Deciphering mechanisms involved in failure of self tolerance to preproinsulin-2 is a key issue in type 1 diabetes. We used nonautoimmune 129SV/Pas mice lacking preproinsulin-2 to study the immune response to preproinsulin-2. In these mice, a T cell response was detected after immunization with several preproinsulin-2 peptides and confirmed by generating hybridomas. Activation of some of these hybridomas by wild-type (wt) islet cells or recombinant murine proinsulin-2 demonstrated that two epitopes can be generated from the naturally expressed protein. Although T cells from wt mice responded to preproinsulin-2 peptides, we could not detect a response to the naturally processed epitopes in these mice. Moreover, after immunization with recombinant whole proinsulin-2, a T cell response was detected in preproinsulin-2-deficient but not in wt mice. This suggests that islet preproinsulin-2-autoreactive T cells are functionally eliminated in wt mice. We used a transplantation model to evaluate the relevance of reactivity to preproinsulin-2 in vivo. Wild-type preproinsulin-2-expressing islets transplanted in preproinsulin-2-deficient mice elicited a mononuclear cell infiltration and insulin Abs. Graft infiltration was further increased by immunization with preproinsulin-2 peptides. Preproinsulin-2 expression thus shapes the immune response and prevents self reactivity to the islet. Moreover, islet preproinsulin-2 primes an immune response to preproinsulin-2 in deficient mice.

Autoimmunity results from the failure of the different mechanisms maintaining immune tolerance to autoantigens (1, 2, 3). Understanding these mechanisms is essential to decipher how autoimmunity occurs and design strategies to prevent its development. Type 1 diabetes, which results from loss of tolerance to β cell autoantigens, has been extensively studied as a model disease for development of autoimmunity (4, 5, 6, 7). Studies in the nonobese diabetic (NOD)3 mouse (4), a spontaneous model of type 1 diabetes, have identified proinsulin, glutamate decarboxylase, tyrosine phosphatase, and several other β cell autoantigens as targets of autoimmunity in this model. Autoimmunity in human diabetes and in the NOD mouse results from a multigenic process in which numerous immune and β cell defects associate to drive the diabetogenic process (5).

Several models have been established to address the issue of self tolerance to specific Ags expressed by β cells in the absence of spontaneous autoimmunity. In a majority of models, transgenes expressed on β cells have been used to study how these new self-Ags shape tolerance (8, 9). These studies have greatly enhanced our understanding of immune response regulation, but generalizing their results is limited by artifacts due to the transgenic process. For instance, variable thymic expression of the same transgene in different transgenic mouse lines was observed, leading to differences in the level of autoreactivity after infection with a virus expressing the same Ag (10). Transgenic expression of B7-1 in β cells was able to induce an autoimmune reaction against a β cell transgene but not against β cell native Ags (11).

Proinsulin is thought to be a major autoantigen in type 1 diabetes for several reasons. Insulin (and proinsulin)-specific Abs and autoreactive T cells have been detected in NOD mice and in patients with diabetes or prediabetes. In young children at risk for developing the disease, insulin autoantibodies are detected first (12). Allelic variations in the variable number of tandem repeats region flanking the insulin gene are important determinants of the genetic susceptibility to diabetes (IDDM2) and influence β cell insulin secretion and proinsulin expression in the thymus (13, 14, 15, 16). Protection from diabetes is obtained in the NOD mouse by injecting insulin, the insulin B-chain or B-chain peptides (17, 18, 19). Understanding how loss of tolerance to preproinsulin occurs and how IDDM2 regulates diabetes autoimmunity is therefore essential. In the NOD mouse, a progressive loss of tolerance to multiple autoantigens occurs, and deciphering the respective role of each of them is complex. Therefore, evaluating self tolerance to autoantigens such as preproinsulin in nonautoimmune animals allows a better understanding of the mechanisms involved.

Rodents carry two different preproinsulin genes on distinct chromosomes, encoding for two different proteins, both biologically active but differentially expressed in the islet and in the thymus (20, 21). Mice that are deficient for expression of preproinsulin-2 gene are fully viable due to the compensatory effect of the remaining gene (22). They provide a unique opportunity to study the role of this isoform in maintenance of immune tolerance to preproinsulin and to β cells. Here, we show that mice deficient in preproinsulin-2 have an enhanced B and T cell reactivity to the missing preproinsulin isoform. Moreover, preproinsulin-2 expressing islets are the target of a nondestructive immune response when transplanted in preproinsulin-2-deficient mice.

We used preproinsulin-2 knockout mice (22) (referred to as proins-2−/−), maintained on a pure 129/Sv/Pas/ICo (129) background and devoid of specific murine pathogens. Proins-2−/− mice did not contain the lacZ expression cassette (22) to avoid the interference of immune response to lacZ. Wild-type (wt) 129/Sv/Pas/ICo mice were obtained from Iffa Credo (l’Arbresle, France). Mice were bred as homozygous lines and typed by PCR on tail DNA. We used the following oligonucleotides for PCR amplification (436 bp): forward 5′-GGTGAGTTCTGCCACTGAATTC-3′ and reverse 5′-GGCATCAGCAGCACAGAAGCAA-3′. All animal studies were approved by our institutional review board.

A set of sixteen 15-mer peptides spanning the sequence of preproinsulin-2 and overlapping by 10-aa residues and relevant homologous preproinsulin-1 peptides were synthesized as described (Ref. 23 ; Table I). In addition, we used two chimeric peptides derived from peptide II56–71: in II56–71–A64E, Ala64 residue was replaced by a glutamic acid residue, as in the preproinsulin-1 isoform; in II56–71–G71S, Gly71 residue was replaced by a serine residue, as in the preproinsulin-1 isoform. As control, we used the human β-adrenergic receptor peptide hβ1AR34–57 known to be efficiently presented by I-Ab molecules (24). Mouse proinsulin-2 produced as a fusion protein with a (His) 6 coding segment immediately upstream of the proinsulin sequence was produced, purified, and checked for the presence of endotoxin as described (25).

Table I.

Sequence of peptides used

PeptideSequencea
Mouse preproinsulin-2  
 II1–15 MALWMRFLPLLALLF 
 II7–23 FLPLLALLFLWESHPTQ 
 II14–30 LFLWESHPTQAFVKQHL 
 II20–35 HPTQAFVKQHLCGSHL 
 II26–41 VKQHLCGSHLVEALYL 
 II33–47 SHLVEALYLVCGERG 
 II40–56 YLVCGERGFFYTPMSRR 
 II46–60 RGFFYTPMSRREVED 
 II50–65 YTPMSRREVEDPQVA
 II56–71 REVEDPQVAQLELGGG 
 II61–78 PQVAQLELGGGPGAGDLQ 
 II66–83 LELGGGPGAGDLQTLALE 
 II71–88 GPGAGDLQTLALEVAQQK 
 II79–94 TLALEVAQQKKRGIVD 
 II88–103 KRGIVDQCCTSICSLY 
 II93–110 DQCCTSICSLYQLENYCN 
  
Mouse preproinsulin-1  
 I1–15 MALLVHFLPLLALLA 
 I14–30 LALWEPKPTQAFVKQHL 
 I26–41 VKQHLCGPHLVEALYL 
 I40–56 YLVCGERGFFYTPKSRR 
 I46–60 RGFFYTPKSRREVED 
 I50–65 YTPKSRREVEDPQVEQ 
 I56–71 REVEDPQVEQLELGGS 
  
Preproinsulin-1 and -2  chimeric peptides  
 II56–71-A64E REVEDPQVEQLELGGG 
 II56–71-G71S REVEDPQVAQLELGGS 
  
Control I-Ab restricted  peptide  
 hβ1AR34–57 VPASPPASLLPPASESPEPLSQQW 
PeptideSequencea
Mouse preproinsulin-2  
 II1–15 MALWMRFLPLLALLF 
 II7–23 FLPLLALLFLWESHPTQ 
 II14–30 LFLWESHPTQAFVKQHL 
 II20–35 HPTQAFVKQHLCGSHL 
 II26–41 VKQHLCGSHLVEALYL 
 II33–47 SHLVEALYLVCGERG 
 II40–56 YLVCGERGFFYTPMSRR 
 II46–60 RGFFYTPMSRREVED 
 II50–65 YTPMSRREVEDPQVA
 II56–71 REVEDPQVAQLELGGG 
 II61–78 PQVAQLELGGGPGAGDLQ 
 II66–83 LELGGGPGAGDLQTLALE 
 II71–88 GPGAGDLQTLALEVAQQK 
 II79–94 TLALEVAQQKKRGIVD 
 II88–103 KRGIVDQCCTSICSLY 
 II93–110 DQCCTSICSLYQLENYCN 
  
Mouse preproinsulin-1  
 I1–15 MALLVHFLPLLALLA 
 I14–30 LALWEPKPTQAFVKQHL 
 I26–41 VKQHLCGPHLVEALYL 
 I40–56 YLVCGERGFFYTPKSRR 
 I46–60 RGFFYTPKSRREVED 
 I50–65 YTPKSRREVEDPQVEQ 
 I56–71 REVEDPQVEQLELGGS 
  
Preproinsulin-1 and -2  chimeric peptides  
 II56–71-A64E REVEDPQVEQLELGGG 
 II56–71-G71S REVEDPQVAQLELGGS 
  
Control I-Ab restricted  peptide  
 hβ1AR34–57 VPASPPASLLPPASESPEPLSQQW 
a

For comparison with the homologous preproinsulin-1 sequence; the differences are in bold and the insertions are underlined.

Individual adult mice (12–15 wk old) were immunized at the base of the tail with 50 μg of peptide or 100 μg of recombinant proinsulin-2 emulsified in CFA. Ten days later, draining lymph nodes and spleen were harvested, and cell suspensions were prepared in DMEM containing 5% FCS. Cell suspensions were incubated in triplicate in 96-well plates for 24 h at 5 × 105 cells/well (spleen) or 1 × 105 cells/well plus 5 × 105 irradiated spleen cells from unprimed mice as APCs (lymph nodes), with peptide at a final concentration of 20 μg/ml. Supernatants were harvested 24 h later for IL-2 measurements, and in some experiments culture supernatants obtained from a different parallel culture were harvested at 72 h for IFN-γ, IL-4, and IL-10 measurements.

IL-2 concentrations were measured by proliferation of the IL-2-dependent cell line CTLL-2 or by ELISA. Proliferation of CTLL-2 cells was measured by thymidine incorporation (in counts per minute) or expressed in units of IL-2, derived from a standard curve run at the same time using recombinant human IL-2 (PeproTech, Rocky Hill, NJ). For ELISA, we used Abs 1A12 (IL-2), AN18 (IFN-γ), 11B11 (IL-4), and JES5-2A5 (IL-10) as capture Abs and biotinylated 5H4 (IL-2), R4-6A2 (IFN-γ), BVD6 (IL-4), and SXC-1 (IL-10) as detection Abs (BD PharMingen, San Diego, CA). Briefly, 50 μl of standard recombinant mouse cytokine (R&D Systems, Minneapolis, MN) or culture supernatants were incubated in coated wells for 2 h, washed, and incubated for 1 h with 50 μl of detection Ab, combined with HRP-streptavidin (AMDEX; Amersham Pharmacia Biotech, Piscataway, NJ) and revealed with OPD substrate solution (Sigma-Aldrich, St. Louis, MO). OD490 was measured using an ELISA reader (MRX Microplate Reader; Dynatech Laboratories, Chantilly, VA). Results are expressed as concentrations of cytokine using recombinant mouse cytokine standards.

Nitrocellulose plates (Millipore MAHA S45) were coated overnight at 37°C with relevant Abs (BD PharMingen: IL-2, 18161D; IL-4, 18191D; IFN-γ, 18181D), 10 mg/ml in PBS. After a washing step with PBS, 2 × 105 lymph nodes cells (LNCs)/well were dispensed, and peptides were added to a final concentration of 10 μg/ml. Control wells contained cells and medium or cells and Con A (2.5 μg/ml). After 18 h at 37°C in 5% CO2, cells were discarded, and biotin-coupled detection Abs (BD PharMingen: IL-2, 18172D; IL-4, 18042D; IFN-γ, 18112D) were added for 3 h, followed by streptavidin-conjugated alkaline phosphatase (ExtrAvidin-PA; Sigma-Aldrich) and an alkaline phosphatase conjugate substrate kit (Bio-Rad, Hercules, CA). Spots with diameters above 30 μm were counted with a computer-assisted device (Zeiss, Oberkochen, Germany).

Mice 12–15 wk old were immunized with peptides as described above. Ten days later, spleen cells and draining LNCs were cultured at 2 × 106/ml in medium with 20 μg/ml peptide. Four days later, blast cells were purified by Ficoll (Amersham Pharmacia Biotech) separation and fused with an hybridoma partner as described (23). After culture in selective medium, hybridomas were tested for reactivity with the immunizing peptide and selected positive hybridomas were cloned by limiting dilution. For testing, hybridoma cells (2 × 104/well) were cultured for 24 h with 5 × 105 irradiated spleen cells and peptide (10 μg/ml) or proinsulin-2 (30 μg/ml), supernatants were harvested, and IL-2 was measured by CTLL-2 proliferation assay. Results are expressed in counts per minute or as the ratio of IL-2 produced with peptide over background IL-2 production by the hybridoma alone. To test the reactivity of hybridomas to islet cells (2 × 104 cells/well), islets were prepared by collagenase digestion as described (26) and dispersed to single cells by treatment with EDTA (10−2 M, 5 min at 37°C) and dispase (0.7 mg/ml; 30 min at 37°C, Boehringer-Mannheim, Mannheim, Germany); islet cell viability was >90% by trypan blue exclusion.

Islets of Langerhans were transplanted under the kidney capsule of proins-2−/− mice as described (26). Six weeks later, the pancreas and graft-bearing kidney were harvested for histology, and spleen cells were tested for their reactivity to proins-2 peptides as described above. In other experiments, proins-2−/− mice were immunized with peptide II46–60 or II56–71 in CFA as described above with three boosts in IFA given every 10 days. Two weeks after the last challenge, the animals were transplanted under the kidney capsule with wt 129 islets and analyzed as described above. Mice 6–10 wk old were used in all transplantation experiments. To quantify islet infiltration in kidney capsule transplants, we designed a scoring system where individual islets scored 0 if not infiltrated, 1 if <50% and 2 if ≥50% of their periphery was surrounded by inflammatory cells. The slides were read by two observers after coding, and a minimum of 15 islets per graft were counted.

Human insulin Abs were measured with a microradioassay adapted from (27), using competition with unlabeled insulin. 125I-insulin (Amersham) was incubated with serum with and without cold insulin for 3 days at 4°C. The samples were then incubated with a 50% protein A, 8% protein G-Sepharose mixture (Pharmacia, Peapack, NJ) in a MultiScreen 96-well filtration plate (Millipore, Bedford, MA). After incubation and washing, radioactivity was measured with a 96-well plate scintillation counter (TopCount; Packard Instrument, Meriden, CT). The result was calculated based on the difference between the well with and the well without cold insulin and was expressed as an index = 100 * [(sample Δcpm − negative control Δcpm)/(positive control Δcpm − negative control Δcpm)]. The positive control sample was obtained from a pool of positive NOD mice, and the upper limit of normal values observed in nonautoimmune mice and human sera was set to 1 as described (27).

To examine whether lack of expression of preproinsulin-2 alters the immune response to preproinsulin and to identify potential epitopes, we immunized individual proins-2−/− mice with peptides spanning the sequence of the deficient preproinsulin isoform. As control, age- and sex-matched wt 129 mice were immunized with the same peptides (Fig. 1). In proins-2−/− mice, 7 of 16 preproinsulin-2 peptides were identified that gave an IL-2 production at least 2-fold over background when lymphocytes were stimulated with the immunizing peptide. Among those peptides, the II56–71 peptide elicited a clear response in proins-2−/− but no response in wt mice. The response to the overlapping II50–65 peptide was barely detectable in proins-2−/− and not in wt mice and the other five peptides gave a similar response in proins-2−/− and wt mice. The difference observed with the II56–71 peptide was confirmed in a separate experiment in which five mice of each type were immunized (Fig. 2): the mean IL-2 concentration after stimulation with II56–71 peptide was 4.1 ± 2.2 IU/ml in proins-2−/− compared with 0.8 ± 0.5 IU/ml in wt mice (p < 0.03, Mann-Whitney test). An ELISPOT assay (Fig. 2,b) confirmed the production of IL-2 by T cells from proins-2−/− but not wt mice immunized with II56–71 in response to the same peptide (p < 0.05, Mann-Whitney test). No response to the homologous I56–71 peptide was observed after immunization with II56–71 peptide (Fig. 2,a) or after direct immunization (data not shown). As expected, wt and proins-2−/− mice responded to the control peptide hβ1AR34–57 (Fig. 1). Thus, expression of preproinsulin-2 in wt mice is associated with a loss of T cell response to II56–71 peptide.

FIGURE 1.

Immunization of wt and proins-2−/− mice with preproinsulin-2 peptides. Spleen cells from adult wt (○) or proins2−/− mice (▪) immunized with each peptide emulsified in CFA were incubated in vitro (5 × 105 cells/well) for 24 h alone or in the presence of the immunizing peptide (20 μg/ml); supernatants were harvested 24 h later, and IL-2 was measured by CTLL-2 proliferation assay (International Units per milliliter); each section represents a different peptide, and each line represents a single mouse. a–h, Peptides eliciting an IL-2 production >2 fold over background; i–l, some peptides eliciting an IL-2 production ≤2-fold over background; IL-2 concentrations with peptides eliciting an IL-2 production ≤2-fold over background are 0.14 ± 0.12 and 0.15 ± 0.09 IU/ml (background, means ± SD), 0.18 ± 0.21 and 0.15 ± 0.08 IU/ml (with peptide) for wt and proins-2−/− mice, respectively; similar results were obtained with LNCs; the results of one of two similar experiments are presented.

FIGURE 1.

Immunization of wt and proins-2−/− mice with preproinsulin-2 peptides. Spleen cells from adult wt (○) or proins2−/− mice (▪) immunized with each peptide emulsified in CFA were incubated in vitro (5 × 105 cells/well) for 24 h alone or in the presence of the immunizing peptide (20 μg/ml); supernatants were harvested 24 h later, and IL-2 was measured by CTLL-2 proliferation assay (International Units per milliliter); each section represents a different peptide, and each line represents a single mouse. a–h, Peptides eliciting an IL-2 production >2 fold over background; i–l, some peptides eliciting an IL-2 production ≤2-fold over background; IL-2 concentrations with peptides eliciting an IL-2 production ≤2-fold over background are 0.14 ± 0.12 and 0.15 ± 0.09 IU/ml (background, means ± SD), 0.18 ± 0.21 and 0.15 ± 0.08 IU/ml (with peptide) for wt and proins-2−/− mice, respectively; similar results were obtained with LNCs; the results of one of two similar experiments are presented.

Close modal
FIGURE 2.

IL-2 response to preproinsulin peptides in mice immunized with II56–71 peptide. a, CTLL-2 assay of IL-2 production by LNCs. LNCs from adult wt (○) or proins-2−/− (▪) mice immunized with II56–71 peptide emulsified in CFA were incubated in vitro (1 × 105 cells/well plus 5 × 105 irradiated spleen cells as APCs) alone or in the presence of the indicated peptide (20 μg/ml); supernatants were harvested 24 h later, and IL-2 was measured in triplicate by CTLL-2 proliferation assay (International Units per milliliter); each line represents a single mouse; p < 0.03 (Mann-Whitney test) for the comparison of IL-2 concentrations with II56–71 peptide between proins-2−/− and wt mice; similar results were obtained with spleen cells. b, ELISPOT assay of IL-2 production by LNCs. LNCs (2 × 105/well) from adult wt (○) or proins-2−/− mice (▪) were dispensed on nitrocellulose plates previously coated with relevant Abs (anti-IL-2, anti-IL-4 or anti-IFN-γ) at 10 mg/ml; after 18 h of incubation, cells were discarded, and biotin-coupled detection Abs (anti-IL-2, anti-IL-4, or anti-IFN-γ) were added, and spots were revealed and counted with a computer-assisted device (Zeiss); the determination was done in triplicate, and each line represents a single mouse; p < 0.05 (Mann-Whitney test) for the comparison of the number of spots with II56–71 peptide between proins-2−/− and wt mice.

FIGURE 2.

IL-2 response to preproinsulin peptides in mice immunized with II56–71 peptide. a, CTLL-2 assay of IL-2 production by LNCs. LNCs from adult wt (○) or proins-2−/− (▪) mice immunized with II56–71 peptide emulsified in CFA were incubated in vitro (1 × 105 cells/well plus 5 × 105 irradiated spleen cells as APCs) alone or in the presence of the indicated peptide (20 μg/ml); supernatants were harvested 24 h later, and IL-2 was measured in triplicate by CTLL-2 proliferation assay (International Units per milliliter); each line represents a single mouse; p < 0.03 (Mann-Whitney test) for the comparison of IL-2 concentrations with II56–71 peptide between proins-2−/− and wt mice; similar results were obtained with spleen cells. b, ELISPOT assay of IL-2 production by LNCs. LNCs (2 × 105/well) from adult wt (○) or proins-2−/− mice (▪) were dispensed on nitrocellulose plates previously coated with relevant Abs (anti-IL-2, anti-IL-4 or anti-IFN-γ) at 10 mg/ml; after 18 h of incubation, cells were discarded, and biotin-coupled detection Abs (anti-IL-2, anti-IL-4, or anti-IFN-γ) were added, and spots were revealed and counted with a computer-assisted device (Zeiss); the determination was done in triplicate, and each line represents a single mouse; p < 0.05 (Mann-Whitney test) for the comparison of the number of spots with II56–71 peptide between proins-2−/− and wt mice.

Close modal

To further study the epitope specificity and Ag presentation of preproinsulin, we generated T cell hybridomas reactive with each of the seven peptides that gave an IL-2 production at least 2-fold over background in proins-2−/− mice (Table II). We used these hybridomas to evaluate whether preproinsulin-2 epitopes identified by peptide immunization were generated in vivo, through processing of endogenous preproinsulin-2. We tested the ability of islet cell preparation from wt (preproinsulin-2 positive) mice to stimulate these hybridomas from proins-2−/− mice (Table II; Fig. 3). None of the hybridomas reactive with II1–15, II14–30, or II26–41 peptides was stimulated by islet cell preparations. In contrast, 10 of 15 II46–60 and 14 of 15 II56–71-reactive hybridomas reacted with wt islet cells, indicating that their respective epitopes could be generated in vivo. There was no relationship between the level of reactivity of hybridomas with peptides and islets. As expected, preproinsulin-2-deficient islet cells from proins-2−/− mice did not stimulate II46–60-reactive and II56–71-reactive hybridomas (data not shown). In additional experiments, selected hybridomas were stimulated with recombinant mouse proinsulin-2 in the presence of irradiated spleen cells. Four hybridomas that reacted with islet cells (respectively, two directed against II46–60 and II56–71 peptides) also reacted with recombinant proinsulin-2. In contrast, two hybridomas against II46–60 peptide, not reacting with islets cells and two hybridomas against II1–15 (outside of proinsulin-2 sequence) were not activated by recombinant proinsulin-2.

Table II.

Generation of preproinsulin-2 peptide-reactive hybridomasa

Mice ImmunizedImmunizing PeptideNo. of Positive HybridomasIL-2 Production Ratio in Response to the Immunizing Peptide (median)Reactivity with Homologous Preproinsulin-1 Peptide (%)No. of Hybridomas Tested with wt Islet CellsNo. of Hybridomas Reactive with wt Islet Cells
Proins-2−/− II1–15 28 32 (10–205)b ND 16 
Proins-2−/− II14–30 55 312 (10–1130) 23 
Proins-2−/− II26–41 200 (31–453) ND 
Proins-2−/− II40–56 13 187 (44–335) 2 (33)c 
Proins-2−/− II46–60 64 63 (12–431) 15 10 (66) 
Proins-2−/− II50–65 35 (12–61) 1d (50) 
Proins-2−/− II56–71 92 229 (11–1458) 15 14 (94) 
       
wt II46–60 14 98 (16–212) 11 
wt II56–71     
Mice ImmunizedImmunizing PeptideNo. of Positive HybridomasIL-2 Production Ratio in Response to the Immunizing Peptide (median)Reactivity with Homologous Preproinsulin-1 Peptide (%)No. of Hybridomas Tested with wt Islet CellsNo. of Hybridomas Reactive with wt Islet Cells
Proins-2−/− II1–15 28 32 (10–205)b ND 16 
Proins-2−/− II14–30 55 312 (10–1130) 23 
Proins-2−/− II26–41 200 (31–453) ND 
Proins-2−/− II40–56 13 187 (44–335) 2 (33)c 
Proins-2−/− II46–60 64 63 (12–431) 15 10 (66) 
Proins-2−/− II50–65 35 (12–61) 1d (50) 
Proins-2−/− II56–71 92 229 (11–1458) 15 14 (94) 
       
wt II46–60 14 98 (16–212) 11 
wt II56–71     
a

Hybridomas were generated from proins-2−/− mice individually immunized with each peptide (one to three mice per peptide); they were tested for IL-2 production in the presence of 129 spleen cells as APCs and the immunizing peptide or proinsulin-2-expressing islet cells (from wt mice); results are expressed as ratios of IL-2 production in the experimental condition over background IL-2 production in the absence of peptide; hybridomas with IL-2 production ratios >10 and peptide-specific IL-2 increments ≥1.5 IU/ml were considered as positive. ND, not determined.

b

Numbers in parentheses, range.

c

Numbers in parentheses, percent.

d

Background 0.1 ± 0.01 IU/ml, with peptide 11.4 ± 0.9 IU/ml, with islets 1.1 ± 0.03 IU/ml.

FIGURE 3.

Reactivity of preproinsulin-2 hybridomas to islet cells. IL-2 production by preproinsulin-2 peptide reactive hybridomas was tested in the presence of APCs and the relevant peptide or APCs and wt islet cells; results are presented for individual hybridomas, in counts per minute of thymidine incorporation by CTLL-2 cells; for each hybridoma, a line connects the two experimental conditions; hybridomas represented by • or by ○ are those reactive and nonreactive with wt islet cells, respectively; spontaneous IL-2 release in the absence of peptide or islet cells was inferior to 3000 cpm in all cases.

FIGURE 3.

Reactivity of preproinsulin-2 hybridomas to islet cells. IL-2 production by preproinsulin-2 peptide reactive hybridomas was tested in the presence of APCs and the relevant peptide or APCs and wt islet cells; results are presented for individual hybridomas, in counts per minute of thymidine incorporation by CTLL-2 cells; for each hybridoma, a line connects the two experimental conditions; hybridomas represented by • or by ○ are those reactive and nonreactive with wt islet cells, respectively; spontaneous IL-2 release in the absence of peptide or islet cells was inferior to 3000 cpm in all cases.

Close modal

Because islets are heterogeneous organs comprising endocrine cells, support cells, and intraislet APCs, we addressed the cell types involved in hybridoma stimulation. We used islets from 129 (H-2b) or BALB/c (H-2d) mice with or without 129 spleen cells to stimulate II56–71-reactive hybridomas (Table III). Wt 129 islet cells could stimulate three of five hybridomas in the absence of spleen cell-derived APCs, although less efficiently. In parallel, spleen cell-derived 129 APCs could efficiently present preproinsulin-2 produced by BALB/c islets to preproinsulin-2-reactive hybridomas, whereas mismatched intraislet APCs from BALB/c mice could not. Therefore, our experiment indicates that both intraislet and spleen cell-derived APCs can efficiently process and present islet cell-derived preproinsulin-2 to T cells.

Table III.

Presentation of islet cell-derived preproinsulin-2 to II56–71 peptide-reactive hybridomas by intra-islet- or spleen cell-derived APCsa

No Islet Cells Background (129 spleen cells)Islet Cells
From wt 129 miceFrom BALB/c mice
129 spleen cellsNo spleen cells129 spleen cellsNo spleen cells
Hybridoma 46.1 0.6 ± 0.2 35 ± 7.5 0.9 ± 0.2 13 ± 8.7 0.5 ± 0.2 
Hybridoma 46.4 0.8 ± 0.2 16 ± 4.5 1.4 ± 0.3 2.8 ± 0.9 0.6 ± 0.01 
Hybridoma 60.6 0.8 ± 0.04 118 ± 39 7.0 ± 4.0 87 ± 19 0.9 ± 0.1 
Hybridoma 63.1 0.9 ± 0.1 67 ± 22 6.4 ± 0.6 25 ± 10 1.0 ± 0.2 
Hybridoma 71b.6 1.2 ± 0.3 187 ± 13 148 ± 9.6 227 ± 8.7 1.1 ± 0.1 
No Islet Cells Background (129 spleen cells)Islet Cells
From wt 129 miceFrom BALB/c mice
129 spleen cellsNo spleen cells129 spleen cellsNo spleen cells
Hybridoma 46.1 0.6 ± 0.2 35 ± 7.5 0.9 ± 0.2 13 ± 8.7 0.5 ± 0.2 
Hybridoma 46.4 0.8 ± 0.2 16 ± 4.5 1.4 ± 0.3 2.8 ± 0.9 0.6 ± 0.01 
Hybridoma 60.6 0.8 ± 0.04 118 ± 39 7.0 ± 4.0 87 ± 19 0.9 ± 0.1 
Hybridoma 63.1 0.9 ± 0.1 67 ± 22 6.4 ± 0.6 25 ± 10 1.0 ± 0.2 
Hybridoma 71b.6 1.2 ± 0.3 187 ± 13 148 ± 9.6 227 ± 8.7 1.1 ± 0.1 
a

Hybridomas were incubated with 129 spleen cells as APCs (background) or stimulated by islet cells from 129 or BALB/c mice, in the presence or absence APCs; IL-2 concentration was measured by CTLL-2 proliferation; means ± SD (10−3 cpm) are shown.

Because hybridomas generated against peptides encompassing the II40–71 region reacted with islet cells, we further studied their reactivity. None of the hybridomas reactive with preproinsulin-2 peptides reacted with the homologous peptide on the preproinsulin-1 isoform. Sequence comparison reveals that II46–60 peptide differs from I46–60 (peptide that is unable to stimulate the hybridomas) only by one single change at position 53, indicating that M53 residue is a critical residue for recognition of II46–60 peptide. Because there are only two amino acid differences in the sequence 56–71 between preproinsulin-1 and −2, we synthesized the two chimeric peptides II56–71–A64E and II56–71–G71S (Table I). II56–71-reactive hybridomas responded to II56–71–G71S similarly to the native peptide, but not to II56–71-A64E. Therefore, A64 is a critical residue within the epitope recognized by II56–71 reactive hybridomas. II46–60 peptide was unable to stimulate hybridomas raised against II56–71 and vice versa. Altogether, these results indicate that at least two epitopes are recognized by hybridomas within the II40–71 region, one spanning residues 46–60, containing the M53 residue, and the other one present within segment 56–71, containing the A64 residue.

We could detect a T cell polyclonal response to II46–60 but not to II56–71 in wt mice. Therefore, we evaluated whether we could generate hybridomas reacting against these two peptides from wt mice (Table II). We were unable to generate hybridomas against II56–71, confirming that immunogenicity of II56–71 peptide is only observed when preproinsulin-2 expression is deficient. In contrast, we successfully generated hybridomas against II46–60 peptide, raising the question of the presence of islet-autoreactive T cells in the peripheral repertoire of wt mice. As shown in Table II and Fig. 3, none of the II46–60 reactive hybridomas from wt mice responded to islet cells, contrasting with the results observed with II46–60-reactive hybridomas generated from proins-2−/− mice. This difference was not explained by different levels of reactivity to II46–60 peptide. This confirms that the II46–60-reactive T cells detected in proins-2−/− mice are a heterogeneous population and indicates that T cells recognizing islet cells are eliminated from the peripheral repertoire of wt mice.

T cells from proins-2−/− mice immunized with recombinant proinsulin-2 secreted IL-2 in response to proinsulin-2 and II46–60 but not II56–71 peptide (Fig. 4). The cytokine response was biased toward IFN-γ and IL-10 but not toward IL-4. In contrast, no response was detected from wt mice.

FIGURE 4.

Immunization of wt and proins-2−/− mice with recombinant proinsulin-2 (PI2). Adult wt mice (○) and proins-2−/− mice (▪) (n = 5 mice per group) were immunized with 100 μg of recombinant proinsulin-2 emulsified in CFA, and spleen cells were stimulated in vitro 10 days later alone or in the presence of the immunizing protein (50 μg/ml) or preproinsulin-2 peptides II46–60, II56–71 or II14–30 (20 μg/ml); IL-2 concentrations were measured in duplicate, 24 h later, by ELISA using recombinant mouse IL-2 as standard; IFN-γ, IL-4 and IL-10 concentrations were measured in duplicate in different sets of culture supernatants, 72 h later, by ELISA using recombinant mouse standards; differences between stimulated and background cytokine concentrations are shown; background cytokine concentrations: IL-2, 6 ± 3 and 12 ± 3 pg/ml; IFN-γ, 8 ± 4 and 24 ± 23 pg/ml; IL-4, 7 ± 3 and 6 ± 3 pg/ml; IL-10, 58 ± 23 and 44 ± 19 pg/ml in wt and proins-2−/− mice, respectively.

FIGURE 4.

Immunization of wt and proins-2−/− mice with recombinant proinsulin-2 (PI2). Adult wt mice (○) and proins-2−/− mice (▪) (n = 5 mice per group) were immunized with 100 μg of recombinant proinsulin-2 emulsified in CFA, and spleen cells were stimulated in vitro 10 days later alone or in the presence of the immunizing protein (50 μg/ml) or preproinsulin-2 peptides II46–60, II56–71 or II14–30 (20 μg/ml); IL-2 concentrations were measured in duplicate, 24 h later, by ELISA using recombinant mouse IL-2 as standard; IFN-γ, IL-4 and IL-10 concentrations were measured in duplicate in different sets of culture supernatants, 72 h later, by ELISA using recombinant mouse standards; differences between stimulated and background cytokine concentrations are shown; background cytokine concentrations: IL-2, 6 ± 3 and 12 ± 3 pg/ml; IFN-γ, 8 ± 4 and 24 ± 23 pg/ml; IL-4, 7 ± 3 and 6 ± 3 pg/ml; IL-10, 58 ± 23 and 44 ± 19 pg/ml in wt and proins-2−/− mice, respectively.

Close modal

Given most experimental models reported thus far, a single Ag expressed by islets is unlikely to elicit an immune response in vivo. We used an islet transplantation model to evaluate whether proins-2−/− mice would mount an immune response against preproinsulin-2-expressing islets (Fig. 5). When wt islets were transplanted in proins-2−/− mice, in four of eight animals a mild and localized inflammatory infiltrate could be detected on serial sections 6 wk after implantation (fIG. 5, b and c). When wt islets were transplanted in proins-2−/− mice who had received multiple challenges with peptide II56–71 or II46–60, peri-insulitis was observed in five of five and three of four mice, respectively (Fig. 5, d–g). In contrast, no cellular infiltrate was detected in two sets of controls, namely proins-2−/− islets grafted in proins-2−/− mice (n = 3) or wt islets grafted in wt mice (n = 5) (Fig. 5, h and i). Because II46–60 peptide is able to induce a response in both types of mice, we also evaluated the effect of immunization with this peptide on the outcome of wt islets grafted in wt mice. No infiltrate and Ab were detected (Fig. 5,a) consistent with absence of response to the mature protein (islets and recombinant proinsulin-2) in T cells from wt mice. Skin grafts from wt to proins-2−/− mice were performed to check their histocompatibility and showed no sign of rejection after 3 mo (n = 3). We could not evaluate in vitro the response to preproinsulin-2 peptides in transplanted mice, due to a nonspecific IL-2 background release. In parallel, we measured the levels of circulating Abs to human insulin in the serum from transplanted mice taken 6 wk after grafting. Significant levels of insulin Abs were detected in 9 of 15 proins-2−/− recipients grafted with wt islets whether they were immunized with II46–60 (4 of 4), II56–71 (2 of 5) or not (3 of 6). Insulin Abs were not detected in control mice (0 of 12; Fig. 4 a). Noticeably, human insulin was used as substrate and corresponds to residues II25–54 (B-chain) and II90–110 (A-chain) of mouse preproinsulin-2 sequence with several substitutions. II56–71 is located in the C peptide and II46–60 is located at the junction between the B-chain and C peptide. Therefore, our method might be insufficient to detect all Ab reactivity to mouse preproinsulin-2. The pancreatic islets of islet transplant recipients were devoid of cellular infiltrate in all cases. Together, these results indicate that grafted preproinsulin-2 expressing islets are able to induce a response in naive proins-2−/− mice. This response is enhanced by immunization against II46–60 and II56–71, suggesting that these peptides are part of the spectrum of response to the native preproinsulin-2.

FIGURE 5.

Transplantation of wt islets in proins-2−/− mice. Islets were transplanted under the kidney capsule of 6 to 10 wk-old mice. a, Insulitis score in individual mice. Islets were scored as noninfiltrated ▩, surrounded by inflammatory cells on <50% (▨) or >50% (◼) of their periphery; insulin autoantibody levels at the time of euthanasia are also shown on the top of the figure and expressed as index with positive results shown in bold (see Materials and Methods for details); b and c, typical histology in proins-2−/− mice transplanted with wt islets; d and e, typical histology in proins-2−/− mice immunized with II56–71 peptide and transplanted with wt islets; f and g, typical histology in proins-2−/− mice immunized with II46–60 peptide and transplanted with wt islets; h and i, typical histology in proins-2−/− mice transplanted with proins-2−/− islets; white arrows, grafted islets; black arrows, inflammatory cells; white arrowheads, kidney cortex; b, d, f, and h, ×10; c, e, g, and i, ×40.

FIGURE 5.

Transplantation of wt islets in proins-2−/− mice. Islets were transplanted under the kidney capsule of 6 to 10 wk-old mice. a, Insulitis score in individual mice. Islets were scored as noninfiltrated ▩, surrounded by inflammatory cells on <50% (▨) or >50% (◼) of their periphery; insulin autoantibody levels at the time of euthanasia are also shown on the top of the figure and expressed as index with positive results shown in bold (see Materials and Methods for details); b and c, typical histology in proins-2−/− mice transplanted with wt islets; d and e, typical histology in proins-2−/− mice immunized with II56–71 peptide and transplanted with wt islets; f and g, typical histology in proins-2−/− mice immunized with II46–60 peptide and transplanted with wt islets; h and i, typical histology in proins-2−/− mice transplanted with proins-2−/− islets; white arrows, grafted islets; black arrows, inflammatory cells; white arrowheads, kidney cortex; b, d, f, and h, ×10; c, e, g, and i, ×40.

Close modal

In these experiments, we have evaluated the immune response to preproinsulin-2 of mice deficient in this isoform. We could identify two epitopes (in II46–60 and II56–71 peptide, respectively) that were generated by APCs from β cell-derived or recombinant mouse proinsulin-2. T cells responding to these epitopes were absent in wt mice expressing endogenous preproinsulin-2. Moreover, preproinsulin-2-expressing islet cells could initiate an immune response or represent the target of a pre-established immune response to preproinsulin-2. In contrast to results obtained in the NOD mouse (23), these experiments allow an appraisal of the mechanisms involved in self tolerance outside the context of spontaneous autoreactivity. Our results have implications for the mechanisms by which self Ags induce immune tolerance and more specifically for the role of insulin in shaping the immune repertoire in type 1 diabetes.

We observed several profiles after preproinsulin-2 peptide immunization. First, with II46–60 peptide, a T cell response could be detected similarly in knockout and wt mice, and an epitope was generated from the mature protein by APCs. This would suggest that autoreactive T cells are present in the peripheral repertoire of wt mice. However, only II46–60 peptide-reactive hybridomas from proins-2−/− mice could react with islet cells. In contrast, all hybridomas from wt mice and some hybridomas from proins-2−/− mice did not react with islet cells. This identifies two populations of II46–60-specific T cells (islet-reactive and non-islet-reactive). Recombinant proinsulin-2 stimulated only islet-reactive II46–60-specific T cells (vs non-islet-reactive) and could prime an immune response in proins-2−/− mice. After immunization with recombinant proinsulin-2, a dominant response to II46–60 peptide was detected only in proins-2−/− mice. The response to II46–60 peptide induced in wt mice did not induce an islet graft infiltrate, in contrast to the findings in proins-2−/− mice. Taken together, these observations indicate that islet-reactive II46–60-specific T cells are functionally eliminated in wt mice. The differences underlying these two populations are currently unknown but might involve the epitope(s) recognized or specificities of TCRs. Second, with II56–71 peptide, the profile was different, because the response was detected in proins-2−/− but not wt mice. The corresponding epitope was generated from the mature protein by APCs. This also suggests that T cells reactive with this epitope are functionally eliminated in wt mice. Last, for peptides encoded in the signal peptide and B chain (II1–15, II14–30, and II26–41), we generated hybridomas but were unable to detect a response with islets. This indicates that these epitopes cannot be generated from the mature protein by APCs and probably have little in vivo relevance. Altogether, proins-2−/− mice allowed us to identify two naturally processed epitopes from mouse preproinsulin-2, and corresponding T cells could not be detected in wt mice.

Similar to in vitro observations, transplanted islets expressing preproinsulin-2 could initiate a B and T cell response, amplified by preimmunization by immunogenic peptides. However, within the 6-wk duration of our experiment, the islets were not rejected. Whether islets differing by a single Ag are recognized or destroyed by the immune system has been evaluated in previous transgenic models. For instance, transplanted islets expressing foreign I-E molecules are not destroyed by the immune system (28), although mouse islets expressing a human complement receptor can be the target of a destructive insulitis (26). Similarly, islets expressing a viral protein can be the target of a destructive insulitis only if the immune system is properly activated by virus infection or in the context of TCR-transgenic mice (10, 29). Altogether these models indicate that a destructive anti-islet immune response can target a single antigenic difference on islets but that additional costimulatory signals are generally needed to initiate such a response (30).

The discussion of the mechanism regulating normal tolerance to preproinsulin is relevant to type 1 diabetes. A first issue is the site of gene expression relevant to the regulation of self tolerance (31). Murine preproinsulin-2 gene is expressed in the thymus, although its cellular localization is still controversial (25, 32, 33, 34). Intrathymic dendritic cells have been proposed (33), but recent data suggest that medullary thymic epithelial cells are the primary thymic cells expressing peripheral Ags, in particular insulin (34, 35). In human fetal thymus, the insulin gene is expressed at the RNA and at the protein level (15, 16, 36). At the cellular level, preproinsulin transcripts and protein have been detected in APCs in the thymus and peripheral lymphoid organs (37). Moreover, variations in the levels of expression of the preproinsulin gene have been associated with susceptibility to type 1 diabetes in mouse and human. NOD mice with invalidated preproinsulin-2 gene develop accelerated diabetes and have increased insulin autoantibodies (38). In the human, diabetes susceptibility alleles of the insulin 5′-variable number of tandem repeats are associated with low thymic gene expression (15, 39). The postulated mechanism is that high levels of thymic expression are associated with deletion of autoreactive precursors and that failure to do so participates in the susceptibility to type 1 diabetes (40). Our experimental setting will allow further dissection of this mechanism, in particular through the establishment of thymus or bone marrow chimeras.

The use of knockout models in which the expression of self Ags is abolished allows the analysis of the mechanisms involved in T cell tolerance. Using a similar approach, Huseby et al. (41) showed that expression of myelin basic protein is associated with tolerance to the Ag through negative selection of Ag-specific thymocytes by bone marrow-derived cells presenting the Ag. In their model, myelin basic protein was not synthesized in APCs but was acquired from other cell types. In another model, B cells specific for desmoglein 3 could efficiently be generated in mice lacking this Ag and induce disease (pemphigus) when transferred to normal mice (42). Similarly, mice deficient in serum amyloid P component are intolerant, and thymic expression of serum amyloid P is sufficient to induce tolerance in bone marrow chimeras (34). These models indicate that different tolerance mechanisms might be involved with various somatically expressed autoantigens. Therefore, our model will allow a precise evaluation of these issues in the case of preproinsulin-2.

In conclusion, proinsulin-2 knockout mice display impaired tolerance to several proinsulin-2 epitopes. Islets expressing proinsulin-2 can prime mice to this Ag and can serve as a target to an established proinsulin-2 response. These findings point out the role of proinsulin-2 gene expression in the regulation of proinsulin-2 autoreactivity. Whether islet-derived proinsulin-2 or ectopically expressed Ag play a role as well as the exact mechanisms involved in these regulatory phenomena remain to be determined.

We thank Agnès Lehuen and Bruno Lucas for helpful comments and suggestions. We acknowledge Karine Vallon-Geoffroy and David Marq for expert animal care and Marie-Noëlle Lotiquet, Sandrine Remeur, and Nadine Giraud for the preparation of histological material.

1

The study was supported by an Institut National de la Santé et de la Recherche Médicale Grant, Juvenile Diabetes Foundation Grant 1-2001-751, Ministère de la Recherche et de la Technologie Action Concertée et Incitative Grant 1A011G and a grant from Association Française des Diabétiques (2002). We thank Association Dia’parole and Comité d’Entreprise des Aéroports de Paris for their generous donations. B.F. was supported by Institut National de la Santé et de la Recherche Médicale and Assistance Publique-Hôpitaux de Paris (poste d’accueil Institut National de la Santé et de la Recherche Médicale 2001).

3

Abbreviations used in this paper: NOD, nonobese diabetic; LNC, lymph node cell; wt, wild type.

1
Nossal, G. J..
1994
. Negative selection of lymphocytes.
Cell
76
:
229
.
2
von Boehmer, H., P. Kisielow.
1990
. Self-nonself discrimination by T cells.
Science
248
:
1369
.
3
Miller, J. F., W. R. Heath.
1993
. Self-ignorance in the peripheral T-cell pool.
Immunol. Rev.
133
:
131
.
4
Delovitch, T. L., B. Singh.
1997
. The nonobese diabetic mouse as a model of autoimmune diabetes: immune dysregulation gets the NOD.
Immunity
7
:
727
.
5
Wong, F. S., C. A. J. Janeway.
1999
. Insulin-dependent diabetes mellitus and its animal models.
Curr. Opin. Immunol.
11
:
643
.
6
Notkins, A. L., A. Lernmark.
2001
. Autoimmune type 1 diabetes: resolved and unresolved issues.
J. Clin. Invest.
108
:
1247
.
7
Rosmalen, J. G., W. van Ewijk, P. J. Leenen.
2002
. T-cell education in autoimmune diabetes: teachers and students.
Trends Immunol.
23
:
40
.
8
Wong, F. S., B. N. Dittel, C. A. J. Janeway.
1999
. Transgenes and knockout mutations in animal models of type 1 diabetes and multiple sclerosis.
Immunol. Rev.
169
:
93
.
9
Vizler, C., N. Bercovici, A. Cornet, C. Cambouris, R. S. Liblau.
1999
. Role of autoreactive CD8+ T cells in organ-specific autoimmune diseases: insight from transgenic mouse models.
Immunol. Rev.
169
:
81
.
10
von Herrath, M. G., C. F. Evans, M. S. Horwitz, M. B. Oldstone.
1996
. Using transgenic mouse models to dissect the pathogenesis of virus-induced autoimmune disorders of the islets of Langerhans and the central nervous system.
Immunol. Rev.
152
:
111
.
11
von Herrath, M. G., S. Guerder, H. Lewicki, R. A. Flavell, M. B. Oldstone.
1995
. Coexpression of B7-1 and viral (“self”) transgenes in pancreatic beta cells can break peripheral ignorance and lead to spontaneous autoimmune diabetes.
Immunity
3
:
727
.
12
Ziegler, A. G., M. Hummel, M. Schenker, E. Bonifacio.
1999
. Autoantibody appearance and risk for development of childhood diabetes in offspring of parents with type 1 diabetes: the 2-year analysis of the German BABYDIAB study.
Diabetes
48
:
460
.
13
Julier, C., R. N. Hyer, J. Davies, F. Merlin, P. Soularue, L. Briant, G. Cathelineau, I. Deschamps, J. I. Rotter, P. Froguel, et al
1991
. Insulin-IGF2 region on chromosome 11p encodes a gene implicated in HLA-DR4-dependent diabetes susceptibility.
Nature
354
:
155
.
14
Kennedy, G. C., M. S. German, W. J. Rutter.
1995
. The minisatellite in the diabetes susceptibility locus IDDM2 regulates insulin transcription.
Nat. Genet.
9
:
293
.
15
Pugliese, A., M. Zeller, A. Fernandez, Jr, L. J. Zalcberg, R. J. Bartlett, C. Ricordi, M. Pietropaolo, G. S. Eisenbarth, S. T. Bennett, D. D. Patel.
1997
. The insulin gene is transcribed in the human thymus and transcription levels correlated with allelic variation at the INS VNTR-IDDM2 susceptibility locus for type 1 diabetes.
Nat. Genet.
15
:
293
.
16
Vafiadis, P., S. T. Bennett, J. A. Todd, J. Nadeau, R. Grabs, C. G. Goodyer, S. Wickramasinghe, E. Colle, C. Polychronakos.
1997
. Insulin expression in human thymus is modulated by INS VNTR alleles at the IDDM2 locus.
Nat. Genet.
15
:
289
.
17
Atkinson, M. A., N. K. Maclaren, R. Luchetta.
1990
. Insulitis and diabetes in NOD mice reduced by prophylactic insulin therapy.
Diabetes
39
:
933
.
18
Muir, A., A. Peck, M. Clare-Salzler, Y. H. Song, J. Cornelius, R. Luchetta, J. P. Krischer, N. K. Maclaren.
1995
. Insulin immunization of nonobese diabetic mice induces a protective insulitis characterized by diminished intraislet interferon-γ transcription.
J. Clin. Invest.
95
:
628
.
19
Carel, J. C., P. F. Bougnères, E. P. S. Group.
1996
. Treatment of prediabetic patients with insulin: experience and future.
Horm. Res.
45S1
:
44
.
20
Wentworth, B. M., I. M. Schaefer, L. Villa-Komaroff, J. M. Chirgwin.
1986
. Characterization of the two nonallelic genes encoding mouse preproinsulin.
J. Mol. Evol.
23
:
305
.
21
Chentoufi, A. A., C. Polychronakos.
2002
. Insulin expression levels in the thymus modulate insulin-specific autoreactive T-cell tolerance: the mechanism by which the IDDM2 locus may predispose to diabetes.
Diabetes
51
:
1383
.
22
Duvillie, B., N. Cordonnier, L. Deltour, F. Dandoy-Dron, J. M. Itier, E. Monthioux, J. Jami, R. L. Joshi, D. Bucchini.
1997
. Phenotypic alterations in insulin-deficient mutant mice.
Proc. Natl. Acad. Sci. USA
94
:
5137
.
23
Halbout, P., J. P. Briand, C. Becourt, S. Muller, C. Boitard.
2002
. T cell response to preproinsulin I and II in the nonobese diabetic mouse.
J. Immunol.
169
:
2436
.
24
Mobini, R., Y. Magnusson, G. Wallukat, M. Viguier, A. Hjalmarson, J. Hoebeke.
1999
. Probing the immunological properties of the extracellular domains of the human β1-adrenoceptor.
J. Autoimmun.
13
:
179
.
25
Chen, W., I. Bergerot, J. F. Elliott, L. C. Harrison, N. Abiru, G. S. Eisenbarth, T. L. Delovitch.
2001
. Evidence that a peptide spanning the B-C junction of proinsulin is an early autoantigen epitope in the pathogenesis of type 1 diabetes.
J. Immunol.
167
:
4926
.
26
Vigneau-Hermellin, M., L. Vitali, I. Tardivel, M. Rabaud, V. M. Holers, J. C. Carel.
1997
. Rejection of islets differing by a single antigen is dependent on donor MHC.
Diabetes
46
:
765
.
27
Yu, L., D. T. Robles, N. Abiru, P. Kaur, M. Rewers, K. Kelemen, G. , S. Eisenbarth.
2000
. Early expression of antiinsulin autoantibodies of humans and the NOD mouse: evidence for early determination of subsequent diabetes.
Proc. Natl. Acad. Sci. USA
97
:
1701
.
28
Markmann, J. F., D. Lo, A. Naji, R. D. Palmiter, R. L. Brinster, E. Heber-Katz.
1988
. Antigen presenting function of class II MHC expressing pancreatic β-cells.
Nature
336
:
476
.
29
Ohashi, P. S., S. Oehen, K. Buerki, H. Pircher, C. T. Ohashi, B. Odermatt, B. Malissen, R. M. Zinkernagel, H. Hengartner.
1991
. Ablation of “tolerance” and induction of diabetes by virus infection in viral antigen transgenic mice.
Cell
65
:
305
.
30
Guerder, S., D. E. Picarella, P. S. Linsley, R. A. Flavell.
1994
. Costimulator B7-1 confers antigen-presenting-cell function to parenchymal tissue and in conjunction with tumor necrosis factor α leads to autoimmunity in transgenic mice.
Proc. Natl. Acad. Sci. USA
91
:
5138
.
31
Hanahan, D..
1998
. Peripheral-antigen-expressing cells in thymic medulla: factors in self-tolerance and autoimmunity.
Curr. Opin. Immunol.
10
:
656
.
32
Heath, V. L., N. C. Moore, S. M. Parnell, D. W. Mason.
1998
. Intrathymic expression of genes involved in organ specific autoimmune disease.
J. Autoimmun.
11
:
309
.
33
Throsby, M., F. Homo-Delarche, D. Chevenne, R. Goya, M. Dardenne, J. M. Pleau.
1998
. Pancreatic hormone expression in the murine thymus: localization in dendritic cells and macrophages.
Endocrinology
139
:
2399
.
34
Derbinski, J., A. Schulte, B. Kyewski, L. Klein.
2001
. Promiscuous gene expression in medullary thymic epithelial cells mirrors the peripheral self.
Nat. Immunol.
2
:
1032
.
35
Kyewski, B., J. Derbinski, J. Gotter, L. Klein.
2002
. Promiscuous gene expression and central T-cell tolerance: more than meets the eye.
Trends Immunol.
23
:
364
.
36
Sospedra, M., X. Ferrer-Francesch, O. Dominguez, M. Juan, M. Foz-Sala, R. Pujol-Borrell.
1998
. Transcription of a broad range of self-antigens in human thymus suggests a role for central mechanisms in tolerance toward peripheral antigens.
J. Immunol.
161
:
5918
.
37
Pugliese, A., D. Brown, D. Garza, D. Murchison, M. Zeller, M. Redondo, J. Diez, G. S. Eisenbarth, D. D. Patel, C. Ricordi.
2001
. Self-antigen-presenting cells expressing diabetes-associated autoantigens exist in both thymus and peripheral lymphoid organs.
J. Clin. Invest.
107
:
555
.
38
Thebault-Baumont, K., D. Dubois-Laforgue, P. Krief, J. P. Briand, P. Halbout, K. Vallon-Geoffroy, J. Morin, V. Laloux, A. Lehuen, J. C. Carel, et al
2003
. Acceleration of type 1 diabetes mellitus in proinsulin 2-deficient NOD mice.
J. Clin. Invest.
111
:
851
.
39
Vafiadis, P., H. Ounissi-Benkalha, M. Palumbo, R. Grabs, M. Rousseau, C. G. Goodyer, C. Polychronakos.
2001
. Class III alleles of the variable number of tandem repeat insulin polymorphism associated with silencing of thymic insulin predispose to type 1 diabetes.
J. Clin. Endocrinol. Metab.
86
:
3705
.
40
Geenen, V., P. J. Lefebvre.
1998
. The intrathymic expression of insulin-related genes: implications for pathophysiology and prevention of type 1 diabetes.
Diabetes Metab. Rev.
14
:
95
.
41
Huseby, E. S., B. Sather, P. G. Huseby, J. Goverman.
2001
. Age-dependent T cell tolerance and autoimmunity to myelin basic protein.
Immunity
14
:
471
.
42
Amagai, M., K. Tsunoda, H. Suzuki, K. Nishifuji, S. Koyasu, T. Nishikawa.
2000
. Use of autoantigen-knockout mice in developing an active autoimmune disease model for pemphigus.
J. Clin. Invest.
105
:
625
.