H. pylori Output Strains from DFMO-Treated Gerbils Display Reduced CagA Translocation.

H. pylori isolates recovered from the stomachs of infected gerbils were used to evaluate T4SS functionality. H. pylori output strains from DFMO-treated gerbils had a significantly reduced capability to translocate CagA, assessed by the ratio of phosphorylated CagA to total CagA, compared with output isolates from untreated gerbils (Fig. 1 D and E). Similarly, H. pylori isolates from DFMO-treated gerbils showed a reduced ability to induce NF-κB activation in AGS cells transfected with a luciferase reporter (Fig. 1F). In parallel, CXCL8 expression, which partially depends on CagA phosphorylation in gastric epithelial cells (22), was also significantly reduced in cells infected with output strains from the DFMO-treated group compared with isolates from infected gerbils without DFMO (Fig. 1G).

Output Strains from DFMO-Treated Gerbils Exhibit cagY Rearrangements.

Using PCR and restriction fragment length polymorphism (RFLP) analysis for the cagY gene, we found that none of the output strains from the untreated animals showed rearrangements in cagY, whereas multiple isolates from DFMO-treated gerbils exhibited alterations in the RFLP pattern compared with the parental strain 7.13 (Fig. 2A). When we extended our analysis to an increased number of output isolates, we determined that 35.7% of the strains recovered from DFMO-treated animals exhibited cagY gene rearrangements, whereas there were no strains with cagY rearrangements isolated from untreated gerbils. Of the 14 DFMO output strains tested, single-molecule real-time (SMRT) sequencing demonstrated an insertion in the middle repeat region (MRR) of the cagY gene in 1 isolate and various deletions in the 2A and 2B motifs of the MRR (12, 13) in four strains (Fig. 2B and SI Appendix, Fig. S2). There were no alterations in the cagY sequence in the output strains from the untreated gerbils (SI Appendix, Fig. S2).

Open in a separate window

Fig. 2.

cagY rearrangements in gerbil output strains. (A) cagY RFLP profile of H. pylori 7.13 parental and output strains from infected gerbils. Arrows indicate strains with rearrangements in cagY compared with the parental strain (DFMO-2, DFMO-4, DFMO-5, and DFMO-8). (B) Schematic representation of the cagY sequencing by SMRT. Repeat domains and identity with VirB10 are shown. The orange and blue boxes represent the repeat motifs designated 2A and 2B. The dashed boxes indicate insertion of a motif. (C) Western blot analysis for CagY in output strains from gerbils in the control and DFMO-treated group. (D) cagY RFLP profiles of DFMO output strains and parental strain 7.13 complemented with a rearranged cagY. (E) Western blot analysis for CagY in output strains from gerbils in DFMO-4, DFMO-8, and complemented strains. (F) Protein levels of CXCL8 quantified by ELISA from supernatants of AGS cells cocultured for 24 h with the parental strain 7.13, DFMO output strains, or parental strain 7.13 complemented with rearranged cagY. Error bars represent SEM. ANOVA with Newman–Keuls multiple comparisons test was used. (G) Western blot for pCagA in AGS cells cocultured for 4 h with parental strain 7.13, DFMO output strains, or parental strain 7.13 complemented with rearranged cagY.

Using an antibody directed against the MRR (9), we found alterations in the CagY protein (Fig. 2C) in the same isolates (DFMO-2, DFMO-4, DFMO-5, and DFMO-8) that had cagY rearrangements in Fig. 2A, compared with the parental strain. It should be noted that for the DFMO-7 output strain, it was not possible to amplify cagY gene and the protein CagY was not detected (Fig. 2 A and C).

Numerous genes in H. pylori besides cagY contain DNA repeats and have the potential to undergo recombination or rearrangements (23). Two of those genes, namely fucosyltransferase (fucT) and N-acetylmuramoyl-l-alanine amidase (amiA), were PCR-amplified from the control and DFMO output strains to assess rearrangements by RFLP analysis. No rearrangements were observed in strains isolated from the control or DFMO-treated gerbils (SI Appendix, Fig. S3). Additionally, none of the output strains from control and DFMO-treated animals showed alterations in cagA sequence (SI Appendix, Fig. S4).

The mutation rate in H. pylori is considerably higher than in other bacteria (24) and leads to elevated frequency of deletions between direct repeats and intergenomic recombination (25). However, we did not find a significant difference in the frequency of spontaneous rifampicin-resistant mutants between control and DFMO output strains (SI Appendix, Table S1).

cagY Rearrangements Induced by DFMO Abrogate H. pylori Virulence and Carcinogenic Potential.

To establish the contribution of cagY rearrangements to the alterations in CagA translocation and subsequent cell signaling in the DFMO output strains, cagY was deleted from the parental strain 7.13 and then complemented with the cagY from two selected DFMO output strains (DFMO-4 and DFMO-8) that exhibited cagY rearrangements. RFLP analysis of cagY (Fig. 2D) and SMRT sequencing (SI Appendix, Fig. S5) confirmed that the strain 7.13 complemented with cagY from DFMO-4 had the same pattern as the output strain DFMO-4; similar results are shown for the complementation with DFMO-8. The complementation was also confirmed at the protein level by Western blotting for CagY (Fig. 2E). CXCL8 secretion was significantly reduced in AGS cells infected with DFMO-4 and DFMO-8 output strains compared with the level of this chemokine induced by the parental strain 7.13 (Fig. 2F). Levels of CXCL8 secretion induced by strain 7.13 complemented with cagY from DFMO-4 and DFMO-8 were also markedly decreased compared with the parental strain 7.13 (Fig. 2G). CagA translocation was abrogated in the output strains (DFMO-4 and DFMO-8) and also in the complemented strains (Fig. 2G).

Mongolian gerbils were then infected with H. pylori parental strain 7.13, the output strain from a DFMO-treated gerbil (DFMO-4), or 7.13 [DFMO-4], the parental strain complemented with the DFMO-4 cagY and analyzed after 12 wk to assess disease outcome. Inflammation scores were significantly decreased in gerbils infected with the output strain DFMO-4 and with the parental strain harboring the cagY gene from the isolate DFMO-4 (Fig. 3A). Consistent with the attenuated inflammatory/immune response, increased colonization levels were observed in animals infected with the strain DFMO-4 or 7.13 [DFMO-4] compared with the parental strain 7.13 (SI Appendix, Fig. S6). Invasive adenocarcinoma and/or dysplasia was observed in 87.5% of the gerbils infected with strain 7.13 (Fig. 3B). There was a marked reduction in carcinogenesis in gerbils infected with the strains with altered cagY. Notably, none of the gerbils infected with DFMO-4 or 7.13 [DFMO-4] developed invasive adenocarcinoma (Fig. 3B). Also, none of the gerbils infected with DFMO-4 developed dysplasia, and only 16.6% of the animals infected with 7.13 [DFMO-4] developed dysplasia (Fig. 3B). Although none of the gerbils infected with 7.13 [DFMO-4] developed cancer, we observed variable levels of inflammation, and this could indicate that other genes besides cagY may be altered in DFMO-4. A representative invasive adenocarcinoma case is shown from a gerbil infected with strain 7.13 (Fig. 3C). In contrast, gerbils infected with DFMO-4 or 7.13 [DFMO-4] exhibited normal histology and gastritis without dysplasia, respectively. Uninfected animals did not develop any disease (SI Appendix, Fig. S1A). When output strains were recovered from these infected gerbils, we observed that in the vast majority of the strains the cagY RFLP profile was conserved compared with the input strain. We found only 1/24 output strains from parental 7.13-infected gerbils, 1/21 DFMO-4 output strains, and 2/24 7.13 [DFMO-4] output strains had new cagY rearrangements (SI Appendix, Fig. S7 A–C). These results indicate that further cagY modifications are not commonly observed in gerbils, even in isolates with rearranged genes, further supporting the concept that DFMO treatment induces a significant increase in the frequency of cagY rearrangements in gerbils.

Open in a separate window

Fig. 3.

Mongolian gerbils infected for 12 wk with H. pylori strain 7.13, DFMO output strain (DFMO-4), or 7.13 parental strain complemented with rearranged cagY (7.13[DFMO-4]). (A) Inflammation score in the gastric tissues of gerbils infected with 7.13, DFMO-4, or 7.13 [DFMO-4]. Error bars represent SEM. ANOVA with Newman–Keuls multiple comparisons test was used for A. (B) Frequency of diagnoses in gerbils. The adjacent numbers correspond to the number of gerbils in each section of the bar. ^§§§P < 0.001 for DFMO-4 and 7.13 [DFMO-4]-infected gerbils versus 7.13-infected, comparing cancer frequency; ^¶¶¶P < 0.001 for DFMO-4 and 7.13 [DFMO-4]-infected gerbils versus 7.13-infected, comparing cancer + dysplasia frequency. For diagnosis comparisons significance was calculated using Fisher’s exact test. (C) H&E staining of gastric tissues, showing invasive adenocarcinoma in a 7.13-infected gerbil, normal histology in a DFMO-4-infected animal, and gastritis in a 7.13 [DFMO-4]-infected gerbil. None of the seven uninfected gerbils developed dysplasia or adenocarcinoma. (Scale bars: 100 μm.)

In Vitro Treatment of H. pylori with DFMO Alters T4SS Functionality by Promoting cagY Rearrangements.

Previous studies have indicated that increased frequency of cagY rearrangements is associated with an increased inflammatory response from the host (26). However, while we observed cagY rearrangements in multiple strains isolated from DFMO-treated gerbils, the inflammation scores in these animals were not significantly different from those in the control group (Fig. 1C). We therefore reasoned that DFMO might directly affect H. pylori T4SS function. H. pylori was thus serially passaged on trypticase soy blood agar plates supplemented or not with DFMO or with ornithine, used as a control. First, we found that DFMO was detected in bacteria grown on DFMO-containing agar plates (SI Appendix, Fig. S8A), evidencing DFMO uptake by H. pylori. We found rearrangements of the cagY gene in the DFMO-treated bacteria as early as passage number 5 (Fig. 4A); this rearrangement was consistently observed up to passage number 20 (Fig. 4A). In contrast, no rearrangements were observed in the strains serially passaged on regular plates or on ornithine-containing plates (Fig. 4A). In addition to strain 7.13, the serial passage of the strain PMSS1 on DFMO plates, but not ornithine-supplemented plates, led to cagY rearrangements (SI Appendix, Fig. S8B). Moreover, we did not find changes in the global frequency of mutations between the strains serially passaged on regular blood agar and on DFMO-containing plates (SI Appendix, Table S2).

Open in a separate window

Fig. 4.

H. pylori serially passaged on DFMO-supplemented plates. (A) cagY RFLP profile of H. pylori 7.13 serially passaged on regular plates (–) or plates supplemented with DFMO (D) or ornithine (O). RFLP profiles are shown for passage 1 (p1), 5 (p5), or 20 (p20). For all of the RFLP studies, four individual colonies per condition were analyzed. (B) NF-κB activation in AGS cells cocultured for 3 h with H. pylori serially passaged 20 times on control, DFMO, or ornithine-containing plates. Western blot (C) and densitometric (D) analysis for pCagA in AGS cells cocultured with gerbil output strains for 4 h. ANOVA with Newman–Keuls multiple comparisons test was used for B and D.

Alterations in cagY after in vitro treatment with DFMO were associated with a decreased ability of H. pylori to induce NF-κB activation, showing more than 50% reduction in NF-κB activation compared with the parental strain 7.13 (Fig. 4B). CagA translocation was also significantly reduced in the DFMO-treated isolates, but not in the control or ornithine-treated bacteria (Fig. 4 C and D).

DFMO Induces Oxidative DNA Damage in H. pylori.

During natural infection, H. pylori is exposed to reactive oxygen and nitrogen species generated by the host that can cause DNA damage in the bacteria (27). These oxidative agents induce DNA changes that include purine oxidation, which tend to be mutagenic (28). Hence, to determine the potential oxidative effects of DFMO on H. pylori, we first assessed the expression of the oxidative stress response genes encoding catalase (katA) and superoxide dismutase (sodB). DFMO treatment increased katA and sodB mRNA expression compared with the untreated bacteria (Fig. 5). Ornithine treatment did not induce a significant change in the expression of these genes.

Open in a separate window

Fig. 5.

Effect of DFMO on sodB and katA mRNA expression in H. pylori. H. pylori strain 7.13 was treated with DFMO or ornithine for 2, 4, or 8 h in liquid culture. sodB (A) and katA (B) mRNA expression, determined by real-time PCR, is expressed as fold increase compared with the untreated control at each time point. ^§§P < 0.01 versus untreated control. ^§§§P < 0.001 versus untreated control. ^¶P < 0.05 versus DFMO-treated. ^¶¶P < 0.01 versus DFMO-treated. ^¶¶¶P < 0.001 versus DFMO-treated. ANOVA with Newman–Keuls multiple comparisons test was used for A and B.

DNA damage was directly assessed by fluorescent staining for 8-oxoguanine using fluorescently labeled avidin, which binds with high specificity to the oxidized base (29). H. pylori treatment with DFMO for 24 h induced DNA damage, detected by confocal microscopy (Fig. 6A), and we also observed a significant increase in the percentage of 8-oxoguanine-positive cells compared with the untreated bacteria by flow cytometry (Fig. 6 B and C). Ornithine treatment did not induce DNA damage in H. pylori (Fig. 6). Exposure to atmospheric concentrations of oxygen was used as a positive control for induction of DNA damage (Fig. 6).

Open in a separate window

Fig. 6.

Oxidative DNA damage in H. pylori. Confocal microscopy (A) and flow cytometry plots (B) for 8-oxoguanine in H. pylori treated with DFMO or ornithine for 24 h, or exposed to oxygen for 4 h. (Scale bars: 5 μm.) (C) Quantification of 8-oxoguanine–positive bacteria by flow cytometry. ANOVA with Newman–Keuls multiple comparisons test was used.

Treatment of Bacteria with DFMO.

H. pylori was serially passaged on TSA blood agar plates supplemented or not with 1% DFMO or 1% ornithine. Bacteria were passaged on new plates after 48–72 h of culture at 37 °C and 5% CO₂.

Ethics Statement.

All experiments were conducted under protocols M/09/147 and M1600091 approved by the Vanderbilt University Institutional Animal Care and Use Committee and Institutional Biosafety Committee. Procedures were performed in accordance with institutional policies, Association for Assessment and Accreditation of Laboratory Animal Care guidelines, the American Veterinary Medical Association Guidelines on Euthanasia, NIH regulations (Guide for the Care and Use of Laboratory Animals, ref. 48), and the United States Animal Welfare Act (1966).

Animals, Infection, and DFMO Treatment.

Male Mongolian gerbils were purchased from Charles River Laboratories and had weights of 41–50 g. AIN-76A rodent diet (Bio-Serv) and water containing or not 1% DFMO was provided ad libitum 1 wk before infection with H. pylori and continued along the experiments. Gerbils were inoculated orogastrically twice with 2 × 10⁹ cfu of H. pylori 7.13 resuspended in Brucella broth. Animals were euthanized after 12 wk of infection. Colonization was assessed by serial dilution and plating of gastric tissue homogenates in TSA plates supplemented with 5% sheep’s blood, 20 μg/mL vancomycin, 30 μg/mL bacitracin, 10 μg/mL nalidixic acid, and 2 μg/mL amphotericin. H. pylori output strains were recovered and frozen for further analysis. Histologic analysis was performed in a blinded manner by a gastrointestinal pathologist (M.B.P.).

RFLP for cagY.

Bacterial DNA was isolated using the DNeasy Blood and Tissue kit (Qiagen). The cagY gene was amplified by PCR using the Expand Long Template PCR System (Roche). Reactions were prepared with 100 ng of DNA, 300 nM primers (sense 5′-AACATCACTCTCCAACCTGC-3′; antisense 5′-ATTGTAACTGGCTAACACAAGG-3′), 500 μM dNTPs, and 1.5 U of Expand Taq DNA polymerase in 20 μL. PCR conditions were as follows: one cycle of 94 °C for 2 min; 10 cycles of 94 °C for 10 s, 53 °C for 30 s, 68 °C for 5 min; 25 cycles of 94 °C for 15 s, 53 °C for 30 s, 68 °C for 5 min with an increase of 20 s per cycle; one cycle of 68 °C for 10 min and final hold at 4 °C. PCR products were precipitated with sodium acetate-ethanol and resuspended in 10 mM TrisHCl, pH 8.5 and digested overnight at 37 °C using DdeI (New England Biolabs). Digested DNA was separated by 9% acrylamide gel electrophoresis and stained with ethidium bromide.

cagY Sequencing.

Amplicons were generated by PCR as described above in the RFLP section. Products were gel-purified, using the QIAquick Gel Extraction kit. SMRTbell libraries were constructed according to the PacBio protocol for preparing Amplicon libraries using barcoded adaptors for multiplex sequencing. Briefly, 33 ng of each sample was used in the end-repair/ligation reactions using reagents from the SMRTbell Damage Repair Kit-SPv3 (PacBio) and the SMRTbell barcoded adapter complete prep kit -96 (PacBio). This was followed by clean-up step with AMPure magnetic beads (Beckman Coulter) at a 1:1 sample:beads ratio. Barcoded samples were pooled for the DNA damage repair step. Ligated-damaged repaired products were exonuclease digested (Exonuclease III and VII) and purified twice with magnetic beads as above. Approximately 250 ng of each library was obtained as quantitated by the QUBIT method (ThermoFisher). These constituted the final libraries ready for sequencing on the PacBio SEQUEL machine. Sequencing was done using the Chemistry Bundle 6.0 (Sequencing and Binding kits 2.1 in combination with the SMRTLink 6.0 software). Each library at 25 pM was applied to the PacBio SEQUEL sample plate by diffusion loading for sequencing on LR-SMRT cells with 20-h data collection and 4-h preextension time. All other steps for sequencing were done according to the recommended protocol by the PacBio sequencing calculator. The raw amplicon reads generated by the PacBio SEQUEL System were processed in PacBio SMRT Analysis (v6.0). The subreads of each sample were processed by Long Amplicon Analysis. The PCR artifacts and chimeric/duplicate sequences were identified by the UCHIME algorithm and filtered. The final consensus sequence for each sample was polished by using the Arrow model.

Sequencing of cagA.

The cagA gene was amplified by PCR using the HotStarTaq Master Mix kit (Qiagen). Reactions were prepared with 100 ng of DNA, 300 nM primers (sense 5′-ATTTTTAGCAGTCTTTGACACC-3′; antisense 5′-ATATGATACCATGAATTGGTAGC-3′), and HotStarTaq Master Mix in 20 μL. Products were gel-purified, using the QIAquick Gel Extraction kit. cagA was sequenced using the Sanger sequencing system. Sequencing primers were designed at ~800-bp intervals and are listed in SI Appendix, Table S3. Sequence fragments were aligned to assemble the full gene.

Analysis of CagA Translocation.

H. pylori strain 7.13 and gerbil output strains were cultured overnight in Brucella broth and then cocultured with AGS cells (ATCC CRL-1739) for 4 h at a multiplicity of infection (MOI) of 50. OD at 600 nm was used to quantify bacterial cultures. Cells were lysed with Tris buffered saline (TBS) containing 0.1% SDS, 1% Nonidet P-40, and 0.5% sodium deoxycholate. Lysate samples were separated by SDS/PAGE, and proteins were transferred to PVDF membranes. Blots were blocked with 1% BSA in TBS overnight at 4 °C. Membranes were then incubated with a mouse monoclonal anti-phospho tyrosine antibody (1/3,000; Santa Cruz Biotechnology), a rabbit polyclonal anti-H. pylori CagA antibody (1/4,000; Austral Biologicals), or a mouse monoclonal anti-human/mouse β-actin antibody (1/20,000; Sigma) for 1 h at room temperature using constant rocking, washed three times with 0.1% Tween-20 in TBS, and incubated with an HRP-labeled goat anti-rabbit IgG antibody or an HRP-labeled goat anti-mouse IgG antibody (1/5,000 for both; Jackson ImmunoResearch). After three more washes, detection was performed using SuperSignal West Dura Substrate for 3 min. Densitometric analysis was performed with ImageJ 1.50i software. The translocation index was calculated by the ratio of phospho-CagA to total CagA, standardized to β-actin.

NF-κB Activation Reporter Assay.

AGS cells stably expressing a luciferase-based NF-κB reporter were generated as described (49). Transfected AGS cells were cocultured with H. pylori at MOI 10 for 3 h, and lysates were prepared using Passive Lysis buffer (Promega). Lysates were mixed with the Luciferase Assay system (Promega) and luminescence was measured using a Biotek Synergy 4 microplate reader. The luminescence units obtained from the uninfected cells were used to calculate the fold increase.

Analysis of mRNA Expression.

H. pylori strain 7.13 and gerbil output strains were cultured overnight in Brucella broth and then cocultured with AGS cells for 4 h at MOI 10. RNA was extracted using the RNeasy mini kit (Quiagen), followed by RQ1 DNase (Promega) treatment. cDNA was prepared using the High-Capacity cDNA reverse transcription kit (Applied Biosystems) and PCR was performed using the LightCycler 480 SYBR Green I Master Mix (Roche) and 250 nM primers (SI Appendix, Table S3). Samples were analyzed in duplicates using a LightCycler 480 II Instrument (Roche). Relative mRNA expression (fold change) was calculated using the comparative cycle threshold (ΔΔCt) analysis method. GAPDH and 16S rDNA were used as housekeeping genes for eukaryotes and prokaryotes, respectively.

CXCL8 ELISA.

H. pylori strain 7.13 and gerbil output strains were cultured overnight in Brucella broth and then cocultured with AGS cells for 24 h at MOI 10. Supernatants were recovered from the cocultures and quantification was performed using the CXCL8/IL-8 DuoSet ELISA (R&D Systems). Briefly, 96-well plates were coated with an anti-CXCL8 capture antibody overnight and then blocked with 1% BSA. The samples or standards were added and incubated for 2 h, followed by incubations with the detection antibody and Streptavidin-HRP. The substrate solution (R&D Systems) was incubated for 20 min and then stopped with 1 M H₂SO₄ (R&D Systems). Optical densities were determined using a Biotek Synergy 4 microplate reader. Samples were normalized to the level of CXCL8 in AGS cells infected with H. pylori strain 7.13 in each experiment.

Western Blot for CagY.

Protein lysates from the different H. pylori strains were separated by SDS/PAGE, and proteins were transferred to PVDF membranes. Membranes were then incubated with a rabbit anti-CagY antiserum (1/10,000) (9). After incubation with the secondary antibody, detection was performed as described for the analysis of CagA translocation.

DFMO Quantification.

H. pylori was grown for 48 h on blood agar plates containing DFMO or not. Bacteria were harvested and washed three times with PBS. Bacterial pellets were extracted using acetonitrile/20 mM NH₄OAc, pH 8 (70:30). Extracts were derivatized by reaction with 20 mM dansyl chloride in 100 mM NaHCO₃ pH 10 for 20 min. DFMO levels were quantified by LC-MS using a Thermo TSQ Vantage Triple Quadrupole instrument operated in positive ion mode. Sample concentration was calculated using a DFMO standard curve.

Analysis of Spontaneous Mutations.

H. pylori liquid cultures were plated in TSA blood agar plates supplemented or not with 20 μg/mL rifampicin. After 10 d, the number of cfu was determined and the frequency of mutation was calculated (50).

8-Oxoguanine Staining.

H. pylori 7.13 was cultured overnight in Brucella broth with 10% FBS, resuspended at an OD₆₀₀ of 0.3, and treated or not with 1% DFMO or 1% ornithine for 24 h. H. pylori was also incubated without CO₂ supplementation and exposed to atmospheric oxygen for 4 h as a positive control for induction of oxidative DNA damage. Bacteria were fixed in Cytofix/Cytoperm buffer (BD Biosciences) and incubated for 30 min at 4 °C, followed by three washes with Perm/Wash buffer (BD Biosciences). Alexa 488-conjugated avidin (4 μg/mL) was added to the bacterial suspensions for 30 min at 4 °C, followed by two washes with Perm/Wash buffer. For confocal microscopy imaging, pellets were resuspended in 300 μL of PBS, and 20 μL of the suspensions were air-dried onto glass slides and mounted with Vectashield Hard Set with DAPI (Vector). Images were acquired using an Olympus FV-1000 inverted confocal microscope. For flow cytometry analysis, bacterial suspensions were incubated with 0.25 μg of 7-AAD (BD Biosciences) to stain the DNA for 10 min before analysis in a Guava easyCyte flow cytometer (Millipore).

We thank Lori M. Hansen (University of California, Davis) for additional help with cagY sequencing and Dr. Rainer Haas (Max von Pettenkofer Institute, Faculty of Medicine, Ludwig Maximilians University, Munich) for providing the CagY antiserum. Immunofluorescence confocal imaging was performed in the Vanderbilt Cell Imaging Shared Resource, supported by the Vanderbilt Digestive Disease Research Center and NIH Grant P30DK058404. SMRT sequencing for cagY was performed at the University of Florida NextGen DNA Sequencing Core. This work was funded by NIH Grants R01AT004821 (to K.T.W.), R01CA190612 (to K.T.W. and D.R.M.), P01CA116087 (to K.T.W. and R.M.P.), and P01CA028842 (to K.T.W.); Veterans Affairs Merit Review Grant I01BX001453 (to K.T.W.); the Thomas F. Frist Sr. Endowment (to K.T.W.); and the Vanderbilt Center for Mucosal Inflammation and Cancer (K.T.W.). This work was also supported by NIH Grant R01AT006896 (to C.S.). P.B.L. was supported by Postdoctoral Fellowship Award 16POST27250138 from the American Heart Association. Mass spectrometry analyses were supported in part by Core Scholarships from the Vanderbilt University Medical Center Digestive Disease Research Center funded by NIH Grant P30DK058404, and the Vanderbilt-Ingram Cancer Center Support Grant P30CA068485.