ena/vasp proteins: regulators of the actin cytoskeleton ... · proteins of the ena/vasp family have...

27 Aug 2003 14:43 AR AR197-CB19-21.tex AR197-CB19-21.sgm LaTeX2e(2002/01/18)P1: GCE10.1146/annurev.cellbio.19.050103.103356

Annu. Rev. Cell Dev. Biol. 2003. 19:541–64doi: 10.1146/annurev.cellbio.19.050103.103356

Copyright c© 2003 by Annual Reviews. All rights reservedFirst published online as a Review in Advance on July 8, 2003

ENA/VASP PROTEINS: Regulators of the ActinCytoskeleton and Cell Migration

Matthias Krause, Erik W. Dent, James E. Bear,Joseph J. Loureiro, and Frank B. GertlerDepartment of Biology, Massachusetts Institute of Technology, Cambridge,Massachusetts 02139; email: [email protected]

Key Words Mena, EVL, filopodia, lamellipodia, focal adhesion

■ Abstract Ena/VASP proteins are a conserved family of actin regulatory pro-teins made up of EVH1, EVH2 domains, and a proline-rich central region. They havebeen implicated in actin-based processes such as fibroblast migration, axon guidance,and T cell polarization and are important for the actin-based motility of the intra-cellular pathogenListeria monocytogenes. Mechanistically, these proteins associatewith barbed ends of actin filaments and antagonize filament capping by capping pro-tein (CapZ). In addition, they reduce the density of Arp2/3-dependent actin filamentbranches and bind Profilin at sites of actin polymerization. Vertebrate Ena/VASP pro-teins are substrates for PKA/PKG serine/threonine kinases. Phosphorylation by thesekinases appears to modulate Ena/VASP function within cells, although the mechanismunderlying this regulation remains to be determined.

CONTENTS

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542DOMAIN AND FUNCTIONAL ORGANIZATION OF THE Ena/VASPPROTEIN FAMILY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 544EVH1 Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 544Proline-Rich Domain: SH3 and WW Interactions and Profilin Binding. . . . . . . . . 545EVH2 Domain: Actin Binding and Tetramerization. . . . . . . . . . . . . . . . . . . . . . . . . 545Isoforms and Unique Features of Mena and EVL. . . . . . . . . . . . . . . . . . . . . . . . . . . 546

FUNCTIONS OF Ena/VASP PROTEINS IN CELLS AND TISSUES. . . . . . . . . . . . 546Role in Fibroblasts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546Role in Actin-Based Movement of Bacterial and Viral Pathogens. . . . . . . . . . . . . . 548Role in the Immune System: T-Cells and Macrophages. . . . . . . . . . . . . . . . . . . . . . 549Role in Neuronal Migration and Axon Guidance. . . . . . . . . . . . . . . . . . . . . . . . . . . 549Role in Endothelia and Epithelia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551Potential Role During Tumor Formation and Metastasis. . . . . . . . . . . . . . . . . . . . . 552

MOLECULAR FUNCTIONS OF Ena/VASP PROTEINS. . . . . . . . . . . . . . . . . . . . . . 552REGULATION OF Ena/VASP PROTEINS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554

Serine/Threonine Phosphorylation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554

1081-0706/03/1115-0541$14.00 541

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.

15 Sep 2003 13:26 AR AR197-CB19-21.tex AR197-CB19-21.sgm LaTeX2e(2002/01/18)P1: GCE

542 KRAUSE ET AL.

Tyrosine Phosphorylation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556FUTURE DIRECTIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 557

INTRODUCTION

One of the fundamental problems in biology is to understand how cells move withintissues. Remodeling of the actin cytoskeleton provides the driving force for cellmigration. Therefore, it is of great importance to unravel the signaling pathwaysthat regulate the actin polymerization cycle in living cells at the molecular level.Proteins of the Ena/VASP family have emerged as regulators of actin assemblyand cell motility in a variety of organisms and cell types for several reasons.

First, Ena/VASP proteins are implicated in the actin-dependent process of axonguidance inDrosophila, Caenorhabditis elegans, and mice.DrosophilaEnabled(Ena), one of the founding members of this protein family, was discovered in agenetic screen for dominant suppressors ofDrosophilaAbelson tyrosine kinase(D-Abl) mutants (Gertler et al. 1990). Hemizygosity forEnaalleviatesD-Abl mu-tant phenotypes. Complete removal ofEnacauses defects in the axonal architectureof the central nervous system (CNS) and the peripheral nervous system (PNS) inthe fly (Gertler et al. 1995). Much genetic and molecular evidence implicatesEnaand unc-34, the C. elegansortholog of Ena, in the pathways downstream of anumber of axon guidance receptors and signaling molecules (Bashaw et al. 2000,Gitai et al. 2003, Wills et al. 1999, Yu et al. 2002). In mice, deletion of mammalianEnabled (Mena), the mammalian Ena ortholog, causes defects in the formation ofnerve fiber tracts in the brain (Lanier et al. 1999).

Second, vasodilator-stimulated phosphoprotein (VASP), the other foundingmember of the Ena/VASP family, was discovered independently as a protein phos-phorylated in platelets in response to agents that elevate cAMP and cGMP (Haffneret al. 1995, Halbrugge & Walter 1989, Waldmann et al. 1987). VASP mutant miceexhibit defects in the actin-dependent process of platelet aggregation (Aszodi et al.1999, Hauser et al. 1999).

Third, Ena/VASP proteins localize within cells to areas of dynamic actin re-organization such as the leading edge of lamellipodia and at the tips of filopodiaand other actin-dependent intracellular structures such as cell-cell contacts, focaladhesions, and in periodic puncta along stress fibers (Gertler et al. 1996, Lam-brechts et al. 2000, Lanier et al. 1999, Reinhard et al. 1992, Rottner et al. 1999)(Figure 1). In addition, Ena/VASP proteins bind directly to F- and G-actin and theactin monomer-binding protein Profilin, suggesting that they could directly reg-ulate actin dynamics (Ahern-Djamali et al. 1999; Bachmann et al. 1999; Gertleret al. 1996; Huttelmaier et al. 1999; Lambrechts et al. 2000; Laurent et al. 1999;Reinhard et al. 1992, 1995a; Walders-Harbeck et al. 2002).

Fourth, Ena/VASP proteins are recruited to the actin tails formed byListeriamonocytogenes, an intracellular, pathogenic bacterium that uses the host’s actincytoskeleton to propel itself through the cytoplasm (Chakraborty et al. 1995).This was the first mammalian actin cytoskeletal-associated protein found to be

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.

27 Aug 2003 14:43 AR AR197-CB19-21.tex AR197-CB19-21.sgm LaTeX2e(2002/01/18)P1: GCE

Ena/VASP FUNCTION 543

TABLE 1 Strategies for modifying Ena/VASP function

Organism orMethod cell type Notes References

Loss-of-function: Drosophila(ena) Genetic lesion that causes no Gertler et al. 1990Random mutation C. elegans(unc-34) protein or nonfunctional protein Gitai et al. 2003

to be produced

Loss-of-function: Mouse (menaand Homologous recombination gene Lanier et al. 1999Gene targeting vasp), Dictyostelium targeting produces a protein-null Aszodi et al. 1999,

(Ddvasp) mutation Hauser et al. 1999Han et al. 2002

FPPPP-Mito Fibroblasts, cortical Sequesters Ena/VASP proteins Bear et al. 2000neurons to the mitochondria. Specificity Goh et al. 2002

control: APPPP-Mito

FPPPP-CAAX Fibroblasts Targets Ena/VASP proteins to the Bear et al. 2000plasma membrane. Specificitycontrol: APPPP-CAAX

FPPPP-Cyto T cells, fibroblasts, Displaces EnaVASP proteins from Bear et al. 2000macrophages interface of T cells with antigen- Krause et al. 2000

presenting cells and phagocytic Coppolino et al. 2001cups of macrophages. Infibroblasts it displaces Ena/VASPproteins from focal adhesions butnot from the leading edge.Specificity control: APPPP-Cyto

Overexpression: Fibroblasts Increased levels of wild-type Liu et al. 1999Full-length protein Ena/VASP protein Bear et al. 2000

Overexpression: Keratinocytes, Expression of the coiled-coil Vasioukhin et al. 2000Tetramerization mouse skin domain of VASP

Domain (TD) Transgenics Specificity control: none

Overexpression: Endothelial cells, Expression of the EVH2 domain Price et al. 2000EVH2 domain Epithelial cells, from various family members Lawrence et al. 2002

fibroblasts Specificity control: FAB deletion Loureiro et al. 2002

Membrane tethering Dictyostelium Direct targeting of DdVASP to Han et al. 2002by lipid anchor the plasma membrane by

N-myristolation

FPPPP peptide Neutrophils, Injection or osmotic shock Anderson et al. 2003loading/injection Listeria-infected loading of FPPPP-based peptides

PtK2 cells Specificity control: irrelevant peptide

Antisense Fibroblasts Random integration of transcription Liu et al. 1999trap retrovirus produces antisensetranscript

recruited toListeria, a system used as a model to study the basic principles of theactin cytoskeleton.

Finally, the use of genetic approaches and strategies to interfere with Ena/VASPfunction (see Table 1) reveals the requirements for Ena/VASP proteins in the devel-opmental and physiological processes of T cell activation, phagocytosis, epithelialmorphogenesis, and migration of neutrophils, fibroblasts, and neurons (Andersonet al. 2003; Bear et al. 2000, 2002; Coppolino et al. 2001; Goh et al. 2002;Grevengoed et al. 2001; Krause et al. 2000; Vasioukhin et al. 2000). All the

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.


544 KRAUSE ET AL.

above-mentioned processes depend upon regulated cytoskeletal remodeling andimplicate Ena/VASP as a key link between signaling pathways and actin dynamics.

DOMAIN AND FUNCTIONAL ORGANIZATIONOF THE Ena/VASP PROTEIN FAMILY

The Ena/VASP family consists ofDrosophilaEna,C. elegansUnc-34,Dictyo-steliumDdVASP, and the three mammalian family members VASP, Mena, andEVL (Ena-VASP-like). All Ena/VASP family members share a conserved domainstructure. An amino-terminal Ena/VASP homology 1 (EVH1) domain followedby a proline-rich central region and a carboxy-terminal Ena/VASP homology 2(EVH2) domain.

EVH1 Domain

The EVH1 domain belongs to the PH domain superfamily (Ball et al. 2000, Barziket al. 2001, Beneken et al. 2000, Fedorov et al. 1999, Prehoda et al. 1999). UnlikePH domains, EVH1 domains do not appear to bind to phosphatidylinositol lipids,although only PI(4,5)P2 has been tested (Volkman et al. 2002). Instead, the EVH1domain, similar to PTB domains (another branch of the PH domain superfamily),binds to peptide ligands with high affinity. EVH1 domains are also found in theWASP family, the Homer/Vesl family, and two other proteins, Spred and SMIF(Brakeman et al. 1997, Callebaut 2002, Callebaut et al. 1998, Gertler et al. 1996,Kato et al. 1997, Symons et al. 1996, Wakioka et al. 2001). Similar to WW and SH3domains, the characterized EVH1 domains appear to bind to peptides containinga poly-proline II helix (PPII helix), a left-handed helix with three residues per turn(Ball et al. 2000, Carl et al. 1999, Fedorov et al. 1999, Niebuhr et al. 1997, Prehodaet al. 1999). Flanking residues in the vicinity of the PPII helix confer specificityof binding to SH3, WW, and EVH1 domains. The EVH1 domain in Ena/VASPproteins binds to the consensus site (F/W/Y/L)PPPPX(D/E)(D/E)(D/E)8 (X= anyamino acid,8= hydrophobic amino acid). Individual prolines in the central corecan be exchanged for F/W/Y and, in some cases, for other amino acids (Ball et al.2000, Carl et al. 1999, Niebuhr et al. 1997). The other EVH1 domains found inthe Homer/Vesl and WASP protein families have binding specificities that differfrom those of the Ena/VASP family (Barzik et al. 2001, Beneken et al. 2000,Tu et al. 1998, Volkman et al. 2002, Zettl & Way 2002).

Functional binding sites for EVH1 domains are found in Zyxin, LPP, Vinculin,Fyb/SLAP, theDrosophilaandC. elegansaxon guidance receptor Robo/Sax-3,and in the ActA protein ofListeria monocytogenes(Bashaw et al. 2000, Brindleet al. 1996, Drees et al. 2000, Krause et al. 2000, Niebuhr et al. 1997, Petit et al.2000, Reinhard et al. 1996, Yu et al. 2002). EVH1 domains recruit Ena/VASPproteins to specific sites within the cell. Ena/VASP proteins are recruited to focaladhesions by Zyxin and Vinculin, to phagocytic cups of macrophages and the

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.



contact site of T cells with antigen-presenting cells by Fyb/SLAP, and toListeriaby ActA. Other known proteins such as Palladin, Migfilin, human FAT, humanAnkyrinG, and several novel proteins also harbor putative EVH1-binding sites(Mykkanen et al. 2001, Niebuhr et al. 1997, Parast & Otey 2000, Tu et al. 2003;M. Krause & F.B. Gertler unpublished data).

Proline-Rich Domain: SH3 and WW Interactionsand Profilin Binding

The central proline-rich region harbors binding sites for SH3 and WW domain–containing proteins and the actin monomer-binding protein Profilin. This regionis the most divergent within the family and therefore may have different bindingpartners and mechanisms of regulation. Ena binds to the SH3 domains of Abl,Src, and the carboxy-terminal SH3 domain of Drk (Ahern-Djamali et al. 1999,Comer et al. 1998, Gertler et al. 1995). Similarly, EVL binds to the SH3 domainsof Lyn, N-Src, Abl, and the WW domain of FE-65 (Lambrechts et al. 2000). Incontrast, Mena does not bind to the SH3 domain of N-Src, but is bound by theSH3 domain of IRSp53, Abl, Arg, Src, and the WW domain of FE65 (Ermekovaet al. 1997, Gertler et al. 1996, Krugmann et al. 2001). It is not clear which ofthese interactions is physiologically significant and shared among all members ofthe Ena/VASP family.

All Ena/VASP family members contain proline-rich-binding sites for the smallG-actin-binding protein Profilin, and this binding is independent of their phospho-rylation status (Ahern-Djamali et al. 1999, Gertler et al. 1996, Lambrechts et al.2000, Reinhard et al. 1995b). Profilin II, the major Profilin isoform expressed inbrain tissues, binds as a dimer with high affinity to VASP but with low affinityto PI(4,5)P2. In contrast, Profilin I displays the opposite preferences (Jonckheereet al. 1999).

EVH2 Domain: Actin Binding and Tetramerization

The EVH2 domain of the Ena/VASP family harbors three conserved blocks: start-ing from the amino terminus, a G-actin-binding site, an F-actin-binding site, anda coiled-coil motif required for oligomerization (Bachmann et al. 1999, Gertleret al. 1996). The G-actin-binding site resembles a motif in the small actin-monomer-binding proteinβ-thymosin (Gertler et al. 1996) and is required for the in vitroactin nucleation activity of VASP (Walders-Harbeck et al. 2002). The secondconserved block in the EVH2 domain of VASP (259–276 aa in human VASP)can co-sediment and bundle F-actin at low salt conditions (33 and 50 mM KCl)(Bachmann et al. 1999, Huttelmaier et al. 1999). The third conserved block (336–380 aa in human VASP) consists of a coiled-coil structure that mediates oligomer-ization of Ena/VASP proteins (Bachmann et al. 1999, Zimmermann et al. 2002).VASP behaves as a tetramer, and the coiled-coil motif is sufficient to mediatetetramer formation (Bachmann et al. 1999, Haffner et al. 1995, Zimmermannet al. 2002). Furthermore, co-immunoprecipitation and yeast two-hybrid analysis

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.


546 KRAUSE ET AL.

indicate that homo- and heterotetramerization among the members of the Ena/VASPfamily can also be observed in vivo (Ahern-Djamali et al. 1998, Carl et al. 1999;F.B. Gertler unpublished data). The presence of the coiled-coil within VASP en-hances in vitro actin binding and bundling, suggesting that oligomerization facil-itates these activities (Bachmann et al. 1999, Walders-Harbeck et al. 2002).

Isoforms and Unique Features of Mena and EVL

Following the EVH1 domain, Mena contains a long insertion not found in otherEna/VASP family members. This insertion in Mena is a striking five-amino acidlong stretch of highly charged basic and acidic amino acids (LERER) that arerepeated 14 times. This repeat might adopt an extended helical structure that func-tion as a protein-protein-binding interface (Gertler et al. 1996). Mena’sDrosophilaortholog Ena contains a glutamine-rich insertion in the same relative part of themolecule with no known function.

At least four different isoforms of Mena can be detected on Western blots oftissue extracts. The smallest isoform of Mena has been found in B-lymphoid cellsand arises from the omission of an exon coding for part of the proline-rich region inthe Mena 80-kDa isoform (Mena-80). Tani and colleagues described this isoformas 80 kDa and named it Mena-s, but because its apparent molecular mass is clearlysmaller than the originally described Mena 80-kDa isoform, we propose to renameit as Mena-75 (Gertler et al. 1996, Tani et al. 2003). The 88-kDa Mena isoformmight arise from the inclusion of a known or unknown exon and is expressed indeveloping tissues (Gertler et al. 1996). All the larger Mena isoforms (Mena-140)contain a large exon that introduces 246 additional amino acids after the LERERrepeats. The amino acid sequence introduced by this exon is proline-rich. SomecDNAs containing this exon also include two additional small exons that introduce4 and 19 amino acids, respectively, directly after the EVH1 domain (Gertler et al.1996). Mena-140 is expressed in developing and adult neuronal tissues (Lanieret al. 1999) and is tyrosine phosphorylated in vivo (Gertler et al. 1996). EVL isexpressed in two isoforms that migrate differently in SDS-gels (Lambrechts et al.2000). The larger isoform, EVL-I, arises from an alternative exon coding for 21amino acids spliced into the EVH2 domain just before the coiled-coil motif.

FUNCTIONS OF Ena/VASP PROTEINS IN CELLSAND TISSUES

Role in Fibroblasts

Fibroblast migration is a complex process in which forward protrusion, attachment,contraction, and rear detachment are integrated to translocate the cell. To studythe role of Ena/VASP proteins in fibroblast motility, cell lines were derived frommouse embryos devoid of Mena and VASP. One fibroblastic cell line from thiscollection that lacked expression of all Ena/VASP proteins (including the third

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.



mammalian family member, EVL) was analyzed by video microscopy. When thiscell line (the MVD7 cells) is rescued with a physiological level of Mena, EVL,or VASP, the cells moved more slowly than the parental null line. Conversely, theoverexpression of Ena/VASP proteins in normal fibroblasts led to slower movement(Bear et al. 2000).

Rescue of the fibroblasts devoid of Ena/VASP proteins with deletion mutantsof Mena allowed the structural requirements of Ena/VASP proteins in fibroblastmotility to be examined (Loureiro et al. 2002). Re-expression of a mutant lackingthe proline-rich region slowed cells down to the same extent as the wild-type pro-tein, indicating that Profilin and other ligand recruitment mediated by this regionare dispensable for random cell migration. However, the proline-rich region is re-quired for other Ena/VASP-dependent processes (see below). The F-actin-bindingsite is essential for proper fibroblast motility and for the ability of Ena/VASPproteins to be recruited effectively to the leading edge.

An alternate approach to block Ena/VASP function is to remove the proteinsfrom their normal localization within cells. As noted above, the EVH1 domain ofEna/VASP proteins binds strongly to the peptide motif (D/E)FPPPX(D/E)(D/E).This interaction can be exploited to sequester Ena/VASP proteins on the surfaceof mitochondria by expression of a targeting construct (called FPPPP-Mito, seeTable 1) that contains four copies of the FPPPP motif linked to a mitochondrialanchoring sequence. FPPPP-Mito blocks the function of all Ena/VASP familymembers by deleting them from their normal sites of function and replicates thephenotypes observed in Ena/VASP-deficient cells (Bear et al. 2000). Expressionof the FPPPP-Mito construct in the null fibroblasts caused no change in cell speed,confirming the specificity of this approach. Taken together, these observations us-ing the null cells and re-localization strategies led to the conclusion that Ena/VASPproteins negatively regulate fibroblast migration speed.

When Ena/VASP proteins are selectively de-localized from focal adhesions, butnot from the leading edge (using FPPPP-Cyto, see Table 1), random fibroblast mi-gration speed on a fibronectin substrate is not affected, indicating that Ena/VASPproteins do not have a critical function in focal adhesions in this type of migrationassay. At a gross level, the composition, number, and morphology of focal adhe-sions is unaffected by loss of Ena/VASP (Bear et al. 2000). At present, the functionof Ena/VASP proteins within focal adhesions is unclear. It will be important to de-termine whether Ena/VASP proteins are required for focal adhesion function inother cell types or when cells are plated on substrates other than fibronectin.

The function of Ena/VASP proteins as negative regulators of fibroblast motility(Bear et al. 2000) seemed paradoxical given that lamellipodial protrusion rate pos-itively correlates with the intensity of GFP-VASP at the leading edge (Rottneret al. 1999). A careful examination of the dynamics of lamellipodia lackingEna/VASP proteins resolved this apparent conflict and indicated that such lamel-lipodia protrude more slowly, but the protrusion persists longer. Conversely, cellsoverexpressing Ena/VASP proteins have lamellipodia that protrude faster than con-trols, but these protrusions are rapidly withdrawn in the form of ruffles. Therefore,

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.


548 KRAUSE ET AL.

Ena/VASP proteins increase lamellipodial protrusion velocity but have a globalnegative effect on cell motility. These observations have led to a more general con-clusion that whole-cell motility rates correlate better with persistence of protrusionthan they do with instantaneous protrusion rate (Bear et al. 2002).

Interestingly, deletion or re-localization of Ena/VASP proteins from the leadingedge causes reorganization of actin filament architecture in lamellipodia (Bear et al.2002). Lamellipodia lacking Ena/VASP function contain actin filament networkswith shorter and more highly branched actin filaments. Excess Ena/VASP leadsto longer but less-branched filaments. These changes in actin network geometryare responsible for the changes in lamellipodial behavior and, ultimately, changesin cell migration rate. Modifying Ena/VASP function in fibroblasts has led to amolecular model of Ena/VASP function that is discussed below.

Role in Actin-Based Movement of Bacterial andViral Pathogens

Like several other bacteria and viruses,Listeria monocytogenesis able to invademammalian cells and use the host cell’s actin cytoskeleton to move within andfrom cell to cell.Listeria requires only one bacterial surface protein, ActA, forits intracellular motility (Domann et al. 1992, Kocks et al. 1992). ActA functionsas a nucleator of F-actin (Pistor et al. 1994) by co-opting the host cells ownArp2/3 complex for F-actin nucleation (Pistor et al. 2000; Skoble et al. 2000;Welch et al. 1997, 1998). The finding that VASP was recruited to and directlybinds, through its EVH1 domain, to proline-rich repeats within ActA was thefirst indication that Ena/VASP proteins might directly regulate actin dynamics(Chakraborty et al. 1995, Gertler et al. 1996, Niebuhr et al. 1997, Pistor et al.1995). Deletion of these proline-rich repeats within ActA led to a reduction ofintracellular speed and reduced pathogenicity (Lasa et al. 1997, Niebuhr et al.1997, Smith et al. 1996). As mentioned above, Ena/VASP proteins bind directlyin vitro to Profilin, suggesting that the recruitment of Profilin by Ena/VASP proteinsmay be one function of this protein family important forListeria motility. Thisview is supported by the reduction ofListeria speeds after depletion experimentsof either Ena/VASP proteins or Profilin from cell extracts and by the restoration ofListeriamovement in both add-back experiments and from reconstitution of pureproteins (Theriot et al. 1994, Kang et al. 1997, Laurent et al. 1999, Loisel et al.1999). Ena/VASP proteins also facilitate ActA-induced Arp2/3-mediated actinnucleation, possibly by increasing the amount of actin monomer available on thebacterial surface (Skoble et al. 2000). Finally, Ena/VASP proteins decrease thenumber of F-actin branches induced by ActA and the Arp2/3 complex in vitro,suggesting that they might regulate the geometry of the actin filament network inthe tails of motile bacteria (Skoble et al. 2000).

Within living cells, Listeria speeds are reduced an order of magnitude in theabsence of Ena/VASP proteins. Expression of Ena/VASP deletion mutants lackingeither Profilin or the G-actin-binding site in Ena/VASP null cells provided only a

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.



modest increase in bacterial speed relative to wild-type protein, indicating that bothdomains are important forListeria motility. In contrast, deletion of the F-actin-binding site increased theListeria speed to higher than wild-type levels, whereasdeletion of the coiled-coil domain had no effect (Geese et al. 2002). Because thesame set of mutants yielded different results when assayed for function in whole-cell motility (described above), the molecular requirements for Ena/VASP functionin ActA-mediated bacterial movement likely differ in some important respects fromtheir function in regulating lamellipodial dynamics. Therefore, it is important toconsider that Ena/VASP proteins probably fulfill different functions depending onthe cellular context, indicating that results obtained assaying a particular functionmay not be entirely translatable to other sytems (Loureiro et al. 2002). Ena/VASPproteins are also localized to the F-actin tail or to surfaces of other intracellularpathogens such asShigella flexneri, Rickettsia, andVacciniavirus, but the role, ifany, of Ena/VASP in the motility of these pathogens is unclear (Chakraborty et al.1995, Frischknecht et al. 1999, Van Kirk et al. 2000).

Role in the Immune System: T-Cells and Macrophages

The physiological function of the immune system depends on dynamic changesof the actin cytoskeleton (Penninger & Crabtree 1999). Actin polymerization isrequired for the interaction of T cells with antigen-presenting cells, phagocytosisby macrophages, and chemotaxis of neutrophils toward bacteria.

Ena/VASP proteins co-localize with the hematopoietic-specific adaptor proteinFyb/SLAP; the adaptor molecules SLP-76, Nck, Vav, WASP; and the F-actin nu-cleator Arp2/3 complex in phagocytic cups of macrophages. They also co-localizeat the interface between T cells and anti-CD3 coated beads (to mimic antigen-presenting cells) (Coppolino et al. 2001, Krause et al. 2000). This localizationof Ena/VASP is mediated by direct binding of the EVH1 domain to Fyb/SLAP(Krause et al. 2000). Ena/VASP proteins are also detected in a protein complex withFyb/SLAP, SLP-76, Nck, Vav, and the Arp2/3 activator WASP upon ligation of theT cell receptor or the Fcγ receptor of macrophages, indicating that two key mod-ulators of the actin dynamics, Ena/VASP and WASP, can be linked in regulated,physiological processes (Coppolino et al. 2001, Krause et al. 2000). Importantly,properly localized Ena/VASP proteins and Arp2/3 are each required for effectiveJurkat T cell polarization and for phagocytosis (Coppolino et al. 2001, Krause et al.2000, May et al. 2000), suggesting that Ena/VASP-WASP-Arp2/3 networks mayfunction together to promote these processes. Recently, peptide-loading exper-iments of neutrophils with putative Ena/VASP function-blocking peptides weredescribed (Anderson et al. 2003) that resulted in reduced migration velocity ofneutrophils (see Table 1).

Role in Neuronal Migration and Axon Guidance

The development of the nervous system involves extensive migration of neurons.To study the role of Ena/VASP proteins in neuronal migration, the FPPPP-Mito

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.


550 KRAUSE ET AL.

approach described above was used to block Ena/VASP function in cortical neu-rons (Goh et al. 2002). Neutralization of Ena/VASP function within neurons thatnormally migrate radially from their birthplace to the deep layers of the cortexresulted in frequent misplacement of these neurons to more superficial layers.Analysis of early stages of neuronal migration indicated that neurons expressingthe FPPPP-Mito construct were already present in more superficial structures andtherefore moved farther than control cells. These data suggest that Ena/VASP pro-teins play a key role in neuronal migration, although their effects on cell speed andmode of migration in cortex remain to be determined.

During development, growth cones, the motile tips of growing processes, musttraverse great distances and continually integrate a plethora of extracellular guid-ance signals into appropriate changes in cytoskeletal dynamics for proper move-ment. There are several reasons to believe that Ena/VASP proteins play impor-tant roles in intracellular signaling cascades that transduce extracellular guidancesignals into actin dynamics within growth cone filopodia and lamellipodia. First,Ena/VASP proteins are concentrated at the tips of growth cone filopodia and lamel-lipodia in primary cultured neurons (Figure 2) (Lanier et al. 1999). Second, all threevertebrate family members are regulated by protein kinase A (PKA), and at leastVASP is known to be phosphorylated by protein kinase G (PKG) (Aszodi et al.1999, Gertler et al. 1996, Halbrugge et al. 1990, Lambrechts et al. 2000, Loureiroet al. 2002). PKA and PKG are two important kinases that modulate signalingdownstream of many axon guidance molecules (Song & Poo 1999). Third, geneticstudies in worms and flies implicate their Ena/VASP homologs in response to spe-cific axonal guidance signals. In worms,unc-34is required for proper response tonetrin, a guidance factor that can act either as an attractant or a repellent dependingon the receptor repertoire and signaling context present within responding neurons(Colavita & Culotti 1998, Gitai et al. 2003). In both flies and worms, Ena/VASPappears to function downstream of the repulsive guidance receptor Robo/Sax3, amolecule that binds directly to Ena/VASP through an EVH1-binding site foundon its cytoplasmic tail (Bashaw et al. 2000, Yu et al. 2002). Genetic data alsosuggest thatDrosophilaEna and D-abl may act antagonistically downstream ofcertain axon guidance receptors (Bashaw et al. 2000, Wills et al. 1999). Otherevidence suggests D-lar, a receptor tyrosine phosphatase that antagonizes D-ablfunction in peripheral nervous system guidance (Wills et al. 1999), may be linkedto Ena function. Deletion of either Ena or D-lar induces similar phenotypes inwhich intersegmental nerve b (ISNb) peripheral nerves fail to branch at the correctposition and instead bypass their muscle target and extend beyond their normalbranching point (Wills et al. 1999). Potential regulation of Ena by D-abl and D-larare discussed below.

In mice, deletion of Mena caused axonal guidance defects in the formation ofthe corpus callosum, hippocampal commissure, and pontocerebellar fiber bundles(Lanier et al. 1999). Because all vertebrate Ena/VASP family members are ex-pressed in the developing vertebrate central nervous system (Lanier et al. 1999),it is likely that the continued presence of EVL and VASP in the nervous system

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.



of Mena mutants permits normal formation of other nerve connections that couldutilize EVL and/or VASP for their guidance. Studies of compound mutant animalsor the use of inhibitory approaches will be required to uncover the full role ofEna/VASP function in nervous system development and specifically in axon guid-ance. It will be particularly interesting to determine whether vertebrate Ena/VASPproteins play a role in conveying the PKA and PKG-dependent responses fromguidance receptors to the cytoskeleton.

Role in Endothelia and Epithelia

Endothelial cells line blood vessels and provide both a barrier function betweenblood and other tissue while still allowing immune cells access to the surround-ing tissue. This barrier function is maintained through tightly controlled cell-celladhesions and is positively regulated by PKA. Interestingly, phospho-VASP hasbeen localized to cell-cell junctions and could be co-immunoprecipitated with thetight-junction marker zonula-occludens-1 (ZO-1) protein from endothelial cells(Comerford et al. 2002).

Similarly, by adhering to each other, epithelial cells form a physical boundarycapable of regulating the transfer of molecules across the epithelial sheet. Thispermits selective exchange of molecules between the organism and its environ-ment, exemplified in the waste removal function of the kidney or nutrient uptakein the gut. In these systems Ena/VASP proteins localize to adherens junctions, anddisruption of their function affects epithelial dynamics (Grevengoed et al. 2001,Lawrence et al. 2002, Reinhard et al. 1992, Vasioukhin & Fuchs 2001). Further-more, VASP can be co-immunoprecipitated with the tight junction protein ZO-1from epithelial cells, and blocking PKA phosphorylation reduces the barrier func-tion of epithelial sheets as measured by electrical conductivity across the epithelialbarrier (Lawrence et al. 2002).

Studies inDrosophilahave demonstrated that reduction of either D-Abl or Enaaffects epithelial morphology and epithelial sheet migration. Genetic interactionssuggest that Ena functions in concert with components of the cadherin-catenincomplex (which mediates cell-cell adhesion) and D-Abl in epithelial sheets at theadherens junction (Grevengoed et al. 2001). In mammals, Ena/VASP proteins arefound in epithelial contacts that are revealed as a double row of dot-like structuresalso containing F-actin, Cadherin, Zyxin, and Vinculin. These cell-cell contactshave been suggested to represent an early stage in cell-cell contact formation andhave been termed the adhesion zipper (Vasioukhin et al. 2000). It has been pro-posed that cell-cell contacts are generated after first contact of lamellipodia and/orfilopodia of opposing cells (Ehrlich et al. 2002, Vaezi et al. 2002). Overexpressionof a fragment of Ena/VASP responsible for oligomerization (see Table 1) disruptsepithelial sheet formation (Vasioukhin et al. 2000). However, the specificity of thisapproach for Ena/VASP proteins has not been demonstrated. Although it seemslikely that Ena/VASP proteins play an important role in epithelial dynamics, furtherwork will be required to clarify their role in such processes.

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.


552 KRAUSE ET AL.

Potential Role During Tumor Formation and Metastasis

Up- or down-regulation of Ena/VASP proteins has not been decisively linked totumor progression in humans. However, it has been reported that either increas-ing (by overexpression) or decreasing (by antisense transcriptional knockdown,see Table 1) levels of VASP protein in NIH 3T3 fibroblasts may cause a tumori-genic phenotype when these cells have been injected into nude mice (Liu et al.1999). This observation is curious given that Mena, VASP, and EVL have broadoverlapping expression patterns and exhibit extensive functional overlap in vitroand in vivo. Furthermore, a similar phenotype has not been reported for VASP orMena knockout mice, which are adult-viable when individually mutated. Giventhat Ena/VASP proteins affect the mitogenic properties of cells, it is tempting tospeculate that loss-of-Ena/VASP-function potentiates metastasis. However, furtherstudies are necessary to clarify the role of Ena/VASP, if any, during tumorigenesis.

MOLECULAR FUNCTIONS OF Ena/VASP PROTEINS

How can the Ena/VASP-dependent phenotypes seen in fibroblast migration,Lis-teria motility, and other actin-dependent processes be explained at the molecularlevel? Early models of Ena/VASP function suggested a role in the de novo nucle-ation of actin filaments. Like many other actin-binding proteins, purified Ena/VASPproteins can shorten the lag phase of actin polymerization in in vitro assays(Bachmann et al. 1999, Huttelmaier et al. 1999, Lambrechts et al. 2000, Laurentet al. 1999). However, no convincing evidence supports this as a physiologicalmechanism for Ena/VASP function in vivo. One report showed that Ena/VASPproteins targeted to the mitochondria could recruit F-actin, supposedly throughnucleation of new filaments (Fradelizi et al. 2001). However, these experimentsused permeabilized cells and added exogenous actin. Several other laboratoriesfailed to observe F-actin recruitment when Ena/VASP proteins were relocalizedto mitochondria in intact cells. Furthermore,Listeria that exhibit defective Arp2/3activation but still recruit normal levels of Ena/VASP proteins fail to assembleor recruit any detectable F-actin (Bear et al. 2000; Lasa et al. 1997; Pistor et al.1994, 1995; Skoble et al. 2000). More recent experiments both in vivo and in vitropoint to three likely molecular mechanisms of Ena/VASP function: anti-capping,anti-branching, and Profilin recruitment.

Ena/VASP proteins may act to prevent or delay capping of barbed ends of actinfilaments by CapZ. Multiple lines of evidence support this anti-capping mecha-nism. Ena/VASP proteins are localized to the leading edge of lamellipodia and thetips of filopodia. These locations contain a high density of actin filament barbedends, and Ena/VASP localization to these places can be displaced by treatmentwith low concentrations of cytochalasin D (CD) (Bear et al. 2002). This drugbinds with high affinity to the barbed ends of actin filaments and presumably dis-places Ena/VASP proteins by competition. This observation is supported by in vitroassays in which Ena/VASP-coated beads can capture uncapped, but not capped,

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.



actin filaments. In addition, purified VASP antagonizes the activity of CapZ in actinfilament elongation assays. Perhaps the most compelling evidence for this modelcomes from studies of fibroblasts with modified Ena/VASP function. Depletion ofEna/VASP leads to shorter actin filaments at the leading edge relative to controls.Conversely, overexpression of Ena/VASP leads to longer than normal filamentsat the leading edge (Bear et al. 2002). These data suggest a mechanism in whichEna/VASP proteins associate with actin filaments at or near their barbed ends toprevent capping by CapZ. This model is supported by experiments in vivo in whichlow doses of CD were used to reverse Ena/VASP-dependent motility phenotypesin fibroblasts. This CD treatment mimics CapZ overexpression and reduced thefilament length to wild-type levels, indicating that a balance between CapZ levelsand Ena/VASP activity regulates actin filament geometry in lamellipodia (Bearet al. 2002).

Ena/VASP proteins may also act to inhibit the process of Arp2/3-mediated actinfilament branching or to reduce branch stability. Results from in vitro experimentsshowed that VASP decreases the number of F-actin branches induced by ActA fromListeriaand the Arp2/3 complex (Skoble et al. 2001). Another study, however, sawno effect of VASP on branching (Boujemaa-Paterski et al. 2001). In support of thework from Skoble and co-workers, modification of Ena/VASP activity in fibroblastsled not only to changes in filament length but also to changes in branching density.High Ena/VASP activity was associated with low branching density, whereas lowEna/VASP activity led to high branching density. The association of Ena/VASPproteins with actin filaments may either compete with Arp2/3 for binding sitesnear the barbed end or condition the filaments in a way to inhibit branching or topromote debranching. Future experiments will be required to understand the anti-branching activity of Ena/VASP proteins on the molecular level. Interestingly, therescue of the fibroblast motility phenotypes by low doses of CD selectively affectedfilament length, but not branching. This suggests that changes in length alone mayexplain the fibroblast motility results (Bear et al. 2002).

Ena/VASP function may also involve binding to Profilin at sites of actin reor-ganization. Genetic experiments in mice indicate that Mena mutants are sensitiveto a 50% reduction in Profilin I levels (Lanier et al. 1999). Normally, Profilin I het-erozygotes are viable despite a twofold reduction in Profilin levels. When bred intoa Mena mutant background, heterozygositity for Profilin I caused Mena mutantsto die perinatally and suffer neurulation defects. This genetic experiment indicatesthat Mena and Profilin I are both involved in the actin-dependent process of neuraltube closure. Ena/VASP proteins bind directly to Profilin in vitro. In vitro assaysalso show that a peptide containing the Profilin-binding site from VASP affects thebiochemical properties of Profilin II, permitting actin nucleation to occur in thepresence of the actin-sequestering protein, thymosin beta-4, which would other-wise block actin nucleation when combined with Profilin (Jonckheere et al. 1999).

As noted above, Profilin recruitment by Ena/VASP plays an important role insupportingListeria motility within living cells. EGFP-Profilin co-localizes onlywith movingListeria, and its intensity correlates positively with the speed of the

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.


554 KRAUSE ET AL.

bacteria (Geese et al. 2000). In addition, injection of cross-linked Profilin-actin(which inhibits the elongation rate of actin filaments in vitro) intoListeria-infectedcells reduces bacterial speed. A mutant version of this cross-linked Profilin-actincomplex that cannot bind to Ena/VASP proteins does not reduce speed (Grenkloet al. 2003). Consistent with this model, Ena/VASP-deficient cells that are com-plemented with forms of Mena or VASP that cannot bind Profilin (the proline-richdeletion) do not supportListeria movement as effectively as wild-type protein(Geese et al. 2002). However, this same experiment indicates that Ena/VASP pro-teins play additional roles inListeria motility that are independent of Profilin re-cruitment. Furthermore, Profilin-binding-defective mutants of Mena fully supportnormal fibroblast movement, indicating that this process does not require Profilinrecruitment by Ena/VASP. Cellular processes other than random fibroblast migra-tion may also be dependent on an Ena/VASP-dependent Profilin recruitment (e.g.,filopodia extension or phagocytosis).

What molecular role might Ena/VASP proteins play in other cellular processessuch as filopodial formation? Recently, Svitkina and colleagues showed that filopo-dia arise from the bundling of preexisting filaments within the actin filament net-work of lamellipodia, rather than from de novo actin filament nucleation. Inter-estingly, GFP-VASP was identified as an early marker of filopodial precursors.Recruitment of VASP was followed by subsequent recruitment of the F-actin-bundling protein, fascin, and filopodia formation (Svitkina et al. 2003). In anotherstudy, a knockout of the sole Ena/VASP family protein inDictyostelium, DdVASP,resulted in the absence of filopodia, providing evidence that Ena/VASP proteinsare required for filopodia formation (Han et al. 2002). In vitro reconstitution offilopodia with purified proteins or cellular extracts demonstrates that lowering theamount of CapZ can transform a dendritic network into a bundled filopodia-likestructure (Vignjevic et al. 2003). According to this model, the anti-capping activityof Ena/VASP would be expected to facilitate filopodial formation.

REGULATION OF Ena/VASP PROTEINS

Serine/Threonine Phosphorylation

The mammalian Ena/VASP proteins are known substrates of cAMP- and cGMP-dependent serine and threonine kinases, PKA and PKG, respectively (Figure 3)(Gertler et al. 1996, Lambrechts et al. 2000, Waldmann et al. 1987). VASP har-bors three phosphorylation sites (Ser-157, Ser-239, and Thr-278), whereas Menacontains the first two, and EVL only the first site (Butt et al. 1994, Gertler et al.1996, Lambrechts et al. 2000). Phosphorylation of Ser-157 of VASP leads to ashift in apparent molecular mass in SDS-PAGE from 46 to 50 kDa, indicating thatthis phosphorylation may cause a change in secondary structure of the molecule(Halbrugge & Walter 1989). Phosphorylation of Mena and EVL at the position thatis equivalent to VASP Ser-157 also induces bandshifts in their mobility (Gertler etal. 1996, Lambrechts et al. 2000).

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.



For VASP, PKG initially phosphorylates the second site then the first, and fi-nally the third site in vitro. Conversely, PKA phosphorylates, in vitro and in intactplatelets, the first, second, and third sites in order (Butt et al. 1994, Halbrugge et al.1990). However, in vivo PKG appears to phosphorylate VASP with different ki-netics: The first two sites are phosphorylated in intact platelets with a similar timecourse but only 50% of VASP shifts from 46 to 50 kDa (Butt et al. 1994).

PKA/PKG phosphorylation of VASP at the various conserved sites has beenobserved in response to a variety of physiological stimuli (Walter et al. 1993). Inplatelets, for example, inhibition of fibrinogen binding to the fibrinogen receptorαIIbβ3 integrin correlates with VASP phosphorylation in platelets, suggesting arole for VASP in the regulation of fibrinogen receptor activation (Halbrugge et al.1990, Horstrup et al. 1994, Waldmann et al. 1987). Studies of mutant mice revealedthat VASP is required for the PKA-mediated inhibition of platelet aggregation(Aszodi et al. 1999, Hauser et al. 1999), suggesting that VASP is the criticalPKA substrate required for this process, and that Ser-157 may be the key site ofregulation.

VASP is also phosphorylated by PKA on its first phosphorylation site upon de-tachment of fibroblasts and is dephosphorylated transiently during reattachment tofibronectin. During subsequent cell spreading, VASP becomes heavily phosphory-lated again (Howe et al. 2002). Whether such phosphorylation reflects the overallphosphorylation of all VASP within the cell or whether VASP at the leading edgeand VASP at focal adhesions are differently phosphorylated remains an importantyet unsolved question.

In random fibroblast migration, phosphorylation of Mena at its first phospho-rylation site (S-236, the equivalent of VASP Ser-157) is essential for its func-tion. A non-phosphorylatable S-236-A Mena mutant localizes normally when ex-pressed in Ena/VASP-deficient cells, yet fails to rescue the hypermotile phenotypeof Ena/VASP null cells (Loureiro et al. 2002). Conversely, a phosphomimeticS-236-D Mena mutant complements the fibroblast phenotype. Therefore, it seemslikely that phosphorylation is not essential for Ena/VASP subcellular targeting butis probably required to regulate Ena/VASP function in lamellipodia. This raises theinteresting possibility that Ena/VASP activity might be regulated by the functionof AKAPs, proteins that anchor and regulate PKA function in discrete subcellularlocations (Diviani & Scott 2001). Taken together, these experiments suggest thatPKA phosphorylation of vertebrate Ena/VASP proteins, particularly at the firstsite, likely elevates their activity.

Interestingly, Drosophila Ena lacks a clear equivalent of VASP Ser-157(Figure 3) (Gertler et al. 1996). When expressed in Ena/VASP-deficient cells, Enalocalizes properly but fails to complement the hypermotility phenotype, possiblybecause it cannot be phosphorylated by PKA (Loureiro et al. 2002). The isolatedEVH2 domains of both Mena and Ena can localize to lamellipodia, though in abroader pattern than the intact protein. The EVH2 domain alone can complementthe hypermotile phenotype of Ena/VASP-deficient cells. Together, these results in-dicate that the EVH2 domain contains the functional elements required to regulate

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.


556 KRAUSE ET AL.

lamellipodial dynamics and suggest that the rest of the protein (including Ser-236of Mena) likely plays a regulatory role. The fact that the Ena EVH2 functions inthis assay suggests that it can perform the molecular functions required to regulatelamellipodial dynamics, but that the intact molecule is inactive, perhaps becausethe Ser-157/-236 equivalent is absent. If correct, this would suggest that PKA reg-ulation of Ena/VASP proteins evolved in the vertebrate system after the divergencewith invertebrates. Interestingly, VASP can largely rescue theDrosophila Enamu-tant phenotype, again suggesting that the basic molecular functions of the familyhave been conserved across evolution but that the vertebrate proteins may requireadditional regulation by PKA/PKG.

What is the effect of serine/threonine phosphorylation on Ena/VASP function?Because phosphorylation of both Ser-157 in VASP and Ser-236 in Mena appearsto correlate with their activity in various processes, it seems likely that this phos-phorylation relieves either an intra- or intermolecular inhibitory interaction. Morework is required to determine the mechanism by which phosphorylation activatesEna/VASP function.

It is known that binding of proteins to the EVH1 domain occurs independently ofphosphorylation (Harbeck et al. 2000). In contrast, binding of certain SH3-domain-containing proteins to the proline-rich central region of Ena/VASP proteins isregulated by serine/threonine phosphorylation. Serine/threonine phosphorylationof EVL by PKA disrupts the binding of EVL to the SH3 domain of c-Abl and nSrc.However, PKA phosphorylation had no effect on the binding of the SH3 domainof Lyn and the WW domain of FE-65 to EVL, and of Profilin to VASP (Harbecket al. 2000, Lambrechts et al. 2000). This is consistent with a recent report thatVASP forms a complex with c-Abl only when VASP is not phosphorylated (Howeet al. 2002).

In the EVH2 domain, the binding to G-actin is greatly reduced by phospho-rylation of the second serine site in VASP (Walders-Harbeck et al. 2002). Thisphosphorylation site is conserved in Mena but absent in EVL, which points todifferences among the family members in regard to their regulation (Gertler et al.1996). The effect of phosphorylation by PKA on F-actin-binding and -bundlingactivity is unclear at present because two studies that used different salt concen-tration in their buffers came to opposite conclusions (Harbeck et al. 2000, Laurentet al. 1999).

Tyrosine Phosphorylation

Drosophila Enawas originally identified as a suppressor of lethality and CNS de-fects associated withD-Abl mutant phenotypes (Gertler et al. 1990). Furthermore,Ena is phosphorylated by D-Abl on multiple tyrosine residues in vitro and whenco-expressed in cells (Comer et al. 1998, Gertler et al. 1995). Ena phosphorylationis decreased inAbl mutant flies, indicating that Abl phosphorylates Ena during de-velopment. However, this also suggests that Ena may be phosphorylated by othertyrosine kinase(s) (Gertler et al. 1995).

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.



What is regulated by tyrosine phosphorylation of Ena? The phosphorylatedtyrosine residues are clustered in the proline-rich region (Figure 3). Consistently,it has been shown that the binding of the SH3 domain of Abl and Src to Ena isabrogated when it is phosphorylated by D-Abl in vitro (Comer et al. 1998). How-ever, at present it is unclear whether tyrosine phosphorylation of Ena is importantfor its function. Expression of a non-phosphorylatable mutant of the six majorsites of Ena inEna mutant flies partially rescued viability (Comer et al. 1998).The same mutant was still capable of rescuing the bypass phenotype in the ISNbperipheral nerve seen in theDrosophila Enamutant (D. Van Vactor, personal com-munication). Furthermore, human VASP, which is not phosphorylated on tyrosineresidues, was able to rescue the embryonic lethality associated with loss ofEnafunction in Drosophila(Ahern-Djamali et al. 1998). The basis forEna suppres-sion ofD-Abl phenotypes remains unclear. It is interesting to note that Abl-familyproteins can bind F- and G-actin and that D-Abl also contains EVH1-binding sites,suggesting that mechanisms other than D-Abl phosphorylation of Ena could beresponsible for the observed genetic interactions.

One difference between vertebrate and invertebrate Ena/VASP proteins is thatthe tyrosine phosphorylation sites in Ena are not conserved in the mammalianEna/VASP proteins (Figure 3). The neuronal-specific Mena-140 isoform has beenshown to be tyrosine phosphorylated in vivo (Gertler et al. 1996). Recently, it wasdemonstrated that Mena-75 and Mena-80 can also be phosphorylated by c-Ablat Tyr-296 when co-expressed with Abi-1 and c-Abl in cells (Tani et al. 2003).Whether this phosphorylation occurs with endogenous levels of c-Abl, Abi-1, andMena is unknown. Mena Tyr-296 is not conserved in VASP or EVL, raising thepossibility that only Mena, but not EVL or VASP, is regulated by c-Abl. Tyr-296 inMena is localized close to the proline-rich central region, as it is in Ena, suggestingthat this phosphorylation could regulate binding of ligands to Mena.

FUTURE DIRECTIONS

Although a great deal has been learned about Ena/VASP proteins since they werediscovered, a number of unanswered questions remain. One important question iswhether the mammalian Ena/VASP proteins are truly functionally interchangeableor whether they have evolved some unique roles. In fibroblast motility andListeriamovement, they function equivalently. However, some cell types or subcellularprocesses may utilize the family members in unique ways. For example, it wasreported recently that cardiac fibroblasts derived from VASP knockout mice, whichexpress normal levels of Mena and EVL, showed increased spreading, prolongedRac and PAK activity, and cell migration defects (Garcia Arguinzonis et al. 2002).The analysis of compound knockouts in all three murine genes should unravelunique functions for the different Ena/VASP proteins in more detail.

The biophysical details of the anti-capping and anti-branching mechanisms re-main unclear. Specifically, how Ena/VASP proteins interact with the barbed endsof actin filaments and yet allow filament elongation or how Ena/VASP proteins

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.


558 KRAUSE ET AL.

inhibit Arp2/3-mediated branching will require detailed structural and kinetic anal-ysis. Furthermore, the relationships among the anti-capping, anti-branching, andProfilin recruitment functions are also unknown. Biochemical and biophysical ex-periments, possibly involving direct observation techniques such as total internalreflection flourescence microscopy, will be needed to understand these questions.

Another mystery concerning Ena/VASP proteins is how they are regulatedin vivo. All evidence points to PKA/PKG phosphorylation as being a criticalregulatory event for mammalian Ena/VASP proteins. It is likely that PKA/PKGphosphorylation regulates Ena/VASP interaction with actin filaments and otherbinding partners. One challenge will be to understand how Ena/VASP proteinsare spatially and temporally phosphorylated within the cell during migration andadhesion. Furthermore, it will be of great interest to understand how tyrosine phos-phorylation by c-Abl and protein-protein interactions, such as those involving SH3-Ena/VASP, may regulate Ena/VASP function within cells. Proteomic analysis ofthe complexes containing Ena/VASP proteins from tissues and cells under differ-ent activation conditions will help to unravel the regulation of Ena/VASP proteins.These experiments and others should clarify signal transduction cascades upstreamof Ena/VASP proteins and ultimately lead to a greater understanding of how cellsregulate the actin cytoskeleton and cell motility.

ACKNOWLEDGMENTS

We thank Geraldine Strasser for the dorsal root ganglion growth cone imagesappearing in Figure 2.

The Annual Review of Cell and Developmental Biologyis online athttp://cellbio.annualreviews.org

LITERATURE CITED

Ahern-Djamali SM, Bachmann C, Hua P,Reddy SK, Kastenmeier AS, et al. 1999.Identification of profilin and src homology3 domains as binding partners forDroso-phila enabled. Proc. Natl. Acad. Sci. USA96:4977–82

Ahern-Djamali SM, Comer AR, Bachmann C,Kastenmeier AS, Reddy SK, et al. 1998.Mutations in Drosophila enabledand res-cue by human vasodilator-stimulated phos-phoprotein (VASP) indicate important func-tional roles for Ena/VASP homology domain1 (EVH1) and EVH2 domains.Mol. Biol.Cell 9:2157–71

Anderson SI, Behrendt B, Machesky LM, In-sall RH, Nash GB. 2003. Linked regulation

of motility and integrin function in activatedmigrating neutrophils revealed by interfer-ence in remodelling of the cytoskeleton.CellMotil. Cytoskelet.54:135–46

Aszodi A, Pfeifer A, Ahmad M, Glauner M,Zhou XH, et al. 1999. The vasodilator-stimulated phosphoprotein (VASP) isinvolved in cGMP- and cAMP-mediatedinhibition of agonist-induced platelet aggre-gation, but is dispensable for smooth musclefunction.EMBO J.18:37–48

Bachmann C, Fischer L, Walter U, ReinhardM. 1999. The EVH2 domain of the vasodila-tor-stimulated phosphoprotein mediates tet-ramerization, F-actin binding, and actin bun-dle formation.J. Biol. Chem.274:23549–57

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.



Ball LJ, Kuhne R, Hoffmann B, Hafner A,Schmieder P, et al. 2000. Dual epitope recog-nition by the VASP EVH1 domain modu-lates polyproline ligand specificity and bind-ing affinity. EMBO J.19:4903–14

Barzik M, Carl UD, Schubert WD, Frank R,Wehland J, Heinz DW. 2001. The N-terminaldomain of Homer/Vesl is a new class II EVH1domain.J. Mol. Biol.309:155–69

Bashaw GJ, Kidd T, Murray D, Pawson T,Goodman CS. 2000. Repulsive axon guid-ance: Abelson and Enabled play opposingroles downstream of the roundabout recep-tor. Cell 101:703–15

Bear JE, Loureiro JJ, Libova I, Fassler R,Wehland J, Gertler FB. 2000. Negative reg-ulation of fibroblast motility by Ena/VASPproteins.Cell 101:717–28

Bear JE, Svitkina TM, Krause M, Schafer DA,Loureiro JJ, et al. 2002. Antagonism betweenEna/VASP proteins and actin filament cap-ping regulates fibroblast motility.Cell 109:509–21

Beneken J, Tu JC, Xiao B, Nuriya M, Yuan JP,et al. 2000. Structure of the Homer EVH1domain-peptide complex reveals a new twistin polyproline recognition.Neuron26:143–54

Boujemaa-Paterski R, Gouin E, Hansen G,Samarin S, Le Clainche C, et al. 2001.Liste-ria protein acta mimics wasp family proteins:It activates filament barbed end branchingby arp2/3 complex.Biochemistry40:11390–404

Brakeman PR, Lanahan AA, O’Brien R, RocheK, Barnes CA, et al. 1997. Homer: a pro-tein that selectively binds metabotropic glu-tamate receptors.Nature386:284–88

Brindle NP, Holt MR, Davies JE, Price CJ, Crit-chley DR. 1996. The focal-adhesion vaso-dilator-stimulated phosphoprotein (VASP)binds to the proline-rich domain in vinculin.Biochem. J.318:753–57

Butt E, Abel K, Krieger M, Palm D, HoppeV, et al. 1994. cAMP- and cGMP-dependentprotein kinase phosphorylation sites of thefocal adhesion vasodilator-stimulated phos-phoprotein (VASP) in vitro and in intact hu-

man platelets.J. Biol. Chem.269:14509–17

Callebaut I. 2002. An EVH1/WH1 domain as akey actor in TGFbeta signalling.FEBS Lett..519:178–80

Callebaut I, Cossart P, Dehoux P. 1998. EVH1/WH1 domains of VASP and WASP pro-teins belong to a large family includingRan-binding domains of the RanBP1 family.FEBS Lett.441:181–85

Carl UD, Pollmann M, Orr E, Gertler FB,Chakraborty T, Wehland J. 1999. Aromaticand basic residues within the EVH1 domainof VASP specify its interaction with proline-rich ligands.Curr. Biol. 9:715–18

Chakraborty T, Ebel F, Domann E, Niebuhr K,Gerstel B, et al. 1995. A focal adhesion factordirectly linking intracellularly motileListe-ria monocytogenesandListeria ivanovii tothe actin-based cytoskeleton of mammaliancells.EMBO J.14:1314–21

Colavita A, Culotti JG. 1998. Suppressors ofectopic UNC-5 growth cone steering iden-tify eight genes involved in axon guidance inCaenorhabditis elegans. Dev. Biol.194:72–85

Comer AR, Ahern-Djamali SM, Juang JL, Jack-son PD, Hoffmann FM. 1998. Phosphoryla-tion of Enabled by theDrosophilaAbelsontyrosine kinase regulates the in vivo functionand protein-protein interactions of Enabled.Mol. Cell. Biol.18:152–60

Comerford KM, Lawrence DW, SynnestvedtK, Levi BP, Colgan SP. 2002. Roleof vasodilator-stimulated phosphoprotein inPKA-induced changes in endothelial junc-tional permeability.FASEB J.16:583–85

Coppolino MG, Krause M, Hagendorff P,Monner DA, Trimble W, et al. 2001. Evi-dence for a molecular complex consisting ofFyb/SLAP, SLP-76, Nck, VASP and WASPthat links the actin cytoskeleton to Fcgammareceptor signalling during phagocytosis.J.Cell Sci.114:4307–18

Diviani D, Scott JD. 2001. AKAP signalingcomplexes at the cytoskeleton.J. Cell Sci.114:1431–7

Domann E, Wehland J, Rohde M, Pistor S,

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.


560 KRAUSE ET AL.

Hartl M, et al. 1992. A novel bacterial vir-ulence gene inListeria monocytogenesre-quired for host cell microfilament interactionwith homology to the proline-rich region ofvinculin. EMBO J.11:1981–90

Drees B, Friederich E, Fradelizi J, Louvard D,Beckerle MC, Golsteyn RM. 2000. Charac-terization of the interaction between zyxinand members of the Ena/vasodilator-stimu-lated phosphoprotein family of proteins.J.Biol. Chem.275:22503–11

Ehrlich JS, Hansen MD, Nelson WJ. 2002.Spatio-temporal regulation of Rac1 localiza-tion and lamellipodia dynamics during ep-ithelial cell-cell adhesion.Dev. Cell3:259–70

Ermekova KS, Zambrano N, Linn H, MinopoliG, Gertler F, et al. 1997. The WW domain ofneural protein FE65 interacts with proline-rich motifs in Mena, the mammalian ho-molog ofDrosophilaEnabled.J. Biol. Chem.272:32869–77

Fedorov AA, Fedorov E, Gertler F, Almo SC.1999. Structure of EVH1, a novel proline-rich ligand-binding module involved in cy-toskeletal dynamics and neural function.Nat.Struct. Biol.6:661–65

Fradelizi J, Noireaux V, Plastino J, Menichi B,Louvard D, et al. 2001. ActA and humanzyxin harbour Arp2/3-independent actin-polymerization activity.Nat. Cell Biol. 3:699–707

Frischknecht F, Moreau V, Rottger S, Gon-floni S, Reckmann I, et al. 1999. Actin-basedmotility of vacciniavirus mimics receptor ty-rosine kinase signalling.Nature401:926–29

Garcia Arguinzonis MI, Galler AB, Walter U,Reinhard M, Simm A. 2002. Increased sprea-ding, Rac/p21-activated kinase (PAK) activ-ity, and compromised cell motility in cellsdeficient in vasodilator-stimulated phospho-protein (VASP).J. Biol. Chem.277:45604–10

Geese M, Loureiro JJ, Bear JE, Wehland J,Gertler FB, Sechi AS. 2002. Contributionof Ena/VASP proteins to intracellular motil-ity of listeria requires phosphorylation andproline-rich core but not F-actin binding or

multimerization.Mol. Biol. Cell 13:2383–96

Geese M, Schluter K, Rothkegel M, JockuschBM, Wehland J, Sechi AS. 2000. Accumu-lation of profilin II at the surface ofListeriais concomitant with the onset of motility andcorrelates with bacterial speed.J. Cell Sci.113:1415–26

Gertler FB, Comer AR, Juang JL, Ahern SM,Clark MJ, et al. 1995. Enabled, a dosage-sensitive suppressor of mutations in theDrosophila Abl tyrosine kinase, encodesan Abl substrate with SH3 domain-bindingproperties.Genes Dev.9:521–33

Gertler FB, Doctor JS, Hoffmann FM. 1990.Genetic suppression of mutations in theDrosophila Abl proto-oncogene homolog.Science248:857–60

Gertler FB, Niebuhr K, Reinhard M, WehlandJ, Soriano P. 1996. Mena, a relative of VASPand Drosophila Enabled, is implicated inthe control of microfilament dynamics.Cell87:227–39

Gitai Z, Yu TW, Lundquist EA, Tessier-LavigneM, Bargmann CI. 2003. The netrin receptorUNC-40/DCC stimulates axon attraction andoutgrowth through Enabled and, in parallel,Rac and UNC-115/AbLIM.Neuron37:53–65

Goh KL, Cai L, Cepko CL, Gertler FB. 2002.Ena/VASP proteins regulate cortical neu-ronal positioning.Curr. Biol. 12:565–69

Grenklo S, Geese M, Lindberg U, Wehland J,Karlsson R, Sechi AS. 2003. A crucial rolefor profilin-actin in the intracellular motilityof Listeria monocytogenes. EMBO Rep.4:1–7

Grevengoed EE, Loureiro JJ, Jesse TL, PeiferM. 2001. Abelson kinase regulates epithelialmorphogenesis inDrosophila. J. Cell Biol.155:1185–98

Haffner C, Jarchau T, Reinhard M, Hoppe J,Lohmann SM, Walter U. 1995. Molecularcloning, structural analysis and functionalexpression of the proline-rich focal adhesionand microfilament-associated protein VASP.EMBO J.14:19–27

Halbrugge M, Friedrich C, Eigenthaler M,

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.



Schanzenbacher P, Walter U. 1990. Stoichio-metric and reversible phosphorylation of a46-kDa protein in human platelets in re-sponse to cGMP- and cAMP-elevating va-sodilators.J. Biol. Chem.265:3088–93

Halbrugge M, Walter U. 1989. Purification ofa vasodilator-regulated phosphoprotein fromhuman platelets.Eur. J. Biochem.185:41–50

Han YH, Chung CY, Wessels D, Stephens S,Titus MA, et al. 2002. Requirement of avasodilator-stimulated phosphoprotein fam-ily member for cell adhesion, the forma-tion of filopodia, and chemotaxis in dic-tyostelium.J. Biol. Chem.277:49877–87

Harbeck B, Huttelmaier S, Schluter K,Jockusch BM, Illenberger S. 2000. Phospho-rylation of the vasodilator-stimulated phos-phoprotein regulates its interaction withactin.J. Biol. Chem.275:30817–25

Hauser W, Knobeloch KP, Eigenthaler M,Gambaryan S, Krenn V, et al. 1999. Mega-karyocyte hyperplasia and enhanced agonist-induced platelet activation in vasodilator-stimulated phosphoprotein knockout mice.Proc. Natl. Acad. Sci. USA96:8120–25

Horstrup K, Jablonka B, Honig-Liedl P, Just M,Kochsiek K, Walter U. 1994. Phosphoryla-tion of focal adhesion vasodilator-stimulatedphosphoprotein at Ser157 in intact humanplatelets correlates with fibrinogen receptorinhibition. Eur. J. Biochem.225:21–27

Howe AK, Hogan BP, Juliano RL. 2002. Reg-ulation of vasodilator-stimulated phospho-protein phosphorylation and interaction withAbl by protein kinase A and cell adhesion.J.Biol. Chem.277:38121–26

Huttelmaier S, Harbeck B, Steffens O, Messer-schmidt T, Illenberger S, Jockusch BM.1999a. Characterization of the actin bind-ing properties of the vasodilator-stimulatedphosphoprotein VASP.FEBS Lett.451:68–74

Jonckheere V, Lambrechts A, VandekerckhoveJ, Ampe C. 1999. Dimerization of profilin IIupon binding the (GP5)3 peptide from VASPovercomes the inhibition of actin nucleationby profilin II and thymosin beta4.FEBS Lett.447:257–63

Kang F, Laine RO, Bubb MR, South-wick FS, Purich DL. 1997. Profilin inter-acts with the Gly-Pro-Pro-Pro-Pro-Pro se-quence of vasodilator-stimulated phosphory-lation (VASP): implications for actin-basedListeria motility. Biochemistry 36:8384–92

Kato A, Ozawa F, Saitoh Y, Hirai K, InokuchiK. 1997. vesl, a gene encoding VASP/Enafamily related protein, is upregulated duringseizure, long-term potentiation and synapto-genesis.FEBS Lett.412:183–9

Kocks C, Gouin E, Tabouref M, Berche P,Ohayon H, Cossart P. 1992.L. monocyto-genes-induced actin assembly requires theactA gene product, a surface protein.Cell68:521–31

Krause M, Sechi AS, Konradt M, MonnerD, Gertler FB, Wehland J. 2000. Fyn-binding protein (Fyb)/SLP-76-associatedprotein (SLAP), Ena/vasodilator-stimulatedphosphoprotein (VASP) proteins and theArp2/3 complex link T cell receptor (TCR)signaling to the actin cytoskeleton.J. CellBiol. 149:181–94

Krugmann S, Jordens I, Gevaert K, DriessensM, Vandekerckhove J, Hall A. 2001. Cdc42induces filopodia by promoting the forma-tion of an IRSp53:Mena complex.Curr. Biol.11:1645–55

Lambrechts A, Kwiatkowski AV, Lanier LM,Bear JE, Vandekerckhove J, et al. 2000.cAMP-dependent protein kinase phosphory-lation of EVL, a Mena/VASP relative, regu-lates its interaction with actin and SH3 do-mains.J. Biol. Chem.275:36143–51

Lanier LM, Gates MA, Witke W, Menzies AS,Wehman AM, et al. 1999. Mena is requiredfor neurulation and commissure formation.Neuron22:313–25

Lasa I, Gouin E, Goethals M, VancompernolleK, David V, et al. 1997. Identification of tworegions in the N-terminal domain of ActAinvolved in the actin comet tail formation byListeria monocytogenes. EMBO J.16:1531–40

Laurent V, Loisel TP, Harbeck B, Wehman A,Grobe L, et al. 1999. Role of proteins of

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.


562 KRAUSE ET AL.

the Ena/VASP family in actin-based motil-ity of Listeria monocytogenes. J. Cell Biol.144:1245–58

Lawrence DW, Comerford KM, Colgan SP.2002. Role of VASP in reestablishmentof epithelial tight junction assembly afterCa2+ switch. Am. J. Physiol. Cell Physiol.282:C1235–45

Liu K, Li L, Nisson PE, Gruber C, JesseeJ, Cohen SN. 1999. Reversible tumorigen-esis induced by deficiency of vasodilator-stimulated phosphoprotein.Mol. Cell. Biol.19:3696–703

Loisel TP, Boujemaa R, Pantaloni D, Car-lier MF. 1999. Reconstitution of actin-basedmotility of Listeria andShigellausing pureproteins.Nature401:613–16

Loureiro JJ, Rubinson DA, Bear JE, Baltus GA,Kwiatkowski AV, Gertler FB. 2002. Criticalroles of phosphorylation and actin bindingmotifs, but not the central proline-rich region,for Ena/vasodilator-stimulated phosphopro-tein (VASP) function during cell migration.Mol. Biol. Cell13:2533–46

May RC, Caron E, Hall A, Machesky LM.2000. Involvement of the Arp2/3 complexin phagocytosis mediated by FcgammaR orCR3.Nat. Cell Biol.2:246–48

Mykkanen OM, Gronholm M, Ronty M, Lalo-wski M, Salmikangas P, et al. 2001. Charac-terization of human palladin, a microfila-ment-associated protein.Mol. Biol. Cell 12:3060–73

Niebuhr K, Ebel F, Frank R, Reinhard M, Do-mann E, et al. 1997. A novel proline-rich mo-tif present in ActA ofListeria monocytogenesand cytoskeletal proteins is the ligand for theEVH1 domain, a protein module present inthe Ena/VASP family.EMBO J. 16:5433–44

Parast MM, Otey CA. 2000. Characterizationof palladin, a novel protein localized tostress fibers and cell adhesions.J. Cell Biol.150:643–56

Penninger JM, Crabtree GR. 1999. The actincytoskeleton and lymphocyte activation.Cell96:9–12

Petit MM, Fradelizi J, Golsteyn RM, Ayoubi

TA, Menichi B, et al. 2000. LPP, an actin cy-toskeleton protein related to zyxin, harborsa nuclear export signal and transcriptionalactivation capacity.Mol. Biol. Cell 11:117–29

Pistor S, Chakraborty T, Niebuhr K, Domann E,Wehland J. 1994. The ActA protein ofListe-ria monocytogenesacts as a nucleator induc-ing reorganization of the actin cytoskeleton.EMBO J.13:758–63

Pistor S, Chakraborty T, Walter U, WehlandJ. 1995. The bacterial actin nucleator pro-tein ActA of Listeria monocytogenescon-tains multiple binding sites for host micro-filament proteins.Curr. Biol. 5:517–25

Pistor S, Grobe L, Sechi AS, Domann E, Ger-stel B, et al. 2000. Mutations of arginineresidues within the 146-KKRRK-150 mo-tif of the ActA protein ofListeria monocy-togenesabolish intracellular motility by in-terfering with the recruitment of the Arp2/3complex.J. Cell Sci.113:3277–87

Prehoda KE, Lee DJ, Lim WA. 1999. Structureof the enabled/VASP homology 1 domain-peptide complex: a key component in the spa-tial control of actin assembly.Cell97:471–80

Reinhard M, Giehl K, Abel K, Haffner C, Jar-chau T, et al. 1995a. The proline-rich focaladhesion and microfilament protein VASP isa ligand for profilins.EMBO J.14:1583–89

Reinhard M, Halbrugge M, Scheer U, Wie-gand C, Jockusch BM, Walter U. 1992. The46/50 kDa phosphoprotein VASP purifiedfrom human platelets is a novel protein asso-ciated with actin filaments and focal contacts.EMBO J.11:2063–70

Reinhard M, Jouvenal K, Tripier D, WalterU. 1995b. Identification, purification, andcharacterization of a zyxin-related proteinthat binds the focal adhesion and microfil-ament protein VASP (vasodilator-stimulatedphosphoprotein).Proc. Natl. Acad. Sci. USA92:7956–60

Reinhard M, Rudiger M, Jockusch BM, WalterU. 1996. VASP interaction with vinculin: arecurring theme of interactions with proline-rich motifs.FEBS Lett.399:103–7

Rottner K, Behrendt B, Small JV, Wehland J.

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.



1999. VASP dynamics during lamellipodiaprotrusion.Nat. Cell Biol.1:321–22

Skoble J, Auerbuch V, Goley ED, Welch MD,Portnoy DA. 2001. Pivotal role of VASP inArp2/3 complex-mediated actin nucleation,actin branch-formation, andListeria mono-cytogenesmotility. J. Cell Biol.155:89–100

Skoble J, Portnoy DA, Welch MD. 2000.Three regions within ActA promote Arp2/3complex-mediated actin nucleation andLis-teria monocytogenesmotility. J. Cell Biol.150:527–38

Smith GA, Theriot JA, Portnoy DA. 1996.The tandem repeat domain in theListeriamonocytogenesActA protein controls therate of actin-based motility, the percentageof moving bacteria, and the localization ofvasodilator-stimulated phosphoprotein andprofilin. J. Cell Biol.135:647–60

Song HJ, Poo MM. 1999. Signal transduc-tion underlying growth cone guidance bydiffusible factors.Curr. Opin. Neurobiol.9:355–63

Svitkina TM, Bulanova EA, Chaga OY, Vign-jevic DM, Kojima S, et al. 2003. Mechanismof filopodia initiation by reorganization of adendritic network.J. Cell Biol.160:409–21

Symons M, Derry JM, Karlak B, Jiang S,Lemahieu V, et al. 1996. Wiskott-Aldrichsyndrome protein, a novel effector for theGTPase CDC42Hs, is implicated in actinpolymerization.Cell 84:723–34

Tani K, Sato S, Sukezane T, Kojima H, HiroseH, et al. 2003. Abl interactor 1 promotes tyro-sine 296 phosphorylation of mammalian En-abled (Mena) by c-Abl kinase.J. Biol. Chem.278(24):21685–92

Theriot JA, Rosenblatt J, Portnoy DA, Gold-schmidt-Clermont PJ, Mitchison TJ. 1994.Involvement of profilin in the actin-basedmotility of L. monocytogenesin cells andcell-free extracts.Cell 76:505–17

Tu JC, Xiao B, Yuan JP, Lanahan AA, LeoffertK, et al. 1998. Homer binds a novel proline-rich motif and links group 1 metabotropicglutamate receptors with IP3 receptors.Neu-ron 21:717–26

Tu Y, Wu S, Shi X, Chen K, Wu C. 2003.

Migfilin and mig-2 link focal adhesions to fil-amin and the actin cytoskeleton and functionin cell shape modulation.Cell 113:37–47

Vaezi A, Bauer C, Vasioukhin V, Fuchs E. 2002.Actin cable dynamics and Rho/Rock orches-trate a polarized cytoskeletal architecture inthe early steps of assembling a stratified ep-ithelium.Dev. Cell3:367–81

Van Kirk LS, Hayes SF, Heinzen RA. 2000. Ul-trastructure ofRickettsia rickettsiiactin tailsand localization of cytoskeletal proteins.In-fect. Immun.68:4706–13

Vasioukhin V, Bauer C, Yin M, Fuchs E. 2000.Directed actin polymerization is the drivingforce for epithelial cell-cell adhesion.Cell100:209–19

Vasioukhin V, Fuchs E. 2001. Actin dynam-ics and cell-cell adhesion in epithelia.Curr.Opin. Cell Biol.13:76–84

Vignjevic D, Yarar D, Welch MD, Peloquin J,Svitkina T, Borisy GG. 2003. Formation offilopodia-like bundles in vitro from a den-dritic network.J. Cell Biol.160:951–62

Volkman BF, Prehoda KE, Scott JA, PetersonFC, Lim WA. 2002. Structure of the N-WASPEVH1 domain-WIP complex: insight intothe molecular basis of Wiskott-Aldrich Syn-drome.Cell 111:565–76

Wakioka T, Sasaki A, Kato R, Shouda T, Mat-sumoto A, et al. 2001. Spred is a Sprouty-related suppressor of Ras signalling.Nature412:647–51

Walders-Harbeck B, Khaitlina SY, HinssenH, Jockusch BM, Illenberger S. 2002. Thevasodilator-stimulated phosphoprotein pro-motes actin polymerisation through directbinding to monomeric actin.FEBS Lett.529:275–80

Waldmann R, Nieberding M, Walter U. 1987.Vasodilator-stimulated protein phosphoryla-tion in platelets is mediated by cAMP- andcGMP-dependent protein kinases.Eur. J.Biochem.167:441–48

Walter U, Eigenthaler M, Geiger J, Reinhard M.1993. Role of cyclic nucleotide-dependentprotein kinases and their common substrateVASP in the regulation of human platelets.Adv. Exp. Med. Biol.344:237–49

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.


564 KRAUSE ET AL.

Welch MD, Iwamatsu A, Mitchison TJ. 1997.Actin polymerization is induced by Arp2/3protein complex at the surface ofListeriamonocytogenes. Nature385:265–69

Welch MD, Rosenblatt J, Skoble J, Portnoy DA,Mitchison TJ. 1998. Interaction of humanArp2/3 complex and theListeria monocyto-genesActA protein in actin filament nucle-ation.Science281:105–8

Wills Z, Bateman J, Korey CA, Comer A, VanVactor D. 1999. The tyrosine kinase Abland its substrate enabled collaborate with thereceptor phosphatase Dlar to control motoraxon guidance.Neuron22:301–12

Yu TW, Hao JC, Lim W, Tessier-Lavigne M,Bargmann CI. 2002. Shared receptors in axonguidance: SAX-3/Robo signals via UNC-34/Enabled and a Netrin-independent UNC-40/DCC function.Nat. Neurosci.5:1147–54

Zettl M, Way M. 2002. The WH1 and EVH1domains of WASP and Ena/VASP familymembers bind distinct sequence motifs.Curr.Biol. 12:1617–22

Zimmermann J, Labudde D, Jarchau T, WalterU, Oschkinat H, Ball LJ. 2002. Relaxation,equilibrium oligomerization, and molecularsymmetry of the VASP (336–380) EVH2tetramer.Biochemistry41:11143–51

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.

Ena/VASP FUNCTION C-1

Figure 1 Expression of Mena in a cultured fibroblast. A fibroblast expressing EGFP-Mena was fixed and labeled with an antibody to Vinculin and with phalloidin (to label actinfilaments). Note the enrichment of Mena at the periphery of the lamellipodia, beyond theF-actin labeling, and in adhesions, which label with anti-Vinculin antibody.

21-HI-RESOLUTION-VERSION 9/15/2003 8:05 PM Page 1

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.

C-2 KRAUSE ET AL.

Figure 2 Expression of Mena in a cultured dorsal root ganglion (DRG) growth cone anda hippocampal growth cone. Mena strongly labels the distal ends of F-actin bundles (phal-lodin stain) in both the DRG and hippocampal growth cones. Anti-Mena labeling alsomarks the ends of actin-rich extensions in the transition region of both growth cones.


Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.

Ena/VASP FUNCTION C-3

Figure 3 Ena/VASP phosphorylation. Cartoons of the six best-characterized members ofthe Ena/VASP family are depicted (not to scale). All share a similar domain structure con-sisting of a central proline-rich region flanked by an amino-terminal EVH1 domain and acarboxy-terminal EVH2 domain. Whereas Ena/VASP proteins are similar in overall struc-ture, in vitro biochemical and in vivo functional data suggest that orthologs from differentspecies have distinct phosphorylation patterns. Although no phosphorylation events havebeen documented for DdVASP or Unc-34, both have a conserved serine just after theEVH1 domain (comparable to Ser-157 of VASP); the conserved serine in Unc-34 is part ofa consensus PKA phosphorylation site (R-R-X-S). Drosophila Enabled does not share thissite, but is tyrosine-phosphorylated in vitro and in vivo (Gertler et al. 1995, Comer et al.1998). Mammalian Ena/VASP proteins all share an amino-terminal PKA/PKG phosphory-lation site (Ser-157 of VASP) that, at least in Mena, affects the regulation of cell motility.A neuronal isoform of Mena (not depicted here) is an in vivo substrate for tyrosine kinaseactivity (Gertler et al. 1996). Recently, more ubiquitous isoforms of Mena have been foundto be tyrosine-phosphorylated in vitro (Tani et al. 2003).


Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.

P1: FRK

September 4, 2003 21:45 Annual Reviews AR197-FM

Annual Review of Cell and Developmental BiologyVolume 19, 2003

CONTENTS

ADULT STEM CELL PLASTICITY: FACT OR ARTIFACT, Martin Raff 1

CYCLIC NUCLEOTIDE-GATED ION CHANNELS, Kimberly Matulef andWilliam N. Zagotta 23

ANTHRAX TOXIN, R. John Collier and John A.T. Young 45

GENES, SIGNALS, AND LINEAGES IN PANCREAS DEVELOPMENT,L. Charles Murtaugh and Douglas A. Melton 71

REGULATION OF MAP KINASE SIGNALING MODULES BY SCAFFOLD PROTEINS

IN MAMMALS, Deborah Morrison and Roger J. Davis 91

FLOWER DEVELOPMENT: INITIATION, DIFFERENTIATION, AND

DIVERSIFICATION, Moriyah Zik and Vivian F. Irish 119

REGULATION OF MEMBRANE PROTEIN TRANSPORT BY UBIQUITIN AND

UBIQUITIN-BINDING PROTEINS, Linda Hicke and Rebecca Dunn 141

POSITIONAL CONTROL OF CELL FATE THROUGH JOINT INTEGRIN/RECEPTOR

PROTEIN KINASE SIGNALING, Filippo G. Giancotti and Guido Tarone 173

CADHERINS AS MODULATORS OF CELLULAR PHENOTYPE,Margaret J. Wheelock and Keith R. Johnson 207

GENOMIC IMPRINTING: INTRICACIES OF EPIGENETIC REGULATION IN

CLUSTERS, Raluca I. Verona, Mellissa R.W. Mann, andMarisa S. Bartolomei 237

THE COP9 SIGNALOSOME, Ning Wei and Xing Wang Deng 261

ACTIN ASSEMBLY AND ENDOCYTOSIS: FROM YEAST TO MAMMALS,Asa E.Y. Engqvist-Goldstein and David G. Drubin 287

TRANSPORT PROTEIN TRAFFICKING IN POLARIZED CELLS, Theodore R. Muthand Michael J. Caplan 333

MODULATION OF NOTCH SIGNALING DURING SOMITOGENESIS,Gerry Weinmaster and Chris Kintner 367

TETRASPANIN PROTEINS MEDIATE CELLULAR PENETRATION, INVASION,AND FUSION EVENTS AND DEFINE A NOVEL TYPE OF MEMBRANE

MICRODOMAIN, Martin E. Hemler 397

INTRAFLAGELLAR TRANSPORT, Jonathan M. Scholey 423

vii

Ann

u. R

ev. C

ell D

ev. B

iol.

2003

.19:

541-

564.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 11

/06/

08. F

or p

erso

nal u

se o

nly.

P1: FRK

September 4, 2003 21:45 Annual Reviews AR197-FM

viii CONTENTS

THE DYNAMIC AND MOTILE PROPERTIES OF INTERMEDIATE FILAMENTS,Brian T. Helfand, Lynne Chang, and Robert D. Goldman 445

PIGMENT CELLS: A MODEL FOR THE STUDY OF ORGANELLE TRANSPORT,Alexandra A. Nascimento, Joseph T. Roland, and Vladimir I. Gelfand 469

SNARE PROTEIN STRUCTURE AND FUNCTION, Daniel Ungar andFrederick M. Hughson 493

STRUCTURE, FUNCTION, AND REGULATION OF BUDDING YEAST

KINETOCHORES, Andrew D. McAinsh, Jessica D. Tytell, and Peter K. Sorger 519

ENA/VASP PROTEINS: REGULATORS OF THE ACTIN CYTOSKELETON AND

CELL MIGRATION, Matthias Krause, Erik W. Dent, James E. Bear,Joseph J. Loureiro, and Frank B. Gertler 541

PROTEOLYSIS IN BACTERIAL REGULATORY CIRCUITS, Susan Gottesman 565

NODAL SIGNALING IN VERTEBRATE DEVELOPMENT, Alexander F. Schier 589

BRANCHING MORPHOGENESIS OF THE DROSOPHILA TRACHEAL SYSTEM,Amin Ghabrial, Stefan Luschnig, Mark M. Metzstein, andMark A. Krasnow 623

QUALITY CONTROL AND PROTEIN FOLDING IN THE SECRETORY PATHWAY,E. Sergio Trombetta and Armando J. Parodi 649

ADHESION-DEPENDENT CELL MECHANOSENSITIVITY,Alexander D. Bershadsky, Nathalie Q. Balaban, and Benjamin Geiger 677

PLASMA MEMBRANE DISRUPTION: REPAIR, PREVENTION, ADAPTATION,Paul L. McNeil and Richard A. Steinhardt 697

INDEXES

Subject Index 733Cumulative Index of Contributing Authors, Volumes 15–19 765Cumulative Index of Chapter Titles, Volumes 15–19 768

ERRATA

An online log of corrections to Annual Review of Cell and DevelopmentalBiology chapters (if any, 1997 to the present) may be found athttp://cellbio.annualreviews.orgA

nnu.

Rev

. Cel

l Dev

. Bio

l. 20

03.1

9:54

1-56

4. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Wis

cons

in -

Mad

ison

on

11/0

6/08

. For

per

sona

l use

onl

y.

ena/vasp proteins: regulators of the actin cytoskeleton ... · proteins of the ena/vasp family have...

Documents