Report

Interaction of Human Arp2/3 Complex and the Listeria monocytogenes ActA Protein in Actin Filament Nucleation

See allHide authors and affiliations

Science  03 Jul 1998:
Vol. 281, Issue 5373, pp. 105-108
DOI: 10.1126/science.281.5373.105

Abstract

Actin filament assembly at the cell surface of the pathogenic bacterium Listeria monocytogenes requires the bacterial ActA surface protein and the host cell Arp2/3 complex. Purified Arp2/3 complex accelerated the nucleation of actin polymerization in vitro, but pure ActA had no effect. However, when combined, the Arp2/3 complex and ActA synergistically stimulated the nucleation of actin filaments. This mechanism of activating the host Arp2/3 complex at the L. monocytogenes surface may be similar to the strategy used by cells to control Arp2/3 complex activity and hence the spatial and temporal distribution of actin polymerization.

The pathogenic bacteriumListeria monocytogenes initiates actin filament polymerization at its cell surface after it gains access to the cytosol of infected host cells (1). Actin polymerization is tightly coupled to intracellular bacterial motility (2) and may provide the motile force (3). Thus the L. monocytogenes cell surface is functionally similar to the leading edge of lamellipodia in locomoting cells, where actin polymerization is linked with membrane protrusion (4). Understanding the mechanism by which polymerization is instigated byL. monocytogenes should shed light both on an essential aspect of bacterial pathogenesis and on the general mechanisms by which actin filament assembly is modulated in cells.

Actin polymerization at the L. monocytogenes surface is mediated by bacterial and host cell factors. The only essential bacterial component is ActA (5, 6), a cell surface protein that recruits host cell factors that promote actin assembly. A critical host factor is the Arp2/3 complex (7), an evolutionarily conserved protein complex that contains actin-related proteins (Arp) in the Arp2 and Arp3 subfamilies as well as five additional proteins (8–10). This protein complex promotes actin assembly at the bacterial surface, mediates bacterial motility in vitro (7), and is localized throughout actin “comet tails” assembled by moving L. monocytogenes in vivo (7, 10). Moreover, the Arp2/3 complex is concentrated in the lamellipodia of mammalian cells (10, 11) and in pseudopodia of Acanthamoeba castellanii (8, 12, 13), which suggests that it is important for membrane protrusion. Genetic analysis in yeast has demonstrated that the Arp2/3 complex is essential for actin function and cell viability (9, 14).

To further understand the biochemical function of the Arp2/3 complex in cells, we sought to determine how it promotes actin polymerization at the L. monocytogenes surface. Structural models of Arp2 and Arp3 (12) suggest that the complex may serve as a nucleating site for the assembly of actin monomer (G-actin). Nucleation is the rate-limiting step in spontaneous actin polymerization and thus represents a kinetic barrier to actin assembly. Alternatively, the Arp2/3 complex may recruit actin filaments (F-actin) (13), which themselves serve as a template for polymerization. To distinguish between these mechanisms, bacteria were incubated with Arp2/3 complex and equal concentrations of rhodamine-labeled G-actin or F-actin (15). Actin clouds were observed surrounding bacteria incubated with Arp2/3 complex and G-actin (Fig. 1A). In contrast, no actin was associated with bacteria in the presence of Arp2/3 complex and F-actin (Fig. 1B). This strongly favors the nucleation model for Arp2/3 complex function on L. monocytogenes.

Figure 1

Function of the Arp2/3 complex at the L. monocytogenes cell surface. (A) Composite image of DAPI-labeled L. monocytogenes (blue) that were incubated with 0.5 μM TMR-labeled G-actin (red) and 0.07 μM Arp2/3 complex (15). Between 30 and 50% of bacteria assembled actin clouds. (B) Composite image L. monocytogenes(blue) that were incubated with 0.5 μM TMR-labeled F-actin (red) and 0.07 μM Arp2/3 complex (15). No actin was associated with bacteria. These data represent a compilation of 17 individual experiments. Bar = 10 μm.

To determine whether the Arp2/3 complex nucleates actin polymerization in the absence of L. monocytogenes, we observed the effect of pure complex on the kinetics of actin assembly. Polymerization kinetics were monitored in vitro by an assay that employs pyrene-actin, a fluorescent derivative of actin that exhibits much higher intensity of fluorescence when present as F-actin than as G-actin (16). In this assay (17), actin alone exhibited typical assembly kinetics marked by an initial lag phase, indicative of the kinetic barrier to nucleation, followed by a period of rapid assembly that represents filament elongation (Fig. 2A). In the presence of the Arp2/3 complex, the kinetics of polymerization were accelerated relative to actin alone (Fig. 2A), but the initial lag phase of assembly was not significantly shortened, even at higher ratios of Arp2/3:actin. This effect on polymerization is consistent with an ability to accelerate actin filament generation by either facilitating nucleation or severing newly formed filaments. However, pure Arp2/3 complex did not affect the rate of filament depolymerization (17), indicating that it does not sever filaments and suggesting that it facilitates nucleation.

Figure 2

Effects of the Arp2/3 complex and ActA on actin polymerization. (A, C, and E) Graphs of fluorescence intensity versus time after initiating polymerization in the pyrene-actin assay (17). Curve 1, 2 μM actin; curve 2, 2 μM actin with 30 nM Arp2/3 complex (1:65 ratio Arp2/3:actin); curve 3, 2μM actin with 30 nM ActA-His (1:65 ratio ActA:actin); curve 4, 2μM actin with 30 nM Arp2/3 complex and 30 nM ActA (1:1:65 ratio Arp2/3:ActA:actin). (B, D, and F) Electron micrographs of grids spotted with polymerization reaction mixtures 30 s after initiating polymerization (21). (B) Actin (2 μM) with 70 nM Arp2/3 complex [corresponding to (A) curve 2]. (D) Actin (2μM) with 70 nM ActA-His [corresponding to (C) curve 3]. (F) (Top) Actin (2 μM) with 70 nM ActA and 70 nM Arp2/3 complex [corresponding to (E) curve 4]. Arrows indicate actin filaments. (Bottom) Higher magnification view of an actin filament from the same reaction. Bars = 500 nm. (G) Expanded view of the initial 40 s of the graph in (E). (H) Graphs from the pyrene-actin assay. Actin (2 μM) with 20 nM Arp2/3 and, from right to left, 0, 0.04, 0.40, 4, 8, 20, 40, and 60 nM ActA-His. (I) Graphs from the pyrene-actin assay in the absence and presence of CD. Curve 0, 4 μM actin; curve 1, 4 μM actin with 1 μM CD; curve 3, 4 μM actin with 18 nM Arp2/3 and 1 μM CD; curve 4, 4μM actin with 18 nM Arp2/3, 10 nM ActA-His, and 1 μM CD.

Actin nucleation by the Arp2/3 complex at the L. monocytogenes surface also requires the bacterial ActA surface protein (7), and synthetic peptides derived from ActA bind to G- and F-actin (18), which suggests that ActA may itself possess nucleating activity. To determine how ActA affects polymerization kinetics, we constructed and purified a variant of ActA called ActA-His (19) (Fig. 3, A and B). The kinetics of actin polymerization in the presence of ActA-His and in the absence of added protein were identical (Fig. 2C), indicating that full-length ActA does not affect actin assembly (18).

Figure 3

ActA derivatives and their effect on actin nucleation. (A) Schematic representation of ActA-His and ActA-N-His (19) showing the signal sequence (SS), the NH2-terminal region that is essential for actin polymerization (hatched box; amino acids 29 to 263) (18,24), the central region containing four proline-rich repeats (gray boxes, PRR; amino acids 263 to 390), and the COOH-terminal tag of six histidine residues (6HIS; replaces the ActA transmembrane domain in ActA-His). (B) Purified ActA-His (lane 1) and ActA-N-His (lane 2) (19) visualized on a SDS–12% polyacrylamide gel stained with Coomassie blue. (C) Graphs from the pyrene-actin assay (17). Curve 1, 2 μM actin with 20 nM ActA-His and 20 nM Arp2/3 complex; curve 2, 2 μM actin with 20 nM ActA-N-His and 20 nM Arp2/3 complex.

In addition to testing the effects of Arp2/3 complex and ActA on actin assembly individually, we monitored polymerization kinetics in the presence of both pure proteins. In the presence of Arp2/3 complex and ActA-His, the initial rate of actin assembly was accelerated up to 50-fold relative to the reactions in the presence of Arp2/3 or ActA alone (Fig. 2E). Moreover, with both factors present, the lag phase of polymerization was eliminated (Fig. 2G), indicating that Arp2/3 complex and ActA function together as a highly efficient nucleating site, which is kinetically comparable to the end of an actin filament. Addition of increasing amounts of ActA to a fixed concentration of Arp2/3 complex caused a dose-dependent acceleration of the kinetics of actin polymerization (Fig. 2H). This effect was specific to ActA-His because addition of an unrelated protein (His-XCTK2 tail, the COOH-terminal domain of a kinesin family protein) (20) to the Arp2/3 complex did not accelerate polymerization kinetics relative to the Arp2/3 complex alone.

To confirm that the assembly kinetics measured by the pyrene-actin assay represented the kinetics of filament formation, we visualized filaments in polymerization reactions by electron microscopy during the lag phase of spontaneous polymerization (30 s after initiating assembly) (21). Filaments were observed in the reaction mixtures containing both ActA and Arp2/3 complex (Fig. 2F) but not those with Arp2/3 complex or ActA alone (Fig. 2, B and D) (21). These results, together with those obtained by the pyrene-actin assay, demonstrate that the Arp2/3 complex and ActA act synergistically to nucleate actin assembly. We suggest that the nucleation activity of the Arp2/3 complex is stimulated by a physical interaction with ActA because the Arp2/3 complex on its own accelerates actin polymerization, whereas ActA does not. However, we cannot rule out the possibility that ActA participates directly in nucleation.

Actin filaments in the L. monocytogenes comet tail are oriented with their barbed (fast growing) ends toward the bacterial surface, and barbed-end elongation is thought to drive motility (22). To determine which end is elongating in filaments nucleated by Arp2/3 complex and ActA, we performed the pyrene-actin assay in the presence of cytochalasin D (CD). This compound prevents assembly at barbed ends (23) and hence limits actin polymerization to pointed ends. When CD was included in the reaction mixture, the kinetics of Arp2/3 and ActA nucleated actin assembly were nearly identical to the kinetics in the presence of the Arp2/3 complex or actin alone (Fig. 2I). This indicates that filaments nucleated by the Arp2/3 complex and ActA elongate predominantly at their barbed ends.

We next sought to determine which domain of ActA is responsible for its activity. The NH2-terminal region (Fig. 3A) is essential for ActA to induce actin assembly (18, 24). In contrast, the four proline-rich repeats in the central region (Fig. 3A) are not essential but enhance the efficiency of polymerization and motility (24, 25). We generated and purified ActA-N-His, which consists only of the NH2-terminal domain of ActA (19) (Fig. 3, A and B). Equal amounts of ActA-N-His and ActA-His were equivalent in their ability to activate Arp2/3 complex nucleation activity in the pyrene-actin assay (17) (Fig. 3C). ActA-N-His alone had no effect on actin polymerization (17). Thus, the Arp2/3 complex interacts with the NH2-terminal region of ActA to form a nucleating activity, and the proline-rich repeats do not contribute to nucleation in this assay. These repeats may enhance actin polymerization and bacterial motility by recruiting the Enabled/vasodilator-stimulated phosphoprotein (Ena/VASP) family of proteins (26) and profilin (27), which may promote the elongation of filaments nucleated by the Arp2/3 complex and ActA.

Our findings indicate that the Arp2/3 complex and ActA function together to nucleate actin assembly at the L. monocytogenescell surface. We propose the following model for the potential role of these two proteins in actin polymerization and L. monocytogenes motility. Before encountering ActA, the Arp2/3 complex only weakly enhances the kinetics of actin polymerization. Upon interacting with the NH2-terminal domain of ActA, the activity of the complex is stimulated and it nucleates actin assembly, generating actin filaments whose elongation propels the bacterium forward (3). Activation of the Arp2/3 complex may occur by two mechanisms. Interaction with ActA may induce a conformational change in the complex. Alternatively, the complex may be activated by self-association facilitated by ActA, which is a dimer on the bacterial surface (28). In either case, the ability of ActA to activate the Arp2/3 complex explains how the complex can generate actin filaments only at the bacterial surface, as is observed in vivo (2), although it is present throughout the actin tails assembled by moving bacteria (7, 10).

Activation of the Arp2/3 complex with a spatially localized factor such as ActA may represent a general strategy used to regulate the distribution of actin polymerization in cells. Cellular proteins with functions similar to ActA may recruit the complex to lamellipodia and activate its nucleating activity, leading to the generation of filaments that elongate to drive membrane protrusion. Although ActA is the only known regulator of the Arp2/3 complex, other factors such as posttranslational modification (7, 10) may also modulate its function. Thus, multiple pathways may operate in concert to regulate Arp2/3 complex activity. A more complete understanding of the cellular mechanisms that control actin polymerization awaits further determination of how Arp2/3-mediated nucleation is regulated and how it is integrated with other processes such as filament uncapping, elongation, cross-linking, and depolymerization.

REFERENCES AND NOTES

View Abstract

Navigate This Article