DNA vaccination to treat autoimmune diabetes

Current state of antigen-specific immunotherapy for type 1 diabetes

PURPOSE OF REVIEW

Update on antigen-specific immunotherapy (ASIT) in type 1 diabetes (T1D) with focus on deoxyribonucleic acid (DNA)-induced immunization and the current obstacles to further research and clinical realization.

RECENT FINDINGS

In T1D, immune system imbalances together with malfunctioning islet-specific processes cause autoreactive immune cells to destroy beta cells in the islets. ASIT may restore self-tolerance; however, the approach has yet to fully meet its promise and may require co-administration of antigen (preproinsulin) and suitable immune response modifiers.

SUMMARY

A self-tolerant immune system may be regained using ASIT where T effector cells are repressed and/or T regulatory cells are induced. Administration of exogenous antigens has been safe in T1D. Conversely, adequate and lasting beta cell preservation has yet to be tested in sufficiently large clinical trials in suitable patients and may require targeting of multiple parts of the immunopathophysiology using combination therapies. DNA-based induction of native antigen expression to ensure important posttranscriptional modifications and presentation to the immune system together with tolerance-enhancing immune response modifiers (i.e., cytokines) may be more efficacious than exogenous antigens given alone. Progress is limited mainly by the scarcity of validated biomarkers to track the effects of ASIT in T1D.

Therapies to Preserve β-Cell Function in Type 1 Diabetes

In spite of modern techniques, the burden for patients with type 1 diabetes mellitus will not disappear, and type 1 diabetes will remain a life-threatening disease causing severe complications and increased mortality. We have to learn of ways to stop the destructive process, preserve residual insulin secretion or even improve the disease via β-cell regeneration. This will give a milder disease, a more stable metabolism, simpler treatment and perhaps even cure. Therapies based on single drugs have not shown sufficient efficacy; however, there are several treatments with encouraging efficacy and no apparent, or rather mild, adverse events. As the disease process is heterogeneous, treatments have to be chosen to fit relevant subgroups of patients, and step by step efficacy can possibly be improved by the use of combination therapies. Thus immunosuppressive therapies like anti-CD3 and anti-CD20 monoclonal antibodies might be combined with fusion proteins such as etanercept [tumor necrosis factor (TNF)-α inhibitor] and/or abatacept (CTLA4-Ig) early after onset to stop the destructive process, supported by β-cell protective agents. The effect may be prolonged by using autoantigen therapy [glutamate decarboxylase (GAD) proinsulin], and by adding agents facilitating β-cell regeneration [e.g. glucagon-like peptide-1 (GLP-1)] there should be a good chance to make the disease milder, perhaps leading to cure in some patients.

The clinical and immunological significance of GAD-specific autoantibody and T-cell responses in type 1 diabetes

Antigen-specific interventions are desirable approaches in Type 1 Diabetes (T1D) as they can alter islet-specific autoimmunity without systemic side effects. Glutamic acid decarboxylase of 65 kDa (GAD65) is a major autoantigen in type 1 diabetes (T1D) and GAD-specific autoimmunity is a common feature of T1D in humans but also in mouse models of the disease. In humans, administration of the GAD65 protein in an alum formulation has been shown to reduce C-peptide decline in recently diagnosed patients, however, these observations were not confirmed in subsequent phase II/III clinical trials. As GAD-based immune interventions in different formulations have successfully been employed to prevent the establishment of T1D in mouse models of T1D, we sought to analyze the efficacy of GAD-alum treatment and the effects on the GAD-specific immune response in two different mouse models of T1D. Consistent with the latest clinical trials, mice treated with GAD-alum were not protected from diabetes, although GAD-alum induced a GAD-specific Th2-deviated immune response in transgenic rat insulin promoter-glycoprotein (RIP-GP) mice. These observations underline the importance of a thorough, preclinical evaluation of potential drugs before the initiation of clinical trials.

Mucosal exposure to antigen: cause or cure of type 1 diabetes?

The human gut offers more than 200 m2 of mucosal surface, where direct interactions between the immune system and foreign antigens take place to eliminate pathogens or induce immune tolerance toward food antigens or normal gut flora. Therefore, mucosally administered antigens can induce tolerance under certain circumstances. In autoimmune diabetes, mucosal vaccination with autoantigens elicits some efficacy in restoring tolerance in mice, but it never succeeded in humans. Furthermore, in some instances autoimmunity can be precipitated upon oral or intranasal autoantigen administration. Therefore, it is difficult to predict the effect of mucosal vaccination on autoimmunity and much effort should be put into establishing better assays to reduce the risk for possible adverse events in humans and enable a rapid and smooth translation.

Gene-gun biolistic immunization encoding glutamic acid decarboxylase: a model for studying Langerhans cell abnormalities and mimicry in the nonobese diabetic mouse

Plasmid-DNA gene-gun immunization may be an efficient approach for investigating the role of skin dendritic cells (DCs) in type 1 diabetes (T1D) pathogenesis and the significance of the presentation of peptides that mimic autoantigenic epitopes in aggravating or modulating the autoimmune reaction. Gene-gun immunization has been described as producing long-lasting immune responses elicited by skin DCs, especially Langerhans cells (LCs). Therefore, we tested the immune response and diabetes modulation in nonobese diabetic (NOD) mice and in control BALB/c mice, by gene-gun administration of plasmid-DNA encoding (1) human 65 kDa glutamic acid decarboxylase (hGAD65) mimicking the crucial mouse autoantigen GAD65 (similarity of 95.7%) or (2) beta-galactosidase (betaGAL) as a negative control. Expression of GAD and betaGAL in skin of pc-GAD- and pc-LacZ-injected mice, respectively, was confirmed. It was surprising that both pc-LacZ-injected BALB/c and NOD mice exhibited a betaGAL-specific Th1 immune response: spleen cells of pc-LacZ mice proliferated specifically to betaGAL (P < 10(-4)) and secreted significant amounts of IFNgamma (P < 10(-4)). pc-LacZ mice also developed a betaGAL-specific Th1-related (IgG2a/2c) and Th2-related (IgG1) humoral response. Although pc-GAD BALB/c mice showed Th2-related GAD-specific IgG1 production and a significant secretion of IL4 (P < .03), pc-GAD NOD mice did not generate either an antibody response or a T cell response specific to GAD. Moreover, gene-gun immunization encoding hGAD65 did not clearly modulate diabetes onset in NOD mice. This absence of detectable GAD-specific response may implicate skin DC deficiencies in NOD mice. The gene-gun technique could thus provide an interesting model for studying skin DC abnormalities in NOD mice and their potential implication of presenting mimetic peptides that modulate the autoimmune response in T1D.

Diverse Plasmid DNA Vectors by Directed Molecular Evolution of Cytomegalovirus Promoters

Genetic vaccinations, gene therapy, and manufacturing of therapeutic proteins would benefit from promoter sequences that provide improved or prolonged expression levels. The cytomegalovirus (CMV) promoter is one of the most potent promoters known to date, and no previous examples of improved activity of this promoter by sequence mutagenesis have been reported. This study describes directed molecular evolution of CMV promoters derived from two human and two nonhuman primate strains of CMV by DNA shuffling and screening. Libraries of chimeric promoters were screened and analyzed for expression levels and immune responses, using plasmid DNA vectors encoding luciferase and beta-galactosidase. The results indicate that high functional diversity among CMV promoters can be generated, and the chimeric promoters selected after two rounds of DNA shuffling and particularly designed screening assays provided approximately 2-fold increased luciferase reporter gene expression and anti-beta-galactoside antibody response in vivo when compared with wild-type promoters. Sequence analysis of the shuffled promoters identified several mutations potentially contributing to the observed enhanced or reduced promoter activities and identified a 42-nucleotide region that appears obsolete for the functioning of the CMV promoter. Taken together, these data demonstrate the feasibility of generating diverse promoter sequences by DNA shuffling and screening methods, and provide novel structure- function information about CMV promoters. DNA shuffling and screening technologies provide a new approach to promoter optimization and development of optimal expression vectors for genetic vaccinations, gene therapy, and protein expression.

Microorganisms and autoimmunity: making the barren field fertile?

Microorganisms induce strong immune responses, most of which are specific for their encoded antigens. However, microbial infections can also trigger responses against self antigens (autoimmunity), and it has been proposed that this phenomenon could underlie several chronic human diseases, such as type 1 diabetes and multiple sclerosis. Nevertheless, despite intensive efforts, it has proven difficult to identify any single microorganism as the cause of a human autoimmune disease, indicating that the 'one organism-one disease' paradigm that is central to Koch's postulates might not invariably apply to microbially induced autoimmune disease. Here, we review the mechanisms by which microorganisms might induce autoimmunity, and we outline a hypothesis that we call the fertile-field hypothesis to explain how a single autoimmune disease could be induced and exacerbated by many different microbial infections.

Intradermal or oral delivery of GAD-encoding genetic vaccines suppresses type 1 diabetes

Genetic vaccines are promising candidates for prevention of type 1 diabetes, an autoimmune disease resulting from cell-mediated destruction of pancreatic beta cells. It is known that the prophylactic effect and immune responses induced by administration of a genetic vaccine can depend on site of delivery. In the work presented here, we used the NOD mouse model for type 1 diabetes to evaluate different routes of delivery for DNA vaccines coding for the beta-cell antigen glutamic acid decarboxylase (GAD). Plasmid DNA coding for intracellular or secreted GAD was given via either the intramuscular (i.m.), intradermal (i.d.), or oral route, using, respectively, 300, 100, or 300 micro g DNA per mouse. Results indicated that both i.d. and oral delivery of GAD-encoding DNA were more effective than i.m. delivery for disease suppression. In addition, cytokine-specific ELISpot analysis indicated that immune responses induced by the different immunization protocols were more dependent on the cellular localization of GAD antigen than on the delivery route, while ELISA of anti-GAD serum antibody isotypes indicated that i.d. delivery of DNA was most likely to induce a Th2-like response. Our results suggest that i.d. or oral delivery of a genetic vaccine for type 1 diabetes might be preferable over the i.m. route in a future clinical setting.

Development of new strategies to prevent type 1 diabetes: the role of animal models

Type 1 diabetes is an immune-mediated disease typically preceded by a long preclinical stage during which a growing number of islet-cell-specific autoantibodies appear in the serum. Although antigen-specific T lymphocytes and cytokines rather than these autoantibodies are the likely executors of beta-cell-destruction, these autoantibodies reflect the existence of autoimmunity that targets islet beta-cells. Abrogation of this autoimmunity during the preclinical stage would be the key to the prevention of type 1 diabetes. However, the quest of protecting islet-cells from the immune attack requires detailed knowledge of mechanisms that control islet-inflammation and beta-cell-destruction, and of mechanisms that control immune tolerance to peripheral self-antigens in general. This knowledge can only be obtained through further innovative research in experimental animal models. In this review, we will first examine how research in non-obese diabetic mice has already led to promising new strategies of diabetes prevention now being tested in human clinical trials. Thereafter, we will discuss how recent advances in understanding the mechanisms that control immune response to peripheral self-antigens such as beta-cell antigens may help to develop even more selective and effective strategies to prevent diabetes in the future.

Decreased insulitis and blood glucose levels after injection of GAD-transduced lymphocytes into NOD mice

We used NOD mice to investigate the effects of injecting transduced lymphocytes on insulitis, nonfasting blood glucose levels, and immune responses. Syngeneic splenocytes were transduced with retroviral particles carrying a cDNA construct encoding the beta cell antigen glutamic acid decarboxylase (GAD65), a secreted form of GAD65 (SGAD55), or secreted alkaline phosphatase (SEAP) as a control antigen. Different multiplicities of infection (m.o.i.) were used with different constructs. Four-week-old NOD mice received intravenous injection of CD4(+) cells isolated from transduced splenocytes, and insulitis and blood glucose levels were determined at 10 weeks of age. No significant effects were observed with lymphocytes transduced with gad65 and sgad55 constructs at low m.o.i. By contrast, at high m.o.i., lymphocytes transduced with the sgad55 and seap constructs caused a decrease in insulitis and blood glucose levels and in insulitis alone, respectively. ELISA of anti-GAD antibody isotypes indicated that GAD-transduced lymphocytes induced similar Th2-like responses at all m.o.i. These results suggest that retroviral particles carrying sgad55 can be used for engineering cell vaccines for type 1 diabetes and provide further evidence that Th2-like responses induced by immunization may not always be a primary cause of diabetes suppression in NOD mice.

Vaccination to prevent type 1 diabetes

Immune-based interventions to treat or prevent autoimmune disorders are opening up entirely novel therapeutic avenues. Presently, systemic immune modulators such as antibodies to cytokines or cell-surface molecules (i.e., costimulators) are undergoing clinical testing or are already in use. The next generation of interventions will function in an antigen-specific way. These strategies are more difficult to design, because successful immunization or vaccination has to be monitored on an individual basis and dose/route/timing considerations play an essential role. Both will be discussed in this chapter and suitable combinatorial approaches will be suggested.

Endogenous expression levels of autoantigens influence success or failure of DNA immunizations to prevent type 1 diabetes: addition of IL-4 increases safety

Administration of autoantigens through DNA immunizations or via the oral route can prevent progression of islet destruction and lower the incidence of type 1 diabetes in animal models. This beneficial effect is mediated by autoreactive regulatory CD4 lymphocytes, and it is known that their induction depends on the precise dose and route of antigen administration. However, it is not clear which endogenous factors determine when such immunizations lead to activation of regulatory versus aggressive autoreactive lymphocytes and how a deleterious outcome can be avoided. Here we describe novel observations made in an animal model for virally induced type 1 diabetes, showing that the endogenous expression levels of the islet antigens and glutamic acid decarboxylase determine whether immunization with these antigens is beneficial or detrimental. Lower expression levels in beta-cells support immune regulation resulting in induction of autoreactive, regulatory cells characterized by increased IL-4 production (Th2-like), whereas higher levels favor Th1-like autoaggressive responses characterized by augmented IFN-gamma generation. Co-immunization with an IL-4-expressing plasmid reduces the risk of augmenting autoaggression and in this way increases the safety margin of this immune-based therapy. Our findings will be of importance for designing safe antigen-specific interventions for human type 1 diabetes.

Effects of plasmid DNA injection on cyclophosphamide-accelerated diabetes in NOD mice

Type 1 diabetes results in most cases from the destruction of insulin-secreting beta cells by the immune system. Several immunization methods based on administration of autoantigenic polypeptides such as insulin and glutamic acid decarboxylase (GAD) have been used to prevent autoimmune diabetes in the non-obese diabetic (NOD) mouse. In the work presented here, a gene-based approach was taken for a similar purpose. A plasmid carrying different cDNAs was used to investigate the effects of injecting naked DNA on cyclophosphamide-accelerated diabetes in female NOD mice. Four-week-old animals received intramuscular injections of plasmid DNA encoding either intracellular GAD, a secreted form of GAD, or a secreted form of a soft coral luciferase. Monitoring of glycosuria and hyperglycemia indicated that injection of plasmid DNA encoding secreted GAD and secreted luciferase could prevent and delay diabetes, respectively. In contrast, injection of DNA encoding intracellular GAD did not suppress the disease significantly. Analysis of anti-GAD IgG(1) antibody titers in animal sera indicated that diabetes prevention after injection of GAD-encoding DNA was possibly associated with increased Th2-type activity. These results suggest that cellular localization of GAD is a factor to consider in the design of GAD-based genetic vaccines for the prevention of autoimmune diabetes.