Insulin Mutations in Diabetes : The Clinical Spectrum

Studies of monogenic disorders of β-cell function have yielded important information on β-cell physiology and have improved the diagnosis and treatment of patients with these rare diseases. These disorders include defects associated with increased insulin secretion, causing hypoglycemia, and decreased insulin secretion, resulting in diabetes. The most common form of monogenic diabetes is so-called maturity-onset diabetes of the young (MODY) syndrome, causing autosomal dominant non–insulin dependent diabetes appearing before the age of 25 years. Mutations in six genes can cause MODY, although in 16–45% of cases the genetic etiology is still unknown (1). Neonatal diabetes mellitus (NDM) is another form of monogenic diabetes, usually defined as overt diabetes diagnosed during the first 6 months of life (2). The disease is rare (incidence 1:300,000–500,000 live births), and ∼50% of patients have transient NDM (TNDM), in which the disease remits after a few months but may reappear months or years later. The other 50% have permanent NDM (PNDM). In a small percentage of these, diabetes is part of a complex clinical syndrome involving organs other than the endocrine pancreas. For those with isolated PNDM, mutations in four different genes have been identified. Inactivating glucokinase mutations were discovered first, but appear to be rather rare causes of this syndrome (3). Activating mutations in either of the two subunits of the β-cell ATP-sensitive K⁺ channel, ABCC8 and KCNJ11, are more frequently seen (4,5).

In a landmark report last year, Stoy et al. (6) described 16 patients with PNDM caused by insulin gene mutations. These mutations appear to cause abnormal protein folding resulting in endoplasmic reticulum (ER) stress, which activates apoptosis pathways leading to β-cell death. Stoy et al. suggested the need for additional studies in larger patient populations to discover the true incidence and clinical spectrum associated with insulin mutations.

In this issue of Diabetes, three groups present their findings after screening different patient populations for INS mutations. Polak et al. (7) found three missense mutations in 38 PNDM patients and one in a patient with childhood-onset nonautoimmune diabetes. Further screening identified three affected relatives. Molven et al. (8) expanded the spectrum of disease associated with INS mutations by screening patients with diabetes onset well after the neonatal period. They identified one mutation in 92 patients with the MODY phenotype and one in 124 patients with autoantibody-negative type 1 diabetes, but none in 99 patients with autoantibody-positive familial type 1 diabetes. Edghill et al. (9) identified 32 mutation carriers among 279 patients with PNDM diagnosed before the age of 6 months, 2 more among 86 patients diagnosed between 6 and 12 months, and none in 58 patients diagnosed between 12 and 24 months of age. In addition, they identified one affected individual among 296 patients with MODY and one in 463 patients with young-onset type 2 diabetes. Taken together, these studies confirm that the vast majority of patients with INS mutations present with severe insulin-dependent diabetes within the first 6 months of age. However, a small minority present as late as 20 years of age and can resemble MODY, early-onset type 2 diabetes, or childhood-onset type 1 diabetes. Similarly, the clinical severity can vary from severe intrauterine insulin deficiency, causing low birth weight, to diet-responsive diabetes phenotypically indistinguishable from type 2 diabetes.

Reviewing the published data from the last several years, the genetic etiology of TNDM and PNDM is becoming clearer (2,10). Currently, the precise genetic cause of TNDM can be determined in almost all patients, whereas for ∼40% of patients with PNDM the genetic etiology is still unknown (Fig. 1). These studies clearly demonstrate that all patients with diabetes onset before the age of 6 months should be investigated for monogenic diabetes, as autoimmune diabetes is exceedingly rare in this age-group. Patients with autoantibody-negative diabetes diagnosed after the age of 6 months may also warrant genetic investigation, although the chance of identifying a disease-causing mutation is still low.

These are not the first insulin gene mutations to be discovered. In the early 1980s, shortly after the discovery of the insulin gene sequence (11), several patients with insulin gene mutations were identified. Some mutations affected insulin-proinsulin processing, resulting in secretion of large amounts of partially processed proinsulin (12–15), whereas others appeared to produce normally processed insulin with subnormal biological activity (16–18). For most of these, there was a clear association between the mutation and marked hyperinsulinemia or proinsulinemia, but not with diabetes, as some mutation carriers were normoglycemic or had variable degrees of glucose intolerance.

It is theoretically possible that other insulin gene mutations could result in increased function and thus hypoglycemia, as has been found for glucokinase (19). To study this, Edghill et al. (9) screened 49 patients with hyperinsulinemic hypoglycemia, but found no INS mutations, suggesting that if such mutations occur they must either be very rare or result in a phenotype that was not included in their cohort of patients with hyperinsulinism of infancy.

Data available today suggest that the insulin gene locus does not contribute significantly to the overall genetic risk of developing type 2 diabetes, although it should be emphasized that the majority of published studies were designed to identify common variants associated with disease and thus cannot exclude the possibility that a large number of rare variants within the gene could contribute significantly to the genetic risk of type 2 diabetes.

Thus, mutations in the insulin gene can cause a spectrum of clinical phenotypes. Mutations affecting receptor binding or postgolgi processing result in elevated serum levels of insulin, proinsulin, or proinsulin split products without necessarily causing hyperglycemia. Mutations affecting ER processing of insulin, in contrast, appear to result in loss of β-cell mass through apoptosis. The severity of the defect, as well as other factors that define the β-cell’s capacity to cope with ER stress, determine the rate of β-cell death and thus the age of onset and severity of clinical disease. At least in some cases, a “window of opportunity” may exist during which therapeutic intervention, aimed at decreasing ER stress or its effect on β-cell apoptosis, may be possible, thus preserving β-cell function and severe insulin-deficient diabetes.

The importance of identifying specific genetic causes of monogenic diabetes is twofold. First, these findings improve our understanding of β-cell physiology and may provide important information leading to a better understanding of type 2 diabetes. Second, for the individual, the identification of the specific genetic cause of disease may have major therapeutic implications. For example, some patients with KCNJ11 or ABCC8 mutations can be treated with oral sulfonylurea drugs, obviating the need for multiple insulin injections and intensive blood glucose monitoring while improving glycemic control and hopefully preventing or delaying long-term complications (20). For patients with INS mutations the therapeutic implications are less obvious, although it may be possible to develop therapies that increase the relative production of the normal insulin, improve the folding of the abnormal peptide, or increase the cell’s tolerance for ER stress.