1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Efficient uptake of Yersinia pseudotuberculosis into cultured mammalian cells is the result of high-affinity binding of invasin to β1 chain integrins. We demonstrate here that uptake requires Rac1 and Arp 2/3 function. Bacterial uptake was stimulated by GTPγS, but was inhibited in mammalian cells transfected with the interfering Rac1-N17 derivative. Rac1 was found to be activated in response to integrin engagement by invasin, whereas Rac1 and Arp 2/3 were found to be intensely localized around phagosomes bearing bacteria, indicating a specific role for Rac1 signalling from the nascent phagosome to downstream effectors. To determine whether the Arp 2/3 complex was a component of this proposed pathway, cells overproducing various derivatives of Scar1/WAVE1, an Arp 2/3-binding protein, were analysed. Sequestration of Arp 2/3 away from the phagocytic cup as a result of Scar1/WAVE1 overproduction dramatically inhibited uptake. To determine whether signalling from Rac1 to Arp 2/3 occurred via N-WASP, uptake was analysed in a cell line lacking expression of WASP and N-WASP. Uptake was unaffected by the absence of these proteins, indicating that β1 integrin signalling from Rac1 to Arp 2/3 can occur in the absence of N-WASP function.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Enteropathogenic Yersinia species cause localized infections in intestinal lymph nodes as well as fatal systemic diseases in mice. Rapid translocation of these microorganisms from the intestinal lumen across M cells overlaying Peyer's patches requires the bacterial outer membrane protein invasin (Pepe et al., 1995), which binds to several integrin heterodimers that contain the β1 chain (Isberg and Leong, 1990). Invasin-mediated uptake of Y. pseudotuberculosis by cultured cells is morphologically similar to zippering phagocytosis and is devoid of large-scale cell surface ruffling (Dersch and Isberg, 1999). This event requires high-affinity receptor–substrate binding, particularly when receptor concentration is limiting (Tran Van Nhieu and Isberg, 1993). Uptake also appears to involve receptor multimerization, as an Ig superfamily-like domain in invasin (Hamburger et al., 1999) that promotes homotypic interaction significantly enhances bacterial internalization (Dersch and Isberg, 1999).

Several pieces of evidence indicate that the integrin transmits signals required for bacterial uptake. Mutations in the cytoplasmic domain of the integrin β1 chain have a variety of phenotypes, ranging from stimulation to total elimination of uptake (Tran Van Nhieu et al., 1996). Furthermore, after receptor engagement, uptake is reduced by tyrosine kinase and lipid kinase inhibitors (Rosenshine et al., 1992), as well as by the Yersinia-encoded YopH tyrosine phosphatase (Black and Bliska, 1997; Persson et al., 1997) and the YopE RhoGap (Von Pawel-Rammingen et al., 2000).

Studies on bacterial uptake systems have provided valuable insight into the mechanisms of signal transduction and actin-based cellular remodelling promoted by members of the Rho family of small GTPases (Chen et al., 1996). Phagocytic cells have been shown to internalize opsonized particles through distinct RhoA- as well as Rac1- and Cdc42-dependent pathways (Cox et al., 1997; Caron and Hall, 1998). Cdc42 and Rac1 also facilitate the entry of Salmonella typhimurium into normally non-phagocytic eukaryotic cells. Uptake of this bacterium is regulated by translocation into mammalian cells of both a Rho family guanosine exchange factor (GEF) and a RhoGAP (Chen et al., 1996; Hardt et al., 1998; Fu and Galan, 1999). Similarly, deposition of the IpaC protein into the mammalian cell membrane by Shigella flexneri appears to be central to bacterial internalization mediated by multiple Rho family GTPases (Tran Van Nhieu et al., 1999). In each of these cases, downstream targets of the GTPases that play a critical role in uptake have not been identified.

Among the Rho family members, Cdc42 has been the most intensely studied in purified systems with regard to its role in promoting actin nucleation. GTP-bound Cdc42 binds directly to a region called the GTPase-binding domain (GBD) on the Wiscott–Aldrich syndrome protein family members WASP and N-WASP (Aspenstrom et al., 1996; Miki et al., 1998a). It has been demonstrated that binding by Cdc42 and 4,5-phosphoinositides allows the acidic C-terminus of WASP family members to recognize the Arp 2/3 complex (Rohatgi et al., 1999; 2000; Higgs and Pollard, 2000). The resulting N-WASP–Arp 2/3 complex is highly active at nucleating actin polymerization. As the GBD site of WASP family members potentially binds Rac1 (Aspenstrom et al., 1996), it is conceivable that Rac1 may function in a similar fashion (Miki et al., 1998b; Edwards et al., 1999; Machesky et al., 1999). However, three lines of investigation suggest potential alternative strategies for Rac1 to control actin dynamics. First, Rac1 is able to bind phosphatidylinositol-4-phosphate-5-kinase, which synthesizes lipids that can dissociate capping proteins from barbed actin ends (Hartwig et al., 1995). Secondly, Rac1 can promote inactivation of the depolymerizing activity of cofilin, via activation of PAK and LIM kinase (Arber et al., 1998; Yang et al., 1998; Edwards et al., 1999). Finally, recent work links Rac1 to activation of the WASP family member WAVE2 via an intermediate protein, IRSp53 (Miki et al., 2000).

In this report, we show a specific requirement for the Rho family GTPase Rac1 in β1 integrin-mediated uptake of Y. pseudotuberculosis. Rac1-promoted actin dynamics during bacterial uptake involves a pathway that includes Arp 2/3 but is able to proceed in the absence of N-WASP.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

GTPγS stimulates invasin-mediated uptake in a permeabilized cell assay

A permeabilized cell assay for invasin-mediated uptake was developed (Fig. 1A and B) to analyse cellular factors that participate in invasin-mediated uptake. Cells with bacteria bound on their surface were mechanically permeabilized to drain off cytosolic components. The permeabilized cells were then allowed to reseal in the presence of exogenous cytosol, and bacterial uptake was assayed by protection from kanamycin killing. The addition of buffer containing ATP and GTP in the absence of cytosol added to the permeabilized cells resulted in background levels of internalization (Fig. 1A; 0 µg of cytosolic extract). The addition of 4–15 µg of cytosol to the permeabilized cells resulted in maximal bacterial uptake (Fig. 1C). Examination of permeabilized cells by electron microscopy confirmed that the addition of cytosolic extract resulted in bacterial internalization (Fig. 1C), whereas in the absence of cytosol, the bound bacteria remained extracellular (compare Fig. 1B and C).


Figure 1. Invasin-mediated entry into permeabilized HEp-2 cells is dependent on the addition of cytosol and is stimulated by GTPγS.

A. Bacterial uptake (cfus) was determined as a function of cytosolic extract from human placenta and GTPγS added to the permeabilized cell mixture containing 0.5 mM ATP.

B. Electron micrograph showing HEp-2 cells with prebound Y. pseudotuberculosis (P, Inv+) after permeabilization and incubation in the presence of 0.2 mM ATP and 0.2 mM GTPγS without the addition of cytosolic extract (Experimental procedures).

C. HEp-2 cells treated as in (B) with cytosolic extract added.

D and E. Comparison of GTPγS with GTP stimulation of uptake in permeabilized cells.

D. Effect of buffer containing ATP alone (no cytosol), buffer with 1.0 µM GTPγS (no cytosol + GTPγS), cytosolic extract alone (cytosol) and cytosolic extract with 1.0 µM GTPγS (cytosol + GTPγS).

E. Effect of buffer containing ATP alone (no cytosol), buffer with GTP (no cytosol + 1.0 µM GTP), cytosolic extract alone (cytosol) and cytosolic extract with GTP (cytosol + 1.0 µM GTP).

The amount of bacterial uptake in each part is normalized to the addition of cytosolic extract alone (cytosol). Results shown are the average of three experiments (nine samples for each condition). Error bars represent SE calculated for an average of three experiments.

Download figure to PowerPoint

Maximal internalization occurred in the presence of the non-hydrolysable analogue of GTP, GTPγS (Fig. 1C and D). In the presence of cytosol, GTPγS addition resulted in 20- to 50-fold stimulation of uptake relative to its absence (Fig. 1A, compare 0 µM GTPγS with 1.6 µM GTPγS). Addition of GTP to the permeabilized cell mixture stimulated uptake (Fig. 1E, compare HPL with cytosol + GTP; P = 0.001), but the effect was not as strong as that observed for GTPγS (Fig. 1D and E; cytosol + GTPγS = 312% versus cytosol + GTP = 160%). Collectively, these results indicate that a GTP-binding protein plays an active role in invasin-mediated uptake.

Differential effects of Clostridium difficile ToxB and Clostridium botulinum C3 toxin on invasin-mediated uptake

Rho family GTPases were analysed next, as members of this family are involved in regulating a variety of cytoskeletal activities, and mutants that assume the GTP-bound form are constitutively activated, mimicking the effects that GTPγS could play in uptake (Hall, 1990). Furthermore, it has been demonstrated recently that the Yersinia cytotoxin YopE, which is able to inhibit phagocytosis, has a RhoGAP activity (Black and Bliska, 2000; Von Pawel-Rammingen et al., 2000). To analyse the role of Rho family members in invasin-promoted uptake, bacterial uptake was analysed in the presence of C. difficile toxin B, which inactivates many members of this family (Just et al., 1995).

COS-1 cells were treated with varying concentrations of purified C. difficile ToxB, challenged with Y. pseudotuberculosis (Inv+), and the effect of toxin on bacterial uptake was quantified by measuring viable counts protected from gentamicin (Experimental procedures). ToxB treatment had a profound effect on the ability of mammalian cells to internalize bacteria, with uptake reduced in a dose-dependent manner (Fig. 2A). This reduction was not caused by decreased bacterial binding to mammalian cells. Cells treated with moderate levels of toxin (Fig. 2B, 0.01–0.1 ng ml−1 ToxB) showed no defect in bacterial binding but were severely impaired for bacterial internalization. The difference in uptake levels did not appear to be the result of a general effect of ToxB on cellular morphology (Just et al., 1995). As with the effect of toxin on bacterial binding, dramatic changes in cell morphology were observed only at the highest concentration of toxin tested (Fig. 2A, 10 ng ml−1 ToxB).


Figure 2. Differential effects of toxins that disrupt Rho family member functions.

A and B. C. difficile ToxB reduces uptake. COS-1 cells were pretreated with varying concentrations of ToxB at 37°C for 90 min and challenged with Y. pseudotuberculosis. Uptake was monitored by gentamicin protection. COS-1 cell rounding and bacterial binding were assessed microscopically. %Uptake, %Association and %Rounding (morphological changes in response to toxin) are shown relative to untreated control cells. Error bars represent the SE from triplicate samples.

C and D. C3 toxin stimulates uptake. COS-1 cells were microinjected with C3 toxin in buffer containing FITC–dextran (molecular weight = 10 000) or buffer alone containing FITC–dextran (Buffer).

C. Uptake was assessed as the percentage of bacteria internalized by injected cells (n = 50 per coverslip).

D. Total number of bacteria internalized by cells microinjected with C3 toxin relative to cells injected with buffer alone. Error bars represent SE from triplicate samples.

E–H. Effect of C3 toxin (in FITC–dextran buffer) on stress fibre content as visualized by rhodamine phalloidin staining. NIH-3T3 fibroblasts injected with buffer alone (E); with C3 toxin (F). COS-1 cells injected with buffer (G) or with C3 toxin (H).

Download figure to PowerPoint

To begin to determine which Rho family members play a role in invasin-promoted uptake, internalization was measured in the presence of microinjected C. botulinum C3 transferase, which specifically inactivates RhoA, RhoB and RhoC. After a 1 h incubation following microinjection to allow toxin action, cells were challenged with Y. pseudotuberculosis and processed for immunofluorescence to determine the efficiency of bacterial uptake. Rho activity was not required for invasin-mediated uptake under these conditions, as cells injected with C3 transferase consistently internalized as many, or more, bacteria than cells injected with buffer (Fig. 2C). In addition, many more bacteria were associated with C3-injected cells than with buffer-injected control cells. As a result, the overall effect of C3 injection was to stimulate uptake ≈ 2.5-fold (Fig. 2D). Identical results were obtained with NIH3T3 cells treated with C3 toxin (data not shown).

As this result conflicts with previous observations regarding phagocytosis via another integrin, αmacβ2 (Caron and Hall, 1998), Rho function was assayed by analysing the presence of stress fibres in cells microinjected with C3 toxin to ascertain whether the toxin was active. In both COS-1 cells and NIH3T3 fibroblasts, microinjection of the toxin followed by a 60 min incubation resulted in complete loss of stress fibres (Figs 2F and H). Toxin treatment altered the elongated shape of NIH3T3 fibroblasts, and the tight bundles of stress fibres seen in untreated cells were eliminated in treated cells (Fig. 2E and F). In COS-1 cells, toxin treatment eliminated the few fine actin filaments seen in these cells (Fig. 2G and H). Therefore, sensitivity of Y. pseudotuberculosis uptake to ToxB cannot result from inactivation of RhoA, RhoB or RhoC.

The Rho family GTPase Rac1 is required for invasin-mediated uptake

In contrast to the above results with C3 toxin, interference with Rac1 function depressed uptake. Cells transfected with wild-type Rac1 (Rac1 WT), dominant interfering Rac1 (Rac1-N17) or a control vector not expressing Rac1 (vector) were examined by immunofluorescence for the relative ability to internalize Y. pseudotuberculosis. The presence of wild-type Rac1 or vector alone had no effect on the ability of cells to internalize bacteria (Fig. 3A). Overexpression of the dominant interfering Rac1-N17, on the other hand, had a dramatic effect on uptake: COS-1 cells overexpressing Rac1-N17 were able to internalize bacteria at < 25% of the efficiency of cells containing vector alone (Fig. 3A). The defect in bacterial uptake was not the result of decreased bacterial association with the target cells, as cells overexpressing wild-type Rac1 or Rac1-N17 had roughly the same number of associated bacteria as cells transfected with vector alone (Fig. 3B).


Figure 3. Role of Rac1 in invasin-mediated uptake.

A. Internalization of Y. pseudotuberculosis (P, Inv+) by transfected COS-1 cells was determined using immunofluorescence protection assay (Experimental procedures). Error bars represent the SE calculated for triplicate coverslips.

B. Bacterial binding was determined in the same immunofluorescence assay as (A), counting bacteria associated with cells expressing Rac1 from vectors. %Cell Associated refers to the amount of bacterial binding to cells relative to vector control.

C. Cdc42 interfering form does not affect invasin-mediated uptake. Entry of Y. pseudotuberculosis (P, Inv+) and S. typhimurium (SL1344) into transfected COS-1 was determined by immunofluorescence protection assay (Experimental procedures).%Uptake Relative to Vector refers to the fraction of cell-associated bacteria internalized by transfectants. Note that, in order to compare the uptake of S. typhimurium with that of Y. pseudotuberculosis (P, Inv+), %Uptake for Cdc42-N17 is expressed as a percentage relative to control cells (vector) of the respective bacterial strains. Error bars represent the SE calculated for triplicate coverslips.

D. Maximal stimulation of GTP loading of Rac1 occurs 5 min after the addition of invasin-coated beads. Latex beads coated with either MBP–Inv497 or MBP (Experimental procedures) were used to challenge Swiss-3T3 cells, and the amount of Rac1-GTP relative to the total Rac1 within the cell was determined.

E. Stimulation of GTP binding by Rac1 requires coating by invasin. Experiment was performed as in (D), comparing the relative amount of Rac1 activation by MBP–Inv497 with MBP at 5 min after infection. Data are the mean of three independent experiments ± SE.

Download figure to PowerPoint

In contrast to the results with Rac1, an interfering form of Cdc42 (Cdc42-N17) had no affect on invasin-promoted uptake (Fig. 3C). In order to demonstrate that the Cdc42-N17 construct had interfering activity, the effect of overexpression of this isoform on S. typhimurium entry into COS-1 cells was tested. As reported previously (Chen et al., 1996), there was strong inhibition of S. typhimurium uptake compared with vector alone (Fig. 3C), indicating that the mutant had inhibitory capabilities.

If Rac1 is required for invasin-promoted uptake, then binding of invasin-coated beads should result in an increase in GTP-bound Rac1 within the cell. To measure Rac1 activation in response to receptor engagement by invasin, beads coated with either an invasin–maltose-binding protein fusion (Inv–MBP497) or control maltose-binding protein (MBP) were used to challenge cells, and the amount of GTP-loaded Rac1 was measured (Experimental procedures). Three minutes after the addition of invasin-coated beads, a significant increase in GTP-bound Rac1 could be observed, with maximal activation occurring 5 min after the addition of beads (Fig. 3D). At maximal activation, incubation with Inv–MBP497 beads resulted in 7.2% of the Rac1 being in the GTP-bound state, compared with 2.2% if cells were incubated with control MBP beads (Fig. 3E).

Consistent with the activation data, overexpression of the constitutively GTP-bound form, Rac1-V12, showed a roughly threefold increase in total uptake relative to transfectants harbouring the empty vector (Fig. 4A). This increase was not the result of an increase in the fraction of cell-associated bacteria that were internalized, as 80% of the cell-associated bacteria were internalized by cells harbouring either RacV12 or empty vector (Fig. 4B). Instead, the total number of bacteria associated with cells expressing Rac1-V12 was much higher relative to the vector control (Fig. 4C), because of either an increased rate of internalization or increased binding.


Figure 4. Overexpression of an activated form of Rac1 (Rac1-V12) increases the amount of invasin-mediated uptake. Transfected COS-1 cells were challenged with Y. pseudotuberculosis (P, Inv+), and bacterial uptake and cell association were determined by the immunofluorescence protection assay (Experimental procedures).

A. Total uptake refers to the number of bacteria found within control (vector) or activated Rac1 (Rac1-V12) transfected cells (expressed as the number of bacteria internalized per 100 transfectants).

B. Internalized/Bound refers to the fraction of the total cell-associated bacteria internalized by either control (vector) or activated Rac1 (Rac1-V12) transfected cells.

C. %Cell Associated refers to the total number of bacteria (inside and outside) associated with activated Rac1 (Rac1-V12) transfected cells relative to control (vector) transfected cells.

Error bars represent the SE of triplicate coverslips.

Download figure to PowerPoint

Rac1 co-localization with nascent phagosomes

Transduction of a Rac1-dependent uptake signal predicts that this protein should be recruited to the nascent phagosome during active phagocytosis. To investigate this possibility, transfectants expressing Rac1 were challenged with Y. pseudotuberculosis and analysed by immunofluorescence microscopy, probing with anti-Y. pseudotuberculosis before permeabilization and anti-Rac1 after permeabilization. In this fashion, bacteria that were external to the cell or in the process of being internalized could be analysed exclusively (Fig. 5A–C). Almost all the bacteria that were being actively internalized, as determined by partial protection from anti-Y. pseudotuberculosis(Fig. 5A–C, arrows), were found to have intense Rac1 staining about the nascent phagosomes (148 out of 151 nascent phagosomes observed). For such phagosomes, the internalized portion of the bacterium could readily be detected by outlines of staining for Rac1 in the absence of staining for Y. pseudotuberculosis(Fig. 5B). When both fully and partially internalized bacteria were probed, 94% of the phagosomes had Rac1 around their surfaces (n = 516 phagosomes observed during a 20 min infection period; Fig. 5D–F).


Figure 5. High-efficiency co-localization of Rac1 and lower efficiency co-localization of Cdc42 with phagosomes bearing Y. pseudotuberculosis. COS-1 cells transfected with plasmids expressing either Rac1-HA or Cdc42 were infected with bacteria, fixed and processed for immunofluorescence microscopy. Samples were probed with anti-Y. pseudotuberculosis either before or after permeabilization (Experimental procedures).

Extracellular: staining of Y. pseudotuberculosis before permeabilization of mammalian cells, revealing only portions of bacteria that have not been internalized.

Total: bacteria stained after permeabilization of mammalian cells, allowing visualization of all cell-associated bacteria.

p21: staining with antibody directed against either Cdc42 or Rac1 after permeabilization of mammalian cells.

A–C. Rac1 association with extracellular bacteria. Arrows indicate a nascent phagosome forming around a bacterium during entry.

D–F. Rac1 association with total (extracellular and internalized) bacteria.

G–I. Cdc42 association with extracellular bacteria. Arrow indicates bacterium in the process of entry; part of the bacterium within the nascent phagosome shows staining with the anti-Cdc42 antibody.

J–L. Cdc42 association with total bacteria. Arrow indicates the only phagosome that clearly stains with anti-Cdc42.

Download figure to PowerPoint

Phagosomes harbouring Y. pseudotuberculosis occasionally co-localized with Cdc42, but the efficiency of co-localization with partially formed phagosomes was lower than that observed for Rac1 (Fig. 5H and I, arrows). The frequency of Cdc42 association with fully formed phagosomes was also lower (Fig. 5J–L), as cells with five or more associated bacteria typically showed only one phagosome with Cdc42 co-localization (Fig. 5K and L, arrow). Twenty minutes after uptake, only 13% of the phagosomes stained positively for Cdc42 (n = 159 phagosomes). These results are consistent with the lack of a strong effect of the dominant interfering form of Cdc42 (Fig. 3C).

Involvement of Arp 2/3 in invasin-promoted uptake

Given the demonstrated importance of Rac1 in promoting invasin-dependent uptake, we wanted to investigate the role of potential downstream effectors that could regulate actin dynamics. Specifically, the role of Arp 2/3 in initiating actin nucleation was investigated. For this reason, uptake was measured in the presence of inhibited Arp 2/3 activity. Overexpression of the WASP family member Scar1/WAVE1 (SPWA construction; Fig. 6) or derivatives containing its C-terminal acidic domain (A domain-containing construction; Fig. 6) have been documented to interfere with a number of Arp 2/3-dependent events within cells (Machesky and Insall, 1998). In vitro, the acidic (A) domain of Scar1/WAVE1 binds Arp 2/3 (Machesky et al., 1999). If Arp 2/3 is required for Y. pseudotuberculosis uptake, then overproduction of Scar1/WAVE derivatives should interfere with internalization.


Figure 6. Overproduction of Scar1 with an intact acidic domain inhibits invasin-promoted uptake.

A and B. COS-1 cells transfected with plasmids encoding intact Scar1 (SPWA), a derivative lacking the acidic domain (SPW) or a derivative with only the W domain (W) were challenged with bacteria. Uptake was quantified for bacteria bound to the transfectants, as well for those bound to the untransfected (UN) cells found on the identical coverslips as the transient transfectants, based on immunofluorescence. Total uptake refers to the number of bacteria found within cells, expressed as the number of bacteria internalized per 100 transfectants. The experiments were performed on separate days.

C. The majority of transfectants harbouring Scar1 had no internalized bacteria. Bacterial uptake was determined for 210 Scar1 transfectants or untransfected cells on three identical coverslips. Uptake is displayed as the number of individual COS-1 cells having the indicated number of internalized bacteria.

Download figure to PowerPoint

Invasin-promoted uptake of Y. pseudotuberculosis was examined in COS-1 cells transiently transfected with various Scar1/WAVE1 derivatives (Experimental procedures). Overproduction of full-length Scar1/WAVE1 (SPWA) greatly depressed bacterial uptake relative to untransfected cells on the same coverslip (SPWA; Fig. 6A). This reduction in uptake efficiency was relieved in transfectants that harboured two different Scar1/WAVE1 derivatives either lacking the acidic domain (SPW; Fig. 6A) or with only the actin-binding domain (W; Fig. 6B). Although the truncation derivative lacking just the acidic region caused a mild reduction in uptake, this resulted from inhibition of bacterial association with the target cells, which was ≈ 70% of that observed for untransfected cells (data not shown). As a result, the fraction of bound bacteria that was internalized in cells harbouring SPW was identical to that seen with untransfected cells.

The significant depression of uptake caused by full-length Scar1/WAVE1 can be visualized by analysing a histogram of 210 cells with bound bacteria. In a typical experiment, 85% of transfectants expressing Scar1/WAVE1 had no internalized bacteria. In contrast, only 14% of untransfected cells on the same coverslip showed no internalized cell-associated bacteria (Fig. 6C). As the Scar1/WAVE1 derivative containing only the actin-binding domain (W) did not inhibit uptake (Fig. 6B), the effects caused by the intact Scar1/WAVE1 cannot result from titration of actin. Rather, removal of Arp 2/3 from the site of bacterial binding is the most plausible cause of this depression in uptake.

Association of Arp 2/3 with nascent phagosomes is influenced by Rac1 and Scar1/WAVE1

The simplest inference from the above results is that overproduction of Scar1/WAVE1 causes sequestration of the Arp 2/3 complex away from the bound bacteria, precluding its participation in invasin-promoted uptake. To test this hypothesis, the cellular distribution of Arp 2/3 was examined using antibodies directed against the 50 kDa Arp3 subunit of the Arp 2/3 complex (Egile et al., 1999). Arp 2/3 was localized to the perinuclear region as well as to the edges of the cell (Fig. 7B, C, E and F). Furthermore, distinct punctate aggregation of Arp 2/3 could be observed adjacent to extracellularly located Y. pseudotuberculosis(Fig. 7A–C). In cells probed for both internalized and extracellular bacteria (total; Fig. 7D), Arp 2/3 was found to be intimately localized around almost all the fully formed phagosomes (Fig. 7D–F). Most strikingly, bacteria captured in the act of entering into COS-1 cells showed dense Arp 2/3 localization around the nascent phagosomes (Fig. 7G, montage of entering bacteria).


Figure 7. Arp 2/3 recruitment during uptake of Y. pseudotuberculosis. COS-1 monolayers infected with bacteria were fixed, immunoprobed for bacteria either before or after, permeabilization and stained for the Arp 3 subunit of the Arp 2/3 complex (Experimental procedures). Small arrows indicate bacteria with co-localized Arp 3, and large arrowheads indicate Arp 3 localization at the edges of mammalian cells.

A–C. Localization of Arp 3 in a COS-1 cell with externally bound bacteria.

D–F. Arp 3 (green) co-localization with internalized bacteria, demonstrated using differential staining of surface-bound (red) and phagocytosed (blue) bacteria (see Experimental procedures).

G. Images of Y. pseudotuberculosis partially engulfed by COS-1 cells. Bacteria (red) were probed before permeabilization, and Arp 3 was probed after permeabilization (green).

Download figure to PowerPoint

Arp 2/3 localization was next examined in the presence or absence of overproduced Scar1/WAVE1. Transfectants expressing full-length Scar1/WAVE1 (SPWA) displayed a dramatic redistribution of Arp 2/3 compared with untransfected cells. Transfected cells showed dense, perinuclear aggregation of Arp 2/3, with little staining observed at the edges of cells (Fig. 8A–C). Untransfected cells and cells expressing the actin-binding Scar1-W fragment retained the wild-type distribution of Arp 2/3, with both Scar1-W and Arp 2/3 localized around the bacteria (Fig. 8D–F). These results were observed in both the presence and the absence of added bacteria or in uninfected and infected cells found on the same coverslips.


Figure 8. Arp 2/3 recruitment during uptake of Y. pseudotuberculosis is inhibited by Scar1/WAVE1 and Rac1-N17,

A–C. Tight perinuclear sequestration of Arp 2/3 in the presence of full-length Scar1/WAVE1.

D–F. Lack of Arp 2/3 aggregation around the nucleus in the presence of the Scar1-W fragment. The arrow indicates an internalized bacterium.

G. Association of Arp 2/3 with extracellularly associated Y. pseudotuberculosis. Bacteria were incubated with transfected COS-1 cells for 20 min. Fifty extracellular bacteria from transfected and untransfected cells were examined for each coverslip. Data are the mean with SE of three coverslips (Experimental procedures).

Download figure to PowerPoint

The consequence of Scar1/WAVE1 overproduction on Arp 2/3 co-localization around nascent phagosomes was assessed quantitatively by probing Arp 2/3 localization around extracellular, bound bacteria. Extracellular bacteria were analysed exclusively because almost all the fully internalized phagosomes had dense Arp 2/3 staining after a 20 min infection. As the uptake efficiency of Scar1/WAVE1 transfectants was greatly diminished relative to controls (Fig. 6), inclusion of the internalized bacterial population in the data set would exaggerate the magnitude of inhibition of Arp 2/3 co-localization caused by Scar1/WAVE1 overproduction. Transfectants overproducing full-length Scar1/WAVE1 (SPWA) were significantly depressed for Arp 2/3 co-localization about extracellular bacteria relative to both Scar1/WAVE1-W transfectants and the untransfected controls (Fig. 8G). The magnitude of the co-localization defect was similar to that observed for transfectants harbouring the dominant interfering Rac1-N17 derivative (Fig. 8G). These results indicate that Scar1/WAVE1 overproduction and the presence of an interfering Rac1 derivative resulted in similar blockades of Arp 2/3 localization. Therefore, recruitment of Arp 2/3 appears to be tightly linked to an integrin-promoted bacterial internalization pathway that involves Rac1 signalling.

N-WASP is not required for invasin-promoted uptake

Activation of Arp 2/3 by Cdc42 in adherent cell lines requires N-WASP function (Miki et al., 1998a; Rohatgi et al., 1999). As Rac1 and Arp 2/3 appear to be central players in invasin-promoted uptake, and Rac1 has been reported to bind to the WASP protein (Aspenstrom et al., 1996), the potential role of N-WASP in linking Rac1 and Arp 2/3 was investigated. Therefore, uptake was analysed in a fibroblast cell line derived from murine embryos that are devoid of WASP and N-WASP expression because of a germline deletion of the N-WASP exon 2 by homologous recombination (Snapper et al., 2001). A variety of N-WASP-dependent activities that require Arp 2/3 function is demonstrably defective in this cell line (Snapper et al., 2001), consistent with the lack of expression of both N-WASP and WASP. When compared with a fibroblast cell line derived from wild-type mouse embryos, invasin-promoted uptake was at least as efficient in the N-WASP mutant line as in the wild-type line (Fig. 9). This result is consistent with a model in which Rac1-dependent phagocytic uptake promoted by Arp 2/3 bypasses N-WASP function.


Figure 9. N-WASP is not required for invasin-promoted uptake. Y. pseudotuberculosis was incubated at an MOI = 20 in the absence of centrifugation with either immortalized fibroblasts having an intact N-WASP gene (N-WASP+) or a truncated N-WASP gene that fails to express any detectable antigen (N-WASP). After either 40 or 80 min of incubation, the coverslips were washed, fixed and probed by immunofluorescence microscopy for internalized bacteria. Data are the results of triplicate coverslips in which 100 infected cells were observed for each coverslip. Total uptake is the number of bacteria that were internalized per 100 cells observed. The means ± SE are shown.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

There are two salient conclusions to this work. First, in contrast to previously proposed models, we have identified conditions in which there was a singular requirement for Rac1 in promoting an integrin-dependent phagocytic event. Secondly, this system has allowed us to demonstrate that Arp 2/3 activation by Rac1 occurs independently of N-WASP function, indicating that Rac1 must induce cytoskeletal rearrangements by some pathway distinct from that described for Cdc42.

It has been argued that there are two types of phagocytosis, one that requires Cdc42 and Rac1, and another that requires Rho (Caron and Hall, 1998). Our results indicate, however, a greater level of diversity with respect to Rho family requirements. Furthermore, although studies have demonstrated that Rho plays a positive role in phagocytosis promoted by the integrin receptor αmacβ2 (Caron and Hall, 1998), this was not observed in our studies involving β1 integrin receptors, which pointed towards a potential negative role for Rho. The notion that Rho could have an inhibitory role (Fig. 2) is consistent with data indicating that the activated form of RhoA, RhoA-N14, strongly inhibits uptake (Black and Bliska, 2000). This result is also reminiscent of work indicating opposing roles for Rac and Rho in neurite extension (Leeuwen et al., 1997). Rho-promoted stress fibre formation in cultured cell lines may interfere with the mobilization of the integrins around the bound bacterium or with the recruitment of downstream cytoskeletal factors critical for invasin-promoted uptake. Based on these results, the Rho family requirement for phagocytosis by integrins depends on the particular receptor involved. Given this diversity, it is also likely that Rho family requirements for promoting uptake could depend significantly on the cell type being analysed.

The distinct and intense association of Rac1 with nascent phagosomes during Y. pseudotuberculosis internalization indicates that recruitment of Rac1 occurs early in the process. The association with phagosomes in the process of actively internalizing Y. pseudotuberculosis argues against this localization being fortuitous and indicates that Rac1 is involved in a dynamic process critical for the subsequent movement and closure of the phagocytic cup. Although the signal from the clustered integrin receptors that elicits Rac1 co-localization is unknown, promising candidates include products of phosphoinositol-3-kinase (PI-3-kinase). Inhibitors of PI-3-kinase interfere with invasin-promoted uptake (Mecsas et al., 1998); this inhibition can be significantly overcome by the presence of the activated Rac1-V12 form (M. Alrutz and R. Isberg, unpublished observations), suggesting that Rac1 is located downstream from the PI-3-kinase signal.

There was less pronounced, but detectable, association of Cdc42 with the phagocytic cup. We have no evidence that this association is functionally important, but it is conceivable that Cdc42 plays a supportive or back-up role in invasin-mediated uptake. In this manner, Cdc42-regulated actin dynamics could enhance the rate of closure of individual phagosomes or stabilize structures necessary to accomplish internalization. Such subtle contributions may not be detectable using the assay systems described here. Alternatively, Cdc42 may be responsible for the residual uptake seen in cells transfected with the interfering Rac1-N17.

The Arp 2/3 complex appears to play an important role in invasin-mediated uptake. Overproduction of the WASP family member Scar1/WAVE1 strongly inhibited uptake, which is presumably the result of sequestration of a critical uptake factor away from the surface-bound bacteria. Our observations are in agreement with the recent study reporting localization of Arp 2/3 around mature phagosomes containing C3- and Ig-coated particles (May et al., 2000). However, as this previous study did not distinguish extracellularly bound particles from intracellular phagosomes, it is not clear whether Arp 2/3 recruitment occurred at the start or completion of phagocytosis. We have shown that, for invasin-mediated bacterial uptake, Arp 2/3 is clearly recruited to phagosomes early in the process and before complete closure (Fig. 7G). A large fraction of extracellularly associated bacteria have Arp 2/3 localized at the site of bacterial binding (Fig. 8G), arguing that recruitment of Arp 2/3 is coincident with phagocytosis. That transfection of Scar1/WAVE1 lowers the recruitment of Arp 2/3 around bound bacteria, and a truncation mutant lacking the acidic tail is significantly relieved of this uptake inhibition, indicates that the uptake defect is caused by lack of availability of Arp 2/3. Inhibition of uptake cannot be attributed to depletion of the soluble actin pool, because overproduction of the actin-binding W domain of Scar1/WAVE1 has no effect on bacterial uptake.

Actin nucleation and branch formation induced in vitro by the Arp 2/3 complex requires activation by WASP, N-WASP or Scar1, all of which have similar carboxyl-terminal structures (Machesky et al., 1999; Rohatgi et al., 1999; Yarar et al., 1999). N-WASP activity requires binding by Cdc42 and phosphoinositol-4,5-phosphate (Egile et al., 1999; Rohatgi et al., 2000), providing one explanation as to how Arp 2/3 activity is stimulated by Cdc42. Based on previous work, it was conceivable that Rac1 could use a similar activation strategy, as the GBD on WASP binds Rac1 (Aspenstrom et al., 1996). Our results demonstrating normal uptake efficiency in an N-WASP null cell line, however, argues against Rac1 working in the same fashion as Cdc42 to promote Arp 2/3 activation. If binding to N-WASP is the primary strategy used by Cdc42 to influence actin dynamics, then the internalization proficiency of the N-WASP null cell line is consistent with our model in which Cdc42 plays, at best, a minor role in invasin-mediated uptake by immortalized cells.

Rac1 may recruit activated Arp 2/3 in an as yet uncharacterized fashion. It has been suggested recently that Rac1 is able to activate WAVE2 by binding to a linking protein, IRSp53 (Miki et al., 2000). This, in turn, may stimulate Arp 2/3-mediated actin nucleation. Our data do not directly address this model, as the overproduced wild-type form of Scar1/WAVE1 inhibits, rather than stimulates, invasin-mediated internalization. Perhaps this overproduced protein interferes with the predominant isoform in these cells, which might be WAVE2 (Miki et al., 2000). As an alternative activation strategy, Rac1 could recruit phosphoinositol-4-phosphate-5-kinases (PIP-5-kinases) to the phagocytic cup. Binding of PIP-5-kinases to Rac1 governs actin dynamics by promoting uncapping of actin filaments (Barkalow et al., 1996) and producing lipids that activate N-WASP (Higgs and Pollard, 2000; Rohatgi et al., 2000). It is also conceivable that PIP-5-kinases may activate other WASP family members, such as Scar/WAVE, in response to GTP-bound Rac1. Further work on invasin-mediated uptake should shed light on these signalling events and on how Rac1 is able to control actin dynamics in response to integrin clustering.

Experimental procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Bacterial and mammalian cell culture

Escherichia coli strain MC4100 (pRI203) (Leong et al., 1995), expressing invasin, was grown in L broth overnight at 28°C. Y. pseudotuberculosis strain 137 (P, Inv+), an uptake-competent strain cured of the virulence plasmid, was grown from a dilution of an overnight culture to an A600 of 1.0 at 28°C in L broth with shaking. S. typhimurium strain SL1344 was grown under conditions that stimulate the type III protein secretion system (Chen et al., 1996).

HEp-2 cells, COS-1 cells and NIH 3T3 cells were grown as described previously (Alrutz and Isberg, 1998). N-WASP-deficient and control cell lines were derived by infecting mouse embryonic fibroblasts from N-WASP-deficient and wild type litter mate control embryos, isolated at day 9.5, with an SV40 large T antigen expressing retrovirus, as described previously (Snapper et al., 2001).

Permeabilized cell assay for bacterial internalization

HEp-2 cells were grown to confluency in a 10 cm dish. The cells were incubated with 1 × 109E. coli strain MC4100 (pRI203) for 1 h at 19°C, washed five times with ice-cold PBS, 0.2 mM Ca2+, 1 mM Mg2+ and scraped into a microfuge tube in a volume of 1 ml. The cells were then pelleted at top speed in a microfuge for 5 min, the supernatant drawn off and the cells resuspended in 1 ml of PBS, 0.2 mM Ca2+, 1 mM Mg2+. Cells were washed four times and finally resuspended in 1 ml of buffer.

The permeabilized cells bound to bacteria (10 µl) were incubated on ice for 30 min in a 96-well microtitre plate with 0.2 mM ATP and varying concentrations of cytosolic extract (cytosol), GTPγS (5′-O-(3-thiotriphosphate); Sigma) or GTP (Sigma) in a total reaction volume of 22 µl. Prewarmed cell culture medium was added to the permeabilized cell mixture, and the plate was then shifted to 37°C for 30 min. The plates were spun for 10 min at 1500 g and the supernatant removed.

Bacterial internalization was determined by kanamycin protection. The pelleted cell mixture was incubated with 100 µl of RPMI, 5% calf serum containing kanamycin (100 µg ml−1) for 45 min at 37°C. The antibiotic was neutralized by 30 min incubation at room temperature in the presence of 10 µl of an S100 extract from DH5α (pHP45ΩKm), prepared essentially as described previously (Tran Van Nhieu and Collatz, 1987), to the reaction mix. The mammalian cells were lysed by the addition of 0.1% Triton X-100, and bacteria were plated onto standard bacteriological media to titre for viable counts.

For cytosolic extract, human placenta (50 g) was homogenized in a Waring blender in a 50 ml volume of ice-cold PBS containing 200 µM phenylmethylsulphonyl fluoride (PMSF) and pelleted at 27 000 g for 30 min at 4°C. The supernatant was removed and dialysed against 100 volumes of PBS at 4°C changing buffer three times to remove the PMSF. After ≈ 18 h of dialysis, the extract was poured over a PD-10 (Pharmacia) desalting column equilibrated with PBS. Protein concentration of the extract was determined by Bradford assay (Bio-Rad).

Internalization assay for ToxB-treated COS-1 cells

COS-1 cells were seeded at a density of 1 × 104 cells per well of a 96-well plate. The next day, the cells were washed and incubated in fresh medium and treated with various concentrations of C. difficile toxin B (a kind gift from Dr T. LaMont) for 90 min at 37°C. After incubation with the toxin, the cells were washed twice with HBS2+ (25 mM HEPES, pH 7.0, 150 mM NaCl, 0.5 mM Ca2+, 1 mM Mg2+) and placed in fresh, prewarmed media. Bacterial internalization was determined by gentamicin protection as described previously (Alrutz and Isberg, 1998).

Ectopic expression of GTPases and Scar derivatives in COS-1 cells

Wild-type Rac1, interfering (N17) and activated (V12) Rac1 and Cdc42 expression vectors that co-express green fluorescent protein (GFP) in dicistronic vectors were kind gifts from J. Galan, as was HA-tagged Rac1 (Chen et al., 1996). Scar derivatives were gifts from Dr L. Machesky (Machesky et al., 1999). For the transfection, 1 × 104 COS-1 cells were seeded onto glass coverslips, and the next day were transfected with 1 µg of purified plasmid DNA, with LipoFectamine Plus according to the manufacturer's protocols (Gibco).

Immunofluorescence protection assay of bacterial uptake

Forty-eight hours after transfection, the cells were challenged at an MOI of ≈ 10 bacteria per cell with Y. pseudotuberculosis 137 (P, Inv+). S. typhimurium infection and assessment of uptake was essentially as described previously (Chen et al., 1996). The cells were incubated for 80 min at 37°C, washed three times with PBS and fixed with 3% paraformaldehyde in PBS at room temperature for 20 min. The coverslips were blocked with PBS + 4% goat serum for 1 h at room temperature, washed with PBS and probed with a 1:1000 diluted rabbit anti-Y. pseudotuberculosis serogroup III serum (a gift from C. Krishnan, Ontario Ministry of Health) for 1 h, followed by washing and probing with anti-rabbit IgG–tetramethyl rhodamine isothiocyanate (TRITC) (1:500 dilution; Boehringer Mannheim) for 1 h at room temperature to detect bacteria bound on the mammalian cell surface. Cells were permeabilized with cold methanol (−20°C) for 10 s, washed and blocked with PBS + 4% goat serum for 1 h at room temperature. The coverslips were then probed again with anti-Y. pseudotuberculosis serum for 1 h at room temperature, washed and probed with goat anti-rabbit IgG–cascade blue (1:500; Molecular Probes) to reveal total cell-associated bacteria.

For most immunofluorescence studies, the percentage uptake was expressed as the number of bound bacteria that resisted staining with TRITC (internalized) relative to the number of cascade blue-staining bacteria (total bound). The percentage uptake of bacteria in cells overexpressing Cdc42 isoforms was calculated as above, but the data are expressed normalized to cells expressing vector alone to be consistent with the presentation of previously published data on the effects of Cdc42 overexpression on S. typhimurium uptake (Chen et al., 1996). The percentage bacterial association represents the total number of cell-associated bacteria normalized to the control (vector transfected or control microinjected).

Assessment of Rac1 activation

The amount of GTP-bound Rac in cells after binding of either invasin-coated latex beads or control beads was determined as described previously (del Pozo et al., 2000). Preparations (100 µg ml−1) of the MBP–invasin fusion MBP–Inv497 or control MBP were used to coat 1 µM latex beads (Sigma; Dersch and Isberg, 1999). Approximately 2 × 108 beads were added in 1 ml of RPMI medium at 37°C to a 100 mm dish supporting 2–3 × 106 Swiss 3T3 cells that had been starved in the absence of FBS for 24 h. At various times after the addition of the beads, the cells were scraped with a rubber policeman in the presence of 1 ml of lysis buffer (50 mM Tris-HCl, pH 7.2, 10 mM MgCl2, 500 mM NaCl, 1%Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM AEBSF, 10 µg ml−1 leupeptin, 10 µg ml−1 aprotinin), harvested by centrifugation and incubated with glutathione S-transferase (GST)–PBD fusion protein (del Pozo et al., 2000) bound to 35 µl of glutathione Sepharose beads. The amount of GTP-bound Rac1 associated with the beads was quantified as described previously (del Pozo et al., 2000).

Co-localization of mammalian proteins with nascent phagosomes

To assess the effects of Scar1 and Rac1 derivatives on Arp 2/3 recruitment to externally bound bacteria, COS-1 cells were transfected with plasmids bearing wild-type Rac1, the dominant-negative mutant Rac1-N17 as well as full-length Myc-Scar1-SPWA and Myc-Scar1-W fragments (kind gifts from Dr L. Machesky). Twenty-four hours after transfection, cells were challenged with Y. pseudotuberculosis YP137 (P Inv+) for 20 min at an MOI of 150. Cells were washed, fixed and probed for external bacteria, as above, before permeabilization and probing for Arp3 (anti-Arp3 peptide antibody, 1:200 dilution, kindly provided by C. Egile; Egile et al., 1999). Scar1 derivatives [anti-Myc 9E10 monoclonal antibody (mAb), 1:250 dilution; Santa Cruz Biotechnology] or Rac1 derivatives (anti-Rac1 mAb, 1:250 dilution; Transduction Labs) were probed simultaneously with anti-Arp3.

Microinjection of mammalian cells

Cells were seeded at a density of 4 × 104 per coverslip in 24-well dishes. On the next day, the cells were washed three times with DME (no serum), and the coverslip was transferred to a 60 mm dish for microinjection. The cells were injected with C3 toxin (30 µg ml−1; Biomol) in microinjection buffer [PBS, 0.6 mg ml−1 lysine-fixable dextran fluorescein isothiocyanate (FITC) conjugate, molecular weight 10 000; Molecular Probes] or microinjection buffer alone. An Eppendorf model 5246 injector was used in conjunction with a TE300 inverted microscope (Nikon). The injection parameters were: Pi, 150 hPa; ti, 0.3 s. Cells were maintained on a 37°C-heated platform during injection and monitored using Hoffman modulation contrast optics with a 40× objective.

After microinjection, the cells were incubated for 60 min at 37°C and challenged with Y. pseudotuberculosis 137 (P, Inv+) at an MOI of 20 bacteria per mammalian cell. The cells were incubated with bacteria for 1 h and 20 min at 37°C, 5% CO2, washed, fixed and processed for the immunofluorescent determination of bacterial uptake by FITC-positive cells, using the fluorescence microscopy assay described above. For stress fibre determination after microinjection, cells were fixed and probed with rhodamine-conjugated phalloidin (Molecular Probes) according to the manufacturer's instructions.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

We thank Dr Guy Tran Van Nhieu for developing the permeabilized cell assay, Drs Ira Herman, Guillaume Duménil and Jonathan Solomon for review of the text, Dr Thomas LaMont for the gift of purified ToxB, Drs Harry Higgs and Coumaran Egile for antibodies directed against Arp 3, and Drs Li Mei Chen and Laura Machesky for advice and kindly providing plasmids. This work was supported by grant RO1-AI23538 from the NIAID to R.I., Program Project Award grant P30DK34928 from the NIDDK and training grant 5T32 AI107422. Work on the N-WASP knock-out cell line was supported by NIH grants HL03749 (S.B.S.) and HL59561 (Frederick W. Alt and S.B.S). R.I. is an Investigator of the Howard Hughes Medical Institute.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  • 1
    Alrutz, M.A. & Isberg, R.R. (1998) Involvement of focal adhesion kinase in invasin-mediated uptake. Proc Natl Acad Sci USA 95: 1365813663DOI: 10.1073/pnas.95.23.13658
  • 2
    Arber, S., Barbayannis, F.A., Hanser, H., Schneider, C., Stanyon, C.A., Bernard, O., Caroni, P. (1998) Regulation of actin dynamics through phosphorylation of cofilin by LIM-kinase. Nature 393: 805809
  • 3
    Aspenstrom, P., Lindberg, U., Hall, A. (1996) Two GTPases, Cdc42 and Rac, bind directly to a protein implicated in the immunodeficiency disorder Wiskott–Aldrich syndrome. Curr Biol 6: 7075
  • 4
    Barkalow, K., Witke, W., Kwiatkowski, D.J., Hartwig, J.H. (1996) Coordinated regulation of platelet actin filament barbed ends by gelsolin and capping protein. J Cell Biol 134: 389399
  • 5
    Black, D.S. & Bliska, J.B. (1997) Identification of p130Cas as a substrate of Yersinia YopH (Yop51), a bacterial protein tyrosine phosphatase that translocates into mammalian cells and targets focal adhesions. EMBO J 16: 27302744
  • 6
    Black, D.S. & Bliska, J.B. (2000) The RhoGAP activity of the Yersinia pseudotuberculosis cytotoxin YopE is required for antiphagocytic function and virulence. Mol Microbiol 37: 515527
  • 7
    Caron, E. & Hall, A. (1998) Identification of two distinct mechanisms of phagocytosis controlled by different Rho GTPases. Science 282: 17171721
  • 8
    Chen, L.M., Hobbie, S., Galan, J.E. (1996) Requirement of CDC42 for Salmonella-induced cytoskeletal and nuclear responses. Science 274: 21152118
  • 9
    Cox, D., Chang, P., Zhang, Q., Reddy, P.G., Bokoch, G.M., Greenberg, S. (1997) Requirements for both Rac1 and Cdc42 in membrane ruffling and phagocytosis in leukocytes. J Exp Med 186: 14871494
  • 10
    Dersch, P. & Isberg, R.R. (1999) A region of the Yersinia pseudotuberculosis invasin protein enhances integrin-mediated uptake into mammalian cells and promotes self- association. EMBO J 18: 11991213
  • 11
    Edwards, D.C., Sanders, L.C., Bokoch, G.M., Gill, G.N. (1999) Activation of LIM-kinase by Pak1 couples Rac/Cdc42 GTPase signalling to actin cytoskeletal dynamics. Nature Cell Biol 1: 253259
  • 12
    Egile, C., Loisel, T.P., Laurent, V., Li, R., Pantaloni, D., Sansonetti, P.J., Carlier, M.F. (1999) Activation of the CDC42 effector N-WASP by the Shigella flexneri IcsA protein promotes actin nucleation by Arp2/3 complex and bacterial actin-based motility. J Cell Biol 146: 13191332
  • 13
    Fu, Y. & Galan, J.E. (1999) A Salmonella protein antagonizes Rac-1 and Cdc42 to mediate host-cell recovery after bacterial invasion. Nature 401: 293297
  • 14
    Hall, A. (1990) The cellular functions of small GTP-binding proteins. Science 249: 597712
  • 15
    Hamburger, Z.A., Brown, M.S., Isberg, R.R., Bjorkman, P.J. (1999) Crystal structure of invasin: a bacterial integrin-binding protein. Science 286: 291295DOI: 10.1126/science.286.5438.291
  • 16
    Hardt, W.D., Chen, L.M., Schuebel, K.E., Bustelo, X.R., Galan, J.E. (1998) S. typhimurium encodes an activator of Rho GTPases that induces membrane ruffling and nuclear responses in host cells. Cell 93: 815826
  • 17
    Hartwig, J.H., Bokoch, G.M., Carpenter, C.L., Janmey, P.A., Taylor, L.A., Toker, A., Stossel, T.P. (1995) Thrombin receptor ligation and activated Rac uncap actin filament barbed ends through phosphoinositide synthesis in permeabilized human platelets. Cell 82: 643653
  • 18
    Higgs, H.N. & Pollard, T.D. (2000) Activation by Cdc42 and PIP(2) of Wiskott–Aldrich syndrome protein (WASp) stimulates actin nucleation by Arp2/3 complex. J Cell Biol 150: 13111320
  • 19
    Isberg, R.R. & Leong, J.M. (1990) Multiple β 1 chain integrins are receptors for invasin, a protein that promotes bacterial penetration into mammalian cells. Cell 60: 861871
  • 20
    Just, I., Selzer, J., Wilm, M., Von Eichel-Streiber, C., Mann, M., Aktories, K. (1995) Glucosylation of Rho proteins by Clostridium difficile toxin B. Nature 375: 500503
  • 21
    Leeuwen, F.N., Kain, H.E., Kammen, R.A., Michiels, F., Kranenburg, O.W., Collard, J.G. (1997) The guanine nucleotide exchange factor Tiam1 affects neuronal morphology; opposing roles for the small GTPases Rac and Rho. J Cell Biol 139: 797807
  • 22
    Leong, J.M., Morrissey, P.E., Marra, A., Isberg, R.R. (1995) An aspartate residue of the Yersinia pseudotuberculosis invasin protein that is critical for integrin binding. EMBO J 14: 422431
  • 23
    Machesky, L.M. & Insall, R.H. (1998) Scar1 and the related Wiskott–Aldrich syndrome protein, WASP, regulate the actin cytoskeleton through the Arp2/3 complex. Curr Biol 8: 13471356
  • 24
    Machesky, L.M., Mullins, R.D., Higgs, H.N., Kaiser, D.A., Blanchoin, L., May, R.C. , et al. (1999) Scar, a WASp-related protein, activates nucleation of actin filaments by the Arp2/3 complex. Proc Natl Acad Sci USA 96: 37393744
  • 25
    May, R.C., Caron, E., Hall, A., Machesky, L.M. (2000) Involvement of the Arp2/3 complex in phagocytosis mediated by FcγR or CR3 process. Nature Cell Biol 2: 246248
  • 26
    Mecsas, J., Raupach, B., Falkow, S. (1998) The Yersinia Yops inhibit invasion of Listeria, Shigella and Edwardsiella but not Salmonella into epithelial cells. Mol Microbiol 28: 12691281
  • 27
    Miki, H., Sasaki, T., Takai, Y., Takenawa, T. (1998a) Induction of filopodium formation by a WASP-related actin-depolymerizing protein N-WASP. Nature 391: 9396
  • 28
    Miki, H., Suetsugu, S., Takenawa, T. (1998b) WAVE, a novel WASP-family protein involved in actin reorganization induced by Rac. EMBO J 17: 69326941DOI: 10.1093/emboj/17.23.6932
  • 29
    Miki, H., Yamaguchi, H., Suetsugu, S., Takenawa, T. (2000) IRSp53 is an essential intermediate between Rac and WAVE in the regulation of membrane ruffling. Nature 408: 732735
  • 30
    Pepe, J.C., Wachtel, M.R., Wagar, E., Miller, V.L. (1995) Pathogenesis of defined invasion mutants of Yersinia enterocolitica in a BALB/c mouse model of infection. Infect Immun 63: 48374848
  • 31
    Persson, C., Carballeira, N., Wolf-Watz, H., Fallman, M. (1997) The PTPase YopH inhibits uptake of Yersinia, tyrosine phosphorylation of p130Cas and FAK, and the associated accumulation of these proteins in peripheral focal adhesions. EMBO J 16: 23072318DOI: 10.1093/emboj/16.9.2307
  • 32
    Del Pozo, M.A., Price, L.S., Alderson, N.B., Ren, X.D., Schwartz, M.A. (2000) Adhesion to the extracellular matrix regulates the coupling of the small GTPase Rac to its effector PAK. EMBO J 19: 20082014DOI: 10.1093/emboj/19.9.2008
  • 33
    Rohatgi, R., Ma, L., Miki, H., Lopez, M., Kirchhausen, T., Takenawa, T., Kirschner, M.W. (1999) The interaction between N-WASP and the Arp2/3 complex links Cdc42- dependent signals to actin assembly. Cell 97: 221231
  • 34
    Rohatgi, R., Ho, H., Kirschner, M.W. (2000) Mechanism of N-WASP activation by CDC42 and phosphatidylinositol 4, 5-bisphosphate. J Cell Biol 150: 12991310
  • 35
    Rosenshine, I., Duronio, V., Finlay, B.B. (1992) Tyrosine protein kinase inhibitors block invasin-promoted bacterial uptake by epithelial cells. Infect Immun 60: 22112217
  • 36
    Snapper, S.B., Takeshima, F., Antón, I., Liu, C.-H., Thomas, S., Nguyen, D. , et al. (2001) N-WASP deficiency reveals distinct pathways for cell surface projections and the actin-based motility of intracellular microorganisms. Nature Cell Biol 3: 897904
  • 37
    Tran Van Nhieu, G. & Collatz, E. (1987) Primary structure of an aminoglycoside 6′-N-acetyltransferase AAC(6′)-4, fused in vivo with the signal peptide of the Tn3-encoded beta-lactamase. J Bacteriol 169: 57085714
  • 38
    Tran Van Nhieu, G. & Isberg, R.R. (1993) Bacterial internalization mediated by β 1 chain integrins is determined by ligand affinity and receptor density. EMBO J 12: 18871895
  • 39
    Tran Van Nhieu, G., Krukonis, E.S., Reszka, A.A., Horwitz, A.F., Isberg, R.R. (1996) Mutations in the cytoplasmic domain of the integrin β1 chain indicate a role for endocytosis factors in bacterial internalization. J Biol Chem 271: 76657672
  • 40
    Tran Van Nhieu, G., Caron, E., Hall, A., Sansonetti, P.J. (1999) IpaC induces actin polymerization and filopodia formation during Shigella entry into epithelial cells. EMBO J 18: 32493262DOI: 10.1093/emboj/18.12.3249
  • 41
    Von Pawel-Rammingen, U., Telepnev, M.V., Schmidt, G., Aktories, K., Wolf-Watz, H., Rosqvist, R. (2000) GAP activity of the Yersinia YopE cytotoxin specifically targets the Rho pathway: a mechanism for disruption of actin microfilament structure. Mol Microbiol 36: 737748DOI: 10.1046/j.1365-2958.2000.01898.x
  • 42
    Yang, N., Higuchi, O., Ohashi, K., Nagata, K., Wada, A., Kangawa, K. , et al. (1998) Cofilin phosphorylation by LIM-kinase 1 and its role in Rac-mediated actin reorganization. Nature 393: 809812
  • 43
    Yarar, D., To, W., Abo, A., Welch, M.D. (1999) The Wiskott–Aldrich syndrome protein directs actin-based motility by stimulating actin nucleation with the Arp2/3 complex. Curr Biol 9: 555558