Outer membrane protein A (OmpA) is a multifaceted predominant outer membrane protein of Escherichia coli and other Enterobacteriaceae whose role in the pathogenesis of various bacterial infections has recently been recognized. Here, the role of OmpA on the virulence of Shigella flexneri has been investigated. An ompA mutant of wild-type S. flexneri 5a strain M90T was constructed (strain HND92) and it was shown to be severely impaired in cell-to-cell spreading since it failed to plaque on HeLa cell monolayers. The lack of OmpA significantly reduced the levels of IcsA while the levels of cell associated and released IcsP-cleaved 95 kDa amino-terminal portion of the mature protein were similar. Nevertheless, the ompA mutant displayed IcsA exposed across the entire bacterial surface. Surprisingly, the ompA mutant produced proper F-actin comet tails, indicating that the aberrant IcsA exposition at bacterial lateral surface did not affect proper activation of actin-nucleating proteins, suggesting that the absence of OmpA likely unmasks mature or cell associated IcsA at bacterial lateral surface. Moreover, the ompA mutant was able to invade and to multiply within HeLa cell monolayers, although internalized bacteria were found to be entrapped within the host cell cytoplasm. We found that the ompA mutant produced significantly less protrusions than the wild-type strain, indicating that this defect could be responsible of its inability to plaque. Although we could not definitely rule out that the ompA mutation might exert pleiotropic effects on other S. flexneri genes, complementation of the ompA mutation with a recombinant plasmid carrying the S. flexneri ompA gene clearly indicated that a functional OmpA protein is required and sufficient for proper IcsA exposition, plaque and protrusion formation. Moreover, an independent ompA mutant was generated. Since we found that both mutants displayed identical virulence profile, these results further supported the findings presented in this study.
Citation: Ambrosi C, Pompili M, Scribano D, Zagaglia C, Ripa S, Nicoletti M (2012) Outer Membrane Protein A (OmpA): A New Player in Shigella flexneri Protrusion Formation and Inter-Cellular Spreading. PLoS ONE 7(11): e49625. doi:10.1371/journal.pone.0049625
Editor: Esteban Chaves-Olarte, Universidad de Costa Rica, Costa Rica
Received: July 19, 2012; Accepted: October 11, 2012; Published: November 14, 2012
This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Funding: This work was supported by Ministero della Ricerca e dell’Istruzione (PRIN projects 2009/KJ9SRT_002 and 2007/W52X9M_003) and by Faculty 60% funds of the University “G. D’Annunzio” of Chieti granted to M.N. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Shigella flexneri is a Gram-negative facultative intracellular pathogen that causes bacillary dysentery, a major public-health problem principally in developing countries. Most of the genes required for invasion of epithelial cells and cell-to-cell spread are located on the large virulence plasmid. Important aspects of Shigella pathogenesis are the ability to penetrate and multiply within colonic epithelial cells , . Invasion of epithelial cells proceeds by bacteria-induced phagocytosis, rapid lysis of the phagocytic vacuole, multiplication inside the cytoplasm of host cells, and subsequently intra- and inter-cellular spreading of the infection within epithelial tissues by polymerizing cell actin and forming long F-actin comet tails which propel bacteria through the cell cytoplasm and to the cell periphery , , , , . Shigella cell-to-cell spread requires the expression and polar surface exposition of IcsA (VirG), a 110-kDa autotransporter protein encoded on the Shigella large virulence plasmid. Once translocated across the outer membrane (OM), IcsA exposes its N-terminal α-domain (the passenger domain) on the bacterial surface interacting with the eukaryotic proteins vinculin and neural Wiskott-Aldrich syndrome protein (N-WASP). N-WASP then recruits the host Arp2/3 complex, which induces polymerization of host globular actin into filamentous actin and cross-links these new filaments at 70° angles , , , , , . In exponentially-growing bacteria IcsA is found to be exclusively exposed at the old bacterial pole , . Although the mechanism driving the polar localization of IcsA is still unclear, several experimental evidences indicate that IcsA inserts directly at the pole , , . Furthermore, Shigella moves to adjacent cells via protrusion formation without leaving the intracellular compartment. Protrusions are membrane-bound cell extensions that are driven by the bacterium and that propel it into adjacent cells. Protrusions, which may extend tens of microns from the cellular surface, are characterized by the presence of a bacterium at its tip . By a process which likely resembles macropinocytosis, contact with the membrane of an adjacent cell is followed by the uptake of the bacterium , leading to the spreading of the infection to neighboring epithelial cells. Several host cell proteins have been implicated in protrusion-mediated Shigella cell-to-cell spread, suggesting that a distinct set of actin regulatory factors interacts with motile bacteria after they contact the plasma membrane . Although actin polymerization and assembly are required for protrusion formation, the specific molecular mechanisms responsible for this phenomenon are poorly defined. Furthermore, recent experimental evidences indicate that actin nucleation processes, involved in protrusion formation, may be independent of the activity of the Arp2/3 complex . In this context, it has been recently reported that protrusion formation and inter-cellular spreading depend on actin polymerization that requires the activation of the Diaphanous formin Dia . Formins are a family of ubiquitous expressed proteins that, in contrast to the Arp2/3 complex, initiate de novo actin polymerization leading to cross-linking of actin polymers in parallel arrays .
OmpA is a 35 kDa monomeric protein embedded in the bacterial OM as a β-barrel protein, highly conserved among Gram-negative bacteria . OmpA is thought to play a pivotal role, along with other bacterial components, in the structural integrity of the OM and it has been reported to be an important virulence factor of several human pathogens , . Recent studies have suggested that OmpA, likely because of its capacity to interact with a variety of host tissues, has to be considered a pleiotropic molecule which plays different roles in the mechanism of virulence of different bacterial pathogens. Among other important features, it has been reported that OmpA may serve as an adhesin/invasin as well as an immune evasin, a participant in biofilm formation, a receptor for several bacteriophages and as a mediator of F-dependent conjugation , , . Furthermore, in an experimental animal model, OmpA of S. flexneri 2a has been identified as a novel molecule coordinating the innate and adaptive immune responses, strongly indicating that OmpA may also represent a promising antigen in vaccine development , . In spite of its importance in the immune response of S. flexneri, to date no reports have documented the role of OmpA on the virulence of S. flexneri.
In this study we have investigated the role of OmpA in the mechanism of virulence of S. flexneri. Our results indicate that OmpA is absolutely required for proper IcsA presentation, plaque and protrusion formation.
Materials and Methods
Bacterial Strains, Plasmids, Cell Lines, and Growth Conditions
Bacterial strains and plasmids used in this study, and their relevant characteristics, are listed in Table 1. S. flexneri strains were routinely plated on tryptic soy broth agar plates containing 0.01% Congo red (CR). Escherichia coli and S. flexneri strains were grown in Luria Bertani broth (LB; Difco). When necessary, antibiotics were used at the following concentrations: chloramphenicol (Cm), 30 µg ml−1, kanamycin (Km), 25 µg ml−1. HeLa cells from ATCC were grown in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum and grown at 37°C in the presence of 5% CO2.
Construction of the ompA Mutant Strain HND92
The ΔompA mutation carried by the E. coli K-12 strain JW0940 was P1 transduced into the S. flexneri 5a wild-type strain M90T (Table 1). Transductants were selected on CR plates supplemented with kanamycin and their structure was verified by PCR analysis (using nearby ompA specific primers; Table 1) and sequencing. The absence of the OmpA protein was verified by Western blot. One transductant, named HND92, was chosen for further studies.
DNA extraction, isolation of plasmids, restriction digestion, electrophoresis, purification of DNA fragments, and electroporation were performed by standard methods . Real Time PCR analysis of icsA expression was performed as previously described . The entire ompA gene, including its promoter region, was amplified by PCR, using the primer pair ompA2FW and ompA2RV (Table 1), designed on available genome sequence of strain M90T (GenBank CM001474.1). The PCR product was BamHI/HindIII digested and cloned into the corresponding sites of the pACYC184 vector (Fermentas), generating plasmid pOA. The correct insertion of the ompA gene in pACYC184 was verified by PCR and sequencing prior to be introduced by electroporation into strains HND92 and HND99 (Table 1) for complementation experiments. OmpA expression was verified by Western blot analysis.
Whole Cell Extracts, Secreted Proteins and Cell Fractionation
Whole cell extracts (WCEs) were prepared by lysing concentrated exponentially-growing bacteria in 1×Laemmli buffer. WCEs of infected HeLa cells were prepared by directly lysing washed infected monolayers with 1×Laemmli buffer. Type III secretion (T3S) system secreted proteins were prepared from bacteria grown to mid-log phase as previously described . Briefly, bacteria were washed and re-suspended in 1×PBS, CR was added to a final concentration of 7 µg ml−1, and bacteria were incubated for 30 min. After incubation, the bacteria were pelleted by centrifugation and the supernatant fraction was filtered through a 0.22 µm-pore filter. Secreted proteins were concentrated by trichloroacetic acid (TCA) treatment. TCA-pellets were re-suspended in 50 µl of 1×Laemmli buffer and stored at −20°C. Crude membrane and cytosolic/periplasmic bacterial fractions were prepared from equal numbers of exponentially-growing bacteria. Bacterial pellets were re-suspended in sonication buffer (150 mM NaH2PO4, 300 mM NaCl, 1 mM PMSF, and 1 mg ml−1 lysozyme), and incubated for 30 min on ice. At the end of the incubation period, cells were disrupted by sonication (Soniprep MSE 150, Sanyo). After removal of cell debris, the cleared-lysates were centrifuged at 18,000×g for 30 min at 4°C, and pellets, corresponding to the crude membrane fraction, were re-suspended in 1×Laemmli buffer, while supernatants, corresponding to the cytosolic and periplasmic fractions, were concentrated by TCA precipitation. TCA-precipitated proteins were washed with 90% acetone and re-suspended in 1×Laemmli buffer. When indicated, fractions were split and incubated at 37°C or at 100°C for 10 min.
Sodium Dodecylsulphate-polyacrylamide Gel Electrophoresis (SDS-PAGE) and Western Blot
Proteins were resolved by 12.5% Tris-glycine SDS-PAGE , and transferred onto PVDF membranes (Hybond-P GE-Heathcare Bio-Sciences). Western blot analysis was performed using the following antibodies: rabbit polyclonal anti-OmpA and anti-IcsA, mouse monoclonal anti-IpaB and anti-IpaC antibodies (gifts from N.V. Prasadarao, M.L. Bernardini and A. Phalipon, respectively); polyclonal anti-host cytosolic phospholipase A2 (cPLA2α) and anti-phospho-cPLA2α antibodies were from Cell Signal Technologies (Danvers, MA) and Santa Cruz Biotechnology (Santa Cruz, CA), respectively. Anti-mouse and anti-rabbit conjugated to horseradish peroxidase (Bio-Rad) were used as secondary antibodies. Blots were visualized by enhanced chemiluminescence system (GE-Healthcare Bio-Sciences).
Preparation of LPS and Serum Sensitivity Assay
LPS were prepared from overnight bacterial cultures essentially as described by Hitchcock and Brown . Aliquots of LPS preparations were incubated at 100°C for 5 min prior to loading on 13% Tris-glycine SDS-PAGE, and silver-stained as described . For serum sensitivity assays, the wild-type strain, the ompA mutant strain HND92, and the E. coli K-12 strain DH10b were grown to mid-log phase, collected by centrifugation, washed and diluted in PBS (pH 7.4). S. flexneri (5×105 CFU ml−1) and E. coli (5×106 CFU ml−1) strains were suspended in the presence of 10% normal human serum (NHS, Lonza), and incubated at 37°C for 2 hrs. The number of CFUs was determined at 30 min intervals by plating serial dilutions onto CR plates. Serum heated at 56°C for 30 min (heat-inactivated normal human serum; HIS) was used as control in all experiments. Percent survival was calculated as the average of CFUs that survived exposure to NHS divided by the CFUs that survived exposure to HIS at each time point×100.
Invasiveness, Plaque Assay, Actin Tails and Protrusion Formation
Invasion of non-confluent HeLa cell monolayers was carried out using the gentamicin protection assay at a multiplicity of infection (MOI) of 20, as previously described . The percentage of cell invasion (percent of HeLa cells containing three or more intracellular bacteria) and the formation of protrusions (defined as extensions of the plasma membrane outside the normal contour of the HeLa cell that extended more than a bacterial length and that contained a bacterium at its tip) were evaluated with Giemsa-stained infected monolayers. The frequency of membrane protrusions was quantified 1 h post-infection by enumerating membrane protrusions divided by the number of infected HeLa cells. A minimum of 300 infected HeLa cells were considered for each experiment. Multiplication of bacteria within HeLa cells was assayed essentially as previously described . Non-confluent HeLa cells were infected (MOI = 100) centrifuged for 20 min at 37°C. Washed monolayers were covered with 2 ml of minimal essential medium containing gentamicin (50 µg ml−1) (time zero). Incubation was usually carried out for 5 h and two plates were removed 1, 3 and 5 hours. One plate was washed and Giemsa stained to calculate the percentage of infected cells at each time point. The other plate was washed, cells were trypsinized, counted and then lysed with 0.5% of sodium deoxycholate in distilled water. Dilutions of bacterial suspensions were plated onto CR plates. The average number of bacteria per infected cell was calculated. Experiments were repeated at least three times for each strain tested. Results represent the means of three experiments. Plaque assays were performed on confluent HeLa cell monolayers, as previously described , using a MOI of 4. After 72 h, cells were fixed and plaques were visualized by Giemsa-staining. To detect F-actin comet tails, infected HeLa cells were fixed for 10 min in 3.7% (w/v) paraformaldehyde, permeabilised with 0.2% Triton X-100 for 10 min and actin tails visualized by staining with rhodamine-conjugated phalloidin (Sigma). Bacterial and cellular DNAs were labelled with 0.2 μg ml−1 4′,6′-diamidino-2-phenylindole (DAPI, Molecular Probes). The frequency of actin tails was calculated at 2 h post-infection, in samples of HeLa cells carrying up 8 to 10 internalized bacteria, as the number of bacteria displaying apparently normal F-actin comet tails divided by the total number of intracellular bacteria. A minimum of 100 infected HeLa cells were considered for each experiments. Experiments were repeated at least three times. Images were recorded with a Leica DM5000B microscope equipped with DFX340/DFX300 camera and processed using the Leica Application Suite 2.7.0.R1 software (Leica).
Quantitative Detection of IcsA
Quantification of the relative amounts of IcsA on intact bacteria was achieved essentially as previously described  as well as by immuno-dot blot analysis. Briefly, equivalent volumes of exponentially-growing bacteria [normalized to the optical density at 600 nm (OD600)] were washed in PBS and fixed for 15 min in 3.7% (w/v) paraformaldehyde. Washed bacteria were then re-suspended in a 1∶100 dilution of S. flexneri type 5 antiserum (Denka Seiken) or in a 1∶100 dilution of rabbit anti-IcsA antiserum and incubated for 1 h. After washing, the bacterial suspensions treated with the anti-IcsA antiserum were suspended in a 1∶100 dilution of HRP-conjugated goat anti-rabbit secondary antibody (Bio-Rad) and incubated for 1 h. After washing, bacteria were re-suspended in 100 µl of peroxidase substrate (Rockland). After 10 min of incubation, the supernatants of each bacterial suspension were transferred to a 96-well flat-bottom plate and absorbance was measured at 370 nm using a microplate reader equipped with Gen5 data analysis software (BioTek Instruments). Immuno-dot blot experiments were also carried out to quantify IcsA on intact bacteria . Briefly, exponentially-growing bacteria were pelleted by centrifugation and re-suspended to equivalent volumes in PBS and 5 µl of each bacterial suspension was spotted onto a nitrocellulose membrane (Whatman) and allowed to air dry. Membranes were then processed using standard Western blotting procedures using diluted (1∶20.000) rabbit polyclonal anti-IcsA antibody. Densitometry was performed using ImageJ software .
Assessment of Membrane Integrity
To assess membrane integrity, the red- and green-fluorescent dyes propidium iodide (PI) and SYTO9 from the LIVE/DEAD BacLight kit (Molecular Probes, Invitrogen) were used in combination, following manufacturer’s instructions. Exponentially-growing bacteria (about 5×108 bacteria ml−1) were washed twice with 0.85% NaCl and re-suspended in 1 ml of 0.85% NaCl containing equal volumes (3 µl each) of PI (20 mM in DMSO) and SYTO 9 (3.34 mM in DMSO). The mixture was incubated at 25°C in the dark for 15 min. Samples of stained cells (200 µl each) were dispensed into separate wells of a 96-well flat-bottom microplate and fluorescence determined with a microplate reader equipped with Gen5 data analysis software (BioTek Instruments). Excitation wavelength was centered at 485±20 for both dyes, while emission wavelengths were centered at 528±20 and 645±40 for SYTO9 and PI, respectively. The intensities of the green (SYTO9) and red (PI) emission of each suspension were acquired as fluorescence intensity units and the green/red ratios were calculated. Isopropanol-treated cells were used as controls. Three independent experiments were performed, each in duplicate.
Construction of the ompA Mutant Strain HND92
To investigate the role of OmpA in S. flexneri virulence, a ΔompA mutation carried by the E. coli K-12 strain JW0940 (Table 1) was introduced into the wild-type S. flexneri strain M90T by P1 transduction. The correct structure of tranductants was verified (see Material and Methods for details) and one of them, named HND92, was used in all experiments. HND92 was able to bind Congo red and did not present growth defects since, in LB medium, there was no difference between the optical density of HND92 and that of the wild-type strain (data not shown). Exploiting the known cross-reactivity between enterobacterial species , , , we used polyclonal rabbit anti-OmpA antibody, raised against the E. coli OmpA protein, to detect OmpA in S. flexneri. A band of the expected size of OmpA (35 kDa) was detected by Western blot of WCE of the wild-type strain while no detectable band corresponding to OmpA was observed in WCE of HND92 (Fig. 1). Complementation of the ompA mutation with plasmid pOA, which carries the ompA gene and its promoter region cloned into the plasmid vector pACYC184 (Table 1), restored OmpA expression (Fig. 1). Moreover, proper localization within the membrane fraction and folding of OmpA was also assayed. Native OmpA of E. coli is known to present typical heat-modifiable mobility of β-barrels in SDS-PAGE which are known to form a compact SDS-resistant structure at low-temperature (37°C) that unfolds upon heating at higher temperature (100°C) , . Indeed, both OmpA of S. flexneri and OmpA expressed by recombinant plasmid pOA behaved as native E. coli OmpA, since both proteins were found to be strictly localized within the membrane fraction and presented the expected heat-variable mobility (Fig. 1 and data not shown). These results indicated that OmpA of S. flexneri presented the same relevant biological characteristics of OmpA of E. coli.
OmpA expression by the wild-type S. flexneri 5a strain M90T (WT), the ompA mutant (ompA), the ompA mutant complemented with plasmid pOA (ompA/pOA) and the ompA mutant complemented with the empty vector pACYC184 (ompA/pv) is shown.WCE, whole cell extract; M, crude membrane fraction; C, cytosolic and periplasmic fraction. Fractions were suspended in 1×Laemmli buffer and heated at 37°C or at 100°C for 10 min before SDS-PAGE to evaluate the unfolded (U) and folded (F) forms of OmpA , . PVDF filters were probed using an anti-OmpA polyclonal antiserum. The positions of U and F forms of OmpA are indicated by arrows at the right side. Size markers (kDa) are shown at the left side.
The ompA Mutation does not Affect IpaABCD Secretion
To assess whether the ompA mutation affected the ability of S. flexneri to synthesize and secrete IpaABCD effector proteins, we compared IpaABCD synthesis and secretion in the wild-type and in the ompA mutant. Exponentially-growing bacteria were incubated in the presence of CR in order to activate the T3S system , . SDS-PAGE analysis of Comassie brilliant blue staining of whole bacterial extracts and of concentrated culture supernatants indicated that neither the production nor the secretion of IpaABCD proteins were affected by the inactivation of the ompA gene (Fig. 2A). Accordingly, immunoblotting using anti-IpaB and anti-IpaC antibodies showed no difference in the expression and secretion of both proteins (Fig. 2B). These results indicated that the ompA mutant has a functional T3S system.
Proteins present in WCEs and concentrated culture supernatants (SN) of the wild-type strain (WT) and of the ompA mutant (ompA), following exposure of bacteria to 7 µg ml−1 of Congo Red (CR), were stained with Comassie brilliant blue (A) or transferred to a PVDF membrane and probed using anti-IpaB and anti-IpaC antibodies (B). The positions and names of the various proteins are indicated at the right side. Size markers (kDa) are shown at the left side.
The ompA Mutant Presents an Altered Invasive Phenotype
The ability of the ompA mutant to invade and to multiply within semi-confluent HeLa cell monolayers was compared to that of wild-type, The ompA mutant was proficient both in host cell invasion and intracellular multiplication (Fig. 3). Interestingly, at 5 h post-infection HND92 displayed a reduced invasion phenotype. On average, we found that the number of HND92 bacteria in the cytoplasm was higher than that of wild-type (216±22 and 158±18 bacteria per infected HeLa cell, respectively). In contrast, whereas the percentage of cells infected by the wild-type strain was 82.1±12.0%, the proportion of cells infected by HND92 was 46.2±6.3% (P≤. 0.05). Moreover, at 3 h post-infection striking differences regarding the intracellular bacterial distribution were observed in Giemsa stained monolayers (Fig. 3). While the wild-type strain displayed the characteristic random distribution within the cytoplasm of Hela cells, the ompA mutant showed a less randomly distributed behavior with bacteria organized in parallel arrays and which seemed to be entrapped within the cell cytoplasm. This intra-cellular distribution was reminiscent of that of mlaA/vacJ S. flexneri mutants impaired in inter-cellular spread , and significantly different to that shown by the icsA mutant strain SC560 (Table 1) which displayed bacteria growing in cluster in the cell cytoplasm (Fig. 3, insets). Moreover, random distribution of bacteria within the cytoplasm of HeLa cells was restored when the ompA mutant was complemented with plasmid pOA, indicating that OmpA is required for normal distribution of S. flexneri in the cell cytoplasm. Noteworthy, differently from wild-type, protrusions were noticed only sporadically in HeLa cells infected with the ompA mutant. This finding prompted us to measure the frequency of protrusion formation. Assuming a relative arbitrary value of 100% for the number of protrusions (per infected HeLa cells) formed by the wild-type strain 1 h post-infection, under these conditions the ompA mutant scored about 6.9±2.4%, while HND92/pOA displayed about 89.7±6.9% (P≤0,01). These results clearly indicated that the lack of OmpA was responsible of the protrusion defect of the ompA mutant. Cytoplasmic entrapment of bacteria and the dramatic decrease in protrusion formation were highly suggestive of impaired actin-based motility (ABM). Therefore, the ability of the ompA mutant to form F-actin comet tails and to plaque on HeLa cells was determined. As for the wild-type, the ompA mutant was found to be proficient in the formation of apparently normal F-actin comet tails (Fig. 4), as determined by measuring the frequency of bacteria presenting apparently normal actin tails per infected HeLa cell (22.2±1.5% and 24.2±2.7% for wild-type strain and the ompA mutant, respectively; P≥0.5). Unexpectedly, despite the formation of proper actin tails, the ompA mutant failed to plaque on confluent HeLa cell monolayers (Fig. 5), even when HeLa cells were infected at higher MOIs or extending the period of incubation (data not shown). The ability to produce plaques at almost wild-type level was rescued by complementing the ompA mutant with plasmid pOA (0.9–1.2 mm against 1.0–1.2 mm of wild-type, after 72 h of incubation) (Fig. 5), indicating that a functional OmpA protein is required and sufficient for the inter-cellular spread of S. flexneri.
Semi-confluent HeLa cell monolayers were infected with the wild-type strain (WT), the ompA mutant (ompA), and the ompA mutant complemented with plasmid pOA (ompA/pOA), at a MOI of 20. Monolayers were fixed with methanol and Giemsa-stained for microscopic evaluation at 1 h and 3 h post-infection (insets). The icsA mutant SC560 strain was included as control (icsA). Representative images of at least four independent experiments are shown. Scale bar, 10 µm.
Semi-confluent HeLa cell monolayers were infected with the wild-type strain (WT), the ompA mutant (ompA), and the ompA mutant complemented with plasmid pOA (ompA/pOA). Actin was stained with rhodamine-conjugated phalloidin (red). Nuclei and bacterial DNA were stained with DAPI (blue). Representative images of double immunofluorescent microscopy are shown. Images are representatives of at least three independent experiments. Images were captured using a Leica camera and processed using Qwin software (Leica). Arrows point representative bacteria exhibiting normal F-actin comet tail formation. Scale bar, 10 µm.
Confluent HeLa cell monolayers were infected with the wild-type strain (WT), the ompA mutant (ompA), and the ompA mutant complemented with plasmid pOA (ompA/pOA), at a MOI of 4. The icsA mutant SC560 strain (icsA) was included as a plaque negative control. Cells were Giemsa-stained and photographed 72 h post-infection. Images are representative of at least three independent experiments.
Previous studies have reported that mutations of genes involved in the biosynthesis of the LPS of S. flexneri lead to the inability to plaque , . To determine whether the lack of OmpA could have altered LPS biosynthesis, thus causing the inability of the ompA mutant to plaque, we first evaluated the sensitivity of the ompA mutant to the killing of normal human serum (NHS), since it has been reported that wild-type LPS of S. flexneri is required to confer resistance to NHS killing . To this end, mid-log cultures of the ompA mutant, of the wild-type strain and of the E. coli K-12 DH10b (used as a serum-sensitive control) were incubated for 2 h (30 min intervals) in the presence of 10% of NHS or with heat-inactivated normal human serum (HIS). The percent survival to killing of the ompA mutant was indistinguishable to that of the wild-type strain while, as expected, the E. coli DH10b strain was highly sensitive to NHS. There was a greater than 5 log difference between the CFUs of E. coli DH10b rescued after 30 min of incubation in 10% NHS and the CFUs of wild-type and the ompA mutant after 2 h of incubation (data not shown). Next, LPS preparations were prepared as described in Materials and Methods. Aliquots were subjected to SDS-13%-PAGE and visualized by silver staining. Figure 6 shows that the ompA mutant produced the typical LPS structure seen in the wild-type strain, thus definitively ruling out that the lack of OmpA might have altered LPS biosynthesis and that the inability to plaque of the ompA mutant was due to LPS alterations.
LPS was extracted as described in Materials and Methods, resolved by SDS-13%PAGE and silver stained.
Polarly-localized IcsA is essential for ABM in Shigella , , . To study the influence of OmpA on IcsA surface exposition at the bacterial pole, indirect immunofluorescence experiments were conducted using a polyclonal IcsA antiserum. Differently from the typical polar IcsA caps displayed by the wild-type strain, the great majority of ompA immunostained bacteria had detectable IcsA on the entire bacterial surface [90.0±8.7% of the ompA mutant displayed IcsA diffused across the entire bacterial surface; 9.5±2.2% displayed apparently normal IcsA caps and 0.5% produced polar IcsA caps smaller than wild-type strain, while 55.3±8.2% of wild-type bacteria displayed typical normal IcsA caps, 26.2±5.9% also displayed IcsA exposed at the old bacterial pole but the caps appeared smaller than typical IcsA caps, and 18.5±3.7% displayed IcsA diffused across the entire bacterial surface (P<0.01). Under these experimental conditions the percentage of immunostained bacteria were 26.5% for the ompA mutant (n = 2950), and 31.3% for wild-type strain (n = 2880) (P>0.5)] (Fig. 7). Noteworthy, the majority (about 62.0±7.8%) of the ompA mutant immunostained bacteria presented IcsA polar reinforcements (data not shown). pOA-complementation restored wild-type IcsA exposition (Fig. 7). Taken together, these results indicated that OmpA, while it is not required for the ability of S. flexneri to invade and multiply within host cells as well as to produce proper F-actin comet tails, it is absolutely required for protrusion formation, cell-to-cell spread and proper IcsA surface exposition.
Exponentially-growing wild-type strain (WT), the ompA mutant (ompA), and the ompA mutant complemented with plasmid pOA (ompA/pOA) were analyzed by indirect immunofluorescence using an anti-IcsA polyclonal antiserum. Representative images from at least three independent experiments are shown. Scale bar, 5 µm.
To determine whether OmpA contributed to the presentation of IcsA on the bacterial surface, as well as to the release of IcsA* (the IcsP-cleaved 95 kDa polypeptide) into culture supernatants , , , WCE and culture supernatants of the ompA mutant, of its pOA-complemented derivative strain and of wild-type were analyzed by Western blot. Unexpectedly, when compared to wild-type, the ompA mutant displayed significant reduced levels of IcsA (56.6±4.8%; P≤0.05) and this reduced expression was due to the lack of OmpA since complementation with plasmid pOA increased these levels to 87.5±5.3% (Fig. 8A). It has been recently reported that a fraction of the IcsP-mediated cleavage products of mature IcsA remains cell associated (ca-IcsA*) so that it can be detected in WCE . Densitometry analysis of WCE and of supernatant fractions revealed that wild-type, the ompA mutant and the pOA-complemented strain displayed similar amounts of IcsA* and of ca-IcsA* (Fig. 8A and B). These results indicated that the reduction of IcsA expression observed in the ompA mutant did not influence the levels of IcsP-cleaved IcsA* and ca-IcsA*.
Western blot analysis of WCE (A) and supernatants (B) of exponentially-growing wild-type strain (WT), the ompA mutant strain HND92 (ompA), the pOA-complemented derivative strain (ompA/pOA) and of the icsA mutant strain SC560 (icsA). Blots were probed with an anti-IcsA antiserum. The positions of full length mature IcsA (IcsA) of IcsP-cleaved fragment (IcsA*) and of cell associated IcsA* (ca-IcsA*), are indicated by arrows at the left side. Quantitative surface immunodetection of IcsA: (C) dot blots of intact bacteria probed with anti-IcsA antibody; and (D) intact bacteria were treated with anti-IcsA and anti-S. flexneri antibodies. The amount of antibody bound to the bacterial surface was determined by labeling with HRP-conjugated secondary antibody and measuring the HRP enzymatic activity (OD370). Means and standard deviation of three independent experiments are shown. Blots are representative. Size markers (kDa) are shown at the right side.
The reduction of IcsA expression in the ompA mutant could be due either to reduced IcsA levels in the cell or by a reduction of the amount of IcsA detectable by antibody on the bacterial surface. To investigate these possibilities, we first investigated whether the reduction of IcsA expression was due to reduced icsA transcription. Real Time PCR analysis , conducted on cDNA preparations of wild-type and the ompA mutant, clearly indicated that the reduction of IcsA expression was not due to reduced transcription (data not shown). Furthermore, the amount of IcsA exposed at the surface of the ompA mutant and of the pOA-complemented derivative strain were determined and compared to that of wild-type by quantifying HRP enzymatic activity  as well as by immuno-dot blot experiments (see Materials and Methods for details). For each strain, the ratio of anti-IcsA-HRP enzymatic activity to anti-Shigella LPS signal was determined in order to normalize the IcsA-HRP signal to the number of bacteria. Interestingly, the HRP immunodetection assay indicated that IcsA levels on the surface of the ompA mutant were approximately 53.5±6.5% (ratio 0.23) compared to that of wild-type bacteria (ratio 0.43) (Fig. 8D). As expected, pOA complementation significantly restored wild-type levels (81.4±7.2; ratio 0.35). The amount of IcsA on the bacterial surface was also determined by quantitative dot-blot analysis. In agreement with the results of HRP enzymatic activity, the levels of IcsA on the bacterial surface, as measured by densitometry analysis, were found to be reduced to 51.5±5.9% of the wild-type strain, while pOA complementation increased these levels to 82.4±4.8% (Fig. 8C). Taken together, these results clearly indicated that the ompA mutant displayed reduced levels of antibody-accessible IcsA on the bacterial surface, that this reduction did not negatively influence the fractions of released IcsA* and of membrane-bound ca-IcsA*, and that it was indeed due to the lack of OmpA.
Membrane Permeability of the ompA Mutant
OmpA has been shown to play an important role in maintaining OM integrity of E. coli , . To study the effect of the lack of OmpA on membrane integrity of S. flexneri, we used the LIVE/DEAD BacLight kit (Invitrogen). Although this kit is commonly used to discriminate between active and dead bacteria, it is also used to differentiate between bacteria presenting intact or damaged membranes. The kit is composed of a mixture of the green- and red-fluorescent nucleic acid stains SYTO9 and propidium iodide (PI), respectively. The two fluorescent molecules differ both in their spectral characteristics and in the ability to penetrate bacterial cells. While SYTO9 enters all cells, PI enters only bacteria presenting altered membrane permeability. When used in combination, the presence of bacteria presenting defects in membrane integrity causes a reduction in the green fluorescence emission . Therefore, exponentially-growing bacteria were stained with a mixture of the fluorescent dyes and the emitted fluorescence was measured and expressed in arbitrary fluorescence intensity units (see Materials and Methods for details). The results obtained indicated that while the wild-type strain displayed a green/red ratio of 20.3±1.2 units, the ompA mutant scored 12.3±1.0, indicating increased membrane permeability (P<0.01). The defect in membrane permeability was due to the lack of OmpA since complementation with pOA, increased the green/red ratio to 19.8±1.1 units, a value comparable to that of the wild-type strain. These data indicated that OmpA is required for S. flexneri membrane integrity.
OmpA is not Required for Host Cytosolic Phospholipase A2 Activation
Several reports have highlighted the involvement of OmpA in pathogen-host-cell interactions , . Recently, it has been reported that activation of host cytosolic phospholipase A2 (cPLA2α) is important for host cell invasion as well as for triggering cell cytoskeletal rearrangements . In particular, using human brain microvascular endothelial cultured cells as a model to study neonatal meningitis, it has been shown that OmpA of an E. coli K1 strain isolated from a patient with neonatal meningitis triggers actin condensation through activation of cPLA2α-dependent pathways . To examine whether the OmpA of S. flexneri is involved in the activation of cPLA2α, HeLa cell monolayers were infected (1 h of infection) with the wild-type and the ompA mutant strains. Western blot analysis of cellular lysates, using anti-p-cPLA2α (which recognizes the phosphorylated active form of cPLA2α) and anti-cPLA2α (which recognizes both the active and the inactive forms of cPLA2α) polyclonal antibodies, failed to display differences between the ompA mutant and the wild-type strain (Fig. S1). These results seemed to exclude that OmpA of S. flexneri could be involved in the activation of cPLA2α.
Generation of an Independent ompA Mutant (HND99) of the S. flexneri Strain M90T
To confirm the results obtained with the ompA mutant strain HND92, we generated strain HND99 (Table 1), another non-polar ompA deletion mutant of the wild-type strain M90T, by using the method described by Datsenko and Wanner  (see Text S1 for technical details). Analysis of the virulence traits expressed by HND99 revealed that HND99 was indistinguishable to HND92. In fact, as HND92, HND99 was able to bind CR, failed to plaque on confluent HeLa cell monolayers while it was able to invade and to multiply within the cell cytoplasm at a rate similar to that of HND92 (39±8% of cells infected, and 252±18 bacteria per infected HeLa cell, 5 h post-infection) and displayed reduced protrusion formation (9.4±. 3.3% of infected cells, 1 h post-infection). Moreover, HND99 displayed IcsA exposed across the entire bacterial surface (93.2±3.0%) with the majority of immunostained bacteria presenting IcsA polar reinforcements (about 73.4±10.8%), and the levels of IcsA on the bacterial surface, as determined by immune-dot blot assay, were approximately 48.0±5.7% that of wild-type bacteria. Finally, complementation with recombinant plasmid pOA restored wild-type plaque size (0.9–1.3 mm), protrusion formation (78.5±12.3%) and proper IcsA exposition.
OmpA is a multifunctional major OM protein of E. coli and other enterobacteria . Since its discovery , OmpA has been the subject of intense study and recently the function of OmpA as a relevant virulence factor in the pathogenesis of various infectious diseases has been recognized . To our knowledge, no study concerning the relevance of OmpA in the pathogenesis of Shigella spp. has been reported. Moreover, it has been recently reported that OmpA of S. flexneri is recognized by the Toll-like receptor 2, indicating that OmpA could play a pivotal role in initiating host innate and adaptive immune responses, thus representing a potential vaccine candidate .
This study was aimed at examining the role of OmpA in the virulence of S. flexneri. To this end, an ompA mutant (strain HND92) of the wild-type S. flexneri 5a strain M90T was generated. The ompA mutant was able to bind Congo red, did not present growth differences with the parental strain and it was able to invade cultured epithelial cells. The percentage of HeLa cells that were invaded by the ompA mutant was significantly lower than wild-type, while the number of intra-cellular bacteria was higher than the number of M90T bacteria in the cytoplasm. This phenotype was highly reminiscent of S. flexneri mutants presenting altered cell-to-cell spread , . Moreover, the lack of OmpA did not affect IpaABCD synthesis and secretion and the formation of apparently proper F-actin comet tails (Fig. 2 and 4), indicating that the lack of OmpA had no influence on the activation of the T3S system as well as on the IcsA-mediated activation of actin-nucleating proteins.
Indeed, the ompA mutant was totally impaired in cell-to-cell spread since it was unable to plaque on HeLa cell monolayers (Fig. 5). The inability to plaque was likely due to the dramatically reduced levels of protrusion formation, seen in HeLa cells infected with the ompA mutant. Moreover, the intra-cellular distribution of the ompA mutant was considerably different to that of wild-type. Bacteria did not display the classical random distribution of wild-type, while appeared to be entrapped within the cytoplasm (Fig. 3, insets). Wild-type plaque size, proper intra-cellular distribution and protrusion formation were restored by introducing plasmid pOA into the ompA mutant, indicating that OmpA is required for of S. flexneri intra- and inter-cellular spread. These findings prompted us to evaluate IcsA exposition by indirect immunofluorescence. Noteworthy, the ompA mutant displayed IcsA exposed across the entire bacterial surface (Fig. 7), with evident IcsA polar reinforcements noticed in the majority of immunostained bacteria. We hypothesized that the presence of these IcsA polar reinforcements may account for the ability of the ompA mutant to form apparently normal F-actin comet tails. The aberrant exposition of IcsA on the entire bacterial surface was not due to increased IcsA expression since we unexpectedly detected even lower levels of the 110-kDa mature IcsA in WCE of the ompA mutant (56.6±4.8%), compared to wild-type (Fig. 8A). Moreover, quantitative surface immuno-dot blot analysis (Fig. 8C) and immuno-assay (Fig. 8D), which measure together the levels of mature IcsA and ca-IcsA* exposed on the surface of intact bacteria, fully confirmed the reduced levels of IcsA expression. Interestingly, despite reduced mature IcsA expression, the levels of ca-IcsA* detected in WCE and of IcsA* in supernatant fractions were similar (Fig. 8A and B). Complementation with plasmid pOA restored wild-type levels of IcsA expression. Furthermore, the reduced levels of mature IcsA detected in the ompA mutant were not due to reduced icsA transcription, as determined by Real-Time PCR analysis of cDNA preparations of the ompA mutant and of wild-type (data not shown), indicating that the reduced IcsA expression could be due to post-trancriptional regulation. Moreover, indirect immunofluorescence experiments showed that, differently from wild-type, IcsA/ca-IcsA* were accessible to the anti-IcsA antibody across the entire bacterial surface (Fig. 7). Therefore, we concluded that the lack of OmpA likely unmasked mature IcsA/ca-IcsA* at bacterial lateral surface.
Intact LPS molecules (smooth LPS) are important for virulence of S. flexneri and O antigen polysaccharide chains have been shown to influence IcsA exposition on the bacterial surface and ABM within host cells , . The inability of the ompA mutant to plaque and IcsA exposition across the entire bacterial surface are phenotypes reminiscent of rough LPS mutants, although rough mutants have been reported to be unable to form proper F-actin comet tails , , a characteristic not shared by the ompA mutant (see above). To unravel this point, the LPS structure of the ompA mutant was evaluated and compared to that of the parental strain either indirectly, by determining the resistance to NHS which is known to correlate with a smooth LPS, or directly by silver staining of LPS preparations. Since the ompA mutant presented the same level of resistance to NHS and displayed the same LPS structure of the parental strain (Fig. 6), we excluded that the inability to plaque and to present IcsA exposed across the entire bacterial surface of the ompA mutant was due to altered LPS biosynthesis.
The current model of the distribution of IcsA on the S. flexneri OM indicates that, following its secretion through the Sec translocon, IcsA is strictly delivered and inserted into the OM at the old bacterial pole by the aid of specific proteins that recognize and bind to IcsA , , , , . Subsequently, membrane-bound IcsA diffuses to lateral bacterial surface by membrane fluidity, forming a gradient that decreases in concentration with distance from the old pole and IcsP is thought to be involved in the sharpening of the IcsA gradient , . IcsA at lateral bacterial surface is likely masked by LPS molecules, probably in combination with specific OM proteins, in order to hamper its function in ABM at those sites , . To explain the formation of typical IcsA caps in S. flexneri it has been hypothesized that LPS molecules of shorter length may be preferentially localized at the old cell pole, thus allowing exposition and function of IcsA only at this site .
Mutations that alter the synthesis or assembly of any of the OM components may cause permeability defects . OmpA, an extremely abundant protein (about 100.000 molecules/cell) and a predominant antigen in enterobacterial OM, has been shown to play an important role in maintaining E. coli OM integrity , , . By using a fluorescence approach (the BacLight kit, Invitrogen), we have shown that the lack of OmpA increased membrane permeability of S. flexneri . Although the mechanisms for targeting and assembling proteins into the OM are largely unknown, it is possible to envisage that the lack of OmpA may have considerably altered the homeostatic control mechanism that coordinates the overall OM assembly process in S. flexneri . In this context, it is conceivable that alteration of the OM composition due to the lack of OmpA (we have shown that OmpA is required for S. flexneri membrane integrity; see above), likely mislocalized newly synthesized lipids (LPS and phospholipids) as well as OM proteins, so that only a fraction of mature IcsA undergoes proper insertion into the OM. We hypothesized that the fraction of mature IcsA molecules which are not assembled into the OM might undergo protein degradation, thus accounting for the reduced IcsA expression seen in the ompA mutant. In line with the finding that the ompA mutant displayed the same LPS structure of the parental strain, it is conceivable to hypothesize OmpA as a possible LPS interactor required for the masking of mature IcsA and/or of ca-IcsA* at lateral bacterial surface.
Noteworthy, while the aberrant IcsA exposition across the bacterial surface did not influence the ability of IcsA to interact with host proteins vinculin and N-WASP to nucleate F-actin polymerization only in correspondence of the old bacterial pole (where we found IcsA polar reinforcements; see above), the lack of OmpA dramatically affected protrusion formation. Although ABM is absolutely required for protrusion formation in Shigella spp., the specific molecular mechanisms are poorly understood . It has been recently suggested that an actin nucleation process, independent of the Arp2/3 complex, may be involved in productive protrusion formation . One intriguing possibility is that OmpA of Shigella might also interact, directly or indirectly, with specific host proteins to trigger invasion of neighboring cells. Several host proteins like the Diaphanous formin Dia, ezrin, vinculin, and cadherins have been reported to be implicated in the process of protrusion-mediated cell-to-cell spread of Shigella spp. . Experiments are underway to evaluate the importance of the interactions between OmpA and these host proteins in the process of protrusion formation.
Moreover, OmpA interaction with host cPLA2α has been also reported . Apart from its role in liberating arachidonic acid from membrane phospholipids, cPLA2α has been shown to be important for host cell invasion as well as for triggering host cell cytoskeletal rearrangements , . In particular, it has been demonstrated that OmpA of a neonatal meningitis causing E. coli K1 strain triggers actin condensation via activation of cPLA2α-dependent pathways . On this basis, we compared the ability of the ompA mutant and of the wild-type strains to activate cPLA2α. Since no difference in cPLA2α activation was observed in extracts of HeLa cells infected with the two strains (Fig. S1), we concluded that, differently from the E. coli K1 strain, the OmpA of S. flexneri was not involved in cPLA2α activation.
Four classes of S. flexneri mutants defective in intercellular spread have been previously recognized . Mutants unable to form protrusions as well as mutants unable to mediate their uptake into adjacent cells via unknown mechanisms were both included into the fourth class ,  In this respect, the ompA mutant might represent a new class of S. flexneri mutants being able to invade and multiply within host cells, able to produce proper F-actin comet tails, presenting a smooth LPS, but exhibiting IcsA exposed across the lateral bacterial surface, impaired in plaque and protrusion formation.
The production of apparently normal F-actin comet tails displayed by the ompA mutant is in contrast with previous studies reporting that S. flexneri mutants displaying IcsA exposed across the bacterial surface presented aberrant bacterial motility , . To explain this discrepancy we argued that the apparently normal F-actin comet tails, displayed by the ompA mutant, might be ineffective to generate those unidirectional movements which have been shown to be necessary for productive protrusion formation. This hypothesis is supported by the findings that not-polarly exposed IcsA drastically affects the force of actin-based motility necessary to deform the plasma membrane thus leading to protrusion formation , and that ca-IcsA* has been shown to be not sufficient to mediate effective ABM . Therefore, we envisage that the apparently normal F-actin comet tails, displayed by the ompA mutant, are not effective to generate productive protrusion formation, thereby blocking the mutant within the cytoplasm.
The observed phenotype of the ompA mutant was reminiscent of that displayed by a mlaA/vacJ S. flexneri mutant (mlaA/vacJ encodes a lipoprotein exposed on the bacterial surface) . It has been reported that the mlaA/vacJ mutation, generated into the S. flexneri 2a strain YSH600T, did not affect the ability of S. flexneri to invade and to multiply within infected MK2 cells and to form protrusions at a rate similar if not identical to that of the wild-type strain, while the mlaA/vacJ mutant was unable to plaque and presented increased membrane permeability. The authors reported that the defect in plaque formation was due to the inability of the mutant to move from protrusions into the cytoplasm of adjacent cells but they did not evaluate the exposition of IcsA on the bacterial surface . This latter point would have been crucial to assess similarity between these two mutants. Anyway, since the authors reported that the mlaA/vacJ mutant displayed wild-type levels of bacterium-containing membranous protrusions, while the ompA mutation dramatically affected protrusion formation, this difference appeared to rule out the possibility that the lack of OmpA may exert a pleiotropic effect on the expression of the mlaA/vacJ gene.
Nevertheless, the results presented in this study have received further support by the analysis of HND99, an independent ompA mutant of the wild-type S. flexneri strain M90T (Table 1; Text S1). HND99 was found to be indistinguishable to the ompA mutant strain HND92. In fact, HND99 was able to invade and to multiply within infected HeLa cells, displayed reduced protrusion formation, failed to plaque and presented IcsA exposed across the entire bacterial surface, with the majority of immunostained bacteria displaying IcsA polar reinforcements. Moreover, as HND92, the introduction of recombinant plasmid pOA into HND99, restored wild-type levels of the considered virulence phenotypes.
Taken together, we concluded that OmpA is required and sufficient for the expression of relevant virulence phenotypes as formation of typical IcsA caps, plaquing ability and protrusion formation and thus for intra-and inter-cellular S. flexneri motility (Fig. 3, 5 and 7). Since the membrane architecture of the spatial arrangement of OmpA depends on the presence and on the interaction with a variety of other molecules , we could not exclude that OmpA might play only an indirect role in the expression of these virulence phenotypes. Experiments are in progress to unravel this point.
In conclusion, in this study we present substantial experimental evidence demonstrating that OmpA plays an active role in relevant S. flexneri virulence traits, like IcsA exposition, cell-to-cell-spread and protrusion formation. These findings complement the recently discovered passive role of OmpA in the induction of the host immune responses , strenghthening its importance also as a promising vaccine candidate.
Construction of HND99, a non-polar ompA deletion mutant of the wild-type S. flexneri strain M90T.
OmpA of S. flexneri is not involved in cPLA2α activation. HeLa cell monolayers were infected with the wild-type (WT) and the ompA mutant (ompA) strains, at a MOI of 20. At 1 h post-infection, cells were lysed and immunoblotted using anti-p-cPLA2α and anti-cPLA2α rabbit polyclonal antibodies. A lysate of uninfected HeLa cells (Ctrl) were included as control.
We thank N.V. Prasadarao, M.L. Bernardini and A. Phalipon for the gift of anti-OmpA, anti-IcsA, anti-IpaB -IpaC antibodies, respectively. The authors are grateful for the valuable help from F. Lombardo for the facility of the Bio-Tek microplate instrument and data analysis.
Conceived and designed the experiments: CA. Performed the experiments: CA MP DS. Analyzed the data: CA MP. Contributed reagents/materials/analysis tools: CA MP DS CZ SR MN. Wrote the paper: CA MN.
- 1. Labrec EH, Schneider H, Magnani TJ, Formal SB (1964) Epithelial cell penetration as an essential step in the pathogenesis of bacillary dysentery. J Bacteriol 88: 1503–1518.
- 2. Sansonetti P, Egile C, Wenneras C (2001) Shigellosis: from disease symptoms to molecular and cellular pathogenesis, p.335–385. In E. A. Groisman (ed.), Principles of Bacterial Pathogenesis. Academic Press, Inc.
- 3. Makino S, Sasakawa C, Kamata K, Kurata T, Yoshikawa M (1986) A genetic determinant required for continuous reinfection of adjacent cells on large plasmid in S. flexneri 2a. Cell 46: 551–555.
- 4. Sansonetti PJ, Ryter A, Clerc P, Maurelli AT, Mounier J (1986) Multiplication of Shigella flexneri within HeLa cells: lysis of the phagocytic vacuole and plasmid-mediated contact hemolysis. Infect Immun 51: 461–469.
- 5. Clerc P, Sansonetti PJ (1987) Entry of Shigella flexneri into HeLa cells: evidence for directed phagocytosis involving actin polymerization and myosin accumulation. Infect Immun 55: 2681–2688.
- 6. Bernardini ML, Mounier J, d'Hauteville H, Coquis-Rondon M, Sansonetti PJ (1989) Identification of icsA, a plasmid locus of Shigella flexneri that governs bacterial intra- and intercellular spread through interaction with F-actin. Proc Natl Acad Sci USA 86: 3867–3871.
- 7. Suzuki T, Miki H, Takenawa T, Sasakawa C (1998) Neural Wiskott-Aldrich syndrome protein is implicated in the actin-based motility of Shigella flexneri. EMBO J 17: 2767–2776.
- 8. Cossart P (2000) Actin-based motility of pathogens: the Arp2/3 complex is a central player. Cell Microbiol 2: 195–205.
- 9. Charles M, Pérez M, Kobil JH, Goldberg MB (2001) Polar targeting of Shigella virulence factor IcsA in Enterobacteriacae and Vibrio. Proc Natl Acad Sci USA. 98: 9871–9876.
- 10. Goldberg MB (2001) Actin-based motility of intracellular microbial pathogens. Microbiol Mol Biol Rev 65: 595–626.
- 11. May KL, Morona R (2008) Mutagenesis of the Shigella flexneri autotransporter IcsA reveals novel functional regions involved in IcsA biogenesis and recruitment of host neural Wiscott-Aldrich syndrome protein. J Bacteriol 190: 4666–4676.
- 12. Heindl JE, Saran I, Yi CR, Lesser CF, Goldberg MB (2010) Requirement for formin-induced actin polymerization during spread of Shigella flexneri. Infect Immun 78: 193–203.
- 13. Goldberg MB, Bârzu O, Parsot C, Sansonetti PJ (1993) Unipolar localization and ATPase activity of IcsA, a Shigella flexneri protein involved in intracellular movement. J Bacteriol 175: 2189–2196.
- 14. Goldberg MB, Theriot JA, Sansonetti PJ (1994) Regulation of surface presentation of IcsA, a Shigella protein essential to intracellular movement and spread, is growth phase dependent. Infect Immun 62: 5664–5668.
- 15. Fukuda I, Suzuki T, Munakata H, Hayashi N, Katayama E, et al. (1995) Cleavage of Shigella surface protein VirG occurs at a specific site, but the secretion is not essential for intracellular spreading. J Bacteriol 177: 1719–1726.
- 16. Steinhauer J, Agha R, Pham T, Varga AW, Goldberg MB (1999) The unipolar Shigella surface protein IcsA is targeted directly to the bacterial old pole: IcsP cleavage of IcsA occurs over the entire bacterial surface. Mol Microbiol 32: 367–377.
- 17. Kadurugamuwa JL, Rohde M, Wehland J, Timmis KN (1991) Intercellular spread of Shigella flexneri through a monolayer mediated by membranous protrusions and associated with reorganization of the cytoskeletal protein vinculin. Infect Immun 59: 3463–3471.
- 18. Haglund CM, Welch MD (2011) Pathogens and polymers: microbe-host interactions illuminate the cytoskeleton. J Cell Biol 195: 7–17.
- 19. Krishnan S, Prasadarao NV (2012) Outer membrane protein A and OprF: versatile roles in Gram-negative bacterial infections. FEBS J. doi: 10.1111/j.1742-4658.2012.08482.x.
- 20. Smith SG, Mahon V, Lambert MA, Fagan RP (2007) A molecular Swiss army knife: OmpA structure, function and expression. FEMS Microbiol Lett 273: 1–11.
- 21. Reusch RN (2012) Insights into the structure and assembly of Escherichia coli outer membrane protein A. FEBS J. doi: 10.1111/j.1742-4658.2012.08484.x.
- 22. Pore D, Chowdhury P, Mahata N, Pal A, Yamasaki S, et al. (2009) Purification and characterization of an immunogenic outer membrane protein of Shigella flexneri 2a. Vaccine 27: 5855–5864.
- 23. Pore D, Mahata N, Chakrabarti MK (2012) Outer membrane protein A (OmpA) of Shigella flexneri 2a links innate and adaptive immunity in a TLR2-dependent manner and involvement of IL-12 and nitric oxide. J Biol Chem 287: 12589–12601.
- 24. Sambrook J, Russell D (2001) Molecular cloning: a laboratory manual, 3nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
- 25. Tran CN, Giangrossi M, Prosseda G, Brandi A, Di Martino ML, et al. (2011) A multifactor regulatory circuit involving H-NS, VirF and an antisense RNA modulates transcription of the virulence gene icsA of Shigella flexneri. Nucleic Acids Res 39: 8122–8134.
- 26. Santapaola D, Del Chierico F, Petrucca A, Uzzau S, Casalino M, et al. (2006) Apyrase, the product of the virulence plasmid-encoded phoN2 (apy) gene of Shigella flexneri, is necessary for proper unipolar IcsA localization and for efficient intercellular spread. J Bacteriol 188: 1620–1627.
- 27. Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680–685.
- 28. Hitchcock PJ, Brown TM (1983) Morphological eterogeneity among Salmonella lipopolysaccharide chemotypes in silver-stained polyacrylamide gels. J Bacteriol 154: 269–277.
- 29. Tsai CM, Frasch CE (1982) A sensitive silver stain for detecting lipopolysaccharides in polyacrylamide gels. Anal Biochem 119: 115–119.
- 30. Hale TL, Formal SB (1981) Protein synthesis in HeLa or Henle 407 cells infected with Shigella dysenteriae 1, Shigella flexneri 2a, or Salmonella typhimurium W118. Infect Immun 32: 137–144.
- 31. Oaks EV, Wingfield ME, Formal SB (1985) Plaque formation by virulent Shigella flexneri. Infect Immun 48: 124–129.
- 32. Purdy GE, Fisher CR, Payne SM (2007) IcsA surface presentation in Shigella flexneri requires the periplasmic chaperones DegP, Skp, and SurA. J Bacteriol 189: 5566–5573.
- 33. Wagner JK, Heindl JE, Gray AN, Jain S, Goldberg MB (2009) Contribution of the periplasmic chaperone Skp to efficient presentation of the autotransporter IcsA on the surface of Shigella flexneri. J Bacteriol 191: 815–821.
- 34. Rasband WS (1997–2007) ImageJ. US National Institutes of Health, Bethesda, Maryland. http://rsb.info.nih.gov/ij/.
- 35. Beher MG, Schnaitman CA, Pugsley AP (1980) Major heat-modifiable outer membrane protein in gram-negative bacteria: comparison with the OmpA protein of Escherichia coli. J Bacteriol 143: 906–913.
- 36. Hofstra H, Dankert J (1980) Major outer membrane proteins: common antigen in Enterobacteriaceae species. J Gen Microbiol 119: 123–131.
- 37. Hofstra H, Van Tol MJD, Dankert J (1980) Crossreactivity of major outer membrane proteins in Enterobacteriaceae studied by crossed immunoelectrophoresis. J Bacteriol 143: 328–337.
- 38. Cock H, van Blokland S, Tommassen J (1996) In vitro insertion and assembly of outer membrane protein PhoE of Escherichia coli K-12 into the outer membrane. Role of Triton X-100. J Biol Chem 271: 12885–12890.
- 39. Sugawara E, Steiert M, Rouhani S, Nikaido H (1996) Secondary structure of the outer membrane proteins OmpA of Escherichia coli and OprF of Pseudomonas aeruginosa.. J Bacteriol 178: 6067–6069.
- 40. Baharani FK, Sansonetti PJ, Parsot C (1997) Secretion of Ipa proteins by Shigella flexneri: inducer molecules and kinetics of activations. Infect Immun 65: 4005–4010.
- 41. Suzuki T, Murai T, Fukuda I, Tobe T, Yoshikawa M, et al. (1994) Identification and characterization of a chromosomal virulence gene, vacJ, required for intercellular spreading of Shigella flexneri. Mol Microbiol 11: 31–41.
- 42. Morona R, Van Den Bosch L (2003) Lipopolysaccharide O antigen chains mask IcsA (VirG) in Shigella flexneri. FEMS Microbiol Lett 221: 173–180.
- 43. Van Den Bosch L, Morona R (2003) The actin-based motility defect of a Shigella flexneri rmlD rough LPS mutant is not due to loss of IcsA polarity. Microb Pathog 35: 11–18.
- 44. Hong M, Payne SM (1997) Effect of mutations in Shigella flexneri chromosomal and plasmid-encoded lipopolysaccharide genes on invasion and serum resistance. Mol Microbiol 24: 779–791.
- 45. Cossart P (2005) Bacterial invasion: a new strategy to dominate cytoskeleton plasticity. Dev Cell 6: 314–315.
- 46. Egile C, d'Hauteville H, Parsot C, Sansonetti PJ (1997) SopA, the outer membrane protease responsible for polar localization of IcsA in Shigella flexneri. Mol Microbiol 23: 1063–1073.
- 47. Prasadarao NV, Wass CA, Weiser JN, Stins MF, Huang SH, et al. (1996) Outer membrane protein A of Escherichia coli contributes to invasion of brain microvascular endothelial cells. Infect Immun 64: 146–153.
- 48. Wang Y (2002) The function of OmpA in Escherichia coli. Biochem Biophys Res Commun 292: 396–401.
- 49. Haugland RP (1996) Handbook of fluorescent probes and research chemicals. 6th ed, Eugene, Oreg, Molecular Probes Inc.
- 50. Maruvada R, Kim KS (2011) Extracellular loops of the Escherichia coli outer membrane protein A contribute to the pathogenesis of meningitis. J Infect Dis 203: 131–140.
- 51. Kandzari DE, Chen J, Goldschmidt-Clermont PJ (1996) Regulation of the actin cytoskeleton by inositol phospholipid pathways. Subcell Biochem 26: 97–114.
- 52. Datsenko KA, Wanner BL (2000) One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A 97: 6640–6645.
- 53. Chai TJ, Foulds J (1977) Purification of protein A, an outer membrane component missing in Escherichia coli K-12 ompA mutants. Biochim Biophys Acta 493: 210–215.
- 54. Robbins JR, Monack D, McCallum SJ, Vegas A, Pham E, et al. (2001) The making of a gradient: IcsA (VirG) polarity in Shigella flexneri. Mol Microbiol 41: 861–872.
- 55. Brandon LD, Goehring N, Janakiraman A, Yan AW, Wu T, et al. (2003) IcsA, a polarly localized autotransporter with an atypical signal peptide, uses the Sec apparatus for secretion, although the Sec apparatus is circumferentially distributed. Mol Microbiol 50: 45–60.
- 56. Tanji K, Ohta Y, Kawato S, Mizushima T, Natori S, et al. (1992) Decrease by psychotropic drugs and local anaesthetics of membrane fluidity measured by fluorescence anisotropy in E scherichia coli. J Pharm Pharmacol 44: 1036–1037.
- 57. Morona R, Daniels C, Van Den Bosh L (2003) Genetic modulation of Shigella flexneri 2a lipopolysaccharide O antigen modal chain length reveals that it has been optimized for virulence. Microbiology 149: 925–939.
- 58. Wu T, Malinverni J, Ruiz N, Kim S, Silhavy TJ, et al. (2005) Identification of a multicomponent complex required for outer membrane biogenesis in Escherichia coli. Cell 121: 235–245.
- 59. Monack DM, Theriot JA (2001) Actin-based motility is sufficient for bacterial membrane protrusion formation and host cell uptake. Cell Microbiol 3: 633–647.
- 60. Rathman M, Jouirhi N, Allaoui A, Sansonetti P, Parsot C, et al. (2000) The development of a FACS-based strategy for the isolation of Shigella flexneri mutants that are deficient in intercellular spread. Mol Microbiol 35: 974–990.
- 61. Sansonetti PJ, Kopecko DJ, Formal SB (1982) Involvement of a plasmid in the invasive ability of Shigella flexneri. Infect Immun 35: 852–860.
- 62. d’Hauteville H, Sansonetti PJ (1992) Phosphorylation of IcsA by cAMP-dependent protein kinase and its effect on intracellular spread of Shigella flexneri. Mol Microbiol 6: 833–841.