Store‐operated calcium entry controls innate and adaptive immune cell function in inflammatory bowel disease

Abstract Inflammatory bowel disease (IBD) is characterized by dysregulated intestinal immune responses. Using mass cytometry (CyTOF) to analyze the immune cell composition in the lamina propria (LP) of patients with ulcerative colitis (UC) and Crohn's disease (CD), we observed an enrichment of CD4+ effector T cells producing IL‐17A and TNF, CD8+ T cells producing IFNγ, T regulatory (Treg) cells, and innate lymphoid cells (ILC). The function of these immune cells is regulated by store‐operated Ca2+ entry (SOCE), which results from the opening of Ca2+ release‐activated Ca2+ (CRAC) channels formed by ORAI and STIM proteins. We observed that the pharmacologic inhibition of SOCE attenuated the production of proinflammatory cytokines including IL‐2, IL‐4, IL‐6, IL‐17A, TNF, and IFNγ by human colonic T cells and ILCs, reduced the production of IL‐6 by B cells and the production of IFNγ by myeloid cells, but had no effect on the viability, differentiation, and function of intestinal epithelial cells. T cell‐specific deletion of CRAC channel genes in mice showed that Orai1, Stim1, and Stim2‐deficient T cells have quantitatively distinct defects in SOCE, which correlate with gradually more pronounced impairment of cytokine production by Th1 and Th17 cells and the severity of IBD. Moreover, the pharmacologic inhibition of SOCE with a selective CRAC channel inhibitor attenuated IBD severity and colitogenic T cell function in mice. Our data indicate that SOCE inhibition may be a suitable new approach for the treatment of IBD.

. Expression levels of Stim1, Stim2, Orai1, Orai2 and Orai3 mRNA in different T helper cell subsets. Naïve CD4 + T cells were isolated from spleen and LNs of WT mice and differentiated into Th1, Th17 and iTreg cells for 3 days. mRNA was isolated and analyzed by quantitative real-time PCR. Bar graphs show mean ± SEM from three mice. Statistical analysis by two-way ANOVA. No statistically significant differences were detected.

Appendix
S2B. LPMCs were isolated from 3 IBD patients and cellular viability in the presence or absence of PMA/Ionomycin (4 h n tro) or additional BTP2 was assessed by quantifying frequencies of early apoptotic (Annexin V + 7AAD -) and late apoptotic cells (Annexin V + 7AAD + ). Statistics were calculated using a paired Wilcoxon matched-pairs signed rank test, *p < 0.05. Error bars represent the standard error of mean (SEM).   Figure S2A. Ca 2+ influx analyses were performed by flow cytometry to assess Ca 2+ influx in PBMCs isolated from healthy donors (n = 3, in duplicate), with or without 1 μM BTP2 treatment.
Statistic was calculated with a paired t test corrected by using the two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli, with Q = 1%. *p-value < 0.05, **p-value < 0.01, ***p-value < 0.001, ****p-value < 0.0001. Bars denote the mean values of influx rate and error bars represent the standard error of mean (SEM). Appendix Figure S3. (A) Gating strategy preceding analyses using the t-distributed stochastic linear embedding (t-SNE) algorithm and FlowSOM/ConsensusClusterPlus self-organizing map. Cells were cleaned from calibration beads and doublets and de-barcoded according to the Cell-ID 20-plex Pd Barcoding Kit and 89Y-CD45 staining. (B) FlowSOM plot of merged FCS files from samples treated with PMA/ionomycin ± 1µM BTP2 (CD: n =5). Colors indicate 9 defined clusters among CD45 + CD3 + LPMCs. Heatmap clusters show the expression levels of the 21 markers used for cluster analysis. viSNE plots display representative LPMCs isolated from one CD patient colored by marker expression levels (blue: low, red: high).  Figure S4. Effects of CRAC channel inhibitor BTP-2 on the function of Th1, Th17 and iTreg cells. (A) Analysis of SOCE in CD4 + T cells cultured in vitro for 3 days under Th0 condition; cells were then treated acutely with BTP-2 for 30 min before calcium measurement. SOCE (peak F340/380) was quantified in (B); shown are means±SEM of cells from two mice measured in triplicates. (C) Analysis of IFNγ, IL-17A and CTLA-4 expression in Th1, Th17 and iTreg cells with CRAC inhibitor BTP-2 treatment. Naïve T cells were cultured under Th1 and Th17 condition for 3 days in the presence of gradient BTP-2, cells were restimulated with PMA/Ionomycin for 4 hours to detect cytokine production by flow cytometry. CLTA4 in iTreg cells was detected by intracellular staining without PMA/Ionomycin stimulation. (D) Quantification of normalized MFI of IFNγ, IL-17A and CTLA-4 by Th1, Th17, iTreg respectively. (E+F) Quantification of normalized MFI of Tbet, RORγt and Foxp3 levels in Th1, Th17 and iTreg in the presence of gradient BTP-2 treatment. Statistical analysis by unpaired student's t-test was performed either between Th1 and iTreg or Th17 and iTreg: ***p<0.001 **p<0.01 *p<0.05. Statistical analysis by unpaired student's t-test: ***p<0.001 **p<0.01 *p<0.05. Data in C-F are the mean ± SEM of 3 mice. Appendix Figure S5. FlowSOM plot of merged FCS files from samples treated with PMA/ ionomycin ± 1µM BTP2 (CD: n =5). Colors indicate 12 defined clusters among CD45 + CD3 -LPMCs. Heatmap clusters show the expression levels of the 21 markers used for cluster analysis. viSNE plots represent LPMCs isolated from one CD patient colored by marker expression levels (blue: low, red: high).  Appendix Table S3. Markers used for viSNE clustering of CD45 + CD3 + and CD45 + CD3 -LPMCs described in Figure 2D-G. Appendix Table S4. Markers used for viSNE clustering of CD45 + LPMCs described in Figure 3. Appendix Table S6. Markers used for viSNE clustering of CD45 + CD3 -LPMCs described in Figure 5.     Figure 1C, (colitis histology score)

C. Statistical summary and p-values
Mann-Whitney test Figure 1E, (frequencies of IFN-g, IL-17A, TNF and Foxp3 CD4 + T cells in mLNs)   A. Statistical Summary and p-values Figure 2B, (calcium Influx Rate in LPMCs), RM one-way ANOVA test, corrected for multiple comparison with Two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli, (Q = 0.05).  Figure 3.

CD4 T cells (LPMC)
A. Statistical summary and p-values Figure 3E, (frequency of CD45 + LPMCs in UC, CD and non-inflamed) EdgeR statistical framework with negative binomial GLM and a false discovery rate adjusted to 10% Benjiamini-Hochberg  Table S11. (A-C) Statistical summary and p-values of Figure 4. Figure 4A, (frequency of CD45 + CD3 + LPMCs in UC, CD and non-inflamed)

A. Statistical summary and p-values
EdgeR statistical framework with negative binomial GLM and a false discovery rate adjusted to 10% Benjiamini-Hochberg  Table S12. (A-C) Statistical summary and p-values of Figure 5.
A. Statistical summary and p-values Figure 5A, (frequency of CD45 + CD3 -LPMCs in UC, CD and non-inflamed) EdgeR statistical framework with negative binomial GLM and a false discovery rate adjusted to 10% Benjiamini-Hochberg Appendix Tables S17-S22. Statistical Summary and p-values of Appendix. Figure S1 (  Figure S2 (A-B) A. Appendix Fig. S2A Figure S4 (A-C) A. Appendix Fig. S4B

S20. Statistical summary and p-values Appendix Figure S9
(Expression levels of TNF in T cell clusters) Wilcoxon matched-pairs signed rank test TNF expression p-value TNF expression p-value cluster 7 vs 1 in UC + P/I 0.0156 cluster 7 vs 1 in CD + P/I 0.0156 cluster 7 vs 3 in UC + P/I 0.0156 cluster 7 vs 3 in CD + P/I 0.0313 cluster 7 vs 10 in UC + P/I 0.0156 cluster 7 vs 10 in CD + P/I 0.0156 cluster 7 vs 5 in UC + P/I 0.0156 cluster 7 vs 5 in CD + P/I 0.0156 cluster 7 vs 6 in UC + P/I 0.0156 cluster 7 vs 6 in CD + P/I 0.0156 cluster 7 vs 15 in UC + P/I 0.1094 cluster 7 vs 15 in CD + P/I 0.1563 cluster 7 vs 12 in UC + P/I 0.5 cluster 7 vs 12 in CD + P/I 0.0469 cluster 7 vs 16 in UC + P/I 0.0156 cluster 7 vs 16 in CD + P/I 0.0156 cluster 7 vs 19 in UC + P/I 0.2188 cluster 7 vs 19 in CD + P/I 0.0156 Figure S10 (A-B) A. Appendix Fig. S10A