In this study, we have investigated the use of plasmid DNA (pDNA) vaccination to elicit Th2 effector cell function in an Ag-specific manner and in turn prevent insulin-dependent diabetes mellitus (IDDM) in nonobese diabetic (NOD) mice. pDNA recombinants were engineered encoding a secreted fusion protein consisting of a fragment of glutamic acid decarboxylase 65 (GAD65) linked to IgGFc, and IL-4. Intramuscular injection of pDNA encoding GAD65-IgGFc and IL-4 effectively prevented diabetes in NOD mice treated at early or late preclinical stages of IDDM. This protection was GAD65-specific since NOD mice immunized with pDNA encoding hen egg lysozyme-IgGFc and IL-4 continued to develop diabetes. Furthermore, disease prevention correlated with suppression of insulitis and induction of GAD65-specific regulatory Th2 cells. Importantly, GAD65-specific immune deviation was dependent on pDNA-encoded IL-4. In fact, GAD65-specific Th1 cell reactivity was significantly enhanced in animals immunized with pDNA encoding only GAD65-IgGFc. Finally, NOD.IL4null mice treated with pDNA encoding GAD65-IgGFc and IL-4 continued to develop diabetes, indicating that endogenous IL-4 was also required for disease prevention. These results demonstrate that pDNA vaccination is an effective strategy to elicit β cell-specific Th2 regulatory cell function for the purpose of preventing IDDM even at a late stage of disease development.

Insulin-dependent diabetes mellitus (IDDM)3 is the result of autoimmune destruction of the insulin-producing β cells found in the islets of Langerhans. The disease process is characterized by a progressive infiltration of lymphocytes and monocytes into the islets (insulitis), culminating in massive destruction of the β cells (1, 2, 3). Studies in the nonobese diabetic (NOD) mouse, a spontaneous murine model for IDDM, have demonstrated that CD4+ and CD8+ T cells are the primary mediators of β cell destruction (for review, see Ref. 4). Furthermore, temporal analyses indicate that only a few β cell autoantigens such as glutamic acid decarboxylase 65 (GAD65) and insulin are targeted in the early stages of disease development by CD4+ T cells (5, 6). However, as progression to IDDM occurs, several other β cell proteins are recognized. Nevertheless, T cell recognition of GAD65 and insulin is believed to be important in the progression of both human and NOD IDDM (7, 8, 9, 10, 11).

Diabetogenic CD4+ T cells typically exhibit a Th1 cell phenotype which is characterized by the secretion of large amounts of IFN-γ and IL-2 (12). Accordingly, it has been suggested that β cell-specific Th2 cells may have a regulatory role in IDDM (12). It is well established that Th2 cells suppress Th1 effector cell differentiation through secretion of IL-4 and IL-10 (13, 14). Furthermore, various studies have provided evidence, albeit indirect, that Th2 cells may regulate the diabetogenic response. For example, recent onset diabetics have been reported to exhibit reduced IL-4 expression in peripheral T cells (15), and an inverse correlation between β cell-specific T cell and Ab reactivity has been detected in these individuals (16). Finally, diabetes can be prevented in NOD mice by a number of conditions which are reported to promote Th2 effector cell function such as systemic administration of IL-4 (17, 18) or immunizing young animals with GAD65 or insulin (5, 6, 19).

Exploiting the regulatory function of β cell-specific Th2 cells may therefore provide an effective means to selectively prevent IDDM in a clinical setting (20). This approach is advantageous since regulatory T cells can be induced in an Ag-specific manner, which in turn suppress the development of other β cell-specific Th1 effector cells through a cytokine-mediated bystander mechanism. Indeed, progression to overt IDDM can be suppressed in NOD mice at late preclinical stages of disease by i.v. or i.p. injection of soluble GAD65 or GAD65-specific peptides prepared in the appropriate adjuvant, respectively (21, 22, 23). However, multiple immunizations with relatively high doses of Ag or peptide are necessary to effectively induce regulatory Th2 cells (22, 23). Extrapolating a similar set of conditions to suppress established β cell autoimmunity in patients may prove to be highly problematic.

Recently, the application of plasmid DNA (pDNA) for the purpose of vaccination has generated a great deal of interest. Immunization with pDNA encoding an immunogen and/or cytokine has proven to be an effective approach to elicit long-term T cell and Ab immunity in infectious disease and cancer models (for review, see Ref. 24). A number of properties including 1) in vivo persistence (25), 2) versatility in the choice of the route of administration, 3) an intrinsic adjuvant effect (26, 27), and 4) the relative ease to construct and produce large amounts of recombinants make the use of pDNAs an appealing strategy for immunotherapy of autoimmune diseases. Indeed, initial reports have demonstrated that induction of experimental autoimmune encephalomyelitis can be prevented by immunizing rodents with pDNAs encoding a variable β-chain segment of an encephalogenic TCR (28) or with epitopes derived from autoantigens used to induce the disease process (29, 30). However, the efficacy of pDNA immunization to suppress established autoimmunity was not examined in these studies.

In the current study, we have assessed the efficacy of pDNA immunization to promote and maintain β cell-specific Th2 regulatory cell function for the purpose of inhibiting IDDM. Here, we show that administration of pDNA vaccines encoding a secreted form of GAD65 and IL-4 can prevent and, more importantly, suppress established β cell autoimmunity in NOD mice. Long-term protection was associated with the induction of GAD65-specific Th2 regulatory cells and was dependent on both exogenous and endogenous IL-4.

The Jw4303 vector employed for this study contains a transcriptional unit composed of a CMV promoter/enhancer element and polyadenylation and transcriptional termination sequences derived from the bovine growth hormone gene (31). To facilitate expression of two cDNA inserts, the Jw4303 vector was modified to contain two individual transcriptional units. cDNA encoding full-length murine IL-4 was cloned from a CD4+ Th2 clone via RT-PCR and the nucleotide sequence confirmed. IgGFc fusion proteins were engineered using the signal pIg vector (R&D Systems, Minneapolis,MN) which contains a human CD33 signal sequence and genomic DNA consisting of the hinge, CH2, and CH3 exons derived from human IgG4. cDNA encoding murine GAD65 or hen egg lysozyme (HEL) spanning nt 656-1070 and 175–270, respectively, were PCR amplified with Pfu DNA polymerase (Strategene, La Jolla, CA). Oligonucleotide primers were used which contained the appropriate flanking restriction enzymes to subclone into the signal pIg vector and a splice donor sequence at the 3′ end of the amplicons to facilitate splicing with IgGFc RNA. The resulting DNA constructs were then subcloned into the wild-type or modified Jw4303 vectors. To test expression and secretion of the GAD65-IgGFc and HEL-IgGFc fusion proteins, COS-7 cells were transfected with the respective pDNAs prepared in Lipofectamine (Life Technologies, Rockville, MD) in accordance with the manufacturer’s directions. Culture supernatants were harvested 2 days after transfection. The IgGFc fusion proteins were immunoprecipitated via protein G-Sepharose (Pharmacia, Piscataway, NJ), resolved on SDS-PAGE, and analyzed by Western blot using a mouse anti-human IgG-HRP conjugate (Jackson ImmunoResearch, West Grove, PA). Levels of IL-4 secretion in COS-7 culture supernatants were determined by an IL-4-specific capture ELISA.

NOD/Lt mice were housed and bred under specific pathogen-free conditions and allowed access to National Institutes of Health diet 31A (Purina, St. Louis, MO). Currently in our colony maintained at the University of North Carolina, IDDM develops in ∼80% of NOD/Lt female mice by one year of age. Establishment and screening of NOD mice homozygous for an inactivated IL-4 gene (NOD.IL4null) have previously been described (23).

Mice were monitored weekly for the development of glycosuria with Ames Diastix (Fisher, Pittsburgh, PA). Glycosuric values of >3 for two successive measurements were considered diagnostic of diabetes onset. Insulitis was assessed by histology. Pancreata were prepared for histology by fixing in neutral-buffered Formalin and then embedding in paraffin. The fixed blocks were sectioned and stained with hematoxylin and eosin. A minimum of five sections, each differing by 90 μm, was cut for each block, and slides were viewed by light microscopy. A minimum of 30 individual islets was scored for each animal. The severity of insulitis was scored as either peri-insulitis (islets surrounded by a few lymphocytes) or intrainsulitis (lymphocytic infiltration into the interior of the islets).

pDNA was prepared from DH5α Escherichia coli using a Qiagen endotoxin-free kit (Qiagen, Chatsworth, CA) and resuspended at 1.0 mg/ml in PBS. NOD or NOD.IL4null female mice 4 wk of age received three i.m. injections of 50 μl (50 μg) of pDNA in each quadricep over 21 days. NOD female mice 12 wk of age were similarly treated with the exception of receiving four i.m. injections over a period of 28 days.

Total cellular RNA was prepared from muscle tissue of individual mice using TRIzol (Life Technologies), and treated with RNase-free DNase. First-strand cDNA was synthesized using the SuperScript preamplification system (Life Technologies) and oligo(dT) primers according to the manufacturer’s instructions. First-strand cDNA encoding either HEL-IgGFc or GAD65-IgGFc was subsequently amplified by PCR using Taq DNA polymerase and oligonucleotide primers specific for the 5′ end of CD33 (CGGAATTCATGCCGCTGCTGCTACTGCTG) and the 3′ end of IgG4 (CGGAATTCTCATTTACCCGGAGACAGGGAGAGGCTCTTCTGCGT). PCR cycling conditions consisted of denaturation at 94°C for 1 min, annealing at 60°C for 1 min, and extension at 72°C for 2 min for 30 cycles, and a final extension cycle at 72°C for 10 min. Using similar cycling conditions, transcripts encoding IL-4 (forward primer, TCGGCATTTTGAACGAGGTC; reverse primer, GAAAAGCCCGAAAGAGTCTC) and β-actin (forward primer, GCATTGTTACCAACTGGG; reverse primer, GTCAGGATCTTCATGAGG) were also amplified. The PCR products were resolved on a 1.5% agarose gel and detected by ethidium bromide staining.

The cloning and preparation of murine β cell autoantigens GAD65, carboxypeptidase H (CPH), and heat shock protein (HSP) 60 have been previously described (5). Briefly, the cDNAs were engineered to encode six histidine residues at the COOH terminus of each protein. Recombinant GAD65 and CPH were expressed in a baculovirus expression system and purified using a Ni2+-conjugated resin (Qiagen). HSP60 was produced in an E. coli expression system and similarly purified. Each recombinant protein was further purified by preparative SDS-PAGE, electroeluted, and dialyzed extensively against PBS.

Lymphocyte cytokine secretion in response to the panel of β cell autoantigens was determined as previously described (23). Briefly, a spleen cell suspension was prepared from individual mice in ice-cold PBS. The spleen cell suspension was immediately centrifuged at 400 × g for 5 min at 4°C, and resuspended at 2.5 × 106 cells/ml in culture medium consisting of RPMI 1640 supplemented with 2% Nutridoma-SP (Boehringer Mannheim, Indianapolis, IN), 5 × 10−5 M 2-ME, 1× nonessential amino acids, 1 mM sodium pyruvate, and 100 U/ml penicillin. Spleen cells (0.2 ml/well) were incubated in a 96-well flat-bottom microtiter plate in the presence of 20 μg/ml β cell autoantigen. Six wells were used for each β cell autoantigen. Culture supernatants were harvested and pooled for each Ag treatment after 48 h, and a capture ELISA was used to measure IFN-γ, IL-4, and IL-5 in 0.1 ml of culture supernatant in duplicate. Abs were obtained from PharMingen (San Diego, CA) and the 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 limit of detection for IFN-γ, IL-4, and IL-5 was 50, 25, and 35 pg/ml, respectively.

ImmunoSpot M200 plates (Cellular Technology, Cleveland, OH) were coated overnight at 4°C with either 2 μg/ml anti-IFN-γ Ab (R4-6A2; PharMingen) or 4 μg/ml anti-IL-4 Ab (11B11; 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. Spleen cells were prepared from individual mice as above with the exception of being resuspended in HL-1 medium (BioWhittaker, Walkersville, MD). Splenocytes were then plated at 5 × 105/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/well with 5 × 105/well irradiated (3000 rad) splenocytes harvested from NOD.IL4null mice. Ag was added to triplicate wells at a final concentration of 20 μg/ml. The plates were incubated for either 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; PharMingen) or anti-IL-4 (BVD6-24G2; PharMingen) was added at 2 and 4 μg/ml, respectively, in 1% BSA/PBS plus 0.025% Tween 20 (0.1 ml/well). After overnight incubation at 4°C, plates were washed three times with 0.025% Tween 20/PBS, and streptavidin-HRP (PharMingen) was added at 1/2000 dilution for 2 h at room temperature. This was followed by three washes with 0.025% Tween 20/PBS and three washes with PBS only. The solution to develop the plates consisted of 0.8 ml of 3-amino-9-carbazole (Sigma, St. Louis, MO; 20 mg dissolved in 2.0 ml of dimethyl formamide) added to 24 ml 0.1 M sodium acetate (pH 5.0) plus 0.12 ml of 3.0% hydrogen peroxide; 0.2 ml of the developing solution was added per well.

Neonatal NOD mice 24–48 h of age received a single i.p. injection of 5 × 106 spleen cells from diabetic donors (32) mixed with or without 106 CD4+ T cells purified from nondiabetic NOD mice treated with pDNA or left untreated. CD4+ T cells were purified from splenocytes by magnetic separation using anti-CD4 Ab conjugated to magnetic beads (Miltenyi Biotec, Auburn, CA). T cells were eluted by flushing the magnetic column with PBS containing 0.5% FCS as recommended by the manufacturer. Typically, the purity of the CD4+ T cells was >95%. Recipient mice were monitored for diabetes up to 10 wk of age.

To investigate the immunotherapeutic efficacy of pDNA vaccination to prevent IDDM in an Ag-specific manner, a pDNA recombinant was established encoding a fusion protein consisting of a portion of murine GAD65 and IgGFc (JwGAD65). The GAD65 fragment contains a minimum of three peptide epitopes which consist of amino acid residues 206–220, 217–236, and 290–309, which are recognized by CD4+ T cell clones derived from unimmunized or GAD65-immunized NOD mice (23, 33, 34). Furthermore, we have recently demonstrated that overt diabetes is prevented in NOD mice coimmunized with the 217–236 and 290–309 peptides prepared in IFA at 4 or 12 wk of age (23). In an attempt to efficiently stimulate CD4+ T cells, the GAD65-specific sequence was fused to a human IgG4Fc molecule. In this way, the fusion protein is secreted, and GAD65-specific epitopes should be preferentially processed and presented via the MHC class II pathway. Additional pDNA recombinants encoding murine IL-4 (JwIL-4) or both IL-4 and GAD65-IgGFc (JwGAD65 + IL4) were established for the purpose of enhancing Th2 effector cell differentiation. Finally, two pDNAs encoding a HEL-IgGFc fusion protein either alone (JwHEL) or in combination with IL-4 (JwHEL + IL-4) were established to serve as controls.

The panel of pDNAs was shown to be functional based on detection of protein secreted by COS-7 cells transfected with the corresponding recombinants (Fig. 1,A). In addition, RNA transcripts encoding HEL-IgGFc, GAD65-IgGFc, or IL-4 were detected via RT-PCR in muscle tissue prepared from NOD mice lacking a functional IL-4 gene (NOD.IL4null) 4 wk after the final of three injections with the corresponding pDNAs (Fig. 1 B).

FIGURE 1.

Detection of pDNA-encoded protein and RNA. A, COS-7 cells were transfected with JwHEL, JwGAD65, JwHEL + IL-4, JwGAD65 + IL-4, or mock transfected (no pDNA) and IgGFc fusion proteins were immunoprecipitated from culture supernatant. Protein was detected via Western blot. B, Four-week-old NOD. IL4null female mice received three i.m. injections of JwHEL, JwGAD65, JwIL-4, JwHEL + IL-4, or JwGAD65 + IL-4 over 21 days. Four weeks after the final immunization, pDNA-encoded RNA transcripts were amplified via RT-PCR from muscle tissue prepared from the site of injection, and the resulting amplicons were resolved on a 1.5% agarose gel containing ethidium bromide. Detected were amplicons (a) representative of appropriately processed, full-length HEL-IgGFc (870 bp) or GAD65-IgGFc (1200 bp) transcripts, or amplicons of the expected molecular size encoding portions of IL-4 (b; 200 bp) or β-actin (c; 350 bp). M, 1-kb ladder (Life Technologies). Amplicons were not detected in the absence of RT.

FIGURE 1.

Detection of pDNA-encoded protein and RNA. A, COS-7 cells were transfected with JwHEL, JwGAD65, JwHEL + IL-4, JwGAD65 + IL-4, or mock transfected (no pDNA) and IgGFc fusion proteins were immunoprecipitated from culture supernatant. Protein was detected via Western blot. B, Four-week-old NOD. IL4null female mice received three i.m. injections of JwHEL, JwGAD65, JwIL-4, JwHEL + IL-4, or JwGAD65 + IL-4 over 21 days. Four weeks after the final immunization, pDNA-encoded RNA transcripts were amplified via RT-PCR from muscle tissue prepared from the site of injection, and the resulting amplicons were resolved on a 1.5% agarose gel containing ethidium bromide. Detected were amplicons (a) representative of appropriately processed, full-length HEL-IgGFc (870 bp) or GAD65-IgGFc (1200 bp) transcripts, or amplicons of the expected molecular size encoding portions of IL-4 (b; 200 bp) or β-actin (c; 350 bp). M, 1-kb ladder (Life Technologies). Amplicons were not detected in the absence of RT.

Close modal

To initially assess the immunotherapeutic efficacy of genetic vaccination, NOD female mice 4 wk of age received three i.m. injections of 50 μg of pDNA in each quadricep over 21 days and were then monitored for diabetes up to 52 wk of age. No significant difference in the onset and frequency of diabetes was observed between the unimmunized group (75%) and NOD mice coimmunized with JwHEL and JwIL-4 (79%) (Fig. 2,A). Notably, the majority of NOD mice receiving JwGAD65 continued to develop diabetes (71%) (Fig. 2,A). Only in NOD mice coimmunized with JwGAD65 and JwIL-4 was there a significant reduction in the frequency of diabetes (29%, p = 0.04 vs unimmunized, p = 0.02 vs JwHEL and JwIL-4, χ2 test) (Fig. 2,A). Furthermore, the majority of islets found in these animals were free of insulitis (Table I). In contrast, a high frequency of intrainsulitis was detected in the pancreata of diabetes-free NOD mice which were left untreated, coimmunized with JwHEL and JwIL-4, or immunized with JwGAD65 only (Table I).

FIGURE 2.

NOD mice are protected from diabetes following immunization with pDNAs encoding GAD65-IgGFc and IL-4. A, Female NOD mice 4 wk of age received three i.m. injections of JwHEL + JwIL-4 (n = 14), JwGAD65 (n = 14), and JwGAD65 + JwIL-4 (n = 14) over 21 days or were left untreated (n = 12). B, Female NOD mice 12 wk of age received four i.m. injections of JwHEL + IL-4 (n = 12), JwGAD65 (n = 12), and JwGAD65 + IL-4 (n = 12) over 28 days or were left untreated (n = 10). The treatment groups were monitored for diabetes on a weekly basis.

FIGURE 2.

NOD mice are protected from diabetes following immunization with pDNAs encoding GAD65-IgGFc and IL-4. A, Female NOD mice 4 wk of age received three i.m. injections of JwHEL + JwIL-4 (n = 14), JwGAD65 (n = 14), and JwGAD65 + JwIL-4 (n = 14) over 21 days or were left untreated (n = 12). B, Female NOD mice 12 wk of age received four i.m. injections of JwHEL + IL-4 (n = 12), JwGAD65 (n = 12), and JwGAD65 + IL-4 (n = 12) over 28 days or were left untreated (n = 10). The treatment groups were monitored for diabetes on a weekly basis.

Close modal
Table I.

Frequency of insulitis in nondiabetic NOD mice immunized at 4 wk of agea

TreatmentNo InfiltrationPeri-insulitisIntrainsulitis
<50%>50%
Unimmunized (n = 3) 5 (4.7%) 8 (7.5%) 19 (18.0%) 74 (69.8%) 
JwHEL+ JwIL-4 (n = 3) 0 (0%) 2 (2.1%) 23 (24.2%) 70 (73.7%) 
JwGAD65 (n = 4) 0 (0%) 0 (0%) 16 (12.5%) 112 (87.5%) 
JwGAD65+ JwIL-4 (n = 10) 249 (71.1%)b 72 (20.6%) 14 (4.0%)b 15 (4.3%)b 
TreatmentNo InfiltrationPeri-insulitisIntrainsulitis
<50%>50%
Unimmunized (n = 3) 5 (4.7%) 8 (7.5%) 19 (18.0%) 74 (69.8%) 
JwHEL+ JwIL-4 (n = 3) 0 (0%) 2 (2.1%) 23 (24.2%) 70 (73.7%) 
JwGAD65 (n = 4) 0 (0%) 0 (0%) 16 (12.5%) 112 (87.5%) 
JwGAD65+ JwIL-4 (n = 10) 249 (71.1%)b 72 (20.6%) 14 (4.0%)b 15 (4.3%)b 
a

Insulitis was determined by counting islets in nondiabetic mice at 52 wk of age.

b

, p < 10−3 vs unimmunized or JwHEL + JwIL-4-treated mice as determined by χ2 analysis.

Of interest was whether pDNA encoding other fragments of GAD65 could mediate a similar protective effect upon immunization. For this purpose, a pDNA (JwGAD651550–1856) encoding the carboxyl terminus of GAD65 (spanning nt 1550–1856) coupled to IgGFc was employed. This GAD65 fragment contains the previously identified T cell epitopes p509–528 and p524–543 (6). As demonstrated in Table II, NOD mice treated at 4 wk of age with JwGAD651550–1856 and JwIL-4 or JwHEL and JwIL-4, but not JwGAD65 and JwIL-4, continued to develop overt diabetes. These results suggest that protection is dependent on the fragment of GAD65 encoded by pDNA.

Table II.

Coadministration of JwGAD651550–1856 and JwIL-4 does not protect NOD mice

TreatmentaDiabetes at 30 wk of Age
Expt. 1  
JwHEL+ JwIL-4 7 /7 
JwGAD651550–1856+ JwIL-4 7 /7 
Expt. 2  
JwHEL+ JwIL-4 4 /5 
JwGAD651550–1856+ JwIL-4 5 /5 
JwGAD65+ JwIL-4 0 /5 
TreatmentaDiabetes at 30 wk of Age
Expt. 1  
JwHEL+ JwIL-4 7 /7 
JwGAD651550–1856+ JwIL-4 7 /7 
Expt. 2  
JwHEL+ JwIL-4 4 /5 
JwGAD651550–1856+ JwIL-4 5 /5 
JwGAD65+ JwIL-4 0 /5 
a

NOD female mice 4 wk of age received three i.m. injections of 50 μg of pDNA in each quadricep over 21 days and were monitored for diabetes up to 30 wk of age.

We next determined whether pDNA vaccination could mediate protection in NOD mice with a greater degree of β cell autoimmunity. Groups of 12-wk-old NOD female mice, which are euglycemic yet exhibit maximal anti-β cell-specific T cell and Ab reactivity and significant insulitis (5), received four i.m. injections of 50 μg ofJwGAD65 + IL-4, JwGAD65, or JwHEL + IL-4 in each quadricep over 28 days. After monitoring animals up to 58 wk of age, the majority of unimmunized NOD mice (70%) or animals immunized with JwHEL + IL-4 (75%) or JwGAD65 (83%) had developed diabetes (Fig. 2,B). In contrast, only 17% of NOD mice immunized with JwGAD65 + IL-4 developed diabetes (p = 0.03 vs unimmunized, p = 0.01 vs JwHEL + IL-4, χ2 test) (Fig. 2,B). Similar results were obtained when 12-wk-old NOD female mice were coimmunized with JwGAD65 and JwIL-4 (data not shown). Histological analysis of pancreata from nondiabetic NOD mice demonstrated that a significant percentage of islets found in JwGAD65 + IL-4-immunized mice remained free of insulitis relative to unimmunized mice or animals immunized with JwHEL + IL-4 or JwGAD65 (Table III). The extent of insulitis observed in the nondiabetic animals treated with JwGAD65 + IL-4 is typical of that detected in untreated 12-wk-old NOD mice (23).

Table III.

Frequency of insulitis in nondiabetic NOD mice immunized at 12 wk of agea

TreatmentNo InfiltrationPeri-insulitisIntrainsulitis
<50%>50%
Unimmunized (n = 3) 0 (0%) 0 (0%) 31 (25.8%) 89 (74.2%) 
JwHEL+ IL-4 (n = 3) 0 (0%) 0 (0%) 34 (32.4%) 71 (67.6%) 
JwGAD65 (n = 2) 0 (0%) 0 (0%) 23 (27%) 62 (73%) 
JwGAD65+ IL-4 (n = 10) 89 (22.8%)b 59 (15.1%) 140 (35.9%) 102 (26.2%)b 
TreatmentNo InfiltrationPeri-insulitisIntrainsulitis
<50%>50%
Unimmunized (n = 3) 0 (0%) 0 (0%) 31 (25.8%) 89 (74.2%) 
JwHEL+ IL-4 (n = 3) 0 (0%) 0 (0%) 34 (32.4%) 71 (67.6%) 
JwGAD65 (n = 2) 0 (0%) 0 (0%) 23 (27%) 62 (73%) 
JwGAD65+ IL-4 (n = 10) 89 (22.8%)b 59 (15.1%) 140 (35.9%) 102 (26.2%)b 
a

Insulitis was determined by counting islets in nondiabetic mice 58 wk of age.

b

, p < 10−3 vs unimmunized, JwHEL + IL-4, or JwGAD65-treated mice as determined by χ2 analysis.

Since prevention of diabetes in NOD mice treated at both 4 and 12 wk of age with JwGAD65 was dependent on coadministration of pDNA-encoding IL-4, this suggested that protection was mediated by GAD65-specific Th2 cells. To test this notion, splenocyte cultures were prepared from nondiabetic or diabetic NOD mice immunized at either 4 and 12 wk of age, and cytokine levels were measured in response to GAD65 and two other candidate β cell autoantigens HSP60 and CPH (Figs. 3 and 4). Cultures established from NOD mice coimmunized with JwHEL and JwIL-4 at 4 wk of age or left untreated contained similar levels of IFN-γ and no detectable IL-4 and IL-5 above background in response to the panel of β cell autoantigens (Fig. 3). Analogous results were obtained with cultures established from NOD mice that were immunized at 12 wk of age with JwHEL + IL-4 or left untreated (Fig. 4). Using RT-PCR, expression of IL-4 but not TGF-β mRNA was detected in anti-CD3 and anti-CD28 Ab-stimulated CD4+ T cells purified from the spleen or pancreatic lymph nodes of NOD female mice coimmunized at 4 wk of age with JwGAD65 and JwIL-4 (data not shown).

FIGURE 3.

Cytokine secretion of T cells from NOD mice immunized at 4 wk of age with pDNA vaccines. Splenocyte cultures were prepared from 52-wk-old nondiabetic mice (A–C) immunized at 4 wk of age with JwHEL + JwIL-4 (n = 3), JwGAD65 (n = 4), and JwGAD65 + JwIL-4 (n = 10) or untreated mice (n = 3). In addition, splenocyte cultures were prepared from diabetic NOD mice (D–F) immunized with JwHEL + JwIL-4 (n = 11), JwGAD65 (n = 10), and JwGAD65 + JwIL-4 (n = 4) or untreated mice (n = 9). Splenocytes were prepared from individual mice and cultured with 20 μg/ml GAD65, HSP60, or CPH. IFN-γ (A and D), IL-4 (B and E), and IL-5 (C and F) in the culture supernatants were measured in duplicate via ELISA. The SE for each duplicate was <10%. The results represent the average of individual mice within a given group. ∗, p < 0.001 vs untreated NOD mice as determined by Student’s t test.

FIGURE 3.

Cytokine secretion of T cells from NOD mice immunized at 4 wk of age with pDNA vaccines. Splenocyte cultures were prepared from 52-wk-old nondiabetic mice (A–C) immunized at 4 wk of age with JwHEL + JwIL-4 (n = 3), JwGAD65 (n = 4), and JwGAD65 + JwIL-4 (n = 10) or untreated mice (n = 3). In addition, splenocyte cultures were prepared from diabetic NOD mice (D–F) immunized with JwHEL + JwIL-4 (n = 11), JwGAD65 (n = 10), and JwGAD65 + JwIL-4 (n = 4) or untreated mice (n = 9). Splenocytes were prepared from individual mice and cultured with 20 μg/ml GAD65, HSP60, or CPH. IFN-γ (A and D), IL-4 (B and E), and IL-5 (C and F) in the culture supernatants were measured in duplicate via ELISA. The SE for each duplicate was <10%. The results represent the average of individual mice within a given group. ∗, p < 0.001 vs untreated NOD mice as determined by Student’s t test.

Close modal
FIGURE 4.

Cytokine secretion of T cells from NOD mice immunized at 12 wk of age with pDNA vaccines. Splenocyte cultures were prepared from 58-wk-old nondiabetic mice (A–C) immunized at 12 wk of age with JwHEL + JwIL-4 (n = 3), JwGAD65 (n = 2), and JwGAD65 + JwIL-4 (n = 10) or untreated mice (n = 3). In addition, splenocyte cultures were prepared from diabetic NOD mice (D–F) immunized with JwHEL + JwIL-4 (n = 9), JwGAD65 (n = 10), and JwGAD65 + JwIL-4 (n = 2) or untreated mice (n = 7). Splenocytes were prepared from individual mice and cultured with 20 μg/ml GAD65, HSP60, or CPH. IFN-γ (A and D), IL-4 (B and E), and IL-5 (C and F) in the culture supernatants were measured in duplicate via ELISA. The SE for each duplicate was <10%. The results represent the average of individual mice within a given group. ∗, p < 0.001 vs untreated NOD mice as determined by Student’s t test.

FIGURE 4.

Cytokine secretion of T cells from NOD mice immunized at 12 wk of age with pDNA vaccines. Splenocyte cultures were prepared from 58-wk-old nondiabetic mice (A–C) immunized at 12 wk of age with JwHEL + JwIL-4 (n = 3), JwGAD65 (n = 2), and JwGAD65 + JwIL-4 (n = 10) or untreated mice (n = 3). In addition, splenocyte cultures were prepared from diabetic NOD mice (D–F) immunized with JwHEL + JwIL-4 (n = 9), JwGAD65 (n = 10), and JwGAD65 + JwIL-4 (n = 2) or untreated mice (n = 7). Splenocytes were prepared from individual mice and cultured with 20 μg/ml GAD65, HSP60, or CPH. IFN-γ (A and D), IL-4 (B and E), and IL-5 (C and F) in the culture supernatants were measured in duplicate via ELISA. The SE for each duplicate was <10%. The results represent the average of individual mice within a given group. ∗, p < 0.001 vs untreated NOD mice as determined by Student’s t test.

Close modal

In comparison, cultures prepared from nondiabetic NOD mice immunized at either 4 or 12 wk of age with the GAD65-IgGFc- and IL-4-encoding pDNAs exhibited significantly reduced levels of IFN-γ and concomitant increased levels of IL-4 and IL-5 in response to GAD65 (Figs. 3 and 4). A similar cytokine profile was also detected in response to HSP60 and CPH, suggesting that intermolecular determinant spreading of the Th2 cell response had occurred (Figs. 3 and 4). In contrast, a Th1 cell cytokine profile in response to the panel of β cell autoantigens was detected in cultures established from the few mice that did develop IDDM following immunization with either JwGAD65 and JwIL-4 (Fig. 3, D–F) or JwGAD65 + IL-4 (Fig. 4, D–F). This indicates that pDNA immunization was ineffective in these animals and may explain why β cell autoimmunity progressed to overt diabetes. Of note, cultures established from nondiabetic or diabetic NOD mice immunized with JwGAD65 only at either 4 or 12 wk of age contained no detectable IL-4 or IL-5 above background, but contained significantly elevated levels of IFN-γ in response to GAD65 relative to untreated mice (Figs. 3 and 4).

To directly determine whether regulatory CD4+ Th2 cells mediated protection, mixtures of diabetogenic spleen cells and CD4+ T cells purified from nondiabetic mice immunized with pDNA at 4 or 12 wk of age were adoptively transferred into NOD neonate recipients. As demonstrated in Table IV, CD4+ T cells prepared from GAD65-IgGFc- and IL-4-vaccinated NOD mice effectively suppressed the adoptive transfer of diabetes to recipients. However, mice receiving cell suspensions containing CD4+ T cells prepared from donor animals left untreated or immunized with HEL-IgGFc and IL-4 or GAD65-IgGFc only developed overt diabetes (Table IV).

Table IV.

CD4+ T cells from NOD mice immunized with pDNAs encoding GAD65-IgGFc and IL-4 inhibit the adoptive transfer of diabetes

CD4+ T CellsDiabetes at 10 wk of Age
NOD mice immunized at 4 wk of age  
Unimmunized 9 /10 
JwHEL+ JwIL-4 10 /10 
JwGAD65+ JwIL-4 0 /10 (p < 10−3
NOD mice immunized at 12 wk of age  
Unimmunized 9 /10 
JwHEL+ IL-4 9 /10 
JwGAD65+ IL-4 3 /10 (p = 0.022) 
CD4+ T CellsDiabetes at 10 wk of Age
NOD mice immunized at 4 wk of age  
Unimmunized 9 /10 
JwHEL+ JwIL-4 10 /10 
JwGAD65+ JwIL-4 0 /10 (p < 10−3
NOD mice immunized at 12 wk of age  
Unimmunized 9 /10 
JwHEL+ IL-4 9 /10 
JwGAD65+ IL-4 3 /10 (p = 0.022) 
a

Neonatal recipient mice (24–48 h old) received diabetogenic spleen cells (5 × 106) plus CD4+ T cells (106) prepared from NOD mice immunized with the respective pDNAs at either 4 or 12 wk of age. χ2 was used for statistical analysis.

Recent studies have demonstrated that pancreatic lymph nodes are key sites for activation of β cell-specific Th cells (35, 36). With this in mind, Th cell reactivity was assessed via ELISPOT in both the spleen and pancreatic lymph nodes of 12-wk-old NOD mice that had been immunized with the panel of pDNAs at 4 wk of age. The frequency of IFN-γ- and IL-4-secreting Th cells detected in splenocyte cultures (Fig. 5, A and C) reflected well the general cytokine profile determined above by ELISA in response to the panel of β cell autoantigens for the respective treatment groups (Figs. 3 and 4). Importantly, an increase and concomitant decrease of IL-4- and IFN-γ-secreting Th cells, respectively, were detected in response to GAD65 in the pancreatic lymph nodes of NOD mice immunized with JwGAD65 and JwIL-4 relative to untreated or JwHEL- and JwIL-4-treated mice (Fig. 5,D). Furthermore, the frequency of IFN-γ-secreting Th1 cells was increased in the pancreatic lymph nodes of NOD mice receiving JwGAD65 only when compared with untreated animals (Fig. 5 B).

FIGURE 5.

GAD65-specific Th cells are detected via ELISPOT in the pancreatic lymph nodes of NOD mice following GAD65-IgGFc pDNA vaccination. Splenocyte (A and C) and pancreatic lymph node (B and D) cultures were prepared from 12-wk-old NOD mice immunized three times beginning at 4 wk of age with JwHEL + JwIL-4 (n = 5), JwGAD65 (n = 5), and JwGAD65 + JwIL-4 (n = 5) or were left untreated (n = 5). ELISPOT was used to determine the frequency of IFN-γ (A and B)- and IL-4 (C and D)-secreting Th cells in response to 20 μg/ml GAD65, HSP60, or CPH in triplicate. The results for the splenocyte cultures represent the average of individual mice, whereas pancreatic lymph nodes were pooled within a given treatment group. ∗, p < 0.001 vs untreated NOD mice as determined by Student’s t test.

FIGURE 5.

GAD65-specific Th cells are detected via ELISPOT in the pancreatic lymph nodes of NOD mice following GAD65-IgGFc pDNA vaccination. Splenocyte (A and C) and pancreatic lymph node (B and D) cultures were prepared from 12-wk-old NOD mice immunized three times beginning at 4 wk of age with JwHEL + JwIL-4 (n = 5), JwGAD65 (n = 5), and JwGAD65 + JwIL-4 (n = 5) or were left untreated (n = 5). ELISPOT was used to determine the frequency of IFN-γ (A and B)- and IL-4 (C and D)-secreting Th cells in response to 20 μg/ml GAD65, HSP60, or CPH in triplicate. The results for the splenocyte cultures represent the average of individual mice, whereas pancreatic lymph nodes were pooled within a given treatment group. ∗, p < 0.001 vs untreated NOD mice as determined by Student’s t test.

Close modal

Next, we examined whether exogenous IL-4 encoded by pDNA was sufficient to elicit GAD65-specific Th2 cell reactivity and mediate protection or whether endogenous IL-4 was also necessary. For this purpose, female NOD.IL4null mice were used which do not express endogenous IL-4 and consequently lack typical Th2 effector cells (23, 37, 38). Groups of 4-wk-old NOD.IL4null female mice were coimmunized with JwGAD65 and JwIL-4, or JwHEL and JwIL-4, and then monitored for overt diabetes up to 48 wk of age. As demonstrated in Fig. 6, the majority of NOD.IL4null mice coimmunized with either JwGAD65 and JwIL-4 (58%) or JwHEL and JwIL-4 (75%) or unimmunized animals (67%) developed overt diabetes. On the other hand, only 25% of wild-type NOD mice coimmunized with JwGAD65 and JwIL-4 developed diabetes (Fig. 6), and these nondiabetic animals exhibited a predominant Th2 cytokine profile in response to the panel of β cell autoantigens in vitro (Fig. 7, A–C). The lack of protection observed in the NOD.IL4null mice coimmunized with JwGAD65 and JwIL-4 correlated with a modest increase in IFN-γ relative to untreated or JwHEL- and JwIL-4-coimmunized mice and no detectable IL-4 above background in response to GAD65 (Fig. 7, A and B). Furthermore, a small but significant increase in IL-5 secretion in response to GAD65 was observed in these mice (Fig. 7,C). Unlike wild-type NOD mice, however, no obvious effect on HSP60- and CPH-specific Th1 cell reactivity was observed in cultures prepared from NOD.IL4null mice coimmunized with JwGAD65 and JwIL-4 (Fig. 7,A). A similar cytokine profile was detected in cultures prepared from NOD.IL4null mice 2 wk after the final injection of JwGAD65 and JwIL-4 (Fig. 7, D–F). Interestingly, the level of IL-4 and IL-5 secretion in response to the panel of β cell autoantigens was significantly increased (p < 0.001) in cultures prepared from wild-type NOD mice shortly after the final pDNA immunization vs wild-type NOD mice that had been followed long term (Fig. 7).

FIGURE 6.

NOD. IL4null mice are not protected from diabetes following immunization with pDNAs encoding GAD65-IgGFc and IL-4. Groups of 12 NOD. IL4null (IL4KO) or wild-type NOD mice were immunized at 4 wk of age with JwHEL + JwIL-4 and JwGAD65 + JwIL-4 or were left untreated, or JwGAD65 + JwIL-4 or were left untreated, respectively, and monitored for diabetes on a weekly basis.

FIGURE 6.

NOD. IL4null mice are not protected from diabetes following immunization with pDNAs encoding GAD65-IgGFc and IL-4. Groups of 12 NOD. IL4null (IL4KO) or wild-type NOD mice were immunized at 4 wk of age with JwHEL + JwIL-4 and JwGAD65 + JwIL-4 or were left untreated, or JwGAD65 + JwIL-4 or were left untreated, respectively, and monitored for diabetes on a weekly basis.

Close modal
FIGURE 7.

Cytokine secretion of T cells from nondiabetic NOD. IL4null or wild-type NOD mice immunized with pDNA at 4 wk of age. Splenocyte cultures were prepared from nondiabetic (A–C) 48-wk-old NOD.IL4null or wild-type NOD mice immunized at 4 wk of age with JwHEL + JwIL-4 (n = 3) and JwGAD65 + JwIL-4 (n = 5) or were left untreated (n = 4), or JwGAD65 + JwIL-4 (n = 9) or were left untreated (n = 3), respectively. Groups of five NOD. IL4null or wild-type NOD mice were similarly treated but splenocyte cultures were established 2 wk after the final immunization (D–F). Splenocytes were prepared from individual mice and cultured with 20 μg/ml GAD65, HSP60, or CPH. IFN-γ (A and D), IL-4 (B and E), and IL-5 (C and F) in the culture supernatants were measured in duplicate via ELISA. The SE for each duplicate was <10%. The results represent the average of individual mice within a given group. +, p < 0.001 vs untreated NOD.IL4null mice. ∗, p < 0.001 vs untreated wild-type NOD mice as determined by Student’s t test.

FIGURE 7.

Cytokine secretion of T cells from nondiabetic NOD. IL4null or wild-type NOD mice immunized with pDNA at 4 wk of age. Splenocyte cultures were prepared from nondiabetic (A–C) 48-wk-old NOD.IL4null or wild-type NOD mice immunized at 4 wk of age with JwHEL + JwIL-4 (n = 3) and JwGAD65 + JwIL-4 (n = 5) or were left untreated (n = 4), or JwGAD65 + JwIL-4 (n = 9) or were left untreated (n = 3), respectively. Groups of five NOD. IL4null or wild-type NOD mice were similarly treated but splenocyte cultures were established 2 wk after the final immunization (D–F). Splenocytes were prepared from individual mice and cultured with 20 μg/ml GAD65, HSP60, or CPH. IFN-γ (A and D), IL-4 (B and E), and IL-5 (C and F) in the culture supernatants were measured in duplicate via ELISA. The SE for each duplicate was <10%. The results represent the average of individual mice within a given group. +, p < 0.001 vs untreated NOD.IL4null mice. ∗, p < 0.001 vs untreated wild-type NOD mice as determined by Student’s t test.

Close modal

The current study demonstrates that pDNA vaccination is an effective strategy to elicit autoantigen-specific Th2 effector cells for the purpose of suppressing β cell-specific autoimmunity. Immunization withpDNAs encoding GAD65-IgGFc- and IL-4-induced protection in NOD mice at early and late preclinical stages of disease (Fig. 2). This protection correlated with 1) suppression of insulitis (Tables I and III), 2) Th2 cell cytokine secretion in response to the panel of β cell autoantigens in both the spleen and pancreatic lymph nodes ( Figs. 3–5 and 7), and 3) induction of CD4+ T cells which inhibited the adoptive transfer of diabetes by diabetogenic spleen cells (Table IV). Importantly, disease suppression was GAD65 specific, since NOD mice immunized with HEL-IgGFc- and IL-4-encoding pDNAs continued to develop diabetes (Fig. 1) and exhibited unaltered β cell-specific T cell reactivity (Figs. 2 and 3). Furthermore, effective immunoregulation of IDDM was dependent on the fragment of GAD65 coupled to IgGFc. Whereas protection was observed in young and older NOD mice immunized with pDNA encoding a fragment of GAD65 spanning nt 656-1070 (Fig. 2), animals continued to develop diabetes following immunization with JwGAD651550–1856 which encodes the carboxyl terminus of GAD65 (Table II). One possible explanation for the lack of immunotherapeutic efficacy of the JwGAD651550–1856 is that the p509–528 and p524–543 epitopes found in this region of the molecule are inefficiently processed and presented, thereby limiting induction of regulatory Th effector cells.

We and others have previously shown that established β cell autoimmunity can be suppressed by immunizing NOD mice with intact GAD65 or GAD65-specific peptides (21, 22, 23). However, to successfully induce GAD65-specific Th2 cells and inhibit disease progression, multiple injections with high doses of protein or peptide prepared in the appropriate adjuvant was required (22, 23). In marked contrast, long-term protection was achieved in this study by immunizing NOD mice over a short period of time with relatively low doses of pDNA prepared in saline. A number of factors may contribute to the immunotherapeutic efficacy we observed with pDNA vaccination. Intramuscular injection of pDNA has been shown to transfect not only APCs and myocytes at the site of injection, but also a variety of cell types and tissues due to the rapid release of free pDNA from the muscle into blood and lymph (39, 40). Consequently, systemic expression of GAD65 may contribute to T cell activation and subsequent Th2 cell differentiation at key peripheral sites. Indeed, GAD65-specific Th2 cell reactivity was detected in the pancreatic lymph nodes of NOD mice immunized with JwGAD65 and JwIL-4 (Fig. 5). Recently, the pancreatic lymph nodes have been shown to be sites of Ag-specific activation and proliferation for diabetogenic T cells (35, 36). Conversely, it is likely that the pancreatic lymph nodes also provide key sites for immunoregulation of the disease process. Experiments are underway to identify additional tissues in which T cells are activated and differentiate into regulatory effector cells following pDNA vaccination. Expression of a secreted GAD65 fusion protein (Fig. 1,B) would also be expected to readily promote CD4+ T cell reactivity. Typically, secreted proteins are processed and presented via the MHC class II pathway. Accordingly, we have found that immunizing 4-wk-old NOD mice with JwIL-4 and a pDNA encoding full-length intracellular GAD65 had only a minimal effect on spontaneous GAD65-specific CD4+ T cell reactivity and disease progression (data not shown). Currently, it is not certain whether the IgGFc portion directly contributes to the immunogenicity of the GAD65 fragment in addition to facilitating secretion of the fusion protein. However, it is clear that the IgGFc does not induce nonspecific protection since NOD mice immunized with pDNA encoding HEL-IgGFc continue to develop diabetes at the same time of onset and frequency as untreated NOD mice. Finally, continuous pDNA expression of GAD65 in vivo may effectively maintain protection once it has been established. Various studies for example have reported in vivo persistence of pDNA up to several months postimmunization (25). It is apparent, however, that once established GAD65- and β cell-specific Th2 cell reactivity does diminish with time. Typically, levels of IL-4 and IL-5 secretion in response to the panel of β cell autoantigens is 2- to 3-fold greater in cultures established shortly after pDNA vaccination vs cultures established from nondiabetic NOD mice followed up for several months posttreatment (Fig. 7). Nevertheless, the Th2 cell reactivity which exists in these older animals was still sufficient to limit progression of insulitis (Table I) and to regulate β cell-specific Th1 cell responses (Fig. 3, A–C).

An important issue associated with Ag-based immunotherapies is the possibility of exacerbating autoimmunity following treatment. Indeed, we found that immunizing 4- or 12-wk-old NOD mice with JwGAD65 alone enhanced GAD65-specific Th1 cell reactivity (Figs. 3, A and D, and 4, A and D). Of importance is that spontaneous GAD65-specific Th1 cell reactivity is first detected in 4-wk-old NOD mice and reaches maximal levels in animals 12 wk of age (5, 6). Two recent experimental autoimmune encephalomyelitis studies have demonstrated that T cells are tolerized via anergy/deletion mechanisms, and autoimmunity is prevented by immunizing rodents with pDNAs encoding peptides derived from the autoantigens used to induce disease (29, 30). In these studies, however, the animals lacked primed autoantigen-specific T cells at the time of pDNA immunization. Our results indicate that once Th1 effector cells have been established, pDNAs encoding both autoantigen and IL-4 are required to 1) minimize expansion of those Th1 effector cells and 2) to promote Th2 cell differentiation ( Figs. 3–5 and 7). Indeed, both unmethylated CpG motifs found in pDNA (26, 27) and saline pDNA immunization tend to preferentially promote Th1 cell differentiation (41). Our results are also consistent with a growing notion that the efficacy at which Th2 cell differentiation can be induced in an Ag-specific manner is dependent on the frequency of established Th1 effector cells and uncommitted T cell precursors. For example, Coon et al. (42) showed that immunization with pDNA-encoding porcine insulin B chain elicited Th2 cell reactivity and prevented diabetes induced by lymphocytic choriomeningitis virus (LCMV) infection in BALB/c mice expressing LCMV nucleoprotein in β cells (42). In contrast, when established T cell effectors already existed, i.m. immunization of pDNA-encoding LCMV nucleoprotein was ineffective. Whether LCMV nucleoprotein-specific Th2 cells could be induced by coimmunizing with pDNA-encoding IL-4 was not examined. Interestingly, we have preliminary data indicating that the majority NOD mice immunized at 4 wk of age with pDNA encoding a murine insulin B chain-IgGFc fusion protein continue to develop overt diabetes (R. Tisch, unpublished data).

Coadministration of pDNAs encoding IL-4 was necessary but not sufficient to effectively induce GAD65-specific Th2 cells and prevent diabetes. Therefore, an endogenous source of IL-4 is also required to mediate protection. This was clearly evident in NOD.IL4null mice, which after coimmunization with JwGAD65 and JwIL4, continued to develop diabetes (Fig. 6). However, a modest yet significant increase in IL-5 secretion was detected in response to GAD65 in cultures prepared from NOD.IL4null mice immunized with JwGAD65 and JwIL-4 (Fig. 7, C and F), suggesting that Th2-like cell differentiation had occurred. Interestingly, continuous i.v. administration of IL-4 has been reported to potentiate Th2 cell development and prevent IDDM in NOD mice (17, 18). In our study, the three or four injections of pDNA may not have been sufficient to achieve systemic levels of IL-4 required to mediate a general suppressive effect. A recent study, for example, demonstrated that diabetes could be prevented in NOD mice by weekly i.m. injections of a pDNA encoding TGF-β (43). Presumably, continuous administration of the pDNA was necessary to achieve and maintain appropriate levels of systemic TGF-β to prevent disease. Based on our findings, we speculate that pDNA-encoded IL-4 is required to initiate GAD65-specific Th2 cell differentiation in the presence of GAD65-IgGFc. Once the appropriate extracellular cytokine milieu has been established locally, other β cell-specific Th2 cells develop which sustain and amplify the regulatory effect via bystander suppression of the GAD65-specific Th2 cells. Persistent expression of GAD65-IgGFc and IL-4 may further maintain GAD65-specific Th2 cell reactivity. The markedly reduced frequency of insulitis in diabetes-free NOD mice treated at either 4 or 12 wk of age suggests that protection is mediated predominantly in the periphery. Interestingly, the frequency of insulitis detected in nondiabetic 58-wk-old NOD mice treated at 12 wk of age is similar to that typically seen in unmanipulated 12-wk-old NOD mice (23), suggesting that at the time of treatment those islets free of insulitis were protected from subsequent infiltration. Furthermore, this protection is Ag specific since immunization with pDNAs encoding HEL-IgGFc and IL-4 had no significant effect on β cell-specific T cell reactivity or disease progression.

Finally, our results suggest that under the appropriate conditions, vaccination with pDNAs encoding IL-4 and GAD65 may prove to be an effective and readily applicable approach to inhibit progression to overt IDDM in individuals at the late preclinical stages of disease development. Employing pDNAs encoding other cytokines such as IL-10 and IL-13 known to promote Th2 cell differentiation or additional β cell autoantigens may further enhance the immunotherapeutic potential of this strategy in a clinical setting.

We thank James Preston for excellent technical assistance.

1

This work was supported by National Institutes of Health Grant 5P01 AI41580.

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; HEL, hen egg lysozyme; pDNA, plasmid DNA; LCMV, lymphocytic choriomeningitis virus.

1
Bach, J. F..
1994
. Insulin-dependent diabetes mellitus as an autoimmune disease.
Endocr. Rev.
15
:
516
2
Castano, L., G. S. Eisenbarth.
1990
. Type-1 diabetes: a chronic autoimmune disease of human, mouse, and rat.
Annu. Rev. Immunol.
8
:
647
3
Tisch, R., H. O. McDevitt.
1996
. Insulin dependent diabetes mellitus.
Cell
85
:
291
4
Delovitch, T. L., B. Singh.
1997
. The nonobese diabetic mouse as a model of autoimmune diabetes: immune dysregulation gets the NOD.
Immunity
7
:
727
5
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
6
Kaufman, D. L., M. Clare-Salzler, J. Tian, T. Forsthuber, G. S. Ting, P. Robinson, M. A. Atkinson, E. E. Sercarz, A. J. Tobin, P. V. Lehmann.
1993
. Spontaneous loss of T-cell tolerance to glutamic acid decarboxylase in murine insulin-dependent diabetes.
Nature
366
:
69
7
Yoon, J. W., C. S. Yoon, H. W. Lim, Q. Q. Huang, Y. Kang, K. H. Pyun, K. Hirasawa, R. S. Sherwin, H. S. Jun.
1999
. Control of autoimmune diabetes in NOD mice by GAD expression or suppression in beta cells.
Science
284
:
1135
8
Geng, L., M. Solimena, R. A. Flavell, R. S. Sherwin, A. C. Hayday.
1998
. Widespread expression of an autoantigen-GAD65 transgene does not tolerize non-obese diabetic mice and can exacerbate disease.
Proc. Natl. Acad. Sci. USA
95
:
10055
9
Atkinson, M. A., M. A. Bowman, L. Campbell, B. L. Harrow, D. L. Kaufman, N. K. Maclaren.
1994
. Cellular immunity to a determinant common to glutamate decarboxylase and Coxsackie virus in insulin-dependent diabetes.
J. Clin. Invest.
94
:
2125
10
Baekkeskov, S., H. J. Aanstoot, S. Christgau, A. Reetz, M. Solimena, M. Cascalho, F. Foli, W. Richter-Olsen, P. de Camilli.
1990
. Identification of the 64K autoantigen in insulin-dependent diabetes as the GABA-synthesizing enzyme glutamic acid decarboxylase.
Nature
347
:
151
11
Daniel, D., R. G. Gill, N. Schloot, D. Wegmann.
1995
. Epitope specificity, cytokine production profile and diabetogenic activity of insulin-specific T cell clones isolated from NOD mice.
Eur. J. Immunol.
25
:
1056
12
Rabinovitch, A..
1994
. Immunoregulatory and cytokine imbalances in the pathogenesis of IDDM: therapeutic intervention by immunostimulation?.
Diabetes
43
:
613
13
O’Garra, A..
1998
. Cytokines induce the development of functionally heterogeneous T helper cell subsets.
Immunity
8
:
275
14
Constant, S. L., K. Bottomly.
1997
. Induction of Th1 and Th2 CD4+ T cell responses: the alternative approaches.
Annu. Rev. Immunol.
15
:
297
15
Berman, M. A., C. I. Sandborg, Z. Wang, K. L. Imfeld, F. Zaldivar, V. Dadufalza, B. A. Buckingham.
1996
. Decreased IL-4 production in new onset type I insulin-dependent diabetes mellitus.
J. Immunol.
157
:
4690
16
Harrison, L. C., M. C. Honeyman, H. J. de Aizpurua, R. S. Schmidli, P. G. Colman, B. D. Tait, D. S. Cram.
1993
. Inverse relation between humoral and cellular immunity to glutamic acid decarboxylase in subjects at risk of insulin-dependent diabetes.
Lancet
341
:
1365
17
Rapoport, M. J., A. Jaramillo, D. Zipris, A. H. Lazarus, D. V. Serreze, E. H. Leiter, P. Cyopick, J. S. Danska, T. L. Delovitch.
1993
. Interleukin 4 reverses T cell proliferative unresponsiveness and prevents the onset of diabetes in nonobese diabetic mice.
J. Exp. Med.
178
:
87
18
Cameron, M. J., G. A. Arreaza, P. Zucker, S. W. Chensue, R. M. Streiter, S. Chakrabarti, 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
19
Muir, A., A. Peck, M. A., Y. H. Clare-Salzler, J. Song, R. Cornelius, J. Luchetta, J. Krischer, N. K. Maclaren.
1995
. Insulin immunization of nonobese diabetic mice induces a protective insulitis characterized by diminished intraislet interferon-γ transcription.
J. Clin. Invest.
95
:
628
20
Tisch, R., H. O. McDevitt.
1994
. Antigen-specific immunotherapy: is it a real possibility to combat T-cell-mediated autoimmunity?.
Proc. Natl. Acad. Sci. USA
91
:
437
21
Tian, J., M. Clare-Salzler, A. Herschenfeld, B. Middleton, D. Newman, R. Mueller, S. Arita, C. Evans, M. A. Atkinson, Y. Mullen, et al
1996
. Modulating autoimmune responses to GAD inhibits disease progression and prolongs islet graft survival in diabetes-prone mice.
Nat. Med.
2
:
1348
22
Tisch, R., R. S. Liblau, X. D. Yang, P. Liblau, H. O. McDevitt.
1998
. Induction of GAD65-specific regulatory T-cells inhibits ongoing autoimmune diabetes in nonobese diabetic mice.
Diabetes
47
:
894
23
Tisch, R., B. Wang, D. V. Serreze.
1999
. Induction of glutamic acid decarboxylase 65-specific Th2 cells and suppression of autoimmune diabetes at late stages of disease is epitope dependent.
J. Immunol.
163
:
1178
24
Donnelly, J. J., J. B. Ulmer, J. W. Shiver, M. A. Liu.
1997
. DNA vaccines.
Annu. Rev. Immunol.
15
:
617
25
Wolff, J. A., J. J. Ludtke, G. Acsadi, P. Williams, J. Agnes.
1992
. Long-term persistence of plasmid DNA and foreign gene expression in mouse muscle.
Hum. Mol. Genet.
1
:
363
26
Chu, R. S., O. S. Targoni, A. M. Krieg, P. V. Lehmann, C. V. Harding.
1997
. CpG oligodeoxynucleotides act as adjuvants that switch on T helper 1 (Th1) immunity.
J. Exp. Med.
186
:
1623
27
Roman, M., E. Martin-Orozco, J. S. Goodman, M. D. Nguyen, Y. Sato, A. Ronaghy, R. S. Kornbluth, D. D. Richmann, D. A. Carson, E. Raz.
1997
. Immunostimulatory DNA sequences function as T helper-1-promoting adjuvants.
Nat. Med.
3
:
849
28
Waisman, A., P. J. Ruiz, D. L. Hirschberg, A. Gelman, J. R. Oksenberg, S. Brocke, F. Mor, I. R. Cohen, L. Steinman.
1996
. Suppressive vaccination with DNA encoding a variable region gene of the T-cell receptor prevents autoimmune encephalomyelitis and activates Th2 immunity.
Nat. Med.
2
:
899
29
Ruiz, P. J., H. Garren, I. U. Ruiz, D. L. Hirschberg, L. V. T. Nguyen, M. V. Karpuj, M. T. Cooper, D. J. Mitchell, C. G. Fathman, L. Steinman.
1999
. Suppressive immunization with DNA encoding a self-peptide prevents autoimmune disease: modulation of T cell costimulation.
J. Immunol.
162
:
3336
30
Lobell, A., R. Weissert, M. K. Storch, C. Svanholm, K. L. De Graaf, H. , R. Lassmann, T. Andersson, T. Olsson, H. Wigzell, H.
1998
. Vaccination with DNA encoding an immunodominant myelin basic protein peptide targeted to Fc of immunoglobulin G suppresses experimental autoimmune encephalomyelitis.
J. Exp. Med.
187
:
1543
31
Yasutomi, Y., H. L. Robinson, S. Lu, F. Mustafa, C. Lekutis, J. Arthos, J. I. Mullens, G. Voss, K. Manson, M. Wyand, N. L. Letvin.
1996
. Simian immunodeficiency virus-specific cytotoxic T-lymphocyte induction through DNA vaccination of rhesus monkeys.
J. Virol.
70
:
678
32
Wang, B., A. Gonzalez, C. Benoist, D. Mathis.
1996
. The role of CD8+ T cells in initiation of insulin dependent diabetes mellitus.
Eur. J. Immunol.
26
:
1762
33
Chao, C. C., H. O. McDevitt.
1997
. Identification of immunogenic epitopes of GAD 65 presented by Ag7 in non-obese diabetic mice.
Immunogenetics
46
:
29
34
Zechel, M. A., J. F. Elliott, M. A. Atkinson, B. Singh.
1998
. Characterization of novel T-cell epitopes on 65 kDa and 67 kDa glutamic acid decarboxylase relevant in autoimmune responses in NOD mice.
J. Autoimmun.
11
:
83
35
Hoglund, P., J. Mintern, C. Waltzinger, W. Heath, C. Benoist, D. Mathis.
1999
. Initiation of autoimmune diabetes by developmentally regulated presentation of islet cell antigens in the pancreatic lymph nodes.
J. Exp. Med.
189
:
331
36
Morgan, D. J., C. Kurts, H. T. C. Kreuwel, K. L. Holst, W. R. Heath, L. A. Sherman.
1999
. Ontogeny of T cell tolerance to peripherally expressed antigens.
Proc. Natl. Acad. Sci. USA
96
:
3854
37
Kuhn, R., K. Rajewsky, W. Muller.
1991
. Generation and analysis of interleukin-4 deficient mice.
Science
254
:
707
38
Wang, B., A. Gonzalez, P. Hoglund, J. D. Katz, C. Benoist, D. Mathis.
1998
. Interleukin-4 deficiency does not exacerbate disease in NOD mice.
Diabetes
47
:
1207
39
Winegar, R. A., J. A. Monforte, K. D. Suing, K. G. O’Loughlin, C. J. Rudd, J. T. MacGregor.
1996
. Determination of tissue distribution of an intramuscular plasmid vaccine using PCR and in situ DNA hybridization.
Hum. Gene Ther.
7
:
2185
40
Torres, C. A. T., A. Iwasaki, B. H. Barber, H. L. Robinson.
1997
. Differential dependence on target site tissue for gene gun and intramuscular DNA immunizations.
J. Immunol.
158
:
4529
41
Feltquate, D. M., S. Heaney, R. G. Webster, H. L. Robinson.
1997
. Different T helper cell types and antibody isotypes generated by saline and gene gun DNA immunization.
J. Immunol.
158
:
2278
42
Coon, B., L. L. An, J. L. Whitton, M. G. von Herrath.
1999
. DNA immunization to prevent autoimmune diabetes.
J. Clin. Invest.
104
:
189
43
Piccirillo, C. A., Y. Chang, G. J. Prud’homme.
1998
. TGF-β1 somatic gene therapy prevents autoimmune disease in nonobese diabetic mice.
J. Immunol.
161
:
3950