Insulin autoantibody polymorphisms with greater discrimination for diabetes in humans

Advances in Type 1 Diabetes Prediction Using Islet Autoantibodies: Beyond a Simple Count

Islet autoantibodies are key markers for the diagnosis of type 1 diabetes. Since their discovery, they have also been recognized for their potential to identify at-risk individuals prior to symptoms. To date, risk prediction using autoantibodies has been based on autoantibody number; it has been robustly shown that nearly all multiple-autoantibody-positive individuals will progress to clinical disease. However, longitudinal studies have demonstrated that the rate of progression among multiple-autoantibody-positive individuals is highly heterogenous. Accurate prediction of the most rapidly progressing individuals is crucial for efficient and informative clinical trials and for identification of candidates most likely to benefit from disease modification. This is increasingly relevant with the recent success in delaying clinical disease in presymptomatic subjects using immunotherapy, and as the field moves toward population-based screening. There have been many studies investigating islet autoantibody characteristics for their predictive potential, beyond a simple categorical count. Predictive features that have emerged include molecular specifics, such as epitope targets and affinity; longitudinal patterns, such as changes in titer and autoantibody reversion; and sequence-dependent risk profiles specific to the autoantibody and the subject's age. These insights are the outworking of decades of prospective cohort studies and international assay standardization efforts and will contribute to the granularity needed for more sensitive and specific preclinical staging. The aim of this review is to identify the dynamic and nuanced manifestations of autoantibodies in type 1 diabetes, and to highlight how these autoantibody features have the potential to improve study design of trials aiming to predict and prevent disease.

Developments in the prediction of type 1 diabetes mellitus, with special reference to insulin autoantibodies

The prodromal phase of type 1 diabetes is characterised by the appearance of multiple islet-cell related autoantibodies (Aab). The major target antigens are islet-cell antigen, glutamic acid decarboxylase (GAD), protein-tyrosine phosphatase-2 (IA-2) and insulin. Insulin autoantibodies (IAA), in contrast to the other autoimmune markers, are the only beta-cell specific antibodies. There is general consensus that the presence of multiple Aab (> or = 3) is associated with a high risk of developing diabetes, where the presence of a single islet-cell-related Aab has usually a low predictive value. The most commonly used assay format for the detection of Aab to GAD, IA-2 and insulin is the fluid-phase radiobinding assay. The RBA does not identify or measure Aab, but merely detects its presence. However, on the basis of molecular studies, disease-specific constructs of GAD and IA-2 have been employed leading to somewhat improved sensitivity and specificity of the RBA. Serological studies have shown epitope restriction of IAA that can differentiate diabetes-related from unrelated IAA, but current assays do not distinguish between disease-predictive and non-predictive IAA or between IAA and insulin antibodies (IA). More recently, phage display technology has been successful in identifying disease-specific anti-idiotopes of insulin. In addition, phage display has facilitated the in vitro production of antibodies with high affinity. Identification of disease-specific anti-idiotopes of insulin should enable the production of a high affinity reagent against the same anti-idiotope. Such a development would form the basis of a disease-specific radioimmunoassay able to identify and measure particular idiotypes, rather than merely detect and titrate IAA.

Epitope analysis of insulin autoantibodies using recombinant Fab

Autoantibodies to insulin are often the first autoantibodies detected in young children with type 1 diabetes and can be present before the onset of clinical diabetes. These autoantibodies and their epitopes are, however, not well characterized. We explored the use of monoclonal antibodies and their recombinant Fab as reagents for epitope analysis. In this study we cloned and characterized the recombinant Fab of the insulin-specific monoclonal antibody CG7C7. We found the epitope of this antibody to be located predominantly at the A-chain loop of the insulin molecule. The recombinant Fab was then used to compete for insulin binding against insulin autoantibodies present in sera from patients with type 1 or type 1.5 diabetes. In competition experiments with sera positive for autoantibodies to insulin the recombinant Fab significantly reduced the binding to [125I]-insulin by sera of type 1 (n = 35) and type 1.5 diabetes [latent autoimmune diabetes in adults (LADA)] (n = 14) patients (P < 0.0001). We conclude that competition between insulin-specific monoclonal antibodies or their recombinant Fab and insulin autoantibodies should prove useful in the epitope analysis of autoantibodies to insulin.

Mature high-affinity immune responses to (pro)insulin anticipate the autoimmune cascade that leads to type 1 diabetes

Children at risk for type 1 diabetes can develop early insulin autoantibodies (IAAs). Many, but not all, of these children subsequently develop multiple islet autoantibodies and diabetes. To determine whether disease progression is reflected by autoantibody maturity, IAA affinity was measured by competitive radiobinding assay in first and subsequent IAA-positive samples from children followed from birth in the BABYDIAB cohort. IAA affinity in first positive samples ranged from less than 10(6) l/mol to more than 10(11) l/mol. High affinity was associated with HLA DRB1*04, young age of IAA appearance, and subsequent progression to multiple islet autoantibodies or type 1 diabetes. IAA affinity in multiple antibody-positive children was on average 100-fold higher than in children who remained single IAA positive or became autoantibody negative. All high-affinity IAAs required conservation of human insulin A chain residues 8-13 and were reactive with proinsulin. In contrast, most lower-affinity IAAs were dependent on COOH-terminal B chain residues and did not bind proinsulin. These data are consistent with the concept that type 1 diabetes is associated with sustained early exposure to (pro)insulin in the context of HLA DR4 and show that high-affinity proinsulin-reactive IAAs identify children with the highest diabetes risk.

The molecular specificity of insulin autoantibodies

Insulin autoantibodies (IAA) are one of several markers for Type I (autoimmune) diabetes, but alone deserve special attention. Unlike the other markers, their ligand is unique to the beta cell. IAA are the first markers to appear during the symptomless period which precedes diabetes and they are present in the vast majority of young children destined to develop diabetes. The primary and tertiary structures of insulin have been known for decades. Binding studies with insulin variants have shown epitope restriction that can distinguish Type 1 diabetes-predictive from non-predictive IAA-positive sera, thereby improving specificity for the test. With two major international Type 1 diabetes prevention trials underway, there is a pressing need to refine markers that reliably indicate the presence of, and remission from, autoimmune insulitis. The binding regions of antibodies are assembled from three multi-gene families, and some of their diversity derives from random mutation during their antigen-driven maturation. There is evidence that mature IAA derive from germline-encoded 'natural' antibodies, and that the gene segments utilised by IAA may be influenced by clinical context. Monoclonal anti-idiotypic (anti-Id) antibodies can serve as probes for antibody variable region determinants, and antibodies to the different epitopes of beef and porcine insulins have already been analysed with monoclonal reagents. Used as antibodies in a radioimmunoassay format, monoclonal anti-Ids will identify and measure autoantibody idiotopes as if they were ligands. The challenge now is to replace the conventional radiobinding assays for IAA, which only detect and titrate, with radioimmunoassays that can be standardised in absolute units. There is sufficient evidence for the existence of Type 1 diabetes-predictive IAA idiotopes to justify the development of idiotope-specific radioimmunoassays which ignore Type 1 diabetes-unrelated IAA.

Relationship of the 37,000- and 40,000-M(r) tryptic fragments of islet antigens in insulin-dependent diabetes to the protein tyrosine phosphatase-like molecule IA-2 (ICA512)

Sera from patients with insulin-dependent diabetes immunoprecipitate 64,000-M(r) proteins, distinct from glutamate decarboxylase, that are cleaved to 37,000- and 40,000-M(r) fragments by trypsin. We investigated possible relationships between 37,000- or 40,000-M(r) fragments of antigen and the tyrosine phosphatase-like protein, IA-2 (ICA512). Antibodies from nondiabetic relatives bound differentially to 37,000- and 40,000-M(r) fragments indicating presence of distinct epitopes. Precursors of these fragments could be separated on immobilized lectins, suggesting different carbohydrate content. Levels of antibodies to 40,000-M(r) fragments were strongly associated with those to the intracellular domain of IA-2. Recombinant intracellular domain of IA-2 blocked binding of antibodies to 40,000-M(r) fragments expressed by insulinoma cells and partially blocked binding to 37,000-M(r) fragments. Furthermore, trypsinization of recombinant intracellular domain of IA-2 generated proteolytic fragments of identical M(r) to the 40,000-M(r) fragments of insulinoma antigen; 37,000-M(r) fragments were not generated. Thus, 40,000-M(r) fragments of islet autoantigen are derived from a protein similar or identical to the tyrosine phosphatase-like molecule, IA-2. The 37,000-M(r) fragments are derived from a different, although related, protein.

Influence of affinity of antibodies upon their detection by liquid phase radiobinding assay and solid phase enzyme linked immunosorbent assay. Demonstration using monoclonal antibodies raised against rDNA human proinsulin

Hybridomas producing proinsulin antibodies were cloned by limiting dilution of cell cultures obtained by fusion of splenocytes of immunized mice with immortal myeloma cells. Some proinsulin monoclonal antibodies crossreacted with labelled insulin but none did with labelled C-peptide indicating that the involved epitopes were at one of the insulin/C-peptide junctions or included in the insulin moiety. Hybridoma supernatants were assayed for IgG concentration by a solid phase assay and for ligand binding by a radiobinding assay and an enzyme linked immunosorbent assay. The half-life of immune complexes formed with radioligand was measured and, as expected, correlated with affinity as measured by the method of Scatchard. Antibody titres determined by enzyme linked immunosorbent assay did not correlate to those measured by radiobinding assay. IgG concentration correlated to enzyme linked immunosorbent assay titres but not to radiobinding assay titres. Finally, a significant correlation was found between radiobinding assay titre and the product of enzyme linked immunosorbent assay titre by the period of immune complexes. It is concluded that, except for very low affinity antibodies, enzyme linked immunosorbent assay is a capacity assay whereas radiobinding assay is influenced by both antibody concentration and affinity. The former assay is thus best suited to detecting low affinity antibodies whereas the latter is more efficient in the presence of low levels of high affinity antibodies.

The relationship between insulin autoantibodies and islet cell histology in the diabetes prone BB rat

The relationship between insulin autoantibodies (IAA) and pancreatic islet cell histology was examined in 71 diabetes prone BB rats from the Toronto colony. Twenty-seven of the 71 became diabetic and of these, 18 (67%) were IAA positive by ELISA. IAA were also detected in 39/44 (89%) which did not develop diabetes, but in none of six control animals at 50-140 days of age. All 27 which became diabetic showed some evidence of lymphocytic infiltration scored + to ++++ histometrically and 26/27 evidence of beta cell degranulation. The frequency of diabetes increased with both intensity of insulitis and degree of beta cell degranulation, but there was no correlation between either and IAA. IAA are a marker for the BB strain of Wistar rat, but do not correlate with islet cell histology and do not predict clinical diabetes.

Clonally restricted insulin autoantibodies in a cohort of 2200 healthy blood donors

Serum samples of 2200 blood donors were screened for anti-insulin IgG by enzyme-linked immunosorbent assay. Specificity of detected antibodies was verified by competition with an excess of insulin and observation that saturated anti-insulin IgG were no longer measurable. The upper limit of measured signal in the population was defined as the mean plus three SD. In the direct assay, 32 sera were positive. Among these, 22 (1%) contained saturable insulin antibodies (true positive) and 10 were non-saturable and considered as non-insulin-specific. The positive blood donors were requested to answer a questionnaire and according to their answers, none had ever received insulin, was a first degree relative of a Type 1 (insulin-dependent) diabetic patient or had experienced a hypoglycaemic episode. None of the 22 true positive sera detected by enzyme-immunosorbent assay bound 125-I-insulin to any significant extent. The nine sera yielding the highest signal were further characterized with regard to heavy and light chains as well as species specificity of ligand. Anti-insulin IgG from healthy blood donors contained only one heavy (gamma 1 2/9;gamma 3 7/9) and one light (kappa 8/9); lambda 1/9) chain. Three sera were human insulin specific; two were non-species-specific; the other four bound insulin in the order human = porcine greater than bovine. These results indicate that low affinity clonally restricted antibodies against insulin are present in unselected blood donors with a prevalence of 1%.

Antibody markers for type 1 diabetes: the issues

The detection of islet-cell-specific autoantibodies before clinical onset may mean that diabetes mellitus can be predicted, with important implications for targeting those to whom preventive treatment might be given. A large number of such markers have been proposed, but only high-titer islet cell antibodies have so far demonstrated the specificity and sensitivity to be of potential clinical value.