Axin1 Prevents Salmonella Invasiveness and Inflammatory Response in Intestinal Epithelial Cells

Background Axin1 and its homolog Axin2 are scaffold proteins essential for regulating Wnt signaling. Axin-dependent regulation of Wnt is important for various developmental processes and human diseases. However, the involvement of Axin1 and Axin2 in host defense and inflammation remains to be determined. Methods/Principal Findings Here, we report that Axin1, but not Axin2, plays an essential role in host-pathogen interaction mediated by the Wnt pathway. Pathogenic Salmonella colonization greatly reduces the level of Axin1 in intestinal epithelial cells. This reduction is regulated at the posttranslational level in early onset of the bacterial infection. Further analysis reveals that the DIX domain and Ser614 of Axin1 are necessary for the Salmonella-mediated modulation through ubiquitination and SUMOylation. Conclusion/Significance Axin1 apparently has a preventive effect on bacterial invasiveness and inflammatory response during the early stages of infection. The results suggest a distinct biological function of Axin1 and Axin2 in infectious disease and intestinal inflammation while they are functionally equivalent in developmental settings.

Bacterial infection is a stress to the host. Pathogens use a type three secretion system (TTSS) to inject bacterial pathogenic proteins, called effectors, into host cells [14]. Virulence bacterial effectors, such as Salmonella AvrA, mimic the activity of a eukaryotic protein and debilitate their target cells [15]. Our publications and those of others demonstrate that b-catenin/Wnt [16,17,18,19,20,21,22], JNK/MAPK [15,23], and p53 [24] are involved in bacterial infection and intestinal inflammation. Axin is a key player for the b-catenin/Wnt, JNK and p53 pathways. However, it is not known whether Axin directly regulates Salmonella-induced inflammation in the intestine.
In this current study, we investigated the molecular mechanism and physiological roles of Axin1-Salmonella interactions. We found that pathogenic Salmonella colonization decreased the Axin1 protein expression in intestinal epithelial cells at the posttranscriptional level. Bacterial protein AvrA expression stabilizes Axin1 protein and beta-catenin/Axin interactions. Axin1 exhibited a specific role in inhibiting Salmonella invasion and bacterial inflammation. It is affected by the level of Axin1, but not Axin2. The resulting data indicate an essential role of intestinal Axin1 in modulating host defense against pathogen-induced inflammation.

Axin1 responds to Salmonella treatment
To determine whether Axin protein plays a role in epithelial-Salmonella interactions, we tested human intestinal epithelial HCT116 cells with wild-type (WT) Salmonella ATCC 14028s. We found that WT Salmonella significantly decreased the total amount of Axin1 in host cells after bacterial colonization for only 1 hour (Fig. 1A. HCT116). To look at the generality of our observation, we further investigated the response in the human colonic epithelial cell lines HT29C19A and CaCo2BBE. We had to use these cell lines because there is no non-cancer and nontransformed colon cell line available in the field. A similar change in Axin1 reduction by pathogenic WT Salmonella was found (Fig. 1A). To test if the response is specific to Salmonella, we also treated cells with inflammatory cytokine TNF-a, human commensal bacterial E. coli F18, and probiotic strain Lactobacillus rhamnosus GG (Fig. 1B). However, we did not see the similar alternation of Axin1. Furthermore, we found that the Salmonellainduced Axin1 reduction occurred in a time-dependent manner. WT Salmonella colonization of the cells for only 30 minutes was able to decrease Axin protein expression, and the effect could last for more than 60 minutes (Fig. 1C).
Using RT-PCR, we investigated Axin mRNA expression in intestinal epithelial cells. The transcriptional levels of Axin1 and 2 were not significantly changed by WT Salmonella in vitro ( Fig. S1A and B). Overall, our data showed that pathogenic Salmonella reduces Axin1 protein.

Reduction of Axin1 protein in the Salmonella-infected mouse intestine in vivo
Although cell lines provide useful information about the effects of Salmonella on Axin1 expression, these models lack the structural and biological relationships that exist in vivo. Therefore, we determined whether bacteria modulate Axin1 expression in streptomycin-pretreated C57BL6 mice colonized with bacterial strains. Streptomycin treatment is known to diminish the intestinal flora and to render mice susceptible to intestinal colonization by various microorganisms. In previous studies, we have established the streptomycin-pretreated Salmonella-colitis mice to understand the host-pathogen interactions [19,20]. Post infection 8 hours is correlated with the early stage of in vitro bacterial colonization. Here, we found that Axin1 protein was significantly reduced by pathogenic Salmonella 8 hours postinfection, the early stage of infection in vivo (Fig. 1D). Additionally, the Axin1 mRNA level was not changed by Salmonella infection in vivo (Fig. S1C).

Axin1 protein synthesis and degradation in epithelial cells colonized with Salmonella
To explore the molecular mechanism of Salmonella-induced Axin 1 reduction, we tested Axin1 protein synthesis using cycloheximide (CHX). CHX is an inhibitor of eukaryotic protein biosynthesis and is commonly used to determine protein half-life. Cells were treated with CHX; Axin1 protein synthesis was reduced after 4 hours of treatment without Salmonella ( Fig. 2A). In contrast, Axin1 was dramatically reduced to 40% 2 hours after CHX and Salmonella treatment. A line chart further shows that the level of Axin1 protein synthesis was higher in untreated cells than in the Salmonella-treated cells. Hence, the rate of Axin1 degradation is faster in Salmonella-treated cells ( Fig. 2A).

Axin1 is regulated by Salmonella at the post-translational level
Because Axin protein destabilization occurred in the early stage of Salmonella invasion, we hypothesized that Axin1 was regulated at the post-transcriptional level upon Salmonella stimulation. It is known that Axin is phosphorylated and ubiquitinated and then is degraded by the proteasome [25]. Hence, we investigated whether Salmonella reduced Axin protein through increased ubiquitination. By Western blot assay, ubiquitinated protein appears as a smear of bands above the regular band of the target protein. We treated epithelial cells with Salmonella and tested Axin expression and ubiquitination using immunoprecipitation. Our data indicated that Salmonella treatment induced more ubiquitinated Axin1 compared to control cells. The ubiquitinated Axin1 (Ub-Axin) was enhanced after Salmonella colonization for 60 minutes (Fig. 2B).
Recent studies demonstrated that Axin is also regulated through SUMOylation [13,25,26]. We tested Axin1 SUMOylation in the Salmonella-treated epithelial cells. We found that the SUMOylated Axin1 was enhanced by treatment with WT Salmonella (Fig. 2C).
To determine whether Axin1 reduction occurs through increased proteasome degradation, cells were treated with the proteasome inhibitor MG262. In the presence of MG262, the level of Axin1 protein was comparable to that of the control cells (Fig. 2D). Taken together, these data showed that increased ubiquitination, SUMOylation, and proteasome degradation of Axin1 occurs during Salmonella infection.

The amino acids specifically required for Salmonellainduced Axin1 degradation
To explore the molecular mechanism of Axin1 degradation in Salmonella-infected cells, we established an Axin overexpression system using HCT116 cells transiently transfected with a c-myctagged Axin1 protein (Fig. S2A). A sketch of the generated Axin1 mutations is shown in Fig. 3A. Axin1 contains a canonical regulator of G protein signaling (RGS) core domain and a DIX domain [8,27]. We tested the Axin1 levels and its stabilization using these Axin1 mutants. We found that the levels of Axin1 DRGS domain mutant were significantly reduced after Salmonella infection (Fig. 3B). In contrast, the level of Axin1DDIX did not decrease after Salmonella infection. In addition, the Axin1 DRGS DDIX double mutation lost the target domain, and its protein level was not reduced by Salmonella infection. Moreover, IP data indicated that there was no change in Axin SUMOylation after Salmonella infection in cells transfected with Axin1 DDIX (Fig. 3C). In contrast, the SUMOylation of Axin DRGS was still enhanced by Salmonella.
We further tested whether Salmonella exploited Axin1 at key amino acid sites that regulate ubiquitination and SUMOylation. The Axin1S614A mutant lost its ability to be ubiquitinated and was not degraded by Salmonella (Fig. 3D). IP data indicated that Salmonella treatment did not change the ubiquitination of AvrA S614A, whereas K954A ubiquitination was enhanced by Salmonella (Fig. 3E). Moreover, Salmonella infection did not enhance Axin1S164A SUMOylation, whereas SUMOylation of Axin K954A was still high (Fig. 3F). We also tested the other Axin1 mutations (Fig. S3A); however, the protein levels of these Axin1 mutants did not change after Salmonella infection ( Fig. S3B-D). Taken together, our data suggested that Salmonella targeted the DDIX domain of Axin1 and ser614 to reduce Axin1 protein level during host-bacterial interactions.
We also tested the function of Axin2 in the HCt116 cells. However, transfected Axin2 did not change after Salmonella treatment (Fig. S2C). In contrast, both endogenous Axin1 and exogenous c-myc tagged Axin1 decreased cells treated with WT Salmonella (Fig. S2B).

Decreased b-catenin/Axin binding in Salmonella colonized cells
Axin can interact with many proteins in the cell [8]. It forms a complex with GSK-3b and b-catenin and promotes GSK-3b-dependent phosphorylation of b-catenin. Using immunoprecipitation and western blot, we found that WT Salmonella decreased the amount of Axin1 in a complex with b-catenin, whereas the binding between Axin1 and b-catenin was not altered by PhoP C colonization (Fig. 4A). Densitometry analysis showed significant decreases in the binding of Axin1 and b-catenin after WT Salmonella colonization.
We also tested another Axin/b-catenin-associated protein, GSK-3b (Fig. 4B). The total amount of Axin bound to GSK-3bwas also reduced by WT Salmonella, whereas PhoP c colonization stabilized the Axin/GSK-3b interaction within the cell. Furthermore, densitometry analysis showed a significant reduction of Axin/GSK-3b interaction induced by WT Salmonella.
WT Salmonella also reduced the amount of APC/Axin-binding proteins, whereas levels of the APC/b-catenin complex were not significantly changed after Salmonella infection (Fig. 4C). Taken together, these IP data indicated that pathogenic Salmonella colonization reduces Axin1 in the intestinal epithelial cells, thereby decreasing the dynamic interaction between Axin and its binding partners.

Axin1 protein protects cells from Salmonella invasion
For the physiological relevance of Axin involved in the Salmonella-host interactions, a green-fluorescence-tagged Salmonella strain was used to detect bacterial invasion in intestinal epithelial cells. There was less green Salmonella in the cells with Axin1 overexpression compared to the normal cells (Fig. 5A). We further counted the numbers of Salmonella invading the HCT116 cells with normal or over-expressed Axin1 protein. We found that Axin1over-expressing epithelial cells had less internalized Salmonella than control cells with regular Axin1 expression (Fig. 5B). We also examined the number of cell-associated bacteria, including bacteria adhered to and/or internalized into the epithelial monolayers. Axin1 expression did not change the number of associated Salmonella in the host cells (Fig. S4A).
We also investigated the distribution of Axin1. Immunostaining data showed that Axin1 protein was dispensed diffusely in the cytoplasm in HCT116 cells without bacterial treatment. Postpathogenic Salmonella infection, Axin1 was scattered as dots in the cytoplasm. Some Axin1 relocated to the cell membrane (Fig. 5C). Axin1 relocation to the cell membrane may help to block the bacterial invasion.

The DIX domain is required for Axin1's function in inhibiting bacterial invasion
We investigated the effect of Axin1 mutations on bacterial invasion. As shown in Fig. 5D, cells with DRGS overexpression had less invaded Salmonella, whereas DDIX-expressing cells had a significantly higher amount of Salmonella invasion. Axin1 mutations did not change the number of bacteria associated with the epithelial cells (Fig. S4B). Taken together, these data further indicate that Axin1DDIX was targeted by Salmonella. Therefore, the mutated Axin1DDIX could not protect cells from bacterial invasion. However, Axin2 did not have any effects on Salmonella attachment and invasion ( Fig. S5A & 5B).

Axin1 protein is directly involved in Salmonella-induced inflammation
We reasoned that Axin1 expression could protect cells from Salmonella-induced inflammation. We tested the inflammatory response of IL-8 in cells overexpressing Axin1. Upon Salmonella colonization, IL-8 mRNA expression decreased in cells overexpressing Axin1 compared to normal HCT116 cells (Fig. 6A). We also found that the IL-8 protein secretion was significantly lower in cells overexpressing Axin1 (Fig. 6B). In addition, Axin2 overexpression did not have any effect on the expression of Salmonellainduced IL-8 at either the transcriptional (data not shown) or the translational level (Fig. S5C).
To further investigate the molecular mechanisms, we blocked Axin1 expression using siRNA. As shown in Fig. 6C-D, the knockdown of Axin1 significantly enhanced IL-8 mRNA expression and IL-8 protein secretion in the cell culture media compared with cells without treatment or with control siRNA. Moreover, the cells overexpressing Axin DRGS produced significantly less IL-8, whereas cells overexpressing DDIX had higher IL-8 secretion induced by Salmonella (Fig. 6E). Taken together, our data indicate that WT Axin1 plays a protective role in bacterial intestinal inflammation. The DIX domain is required for Axin1's functions.

Discussion
In the current study, we showed that Axin1, but not Axin2, plays an important role in inhibiting bacterial invasion and inflammation in intestinal epithelial cells. The reduction of Axin1 occurred at an early stage of Salmonella infection (Fig. 6F). Axin1 protein was modified by Salmonella-colonization through ubiquitination and SUMOylation. Our findings further demonstrated that the protein expression level of Axin may determine the host's inflammatory responses to Salmonella. Salmonella-induced IL-8 was significantly lower in cells with Axin overexpression compared to those in the normal intestinal epithelial cells. In contrast, when Axin1 was knockdown, Salmonella induced significantly higher level of IL-8 expression. Axin1DDIX domain and ser164A is specifically required for Salmonella-induced degradation of Axin1. Although Axin 1 and 2 are functionally equivalent in developmental settings [28,29,30], a distinct biological function of Axin1 and 2 was identified in host defense. We reported that Axin2 was not involved in inhibiting bacterial invasion and inflammation, whereas Axin1 plays a protective role for host defense against pathogenic bacteria.
Our previous studies demonstrated that Wnt/b-catenin is involved in Salmonella-induced intestinal inflammation [19,20,21,22,31]. Classically, Axin1 is viewed as a negative regulator of the Wnt/b-catenin pathway in development and cancer. The degradation of Axin1 leads to activation of the Wnt/ b-catenin pathway. The binding between b-catenin and Axin is modified by Salmonella colonization. This may further explain our previous observation that the Wnt/b-catenin pathway is involved in Salmonella-induced intestinal inflammation. In in vitro studies, we used HCT116 cells with full length APC and HT29 and Caco2BBE with truncated APC. We found that the length of APC may influence the abundance of Axin1 in the cells. However, it does not change the effect of Salmonella-induced downregulation of Axin1. Recent studies demonstrate that Wnt/b-catenin signaling is regulated through both Axin-dependent proteolysis and Axin-independent proteolysis [32]. Hence, some of the effects we observed, such as Axin1 overexpression inhibiting Salmonella invasion, could be b-catenin-independent.
In Axin1 over-expressing cells, there was less Salmonella-induced IL-8 secretion. This finding indicates that Axin1 expression can protect cells from inflammation. Recently, studies have implied the Wnt signaling in the activation of proinflammatory mediators in inflammatory disorders [33]. Patients with inflammatory bowel diseases (IBD) have significantly higher expression of Wnts compared to non-IBD patients [34]. There was reduced expression of the Wnt-signaling pathway transcription factor TCF-4 in ileal Crohn's disease [35]. However, some of these studies have focused on the responses of monocytes to the inflammatory cytokines in the immune system through Wnts [36]. The effects and molecular mechanisms of Axin, the key scaffold of Wnt signaling, in regulating intestinal inflammation remain unknown.
Axin1 has been shown to interact with many critical proteins, including b-catenin, GSK3 b, LRP5, and DVL1 [25,37,38,39]. In addition, Axin1 expression determines p53 activation in cell development [12,40]. It is still unknown how Axin1 may synergistically regulate multiple signaling pathways, thereby altering the host response to bacterial infection. A limitation of the current work is that we could not use whole Axin1 knockout mice for further in vivo studies, because its phenotype in mice is lethal. The Axin1 conditional mutant mice are available now [41], we will focus on the physiological role of intestinal Axin in in vivo in anti-inflammation and anti-infection for future studies.
In summary, the current study determines the protective role of Axin1 in the Salmonella invasion and bacterial-induced inflammation. Our results indicate that a previously undefined role of Axin1 in host-pathogen interactions and inflammation.

Ethics statement
All animal work was approved by the University of Rochester University Committee on Animal Resources (UCAR) committee (UCAR 2007-065). If a mouse showed that it had aspirated fluid or significant body weight loss (10% or more), and did not die immediately, the mouse was humanely euthanized.

Bacterial strains and growth condition
Bacterial strains used in this study included Salmonella typhimurium wild-type strain ATCC14028 (WT-SL) and non-pathogenic Salmonella mutant strain PhoP c [42]. Non-agitated microaerophilic bacterial cultures were prepared as previously described [22].

Streptomycin pre-treated mouse model
Animal experiments were performed using specific-pathogenfree female C57BL/6 mice (Taconic, Hudson, NY) that were 6-7 weeks old, as previously described [20,43,44]. The protocol was approved by the University of Rochester University Committee on Animal Resources (UCAR). Water and food were withdrawn 4 hours before oral gavage with 7.5 mg/mouse of streptomycin (100 ml of sterile solution or 100 ml of sterile water in control). Afterwards, animals were supplied with water and food ad libitum. Twenty hours after streptomycin treatment, water and food were withdrawn again for 4 hours before the mice were infected with 1610 7 CFU of S. Typhimurium (100 ml suspension in HBSS) or treated with sterile HBSS (control). Eight hours after infection, mice were sacrificed, and tissue samples from the intestinal tracts were removed for analysis.

Axin1 siRNA
HCT116 cells were grown in 12-well plates. The cells were transfected with on-Target plus smart pool human Axin 1 siRNA (Dharmacon Inc., Lafayette, MO) or scrambled siRNA control (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) using Surefect reagent (SABiosciences, Frederick, MD). After transfection for 72 hours, cells were colonized by Salmonella for 30 min, washed, and incubated for 30 min in DMEM with Gentamicin (500 mg/ml), and then the levels of Axin and b-actin were assessed by western blot.

Immunoblotting
Mouse epithelial cells were scraped and lysed 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 then the protein concentration was measured. 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 then sonicated. Equal amounts of protein were separated by  Membranes that were probed with more than one antibody were stripped before re-probing.

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 incubated with 2 mg of primary antibodies for 1 hour at 4uC. Protein A-agarose was added to the lysate and incubated for 30 min with agitation at 4uC before being washed with cold immunoprecipitation buffer. The pellet was resuspended in 0.1 M glycine, pH 2.5, and incubated with agitation for 10 min at 4uC and then centrifuged at 9,000 g for 2 minutes. The supernatant was removed and neutralized with 1 M Tris?HCl, pH 8.0. The samples were diluted with concentrated (56) electrophoresis sample buffer (125 mM Tris, pH 6.8, 4% SDS, 10% glycerol, 0.006% bromophenol blue, 2% b-mercaptoethanol), boiled for 5 minutes, separated by SDS-polyacrylamide gel electrophoresis, and transferred onto a nitrocellulose membrane. Membrane blots were probed with secondary antibody and visualized by ECL.

Immunofluorescence
Colonic tissues from the proximal and distal portion of the colon were freshly isolated and paraffin-embedded after fixation with 10% neutral buffered formalin. Immunofluorescence was performed on paraffin-embedded sections (1 mm) of mouse colons. After preparation of the slides as described previously [19], slides were incubated in 3% hydrogen peroxide for 20 minutes at room temperature to block endogenous peroxidase activity, followed by incubation for 20 minutes in 5% BSA with 0.1% saponin in PBS to reduce nonspecific background.
Cultured cells were rinsed three times in HBSS, fixed for 20 minutes in 100% cold ETOH, and permeabilized for 10 minutes with 0.2% Triton X-100, followed by three rinses with Hanks HBSS, and incubation for 30 minutes in 10% BSA in HBSS to reduce nonspecific background. The permeabilized cells or tissue samples were incubated with primary antibodies for 90 minutes at room temperature. Samples were then incubated with goat anti-rabbit Alexa Fluor 488 or goat anti-mouse Alexa Fluor 594 (Molecular Probes, CA; 1:200) and DAPI (Molecular Probes 1:10,000) for 1 hour at room temperature. Tissues or cells were mounted with SlowFade (SlowFadeH 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.

S. typhimurium invasion of human epithelial monolayers
Infection of HCT116 cells was performed by a previously described method [44,45]. Bacterial solution was added, and bacterial invasion was assessed after 30 minutes. Cell-associated bacteria, representing bacteria adhered to and/or internalized into the monolayers, were released by incubation with 100 ml 1% Triton X-100 (Sigma). Internalized bacteria were those obtained from lysis of the epithelial cells with 1% Triton X-100 30 minutes after the addition of gentamicin (50 mg/ml). For both cellassociated and internalized bacteria, 0.9 ml LB broth was added, and each sample was vigorously mixed and quantitated by plating for CFU on MacConkey agar medium.

Salmonella-induced human IL-8 secretion
HCT116 cells were cultured in DMEM, followed by Salmonellacontaining HBSS for 30 minutes, washed 3 times in HBSS, and incubated at 37uC for 6 hours. Cell supernatants were removed and assayed for IL-8 by ELISA in 96-well plates.

Real Time quantitative PCR
Total RNA was extracted from epithelial cell monolayers or mouse colonic epithelial cells using TRIzol reagent (Invitrogen, Carlsbad, CA). The RNA integrity was verified by gel electrophoresis. RNA reverse transcription was done 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 singlecolor real-time PCR detection system (Bio-Rad) and iQ SYBR green supermix (Bio-Rad) according to the manufacturer's directions. All expression levels were normalized to b-actin levels of the same sample. 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 using Lasergene software (Table 1)

Statistical Analysis
Data are expressed as mean 6 SD. Differences between two samples were analyzed by Student's t test. P-values of 0.05 or less were considered statistically significant. Differences among three or more groups were analyzed using ANOVA (SAS 9.2 version, SAS Institute Inc., Cary, NC).