that Lys14, which is known to be acetylated prior to nuclear
import, is important for Importin‐histone interaction. Using acetylation mimic mutation histone H3 Tail
K14Q, we show that acetylation of Lys14 and Lys18 likely disrupts binding to Importins. In contrast,
histone H4 Tail acetylation mutation mimics at Lys5/Lys12 did not affect binding. In summary, our
results show that the Importins Impβ, Kapβ2, Imp4, Imp5, Imp7, and Imp9 recognize a ‘universal NLS’ in
the N‐terminal tail of histones H3 and H4, and that acetylation of the H3 histone tail prior to nuclear
import affects binding to the Importins.
Protein turnover and dynamics spanning seconds to months in non‐dividing cells.
B.H. Toyama1, M.W. Hetzer1; 1Molecular and Cell Biology Laboratory, The Salk Institute for Biological
Studies, La Jolla, CA
Proteins in the cell are in constant flux and subject to association, assembly, disassembly, synthesis, and
degradation. The timescale on which these processes occur can span decades, however, most proteins
turnover in hours to a few days. These relatively short half‐lives result in a “young” and functional
proteome, turning over proteins before they accumulate damage. Rare long‐lived proteins which persist
in the cell far longer than the typical protein, are thus prone to damage accumulation and loss of
function, as is seen with long‐lived eye‐lens crystallin and cataract formation. In previous studies we
made the surprising discovery that a number of proteins have lifespans in excess of 6 months, including
proteins of the nuclear pore complex (Nups), histone octamer, and myelin sheath. As these proteins play
essential roles in a myriad of cellular processes, loss of their function can have a devastating impact on
the cell. How the cell maintains these long‐lived proteins (LLPs), where they reside, and the impact of
their longevity, are still not understood. In an effort to study the mechanisms underlying the stability
and maintenance of LLPs in the cell, we adapted a tag‐switching technique called RITE (recombination
induced tag exchange) to study protein dynamics in mammalian cells. In brief, cre recombinase is used
to switch expression of a c‐terminal tag on a protein of interest from one tag (“old”) to a different tag
(“new”), allowing the distinction between old and new protein at a much later date. We tag‐switched
the scaffold Nups Nup93, Nup96, Nup133, which have long lifespans, and Pom121, which is rapidly
turned over, and found in dividing C2C12 myoblasts, all tested Nups had lifespans on the order of 1 day.
However, in differentiated non‐dividing myotubes, Pom121 had a lifespan of 2‐3 days, Nup133 of 7 days,
and Nup93 and Nup96 of over 14 days. The difference in lifespan between Nup133 and Nup96 is
surprising as both proteins belong to the same subcomplex, suggesting multiple dynamic states even
within this subcomplex. Similarly, histone H2B, H3.1, H3.3, and H4 all had lifespans on the order of 1 day
in dividing myoblasts, but all persisted much longer in myotubes with H3.1 at less than 15% turned over
after 7 days and H2B turning over 50% after 2 days. Interestingly, histones associated with chromatin
regions marked by H3K9me3 foci were more stable, while active regions exhibited more rapid turnover.
Altogether, our studies have surprisingly revealed that the same protein can have both dynamic and
highly stable states depending on its cellular context and localization. Understanding how stability is
established and maintained will be key as these long‐lived populations of proteins will be most prone to
the deleterious affects of aging.
Slicing Activity of the Argonaute CSR‐1 Tunes Expression of Germline Genes to Control
Embryonic Cell Division and Germline Development.
A. Gerson‐Gurwitz1, S. Wang1, R.A. Green1, K. Oegema1, A.B. Desai1; 1Dep. of Cellular Molecular
Medicine, Ludwig Cancer Research, San Diego, CA
The Argonaute protein family plays a central role in diverse processes involving small RNAs. In C.
elegans, which has 27 Argonautes, inhibition of only the Argonaute CSR‐1 results in severe defects in
embryonic chromosome segregation and penetrant embryonic lethality. CSR‐1 interacts with 22G small
RNAs derived from germline‐expressed transcripts. Many functions have been proposed for CSR‐1,
including control of chromosome architecture, protection of germline gene expression, and histone
mRNA maturation. However, despite robust mRNA slicing activity in vitro, a role for CSR‐1 in controlling
target expression has been disfavored. Here, we show that selective mutation of CSR‐1 slicing activity
phenocopies CSR‐1 removal. Quantitative immunoblotting using 44 validated antibodies against
germline‐expressed proteins revealed that loss of CSR‐1 slicing activity led to 1.5‐to‐3‐fold elevation of
the levels of 19 of the tested proteins, with the magnitude of the elevation correlating with the
abundance of target‐specific CSR‐1‐22G RNA complexes; the mRNA levels of a subset of the 44 tested
targets were also significantly elevated. A phenotypic hallmark of CSR‐1 inhibition is greatly reduced
microtubule assembly in embryos, which we associated with ~3‐fold elevated expression of the
microtubule depolymerase MCAK/KLP‐7. Reduction of the elevation in MCAK/KLP‐7 levels suppressed
the prominent microtubule assembly defect observed in the absence of CSR‐1. Thus, fine control of
target expression by the CSR‐1‐22G RNA system is essential for tuning the activity of proteins such as
MCAK/KLP‐7, which is a potent enzyme. We propose that the CSR‐1‐22G RNA system functions in a
catalytic activity‐dependent manner to fine‐tune the expression of a large number of germline‐
expressed genes, and that this control is essential for embryonic divisions and germline development.
NuMA regulates the mobility of 53BP1 in the cell nucleus and its accumulation at DNA damage
P. Vidi1, J. Liu2, M. Gray3, L. Parker4, J. Irudayaraj2, S.A. Lelievre3; 1Cancer Biology, Wake Forest School of
Medicine, Winston‐Salem, NC, 2Agricultural and Biological Engineering, Purdue University, West
Lafayette, IN, 3Basic Medical Sciences, Purdue University, West Lafayette, IN, 4Biochemistry, Molecular
Biology and Biophysics, University of Minnesota, Saint Paul, MN
Following genomic insults, mammalian cells trigger a rapid and comprehensive DNA damage response
(DDR), which enables the recruitment of factors for DNA repair, as well as checkpoint signaling for cell
cycle arrest. The recruitment of 53BP1 at DNA double‐strand breaks mediates DNA repair pathway
choice and checkpoint activation. Whereas 53BP1 repair foci formation at DNA damage sites is well
characterized, if and how 53BP1 diffusion is regulated outside repair foci is not known. We developed an
in silico model that predicts protein diffusion in the viscoelastic environment of the cell nucleus.
Comparing values generated by this model with the actual diffusion of 53BP1 measured by fluorescence
correlation spectroscopy indicates constrained 53BP1 movements. Importantly, induction of DNA
damage lead to increased 53BP1 mobility. Using proteomics, we found that 53BP1 interacts with the
nucleoskeletal protein NuMA at steady state. This interaction weakens in response to DNA damage.
Accordingly, NuMA‐depleted cells display increased 53BP1 diffusion and increased 53BP1 accumulation
at DNA damage sites, whereas NuMA overexpression has the opposite effect. Colocalization analysis in
cultured mammary epithelial cells as well as in human mammary tumors indicates decreased overlap
between the two proteins upon DNA damage. We propose a new mechanism that involves the
sequestration of 53BP1 by NuMA in the absence of DNA damage and 53BP1 release triggered within the
DDR. Such mechanism regulating the access of 53BP1 to the chromatin may have evolved in higher
eukaryotes to disable repair functions in the absence of DNA damage.
Microsymposium 17: Spindle Assembly and Chromosome Dynamics
DNA Replication‐timing Maps Reveal Dynamic Chromatin Domains in the Developing Zebrafish
J.C. Siefert1,2, C.L. Sansam1,2, A. Koren3; 1Cell Cycle Cancer Biology Research Program, Oklahoma Medical
Research Foundation, Oklahoma City, OK, 2Department of Cell Biology, University of Oklahoma Health
Science Center, Oklahoma City, OK, 3Department of Molecular Biology and Genetics, Cornell University,
Ithaca, NY
Every dividing cell type in metazoan organisms replicates its genome in a defined temporal order.
Transcriptionally active, open euchromatin domains generally replicate during early S‐phase, while
transcriptionally inactive, closed heterochromatin domains generally replicate late in S‐phase. Little is
known about how the replication timing program is established or how and why it changes with cellular
differentiation. To better understand how DNA replication is developmentally regulated, we have
generated genome‐wide replication timing maps for a zebrafish fibroblast cell line and zebrafish
embryos at five key stages of early development corresponding to different transcriptional and
differentiation statuses. While the overall replication timing program changes gradually during early
development, specific genomic regions display strong and abrupt changes. One particularly striking
change occurs between shield and bud stages, at the same time that germ layer fates are determined.
Similar replication timing changes have been reported for mammalian embryonic stem cells as they
undergo differentiation and widespread chromatin restructuring. These genome‐wide replication timing
profiles provide an unprecedented view of changing DNA replication patterns during development, and
they also serve as spatiotemporal maps of heterochromatin formation. Thus, our data provides a
foundation to use the zebrafish to study the mechanisms and functions of developmental changes in
DNA replication and heterochromatin. Furthermore, ongoing analysis comparing zebrafish replication
timing with different genetic and epigenetic features has revealed the extent to which correlations with
these features are conserved among vertebrates and reflects their contribution to replication timing
regulation and maintenance.
Altered epidermal development and growth control, without perturbed homeostasis or
tumorigenesis, in the presence of centrosome amplification.
A. Kulukian1, A. Holland2, B.D. Vitre2, D.W. Cleveland2, E. Fuchs1; 1Mammalian Cell Biology and
Development, Rockefeller University, New York, NY, 2Ludwig Institute for Cancer Research; Dept of
Cellular and Molecular Medicine, University of California San Diego, La Jolla, CA
As nucleators of the mitotic spindle and of the primary cilium, centrosomes play a crucial role in equal
segregation of DNA content to daughter cells, in the coordination of growth and differentiation, and in
the transduction of homeostatic cues. Centrosome amplification, or the condition of having more than
two centrosomes per cell, has been suggested to contribute to chromosomal instability, an imbalance in
asymmetric divisions, and reorganization of tissue architecture. However, the degree to which these
conditions are a direct cause or simply consequence to human disease, including cancer, is poorly
understood. We address this issue by generating a mouse model inducing centrosome amplification in a
naturally proliferative epithelial tissue by elevating polo‐like kinase 4 (Plk4) expression in the skin
epidermis. By altering centrosome numbers, we observed multi‐ciliated cells, spindle orientation errors,
and chromosome segregation defects within the developing epidermis. However, none of these defects
were sufficient to impart a proliferative advantage within the tissue. Rather, impaired mitoses led to
p53‐mediated cell death and contribute to defective growth and stratification. Despite these
abnormalities, mice remained viable and healthy. Moreover, centrosome amplification was insufficient
to disrupt homeostasis, initiate spontaneous tumorigenesis, or enhance chemical carcinogenesis.
Operational characteristics of the kinetochore‐based signaling reactions of the Spindle
Assembly Checkpoint.
P. Aravamudhan1, R. Chen2, J.H. Sim2, A.P. Joglekar2; 1Biophysics, University of Michigan, Ann Arbor, MI,
Cell and Developmental Biology, University of Michigan, Ann Arbor, MI
The Spindle Assembly Checkpoint (SAC) is a fundamental cell cycle control that arrests the progress of
cell division if the dividing cell contains one or more unattached kinetochores. Cell biologically optimal
operation of the SAC requires that a single unattached kinetochore is capable of arresting cell division
robustly, and at the same time, the cell can rapidly overcome the arrest and divide, after all
kinetochores are attached to the spindle apparatus. Whether or not the SAC meets these requirements
is dictated by a large set of parameters that govern the biochemical signaling cascade of the SAC. The
ability of SAC to meet these requirements has significant implications on genome stability during
development and aging, and on the ability of cancer cells to proliferate.
Although much is known about the cytosolic reactions of the SAC, the nature of its kinetochore‐based
reactions and their biochemical regulation is poorly understood. These operational characteristics
ultimately determine whether or not the SAC is cell biologically effective. To define these
characteristics, we systematically analyzed the steady‐state operation of the kinetochore‐based SAC
reactions under a wide range of parameters in the budding yeast Saccharomyces cerevisiae. We find
that the Mps1 kinase maximally phosphorylates the kinetochore protein Spc105 (KNL‐1 orthologue), and
thus primes it for binding the maximal number of the Bub3‐Bub1 complex. However, two mechanisms
strongly dampen Bub3‐Bub1 recruitment and hence the signaling strength of the kinetochore: negative
cooperativity in binding more than one Bub3‐Bub1 molecule to the same Spc105, and low Bub1
expression, which limits the concentration of Bub3‐Bub1 available for signaling. Perhaps to counter this
modulation, the yeast kinetochore is capable of recruiting two Mad1‐Mad2 dimers for every Bub3‐Bub1
molecule. We propose that these characteristics of the kinetochore‐based signaling reactions balance
the conflicting requirements of ensuring robust signal generation from each unattached kinetochore,
and limiting the cumulative signal generated by many kinetochores. Our data establish a quantitative
framework for predicting the impact of adaptive or aberrant changes in cellular parameters on the
ability of the unattached kinetochore to activate the SAC, and also provides a road‐map for similar
studies in other organisms.
Probing the Physical Inputs Controlling the Spindle Assembly Checkpoint.
J.A. Kuhn1,2, S. Dumont1,3; 1Cell and Tissue Biology, University of California, San Francisco, San Francisco,
CA, 2Tetrad Graduate Program, University of California, San Francisco, San Francisco, CA, 3Cellular and
Molecular Pharmacology, University of California, San Francisco, San Francisco, CA
To ensure accurate chromosome segregation, the cell employs a signaling system known as the spindle
assembly checkpoint (SAC). The SAC prevents the final separation of sister chromatids until all
chromosomes are correctly attached to opposite sides of the spindle. The kinetochore, which attaches
chromosomes to the spindle, has been proposed to sense both biochemical and physical cues to control
the SAC, but the contributions of these inputs are not understood.
To probe how the mammalian kinetochore senses correct chromosome attachments, we developed a
simple fluorescent sensor and imaging platform to capture SAC activation levels at individual
kinetochores in real time. We quantify temporal SAC signal dynamics as individual kinetochores attach
to an assembling spindle, and quantify concurrent spindle structural changes occurring in different
scenarios. We find that Mad1 loss from kinetochores – which is necessary and sufficient for SAC
satisfaction – occurs with single exponential kinetics whose rate is independent of different spindle
attachment scenarios. Mad1 loss only begins after both kinetochores are attached, and after tension
(centromere stretch) is established. Yet, Mad1 does not leave laterally attached kinetochores under
tension, and both sister kinetochores can thus satisfy the SAC minutes apart. Mad1 begins to leave
kinetochores immediately as an end‐on attachment forms, and before a full complement of
kinetochore‐microtubules is bound. Together, this work quantitatively maps the dynamic relationship
between kinetochore‐to‐spindle attachments and SAC signaling that ultimately regulates mitotic
Differential chromatin states can regulate chromosome length scaling.
A. Ladouceur1, L. Smith1, J.G. Lawrimore1, K.S. Bloom1, P.S. Maddox1; 1Biology, University of North
Carolina, Chapel Hill, NC
During mitosis, the longest chromosome arm must be shorter than half of the mitotic spindle for proper
chromosome segregation. Multicellular development requires that cells reduce in size due to
consecutive cell divisions without an increase in embryo volume. To maintain cellular integrity,
organelles including nuclei, mitotic spindles, centrosomes, and chromosomes adapt to cell size
throughout development. Using high‐resolution time‐lapse microscopy of living C. elegans embryos, we
found that chromosome length predictably scaled through cell size and nuclear size; larger cells have
longer chromosomes. Cell size based control is saturated in very large cells present during the first
divisions as chromosome length reaches a maximum and does not scale, a phenomenon we term
“plateau”. We suggest a limiting component model where a theoretical inhibitor of chromosome
compaction is in excess in the early embryo, restricting chromosome compaction early during
development. In order to identify regulators of chromosome scaling, we designed a large‐scale RNAi
screen in C. elegans based on differential embryonic lethality of a sensitized worm strain compared to
wild‐type. The sensitized animals have a telomeric fusion between the 2 longest chromosomes resulting
in an abnormally long chromosome without changing total amount of DNA. We hypothesized those
worms would be more susceptible to any defect in chromosome scaling compared to control. We used a
sub‐library of the whole genome RNAi library that includes 438 genes known or predicted to be
chromatin binding and/or modifying enzymes. The screen resulted in a total of 15 hits, which, half are
known to be required for assembling and/or regulating mitotic chromosome assembly. RNAi depletion
of individual hits followed by high‐resolution imaging revealed that partial depletion of the histone H3
variant CENP‐A reduces chromosome length as early as the one‐cell stage. We propose CENP‐A
chromatin acts as a compaction inhibitor on holocentric chromosomes as predicted by our model of
chromosome size regulation. CENP‐A chromatin could be more (or less) flexible, thus resulting in altered
chromosome structure when the absolute amount of CENP‐A is reduced. In support of this we find that
full depletion of CENP‐A disrupts chromosome formation. We are currently mechanistically testing the
hypothesis that CENP‐A nucleosomes modulate the biophysical properties of the chromatin compared
to H3 nucleosomes. In summary, we are using large‐scale RNAi depletion and high resolution imaging to
determine the mechanisms of mitotic chromosome size regulation. Our research exploits a normal
developmental process to understand the cell biological problem of chromosome condensation.
Cooperation between kinesin motors promotes spindle symmetry and chromosome
organization in oocytes.
S.J. Radford1, A.M. Go1,2, K.S. McKim1,2; 1Waksman Institute, Rutgers University, Piscataway, NJ,
Department of Genetics, Rutgers University, Piscataway, NJ
During cell division, chromosomes interact with a bipolar array of microtubules that constitute the
spindle to direct their segregation. In most animals, including Drosophila and humans, oocytes face
several challenges to ensure the accuracy of chromosome segregation. First, the meiotic spindle is
assembled in the absence of the microtubule‐organizing centers called centrosomes. Second, the
volume of the oocyte is large relative to the size of the chromosomes and spindle. Finally, the oocyte
divisions are asymmetric, resulting in a large egg and smaller polar bodies. Without the organization
provided by centrosomes, we have hypothesized that acentrosomal meiotic spindle organization relies
on the bundling of microtubules by kinesin motor proteins. It has been known for many years that two
microtubule‐bundling kinesins – the plus‐end directed kinesin‐6 Subito and the minus‐end directed
kinesin‐14 NCD – are required during Drosophila oocyte spindle assembly. How these microtubule‐
bundling activities cooperate, and whether additional microtubule‐bundling kinesins function in oocytes
is not known. The most prominent microtubule‐bundling kinesin in centrosomal cells is the plus‐end
directed kinesin‐5 KLP61F. We found that loss of KLP61F frequently leads to a dispersal of chromosomes
throughout the entire volume of the oocyte. This phenotype is suppressed by loss of NCD, but enhanced
by loss of Subito. We also found that loss of KLP61F leads to a distinct asymmetry of the acentrosomal
spindle with one half spindle markedly weaker than the other. This correlates with an asymmetric
distribution of centromeres such that fewer centromeres are oriented towards the weaker spindle pole.
This phenotype is suppressed by loss of either NCD or ASP, a microcephaly protein known to function in
asymmetric cell division. We are currently investigating the hypothesis that KLP61F and ASP cooperate
to direct the asymmetric oocyte cell division. Together, these results suggest a model in which
chromosomes are maintained at the center of a symmetric bipolar spindle through a balance of kinesin‐
based forces. The plus‐end‐directed kinesins KLP61F and Subito are partially redundant and antagonistic
to the minus‐end‐directed kinesin NCD, resulting in proper spindle and chromosome organization in
acentrosomal oocytes.
Timing and pattern of release of chromosome cohesion differs between species with
monocentric and holocentric chromosomes.
K.D. Felt1, S. Thibault1, A.M. Martens1, L.V. Paliulis1; 1Biology Department, Bucknell University,
Lewisburg, PA
While anaphase chromosome separation appears rapid and at times synchronous along the length of
the chromosome, persistent connections through late anaphase of mitotic and meiotic cell divisions
have been observed in many species. In this study, we aimed to determine whether patterns of release
of sister chromatid cohesion were similar in organisms with very different modes of chromosome
attachment to the spindle. We compared this release of chromosome cohesion in two organisms with
monocentric chromosomes (the grasshopper Melanoplus sanguinipes and the cricket Acheta
domesticus) with release of chromosome cohesion in an organism that has holocentric chromosomes in
mitosis but functionally monocentric chromosomes in meiosis (the milkweed bug Oncopeltus fasciatus).
After observing anaphase chromosome movements in these species, we determined that complete
release of chromosome cohesion in anaphase I of meiosis and in mitotic anaphase takes an average of 7
minutes in cells with monocentric chromosomes. In Oncopeltus fasciatus, complete release of cohesion
also takes an average of 7 minutes in meiosis I, where chromosomes appear monocentric, but takes an
average of only 30 seconds in mitosis, during which chromosomes are holocentric. Using
micromanipulation, we pushed one chromosome and observed the effect on its partner in anaphase.
We found that persistent but invisible connections existed between partner chromosomes through late
anaphase in cells with monocentric chromosomes, but that these connections could not be observed
during anaphase in cells with holocentric chromosomes. This suggests that chromatin linkages are likely
released more gradually along monocentric chromosomes, where only a singular centromere is present,
and more simultaneously along holocentric chromosomes, where there are kinetochore proteins along
the entire length of the chromosomes.
Microsymposium 18: Cell Mechanics and Adhesion
AMPK: A Novel Link Between E‐cadherin Force Transmission and Cell Metabolism.
J.L. Bays1, C. Heidema1, W. Hacker1, K.A. DeMali1; 1Department of Biochemistry, University of Iowa, Iowa
City, IA
Cells experience forces throughout their lifetime. These forces are sensed by cell surface adhesion
receptors and trigger robust actin cytoskeletal rearrangements and growth of the associated adhesion
complex to counter the applied force. This process is known as cell stiffening or reinforcement. The actin
re‐arrangements necessary for stiffening are energetically costly suggesting that the mechanisms
coupling force transduction and energy production might exist. Here, we show that application of force
on E‐cadherin increases glucose uptake, elevates cellular ATP levels, and activates the master glucose
regulator, AMP‐dependent kinase (AMPK). We further present evidence that inhibition of AMPK
diminishes glucose uptake and impairs the cytoskeletal rearrangements necessary for cell stiffening. We
investigated how AMPK activation is linked to these cytoskeletal changes and observed that AMPK
forms a complex with Abl tyrosine kinase thereby activating Abl kinase to phosphorylate vinculin Y822.
When this pathway is perturbed, cell‐cell adhesion and cell stiffening are impaired. Collectively, these
findings suggest a model whereby force on E‐cadherin activates AMPK which in turn promotes both the
uptake of glucose and the production of ATP thereby providing the energy necessary for the actin
cytoskeletal rearrangements that support cell stiffening.
Force regulation of talin unfolding by a molecular clutch defines cell rigidity sensing and
A. Elosegui‐Artola1, R. Oria1,2, Y. Chen3,4, A. Kosmalska1,2, C. Pérez‐González1,2, N. Castro1, C. Zhu3,4,5, X.
Trepat1,2,6, P. Roca‐Cusachs1,2; 1Institute for Bioengineering of Catalonia, Barcelona, Spain, 2University of
Barcelona, Barcelona, Spain, 3Woodruff School of Mechanical Engineering, Georgia Institute of
Technology, Atlanta, GA, 4Parker H. Petit Institute for Bioengineering and Bioscience, Georgia Institute of
Technology, Atlanta, GA, 5Coulter Department of Biomedical Engineering, Georgia Institute of
Technology, Atlanta, GA, 6Institució Catalana de Recerca i Estudis Avançats , Barcelona, Spain
Cell function is highly dependent on substrate rigidity, which cells probe by first applying forces, and
then transducing them into biochemical signals. However, the molecular mechanisms that drive this
two‐step process remain unclear. Here we addressed this problem by combining cell force
measurements, molecular biology, theoretical modelling, and single molecule force spectroscopy. Our
results show that, as cells pull on their substrate through the actin‐talin‐integrin‐fibronectin link, the
mechanical properties of the link result in a rigidity threshold. Below the threshold, the link unbinds
before talin can unfold. Above the threshold, talin unfolds, leading to vinculin binding, adhesion growth,
force transmission, and activation of YAP signaling. The force/rigidity relationship, and the rigidity
threshold for unfolding, can be tuned according to a quantitative clutch model by regulating talin
mechanical stability, cell contractility, integrin ligation, or ECM coating density. Our results clarify the
molecular mechanism by which substrate rigidity and composition regulate force transmission, and how
force transmission in turn triggers mechanotransduction.
T cell membrane adhesion and organization on engineered surrogate Antigen Presenting Cells.
P. Dillard1,2, F. Pi2, L. Limozin1, K. Sengupta2; 1INSERM UMR 1067 AMU‐CNRS UMR 7333, LAI, Marseille,
France, 2AMU‐CNRS UMR 7325, CINaM, Marseille, France
T cells play a central role in cell mediated adaptive immune response. They carry distinctive receptors
called T cell receptors (TCR), and are responsible for specific recognition of foreign peptides displayed on
antigen presenting cells (APC). One of the prerequisites for this recognition is the adhesion of the T cell
to the APC. It has been shown that most diseases of the immune system, including AIDS and auto‐
immune diseases, impact this process. It is therefore very important to understand the relation between
the ligands at the molecular level and the biophysical behavior of the cell.
Here we focus on two important aspects of ligand presentation ‐ their mobility and spatial distribution.
The mobility of the TCR ligands on the surface of the APC depends on their nature, for example, the
diffusion is practically zero for dendritic cells, and much higher for B lymphocytes. Furthermore, it was
shown recently that the ligands are presented in the form of sub‐micronic clusters. We re‐created these
specific features of the T cell micro‐environment on synthetic bio‐mimetic substrates, either coated with
mobile or immobilized anti‐CD3 (targeting the TCR complex) [Dillard et al. 2014], or presenting nano‐
dots of anti‐CD3 [Pi, Dillard et al. 2013 and 2015]. We study the morphodynamics of human T
lymphocytes on these surrogate Antigen Presenting Cells.
The extent of cell spreading is about four fold greater on substrates carrying fixed ligands compared to
those with mobile ligands. Interestingly, in the absence of myosin activity or in the presence of ICAM,
the differences between mobile and immobile case disappear. Based on the quantification of spreading
dynamics, and actin retrograde flow, we propose a theoretical model that links the dynamics of the
leading edge of the spreading T cell, driven by actin polymerization, to the friction generated at the
surface by dragging or pinning of the ligands.
In a second series of experiments we explore the adhesion of T cells to submicron‐scale islands of anti‐
CD3 embedded in a non‐adhesive background of PEG. Global characteristics like cell‐substrate contact
area and cell shape are determined mainly by PEG density, but the molecular organization of the TCRs as
well as the kinase ZAP‐70 are strongly modulated by the dots. The nature of TCR cluster recruitment to
the nano‐dots supports the idea that the TCRs may be pre‐clustered on quiescent cells. Dynamic early
time data indicates that the ZAP‐70 microclusters are directly recruited to the site of the antibody dots
and this process is concomitant with membrane adhesion. These results together point to a complex
interplay of adhesion, molecular organization and activation in response to spatially modulated
Mechano‐sensitivity of nascent adhesions on soft substrates revealed by fluorescence
fluctuation analysis and traction microscopy.
S.J. Han1, A. Bachir2, A.R. Horwitz2, G. Danuser1; 1Department of Cell Biology, University of Texas
Southwestern Medical Center, Dallas, TX, 2Department of Cell Biology, University of Virginia,
Charlottesville, VA
Cell‐matrix adhesions are a key mechano‐transducer that converts mechanical signals (e.g. rigidity) in
extracellular matrix (ECM) into biochemical signals, which in turn modulate critical cellular processes
including differentiation, survival and migration. Mechanotransduction via cell‐matrix adhesions
depends on an intricate network of putative molecular interactions, some of which, in living cells, have
been characterized at the level of large, mature focal adhesions. However, it has been elusive whether
the mechano‐sensing nature of adhesions also exists before they reach their full maturation: i.e. in the
state of nascent adhesions, which are diffraction‐limited complexes coupled to the lamellipodial F‐actin
meshwork. In this study we examine if ECM stiffness is sensed at the level of nascent adhesions through
altered force transmission, protein localization and interactions. By the rationale that there should be
force application on the ECM for adhesions to sense its stiffness, we used high‐resolution traction
microscopy1 reported recently to investigate whether and how much nascent adhesions transmit
localized force to the substrate. We also use high‐resolution fluorescence fluctuation spectroscopy2 to
map the formation and stoichiometry of integrin‐associated complexes in the adhesions, in which we
focus on putative integrin activating (kindlin and talin) and actin‐linking (talin, vinculin and α‐actinin)
molecules. We show that all molecules are present in adhesions as soon as they are visible, but their
association to each other in complexes was hierarchical, with different time delays and stoichiometry
that changed as the adhesion matures into larger morphologies. While these quantitative characteristics
were consistent on both glass and soft substrates, the number of vinculin and talin decreased while the
number of integrin and kindlin increased on soft substrates compared to adhesions in glass substrates.
We also show that individual and newly‐formed nascent adhesions at the leading edge of protruding
cells transmit traction forces. Among those ~30 % of exhibit forces that are highly correlated with the
integrated fluorescence intensity and the total number of early adhesion molecules. These
measurements suggest evidence of mechano‐sensitivity of a significant subset of nascent adhesions and
provide further novel information on molecular complex formation as adhesions evolve and respond to
substrate rigidity.
SJ Han, Y Oak, A Groisman, G Danuser, Traction microscopy to identify force modulation in subresolution
adhesions, 2015, Nat Methods.
AI Bachir, J Zareno, K Moissoglu, EF Plow, E Gratton, AR Horwitz, Integrin‐Associated Complexes Form
Hierarchically with Variable Stoichiometry in Nascent Adhesions, 2014, Curr Biol.
Protein Tyrosine Kinases Control Local Pinching Involved in Adhesion‐Dependent
B. Yang1, Z. Lieu1, H. Wolfenson2, F.M. Hameed 1, A.D. Bershadsky1,3, M.P. Sheetz1,2; 1Mechanobiology
Institute, National University of Singapore, Singapore, Singapore, 2Columbia University, Department of
Biological Sciences, New York, NY, 3Weizmann Institute of Science, Department of Molecular Cell
Biology, Rehovot, Israel
Matrix rigidity is an important physical aspect of cell microenvironments. Many neoplastically
transformed cells can ignore the softness of their microenvironment (1). The mechanism of how cell
tests substrate rigidity is not clear. Submicron pillar studies suggest that cells may sense rigidity by
measuring the forces required for local standard contractions at the cell periphery (pinching activity) (2).
In a siRNA screen of the human tyrosine kinases, several kinases were found to be required for the cell
rigidity sensing (3). We have focused on two kinases, AXL and ROR2, which play important roles in
neoplastic transformation and development, respectively. Knockdown of either tyrosine kinase causes
cells to build large focal adhesions on soft surfaces as well as on rigid, even though overall cell
contractility on the soft substrate remains weaker (3). Surprisingly, the mechanical characteristics of
local contractions, unlike the global ones, are significantly altered by AXL or ROR2 knockdown,
increasing either magnitude or time of contraction. Further, phospho‐AXL and ROR2 localize to
sarcomeric contraction units and associate with the major contractile components, tropomyosin‐1 and
myosin IIA (AXL), or filamin A (ROR2). Thus, we suggest that these tyrosine kinases affect adhesion‐
dependent mechanosensitivity and consequently metastasizing and morphology changes in
development through their regulation of local mechanosensory contractions.
D. T. Butcher, T. Alliston, V. M. Weaver, A tense situation: forcing tumour progression. Nat Rev Cancer 9,
108‐122 (2009); published online EpubFeb (10.1038/nrc2544).
S. Ghassemi, G. Meacci, S. Liu, A. A. Gondarenko, A. Mathur, P. Roca‐Cusachs, M. P. Sheetz, J. Hone, Cells
test substrate rigidity by local contractions on submicrometer pillars. Proc Natl Acad Sci U S A 109, 5328‐
5333 (2012); published online EpubApr 3 (10.1073/pnas.1119886109).
M. Prager‐Khoutorsky, A. Lichtenstein, R. Krishnan, K. Rajendran, A. Mayo, Z. Kam, B. Geiger, A. D.
Bershadsky, Fibroblast polarization is a matrix‐rigidity‐dependent process controlled by focal adhesion
mechanosensing. Nat Cell Biol 13, 1457‐1465 (2011); published online EpubDec (10.1038/ncb2370).
Catch‐bond adhesion of filopodia: involvement of myosin II and formins.
N.O. Alieva1, A.K. Efremov1, S. Hu1, M. Natarajan1, A.D. Bershadsky1,2; 1National University of Singapore,
Mechanobiology Institute, Singapore, Singapore, 2Weizmann Institute of Science, Rehovot, Israel
Filopodia are dynamic membrane protrusions driven by the assembly of actin filament bundles (actin
cores). They are important in many cellular activities including cell migration, neuronal growth, cell‐cell
junction formation and tumor invasion. Unconventional myosin X (MYO10), known to interact with
integrins, is a powerful inducer of filopodia formation and elongation. MYO10 is present at the filopodia
tip as a bright puncta and remains there as the filopodia extend and retract. Several members of the
formin family have been found in filopodia, though the exact roles they play in filopodia dynamics are
not fully understood. Here, we studied the effects of mechanical force application on filopodia growth
and adhesion to the extracellular matrix. By using optical tweezers to immobilize fibronectin coated
latex beads attached to the tips of MYO10‐induced filopodia, and applying force by slow movement of
the microscope stage, we succeeded in maintaining sustained filopodia growth. Strikingly, inhibition of
myosin II light chain phosphorylation by the ROCK inhibitor Y‐27632 interfered with the filopodia growth
by apparent reducing filopodia tip affinity for the bead. Using structured illumination microscopy (SIM),
we demonstrated that in non‐treated cells myosin II filaments were localized to the base region of
MYO10‐induced filopodia. Small molecular formin inhibitor SMIFH2 also suppressed affinity between
filopodia and the bead and decreased the ability of filopodia to grow upon stretching. Moreover,
treatment by SMIFH2 decreased the length of MYO10‐induced filopodia and caused gradual
deterioration of MYO10‐enriched puncta at the filopodia tip. Such deterioration occurred via pinching of
MYO10‐positive patches out of the MYO10 puncta at the filopodia tip, followed by quick removal of
these patches from the filopodia by retrograde movement towards the cell body. This retrograde
movement was myosin‐II‐dependent and could be stopped by Y‐27632 treatment. It appears formin
inhibition disrupted an association between actin filaments and the filopodia tip. Thus, our experiments
revealed a strong synergistic cooperation between formins and myosin II, required for the mechanical
tension development in filopodia actin bundles. It evokes a catch‐bond like effect in filopodia adhesion:
once internal tension (estimated to be around several pN) was maintained, filopodia successfully
attached to the bead and grew, while upon tension release, such as in cells treated with myosin II or
formin inhibitors, filopodia were no longer able to grow and subsequently detached from the bead. This
catch‐bond adhesion behavior could play an important role in filopodia‐mediated matrix rigidity sensing
and cell durotaxis.
Targeting adhesive molecules to prevent protective cancer cell niches.
J.L. Young1, S. Klar2, J.P. Spatz1,3; 1New Materials and Biosystems, Max Planck Institute for Intelligent
Systems, Stuttgart, Germany, 2Biomedical Sciences, Reutlingen University, Reutlingen, Germany,
Biophysical Chemistry, Heidelberg University, Heidelberg, Germany
Tumor progression and metastasis are ultimately driven by a disruption in the organization and
composition of the extracellular matrix (ECM). Cancer cell‐ECM interactions have been shown to
positively influence cancer cell survival and invasion by conferring adhesion‐based resistance to
chemotherapeutic drugs and subsequently upregulating pathways driving metastasis into surrounding
tissues. The frequent inability of chemotherapy to provide a long‐term, complete cure may be
attributed to this specific, adhesion‐mediated resistance of malignant cells to chemotherapeutic
interventions. Here, we present a platform designed to identify and perturb protective matrix
properties in a high‐throughput manner, allowing for the conversion of chemoresistant cells into
chemosensitive ones, and subsequently examining these properties in a novel 3D microchannel system.
Block copolymer micelle nanolithography (BCML) was utilized to create scalable arrays of defined ligand
presentation with 50‐100 nm spacing, allowing for 48 culture conditions per sample. Multiple breast
cancer cell lines were then treated with various chemotherapeutic drugs in order to identify the most
protective ligand interactions, as determined by cell survival assays and immunofluorescence. By
utilizing BCML in conjunction with soft polymer transfer nanolithography, the influence of mechanical
matrix properties in parallel with ligand presentation was analyzed on 2D substrates and inside 3D
microchannels of 125 µm diameter. Of the various peptides screened, Laminin and the cell‐binding
domain of Fibronectin were found to provide up to 2‐fold chemoresistance to 5‐Fluorouracil treatment
compared to VCAM1 or the heparin‐binding domain of Fibronectin. Immunofluoresence imaging
revealed a positive correlation between focal adhesion formation and cell survival, in that more robust
adhesions were found in chemoresistant cells. In both 2D and 3D substrates, stiff 50 kPa hydrogels were
more effective than soft 10 kPa hydrogels in upregulating cell area and polarity, both of which have
been hypothesized to play a role in chemoresistance. This system provides a scalable, high‐throughput
platform for screening large peptide libraries for the ability to promote cancer cell chemoresistance,
where survival is dependent on the composition and presentation of extracellular ligands. These
identified interactions can be perturbed in order to convert chemoresistant cells to chemosensitive cells,
aiding in the future establishment of effective combinatorial chemotherapeutic clinical strategies.
Minisymposium 13: Applications of Cell Biology 1
A synthetic transcriptional program for preventing proteostasis collapse derived from the
ancient function of Heat Shock Factor 1.
V. Denic1; 1Molecular and Cellular Biology, Harvard University, Cambridge, MA
Cells maintain protein homeostasis (proteostasis) by balancing protein synthesis with protein folding
and turnover. The extent to which gene control regulates this balance is poorly understood. Hsf1 is a
eukaryotic transcription factor that drives changes in heat shock protein (Hsp) gene expression to
counteract cytoplasmic protein aggregation following heat shock. Hsp genes are also controlled by other
heat‐activated transcription factors but it remains unknown why, among them, only Hsf1 is essential for
cell proliferation of yeast and many human cancer cell lines. Here we induced rapid nuclear export of
yeast Hsf1 by chemical genetics and detected by NET‐seq immediate transcriptional attenuation (within
15’ of export) of 18 Hsf1 gene targets out of 38, which we defined by ChIP‐seq. CRISPR/Cas9‐mediated
ablation of mammalian Hsf1 in mouse embryonic fibroblasts and mouse embryonic stem cells
attenuated heat activation of homologous gene targets. Using this ancient function of Hsf1 as a
blueprint, we designed a synthetic transcriptional program that enabled creation of cells lacking Hsf1,
which robustly proliferated without apparent protein aggregation. The minimal version of this synthetic
program comprised two Hsf1 gene targets‐encoding the Hsp70 and Hsp90 chaperones‐that are essential
and not controlled by the other two heat‐activated transcription factors in yeast. Our work establishes
the primacy of protein folding gene control in preventing proteostasis collapse. It also provides a
compact list of protein folding effectors for studying the oncogenic, neuroprotective, and anti‐aging
effects of Hsf1.
Strategies for Targeting SMARCA4 Mutant Cancer.
L.D. Belmont1; 1Discovery Oncology, Genentech, Inc., South San Francisco, CA
The SWI/SNF (BAF) complex is a chromatin remodeling complex that repositions nucleosomes. The
composition of SWI/SNF complexes varies by cell type but all complexes contain one of two mutually
exclusive ATP dependent helicases, SMARCA4 (BRG1) or SMARCA2 (BRM), that disrupt nucleosome/DNA
contacts to allow repositioning. Recently, large scale cancer sequencing efforts have revealed that 20%
of all cancers harbor a mutation in at least one subunit of the SWI/SNF complex, including 5 to 10% of
non‐small cell lung cancers that have lost functional BRG1 protein. We have confirmed that loss of
BRG1 renders the cancer cells dependent upon BRM for survival, making it an appealing drug target for
BRG1 mutant cancers. However, there are challenges to selectively inhibiting BRM. As such, we are
evaluating multiple approaches to targeting BRG1 cancer. The approaches include novel strategies for
inhibiting BRM and running a CRISPR/Cas9 dropout screen to identify additional targets in BRG1 mutant
Mining Cellular Heterogeneity for Mechanistic Insights in Phenotypic Profiling and Drug
A.H. Gough1,2, T. Shun2, D.L. Taylor1,2,3, M. Schurdak1,2,3; 1Computational and Systems Biology, University
of Pittsburgh, Pittsburgh, PA, 2Drug Discovery Institute, University of Pittsburgh, Pittsburgh, PA, 3Cancer
Institute, University of Pittsburgh, Pittsburgh, PA
Heterogeneity is well recognized as a common property of cellular systems that impacts biomedical
research and the development of therapeutics and diagnostics. Several studies have shown that
analysis of heterogeneity gives insight into mechanisms of action of perturbagens, can be used to
predict optimal combination therapies, and to quantify heterogeneity in tumors where heterogeneity is
believed to be associated with adaptation and resistance. Cytometry methods including high content
screening (HCS), high throughput microscopy, flow cytometry, mass spec imaging and digital pathology
capture cell level data for populations of cells. However it is often assumed that the population
response is normally distributed and therefore the population average adequately describes the result
of the measurement. A deeper understanding of the biological mechanisms, and a more effective
comparison of perturbagen effects, require analysis that takes into account the distribution of the
cellular phenotypes. However, the reproducibility of heterogeneous data collected on different days,
and in different plates has not previously been evaluated. Here we show that conventional assay quality
metrics alone are not adequate for quality control of the heterogeneity in the cellular response. To
enable routine analysis of heterogeneity in screening/profiling we have developed an effective means to
normalize population distributions from plate‐to‐plate; a standard approach to quality control of
heterogeneity in large‐scale biology projects using the Kolmogorov Smirnov (KS) statistic; and
demonstrate the use of a set of three heterogeneity indices that measure diversity, normality and
outliers in the population to quantify, compare and review variations in heterogeneity in thousands of
distributions resulting from treatment with perturbagens. We apply these methods in a retrospective
analysis of heterogeneity in an SAR screen for inhibitors of STAT3 activation by IL‐6. Results of this
analysis show that there are: reproducible and compound dependent variations in the distribution of
cellular responses; some compounds only inhibit STAT3 activation in a subpopulations of cells leading to
a bimodal distribution of activity; heterogeneity in the inhibition of STAT3 may result in a misleadingly
potent IC50 determination; and that compounds with different mechanisms of action exhibit differences
in the distribution of cellular responses. These results indicate that quantifying and comparing
heterogeneity may lead to a better understanding of compound mechanisms, help prioritize compounds
for development and help guide an SAR on a particular mechanism, even when the target is not known.
The metrics and methods are presented as a workflow for analysis of heterogeneity.
Development of cellular morphology‐based separation system for three‐dimensional culture.
H. Matsui1, M. Tamura2, S. Sugiura2, R. Kato3, M. Yanagisawa4, T. Kanamori2; 1Faculty of Medical
Sciences, University of Tsukuba, Tsukuba, Japan, 2National Institute of Advanced Industrial Science and
Technology (AIST), Tsukuba, Japan, 3Nagoya Unversity, Nagoyaa, Japan, 4Enggineering System Co., Ltd.,
Matsumoto, Japan
The present study includes proof concept of morphology‐based cell separation and development of
automatic cell separation system. 3D culture environment with a specific extracellular matrix regulates
cellular function and phenotype. In addition, cancer cell morphology changes depending on its
malignancy in 3D culture environment. The cell separation system in 3D culture environment should
need to obtain the cells according to its morphology, which includes cell phenotypes. Recently, we
developed gelatin‐based photodegradable hydrogels, and applied this hydrogels to optical cell
separation. The target cells in the photodegradable hydrogels were successfully separated by the optical
cell separation, the separated cells was growth on another dish. On the other hand, we recently
developed the predication model of stem cell differentiation by image analysis. The image analysis
technique and the photodegradable gelatin hydrogels are included in the automated morphology‐based
cell separation system in the 3D culture environment. For forming cell encapsulated‐photodegradable
hydrogels, suspension of cells including heterogeneous population is mixed with pregel solutions and
cells are encapsulated in the gelatin‐based photodegradable hydrogels. After the culture in 3D
environment, microscopic images of the cells are captured. The captured images are analyzed to
distinguish the target cells from the other cells by using the image analysis algorithm, which we
previously developed for analyzing stem cells. The hydrogels around the target area is irradiated with UV
light. The cells in the irradiated area are collected by automated pipetting system. We developed
automated system for this optical cell separation procedure, including cultivation, image acquisition,
image analysis, light irradiation, and pipetting for cell collection. We demonstrated automated optical
cell separation using the model culture system. Normal gastric mucosal cells were cultured in the
photodegradable hydrogels. After cultivation for 1 week, the cells were irradiated UV light for 5 to 20
min. The cells in the irradiated area were collected by automated pipetting and transferred into a
collection dish. The collected cells were viable and attached in the collection dish after collection. We
are currently developing an automated image analysis algorithm to distinguish cancer cells from normal
cells under 3D environment. The automated optical cell separation system with image analysis algorithm
will be applied to the establishment of novel cancer‐cell lines from clinical samples such as biopsy tissue.
Applying m‘TORC’1 to prolong vision in Retinitis pigmentosa.
A. Venkatesh1, S. Ma1, C. Punzo1; 1Ophthalmology and Gene Therapy Center, University of
Massachusetts Medical School, Worcester, MA
Retinitis pigmentosa (RP) is an inherited retinal disease characterized by an initial loss of night‐active rod
photoreceptors, followed by the degeneration of cones, leading to complete loss of daylight vision. The
mechanism of secondary cone death in RP has remained enigmatic and its understanding is critical to
devise strategies to prolong vision in the disease. We previously proposed that following the initial loss
of rods, cones suffer from metabolic imbalances, induced by a shortage of nutrients, particularly of
glucose (Punzo et al). Our preliminary data suggested that modulation of the insulin/mammalian target
of rapamycin (mTOR) pathway, a key pathway regulating cell growth, metabolism and survival, would
promote cone survival.
We therefore performed an in depth analysis by crossing mice carrying conditional knockout alleles of
various genes of the pathway into the retinal degeneration‐1 (rd1) mouse model of RP. These mice were
then crossed to a cone‐specific Cre driver line to achieve deletion in cones. The genes deleted were as
Pten (loss activates both mTOR complexes, the pro‐growth mTOR complex I (mTORC1) and the pro‐
survival mTOR complex II (mTORC2))
Raptor (loss eliminates mTORC1 activity)
Rictor (loss eliminates mTORC2 activity)
Tsc1 (loss activates only mTORC1)
Concurrent loss of Pten & Raptor
Concurrent loss of Tsc1 & Raptor
Concurrent loss of Tsc1 & Rictor
Our comprehensive analysis revealed that activation of mTORC1 upon loss of Pten or Tsc1 was both
required and sufficient to promote long‐term cone survival, while mTORC2 was dispensable for cones.
mTORC1 activation increased expression of genes involved in glucose metabolism, enabling cones to
better regulate their metabolic demands to overcome the nutrient shortage. Since removal of Tsc1 is a
more robust activator of mTORC1, its loss conferred an initial stronger survival advantage to cones,
when compared to removal of Pten.
However, because mTORC1 is a negative regulator of autophagy, Tsc1 loss also resulted in an
accumulation of autophagic aggregates positive for ubiquitin and p62, inducing a shortage of free amino
acids in cones. This caused an eventual decline in the efficiency of cone survival mediated by Tsc1 loss.
On the other hand, while loss of Pten caused only a moderate increase in mTORC1 activity, autophagy
was not inhibited, resulting in a more sustained survival effect over‐time.
Our work suggests that to achieve long‐term cone survival in RP, mTORC1 activity should be maintained
at levels that balance both the expression of key metabolic genes as well as autophagy. Besides
providing a novel therapeutic approach to prolong vision in RP, our work presents a detailed
characterization of various effects of manipulating mTOR signaling in cones.
Using tardigrades to investigate mechanisms and applications of desiccation tolerance.
T.C. Boothby1, H. Tapia2, A.H. Brozena3, D.E. Koshland2, B. Goldstein1; 1Department of Biology, University
of North Carolina, Chapel Hill, NC, 2Department of Molecular Cell Biology, University of California,
Berkeley, Berkeley, CA, 3Department of Chemical and Biomolecular Engineering, North Carolina State
University, Raleigh, NC
Most life requires water. However, there are some organisms that are able to survive essentially 100%
water loss. Understanding desiccation tolerance promises to contribute to many applications, such as
engineering of drought tolerant plants and the stabilization of biomaterials. Being able to stabilize
biological materials dry at room temperature would be economically and logistically valuable. For
example, ~80% of the costs of vaccination programs in developing countries comes from having to keep
vaccines cold. To consider new strategies for engineering of crops and reducing our dependency on the
cold chain, we are studying rare cases where nature already achieves this feat. To this end, we are
investigating mechanisms of desiccation tolerance in tardigrades (water bears), tiny invertebrate
animals capable of surviving extreme abiotic stresses including prolonged desiccation, freezing and
boiling temperatures, intense ionizing radiation, and even the vacuum of outer space. We have found
that multiple tardigrade‐specific genes that encode intrinsically disordered proteins (IDPs) are
upregulated during drying, and that these genes are required for tardigrades to survive drying.
Interestingly these genes do not appear to be essential for general survival or for surviving other
stresses. CD spectroscopy suggests that these IDPs and others like them lack secondary structure while
in aqueous solution, but adopt well‐defined alpha helices during water loss. Consistent with this, we
found that when expressed in HeLa cells, desiccation induced a relocalization of these IDPs, which under
hydrated conditions appeared diffuse throughout cells’ cytoplasm, to specific cytoplasmic organelles –
suggesting that individual proteins are targeted to different parts of cells, perhaps protecting specific
cellular compartments. We found in vitro these proteins formed biological glasses when dried.
Bioglasses have been proposed to help prevent the denaturation and aggregation of other proteins as
well as maintain the integrity of membranes in dried cells. To test whether these proteins can act
exogenously, we expressed these proteins in bacteria and yeast. We found that this increased the
desiccation tolerance of these cells by 1 to 2 orders of magnitude and resulted in the formation of novel
glassy materials in vivo. We are currently investigating the mechanisms by which these IDPs protect cells
and their purified components (e.g., proteins and membranes) during desiccation. We hope to answer
fundamental questions about biological materials can survive desiccation, which could contribute to the
engineering of drought resistant crops and provide an avenue for pursuing technologies for the
stabilization of biomedical materials in a dried state.
Leveraging the Cell Biology of Metabolic Enzymes to Uncover New Insights Into Orphan Genetic
R.M. Broyer1, C. Noree1, E. Monfort1, J.E. Wilhelm1; 1Cell and Developmental Biology, UC San Diego, La
Jolla, CA
There are approximately 7000 orphan diseases affecting 24 million people in the United States.
Although the biochemical basis for many of these “inborn errors of metabolism” is well established,
most of these diseases have symptoms that are difficult to explain in terms of the known biochemical
function of the enzyme. As a result, most efforts to treat these diseases have focused on gene therapy
or enzyme replacement. We have focused on leveraging the insights we have gained from our work on
the cell biology of metabolic pathway organization to develop a new framework for considering these
diseases that will allow the development of new treatment approaches. Recently, we have identified 59
metabolic enzymes that form novel intracellular structures in yeast raising the possibility that some of
the phenotypes observed in these genetic diseases might be due to functions that are separable from
the catalytic activity of the enzyme. To test this possibility, we have focused on human PRPP synthase
(PRPS1) since the genetics of this disease have long been perplexing because loss of function and
superactvity/feedback resistance mutations both cause the same neurological syndrome: ataxia and
deafness. We have found a potential solution to this paradox. Endogenous PRPS1 forms a novel nuclear
filament in wild type cells. In contrast, the cytoplasm fills with PRPS1 filaments in fibroblasts from
patients with feedback resistance mutations in PRPS1. This argues that the cytoplasmic filaments may
dominantly interfere with the nuclear function of the PRPS1 filament. Furthermore, we have identified
several small molecule treatments that can be used to selectively eliminate the cytoplasmic filaments
while leaving the nuclear filament unaffected.
In order to define the mechanism that leads from alteration in PRPS1 polymerization to deafness and
ataxia, we have identified a novel regulator of PRPS1 enzyme activity that binds to PRPS1 filaments and
that also regulates the actin cytoskeleton. We propose that alterations in PRPS1 polymerization disrupt
actin organization via this protein leading to deafness and ataxia. We have also identified a family with a
novel dominant form of PRPS1 superactivity disorder where the mutation lies in a gene other than
PRPS1. We are currently testing whether the mutation disrupts our PRPS1 filament binding protein or
another uncharacterized gene in the pathway. These results argue that orphan diseases may hold the
key to uncovering novel aspects of cell biology that can be leveraged for the diagnosis and treatment of
this understudied set of diseases.
Reconstituting chemically induced dimerization system as a potential artificial chemoattractant
sensing mechanism in giant liposomes.
S. Razavi1,2, L. Tianzhi1, D.N. Robinson1,3, T. Inoue1,2,4; 1Cell Biology, Center for Cell Dynamics, Johns
Hopkins School of Medicine, Baltimore, MD, 2Biomedical Engineering, Johns Hopkins University School
of Medicine, Baltimore, MD, 3Chemical and Biomolecular Engineering, Whiting School of Engineering,
Johns Hopkins University, Baltimore, MD, 4Precursory Research for Embryonic Science and Technology
Investigator, Tokyo, Japan
In the highly intertwined cellular milieu it is difficult to extract the minimal signaling events that play a
key role in chemotaxis. We aim to bypass this challenge by taking a reductionist approach through
engineering an artificial cell that is built one molecule type at a time, is minimal in the number of
components, and can migrate towards a desired target. Given our overall design of artificial migratory
cells, we needed to engineer these cells such that they can sense rapamycin as the potential
chemoattractant. This could allow us to obviate the need to reconstitute membrane proteins as the
sensors of the chamoattractive signal. In this work we made giant unilamellar vesicles (GUVs) that are
able to respond to rapamycin. Rapamycin is known to lead to the dimerization of the FKBP and FRB
protein pairs. We fabricated GUVs that contain the two protein components of the chemically induced
dimerization (CID) system: purified FKBP and FRB proteins expressed in bacteria. Upon addition of
rapamycin, the fluorescently‐tagged FKBP and FRB reconstituted in the GUV lumen dimerized and
created a FRET output on a time‐scale of less than one minute. We further designed the FKBP fused with
the MARCKS positively charged lipid binding domain. This allowed us to anchor FKBP to the negatively
charged inner leaflet of the GUV membrane while keeping FRB luminal. Rapamycin addition resulted in
fast translocation of the FRB towards the membrane‐bound FKBP. This indicates that rapamycin could
penetrate the artificial vesicles and the purified FKBP and FRB proteins remained functional after
reconstitution in GUVs. This manipulation technique can be used to bring various reconstituted signaling
molecules to the GUV membrane. In turn, leading to the activation of processes such as actin
polymerization and GUV membrane deformation as the initial steps needed to trigger motility in
artificial cells. Recapitulation of the CID system in GUVs can also be used to study the spatio‐temporal
topology of signaling molecules, in realtime, within minimal model systems.
Cardiac‐inducing RNAs direct the differentiation of stem cells into cardiomyocytes for heart
A. Kochegarov1, M.S. Neal1, A. Davis1, L. Mitchell1, N. Scarcelli1, G. Vaughn1, L.F. Lemanski1, H. Fetters1;
Biological and Environemental Sciences, Texas AM University, Commerce, TX
Cardiomyocytes are replaced by fibrous scar tissue rather than new cardiac muscle cells after a
myocardial infarction. This weakens the organ’s ability to contract normally and increases the risk of
additional heart attacks. In the present studies, we developed two human‐derived functional homologs
of an RNA that promotes heart development in cardiac non‐function mutant salamanders, Ambystoma
mexicanum, (Zhang et al., 2009, J. Biomed Sci, 16;81). These RNAs are termed Cardiac‐Inducing RNAs
(CIR). One of the RNAs is associated with the mitochondrial cytochrome c oxidase gene and a second
with the caspace recruitment gene. We have found that both are capable of inducing tropomyosin
synthesis and myofibril formation in mutant axolotl myocardial cells which ordinarily do not synthesize
tropomyosin or form myofibrils. We have found further that mouse embryonic fibroblasts transfected
with the human CIRs significantly increase the expression of specific proteins after only one week
incubation demonstrating the fibroblast’s induction into cardiac cell lineages. This approach may offer a
novel way to treat myocardial infarctions by reprogramming an individual’s fibroblasts to form into
functioning cardiomyocytes. Stem cells or fibroblasts transfected with the human CIRs form into cells
characteristic of early developing cardiomyocytes, and express cardiac protein markers including cardiac
specific cardiac troponin‐T, tropomyosin and α‐actinin as detected by immunohistochemical staining.
Furthermore, these contractile proteins organize into sarcomeric myofibrils characteristic of striated
cardiac muscle cells. Computer analyses of the CIR secondary structures reveal significant similarities to
the myofibril‐inducing RNA (MIR) secondary structure described in salamander, that also promotes
nonmuscle cells to differentiate into cardiac muscle. Thus, the axolotl and human RNAs appear to have
evolutionarily conserved secondary structures suggesting that both play major roles in vertebrate heart
development and in the differentiation of cardiomyocytes from non‐muscle cells during development.
Our current goal is to generate a myocardial infarction model in mice by coronary artery ligation and
inject CIR‐treated stem‐cell‐derived cardiomyocytes to regenerate the damaged areas of the infarcted
mouse hearts leading to a full recovery of the damaged heart muscle tissue. We hope in the future that
this type of approach can be used to treat myocardial infarctions in human patients such that heart
attack victims could return to normal pre‐heart‐attack activity levels. (NIH grant HL61246, NSFRU1 grant
1121151 and an American Heart Association grant to LFL).
Minisymposium 14: Cytokinesis
Nanoscale architecture of the actomyosin cortex during cell division.
B. TRUONG QUANG1,2, P. Chugh1,2, M. Smith1,2, G. Salbreux3, E.K. Paluch1,2; 1UCL, MRC Laboratory for
Molecular Cell Biology, London WC1E 6BT, UK, 2UCL, Institute for the Physics of Living Systems, London
WC1E 6BT, UK, 344 Lincoln's Inn Fields, The Francis Crick Institute, London WC2A 3LY, UK
Animal cells undergo dramatic cell shape changes during cell division. Mitotic cell rounding and cleavage
furrow ingression in cytokinesis are mediated by precisely regulated changes in the contractility of cell
cortex, a thin actomyosin meshwork directly beneath the plasma membrane. To understand the
molecular control of cortical contractility, we need to understand the nanoscopic organisation of the
major cortical components, such as actin and myosin II. Due to the tiny size of the cortical components
and the limited resolution of conventional fluorescent microscopy methods, the spatial organization of
the cortical network remains mostly unknown. We used superresolution optical microscopy to image
the transversal distribution of cortical actomyosin in live cells. We found that the cortex undergoes
significant thinning between interphase and mitosis. Interestingly, we found that myosin II does not
overlap with the entire thickness of the actin layer, suggesting the existence of a highly dense actin
region in the vicinity of the plasma membrane from which myosin minifilaments might be sterically
excluded. Furthermore, by imaging individual myosin II minifilaments with structured illumination
microscopy, we showed that the spatial orientation of myosin minifilaments in the cortex changes
towards more contractile configurations as cells transition from interphase to mitosis. Together with an
agent‐based physical model of the cortex, our data bring new insight into how the nanoscale
architecture of the actomyosin cortical network controls the generation of contractile tension at the cell
RhoA activation is sufficient to induce cleavage furrow formation in metaphase and anaphase:
an optogenetic analysis of cytokinesis.
E. Wagner1, M. Glotzer1; 1Department of Molecular Genetics and Cell Biology, The University of Chicago,
Chicago, IL
The small GTPase RhoA is required for actomyosin‐based contractile ring assembly and furrow
ingression during cytokinesis. Positioning of RhoA activity is mediated by the mitotic spindle and is
critical to ensure that cells divide properly and generate two genetically equivalent daughter cells.
However, several key questions remain unanswered. Is a local zone of RhoA activity sufficient to give
rise to furrow formation? Is the entire cortex equally responsive or does the spindle modulate the
response to active RhoA? Does active RhoA induce positive feedback? Does the cell cycle play a role in
regulating the response to RhoA activation? To address these questions we engineered the optogenetic
system TULIPs to gain tight spatial and temporal control of RhoA activity in mammalian cells. Light‐
mediated recruitment of a RhoGEF domain to the plasma membrane leads to rapid induction of RhoA
activity, as indicated by accumulation of a RhoA biosensor, myosin II, and F‐actin. During anaphase,
light‐mediated RhoA activation at the cell poles induces contractility that slows or abrogates ingression
of the endogenous furrow. This demonstrates our ability to generate physiologically relevant levels of
RhoA activity. In anaphase cells in which we have blocked the endogenous upstream activation
pathway, light‐mediated RhoA activation at either the equatorial or polar cortex generates similar
furrowing behavior, suggesting that astral microtubules do not inhibit RhoA‐induced contractile
behavior. However the kinetics and extent of ingression (~35%) are are slower and less complete
compared to normal cytokinesis. When the light stimulus is removed, furrows regress, indicating that
active RhoA does not strongly induce positive feedback that maintains furrow ingression. Cells furrow
both in metaphase and anaphase indicating the contractile response downstream of RhoA activation is
not cell cycle regulated. These results reveal that RhoA activation is sufficient to drive cytokinetic furrow
formation, that the induction of cytokinetic furrowing is primarily controlled at the level of RhoA
activation, and suggests that astral microtubule‐mediated inhibition of furrowing acts on RhoA or its
The cell cortex is an excitable medium.
W.M. Bement1, A.L. Miller2, M. Leda3, A. Goryachev3, G. von Dassow4; 1Laboratory of Cell and Molecular
Biology, University of Wisconsin‐Madison USA, Madison, WI, 2Department of Molecular, Cellular and
Developmental Biology, University of Michigan, Ann Arbor, MI, 3Centre for Synthetic and Systems
Biology, University of Edinburgh, Edinburgh, United Kingdom, 4Oregon Institute for Marine Biology,
Charleston, OR
Animal cell cytokinesis results from patterned activation of the small GTPase Rho, which directs
assembly of actomyosin in the equatorial cortex. Cytokinesis is restricted to a portion of the cell cycle
following anaphase onset in which the cortex is responsive to signals from the spindle. We show that
shortly after anaphase onset oocytes and embryonic cells of frogs and echinoderms exhibit cortical
waves of Rho activity and F‐actin polymerization. The waves are modulated by cyclin‐dependent kinase
1 (Cdk1) activity and require the Rho GEF (guanine nucleotide exchange factor), Ect2. Surprisingly,
during wave propagation, while Rho activity elicits F‐actin assembly, F‐actin subsequently inactivates
Rho. Experimental and modeling results show that waves represent excitable dynamics of a reaction
diffusion system with Rho as the activator and F‐actin the inhibitor. We propose that cortical excitability
explains fundamental features of cytokinesis including its cell cycle regulation.
A Dynamic Steady State within ESCRT‐III Polymers at the Cytokinetic Abscission Site.
B. Mierzwa*1,2, N. Chiaruttini*3, J. König1,4, I. Poser5, A.A. Hyman5, T. Müller‐Reichert1,4, A. Roux3, D.W.
Gerlich1,2; 1Marine Biological Laboratory, Woods Hole, MA, 2Institute of Molecular Biotechnology of the
Austrian Academy of Sciences (IMBA), Vienna, Austria, 3Department of Biochemistry, University of
Geneva, Geneva, Switzerland, 4Medical Theoretical Center, Dresden University of Technology, Dresden,
Germany, 5Max Planck Institute for Molecular Cell Biology and Genetics, Dresden, Germany
*These authors contributed equally to this work.
The Endosomal Sorting Complex Required for Transport (ESCRT)‐III mediates membrane fission in
various cellular processes including multivesicular body formation, cytokinetic abscission, and virus
budding. ESCRT‐III is thought to function as a filamentous polymer structure, which initially deforms flat
membranes into tubes and subsequently constricts tubes until the membranes split. Membrane
constriction and ESCRT‐III subunit recycling depend on VPS4, an ATPase with ESCRT‐III depolymerization
activity. Prevailing models for ESCRT‐III‐mediated membrane fission assume sequential phases of ESCRT‐
III polymerization, VPS4 recruitment, membrane neck constriction, and ESCRT‐III polymer disassembly,
yet how these sequential phases are coordinated is poorly understood. It is also not known whether
ESCRT‐III polymers exchange their subunits with a soluble cytoplasmic pool, which is a common feature
of other force‐generating filament systems like actin and tubulin.
Here, we performed a systematic analysis of ESCRT‐III polymer dynamics in live human cells and in an in
vitro polymerization assay. We found that at cytokinetic abscission sites ESCRT‐III polymers undergo
rapid and complete subunit exchange during net polymer growth, about an order of magnitude faster
than net accumulation of ESCRT‐III. The dynamic steady state of ESCRT‐III depended on VPS4, which
accumulated simultaneously with ESCRT‐III at the abscission site. To study how individual ESCRT‐III
components contribute to polymer steady‐state dynamics, we reconstituted polymerization of
recombinant budding yeast ESCRT‐III subunits on supported lipid membranes. The scaffold subunit Snf7
alone polymerized into stable patches, and its polymerization was suppressed by the putative ESCRT‐III
capping subunits Vps2/Vps24. When Vps4 was added in presence of soluble Snf7, the polymers were
maintained in a dynamic steady state, with subunit exchange kinetics similar to those observed with
human ESCRT‐III subunits at the cytokinetic abscission site.
Thus, as known for many other cellular filament systems, ESCRT‐III polymers undergo continuous
steady‐state remodeling, mediated by a constitutive activity of VPS4. The high rates of subunit exchange
might leverage the amount of energy fueled into membrane deformation, and it might contribute to the
adaptation of the ESCRT‐III system to variable membrane geometries such as nanometer‐sized vesicles
and micrometer‐sized tubes as observed during abscission.
Fak‐Src signaling pathway controls the timing of abscission by decelerating Plk1 degradation
and subsequent recruitment of Cep55 at mid‐body.
S.A. Kamranvar 1, D.K. Gupta1, S. Johansson1; 1Department of Medical Biochemistry and Microbiology,
Biomedical Center, Uppsala University , Uppsala, Sweden
Cell adhesion to extracellular matrix (ECM) is required for progression through the G1 and cytokinesis
phases. However, cancer cells acquire the ability to circumvent this control mechanism, a malignant
phenotype known as anchorage‐independent growth.
This study aimed to identify adhesion‐dependent steps regulating cytokinesis of non‐transformed
human fibroblasts. We show that cell anchorage to ECM is required for the completion of abscission at
the final stage of fibroblast cytokinesis, during which the endosomal sorting complex required for
transport (ESCRT) machinery cleaves the thin intercellular membrane bridge connecting two nascent
daughter cells. Centrosomal protein 55 (Cep55), a key protein involved in the abscission process, was
localized to the mid‐body of both adherent and non‐adherent fibroblasts but unable to efficiently recruit
Alix, TSG101 and the ESCRT III subunit Chmp4B in the non‐adherent cells. Polo‐like kinase 1 (Plk1), which
plays a critical role in abscission timing by preventing premature recruitment of Cep55 to the mid‐body,
was degraded more rapidly in the non‐adherent cells. The Fak‐Src signaling pathway, downstream of
integrin‐mediated cell anchorage to ECM, was found to be critically involved in the timing of mid‐body
maturation by decelerating Plk1 degradation and Cep55 accumulation at the mid‐body required for the
successful abscission in the fibroblast cells.
Loss of p120‐catenin induces cancer multinucleation and polyploidy through cytokinesis failure.
R. van de Ven1, J. de Groot1, D. Park2, R. van Domselaar1, D. de Jong3, K. Szuhai3, E. Sahai2, P.J. van Diest1,
M.W. Hetzer4, P.W. Derksen1, E. van der Wall5; 1Pathology, UMC Utrecht, Utrecht, Netherlands, 2Tumour
Cell Biology Laboratory, Cancer Research UK London Research Institute, London, UK, 3Molecular Cell
Biology, Leiden University Medical Center, Leiden, Netherlands, 4Molecular and Cell Biology Laboratory,
Salk Institute for Biological Studies, La Jolla, CA, 5Internal Medicine, UMC Utrecht, Utrecht, Netherlands
High‐grade invasive tumors often show loss of cell‐cell adhesion and are characterized by nuclear atypia
and aneuploidy. We have established that mammary‐specific inactivation of the adherens junction (AJ)
through loss of either E‐cadherin or p120‐catenin (p120) causes tumor invasion and metastatic
dissemination. However, in sharp contrast to E‐cadherin loss, we observed that inactivation of p120
leads to mammary carcinomas that are characterized by multinucleation and nuclear aytpia. Our
experiments in mouse models and human cancer cells now show that p120 controls cytokinesis and
suppresses the induction of tumor polyploidy. Our data point to a mechanism whereby p120 acts as a
Rho GDP dissociation inhibitor to control spatiotemporal RhoA localization, actomyosin contraction and
subsequent furrow ingression at the equatorial cortex during anaphase‐telophase. For the correct
localization and regulation of RhoA‐dependent cleavage furrow ingression, membrane‐uncoupled p120
depends on concomitant binding to RhoA and a core component of cell division: the centralspindlin
component MKLP1. In conclusion, our data demonstrate that loss of p120, a frequently occurring event
in human cancer that correlates with a poor prognosis, not only induces metastatic dissemination
through destabilization of cell‐cell junctions, but also induces tumor progression through aberrant
cytokinesis, leading to multinucleation and chromosomal instability.
Tension generated by cytokinesis remodels cell‐cell junctions and recruits Anillin.
T.R. Arnold1, T. Higashi1, K.M. Dinshaw1, R.E. Stephenson1, A.L. Miller1; 1Molecular Cellular and
Developmental Biology, University of Michigan, Ann Arbor, MI
During cytokinesis in an epithelial tissue, a contractile ring generates force in order to pinch the dividing
cell in two. The impact that the tension generated by the contractile ring has on cell‐cell junctions
remains poorly understood, despite the fact that it has implications for tissue integrity and epithelial
barrier function. Here, we investigate how tension from the contractile ring affects cell‐cell junctions
using live imaging in gastrula‐stage Xenopus laevis embryos. First, using fluorescence recovery after
photobleaching, we found that adherens junction proteins are stabilized at the cleavage furrow of
dividing cells, whereas the stability of tight junction proteins is not affected. This suggests that tension
from the contractile ring is transmitted to adherens junctions, and in response, the adherens junctions
strengthen the adhesive connection between the dividing cell and neighboring cell to maintain tissue
integrity throughout cytokinesis. Second, we investigated how the scaffolding protein Anillin responds
to increased junctional tension. We recently reported that Anillin localizes to cell‐cell junctions and is
important for junction structure and function; however, it is not known how Anillin is recruited to
junctions. Using tagged Anillin and a fluorescent tension‐sensing probe, we found that Anillin is
recruited to junctions experiencing tension generated by the cytokinetic cell. Anillin also accumulates at
junctions that are experiencing increased tension generated by other means, such as laser‐induced cell
wounding. Work in progress addresses which of Anillin’s domains are required for its tension‐
dependent recruitment to junctions and the role of Anillin accumulation in response to increased
junctional tension. The data presented here demonstrate that tension generated by cytokinesis
promotes molecular remodeling of cell‐cell junctions and that Anillin may be important in mediating
these responses.
Minisymposium 15: Endo‐Lysosome Trafficking in Development and Disease
The protein architecture of the yeast endocytic machinery analyzed by FRET.
M. Skruzny1,2, G. Malengo1,2, V. Sourjik1,2; 1Max Planck Institute for Terrestrial Microbiology, Marburg,
Germany, 2LOEWE Center for Sythetic Microbiology (SYNMIKRO), Marburg, Germany
Clathrin‐mediated endocytosis is a key cellular trafficking pathway responsible for homeostasis of the
plasma membrane, uptake of hormones and nutrients, and regulation of many signaling pathways. To
form an endocytic vesicle from a small piece of the plasma membrane dozens of endocytic proteins
have to assemble at the endocytic site in highly coordinated manner. Although the timeline of their
assembly has been described in great detail, their functional arrangement during membrane
invagination is still poorly understood. Such information however is critical for understanding the
process of vesicle formation and its regulation.
Analysis of the protein architecture of the endocytic site is very complicated task given that ~40‐60
proteins (with ~10‐200 copies of each) localize in a diffraction‐limited spot for a limited time (seconds to
minutes). Correlative electron microscopy and subpixel‐resolution methods provided recently some
important insights into the organization of the endocytic site. However they were only partially
successful due to their limited spatial resolution. Importantly, protein densities and copy numbers at the
endocytic site seem to be well suited for mapping their organization by Förster (or fluorescence)
resonance energy transfer (FRET). In addition, FRET, which can occur between two fluorophores
separated by less then 10 nm, was recently successfully applied to map the architecture of similarly
complex yeast kinetochore.
Here we present our results of systematic FRET‐based proximity mapping of 12 conserved endocytic
factors with important roles in endocytosis. We endogenously tagged yeast endocytic accessory proteins
Ede1 and Syp1 (Eps15 and FCHO1/2 in human); clathrin adaptors Yap1801/2, Ent1/2, Sla2 (CALM/AP180,
epsins, Hip1R); intersectin/CIN85 functional homologs Pan1, End3 and Sla1; and regulators of actin
polymerization Las17 and Lsb3 (WASP, SH3YL1) by GFP and mCherry on their N‐ and C‐termini. We then
assessed the proximity of various protein pairs by acceptor photobleaching, a robust FRET technique.
We obtained several highly specific and strong FRET signals between individual protein pairs, many of
them not yet been recognized by other (e.g. protein interaction) methods. Ongoing experiments are
focused on characterizing these “proximity pairs” through genetic means and on studying their potential
rearrangement during endocytosis by dynamic FRET methods. In summary, our results show that FRET‐
based proximity mapping is a highly valuable tool for providing important data about the protein
architecture of the endocytic machinery.
Mechanisms of endocytosis: shape and size dependence.
M.S. Magon1,2,3, G. Battaglia1,3; 1MRC/UCL Centre for Medical Molecular Virology, University College
London, London, United Kingdom, 2BBSRC London Interdisciplinary Biosciences DTP, University College
London, London, United Kingdom, 3Department of Chemistry, University College London, London,
United Kingdom
Viruses and intracellular bacteria evolved to invade cells by hijacking a variety of cellular mechanisms
(Cossart & Helenius 2014). The variability of their shape and size hints at the importance of these
physical properties in endocytic processes. Using nanotechnological surrogates in place of systems from
nature, with an advantage in fine control of their physical properties, we aim to investigate the basic
biophysical mechanisms of endocytosis, in particular the underlying role of size and shape of cell‐
entering particles.
Polymersomes, the virus‐mimicking nanoscopic vesicles, made of pH‐sensitive amphiphilic block
copolymers, have been developed in our lab for various biomedical applications (Canton et al.2013,
Colley et al.2014, Lomas et al.2007, Robertson et al. 2014, Pegoraro et al.2013, Akinc & Battaglia 2013,
Canton & Battaglia 2012). We showed that polymersomes interact with cells in a similar manner to
biological particles by recognising Scavenger Receptor Class B receptors on the cell surface, entering the
cells, and penetrating into the cytoplasm (Canton et al.2013, Colley et al.2014).
In these studies we focus on the biophysics of endocytic processes. We present fast time‐lapse 3D
imaging techniques and analysis methods, based on laser scanning confocal microscopy, that enable
monitoring and quantitation of dynamics of nanoparticle interactions at the plasma membrane and
endosome interfaces in live cells. The results for spherical, tubular and high genus‐shaped nanoparticles,
of the same chemistry and homogenous topology, demonstrate the shape‐dependent control of plasma
membrane binding and cellular entry kinetics. The comparative study with small (20‐30 nm diameter)
spherical micelles and larger spherical vesicles (100‐200 nm diameter) indicate the size dependence.
Going a step further, we also explored the particle trafficking pathway inside the cell. While the
spherical particles entered the first acidic compartment, the early endosome, the tubes and genus did
not, as demonstrated in the time‐lapse colocalisation analysis. The fast dynamics of the events at the
early endosome interface for spherical vesicles emphasised that the synthetic particles behave like
natural viral particles by entering the early endosome, rearranging/disassembling themselves at the
acidic pH and penetrating into the cytoplasm.
The studies highlight the importance of the basic physical interactions with cellular membranes in the
dynamics of endocytic processes and the intracellular trafficking routes. We aim to examine further the
effect of shape and size of particles on the molecular mechanisms of endocytosis. To shed more light on
the biophysics of cell‐particle interactions we also aim to investigate the role of surface topology of
particles in endocytosis.
Differential control of death‐receptor endocytosis and apoptosis by dynamins.
C.R. Reis1, S.L. Schmid1; 1Cell Biology, UT Southwestern Medical Center, Dallas, TX
Introduction: The remarkable tumor‐specific apoptosis inducing properties of the TNF‐related apoptosis
inducing ligand (TRAIL) result in death‐receptor mediated activation of caspases in cancer cells and have
raised considerable interest in its use as an anti‐cancer therapeutic. However, many cancer cells still
display resistance to TRAIL‐induced apoptosis. Although it is well established that TRAIL‐death receptors
undergo ligand‐induced clathrin‐mediated endocytosis (CME), the biological significance of this event for
apoptosis remains unresolved. Recently we have uncovered new isoform specific roles for dynamins in
regulating endocytosis, through a feed‐forward pathway that links signaling from surface receptors to
the regulation of CME (Reis CR et al, EMBO J. in press).
Methods and Results: Using siRNA‐mediated knockdown, CRISPR‐Cas9n‐mediated knockout of different
dynamin isoforms and reconstitution studies, we set out to investigate the role of dynamin isoforms in
the regulation of death receptor trafficking and signaling mediated via TRAIL. Our data establishes that
dynamin isoforms differentially regulate ligand‐induced death receptor endocytosis and apoptosis.
Whereas transferrin receptor (TfnR) endocytosis via CME depends primarily on dynamin‐2, we find that
TRAIL‐induced death‐receptor endocytosis requires dynamin‐1 (previously thought to be neuronal‐
specific), but not dynamin‐2. Dynamin‐1‐dependent CME prevents caspase activation via these
receptors in both TRAIL sensitive and resistant cancer cells. Conversely, knockdown or knockout of
dynamin‐1 sensitizes cells to TRAIL‐induced apoptosis. Our results further suggest that this specificity
may be related to a calcium sensing mechanism, involving the activation of the phosphatase calcineurin,
and its regulatory role in inducing endocytosis through dynamin‐1 activation.
Conclusion: These findings suggest the very intriguing possibility that distinct dynamin isoforms can
differentially regulate the endocytosis of specific cargo receptors through CME, and establish new
functional links between the large GTPase dynamin, calcium signaling, and the execution of
programmed cell death. Insight into the molecular mechanisms underlying cell fate decisions in normal
vs transformed cells, will not only enhance our knowledge of the relationship between endocytic
trafficking and signaling, but will also aid in the development of new combination strategies, aiming at
specifically killing cancer cells in a highly effective manner.
Polarized endosome dynamics by spindle asymmetry during asymmetric cell division.
E. Derivery1, C. Seum1, A. Daeden1, S. Loubéry1, L. Holtzer1, F. Julicher2, M. Gonzalez‐Gaitan1;
Biochemistry, University of Geneva, Geneva, Switzerland, 2Max Planck Institute for the Physics of
Complex Systems, Dresden, Germany
During asymmetric division, fate determinants at the cell cortex segregate unequally into the two
daughter cells. It has recently been shown that Sara signaling endosomes in the cytoplasm also
segregate asymmetrically during asymmetric division. Biased dispatch of Sara endosomes mediates
asymmetric Notch/Delta signaling during the asymmetric division of sensory organ precursors in
Drosophila. In flies, this has been generalized to stem cells in the gut and the central nervous system
and, in zebrafish, to neural precursors of the spinal cord. However, the mechanism of asymmetric
endosome segregation is not known. Here we unravelled this mechanism. The plus‐end kinesin motor
Klp98A targets Sara endosomes to the central spindle. At the central spindle, endosomes move
bidirectionally on an antiparallel array of microtubules. The microtubule depolymerising kinesin Klp10A
and its antagonist Patronin generate central spindle asymmetry. The asymmetric spindle, in turn,
polarizes endosome motility, ultimately causing asymmetric endosome dispatch into one daughter cell.
To demonstrate this, we inverted the polarity of the spindle by polar targeting of Patronin with
nanobodies. Spindle inversion targets the endosomes to the wrong cell. Our data uncovers the
molecular and physical mechanism by which organelles localized away from the cellular cortex can be
dispatched asymmetrically during asymmetric division.
LMA‐1 maintains lysosome integrity and normal adult life span in C. elegans.
Y. Li1, W. Zou1, B. Chen1, X. Wang2, Y. Sun1, C. Yang2, X. Wang1; 1National Institute of Biological Sciences,
Beijing, China, 2State Key laboratory of Molecular Developmental Biology, Institute of Genetics and
Developmental Biology, Chinese Academy of Science, Beijing, China
Lysosomes are specialized membrane bound organelles that degrade macromolecular cargoes derived
from endocytosis, phagocytosis or autophagy. The resulting catabolites are exported by membrane
transporters on lysosomes for re‐utilization in cellular metabolism. Over 50 human disorders have been
identified due to dysfunctions of lysosomes, indicating important roles of lysosomes in cell homeostasis
and animal development. In C. elegans, lysosomes appear as both small puncta and thin tubules,
displaying highly dynamic patterns during development, in aging process and under different culture
conditions. qx193 is a recessive mutation that causes abnormally enlarged globular lysosomes and lack
of the tubular ones. The gene affected in qx193 is named lma‐1 as lysosomal morphology abnormal 1.
We found that lysosomes, which are quite dynamic in wild type, become static in lma‐1(qx193) worms
and are defective in degrading cargoes derived from both endocytosis and autophagy. Thus LMA‐1 is
important for maintaining lysosome morphology, motility and function. Our further cell biological and
genetic analyses revealed that loss of its function causes accumulation of damaged lysosomes, which
can be cleared by autophagy. qx193 mutants display a delayed larval development and significantly
shortened life span, suggesting that lysosomal integrity maintained by LMA‐1 is important for normal
larval development and adult life span. Identification and characterization of LMA‐1 will help us to
understand how lysosome integrity is maintained and how this contributes to cell homeostasis and
animal development.
Mechanisms of mTORC1‐independent lysosome biogenesis.
Y. Li1, M. Xu1, X. Ding2, Z. Song3, X. Huang1, Y. Jian1, G. Tang2, C. Yang3, Y. Di1, X. Liu1, K. Liu1, T. LI1, Y.
Wang1, X. Hao2,3, C. Yang1; 1Center for Developmental Biology, Institute of Genetics and Developmental
Biology, Chinese Academy of Sciences, Beijing, China, 2Kunming Institute of Botany, Chinese Academy of
Sciences, Kunming, China, 3The Key Laboratory of Chemistry for Natural Product of Guizhou Province
and CAS, Guiyang, China
Lysosomes serve as the major degradative sites and hubs of metabolism signaling within the cell.
Biogenesis of lysosomes is transcriptionally activated by TFEB and TFE3 transcription factors but
suppressed by the ZKSCAN3 transcription repressor. While TFEB and TFE3 is known to be regulated by
mTORC1‐dependent phosphorylation in response to nutrient supply, mechanisms other than nutrient
signaling that control this important cellular event remain to be elucidated. By screening lysosome‐
inducing natural small‐molecule compounds, here we show that protein kinase C (PKC) is a master
switch of lysosome biogenesis by coupling TFEB activation with ZKSCAN3 inactivation through two
parallel signaling cascades. PKC activation resulted in inactivation of GSK3β, leading to reduced
phosphorylation, nuclear translocation and activation of TFEB, but not TFE3. On the other hand, PKC
activation of JNK and p38 MAPK phosphorylates ZKSCAN3, leading to its inactivation by translocation out
of the nucleus. These findings reveal mechanisms of lysosome biogenesis in response to a wide range of
environmental or physiological cues mediated by PKC without compromising mTORC1, suggesting that
PKC activators may be viable treatment options for lysosome‐related disorders.
Real time analysis of Clathrin and Dynamin‐independent endocytic vesicle formation: Role of
actin nucleators and BAR domain proteins.
M. Sathe1, G. Muthukrishnan1, M. Thattai1, S. Mayor1; 1National Center for Biological Sciences,
Bangalore, India
Eukaryotic cells employ multiple endocytic routes to internalize cell surface receptors, nutrients,
membrane proteins/lipids and extracellular fluid. Clathrin dependent endocytic pathways are studied
extensively in the context of receptor mediated uptake. Clathrin‐ independent endocytic pathways are
emerging as critical players in many cellular processes. The Clathrin Independent Carriers/GPI‐AP
Enriched Endosomal Compartments (CLIC/GEEC; CG) internalizes a major fraction of the fluid‐phase in
many (but not all) cell types, and is responsible for the rapid turnover of the plasma membrane in
fibroblasts. CG pathway is sensitive to perturbation of actin polymerization and cholesterol levels at the
plasma membrane1 . Although CG pathway is regulated by small GTPases such as CDC42 and Arf1, in the
absence of any discernible coats on the CLICs or known scission factors, the structural components that
assist in CG vesicle generation are yet to be understood. To find molecules which could assist in CG
endosome formation, we focused our attention on BAR domain proteins. This family of proteins
contains a curvature sensing or stabilizing domain along with several accessory domains, which regulate
its localization via interaction with specific lipids, influence actin nucleators, and may recruit other
proteins to the site of vesicle generation2 . We screened all the Drosophila BAR domain proteins for its
role in CG pathway and have found a few potential candidates. We are exploring the role of these
proteins (for example, PICK1; protein interacting with PRKCA 1 which is both a negative regulator of
ARP2/3 mediated actin polymerization and an Arf1 effector). Real time imaging of the formation of CG
endosomes via a pH‐pulse assay3 , shows that there is an orchestrated sequence in which GBF1, Arf1,
Cdc42 and specific actin nucleators are recruited to the forming endocytic sites. Predictably, neither
clathrin nor dynamin are recruited to these sites. Using this assay we show that a medley of BAR
domain‐containing proteins are recruited to the forming endosome in a specific sequence, suggesting a
complex mechanism for the formation of this widely used pathway.
Mayor, S. & Pagano, R. E. Pathways of clathrin‐independent endocytosis. Nat. Rev. Mol. Cell Biol. 8, 603–
12 (2007).
Frost, A., Unger, V. M. & De Camilli, P. The BAR domain superfamily: membrane‐molding
macromolecules. Cell 137, 191–6 (2009).
Merrifield, C. J., Perrais, D. & Zenisek, D. Coupling between clathrin‐coated‐pit invagination, cortactin
recruitment, and membrane scission observed in live cells. Cell 121, 593–606 (2005).
Mdm1/Snx13 is a novel inter‐organelle membrane contact site tethering protein.
M. Henne1, L. Zhu2, Z. Balogi2, C. Stefan2, J. Pleiss2, S. Emr2; 1Cell Biology, UT Southwestern, Dallas, TX,
Weill Institute, Cornell University, Ithaca, NY
Although endolysosomal trafficking is well defined, how it is regulated and coordinates with cellular
metabolism is unclear. In order to identify genes governing endolysosomal dynamics, we conducted a
global fluorescence‐based screen in yeast to reveal endomembrane effector genes. Screening implicated
Phox (PX) domain‐containing protein Mdm1 in endomembrane dynamics. Surprisingly, we demonstrate
that Mdm1 is a novel inter‐organelle tethering protein that localizes to ER‐vacuole/lysosome Membrane
Contact Sites (MCSs). We show that Mdm1 is ER‐anchored and contacts the vacuole surface in trans via
its lipid‐binding PX domain. Strikingly, over‐expression of Mdm1 induced ER‐vacuole hyper‐tethering,
underscoring its role as an inter‐organelle tether. We also show that Mdm1 and its paralog Ydr179w‐a
(named Nvj3 here) localize to ER‐vacuole MCSs independently of established tether Nvj1. Finally, we find
that Mdm1 truncations analogous to neurological disease‐associated SNX14 alleles fail to tether the ER
and vacuole and perturb sphingolipid metabolism. Our work suggests Mdm1 homologs, which are highly
conserved in humans, may play previously unappreciated roles in inter‐organelle communication, lipid
metabolism, and neurological disease.
Lysophagy: Novel selective autophagy eliminating damanged oraganelles and suppressing
M. Hamasaki1, H. Teranishi1, I. Maejima1, T. Yoshimori1; 1Department of Genetics Graduate School of
Medicine, Osaka University, Osaka, Japan
Autophagy is a tightly regulated intracellular degradation system that plays fundamental roles in cellular
homeostasis. Autophagy is induced by starvation and also by elimination of bacteria, accumulation of
aggregate proteins, damaged organelles and so on. We recently identified that damaged lysosomes are
also eliminated by autophagy and named as lysophagy. We used the lysosomotropic agents called
LLoMe digested in lysosomes to become membranolytic form thus damages the lysosome membrane.
These damaged lysosomes are ubiquitinated first then autophagic machinery is recruited, which are
then engulfed by autophagosomes. Loss of lysophagy lead to hyperuricemic nepharopathy. To identify
what is recognizing the damage of organelles, we performed proteomics assay on damaged membrane.
Here we identified E3 ligase that is putatively involving at the initial stage of lysophagy.
Minisymposium 16: Morphogenesis
Surface cell expansion triggers radial cell intercalations in zebrafish gastrulation.
C. Heisenberg1; 1‐, IST Austria, Klosterneuburg, Austria
Radial cell intercalations are commonly associated with tissue spreading in many developmental and
disease‐related processes, such as vertebrate gastrulation and tumor metastasis. Yet, how radial cell
intercalations are controlled and function in tissue spreading remains unknown. Here, we use a
combination of experiments and theory to analyze radial cell intercalations during doming, the initial
spreading of the blastoderm over the yolk cell at early zebrafish gastrulation. Strikingly, we found that
radial cell intercalations do not drive doming, but rather determine the resistance of the blastoderm to
deformation by tensile stress at the surface of the tissue. We further show that the expansion of surface
epithelial cells reduces tissue surface tension (TST), which in turn triggers radial cell intercalations and
doming. Thus, radial cell intercalations are required for translating changes in TST into tissue
Local and tissue scale forces drive oriented junction growth during tissue extension.
C. Collinet1, M. Rauzi1, P. Lenne1, T. Lecuit1; 1CNRS, IBDM Institute de Biologie du Développement de
Marseille, Marseille, France
Convergence‐extension is a widespread morphogenetic process driven by polarized cell intercalation. In
the Drosophila germband, epithelial intercalation comprises loss of junctions between antero‐posterior
neighbors followed by growth of new junctions between dorsal‐ventral neighbors. Much is known about
how active stresses drive polarized junction shrinkage. However, it is unclear how tissue convergence‐
extension emerges from local junction remodeling and what is the specific role, if any, of junction
growth. Here we report that tissue convergence and extension correlate mostly with new junction
growth. Simulations and in vivo mechanical perturbations reveal that junction growth is due to local
polarized stresses driven by medial actomyosin contractions. Moreover, we find that tissue‐scale pulling
forces at the boundary with the invaginating posterior midgut actively participate in tissue extension by
orienting junction growth. Thus, tissue extension is akin to a polarized fluid flow that requires parallel
and concerted local and tissue‐scale forces to drive junction growth and cell‐cell displacement.
Compliance sensing by actomyosin self‐organization coordinates epithelial tension and tissue
S. Chanet1, C. Miller2, C. Vasquez1, B. Ermentrout3, L.A. Davidson2, A. Martin1; 1Biology, MIT, Cambridge,
MA, 2Bioengineering, Pittsburgh University, Pittsburgh, PA, 3Mathematics, Pittsburgh University,
Pittsburgh, PA
Generating proper organ and organism shape requires that cells generate and orient forces, resulting in
patterns of tissue tension critical to bend and fold epithelial tissues. How individual cells embedded
within an epithelium coordinate the direction and magnitude of force to elicit proper tissue form is
unknown. During Drosophila gastrulation, actomyosin contraction in ventral cells generates a long,
narrow epithelial furrow, termed the ventral furrow, in which tension is directed along the length of the
furrow. Here we demonstrate that cellular force generation is coordinated by the compliance of the
surrounding tissue with tension directed along the axis of least compliance. We show that actomyosin
meshworks adopt different morphologies and generate force with different direction and magnitude
depending on external constraints to contraction. Our demonstration of the mechanosensitive nature of
dynamic actomyosin meshworks establishes a role for mechanical patterning in tissues. Systems‐level
physical properties derived from embryo/organ shape and resulting constraints to deformation
influence cellular force generation via a mechanical feedback loop centered on actomyosin meshwork
self‐organization. Because dynamic actomyosin meshworks drive diverse morphogenetic events, there
are likely many examples of mechanical patterning where force generation is specified by mechanical
Pressure in developmental size control.
I.A. Swinburne1, K.R. Mosaliganti1, A. Green1, T. Hiscock1, L. Mahadevan2,3, S.G. Megason1; 1Department
of Systems Biology, Harvard University, Boston, MA, 2Department of Physics, Harvard University,
Cambridge, MA, 3School of Engineering and Applied Science, Harvard University, Cambridge, MA
Animals develop tissues of precise size, shape and symmetry despite noise in underlying molecular and
cellular processes. How tissue and organ‐level feedback regulates this noise is largely unknown. Here,
we combine quantitative imaging, physical theory and perturbations in zebrafish to study size control of
the developing inner ear. We find that transepithelial fluid flow creates hydrostatic pressure in the
lumen leading to stress in the otic epithelium and expansion of the otic vesicle. Pressure, in turn, inhibits
endolymph transport into the lumen. This negative feedback loop between pressure and fluid flux
allows the otic vesicle to change growth rate in order to regulate natural or experimentally induced size
variation. Furthermore, the shape of the inner ear is modulated by spatial‐temporal patterning of
actomyosin contractility allowing a common lumenal pressure to drive varying local epithelial strain
rates. Once formed it is essential that the inner ear maintains a homeostatic pressure for its proper
function. We discovered that pressure is maintained by a cell‐based pressure relief valve made of novel
epithelial junctions we term "basal lamellar junctions" that break open under pressure. Our work
uncovers how molecular driven mechanisms such as osmotic force generation and actomyosin tension
can regulate tissue level morphogenesis via hydraulic feedback to ensure robust control of organ size.
We also find a key role for pressure in size control of the neural tube, but through a very different
mechanism. The pressure of cells dividing during mitosis regulates the propensity of neighboring cells to
differentiate. This mechanical feedback implements derivate control which we show yields robust size
control during growth.
Stochastic fluctuations in oxidative stress signaling induce anisotropies in adhesion and
cytoskeletal organization to influence cell behavior and spatial patterning in a Drosophila
M. Narasimha1, S. Muliyil1,2, S. Saravanan1; 1Department of Biological Sciences, Tata Institute of
Fundamental Research, Mumbai, India, 2Sir William Dunn School of Pathology, University of Oxford,
Oxford, United Kingdom
The morphogenesis and maintenance of tissues relies on dynamic and heterogeneous cell behaviors.
The origin of these heterogeneities and their coordination remain poorly understood. We investigate
how heterogeneities in cell behavior (pulsed or unpulsed cell constriction and cell delamination) are
patterned using the amnioserosa, an active participant during Drosophila dorsal closure as our model
and seek to understand the cellular, molecular and physical bases of individual behaviours (both
stochastic and collective) and their coordination that ensures the stereotypical dynamics of this tissue.
Using targeted (single cell, patchy or whole tissue) genetic and nanoscale laser perturbations, 4D live
confocal microscopy and quantitative morphometric analysis, we show that differences in cell behavior
result from differences in the spatial organization of cell‐cell adhesion and the actomyosin and
microtubule cytoskeleton (1 and unpulished observations). These anisotropies are in turn generated by
stochastic fluctuations in mitochondrial ROS that we find act both autonomously and non‐
autonomously. We have identified a pathway operating downstream of ROS signalling (Rho‐ROCK‐MLC)
that generates asymmetries in actomyosin organization through its localization to distinct (cortical or
medial) actomyosin containing structures. We are currently investigating the basis of its effects on
anisotropies in microtubule cytoskeleton organisation and in cell‐cell and cell‐substrate adhesion and
also the origin of the stochasticities (2 and unpublished observations). Our findings are beginning to
provide insights into the local control of cell behavior and their influence on the spatial patterning of
tissues during morphogenesis. They also provide an explanation for compromised tissue integrity in
metabolic and oncological pathologies and hint at the intricate interplay between mechanical, metabolic
and chemical cues in multicellular sensing.
Saravanan S, Meghana C, Narasimha M (2013) Local, cell non‐autonomous feedback regulation of
myosin dynamics patterns transitions in cell behaviour; a role for tension and geometry? Molecular
Biology of the Cell 24: 2350.
Muliyil S, Narasimha M* (2014) Mitochondrial ROS Regulates Cytoskeletal and Mitochondrial
Remodeling to Tune Cell and Tissue Dynamics in a Model for Wound Healing. Developmental Cell 28:
aMOTIV microscopy: mechanical characterization of the in vitro and in vivo tissue
J.R. Staunton1, B.H. Blehm1, A. Devine1, K. Tanner1; 1National Cancer Institute, National Institutes of
Health, Bethesda, MD
Microscale heterogeneities in tissue properties such as stiffness and viscosity strongly influence cell fate
and malignancy. However, outstanding questions about the timescales of interactions (measured as a
range of frequencies), length scales and type of interactions sensed by cells within tissues that are
physiologically relevant remain unanswered. What is needed is the ability to resolve and quantitate
minute forces that cells sense in the local environment (on the order of microns) within thick tissue
(~mm) and 3D culture models, that approximate clinically relevant in vivo architecture and signaling
cues, allowing for real time characterization of cell‐ECM dynamics.
We performed active Microrheology using an Optical Trap In Vivo (aMOTIV) microscopy using an in situ
calibration method to obtain exact trap stiffness at each probe to quantify local applied forces with high
spatial and temporal resolutions. This allowed us to determine tissue mechanics at length scales (nm‐
μm) and frequencies (1‐10,000’s Hz) unobtainable by bulk rheology, which misses the cell‐scale
heterogeneities, or with passive microrheology, which misses the interesting non‐linear stress/strain
curves seen with active probing then apply defined strains, Applicable to thick tissue, this technique
allowed us to distinguish mechanical heterogeneities with micrometer spatial resolution at penetration
depths up to 500 um.
Initial measurements in Matrigel and Hyaluronic Acid (HA) indicate they are both very soft, similar to
those observed for bulk rheology of mammalian tissues. However, our microrheological technique
measures significantly different rheological characteristics at low stresses and strains (probed by
reducing oscillation amplitude or trap stiffness). We also demonstrate significant improvement over
prior microrheological techniques, displaying 2 to 10‐fold improvements by calibrating in situ. After
initial characterization of 3D cell culture gels, we applied our technique in zebrafish, Danio rerio; the first
time in situ calibration and microrheology has been applied to a living model vertebrate organism. Our
initial data indicates a broad range of elastic moduli, with measurements in the tail ranging ~ 10s to
1000s of Pascals, while the brain is significantly softer (10’s to 100’s Pa).
We show here an advanced technique for micro‐rheological characterization in vitro and in vivo, to
accurately quantify physical determinants of the local microenvironment. This allows for accurate
measurements of mechanical properties in living organisms and in tissue models to definitively account
for microenvironmental impact on individual cells and organogenesis.
Actomyosin force generation directs hydra regeneration.
K. Keren1, A. Livshitz1, L. Zerbib1, E. Braun1; 1Physics, Technion‐ Israel Institute of Technology, Haifa,
Hydra is a multicellular fresh‐water polyp with uniaxial symmetry that exhibits remarkable regeneration
properties. Small tissue fragments excised from mature hydra can regenerate into functional animals
within 2‐3 days, and thus form an excellent model system for studying animal morphogenesis. We
follow the actin dynamics during the regeneration process using transgenic hydra strains expressing
lifeact‐GFP. We find that the polar supra‐cellular actin organization in freshly excised tissue fragments,
which originates from the donor hydra, persists throughout the regeneration process, and defines the
body axis in the regenerating hydra. Moreover, defects in the regeneration process that lead to the
formation of hydra with multiple body axes (i.e. more than one head and/or foot) can be traced to the
lack of global alignment of the actin cytoskeleton at the early stages of regeneration. The role of the
actomyosin cytoskeleton in hydra regeneration is also examined using biochemical inhibitors. We find
that inhibition of either actin dynamics or myosin activity prevents morphogenesis. In particular, even
partial inhibition of actomyosin force generation with low doses (~1µM) of blebbistatin is sufficient for
blocking regeneration, with treated tissue fragments remaining spherical for days and eventually
disintegrating. To further examine the influence of mechanical forces on morphogenesis, we subject
regenerating hydra to external forces. Remarkably, we find that we can redirect the body axis of
regenerating hydra in the direction of the applied force. Overall, our results demonstrate the
importance of mechanical forces during the early stages of axis formation in hydra, highlighting the
intimate coupling between mechanics and biochemistry in morphogenesis.
Conserved roles for cytoskeletal components in determining laterality.
G.S. McDowell1,2, J. Lemire1,2, J. Pare1,2, M. Levin1,2; 1Center for Regenerative and Developmental Biology,
Tufts University, Medford, MA, 2Biology Department, Tufts University, Medford, MA
The development of the left‐right axis in bilateral organisms is required for the proper placement of
internal organs, and is a paradigm case of how the functional biology of individual cells sets the large‐
scale structure of an entire organism. In stark contrast to the other two body axes, there is considerable
controversy about the earliest mechanisms of consistent embryonic laterality. Left‐right symmetry‐
breaking is commonly attributed to extracellular cilia‐dependent fluid flow in the neurula‐stage embryo.
However, there are many exceptions and caveats to this model. Importantly, numerous organisms,
including even amniotes such as chick and pig, do not have ciliated structures capable of generating
chiral flow but nevertheless develop a consistently asymmetric bodyplan using the same downstream
asymmetric gene cascade. In contrast to the wide evolutionary divergence of early mechanisms that is
required by a ciliary model, we have previously proposed that asymmetry is dictated by an intracellular,
well‐conserved mechanism. To resolve the question of conservation of the early steps of LR patterning,
we used the Xenopus laevis (frog) embryo to functionally test a number of protein targets known to
direct asymmetry in plants, fruit fly, and rodent. Using the same reagents that randomize asymmetry in
Arabidopsis, Drosophila, and mouse embryos, we show that manipulation of the microtubule and actin
cytoskeleton immediately post‐fertilization, but not later, results in laterality defects in the frog embryo.
Moreover, our data reveal novel aspects of the linkage between biophysical cues and downstream
asymmetric gene expression cascades, including a remarkable “repair” process that corrects aberrant
gene expression to result in normal organ situs. Our data reveal a well‐conserved role for the
cytoskeleton across Kingdoms of life in regulating left‐right axis formation immediately after fertilization
of the egg.
Development plain and simple: using cartography to analyze forces driving morphogenesis.
S.J. Streichan1, I. Heemskerk1, M.F. Lefebvre2, E. Wieschaus2, B. Shraiman1; 1Kavli Institute for Theoretical
Physics, University of California Santa Barbara, Santa Barbara, CA, 2Molecular Biology, Princeton
University , Princeton, NJ
During embryogenesis, cells and tissues are transformed into complex yet highly reproducible shapes.
Genetics successfully revealed the existence of a direct relation between morphogens expressed in the
embryo and its final form. Morphogens setup a coordinate system and concert tissue flows on the
embryo to conduct morphogenesis.These flows are subject to non‐trivial constraints by topology, and
the interplay between patterned signaling molecules with physical forces required for such movements
remains elusive. Here we study genetically instructed tissue flows using D. melanogaster as a model
system.We deploy multi view light sheet microscopy to generate high resolution in toto time‐lapse
recordings of embryos beginning with gastrulation until completion of larva formation while monitoring
gene expression. To facilitate data analysis and extract tissue flows, we designed the Image Surface
Analysis Environment (ImSAnE), a cartography toolbox tailored towards analysis of layered and curved
bio‐image data. Combined with a model driven approach we infer stress patterns that generate the
observed tissue flows and identify error correction mechanisms that together enable highly
reproducible collective cell movements in embryos.
Minisymposium 17: New Technologies and their Application to Probe the Spatial
Organization of the Cell
Mapping the subcellular proteome and determining dynamic subcellular rearrangements using
quantitative mass spectrometry.
K.S. Lilley1, A. Christoforou1, C.M. Mulvey1, L.M. Breckels1, L. Gatto1, A. Geladaki1, T. Hurrell1, D.J.
Nightingale1, H. Zhou1, A. Martinez‐Arias2; 1Department of Biochemistry, University of Cambridge,
Cambridge, United Kingdom, 2Department of Genetics, University of Cambridge, Cambridge, United
Intracellular proteins exist in controlled micro‐environments, such as organelles, sub‐organellar
compartments, clusters of membrane proteins and multi‐protein complexes, where they carry out
different roles dependent on their local environment. To gain a complete functional analysis of the
proteome, it is vital to determine the possible sub‐cellular niches in which a protein may reside.
Moreover, the ability to map changes in location in response to drug treatment, developmental stage
and disease progression is of paramount importance to the elucidation of cellular mechanisms.
We have created hyperLOPIT, which couples quantitative mass spectrometry methods and advanced
machine‐learning tools. This method enables the simultaneous assignement the steady‐state location of
thousands of proteins to multiple subcellular compartments to create a high resolution map of a cell.
HyperLOPIT maps have revealed sub‐organellar detail and the steady state location of hundreds of
protein complexes. Moreover the method can be used in a comparative manner, where the intracellular
maps of cells in two or more conditions can be assesed. HyperLOPIT workflows can be used to
determine the effect of post transcriptional and post translational modification upon localization.
HyperLOPIT allows interrogation of the dynamic subcellular proteome on an unprecedented scale, and is
complementary to immunocytochemistry, and other mass spectrometry based methods including
proximity labelling and affinity purification of protein complexes.
Discovery and characterization of novel synaptic and mitochondrial proteins via peroxidase‐
mediated live cell proteomic mapping.
A.Y. Ting1; 1Chemistry, Massachusetts Institute of Technology, Cambridge, MA
Membrane‐membrane contact sites, including the synaptic cleft and mitochondria‐endoplasmic
reticulum (ER) contact sites, are of central importance in biology, but they are difficult to characterize
and study. One reason is that these structures cannot be purified and isolated for mass spectrometric
analysis to comprehensively determine their proteomic content. We have developed a methodology
that bypasses the need for organelle purification but can obtain highly specific maps of defined cellular
regions. The method relies on a genetically targeted peroxidase enzyme, called APEX2, that can
promiscuously biotinylate endogenous proteins within a few nanometers of it, in living cells, over a 1
minute time window. This approach was used to obtain highly specific proteomic maps of excitatory and
inhibitory synaptic clefts, as well as mitochondria‐ER junctions in mammalian cells. We will describe
these proteomic inventories as well as the novel proteins and functions we have discovered using these
Spectrally resolved super‐resolution microscopy.
Z. Zhang1,2, S.J. Kenny1, M. Hauser1, W. Li1, K. Xu1,2; 1Department of Chemistry, University of California,
Berkeley, CA, 2Life Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA
Emerging super‐resolution microscopy methods offer outstanding spatial resolution but no spectral
information. As a result, high‐quality multicolor 3D super‐resolution microscopy remains a challenge,
and issues like color crosstalk, compromised image quality, and difficulties in aligning the images
obtained by different color channels abound. We have developed a new method, which we named
Spectrally Resolved ‐ Stochastic Optical Reconstruction Microscopy (SR‐STORM), to synchronously
obtain the fluorescence spectra and positions of millions of single molecules in densely labeled cell
samples in minutes, and hence the concept and realization of spectrally resolved, “true‐color” super‐
resolution microscopy. Remarkably, we found that contrary to previous results on dye molecules
immobilized at solid surfaces, in labeled cells single molecules of most dyes exhibited highly uniform
emission spectra. This allowed us to unambiguously identify single molecules of different dyes that
overlapped heavily in spectrum. Crosstalk‐free 3D super‐resolution microscopy was thus achieved for
four dyes that were only 10 nm apart in emission spectrum; excellent resolution was obtained for every
channel, and the 3D localizations of all molecules were automatically aligned within one imaging path.
The application of SR‐STORM to live cells, in combination with the design of STORM‐compatible
fluorophores that are spectrally responsive to local environments, represents exciting future challenges.
This work is just accepted by Nature Methods.
Exploring the native molecular architecture of organelles with in situ cryo‐electron tomography.
B.D. Engel1, M. Schaffer1, S. Albert1, S. Asano1, L. Kuhn Cuellar1, S. Pfeffer1, J.M. Plitzko1, W. Baumeister1;
Molecular Structural Biology, Max Planck Institute of Biochemistry, Martinsried, Germany
We are leveraging new advances in cryo‐electron tomography (cryo‐ET) to investigate how
macromolecular complexes establish cellular architecture. Thin slices of vitrified Chlamydomonas
reinhardtii cells were prepared by cryo‐focused ion beam (cryo‐FIB) milling and then imaged by state‐of‐
the‐art cryo‐ET. The resulting 3D views of the native cellular environment have provided new insights
into the molecular organization of organelles, including the nucleus, Golgi, ER, mitochondrion, and
chloroplast. These insights include the de novo identification of: 1) protein arrays that establish the
narrow membrane spacing of the trans‐Golgi cisternae, 2) linker proteins that bind RuBisCO complexes
together within the chloroplast’s pyrenoid, and 3) fine membrane tubules that likely serve as conduits
for the directed diffusion of metabolites between the pyrenoid and the chloroplast stroma. Our
tomograms have also enabled a structural comparison of the translocon machinery associated with
membrane‐bound ribosomes on the ER and nuclear envelope. In addition to the in situ characterization
of individual macromolecular complexes, in the future we will aim for a visual proteomics approach by
applying machine learning algorithms to identify and classify every macromolecule within the cellular
The Balbiani Body in Xenopus Forms by Amyloid‐like Aggregation.
E. Boke1, R. Lemaitre2, S. Alberti2, A.A. Hyman2, D.N. Drechsel2, T.J. Mitchison1; 1Department of Systems
Biology, Harvard Medical School, Boston, MA, 2Max Planck Institute for Molecular Cell Biology and
Genetics, Dresden, Germany
There are two basic types of cells in animals; somatic cells and germ cells. Germ cells generate the
sperm and eggs that are necessary for propagation of the next generation. A key feature of germ cells is
that the complement of mitochondria and RNA is kept intact for decades before fertilization. However,
we have little idea how these components are protected for such long periods of time. A common
feature of germ line specification in most vertebrate oocytes (including mammals) is the Balbiani body,
which is a non‐membrane bound compartment packed full of mitochondria, along with RNA and other
organelles. Understanding the structure and function of the Balbiani body will be crucial for
understanding how mitochondria and RNA are preserved in oocytes for a long time.
We have studied the organization of the Balbiani body in Xenopus, the large size of which makes it
amenable to biochemistry. We show by quantitative mass spectrometry that the most enriched protein
in Balbiani body of Xenopus is Xvelo, which is a homolog of the germ line protein bucky ball in zebrafish.
Xvelo has a prion like domain in its N terminus. Xvelo protein forms micron‐sized networks in vitro over
time and that the network dynamics can be tuned with RNA. Xvelo networks are SDS‐resistant and stain
positive with the dyes that recognize an amyloid state. Disruption of the prion‐like domain of Xvelo
changes its association dynamics with the Balbiani body in vivo, and prompts the disassembly of Balbiani
body with its mitochondrial content. We propose that the Balbiani body forms by an amyloid‐like
aggregation mechanism. Because the prion‐like domain of Xvelo is conserved in different germ plasm‐
related proteins in other species, Balbiani body formation by amyloid like aggregation could be a
conserved feature in evolution for maintaining the immortal character of germ cells.
The nucleolus as an active multiphase droplet.
M. Feric1, N. Vaidya1, T.M. Richardson1, C.P. Brangwynne1; 1Department of Chemical Biological
Engineering, Princeton University, Princeton, NJ
Increasing evidence supports the concept that membrane‐less RNA/protein (RNP) bodies assemble
through a type of liquid‐liquid phase separation. However, RNP bodies are highly complex, consisting of
hundreds of different protein and RNA molecules, many of which exhibit ATP‐dependent activity. The
nucleolus is one particularly interesting example, which despite exhibiting liquid‐like properties,
nevertheless has distinct substructures that challenge the simple phase separation concept. Here, we
use the large X. laevis oocyte to show that the different components of the nucleolus (granular, dense
fibrillar and fibrillar center) act as distinct liquid‐like subphases that are kinetically stabilized by a nuclear
actin network. Actin disruption leads to global coarsening of the nucleolar structure as a result of
homotypic fusion events between each component. Furthermore, the typically fast, liquid‐like recovery
of each component after photobleaching is sensitive to ATP depletion, actin disruption, and biochemical
disruption with ALS‐related peptides. Consistent with the phase separation framework, increasing the
concentration of individual nucleolar proteins by microinjection leads to corresponding changes in the
volume fraction of nucleolar subphases, including formation of extranucleolar droplets. Several purified
nucleolar proteins form droplets in vitro, providing insight into the structural features that give rise to
coexisting nucleolar subphases. These results suggest that phase transitions may underlie not only
intracellular organization, but also intraorganelle organization.
Genetically encoded nanoparticles reveal mechanisms that control cellular biophysics.
K.J. Kennedy1, J. Guttierez1, I.V. Surovtsev2, C. Renou1, C. Jacobs‐Wagner2, L.J. Holt1; 1Molecular and Cell
Biology, University of California Berkeley, Berkeley, CA, 2Department of Molecular, Cellular, and
Developmental Biology, Yale University, New Haven, CT
Cells use active, motor‐directed, transport for distribution of some molecules, but the majority of
cytoplasmic molecules diffuse by passive transport. Here we present evidence that the cell regulates
passive transport by modulating the rate of diffusion. We have developed genetically encoded
microrheology nanoparticles (GEMs) to survey viscoelastic properties inside living cells. GEMs are
comprised of the Pyrococcus furiosus encapsulin protein fused to the GFP variant sapphire. 180 copies of
this protein self‐assemble into stable spherical particles 30 nm in diameter. These stereotyped
nanoparticles facilitate passive microrheology as a method to probe the biophysics of the cell. We have
used this system to discover signaling pathways that modulate the global rate of diffusion in the
cytoplasm and nucleus in both yeast and mammalian cells. We speculate that these pathways may be
involved in the broader control of the viscoelastic properties of the cell.
Proteomic Clues to Cell Organization.
M. Wühr1,2, T. Güttler2, L. Peshkin1, G.C. McAlister2, M. Sonnett1,2, K. Ishihara1, A.C. Groen1, M. Presler1,
B.K. Erickson2, T.J. Mitchison1, S.P. Gygi2, M.W. Kirschner1; 1Department of Systems Biology, Harvard
Medical School, Boston, MA, 2Department of Cell Biology, Harvard Medical School, Boston, MA
The composition of the nucleoplasm determines the behavior of key processes such as transcription, yet
there is still no reliable and quantitative resource of nuclear proteins. Furthermore, it is still unclear how
the distinct nuclear and cytoplasmic compositions are maintained. To describe the nuclear proteome
quantitatively, we hand isolated the large nuclei of frog oocytes and measured the nucleocytoplasmic
partitioning of ~9000 proteins by mass spectrometry. Most proteins localize entirely to either nucleus or
cytoplasm, only ~17% partition equally. Native size but not polypeptide‐molecular‐weight is predictive
of localization: partitioned proteins exhibit native size larger than 100 kDa. To evaluate the role of
nuclear export in maintaining localization, we inhibited Exportin 1. This resulted in the expected
relocation of proteins towards the nucleus, but only 3% of the proteome was affected. Thus, complex
assembly and passive retention, not continuous active transport, is the dominant mechanism for the
maintenance of distinct nuclear and cytoplasmic compositions.
Optogenetic control of molecular motors and organelle distributions in cells.
L. Duan1, D. Che1, K. Zhang2, Q. Ong1, S. Guo1, B. Cui1; 1Chemistry, Stanford University, Stanford, CA,
Biochemistry, University of Illinois Urbana‐Champaign, Champaign, IL
Organelle transport and distribution play important roles in various cellular activities, including cell
polarization, intracellular signaling, cell survival, and apoptosis. However, establishing a direct link
between organelle distribution and cellular functions is hindered by the lack of means to manipulate
molecular motors in the complex intracellular environment. Here, we provide an optogenetic strategy to
control the transport and distribution of organelles by light. This is achieved by optically recruiting
molecular motors onto organelles through the heterodimerization of Arabidopsis thaliana cryptochrome
2 (CRY2) with its interacting partner CIB1. Upon exposure to blue light, CRY2 and CIB1 dimerize within
subseconds, which requires no exogenous ligands and low intensity of light. We demonstrate that
various organelles, including mitochondria, peroxisomes, and lysosomes, can be driven toward the cell
periphery or the cell nucleus upon recruitment of specific molecular motors. Light‐induced motor
recruitment and organelle movements are repeatable, reversible, and can be achieved at subcellular
regions. This optogenetic strategy provides a valuable tool to unveil the roles of distributions of various
organelles in cellular functions in living cells.
Minisymposium 18: Regulation and Integrated Functions of Actin Cytoskeleton
Structural Investigation of Cooperative Actin Disassembly.
V. Tang1, A. Nadkarni1, W.M. Brieher1; 1Cell and Developmental Biology, University of Illinois, Urbana‐
Champaing, Urbana, IL
Actin depolymerization by cofilin, coronin, and Aip1 presents a new mode of filament disassembly that is
faster than cofilin‐mediated severing alone. We investigate this new mode of filament destabilization by
electron microscopy. Filament structure was disorganized five seconds after adding the
depolymerization factors. By 10 seconds, canonical features of the filaments were no longer
recognizable. The disassembly reaction could be divided into a series of subreactions. The first step
involves coronin mediated acceleration of phosphate release from the actin filaments. This promotes
highly cooperative cofilin loading to produce long stretches of polymer that are saturated with cofilin
and hypertwisted. These hypertwisted filaments are stable, but they are highly unstable in the presence
of Aip1 which induces a fast and concerted obliteration of hypertwisted filaments in seconds. Our result
implies a mechanism of filament disassembly that is not simply severing but a cooperative dissolution of
subunit interactions within a hypertwisted filament.
apCAM adhesion sites are mechanically isolated from retrograde actin flow by by local Arp 2/3
complex‐dependent actin assembly during neurite growth.
K.B. Buck1, A.W. Schaefer1, V.T. Schoonderwoert1, M.S. Creamer1, E.R. Dufresne2, P. Forscher1;
Molecular Cell and Developmental Biology, Yale University, New Haven, CT, 2Mechanical Engineering,
Yale University, New Haven, CO
Homophilic binding of Ig superfamily molecules, such as apCAM, leads to actin filament assembly near
nascent adhesion sites. Such actin assembly can generate significant localized forces that have not been
characterized in the larger context of axon growth and guidance. Here we have used apCAM coated
bead substrates applied to the surface of neuronal growth cones to investigate the roles of Arp 2/3
complex and Rac localized near apCAM bead binding sites during growth responses to physically
restrained beads. Pharmacological inhibition of the Arp 2/3 complex or Rac attenuated F‐actin assembly
near bead binding sites, decreased the efficacy of growth responses, and blocked accumulation of
signaling molecules associated with nascent adhesions. Actin assembly‐driven propulsive motility of
unrestrained apCAM coated beads matched or exceeded rates of retrograde network flow and required
Arp 2/3 complex activity. Local Arp2/3 dependent F‐actin assembly modulated force production
between tractive and propulsive modes, and could buffer nascent adhesion sites from the relatively
strong mechanical effects of retrograde flow. These studies introduce a new paradigm for regulation of
traction force by local Arp2/3 and Rac dependent actin assembly.
Cellular control of cortical actin nucleation.
M. Bovellan1, A. Yonis1, Y. Romeo2, M. Biro3, A. Boden4, P. Chugh4,5, M. Vaghela1, M. Fritzsche1, D.
Moulding6, A. Jegou7, A.J. Thrasher6, G. Romet‐Lemonne7, E.K. Paluch4,5, P.P. Roux2, G. Charras1; 1London
Centre for Nanotechnology, University College London, London, United Kingdom, 2IRIC, Universite de
Montreal, Montreal, Canada, 3Centenary Institute of Cancer Medicine and Cell Biology, University of
Sydney, Sydney, Australia, 4Molecular Cell Biology and Genetics, Max Planck Institute , Dresden,
Germany, 5MRC‐LMCB, University College London, London, United Kingdom, 6Institute of Child Health,
University College London, London, United Kingdom, 7LEBS, CNRS, Gif sur Yvette, France
The contractile actin cortex is a thin layer of actin, myosin, and actin‐binding proteins that subtends the
membrane of animal cells. The cortex is the main determinant of cell shape and plays a fundamental
role in cell division, migration, and tissue morphogenesis. For example, cortex contractility plays a
crucial role in amoeboid migration of metastatic cells and during division, where its misregulation can
lead to aneuploidy. Despite its importance, our knowledge of the cortex is poor and even the proteins
nucleating it remain unknown, though a number of candidates have been proposed based on indirect
evidence. Here, we used two independent approaches to identify cortical actin nucleators: a proteomic
analysis using cortex‐rich isolated blebs and a localization/shRNA screen searching for phenotypes with
a weakened cortex or altered contractility. This unbiased study revealed that two proteins generated
the majority of cortical actin: the formin mDia1 and the Arp2/3 complex. Each nucleator contributed a
similar amount of F‐actin to the cortex, but had very different accumulation kinetics. Electron
microscopy examination revealed that each nucleator affected cortical network architecture differently.
mDia1 depletion led to failure in division but not Arp2/3. Interestingly, despite not affecting division on
its own, Arp2/3 inhibition potentiated the effect of mDia1 depletion. Our findings indicate that the bulk
of the actin cortex is nucleated by mDia1 and Arp2/3 and suggest a mechanism for rapid fine‐tuning of
cortex structure and mechanics by adjusting the relative contributions of each nucleator. To investigate
potential mechanisms of nucleator activity regulation, we searched for known regulators of Diaph1 and
Arp2/3 within the proteins identified by proteomics. This identified four candidate nucleation promoting
factors (Flightless‐I, IQGAP1, the Wave complex, and NCKIPSD) and we assessed their role in actin
cortex growth using a localisation/shRNA screen in constitutively blebbing cells as well as long term
imaging of HeLa cells undergoing mitosis. Depletion of IQGAP1 phenocopied Diaph1 depletion, while
knockdown of Wave complex subunit resulted in phenotypes similar to Arp2/3 depletion. As two of the
NPFs (IQGAP1 and NCKIPSD) can interact with both Diaph1 and the Arp2/3 complex, we are using
truncation mutants to determine how they regulate the interplay between the two nucleators. Finally,
we are confirming these results using biotin proximity labelling approaches to identify other NPFs
participating to the regulation of the activity of actin nucleation in the cortex.
Chromosome transport during starfish oocyte meiosis: A model for 3D contraction generated
by actin filament dynamics.
P. Bun1, S. Dmitrieff1, M. Mori2, F. Nedelec1, P. Lénárt1; 1Cell Biology and Biophysics Unit, European
Molecular Biology Laboratory, Heidelberg, Germany, 2Genome Information Research Center, Osaka
University, Osaka, Japan
Actomyosin contractility is well understood in highly ordered muscle sarcomeres. However, it is much
less clear how contractility is generated in unordered actin networks. Recently, in vitro reconstituted
networks and studies on in vivo cortical networks provided important insights into possible mechanisms
of contractility in 2D, but contraction in a randomly oriented 3D network have not yet been
characterized. We have characterized earlier a 3D contractile F‐actin meshwork that mediates long‐
range, directional transport of chromosomes in meiotic starfish oocytes (Lenart et al, 2005; Mori et al,
2011). The F‐actin meshwork contracts homogeneously, while the directionality is provided by its
anchoring to the cortex. Chromosomes are larger than the mesh size of the network, and thus
transported by passive sieving (Mori et al, 2011). Here, we use this system as a model to understand
contractility in a 3D F‐actin meshwork. To this end, we monitor the F‐actin meshwork dynamics as well
as the transport of scattered chromosomes while performing biochemical and/or mechanical
manipulations. Surprisingly, using small‐molecule inhibitors against non‐muscle myosin II (NMY‐2) and
its activators we show that chromosome speed and directionality are not perturbed during the
transport. This observation suggests that NMY‐2 is not involved in contraction. Generation of contractile
forces could alternatively rely on filament dynamics, as recently hypothesized by others (Sun et al.,
2010)(Jégou et al., 2013). To test this possibility, we monitor the F‐actin density during the chromosome
transport. Quantitative analysis indicates a prominent depolymerization activity within the F‐actin
meshwork, as evidenced by a decrease of F‐actin mass concomitant with chromosome transport. This
results in a plateauing of the concentration of F‐actin, indicating a tight coordination between
contraction of the meshwork and F‐actin depolymerization rate. Based on these observations, we
propose a model for a force‐generating mechanism based on the pressure generated by pre‐stressed
cytoplasmic F‐actin, pushing the cross‐linked F‐actin meshwork already present in the nuclear region.
We hope to support our model by identifying potential actin cross‐linkers and bundlers and by
characterizing the structural dynamic of the F‐actin meshwork organization at high temporal resolution.
Pulsatile Contractions are an Intrinsic Property of MyosinIIa in Adherent Cells.
M.A. Baird1, R.S. Fischer1, A. Wang 1, R.S. Adelstein1, C.M. Waterman1; 1NHLBI, National Institutes of
Health , Bethesda, MD
Non‐muscle myosin‐IIA is vital to many cellular and developmental processes including cell adhesion,
migration, cytokinesis, and tissue morphogenesis. Myosin‐II activity can be observed in steady
contraction events such as during cytokinesis, but may also exhibit pulsatile contraction in other
contexts. For example, during embryogenesis in Drosophila, myosin‐II is recruited to the cell cortex to
initiate pulsatile contractions, which directly correlate with cell constriction thereby inducing large
structural changes in tissue. Here, we find that myosin‐II pulses also occur in a variety of adherent cells
in culture, suggesting that this behavior is an intrinsic property of the actomyosin cortex, and is not
unique to cells in a tissue. Each pulsatile event displays a consistent duration, while intervals between
contraction pulses can vary. Utilizing a photoactivatable myosin‐IIA probe, we show that these pulses
are the result of local recruitment of new myosin II mini‐filaments to sites of contraction pulses, and F‐
actin is locally concentrated concomitantly with the contraction pulse. Small molecule perturbations of
myosin II activators demonstrate that myosin II phosphorylation is required for the pulsatile behavior.
The contraction pulses occur independently of integrin activation. Using gadolinium inhibition we find
that the pulsatile contractions require the activity of stretch‐activated calcium channels. Additionally,
we show that this pulsatile behavior is unique to myosin IIA and not IIB. Utilizing myosin‐IIA / IIB
chimeras, we show that the myosin IIA motor domain is required for the pulsatile behavior. Thus, the
contractile pulses are a result of the kinetics of the myosin‐IIA motor, whereas the myosin‐IIB motor is
insufficient to induce this dynamic behavior. We conclude that this pulsatile contraction is an inherent
and intrinsic property of myosin‐IIA / F‐actin networks in adherent cells independent of their assembly
into organized tissues.
From Molecules to Meshes: Generating Tension in an Actin Cortex.
M.B. Smith1, P. Chugh1, D. Cassani1, G. Salbreux2, E.K. Paluch1; 1LMCB, UCL, London, England,
Theoretical physics of biology, The Francis Crick Institute, London, England
Cell shape is largely controlled by mechanical forces at the cell surface. The actin cortex is a thin
polymer network just beneath the plasma membrane of many eukaryotic cells. The actin cortex can
provide different amounts of tension during cell shape changes. For example, when cells round up for
mitosis their surface tension increases. It is our goal to understand how the molecular behavior of
proteins in the actin cortex influences the mechanical forces.
To look at the the cell‐scale properties of the actin cortex we measured thickness and tension of cells in
various conditions. We show that cells entering mitosis have a higher surface tension, and a thinner cell
cortex compared to cells in interphase. We also genetically perturbed actin related proteins which lead
to changes in cortex thickness and tension. We use electron microscopy to show the membrane facing
structure of the actin cortex. We have created a model to assimilate experimental results and answer
the questions: How does an isotropic cross‐linked network generate tension? How can filament
organization affect tension generation?
The model is a simulation of a cross‐linked actin filament network pulled on by myosin mini‐filaments.
The model contains three representative structural elements of the actin cortex: actin filaments, myosin
minifilaments, and actin cross linkers. Changes in model parameters can capture the changes observed
in live cell experiments. Further, the model exhibits an optimum condition for generating tension. The
model will help understanding tension generation across scales, from molecular processes to actin
network organization and up.
Protrusive Waves Guide 3D Cell Migration along Nanofibers.
C. Guetta‐Terrier1, H. Long2,3, P. Monzo1, J. Zhu4, Z. Yue5,6, P. Wang5,6, S.Y. Chew2,3, A. Mogilner7, B.
Ladoux1,8, N.C. Gauthier1; 1Mechanobiology Institute, National University of Singapore, Singapore,
Singapore, 2School of Chemical Biomedical Engineering, Nanyang Technological University, Singapore,
Singapore, 3Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore,
Singapore, 4Cellular and Molecular Physiology, Yale University, New Haven, CT, 5Cardiovascular Research
Institute, National University Health System, Singapore, Singapore, 6Department of Medicine, Yong Loo
Lin School of Medicine, National University of Singapore, Singapore, Singapore, 7Courant Institute and
Department of Biology, New York University, New York, NY, 8Institut Jacques Monod, CNRS UMR 7592
Université Paris Diderot, Paris, France
Three‐dimensional cell migration on fibrous matrices is actively investigated. However environment
complexity renders study of the key molecular and physical mechanisms implicated difficult. Using
reductionist approaches based on 3D electrospun fibers, we report for various cell types that single cell
migration along fibronectin‐coated nanofibers is robustly driven by lateral actin‐based waves. Those
cyclical waves propagate up to several hundred micrometers from the cell body to the leading edge and
thus promote highly persistent movement. They harbor a fin‐like shape and depend on a balance
between Rac1/N‐WASP/Arp2/3 and Rho/formins pathways. The waves originate from one major
adhesion site at leading end of the cell body. This adhesion site is linked through acto‐myosin based
contractility to another site at the back of the cell, allowing force generation and matrix deformation.
From experiments and simulations, we unveil that cells migrating on fibrous environment developed fin‐
like protrusions. These dynamic actin waves appear as versatile structures to achieve proper migration
within complex fibrillar 3D scaffolds.
Emergence of an apical epithelial cell surface in vivo.
J. Sedzinski1, E. Hannezo2, F. Tu1, M. Biro3, J.B. Wallingford1; 1Molecular Biosciences, University of Texas,
AUSTIN, TX, 2Cavendish Laboratory, Department of Physics, University of Cambridge, Cambridge, UK,
University of Sydney, Sydney, Australia
Epithelial sheets are crucial components of all metazoan animals, enclosing organs and protecting the
animal from its environment. Epithelial homeostasis poses unique challenges, as addition of new cells
and loss of old cells must be achieved without disrupting the fluid‐tight barrier and apicobasal polarity of
the epithelial sheet. Several studies have identified genetic and cell biological mechanisms underlying
extrusion and delamination of cells from epithelia, but far less is known of the converse mechanism by
which new cells are added. Here, we combine molecular, pharmacological and laser‐dissection
experiments with quantitative physical modeling to characterize forces driving emergence of a new
apical surface as nascent cells are added to a vertebrate epithelium in vivo. We find that this process
involves an interplay between cell‐autonomous actin‐generated forces in the emerging cell and the
mechanical properties of neighboring cells. Our findings define the forces driving a novel cell behavior,
and by complementing previous studies of delamination and extrusion, they provide a more
comprehensive understanding of epithelial homeostasis.
Filament spacing in actin bundles is an architectural feature that drives protein sorting.
J.D. Winkelman1, C. Suarez1, A.J. Harker2, G.M. Hocky3, J.R. Christensen1, A.N. Morganthaler1, J.R.
Bartles4, D.R. Kovar1; 1Molecular Genetics and Cell Biology, University of Chicago, Chicago, IL,
Biochemistry and Molecular Biology, University of Chicago, Chicago, IL, 3Chemistry, University of
Chicago, Chicago, IL, 4Feinberg School of Medicine, Northwestern University, Chicago, IL
Cells assemble multiple actin filament networks within a common cytoplasm that facilitate diverse
processes including division, polarization and motility. Networks with different actin filament
organizations are assembled and maintained by unique sets of actin binding proteins (ABPs) with
complementary properties. Cells contain two functionally distinct types of bundled actin networks that
display fundamentally different mechanical properties. The first network type comprises actin filaments
aligned in parallel and tightly packed that can protrude from the membrane without buckling, e.g.
filopodia and microvilli. These networks are associated with bundlers such as fascin, espin and fimbrin
which generate bundles of filaments that are spaced ~ 10 nm apart. Filaments in the second bundled
network type contain widely‐spaced antiparallel filaments that generate pulling forces by myosin
minifilament‐mediated contraction. These contractile networks are associated with the bundler α‐
actinin which generates mixed‐polarity bundles of filaments spaced ~35 nm apart. How particular sets of
ABPs, including Fascin and α‐actinin, sort to these networks and tailor their properties is critical to
understanding how a dynamic actin cytoskeleton self‐organizes. Using in vitro techniques such as single
molecule three‐color TIRF microscopy we show that when mixed, these two classes of bundlers
intrinsically segregate and form distinct bundled domains with either dense‐parallel‐ or widely‐spaced
filaments. Importantly, other actin binding proteins strongly discriminate between these two bundle
types within the same reaction. Here we directly observe an unappreciated mechanism that drives the
intrinsic sorting of actin binding proteins by recognizing a specific architectural feature (filament
spacing) in actin filament networks.
Gibco Award Presentation and E.B. Wilson Medal Presentation and Address
Skin Stem Cells: Where They Come From, How They Make Tissues and Who Controls Their
E. Fuchs1,2; 1Mammalian Cell Biology Development, The Rockefeller University, New York, NY, 2Howard
Hughes Medical Institute, New York, United States
As a cell biologist, I’ve long been interested in stem cells, which have the ability to self‐renew long term
and differentiate into one or more tissues. Typically, stem cells are used sparingly to replenish cells
during normal homeostasis. However, even stem cells that are quiescent must be able to respond
quickly to injury in order to fuel rapid tissue regeneration. How stem cells balance self‐renewal and
differentiation is of fundamental importance to our understanding of normal tissue maintenance and
wound repair. Increasing evidence suggests that the regulatory circuitry governing this balancing act is
at the root of some types of cancers. Skin is an excellent model system to understand where stem cells
come from, how they function in normal tissue generation and how this process goes awry in cancer.
We’ve identified and characterized at a molecular level the stem cell niches within the epidermis, hair
follicle, and glands of the skin and traced their embryonic origins. We’ve learned that during
development and homeostasis, stem cell behavior is controlled not only through cues received from
their microenvironment but also through signals emanating from their differentiating lineages. We’ve
been dissecting how extrinsic niche cues, particularly those eliciting changes in BMP, WNT and SHH
signaling, triggers a cascade of chromatin‐associated and transcriptional changes within stem cells that
governs their activation during tissue development, homeostasis, and hair regeneration. We’ve applied
this knowledge in exploring how stem cells change as they exit their niche and embark upon a specific
lineage program or alternatively participate in wound‐repair following injury and in cancer.
LGBTQ Diversity Session
Mobilization of the actin cytoskeleton by microbial pathogens.
M.D. Welch1; 1Molecular Cell Biology, University of California, Berkeley, Berkeley, CA
The ability of diverse microbial pathogens to hijack the actin cytoskeleton of their host cells is crucial for
pathogenic processes such as attachment, invasion, transport and spread. To better understand this
recurrent mechanism of pathogenesis, and to uncover pathways of actin regulation in uninfected cells,
we have worked to elucidate the molecular mechanisms used by pathogens to harness actin
polymerization for intracellular actin‐based motility. Our studies have revealed that pathogen proteins
mimic an unexpectedly diverse array of host actin polymerization pathways. Listeria monocytogenes
ActA and baculovirus p78/83 mimic host nucleation promoting factors (NPFs) that activate the Arp2/3
complex and assemble branched actin networks, similar to actin assembly in cellular lamellipodia.
Burkholderia species BimA exhibits species‐specific differences in its mechanism of action; BimA from B.
thailandensis mimics host NPFs, whereas BimA from B. pseudomallei and B. mallei mimic host Ena/VASP
actin polymerases to assemble bundled actin arrays like those in cellular filopodia. Rickettsia species
produce two actin assembly proteins that act in distinct phases of infection; early phase infection uses
RickA to activate the Arp2/3 complex, and late phase infection uses Sca2 to mimic host formin proteins
and polymerize bundled filament arrays. Differences in the mechanism of actin assembly by each
pathogen protein lead to distinct efficiencies of motility initiation, and distinct biophysical characteristics
of motility, with Arp2/3 driven movement occurring in curved and meandering paths, and Ena/VASP‐ or
formin‐driven movement occurring in straighter paths. Beyond the diversity of mechanisms, we have
found that pathogen actin‐based motility is also used for disparate infectious processes, ranging from
cell‐to‐cell spread for Listeria, Rickettsia and Burkholderia, to transit into and out of the nucleus for
baculoviruses. Future studies of how and why pathogens mobilize the actin cytoskeleton will continue to
reveal new molecular mechanisms of pathogenesis, and may uncover new mechanisms for regulating
dynamic actin processes in cells.
Oral Presentations‐Wednesday, December 16
Symposium 7: Beyond the Five Senses: Detection of Magnetic and Electric Fields
Cell Biology of Magnetic Particle Formation in Magnetotactic Bacteria.
E. Cornejo1, D. Hershey1, P. Subramanian2, X. Ren3, J. Hurley3, G.J. Jensen2, A. Komeili1,3; 1Plant and
Microbial Biology, University of California, Berkeley, Berkeley, CA, 2Division of Biology and Biological
Engineering, California Institute of Technology, Pasadena, CA, 3Molecular and Cell Biology, University of
California, Berkeley, Berkeley, CA
Many living organisms depend on the production of highly ordered biominerals for their growth and
survival. The products of biomineralization processes serve as skeletons, teeth, defensive structures and
navigational tools. While these features are generally composed of inorganic materials, their formation,
subcellular positioning and display depend on intriguing cell biological processes that are poorly
understood. In my group we examine the formation of magnetic nanoparticles within magnetosome
organelles of magnetotactic bacteria as a model for the cell biology of biomineralization. Magnetosomes
are invaginations of the inner cell membrane and contain a unique set of proteins that direct the
biomineralization of highly pure and precisely sculpted magnetic iron oxide (magnetite) or iron sulfide
(greigite) crystals. Multiple magnetosomes are arranged into chains with the help of a dedicated
cytoskeletal network, allowing the cell to orient in geomagnetic fields and efficiently find the proper
redox conditions in aquatic environments. The main focus of my group is to understand the molecular
mechanisms that control the biogenesis of the magnetosome membrane, its assembly within the cell
and the biomineralization of magnetic nanoparticles. Using genetic, we have identified large number of
genes that participate in every step of magnetosome formation. Here, I will describe our recent efforts
to examine the process of magnetosome formation in real‐time. These studies have revealed that the
biophysical features of the magnetosome membrane are directly tied to checkpoints that monitor the
progress of mineral formation. I will also describe the biochemical and structural characterization of
factors involved in the formation of magnetic nanoparticles that have led to novel insights into the
evolution of iron‐binding proteins.
Individual and collective cell polarization and migration in electric field.
Y. Sun1, K. Zhu1, Y. Sun1, B. Reid1, F. da Silva Ferreira1, Y. Li1, X. Gao1, M. Ying1, B.W. Draper1, M. Zhao1, A.
Mogilner2; 1Medical School, University of California, Davis, United States, 2Courant Institute and Dept of
Biology, New York University, New York, United States
Cells are electrical units: they transport ions across membranes, and as a result are surrounded and
regulated by electrical fields and currents. Ability of cells to sense the direction of an electric field (EF) –
galvanotaxis – plays important roles in wound healing, development and regeneration. Galvanotaxis is,
arguably, as important as chemotaxis, but our understanding of how cells use EF to navigate is in its
infancy. We used individual zebrafish keratocyte cells and cohesive groups of cells to investigate their
polarization, migration and directional sensing in EF. Individual cells polarized slowly spontaneously; EF
drastically accelerated the motility initiation. While pharmacological perturbations of myosin or PI3K‐
signaling disabled the spontaneous polarization, EF induced polarization of the perturbed cells.
Spontaneously polarized cells moved persistently for hours; if EF was switched on, moving cells started
to migrate along EF, and after EF was switched off, cells continued to migrate randomly. Surprisingly, if a
cell was stationary and if its polarization was EF‐induced, the motile cell lost polarity and became
stationary as soon as EF was switched off. Microscopy demonstrated that adhesion (vinculin) spatial
distributions were different in motile cells after spontaneous versus EF‐induced polarization.
Large cohesive groups of cells did not migrate persistently. When EF was switched on, cells at the
group’s leading edge accelerated to cathode, and gradually all cells followed with the same speed.
Surprisingly, when PI3K‐inhibited cell groups started to move to cathode in EF, not the leading, but the
rear edge cells accelerated to the cathode, pushing the cells in the middle and leading edge in front of
them. Groups of myosin‐inhibited cells could not move persistently in EF, but rather the leading cells
crawled to cathode, while the rear cells crawled to anode, and the group just deformed. Even more
surprisingly, groups of cells with both PI3K and myosin inhibited sensed EF readily and rapidly moved to
cathode. Microscopy showed that the polarized cell groups organized like a giant motile cell, with
protrusive front of the group and contractile rear. We use these data to formulate a model of individual
and collective cell directional response. According to this model, electro‐chemical PI3K‐mediated
‘frontness’ pathway, electro‐mechano‐chemical myosin‐mediated ‘backness’ pathway, and mechano‐
chemical contact inhibition pathway are engaged in a tag‐of‐war to orient cells individually and
collectively in EF.
Minisymposium 19: A Tribute to Alan Hall: Rho GTPase Signaling
P‐cadherin/β‐PIX/Cdc42 promotes collective cell migration through increase in the anisotropy
and magnitude of mechanical forces.
C. Gauthier‐Rouviere1, C. Plutoni1, E. Bazellierres2, M. Le Borgne‐Rochet1, F. Comunale1, A. Brugués2, D.
Planchon1, N.S. Morin1, S. Bodin1, X. Trepat2; 1CRBM, CNRS, MONTPELLIER, France, 2IBEC, BARCELONA,
Collective cell migration is essential for organism development and wound healing as well as for
metastatic transition, the primary cause of cancer‐related death, and, involves cell‐cell adhesion
molecules of the cadherin family. Increased levels of P‐cadherin expression are correlated with tumor
aggressiveness in carcinoma and aggressive sarcoma; however, how P‐cadherin promotes tumor
malignancy remains unknown. In this study, using integrated cell biology and biophysical approaches,
we determined that P‐cadherin specifically induces the collective cell migration through the generation
of biochemical cues that mechanically favor collective cell motion. We identified a P‐cadherin/β‐
PIX/Cdc42 axis that can induce focal adhesions and cell layer polarization in the direction of migration,
thereby increasing intercellular and traction force anisotropy and strength. Our results identify a novel
Cdc42‐mediated mechanism for controlling the anisotropy and strength of mechanical forces during
collective cell migration. Moreover, we demonstrate that P‐cadherin activates Cdc42, which in turn
regulates polarity and forces anisotropy, to drive collective cell migration. This is a is a cross‐disciplinary
study at the croos‐road of cell biology and biophysics that will attrack cell and cancer biologists and
physicists interested in mechanobiology.
The small GTPase Rac3 is essential for invadopodia maturation and function in breast cancer
S.K. Donnelly1,2, J. Bravo Cordero1,2, J.S. Condeelis1,2, L. Hodgson1,2; 1Gruss Lipper Biophotonics Center,
Albert Einstein College of Medicine, Bronx, NY, 2Anatomy and Structural Biology, Albert Einstein College
of Medicine, Bronx, NY
Invadopodia are actin‐based protrusions that mediate the extracellular matrix (ECM) degradation
necessary for tumor cell invasion. RhoGTPases are a large protein family that has well‐established roles
in cytoskeletal regulation. They also play key roles in the function and formation of invadopodia. While
RhoA, Rac1 and Cdc42 are the best‐studied RhoGTPases, diverse roles have emerged for closely related
isoforms, RhoC and RhoG. However the functional importance of many other isoforms remains
unknown. Rac3 is 93% identical to Rac1 and both isoforms localize to the plasma membrane when
activated. Here we show that Rac3 accumulates at invadopodia and plays a critical role in regulating the
maturation of these structures and their ability to degrade the ECM. This is in direct contrast to the
recently published role of Rac1 in controlling invadopodia disassembly. Our data suggests that Rac3
regulates invadopodia maturation and function in two ways. Firstly Rac3 modulates the anchoring of
invadopodia to the ECM via a CIB1‐integrin signaling axis. Depletion of these components results in
short‐lived invadopodia that cannot degrade the ECM. Secondly, Rac3 regulates the localization of the
Arf6‐GAP, GIT1 to the core of invadopodia, which could affect MT1‐MMP trafficking to these structures.
Indeed, Rac3 depletion reduces MT1‐MMP localization at invadopodia. The striking sequence similarity
between Rac3 and Rac1 suggest that their different functions at invadopodia are due to highly regulated
spatio‐temporal control of their localization and activity. To investigate this hypothesis, we have
generated and validated a novel single‐chain FRET biosensor that reports on the real time activation
status of Rac3. Furthermore, we have developed a photo‐activatable Rac3, allowing us to directly
interrogate (biosensor) and then perturb (PA‐Rac3) Rac3 activity at invadopodia in live cancer cells.
Together our data suggests that Rac3 is an important molecular switch that regulates the transition of
invadopodia from precursor to mature functional structure.
Patterning Rho signaling at the epithelial Zonula Adherens: a tale of feedback loops.
R. Priya1, G.A. Gomez1, S. Budnar1, S. Verma1, N.A. Hamilton1, A.S. Yap1; 1Division of Cell Biology and
Molecular Medicine, Institute for Molecular Bioscience, The University of Queensland, Brisbane, QLD,
The apical cadherin junctions of epithelial cells (zonulae adherente, ZA) are sites that couple E‐cadherin
and actomyosin to generate tissue‐level tension for epithelial integrity and morphogenesis. The Rho
GTPase plays a key role in coordinating adhesion and contractility at the ZA. By imaging GTP‐Rho with
location biosensors we now find that the ZA constitutes a mesoscopic zone of Rho signaling that is stable
over time scales much longer (> 10 fold) than its constituent molecules. This poses the challenging
question of how cells generate stable zones of signaling from populations of molecules whose individual
elements are much more dynamic. Combining quantitative optical microscopy with mathematical
modeling, we uncovered a novel signaling apparatus that ensures the fidelity of the Rho zone at the ZA
and buffers it against stochastic biological noise. This involves an unexpected feedback loop from
Myosin II to Rho. The key is scaffolding of ROCK‐1 at the ZA by the rod domain of Myosin II, which serves
to prevent the junctional localization of the Rho antagonist, p190B RhoGAP. This is achieved via ROCK‐1
mediated phosphorylation and inhibition of Rnd3, which otherwise recruits p190B to the ZA. Myosin II
thus preserves Rho signaling at the ZA by defining a zone that excludes the Rnd3/p190B Rho inhibitory
apparatus. Mathematical modeling revealed that this feedback network has the capacity to confer
bistability on signaling outcomes at the level of populations of Rho molecules. This is predicted to
promote robustness of the Rho zone. The spatial definition of the Rho zone can be further understood
as the product of a reaction‐diffusion model that is determined by local stabilization of cortical Myosin
II. These findings carry wide implications for diverse cellular processes relying on actomyosin remodeling
and Rho signaling.
Spontaneous vs. light‐induced symmetry breaking – characterizing the relation of mechanical
traction forces and morphological events during cell migration.
K. Hennig1, O. Destaing2, C. Albiges‐Rizo2, M. Balland1; 1Materials, Optics and Instrumental Techniques
for the Life Sciences (MOTIV), Laboratory of Interdisciplinary Physics (LiPhy), Grenoble, France,
Differentiation and Cell Transformation, Institute Albert Bonniot (IAB) Inserm U823, Grenoble, France
The initiation of cellular migration requires breaking of symmetry which can be triggered by external
cues or occur spontaneously in uniform stimulant. Critical factors are RhoGTPase‐mediated cytoskeleton
reorganization and asymmetric acto‐myosin contraction as well as cellular traction forces. Nevertheless,
this force‐motion relation of migration remains so far unknown. The purpose of our study is to
quantitatively examine the spatio‐temporal coordination of front‐rear polarization during initiation of
cellular migration and to relate this sequence of morphological events to spatial changes in traction
force patterns. Traction force microscopy is employed to follow those dynamic force modifications
during symmetry breaking. Microfabricated 1D adhesive fibronectin line patterns prepolarize adherent
NIH 3T3 fibroblasts, restrict direction of migration and mimic a 3D fibrillar matrix. Adhesive lines of
limited length allow fibroblasts to fully spread while migration is inhibited. Observed are inward
directed, contractile traction forces in the pico‐joule range at cell edges. The maximum magnitude of
stress is equal on either side leading to a symmetric spatial traction profile. Non‐restricted NIH 3T3 are
able to initiate cell migration. During symmetry breaking the traction force distribution becomes
anisotropic. The spatial force pattern during migration is characterized by a spherical, small region of
high stress at the cell front, while the rear exhibits an elongated force distribution. The nucleus
physically divides those two distinct compartments of traction and is continuously displaced along the
axis of migration. Simultaneously the cell coordinates protrusion formation at the leading edge and rear‐
retraction. The latter is mainly mediated by RhoA. In order to control breaking of symmetry and hence
guide migration an optogenetic tool is employed which perturbs the mechanical behavior of an initially
non‐motile fibroblast. The optogenetic system is based on light‐induced dimerization of Cry2 and CIBN
and allows high spatial and temporal control of localization as well as activation of our target protein
RhoA. Upon local blue light stimulation RhoA‐mediated acto‐myosin network reorganization drives rear
contraction and initiates polarization. Consequently, imbalanced traction forces and morphological
changes trigger cell motility. To conclude, we developed a straightforward approach to relate a
sequence of morphological events to cellular mechanical forces which show that motility does not just
involves a global gradient of cytoskeleton and molecular compounds but also of traction forces.
Furthermore, local optogenetic RhoA‐induced contraction enables us to initiate symmetry breaking and
direct cell migration.
Stress Fibers Store Contractile Energy to Resist Changes in Cell Shape.
P.W. Oakes1, E. Wagner2, C.A. Brand3, D. Probst3, M. Linke3, U.S. Schwarz3, M. Glotzer2, M.L. Gardel1;
Institute for Biophysical Dynamics, James Franck Institute and Department of Physics, University of
Chicago, Chicago, IL, 2Department of Molecular Genetics and Cell Biology, University of Chicago,
Chicago, IL, 3Institute for Theoretical Physics and BioQuant, Heidelberg University, Heidelberg, Germany
Contractile stresses generated by the actomyosin cortex are vital to regulating shape changes in cells,
facilitating diverse processes like migration and cell division. Coordination of these contractile stresses is
driven by the spatial regulation of RhoA activity, which mediates the assembly of actomyosin bundles
through its downstream effectors, Dia and ROCK. To explore the spatiotemporal regulation of tension in
adherent cells, we have developed an optogenetic approach using the LOVpep system to locally control
RhoA activity in both space and time. Upon activation of RhoA we see local stress fiber contraction,
increases in traction stress, and accumulation of both myosin and actin in the activation region. This
accumulation is accompanied by long‐range flow along stress fibers of myosin towards the activation
region. Surprisingly, despite having ample time to remodel the cytoskeletal network, the RhoA‐induced
contractility is completely reversible upon removal of the activating light. The cytoskeleton relaxes
elastically, with the flow of cytoskeletal components reversing away from the region of activation. We
also find that local activation of RhoA does not lead to de novo stress fiber or focal adhesion assembly,
even at long time scales. Perturbations to the stress fiber architecture, however, including inhibition of
Dia or knockdown of myosin IIB, result in a loss of this elastic behavior. Under these conditions local
activation of RhoA permanently deforms and remodels the actin cytoskeleton. We find that our
experimental results can be captured by modeling the cell cortex as a viscous material embedded with
elastic stress fibers anchored to the substrate. The stress fibers thus serve to both transmit stress over
cellular length scales, and to elastically store the contractile energy of the locally activated RhoA regions.
These findings demonstrate that stress fibers enable cells to behave mechanically as solid‐like materials
over timescales traditionally associated with fluid‐like behavior, thus allowing cells to maintain their
shape under externally applied stresses.
RhoA activity cycling promotes dynamic cytoskeletal contractions during epithelial invagination.
F.M. Mason1, S. Xie1, M. Tworoger1, A. Martin1; 1Biology, Massachusetts Institute of Technology,
Cambridge, MA
The Rho family of small GTPases influences force generation by regulating cytoskeletal dynamics.
However, little is known about how RhoA activity is spatiotemporally regulated to organize actin‐myosin
cytoskeletal dynamics and promote changes in cell and tissue shape. In Drosophila, apically constricting
cells exhibit polarized RhoA signaling within their apical domain, as well as temporally dynamic or
pulsatile accumulation of RhoA effectors Rho‐associated kinase (ROCK) and Myosin‐II. Pulsatile RhoA
pathway activity suggests that RhoA activity is not simply turned “on” to drive contraction, but that
RhoA must be dynamically regulated in space and time. We observe that the RhoA activator, RhoGEF2,
undergoes pulsing similar to Myosin‐II. In addition, constitutive RhoA activation does not promote
pulsatile contraction or apical constriction, suggesting that cycling of RhoA activity is essential for
constriction. We have identified a Rho GTPase Activating Protein (GAP), which we name Cumberland
GAP (C‐GAP), that polarizes ROCK and Myosin‐II to the center of the apical cortex. Depletion or
overexpression of C‐GAP also results in altered Myosin‐II structure and pulsation dynamics, revealing
that C‐GAP regulates both the spatial and temporal activity of the RhoA pathway. Ultimately, loss of C‐
GAP results in delayed or failed tissue invagination. Interestingly, C‐GAP expression is upregulated by
transcription factors that promote RhoA activation and apical constriction. Thus, our data demonstrate
that simply turning RhoA activity “on” is insufficient to organize actin‐myosin contractility, and that
inactivation of RhoA signaling is as important as its activation to promote contractility in tissues.
Roles of guanine nucleotide exchange factors in regulating collective cell migration.
A. Zaritsky1, Y. Tseng2, M. Rabadán2, M. Overholtzer2, G. Danuser1, A. Hall2; 1Department of Cell Biology,
UT Southwestern Medical Center, Dallas, TX, 2Cell Biology Program, Memorial Sloan‐Kettering Cancer
Center, New York, NY
Guanine nucleotide exchange factors (GEFs), activators of Rho GTPases, regulate cell migration, but little
is known about their roles in collective cell migration, a motility mode that drives development,
regeneration and cancer invasion. We conducted a comprehensive targeted screen using 3 distinct and
validated shRNA constructs against the known 82 human genes encoding GEFs. Live cell imaging of in
vitro wound healing assays of human bronchial epithelial cells were performed and accompanied by
highly‐sensitive, objective spatiotemporal quantification of single cell behavior in in the collective to
identify and characterize GEFs that regulate various properties of collective cell migration.
A robust algorithmic pipeline was devised for phase contrast time‐lapse data to extract spatiotemporal
measures for cellular speed, directionality and coordination, and used to analyze over 1500 time‐lapse
experiments. Dimensionality reduction and clustering analysis were used to detect altered migration
phenotypes upon GEFs knock‐down and to gain insight into their functionality.
Knock‐down of TRIO, a RAC1, RHOG and RHOA specific GEF, had the most dramatic inhibition of cell
speed, directionality and coordination. Knock‐down of SOS1, a dual GEF for RAC and RAS, also induces a
reduction in speed and directionality. TUBA showed a dose‐dependent reduction in cell speed.
A group of RHOA‐specific paralog GEFs appeared to alter cell‐cell communication traits. Depletion of
ARHGEF3, ARHGEF11 or ARHGEF28 enhanced directed or coordinated migration speculatively through
increase in RAC1 activation strengthening adherent junctions. Depletion of ARHGEF1 or ARHGEF18
accelerated the transmission of a backward‐propagation wave, initiated at the monolayer edge that
transitioned cells from a non‐motile to a motile state. This resembled, with a weakened effect, the
alteration induced by depletion of RHOA, simultaneously transitioning cells at the front and back of the
monolayer from a non‐motile to a motile state, but with reduced final cell speed. This simultaneous but
weaker transition of RHOA‐depleted cells suggests a rapid front‐back transmission of mechanical cue
through the cell monolayer, speculatively through down‐regulation of Myosin‐II‐dependent actomyosin
contractility that enable passive cell‐cell force transmission. This work highlights in a systematic screen
the delicate balance between actin assembly and adhesion on the one hand and actomyosin
contractility on the other hand in driving collective migration.
We dedicate this work to our mentor and colleague Alan Hall, who has initiated this project in the
context of a NIGMS‐funded program project grant.
Minisymposium 20: Applications of Cell Biology 2
Increased spatiotemporal resolution reveals highly dynamic, tubular lattices in the peripheral
endoplasmic reticulum.
J. Nixon‐Abell1,2, C.J. Obara3, A.V. Weigel3, D. Li4, W.R. Legant4, K. Harvey2, E. Betzig4, J. Lippincott‐
Schwartz3, C.D. Blackstone1; 1Cell Neurology Section, National Institute of Neurological Disorders and
Stroke, Bethesda, MD, 2Department of Pharmacology, UCL School of Pharmacy, London, United
Kingdom, 3Section on Organelle Biology, Eunice Kennedy Shriver National Institute of Child Health and
Human Development, Bethesda, MD, 4Janelia Research Campus, Howard Hughes Medical Institute,
Ashburn, VA
The endoplasmic reticulum (ER) is an expansive, membrane‐enclosed organelle implicated in diverse
cellular functions, with complex structural elements. In most cells, the ER consists of the nuclear
envelope interconnected with ribosome‐studded perinuclear sheets and an extensive polygonal network
of highly dynamic tubules and peripheral sheets, whose functions remain poorly understood. Numerous
proteins are involved in maintaining this complex structural organization, and a large number are
mutated in neurologic disorders such as hereditary spastic paraplegia. Thus, understanding the
morphology of the ER and how it relates to various functions is critical. Here, we employ a variety of
emerging super‐resolution imaging approaches including point accumulation for imaging in nanoscale
topography (PAINT) and total internal reflection fluorescence‐structured illumination microscopy (TIRF‐
SIM) in both live and fixed cells to better understand the distinct morphologies and dynamics of the ER.
Gains in either spatial or temporal resolution reveal an increasingly dense clustering of subdiffraction‐
limited holes (~150‐250 nm diameter) with highly transient lifespans (~250 msec) in peripheral sheets.
This data suggests that many structures identified as sheets by confocal microscopy are actually
comprised of a tightly‐latticed meshwork of highly dynamic tubules. Accordingly, these lattices harbor a
number of curvature‐stabilizing ER proteins including multiple reticulon isoforms and proteins
associated with three way junction formation (i.e., atlastin‐1) or stabilization (i.e., lunapark). Lattices
also undergo rapid transitions to clear tubular networks whilst retaining proteins associated with the
stabilization and characterization of sheets (e.g., CLIMP‐63). Together, our results provide novel insight
into the structure of the peripheral ER. Specifically, many structures previously identified as peripheral
sheets are composed of a lattice of cross‐linked ER tubules, suggesting a complex, highly dynamic
system in which tubules and sheets may be more difficult to differentiate than previously thought.
Kinesin‐1 defects lead to altered axonal transport in the important hippocampal to basal
forebrain memory circuit in living intact mouse brain.
C.S. Medina1, O. Biris2, F. Chaves1, A.J. Zimmerman1, T. Falzone3, R.E. Jacobs4, E.L. Bearer1; 1Pathology,
University of New Mexico Health Sciences Center, Albuquerque, NM, 2Division of Engineering, Brown
University, Providence, RI, 3IBCN‐Facultad de Medicina, Universidad de Buenos Aires, Buenos Aires,
Argentina, 4Beckman Institute, California Institute of Technology, Pasadena, CA
Microtubule‐based motors carry cargo back and forth from the synaptic region to the cell body. We are
using magnetic resonance imaging to study cell biological processes, such as transport, in the brain of
intact living animals. Mutations in KIF5A, the human gene encoding the heavy chain of conventional
kinesin‐1, result in axonal transport causing peripheral neuropathies. Some mutations in human KIF5A
also cause severe central nervous system defects, yet kinesin‐1 activity is involved in any particular tract
in the central nervous system is unknown. While transport dynamics in the peripheral nervous system
are well characterized experimentally, transport in the central nervous system is less experimentally
accessible. Here we study transport dynamics within the central nervous system of living mice with and
without kinesin light chain 1 (KO KLC‐1). We injected an MRI tract tracer, Mn2+, together with a classic
histologic tracer, rhodamine dextran, into CA3 of the posterior hippocampus. We imaged axonal
transport in vivo by capturing whole‐brain 3D magnetic resonance images (MRI) at discrete time points
after injection. To determine histologic location of tracer, the mice were sacrificed, perfusion‐fixed and
embedded for serial whole brain sectioning in register and examined by fluorescence microscopy and
histologic staining of alternate sections. The MR images were skull‐tripped, align‐warped and analyzed
by statistical parametric mapping with ANOVA to compare intensities at successive time points within
genotype and between genotypes. The position of the Mn2+‐enhanced MR signal as it proceeded from
the injection site into the forebrain, the expected projection from CA3 was delayed in KLC‐KO mice at
early time points, yet achieved destinations comparable to wild type littermates at 25 hr, similar to our
previous publication on transport in the optic tract (Bearer et al. 2007). Histologic examination
confirmed precision of the injection site and revealed smaller brain size with normal anatomy in the KLC‐
KO compared to wild type, while color‐coded rendering of diffusion tensor MR images demonstrated
relatively normal anatomy. We conclude that kinesin‐1 defects diminish the rate of axonal transport
within the living brain and thus cause cognitive impairments, possibly due to decreased synaptic vesicle
replenishment at active synapses. Finally this study demonstrates the power of MEMRI to observe and
measure transport dynamics in the central nervous system caused by defects in motors which may
result from or lead to brain pathology. Supported by NS062184, NS046810, MH087660 and the Harvey
Family Endowment (ELB).
Poly(ADP‐Ribosylation) Regulates Axon Regeneration.
A.B. Byrne1,2, Y. Sekine1,3, R.D. McWhirter4, S.M. Strittmatter1,3, D.M. Miller III4, M. Hammarlund1,2;
CNNR, Yale University, New Haven, CT, 2Genetics, Yale University, New Haven, CT, 3Neurology, Yale
University, New Haven, CT, 4Cell and Developmental Biology, Vanderbilt University, Nashville, TN
The ability of a neuron to regenerate its axon after injury depends significantly on its intrinsic
regenerative potential. This potential is determined by signaling pathways, including the conserved DLK
MAP kinase pathway, that regulate neuronal gene transcription. However, it remains unclear which
transcriptional targets are functionally relevant and how these targets affect the cell biology of injured
Activation of DLK signaling is critical for regeneration. We analyzed RNAseq profiles of FACS‐sorted C.
elegans GABA motor neurons with activated DLK‐1 signaling to identify transcriptional targets of the DLK
pathway. We found that the poly(ADP‐ribose) glycohydrolases (PARGs) parg‐1 and parg‐2 are
upregulated by DLK‐1 signaling, and that PARGs are required for regeneration.
PARGs degrade poly(ADP‐ribose), which is synthesized by poly(ADP‐ribose) polymerases (PARPs). The
balance between PARG and PARP activity determines cellular levels of poly(ADP‐ribose) (PAR). We found
that loss of function of the PARP genes parp‐1 and parp‐2 results in increased axon regeneration. Loss of
PARP also increased regeneration in mammalian cortical neurons.
PARP inhibitors are currently in clinical trials for indications including cancer and stroke. We found that
animals treated with PARP inhibitors after injury showed enhanced axon regeneration. Thus, PARylation
regulates the acute response of neurons to axon injury, and PARP inhibition after injury is sufficient to
improve regeneration.
Together, our results indicate that regulation of poly(ADP‐ribose) levels is a critical function of the DLK
regeneration pathway, that poly(ADP‐ribosylation) inhibits axon regeneration across species, and that
chemical inhibition of PARPs can elicit axon regeneration.
Dynamics of thrombus formation in mouse testicular surface vein by new collaborative analysis
with live‐imaging in vivo and following TEM observation.
A. Sawaguchi1, S. Nishimura2,3,4; 1Anatomy, University of Miyazaki, Miyazaki, Japan, 2Cardiovascular
Medicine, The University of Tokyo, Tokyo, Japan, 3Translational Systems Biology and Medicine Initiative,
The University of Tokyo, Tokyo, Japan, 4Center for Molecular Medicine, Jichi Medical University, Tochigi,
One of the goals of biomedical microscopy is to elucidate a functional morphology in vivo, leading to
associated pathophysiology and further clinical investigation. The dynamics of thrombus formation has
yet to be elucidated to prevent human diseases, such as infarction in the heart and the brain, etc. Most
recent progress of in vivo multi‐photon laser scanning microscopy (LSM) enables us to visualize
experimentally induced vascular damage followed by thrombus formation in blood vessels of mice.
However, a new reliable method remains to be developed to acquire a transmission electron
microscopic (TEM) image of the experimentally formed thrombus in situ, hampered by difficulties in
locating and processing the thrombus to be cut into ultrathin sections. To address this problem, we
newly developed a “Vascular Mapping” method for correlative light and electron microscopy (CLEM) of
thrombus formation in vivo. Briefly, after anesthesia, Texas‐Red–dextran was injected to visualize blood
flow, and then the mouse testicular surface vein was exposed onto a multi‐photon LSM. Following a
capture of complex vascular pattern at lower magnification, the focused region was damaged by laser
irradiation to obtain sequential images of thrombus formation in situ.
For TEM image preparation, the testis was excised and immersed into half‐strength Karnovsky fixative.
Then, the focused region was excised referring the captured image of complex vascular “map”, and
processed to be embedded into epoxy resin. Next, semi‐thin sections (3 µm in thickness) were
sequentially cut tracing upstream of the damaged vein to reach the thrombus. The obtained semi‐thin
section of thrombus was re‐embedded into epoxy resin to be cut into ultrathin sections (70 nm in
thickness) using capsule‐supporting ring.
As results, the present “vascular mapping” approach succeeded in TEM observation of the thrombus,
yielding live‐CLEM imaging in vivo. The preliminary observation demonstrated a noteworthy attachment
of neutrophils onto the thrombus as well as fine structure of the aggregated platelets to arrest the
bleeding. It is highly anticipated that further application will clarify not only thrombus formation but also
the following fibrinogenolysis and blood vessel repair, leading to a goal of biomedical microscopy for
further clinical investigation such as antithrombotic treatment under “live” CLEM imaging.
We demonstrate step‐by‐step procedures from “live” to “TEM” imaging of thrombus formation in vivo
by newly devised “Vascular mapping” method.
Microscopic detection of vulnerable sites exposed on cell‐bound HIV to inform better vaccine
M. Mengistu1, G.K. Lewis1, A.L. DeVico1; 1Institute of Human Virology, University of Maryland School of
Medicine, Baltimore, MD
The pathways of protein domain exposure on the HIV surface delimit relationships between infection
and humoral immunity and provide an important framework for developing approaches to treat or
prevent HIV infection. The gp120 and gp41 components of HIV envelope trimers have been intensely
studied in the context of free virions and it is understood how these proteins must experience multiple
structural rearrangements during the course of host cell receptor (CD4 and coreceptor) engagement.
However, the impact of such changes on the structural features of an intact virion bound to a host cell
remains an ongoing, critical question. Previous predictions held that conserved envelope domains
operating within cell contact zones encounter a variety of spatial constraints that occlude their exposure
and/or immunoreactivity. However, certain lines of evidence, including the partially successful RV144
vaccine trial, are discordant with these predictions. Although transition state gp120 domains absent on
free virions but induced by CD4 binding (CD4i) are predicted to be profoundly occluded on cell‐bound
HIV, cognate anti‐CD4i monoclonal antibodies reproducibly mediate antibody dependent cell cytotoxic
activity (ADCC) against target cells presenting attached virions. Prompted by such discrepancies, we
applied confocal and three‐dimensional superresolution microscopic techniques to interrogate virions
bound to entry‐competent target cells. Surprisingly, these analyses showed that CD4i epitopes on gp120
are visibly exposed and reactive with whole cognate antibodies on bound virion surfaces for a period of
hours. These exposure patterns resemble what is observed for constitutively expressed neutralizing
epitopes on cell‐bound virions. Further, three‐dimensional direct stochastic optical reconstruction
microscopy (dSTORM) showed that CD4i epitopes were unexpectedly exposed distal to the HIV‐cell
contact interface in a manner readily accessible to circulating antibodies. Such distal exposures of
certain CD4i epitopes was abrogated on mutant virions with disrupted envelope – matrix connections.
Collectively these observations suggest that previously unsuspected structural dynamics emerge on HIV
during host cell attachment. Such processes may provide new windows of vulnerability to antiviral
countermeasures against freshly targeted host cells.
Neighbor‐killing via the Type‐VI Secretion System enables high‐efficiency, cross‐species
acquisition of antibiotic resistance genes in competent Acinetobacter bacteria.
R.M. Cooper1, L. Tsimring1, J. Hasty1; 1Biocircuits Institute, UC San Diego, San Diego, CA
Spreading antibiotic resistance is one of the most significant threats facing our current healthcare
system. In addition to treating existing infections, reliable antibiotics are also essential for enabling
complex surgeries. However, antibiotic discovery has slowed even as resistance is spreading, and the
dearth of new drugs in the pipeline is recognized as a major challenge. One of the CDC’s top six drug‐
resistant pathogens of concern is Acinetobacter baumanii, which only a few decades ago was dismissed
as “relatively avirulent” and “infrequently pathogenic”. The main reason for A. baumanii’s surprising
emergence as a major global pathogen is its remarkably rapid acquisition of new resistance and
virulence genes by horizontal gene transfer (HGT). In addition to infecting human hosts, A. baumanii
kills adjacent bacteria by directly injecting toxins via a Type VI Secretion System (T6SS). Despite being
widespread among bacteria, the T6SS was only recently discovered, and its ecological implications are
still unclear. While studying the closely related A. baylyi, we have found that T6SS‐mediated killing of
neighboring bacteria releases a locally high concentration of DNA, which can then be acquired by the
predatory cell. This killing‐mediated, high efficiency HGT requires either a replicating plasmid or
genomic homology, but the bait gene within the prey cell can be carried either on a plasmid or
genomically. A mutant that cannot kill neighboring cells acquires foreign resistance genes at much
lower efficiency. Killing‐mediated HGT is so efficient that we were able to visualize functional
acquisition and expression of GFP from neighboring E. coli in real‐time, with multiple HGT events
occurring in each ~104 um2 microfluidic trap. Upon addition of kanamycin to simulate treatment of a
patient, the parent Acinetobacter strain ceased growing, but the newly drug‐resistant, GFP‐positive cells
quickly dominated the chips. A dynamical model of growth, killing, and transformation enabled
experimental determination of key parameters, and accurately predicts gene transfer in a biofilm. Both
competence and contact‐dependent killing are common throughout Acinetobacter, including A.
baumanii. These results suggest that “active gene snatching”, in which cells actively generate locally
high DNA concentrations by lysing their neighbors, as opposed to passively waiting for chance
encounters with environmental DNA, may help explain rapid transfer of resistance genes. This
demonstrates a key role for the T6SS in HGT, helping explain its role in biofilm dynamics. Understanding
this high‐efficiency mechanism for HGT could help us inhibit the spread of antibiotic resistance, and
harnessing it could even potentially help restore susceptibility.
The Retromer Complex Regulates Exosomal APP at the Drosophila Neuromuscular Junction.
R.B. Walsh1, M.J. Zunitch1, A.N. Becalska1, J. Gittzus1, A.A. Rodal1; 1Department of Biology, Brandeis
University, Waltham, MA
Alzheimer’s Disease (AD) is one of the most common neurological diseases of the elderly, and is
characterized by severe neurodegeneration, progressive dementia and death. A pathophysiological
hallmark of AD is the accumulation of extracellular proteinaceous aggregates called amyloid plaques,
which are composed primarily of a fragment of the Amyloid Precursor Protein (APP) called Amyloid‐β
(Aβ). APP and its processing proteases reside on the plasma membrane and in specific endosomal
compartments, so the intracellular trafficking of APP is likely to play a significant role in the amount of
Aβ produced. Further, although it is known that elevated levels of Aβ are neurotoxic, it is not well
understood how Aβ propagates between neurons and escapes the cell to aggregate in plaques.
Exosomes are small, endosome‐derived secreted vesicles which contain proteins, RNAs, and other
macromolecules, and may contribute to the propagation of pathological proteins in the brain. Here, we
show at the Drosophila neuromuscular junction (NMJ) that a major trafficking trajectory of APP is
towards motor‐neuron derived exosomes, and that these exosomes are enriched for an APP C‐terminal
fragment. We found that the endosomal membrane coat complex Retromer is a modulator of APP‐
positive exosome production. Retromer is composed of two subunits – a highly‐conserved cargo‐
selective subunit (the CSC, or the “core” Retromer) and a membrane‐binding subunit. We have shown
that loss of VPS35 (a CSC component) causes an increase in total as well as exosomal APP, and that loss
of SNX1 (a SNX‐BAR membrane‐binding component) decreases the amount of APP‐labeled exosomes at
the Drosophila NMJ. These results suggest that sequential steps in Retromer function may play opposing
roles in exosome processing. Loss of VPS35 has been implicated in late‐onset AD, and our results
indicate that increased exosomes may be part of the pathological mechanism of extracellular Aβ
Gamma‐Secretase Inhibitors (GSIs) and Modulators (GSMs) Induce Distinct Conformational
Changes in the Active Sites of Gamma‐Secretase and Signal Peptide Peptidase (SPP).
N. Gertsik1, D. Chau2, Y. Li3; 1Biochemistry and Structural Biology, Weill Cornell Medical College, New
York, NY, 2Pharmacology, Weill Cornell Medical College, New York, NY, 3Pharmacology, Memorial Sloan
Kettering Cancer Center, New York, NY
Signal peptide peptidase (SPP) and gamma‐secretase are aspartyl proteases that catalyze regulated
intramembrane proteolysis, a process that controls the activity and function of membrane proteins in all
living systems studied to date. SPP is implicated in viral lifecycle and gamma‐secretase is a frontline
target for cancer and Alzheimer’s disease (AD) drug development. Both enzymes are multi‐span
transmembrane proteins containing the YD/GXGD catalytic motif, but are structurally and functionally
distinct. The key differences include their opposite orientations within the membrane, limited sequence
homology, and requirement for protein cofactors. Despite these differences, SPP and gamma‐secretase
share important pharmacological characteristics. For example, transition‐state gamma‐secretase
inhibitor L‐685,458 inhibits SPP activity and L‐685,458‐based photoaffinity probe L‐646 specifically labels
SPP. These data suggest overall structural similarity within the active sites of SPP and gamma‐secretase.
However, detailed analysis and comparison of active site conformation has never been performed. In
this study, we used a chemical biology approach to probe the active sites of SPP and gamma‐secretase.
Through our “photophore walking” technique, which allows us to target the various sub‐pockets of the
enzyme active site, we were able to detect subtle structural differences between the enzyme active
sites. Furthermore, we used the photophore walking method to determine the effects of GSIs and
GSMs, compounds used in clinical trials for cancer and Alzheimer’s disease, on the active site
conformations of SPP and gamma‐secretase. This marks an important step toward understanding
potential SPP‐associated off‐target effects of GSI/GSM treatment. Finally, we provide additional
evidence that active SPP exists as a homodimer.
Directive Nanoscale Cues for Regenerative Neural Cell Systems.
V.M. Ayres1, V.M. Tiryaki2, I. Ahmed3, D.I. Shreiber3; 1Electronic Biological Nanostructures Laboratory,
Michigan State University, East Lansing, MI, 2Computer Engineering, Siirt University, Siirt, Turkey,
Biomedical Engineering, Rutgers, The State University of New Jersey, Piscataway, NJ
There is great promise and excitement in the field of regenerative medicine resulting from experimental
observations that when injury sites are supplied with scaffold‐based bridge environments that are
physically and biochemically mimetic for the extracellular matrix /basement membrane, endogenous
cell populations can regenerate and re‐establish functional connections with healthy surrounding tissue.
Alternatively such scaffolds can be used to replace lost tissue through grafting of exogenous cells,
including pluripotent stem cells that receive their selective development cues from a scaffold bridge
Cell systems that are influenced by scaffold‐based environments are ubiquitous throughout the body
and in our research are represented by key components of the central nervous system (CNS). Repair of
CNS traumatic injury remains a challenging problem, with limited treatment options beyond costly and
uncertain rehabilitation. One of the hurdles that must be overcome is the formation of the glial scar,
established after traumatic central nervous system injury in mammalian systems. Reactive astrocytes at
a wound site undergo a morphological change, extending interwoven processes that form chain‐like
clusters while expressing inhibitory proteoglycans and tenascins. Formation of the glial scar therefore
biomechanically and biochemically blocks axonal elongation and reconnection.
Our recent research indicates that one powerful approach to control astrocyte behavior is to engineer
biomaterials at the nanoscale to have instructive properties that steer the cells away from the reactive
state. Our group has identified implantable scaffolds composed of electrospun nanofibers that appear to
have promising in vivo wound healing properties, including mitigation of astrocytic scarring. I will
present our latest research using advanced scanning probe microscopy, electron microscopies and
contact angle measurements to quantitatively describe the nanoscale elasticity, surface roughness, work
of adhesion and surface polarity of the promising scaffold environments. I will also present our latest
research using super‐resolution microscopy‐based immunocytochemistry and atomic force microscopy
to evaluate neural cell morphological responses and protein expressions, along with early results from
our investigations of how and which integrin‐family receptors are possibly involved. Finally, I will present
our novel application of cluster methods for analysis of uncorrelated data across multiple experimental
modalities and its use in the quantification of synergistic scaffold properties and cell responses. Our goal
is to generate an integrated description and understanding of how a biomaterials‐based intervention
can achieve its goals.
Minisymposium 21: Cell Cycle Regulation
Mechanisms of mitotic regulation by the APC/C.
D. Lu1, J.R. Girard1, A. Mizrak1, D.O. Morgan1; 1Physiology, University of California, San Francisco, San
Francisco, CA
The dramatic and beautiful events of chromosome segregation are governed by a multi‐subunit
ubiquitin ligase called the anaphase‐promoting complex or cyclosome (APC/C), which catalyzes the
ubiquitination and destruction of securin, cyclins, and other important regulatory proteins. The APC/C is
activated at specific cell‐cycle stages by association with an activator subunit, Cdc20 or Cdh1, which
provides binding sites for specific substrate sequence motifs, or degrons. Like other members of the
RING family of E3s, the APC/C catalyzes direct ubiquitin transfer from an E2‐ubiquitin conjugate (E2‐Ub)
to lysine residues on the protein substrate.
We are using experimental and computational methods to dissect the mechanisms underlying the
ordered degradation of APC/C substrates. Our single‐cell analyses of GFP‐tagged APC/C substrates in
budding yeast suggest that the timing of substrate degradation is determined by changes in substrate
degrons and other mechanisms that govern substrate interaction with the APC/C. Biochemical
experiments show that degrons might not simply enhance substrate affinity for the APC/C but also
enhance the enzyme’s catalytic function: for example, mutation of a specific degron motif in the S cyclin,
Clb5, reduces maximal catalytic rate with that substrate. We have also used computational modeling to
develop a more complete understanding of how changes in substrate affinity and catalytic rate can
provide the precisely ordered degradation of different APC/C‐Cdc20 substrates. These studies reveal
that following APC/C activation, the multi‐step ubiquitination process provides an adjustable delay in
the onset of substrate degradation. Small variations in substrate affinity and/or catalytic rate can have a
major impact on the timing of degradation onset, and robust degradation timing can be achieved over a
broad range of parameters. When two substrates share the same pool of APC/C‐Cdc20, their relative
enzyme affinities and rates of turnover can influence the partitioning of APC/C‐Cdc20 among substrates,
and competition between substrates can occur. However, increased expression of the early APC/C‐
Cdc20 substrate Clb5 does not delay the degradation of the later substrate securin, arguing against a
role for competition in establishing securin degradation timing. These studies provide a conceptual
framework for understanding the factors that determine the timing of substrate modification in the cell
Mechanical stretch triggers rapid epithelial cell division through the stretch‐activated channel
J. Lindblom1, P.D. Loftus1, K. Edes1, M.J. Redd1, J. Rosenblatt1; 1Oncological Science, Huntsman Cancer
Institute, Salt Lake City, UT
Epithelial cells turnover at some of the fastest rates in the body, despite acting as a barrier for the
organs and organisms they encase. How do the number of dying cells match those dividing to maintain
constant numbers and an impermeable barrier? We previously found that epithelial cell crowding
results in extrusion of epithelial cells that later die via the stretch‐activated channel Piezo11. Conversely,
what controls epithelial cell division to balance cell death at steady state? Here, we find that cell division
occurs in sparse regions of epithelia, suggesting that stretch could control proliferation. We show that
stretching epithelia either mechanically or by wounding causes cells to rapidly enter mitosis and that
this effect is again mediated by Piezo1. Stretch‐activation of Piezo1 triggers a population of epithelial
cells in G2 that are poised for repair to accumulate cyclin B and enter mitosis following stretch.
Furthermore, unlike wild type epithelia, monolayers lacking Piezo1 do not stop migrating upon wound
closure, suggesting a role for Piezo1 in controlling both contact inhibition of epithelial cell proliferation
and migration. Piezo1 controls mitosis only once epithelia reach correct functional densities, presumably
by localizing to the plasma membrane, where it could mechanically sense tensions and activate calcium
currents. Because Piezo1 senses both mechanical crowding and stretch, it may act as a homeostatic
sensor to control epithelial cell numbers by driving extrusion and apoptosis when cells are too crowded
and cell division when they are too sparse.
Lattice Light‐Sheet Microscopy of Dividing Cells in Culture and in Live Zebrafish Embryos With
High Spatiotemporal Resolution Demonstrates Similar Membrane and Cell Volume Dynamics.
S. Upadhyayula1, F. Aguet1, R. Gaudin1, E. Coccuci1, K. He1, B. Chen2, K.R. Mosaliganti3, W.R. Legant2, T.
Liu2, E. Marino1, G. Danuser4, S.G. Megason3, E. Betzig2, T. Kirchhausen1; 1Cell Biology, Harvard Medical
School, Boston, MA, 2Howard Hughes Medical Institute , Janelia Research Campus, Ashburn, VA,
Systems Biology, Harvard Medical School, Boston, MA, 4Cell Biology, UT Southwestern, Dallas, TX
Lattice light‐sheet microscopy (LLSM) using ultrathin light sheets is a minimally invasive, new 3D live‐cell
imaging method, that can be used for long durations before altering the physiological state of the
biological specimen (Chen et al. , 2014). Here we used LLSM to investigate the membrane dynamics
during cell division of mammalian cells grown ex‐vivo in culture and of cells growing in situ in the
developing eye of zebrafish embryos. Using gene‐edited human cells expressing the endocytic clathrin
adaptor AP2, we simultaneously tracked every clathrin‐based endocytic event within a single cell and
established that all coated pits have the same assembly dynamics regardless of whether they formed at
the dorsal (free) or ventral (attached) plasma membranes. We then confirmed that clathrin‐dependent
endocytosis persists during all stages of cell cycle, transiently decreasing during mitosis.
We also used the volumetric imaging data sets obtained with LLSM to determine the volume and surface
area of the cultured cells during all stages of cell division. In agreement with earlier observations
(Stewart et al. , 2011), we found that the volume of cells undergoing mitosis remained constant and was
similar to the volume of most cells in interphase, and also similar to the combined volume of the two
daughter cells undergoing cytokinesis. In contrast, mitotic cells had less surface area than interphase
cells and they rapidly recovered it at the onset of cytokinesis. We extended the 4D imaging to the
rapidly dividing cells in the eye and spine of zebrafish embryos analyzed 17‐24 hrs post fertilization.
Using automated cell segmentation and tracking, we established the boundaries of all cells from which
we then extracted the surface area and volume data – thus allowing us to compare the values from
interphase cells with cells during all stages of division. As found ex‐vivo , within the tissue the volume of
dividing cells also remained constant, whereas the surface area transiently decreases during metaphase.
From these non‐invasive observations we propose two general principles for animal cells undergoing cell
division: (1) fine‐tuning of the homeostatic balance of membrane traffic allowing the transient decrease
of the surface area of rounded mitotic cells, principally by slowing down the recycling step, and (2)
maintenance of cell volume during mitosis, at a time when the nucleus has lost its nuclear envelope
while undergoing massive topological rearrangements of the chromatin.
Chen, B.‐C. et al. (2014). Lattice light‐sheet microscopy: imaging molecules to embryos at high
spatiotemporal resolution. Science 346 , 1257998.
Stewart, M. P., Helenius, J., Toyoda, Y., Ramanathan, S. P., Muller, D. J., and Hyman, A. A. (2011).
Hydrostatic pressure and the acto
Direct visualization of nuclear envelope rupture.
N. Wesolowska1, P. Lénárt1, M. Mori2; 1Cell Biology and Biophysics, European Molecular Biology
Laboratory, Heidelberg, Germany, 2Osaka University, Osaka, Japan
In metazoa the nuclear envelope (NE) is rapidly disassembled at the onset of mitosis in a process called
nuclear envelope breakdown (NEBD). The breakdown, critical to faithful and efficient chromosome
segregation, requires complete dismantling of the complex NE assembly of two membranes interlinked
with the underlying lamina, the proteins of the inner and outer NE and the nuclear pore complexes
(NPCs). A phase of sudden rupture appears to be a general feature of NEBD suggesting involvement of
mechanical forces. In cultured mammalian cells microtubules have been shown to tear NE membranes.
In contrast, we have recently uncovered that in starfish oocytes a cortex‐like actin shell rapidly
assembles on the inner side of the NE to promote this event. We hypothesize that, independent of
whether rupture is generated by the actin or microtubule cytoskeleton, pulling forces are involved.
However, the way forces are transduced to the NE and the type of membrane rearrangements that
result from it remain completely unknown. The aim of our research is to visualize the disruption of the
NE organization during this essential and conserved event. NEBD proceeds as a spatial wave along the
enormous nucleus of the starfish oocyte. This affords us the possibility of inspecting this rapid process
through distinct intermediates spatially ordered in a single sample. We have performed 3D electron
microscopy imaging (using Focused Ion Beam Scanning Electron Microscopy) of the nuclei at the exact
time of rupture by direct correlation to fluorescent markers. These markers, which indicate the moment
and site of breach of the permeability barrier, can be used for targeting to the site of NE rupture at the
subcellular resolution. This approach will allow us to understand the nature of membrane disruption at
the site of rupture and how the initial crisis propagates through the NE. Furthermore, through
immunostaining we have found that the F‐actin shell intercalates within the still‐intact lamina and
extends spike‐like protrusions towards the NE. This suggests that the NE is pulled apart from the lamina,
potentially leading to a necessary destabilization of the NE structure. Through our work we hope to
elucidate with unprecedented spatial and temporal resolution the global reorganization of membranes
and protein assemblies during NEBD.
Sphingolipids Activate an ER Stress Surveillance (ERSU) Pathway that Monitors the Proper
Inheritance of Functional Endoplasmic Reticulum (ER) During the Yeast Cell Cycle.
M. Niwa1, F. Yagisawa1, F. Pina‐Nunez1, J.T. Chao1, A.B. Tam1; 1Molecular Biology, UCSD, La Jolla, CA
During a normal cell cycle, various checkpoints and regulatory mechanisms are in place to ensure proper
segregation of nuclear genomic material. Currently, much less is known about whether similar types of
regulatory mechanisms or checkpoints operate for segregation of cytoplasmic components, including
organelles. The ER is a vital organelle as it serves as a gateway for secretory and membrane proteins and
is a major site for both lipid synthesis and intracellular calcium regulation. The ER cannot be synthesized
de novo but arises only from existing ER; therefore, each daughter cell must inherit the ER from the
mother cell at every division.
Previously, we reported the existence of an ER surveillance mechanism (ERSU) that monitors and
ensures the inheritance of a functional ER by the daughter cell. When the demand on ER function is high
(known as ‘ER stress’), transmission of the cortical ER (cER) into the daughter cell is blocked, and
concomitantly, the septin complex at the bud neck is altered, preventing the cell from proceeding to
cytokinesis. Paradoxically, these events are not activated by components of the well‐known Unfolded
Protein Response (UPR), but rather by the cell surface receptor Wsc1 and the MAP kinase Slt2. An
illustration of the importance of this ERSU pathway is that cells deficient in slt2 (slt2Δ) fail to delay
inheritance of functionally stressed ER and undergo cell death. Conversely, preventing the stressed ER
from entering the daughter cell rescues slt2Δ cell growth. Thus, the ERSU pathway is a cell cycle
checkpoint for ensuring that the yeast daughter cell inherits a functional ER.
One important question about the ERSU pathway is how it becomes activated in response to ER stress.
Since neither Wsc1 nor Slt2 is localized in the ER, we hypothesized that an ER‐resident component(s)
initiates the ERSU pathway according to the level of ER stress. Of note, while WSC1 and SLT2 are both
involved in the Cell Wall Integrity (CWI) pathway, we previously showed that the ERSU pathway is
independent of the CWI.
We recently discovered that cells defective in sphingolipid synthesis, like ERSU‐deficient slt2Δ cells,
failed to block the inheritance of the functionally stressed ER by the daughter cell and mislocalized the
septin complex from the bud neck. Sphingolipid synthesis is initiated in the ER and cells defective in
sphingolipid synthesis were able to induce the UPR normally, demonstrating that the inability to initiate
the ERSU pathway was not due to a lack of ER stress. Further analyses of how sphingolipid/ceramide
activates the ERSU pathway will be discussed.
Spatial Regulation of Greatwall by Cdk1 and PP2A‐Tws in the Cell Cycle.
P. Wang1,2, M. Larouche1,2, H. Mehsen1,2, K. Normandin2, D. Kachaner1,2, G. Emery2,3, V. Archambault1,2;
Department of Biochemistry and Molecular Medicine, Université de Montréal, Montréal, QC, 2Institute
for Research in Immunology and Cancer, Montréal, QC, 3Department of Pathology and Cell Biology,
Université de Montréal, Montréal, QC
Entry into mitosis requires the phosphorylation of multiple substrates by cyclin B‐Cdk1, while exit from
mitosis requires their dephosphorylation, which depends largely on the phosphatase PP2A in complex
with its B55 regulatory subunit (Tws in Drosophila). At mitotic entry, cyclin B‐Cdk1 activates the
Greatwall (Gwl) kinase, which phosphorylates Endosulfine proteins, thereby activating their ability to
inhibit PP2A‐B55 competitively. The inhibition of PP2A‐B55 by the Gwl‐Endosulfine axis at mitotic entry
facilitates the accumulation of phosphorylated Cdk1 substrates, and failure in this process results in
mitotic collapse after nuclear envelope breakdown. The spatial coordination of these enzymes is crucial
for this mechanism. In interphase, Gwl is nuclear, while PP2A‐B55 is cytoplasmic. We and others have
recently shown that Gwl suddenly relocalizes from the nucleus to the cytoplasm in prophase, before
nuclear envelope breakdown and that this controlled localization of Gwl is required for its function.
Phosphorylation of Gwl by cyclin B‐Cdk1 at multiple sites appeared to be required for its nuclear
exclusion, but the precise mechanisms remained unclear. In addition, how Gwl returns to its nuclear
localization was not explored.
We have investigated these mechanisms in Drosophila. We found that cyclin B‐Cdk1 directly inactivates
a Nuclear Localization Signal in the central region of Gwl. This phosphorylation facilitates the
cytoplasmic retention of Gwl, which is exported to the cytoplasm in a Crm1‐dependent manner.
Inactivation of either event prevents the efficient relocalization of Gwl from the nucleus to the
cytoplasm in prophase. In addition, we found that PP2A‐Tws promotes the return of Gwl to its nuclear
localization during cytokinesis.
Our results indicate that the cyclic changes in Gwl localization at mitotic entry and exit are directly
regulated by the antagonistic cyclin B‐Cdk1 and PP2A‐Tws enzymes. These findings highlight the
importance of spatial control in the reciprocal coordination between master regulators of the cell cycle.
Bistability of a coupled Aurora B kinase‐phosphatase system in cell division.
A.V. Zaytsev1, D. Segura‐Pena2, E.R. Ballister2, A. Calderon2, A.M. Mayo2, R. Stamatov2, M. Godzi1, L.
Peterson3, B.E. Black4, F.I. Ataullakhanov5,6,7, M.A. Lampson2, E.L. Grishchuk1; 1Department of Physiology,
University of Pennsylvania, Perelman School of Medicine, Philadelphia, PA, 2Department of Biology,
University of Pennsylvania, Philadelphia, PA, 3Department of Biology and Department of Chemistry,
Massachusetts Institute of Technology, Cambridge, MA, 4Department of Biochemistry and Biophysics,
University of Pennsylvania, Perelman School of Medicine, Philadelphia, PA, 5Department of Physics,
Moscow State University, Moscow, Russia, 6Center for Theoretical Problems of Physicochemical
Pharmacology, Russian Academy of Sciences, Moscow, Russia, 7Federal Research and Clinical Centre of
Pediatric Hematology, Oncology, Moscow, Russia
Aurora B is well established as a key regulator of multiple processes in cell division, including
kinetochore‐microtubule interactions and cytokinesis. Because of its dramatic redistribution from
centromeres to the spindle midzone at anaphase onset, mechanisms that regulate Aurora B localization
have been intensively studied. Despite these efforts, a paradox remains: major sites of Aurora B
localization in mammalian cells do not overlap with many of its substrates. How spatial patterns of
activity are established, enabling Aurora B to reach its substrates at a distance, is an outstanding
question of fundamental significance for cell division. We have combined quantitative approaches in
vivo, reconstitution of a simplified system in vitro, and mathematical modeling to provide evidence that
this activity at a distance depends on both sites of high kinase concentration and the bistability of a
coupled kinase‐phosphatase system. First, with a molecular system to control Aurora B localization in
human cells with high temporal precision, we show that localizing Aurora B at centromeres produces
soluble active kinase that phosphorylates distant substrates. Second, using purified components in vitro
we demonstrate that a coupled Aurora B kinase‐phosphatase system exhibits bistability, switching in a
hysteretic manner between “high” and “low” states of kinase activity, which is a hallmark of non‐linear
systems capable of generating a persistent source of active kinase. To our knowledge, this is the first
such coupled kinase‐phosphatase system reconstituted and theoretically analyzed. Finally, we
demonstrate bistability and hysteresis of Aurora B‐dependent phosphorylation in live mitotic cells, in
remarkable consistency with our predictions. Together, these results suggest a novel mechanism for
spatiotemporal organization of mitosis, whereby a bistable reaction‐diffusion system is responsible for a
generation of spatial patterns of Aurora B activity around its localization sites.
Minisymposium 22: Mammalian Cell Signaling
High resolution ordering of single cells along developmental trajectories with branches.
D. Pe'er1; 1Dept of Biological Sciences, Columbia University, New York, NY
We present novel methods to trace developmental trajectories with branches from high dimensional
single cell data. As a first step, we present Wishbone, an algorithm for identification of bifurcating
developmental trajectories.
As input, Wishbone receives multi‐dimensional single cell data and orders cells according to their
developmental trajectory. Additionally, Wishbone labels each cell as belonging either to the trunk or
one of two cell fates. Thus Wishbone can pinpoint where the bifurcation happens and characterize the
differences following bifurcation. This allows us to trace the expression of lineage markers along the
branching trajectory and characterize the decision making process for lineage commitment in single
cells. Wishbone is robust to the free parameters used and significantly outperforms existing methods
such as Monocle and Scuba in identification of both ordering and branching of cells. We demonstrate
Wishbone on T cell development in the mouse thymus, as this is an ideal system to study developmental
trajectories with branching. In this system, CD4+ helper T cells and CD8+ cytotoxic T cells develop from
lymphoid progenitors seeded from the bone marrow. We applied Wishbone to 42 channel mass
cytometry data and accurately recovered the various known stages in T cell development including the
bifurcation point. Wishbone identified differential roles for a number of key T cell transcription factors in
development of the two branches and also identified a potential preference for different signaling
pathways in the two branches.
Encouraged by accurate results on the well‐studied T‐cell differentiation, we sought to tackle human
myeloid differentiation. While the various terminal differentiated cell types in bone marrow
development have been well characterized, the different progenitors giving rise to these differentiated
cell types, particularly the monocyte and dendritic cell lineages are largely unknown. Using principles of
Wishbone, we expanded our algorithm to detect multiple branches to automatically detect the number
of branches and the hierarchical organization of these branches. This allows a high resolution mapping
of the different progenitors in human myeloid development and identification of differences and
similarities compared to myeloid development in mice. With the even increasing volume of single cell
datasets in both RNA and protein measurements, we anticipate that our algorithms will provide novel
insights to processes driving differentiation and eventually their dysregulation in disease.
Paracrine Communication Maximizes Cellular Response Fidelity in Wound Signaling.
L.N. Handly1, A. Pilko1, R. Wollman1,2,3; 1Chemistry and Biochemistry, University of California, San Diego,
La Jolla, CA, 2Cell and Developmental Biology, University of California, San Diego, La Jolla, CA, 3San Diego
Center for Systems Biology, La Jolla, CA
Population averaging due to paracrine communication can arbitrarily reduce cellular response
variability. Yet, variability is ubiquitously observed, suggesting limits to paracrine averaging. It remains
unclear whether and how biological systems may be affected by such limits of paracrine signaling. To
address this question, we quantified the signal and noise of Ca2+ and ERK spatial gradients in response to
a wound within a novel microfluidics‐based device. We found that while paracrine communication
reduces gradient noise, it also reduced the gradient magnitude. Accordingly we predicted the existence
of a maximum gradient signal to noise ratio. Direct measurement of paracrine communication verified
these predictions and revealed that cells utilize optimal levels of paracrine signaling to maximize the
accuracy of gradient‐based positional information. Our results demonstrate the limits of population
averaging and show the inherent tradeoff in utilizing paracrine communication to regulate cellular
response fidelity.
Causes and consequences of variation in p53 signaling dynamics in tissues and tumors.
J. Stewart‐Ornstein1, G. Lahav1; 1Systems Biology, Harvard Medical School, Boston, MA
Some tissues and tumors show great sensitivity to DNA damage, others do not. Experiments in mice
have shown that the transcription factor p53 regulates much of this differential sensitivity. To define
what features correlate with and what factors regulate p53 activity we combined live cell imaging of a
panel of human cancer cell lines with measurements of p53 in mouse tissues. By measuring the time
course of p53 signaling in each tissue in response to DNA damage we define how the dynamics of p53
vary and link this variation to higher radio‐sensitivity of the hematopoietic tissues. To study the
regulatory pathways that alter p53 signaling dynamics we use live cell imaging and a kinase inhibitor
collection of small molecules to identify regulators of p53 dynamics and investigate the differential
regulation of these pathways in a panel of cancer cell lines. We find that variation in DNA repair and
ATM signaling across cell lines explain substantial variance in p53 signaling among cell lines, but not
tissues. These results suggest that p53 signaling dynamics vary greatly across tissues and tumors but
that the processes that induce this variation in normal tissues and tumors may be different.
A circadian code for fat cell differentiation.
Z. Bahrami‐Nejad1, M.L. Zhao1, K. Tkach1, S. von Schie1, M.N. Teruel1; 1Chemical and Systems Biology,
Stanford University, Stanford, CA
Adipogenesis, the process through which precursor cells differentiate into fat cells (adipocytes), is
strongly driven by glucocorticoid signaling and occurs at a low rate of 10% turnover per year (1).
Mammals show significant oscillations in glucocorticoid and other differentiation‐inducing hormones
with frequencies of minutes to days (2). Since adipogenesis is induced through a highly cooperative
network of positive feedback loops centered around the master transcription factor PPARG (3,4), we
hypothesized that the time delays between these loops could filter and exploit the oscillatory rhythms in
input hormones to regulate the number of fat cells that are produced. Specifically, we wanted to test
whether certain pulse patterns of glucocorticoids selectively regulate the all‐or‐none decision to switch
from proliferating preadipocytes into non‐dividing, lipid‐accumulating adipocytes, and that this control
would be lost under constant stimuli. Strikingly we found that in most cells, the adipogenic network
rejects short or circadian glucocorticoid inputs, yielding a low rate of differentiation. In contrast, fat cell
production occurs at high rates in response to sustained (>12 hour) inputs. Using single‐cell imaging,
computational modeling, and live‐cell tracking of endogenously‐tagged PPARG, we show that with short
or circadian inputs, most cells are unable to reach the PPARG threshold needed to trigger the bistable
switch into the differentiated state. In addition, we show that the filtering of glucocorticoid inputs
occurs at the level of early, upstream regulators of the adipogenesis network, and that the slow timing
of the feedback from PPARG back to CEBPB is a main mechanism for the filtering. Together, we provide
a framework for understanding how circadian secretion of hormones could be molecularly linked to
body weight which may help explain why aging, Cushing's disease, overeating, and other conditions in
which the secretion of glucocorticoids loses its characteristic pulsatility, are so strongly correlated with
Spalding, K. L. et al. Dynamics of fat cell turnover in humans. Nature 453, 783–787 (2008).
Spiga, F., Walker, J. J., Terry, J. R. & Lightman, S. L. HPA axis‐rhythms. Comprehensive Physiology 4,
1273–1298 (2014).
Park, B. O., Ahrends, R. & Teruel, M. N. Consecutive positive feedback loops create a bistable switch that
controls preadipocyte‐to‐adipocyte conversion. Cell reports 2, 976–990 (2012).
Ahrends, R.,.., & Teruel, M.N. Controlling low rates of cell differentiation through noise and ultrahigh
feedback. Science 344, 1384–1389 (2014).
Modulation of Macrophage Inflammatory NF‐κB Signaling by Intracellular Cryptococcus
J.B. Hayes1, L.E. Heusinkveld1, R. Leander2, W. Ding2, E.E. McClelland1, D.E. Nelson1; 1Biology, Middle
Tennessee State University, Murfreesboro, TN, 2Mathematics, Middle Tennessee State University,
Murfreesboro, TN
Cryptococcus neoformans (Cn) is an incredibly common facultative intracellular pathogen that is
particularly dangerous to immunocompromised individuals, and is capable of causing life‐threatening
fungal meningitis. However, in healthy individuals, Cn can establish chronic infections that may last
many decades or even a lifetime. In order to do so, Cn has evolved strategies to attenuate and avoid the
host immune system. These involve the exploitation of host macrophages as an intracellular growth
niche and the manipulation NF‐κB signaling, thereby altering macrophage cell fate (i.e. cell cycle and
apoptosis) and the expression of genes involved in the inflammatory response (e.g. TNFα). While
extracellular Cn has been shown to suppress macrophage inflammatory NF‐κB through shed
immunomodulatory capsular polysaccharides that bind Fcγ‐RIIB receptors at the cell surface, it is
currently unclear whether intracellular Cn is also able to influence NF‐κB signaling. To investigate this we
utilized a live cell imaging approach. This enabled us follow individual macrophages after phagocytosis of
Cn, monitoring the changing intracellular Cn burden (number of Cn per macrophage) and the temporal
dynamics of NF‐κB signaling, as reported by the localization of EGFP‐tagged p65 proteins and the
expression of an mCherry reporter of NF‐κB‐dependent gene transcription. Computational models of the
NF‐κB pathway were used to aid the interpretation of single cell data. In agreement with earlier studies,
we found that extracellular glucuronoxylomannan (GXM), the major Cn capsular polysaccharide,
inhibited LPS‐induced nuclear translocation of p65 in RAW 264.7 murine macrophages. In contrast, we
now show that a low intracellular Cn burden (1‐3 Cn per macrophage) delayed but did not inhibit LPS‐
induced p65 nuclear translocation. However, at higher intracellular burdens (>4 Cn per macrophage) Cn
was sufficient to induce stable nuclear localization of p65 (lasting >4 hours) in the absence of additional
pro‐inflammatory stimuli, and was not accompanied by expression from NF‐κB‐responsive promoters
(i.e. TNFα). In conclusion, the distinct effects of Cn at low and high burden suggest the existence of two
different mechanisms through which intracellular Cn alters NF‐κB signaling. The absence of NF‐κB‐
dependent gene expression despite p65 translocation in cells containing a high Cn burden may also
suggest that Cn promotes the formation of transcriptionally repressive NF‐κB species, thereby
suppressing pro‐inflammatory signaling in infected macrophages.
Chromatin modifies the transmission of TNF‐induced NF‐κB signals in single cells.
V.C. Wong1, A.K. Chavali2, R.E. Lee3,4,5, W. Mothes6, S. Gaudet4,5, K. Miller‐Jensen1,2; 1Department of
Molecular, Cellular, and Developmental Biology, Yale University, New Haven, CT, 2Department of
Biomedical Engineering, Yale University, New Haven, CT, 3Department of Computational and Systems
Biology, University of Pittsburgh School of Medicine, Pittsburgh, PA, 4Department of Cancer Biology and
Center for Cancer Systems Biology, Dana‐Farber Cancer Institute, Boston, MA, 5Department of Genetics,
Harvard Medical School, Boston, MA, 6Department of Microbial Pathogenesis, Yale University School of
Medicine, New Haven, CT
Tumor necrosis factor (TNF)‐mediated activation of NF‐κB promotes inflammatory and stress‐responsive
gene transcription. Because nuclear NF‐κB levels exhibit substantial cell‐to‐cell variability, regulatory
mechanisms must be present to maintain precise activation of target genes. Chromatin accessibility and
modifications provide one mechanism by which promoters tune interactions with activating
transcription factors, but it is unclear whether chromatin has a role in buffering against signaling
variability. To examine if chromatin modifies how transcription‐inducing signals are interpreted, we
studied NF‐κB‐mediated activation of HIV LTR‐driven gene expression for LTR promoters integrated into
permissive or restrictive chromatin environments, which exhibit high or low responsiveness to TNF,
respectively (LTRhigh and LTRlow). Using live‐cell microscopy, we tracked nuclear translocation of NF‐κB
and LTR‐driven GFP expression in the same single cells. In a permissive chromatin environment, we
show that GFP expression is strongly correlated with fold‐change in nuclear NF‐κB rather than absolute
nuclear abundance, similar to highly inducible endogenous genes. However, repressive chromatin
significantly weakens this fold‐change sensitivity. To better understand why chromatin weakens
response to the NF‐κB signal, we used RNA FISH to quantify transcriptional heterogeneity following TNF
stimulation. We observe that LTRhigh and LTRlow produce the same average number of transcripts before
and shortly after TNF treatment, but they exhibit significant differences in noise. By fitting the transcript
distributions to a mathematical model of transcription, we find that TNF primarily increases the size of
transcriptional bursts at LTRhigh but increases the frequency of bursts at LTRlow, which results in different
sensitivities to NF‐κB activation. Our results demonstrate that the mechanism by which NF‐κB acts on
the LTR differs depending on chromatin environment, and we observe that these trends are conserved
at other NF‐κB‐target promoters. Taken together, these data suggest that fold‐change sensitivity to
nuclear NF‐κB may be generally applicable to target genes, but reveal that chromatin provides an
additional layer of regulation that modifies information encoded in the NF‐κB signal.
High‐sensitivity measurements of multiple kinase activities in live single cells.
M.W. Covert1, S. Regot1, J. Hughey1, B. Bajar1, S. Carrasco1; 1Bioengineering, Stanford University ,
Stanford, CA
Increasing evidence has shown that population dynamics are qualitatively different from single cell
behaviors. Reporters to probe dynamic, single cell behaviors are desirable, yet relatively scarce. Here we
describe an easy‐to‐implement and generalizable technology to generate reporters of kinase activity for
individual cells. Our technology converts phosphorylation into a nucleocytoplasmic shuttling event that
can be measured by epifluorescence microscopy. Our reporters reproduce kinase activity for multiple
types of kinases in a variety of cell lines, and allow for calculation of active kinase concentrations via a
mathematical model. Using this technology, we made several experimental observations that had
previously been technically unfeasible, including stimulus‐dependent patterns of c‐Jun N‐Terminal
Kinase (JNK) and Nuclear Factor kappa B (NF‐κB) activation. We also measured JNK, p38 and ERK
activities simultaneously, finding that p38 regulates the peak number, but not the intensity, of ERK
fluctuations. Our approach opens the possibility of analyzing a wide range of kinase‐mediated processes
in individual cells.
Minisymposium 23: Motility and Cytoskeleton of Microbes
Gliding mechanism of Mycoplasma, the smallest bacteria.
M. Miyata1; 1Department of Biology, Graduate School of Science, Osaka City University, Osaka, Japan
Mycoplasma is a group of parasitic bacteria lacking the peptidoglycan layer, featured by small genome
and cell sizes. More than ten species of Mycoplasmas including Mycoplasma pneumoniae, a human
pathogen form a membrane protrusion, the gliding machinery at a cell pole, bind host cell surfaces by
this machinery and perform "gliding". This motility is not related to other bacterial motility systems
including flagella or pili etc, or the conventional motor proteins such as myosin, kinesin or dynein. Since
1997, we have been studying about the gliding mechanism of the fastest species, Mycoplasma mobile
which is 0.7 µm long and glides up to 4 µm per s, exerting a force up to 27 pN. We have clarified the
structures and proteins of the gliding machinery, the direct energy source and binding molecules,
analyzed inhibitory antibodies and mutations, and analyzed the mechanical characters, and then
proposed a working "centipede" model, “The movements generated by ATP hydrolysis by the internal
structure transmits to a "leg" protein through a "crank" protein, resulting in the repeated catching,
pulling, and releasing of sialylated oligosaccharides fixed on the surface (Miyata M, Annual Review of
Microbiology, 2010, 64, 519‐537)." Recently, we obtained some experimental results suggesting that
the internal structure of the machinery may be developed from an F‐type rotary ATPase/synthase.
Mechanisms of an acid‐actuated protein lancet.
J.K. Polka1,2, M.D. Vahey3, D.A. Fletcher3, J.M. Kollman4, T.J. Mitchison2, P.A. Silver1,2; 1Wyss Institute for
Biologically Inspired Engineering, Boston, MA, 2Systems Biology, Harvard Medical School, Boston, MA,
Bioengineering, University of California, Berkeley, Berkekey, CA, 4Biochemistry, University of
Washington, Seattle, WA
Moving cargos across cell membranes is a critical challenge for drug and gene delivery, but it is also an
important process in pathogenesis. R bodies are polymers that address this problem with brute force.
They are nucleotide‐independent protein ribbons are driven by pH changes to shift between two states:
1) a compact coil 500nm in diameter, and 2) an extended tube up to 20 microns long. R bodies are
thought to puncture the membranes of phagosomal compartments in Paramecium to deliver a
bacterium‐encoded toxin into the cytoplasm of the predator.
Unlike cytoskeletal filaments that do their work by assembling and disassembling, these actuators are
static in terms of their arrangement of monomers. However, much like shape‐memory alloys, they
undergo dramatic and powerful conformational rearrangements in response to environmental changes.
The shift can occur in a fraction of a second and is fully reversible. Yet, R bodies are assembled from only
two proteins, are cell‐independent, and function in a variety of buffer conditions.
Using recombinantly expressed and purified R bodies, we find that their constituent proteins are driven
by electrostatic forces to undergo dramatic secondary structure changes in response to pH. Using a
visual screen, we have created R body mutants with altered assembly and pH response, thereby
identifying regions crucial for the structural shift. With this information as well as structural data from
site‐specific labeling and cryoelectron microscopy, we propose a mechanical model for R body
actuation. Finally, we show that R bodies are capable of delivering contents across membrane barriers
by the disruptive force of their extension in response to an external trigger.
Integrated systems biology underlying the final steps of bacterial cell growth.
E.R. Rojas1, K.C. Huang2, J.A. Theriot1; 1Biochemistry, Stanford University, Stanford, CA, 2Bioengineering,
Stanford University, Stanford, CA
Cell growth is a complex process that requires coordination between synthesis of all the cell’s
constituents and physical enlargement of the cell. While biosynthetic pathways are well described, the
feedback mechanisms controlling physical enlargement of the cell are poorly understood. We used
microfluidic devices to apply osmotic shock to Bacillus subtilis cells as a means of perturbing their size
during growth, and measured the growth‐response to this perturbation using time‐lapse microscopy.
Interestingly, we observed a complex response to hypoosmotic shock, in which cells exhibited damped
growth oscillations. The period of these oscillations was linearly proportional to the shock magnitude,
with a minimum period of approximately 70 s for small shocks. We discovered that these oscillations
are initiated by underlying oscillations in membrane potential and in the dynamics of MreB, a protein
that coordinates peptidoglycan synthesis. Indeed, the period of growth oscillations could be tuned by
adding potassium exogenously, which affects the dynamics of membrane potential homeostasis.
Furthermore, reduction of membrane tension prior to hypoosmotic shock precluded growth oscillations.
The sum of these data could be united under a single mechanistic, mathematical model wherein excess
membrane tension inhibits peptidoglycan synthesis via membrane depolarization, thereby temporarily
stopping cell growth. We argue that this negative feedback loop is not simply an acute response to
hypoosmotic shock, but an integrated system that regulates the final steps of peptidoglycan expansion,
and therefore cell growth. Although Escherichia coli does not exhibit growth‐oscillations in response to
hypoosmotic shock, both Listeria monocytogenes and Clostridium perfringens behave as B. subtilis,
suggesting that this mechanism of growth regulation may be universal to Gram‐positive rod‐shaped
IFT‐independent translocation of the +TIP protein EB1 in Chlamydomonas flagella.
J.A. Harris1, Y. Liu2, P. Yang2, P. Kner3, K.F. Lechtreck1; 1Cellular Biology, University of Georgia, Athens,
GA, 2Biological Sciences, Marquette University, Milwaukee, WI, 3Engineering, University of Georgia,
Athens, GA
Ciliary assembly requires the translocation of vast amounts of proteins to the distal growing end of the
organelle and intraflagellar transport (IFT) is thought to be the predominant protein transport system in
cilia and flagella. Previously, it was shown that the microtubule plus‐end tracking protein EB1 is present
at the tip of growing, steady‐state, and shrinking cilia and remains at the tip in the absence of IFT
(Pedersen et al. 2003). These observations could be explained by firm binding of EB1 to the ciliary tip.
We used in vivo imaging of EB1‐GFP in C. reinhardtii to address the question of how EB1 moves to and
accumulates at the ciliary tip. In the cell body, EB1‐GFP formed the characteristic transient comet‐like
structures indicating that C. reinhardtii EB1‐GFP functions similarly to EB1 in other systems. After
photobleaching, the EB1‐GFP signal at the ciliary tip recovered within a few minutes indicative for a
constant exchange with unbleached EB1 diffusing in the cilium. Repeated bleaching of the tip or entire
cilium revealed that EB1 rapidly enters cilia from the cell body and that EB1 FRAP at the tip is limited by
the rate at which bleached EB1 leaves the tip rather than by EB1 supply via diffusion. The recovery of
the EB1‐GFP signal at the ciliary tip did not require the IFT pathway and two‐color imaging showed that
EB1 moves without association to IFT; instead, its entry into cilia and translocation to the tip relies on
diffusion. Single molecule imaging showed that EB1‐GFP is highly mobile along the ciliary shaft but
displayed a markedly reduced mobility in the ciliary tip segment interspersed with brief stationary
phases indicative for microtubule binding. Individual EB1 particles dwelled for several seconds at the tip
probably indicating the presence of stable EB1 binding sites along the ciliary tip region. Simulations
showed that these two distinct mobilities are sufficient to explain EB1 accumulation at the ciliary tip.
Our data show that EB1 enters and accumulates in cilia independently of IFT. Diffusion and regulated
binding of proteins at the tip could also explain the rapid capture and release of proteins from cilia as it
occurs during signaling.
Evolutionary retention of two actin nucleation promoting factors, WAVE and WASP, predicts
amoeboid motility in the amphibian chytrid fungus.
L. Fritz‐Laylin1, S. Lord1, R.D. Mullins1; 1Cellular and Molecular Pharmacology, University of California, San
Francisco, San Francisco, CA
Actin‐based ‘amoeboid’ motility is used by unicellular amoebae and white blood cells alike to seek and
engulf bacteria using dynamic protrusions known as pseudopods. Although a few cell types form
pseudopods by “blebbing” (delamination of the cell membrane resulting in bubble‐like protrusions), the
vast majority form pseudopods using branched ‘dendritic’ actin networks built by the Arp2/3 complex,
that is in turn activated by Nucleation Promoting Factors (NPFs). Although widespread among the
eukaryotes, the origin of amoeboid motility remains murky: did it evolve once, or were the underlying
branched actin networks—also used for various other cell processes like endocytosis— co‐opted for
motility multiple times? This uncertainty is in part due to a lack of genomic signatures specific for
amoeboid motility. Here we present the first such genomic marker: only organisms that retain genes
encoding two NPFS, WASP and SCAR/WAVE, form pseudopods. Further, based on the presence of both
WASP and WAVE genes, we predict that Batrachochytrium dendrobatidis (Bd), a chytrid fungus
responsible for massive amphibian death worldwide, is capable of amoeboid motility. We confirm this
prediction by demonstrating that Bd zoospores form Arp2/3‐dependent pseudopods. We also show
that, like the WAVE complex, WASP localizes to dynamic “waves” at the leading edge of migrating
amoeboid neutrophils, and is required for rapid actin assembly, pseudopod formation, and cell motility.
Together, these data indicate that WAVE and WASP proteins play non‐overlapping roles in pseudopod
assembly, and support the hypothesis that ameoboid motility arose once during eukaryotic evolution,
and was subsequently lost in various lineages, including multicellular fungi.
The torsinA homolog tsin is required for the multicellular development of Dictyostelium
C.A. Saunders1, J.R. Erickson1, B.M. Woolums1, H. Bauer1, M.A. Titus1, G. Luxton1; 1Genetics, Cell Biology,
and Development, University of Minnesota, Minneapolis, MN
DYT1 dystonia is a neurological movement disorder characterized by repetitive muscle contractions that
result in involuntary twisting of the extremities and abnormal posturing. DYT1 dystonia is caused by a
mutation within the DYT1/Tor1a gene that deletes a single glutamic acid residue (ΔE302/303 or ΔE) from
the evolutionarily conserved AAA+ protein torsinA that localizes to the shared lumen of the endoplasmic
reticulum and nuclear envelope. The mechanism through which the ΔE mutation causes DYT1 dystonia
is unclear because the basic cellular function of torsinA is unknown. A significant hindrance in our
understanding of torsinA function has been the lack of torsinA homologs in well‐established,
experimentally tractable single‐cell model systems like yeast. Here we report the identification of a
torsinA homolog, TorSIN (tsin), in the social amoeba Dictyostelium discoideum (Dicty). GFP‐tagged wild
type tsin localizes to the endoplasmic reticulum and nuclear envelope. Like torsinA, mutation of the
predicted Walker B motif that is responsible for ATP‐hydrolysis in AAA+ proteins results in accumulation
of GFP‐tsin within the nuclear envelope indicating that potential tsin substrates reside within this
subcellular compartment. Under favorable environmental conditions, dicty exist as free‐living amoeba
that grow and divide by mitosis. Upon starvation, individual amoeba will aggregate by polarizing and
migrating towards a secreted cAMP chemoattractant resulting in the formation of a small mound that
differentiates to form a spore‐containing fruiting body. To test the potential function of tsin during dicty
migration and development, we used homologous recombination to disrupt tsin expression. Tsin‐null
cells exhibit striking defects in during early aggregation, with significantly smaller aggregation territories
than wild type controls. Additionally, while tsin mutant cells can aggregate and eventually form mounds,
they have a defect in the transition out of the mound phase, with only a fraction of mounds developing
into fruiting bodies. Surprisingly, tsin‐null cells migrate faster than wild type cells and are competent for
chemotaxis when cAMP is experimentally provided. Taken together, these results suggest that tsin is
required for signaling propagation during multicellular development in Dicty. Furthermore, they
establish Dicty as a powerful model system for the dissection of the evolutionarily conserved functions
of torsinA.
A Gα stimulated Ras/Rap switch regulates Dictyostelium chemotaxis.
J. Lacal1, Y. Liu2, D.M. Veltman3, I. Keizer‐Gunnink2, F. Fusetti4, P.J. van Haastert2, R.A. Firtel1, A. Kortholt2;
Division of Biological Sciences, University of California, San Diego, CA, 2Department of Cell Biochemistry,
University of Groningen, Groningen, Netherlands, 3Laboratory of Molecular Biology, Medical Research
Council, Cambridge, UK, 4Department of Biochemistry and Netherlands Proteomics Centre, University of
Groningen, Groningen, Netherlands
Chemotaxis, or directional movement towards an extracellular chemical gradient, is an important
property of cells that is mediated through G‐protein coupled receptors (GPCRs) in many cell types. While
many chemotaxis pathways downstream of Gβγ have been identified, few Gα effectors are known. Gα
effectors are of particular interest as they provide the cell from distinguishing signals downstream from
distinct chemoattractant GPCR. Here, we identify a novel RasGEF and RhoGAP containing protein, GEF‐
like protein B (GflB), as a Dictyostelium Gα2 binding partner. GflB localizes to the leading edge and
functions as a Gα stimulated, Rap1‐specific GEF that is important for the balance between Ras and Rap
signaling during chemotaxis. The kinetics of GflB translocation are fine‐tuned by GSK‐3 mediated
phosphorylation. Cells lacking gflB display impaired actin and myosin dynamics, resulting in defective
chemotaxis. Our observations demonstrate that GflB is a central, upstream regulator of
chemoattractant‐mediated cell polarity and cytoskeletal reorganizations by directly linking Gα activation
to monomeric G‐protein signaling.
Mechanism of actin filament assembly by the Vibrio virulence factors VopF and VopL.
T.A. Burke1, E. Kerkhoff2, M.K. Rosen3, R. Dominguez4, D.R. Kovar1,5; 1Molecular Genetics and Cell
Biology, The University of Chicago, Chicago, IL, 2Neurology, University Hospital Regensburg, Regensburg,
Germany, 3Biophysics, Biochemistry, Green Center for Systems Biology, University of Texas
Southwestern Medical Center, Dallas, TX, 4Physiology, University of Pennsylvania Perelman School of
Medicine, Philadelphia, PA, 5Biochemistry and Molecular Biology, The University of Chicago, Chicago, IL
VopF and VopL are actin assembly factors expressed by the infectious bacteria Vibrio cholerae and Vibrio
parahaemolyticus to promote virulence. Unlike other infectious bacteria that secrete proteins to
activate host‐cell nucleation factors such as Listeria (ActA) and Shigella (IcsA) to hijack the host‐cell’s
actin cytoskeleton, VopF and VopL assemble the host cell’s actin directly. VopF and VopL are
homodimeric actin assembly factors, and are categorized into a group of nucleators that rely on multiple
monomeric G‐actin binding WH2 domains for their activity. VopF and VopL have a conserved domain
organization with 72% protein sequence similarity, and while the actin assembly capabilities of VopF and
VopL are undisputed, recent studies have been in conflict over the exact mechanism employed. A
particularly crucial question is whether VopF and VopL associate with the slow growing pointed end or
fast growing barbed end of the actin filament. Given the similarity between the proteins and our
previous studies, we hypothesize that both assembly factors associate with the pointed end of actin
filaments, stimulating rapid cycles of nucleation. To unequivocally answer this question we are using
multicolor evanescent wave fluorescence microscopy imaging to follow individual VopF and VopL
molecules on single actin filaments. F‐actin bound by labeled VopF and VopL do not exhibit a change in
elongation rate from unbound control filaments implying the faster growing barbed end is free.
Experiments where the F‐actin is purposefully bleached show VopF and VopL are always associated with
the dimmer, slower growing end. Conversely, barbed end associated formin mDia2 associates with the
brighter, faster growing end. In agreement with this result, we can visualize filaments simultaneously
bound with mDia2 on one end and either VopF or VopL on the other end. These results strongly indicate
both VopF and VopL only bind to the slower growing pointed end of filaments, stimulating rapid cycles
of nucleation. This agrees with the template filament nucleation model of VopF and VopL wherein the
VCD domain organizes actin monomers provided by the WH2 domains into filament‐like structures.
Latrunculin‐resistant F‐actin and cleavage furrows without myosin II in Chlamydomonas.
M. Onishi1, F.R. Cross2, J.R. Pringle1; 1Department of Genetics, Stanford University School of Medicine,
Stanford, CA, 2The Rockefeller University, New York, NY
Cytokinesis in animals and fungi is achieved by ingression of the cleavage furrow, which involves a ring
containing F‐actin, type‐II myosin, and functionally related proteins. This “contractile actomyosin ring”
has long been thought to provide the force for ingression by a sliding interaction between actin and
myosin filaments. However, although type‐II myosins are found only in fungi, animals, and slime molds,
most other eukaryotes also divide by furrowing. To address this paradox, we are investigating the
mechanisms of cleavage‐furrow ingression in one such eukaryote, the green alga Chlamydomonas
reinhardtii. We have demonstrated the existence of F‐actin in the cleavage furrow of Chlamydomonas
by developing a novel method to overexpress a Lifeact‐Venus probe, which labels only F‐actin. This
observation suggests a myosin‐II‐independent, ancient role for F‐actin in eukaryotic cell division.
However, neither an actin‐null mutation (ida5) or latrunculin drugs disrupted this F‐actin‐like structure,
apparently because of a surprising defense mechanism against a loss of F‐actin that involves the
transcriptional upregulation of many genes including the unconventional actin gene NAP1. Despite
being only ~65% identical to conventional actins, Nap1 appears to be capable of forming filaments;
moreover, these filaments are significantly more resistant to latrunculin than conventional actin
filaments. Possibly as a result of this mechanism, both ida5 and latrunculin‐treated cells continue to
grow at nearly normal rates with no apparent defect in cell division. To search for the pathways
involved in proliferation without conventional F‐actin, we screened for mutants sensitive to latrunculin
and found 40 mutations in four complementation groups, lat1‐lat4. By a novel overlapping‐pool next‐
generation sequencing strategy, the four lat loci were found to encode two kinases (LAT1 and LAT3), one
uncharacterized protein (LAT2), and Nap1 (LAT4). The lat1‐lat3 mutants are defective in the normal
upregulation of NAP1 upon latrunculin treatment, and all lat mutants were rescued by expression of
NAP1 from a constitutive promoter, suggesting a quasi‐linear pathway. lat mutants treated with
latrunculin were immediately blocked for growth and cell division, and lat ida5 double mutants were
inviable with apparent cytokinesis defects, demonstrating for the first time that F‐actin has an essential
role in Chlamydomonas cell viability and proliferation. The further characterization of these mutants
and of furrow formation in the absence of myosin II will be discussed.
MSL8, A Mechanosensitive Ion Channel That Protects Cells From Developmentally Imposed
Osmotic Shock.
E.S. Haswell1, E.S. Hamilton1, G. Maksaev1, G.S. Jensen1; 1Biology, Washington University in Saint Louis,
Saint Louis, MO
How cells adjust to changes in environmental osmolarity is a fundamental question in cell biology, and
much is known about cellular volume control and osmotic shock signaling. Much less is known about
how cells handle osmotic changes that are inherent to normal growth and development, such as the
rehydration of spores, the elaboration of the vasculature or the germination of seeds. One molecular
mechanism for the perception and release of osmotic stress involves mechanosensitive ion channels.
Mechanosensitive ion channels are protein pores in the membrane that allow the transport of ions and
other osmolytes across a membrane in the presence of mechanical force. Numerous families of
mechanosensitive channels have been identified in animal systems (such as Piezo, K2P, TRP, Dec/ENaC)
and two families are found in bacteria (MscL and MscS). Relatively little is known about
mechanosensitive ion channels and osmotic stress in plants, except for a family of plant proteins related
to the Mechanosensitive channel of Small conductance from Escherichia coli (MscS)
Ec‐MscS is a non‐selective mechanosensitive ion channel that is gated directly through tension in the
membrane, and serves as an osmotic release valve during hypoosmotic swelling. Thirty years of site‐
directed mutagenesis, in vivo physiology, crystallography and biophysical modeling experiments on Ec ‐
MscS provide a strong foundation for investigations into MscS homologs, which are found in all three
kingdoms of life. There are ten MscS‐Like (MSL) proteins in the model plant Arabidopsis thaliana, and
we have shown they are required for organelle osmoregulation (Current Biology 22:408‐413, 2012) and
to trigger programmed cell death (Plant Cell 26:3115‐31, 2014). Here we describe a new role for MSL
proteins in pollen, the male gametes of plants. During fertilization, pollen grains must undergo dramatic
changes in cellular water potential in order to successfully deliver the male germline to female gametes.
We used molecular genetics, single channel patch‐clamp electrophysiology, live imaging, and novel in
vitro assays for pollen function to identify and characterize MSL8, a pollen‐specific, membrane tension‐
gated ion channel required for pollen to survive the hypoosmotic shock of rehydration and for full male
fertility. MSL8 negatively regulates pollen germination, but is required for cellular integrity during
germination and tube growth. In conclusion, a plant homolog of the bacterial mechanosensitive ion
channel Ec‐MscS is required to respond to multiple osmotic challenges during pollen hydration and
germination. These data further suggest that homologs of Ec‐MscS have been repurposed in eukaryotes
to sense and respond to mechanical stimuli in a variety of developmental and environmental contexts.
FtsZ minirings curvature is the opposite of tubulin rings.
M. Housman1, M. Osawa1, H.P. Erickson1; 1Cell Biology, Duke Univ, Durham, NC
Bacterial tubulin homolog FtsZ forms a Z ring at the center of the bacterium, which provides the
cytoskeletal framework for cytokinesis,. The Z ring constricts to divide the cell. In vitro FtsZ assembles
straight protofilaments (pfs) that are structural analogs of the pfs that make the microtubule (MT) wall.
These pfs can adopt a curved conformation forming a miniring or spiral tube (tubes are assembled in
DEAE dextran) 24 nm in diameter. Tubulin pfs also have a curved conformation, forming 42 nm tubulin
rings. We have previously provided evidence that FtsZ generates the constriction force by switching
from straight pfs to the curved conformation, generating a bending force on the membrane. The
membrane tether of FtsZ exits from the C terminus of the globular FtsZ, which would have to be on the
outside of the curved pf to meet the concave membrane inside the bacterium. However, it is well
established that tubulin rings have the C terminus, corresponding to the outside of the MT, on the inside
of the ring. Could FtsZ and tubulin rings have the opposite curvature? In the present study we explored
the direction of curvature of FtsZ rings by fusing FN7‐10 to the N or C terminus of the FtsZ globular
domain. FN7‐10 is a rod‐shaped protein segment ~14 nm long. FN‐FtsZ, with the FN on the N terminus,
did not assemble tubes in DEAE dextran. This was expected if the N terminus is on the inside, because
the FN extension is too big to fit in the interior of the tube. FtsZ‐FN assembled normal tubes, however
the FN extensin was not visible in negative stain. We did thin section EM after pelleting the tubes, and
found that the tubes associated in parallel arrays. Tubes of native FtsZ packed together with their dark‐
stained protein walls in contact. Tubes of FtsZ‐FN were surrounded by a pale halo (the FN extensions)
and the walls of the tubes were separated by ~10 nm. Similar results were obtained with N‐ and C‐
terminal fusions of YFP. We conclude that the C‐terminal FN is on the outside of the tubes, and they
interdigitate to form the 10 nm separation. This has interesting implications for the evolution of tubulin.
It seems likely that tubulin began with the curvature of FtsZ, which would have resulted in pfs curving
toward the interior of a disassembling MT. Evolution not only eliminated this undesirable curvature, but
managed to reverse direction to produce the outward curving rings that are useful for pulling
Minisymposium 24: New Insights Into Secretory Trafficking Mechanisms
Visualizing the large procollagen I carrying COPII vesicles by super resolution fluorescence
L. Yuan1, A. Gorur1, S. Baba1, S.J. Kenny2, K. Xu2, R.W. Schekman1; 1MCB, University of California‐
Berkeley, Berkeley, CA, 2Chemistry, University of California‐Berkeley, Berkeley, CA
As an essential step in conventional protein secretion, coat protein complex II (COPII) mediates vesicular
transport from the endoplasmic reticulum (ER) to the Golgi apparatus in all eukaryotes. Mutations in
COPII genes lead to several human genetic diseases that demonstrate the requirement of COPII to
secrete bulky cargos, such as procollagen I (PC1). However, it has not been clear how COPII proteins,
which usually produce budded vesicles approximately 80nm in diameter, participate in the transport of
large cargo such as the 300nm long rigid rod of PC1. We visualized COPII vesicles by stochastic optical
reconstruction microscopy (STORM) and found large PC1‐carrying vesicles that are encaged by COPII
lattices. The PC1‐containing vesicles are physically separated from ER membranes and are devoid of ER
marker proteins. Further, we used a cell‐free reaction to reconstitute the COPII‐mediated budding of
PC1‐containing vesicles from the ER of PC1‐secreting cells. This in vitro reaction requires purified COPII
proteins, nucleotides and cytosol. Structured illumination microscopy (SIM) of vesicles generated during
the cell‐free reaction revealed the ultrastructure of large COPII‐coated vesicles that carry PC1 as well as
of smaller COPII‐coated vesicles devoid of PC1. We conclude that the bulky cargo PC1 is exported out of
the ER in large COPII‐coated vesicles.
Recruitment of ERGIC membranes by TANGO1 and their fusion with the endoplasmic reticulum
is required for export of the bulky cargo Collagen VII.
I. Raote1,2, A.J. Santos1,2, M. Scarpa1,2, N. Brouwers1,2, V. Malhotra1,2; 1Universitat Pompeu Fabra (UPF),
Barcelona, Spain, 2Cell and Developmental Biology, Centre for Genomic Regulation (CRG), Barcelona,
Procollagen VII export from the Endoplasmic Reticulum (ER) requires TANGO1, which interacts with
procollagen VII in the lumen and with cTAGE5 and COPII components on the cytoplasmic face, at an ER
exit site. We have previously shown that the ER‐resident t‐SNARE Syntaxin 18 was required for export of
procollagen VII, indicating that a membrane fusion event into the ER was necessary. We now show the
complete complement of ER t‐SNAREs acting in coordination with Syntaxin‐18 in this fusion.
Identification of the corresponding v‐SNARE, and its localization to ERGIC membranes, provide an
indication of the source of membranes involved in this process. Interestingly, a sub‐population of ERGIC
membranes are recruited to collagen VII‐enriched patches in the ER by a TANGO1‐mediated process.
Here we show that a specific domain of TANGO1, comprising approximately 200 amino acids, when
targeted to mitochondrial membranes, recruited ERGIC membranes. The domain, which we have named
TEER (Tether of ERGIC at the ER), does not recruit COPII coats or Golgi membranes, revealing its specific
affinity for ERGIC membranes. This provides a likely mechanism to ensure specificity for the recruitment
of a subset of ERGIC membranes to a carrier for bulky cargoes like Collagen VII. Based on our new
findings, we suggest that lumenal TANGO1 concentrates collagen VII at an ER exit site and
concomittantly, on the cytoplasmic side, recruits ERGIC membranes via its TEER domain. These ERGIC
membranes fuse with a collagen‐enriched domain of ER and provide membranes for the growth of a
mega carrier to export collagen VII.
COPI selectively drives maturation of the early Golgi.
E. Papanikou1, K.J. Day1, J. Austin II1, B.S. Glick1; 1Molecular Genetics Cell Biology, University of Chicago,
Chicago, IL
COPI coated vesicles bud from Golgi membranes, but the role of COPI in the secretory pathway has been
ambiguous. Previous studies of thermosensitive yeast COPI mutants yielded the surprising conclusion
that COPI was dispensable both for the secretion of certain proteins and for Golgi cisternal maturation.
To revisit these issues, we optimized the anchor‐away method, which allows peripheral membrane
proteins such as COPI to be sequestered rapidly by adding rapamycin. When COPI is inactivated in this
manner, secretion continues, but at variable levels for different proteins. Video fluorescence microscopy
revealed that COPI inactivation causes an early Golgi protein to remain in place while late Golgi proteins
undergo cycles of arrival and departure. These dynamics generate hybrid Golgi structures that remain
partially functional for secretion. Our findings suggest that cisternal maturation involves both COPI‐
dependent recycling of early Golgi proteins and COPI‐independent recycling of late Golgi proteins. We
suggest that a previously unappreciated pathway mediates the intra‐organellar recycling of late Golgi
membrane proteins.
Golgin Tether‐tSNARE Interaction is required for Golgi Stacking.
I. LEE1, J.E. Rothman1; 1Shanghai Institute for Advanced Immunochemical Studies, Shanghaitech
University, Shanghai, China
Golgin tethers and SNARE proteins play essential roles in specificity of docking and membrane fusion of
cargo transport intermediates to target membranes. These proteins have also been implicated in Golgi
cisternal stacking, although exactly how Golgins and SNAREs might contribute to Golgi stacking has
remained elusive. We show here that two Golgin tethers, GM130 and Golgin‐45, directly interact with
Golgi tSNARE, Syntaxin‐5, using their coiled‐coil domains, and that this interaction facilitates cisternal
stacking and inhibits SNARE‐mediated membrane fusion among stacked cisternae. Exogenous
expression of a mutant Golgin‐45 that abrogates its interaction with Syntaxin‐5 induces inter‐cisternal
fusion. Further, this mutant Golgin‐45 fails to rescue Golgi cisternal morphology in GRASPs‐depleted
cells (Lee et al, PNAS, 2014). Strikingly, a small molecule that re‐distributes Syntaxin‐5 from the Golgi
causes disassembly of Golgi stacks within a few hours only in GRASP‐double depleted cells, indicating
that Golgin/SNARE interaction is likely to provide a GRASP‐independent membrane adhesion
mechanism for Golgi stacking. We propose that Golgin‐mediated Syntaxin‐5 sequestration might be a
key regulatory mechanism for inhibition of uncontrolled membrane fusion within the Golgi stacks.
The molecular organization of the exocyst determined by live cell imaging.
A. Picco1,2, O. Gallego3, I. Irastorza4, T. Specht1, D. Devos4, M. Kaksonen1,2; 1Cell Biology and Biophysics,
European Molecular Biology Laboratory (EMBL), Heidelberg, Germany, 2Department of Biochemistry,
University of Geneva, Geneva, Switzerland, 3Institute for Research in Biomedicine (IRB), Barcelona,
Spain, 4Centro Andaluz de Biología del Desarrollo (CABD), Universidad Pablo de Olavide‐CSIC, Sevilla,
During exocytosis the secretory vesicles are tethered to the plasma membrane and subsequently fuse
with it. The exocyst is a hetero‐octameric protein complex that mediates the tethering step and is
essential for the viability of eukaryotic cells. The structure of the exocyst has been extensively
investigated but its molecular organization remained unknown, probably because of the large size and
the flexible nature of the complex.
To determine the molecular organization of the exocyst we designed a hybrid approach that integrates
live cell imaging with the structural information available for each of the exocyst subunits. First, we
engineered yeast cells to express defined anchor platforms on their plasma membrane to which we
recruited the exocyst in controlled orientation. The anchor was labeled with a fluorescent protein to act
as a landmark and the exocyst subunits were tagged one at the time with a different fluorophore at
their N‐ or C‐termini. We imaged yeast cells at their equatorial plane where the recruited exocyst and
the anchor platforms appeared as colocalising fluorescent spots on the plasma membrane. We
measured the centroid positions of the spots and, by observing large numbers of spots, we estimated
the average distance between the centroids of the anchor platform and of each subunit with a precision
of 5 nm or better. These distances positioned the termini of each subunit in respect to the anchor
platform. We then used these distances, which we integrated with structural information about the
subunits, as constraints to reconstruct the molecular organization of the exocyst.
Our results show that the exocyst subunits are rod‐shaped and they use one end of the rod to form the
core of the complex where all the subunits are connected. The other ends of the subunits protrude
outward from the core to mediate vesicle tethering. We also reconstructed the exocyst complex with an
attached vesicle. This reconstruction revealed how the exocyst complex can tether the vesicle to the
plasma membrane without hindering the subsequent fusion step.
Functional interactions among Sec17/alpha‐SNAP, SM proteins, tethers, and SNAREs in
membrane fusion in vitro and in vivo.
A.J. Merz1, A. Guitierrez1, B. Lobingier1, D.P. Nickerson1,2, R. Plemel1, M.L. Schwartz1, M. Zick3;
Biochemistry, University of Washington, Seattle, WA, 2Biology, California State University, San
Bernadino, San Bernadino, CA, 3Biochemistry, Geisel School of Medicine, Hanover, NH
Although SNARE proteins are broadly conserved catalysts of intracellular membrane fusion and
exocytosis, the full sequence of regulatory and catalytic events has not yet been delineated for the full
SNARE cycle in membrane trafficking reaction. We study two SM (Sec1/Munc18 family) proteins, Vps33
and Sly1. Experiments with purified systems, intact organelles in vitro, and genetic manipulations in vivo
show these SM proteins have the properties of true enzymes: (1) SMs capture substrates (unpaired
SNARE proteins) and bring them together in a configuration that facilitates the forward fusion reaction
(trans‐SNARE pairing); (2) SMs limit or prevent off‐pathway reactions including reverse reactions (SNARE
disassembly). In addition we show that SMs and the SNARE disassembly chaperone Sec17/alpha‐SNAP
can physically and functionally interact, in vitro and in vivo.
Fluorescence based analysis of atlastin crossover and fusion kinetics.
J. Winsor1, T.H. Lee1; 1Department of Biological Sciences, Carnegie Mellon University, Pittsburgh, PA
Atlastins are membrane‐anchored GTPases that utilize the energy of nucleotide hydrolysis to catalyze
the homotypic fusion of endoplasmic reticulum (ER) tubules. GTP hydrolysis triggers trans dimer
formation between opposing GTPase heads and a large scale crossover conformational change of the
middle domains predicted to draw opposing membranes into close proximity. However, fusion also
depends on two closely spaced TM helices that anchor the GTPase in the membrane and an amphipathic
helix tail that serves to destabilize the lipid bilayer. The requirements for atlastin‐catalyzed fusion are
agreed upon; however, the actual driving force for membrane fusion is currently debated. For instance,
crossover may power fusion directly; or it may simply serve to tether or draw together opposing
membranes in preparation for fusion, with fusion being powered by the TM helices and tail. To resolve
this debate we have adapted Protein Induced Fluorescence Enhancement (PIFE) to the analysis of
atlastin crossover kinetics. PIFE, a method previously used to monitor protein/DNA interactions, takes
advantage of interactions between an attached fluorophore and its environment to report on local intra
and/or intermolecular interactions in real time. Here we used PIFE to monitor the kinetics of middle
domain crossover in a panel of atlastin mutant variants with variable fusion kinetics. Thus far, our results
reveal a close correlation between crossover kinetics and lipid mixing kinetics, consistent with a role for
crossover in powering atlastin‐catalyzed membrane fusion.
A novel imaging method for quantitative localization of Golgi proteins at nanometer resolution.
H. Tie1, D. Mahajan1, C. Li2, L. Lu1; 1School of Biological Sciences, Nanyang Technological University,
Singapore, Singapore, 2Bioinformatics Institute, Singapore, Singapore
The Golgi complex comprises tightly adjacent membrane cisternae and it is highly controversial
regarding how it works. The dynamic sub‐Golgi localization data of resident proteins or transiting cargos
is importance in resolving current conflicting models and building new ones. However, there is a lack of
method to rapidly and quantitatively achieve nanometer localization accuracy of Golgi proteins. Here,
we present a method called the Golgi localization by imaging center of mass (GLIM) to fill in this gap.
Through GLIM, we can rapidly acquire the localization and axial distribution of a Golgi protein using
conventional light microscopes. GLIM is validated by ~20 Golgi proteins whose cisternal localizations
were known from electron microscopic studies. With GLIM, we demonstrated that it is able to monitor
the intra‐Golgi trafficking of cargos with unprecedented resolution and convenience. Using synchronized
secretory membrane cargos, we directly resolved the sequential transition of cargos from the cis to
trans side of the Golgi, supporting the cisternal progression model. We further found that the efflux of
constitutive cargos is restricted at the Golgi stack and only cargos with sorting signals could target to the
Quality control of GPI‐anchored proteins at the plasma membrane.
P. Satpute‐krishnan1, B.S. Park1,2, J. Lippincott‐Schwartz1; 1Cell Biology and Metabolism Program, NICHD,
National Institutes of Health, Bethesda, MD, 2Thomas Jefferson High School for Science and Technology,
Alexandria, VA
Secretory pathway substrates, including most membrane and soluble proteins, play essential roles on
the cell surface or extracellularly to cull nutrients and to mediate cell‐to‐cell interactions for proper
development and physiology in metazoans. To ensure that only properly folded proteins are presented
at the cell surface and to minimize the secretion of potentially disruptive, aggregation‐prone misfolded
proteins, cells have evolved protein quality control mechanisms to target misfolded proteins for
degradation at their point of origin, in the endoplasmic reticulum (ER). Recently, however, we
discovered that misfolded glycosylphosphatidylinositol‐anchored proteins (GPI‐APs) bypass ER
degradation pathways and instead are constitutively released from the ER to the cell surface via the
secretory pathway. Unlike wild type GPI‐APs that are stably localized to the cell surface and undergo
constitutive recycling between the plasma membrane and recycling endosomes, misfolded GPI‐APs are
rapidly internalized and degraded in lysosomes, implicating the action of a plasma membrane level
quality control pathway. Since GPI‐APs are integrated into the outer leaflet of the plasma membrane,
they provide an experimentally tractable system to study how cells recognize proteins with misfolded
extracellular domains. Although misfolded transmembrane proteins have been shown to be down‐
regulated from the plasma membrane after ubiquitination on their cytosolic regions, GPI‐APs lack a
cytosolic domain, suggesting that GPI‐APs engage a distinct pathway for rapid down‐regulation.
To elucidate how cells target misfolded GPI‐APs on the plasma membrane for lysosomal degradation, we
created a panel of fluorescently‐tagged misfolded variants of diverse GPI‐APs. Using our system for
synchronized trafficking, we monitored them by live cell microscopy combined with biochemical
analysis. Our data suggests that down‐regulation appears to be mediated via a clathrin‐independent and
dynamin‐independent endocytic pathway. We will present our progress in identifying the key endocytic
machinery that direct misfolded GPI‐APs to lysosomes and the role of cellular quality control factors that
associate with misfolded GPI‐APs on the cell surface. GPI‐APs encompass a major class of proteins that
play essential roles in fundamental life processes including embryogenesis, fertilization, neurogenesis,
and immunity and the GPI‐anchor is a post‐translational modification conserved in all eukaryotes. Thus,
we expect that the findings presented here will have broad implications from normal cellular function to
organismal health and disease.
Minisymposium 25: Organelle Homeostasis and Turnover
Dynamic remodeling of the magnetosome membrane is triggered by the initiation of
E. Cornejo1, P. Subramanian2, G.J. Jensen2, A. Komeili1; 1Department of Plant and Microbial Biology,
University of California Berkeley, Berkeley, CA, 2Department of Biology, California Institute of
Technology, Pasadena, CA
Bacterial organelles are diverse in structure and function, but the mechanisms that govern membrane
remodeling and their intracellular organization are not entirely clear. Magnetospirillum magneticum
AMB‐1 is a member of a phylogenetically diverse group of bacteria that synthesize magnetic
nanoparticles inside membrane‐bound organelles called magnetosomes. Magnetosomes organize into a
chain with the help of the bacterial actin, MamK, allowing the bacterium to swim along geomagnetic
fields. Previous studies have identified a number of genes that are required at different stages of
magnetosome formation, but the timing and coordination of these steps are not known. We designed a
genetically inducible system to control magnetosome formation and followed organelle biogenesis,
organization, and magnetic particle synthesis by fluorescence microscopy and measurement of the
magnetic response of the bacterial culture. Using electron cryotomography, we show that clusters of
magnetosomes can align over a long‐range with significantly large gaps that are closed by the bacterial
actin MamK. Additionally, we find that individual magnetosome membranes are capable of growing in
size in the inducible as well as the wildtype strains. Interestingly, the growth of the membrane is
intimately linked to the initiation of biomineralization. Magnetosome membranes that have not
produced a mineral do not grow beyond a diameter of ~50 nanometers (nm). In contrast, in
magnetosomes that harbor crystals, there is a linear relationship between crystal size and membrane
size and magnetosomes can grow as large as ~75 nm. Membrane expansion does not appear to be
directly driven by the growth of the mineral, since magnetosome membranes of the ΔmmsF strain, a
mutant that makes small crystals, are not growth‐limited . Our findings show that the magnetosome is a
dynamic organelle that experiences multiple stages of membrane remodeling linked to the synthesis of
magnetic nanoparticles.
The SND proteins target SRP‐independent substrates to the endoplasmic reticulum.
N. Aviram1, T. Ast1, S. Haßdenteufel2, E.A. Costa3, E.C. Arakel4, S. Schorr2, S.G. Chuartzman1, C.H. Jan3, B.
Schwappach4, R. Zimmermann2, J.S. Weissman3, M. Schuldiner1; 1Department of Molecular Genetics,
Weizmann Institute of Science, Rehovot, Israel, 2Department of Medical Biochemistry and Molecular
Biology, Saarland University, Homburg, Germany, 3Department of Cellular and Molecular Pharmacology,
University of California, San Francisco and Howard Hughes Medical Institute, San Francisco, CA,
Department of Molecular Biology, University Medical Center Göttingen, Göttingen, Germany
Targeting to the endoplasmic reticulum (ER) is a critical step in the biogenesis of thousands of
endomembrane proteins. Mounting evidence indicates that the Signal Recognition Particle (SRP)
pathway cannot target all endomembrane proteins, stipulating the existence of at least one additional
relay mechanism. To identify new targeting factors we conducted a systematic high‐content screen in
yeast and revealed three uncharacterized proteins, which we named SND (Srp‐iNDependent), that are
needed for the correct targeting of SRP‐independent substrates both in‐vivo and in‐vitro . Since at least
one SND protein has a clear human homologue our results suggest that SRP‐independent targeting is
evolutionarily conserved. We uncover that the SND proteins cater to substrates that have downstream
transmembrane domains by using proximity specific ribosome profiling and substrate reengineering
approaches. Our results highlight the missing link for endomembrane protein targeting and map the
complex interplay between pathways that makes ER import a robust cellular phenomenon.
Endogenous Parkin Preserves Dopaminergic Substantia Nigral Neurons following Mitochondrial
DNA Mutagenic Stress.
A.M. Pickrell1, S.R. Kennedy2, C. Huang1, A. Ordureau3, D.P. Sideris1, J. Hoekstra2, J.W. Harper3, R.J.
Youle1; 1National Institutes of Neurological Disorders and Stroke, National Institutes of Health, Bethesda,
MD, 2Department of Pathology, University of Washington, Seattle, WA, 3Department of Cell Biology,
Harvard Medical School, Boston, MA
Parkinson's disease (PD) is a neurodegenerative disease caused by the loss of dopaminergic neurons in
the substantia nigra. PARK2 mutations cause early‐onset forms of PD. PARK2 encodes an E3 ubiquitin
ligase, Parkin. During mitochondrial depolarization, PINK1, a serine/threonine‐protein kinase, is
stabilized on mitochondria. PINK1 phosphorylates both ubiquitin and Parkin at serine 65 leading to
Parkin’s activation and retention on damaged mitochondria. These phosphorylation events trigger the
translocation of Parkin specifically to damaged mitochondria causing ubiquitylation of protein targets on
the outer mitochondrial membrane ultimately inducing the autophagic removal of the damaged
mitochondria. However, Parkin knockout (KO) mice do not display signs of neurodegeneration. To assess
Parkin function in vivo, we utilized a mouse model that accumulates dysfunctional mitochondria caused
by an accelerated generation of mtDNA mutations (Mutator mice). In the absence of Parkin,
dopaminergic neurons in Mutator mice degenerated causing an L‐DOPA reversible motor deficit. Other
neuronal populations were unaffected. Phosphorylated ubiquitin at serine 65 was increased in the
brains of Mutator mice indicating PINK1‐Parkin activation. Parkin loss caused mitochondrial dysfunction
and affected the pathogenicity but not the levels of mtDNA somatic mutations. A systemic loss of Parkin
synergizes with mitochondrial dysfunction causing dopaminergic neuron death modeling PD pathogenic
Pathogenic LRRK2 Impairs Miro Degradation and Mitochondrial Transport in a Pathway also
Present in Sporadic Parkinson’s.
C. Hsieh1, A. Shaltouki1, X. Wang1; 1Neurosurgery, Stanford University, Palo alto, CA
Mutations in LRRK2 are the most frequent cause of Parkinson’s disease (PD). However, how pathogenic
LRRK2 causes neurodegeneration remains elusive. Miro is an outer mitochondrial membrane protein
controlling mitochondrial motility. Here we report that LRRK2 interacts with Miro upon mitochondrial
depolarization to remove Miro from damaged mitochondria and to arrest mitochondrial motility. The
pathogenic G2019S mutation in LRRK2 abolishes this binding to Miro, causing failure to remove Miro, to
arrest damaged mitochondrial movement, and to initiate mitophagy. These functional deficits are
present in multiple independent disease models, including induced pluripotent stem cell (iPSC)‐derived
neurons and skin fibroblasts from familial PD patients. Furthermore, partial reduction of Miro in
LRRK2G2019S iPSC‐derived neurons restores mitophagy and alleviates neurodegeneration. We also
identify impairments in Miro degradation in sporadic PD patients. Thus, failure to degrade Miro and the
downstream cellular consequences that ensue may constitute a central component of PD pathogenesis.
Axonal autophagosomes recruit dynein for retrograde transport through fusion with late
X. Cheng1, B. Zhou1, M. Lin1, Z. Sheng1; 1NINDS, NIH, Bethesda, MD
Autophagy is an important cellular process to eliminate defective organelles and aggregated proteins
over a neuron’s lifetime. Autophagy undergoes stepwise maturation: bulk cytoplasmic components and
organelles are engulfed within double‐membrane organelles termed autophagosomes, followed by
fusion with late endosomes (LEs) into hybrid organelles called amphisomes, or fusion with lysosomes
into autolysosomes fordegradation. Efficient degradation of autophagic vacuoles (AVs) via lysosomes is
critical for cellular homeostasis. Neurons are highly polarized cells with long axons, thus facing special
challenges to transport AVs generated at distal processes toward the soma where mature lysosomes are
relatively enriched. Although dynein‐driven retrograde transport of AVs has been suggested, a
fundamental question remains: How do autophagosomes generated at distal axons acquire dynein for
retrograde transport to the soma. Addressing this issue will advance our understanding of several
neurodegenerative diseases associated with autophagic stress in distal axons and at synapses. Our
previous study reveals that snapin acts as an adaptor recruiting dynein motors to LEs and plays a key
role in coordinating dynein‐driven LE retrograde transport (Cai et al., Neuron 2010).
Here, we reveal a new mechanism underlying recruitment of dynein motors to amphisomes, thus driving
AV retrograde transport toward the soma. (Cheng et al., JCB 2015). We demonstrate that LE‐loaded
dynein‐snapin complexes drive AV retrograde transport in axons upon fusion of autophagosomes with
LEs into amphisomes. Blocking this fusion event with syntaxin17 knockdown reduced recruitment of
dynein motors to AVs, thus immobilizing them in axons. Deficiency in dynein‐snapin coupling impaired
AV transport, resulting in AV accumulation in neurites and synaptic terminals. Intriguingly, we captured
dynamic de novo autophagosomal biogenesis, fusion with LEs, and retrograde transport in growth cones
and distal axon shafts of live DRG neurons immediately after starvation. This trafficking route is crucial
for neurons to maintain effective autophagic clearance through lysosomal degradation in the soma.
Thus, our study reveals, for the first time, the motor‐adaptor sharing mechanism that allows 2 different
organelles to take the “ride‐on service” for long‐distance trafficking by forming hybrid intermediate
organelles. This mechanism enables neurons to efficiently remove distal AVs engulfing misfolded mutant
proteins and dysfunctional organelles associated with several major neurodegenerative diseases.
(Supported by the Intramural Research Program of NINDS, NIH)
Cheng, X. et al. Axonal autophagosomes recruit dynein for retrograde transport through fusion with late
endosomes. Journal of Cell Biology 209, 377‐386.
Role of Fis1 in mitochondrial fission and stress.
A.M. Van Der Bliek1, R. Youle2; 1Biological Chemistry, David Geffen School of Medicine at UCLA, Los
Angeles, CA, 2National Institutes of Neurological Disorders and Stroke, National Institutes of Health,
Bethesda, MD
Functional studies of the mitochondrial outer membrane protein Fis1 uncovered a novel regulatory
pathway with unexpected effects on mitophagy. The main mitochondrial fission protein in metazoans,
Drp1, is recruited to mitochondria by the outer membrane proteins Mff, Mid51 and Mid49. Fis1 was
previously also thought to be a recruitment factor for Drp1, but Fis1 ‐/‐ cells have only mild or no fission
defects. Our results show that Fis1 is still part of the mitochondrial fission complex in metazoan cells,
but it most strongly affects downstream events, helping to guide the selective elimination of damaged
mitochondria during mitophagy. Drp1 first binds to Mff on the surface of mitochondria and then enters
into a complex that includes Fis1 and ER proteins at the ER‐mitochondrial interface. Mutations in Fis1 do
not normally affect fission but they can disrupt downstream degradation events when specific
mitochondrial toxins are used to induce fission. The disruptions caused by mutations in Fis1 lead to an
accumulation of large LC3 aggregates. Fis1 therefore acts in sequence with Mff at the ER‐mitochondrial
interface to couple stress‐induced mitochondrial fission with downstream degradation processes (Shen
et al. 2014, MBoC 25, 145‐59). An important clue to the downstream functions of Fis1 came from the
discovery of binding interactions between the Fis1 and the Rab7 GAP TBC1D15 by the Ishihara lab
(Onoue et al. 2013, J. Cell Sci. 126, 176‐185). Our collaborators in the lab of Richard Youle (NIH) showed
that Knock out of the TBC1D15 gene gave rise to the same types of LC3 aggregates that were observed
in Fis1 ‐/‐ cells. Additional experiments showed that TBC1D15 inhibits Rab7 activity and associates with
both the mitochondria through binding Fis1 and the isolation membrane through the interactions with
LC3/GABARAP family members. A homologue of TBC1D15, TBC1D17, also participates in mitophagy and
forms homodimers and heterodimers with TBC1D15. These results show that TBC1D15 and TBC1D17
direct the autophagic encapsulation of mitochondria by regulating Rab7 activity at the interface
between mitochondria and isolation membranes (Yamano et al. 2014, eLife 3, e01612). The requirement
for Fis1 is triggered by fission inducing conditions that go hand in hand with ROS production, such as
treatments with Antimycin A or Paraquat. Use of Jnk inhibitors and siRNA for the mitochondrial Jnk
binding partner SAB suppress the LC3 aggregates in Fis1 ‐/‐ cells, suggesting that the Jnk kinase signaling
pathway modifies the fission machinery. We are currently testing possible Jnk kinase targets involved in
this specific process and we are looking for upstream regulators and downstream effectors in the
Synchronous regulation of nuclear and mitochondrial translation during mitochondrial
M. Couvillion1, S. Churchman1; 1Department of Genetics, Harvard Medical School, Boston, MA
The synthesis and assembly of oxidative phosphorylation (OXPHOS) complexes pose a unique challenge
to the cell, because their subunits are encoded on two different genomes, the nuclear genome (nDNA)
and the mitochondrial genome (mtDNA). As eukaryotic cells evolved, mtDNA diverged dramatically from
its prokaryotic counterpart, and is expressed by a mitochondrial‐specific single‐subunit RNA polymerase
and a dedicated ribosome. Due to challenges in monitoring gene expression across compartments, it
remains unknown whether nDNA‐ and mtDNA‐encoded OXPHOS genes coordinate their expression, and
if they do, how it is achieved. Through adapting and applying quantitative genomic approaches to the
mitochondrial and nuclear genomes, we globally observed distinct stages of gene expression during
mitochondrial biogenesis in Saccharomyces cerevisiae. Interestingly, mRNA levels of mtDNA‐ and nDNA‐
encoded OXPHOS subunits do not increase concordantly, as is typically the case for cellular protein
complexes. We found that synthesis of mtDNA‐encoded OXPHOS subunits within the mitochondrial
matrix is dynamically regulated during mitochondrial biogenesis, with the translation efficiency of
cytochrome c oxidase subunits increasing early accompanied by suppressed translation efficiency of ATP
synthase subunits. Strikingly, nDNA‐encoded OXPHOS subunits are translationally regulated in a similar
pattern, indicating that the mitochondrial and nuclear genomes coordinate their expression during
translation. Finally, we show that the mitochondrial translation response requires cytoplasmic
translation, suggesting that the coordination is actively maintained. Despite orthogonal gene expression
machinery and genomes of distinct origin and subcellular localization, the eukaryotic cell remarkably
synchronizes translation regulation across compartments to promote efficient synthesis of OXPHOS
A cellular molecular timer measures synaptic vesicle use and prevents the participation of aged
vesicles in synaptic transmission.
S. Truckenbrodt1, A. Viplav2, A. Denker3, A. Vogts4, E.F. Fornasiero1, S.O. Rizzoli1; 1Department for Neuro‐
and Sensory Physiology, University of Göttingen Medical School, Göttingen, Germany, 2Cells in Motion
Cluster of Excellence, University of Münster, Münster, Germany, 3Hubrecht Institute, Utrecht,
Netherlands, 4Leibniz‐Institute for Baltic Sea Research, Rostock, Germany
Old organelles can become a hazard to cellular function, due to the accumulation of damaged
molecules. Mechanisms that identify old organelles and prevent them from participating in cellular
reactions are therefore necessary. The prevailing assumption is that organelles are functionally
employed in cells until degradation via damage response mechanisms. We have identified an additional
mechanism that measures the use of synaptic vesicles, rather than damage, and functionally inactivates
them long before degradation. Live‐cell immunocytochemistry allowed us to track synaptic vesicles for
up to 10 days in hippocampal cultures. We determined that synaptic vesicles recycle ~300 times over 12‐
24 hours, before entering a state of functional inactivity for 1‐2 days, ultimately followed by
degradation. We performed metabolic imaging, using nanoSIMS and FUNCAT, and found that newly
synthesized synaptic vesicles are preferentially employed in neurotransmitter release. We probed the
changes in molecular composition of synaptic vesicles over time, using correlative two‐color super‐
resolution STED microscopy, and found that they become contaminated with certain molecules from the
plasma membrane during ageing, which interfere with vital components of the synaptic vesicle release
machinery. This renders the old vesicles less competent to release than their newly synthesized
counterparts and leads to their inactivation. This process can be accelerated by increasing the frequency
of neurotransmitter release, suggesting that the mechanism we identified directly measures the number
of times a vesicle has been used, rather than its temporal age. When depleting the supply of newly
produced vesicles or applying supra‐physiological stimulation to force the release of aged synaptic
vesicles, they perform more poorly than their young counterparts and display a loss of molecular
cohesion during recycling. The old vesicles are eventually targeted for degradation, but their functional
inactivation precedes degradation by several days. We conclude that we have identified a molecular
timing mechanism, evolved to ensure that old vesicles are not used in neurotransmitter release.
Synaptic transmission is presumably too sensitive to tolerate progressive accumulation of damage in
synaptic vesicles that are still in use. They need to be culled from neurotransmission even before they
accumulate enough damage to be recognized by classical damage response mechanisms, necessitating
this preventive functional inactivation to avoid the use of even slightly damaged organelles.
Dissecting the mechanisms of liquid to solid phase transition of the ALS protein FUS.
A. Patel1, L. Jawerth1, T.M. Franzmann1, R. Wheeler1, A.A. Hyman1; 1Hyman Lab, Max‐Planck‐Institute of
Cell Biology and Genetics, Dresden, Germany
FUS/TLS is a prion‐like protein that contains intrinsically disordered domains and is associated with
neurodegenerative disease. We recently showed that intracellular FUS/TLS compartments form under
various cellular conditions and that these compartments exhibit liquid‐like properties in vivo and in
vitro. “Aging” experiments revealed that FUS/TLS liquid droplets undergo a phase transition to a solid‐
like state which is accelerated by disease mutations (Patel, Lee et al. 2015). We discovered that
concentrating proteins by phase separation comes with the trade‐off that can also promote protein
aggregation. Solid‐like aggregates of prion‐like proteins are a hallmark of many aging‐associated
diseases. Aberrant phase transitions might be one trigger causing aging‐associated diseases. However,
the molecular mechanisms underlying this aberrant phase transition and the strategies cells have
developed to sustain the function of these aggregation‐prone proteins remain largely enigmatic. Here,
we present recent advances we made to dissect out the mechanisms of liquid‐liquid and liquid‐solid
phase transitions by combining a wide range of biochemical, biophysical and cell biology techniques. We
find that electrolytes, small compounds and protein interactors affect the liquid‐liquid, as well as liquid‐
solid transitions. Insights gained from studying liquid‐solid phase transition might help us developing
drugs targeted to treat age‐associated diseases. References:Patel, A., et al. (2015). "A liquid to solid
phase transition of the ALS protein FUS correlates with disease." Cell in press.
Minisymposium 26: Tissue Biology
Mechanisms underlying the age‐related decline in the regenerative capacity of Drosophila
imaginal discs.
R.E. Harris1, L. Setiawan1, J. Saul1, I.K. Hariharan1; 1Molecular and Cell Biology, University of California,
Berkeley, Berkeley, CA
For many organisms, the ability to regenerate damaged tissues diminishes with increasing maturity. The
mechanisms responsible are not known, but their identification is essential for the design of strategies
to promote tissue regeneration in adult organisms. Drosophila imaginal discs are larval precursors to
adult tissues that, when damaged in situ, are able to regenerate. However this ability diminishes
dramatically during the final larval stage. This decline in regenerative capacity correlates with a
reduction in the ability to upregulate a number of genes associated with regeneration, including
wingless (wg, Drosophila WNT1). Since WNT proteins are upregulated following tissue damage in
diverse taxa, the mechanisms that regulate WNT expression following tissue damage are likely to be
evolutionarily conserved. To understand how damage‐induced wg expression is activated, we have
characterized an enhancer located in the WNT cluster that regulates the expression of two flanking WNT
genes; wg and Wnt6, during regeneration. Deletion of this enhancer permits normal development, but
compromises regeneration. Dissection of the genomic region reveals a bipartite structure that includes a
damage‐responsive module and a silencing element. The damage responsive module is robustly
activated following various damaging stimuli, and importantly its activity is undiminished with increasing
maturity. In contrast, the adjacent silencer element has no enhancer activity on its own, but can
attenuate expression mediated by the damage‐responsive module in cis, and moreover, is responsible
for the age‐related decline of expression in older larvae. Thus the signaling mechanisms that mediate
damage‐responsive WNT expression are intact in older larvae, but are actively repressed by a silencing
mechanism. We show this silencing mechanism is epigenetic in nature, and furthermore we have used
genome editing to circumvent this silencing, to restore regenerative WNT signaling in older larvae.
Finally, targeting Myc expression specifically to the regeneration blastema can improve regenerative
ability even in mature larvae, thus demonstrating a method by which the age‐dependent decline in
regenerative capacity can be overcome.
Polarization of junctions, cytoskeleton and signal receptors in multi‐layered epithelia.
C.M. Niessen1,2,3, M. Rübsam1,2,3, S. Vorhagen1,2,3, F. Tellkamp1,2,3, J. Xia1,2,3, J. Nafizi1,2,3, B. Boggetti1,2,3;
Center for Molecular Medicine Cologne, University of Cologne, Cologne, Germany, 2Cologne Excellence
Cluster on Cellular Stress Responses in Aging‐Associated Diseases (CECAD), University of Cologne,
Cologne, Germany, 3Department of Dermatology, University of Cologne, Cologne, Germany
Intercellular adhesion and cell polarity are crucial determinants of tissue architecture and
morphogenesis. We ask through which mechanisms intercellular adhesion and polarity regulate the
establishment and maintenance of self‐renewing, multi‐layered epithelia, such as the skin epidermis. In
contrast to apico‐basolateral polarity found in simple epithelia, multi‐layered epithelia polarize not
within a single cell but across layers. How this is controlled and whether this involves cadherins and
polarity proteins such as atypical protein kinase C is not known. Using mouse transgenics we have
identified a crucial role for classical cadherins in epidermal cohesion and barrier function and showed
that they serve as important regulators of mammalian aPKCs. Whole mount analysis revealed that E‐
cadherin coordinates the polarized organization of junctions, the cytoskeleton and signal receptors
across layers to control differential mechano‐transduction. This polarization of tension across layers is
crucial for epidermal barrier formation and function. Our data also identify mammalian aPKCs as key
regulators of cell fate decisions, likely by coupling fate determinant signaling to spindle orientation.
Inactivation of epidermal aPKCλ leads to a gradual loss of epidermal stem cells and premature skin
aging. Using SILAC and proteomics we have identified a range of binding partners and novel substrates
for mammalian aPKCs and at present we are analyzing their relevance in cell fate regulation and
polarized barrier function. Overall, our data suggest that regulation of cell architecture is crucial for
epidermal stem cell behavior, morphogenesis and homeostasis.
A novel, noncanonical BMP pathway modulates synapse maturation at the Drosophila
neuromuscular junction.
M.J. Sulkowski1, T. Han1, C.M. Ott1, E.M. Verheyen2, J. Lippincott‐Schwartz1, M. Serpe1; 1NICHD, NIH,
Bethesda, MD, 2Department of Molecular Biology and Biochemistry, Simon Fraser University, Burnaby,
Synapse activity and synapse development are intimately linked, but our understanding of the coupling
mechanisms remains limited. At the Drosophila NMJ, BMP signaling is critical for synapse growth and
homeostasis. BMP signaling to motor neurons triggers a canonical pathway ‐ which modulates
transcription of BMP target genes and promotes NMJ growth, and a noncanonical pathway ‐ which
connects local BMP/BMP receptor complexes with the cytoskeleton and ensures synapse stability. Here
we describe a novel noncanonical BMP pathway, which is genetically distinguishable from all other
known BMP signaling cascades. This pathway does not contribute to NMJ growth, and instead influences
synapse formation and maturation in an activity‐dependent manner. Specifically, phosphorylated Smad
(pMad in flies), the BMP pathway effector, accumulates at active zone in response to active postsynaptic
type‐A glutamate receptors, a specific receptor subtype. In turn, synaptic pMad functions to promote
the recruitment of type‐A receptors at synaptic sites. This positive feedback loop provides a molecular
switch controlling which flavor of glutamate receptors will be stabilized at synaptic locations as a
function of synapse status. Since BMPs also control NMJ growth and stability, BMP signaling offers an
exquisite means to monitor the status of synapse activity and coordinate NMJ growth with synapse
maturation and stabilization.
Discovery of a hormone requirement for cell growth allows visualization of growth dynamics in
a proliferating epithelial tissue explant.
N.A. Dye1, S. Eaton1; 1MPI‐CBG, Dresden, Germany
We are using the Drosophila larval wing to study how morphogen gradients, mechanics and systemically
circulating hormones regulate the amount and orientation of cell proliferation in an epithelial tissue.
Spatially homogeneous proliferation in the wing disc is under the control of three signaling pathways
(Dpp, Wingless, and Fat/Daschous) with graded distributions of activity throughout the tissue. To learn
how these gradients regulate growth and tissue mechanics, we aim to quantify dynamic cell behaviors
during tissue growth in wild type and mutant genotypes and mechanically perturb the tissue. Toward
this aim, we needed to first to improve methods for live imaging of this tissue, as previously established
methods support only limited growth. We discovered that low levels of the steroid hormone ecdysone
are sufficient to support long‐term proliferation of wing discs in culture, extending the observable
window of growth 2‐3x longer than previously described methods using insulin. We used spinning disc
microscopy to image discs expressing E‐Cadherin‐GFP at cellular resolution and found that those
cultured with ecdysone exhibit a tissue growth pattern that is consistent with in vivo clonal analysis. We
then quantified the contributions of cell divisions, rearrangements, extrusions, and cell shape changes to
tissue growth and found that cell divisions are not the only process underlying oriented tissue growth,
contrary to what has been inferred from analysis of fixed tissues. Indeed, oriented cell rearrangements
and cell shape changes make quantitatively similar contributions. Interestingly, discs cultured in the
absence of ecdysone but with insulin do not exhibit properly oriented growth, indicating that the lack of
ecdysone may disrupt the signaling gradients that regulate the growth pattern. We demonstrate that
ecdysone is required not just in culture but also in vivo . Genetically blocking ecdysone production by the
brain during the third instar inhibits proliferation in the wing disc and reduces morphogen gradient
production and signaling. The effect is mimicked by genetically perturbing the ecdysone receptor in the
wing disc, indicating that ecydsone acts directly on the wing. To gain insight into the mechanism by
which ecdysone promotes growth, we sequenced the transcriptome of discs cultured in ecdysone or
insulin. Our data suggest that ecdysone (but not insulin) regulates numerous genes with functional
classifications related to cell growth and tissue development, including genes with linked to the Wg,
Notch and Dpp pathways. Our findings identify a mechanism for tissue growth regulation that helps to
integrate local cues with hormonal signals and also help overcome a long‐standing technical barrier to
studying growth in this model system.
Emergent mechanisms of collective cell durotaxis.
R. Sunyer1, V. Conte1, J. Escribano2, J.M. García Aznar2, J. Muñoz3, P. Roca‐Cusachs1,4, X. Trepat1,4;
Institute for Bioengineering of Catalonia, Barcelona, Spain, 2University of Zaragoza, Zaragoza, Spain,
Polytechnic University of Catalonia, Barcelona, Spain, 4University of Barcelona, Barcelona, Spain
The ability of single cells to follow gradients of extracellular matrix stiffness ‐durotaxis‐ has been
implicated in development, fibrosis and cancer. However, this type of directed migration is not universal
across cell types and its efficiency is poor. In contrast with the case of single cells, here we show that
collectively migrating cell clusters exhibit efficient durotaxis, a phenomenon that is emergent and
appears to be universal. We provide the scaling laws of collective durotaxis as a function of the size of
the cluster and the shape of the gradient. Collective durotaxis requires the action of myosin motors and
the integrity of cell‐cell junctions. By extending traction microscopy and monolayer stress microscopy to
substrates of arbitrary stiffness profiles, we demonstrate force transmission across the full length of the
cell clusters. Finally, we show that that the emergence of collective durotaxis is naturally captured by a
generalized clutch‐model in which local cell‐matrix dynamics are coupled though cell‐cell junctions.
Overall, our study reveals that coordinated cluster dynamics enhances the capacity of cells to sense and
respond to mechanical cues.
A novel bidirectional signaling pathway regulates collective cell migration in the Drosophila egg
K. Barlan1, M. Cetera1, S. Horne‐Badovinac1; 1Molecular Genetics Cell Biology, University of Chicago,
Chicago, IL
Collective cell migration is critical to the formation of diverse tissues and organs during animal
development. Collectively migrating cells use the same migration machinery as individual cells. What
sets collective cell migration apart is that each cell must both react to and influence the behavior of its
neighbors through direct cell‐cell signaling. The necessity for communication among collectively
migrating cells is perhaps best exemplified by the migration of an epithelial sheet, where leading edge
actin‐based protrusions of one cell extend along the extracellular matrix (ECM) directly beneath the
trailing edges of the cells in front. It is essential, therefore, that trailing edge retraction in one cell be
coordinated with protrusion formation in the cell(s) behind so that cells don’t compete to bind the same
region of ECM. How this high level of organization is achieved is unknown. In the Drosophila egg
chamber, follicular epithelial cells migrate collectively as a sheet; this migration causes the entire egg
chamber to rotate, which helps to produce the elongated shape of the egg. Using this model of
collective cell migration, we have identified a novel, bidirectional signaling system that operates at the
basal epithelial surface to coordinate leading and trailing edge behaviors between neighboring follicle
cells. We have found that the Fat2 cadherin, which localizes to each cell’s trailing edge, signals to the
cell(s) behind to induce their leading edge protrusions. Conversely, the receptor tyrosine phosphatase
Lar, which appears to localize to each cell’s leading edge, signals to the cell(s) in front to stimulate their
trailing edges to retract. Fat2 and Lar also have cell‐autonomous functions that complement their
signaling functions. Fat2 is required in the cell where it is expressed for trailing edge retraction, while Lar
is required in the cell where it is expressed for protrusion formation. These complementary localization
patterns and phenotypes suggest that Fat2 and Lar may form a signaling complex across cell‐cell
boundaries. In support of this notion, we find that Fat2 and Lar often colocalize in punctae at the basal
surface. Further, Fat2 functions non‐cell‐autonomously to localize Lar, and Fat2’s extracellular domain is
sufficient to recruit Lar. Together these data suggest a model in which Fat2 and Lar work together to
promote epithelial motility by coordinating leading edge and trailing edge behaviors between
neighboring cells.
Vertebrate embryos in suspended animation ‐ Characterizing cellular and molecular
mechanisms of diapause which put embryo development on hold to survive through adverse
C. Hu1, A. Brunet1; 1Genetics, Stanford University, Stanford, CA
Organisms experience seasonal changes in the wild, and extreme conditions give rise to extreme
adaptions. Diapause is a unique surviving approach by which embryos put all developmental processes
in temporary arrest to live through adverse environment. Similar dormancies are widely employed from
plants, insects, to mammals. While significant knowledge has been acquired from studying diapause in
invertebrates such as fruit flies and nematodes, still little is understood about the underlying
mechanisms governing diapause in a more complex vertebrate system. Among vertebrates, African
killifish Nothobranchius furzeri (N. furzeri) possess a rare ability to enter diapause at the mid‐stage of
embryogenesis, where embryos has already equipped with well‐developed neural, muscle, and
circulation systems. N. furzeri naturally lives in ephemeral ponds, with newly laid embryos preparing for
the coming drought by entering diapause. This diapause might last months or sometime even years until
another rain season refills the ponds with water. During diapause, embryos show higher tolerance to
various stresses, and the time stayed in diapause does not affect the lifespan afterward, suggesting that
cells in diapause are protected from deleterious effects of both environment and senescence. These
unique protective features distinguish cells in diapause from regular cell cycle arrest. Although N. furzeri
embryos enter diapause by obligation, this fate can be manipulated by maternal effects on selected
embryos to skip diapause. Rather than at a specific developmental point, embryonic cells enter diapause
within a window around the end of somitogenesis, indicating that diapause is not a hard‐wired program
essential for embryo development. At cellular level, cell proliferation is completely ceased once
diapause is initiated, with circulation system and maternal hormone potentially involved to coordinate
cell activities and proliferation within embryos. At molecular level, diapause cells are epigenomically
reprogrammed to selectively express functional diapause genes while keep global gene expression in
quiescence. A thorough RNA‐Seq profiling was conducted to compare embryos at specific stages before,
during, and after diapause, and to identify candidate genes and pathways responsible for unique
features of diapause. Together, this study provides a unique opportunity to understand how suspended
animation works in complex vertebrate systems, how distinct cell types under a robust embryogenesis
process can all be halted simultaneously, and importantly, for an extensive length of time without
suffering the harmful consequences from various stresses and senescence.
Aberrant cell segregation driven by ephrin‐B1 mosaicism relies on Eph/ROCK signaling and
involves changes in actin polymerization.
A.K. ONeill1,2, A.R. Larson1,2,3, T.K. Niethamer1,2,4, J.O. Bush1,2,5; 1Program in Craniofacial Biology,
University of California San Francisco, San Francisco, CA, 2Cell and Tissue Biology, University of California
San Francisco, San Francisco, CA, 3School of Medicine, University of California San Francisco, San
Francisco, CA, 4Biomedical Sciences Graduate Program, University of California San Francisco, San
Francisco, CA, 5Institute for Human Genetics, University of California San Francisco, San Francisco, CA
Proper cell segregation is critical for many aspects of development. The Eph/ephrin family of signaling
proteins mediate cell segregation in many developmental contexts via complex interactions between
Eph‐ and ephrin‐expressing cells that lead to the formation of distinct domains. Ephs and ephrins are
unusual in that they utilize both forward signaling (into the Eph‐expressing cells) and reverse signaling
(into the ephrin‐expressing cells); this bidirectional signaling has been proposed to mediate Eph/ephrin‐
driven cell segregation. Aberrant cell segregation also results in congenital disease. Craniofrontonasal
syndome (CFNS), which affects multiple aspects of craniofacial, skeletal, and neurological development,
is caused by mosaicism for ephrin‐B1, which results in partitioning of the embryo into ephrin‐B1‐positive
and ‐negative domains. Here, we elucidate the developmental contexts, molecular mediators, and
cellular behaviors associated with this aberrant segregation. By generating compound mutant mouse
and cell culture models, we show that cell segregation occurs in the neuroepithelium, a progenitor
population for cells of the craniofacial skeleton, and is driven by unidirectional signaling to regulate the
actin cytoskeleton. These results delineate the developmental contexts of pathological cell segregation,
shed light on the roles of forward and reverse signaling in this process, and begin to identify the cellular
mechanisms involved.
Planar cell polarity signaling in airway epithelial homeostasis and disease.
E.K. Vladar1, J.V. Nayak2, J.D. Axelrod1; 1Pathology, Stanford University School of Medicine, Stanford, CA,
Otolaryngology‐Head and Neck Surgery, Stanford University School of Medicine, Stanford, CA
Planar cell polarity (PCP) signaling is a key developmental pathway that regulates directional cell
behaviors including migration, oriented division, and polarized cellular morphologies. We have recently
demonstrated that airway cilia, whose whip‐like movement propels inhaled contaminants out of the
lung, are physically oriented in a common direction by PCP signaling. The airway epithelium
differentiates during embryonic development and acquires PCP in a two step process: 1. prior to
ciliogenesis, all cells participate in cell‐cell communication to align along the lung‐oral tissue axis; and 2.
after ciliogenesis, multiciliated cells (MCCs) align their cilia via the microtubule cytoskeleton towards the
oral direction. At the heart of this mechanism are “crescents” of asymmetrically distributed PCP protein
complexes at the oral and the lung side adherens junctions that both mediate cell‐cell alignment (Step 1)
and tether the microtubules that orient cilia in the oral direction (Step 2).
We observe that crescents begin to form, but fail to maintain their oral‐lung asymmetric localization if
MCC differentiation is blocked in in vitro differentiated primary mouse airway epithelial cells, indicating
that MCCs are necessary for PCP signaling in the airway epithelium. We show that PCP does not require
the physical presence of cilia, and therefore does not depend on cilium‐induced flow, but does require
the motile ciliogenesis transcriptional program, which drives the expression of MCC‐specific crescent
components. Exploring the consequences of PCP dysfunction, we find that PCP mutant mouse airways
not only have misaligned cilia, but also have less stable epithelial junctions and do not effectively
undergo wound healing after injury. This suggests that PCP signaling requires proper airway epithelial
differentiation and regulates multiple aspects of its homeostasis. Respiratory diseases with chronic
inflammation, such as asthma, rhinosinusitis and cystic fibrosis lead to the loss of MCCs, and we observe
that human airway epithelia from patients with these diseases also lack PCP crescents. Correspondingly,
cultures from these specimens have defective junctions and wound healing response, which may
contribute to disease phenotypes. PCP dysfunction in the diseased epithelia is likely due to defective
MCC differentiation in response to inflammatory cytokines. Importantly, we find that pharmacological
manipulation of the motile ciliogenesis pathway can reinstate MCCs, restoring PCP activity and other
epithelial functions to diseased epithelia. Our results indicate that PCP signaling is a novel factor in
epithelial homeostasis in normal airways and may drive reversible pathogenic cellular changes in
Asymmetric partitioning of WNT and SHH signaling regulates the specification of hair follicle
stem cells.
T. Ouspenskaia1, I. Matos1, A.F. Mertz1, J. Levorse1, L. Polak1, E. Fuchs1; 1Laboratory of Mammalian Cell
Biology and Development, HHMI, Rockefeller University, New York, NY
Organogenesis is a complex process where cell proliferation must be balanced with cell fate
specification to generate a functional tissue, and stem cells (SC) need to be set apart for tissue
homeostasis and repair. Hair follicles (HF) provide a unique system to study the origin of SCs during
development and the mechanisms that regulate their specification.
Little is known about the early stages of HF morphogenesis, other than established links to both WNT‐
and SHH‐signaling. Interestingly, SOX9 – a key transcription factor in adult HFSCs – is expressed as soon
as the first hair bud appears in the basal layer of the epidermis. This prompted us to hypothesize that SC
specification might occur during that time.
Using live imaging and immunofluorescence microscopy, we show that early cell divisions in hair buds
are exclusively asymmetric, generating one basal daughter high for WNT‐signaling and one suprabasal
daughter high for SHH‐signaling. Using in utero epidermis‐specific lentiviral gene delivery, we create
juxtaposing regions of high and low WNT or SHH activity, and show that there is an extensive crosstalk
between these pathways that regulates the balance of cell proliferation and cell fate commitment.
Finally, we use long‐term lineage tracing of cells in the hair buds to identify the earliest cell types that
contribute to the adult SC niche.
Niche architecture and stiffness supports satellite cell self‐renewal.
R. Cheng1,2, H. Liu3, S. Davoudi1,2, C. Simmons2,3, P.M. Gilbert1,2; 1Donnelly Centre for Cellular and
Biomolecular Research, University of Toronto, Toronto, ON, 2Institute of Biomaterials and Biomedical
Engineering, University of Toronto, Toronto, ON, 3Department of Mechanical Engineering, University of
Toronto, Toronto, ON
Satellite cells, the resident adult stem cells of skeletal muscle, are required for tissue repair throughout
life. Our prior studies showed that a soft environment maintains tissue homeostasis by supporting
asymmetric self‐renewal divisions, though the molecular mechanisms remain unclear. In the current
study, we quantified the biomechanical changes that accompany skeletal muscle regeneration and
determined the cellular and molecular implications of those changes on the fate of satellite cells. Using
shear rheometry and atomic force microscopy measurements, we quantified a transient 3‐fold stiffening
of the satellite cell niche that accompanied time‐points associated with satellite cell activation and
symmetric self‐renewal divisions. Immunohistochemical and western blot analysis suggests that this is
due in part to increased extracellular matrix deposition and reorganization at these timepoints. Our
subsequent studies indicated that the transient increase in stiffness serves to prime the non‐canonical
Wnt/planar cell polarity (PCP) pathway, which was shown by others to control satellite cell symmetric
self‐renewal divisions. Firstly, we found that increased stiffness serves to induce transcription of key
molecular mediators of the PCP pathway. Secondly, using a tunable three‐dimensional model of the
satellite cell niche, together with time‐lapse microscopy, we found that a stiff niche forces spindle pole
alignment by constraining the architecture of the niche to support planar orientation divisions. We
further found that three‐dimensional niche stiffness synergizes with Wnt ligands to enable self‐renewal
divisions. In summary, we find that dynamic changes in the biomechanical properties of the stem cell
niche serve two important roles key to supporting satellite cell self‐renewal divisions;
mechanotransduction and niche architecture control. Our results provide new insights into the role of
niche biomechanics as they relate to satellite cell self‐renewal and the planar cell polarity pathway.
Random flashcards
State Flags

50 Cards Education

Countries of Europe

44 Cards Education

Art History

20 Cards StudyJedi

Sign language alphabet

26 Cards StudyJedi

Create flashcards