Impaired TIGIT expression on B cells drives circulating follicular helper T cell expansion in multiple sclerosis

Abstract

B cell depletion in patients with relapsing-remitting multiple sclerosis (RRMS) markedly prevents new MRI-detected lesions and disease activity, suggesting the hypothesis that altered B cell function leads to the activation of T cells driving disease pathogenesis. Here, we performed comprehensive analyses of CD40 ligand- (CD40L-) and IL-21-stimulated memory B cells from patients with MS and healthy age-matched controls, modeling the help of follicular helper T cells (Tfh cells), and found a differential gene expression signature in multiple B cell pathways. Most striking was the impaired TIGIT expression on MS-derived B cells mediated by dysregulation of the transcription factor TCF4. Activated circulating Tfh cells (cTfh cells) expressed CD155, the ligand of TIGIT, and TIGIT on B cells revealed their capacity to suppress the proliferation of IL-17-producing cTfh cells via the TIGIT/CD155 axis. Finally, CCR6+ cTfh cells were significantly increased in patients with MS, and their frequency was inversely correlated with that of TIGIT+ B cells. Together, these data suggest that the dysregulation of negative feedback loops between TIGIT+ memory B cells and cTfh cells in MS drives the activated immune system in this disease.

Keywords: Adaptive immunity; Autoimmunity; Immunology.

Figure 1. TIGIT is downregulated on memory B cells in MS.

(A–C) Sorted CD20⁺CD27⁺ memory B cells from patients with MS (n = 8) and healthy donors (HD) (n = 9) were cultured with CD40L and IL-21 for 2 days, and RNA-Seq was performed. (A) Heatmap of DEGs (|log₂ FC| >0.5, FDR < 0.1) in patients with MS and healthy donors. (B) Volcano plot depicting DEGs in memory B cells. Red dots represent significantly upregulated genes in MS-derived memory B cells, and blue dots represent significantly downregulated genes. Genes whose location is categorized as “plasma membrane” by IPA software are labeled. (C) IPA was performed to identify signatures related to altered molecular and cellular functions. Functions whose –log (Benjamini-Hochberg [B-H] P value) values were greater than 1.8 are shown. (D–F) Sorted CD20⁺CD27⁺ memory B cells from patients with MS and healthy donors (n = 12 each) were cultured with CD40L and IL-21 for 2 days. Gene expression was measured relative to B2M by qPCR (D and E). Representative flow data for TIGIT expression (F, left) and proportion of TIGIT⁺ cells (F, right). mB, memory B cells. Data are presented as the mean ± SEM and were evaluated by 2-tailed, unpaired Student’s t test (D–F). FSC-W, forward scatter width.

Figure 2. TIGIT is clearly separated from IL-10 expression.

(A) Representative flow data of TIGIT and IL-10 expression in CD40L+IL-21–stimulated memory B cells. (B) The proportions of TIGIT⁺IL-10⁺, TIGIT⁺IL-10^–, and TIGIT^–IL-10⁺ (n = 23) cells were evaluated by 2-way ANOVA with Dunn’s multiple-comparison test. Data are presented as the mean ± SEM. (C) Experimental workflow for RNA-Seq with DN (TIGIT^–IL-10^–) cells, TIGIT (TIGIT⁺IL-10^–) cells, and IL-10 (TIGIT^–IL-10⁺) cells. (D) PCA of RNA-Seq transcriptomes (n = 3 healthy donors). (E) Heatmap of representative genes that were differentially expressed (|log₂ FC| >0.5, FDR < 0.1) among 3 groups. (F) Representative flow data for PD-1 and CD226 (left) and their frequencies among TIGIT⁺ and TIGIT^– cells (right). (G) Representative flow data for FGL2 (left) and its frequency among TIGIT⁺ and TIGIT^– cells (right). Data are presented as the mean ± SEM. Significance was determined by 2-tailed, unpaired Student’s t test (F and G).

Figure 3. IL-4 treatment suppresses TIGIT expression on B cells.

(A) Healthy donor–derived CD19⁺ B cells (n = 10) were stimulated with CD40L in the presence of the indicated cytokines. The frequencies of TIGIT⁺ cells were measured by flow cytometry. Data are presented as the mean ± SEM and were evaluated by Dunnett’s multiple-comparison test. (B) Sorted CD20⁺CD27⁺ memory B cells from healthy donors were cultured with CD40L or CD40L+IL-4 for 2 days, and TIGIT mRNA expression was measured relative to B2M by qPCR (n = 11). Data were evaluated by 2-tailed, unpaired Student’s t test. (C and D) Sorted CD20⁺CD27⁺ memory B cells from healthy donors (n = 3) were cultured with CD40L or CD40L+IL-4 for 2 days, and RNA-Seq was performed. Heatmap of DEGs (|log₂ FC| >0.5, FDR < 0.1; 736 genes) between CD40L and CD40L+IL-4 conditions. (C) Representative genes are depicted. (D) Coinhibitory receptor expression pattern in CD40L or CD40L+IL-4–stimulated memory B cells. *FDR < 0.1.

Figure 4. TCF4 induces TIGIT expression on memory B cells.

(A) Venn diagrams showing the overlapped genes. Significantly upregulated genes (log₂ FC >0.5, FDR < 0.1) in TIGIT⁺IL-10^– cells compared with TIGIT^–IL-10⁺ cells and TIGIT^–IL-10^– cells were evaluated. Among the 73 overlapped genes, transcription factors are highlighted, and heatmaps are depicted on basis of the log₂ FC under the IL-4–stimulated condition. *FDR < 0.1, **FDR < 0.01, and ***FDR < 0.001. (B) TIGIT⁺ and TIGIT^– cells were sorted from CD20⁺CD27⁺ memory B cells stimulated with CD40L+IL-21 for 2 days, and TCF4 mRNA expression was measured relative to B2M by qPCR (n = 6). Significance was determined by 2-tailed, unpaired Student’s t test. (C) Representative histogram of TCF4 expression by flow cytometric analysis of TIGIT⁺ cells, TIGIT^– cells, and control (no staining). Max, maximum. (D) mRNA expression kinetics of TCF4 and TIGIT from 7 different time points (n = 7). Data are presented as the mean ± SEM. (E–G) CD20⁺CD27⁺ memory B cells were transfected with an siRNA targeting TCF4 (siTCF4) or the control (siCtrl). TCF4 expression was measured relative to B2M by qPCR, and 51% knockdown efficiency was confirmed (E). Representative flow data for TIGIT expression (F) and the proportion of TIGIT⁺ cells (G). Data are presented as the mean ± SEM and were evaluated by 2-tailed, unpaired Student’s t test (E and G).

Figure 5. The CD40/TCF4/TIGIT axis is dysregulated on MS-derived memory B cells.

(A–C) CD20⁺CD27⁺ memory B cells were transfected with siRNAs targeting ID2 and ID3 (siID2/ID3) or with siCtrl (n = 4). (A) Representative flow data for TIGIT expression. (B) Proportion of TIGIT⁺ cells by flow cytometric analysis. (C) TIGIT, TCF4, ID2, and ID3 expression levels were measured relative to B2M by qPCR (FC versus the siRNA/siXBP1 condition). (D and E) Sorted CD20⁺CD27⁺ memory B cells from healthy donors (n = 12) and patients with MS (n = 12) were cultured with CD40L+IL-21 for 2 days, and ID2, ID3 (D), and TCF4 (E) expression levels were measured relative to B2M by qPCR. All data are presented as the mean ± SEM and were evaluated by 2-tailed, unpaired Student’s t test (B, D, and E).

Figure 6. TIGIT⁺ B cells suppress the proliferation of CCR6⁺ Tfh cells.

(A–C) Sorted CD4⁺CD45RA^–CXCR5⁺ Tfh cells and CD4⁺CD45RA^–CXCR5^– non-Tfh cells were stimulated with anti-CD3/anti-CD28 antibodies (each 1 μg/mL) for 3 days. Representative flow data for CD155 expression (A, left) and the proportion of CD155⁺ cells (A, right) (n = 10) are shown. (B) PVR mRNA expression was measured relative to B2M by qPCR (n = 7). (C) Proportion of CD112⁺ cells (n = 10). (D) Experimental workflow for coculture assays with sorted CD20⁺CD27⁺ memory B cells and CD4⁺CD45RA^–CXCR5⁺ Tfh cells. (E) Representative flow data for Tfh cell proliferation. (F) Proportion of proliferated CellTrace Violet⁺ (CTV⁺) cells. (G) IL-17 and IFN-γ expression in the supernatants of coculture assays was evaluated by ELISA. (H) Correlation between CD4⁺CD45RA^–CXCR5⁺CCR6⁺ Tfh cells (percentage of CD4⁺CD45RA^–CXCR5⁺ Tfh cells) and TIGIT⁺ cells (percentage of CD20⁺CD27⁺ memory B cells). Data for healthy donors are indicated by blue dots (n = 15) and by red dots for patients with MS (n = 16). Linear regression is shown with a 95% CI (pink area). (I) Proportion of CCR6⁺ Tfh cells between healthy donors (n = 18) and patients with MS (n = 17). Data are presented as the mean ± SEM and were evaluated by 2-tailed, unpaired Student’s t test (A–C and I) or Wilcoxon matched-pairs, signed-rank test (F and G).