Fig. 7: HLA-DQ8-restricted C19S-specific CD4+ T cells expand in patients with T1D and acquire persistent memory activation during disease progression.
a, Bar graph (mean ± s.e.m.) showing the numbers of indicated DQ8 tetramer-binding T cell populations from all 27 individuals across non-diabetic (n = 7), recent-onset (n = 16) and established T1D (n = 4) cohorts. b, Representative flow cytometry plots (left) and bar graph (right; mean ± s.e.m.) showing the frequencies of naïve (CD45RA+CD95⁻) and memory (CD45RA⁻CD95+) cells within the indicated tetramer-binding CD4⁺ T cell populations in the non-diabetic (n = 5), recent-onset (n = 14) and established T1D (n = 4) cohorts. c, t-SNE plot showing unsupervised clustering of merged tetramer-binding and Tet− cells from all 27 participants in the three cohorts. d, Stacked bar graph showing the distribution of each cluster in each of the three T cell populations across non-diabetic, recent-onset and established cohorts. e, Representative flow cytometry plots (left) and longitudinal quantification (right; mean ± s.e.m.) showing CXCR3+CD95+ activated memory cells within CD127hiCD25⁻ conventional T cells among InsB12-20:DQ8 or InsB12-20(C19S):DQ8 tetramer-binding CD4⁺ T cells. f, Longitudinal quantification (mean ± s.e.m.) of CD25+CD127lo regulatory T cells (Treg cells) within InsB12-20:DQ8 or InsB12-20(C19S):DQ8 tetramer-binding CD4⁺ T cells in the three cohorts. g, Bar graph (mean ± s.e.m.) showing frequencies of individual activation markers (CD11a, CD49d and KLRG1) among CD95+ tetramer-binding CD4+ T cells in the non-diabetic (n = 5), recent-onset (n = 14) and established T1D (n = 4) cohorts. In a, b and g, n is the number of human subjects; each dot is one participant. In b, e, f and g, samples without detectable InsB12-20:DQ8 or InsB12-20(C19S):DQ8 tetramer-binding CD4⁺ T cells were excluded from the analysis. For statistical analysis, two-tailed Mann-Whitney tests were performed for between-cohort comparisons in a. Two-tailed paired t-tests were performed for within-subject comparisons in a, b and e–g. The data (a, b, e–g) represent 14 independent experiments.
