Recurrent autoimmune destruction of the insulin-producing β cells is a key factor limiting successful islet graft transplantation in type I diabetic patients. In this study, we investigated the feasibility of using an Ag-specific plasmid DNA (pDNA)-based strategy to protect pro-islets that had developed from a neonatal pancreas implanted under the kidney capsule of nonobese diabetic (NOD) mice. NOD recipient mice immunized with pDNA encoding a glutamic acid decarboxylase 65 (GAD65)-IgFc fusion protein (JwGAD65), IL-4 (JwIL4), and IL-10 (pIL10) exhibited an increased number of intact pro-islets expressing high levels of insulin 15 wk posttransplant, relative to NOD recipient mice immunized with pDNA encoding a hen egg lysozyme (HEL)-IgFc fusion protein (JwHEL)+JwIL4 and pIL10 or left untreated. Notably, the majority of grafted pro-islets detected in JwGAD65+JwIL4- plus pIL10-treated recipients was free of insulitis. In addition, administration of JwGAD65+JwIL4+pIL10 provided optimal protection for engrafted islets compared with recipient NOD mice treated with JwGAD65+JwIL4 or JwGAD65+pIL10, despite effective protection of endogenous islets mediated by the respective pDNA treatments. Efficient protection of pro-islet grafts correlated with a marked reduction in GAD65-specific IFN-γ reactivity and an increase in IL-10-secreting T cells. These results demonstrate that pDNA vaccination can be an effective strategy to mediate long-term protection of pro-islet grafts in an Ag-specific manner and that conditions are more stringent to suppress autoimmune destruction of grafted vs endogenous islets.
Type I diabetes (T1D)5 is an autoimmune disorder characterized by destruction of the insulin-producing β cells found in the pancreatic islets of Langerhans (1, 2, 3). Autoimmunity is manifested by a chronic inflammatory response involving islet infiltration (insulitis) by lymphocytes and monocytes. In the nonobese diabetic (NOD) mouse, a spontaneous model for T1D, insulitis is initially detected in the peri-islet regions (peri-insulitis) at ∼3–5 wk of age and is followed by progressive infiltration of the islets (intrainsulitis). Clinical onset of T1D is first detected at ∼12 wk of age once a sufficient number of β cells have been destroyed, and by 30 wk the majority of NOD female mice are diabetic.
The primary mediators of β cell destruction are CD4+ and CD8+ T cells (4, 5). Pathogenic CD4+ T cells typically exhibit a Th1 cell phenotype characterized by high levels of IFN-γ secretion. Furthermore, a number of β cell autoantigens including insulin, IA-2, and glutamic acid decarboxylase 65 (GAD65) have been identified as targets of CD4+ Th1 cells (6). A key role for GAD65 in human T1D is suggested by studies demonstrating a positive correlation between the detection of GAD65-specific autoantibodies in prediabetic individuals and the eventual progression toward overt diabetes (6, 7). In NOD mice, GAD65-specific T cell reactivity is detected as early as 4 wk of age, coinciding with the initial T cell response to an islet extract (8, 9, 10).
We and others have demonstrated that administration of β cell autoantigens or derived peptides can effectively prevent initiation of the disease process and suppress established β cell autoimmunity under certain conditions in NOD mice (8, 10, 11, 12, 13, 14). In many of these studies, protection is mediated through the induction of β cell-specific regulatory Th cells. Th2 and/or Th3 cells suppress Th1 cell differentiation through a bystander mechanism involving secretion of regulatory cytokines including IL-4, IL-10, and TGFβ1 (15, 16). These cytokines act either directly on naive Th cells or modulate APC function. The immunotherapeutic efficacy associated with a given autoantigen is thought to be partly due to the frequency of naive precursor Th cells which differentiate into regulatory Th2 and/or Th3 cells, in addition to the relative efficacy of a given treatment modality to establish the appropriate in vivo conditions promoting regulatory T cell differentiation (17, 18, 19).
A great deal of interest has focused on the use of plasmid DNAs (pDNAs) to elicit cellular and humoral immunity in the context of infectious diseases and cancer immunotherapy (20). The application of pDNA has a number of advantages compared with parenteral administration of intact Ag or peptide. Notably, combinations of pDNAs encoding Ag and different cytokines can be readily used to influence the nature and magnitude of the immune response. Recently, pDNAs have been applied for the prevention of autoimmunity in various models (21, 22, 23). Indeed, we and others have demonstrated that administration of pDNA encoding β cell autoantigens and/or appropriate regulatory cytokines to young NOD mice can prevent diabetes (19, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34). Importantly, overt diabetes can also be prevented in NOD mice at a late preclinical stage of T1D by coimmunization with pDNA encoding a GAD65-IgFc fusion protein and IL-4 (24). With this in mind, the focus of the current study was to assess the feasibility of using an Ag-specific pDNA-based vaccine strategy to mediate long-term protection for syngeneic islet grafts in NOD recipient mice.
Islet transplantation provides one approach for β cell replacement in diabetic individuals. However, because of allogeneic- and autoimmune-mediated destruction of β cells, successful islet engraftment has relied on continuous administration of immunosuppressive drugs to patients, which over time could lead to unwanted complications (35, 36). A preferred strategy is to selectively suppress islet immunity while maintaining normal immune function. Although overlap exists between the two processes, allogenic- and autoimmune-mediated islet destruction differ in terms of Ag recognition and effector mechanisms (37). Therefore, achieving effective long-term protection of transplanted islets may require strategies that individually target allogeneic and autoimmune reactivity. Presumably, strategies used to prevent diabetes can be applied in the latter case. However, whether Ag-specific immunotherapies used as a prophylactic can also be applied for long-term protection of an islet graft has yet to be established. For example, location and possible differences in the repertoire of effector T cells may necessitate conditions unique to transplanted vs endogenous islets. In this study, we demonstrate that administration of pDNA encoding GAD65-IgFc and regulatory cytokines can effectively protect syngeneic islets in NOD recipients but that conditions are more stringent than those required to suppress β cell autoimmunity in the endogenous pancreas.
Materials and Methods
NOD/Lt mice were housed and bred under specific pathogen-free conditions and fed NIH diet 31A (Purina, St. Louis, MO). Currently, T1D develops in ∼80% of female NOD/Lt mice by 1 year of age.
pDNA preparation and administration
Construction of pDNAs encoding GAD65-IgFc (JwGAD65), hen egg lysozyme (HEL)-IgFc (JwHEL), and murine IL-4 (JwIL4) have previously been described (24). The pNGVL3-IL10 construct (pIL10) encoding murine IL-10 was obtained from the National Gene Vector laboratory at the University of Michigan (Ann Arbor, MI). pDNA was prepared from DH5α Escherichia coli using a Qiagen endotoxin free kit (Qiagen, Chatsworth, CA) followed by phenol/chloroform extraction and ethanol precipitation. pDNA was resuspended at 1.0 mg/ml in sterile PBS. NOD/Lt female mice 10–11 wk of age received a total of four i.m. injections in each quadricep with 50 μg of a given pDNA in a final volume of 50 μl on a weekly basis.
Transplantation of syngeneic neonatal pancreata
One week following the final pDNA vaccination, mice received syngeneic neonatal pancreatic transplants. NOD mice 24 h of age were used as pancreas donors. Whole pancreata were removed and placed in sterile PBS on ice. Under general anesthesia via isoflurane inhalation (Halocarbon Laboratories, Riveredge, NJ), a small dorsal to ventral incision under the recipient’s rib cage was made, allowing the left kidney to be exposed. The donor pancreas was inserted under the kidney capsule through a small incision, after which the kidney was repositioned into the abdominal cavity. The peritoneal and muscular layers were closed using absorbing plain gut sutures (Ethicon, Summerville, NT) while sterile 9-mm wound clips (BD Biosciences, San Diego, CA) were used to close the epidermal incision. Recipient mice were monitored up to 30 wk of age (15 wk posttransplant), after which time they were sacrificed and all analyses were performed.
Assessment of diabetes and insulitis
Mice were monitored weekly for the development of urine glucose with Diastix (Ames, Elkhart, IN). Glycosuric values of >3 for two successive measurements were considered diagnostic of diabetes onset. Insulitis was assessed by histology. Pancreata were fixed in neutral-buffered Formalin, embedded in paraffin, sectioned, and stained with H&E. 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 were scored for each animal. The severity of insulitis was scored as no insulitis, peri-insulitis, or intrainsulitis less or more than 50%.
Characterization of function and insulitis of the neonatal pancreatic graft
Insulitis and insulin expression of the neonatal pancreatic grafts were determined by histology. Kidneys from individual mice were fixed in 4% paraformaldehyde and embedded in paraffin. The entire kidney was sectioned. H&E staining was used to determine the presence and degree of insulitis as described above. Insulin expression was assessed using an anti-insulin guinea pig polyclonal Ab (Zymed Laboratories, San Francisco, CA) and Ab binding was detected with a Vecta Stain Avidin-Biotin Complex kit (Vector Laboratories, Burlingame, CA). H&E and insulin staining were separately performed on serial sections 5 μm apart to correlate islet infiltration with β cell function. Alternatively, sections were stained with anti-insulin Ab and counterstained with hematoxylin. Two forms of islet morphology were observed in a given transplant. Intact grafted pro-islets were identified by typical islet architecture, while remnants of pro-islet destruction were identified by a few localized β cells lacking organized islet architecture. Total grafted pro-islets and pro-islet remnants were counted and scored based on insulin production and cellular infiltration.
ELISPOT was conducted as previously described (19) using Abs purchased from BD PharMingen (San Diego, CA). Briefly, ImmunoSpot M200 plates (Cellular Technology, Cleveland, OH) were coated overnight at 4°C with either 2 μg/ml purified rat anti-mouse IFN-γ, 2 μg/ml purified rat anti-mouse IL-4, or 2 μg/ml purified rat anti-mouse IL-10 prepared in PBS (0.1 ml/well). Plates were blocked with 1% BSA/PBS for a minimum of 1 h at room temperature and washed four times with sterile PBS. Spleen cells were prepared from individual mice as previously described (19). Briefly, spleen cell suspensions were prepared in ice-cold PBS and immediately centrifuged at 400 × g for 5 min at 4°C. RBCs were lysed and splenocytes were washed, resuspended at 5 × 106 cells/ml in HL-1 medium (BioWhittaker, Walkersville, MD), and then plated at 1 × 106/well (0.2 ml/well). Ag was added to triplicate wells at a final concentration of 5 μg/ml. The plates were then incubated for either 24 h (IFN-γ, IL-10) or 48 h (IL-4) at 37°C in 5.5% CO2. Wells were initially washed three times with PBS and then three times with 0.05% Tween 20/PBS. Biotinylated rat anti-mouse IFN-γ, biotinylated rat anti-mouse IL-4, or biotinylated rat anti-mouse IL-10 were added at 1 μg/ml prepared in 1% BSA/0.05% Tween 20/PBS (0.1 ml/well). After overnight incubation at 4°C, plates were washed three times with 0.05% Tween 20/PBS. Streptavidin-HRP (BD PharMingen) was added at 1/2000 dilution for 2 h at room temperature. This was followed by three washes with 0.05% Tween 20/PBS and three washes with PBS only. Development solution consisted of 0.8 ml of 20 mg/ml 3-amino-9-ethylcarbazole (Sigma-Aldrich, St. Louis, MO) dissolved in 2.0 ml of dimethyl formamide 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 of the developing solution was added per well. The reaction was stopped by rinsing the wells in ddH2O. An ImmunoSpot plate reader (Cellular Technology) was used to quantitate the number of spot-forming cells (SFC) per well.
The cloning and preparation of murine GAD65 have previously been described (8). Briefly, cDNA encoding murine GAD65 was engineered to encode six histidine residues at the C terminus of the protein. Recombinant GAD65 was expressed in a baculovirus expression system and purified using a Ni2+-conjugated resin (Qiagen). An additional purification step involved preparative SDS-PAGE, after which recombinant GAD65 was electroeluted and dialyzed extensively against PBS.
Pancreatic islet isolation
Pancreatic islets were isolated as previously described (9). Briefly, pancreases from NOD.scid mice were perfused with 1.75 mg/ml collagenase P (Roche, Indianapolis, IN) and digested for 20 min at 37°C. Islets were purified from digested tissues using a Ficoll gradient and then hand picked. Before transplantation, purified islets were washed three times in PBS and resuspended in a final volume of 20–30 μl of sterile PBS.
Immunization with pDNAs encoding GAD65-IgFc, IL-4, and IL-10 protect syngeneic islet grafts
The general aim of this study was to determine whether an Ag-specific pDNA strategy could effectively prevent autoimmune-mediated destruction of grafted syngeneic β cells. For this purpose, a neonatal pancreas transplant model was used. Within 2–3 wk of implanting a neonatal pancreas under the kidney capsule of a syngeneic nonautoimmune recipient, the exocrine tissue atrophies while endocrine cells grow and differentiate into pro-islets identical to those found in the endogenous pancreas of an adult mouse (38). pDNA vaccination entailed four i.m. injections of 50 μg of each pDNA into the quadriceps of 10-wk-old NOD female mice over 4 wk. At 10 wk of age, NOD female mice are typically euglycemic but exhibit maximal β cell-specific T cell reactivity (8). pDNAs encoding IgFc fusion proteins of either a fragment of murine GAD65 (JwGAD65) or HEL (JwHEL)+pDNAs encoding murine IL-4 (JwIL4) and IL-10 (pIL10) were used for vaccination. A third group of NOD mice was left untreated. One week after the final pDNA immunization, a neonatal NOD pancreas was implanted under the kidney capsule of individual mice. Recipient mice were then monitored for the development of overt diabetes up to 30 wk of age. Consistent with previous work demonstrating the immunotherapeutic efficacy of JwGAD65+JwIL4 administration (24), the majority of recipient mice treated with JwGAD65+ JwIL4+pIL10 (9 of 10) remained diabetes free (Fig. 1⇓). Analogous protection was observed in NOD mice vaccinated with JwGAD65+pIL10 (data not shown). In contrast, all of the JwHEL+JwIL4+JwIL10-immunized (6 of 6) and untreated (5 of 5) recipient mice developed diabetes (Fig. 1⇓). A significant delay in the onset of diabetes, however, was observed for JwHEL+JwIL4+JwIL10-immunized vs untreated recipient mice (p = 0.0013, Kaplan-Meier).
Next, the status of the neonatal pancreas transplants in the three experimental groups of recipient mice was examined. Engrafted pro-islets were assessed for architecture, insulin expression, and the frequency of insulitis. Typically, 25–40 pro-islets are detected in individual mice 2–3 wk postimplantation of the neonatal pancreas regardless of the treatment, although pro-islets in untreated recipients exhibit significant insulitis at this time (data not shown). Two general types of pro-islet morphology were observed in sections of the engrafted kidney. Intact pro-islets were identified by characteristic islet architecture, while remnants of pro-islet destruction were identified by a few localized β cells (Fig. 2⇓). Insulin expression was defined as high, low, or not detectable as determined by immunohistochemical staining (Fig. 2⇓). In recipient mice receiving no pDNA treatment, the graft site contained only a few remnants and no intact pro-islets (Fig. 3⇓A). The graft site of JwHEL+JwIL4+pIL10-immunized recipient mice consisted primarily of remnants with some intact pro-islets. This contrasted with the JwGAD65+JwIL4+pIL10-immunized group in which there was a significant increase in intact pro-islets (p < 10−4, χ2) and fewer remnants (p < 10−4, χ2) in individual recipients (Fig. 3⇓A). An average of ∼32 intact pro-islets per graft site was found in the JwGAD65+JwIL4+pIL10-immunized group of mice, suggesting that the majority of pro-islets derived from the transplanted endocrine tissue persisted.
Insulin expression was detected in all of the intact pro-islets of JwGAD65+JwIL4+pIL10-immunized mice with the majority (274 of 292) expressing high levels (Fig. 3⇑B). This differed from JwHEL+JwIL4+pIL10-treated recipient mice in which the majority of intact pro-islets expressed either low (18 of 48) or undetectable (15 of 48) levels of insulin (Fig. 3⇑B). Insulin expression was not observed in pro-islet remnants for the three experimental groups. Strikingly, almost all (287 of 292) of the intact pro-islets found in recipient mice treated with JwGAD65+JwIL4+pIL10 were free of insulitis, whereas only a few (7 of 48) of the intact pro-islets detected in JwHEL+JwIL4+pIL10-immunized animals were insulitis free (Fig. 3⇑C). These results demonstrate that an increased frequency of intact and functional pro-islets can be detected up to 15 wk postimplantation in NOD mice treated with JwGAD65+JwIL4+pIL10 relative to JwHEL+JwIL4+pIL10 or untreated recipient animals.
To ensure that JwGAD65+JwIL4+pIL10 treatment also mediated protection for isolated islets, NOD mice were immunized with either JwGAD65+JwIL4+pIL10 or JwHEL+JwIL4+pIL10 and then received islets prepared from NOD.scid donor mice. Similar to findings made with neonatal pancreas transplants, grafts consisting of isolated islets in JwGAD65+JwIL4+pIL10-treated recipients were completely free of insulitis (Fig. 2⇑D) and expressed high levels of insulin. In contrast, islet grafts were heavily infiltrated (Fig. 2⇑E) and lacked insulin expression in JwHEL+JwIL4+pIL10-treated recipients. This observation demonstrates that there is no difference in the immunotherapeutic efficacy of JwGAD65+JwIL4+pIL10 to mediate protection for pro- vs isolated islets.
Reduced IFN-γ and increased IL-4 secretion in response to GAD65 is detected in JwGAD65+JwIL4+pIL10-treated recipients
Earlier work demonstrated that JwGAD65+JwIL4 immunization suppressed β cell autoimmunity in NOD mice via induction of GAD65-specific Th2-like reactivity (19, 24). Accordingly, the frequency of IFN-γ- and IL-4-secreting CD4+ T cells in response to GAD65 was investigated in splenocyte cultures prepared from the three experimental groups of NOD recipients. Untreated NOD recipients exhibited a typical Th1-like response in which T cells secreting IFN-γ but not IL-4 were detected in response to GAD65 (Fig. 4⇓). In contrast to the untreated group, NOD recipient mice receiving JwGAD65+JwIL4+pIL10 or JwHEL+JwIL4+pIL10 exhibited a significantly reduced frequency of T cells secreting IFN-γ (p < 0.004, Student’s t test) and a concomitant increase in IL-4 secreting T cells (p < 0.004, Student’s t test) in response to GAD65 (Fig. 4⇓). Notably, cultures prepared from JwGAD65+ JwIL4+pIL10 treated recipients displayed a greater reduction in the IFN-γ response (p < 0.004, Student’s t test), and an enhanced frequency of IL-4-secreting T cells (p = 0.01, Student’s t test) relative to JwHEL+JwIL4+pIL10-immunized recipients (Fig. 4⇓). These findings indicate that protection of pro-islet grafts mediated by JwGAD65+JwIL4+pIL10 treatment correlated with a markedly reduced GAD65-specific Th1 cell response and an increased frequency of IL-4-secreting T cells.
Administration of pDNAs encoding GAD65-IgFc, IL-4, and IL-10 is required for efficient protection of islet grafts
Experiments were conducted to define the minimal requirements for islet graft protection associated with JwGAD65+JwIL4 plus pIL10 administration. In this study, comparisons were drawn regarding the efficacy of JwGAD65+JwIL4+pIL10, JwGAD65+JwIL4, and JwGAD65+pIL10 treatments to protect grafted pro-islets. Groups of 10-wk-old NOD female mice were left untreated or vaccinated weekly with the panel of pDNAs for 4 wk, after which NOD neonatal pancreata were implanted under the kidney capsule and the status of the grafted pro-islets was determined 15 wk posttransplant. In untreated NOD recipients, only a few intact pro-islets were detected which in turn lacked insulin expression (Fig. 5⇓A). Similar to results described above, a significant increase in the number of intact pro-islets was observed in individual recipient mice immunized with JwGAD65+JwIL4+pIL10 relative to untreated recipients (p = 0.006, χ2; Fig. 5⇓A). Furthermore, the majority of these intact pro-islets (70 of 81) expressed high levels of insulin (Fig. 5⇓B). Pro-islet graft protection mediated by JwGAD65+JwIL4 vaccination was noticeably less efficient relative to JwGAD65+JwIL4-plus pIL10-treated recipient NOD mice, characterized by a significantly reduced number of intact pro-islets (p = 0.022, χ2; Fig. 5⇓A). In contrast, a comparable number of intact pro-islets was detected in individual recipients treated with JwGAD65+pIL10 and JwGAD65+JwIL4+pIL10 (Fig. 5⇓A). However, a higher percentage of intact pro-islets lacked insulin expression (23.6% vs 0%; Fig. 5⇓B) and exhibited insulitis (Fig. 5⇓C) in JwGAD65+pIL10- vs JwGAD65+JwIL4+pIL10-treated recipient mice. Taken together, these results demonstrate that complete protection of pro-islet grafts required the administration of pDNAs encoding GAD65-IgFc, IL-4, and IL-10.
Previous work demonstrated that administration of JwGAD65+ JwIL4 to 10- to 12-wk-old NOD female mice effectively inhibited further progression of insulitis in the pancreas long term (24). With this in mind, the frequency of insulitis in the endogenous pancreas of the four groups of NOD recipients was also examined. Despite only marginal protection of pro-islet grafts, a significant increase in islets lacking infiltration was detected in the pancreas of JwGAD65+JwIL4-immunized recipients relative to the untreated group (p < 10−4, χ2; Table I⇓). Furthermore, the frequency of endogenous islets free of insulitis was similar for recipient mice receiving either JwGAD65+JwIL4 or JwGAD65+pIL10 (Table I⇓). Nevertheless, a significant increase in endogenous islets free of infiltration (p < 10−4, χ2) was observed in JwGAD65+JwIL4+pIL10 vs JwGAD65+JwIL4- or JwGAD65+pIL10-immunized mice (Table I⇓). These findings indicate that the three conditions of JwGAD65 treatment suppressed insulitis in the endogenous pancreas despite varying efficacies to protect pro-islet grafts.
To gain insight into why the different pDNA treatments varied in pro-islet graft protection, 10-wk-old NOD female were immunized with JwGAD65+JwIL4+pIL10, JwGAD65+JwIL4, or JwGAD65+pIL10 and the frequency of IFN-γ-, IL-4-, and IL-10-secreting T cells in response to GAD65 was compared at 21 wk of age. As controls, NOD mice received JwHEL+JwIL4+pIL10 or were left untreated. Consistent with results described above, a significant increase in IL-4-secreting T cells (p = 0.008, Student’s t test) and a notable decrease in the IFN-γ response (p < 10−4, Student’s t test) was detected in cultures prepared from JwHEL+JwIL4+pIL10-immunized mice vs untreated animals (Fig. 6⇓). Furthermore, there was a trend toward an increased frequency of IL-10-secreting T cells in JwHEL+JwIL4+pIL10 cultures, although this response was not significantly different from that seen in cultures prepared from untreated animals (p = 0.08, Student’s t test). In contrast, cultures established from JwGAD65+JwIL4+pIL10-vaccinated animals exhibited a significant increase in the frequency of IL-4 (p < 10−4, Student’s t test)- and IL-10 (p < 0.004, Student’s t test)-secreting T cells and a marked decrease in IFN-γ reactivity (p < 10−4, Student’s t test) relative to either untreated or JwHEL+JwIL4+pIL10-immunized mice (Fig. 6⇓). On the other hand, cultures prepared from JwGAD65+JwIL4-immunized mice exhibited a comparable IL-4 response, increased IFN-γ reactivity (p = 0.0006, Student’s t test), and a markedly reduced frequency of IL-10-secreting T cells (p < 10−4, Student’s t test) relative to JwGAD65+JwIL4+pIL10-immunized mice (Fig. 6⇓). Interestingly, a similar cytokine profile in response to GAD65 was detected in cultures prepared from JwGAD65+JwIL4 plus IL10- and JwGAD65+pIL10-vaccinated animals (Fig. 6⇓). These results indicate that the degree of pro-islet graft protection associated with the different combinations of JwGAD65, JwIL4, and pIL10 correlated more closely with the frequency of GAD65-specific IFN-γ- and IL-10-secreting T cells.
We previously demonstrated that administration of pDNA encoding GAD65-IgFc and IL-4 effectively inhibited ongoing β cell destruction in the pancreas of NOD mice through induction of GAD65-specific Th2-like reactivity (24). Furthermore, Tian et al. (39) reported that treatment with GAD65 prepared in IFA enhanced survival of syngeneic islet grafts in NOD recipients, although protection was short-lived. The focus of this study was to determine whether pDNA vaccination could inhibit autoimmune-mediated destruction of islet grafts and to define the conditions for efficient and long-term protection. Two key observations were made in this study. First, an Ag-specific pDNA-based strategy can be applied for long-term suppression of autoimmune destruction of pro-islet grafts. Survival of functional β cells found in the graft site of recipients vaccinated with JwGAD65+JwIL4+pIL10 correlated with the induction of GAD65-specific T cells secreting IL-4 and IL-10 and a concomitant decrease in IFN-γ-producing T cells (Figs. 4⇑ and 6⇑). Similarly, isolated islets implanted under the kidney capsule of recipient NOD mice immunized with JwGAD65+ JwIL4+pIL10 also remained insulitis free (Fig. 2⇑D). Second, conditions needed to inhibit β cell autoimmunity differ in the context of a pro-islet graft vs the intact pancreas. Whereas progression of insulitis in the endogenous pancreas was equally suppressed by immunization with JwGAD65+JwIL4 or JwGAD65+pIL10 (Table I⇑), protection of the pro-islet grafts varied between the respective treatment groups (Fig. 5⇑).
Protection of pro-islet grafts in JwGAD65+JwIL4+pIL10-immunized recipients was characterized by an increased number of intact pro-islets relative to the JwHEL+JwIL4+pIL10 or untreated recipient mice 15 wk postimplantation (Figs. 3⇑A and 5⇑A). The fact that the number of pro-islets detected in JwGAD65+JwIL4+pIL10-immunized recipients at 15 wk postimplantation was comparable to that typically found in recipients 2–3 wk after transplantation indicated that the majority of pro-islets that differentiated from the neonatal endocrine tissue had survived. Furthermore, virtually all of the intact pro-islets found in JwGAD65+JwIL4+pIL10-vaccinated recipients expressed high levels of insulin, thereby confirming the function of the grafted β cells (Figs. 3⇑B and 5⇑B). The number of pro-islets which arise from the neonatal pancreas are too few to contribute significantly to insulin blood levels so that function of the grafted β cells in this model is limited to immunohistological analyses. Strikingly, the intact pro-islets detected in the graft site of JwGAD65+JwIL4+pIL10-treated recipients were almost completely free of insulitis (Figs. 3⇑C and 5⇑C). This contrasted with JwHEL+JwIL4+pIL10-immunized recipients in which the few intact islets found in the graft site typically exhibited insulitis (Fig. 3⇑C). Lack of infiltrating monocytes and T cells would prevent direct β cell destruction and also negate the effects of proinflammatory cytokines such as IL-1, TNF-α, and IFN-γ which have been shown to influence the ability of β cells to secrete insulin in vitro (40). The latter likely explains why a significant number of intact pro-islets detected in mice immunized with JwHEL+JwIL4+pIL10 expressed either low or no detectable insulin (Fig. 3⇑B). The absence of insulitis also suggests that protection associated with JwGAD65+JwIL4+pIL10 treatment is primarily mediated in the periphery and not at the graft site. This finding is consistent with earlier work (19, 24) and herein (Table I⇑) demonstrating that the progression of insulitis is effectively inhibited in the pancreas by pDNA encoding GAD65-IgFc and IL-4 and/or IL-10. In part, this effect may be due to the ability of GAD65-specific T cells secreting IL-4 and IL-10 to reduce the frequency of naive β cell-specific precursors differentiating into Th1 cells (Figs. 4⇑ and 6⇑). On the other hand, the capacity of IL-4 and IL-10 to modulate production of various chemokines and expression of chemokine receptors by APC and/or T cells may block trafficking of established β cell-specific Th1 cells (41, 42, 43, 44). Efforts are ongoing to characterize the events involved in T cell migration to an islet graft and in turn how pDNA vaccination can alter this process.
Effective pro-islet graft protection required immunization with pDNA encoding GAD65-IgFc. Immunization with JwGAD65+JwIL4+pIL10 but not JwHEL+JwIL4+pIL10 resulted in increased numbers of intact, functional pro-islets free of insulitis. Nevertheless, administration of JwHEL+JwIL4+pIL10 delayed the onset of diabetes (Fig. 1⇑) and increased the number of intact pro-islets found in the graft site relative to untreated NOD recipients (Fig. 3⇑A). Notably, immunization of 10-wk-old NOD female mice with JwIL4+pIL10 alone elicited a similar profile of IFN-γ, IL-4, and IL-10 GAD65-specific reactivity compared with JwHEL+JwIL4+pIL10-treated mice (S.P. and R.T., unpublished data). The modulatory effect associated with JwIL4+pIL10 immunization is in agreement with other studies that have demonstrated that treatment of young NOD mice with IL-4 or IL-10 either as a recombinant protein or encoded by a genetic vaccine can modulate disease progression through induction of regulatory Th2 cells (29, 31, 45, 46, 47, 48). However, diabetes prevention via cytokine treatment-only has generally been successful when applied at relatively early stages of T1D progression. The limited immunotherapeutic efficacy associated with JwHEL+JwIL4+pIL10 treatment correlated with a significantly reduced frequency of GAD65-specific T cells secreting IL-4 and IL-10 compared with JwGAD65+JwIL4+IL10 (or JwGAD65+pIL10)-immunized animals (Figs. 4⇑ and 6⇑). We have previously shown that induction of Th2 cells at late preclinical stages of T1D is in itself not sufficient to prevent diabetes, but that the frequency of Th2 cells appears to be the determining parameter for effective immunoregulation and suppression of disease progression (18, 19).
Another important observation made in this study was that the requirements for effective protection of grafted vs endogenous islets differed. More stringent conditions were needed to prevent β cell destruction in grafted pro-islets compared with islets found in the pancreas. For instance, administration of JwGAD65+JwIL4 or JwGAD65+pIL10 proved to be effective in protecting endogenous islets from further infiltration (Table I⇑). Indeed, the frequency of insulitis was typical of that seen in 10- to 12-wk-old NOD female mice, suggesting that at the time of JwGAD65+JwIL4 or JwGAD65+pIL10 vaccination, the progression of pancreatic infiltration was effectively suppressed. However, these two treatment regimens had only limited efficacy in protecting pro-islet grafts when compared with recipients treated with JwGAD65+JwIL4+pIL10 (Fig. 5⇑). Differences were detected among the three treatment groups regarding the frequency of intact pro-islets, insulitis found within the graft site, and the phenotype of GAD65-specific T cells. For example, a significant reduction in the number of intact pro-islets was detected in NOD recipients treated with JwGAD65+JwIL4 vs JwGAD65+JwIL4+pIL10 (Fig. 5⇑A). The limited efficacy associated with JwGAD65+JwIL4 immunization correlated with a reduced frequency of GAD65-specific T cells secreting IL-10 and an increased IFN-γ response compared with JwGAD65+JwIL4+pIL10-immunized animals (Fig. 6⇑). On the other hand, a comparable frequency of GAD65-specific T cells secreting IL-4 was detected for the two different treatment groups (Fig. 6⇑). IL-10 is generally considered to have a broader range of immunosuppressive properties than IL-4, which in turn may contribute to islet graft protection. For instance, IL-10 inhibits cytokine production and Ag presentation by macrophages and dendritic cells and can reduce IFN-γ and IL-2 secretion by T cells (49). In addition, IL-10 has been shown to induce other subsets of immunoregulatory T cells such as T regulatory 1 cells that are characterized by high levels of IL-10 secretion but low IL-4 production (50). Furthermore, IL-10-secreting T cells have been reported to contribute to suppression of autoimmunity and allograft destruction (51, 52, 53). In contrast to the JwGAD65+JwIL4 group, JwGAD65+pIL10-treated recipients contained a similar number of intact pro-islets relative to recipients immunized with JwGAD65+JwIL4+pIL10 (Fig. 5⇑A). However, a marked increase in the frequency of insulitis in the graft site, coupled with reduced insulin expression, was apparent in JwGAD65+pIL10-treated recipients compared with JwGAD65+JwIL4+pIL10-immunized mice (Fig. 5⇑, A and B). Interestingly, the T cell cytokine profile in response to GAD65 was similar for the two treatment groups (Fig. 6⇑). This finding suggests that the frequencies of GAD65-specific T cells secreting IL-4 and IL-10 do not entirely account for the observed protection mediated by JwGAD65+JwIL4+pIL10 vaccination. Islet graft protection may be influenced by other regulatory cells such as CD4+CD25+ T cells or NK T cells and/or “modified” APCs which exhibit a regulatory phenotype (54, 55, 56). Another interesting aspect of this study is how immunoregulatory CD4+ T cells induced via pDNA administration influence β cell-specific CD8+ T cell reactivity. The expansion, effector function, and/or trafficking of key β cell-specific CD8+ clonotypes may be altered and/or CD8+ T cells exhibiting a regulatory function induced following JwGAD65+JwIL4+pIL10. Together, these results suggest that effective suppression of autoimmune-mediated destruction of islet grafts entails a complex set of events. Work is currently ongoing to further define the role and phenotype of regulatory T cells, the possible impact of “regulatory” APCs in islet graft protection induced by JwGAD65+JwIL4+pIL10, and the interplay between immunoregulatory CD4+ and β cell-specific CD8+ T cells.
In summary, we provide evidence that an Ag-specific pDNA-based strategy can be an effective approach to protect grafted islets from autoimmune-mediated destruction. In general, pDNA vaccination is well suited for the purpose of inducing transplantation tolerance. As shown here, the nature and magnitude of a response can be readily manipulated through the use of pDNAs encoding Ag and appropriate cytokines. Importantly, more stringent conditions are required to protect grafted islets than those found in the pancreas. Therefore, strategies used in the clinic as a prophylactic may not necessarily be applicable for the suppression of β cell autoimmunity targeting transplanted islets.
↵1 This study was supported by a grant from the Juvenile Diabetes Research Foundation. C.P.W. was supported by National Institute of Allergy and Infectious Disease Training Grant 5-T32-AI07273.
↵2 C.S. and S.P. contributed equally to this work.
↵3 Current address: Department of Medicine I, Friedrich-Alexander University, Ulmenweg 18, 91054 Erlangen, Germany.
↵4 Address correspondence and reprint requests to Dr. Roland Tisch, Department of Microbiology and Immunology, 804 Mary Ellen Jones Building, CB 7290, University of North Carolina, Chapel Hill, NC 27599-7290. E-mail address:
↵5 Abbreviations used in this paper: T1D, type I diabetes; GAD65, glutamic acid decarboxylase 65; HEL, hen egg lysozyme: NOD, nonobese diabetic; pDNA, plasmid DNA; SFC, spot-forming cell.
- Received December 30, 2002.
- Accepted March 5, 2003.
- Copyright © 2003 by The American Association of Immunologists