A gut pathobiont synergizes with the microbiota to instigate inflammatory disease marked by immunoreactivity against other symbionts but not itself

Inflammatory bowel diseases (IBD) are likely driven by aberrant immune responses directed against the resident microbiota. Although IBD is commonly associated with a dysbiotic microbiota enriched in putative pathobionts, the etiological agents of IBD remain unknown. Using a pathobiont-induced intestinal inflammation model and a defined bacterial community, we provide new insights into the immune-microbiota interactions during disease. In this model system, the pathobiont Helicobacter bilis instigates disease following sub-pathological dextran sulfate sodium treatment. We show that H. bilis causes mild inflammation in mono-associated mice, but severe disease in the presence of a microbiota, demonstrating synergy between the pathobiont and microbiota in exacerbating pathology. Remarkably, inflammation depends on the presence of H. bilis, but is marked by a predominant Th17 response against specific members of the microbiota and not the pathobiont, even upon the removal of the most immune-dominant taxa. Neither increases in pathobiont burden nor unique changes in immune-targeted microbiota member abundances are observed during disease. Collectively, our findings demonstrate that a pathobiont instigates inflammation without being the primary target of a Th17 response or by altering the microbiota community structure. Moreover, our findings point toward monitoring pathobiont-induced changes in microbiota immune targeting as a new concept in IBD diagnotics.

concentration of 200 µg/mL. Cells were incubated overnight at 37 o C with 5% CO 2 . Prior to incubation with purified CD4 + T cells, all feeder cells were treated with 50 µg/mL filteredsterilized mitomycin C solution (Sigma) as described 4 and incubated at 37 o C for 20 min with 5% CO 2 . Thereafter, feeder cells were washed 5X with an excess volume of CTCM. Feeder cell suspensions were adjusted to a final concentration of 2 million live cells/mL prior to plating 0.2 million feeder cells in 100 µL CTCM per well of a 96-well U-bottom plate with purified CD4 + T cells.
Evaluation of EM CD4 + T cells secreting IL17A + and/or IFN-γ + included staining anti-CD62L and anti-CD44 as described above.
Cells were analyzed using a BD FACSCanto II (BD Biosciences, San Jose, CA). Data were analyzed using FlowJo v10.2 (FlowJo™, LLC, Ashland, OR). During data analysis, gates were first drawn on singlet events (FSC-H vs FSC-A) and then on lymphocytes (FSC-A vs SSC-A). Populations of interest were subsequently evaluated; gates were drawn based on fluorescence minus one controls. Representative graphs of depicting all gating strategies can be seen in Fig. S11A-E and S12A-F. All samples were kept at -20 o C prior to bacterial quantification. The pre-sorted samples were also centrifuged at 10,000 x g for 10 min at room temperature prior to adding 400 µL of cold 1X PBS before DNA extraction and bacterial quantification. Thereafter, samples were homogenized by bead beating for 2 min (Minibeadbeater; Biospec), 2 min on ice, and then another 2 min of bead beating; samples were placed on ice prior to centrifugation at 6,000 x g for 5 min at 4 o C. The top layer was then transferred to a 2 mL sterile snap-cap tube (Thermo Fisher Scientific Inc.) at a ratio of 1:1 with PCI (300 µL of each) prior to homogenization. Samples were centrifuged at 16,100 x g for 3 min at room temperature and the top layer was transferred into 2 mL sterile snap-cap tubes. Two volumes of 100% ethanol (Thermo Fisher Scientific Inc.) were added prior to storing samples at -80°C for 2 hours.
Samples were then centrifuged at 21,130 x g for 20 min and the supernatant discarded without disturbing the DNA pellet. Pellets were washed by adding 500 µL 70% ethanol and centrifuging at 21,130 x g for 20 min before drying the tube for 30 min at room temperature. Lastly, DNA was eluted in 200µL of Tris-EDTA Buffer (Thermo Fisher Scientific Inc.) and stored at -20°C until quantification and qPCR analysis.
ASF quantification of DNA samples from pre-sorted, IgA+ or IgA-fractions was performed as previously described 1 . H. bilis was quantified as described above (see H. bilis qPCR assay) using the following predictive linear equation for calculations: Y = -2.9125X + 34.313, where Y = mean cycle threshold value and X = log 10 total bacterial abundance, R 2 = 0.995 and efficiency = 1.20, with the limit of detection determined to be 97 bacteria per sample.
All qPCR reactions were run using the Maxima SYBR Green qPCR Master Mix 2X (Thermo Fisher Scientific Inc.) as previously described 1 . Prior to qPCR, DNA was quantified using the Quant-iT™ PicoGreen® dsDNA Broad Range and High Sensitivity Reagents (Thermo Fisher Scientific Inc.) as previously described 1 . For pre-sorted samples, 10 ng of DNA template (1 µL/sample) was used per reaction; 10 µL of IgA+ and IgA -fractions were used per reaction. A total reaction volume of 25 µL was used for all samples, which were tested in duplicate.
Final calculations for the relative abundance of each ASF member and H. bilis in all samples were made on an individual animal basis. The log 10 total number of bacteria was initially calculated using linear equations followed by normalization to 100 ng of DNA template.
Thereafter, the mean relative abundance (% of each taxon) was estimated. The ratio between the mean % of each taxon for the IgA positive and IgA negative fractions for a given animal (i.e., IgA index) was then used for statistical analysis to make comparisons between DSS treated versus control treatments. A ratio of 1 indicated no difference in relative abundance of a given bacterium between the IgA positive fraction and the corresponding IgA negative fraction.
Statistical analysis. Summary statistics were calculated for all treatments to assess the overall quality of the data set, including normality. Outliers were statistically identified using the ROUT test (Q = 1%); however, they were only removed if the value was biologically implausible given the methodology used for its measurement. An unpaired non-parametric Mann-Whitney Test was used to analyze targeted pairwise differences within the following datasets: gross and histopathological cecal scores, H. bilis and individual ASF abundances, proportion of EM CD4 + T cells, and chemokine and cytokine concentrations. A parametric, unpaired, two-tailed T-test was used for comparisons made for cellularity of effector (IL-17A + and/or IFN-γ + ), regulatory CD4 + T cells in addition to active B cells, and for comparing the IgA index across DSS treated vs non-treated groups in the assembly experiment. A non-parametric Kruskal-Wallis one-way ANOVA followed by post-hoc pairwise comparison with Dunn's test was used to compare gross cecal scores and H. bilis abundance across DSS treated groups, and for individual ASF abundances in cecal contents and tissues. All those statistical analyses were performed using GraphPad Prism 7 (version 7.0a, 2016, GraphPad Software, Inc., La Jolla, CA) using the significance cut-off of P < 0.05.
The following statistical analyses were all conducted using R software, version 3. between and within groups for the ASF abundance. This analysis generated an R statistic ranging from -1 to 1. The closer the value to 1, the greater the dissimilarity between groups than within groups. Values below 0 suggested that dissimilarities were greater within rather than between groups, indicating that treatment was not influencing the outcome. The BCDC value was calculated using the vegdist() function. The ANOSIM was performed using the anosim() function where BCDC was the outcome of interest and treatment groups were the explanatory variables. iii) PERMANOVA analysis was performed using the adonis() function to identify the contribution of explanatory variables to the overall variability in the dataset. A variable would have a strong contribution depending on the interpretation of two values: R 2 and calculated Pvalue. All these analyses were completed using the Vegan package. The significance cut-off used across all analyses was P < 0.05. A Spearman's rank correlation analysis was performed to determine if there were correlations between the pathobiont and any other ASF member abundance in diseased versus non-diseased animals for both cecal contents and tissues. An arbitrary cut-off value of +0.7 was used to define a strong positive correlation between two bacterial species. Figure S1. The Helicobacter pathobiont synergized with the resident microbiota (ASF) to promote specific histopathological alterations and a pro-inflammatory response during disease. (A) Histopathological parameters in cecal tissues from germ-free (GF), ASF-bearing, H. bilis mono-associated and H. bilis colonized ASF-bearing mice treated with 1.5% DSS or left untreated (n = 9-15 animals per treatment). (B-E) Pro-inflammatory cytokines in cecal explants from GF, ASF-bearing, H. bilis mono-associated and H. bilis colonized ASF bearing mice treated with 1.5% DSS or left untreated (n = 9-12 animals per treatment). (B-E) Horizontal bars represent group means in all graphs. (A-E) Asterisks depict the degree of significance for differences as determined by a non-parametric unpaired Mann-Whitney test (i.e., cytokines) or using unpaired parametric T-test (i.e., total number of Th17 cells) with a two-tailed distribution for P-value calculations (* P < 0.05, ** P ≤ 0.01, *** P ≤ 0.001 and **** P ≤ 0.0001). Only significant differences between treatments are presented in the graphs. Experiments were performed using male and female C3H/HeN mice at 8-10 wks of age; mice were colonized with H. bilis for 3 wks. Figure S2. The absence of an adaptive immune system limited development of a proinflammatory response during pathobiont-induced disease. (A-H) Pro-inflammatory cytokines and chemokines in cecal explants from ASF-bearing C57BL/6 wild type (WT) and Rag1 -/mice harboring the ASF and colonized with or without H. bilis and either treated with 2% DSS or left untreated (n = 8-18 animals per treatment). Horizontal bars represent group means in all graphs. Asterisks depict the degree of significance for differences as determined by a non-parametric unpaired Mann-Whitney test using a two-tailed distribution for P-value calculations (* P < 0.05, ** P ≤ 0.01, *** P ≤ 0.001 and **** P ≤ 0.0001 and NS = not significant P ≥ 0.05). Experiments were performed using male and female C57BL/6 (WT or Rag1 -/-) mice at 8-10 wks of age; mice were colonized with H. bilis for 3 wks. Figure S3. Effector memory Th17, active B cells and iTregs, but not nTregs or IFN-γ + CD4 + T cells, expanded during pathobiont-induced intestinal inflammation. (A-F) Total number of effector memory (EM) CD44 high CD62L low IL-17A + CD4 + T cells (A), active CD19 + CD23 -B cells (B), iTreg cells (CD25 + Foxp3 + Neuropilin low Helios low CD4 + ) (C), total number of IFN-γ + CD4 + T cells (D), EM IFN-γ + CD4 + T cells (E), total IFN-γ + IL-17A + CD4 + T cells (F), EM IFN-γ + IL-17A + CD4 + T cells (G), total CD25 + Foxp3 + CD4 + Tregs (H), nTregs (CD25 + Foxp3 + Neuropilin high Helios high CD4 + ) (I) and CD8b + T cells (J) in mesenteric lymph nodes (MLN; n = 9-18 animals per treatment). (A-J) Horizontal bars represent treatment means in all graphs. Differences in the absolute immune cell counts were tested using an unpaired parametric T-test using a two-tailed distribution for P-value calculations. Asterisks depict the degree of significance for the difference in the total number of immune cells (* P < 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001 and NS = not significant P ≥ 0.05). Experiments were performed using male and female C57BL/6 mice at 8-10 wks of age; mice were colonized with H. bilis for 3 wks. . (A-E) Asterisks depict the degree of significance determined by a non-parametric unpaired Mann-Whitney test using a two-tailed distribution for P-value calculations (* P < 0.05, ** P ≤ 0.01, *** P ≤ 0.001 and **** P ≤ 0.0001 and NS = not significant P ≥ 0.05). (F) Differing superscript letters indicate significant differences across treatments as per a nonparametric Kruskal-Wallis one-way ANOVA, followed by a post-hoc test (Dunn's test, P < 0.05). ASF bacterial abundances were measured using species-specific qPCR assays. All experiments were performed using male and female C3H/HeN mice. Asterisks depict the degree of significance for differences as determined by a non-parametric unpaired Mann-Whitney test using a two-tailed distribution for P-value calculations (* P < 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001 and NS = not significant P ≥ 0.05). Experiments were performed using male and female C3H/HeN mice at 8-10 wks of age; mice were colonized with H. bilis for 3 wks. . Asterisks depict the degree of significance for differences as determined by a non-parametric unpaired Mann-Whitney test using a two-tailed distribution for P-value calculations (* P < 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001 and NS = not significant P ≥ 0.05). Experiments were performed using male and female C3H/HeN mice at 8-10 wks of age; mice were colonized with H. bilis for 3 wks. . Asterisks depict the degree of significance for differences as determined by a non-parametric unpaired Mann-Whitney test using a two-tailed distribution for P-value calculations (* P < 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001 and NS = not significant P ≥ 0.05). Experiments were performed using male and female C3H/HeN mice at 8-10 wks of age; mice were colonized with H. bilis for 3 wks. . Asterisks depict the degree of significance for differences as determined by a non-parametric unpaired Mann-Whitney test using a two-tailed distribution for P-value calculations (* P < 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001 and NS = not significant P ≥ 0.05). Experiments were performed using male and female C3H/HeN mice at 8-10 wks of age; mice were colonized with H. bilis for 3 wks. Figure S9. Removal of immune-dominant taxa from the resident microbiota does not alter the severity of pathobiont-mediated intestinal inflammation or eliminate Th17 immunoreactivity against the remaining gut symbionts. IL-17A (A-D) and IL-17F (E-M) secretion from CD4 + T cells isolated from the mesenteric lymph nodes of mice in all treatments except for animals colonized with the ASF minus 457. CD4 + T cells were either left unstimulated (NS) or stimulated with individual whole-cell sonicate ASF antigens for 72 hrs (n = 2-24 pools of 2-3 animals per pool per treatment). (A-M) Asterisks depict the degree of significance for differences as determined by a non-parametric unpaired Mann-Whitney test using a two-tailed distribution for P-value calculations (* P < 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001 and NS = not significant P ≥ 0.05). Differing superscript letters indicate significant differences across treatments as per a non-parametric Kruskal-Wallis one-way ANOVA, followed by a posthoc test (Dunn's test, P < 0.05). Experiments were initiated when male and female C3H/HeN mice were 4-5 wks of age. Figure S10. Absence of enhanced IgA coating of intestinal bacteria during pathobiont-mediated intestinal inflammation. (A) FACS gating strategy to sort the bacterial IgA+ fraction from cecal contents for subsequent quantification of the ASF and H. bilis relative abundances by qPCR. FACS was performed following pre-enrichment of the IgA+ fraction using magnetic cell sorting. (B-K) Intestinal inflammation triggered by H. bilis and DSS treatment did not significantly affect the ratio of bacteria detected in the IgA+/-fractions (n = 5-13 individual cecal samples per treatment). A ratio of 1 between IgA+ and -fractions indicates equal bacterial relative abundances (represented by the horizontal dotted line in Fig. S10B-J). (B-J) Asterisks depict the degree of significance for differences as determined by a parametric unpaired T-test using a twotailed distribution for P-value calculations (* P < 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001 and NS = not significant P ≥ 0.05). Experiments were initiated when male and female C3H/HeN mice were 4-5 wks of age. Experimental design is described in Fig. 5A. Gating on total Tregs (CD25 + Foxp3 + CD4 + ), nTregs (CD25 + Foxp3 + Neuropilin-1 high Helios high CD4 + ) and iTregs (CD25 + Foxp3 + Neuropilin-1 low Helios low CD4 + ). Figure S12. Flow cytometry gating strategy for quantification of Th1 and Th17 cells in the mesenteric lymph nodes during pathobiont-elicited intestinal inflammation. (A-B) Gating on singlets and lymphocytes, respectively. (C-E) Gating on total CD4 + T cells, Th17 (IL-17A + CD4 + ) and Th1 cells (IFN-γ + CD4 + ). (F) Gating on effector memory Th17 (CD62L low CD44 high IL-17A + CD4 + ) and Th1 cells (CD62L low CD44 high IFN-γ + CD4 + ) .