Btla signaling in conventional and regulatory lymphocytes coordinately tempers humoral immunity in the intestinal mucosa

Abstract

The Btla inhibitory receptor limits innate and adaptive immune responses, both preventing the development of autoimmune disease and restraining anti-viral and anti-tumor responses. It remains unclear how the functions of Btla in diverse lymphocytes contribute to immunoregulation. Here, we show that Btla inhibits activation of genes regulating metabolism and cytokine signaling, including Il6 and Hif1a, indicating a regulatory role in humoral immunity. Within mucosal Peyer's patches, we find T-cell-expressed Btla-regulated Tfh cells, while Btla in T or B cells regulates GC B cell numbers. Treg-expressed Btla is required for cell-intrinsic Treg homeostasis that subsequently controls GC B cells. Loss of Btla in lymphocytes results in increased IgA bound to intestinal bacteria, correlating with altered microbial homeostasis and elevations in commensal and pathogenic bacteria. Together our studies provide important insights into how Btla functions as a checkpoint in diverse conventional and regulatory lymphocyte subsets to influence systemic immune responses.

Keywords: B cell; CP: Immunology; CP: Microbiology; IgA; Tfh; Tfr; Treg; autoimmunity; germinal center; inhibitory receptor; microbiome.

Figure 1.. Anti-Btla inhibits inflammatory signaling in T and B cells

(A) Setup for RNAseq experiment with C57BL/6 animals. (B) Volcano plots indicating genes significantly altered in anti-Btla-treated cells compared with control in anti-CD3-activated CD4⁺ T cells or anti-IgM/IgG activated. (C) Heatmaps of significantly altered genes in CD4⁺ T or B cells with GSEA hallmark pathways shown. (D) Hallmark GSEA pathways in CD4⁺ T and B cells. (E) Graphs of gene expression in germinal center B cells in SRBC-immunized wild-type or Btla^−/− hosts. Data analyzed using Student’s t test. *p < 0.05; ****p < 0.0001. Representative two experiments with n = 4 replicates for each condition. See also Figure S1.

Figure 2.. Btla regulates spontaneous GC reactions in aged animals

(A) Spleen sections from 9-month-old wild-type or Btla^−/− animals stained with IgD and peanut agglutinin (PNA). Scale bar equals 1 mm. (B) Plots and graph of CD19⁺IgD^lowGL7⁺Fas⁺ spleen GC B cells in aged wild-type and Btla^−/− animals. (C) Stacked bar graph of total area of GC reactions in (A). (D and E) Graphs of total serum Ig (D) and isotype-specific antibody titers (E) in aged wild-type and Btla^−/− animals. Data analyzed using Student’s t test. Error bars indicate SD. *p < 0.05; **p < 0.01. Representative of four separate experiments with n = 4 replicates for each condition. See also Figures S3A–S3I.

Figure 3.. Btla expression in T cells limits serum IgA in aged animals

(A) Spleen sections from 11-month-old Btla^flox and Btla^ΔCd4 animals stained for PNA and DAPI or B220, CD3, IgA, and DAPI. Scale bar equals 100 μm. (B) Graph of number of PNA⁺ GC in (A). (C) Plots and graph of IgD^lowCD38^lowGL7⁺Fas⁺ GC B cells in CD19⁺ B cells. (D) Plots of Cxcr4^hiCD86^lo DZ and Cxcr4^loCD86^hi LZ cells and graph of DZ/LZ cell ratio in spleens of aged Btla^flox and Btla^ΔCd4 animals. (E) Graph of serum IgA in aged Btla^flox, Btla^ΔCd4, and Btla^ΔCd19 animals. Data analyzed using multi-parameter linear modeling ANOVA. *p < 0.05; **p < 0.01. Representative of four separate experiments with n = 4 replicates for each condition. See also Figures S2 and S3J–S3M.

Figure 4.. Btla expression in T and B cells regulates GC reactions in PPs

(A and B) Plots and graphs of Foxp3⁻PD1⁺Cxcr5⁺ Tfh, Foxp3⁺PD1⁻Cxcr5⁻ Treg, and Foxp3⁺PD1⁺Cxcr5⁺ Tfr cells (A) and IgD^lowCD38^hi memory B cells, IgD^lowCD38^lowGL7⁺Fas⁺ GC B cells, GC B cell cellularity, and PP cellularity (B) in PPs of 8-week-old Btla^flox, Btla^ΔCd4, and Btla^ΔCd19 animals. (C) PP sections from 8-week-old Btla^flox and Btla^ΔCd4 animals stained with anti-CD4, anti-B220, anti-IgA, and DAPI. Scale bar equals 100 μm. (D) Plots of Cxcr4^hiCD86^lo DZ and Cxcr4^loCD86^hi LZ cells and graph of DZ/LZ cell ratio in PPs of 8-week-old Btla^flox, Btla^ΔCd4, and Btla^ΔCd19 animals. Data analyzed using multi-parameter linear modeling ANOVA. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. Representative of three separate experiments with n = 4 replicates for each condition. See also Figures S4A–S4H.

Figure 5.. Hvem expression in T and B cells regulates GC reactions in PPs

(A and B) Plots and graphs of Foxp3⁻PD1⁺Cxcr5⁺ Tfh, Foxp3⁺PD1⁻Cxcr5⁻ Treg, and Foxp3⁺PD1⁺Cxcr5⁺ Tfr cells (A) and IgD^lowCD38^lowGL7⁺Fas⁺ GC B cells, GC B cellularity, and PP cellularity in PPs of 8-week-old Tnfrsf14^flox, Tnfrsf14^ΔCd4, and Tnfrsf14^ΔCd19 animals (B). (C) PP sections from 8-week-old Tnfrsf14^flox and Tnfrsf14^ΔCd19 animals stained with anti-CD4, anti-B220, anti-IgA, and DAPI. Scale bar equals 100 μm. (D) Plots of Cxcr4^hiCD86^lo DZ and Cxcr4^loCD86^hi LZ cells and graph of DZ/LZ cell ratio in PPs of 8-week-old Tnfrsf14^flox, Tnfrsf14^ΔCd4, and Tnfrsf14^ΔCd19 animals. Data analyzed using multi-parameter linear modeling ANOVA. *p < 0.05; **p < 0.01; ***p < 0.001. Representative of three separate experiments with n = 4 replicates for each condition. See also Figures S4I and S4J.

Figure 6.. Cell-intrinsic maintenance of Treg cell homeostasis and function by Btla in PPs

(A–C) Graphs of T cell subsets in PPs of control or anti-Btla-treated wild-type (A), Tnfrsfl14^−/− (B) animals, or of untreated Btla^flox and Btla^ΔFoxp3 animals (C). (D) Plots and graph of FoxP3⁺YFP^+/− Treg cells in PPs of untreated wild-type Foxp3^+/cre and Btla^flox/floxFoxp3^+/cre animals. (E and F) Plots and graphs of Foxp3⁻PD1⁺Cxcr5⁺ Tfh, Foxp3⁺PD1⁻Cxcr5⁻ Treg, and Foxp3⁺PD1⁺Cxcr5⁺ Tfr cells (E) and IgD^lowCD38^lowGL7⁺Fas⁺ GC B cells (F) in PPs of untreated wild-type, Btla^−/−, and tamoxifen-treated Btla^flox and Btla^ΔFoxp3ERT2 animals. Tam, yamoxifen. Data analyzed using multi-parameter linear modeling ANOVA. *p <0.05; **p<0.01; ***p < 0.001; ****p < 0.0001. Representative of six(wild-type) experiments, with at least n = 5 replicates for each condition; two (Tnfrsf14^−/−) experiments, with at least n = 7 replicates for each condition in antibody treatment experiments; and four (constitutive) experiments, with at least n = 3 replicates for each condition and two (induced) Treg cells Btla deletion experiments, with at least n = 8 replicates for each condition. See also Figures S5 and S6.

Figure 7.. Lymphocyte expressed Btla and Hvem regulate microbial homeostasis

(A and B) Graphs of IgA in fecal pellets of Btla^flox, Btla^ΔCd4, and Btla^ΔCd19 animals (A), and Tnfrsf14^flox, Tnfrsf14^ΔCd4, and Tnfrsf14^ΔCd19 animals (B). (C and D) Plots and graphs of IgA⁺DAPI⁺ bacteria in fecal pellets of Rag1^−/−, Btla^flox, Btla^ΔCd4, and Btla^ΔCd19 animals (C), and Rag1^−/−, Tnfrsf14^flox, Tnfrsf14^ΔCd4, and Tnfrsf14^ΔCd19 animals (D). (E and F) PcoA plot (E) and graph of microbial β diversity (F) in fecal pellets of Btla^flox, Btla^ΔCd4, and Btla^ΔCd19 animals. (G–I) Experimental setup (G), plots of sorting strategy (H), and qRT-PCR graphs of total and IgA⁺Igκ⁺ bacteria species (I) in cohoused Btla^flox and Btla^ΔCd4 animals. MIB, mouse intestinal Bacteroides; SFB, segmented filamentous bacteria; data analyzed using multi-parameter linear modeling ANOVA. *p < 0.05; **p < 0.01; ***p < 0.001. Representative of five (Btla^flox, Btla^ΔCd4, and Btla^ΔCd19) experiments with at least n = 4 replicates for each condition, three (Tnfrsf14^flox, Tnfrsf14^ΔCd4 and Tnfrsf14^ΔCd19) experiment swith at least n = 4 replicates for each condition, and two (cohoused Btla^flox and Btla^ΔCd4) experiments with at least n = 4 replicates for each condition. See also Figure S7.