Dynamic Immune Phenotypes of B and T Helper Cells Mark Distinct Stages of T1D Progression

Abstract

Multiple studies of B- and T-cell compartments and their response to stimuli demonstrate alterations in established type 1 diabetes (T1D). Yet it is not known whether these alterations reflect immune mechanisms that initiate islet autoimmunity, promote disease progression, or are secondary to disease. To address these questions, we used samples from the TrialNet Pathway to Prevention study to investigate T-cell responses to interleukin (IL)-2 and regulatory T cell-mediated suppression, the composition of the B-cell compartment, and B-cell responses to B-cell receptor and IL-21 receptor engagement. These studies revealed stage-dependent T- and B-cell functional and immune phenotypes; namely, early features that differentiate autoantibody-positive at-risk first-degree relatives (FDRs) from autoantibody-negative FDRs and persisted through clinical diagnosis; late features that arose at or near T1D diagnosis; and dynamic features that were enhanced early and blunted at later disease stages, indicating evolving responses along the continuum of T1D. We further explored how these specific phenotypes are influenced by therapeutic interventions. Our integrated studies provide unique insights into stable and dynamic stage-specific immune states and define novel immune phenotypes of potential clinical relevance.

Figure 1

Teff response to IL-2 and Teff suppression by Treg is decreased in autoAb⁺ subjects. A: IL-2–mediated STAT5 phosphorylation in autoAb^neg FDRs was compared with autoAb⁺ nonprogressors (Non-Prog) and autoAb⁺ progressors (Prog) in CD4⁺CD25^dimCD45RO⁺ memory T cells at time 1, ages ≤18 years. Cells were activated with 25 IU/mL IL-2 for 10 min and then fixed, permeabilized, and stained with antibodies against pSTAT5 (pY694), CD25, CD4, and CD45RO. Percentage positive pSTAT5 was determined based on media stimulation of the same sample as shown in Supplementary Fig. 1A. Significance was determined using a Mann-Whitney two-tailed test. B: IL-2/pSTAT5 response of Teff stratified by number of autoantibodies (numeric 0–4: includes autoAb^neg FDRs, autoAb⁺ nonprogressors, and autoAb⁺ progressors) at time 1, ages ≤18 years. AutoAbs include microinsulin autoAb, GAD65 standard, IA-2, ICA, ICA512, and Zn transporter (if assayed). Significance was determined using a one-way ANOVA (Kruskal-Wallace test) with Benjamini-Hochberg procedure for controlling false discovery rate. None of the comparisons was statistically significant. C: Treg suppression of CD4⁺CD25^dim Teff in autoAb^neg FDRs was compared with autoAb⁺ nonprogressors and autoAb⁺ progressors in total CD4⁺ T cells at time 1. Percent suppression was determined as in Supplementary Fig. 1B. Significance was determined using Mann-Whitney two-tailed tests. None of the two-way comparisons were statistically significant. D (left panel): Paired analysis of percent suppression between time 1 and time 2 for all autoAb⁺ subjects (nonprogressors and progressors) seropositive for two or more autoAbs. Significance was determined using a two-tailed Wilcoxon matched-pairs signed rank test. D (right panel): Percent suppression of Teff stratified by number of autoantibodies (numeric 0–4: includes autoAb^neg FDRs, autoAb⁺ nonprogressors, and autoAb⁺ progressors). AutoAbs include microinsulin autoAb, GAD65 standard, IA-2, ICA, ICA512, and Zn transporter (if assayed). Significance was determined using a one-way ANOVA (Kruskal-Wallace test) with Benjamini-Hochberg procedure for controlling false discovery rate. 0 vs. 3 autoAbs: P = 0.014; q = 0.06. All other comparisons were not statistically significant.

Figure 2

Naive B-cell homeostasis is altered in autoAb⁺ at-risk FDRs. A: Transitional B-cell frequency (as defined in Supplementary Fig. 2A) in autoAb^neg FDR was compared with autoAb⁺ nonprogressors (Non-Prog) and autoAb⁺progressors (Prog) at time 1, ages ≤18 years. Significance was determined using a Mann-Whitney two-tailed test. B: B_ND B-cell frequency (as defined in Supplementary Fig. 2A) in young autoAb⁺ nonprogressors and autoAb⁺ progressors as compared with autoAb^neg FDRs at time 2, ages ≤18 years. Significance was determined using a two-tailed Mann-Whitney test and one-way ANOVA (Kruskal-Wallace test). C: Stratification of BCR response by autoAb status at time 1, ages ≤18 years. Signaling was quantified as in Supplementary Fig. 3A. Significance was determined using a Mann-Whitney test. MFI, mean fluorescence intensity.

Figure 3

Enhanced IL-21 responsiveness in naive B cells from autoAb⁺ at-risk FDRs and diminished IL-21 signaling in autoAb⁺ FDRs who progress to T1D. Thawed PBMCs were rested in serum-free X-VIVO 15 medium for 1 h, washed with PBS, and stimulated with recombinant human IL-21 (0.1 ng/mL or 50 ng/mL) (Miltenyi Biotec) for 10 min. Cells were fixed, permeabilized, and stained with antibodies against pSTAT3 (pY705), CD3, CD4, CD45RA, CD8, CD20, and CD27. A and B: Multivariable statistical analysis on combined time point data for naive B-cell response to 0.1 ng/mL IL-21 was performed using a mixed-effects linear model of log₂-transformed fold change mean fluorescence intensity (MFI) pSTAT3 log₂(FC pSTAT3) with variables including cohort (autoAb^neg FDR, autoAb⁺ non-progressor [Non-Prog], and autoAb⁺ progressor [Prog]), number of autoantibodies (numeric, 0–4), time point, age, glucose, and 2-h AUC C-peptide. A: Modeled by cohort. Statistical significance was determined by ANOVA: P = 0.0273; pairwise comparisons were done using a Mann-Whitney two-tailed test. B: Modeled by number of autoantibodies. Statistical significance was determined by ANOVA with multiple comparisons testing: 0 vs. 2 autoAbs, P = 0.030; 1 vs. 2 autoAbs, P = 0.052; 0 vs. 3 autoAbs, P = 0.123; and 0 vs. 4 autoAbs, P = 0.307. Time point, age, and metabolic parameters were not significant in this model. C: Paired analysis of fold change MFI pSTAT3 by time point for autoAb⁺ cohorts following stimulation with 0.1 ng/mL IL-21. Significance was determined using two-tailed Mann-Whitney and Wilcoxon signed-rank tests. D: Analysis of IL-21/pSTAT3 response in adult healthy control subjects (N = 100) and subjects with T1D (n = 100) matched for sex, ethnicity, and age (control subjects, 18–49 years, mean 32.6 years; and T1D, 18–48 years, mean 32.5 years) from the BRI screened control and T1D repositories. Statistical significance was determined using a Mann-Whitney two-tailed test adjusted for multiple testing using the Benjamini-Hochberg method.

Figure 4

BCR response is significantly increased in the rituximab trial participants with sustained C-peptide at 1 year. A: Rituximab trial participant samples from pretreatment (BL) and week 52 (Wk 52) time points were assayed for BCR response as described in Supplementary Fig. 3A. Significantly increased fold change pPLCγ2 in total B cells from active treatment responders at 1-year posttherapy is shown (determined by a paired Wilcoxon signed-rank test). B: T1DAL trial participant samples from pretreatment and week 52 time points were assayed for BCR response as described in Supplementary Fig 3A. Statistical significance was determined by a paired Wilcoxon signed-rank test. Responder/nonresponder status is defined using the rituximab trial definition based on 2-h mixed-meal tolerance test assay results at week 24 (48).

Figure 5

Sequential model of the evolution of T- and B-cell phenotypes. Our findings yield unique insight into whether specific traits are “fixed” (arise early and persist into later disease stages; e.g., IL-2/pSTAT5 and increased transitional B cells), potentially because of genetic/epigenetic programs or early inflammation, or “acquired” (develop near or at disease manifestation, e.g., increased Teff resistance to Treg suppression and decreased B_ND B cells). These findings further reveal dynamic responses in B cells along the T1D continuum (BCR signaling and IL-21/pSTAT3 response), providing a better understanding of how and when these potentially cooperative B- and T- cell processes may be optimally targeted with combination therapies. Black font indicates findings published previously; red font indicates findings described in this study.