The JI PBL Intereron Source
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chen, W.
Right arrow Articles by Delovitch, T. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chen, W.
Right arrow Articles by Delovitch, T. L.
The Journal of Immunology, 2001, 167: 4926-4935.
Copyright © 2001 by The American Association of Immunologists

Evidence That a Peptide Spanning the B-C Junction of Proinsulin Is an Early Autoantigen Epitope in the Pathogenesis of Type 1 Diabetes1

Wei Chen2,*, Isabelle Bergerot2,*, John F. Elliott{ddagger}, Leonard C. Harrison§, Norio Abiru, George S. Eisenbarth and Terry L. Delovitch3,*,{dagger}

* Autoimmunity/Diabetes Group, The John P. Robarts Research Institute and {dagger} Departments of Microbiology and Immunology, and Medicine, University of Western Ontario, London, Ontario, Canada; {ddagger} Department of Medical Microbiology and Immunology, University of Alberta, Edmonton, Canada; § Autoimmunity and Transplantation Division, The Walter and Eliza Hall Institute of Medical Research, Royal Melbourne Hospital PO, Parkville, Victoria, Australia; and Barbara Davis Center for Childhood Diabetes, University of Colorado Health Sciences Center, Denver, CO 80262


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
The expression of pro(insulin) in the thymus may lead to the negative selection of pro(insulin) autoreactive T cells and peripheral tolerance to this autoantigen in type 1 diabetes (T1D). We investigated whether proinsulin is expressed in the thymus of young nonobese diabetic (NOD) mice, whether T cells from naive NOD female mice at weaning are reactive to mouse proinsulin, and the role of proinsulin as a pathogenic autoantigen in T1D. Proinsulin II mRNA transcripts were detected in the thymus of 2-wk-old NOD mice at similar levels to other control strains. Despite this expression, proinsulin autoreactive T cells were detected in the periphery of 2- to 3-wk-old naive NOD mice. Peripheral T cells reactive to the insulin, glutamic acid decarboxylase 65 (GAD65), GAD67, and islet cell Ag p69 autoantigens were also detected in these mice, indicating that NOD mice are not tolerant to any of these islet autoantigens at this young age. T cell reactivities to proinsulin and islet cell Ag p69 exceeded those to GAD67, and T cell reactivity to proinsulin in the spleen and pancreatic lymph nodes was directed mainly against a p24–33 epitope that spans the B chain/C peptide junction. Intraperitoneal immunization with proinsulin perinatally beginning at 18 days of age delayed the onset and reduced the incidence of T1D. However, s.c. immunization with proinsulin initiated at 5 wk of age accelerated diabetes in female NOD mice. Our findings support the notion that proinsulin p24–33 may be a primary autoantigen epitope in the pathogenesis of T1D in NOD mice.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Autoimmunity to insulin and/or isoforms of glutamic acid decarboxylase (GAD65 and GAD67)4 is detectable in humans and nonobese diabetic (NOD) mice with type 1 diabetes (T1D) (1). Insulin-specific T cells are found preferentially in islet infiltrates in NOD mice between 4 and 12 wk of age, and insulin B chain (B9–23)-specific T cell clones both accelerate diabetes in young NOD mice and adoptively transfer T1D to NOD.scid mice (2, 3). GAD65-reactive splenic T cells are detectable in NOD mice as early as 3 wk, and tolerization of NOD mice with GAD65 delays/prevents insulitis and T1D (4, 5). However, NOD T cell clones exist that can induce insulitis and T1D and yet do not react with GAD65 or insulin. Thus, it remains unclear whether insulin, GAD65, or another protein is an early autoantigen that elicits T cell-mediated islet {beta} cell destruction.

Accumulating evidence suggests that proinsulin may be an early autoantigen in the pathogenesis of T1D. Proinsulin gene transcripts have been found in the neonatal mouse and human thymus (6, 7). Furthermore, proinsulin is the only islet {beta} cell-specific autoantigen, and aberrant transcription or processing of a cell-specific product would be a logical explanation for targeted autoimmunity. Indeed, we have shown that altered processing of human insulin occurs in B cell APC from a patient with T1D (8). This notion is further supported by evidence that T cell epitopes of insulin defined in HLA-DR4 transgenic NOD mice are actually derived from preproinsulin and proinsulin (9). A 13-aa sequence homology between human proinsulin residues 24–36 and human GAD65 (hGAD65) residues 506–518 was identified (10). This 13-mer sequence of proinsulin contains a binding motif for the I-Ag7 MHC class II molecule of NOD mice (11). T cell reactivity to this proinsulin peptide may therefore represent an early autoimmune event in T1D, and as a result of molecular mimicry and cross-reactivity may give rise to T cell reactivity to the similar GAD peptide (5). Interestingly, rat CD4+ T cell lines specific for proinsulin peptides (located between B chain and C peptide of proinsulin), but not the similar GAD65 peptide, adoptively transfer insulitis to syngeneic naive rats (12). Moreover, GAD65 and insulin B chain peptide (9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23) are not primary autoantigens for the development of T1D in Bio-Breeding rats (13). T cell reactivity to proinsulin occurs in individuals at risk for T1D (10, 14), and transgenic expression of mouse proinsulin by MHC class II-bearing cells prevents T1D in NOD mice (15). These findings support the possibility that proinsulin may be a pathogenic autoantigen early in the development of T1D.

A disproportionately elevated release of proinsulin, but not insulin, is found in human islets in response to cytokines (16). It may be clinically relevant that circulating proinsulin levels can be more than 2-fold higher in recently diagnosed diabetic patients than in normal healthy control individuals (17, 18). As proinsulin is expressed in the thymus and in highest concentration in islet {beta} cells, proinsulin rather than its processed products (insulin and C peptide) might be a better candidate for an early autoantigen in T1D. To test this possibility, we analyzed whether peripheral T cells in perinatal vs adult female NOD mice respond to proinsulin and its immunodominant peptides, and whether perinatal immunization with proinsulin accelerates or delays progression to the onset of T1D.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Mice

NOD and NOD.scid mice were bred under specific pathogen-free conditions in our animal facility. In our colony of NOD mice, the onset of insulitis occurs at 4–5 wk of age, and the incidence of diabetes is 70–90% in females and 20–40% in males at 30 wk of age. NOD.scid mice were generously provided by L. Shultz (The Jackson Laboratory, Bar Harbor, ME) and were bred in our animal facility. The blood glucose level (BGL) was monitored weekly with a Glucometer Encore (Miles/Bayer, Toronto, Ontario, Canada), and mice with a BGL >11.1 mmol/L (200 mg/dl) for 2 consecutive days were considered diabetic. BALB/cJ and C57BL/6J (B6) mice were purchased from The Jackson Laboratory.

Production and purification of recombinant mouse proinsulin II

Mouse proinsulin II was expressed in Escherichia coli using the vector pT7jlh (19). A cDNA fragment encoding proinsulin II was obtained from a mouse islet cDNA library by PCR using Pfu DNA polymerase (Stratagene, La Jolla, CA). Primers were designed so as to place a XbaI site immediately 5' of the first codon of the mature protein and a NotI site immediately 3' of the termination codon. This fragment was cloned into pT7jlh using the XbaI and NotI sites, so as to generate a fusion with the (His) 6 coding segment immediately upstream of the XbaI site (N terminus of the recombinant protein is MRGSHHHHHHSRFVKQ; proinsulin sequence is underlined). The final MuProinsulin2/pT7jlh construct was transformed into E. coli MC1061/pT7POL23 (20), and 10 random colonies that grew at 28°C on ampicillin plus kanamycin plates were transferred to 5-ml cultures and grown with shaking at 28°C in 2x YT media containing ampicillin and kanamycin (21). After an OD600 of 0.40 was reached, 1 ml of each culture was added to 250 µl 50% glycerol to generate glycerol stocks (stored at -70°C). The remainder of each culture was split equally, with one-half continuing to be shaken at 28°C (noninduced) and the other half being placed at 42°C with shaking to induce expression of recombinant protein. After 3 h of culture, noninduced and induced bacterial pellets were analyzed by 15% SDS-PAGE. For large-scale protein expression, a sample from the glycerol stock corresponding to the isolate that yielded the highest levels of proinsulin expression was used to generate a 1-L culture (grown at 28°C in 2x YT media containing ampicillin and kanamycin). This culture was then used to inoculate a 50-L fermentor containing the same medium. Once the OD600 of the fermentor culture reached 0.40, the temperature was rapidly shifted from 28°C to 42°C by heating the jacket with steam, and the culture continued at 42°C for 3 h. The biomass was harvested by centrifugation; bacterial cells opened using a French press (twice at 16,000 psi) in the presence of DNase (4 µg/ml), RNase (2 µg/ml), and PMSF (1 mM); and inclusion bodies were harvested by centrifugation (5000 x g for 30 min). The inclusion bodies were dissolved in PBS/20 mM Tris-HCl, pH 8/6 M guanidium hydrochloride.

The recombinant protein was purified by affinity chromatography on a nickel-chelating Ni-NTA column, and eluted using a stepwise pH gradient (pH 8, 6.3, 5.8, 5.3, 4.5, and 3.5) in 8 M urea, as described (19). After washing the column several times with 1% SDS to remove contaminants, the protein that eluted at pH 5.8 and 5.3 was dialyzed against Laemmli running buffer, against the same buffer with one-tenth the standard amount of SDS, and finally against 4 mM HEPES, pH 7.4. The dialyzed recombinant protein was tested for the presence of inhibitory concentrations of SDS by analyzing the proliferative capacity of human PBMC in PHA, and was found to have no effect on cell proliferation at concentrations up to 100 µg/ml (data not shown). The purified protein was concentrated to ~1 mg/ml by lyophilization and stored at -70°C.

SDS-PAGE (4–20% gradient gel; Invitrogen Canada, Buglington, Canada) analysis of the purified recombinant mouse proinsulin II under nonreducing conditions demonstrated that it consists mainly of monomers, a lesser amount of dimers, and a rather small amount of multimers (includes hexamers) (Fig. 1GoA). In contrast, human proinsulin (Sigma, St. Louis, MO; expressed in E. coli and purified by HPLC) consists mainly of hexamers. The mouse proinsulin II preparation did not contain another detectable recombinant protein(s). Considering that a low percentage of proinsulin multimers may alter the immunogenicity of the protein preparation, mouse proinsulin II was further purified by electroelution from a SDS-PAGE gel. The electroeluted protein was dialyzed against >200 vol of 20 mM sodium carbonate, 4 mM DTT (pH 10.6) at 23°C (two steps of 4 h each) and then at 4°C (two steps of 4 h each). Purified proinsulin II was present in monomeric form (Fig. 1GoB) under reducing conditions (1 mM DTT).



View larger version (15K):
[in this window]
[in a new window]
  		FIGURE 1. A, Coomassie blue-stained SDS-PAGE (4–20% gradient gel) of purified recombinant mouse proinsuiln and recombinant human proinsulin under nonreducing conditions. Lane 1, Molecular mass markers. Lanes 2 and 3, Mouse and human proinsulin, respectively. B, Coomassie blue-stained SDS-PAGE (15% gel) of electrophoretically purified recombinant mouse proinsulin II. Lane 1, Molecular mass markers. Lane 2, Recombinant mouse proinsulin II electrophoresed under reducing conditions (20 mM Na2CO3 + 1 mM DTT).

 
Analysis of thymic proinsulin gene expression

Thymii and pancreatic islets were collected from 2-wk-old female NOD, BALB/cJ, and B6 mice. Total RNA was purified from fresh tissues by the guanidinium isothiocyanate/silica gel-based membrane RNeasy method (Qiagen, Valencia, CA), according to the manufacturer’s recommended procedure, with the exception that 0.14 M 2-ME was used during the isolation of total thymic or pancreatic RNA. Total RNA was treated with RNase-free DNase (Life Technologies, Gaithersburg, MD) to remove contaminating genomic DNA, phenol/chloroform extracted, and ethanol precipitated. Determination and quantification of proinsulin II mRNA levels were performed by RT-PCR. Briefly, reverse transcription of RNA (1 µg) was performed using Superscript II (Life Technologies) and oligo(dT) primers. Standardization of cDNA was determined by PCR amplification of GAPDH in each sample. The primers used for GAPDH were previously described (22). Equivalent amounts of cDNA (according to GAPDH standardization) were added for determination of proinsulin mRNA, and PCR amplification of cDNA was performed using Taq polymerase (Life Technologies). For proinsulin gene expression, a single pair of oligonucleotide primers (23) was used because they were homologous to both the proinsulin I and proinsulin II genes. After 30 cycles of PCR, each cycle consisting of 1 min at 94°C, 1.5 min at 59°C, and 1 min at 72°C, the products were separated on an agarose gel. Ethidium bromide-stained product band intensities of the cDNA targets were quantified by Gel Doc 1000 video gel documentation (Bio-Rad, Hercules, CA). The relative amount of proinsulin mRNA was calibrated to the amount of GAPDH mRNA, which was set equal to 1. Controls verified a linear relationship between the quantity of mRNA analyzed and the signal intensity determined by densitometric analysis (Gel Doc 1000) of cDNA.

Immunization with Ag

Purified mouse proinsulin II or porcine insulin (Sigma; 100 µg) in PBS was emulsified in IFA (Sigma) (1:1 v/v) and then injected (100 µl) i.p. or s.c. into NOD female mice of various ages. Control mice received either OVA or PBS + IFA.

T cell preparation

Suspensions of lymphocytes from spleens and pancreatic lymph nodes (PLN) of mice were prepared according to standard procedures. T cell populations were obtained by filtration through a murine T cell enrichment column (R&D Systems, Minneapolis, MN), which yielded cell populations consisting of >=90% CD3+ T cells determined by flow cytometry. For CD4+ T cell enrichment, lymphocyte suspensions were incubated (15 min, 23°C) with an anti-CD8 mAb and were then filtered through a T cell subset column (R&D Systems). The resulting T cell subpopulation was demonstrated by flow cytometry using a FITC-labeled anti-mouse CD4 (clone H129.19; BD PharMingen, San Diego, CA) or FITC-labeled anti-mouse CD8{alpha} (clone 53-6.7; BD PharMingen) mAb to consist of >=90% CD4+ and <=0.3% CD8+ T cells.

T cell proliferative response to Ags

Spleen and PLN T cell cultures were established in complete RPMI 1640 (RPMI 1640 containing 200 U/ml penicillin, 200 µg/ml streptomycin, 10 mM HEPES, 0.06 µg/ml L-glutamine, 10-5 mM 2-ME, and 10% FCS) in a final volume of 200 µl in triplicate wells of round-bottom 96-well microtiter plates. Spleen or PLN T cells (5 x 105) were cocultured with irradiated (3000 R) splenocytes (5 x 105) as APC for 5 days at 37°C in the presence or absence of 20 µg/ml of either OVA, porcine insulin, bovine insulin A chain or B chain (Sigma), mouse proinsulin, mouse GAD67 (mGAD67), hGAD65, or hGAD65 peptide (509–528) (kindly provided by B. Singh, University of Western Ontario, London, Ontario, Canada). T cell responses to mouse insulin B chain peptide 9–23 (kindly provided by G. S. Eisenbarth, Barbara Davis Center for Childhood Diabetes, Denver, CO), proinsulin II peptides, and human proinsulin peptide B24-C33 were also determined. The latter peptides were synthesized by F-moc chemistry on solid phase (Mimotopes, Melbourne, Australia), purified by HPLC, analyzed by mass spectrometry, and used at designated concentrations. E. coli bacterial LPS purchased from Sigma was used as a positive control Ag in some experiments. During the final 18 h of culture, 1 µCi [3H]thymidine was added in each well, after which the cells were harvested using a 96-well cell harvester (Tomtec, Orange, CT). The amount of [3H]thymidine incorporation was quantified using a 1450 Microbeta counter (Wallac, Turku, Finland). Data are expressed either as the mean ± SD or as a stimulation index (SI) = mean cpm (stimulated response)/mean cpm (unstimulated response) for better comparison and readability.

Cytokine assays

Splenic T cells (5 x 105) were cocultured with irradiated (3000 rad) splenocytes (5 x 105) as APC in round-bottom 96-well plates in the presence of either mouse proinsulin, mouse proinsulin peptides, or control Ags in complete RPMI 1640 medium. After 48 h, culture supernatants were assayed for IL-2, IL-4, IL-10, and IFN-{gamma} by ELISA, as previously described (24). Briefly, 96-well Nunc-Immuno plates (Nunc, Roskilde, Denmark) were coated (16 h, 4°C) with an anti-cytokine mAb (1 µg/ml in 0.1 mM NaHCO3, 50 µl/well), and nonspecific binding sites were blocked (2 h, 23°C) with 3% BSA in PBS. Standards (25–2000 pg/ml, 100 µl/ml) or samples were added for 16 h at 4°C, and after washing biotinylated anti-cytokine mAb (1 µg/ml, 50 µl/ml) was added for 1 h at 23°C. A streptavidin-peroxidase conjugate (0.2 U/ml; Boehringer Mannheim) and chromogen substrate (p-nitrophenyl phosphate, 1 mg/ml; Sigma) in diethanolamine buffer were added to develop the reaction, and the absorbance at 405 nm was determined using an automated microplate reader (Bio-Rad). The rat anti-mouse cytokine mAb pairs (BD PharMingen) used and their clone designations were as follows: JES6-1A12 and biotinylated JES6-5H4 for IL-2; 11B11 and biotinylated BVD6-24G2 for IL-4; JES5-2A5 and biotinylated SXC-1 for IL-10; and R6-6A2 and biotinylated XMG1.2 for IFN-{gamma}.

Insulin autoantibody assay

Levels of serum insulin autoantibody (IAA) expression were evaluated prospectively in immunized mice. Serum samples were collected at different time points after immunization with either mouse proinsulin II or control Ags. IAA expression was measured using a 96-well filtration-plate micro IAA assay, as described (25). Levels of IAA were expressed as a relative index as follows: IAA index = (sample {Delta}cpm - negative control {Delta}cpm)/(positive control {Delta}cpm - negative control {Delta}cpm). An IAA index value = 0.01 was chosen as background level of the assay, as described (25).

Endotoxin assay

The concentrations of E. coli-derived bacterial endotoxin in the various Ag preparations as well as the proinsulin peptides analyzed were determined using a Limulus Amebocyte Lysate Assay (BioWhittaker, Walkersville, MD), according to the manufacturer’s instructions. A maximum concentration of 130 pg/ml bacterial endotoxin was detected in the peptide samples (1 mg/ml) used. To remove endotoxin from the Ag preparations, the Ags were passed through a column of Affi-Prep Polymyxin B Sulfate (PmB) Matrix (Bio-Rad).

Statistical analysis

Statistical analyses were performed using either the Student’s t test, Fisher’s exact test, or log-rank tests. Differences were considered statistically significant with p < 0.05.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Proinsulin mRNA transcripts in the thymus of NOD mice

To determine whether proinsulin genes are expressed in the thymus of NOD mice, total thymic RNA from 2-wk-old female NOD mice and age- and gender-matched BALB/cJ and B6 mice were analyzed by RT-PCR. Proinsulin mRNA transcripts were detectable in the thymus of NOD mice at similar levels to other mouse strains tested (Fig. 2Go). Note that both proinsulin I and proinsulin II transcripts are expressed in the thymus, as assayed by sensitivity of their RT-PCR cDNA products to digestion by MspI. This enzyme cuts at different sites in each of the two genes, generating subfragments of 34, 71, and 77 bp for proinsulin I and 76, and 112 bp for proinsulin II (23) (data not shown). Although proinsulin genes are expressed in the thymus of NOD mice, the level of proinsulin mRNA observed in the thymus was ~40-fold less than that found in the islets. To our knowledge, this is the first demonstration that proinsulin genes are expressed in the thymus of NOD mice.



View larger version (29K):
[in this window]
[in a new window]
  		FIGURE 2. Proinsulin gene expression in the thymus of NOD mice. A, Thymic proinsulin mRNA in NOD, BALB/cJ, and B6 mice. The data represent mean intensities of RT-PCR signals for proinsulin mRNA, as determined by densitometry of cDNA, followed by normalization to the signals of GAPDH of individual samples of three experiments ± SD. B, Results shown are the PCR products of proinsulin and GAPDH mRNA expression separated in an ethidium bromide-stained agarose gel.

 
T cell reactivity to mouse proinsulin and other islet {beta} cell autoantigens in perinatal NOD mice

To investigate whether NOD female mice are tolerant to proinsulin at a perinatal age, we compared the peripheral T cell response to this autoantigen with that to other control Ags in young NOD mice. Note that preparations of mouse proinsulin II containing only monomers or both monomers and multimers were found to stimulate T cell proliferative responses of very similar magnitude (Fig. 3GoA). This demonstrates that the immunogenicity of these two proinsulin preparations does not differ. With the exception of islet cell Ag p69 (ICA69), the response of naive splenic T cells from 3- to 4-wk-old NOD female mice to mouse proinsulin (20 µg/ml) in the presence of splenic APC was ~3- to 6-fold greater than that to the control Ags, including mGAD67 (Fig. 3GoB). Thus, peripheral T cell responses to proinsulin and ICA69 both emerge early in perinatal female NOD mice, indicating that these mice are not tolerant to these two autoantigens. Since each of these Ags was derived from the same E. coli expression system, the T cell proliferation observed may have resulted from stimulation by a contaminating bacterial endotoxin. To rule out this possibility, we analyzed the levels of endotoxin contamination in these Ag preparations. After passage through a PmB column, the endotoxin content was very low (100~440 pg/ml) in each Ag preparation used.



View larger version (28K):
[in this window]
[in a new window]
  		FIGURE 3. Mouse proinsulin stimulates T cell proliferation in NOD mice. A and B, Splenic T cells from 4- to 5-wk-old NOD female mice were stimulated in the presence of 20 µg/ml of either mouse proinsulin (contains only monomers or monomers plus multimers), mouse ICA69, OVA, malaria protein, mGAD67, hGAD65, or hGAD65 peptide (509–528), respectively. Results are expressed as mean cpm ± SD of two independent experiments, each performed in triplicate. Significant increases of proliferation relative to the PBS control are indicated; *, p < 0.001, or **, p < 0.0005, respectively. C, T cell responsiveness to mouse proinsulin is not inhibited by PmB. Splenic T cells from 4- to 5-wk-old NOD female mice were stimulated in vitro with 20 µg/ml of either mouse proinsulin or other control Ags. LPS was used at a concentration of 20 ng/ml. All cultures were maintained in the absence ({square}) or presence ({blacksquare}) of 1 µg/ml PmB. T cell responses to Ags are expressed as in A. Data are expressed as mean cpm ± SD of two independent experiments, each performed in triplicate. Significantly reduced (p < 0.005) proliferation to Ags in the presence of PmB is indicated with an asterisk.

 
To further ensure that any residual endotoxin was removed from the Ag preparations, these preparations were treated with PmB (1 µg/ml) 2 h before addition to cell cultures. Treatment with PmB suppressed the T cell responses not only to bacterial LPS, but also to OVA + LPS (Fig. 3GoC). The proliferative response to OVA + LPS in the presence of PmB was equivalent to the response to OVA obtained in the absence of PmB, demonstrating that the OVA preparation used was relatively free of endotoxin contamination. In contrast, PmB did not suppress the T cell responses induced by either proinsulin or malaria protein. These observations indicate that our recombinant Ag preparations contain relatively little, if any, contaminating bacterial endotoxin.

Characterization of T cell response to proinsulin in NOD mice

We determined the influence of age, tissue localization, and T cell subset on the peripheral T cell response to mouse proinsulin in perinatal NOD mice. Between 3 wk of age and the onset of diabetes (20–30 wk of age), spleen T cell responses to proinsulin were elevated considerably (SI {approx} 7–43) in female NOD mice relative to control BALB/c mice (SI <= 3) (Fig. 4GoA). Note that at 3–4 wk of age, NOD T cell responses to proinsulin (SI {approx} 12) were 2-fold greater than the responses to either GAD65 (SI {approx} 6, p < 0.01) (Fig. 4GoB) or insulin (SI {approx} 6–7, p < 0.01) (Fig. 4GoC). At 6–8 wk of age, NOD T cell responses to hGAD65 persisted and were even increased (SI {approx} 36–45) (Fig. 4GoB). In contrast, T cell reactivity to insulin (SI {approx} 3–6) (Fig. 4GoC), insulin A chain (SI < 3) (Fig. 4GoF), insulin B chain (SI < 5) (Fig. 4GoD), and OVA (SI < 3) (Fig. 4GoE) remained relatively weak. Three important findings emerge from these studies. First, spleen T cell responses to proinsulin and GAD65 are present in female NOD mice as early as 3 wk of age, but the response to proinsulin exceeds that of GAD65 at this time. Second, spleen T cell responses to proinsulin and GAD65 in 6- to 8-wk-old female NOD mice exceed the responses to the other Ags tested. Third, NOD spleen T cells recognize a proinsulin epitope(s) that is absent from insulin.



View larger version (23K):
[in this window]
[in a new window]
  		FIGURE 4. Age dependency of NOD splenic T cell proliferative response to proinsulin and other Ags. Diabetes-susceptible NOD female mice were tested for their T cell reactivity to the proinsulin, GAD65, insulin, insulin A chain, insulin B chain, and OVA Ags. Mice were sacrificed at either 3, 4, 6, and 8 wk of age or at 2 days after diabetes onset (diabetic). Splenic T cells were cultured with irradiated syngeneic splenocytes for 5 days in the presence of Ag (20 µg/ml). BALB/cJ female mice (6–8 wk old) were used as a control strain; no differences were noted when T cells of BALB/cJ female mice at different ages were challenged with proinsulin. T cell responses to Ags are expressed on a logarithmic scale as a stimulation index (SI ± SD). The arrow on this scale indicates the threshold of significance (SI = 3). The range of background values of [3H]thymidine incorporation observed in medium-only controls for individual mice tested in each group was as follows: 3 wk old, 904-1658 cpm; 4 wk old, 1268–2329 cpm; 6 wk old, 1896–3284 cpm; 8 wk old, 1773–2895 cpm; diabetic, 1487–2367 cpm; and 6 wk old, BALB/cJ, 1782–2967 cpm. Results of three combined independent experiments are shown, with a total of >=12 mice/group used at each time point of analysis. Significant differences relative to the BALB/cJ control values in each group from three independent experiments are indicated as *, p < 0.01, or **, p < 0.001.

 
PLN T cells from 3- to 8-wk-old and diabetic NOD mice responded to proinsulin (SI {approx} 6–19) (Fig. 5GoA) and GAD65 (SI {approx} 7–21) (Fig. 5GoB). Interestingly, these PLN T cells also responded to insulin (SI {approx} 5–11) (Fig. 5GoC), but not to insulin A chain (SI < 3) (Fig. 5GoF), insulin B chain (SI < 3) (Fig. 5GoD), or OVA (SI < 3) (Fig. 5GoE). Purified NOD splenic CD4+ T cells also displayed a strong T cell proliferative response to proinsulin (SI {approx} 6–17) (Fig. 6GoA), and this response was completely inhibited in the presence of a blocking anti-CD4 mAb (Fig. 6GoB). Thus, CD4+ T cells mediate this cellular immune response to proinsulin.



View larger version (26K):
[in this window]
[in a new window]
  		FIGURE 5. Age dependency of NOD PLN T cell proliferative response to proinsulin and other Ags. The in vitro proliferative responses of NOD PLN T cells to various Ags were determined as described in Fig. 4Go. The background values of [3H]thymidine incorporation observed in medium-only controls for individual mice tested in each group were as follows: 3 wk old, 873-2145 cpm; 4 wk old, 2389–3103 cpm; 6 wk old, 1359–3781 cpm; 8 wk old, 1943–3210 cpm; diabetic, 1502–3290 cpm; and BALB/cJ mice, 1257–3478 cpm. The combined results of three independent experiments are shown, with a total of >12 mice per group used at each time point of analysis. No significant differences were noted between NOD PLN T cells and BALB/cJ PLN T cells in response to proinsulin, insulin, and GAD65 at the different time points. Significant differences relative to the medium control values in each group from three independent experiments are indicated as *, p < 0.01, or **, p < 0.001.

 


View larger version (14K):
[in this window]
[in a new window]
  		FIGURE 6. NOD splenic T cell reactivity to proinsulin is CD4+ T cell dependent. A, NOD female mice were sacrificed at 3, 4, 6, and 8 wk of age or at 2 days after diabetes onset (diabetic). Splenic CD4+ T cells from these mice were cultured (5 x 105/well) in the presence of proinsulin (20 µg/ml) and irradiated splenic APC (5 x 105/well). T cell responses to Ag are expressed as in Fig. 4Go. The background values of [3H]thymidine incorporation observed in medium-only controls for individual mice tested in each group were as follows: 3 wk old, 1045–2038 cpm; 4 wk old, 1290–3048 cpm; 6 wk old, 1304–2903 cpm; 8 wk old, 978-3904 cpm; 8 wk old, 1589–3467 cpm; and diabetic, 1532–3011 cpm. Data shown are the results combined from three independent experiments. Significant differences relative to the medium control values in each group are indicated as *, p < 0.005. No significant differences were noted among the different age groups. B, Splenic T cells from 8-wk-old NOD mice were incubated in the absence ({square}) or presence ({blacksquare}) of proinsulin (20 µg/ml), irradiated splenic APC, and various dilutions of a blocking anti-CD4 mAb (clone GK 1.5). Results are expressed as mean cpm ± SD of two independent experiments performed in triplicate. Significant differences relative to the medium control values are indicated as *, p < 0.001.

 
To determine the strain specificity of the observed T cell autoreactivity to proinsulin, we compared the responses to proinsulin of T cells from 8-wk-old NOD female mice with those from diabetes-resistant and age-matched BALB/c and B6 female mice. Fig. 7GoA illustrates that spleen T cells from NOD (SI = 29), but neither BALB/c (SI < 3) nor B6 (SI < 3) mice recognize proinsulin. In contrast, PLN T cells from all three strains respond well to proinsulin (SI > 7) (Fig. 7GoB). These results and those presented above in Fig. 5Go demonstrate that T cells localized in PLN draining the pancreas of diabetes-susceptible and diabetes-resistant mice are autoreactive to proinsulin, insulin, and GAD65. However, proinsulin and GAD65 are recognized as autoantigens by spleen CD4+ T cells only in diabetes-susceptible NOD mice.



View larger version (13K):
[in this window]
[in a new window]
  		FIGURE 7. Strain and tissue dependency of T cell proliferative response to proinsulin. Splenic (A) or PLN (B) T cells from 8-wk-old NOD, BALB/cJ, or B6 female mice were tested for their responsiveness to proinsulin. T cell responses to Ags are expressed as in Fig. 4Go. Significantly different (A) splenic T cell proliferative responses were noted between NOD, BALB/cJ, and B6 mice (*, p < 0.001). Background values of [3H]thymidine incorporation for medium-only controls of different mouse strains were: NOD splenic and PLN T cells, 1346–2530 cpm and 905-2359 cpm, respectively; BALB/cJ splenic and PLN T cells, 2315–3104 cpm; and B6 splenic and PLN T cells, 1289–2120 cpm and 1302–3501 cpm, respectively. Data shown are results from three independent experiments. Significant NOD, BALB/cJ, and B6 PLN T cell proliferative responses were observed compared with the medium control groups (*, p < 0.005), but no significant differences were observed between the PLN T cell responses of the three mouse strains.

 
Perinatal NOD T cell response to a dominant mouse proinsulin epitope

We next determined the epitope specificity of T cells responsive to proinsulin in perinatal NOD mice. Both splenic and PLN T cells from 2.5- to 4-wk-old NOD female mice were cocultured with splenic APC in the presence of 20 µg/ml of either mouse proinsulin (Fig. 8GoA), mouse proinsulin peptides, or control Ags. Splenic T cells from 2.5-wk-old mice responded (SI >= 3) only to mouse proinsulin, but not the mouse proinsulin peptides (data not shown). At 3–4 wk, both splenic and PLN T cells from female NOD mice showed high responses (SI >= 4) to the mouse proinsulin peptides B24-C33 and B24-C38 at a dose of 20 µg/ml, in addition to proinsulin (Fig. 8Go, B and C; p < 0.01). High responses to these peptides in the 2–50 µg/ml dose range were also observed (data not shown). In addition, responses to mouse proinsulin and LPS, but not OVA, were detected. Splenic and PLN T cells from 2.5- or 3- to 4-wk-old NOD mice did not respond to the GAD65 peptide 509–528 (data not shown). Thus, peptide B24-C33 seems to represent a predominant mouse proinsulin epitope recognized by both splenic and PLN T cells from naive perinatal NOD mice.



View larger version (26K):
[in this window]
[in a new window]
  		FIGURE 8. Perinatal NOD T cell reactivity to proinsulin (PI) peptides. A, Amino acid sequence of mouse proinsulin II. From left to right is the leader sequence (1–24), B chain (25–54), C peptide (57–87), and A chain (90–110). B and C, Splenic and PLN T cell reactivity to proinsulin peptides in 3- to ~4-wk-old female NOD mice. Spleen T cells from NOD female mice were stimulated in vitro with 20 µg/ml mouse proinsulin, 20 µg/ml mouse proinsulin peptides, mouse insulin B chain peptide, and human proinsulin peptide. LPS (10 ng/ml) was used as control Ag. Results are expressed as mean ± SD of two independent experiments performed in triplicate. Significant differences from the OVA control are indicated as *, p < 0.01; **, p < 0.001.

 
Responses of T cells from NOD mice immunized with mouse proinsulin II may be recalled by mouse proinsulin B24-C33

Having shown that peripheral T cells from naive perinatal NOD mice respond to mouse proinsulin II and proinsulin B24-C33, we next determined whether responses of T cells from NOD mice immunized with mouse proinsulin II may be recalled in vitro by mouse proinsulin B24-C33. We reasoned that such a result would further demonstrate that NOD T cells that respond to our mouse proinsulin II preparation are indeed specific for a mouse proinsulin T cell epitope rather than a contaminant protein(s) and/or peptide(s). The in vitro recall responses of draining lymph node T cells to intact proinsulin II or peptide B24-C33 from NOD female mice (4- to 5-wk-old; three to four mice/group) immunized intrafootpad with mouse proinsulin or proinsulin B24-C33 emulsified in CFA were analyzed. Indeed, the responses of NOD T cells primed to mouse proinsulin or mouse proinsulin B24-C33 were recalled in vitro by mouse proinsulin B24-C33 (Fig. 9Go). However, these proinsulin-primed T cells did not respond to mouse proinsulin B26-C34, hGAD65, hGAD65 509–528, or OVA. Thus, perinatal NOD proinsulin-specific T cells appear to react to an immunodominant B24-C33 peptide, demonstrating the mouse proinsulin T cell epitope specificity of NOD T cells that respond to our preparation of mouse proinsulin II.



View larger version (17K):
[in this window]
[in a new window]
  		FIGURE 9. Responses of T cells from NOD mice immunized with mouse proinsulin (PI) may be recalled by mouse proinsulin B24-C33. Female NOD mice (4–5 wk old; three to four mice/group) were immunized intrafootpad with either 50 µg mouse proinsulin ({blacksquare}) or proinsulin B24-C33 ({square}) emulsified in CFA. After 10 days, in vitro proliferative responses of T cells from the draining lymph nodes to mouse proinsulin (0.1 µg/ml), proinsulin B24-C33 (0.1 µg/ml), and other control proteins were evaluated by [3H]thymidine incorporation, as in Fig. 3Go. Background values of [3H]thymidine incorporation for medium-only controls from each experiment were 1120–1890 cpm. Data from three separate experiments expressed as mean SI ± SD are shown. Significant differences from the OVA controls are indicated as *, p < 0.001.

 
Cytokine production in response to mouse proinsulin and proinsulin peptides by peripheral T cells from perinatal NOD mice

To further characterize the T cell response to mouse proinsulin and its immunodominant peptides, we analyzed the cytokine secretion profile of both splenic and PLN T cells from 3- to 4-wk-old NOD female mice after in vitro stimulation with the peptides described above in Fig. 8Go. After 72 h, splenic T cells stimulated either by mouse proinsulin and its peptides (B24-C38, B24-C33, and B25-C33) or by LPS secreted 3-fold more IFN-{gamma} compared with control cultures with PBS or OVA (p = 0.001) (Fig. 10GoA). It is interesting to note that mouse proinsulin and mouse insulin B9–23 stimulated the secretion of similar levels of IFN-{gamma}. Interestingly, upon stimulation by either mouse proinsulin, mouse proinsulin B24-C38 and B24-C33, or human proinsulin B24-C33, PLN T cells secreted significant (p < 0.05) amounts of IFN-{gamma} that were comparable with those observed for splenic T cells (Fig. 10GoB). A significant increase in IL-2 secretion was also observed upon stimulation of splenic T cells by mouse proinsulin (p < 0.005), mouse proinsulin B24-C33 (p < 0.05), and LPS (p < 0.005), but not by other mouse proinsulin peptides (Fig. 10GoC). Similar increases in IL-2 production were also observed for PLN T cells after stimulation by mouse proinsulin (p < 0.001), mouse proinsulin B24-C38 (p < 0.005) and B24-C33 (p < 0.01), and human proinsulin B24-C33 (p < 0.01) (Fig. 10GoD). T cell secretion of IL-4 and IL-10 was not detectable under these conditions (data not shown), suggesting that these mouse proinsulin-activated T cells are of the Th1 phenotype. These results indicate that peripheral T cells from perinatal NOD mice are polarized toward a Th1-type immune response upon stimulation by mouse proinsulin and its peptides.



View larger version (23K):
[in this window]
[in a new window]
  		FIGURE 10. NOD T cell secretion of Th1 cytokines induced by mouse proinsulin (PI) and proinsulin peptides. Splenic and PLN T cells from 3- to 4-wk-old female NOD mice were cocultured as described in Fig. 8Go. After incubation for 72 h, the concentrations of IFN-{gamma} in the supernatants of cultures of splenic T cells (A) and PLN (B) T cells or of IL-2 secreted by splenic T cells (C) and PLN T cells (D) were assayed by ELISA. Data represent the mean cytokine concentrations in the supernatants ± SD of three independent experiments performed in triplicate. Significant differences from the OVA or PBS controls are indicated as *, p < 0.05, or **, p < 0.005.

 
Effect of immunization with proinsulin perinatally on the development of diabetes in NOD mice

Since T cell reactivity to proinsulin is detectable as early as 2.5 wk of age in NOD mice, we tested whether the course of development of T1D in these mice can be altered by immunization with proinsulin beginning at this time. Female NOD mice were administered (i.p.) 100 µg of either proinsulin or PBS emulsified in IFA at 18 days (before detectable insulitis and before weaning), 28 days, and 56 days of age, respectively. Diabetes was detectable by 15 and 19 wk of age in PBS/IFA- and proinsulin/IFA-immunized mice, respectively, demonstrating that the onset of T1D is delayed by ~4 wk after early immunization with proinsulin (Fig. 11GoA). By 20 wk of age, the incidence of T1D was 14.3% (2 of 14) in mice immunized with proinsulin and 35.7% (5 of 14) in control mice injected with PBS. At 30 wk of age, the incidence of T1D was 28.6% (4 of 14) in proinsulin-immunized mice and 85.7% (12 of 14) in PBS-treated control mice. Thus, immunization with mouse proinsulin before the onset of insulitis delays the onset of T1D by ~1 mo and protects against T1D in NOD mice (p < 0.01). In contrast, i.p. immunization of NOD female mice beginning at 5 wk of age with either proinsulin or insulin emulsified in IFA did not alter the kinetics of onset or incidence of T1D (data not shown).



View larger version (22K):
[in this window]
[in a new window]
  		FIGURE 11. Effect of immunization of NOD mice with proinsulin (PI) on the development of T1D is age and route dependent. A, Immunization with proinsulin before the development of insulitis protects against T1D in NOD mice. Female NOD mice (n = 14 mice/group) were immunized i.p. with either mouse proinsulin ({blacksquare}) or PBS ({circ}) emulsified in IFA at 18 days (before weaning), 28 days, and 56 days of age. Data are representative of two pooled experiments. B, Subcutaneous immunization with proinsulin after the onset of insulitis accelerates progression to diabetes. At 5 and 9 wk of age (indicated by arrows), NOD female mice (n = 8/group) were immunized s.c. with 100 µg of either mouse proinsulin ({diamondsuit}), porcine insulin (•), or OVA ({triangleup}) emulsified in IFA (Ag-IFA). Mice with BGL >200 mg/dl on 2 consecutive wk were considered diabetic. Data are representative of two independent experiments. C, Levels of serum IAA are enhanced after immunization (s.c.) with proinsulin. The sera of immunized mice (n = 6–8/group) were collected at various time points after immunization, and the levels of IAA were measured as described in Materials and Methods. Data are expressed as mean SI ± SD.

 
The influence of immunization with proinsulin on the onset of T1D also varies according to the route of immunization. Since s.c. immunization of NOD mice with insulin protects against T1D (26, 27), we tested whether s.c. treatment of NOD mice with proinsulin also protects them from T1D. Immunization (s.c.) of NOD mice with 100 µg of either proinsulin, insulin, or OVA emulsified in IFA at 5 wk of age and a booster shot 10 days later elicited the onset of T1D in each of these mice by 16 wk of age (Fig. 11GoB). At 19 wk of age, an unexpectedly high incidence of T1D (83%) was observed in mice immunized with proinsulin. In contrast, note that the incidence of T1D at this time was only 33% in OVA- and 17% in insulin-immunized mice, respectively. By 24 wk of age, the incidence of T1D was still significantly greater in proinsulin (83%) compared with OVA (50%, p < 0.05)- and insulin (17%, p < 0.01)-immunized NOD mice, respectively. However, by 30 wk of age, although the incidence of T1D did not differ significantly between proinsulin (83%)- and OVA (67%)-immunized mice, the differences between insulin (33%)- and OVA (67%)- or proinsulin (83%)-immunized mice were statistically significant (p < 0.05). These data demonstrate that s.c. administration of proinsulin accelerates the development of T1D in female NOD mice.

Recently, it was reported that NOD mice exhibit detectable levels of serum IAA beginning at 4 wk of age, and that the early appearance of IAA at 4–8 wk is strongly associated with the subsequent early onset of T1D (25). Given our finding that mouse proinsulin can accelerate the development of T1D, we evaluated the levels of serum IAA expression prospectively in NOD mice immunized s.c. with either proinsulin or control Ags. Administration of proinsulin with IFA at 4 wk of age significantly (p < 0.001) enhanced the level of IAA expression at 8 wk of age (IAA index = 0.12 ± 0.07) compared with that observed in mice immunized with OVA (IAA index = 0.03 ± 0.01) (Fig. 11GoC). A rapidly enhanced level of IAA expression was also observed in insulin/IFA-treated NOD mice (IAA index = 0.56 ± 0.36; p < 0.0001) compared with that seen in OVA/IFA-treated mice. After 8 wk of immunization, a further 10-fold rise in IAA expression was observed in mice that received proinsulin (IAA index = 1.38 ± 0.92) or OVA (IAA index = 0.65 ± 0.34) compared with the levels of IAA detected at 4 wk after immunization. Thus, enhanced IAA expression after immunization with proinsulin correlates with an accelerated development of T1D in NOD mice. These findings further implicate proinsulin as a pathogenic autoantigen during the perinatal period in the development of T1D.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Insulin and its precursor molecule, proinsulin, are the only known islet {beta} cell-specific autoantigens involved in T1D. Considerable evidence suggests a role for these autoantigens in the pathogenesis of T1D in humans (9, 14, 17, 28, 29, 30, 31), NOD mice (2, 3, 15, 21, 32, 33, 34, 35, 36), and Bio-Breeding rats (12). In NOD mice, T cell autoreactivity to hGAD65 (37, 38) and mGAD67 (21) was reported to precede that to insulin (3), indicating that insulin may not be an early target autoantigen in T1D. However, the latter conclusion does not exclude the possibility that a T cell proinsulin epitope(s), not present in insulin, is involved in the amplification of potentially autoreactive T cells in the periphery of diabetes-susceptible NOD mice. Thus, in this study, we analyzed the NOD T cell response to mouse proinsulin at different stages of development of T1D.

Our results demonstrate that whereas T cells from peripheral lymphoid tissues (e.g., spleen) of perinatal NOD mice do not react against insulin or insulin B chain 9–23, these T cells do respond strongly to proinsulin with the same kinetics as the GAD65-specific T cell response. The decamer peptide p24–33 of mouse proinsulin was found to be an immunodominant epitope in these NOD mouse T cell responses. In ongoing studies, we are deriving proinsulin-reactive NOD splenic T hybridomas to obtain a better estimate of the frequency of p24–33-responsive T cells and confirm the immunodominance of this proinsulin peptide. Our findings to date on NOD mouse T cell responses agree closely with those reported for subjects at high risk for T1D in whom T cell reactivity occurs to the proinsulin peptide B24-C36 (10, 14). This peptide possesses marked amino acid sequence similarity to peptides in hGAD65 and mGAD67 (39). Although we found that the proinsulin peptide is an early target autoantigen epitope, we found no evidence for NOD spleen T cell reactivity to the similar GAD65 peptide 506–518.

T cells derived from PLN draining the pancreas differ from splenic T cells in that they respond to GAD65, insulin, and proinsulin both in diabetes-susceptible NOD mice and diabetes-resistant BALB/c and B6 mice. These results agree closely with the reports that PLN T cells from diabetes-prone and diabetes-resistant mice are not tolerant to pancreatic islet {beta} cell Ags (40), and that islet-reactive peripheral blood T cells are detectable in normal healthy individuals (41). In double transgenic mice that express both a given Ag in their islet {beta} cells and the associated Ag-specific TCR on a large proportion of T cells, these autoreactive T cells function in a pancreas expressing the target autoantigen in islet {beta} cells without developing either insulitis or diabetes (42). Thus, peripheral regulatory mechanisms may control this autoreactivity in diabetes-resistant mice (27, 32). The failure of immunoregulation leading to T1D in NOD mice (1, 24) is likely to be reflected by the expansion of islet {beta} cell Ag-reactive T cells present in peripheral lymphoid tissues, such as the spleen. Thus, T cell autoreactivity in the spleen may be more informative about the extent of autoimmunity and progression to overt T1D.

In this study, mouse proinsulin II was produced in bacterial expression system. During the purification procedure, the protein was denatured with buffers containing SDS. Nonetheless, we showed in T cell proliferation assays that the final product was not toxic to T cells. A protein (e.g., proinsulin) may be modified posttranslationally by the nonenzymatic deamidation of asparagine and glutamine (43), and this may result in the altered T cell immunogenicity of this protein. This phenomenon was reported for the T cell recognition of dominant I-Ak-restricted hen egg lysozyme epitope, in which T cells were found to respond to a deamidated form of a hen egg lysozyme peptide, but not the native protein (43). Although we did not examine the deamidation of our proinsulin preparations, we do not think that this mechanism of posttranslational modification resulted in the altered immunogenicity of these preparations, for several reasons. First, NOD mice have a specific I-Ag7 MHC class II molecule in which the peptide-binding motif is mainly dictated by residues 6 (p6) and 9 (p9) (11). P6 is usually a large, hydrophobic residue (Leu, IIe, Met, or Val), whereas p9 is aromatic and hydrophobic (Tyr, Phe). The Asn or Gln residues are not found in these positions. Second, in support of this notion, human proinsulin peptide 17–31 binds tightly to the I-Ag7 motif. Third, we observed that mouse proinsulin peptide B24-C33, which does not contain any Asn or Gln residues, stimulated T cell proliferation. Finally, our in vivo data did not show any discordance of T cell responses to proinsulin peptides or whole proinsulin after immunization.

Previously, it was believed that immune tolerance to a protein Ag challenge was readily inducible during the neonatal period. However, it is now known that this immune privilege does not necessarily exist in neonates. Similar to T cells from adult mice, T cells from neonatal mice may be activated by the appropriate APC, costimulatory signals, and dose of Ags (44, 45, 46). These findings would predict the absence of T cell tolerance to islet autoantigens in neonatal autoimmune NOD mice. Consistent with this prediction, we found that T cell reactivity to the proinsulin, GAD67, and ICA69 autoantigens is indeed detectable in perinatal female NOD mice. The latter result agrees with the report of a loss of self-tolerance to ICA69 in NOD mice (47).

In neonates, Th1- or Th2-type immune responses are inducible depending on the nature of the APC, the type of adjuvant, and the dose of Ag (44, 45, 46). Low dose Ag administration to neonates may modulate Ag-activated T cells more easily than in adults because of a low frequency of Ag-responsive T cells. Since perinatal tolerization of NOD mice to GAD65 blocks the development of T1D (4, 37, 38), we examined the ability of perinatal administration of proinsulin to interfere with disease progression. Multiple immunization of perinatal NOD female mice with proinsulin in IFA before the onset of insulitis afforded significant protection against T1D. In contrast, when immunization with proinsulin was initiated after the development of insulitis, NOD mice were not protected from T1D and, moreover, displayed an accelerated onset of disease. Our result that perinatal vaccination with proinsulin blocks the development of T1D suggests that proinsulin may play an important role in the pathogenesis of T1D in NOD mice.

The ability of transgenic proinsulin expressed by MHC class II-bearing cells to prevent diabetes in NOD mice does not involve the generation of regulatory T cells (15). These mice are unable to generate diabetogenic T cells in thymic cultures, consistent with deletion of proinsulin-specific T cells induced by a high level of proinsulin expression in thymic APC. Nevertheless, T cells from these proinsulin transgenic mice respond to the proinsulin 9–36 and GAD65 524–543 peptides after in vivo immunization, although the T cell response to proinsulin p24–36 was absent, suggesting that high affinity pathogenic T cells reactive to this epitope are efficiently deleted (15).

Proinsulin is expressed by a rare subpopulation of follicular dendritic cells in the thymic medulla in normal mice (6), and proinsulin is expressed in the human thymus at a 4-fold higher level than insulin (7). It is conceivable, therefore, that thymic expression of proinsulin dictates negative selection of pro(insulin)-autoreactive T cells. Interestingly, however, we did not observe any differences in the level of proinsulin mRNA expressed in diabetes-susceptible NOD mice and diabetes-resistant strains. As a consequence of promiscuous or altered peptide binding to I-Ag7 (48), these autoreactive T cells may be expanded in the periphery upon stimulation with a cross-reactive autoantigen(s) such as GAD65, ICA69, or IA-2, and result in a high frequency of (pro)insulin-specific T cells. This could be critical for the control of susceptibility to T1D in NOD mice.

A similar mechanism was recently proposed for the control of development of experimental allergic encephalomyelitis (EAE), another mouse model of experimental autoimmune disease (49). In this model, although two major encephalitogenic epitopes of myelin proteolipid protein (PLP) 139–151 and 178–191 bind to I-As with similar affinity, the immune response to the PLP 139–151 epitope is always dominant. The immunodominance of this epitope in SJL mice appears to be due to the presence of expanded numbers of T cells reactive to PLP 139–151 in the peripheral repertoire of naive mice. This repertoire expansion is not associated with the PLP autoantigen or infectious environmental agents. The high frequency of PLP 139–151-reactive T cells in SJL mice is partly due to the lack of thymic deletion in response to PLP 139–151, as the DM20 isoform of PLP that lacks residues 116–150 is more abundantly expressed in the thymus than full-length PLP. Reexpression of PLP 139–151 in the embryonic thymus results in a significant reduction of PLP 139–151-reactive precursors in naive mice. As a result, the incidence of EAE is also significantly reduced. As an experimental model of an autoimmune disease, T1D in NOD mice presents with more immune defects than does EAE in SJL mice. Thus, further experimentation is required to test this model of thymic deletion and determine the role of thymic expression of proinsulin in the development of T1D in NOD mice.

Finally, it is important to note that in newly diagnosed patients with T1D, the magnitude of the peripheral T cell response to proinsulin was shown to be similar to that in islet autoantibody-negative relatives (50, 51). This result is in agreement with our finding that T cell reactivity to proinsulin decreases during progression to diabetes in NOD mice. The ability to protect NOD mice from diabetes by perinatal vaccination with proinsulin initiated at 18 days of age, i.e., before weaning, suggests that T cell reactivity to proinsulin plays an important role early in the development and pathogenesis of T1D in NOD mice. This notion is consistent with the capacity of T cells to respond to proinsulin in islet autoantibody-positive relatives of patients with T1D (10).


   Acknowledgments
 
We kindly thank Dr. Len Shultz for his generous gift of NOD.scid mice; all members of our laboratory for their valuable advice and encouragement; and Bibi Pettypiece for her expert assistance with the preparation of this manuscript.


   Footnotes
 
1 This work was supported by grants from the Canadian Institutes of Health Research (MT-5729), Juvenile Diabetes Research Foundation International (JDRFI), Canadian Institutes of Health Research /JDRFI Diabetes Interdisciplinary Research Program, and the National Health and Medical Research Council of Australia. W.C. was the recipient of a postdoctoral fellowship from the JDRFI. I.B. was the recipient of postdoctoral fellowships from the Foundation pour la Recherché Médicale and Canadian Diabetes Association. Back

2 W.C. and I.B. contributed equally to this work. Back

3 Address correspondence and reprint requests to Dr. Terry L. Delovitch, Autoimmunity/Diabetes Group, The John P. Robarts Research Institute, 1400 Western Road, London, Ontario, Canada N6G 2V4. E-mail address: del{at}rri.ca Back

4 Abbreviations used in this paper: GAD, glutamic acid decarboxylase; BGL, blood glucose level; EAE, experimental allergic encephalomyelitis; hGAD, human GAD; IAA, insulin autoantibody; ICA69, islet cell Ag p69; mGAD, mouse GAD; NOD, nonobese diabetic; PLN, pancreatic lymph node; PLP, proteolipid protein; PmB, Polymyxin B Sulfate; SI, stimulation index; T1D, type 1 diabetes. Back

Received for publication March 8, 2001. Accepted for publication August 24, 2001.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Delovitch, T. L., B. Singh. 1997. The nonobese diabetic mouse model of autoimmune diabetes: immune dysregulation gets the NOD. Immunity 7:727.[Medline]
  2. Wegmann, D. R., M. Norbury-Glasser, D. Daniel. 1994. Insulin-specific T cells are a predominant component of islet infiltrates in pre-diabetic NOD mice. Eur. J. Immunol. 24:1853.[Medline]
  3. Wong, F. S., J. Karttunen, C. Dumont, L. Wen, I. Visintin, I. M. Pilip, N. Shastri, E. G. Pamer, Jr C. A. Janeway. 1999. Identification of an MHC class I-restricted autoantigen in type 1 diabetes by screening an organ-specific cDNA library. Nat. Med. 5:1026.[Medline]
  4. Kaufman, D. L., M. Clare-Salzler, J. Tian, T. Forsthuber, G. S. Ting, P. Robinson, M. A. Aktinson, E. E. Sercarz, A. J. Tobin, P. V. Lehmann. 1993. Spontaneous loss of T cell tolerance to glutamic decarboxylase in murine insulin-dependent diabetes. Nature 366:69.[Medline]
  5. Tisch, R., X. D. Yang, S. M. Singer, R. S. Liblau, L. Fugger, H. O. McDevitt. 1993. Immune response to glutamic acid decarboxylase correlates with insulitis in non-obese diabetic mice. Nature 366:72.[Medline]
  6. Throsby, M., F. Homo-Delarche, D. Chevenne, R. Goya, M. Dardenne, J. M. Peau. 1998. Pancreatic hormone expression in the murine thymus: localization in dendritic cells and macrophages. Endocrinology 139:2399.[Abstract/Free Full Text]
  7. Pugliese, A., M. Zeller, A. Fernandez, J. L. 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 correlate with alleic variation at the INS VNTR-IDDM2 susceptibility locus for type 1 diabetes. Nat. Genet. 15:293.[Medline]
  8. Semple, J. W., T. L. Delovitch. 1991. Altered processing of human insulin by B lymphocytes from an immunologically insulin-resistant type I diabetic patient. J. Autoimmun. 4:277.[Medline]
  9. Congia, M., S. Patel, A. P. Cope, S. De Virgiliis, G. Sønderstrup. 1998. T cell epitopes of insulin defined in HLA-DR4 transgenic mice are derived from preproinsulin and proinsulin. Proc. Natl. Acad. Sci. USA 95:3833.[Abstract/Free Full Text]
  10. Rudy, G., N. Stone, L. C. Harrison, P. G. Colman, P. McNair, V. Brusic, M. B. French, M. G. Honeymoon, B. Tait, A. M. Lew. 1995. Similar peptides from two {beta} cell autoantigens, proinsulin and glutamic acid decarboxylase, stimulate T cells of individuals at risk for insulin-dependent diabetes. Mol. Med. 1:625.[Medline]
  11. Harrison, L. C., M. C. Honeyman, S. Trembleau, S. Gregori, F. Gallazzi, P. Augstein, V. Brusic, J. Hammer, L. Adorini. 1997. A peptide-binding motif for I-Ag7, the class II major histocompatibility complex (MHC) molecule of NOD and Biozzi AB/H mice. J. Exp. Med. 185:1013.[Abstract/Free Full Text]
  12. Griffin, A. C., W. Zhao, K. W. Wegmann, W. F. Hickley. 1995. Experimental autoimmune insulitis: induction by T lymphocytes specific for a peptide of proinsulin. Am. J. Pathol. 147:845.[Abstract]
  13. Bieg, S., C. Hanlon, C. S. Hampe, D. Benjamin, C. P. Mahoney. 1999. GAD65 and insulin B chain peptide (9–23) are not primary autoantigens in the type 1 diabetes syndrome of the BB rat. Autoimmunity 31:15.[Medline]
  14. Dubois-LaForgue, D., J. C. Carel, P. F. Bougnères, J. G. Guillet, C. Boitard. 1999. T-cell response to proinsulin and insulin in type 1 and pretype 1 diabetes. J. Clin. Immunol. 19:127.[Medline]
  15. French, M. B., J. Allison, D. S. Cram, H. E. Thomas, M. Dempsey-Collier, A. Silva, H. M. Georgiou, T. W. Kay, L. C. Harrison, A. M. Lew. 1997. Transgenic expression of mouse proinsulin II prevents diabetes in nonobese diabetic mice. Diabetes 46:34.[Abstract]
  16. Hostens, K., D. Pavlovic, Y. Zambre, Z. Li, C. Van Schravendijk, L. D. Eizirik, G. Pipeleers. 1999. Exposure of human islets to cytokines can result in disproportionately elevated proinsulin release. J. Clin. Invest. 104:67.[Medline]
  17. Heaton, D. A., B. A. Millward, I. P. Gray, Y. Tun, C. N. Hales, D. A. Pyke, R. D. Leslie. 1988. Increased proinsulin levels as an early indicator of {beta}-cell dysfunction in non-diabetic twins of type 1 diabetic patients. Diabetologia 31:182.[Medline]
  18. Heding, L., J. Ludvigsson, T. Kasperska-Czyzykowa. 1981. {beta} cell secretion in non-diabetics and insulin-dependent diabetics. Acta Med. Scand. 656:5.
  19. Mertens, N., E. Remaut, W. Fiers. 1995. Tight transcriptional control mechanism ensures stable high-level expression from T7 promoter-based expression plasmids. Bio/Technology 13:175.[Medline]
  20. Sambrook, J., E. F. Fritisch, T. Maniatis. 1989. Molecular Cloning: A Laboratory Manual 2nd Ed. Cold Spring Harbor Press, Cold Spring Harbor.
  21. Elliott, J. F., H. Y. Qin, S. Bhatti, D. K. Smith, R. K. Singh, E. Fraga, T. Dillon, J. Lauzon, B. Singh. 1994. Immunization with the larger isoform of mouse glutamic acid decarboxylase (GAD67) prevents autoimmune diabetes in NOD mice. Diabetes 43:1494.[Abstract]
  22. Cameron, M. J., G. A. Arreaza, M. Grattan, C. Meagher, S. Sharif, M. D. Burdick, R. M. Strieter, D. N. Cook, T. L. Delovitch. 2000. Differential expression of CC chemokines and the CCR5 receptor in the pancreas is associated with progression to type 1 diabetes. J. Immunol. 165:1102.[Abstract/Free Full Text]
  23. Deltour, L., P. Leduque, N. Blume, O. Madsen, P. Dubois, J. Jami, D. Bucchini. 1993. Differential expression of the two nonallelic proinsulin genes in the developing mouse embryo. Proc. Natl. Acad. Sci. USA 90:527.[Abstract/Free Full Text]
  24. Cameron, M. J., G. A. Arreaza, P. Zucker, S. W. Chensue, R. M. Strieter, S. Chakarabarti, T. L. Delovitch. 1997. IL-4 prevents insulitis and insulin-dependent diabetes mellitus in nonobese diabetic mice by potentiation of regulatory T helper-2 cell function. J. Immunol. 159:4686.[Abstract]
  25. Yu, L., D. T. Robles, N. Abiru, P. Kaur, M. Rewers, K. Kelemen, G. S. Eisenbarth. 2000. Early expression of anti-insulin antibodies of humans and the NOD mouse: evidence for early determination of subsequent diabetes. Proc. Natl. Acad. Sci. USA 97:1701.[Abstract/Free Full Text]
  26. Daniel, D., D. R. Wegmann. 1996. Protection of nonobese diabetic mice from diabetes by intranasal or subcutaneous administration of insulin peptide B (9–23). Proc. Natl. Acad. Sci. USA 93:956.[Abstract/Free Full Text]
  27. Muir, A., A. Peck, M. Clare-Salzler, Y. H. Song, J. Cornelius, R. Luchetta, J. Krischer, N. Maclaren. 1995. Insulin immunization of nonobese diabetic mice induces a protective insulitis characterized by diminished intraislet interferon-{gamma} transcription. J. Clin. Invest. 95:628.
  28. Naquet, P., J. Ellis, D. Tibensky, A. Kenshole, B. Singh, R. Hodges, T. L. Delovitch. 1988. T cell autoreactivity to insulin in diabetic and related nondiabetic individuals. J. Immunol. 140:2569.[Abstract]
  29. Kuglin, B., F. A. Gries, H. Kolb. 1988. Evidence of IgG autoantibodies against human proinsulin in patients with IDDM before insulin treatment. Diabetes 37:130.[Abstract]
  30. Spinas, G. A., O. Snorgaard, S. G. Harrling, M. Oberholzer, W. Berger. 1992. Elevated proinsulin levels related to islet cell antibodies in first-degree relatives of IDDM patients. Diabetes Care 15:632.[Abstract]
  31. Bergerot, I., N. Abien, V. Maguer, C. Thivolet. 1994. Oral administration of human insulin to NOD mice generates CD4+ T cells that suppress adoptive transfer of diabetes. J. Autoimmun. 8:323.
  32. Bergerot, I., G. A. Arreaza, M. Cameron, H. Chou, T. L. Delovitch. 1997. Role of T cell anergy and suppression in susceptibility to IDDM. Res. Immunol. 148:281.
  33. Atkinson, M. A., N. K. Maclaren, R. Luchetta. 1990. Insulitis and diabetes in NOD mice reduced by prophylactic insulin therapy. Diabetes 39:933.[Abstract]
  34. Zhang, Z. J., L. Davidson, G. Eisenbarth, H. L. Weiner. 1991. Suppression of diabetes in nonobese diabetic mice by oral administration of porcine insulin. Proc. Natl. Acad. Sci. USA 88:10252.[Abstract/Free Full Text]
  35. Harrison, L. C., M. Dempsey-Collier, D. R. Kramer, K. Takahashi. 1996. Aerosol insulin induces regulatory CD8 {gamma}{delta} T cells that prevent murine insulin-dependent diabetes. J. Exp. Med. 184:2167.[Abstract/Free Full Text]
  36. Sai, P., A. S. Rivereau. 1996. Prevention of diabetes in the nonobese diabetic mouse by oral immunological treatments. Diabetes Metab. 22:341.[Medline]
  37. Zechel, M. A., J. F. Elliott, M. A. Atkinson, B. Singh. 1998. Characterization of novel T cell epitopes on 65kDa glutamic acid decarboxylase relevant in autoimmune responses in NOD mice. J. Autoimmun. 11:83.[Medline]
  38. Petersen, J. P., A. E. Karrlsen,