Defining a Core Set of Actin Cytoskeletal Proteins Critical For

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

provided by Elsevier - Publisher Connector Cell Host & Microbe Article

Deﬁning a Core Set of Actin Cytoskeletal Proteins Critical for Actin-Based Motility of Rickettsia

Alisa W. Serio,1 Robert L. Jeng,1,3 Cat M. Haglund,1 Shawna C. Reed,2 and Matthew D. Welch1,* 1Department of Molecular and Cell Biology 2Microbiology Graduate Group University of California, Berkeley, Berkeley, CA 94720, USA 3Present address: Burnham Institute for Medical Research, 10901 North Torrey Pines Road, La Jolla, CA 92037, USA *Correspondence: [email protected] DOI 10.1016/j.chom.2010.04.008

SUMMARY monocytogenes, Shigella flexneri, Mycobacterium marinum, and Burkholderia pseudomallei (Gouin et al., 2005). Many Rickettsia species are intracellular bacterial Interestingly, the filament architecture in SFG Rickettsia actin pathogens that use actin-based motility for spread tails differs substantially from that of other pathogens. Actin fila- during infection. However, while other bacteria ments in Rickettsia tails are long and organized into parallel assemble actin tails consisting of branched bundles, similar to actin in cellular protrusions such as filopodia networks, Rickettsia assemble long parallel actin and microvilli (Gouin et al., 1999; Van Kirk et al., 2000). In bundles, suggesting the use of a distinct mechanism contrast, L. monocytogenes tails are composed of a sheath of longer bundled filaments surrounding a core of shorter filaments for exploiting actin. To identify the underlying mech- organized in Y-branched arrays (Brieher et al., 2004; Cameron anisms and host factors involved in Rickettsia parkeri et al., 2001; Sechi et al., 1997), similar to actin in lamellipodia actin-based motility, we performed an RNAi screen (Svitkina and Borisy, 1999). These differences in actin organiza- targeting 115 actin cytoskeletal genes in Drosophila tion presumably reflect differences in how actin filaments are cells. The screen delineated a set of four core assembled and organized at the bacterial surface. proteins—profilin, fimbrin/T-plastin, capping protein, Well-studied pathogens such as L. monocytogenes and S. and cofilin—as crucial for determining actin tail flexneri assemble and organize actin by recruiting and activating length, organizing filament architecture, and the host Arp2/3 complex, an actin-nucleating factor that poly- enabling motility. In mammalian cells, these proteins merizes Y-branched filament networks and is localized to bacte- were localized throughout R. parkeri tails, consistent rial actin tails (Gouin et al., 2005). Both pathogens activate the with a role in motility. Profilin and fimbrin/T-plastin Arp2/3 complex by expressing surface proteins that either mimic or recruit host nucleation-promoting factors (NPFs) (Boujemaa- were critical for the motility of R. parkeri but not Paterski et al., 2001; Egile et al., 1999; Skoble et al., 2001; Suzuki Listeria monocytogenes. Our results highlight key et al., 1998). L. monocytogenes and S. flexneri motility can be re- distinctions between the evolutionary strategies constituted in vitro in a mix of purified proteins including actin, and molecular mechanisms employed by bacterial Arp2/3 complex, capping protein, and actin depolymerizing pathogens to assemble and organize actin. factor (ADF/cofilin) (Loisel et al., 1999). Thus, a small set of core proteins that control actin dynamics and organization is sufficient to drive motility. INTRODUCTION In contrast, for Rickettsia, the molecular mechanism of actin assembly and organization is not well understood. The Rickettsia are Gram-negative, obligate-intracellular bacteria that R. rickettsii sca2 gene was recently reported to be required for are transmitted to mammals by blood-feeding arthropod motility and virulence, but the mechanism was not investigated vectors, and many species cause diseases such as spotted fever (Kleba et al., 2010). In addition, most SFG Rickettsia genomes and typhus (Walker and Ismail, 2008). During infection in encode an NPF for the Arp2/3 complex called RickA, which mammals, Rickettsia induce phagocytosis by endothelial cells can direct Arp2/3 complex-dependent motility of plastic beads and then rapidly escape from the phagosome into the cytosol, in cell extracts, but RickA expression is not directly correlated where they replicate. Once in the cytosol, most spotted fever with Rickettsia spp. pathogenicity (Gouin et al., 2004; Jeng group (SFG) Rickettsia species, as well as the typhus group et al., 2004; Ogata et al., 2001). There is also conflicting evidence (TG) species R. typhi, assemble actin tails at their surface and as to whether the Arp2/3 complex is (Gouin et al., 2004) or is not undergo actin-based motility (Heinzen et al., 1993; Teysseire (Gouin et al., 1999; Van Kirk et al., 2000) associated with motile et al., 1992). Rickettsia are then propelled throughout the cell Rickettsia, and whether the complex is (Gouin et al., 2004)oris and into protrusions, mediating cell-to-cell spread and not (Balraj et al., 2008; Harlander et al., 2003; Heinzen, 2003) enhancing virulence (Kleba et al., 2010). The ability to hijack functionally important for bacterial motility. Numerous other the actin cytoskeleton to drive movement is shared between actin-binding proteins have been localized to Rickettsia tails Rickettsia and other bacterial pathogens including Listeria including profilin, VASP, a-actinin, and filamin (Gouin et al.,

388 Cell Host & Microbe 7, 388–398, May 20, 2010 ª2010 Elsevier Inc. Cell Host & Microbe Host Proteins Critical for Rickettsia Motility

1999; Van Kirk et al., 2000). However, their function in Rickettsia adapters and membrane-cytoskeleton linkers; (5) myosin motility has not yet been investigated, and the actin cytoskeletal motors; and (6) cytoskeletal signaling proteins. proteins required for Rickettsia motility have yet to be defined. Of the 18 genes implicated in infection, RNAi targeting of eight Based on the unique filament architecture in Rickettsia actin resulted in a decrease in bacteria per cell, including those encod- tails, we hypothesized that a distinct set of host cytoskeletal ing the Arp2/3 complex subunit Arp2, the NPFs SCAR and proteins is critical for Rickettsia motility compared to other path- WASP, the SH2/SH3 adaptor protein Drk, the GTPase Mtl, the ogens. To identify the host factors that are required for Rickettsia motors myosin V and myosin VI, and the actin-associated motility, we employed RNA interference (RNAi) to examine the protein Dah. These proteins might enhance R. parkeri infection function of over 100 actin cytoskeletal proteins in Drosophila by promoting internalization, motility or spread. RNAi of ten melanogaster cells. We identified numerous proteins that play others resulted in an increase in bacteria per cell, which was roles in Rickettsia infection and actin-based motility. In partic- caused by an increase in cell size in some cases (allowing ular, profilin, fimbrin/T-plastin, capping protein, and ADF/cofilin more bacteria to accumulate in a single cell), or could also reflect were essential for actin tail morphology and motility. In mamma- more efficient bacterial internalization, replication, motility, lian cells, profilin and fimbrin/T-plastin were specifically impor- and/or spread. tant for motility of R. parkeri but not for L. monocytogenes. The screen also identified 20 genes implicated in actin-based Thus, our results identify a distinct set of core host proteins motility. These included genes encoding Arp2/3 complex that play a role in Rickettsia actin-based motility. subunits (ARPC3, ARPC4, and ARPC5), the G-actin-binding proteins profilin and cyclase-associated protein (CAP), the severing/depolymerizing protein cofilin, the cofilin phosphatase RESULTS slingshot, the actin bundling proteins fimbrin and a-actinin, the GTPases Rac1 and Rac2, the kinase ROCK, the motors myosin RNAi Screening Identifies Proteins Important IA and II, and the endocytic protein Hip1R. Five of the 20 genes— for R. parkeri Infection and Actin Tail Formation encoding SCAR, Dah, Drk, Mtl, and myosin VI—were also impli- Because of the amenability of Drosophila S2R+ cells to RNAi- cated in infection, suggesting that defects in tail formation and mediated gene silencing and the natural ability of Rickettsia to infection can result from inhibition of common pathways. Thus, infect arthropod cells, we used these cells for RNAi screening the RNAi screen identified numerous candidate proteins that to identify host proteins that play a role in Rickettsia actin-based might perform important functions in actin-based motility. motility. We found that R. parkeri, a member of the SFG that causes mild human disease (Paddock et al., 2008), readily RNAi of Numerous Targets Causes Defects in Actin Tail invaded and replicated in S2R+ cells. Moreover, R. parkeri Length and Morphology motility occurred at similar rates in S2R+ (13.0 ± 0.3 mm/min, To identify those target genes that function specifically in actin- n = 23) and mammalian COS7 (13.8 ± 0.43 mm/min, n = 26) cells, based motility, we sought a phenotypic parameter more directly resulting in the formation of actin tails with similar morphology in linked to this process. Previous work with L. monocytogenes both cell types (see Figure S1A available online). This suggests showed that the length of actin tails associated with moving that the mechanism of motility is conserved between species. bacteria directly correlates with the rate of movement (Theriot We conducted an RNAi screen by targeting 115 actin cyto- et al., 1992). We therefore targeted the 34 genes identified in skeletal genes and two noncytoskeletal control genes the original screen using RNAi and measured the effect on actin (Table S1). The list of targets was expanded from a previous tail length. RNAi of 20 genes resulted in significantly shorter actin screen that assessed the role of cytoskeletal proteins in lamellae tails than those in untreated control cells, with the most dramatic formation (Rogers et al., 2003). Cells were treated with dsRNA for reductions (>66%) observed after targeting profilin, fimbrin, and 4 days, infected with R. parkeri for 2 days, and then bacteria and capping protein (Figure 1A). Moreover, RNAi of seven genes re- actin visualized by fluorescence microscopy (Figure S1B). sulted in significantly longer tails, with the largest increases Protein depletion was confirmed for a subset of targets by immu- (30%–50%) observed after targeting SCAR, Hip1R, and cofilin noblotting (Figure S1C). Two phenotypes were then quantified: (Figure 1A). RNAi of Arp2/3 complex subunits did not have the percentage of bacteria associated with actin tails as a dramatic impact on tail length, suggesting that the complex a measure of actin-based motility and the number of bacteria might not play a direct role in Rickettsia motility. Importantly, per cell as a measure of overall infection efficiency. four of the six RNAi targets that produced the most pronounced Thirty-four targets were identified as having potential functions phenotypes—profilin, fimbrin, capping protein, and cofilin—are in actin-based motility and/or infection efficiency (Table 1). RNAi known to play key roles in controlling actin dynamics and orga- of 20 genes resulted in a significant difference in the percentage nization, suggesting these might constitute a core set of proteins of bacteria associated with actin tails, and RNAi of 18 resulted in that are crucial for Rickettsia motility. a significant difference in the number of bacteria per cell, with RNAi of the above-mentioned targets often affected aspects five exhibiting differences for both phenotypes (Table 1, of tail morphology other than length (Figure 1C). For example, Figure S1D). RNAi of capping protein did not result in a significant RNAi of the actin bundling protein fimbrin resulted in short tails difference in either category but resulted in notably short tails so that were significantly wider than in control cells and had loose it was included in further analyses. These 34 proteins fell into six and splayed filament strands (Figures 1B and 1C). This suggests major classes (Figure S1D) including: (1) Arp2/3 complex that fimbrin plays an important role in organizing actin filaments subunits, NPFs, and NPF-binding proteins; (2) actin-monomer in Rickettsia tails. In addition, RNAi of the actin monomer-binding availability proteins; (3) actin-bundling proteins; (4) endocytic protein profilin resulted in very short tails, short actin spikes on

Cell Host & Microbe 7, 388–398, May 20, 2010 ª2010 Elsevier Inc. 389 Cell Host & Microbe Host Proteins Critical for Rickettsia Motility

Table 1. Proteins Implicated in Rickettsia Actin Tail Formation the bacterial surface, or a failure to recruit actin (Figure 1C). and Infection by RNAi Screening When present, actin tails were also significantly wider than those Proteins Implicated Bacteria with in the untreated control cells (Figure 1B). This dramatic pheno- in Tail Formation Tails (%) P Value type suggests that profilin plays an essential role in actin Untreated control 25 ± 13 n/a assembly and organization by Rickettsia. Importantly, targeting profilin, fimbrin, capping protein, and cofilin with a second set a-actinin 9.2 ± 6.7 0.047 of dsRNAs produced nearly identical results, suggesting that ARPC3 7.3 ± 8.2 0.029 the observed phenotypes were not due to off-target effects ARPC4 8.1 ± 5.9 0.035 (Figures 1A and 1B). ARPC5 3.9 ± 1.5 0.0095 CAP 5.4 ± 1.7 0.015 Profilin, Fimbrin, Capping Protein, and Cofilin Play Capping proteina 16 ± 6.9 0.19 Direct Roles in Actin-Based Motility Cofilin 4.2 ± 1.6 0.010 The defects in actin tail length and morphology observed upon Dahb 6.7 ± 4.9 0.0095 RNAi targeting of profilin, fimbrin, capping protein, and cofilin suggested that these factors play a direct role in motility. To Drk b 11 ± 5.1 0.042 test this, we measured the rates of actin tail growth in live R. par- DROK 7.6 ± 7.6 0.014 keri-infected S2R+ cells expressing Lifeact-GFP, an F-actin- Fimbrin 8.3 ± 5.0 0.017 binding probe (Riedl et al., 2008). RNAi targeting of fimbrin, Hip1R 8.6 ± 4.4 0.039 capping protein, and cofilin caused a moderate and statistically Mtl b 5.8 ± 3.2 0.0066 significant reduction in motility rates compared to controls Myosin IA 11 ± 7.3 0.045 (Figure 2A). Moreover, RNAi of profilin caused a dramatic Myosin II 7.3 ± 8.1 0.012 reduction in motility rates (Figure 2A). Thus, all four proteins Myosin VI b 8.8 ± 7.0 0.021 are important for actin-based motility, and profilin plays a partic- ularly critical role in this process. Profilin 0.0 ± 0.0 0.00 To further examine the role of these proteins in actin-based Rac1 8.9 ± 7.9 0.022 motility, we explored the relationship between the motility rate Rac2 6.5 ± 4.6 0.021 and tail length. It was previously reported that L. monocytogenes b SCAR 2.6 ± 1.9 0.0018 motility rates (3–21 mm/min) and tail lengths (1–16 mm) varied Slingshot 7.3 ± 11 0.030 widely, and a linear correlation (R2 > 0.96) between the two Proteins Implicated parameters could be observed (Theriot et al., 1992). We there- in Infection Bacteria/Cell P Value fore measured both the motility rate and mean tail length for Untreated control 6.9 ± 4.0 n/a each moving bacterium over a 1–3 min interval in control and Anillin 18 ± 7.5 <0.0001 RNAi-treated cells. In untreated controls, there was a compact Arp2 2.1 ± 1.5 0.026 distribution of motility rates (10.5–15.5 mm/min) and tail lengths (9–22 mm), and no linear correlation between the two parameters Arp3 13 ± 9.9 0.021 was observed (R2 = 0.02) (Figure 2B). However, when values for Band 4.1 FERM-like 12 ± 5.4 0.030 control and RNAi-targeted cells were taken together, there was b Dah 3.1 ± 0.95 0.044 a stronger linear correlation (R2 = 0.30). Interestingly, the values Drk b 2.8 ± 1.2 0.033 for fimbrin and cofilin appeared to be outliers, because omitting MIM-like 20 ± 7.8 <0.0001 these from the analysis resulted in an even stronger linear corre- 2 Mtl b 2.9 ± 2.1 0.038 lation (R = 0.80), suggesting that silencing these proteins Myosin V 2.5 ± 1.2 0.039 caused a deviation in the relationship between tail length and Myosin VI b 2.9 ± 0.69 0.035 motility rate. In particular, fimbrin RNAi caused a 23% decrease in motility rate but a 62% decrease in tail length (Figure 2B). This Pp2A 18 ± 22 0.011 suggests that fimbrin maintains the structure of the tail but does Rab5 14 ± 10 0.0071 not dramatically influence the rate of actin assembly. Moreover, RacGAP50C 26 ± 31 0.0011 cofilin RNAi caused a 33% reduction in motility rate but had no SCAR b 1.8 ± 1.0 0.0086 significant impact on mean tail length, although the distribution Twinfillin 14 ± 12 0.012 of lengths was broadened and longer tails were observed Villin-like 17 ± 15 0.0026 (Figure 2B). This suggests that actin disassembly by cofilin might Vinculin 13 ± 13 0.035 replenish the monomer pool to fuel actin assembly and bacterial WASP 2.7 ± 1.6 0.046 movement. The percentage of bacteria with actin tails and the number of bacteria per cell are presented as the mean ± standard deviation. The p value was a No statistically significant difference in the percentage of bacteria with determined by pairwise comparison with the untreated control using tails was observed, but tails were uniformly shorter. Student’s t test. p < 0.05 was considered statistically significant. n/a b Targets for which both the percentage of bacteria with tails and the means not applicable. For related supplemental information, see also number of bacteria per cell were statistically different from the control Figure S1 and Table S1. as defined by the p value.

390 Cell Host & Microbe 7, 388–398, May 20, 2010 ª2010 Elsevier Inc. Cell Host & Microbe Host Proteins Critical for Rickettsia Motility

Figure 1. RNAi Identified Numerous Targets that Affect Actin Tail Length and Morphology (A and B) Graphs of actin tail lengths (A) and widths (B) following RNAi of the indicated targets. Data are mean ± SEM. Black bar indicates the untreated control, gray bars indicate no difference from the control based on a pairwise Student’s t test (p > 0.05), and white bars indicate a statistically significant difference from the control (p < 0.05). In cases where two distinct dsRNAs were used, targets are designated with À1 and À2. (C) Z sections from deconvolved images of untreated cells, or cells treated with dsRNAs targeting profilin, fimbrin, and capping protein. R. parkeri (red) was visualized by immunofluorescence, and actin tails (green) were visualized with Alexa Fluor 488 phalloidin. Scale bar, 10 mm.

Profilin, T-plastin, capping protein, and cofilin all localized to tails formed by both pathogens (Figure 3; Movies S4–S7). Plastin-GFP exhibited the most promi- nent localization to R. parkeri and L. monocytogenes tails (Movie S4). Profi- lin-GFP (Movie S5), GFP-capping protein Next, we examined the morphology of actin tails in live S2R+ (Movie S6), and GFP-cofilin (Movie S7) were visible in R. parkeri cells expressing Lifeact-GFP to label actin (Movies S1–S3). tails, but the intensity of fluorescence was less than in L. mono- Compared with controls (Movie S1), actin tails in capping protein cytogenes tails, perhaps due to a lower overall density of actin (Movie S1) and cofilin RNAi cells (Movie S2) were morphologi- filaments in R. parkeri tails. GFP-Arp3 localized only to cally normal but were shorter after RNAi of capping protein. In L. monocytogenes tails, and GFP did not localize to any tails contrast, tails in fimbrin RNAi cells were short and disorganized, (Movie S8). The localization of profilin, T-plastin (fimbrin), with loose filament strands splaying out from the bacterial capping protein, and cofilin to actin tails is consistent with the surface (Movie S2). In profilin RNAi cells, bacteria were either hypothesis that they play a direct role in actin polymerization associated with actin clouds and failed to move, or formed and organization during actin-based motility. very short actin tails (Movie S3). These results confirm that profilin, fimbrin, capping protein, and cofilin comprise a core set of Fimbrin/T-Plastin Is Important for R. parkeri Actin Tail proteins that are important for R. parkeri actin tail assembly Formation and Actin-Based Motility in Mammalian Cells and motility, and that each likely plays a distinct role in this The importance of fimbrin for Rickettsia actin tail organization process. and motility in Drosophila cells, coupled with the localization of fimbrin/T-plastin to actin tails, suggested that this protein might play an important role in R. parkeri motility in mammalian cells. Profilin, Fimbrin, Capping Protein, and Cofilin Are To evaluate the function of fimbrin/T-plastin, COS7 cells were Localized to R. parkeri Actin Tails in Mammalian Cells transfected with two siRNAs targeting T-plastin, or with control We hypothesized that the core proteins critical for Rickettsia nonspecific or GAPDH siRNAs, and silencing was confirmed motility in insect cells were also likely to perform an important by immunoblotting (Figure S2). Cells were then infected with R. role in mammalian cells. To begin to test this, we observed parkeri or L. monocytogenes, and the percentage of bacteria whether each protein was localized to Rickettsia actin tails. associated with actin tails was quantified. Strikingly, fimbrin/T- Mammalian COS7 cells were transfected with plasmids ex- plastin knockdown caused a 50% reduction in the percentage pressing GFP-tagged profilin II (profilin-GFP), T-plastin (plas- of R. parkeri associated with tails (Figures 4A and 4G) but had tin-GFP; the mammalian equivalent of fimbrin), capping no effect on L. monocytogenes (Figure 4D). Thus, fimbrin/T-plas- protein b subunit (GFP-capping protein), and cofilin (GFP-cofi- tin plays an important role in the initiation and/or maintenance of lin). For controls, we used plasmids expressing GFP or GFP- R. parkeri, but not L. monocytogenes, actin-based motility. Arp3 (Arp3 was reported to localize to L. monocytogenes, We next tested whether silencing of fimbrin/T-plastin affected but not to R. conorii or R. rickettsii actin tails) (Gouin et al., the length of bacterial actin tails or the rate of motility. COS7 cells 1999; Harlander et al., 2003). Cells were subsequently infected were cotransfected with fimbrin/T-plastin or nonspecific siRNAs, with R. parkeri or L. monocytogenes, and the localization of together with a plasmid expressing Lifeact-GFP, and then in- proteins was visualized by time-lapse fluorescence micros- fected with R. parkeri or L. monocytogenes. Actin tail length copy. and motility rate were measured over a 1–3 min interval by

Cell Host & Microbe 7, 388–398, May 20, 2010 ª2010 Elsevier Inc. 391 Cell Host & Microbe Host Proteins Critical for Rickettsia Motility

Figure 2. RNAi of Profilin, Fimbrin, Capping Protein, and Cofilin Resulted in Reduced Motility Rates (A) Graph of R. parkeri motility rates (mean ± SD) in untreated cells, and cells treated with dsRNAs targeting profilin, fimbrin, capping protein, and cofilin. Rates for all dsRNA-treated cells are different from the untreated control (Kruskal-Wallis and Dunn’s multiple comparison tests; p < 0.05), and rates for profilin RNAi are different from all other RNAi targets (p < 0.01). (B) Scatter plot showing the mean tail length and the corresponding mean motility rate for individual bacteria in untreated control cells (black) or cells treated with dsRNAs targeting profilin (red), fimbrin (blue), capping protein (yellow), and cofilin (green). The gray line is the best linear fit for the combined data set. See also Movies S1–S3.

time-lapse fluorescence microscopy. Both Rickettsia and reported importance of profilin for L. monocytogenes motility Listeria actin tail lengths were indistinguishable in cells silenced (Grenklo et al., 2003; Loisel et al., 1999; Theriot et al., 1994), no for fimbrin/T-plastin expression versus their respective control significant difference was observed in the percentage of bacteria cells (Figures 4B, 4E, and 4G). However, the mean rate of R. par- associated with actin, the length of tails, or the rate of motility in keri motility was significantly reduced from 14 mm/min in control profilin 1 RNAi cells versus controls (Figures 5D–5F). Together cells to 10 mm/min in cells silenced for fimbrin-T-plastin these observations indicate that profilin is specifically required (Figure 4C; Movies S9 and S10; p < 0.0001 by Student’s t for rapid and efficient Rickettsia actin-based motility, and that test). There was no significant difference in the rate of L. mono- Rickettsia motility is exquisitely sensitive to the levels of profilin cytogenes motility in fimbrin/T-plastin RNAi versus control cells compared to Listeria. (Figure 4F). These results confirm that fimbrin/T-plastin plays an important and specific role in R. parkeri, but not L. monocyto- DISCUSSION genes, actin-based motility in mammalian cells.

The unique actin filament architecture in Rickettsia actin tails Profilin Is Crucial for Rapid Rickettsia Actin-Based suggests that they use a distinct set of host cytoskeletal proteins Motility in Mammalian Cells to assemble and organize actin filaments. Using RNAi screening The observations that profilin is required for rapid R. parkeri we identified numerous cytoskeletal proteins that play a role in motility in Drosophila cells and is localized to actin tails suggests Rickettsia motility, including a set of four core proteins—profilin, that it might also play an important role in Rickettsia motility in fimbrin/T-plastin, capping protein, and cofilin—that are impor- mammalian cells. To examine the functional importance of pro- tant for this process. Two of these proteins, profilin and fim- filin, COS7 cells were transfected with two siRNAs targeting pro- brin/T-plastin, are specifically required for the actin-based filin 1 (the ubiquitously-expressed isoform) (Jockusch et al., motility of R. parkeri, but not L. monocytogenes. These results 2007), or with nonspecific or GAPDH siRNAs as controls, and provide a comprehensive analysis of the molecular requirements silencing was confirmed by immunoblotting (Figure S2). Cells for Rickettsia motility and highlight key distinctions in the mech- were then infected with R. parkeri or L. monocytogenes and anism of actin assembly and organization between different the percentage of bacteria associated with actin, tail lengths, bacterial pathogens. and motility rates was measured as described above for fim- Comparison of the core proteins required for R. parkeri, L. brin/T-plastin. monocytogenes, and S. flexneri motility revealed interesting RNAi silencing of profilin 1 in COS7 cells had no impact on the similarities and differences. Reconstitution of L. monocytogenes percentage of R. parkeri associated with actin tails compared and S. flexneri motility requires three factors: the Arp2/3 with controls (Figure 5A) but had a dramatic impact on R. parkeri complex, capping protein, and cofilin (Loisel et al., 1999). Our tail morphology and motility. Compared to controls, Rickettsia results suggest that the Arp2/3 complex is not required for Rick- tails were 66% shorter in profilin 1 RNAi cells (Figures 5B, 5C, ettsia motility (discussed below). On the other hand, capping and 5G; Movies S9 and S11), and there was a 60% reduction protein and cofilin play an important role during Rickettsia in motility rate (Figure 5C; mean 5.4 mm/min in profilin 1 RNAi; motility in various cell types. 13.8 mm/min in control cells; p < 0.0001 by Student’s t test). Inter- The importance of capping protein for maintaining rapid Rick- estingly, Rickettsia in profilin 1 knockdown cells were often ettsia motility and long comet tails is consistent with its role in observed in microcolonies and were associated with actin in enhancing L. monocytogenes and bead motility in vitro (Akin a rosette-like pattern (Figure 5G), suggesting that bacterial repli- and Mullins, 2008; Loisel et al., 1999). Two models have been cation occurred without significant movement. Despite the proposed to explain the importance of capping protein in

392 Cell Host & Microbe 7, 388–398, May 20, 2010 ª2010 Elsevier Inc. Cell Host & Microbe Host Proteins Critical for Rickettsia Motility

Figure 3. Localization of GFP-Tagged Cytoskeletal Proteins to R. parkeri and L. monocytogenes Tails Images taken from movies of COS7 cells expressing the indicated GFP-tagged proteins or GFP alone. Cells were either infected with R. parkeri or L. monocytogenes. Arrowheads indicate motile bacteria. Scale bar, 5 mm. See also Movies S4–S8. actin-based motility. In the actin funneling model (Carlier and 1997) and that addition of cofilin accelerates L. monocytogenes Pantaloni, 1997), capping protein caps most filament barbed motility (Carlier et al., 1997). Cofilin functions to accelerate actin ends in cells, increasing the concentration of free actin mono- dynamics (Van Troys et al., 2008) by both severing ADP-bound mers that funnel onto uncapped barbed ends to promote rapid filaments at low cofilin:actin ratios and nucleating new filaments elongation. In the monomer gating model (Akin and Mullins, at high cofilin:actin ratios (Andrianantoandro and Pollard, 2006). 2008), capping protein acts as a switch between elongation Given the low levels of GFP-cofilin observed in Rickettsia tails and nucleation by terminating elongation and biasing new mono- and the increased tail lengths observed after cofilin depletion, mers to participate in filament nucleation by the Arp2/3 complex. it is likely that the key role of cofilin during Rickettsia motility is Although the monomer gating model is applicable to L. monocy- to sever/disassemble actin filaments in actin tails and elsewhere, togenes motility, which requires the Arp2/3 complex as a nucle- and to regenerate the actin monomer pool to fuel assembly and ator (Gouin et al., 2005), our evidence fails to support a major role bacterial movement. for Arp2/3 complex in Rickettsia motility (discussed below). Our results also indicate that Rickettsia motility involves two Thus, during Rickettsia motility, capping protein might act as additional core proteins, fimbrin/T-plastin and profilin, which a monomer gate with a different nucleator, or might cap the are not required to reconstitute L. monocytogenes and S. flexneri bulk filament population and enable monomer funneling onto un- motility. Fimbrin/T-plastin bundles actin filaments that are capped ends adjacent to the bacterium. oriented with uniform polarity (Delanote et al., 2005) like those Our observation that cofilin is important for shortening Rick- in Rickettsia tails (Gouin et al., 2004; Van Kirk et al., 2000). Inter- ettsia actin tails and maintaining rapid motility is similar to estingly, other actin bundling proteins including a-actinin, fascin, previous reports that depletion of cofilin from cell extracts and filamin reportedly localize to Rickettsia tails (Gouin et al., increases L. monocytogenes tail length (Rosenblatt et al., 2004; Van Kirk et al., 2000), but we did not observe any effect

Cell Host & Microbe 7, 388–398, May 20, 2010 ª2010 Elsevier Inc. 393 Cell Host & Microbe Host Proteins Critical for Rickettsia Motility

Figure 4. RNAi of T-Plastin in Mammalian Cells Reduces the Frequency and Rate of Motility COS7 cells were treated with siRNA targeting T-plastin (PLS3-1, À2), or control GAPDH (GAPD) or nonspecific (NS) siRNAs, and infected with either R. parkeri (A–C and G) or L. monocytogenes (D–F). (A and D) Graphs of the percentage of bacteria (mean ± SD) associated with actin tails after transfection with the indicated siRNA. Triple asterisks indicate a statistically significant difference after pairwise comparison to the nonspecific siRNA control (p < 0.001, Student’s t test). (B and E) Scatter plots of tail length after transfection with the indicated siRNA (red bar indicates mean). (C and F) Scatter plots of the mean tail length and the corresponding mean motility rate for individual bacteria in cells transfected with control nonspecific (black) or T-plastin siRNA (blue). (G) R. parkeri (red) visualized by immunofluorescence and actin (green) visualized with Alexa Fluor 488 phalloidin in cells transfected with nonspecific (NS) or T-plastin (PLS3-1) siRNAs. Scale bar, 5 mm. See also Figure S2 and Movies S9 and S10.

rate of L. monocytogenes motility in vitro and in cells (Geese et al., 2002; Kang et al., 1997; Loisel et al., 1999; Theriot et al., 1994). Regardless, our results suggest that Rickettsia motility has a much more stringent dependence on profilin, highlighting a key difference in the mechanisms of actin assembly between these pathogens. Profilin possesses several biochemical activities that are relevant to bacterial actin assembly, including the ability to lower the critical concentration for barbed end actin assembly, enhance dynamics on Rickettsia tail formation and motility after RNAi targeting these at barbed ends, and accelerate the exchange of ADP for ATP proteins. Although silencing fimbrin/T-plastin decreased the on actin (Yarmola and Bubb, 2006). We propose that, unlike for frequency and rate of Rickettsia motility and caused disorga- L. monocytogenes, actin nucleation and/or elongation at the nized comet tails, it had no effect on L. monocytogenes tail Rickettsia surface requires one or more of these activities of formation or motility. Together these results indicate that fim- profilin. brin/T-plastin plays a primary role in bundling actin and estab- Surprisingly, our RNAi screen did not identify a cellular actin lishing the unique architecture of the parallel actin filaments in nucleator that is required for R. parkeri motility. Although SFG Rickettsia tails, and in initiating and maintaining motility by form- Rickettsia express an activator of the Arp2/3 complex called ing tight bundles that provide mechanical rigidity for force gener- RickA, RNAi targeting of Arp2/3 complex subunits in Drosophila ation. Additionally, fimbrin/T-plastin might inhibit the filament cells had no dramatic effect on actin tail length. This is consistent severing or disassembling activity of cofilin (Giganti et al., with a report that overexpressing the Arp2/3-binding WCA 2005), further stabilizing filaments in Rickettsia actin tails. domain of the NPF SCAR had little effect on R. rickettsii motility Interestingly, our results indicate that the actin monomer but inhibited S. flexneri motility (Harlander et al., 2003), although binding protein profilin plays a profoundly important role in Rick- another group reported that SCAR WCA expression inhibited ettsia motility, and is crucial for actin tail formation and rapid R. conorii motility (Gouin et al., 2004). These results, together movement. In contrast, we did not detect a role for profilin in L. with reports that the Arp2/3 complex is absent from Rickettsia monocytogenes actin tail formation or motility. Although this is tails (Gouin et al., 1999; Van Kirk et al., 2000; this study), suggest consistent with a previous report that depleting profilin from that the Arp2/3 complex is not required for Rickettsia motility, in cell extract had no impact on L. monocytogenes motility (March- contrast with L. monocytogenes (Loisel et al., 1999). This is also and et al., 1995), other reports suggest that profilin increases the consistent with reports that RickA expression does not always

394 Cell Host & Microbe 7, 388–398, May 20, 2010 ª2010 Elsevier Inc. Cell Host & Microbe Host Proteins Critical for Rickettsia Motility

Figure 5. RNAi of Profilin 1 in Mammalian Cells Reduces the Length of R. parkeri Actin Tails and the Motility Rate COS7 cells were treated with siRNA targeting profilin 1 (PFN1-1, À2), or control GAPDH (GAPD) or nonspecific (NS) siRNAs, and infected with either R. parkeri (A–C and G) or L. monocytogenes (D–F). (A and D) Graphs of the percentage of bacteria associated with actin tails (mean ± SD) after transfection with the indicated siRNA. (B and E) Scatter plots of the length of actin tails after transfection with the indicated siRNA (red bar indicates mean). Triple asterisks indicate a statistically significant difference after pairwise comparison to the nonspecific siRNA control (p < 0.001); Student’s t test). (C and F) Scatter plots showing the mean tail length and the corresponding mean motility rate for individual bacteria in cells transfected with control nonspecific siRNA (black) or with profilin 1 siRNA (red). (G) R. parkeri (red) visualized by immunofluorescence and actin (green) visualized with Alexa Fluor 488 phalloidin in cells transfected with nonspecific (NS) or profilin 1 (PFN1-1) siRNAs. Scale bar, 5 mm. See also Figure S2 and Movies S9 and S11.

2010), as well as four core host proteins that include profilin, fimbrin/T-plastin, capping protein, and cofilin. Profilin is critical for delivering ATP-bound actin monomers to barbed ends at the bacterial surface, enhancing filament elongation and force generation. Capping protein blocks monomer addition on older barbed ends, funneling monomers onto, or gating nucleation of, newer filaments at the bacterial surface. Fimbrin/ T-plastin bundles filaments into long helical arrays, generating a structural correlate with Rickettsia motility or virulence (Balraj et al., 2008). scaffold that supports force generation and also competes On the other hand, silencing of Arp2/3 complex subunits as well with cofilin to enhance filament stability. Ultimately, cofilin as the NPFs SCAR and WASP decreased the efficiency of R. par- severs older ADP-bound filaments, replenishing the monomer keri infection, suggesting that the Arp2/3 complex might function pool and enhancing the rate of actin assembly and motility. In in host cell entry or cell-to-cell spread, consistent with previous addition, our screen identified a number of other proteins, results (Martinez and Cossart, 2004). including those that participate in actin dynamics and organiza- RNAi targeting of other cellular actin nucleation and elongation, endocytosis, membrane linkage, and signaling, that might tion factors that form unbranched filaments (Campellone and play an accessory role in actin-based motility and infection or Welch, 2010), including all six Drosophila formin family might participate indirectly via an effect on bacterial entry or proteins, the WH2-domain nucleator Spire, and Ena/VASP spread. Based on the success of our targeted screen, proteins, also had no effect on Rickettsia actin tails. These genome-wide RNAi screening will likely identify additional host results suggest that no single host protein nucleates actin fila- proteins involved in Rickettsia infection, replication, motility, ments during Rickettsia motility, and point to the possible exis- and spread. tence of a bacterial actin nucleator. It was recently reported Our work demonstrates that the molecular requirements for that the R. rickettsii Sca2 protein is required for actin assembly actin assembly and organization into the distinct filament and motility, suggesting that it is an actin nucleation factor bundles observed in Rickettsia tails include proteins that play (Kleba et al., 2010). a widespread role in the actin-based motility of bacterial patho- We propose a model for the mechanism of actin assembly gens, as well as proteins that play a uniquely important role for and organization during Rickettsia actin-based motility (Figure 6). Rickettsia. Because the organization of actin in Rickettsia tails In this model, R. parkeri tail assembly and motility requires is similar to that in cell-surface protrusions such as filopodia a bacterial actin nucleator that is likely Sca2 (Kleba et al., and microvilli, Rickettsia motility might serve as a model for

Cell Host & Microbe 7, 388–398, May 20, 2010 ª2010 Elsevier Inc. 395 Cell Host & Microbe Host Proteins Critical for Rickettsia Motility

To generate a plasmid that expresses Lifeact-GFP in mammalian cells, complementary DNA oligomers encoding Lifeact (Riedl et al., 2008) were annealed and subcloned into pEGFP-N1 (Clontech) to generate an in-frame Life- act-GFP fusion gene. To generate a Lifeact-GFP expressing baculovirus for insect cell transduction, theLifeact-GFPlocus wasﬁrstsubclonedinto a plasmid containing the baculovirus ie-1 promoter and then subcloned into a plasmid containing ﬂanking regions of homology with the baculovirus genome, including the essential viral p78/83 gene. Cotransfection of this plasmid with a bacmid containing a p78/83 deletion resulted in the restoration of the p78/83 gene and the production of a baculovirus expressing ie-1 Lifeact-GFP.

Bacterial and Cell Growth and Infection Bacteria and cells were obtained from the following sources: R. parkeri Ports- mouth strain (C. Paddock, CDC) (Paddock et al., 2004), L. monocytogenes 10403S (D. Portnoy, University of California, Berkeley) (Bishop and Hinrichs, 1987), Drosophila S2R+ cells (R. Tjian, U. C. Berkeley), and COS7 and Vero cells (UC Berkeley, tissue culture facility). S2R+ cells were grown at 28C in M3 Shields and Sang media (Sigma) sup-

plemented with 0.5 g/L KHCO3, 1 g/L yeast extract, 2.5 g/L meat peptone extract, and 10% fetal bovine serum (FBS). COS7 and Vero cells were grown Figure 6. Model of the Molecular Mechanism of Actin-Based Rick- in DMEM supplemented with 2%–10% FBS at 37 C with 5% CO2. R. parkeri ettsia Motility was propagated in Vero cells at 33 C with 5% CO2 and purified by Renografin Actin filament assembly at the bacterial surface is induced by a nucleator that density gradient centrifugation (Hackstadt et al., 1992). L. monocytogenes was is likely of bacterial origin (red drop). Actin monomers (green circles) associ- grown in liquid brain heart infusion media at 37 C. ated with profilin (blue circles) assemble onto filament barbed ends at the R. parkeri was infected at an moi of 0.2–1.0. Cells to be fixed were infected bacterial surface, generating motile force. Fimbrin/T-plastin (yellow cross) for 48 hr, while cells to be viewed live were infected for 24 hr. For infection with bundles the growing filaments into long helical strands, while displacing cofilin L. monocytogenes, bacteria were preincubated in DMEM for 1 hr at 33 C with (orange triangles). Cofilin binds to older ADP-actin filaments and facilitates 5% CO2, then added at an moi of 5–10 for 3.5 hr at 33 C with 5% CO2, and m severing and disassembly at the pointed ends, replenishing the monomer finally treated with 10 g/ml gentamicin for 0.5 hr prior to fixation to kill extra- pool. Capping protein (purple arches) blocks addition of ATP-actin monomers cellular bacteria. to the barbed ends of older filaments, enhancing the flux of profilin-bound actin monomers onto newer barbed ends. RNA Synthesis and RNAi Screening in S2R+ Cells Primers for PCR of target genes were derived from those published by Rogers et al. (2003) or from the Drosophila RNAi Screening Center (DRSC, understanding the mechanisms of actin function in these cellular http://flyrnai.org/). Primers for the second set of dsRNAs for profilin, fimbrin, structures. Future work will reveal additional details of the molec- capping protein, and cofilin were recommended by the DRSC based on algo- ular mechanism of Rickettsia motility and spread and might rithms used to minimize potential off-target effects. Each primer contained uncover new paradigms for the cellular regulation of actin. a T7 RNA polymerase-binding site. Following PCR from Drosophila genomic DNA, the sequences were confirmed. In vitro transcription was performed in

80 mM Tris (pH 8.0), 10 mM DTT, 2 mM Spermidine-HCl, and 20 mM MgCl2, EXPERIMENTAL PROCEDURES with 7.5 mM each NTPs and 109 U RNAsin, 0.4 U yeast inorganic pyrophos- phatase, and 91 U T7 RNA polymerase. The reactions were incubated for Antibodies 6hrat37 C. RNA products were LiCl2 purified, confirmed to be of the correct Antibodies were obtained from the following sources: mouse anti-Dm-profilin size by agarose gel electrophoresis, and their concentration determined by (developed by L. Cooley at Yale School of Medicine, maintained by the Devel- spectrophotometry. opmental Studies Hybridoma Bank [DSHB] at University of Iowa), rabbit anti- For RNAi screening, monolayers of S2R+ cells were treated with dsRNA at Dm-anillin (C. Field, Harvard Medical School) (Field and Alberts, 1995), rabbit a final concentration of 20 mg/ml for 4 days and then infected with R. parkeri anti-Dm-cofilin serum (M.L. Goldberg, Cornell University), rat anti-Dm-Arp3 for 2 days. RNAi was performed for each target in triplicate. Cells were fixed (L. Cooley) (Hudson and Cooley, 2002), guinea pig anti-Dm-SCAR (J. Zallen, and processed for immunofluorescence microscopy, as described below. Sloan-Kettering Institute) (Zallen et al., 2002), mouse anti-Dm-a-tubulin (GE The number of bacteria per infected cell and the number of tails per bacterium Healthcare), rabbit anti-Hs-profilin-1 (Cell Signaling Technology), mouse were counted in 50–100 cells in over at least ten different fields of view. The anti-GAPDH (Ambion), rabbit anti-Hs-T-plastin C-15 (Santa Cruz Biotech- values for each target were compared pairwise with the untreated control by nology), mouse anti-tubulin (developed by M. Klymkowsky at the University the Student’s t test using Prism (Graphpad Software). Differences were of Colorado, Boulder, maintained by the DSHB), mouse anti-Rickettsia considered to be significant if p < 0.05. M14-13 and rabbit anti-Rickettsia antibody R4668 (T. Hackstadt, NIH/NIAID) (Anacker et al., 1987; Policastro et al., 1997), rabbit anti-Listeria O antibody Transfection of Mammalian Cells (BD Difco), horseradish peroxidase-conjugated secondary antibodies for For expression of GFP-tagged proteins, 0.4–0.6 mg of plasmid DNA was tran- immunoblotting (GE Healthcare), and Alexa Fluor 568-conjugated secondary siently transfected into COS7 cells using Lipofectamine 2000 (Invitrogen). For antibodies for immunofluorescence (Invitrogen). Abbreviations are as follows: gene silencing, siRNAs (Ambion) at a final concentration of 50 nM were trans- Drosophila melanogaster, Dm; Homo sapiens, Hs. fected into COS7 cells using Lipofectamine-RNAi-MAX (Invitrogen) in serum- free DMEM. At 48 hr posttransfection, cells were infected with R. parkeri,or Plasmids and Baculovirus Strains at 92 hr cells were infected with L. monocytogenes. Cells were fixed at 96 hr Plasmids were obtained from the following sources: pEGFP-N1-profilin II and were processed for immunoblotting or immunofluorescence, as described (A. Sechi, Universita¨ tsklinikum Aachen) (Geese et al., 2000), pCDNA3.1- below. smRS-GFP-Plastin (T. Timmers, INRA/CNRS) (Timmers et al., 2002), pEGFP-C1-capping protein b2 subunit (Addgene) (Schafer et al., 1998), Immunoblotting, Immunofluorescence, and Phenotypic Analysis pEGFP-C2-cofilin (H.G. Mannherz, Ruhr University) (Mannherz et al., 2005), For immunoblotting, protein samples were separated by SDS-PAGE, trans- and pEGFP-N1-ACTR3 (Welch et al., 1997). ferred to nitrocellulose membranes, probed with primary and secondary

396 Cell Host & Microbe 7, 388–398, May 20, 2010 ª2010 Elsevier Inc. Cell Host & Microbe Host Proteins Critical for Rickettsia Motility

antibodies, and visualized by enhanced chemiluminescence (GE Healthcare). Anacker, R.L., Mann, R.E., and Gonzales, C. (1987). Reactivity of monoclonal For immunofluorescence, cells were fixed with 4% formaldehyde in PBS for antibodies to Rickettsia rickettsii with spotted fever and typhus group rickett- 20 min, permeabilized in PBS with 1% Triton X-100, and then incubated with siae. J. Clin. Microbiol. 25, 167–171. primary antibodies followed by Alexa-conjugated secondary antibodies in Andrianantoandro, E., and Pollard, T.D. (2006). Mechanism of actin filament PBS with 2% bovine serum albumin (BSA) and 0.5% Triton X-100. Actin was turnover by severing and nucleation at different concentrations of ADF/cofilin. stained with 0.04 U/ml of Alexa-conjugated phalloidin (Invitrogen). Cells were Mol. Cell 24, 13–23. mounted with Prolong Gold antifade (Invitrogen). Balraj, P., El Karkouri, K., Vestris, G., Espinosa, L., Raoult, D., and Renesto, P. Images were captured using an Olympus IX71 microscope with a 1003 (2008). RickA expression is not sufficient to promote actin-based motility of (1.35 NA) PlanApo objective lens and a Photometrics CoolSNAP HQ camera. Rickettsia raoultii. PLoS ONE 3, e2582. 10.1371/journal.pone.0002582. Images were captured as 16-bit TIFF files using MetaMorph software (Molec- ular Devices), converted to 8-bit files using Adobe Photoshop, and the bright- Bishop, D.K., and Hinrichs, D.J. (1987). Adoptive transfer of immunity to Liste- ness/contrast levels were adjusted. Actin tail lengths were measured from the ria monocytogenes. The influence of in vitro stimulation on lymphocyte subset bacterial pole to the end of the tail for over 200 examples in at least 30 images requirements. J. Immunol. 139, 2005–2009. using ImageJ software. Tail lengths were compared pairwise to the control by Boujemaa-Paterski, R., Gouin, E., Hansen, G., Samarin, S., Le Clainche, C., the Student’s t test using Prism. Differences were considered to be significant Didry, D., Dehoux, P., Cossart, P., Kocks, C., Carlier, M.F., et al. (2001). Listeria if p < 0.05. Deconvolution images were taken using an Applied Precision protein ActA mimics WASp family proteins: it activates filament barbed end DeltaVision 4 Spectris microscope with a 1003 (1.4 NA) PlanApo objective branching by Arp2/3 complex. Biochemistry 40, 11390–11404. equipped with a Photometrics CH350 CCD camera. Images were captured Brieher, W.M., Coughlin, M., and Mitchison, T.J. (2004). Fascin-mediated pro- using SoftWoRx v3.3.6 software (Applied Precision) and were deconvolved pulsion of Listeria monocytogenes independent of frequent nucleation by the with Huygens Professional v3.1.0p0 software (Scientific Volume Imaging). Arp2/3 complex. J. Cell Biol. 165, 233–242. Cameron, L.A., Svitkina, T.M., Vignjevic, D., Theriot, J.A., and Borisy, G.G. Live-Cell Imaging (2001). Dendritic organization of actin comet tails. Curr. Biol. 11, 130–135. Cells were plated on glass-bottomed 35 mm dishes (MatTek). Actin was visu- Campellone, K.G., and Welch, M.D. (2010). A nucleator arms race: cellular alized by infecting S2R+ cells with Lifeact-GFP-expressing-Baculovirus at an control of actin assembly. Natl. Rev. 11, 237–251. moi of 10, treating with the dsRNA of the target of interest, and then infecting Carlier, M.F., and Pantaloni, D. (1997). Control of actin dynamics in cell motility. with R. parkeri for 48 hr. COS7 cells were cotransfected with the Lifeact-GFP J. Mol. Biol. 269, 459–467. plasmid and the relevant siRNA using Lipofectamine 2000 and then infected with R. parkeri for 48 hr, or incubated for 44 hr and infected with L. monocyto- Carlier, M.F., Laurent, V., Santolini, J., Melki, R., Didry, D., Xia, G.X., Hong, Y., genes for 4 hr. To visualize other GFP-tagged proteins, COS7 cells were trans- Chua, N.H., and Pantaloni, D. (1997). Actin depolymerizing factor (ADF/cofilin) fected with the relevant plasmid and then infected with R. parkeri for 24 hr, or enhances the rate of filament turnover: implication in actin-based motility. incubated for 20 hr and infected with L. monocytogenes for 4 hr. Image acqui- J. Cell Biol. 136, 1307–1322. sition was performed as described above, and movies were assembled using Delanote, V., Vandekerckhove, J., and Gettemans, J. (2005). Plastins: versatile ImageJ and Adobe Photoshop. The rate of movement was measured by modulators of actin organization in (patho)physiological cellular processes. manually tracking 15–30 individual moving bacteria for 1–3 min at 10 s intervals Acta Pharmacol. Sin. 26, 769–779. using the Manual Tracking plugin in ImageJ. The length of the tail associated Egile, C., Loisel, T.P., Laurent, V., Li, R., Pantaloni, D., Sansonetti, P.J., and with each moving bacterium was measured in each frame using ImageJ, Carlier, M.F. (1999). Activation of the CDC42 effector N-WASP by the Shigella and the mean tail length over the time interval was determined. Correlation flexneri IcsA protein promotes actin nucleation by Arp2/3 complex and bacte- coefficients were calculated by linear regression analysis using Prism. rial actin-based motility. J. Cell Biol. 146, 1319–1332. Field, C.M., and Alberts, B.M. (1995). Anillin, a contractile ring protein that SUPPLEMENTAL INFORMATION cycles from the nucleus to the cell cortex. J. Cell Biol. 131, 165–178. Geese, M., Schluter, K., Rothkegel, M., Jockusch, B.M., Wehland, J., and Supplemental Information includes two figures, one table, and eleven Sechi, A.S. (2000). Accumulation of profilin II at the surface of Listeria is movies and can be found with this article online at doi:10.1016/j.chom.2010. concomitant with the onset of motility and correlates with bacterial speed. 04.008. J. Cell Sci. 113, 1415–1426. Geese, M., Loureiro, J.J., Bear, J.E., Wehland, J., Gertler, F.B., and Sechi, A.S. ACKNOWLEDGMENTS (2002). Contribution of Ena/VASP proteins to intracellular motility of listeria requires phosphorylation and proline-rich core but not F-actin binding or multi- We thank members of the Welch lab for technical assistance and comments merization. Mol. Biol. Cell 13, 2383–2396. on the manuscript. We are grateful to R. Tjian, L. Cooley, C. Field, M. L. Gold- Giganti, A., Plastino, J., Janji, B., Van Troys, M., Lentz, D., Ampe, C., Sykes, C., berg, J. Zallen, M. Klymkowsky, T. Hackstadt, D. Portnoy, A. Sechi, T. and Friederich, E. (2005). Actin-filament cross-linking protein T-plastin Timmers, J. Cooper, H. G. Mannherz, T. Ohkawa, and C. Paddock for increases Arp2/3-mediated actin-based movement. J. Cell Sci. 118, 1255– providing strains or reagents. A.W.S. was supported by National Institutes of 1265. Health/National Institute of General Medical Sciences (NIH/NIGMS) Ruth L. Gouin, E., Gantelet, H., Egile, C., Lasa, I., Ohayon, H., Villiers, V., Gounon, P., Kirschstein National Research Service Award 5F32GM084359 and American Sansonetti, P.J., and Cossart, P. (1999). A comparative study of the actin- Heart Association Postdoctoral Fellowship 09POSY2160022. M.D.W. was based motilities of the pathogenic bacteria Listeria monocytogenes, Shigella supported by NIH/NIAID grant AI 074760. flexneri and Rickettsia conorii. J. Cell Sci. 112, 1697–1708. Gouin, E., Egile, C., Dehoux, P., Villiers, V., Adams, J., Gertler, F., Li, R., and Received: December 4, 2009 Cossart, P. (2004). The RickA protein of Rickettsia conorii activates the Revised: March 9, 2010 Arp2/3 complex. Nature 427, 457–461. Accepted: March 25, 2010 Published: May 19, 2010 Gouin, E., Welch, M.D., and Cossart, P. (2005). Actin-based motility of intracellular pathogens. Curr. Opin. Microbiol. 8, 35–45. REFERENCES Grenklo, S., Geese, M., Lindberg, U., Wehland, J., Karlsson, R., and Sechi, A.S. (2003). A crucial role for profilin-actin in the intracellular motility of Listeria Akin, O., and Mullins, R.D. (2008). Capping protein increases the rate of actin- monocytogenes. EMBO Rep. 4, 523–529. based motility by promoting filament nucleation by the Arp2/3 complex. Cell Hackstadt, T., Messer, R., Cieplak, W., and Peacock, M.G. (1992). Evidence 133, 841–851. for proteolytic cleavage of the 120-kilodalton outer membrane protein of

Cell Host & Microbe 7, 388–398, May 20, 2010 ª2010 Elsevier Inc. 397 Cell Host & Microbe Host Proteins Critical for Rickettsia Motility

rickettsiae: identification of an avirulent mutant deficient in processing. Infect. Rogers, S.L., Wiedemann, U., Stuurman, N., and Vale, R.D. (2003). Molecular Immun. 60, 159–165. requirements for actin-based lamella formation in Drosophila S2 cells. J. Cell Harlander, R.S., Way, M., Ren, Q., Howe, D., Grieshaber, S.S., and Heinzen, Biol. 162, 1079–1088. R.A. (2003). Effects of ectopically expressed neuronal Wiskott-Aldrich Rosenblatt, J., Agnew, B.J., Abe, H., Bamburg, J.R., and Mitchison, T.J. syndrome protein domains on Rickettsia rickettsii actin-based motility. Infect. (1997). Xenopus actin depolymerizing factor/cofilin (XAC) is responsible for Immun. 71, 1551–1556. the turnover of actin filaments in Listeria monocytogenes tails. J. Cell Biol. Heinzen, R.A. (2003). Rickettsial actin-based motility: behavior and involve- 136, 1323–1332. ment of cytoskeletal regulators. Ann. N Y Acad. Sci. 990, 535–547. Schafer, D.A., Welch, M.D., Machesky, L.M., Bridgman, P.C., Meyer, S.M., Heinzen, R.A., Hayes, S.F., Peacock, M.G., and Hackstadt, T. (1993). Direc- and Cooper, J.A. (1998). Visualization and molecular analysis of actin tional actin polymerization associated with spotted fever group Rickettsia assembly in living cells. J. Cell Biol. 143, 1919–1930. infection of Vero cells. Infect. Immun. 61, 1926–1935. Sechi, A.S., Wehland, J., and Small, J.V. (1997). The isolated comet tail pseu- Hudson, A.M., and Cooley, L. (2002). A subset of dynamic actin rearrange- dopodium of Listeria monocytogenes: a tail of two actin filament populations, ments in Drosophila requires the Arp2/3 complex. J. Cell Biol. 156, 677–687. long and axial and short and random. J. Cell Biol. 137, 155–167. Jeng, R.L., Goley, E.D., D’Alessio, J.A., Chaga, O.Y., Svitkina, T.M., Borisy, Skoble, J., Auerbuch, V., Goley, E.D., Welch, M.D., and Portnoy, D.A. (2001). G.G., Heinzen, R.A., and Welch, M.D. (2004). A Rickettsia WASP-like protein Pivotal role of VASP in Arp2/3 complex-mediated actin nucleation, actin activates the Arp2/3 complex and mediates actin-based motility. Cell. Micro- branch-formation, and Listeria monocytogenes motility. J. Cell Biol. 155, biol. 6, 761–769. 89–100. Jockusch, B.M., Murk, K., and Rothkegel, M. (2007). The profile of profilins. Rev. Physiol. Biochem. Pharmacol. 159, 131–149. Suzuki, T., Miki, H., Takenawa, T., and Sasakawa, C. (1998). Neural Wiskott- Aldrich syndrome protein is implicated in the actin-based motility of Shigella Kang, F., Laine, R.O., Bubb, M.R., Southwick, F.S., and Purich, D.L. (1997). flexneri. EMBO J. 17, 2767–2776. Profilin interacts with the Gly-Pro-Pro-Pro-Pro-Pro sequences of vasodi- lator-stimulated phosphoprotein (VASP): implications for actin-based Listeria Svitkina, T.M., and Borisy, G.G. (1999). Arp2/3 complex and actin depolyme- motility. Biochemistry 36, 8384–8392. rizing factor/cofilin in dendritic organization and treadmilling of actin filament Kleba, B., Clark, T.R., Lutter, E.I., Ellison, D.W., and Hackstadt, T. (2010). array in lamellipodia. J. Cell Biol. 145, 1009–1026. Disruption of the Rickettsia rickettsii Sca2 autotransporter inhibits actin based Teysseire, N., Chiche-Portiche, C., and Raoult, D. (1992). Intracellular move- motility. Infect. Immun. 78, 2240–2247. ments of Rickettsia conorii and R. typhi based on actin polymerization. Res. Loisel, T.P., Boujemaa, R., Pantaloni, D., and Carlier, M.F. (1999). Reconstitu- Microbiol. 143, 821–829. tion of actin-based motility of Listeria and Shigella using pure proteins. Nature Theriot, J.A., Mitchison, T.J., Tilney, L.G., and Portnoy, D.A. (1992). The rate of 401, 613–616. actin-based motility of intracellular Listeria monocytogenes equals the rate of Mannherz, H.G., Gonsior, S.M., Gremm, D., Wu, X., Pope, B.J., and Weeds, actin polymerization. Nature 357, 257–260. A.G. (2005). Activated cofilin colocalises with Arp2/3 complex in apoptotic Theriot, J.A., Rosenblatt, J., Portnoy, D.A., Goldschmidt-Clermont, P.J., and blebs during programmed cell death. Eur. J. Cell Biol. 84, 503–515. Mitchison, T.J. (1994). Involvement of profilin in the actin-based motility of Marchand, J.B., Moreau, P., Paoletti, A., Cossart, P., Carlier, M.F., and Panta- L. monocytogenes in cells and in cell-free extracts. Cell 76, 505–517. loni, D. (1995). Actin-based movement of Listeria monocytogenes: actin assembly results from the local maintenance of uncapped filament barbed Timmers, A.C., Niebel, A., Balague, C., and Dagkesamanskaya, A. (2002). ends at the bacterium surface. J. Cell Biol. 130, 331–343. Differential localisation of GFP fusions to cytoskeleton-binding proteins in animal, plant, and yeast cells. Protoplasma 220, 69–78. Martinez, J.J., and Cossart, P. (2004). Early signaling events involved in the entry of Rickettsia conorii into mammalian cells. J. Cell Sci. 117, 5097–5106. Van Kirk, L.S., Hayes, S.F., and Heinzen, R.A. (2000). Ultrastructure of Ogata, H., Audic, S., Renesto-Audiffren, P., Fournier, P.E., Barbe, V., Samson, Rickettsia rickettsii actin tails and localization of cytoskeletal proteins. Infect. D., Roux, V., Cossart, P., Weissenbach, J., Claverie, J.M., et al. (2001). Mech- Immun. 68, 4706–4713. anisms of evolution in Rickettsia conorii and R. prowazekii. Science 293, 2093– Van Troys, M., Huyck, L., Leyman, S., Dhaese, S., Vandekerkhove, J., and 2098. Ampe, C. (2008). Ins and outs of ADF/cofilin activity and regulation. Eur. Paddock, C.D., Sumner, J.W., Comer, J.A., Zaki, S.R., Goldsmith, C.S., God- J. Cell Biol. 87, 649–667. dard, J., McLellan, S.L., Tamminga, C.L., and Ohl, C.A. (2004). Rickettsia par- Walker, D.H., and Ismail, N. (2008). Emerging and re-emerging rickettsioses: keri: a newly recognized cause of spotted fever rickettsiosis in the United endothelial cell infection and early disease events. Nat. Rev. Microbiol. 6, States. Clin. Infect. Dis. 38, 805–811. 375–386. Paddock, C.D., Finley, R.W., Wright, C.S., Robinson, H.N., Schrodt, B.J., Welch, M.D., DePace, A.H., Verma, S., Iwamatsu, A., and Mitchison, T.J. Lane, C.C., Ekenna, O., Blass, M.A., Tamminga, C.L., Ohl, C.A., et al. (2008). (1997). The human Arp2/3 complex is composed of evolutionarily conserved Rickettsia parkeri rickettsiosis and its clinical distinction from Rocky Mountain subunits and is localized to cellular regions of dynamic actin filament spotted fever. Clin. Infect. Dis. 47, 1188–1196. assembly. J. Cell Biol. 138, 375–384. Policastro, P.F., Munderloh, U.G., Fischer, E.R., and Hackstadt, T. (1997). Rickettsia rickettsii growth and temperature-inducible protein expression in Yarmola, E.G., and Bubb, M.R. (2006). Profilin: emerging concepts and embryonic tick cell lines. J. Med. Microbiol. 46, 839–845. lingering misconceptions. Trends Biochem. Sci. 31, 197–205. Riedl, J., Crevenna, A.H., Kessenbrock, K., Yu, J.H., Neukirchen, D., Bista, M., Zallen, J.A., Cohen, Y., Hudson, A.M., Cooley, L., Wieschaus, E., and Schejter, Bradke, F., Jenne, D., Holak, T.A., Werb, Z., et al. (2008). Lifeact: a versatile E.D. (2002). SCAR is a primary regulator of Arp2/3-dependent morphological marker to visualize F-actin. Nat. Methods 5, 605–607. events in Drosophila. J. Cell Biol. 156, 689–701.

398 Cell Host & Microbe 7, 388–398, May 20, 2010 ª2010 Elsevier Inc.