Fig 1.
Chemical inhibition of formins decreases actin pedestal assembly by EPEC, but not by KC12+EspFU.
(A) HeLa cells were pretreated with DMSO, CK666+CK869 (CKCK), Wiskostatin (WISKO), or SMIFH2, and infected with either EPEC or KC12+EspFU for 3.5 h. Cells were fixed and treated with antibodies to detect HA-Tir (red), phalloidin to visualize F-actin (green), and DAPI to stain DNA (blue). Scale bar, 25 µm. (B-C) The % of adherent EPEC or KC12+EspFU (defined by Tir staining) that associated with pedestals was quantified from experiments performed as in A. Each point represents a single infected cell (n = 25–30) harboring 10–50 bacteria, and bars display the mean % +/- SEM. (D-E) The F-actin pixel intensity in the pedestal-forming region (indicated by Tir staining) was quantified and normalized to adjacent pedestal-free areas of the cell, which were set to 1. Each point represents a single EPEC or KC12+EspFU pedestal-forming region, and lines show the mean intensity +/- 95% CI (n = 150 pedestal-forming regions, 15 cells). ** p<0.01, *** p<0.001, ns = not significant (ANOVA, Tukey post-hoc tests). Significance asterisks in B and C are in reference to the DMSO treated conditions.
Fig 2.
Chemical inhibition of formins impairs EPEC motility and colonization.
(A) NIH3T3 cells stably expressing mCherry-actin (red) were infected with EPEC (left) or KC12+EspFU (right) for 3.5–4.0 h, treated with the indicated inhibitors, and imaged live for 20–30 min. Bacteria with pedestals were tracked over time (top panels, scale bar, 2 µm), and white lines highlight the paths taken by representative bacteria during imaging. These experiments were used to determine actin-based motility rates (lower panels). Each point represents a single bacterium associated with a pedestal, and lines show the mean speed +/- SEM (n = 25–40 pedestals, 3–4 cells per condition). *** p < 0.001 (ANOVA, Tukey post-hoc tests). (B) Polarized Caco-2 monolayers were treated with the indicated inhibitors for 15 min prior to and during infection with EPEC or KC12+EspFU for 6 h. Monolayers were then fixed and treated with antibodies to detect LPS (red), phalloidin to visualize F-actin (green), and DAPI to label DNA (blue). Scale circles, 500, 100 µm2. (C-D) Experiments shown in B were quantified. Each bar represents the mean macrocolony area +/- SEM (n = 70–124 EPEC colonies, 750–1177 KC12+EspFU colonies). Only macrocolonies larger than 100 µm2 were included in the analysis. * p < 0.05, ** p < 0.01, *** p<0.001 (ANOVA, Dunnett’s multiple comparison test). Significance asterisks in A, C, and D are in reference to the DMSO treated conditions.
Fig 3.
An siRNA screen reveals that targeting mDia1 inhibits actin pedestal formation by EPEC.
(A) HeLa cells were treated with siRNA pairs to the indicated targets, infected with EPEC for 4 h, fixed, and stained with antibodies to detect HA-Tir (magenta) and with phalloidin to visualize F-actin (green). Scale bar, 10 µm. (B-C) The % of adherent, Tir-positive EPEC (B) or KC12+EspFU (C) that were associated with pedestals was quantified from experiments performed in A. Each bar shows the mean % +/- SEM of pedestal-forming bacteria, while black data points represent mean %s from each infected cell that harbored 10–50 bacteria (n = 15–53 cells). * p < 0.05, *** p<0.001 (ANOVA, Dunnett’s multiple comparison test). Significance asterisks in B and C are in reference to the control.
Fig 4.
Depletion of mDia1 inhibits actin pedestal assembly by EPEC but not by KC12+EspFU.
(A) HeLa cells were treated with control siRNAs or independent siRNAs targeting mDia1 and infected with EPEC or KC12+EspFU for 4 h. Cells were fixed and treated with antibodies to detect HA-Tir (red) and mDia1 (green), and with phalloidin to visualize F-actin (magenta). Scale bar, 10 µm. (B) The whole cell fluorescence intensity for mDia1 from experiments shown in A was measured. Each point represents the average pixel intensity of a single cell, and black lines show the mean intensity +/- SD for 28–32 cells. *** p<0.001, ns = not significant (ANOVA, Tukey post-hoc tests). (C) Lysates from uninfected cells treated in parallel to those in A were analyzed by immunoblotting with antibodies to detect mDia1, actin, and GAPDH. mDia1 band intensity was calculated relative to actin and GAPDH and normalized to 100 in control extracts. Data represent the mean +/- SEM quantified from 9 blots encompassing 7 experiments. (D) The % of adherent bacteria (determined by HA-Tir staining) associated with pedestals was quantified from experiments performed in A. Each point represents a single infected cell (n = 15–20 cells) harboring 10–50 bacteria, and lines display the mean % +/- SD. *** p<0.001, ns = not significant (ANOVA, Tukey post-hoc tests). (E-F) The whole-cell intensity of mDia1 staining (for control and mDia1 depleted cells) was plotted against the % of EPEC (E) or KC12+EspFU (F) forming pedestals on that cell. Each point represents a single cell (n = 45–46 cells). Data were subjected to linear regression analysis, and linear trend lines, including R2 values, are displayed on the plot with p-values describing whether the slopes are significantly non-zero. (G) Cells were treated with control siRNAs or individual siRNAs targeting ArpC4 or mDia1. After fixation, cells were treated with antibodies to detect HA-Tir (red), with phalloidin to visualize F-actin (green), and with DAPI to label DNA (blue). Scale bar, 2 µm. (H) Pixel intensity plots were generated from cells treated as in G. Lines were drawn through the pedestal-forming region (indicated by Tir staining and displayed as white lines in panel G) and F-actin intensity along the 3 µm line was plotted. All plots were normalized so that a distance of 0 represents the brightest fluorescence of HA-Tir, with the bacteria positioned to the left of 0. Points represent the normalized mean fluorescence of F-actin +/- SEM (n = 21 pedestal-forming regions per condition from 4–5 cells).
Fig 5.
mDia1 is recruited to EPEC pedestals in a Tir phosphotyrosine-mediated manner.
(A) HeLa cells (upper panels), or HeLa cells transiently expressing GFP-mDia1 (lower panels) were infected with EPEC for 4 h, fixed, and treated with antibodies to detect HA-Tir, antibodies to stain mDia1 (top panels only), and with phalloidin to visualize F-actin. Arrowheads highlight mDia1-positive pedestals. Scale bars, 10 µm. (B) HeLa cells were infected, fixed, and stained as in A. Lines were drawn through pedestals to measure pixel intensity profiles for HA-Tir, F-actin, and mDia1. The brightest pixel from HA-Tir staining was set to a distance of 0 (indicated by the black triangle) for each pedestal to normalize the F-actin and mDia1 curves (n = 15 pedestals, 4 cells). Each point represents the mean normalized pixel intensity (+/- 95% CI), and black points indicate the maxima. (C) The % of Tir-positive EPEC or KC12+EspFU associated with an enrichment in mDia1 staining was calculated from control siRNA experiments described in Fig 4A. Each point represents a single infected cell, and the mean % +/- SEM is indicated in black (n = 16–20 cells). *p<0.05 (unpaired t test). (D) HeLa cells were infected with the indicated strains of EPEC or KC12 for 3.5 h, fixed, and treated with antibodies to detect HA-Tir (red) and mDia1 (green), with phalloidin to visualize F-actin (magenta), and with DAPI to label DNA. Scale bar, 5 µm. (E) 3 adherent bacteria from each panel in D were selected for line scan analysis as in A-B. Each point displays the normalized pixel intensity of F-actin (magenta) or mDia1 (green) along the 2–3 µm line in the corresponding panel. (F) HeLa cells treated with siRNAs against mDia1 were infected with EPEC expressing EspFU-myc, then fixed and stained with antibodies to detect mDia1 (green) and EspFU-myc (red), and with phalloidin to visualize F-actin (magenta). The “+” indicates a cell with detectable levels of mDia1, and the “-” indicates a cell with substantially less mDia1 because of successful knockdown. Scale bar, 25 µm; inset, 5 µm.
Fig 6.
The Arp2/3 complex is essential for all forms of actin pedestal assembly.
(A) ArpC2-floxed mouse fibroblasts were treated with DMSO (ArpC2 Flox) or tamoxifen (ArpC2 KO) for 6 days to delete arpC2 and deplete the Arp2/3 complex. Following 2–6 days of culturing in normal growth media, cells were infected with EPEC or KC12+EspFU for 4 h, fixed, and stained with antibodies to detect HA-Tir (magenta), phalloidin to visualize F-actin (green), and DAPI to label DNA (blue). The arrow denotes a weak actin punctum. The bottom row shows an F-actin basket. Scale bar, 5 µm. (B) ArpC2 Flox and KO cell populations cultured as in A were treated with control siRNAs or independent siRNAs targeting mouse mDia1. Cell lysates were analyzed by immunoblotting with antibodies to detect mDia1, ArpC2, tubulin, and actin. Mean mDia1 and ArpC2 band intensities +/- SEM were normalized to tubulin and actin and quantified from 2 experiments. ND, not detectable. (C) ArpC2 KO cells were treated with control or mDia1 siRNAs, infected with EPEC, fixed, and stained with antibodies to detect HA-Tir (red), phalloidin to visualize F-actin (green), and DAPI to label DNA (blue). Scale bar, 2 µm. (D) Pixel intensity plots were generated from cells treated as in C. Lines were drawn through the pedestal-forming region (indicated by Tir staining and displayed as white lines in panel C) and F-actin intensity along the line was plotted. All lines were normalized so that a distance of 0 represents the brightest fluorescence of HA-Tir, with the bacteria positioned to the left of 0. Points represent the normalized mean fluorescence of F-actin +/- SEM (n = 9 pedestal-forming regions per condition from 3 cells).
Fig 7.
mDia1 and the Arp2/3 complex localize to distinct subregions within EPEC pedestals.
(A) HeLa cells were infected with EPEC for 4 h, fixed, and treated with antibodies to detect mDia1 (green) and Arp3 (red), as well as phalloidin to visualize F-actin (magenta), and DAPI to label DNA. Scale bars, 2 µm. (B) Pixel intensity plots were generated from 3 actin pedestals selected from A (white lines i, ii, and iii). Each point represents the relative fluorescence intensity of Arp3 (red) or mDia1 (green) along the 2–3 µm line. All lines were normalized so that a distance of 0 represents the brightest fluorescence of DNA, with the pedestal positioned to the right of 0, and black points represent the maxima. (C) Line scan analyses of EPEC pedestals stained as in A and analyzed as in B were used to calculate the distance from the maximum DNA signal (positioned at 0) to the distance of the maximum signal for F-actin, Arp3, and mDia1 staining. Each point represents the distance within an individual pedestal (n = 18 pedestals), and black lines indicate the mean distance +/- SEM. ** p<0.01 (ANOVA, Tukey post-hoc test). (D) ArpC2 Flox and KO cells were treated with control or mouse mDia1 siRNAs, infected with EPEC, fixed, and stained with antibodies to detect HA-Tir (red) and mDia1 (green), as well as phalloidin to visualize F-actin (magenta), and DAPI to label DNA (blue). The arrow indicates mDia1 enrichment near an accumulation of F-actin adjacent to Tir. Scale bar, 2 µm. (E) HeLa cells were pre-infected for 4 h with EPECΔeae, then washed and challenged with E. coli expressing intimin for 10 min. Cells were fixed and stained with antibodies to detect mDia1 (green) and Arp3 (red), phalloidin to visualize F-actin (magenta), and DAPI to label bacterial DNA. The arrow points to a bacterium lacking a pedestal, and the arrowhead indicates a pedestal. Scale bar, 2 µm.
Fig 8.
mDia1 is important for Src-family kinase activation and phosphorylation of EPEC Tir Y474.
(A) HeLa cells were treated with control or mDia1 siRNAs and infected with EPEC for 4 h. Cells were fixed and stained with antibodies to detect HA-Tir (red) and phosphotyrosine residues (pY, green). Scale bar, 10 µm. (B) The fluorescence intensities of HA-Tir and pY from cells treated as in A were quantified in the pedestal forming regions by outlining the area encompassing HA-Tir and measuring the average pixel intensity within that area. Intensities were normalized to a bacteria-free area of the cell, and then to the mean for control siRNA treated cells, which was set to 1. Each point represents the pY/HA-Tir intensity ratio for an individual bacterium, and black lines show the mean intensity +/- SEM for 131–165 Tir foci from 10 cells per condition. *** p<0.001, ns = not significant (ANOVA, Tukey post-hoc tests). (C) Cells treated with an mDia1 siRNA were fixed and stained with antibodies to detect mDia1 (green) and pY (red), as well as phalloidin to visualize F-actin (magenta), and DAPI to label DNA (blue). The “+” indicates a cell with detectable levels of mDia1, and the “-” indicates a cell with substantially less mDia1 because of successful knockdown. The arrow denotes a bacterium with no pY staining or actin assembly in an mDia1-depleted cell, and the arrowhead highlights a bacterium with pY staining and an actin pedestal in an adjacent mDia1-expressing cell. Scale bar, 25 µm. (D) Cells were treated as in A, but stained with antibodies to phosphotyrosine 416 in active Src-family kinases (pSrc, green). Scale bar, 25 µm. (E) The normalized HA-Tir intensity and pSrc intensity at pedestal forming regions from cells treated as in D were quantified and normalized as in B. Each point represents the normalized pSrc/HA-Tir intensity ratio for an individual bacterium, and black lines show the mean intensity +/- SEM for 70–94 Tir foci from 8 cells per condition. *** p<0.001, ** p<0.01, ns = not significant (ANOVA, Tukey post-hoc tests). (F) HeLa cells were treated with control siRNAs or independent siRNAs targeting mDia1 and either left uninfected (“-”) or infected with EPEC expressing wild type Tir (“WT”) or the Tir Y474F mutant (“F”). Cell lysates were analyzed by immunoblotting with antibodies to mDia1, phosphotyrosine (pTir), HA-Tir, Src phosphotyrosine-416 (pSrc), total Src, actin, and GAPDH. (G) mDia1 band intensities from immunoblots described in F were quantified relative to actin and GAPDH and normalized to the control, which was set to 100. Each point represents the intensity calculated for a single mDia1 band, and the mean +/- SEM is indicated in black. *** p<0.001 (ANOVA, Tukey post-hoc tests). (H) A Tir phosphorylation index was determined from immunoblot band intensities described in F by calculating the pTir to Tir ratio for each condition and normalizing to the control, which was set to 1.0. Each point represents the Tir phosphorylation index within a single sample, and the mean +/- SEM is indicated in black. *** p<0.001 (ANOVA, Tukey post-hoc tests). (I) The mDia1 densitometry points from G were plotted against the Tir phosphorylation indices from H. Data were subjected to linear regression analysis, and the linear trend line and R2 value are displayed on the plot, with the p-value indicating that the slope is significantly non-zero. (J) A Src activation index was determined by calculating the pSrc to Src ratio for each condition and normalizing to the control, which was set to 1.0. Each point represents the Src activation index within a single sample, and the mean +/- SEM is indicated in black. *** p<0.001 (ANOVA, Tukey post-hoc tests).
Fig 9.
Model for mDia1 functions in EPEC pedestals.
(A) We propose that mDia1 activity at the base of the pedestal assembles linear actin filaments, which can be used as mother filaments by the Arp2/3 complex to generate a densely branched network towards the pedestal tip. In the absence of mDia1, pedestals are less numerous and weaker due to fewer seed filaments and lower levels of Tir phosphorylation. In the absence of the Arp2/3 complex, any actin that is nucleated by mDia1 (as weak puncta or actin baskets) is unable to be reorganized and focused into a pedestal. See Discussion for details. (B) A previously established signaling pathway for pedestal formation (black solid arrows) is driven by Tir clustering, tyrosine phosphorylation by Src and Abl family kinases, recruitment of Nck1/2 and N-WASP, and activation of the Arp2/3 complex to polymerize branched actin networks. Our current results expand on this model to include (1) mDia1-associated actin assembly (blue solid arrow) possibly providing seed filaments for the Arp2/3 complex (blue dashed arrow), and (2) mDia1-mediated enhancement of Src activation in pedestals (blue solid arrow) to promote and sustain Tir phosphorylation. It is unclear if mDia1 is recruited by SFKs, pY474, or the Arp2/3 complex, or if Arp2/3-based actin networks that disassemble at the base of the pedestal are recycled by mDia1 (gray dashed arrows). See Discussion for details.