Type 1 diabetes is an autoimmune disease in which the body’s immune system attacks the body’s own components, resulting ultimately in the stress and death of insulin-producing beta cells in the pancreas. The exact impetus for the autoimmune breakdown is unclear, but the theory goes something like:
1. For some reason, the immune system becomes sensitized against pancreatic beta cell proteins like insulin and the enzyme GAD65, to which it makes antibodies (An antibody is a special protein released by the immune system that is designed to target and disable invading pathogens like bacteria and viruses; an autoantibody, then, is an antibody directed towards the self.)
2. This internal immune struggle against individual proteins expands, resulting in a very inflammatory environment in the tissues of the pancreas.
3. The stress is too much for beta cells, and they undergo apoptosis, or cell death.
4. As the beta cells die off, the body loses its ability to produce insulin. The result is diabetes.
If this is the general progression of type 1 diabetes, then it stands to reason that ameliorating the initial inflammatory reaction– against insulin, GAD65 and so on– could help prevent the gradual increase of inflammation and ultimate beta cell apoptosis. Thus, a number of researchers, including Leonard C. Harrison at the Walter and Eliza Hall Institute of Medical Research in Victoria, Australia, have been investigating the possibility that type 1 diabetes can be stopped in its early pre-clinical stage.
How does one prevent the immune system attack on these proteins? The general idea is a kind of reverse-vaccine: inundate the body with the protein of interest in a particular way, and the immune system will be normalized, accepting the protein as self rather than foreign invader.
A number of different approaches have been tried along this vein, with different proteins and different means of administration, but Harrison and his team have focused on nasal administration of the insulin protein. After seeing the success in mice of insulin administration [1 – 3], Harrison conducted a trial in which insulin was administered nasally to 52 recent-onset type 1 diabetics [4]. Unfortunately, the treatment did not prevent or significantly slow the death of beta cells in the participants. However, the treatment did desensitize the immune system so that when insulin was later injected the amount of antibodies made to the injected insulin was significantly reduced.
And in that reduction, there is hope for people who don’t have type 1 diabetes but who are at risk. Harrison and his team have therefore pushed forward with a trial of the nasal insulin vaccine in high risk children and young adults, believing they are on the path to prevent the autoimmune destruction of the beta cells, or at least slowing the progression to diabetes. We spoke to Dr. Harrison, hoping to find out more about the research he has done, and what comes next.
How will the success of the vaccine be measured? In trials of autoantigen delivery to newly-diagnosed diabetics, C-peptide levels serve as a measure of beta-cell ablation abatement. What do you measure when administering a vaccine to non-diabetics?
Outcome measures will be the participants’ first phase insulin response (FPIR) to glucose delivered intravenously and diabetes-free status 5 years after randomization (that is, whether the predicted 5 year proportion of participants with diabetes is reduced from 40 –50% by at least 50%).
Insulin is one of many candidate autoantigens that have been tried in the hopes of inducing immune tolerance. Why have you decided to administer insulin nasally, rather than proinsulin, GAD65, or any other peptide?
Several reasons: the evidence in mice and human children is that insulin is the primary autoantigen; insulin has been around for years and in fact was given nasally in the 1920s soon after it was first extracted from the animal pancreas to see if it would work that way for the treatment of diabetes. Therefore we had only minimal regulatory hurdles to overcome; in the non-obese diabetic mouse model of spontaneous type 1 diabetes we had found that aerosol/nasal insulin induced regulatory T cells that worked in a ‘bystander’ manner to suppress immune responses to another beta-cell autoantigen (GAD65).
Why have you decided to administer the vaccine nasally instead of orally or otherwise?
Insulin delivered nasally reaches the mucosa directly without being degraded, in contrast to insulin delivered orally, and is therefore more effective dose-for-dose. In addition, we had found that vaccination via the nasal route was able to induce changes in the mucosal immune system at other mucosal sites that we didn’t see via the oral route.
Trials of autoantigen delivery (intradermally, orally, intravenously, intranasally, etc.) have collectively had disappointing results, with some resulting in temporary beta-cell life extension, but ultimately not seeing any effect on C-peptide levels. What makes your approach different?
We are not starting at the end-stage disease (as with trials in rheumatoid arthritis, multiple sclerosis, and clinical type 1diabetes). The intranasal insulin vaccine trial II, currently underway, recruits children and young adults with two islet autoantibodies who have normal FPIR and oral glucose tolerance test results,and a 5 year risk of developing diabetes of 40-50%.
There are many unknowns and therefore many variables in trying to manipulate the immune system– delivery mechanism, dosing, timing, and so on. How have you determined the details of the trial– when to deliver insulin, how much to deliver, how it should be delivered?
It would take decades to refine all the variables in humans. We have been guided by studies, mostly unpublished, in the non-obese diabetic mouse and were fortunate in the Intranasal Insulin Trial (INIT I) (Harrison LC, et al Diabetes Care, 27:2348-55, 2004) to find a minimal dose that had bioactivity (elicited an immune response), something that had not been shown in previous trials of mucosally-administered autoantigens.
One of the concerns with immune-modifying therapies like this is that instead of inducing tolerance, the autoantigens will instead induce an immune response, resulting in more rapid autoimmune inflammation, or even autoimmunity where there may not have been otherwise. In your trials, is there any evidence that this sort of opposite-of-intent effect occurs? Is there any way to protect against it? How will you be measuring the possibility that vaccine induces diabetes?
We are very aware of this and have indeed published evidence for this ‘double-edged sword’ using a model system in mice (Hanninen A, et al J ClinInvest 109:261-7, 2002). There are ways of blocking the generation of potentially pathogenic CD8 T cells after mucosal antigen, but we haven’t had to consider these because there was no evidence that nasal insulin accelerated diabetes development in INIT I. Nevertheless, it is possible that removing parts of a protein that elicit potentially pathogenic T cells might make a protective vaccine more effective as we showed for a peptide from proinsulin (Martinez N,et al J ClinInvest 111: 1365-71, 2003).
There is some talk of the possible benefits of combination therapy– autoantigens plus anti-CD3 treatment, for example, as in the work of Bresson and von Herrath. Have you considered this possibility in terms of a vaccine? Is it viable and potentially beneficial?
Combinatorial treatment, incorporating a ‘reverse vaccine’ is on the drawing board. Note though that recent trials of anti-CD3 antibodies by Lilly and GlaxoSmithKline recent-onset type 1diabetics have not lived up to promise, but a variety of other agents are available to use in combination.
Are you monitoring T-cell levels and subpopulation distribution throughout the trial?
Yes, 3-monthly on blood cells frozen and batch assayed.
With a vaccine like this, it seems that the particular type of genetic predisposition to diabetes might be very relevant in defining patient response. That is, if a potential diabetic has the HLA-locus mutation, perhaps they will respond favorably to the autoantigen, while a patient with the insulin gene mutation might not, given that the alleged source of the problem is T-cell selection deep in the thymus. Are you controlling for genetic mutation types in your trial? Are participants a homogenous or heterogenous population in terms of their genotype?
The majority of participants have high or intermediate risk HLA genes for type 1diabetes. We are typing for HLAand for the insulin gene locus VNTR, as well as for 19 SNPs associated with type 1 diabetes. At this time we are not stratifying by genotype but it is almost certain that genetic heterogeneity will influence response to autoantigen-specific immunotherapy. We have identified proinsulin peptides that bind to the HLA molecules carried by nearly all people at risk for type 1 diabetes, and are thereby recognized by Tcells, and envisage that these could be used together to create a highly specific and safe vaccine.
For more information about the Diabetes Vaccine Development Centre and current trials visit: www.stopdiabetes.com.au
1. Zhang ZJ, Davidson L, Eisenbarth G, Weiner HL. Suppression of diabetes in nonobese diabetic mice by oral administration of porcine insulin. Proc Natl Acad Sci U S A 1991;88:10252–10256
2. Bergerot I, Fabien N, Maguer V, Thivolet C. Oral administration of human insulin to NOD mice generates CD4+ T cells that suppress adoptive transfer of diabetes. J Autoimmun 1994;7:655–663
3. Harrison LC, Dempsey-Collier M, Kramer DR, Takahashi K. Aerosol insulin induces regulatory CD8 gamma delta T cells that prevent murine insulin-dependent diabetes. J Exp Med 1996;184:2167–2174
4. Fourlanos S, Perry C, Gellert SA, Martinuzzi E, Mallone R, Butler J, Colman PG, Harrison LC. Evidence that nasal insulin induces immune tolerance to insulin in adults with autoimmune diabetes. Diabetes. 2011 Apr;60(4):1237-45
Len Harrison is a Senior Principal Research Fellow of the National Health and Medical Research Council (NHMRC) Australia and a Professor in the Department of Medical Biology at the University of Melbourne. He heads the Diabetes Laboratory at the Walter and Eliza Hall Institute of Medical Research and the Burnet Clinical Research Unit at the Royal Melbourne Hospital. He has authored 480 research publications on the actions of hormones and immune mechanisms of disease. His research is focused on the pathogenesis and prevention of diabetes. His awards for research include: C.J. Martin Fellowship, NHMRC; Wellcome (Glaxo) Australia Medal; Susman Prize, Australasian College of Physicians; Kellion Medal, Australian Diabetes Society; Rumbough Award for scientific excellence, Juvenile Diabetes Research Foundation (JDRF). His service commitments include: NHMRC Medical Research Committee and Health Research Ethics Committee; Secretary and President, Australian Diabetes Society; Director, Australian Society for Medical Research; Chair, JDRF Scientific Review Committee; Chair, Professional Advisory Panel JDRF Australia; President, Immunology of Diabetes Society.
