Several studies have provided indirect evidence in support of a role for β cell-specific Th2 cells in regulating insulin-dependent diabetes (IDDM). Whether a homogeneous population of Th2 cells having a defined β cell Ag specificity can prevent or suppress autoimmune diabetes is still unclear. In fact, recent studies have demonstrated that β cell-specific Th2 cell clones can induce IDDM. In this study we have established Th cell clones specific for glutamic acid decarboxylase 65 (GAD65), a known β cell autoantigen, from young unimmunized nonobese diabetic (NOD) mice. Adoptive transfer of a GAD65-specific Th2 cell clone (characterized by the secretion of IL-4, IL-5, and IL-10, but not IFN-γ or TGF-β) into 2- or 12-wk-old NOD female recipients prevented the progression of insulitis and subsequent development of overt IDDM. This prevention was marked by the establishment of a Th2-like cytokine profile in response to a panel of β cell autoantigens in cultures established from the spleen and pancreatic lymph nodes of recipient mice. The immunoregulatory function of a given Th cell clone was dependent on the relative levels of IFN-γ vs IL-4 and IL-10 secreted. These results provide direct evidence that β cell-specific Th2 cells can indeed prevent and suppress autoimmune diabetes in NOD mice.
Insulin-dependent diabetes mellitus (IDDM)3 is characterized by autoimmune-mediated destruction of the insulin-secreting β cells found in the islets of Langerhans located in the pancreas. Both β cell-specific Ab and T cell responses can be detected in persons with IDDM or at increased risk for the disease and in spontaneous animal models for IDDM such as the nonobese diabetic (NOD) mouse (1, 2, 3). Studies in the NOD mouse have established that the primary mediators of β cell destruction are CD4+ and CD8+ T cells (for review, see Ref. 4). Currently, the events that lead to the breakdown of self tolerance within the T cell compartment and subsequent progression of the diabetogenic response remain largely ill defined. However, several studies suggest that β cell-specific autoimmunity is in part the result of defective peripheral immunoregulation (3, 4, 5). Indeed, pathogenic β cell-specific CD4+ T cells typically exhibit a Th1 cell phenotype, and injection of cytokines that promote Th1 cell development and effector function exacerbate disease in NOD mice (6).
Conversely, CD4+ Th2 cells that inhibit Th1 cell differentiation through the secretion of IL-4 and IL-10 (7, 8) are believed to have a regulatory role in IDDM (3, 4, 5). This hypothesis is supported by studies in the NOD mouse that have demonstrated that β cell autoimmunity is prevented or suppressed following induction of Th2 cell reactivity by a variety of means, including 1) systemic administration of IL-4 (9, 10), 2) transgene mediated secretion of IL-4 by β cells (11, 12), and 3) immunization with specific β cell autoantigens such as insulin (13). Furthermore, analysis of cytokine RNA from the pancreas demonstrated that IL-4 expression at the onset of islet infiltration corresponds to a nondestructive type of insulitis found predominantly in NOD male mice, which typically develop overt diabetes at a lower frequency than females (14). In contrast, the breakdown of immunoregulation involving β cell-specific Th1 cells is thought to be partly due to a defect in efficient induction of Th2 cells (9, 15).
Evidence demonstrating that a homogeneous population of Th2 cells with a defined β cell Ag specificity can immunoregulate IDDM is currently lacking. For example, in studies where protection has been induced in NOD mice, the population of Th2 effector cells is typically heterogeneous in terms of β cell Ag specificity and may consist of other types of regulatory Th cells. Similar caveats exist for studies demonstrating that diabetes can be prevented in young NOD mice upon adoptive transfer of short term Th2 cell lines established from immunized animals. Finally, the role of Th2 cells in IDDM has been brought into question by observations that 1) disease progression is not affected in NOD mice lacking a functional IL-4 gene (16, 17); and 2) IDDM is exacerbated in young NOD mice receiving β cell-specific Th2 cell clones established from NOD mice transgenic for the BDC2.5 clonotypic TCR (18).
In the present study, we have investigated the potential immunoregulatory function of β cell-specific Th2 cells in autoimmune diabetes through the use of a panel of Th cell clones specific for glutamic acid decarboxylase 65 (GAD65). GAD65 is among the first β cell autoantigens targeted in the NOD mouse and is believed to have a key, albeit undefined, role in both murine and human IDDM (19, 20, 21, 22, 23, 24). Furthermore, β cell autoimmunity is prevented or suppressed in NOD mice immunized at various ages with either GAD65 protein or peptides (17, 19, 20, 25, 26). The protection observed in these studies correlated with the induction of GAD65-specific Th2 cell reactivity. Here we provide evidence that adoptive transfer of a GAD65-specific Th2 cell clone established from young unimmunized NOD mice can effectively prevent and/or suppress established β cell autoimmunity in recipient mice.
Materials and Methods
NOD/Lt and NOD.scid mice were housed and bred under specific pathogen-free conditions and allowed access to NIH diet 31A (Ralston Purina, St. Louis, MO). Currently, IDDM develops in ∼85% of NOD/Lt female mice by 1 yr of age in our colony maintained at University of North Carolina (Chapel Hill, NC). The establishment and screening of NOD mice homozygous for an inactivated IL-4 gene (NOD.IL4null) have previously been described (17).
Assessment of diabetes and insulitis
Mice were monitored weekly for the development of glycosuria with Diastix (Ames, Elkhart, IN). Two successive positive measurements were considered diagnostic of diabetes onset. Insulitis was assessed by histology. Pancreases were prepared for histology by fixation in neutral buffered formalin followed by paraffin embedding. A minimum of five sections, 90 μm apart, were cut from each block, stained with hematoxylin and eosin, and viewed by light microscopy. A minimum of 30 islets were scored for each animal. The severity of insulitis was scored as either peri-insulitis (islets surrounded by a few lymphocytes) or intrainsulitis (lymphocytic infiltration of the islets).
The cloning and preparation of murine β cell autoantigens GAD65, heat shock protein 60 (HSP60), and carboxypeptidase H (CPH) have been previously described (19). Briefly, the cDNAs were engineered to encode six histidine residues at the COOH terminus of each protein. Recombinant proteins were expressed in baculovirus (GAD65, CPH) and Escherichia coli (HSP60) expression systems and were purified using an Ni2+-conjugated resin (Qiagen, Chatsworth, CA). Each recombinant protein was further purified by preparative SDS-PAGE, electroelution, and extensive dialysis in PBS. Peptides were synthesized using standard F-moc chemistry on a Rainin Symphony (Rainin Instruments, Emeryville, CA) at the peptide synthesis facility of University of North Carolina. The purity of the peptides was verified by reverse phase HPLC and mass spectroscopy.
Establishment of GAD65-specific CD4+ Th cell clones
GAD65-specific T cell clones were established by culturing 7.5 × 106 splenocytes from 4-wk-old unimmunized NOD female mice in a 24-well plate in 1.5 ml of RPMI 1640, 5 × 10−5 M 2-ME, 1 mM sodium pyruvate, 1× nonessential amino acids, 1 mM glutamine, 100 U/ml of penicillin/streptomycin (base medium) supplemented with 1.0% NOD serum, and 10 μg/ml of murine GAD65 for 7 days. T cells (1 × 106) harvested on a Lympholyte M gradient (Cedarlane Laboratories, Hornby, Canada) were cultured with 5 × 106 irradiated (3000 rad) NOD splenocytes in 1.5 ml of the above medium and 10 μg/ml GAD65 in a 24-well plate. Three days later the cultures were supplemented with base medium containing 20 U/ml of murine IL-2 (BD PharMingen, San Diego, CA) and 10% FBS and maintained for an additional 3 days. CD4+ T cells were purified by magnetic separation using anti-CD4 magnetic beads (Miltenyi Biotec, Auburn, CA) as recommended by the manufacturer. GAD65-specific CD4+ T cell clones were established by culturing 0.5 and 1.0 T cells with 2 × 105 irradiated NOD splenocytes/well in 96-well round-bottom plates in base medium containing 10% FBS, 20 U/ml IL-2, and 10 μg/ml GAD65. Subclones of the parental 6H1 T cell clone were established in an identical manner. Following expansion, the monoclonality of a given T cell clone was determined by flow cytometric analysis of TCR Vα 2 and Vβ14 chain usage. 1F2 cells were established in a similar manner, with the exception of using the GAD65-specific p524–543 peptide to stimulate and propagate T cells instead of intact GAD65. Once established, T cell clones were maintained in culture on a 21-day growth cycle. Briefly, 2 × 106 T cell clones were stimulated with 2 × 107 irradiated (3000 rad) splenocytes and 10 μg/ml GAD65 or peptide in an upright T-25 tissue culture flask. On day 3, base medium containing 10% FBS, 20 U/ml IL-2 plus 40 U/ml IL-4 (BD PharMingen) was added to the cultures, and T cell clones were expanded accordingly up to 21 days.
T cell clone proliferation and cytokine assays
Th cell clone proliferation was measured by culturing in triplicate 2.0 × 104 T cells plus 2.0 × 105 irradiated (3000 rad) splenocytes/well (0.2 ml) in a 96-well round-bottom microtiter plate for 72 h with the appropriate Ag. Proliferation was assessed by measuring the amount of [3H]thymidine incorporation (1 μCi/well) during the last 18 h of culture and was expressed as a stimulation index (mean cpm of response to Ag divided by mean cpm with medium only). To measure cytokine secretion, supernatants from cultures established as described above were harvested after 48 h, pooled, and used in a capture ELISA system. The levels of IFN-γ, IL-4, IL-5, and IL-10 were determined in triplicate in 0.1 ml of supernatant. Abs were obtained from BD PharMingen, and ELISA was conducted as recommended by the manufacturer. Standard curves were established to quantitate the amount of the respective cytokines in the culture supernatants. The lower limits of detection for IFN-γ, IL-4, IL-5, and IL-10 were typically 50, 25, 30, and 30 pg/ml, respectively.
Adoptive transfer of GAD65-specific T cell clones
For adoptive transfer experiments, Th cell clones were harvested on a Lympholyte M gradient 7 days after Ag stimulation, extensively washed, and resuspended in PBS. Two-week-old NOD or 4-wk-old NOD.scid female mice received two i.p. or i.v. (tail vein) injections of 107 cells (0.1 ml) within 14 days. Twelve-week-old NOD or NOD.IL4null female mice received two i.v. injections of 2 × 107 cells (0.1 ml) over 14 days. Animals were then monitored for diabetes.
In vitro cytokine assay
Lymphocyte cytokine secretion in cultures prepared from adoptive transfer recipients in response to the panel of β cell autoantigens was determined as previously described (23). Briefly, a splenocyte suspension was prepared from individual mice in ice cold PBS. The splenocyte suspension was immediately centrifuged at 400 × g for 5 min at 4°C and resuspended at 2.5 × 106 cells/ml in base medium supplemented with 2% Nutridoma-SP (Roche, Indianapolis, IN). Splenocytes (0.2 ml/well) were incubated in 96-well flat-bottom microtiter plates in the presence of 10 μg/ml of GAD65, HSP60, or CPH or 25 μg/ml of peptide. Six wells were used for each β cell autoantigen. Culture supernatants were harvested and pooled for each Ag treatment after 48 h, with a capture ELISA used to measure IFN-γ and IL-4 in 0.1 ml of culture supernatant (in triplicate) as described above.
Enzyme-linked immunospot (ELISPOT)
ImmunoSpot M200 plates (Cellular Technology, Cleveland, OH) were coated overnight at 4°C with either 2 μg/ml anti-IFN-γ Ab (R4-6A2, BD PharMingen) or 4 μg/ml anti-IL-4 Ab (11B11, BD PharMingen) prepared in PBS. Plates were blocked with 1% BSA-PBS for a minimum of 2 h at room temperature and then washed four times with PBS. Splenocytes were prepared from individual mice as described above, with the exception of being resuspended in HL-1 medium (BioWhittaker, Walkersville, MD). Splenocytes were then plated at 5 × 105 cells/well (0.2 ml/well). Pancreatic lymph nodes were pooled within a given treatment group, and the resulting suspension was prepared in HL-1 medium and plated at 2.5 × 105 cells/well with 5 × 105 cells/well irradiated (3000 rad) splenocytes harvested from NOD.IL4null mice. Ag or peptide was added to triplicate wells at a final concentration of 10 or 25 μg/ml, respectively. The plates were incubated for 24 h (IFN-γ) or 48 h (IL-4) at 37°C in 5.5% CO2 and then washed three times with PBS followed by an additional three washes with 0.025% Tween 20-PBS. Biotinylated anti-IFN-γ (XMG1.2, BD PharMingen) or anti-IL-4 (BVD6-24G2, BD PharMingen) was added at 2 and 4 μg/ml, respectively, in 1% BSA-PBS (0.1 ml/well). After overnight incubation at 4°C, plates were washed three times with 0.025% Tween 20-PBS and incubated with streptavidin-HRP (BD PharMingen; 1/2000) for 2 h at room temperature. This was followed by three washes with 0.025% Tween 20-PBS and three washes with PBS only. Development solution consisted of 0.8 ml of 3-amino-9-ethyl-carbazole (Sigma, St. Louis, MO; 20 mg dissolved in 2.0 ml of dimethylformamide) added to 24 ml of 0.1 M sodium acetate (pH 5.0), plus 0.12 ml of 3.0% hydrogen peroxide; 0.2 ml was added per well.
Establishment and characterization of GAD65-specific Th cell clones from NOD mice
We and others have previously shown that GAD65-specific CD4+ T cell reactivity is first detected in 4-wk-old NOD female mice (19, 20). To gain insight into the role of these Th cells in disease progression and the peptide epitope(s) targeted, we established GAD65-specific Th cell clones from the spleens of unimmunized 4-wk-old NOD female mice using intact murine GAD65. Our conditions included supplementing the cultures with IL-2, but no effort was made to selectively promote either Th1 or Th2 cell clones. Of the 89 GAD65-specific Th cell clones that were established, 88 were shown to be specific for a peptide epitope spanning amino acid residues 217–236 (p217–236). A minimum of 10 distinct clonotypes were detected based on TCR Vβ-chain usage. The p217–236-specific Th cell clones exhibited either a Th0 or a Th1 cell phenotype as determined by IFN-γ, IL-4, IL-5, and IL-10 secretion. The remaining Th cell clone, designated 6H1, differed from the rest in both peptide specificity and phenotype. 6H1 cells proliferated in response to intact GAD65 or p290–309 and secreted IL-4, IL-5, and IL-10, but not IFN-γ, indicating a Th2 cell phenotype. Further cloning of the 6H1 Th2 cells via limiting dilution resulted in 40 subclones. Thirty-nine of these, represented by the 6H1E subclone, proliferated in response to GAD65 and p290 and exhibited a typical Th2 cytokine secretion profile (Fig. 1⇓). Interestingly, one subclone, designated 6H1L, secreted significant amounts of IFN-γ in addition to IL-4, IL-5, and IL-10 (Fig. 1B⇓). Messenger RNA encoding TGF-β could not be detected in either the 6H1 parental cells or the 6H1E and 6H1L subclones as determined by RT-PCR.
The 1F2 Th2 cell clone, established from the spleens of 8-wk-old unimmunized NOD female mice, was also used in this study. 1F2 cells secreted IL-4, IL-5, and IL-10, but not IFN-γ, in response to the GAD65-specific peptide p524–543 (Fig. 1⇑B). It is noteworthy that 1F2 Th2 cells did not respond to intact GAD65 (Fig. 1⇑A).
To assess the short-term in vivo function of the 6H1E, 6H1L, and 1F2 Th cell clones, 2-wk-old NOD female mice received two i.v. injections of 107 cells over 14 days or were left untreated. Four weeks after the last injection, spleen and pancreatic lymph node cultures were prepared, and cytokine secretion in response to GAD65 protein or GAD65-derived peptides was measured. Cultures prepared from the spleen or pancreatic lymph nodes of untreated NOD mice yielded a typical Th1 cell cytokine profile in response to GAD65 or p290–309 and only a marginal response to p524–543 (Fig. 2⇓A). In marked contrast, significantly reduced levels of IFN-γ and a concomitant increase in IL-4 were detected in response to GAD65 by cultures prepared from 6H1E recipient NOD mice (Fig. 2⇓). Furthermore, enhanced IL-4 secretion in response to p290–309 was detected in these cultures relative to that in cultures prepared from untreated animals (Fig. 2⇓B). Cultures prepared from 1F2 recipient mice exhibited a modest decrease in IFN-γ secretion in response to GAD65 and only limited, but significant, IL-4 secretion in response to p524–543 (Fig. 2⇓). Furthermore, these cultures did not secrete IL-4 in response to whole GAD65 (Fig. 2B⇓), consistent with data demonstrating that the 1F2 Th2 cell clone responds to p524–543, but not to intact Ag (Fig. 1⇑A). A distinct profile of cytokine secretion was observed in cultures prepared from mice receiving the 6H1L Th cell clone. In these cultures, both IFN-γ and IL-4 secretions were increased in response to GAD65 and p290–309 relative to those in cultures established from untreated animals (Fig. 2⇓).
To determine the relative frequency of the Th cell clones in vivo, ELISPOT analyses were performed on splenocyte and pancreatic lymph node cultures prepared from recipient mice. The general cytokine profiles detected via ELISPOT in cultures prepared from the respective treatment groups were consistent with the ELISA data (compare Figs. 2⇑ and 3⇓). A greater frequency of 6H1E, 6H1L, and 1F2 cells was detected in splenocyte vs pancreatic lymph node cultures, as determined by responses to p290–309 and p524–543, respectively. Furthermore, the frequency of IL-4-secreting Th cells in response to p290–309 in either spleen or pancreatic lymph node cultures was similar for the 6H1E and 6H1L Th cell clone recipients (Fig. 3⇓B). These frequencies of IL-4-secreting Th cells were consistently greater than those detected in spleen and pancreatic lymph node cultures established from recipients of the 1F2 Th cell clone in response to p524–543 (Fig. 3⇓B).
Adoptive transfer of the GAD65-specific 6H1E Th2 cell clone inhibits insulitis and the development of diabetes in NOD recipient mice
To investigate the effect of the GAD65-specific Th cell clones on IDDM, adoptive transfer experiments were conducted in NOD female mice exhibiting different stages of disease progression. As described above, 2-wk-old NOD female mice received two i.v. injections of 107 cells of a given clone and were subsequently monitored for diabetes up to 35 wk of age. NOD mice at this young age lack detectable β cell-specific T cell and Ab reactivity and show no infiltration of the pancreas. The majority of mice (12 of 15) receiving 1F2 cells developed IDDM at a frequency and time of onset indistinguishable from untreated NOD female mice (12 of 15; Fig. 4⇓A). In contrast, a significant reduction in the frequency of overt diabetes was observed in the group of mice receiving the 6H1E Th2 cell clone (2 of 15; p < 10−3, by χ2 test) relative to that in untreated animals (Fig. 4⇓A). Interestingly, adoptive transfer of the 6H1L cells had no significant effect on the onset and frequency (11 of 15) of diabetes in recipient mice (Fig. 4⇓A), suggesting that secretion of IFN-γ blocked the immunoregulatory function of the Th cell clone. We also performed adoptive transfer experiments with four p217–236 Th1 cell clones and found that the development of insulitis was, in fact, enhanced in recipient mice (data not shown).
We next investigated whether the 6H1E Th2 cell clone, in addition to preventing diabetes, could suppress progression of established β cell-specific autoimmunity. Euglycemic NOD female mice, 12 wk of age, were used as recipients. Typically, 12-wk-old NOD female mice exhibit maximal β cell-specific T cell and Ab reactivity in addition to significant insulitis. Mice received two i.v. injections of 2 × 107 1F2, 6H1E, or 6H1L cells over 14 days and were monitored for overt diabetes up to 35 wk of age. The group of mice receiving the 6H1E Th2 cell clone exhibited a significantly reduced frequency of diabetes (1 of 10; p = 0.009, by χ2 test) compared with untreated mice (9 of 12; Fig. 4⇑B). However, the majority of mice injected with the 6H1L (8 of 10) or 1F2 (9 of 10) Th cell clones continued to develop diabetes (Fig. 4⇑B).
Histological analysis of pancreases obtained from nondiabetic mice receiving the respective Th cell clones at 2 or 12 wk of age further confirmed the protective effect associated with the 6H1E Th2 cell clone. Whereas the majority of islets exhibited extensive intrainsulitis in the pancreases of mice injected with the 1F2 or 6H1L Th cell clones, a significant number of islets remained free of cellular infiltration in the pancreases of 6H1E cell recipient mice in either age group tested (Tables I⇓ and II⇓). Strikingly, while a significant number of islets exhibited peri- and intrainsulitis, further progression of insulitis was suppressed in mice receiving the 6H1E Th2 cell clone at 12 wk of age; the frequency of insulitis in 35-wk-old 6H1E cell recipient mice was similar to that typically observed in untreated 12-wk-old NOD female mice.
Diabetes prevention in 6H1E recipient mice correlates with a GAD65-specific Th2 cell cytokine profile in vitro
To determine the immune status of NOD mice receiving the 1F2, 6H1E, or 6H1L Th cell clones at 2 or 12 wk of age, splenocyte cultures were prepared from recipient mice, and cytokine secretion in response to GAD65, p290–309, or p524–543 was measured. As demonstrated in Figs. 5⇓A and 6A, similar amounts of IFN-γ secretion in response to GAD65 and p290–309 were detected in cultures prepared from both nondiabetic and diabetic mice left untreated or receiving the 1F2 cells. In addition, IL-4 was not detected in response to p524–543 (Figs. 5⇓B and 6B). As in the short term adoptive transfer experiments described above (Figs. 2⇑ and 3⇑), reduced IFN-γ and increased IL-4 secretion in response to GAD65 and p290–309 were detected in cultures prepared from nondiabetic 6H1E cell recipient mice at 2 wk (Fig. 5⇓) or 12 wk (Fig. 6⇓) of age. Furthermore, a distinct Th1 cell cytokine profile similar to that in untreated animals was detected in cultures prepared from the few 6H1E cell recipients that developed overt diabetes (Figs. 5⇓ and 6⇓). This result suggests that IDDM progressed in these animals due to the lack of in vivo persistence or sufficient expansion of the 6H1E Th2 cell clone. This contrasted with the 6H1L cell recipient mice, in which cultures prepared from both nondiabetic and diabetic recipients displayed enhanced IFN-γ and IL-4 secretion in response to GAD65 and p290–309 relative to untreated mice (Figs. 5⇓ and 6⇓).
GAD65-specific Th cell clones do not traffick to the islets of NOD.scid recipient mice
The reduced frequency of insulitis observed in NOD mice receiving the 6H1E Th2 cell clone at 2 or 12 wk of age (Tables I⇑ and II⇑) strongly suggested that protection was primarily mediated in the periphery as opposed to the pancreatic islets. However, cells found infiltrating the pancreas of nondiabetic recipient mice may have nevertheless consisted of 6H1E Th2 cells. To investigate this issue further the 1F2, 6H1E, and 6H1L Th cell clones were adoptively transferred into NOD.scid mice. Six weeks after the final injection of the Th cell clones, NOD.scid recipients were assessed for insulitis in addition to in vitro T cell reactivity in the spleen and pancreatic lymph nodes. Histological examination of the pancreases prepared from NOD.scid mice receiving any of the three Th cell clones showed no detectable insulitis (Table III⇓). This differed from NOD.scid mice, which exhibited significant intrainsulitis after receiving a suboptimal dose of splenocytes from diabetic NOD donor mice (Table III⇓). Importantly, cytokine secretion in response to GAD65 and p290–309 was detected in both splenocyte and pancreatic lymph node cultures prepared from NOD.scid mice receiving the 6H1E or 6H1L Th cell clones (Fig. 7⇓). In agreement with results obtained from wild-type NOD recipients, modest IL-4 secretion in response to p524–543 was detected in splenocyte or pancreatic lymph node cultures established from NOD.scid mice receiving the 1F2 Th2 cell clone.
Adoptive transfer of the 6H1E Th2 cell clone delays the onset of diabetes in NOD.IL4null recipient mice
Next, we determined whether β cell-specific Th1 cell reactivity was directly suppressed by the 6H1E Th2 cell clone or required recruitment of other β cell-specific regulatory Th2 cells. For this set of experiments, we used NOD.IL4null mice, which lack a functional IL-4 gene and consequently are unable to generate typical Th2 effector cells (16, 17). Twelve-week-old NOD.IL4null or wild-type NOD mice received two injections of the 6H1E Th2 cell clone as described above and were monitored for diabetes up to 45 wk of age. As expected, the majority of wild-type NOD mice receiving the 6H1E Th2 cell clone remained free of diabetes by 45 wk of age (Fig. 8⇓A). Furthermore, a significant reduction in IFN-γ and a corresponding increase in IL-4 secretion were detected in response to GAD65 and p290–309 (Fig. 8⇓B). As described above, IFN-γ, but not IL-4 secretion in response to GAD65 or p290–309 was detected in cultures prepared from the few wild-type NOD mice that had received the 6H1E cells and developed diabetes (Fig. 8⇓B).
A significant delay in the development of diabetes was observed in the NOD.IL4null mice receiving the 6H1E Th2 cell clone compared with untreated NOD.IL4null mice (Fig. 8⇑A). However, the majority of the 6H1E NOD.IL4null recipients (7 of 10) eventually developed overt diabetes. Adoptive transfer of the 1F2 Th2 cell clone had no effect on the onset or frequency of diabetes in NOD.IL4null recipients (data not shown). Interestingly, enhanced secretion of both IL-4 and IFN-γ was detected in response to p290–309 in cultures prepared from either nondiabetic or diabetic NOD.IL4null recipients monitored long term (Fig. 8⇑C). Furthermore, this cytokine profile, which resembles that observed with 6H1L cells, suggests that endogenous IL-4 was necessary to maintain the Th2 cell phenotype of the 6H1E cells in vivo and, in turn, may explain why overt diabetes eventually developed in recipient animals. We established individual Th cell clones from the NOD.IL4null recipient mice and confirmed that most of the transferred 6H1E cells obtained the capacity to secrete IFN-γ (data not shown).
We recently demonstrated that immunizing NOD mice with the GAD65-specific peptides p217–236 and p290–309 prepared in IFA prevented IDDM and correlated with the induction of GAD65-specific regulatory Th2 cells (17). This Th2 response was not limited to GAD65, but had spread to other β cell autoantigens. However, peptide immunization of NOD.IL4null mice mediated no protection or Th2 responses to either GAD65 or other β cell autoantigens. To determine whether a similar set of events could be observed here, IL-4 secretion in response to p217–236 and the candidate β cell autoantigens HSP60 and CPH was measured in splenocyte cultures prepared from NOD or NOD.IL4null mice 4 wk after adoptive transfer of 6H1E cells. As demonstrated in Fig. 9⇓, cultures prepared from wild-type NOD mice receiving the 6H1E cells exhibited elevated levels of IL-4 relative to untreated NOD mice in response to p290–309 as well as p217–236, HSP60, and CPH. In contrast, IL-4 secretion was detected in cultures prepared from NOD.IL4null recipients only in response to p290–309 and not to the other β cell autoantigens assayed.
The relative contribution of β cell-specific Th2 cells in regulating IDDM is currently unclear. In part, this is due to a lack of evidence demonstrating that a homogeneous population of Th2 cells with a defined β cell Ag specificity can prevent or suppress the diabetogenic response. In the present study GAD65-specific Th2 cell clones were established from unimmunized NOD mice using culture conditions that promote in vitro expansion of Th cells that have encountered β cell autoantigen in vivo. Accordingly, the Th cell clones established in this manner reflect GAD65-specific clonotypes that occur spontaneously in vivo and provide evidence that β cell-specific Th2 cells normally exist in unmanipulated NOD mice. However, the frequency of GAD65-specifiic Th2 cells is low, consistent with our work and others demonstrating that Th1-like responses to GAD65 or other β cell autoantigens predominate (19, 20, 25, 26). Indeed, efficient induction of regulatory Th2 cells in NOD mice has often required an exogenous source of IL-4 (9, 10, 27) That β cell-specific Th2 cells can indeed immunoregulate IDDM was demonstrated by adoptive transfer of the 6H1E Th2 cell clone, which prevented the progression of insulitis (Tables I⇑ and II⇑) and the development of diabetes in NOD recipient mice (Figs. 4⇑ and 8A⇑). The immunoregulatory function of this GAD65-specific Th2 cell clone was effective either before or after the establishment of β cell-specific Th1 cells. Prevention of IDDM by the Th cell clones correlated with a Th2-like recall response to p290-309 and GAD65 (Figs. 5⇑, 6⇑, and 8⇑) in addition to other β cell autoantigens (Fig. 9⇑). Notably, this T cell reactivity was dependent on the relative levels of IFN-γ vs Th2 cell cytokines secreted by the Th cell clone.
A cytokine profile indicative of Th2 cell reactivity was detected in splenocyte and pancreatic lymph node cultures established from nondiabetic NOD mice monitored either short or long term following adoptive transfer of the 6H1E Th2 cell clone (Figs. 2⇑, 3⇑, 5⇑, 6⇑, and 8⇑). These cultures typically exhibited reduced Th1 cell reactivity and enhanced IL-4 secretion in response to GAD65 and p290–309. Strikingly, a Th2-like cytokine profile was also established in cultures prepared from 12-wk-old NOD recipients, an age at which β cell autoimmunity is well underway. The existing anti-GAD65-specific Th1 cell reactivity observed in these cultures presumably represented established Th1 effector cells at the time of adoptive transfer of 6H1E cells. The immunoregulatory function of the transferred 6H1E cells was also characterized by induction of Th2 cell reactivity to other β cell autoantigen determinants, such as HSP60, CPH, and an additional GAD65-specific epitope, p217–236 (Fig. 9⇑). This intermolecular determinant spread of the Th2 cell phenotype has previously been shown to be important for IDDM prevention in NOD mice immunized with GAD65-derived peptides (17). The inability of the 6H1E cells to promote Th2 reactivity to other β cell autoantigens in NOD.IL4null mice probably explains why diabetes eventually developed in the recipients. However, in wild-type NOD recipients the recruitment of additional β cell-specific Th2 cells may have aided in maintaining the Th2 cell phenotype of the 6H1E cells by providing an endogenous source of IL-4 in addition to amplifying immunoregulatory effects. Nevertheless, the delay in the onset of diabetes in NOD.IL4null recipients indicates that the frequency of 6H1E cells and the level of IL-4 secretion were sufficient to directly regulate the diabetogenic response shortly after transfer. The failure of the 1F2 Th2 cell clone to prevent IDDM despite secreting higher levels of IL-4 and IL-10 compared with 6H1E cells (Fig. 1⇑, C and E) may reflect an inability of the T cells to persist or sufficiently expand in vivo. This idea is consistent with the observation that 1F2 cells do not respond to GAD65 protein in vitro (Fig. 1⇑A).
In contrast, failure of 6H1L cells to protect NOD recipients from overt diabetes demonstrated the importance of the relative balance between IFN-γ and Th2-like cytokines in determining the immunoregulatory function of Th cell clones (Fig. 4⇑). In addition to secreting equivalent levels of IL-4 and IL-10 in vitro (Fig. 1⇑B), the 6H1L and 6H1E Th cell clones trafficked to the spleen and pancreatic lymph nodes to the same extent upon adoptive transfer, as seen by the frequency of IL-4-secreting Th cells responding to p290–309 (Fig. 3⇑B). Nevertheless, NOD mice receiving the 6H1L Th cell clone developed diabetes with a similar onset and frequency as untreated animals (Fig. 4⇑). The inability of the 6H1E Th2 cell clone to protect NOD.IL4null mice long term provided additional evidence that the relative levels of Th1 vs Th2 cell cytokines are the key to determining the immunoregulatory efficacy of Th cell clones. In the absence of endogenous IL-4, 6H1E cells began to secrete significant levels of IFN-γ while continuing to secrete IL-4 (Fig. 8⇑C) and IL-10 (data not shown) in response to GAD65 and p290–309. Consequently, the majority of recipient mice developed overt diabetes, albeit with a delayed onset relative to untreated NOD.IL4null mice (Fig. 8⇑A) or animals receiving the 1F2 Th2 cell clone (R. Tisch, unpublished observations). This delayed onset of diabetes probably reflected the time during which levels of IFN-γ secretion became sufficient to suppress the immunoregulatory capacity of the 6H1E cells. We have established Th cell clones from NOD.IL4null recipient mice and confirmed that the enhanced IFN-γ secretion detected in response to p290–309 was derived from 6H1E cells based on expression of the 6H1 clonotypic Vα2/Vβ14 TCR (R. Tisch, unpublished observations). Interestingly, adoptive transfer of an insulin-specific Th1 cell clone has been reported to prevent diabetes in NOD recipient mice (28). However, in this instance, the immunoregulatory function of the Th1 cell clone was due to secretion of TGF-β, which is not expressed by 6H1E or 6H1L cells.
Another important aspect of the protection associated with the 6H1E Th2 cell clone was the inhibition of insulitis in recipient animals. NOD mice receiving 6H1E cells at 2 or 12 wk of age exhibited significantly reduced frequencies of insulitis relative to mice left untreated or receiving the 1F2 or 6H1L Th cell clones (Tables I⇑ and II⇑). This observation is in agreement with a recent study demonstrating that GAD65-specific Th cell lines established from NOD mice expressing IL-4 in β cells limited the capacity of diabetogenic splenocytes to infiltrate the pancreas of recipient animals in coadoptive experiments (12). Based on the lack of detectable infiltration in NOD.scid recipients (Table III⇑) 6H1E cells do not appear to traffick to the islets. This contrasts with our GAD65-specific Th cell clones that exhibit a classical Th1 cell phenotype, i.e., IFN-γ+, IL-4−, IL-10−, and TGF-β−, which do infiltrate the islets of NOD.scid recipients (R. Tisch, unpublished results). Together, these findings demonstrate that immunoregulation by the 6H1E Th2 cell clone occurred largely in the periphery of recipient mice. The pancreatic lymph nodes in which activation and regulation of β cell-specific T cells have been shown to occur (29, 30, 31) probably provide key sites for immunoregulation by 6H1E cells. Indeed, a low, but significant frequency of IL-4-secreting T cells was detected in response to GAD65 and p290–309 in pancreatic lymph node cultures prepared from NOD mice receiving 6H1E cells (Fig. 3⇑B).
We suggest the following scenario to explain the immunoregulatory effect of the 6H1E Th2 cell clone. Upon adoptive transfer, 6H1E cells seed peripheral sites, including the spleen and pancreatic lymph nodes. At these sites a low frequency of 6H1E cells sufficient to establish a microenvironment conducive to Th2 cell development persist through recognition and stimulation by endogenous GAD65. The secretion of IL-4 would promote Th2 while suppressing Th1 differentiation of uncommitted β cell-specific Th cell precursors. Additionally, IL-4 and IL-10 would modify the function of APCs to further limit commitment along the Th1 pathway (31, 32). Induction of other β cell-specific Th2 cells amplifies the immunoregulatory effect, in addition to providing a source of endogenous IL-4, which is required to maintain the phenotype of 6H1E cells in vivo. In this manner, infiltration of β cell-specific Th1 cells is effectively inhibited. As indicated by the 6H1L cell transfer data, secretion of IFN-γ inhibits efficient Th2 cell commitment despite the presence of IL-4 and would also be expected to enhance the responsiveness of naive Th cell precursors to IL-12 to promote Th1 cell differentiation (33, 34).
Recent work by Poulin and Haskins has demonstrated that Th2 cell clones established from the BDC2.5 TCR transgenic mouse accelerated the onset of diabetes upon adoptive transfer into young NOD recipients (18). Currently, it is not clear why the GAD65-specific and the BDC2.5 derived Th2 cell clones prevent and promote IDDM, respectively. One obvious difference is the specificity of the respective Th2 cell clones. The identity of the autoantigen recognized by BDC2.5 Th cells has yet to be determined; however, differences in the level and anatomical presentation of the corresponding peptide epitopes may impact on a variety of in vivo T cell properties, including trafficking and effector function. A direct comparison between the GAD65-specific and BDC2.5-derived Th2 cell clones may provide further insight into the factors that govern regulatory Th2 cell function.
In summary, this study provides evidence that GAD65-specific Th2 cells exist in unmanipulated NOD mice, and that adoptive transfer of a GAD65-specific Th2 cell clone can effectively prevent and suppress IDDM. Nevertheless, it is apparent that immunoregulation of IDDM is complex, involving not only Th2 cells, but also a number of types of regulatory T and non-T cells (28, 35, 36, 37, 38, 39, 40, 41, 42). The task at hand is to determine the relative contributions and potential interactions between these different regulatory cells to gain a complete understanding of the events leading to the breakdown of self tolerance in IDDM and other T cell-mediated autoimmune diseases.
↵1 This work was supported by National Institutes of Health Grant 5RO1DK52365.
↵2 Address correspondence and reprint requests to Dr. Roland Tisch, Department of Microbiology and Immunology, Mary Ellen Jones Building, Room 804, Campus Box 7290, University of North Carolina, Chapel Hill, NC 27599-7290. E-mail address:
↵3 Abbreviations used in this paper: IDDM, insulin-dependent diabetes mellitus; NOD, nonobese diabetic; CPH, carboxypeptidase H; ELISPOT, enzyme-linked immunospot; GAD65, glutamic acid decarboxylase 65; hsp60, heat shock protein 60.
- Received February 7, 2001.
- Accepted March 22, 2001.
- Copyright © 2001 by The American Association of Immunologists