Cdc42-dependent actin polymerization during compensatory...

L E T T E R S

NATURE CELL BIOLOGY VOLUME 5 | NUMBER 8 | AUGUST 2003 727

Cdc42-dependent actin polymerization duringcompensatory endocytosis in Xenopus eggsAnna Marie Sokac1,5, Carl Co2, Jack Taunton3 and William Bement1,4

The actin filament (F-actin) cytoskeleton associatesdynamically with the plasma membrane and is thus ideallypositioned to participate in endocytosis. Indeed, a wealth ofgenetic and biochemical evidence has confirmed that actininteracts with components of the endocytic machinery1,although its precise function in endocytosis remains unclear.Here, we use 4D microscopy to visualize the contribution ofactin during compensatory endocytosis in Xenopus laevis eggs.We show that the actin cytoskeleton maintains exocytosingcortical granules as discrete invaginated compartments, suchthat when actin is disrupted, they collapse into the plasmamembrane. Invaginated, exocytosing cortical granulecompartments are directly retrieved from the plasmamembrane by F-actin coats that assemble on their surface.These dynamic F-actin coats seem to drive closure of theexocytic fusion pores and ultimately compress the corticalgranule compartments. Active Cdc42 and N-WASP arerecruited to exocytosing cortical granule membranes before F-actin coat assembly and coats assemble by Cdc42-dependent, de novo actin polymerization. Thus, F-actin maypower fusion pore resealing and function in two novelendocytic capacities: the maintenance of invaginatedcompartments and the processing of endosomes.

In yeast, actin is firmly implicated as a mediator of endocytosis, as

genes necessary for a viable actin cytoskeleton are required for endocy-

tosis and vice versa2. In vertebrate cells, proteins associated with the

endocytic machinery interact with F-actin or with activators of

Arp2/3-dependent actin polymerization1. Conversely, the membranes

of endosomes and pinosomes are endowed with activators of actin

polymerization and can stimulate actin polymerization3–9. These

molecular links have led to numerous proposed roles for actin in

endocytosis10: for example, F-actin may function as a scaffold to con-

centrate endocytic machinery at sites of uptake. Alternatively, actin

polymerization and/or myosin-dependent processes may generate

force to drive membrane invagination, membrane fission and endo-

some trafficking. Indeed, temporal and spatial high-resolution analy-

ses of actin during endocytosis are consistent with these proposed

roles. In Swiss 3T3 cells11, actin gathers at clathrin-coated pits after

dynamin and coincident with the departure of pits/nascent coated

vesicles from the plasma membrane. If this actin is in the form of

dynamic comet tails, it may propel coated vesicles away from the

plasma membrane and into the cell interior, as previously described

for trafficking of macropinosomes in rat basophilic leukaemia cells12.

In Xenopus oocytes and eggs, the en masse membrane insertion

incurred during cortical granule exocytosis is countered by compensa-

tory endocytosis that generates large endosomes of the same size as

cortical granules13, possibly by resealing the exocytic fusion pores.

Such a mechanism, whereby emptied secretory granule membranes

are directly retrieved from the plasma membrane by reversal of fusion

pore formation, was previously described for sea urchin eggs14, as well

as mammalian neurons and endocrine cells10, and is implicit in the

‘kiss-and-run’ model of compensatory endocytosis15. To determine

whether large endosomes are directly derived from exocytosed cortical

granules in Xenopus eggs, we followed formation of large endosomes

during cortical granule exocytosis. In these cells, a wave of cortical

granule exocytosis will emanate from a site of local calcium increase16.

Thus, eggs were pricked in calcium-containing buffer supplemented

with fluorescein–dextran (F–dex) and then fixed and stained for corti-

cal granules with rhodamine–lectin. Conventional confocal

microscopy showed that cortical granule exocytosis near the prick site

correlates with the appearance of F–dex-filled large endosomes

(Fig. 1a). Large endosomes were approximately the same size as corti-

cal granules and, as revealed by imaging regions far from the prick site,

large endosome formation required earlier cortical granule exocytosis.

Strikingly, some large endosomes contained both F–dex and rho-

damine–lectin, demonstrating that large endosomes are derived from

cortical granules and so represent the exocytosis–endocytosis interme-

diate generated when emptied cortical granule membranes are directly

retrieved from the plasma membrane.

To investigate the role of actin in large endosome formation, eggs

pricked in extracellular Texas Red–dextran (TR–dex) were fixed and

stained for F-actin with Alexa488–phalloidin. In single optical plane,

en face views, F-actin rings encircled large endosomes (Fig. 1b). Z-

sections revealed that the F-actin rings are spherical coats that encase

the large endosomes (Fig. 1c).

Cortical actin dynamics during large endosome formation were

then examined in living eggs by time-lapse confocal microscopy in a

single optical plane. Eggs were injected with Alexa488-labelled G-actin

(Alexa488–G-actin) and caged inositol-1,4,5-trisphosphate (caged

1Department of Zoology, University of Wisconsin, Madison, WI 53706, USA. 2Program in Biological Sciences, University of California, San Francisco, CA 94107,USA. 3Department of Cellular and Molecular Pharmacology, University of California, San Francisco, CA 94107, USA. 4Program in Molecular and Cellular Biology,University of Wisconsin, Madison, WI 53706, USA. 5Current address: Department of Molecular Biology, Princeton University, Princeton, NJ 08544, USA.6Correspondence should be addressed to A.M.S. ([email protected]).

© 2003 Nature Publishing Group

L E T T E R S

728 NATURE CELL BIOLOGY VOLUME 5 | NUMBER 8 | AUGUST 2003

InsP3), which when photolysed, increases local calcium levels17 and so

induces cortical granule exocytosis. Eggs were mounted in TR–dex and

imaged. Almost immediately after caged-InsP3 photolysis, exocytosing

cortical granules filled with TR–dex (Fig. 1d; also see Supplementary

Information, Movie 1). Within 16.3 ± 0.2 s of filling, actin encircled

the exocytosing cortical granules (mean ± s.e.m., n = 60 exocytosing

cortical granules from three experiments). The actin rings grew more

robust and then constricted until the TR–dex-filled compartments

disappeared and only plugs of actin remained.

Although single optical plane, time-lapse analysis demonstrated that

F-actin coats are dynamic and transient, it did not allow us to ade-

quately examine their function. Hence, we collected image stacks of

cortical actin during cortical granule exocytosis/large endosome forma-

tion and then prepared these data as time-lapse volumetric renderings

(4D). In 4D renderings, looking on to the external surface of the plasma

membrane (apical; Fig. 2a, b; also see Supplementary Information,

Movies 2 and 3), exocytosing cortical granules incorporated TR–dex,

around which actin accumulated in circumferential ridges. As the

ridges heightened and constricted, sheets of actin were also observed

sweeping over the TR–dex-filled compartments. At higher magnifica-

tion (Fig. 2b, note 40-s timepoint; also see Supplementary Information,

Movie 3), it was apparent that these actin structures closed over the

TR–dex-filled compartments to form large endosomes.

For an alternative perspective, 4D volumes were rotated 180° from

apical, (basal; Fig. 2c; also see Supplementary Information, Movie 4),

thus providing a view from the cell interior towards the inner surface

of the plasma membrane. In this view, exocytosing cortical granules

were enveloped by actin sheets that extended progressively along their

surfaces, meeting basally to form complete coats. Assembled actin

coats compressed apically to squeeze the compartments until TR–dex

was no longer detectable and the spent coat remained as only an actin

ball at the plasma membrane. Although some TR–dex was most prob-

ably expelled extracellularly during coat compression, a portion was

also endocytosed and trafficked into a sub-cortical membrane com-

partment detectable in fixed eggs (data not shown).

Actin coat dynamics appeared rather similar from both the apical

and basal views, even though the coats encounter different tasks

depending on the perspective: In the apical view, fusion of the cortical

granule with the plasma membrane is followed by closure of actin over

the fusion pore. In the basal view, the entire granule surface becomes

encased in actin and is squeezed. The coat activities observed from

these distinct views are summarized in 4D cross-sections (Fig. 2d; also

see Supplementary Information, Movie 5), where actin coat constric-

tion within the plane of the plasma membrane was obvious, as well as

the apically directed compression throughout the entire coat.

Interestingly, coat assembly always started at the apical surface of the

exocytosing cortical granule (for example see Fig. 2c; also see

Supplementary Information, Movie 4), where components of the cor-

tical granule membrane and plasma membrane could mix most rap-

idly after fusion, and actin coats could simultaneously initiate their

apical constriction and basal encasement.

To further assess the role of F-actin in the recovery of exocytosing

cortical granule compartments, eggs injected with Alexa488–G-actin

and caged-InsP3 were treated with 5 µM latrunculin B and imaged as

above. Single optical plane, time-lapse analysis demonstrated that

latrunculin blocked coat assembly, but not cortical granule exocytosis

(Fig. 3a). When F-actin was disrupted, adjacent exocytosing cortical

granules seemed to fuse together, an event rarely observed in control

eggs (118 ± 11.4 fusions per 85 µm2 field for latrunculin versus 0.6 ± 0.3

for controls; mean ±s.e.m., n ≥ 3 eggs from three experiments;

P < 0.005). In 4D cross-section (Fig. 3b), some exocytosing cortical

granules rapidly collapsed into the plasma membrane, rather than

remaining as invaginated compartments. Whether fusion between adja-

cent exocytosing cortical granules represents a distinct phenomenon or

is simply a different manifestation of collapse into the plasma mem-

brane is currently unresolved.

Apical view, 4D renderings then demonstrated that latrunculin

treatment and the resulting failure to assemble actin coats resulted in

massive cortical disintegration after cortical granule exocytosis

(Fig. 3c), presumably because of the uncompensated insertion of

excess membrane into the plasma membrane. Thus, F-actin maintains

exocytosing cortical granules as discrete invaginated compartments

that are subsequently retrieved by dynamic actin coats. Although the

actin coats could themselves maintain the invaginated compartments,

the cortical F-actin network most probably also imposes physical

restraints that favour maintenance rather than collapse.

F-actin coats may assemble on the exocytosing cortical granule

compartments by recruiting pre-existing F-actin or by de novo poly-

merization. To discriminate between these possibilities18,19, eggs

were co-injected with Alexa488–G-actin and Alexa568–phalloidin.

As the association rate of phalloidin for F-actin is relatively slow20,

Alexa488–G-actin should selectively label newly polymerized

F-actin before phalloidin. Single optical plane, time-lapse imaging

revealed that Alexa488–G-actin incorporates into coats 2.3 ± 0.4 s

Figure 1 Large endosomes are surrounded by dynamic coats of F-actin.(a) Single-plane confocal images showing that large endosomes (F–dex,green) form as cortical granules (rhodamine–lectin, red) are exocytosed.Some large endosomes contain both F–dex and rhodamine–lectin(arrowheads). (b) Single-plane confocal images showing that F-actinrings (Alexa488–phalloidin, green) encircle large endosomes (TR–dex,

red). (c) A z-section showing F-actin coats (green) surrounding largeendosomes (red). (d) Single-plane, time-lapse confocal images showingexocytosing cortical granules (TR–dex, red) being rapidly encircled byactin rings that constrict (Alexa488–G-actin, green, arrowheads).Photolysis was performed at 0 s. Scale bars represent 5 µm in a, b andd, and 2.5 µm in c.


L E T T E R S


before Alexa568–phalloidin (mean ± s.e.m., n = 65 coats from three

experiments; Fig. 4a, b; P < 0.005). Thus, actin polymerization

drives coat assembly.

In Xenopus eggs6 and egg extracts6,21–23, actin polymerizes on vesi-

cles through the Cdc42–N-WASP–Arp2/3 pathway. We therefore

probed for active Cdc42 during compensatory endocytosis with the

GTPase-binding domain of WASP (wGBD), which binds only the

active (GTP-bound) form of Xenopus Cdc42 (ref. 24). Single optical

plane, time-lapse imaging revealed that enhanced green fluorescent

protein (eGFP)-tagged wGBD is recruited to exocytosing cortical

granule membranes 10.0 ± 0.4 s after TR–dex filling (Fig. 4c; mean ±s.e.m., n = 60 exocytosing cortical granules from three experiments). A

marker for endogenous N-WASP, Alexa488–N-WASP-truncate, was

similarly recruited to exocytosing cortical granules 11.1 ± 0.2 s after

TR–dex filling (Fig. 4d; mean ± s.e.m., n = 60 exocytosing cortical

granules from three experiments).

The recruitment of active Cdc42 and N-WASP to exocytosing corti-

cal granules precedes coat assembly (see above), as predicted if the

Cdc42–N-WASP–Arp2/3 pathway effects coat assembly25. To function-

ally implicate this pathway, we injected eggs with dominant-negative

Figure 2 F-actin coats close over and compress exocytosing corticalgranule compartments. (a) A low-magnification, apical view (as if lookingonto the external surface of the plasma membrane from outside of thecell), 4D image showing actin coats (Alexa488–G-actin, green) manifestedas circumferential ridges that form around and close over exocytosingcortical granules (TR–dex, red; arrowheads). (b) High-magnification,apical view, 4D images showing that the actin ridge rises then constrictsto close over the cortical granule compartment (arrowhead). (c) High-magnification, basal view (rotated 180° from apical; as if looking onto the

internal surface of the plasma membrane from inside of the cell), 4Dimages of same exocytosing cortical granule, showing that the actin coatencases the compartment and compresses apically. (d) High-magnification, 4D images of a coat in cross-section, showing the coatdynamics seen in both the apical and basal views and indicatingconstriction in the plane of the plasma membrane, as well as apicallydirected compression throughout. Photolysis was performed at 0 s. Scalebar represents 5 µm in a, and 2.5 µm in b and c. Scale in d is identical tothe scale in b and c.


L E T T E R S


Cdc42 (Cdc42T17N)26. As viewed by single optical plane, time-lapse

microscopy, Cdc42T17N blocked coat assembly, but not cortical granule

exocytosis (Fig. 4e). The Cdc42T17N phenotype mimicked that induced

by latrunculin, resulting in fusions between adjacent exocytosing cortical

granules (45 ± 8.8 fusions per 85 µm2 field for Cdc42T17N versus 0.6 ± 0.3

for controls; mean ± s.e.m., n ≥ 3 eggs from three experiments;

P < 0.005). Thus, a Cdc42-dependent pathway drives coat assembly.

For propagation of the Cdc42–N-WASP–Arp2/3 pathway, N-WASP

requires input from both active Cdc42 and phosphatidylinositol-4,5-

bisphosphate (PtdInsP2)25. In the egg, PtdInsP2 is restricted to the

plasma membrane (A.M.S., unpublished observations), but may dif-

fuse laterally into cortical granule membranes immediately after exo-

cytosis to trigger actin coat assembly. Such ‘compartment mixing’

could represent a general mechanism to target local actin polymeriza-

tion on membranes. In this case, targeted actin polymerization on the

cortical granule membrane is associated with normal compensatory

endocytosis, as the actin coat seems to close over the fusion pore and

clearly compresses the exocytosing cortical granule compartment, per-

haps as a first step in endosomal processing. Additional insight into the

role of actin in compensatory endocytosis will certainly be gained by

simultaneously visualizing actin, the plasma membrane and cortical

granule/large endosome membranes throughout the process.

Unfortunately, we have thus far found actin coat assembly to be

severely inhibited by conventional lipid markers (data not shown).

For actin to achieve its compensatory endocytic roles, we suggest

that actin coat polymerization must be coupled to the membranes of

fusion pores and exocytosing cortical granule compartments so as to

exert force on the membranes, as previously described in cell migra-

tion, phagocytosis, and pathogen and vesicle motility1. The exerted

force is especially remarkable during coat compression, when the vol-

ume of TR–dex-filled cortical granule compartments is reduced by

0.5–1.2 µm3 s (n = 7 compartments from three eggs), and may provide

the first in vivo confirmation of the recent in vitro observations that

actin polymerization directly squeezes plastic beads27 and synthetic

lipid vesicles28,29. However, we cannot discount the possibility that

coat polymerization may also indirectly contribute to force production

by providing substrate for myosin activity30,31.

METHODSEgg procurement and injection. Oocytes were harvested as previously

described32 and cultured in OR2 (82.5 mM sodium chloride, 2.5 mM potas-

sium chloride, 1 mM calcium chloride, 1 mM magnesium chloride, 1 mM

Na2HPO4 and 5 mM Hepes at pH 7.4). If required, injections of 40 nl per

oocyte were delivered using a PLI-100 picoinjector (Medical Systems,

Greenvale, NY). To obtain eggs, meiotic maturation was induced by incubating

oocytes in OR2 with 5 µg ml−1 progesterone for 10–14 h.

Egg fixation and staining. Eggs were pricked with a fine glass needle in OR2

supplemented with 1 mM 3,000 Mr, lysine-fixable TR–dex or F–dex (Molecular

Probes, Eugene, OR). After 2 min, cells were washed vigorously and then fixed

for 4 h in 3.7% paraformaldehyde, 0.1% glutaraldehyde, 10 mM EGTA, 100mM

potassium chloride, 3 mM magnesium chlrode, 10 mM Hepes at pH 7.6 and

150 mM sucrose. Fix also included 0.4 units ml−1 Alexa488–phalloidin

(Molecular Probes) or 0.5% Triton X-100 for F-actin or cortical granule stain-

ing, respectively. Eggs were bisected before staining. F-actin was labelled with

1U ml−1 Alexa488–phalloidin. The heavily glycosylated cortical granule con-

tents were labelled with 10 µg ml−1 rhodamine–Dolichos biflorus Agglutinin

(Vector Labs, Burlingame, CA).

Probe preparation. Human WASP GBD (residues 219–311; provided by H.

Higgs, Dartmouth Medical School) was cloned carboxy-terminal to eGFP in the

custom vector pCS2+-eGFP, and capped eGFP-wGBD mRNA was synthesized

from this construct in vitro (mMessage mMachine transcription kit; Ambion,

Austin, TX).

Truncated rat N-WASP (residues 151–501) was cloned carboxy-terminal to a

His6-tag followed immediately by a Tobacco Etch Virus (TEV) protease cleavage

site in the expression vector pBH4. Protein expressed in Escherichia coli strain

BL21 was batch-purified using Nickel–NTA agarose (Qiagen, Valencia, CA) and

eluted with 250 mM imidazole. The His6-tag was removed by TEV protease

cleavage and the N-WASP protein then conjugated to Alexa488. This probe

mimics full-length N-WASP in that it rescues comet formation in Xenopus egg

extracts immunodepleted for endogenous N-WASP (data not shown).

Cell preparation for live imaging. Oocytes were injected with NPE caged-InsP3

(Molecular Probes) to attain a final intracellular concentration of 10 µM.

Protein probes and inhibitor were simultaneously injected with caged InsP3 to

attain final intracellular concentrations as follows: 96 µg ml−1 Alexa488–G-

actin (Molecular Probes), 6.7U ml−1 Alexa568–phalloidin (Molecular Probes),

2 µM Alexa488–N-WASP truncate and 3.2 µM human Cdc42T17N–glutathione

Figure 3 Retrieval of cortical granule membranes is F-actin-dependent.(a) Single-plane, time-lapse confocal images showing that latrunculininhibits actin (Alexa488–G-actin, green) coat assembly and inducesfusion between adjacent exocytosing cortical granules (TR–dex, red;arrowheads). (b) High-magnification, 4D images of an exocytosingcortical granule in cross-section, showing that latrunculin treatment

induces collapse of exocytosing cortical granules into the plasmamembrane. (c) Apical view, 4D images, showing complete disintegrationof the cortex in latrunculin-treated eggs after cortical granule exocytosis.Exocytosing cortical granules fuse with adjacent compartments(arrowheads). Photolysis was performed at 0 s. Scale bar represents10 µm in a, and 5 µm in b and c.


L E T T E R S


S-transferase (GST; Cytoskeleton, Denver, CO). Capped eGFP-wGBD mRNA

was injected at 6 ng per cell after caged InsP3 injection and expression pro-

ceeded concomitant with meiotic maturation. For latrunculin analysis, eggs

were pre-treated with 5 µM latrunculin B for 30 min then imaged in the pres-

ence of 5 µM latrunculin B.

Microscopy and image analysis. Imaging was performed with a Zeiss Axiovert

100 M microscope (Carl Zeiss, Thornwood, NY) with Bio-Rad 1024 × 1024

Lasersharp Confocal software (Bio-Rad, Hercules, CA) using a numerical aper-

ture 1.4, 63× objective lens. Eggs were mounted in OR2 with 100 µg ml−1 wheat

germ agglutinin (Sigma, St Louis, MO) to suppress cortical flow32. When

appropriate, mounting media also contained 100 µM 3,000 Mr neutral TR–dex

(Molecular Probes). After imaging commenced, caged InsP3 was photolysed by

focusing UV light through the objective and on to the egg cortex for 3–5 s. UV

light was derived from an HBO 100W AttoArc power supply (Atto Bioscience,

Rockville, MD) set at 15% intensity. The time-lapse, single optical plane data

was collected at 1.5 s intervals and the 4D data collected at 5 s intervals. The z-

step for 4D data collection ranged from 0.3–0.5 µm. For quantification of fluo-

rescence intensities, the NIH Object-Image v2.06 software was used. Z-series

and time-lapse, single optical plane movies and 4D renderings were generated

with Volocity v10 software (Improvision, Lexington, MA). All other image pro-

cessing was performed with Adobe Photoshop 3.0.

Note: Supplementary Information is available on the Nature Cell Biology website.

ACKNOWLEDGEMENTS

We thank H. Higgs for providing the original WASP GBD clone and P. Krieg for the

backbone Xenopus expression plasmid pCS2+. This work was supported by grants

from the National Institutes of Health to both W.B. and J.T.

COMPETING FINANCIAL INTERESTS

The authors declare that they have no competing financial interests.

Received 30 April 2003; Accepted 27 June 2003;Published online: 20 July 2003; DOI: 10.1038/ncb1025.

1. Welch, M. D. & Mullins, R. D. Cellular control of actin nucleation. Annu. Rev. Cell Dev.Biol. 18, 247–288 (2002).

2. Geli, M. I. & Riezman, H. Endocytic internalization in yeast and animal cells: similarand different. J. Cell Sci. 111, 1031–1037 (1998).

3. Kaksonen, M., Peng, H. B. & Rauvala, H. Association of cortactin with dynamic actinin lamellipodia and on endosomal vesicles. J. Cell Sci. 113, 4421–4426 (2000).

4. Rozelle, A. L., et al. Phosphatidylinositol 4,5-bisphosphate induces actin-based move-ment of raft-enriched vesicles through WASP-Arp2/3. Curr. Biol. 10, 311–320 (2000).

5. Schafer, D. A., D’Souza-Schorey, C. & Cooper, J. A. Actin assembly at membranes con-trolled by ARF6. Traffic. 1, 896–907 (2000).

6. Taunton, J. et al. Actin-dependent propulsion of endosomes and lysosomes by recruit-ment of N-WASP. J. Cell Biol. 148, 519–530 (2000).

7. Merrifield, C. J. et al. Annexin 2 has an essential role in actin-based macropinocyticrocketing. Curr. Biol. 11, 1136–1141 (2001).

8. Lee, E. & De Camilli, P. Dynamin at actin tails. Proc. Natl Acad. Sci. USA 99,161–166 (2002).

9. Orth, J. D., Krueger, E. W., Cao, H. & McNiven, M. A. The large GTPase dynamin regu-lates actin comet formation and movement in living cells. Proc. Natl Acad. Sci. USA99, 167–172 (2002).

10. Gundelfinger, E., Kessels, M. & Qualmann, B. Temporal and spatial coordination ofexocytosis and endocytosis. Nature Rev. Mol. Cell Biol. 4, 127–139 (2003).

11. Merrifield, C. J., Feldman, M. E., Wan, L. & Almers, W. Imaging actin and dynaminrecruitment during invagination of single clathrin-coated pits. Nature Cell Biol. 4,691–698 (2002).

12. Merrifield, C. J. et al. Endocytic vesicles move at the tips of actin tails in cultured mastcells. Nature Cell Biol. 1, 72–74 (1999).

13. Bement, W. M., Benink, H., Mandato, C. A. & Swelstad, B. B. Evidence for directmembrane retrieval following cortical granule exocytosis in Xenopus oocytes and eggs.J. Exp. Zool. 286, 767–775 (2000).

14. Whalley, T., Terasaki, M., Cho, M. S. & Vogel, S. S. Direct membrane retrieval into largevesicles after exocytosis in sea urchin eggs. J. Cell Biol. 131, 1183–1192 (1995).

15. Fesce, R., Grohovaz, F., Valtorta, F. & Meldolesi, J. Neurotransmitter release: fusion or‘kiss-and-run’? Trends Cell Biol. 4, 1–4 (1994).

16. Kline, D. & Nuccitelli, R. The wave of activation current in the Xenopus egg. Dev. Biol.111, 471–487 (1985).

17. Terasaki, M., Runft, L. L. & Hand, A. R. Changes in organization of the endoplasmic

Sig

nal i

nten

sity

(a

rbitr

ary

units

)

Figure 4 Actin coats assemble by de novo polymerization in associationwith active Cdc42 and N-WASP. (a) Single-plane, time-lapse confocalimages showing G-actin (Alexa488–G-actin, green) incorporates into coatsbefore phalloidin (Alexa568–phalloidin, red). (b) Incorporation of G-actin(G-act) or phalloidin (Phall), measured as signal intensity. Results arerepresentative of three experiments. Inset demonstrates incorporationduring earliest coat assembly. 0 s is just before coat assembly in a and b.(c) Single-plane, time-lapse confocal images showing active Cdc42

(eGFP-wGBD, green) recruited to exocytosing cortical granules (TR–dex,red). (d) Single-plane, time-lapse confocal images showing N-WASP(Alexa488–N-WASP truncate, green) recruited to exocytosing corticalgranules. (e) Single-plane, time-lapse confocal images showing thatCdc42T17N inhibits actin (Alexa488–G-actin, green) coat assembly,promoting fusion between adjacent exocytosing cortical granules(arrowheads). Scale bar represents 2.5 µm in a and 5 µm in c–e.Photolysis was performed at 0 s in c–e.


L E T T E R S


reticulum during Xenopus oocyte maturation and activation. Mol. Biol. Cell 12,1103–1116 (2001).

18. Cao, L. G. & Wang, Y. L. Mechanism of the formation of contractile ring in dividing cul-tured animal cells. I. Recruitment of preexisting actin filaments into the cleavage fur-row. J. Cell Biol. 110, 1089–1095 (1990).

19. Mandato, C. A. & Bement, W. M. Contraction and polymerization cooperate to assem-ble and close actomyosin rings around Xenopus oocyte wounds. J. Cell Biol. 154,785–797 (2001).

20. De La Cruz, E. & Pollard, T. Transient kinetic analysis of rhodamine phalloidin bindingto actin filaments. Biochemistry 33, 14387–14392 (1994).

21. Ma, L., Cantley, L. C., Janmey, P. A. & Kirschner, M. W. Corequirement of specificphosphoinositides and small GTP-binding protein Cdc42 in inducing actin assembly inXenopus egg extracts. J. Cell Biol. 140, 1125–1136 (1998).

22. Moreau, V., & Way, M. Cdc42 is required for membrane-dependent actin polymeriza-tion in vitro. FEBS Lett. 427, 353–356 (1998).

23. Rohatgi, R., et al. The interaction between N-WASP and the Arp2/3 complex linksCdc42-dependent signals to actin assembly. Cell. 97, 221–231 (1999).

24. Li, Z., Aizenman, C. D. & Cline, H. T. Regulation of rho GTPases by crosstalk and neu-ronal activity in vivo. Neuron 33, 741–750 (2002).

25. Higgs, H. N. & Pollard, T. D. Regulation of actin filament network formation through

Arp2/3 complex: activation by a diverse array of proteins. Annu. Rev. Biochem. 70,649–676 (2001).

26. Kozma, R., Ahmed, S., Best, A. & Lim, L. The Ras-related protein Cdc42Hs andbradykinin promote formation of peripheral actin microspikes and filopodia in Swiss3T3 fibroblasts. Mol. Cell. Biol. 15, 1942–1952 (1995).

27. Bernheim-Groswasser, A., Wiesner, S., Golsteyn, R., Carlier, M. & Sykes, C. Thedynamics of actin-based motility depend on surface parameters. Nature 417,308–311 (2002).

28. Upadhyaya, A., Chabot, J. R., Andreeva, A., Samadani, A. & van Oudenaarden, A.Probing polymerization forces by using actin-propelled lipid vesicles. Proc. Natl Acad.Sci. USA 100, 4521–4526 (2003).

29. Giardini, P. A., Fletcher, D. A. & Theriot, J. A. Compression forces generated by actincomet tails on lipid vesicles. Proc. Natl Acad. Sci. USA 100, 6493–6498 (2003).

30. Becker, K. A. & Hart, N. H. Reorganization of filamentous actin and myosin-II inzebrafish eggs correlates temporally and spatially with cortical granule exocytosis. J.Cell Sci. 112, 97–110 (1999).

31. Lee, E. & Knecht, D. A. Visualization of actin dynamics during macropinocytosis andexocytosis. Traffic 3, 186–192 (2002).

32. Canman, J. C. & Bement, W. M. Microtubules suppress actomyosin-based cortical flowin Xenopus oocytes. J. Cell Sci. 110, 1907–1917 (1997).


supplementary information

1

Movie 1. F-actin coats are highly dynamic. Single-plane, time-lapse confocal imag-ing shows that prior to CG exocytosis, cortical actin (Alexa488-G-actin, green)appears as a honeycomb, the CGs as black silhouettes. The black flash indicatescaged-IP3 photolysis to stimulate CG exocytosis. Actin coats assemble around exo-cytosing CGs (TR-dex, red) and constrict. Images collected ~2 µm in from the PM.(1.5 s intervals over 01:36).

Movie 2. Actin coats close over exocytosing CG compartments. Low-magnifica-tion, apical view, 4D imaging shows that the apical edges of the actin coats(Alexa488-G-actin, green) rise in ridges around the exocytosing CGs (TR-dex, red).The circumferential ridges constrict and actin sheets close over the compartmentinteriors. The black flash indicates caged-IP3 photolysis. (5 s intervals over 01:40)

Movie 3. Actin coats close over exocytosing CG compartments. High-magnification,apical view, 4D imaging shows that the circumferential ridge of the actin coat(Alexa488-G-actin, green) rises around the exocytosing CG (TR-dex, red) and con-stricts to close over the compartment. The black flash indicates caged-IP3 photoly-sis. (5 s intervals over 01:40)

Movie 4. Actin coats compress exocytosing CG compartments. High-magnification,basal view, 4D imaging shows that actin (Alexa488-G-actin, green) extends basallyalong the surface of the exocytosing CG (TR-dex, red) to form the coat that com-presses apically. The black flash indicates caged-IP3 photolysis. (5 s intervals over01:40)

Movie 5. Actin coats close over and compress exocytosing CG compartments.High-magnification 4D imaging of an exocytosing CG (TR-dex, red) in cross-sectionshows that the actin coat (Alexa488-G-actin, green) constricts in the plane of thePM and apically compresses throughout its entirety to squeeze the CG compart-ment. The black flash indicates caged-IP3 photolysis. (5 s intervals over 01:40)


Cdc42-dependent actin polymerization during compensatory...

Documents

Transcript of Cdc42-dependent actin polymerization during compensatory...