Dispatch
R43
14. Garrett, A., and Johnson, K. (2013). Phonetic
bias in sound change. In Origins of Sound
Change: Approaches to Phonologization,
A.C.L. Yu, ed. (Oxford: Oxford University
Press), pp. 51–96.
15. Browman, D.L., and Goldstein, L. (1992).
Articulatory Phonology: An Overview.
Phonetica 49, 155–180.
16. Johanson, L., and Csató, É.Á., eds. The Turkic
languages (Routledge).
17. Pagel, M., Atkinson, Q.D., Calude, A.S., and
Meade, A. (2013). Ultraconserved words point
to deep language ancestry across Eurasia.
PNAS 110, 8471–8476. http://dx.doi.org/
10.1073/pnas.1218726110.
18. Bowern, C. (2013). Relatedness as a Factor in
Language Contact. Journal of Language
Contact 6, 411–432.
19. Atkinson, Q.D., Meade, A., Venditti, C.,
Greenhill, S.J., and Pagel, M. (2008). Languages
Epithelial Cell Division: Aurora Kicks
Lgl to the Cytoplasmic Curb
The Drosophila neoplastic tumor suppressor Lethal giant larvae (Lgl) regulates
apico–basal polarity in epithelia as well as the asymmetric segregation of cell
fate in neural progenitors. Two new studies uncover a new facet of its
regulation in epithelia, where Aurora-dependent phosphorylation triggers Lgl
dissociation from the basolateral cortex to facilitate planar orientation of the
mitotic spindle.
Yu-ichiro Nakajima1
and Matthew C. Gibson1,2,*
Epithelia are the fundamental building
blocks of animal organ and appendage
structures, and thus play prominent
roles in both development and disease.
In broad terms, epithelial architecture
requires the localized assembly of
adhesive junctions in concert with the
apico-basal polarity of each cell.
Considering the complexity associated
with establishing and maintaining this
degree of structural order, how do
proliferating epithelial cells maintain
polarity and tissue integrity while
cyclically disassembling their
interphase morphologies, rounding up,
and dividing into co-equal daughters?
One key hypothesis is that during
division, the mitotic spindle is oriented
to the plane of the epithelium in order to
facilitate the conservation of cell
junctions and the correct integration of
post-mitotic cells into the monolayer.
While both classic papers and more
recent studies have implicated polarity
determinants as cues for planar spindle
orientation [1], precisely how epithelial
polarity is modulated during mitosis
in vivo remains poorly understood.
Addressing this problem head-on, two
reports in this issue of Current Biology
reveal a novel mechanism for the
mitotic regulation of the conserved
polarity regulator Lethal giant larvae
(Lgl) in Drosophila epithelia [2,3].
Lgl was the first reported tumor
suppressor in Drosophila, named for
mutant larvae that exhibit dramatic
overgrowth and a corresponding
disruption of tissue architecture in the
imaginal discs and neuroblasts [4].
Similar phenotypes are observed in
mutants of discs large (dlg) and
scribble (scrib), and subsequent work
has demonstrated that these three
neoplastic tumor suppressor genes
function in the same genetic pathway
[5]. In epithelia, the protein products of
lgl, dlg, and scrib co-localize at the
basolateral membrane and work
together as a protein complex that
controls cell polarity (the Scrib
complex) [5]. Consistent with its
neoplastic phenotypes, Lgl is
implicated in the regulation of
apico-basal polarity in epithelia and
asymmetric cell division in neuroblasts
[6]. In the last decade, further studies
have suggested a contribution of Scrib
complex mutations to tumorigenesis
by investigating Drosophila models
and by exploring the association of
mutations in human orthologs with
cancer [7]. Lgl is a cytoskeletal protein
that primarily localizes at the cell cortex
and plasma membrane, but it is also
found in the cytoplasm [6]. In epithelia,
basolateral Lgl, Dlg and Scrib regulate
cell polarity through mutually
antagonistic interactions with the
apical Par (Par3–Par6–aPKC) and
Crumbs (Crumbs–PatJ–Stardust)
complexes [8]. Similarly, during
asymmetric cell division of
neuroblasts, Lgl targets fate
determinants to the basal cortex by
mutually inhibiting the activity of the
Par complex [9]. Thus, in two very
evolve in punctuational bursts. Science 319,
588–588.
Department of Linguistics, 370 Temple St,
Rm 313, New Haven, CT 06511, USA.
E-mail: claire.bowern@yale.edu
http://dx.doi.org/10.1016/j.cub.2014.11.053
different cellular contexts, the role of
Lgl is to restrict the spatial localization
and activity of polarity determinants
along the cortex and membrane.
The subcellular localization of Lgl is
controlled, in part, by aPKC-dependent
phosphorylation at three conserved
serine residues (S656, S660, and S664)
(Figure 1A). Upon phosphorylation, Lgl
dissociates from the cell cortex,
leading to its cytoplasmic localization
and inactivation [10]. In epithelia and
asymmetrically dividing neuroblasts,
Lgl is excluded from the apical cell
cortex by aPKC-dependent
phosphorylation, which is necessary to
maintain epithelial polarity and direct
fate determinants, respectively [10,11].
Interestingly, during neuroblast cell
division Lgl translocates from the cell
cortex to the cytoplasm at prophase
entry [12]. This event, termed Lgl
cortical release, is triggered by the
mitotic kinase Aurora A (AurA). At the
onset of mitosis, AurA activates aPKC
by directly phosphorylating Par-6, thus
relieving aPKC from negative
regulation by Par-6. The mitotically
activated aPKC then phosphorylates
Lgl and remodels the Par complex [12].
Combined, studies from neuroblast cell
division indicate that protein
localization is dynamically reorganized
to coordinate cell polarity with mitosis.
Until now, however, whether and how
polarity determinants are remodeled
during epithelial cell division has
remained poorly understood.
Using genetic analysis and in vivo
live-imaging, Carvalho et al. [3] and Bell
et al. [2] share the finding that Lgl
relocalizes from the cortex to the
cytoplasm during mitosis in imaginal
and follicular epithelia. How is Lgl
relocalization controlled? Like in
neuroblasts, cortical release of
epithelial Lgl depends on its aPKC
phosphorylation motifs. Further, Lgl
relocalization is strongly delayed in
aurA kinase-defective mutants,
suggesting that AurA activity is
required for Lgl cortical release at
Current Biology Vol 25 No 1
R44
A
B
C
aPKC
S656 S660
P S664
AurA
S656
Lgl
P P
Lgl
P S664
Lgl
or
P P P
Lgl
AurA
AurA
AurA
AurA
AurA
Interphase
Prophase
Metaphase
Current Biology
Figure 1. Lgl relocalization is triggered by AurA-mediated phosphorylation at the onset of
mitosis and its redistribution is necessary for planar spindle orientation in epithelia.
(A) During interphase, Lgl is restricted from the apical cortex by aPKC-mediated phosphorylation. (B) Upon prophase entry, AurA directly phosphorylates specific serine residues
(S656, S664), releasing Lgl from the basolateral cortex into the cytoplasm. (C) Lgl relocalization
is hypothesized to allow mitotic spindles to align at the basolateral region, where Dlg and the
Pins complex may interact. Grey, Adherens junctions; green, basolateral cortex.
mitotic entry [3]. While prior neuroblast
work has implicated AurA in the
indirect control of Lgl localization
through aPKC [12], the present studies
suggest a direct phosphorylation.
Indeed, in epithelia the kinase
activity of aPKC is dispensable for
Lgl release, and AurA directly
phosphorylates the putative aPKC
phosphorylation motif in Lgl
(Figure 1B). Using Drosophila S2 cells,
Carvalho et al. show that double
phosphorylation of Lgl is necessary for
relocalization and that AurA
phosphorylates S656 and S664 during
mitosis [3]. Bell et al. reach a similar
conclusion by combining in vitro kinase
assays with in vivo localization studies
in both epithelia and neuroblasts [2].
What is the function of Lgl cortical
release, if any? Both Dlg and Scrib
remain associated with the cortex
during mitosis and have been
implicated in controlling planar
alignment of the mitotic spindle [13,14].
In imaginal discs and follicular
epithelia, this activity appears to be
distinct from the role of the Scrib
complex in polarity. Similarly, both Bell
et al. [2] and Carvalho et al. [3] show
that AurA-mediated Lgl
phosphorylation is not required for
overall epithelial polarity. Instead, Lgl
cortical release appears to be
important only during mitosis.
Following the expression of
AurA-insensitive forms of Lgl that
cannot relocalize to the cytoplasm,
mitotic spindle orientation becomes
randomized in the follicular epithelium
[3]. Abnormal spindle orientation
phenotypes are also observed in wing
epithelial cells expressing
AurA-insensitive or
membrane-tethered forms of Lgl [2].
Altogether, these results demonstrate
that Lgl relocalization is required for
planar spindle orientation in the
proliferating Drosophila epithelia.
Mechanistically, how does removal
of Lgl from the cortex contribute to
planar spindle orientation? One
hypothesis is that Lgl release facilitates
a Dlg-Pins (Partner of Inscuteable)
interaction at the lateral cortex during
mitosis. Pins (known as LGN in
vertebrates) is a conserved plasma
membrane-associated protein that
regulates mitotic spindle orientation
[1]. In the Drosophila follicular
epithelium and chick neuroepithelium,
genetic and cell biological evidence
suggests that Dlg directs Pins/LGN
basolateral localization through its
guanylate kinase (GUK) domain [13,15].
Recent structural and biochemical
studies using mammalian homologues
show that the Dlg GUK domain directly
interacts with both phosphorylated Lgl
and phosphorylated Pins/LGN [16,17].
Given that Pins/LGN is mitotically
phosphorylated by AurA [18] and the
Lgl–Dlg interaction is weaker than the
Pins/LGN–Dlg interaction [16,17], Lgl
relocalization may promote planar
spindle orientation by transiently
freeing Dlg to interact with Pins/LGN or
other spindle-associated factors
(Figure 1C). Looking forward, future
studies should seek to define the
biochemical links between the spindle
and the Scrib complex and explore the
possibility that similar events occur in
other organismal systems. Intriguingly,
AurA is highly expressed in human
cancers [19] and regulates mitotic
spindle orientation in mammary
epithelial cells [20]. A deeper
understanding of the modulation of cell
polarity during epithelial cell division
could thus provide broad new insights
into both epithelial homeostasis and
disease.
References
1. Morin, X., and Bellaı̈che, Y. (2011). Mitotic
spindle orientation in asymmetric and
symmetric cell divisions during animal
development. Dev. Cell 21, 102–119.
2. Bell, G.P., Fletcher, G.C., Brain, R., and
Thompson, B.J. (2015). Aurora kinases
phosphorylate Lgl to induce mitotic spindle
orientation in Drosophila epithelia. Curr. Biol.
25, 61–68.
3. Carvalho, C.A., Moreira, S., Ventura, G.,
Sunkel, C.E., and Morais-de-Sá, E. (2015).
Aurora A triggers Lgl cortical release during
symmetric division to control planar spindle
orientation. Curr. Biol. 25, 53–60.
4. Gateff, E. (1978). Malignant neoplasms of
genetic origin in Drosophila melanogaster.
Science 200, 1448–1459.
5. Bilder, D. (2004). Epithelial polarity and
proliferation control: links from the Drosophila
neoplastic tumor suppressors. Genes Dev. 18,
1909–1925.
6. Wirtz-Peitz, F., and Knoblich, J.A. (2006). Lethal
giant larvae take on a life of their own. Trends
Cell Biol. 16, 234–241.
7. Humbert, P.O., Grzeschik, N.A., Brumby, A.M.,
Galea, R., Elsum, I., and Richardson, H.E.
(2008). Control of tumourigenesis by the
Scribble/Dlg/Lgl polarity module. Oncogene 27,
6888–6907.
8. St Johnston, D., and Ahringer, J. (2010). Cell
polarity in eggs and epithelia: parallels and
diversity. Cell 141, 757–774.
9. Knoblich, J.A. (2010). Asymmetric cell division:
recent developments and their implications for
tumour biology. Nat. Rev. Mol. Cell Biol. 11,
849–860.
10. Betschinger, J., Mechtler, K., and Knoblich, J.A.
(2003). The Par complex directs asymmetric
cell division by phosphorylating the
cytoskeletal protein Lgl. Nature 422, 326–330.
11. Hutterer, A., Betschinger, J., Petronczki, M.,
and Knoblich, J.A. (2004). Sequential roles of
Cdc42, Par-6, aPKC, and Lgl in the
establishment of epithelial polarity during
Drosophila embryogenesis. Dev. Cell 6,
845–854.
12. Wirtz-Peitz, F., Nishimura, T., and
Knoblich, J.A. (2008). Linking cell cycle to
asymmetric division: Aurora-A phosphorylates
the Par complex to regulate Numb localization.
Cell 135, 161–173.
13. Bergstralh, D.T., Lovegrove, H.E., and St
Johnston, D. (2013). Discs large links spindle
orientation to apical-basal polarity in
Drosophila epithelia. Curr. Biol. 23, 1707–1712.
14. Nakajima, Y., Meyer, E.J., Kroesen, A.,
McKinney, S.A., and Gibson, M.C. (2013).
Epithelial junctions maintain tissue architecture
by directing planar spindle orientation. Nature
500, 359–362.
Dispatch
R45
15. Saadaoui, M., Machicoane, M., di Pietro, F.,
Etoc, F., Echard, A., and Morin, X. (2014). Dlg1
controls planar spindle orientation in the
neuroepithelium through direct interaction with
LGN. J. Cell Biol. 206, 707–717.
16. Zhu, J., Shang, Y., Xia, C., Wang, W., Wen, W.,
and Zhang, M. (2011). Guanylate kinase
domains of the MAGUK family scaffold proteins
as specific phospho-protein-binding modules.
Embo J. 30, 4986–4997.
17. Zhu, J., Shang, Y., Wan, Q., Xia, Y., Chen, J.,
Du, Q., and Zhang, M. (2014). Phosphorylationdependent interaction between tumor
suppressors Dlg and Lgl. Cell Res. 24,
451–463.
18. Johnston, C.A., Hirono, K., Prehoda, K.E., and
Doe, C.Q. (2009). Identification of an Aurora-A/
PinsLINKER/Dlg spindle orientation pathway
using induced cell polarity in S2 cells. Cell 138,
1150–1163.
19. Fu, J., Bian, M., Jiang, Q., and Zhang, C. (2007).
Roles of Aurora kinases in mitosis and
tumorigenesis. Mol. Cancer Res. 5,
1–10.
20. Regan, J.L., Sourisseau, T., Soady, K.,
Kendrick, H., McCarthy, A., Tang, C.,
Brennan, K., Linardopoulos, S., White, D.E., and
Smalley, M.J. (2013). Aurora A kinase regulates
mammary epithelial cell fate by determining
mitotic spindle orientation in a
DNA Damage Responses: Beyond
Double-Strand Break Repair
The RAG endonuclease generates DNA double strand breaks during antigen
receptor gene assembly, an essential process for B- and T-lymphocyte
development. However, a recent study reveals that RAG endonuclease activity
affects natural killer cell function, demonstrating that such double strand
breaks, and the responses they elicit, may have broad cellular effects.
Andrea L. Bredemeyer
and Barry P. Sleckman*
DNA double-strand breaks (DSBs) are
generated by genotoxic agents and as
intermediates in several physiological
processes. These DSBs activate the
ATM kinase, which orchestrates a
canonical DNA damage response
(DDR) that includes activation of
cell-cycle checkpoints, initiation of
DSB repair and the activation of cell
death pathways when DSBs persist
unrepaired [1]. However, recent
studies [2–5], including one published
in Cell by Karo et al. [2], reveal that
signals from DNA DSBs may have
broader effects on cellular functions
that persist long after the DSB has been
repaired.
During development, B and T
lymphocytes must assemble antigen
receptor genes through the process
of V(D)J recombination [6]. This
reaction is initiated when the RAG1
and RAG2 proteins, which together
form the RAG endonuclease, introduce
DNA DSBs at the border of two
recombining gene segments (V, D or J),
and their flanking RAG recognition
sequences [6]. RAG is expressed only
in developing lymphocytes and their
immediate precursor, the common
lymphoid progenitor [7]. RAG DSBs
are processed and repaired by the
non-homologous end-joining (NHEJ)
pathway of DNA DSB repair [8].
Assembly of antigen receptor
genes is an absolute requirement for
B- and T-lymphocyte development,
and mice and humans deficient in
RAG1, RAG2 or components of the
NHEJ pathway are severely
lymphopenic. Common lymphoid
progenitors also give rise to natural
killer (NK) cells that serve critical
functions in early immune responses.
Unlike B and T lymphocytes, NK
cell development does not depend
on antigen receptor gene assembly,
and mice and humans deficient
in RAG1, RAG2 or NHEJ proteins
have normal numbers of mature NK
cells.
Using mice that allow for fate
mapping of cells that have expressed
RAG1 [9], Karo et al. [2] show that a
significant fraction of mature NK
cells have expressed RAG1 during
development. As expected, RAG1 is
not expressed in mature NK cells.
Strikingly, there were clear
phenotypic differences between
mature NK cells that had expressed
RAG1 during development, and
those that had not. Mature NK cells
that had never expressed RAG1
appeared to be more activated and
terminally differentiated, in addition to
exhibiting higher levels of cytotoxicity.
Analysis of NK cells from RAG1- and
RAG2-deficient mice revealed
functional phenotypes similar to
wild-type NK cells with no history of
Notch-dependent manner. Cell Rep. 4,
110–123.
1Stowers
Institute for Medical Research,
1000 East 50th Street, Kansas City, MO
64110, USA. 2Department of Anatomy and
Cell Biology, University of Kansas Medical
Center 3901 Rainbow Boulevard, Kansas
City, Kansas 66160, USA.
*E-mail: MG2@stowers.org
http://dx.doi.org/10.1016/j.cub.2014.11.052
RAG expression. Moreover, compared
with NK cells from wild-type mice,
those from RAG-deficient mice
failed to expand and persist
following mouse cytomegalovirus
infection, due to an increased
susceptibility to apoptosis. Thus,
expression of RAG in developing NK
cells had a remarkable effect on the
activity of mature NK cells, even though
these cells do not require antigen
receptor gene assembly — the only
known RAG activity — for their
development.
How is it that the RAG expression in
developing NK cells can affect the
function of mature NK cells?
Presumably this occurs through a
RAG-dependent alteration in the
genetic program in developing NK cells
that persists in mature NK cells.
Indeed, Karo et al. [2] find that the
expression of several genes encoding
proteins involved in the DDR, including
DNA-PKcs (Prkdc), Ku80 (Xrcc5), Chk2
(Chek2) and Atm, in mature NK cells
depends on prior RAG expression.
Moreover, compared with mature NK
cells from wild-type mice, those from
RAG-deficient mice exhibit
perturbations in the DDR, indicating
that these gene expression changes
have functional consequences. In
agreement with this notion, the authors
find that mature NK cells from
DNA-PKcs-deficient mice have
similar phenotypes to those
from RAG-deficient mice. Thus,
RAG expression in developing NK
cells is required, at least in part, to
promote a normal DDR in mature NK
cells.
RAG1 and RAG2 have no known
independent functions and together
their only known activity is as an
endonuclease. However, RAG2 has a
plant homeodomain that binds
broadly throughout the genome to