Chapter 1
Etiology of diabetes mellitus

Ravichandran Ramasamy, PhD and Ann Marie Schmidt, MD

Introduction

The defining characteristics of diabetes, irrespective of the precise etiology, relate to the presence of hyperglycemia. The American Diabetes Association (ADA) has set forth specific criteria for the definition of diabetes. In the ADA guidelines, the following are necessary for the diagnosis of diabetes: (1) hemoglobin A1c (HbA1c) equal to or greater than 6.5% OR (2) fasting plasma glucose equal to or greater than 126 mg/dl OR (3) two-hour plasma glucose equal to or greater than 200 mg/dl during an oral glucose tolerance test (OGTT) (glucose load containing 75 grams anhydrous glucose dissolved in water) OR (4) in a patient with classic symptoms of diabetes or during a hyperglycemic crisis, a random glucose of equal to or greater than 200 mg/dl suffices to diagnose diabetes [1].

In this chapter, we will review the major types of diabetes and the etiologic factors that are known to or are speculated to contribute to these disorders. Furthermore, we will take the opportunity to present an overview of emerging theories underlying the pathogenesis of type 1 and type 2 diabetes. Types 1 and 2 diabetes constitute the vast majority of diabetes cases. Interestingly, both of these types of diabetes are on the rise worldwide [2, 3]. In addition to types 1 and 2 diabetes, we will also discuss gestational diabetes. Often a harbinger to the ultimate development of frank type 2 diabetes in the mother, this form of diabetes is potentially dangerous to both the mother and the developing fetus. Finally, we will discuss the syndromes known as MODY or maturity onset diabetes of the young. The disorders underlying MODY have very strong genetic components and are due to mutations in multiple distinct genes.

The greatest long-term danger of diabetes, irrespective of the etiology, lies in the potential for complications. The complications of the disease are insidious, deadly, and difficult to treat or reverse; hence, there is great urgency to identify specific means to prevent or mitigate these most common types of diabetes.

Type 1 diabetes

Type 1 diabetes accounts for approximately 5–10% of all cases of diabetes [1]. The countries with the highest incidence of type 1 diabetes include Finland and Sardinia [4]. Type 1 diabetes is usually diagnosed in childhood, hence the original classification “juvenile onset diabetes.” Indeed, type 1 diabetes accounts for more than 90% of diabetes diagnosed in children and adolescents. Given that the disease is often diagnosed in adults, however, even into advanced age, the term “type 1 diabetes” has been adopted to more accurately reflect the diversity of affected ages. In type 1 diabetes, the primary etiology is due to a cellular-mediated autoimmune-mediated destruction of the β cells of the pancreas. Traditionally, in subjects with type 1 diabetes, autoantibodies may be detected that reflect the underlying attack against these cells [5]. These include autoantibodies to insulin, to GAD65, and to IA-2 and IA-2β (the latter two are tyrosine phosphatases). These antibodies are often detected up to years before the diagnosis of type 1 diabetes [6]. In most subjects with type 1 diabetes, one or more of these antibodies is evident. Indeed, in vulnerable subjects, such as first-degree relatives of affected individuals, the presence of these autoantibodies is often, but not always, a harbinger of the eventual diagnosis of diabetes. Hence, these antibody profiles may be used to predict the risk of diabetes in the siblings and relatives of affected subjects with type 1 diabetes [6].

Genetics of type 1 diabetes

More than forty years ago, type 1 diabetes was found to have very strong links to the human leukocyte antigen (HLA)-encoding genes [7]. The largest study to address this issue was known as the Type 1 Diabetes Genetics Consortium (T1DGC). This group was composed of an international collaboration and amassed more than 14,000 samples [8]. By far, the greatest association to type 1 diabetes was found in the HLA, particularly in the HLA DR-DQ haplotypes. Furthermore, other genes found to have strong genetic association were in polymorphisms identified in the insulin gene [9]. The researchers of T1DGC earlier reported that beyond these two associations, two other loci were found to have odds ratios (ORs) greater than 1.5, and included PTPN22 and IL2RA [9]. However, the ORs for these genes were relatively much lower than that of the HLA region, consistent therefore with the overall strong role of the HLA in the susceptibility to type 1 diabetes.

A number of groups have published the results of genome wide association studies (GWAS) in type 1 diabetes and identified more than 40 potential susceptibility loci in the disease [11]. Candidate genes identified in this approach included those encoding IL10, IL19, IL20, GLIS3, CD69, and IL27; these are all genes strongly linked to the immune/inflammatory response [10]. In their report, Bergholdt and colleagues integrated the data from these GWAS studies and translated them to a more functional level, that is protein-protein interactions and, finally, they tested their relevance in human islets and in a β cell line, INS-1 cells (rat insulimona-derived cells) [11]. First, they performed a meta-analysis of the type 1 diabetes genome wide Association studies that were available. From these, they identified 44 type 1 diabetes non-major histocompatibility complex (MHC) low density (LD) regions with significance; these regions contained more than 395 candidate genes. They then performed network analysis studies with the intention to more deeply probe network connections and protein-protein interactions. From this work, 17 protein networks were identified (which contained 235 nodes) containing at least two genes from different type 1 diabetes LD regions [11].

To follow up on these findings, human islets were exposed to pro-inflammatory cytokines and comparisons were made between the treated and untreated human islets (retrieved from eight donors). From this, the following genes were found to be significantly impacted by the cytokine stimulation in the human islets: IL17RD, CD83, IFNGR1, TRAF3IP2, IL27RA, PLCG2, MYO1B, and CXCR7. Interestingly, the study design suggested that perhaps these traditionally inflammation-associated factors were being produced by pancreatic β cells and not necessarily solely by immune cells. To test this specific point, rat INS-1 cells were treated with cytokines and the above eight genes were examined. Indeed, all but IL27ra were identified in the stimulated INS-1 cells [11]. In the case of cultured INS-1 cells, no immune cells are present, therefore suggesting the interesting possibility that these factors may be produced both by islet β cells themselves as well, likely, by infiltrating inflammatory cells. Examples of non-HLA genes linked to type 1 diabetes are illustrated in Table 1.1.

Table 1.1 Examples of non-HLA type 1 diabetes-associated loci.

Pathogenesis of type 1 diabetes

There is strong evidence that links the pathogenesis of type 1 diabetes to immune-mediated mechanisms of β cell destruction, including the detection of insulitis, the presence of islet cell autoantibodies, activated β cell-specific T lymphocytes and, as considered above, association of the disease with a restricted set of class II major histocompatibility alleles [12]. Importantly, the rate of the development of type 1 diabetes after the appearance of autoantibodies may be quite variable, reflecting perhaps the contribution of protective mechanisms (such as CD4 + -T regulatory cells and other regulatory cells such as invariant natural killer T...