Gut epithelial Interleukin-17 receptor A signaling can modulate distant tumors growth through microbial regulation

Abstract

Microbes influence cancer initiation, progression and therapy responsiveness. IL-17 signaling contributes to gut barrier immunity by regulating microbes but also drives tumor growth. A knowledge gap remains regarding the influence of enteric IL-17-IL-17RA signaling and their microbial regulation on the behavior of distant tumors. We demonstrate that gut dysbiosis induced by systemic or gut epithelial deletion of IL-17RA induces growth of pancreatic and brain tumors due to excessive development of Th17, primary source of IL-17 in human and mouse pancreatic ductal adenocarcinoma, as well as B cells that circulate to distant tumors. Microbial dependent IL-17 signaling increases DUOX2 signaling in tumor cells. Inefficacy of pharmacological inhibition of IL-17RA is overcome with targeted microbial ablation that blocks the compensatory loop. These findings demonstrate the complexities of IL-17-IL-17RA signaling in different compartments and the relevance for accounting for its homeostatic host defense function during cancer therapy.

Keywords: B cells; IL-17; intestinal homeostasis; microbiome; pancreatic cancer.

Figure 1:. Microbially driven elevation of IL-17 promotes pancreatic tumor growth

A.Experimental outline of murine PDAC model. Pancreata (orthotopic) or skin (subcutaneous) of Il17ra^+/+ and Il17ra^−/− mice were implanted with KPC cells and sacrificed after 4–5 weeks. B. Representative images (duplicates) of PDAC orthotopic tumors from Il17ra^+/+ and Il17ra^−/− mice at endpoint. Scale bar represents 3 mm. C-D. Orthotopic tumor volumes (C) and weight (D) at endpoint. E-F. Subcutaneous tumor growth curves (E) and end-point tumor weight (F). G. Volcano plot showing upregulated genes from RNA-seq analysis of orthotopic tumors in Il17ra^−/− mice compared to Il17ra^+/+ mice (n=4/group). H. z-score of selected IL-17 related pathways in orthotopic tumors and ilea of Il17ra^−/− mice, relative to Il17ra^+/+ mice from Ingenuity Pathway Analysis (IPA). I. Relative mRNA expression measured by qPCR in orthotopic tumors (left) and ilea (right) at endpoint. J. Representative images showing Il17a RNAScope staining in orthotopic tumors (left) and ileum (right). Scale bars represent 40 μm. K.Heatmap showing foldchange (FC) of plasma cytokine levels in tumor-bearing Il17ra^−/− mice compared to Il17ra^+/+ mice at endpoint, measured by Luminex (n=3/group). L. Pre and post treatment plasma IL-17 levels in mice treated with either single or cocktail of antibiotics (ampicillin, vancomycin, neomycin, metronidazole and amphotericin B) in drinking water for two weeks. M. Experimental outline of IL-17RA blockade in combination with antibiotics in PDAC subcutaneous tumor model. Starting 10 days after tumor implantation, mice were given four doses of isotype IgG or αIL-17RA monoclonal antibody with or without ampicillin treatment in drinking water until endpoint. N. Plasma IL-17 levels at endpoint. O. Subcutaneous tumor growth curves IL-17RA blockade in combination with ampicillin. Data is reported as mean ± SEM; P values were calculated using Student’s t test (unpaired, two tailed) for 2 groups and ANOVA for ≥2 groups; *P < 0.05, **P < 0.01, ***P < 0.001, or ****P < 0.0001. See also Figure S1.

Figure 2:. Enteric IL-17RA genetic deletion promotes microbially driven IL-17 signaling dependent pancreatic tumor growth.

A. Baseline plasma IL-17 levels in Il17ra^+/+ and Il17ra^fl/fl;Villin-Cre mice. B. Experimental outline of murine PDAC orthotopic model. Mice were treated with ampicillin in drinking water starting from one week prior to tumor implantation. Pancreata of Il17ra^+/+ and Il17ra^fl/fl;Villin-Cre mice were implanted with KPC^Il17ra+/+ and KPC^{Il17ra−/−} cells and sacrificed after 4 weeks. C. Representative images (duplicates) of PDAC orthotopic tumors at endpoint. Scale bar represents 3 mm. D-E. Orthotopic tumor volumes (D) and weight (E) at endpoint. F. Plasma IL-17 levels in tumor-bearing Il17ra^fl/fl;Villin-Cre mice at endpoint with or without ampicillin treatment from 2B (n≥4/group). G. Experimental outline of PDAC subcutaneous tumor model implanted with KPC^Act1+/+ and KPC^Act1−/− cells. H-I. Subcutaneous tumor growth curves (H) and endpoint tumor weight (I) of KPC^Act1+/+ and KPC^Act1−/− tumors in Il17ra^+/+ and Il17ra^fl/fl;Villin-Cre mice. J. z-score of the IL-17 signaling pathway in orthotopic tumors and ilea (gut) of Il17ra^−/− and Il17ra^fl/fl;Villin-Cre mice, relative to Il17ra^+/+ mice calculated from IPA. K. Heatmap showing log2FC of overlapping genes in tumors and ilea (gut) of Il17ra^−/− and Il17ra^fl/fl;Villin-Cre mice, relative to Il17ra^+/+ mice from RNA-seq (n=4/group). L. Baseline plasma IL-17F levels in tumor-bearing Il17ra^+/+ and Il17ra^fl/fl;Villin-Cre mice at endpoint measured by Luminex (n ≥3/group). M. Bray-Curtis PCoA plot of beta diversity showing distance between stool samples of Il17ra^+/+ and Il17ra^fl/fl;Villin-Cre mice. Data is reported as mean ± SEM; P values were calculated using Student’s t test (unpaired, two tailed) for 2 groups and ANOVA for ≥2 groups; *P < 0.05, **P < 0.01, ***P < 0.001, or ****P < 0.0001. See also Figures S2, S3 and S4.

Figure 3:. Intestinal IL-17-IL-17RA dependent microbial dysbiosis can affect tumor growth remotely

A. Experimental outline of murine GBM orthotopic model. Guided screws were intra-cranially implanted one week prior to tumor implantation. Brains of Il17ra^+/+, Il17ra^−/− and Il17ra^fl/fl;Villin-Cre mice were implanted with murine GBM cells Gl261 and sacrificed after three weeks. B. Representative bioluminescent images showing cranial luciferin expression in Il17ra^+/+, Il17ra^−/− and Il17ra^fl/fl;Villin-Cre mice at day 1 (left), day 11 (middle) and day 22 (right) post implantation. Luminescence is represented as a heatmap in radians. C. Tumor growth curves measured by bioluminescent imaging. D. Representative images (triplicates) of GBM orthotopic tumors from Il17ra^+/+, Il17ra^−/− and Il17ra^fl/fl;Villin-Cre mice at endpoint. Yellow arrows indicate site of tumor injection. Scale bar represents 3 mm. E. Tumor volumes at endpoint measured by bioluminescent imaging. F-G. Neurological symptoms- reflex score (F) and presence of seizures (G) in Il17ra^+/+, Il17ra^−/− and Il17ra^fl/fl;Villin-Cre mice at endpoint. H. Heatmap showing log2FC of genes in GBM orthotopic tumors and ilea of Il17ra^−/− and Il17ra^fl/fl;Villin-Cre mice, relative to tumors and ilea of Il17ra^+/+ mice assessed from RNA-seq (n=4/group). Data is reported as mean ± SEM; P values were calculated using Student’s t test (unpaired, two tailed) for 2 groups and ANOVA for ≥2 groups; *P < 0.05, **P < 0.01, ***P < 0.001, or ****P < 0.0001. See also Figure S4.

Figure 4:. Enteric IL-17RA deficiency triggers B cells development

A. Experimental outline of murine PDAC orthotopic model. Pancreata of Il17ra^+/+ and Il17ra^fl/fl;Villin-Cre mice were implanted with KPC cells sacrificed after 3 weeks for immunoprofiling of tumor, lamina propria (LP), Peyer’s patches (PP), mesenteric lymph nodes (mLN), pancreatic lymph nodes (pLN) and spleen; via CyTOF and flow cytometry. B-C. viSNE plot (B) and lymphocytic populations (C) in LP from Il17ra^+/+ and Il17ra^fl/fl;Villin-Cre mice (n=3/group). D-E. viSNE plot (D) and lymphocytic populations (E) in tumors from Il17ra^+/+ and Il17ra^fl/fl;Villin-Cre mice (n=3/group). F. Representative flow cytometric plot showing immune populations in orthotopic tumors of Il17ra^+/+ and Il17ra^fl/fl;Villin-Cre mice. G. Quantification of B cells abundance in LP, mLN, spleen and tumors by flow cytometric analysis, reported as a percent of viable CD45⁺ cells. H. Representative images showing CD20 IHC staining in orthotopic KPC tumors from Il17ra^+/+, Il17ra^−/− and Il17ra^fl/fl;Villin-Cre mice. Scale bar represents 50 μm. I. Experimental outline of in vivo B cell depletion in PDAC subcutaneous tumor model using neutralizing antibodies against αCD20. J. Levels of B cells in tumors reported as a percent of viable CD45⁺ cells, measured by flow cytometry. K. Growth curves of subcutaneous tumors with B cell depletion in Il17ra^+/+ and Il17ra^fl/fl;Villin-Cre mice. Data is reported as mean ± SEM; P values were calculated using Student’s t test (unpaired, two tailed) for 2 groups and ANOVA for ≥2 groups; *P < 0.05, **P < 0.01, ***P < 0.001, or ****P < 0.0001. See also Figure S5 and Table S1.

Figure 5:. Enteric IL-17RA deficiency driven dysbiosis rewires immune and tumor cells

A. Experimental outline of murine PDAC orthotopic model. Pancreata of Il17ra^+/+ and Il17ra^fl/fl;Villin-Cre mice were implanted with KPC cells sacrificed after 3 weeks for single cell RNA sequencing (scRNA-seq) of matched tumor and lamina propria (gut) samples. B. UMAP plot showing cell clusters in ilea (gut) samples of tumor bearing Il17ra^fl/fl;Villin-Cre and Il17ra^+/+ mice (n=2/group). C. UMAP plot showing cell clusters in matched tumors samples of Il17ra^fl/fl;Villin-Cre and Il17ra^+/+ mice (n=3/group). D. Dot plots showing Il17a (left) and Il17f (right) average expression in Th17 and B cells in gut (top) and tumor (bottom) samples from Il17ra^fl/fl;Villin-Cre mice assessed from single cell analysis. E. RNAScope staining showing Il17a and Il17f expression in pancreatic orthotopic tumors from Il17ra^+/+ and Il17ra^fl/fl;Villin-Cre mice. Scale bar represents 40 μm. F. RNAScope staining showing Il17f and CD20+ cells in pancreatic orthotopic tumors from Il17ra^fl/fl;Villin-Cre mice. Scale bar represents 20 μm. G. Enriched pathway in tumor B cells cluster from GSEA analysis in Il17ra^fl/fl;Villin-Cre vs. Il17ra^+/+ mice. H. Venn diagram showing overlapping genes in IL-17 driven pancreatic datasets (Table S2): a) upregulated genes in ‘Tumor cells1’ cluster from single cell analysis (scRNA-seq) from advanced pancreatic tumors of Il17ra^fl/fl;Villin-Cre, b) downregulated genes in PanINs (premalignant pancreas) of KC^iMist mice receiving IL-17 blockade and c) downregulated genes in advanced pancreatic tumors of WT mice receiving IL-17 blockade. I. Violin plot showing Duox2 expression in sub clusters of ‘Tumor cells1’ divided by Duox2 expression in Il17ra^+/+ and Il17ra^fl/fl;Villin-Cre mice. J. Enriched pathways from GSEA analysis in ‘Tumor cells1’ sub cluster 1 (Duox2 High) in Il17ra^fl/fl;Villin-Cre vs Il17ra^+/+ mice. K. Experimental outline of murine pancreatic orthotopic model. Pancreata of whole body tdTomato+ mice were implanted with KPC cells and treated with αIL-17+αIL-17RA twice a week starting at day 10. At sacrifice, tdTomato negative (tumor cells) were enriched by sorting. L. Relative mRNA expression of Duox2 and Duoxa2 measured by qPCR in tumors. M. Quantification of H₂O₂ released over time by KPC^Il17ra+/+ and KPC^{Il17ra−/−} cells treated with IL-17 in vitro, measured by Amplex Red assay. Experiment was repeated at least thrice. Data is reported as mean ± SEM; P values were calculated using Student’s t test (unpaired, two tailed) for 2 groups and ANOVA for ≥2 groups; *P < 0.05, **P < 0.01, or ****P < 0.0001. See also Figures S6, S7 and Table S2.

Figure 6:. Human PDAC has circulating and tumor infiltrating IL-17-expressing cells that associate with DUOX2

A. Experimental outline of spatial single cell transcriptomic profiling of clinical PDAC primary tumors by CosMx Spatial Molecular Profiling (SMI). B. Representative image showing IL17A transcripts (white) across cell types in selected ROIs. Scale bar represents 125 μm (left) and 25 μm (right, insets). C. Bubble plot showing average normalized counts of IL17A expression in each cell type cluster from CosMx SMI analysis. D. DUOX2 expression in TCGA PDAC patients stratified according to average IL17A expression. E-F. Pathway analysis from GSEA analysis showing enriched pathways (E) and cnetplot (F) in TCGA PDAC patients stratified according to high vs. low DUOX2 expression. Data is reported as mean ± SEM; P values were calculated using Student’s t test (unpaired, two tailed); ****P < 0.0001. See also Figure S7.