Enteropathogenic Escherichia coli Recruits the Cellular Inositol Phosphatase SHIP2 to Regulate Actin-Pedestal Formation  Katherine Smith, Daniel Humphreys, Peter J. Hume, Vassilis Koronakis  Cell Host & Microbe  Volume 7, Issue 1, Pages 13-24 (January 2010) DOI: 10.1016/j.chom.2009.12.004 Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 1 Control of Actin-Pedestal Morphology by EPEC Tir Y483 and Y511 (A) A schematic indicating the position of C-terminal tyrosine residues in EPEC (left) and EHEC (right) Tir that are exposed to the host cell cytoplasm. Pedestal formation by both EPEC and EHEC requires N-WASP-triggered activation of Arp2/3 for actin nucleation. Clustered EPEC Tir in the cellular plasma membrane is phosphorylated at residue Y474 triggering host adaptor Nck recruitment, which binds and stimulates N-WASP. EPEC can also generate pedestals through lower efficiency Nck-independent pathways (dashed black line). EHEC Tir clustering promotes interactions between Tir Y458 and the bacterial effector EspFu, which binds and activates N-WASP. No role has been assigned for Tir tyrosine residues Y483 and Y511 in EPEC Tir and Y490 and Y519 in EHEC Tir in pedestal formation. The regions surrounding these residues share sequence similarity to the eukaryotic regulatory motifs ITIMs and are referred to as tandem ITIM-like sequences (TILS). Conserved tyrosine residues in EPEC and EHEC Tir appear in the same color. EHEC Tir does not contain a tyrosine equivalent to EPEC Tir Y474. Right-hand panel: Amino acid sequence alignment of regions surrounding the tyrosine residues of the putative TILS motifs of EPEC Tir (Y483 and Y511) and EHEC Tir (Y490 and Y519), with the ITIM consensus sequence. Residues that correspond to the cellular ITIM consensus sequence are indicated in bold. (B) Immunofluorescence micrographs showing actin-pedestal morphology in HeLa cells primed (3 hr) with TirEPEC or TirEPEC Y483 511F and challenged (1 hr) with laboratory E. coli expressing intimin. Pedestals were visualized on fixed cells by staining for Tir (green) using an anti-Tir antibody, F-actin (red) using Texas-red phalloidin and bacteria (blue) with DAPI. The actin channel alone is shown in grayscale. These are representative data from five independent experiments. Equivalent data were obtained with NIH 3T3 fibroblasts (data not shown). Scale bar is 2 μm. (C) Scanning electron images of pedestals induced on HeLa cells by priming with TirEPEC (upper) or TirEPEC Y483F Y511F (lower) (3 hr) before a challenge with E. coli expressing intimin (1 hr). The pedestal components F-actin and bacteria are faux colored red and blue, respectively, using Adobe Photoshop software. Scale bar is 2 μm. (D) Images showing actin pedestals formed on HeLa cells following a prime (3 hr) with TirEHEC or TirEHEC Y490F Y519F and a challenge (1 hr) with E.coli expressing intimin. Fixed cells were processed for immunofluorescence microscopy as in (B). The actin channel is shown alone in grayscale. Scale bar is 2 μm. See also Figure S1. Cell Host & Microbe 2010 7, 13-24DOI: (10.1016/j.chom.2009.12.004) Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 2 Recruitment of the Inositol Phosphatase SHIP2 to EPEC-Induced Pedestals (A) The cellular distribution of the SH2-domain containing phosphatase derivatives SHP1-YFP, SHP2-YFP, SHIP1-YFP, or SHIP2-YFP (green) in HeLa cells. The localization of each phosphatase was monitored after infection with wild-type EPEC (3 hr) 16 hr post transfection. Fixed cells were stained with Texas-red conjugated phalloidin to visualize F-actin (red) and DAPI to visualize cell nuclei and bacteria (blue). The YFP channel is shown alone in grayscale. Scale bar is 4 μm. Insets show pedestals at a higher magnification. (B) Endogenous SHIP2 localization in HeLa cells after infection (3 hr) with wild-type EPEC (upper) or wild-type EHEC (lower). Fixed cells were stained for cellular SHIP2 (green) using an anti-SHIP2 antibody, F-actin (red), and bacteria (blue) using Texas-red conjugated phalloidin and DAPI respectively before visualization using fluorescence microscopy. The grayscale channel shows endogenous SHIP2 alone. Scale bar 4 μm. Insets show pedestals at a higher magnification. See also Figure S2. Cell Host & Microbe 2010 7, 13-24DOI: (10.1016/j.chom.2009.12.004) Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 3 Recruitment of SHIP2 by EPEC Tir Y483 and Y511 and Its Requirement for EPEC Actin-Pedestal Formation (A) Recruitment of endogenous SHIP2 in HeLa cells primed (3 hr) with TirEPEC, TirEPEC Y483 511F, TirEPEC Y483F, or TirEPEC Y511F before a challenge (1 hr) with E.coli expressing intimin. Fixed cells were stained with anti-SHIP2 (green), Texas-red-conjugated phalloidin to visualize F-actin (red) and DAPI, which localizes to cell nuclei and bacteria (blue). Grayscale shows endogenous SHIP2. Scale bar is 4 μm. (B) Representative images showing the impact of SHIP2 depletion on EPEC actin-pedestal morphology in HeLa cells. Cells were either mock transfected (upper) or transfected with siRNA specifically targeted against SHIP2 (lower) (72 hr). Transfected cells were either primed (3 hr) with TirEPEC before a challenge with E. coli expressing intimin (1 hr) (left) or infected (3 hr) with wild-type EPEC (right). Pedestal morphology was observed through phalloidin labeling of F-actin (red) and DAPI to show extracellular bacteria (blue). The actin channel alone is shown in grayscale. Scale bar is 4 μm. Equivalent results were obtained from multiple oligonucleotides targeted against SHIP2 whereas scrambled controls had no effect on pedestal structure. Normal pedestal morphology was restored following the expression of plasmid-encoded SHIP2, thus confirming the specificity of knockdown (data not shown). See also Figure S3. Cell Host & Microbe 2010 7, 13-24DOI: (10.1016/j.chom.2009.12.004) Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 4 Influence of SHC on EPEC Actin-Pedestal Formation (A) Immunofluorescence images of SHC localization in HeLa cells infected (3 hr) with wild-type EPEC. Fixed cells in the upper panels were stained with anti-SHC (green), Texas-red conjugated phalloidin to show F-actin (red), and DAPI to visualize bacteria (blue). The SHC channel is shown in grayscale. The lower panels were triple stained with anti-SHC (green), anti-SHIP2 (red), and with DAPI to visualize bacteria and cellular nuclei. The SHC and SHIP2 channels are shown alone in grayscale. Scale bar is 4 μm. Insets show pedestals at a higher magnification. (B) Immunofluorescence images showing colocalization of SHIP2 or SHC with Tir in HeLa cells primed (3 hr) with TirEPEC or the non-pedestal-forming Tir mutant TirEPEC Y474F before a challenge (1 hr) with E.coli expressing intimin. The left panels show cells costained with anti-SHIP2 (left, green) and anti-HA, to detect Tir (red). The right panels show cells costained with anti-SHC (right, green) and anti-Tir (red). Bacteria were also labeled by staining fixed cells with DAPI (blue). Scale bar is 4 μm. SHIP2 (left) or SHC (right) are shown individually in the grayscale channel. Immunofluorescence images were used to quantify SHIP2 and SHC colocalization with Tir. Over 50 stained cells were analyzed in three independent experiments. (C) Immunofluorescence micrographs showing the effect of RNAi against SHC on pedestal morphology. Seventy-two hours after transfection with siRNA against SHC, HeLa cells were infected with wild-type EPEC (3 hr). Images display F-actin (red) pedestal morphology by staining with Texas-red-conjugated phalloidin as well as DAPI to show bacteria (blue). The actin channel is shown alone in grayscale (right). Scale bar is 4 μm. Equivalent results were obtained from multiple oligonucleotides targeted against SHC whereas scrambled controls had no effect (data not shown). (D) Images showing the impact of RNAi treatment against SHIP2 and SHC on SHC recruitment. Seventy-two hours after transfection with siRNA against SHC (upper) or SHIP2 (lower), HeLa cells were primed (3 hr) with TirEPEC before a challenge with E. coli expressing intimin (1 hr). Fixed cells were stained with anti-SHIP2 (upper, green) or with anti-SHC (green, lower) antibodies. Cells were additionally stained with Texas-red-conjugated phalloidin to visualize actin (red) as well as DAPI to show bacteria (blue). The SHIP2 channel (upper) or the SHC channel (lower) is in grayscale. Scale bar is 4 μm. Equivalent results were obtained from multiple oligonucleotides targeted against SHC. Normal pedestal morphology was restored following the expression of plasmid-encoded SHC, thus confirming the specificity of knockdown (data not shown). See also Figure S4. Cell Host & Microbe 2010 7, 13-24DOI: (10.1016/j.chom.2009.12.004) Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 5 SHIP2 Inositol-5 Phosphatase Activity at EPEC-Induced Pedestals (A) Immunofluorescence images depicting the phosphoinositide composition of wild-type EPEC pedestals. HeLa cells were transiently transfected with the phosphoinositide probes TAPP1-YFP, which specifically recognizes PI(3,4)P2 and Btk-YFP, which localizes to PI(3,4,5)P3 (green) (16 hr). Posttransfection cells were infected with wild-type EPEC (3 hr) and probe localization was monitored. Cell nuclei and bacteria were observed by staining fixed cells with DAPI (blue). Scale bar is 4 μm. Insets show adherent bacteria at a higher magnification. (B) The distribution of the transiently transfected phosphoinositide probes Btk-YFP (PI(3,4,5)P3) (left, green) or TAPP1-YFP, (PI(3,4)P2) (right, green) in HeLa cells primed (3 hr) with TirEPEC, TirEPEC Y483F Y511F, TirEPEC Y483F, or TirEPEC Y511F and challenged with E. coli expressing intimin (1 hr) at 16 hr post-transfection. Fixed cells were stained with DAPI to visualize cell nuclei and bacteria (blue). The phosphoinositide probe alone is shown in grayscale. Scale bar is 4 μm. Insets show adherent bacteria at a higher magnification. See also Figure S5. Cell Host & Microbe 2010 7, 13-24DOI: (10.1016/j.chom.2009.12.004) Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 6 Lamellipodin Recruitment and Activity at EPEC Pedestals (A) The localization of LPD in HeLa cells infected with wild-type EPEC or EHEC or primed (3 hr) with TirEPEC or TirEPEC Y483F Y511F and challenged (1 hr) with laboratory E. coli expressing intimin. HeLa cells were stained for endogenous lamellipodin (LPD; green), an adaptor that is specifically recruited to the SHIP2 product PI(3,4)P2, 16 hr before infection. F-actin (red) and bacteria/cell nuclei (blue) were observed by staining with Texas red phalloidin and DAPI, respectively. LPD is shown alone in grayscale. Scale bar is 4 μm. (B) Immunofluorescence micrographs to show the impact of LPD depletion on EPEC pedestal morphology. HeLa cells were either mock transfected or transfected with siRNA against LPD before a prime (3 hr) with TirEPEC and challenge (1 hr) with laboratory E. coli expressing intimin (left) or infection with wild-type EPEC (3 hr) (right). Pedestals were observed by phalloidin staining of fixed cells to show F-actin (red) and staining with DAPI to display bacteria and cell nuclei (blue). The actin channel is shown alone in grayscale. Scale bar is 4 μm. Equivalent results were obtained from multiple oligonucleotides targeted against LPD whereas scrambled controls had no effect. Normal pedestal morphology was restored following the expression of plasmid-encoded LPD, confirming the specificity of knockdown (data not shown). See also Figure S6. Cell Host & Microbe 2010 7, 13-24DOI: (10.1016/j.chom.2009.12.004) Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 7 The Control of Actin-Pedestal Morphology by EPEC (A) Immunofluorescence images showing the recruitment of SHIP2, SHC, and LPD to wild-type EPEC. Cells were treated with siRNA against SHIP2, SHC, and LPD for 72 hr prior to infection (3 hr) with wild-type EPEC before being processed for immunofluorescence microscopy. Fixed cells were costained with anti-SHIP2, anti-SHC, and anti-LPD antibodies as appropriate and bacteria (blue) were visualized by DAPI staining. SHIP2, SHC, and LPD are shown individually in the grayscale channel. Scale bar is 4 μm. Insets show adherent bacteria at a higher magnification. (B) EPEC Tir is translocated into the epithelial cell in a type III secretion system-dependent manner. Once delivered, Tir localizes to the plasma membrane, where it is clustered by the bacterial outer membrane protein intimin. SHIP2 is recruited to clustered EPEC Tir either through direct or indirect interactions with Y483 and Y511. Phosphorylation of Tir at Y474 by host kinases promotes recruitment of SHC to SHIP2 at Tir, as well as inducing the signaling cascade necessary for actin-pedestal rearrangements. SHIP2 also has inositol-5 phosphatase activity and cleaves PI(3,4,5)P3 into PI(3,4)P2 providing a lipid signaling platform essential for the specific recruitment of the cytoskeletal regulator LPD to control actin pedestal morphology. See also Figure S7. Cell Host & Microbe 2010 7, 13-24DOI: (10.1016/j.chom.2009.12.004) Copyright © 2010 Elsevier Inc. Terms and Conditions