Responders to low-dose ATG induce CD4+ T cell exhaustion in type 1 diabetes

Abstract

BACKGROUNDLow-dose anti-thymocyte globulin (ATG) transiently preserves C-peptide and lowers HbA1c in individuals with recent-onset type 1 diabetes (T1D); however, the mechanisms of action and features of the response remain unclear. Here, we characterized the post hoc immunological outcomes of ATG administration and their potential use as biomarkers of metabolic response to therapy (i.e., improved preservation of endogenous insulin production).METHODSWe assessed gene and protein expression, targeted gene methylation, and cytokine concentrations in peripheral blood following treatment with ATG (n = 29), ATG plus granulocyte colony-stimulating factor (ATG/G-CSF, n = 28), or placebo (n = 31).RESULTSTreatment with low-dose ATG preserved regulatory T cells (Tregs), as measured by stable methylation of FOXP3 Treg-specific demethylation region (TSDR) and increased proportions of CD4+FOXP3+ Tregs (P < 0.001) identified by flow cytometry. While treatment effects were consistent across participants, not all maintained C-peptide. Responders exhibited a transient rise in IL-6, IP-10, and TNF-α (P < 0.05 for all) 2 weeks after treatment and a durable CD4+ exhaustion phenotype (increased PD-1+KLRG1+CD57- on CD4+ T cells [P = 0.011] and PD1+CD4+ Temra MFI [P < 0.001] at 12 weeks, following ATG and ATG/G-CSF, respectively). ATG nonresponders displayed higher proportions of senescent T cells (at baseline and after treatment) and increased methylation of EOMES (i.e., less expression of this exhaustion marker).CONCLUSIONAltogether in these exploratory analyses, Th1 inflammation-associated serum and CD4+ exhaustion transcript and cellular phenotyping profiles may be useful for identifying signatures of clinical response to ATG in T1D.TRIAL REGISTRATIONClinicalTrials.gov NCT02215200.FUNDINGThe Leona M. and Harry B. Helmsley Charitable Trust (2019PG-T1D011), the NIH (R01 DK106191 Supplement, K08 DK128628), NIH TrialNet (U01 DK085461), and the NIH NIAID (P01 AI042288).

Keywords: Autoimmune diseases; Diabetes; Endocrinology; Immunology; Immunotherapy.

Figure 1. C-peptide AUC change over time for responders and nonresponders.

The mean C-peptide AUC values for each treatment arm (ATG, blue; ATG/G-CSF, green; placebo, red) were disaggregated by responders (solid line) or nonresponders (dashed line). Response is based on the definition of an individual participant slope from baseline to 48 weeks being above the median (responders) or below the median (nonresponders) for the entire cohort, with the number of participants listed above or below the graph, respectively. The model included participant-level random effects terms for the intercept and slopes (16). The remaining clinical trial C-peptide AUC data from 48 weeks to 96 weeks are shaded in gray, as they were not part of the response definition. There were 582 samples collected, including 196 in the placebo group, 197 in the ATG group, and 189 in the ATG/G-CSF group. Among these, 279 samples were from the nonresponder group and 303 samples were from the responder group.

Figure 2. ATG and ATG/G-CSF transiently induce Th1 and innate cell cytokines.

Cytokine log₂ concentration (pg/mL) over time by treatment arm and responder/nonresponder status for (A) INF-γ, (B) IL-17A, (C) IL-1β, (D) IL-2, (E) IL-12p70, (F) MIP-1α, (G) IL-6, (H) G-CSF, (I) IL-10, (J) IL-4, (K) IP-10, (L) TNF-α, and (M) sIL-2RA. Blue lines represent the ATG arm, green lines the ATG/G-CSF arm, and red lines the placebo arm. Post hoc ANOVA (concentration log₂ transformed) testing was done to determine differences in ATG versus placebo and ATG/G-CSF versus placebo. *P < 0.05, **P < 0.001. Solid lines denote responders and dashed lines nonresponders. Statistical comparisons between responders and nonresponders are not shown here. For panels A–M, there were 348 samples, comprising 87 samples at baseline, 87 samples at week 2, 87 samples at week 12, and 86 samples at week 24.

Figure 3. Th17 and Th1 augmentation following ATG and ATG/G-CSF denote this treatment effect is part of the mechanism of action in T1D.

Blue lines represent the ATG arm, green lines the ATG/G-CSF arm, and red lines the placebo arm. Solid lines denote responders and dashed lines nonresponders. Panels depict (A) activated Th17 cells (percentage CD38⁺ of non-naive CCR6⁺CD4⁺ T cells) via flow cytometry (n = 322 samples over 4 time points), (B) median percentage methylation of the CXCR3 5′ upstream region (in the promoter) (n = 257 samples over 3 time points), (C) percentage of CXCR3⁺CD45RO⁺CD4⁺ T cells (P = 322), (D) percentage of activated CD4⁺ T cells (percentage CD38⁺ of non-naive CXCR3⁺CD4⁺ T cells) (n = 322), (E) percentage of activated CD8⁺ T cells (percentage CD38⁺ of non-naive CXCR3⁺CD8⁺ T cells) (n = 322), and (F) methylation of the ITGAL 5′ upstream region (n = 259). Several treatment effects were noted via post hoc ANCOVA (*P < 0.05, **P < 0.001). Statistical testing comparing responders and nonresponders is not demonstrated here.

Figure 4. Differential effect of ATG and ATG/G-CSF on Tregs.

Blue lines represent the ATG arm, green lines the ATG/G-CSF arm, and red lines the placebo arm; solid lines denote responders and dashed lines nonresponders. Panels depict (A) the median percentage methylation of the TSDR region of FOXP3 (n = 255 samples over 3 time points) and (B) the percentage of FOXP3⁺CD4⁺ T cells via flow cytometry (n = 279 samples over 4 time points). Treatment effects were assessed via post hoc ANCOVA (*P < 0.05, **P < 0.001). In panel A, there was not a significant difference between ATG-treated participants and placebo- or ATG/G-CSF–treated and placebo at any single time point. However, there was a marked rise in the percentage methylation of the FOXP3 TSDR from baseline to 12 weeks (and baseline to 48 weeks) in those in the ATG/G-CSF arm (P < 0.001 for both).

Figure 5. Responders identified via transient cytokine surge and exhaustion but not senescent T cell phenotype.

Blue lines represent the ATG arm, green lines the ATG/G-CSF arm, and red lines the placebo arm. Solid lines denote responders and dashed lines nonresponders. The panels depict (A) IL-6 log₂ concentration (n = 348 samples over 4 time points), (B) IP-10 log₂ concentration (n = 348), (C) TNF-α log₂ concentration (n = 348), (D) exhausted CD4⁺ T cells (percentage PD-1⁺KLRG1⁺CD57^– on CD4⁺ T cells) (n = 279), (E) median fluorescence intensity (MFI) of PD-1 on Temra CD4⁺ T cells (n = 264), (F) percentage TIGIT⁺ on Temra CD8⁺ T cells (n = 276), (G) median percentage methylation of EOMES 5′ upstream region (n = 252 samples over 3 time points), (H) percentage CD57⁺ of Tcm CD8⁺ T cells (n = 275), (I) percentage CD57⁺ on Tcm non-Treg CD4⁺ T cells (n = 276), (J) percentage CD57⁺ on Tem non-Treg CD4⁺ T cells (n = 277), (K) senescent CD4⁺ T cells (percentage CD57⁺KLRG1⁺ of PD-1^–CD4⁺ T cells) (n = 279), and (L) naive CD4⁺ T cells (percentage CD45RA⁺ of non-Treg CD4⁺ T cells) (n = 322). *P < 0.05, **P < 0.001 by post hoc ANCOVA testing for comparisons between responders and nonresponders within a treatment arm. For panels A and B, there was a significant increase in cytokine concentration from baseline to 2 weeks and significant fall from 2 weeks to both 12 and 24 weeks in responders (P < 0.001 for all). In panel C, responders also demonstrated a significant rise from baseline to 2 weeks (P = 0.013) and significant fall from 2 weeks to 24 weeks (P = 0.004).