Figures and data
Unique phenotypic traits of naïve NOD CD8+ T cells are stratified by CD5 expression.
Flow cytometry analysis of naïve CD8+ T cells with the upper 5% (CD5hiCD8+) and lower 5% (CD5loCD8+) CD5 expression from the spleen and PLNs of normoglycemia female NOD mice at the age of 6 to 8 weeks old. (A) Representative flow cytometry histograms and the corresponding (B) geometric mean fluorescence intensity (gMFI) of CD5, p-CD3ζ and p-Erk expression shown in the indicated samples. Additionally, the percentages of (C) transcription factors T-bet and Eomes, along with (D) cytokines Granzyme B, TNF-α, IFN-γ and IL-2 in CD8+ T cells are also presented in the indicated samples. The gMFI for effector/memory T cell-related markers includes (E) CXCR3, CD122, CD127 and (F) CD69, PD-1, KLRG-1, CD44 and CD25. (G) Representative flow cytometry plots (left) and gMFI (right) of positive IGRP-tetramer staining in CD8+ T cells compared among naive total CD8+ T cells, CD5hi and CD5lo populations. In (A) to (G), the naïve total CD8 (grey), CD5hi (red) and CD5lo (blue) populations are represented. Data represent mean ± SD. The data presented in (A) to (F) represent two experiments. The sample size was n = 5-10 (B), n = 6-10 (C), n = 7-10 (D), n = 5-10 (E) n = 6-10 (F) and n = 3 (G) per group. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, by one-way ANOVA (B-D and G) and unpaired, two-tailed t test (E and F).
The gene expression profile of naïve CD5hiCD8+ T cells reveals a poised phenotype for effector T cells in autoimmune diseases.
(A) RNA-Seq performed on the sorted 10% upper and lower CD5-expressing of naïve CD8+ T cells (CD5hiCD8+ and CD5loCD8+) from 6-8-week-old normoglycemia female NOD mice to understand the gene expression profile of CD5hiCD8+ T cells and comparing that to the CD5lo counterpart to gain insights into their potential role in autoimmune diseases. (B) The volcano plot displaying the transcripts that are upregulated and downregulated by RNA sequencing of naïve CD5hiCD8+ T cells compared to CD5loCD8+ T cells. The plot shows the log10 fold change of each transcript and the significance of the change, measured by the negative log10 of the adjusted P-value (adj.p). Genes with higher fold changes and lower adj.p values (red plots) indicate a significant change in expression between the two groups. (C) The gene set enrichment score plot for gene ontology terms (T cell activation, external side of plasma membrane, positive regulation of immune system process, and regulation of catalytic activity). Higher scores indicate a higher degree of enrichment of these processes in naïve CD5hiCD8+ T cells compared to the CD5lo counterpart. (D) The dot plots of the significant KEGG gene set enrichment in the top 10 enriched pathways, reinforcing the presence of a poised phenotype for effector T cells in autoimmune diseases in naïve CD5hiCD8+ T cells. The data presented in (A) to (D) represent two experiments. In (A) to (D), n=10 mice per CD5hiCD8+ or CD5loCD8+ group, with pooled spleen and PLN cells for sorting obtained from a total of 20 normoglycemia female NOD mice aged 6 to 8 weeks old. Significance is determined using a false discovery rate (FDR) adjusted P-value < 0.05 and validated using multiple testing correction method Benjamini-Hochberg correction.
Naïve CD5hiCD8+ and CD5loCD8+ T cells exhibit distinct expression in effector/memory T cell gene-expression clusters.
CD5 expression distinguishes CD8+ T cell subsets with differential activation and proliferation.
(A) Diagram depicting naïve CD8+ T cells from normoglycemic female NOD mice aged 6 to 8 weeks, stimulated with indicated concentrations of anti-CD3 (0-10 µg/ml) and anti-CD28 (2 µg/ml; 0 µg/ml for control) for 2 days. Flow cytometry diagrams (B and G, upper and left panel, respectively) and statistical results (B-H) were derived from FLOW analysis. Comparison of CD5 (B), p-CD3ζ (C), p-Erk (D), CD69 (E), CD44 (F) expression, proliferation levels (CTV dilution percentage) (G), and CD127 expression (H) between 5% of upper (CD5hi) and lower (CD5lo) CD5-expressing CD8+ populations was conducted using gMFI of indicated markers. (I) Histogram comparing CD69 expression between the upper and lower 10% of CD5-expressing CD8+ T cells (CD5hiCD8+ and CD5loCD8+). Naïve T cells from normoglycemic female NOD mice aged 6 to 8 weeks were stimulated with anti-CD3/CD28 dynabeads or PMA plus Ionomycin for specified durations and analyzed for CD69 expression by flow cytometry. (J) Comparison of Ki-67 expression and IFN-γ production by flow cytometry between CD5hiCD8+ and CD5loCD8+ T cells isolated from normoglycemic female NOD mice aged 6 to 8 weeks. (K) Stimulation of CD5hiCD8+ and CD5loCD8+ T cells isolated from normoglycemic female NOD mice aged 6 to 8 weeks with indicated amounts of anti-CD3/CD28 dynabeads or PMA plus Ionomycin for 2 days. Cell proliferation was assessed by [methyl-3H] thymidine incorporation (cpm; counts per minute). (L) Comparison of CD5 gMFI in CD8+CD69+ (upper gate of flow dot plot) to that in CD8+CD69− (lower gate of flow dot plot) population by histogram (CD8+CD69+ in solid red line; CD8+CD69− in grey), across spleen and PLNs from normoglycemic female NOD mice aged 6 to 8 weeks. The statistical results of the percentages of activated (CD8+CD69+) population (left) and the ratio of activated to non-activated CD5 gMFI (right) are shown. Data represent mean ± SD. The data presented in (B) to (H) represent two experiments. The sample size was n = 10 (B-H), n = 3 (J), n = 3-7 (K), and n = 4 (L) per group. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, by two-way ANOVA (B-K) and unpaired, two-tailed t test (L).
CD5hiCD8+ T cells-transferred group exhibits severe insulitis and disease phenotype.
(A) Diabetes incidence (left) in NOD Rag1−/− mice transferred with CD5hiCD8+, CD5loCD8+, or CD8+ T cells, combined with CD4+ T cells, respectively. The islet histological analyses and insulitis scoring (right) for the mice after a 4-week transfer in the respective transferred groups showed a higher diabetes frequency in the CD5hiCD8+ group compared to the CD5loCD8+ group. The cells were sorted from the 10% upper and lower of CD5 expression of naïve CD8+ T cells (CD5hiCD8+ and CD5loCD8+ T cells) from 6-8-week-old normoglycemic female NOD mice and adoptively transferred to 6-8-week-old female NOD Rag1−/− recipients via i.p. injection. Urine glucose concentrations of the four groups of mice were monitored weekly for diabetes incidence. Diabetes was defined as glycosuria > 500 mg/dl in two consecutive tests. Insulitis scoring criteria can be found in the supplementary materials and methods. (B) Percentages and cell numbers of CD8+ T cells in spleen and PLNs, analyzed from CD5hiCD8+ and CD5loCD8+ cells-transferred group, respectively, after 4-week adoptive transfer. (C) Percentages of TNF-α, IFN-γ and IL-2 production in CD5hiCD8+ and CD5loCD8+ group, respectively, from spleen or PLNs, analyzed by flow cytometry, with representative flow dot plots (left) and statistical results (right). (D) Percentages of T-bet, Eomes and Granzyme B-expressing in CD5hiCD8+ and CD5loCD8+ group, respectively, from spleen or PLNs, analyzed by flow cytometry, with representative flow dot plots (left) and statistical results (right). (E) Comparison of differential gMFI expression of CD5, CD25, CD44, CD69, CD122, CD127, KLRG1 and PD-1 between CD5hiCD8+ and CD5loCD8+ group from spleen or PLNs. (F) Comparison of percentages of central memory T cells (defined by CD44hiCD62Lhi) (upper) and positive IGRP-tetramer staining (lower) between CD5hiCD8+ and CD5loCD8+ groups from PLNs, with representative flow dot plots (left) and statistical results (right). Data represent mean ± SD. Sample size was n = 5-11 (A), and n = 3 (B-F) per group. *P < 0.05, **P < 0.01, ***P < 0.001, by log-rank test (A), by two-way ANOVA (B) and (E) and unpaired, two-tailed t test (C), (D) and (F).
The high CD5-linked TCR basal signals positively correlate with thymic selection in autoimmune-poised NOD8.3 mice.
(A) CD5, p-CD3ζ, p-Erk and IL-2 expression levels compared between the lower 5% (hollow blue) and upper 5% (hollow red) CD5 expression in thymocytes of normoglycemia female NOD mice aged 6 to 8 weeks old across DN, DP and CD8SP stages. (B-H) Comparison of various marker expression levels between thymocytes of normoglycemia female NOD (solid blue) and NOD8.3 (solid red) mice aged 6 to 8 weeks old across different selection stages. (B) CD5 gMFI in NOD and NOD8.3 thymocytes across DN, DP and CD8SP stages. (C) Percentages of p-CD3ζ (left) and the corresponding gMFI of p-CD3 (right) compared between NOD and NOD8.3 thymocytes in the indicated stages DN, DP and CD8SP. (D) Percentages of DN, DP and CD8SP composition compared between NOD and NOD8.3 thymocytes. (E) Percentages of indicated DN stages (DN1, DN2, DN3 and DN4) in the DN subset compared between NOD and NOD8.3 mice. DN stages are defined as follows: DN1 (CD44+CD25−), DN2 (CD44+CD25+), DN3 (CD44−CD25+) and DN4 (CD44−CD25−), as depicted in the gating shown in Supplementary Figure 3A, middle panels. (F) CD5 gMFI (upper) and corresponding percentages (lower) in the pre-positive selection (TCRβ−CD69−) and post-positive selection (TCRβ+CD69+) populations in DP thymocytes compared between NOD and NOD8.3 mice. The presence of higher TCR basal signals (indicated by higher CD5 gMFI) in NOD8.3 in the pre-positive selection stage leads to an increased percentage of TCRβ+CD69+ cells in the post-selection stage compared to NOD mice, as shown in Supplementary Figure 3F. (G) gMFI of CD122, CD127, CD44 and CD25 assessed in DP thymocytes compared between NOD and NOD8.3 mice. (H) Percentages of virtual memory (Vmemory) cells in CD8SP thymocytes compared between NOD and NOD8.3 mice. Data represent mean ± SD. The data presented in (B) to (H) represent two experiments. The sample size was n = 10 (A), n = 10-11 (B), n = 3-6 (C), n = 9-11 (D), n = 7-11 (E), n = 8-11 (F) n = 10-11 (G) and n = 8-9 (H) per group. *P < 0.05, **P < 0.01, ****P < 0.0001, by unpaired, two-tailed t test (A-H).
The protective effect of transgenic Pep on diabetogenesis is negated in dLPC/NOD8.3 mice.
(A) Spontaneous diabetes incidence (left) and insulitis scoring (right) of female non-Tg NOD, dLPC/NOD, NOD8.3 and dLPC/NOD8.3 mice. Insulitis scoring was performed on the islets from the pancreata of 6-8-week-old mice before the onset of diabetes in each group. Insulitis scoring criteria can be found in the supplementary materials and methods. Urine glucose concentrations of mice were monitored weekly for spontaneous diabetes incidence. Diabetes was defined as glycosuria > 500 mg/dl at two consecutive tests. (B-G) Characterization of various aspects of TCR signal strength and effector/memory T cell phenotypes among female normoglycemia non-Tg NOD (blue), dLPC/NOD (black), NOD8.3 (red) and dLPC/NOD8.3 (grey) mice at the age of 6 to 8 weeks old. (B) CD5 gMFI of CD8+ T cells in the spleen (left) and PLNs (right). (C) Percentage of p-CD3ζ+CD8+ T cells in the spleen (left) and PLNs (right). (D) Percentage of p-Erk+CD8+ T cells in the spleen (left) and PLNs (right). (E) gMFI of effector T cell markers CD69 and KLRG1 in CD8+ T cells from the spleen (upper) and PLNs (lower). (F) gMFI of memory T cell markers CD25, CD122 and CD44 in CD8+ T cells from the spleen (upper) and PLNs (lower). (G) Percentage of IL-2+-, TNF-α+-, and IFN-γ+CD8+ T cells in PLNs. Data in (B) to (G) represent mean ± SD. The data presented in (B) to (G) represent two experiments. The sample size was n = 7-10 (A), n = 4-11 (B), n = 3-7 (C), n = 3-7 (D), n = 5-11 (E), n = 5-11 (F) and n = 3-9 (G) per group. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, by log-rank test (A), by one-way ANOVA (B-F) and unpaired, two-tailed t test (G).
Examination of the CD5hi population reveals TCR repertoires linked to autoimmune disorders.
(A) Comparison of TCR repertoire differences between CD5hiCD8+ and CD5loCD8+ T cell populations conducted using bulk RNA sequencing, followed by TCR repertoire analysis using the MiXCR pipeline. Naïve CD5hiCD8+ and CD5loCD8+ cells were sorted from female normoglycemia NOD mice aged 6 to 8 weeks old following a similar procedure as outlined in Figure 2A, with 1.5 × 106 cells per group. The TCR repertoire of naïve CD5hiCD8+ T cells exhibited shorter CDR3 lengths, higher clone overlap values within group comparisons, pathogenic-related CDR3 motifs, and greater hydrophobicity and diversity in the TRA chain compared to CD8+CD5lo cells. (B-E) Comparison of the TRAV gene segment usage (B), TRAJ gene segment usage (C), TRBV gene segment usage (D), and TRBJ gene segment usage (E) within TRA variable and TRB variable genes between CD5hiCD8+ (red) and CD5loCD8+ (blue) cells. (F-J) Comparison of CDR3α (left) or CDR3β (right) length in amino acids (F), CDR3 clonal overlap by overlapped count matrix (G), top 20 CDR3 motifs (H), average hydrophobicity score (I) and diversity (J) of CDR3α (left) and CDR3β (right) between CD5hiCD8+ and CD5loCD8+ T cells. In (G), number in cell indicates count number shared within the comparison group. (H) Occurrence frequencies of the top 20 CDR3 motifs from mapped-CDR3α (left) and CDR3β (right) sequencings in CD5hiCD8+ and CD5loCD8+ cells, with the high-occurrence rate of the CDR3α motif “SGGSNYKLTF” (underlined in red) specific to T1D-related IGRP peptide in the CD8+CD5hi population. (I) Average hydrophobicity score defined by summing up the hydrophobicity score in each amino acid in the CDR3 motif and dividing by the CDR3 length, then averaging by total counts in each group. Diversity definition in (J) is shown in Materials and Methods section Clonotype repertoire metrics formulas. In (B-E), each point represents 10 mice, with two points for each group, totaling n = 20 mice (F-J) for all CD5hi or CD5lo samples. *P < 0.05, **P < 0.01, ***P < 0.001, by two-way ANOVA (B-E), by Welch two sample t-test (F), and unpaired, two-tailed t test (I) and (J).
