Prospects of a type 1 diabetes vaccine

KEYWORDS:

• Vaccine
• type 1 diabetes
• immune modulation
• combination therapy
• antigen specific

1. Introduction

​The hallmark characteristic of a vaccine is either prevention or cure of a disease by sensitizing or desensitizing the immune system to specific antigen(s) (Ags) and avoiding propagation of the pathogenic immune responses. In the case of autoimmune type 1 diabetes (T1D), in which the insulin-producing beta-cells within the pancreatic islets of Langerhans are selectively destroyed by the immune system (primarily the pathogenic T-cells), this entails reinstallment of immune tolerance to beta-cell Ags or peptides.

Ag-specific therapies have the potential to reset the immune system toward long-standing tolerance to the involved Ags, if applied correctly. Tolerance induction depends heavily on the route of administration (i.e. oral or parenteral), the time of administration with respect to disease stages and the Ag dose. High doses of Ag preferentially promote anergy or apoptosis of pathogenic T-cells, while low doses favor regulatory T-cell (Treg) induction or prevent immune-cell interactions [1]. Although the primary Ag in T1D is not yet identified, several exogenous Ags like dietary wheat and cow’s milk-derived proteins, displaying molecular mimicry with beta-cell surface proteins, in addition to endogenous (pro)insulin, glutamic acid decarboxylase (GAD65), the tyrosine phosphatase-related islet antigen (IA-2), heat shock protein (HSP60), islet-specific glucose-6-phosphatase catalytic subunit-related protein (IGRP), and zinc transporter (ZnT8) have been linked to T1D initiation and progression and are, therefore, potential therapeutic targets. Also environmental infectious agents, like viruses, contribute to triggering T1D; true vaccination against these agents could help in preventing disease.

The ideal timing for a prophylactic vaccine in T1D is probably before disease-onset during the presymptomatic stages, particularly before any immune attack against the beta-cell has been initiated [2]. Intervention at this stage is often studied in T1D animal models, like the nonobese diabetic (NOD) mouse, and Ag-specific interventions with administration of the suspected autoAgs, for example insulin, are effective in arresting disease. Unfortunately, clinical translation of these findings remains tough, as timing of Ag administration, Ag dose, repetition of Ag administration etc. are crucial factors in determining success or failure. Moreover, treating humans at this stage of the disease means treating at-risk populations without clinical symptoms, namely neonates or small children (often first-degree relatives of T1D patients) identified through genetic screening with high-risk HLA genotypes and autoantibody positivity. In this population, an Ag-based vaccine should not only prevent activation of naive T-cells but also control activated effector T-cells that are already damaging the beta-cell, as shown by autoantibody production from bystander B-cells, in order to prevent progression of the autoimmune attack and onset of disease. Importantly, although the majority of high-risk individuals are subjected to one or multiple environmental triggers, only a limited group will progress to disease. Once hyperglycemia is present, the challenge for successful vaccination becomes even greater as not only reinstallation and regulation of tolerance is needed, but the therapeutic success will depend on the moment of intervention and the amount of functioning beta-cells around to be saved. In the two latter stages of disease, when autoimmunity is happening, pure vaccination as in ‘administration of Ag in a tolerogenic way’ will probably not suffice to achieve disease prevention. In these cases, it is to be expected that combinations with immunomodulators will be necessary to arrest progression of disease.

2. Prophylactic vaccines

Several researchers have found that infections with members of the group B coxsackieviruses (i.e. coxsackievirus B1 [CVB1]) are associated with increased T1D risk, especially in genetically susceptible subjects before appearance of autoantibodies (reviewed in [3]). This observation may initiate the development of vaccines against confined viral peptides to prevent specific cases of T1D when applied early enough in life. In line with this strategy, vaccination of diabetes-prone NOD mice with formalin-fixed CVB1 prevented both viral replication and the acceleration of T1D onset.

Early introduction of infant formulas based on milk or consumption of cow’s milk products are associated with higher hazard ratio of incident T1D [4]. Increased intestinal permeability, known to be present in T1D subjects and even in those genetically at-risk, especially the HLA-A1-B8-DR3-DQ2 type, may expose the highly activated intestinal immune system to dietary Ags with epitopes mimicking beta-cell peptides. Furthermore, the lack of intestinal Tregs may facilitate an overreaction to harmless Ags, ultimately resulting in beta-cell autoimmunity in selected individuals [5]. How these findings can translate in a preventive vaccine need thoughtful consideration. Nevertheless, breast milk which contains growth factors, cytokines, and additional immunomodulatory molecules can stimulate maturation of the mucosa-associated lymphoid tissue, thereby encouraging the development of oral tolerance to beta-cell autoAgs, like insulin, in breast milk. Moreover, high-risk newborns weaned to strongly hydrolyzed casein-based formulae were less prone to T1D development by the age of 10 years, but so far the mechanisms of protection are not fully understood.

Insulin, as a key autoAg in T1D, has been tested as preventive vaccine in high-risk individuals with varying outcomes. Unfortunately, none are living up to the high expectations created by preclinical data where oral, subcutaneous, intravenous, and intranasal administration of insulin could prevent or at least delay T1D onset in the NOD mouse model. Noteworthy, post hoc analyses of the Diabetes Prevention Trial-Type 1 (DPT-1) with oral insulin discovered beneficial effects of the vaccine in a subgroup possessing high insulin autoantibody levels [6], which prompted a new trial specifically for this population [7]. When high-risk children, without prior beta-cell autoimmunity, were fed high doses of insulin, a regulatory immune response could be observed. Successful induction of oral tolerance was indicated by the increased saliva IgA binding to insulin and regulatory profile in insulin-responsive T-cells in treated subjects [8]. These encouraging findings are from a small cohort and are undergoing further investigation in a larger phase II trial [9].

3. Therapeutic vaccines

Subcutaneous vaccination with an immunomodulatory protein derived from HSP60 (i.e. DiaPep277) in new-onset T1D patients showed improved metabolic parameters associated with a shift in the autoimmune response from T helper 1 (Th1) to Th2 profile. However, clinical studies could never show therapeutic effect in humans, with initial reports retracted later on for fraud [10].

Vaccinations with GAD65 formulated with aluminum hydroxide (GAD-alum) showed the first promising results in new-onset T1D patients, likely due to induction of GAD-specific Tregs [11]. However, these results could not be confirmed in larger trials.

Intervention trials with intradermal injections of a single HLA-DR4-restricted proinsulin epitope were not associated with risk of systemic allergic hypersensitivity nor with aggravated pro-inflammatory T-cell responses in new-onset T1D patients or those with established disease (MonoPepT1De trial [12]). Moreover, intradermal delivery of multiple beta-cell peptides (MultiPepT1De) into humanized HLA-DRB1*0401 transgenic mice enhanced proliferation of proinsulin-specific Tregs as well as modulated auto-inflammatory responses after breakdown of proinsulin tolerance [13]. These promising outcomes opened the way for an ongoing intervention trial in new-onset T1D patients [14]. Promising results were also obtained using a plasmid DNA vaccine encoding proinsulin in patients with T1D for up to 5 years. Reduced insulin-specific CD8⁺ T-cells temporarily preserved beta-cell function; however, this effect diminished after discontinutation of therapy [15].

Another interesting approach is the exploitation of nanoparticle (NP) vaccines coated with T1D-relevant peptide-MHC (pMHC) complexes (Parvus Therapeutics Inc, Calgary, Alberta, Canada). NOD mice treated with pMHC-NPs containing IGRP epitopes expanded cognate Tregs with memory phenotype, induced anery in noncognate T-cells and reverted new-onset diabetes [16]. Initial steps have been taken to embark on a phase I trial in new-onset T1D patients.

Other cellular therapies, such as autologous transfers of ex vivo expandend Tregs and tolerogenic dendritic cells, also hold great potential; however, these go beyond the scope of this review and are discussed elsewhere [17,18].

To date, the prophylactic and therapeutic vaccines tested in clinic only demonstrated temporary success and they often lacked the efficacy to entirely prevent or revert diabetes. Each of these monotherapies only targets one aspect of this complex multifactorial disease, thus providing a rational for the insufficient therapeutic outcome.

4. Expert opinion

The success (or failure) of a prevention or intervention trial with an immune vaccine for T1D seems to depend on correct inclusion criteria and primary trial end points. Crucial to these aspects is the discovery or development of appropriate biomarkers. As was concluded from the DPT-1 trial, the use of immune biomarkers such as autoantibody positivity or presence of autoreactive T-cells may be essential to including the right patients into the right trials. Indeed, perhaps not all patients will be responsive to a similar therapy and in that case, we will need a large repertoire of Ag-specific vaccines to be able to achieve this kind of personalized medicine.

Over the past decade, increasing evidence demonstrated that stress-induced posttranslational modifications can result in the generation of neo-epitopes capable of eliciting both innate and adaptive beta-cell autoimmune responses [19]. These observations can further fuel the Ag repertoire for potential therapeutics. Still, while Ag-specific therapies hold great potential, they all currently lack the necessary robustness to reverse beta-cell autoimmunity at the moment of onset or beyond. At those stages, in animal models only combination therapies, consisting of Ag administration with general immunomodulators, are able to safely inhibit or even stably inverse the disease process. A promising intervention is anti-CD3, which can stably revert established (new-onset) diabetes in NOD mice. In T1D humans, the application of the anti-CD3 antibodies otelixuzumab or teplizumab is currently hampered by observations of Epstein–Barr virus reactivation and flu-like symptoms under doses used [20]. Today, dose-finding studies are undertaken to find a safe and effective dose in new-onset T1D patients [21]. Of note, although subtherapeutic doses of anti-CD3 antibodies may not cure disease, they can still induce Treg accumulation, thereby opening new avenues for these T-cell targeting agents combined with Ag-specific vaccines.

In our eyes, one such strategy is the combination of low-dose anti-CD3 with selected beta-cell Ags (i.e. proinsulin, GAD65, IA-2) and anti-inflammatory cytokine IL10 administered via the mucosal route using the bacterial vehicle Lactococcus lactis (L. lactis, Intrexon Actobiotics nv, Zwijnaarde, Belgium). Genetically modified L. lactis have been developed as oral or intranasal Ag carriers delivering precise doses of bioactive molecules. Our studies demonstrated that this combination, with either proinsulin or GAD65 as beta-cell Ag, could stably install long-standing tolerance in ~60% of new-onset diabetic NOD mice [22,23]. Disease reversal was dependent on the presence and functionality of CD4⁺Foxp3⁺ T cells [24]. In a phase I trial for Crohn’s disease, L. lactis secreting IL10 displayed an excellent safety profile, being well-tolerated and biologically contained [25]. A great advantage of this combination is its excellent clinical applicability and safety profile. Low-dose anti-CD3, as is recommended here, is unlikely to induce unwanted side effects such as those seen in phase I and II trials. Diabetes reversal correlated strongly with residual beta-cell mass at therapy initiation and reactivity to the autoAg. These parameters will strongly influence patient eligibility in potential future trials. Moreover, a decline in insulin autoantibody positivity was an immune biomarker of therapeutic outcome that could be a useful tool during patient follow-up.

Besides anti-CD3 antibodies, also other immunomodulators could be excellent candidates in combinations with Ag-based vaccines. Although T1D is a T-lymphocyte-mediated disease targeting B-cells, through transient depletion with the anti-CD20 monoclonal antibody rituximab, was also able to partially preserve beta-cell function in early T1D [26]. Subcutaneous administration of denatured insulin encapsulated in poly(lactic-co-glycolic acid) microparticles together with a peptide hydrogel containing granulocyte macrophage colony-stimulating factor and CpG-DNA protected 40% of NOD mice from developing T1D [27]. Improving immune regulation, i.e. by systemic administration of recombinant low-dose IL2 in conjunction with Ag vaccination may be another approach to expand the Ag-specific Treg repertoire. Low-dose IL2 is well tolerated and immune effective in new-onset diabetic mice and men [28]. It is currently being tested as a monotherapy in new-onset T1D patients [29].

Although many therapies have already been explored and were not as successful in clinical practice as was expected based on preclinical data and preliminary trials, we have learned a great deal from these endeavors. While combination therapies may be difficult to conduct in patient settings, they clearly hold the most therapeutic potential.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Additional information

Funding

The authors have received grants from the University of Leuven [KULEUVEN GOA 2014-10], the European Foundation for the Study of Diabetes [EFSD Sanofi 2014], and the Research Fund of Flanders (FWO) [grant no. FWO G 0554 13N]. DP Cook also has received an FWO scholarship [1 1Y67 16N].