The Serine/Threonine Kinase STK11 Promotes Shigella flexneri Dissemination through Establishment of Cell-Cell Contacts Competent for Tyrosine Kinase Signaling

Abstract

Shigella flexneri is an intracellular pathogen that disseminates in the intestinal epithelium by displaying actin-based motility. We found that although S. flexneri displayed comparable actin-based motilities in the cytosols of HeLa229 and HT-29 epithelial cell lines, the overall dissemination process was much more efficient in HT-29 cells. Time-lapse microscopy demonstrated that as motile bacteria reached the cell cortex in HT-29 cells, they formed membrane protrusions that resolved into vacuoles, from which the bacteria escaped and gained access to the cytosol of adjacent cells. In HeLa229 cells, S. flexneri also formed membrane protrusions that extended into adjacent cells, but the protrusions rarely resolved into vacuoles. Instead, the formed protrusions collapsed and retracted, bringing the bacteria back to the cytosol of the primary infected cells. Silencing the serine/threonine kinase STK11 (also known as LKB1) in HT-29 cells decreased the efficiency of protrusion resolution into vacuoles. Conversely, expressing STK11 in HeLa229 cells, which lack the STK11 locus, dramatically increased the efficiency of protrusion resolution into vacuoles. S. flexneri dissemination in HT-29 cells led to the local phosphorylation of tyrosine residues in protrusions, a signaling event that was not observed in HeLa229 cells but was restored in STK11-expressing HeLa229 cells. Treatment of HT-29 cells with the tyrosine kinase inhibitor imatinib abrogated tyrosine kinase signaling in protrusions, which correlated with a severe decrease in the efficiency of protrusion resolution into vacuoles. We suggest that the formation of STK11-dependent lateral cell-cell contacts competent for tyrosine kinase signaling promotes S. flexneri dissemination in epithelial cells.

INTRODUCTION

The intracellular pathogen Shigella flexneri is the causative agent of bacillary dysentery in humans (1). The disease is characterized by bacterial invasion of the colonic epithelium and associated with inflammation and destruction of the epithelial mucosa. The invasion process relies on the bacterial type three secretion system, which triggers uptake of the bacterium via massive cytoskeletal rearrangements (2). Once the pathogen escapes the vacuole compartment, it replicates in the cytosol of infected cells, where it displays actin-based motility (3, 4). As motile bacteria encounter the plasma membrane of the primary infected cell, they form membrane protrusions that extend into adjacent cells (5). The Shigella-containing protrusions resolve into Shigella-containing double membrane vacuoles (DMV) in adjacent cells. The pathogen then escapes the DMV and gains access to the cytosol of adjacent cells, where it continues its infectious cycle by replicating and displaying actin-based motility. The ability to spread from cell to cell is a central determinant of S. flexneri pathogenesis, and spreading-defective mutants are avirulent (3, 4).

The mechanisms supporting actin-based motility of intracellular pathogens were deciphered using Listeria monocytogenes as a model system (6). L. monocytogenes displays actin-based motility in the cytosol of infected cells through recruitment of the ARP2/3 complex, a major component of the assembly machinery (7). The ARP2/3 complex binds to existing actin filaments and nucleates the formation of daughter filaments, whose elongation leads to the formation of a branched network (8). The expansion of the network generates forces at the bacterial surface that propel the pathogen throughout the cytosol. The recruitment of the ARP2/3 complex to the bacterial surface relies on the expression of the bacterial factor ActA (9, 10), which mimics the activity of N-WASP (11, 12), an endogenous cytoskeleton regulator that promotes the activity of the ARP2/3 complex. S. flexneri actin-based motility also relies on the ARP2/3 complex (5, 13). The recruitment of the nucleator to the bacterial surface is indirectly mediated by the bacterial factor IcsA through recruitment of the nucleation-promoting factor N-WASP (14, 15).

In contrast to the mechanisms supporting cytosolic motility, the mechanisms supporting the actual spread from cell to cell are poorly understood. Seminal electron microscopy studies on cells infected with Listeria monocytogenes revealed that intracellular bacteria displaying actin-based motility in the cytosol of infected cells spread from cell to cell through formation of membrane protrusions that resolve into double membrane vacuoles in adjacent cells (16). Similarly, electron microscopy studies established that S. flexneri also forms membrane protrusions that resolve into double membrane vacuoles in adjacent cells (5, 17). Studies using a mouse fibroblastic sarcoma cell line that does not produce cell adhesion molecules indicated that expression of E-cadherin is required for efficient S. flexneri spread from cell to cell (17). These observations led to a model of S. flexneri dissemination in which cell adhesion molecules not only contribute to the rigidity of protrusions through association of actin tails with the plasma membrane but also promote efficient uptake of protrusions by adjacent cells through intercellular homotypic interactions. Mechanistically, a recent study has revealed that in addition to the requirement of the ARP2/3 complex for actin-based motility, S. flexneri requires the activity of formins in order to form protrusions (18). In addition, a role for endocytosis in the uptake of protrusions by adjacent cells was recently proposed on the basis of experiments involving pharmacological inhibition or depletion of components of the endocytic machinery (19). In this study, we conducted a comparative examination of S. flexneri dissemination in HT-29 and HeLa229 cells, and we report here a role for the serine/threonine kinase STK11 in the formation of cell-cell contacts competent for tyrosine kinase signaling, which promotes bacterial dissemination through resolution of protrusions into vacuoles.

MATERIALS AND METHODS

Cell lines and bacterial strains.

HeLa229 cells (ATCC CCL-2.1) and HT-29 cells (ATCC HTB-38) were cultured at 37°C with 5% CO₂ in Dulbecco modified Eagle medium (DMEM; Gibco) and McCoy's 5A medium (Gibco), respectively, supplemented with 10% heat-inactivated fetal bovine serum (FBS; Invitrogen). The Shigella flexneri strain used in this study was serotype 2a strain 2457T (20).

Bacterial infection.

S. flexneri was grown overnight in LB broth at 37°C with agitation. Twenty microliters of stationary-phase culture was used to inoculate 2 ml of LB broth, and the bacteria were grown to exponential phase for approximately 3 h at 37°C. Cells were infected with S. flexneri expressing green fluorescent protein (GFP), yellow fluorescent protein (YFP), or red fluorescent protein (RFP) under the control of an isopropyl-β-d-thiogalactopyranoside (IPTG)-inducible promoter. Infection was initiated by centrifuging the plate at 1,000 rpm for 5 min, and internalization of the bacteria was allowed to proceed for 1 h at 37°C before gentamicin (50 μM final) was added in order to kill the extracellular bacteria. Two hours before the infection was stopped, IPTG (4 mM final) was added to the medium to induce GFP, YFP, or RFP expression in intracellular bacteria. For protrusion formation, vacuole resolution, and actin tail quantification, infected cells were incubated at 37°C for 4 h and fixed in phosphate-buffered saline (PBS)–4% formaldehyde at 4°C overnight before immunostaining. For S. flexneri focus size analysis, infected cells were incubated at 37°C for 8 h. For tyrosine kinase inhibition, imatinib (100 nM) or vehicle (dimethyl sulfoxide [DMSO], mock) was added 1.5 h after the initiation of infection.

DNA constructs and cell transfection.

HT-29 cell lines stably expressing cyan fluorescent protein (CFP) membrane, RFP membrane, or GFP membrane markers were generated using the pLB vector from Addgene (Addgene plasmid 11619). The corresponding lentiviruses were generated in 293E cells cotransfected with the packaging constructs pCMVΔ8.2Δvpr (HIV helix packaging system) and pMD2.G (a vesicular stomatitis virus glycoprotein) as previously described (21). In order to generate GFP-Mb-pLB and RFP-Mb-pLB, the original pAcGFP1-Mem and pDsRed-Monomer-Mem vectors from Clontech were used for subcloning into the EcoRI and AgeI restriction sites of pLB. The CFP-Mb-pLB vector was generated by cloning CFP into the NotI and KasI sites of pECFP-N1 and further subcloning into the EcoRI and AgeI sites of pLB.

STK11 cDNA was amplified using primers ATTCTCGAGCTGAGGTGGTGGACCCGCAGC and TTAGAATTCTCACTGCTGCTTGCAGGCCGAC and cloned into the XhoI and EcoRI restriction sites (underlined in the primer sequence) of the pDsRed-Monomer C1 vector (Clontech). In order to generate the STK11-pLB, the DNA fragment corresponding to STK11-DsRed was subcloned into the EcoRI and AgeI sites of pLB.

Immunofluorescence and antibodies.

Cells were fixed and permeabilized in methanol (5 min at −20°C) for staining with anti-E-cadherin (1:1,000; BD Transduction Laboratories) and anti-occludin (1:100, Invitrogen) antibodies. Cells were fixed in PBS–4% formaldehyde (overnight at 4°C) and permeabilized in PBS–0.1% Triton for staining with antiphosphotyrosine (1:500; Cell Signaling) antibody and Alexa Fluor phalloidin (1:1,000; Invitrogen). The secondary antibodies used were anti-mouse Alexa 594, anti-mouse Alexa 488, anti-rabbit Alexa 488, and anti-mouse 514 (1:1,000; Molecular Probes).

Size of infection foci.

The size of infection foci formed in HT-29 and HeLa229 cells infected with GFP-expressing S. flexneri was determined in a 384-well plate format. After fixation and 4′,6-diamidino-2-phenylindole (DAPI) staining, the plates were imaged using a TE2000 automated microscope (Nikon) equipped with a motorized stage (Prior), motorized filter wheels (Sutter Instrument, Inc.), and a 10× objective (Nikon) mounted on a piezo focus drive system (Physik Instrumente). Image acquisition was conducted using MetaMorph 7.1 software (Molecular Devices, Inc.). For image analysis (see Fig. S1 in the supplemental material), high-intensity pixels corresponding to the bacteria were selected against the low-intensity pixels corresponding to background levels by using the threshold function of the MetaMorph 7.1 imaging software. Selected pixels were clustered by proximity using the “close” morphological filter in order to create a single object that represented a given infection focus. The integrated morphometry analysis (IMA) module was used to determine the area of each infection focus in a given image. Image analyses were conducted on at least 25 infection foci in three independent experiments.

Cytosolic velocity and dynamics of cell-to-cell spread.

For time-lapse microscopy, HeLa229 and HT-29 cells were grown on 35-mm imaging dishes at 37°C in 5% CO₂ (MatTek, Ashland, MA). Cells were infected with S. flexneri and imaged with a Nikon TE2000 spinning-disc confocal microscope driven by the Volocity software package (Improvision). For analysis of cytosolic velocity, Z-stacks were captured every minute for at least 30 min and individual bacteria were tracked in four dimensions using the tracking module of the Volocity software. The average speed was computed for bacteria undergoing motility (>0.01 μm/s) for at least 60 consecutive seconds.

For analysis of protrusion and vacuole formation, and vacuole escape, z-stacks were captured 2.5 h postinfection every 6 min for at least 150 min. For clarity, the Z plans corresponding to the basal plasma membrane were subsequently excluded from the merged z-stacks, as presented in Fig. 2 and 3 and Fig. S3 in the supplemental material. Protrusions were defined as plasma membrane extensions that formed as a result of motile bacteria reaching the cell cortex and progressing further into adjacent cells. Vacuoles were defined as membrane-bound compartments that derived from protrusions and displayed a continuous lining of the plasma membrane surrounding the bacteria. As opposed to protrusions, vacuoles were therefore no longer connected to the sending cell through a membranous neck. Vacuole escape was defined as the appearance of a discontinuity in the otherwise continuous lining of the plasma membrane surrounding bacteria in vacuoles. Free bacteria were defined as bacteria that were previously observed in vacuoles but were no longer surrounded by a continuous lining of the plasma membrane. Membrane remnants were defined as plasma membrane structures that remained associated with free bacteria after vacuole escape.

FIG 2.

Dynamics of S. flexneri dissemination in HT-29 cells. (A to C) Time-lapse microscopy of HT-29 cells (green, GFP membrane marker) infected with RFP-expressing S. flexneri (red, Shigella). Representative images show three bacteria (arrows) forming membrane protrusions (protrusion) that resolve into vacuoles (vacuole), from which the bacteria escape (vacuole escape) and gain access to the cytosol of adjacent cells (free bacteria). For each panel, the top images are low-magnification images of infected cells and the bottom images are enlargements of the tracked bacterium (merged bacterium and membrane channels, left; membrane channel only, right). (D) Tracking analysis of 40 bacteria (including the ones displayed in panels A, B, and C) that formed protrusions in 10 independent foci. All bacteria were tracked for 150 min, and the progression of the dissemination process was depicted using the color key displayed at the bottom: protrusion, light blue; vacuole, yellow; free bacteria in the cytosol of adjacent cells, red; and retraction to the primary infected cells, dark blue. In HT-29 cells, more than 70% (29/40) of the bacteria that formed protrusions gained access to the cytosol of adjacent cells.

FIG 3.

Dynamics of S. flexneri dissemination in HeLa229 cells. (A to C) Time-lapse microscopy of HeLa229 cells (green, GFP membrane marker) infected with RFP-expressing S. flexneri (red, Shigella). A time-lapse sequence shows representative examples of bacteria forming protrusions (protrusion, arrow) that subsequently retracted to the primary infected cells (primary cell, arrow). For each panel, the top images are low-magnification images of infected cells and the bottom images are enlargements of the tracked bacterium (merged bacterium and membrane channels, left; membrane channel only, right). (D) Tracking analysis of 40 bacteria (including the ones displayed in panels A, B, and C) that formed protrusions in 10 independent foci. All bacteria were tracked for 150 min, and the progression of the dissemination process was depicted using the color key displayed at the bottom: protrusion, light blue; vacuole, yellow; free bacteria in the cytosol of adjacent cells, red; and retraction to the primary infected cells, dark blue. In HeLa229 cells, 95% (38/40) of the bacteria that formed protrusions failed to gain access to the cytosol of adjacent cells.

RESULTS

Quantification of S. flexneri dissemination in HT-29 cells and HeLa229 cells.

We have recently shown that the intestinal HT-29 cell line constitutes a viable system for modeling Shigella flexneri actin-based motility in the cytosol of human intestinal cells (22). In this experimental system, motile bacteria spread from cell to cell and lead to the formation of large infection foci, as illustrated in Fig. 1 by the low-magnification view of infected cells (Fig. 1A, top row) and the high-magnification view of a representative infection focus (Fig. 1A, bottom row). As a measure of bacterial dissemination, we used computer-assisted image analysis to quantify the size of the formed infection foci (see Fig. S1 in the supplemental material). Statistical analyses confirmed the expected increase in the size of infection foci as the infection progressed in HT-29 cells (Fig. 1A and B, 4, 6, and 8 h). We also used computer-assisted image analysis to quantify the size of the infection foci formed in HeLa229 cells. HeLa229 cells supported the invasion process as well as HT-29 cells, as demonstrated by the presence of numerous infected cells 4 h postinfection (Fig. 1C, 4 h). However, the bacteria failed to spread from cell to cell (Fig. 1C, 4, 6, and 8 h) and the average size of the formed infection foci did not increase during the course of infection (Fig. 1D, 4, 6, and 8 h). Thus, S. flexneri disseminated poorly in HeLa229 cells, as suggested in previous studies using HeLa and Caco-2 cells (23).

FIG 1.

Quantification of S. flexneri dissemination in HT-29 and HeLa229 cells. (A and C) Infection of HT-29 cells (A) and HeLa229 cells (C) with GFP-expressing S. flexneri. Representative images of time course experiments show the increase in the size of infection foci in HT-29 cells (A, infection focus, 4 h, 6 h, and 8 h) but not HeLa229 cells (C, infection focus, 4 h, 6 h, and 8 h) (green, Shigella; red, DNA). Scale bar, 100 μm. (B and D) Computer-assisted image analysis was used to quantify the size of the infection foci in HT-29 cells (B) and HeLa229 cells (D). The average focus size was determined (****, P < 0.0001, unpaired t test).

Quantification of actin-based motility in HT-29 and HeLa229 cells.

HeLa cells have been widely used to decipher the mechanisms supporting S. flexneri actin-based motility. It was therefore unlikely that the striking differences in dissemination efficiency observed in HeLa229 and HT-29 cells originated from differences in cytosolic velocity. In agreement with this assumption, the numbers of bacteria associated with actin tails in the cytosol of infected cells were similar in HeLa229 and HT-29 cells (see Fig. S2A to C in the supplemental material). As expected, the velocities of the bacteria in the cytosol of infected cells were also similar in HeLa229 and HT-29 cells (see Fig. S2D). These experiments thus confirmed that the greater efficiency of bacterial spread observed in HT-29 cells was not related to differences in cytosolic motility.

Dynamics of protrusion and vacuole formation in HT-29 cells.

We next defined the sequence of events occurring in HT-29 cells upon S. flexneri spread from cell to cell. To visualize the formation of protrusions and vacuoles, we generated a stable HT-29 cell line that expressed a plasma membrane-targeted version of GFP. In order to track individual bacteria, we used a low multiplicity of infection resulting in the presence of 3 to 5 bacteria in the cytosol of infected cells 2.5 h postinfection. To minimize the deleterious effects of high-power laser illumination of host cells and bacteria, we acquired images of infected cells every 6 min. As shown in Movie S1 in the supplemental material and Fig. 2, as cytosolic and motile bacteria reached the cell cortex of primary infected cells, they formed plasma membrane protrusions that elongated into adjacent cells but remained connected to the sending cell through a membranous neck (Fig. 2A to C, protrusion). Membrane protrusions resolved into vacuoles, which displayed a continuous lining of the plasma membrane surrounding bacteria and were therefore no longer connected to the sending cell through a membranous neck (Fig. 2A to C, vacuole). The subsequent appearance of discontinuities within the otherwise continuous lining of the vacuole membrane probably reflected membrane rupture, which defined the moment at which the pathogen escaped vacuoles and gained access to the cytosol of the receiving cells (Fig. 2A to C, vacuole escape, arrowheads). Bacteria that escaped vacuoles resumed cytosolic motility (Fig. 2A, free bacteria, arrow) or remained associated with remnants of the vacuole membrane (Fig. 2B and C, free bacteria, arrow). Our tracking data of 40 bacteria in 10 infection foci indicated that more than 70% (29/40) of the bacteria formed productive protrusions (Fig. 2D, light blue lines) that resolved into vacuoles (Fig. 2D, yellow lines) from which the pathogen escaped and gained access to the cytosol of adjacent cells (Fig. 2D, red lines). Among those bacteria that gained access to the cytosol of adjacent cells, more than 60% (18/29) of the bacteria resumed cytosolic motility within the time frame of observation. Due to their great velocity and the 6-min time interval between images, most bacteria that resumed cytosolic motility were no longer tractable under our experimental conditions. These bacteria were depicted as free bacteria from the moment they escaped vacuoles and for the remaining time of the movie (Fig. 2D, red lines, 150-min time point). The remaining ∼30% (11/40) of the bacteria that did not gain access to the cytosol of adjacent cells either formed vacuoles from which the pathogen did not escape (Fig. 2D, yellow lines, 150-min time point) or formed protrusions that failed to resolve into vacuoles (Fig. 2D, blue lines, 150-min time point). These nonproductive protrusions stopped elongating and remained unresolved for the entire duration of the experiment (Fig. 2D, light blue lines, 150-min time point) or completely collapsed and retracted toward the sending cells, which brought the pathogen back to the cytosol of the primary infected cells (Fig. 2, dark blue lines, 150-min time point). On average, the dissemination process from the cytosol of the sending cell to the cytosol of the receiving cell was ∼60 ± 20 min long, during which period the bacteria first spent ∼20 ± 10 min in protrusions (Fig. 2D; protrusions, light blue lines) and then ∼35 ± 15 min in secondary vacuoles (Fig. 2D; vacuoles, yellow lines). Together, these results provide a characterization of the sequence and timing of events occurring upon S. flexneri dissemination in HT-29 cells and uncover the previously unappreciated notion that upon dissemination, the cytosolic pathogen S. flexneri spends a significant amount of time (∼60 min) in membrane-bound compartments, including protrusions and vacuoles.

Dynamics of protrusion and vacuole formation in HeLa229 cells.

We next used time-lapse confocal microscopy to investigate S. flexneri dissemination in HeLa229 cells. As shown in Movie S2 in the supplemental material and corresponding Fig. 3, HeLa229 cells supported the formation of membrane protrusions that elongated into neighboring cells (Fig. 3A to C, protrusions). However, in stark contrast with the situation observed in HT-29 cells, the formed protrusions did not resolve into vacuoles. Instead, elongated protrusions collapsed and retracted, which brought the bacteria back to the cytosol of the sending cell, where the pathogen multiplied (Fig. 3A to C, primary cell). These observations are in agreement with the observed accumulation of bacteria in primary infected cells at later time points, as shown in Fig. 1 (6- and 8-h time points). Our tracking analysis of 40 bacteria in 10 independent foci of infection revealed that 95% (38/40) of the protrusions formed in HeLa229 cells failed to resolve into vacuoles (Fig. 3D, blue lines). These results indicate that the spreading defect observed in HeLa229 cells infected with S. flexneri (Fig. 1) was due to a dramatic failure to resolve protrusions into vacuoles.

Protrusion and vacuole counts reflect the efficiency of the dissemination process.

Our previous results established that the vast majority of the protrusions formed in HT-29 cells resolved into vacuoles, from which the bacteria escaped and gained access to the cytosol of the receiving cells (Fig. 2D). All the various steps of this efficient spreading process, which include protrusions, vacuoles, and free bacteria in adjacent cells, were equally represented 4 h postinfection (90-min time points in Fig. 2D). In contrast, the striking spreading defect observed in HeLa229 cells was reflected by the overrepresentation of protrusions and the near absence of vacuoles and free bacteria 4 h postinfection (90-min time points in Fig. 3D). We confirmed the statistical significance of these observations by counting protrusions, vacuoles, and free bacteria in adjacent cells on samples collected and fixed 4 h postinfection (see Fig. S3 in the supplemental material). Thus, the numeration of protrusions, vacuoles, and free bacteria 4 h postinfection constitutes an accurate readout of dissemination efficiency.

STK11 is required for efficient S. flexneri dissemination in HT-29 cells.

We next investigated the potential origin of the differences observed between HeLa229 and HT-29 cells with respect to S. flexneri dissemination efficiency. In an RNA interference (RNAi) screen for host factors required for S. flexneri infection, we identified the cell polarity kinase STK11 (also known as LKB1) as a serine/threonine kinase required for bacterial dissemination. Interestingly, the genome of HeLa cells has undergone substantial rearrangements, and the locus encoding STK11 is no longer present in this cell line (24). Accordingly, STK11 expression was detectable by Western blot analysis in HT-29 cells but not in HeLa229 cells (see Fig. S4 in the supplemental material). To further validate the role of STK11 in S. flexneri dissemination in HT-29 cells, we used four independent small interfering RNA (siRNA) duplexes targeting STK11 expression and quantified their ability to silence STK11 expression and to confer a spreading defect using the assay presented in Fig. 1. We found that the STK11-targeting duplexes D2 and D3 conferred the strongest spreading defects (50 to 60% reduction in the size of infection foci) (Fig. 4A and B), which correlated with the strongest silencing efficiency at the mRNA and protein levels (Fig. 4C; see also Fig. S4). The correlation between silencing efficiency and strength of the observed spreading defect demonstrated that STK11 is specifically required for S. flexneri dissemination in HT-29 cells. We next analyzed protrusion and vacuole formation 4 h postinfection and found that both mock-treated and STK11-depleted cells supported the formation of protrusions that extended into adjacent cells. However, we observed a significant decrease in the numbers of the vacuoles formed in STK11-depleted cells (Fig. 4D). Altogether, these results indicate that SKT11 supports S. flexneri dissemination through the resolution of protrusions into vacuoles.

FIG 4.

STK11 is required for S. flexneri dissemination in HT-29 cells. (A) Representative images of HT-29 cells mock-treated (mock) or transfected with four individual siRNA duplexes (D1, D2, D3, and D4) targeting STK11 and infected with GFP-expressing S. flexneri for 8 h (green, Shigella; red, DNA). Scale bar, 100 μm. (B) Graph showing the relative size of infection foci in mock-treated and STK11-depleted cells for each individual siRNA duplexes (D1, D2, D3, and D4). *, P < 0.05; **, P < 0.0025. (C) Graph showing the relative silencing efficiency of individual siRNA duplexes as used in panel B. Values represent the means ± SDs of three independent experiments. *, P < 0.05; **, P < 0.0025. (D) Graph showing the counts of protrusions, vacuoles, and free bacteria in receiving cells in mock-treated (mock) and STK11-depleted (STK11 KD) cells using a 50 nM pool of siRNA duplexes D1, D2, D3, and D4 (12.5 nM each). Values represent the mean of 3 independent experiments. Statistical analysis: mock versus STK11 KD protrusions, P < 0.0001; mock versus STK11 KD vacuoles, P < 0.001; mock versus STK11 KD free bacteria, P < 0.0001 (unpaired t test).

STK11 expression restores S. flexneri dissemination in HeLa229 cells.

To determine whether reexpression of STK11 in HeLa229 cells would increase the efficiency of S. flexneri dissemination, we engineered a retroviral construct expressing the STK11 gene under the control of the cytomegalovirus (CMV) promoter. We infected HeLa229 cells with the corresponding retroviral construct and derived stable cell lines that displayed STK11 expression in HeLa229 cells (HeLa229-STK11 cells). The expression of STK11 in HeLa229 cells led to a striking increase in the efficiency of S. flexneri dissemination (Fig. 5A and B). As expected, treatment with siRNA duplexes targeting STK11 expression did not affect S. flexneri dissemination in HeLa229 cells (see Fig. S5A in the supplemental material). However, STK11 silencing abolished the increase in S. flexneri dissemination observed in STK11-expressing HeLa229 cells (see Fig. S5B), which correlated with efficient silencing (see Fig. S5C). The difference in dissemination efficiency in HeLa229 and STK11-expressing HeLa229 cells was not related to differences in actin tail formation (see Fig. S6). In stark contrast with the situation observed in the parental HeLa229 cells, S. flexneri formed numerous vacuoles in HeLa229-STK11 cells (Fig. 5C). Furthermore, the presence of free bacteria indicated that the pathogen efficiently escaped from the formed vacuoles (Fig. 5C). Thus, STK11 expression in HeLa229 cells was sufficient to promote the resolution of protrusions into vacuoles, which dramatically increased the efficiency of S. flexneri dissemination.

FIG 5.

STK11 expression restores S. flexneri dissemination in HeLa229 cells. (A) Representative images show HeLa229 and STK11-expressing HeLa229 cells (HeLa229-STK11) infected with S. flexneri for 8 h (green, Shigella; red, DNA). Scale bar, 50 μm. (B) Graph showing the quantification of the size of infection foci as shown in panel A. The average focus size was determined (****, P < 0.0001; unpaired t test). (C) Graph showing the counts of protrusions, vacuoles, and free bacteria in receiving cells among HeLa229 and STK11-expressing HeLa229 cells (HeLa229-STK11). Values represent the means of 3 independent experiments. Statistical analysis: HeLa229 versus HeLa229-STK11 protrusions, P < 0.0001; HeLa229 versus HeLa229-STK11 vacuoles, P < 0.0001; HeLa229 versus HeLa229-STK11 free bacteria, P < 0.0001 (unpaired t test).

STK11 controls the expansion of lateral cell-cell contacts in epithelial cells.

STK11 is a pleiotropic serine/threonine kinase that displays numerous functions in epithelial cells, including the establishment and maintenance of apical-basal polarity. Under our experimental conditions, STK11 depletion did not dramatically affect the formation of tight junctions, as reflected by the localization of the tight junction marker occludin (Fig. 6A). The identity of the apical domain was apparently maintained, as reflected by the localization of the apical marker villin (see Fig. S7 in the supplemental material). Although lateral cell-cell contacts were clearly established, we noticed that the height of the cells was markedly decreased (40%), reflecting a role for STK11 in promoting the expansion of the lateral domain in HT-29 cells (Fig. 6B). Interestingly, we found that STK11 expression in HeLa229 cells promoted a dramatic expansion of the lateral cell-cell contacts (Fig. 6C and D). It was previously suggested that E-cadherin, a major component of the lateral domain in epithelial cells, constitutes an important determinant of S. flexneri spread from cell to cell (17). We thus tested whether STK11 depletion may affect S. flexneri dissemination through modulation of E-cadherin expression. Immunofluorescence staining did not reveal striking differences in the amounts and localization of E-cadherin in mock-treated and STK11-depleted cells (Fig. 6A). As opposed to the situation observed in STK11-depleted cells, E-cadherin silencing with four independent siRNA duplexes completely prevented cell-cell contact formation and led to the formation of rounded cells that, perhaps not surprisingly, did not support S. flexneri dissemination any longer (see Fig. S8). Given the phenotypic differences observed in STK11- and E-cadherin-depleted cells, it is unlikely that STK11 depletion affects S. flexneri dissemination through modulation of E-cadherin expression. Together, these experiments suggest that STK11 may promote S. flexneri dissemination by specifying lateral cell-cell contacts that support the resolution of protrusions into vacuoles.

FIG 6.

STK11 controls the expansion of lateral cell-cell contacts. (A) Mock-treated (mock) and STK11-depleted (STK11 KD) HT-29 cells stained for E-cadherin and occludin (red, E-cadherin; green, occludin). E-cadherin staining shows the protein localization at the basolateral surface. Occludin staining shows the protein localization at the tight junctions. Scale bar, 5 μm. (B) Graph showing the relative height of mock-treated and STK11-depleted HT-29 cells as shown in panel A. (C) HeLa229 and STK11-expressing HeLa229 (HeLa229-STK11) cells were stained for actin (F-actin) (red, STK11; green, F-actin). Scale bar, 5 μm. (D) Graph showing the relative height of HeLa229 and STK11-expressing HeLa229 cells as shown in panel C. Values represent the means and standard deviations of 3 independent experiments (*, P < 0.05; unpaired t test).

Tyrosine kinase signaling in S. flexneri protrusions.

In a survey for potential markers of S. flexneri dissemination, we noticed that the membrane protrusions formed in HT-29 cells were highly enriched in phosphotyrosine residues, suggesting that the formation of protrusions correlates with high levels of tyrosine kinase signaling [Fig. 7A, HT-29 (Mock); see also Fig. S9 in the supplemental material]. Interestingly, phosphotyrosine residues were not detected in the protrusions formed in HeLa229 cells (Fig. 7B, HeLa2229; see also Fig. S9). However, STK11 expression restored high levels of phosphotyrosine residues in the protrusions formed in HeLa229 cells (Fig. 7B, HeLa229-STK11; see also Fig. S9). Quantification of the proportion of protrusions displaying phosphotyrosine residues revealed that 75% of the protrusions formed in HT-29 cells were phosphotyrosine positive (Fig. 7A, graph, mock), whereas 90% of the protrusions formed in HeLa229 were phosphotyrosine negative (Fig. 7B, graph, Hela229). Strikingly, more than 60% of the protrusions formed in HeLa229-STK11 were phosphotyrosine positive (Fig. 7B, graph, HeLa229-STK11). These results thus suggested a correlation between the levels of tyrosine kinase signaling in protrusions and the efficiency of protrusion resolution into vacuoles. To further establish the functional importance of tyrosine kinase signaling in protrusions, we tested the impact of several tyrosine kinase inhibitors on S. flexneri dissemination. Most of the tested inhibitors displayed obvious cell toxicity and were disregarded from further analyses. Interestingly, treatment of HT-29 cells with imatinib, also known as Gleevec, displayed no apparent signs of adverse effects on the establishment of cell-cell contact and the maintenance of cell polarity (see Fig. S10). However, more than 70% of the protrusions formed in imatinib-treated cells were phosphotyrosine negative [Fig. 7A, HT-29 (Imatinib) and graph, Mock versus Imatinib], and bacterial dissemination was severely impaired (Fig. 7C). This inhibition of bacterial dissemination was not related to defects in cytosolic tail formation (see Fig. S10). Importantly, imatinib treatment led to a significant decrease in the numbers of vacuoles formed in adjacent cells (Fig. 7D). Thus, the formation of STK11-dependent lateral cell-cell contacts competent for tyrosine kinase signaling promotes the resolution of S. flexneri protrusions into vacuoles in epithelial cells.

FIG 7.

Tyrosine kinase signaling is required for S. flexneri dissemination. (A) Representative images of dimethyl sulfoxide (DMSO)-treated (mock) and imatinib (100 nM)-treated HT-29 cells expressing a CFP membrane marker (blue), infected with CFP-expressing Shigella (blue), and stained for phosphotyrosine residues (yellow). The graph shows the counts of phosphotyrosine-positive and -negative protrusions in DMSO-treated cells (mock) or cells treated with imatinib. (B) Representative images of HeLa229 cells and HeLa229-STK11 cells expressing a CFP membrane marker (blue) infected with CFP-expressing Shigella (blue) and stained for phosphotyrosine residues (yellow). The graph shows the quantification of phosphotyrosine-positive and -negative protrusions in HeLa229 and STK11-expressing Hela229 cells (HeLa229-STK11). (C) Representative images of HT-29 cells treated with DMSO (mock) or treated with imatinib and infected with GFP-expressing S. flexneri (green, Shigella; red, DNA). Scale bar, 50 μm. The graph showings the relative size of the infection foci formed in DMSO-treated (mock) and imatinib-treated cells. The average focus size was determined (****, P < 0.0001; unpaired t test). (D) Graph showing the counts of protrusions, vacuoles, and free bacteria in receiving cells in DMSO-treated (mock) and imatinib (100 nM)-treated HT-29 cells. Values represent the means of 3 independent experiments. Statistical analysis: mock versus imatinib protrusions, P < 0.0001; mock versus imatinib vacuoles, P < 0.001; mock versus imatinib free bacteria, P < 0.0001 (unpaired t test).

DISCUSSION

The dissemination of Shigella flexneri relies on actin-based motility in the cytosol of infected cells and formation of plasma membrane protrusions that resolve into vacuoles in adjacent cells. While the mechanisms supporting actin-based motility are now fairly well understood, the cellular processes supporting the formation and resolution of S. flexneri protrusions are comparatively unknown. To address this knowledge gap, we have developed several assays to quantify S. flexneri dissemination (Fig. 1) and capture the dynamics of protrusion formation, resolution into vacuoles, and bacterial escape from the formed vacuoles (Fig. 2, 3, and 4). The comparative analysis of S. flexneri dissemination in HT-29 and HeLa229 cells revealed that S. flexneri disseminates poorly in HeLa229 cells (Fig. 1). As expected, the bacteria displayed actin-based motility in HeLa229 cells, but the resolution of protrusions into vacuoles failed dramatically. Since our RNAi-based genetic investigations with HT-29 cells revealed a role for STK11 in S. flexneri infection and since previous studies reported that STK11 is not expressed in HeLa cells, we further investigated the role of STK11 in bacterial dissemination. We found that depletion of STK11 decreased the efficiency of S. flexneri dissemination in HT-29 cells. Moreover, we showed that restoring STK11 expression improved bacterial dissemination in HeLa229 cells. Our results revealed a striking correlation between STK11 expression and the expansion of the lateral domain of cells, which may contribute to the efficiency of S. flexneri dissemination by increasing the surface of cell-cell contacts available for the formation of protrusions that efficiently project into adjacent cells. However, since S. flexneri formed as many protrusions in HeLa229 cells and HT-29 cells, it is likely that the main role of STK11 in S. flexneri dissemination is to specify the identity of the lateral domains and therefore the components of cell-cell contacts. This notion is in agreement with our observation that STK11-expressing cells form protrusions that are competent for tyrosine kinase signaling, as opposed to cells that do not express STK11, such as HeLa229 cells. It is noteworthy that the actin tails formed by S. flexneri in the cytosol do not display phosphotyrosine residues (25), suggesting that the tyrosine kinase signaling events observed in protrusions are most likely supported by the presence of the plasma membrane. Importantly, various receptor tyrosine kinases (RTKs) have been reported to specifically localize to the basolateral domain of polarized cells, a process that relies on regulated vesicular trafficking (26). We therefore speculate that the role of STK11 in S. flexneri dissemination is to orchestrate basolateral vesicular trafficking and target RTKs to cell-cell contacts, where the protrusions formed by S. flexneri project into adjacent cells.

The functional relevance of RTK signaling was further substantiated by the observation that treatment with the RTK inhibitor imatinib strongly inhibited S. flexneri dissemination. In contrast to the situation observed with other pathogens, such as vaccinia virus (27), imatinib did not inhibit S. flexneri dissemination by interfering with actin-based motility. Our results indicate that imatinib inhibited S. flexneri dissemination by interfering with vacuole formation through a mechanism that remains to be determined. Imatinib was discovered in a program designed to identify inhibitors of the platelet-derived growth factor (PDGF) receptor in the context of cancer therapy (28). However, imatinib displays potent inhibitory activities on additional tyrosine kinases, including ABL and KIT (29, 30). Additional work will thus be required to identify the tyrosine kinase(s) and the downstream effectors potentially involved in S. flexneri dissemination.

Numerous studies have emphasized the central role of tyrosine kinase signaling in the regulation of cellular responses in health and disease, and various drugs are now available to target the activity of the involved signaling components (30). We have recently shown that treatment with ibrutinib, a specific inhibitor of the tyrosine kinase Btk, inhibited S. flexneri dissemination by interfering with actin-based motility (22). Here, we showed that treatment with the tyrosine kinase inhibitor imatinib inhibited S. flexneri dissemination by interfering with the resolution of protrusions into vacuoles. Thus, tyrosine kinase signaling emerges as a central theme in S. flexneri dissemination. We suggest that in addition to standard antimicrobial compound-based therapy, treatment with tyrosine kinase inhibitors, such as imatinib, may represent an effective strategy to interfere with S. flexneri infection, as previously proposed for the treatment of smallpox (27).