Self-Assembly of Filopodia-Like Structures
on Supported Lipid Bilayers
by
MASSACHUSETTS INSTITUTE
OF TECHN%'OLOGY
Kwonmoo Lee
NOV 18 2010
M.S. Physics
of
Science and Technology, 1998
Pohang University
LIBRARIES
Submitted to the Department of Physics
in partial fulfillment of the requirements for the degree of
ARCH! VFS
Doctor of Philosophy in Physics
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
June 2010
0 2010 Massachusetts Institute of Technology. All rights reserved.
Signature of Author:
/,
A
Department of Physics
April 20, 2010
Certified by:
Marc W. Kirschner
John Franklin Enders University Professor/Chair of Systems Biology
Harvard Medical School
Thesis Supervisor
Certified by:
Alexander van Oudenaarden
Professor of Physics
Co-supervisor
Accepted by:
Krishna Rajagopal
Professor of Physics
Associate Department Head for Education
Self-Assembly of Filopodia-Like Structures
on Supported Lipid Bilayers
by
Kwonmoo Lee
Submitted to the Department of Physics
on April 30, 2010, in partial fulfillment of
the requirements for the degree of
Doctor of Philosophy in Physics
Abstract
Filopodia are finger-like protrusive structures of cells, comprised of actin bundles, which
can serve as sensory organelles. To probe their pathway of assembly we have
reconstitutued filopodia-like structures (FLSs) by applying frog egg extracts to supported
lipid bilayers containing phosphatidylinositol(4,5)bisphosphate, PI(4,5)P 2. The FLSs
recapitulate important characteristics of filopodia - they assemble parallel actin bundles
from the lipid membrane and they form in the presence of capping activity. Known
filopodial tip components such as Diaphanous-related formin and VASP localize to the
membrane base of the structures, and bundling protein fascin to the shaft. Actin subunits
assemble at the tip and translocate into the shaft. FLS assembly requires negativelycharged lipid membranes, with specific requirements for PI(4,5)P 2 and, for maximal
efficiency, phosphatidyl-serine. The focal nature of FLSs is not a result of templating by
PI(4,5)P 2 microdomains but instead by the self-organization of tip complex assembly on
uniform PI(4,5)P 2-enriched regions. BAR domain protein toca-1 recruits N-WASP then
the Arp2/3 complex and actin assembly follow. Elongation proteins Diaphanous-related
formin, VASP and fascin are recruited later. The Arp2/3 complex is absolutely required
for FLS initiation but is not required for elongation, which may involve multiple factors
including formins. We propose a model for filopodia formation involving an initial
clustering of Arp 2/3 complex regulators, self-assembly of filopodial tip complexes on
the membrane, resulting in the outgrowth of parallel actin bundles.
Thesis Supervisor: Marc W. Kirschner
Title: John Franklin Enders University Professor/Chair of Systems Biology, Harvard
Medical School
Table of Contents
Title
Abstract
Table of Contents
Acknowledgments
List of Figures
1. General Introduction
2. In Vitro Reconstitution of Filopodia-Like Structures (FLSs)
Introduction
Results
Discussion
3. Membrane Requirements for FLS Formation
Introduction
Results
Discussion
4. Temporal Coordination of FLS Tip Assembly
Introduction
Results
Discussion
64
5. A Clustering-Outgrowth Model of Filopodia Formation
Introduction
67
Results
68
Discussion
77
6. Conclusion
80
7. Materials and Methods
85
8. References
97
To My Family
Acknowledgments
I have had the great fortune to have Marc Kirschner as my thesis supervisor. I
would like to thank him for his support, advices, and inspiration through the years. His
intuition and enthusiasm have had an immeasurable impact on my research. I am always
amazed by his breadth and depth in science.
I would like to thank Alexander van Oudenaarden, my co-supervisor who
supported my decision to work with Marc Kirschner as a physics student. I also would
like to thank the members of my thesis committee, George Benedek, Mehran Kardar, and
Frank Gertler for their invaluable advices and service.
I have been very fortunate to collaborate with Jenny Gallop, a post-doc in the lab,
who made great contributions to this project. Her exceptional expertise on biochemistry
and lipid signaling has been a great asset to this work. I also like to thank Komal
Rambani for her helping me with my experiments. I would like to thank Orion Weiner,
Henry Ho, and Andres Lebensohn, who have taught me all the experimental techniques
and giving me countless advices, Michael Gage, Scott Gruver, and Victor Li for reading
this thesis.
I also would like to thank Euiheon Chung for giving me invaluable advices about
how to build solid biological knowledge in my early stage of graduate study, Hyungsuk
Lee for his encouragement for my research.
Finally, I thank my parents, parents-in-law, and my wife Namnim for their
unconditional love and supports, and my son, Ian Bumsoo for growing healthily and
cheerfully. Without them, I could not finish this thesis.
List of Figures
Figure 2.1
Assembly and dynamics of filopodia-like structures formed on supported
lipid bilayers containing 45% PC, 45% PI, 10% PI(4,5)P2. ------------ 33
Figure 2.2
FLS diameter histogram, electron microscopy and quantitation of pulse
chase elongation rate -----------------------------------------------------
Figure 2.3
35
Immunostaining and use of fluorescently tagged proteins shows that
filopodia-like structures contain bundling protein fascin and filopodial tip
complex proteins -------------------------------------------
37
Figure 3.1
Membrane requirements for FLS formation ------------------------
46
Figure 3.2
Control experiments for GFP-PLC8 PH domain binding to the supported
bilayer ---------------------------------------------
48
Figure 3.3
Domain formation and FLS distribution in supported bilayers --------- 50
Figure 3.4
Rescue of FLS formation from fluid membranes by increased PI(4,5)P2
and extract concentrations -------------------------------
Figure 4.1
52
Kinetics of signaling protein recruitment to filopodia-like structures Arp2/3 complex signaling proteins are recruited before actin and formin,
VASP and fascin are recruited later ------------------------
61
Figure 4.2
SDS-PAGE of the purified fluorescently tagged proteins ---------- 63
Figure 5.1
Initiation and elongation of FLSs occur by separable molecular
mechanisms -----------------------------------------
71
Figure 5.2
The effects of GST-CA and immunodepletion of N-WASP and toca-1 on
FLSs -----------------------------------------------------------------------------
Figure. 5.3
73
Quantification of the Arp2/3 complex fluorescence of the GST-CA
inhibition experiment from Fig. 5.1 G-H. Inhibition of FLS Arp2/3
complex independent elongation by dominant-negative RhoA ----------- 75
Figure 5.4
A clustering-outgrowth model for filopodia formation --------------------
79
Chapter 1
General Introduction
Actin Cytoskeleton
Actin is the most abundant protein in eukaryotic cells and is well conserved
throughout all eukaryotic cells. It plays essential roles in cell shape change and
locomotion, fundamental physiological processes in development, wound healing,
neuronal pathfinding, and immune responses (Pollard and Cooper, 2009). Actin filaments
provide cells with mechanical support as well as force generation for movement,
indispensable to many cellular processes such as mechanosensation, endocytosis,
motility, polarity and cytokinesis (Stossel, 1993;Drubin et al., 1996;Field et al,
1999;Pollard and Cooper, 2009).
Under physiological salt conditions, actin monomers undergo spontaneous
polymerization,
resulting
in long helical filaments (Pollard, 2007).
The helical
arrangement makes the filament more resistant to spontaneous breakage. The rate
limiting step of the polymerization is a nucleation process where stable actin oligomers
are created (Pollard and Cooper, 1986). Once nucleation happens, actin polymerization
happens rapidly. Actin filaments have structural polarity which determines the
directionality of myosin motor proteins on the actin filaments. In addition, one end of the
filament called "barbed end" grows much faster than the other end called "pointed end".
It is the barbed ends that contribute to force generation toward the cell membrane,
generating cell protrusions (Pollard, 1986).
Actin has a binding site for ATP or ADP. Soon after a monomeric ATP-actin is
incorporated into filament, ATP is hydrolyzed to ADP, followed by phosphate release.
This reaction causes subtle structural changes in the actin filament which is more prone to
actin disassembly (Carlier, 1988;De La Cruz et al., 2000). In this way, ATP hydrolysis in
actin can serve as a molecular timer which differentiates old and new actin filaments.
These properties of polarity and ATP hydrolysis allow actin filaments to undergo
"treadmilling" where actin assembly at the barbed end is balanced with actin disassemly
at the pointed end. In this nonequilibrium steady state, actin filament can move toward
one direction until ATP is consumed completely, which is a fundamental process in cell
migration (Neuhaus et al., 1983).
Cell protrusion is coupled with the polymerization of actin filaments toward the
leading edge plasma membrane (Pollard and Borisy, 2003). Current biophysical models
state that filament elongation acts as a ratchet to rectify the thermal fluctuations of actin
filaments in the forward direction (Peskin et al., 1993;Mogilner and Oster, 1996;Mogilner
and Oster, 2003). The elastic actin filaments constantly undergo bending due to thermal
fluctuations (Mogilner and Oster, 1996). When this thermal bending generates the gap
between actin filaments and membrane, an additional monomeric actin can be inserted
into the filaments. When these filaments straighten, the restoring force is exerted against
the membrane, resulting in cell protrusions. Many studies of F-actin dynamics in
migrating cells (Theriot and Mitchison, 1991;Watanabe and Mitchison, 2002;Zicha et al.,
2003) have resulted in a model where leading edge movements depend on the
coordinated assembly and disassembly of a branched network of actin filaments in a
treadmill-like fashion (Pollard and Cooper, 2009;Pollard and Borisy, 2003).
The actin cytoskeleton also plays important roles in pathophysiologcal processes
such as cancer metastasis. Protrusion activation, likely in cross-talk with the loss of cellcell adhesion, is the earliest event in the epithelial-to-mesenchymal transformation
(EMT), which underlies the metastasis of epithelial carcinoma (Yilmaz and Christofori,
2009). In the epithelial sheet model which recapitulates many of the early steps of EMT,
it has been shown that physical force such as traction force generated by integrin-
dependent actomyosin was sufficient to disrupt cell-cell adhesion during epithelial
scattering (de Rooij et al., 2005). Since there are vigorous protrusion responses long
before cells break cell adhesions to crawl, it is likely that force generation from cell
protrusion can contribute directly to disruption of cell junctions. Therefore, understanding
how cell protrusion driven by actin polymerization is acquired in a variety of chemical
and mechanical situations will shed light on metastasis in addition to normal cell
migration .
Organizing the Actin Cytoskeleton
Cells have a variety of actin binding proteins to organize the actin cytoskeleton.
To achieve rapid actin polymerization for cell protrusion and locomotion, cells have to
maintain a highly concentrated poll of monomeric actin and spontaneous actin
polymerization should be suppressed. To achieve these goals, cells use actin monomer
binding proteins, thymosin and profilin. By sequestering monomeric actin, thymosin
plays a role in global suppression of actin polymerization (Safer and Nachmias, 1994).
Profilin which competes with thymosin for binding actin monomer, can deliver
monomeric actin to barbed ends of actin filaments (Pring et al., 1992). Therefore, local
activation of profilin leads to a high concentration of monomeric actin available for
barbed end incorporation.
Since actin polymerization is well-suppressed in vivo, one of the most important
questions is how new actin filaments are nucleated in the first place. This is done by two
major actin nucleators, Arp2/3 complex and formin. Actin monomers are polymerized
very slowly because formation of dimers and trimers is highly unfavorable compared
with polymer elongation (Pollard and Cooper, 2009). The Arp2/3 complex is composed
12
of two-actin related proteins, Arp2 and Arp3, with five other subunits (Arc4 1,
Arc34,Arc20, Arc16) (Welch et al., 1997;Ma et al., 1998a). Nucleation-promoting factors
such as N-WASP deliver an actin monomer to the Arp2/3 complex so that activated
Arp2/3 complex mimic stable actin trimer, thereby dramatically increasing the actin
polymerization rate. Binding to the side of a pre-existing filament completes Arp2/3
activation, and the barbed end of the daughter filament grows from the Arp2/3 complex.
The Arp2/3 complex also binds pre-existing actin filament at an angle of 70 degree to
form actin meshwork filaments (Pollard and Cooper, 2009). This branching is spatially
controlled by the preference of the Arp2/3 complex for binding to newly polymerized
ATP-actin filaments, meaning that branching occurs only close to the membrane
(Rafelski and Theriot, 2004). In addition to the actin filament branching, debranching is
also spatio-temporally controlled. The Arp2/3 complex dissociates from the filament
when phosphate is released upon ATP hydrolysis by actin (Pollard and Borisy, 2003).
The Arp2/3 complex itself also hydrolyzes ATP. This ATP hydrolysis is also thought to
be related with debranching of actin meshwork (Rafelski and Theriot, 2004;Le Clainche
et al., 2003). Formins initiate polymerization from free actin monomers and remain
associated with the growing barbed end. Profilin-actin binds to formin and transfers actin
onto the barbed end of the filament(Pollard and Cooper, 2009).
For spatial organization of the actin cytoskeleton and efficient force generation,
cells also need to stop the actin polymerization. The major role of capping protein is to
bind to the growing barbed ends of actin filaments to terminate further elongation
(Pollard and Borisy, 2003). Because capping protein prevents old polymerized filaments
from growing, they allow monomeric actin to be incorporated into barbed ends of newly
polymerized actin filaments. This results in a very efficient tread-milling of the actin
meshwork (Pantaloni et al., 2001). Interestingly, capping protein helps to form the actin
meshwork of the leading edge, the lamellipodium. By capping old filament, capping
proteins make only newly created filament barbed ends grow. This makes highly
branched actin meshwork which can push the leading edge of the cells (Carlier,
1998;Condeelis et al., 2005). Since actin filaments have finite stiffness, they undergo a
buckling transition when too much load is exerted. Once the filament is buckled, it is no
longer useful in force generation. In order to minimize this buckling, the length of actin
filament has to be short. Capping proteins terminate the filament elongation by capping
the barbed ends of actin filaments. This capping process takes place very rapidly because
capping proteins are abundant in most cells and have high affinity for the barbed end
(DiNubile et al., 1995). Capping proteins also help with inhibiting filament elongation
from accidental nucleation.
Rapid actin polymerization will quickly deplete the pool of actin monomers.
Therefore, in order to maintain efficient migration, cells need to constantly replete actin
monomers
by depolymerizing old actin filaments. Cofilin is the major actin
depolymerization factor, which selectively binds ADP-actin filaments (old filament) and
promote their depolymerization (Pollard and Borisy, 2003). Binding of Cofilins to ADPactin filaments changes the twist of the actin (McGough et al., 1997) and promotes the
severing of the filaments into short segments (Pollard and Borisy, 2003). Cofilin's
severing and depolymerization activities are inhibited by phosphorylation, monomeric
actin binding, and binding to PI(4,5)P 2 (DesMarais et al., 2005;Paavilainen et al., 2004).
In addition to depolymerization, Cofilin can contribute to the nucleation of dendritic
networks both in vitro (Ichetovkin et al., 2002) and in vivo (DesMarais et al., 2004)
synergistically with Arp2/3. This synergy results from the amplification of the Arp2/3
complex's nucleation activity by cofilin's severing activity, which creates barbed ends
that elongate to form newly polymerized actin filaments (Ichetovkin et al., 2002). The
newly polymerized filaments are the preferred filament type for Arp2/3 complexmediated branching (DesMarais et al., 2005;Ichetovkin et al., 2002;DesMarais et al.,
2004). Cofilin also has been found to amplify and stabilize N-WASP generated
invadopods, suggesting that the synergistic interaction between the cofilin and Arp2/3
complex pathways described above is at work during cancer cell invasion (Ghosh et al.,
2004;Yamaguchi et al., 2005).
Crosslinking proteins also affect the structural organization of cytoskeletal
networks and the viscoelastic properties of the cytoplasm by assembling networks or
bundles of actin filaments. Fascin and fimbrin stabilizes parallel bundles of filaments
such as those in filopodia (Vignjevic et al., 2006a). Filamin, a-actinin, and spectrin can
stabilize orthogonal networks. At low concentrations a-actinin links actin filaments to
form an isotropic gel whereas at high concentrations, a-actinin assembles actin filaments
into parallel or anti-parallel bundles (Fletcher and Mullins, 2010).
Molecular Mechanism of Formin-Mediated Actin Nucleation
In filopodia or contractile rings of cytokinesis, we can find the rapid assembly of
long unbranched actin filaments. The mechanism by which long filaments can be formed
without being capped is provided by formin which is an actin nucleator and remains
processively at the barbed end for protection from capping proteins (Pruyne et al., 2002).
Formin also adds profilin-actin to the barbed end at rates about 5-10 fold faster than free
barbed ends (Romero et al., 2004). In addition, it has been shown that formin is a motor
protein which harnesses ATP hydrolysis of profilin-actin to allow rapid elongation and
strong force generation (Romero et al., 2004).
In budding yeast, formins Bnrl and Bnil are localized to the bud neck and the tip
to assemble actin cables where myosin-V delivers other proteins and vesicles to the bud
tip (Kovar, 2006). For filopodium formation, mouse formin mDia2 (Pellegrin and Mellor,
2005) and Dictyostelium formin dDia2 (Schirenbeck et al., 2005) generate actin bundles
at the leading edge of the cell.
The common feature of various isoforms of formins is a FH2 domain composed
of 400 amino acids (Higgs and Peterson, 2005) and the adjacent variable length prolinerich FH1 domain which binds to profilin-actin to mediate actin assembly (Kovar, 2006).
When formin is inactive, the regulatory region binds to FH2 domain for autoinhibition.
When Rho GTPase binds to the formin regulatory region, the FH2 domain is exposed so
that it can form a dimer with other active formin molecules (Kovar, 2006).
It is suggested that each half of the FH2 homodimer interacts with only one actin
subunit (one half with the ultimate subunit and the other with the penultimate subunit)
and that it undergoes a stair stepping motion with the elongation of barbed end. Structural
and biochemical experiments revealed that the FH2 dimer exists in equilibrium between
the closed and open states. The closed state does not allow actin monomer addition, and
the open state does. Movement of the lagging unit of the FH2 dimer towards the barbed
end induces transition from closed to open state (Otomo et al., 2005).
Flexible FH1 domains adjacent to the FH2 domain have variable consecutive
proline residues that bind profilin (Chang et al., 1997) to increase the probability that
actin is recruited to the barbed end (Kovar, 2006). Here, profilin acts as a scaffold to
bring FH1 and actin together (Sagot et al., 2002). Even though proflin binding sites are
quite variable, profilin seems to increase the barbed end elongation rate for all kinds of
formin FH1-FH2 domains (Kovar, 2006).
High-Order Actin Structures in Cell Migration
Cell migration requires mechanochemical cycles of leading edge protrusion, front
adhesion formation, rear adhesion disruption, and cell body contraction (Lauffenburger
and Horwitz, 1996). Deciphering the mechanism of cell protrusion is crucial for the
understanding of cell migration, since it seems to initiate the cycles (Pankov et al.,
2005;Small et al., 2002).
Cell protrusion is driven by actin polymerization pushing the leading edge of
plasma membrane (Pollard and Borisy, 2003). There are two distinct actin structures
involved in cell protrusion. The lamellipodium spans a 2 ~ 4 pim wide region from the
leading edge. Rapid actin assembly and disassembly occurs in conjunction with fast
retrograde flow(Ponti et al., 2004), driven by the activation of Arp2/3 which nucleates
actin filaments from pre-existing filaments into a characteristic dendritic network.
Current models postulate that the GTPase Rac1 activation promotes lamellipodium
assembly (Oikawa et al., 2004). The lamella spans a wider region behind the
lamellipodia, with lower rates of F-actin turnover and retrograde flow (Ponti et al., 2004).
Lamella actin bundles have been thought to be responsible mainly for the generation of
cell body traction, in response to activation of myosin II downstream of the GTPase
RhoA. However, recent works suggest that the lamella also play a major role in leading
edge protrusion (Gupton et al., 2005). One model suggests that the lamella spatially
overlap with lamellipodia (Ponti et al., 2004). As further evidence for this model, RhoA
is also active at the leading edge (Pertz et al., 2006). In fact, RhoA activation precedes
Rac 1 activation at the onset of a protrusion event (Machacek et al., 2009), likely via its
ability to activate the actin nucleator formin (Narumiya et al., 1997;Yamana et al., 2006).
Filopodia are finger-like protruding F-actin-based structure at the leading edge,
composed of parallel bundles of actin filaments(Davenport et al., 1993;Lendvai et al.,
2000). They are often emerging from lamellipodia, larger sheet-like regions of actin
polymerization activity (Svitkina et al., 2003). Instead of generating force for migration
like lamellipodia, filopodia are thought to play important roles in sensing external
chemical and mechanical cues as they determine the direction in which cells and axons
move and they contain various adhesion and signaling receptors(Lebrand et al., 2004).
Filopodia contain bundled parallel actin filaments, which suggests that regulatory
proteins with anti-capping and/or processive elongation activity (e.g. VASP and formins)
might underlie their formation (Romero et al., 2004;Lebrand et al., 2004;Applewhite et
al., 2007;Breitsprecher et al., 2008). This structure will be reviewed in great detail in the
following section.
Stress fibers are structures consisting of anti-parallel actin bundles, and myosin II
motors. The anti-parallel actin bundles are slid over each other by the action of biopolar
myosin II filaments, resulting in stress fiber contraction (Huxley, 1985). This contractile
force is responsible for the detachment of the rear edge of migrating cells. Stress fibers
are connected to focal adhesion where traction force is exerted on extracellular matrix
(ECM). Adhesions consist of dynamic clusters of transmembrane ECM-receptors, called
integrins, which, together with over 50 other proteins, form a structural link between the
ECM and the F-actin network (Webb et al., 2002;Schwartz, 2001;Geiger et al., 2001).
Adhesions are mechanically adaptive, undergoing structural changes in response to force
stimulation (Galbraith et al., 2002). Contraction against these fixed external substrates is
what allows the force generated by myosin motors and filament growth and
rearrangement to move and reshape the cell.
Podosomes are dynamical structure of the actin cytoskeleton of monocyte-derived
cells such as macrophages, osteoclasts and dendritic cells (Bums et al., 2001;Linder et al.,
1999). Podosomes consist of dot-like, F-actin-rich close contacts, whose diameter is
about 1-2 pim. They are localized at the substrate-attached part of the cell. Within each
podosome, the actin core is surrounded by a ring enriched in vinculin and talin (Buccione
et al., 2004). Usually, podosomes are found in cells crossing tissue boundaries such as
macrophages and DCs (Buccione et al., 2004). It has been found that podosomes degrade
ECM (Extracellular Matrix) (Mizutani et al., 2002) which supports that they play a role in
cell invasion. Moreover, metalloproteases such as MTl-MMP or MMP-9 which degrade
ECM are found to be localized at podosomes in osteoclasts (Bums et al., 2001).
Podosomes are dynamic structures with a half life time of 2-12 min (Destaing et al.,
2003). They are mostly clustered into ordered groups, which undergo constant
rearrangements. The individual podosomes do not move, but they are disassebmled at the
rear and assembled at the front, thereby achieving net movement (Destaing et al., 2003).
It has been also shown that podosomes have an even faster internal dynamics, which
means actin turnover in the core is 2-3 times higher than the life time of podosomes
(Destaing et al., 2003), One of the best-investigated pathways for actin assembly involves
Cdc42, WASP/N-WASP, and Arp2/3 complex, which have been all localized at the
podosome core (Linder et al., 1999;Mizutani et al., 2002). Microinjection of both
constitutively active and inactive mutants of Cdc42 abolished podosomes in human
macrophages (Linder et al., 1999) and dendritic cells (Burns et al., 2001), supporting that
fine-tuned GDP-GTP cycling of RhoGTPases (Symons et al, 2000) is necessary for
control. Podosomes are also controlled by other RhoGTPases such as Rho and Rac
(Linder et al., 1999). Rapid actin turnover also requires regulated filament disassembly.
Cofilin is localized at podosomes (Linder and Aepfelbacher, 2003).
Signaling to the Actin Cytoskeleton
RhoGTPases including Rac, Cdc42, and RhoA are known to regulate pathways
mediating the dynamics of the actin cytoskeleton. GTPases function as molecular
switches between a GDP-bound inactive state and a GTP-bound active state (Hall, 1998).
When RhoGTPases are inactive with GDP, they are typically in the cytosol in RhoGDI
bound form (Hoffman et al., 2000). Once they become GTP bound, RhoGTPases are
dissociated from RhoGDI and inset their prenylation tails into plasma membrane where
they activate downstream effectors and initiate a variety of cellular responses. After
RhoGTPases become inactivated by GTP hydrolysis, RhoGDI extracts them back to
cytosol. Guanine nucleotide exchange factors (GEFs) promote the exchange of GDP for
GTP, and function to maintain GTPases in the active form (Ma et al., 1998b). GTPase
activating proteins (GAPs) stimulate GTPase activity, and promote the inactive, GDPbound form. The balance between activities of GEF and GAP regulates the activity of a
specific RhoGTPases.
Currently, in mammals, the Rho GTPase family are composed of over 20
members (Boureux et al., 2007). The three best characterized GTPases are Rho, Rac and
Cdc42, and all play an important role in regulating cell motility. Earlier studies in
fibroblasts suggested that the three GTPases control the formation of specific actin
structures. Rho regulates the assembly of stress fibers and focal adhesions, Rac controls
the formation of lamellipodia, and Cdc42 regulates the production of filopodia (Nobes
and Hall, 1995). Even though this study suggest each RhoGTPase has its distinct role in
organizing actin cytoskeleton, recent biosensor study revealed that they are coordinated
in cell protrusions with RhoA activation preceding Rac and Cdc42 activation (Machacek
et al., 2009).
N-WASP is one of nucleation promoting factors which activate the Arp2/3
complex. N-WASP has a C-terminal VCA domain which binds the Arp2/3 complex and
monomeric actin and deliver the monomeric actin to the Arp2/3 complex. Cdc42 is a
well-known upstream activator of N-WASP and toca-1 activation is also required for NWASP activation(Rohatgi et al., 1999). The regulators of N-WASP activity appear to
operate by either stabilizing or destabilizing its autoinhibitory conformation (Rohatgi et
al., 2000;Ho et al., 2004).
WAVE found in heteropentameric complexes is another nucleation promoting
factor containing C-terminal VCA domain. Differing from N-WASP, it is activated by
Rac. It may be a main player in the organization of the lamellipodium structure in
migrating cells by activating Arp2/3 complex. Recently, its specific phosphorylation state
and acidic lipid environment were shown to be required for the activation of WAVE
complex (Lebensohn and Kirschner, 2009).
RhoA can activate various downstream effectors involved in actin regulation. It
activates Diaphanous-related formins to generate actin bundles (Romero et al.,
2004;Wallar and Alberts, 2003). It can down-regulate cofilin activity by activating
ROCK (Maekawa et al., 1999). Activating ROCK by RhoA can also increase myosin
activity by phosphorylation (Kamm and Stull, 2001).
Phosphoinositides are also important signaling intermediates which regulate the
actin cytoskeleton. PI(4,5)P 2 can bind profilin. In addition, PI(4,5)P 2 was shown to bind
gelsolin and inhibit its severing and capping activity. It also contributes to N-WASP
activation by direct binding(Rohatgi et al., 2000). Most importantly, it is suggested that
PI(4,5)P 2 is upstream of Cdc42, thereby activating Cdc42-mediated actin assembly(Ho et
al., 2004).
PI(3,4,5)P 3 plays pivotal roles in cell migration. PI(3,4,5)P 3 production by P13kinase leads to Rac activation and cell polarization. Interestingly, activated Rac can
stimmulate more production of PI(3,4,5)P 3 by activating P13-kinase, constituting a RacPI(3,4,5)P 3 positive feedback loop. In neutrophil chemotaxis, this positive feedback in the
leading edge was reported by showing that the cell generated endogenous PI(3,4,5)P 3
when exogenous PI(3,4,5)P 3 was added(Weiner et al., 2002). This Rac-PI(3,4,5)P3
positive feedback is a core circuit for neutrophil polarity and amplification of the small
difference of chemoattractant.
Filopodia
Filopodia composed of bundled parallel actin filaments are central to several
fundamental physiological processes such as synapse formation, directionality of growth
cone movement, wound healing and cell motility in a variety of contexts(Gupton and
Gertler, 2007). They serve as organelles for environment sensing (Davenport et al.,
1993;Goodhill et al., 2004), thereby playing a role in signal transduction. One of the most
prominent examples is that filopodia play a vital role in brain development in that they
guide the neuronal growth cones (Bentley and Toroian-Raymond, 1986;Zheng et al.,
1996). Another example is that dorsal and ventral closure during development, where
filopodia are important in zippering epithelial sheets (Jacinto et al., 2000;Raich et al.,
1999;Vasioukhin et al., 2000). In addition, filopodia can directly contribute to cell
22
migration by providing adhesion sites, which are required for cell body contraction
(Bridgman et al., 2001;Heidemann et al., 1990) because they contain molecules involved
with adhesion, such as integrins or cadherins (Letourneau and Shattuck, 1989;Steketee
and Tosney, 2002).
It has been hypothesized that there is a central regulator of filopodial formation,
located at the tips of filopodia, called the "tip complex". Electron microscopy revealed
that there are dense protein aggregations at filopodial tips, indirect evidence of the
existence of a tip complex (Svitkina et al., 2003;Fiala et al., 1998). But, its molecular
identity and compositions have been elusive. The following are the proteins known to be
involved in filopodial assembly, some of which may comprise the tip complex.
Cdc42 is a well-known Rho GTPase that promotes filiopodia (Nobes and Hall,
1995).
It can activates N-WASP/WASP, upstream activators of
Arp2/3 complex
(Rohatgi et al., 1999;Rohatgi et al., 2000). Microscopy studies revealed the importance of
Arp2/3 complex in filopodial formation even thought it is known to assemble actin
networks, which are not compatible with actin bundles in filopodia (Svitkina et al., 2003),
suggesting that filopodia arise from the branched actin network of the lamellipodia.
However, other studies suggested that the Arp2/3 complex is dispensable and formin
plays an essential role, suggesting fomin rather than Arp2/3 complex comprises the core
machinery of filopodia (Steffen et al., 2006). A knock-out study of N-WASP also
suggested that N-WASP is not essential in filopodial formation (Lommel et al., 2001).
Therefore, there have been disputes about the role of Arp2/3 complex in filopodia
assembly.
The role of formin in filopodial assembly has attracted great attention since it can
generate de novo actin bundles and it localizes at the tips of filopodia. The most well-
known formin implicated in filopodia assembly is the Cdc42- and Rif-effector
mDia2/Drf3 (Pellegrin and Mellor, 2005;Peng et al., 2003). mDial/Drfl also seems to be
involved in filopodial formation (Sarmiento et al., 2008) even though its activator is
RhoA, not Cdc42. The most important functions of formin in filopodia are to nucleate
long actin filaments (Kovar et al., 2006) and protect barbed ends from capping proteins
(Zigmond et al., 2003).
Importantly, knockout studies revealed the essential role of
mDia2 for filopodium formation in Dictyostelium (Schirenbeck et al., 2005) leading to de
novo nucleation model of filopodia formation (Faix et al., 2009).
VASP is also important in the formation
of filopodia (Lebrand et al.,
2004;Applewhite et al., 2007;Mejillano et al., 2004). VASP localizes at the tips of
filopodia (Svitkina et al., 2003) and promotes the growth of filopodia in many cell types
(Lebrand et al., 2004;Mejillano et al., 2004). The molecular mechanism of VASP in
filopodial assembly is not clear. It has three domains: an N-terminal EVH1 domain for
subcellular localization, a central proline-rich domain, and a C-terminal EVH2 domain
containing GAB (G-actin-binding domain), FAB (F-actin-binding domain), and coiledcoil region mediating tetramerization (Krause et al., 2003). VASP has been suggested to
have anti-capping activity which allows long actin filament elongation by protecting
barbed ends from capping protein (Bear et al., 2002) although this is disputed
(Schirenbeck et al., 2006). Recent studies suggests that the roles of VASP is processive
elongation and filament tethering to the membrane via the clustering of barbed ends
(Applewhite et al., 2007;Breitsprecher et al., 2008). VASP proteins also bundle F-actin
(Schirenbeck et al., 2006;Barzik et al., 2005).
Fascin is actin bundling protein and localizes at the shaft of filopodia (DeRosier
and Edds, 1980). It plays critical roles in the formation of filopodia because knock-down
of fasin leads to inhibition of filopodia (Vignjevic et al., 2006a). FRAP study revealed the
dynamic nature of fascin binding to actin filaments, allowing easy remodeling actin
structures (Aratyn et al., 2007).
Aims of This Thesis
In vitro reconstitution of complex biological process provides a complementary
approach to in vivo studies in that it can reveal detailed mechanistic understanding of the
cellular processes in much more simplified environments. Despite the fact that in vitro
reconstitution of lamellipodial-like structure using Listeria or Acta beads allowed us to
know a great deal of biochemical and biophysical mechanisms of actin organization, an
in vitro model which faithfully recapitulates physiological aspects of filopodia is not well
established. The aim of this thesis is to reconstitute filopodia-like structures (FLSs) using
Xenopus egg extracts containing a great amount of actin regulatory proteins and reveal
the mechanism of filopodial assembly using reconstituted FLSs.
In Chapter 2, I will describe how FLSs can be reconstituted on supported lipid
bilayers using Xenopus egg extracts and provide evidence that FLSs recapitulate
filopodia. In Chapter 3, I will explain the requirements of specific lipid environments
leading to FLSs. In Chapter 4, in order to get the mechanistic insight in FLS assembly, I
will show the kinetic data which reveal temporal coordination of FLS tip proteins. In
Chapter 5, by intervening during FLS assembly in initiation and elongation phases, I will
describe our investigation into the mechanistic understanding of how FLSs arise. In
conjunction with kinetic data in Chapter 4, a new model of filopodial assembly will be
proposed.
Chapter 2
In Vitro Reconstitution of
Filopodia-Like Structures
ATTRIBUTIONS: This chapter contains the part of the manuscript "Self-Assembly of
Filopodia-Like Structure on Supported Lipid Bilayers, K. Lee, J. L Gallop, K. Rambani,
and M. W. Kirschner" submitted for the publication.
Introduction
Actin assembly underlies cell shape and movement in almost all eukaryotic cells
(Pollard and Cooper, 2009). Different assemblies of actin filaments underlie a range of
morphological structures, where regulatory proteins presumably play a major role. Of
particular interest are filopodia, which are central to synapse formation, directionality of
growth cone movement and cell motility in a variety of contexts (Davenport et al.,
1993;Lendvai et al., 2000). Filopodia contain bundled parallel actin filaments, very
distinct from the dendritic network found in other actin assemblies. The persistence of
actin growth in a parallel bundle suggests that proteins with anti-capping and/or
processive elongation activity (e.g. VASP and formins) underlie their formation.
Although formins are implicated in filopodia, they are also important in very different
structures e.g. lamellipodia (Romero et al., 2004;Svitkina et al., 2003;Lebrand et al.,
2004;Steffen et al., 2006;Krugmann et al., 2001;Yang et al., 2007). The role of the NWASP/Arp2/3 complex pathway in filopodia formation is unclear. Though it can
stimulate filopodia formation in cells, some RNAi and knockout studies suggest it is not
required (Steffen et al., 2006;Lommel et al., 2001;Miki et al., 1998;Snapper et al.,
2001 ;Nicholson-Dykstra and Higgs, 2008;Korobova and Svitkina, 2008).
Despite the tremendous progress made during the past decade in identifying the
molecular players in the actin cytoskeleton and their interactions, elucidating the specific
mechanisms by which these pathways affect the dynamics and spatial organization of the
cytoskeleton has remained challenging. This is because the cytoskeleton is a complex
dynamic structure, within which more than 150 actin binding proteins cooperate by
performing various functions, some of which overlap (Pollard and Cooper, 2009). Hence,
it is very difficult to dissect the mechanisms using genetic perturbation in vivo.
Furthermore, genetic studies can not demonstrate how these various proteins integrate
into a functional whole.
In the past, biochemical has greatly advanced our understanding of complex
biological processes in particular, the complex signaling and membrane pathways that
regulate actin assembly. For example, Xenopus egg extracts have proven to be a powerful
cell-free system for reconstituting complex cellular processes, including nuclear assembly
and disassembly, chromosome condensation, spindle assembly, DNA replication, and the
control of cell cycle progression (Liu and Fletcher, 2009). Using these extracts, Theriot
and Mitchison first reconstituted in vitro actin assembly for Listeria comet tails in a very
similar nature as that found in infected cells (Theriot et al., 1994). When Listeria enter the
cytoplasm of a host cell, they hijack the cell's actin machinery and develop an actin tail
of dense polymerized actin network, providing the force for bacterial propulsion (Tilney
and Portnoy, 1989). The actin comet tail recapitulates a simplified lamellipodium and the
bacterial surface represents the membrane at the leading edge (Upadhyaya and van
Oudenaarden, 2003). On the bacterial surface there are ActA proteins which mimic
WASP or N/WASP proteins which activates Arp2/3 complex, thereby nucleating actin
polymerization (Kocks et al., 1995). Beads (Cameron et al., 1999;van Oudenaarden and
Theriot, 1999) or phospholipid vesicles (Upadhyaya and van Oudenaarden, 2003) coated
with ActA showed similar motility behavior. Our lab has also reconstituted PI(4,5)P 2 and
Cdc42-induced actin polymerization in concentrated cytoplasmic extracts, systems that
permit detailed biochemical investigations into both problems of signaling and actin
assembly (Ma et al., 1998b;Ho et al., 2004). This in vitro system provides a simple model
of actin based motility and allows easier manipulation. These physiological systems have
also allowed us to study the biophysical mechanism of lamellipodium force generation
(Upadhyaya and van Oudenaarden, 2003;Parekh et al., 2005).
There have been also in vitro models of actin bundling. When WASP coated
beads were in the brain extracts where capping proteins were depleted, the bead
generated actin bundles instead of actin networks, suggesting anti-capping is crucial to
filopodia formation (Vignjevic et al., 2003). In this system, Arp2/3 generated long actin
filaments. Subsequently, fascin bound the filament and reorganized disorganized actin
networks to parallel bundles. When the energy gain from bundling overcomes the
bending energy of actin filaments of the networks, fascin can mediate the transitions from
actin networks to bundles (Ideses et al., 2008). Moreover, it was reported that elastic
interaction between membrane and actin filaments can substitute fascin's role to induce
the same transitions (Liu et al., 2008). Listeria are also know to generate actin bundles in
vitro when Arp2/3 complex is inhibited after Arp2/3-mediated actin nucleation starts on
the bacterial surface (Brieher et al., 2004). Although a significant amount of works has
been done for in vitro models of actin bundling, no current systems recapitulate two
physiological aspects of filopodia: (1) spontaneous formation of actin bundles in the
presence of capping activity, which is known to be prevalent in cytosol and (2) the
assembly of a membrane-localized tip complex. The critical feature of the assembly of
actin bundles at particular regions of membranes is the least understood aspect of their
function.
Results
To investigate actin polymerization from membranes and to exploit advanced
microscopic techniques like confocal and TIRF microscopy, we replaced liposomes as a
source of lipids with supported lipid bilayers (Chan and Boxer, 2007). Purified
Cdc42.GTPyS, N-WASP-WIP, Toca-1, Arp2/3, and actin comprise a minimal set of
proteins for stimulation of actin nucleation by PI(4,5)P 2-containing liposomes in vitro
(Ho et al., 2004). A lipid bilayer, underlain by an aqueous layer and glass support was
made according to standard methods.
We supplied prenylated Cdc42.GTPYS to the
membrane from the reaction mixture using the RhoGDI bound form and the EDTA
exchange reaction; actin and the other proteins were in the aqueous reaction mixture. A
thin actin layer with some heterogeneity was generated, as expected (Fig. 2.1A and B).
However when we substitute concentrated frog egg extracts for the purified proteins, a
strikingly different picture is observed: dense, focal and long actin structures, with a
diameter of 0.3-1.5 pm, rise from the surface of the bilayer (Fig. 2.1C-E, Fig. 2.2A).
These novel structures have not been previously reported.
For reasons we elaborate below, we call these novel structures filopodia-like
structures (FLSs). Filopodia are characterized by parallel arrays of actin filaments. To
observe the ultrastructure of the FLSs by electron microscopy, we use a protocol which
minimizes convection, and allows extraction of soluble proteins without fragmenting the
FLSs. This problem is not usually encountered fixing intact cells but because the FLSs
only have a small point of adherence to the glass and do not have surrounding membrane,
they are vulnerable to pipetting. By negative stain we observe bundled actin filaments,
distinct from the dendritic networks typically made by Listeria or ActA beads (Fig. 2. 1FG). The organized packing and parallel alignment of actin filaments is clearly revealed
when smaller FLSs spread out two dimensionally (Fig. 2.1 G). We observe actin bundles
of the complete size range seen by light microscopy (Fig. 2-2B and C). Prefixing the
FLSs using glutaraldehyde (0.1%) also shows similar actin bundles (Fig. 2-2D).
If these structures recapitulated filopodial structure and assembly, we would
expect new actin monomers to be added at the membrane-localized tip and the dynamics
to be commensurate
with that of filopodia in vivo (Mooseker and Tilney,
1975;Mallavarapu and Mitchison, 1999). Time-lapse confocal imaging and z-stack
reconstructions show that the typical initial rate of FLS growth is 2.5 pm/min, within the
range of filopodia (Fig. 2.1H). A pulse-chase experiment starting actin growth with Alexa
647-actin and adding Alexa 488-actin after 20 min shows that actin polymerization
occurs at the membrane-localized tip of the structure (Fig. 2.11-N). New actin monomers
are added at a rate of 2.8 pim/min (Fig. 2.2E). Growth into the membrane is of course
prevented by the underlying glass support; hence assembly is constrained to occur away
from the membrane. With this difference, the assembly of bundled and parallel actin
filaments at the membrane and the driving of actin into the shaft faithfully recapitulates
filopodial formation and growth.
At the junction of the membrane and the shaft in vivo is a collection of proteins
called the tip complex. To determine if such a structure is assembled for the FLSs, we
immunostained for known filopodial components or expressed tagged versions. Like
filopodia, FLSs stained for fascin along the length of the structure (Vignjevic et al.,
2006a) (Fig. 2.3A antibody specificity is shown in Fig. 2.2F). The membrane-localized
base of the structures contains characteristic proteins that localize to the tips of filopodia diaphanous related
formins, VASP, profilin, N-WASP
(Pellegrin
and Mellor,
2005;Svitkina et al., 2003;Steffen et al., 2006;Sarmiento et al., 2008;Ho et al.,
2001;Arasada et al., 2007) (Fig. 2.3 B, C, D, F). Toca-1 and Cdc42 also localize to the tip
(Fig 2.3E, G). Alexa-568 labeled Arp2/3 decorated the actin shaft (Fig 2.3H). In some
filopodia Arp2/3 is excluded from the shaft, in others it is present (Svitkina et al.,
2003;Johnston et al., 2008).
In summary, these results suggest that not only does the Cdc42, N-WASP, toca1, and Arp2/3 complex pathway work on supported bilayers, as predicted by previous
work with liposomes, but also that the supported bilayer configuration also leads to
stimulation of additional molecular pathways in extracts that resemble those occurring in
filopodia.
..
:::.:..:::::
. ......
..........
............
..........
..
...........
. ..
I
Figure 2.1 Assembly and dynamics of filopodia-like structures formed on supported
lipid bilayers containing 45% PC, 45% PI, 10% PI(4,5)P 2. (A) The purified system
containing N-WASP-WIP, toca-1, Arp2/3 complex, and actin generates uniform, short
polymerized actin. (B) Z-stack reconstruction showing growth of actin in the z-axis. (C)
Xenopus egg extracts generate focal structures, seen in x-y. (D) Z-stack reconstruction of
x-z shows the height of the actin structures. (E) 3D reconstruction. Bars: 5 pm. (F-G)
Negative-stain electron microscopy of the phalloidin-stabilized actin structures shows
that these are made of bundled, unbranched actin filaments Bars: 100 nm (H) Time-lapse
sequence of FLS formation seen in x-z. Bars: 2 gm. (I-N) Pulse-chase experiments show
that actin polymerization occurs from the supported bilayer. The reaction is started with
Alexa 647-actin (red) and chased by Alexa 488-actin (green). Bars: 2 gm. (I-K) 1 min
timepoint (I) First color actin (red) (J) Second color actin (green) (K) overlay. (L-N) 2
min 40 sec timepoint (L) First color actin (red) (M) Second color actin (green) (N)
overlay, bars: 2 jim.
. . .......................
.
.....................
......................
A
B
0
70 40 -
30 60
20 10-
0-FLSunamter (m
D
F
kD
207
120
99
57
Xenopus
Fascin
38
29
20
7
Tb.. OW
Figure 2.2 FLS diameter histogram, electron microscopy and quantitation of pulse
chase elongation rate The lipid composition used was 45% PC, 45% PI, 10% PI(4,5)P 2
(A) The distribution of FLS diameters. Due to the diffraction limit, the value for the 0.3
35
gm bin includes all FLSs whose diameters are below 0.3 gm. (B-C) Negative-stain
electron microscopy of the phalloidin-stabilized actin structures shows actin bundles of
typical FLS size. Bars: 100 nm (D) Negative-stain electron microscopy of the
glutaraldehyde-fixed actin structures shows similar actin bundles to the unfixed sample.
Bar: 100nm (E) The elongation curve for the second color of actin in pulse chase
experiments. The linear fitting line shows the elongation rate is 2.8 jm/min. Error bars
are s.d. (F) Western blot of Xenopus extract using the rabbit polyclonal antibody raised
and affinity purified against Xenopus fascin.
Figure 2.3 Immunostaining and use of fluorescently tagged proteins shows that
filopodia-like structures contain bundling protein fascin and filopodial tip complex
proteins. (A) Immunostained FLSs with side view, green: fascin immunostain, red:
phalloidin (B) tilted view, green: Drfl immunostain, red: phalloidin (C) tilted view,
green: GFP-VASP, red: Alexa 647-actin (D) tilted view, green: profilin immunostain,
red: phalloidin (E) top view, red: mCherry-Cdc42, green: Alexa 488-actin (F) tilted
view, green: N-WASP immunostain, red: phalloidin (G) tilted view, green: GFP-toca-1,
red: Alexa 647-actin (H) side view, red: Alexa 568-Arp2/3 complex green: Alexa 647actin. Bars: 2 gm. The lipid composition was 45% PC, 45% PI, 10% PI(4,5)P 2
Discussion
Using extracts and supported lipid bilayers we have produced actin structures in
vitro that recapitulate key features of filopodia - production of a bundled actin structure
and formation of a 'tip complex' at the membrane-localized, growing end of the actin
structure. The supported lipid bilayer lacks receptors and no additional modifications of
lipid or protein composition were necessary for formation of the FLSs, meaning that the
formation of filopodia itself does not rely on receptor signaling. And it further suggests
that significant information about the formation of filopodia is encoded within the
protein-protein and protein-lipid interactions that comprise the tip complex.
Since FLSs are assembled on the hard flat membrane, they grow backwards
instead of deforming and protruding the membrane. Therefore, some processes involved
in membrane protrusion are not recapitulated in our in vitro system. For example, IRSp53
is known to help with filopodia assembly by deforming the plasma membrane using its
inverse BAR domain (Mattila et al., 2007). It has also been shown that finger-like
membrane protrusions biophysically facilitate actin bundling by aligning actin filaments
(Liu et al., 2008). Therefore, it may be difficult to study processes coupled to membrane
deformation using reconstituted FLSs. However, reconstituted FLSs suggest that the tip
complex assembly itself may be independent of membrane deformation. Therefore, they
allow us to study the tip complex in much simpler circumstances than in vivo.
Chapter 3
Membrane Requirements for FLS Formation
ATTRIBUTIONS: This chapter contains the collaborative work with Jennifer L. Gallop
and the part of the manuscript "Self-Assembly of Filopodia-Like Structure on Supported
Lipid Bilayers, K. Lee, J. L Gallop, K. Rambani, and M. W. Kirschner" submitted for the
publication.
Introduction
Membranes not only function as passive barriers, they also actively participate in
biological regulations via lipid modifications, protein recruitment, membrane proteins,
etc. In particular,
they are known to play important roles in actin assembly. The
production of PI(4,5)P 2 and PI(3,4,5)P 3 by phosphorylation at the plasma membrane
activates and recruits small RhoGTPases, resulting in rapid actin polymerization. Specific
membrane characteristics such as composition, charge and curvature are also important in
the regulation of nucleation promotion factors including N-WASP, WAVE, and adaptor
proteins such as IRSp53 and toca-1 (Rohatgi et al., 1999;Lebensohn and Kirschner,
2009;Itoh et al., 2005;Suetsugu et al., 2009).
The cell membrane also laterally compartmentalize its constituents to perform
cellular functions, leading to membrane heterogeneity(Lingwood and Simons, 2010).
Lipid raft based heterogeneity has been suggested to arise from dynamic nano-scale
assemblies of spingolipids, cholesterol, and some specific proteins, leading to the
compartmentalization of biological processes within the membrane. Synthetic lipid
bilayers can also form heterogeneity via phase separation. In this model membrane, there
exist three distinct phases (Schwille et al., 2005); the gel (solid) phase characterized by
both high conformational and translational order of the lipid chains, the liquid-disordered
phase by conformational and translational disorder, and the liquid-ordered phase
charaterized by lipid chains order due to cholesterol and translational disorder, allowing
for lateral diffusion. Even though the phase separation behavior cannot be directly
translated into in vivo situations,
it provides a framework for understanding how
heterogeneity of cell membranes arises (Feigenson, 2009).
Use of artificial lipid bilayers in combination with purified proteins and extracts
provides a tractable biochemical model for investigating protein regulations at the
membrane-cytosol interface (Itoh et al., 2005). In the previous chapter, we have shown
that filopodia-like structures (FLSs) on PI(4,5)P 2 supported lipid bilayers using Xenopus
egg extracts can recapitulate several key features of filopodia - their focal nature, rate of
growth, their bundled parallel actin filaments and the localization of filopodial proteins at
the tips of the structures (similar to the proposed 'tip complex' of filopodia). In terms of
the roles of lipid, the focal nature of FLSs is of particular interest since it suggests that
filopodia are the outcome of self-organizing properties of protein machinery to break
spatial symmetry. But, another possibility is that the exact size and location of the FLSs
are determined by lipid templates in the supported bilayer related to lipid heterogeneity.
There are several examples of symmetry breaking in biology. One of the wellknown phenomena is symmetry breaking of Listeria actin comet tails. Shortly after
entering the host cell, a Listeria bacterium is surrounded by a symmetric cloud of hostcell actin filaments. However, asymmetrical distribution of ActA rearranges the actin
cloud to form an asymmetric tail (Tilney and Portnoy, 1989;Kocks et al., 1995) that
pushes the Listeria forward. The same spontaneous symmetry breaking occurs in the case
of the uniform distribution of ActA on the beads (van Oudenaarden and Theriot, 1999).
A few theories have been suggested to explain the underlying mechanism of this
symmetry breaking. van Oudenaarden and Theriot proposed that there is cooperativity of
growing actin filaments (van Oudenaarden and Theriot, 1999). When one filament grows
on one side of the bead, it facilitates neighboring filaments to grow by making room for
them. On the other hand, the growth of filaments on the other side of the bead is inhibited
due to being little room for their growth. This cooperativity generates local positive
feedback and global inhibition of filament growth, resulting in a symmetry breaking
event.
Most of in vitro studies of actin polymerization have previously used liposomes,
which are more relevant to membrane trafficking vesicles (Ma et al., 1998b;Suetsugu et
al., 2009). Cell-sized giant vesicles are more suitable for studying actin assembly near
the plasma membrane due to their much shallower curvature. They have been used in
combination with purified proteins to study actin dependent phase partitioning and actin
bundling induced by membrane deformation (Liu and Fletcher, 2006; Liu et al., 2008). In
this thesis, we used flat supported lipid bilayers which are suitable for various
microscopic techniques, but they have not been fully characterized when exposed to
extracts. Here, we explore the biochemical and biophysical features of supported bilayers
that are important for filopodia-like structure formation from Xenopus egg extracts.
Results
Tip complex assembly relies on the membrane as a scaffold as no such structures
form directly on glass. Phosphatidycholine (PC) alone, which is the major structural lipid
within cell membranes, forms a supported bilayer but does not stimulate FLS assembly.
Addition of phosphatidylserine (PS) or phosphatidylinositol (PI), which have a net charge
of -1 at physiological pH supports the nucleation of a small number of FLSs in a doseresponsive manner (Fig. 3.lA, B). However a direct comparision of PC/PS and
PC/PS/Phosphatidylinositol(4,5)bisphosphate (PI(4,5)P2 ) membranes of equivalent net
charge shows that there is PI(4,5)P 2 specificity, as the density of FLSs formed is 2.5-fold
higher at steady state (20 min) and their initial rate of appearance is 5-fold faster than
with just PS (the 60% PS and 30% PS, 10% PI(4,5)P2 conditions, shown in cyan and
orange, Fig. 3.1A). Substitution of PS for PI in the original composition produces a
similar final number FLSs but leads to a 2-fold increase in their rate of appearance (Fig.
3.IA). This may reflect the role of PS in binding some F-BAR proteins (Itoh et al., 2005).
All phosphoinositides (PIPs) nucleate actin spots on the membrane (bis/tris-PIPs are
shown in Fig. 3.1B and monophosphorylated PIPs in Supporting Fig. 3.2A). Few of the
actin spots nucleated by monophosphorylated PIPs produce elongated actin structures.
For bisphosphorylated PIPs and PI(3,4,5)P 3 , PI(4,5)P 2 is most effective at elongation and
also supports the highest number of nucleation sites at the membrane (Fig. 3.1B and C).
One caveat in comparing the different PIPs is that the extract may convert one form into
another, but overall these data show that FLS formation takes place optimally on a lipid
composition consistent with that of the plasma membrane.
The tendency of lipid mixtures to segregate into domains suggests two very
different models for tip assembly and FLS nucleation. In the first, the FLSs would be
templated by small domains enriched in PI(4,5)P 2 . In this model the size of the tips and
the girth of the FLSs should be commensurate with the size of the domains. In the
second possible model, tip complexes would self-assemble on the lipid bilayer in
relatively homogeneous lipid domains. To distinguish between these modes of
nucleation, templating versus self-assembly, we use the GFP-PLCS PH domain, which
binds PI(4,5)P 2 , to detect the distribution of PI(4,5)P 2 The GFP-PLCS PH domain is
evenly distributed over relatively large domains (Fig. 3.1D). We do not see an enrichment
of GFP-PLC6 PH domain at sites of FLS growth. Binding of the PH domain to PI(4,5)P 2
is significantly higher than to membranes containing PI and PC alone (Fig. 3.2B-D). In
addition, the kinetics of PH domain binding to the bilayer does not reveal hotspots of
PI(4,5)P 2 (Fig. 3.2E and F). We asked whether the regions depleted in PH domain
binding represent physical holes in the bilayer. To detect overall lipid deposition we
included rhodamine-PE in the liposomes. Surprisingly, rhodamine-PE has the inverse
distribution to GFP-PH domain, confirming that lipid is present in regions depleted of the
GFP-PLCS PH domain but also indicating that either PI(4,5)P 2 partitions differently from
rhodamine-PE, or that the PH domain cannot bind PI(4,5)P
2
within these domains (Fig.
3.1D). Domains of rhodamine-PE enrichment vary in size with 53% less than 10
m2 ,
38% 10-100 jim 2 , 7% 100-1000 jim 2 and 2% >1000 gm 2 . Rhodamine-PE enriched
regions nucleate 20-fold fewer FLSs than the rhodamine-PE depleted regions (Fig. 3. 1E).
Rhodamine-PE has been observed to label the liquid-disordered phase of membranes,
which has high fluidity (Crane et al., 2004). DiI, another fluorescent membrane marker
that labels the disordered phase, also localizes to areas with fewer FLSs (Fig. 3.3A).
Fluorescence recovery after photobleaching (FRAP) experiments on our supported
bilayers confirm that rhodamine-PE positive regions are fluid (Fig. 3.3B and C). The
irregular boundaries between rhodamine-PE regions and GFP-PH domain regions suggest
the coexistence of the liquid-disordered (rhodamine-PE) and gel phase (GFP-PH domain)
(Schwille et al., 2005). FRAP of GFP-PH domain confirms the low fluidity of these
regions (Fig 3.3D and E). Thus FLS formation occurs preferentially in the gel phase, but
not exclusively. The tendency of FLS not to form from the liquid-disordered phase can be
overcome by increasing the mole fraction of PI(4,5)P 2 and/or by increasing the extract
concentration (Fig. 3.1F and Fig. 3.4). Thus the gel phase itself is not an absolute
prerequisite for FLS formation; it could reflect a higher local concentration of PI(4,5)P 2 .
These properties of the supported bilayer demonstrate that assembly of the FLSs
does not occur by direct templating through a preformed domains in the membrane, but
instead occurs by a process of self-assembly driven by proteins at a permissive membrane
surface. There are several potential mechanistic interpretations of these results, all of
which could contribute to FLS formation: (1) Clustering of oligomeric activation proteins
(e.g. BAR domain proteins) (Padrick et al, 2008); (2) Positive feedback on small Arp2/3
complex-catalyzed clusters of polymerized actin through recruitment of more nucleation
factors (Co et al., 2007) e.g. by the interaction of N-WASP with barbed ends of actin
filaments, as occurs during symmetry breaking by N-WASP on lipid coated glass beads;
(3) Lattices generated by cooperative protein-protein interactions between signaling
molecules; (4) Cooperative association of proteins with PI(4,5)P 2 , as has previously been
shown for N-WASP (Papayannopoulos et al, 2005). Any small local fluctuations in
PI(4,5)P 2 could be magnified in such a mechanism.
. ... ........
.........
. ...
... ......... .........
..
AB
A
C3
900-3
400
800-
0
700600.0
250-601
500.
200
400-
15
300-
100
200
100
50
0.
0
2
Time (min)
4
0MMMMN
0
0
2
Tie (min)
PI(3.5)PP1(4,5)P2
(3.4,5)F
F N0-50
E
D
PI(3,4)
4
10 E200-250
WNOW100E 100-150
I
35-0 3150-2003m250300 E300
30
15
3
12
6
Extractconcentration (mghn)
25
Figure 3.1 Membrane requirements for FLS formation (A) Time course of FLS
appearance shows negatively-charged lipids are essential for FLS formation and there is
specificity for PI(4,5)P 2 and PS. Cyan: 60% PC/30% PS/10% PI(4,5)P 2, purple: 60%
PC/30% PI/10% PI(4,5)P 2, orange:40% PC/60% PS, red: 55% PC/45% PS, green: 70%
PC/30% PS. (B) Time course of FLS appearance. All PIPs can nucleate FLSs but there is
preference
for
PI(4,5)P 2 . Compositions:
60%
PC/30% PS/10%
PIP,
where
lime:PI(4,5)P 2, gray:PI(3,4)P 2 magenta:PI(3,5)P 2 navy:PI(3,4,5)P3 , brown:PI. Data are
the mean of 3 timecourses normalized to the average number of structures from the 3
experiments from 5 or more pictures over each supported bilayer. (C) Rate of actin
addition (using the pulse-chase approach) for the different PIPs shows headgroup
specificity, error bars are s.d., n = 18, 19, 16, 20. (4-way ANOVA, p<0.001). (D) GFP-
PH domain (green) addition to supported bilayers including rhodamine-PE (red) shows
membrane domains. Bar: 2 gm. The lipid composition was 45% PC, 45% PI, 10%
PI(4,5)P 2 , similar data was obtained with 60% PC, 30% PS, 10% PI(4,5)P 2 . (E) FLSs
grow preferentially from rhodamine-PE depleted (PH domain binding) regions. Alexa647 actin is in green, rhodamine-PE in red. Bar: 2 pm (F) Contour plot of the number of
FLSs per field of view at steady state (20 min) from fluid membranes in response to
PI(4,5)P 2 and extract concentrations. The lipid composition was 45% PC, 45% PI, 10%
PI(4,5)P 2 . For comparison, overlaid single points show the number of FLSs formed from
the gel phase. Data is plotted logarithmically. Example pictures and FRAP data are in
Fig. 3.4A.
I..","
..
.....
......
-:::
...............
............
.....
:,.:-----
-.. -
.
A ,oo
800
PI(3)P
PI(4)P
500
S4002500
300
16 200-
1
2
1000
0
2
4
Time (min)
BC
D-1A
BC
01
1
rPI(4,5)P2
=
O
0
10 350 560 7090 1150 1360
E
650
j350
250
150.
00:00.0
F omin
1m n
05:45.6 11:31.2 17:16.8
7min
11 min
18 min
Figure 3.2 Control experiments for GFP-PLCS PH domain binding to the supported
bilayer. (A) Time course of actin spot appearance with monophosphorylated PIPs.
Compositions: 60% PC/30% PS/10% PIP, where pink:PI(3)P, blue:PI(4)P green:PI(5)P.
Data are the mean of 3 timecourses normalized to the average number of structures per
. ..........
......
........
--
experiment. Most monophosphorylated PIP actin spots fail to elongate. (B) 50 nM GFPPLC6 PH domain binding to supported bilayer made from liposomes containing 45% PC,
45% P1, 10% PI(4,5)P 2. (C) 50 nM GFP PLCS PH domain binding to supported bilayer
made from liposomes containing 50% PC, 50% PI. (D) Image intensity histogram of PI
and PI(4,5)P 2 bilayers. (E) Kinetics of GFP-PLCS PH domain binding to PI(4,5)P 2
supported bilayer and dissociation after washing. (F) Time-lapse pictures of GFP-PLCS
PH domain binding to PI(4,5)P 2 supported bilayer from the graph shown in (E) show that
no PI(4,5)P 2 punctae are revealed by kinetic analysis. All bars: 2 gm. Similar
characteristics of GFP-PH domain binding was observed for 60% PC/30% PS/10%
PI(4,5)P 2, and background binding similar to PI was observed for the other
bisphosphorylated phosphoinositides, confirming the specificity of the domain (data not
shown).
..................
.......
. ...........
..........................
.......................
120Rhd
TmmPE
110100 o90s0o70800
-2
(,PPH
E
2
Time (s)
4
4
a
6
104
102
100
.98
~96
:94
0
50
100
150
200 250
300
Time (s)
Figure 3.3 Domain formation and FLS distribution in supported bilayers. (A)
Additional fluorescent markers that partition into the liquid disordered phase and thus do
not label areas where FLSs grow: TopFluor PI(4,5)P 2 (green) and DiI (red). Shown with
actin structures (blue). (B,C) Rhodamine-PE is fluid within enriched areas. Fluorescence
recovery after photobleaching experiment showing example image in (B) and
quantitation of five independent experiments in (C). (D,E) Fluorescence recovery after
..................
..............
......
......
photobleaching experiment of GFP-PH domain showing the much slower recovery
compared with rhodamine-PE, confirming that this domain binds PI(4,5)P 2 that is within
the gel phase. All bars: 2 p.m. The lipid composition used was 45% PC, 45% PI, 10%
PI(4,5)P 2 . The partitioning and fluidity of rhodamine-PE was tested and found to be
similar for all lipid compositions described in the paper.
..........
................
8% PI(4,5)P2
D120
15% PI(4,5)P2
100
30% PI(4,5)P2, 25 g/mi extract
60
40
20
0
-4
-2
0
2
4
6
Time (s)
8
10
12
-4
-2
0
2
4
6
8
10
12
Time (s)
Figure. 3.4 Rescue of FLS formation from fluid membranes by increased PI(4,5)P
2
and extract concentrations. (A) Single representative compressed z-stacks shown in x-y
orientation for data the quantified in Fig. 3.1F. The quantification was an average of 3
independent experiments, 3-6 images from each using 8, 15, 30 % PI(4,5)P 2 by mols with
background lipid composition of 45% PC and pro-rated mol fraction PI. Each condition
was tested with 3.1, 6.3, 12.5 and 25 mg/ml extract concentration. Increasing the mol
fraction of PI(4,5)P 2 and the extract concentration rescues FLS formation from fluid
membranes. No structures are seen for 15% PI(4,5)P 2 at less than 12.5 mg/ml extract
concentration. (B) Supported bilayer from the 30%, 25 mg/ml condition as an example,
showing the distribution of rhodamine-PE. (C) FRAP experiments of rhodamine PE at
the different PI(4,5)P2 concentrations with buffer or 25 mg/ml extract, as annotated
showing that it remains unaltered. Data are the mean of 5 experiments, error bars are the
standard deviation. To allow comparisons between the conditions, the initial intensity
levels were normalized (to 100%) and the bleached intensity level, which was -80-90%
of initital intensity, normalized to 0%. (D) FRAP experiments of rhdoamine-PE at
increasing extract concentrations at 10% PI(4,5)P2, again showing that rhodamine-PE
fluidity remains unchanged. However we would expect that protein binding to the
membrane will change the mobilities of some lipids and proteins. Data are the mean of 5
experiments, error bars are the standard deviation. To allow comparisons between the
conditions, the initial intensity levels were normalized (to 100%) and the bleached
intensity level, which was ~80-90% of initital intensity, normalized to 0%. Bars: 5 gm.
Discussion
Our experiments demonstrate the complex nature of lipid-extract interactions and
provide valuable insight into how lipid signaling can contribute to the formation of
filopodia. We tried to understand the role of lipid heterogeneity within supported lipid
bilayers in promoting FLS generation. Instead of pre-formed instructional templates in
the membrane, we found that uniformly distributed permissive membranes allows the
FLS formation via self-assembly. The permissive rather than instructive role for PI(4,5)P 2
without distinct patches means that the focal nature of FLSs is emerged from a highly
cooperative protein-protein interaction machinery. In other words, the role of lipid
bilayers is to only recruit the necessary proteins without any spatial organization. Then,
the proteins self-organize into distinct focal structures. This could be facilitated by the
clustering of oligomeric activation proteins (e.g. BAR domain proteins), followed by
large-scale protein complex formation between signaling molecules (e.g. toca-Cdc42-NWASP). Also, non-linear association of N-WASP and/or other proteins with PI(4,5)P 2
could make such contributions. It is also possible that some positive feed-back loop
mechanisms can reinforce this cluster formation. Local small fluctuations in PI(4,5)P 2
would be magnified in such a mechanism. At this point, it is unclear how the diameter of
FLS is determined. One possibility is that local depletion of necessary proteins may limit
the FLS diameter. It is also possible that competition with other membrane binding
proteins in the extract limits the diameter of the FLSs.
In addition, we showed the effectiveness of FLS formation is highly dependent
upon the fluidity of supported lipid bilayers. We found that fluidic liquid-disorderd
regions of supported bilayers had decreased tendency to form FLSs. On the other hand,
the gel-phase regions of membrane bind GFP-PH most effectively and nucleate FLSs.
Interestingly, filopodia have previously been seen to preferentially occur from
membranes with raft-type characteristics (Gaus et al., 2003). Therefore, this lipid domain
dependence of FLS formation may have in vivo significance. Actin polymerization in
vitro has also been shown to induce phase separations in giant vesicles and our extract
and PI(4,5)P 2 titration experiments support that the FLS signaling machinery can both
respond to and generate lipid inhomogeneities.
In terms of the lipid requirements for FLS formation, we find that FLSs form
most effectively on PI(4,5)P 2 containing bilayers with negative charge background and
that in addition to the charge effects, PS also has a specific role in accelerating the
nucleation process. We think this type of lipid composition is broadly consistent with
that of the inner leaflet of plasma membrane.
In order to make more refined measurements, defined and simpler compositions
of lipid bilayers may be required since the lipid composition that we have used is highly
complex acyl chain compositions from natural lipids. Addition of extract will cause lipid
modification as well as binding events so using purified protein components will reveal
clearer pictures of the formation. Even if FLSs can form without any receptors, in vivo
transmembrane proteins are also likely to influence the propensity to form filopodia and
use of polymer-cushioned bilayers and proteoliposomes represent an exciting extension
to linking together receptor signaling with actin polymerization.
Chapter 4
Temporal Coordination of FLS Tip Assembly
ATTRIBUTIONS: This chapter contains the collaborative works with Jennifer L. Gallop
and Komal Rambani, and the part of the manuscript "Self-Assembly of Filopodia-Like
Structure on Supported Lipid Bilayers, K. Lee, J. L Gallop, K. Rambani, and M. W.
Kirschner" submitted for the publication.
Introduction
We have made the first insight into how signals at the membrane organize
bundled actin structures using supported bilayers in conjunction with Xenopus egg
extracts, using confocal microscopy. We have an initial understanding of these structures
using immunostaining, and observe that characteristic components of filopodia are
localized at the membrane-localized tip of the structures, including VASP, diaphanousrelated formin and N-WASP.
Fascin and Arp2/3 complex are present throughout the
length of the bundles. At non-fluidic regions of the membrane bundles of multiple 100
nm wide filopodia-like structures (FLSs) are nucleated and grow from the membrane
surface. This in vitro assay is perfect for investigation by Total internal reflection
fluorescence (TIRF) microscopy as the key signaling events occur at the membraneextract interface.
Most of our understanding of cell regulation has been derived from cellular
responses to genetic or chemical perturbations with the caveat that cellular pathways may
be driven far outside their normal physiological ranges. Particularly, cellular
compensatory mechanisms to such perturbations make it hard to infer intact physiological
processes. For instance, when some proteins are knocked-out or depleted, the
homologous versions of the proteins can substitute their roles. Kinetic study of
fluorescent proteins in conjunction with sensitive imaging techniques provides an
alternative way to probe the regulation of complex biological processes with a minimallyperturbing approach (Nalbant et al., 2004;Hodgson et al., 2008). Specifically, biosensor
studies for RhoGTPases revealed the spatiotemporal coordination of RhoGTPases at the
order of seconds and sub-micron resolution (Pertz et al., 2006;Machacek et al., 2009).
One advantage of reconstituted FLSs is that monitoring the activities of various
tip proteins is readily possible when purified fluorescently tagged tip proteins are added
into the extracts. The pulse-chase experiment in Chapter 2, showed that FLS tips have
ongoing actin polymerization activity. This dynamic nature of FLS tips allows us to
translate the recruitment of tip protiens into their corresponding activities. This assay also
provides more homogeneous population and higher number of the structures than in vivo
cases. Flat lipid bilayers also provide convenience of monitoring and they are compatible
with TIRF microscopy.
TIRF microscopy has emerged as the method of choice to probe cellular processes
near the basal plasma membrane of adherent cells due to its unique capability of
illuminating a very thin region on the order of 100 nm (Axelrod, 2001). The evanescent
wave intensity decays exponentially from the interface and this near-field excitation
volume allows intrinsic optical sectioning to less than one-fifth of the excitation
wavelength. The selective excitation of TIRF removes the out-of-focus noise, reduces
photobleaching of fluorophores outside the focal plan, and is thus ideal for single
molecule imaging (Tokunaga et al., 1997;Webb et al., 2006). This sensitive imaging
technique allows us to add very low concentration of purified fluorescent proteins into
extracts so that the recruitment of tip proteins can be monitored without perturbing
endogenous activities.
Here, the recruitment of various filopodium tip proteins to sites of growth of the
FLSs will be monitored and their relative timing of the recruitment will be measured.
Characterization of the time of recruitment of all these proteins will allow us to see how
tip assembly is temporally coordinated, leading to the mechanistic insights of FLS tip
assembly.
Results
In order to measure the kinetics of FLS tip assembly, we purified GFP tagged
FLS tip proteins from 293F cells. We added each GFP fusion protein (5nM) into
Xenopus egg extracts with fluorescently tagged actin and monitored the recruitment of
each protein at FLS tips using TIRF microscope. Using the appearance time of actin
assembly as a common reference between different tip proteins, time shifts were
determined to compare their appearance times. Since the nucleation of each FLS start at
different time points, this approach also allows us to eliminate this variability between
different FLSs within the same experiment. Similar method called "computational
multiplexing" has been applied for comparing signals from different RhoGTPase
biosensors during cell protrusions (Machacek et al., 2009).
To investigate the self-assembly process of the FLSs, we have added fluorescently
labelled known filopodial proteins, to the extract and followed their recruitment to sites
of FLS formation by total internal reflection fluorescence microscopy (TIRFM) (Fig. 4.1
and 4.2). In these experiments we have normalized the recruitment time of each
component to the recruitment time of labeled actin and plotted the time differences as a
histogram, to show the start of the particular biochemical event for each FLS relative to
the first appearance of actin (Fig. 4.1). We find that this is a more informative way to
quantify the kinetic data than, for example, the time of half-maximal accumulation. The
latter is biased by the largest structures, is more sensitive to photobleaching and conflates
the time of recruitment with the extent of assembly. In these experiments we observe that
membrane-binding, F-BAR domain protein toca-1 is recruited earliest, as defined spots,
at sites that later go on to form FLSs (Fig. 4.1A and G). Some toca-1 spots also do not
develop actin colocalization. After a variable time period, N-WASP, the key activator of
the Arp2/3 complex is recruited to these sites, again before actin (Fig. 4. 1B and G). The
Arp2/3 complex is recruited concomitantly with actin (Fig. 4.IC and G). VASP and
mDia2 proteins that are implicated in the formation of long, unbranched actin filaments
are recruited to the tip complex after the first appearance of actin (Fig. 4. lD,E and H).
The bundling protein fascin is recruited last (Fig. 4.1 F and H).
These kinetic data suggest a mechanism where an Arp2/3 complex-driven
initiation step nucleates an initial branched actin structure in a small patch and this
stimulates the recruitment of filament elongation and bundling factors.
Toca-1 and N-WASP were recruited before actin polymerization. They are wellknown to activate Arp2/3 complex, leading to disorganized actin networks. This suggest
that Arp2/3-mediated actin networks play a vital role in FLS nucleation. Interestingly,
toca-1 recruitment was significantly ahead of N-WASP, suggesting toca-1 may be
important in organizing focal formation of FLS. On the other hand, mDia2, VASP, and
fascin were recruited to FLS tips after actin polymerization. They are known to be
directly related with filopodial tip complex. This data suggest that the N-WASP, toca-1
and Arp2/3 pathway provides the initial nucleation event and then there is a switch to tip
complex formation after nucleation.
In summary, these kinetic analysis of FLS tip assembly strongly suggest that the
following steps of filopodial assembly; 1) Actin network nucleation by Arp2/3 complex.
2) Actin elongation by formin and VASP. 3) Actin bundling by fascin.
............
in
AToca-1
B N-WASP
G
0.3
L 0.2
%I-
0
C
0
CU
C A 2/3 c
e
I-
LL
0
-4.1
D
Eu
n
-2.0
-1.0
0.0
1.0
mDia2 formin
E VASP
F
-3.0
Fasain
0.3
0.2
0.1
-1.5
-0.5
0.5
1.5
2.5
Time of first appearance relative to actin (min)
Figure 4.1 Kinetics of signaling protein recruitment to filopodia-like structures Arp2/3 complex signaling proteins are recruited before actin and formin, VASP and
fascin are recruited later. (A-F) Total internal reflection fluorescence images
illustrating the tip of single FLSs show fluorescently labeled proteins (GFP-toca-1, GFPN-WASP, Alexa568-Arp2/3 complex, GFP-mDia2, GFP-VASP, GFP-fascin) in green
and Alexa-647 actin in red. 20 s time intervals are shown at the time of first recruitment
of the signaling protein or actin. 5 nM labeled proteins were added to the extracts. (G)
Histogram showing the relative time of first recruitment of Arp2/3 complex signaling
proteins toca- 1 and N-WASP compared to the first appearance of actin, n=26 (toca- 1), 34
(N-WASP), 48 (Arp2/3 complex). (H) Histogram showing the relative time of first
recruitment for filopodia-linked elongation and bunding proteins, VASP, mDia2 and
fascin compared to actin, n=35 (mDia2), 34 (VASP), 34 (fascin). Using the KS-test,
p=0.000 for toca-1-N-WASP, N-WASP-Arp2/3 complex, Arp2/3 complex-VASP;
p=0.005 for Arp2/3 complex-mDia2; p=0.003 for mDia2-fascin; p=0.004 for VASPfascin and no significant difference for mDia2-VASP. The lipid composition was 60%
PC, 30% PS, 10% PI(4,5)P 2*
I ...........
. .......
.
OR,'
B
-250
-150
*
*
-~ 100
250 C
-150
-100
-200
D
- 116
97
75
66
-
75
- 50
-
45
-
31
-
21
-
14
- 37
- 25
- 20
15
Figure 4.2 SDS-PAGE of the purified fluorescently tagged proteins. The tagged
protein is labeled with an asterisk. (A) N-terminally his and GFP or mCherry toca-1,
expressed in 293F cells and purified using Ni-NTA agarose, washing and elution with
imidazole. N-terminally GFP-tagged N-WASP was purified as N-WASP/WIP complex
using co-expression of ZZ-tagged WIP and IgG sepharose. The abundant band at 40 kDa
labeled with an arrow is actin. (B) VCA-purified Arp2/3 complex. Double asterisk
indicates subunits that have similar mobility on this gel (4-12% gradient) (C,D) His-GFPmDia2, his-GFP VASP and his-GFP fascin were expressed and purified similarly to toca1.
Discussion
Using cell extracts, combined with supported lipid bilayers and TIRF technology,
our experiments gave original insights into how the spatial organization and dynamics of
signaling from the membrane controls the polymerization of actin into specific structures.
The ease of addition of labeled components for multicolor imaging and the fact that this
assay exploits events at the extract/membrane/glass interface means that we are poised to
use the full capability of TIRF microscope.
Monitoring spatiotemporal coordination of protein regulations using microscopy
is opening up a new way for studying complex cellular processes with minimal
perturbations. Usually, the development of reliable biosensors are the rate-limiting step in
this type of study. The fact that important events occur at the extract/membrane/glass
interface allows us to read out the activities by using conventional fluorescent proteins.
Here, using TIRF microscopy and only conventional fluorescent proteins, we could
observe the temporal coordination of various tip proteins in the time-scale of 10 seconds,
which lead us to a important mechanistic insight on FLS nucleation. Cells have a variety
of actin structures and filopodia in vivo seem to have multiple regulatory pathways. The
advantage of FLSs in this type of experiment is that in vitro systems are much more
simple than in vivo case so less heterogeneity of recruitment characteristics may make it
possible to capture this temporal coordination. In combination with inhibition/delepletion
strategies and mathematical framework to analyze the various kinetic data, monitoring
FLS tip assembly will reveal detailed mechanism of filopodia assembly.
Here, by measuring the time of recruitment, we can divide FLS tip proteins into
two kind of modules involved in FLS assembly. One group is recruited earlier and
composed of N-WASP, toca-1, and Arp2/3 complex which mediate actin networks. The
other one is recruited later and composed of mDia2, VASP, and fascin, known to be parts
of flipodial elongation machinery or tip complexes. This suggest that actin network
nucleation by Arp2/3 complex precede tip complex formation and there may be a
structural transitions from actin networks to bundles during FLS assembly. This
hypothesis will be further tested in the following chapter with different inhibitors.
Chapter 5
A Clustering-Outgrowth Model of Filopodia
Formation
ATTRIBUTIONS: This chapter contains the part of the manuscript "Self-Assembly of
Filopodia-Like Structure on Supported Lipid Bilayers, K. Lee, J. L Gallop, K. Rambani,
and M. W. Kirschner" submitted for the publication.
Introduction
Currently, there are two main models for filopodial assembly - convergent
elongation (Svitkina et al., 2003) and de novo nucleation (Faix et al., 2009). In the
former, Arp2/3 complex activity plays an important role in the filopodial initiation
process. Arp2/3-mediated actin polymerization leads to long actin filaments via anticapping activity by VASP.
Association of VASP proteins with the barbed ends of
filaments may mark F-actin for subsequent filopodium formation, presumably by
clustering barbed ends together, protecting them from capping, and permitting rapid
polymerization (Bear et al., 2002;Applewhite et al., 2007), and possibly bundling
filaments (Schirenbeck et al., 2006;Barzik et al., 2005). Fascin plays a role in bundling
these long actin filaments (Vignjevic et al., 2006).
Another possible mechanism involves de novo filament nucleation, likely by Drf3
or other formin proteins, the subsequent polymerization, and bundling of filaments. This
model proposes both distinct signaling pathways and separable core machinery driving
the formation of lamellipodia and filopodia (see also Faix and Rottner, 2006). In such a
scenario, initiation and continuous protrusion of a filopodium would be driven by de
novo nucleation of filopodial filaments, e.g. by a formin.
Therefore, it has been suggested that there are multiple pathways to regulate
filopodia assembly and different cell types can use different mechanisms to form
filopodia (Gupton and Gertler, 2007). Therefore, both models may be valid. However, the
group which originally came up with formin nucleation model still favor the idea that
there are only one formin dependent core machinery in filopodial formation (Faix et al.,
2009).
The experimental results in Chapter 2 and 4 suggests that both Arp2/3 complex
and formin are involved in FLS formation. Particularly, the kinetic study in Chapter 4
revealed the possible switching mechanism from Arp2/3 to formin mediated actin
assembly. In order to address this issue, the functional study using inhibitors will be
presented in this chapter.
Results
The kinetic data suggest a mechanism where an Arp2/3 complex-driven initiation
step nucleates an initial branched actin structure in a small patch and this stimulates the
recruitment of filament elongation and bundling factors. To test this mechanism further
we looked at the recruitment of Arp2/3 complex and mDia2 to toca-1 spots in the
presence of actin monomer sequestering drug, latrunculin B. The number of toca-1 spots
that have Arp2/3 complex are unaffected by latrunculin (Fig. 5.lA). In stark contrast,
latrunculin completely blocks toca-1-mDia2 colocalization (Fig. 5.1A).
To probe the role of the Arp2/3 complex we add GST-CA domain, which inhibits
N-WASP activation of the Arp2/3 complex, to the extract; this reduces the number of
nucleation sites and abolishes FLS elongation (Fig. 5.1B, Fig. 5.2A-B). Immunodepletion
of N-WASP significantly decreases but does not completely inhibit FLS formation and
elongation (Fig. 5.2D, H-I). This is consistent with studies in cultured cells suggesting the
involvement of other Arp2/3 complex nucleation-promoting factors (Snapper et al.,
2001;Lommel et al., 2001). Immunodepletion of toca-1 has only a minor effect on
elongation (Fig. 5.2E, H-I), however there are more than a dozen candidate BAR domain
proteins that could compensate for loss of toca-1. We conclude that signaling through the
Arp2/3 complex plays a vital role in the initiation of FLS formation, although these
proteins alone are not sufficient to generate FLSs (Fig. 2.1A and B).
The known product of Arp2/3 complex activation is an array of disorganized or
branched actin structures, rather than organized parallel actin bundles. The kinetics of
protein recruitment to the nascent FLSs suggests that a formin-driven elongation process
occurs after the formation of the first actin nucleus. To test whether the elongation phase
is independent of the Arp2/3 complex, we start the reaction with Alexa-647 actin and
Alexa-568 Arp2/3 complex and after 20 min add GST-CA to block further Arp2/3
complex function; Alexa-488 actin is added at the same time to record any further actin
polymerization (Fig. 5.1 C-H). We find that new actin monomers are still added at the
FLS tip after Arp2/3 complex inhibition (Fig. 5.1 C-F). Significantly, in the presence of
GST-CA, there is no further incorporatation of Arp2/3 complex into the FLS. The region
of newly assembled actin in the shaft lacks Arp2/3 complexes (Fig. 5.1E and F, Fig.
5.3A). After ~5 min addition of new actin to the FLS slows, indicating a requirement for
Arp2/3 complex to maintain FLS growth over the long run (Fig. 5.1G). At high
concentrations of GST-CA, there is a noticeable lag in the incorporation of new actin
monomers (Fig. 5.1G). This suggests that the occupancy of GST-CA on N-WASP
binding sites of the Arp2/3 complex stimulates a reorganization of the tip complex. Even
at high doses of GST-CA, the Arp2/3 complex-independent component of elongation is
maintained (Fig. 5.1H).
Diaphanous-related formins are the logical engine for filopodial growth in the
absence of the Arp2/3 complex, as formins are thought to drive filopodial elongation in
different cell types. Immunodepletion of the formin Drf3 from the extract, does not
significantly affect either initiation or elongation of the FLSs (data not shown). As there
are many formin proteins, any of which may be compensating, we are not be surprised
that any one of them would be unnecessary (Higgs et al., 2005). To circumvent the
problem with the large number of formins, employed a dominant-negative approach
known to inhibit filopodia formation (Eisenmann et al., 2007). The leucine-rich region
(LRR) of diaphanous interacting protein (DIP) binds and inhibits Drfl and Drf3
(Eisenmann et al., 2007). When GST-DIP-LRR is added to extracts, FLS formation still
occurs with no significant reduction in elongation rate or the number of structures.
However, when Arp2/3 complex function is inhibited by GST-CA in the presence of
GST-DIP-LRR, the number of structures that incorporates new actin monomers is
significantly reduced (Fig. 5.1I-M). This is accompanied by the detachment of many
FLSs from the lipid bilayer, suggesting that the tip complex undergoes conformational
changes, and that continuing reorganization of the tip complex cannot occur in the
presence of GST-LRR (Fig. 5.1M). Furthermore, as diaphanous-related formins are
activated by RhoA, we can also test dominant negative RhoA, GST-RhoA-N19. This
produces similar results to GST-DIP-LRR (Fig. 5.3B-D). Thus over the long run there is
a need for the continuing function of both Arp2/3 complex and formins. In the presence
of the high level of capping activities, there may be a requirement for the Arp2/3 complex
to generate new actin nuclei (Romero et al., 2004;Mejillano et al., 2004).
..
....
. .......
. ...................
..
140
120
100
so
6040-
0
Control
Lat. B
Contml
G0
Lat. B
z
30
40
GST-CA (jiM)
2D
18
Je
Time (min)
10
20
GST-CA (sM)
OM"f
GsT-ORM
Figure 5.1 Initiation and elongation of FLSs occur by separable molecular
mechanisms (A) Latrunculin B does not affect the co-localization of Arp2/3 complex
with toca-1, but completely inhibits mDia2 recruitment to toca- 1 sites. (B) Dose-response
of FLS initial elongation rate with GST-CA preincubated in the extract before addition to
the supported bilayer (measurement made after 7 min). The dotted-line indicates the
minimum elongation rate (0.1 pm/min) due to the axial resolution limit of confocal
microscope (-0.7 jim) (C-F) Pulse-chase experiment starting with Alexa-647 actin and
Alexa 568-labeled Arp2/3 complex in the extract, with later addition of 40 gM GST-CA
and Alexa-488 actin after 20 mins. Panels show Alexa 647-actin (C, blue); Alexa 488actin (D, green); Alexa-568-Arp2/3 complex (E, red) and three color overlay (F).
Addition of new actin monomers continues in the absence of the Arp2/3 complex
recruitment into the FLS Bars: 2 gm. (G) The time course of FLS elongation (measured
by the second color of actin) at increasing GST-CA concentrations. (H) Dependence of
maximum FLS elongation rate on the concentration of GST-CA shows that elongation
occurs independently of Arp2/3 complex activity. (I-M) Similar pulse-chase experiment
explained in (C-H), with the additional use of GST-LRR to inhibit formin activity. (I and
K) Control addition of GST-CA plus the second color of actin showing the z-stack
reconstruction in x-z. Alexa-647 (first, red) and Alexa-488 (second, green). (J and L)
Inclusion of 5 gM GST-LRR in the extract then addition of GST-CA and Alexa-488
actin. GST-LRR leads to the detachment of the FLSs so fewer punctae are present at the
membrane surface. Bars: 5 gm (I,J); 2 jm (K,L) (M) Quantification of GST-DIP-LRR
addition. *p<0.001. All error bars are s.d. The lipid composition was 45% PC, 45% PI,
10% PI(4,5)P2'
..
...........
...
...
Mock
N-WASP
Depletion
Mock
Toca-1
Depletion
800
am
Hf
hck
N-WASP
Toca-1
Toca-i
Figure 5.2 The effects of GST-CA and immunodepletion of N-WASP and toca-1 on
FLSs. The side view of FLSs at 7 min with 0 gM (A) and 40gM (B) GST-CA. The side
view of FLSs at 20 min with (C) mock (rabbit IgG), (D) N-WASP (E) toca-1 depletion.
Bars: 10 gm. (F-G) Western blots of depletions. (H) The number of focal actin structures
73
after depletion *p = 0.008. (I) The length of FLSs with depleted extracts, *p < 0.001, **p
= 0.001. All error bars are s.d. In this experiment the lipid composition was 45% PC,
45% P1, 10% PI(4,5)P 2 ; similar results are obtained with 60% PC, 30% PS and 10%
PI(4,5)P 2 '
.................
....
. ..................
. .. . ..
. ..........
.....
....
............
:..........
:................
.-
60
50
4
40
0
10
20
GST-CA (pM)
30
40
QC3#rd
GST4oA-NI9
Fig. 5.3 (A) Quantification of the Arp2/3 complex fluorescence of the GST-CA
inhibition experiment from Fig. 5.1 G-H. The fluorescence of Alexa568-Arp2/3
complex was measured 1gm from the bilayer and is eliminated by GST-CA. (B-F)
Inhibition of FLS Arp2/3 complex independent elongation by dominant-negative
RhoA. The number of FLSs with Arp2/3-complex-independent elongation was reduced
in the presence of 4 gM GST-RhoA-N 19, similarly to our findings with GST-LRR (Fig.
5.1I-M). (B) Quantification of GST-RhoA-N19 addition, *p = 0.016. Side views of first
actin in green and second in red for (C) control and (D) GST-RhoA-N 19 addition. Actin
on the bilayer, first in green and second in red for (E) control and (FE) GST-RhoA-N19
addition. Bars: 5 gm. All error bars are s.d. The lipid composition was 45% PC, 45% P1,
10% PI(4,5)P 2 -
Discussion
Currently there are two main models for filopodial assembly - convergent
elongation (Svitkina et al., 2003) and de novo nucleation (Faix et al., 2009). In the former
Arp2/3 complex driven actin assembly continually coalesces into a parallel shaft of actin
by the continuous action of bundling proteins like fascin. In the latter, filopodia are
fundamentally different from the beginning via the establishment of a tip complex of
formins, which produces long filaments which are then bundled. We propose a
clustering-outgrowth model for filopodia formation (Fig. 5.4), and a non-static tip
complex. In this model, signaling by a permissive lipid environment first activates the
Arp2/3 complex via the clustering of BAR domain proteins and N-WASP or other
nucleation promoting factors at the membrane, leading to a small patch of short actin
filaments. This represents the key difference from the convergent elongation model, as
symmetry breaking occurs by focal recruitment of Arp2/3 complex activating proteins
rather than by a coalescence of actin barbed ends. The local assembly of actin initiated by
the Arp2/3 complex is converted into a filopodial tip complex by the recruitment of
formins and VASP. This recruitment enables outgrowth of the filopodium, where the
short actin filaments are elongated by formins and/or VASP and bundled by fascin
(Vignjevic et al., 2003; Brieher et al., 2004). We believe that actin filaments generated by
the Arp2/3 complex are continually required to feed the elongation process, though we
would expect that other actin nucleators could fulfill a similar role. The observation of
short actin filaments at the tips of filopodia in Dictyostelium by electron tomography is
consistent with our model (Medalia et al., 2007). The twin processes of the clustering of
actin assembly proteins and outgrowth by elongation factors may be served by different
pathways: the overall mechanism of filopodia may be conserved but there may be
flexibility in their composition, as has previously been proposed (Gupton et al., 2007).
The approach used in this study provides a way to dissect the different qualitative and
quantitative attributes of filopodia and may help reveal their role as sensory organelles
during synapse formation, cell migration and morphogenesis.
...
.......
........
....
..
...
.......
Key
Sh
initiation proteins
eg toca, Arp2/3 complex
bundling protein
actin filament al eg
fascin
elongation proteins
eg forminsVASP
membrane
0* Si
Figure 5.4 A clustering-outgrowth model for filopodia formation. Stage 1: symmetry-
breaking via BAR domain proteins and stage 2: initial actin polymerization via N-WASP
and Arp2/3 complex, indicated by blue/green shapes. Stage 3: recruitment of elongation
factors and stage 4: bundling proteins, indicated by orange/pink shapes. The result is
focal actin protrusion, characteristic of filopodia.
Chapter 6
Conclusion
In this thesis, I have provided the experimental evidence that the reconstituted
FLSs share critical physiological aspects with the formation of filopodia, such as tip
complex assembly. EM and immuno-fluorescence studies have revealed that FLSs
contain parallel bundle structures with the known tip components of filopodia. The
results from FLSs are not only consistent with previous in vivo data but have also brought
new mechanistic insight to the underlying process of filopodia assembly.
FLSs seem to assemble the focal structures without direct instructive signals.
Instead, permissive PI(4,5)P 2 environments may allow the protein machinery to selforganize and generate focal FLSs. In general, the actin cytoskeleton can be seen as a
self-organizing system where minute inputs can induce massive structural changes and
spatial symmetry breaking can occur spontaneously. It has been shown that the
interaction between N-WASP and barbed ends of actin filaments is responsible for the
symmetry breaking of N-WASP distribution on the lipid coated glass beads (Co et al.,
2007), suggesting that N-WASP and the Arp2/3 complex system has autocatalytic
activity (positive feedback) which leads to local excitation of actin nucleation. However,
the fact that toca-1 binding to membrane significantly preceded the recruitment of NWASP from the kinetic experiments suggests a central role of toca-1 in this focal
structure formation, which has not been reported. It is possible that there are multiple
amplification steps involved in this focal structure formation. First, initial toca-1 forms
clusters via an unknown mechanism, serving as scaffolds for N-WASP recruitment.
Then, N-WASP and Arp2/3 undergo a positive feedback loop inducing further
amplification, leading to highly focal structures.
Subsequently, this intermediate
structure will undergo reorganization by filopodial elongation machinery.
FLSs are assembled via a two-stage process where the initial actin networks
stimulate tip complex assembly, leading to parallel actin bundles.
The kinetic data
presented in Chapter 4 and inhibition experiments done in Chapter 5 are all consistent
with this two-stage model of filopdial assembly. The Arp2/3 complex is known to
promote actin network structures by binding to pre-existing actin filaments. However, the
Arp2/3 complex is absolutely required for FLS initiation even if FLS has actin bundles.
This suggests that there may be structural transitions from actin networks to bundles. In
vitro, fascin is known to contribute to this kind of transition (Vignjevic et al., 2003;Ideses
et al., 2008). When the energy gain from bundling overcomes the bending energy of the
actin filaments in the networks, fascin can mediate such transitions (Ideses et al., 2008).
Moreover, it was reported that elastic interaction between the membrane and actin
filaments can substitute for fascin's role and induce the same transitions (Liu et al.,
2008). Another possibility for Arp2/3's role is that Arp2/3 mediated actin networks
provide the nucleation sites of FLS where tip complexes including formin are assembled,
allowing de novo bundle generation. Here, fascin's role may be merely to further stabilize
actin bundles. In this scenario, the structural transition is the result of biochemical
switching from Arp2/3 to other actin elongation activities, possibly mediated by formin.
The kinetic experiments done in Chapter 4 highly suggest this kind of biochemical
switching should contribute to FLS assembly. Therefore, during filopodial assembly there
could be two distinct mechanisms to drive the structural transitions. One is a biophysical
transition where fascin's mechanical properties play a central role. The other is a
biochemical transition where the tip complex generates de novo bundles from the
networks.
How N-WASP and the Arp2/3 complex contribute to filopodia formation has
been controversial. The original convergent elongation model and the subsequent study
support a significant role of Arp2/3 complex in filopodial assembly (Svitkina et al.,
2003;Korobova and Svitkina, 2008). However, other studies suggested that the Arp2/3
complex is dispensable and that instead formin plays an essential role. These studies
further suggest that fomin rather than the Arp2/3 complex comprises the core machinery
of filpodia. This is the formin nucleation model (Steffen et al., 2006). As for N-WASP,
the activator of Arp2/3 complex, there has been a similar dispute for N-WASP's role in
filopodia formation. In this thesis, I presented evidence where N-WASP and Arp2/3
driven actin networks were transformed into bundles during FLS formation, consistent
with the convergent elongation model. On the other hand, our kinetic and inhibition
experiments also suggest that the roles of N-WASP and Arp2/3 complex are distinct from
those of tip complex in FLS assembly. Explosive local actin polymerization driven by NWASP and Arp2/3 complex, shown to be a prerequisite for tip complex formation by
latrunculin experiments, can provide scaffolds for tip complex recruitment, resulting in
numerous filopodia. However, it is also possible that the compensatory mechanisms of
the cell in the absence of N-WASP and/or Arp2/3 complex can produce such local actin
scaffolds to bring elongation machinery to the site of filopodia assembly.
By showing the collaborative relation between the Arp2/3 complex and formin,
we think our two-stage model unifies the two current competing models which have been
previously thought to be incompatible. While our model has some similarity to the
convergent elongation model , the symmetry breaking (focal structure formation) step
occurs before any actin assembly at that site while Arp2/3 mediated actin assembly
induces symmetry breaking in that model. Therefore our model does not support the type
of actin structural transformation suggested by convergent elongation model, but
emphasizes the importance of membrane-binding proteins such as BAR-domain proteins
for the initiation. Unlike the formin nucleation model where the tips pre-exist and are
well-defined, our model suggests that the tip is instead assembled through several phases
during the formation of filopodia.
In summary, the use of supported lipid bilayers in conjunction with extracts
creates an informative biochemical system that reveals the mechanisms of filopodia
formation. The reconstituted filopodia-like structures will allow us to study not only
biochemical nature, but also the dynamic aspects of filopodial assembly. This work
opens up the new avenues toward the reconstitution of filopodial tips using purified
components leading to a comprehensive biochemical understanding from initial signaling
to filopodial assembly.
Chapter 7
Materials and Methods
Extract preparation
High speed Xenopus egg extracts were prepared as previously described with
modifications of the centrifugation conditions (Lebensohn et al, 2006). Briefly, low speed
CSF-arrested extracts were prepared by crushing dejellied Xenopus eggs in CSF-XB (10
mM K-HEPES, pH 7.7, at 16'C, 100 mM KCl, 2 mM MgCl 2 , 50 mM sucrose, 5 mM
EGTA) supplemented with protease inhibitors and 1mM DTT at 10,000g for 10 min at
4'C. Then, the low speed extracts were diluted 10-fold in the same buffer overlaid with
mineral oil and then centrifuged at 200,000g for 1 h at 4'C to remove the internal
membranes. The clear supernatant was reconcentrated to its original volume in
Centriprep YM-10 concentrators (Milipore 4304). The usual final concentration is about
25 mg/ml. High speed extracts were supplemented with 200 mM sucrose and CSF-energy
mix containing 1 mM ATP, 1 mM MgCl 2 , and 7.5 mM creatine phosphate, snap frozen
and stored at -80'C.
Plasmid construction
In order to construct plasmids for fluorescent fusion proteins, we obtained
Xenopus full-length Xenopus VASP clone from Open Biosystems. Mouse mDia2
plasmid was a gift from Arthur Alberts, human fascin plasmid from Danijella Vignjevic,
and mCherry plasmid from Roger Tsien. Human Cdc42, toca- 1 (Ho et al, 2006), mouse
mDia2, Xenopus VASP , and human fascin were subcloned into pCS2 vector with Nterminal hexahistadine tag and eGFP/mCherry. EcoR I and Xho I restricton sites were
used for Cdc42, toca-1, VASP, and fascin, BspE I and Kpn I for mDia2, Bovine N-
WASP was subcloned into pCS2 vector without tag using Fse I and Asc I restriction sites.
Human WIPI was subcloned into pCS2 vector with N-terminal zz-tag and TEV protease
site. For the non-fluorescent version, human Cdc42 was subcloned into pCS2 vector with
N-terminal hexahistadine tag. Diaphanous interacting protein-LRR fragment (residues
507-722) was prepared by PCR using Human full-length DIP obtained from Open
Biosystems
and
primers,
gatcggatccatcctggccatggtcttctc
and
gatcctcgagctagctgggagcctccccca and subcloned into BamHI and Xho I sites of pGEX
vector. This region does not contain the reported Arp2/3 complex activation sequence
(Kim et al, 2006). The GST-CA plasmid has been previously described (Miki et al,
1996). GST-RhoA-N19 plasmid was a gift from Gary Bokoch (Addgene plasmid
12960). pEGFP GFP-PLC6 PH domain plasmid was a gift from Seth Field and was
subcloned using EcoRI and NotI restriction sites into pGEX-4T2. RhoGDI plasmid for
mammalian expression was a gift from Orion Weiner.
Protein purification and labeling
N-WASP-WIP, toca- 1, and Arp2/3 complex were purified as previously described
(Ho et al, 2006). Cdc42-RhoGDI, mCherry-Cdc42-RhoGDI,
GFP-N-WASP-WIP,
GFP/mCherry-toca-1, GFP-mDia2, GFP-VASP, and GFP-fascin were expressed in 293F
cells using 293Fectin (Invitrogen) for 2 days. GFP/mCherry-toca-1 was purified in the
same way as unlabeled toca-1. For the other proteins, the culture was resuspended in
phosphate-buffered saline (PBS) with 0.5 mM DTT, protease inhibitor tablets (Roche),
and 10 mM imidazole. For GFP-mDia2, 1%NP-40 was supplemented. For GFP-fascin,
0.5% Triton X-100 was supplemented. For GFP-N-WASP-WIP, it was in PBS with 1
mM DTT, protease inhibitor tablets.After the sonication, the lysate was cleared by
centrifugation at 200,000g for 30 min at 4'C and incubated with Ni-NTA-agarose beads
(Qiagen) for 2 hours at 4'C. For GFP-N-WASP-WIP, the cleared lysate was incubated
with IgG sepharose beads (GE Healthcare) for 3 hours at 4'C. For the his-tagged GFP
fusion proteins (GFP-mDia2, GFP-VASP, and GFP-fascin), the beads were washed with
cold PBS with 0.5 mM DTT and 20 mM imidazole, eluted with 300 mM imidazole in
PBS with 0.5 mM DTT, and dialyzed against XB (20 mM HEPES, pH 7.6, 100 mM KCl,
1 mM MgCl 2, 0.1 mM EDTA, 1 mM DTT) with 10 % Glycerol. For GFP-fascin, the
dialysis was done against PBS supplemented 1 mM DTT and 10% Glycerol. For GFP-NWASP-WIP, the beads were washed with cold PBS with 1mM DTT and incubated with
0.03mg/ml GST-TEV protease in the same buffer for 1 hour at room temperature,
followed by removal of GST-TEV protease using glutathione-sepharose beads (GE
Healthcare) and dialysis against XB with 10% Glycerol. pGEX GST-CA, RhoA-N19
vector were transformed into BL21 Codon Plus (DE-3)-RP (Stratagene). The bacteria
were grown in LB media at 37'C until OD600 reached 0.6, then induced with 0.5mM
IPTG and incubated at 24 'C overnight. The culture was resuspended in PBS with 1 mM
DTT, protease inhibitor tablets (Roche), and 1 mg/ml lysozyme and incubated for 20 min
on ice before sonication. The lysate was cleared by centrifugation and incubated with
glutathione-sepharose beads (GE Healthcare) for 4 hours at 4 'C. The beads were washed
with cold PBS with 1 mM DTT, eluted with 50 mM glutathione in PBS with 1 mM DTT,
and dialyzed against XB with 10% Glycerol. GST-LRR and GST-GFP-PLC6 PH domain
were transformed into the same BL21s, grown in TB media until log phase, induced with
0.5 mM IPTG and then incubated at 19 'C overnight. The purification was performed
similarly to the other GST tagged proteins, except 150 mM NaCl, 20 mM HEPES pH 7.4,
2 mM EDTA and 2 mM DTT was used instead of PBS. In the case of RhoA-N19, all the
buffers were supplemented with 1 mM MgCl 2 . The labeling of Arp2/3 complex with
Alexa568-maleimide (Invitrogen) has been previously described (Zalevsky et al, 2001).
Alexa 488 or 647 labeled actin was purchased from Invitrogen. All proteins were snap
frozen and stored at -80 'C.
Liposome preparation
Porcine brain PC, bovinc liver PI, porcine brain PS and porcine brain PI(4,5)P
2
were used for most experiments. Di-oleoyl phosphoinositides were used for the PIP
specificity
experiments
and
protonated
before
use
by
resuspension
in
chloroform:methanol: water 20:9:1, addition of water acidified with HCl to pH 2.5, and
taking the lower chloroform layer. These lipids, rhodamine-PE and TopFluorPI(4,5)P
2
were purchased from Avanti Polar Lipids. Fluorescent lipids were used at 1%. To make
liposomes, lipid mixtures were rapidly dried in glass tubes under a stream of dry nitrogen
gas and further dried under vacuum for lhr to remove chloroform completely. The dried
lipid mixture was hydrated in XB buffer to a final concentration of 2 mM, bath sonicated
for 1 min and then filtered using a mini-extruder sequentially through 800 nm then 100
nm pore-size polycarbonate membranes (Whatman). DiI (1,1'-dioctadecyl-3,3,3',3'tetramethylindocarbocyanine perchlorate) was purchased from Invitrogen.
Antibody preparation
To raise a Xenopus fascin antibody, the full-length Xenopus fascin clone was
obtained from Open Biosystems and subcloned into the pGEX vector. Using BL21 as
above, Xenopus fascin was expressed and purified as previously described (Vignjevic et
al, 2006). Purified full-length Xenopus fascin was used to raise antisera in rabbits
(Cocalico, Reamstown, PA). The antibodies were affinity purified using the same fascin
proteins according to Harlow et al, 1999. Drfl and profilin antibodies were purchased
from Axxora. The N-WASP antibody has been previously described (Rohatgi et al,
1999).
FLS assays
To make the supported bilayers, No. 1.5 glass coverslips were incubated with
freshly prepared liposomes containing 45 % PC, 45 % PI, and 10 % PI(4,5)P 2 in XB
buffer for 20 min at room temperature, followed by extensive washing with XB buffer.
Membrane phase separation was variable and was largely influenced by the particular
batch of glass. Rigorous washing of the coverslips with hot detergent also promoted the
liquid disordered phase. All assays were carried out at room temperature (-22 C). For
the purified system experiments, prenylated Cdc42.GTPyS was supplied to the lipid
bilayer from 100 nM Cdc42-RhoGDI in solution using the EDTA exchange reaction
(Read et al, 2000). The reaction mixture including N-WASP-WIP, toca-1, Arp2/3
complex, and actin as previously described (Ho et al, 2006) was added after Cdc42
loading. Typical FLS reactions (50 ptl volume) contained a 2-fold dilution of Xenopus
egg extract, 4 ptM Alexa 647 actin (10% labeling efficiency, rabbit skeleton muscle
actin), 0.35 M sucrose, 1 mM ATP, 1 mM MgCl 2, 7.5 mM phosphocreatine in XB buffer.
The reaction mixtures were added on top of the freshly prepared supported bilayer and
monitored with a spinning disk confocal microscope. For the pulse chase experiments,
the second reaction (5 pl volume) Xenopus egg extract, 12 gM Alexa-488 actin (5%
labeling efficiency, rabbit skeletal muscle actin), 1 mM ATP, 7.5 mM phosphocreatine in
XB and 5 ptl was added gently on top of the first reactions. For dose response of FLS
initial elongation, the reaction mixture was supplement with different dose of GST-CA
and images were taken after 7 min. For Arp2/3 complex independent elongation
experiments, the first reaction was supplemented with 50 nM Alexa568-Arp2/3 complex.
The second reaction is assembled similarly to the pulse-chase experiments with different
doses of GST-CA. For the GFP-PH domain experiments, 50-300 nM was used. For
timelapse movies of FLS growth, an oxygen scavenger mix was added which contained:
4.5 mg/ml glucose, 0.5 % 2-mercaptoethanol, 0.2 mg/ml glucose oxidase (SigmaAldrich), 35 jg/ml catalase (Sigma-Aldrich).
Light microscopy
Microscopy for chapters 2 and 5 was performed using an inverted Nikon
TE2000U microscope with a 100x, 1.4 NA Plan Apochromat objective lens and
motorized stage and focus motor from Prior. Confocal images were obtained using a
Yokogawa CSU-10 spinning disk confocal head with Prairie laser launch with a 2.5 W
water-cooled Coherent Argon-Krypton laser. Excitation and emission wavelengths were
selected and attenuated with an AOTF and a triple 488/568/647 dichroic mirror from
Chroma. GFP and Alexa-488 were visualized using the 488 laser line and 525/50
emission filter; Alexa-568 was visualized using the 568 laser line and 600/45 emission
filter; Alexa-647 was visualized by the 647 laser line and 700/75 emission filter
(Chroma). Images were collected with a ORCA-AG cooled CCD camera from
Hamamatsu and Metamorph software v7.6 (Molecular Devices). Exposure times were
typically 100~400 ms using 25-50% laser power and a bin of 2x2. Z-stacks were
collected with a step size of 0.5 im. Light microscopy for chapter 3 and 4 was performed
using an inverted Nikon Ti-E microscope with a 100x, 1.4 NA Plan Apochromat
objective lens and motorized stage from Prior. Confocal images were obtained using
a Yokogawa CSU- 10 spinning disk confocal head with 100 mW Argon-Krypton laser
from Melles Griot. Excitation and emission wavelengths were selected using Sutter filter
wheels and a triple 488/568/647 dichroic mirror from Chroma. Images were collected
with an ORCA-ER cooled CCD camera from Hamamatsu and Metamorph software v7.6
(Molecular Devices). GFP was visualized using the 488 laser line selected with a 488/10
excitation filter and 525/50 emission filter; rhodamine and Alexa-568 were visualized
using the 568 laser line selected with a 568/10 excitation filter and 620/60 emission filter;
Alexa-647 was visualized by the 647 laser line selected with a 647/10 filter, and 647/10
emission filter (Chroma). Exposure times were typically 200 ms using a bin of 2x2. For
time-lapse experiments of FLS initiation, the Perfect Focus System (Nikon) was used to
maintain focus, and images were acquired every 10 s for 10 minutes. Z-stacks were
acquired with a step size of 1 pm. For the fluorescence recovery after photobleaching
experiments of the supported bilayer, wide-field epifluorescence illumination was used
(with a Hamamatsu ORCA-R2 cooled CCD camera and an X-Cite series 120 light
source) and
rhodamine-PE was photobleached to 80-90% of initial intensity with 515 nm light from a
nitrogen pulse laser (Photonic Instruments Micropoint system) focused to a spot less than
1 micron in diameter. The filter was Y-2E/C (excitation: 560/40 dichroic: 595 emission;
630/60) from Nikon. The exposure time was 25 ms, and images were typically acquired
every 1 s for 1-20 min. For the multispectral total internal reflection fluorescence
microscopy in chapter 4, we used a Nikon Ti-E inverted motorized microscope with
integrated Perfect Focus System, Nikon 100x 1.49 NA TIRF DIC objective lens, Nikon
halogen trans illuminator with 0.52 NA LWD and 0.85 NA Dry condenser, Nikon dualport TIRF/Epi illuminator with motorized laser incident angle adjustment and motorized
switching between TIRF and epi-illumination. For lasers, a Solamere laser launch was
used with 100mW 491nm, 75mW 561nm and 30mW 640nm solid state lasers with a
fiber-optic delivery system and 4-channel AOTF. A Prior Proscan II controller was used
for fast excitation and emission filter wheels, fast transmitted and epi-fluorescence light
path shutters, and a linear-encoded motorized stage. A Chroma zet405/491/561/638
dichroic mirror was used with a 491nm laser line and a 525/50 emission filter for GFP; a
561 laser line and 600/50 emission filter for Alexa568; and a 640 laser line and a 700/75
emission filter for Alexa647 . In addition to emission filters, a custom Chroma laser notch
filter was used in the emission path to further block the illumination light from reaching
the camera and to minimize interference patterns. Images were collected with a
Hamamatsu ImagEM 512x512 back-thinned electron multiplying cooled CCD camera
and MetaMorph v7.7 (Molecular Devices). Exposure times were typically ~100 ms using
25-50% laser power.
Electron microscopy
In order to prevent the fragmentation of FLSs from the convective flow, it is
important to add sufficient amount of sucrose to the reactions and add stabilization
solution on top of the reaction very gently and let it diffuse into the reaction. Typically 11
% (w/v) sucrose is included in the reaction FLSs are stabilized and arrested by the
incubation with 20gM each of phalloidin and latrunculin B for 1 min (Akin and Mullins,
2008), followed by repeated gentle dilution using XB buffer to remove soluble proteins
and then fixed with 0.1% glutaraldehyde in XB for 20min. Use of negative stain to
visualize actin structures has been described previously (Auinger and Small, 2008).
Briefly, fixed FLSs were rinsed three times in XB, and the actin filaments were furthur
stabilized by incubation in 10 gg/ml phalloidin in the same buffer for at least 20 min or
until use. To detach FLSs, the glass surface was scratched using a scalpel or a pipet tip.
The detached FLS were adsorbed onto glow-discharged formvar-carbon coated grids, and
negatively stained with aqueous 3 % sodium silico-tungstate. The electron microscope
was a Tecnai G2 Spirit BioTWIN electron microscope operating at 80 kV. For unfixed
FLSs, they were first stabilized and arrested by 20 gM each of phalloidin and latrunculin
B for 1 min. Then, soluble proteins were removed by repeated gentle dilution with XB.
These FLSs were easily detached from the lipid bilayer by pipetting for the same negative
staining. Alternative buffer conditions or additives for FLS fixation are also described
elsewhere (Svitkina, 2007; Auinger and Small, 2008).
Immunostaining
FLSs were fixed using 4 % formaldehyde in CB for 40 min. It is important to add
the fixation mixture very gently on the reactions. Typically, 200 pl of the fixation
solution was added to a 50 pl reaction. 2 % BSA was used for blocking, the FLSs were
incubated with primary antibody or anti-serum (1:100 dilution), followed by incubation
with Alexa 488-conjugated goat anti-rabbit secondary antibody (1:200 dilution)
(Invitrogen). After extensive washing, actin was stained with Alexa-568 conjugated
phalloidin (Invitrogen).
Immunodepletion
Immunodepletion of N-WASP and toca- 1 from Xenopus egg extracts is described
elsewhere (Lebensohn et al, 2006).
Image Analysis
Image analysis was performed using MetaMorph (Molecular Devices). FLS
elongation was measured using reconstructed side view images. The linescan method in
MetaMorph was used measure actin fluorescence intensity along the tails. In the cases of
pulse chase and Arp2/3 complex independent elongation experiments, the second actin
signals were divided by the first actin signals along the FLS tail lengths. The elongation
of the second color of actin was measured at the point it disappeared. We sampled 10-20
FLSs per field of view at each time point to calculate mean and standard deviation of
elongations. The standard deviation of elongation rate between time t, and t2 is calculated
as
a "+ a ' (t 2 - t)
where o-, and a2 are the standard deviations of FLS elongation at
time t, and t2 under the assumption that the measurements at t and t2 are statistically
independent. For quantitation of Alexa568-Arp2/3
complex in Arp2/3 complex
independent elongation experiments, the fluorescence at 1 gm from bilayers was
measured from reconstructed side view images of Arp2/3 complex and was subtracted by
the nearby background fluorescence. For quantitation of the density and rate of
appearance of FLS nuclei, the Transfluor quantitation tool of MetaMorph was used . The
typical parameter values of the analysis were a threshold of 200 intensity values and a
size range of 0.5-10 ptm. In order to quantify the diameter of FLSs, we used Cell Scoring
tool of Metamorph for segmentation and the cross-sectional are of FLSs were measured
using Morphometry Analysis tool. Then, the effective diameters of FLSs were calculated
as 2 A/r where A is a FLS cross-sectional area. For the fluorescent protein recruitment
experiments, FLSs were identified at the end of the time-lapse sequence and by backtracking and observation by eye, the first frame where the fluorescence for actin and the
candidate
protein
is
higher
than
background
was
noted.
The
website
www.physics.csbsju.edu/stats was used for ANOVA, KS and t-tests. Microsoft Excel was
used to make the graphs and contour plot.
Chapter 8
References
Applewhite, DA., Barzik, M., Kojima, S. et al. (2007). Ena/VASP proteins have an anticapping independent function in filopodia formation. Mol. Biol. Cell 18, 2579-2591.
Arasada, R., Gloss, A., Tunggal, B. et al. (2007). Profilin isoforms in Dictyostelium
discoideum. Biochim. Biophys. Acta 1773, 631-641.
Aratyn, YS., Schaus, TE., Taylor, EW. et al. (2007). Intrinsic dynamic behavior of fascin
in filopodia. Mol. Biol. Cell 18, 3928-3940.
Auinger, S. and Small, JV. (2008). Correlated light and electron microscopy of the
cytoskeleton. Methods Cell Biol. 88, 257-272.
Axelrod, D. (2001). Total internal reflection fluorescence microscopy in cell biology.
Traffic 2, 764-774.
Barzik, M., Kotova, TI., Higgs, HN. et al. (2005). Ena/VASP proteins enhance actin
polymerization in the presence of barbed end capping proteins. J. Biol. Chem. 280,
28653-28662.
Bear, JE., Svitkina, TM., Krause, M. et al. (2002). Antagonism between Ena/VASP
proteins and actin filament capping regulates fibroblast motility. Cell 109, 509-521.
Bentley, D. and Toroian-Raymond, A. (1986). Disoriented pathfinding by pioneer
neurone growth cones deprived of filopodia by cytochalasin treatment. Nature 323, 712715.
Boureux, A., Vignal, E., Faure, S. et al. (2007). Evolution of the Rho family of ras-like
GTPases in eukaryotes. Mol. Biol. Evol. 24, 203-216.
Breitsprecher, D., Kiesewetter, AK., Linkner, J. et al. (2008). Clustering of VASP
actively drives processive, WH2 domain-mediated actin filament elongation. EMBO J.
27, 2943-2954.
Bridgman, PC., Dave, S., Asnes, CF. et al. (2001). Myosin IIB is required for growth
cone motility. J.Neurosci. 21, 6159-6169.
Brieher, WM., Coughlin, M. and Mitchison, TJ. (2004). Fascin-mediated propulsion of
Listeria monocytogenes independent of frequent nucleation by the Arp2/3 complex. J.
Cell Biol. 165, 233-242.
Buccione, R., Orth, JD. and McNiven, MA. (2004). Foot and mouth: podosomes,
invadopodia and circular dorsal ruffles. Nat. Rev. Mol. Cell Biol. 5, 647-657.
Bums, S., Thrasher, AJ., Blundell, MP. et al. (2001). Configuration of human dendritic
cell cytoskeleton by Rho GTPases, the WAS protein, and differentiation. Blood 98, 11421149.
Cameron, LA., Footer, MJ., van Oudenaarden, A. et al. (1999). Motility of ActA proteincoated microspheres driven by actin polymerization. Proc. Natl. Acad. Sci. U.S.A. 96,
4908-4913.
Carlier, MF. (1988). Role of nucleotide hydrolysis in the polymerization of actin and
tubulin. Cell Biophys 12, 105-117.
Carlier, MF. (1998). Control of actin dynamics. Curr. Opin. Cell Biol. 10, 45-51.
Chan, YM. and Boxer, SG. (2007). Model membrane systems and their applications. Curr
Opin Chem Biol 11, 581-587.
Chang, F., Drubin, D. and Nurse, P. (1997). cdcl2p, a protein required for cytokinesis in
fission yeast, is a component of the cell division ring and interacts with profilin. J. Cell
Biol. 137, 169-182.
Co, C., Wong, DT., Gierke, S. et al. (2007). Mechanism of actin network attachment to
moving membranes: barbed end capture by N-WASP WH2 domains. Cell 128, 901-913.
Condeelis, J., Singer, RH. and Segall, JE. (2005). The great escape: when cancer cells
hijack the genes for chemotaxis and motility. Annu. Rev. Cell Dev. Biol. 21, 695-718.
100
Crane, JM. and Tamm, LK. (2004). Role of cholesterol in the formation and nature of
lipid rafts in planar and spherical model membranes. Biophys. J. 86, 2965-2979.
Davenport, RW., Dou, P., Rehder, V. et al. (1993). A sensory role for neuronal growth
cone filopodia. Nature 361, 721-724.
De La Cruz , EM., Mandinova, A., Steinmetz MO., Stoffler, D., Aebi, U., and Pollard
TD. (2000). Polymerization and structure of nucleotide-free actin filaments. J. Mol. Biol.
295, 517-526.
de Rooij, J., Kerstens, A., Danuser, G. et al. (2005). Integrin-dependent actomyosin
contraction regulates epithelial cell scattering. J. Cell Biol. 171, 153-164.
DeRosier, DJ. and Edds, KT. (1980). Evidence for fascin cross-links between the actin
filaments in coelomocyte filopodia. Exp. Cell Res. 126, 490-494.
DesMarais, V., Macaluso, F., Condeelis, J. et al. (2004). Synergistic interaction between
the Arp2/3 complex and cofilin drives stimulated lamellipod extension. J. Cell. Sci. 117,
3499-3510.
DesMarais, V., Ghosh, M., Eddy, R. et al. (2005). Cofilin takes the lead. J. Cell. Sci. 118,
19-26.
101
Destaing, 0., Saltel, F., G6minard, J. et al. (2003). Podosomes display actin turnover and
dynamic self-organization in osteoclasts expressing actin-green fluorescent protein. Mol.
Biol. Cell 14, 407-416.
DiNubile, MJ., Cassimeris, L., Joyce, M. et al. (1995). Actin filament barbed-end capping
activity in neutrophil lysates: the role of capping protein-beta 2. Mol. Biol. Cell 6, 16591671.
Drubin, DG., and Nelson, WJ. (1996). Origin of cell polarity. Cell. 84, 335-344.
Eisenmann, KM., Harris, ES., Kitchen, SM. et al. (2007). Dia-interacting protein
modulates formin-mediated actin assembly at the cell cortex. Curr. Biol. 17, 579-591.
Faix, J., Breitsprecher, D., Stradal, TEB. et al. (2009). Filopodia: Complex models for
simple rods. Int. J. Biochem. Cell Biol. 41, 1656-1664.
Feigenson, GW. (2009). Phase diagrams and lipid domains in multicomponent lipid
bilayer mixtures. Biochim. Biophys. Acta 1788, 47-52.
Fiala, JC., Feinberg, M., Popov, V. et al. (1998). Synaptogenesis via dendritic filopodia
in developing hippocampal area CAl. J. Neurosci. 18, 8900-8911.
Field, C., Li R., and Oegema K. (1999). Cytokinesis in eukaryotes: a mechanistic
comparison, Curr. Opinion in Cell Biol. 11:68-80.
102
Fletcher, DA. and Mullins, RD. (2010). Cell mechanics and the cytoskeleton. Nature 463,
485-492.
Galbraith, CG., Yamada, KM. and Sheetz, MP. (2002). The relationship between force
and focal complex development. J. Cell Biol. 159, 695-705.
Gaus, K., Gratton, E., Kable, EPW. et al. (2003). Visualizing lipid structure and raft
domains in living cells with two-photon microscopy. Proc. Natl. Acad. Sci. U.S.A. 100,
15554-15559.
Geiger, B., Bershadsky, A., Pankov, R. et al. (2001). Transmembrane crosstalk between
the extracellular matrix--cytoskeleton crosstalk. Nat. Rev. Mol. Cell Biol. 2, 793-805.
Ghosh, M., Song, X., Mouneimne, G. et al. (2004). Cofilin promotes actin polymerization
and defines the direction of cell motility. Science 304, 743-746.
Goodhill, GJ., Gu, M. and Urbach, JS. (2004). Predicting axonal response to molecular
gradients with a computational model of filopodial dynamics. Neural Comput 16, 22212243.
Gungabissoon, RA. and Bamburg, JR. (2003). Regulation of growth cone actin dynamics
by ADF/cofilin. J. Histochem. Cytochem. 51, 411-420.
103
Gupton, SL., Anderson, KL., Kole, TP. et al. (2005). Cell migration without a
lamellipodium: translation of actin dynamics into cell movement mediated by
tropomyosin. J. Cell Biol. 168, 619-63 1.
Gupton, SL. and Gertler, FB. (2007). Filopodia: the fingers that do the walking. Sci.
STKE 2007, re5.
Hall, A. (1998). Rho GTPases and the actin cytoskeleton. Science. 279, 509-514.
Harlow, E. and Lane, D. Using Antibodies: A Laboratory Manual. . Cold Spring Harbor
Laboratory Press, Cold Spring Harbor, NY, 1999.
Heidemann, SR., Lamoureux, P. and Buxbaum, RE. (1990). Growth cone behavior and
production of traction force. J. Cell Biol. 111, 1949-1957.
Higgs, HN. and Peterson, KJ. (2005). Phylogenetic analysis of the formin homology 2
domain. Mol. Biol. Cell 16, 1-13.
Ho, HY., Rohatgi, R., Ma, L. et al. (2001). CR16 forms a complex with N-WASP in
brain and is a novel member of a conserved proline-rich actin-binding protein family.
Proc. Natl. Acad. Sci. U.S.A. 98, 11306-11311.
Ho, HH., Rohatgi, R., Lebensohn, AM. et al. (2004). Toca-1 mediates Cdc42-dependent
actin nucleation by activating the N-WASP-WIP complex. Cell 118, 203-216.
104
Ho, HH., Rohatgi, R., Lebensohn, AM. et al. (2006). In vitro reconstitution of cdc42mediated actin assembly using purified components. Meth. Enzymol. 406, 174-190.
Hodgson, L., Pertz, 0. and Hahn, KM. (2008). Design and optimization of genetically
encoded fluorescent biosensors: GTPase biosensors. Methods Cell Biol. 85, 63-81.
Hoffman, GR., Nassar, N. and Cerione, RA. (2000). Structure of the Rho family GTPbinding protein Cdc42 in complex with the multifunctional regulator RhoGDI. Cell 100,
345-356.
Huxley, HE. (1985). The crossbridge mechanism of muscular contraction and its
implications. J Exp. Biol. 115, 17-30.
Ichetovkin, I., Grant, W. and Condeelis, J. (2002). Cofilin produces newly polymerized
actin filaments that are preferred for dendritic nucleation by the Arp2/3 complex. Curr.
Biol. 12, 79-84.
Ideses, Y., Brill-Karniely, Y., Haviv, L. et al. (2008). Arp2/3 branched actin network
mediates filopodia-like bundles formation in vitro. PLoS ONE 3, e3297.
Itoh, T., Erdmann, KS., Roux, A. et al. (2005). Dynamin and the actin cytoskeleton
cooperatively regulate plasma membrane invagination by BAR and F-BAR proteins.
Dev. Cell 9, 791-804.
105
Jacinto, A., Wood, W., Balayo, T. et al. (2000). Dynamic actin-based epithelial adhesion
and cell matching during Drosophila dorsal closure. Curr. Biol. 10, 1420-1426.
Johnston, SA., Bramble, JP., Yeung, CL. et al. (2008). Arp2/3 complex activity in
filopodia of spreading cells. BMC Cell Biol. 9, 65.
Kamm, KE. and Stull, JT. (2001). Dedicated myosin light chain kinases with diverse
cellular functions. J. Biol. Chem. 276, 4527-4530.
Kim, DJ., Kim, SH., Lim, CS. et al. (2006). Interaction of SPIN90 with the Arp2/3
complex mediates lamellipodia and actin comet tail formation. J. Biol. Chem. 281, 617625.
Kocks, C., Marchand, JB., Gouin, E. et al. (1995). The unrelated surface proteins ActA of
Listeria monocytogenes and IcsA of Shigella flexneri are sufficient to confer actin-based
motility on Listeria innocua and Escherichia coli respectively. Mol. Microbiol. 18, 413423.
Korobova, F. and Svitkina, T. (2008). Arp2/3 complex is important for filopodia
formation, growth cone motility, and neuritogenesis in neuronal cells. Mol. Biol. Cell 19,
1561-1574.
106
Kovar, DR. (2006). Molecular details of formin-mediated actin assembly. Curr. Opin.
Cell Biol. 18, 11-17.
Kovar, DR., Harris, ES., Mahaffy, R. et al. (2006). Control of the assembly of ATP- and
ADP-actin by formins and profilin. Cell 124, 423-435.
Krause, M., Dent, EW., Bear, JE. et al. (2003). Ena/VASP proteins: regulators of the
actin cytoskeleton and cell migration. Annu. Rev. Cell Dev. Biol. 19, 541-564.
Krugmann, S., Jordens, I., Gevaert, K. et al. (2001). Cdc42 induces filopodia by
promoting the formation of an IRSp53:Mena complex. Curr. Biol. 11, 1645-1655.
Lauffenburger, DA. and Horwitz, AF. (1996). Cell migration: a physically integrated
molecular process. Cell 84, 359-369.
Le Clainche, C., Pantaloni, D. and Carlier, M. (2003). ATP hydrolysis on actin-related
protein 2/3 complex causes debranching of dendritic actin arrays. Proc. Natl. Acad. Sci.
U.S.A. 100, 6337-6342.
Lebensohn, AM., Ma, L., Ho, HH. et al. (2006). Cdc42 and PI(4,5)P2-induced actin
assembly in Xenopus egg extracts. Meth. Enzymol. 406, 156-173.
Lebensohn, AM. and Kirschner, MW. (2009). Activation of the WAVE complex by
coincident signals controls actin assembly. Mol. Cell 36, 512-524.
107
Lebrand, C., Dent, EW., Strasser, GA. et al. (2004). Critical role of Ena/VASP proteins
for filopodia formation in neurons and in function downstream of netrin-1. Neuron 42,
37-49.
Lendvai, B., Stern, EA., Chen, B. et al. (2000). Experience-dependent plasticity of
dendritic spines in the developing rat barrel cortex in vivo. Nature 404, 876-881.
Letourneau, PC. and Shattuck, TA. (1989). Distribution and possible interactions of
actin-associated
proteins and cell adhesion molecules of nerve growth cones.
Development 105, 505-519.
Linder, S., Nelson, D., Weiss, M. et al. (1999). Wiskott-Aldrich syndrome protein
regulates podosomes in primary human macrophages. Proc. Natl. Acad. Sci. U.S.A. 96,
9648-9653.
Linder, S. and Aepfelbacher, M. (2003). Podosomes: adhesion hot-spots of invasive cells.
Trends Cell Biol. 13, 376-385.
Lingwood, D. and Simons, K. (2010). Lipid rafts as a membrane-organizing principle.
Science 327, 46-50.
Liu, AP. and Fletcher, DA. (2006). Actin polymerization serves as a membrane domain
switch in model lipid bilayers. Biophys. J. 91, 4064-4070.
108
Liu, AP., Richmond, DL., Maibaum, L. et al. (2008). Membrane-induced bundling of
actin filaments.. Nature Physics 4, 789-793.
Liu, AP. and Fletcher, DA. (2009). Biology under construction: in vitro reconstitution of
cellular function. Nat. Rev. Mol. Cell Biol. 10, 644-650.
Lommel, S., Benesch, S., Rottner, K. et al. (2001). Actin pedestal formation by
enteropathogenic Escherichia coli and intracellular motility of Shigella flexneri are
abolished in N-WASP-defective cells. EMBO Rep. 2, 850-857.
Ma, L., Cantley, LC., Janmey, PA. et al. (1998). Corequirement of specific
phosphoinositides and small GTP-binding protein Cdc42 in inducing actin assembly in
Xenopus egg extracts. J. Cell Biol. 140, 1125-1136.
Ma, L., Rohatgi, R. and Kirschner, MW. (1998). The Arp2/3 complex mediates actin
polymerization induced by the small GTP-binding protein Cdc42. Proc. Natl. Acad. Sci.
U.S.A. 95, 15362-15367.
Machacek, M., Hodgson, L., Welch, C. et al. (2009). Coordination of Rho GTPase
activities during cell protrusion. Nature 461, 99-103.
Maekawa, M., Ishizaki, T., Boku, S. et al. (1999). Signaling from Rho to the actin
cytoskeleton through protein kinases ROCK and LIM-kinase. Science 285, 895-898.
109
Mallavarapu, A. and Mitchison, T. (1999). Regulated actin cytoskeleton assembly at
filopodium tips controls their extension and retraction. J. Cell Biol. 146, 1097-1106.
Mattila, PK., Pykildinen, A., Saarikangas, J. et al. (2007). Missing-in-metastasis and
IRSp53 deform PI(4,5)P2-rich membranes by an inverse BAR domain-like mechanism. J.
Cell Biol. 176, 953-964.
McGough, A., Pope, B., Chiu, W. et al. (1997). Cofilin changes the twist of F-actin:
implications for actin filament dynamics and cellular function. J. Cell Biol. 138, 771-781.
Mejillano, MR., Kojima, S., Applewhite, DA. et al. (2004). Lamellipodial versus
filopodial mode of the actin nanomachinery: pivotal role of the filament barbed end. Cell
118, 363-373.
Miki, H., Miura, K. and Takenawa, T. (1996). N-WASP, a novel actin-depolymerizing
protein, regulates the cortical cytoskeletal rearrangement in a PIP2-dependent manner
downstream of tyrosine kinases. EMBO J. 15, 5326-5335.
Miki, H., Sasaki, T., Takai, Y. et al. (1998). Induction of filopodium formation by a
WASP-related actin-depolymerizing protein N-WASP. Nature 391, 93-96.
110
Mizutani, K., Miki, H., He, H. et al. (2002). Essential role of neural Wiskott-Aldrich
syndrome protein in podosome formation and degradation of extracellular matrix in srctransformed fibroblasts. Cancer Res. 62, 669-674.
Mogilner, A. and Oster, G. (1996). Cell motility driven by actin polymerization. Biophys.
J. 71, 3030-3045.
Mogilner, A. and Oster, G. (2003). Force generation by actin polymerization II: the
elastic ratchet and tethered filaments. Biophys. J. 84, 1591-1605.
Mooseker MS. and Tilney LG (1975). Organization of an actin filament-membrane
complex. Filament polarity and membrane attachment in the microvilli of intestinal
epithelial cells. J Cell Biol. 67, 725-743.
Nalbant, P., Hodgson, L., Kraynov, V. et al. (2004). Activation of endogenous Cdc42
visualized in living cells. Science 305, 1615-1619.
Narumiya, S., Ishizaki, T. and Watanabe, N. (1997). Rho effectors and reorganization of
actin cytoskeleton. FEBS Lett. 410, 68-72.
Neuhaus JM., Wanger M., Keiser T., and Wegner A. (1983). Treadmilling of actin. J
Muscle Res Cell Motil. 4, 507-527.
111
Nicholson-Dykstra, SM. and Higgs, HN. (2008). Arp2 depletion inhibits sheet-like
protrusions but not linear protrusions of fibroblasts and lymphocytes. Cell Motil.
Cytoskeleton 65, 904-922.
Nobes, CD. and Hall, A. (1995). Rho, rac, and cdc42 GTPases regulate the assembly of
multimolecular focal complexes associated with actin stress fibers, lamellipodia, and
filopodia. Cell 81, 53-62.
Oikawa, T., Yamaguchi, H., Itoh, T. et al. (2004). Ptdlns(3,4,5)P3 binding is necessary
for WAVE2-induced formation of lamellipodia. Nat. Cell Biol. 6, 420-426.
Otomo, T., Tomchick, DR., Otomo, C. et al. (2005). Structural basis of actin filament
nucleation and processive capping by a formin homology 2 domain. Nature 433, 488494.
Paavilainen, VO., Bertling, E., Falck, S. et al. (2004). Regulation of cytoskeletal
dynamics by actin-monomer-binding proteins. Trends Cell Biol. 14, 386-394.
Padrick, SB., Cheng, H., Ismail, AM. et al. (2008). Hierarchical regulation of
WASP/WAVE proteins. Mol. Cell 32, 426-438.
Pankov, R., Endo, Y., Even-Ram, S. et al. (2005). A Rac switch regulates random versus
directionally persistent cell migration. J. Cell Biol. 170, 793-802.
112
Pantaloni, D., Le Clainche, C. and Carlier, MF. (2001). Mechanism of actin-based
motility. Science 292, 1502-1506.
Papayannopoulos, V., Co, C., Prehoda, KE. et al. (2005). A polybasic motif allows NWASP to act as a sensor of PIP(2) density. Mol. Cell 17, 181-191.
Parekh, SH., Chaudhuri, 0., Theriot, JA. et al. (2005). Loading history determines the
velocity of actin-network growth. Nat. Cell Biol. 7, 1219-1223.
Pellegrin, S. and Mellor, H. (2005). The Rho family GTPase Rif induces filopodia
through mDia2. Curr. Biol. 15, 129-133.
Peng, J., Wallar, BJ., Flanders, A. et al. (2003). Disruption of the Diaphanous-related
formin Drfl gene encoding mDial reveals a role for Drf3 as an effector for Cdc42. Curr.
Biol. 13, 534-545.
Pertz, 0., Hodgson, L., Klemke, RL. et al. (2006). Spatiotemporal dynamics of RhoA
activity in migrating cells. Nature 440, 1069-1072.
Peskin, CS., Odell, GM. and Oster, GF. (1993). Cellular motions and thermal
fluctuations: the Brownian ratchet. Biophys. J. 65, 316-324.
Pollard, TD. (1986) Rate constants for the reactions of ATP- and ADP-actin with the
ends of actin filaments. J Cell Biol 103,987-1035.
113
Pollard, TD. and Cooper JA. (1986). Actin and actin-binding proteins. A critical
evalutaion of mechanisms and functions. Annu Rev Biochem 55, 987-1035.
Pollard, TD. and Borisy, GG. (2003). Cellular motility driven by assembly and
disassembly of actin filaments. Cell 112, 453-465.
Pollard, TD. (2007). Regulation of actin filament assembly by Arp2/3 complex and
formins. Annu Rev Biophys Biomol Struct 36, 451-477.
Pollard, TD. and Cooper, JA. (2009). Actin, a central player in cell shape and movement.
Science 326, 1208-1212.
Ponti, A., Machacek, M., Gupton, SL. et al. (2004). Two distinct actin networks drive the
protrusion of migrating cells. Science 305, 1782-1786.
Pring, M., Weber, A. and Bubb, MR. (1992). Profilin-actin complexes directly elongate
actin filaments at the barbed end. Biochemistry 31, 1827-1836.
Pruyne, D., Evangelista, M., Yang, C. et al. (2002). Role of formins in actin assembly:
nucleation and barbed-end association. Science 297, 612-615.
Rafelski, SM. and Theriot, JA. (2004). Crawling toward a unified model of cell mobility:
spatial and temporal regulation of actin dynamics. Annu. Rev. Biochem. 73, 209-239.
114
Raich, WB., Agbunag, C. and Hardin, J. (1999). Rapid epithelial-sheet sealing in the
Caenorhabditis elegans embryo requires cadherin-dependent filopodial priming. Curr.
Biol. 9, 1139-1146.
Read, PW., Liu, X., Longenecker, K. et al. (2000). Human RhoA/RhoGDI complex
expressed in yeast: GTP exchange is sufficient for translocation of RhoA to liposomes.
Protein Sci. 9, 376-386.
Rohatgi, R., Ma, L., Miki, H. et al. (1999). The interaction between N-WASP and the
Arp2/3 complex links Cdc42-dependent signals to actin assembly. Cell 97, 221-231.
Rohatgi, R., Ho, HY. and Kirschner, MW. (2000). Mechanism of N-WASP activation by
CDC42 and phosphatidylinositol 4, 5-bisphosphate. J. Cell Biol. 150, 1299-13 10.
Romero, S., Le Clainche, C., Didry, D. et al. (2004). Formin is a processive motor that
requires profilin to accelerate actin assembly and associated ATP hydrolysis. Cell 119,
419-429.
Safer, D. and Nachmias, VT. (1994). Beta thymosins as actin binding peptides. Bioessays
16, 473-479.
Sagot, I., Rodal, AA., Moseley, J. et al. (2002). An actin nucleation mechanism mediated
by Bnil and profilin. Nat. Cell Biol. 4, 626-63 1.
115
Sarmiento, C., Wang, W., Dovas, A. et al. (2008). WASP family members and formin
proteins coordinate regulation of cell protrusions in carcinoma cells. J. Cell Biol. 180,
1245-1260.
Schirenbeck, A., Bretschneider, T., Arasada, R. et al. (2005). The Diaphanous-related
formin dDia2 is required for the formation and maintenance of filopodia. Nat. Cell Biol.
7, 619-625.
Schirenbeck, A., Arasada, R., Bretschneider, T. et al. (2006). The bundling activity of
vasodilator-stimulated phosphoprotein is required for filopodium formation. Proc. Natl.
Acad. Sci. U.S.A. 103, 7694-7699.
Schwartz, MA. (2001). Integrin signaling revisited. Trends Cell Biol. 11, 466-470.
Schwille, P., Kahya, N. and Bacia, K. (2005). in Protein-Lipid Interactions: From
Membrane Domains to Cellular Networks L. Tamm, Ed. (Wiley-VCH, Weinheim,
2005) . , 337-365.
Small, JV., Stradal, T., Vignal, E. et al. (2002). The lamellipodium: where motility
begins. Trends Cell Biol. 12, 112-120.
116
Snapper, SB., Takeshima, F., Ant6n, I. et al. (2001). N-WASP deficiency reveals distinct
pathways for cell surface projections and microbial actin-based motility. Nat. Cell Biol.
3, 897-904.
Steffen, A., Faix, J., Resch, GP. et al. (2006). Filopodia formation in the absence of
functional WAVE- and Arp2/3-complexes. Mol. Biol. Cell 17, 2581-2591.
Steketee, MB. and Tosney, KW. (2002). Three functionally distinct adhesions in
filopodia: shaft adhesions control lamellar extension. J. Neurosci. 22, 8071-8083.
Stossel, TP., (1993). On the crawling of animal cells. Science. 260, 1086-1094.
Suetsugu, S., Toyooka, K. and Senju, Y. (2009). Subcellular membrane curvature
mediated by the BAR domain superfamily proteins. Semin. Cell Dev. Biol. , .
Svitkina, TM., Bulanova, EA., Chaga, OY. et al. (2003). Mechanism of filopodia
initiation by reorganization of a dendritic network. J. Cell Biol. 160, 409-421.
Theriot, JA. and Mitchison, TJ. (1991). Actin microfilament dynamics in locomoting
cells. Nature 352, 126-13 1.
Theriot, JA., Rosenblatt, J., Portnoy, DA. et al. (1994). Involvement of profilin in the
actin-based motility of L. monocytogenes in cells and in cell-free extracts. Cell 76, 505517.
117
Tilney, LG. and Portnoy, DA. (1989). Actin filaments and the growth, movement, and
spread of the intracellular bacterial parasite, Listeria monocytogenes. J. Cell Biol. 109,
1597-1608.
Tokunaga, M., Kitamura, K., Saito, K. et al. (1997). Single molecule imaging of
fluorophores and enzymatic reactions achieved by objective-type total internal reflection
fluorescence microscopy. Biochem. Biophys. Res. Commun. 235, 47-53.
Upadhyaya, A. and van Oudenaarden, A. (2003). Biomimetic systems for studying actinbased motility. Curr. Biol. 13, R734-44.
van Oudenaarden, A. and Theriot, JA. (1999). Cooperative symmetry-breaking by actin
polymerization in a model for cell motility. Nat. Cell Biol. 1, 493-499.
Vasioukhin, V., Bauer, C., Yin, M. et al. (2000). Directed actin polymerization is the
driving force for epithelial cell-cell adhesion. Cell 100, 209-219.
Vignjevic, D., Yarar, D., Welch, MD. et al. (2003). Formation of filopodia-like bundles
in vitro from a dendritic network. J. Cell Biol. 160, 951-962.
Vignjevic, D., Kojima, S., Aratyn, Y. et al. (2006). Role of fascin in filopodial protrusion.
J. Cell Biol. 174, 863-875.
118
Vignjevic, D., Peloquin, J. and Borisy, GG. (2006). In vitro assembly of filopodia-like
bundles. Meth. Enzymol. 406, 727-739.
Wallar, BJ. and Alberts, AS. (2003). The formins: active scaffolds that remodel the
cytoskeleton. Trends Cell Biol. 13, 435-446.
Watanabe, N. and Mitchison, TJ. (2002). Single-molecule speckle analysis of actin
filament turnover in lamellipodia. Science 295, 1083-1086.
Webb, DJ., Parsons, JT. and Horwitz, AF. (2002). Adhesion assembly, disassembly and
turnover in migrating cells -- over and over and over again. Nat. Cell Biol. 4, E97-100.
Webb, SED., Needham, SR., Roberts, SK. et al. (2006). Multidimensional singlemolecule imaging in live cells using total-internal-reflection fluorescence microscopy.
Opt Lett 31, 2157-2159.
Weiner, OD., Neilsen, PO., Prestwich, GD. et al. (2002). A PtdlnsP(3)- and Rho GTPasemediated positive feedback loop regulates neutrophil polarity. Nat. Cell Biol. 4, 509-513.
Welch, MD., Iwamatsu, A. and Mitchison, TJ. (1997). Actin polymerization is induced
by Arp2/3 protein complex at the surface of Listeria monocytogenes. Nature 385, 265269.
119
Yamaguchi, H., Lorenz, M., Kempiak, S. et al. (2005). Molecular mechanisms of
invadopodium formation: the role of the N-WASP-Arp2/3 complex pathway and cofilin.
J. Cell Biol. 168, 441-452.
Yamana, N., Arakawa, Y., Nishino, T. et al. (2006). The Rho-mDial pathway regulates
cell polarity and focal adhesion turnover in migrating cells through mobilizing Apc and
c-Src. Mol. Cell. Biol. 26, 6844-6858.
Yang, C., Czech, L., Gerboth, S. et al. (2007). Novel roles of formin mDia2 in
lamellipodia and filopodia formation in motile cells. PLoS Biol. 5, e317.
Yilmaz, M. and Christofori, G. (2009). EMT, the cytoskeleton, and cancer cell invasion.
Cancer Metastasis Rev. 28, 15-33.
Zalevsky, J., Grigorova, I. and Mullins, RD. (2001). Activation of the Arp2/3 complex by
the Listeria acta protein. Acta binds two actin monomers and three subunits of the Arp2/3
complex. J. Biol. Chem. 276, 3468-3475.
Zheng, JQ., Wan, JJ. and Poo, MM. (1996). Essential role of filopodia in chemotropic
turning of nerve growth cone induced by a glutamate gradient. J. Neurosci. 16, 11401149.
Zicha, D., Dobbie, IM., Holt, MR. et al. (2003). Rapid actin transport during cell
protrusion. Science 300, 142-145.
120
Zigmond, SH., Evangelista, M., Boone, C. et al. (2003). Formin leaky cap allows
elongation in the presence of tight capping proteins. Curr. Biol. 13, 1820-1823.
121