Dendritic Cell–Dependent In Vivo Generation of Autoregulatory T Cells by Antidiabetogenic MHC Class II

Sue Tsai, Pau Serra, Xavier Clemente-Casares, Robyn M. Slattery and Pere Santamaria
J Immunol July 1, 2013, 191 (1) 70-82; DOI: https://doi.org/10.4049/jimmunol.1300168

Abstract

Several mechanisms have been proposed to explain how certain MHC class II molecules afford dominant resistance to autoimmune diseases like type 1 diabetes (T1D). However, it remains unclear how protective MHC types can blunt autoreactive T cell responses directed against a diverse repertoire of autoantigenic epitopes presented by disease-promoting MHCs. In this study, we show that expression of I-E on dendritic cells (DCs) of NOD mice promotes the differentiation of MHC promiscuous autoreactive CD4+ clonotypes into antidiabetogenic autoregulatory T cells. We expressed an I-EαkloxP transgene in NOD mice and used cell type–specific I-E ablation to show that I-E–expressing DCs, but not B cells, promote the generation of autoreactive CD4+Foxp3+ regulatory T cells (Tregs) and their accumulation in the pancreas-draining lymph nodes. There, these Tregs suppress the presentation of β cell Ags to naive autoreactive CD4+ and CD8+ T cells restricted by diabetogenic MHC molecules in an I-E–independent manner. Whereas selective removal of I-E on DCs abrogated autoregulatory Treg formation and T1D protection, selective removal of I-E on B cells was inconsequential. These results explain how certain MHC class II molecules can completely suppress antigenically complex autoimmune responses in an Ag-nonspecific manner.

Introduction

The HLA gene region on human chromosome 6 accounts for >50% of the genetic risk for autoimmune diseases, including type 1 diabetes (T1D) (1–6). Genes encoding HLA-DQ and -DR or their murine counterparts (I-A and I-E, respectively) (7), as well as MHC class I genes (5, 8), play key roles, but the underlying mechanisms remain poorly understood. Human T1D is primarily associated with HLA-DQB1. Alleles encoding Ser, Ala, or Val at position 57 provide risk, whereas those encoding Asp at this position afford protection (reviewed in Refs. 9 and 10)). The NOD mouse is homozygous for a unique H-2 haplotype (H-2g7) that carries a nonproductive I-Eα gene and encodes an I-Aαd/I-Aβg7 heterodimer in which the Pro and Asp found at positions 56 and 57 in most I-A β chains are replaced by His and Ser, respectively (11). Early studies of NOD mice expressing non-NOD MHC haplotypes or I-Eαd, I-Eαk, modified I-Aβg7, I-Aαk/I-Aβk, or I-Aβd transgenes have proven that class II molecules play a direct role. However, whereas some authors suggested that protective alleles might afford resistance to T1D by tolerizing autoreactive T cells (12–15), others argued against this (16–21).

Our studies of NOD mice expressing a diabetogenic, I-Ag7–restricted TCR of unknown specificity (4.1-NOD mice (22)) documented an unexpected relationship between thymic tolerance of certain highly pathogenic I-Ag7–restricted CD4+ T cells and the MHC-linked resistance to T1D (23–25). We found that 4.1-TCR–transgenic thymocytes undergo deletion in H-2g7/b, H-2g7/k, H-2g7/q, and H-2g7/nbl NOD mice (T1D-resistant strains) and that they do so by engaging protective MHC class II molecules on an hematopoietic cell type, upon undergoing thymic positive selection on peptide–I-Ag7 complexes expressed on thymic cortical epithelial cells (23, 24). Likewise, expression of protective class II transgenes, such as I-Eαk (which restores expression of endogenous I-Eβg7), I-Aβd, and I-Aβg7PD (I-A g7 carrying Pro-56 and Asp-57) in 4.1-NOD mice, led to various degrees of 4.1-thymocyte tolerance or anergy and T1D resistance (25). On the basis of these and other observations (26, 27), we proposed that protective MHC class II molecules afford T1D resistance by tolerizing “4.1-like” (autoreactive, MHC-promiscuous, and diabetogenic) CD4+ T cells. This interpretation was consistent with the observation that the antidiabetogenic activity of protective MHC class II molecules in both 4.1-NOD and NOD mice maps to residues around I-Aβ chain position 57 (25), which are also implicated in the DQB1-linked resistance to human T1D (9, 10). However, this interpretation could not explain how deletion of 4.1-like CD4+ T cells is able to prevent the development of an autoimmune disease that, in non–TCR-transgenic NOD mice, is driven by a diverse repertoire of autoreactive CD4+ and CD8+ T cell specificities.

Because a significant fraction of 4.1-NOD.I-Eαk mice are resistant to insulitis and diabetes despite exporting functional autoreactive 4.1-CD4+ T cells to the periphery (25), we reasoned that another mechanism by which protective MHC class II molecules might afford T1D resistance would be to promote the differentiation of 4.1-like MHC class II-promiscuous autoreactive CD4+ thymocytes into autoregulatory T cell types via one or more hematopoietic APC types. To investigate this, we expressed an I-EαkloxP transgene in NOD mice, alone or in combination with highly diabetogenic I-Ag7– or Kd–restricted TCRs and/or ubiquitous or cell type–specific Cre recombinase transgenes, and explored the effects of transgene expression on effector versus regulatory T cell (Treg) development, activation, and recruitment. We report that I-E expression on CD11c+ dendritic cells (DCs), but not B cells or other I-E–expressing APCs, promotes the generation of autoregulatory CD4+Foxp3+ Tregs from endogenous precursors and mobilizes their accumulation in the pancreatic draining lymph nodes (PLNs), where they suppress the presentation of cognate and noncognate β cell Ags to diabetogenic CD4+ and CD8+ T cells by local APCs. These results provide an explanation as to how certain MHC class II molecules can suppress antigenically complex autoimmune responses.

Materials and Methods

Mice

NOD/ShiLt, NOD.scid, NOD.mβ2m−/−, and BDC2.5-NOD mice were purchased from The Jackson Laboratory (Bar Harbor, ME). 4.1-NOD, 8.3-NOD, 4.1-NOD.Rag2−/−, and 8.3-NOD.Rag2−/− mice have been described previously (22, 23). 8.3-NOD.I-Eαk.mβ2m−/−.Rag2−/− mice were generated by intercrossing the offspring of 8.3-NOD.mβ2m−/− × NOD.I-Eαk.Rag2−/− crosses. All mice were followed for development of spontaneous diabetes by measuring urine glucose levels with Diastix strips (Miles, ON, Canada). Mice were considered overtly diabetic when their urine glucose levels were 3+ (>16 mM blood glucose). NOD.I-Eαk mice were produced by backcrossing the I-Eαk transgene of C57BL/6.I-Eα16 mice (The Jackson Laboratory) (17) onto the NOD.ShiLtj background for at least 10 generations. NOD.CD19P-Cre mice were produced by backcrossing a targeted CD19-Cre transgene (28) onto the NOD.ShiLtj background for >10 generations. Briefly, a cassette encoding Cre recombinase was placed under the control of the CD19 promoter such that it replaced the second exon of the endogenous CD19 coding sequence. Mice that are hemizygous for the Cre recombinase insertion and thus carry only one copy of the CD19 allele are phenotypically normal. All NOD.CD19P-Cre–transgenic mice studied in this paper were hemizygous for the transgene. CD11cP-Cre–transgenic C57BL/6 mice were from B. Reizis (Columbia University, New York, NY) (29). NOD.CD11P-Cre mice were produced by backcrossing this transgene onto the NOD.ShiLtj background for at least 10 generations. NOD.ACTBP-Cre mice were produced by backcrossing the ACTB-Cre transgene of C57BL/6.ACTB-Cre mice (The Jackson Laboratory) onto the NOD.ShiLtj background for >10 generations. NOD.I-EkloxP mice were produced by microinjection of an 8.5-kb floxed IEαk (I-EkloxP) transgene into NOD zygotes. The transgene was generated using the 8.2-kb HindIII fragment of the IEαk gene as template. Dormant 120-bp loxP sites were inserted at the 854-bp XbaI site and between exons 1 and 2 at the 3573-bp Eco47III site. The animal studies described in this paper were approved by the University of Calgary Faculty of Medicine’s Animal Care Committee and followed the guidelines of the Canadian Council of Animal Care.

Flow cytometry reagents

FITC-, PE-, PerCP-, or biotin-conjugated Abs to CD4 (RM4-5), CD8α (53-6.7), CD11c (HL3), CD19 (1D3), CD80 (16-10A1), B220 (RA36B2), TCRβ (H56-597), Vβ8 (MR5-2), Vβ11 (RR3-15), I-Ag7 (OX6), I-Ek (17-3-3), and streptavidin–PerCP were purchased from BD Pharmingen (San Diego, CA). Anti-Foxp3 (clone FJK-16s) and anti-Helios (clone 22F6) Abs were from eBioscience (San Diego, CA). CFSE was from Molecular Probes (Eugene, OR).

Peptides

The peptides NRP-A7 (KYNKANAFL) (30) and 2.5 mimetope (2.5mi) (31) were purchased from Mimotopes (Clayton, VIC, Australia).

Flow cytometry

Single-cell suspensions were stained with Abs diluted 1:100 in FACS buffer (PBS containing 1% FBS and 0.05% NaN3), washed, fixed in 1% paraformaldehyde in PBS, and analyzed with FACScan, FACSaria, or BD LSRII flow cytometers. All Ab stainings of DCs were done in the presence of an anti-CD16/CD32 mAb (2.4G2; BD Pharmingen) to block FcRs. For intranuclear staining, cells were first stained for surface markers, followed by permeabilization using the Foxp3 staining buffer set (eBioscience), according to the manufacturer’s instructions. To enumerate Helios+Foxp3+ cells in the CD4+ subset, cells from the CD4+CD8− gate were subjected to gating on the Helios+Foxp3+ subsets using cutoffs set for thymic Foxp3+ (Helios+) and Foxp3− (Helios−) populations.

Analysis of Treg TCRα and TCRβ repertoire by PCR

CD4+CD25+ Tregs and CD4+CD25− T cells were isolated from splenocytes and PLN cells by FACS and used as a source of mRNA using the RNeasy Mini Kit (Qiagen, Gaithersburg, MD). cDNA was synthesized by oligo(dT)-primed reverse transcription using Moloney murine leukemia virus reverse transcriptase (Invitrogen), followed by two rounds of PCR amplication using TCR-Cα– or Cβ- and Vα- and Vβ-chain–specific primers, as described previously (32).

Histochemistry

Scoring of insulitic lesions was performed on H&E-stained pancreas sections as described previously (33). Insulitis scores were determined using the following criteria: 0, no insulitis; 1, peri-insulitis; 2, insulitic lesion covering <25% of the islet surface area; and 3 and 4, insulitic lesion covering 25–50 and >50% of the total islet surface area, respectively.

T cell isolation

CD4+ T cells were purified from the spleens of 8- to 12-wk-old NOD or 4.1-NOD mice using purification systems from Miltenyi Biotec (Auburn, CA) or BD IMAG following the manufacturer’s instructions. CD4+CD25+ and CD4+CD25− T cells were isolated using the EasySep Mouse Treg isolation kit from StemCell Technologies as per the manufacturer’s instructions. The purities of each population were confirmed to be >90% by flow cytometry. CD4+ and CD8+ T cells used in adoptive transfer experiments were isolated from the spleens of 6- to 8-wk-old NOD females using mAb-coated BD IMAG magnetic beads (BD Biosciences). Islet-associated CD4+ T cells were drawn out from the islets by culturing islets in 0.5 U/ml Takeda rIL-2– and FBS (10% v/v)-supplemented RPMI 1640 medium (Life Technologies) overnight. Total cells were collected the next day for flow cytometric analysis.

DC isolation

Spleens or thymi were immersed in HBSS (Life Technologies) containing collagenase D (0.25% w/v; Roche, Mannheim, Germany) and DNAseI (0.01% w/v; Roche) and shredded into small pieces with two syringe needles to facilitate digestion (in a 37°C water bath for 15 min, with shaking). Digested spleens were passed through a 40-μm cell strainer and subjected to DC isolation using anti-mouse CD11c-Ab–coated magnetic beads (Miltenyi Biotec). Purified splenic DCs were used immediately in flow cytometry analysis, or proliferation or coculture assays without further culturing. Bone marrow–derived DCs (BMDCs) were generated by culturing bone marrow cells in the presence of recombinant mouse IL-4 (5 ng/ml; R&D Systems, Minneapolis, MN) and GM-CSF (5 ng/ml; R&D Systems) for 7 d, followed by magnetic isolation using CD11c-Ab–coated magnetic beads (Miltenyi Biotec). LPS (1 μg/ml) were added to the culture for the last 24 h to generate mature BMDCs.

In vivo CFSE dilution

Purified CD8+ T cells were stained with CFSE (0.5 μM) for 4 min at 37°C, washed twice with 15 ml RPMI 1640 medium (Life Technologies) containing 10% FBS (Sigma-Aldrich), and enumerated. Cells (5 × 106) were injected i.v. into NOD or transgenic hosts, and spleens plus mesenteric draining LNs (MLNs) and PLNs were collected for flow cytometric analysis 7 d after injection.

Functional in vitro T cell assays/I-Ek allorecognition assay

For Ag-stimulated T cell proliferation, 8.3-CD8+ T cells (2 × 104) or BDC2.5-CD4+ T cells (4 × 104) were cocultured with DCs (104) and various concentrations of NRP-A7 or 2.5 mimotope, respectively. To test I-Ek allorecognition, 8.3-CD8+ T cells (2 × 104) were cocultured with NOD versus NOD.I-Eαk DCs (104) and in the absence of NRP-A7 peptide. 4.1-CD4+ T cells (4 × 104) were cocultured with dissociated islet cells (2 × 105), which contained autoantigen-loaded APCs. Culture supernatants were collected at 48 h for cytokine determination by ELISA (R&D Systems). Culture wells were pulsed with [3H]thymidine at this point (1 μCi/well; Molecular Probes) and harvested for measurement of 3H incorporation at 72 h. For polyclonal T cell stimulation, Dynabeads Mouse T activator anti-CD3/CD28 beads (Invitrogen Dynal, Oslo, Norway) were used accordingly with the manufacturer’s recommendations.

In vitro Treg suppression assay

CD4+CD25+ or CD4+CD25− T cells were activated with plate-bound anti-CD3 mAb (3 μg/ml) and Takeda rIL-2 (0.5 U/ml) and cocultured at different ratios with responder 8.3-CD8+ T cells (2 × 104) and gamma-irradiated (1200 rad) NOD splenocytes (105) in the presence of 1 μg/ml NRP-A7 for 48 h. The supernatant was collected for IFN-γ determination by ELISA.

In vitro thymocyte/DC coculture

FACS-sorted CD4+CD8− single-positive (SP) thymocytes (4 × 104) were cultured with thymic DCs (104) in the presence or absence of murine thymic stromal lymphopoietin (TSLP) (20 ng/ml; R&D Systems) for 4 or 7 d, at which point the cultured cells were harvested and analyzed by flow cytometry for CD4 and Foxp3 expression.

In vivo Treg suppression of DCs

TCR-transgenic CD4+CD25+ Tregs (5 × 105) were injected i.v. into 15-wk-old NOD females, and the spleens and PLNs were collected 4 d later for analysis by flow cytometry. CD11c+ splenic DCs were isolated as described above; PLNs were subjected to collagenase-digestion and stained for CD11c, CD80, and I-Ag7 directly.

In vivo Treg recruitment

To assay relative recruitment of NOD versus NOD.I-Eαk Tregs, FACS-sorted CD4+CD25+ Tregs (106) from NOD and NOD.I-Eαk mice were stained with 10 μM and 1 mM CFSE and injected 1:1 into NOD hosts. MLNs and PLNs were analyzed for presence of CFSE+ cells 4 d later. To assay relative recruitment by NOD versus NOD.I-Eαk hosts, FACS-sorted CD4+CD25+ Tregs (106) from NOD.I-Eαk mice were stained with 10 μM CFSE and injected into NOD and NOD.I-Eαk hosts, and the frequency of CFSE+ cells in the MLNs and PLNs was quantified by flow cytometry 4 d later.

In vivo Treg suppression of diabetes

TCR-transgenic CD4+CD25+ Tregs (2 × 105), or non–TCR-transgenic Tregs (5 × 105)were injected i.v. into 5- to 6-wk-old 8.3-NOD.Rag2−/− or 6- to 8-wk-old NOD.scid females. Injected 8.3-NOD.Rag2−/− mice were monitored weekly for development of glycosuria using Diastix strips for at least 100 d. Treg-injected NOD.scid mice were infused with NOD splenic CD4+ and CD8+ T cells (20 × 106) 24 h later and followed for glycosuria for 100 d.

Statistical analyses

Data were compared by Student t, Mann–Whitney U, or two-way ANOVA tests using GraphPad software. Survival curve comparisons were performed using the log-rank test, and comparison of final diabetes incidence was done using χ2 tests.

Results

Diabetes resistance and enhanced development and recruitment of Tregs in oligoclonal but not monoclonal 4.1-NOD.I-Eαk and 8.3-NOD.I-Eαk mice

To investigate whether protective MHC class II molecules might afford T1D resistance by promoting the differentiation of 4.1-like MHC class II promiscuous autoreactive CD4+ thymocytes into autoregulatory T cell types, we followed the fate of 4.1-CD4+ T cells in 4.1-NOD.I-Eαk mice. As expected, 4.1-NOD.I-Eαk mice were almost completely protected from T1D as compared with 4.1-NOD mice (Fig. 1A). The thymocyte profiles of both strains, including the absolute and relative numbers of the different thymocyte subsets and the levels of the transgenic TCR on CD4+CD8+ (double-positive [DP]) and CD4+CD8− (SP) thymocytes, were similar (Fig. 1B, 1C; data not shown). Likewise, the peripheral CD4+CD25− T cells of 4.1-NOD.I-Eαk and 4.1-NOD mice proliferated to a similar extent and secreted similar levels of IFN-γ in response to plate-bound anti-CD3 mAb or islet cell suspensions (Fig. 1D), indicating that the resistance of 4.1-NOD.I-Eαk mice to T1D was not caused by deletional tolerance or functional inactivation. 4.1-NOD.I-Eαk mice, however, harbored increased percentages of Foxp3+CD4+ T cells in the thymus and the pancreatic (but not mesenteric) lymph nodes (Fig. 1E). The absolute numbers of Foxp3+CD4+ T cells were also significantly higher in the thymic CD4+CD8− subset of 4.1-NOD.I-Eαk versus 4.1-NOD mice (Supplemental Fig. 1A), albeit not in the PLNs (owing to lower cellularity and absence of local inflammation in the former; data not shown). Helios staining confirmed that these differences in Treg frequencies originated in the thymus (Fig. 1F, Supplemental Fig. 1B, 1C).

FIGURE 1.
FIGURE 1.

Diabetes resistance and enhanced development and recruitment of Tregs in oligoclonal but not monoclonal 4.1-NOD.I-Eαk mice. (A) Cohorts of 4.1-NOD (n = 25) and 4.1-NOD.I-Eαk (n = 12) females were followed for diabetes development until 32 wk of age. (B) Representative thymic flow cytometric profiles and average percentages of CD4+, CD8+, DP, and double-negative (DN) thymocytes from 4.1-NOD (n = 27) and 4.1-NOD.I-Eαk (n = 11) mice. Data are presented as average ± SEM. (C) Representative anti-Vβ11 staining profiles of non–TCR-transgenic NOD thymocytes overlaid with those of 4.1-NOD and 4.1-NOD.I-Eαk thymocytes (left panel) and average Vβ11 expression levels on CD4+, CD8+, and DP thymocytes from 4.1-NOD (n = 6) and 4.1-NOD.I-Eαk (n = 6) mice (right panel). Data are presented as mean fluorescence intensity (MFI) normalized to the levels seen in 4.1-NOD controls (top right) or as unnormalized absolute values (bottom right). Normalized MFI was calculated by dividing the absolute MFI of the 4.1-NOD.I-Eαk by corresponding values from 4.1-NOD. Data are presented as average ± SEM. (D) Anti–CD3-induced (top) or islet cell–induced (bottom) IFN-γ secretion and proliferation of splenic CD4+ T cells isolated from 4.1-NOD and 4.1-NOD.I-Eαk mice. Data at the top correspond to seven mice per strain, assayed in four separate experiments. Data at the bottom correspond to three mice per strain, assayed in two separate experiments. (E) Percentages of Foxp3+ T cells in CD4+ T cells from the thymus (THY), MLNs, and PLNs of 4.1-NOD (n = 9) versus 4.1-NOD.I-Eαk (n = 7) mice. Data are presented as mean ± SEM. (F) Percentages of Foxp3+Helios+ T cells in thymic (THY) CD4+CD8− cells, MLNs, and PLNs of 4.1-NOD versus 4.1-NOD.I-Eαk mice. n = 5–7. Supplemental Fig. 1B and 1C depict representative stains. (G) Cohorts of 8.3-NOD.Rag2−/− females were injected with purified splenic CD4+CD25+ Tregs from 4.1-NOD (n = 9) and 4.1-NOD.I-Eαk (n = 7) mice and followed for diabetes development until 100 d postinjection. (H) Cohorts of 4.1-NOD.Rag2−/− mice (n = 18) and 4.1-NOD.I-Eαk.Rag2−/− mice (n = 35) females were followed for diabetes development until 32 wk of age.

To ascertain whether the increased peripheral frequency of Tregs in 4.1-NOD.I-Eαk mice was associated with increased regulatory activity, we compared the ability of splenic Tregs purified from 4.1-NOD.I-Eαk and 4.1-NOD donor mice to suppress the development of diabetes in two different adoptive transfer models. As shown in Fig. 1G and Supplemental Fig. 1D, the former had superior antidiabetogenic properties than the latter. Clearly, both phenotypes (diabetes resistance and increased Treg development and function) were 4.1-TCR transgene independent, because 4.1-NOD.I-Eαk.Rag2−/− mice did not export Foxp3+4.1-CD4+ T cells (data not shown) and developed diabetes essentially like their RAG2-competent counterparts (Fig. 1H). These observations indicated that the antidiabetogenic and pro-Treg developmental effects of I-E expression in 4.1-NOD.I-Eαk mice require an interaction between endogenous (nontransgenic) TCRs and I-Eαkβg7 and are dissociated from epitope stealing (i.e., capture of the cognate I-Ag7–binding epitope by I-E (34, 35) or blockade of the I-Ag7 peptide binding site by a peptide derived from I-E (36)).

To steer clear of potential effects of I-Eαk on CD4+ effector T cell development and function altogether and to rule out the possibility that the observations described above were a peculiarity of the 4.1-TCR, we introduced the I-Eαk transgene into 8.3-TCR–transgenic NOD (8.3-NOD) mice (22), expressing a diabetogenic H-2Kd–restricted TCR that is specific for the β cell epitope IGRP206–214 (30, 37, 38). Similar to 4.1-NOD.I-Eαk mice, 8.3-NOD.I-Eαk mice showed near-complete resistance to T1D (Fig. 2A). Thymocyte flow cytometric profiles also showed no evidence of clonal deletion of 8.3-TCR–transgenic thymocytes; the absolute and relative numbers of CD4−CD8−, CD4+CD8+, CD4+CD8−, or CD4−CD8+ thymocytes in 8.3-NOD.I-Eαk mice were similar to those seen in 8.3-NOD mice (Fig. 2B; data not shown).

FIGURE 2.
FIGURE 2.

Diabetes resistance and enhanced development and recruitment of Tregs in oligoclonal but not monoclonal 8.3-NOD.I-Eαk mice. (A) 8.3-NOD (n = 598) and 8.3-NOD.I-Eαk (n = 30) females were followed for diabetes development until 32 wk of age. (B) Representative thymic flow cytometric profiles and average percentages of CD4+, CD8+, DP, and double-negative (DN) thymocytes from 8.3-NOD (n = 6) and 8.3-NOD.I-Eαk (n = 6) mice. Data are presented as mean ± SEM. (C) Representative anti-Vβ8 staining profiles of non–TCR-nontransgenic NOD thymocytes overlaid with those of 8.3-NOD and 8.3-NOD.I-Eαk thymocytes (left panel), and average Vβ8 expression levels on CD4+, CD8+, and DP thymocytes from 8.3-NOD (n = 5) and 8.3-NOD.I-Eαk (n = 6) mice (right panel). Data are presented as absolute mean fluorescence intensity (MFI) or as MFI normalized to the levels seen in 8.3-NOD controls, calculated by dividing the absolute MFI of 8.3-NOD.I-Eαk mice by corresponding values from 8.3-NOD mice analyzed in each independent experiment. Data are presented as average ± SEM. (D) Percentages of Foxp3+ T cells in the CD4+ subset from the thymus (THY), MLN, and PLN of 8.3-NOD (n = 5) versus 8.3-NOD.I-Eαk (n = 6) mice. Data are presented as means ± SEM. (E) Cohorts of 8.3-NOD.Rag2−/− (n = 106) and 8.3-NOD.I-Eαk.Rag2−/− (n = 18) females were followed for diabetes development until 32 wk of age. (F) Thymic DCs were isolated from young NOD or NOD.I-Eαk mice and cocultured (105) with 8.3-TCR+ CD4-SP thymocytes (2 × 105) in the presence or absence of mouse TSLP (20 ng/ml) for 7 d, at which point Foxp3+ percentages were determined by flow cytometry. Top panel compares data presented as mean fold increase, calculated by dividing the percentage of Foxp3+ of the cultures using I-Eαk+ DCs by the corresponding values in cultures using I-Eαk− DCs. This experiment was repeated three times with one to two mice per experiment. Bottom panel shows representative FACS staining profiles from two mice.

Interestingly, however, the CD4+CD8+ and CD4+CD8− thymocyte subsets of 8.3-NOD.I-Eαk mice expressed slightly lower levels of the transgenic TCRβ-chain (Fig. 2C), suggesting a potential cognate interaction between the transgenic MHC class I–restricted TCR on CD4+ thymocytes and I-Eαkβg7 leading to 8.3-TCR downregulation. This is an intriguing observation, because one would not normally expect a MHC class I–restricted TCR to cross-react with a MHC class II molecule. However, we and others have previously observed peptide-MHC (pMHC)-binding promiscuity to be a feature of other pathogenic CD4+ and CD8+ T cell clonotypes (23, 24, 39), one that might presumably allow antidiabetogenic MHC class II molecules to effect disease resistance. Alternatively, this downregulation is due to the pairing of the transgenic TCRβ-chain with endogenous α-chains capable of enhancing this TCR’s intrinsic MHC promiscuity. Whatever the explanation, 8.3-TCR downregulation in thymocytes from 8.3-NOD.I-Eαk was not inherited by peripheral 8.3-CD8+ T cells (data not shown). In addition, the peripheral CD8+ T cells of 8.3-NOD.I-Eαk (and 8.3-NOD mice) did not cross-react with I-Ek expressed on splenic DCs from NOD.I-Eαk (Supplemental Fig. 2A, top), and I-Ek expression on the DCs did not affect the in vitro responsiveness of the CD8+ T cells of either mouse strain to the cognate autoantigenic epitope (Supplemental Fig. 2A, bottom). Likewise, the splenic CD8+ T cells of 8.3-NOD.I-Eαk and 8.3-NOD mice secreted similar levels of IFN-γ and proliferated equally well in response to NRP-V7 peptide–pulsed NOD DCs in vitro (Supplemental Fig. 2B), indicating that they are not hyporesponsive to antigenic stimulation.

Like 4.1-NOD.I-Eαk mice, 8.3-NOD.I-Eαk mice harbored increased percentages of CD4+Foxp3+ T cells in both the thymus and PLNs (Fig. 2D). As was the case with 4.1-TCR–transgenic mice, introduction of a RAG-2 deficiency into 8.3-NOD.I-Eαk mice abrogated both Foxp3+CD4+ Treg development (Supplemental Fig. 2C) and their diabetes resistance (Fig. 2E), suggesting essential roles for both endogenous TCR rearrangements and Foxp3+ Tregs in I-Ek–mediated diabetes protection. This was further substantiated by introducing a β2-microglobulin deficiency into 8.3-NOD.I-Eαk.Rag2−/− mice; as shown in Supplemental Fig. 2C, the 8.3-TCR was incapable of promoting Foxp3+CD4+ (or CD4+) T cell development via I-Eαk, even in the absence of MHC class I molecules.

Taken together, the above data suggested that, in both types of TCR-transgenic mice, expression of I-Eαk promoted Foxp3+CD4+ Treg development in the thymus. To explore this further, we queried the ability of purified thymic DCs from NOD and NOD.I-Eαk mice to promote the development of Foxp3+CD4+ T cells from purified 8.3-CD4+CD8− thymocytes in vitro. Remarkably, DCs from NOD.I-Eαk DCs induced significantly higher yields of Foxp3+CD4+ T cells than those isolated from NOD mice (Fig. 2F, Supplemental Fig. 2D), and these differences were augmented by the addition of TSLP, known to foster Treg development in human thymic cultures (40).

Collectively, these observations suggest that I-Eαk–expressing thymic DCs promote thymic Treg formation, either by promoting the conversion of uncommitted thymocytes into Tregs or by triggering the expansion of preformed Tregs. Because up to 20% of Foxp3+CD4+ thymocytes lack CD25 expression (see representative example in Supplemental Fig. 2E), it is not possible to unambiguously determine which of these two alternative possibilities is at play by doing these experiments using CD25-depleted thymocytes.

Enhanced development and recruitment of Tregs in non–TCR-transgenic NOD.I-Eαk mice

The I-Eαk–associated increase in thymic Treg development seen in both 4.1- and 8.3-TCR–transgenic mice was not seen in thymi from non–TCR-transgenic NOD.I-Eαk mice (Fig. 3A), suggesting that it is either driven by autoreactive TCRs or is only visible in TCR-transgenic animals owing to the oligoclonal nature of their repertoires. However, like the PLNs of their TCR-transgenic counterparts, the PLNs of non–TCR-transgenic NOD.I-Eαk mice contained significantly increased percentages of Foxp3+CD4+ T cells as compared with NOD mice (Fig. 3A). Furthermore, Tregs purified from pooled PLNs (but not spleen) of NOD.I-Eαk mice exhibited different TCR usage than those isolated from the PLNs of NOD mice, as determined using a TCRα- or TCRβ-specific nested PCR approach (Supplemental Fig. 3) (32), followed by sequencing to validate the specificity of the PCR approach (data not shown). These two separate lines of evidence suggested that the Foxp3+CD4+ Treg subset developing in non–TCR-transgenic NOD.I-Eαk mice is also enriched for autoreactive Treg cell specificities, leading to enhanced recruitment and retention in the pancreas-draining LNs. To investigate this further, we compared the ability of equal numbers of purified splenic Tregs from NOD.I-Eαk versus NOD mice to protect 8.3.NOD.Rag2−/− hosts from developing diabetes. As shown in Fig. 3B, Tregs from NOD.I-Eαk donors suppressed disease more efficiently than those derived from NOD mice. To substantiate this further, we compared the accumulation of splenic CFSE-labeled Tregs from NOD.I-Eαk and NOD donors (first labeled with low and high CFSE concentrations, respectively, and then pooled at 1:1 ratio) into the MLNs and PLNs of wild-type NOD hosts. As shown in Fig. 3C, NOD.I-Eαk–derived Tregs accumulated in the PLNs more efficiently than NOD-derived Tregs, supporting the idea that I-E expression promotes the development of islet–Ag-autoreactive Tregs but is dispensable for enhanced recruitment of these T cells into the PLNs. That I-E expression on peripheral DCs plays a dispensable role was further supported by the observation that splenic Tregs purified from NOD.I-Eαk donors accumulate similarly in the MLNs and PLNs of NOD and NOD.I-Eαk hosts (data not shown).

FIGURE 3.
FIGURE 3.

Enhanced development and recruitment of Tregs in NOD.I-Eαk mice. (A) Percentages of Foxp3+ T cells in CD4+CD8− T cells from the thymus (THY), MLN, and PLN of NOD versus I-Eαk+ NOD mice. Data correspond to n = 6 and 8, respectively. (B) Cohorts of 8.3-NOD.Rag2−/− females were injected with CD4+CD25+ Tregs purified from the spleens of NOD (n = 17) and I-Eαk+ NOD mice (n = 12) and followed for diabetes development until 100 d postinjection. (C) Tregs from NOD.I-Eαk PLNs are enriched for autoreactive clonotypes (i.e., recognizing Ags draining to the PLN) as compared with NOD PLN–derived Tregs. CFSE-stained NOD (CFSEhigh) and NOD.I-Eαk (CFSElow) splenic Tregs were injected into NOD hosts, and their localization was quantified by flow cytometry 4 d postinjection. Data correspond to two experiments of n = 2 mice each. (D) 4.1-TCR+Foxp3+CD4+ Tregs downregulate MHC class II and CD80 expression on PLN DCs in vivo. FACS-sorted 4.1-TCR+ Tregs (5 × 105) were injected into NOD hosts. The splenic and PLN DCs of Treg-injected or uninjected control mice were isolated 4 d later and analyzed for MHC class II (I-Ag7) (left panels) and CD80 (right panels) expression. Shown are representative flow cytometric profiles (top panels) and plots comparing relative mean fluorescence intensity (MFI) of staining (middle panels). To correct for interexperimental variability in flow cytometer settings, MFIs were normalized to those corresponding to samples of untreated NOD mice analyzed in the same experiment. Data correspond to n = 3–4 mice studied in two separate experiments. Bottom panel shows the absolute MFI values corresponding to four mice analyzed in the same experiment. (E) Expression levels of CD80 and I-Ag7 on PLN DCs of unmanipulated NOD versus NOD.I- Eαk mice. Shown are representative profiles (top panels), normalized values (middle panels), and absolute MFI values corresponding to mice analyzed in the same experiment (bottom panel). (F) In vivo proliferation of naive CFSE-labeled 8.3-CD8+ T cells in the PLN (n = 5–6 animals each) and MLN (n = 3–6) of NOD versus NOD.I-Eαk hosts.

We and others have previously shown that an important outcome of increased recruitment of autoreactive Tregs to the PLNs is the inhibition of autoantigen presentation by local DCs (41–44). Consistent with this, purified 4.1-NOD Tregs downregulated MHC class II (I-Ag7) and costimulatory molecule (i.e., CD80) expression specifically on PLN (but not MLN or splenic) DCs upon transfer into wild-type NOD hosts (Fig. 3D). Furthermore, the PLN DCs of unmanipulated NOD.I-Eαk mice expressed significantly lower levels of both I-Ag7 and CD80 than the PLN DCs of unmanipulated NOD mice (Fig. 3E). In addition, the PLNs of NOD.I-Eαk hosts exhibited a significantly reduced capacity to cross-present endogenous IGRP206–214 to CFSE-labeled 8.3-CD8+ T cells, as compared with the PLNs of wild-type NOD hosts (Fig. 3D). Downregulation of I-Ag7 and CD80 on the PLN DCs of NOD.I-Eαk versus NOD mice was specific for PLN DCs and not an artifact of I-Eαk transgenesis because neither splenic nor BMDCs exhibited this phenotype, even after LPS stimulation (data not shown).

Taken together, these data suggest that I-Eαk expression inhibits T1D development by promoting the generation of autoreactive Foxp3+CD4+ thymocytes and their recruitment to the PLNs.

I-Eαk affords T1D resistance exclusively via DCs

Previous studies in bone marrow chimeras demonstrated that I-E expression affords diabetes resistance through a hematopoietic cell type (21, 45). Additional studies of NOD mice expressing promoter-modified I-Ek transgenes to compartmentalize I-E expression on specific subsets of APCs provided additional supporting evidence in this regard (17, 46).

We have recently shown that the antidiabetogenic effects of another MHC class II molecule, I-Ab, requires that it be expressed on DCs and macrophages but not B cells or thymic epithelial cells (47). This prompted us to ask whether the antidiabetogenic ability of I-E might also be mediated via DCs. To this end, we produced transgenic NOD mice expressing a loxP-flanked I-Eαk transgene (Fig. 4A) by direct microinjection of NOD oocytes (NOD.I-EkloxP). This design resulted in lower levels of expression of I-Eαkβg7 as compared with those seen in the NOD.I-Eαk–transgenic line (Fig. 4B) but nonetheless afforded significant diabetes protection to NOD.I-EkloxP, 4.1-NOD.I-EkloxP, and 8.3-NOD.I-EkloxP mice (Fig. 5). The diabetes resistance of these mice was mediated by I-EkloxP expression rather than by a site-specific transgene integration effect because NOD.I-EkloxP mice expressing a β-actin promoter–driven Cre recombinase transgene developed diabetes with the same incidence and tempo as wild-type NOD mice (Fig. 5A).

FIGURE 4.
FIGURE 4.

Selective removal of an I-EαkloxP transgene in CD11c+ or CD19+ cells of NOD.I-EαkloxP by coexpression of CD11cP-Cre and CD19P-Cre transgenes. (A) An 8.2-kb HindIII fragment of the IEαk-gene was used to clone the final 8.5-kb floxed IEαk transgene, which was injected into NOD zygotes to generate I-EαkloxP–transgenic founder mice. Dormant 120-bp loxP sites (triangles) were inserted at the 854-bp XbaI site and between exons 1 and 2 at the 3573-bp Eco47III site (thin lines: introns; black boxes: exons). (B) I-E expression on purified splenic DCs as determined by flow cytometric analysis of cells stained with an anti–I-Ek Ab (n = 2 per strain type). Mean fluorescence intensities (MFIs) are normalized to those of unstained cells (n = 2 per strain type). (C) Top panels show representative I-Ek staining profiles of splenic CD11c+ DCs or splenic CD19+ B cells from NOD, NOD.I-EkloxP, NOD.I-EkloxP/CD11cP-Cre, and NOD.I-EkloxP/CD19P-Cre mice (in this paper referred to as NOD.I-EkloxP, NOD.I-EDC flox, and NOD.I-EB flox, respectively) and the corresponding percentages of I-E+ cells. Bottom panels show the levels of I-E on the DCs or B cells of the different strains, normalized to the corresponding values on cells from the positive control strain NOD.I-Eαk, which reflect the extent of Cre-mediated removal of I-EkloxP in CD19+ and CD11c+ cells of NOD.I-EDC flox and NOD.I-EB flox mice. Data correspond to four to five mice per strain type. (D) I-Ag7 staining profiles of splenic DCs from NOD.I-EkloxP (top panels), NOD.I-EB flox (middle panels), and NOD.I-EDC flox (bottom panels) mice as compared with NOD splenic DCs (gray histogram). Data are representative of three experiments (n = 3). (E and F) IFN-γ secretion and proliferation of splenic BDC2.5-CD4+ (E) or 8.3-CD8+ T cells (F) from the corresponding TCR-transgenic strain in response to peptide-pulsed DCs from NOD, NOD.I-EkloxP, NOD.I-EB flox, and NOD.I-EDC flox donor mice. Data are representative of four separate experiments using four to eight mice per strain. (G) In vivo proliferation of naive CFSE-labeled 8.3-CD8+ T cells in the PLN and MLN of non–I-EkloxP–transgenic NOD.CD19P-Cre and NOD.CD11cP-Cre mice. Data correspond to four to five mice per strain type.

FIGURE 5.
FIGURE 5.

I-Eαk affords T1D resistance exclusively via DCs. (A) Systemic removal of I-EkloxP via Cre driven by the β-actin promoter (ACTBP) abrogated I-EkloxP–mediated T1D protection. Data correspond to n = 56 NOD, n = 99 I-EkloxP, and n = 8 for ACTBP-Cre.I-EkloxP mice. (B) Selective removal of I-EkloxP in DCs but not B cells abrogates I-EkloxP–mediated T1D protection in non–TCR-transgenic NOD.I-EkloxP mice. Wild-type NOD (n = 56), NOD.I-EkloxP (n = 99), NOD.I-EB flox (n = 40), and NOD.I-EDC flox (n = 14) females were followed for diabetes until 32 wk of age. (C) Selective removal of I-EkloxP in DCs but not B cells abrogates I-EkloxP–mediated resistance to insulitis in non–TCR-transgenic NOD.I-EkloxP mice. Pancreata from-32-wk-old wild-type NOD (n = 3), NOD.I-Eαk (n = 2), NOD.I-EkloxP (n = 5), NOD.I-EB flox (n = 5), and NOD.I-EDC flox (n = 7) females were sectioned, stained with H&E, and scored for insulitis. The average insulitis scores for the different groups were 1.95 ± 0.32, 0.55 ± 0.25, 0.48 ± 0.23, 0.62 ± 0.37, and 2.93 ± 0.19, respectively. The insulitis scores of NOD.I-Eαk and NOD.I-EkloxP mice were pooled for statistical analysis. (D and E) Selective removal of I-EkloxP in DCs also abrogates I-EkloxP–mediated T1D protection in 4.1-NOD.I-EkloxP (D) and 8.3-NOD.I-EkloxP mice (E). Data in (D) correspond to 4.1-NOD (n = 25), 4.1-NOD.I-EkloxP (n = 65), 4.1-NOD.I-EB flox (n = 23), and 4.1-NOD.I-EDC flox (n = 6) mice. Data in (E) correspond to 8.3-NOD (n = 598), 8.3-NOD.I-EkloxP (n = 79), 8.3-NOD.I-EB flox (n = 16), and 8.3-NOD.I-EDC flox (n = 13) mice.

To be able to gauge the contribution of CD11c+ or CD19+ APCs to I-EkloxP–induced T1D resistance, we crossed NOD.I-EkloxP mice with NOD mice expressing the Cre recombinase under the CD19 or CD11c promoters (CD19P-Cre and CD11cP-Cre, respectively) and compared the levels of I-E on the surface of splenic CD11c+ and CD19+ cells from NOD.I-EkloxP, NOD.I-EkloxP/CD11cP-Cre, and NOD.I-EkloxP/CD19P-Cre mice (in this paper referred to as NOD.I-EkloxP, NOD.I-EDC flox, and NOD.I-EB flox, respectively). As shown in Fig. 4C, the CD11cP-Cre and CD19P-Cre transgenes specifically reduced I-E expression by 10- to 20-fold on the CD11c+ and CD19+ cells of NOD.I-EDC flox and I-EB flox mice, respectively, without affecting endogenous I-Ag7 expression (Fig. 4D) or I-Ag7–restricted or Kd-restricted Ag presentation (Fig. 4E, 4F). As expected, the expression of Cre in CD11c+ or CD19+ cells of non–I-EkloxP–transgenic NOD.CD11cP-Cre and NOD.CD19P-Cre mice did not alter local presentation of β cell autoantigens in the PLNs in vivo, as measured by dilution of CFSE in adoptively transferred naive 8.3-CD8+ T cells (Fig. 4G). Remarkably, however, selective removal of I-EkloxP in CD11c+, but not CD19+ cells, abrogated diabetes protection in non–TCR-transgenic NOD mice in terms of both disease incidence (Fig. 5B) and islet inflammation (Fig. 5C). Selective removal of I-EkloxP in DCs also abrogated I-E–associated diabetes resistance in 4.1- and 8.3-TCR–transgenic animals (Fig. 5D, 5E). Thus, I-E affords T1D resistance specifically via CD11c+ cells.

I-EkloxP expression in DCs does not impair the intrinsic diabetogenic activity of autoreactive T cells or the ability of DCs to prime naive autoreactive T cells

We next investigated whether Cre and/or I-EkloxP expression suppressed T1D by impairing the development or intrinsic diabetogenic properties of autoreactive T cells by profiling thymic and peripheral T cells from 4.1-NOD mice expressing different combinations of Cre and I-EkloxP transgenes. As shown in Supplementary Fig. 4, 4.1-NOD mice expressing I-EkloxP in the presence or absence of CD19P-Cre or CD11cP-Cre transgenes did not exhibit altered T cell development profiles or TCR expression levels in the thymus (Supplemental Fig. 4A–C) or the spleen (Supplemental Fig. 4D, 4E). Furthermore, CD4+CD25− T cells from each of these strains proliferated equally well and secreted similar levels of IFN-γ in response to stimulation with islet cell suspension in vitro (Supplemental Fig. 4F) and transferred diabetes to NOD.scid mice with an efficiency similar to that seen using CD4+CD25− T cells from 4.1-NOD mice (Supplemental Fig. 4G). Taken together, these data indicated that I-EkloxP expression in DCs impairs neither the intrinsic diabetogenic activity of autoreactive T cells nor the ability of DCs to prime naive autoreactive T cells.

Selective removal of I-E on CD11c+ cells abrogated I-E–potentiated Treg development

We next compared the frequency and function of Foxp3+CD4+ Tregs in 4.1-NOD.I-EkloxP, 4.1-NOD.I-EDC flox, and 4.1-NOD.I-EB flox mice. Like 4.1-NOD.I-Eαk mice, 4.1-NOD.I-EkloxP mice harbored increased percentages of Foxp3+CD4+ in the PLNs, albeit not to the same extent and not in the thymus (compare Figs. 6A and 1E). The lower levels of I-Eαk transgene expression in NOD.I-EkloxP versus NOD.I-Eαk mice (Fig. 4B) are likely responsible for both this observation as well as the lower diabetes resistance of 4.1-NOD.I-EkloxP or 8.3-NOD.I-EkloxP versus 4.1-NOD.I-Eαk and 8.3-NOD.I-Eαk mice (compare Figs. 5D and 1A and Figs. 5E and 2A, respectively), as it would be expected based on previous observations (45). Importantly, selective abrogation of I-EkloxP–induced T1D resistance in 4.1-NOD.I-EDC flox mice (Fig. 5D) correlated with normalization of the Foxp3+CD4+ Treg content in the PLNs (Fig. 6A). Notably, this did not happen in 4.1-NOD.I-EB flox mice (Fig. 6A), which develop T1D essentially like 4.1-NOD.I-EkloxP mice (Fig. 5D). Thus, I-E expression in TCR-transgenic NOD mice potentiates Treg development exclusively via CD11c+ cells.

FIGURE 6.
FIGURE 6.

DC expression of I-E in NOD.I-EkloxP mice potentiates Treg-dependent suppression of autoantigen presentation in the PLNs. (A) Percentages of Foxp3+ T cells in the MLNs, PLNs, and thymic (THY) CD4+CD8− T cells of 4.1-NOD (n = 9), 4.1-NOD.I-EkloxP(n = 12), 4.1-NOD.I-EB flox (n = 11), and 4.1-NOD.I-EDC flox (n = 6) mice. Data presented as means ± SEM. (B) Antidiabetogenic activity of splenic CD4+CD25+ T cells from NOD (n = 17), NOD.I-EkloxP (n = 12), NOD.I-EB flox (n = 7), and NOD.I-EDC flox (n = 7) mice in 8.3-NOD.Rag2−/− hosts, followed for 100 d posttransfer. (C) In vivo proliferation of naive CFSE-labeled 8.3-CD8+ T cells in the PLN and MLN of NOD (n = 5) versus NOD.I-EkloxP (n = 5), NOD.I-EB flox (n = 7), and NOD.I-EDC flox mice (n = 4) at day 7 after cell injection. Shown are representative flow cytometric profiles (left panel) and averaged percentages of CFSE-diluted cells ± SEM (right panel).

I-E expression potentiates Foxp3+CD4+ Treg function in non–TCR-transgenic NOD.I-EkloxP mice

As noted above, the I-EkloxP–induced increase in Treg development/recruitment seen in TCR-transgenic NOD.I-EkloxP mice was less pronounced than that seen in their TCR-transgenic NOD.I-Eαk counterparts, suggesting that the magnitude of the I-EkloxP–induced phenotype is too small to be detected by flow cytometry in the polyclonal PLN repertoire of non–TCR-transgenic mice (data not shown). To ascertain whether NOD.I-EkloxP mice harbor increased frequencies of islet Ag-autoreactive Tregs, as is the case for NOD.I-Eαk versus NOD mice, we compared the ability of purified splenic Tregs from NOD, NOD.I-EkloxP, NOD.I-EB flox, and NOD.I-EDC flox mice to suppress the development of spontaneous autoimmune diabetes in RAG2-deficient (Treg-free) 8.3-NOD mice in an I-E–independent manner. Remarkably, Tregs isolated from strains expressing I-EkloxP on DCs (I-EkloxP and NOD.I-EB flox) were significantly more potent at suppressing 8.3-CD8+ T cell–induced diabetes than those isolated from strains whose DCs do not express I-EkloxP (NOD and NOD.I-EDC flox mice) (Fig. 6B), confirming that I-E induces regulation by promoting, via DCs, the development of endogenous autoregulatory Foxp3+CD4+ Tregs.

Selective removal of I-E on CD11c+ cells of non–TCR-transgenic NOD mice abrogated I-E–potentiated suppression of autoantigen presentation in the PLNs

We have shown that Foxp3+CD4+ Tregs suppress T1D development, at least in part, by inhibiting the maturation of autoantigen-loaded DCs in vivo (41), a concept substantiated by the demonstration that PLN DCs are major cellular targets of autoreactive Tregs in vivo (42). We have also shown in this study that the Foxp3+CD4+ Tregs developing in NOD.I-Eαk mice downregulate MHC class II and CD80 on PLN DCs (Fig. 3D) and that the PLNs of NOD.I-EkloxP mice have a reduced capacity to support the cross-presentation of CFSE-labeled autoreactive CD8+ T cells (Fig. 3D). To investigate whether the latter phenotype might be a consequence of the expression of I-E on CD11c+ cells versus other cell types, we compared the ability of the PLNs of NOD, NOD.I-EkloxP, NOD.I-EB flox, and NOD.I-EDC flox mice to cross-present endogenous IGRP206–214 to CFSE-labeled 8.3-CD8+ T cells. As shown in Fig. 6C, DC-specific removal of I-EkloxP in NOD.I-EDC flox mice restored the capacity of the PLNs of NOD.I-EkloxP mice to support the proliferation of CFSE-labeled 8.3-CD8+ T cells. In contrast, B cell–specific removal of I-EkloxP in NOD.I-EB flox was inconsequential.

Taken together, these data demonstrate that I-E potentiates the development of autoregulatory Tregs via DCs, fueling a feed-forward negative feedback regulatory loop that keeps autoantigen-loaded DCs in an immature state, hence, unable to promote the activation and recruitment of diabetogenic CD4+ and CD8+ T cells.

Discussion

Several mechanisms have been previously proposed to explain how certain MHC class II molecules, such as I-E, afford resistance to certain autoimmune diseases including autoimmune diabetes. The finding that thymocytes expressing Vβ5, Vβ11, Vβ16, or Vβ17a elements undergo central and/or peripheral deletion in I-Eαd or I-Eαk–transgenic mice suggested that I-E might afford T1D resistance by triggering the deletion of certain key autoreactive T cell specificity/ies via endogenous superantigens (14, 48–51). Subsequent reports, however, provided evidence that argued against this view (16, 17). Other lines of evidence suggested that I-E might afford diabetes resistance by skewing the development of certain autoreactive CD4+ T cell specificities into antidiabetogenic IL-4 producers (45, 52). It was later shown, however, that I-E–induced T1D resistance is dissociated from IL-4 expression (53). It was also hypothesized that I-E might inhibit the presentation of key autoantigenic epitopes by the diabetogenic MHC class II molecule I-Ag7 via 1) determinant-stealing/capture, whereby I-E would compete with I-Ag7 for the same epitope or for a polypeptide chain containing both a key I-Ag7–restricted epitope and a separate I-E–restricted one (34, 35), or 2) through blockade of the I-Ag7 peptide binding site by a peptide derived from I-E (36). However, there is no formal evidence that this mechanism plays a role in the resistance of I-E–transgenic NOD mice to developing spontaneous T1D. In fact, APCs from both I-E–transgenic and nontransgenic NOD mice can stimulate diabetogenic CD4+ T cell clones with similar efficiency (50), and I-E expression cannot protect mice against diabetes driven by the I-Ag7–restricted BDC2.5-CD4 T cell clone (16, 54). T cell–transferable dominant tolerance is yet another mechanism that was proposed to account for I-E’s antidiabetogenic properties (20, 55). However, I-E–transgenic and nontransgenic NOD mice appear to harbor similar numbers of Tregs with comparable in vitro regulatory potency (54). In this paper, we revisit this issue using alternative approaches and provide compelling evidence indicating that I-E affords T1D resistance by promoting the differentiation of MHC class II promiscuous (i.e., capable of recognizing more than one MHC class II molecule) autoreactive T cell specificities into autoregulatory CD4+ T cells capable of blunting diabetogenesis in a disease-specific manner, akin to the type of T cell–transferable tolerance proposed by Singer et al. (20, 55) in the 1990s.

Our initial studies of NOD mice expressing a diabetogenic, I-Ag7–restricted TCR (4.1-NOD mice (22)) revealed an association between the MHC class II promiscuity of certain autoreactive CD4+ T cells and the MHC-linked resistance to T1D (23–25). Because a significant fraction of 4.1-NOD.I-Eαk mice are resistant to T1D despite exporting functional autoreactive 4.1-CD4+ T cells to the periphery (25), we investigated whether I-E expression inhibits T1D by promoting the differentiation of other MHC class II promiscuous autoreactive CD4+ thymocytes (4.1-like, but not 4.1) into autoregulatory T cell types. Furthermore, because previous studies had implicated different APC subsets as mediators of I-E–associated resistance to T1D (17, 25, 46, 53, 56, 57), we sought to address whether I-E expression afforded T1D resistance via one or more APC types. This was done by expressing an I-EαkloxP transgene in NOD mice, alone and in combination with highly diabetogenic I-Ag7– or Kd–restricted TCRs and/or ubiquitous or cell type–specific Cre recombinase transgenes, and by exploring the effects of transgene expression on Treg development vis-à-vis autoreactive CD4+ and CD8+ T cell development, activation, and recruitment. Our data show that I-E affords dominant resistance to T1D to both TCR-transgenic and nontransgenic NOD mice in a DC-dependent manner by promoting the development of autoregulatory CD4+Foxp3+ Tregs from endogenous MHC class II promiscuous precursors. Our data clearly exclude a role for professional APC types proposed in earlier studies, including B cells, CD11c− macrophages, and thymic epithelial cells, although they cannot tell us whether the effects of I-E are mediated by a specific subset of DCs. Nevertheless, the observations that I-Eαk expression is associated with enhanced development of thymic Tregs, and the increased ability of thymic DCs purified from NOD.I-Eαk mice to promote thymocyte Treg development in vitro suggests that thymic DCs play a critical role.

Using two different, T1D-relevant TCR-transgenic NOD models, we show that I-E expression on DCs does not have intrinsic effects on DC function, on I-Ag7– or Kd-restricted autoantigen presentation, or on the diabetogenic potential of effector CD8+ or 4.1-CD4+CD25− T cells. In fact, I-E–mediated Treg induction and diabetes resistance in these TCR-transgenic mice required the expression of endogenous TCR rearrangements. We note that this phenotype differs from the one we have seen in I-Ab–transgenic 4.1-NOD mice, where selective expression of I-Ab on CD11c+ cells triggers the deletion of 4.1 thymocytes and their differentiation into Tregs in both RAG-sufficient and RAG-deficient 4.1-TCR–transgenic mice (47). The preferential accumulation of these I-E–induced Tregs in the PLNs support the idea that these Tregs are β cell–autoreactive, hence, retained by PLN-associated, autoantigen-loaded APCs. This would explain why these Tregs can suppress the presentation of both cognate and noncognate β cell Ags to diabetogenic CD4+ and CD8+ T cells by local APCs, as we and others have documented experimentally both in this study and elsewhere (41, 42, 44). In support of this interpretation of the data, polyclonal Tregs arising in NOD.I-Eαk mice accumulate in the PLNs of wild-type NOD hosts significantly more efficiently than Tregs arising in NOD mice, suggesting that the former harbor increased frequencies of islet Ag-autoreactive Tregs. This interpretation is consistent with the observation that islet Ag-specific Tregs have superior ability to accumulate in the PLNs (58) and are superior at suppressing autoimmunity as compared with polyclonal, nonspecific Tregs (42, 59). Our data further suggest that these autoreactive Tregs need not have to engage I-E on local DCs to suppress autoantigen presentation and that the role of I-E is purely developmental in nature (i.e., promoting Treg development but not required for Treg function). Collectively, our data indicate that certain I-Ag7–restricted autoreactive CD4+ T cells (i.e., positively selected by peptide–I-Ag7 complexes expressed on thymic cortical epithelial cells) differentiate into autoreactive regulatory (autoregulatory) CD4+ T cells upon engaging cross-reactive I-E complexes on DCs.

There is evidence suggesting that MHC promiscuity is not rare among autoreactive TCRs and certainly that it is not a peculiarity of the 4.1-TCR. The diabetogenic MHC class I-restricted AI4-TCR is thought to recognize I-E (39). Selective downregulation of the 8.3-TCR in thymocytes from 8.3-NOD.I-Eαk mice suggests that the 8.3-TCR also crossreacts with I-E. This is not caused by an endogenous superantigen binding to the 8.3-TCRβ-chain, because this phenotype has not been seen in transgenic NOD mice expressing a low affinity IGRP206–214-specific TCR that uses the same TCRβ-chain and an almost identical TCRα rearrangement as the 8.3-TCR (S. Tsai and P. Santamaria, unpublished observations). Furthermore, an encephalitogenic, myelin proteolipid protein–specific TCR has been shown to cause autoimmune disease in the context of HLA-A3 but to undergo tolerance in the presence of HLA-A2 (60). In addition, structural studies have shown that, as compared with foreign Ag-specific TCRs, autoreactive T cells tend to either bind their cognate pMHC ligands with a unique topology or to recognize a peptide that occupies only part of the peptide binding site of the presenting MHC molecule (61). These TCRs make fewer contacts with peptide residues (61) and appear to catalyze the formation of unconventional synapse structures as compared with foreign Ag-specific TCRs (62). Such an altered mode of binding may account for the MHC promiscuity of certain autoreactive TCRs, whereby polymorphic MHC residues on other (nonselecting) MHC molecules would compensate for the paucity of molecular contacts with the selecting pMHC complex(es). Certain features of known T1D-relevant TCRs support the existence of such pMHC recognition anomalies. For example, nearly all IGRP206–214-reactive TCRs use identical TCRα chains (including their CDR3α regions) but a highly diverse repertoire of TCRβ-chains (30, 63), suggesting that most contacts with peptide are made via the TCRα-chain. Likewise, B:9-23 autoreactive TCRs generally use a single Vα element in their TCRα-chain (64), and the BDC2.5-TCR recognizes several chromogranin A–derived peptides (64, 65), some of which bind to the I-Ag7 molecule in a very unusual fashion (64).

Whatever the structural mechanisms, we demonstrate that the diabetes-protective MHC class II I-E mediates disease protection by inducing immune regulation via autoreactive Tregs, a process in which DCs play a centric role. In addition, we propose a unifying concept for the phenomena of MHC-linked T1D resistance and susceptibility. That is, protective MHC class II molecules act by engaging autoreactive TCRs that are intrinsically MHC promiscuous. This built-in feature (MHC promiscuity) may contribute to the pathogenicity of these TCRs in genetic backgrounds that fail to censor their development but affords protective MHC class II molecules the opportunity to subvert their autoreactivity for the benefit of the host—by turning pathogenic autoreactivity into autoregulation. It is undisputed that genetic diversity at the MHC loci evolved to ensure survival by accommodating immune recognition of a diverse array of infectious agents; from a different perspective, the same MHC diversity may also be viewed as nature’s design against T cell–mediated autoimmunity.

Disclosures

The authors have no financial conflicts of interest.

Acknowledgments

We thank J. Luces and R. Barasi for technical assistance, L. Kennedy, L. Robertson, and Y. Liu for flow cytometry, B. Reizis for providing the C57BL/6.CD11cP-Cre mice, and A. Ziegler for preparing Fig. 3A.

Footnotes

  • This work was funded by the Canadian Institutes of Health Research. S.T. and X.C.-C. were supported by studentships from Alberta Innovates – Health Solutions (formerly the Alberta Heritage Foundation for Medical Research) and the AXA Research Fund, respectively. P. Serra is an investigator of the Ramon y Cajal Reintegration Program of the Ministerio de Economia y Competitividad of Spain. P. Santamaria is a scientist of Alberta Innovates – Health Solutions and a scholar of the Juvenile Diabetes Research Foundation and the Instituto de Investigaciones Sanitarias Carlos III. The Julia McFarlane Diabetes Research Centre was supported by the Diabetes Association (Foothills) and the Canadian Diabetes Association.

  • The online version of this article contains supplemental material.

  • Abbreviations used in this article:

    DC
    dendritic cell
    DP
    double-positive
    MLN
    mesenteric draining lymph node
    PLN
    pancreatic draining lymph node
    pMHC
    peptide–MHC
    SP
    single-positive
    T1D
    type 1 diabetes
    THY
    thymic/thymocyte
    Treg
    regulatory T cell
    TSLP
    thymic stromal lymphopoietin.

  • Received January 16, 2013.
  • Accepted April 26, 2013.

References

  1. ↵
    1. Wang W. Y.,
    2. B. J. Barratt,
    3. D. G. Clayton,
    4. J. A. Todd
    . 2005. Genome-wide association studies: theoretical and practical concerns. Nat. Rev. Genet. 6: 109–118.
    OpenUrlCrossRefPubMed
    1. Barton N. H.,
    2. P. D. Keightley
    . 2002. Understanding quantitative genetic variation. Nat. Rev. Genet. 3: 11–21.
    OpenUrlCrossRefPubMed
    1. MacKay K.,
    2. S. Eyre,
    3. A. Myerscough,
    4. A. Milicic,
    5. A. Barton,
    6. S. Laval,
    7. J. Barrett,
    8. D. Lee,
    9. S. White,
    10. S. John,
    11. et al
    . 2002. Whole-genome linkage analysis of rheumatoid arthritis susceptibility loci in 252 affected sibling pairs in the United Kingdom. Arthritis Rheum. 46: 632–639.
    OpenUrlCrossRefPubMed
    1. Hafler D. A.,
    2. A. Compston,
    3. S. Sawcer,
    4. E. S. Lander,
    5. M. J. Daly,
    6. P. L. De Jager,
    7. P. I. de Bakker,
    8. S. B. Gabriel,
    9. D. B. Mirel,
    10. A. J. Ivinson,
    11. et al
    . 2007. Risk alleles for multiple sclerosis identified by a genomewide study. N. Engl. J. Med. 357: 851–862.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Nejentsev S.,
    2. J. M. Howson,
    3. N. M. Walker,
    4. J. Szeszko,
    5. S. F. Field,
    6. H. E. Stevens,
    7. P. Reynolds,
    8. M. Hardy,
    9. E. King,
    10. J. Masters,
    11. et al
    . 2007. Localization of type 1 diabetes susceptibility to the MHC class I genes HLA-B and HLA-A. Nature 450: 887–892.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Erlich H.,
    2. A. M. Valdes,
    3. J. Noble,
    4. J. A. Carlson,
    5. M. Varney,
    6. P. Concannon,
    7. J. C. Mychaleckyj,
    8. J. A. Todd,
    9. P. Bonella,
    10. A. L. Fear,
    11. et al
    . 2008. HLA DR-DQ haplotypes and genotypes and type 1 diabetes risk: analysis of the type 1 diabetes genetics consortium families. Diabetes 57: 1084–1092.
    OpenUrlAbstract/FREE Full Text
  7. ↵
    1. Todd J. A.,
    2. J. I. Bell,
    3. H. O. McDevitt
    . 1987. HLA-DQ β gene contributes to susceptibility and resistance to insulin-dependent diabetes mellitus. Nature 329: 599–604.
    OpenUrlCrossRefPubMed
  8. ↵
    1. Howson J. M.,
    2. N. M. Walker,
    3. D. Clayton,
    4. J. A. Todd, and,
    5. Type 1 Diabetes Genetics Consortium
    . 2009. Confirmation of HLA class II independent type 1 diabetes associations in the major histocompatibility complex including HLA-B and HLA-A. Diabetes Obes. Metab. 11(Suppl. 1): 31–45.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Tisch R.,
    2. H. McDevitt
    . 1996. Insulin-dependent diabetes mellitus. Cell 85: 291–297.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Vyse T. J.,
    2. J. A. Todd
    . 1996. Genetic analysis of autoimmune disease. Cell 85: 311–318.
    OpenUrlCrossRefPubMed
  11. ↵
    1. Acha-Orbea H.,
    2. H. O. McDevitt
    . 1987. The first external domain of the nonobese diabetic mouse class II I-A β chain is unique. Proc. Natl. Acad. Sci. USA 84: 2435–2439.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    1. Uehira M.,
    2. M. Uno,
    3. T. Kürner,
    4. H. Kikutani,
    5. K. Mori,
    6. T. Inomoto,
    7. T. Uede,
    8. J. Miyazaki,
    9. H. Nishimoto,
    10. T. Kishimoto,
    11. et al
    . 1989. Development of autoimmune insulitis is prevented in Eα d but not in Aβk NOD transgenic mice. Int. Immunol. 1: 209–213.
    OpenUrlAbstract/FREE Full Text
    1. Lund T.,
    2. L. O’Reilly,
    3. P. Hutchings,
    4. O. Kanagawa,
    5. E. Simpson,
    6. R. Gravely,
    7. P. Chandler,
    8. J. Dyson,
    9. J. K. Picard,
    10. A. Edwards,
    11. et al
    . 1990. Prevention of insulin-dependent diabetes mellitus in non-obese diabetic mice by transgenes encoding modified I-A β-chain or normal I-E α-chain. Nature 345: 727–729.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Reich E. P.,
    2. R. S. Sherwin,
    3. O. Kanagawa,
    4. C. A. Janeway Jr..
    1989. An explanation for the protective effect of the MHC class II I-E molecule in murine diabetes. Nature 341: 326–328.
    OpenUrlCrossRefPubMed
  15. ↵
    1. Miyazaki T.,
    2. M. Uno,
    3. M. Uehira,
    4. H. Kikutani,
    5. T. Kishimoto,
    6. M. Kimoto,
    7. H. Nishimoto,
    8. J. Miyazaki,
    9. K. Yamamura
    . 1990. Direct evidence for the contribution of the unique I-ANOD to the development of insulitis in non-obese diabetic mice. Nature 345: 722–724.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Katz J. D.,
    2. B. Wang,
    3. K. Haskins,
    4. C. Benoist,
    5. D. Mathis
    . 1993. Following a diabetogenic T cell from genesis through pathogenesis. Cell 74: 1089–1100.
    OpenUrlCrossRefPubMed
  17. ↵
    1. Böhme J.,
    2. B. Schuhbaur,
    3. O. Kanagawa,
    4. C. Benoist,
    5. D. Mathis
    . 1990. MHC-linked protection from diabetes dissociated from clonal deletion of T cells. Science 249: 293–295.
    OpenUrlAbstract/FREE Full Text
    1. Slattery R. M.,
    2. L. Kjer-Nielsen,
    3. J. Allison,
    4. B. Charlton,
    5. T. E. Mandel,
    6. J. F. Miller
    . 1990. Prevention of diabetes in non-obese diabetic I-Ak transgenic mice. Nature 345: 724–726.
    OpenUrlCrossRefPubMed
    1. Wicker L. S.,
    2. M. C. Appel,
    3. F. Dotta,
    4. A. Pressey,
    5. B. J. Miller,
    6. N. H. DeLarato,
    7. P. A. Fischer,
    8. R. C. Boltz Jr..,
    9. L. B. Peterson
    . 1992. Autoimmune syndromes in major histocompatibility complex (MHC) congenic strains of nonobese diabetic (NOD) mice: the NOD MHC is dominant for insulitis and cyclophosphamide-induced diabetes. J. Exp. Med. 176: 67–77.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Singer S. M.,
    2. R. Tisch,
    3. X. D. Yang,
    4. H. O. McDevitt
    . 1993. An Abd transgene prevents diabetes in nonobese diabetic mice by inducing regulatory T cells. Proc. Natl. Acad. Sci. USA 90: 9566–9570.
    OpenUrlAbstract/FREE Full Text
  21. ↵
    1. Parish N. M.,
    2. P. Chandler,
    3. R. Quartey-Papafio,
    4. E. Simpson,
    5. A. Cooke
    . 1993. The effect of bone marrow and thymus chimerism between non-obese diabetic (NOD) and NOD-E transgenic mice, on the expression and prevention of diabetes. Eur. J. Immunol. 23: 2667–2675.
    OpenUrlCrossRefPubMed
  22. ↵
    1. Verdaguer J.,
    2. D. Schmidt,
    3. A. Amrani,
    4. B. Anderson,
    5. N. Averill,
    6. P. Santamaria
    . 1997. Spontaneous autoimmune diabetes in monoclonal T cell nonobese diabetic mice. J. Exp. Med. 186: 1663–1676.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    1. Schmidt D.,
    2. J. Verdaguer,
    3. N. Averill,
    4. P. Santamaria
    . 1997. A mechanism for the major histocompatibility complex-linked resistance to autoimmunity. J. Exp. Med. 186: 1059–1075.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Schmidt D.,
    2. A. Amrani,
    3. J. Verdaguer,
    4. S. Bou,
    5. P. Santamaria
    . 1999. Autoantigen-independent deletion of diabetogenic CD4+ thymocytes by protective MHC class II molecules. J. Immunol. 162: 4627–4636.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Thiessen S.,
    2. P. Serra,
    3. A. Amrani,
    4. J. Verdaguer,
    5. P. Santamaria
    . 2002. T-cell tolerance by dendritic cells and macrophages as a mechanism for the major histocompatibility complex-linked resistance to autoimmune diabetes. Diabetes 51: 325–338.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Kanagawa O.,
    2. B. A. Vaupel,
    3. G. Xu,
    4. E. R. Unanue,
    5. J. D. Katz
    . 1998. Thymic positive selection and peripheral activation of islet antigen-specific T cells: separation of two diabetogenic steps by an I-A(g7) class II MHC beta-chain mutant. J. Immunol. 161: 4489–4492.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    1. Lühder F.,
    2. J. Katz,
    3. C. Benoist,
    4. D. Mathis
    . 1998. Major histocompatibility complex class II molecules can protect from diabetes by positively selecting T cells with additional specificities. J. Exp. Med. 187: 379–387.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Rickert R. C.,
    2. J. Roes,
    3. K. Rajewsky
    . 1997. B lymphocyte-specific, Cre-mediated mutagenesis in mice. Nucleic Acids Res. 25: 1317–1318.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Caton M. L.,
    2. M. R. Smith-Raska,
    3. B. Reizis
    . 2007. Notch-RBP-J signaling controls the homeostasis of CD8- dendritic cells in the spleen. J. Exp. Med. 204: 1653–1664.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Anderson B.,
    2. B. J. Park,
    3. J. Verdaguer,
    4. A. Amrani,
    5. P. Santamaria
    . 1999. Prevalent CD8+ T cell response against one peptide/MHC complex in autoimmune diabetes. Proc. Natl. Acad. Sci. USA 96: 9311–9316.
    OpenUrlAbstract/FREE Full Text
  31. ↵
    1. Stratmann T.,
    2. N. Martin-Orozco,
    3. V. Malet-Designe,
    4. L. Poirot,
    5. D. McGavern,
    6. G. Losyev,
    7. C. M. Dobbs,
    8. M. B. Oldstone,
    9. K. Yoshida,
    10. H. Kikutani,
    11. et al
    . 2003. Susceptible MHC alleles, not background genes, select an autoimmune T cell reactivity. J. Clin. Invest. 112: 902–914.
    OpenUrlCrossRefPubMed
  32. ↵
    1. Dash P.,
    2. J. L. McClaren,
    3. T. H. Oguin III.,
    4. W. Rothwell,
    5. B. Todd,
    6. M. Y. Morris,
    7. J. Becksfort,
    8. C. Reynolds,
    9. S. A. Brown,
    10. P. C. Doherty,
    11. P. G. Thomas
    . 2011. Paired analysis of TCRα and TCRβ chains at the single-cell level in mice. J. Clin. Invest. 121: 288–295.
    OpenUrlCrossRefPubMed
  33. ↵
    1. Verdaguer J.,
    2. J. W. Yoon,
    3. B. Anderson,
    4. N. Averill,
    5. T. Utsugi,
    6. B. J. Park,
    7. P. Santamaria
    . 1996. Acceleration of spontaneous diabetes in TCR-β-transgenic nonobese diabetic mice by β-cell cytotoxic CD8+ T cells expressing identical endogenous TCR-α chains. J. Immunol. 157: 4726–4735.
    OpenUrlAbstract
  34. ↵
    1. Deng H.,
    2. R. Apple,
    3. M. Clare-Salzler,
    4. S. Trembleau,
    5. D. Mathis,
    6. L. Adorini,
    7. E. Sercarz
    . 1993. Determinant capture as a possible mechanism of protection afforded by major histocompatibility complex class II molecules in autoimmune disease. J. Exp. Med. 178: 1675–1680.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Roberts S. A.,
    2. G. Barbour,
    3. M. R. Matarrese,
    4. D. L. Mason,
    5. E. H. Leiter,
    6. K. Haskins,
    7. M. S. Hanson
    . 2003. Adoptive transfer of islet antigen-autoreactive T cell clones to transgenic NOD.Ea(d)mice induces diabetes indicating a lack of I-E mediated protection against activated effector T cells. J. Autoimmun. 21: 139–147.
    OpenUrlCrossRefPubMed
  36. ↵
    1. Trembleau S.,
    2. S. Gregori,
    3. G. Penna,
    4. I. Gorny,
    5. L. Adorini
    . 2001. IL-12 administration reveals diabetogenic T cells in genetically resistant I-Eα-transgenic nonobese diabetic mice: resistance to autoimmune diabetes is associated with binding of Eα-derived peptides to the I-A(g7) molecule. J. Immunol. 167: 4104–4114.
    OpenUrlAbstract/FREE Full Text
  37. ↵
    1. Amrani A.,
    2. J. Verdaguer,
    3. P. Serra,
    4. S. Tafuro,
    5. R. Tan,
    6. P. Santamaria
    . 2000. Progression of autoimmune diabetes driven by avidity maturation of a T-cell population. Nature 406: 739–742.
    OpenUrlCrossRefPubMed
  38. ↵
    1. Lieberman S. M.,
    2. A. M. Evans,
    3. B. Han,
    4. T. Takaki,
    5. Y. Vinnitskaya,
    6. J. A. Caldwell,
    7. D. V. Serreze,
    8. J. Shabanowitz,
    9. D. F. Hunt,
    10. S. G. Nathenson,
    11. et al
    . 2003. Identification of the β cell antigen targeted by a prevalent population of pathogenic CD8+ T cells in autoimmune diabetes. Proc. Natl. Acad. Sci. USA 100: 8384–8388.
    OpenUrlAbstract/FREE Full Text
  39. ↵
    1. Serreze D. V.,
    2. T. M. Holl,
    3. M. P. Marron,
    4. R. T. Graser,
    5. E. A. Johnson,
    6. C. Choisy-Rossi,
    7. R. M. Slattery,
    8. S. M. Lieberman,
    9. T. P. DiLorenzo
    . 2004. MHC class II molecules play a role in the selection of autoreactive class I-restricted CD8 T cells that are essential contributors to type 1 diabetes development in nonobese diabetic mice. J. Immunol. 172: 871–879.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Hanabuchi S.,
    2. T. Ito,
    3. W. R. Park,
    4. N. Watanabe,
    5. J. L. Shaw,
    6. E. Roman,
    7. K. Arima,
    8. Y. H. Wang,
    9. K. S. Voo,
    10. W. Cao,
    11. Y. J. Liu
    . 2010. Thymic stromal lymphopoietin-activated plasmacytoid dendritic cells induce the generation of FOXP3+ regulatory T cells in human thymus. J. Immunol. 184: 2999–3007.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    1. Serra P.,
    2. A. Amrani,
    3. J. Yamanouchi,
    4. B. Han,
    5. S. Thiessen,
    6. T. Utsugi,
    7. J. Verdaguer,
    8. P. Santamaria
    . 2003. CD40 ligation releases immature dendritic cells from the control of regulatory CD4+CD25+ T cells. Immunity 19: 877–889.
    OpenUrlCrossRefPubMed
  42. ↵
    1. Tang Q.,
    2. J. Y. Adams,
    3. A. J. Tooley,
    4. M. Bi,
    5. B. T. Fife,
    6. P. Serra,
    7. P. Santamaria,
    8. R. M. Locksley,
    9. M. F. Krummel,
    10. J. A. Bluestone
    . 2006. Visualizing regulatory T cell control of autoimmune responses in nonobese diabetic mice. Nat. Immunol. 7: 83–92.
    OpenUrlCrossRefPubMed
    1. Yamanouchi J.,
    2. M. C. Puertas,
    3. J. Verdaguer,
    4. P. A. Lyons,
    5. D. B. Rainbow,
    6. G. Chamberlain,
    7. K. M. Hunter,
    8. L. B. Peterson,
    9. L. S. Wicker,
    10. P. Santamaria
    . 2010. Idd9.1 locus controls the suppressive activity of FoxP3+CD4+CD25+ regulatory T-cells. Diabetes 59: 272–281.
    OpenUrlAbstract/FREE Full Text
  44. ↵
    1. Yamanouchi J.,
    2. D. Rainbow,
    3. P. Serra,
    4. S. Howlett,
    5. K. Hunter,
    6. V. E. Garner,
    7. A. Gonzalez-Munoz,
    8. J. Clark,
    9. R. Veijola,
    10. R. Cubbon,
    11. et al
    . 2007. Interleukin-2 gene variation impairs regulatory T cell function and causes autoimmunity. Nat. Genet. 39: 329–337.
    OpenUrlCrossRefPubMed
  45. ↵
    1. Hanson M. S.,
    2. M. Cetkovic-Cvrlje,
    3. V. K. Ramiya,
    4. M. A. Atkinson,
    5. N. K. Maclaren,
    6. B. Singh,
    7. J. F. Elliott,
    8. D. V. Serreze,
    9. E. H. Leiter
    . 1996. Quantitative thresholds of MHC class II I-E expressed on hemopoietically derived antigen-presenting cells in transgenic NOD/Lt mice determine level of diabetes resistance and indicate mechanism of protection. J. Immunol. 157: 1279–1287.
    OpenUrlAbstract
  46. ↵
    1. Pilström B.,
    2. J. Böhme
    . 1997. Alleviation of insulitis in NOD mice is associated with expression of transgenic MHC E molecules on primary antigen-presenting cells. Immunology 90: 483–488.
    OpenUrlCrossRefPubMed
  47. ↵
    1. Tsai S.,
    2. P. Serra,
    3. X. Clemente-Casares,
    4. J. Yamanouchi,
    5. S. Thiessen,
    6. R. M. Slattery,
    7. J. F. Elliott,
    8. P. Santamaria
    . 2013. Antidiabetogenic MHC class II promotes the differentiation of MHC-promiscuous autoreactive T cells into FOXP3+ regulatory T cells. Proc. Natl. Acad. Sci. USA 110: 3471–3476.
    OpenUrlAbstract/FREE Full Text
  48. ↵
    1. Bill J.,
    2. V. B. Appel,
    3. E. Palmer
    . 1988. An analysis of T-cell receptor variable region gene expression in major histocompatibility complex disparate mice. Proc. Natl. Acad. Sci. USA 85: 9184–9188.
    OpenUrlAbstract/FREE Full Text
    1. Tomonari K.,
    2. E. Lovering
    . 1988. T-cell receptor-specific monoclonal antibodies against a Vβ11-positive mouse T-cell clone. Immunogenetics 28: 445–451.
    OpenUrlCrossRefPubMed
  50. ↵
    1. Nakano N.,
    2. H. Kikutani,
    3. H. Nishimoto,
    4. T. Kishimoto
    . 1991. T cell receptor V gene usage of islet β cell-reactive T cells is not restricted in non-obese diabetic mice. J. Exp. Med. 173: 1091–1097.
    OpenUrlAbstract/FREE Full Text
  51. ↵
    1. Kappler J. W.,
    2. N. Roehm,
    3. P. Marrack
    . 1987. T cell tolerance by clonal elimination in the thymus. Cell 49: 273–280.
    OpenUrlCrossRefPubMed
  52. ↵
    1. Brenden N.,
    2. J. Böhme
    . 1998. Disease-protected major histocompatibility complex Ea-transgenic non-obese diabetic (NOD) mice show interleukin-4 production not seen in susceptible Ea-transgenic and non-transgenic NOD mice. Immunology 95: 1–7.
    OpenUrlCrossRefPubMed
  53. ↵
    1. Johnson E. A.,
    2. P. Silveira,
    3. H. D. Chapman,
    4. E. H. Leiter,
    5. D. V. Serreze
    . 2001. Inhibition of autoimmune diabetes in nonobese diabetic mice by transgenic restoration of H2-E MHC class II expression: additive, but unequal, involvement of multiple APC subtypes. J. Immunol. 167: 2404–2410.
    OpenUrlAbstract/FREE Full Text
  54. ↵
    1. Mellanby R. J.,
    2. J. M. Phillips,
    3. N. M. Parish,
    4. A. Cooke
    . 2008. Both central and peripheral tolerance mechanisms play roles in diabetes prevention in NOD-E transgenic mice. Autoimmunity 41: 383–394.
    OpenUrlCrossRefPubMed
  55. ↵
    1. Singer S. M.,
    2. R. Tisch,
    3. X. D. Yang,
    4. H. K. Sytwu,
    5. R. Liblau,
    6. H. O. McDevitt
    . 1998. Prevention of diabetes in NOD mice by a mutated I-Ab transgene. Diabetes 47: 1570–1577.
    OpenUrlAbstract
  56. ↵
    1. O’Shea H.,
    2. N. Yousaf,
    3. D. Altmann,
    4. Z. Fehervari,
    5. P. Tonks,
    6. C. Hetherington,
    7. S. Harach,
    8. C. Bland,
    9. A. Cooke,
    10. T. Lund
    . 2006. Effect of X- and Y-box deletions on the development of diabetes in H-2Eα-chain transgenic nonobese diabetic mice. Scand. J. Immunol. 63: 17–25.
    OpenUrlCrossRefPubMed
  57. ↵
    1. Chen Y. G.,
    2. P. A. Silveira,
    3. M. A. Osborne,
    4. H. D. Chapman,
    5. D. V. Serreze
    . 2007. Cellular expression requirements for inhibition of type 1 diabetes by a dominantly protective major histocompatibility complex haplotype. Diabetes 56: 424–430.
    OpenUrlAbstract/FREE Full Text
  58. ↵
    1. Samy E. T.,
    2. L. A. Parker,
    3. C. P. Sharp,
    4. K. S. Tung
    . 2005. Continuous control of autoimmune disease by antigen-dependent polyclonal CD4+CD25+ regulatory T cells in the regional lymph node. J. Exp. Med. 202: 771–781.
    OpenUrlAbstract/FREE Full Text
  59. ↵
    1. Tarbell K. V.,
    2. S. Yamazaki,
    3. K. Olson,
    4. P. Toy,
    5. R. M. Steinman
    . 2004. CD25+CD4+ T cells, expanded with dendritic cells presenting a single autoantigenic peptide, suppress autoimmune diabetes. J. Exp. Med. 199: 1467–1477.
    OpenUrlAbstract/FREE Full Text
  60. ↵
    1. Friese M. A.,
    2. K. B. Jakobsen,
    3. L. Friis,
    4. R. Etzensperger,
    5. M. J. Craner,
    6. R. M. McMahon,
    7. L. T. Jensen,
    8. V. Huygelen,
    9. E. Y. Jones,
    10. J. I. Bell,
    11. L. Fugger
    . 2008. Opposing effects of HLA class I molecules in tuning autoreactive CD8+ T cells in multiple sclerosis. Nat. Med. 14: 1227–1235.
    OpenUrlCrossRefPubMed
  61. ↵
    1. Wucherpfennig K. W.,
    2. M. J. Call,
    3. L. Deng,
    4. R. Mariuzza
    . 2009. Structural alterations in peptide-MHC recognition by self-reactive T cell receptors. Curr. Opin. Immunol. 21: 590–595.
    OpenUrlCrossRefPubMed
  62. ↵
    1. Schubert D. A.,
    2. S. Gordo,
    3. J. J. Sabatino Jr..,
    4. S. Vardhana,
    5. E. Gagnon,
    6. D. K. Sethi,
    7. N. P. Seth,
    8. K. Choudhuri,
    9. H. Reijonen,
    10. G. T. Nepom,
    11. et al
    . 2012. Self-reactive human CD4 T cell clones form unusual immunological synapses. J. Exp. Med. 209: 335–352.
    OpenUrlAbstract/FREE Full Text
  63. ↵
    1. Han B.,
    2. P. Serra,
    3. J. Yamanouchi,
    4. A. Amrani,
    5. J. F. Elliott,
    6. P. Dickie,
    7. T. P. Dilorenzo,
    8. P. Santamaria
    . 2005. Developmental control of CD8 T cell-avidity maturation in autoimmune diabetes. J. Clin. Invest. 115: 1879–1887.
    OpenUrlCrossRefPubMed
  64. ↵
    1. Stadinski B. D.,
    2. T. Delong,
    3. N. Reisdorph,
    4. R. Reisdorph,
    5. R. L. Powell,
    6. M. Armstrong,
    7. J. D. Piganelli,
    8. G. Barbour,
    9. B. Bradley,
    10. F. Crawford,
    11. et al
    . 2010. Chromogranin A is an autoantigen in type 1 diabetes. Nat. Immunol. 11: 225–231.
    OpenUrlCrossRefPubMed
  65. ↵
    1. Nikoopour E.,
    2. C. Sandrock,
    3. K. Huszarik,
    4. O. Krougly,
    5. E. Lee-Chan,
    6. E. L. Masteller,
    7. J. A. Bluestone,
    8. B. Singh
    . 2011. Cutting edge: vasostatin-1‑derived peptide ChgA29-42 is an antigenic epitope of diabetogenic BDC2.5 T cells in nonobese diabetic mice. J. Immunol. 186: 3831–3835.
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section