 |
Figure 1 N-WASP accumulation at one pole of intracellular Shigella. HeLa cells were infected with S.flexneri YSH6000 (A–D) or L.monocytogenes 1/2a EGD (E–G). Triple immunofluorescence stainings (A–D) were performed with rhodamine–phalloidin (A), Cy2-labeled anti-N-WASP antibody (B) and Cy5-labeled rat VRG-N2 antibodies (C). Double immunofluorescence stainings (E–G) were performed with rhodamine–phalloidin (E) and Cy2-labeled anti-N-WASP antibody (F). The yellow color in the combined image shown in (D) and (G) indicates co-localization between F-actin (red) and N-WASP (green). Arrows indicate an intracellular bacterium recruiting N-WASP and F-actin. Bar, 10  m.
|
 | Figure 2 Association of N-WASP with VirG is mediated by its verprolin-homology region. (A) Schematic representation of the structure of N-WASP and its variants. Bars indicate deleted constructs expressed in Cos-7 or recombinant GST fusion proteins. (B) Cell lysates from HeLa cells were incubated with Sepharose-immobilized GST–fusion proteins (20 g). Proteins bound to the beads were analyzed by immunoblotting with anti-N-WASP antibody. (C) Cell lysates from Cos-7 cells expressing N-WASP variants were used for a binding assay with GST– 1. Immunoblots of whole cell lysates are shown in the left panel. The 65 kDa protein bands appearing in all the lanes indicate endogenous N-WASP expressed in Cos-7 cells. The thick 65 ,  47 and 59 kDa bands correspond to full-length,  VCA and V constructs, respectively. Immunoblots of N-WASP bound to GST–1 beads are shown in the right panel. (D) Recombinant GST fusion proteins of N-WASP (5 g/lane) were analyzed by a gel blot overlay assay with VirG-1. Coomassie blue staining and immunoblots with anti-GST antibody are shown in the left and middle panel, respectively. Proteins transferred onto nitrocellulose membrane were overlaid with recombinant VirG-1 (10 g/ml), and bound protein was detected with rabbit VRG-N2 antibody.
|
 |
Figure 3 Targeting of GST–verprolin-homology region of N-WASP to surface-exposed VirG on intracellular Shigella. PtK2 cells infected with S.flexneri (M94 carrying pD10-1) were microinjected with GST–V fusion protein and immunostained using Cy2-labeled anti-GST (A), Cy5-labeled rabbit VRG-N2 antibodies (B) and rhodamine–phalloidin (C). Bar, 10 m.
|
 | Figure 4 VirG interacts with N-WASP via its glycine-rich repeat region. (A) Schematic representation of the construction of GST– domain fusion proteins and summary of N-WASP binding assays. (B) Cell lysates from HeLa cells were incubated with GST–VirG fusions immobilized on Sepharose beads. Proteins bound to the beads were analyzed by immunoblotting with anti-N-WASP antibody.
|
 |
Figure 5 Actin assembly by Shigella is inhibited by transient expression of dominant-negative N-WASP in Cos-7 cells. (A) Cell lysates from Cos-7 cells expressing N-WASP variants were used for a binding assay with GST–1. Immunoblots of whole cell lysates are shown in the upper panel. Immunoblots of N-WASP bound to GST–1 beads are shown in the lower panel. (B) Quantitation of the actin-associated intracellular bacteria in transfected cells. The transfected cells were infected with S.flexneri (M94 carrying pD10-1) or L.monocytogenes 1/2a EGD. The data shown are the means of triplicate experiments. The top bars show the standard deviation of the mean. (C) Cos-7 cells overexpressing full-length N-WASP (a, c and e) or cof mutant (b, d and f) were infected with Shigella and immunostained using anti-N-WASP antibody (a and b) and rhodamine–phalloidin (c and d). The yellow color in the combined image shown in (e) and (f) indicates co-localization between N-WASP (green) and F-actin (red). Bars, 10 m.
|
 | Figure 6 N-WASP is essential for VirG-induced actin assembly in Xenopus egg extracts. (A) Pre-immune IgG and anti-N-WASP antibody were used for immunodepletion of N-WASP from Xenopus egg extracts. Depleted extracts and precipitated proteins bound to the beads were analyzed by immunoblotting with anti-N-WASP antibody. (B) Quantitation of the actin-associated bacteria in N-WASP-depleted extracts. The data shown are the means of triplicate experiments. The top bars show the standard deviation of the mean. (C) Tail formation in N-WASP-depleted Xenopus egg extracts. Bacteria and actin tails were visualized by adding DAPI-labeled E.coli expressing VirG (blue) and rhodamine-labeled G-actin (red), respectively: (a) mock-depleted extracts, (b) N-WASP-depleted extracts, (c) N-WASP-depleted extracts plus purified N-WASP. Insets in (c) show that the bacteria possessed a short actin tail. Bar, 1 m.
|
 |
Figure 7 Hypothetical model for VirG-induced actin polymerization on Shigella in infected mammalian cells. The VirG -domain exposed on the bacterial surface binds vinculin and N-WASP. The vinculin could then interact with actin filaments and/or other signaling molecules such as VASP. The actin filaments recruited may be partially depolymerized by N-WASP and the uncapped barbed ends become exposed. At the uncapped barbed ends, actin polymerization would occur, possibly promoted by profilin. In this model, we are not sure whether or not the vinculin and N-WASP are recruited by VirG at the same time and by the same polypeptide.
|
