AIP1

Enterococcus faecalis Strains Differentially Regulate Alix/AIP1 Protein Expression and ERK 1/2 Activation in Intestinal Epithelial Cells in the Context of Chronic Experimental Colitis Micha Hoffmann,† Sandra C. Kim,‡ R. Balfour Sartor,‡ and Dirk Haller*,† Chair for Biofunctionality, ZIEL-Research Center for Nutrition and Food Science, Technische Universita¨t Mu ¨ nchen, 85350 Freising-Weihenstephan, Germany, and Department of Pediatrics and Medicine, Center for Gastrointestinal Biology and Disease, University of North Carolina, Chapel Hill, North Carolina 27599 Received September 16, 2008

Monoassociation of germfree Interleukin 10 gene deficient (IL-10-/-) 129SvEv but not wild-type mice with Enterococcus faecalis induces severe chronic colitis. Bacterial strain-specific effects on the development of chronic intestinal inflammation are not understood. We investigated the molecular mechanisms of E. faecalis OG1RF (human clinical isolate, colitogenic) and E. faecalis ms2 (endogenous isolate from an IL-10-/- mouse) in initiating chronic experimental colitis using IL-10-/- mice. Monoassociation of IL-10-/- mice for 14 weeks revealed significant differences in colonic inflammation (3.6 ( 0.2 and 2.4 ( 0.6 for OG1RF and ms2, respectively) (n ) 5 mice in each group) (histological scoring (0-4)). Consistent with the tissue pathology, gene expression of the pro-inflammatory chemokine interferon-gamma inducible protein-10 (IP-10) was significantly higher in intestinal epithelial cells (IEC) derived from E. faecalis OG1RF monoassociated IL-10-/- mice. We further compared the differentially E. faecalis induced colitis on the epithelial level by 2D-SDS PAGE coupled with MALDITOF MS. Proteome analysis identified 13 proteins which were differentially regulated during disease progression in the epithelium of E. faecalis-monoassociated IL-10-/- mice. Regulation of Alix/AIP1 protein expression and ERK1/2 phosphorylation was validated in primary IEC and epithelial cell lines, suggesting a protective role for Alix/AIP1 in the process of disease progression. Alix/AIP1 protein expression was further characterized in epithelial cell lines using siRNA-mediated knock-down. Our study demonstrates E. faecalis strain-specific induction of colitis in IL-10-/- mice after 14 weeks of monoassociation. Our study suggests that Alix/AIP1 protein expression and ERK1/2 activation are decreased in severe colitis. Keywords: inflammatory bowel disease • ulcerative colitis • Enterococcus faecalis strains • intestinal epithelial cells • chronic intestinal inflammation • IL-10-/- mice • Alix • ERK • IP-10

Inflammatory bowel diseases (IBD) including Crohn’s disease (CD) and ulcerative colitis (UC) are idiopathic, spontaneously relapsing or chronic inflammatory disorders of the gastrointestinal tract mainly affecting ileum and colon. The genetic predisposition to dysregulated mucosal immune responses and the concurrent prevalence of certain environmental triggers in developed countries are strong etiologic factors for the disease pathogenesis.1-3 Different findings emphasize that the pathogenfree enteric microbiota plays a major role in the initiation and perpetuation of IBD: (i) diseases occur mainly at areas of the highest anaerobic bacterial population, either the colon (UC) or the terminal ileum (CD); (ii) probiotic bacteria help to

prevent relapses; and (iii) some of the disease are successfully treatable with antibiotics.4,5 In addition, the selective colonization of germfree rodent models for experimental colitis including IL-10 deficient mice (IL-10-/-), IL-2 deficient mice (IL2-/-) and HLA-B27 transgenic rats (HLA-B27tg) implicate Enterococcus faecalis, Escherichia coli and Bacteroides vulgatus as particularly important to the induction of colitis in these models.6-8 These mice develop chronic colonic inflammation in the presence of a specific pathogen-free microbiota but remain disease-free in the germfree environment.9,10 Recent findings clearly suggest that the epithelial cell interface of the intestine is critically important in regulating mucosal homeostasis and the development of chronic inflammation.11

* To whom correspondence should be addressed. Haller Dirk, Ph.D., Professor, Chair for Biofunctionality, ZIEL-Research Center for Nutrition and Food Science, Technische Universita¨t Mu ¨nchen, Am Forum 5, 85350 FreisingWeihenstephan, Germany. E-mail: [email protected]. Phone: ++49-816171-2026. Fax; + +49-8161-71-2824. † Technische Universita¨t Mu ¨ nchen. ‡ University of North Carolina.

E. faecalis is a commensal lactic acid bacterium which can be found in the human intestinal tract.12 Some strains of E. faecalis are used as probiotics to enhance the host immune response.13 On the other hand, strains of enterococci possess multiple antibiotic resistances and they are a leading cause of hospital acquired infections. Glycopetide resistance, especially

Introduction

10.1021/pr800785m CCC: $40.75

 2009 American Chemical Society

Journal of Proteome Research 2009, 8, 1183–1192 1183 Published on Web 01/23/2009

research articles vancomycin resistance is a great concern, since these are the antibiotics of the last choice in serious nosocomial infections with enterococci.14,15 In addition, E. faecalis plays a role in several other infections, for example, edocarditis,16 bacteremia,17 root canal18 and urinary tract infections.19 Because enterococci are both commensal intestinal bacteria and opportunistic pathogens, we asked the question whether different isolates of E. faecalis differentially induce colitis in germfree IL-10-/- mice. In this study, we monoassociated wild-type and IL-10-/- mice with E. faecalis OG1RF and the endogenous isolate E. faecalis isolate (ms2) originating from an inflamed IL-10-/- mouse. We compared the different grades of colitis in intestinal epithelial cells (IEC) using differential proteomics. As one of 13 differentially regulated proteins, we found the multifunctional adapter protein Alix/AIP1 to be down-regulated in E. faecalis OG1RF-induced severe inflammation.

Materials and Methods Mice and Bacterial Monoassociation. Germfree 129 SvEv TAC mice and germfree IL-10 gene-deficient (-/-) 129 SvEv TAC mice (derived by Dr. Edward Balish, University of Wisconsin, Madison, WI) were monoassociated at 12-16 weeks of age with the human oral isolate E. faecalis OG1RF (a generous gift from M. Huycke, University of Oklahoma, Oklahoma City, OK) (in the figures “OG1”) and with strain E. faecalis ms2, which is an endogenous isolate derived from an IL-10 -/- mouse. In collaboration with Dr. R. Balfour Sartor (University of North Carolina, Chapel Hill, NC), the mice were maintained in the National Gnotobiotic Rodent Resource Center at the University of North Carolina, Chapel Hill. Bacterial monoassociation and absence of contamination by other bacterial species were confirmed by culturing samples from the small and large intestine at necropsy and culturing serial fecal samples. Animal use protocols were approved by the Institutional Animal Care and Use Committee (IACUC), North Carolina State University. Mice were killed 1 or 14 weeks after initial bacterial colonization. Bacterial monoassociation was evaluated by plating 10fold dilution series of cecal content of all mice. Germfree mice were used as controls. Histology. Sections of proximal colon were fixed in 10% neutral buffered formalin. The fixed tissue was embedded in paraffin. Histology scoring (0-4) was analyzed by blindly assessing the degree of lamina propria mononuclear cell infiltration, crypt hyperplasia, goblet cell depletion, and architectural distortion, as previously described.20 Isolation of Primary Mouse Epithelial Cells. Colon and cecum of germfree as well as monoassociated mice were removed, opened longitudinally, washed in calcium/magnesium free PBS (Invitrogen), cut into approximately 5 mm pieces and incubated for 30 min at 37 °C in 30 mL of Dulbecco’s Modified Eagle Medium (DMEM) containing 5% FBS, 1% antibiotic antimycotic solution, and 1 mM dithiothreitol (Sigma). During incubation, samples were vortexed 3 times for 60 s. After incubation, samples were filtered and the filtrate was centrifuged for 7 min at 300g. The pellet was resuspended in 5 mL of DMEM containing 5% FBS as well as 1% antibiotic antimycotic solution. The pieces remaining on the filter were incubated for 10 min at 37 °C in 30 mL of calcium/magnesium free PBS containing 1.5 mM EDTA (Merck), vortexed for 60 s before and after incubation and filtered. The filtrate was centrifuged for 7 min at 300g and the pellet was resuspended in the same 5 mL of DMEM already mentioned above. This IEC suspension was filled on top of a 20%/40% discontinuous Percoll gradient 1184

Journal of Proteome Research • Vol. 8, No. 3, 2009

Hoffmann et al. (Sigma) and centrifuged at 600g for 30 min to purify and collect primary mouse IEC. Finally, one part of the primary IEC was lysed in Trizol Reagent for RNA isolation, the other part in Proteome Lysis Buffer for Western blotting as well as 2D-Gel Electrophoresis and stored at -80 °C. Absence of T-cell contamination was assessed by Western blotting for CD3 to confirm purity of the IEC as recently published.21 Bacteria Culture Conditions. Twenty-five milliliters of Luria Broth Medium containing 1% tryptone, 0.5% yeast extract (both Becton Dickinson, Sparks, MD) and 0.5% NaCl (Roth, Karlsruhe, Germany) was inoculated with 5% of an E. faecalis overnight culture and grown aerobically at 37 °C on a horizontal shaker. After 24 h, OD600 was determined with a biophotometer (Eppendorf), and bacteria were harvested in the stationary growth phase at 4500g for 5 min and resuspended in Mode K cell culture medium. Mode K cells were stimulated with E. faecalis using a moi of 30 (moi (multiplicity of infection) ) bacteria to epithelial cell ratio). Cell Culture. The mouse intestinal epithelial cell line Mode K (passage 15-35) was cultured in flasks, six-well or 24-well culture plates (Greiner Bio-One, Frickenhausen, Germany) in a humidified 5% CO2 atmosphere at 37 °C, as previously described.22 Cells were stimulated with bacteria (multiplicity of infection, moi 30) and with Epidermal Growth Factor (EGF) (20 ng/mL) (R&D Systems, Minneapolis, MN) as indicated. ELISA Analysis. For Enzyme Linked Immuno Sorbent Assay (ELISA), we used an Interferon-gamma inducible protein-10 (IP-10) ELISA Kit (R&D Systems, Minneapolis, MN) according to the manufacturer’s protocol and 96-well Polystyrene microplates (Greiner Bio-One, Frickenhausen, Germany). RNA Isolation and Quantitative PCR. Primary IEC from wild-type as well as IL-10 -/- mice were lysed in Trizol Reagent (Invitrogen, Karlsruhe, Germany). RNA was isolated according to the manufacturer’s protocol and dissolved in 20 µL of water containing 0,1% diethyl-pyrocarbonate. Transcription to cDNA, as well as Real Time PCR, was performed as described previously23 using a Light Cycler System (Roche Diagnostics) and 1 µg of cDNA. Primer sequences were as follows: IP-10 forward, 5′-tccctctcgcaaggac-3′ and IP-10 reverse, 5′-ttggctaaacgctttcat-3′; GAPDH forward, 5′-atcccagagctgaacg3′ and GAPDH reverse, 5′-gaagtcgcaggagaca-3′. Protein from Primary IEC. Purified IEC were lysed in a buffer containing 7 M Urea, 2 M Thiourea, 2% CHAPS, (all from Roth, Karlsruhe, Germany) and 2% Pharmalyte pH 3-10 (GE Healthcare, Freiburg, Germany). The 1% DTT (Roth Karlsruhe, Germany) and Complete Mini Protease Inhibitor (Roche Diagnostics, Mannheim, Germany) were freshly added prior to lysis. Complete Mini Protease Inhibitor was used in a ratio of 1 tablet for every 12.5 mL of extraction volume. Every sample was subjected to 10 ultrasonic impulses on ice (amplitude 35, cycle 0.5) using an UP200s ultrasonic processor (Hielscher Ultrasonics GmbH, Teltow, Germany) followed by 30 min incubation on ice and 30 min centrifugation at 18 000g and 4 °C for 30 min. Supernatants were stored at -80 °C. Protein concentrations were determined using the Bio-Rad Protein Assay (Bio Rad, Munich, Germany) as recommended in the manufacturer’s protocol. Isoelectric Focusing and 2D- SDS PAGE. Immobilized pH gradient (IPG) dry strips (pH 3-10, 18 cm, GE Healthcare, Freiburg, Germany) were rehydrated overnight using 350 µL of buffer containing 8 M Urea, 0.5% CHAPS, 15 mM DTT and 0.5% Pharmalyte pH 3-10. Isoelectric focusing (IEF) was performed with an IPGphor 2 (GE Healthcare, Freiburg,

research articles

E. faecalis Strains and Experimental Colitis Germany) using anodic cup loading of 500 µg of protein per IPG strip. IEF conditions were as follows: 500 V (1 min, gradient), 4000 V (1.5 h, gradient), 8000 V (28 000 Vh, Step-nhold). For the second dimension, 12.5% SDS-polyacrylamide gels were used in an Ettan-Dalt II System (GE Healthcare, Freiburg, Germany). Prior to electrophoresis, IPG-strips were incubated for 15 min in an equilibration buffer (6 M Urea, 1.5 M TrisHCl, pH 8.8, 26% glycerol, 2% SDS; all from Roth, Karlsruhe, Germany) containing 1% DTT and further 15 min in the same buffer containing 4% iodacetamide (Sigma Aldrich) instead of the DTT. IPG stripes were then embedded in 0.5% agarose gels on top of the SDS-polyacrylamide gels and electrophoresis was carried out at 4 mA per gel for 1 h followed by 12 mA per gel overnight. For protein staining, gels were fixed in 40% ethanol and 10% acetic acid (both from Roth, Karlsruhe, Germany) for 6 h followed by an overnight exposure to a Coomassie solution containing 10% (NH4)2SO4, 2% phosphoric acid, 25% methanol, and 0.625% Coomassie Brillant blue G-250 (all from Roth, Karlsruhe, Germany). Destaining of the gels was performed in double distilled water until the background was completely clear. All compared gels were simultaneously subjected to all steps of 2D-gel electrophoresis including IEF, SDS-PAGE, Coomassie staining and quantitative analysis in order to minimize variability between gels. Image Analysis and Mass Spectrometry. Coomassie-stained gels were scanned (ImageScanner, GE Healthcare, Freiburg, Germany) and analyzed by Proteom Weaver software (Definiens, Munich, Germany) including background subtraction and normalization. Reference gels from pooled IEC samples of all 5 E. faecalis OG1RF monoassociated IL-10-/- mice or wildtype mice were generated and compared with 5 gels from single E. faecalis ms2-monoassociated IL-10-/- mice or E. faecalis ms2-monoassociated wild-type mice, respectively. Only spots present in all compared gels were considered as relevant for further analyses. Spots with at least 2-fold differences in protein intensity present in at least 3 out of 5 gels were submitted to MALDI-TOF MS and 4 spots were excised. Trypsin digestion of the spots and subsequent MALDI-TOF-mass spectrometry analysis were performed as described earlier.21 Briefly, spots were alternately washed using 50 mmol/L NH4HCO3 and a 1:1 mixture of acetonitrile and 50 mmol/L NH4HCO3 (both from Roth, Karlsruhe, Germany) followed by 100% acetonitrile. Spots were digested with 6 µL of 0.02 µg/µL trypsin (Promega, Mannheim, Germany) at 37 °C overnight. Protein fragments were spotted on a MTP AnchorChip 600/384 target using the thin-layer affinity HCCA AnchorChip preparation by Bruker Daltonics. Peptide mass fingerprints were generated using an Autoflex mass spectrometer (Bruker Daltonics, Bremen, Germany) and the Flex Control as well as Flex Analysis software. Proteins were identified using the Mascot Server 1.9 based on mass searches within murine sequences only. The search parameters allowed the carboxyamidomethylation of cysteine and 1 missing cleavage. The mass accuracy was set to 100 ppm. We used the Mascot Mowse score for significant identification of proteins. The Mowse score is -10 × Log(P), where P is the probability that the observed match is a random event. Mascot defines the value for a significant hit to >61 when using the search criteria mentioned above. In Table 2, only proteins which were identified 3-4 times by MALDI-TOF MS are listed. Western Blotting. For SDS-PAGE, a Protean Mini cell (BioRad, Munich, Germany) was used according to the manufacturer’s instructions. Equivalent protein amounts from primary

IEC from 5 equally treated mice were pooled and mixed 1:1 with Laemmli Buffer before use. Stimulated Mode-K cells were lysed in 1× Laemmli. A total of 50 µg of protein (primary IEC) or 20 µL protein lysates (Mode K cells) was subjected to 10% SDS-PAGE gels. Size was confirmed with the Precision Plus Protein Dual Color Standards (Bio-Rad, Munich, Germany). Proteins were blotted on PVDF Membranes (Roth, Karlsruhe, Germany) using a Trans Blot SD Semi Dry Transfer Cell (BioRad, Munich, Germany) and 0.2 A for 50 min. Membranes were blocked with a solution of 5% (w/v) dry milk in Tris-buffered saline (all Roth, Karlsruhe, Germany) containing 0.1% Tween20 (Sigma Aldrich) for 1 h at RT. All antibodies were diluted in 5% skim milk powder as follows: Anti-Alix/AIP1 (BD Biosciences, Santa Cruz, Europe) 1:250; 1:1000, respectively, antiβ-actin (ICN, Costa Mesa, CA) 1:5000, anti-phospho-ERK (phosphorylation at Thr202 and Tyr204) 1:1000 and anti-ERK 1:1000 (Cell Signaling, Beverly, MA) and anti-CD3 (Santa Cruz, Europe) 1:1000. The antibodies were detected using anti-rabbit IgG, anti-mouse IgG and anti-goat IgG conjugated with horseradish peroxidase at a dilution of 1:1000 and an enhanced chemiluminescence light (ECL) detection kit (all GE Healthcare, Freiburg, Germany) as recommended by the manufacturer. Small Interference RNA Transfection. Mode-K cells were cultured in 24-well cell culture plates using 400 µL of DMEM per well containing 5% FBS as well as 1% antibiotic antimycotic solution and 1% L-glutamin (all Invitrogen, Karlsruhe, Germany). After reaching 50-80% confluency, cells were transfected using 10 nM synthetic Alix/AIP1 siRNA (Mm_Pdcd6ip_5_HP or Mm_Pdcd6ip_4_HP) or control siRNA (all star negative control siRNA) and 4.5 µL of High Perfect transfection reagent (all Qiagen, Hilden, Germany) according to the manufacturer’s protocol. The Mm_Pdcd6ip_5_HP si RNA resulted in better Alix/ AIP1 protein knock down and was used for all experiments. In Figure 5B, “Alix si 1” represents an experiment with 1.5 µL of High Perfect transfection reagent and “Alix si 2” an experiment using 4.5 µL. After 72 h of transfection, cells were stimulated with recombinant mouse Epidermal Growth Factor (EGF) (R&D Systems, Minneapolis, MN) for 5 min or with the E. faecalis strains at moi 30 for 1 h. For ELISA, cell culture supernatants were taken after 12 h of stimulation and stored at -20 °C. For Western blot, cells were lysed in 70 µL of La¨mmli Buffer, stored at -20 °C and analyzed as described above. Statistical Analyses. Data are expressed as the mean ( SD. Statistical analysis was performed by the two-tailed Student’s t test and considered significant if p-values were