Editorial: A Tale of Two Cousins: Type 1 and Type 2 Diabetes

Insulin secretion and β-cell biology are the main theme of the Servier-IGIS Meetings that have been held yearly since 2000 in St. Jean Cap Ferrat in southern France (1–5). Previous symposia have mostly focused on insulin secretion and its relationship with β-cell defects in type 2 diabetes. Type 1 diabetes has not previously been considered, since it is generally viewed exclusively as an autoimmune disease. However, the pathophysiology of either disease is centered on β-cells, and type 1 and type 2 diabetes share many concerns, especially the goal of preserving or restoring a normal functional β-cell mass.

Evidence has accumulated that immune tolerance reflects a permanent cross talk between the different cells involved in immune survey functions and peripheral tissues, as in the case of β-cells in type 1 diabetes. As in most common autoimmune diseases, the problem of the antigenic specificity of the autoimmune reaction that drives the β-cell damage has been one with most unpredictable issues over the past 20 years. Whether the activation of autoimmunity proceeds from intrinsic immune dysregulation or requires the presence of β-cells is now no longer a question. Against all predictions, the search for the “diabetes autoantigen” has resulted in a long list of candidates with rather loose identification criteria. All presently high-ranking β-cell autoantigens are defined as proteins expressed by β-cells, not as β-cell–specific antigens. Indeed, evidence has accumulated to indicate that B- and T-cells recognize many autoantigens, rather than a single autoantigen, during diabetes development and that most autoantigens are not β-cell specific. This makes type 1 diabetes a β-cell disease rather than an antigen-specific immune disease.

As for type 2 diabetes, there is overwhelming evidence that it cannot be considered exclusively from the angle of the β-cell. β-Cells are at the center of multiple physiological loops that send signals and receive information to control energy disposal and storage. A key loop is the one between β-cells and adipocytes, establishing a direct link between metabolic and immune pathways. Many hormones, cytokines, transcription factors, and bioactive lipids are common to metabolic and immune pathways. Macrophages and adipocytes are overlapping cells that establish a functional link between energy storage functions and defense against pathogens. Not unexpectedly, obesity is associated with a state of chronic inflammation that has direct implications for the pathophysiology of type 2 diabetes. The close evolutionary links between immune and metabolic pathways raise the possibility that type 1 and type 2 diabetes may not be as different as initially thought. Moreover, uncommon forms of diabetes, namely latent autoimmune diabetes in adults (LADA), have been considered as possible overlaps between type 1 and type 2 diabetes. The lack, however, of clear biological criteria for the diagnosis of type 2 diabetes leaves open the issue of overlapping forms of diabetes.

In addition to inflammation, which constitutes a possible link between type 1 and type 2 diabetes, both diseases are highly polygenic. Among the several problems that await solutions in type 1 and type 2 diabetes are the initial triggers that lead to failure of immune tolerance in type 1 diabetes and to failure of insulin secretion in type 2 diabetes, the relative contribution of environmental factors and of genetic variants to both diseases on a highly polygenic background, and others.

This volume, the sixth of a series published in Diabetes, collects the proceedings of the Sixth Servier-IGIS Symposium, which reviewed the evidence for or against a common basis for the two forms of diabetes. Section I details issues that relate to the initiation of the autoimmune process in type 1 diabetes and issues that must be considered in the pathophysiology of type 2 diabetes. Hugh McDevitt, in his introductory talk, set forth the molecular basis of the mechanisms that maintain immune tolerance to self and its failure in type 1 diabetes. Most forms of autoimmunity start with the presentation of autoantigens, in the form of short peptides, to T-cells. Class II major histocompatibility molecules are the “molecular filter” that determine the array of antigenic peptides presented to T-cells. Expression of the right class II variant is a prerequisite for an individual to develop an immune reaction of a defined specificity, i.e., an autoimmune reaction. Genes that encode restricting class II molecules are the strongest genetic determinant of the development of autoimmunity, as best exemplified in type 1 diabetes. The demonstration that major tissue-specific autoantigens are expressed in the thymic medullary epithelium under the control of the AIRE gene is a breakthrough in the understanding of the mechanisms that control the education of T-cells in the thymus. Furthermore, educated regulatory T-cells that express the FoxP3 gene and secrete regulatory lymphokines are key cells in controlling the activation of autoreactive T-cells in the periphery. Insulin is among the autoantigens expressed within the thymic medulla. The role of thymic expression of insulin is developed in a second article in relation with the role of the variable number of tandem repeats (VNTR) polymorphism located upstream of the insulin promoter. This VNTR is strongly associated with type 1 diabetes. A two- to threefold higher rate of insulin gene transcription is observed in individuals expressing class III VNTRs than in individuals expressing class I VNTRs in the thymus. This article provides preliminary evidence in the human that class III alleles, which confer dominant protection from the development of type 1 diabetes, are associated with higher interleukin-10 T-cell responses to proinsulin than class I alleles. The level of thymic expression of proinsulin may thus directly impact on the repertoire and function of insulin-specific T-cells in the periphery, as evidenced in the mouse.

As underscored in this section, the etiology of type 1 diabetes remains unknown. One intringuing observation in humans is the very early detection of islet cell autoantibodies, in the first-line anti-insulin autoantibodies, in individuals at risk for diabetes, as shown in the BABYDIAB study. Once islet cell autoantibodies have developed, the progression to diabetes is determined by the age at which antibodies appear and by the magnitude of the autoimmune response, in turn related to the age of the subject. Although it is the most comprehensive follow-up study of individuals at risk for type 1 diabetes, this study, like previous studies, has so far not been able to definitively identify the environmental factors that may initiate the autoimmune process. This is quite unexpected according to the dominant hypothesis in autoimmunity that, on a highly polygenic susceptibility background, an environmental factor is the triggering event that initiates the failure of immune tolerance to β-cells. This model is challenged by both the failure to identify a unique environmental factor in humans and by animal models of type 1 diabetes, raising the possibility that stochastic events influence diabetes development, i.e., multiple environmental factors may intervene in disease evolution, and that on a genetic susceptibility background, protective environmental factors may be even more important. None of these hypotheses are presently incompatible with the BABYDIAB data. Studies of insulin production along the natural history of type 1 diabetes were reviewed in this section. There is progressive impairment in insulin secretory responses before diabetes, but the β-cell reserve to physiologic stimuli may still be substantial at the time of diagnosis, although maximal responses are more impaired. Other factors, including insulin resistance, may play a role in the timing of the clinical presentation of diabetes along this continuum. The presence of islet cell autoantibodies predicts future decline, and age is a determinant of residual insulin production at diagnosis. This is important since historical as well as recent clinical experience has emphasized the importance of residual insulin production for glycemic control and prevention of end-organ damage. As for the role of an intrinsic (nonimmune) β-cell defect in the development of type 1 diabetes, the experimental evidence remains indirect and inconclusive. This, nevertheless, sets up a possible link between type 1 and type 2 diabetes. Familial clustering of type 1 and type 2 diabetes has been reported in many studies, suggesting common genetic susceptibility. Patients with double genetic predisposition have an intermediate phenotype that partly overlaps with LADA. The striking association of type 2 diabetes with HLA DR-4 haplotypes, however, does not implicate class II genes, suggesting that other genes on the short arm of chromosome 6 may be shared. It is proposed that a proportion of diabetic patients may have both type 1 and type 2 processes contributing to their diabetic phenotype. As for the role of the insulin gene VNTR, data have been controversial in obesity and type 2 diabetes, underscoring the need for functional and genetic studies in large, carefully phenotyped cohorts.

Section II overviews type 1–related forms of diabetes that may proceed from mechanisms at least in part different from those at play in the common forms of type 1 diabetes. Characterization of autoantigens that are recognized by T- and B-cells all along the development of type 1 diabetes has proved a major advance in β-cell biology. A first report described the most “recent” of the major β-cell autoantigens, insulinoma-associated protein 2 (IA-2) and IA-2β, members of the transmembrane protein tyrosine phosphatase family located in dense core vesicles of neuroendocrine cells, including the β-cells of pancreatic islets. As developed in Section III for insulin, gene knock out is a powerful approach in defining the role of candidate autoantigens in β-cell physiology and, in the presence of a diabetes susceptibility background, in type 1 diabetes. The report in this volume describes double IA-2^−/−/IA-2β^−/− knockout mice, which show glucose intolerance and an absent first-phase insulin response, indicating that the dense core vesicle proteins IA-2 and IA-2β, alone or in combination, are involved in insulin secretion. But neither alone nor in combination are they required for the development of diabetes in NOD mice. From an immunological standpoint, islet cell autoantibodies are strongly associated with the development of type 1 diabetes, as reviewed in the same section. In the absence of reliable T-cell tests, dissection of autoantibody responses in subjects at genetic risk has so far proved the best marker for autoimmunity, including in uncommon forms of diabetes, as best exemplified by LADA. Despite differences in the pattern of autoantibodies, specificity of autoreactive T-cells, and level of insulin resistance detected in both clinical entities, the issue of a common versus distinct pathophysiological background in type 1 diabetes and LADA remains open. LADA is a disorder in which, despite the presence of islet antibodies at diagnosis of diabetes, slowly progressive β-cell failure dictates the evolution of diabetes. The choice of insulin as a first-line treatment of LADA is discussed in this section.

Section III covers the basis for inflammation and its role in β-cell damage in type 1, and possibly in type 2, diabetes. Since the seminal demonstration in 1993 by G. Hotamisligil that adipocytes produce tumor necrosis factor (TNF)-α and that TNF-α expression is increased in adipocytes of obese animals, accumulating evidence has associated obesity with low-grade inflammation and inflammation with type 2 diabetes. The roles of proinflammatory cytokines in modifying insulin signaling, and of obesity-related endoplasmic reticulum stress in generating insulin resistance, have since been largely elucidated. The central role of β-cells in diabetes, as contributor to the generation of the autoimmune reaction in type 1 diabetes, and as victim of inflammatory attacks in both type 1 and type 2 diabetes, was further developed in this section. Experiments in the NOD mouse and epidemiological evidence in the human point to proinsulin as a key autoantigen in type 1 diabetes. The functional importance of insulin, the high number of autoantigens characterized at different stages of diabetes, and their clustering within β-cell subparticles point to the islet as a starting point for the initiation of the disease. The knock out of either of the proinsulin genes in the mouse shows a direct impact on the repertoire of proinsulin-specific T-cells and/or modulation of diabetes susceptibility on the NOD background. Many genes likely contribute to diabetes by directing the autoimmune reaction toward the β-cell target. As for mechanisms involved in β-cell damage, T-cells and cytokines are in the front line in type 1 diabetes, while hyperglycemia, increased free fatty acids, and proinflammatory cytokines contribute to β-cell damage in type 2 diabetes, including modified insulin secretion and reduced β-cell survival. Some of the pathways that contribute to insulin resistance are shared by pathways controlling β-cell function and survival, namely insulin receptor substrate-2, nuclear factor-κB, endoplasmic reticulum stress, and mitochondrial dysfunction. There is some controversy, however, in the extent of overlap of mechanisms involved in type 1 and type 2 diabetes. Cytokines and nutrients may trigger β-cell death by different mechanisms, with differential contribution of the nuclear factor-κB pathway. Interleukin-1β, in addition to TNF-α and interferon-γ, induces β-cell apoptosis in type 1 diabetes. But its production by β-cells under high glucose concentrations in type 2 diabetes is debated. In any case, the role of hyperglycemia and “adipocyte byproducts” in reducing β-cell mass in type 2 diabetes only comes in addition to a functional intrinsic primary defect of insulin secretion in type 2 diabetes, while β-cell destruction is central to the type 1 diabetes process. The controversies regarding these questions are exemplified by the articles of M. Donath and D.L. Eizirik, presented in this section.

Section IV raises the issue of the relative contribution of genetic and environmental factors in both type 1 and type 2 diabetes. Evidence that supports a critical role of exogenous factors in the development of type 1 diabetes was first developed. The trigger-booster hypothesis claiming that the diabetic disease process is triggered by an exogenous factor with definite seasonal variation and driven by one or several other environmental determinants is further discussed. In type 2 diabetes, the role of environmental factors is important at least for the increase in β-cell demand imposed by insulin resistance and obesity. The gerbil Psammomys obesus is a good model of gene versus environment interactions in type 2 diabetes. On a high-calorie diet, P. obesus develops moderate obesity and postprandial hyperglycemia. Continued dietary load combined with innate insulin resistance results in inappropriate preproinsulin gene response to increased needs and consequently in depletion of insulin stores. Another example of the role of the environment is the association of fetal or infant growth defects and diabetes. In the Hertfordshire Cohort Study, birth weight, but not weight at 1 year of age, is inversely related to the overall prevalence of diabetes in men and women. Higher insulin and glucose concentrations are seen in people who were small at birth or during infancy. In humans, both type 1 and type 2 diabetes exemplify genetically heterogeneous diseases in which epigenetic factors add risk to the underlying genetic susceptibility. The familial clustering of type 1 and type 2 diabetes suggests, as developed in Section I, the contribution of common genes, or genetic regions, in both forms of diabetes. Systemically expressed genes regulating β-cell ability to withstand chronic diabetogenic stress may represent a component of shared susceptibility to both major disease forms. But crossing of animal models developing either form of diabetes shows quite distinct genetic bases. However, limitations in the spectrum of diabetic phenotypes available in animal models leave open the possibility of overlap in the genetic backgrounds that may be at play in humans.