salmonella is a well-armed pathogen that produces a diverse array of pathogenic factors and causes infection. It uses a type three secretion system (TTSS), a needle system that injects bacterial pathogenic proteins into host cells (36). The virulence proteins injected by the TTSS are called effectors. Avirulence factor for Salmonella (AvrA) is a newly described Salmonella effector translocated into host cells (10). AvrA gene is present in 80% of Salmonella enterica serovar Typhimurium (39).

Although the exact function and mechanism of AvrA is not entirely clear, it is known that AvrA is a multifunctional deubiquitinase that influences eukaryotic cell pathways that utilize ubiquitin (48). AvrA also possesses acetyltransferase activity toward specific mitogen-activated protein kinase kinases and potently inhibits c-Jun NH₂-terminal kinase and NF-κB signaling pathways (5, 18). Yersinia outer protein J (YopJ), an AvrA family member, is also assigned multiple functions. Some studies demonstrated that YopJ is a deubiquitinase that negatively regulates signaling by removing ubiquitin moieties from critical proteins, such as TRAF2, TRAF6, and IκBα (42, 49). Other studies have shown that YopJ is an acetyltransferase (31, 32). Therefore, bacterial effectors may have multiple protease activities to modify different eukaryotic proteins and maximize their ability to modulate host cellular functions, just like the eukaryotic protein A20, which has two enzyme activities (4, 14, 45).

Recent studies show that AvrA belongs to a family of proteases which regulates diverse bacterial-host interactions (7, 18, 22, 48). Interestingly, after being injected into the host cells, AvrA is phosphorylated by a TTSS-effector-activated ERK pathway in mammalian cells (7). AvrA may reverse the activation of signaling pathways induced by other Salmonella effectors via the same TTSS (7). Previous studies suggest that other family members related to AvrA include the Yersinia virulence factor YopJ and the Xanthomonas campestris pv. vesicatoria protein AvrBsT (35). Recent analysis with MEROPS database showed that AvrA belongs to YopJ-like proteins and genes (family C55) in bacterial species (see details in http://merops.sanger.ac.uk). These bacterial effectors mimic the activity of a eukaryotic protein, such as acetyltransferase, and debilitate their target cells. Therefore, it is important to elucidate the mechanisms through which bacterial effectors exert their effects.

The p53 protein is a transcription factor known as a “guardian of the genome” because of its crucial role in coordinating cellular responses to stress (13, 19–21, 28). The tumor suppression effects of p53 are mediated by a variety of mechanisms, including regulation of cell cycle arrest, apoptosis, and cellular senescence (34, 43). Expression of p53 is tightly controlled, so that its protein product usually exists in a latent form, and at low levels in unstressed cells. However, the steady-state levels and transcriptional activity of p53 increase dramatically in cells that undergo various types of stress (44). Although the precise mechanisms of p53 activation are not fully understood, they involve posttranslational modification, including ubiquitination, acetylation, phosphorylation, sumoylation, neddylation, methylation, and glycosylation of the p53 polypeptide (19, 44).

Microbial infection constitutes a stress to the host. Viruses (6, 15, 16, 21, 26, 33) and mycoplasma, an intracellular pathogen (25), are known to interact with the p53 pathway. Mycoplasma infection plays the role of a p53-suppressing oncogene that cooperates with Ras in cell transformation (25). It is unknown whether and how infection due to Salmonella, another intracellular pathogen, is involved in the posttranslational modification of p53. We reasoned that the p53 pathway should respond to Salmonella infection, which is a nongenotoxic stress. Although the exact function and mechanism of AvrA is not entirely clear, it is known that AvrA influences eukaryotic cell pathways that utilize ubiquitin (48) and acetylation (18). Acetylation activates p53 and permit p53 to arrest cell cycle (13, 47). Therefore, we hypothesized that Salmonella infection activates the p53 pathway and Salmonella AvrA possesses acetyltransferase activity to increase the acetylation of p53, thereby inducing cell cycle arrest.

Using in vitro and in vivo models, we investigated whether the Salmonella effector protein AvrA plays a role in the regulation of p53 when the host is infected with Salmonella. We demonstrate that this is indeed the case; AvrA is an acetyltransferase that increases the acetylation of p53 in a cell-free system. Expression of AvrA in the intestinal epithelial cells activates p53 transcriptional activity. In addition, intestinal epithelial p53 acetylation is increased by bacterial AvrA expression in a mouse model of Salmonella infection. Functionally, AvrA expression induces cell cycle arrest with increased cell numbers at the G₀/G₁ phase and decreases cell numbers at the G₂/M phase in host cells. Without p53 as a bacterial target, intestinal epithelial HCT116 p53−/− cells were less susceptible to inflammatory responses. Our previous study demonstrated the inhibitory function of AvrA on inflammation in the host (22, 40, 48). In this study we further investigate whether AvrA's inhibitory effect on the inflammatory response is beneficial or harmful for the host. Using a mouse model of Salmonella infection, we showed that AvrA overexpression in wild-type (WT) Salmonella decreases survival. Overall, this study provides a mechanism for how a Salmonella protein interferes with host responses via p53 in intestinal epithelial cells.

MATERIALS AND METHODS

Bacterial strains and growth condition.

Bacterial strains used in this study included Salmonella typhimurium WT strain ATCC14028, Escherichia coli F18, and nonpathogenic Salmonella mutant strain PhoP^c (29), PhoP^c AvrA−, and PhoP^cAvrA−/AvrA+, PhoP^c AvrA+, WT Salmonella SL1344 (SB300), and its AvrA mutant strain SB1117. WT 14028AvrA+ and PhoP^c AvrA+ were generated by overexpressing AvrA gene in a pWSK29 low-copy plasmid (Table 1). Nonagitated microaerophilic bacterial cultures were prepared as previously described (41).

AvrA point mutation.

The AvrA gene is from WT S. typhimurium strain SL3201. AvrA point mutations were generated by using QuikChange II Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA). DNA sequencing was performed by the Functional Genomics Center of University of Rochester.

AvrA and its mutant protein purification.

Salmonella full-length gene AvrA and single point mutants AvrA(C186A), AvrA(R180G), AvrA(C179A), AvrA(E142A), or AvrA(H123A) were cloned into NH₂-terminal glutathione S-transferase-fused Vector pEGX-4T2 (Invitrogen) and transformed into E. coli strain BL21 (DE3). Affinity purification was performed by using a glutathione-Sepharose resin (Amersham Bioscience, Piscataway, NJ) as previously described (48).

Cell culture.

Human embryonic kidney 293 cells, Caco2-BBE, HT29Cl.29A, epithelial HeLa cells, and MEFs were maintained in DMEM supplemented with 10% FCS, penicillin-streptomycin, and l-glutamine. Human colonic epithelial HCT116 cells and the p53−/− knockout cells were cultured in McCoy's 5A medium supplemented with 10% (vol/vol) fetal bovine serum. The rat small intestinal IEC-18 cell line was grown in DMEM (high glucose, 4.5 g/l) containing 5% (vol/vol) fetal bovine serum, 0.1 U/ml insulin, 50 μg/ml streptomycin, and 50 U/ml penicillin.

Streptomycin-pretreated mouse model.

Animal experiments were performed with specific-pathogen-free female C57BL/6 mice (Taconic) that were 6–7 wk old as previously described (8). Water and food were withdrawn 4 h before oral gavage with 7.5 mg/mouse of streptomycin (100 μl of sterile solution or 100 μl of sterile water in control). Afterward, animals were supplied with water and food ad libitum. Twenty hours after streptomycin treatment, water and food were withdrawn again for 4 h before the mice were infected with 1 × 10⁷ colony-forming units (CFU) of S. typhimurium (100 μl suspension in HBSS) or treated with sterile HBSS (control) by oral gavage as previously described (48). At indicated time after infection, mice were euthanized and tissue samples from the intestinal tracts were removed for analysis. For the survival rate of mice post-S. typhimurium infection, mice (n = 20 per group) were infected with 1 × 10⁸ CFU WT S. typhimurium strain 14028s with insufficient AvrA expression (WT) or 14028s with AvrA overexpression (WTAvrA+) and observed for 7 days. If a mouse showed indication that it had aspirated fluid or significant body weight loss (10% or more), and did not die immediately, the mouse was humanely euthanized. The protocol was approved by the University of Rochester University Committee on Animal Resources.

Mouse colonic epithelial cells.

Mouse colonic epithelial cells were collected by scraping the mouse colon including proximal and distal regions. Cells were sonicated in lysis buffer (1% Triton X-100, 150 mM NaCl, 10 mM Tris pH 7.4, 1 mM EDTA, 1 mM EGTA, pH 8.0, 0.2 mM sodium orthovanadate, protease inhibitor cocktail), and the protein concentration was determined (Bio-Rad, Hercules, CA). β-Actin was used as the loading control for all Western blots. Villin, an accepted marker of epithelial cells (9), was used as the control for epithelial cell protein content in all Western blots.

Transient transfections.

Transient transfections were performed with LipofectAMINE2000 (Invitrogen, San Diego, CA) in accordance with the manufacturer's instructions. Briefly, 6×10⁵ epithelial cells were seeded on 60-mm dishes overnight before cotransfection with 4 μg c-Myc-tagged AvrA or C186A with 2 μg green fluorescent protein (GFP)-tagged p53, or 2 μg HA-tagged p53. DNA was mixed with the liposome reagent at a ratio of 1:1 before adding to cells. After a 24-h transfection, proteins were extracted with RIPA buffer (50 μM Tris·HCl, pH 8.0 with 150 mM sodium chloride, 1.0% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS) for immunoblotting.

Immunoblotting.

Mouse colonic epithelial cells were collected by scraping the mouse colon including proximal and distal regions (8). Cells were sonicated in lysis buffer (1% Triton X-100, 150 mM NaCl, 10 mM Tris pH 7.4, 1 mM EDTA, 1 mM EGTA pH 8.0, 0.2 mM sodium orthovanadate, protease inhibitor cocktail). The protein concentration was measured with Bio-Rad reagent (Bio-Rad). Cultured cells were rinsed twice in ice-cold HBSS, lysed in protein loading buffer (50 mM Tris, pH 6.8, 100 mM dithiothreitol, 2% SDS, 0.1% bromophenol blue, 10% glycerol), and sonicated. Equal amounts of protein were separated by SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose, and immunoblotted with primary antibodies. The following antibodies were used: monoclonal mouse anti-c-Myc or HA (Santa Cruz Biotechnology, Santa Cruz, CA), monoclonal mouse anti-GFP, monoclonal mouse anti-p53, monoclonal mouse anti-Bax, anti-Bcl-2, anti-p21 (Santa Cruz), anti-acetylated p53(373), p53(373/382), or phosphorylated p53 at ser9 (Cell Signaling Technology, Beverly, MA), anti-acetyl-lysine, anti-acetylated p53(320) (Upstate, Temecula, CA), monoclonal mouse anti-p14ARF, monoclonal mouse anti-MDM2 (Abcam, Cambridge, MA), or anti-β-actin (Sigma-Aldrich, Milwaukee, WI).

Coimmunoprecipitation assay.

Cells were rinsed twice in ice-cold HBSS and lysed in cold immunoprecipitation buffer (1% Triton X-100, 150 mM NaCl, 10 mM Tris·HCl, pH 7.4, 1 mM EDTA, 1 mM EGTA, pH 8.0, 0.2 mM sodium orthovanadate) containing protease inhibitor cocktail. Samples were precleared with protein A-agarose. Precleared lysates were then incubated with indicated antibody. Coimmunoprecipitation samples were separated by SDS-polyacrylamide gel electrophoresis and transferred onto a nitrocellulose membrane (40, 41, 48). Membrane blots were probed with anti-c-myc or anti-p53 antibody and visualized by enhanced chemiluminescence.

Immunofluorescence.

Colonic tissues from the proximal and distal portion of the colon were freshly isolated and embedded in paraffin wax after fixation with 10% neutral buffered formalin. Immunohistochemistry was performed on paraffin-embedded sections (1 μm) of mouse colons. Tissue samples or cultured cells were processed for immunofluorescence as described previously (22). The slides were stained with anti-acetylated p53 antibody. Cells or tissues were mounted with SlowFade (SlowFade AntiFade Kit, Molecular Probes) followed by a coverslip, and the edges were sealed to prevent drying. Specimens were examined with a Leica SP5 Laser Scanning confocal microscope.

In vitro AvrA transacetylase assays.

For the cell-free AvrA transacetylase assay, we used purified WT p53 protein (Santa Cruz, sc4246) as the substrate. The WT AvrA, AvrA mutant proteins were purified from E. coli strain BL21(DE3). The 50-μl reactions contained 10 μl 5× reaction buffer (250 mM Tris, pH 8.0, 50% glycerol, 0.5 mM EDTA, 5 mM DTT), 5 μl 0.1 mM acetyl-CoA (Sigma), p53, and AvrA protein. Transacetylase p300-CBP-associated factor (PCAF; Upstate) was used as a positive control. The reaction mixture was incubated at 30°C for 30 min. The reactions were stopped by the addition of an equal volume of SDS-gel sample buffer. It was then denatured and immunoblotted for detection.

p53 transcriptional activation.

HCT116 p53−/− cells were grown in 12-well plates in triplicate. The cells were transfected with different plasmids. The plasmid groups include a pFC-p53 positive control plasmid (Fc), a p53-Luc-cis reporter plasmid (p53) (Stratagene, La Jolla, CA), an internal control pRL-TK plasmid (TK), and a pCMV-Myc-AvrA, using LipofectAMINE2000. The control plasmid pRL-TK contains a Renilla reporter gene driven by the thymidine kinase promoter (Promega, Madison, WI). After transfection for 24 h, cells were lysed, and luciferase activity was determined by using the Dual Luciferase Reporter Assay System (Promega) with a TD-20/20 luminometer (Turner Designs, Sunnyvale, CA). Firefly luciferase activity was normalized to Renilla luminescence activity, and then activity was expressed as relative units.

Cell cycle.

Cell cycle was analyzed by DNA Content (propidium iodide staining). IEC-18 cells were colonized by the indicated bacterial strains at 37°C for 30 min, washed, incubated for 4h in DMEM with gentamicin. Forskolin (50 μM) treatment was used as a positive control. Cells were collected, washed twice with Hanks', and fixed for 1 h by using 1 ml methanol prechilled at −20°C. Cell samples were resuspended in 400 μl PI/RNase staining buffer (BD, Franklin Lakes, NJ) at room temperature for 10 min, then subjected to flow cytometry in a Beckman Coulter Epics XL MCL, and 50,000 cells were collected for cell cycle analysis.

Salmonella-induced human IL-8 secretion.

HCT116 p53−/− or p53+/+ cells were cultured in DMEM, followed by Salmonella-containing HBSS (1.6 × 10¹⁰ bacteria/ml) for 30 min, washed three times in HBSS, and incubated at 37°C for 6 h. Cell supernatants were removed and assayed for IL-8 by ELISA in 96-well plates as described previously (27).

Real-time quantitative PCR analysis of IL-8.

Total RNA was extracted from epithelial cell monolayers by use of TRIzol reagent (Invitrogen, Carlsbad, CA). The RNA integrity was verified by electrophoresis gel. RNA reverse transcription was done by using the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA) according to the manufacturer's directions. The RT cDNA reaction products were subjected to quantitative real-time PCR using the MyiQ single-color real-time PCR detection system (Bio-Rad) and iQ SYBR Green Supermix (Bio-Rad) according to the manufacturer's directions. IL-8 cDNA was amplified by using primers to the human IL-8 gene that are complementary to regions in exon 1 (5′-TGCATAAAGACATACTCCAAACCT) and overlapping the splice site between exons 3 and 4 (5′-AATTCTCAGCCCTCTTCAAAAA). All expression levels were normalized to the GAPDH levels of the same sample, by using forward (5′-CTTCACCACCATGGAGAAGGC) and reverse (5′-GGCATGGACTGTGGTCATGAG) primers for GAPDH. Percent expression was calculated as the ratio of the normalized value of each sample to that of the corresponding untreated control cells. All real-time PCR reactions were performed in triplicate. All PCR primers were designed by use of Lasergene software (DNAStar, Madison, WI).

Statistical analysis.

Data are expressed as means ± SD. Differences between two samples were analyzed by Student's t-test. Differences between groups were analyzed by ANOVA (SAS 9.2 version). P values of 0.05 or less were considered significant.

RESULTS

Salmonella but not TNF increase acetylation of p53.

To determine whether Salmonella plays a role in modulating the p53 pathway, we colonized epithelial HCT116 cells [possesses WTp53 (24)] with WT Salmonella or TNF, a proinflammatory cytokine. Salmonella increased the p53 acetylation in host cells after bacterial colonization for only 1 h (Fig. 1A HCT116 p53+/+). In contrast, TNF had less effect on p53 acetylation. Densitometry analysis of p53 acetylation bands showed that the increase in p53 acetylation by Salmonella was significant. HCT116 p53−/− p53 knockout cells were used as a negative control. There was no total or acetylated p53 expression in HCT116 p53−/− cells (Fig. 1A, HCT116 p53−/−). We also investigated the response of the human colonic epithelial cells HT29 Cl.19A or Caco-2 BBE. Similar changes of Salmonella-induced p53 acetylation were also found (supplemental data for this article are available online at the American Journal of Physiology Gastrointestinal and Liver Physiology website in Supplemental Fig. S1). To further confirm our findings, we examined normal epithelial IEC-18 cells, a nontransformed cell line, and MEFs. As shown in Fig. 1, Salmonella colonization also increased the acetylated form of p53 at position 373 in these cell lines, whereas TNF had no effect on the acetylation of p53 at position 373. The total acetylated lysine was also determined (Fig. 1). Whereas Salmonella increased acetylation of p53 at position 373, total acetylated lysine was only slightly increased. Densitometry data indicated the significant increase in p53 acetylation at 373 by Salmonella [Fig. 1B, IEC-18 and mouse embryonic fibroblast (MEF) cells].

p53 has nine acetylation sites related to p53 activation and stability (19). We chose acetylation of p53 at position 373 because of the availability of the antibody. We also tested p53 acetylation at amino acid position 320 and did not find any change induced by Salmonella colonization. p53 NH₂-terminal phosphorylation, such as ser 9, is activated by DNA damage, UV light, ionizing radiation, replicative senescence, or phosphatidylcholines (19). However, Salmonella colonization did not change the phosphorylated p53 at ser9 either (data not shown).

Acetylated p53 is located in the nuclei. Using immunofluorescence staining, we also investigated the location of acetylated p53 in the intestinal epithelial cells (Fig. 1C). There was little acetylated p53 staining in the control cells without treatment. Salmonella colonization induced the nuclear acetylated p53. Overall, these data indicate that the change in p53 acetylation was induced by Salmonella.

AvrA expression increases acetylated p53.

Salmonella effector AvrA is a multiple functional protein that hijacks the host signaling pathway using ubiquitination and acetylation (7, 18, 48). We then asked whether AvrA was involved in the acetylation of p53. We focused on bacterial strains sufficient or deficient in AvrA. The Western blot assay in Fig. 2A shows the AvrA protein levels in strains used in our studies. The WT strain ATCC14028 has very low AvrA protein expression. PhoP^c is a mutant strain derived from WT ATCC14028 with sufficient AvrA protein expression (48). We have used parental PhoP^c AvrA mutant (AvrA−), or the AvrA-complemented strain (PhoP^c AvrA−/AvrA+) in previous studies (48). After bacterial colonization, AvrA expression in the host cells was determined by PCR (Fig. 2B). Please note that PhoP^c AvrA−/AvrA+ is with AvrA complemented in an AvrA−pWSK29 plasmid; therefore, we did not detect the AvrA mRNA by extracting the total bacterial RNA. The AvrA protein is detectable in PhoP^c AvrA−/AvrA+, as shown in Fig. 2A. In IEC-18 cells, parental PhoP^c strain colonization increased acetylated p53 whereas cells treated with PhoP^c AvrA− bacteria and WT strain ATCC14028, which has decreased levels of AvrA protein expression (3, 22), displayed reduced levels of acetylated p53. In cells colonized with PhoP^c AvrA−/AvrA+, the complementary AvrA expression increased the acetylated forms of p53 at amino acid positions 382 and 373 (Fig. 2C, IEC-18).

We further tested the Salmonella-induced p53 acetylation in different cell lines of mouse and human origin. Human epithelial HeLa cells were treated with AvrA-sufficient or -deficient Salmonella strains for 6 h, and the AvrA+ strain increased p53 acetylation at 373 and 382 sites (Fig. 2D, HeLa). Using transient transfection of a WT pCMV-Myc-AvrA plasmid or a mutant AvrAC186A plasmid into human colonic epithelial cells, we further determined the role of AvrA in enhancing p53 acetylation in host cells. The HCT116 p53−/− cells were transfected with the AvrA, p53 or AvrA+p53 plasmids (Fig. 2E) to further test the direct effect of AvrA on the p53 acetylation. There was no acetylated p53 expression in the control cells, cells transfected with L2K, or empty plasmid. The AvrA or p53 transfection added a slight background in the Western blot for p53 acetylation. In contrast, AvrA and p53 cotransfection dramatically enhanced the p53 acetylation in the HCT116 p53−/− cells. In addition, intestinal epithelial Caco2BBE cells were transfected with either the pCMV-Myc-AvrA or AvrA mutant C186A plasmid for 24 h. C186A is the key amino acid for the deubiquitinase activity of AvrA (5). Total cell lysates were analyzed for protein levels by immunoblot. We found that overexpression of AvrA increased p53 acetylation in epithelial cells (Supplemental Fig. S2A). The AvrA mutant C186A had less effect on p53 acetylation compared with the WT AvrA. In the MEF cells, WT strain colonization increased acetylated p53 at position 373, whereas PhoP^c bacteria (with sufficient AvrA expression) increased even more acetylated p53 compared with the WT strain (with less AvrA protein) (Supplemental Fig. S2B). Taken together, our data demonstrate that Salmonella protein AvrA induces acetylation of p53 in host cells.

AvrA displays a physical interaction with p53.

To establish whether the increased p53 acetylation was occurring through physical interaction between AvrA and p53, we first determined whether AvrA can form a complex with p53 in epithelial cells. A c-myc-AvrA plasmid DNA was cotransfected with the hemagglutinin epitope YPYDVPDYA (HA tag) p53 for 24 h, then immunoprecipitated with HA. Immunoblot with HA clearly showed that AvrA interacts with p53, whereas less p53 was observed to be bound to the AvrA mutant C186A which lacks deubiquitin activity (Fig. 3A). Densitometry data further showed that AvrA mutant C186A significantly reduced the p53/AvrA binding. In addition, GFP-p53 was cotransfected with c-Myc-AvrA for 24 h. Immunoblot with c-Myc also showed AvrA interaction with p53 (Supplemental Fig. S3). We further examined the role of AvrA in targeting the acetylated p53 at positions 373 and 382 during Salmonella infection. The c-Myc-tagged AvrA was cotransfected with HA-p53, and then HA beads were used to perform pull-down assays. The acetylation of p53 at sites 382 and 373 was detected by Western blot (Fig. 3B). However, there was no association of AvrA with acetylated p53 at position 320. Moreover, densitometry data indicated that AvrA mutant C186A significantly decreased the p53 acetylation.

AvrA protein directly increases p53 acetylation.

Studies on effector protein showed that bacterial effector could function independent of TTSS as a protease (17). Hence, we purified the AvrA WT protein and tested its effects on p53 acetylation. We directly treated the HCT116 cells with the AvrA protein in the culture media. We found that AvrA protein treatment was able to increase p53 acetylation in the intestinal epithelial cells, whereas TNF-α or BSA proteins did not change p53 acetylation (Fig. 4A). In a cell-free system for the AvrA transacetylase assay, purified WT p53 protein was used as the substrate. Transacetylase PCAF was used as a positive control. As shown in Fig. 4B, AvrA was able to acetylate p53 in the cell-free system. We used AvrA at different concentrations and mixed it with p53 in the reaction buffer (Fig. 4B). The level of acetylated p53 enhanced with increasing AvrA concentrations in the cell-free system. These data indicated that the AvrA's effect on p53 acetylation is dose dependent.

AvrA increases p53 transcriptional activity and modifies target genes of p53.

p53 protein has a relatively short half-life, ∼20 min. In unstressed cells, p53 usually exists in a latent form, and at low levels. p53 is activated rapidly in response to stress including microbial infection. p53 activation involves posttranslational modification, including ubiquitination and acetylation (19, 44). Under various types of stress, the transcriptional activity of p53 increases dramatically.

We then investigated the effects of AvrA on the regulation of p53 transcriptional activity. HCT116p53−/− cells were transfected with a pFC-p53 positive control plasmid (Fc), a p53-Luc-cis reporter plasmid (p53), an internal control pRL-TK plasmid (TK) or a pCMV-Myc-AvrA. As shown in Fig. 5D, single pFC-p53, TK, or p53+TK cotransfection had no effect on the transcriptional activity, whereas AvrA and p53 cotransfection significantly increased p53 transcription activity to a level comparable to the positive control group with the p53 and pFC-p53 cotransfection (Fig. 5A). These data indicated that the expression of bacterial AvrA was able to increase p53 transcriptional activity in intestinal epithelial cells.

We further investigated the effect of AvrA on the p53's target gene expression in host cells. Human epithelial cells were colonized with the bacterial strains PhoP^C, PhoP^C with AvrA overexpression (PhoP^CAvrA+), and AvrA mutation. The protein levels of downstream target genes of p53, including p21, BAX, p14ARF, and bcl-2, were investigated by Western blot (Fig. 5B). Etoposide and staurosporine (STS) were used as controls. Our data showed etoposide, acting primarily in the G₂ and S phases of the cell cycle, increased p53 expression, whereas STS, a nonselective protein kinase inhibitor that has been shown to induce apoptosis, decreased p53 expression. With AvrA overexpression in the PhoP^cAvrA+ strain, the protein levels of p21 and BAX were decreased (Fig. 5B). AvrA expression had no effect on p14ARF and BCL-2 expression. MDM2, a p53-specific E3 ubiquitin ligase, is the principal cellular antagonist of p53, acting to limit the p53 growth-suppressive function in unstressed cells (30). Interestingly, cells colonized with bacterial strains PhoP^C, PhoP^C with AvrA overexpression, and AvrA mutation all had elevated MDM2 expression compared with the cells without treatment or cells treated with etoposide or STS. PhoP^CAvrA with AvrA overexpression showed a slightly less MDM2 expression compared with the parental PhoP^C group. We also examined p21, which mediates the p53-dependent cell cycle G₁ phase arrest in response to a variety of stress stimuli (11). Cells colonized with bacterial PhoP^c were able to increase p21 expression, as did the etoposide treatment, whereas PhoP^cAvrA+ decreased p21 expression.

Salmonella infection increases the acetylation of p53 in a mouse model.

A streptomycin-pretreated mouse model of Salmonella infection (2) was used to confirm our in vitro findings. One concern for the in vivo study is confirmation of the bacterial colonization and invasion ability in the intestinal epithelial cells. To address the location by which S. typhimurium infects the intestine, we stained infected colon tissues for Salmonella lipopolysaccharide by immunofluorescence. As shown in Fig. 6B, S. typhimurium bacteria were found in the mucosa. This localization indicates that Salmonella invaded intestinal epithelium. Mice colonic epithelial cells were collected post-Salmonella infection. We found that WT Salmonella increased the acetylation of p53 postinfection for 2 h and was persistent for over 6 h (Fig. 6A). In addition, immunofluorescence staining showed increased acetylated p53 (red staining) in the epithelial cell nuclei after Salmonella colonization (mouse colons Fig. 6C). Since we do not have the AvrA mutant generated from the SL14028 genetic background, we used WT Salmonella SB300 and its AvrA mutant strain SB1117 (AvrA−) to test the effects of AvrA in regulating p53 acetylation in vivo. Salmonella SB300 with AvrA expression significantly increased acetylated p53, and SB1117 (AvrA−) did not change the level of p53 acetylation (Fig. 6D). Because the acetylation of p53 at the COOH terminus is related to p53 stability (19), we also examined the level of total p53, related regulator MDM2, and target gene p21. Total p53 decreased in Salmonella-infected mice over 4 days. The MDM2 and p21 expression were increased by SB300 with AvrA, whereas the AvrA deficient SB1117 did not change the expression of MDM2 and p21 (Fig. 6D).

AvrA expression induces cell cycle arrest.

p53 pathway activation directly increases p21 for cell cycle regulation and leads to the cell cycle arrest in both G₀ and G₁ (12). We further investigated the cell physiological function of AvrA in regulating the cell cycle in intestinal epithelial cells. Intestinal epithelial IEC-18 cells were infected with bacteria. The experimental groups include C (normal IEC-18 cells without any treatment), F (positive control forskolin), T (TNF-α), S (WT Salmonella ATCC14028), P (PhoP^C with AvrA expression), A− (AvrA−), A+ (AvrA− restored with AvrA in plasmid), SB300 (WT Salmonella SL1344), and SB1117 (AvrA mutant from SL1344). Comparing PhoP^c to PhoP^c AvrA− or SB300 to SB1117 with deficient AvrA, we found that AvrA expression significantly increased cell numbers at G₀/G₁ phase and decreased cell numbers at G₂/M phase (Fig. 7A; AvrA-sufficient group vs. AvrA deficient group, P < 0.01). These data indicate that AvrA expression was able to induce cell cycle arrest at G₀/G₁ in the host cells.

p53−/− cells are less susceptible to Salmonella infection.

We reasoned that p53 plays a key role in response to Salmonella stress signal and inflammation. Therefore, without p53 expression in the cells, Salmonella loses a target protein and p53−/− cells could be less susceptible to the bacterial stimulation. We assessed the effect of p53 expression on the IL-8 secretion in cells colonized with WT Salmonella. As shown here (Fig. 7B), there is a significant difference of IL-8 secretion in the cell lines with different status of p53 expression. HCT116 cells with p53 significantly increased the IL-8 protein secreted in the cell media after Salmonella colonization for 6 h. In contrast, the p53−/− HCT116 cells had less inflammatory IL-8 protein secretion after Salmonella colonization (Fig. 7B). In addition, IL-8 real-time PCR showed that the IL-8 mRNA was significantly lower with the p53−/− cells compared with the p53+/+ HCT116 cells with Salmonella colonization (Fig. 7B). These data indicate that intestinal epithelial cells without p53 are less susceptible to Salmonella infection.

AvrA overexpression in the WT Salmonella decreases the mice survival rate.

Since the inhibitory function of AvrA on inflammation in the host is observed at early stages of infection (22, 48), we also investigated whether AvrA's inhibitory effect on inflammatory response is beneficial or harmful for the host. We infected mice with WT S. typhimurium strain 14028s (WT) with insufficient AvrA expression or WT 14028s with AvrA overexpression (WTAvrA+). The survival rate of mice post-S. typhimurium infection for over 7 days is shown in Fig. 7C. At day 2, the survival percentage of the WT group is 100%, whereas the WTAvrA+ group is ∼60%. Overall, the mice infected with WT S. typhimurium strain 14028s (WT) with insufficient AvrA expression survived longer than the mice with the 14028s with AvrA overexpression (WTAvrA+) (Fig. 7C). These data indicate that AvrA overexpression renders Salmonella highly virulent, resulting in more severe infection and decreased overall survival even after 7 days of infection.

DISCUSSION

In this study, we have demonstrated that Salmonella infection increases p53 acetylation via the acetyltransferase activity of the Salmonella effector protein AvrA. Functionally, AvrA expression increased p53 transcriptional activity and induced cell cycle arrest at the G₀/G₁ phase. In addition, bacterial AvrA expression increased p53 acetylation in intestinal epithelial cells in a mouse model. To our knowledge, this is the first study that provides direct evidence and a mechanism for how a bacterial protein interferes with host responses via p53 acetylation in intestinal epithelial cells using both in vitro and in vivo systems.

Our data also indicate that Salmonella infection is able to increase the expression of MDM2, an E3 ligase for p53 which promotes the ubiquitination and proteasomal degradation of p53. Global microarray analysis on the mouse mucosa infected with Salmonella further showed that Salmonella significantly activated the p53 pathway and its target genes (Supplement Table S1). These p53 target genes respond to a variety of stress signals that impact cellular homeostatic mechanisms that monitor DNA replication, chromosome segregation, and cell division (43). p53 was first shown to be degraded through ubiquitination by the human papilloma virus E6-associated cellular protein E6AP (37). E6AP efficiently ubiquitinated and degraded p53 to replicate in the host. Taken together, these studies demonstrate that bacteria and virus exploit the p53 pathway to modulate the host response.

We generated AvrA point mutations at positions 123, 142, 179, 186 (key amino acid sites), and 180 (nonspecific amino acid site) to investigate the relative contributions of the potential catalytic residues in AvrA function. Whereas the point mutations of AvrA at positions 180 and 123 did not change the acetyltransferase activity, point mutations of AvrA at amino acid sites 186, 142, and 179 reduced its acetyltransferase activity (Supplemental Fig. S4). Although protein sequence and domain prediction can help to determine the potential catalytic amino acid (34), they are not 100% accurate. Our data indicate that AvrA could be an acetyltransferase with multiple protease domains and one amino acid point mutation may not be enough to completely abolish the acetyltransferase activity of AvrA. Moreover, recent study indicated a different initiation codon for AvrA translation (7). The initiation codon in the constructed plasmid may affect the activity of AvrA protein in vitro.

AvrA at site C186 is the key amino acid for the deubiquitinase activity of AvrA (48). Our data demonstrate that AvrA mutant C186A changed the physical binding of AvrA and p53 (Fig. 3). C186A also partially abolished the effect of AvrA on reduction of total p53 protein (Fig. 5, A and B). Our data from AvrA point mutation experiments and cell-free system assays indicate that different key amino acids of AvrA contribute to its different enzyme activities. The amino acid at position 186 may primarily control the deubiquitinase activity, whereas positions 142 and 179 regulate primarily acetyltransferase activity. AvrA enzymatically modifies diverse target proteins, including IκBα, β-catenin (48), MKK7 (7), and p53. Hence, like other deubiquitylating enzymes and acetyltransferases, AvrA appears to act on multiple substrates. Further investigation is needed to understand how AvrA plays its dual roles as a deubiquitinase and as an acetyltransferase in regulation of the signaling pathways in the host.

AvrA has multiple effects on p53 including physical binding, acetylation, and cell cycle modulation. Some of our findings are consistent with a study of the Epstein-Barr virus immediate-early protein BZLF1, which had numerous effects on p53 posttranslational modification (26). Infection of cells with this BZLF1 vector increased the level of cellular p53 but prevented the induction of p53-dependent cellular target genes, such as p21 and MDM2. BZLF1-expressing cells displayed increased p53 phosphorylation at multiple residues, and increased acetylation at lysine 320 and lysine 382. We demonstrate that Salmonella infection increases p53 acetylation. However, although some of our findings are similar, our data indicate that AvrA has no effect on the acetylation of p53 at lysine 320. Although the AvrA-deficient Salmonella strains are able to reduce p53 acetylation, they cannot completely abolish p53 acetylation. This indicates that other bacterial proteins may also participate in the modification of p53.

We recognized limitations of using in vitro cell culture. Those limitations include: 1) the use of transformed cell lines that may not necessarily reflect behavior of normal epithelial cells; 2) technical difficulties in performing biological assays beyond 48 h because cells become confluent; 3) the potential uncontrolled bacterial growth over 24 h, which may damage cells nonspecifically even with extensive washing in the presence of antibiotics; and 4) a transient transfection system that does not fully mimic the TTSS system, which normally delivers bacterial effectors into the host cell. Therefore, we further assessed in vivo role of AvrA using mouse models.

On the basis of our present findings, we speculate that Salmonella uses bacterial effectors including AvrA to increase the acetylation of p53 (Fig. 8). Bacterial effectors are injected by the Salmonella TTSS into the intestinal epithelial cells. Once in the host cells, AvrA may target p53 and act as a transacetylase to induce acetylation of p53, thus activating the p53 pathway and increasing the downstream target genes such as cyclin-dependent kinase inhibitor p21^{CIP1/WAF1/SDI1} (p21). By acting on its target genes, AvrA induces intestinal epithelial cell cycle arrest, inhibits apoptosis, and allows the intestinal epithelial cells to survive (Fig. 8). Eventually, AvrA inhibits the host's inflammatory responses and induces severe infection in the host. The cell cycle arrest is one of the biological impacts induced by Salmonella AvrA. AvrA's impacts on intestinal epithelial cell apoptosis, proliferation, and inhibition of inflammation in vivo have been already published in the previous study (22, 23, 48). p53 is known to regulate cell apoptosis and inflammation; therefore, the impacts of AvrA that we observed in the previous study may be partly due to the AvrA activation of the p53 pathway.

Bacteria play a key role in intestinal homeostasis (1, 38, 46). Although the p53 tumor suppressor has been extensively studied, many critical questions remain unanswered about the biological functions of p53. Investigating the effect of bacteria on p53 acetylation will provide further insights into beneficial and pathogenic roles of bacteria important for understanding health as well as disease states such as inflammatory bowel diseases and cancer.

GRANTS

This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant KO1 DK075386 to J. Sun.