Antigen-specific immunotherapy with a CD4+ T cell neoepitope restrains CD8+ T cell differentiation in murine pancreatic islet grafts

DiLisio, James E.; Beard, K. Scott; Neef, Tobias; Miller, Stephen D.; Wenzlau, Janet M.; Nakayama, Maki; D’Antonio, Marc A.; Jamison, Braxton L.; Nicholson, Kelli S.; Baker, Rocky L.; Brunetti, Tonya M.; Gapin, Laurent; Haskins, Kathryn

doi:10.1038/s41467-026-70878-2

Download PDF

Article
Open access
Published: 24 March 2026

Antigen-specific immunotherapy with a CD4⁺ T cell neoepitope restrains CD8⁺ T cell differentiation in murine pancreatic islet grafts

Nature Communications volume 17, Article number: 4355 (2026) Cite this article

6292 Accesses
11 Altmetric
Metrics details

Subjects

Abstract

Pancreatic islet transplantation offers a potential cure for autoimmune type 1 diabetes (T1D) but requires immunosuppression to prevent recurrent, T cell-mediated autoimmunity. Antigen-specific immunotherapy is a targeted alternative that induces peripheral tolerance in autoreactive T cells without broad immunosuppression. Tolerance induction with antigen-specific therapies may induce dominant tolerance, extending to T cell clones other than those targeted by the therapy. However, the mechanism of this suppression remains unclear. Using pancreatic islet transplantation in NOD mice, we show that induction of tolerance to a CD4⁺ T cell hybrid insulin peptide neoepitope reprograms the fate of antigen-specific CD8⁺ T cells within the grafts. Induction of tolerance to a single hybrid insulin peptide transiently protected islet grafts from autoimmune destruction, and graft survival could be extended with continued therapeutic dosing. Treatment limited a cytolytic/interferon-stimulated differentiation program in antigen-specific CD8⁺ T cells and reduced TCR avidity within grafts. This was mediated in part by IL-10-producing regulatory CD4⁺ T cells that suppressed dendritic cell activation. IL-10 blockade reversed these effects, restoring CD8⁺ T cell differentiation, as well as the licensing of dendritic cells. These findings reveal an IL-10-dependent mechanism by which CD4⁺ T cell-targeted immunotherapy restrains pathogenic CD8⁺ T cell fates through the suppression of dendritic cells in islet grafts.

Enhanced CD4⁺ and CD8⁺ T cell infiltrate within convex hull defined pancreatic islet borders as autoimmune diabetes progresses

Article Open access 25 August 2021

Autoimmune diseases: targets, biology, and drug discovery

Article 14 December 2023

Single-cell multi-omics analysis of human pancreatic islets reveals novel cellular states in type 1 diabetes

Article 28 February 2022

Introduction

Type 1 diabetes (T1D) is an autoimmune disease characterized by the destruction of insulin-producing β-cells driven primarily by autoreactive T cells^1,2. Recent advances in stem cell-derived islet transplantation have offered a potential cure for diabetes^3,4,5,6. However, most recipients remain dependent on life-long immunosuppressive regimens, which over time increase the risk of nephrotoxicity⁷, infection⁸, and malignancy⁹. By contrast, antigen-specific immunotherapy (ASI) aims to selectively silence pathogenic T cell reactivity in target tissues while sparing the majority of the T cell repertoire^10,11. This offers a precision approach to restoring peripheral T cell tolerance in autoimmune patients in the absence of indiscriminate immunosuppression. In T1D and other T cell-mediated autoimmune diseases, precision delivery of autoantigen peptide for antigen-specific immunotherapy via tolerogenic vehicles such as nanoparticles has emerged as a versatile method to induce antigen-specific peripheral tolerance without eliciting broad immunosuppression^12,13,14,15.

A natural mechanism of peripheral tolerance is mediated by antigen-presenting cell (APC) uptake and presentation of self-peptides without co-stimulatory signals such as CD80 and CD40^16,17. Early efforts to therapeutically harness this pathway demonstrated that conjugating autoantigenic peptides to apoptotic cells could suppress autoreactive T cell responses and prevent autoimmunity^18,19. To improve translational potential through greater stability and uniformity, biodegradable poly(lactic-co-glycolic acid) (PLG) nanoparticles (NPs) were engineered to mimic apoptotic cells as tolerogenic, antigen-carrying vehicles²⁰. Epitope-conjugated PLG nanoparticles have since been shown to induce anergy in effector T cell responses and to promote regulatory T cell expansion in multiple autoimmune disease models^21,22,23.

Identifying the specificities of islet-infiltrating T cells has been pivotal for understanding disease progression and guiding the selection of ASI target peptides. Genetic susceptibility in both human T1D and the non-obese diabetic (NOD) mouse is predominantly linked to certain MHC class II alleles²⁴, highlighting the central role of autoreactive CD4⁺ T cells in disease initiation and progression²⁵. Early work in the NOD mouse identified a panel of CD4⁺ T cell clones, including the prototype T cell clone BDC-2.5, that could infiltrate islets and markedly accelerate disease²⁶. These clones enabled the identification of cognate autoantigens for islet-specific CD4⁺ T cells, culminating in the discovery of a novel class of post-translationally modified neoantigens, termed hybrid insulin peptides (HIPs)²⁷. HIPs are generated by covalent fusion of proinsulin C-peptide fragments to other β-cell granule-derived peptides^27,28. Among the HIPs identified in NOD islets, the 2.5HIP epitope (Ins2_(77–82)/ChgA_(358-362)) elicits a potent pathogenic CD4⁺ T cell response. 2.5HIP-reactive T cells are enriched in the islets of prediabetic NOD mice, increase in frequency during disease progression²⁹, and have been targeted in other preclinical antigen-specific immunotherapies^15,30,31. HIPs have also been detected in human islets, and HIP-reactive CD4⁺ T cells, including those specific for the human analog of the 2.5HIP, have been identified in patients with T1D^{27,32,33,34,35,36,37,38,39,40}.

CD4⁺ T cells typically act as helpers, supporting the activation, expansion, and differentiation of cytolytic CD8⁺ T cells through both direct and indirect mechanisms^41,42,43,44 including the licensing of dendritic cells (DCs). However, the specific pathways through which HIP-specific CD4⁺ T cells influence the fate of autoreactive CD8⁺ T cells, and how this instructional role is altered following tolerance induction, is not entirely understood. Recent studies have shown that antigen-specific autoreactive CD8⁺ T cells undergo a hierarchical differentiation process, beginning as self-renewing, stem-like precursors in pancreatic lymph nodes and progressing into terminal effectors that accumulate within the islets^45,46. The two populations can be readily distinguished by the expression of the self-renewal transcription factor, TCF1. Other work has demonstrated that a proportion of bulk CD8⁺ T cells in islets remain TCF1^hi and can differentiate into TCF1^lo effectors in situ^47,48. Some studies have shown that many islet-specific CD8⁺ T cells are restrained by an exhaustion-like, terminal differentiation program akin to that seen in cancer and chronic viral infection^49,50. This exhaustion-like program can be induced in islet-specific CD8⁺ T cells as demonstrated by extra-islet expression of the CD8⁺ T cell epitope, IGRP⁵¹. However, an unanswered question has been how antigen-specific immunotherapy with a CD4⁺ T cell epitope influences the differentiation fate in the CD4⁺ T cells targeted by the therapy and in other islet-specific T cells, including cytotoxic CD8⁺ effectors. It has been proposed that tolerance to CD4⁺ epitopes may constrain CD8⁺ T cell effector function, and therefore tissue differentiation, via a form of dominant tolerance mediated by the suppression of antigen-presenting cells^52,53,54,55. The APCs may lack proper CD4⁺ licensing, generating a helpless CD8⁺ fate, and/or could be actively suppressed by regulatory T cells. In our previous transplant tolerance study⁵⁶ and in another tolerance induction study during spontaneous disease progression⁵⁷, initial evidence of suppression in CD8⁺ T cells following CD4-targeted antigen-specific immunotherapy has been observed, but the precise nature of the defect in antigen-specific CD8⁺ T cells and the mechanism by which this peripheral tolerance occurs remains undefined.

Here, we use antigen-specific immunotherapy with a single CD4⁺ T cell hybrid insulin peptide neoepitope, the 2.5HIP, to mechanistically determine how peripheral tolerance shapes the tissue differentiation fate and function of islet-specific CD8⁺ T cells in a polyclonal, autoimmune syngeneic islet graft setting.

Results

Antigen-specific CD4⁺ and CD8⁺ T cells undergo stem-like to effector differentiation in islet grafts

Antigen-specific CD8⁺ T cells exhibiting stem-like features occupy the pancreatic lymph node and terminally differentiate into effector cells upon infiltration into islets, where antigen is abundant^45,46. This anatomical compartmentalization of cell states can be delineated by the expression of the self-renewal transcription factor, T cell factor 1 or TCF1 (gene Tcf7). Stem-like T cells express high levels of TCF1 (TCF1^hi) and terminally differentiated T cells lose TCF1 (TCF1^lo), to adopt an effector-like program in tissue. We hypothesized that antigen-specific CD8⁺ and CD4⁺ T cells infiltrating islet grafts undergo the same stem-to-effector transition. To test this hypothesis, we used an autoimmune transplant model in which spontaneously diabetic NOD mice receive syngeneic islet transplants under the kidney capsule from NOD.scid donors (Fig. 1a). Normoglycemia is transiently restored until the graft is destroyed by recurrent T cell mediated autoimmunity^58,59,60 and animals revert to hyperglycemia (Fig. 1b). Peptide-loaded MHC tetramers were used to examine antigen-specific T cells recognizing immunodominant CD4⁺ (2.5HIP presented by I-Ag⁷)²⁹ and CD8⁺ (IGRP_(206-214) presented by H2-K^d)⁶¹ T cell epitopes in the islet graft and the kidney-draining lymph node (kLN).

Fig. 1: Antigen-specific CD4+ and CD8+ T cells undergo a stem-like to effector differentiation in diabetic NOD mouse islet grafts. — **Fig. 1: Antigen-specific CD4⁺ and CD8⁺ T cells undergo a stem-like to effector differentiation in diabetic NOD mouse islet grafts.**

Around the time of graft failure at day 8–9 post-transplant, tetramer-positive T cells from islet grafts and kLN were assessed for antigen encounter and effector differentiation by flow cytometry (gating strategy in Fig. S1a). Both 2.5HIP-tetramer+ CD4⁺ (2.5HIP_CD4) and IGRP-tetramer+ CD8⁺ (IGRP_CD8) T cells were more abundant in the graft than in the draining lymph node, in frequency and absolute number (Fig. 1c). Similarly, 2.5HIP_CD4 and IGRP_CD8 T cells (>70%) were TCF1^hi in the kLN, whereas tetramer-positive T cells in the graft were predominantly TCF^lo, terminally differentiated effector T cells (Fig. 1d). These data indicate that both antigen-specific CD4⁺ and CD8⁺ T cells undergo a similar stem-like-to-effector transition from draining lymph node to graft. Levels of inhibitory receptors PD1 and TIM3 (as proxies for TCR activation and effector differentiation) and the proliferation marker Ki67 were higher in graft-infiltrating antigen-specific CD4⁺ and CD8⁺ T cells compared to the same specificities in the kLN, more evidence indicative of antigen encounter and subsequent differentiation in islet graft tissue (Fig. 1e).

To further understand the kinetics of effector differentiation leading up to graft destruction, the proportion and number of TCF1^hi/lo 2.5HIP_CD4 and IGRP_CD8 T cells were examined in islet grafts on day 3, 6, and 9 post-transplant (Fig. 1f, g). There was no significant change in the proportion of 2.5HIP_CD4 T cells of total CD4⁺ T cells, or as a fraction of TCF1^hi T cells, over the course of graft destruction. Yet, there is a significant accumulation in the absolute number of both TCF1^hi and TCF1^lo 2.5HIP_CD4 T cells as the graft progresses toward destruction. By contrast, IGRP_CD8 T cells increase as a proportion of total CD8⁺ T cells over time. Few IGRP_CD8 T cells can be detected at day 3 post-transplant, most of them being TCF1^lo effectors. By day 6, there is a reasonable percentage of (~15%) of TCF1^hi IGRP_CD8 T cells. While the absolute number of TCF1^hi IGRP_CD8 is established by day 6 and does not increase by day 9, there is a steady and significant accumulation of IGRP_CD8 TCF1^lo effector T cells between all timepoints. Together, these data demonstrate that both 2.5HIP_CD4 and IGRP_CD8 T cells are predominantly TCF1^hi in the graft-draining lymph node and differentiate into TCF1^lo effectors in islet grafts, aligning with previous observations of antigen-specific CD8⁺ T cells in pancreatic islets and the draining lymph node^45,46.

Induction of tolerance to a hybrid insulin peptide restrains functional effector differentiation of graft-infiltrating, antigen-specific T cells

Graft infiltration of antigen-specific T cells is a critical step in the destruction of transplanted islet grafts^62,63,64. We hypothesized that tolerance induction to a dominant CD4 neoepitope, the 2.5HIP (Ins2_(77-82)/ChgA_(358-362)), would diminish the magnitude and quality of effector fates in graft-infiltrating, antigen-specific CD4⁺ and CD8⁺ T cells. To test this hypothesis, tolerogenic PLG nanoparticles (NPs), covalently conjugated to either the 2.5HIP or an irrelevant control peptide from hen egg lysozyme (HEL_(11-25)) (Fig. 2a), were administered to spontaneously diabetic NOD mice before and after syngeneic islet transplantation (−7d, +1d, +7d relative to transplant) (Fig. 2b). A separate group of 2.5HIP NP tolerized mice were maintained on 2.5HIP NP treatment administered every 10 days for a total of 9 injections to determine if continued therapeutic treatment could extend graft survival.

**Fig. 2: Induction of tolerance to a hybrid insulin peptide restrains functional effector differentiation of graft-infiltrating, antigen-specific T cells.**

As previously demonstrated⁵⁶, three doses of 2.5HIP NPs were able to transiently extend graft survival (~45 days) over that of HEL NP control treated mice (~10 days) (Fig. 2c). Continued treatment with 2.5HIP NPs every 10 days until day 67 (9 total doses), increased graft survival even further to ~100 days and nearing the cured time-point in mice. This result indicates that graft survival can be extended commensurate with the duration of continued antigen-specific therapy.

To understand how tolerance was altering effector differentiation, non-survival mice were euthanized 8–9 days post-transplant, and graft-infiltrating 2.5HIP_CD4 (target of nanoparticle therapy) and IGRP_CD8 (immunodominant CD8⁺ T cell epitope) T cells were analyzed from islet grafts and kidney-draining lymph nodes (kLN). Treatment with 2.5HIP-NPs significantly reduced the frequency and absolute number of 2.5HIP_CD4 and IGRP_CD8 T cells in islet grafts (Fig. 2d). In the kLN, however, 2.5HIP-NP treatment slightly increased the proportion of 2.5HIP_CD4 T cells with no significant differences observed in IGRP_CD8 T cells (Fig. S2a, b). To assess the impact of antigen-specific tolerance on effector differentiation, we examined the stem-like (TCF1^hi) and effector-like (TCF1ˡᵒ) subsets among 2.5HIP_CD4 and IGRP_CD8 T cells in both the kLN (Fig. S2.c) and islet grafts (Fig. 2e). In islet grafts from 2.5HIP-NP tolerized animals, a greater proportion of antigen-specific T cells retained TCF1 expression compared to HEL-NP controls (Fig. 2e). 2.5HIP-NP treatment markedly reduced the number of TCF1⁻ effector cells in both 2.5HIP_CD4 and IGRP_CD8 T cells, despite similar absolute numbers of TCF1⁺ cells (Fig. 2f). In comparison, there were no changes in the kLN in the portion of TCF1^hi cells for either specificity in the kLN (Fig. S2c), or in the fraction of TCF1^hiCD62L⁺ T cells (Fig. S2d), a subset present in LNs thought to have the most stem-like properties⁴⁶. These data suggest the effects of tolerance induction by 2.5HIP-NPs are manifested as a lack in the absolute numbers of TCF1^lo graft-infiltrating, antigen-specific T cells with minimal impact on the same population in the draining lymph node at this equivalent timepoint.

To assess antigen encounter and exhaustion-like differentiation fates in islet grafts, expression levels of inhibitory receptors PD1 and TIM3 and Ki67 were measured within TCF1^hi and TCF1ˡᵒ subsets. The inhibitory receptor, PD1, was highly expressed in nearly all graft-infiltrating antigen-specific T cells and was significantly decreased in both TCF1^hi/lo 2.5HIP_CD4 populations and TCF1^hi IGRP_CD8 T cells following tolerance induction. By contrast, TIM3 surface expression was reduced exclusively on 2.5HIP_CD4 T cells in grafts upon 2.5HIP-NP tolerance induction, with little change in IGRP_CD8 T cells (Fig. 2g, h). These data may indicate that 2.5HIP-NP-treatment limits terminal effector differentiation and functional antigen-encounter in antigen-specific CD4⁺ and CD8⁺ T cells, as opposed to inducing an exhaustion-like program characterized by higher IR expression. Ki67 expression did not significantly differ between treatments for either specificity (Fig. S2e), suggesting local proliferation in the graft was unaffected at this time point. To examine how tolerance induction impacted the functional capacity of T cells, suspensions from grafts were stimulated ex vivo with PMA and ionomycin and stained for inflammatory cytokine accumulation. In our previous study⁵⁶, we observed impaired cytokine production upon graft-failure but did not resolve whether stem-like, effector, or both T cell populations were functionally disrupted, nor did we examine responses at equivalent timepoints. Antigen-specific T cells were gated on TCF1 by TIM3 to examine how functional capacity changes through progression to terminal differentiation. Tolerance induction reduced IFNγ⁺TNF⁺ double-producing cells among both 2.5HIP_CD4 (Fig. 2i, j) and IGRP_CD8 T cells (Fig. 2k, l). The significance and magnitude of these differences were most exaggerated in terminally differentiated (TCF1^loTIM3^hi) T cells compared to their stem-like counterparts. This indicates that, as T cells differentiate within a tolerogenic environment, they are progressively diverted toward an altered, hypofunctional terminal effector state with reduced inhibitory receptor expression, distinct from an exhaustion-like phenotype and exhibiting limited differentiation. Together, these findings demonstrate that 2.5HIP-NP-induced tolerance reduces the number and functional differentiation of TCF1^lo 2.5HIP_CD4 and IGRP_CD8 T cells in islet grafts.

IGRP-specific CD8⁺ T cells display a reduced interferon-stimulated gene (ISG) signature and cytolytic effector fate in grafts following 2.5HIP-NP tolerance induction

To further interrogate how 2.5HIP tolerance reshapes graft-infiltrating antigen-specific CD8⁺ T cell effector fates, IGRP_CD8 T cells were sorted (gating shown in Fig. S1b) from islet grafts of 2.5HIP-NP- and HEL-NP-treated NOD mice eight days post-transplant and subjected to single-cell RNA sequencing (Fig. 3a). Dimensionality reduction by UMAP on all sequenced cells, excluding TCR genes, revealed eight transcriptionally discrete clusters (Figs. 3b, S3a) annotated within the framework of effector differentiation, dominant gene signatures, and inferred function using differences in cluster-defining genes (Figs. 3c, S3b). Briefly, we found two stem-like clusters characterized by high expression of self-renewal genes (Tcf7, Ccr7, Slamf6), the larger Tcf7_hi cluster and a smaller subset, Tcf7_ISG, which additionally exhibited an interferon-stimulated gene (ISG) signature (Ifit3, Isg15, Ly6a). The six Tcf7 low clusters included four effector clusters Eff_ISG, Eff_Ifng, Eff_early, Eff_late characterized by high inhibitory receptor (Lag3, Pdcd1) and chemokine receptor expression (Ccr5, Cxcr6), a proliferating cluster Prolif (Mki67, Pclaf, Cdk1), and recently activated cluster Activated (Nfatc1, Nfatc3, Tox). While Tox transcripts were detected at high levels only in the Activated cluster, Tox protein was expressed at high levels in all graft-infiltrating IGRP_CD8 T cells and at higher levels than in IGRP-tetramer-negative CD8⁺ T cells (Fig. S3k). Further description of the differences between the four effector clusters are elaborated in Figure legend 3.

Fig. 3: IGRP-specific CD8+ T cells display a reduced interferon-stimulated gene (ISG) signature and cytolytic effector fate in grafts following 2.5HIP-NP tolerance induction. — **Fig. 3: IGRP-specific CD8⁺ T cells display a reduced interferon-stimulated gene (ISG) signature and cytolytic effector fate in grafts following 2.5HIP-NP tolerance induction.**

Notably, the Eff_ISG cluster showed elevated ISG expression, including high expression of the surface molecule Ly6a, and the highest levels of Gzmb, a key cytolytic effector molecule (Fig. 3d). The Eff_ISG cluster scored significantly above all other clusters for cytolytic gene modules (Fig. S3c, d) and ISG modules (except for the small Tcf7_ISG cluster), which were positively correlated together (Fig. S3e, f). Collectively, these data indicate that cytolytic and ISG signatures may represent a linked cell fate during the functional differentiation of graft-infiltrating IGRP_CD8 T cells.

Following cluster annotation, we determined how treatment impacted IGRP_CD8 T cells in grafts. Mirroring flow cytometry results (Fig. 2e, f), 2.5HIP-NP treatment led to a significant proportional increase in the Tcf7_hi stem-like cluster (Fig. S3g, h). There was also a concomitant reduction in the relatively small Tcf7_ISG and Activated clusters (Fig. S3g, h). To broadly assess treatment-induced transcriptional changes, pseudo-bulk differential expression revealed that approximately 30% of genes downregulated upon 2.5HIP-NP treatment were ISGs (Ly6a, Ifit3, Ifitm2, Ifitm1), alongside significant decreases in the main cytolytic effector molecules Gzma and Gzmb (Fig. 3e). This indicates the arrest of the aforementioned highly functional ISG/cytolytic IGRP_CD8 fate by 2.5HIP-NP treatment. Gene set enrichment analysis (GSEA) confirmed marked depletion of ISG activation in IGRP_CD8 T cells with significant reduction in an Interferon Induced Antiviral Module and significant reduction in an Antigen Response pathway (Figs. 3f, S3i). Single-cell module scoring demonstrated bulk and cluster-specific reductions in Ly6a and Gzmb expression following 2.5HIP tolerance induction (Figs. 3g, S3j).

To validate the transcriptional reduction in the ISG signature in IGRP_CD8 T cells, Ly6A surface expression, readily detected by flow cytometry compared to other ISGs, was measured on IGRP_CD8 T cells from grafts of control (HEL-NP) and tolerized (2.5HIP-NP) mice (Fig. 3i). Relative levels of Ly6A progressively increased on IGRP_CD8 T cells from spleen to pancreatic lymph node and peaked in the graft, indicating Ly6A may serve as a marker of effector differentiation and survival in graft tissue (Fig. 3h). Reflecting the scRNAseq data, Ly6A surface levels were significantly lower in both TCF1^lo and TCF1^hi IGRP_CD8 T cells following 2.5HIP tolerance induction (Fig. 3i). To further dissect how peripheral tolerance alters CD8⁺ T cell fate, we applied Monocle3 pseudotime analysis⁶⁵ to infer differentiation trajectories of IGRP_CD8 T cells within the graft (Fig. 3j). In control mice, cells progressed through the highly functional Eff_Ifng cluster and predominantly terminated in the cytolytic Eff_ISG state. By contrast, after 2.5HIP-NP treatment, cells followed an alternative trajectory through the Eff_early cluster and instead terminated partway into the less functional Eff_late state, largely bypassing the Eff_Ifng and Eff_ISG branches. These findings suggest that 2.5HIP-induced peripheral tolerance diverts autoreactive CD8⁺ T cells away from a cytolytic/ISG fate toward a hypofunctional differentiation state.

IGRP-specific CD8⁺ T cells have a low-avidity TCR repertoire following 2.5HIP-NP tolerance induction

Single-cell RNA sequencing analysis revealed that 2.5HIP-NP-mediated tolerance induction dampens TCR-driven transcriptional programs in IGRP-specific CD8 T cells. Expression of Rln3, a downstream target of the TCR-regulated transcription factor Nur77^66,67, was significantly reduced (Fig. 3e). In parallel, expression of Cd5, a gene whose surface expression on T cells correlates with cumulative TCR signal strength and apparent TCR avidity⁶⁸, was also decreased in tolerized cells (Figs. 3e, 4a). Additionally, GSEA revealed a decrease in an Antigen Response pathway following 2.5HIP-NP tolerance induction (Figs. 3f, S3i). Together, these findings suggest that induction of tolerance to the 2.5HIP modifies the TCR signaling axis in graft-infiltrating IGRP_CD8 T cells. Therefore, we hypothesized that 2.5HIP-NP-induced tolerance might impact the relative avidity/affinity of the IGRP_CD8 TCR repertoire in islet grafts.

Fig. 4: IGRP-specific CD8+ T cells have a modified TCR repertoire following 2.5HIP-NP tolerance induction. — **Fig. 4: IGRP-specific CD8⁺ T cells have a modified TCR repertoire following 2.5HIP-NP tolerance induction.**

Surface staining of IGRP_CD8 T cells confirmed treatment-induced reduction in CD5 levels (Fig. 4b) that were observed in the transcriptional results. Tetramer staining intensity for similar levels of surface TCR expression approximates the relative avidity of TCRs^69,70. Indeed, IGRP_CD8 T cells from 2.5HIP-NP tolerized grafts had a lower average IGRP-H2-Kd tetramer gMFI than IGRP_CD8 T cells from control, HEL-NP treated grafts (Fig. 4c). To ensure that down-regulation of surface TCR levels was not responsible for the apparent reduction in tetramer avidity, we examined the total level of the TCR on IGRP_CD8 T cells and found no significant differences in the gMFI for surface TCR levels (Fig. 4d). These results indicate that the repertoire of IGRP_CD8 T cells present in 2.5HIP-NP tolerized grafts may have a lower apparent avidity than the CD8 T cells in control grafts.

To assess whether 2.5HIP-NP tolerance alters the diversity and public nature of the IGRP_CD8 T cell TCR repertoire, we analyzed TCR sequences from 5’ scRNA-seq data. Analysis of expanded clones revealed highly restricted V gene usage, with a single Vα (TRAV16N) and a single Vβ (TRBV13-3) comprising over 50% of variable chain usage across all animals regardless of treatment (Fig. S4a), consistent with prior reports of islet-infiltrating IGRP_CD8 T cells^48,71. Every animal, whether treated with 2.5HIP-NPs or not, had 3 to 4 highly expanded TCR clones occupying ~60-90% of their respective repertoire (Fig. 4e). Expanded clones were largely private among control animals, with minimal sharing observed (Fig. 4e, f). By contrast, a single public clone comprised over 50% of the repertoire in two of the 2.5HIP-NP-treated animals (HIP_1, HIP_2) and was significantly expanded in the third (HIP_3) (~10%) (Fig. 4e, f). The combined α and β CDR3 sequences of this treatment-expanded clone (Fig. S4b) included at least two distinct nucleotide variants across all three 2.5HIP-NP-treated animals (Fig. S4c), indicating convergent TCR recombination. Repertoire diversity and public TCRs were assessed by the Shannon Diversity index and the Jaccard similarity index, respectively. While not statistically significant, IGRP_CD8 T cells from 2.5HIP-NP tolerized grafts trended towards lower diversity (Fig. S4d) and the Jaccard similarity index between all animals indicated more sharing between tolerized animals (Fig. S4e). 2.5HIP NP treatment seemed to restrict the repertoire of IGRP_CD8 T cells to a single shared clonotype.

To further assess TCR repertoire changes induced by tolerance induction, we examined the sequences of expanded clones from either treatment. Weighted logo plot comparisons of CDR3α and CDR3β sequences revealed strong consensus in CDR3α across treatments, and near-consensus in CDR3β except for variation at two central residues (positions 6 and 7) (Fig. 4h). After aligning the CDR3 sequence of the expanded clones, the most abundant clone from control animals, from animal HEL_4, shared all but the two central amino acids (PG) in the CDR3β region with the expanded clone (SQ) found in 2.5HIP-NP tolerized animals (Fig. S4f). To validate the difference in apparent avidity, the HEL_4 TCR and HIP_1 TCR were reconstructed from 5’ scRNA-sequencing data and expressed in immortalized 5KC mouse T cells (gating shown in Fig. S1c). Cells were then stained with the IGRP tetramer versus total TCRβ surface expression. The total level of TCR on the surface was no different between the two TCRs (Fig. 4. h). At a given median total TCR gMFI slice, there is higher binding of the IGRP tetramer to HEL_4 TCR-expressing cells than to the HIP_1-expressing cells. This demonstrates that the expanded clone in HIP-tolerized grafts has a lower affinity than an equivalent, highly expanded clone in HEL control grafts (Fig. 4i, S4g). These data suggest that 2.5HIP-NP tolerance induction limits the repertoire of IGRP_CD8 clones present in treated grafts to a low-avidity population.

IL-10⁺ 2.5HIP_CD4 T cells are expanded in islet grafts following 2.5HIP-NP tolerance induction

To investigate how 2.5HIP-NPs may actively suppress CD8 T cell effector differentiation within islet grafts, we tracked the regulatory fate of graft-infiltrating 2.5HIP_CD4 T cells (Fig. 5a) in response to treatment. We focused on two key regulatory subsets: FOXP3⁺ T regulatory cells (Tregs) and FOXP3⁻CD25⁻ IL-10-producing Tr1-like cells. Tolerance induction with 2.5HIP-NPs significantly increased the frequency of FOXP3⁺ 2.5HIP⁺CD4⁺ T cells within the graft (Fig. 6b), consistent with a trend we previously observed at the same timepoint⁵⁶, though this population comprised only ~10% of the total antigen-specific CD4⁺ T cells.

Fig. 6: IL-10 blockade compromises prolonged graft survival, promotes effector differentiation of antigen-specific CD8+ T cells, and restores CD40 expression on intra-graft cDC1s. — **Fig. 6: IL-10 blockade compromises prolonged graft survival, promotes effector differentiation of antigen-specific CD8⁺ T cells, and restores CD40 expression on intra-graft cDC1s.**

Previous studies employing antigen-specific therapies in NOD mice have demonstrated the critical role of IL-10 producing Tr1-like FOXP3⁻CD25⁻ regulatory T cells in establishing peripheral tolerance⁷². Additionally, Tr1 cells have been associated with positive outcomes in transplantation⁷³. In our previous study⁵⁶, we also observed that the majority of 2.5HIP_CD4 T cells adopted an anergic phenotype, which may give rise to IL10-producing Tr1-like T cells. To assess IL-10 expression in both regulatory subsets following treatment, we utilized the tiger (interleukin-ten ires gfp-enhanced reporter)⁷⁴ mouse backcrossed to the NOD mouse (NOD/Tiger)⁷⁵ for islet transplant and tolerance induction experiments. To detect IL-10^GFP/mRNA without fixation, surrogate surface markers for FOXP3⁺ Tregs (CD25⁺CD127⁻) and conventional CD4 T cells (CD25⁻) were validated (Fig. S5a) and used to phenotype graft-infiltrating 2.5HIP⁺ CD4 T cells following tolerance induction. Diabetic NOD/Tiger mice were treated as before with 2.5HIP-NPs or HEL-NPs and graft-infiltrating 2.5HIP_CD4 T cells were examined for levels of GFP (Fig. 5c). CD4 T cells comprised ~80% of IL-10^GFP/mRNA⁺ cells in islet grafts from both 2.5HIP-NP-tolerized and HEL-NP control mice (Fig. 5d), suggesting they are the primary source of IL-10 at this time point. Graft-infiltrating IL-10⁺CD25^- CD4 T cells expressed higher levels of canonical markers of Tr1 CD4 T cells (Lag3, and CD49b) than their IL-10^- CD25^- CD4 counterparts (Fig. 5e), indicating these cells have a Tr1-like surface phenotype.

When examining these populations in response to treatment, tolerance induction to the 2.5HIP significantly increased the proportion of CD25⁺CD127^- Tregs cells expressing IL-10 (Fig. 5f). 2.5HIP-NP treatment also expanded a population of IL-10⁺ CD25^- Tr1-like CD4 T cells (Fig. 5g). However, as a fraction of all 2.5HIP_CD4 T cells, CD25^-IL-10⁺ Tr1-like cells accounted for nearly half of the total 2.5HIP_CD4 T cells in the graft following treatment, 3 to -4-fold more than their FOXP3⁺ counterparts. This indicates CD25^-IL10⁺ 2.5HIP_CD4 T cells are the predominant regulatory population expanded by 2.5HIP-NP tolerance induction in islet grafts. To replicate these findings in a different model, pre-diabetic NOD/Tiger mice were treated with 2.5HIP- or HEL-NPs to determine if islet-infiltrating 2.5HIP_CD4 T cells exhibited the same increase in IL-10^mRNA/GFP expression (Fig. S5b). Indeed, both CD25⁺CD127^- Treg and CD25^- populations of 2.5HIP_CD4 T cells were expanded in the islets (Fig. S5c, d), lymph nodes (only CD25^- 2.5HIP_CD4 T cells recovered), and spleens of 2.5HIP-NP tolerized animals (Fig. S5d). Taken together, these data demonstrate a robust expansion of IL-10⁺ 2.5HIP_CD4 T cells in 2.5HIP-NP-tolerized grafts, with the majority comprising Tr1-like suppressive cells and a smaller FOXP3⁺ Treg subset.

IL-10 blockade compromises prolonged graft survival, promotes effector differentiation of antigen-specific CD8 T cells, and restores CD40 expression on intra-graft cDC1s

Increased IL-10 production by 2.5HIP_CD4 T cells may affect differentiation of IGRP_CD8 T cell effector fate directly⁷⁶ and/or indirectly via suppression of antigen-presenting cells such as DCs^77,78. To investigate whether IL-10 is required for prolonged graft survival following 2.5HIP-NP tolerance induction, IL-10 was depleted using an αIL-10 monoclonal antibody during the induction of tolerance (Fig. 6a). Blockade of IL-10 abrogated the prolonged graft survival afforded by 2.5HIP-NP-induced tolerance, with the time to disease recurrence mirroring that of HEL-NP isotype control and HEL-NP αIL-10-treated animals (Fig. 6b, c). To determine if IL-10 blockade restored the effector fate of graft-infiltrating, antigen-specific T cells, grafts were isolated at 8-9 days post-transplant from HEL-NP or 2.5HIP-NP tolerized mice and were treated with αIL-10 antibody or an αisotype control Ab. Inhibiting IL-10 during tolerance induction had no impact on the expansion of FOXP3⁺ 2.5HIP_CD4 Tregs in islet grafts (Fig. 6d). There was also no apparent effect on the fraction of TCF^hi 2.5HIP_CD4 T cells (Fig. 6e) or absolute numbers of TCF1^lo terminal effectors (Fig. 6f). By contrast, blockade of IL-10 restored the fraction of IGRP_CD8 T cells to levels similar to those found in HEL-NP isotype control grafts (Fig. 6g). IL-10 blockade reduced the proportion of TCF1^hi IGRP_CD8 T cells to levels seen in HEL-NP controls (~10–20%) (Fig. 6h), which was accompanied by an increase in TCF1⁻ effector cell numbers (Fig. 6i). The elevated ISG signature (measured by Ly6A surface expression) (Fig. 6j, k) and relative TCR avidity (measured by normalized IGRP staining) (Fig. 6l) were also restored to control levels upon IL-10 blockade in IGRP_CD8 T cells. Collectively, these data indicate that IL-10 is essential for arrest of IGRP_CD8 T cell effector differentiation induced by 2.5HIP tolerance, but that this effect occurs downstream of both the expansion of FOXP3⁺ 2.5HIP_CD4 regulatory T cells and the arrest of graft-infiltrating effector 2.5HIP_CD4 T cells.

Given the reduced ISG signature, altered TCR repertoire of IGRP_CD8 T cells, and the known inhibition of antigen presentation by IL-10^77,78, we next asked whether 2.5HIP-NP therapy impairs APC licensing, thereby indirectly limiting CD8 T cell effector fates. Professional APCs, especially CD4-licensed dendritic cells, are necessary for robust CD8 effector differentiation^79,80. CD4 T cell-mediated licensing of DCs by CD40-CD40L ligation is critical to achieve full cytolytic potential in CD8 T cells^79,81. To determine if 2.5HIP-NP tolerance induction had any impact on APC populations, prediabetic NOD mice (<12 weeks old) were treated with 2.5HIP- or HEL-NPs, and APC subsets were examined in pancreatic islets and the pancreatic lymph node (pLN) (Fig. S6a, gating strategy in Fig. S6b). Flow cytometry revealed there was a significant decrease in CD40 levels exclusively on migratory cDC1s (CD11c MHCII^hiXCR1⁺) in the pancreatic LN of 2.5HIP-NP tolerized mice but not in resident cDC1s or migratory cDC2s (Fig. S6d). These cells likely migrated from the islets, indicating that 2.5HIP-NP tolerance induction suppresses cDC1 activation in islets.

Next, we assessed whether DCs tolerized in situ by 2.5HIP-NPs can independently modulate IGRP_CD8 T cell activation. DCs were isolated from the islets and spleens of NOD.scid mice adoptively transferred with 2.5HIP-specific transgenic T cells (BDC-2.5 TCR-Tg) and tolerized with either 2.5HIP- or HEL-conjugated nanoparticles (Fig. S7a). When co-cultured with transgenic IGRP-specific CD8 T cells (NY8.3), islet-derived DCs from either treatment were able to activate NY8.3T cells based on the increased proportion of CD25⁺CD69⁺ double-positive T cells (Fig. S7b). However, there was a significant decrease in the gMFI of CD69 and Ly6A on NY8.3 T cells incubated with islet-derived DCs from 2.5HIP-NP-treated animals (Fig. S7c, d). This difference was not observed with spleen-derived DCs from either treatment, indicating that 2.5HIP-NP tolerance induction can specifically impact the ability of intra-islet DCs to activate IGRP-specific T cells and express Ly6A.

Finally, we determined if DCs are suppressed by 2.5HIP tolerance induction in islet grafts and whether this suppression is IL-10 dependent. The cDC1 and cDC2 subsets (Fig. 6m) were examined in the grafts of HEL-NP or 2.5HIP-NP-tolerized mice, with or without IL-10 blockade at day 9 post-transplant. Tolerance induction selectively reduced the fraction of CD40^hi intra-islet cDC1s but not cDC2 cells (Fig. 6n). Following IL-10 blockade, CD40^hi cDC1s were restored to levels found in control grafts. In summary, these findings demonstrate that 2.5HIP-specific tolerance reshapes the graft microenvironment by inducing IL-10-dependent suppression of DC activation, thereby impairing the ability of DCs to license CD8 T cells and promote ISG/cytolytic differentiation.

Fate of T cells and dendritic cells during tolerance and following graft destruction in 2.5HIP NP treated mice

While therapeutic induction of tolerance with 2.5HIP NPs during islet transplant prolongs graft survival (~45 days), and continued treatment extends this protection (~90 days) (Fig. 2c), eventually the grafts are destroyed. In order to understand where tolerance breaks, antigen-specific T cells were examined in tolerized grafts and the draining lymph node at a time point when treated grafts are protected (day 15) and again upon graft failure (~day 45) (Fig. 7a).

**Fig. 7: Fate of T cells and dendritic cells during tolerance and following graft destruction.**

In the kLN, we observed a numeric accumulation of 2.5HIP_CD4 T cells at d15 (Fig. 7b), higher than the number at graft failure and d9 (Fig. S2b). This was true for 2.5HIP tetramer-negative CD4 T cells as well (Fig. S8a) and continued a trend observed in tolerized grafts at d9 (Fig. S2b). At d15 kLN 2.5HIP_CD4 T cells were similar or less TCF1^hi compared to the same cells at graft failure (Fig. 7c). Conversely, in the graft, there was a non-significant increase in the total number of 2.5HIP_CD4 T cells upon graft failure compared to d15 (Fig. 7b), as with 2.5HIP tetramer-negative CD4 T cells (Fig. S8a). Interestingly, upon graft failure, 2.5HIP_CD4 T cells were equally or more TCF1^hi compared to the equivalent cells at d15 (Fig. 7c). IGRP_CD8 T cells exhibited a very similar phenotype to the 2.5HIP_CD4 T cells at both timepoints, albeit muted in magnitude. At d15, there was a trending accumulation of IGRP_CD8 T cells in the kLN (Fig. 7d), with a significant increase for IGRP tetramer-negative CD8 T cells (Fig. S8a). Most of the IGRP_CD8 T cells in the kLN were TCF1^hi (Fig. 7e) at either time point. Upon graft failure, there was a near-significant increase of IGRP_CD8 T cells in the islet graft compared to d15 (Fig. 7d); this was also the case for IGRP tetramer-negative CD8 T cells (Fig. S8a). At graft failure, IGRP_CD8 T cells were significantly more TCF1^hi than at d15. Collectively, there was an accumulation of tetramer-positive and tetramer-negative T cells in the kLN at d15 when grafts were protected. Upon graft destruction, there was an increase in the number of antigen-specific and non-specific CD4 and CD8 T cells in the graft, which may indicate epitope spread to T cells of other specificities, including insulin B-chain. Interestingly, TCF1^hi cells were more abundant in antigen-specific CD4 and CD8 T cells upon graft failure. This is possibly the result of the accumulation of stem-like TCF1^hi cells during the time that the graft is still protected. Tolerance may induce a helpless fate⁸² in CD8 T cells that promotes memory formation at the expense of effector function, resembling a tertiary lymphoid structure at the graft.

To determine if treatment-induced regulatory 2.5HIP_CD4 T cells were maintained in the tissue, we assessed the proportion of FOXP3⁺ 2.5HIP_CD4 T cells at the d15 tolerance point and upon graft failure. While there was no difference in percentage of FOXP3⁺ 2.5HIP_CD4 T cells in the kLN, there was a significant decrease in regulatory 2.5HIP_CD4 T cells in the graft upon failure (Fig. 7f), returning to levels found in d9 control grafts (Fig. 5b). This indicates the graft regulatory landscape induced by 2.5HIP NP tolerance induction is eroded upon graft failure, with the regulatory cells likely dependent on continued exogenous antigen-NP treatment. Finally, dendritic cells were assessed to determine if CD40^hi DC are suppressed at d15 and return upon graft failure. At d15, the percentage of CD40^hi cDC1s was similar to that found at d9 tolerized grafts, but returned to significantly higher levels upon graft failure, similar to those found in d9, control grafts (Fig. 6n). Without continued active suppression by regulatory 2.5HIP_CD4 T cells, intra-graft CD40^hi cDC1s are re-established, driving functional antigen-presentation to CD8 T cells within the islet graft. We posit the graft ultimately fails due to lack of treatment-dependent, regulatory 2.5HIP_CD4 T cells in the graft, allowing for effective antigen-presentation to CD8 T cells of multiple specificities, eventually reaching a threshold of sufficient number and function of cytotoxic T cells that destroy the graft.

The mechanistic basis of the dominant tolerance observed in treated islet grafts, whereby CD4 antigen-specific therapy restrains CD8 T cell effector differentiation by limiting effective antigen presentation by cDC1s, is summarized in the graphical abstract (Fig. 8).

**Fig. 8: Graphical abstract. The comparison of antigen-specific T cell differentiation in control HEL NP and tolerized 2.5HIP NP treatment settings.**

Discussion

The goal of antigen-specific immunotherapy is to induce tolerance by restraining pathogenic T cells and thereby protecting pancreatic or transplanted islets in the absence of broad immunosuppression. In the autoimmune islet graft model, antigen-specific 2.5HIP_CD4 and IGRP_CD8 T cells exhibited a differentiation trajectory from TCF1^hi stem-like precursors in the draining lymph node to TCF1^lo terminal effectors within the islet graft. Antigen-specific immunotherapy with 2.5HIP-NPs prevented this effector transition in both 2.5HIP_CD4 and IGRP_CD8 T cells infiltrating the graft and transiently prolonged graft survival, although this survival could be greatly extended with continued HIP-NP treatment. Notably, tolerance induction prevented IGRP_CD8 T cells from adopting an ISG/cytolytic transcriptional program and restricted their clonal repertoire to a low-diversity, public pool of TCRs with lower avidity relative to controls. In parallel, 2.5HIP-NP treatment drove a marked expansion of regulatory IL-10-producing CD4 T cells, including both FOXP3⁺ Tregs and CD25⁻IL-10⁺ Tr1-like 2.5HIP_CD4 T cells. IL-10 was required for the prolonged graft survival, attenuation of IGRP_CD8 differentiation, and reduced cDC1 licensing conferred by 2.5HIP-NP tolerance induction. Collectively, these data help define how peripheral tolerance induced by a single dominant CD4 neoepitope constrains the effector fate of autoreactive CD8 T cells in islet grafts.

A growing consensus from studies in infection and cancer holds that antigen-specific T cells follow a stem-like to effector differentiation fate as they transit from lymphoid niches into antigen-rich tissues^83,84. In autoimmune diabetes, this same fate transition occurs, where CD8 T cells occupy a TCF1^hi reservoir in pancreatic lymph nodes before differentiating into TCF1^lo effectors in islets. Recent studies extended this paradigm and identified a stem-like niche of TCF1^hi 2.5HIP_CD4 T cells in pancreatic LNs and the islets of NOD mice⁸⁵. We now show that in syngeneic islet grafts, both 2.5HIP-specific CD4 and IGRP-specific CD8 T cells traverse the same TCF1^hi-to-TCF1^lo trajectory, indicating a unified differentiation program and tissue compartmentalization for autoreactive T cells in β-cell autoimmunity.

Induction of peripheral tolerance to the 2.5HIP epitope was able to arrest functional effector fates in both 2.5HIP_CD4 and IGRP_CD8 graft-infiltrating T cells. Further interrogation of IGRP_CD8 T cells revealed a collapse of type I interferon and cytolytic signatures, with fate trajectories terminating short of the hypo-functional arrested or restrained subset. A similar population of LAG^hi IGRP_CD8 T cells has been described elsewhere⁴⁹. Nearly one-third of downregulated genes were canonical ISGs, and module scoring confirmed broad suppression of both ISG and granzyme pathways. Induction of tolerance by 2.5HIP-NPs likely interrupts T cell effector maturation at a critical inflection in the graft, preventing functional differentiation of ISG/cytolytic signature CD8 T cells and instead promoting an alternative, hypofunctional fate in tissue. The TCR repertoire of tolerized IGRP_CD8 T cells had a lower apparent avidity, dominated by a single, public TCR. We postulate this may arise for a number of reasons: high-avidity clones undergo activation-induced cell death after suboptimal antigen-encounter from APCs in tolerized grafts⁸⁶, low-avidity clones may preferentially expand and persist in tolerized grafts as has been shown in tumors⁸⁷, or it is possible that high-affinity clones are suppressed in lymph nodes and fail to ever migrate into tolerized grafts.

The induction of tolerance to the 2.5HIP epitope also expanded a population of IL-10-producing CD4 regulatory T cell subsets within grafts, including Foxp3⁺ Tregs and CD25-negative Tr1-like cells. Santamaria and colleagues demonstrated that antigen-specific interventions in NOD mice shift CD4 differentiation from Tfh precursors to IL-10-producing Tr1 cells⁸⁸. Observing a similar fate of 2.5HIP_CD4 T cells in our system, we predict that IL-10 may act directly and/or indirectly on the fate of CD8 T cells. Directly, IL-10 could engage the IL-10 receptor on IGRP_CD8 T cells, limiting proliferation, and raising the threshold for TCR signaling mechanisms known to blunt effector differentiation of CD8 T cell populations⁷⁶. While IL-10 has been shown to be important for memory formation⁸⁹, we believe it mainly acts to limit effector differentiation of graft-infiltrating antigen-specific CD8 T cells that are fated for terminal differentiation.

IL-10 was originally described as a cytokine acting on antigen-presenting cells, suppressing their capacity to support Th1 responses⁹⁰. Ligation of the IL-10 receptor downregulates co-stimulatory molecules on dendritic cells, such as CD40⁹¹, effectively limiting their ability to license CD8 T cells. The suppression of APCs by regulatory CD4 T cells has been thought of as a cellular intermediary, conferring tolerance to other specificities^92,93,94. We propose in our system, that 2.5HIP_CD4-derived IL-10 acts indirectly by suppressing cDC1 licensing in grafts, thereby limiting IGRP_CD8 T cell differentiation and function in the target tissue. Suppression of CD40^hi cDC1s was also observed in the pLN of prediabetic NOD mice treated with 2.5HIP NPs (Fig. S6d), indicating this mechanism of suppression may function during normal disease progression and potentially in response to alloantigen as well. 2.5HIP_CD4 Tr1-like cells may also directly kill cDC1s in the islet graft, as has been shown in tumor-infiltrating, neoantigen-specific Tr1 CD4 T cells⁹⁵.

Taken together, this study helps explain how therapeutic tolerance induction with a CD4 neoepitope restrains the differentiation fate and function of high-affinity IGRP_CD8 T cells through a dendritic cell intermediate. This form of dominant tolerance shows promise for human therapy, wherein the identification and induction of tolerance to key epitopes may be sufficient to protect grafts and lead to the tolerance of other islet-specific T cells. These findings raise further questions about how exactly suppression operates in the graft microenvironment. Must CD4 and CD8 T cells engage the same DC in a cellular triad to establish linked suppression, or is bystander paracrine IL-10 production sufficient? What is the relative impact, if any, of direct IL-10r signaling on CD8 T cells? How do T cells of other antigen specificities play a role in tolerance induction and the eventual loss of tolerance following treatment cessation? Why do T cells transiently accumulate in the kLN of protected grafts? In tolerized grafts that eventually fail, there is an accumulation of stem-like TCF1^hi T cells at the graft site over time. Why do these memory T cells still accumulate at the graft, and can their infiltration also be stopped? How can durable antigen-specific tolerance be achieved with limited treatment duration in the absence of indiscriminate immunosuppression? Could continued treatment sustain graft survival for the lifetime of the recipient? Addressing these questions will be essential for refining strategies to induce peripheral tolerance through the targeting of CD4 hybrid insulin peptide neoepitopes and, in general, to inform antigen-specific immunotherapies in transplantation to eventually replace broad immunosuppression.

Methods

Animal studies

All animal study experimental designs, treatments, husbandry, group numbers, statistical cut-offs, and power calculations were submitted to the IACUC at the University of Colorado Anschutz Medical Campus for approval prior to beginning experiments. All mice were maintained and bred in a specific pathogen-free facility at the University of Colorado Anschutz Medical Campus Vivarium (Aurora, CO, USA) (12/12 light/dark cycles, temperature 18-26 °C, humidity 30-70%). Experimental and control animals were co-housed, and littermates were used for different treatments when possible. Mice were euthanized using carbon dioxide (CO₂) inhalation followed by a secondary method of cervical dislocation in accordance with institutional guidelines. NOD.ShiLtJ (Jackson lab #001976), NOD.scid (Jackson Lab #001303), and NOD.BDC-2.5 TCR-transgenic (Tg/BDC-2.5) (Jackson Lab #004460) mice were obtained from The Jackson Laboratory for breeding, and new breeders were ordered every 4–5 generations. NOD/tiger (interleukin-ten ires gfp-enhanced reporter⁷⁴ (Jackson Lab #034192) backcrossed onto the NOD.ShiLtJ background for >10 generations) and NOD/NY8.3 TCR-transgenic (Tg/NY8.3) (Jackson Lab #005868) breeders were obtained from Dr. Rachel Friedman and were bred as heterozygous animals with NOD.ShiLtJ mice from The Jackson Laboratory. NOD/BDC-2.5 TCR-transgenic (Tg/BDC-2.5) mice were also bred heterozygously with NOD.ShiLtJ. All mice were maintained on standard chow and water ad libitum except NOD.scid mice, which were maintained on a 50/50 mixture of standard chow and antibiotic chow ad libitum to prevent infection in the immunocompromised strain.

Polyclonal NOD and NOD/Tiger mice were monitored for diabetes by weekly random urine glucose measurements; high readings were confirmed for hyperglycemia by tail vein blood glucose readings >250 mg/dl. Following the onset of hyperglycemia, mice were maintained on daily exogenous insulin, Humulin R (0.5-1U/day) until transplant to prevent overt metabolic disease. On the indicated days, mice were transplanted by inserting under the kidney capsule 500 islet equivalents (~300 individual islets) from multiple NOD.scid donor mice. A single experiment is defined as the diabetic recipient mice receiving an islet transplant on the same day from the same pool of NOD.scid islets. For transplant surgery, mice were given Buprenorphine SR (1.0 mg/kg by body weight) analgesic subcutaneously, ~30 minutes prior to surgery. Under anesthesia (isoflurane, 3–5% induction maintained with 1–2.5% oxygen), an incision was made to expose the left kidney. An opening was made in the kidney capsule with a needle, and the islets were expelled beneath the capsule. The retroperitoneal wall was closed with dissolving sutures, and the skin was closed with autoclips. Animals were closely monitored for recovery in a temporary cage on a heating pad until determined to be bright, alert, and responsive. If needed, post-op fluid supplementation (0.5–1 mL saline solution) was given subcutaneously to facilitate recovery. After transplanting, mice were monitored by random blood glucose measurements for recurrence of disease (>250 mg/dl). For IL-10 antibody depletion experiments, diabetic mice were injected intraperitoneally with 100 μg/dose of αIL-10 mAb (JES5-2A5 or an isotype control, Bioxcell) on the indicated days. Treatment with αIL-10 was given 4 hours before NP treatment if given on the same day.

For injections of nanoparticles and transgenic T cells, nanoparticles or cells were resuspended and filtered with 45μm mesh before retro-orbital intravenous (i.v.) injection into mice under isoflurane anesthesia. For islet transplant experiments, diabetic female mice were used as recipients based on the sex-specific increased disease incidence. For prediabetic and transgenic adoptive transfer experiments, age-matched (8–12-weeks-old) male mice were used as donors.

Nanoparticle synthesis, peptide conjugation, and treatment

Biodegradable poly(lactic-co-glycolic acid) nanoparticles (NPs) were provided by the Miller lab, generated using a single emulsion followed by solvent evaporation as previously described²⁰. Nanoparticles were then washed three times with PBS and conjugated covalently to a soluble form of the 2.5HIP (RGG-LQTLALWSRMD-GGR) or the HEL (AMKRHGLDNYRGYSL) peptide by incubating with 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide (ECDI) and then washed three times again. Nanoparticles were prepared fresh for each experiment and immediately administered by retro-orbital i.v. injection (2.5 mg of nanoparticles in 200 μl of sterile PBS) on the indicated days.

Flow cytometry MHC-tetramer and antibody staining

For tetramer and antibody staining for flow cytometry analysis, single-cell suspensions were generated from various organs. Lymph nodes and spleen were minced with scissors and filtered through a 70 μm cell strainer with RBC lysis if needed. Grafts were resected from the kidney and digested in Dispase II (Gibco 17105-041) at 2 U/ml for 7–10 minutes with intermittent mixing by pipetting. Following digestion, graft suspensions were also filtered through a 70 μm cell strainer. After a wash, cells were resuspended in complete media and stained with MHC tetramers at 37 °C for 20 minutes with the 2.5HIP I-Ag7 tetramer or IGRP_(206-214) H2-Kd tetramer in PE or APC. Surface antibody cocktails were then added to the tetramer stain and incubated for a further 40 minutes. The following antibodies were used: rat IgG2b anti-mouse CD3 APC/Fire 750, clone:17A2, Biolegend, cat #100267, 1:400; mouse IgG1 anti-rat RT1B (MHCII) FITC, clone:OX-6, Biolegend, cat #205305, 1:600; mouse IgG2b anti-mouse/rat XCR1 APC, clone:ZET, Biolegend, cat #148205, 1:200; rat IgG1 anti-mouse CD172a (SIRPα), PE clone:P84, Biolegend, cat #124629, 1:400; rat IgG2a anti-mouse F4/80 BV711, clone:BM8, Biolegend, cat #103057, 1:200; rat IgG2a anti-mouse CD40 PE/Dazzle 594, clone:3/23, Biolegend, cat #124629, 1:400; Armenian hamster IgG anti-mouse CD11c BV705, clone: N418, Biolegend, cat #117333, 1:200; Armenian hamster IgG anti-mouse CD11c BB700, clone:HL3, BD Biosciences, cat #745899; rat IgG2a anti-mouse CD19 BB700, clone:1D3, BD Biosciences, cat #566412; rat IgG2b anti-CD11b BB700, clone:M1/70, BD Biosciences cat #566416; rat IgG2a anti-mouse Ly6G BB700, clone:1A8, BD Biosciences cat #566435; rat IgG2a anti-mouse CD5 PE, clone:53-7.3, Biolegend, cat #100607, 1:400; rat IgG2a anti-mouse Ly6A/E BV421, clone:D7, Biolegend, cat #108127, 1:400; Armenian hamster IgG anti-mouse PD1 BUV615, clone:J43, BD Biosciences, cat #752299, 1:200; hamster anti-mouse CD279 (PD-1) BUV615, clone:J43, Biolegend 1:200; rat IgG2a anti-mouse CD8a PerCP-eFluor710, clone:53-6.7, eBioscience, cat #46-0081-82, 1:400; rat IgG2a anti-mouse CD8a BB515, clone:53-6.7, BD Biosciences, cat #564459; recombinant human IgG1 anti-human/mouse TOX PE, clone:REA473, Miltenyi Biotec, cat #130-120-785, 1:200; rat IgG1 anti-mouse CD25 BV650, clone:PC61, Biolegend, cat #102038, 1:200; rat IgG2a anti-mouse CD127 PE Cy7, clone:A7R34, Biolegend, cat #135013, 1:400; rat IgG1 anti-mouse LAG3 BV786, clone:C9B7W, Biolegend, cat #125219; rat IgG2a anti-mouse CD4 BV605 and BV421, clone:RM4-5, Biolegend, cat #100547, #116023, 1:400; rat IgG2a anti-mouse CD366 (TIM3) BV785, clone:RMT3-23, Biolegend, cat #119725, 1:200; rat IgM anti-mouse CD49b APC Fire750, clone:DX5, Biolegend, cat #108925, 1:200; rat IgG2a anti-mouse/human CD44 BV771, clone:IM7, Biolegend, cat #103057, 1:400; rat IgG2a anti-mouse CD62L FITC, clone:MEL-14, Biolegend, cat #104406, 1:400; rat IgG2a anti-mouse Ki67 FITC, clone:SolA15, Biolegend, cat #151211, 1:600; rabbit IgG anti-TCF1/TCF7 AlexaFluor 647, clone:E601K, Cell Signaling Technologies, cat #83268S, 1:400; rabbit IgG anti-TCF1/TCF7 AlexaFluor488, clone:E601K Cell Signaling Technologies, cat #50639S, 1:400; rat IgG2a anti-mouse FOXP3 eFluor, 450 clone:FJK-16s, Invitrogen, cat #48-5773-82, 1:200; rat IgG2a anti-mouse IFNγ eFluor, 450 clone:XMG1.2, eBioscience, cat #48-7311-80 1:400; rat IgG1 anti-mouse, TNF PE Cy7, clone:MP6-XT22, Biolegend cat #506323, 1:400; hamster anti-mouse TCRβ BV605, clone:H57-597, Biolegend, cat #109241, 1:400; and rat IgG2a anti-mouse CD16/32 (Fc-block), clone:93, Biolegend, cat #101302, 1:500.

Cells were washed, stained with viability dye in PBS, and fixed with the fixation buffer from the FOXP3 eBioscience Intracellular Staining Kit (#00-5523-00). Following overnight fixation, cells were permeabilized and stained for intracellular targets. After two hours, cells were washed and resuspended in FACS buffer (2% FBS, 1 mM EDTA in PBS). Cells were then run on a 5-laser Cytek Aurora Spectral flow Cytometer and analyzed using Flowjo v10.10. To ensure reproducibility, cells were stained in the same antibody concentrations, incubated for the same period of time, and analyzed on the same cytometer.

5’ scRNA-sequencing of IGRP_CD8 T cells

IGRP_CD8 T cells were sorted by FACS from day 8 islet grafts (lymphocytes, live, CD45+, lineage-(CD11c, CD11b, CD19, Ly6G, CD4), IGRP tetramer+). After sorting, cells from individual animals were barcoded with hashtag antibodies to deconvolute after sequencing (Biolegend: TotalSeq -C0301 anti-mouse Hashtag 1 ACCCACCAGTAAGAC cat #155861, TotalSeq -C0301 anti-mouse Hashtag 2 GGTCGAGAGCATTCA cat #155863, TotalSeq -C0301 anti-mouse Hashtag 3 CTTGCCGCATGTCAT cat #155865, TotalSeq -C0301 anti-mouse Hashtag 4 AAAGCATTCTTCACG cat #155867 all 1:100). Single cell gene expression profile, cell surface proteomics profile, and T cell receptor (TCR) profile were obtained using the Chromium Single-Cell 5’ NextGEM Library and Gel Bead Kit from 10X Genomics. Sorted and hashed cells were partitioned on the 10X Genomics Chromium X instrument for single cell capture, followed by amplification of cDNA and library preparation to recover 10,000 - 12,000 cells per sample. Individual libraries corresponding to gene expression, cell surface protein, and TCR were quantified, pooled, and sequenced on the Illumina NovaSeq 6000 sequencer at a depth of 50,000 reads/cell for gene expression, 5000 reads/cell for cell surface protein, and 5000 reads/cell for TCR, respectively. Following sequencing, standard quality control, filtering, and normalization were applied using default CellRanger pipelines. TCR sequences were annotated using the scRepertoire package. All cells from both treatments were used to generate a UMAP reduction. Clustering was then performed using SNN nearest-neighbor analysis while masking TCR genes from the clustering. Pseudobulk differential gene expression (DEG) analysis was used to generate the Volcano plots with the ProjecTIL R package to compare between treatments after excluding TCR genes. Gene set enrichment analysis (GSEA) was run on differentially expressed genes (also after excluding TCR genes) using the fgsea R package from the molecular signature database (MSigDB) C2 curated gene sets, subcategory CGP, with a normalized enrichment score NES cutoff of 1 and adjusted p-value (q) of 0.05. All packages required for data processing and analysis and version number: scProportionTest (0.0.4), ggseqlogo (0.2), immunarch (0.9.1), tidyverse (2.0.0), SingleCellExperiment (1.30.1), Seurat (5.3.0), Matrix (1.6-5), scale (1.2.1), cowplot (1.2.0), RCurl (1.98-1.16), AnnotationHub (3.16.1), AnnotationDbi (1.64.1), ensembldb (2.26.0), harmony (1.2.3), ggpubr (0.6.0), gridExtra (2.3), clusterProfiler (4.10.1), Startrac (0.1.0), ggplot2 (4.0.0), msigdbr (7.5.1), scales (1.4.0), fgsea (1.28.0), tradeSeq (1.22.0), viridis (0.6.5), SignatuR (0.3.0), UCell (2.8.0), circlize (0.4.16), ggrepel (0.9.5), ggsci (3.2.0), gprofiler2 (0.2.3), MetBrewer (0.2.0), wesanderson (0.3.7), scRepertoire (2.4.0), SCpubr (2.0.0), dittoSeq (1.16.0), EnhancedVolcano (1.20.0), SeuratExtend (1.2.5), Nebulosa (1.14.0), impute (1.78.0), ComplexHeatmap (2.24.1), sscVis (0.4.0), ggsignif (0.6.4), Biostrings (2.76.0), dplyr (1.1.4), SeuratWrapper (0.4.0), monocle3 (1.4.26), vegan (2.8.0), psych (2.4.6), ProjecTIL (2.4.1), metap (1.11).

TCR reconstruction and expression on TCR^ko 5KC mouse T cell line

The top expanded IGRP-specific clones from each treatment (HEL_4, HIP_1) were reconstructed and expressed in the immortalized, TCR^ko 5KC mouse T cell line. Briefly, exact nucleotide sequences were reconstructed from the standard 10X Genomics 5’ scRNAseq TCR library for each of the alpha and beta of each TCR. Sequences from 5 separate barcoded cell IDs were aligned for each TCR to ensure correct and consensus sequences. The alpha and beta chains were connected by a P2A and cloned into the pMIGII⁹⁶ vector (addgene #52107) by restriction enzymes and grown up in chemically competent E. coli. Gamma-retroviral stocks containing the IGRP^TCR-pMIGII plasmid were packaged using Platinum-E cells. 5KC were transduced using polybrene spinfection. Resulting cells were cultured in Iscove’s Modification of DMEM with 10% fetal calf serum, 1% pen/strep, and 50μM 2-ME at 37 °C in 5% CO₂. Cells were split daily and assessed after 5 days by flow cytometry for IGRP tetramer staining and total surface TCR expression.

Tg/BDC-2.5 T cell adoptive transfer, NP treatment, and dendritic cell isolation

BDC-2.5 transgenic CD4 T cells were isolated from the spleen and lymph nodes (pancreatic, inguinal, axillary, submandibular) by negative selection (CD4 mojo sort Biolegend). Single cell suspensions of T cells were stimulated in complete RPMI media (10% FBS, 1% pen/strep, Na pyruvate, glutamine, 2-ME, and NEAA) with 20 U/ml recombinant murine IL2 (Peprotech # 212-12-20UG) and αCD3/CD28 coated-beads (Dynabeads Thermo Fischer #11452D) at a 1:2 ratio (T cell:bead). Following three days of stimulation, beads were removed, and cells were washed and counted. Tg/BDC-2.5 T cells (2.5 million) were adoptively transferred IV to 8-12-week-old NOD.scid recipients. One day after adoptive cell transfer, mice received 2.5 mg of either HEL- or 2.5HIP-NPs i.v.; four days later, mice were euthanized and dendritic cells isolated by negative selection (EasySep pan-DC #19863) from islets and spleen.

DC and Tg/NY8.3 T cell co-culture

Isolated, primary DCs were incubated in complete RMPI and pulsed with 2ug/ml IGRP_206-214 for 4 hours. DCs were washed 4 times before incubation with Tg/NY8.3 T cells to remove residual peptide. Tg/NY8.3 were isolated by CD8 negative selection (Biolegend #480008) and stimulated as Tg/BDC-2.5 T cells with the addition of recombinant murine IL7 (PeproTech # 217-17-50UG) for three days. Beads were removed, and Tg/NY8.3 T cells were washed two times and incubated with unpulsed or peptide-pulsed DCs isolated from the spleen or islets of HEL- or 2.5HIP-NP-treated mice at a 1:1 ratio.

Statistical analysis

Sample sizes were determined based on previous tolerance and transplant studies as well as preliminary experiments, and on an expected effect size of 1.2 (difference in means standardized by pooled standard deviation), a significance level (α) of 0.05, and a desired power of 80% (β = 0.2). Power analysis was performed using G*Power v3.1.9.7. Data collection was stopped after an appropriate number of n were analyzed based on the aforementioned power calculations. Data from transplanted animals were excluded if the recipient mice did not normalize to <250 mg/dl within two days post-transplant. Flow cytometry data with 200 events of intact cells or fewer were excluded from analysis. Transplant experiments were repeated 3-6 times with 1-3 transplanted animals per experiment. Experiments with prediabetic and transgenic mice were repeated at least 2 times with 3-4 animals per experiment. Data are presented as mean ± s.e.m. for continuous variables or mean with box and whisker plots for interquartile range (IQR) and 1.5*IQR, respectively. Analysis was performed in GraphPad Prism v10.4.2 or RStudio version 2023.12.1 + 402. Shaprio-Wilk test was used to test for normality in n < 20 datasets. Statistical significance was determined by a non-parametric unpaired two-tailed T test for comparison of a single continuous variable between two groups with n < 20. For a continuous variable with n > 20, the Wilcoxon rank-sum test was used with Bonferroni correction if necessary. The Kaplan-Meier method and the Multiple Log-Rank test were used to determine statistical significance for survival analysis with Bonferroni correction for three groups or more. One-way analysis of variance (ANOVA) was computed to test statistical differences between two or more groups. Levene’s test was used to determine the variances between multiple groups. Šídák’s multiple comparison test was computed when performing pairwise-comparison between cell subsets of different treatments (i.e., TCF1^lo IGRP_CD8 in 2.5HIP-NP vs HEL-NP). Tukey Honest Significant Differences was used to compare multiple means from different treatment groups (i.e., HEL-NP αisotype vs 2.5HIP-NP αisotype vs 2.5HIP-NP αIL10). Details, statistical tests, and analysis for specific experiments can be found in the figure legends. The alpha value for the threshold of significance was 0.05, including corrected p-values.

Declaration of generative AI and AI-assisted technologies

During the preparation of this work, the authors used ChatGPT in order to clarify prewritten text. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The single-cell RNA-seq data generated in this study have been deposited in the GEO database under accession code GSE299087. The processed and annotated.rds file (CD8_C.rds) is available for ease of analysis at Figshare [https://doi.org/10.6084/m9.figshare.31255354], and treatments are identified by ‘orig.ident’. Source data are provided with this paper.

Code availability

All code used to analyze the scRNA-sequencing for this manuscript has been deposited at Figshare [https://doi.org/10.6084/m9.figshare.31255354].

References

In’t Veld, P. Insulitis in human type 1 diabetes: a comparison between patients and animal models. Semin. Immunopathol. 36, 569–579 (2014).
Article PubMed PubMed Central Google Scholar
Pugliese, A. Autoreactive T cells in type 1 diabetes. J. Clin. Invest. 127, 2881–2891 (2017).
Article PubMed PubMed Central Google Scholar
Wang, S. et al. Transplantation of chemically induced pluripotent stem-cell-derived islets under abdominal anterior rectus sheath in a type 1 diabetes patient. Cell 187, 6152–6164.e6118 (2024).
Article CAS PubMed Google Scholar
Keymeulen, B. et al. Encapsulated stem cell-derived beta cells exert glucose control in patients with type 1 diabetes. Nat. Biotechnol. 42, 1507–1514 (2024).
Article CAS PubMed Google Scholar
Reichman, T. W. et al. Stem cell-derived, fully differentiated islets for Type 1 diabetes. N. Engl. J. Med. https://doi.org/10.1056/NEJMoa2506549
Hu, X. et al. Hypoimmune islets achieve insulin independence after allogeneic transplantation in a fully immunocompetent non-human primate. Cell Stem Cell 31, 334–340.e335 (2024).
Article CAS PubMed Google Scholar
de Mattos, A. M., Olyaei, A. J. & Bennett, W. M. Nephrotoxicity of immunosuppressive drugs: long-term consequences and challenges for the future. Am. J. Kidney Dis. 35, 333–346 (2000).
Article PubMed Google Scholar
Roberts, M. B. & Fishman, J. A. Immunosuppressive Agents and Infectious Risk in Transplantation: Managing the “Net State of Immunosuppression”. Clin. Infect. Dis. 73, e1302–e1317 (2021).
Article PubMed PubMed Central Google Scholar
Gutierrez-Dalmau, A. & Campistol, J. M. Immunosuppressive therapy and malignancy in organ transplant recipients: a systematic review. Drugs 67, 1167–1198 (2007).
Article CAS PubMed Google Scholar
Peakman, M. & Dayan, C. M. Antigen-specific immunotherapy for autoimmune disease: fighting fire with fire? Immunology 104, 361–366 (2001).
Article CAS PubMed PubMed Central Google Scholar
Buckner, J. H. Antigen-specific immunotherapies for autoimmune disease. Nat. Rev. Rheumatol. 21, 88–97 (2025).
Article CAS PubMed Google Scholar
Est-Witte, S. E., Livingston, N. K., Omotoso, M. O., Green, J. J. & Schneck, J. P. Nanoparticles for generating antigen-specific T cells for immunotherapy. Semin. Immunol. 56, 101541 (2021).
Article CAS PubMed PubMed Central Google Scholar
Kreiner, F. F., von Scholten, B. J., Coppieters, K. & von Herrath, M. Current state of antigen-specific immunotherapy for type 1 diabetes. Curr. Opin. Endocrinol. Diab Obes. 28, 411–418 (2021).
Article CAS Google Scholar
Firdessa-Fite, R. et al. Soluble antigen arrays provide increased efficacy and safety over free peptides for tolerogenic immunotherapy. Front. Immunol. 15, 1258369 (2024).
Article CAS PubMed PubMed Central Google Scholar
Gammon, J. M. et al. Engineering the lymph node environment promotes antigen-specific efficacy in type 1 diabetes and islet transplantation. Nat. Commun. 14, 681 (2023).
Article ADS CAS PubMed PubMed Central Google Scholar
Poon, I. K., Lucas, C. D., Rossi, A. G. & Ravichandran, K. S. Apoptotic cell clearance: basic biology and therapeutic potential. Nat. Rev. Immunol. 14, 166–180 (2014).
Article CAS PubMed PubMed Central Google Scholar
Steinman, R. M., Turley, S., Mellman, I. & Inaba, K. The induction of tolerance by dendritic cells that have captured apoptotic cells. J. Exp. Med. 191, 411–416 (2000).
Article CAS PubMed PubMed Central Google Scholar
Getts, D. R., McCarthy, D. P. & Miller, S. D. Exploiting apoptosis for therapeutic tolerance induction. J. Immunol. 191, 5341–5346 (2013).
Article CAS PubMed PubMed Central Google Scholar
Prasad, S., Kohm, A. P., McMahon, J. S., Luo, X. & Miller, S. D. Pathogenesis of NOD diabetes is initiated by reactivity to the insulin B chain 9-23 epitope and involves functional epitope spreading. J. Autoimmun. 39, 347–353 (2012).
Article CAS PubMed PubMed Central Google Scholar
Hunter, Z. et al. A biodegradable nanoparticle platform for the induction of antigen-specific immune tolerance for treatment of autoimmune disease. ACS Nano 8, 2148–2160 (2014).
Article CAS PubMed PubMed Central Google Scholar
Podojil, J. R. et al. Tolerogenic immune-modifying nanoparticles encapsulating multiple recombinant pancreatic beta cell proteins prevent onset and progression of Type 1 diabetes in nonobese diabetic mice. J. Immunol. 209, 465–475 (2022).
Article CAS PubMed PubMed Central Google Scholar
Benne, N., Ter Braake, D., Stoppelenburg, A. J. & Broere, F. Nanoparticles for inducing antigen-specific T cell tolerance in autoimmune diseases. Front. Immunol. 13, 864403 (2022).
Article CAS PubMed PubMed Central Google Scholar
Saito, E. et al. Design of biodegradable nanoparticles to modulate phenotypes of antigen-presenting cells for antigen-specific treatment of autoimmune disease. Biomaterials 222, 119432 (2019).
Article CAS PubMed PubMed Central Google Scholar
Wicker, L. S. et al. Type 1 diabetes genes and pathways shared by humans and NOD mice. J. Autoimmun. 25, 29–33 (2005).
Article ADS CAS PubMed Google Scholar
Walker, L. S. & von Herrath, M. CD4 T cell differentiation in type 1 diabetes. Clin. Exp. Immunol. 183, 16–29 (2016).
Article CAS PubMed Google Scholar
Haskins, K. Pathogenic T-cell clones in autoimmune diabetes: more lessons from the NOD mouse. Adv. Immunol. 87, 123–162 (2005).
Article CAS PubMed Google Scholar
Delong, T. et al. Pathogenic CD4 T cells in type 1 diabetes recognize epitopes formed by peptide fusion. Science 351, 711–714 (2016).
Article ADS CAS PubMed PubMed Central Google Scholar
Crawford, S. A. et al. Cathepsin D drives the formation of hybrid insulin peptides relevant to the pathogenesis of Type 1 Diabetes. Diabetes 71, 2793–2803 (2022).
Article CAS PubMed PubMed Central Google Scholar
Baker, R. L. et al. CD4 T cells reactive to hybrid insulin peptides are indicators of disease activity in the NOD mouse. Diabetes 67, 1836–1846 (2018).
Article CAS PubMed PubMed Central Google Scholar
Dastagir, S. R. et al. Efficient presentation of multiple endogenous epitopes to both CD4(+) and CD8(+) diabetogenic T cells for tolerance. Mol. Ther. Methods Clin. Dev. 4, 27–38 (2017).
Article CAS PubMed Google Scholar
Jamison, B. L. et al. Nanoparticles containing an Insulin-ChgA hybrid peptide protect from transfer of autoimmune diabetes by shifting the balance between effector T cells and regulatory T cells. J. Immunol. 203, 48–57 (2019).
Article CAS PubMed PubMed Central Google Scholar
Mannering, S. I., Rubin, A. F., Wang, R. & Bhattacharjee, P. Identifying New Hybrid Insulin Peptides (HIPs) in Type 1 Diabetes. Front. Immunol. 12, 667870 (2021).
Article CAS PubMed PubMed Central Google Scholar
Wiles, T. A. et al. Identification of Hybrid Insulin Peptides (HIPs) in Mouse and Human Islets by Mass Spectrometry. J. Proteome Res. 18, 814–825 (2019).
Article PubMed PubMed Central Google Scholar
Baker, R. L. et al. Hybrid insulin peptides are autoantigens in Type 1 Diabetes. Diabetes 68, 1830–1840 (2019).
Article CAS PubMed PubMed Central Google Scholar
Hohenstein, A. C. et al. Novel T-cell reactivities to hybrid insulin peptides in islet autoantibody-positive at-risk individuals. Diabetes 74, 933–942 (2025).
Article CAS PubMed PubMed Central Google Scholar
Arribas-Layton, D. et al. Hybrid insulin peptides are recognized by human t cells in the context of DRB1*04:01. Diabetes 69, 1492–1502 (2020).
Article CAS PubMed PubMed Central Google Scholar
Tran, M. T. et al. T cell receptor recognition of hybrid insulin peptides bound to HLA-DQ8. Nat. Commun. 12, 5110 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Callebaut, A. et al. An Insulin-Chromogranin a hybrid peptide activates DR11-restricted T cells in human Type 1 diabetes. Diabetes 73, 743–750 (2024).
Article CAS PubMed PubMed Central Google Scholar
Mitchell, A. M. et al. T-cell responses to hybrid insulin peptides prior to type 1 diabetes development. Proc. Natl. Acad Sci. USA 118, https://doi.org/10.1073/pnas.2019129118
Bhattacharjee, P. et al. Proinsulin C-peptide is a major source of HLA-DQ8 restricted hybrid insulin peptides recognized by human islet-infiltrating CD4(+) T cells. PNAS Nexus 3, pgae491 (2024).
Article CAS PubMed PubMed Central Google Scholar
Espinosa-Carrasco, G. et al. CD4(+) T helper cells play a key role in maintaining diabetogenic CD8(+) T cell function in the pancreas. Front. Immunol. 8, 2001 (2017).
Article PubMed Google Scholar
Laidlaw, B. J., Craft, J. E. & Kaech, S. M. The multifaceted role of CD4(+) T cells in CD8(+) T cell memory. Nat. Rev. Immunol. 16, 102–111 (2016).
Article CAS PubMed PubMed Central Google Scholar
Zhang, S., Zhang, H. & Zhao, J. The role of CD4 T cell help for CD8 CTL activation. Biochem. Biophys. Res. Commun. 384, 405–408 (2009).
Article ADS CAS PubMed Google Scholar
Smith, C. M. et al. Cognate CD4(+) T cell licensing of dendritic cells in CD8(+) T cell immunity. Nat. Immunol. 5, 1143–1148 (2004).
Article CAS PubMed Google Scholar
Abdelsamed, H. A. et al. Beta cell-specific CD8(+) T cells maintain stem cell memory-associated epigenetic programs during type 1 diabetes. Nat. Immunol. 21, 578–587 (2020).
Article CAS PubMed PubMed Central Google Scholar
Gearty, S. V. et al. An autoimmune stem-like CD8 T cell population drives type 1 diabetes. Nature 602, 156–161 (2022).
Article ADS CAS PubMed Google Scholar
Zakharov, P. N., Hu, H., Wan, X. & Unanue, E. R. Single-cell RNA sequencing of murine islets shows high cellular complexity at all stages of autoimmune diabetes. J. Exp. Med. 217. https://doi.org/10.1084/jem.20192362
Kasmani, M. Y. et al. Autoreactive CD8 T cells in NOD mice exhibit phenotypic heterogeneity but restricted TCR gene usage. Life Sci. Alliance 5. https://doi.org/10.26508/lsa.202201503
Grebinoski, S. et al. Autoreactive CD8(+) T cells are restrained by an exhaustion-like program that is maintained by LAG3. Nat. Immunol. 23, 868–877 (2022).
Article CAS PubMed PubMed Central Google Scholar
Wiedeman, A. E. et al. Autoreactive CD8+ T cell exhaustion distinguishes subjects with slow type 1 diabetes progression. J. Clin. Invest. 130, 480–490 (2020).
Article CAS PubMed PubMed Central Google Scholar
Selck, C. et al. Extraislet expression of islet antigen boosts T cell exhaustion to partially prevent autoimmune diabetes. Proc. Natl. Acad. Sci. USA 121, e2315419121 (2024).
Article CAS PubMed PubMed Central Google Scholar
Wise, M. P., Bemelman, F., Cobbold, S. P. & Waldmann, H. Linked suppression of skin graft rejection can operate through indirect recognition. J. Immunol. 161, 5813–5816 (1998).
Article CAS PubMed Google Scholar
Hoyne, G. F., Dallman, M. J. & Lamb, J. R. Linked suppression in peripheral T cell tolerance to the house dust mite-derived allergen Der p 1. Int. Arch. Allergy Immunol. 118, 122–124 (1999).
Article CAS PubMed Google Scholar
Marsland, B. J. et al. Bystander suppression of allergic airway inflammation by lung resident memory CD8+ T cells. Proc. Natl. Acad. Sci. USA 101, 6116–6121 (2004).
Article ADS CAS PubMed PubMed Central Google Scholar
Gravano, D. M. & Vignali, D. A. The battle against immunopathology: infectious tolerance mediated by regulatory T cells. Cell Mol. Life Sci. 69, 1997–2008 (2012).
Article CAS PubMed Google Scholar
Jamison, B. L. et al. Tolerogenic delivery of a hybrid insulin peptide markedly prolongs islet graft survival in the NOD mouse. Diabetes 71, 483–496 (2022).
Article CAS PubMed PubMed Central Google Scholar
Neef, T. et al. Tolerance Induced by Antigen-Loaded PLG Nanoparticles Affects the Phenotype and Trafficking of Transgenic CD4(+) and CD8(+) T Cells. Cells 10. https://doi.org/10.3390/cells10123445
Stegall, M. D., Loberman, Z., Ostrowska, A., Coulombe, M. & Gill, R. G. Autoimmune destruction of islet grafts in the NOD mouse is resistant to 15-deoxyspergualin but sensitive to anti-CD4 antibody. J. Surg. Res. 64, 156–160 (1996).
Article CAS PubMed Google Scholar
Wegmann, D. R. et al. Establishment of islet-specific T-cell lines and clones from islet isografts placed in spontaneously diabetic NOD mice. J. Autoimmun. 6, 517–527 (1993).
Article CAS PubMed Google Scholar
Wang, Y., Pontesilli, O., Gill, R. G., La Rosa, F. G. & Lafferty, K. J. The role of CD4+ and CD8+ T cells in the destruction of islet grafts by spontaneously diabetic mice. Proc. Natl. Acad. Sci. USA 88, 527–531 (1991).
Article ADS CAS PubMed PubMed Central Google Scholar
Lieberman, S. M. et al. Identification of the beta cell antigen targeted by a prevalent population of pathogenic CD8+ T cells in autoimmune diabetes. Proc. Natl. Acad. Sci. USA 100, 8384–8388 (2003).
Article ADS CAS PubMed PubMed Central Google Scholar
Fu, F. et al. Long-term islet graft survival in streptozotocin- and autoimmune-induced diabetes models by immunosuppressive and potential insulinotropic agent FTY720. Transplantation 73, 1425–1430 (2002).
Article CAS PubMed Google Scholar
Suarez-Pinzon, W., Rajotte, R. V., Mosmann, T. R. & Rabinovitch, A. Both CD4+ and CD8+ T-cells in syngeneic islet grafts in NOD mice produce interferon-gamma during beta-cell destruction. Diabetes 45, 1350–1357 (1996).
Article CAS PubMed Google Scholar
Wang, Q. et al. Pancreatic islet transplantation: current advances and challenges. Front. Immunol. 15, 1391504 (2024).
Article CAS PubMed PubMed Central Google Scholar
Trapnell, C. et al. The dynamics and regulators of cell fate decisions are revealed by pseudotemporal ordering of single cells. Nat. Biotechnol. 32, 381–386 (2014).
Article CAS PubMed PubMed Central Google Scholar
Mohankumar, K., Shrestha, R. & Safe, S. Nuclear receptor 4A1 (NR4A1) antagonists target paraspeckle component 1 (PSPC1) in cancer cells. Mol. Carcinog. 61, 73–84 (2022).
Article CAS PubMed Google Scholar
You, X. et al. Transcriptional up-regulation of relaxin-3 by Nur77 attenuates beta-adrenergic agonist-induced apoptosis in cardiomyocytes. J. Biol. Chem. 293, 14001–14011 (2018).
Article CAS PubMed PubMed Central Google Scholar
Azzam, H. S. et al. CD5 expression is developmentally regulated by T cell receptor (TCR) signals and TCR avidity. J. Exp. Med. 188, 2301–2311 (1998).
Article CAS PubMed PubMed Central Google Scholar
Fasso, M. et al. T cell receptor (TCR)-mediated repertoire selection and loss of TCR vbeta diversity during the initiation of a CD4(+) T cell response in vivo. J. Exp. Med. 192, 1719–1730 (2000).
Article CAS PubMed PubMed Central Google Scholar
Campillo-Davo, D., Flumens, D. & Lion, E. The quest for the best: how TCR affinity, avidity, and functional avidity affect TCR-engineered T-cell antitumor responses. Cells 9. https://doi.org/10.3390/cells9071720
Fuchs, Y. F. et al. CD8(+) T cells specific for the islet autoantigen IGRP are restricted in their T cell receptor chain usage. Sci. Rep. 7, 44661 (2017).
Article ADS PubMed PubMed Central Google Scholar
Clemente-Casares, X. et al. Expanding antigen-specific regulatory networks to treat autoimmunity. Nature 530, 434–440 (2016).
Article ADS CAS PubMed Google Scholar
Freeborn, R. A., Strubbe, S. & Roncarolo, M. G. Type 1 regulatory T cell-mediated tolerance in health and disease. Front. Immunol. 13, 1032575 (2022).
Article CAS PubMed PubMed Central Google Scholar
Kamanaka, M. et al. Expression of interleukin-10 in intestinal lymphocytes detected by an interleukin-10 reporter knockin tiger mouse. Immunity 25, 941–952 (2006).
Article CAS PubMed Google Scholar
Yu, H. et al. Intestinal type 1 regulatory T cells migrate to periphery to suppress diabetogenic T cells and prevent diabetes development. Proc. Natl. Acad. Sci. USA 114, 10443–10448 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Smith, L. K. et al. Interleukin-10 Directly Inhibits CD8(+) T cell function by enhancing N-Glycan branching to decrease antigen sensitivity. Immunity 48, 299–312.e295 (2018).
Article CAS PubMed PubMed Central Google Scholar
Mittal, S. K. & Roche, P. A. Suppression of antigen presentation by IL-10. Curr. Opin. Immunol. 34, 22–27 (2015).
Article CAS PubMed PubMed Central Google Scholar
Bhattacharyya, S. et al. Immunoregulation of dendritic cells by IL-10 is mediated through suppression of the PI3K/Akt pathway and of IkappaB kinase activity. Blood 104, 1100–1109 (2004).
Article CAS PubMed Google Scholar
Ferris, S. T. et al. cDC1 prime and are licensed by CD4(+) T cells to induce anti-tumour immunity. Nature 584, 624–629 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Peterson, J. N. & Alpert, N. R. Molecular motor mechanics in the contracting heart. V1 versus V3 myosin heavy chain. Ann. N. Y Acad. Sci. 793, 54–63 (1996).
Article ADS CAS PubMed Google Scholar
Lei, X. et al. CD4(+) helper T cells endow cDC1 with cancer-impeding functions in the human tumor micro-environment. Nat. Commun. 14, 217 (2023).
Article ADS CAS PubMed PubMed Central Google Scholar
Busselaar, J., Tian, S., van Eenennaam, H. & Borst, J. Helpless priming sends CD8(+) T cells on the road to exhaustion. Front Immunol. 11, 592569 (2020).
Article CAS PubMed PubMed Central Google Scholar
Paley, M. A. et al. Progenitor and terminal subsets of CD8+ T cells cooperate to contain chronic viral infection. Science 338, 1220–1225 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Utzschneider, D. T. et al. T cell Factor 1-Expressing memory-like CD8(+) T cells sustain the immune response to chronic viral infections. Immunity 45, 415–427 (2016).
Aljobaily, N. et al. Autoimmune CD4(+) T cells fine-tune TCF1 expression to maintain function and survive persistent antigen exposure during diabetes. Immunity 57, 2583–2596.e2586 (2024).
Article CAS PubMed PubMed Central Google Scholar
Arakaki, R., Yamada, A., Kudo, Y., Hayashi, Y. & Ishimaru, N. Mechanism of activation-induced cell death of T cells and regulation of FasL expression. Crit. Rev. Immunol. 34, 301–314 (2014).
Article CAS PubMed Google Scholar
Singhaviranon, S., Dempsey, J. P., Hagymasi, A. T., Mandoiu, I. I. & Srivastava, P. K. Low-avidity T cells drive endogenous tumor immunity in mice and humans. Nat. Immunol. 26, 240–251 (2025).
Article CAS PubMed PubMed Central Google Scholar
Sole, P. et al. A T follicular helper cell origin for T regulatory type 1 cells. Cell Mol. Immunol. https://doi.org/10.1038/s41423-023-00989-z
Tian, Y., Mollo, S. B., Harrington, L. E. & Zajac, A. J. IL-10 regulates memory T cell development and the balance between Th1 and follicular Th cell responses during an acute viral infection. J. Immunol. 197, 1308–1321 (2016).
Article CAS PubMed PubMed Central Google Scholar
Fiorentino, D. F. et al. IL-10 acts on the antigen-presenting cell to inhibit cytokine production by Th1 cells. J. Immunol. 146, 3444–3451 (1991).
Article CAS PubMed Google Scholar
Qin, H. et al. IL-10 inhibits lipopolysaccharide-induced CD40 gene expression through induction of suppressor of cytokine signaling-3. J. Immunol. 177, 7761–7771 (2006).
Article CAS PubMed Google Scholar
Kendal, A. R. & Waldmann, H. Infectious tolerance: therapeutic potential. Curr. Opin. Immunol. 22, 560–565 (2010).
Article CAS PubMed Google Scholar
Qin, S. et al. Infectious” transplantation tolerance. Science 259, 974–977 (1993).
Article ADS CAS PubMed Google Scholar
Cobbold, S. & Waldmann, H. Infectious tolerance. Curr. Opin. Immunol. 10, 518–524 (1998).
Article CAS PubMed Google Scholar
Sultan, H. et al. Neoantigen-specific cytotoxic Tr1 CD4 T cells suppress cancer immunotherapy. Nature 632, 182–191 (2024).
Article ADS CAS PubMed PubMed Central Google Scholar
Holst, J. et al. Generation of T-cell receptor retrogenic mice. Nat. Protoc. 1, 406–417 (2006).
Article CAS PubMed Google Scholar

Download references

Acknowledgements

We thank all of the funding sources that contributed to this work, including: National Institute of Health grant R01-DK081166 (KH), National Institute of Health grant F31-AI181452 (JED), Breakthrough T1D grant 2-SRA-2020-907-S-B (KH, SDM), Breakthrough T1D grant 2-SRA-2022 1107-S-B (KH), and the Beatson Foundation grant #2024-009 (KH). We would like to acknowledge Dr. Rachel Friedman for kindly providing NOD/Tiger and Tg/NY8.3 breeder mice. We would also like to acknowledge the Barbara Davis Center Tissue Procurement and Processing Core at the University of Colorado Diabetes Research Center, NIDDK Grant P30-DK116073. Next, we would like to thank the Cancer Center Flow Cytometry Core and Genomics Core under the NIH Cancer Center Support Grant P30CA046934. We thank Dr. Kole Degolier and Dr. Terry Fry for retroviral transduction cells and protocols. We are appreciative of the open-source images available from the NIAID NIH BioArt Source. Finally, these experiments would not be possible without critical reagents (2.5HIP and IGRP tetramer) generously provided by the NIH Tetramer Core Facility (NIH Contract 75N93020D00005 and RRID:SCR_026557).

Author information

Authors and Affiliations

Department of Immunology and Microbiology, School of Medicine, University of Colorado, Aurora, CO, USA
James E. DiLisio, Janet M. Wenzlau, Maki Nakayama, Marc A. D’Antonio, Kelli S. Nicholson, Rocky L. Baker, Tonya M. Brunetti, Laurent Gapin & Kathryn Haskins
Barbara Davis Center for Childhood Diabetes, Aurora, CO, USA
K. Scott Beard & Maki Nakayama
Department of Microbiology-Immunology, School of Medicine, Northwestern University Feinberg School of Medicine, Chicago, IL, USA
Tobias Neef & Stephen D. Miller
Allen Institute for Immunology, Seattle, WA, USA
Braxton L. Jamison

Authors

James E. DiLisio
View author publications
Search author on:PubMed Google Scholar
K. Scott Beard
View author publications
Search author on:PubMed Google Scholar
Tobias Neef
View author publications
Search author on:PubMed Google Scholar
Stephen D. Miller
View author publications
Search author on:PubMed Google Scholar
Janet M. Wenzlau
View author publications
Search author on:PubMed Google Scholar
Maki Nakayama
View author publications
Search author on:PubMed Google Scholar
Marc A. D’Antonio
View author publications
Search author on:PubMed Google Scholar
Braxton L. Jamison
View author publications
Search author on:PubMed Google Scholar
Kelli S. Nicholson
View author publications
Search author on:PubMed Google Scholar
Rocky L. Baker
View author publications
Search author on:PubMed Google Scholar
Tonya M. Brunetti
View author publications
Search author on:PubMed Google Scholar
Laurent Gapin
View author publications
Search author on:PubMed Google Scholar
Kathryn Haskins
View author publications
Search author on:PubMed Google Scholar

Contributions

Conceptualization: J.E.D., R.L.B., K.H., Methodology: J.E.D., K.S.B., T.N., M.N., Investigation: J.E.D., K.S.B., M.A.D., K.S.N., B.L.J., J.M.W., Visualization: J.E.D., T.M.B., L.G., Funding acquisition: K.H., J.E.D., S.D.M., Project administration: K.H., Supervision: K.H., S.D.M., Writing – original draft: J.E.D., Writing – review & editing: J.E.D., R.L.B., L.G., K.H.

Corresponding authors

Correspondence to James E. DiLisio or Kathryn Haskins.

Ethics declarations

Competing interests

SDM is a cofounder, member of the scientific advisory board, a paid consultant, grantee, and holds stock in Cour Pharmaceuticals Development Company; a paid consultant and member of the scientific advisory board of NextCure, Inc.; and a paid consultant for Takeda Pharmaceuticals International Co. No other authors reported conflicts of interest.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information (download PDF )

Transparent Peer Review file (download PDF )

Reporting Summary (download PDF )

Source data

Source Data (download XLSX )

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

DiLisio, J.E., Beard, K.S., Neef, T. et al. Antigen-specific immunotherapy with a CD4⁺ T cell neoepitope restrains CD8⁺ T cell differentiation in murine pancreatic islet grafts. Nat Commun 17, 4355 (2026). https://doi.org/10.1038/s41467-026-70878-2

Download citation

Received: 15 July 2025
Accepted: 03 March 2026
Published: 24 March 2026
Version of record: 14 May 2026
DOI: https://doi.org/10.1038/s41467-026-70878-2