Current Biology, Vol. 13, R128–R130, February 18, 2003, ©2003 Elsevier Science Ltd. All rights reserved. PII S0960-9822(03)00072-1
Polarised Migration: Cofilin Holds the
Front
Maryse Bailly1 and Gareth E. Jones2
Persistent cell locomotion is a key feature of
eukaryotic cells responding to diverse physiological
cues. New work now directly implicates ADF/cofilin
proteins as essential regulators of polarised cell
migration.
Many eukaryotic cells have the capacity to polarise and
migrate in response to external cues. The ability of a
cell to recognise and adapt to such ‘motogenic’ signals
is critical for directing cell movements, for example in
response to gradients of diffusible chemoattractants, as
is seen in cells of the immune system, or in the cell
migration processes involved in development, wound
healing and tumour metastasis [1]. Recent work on Dictyostelium and neutrophils (reviewed in [2]) has pointed
to a central role for lipid products of PI3 kinase in transducing information from activated surface receptors to
the machinery of cell locomotion. The concept of a
‘chemical compass’ [2] able to point a cell in any suitable direction and direct its navigation to a defined
location is a seductive idea, made all the better for
being open to experimental investigation.
While the role of microtubules in maintaining a cell
axis is well established, most would agree that the
actin cytoskeleton is the major player in generating a
polarised cell shape and subsequently driving forward
the leading lamellipodium of polarising cells (reviewed
in [3]). One recently promoted paradigm brings
together a large body of work on signalling from
surface receptors with data on the biochemistry of
actin filament turnover to provide an attractive model
of actin dynamics at the leading edge of locomoting
cells [4]. A striking feature of this model is the essential
role played by ADF/cofilin in actin filament disassembly. New work from Dawe et al. [5], published recently
in Current Biology, now shines more light on the indispensability of cofilin for polarised cell movement.
ADF/cofilin proteins are ubiquitous small (19 kDa)
polypeptides which regulate actin filament dynamics
and are potentially involved in a wide range of human
diseases (reviewed in [6]). There are three different isoforms in mammals — ADF, non-muscle cofilin and
muscle cofilin — which bind to both monomeric (G)
and filamentous (F) actin, with a higher affinity for ADPbound actin subunits. The binding of ADF/cofilin to filamentous actin is cooperative and stabilises a twisted
form of the actin filament. This conformational change
promotes both fragmentation of the filament, via the
severing activity of ADF/cofilin, and also increases the
1Division
of Cell Biology, Institute of Ophthalmology,
University College London, London, UK. 2Randall Centre,
King’s College London, Guy’s Campus, London SE1 1UL, UK.
E-mail: [email protected]
Dispatch
rate of subunit loss from the pointed end of the actin
filament, via ADF/cofilin’s depolymerisation activity.
Both properties can contribute to either increasing the
turnover of the actin filaments or promoting growth
from new filament ends, depending on the local availability of actin monomers. When bound to actin
monomers, ADF/cofilin inhibits nucleotide exchange,
consequently regulating the recycling of the monomer
pool into ATP-actin (Figure 1).
Because ADF/cofilin proteins can have such tremendous effects on actin filament dynamics, their regulation is understandably tightly coordinated in cells. For
deactivation, the severing/depolymerising activity of
ADF/cofilin is primarily regulated by phosphorylation
on serine 3, which typically abolishes actin binding.
Four kinases have been identified so far that specifically target that residue and block cofilin activity: LIM
kinases 1 and 2, and TES kinases 1 and 2 [6]. All are
rather ubiquitous, and appear to function downstream
of the Rho-family GTPases, although probably not
through the same pathway.
The pathway leading to dephosphorylation and thus
activation of ADF/cofilin is less clear, but at least one
phosphatase, Slingshot, has been identified recently
that specifically dephosphorylates ADF/cofilin both
in vivo and in vitro [7]. Another essential regulator
of ADF/cofilins is pH, their depolymerising activity
increasing upon alkalisation in vivo as well as in vitro
[8]. In addition, ADF/cofilin activity is directly inhibited
by phosphatidylinositol-4,5-bisphosphate (PIP2) [9],
possibly under the control of a phospholipase Cmediated pathway .
Although it is generally agreed that ADF/cofilin
proteins have an essential role in cell motility, the
supporting evidence is still largely circumstantial. In
vivo data have long been scarce because deleting or
expressing functionally impaired mutants of ADF/cofilin
in cells is usually lethal. Recently, the use of temperature-sensitive mutants of the Drosophila homologue of
cofilin Twinstar has demonstrated that ADF/cofilin proteins are required for cell motility during development
[10], although their role in polarised migration remains
confusing. Cofilin phosphorylation following LIM kinase
activation appears essential for T-cell chemotaxis [11],
and levels of phospho-cofilin increase after transient
alkalisation in motile cells [8]. But activation of cofilin is
clearly required for the initial phase of chemoattractantmodulated protrusion [12,13].
The work of Dawe et al. [5] builds upon these
findings to show that the maintenance of a local pool
of active ADF/cofilin proteins at the leading edge of a
migrating cell is required for the cell to maintain an
initial polarity and achieve sustained polarised
protrusion in the direction of migration. This group
previously showed that blocking the ability of actin
filaments to depolymerise inhibits polarised (but not
random) lamellipodium extension, suggesting that
sustained polarised migration requires concomitant
Current Biology
R129
Figure 1. Regulation of ADF/cofilin
activity.
ADF/cofilins bind cooperatively to F-actin,
with a higher affinity for the ADP-actin
subunits. Upon a rise in pH, their activity is
stimulated and they disassemble actin filaments by severing the filaments and
enhancing the rate of subunit dissociation
from the pointed (–) ends of the filaments.
The actin monomers released (G actin)
undergo nucleotide exchange and are
recycled back towards polymerisation at
the barbed (+) end of the filaments [4].
ADF/cofilins are inactivated upon phosphorylation through LIM or TES kinases, or
upon binding to phosphoinositols. Their
activity is restored upon dephosphorylation — by Slingshot or other phosphatases
– or PIP2 hydrolysis, respectively.
PIP2
PIP2
+
–
∆pH
Kinases
(LIM or TES)
PI
Phosphatases
(Slinghot, others)
ADP
ATP
ATP-actin
ADP-Pi-actin
ADP-actin
ADF/cofilin
Current Biology
depolymerisation in the lamellipodium [14]. Their latest
work now shows that phospho-cofilin, the inactive
form, is below detectable levels in the sub-membranous compartment at the edge of polarising lamellipodia. They then altered the total level of inactive
(phosphorylated) ADF/cofilin by overexpressing either
constitutively active LIM kinase or a pseudophosphorylated form of ADF/cofilin. In both cases, the increase
in phospho-ADF/cofilin resulted in loss of cell polarity,
and the smaller lamellipodia still present around the
cell perimeter were enriched in phospho-ADF/cofilin,
as opposed to the non-phosphorylated ADF/cofilin
seen in polarising lamellipodia.
How can cells spatially restrict a pool of active
ADF/cofilin to the leading edge? Physiological signals
that generate cell polarity are known to be accompanied by meaningful localised changes in intracellular
pH. This phenomenon is further amplified at or near the
membrane in the lamellipodium, where frequent ion
transients can create microdomains where pH changes
are much greater than those measured globally [15]. An
initial polarising stimulus, such as the creation of a
wound in a monolayer, can trigger a transient, spatially
restricted, intracellular alkalisation which can modulate
ADF/cofilin behaviour [8]. The consequent increase in
ADF/cofilin depolymerisation/severing activity could
locally increase the level of monomeric actin in cells,
where G actin is limiting, and increase the density of
barbed ends in cells with abundant G actin, further
fuelling lamellipodial extension [4,13], and thereby creating a ‘leading edge’. ADF/cofilin’s depolymerising
activity would decrease as the pH drops back to
normal, and regular actin polymerisation and treadmilling could resume to sustain protrusion activity.
It is thus tempting to speculate that a transient burst
in ADF/cofilin activity at the leading edge is the major
determinant of protrusion initiation and cell polarisation.
This idea is supported by earlier data demonstrating
that transient ADF/cofilin severing activity at the
leading edge is necessary for lamellipodium extension
[12,13]. Such unique ‘cofilin-friendly’ intracellular
microdomains can also be perpetuated in the cells
thanks to the spatial restriction of actin and actinbinding protein isoforms [16]. In particular, tropomyosin, an F-actin-binding protein, has recently been
shown not only to stabilise actin filaments, but also to
directly antagonise the function of ADF/cofilin proteins
(reviewed in [17]). New findings showing that tropomyosin is excluded from the sub-membranous compartment at the leading edge of growing lamellipodia
[18] provide further insights into how it could help to
spatially restrict ADF/cofilin activity to the protruding
rim of a forming lamellipodium.
How is ADF/cofilin maintained in the non-phosphorylated active state at the leading edge? On the basis of
previous data suggesting that the amount of monomeric actin at the leading edge is limited [19], and that
the level of phosphorylation of ADF/cofilin proteins rises
with increasing amount of monomeric actin in the cells
[20], one might envisage some feedback mechanism by
which rapid polymerisation at the leading edge maintains low levels of monomeric actin, which in turn maintains ADF/cofilin in an active/dephosphorylated state
and allows — through severing/depolymerisation — for
a continuous supply of monomers to fuel the protrusion. As this mechanism would be spatially restricted to
the rapidly polymerising filaments at the leading edge,
it could help to sustain the polarity of the protrusion.
Any alteration in the level of phosphorylation of
ADF/cofilin at this front would dramatically shift the
equilibrium and perturb protrusion.
Conversely, new protrusion could then be randomly
initiated elsewhere, as actin monomer stores are
abundant throughout the cell body [14], leading to the
Dispatch
R130
phenotype observed in non-migrating cells [19]. Other
possible mechanisms that might contribute to the
maintenance of a specific pool of active cofilin at the
leading edge might involve either increased localised
ADF/cofilin-specific phosphatase activity, decreased
kinase activity, or stabilisation of the phosphorylated
form of ADF/cofilin outside the lamellipodium through
accessory molecules such as protein 14-3-3ζζ
(reviewed in [6]).
One of the most challenging research endeavours
now open to the field is to identify how the process is
spatially and temporally regulated. First, it is not clear
why actin monomers are somewhat depleted at the
leading edge of migrating cells. This could be the
consequence of rapid incorporation of monomer into
the expanding lamellipodial filament network. Indeed,
while the advance of the leading edge is much slower
than the intracellular rates of actin monomer diffusion,
filament assembly is a fast process. Second, while the
maintenance of a localised pool of active cofilin at
the leading edge of polarising lamellipodia remains
unexplained, one cannot help but wonder what makes
the sub-membranous region at the leading edge so
special, and how it is different in polarising lamellipodia compared with random ones.
While the answer to the first of these two questions
is still unclear, Dawe et al. [5] propose for the second
that the different mechanisms for lamellipodial protrusion are distinguished by the way the actin monomers
are supplied for polymerisation. In migrating cells,
rapid treadmilling of actin filaments provides the
monomers necessary for protrusion, whereas in nonmigrating cells the supply comes from monomer
stores [14]. Thus, only the former mechanism would
strongly depend on ADF/cofilin activity. Clearly we still
have a lot more to learn about the molecular mechanisms regulating actin polymerisation in vivo, and we
are now facing challenging times in unravelling in
detail the biology of directed cell locomotion.
References
1. Franz, C.M., Jones, G.E. and Ridley, A.J. (2002). Cell migration in
development and disease. Dev. Cell 2, 153–158.
2. Weiner, O.D. (2002). Regulation of cell polarity during eukaryotic
chemotaxis: the chemotactic compass. Curr. Opin. Cell Biol. 14,
196–202.
3. Small, J.V., Geiger, B., Kaverina, I. and Bershadsky, A. (2002).
Opinion: How do microtubules guide migrating cells? Nat. Rev. Mol.
Cell Biol. 3, 957–964.
4. Pollard, T.D., Blanchoin, L. and Mullins, R.D. (2001). Actin dynamics. J. Cell Sci. 114, 3–4.
5. Dawe, H.M., Minamide, L.S., Bamburg, J. and Cramer, L.P. (2003).
ADF/cofilin controls cell polarity during fibroblast migration. Curr.
Biol. 13, 4th February issue.
6. Bamburg, J.R. and Wiggan, O.P. (2002). ADF/cofilin and actin
dynamics in disease. Trends Cell Biol. 12, 598–605.
7. Niwar, R., Nagata-Ohashi, K., Takeichi, M., Mizuno, K. and Uemura,
T. (2002). Control of actin reorganization by Slingshot, a family of
phosphatases that dephosphorylate ADF/cofilin. Cell 108, 233–246.
8. Bernstein, B.W., Painter, W.B., Chen, H., Minamide, L.S., Abe, H.
and Bamburg, G., Jr. (2000). Intracellular pH modulation of
ADF/cofilin proteins. Cell Motil. Cytoskeleton 47, 319–336.
9. Ojala, P.J., Paavilainen, V. and Lappalainen, P. (2001). Identification
of yeast cofilin residues specific for actin monomer and PIP2
binding. Biochemistry 40, 15562–15569.
10. Chen, J., Godt, D., Gunsalus, K., Kiss, I., Goldberg, M. and Laski,
F.A. (2001). Cofilin/ADF is required for cell motility during Drosophila
ovary development and oogenesis. Nat. Cell Biol. 3, 204–209.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Nishita, M., Aizawa, H. and Mizuno, K. (2002). Stromal cell-derived
factor 1alpha activates LIM kinase 1 and induces cofilin phosphorylation for T-cell chemotaxis. Mol. Cell. Biol. 22, 774–783.
Zhan, Q., Bamburg, J.R. and Badwey, J.A. (2003). Products of phosphoinositide specific phospholipase C can trigger dephosphorylation of cofilin in chemoattractant stimulated neutrophils. Cell Motil.
Cytoskeleton 54, 1–15.
Zebda, N., Bernard, O., Bailly, M., Welti, S., Lawrence, D.S. and
Condeelis, J.S. (2000). Phosphorylation of ADF/cofilin abolishes
EGF-induced actin nucleation at the leading edge and subsequent
lamellipod extension. J. Cell Biol. 151, 1119–1128.
Cramer, L.P. (1999). Role of actin-filament disassembly in lamellipodium protrusion in motile cells revealed using the drug jasplakinolide. Curr. Biol. 9, 1095–1105.
Schwiening, C.J. and Willoughby, D. (2002). Depolarization-induced
pH microdomains and their relationship to calcium transients in isolated snail neurones. J. Physiol. 538, 371–382.
Gunning, P., Hardeman, E., Jeffrey, P. and Weinberger, R. (1998).
Creating intracellular structural domains: spatial segregation of
actin and tropomyosin isoforms in neurons. Bioessays 20, 892–900.
Cooper, J.A. (2002). Actin dynamics: tropomyosin provides stability. Curr. Biol. 12, R523–525.
Desmarais, V., Ichetovkin, I., Condeelis, J. and Hitchcock-Degregori, S.E. (2002). Spatial regulation of actin dynamics: a
tropomyosin-free, actin-rich compartment at the leading edge. J.
Cell Sci. 115, 4649–4660.
Cramer, L.P., Briggs, L.J. and Dawe, H.R. (2002). Use of fluorescently labelled deoxyribonuclease I to spatially measure G-actin
levels in migrating and non-migrating cells. Cell Motil. Cytoskeleton
51, 27–38.
Minamide, L.S., Painter, W.B., Schevzov, G., Gunning, P. and
Bamburg, J.R. (1997). Differential regulation of actin depolymerizing
factor and cofilin in response to alterations in the actin monomer
pool. J. Biol. Chem. 272, 8303–8309.