Spatial Control of Actin Filament Review Assembly: Lessons from Listeria

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Cell, Vol. 95, 741–748, December 11, 1998, Copyright 1998 by Cell Press Spatial Control of Actin Filament Review Assembly: Lessons from Listeria

Mary C. Beckerle (Figure 1A). In striated muscle, semicrystalline arrays Department of Biology of actin and myosin are constructed to establish the University of Utah contractile machinery (Figure 1B). Although many cells Salt Lake City, Utah 84103 lack such striking actin filament displays, they nevertheless exhibit a similar capacity to define zones of actin assembly. Processes such as cell migration depend on the ability of cells to fix sites of actin assembly with The actin cytoskeleton of a eukaryotic cell is critical for a high degree of spatial resolution. For example, for many functions, including cell locomotion, cytokinesis, vectorial cell locomotion of a fibroblast to occur, actin intracellular transport, establishment of cell polarity, and assembly must be focused at the leading edge (Figure the elaboration of cell surface projections that increase 1C), a broad membrane protrusion that is comprised of cell surface area and sense the environment. Arrays a meshwork of actin filaments that are generally oriented of actin filaments must be assembled within cells in a with their barbed ends directed toward the cell pe- fashion that is both spatially and temporally controlled riphery. in order to contribute to such a diversity of cellular pro- The situation for actin contrasts with that of its sister cesses. polymer, the microtubule. In animal cells, there is typi- Cells face a number of challenges when it comes to cally a single cellular site called the microtubule organiz- actin. Actin filaments are structurally and functionally ing center that provides the primary focus for microtu- polar polymers that assemble from monomers that come bule assembly. An excess of microtubules is assembled together in a head to tail fashion. The intrinsic polarity with apparent spatial abandon, and only selected micro- of an actin filament can be visualized by decoration of tubules are retained by stabilization, for example by the filament with myosin fragments that coat it to gener- binding to specialized capture sites on mitotic chromo- ate an arrowhead appearance. One end of the filament, somes. For the actin cytoskeleton, assembly appears referred to as the barbed or “plus” end is the preferred to follow site selection. But how are suitable sites for site for monomer addition; the other end, referred to as actin assembly specified? the “minus” end, is the slow end with respect to growth. Actin polymerization occurs in two phases. The rate- limiting step in actin filament formation involves the gen- Listeria as a Model System for Dissecting eration of an assembly-competent actin nucleus such the Actin Assembly Machinery as an actin trimer by de novo assembly or a free actin One particularly useful model organism for deciphering filament end by uncapping or severing. When an actin the mechanism by which spatially controlled actin as- nucleus is present and the concentration of monomer is sembly is accomplished in mammalian cells is the intra- sufficient, spontaneous elongation of a filament occurs. cellular bacterial pathogen, Listeria (Tilney and Portnoy, Within cells, monomeric actin is present at concentra- 1989). Listeria is a Gram-positive food-borne pathogen tions well above what is required for rapid polymeriza- that can cause life-threatening illnesses such as en- tion at physiological ionic strength. So, what is it that cephalitis, particularly in immune-compromised individ- prevents our cells from filling up with tangled masses uals. Listeria has the capacity to invade mammalian of filamentous actin? cells. Once internalized, the bacterium escapes the One way that cells control actin assembly is by se- membrane-bound endosome, replicates, and resides questering monomeric actin subunits through stoichio- happily within the host cytoplasm. There, the bacterium metric binding to protein partners such as thymosin harnesses the host machinery required for actin assem- ␤4 and profilin that control the availability of the actin bly and generates an actin “comet” tail (Figure 1D) that monomers for polymerization. Cells also produce pro- propels the bacterium within the host cell. The intracellu- teins such as CapZ and gelsolin that can cap the fast- lar motion of Listeria is directly coupled to actin assem- growing (barbed) ends of actin filaments and thus con- bly and is independent of myosin function (Sanger et al., trol the elongation of preexisting filaments. Control of 1992; Theriot et al., 1992). It is this actin-based motility of nucleation and regulation of available actin monomer the bacterium that enables it to enter an adjacent cell levels thus provide important mechanisms for regulating via a membrane-bound projection. the amount of actin polymer inside cells, but they are Because the ability of Listeria to be transferred from not alone sufficient to answer the fundamental question one cell to the next, and thus to spread the infection, is of how these actin filaments come to be assembled at absolutely dependent on its ability to build an actin- appropriate locations within the cell. rich comet tail, microbiologists have sought to identify bacterial genes that are critical for this process. Interest- ingly, just a single bacterial factor, the ActA protein, has Diversity of Actin Assemblies been shown to be required for the ability of Listeria to Simple inspection of cellular architecture makes it clear assemble filamentous actin on its surface within the host that cells have exquisite control over where actin fila- cytoplasm (Domann et al., 1992; Kocks et al., 1992). ActA ments will accumulate. In cochlear hair cells, highly or- is asymmetrically distributed on the bacterial surface, dered apical bundles of actin filaments form the struc- concentrated on the end of the pathogen that is associ- tural backbone of the stereocilia that sense vibration ated with the actin comet tail. Although the ActA protein Cell 742

Figure 1. Actin Assemblages (A) Actin-rich projections in the cochlea exhibit defined lengths and positions. (B) Semicrystalline actin arrays in striated muscle. (C) Zones of actin assembly (arrows) detected by incorporation of fluorescently labeled actin (red) at the leading edge of a migrating cell. (D) The actin-rich comet tails (arrows) of Liste- ria monocytogenes within host cytoplasm. Images shown here have been reproduced from the following sources: (A) Tilney et al., 1983, by copyright permission of the Rocke- feller University Press; (B) Smith, 1972, by copyright permission of Academic Press, Inc.; (C) Chan et al., 1998 by copyright permission of the Company of Biologists, Ltd; (D) Alberts et al., 1994, courtesy of Tim Mitchison and Julie Theriot.

is not alone sufficient to stimulate the assembly of bio- actin assembly. Deletions in the proline region of ActA chemically purified actin, a number of lines of evidence slow the rate of bacterial motility, a process that is di- suggest that it drives actin assembly within the context rectly coupled to actin polymerization, thus there is a of eukaryotic cytoplasm. For example, when ActA bear- functional link between the proline-rich region and actin ing a C-terminal CAAX sequence that signals farnesyla- assembly. The mechanism by which the proline-rich do- tion and carboxymethylation is targeted to the inner main augments net actin polymerization on the bacterial leaflet of the plasma membrane of uninfected mamma- surface is not well understood. Several possibilities re- lian cells, it induces the elaboration of actin-rich cell main consistent with experimental observations. For ex- surface projections (Friederich et al., 1995). Likewise, ample, this region could act directly or indirectly to en- targeting of ActA to the surface of mitochondria stimu- hance the nucleation potential of the N-terminal domain lates organelle-associated actin assembly (Pistor et al., of ActA, could stimulate actin filament elongation by 1994), and the expression of ActA by a nonpathogenic promoting filament uncapping or monomer addition, or bacterium is sufficient to promote comet tail formation could somehow enhance local levels of polymerization- (Kocks et al., 1995). Because of the ability of ActA to competent monomer. Reconstitution of ActA-depen- interact so productively with host factors involved in dent actin assembly with purified components will be actin assembly, it has been speculated that ActA is a molecular mimic of a mammalian protein or proteins that play central roles in the process of actin assembly. Thus, understanding precisely how ActA works is likely to provide substantial insight into how cells define sites of actin assembly. Recently, much progress has been made toward de- termining the role of the ActA protein and its relationship to the endogenous cellular machinery involved in actin assembly. The functional domains of the ActA protein have been defined by mutagenesis (Figure 2) (Lasa et al., 1995; Pistor et al., 1995; Smith et al., 1996; Mourrain et al., 1997). The ActA protein, which exists as a dimer, is tethered to the bacterial surface via a C-terminal anchor. As might be predicted for a protein that stimulates the assembly of such glorious actin arrays within cells (Figure 1D), ActA displays several regions that promote Figure 2. Schematic Representation of the Listeria ActA Protein actin assembly. An N-terminal ActA domain is absolutely ActA displays a signal sequence, an N-terminal region that appears required for the ability of the bacterium to elaborate to interact with the Arp2/3 complex to nucleate actin assembly, a actin filament arrays on its surface. As will be discussed central series of proline-rich repeats that bind Ena/VASP proteins, in greater detail below, this region of ActA has recently and a C-terminal membrane anchor. No eukaryotic proteins with been implicated in facilitating actin nucleation, the rate- substantial similarity to the “nucleation” domain of ActA have been identified. In contrast, both vinculin and zyxin display sequences limiting step in actin assembly. A central proline-rich that are strikingly related to the “assembly-accelerating” domain of domain of ActA is also necessary for efficient assembly ActA. Although not depicted here, ActA is a dimer in vivo (Mourrain of the comet tail but is not in itself sufficient to trigger et al., 1997). Review 743

necessary to distinguish among these (and other) possi- it is pleasing that the domain of ActA that was found to bilities. synergize with Arp2/3 to nucleate actin assembly in vitro It is clear that ActA requires the cooperation of several maps to the N terminus of the protein, the same region host factors to stimulate actin polymerization. One inter- of ActA that has been shown to be essential for nucle- esting view is that the ActA protein represents a hybrid ation of actin assembly on the surface of the bacterium protein that exhibits functions found in multiple eukary- within the cytoplasm of a mammalian cell. otic proteins; noncovalent association of these cellular The precise mechanism by which the Arp2/3 complex proteins would reconstitute full “ActA activity” and pro- stimulates actin assembly is not understood. Although vide a cellular machine to regulate actin assembly. Per- the Arp2/3 complex does not act as a preformed nucleus haps ancestors of Listeria acquired genes related to for polymerization on its own, it may have this property those of their early mammalian hosts by nucleic acid in the presence of key accessory factors like Listeria transfer. If that were the case, then one would expect ActA. Indeed, when ActA and Arp2/3 are mixed together to be able to identify relatives of ActA in mammalian with monomeric actin, the nucleation phase of actin cells. Alternatively, it is also possible that Listeria assembly, which is normally slow, is essentially instanta- evolved entirely novel strategies for interacting with the neous (Welch et al., 1998). The ability of purified Arp2/3 host machinery for actin assembly. In either case, identi- to bind to the pointed, slow-growing ends of actin fila- fication of the host components that associate with ActA ments has also led to the suggestion that the Arp2/3 has already provided insight into how cells regulate the complex may facilitate actin assembly by docking and polymerization of actin. stabilizing an actin nucleus such as a dimer or trimer (Mullins et al., 1998a). Arp2/3-dependent capping of actin filaments may also enhance their stability by slowing The Arp2/3 Complex: A Host Factor that Facilitates disassembly from the pointed ends of the polymers. Actin Assembly on the Listeria Surface The ability of the Arp2/3 complex to cap the pointed Reconstitution of actin assembly and Listeria motility in ends of actin filaments may help to explain why Arp2/3 is cytoplasmic extracts has led to the identification of host found throughout the actin comet tail of a motile Listeria, proteins that collaborate with ActA to promote actin rather than just at the site of actin growth near the bacte- assembly in vivo. Using human platelet extracts to sup- rial surface. Many of the actin filaments found in the port ActA-dependent actin assembly on the surface of Listeria comet tail are short relative to the length of the Listeria, Welch and colleagues showed that a complex tail. If short actin filaments nucleated by Arp2/3 in close of proteins called the Arp2/3 complex is both necessary proximity to ActA were subsequently released with an and sufficient for ActA’s ability to provide a focus for Arp2/3 cap, the Arp2/3 complex might persist in the actin assembly (Welch et al., 1997a). The Arp2/3 com- older portions of the tail as ActA-associated actin as- plex is highly conserved from Acanthamoeba to humans sembly progressed near the bacterial surface. (Kelleher et al., 1995; Machesky et al., 1997; Welch et In addition to its involvement in the nucleation of actin al., 1997b) and contains seven protein constituents, in- assembly, the Arp2/3 complex has also been implicated cluding two actin-related proteins, Arp2 and Arp3. As in the organization of actin arrays. In vitro, the Arp2/3 would be expected for a collection of proteins that play complex promotes the assembly of branched actin fila- a central role in actin polymerization, the Arp2/3 com- ment arrays. It has been suggested that the Arp2/3 complex is enriched at sites of actin assembly such as the plex may support the assembly of a new filament off leading edges of motile cells (Kelleher et al., 1995; Ma- the side of a preexisting polymer (Mullins et al., 1998a). chesky et al., 1997; Welch et al., 1997b). Arp2 and Arp3 The morphology of actin filament arrays that are pre- have also been identified in yeast where they have been pared in the presence of Arp2/3 is similar to the filament shown to be localized in actin-rich cortical structures architecture observed at the leading edge of cells (Svit- and are required for actin-dependent functions (McCol- kina et al., 1997), an observation that has suggested a lum et al., 1996; Moreau et al., 1996). role for Arp2/3 proteins in lamellipodial cytoarchitecture. Because of the presence of two actin-related proteins The illustration that Arp2/3 and ActA cooperate to in the Arp2/3 complex, it was postulated that the Arp2/3 stimulate actin nucleation suggests that one way to con- complex might stimulate actin assembly by mimicking trol actin assembly would be to control the access of the structure of an actin nucleus or free filament end these two components to each other. Localization of (Kelleher et al., 1995). However, in recent biochemical the Arp2/3 complex, its cellular activator(s), or both to studies of Acanthamoeba and human Arp2/3, the com- specific cellular sites is likely to be critical for defining plex failed to reduce significantly the rate-limiting step fertile zones for actin assembly in uninfected cells as for actin assembly (Mullins et al., 1998a; Welch et al., well. Identification of the endogenous cellular factors 1998); thus, it does not appear to act as a preformed that cooperate with the Arp2/3 complex to nucleate actin nucleus for actin assembly, at least not on its own. The polymerization will surely provide substantial insight into fact that the Arp2/3 complex is an inefficient nucleator how actin assembly sites are generated within cells. of actin assembly when it is in a highly purified form provides a potentially powerful mechanism for a cell to control its activity by regulating the availability and/or ActA Domains that Accelerate Actin Assembly: localization of essential cofactors. In a Listeria-infected Relationship to Zyxin and Vinculin cell, the ActA protein on the bacterial surface appears Examination of the region of ActA involved in accelera- to function as the cofactor that stimulates the nucleating tion of actin polymerization has also provided insight activity of Arp2/3 (Welch et al., 1998). Although direct into the cellular mechanisms required for assembly of binding of Arp2/3 to ActA has not been demonstrated, actin filaments. The assembly-enhancing activity of Cell 744

ActA has been mapped to a series of short proline- 1998b), though the biological consequences of this asso- rich repeats characterized by the consensus D/EFPPPP. ciation, if any, have not been elucidated. Mutations that affect these proline repeats result in a Two potential counterparts of the proline-rich region reduction in actin polymer on the surface of the bacte- of the ActA protein have been identified in metazoans, rium and a reduced rate of motility of the bacterium including mammals. A proline repeat similar to those within the host cytoplasm (Lasa et al., 1995; Pistor et found in ActA is present in vinculin, a component of al., 1995; Smith et al., 1996). The proline repeats of ActA adhesion plaques that also binds VASP (Brindle et al., have been shown to serve as docking sites for host 1996); however, the ability of vinculin to cooperate with proteins in the Enabled/Vasodilator-stimulated phos- Ena/VASP family members to regulate actin assembly phoprotein (Ena/VASP) family (Smith et al., 1996; Nie- has not yet been probed substantially. In contrast, sev- buhr et al., 1997). To date, Ena/VASP family members eral approaches, including biochemical, immunological, are the only known ligands for these bioactive proline- and functional studies, suggest that one mammalian rich sequences, so it is reasonable to imagine that the protein mimicked by ActA is zyxin (Sadler et al., 1992; enhancement of actin assembly by the proline repeats Beckerle, 1997). As is the case for ActA, zyxin binds involves Ena/VASP activity. In support of this view, at members of the Ena/VASP family (Reinhard et al., 1995b; least one mammalian Ena/VASP family member triggers Gertler et al., 1996; Ahern-Djamali et al., 1998). More- actin assembly when expressed in mammalian cells over, antibodies raised against Listeria ActA recognize (Gertler et al., 1996). However, it should be noted that only zyxin in uninfected mammalian cell lysates, a result no direct evidence that Ena/VASP proteins are essential that identifies zyxin as the mammalian protein that is for optimal assembly of the Listeria comet tail has yet perhaps most similar to ActA (Golsteyn et al., 1997). Not been presented. A demonstration that cells or extracts surprisingly given the antibody cross-reactivity and the lacking Ena/VASP activity show reduced ability to sup- shared ability to bind Ena/VASP family members, zyxin port actin assembly and bacterial motility would confirm displays a collection of four proline-rich sequences that directly a role for these proteins in comet tail formation. are very similar to those found in the ActA protein. Con- Nevertheless, given the reasonable hypothesis that sistent with the possibility that zyxin may play a role in Ena/VASP proteins are central to Listeria comet tail as- actin assembly or organization, zyxin is found at cellular sembly, how might they enhance actin polymerization locations that are enriched in actin filaments, including at the bacterial surface? One potential mechanistic link the leading edge and adhesion plaques, many of which between Ena/VASP proteins and actin polymerization is may be elaborated from structures originally generated derived from the ability of these proteins to multimerize at the leading edge. In general, sites of zyxin concentra- and dock profilin, an actin monomer binding protein tion are typically coincident with regions where the fast- that stimulates the assembly of actin filaments at their growing ends of actin filaments are amassed. barbed ends (Reinhard et al., 1995a; Kang et al., 1997). Although cells carrying loss-of-function mutations in Localization studies have revealed that both Ena/VASP the zyxin gene have not been described, efforts to ex- proteins and profilin are closely associated with the actin assembly-competent end of a Listeria cell (Theriot et plore the function of zyxin in cells have been initiated. al., 1994; Gertler et al., 1996; Smith et al., 1996). It has Targeting of zyxin sequences to the inner leaflet of the been suggested that the ActA-dependent recruitment plasma membrane by incorporation of a CAAX tag in- of Ena/VASP proteins and their partners could serve to duces the assembly of actin-rich cell surface projec- concentrate actin monomers at the bacterial surface, tions, consistent with the view that localization of zyxin close to the region where assembly-competent actin could serve to tether machinery required for actin as- nuclei are found. In theory, ActA could serve as a dock- sembly (Golsteyn et al., 1997). The N-terminal 350 aa ing site for up to four Ena/VASP family members, since it region of zyxin, which contains the proline-rich ActA displays four proline-rich repeats and VASP is a tetramer repeats, is sufficient to induce actin assembly and reor- with numerous potential profilin-binding sites; therefore, ganization when positioned proximal to the plasma if all VASP-binding and profilin-binding sites were fully membrane; however, it has not been directly demon- occupied, a dramatic increase in local profilin–actin lev- strated that the proline-rich repeats themselves are nec- els could be achieved. This view is appealing in its sim- essary for this activity. plicity, but it is not clear how the accumulation of pro- As noted above, productive actin assembly requires filin–actin bound to Ena/VASP family members might the presence of actin nuclei or free filament ends and the contribute to enhancing local levels of soluble profilin– ability to elongate the filaments by addition of monomer. actin pools that would be available to participate in actin Examination of zyxin’s binding partner repertoire may filament elongation. Moreover, the very role of profilin provide some insight into how zyxin could trigger actin in the assembly of the Listeria tail remains controversial, assembly (Figure 3). The proline region of zyxin that has since depletion of profilin has not consistently abolished the capacity to stimulate actin assembly at the inner comet tail assembly (Theriot et al., 1994; Marchand et surface of the plasma membrane has three known bind- al., 1995). Clearly, there are some missing links in our ing partners: ␣-actinin, vav, and Ena/VASP proteins. All understanding of the biological roles of Ena/VASP pro- of these partners have themselves been implicated in teins. Perhaps they act catalytically to enhance the as- actin assembly or organization. ␣-actinin could serve to sembly potential of actin–profilin complexes in their vi- cross-link actin filaments and to tether these cytoskele- cinity or to transfer actin monomers directly to nearby tal elements directly to integrin extracellular matrix re- filament ends. Alternatively, the Ena/VASP–profilin com- ceptors. Vav is a protooncogene product that interacts plex could have an impact on Arp2/3’s nucleation activ- with zyxin via its C-terminal SH3 domain (Hobert et al., ity; Acanthamoeba Arp2/3 binds profilin (Mullins et al., 1996); vav exhibits guanine nucleotide exchange activity Review 745

Figure 3. Similarities between Actin Assembly on the Listeria Surface and at the Cytoplasmic Face of the Plasma Membrane A comparison of molecular models for actin assembly by Listeria within its eukaryotic host (top panel) and at the surface of an uninfected mammalian cell (bottom panel). The Listeria ActA protein appears to stimulate actin assembly by interacting with the Arp2/3 complex and members of the Ena/VASP family (E/V) that recruit profilin (P). In an uninfected mammalian cell, for example, Arp2/3 may be localized and activated at particular regions of the cell where actin assembly is to be favored. One region of the zyxin protein appears to function like ActA to recruit Ena/VASP and ultimately profilin. Zyxin also binds ␣-actinin, which could facilitate the cross-linking of newly assembled actin filaments as well as link zyxin directly to cell surface receptors such as integrins. When active, vav bound to zyxin may locally reduce actin filament capping, thus enhancing filament elongation. Cell adhesion to extracellular matrix can stimulate vectorial cell migration, but how is the adhesive event coupled to directional membrane protrusion? The diagram illustrates a potential mechanism for insuring the spatial coincidence of sites of substratum recognition by integrin receptors and the actin assembly machinery necessary for cell surface extension. Details are described in the text. for Rho family members (Olson et al., 1996; Crespo et (Ahern-Djamali et al., 1998). One straightforward inter- al., 1997) and thus may contribute significantly to zyxin’s pretation of this observation is that zyxin is required to ability to stimulate changes in the actin cytoskeleton. dock Ena at particular subcellular locations in order to Rho family members, including Rho, Rac, and Cdc42, stimulate the local actin assembly necessary for nerve have well-established roles in the control of actin assem- cell migration during development. Indeed, mutant Ena bly and organization, regulating such actin-dependent protein that lacks zyxin binding capacity in vitro fails to processes as adhesion, ruffling, and filopod formation. localize appropriately when expressed in mammalian It is of particular interest that activation of certain Rho cells (Ahern-Djamali et al., 1998). As was described family members can, among other things, enhance the above for Listeria ActA, binding of Ena/VASP proteins production of phosphatidyl inositol (4,5) bisphosphate to zyxin could facilitate the recruitment of profilin–actin (PIP2). Phosphoinositides have been shown to be critical complexes to sites where assembly-competent barbed for membrane-associated actin assembly in Xenopus filament ends are concentrated. Within the mammalian extracts (Ma et al., 1998), and they could contribute to cytoplasm, there may be some mechanism for control- the regulation of actin assembly by inducing the local ling whether Ena/VASP family members recruit profilin, uncapping of actin filaments (Hartwig et al., 1995) or since codistribution of these proteins does not appear by affecting the interaction between profilin and actin to be obligatory at steady state. (Lassing and Lindberg, 1985). A third group of zyxin-binding partners is the Ena/ VASP family. Recent genetic studies suggest that the Coordinating Actin Assembly ability of Ena/VASP family members to associate with with Information Input zyxin is essential to their function. The Ena protein is In a migrating cell, the actin assembly machinery must required for nervous system development in Drosophila be exquisitely sensitive to cues from the environment. and is thought to be important for nerve cell migration Indeed, cell binding to extracellular matrix molecules and fasciculation. Characterization of the genetic defect can stimulate cell locomotion, and disrupted expression present in a lethal ena allele has revealed a point muta- of a major extracellular matrix receptor subunit, ␤1 inte- tion resulting in a single conservative amino acid substi- grin, results in decreased cell migration by embryonic tution that obliterates the protein’s ability to bind zyxin stem cells and neuroblasts (Galileo et al., 1992; Fassler Cell 746

et al., 1995). Although integrins are typically thought to eukaryotic cells regulate actin polymerization. The ActA be required for cell motility because of the need for protein of Listeria appears to act by interacting with the cell adhesion to generate traction forces, other roles for Arp2/3 complex and members of the Ena/VASP family integrins may also exist. For example, it is possible that to facilitate the assembly of actin filaments. Cooperative actin assembly is stimulated by integrin–ligand binding, localization of these proteins in an uninfected eukaryotic thus providing a means to connect recognition of envi- cell would similarly be expected to create an actin as- ronmental cues to vectorial cell surface projection. sembly zone. One eukaryotic protein that appears to be One mechanism by which integrins could participate mimicked by ActA is zyxin. Zyxin acts to dock members in the control of cell locomotion by regulating actin as- of the Ena/VASP family at specific sites within the cyto- sembly is illustrated in Figure 3. As discussed in greater plasm and thus may play a central role in facilitating detail below, zyxin may serve to cluster a collection of their functions. Endogenous cellular proteins that acti- response proteins, including stimulators of actin un- vate, and perhaps localize, the Arp2/3 complex have capping, polymerization, and cross-linking, near the in- not yet been identified, but the study of Listeria suggests tegrin receptors. that such factors exist and are likely to be central regula- Actin filaments are rapidly capped when added to cell tors of actin assembly. extracts (Zigmond et al., 1998), so the cell must have The molecular machinery underlying the spatial regu- some mechanism to inhibit capping or stimulate un- lation of actin assembly has been unveiled through the capping in a spatially controlled fashion to enhance the study of a simple bacterium within its mammalian host. local availability of filament ends for elongation. Even In addition to Listeria, many viral and bacterial patho- in the case of the Listeria tail, local maintenance of gens, including Shigella, Salmonella, and Vaccinia, har- uncapped actin filaments near the bacterial surface is ness the actin assembly machinery of their host cells postulated to be essential for optimal growth of the during their infectious cycles. Although substantial comet tail (Marchand et al., 1995). In an uninfected mam- progress has been made toward dissecting the cellular malian cell, a guanine nucleotide exchange factor such machinery that defines sites of actin assembly, virtually as vav might contribute to adhesion-stimulated actin nothing is known about how specification of the diversity assembly by regulating filament capping. Vav could be of actin filament patterns (see Figure 1 for examples) is localized to integrin-rich sites by virtue of its ability to achieved. Discovery of the myriad ways in which patho- bind zyxin, and its activity could be controlled by recep- gens access cellular pathways that regulate the actin tor-mediated signaling. Engagement of integrin recep- cytoskeleton should contribute substantially to our un- tors activates tyrosine kinases and results in the phos- derstanding of the molecular machineries that control phorylation and activation of vav (Zheng et al., 1996; cell surface extension and locomotion. Crespo et al., 1997). Local activation of zyxin-tethered vav would theoretically lead to local activation of Rho References family members and outputs, including phosphatidyl inositol–induced uncapping of actin filaments. Uncap- Ahern-Djamali, S.M., Comer, A.R., Bachmann, C., Kastenmeier, A.S., ping of preexisting actin filaments could provide a rapid Reddy, S.K., Beckerle, M.C., Walter, U., and Hoffman, F.M. (1998). mechanism for generating elongation-competent sites Mutations in Drosophila Enabled and rescue by human vasodilator- or could contribute to keeping newly generated filament stimulated phosphoprotein (VASP) indicate important functional roles for Ena/VASP homology domain 1 (EVH1) and EVH2 domains. ends from becoming capped. This is an attractive idea, Mol. Biol. Cell 9, 2157–2171. but it should be noted that the form of vav that has Alberts, B., Bray, D., Lewis, J., Raff, M., Roberts, K., and Watson, been shown to bind zyxin is expressed exclusively in J.D. (1994). Molecular Biology of the Cell, Third Edition (New York: hematopoeitic cells, so the broad relevance of this inter- Garland Publishing, Inc.), p. 786. action for cell behavior remains to be determined. Never- Beckerle, M.C. (1997). Zyxin: zinc fingers at sites of cell adhesion. theless, a ubiquitous form of vav has been identified, Bioessays 19, 949–957. and it displays a C-terminal SH3 domain that is very Brindle, N.P.J., Holt, M.R., Davies, J.E., Price, C.J., and Critchley, similar to the originally described zyxin-binding site. D.R. (1996). The focal-adhesion vasodilator-stimulated phospho- By analogy to what is postulated to occur in associa- protein (VASP) binds to the proline-rich domain in vinculin. Biochem. tion with the Listeria surface, the association of Ena/ J. 318, 753–757. VASP proteins with zyxin would be expected to promote Chan, A.Y., Raft, S., Bailly, M., Wyckoff, J.B., Segall, J.E., and Con- actin assembly at these sites. As mentioned above, the deelis, J.S. (1998). EGF stimulates an increase in actin nucleation and filament number at the leading edge of the lamellipod in mam- detailed molecular mechanism by which Ena/VASP pro- mary adenocarcinoma cells. J. Cell Sci. 111, p. 204. teins might enhance actin assembly remains to be de- Crespo, P., Schuebel, K.E., Ostrom, A.A., Gutkind, J.S., and Bustelo, fined. As depicted in Figure 3, the Arp2/3 complex is X.R. (1997). Phosphotyrosine-dependent activation of Rac-1 GDP/ likely to be involved in the nucleation phase of actin GTP exchange by the vav proto-oncogene product. Nature 385, assembly and may cooperate with zyxin and its partners 169–172. to maximize the efficiency of actin assembly. Many as- Domann, E., Wehland, J., Rohde, M., Pistor, S., Hartl, M., Goebel, pects of this model remain to be tested, but it provides W., Leimeister-Wa¨ chter, M., Wuenschner, M., and Chakraborty, T. a molecular means by which a cell adhesion event might (1992). A novel bacterial gene in Listeria monocytogenes required be coupled to changes in the cytoskeleton to elicit a for host cell microfilament interaction with homology to the proline- rich region of vinculin. EMBO J. 11, 1981–1990. response such as cell migration. Fassler, R., Pfaff, M., Murphy, J., Noegel, A.A., Johansson, S., Timpl, R., and Albrecht, R. (1995). Lack of ␤-1 integrin gene in embryonic Lessons Learned stem cells affects morphology, adhesion, and migration but not Listeria is a master at actin assembly, and the study integration into the inner cell mass of blastocysts. J. Cell Biol. 128, of its craft has provided substantial insights into how 979–988. Review 747

Friederich, E., Gouin, E., Hellio, R., Kocks, C., Cossart, P., and Lou- capping, and formation of branching networks of filaments. PNAS vard, D. (1995). Targeting of Listeria monocytogenes ActA protein 95, 6181–6186. to the plasma membrane as a tool to dissect both actin-based cell Mullins, R.D., Kelleher, J.F., Xu, J., and Pollard, T.O. (1998b). Arp2/3 morphogenesis and ActA function. EMBO J. 14, 2731–2744. complex from Acanthamoeba binds profilin and cross-links actin Galileo, D.S., Majors, J., Horwitz, A.F., and Sanes, J.R. (1992). Retro- filaments. Mol. Cell Biol. 9, 841–852. virally introduced antisense integrin RNA inhibits neuroblast migra- Niebuhr, K., Ebel, F., Frank, R., Reinhard, M., Domann, E., Carl, U.D., tion in vivo. Neuron 9, 1117–1131. Walter, U., Gertler, F.B., Wehland, J., and Chakraborty, T. (1997). A Gertler, F.B., Niebuhr, K., Reinhard, M., Wehland, J., and Soriano, novel proline-rich motif present in ActA of Listeria monocytogenes P. (1996). Mena, a relative of VASP and Drosophila Enabled, is impli- and cytoskeletal proteins is the ligand for the EVH1 domain, a protein cated in the control of microfilament dynamics. Cell 87, 227–239. module present in the Ena/VASP family. EMBO J. 16, 5433–5444. Golsteyn, R., Beckerle, M.C., Koay, T., Louvard, D., and Friederich, Olson, M.F., Pasteris, N.G., Gorski, J.L., and Hall, A. (1996). Facio- E. (1997). Structural and functional similarities between the human genital dysplasia protein (FGD1) and vav, two related proteins re- cytoskeletal protein, zyxin, and the ActA protein of Listeria mono- quired for normal embryonic development, are upstream regulators cytogenes. J. Cell Sci. 10, 1893–1906. of Rho GTPases. Curr. Biol. 6, 1628–1633. Hartwig, J.H., Bokoch, G.M., Carpenter, C.L., Janmey, P.A., Taylor, Pistor, S., Chakraborty, T., Niebuhr, K., Domann, E., and Wehland, L.A., Toker, A., and Stossel, T.P. (1995). Thrombin receptor ligation J. (1994). The ActA protein of Listeria monocytogenes acts as a and activated Rac uncap actin filament barbed ends through phos- nucleator inducing reorganization of the actin cytoskeleton. EMBO phoinositide synthesis in permeabilized human platelets. Cell 82, J. 13, 758–763. 643–653. Pistor, S., Chakraborty, T., Walter, U., and Wehland, J. (1995). The Hobert, O., Schilling, J.W., Beckerle, M.C., Ullrich, A., and Jallal, bacterial actin nucleator protein ActA of Listeria monocytogenes B. (1996). SH3 domain-dependent interaction of vav with the LIM- contains multiple binding sites for host microfilament proteins. Curr. domain protein zyxin. Oncogene 12, 1577–1581. Biol. 5, 517–525. Kang, F., Laine, R.O., Rubb, J.R., Southwick, F.S., and Purich, D.L. Reinhard, M., Giehl, C., Abel, K., Haffner, C., Jarchau, T., Hoppe, (1997). Profilin interacts with the Gly-Pro-Pro-Pro-Pro-Pro se- V., Jockusch, B.M., and Walter, U. (1995a). The proline-rich focal quences of vasodilator-stimulated phosphoprotein (VASP): implica- adhesion and microfilament protein VASP is a ligand for profilins. tions for actin-based Listeria motility. Biochemistry 36, 8384–8392. EMBO J. 14, 1583–1589. Kelleher, J.F., Atkinson, S.J., and Pollard, T.D. (1995). Sequences, Reinhard, M., Jouvenal, K., Tripier, D., and Walter, U. (1995b). Identi- structural models, and cellular localization of the actin-related pro- fication, purification, and characterization of a zyxin-related protein teins Arp2 and Arp3 from Acanthamoeba. J. Cell Biol. 131, 385–397. that binds the focal adhesion and microfilament protein VASP (vaso- Kocks, C., Gouin, E., Tabouret, M., Berche, P., Ohayon, H., and dilator-stimulated phosphoprotein). PNAS 92, 7956–7960. Cossart, P. (1992). Listeria monocytogenes-induced actin assembly Sadler, I., Crawford, A.W., Michelsen, J.W., and Beckerle, M.C. requires the actA gene product, a surface protein. Cell 68, 521–531. (1992). Zyxin and cCRP: two interactive LIM domain proteins associ- Kocks, C., Marchand, J.B., Gouin, G., d’Hauteville, H., Sansonetti, ated with the cytoskeleton. J. Cell Biol. 119, 1573–1587. P.J., Carlier, M.F., and Cossart, P. (1995). The unrelated surface Sanger, J.M., Sanger, J.W., and Southwick, F.S. (1992). Host cell proteins ActA of Listeria monocytogenes and IcsA of Shigella flex- actin assembly is necessary and likely to provide the propulsive neri are sufficient to confer actin-based motility on Listeria innocula force for intracellular movement of Listeria monocytogenes. Infect. and Escherichia coli respectively. Mol. Microbiol. 18, 413–423. Immun. 60, 3609–3619. Lasa, I., David, V., Gouin, E., Marchand, J.P., and Cossart, P. (1995). Smith, D.S. (1972). Muscle. (New York: Academic Press, Inc.), p. 9. The amino-terminal part of ActA is critical for the actin-based motility Smith, G.A., Theriot, J.A., and Portnoy, D.A. (1996). The tandem of Listeria monocytogenes; the central proline-rich region acts as repeat domain in the Listeria monocytogenes ActA protein controls a stimulator. Mol. Microbiol. 18, 425–426. the rate of actin-based motility, the percentage of moving bacteria, Lassing, I., and Lindberg, U. (1985). Specific interaction between and the localization of vasodilator-stimulated phosphoprotein and phosphatidylinositol 4,5-bisphosphate and profilactin. Nature 314, profilin. J. Cell Biol. 135, 647–660. 472–474. Svitkina, T.M., Verkhovsky, A.B., McQuade, K.M., and Borisy, G.G. Ma, L., Cantley, L.C., Janmey, P.A., and Kirschner, M.W. (1998). (1997). Analysis of the actin-myosin II system in fish epidermal kera- Corequirement of specific phosphoinositides and small GTP-bind- tocyctes: mechanism of cell body translocation. J. Cell. Biol. 139, ing protein Cdc42 inducing actin assembly in Xenopus egg extracts. 397–415. J. Cell Biol. 140, 1125–1136. Theriot, J.A., Mitchison, T.J., Tilney, L.G., and Portnoy, D.A. (1992). Machesky, L.M., Reeves, E., Wientjes, F., Mattheyse, F.J., Grogan, The rate of actin-based motility of intracellular Listeria monocyto- A., Totty, N.F., Burlingame, A.L., Hsuan, J.J., and Segal, A.W. (1997). genes equals the rate of actin polymerization. Nature 357, 257–260. Mammalian actin-related protein 2/3 complex localizes to regions Theriot, J.A., Rosenblatt, J., Portnoy, D.A., Goldschimdt-Clermont, of lamellipodial protrusion and is composed of evolutionarily con- P.J., and Mitchison, T.J. (1994). Involvement of profilin in the actin- served proteins. Biochem. J. 328, 105–112. based motility of L. monocytogenes in cells and cell-free extracts. Marchand, J.-P., Moreau, P., Paoletti, A., Cossart, P., Carlier, M.-F., Cell 76, 505–517. and Pantaloni, D. (1995). Actin-based movement of Listeria mono- Tilney, L.G., and Portnoy, D.A. (1989). Actin filaments and the growth, cytogenes: actin assembly results from the local maintenance of movement, and spread of the intracellular bacterial parasite Listeria uncapped filament barbed ends at the bacterium surface. J. Cell monocytogenes. J. Cell Biol. 109, 1597–1608. Biol. 130, 331–343. Tilney, L.G., Egelman, E.H., DeRosier, D.J., and Saunder, J.C. (1983). McCollum, D., Feoktistova, A., Morphew, M., Balasubramanian, M., Actin filaments, stereocilia, and hair cells of the bird cochlea. II. and Gould, K.L. (1996). The Schizosaccharomyces pombe actin- packing of actin filaments in the stereocilia and in the cuticular plate related protein, Arp3, is a component of the cortical actin cytoskele- and what happens to the organization when the stereocilia are bent. ton and interacts with profilin. EMBO J. 15, 6438–6446. J. Cell Biol. 96, p. 824. Moreau, V., Madania, A., Martin, R.P., and Winson, B. (1996). The Welch, M.D., Iwamatsu, A., and Mitchison, T.J. (1997a). Actin poly- Saccharomyces cerevisiae actin-related protein Arp2 is involved in merization is induced by Arp2/3 protein complex at the surface of the actin cytoskeleton. J. Cell Biol. 134, 117–132. Listeria monocytogenes. Nature 385, 265–269. Mourrain, P., Lasa, I., Gautreau, A., Gouin, E., Pugsley, A., and Cos- Welch, M.D., DePace, A.H., Verma, S., Iwamatsu, A., and Mitchison, sart, P. (1997). ActA is a dimer. PNAS 94, 10034–10039. T.J. (1997b). The human Arp2/3 complex is composed of evolution- Mullins, R.D., Heuser, J.A., and Pollard, T.D. (1998a). The interaction arily conserved subunits and is localized to cellular regions of dy- of Arp2/3 complex with actin: nucleation, high affinity pointed end namic actin filament assembly. J. Cell Biol. 138, 375–384. Cell 748

Welch, M.D., Rosenblatt, J., Skoble, J., Portnoy, D.A., and Mitchison, T.J. (1998). Interaction of human Arp2/3 complex and the Listeria monocytogenes ActA protein in actin filament nucleation. Science 281, 105–108. Zheng, L., Sjo¨ lander, A., Eckerdal, J., and Andersson, T. (1996).

Antibody-induced engagement of ␤2 integrins on adherent human neutrophils triggers activation of p21ras through tyrosine phosphory- lation of the protooncogene product Vav. PNAS 93, 8431–8436. Zigmond, S.H., Joyce, M., Yang, C., Brown, K., Huang, M., and Pring, M. (1998). Mechanism of Cdc42-induced actin polymerization in neutrophil extracts. J. Cell Biol. 142, 1001–1012.