Thygeson superficial punctate keratitis (TSPK) is clinically characterized by exacerbations and remissions of gray-white opacities within the corneal epithelium, most often bilateral but may be asymmetric. Symptoms typically include photophobia, tearing, blurring, and eye irritation. Although disease progression and prognosis are well described, the exact cause is unknown. Hypotheses exist implicating virus-mediated immunity as the cause of TSPK following cases of viral keratitis; however, several polymerase chain reaction studies refute the infectious process concurrently with symptomatic TSPK. This is further supported by the consistent lack of response to antiviral and antibacterial treatment. A subset of dendritic cells known as Langerhans cells (LC) found within the corneal epithelium has been positively correlated with exacerbations of TSPK. Langerhans cells proliferate to protect and mitigate the cornea's inflammatory response, but the inflammatory triggers and relapses associated with TSPK are not well understood. Several topical drugs exist to treat inflammation related to TSPK; however, drug delivery is a major barrier to treatment because of the tear film and epithelial barrier. Drug-eluting contact lenses that target intermediates of inflammation could serve as a more effective treatment modality because of the increased bioavailability of the drugs. This review is an in-depth survey of the literature regarding the relationship between the origin and pathophysiology of LC and TSPK at the immunologic level. We also discuss potential pharmacotherapeutic interventions for TSPK prevention and treatment.

HLA-DR4, a common antigen of HLA-DRB1, has multiple subtypes that are strongly associated with risk of type 1 diabetes (T1D); however, some are risk neutral or resistant. The pathobiological mechanism of HLA-DR4 subtypes remains to be elucidated.

We used a population-based case-control study of T1D (962 patients and 636 controls) to decipher genetic associations of HLA-DR4 subtypes and specific residues with susceptibility to T1D. Using a birth cohort of 7865 children with periodically measured islet autoantibodies (GADA, IAA or IA-2A), we proposed to validate discovered genetic associations with a totally different study design and time-to-seroconversions prior to clinical onset of T1D. A novel analytic strategy hierarchically organized the HLA-DRB1 alleles by sequence similarity and identified critical amino acid residues by minimizing local genomic architecture and higher-order interactions.

Three amino acid residues of HLA-DRB1 (β71, β74, β86) were found to be predictive of T1D risk in the population-based study. The "KAG" motif, corresponding to HLA-DRB1×04:01, was most strongly associated with T1D risk ([O]dds [R]atio=3.64, p = 3.19 × 10-64). Three less frequent motifs ("EAV", OR = 2.55, p = 0.025; "RAG", OR = 1.93, p = 0.043; and "RAV", OR = 1.56, p = 0.003) were associated with T1D risk, while two motifs ("REG" and "REV") were equally protective (OR = 0.11, p = 4.23 × 10-4). In an independent birth cohort of HLA-DR3 and HLA-DR4 subjects, those having the "KAG" motif had increased risk for time-to-seroconversion (Hazard Ratio = 1.74, p = 6.51 × 10-14) after adjusting potential confounders.

DNA sequence variation in HLA-DRB1 at positions β71, β74, and β86 are non-conservative (β74 A→E, β71 E vs K vs R and β86 G vs V). They result in substantial differences in peptide antigen anchor pocket preferences at p1, p4 and potentially neighboring regions such as pocket p7. Differential peptide antigen binding is likely to be affected. These sequence substitutions may account for most of the HLA-DR4 contribution to T1D risk as illustrated in two HLA-peptide model complexes of the T1D autoantigens preproinsulin and GAD65.

National Institute of Diabetes and Digestive and Kidney Diseases and the Swedish Child Diabetes Foundation and the Swedish Research Council.

Neonatal diabetes (ND) appears during the first months of life and is caused by a single gene mutation. It is heterogenous and very different compared to other forms of multi-factorial or polygenic diabetes. Clinically, this form is extremely severe, however, early genetic diagnosis is pivotal for successful therapy. A large palette of genes is demonstrated to be a cause of ND, however, the mechanisms of permanent hyperglycemia are different. This review will give an overview of more frequent genetic mutations causing ND, including the function of the mutated genes and the specific therapy for certain sub-forms.

Single nucleotide polymorphisms (SNPs) are usually the most frequent genomic variants. Directly pedigree-phased multi-SNP haplotypes provide a more accurate view of polymorphic population genomic structure than individual SNPs. The former are, therefore, more useful in genetic correlation with subject phenotype. We describe a new pedigree-based methodology for generating non-ambiguous SNP haplotypes for genetic study. SNP data for haplotype analysis were extracted from a larger Type 1 Diabetes Genetics Consortium SNP dataset based on minor allele frequency variation and redundancy, coverage rate (the frequency of phased haplotypes in which each SNP is defined) and genomic location. Redundant SNPs were eliminated, overall haplotype polymorphism was optimized and the number of undefined haplotypes was minimized. These edited SNP haplotypes from a region containing HLA-DRB1 (DR) and HLA-DQB1 (DQ) both correlated well with HLA-typed DR,DQ haplotypes and differentiated HLA-DR,DQ fragments shared by three pairs of previously identified megabase-length conserved extended haplotypes. In a pedigree-based genetic association assay for type 1 diabetes, edited SNP haplotypes and HLA-typed HLA-DR,DQ haplotypes from the same families generated essentially identical qualitative and quantitative results. Therefore, this edited SNP haplotype method is useful for both genomic polymorphic architecture and genetic association evaluation using SNP markers with diverse minor allele frequencies.

Type 1 diabetes mellitus (T1DM) has increased dramatically over the last two decades with global variation greater than 350-fold difference reflecting the ethnic, racial, and geographical variation. Diabetic patients remain at a higher risk of cardiovascular mortality than those without diabetes. Therefore, it is vital for clinicians to have in-depth knowledge of T1DM statistics and their impact on people health and health resources.

This review will cover the epidemiologic characteristics of T1DM and the influence of race, ethnicity, and geographical variation on the incidence and the outcome. The minority populations health disparities in the clinical presentation and outcomes among youth with T1DM, the long-term glycemic control patterns in racially and ethnically diverse youth, and the long-term influence of these factors on cardiovascular outcomes will be elucidated. The PubMed database was searched using the terms: T1DM ± incidence, Race, ethnicity, and Genetic.

Understanding the epidemiological characteristics of T1DM including race, ethnicity and the genetic predisposition will help to develop guidelines target these higher risk patients of an unfavorable outcome. Further research and interventional strategies to identify infants at genetic risk of T1DM may help to prevent, stop or retard the destructive autoimmune process leading to T1DM.

Type 1 diabetes (T1D) is a complex autoimmune disorder that results from the T cell-mediated destruction of the pancreatic β cells and is due to interactions between environmental and genetic factors. Although Arabs have one of the highest global incidence and prevalence rates of T1D, unfortunately, there is a dearth of information regarding the genetic epidemiology of T1D in the Arab world. Arabs share several HLA haplotypes with other ethnic groups, which confer either susceptibility or protection to T1D, but they have specific haplotypes that are distinctive from other ethnicities. Among different Arab countries, several non-HLA genes were reported to be associated with susceptibility to T1D, including CTLA4, CD28, PTPN22, TCRβ, CD3z, IL15, BANK1, and ZAP70. In Arab countries, consanguinity, endogamy, and first-cousin marriage rates are some of the highest reported worldwide and are responsible for the creation of several inbreeding communities within the Arab world that have led to an increase in homozygosity of both the HLA haplotypes and non-HLA genes associated with either protection or susceptibility to T1D among Arabs. Homozygosity reduces the HLA complexity and is expected to facilitate our understanding of the mode of inheritance of HLA haplotypes and provide valuable insight into the intricate genotype-phenotype correlations in T1D patients. In this review, based on literature studies, I will discuss the current epidemiological profile and molecular genetic risks of Arabs with T1D.

This study evaluated whether people with ankylosing spondylitis (AS) and spondyloarthritis are at higher risk of type 2 diabetes mellitus (T2DM). We used a sub-dataset of the National Health Insurance Research Database from 1996 to 2010 to established a AS cohort consisting new patients with AS or spondyloarthritis (N = 7,778) and a non-AS cohort without the diseases (N = 31,112). Incidences of T2DM in the two cohorts, hazard ratios (HRs) of risk of T2DM in association with AS, and cumulative probability of having T2DM were estimated by the end of 2010. The incidence of T2DM was 1.17-fold higher in the AS cohort than in the non-AS cohort (13.5 vs. 11.5, per 1,000 person-years), with an adjusted HR of 1.16 (95 % CI = 1.05-1.29). The T2DM incidence was higher for women than for men; while the Cox model measured sex-specific adjusted HR of T2DM was higher for men than for women. The incidence rate of T2DM increased with age in both cohorts, while the age-specific measures showed that the adjusted HR of T2DM was higher in young AS patients (≤50 years of age) than older ones, compared to their peers of non-AS group. The plot of Kaplan-Meier analysis showed that the overall probability of having T2DM was 2 % higher in the AS cohort than in the non-AS cohort (log-rank test: p < 0.0001). Patients with AS and spondyloarthritis have an increased risk of developing T2DM.

The pancreatic B-cell function (glucose tolerance, C-peptide release) and organ-specific autoantibodies, including islet cell cytoplasmic and cell surface (mouse), were studied in 45 first-degree relatives of patients with insulin-dependent diabetes mellitus diagnosed before the age of 30 years. Compared to 107 healthy persons without any family history of either insulin-dependent or non-insulin-dependent diabetes mellitus, the prevalence of autoantibodies was increased among the relatives. The prevalence of islet cell antibodies did not differ between relatives and controls and none of the individuals had complement-fixing islet cell antibodies. There was no difference in glucose tolerance or C-peptide release between relatives and controls, whether they had autoantibodies or not. At a three-year follow-up, none of the individuals had developed insulin-dependent diabetes.

Ketosis-prone diabetes (KPD) comprises four subgroups based on the presence or absence of beta-cell autoantibodies (A+ or A-) and beta-cell functional reserve (beta+ or beta-). Genetic factors could contribute to their distinctive phenotypes. Our aim was to specify the role of HLA class II alleles associated with susceptibility or resistance to autoimmune type 1 diabetes in determining KPD phenotypes.

A total of 185 adults presenting with diabetic ketoacidosis were followed longitudinally for a mean of 5.5 years, with measurements of autoantibodies, beta-cell functional reserve, insulin sensitivity, and insulin requirement. Frequencies of susceptibility and resistance alleles at HLA DQA1, DQB1, and DRB1 loci were correlated with clinical and phenotypic features of KPD subgroups and compared with those of ethnic-specific population control subjects.

Susceptibility alleles were more frequent (P < 0.0001) in the two A+ than the two A- KPD subgroups; in the latter, the frequency was no greater than in population control subjects (except for DQB1*0302). Susceptibility alleles differentiated the two clinically similar beta- subgroups (more frequent in A+beta- than A-beta- KPD; P < 0.01). Resistance alleles were more frequent in the two beta+ than the two beta- KPD subgroups (P < 0.01). The frequencies of certain susceptibility (e.g., DQB1*02) and resistance (DQB1*0602) alleles were higher in African-American A-beta+ KPD patients than in African-American control subjects. DQB1*0302 was more frequent in all KPD subgroups compared with control subjects.

HLA class II alleles associated with susceptibility or resistance to autoimmune type 1 diabetes help specify the four subgroups of KPD. Inheritance of these alleles may influence long-term beta-cell functional reserve.

CD8(+) T cells play an important role in the development of type 1 diabetes (T1D) in NOD mice and humans. IGRP (islet-specific glucose-6-phosphatase catalytic subunit-related protein) has emerged in recent years as a major antigen in NOD mice. Therefore, we aimed to determine if IGRP is an antigen in T1D patients and to identify the HLA-A2-restricted IGRP epitopes targeted. Using IFN-gamma ELISPOT assay, we tested PBMC from recent-onset pediatric T1D patients and healthy controls for reactivity to four IGRP peptides directly ex vivo. Importantly, 65% of patients and 0% of controls were positive for at least one IGRP peptide. Two of these have not been reported previously. These data provide evidence that IGRP is a CD8(+) T cell antigen in humans, contributing to the understanding of the underlying disease process as well as to future directions for diagnosis and monitoring disease progression in T1D patients.

Breed differences in susceptibility to diabetes mellitus in dogs suggest an underlying genetic component to the pathogenesis of the disease. There is little evidence for an equivalent of human type 2 diabetes in dogs, and it has been proposed that canine diabetes is more comparable to the type 1 form of the disease. Certain immune response genes, particularly those encoding major histocompatibility complex molecules involved in antigen presentation, are important in determining susceptibility to human type 1 diabetes. We tested the hypothesis that canine major histocompatibility complex genes (known as the dog leucocyte antigen) are associated with diabetes in dogs. A total of 530 diabetic dogs and more than 1000 controls were typed for dog leucocyte antigen, and associations were found with three specific haplotypes. The DLA-DRB1*009/DQA1*001/DQB1*008 haplotype shows the strongest association with diabetes in the UK dog population. This haplotype is common in diabetes-prone breeds (Samoyed, cairn terrier and Tibetan terrier) but rare in diabetes-resistant breeds (boxer, German shepherd dog and golden retriever), which could explain differences in the prevalence of diabetes in these different breeds. There is evidence that the DLA-DQA1*001 allele is also associated with hypothyroidism, suggesting that this could represent a common susceptibility allele for canine immune-mediated endocrinopathies.

Although the HLA class II alleles and immunological abnormalities are associated with type 1 diabetes mellitus (T1DM) in all racial groups, there are considerable variations in the genotypes and the prevalence of autoantibodies. In order to investigate the characteristics of the immunogenetic patterns and to use these as an early diagnostic tool and guideline for a therapeutic plan, we examined the clinical characteristics and the patterns of anti-GAD antibody (GADA), IA-2 antibody (IA-2A), HLA-DR and HLA-DQ in Korean adult-onset T1DM patients. Adult-onset patients had higher serum C-peptide levels than child-onset patients. In adult-onset patients, the prevalence of GADA and IA-2A were 59.5% and 15.3% respectively, and increased frequencies of HLADR4 and-DR9 were found. The frequencies of HLADQA1,-DQB1 and-DQ heterodimers were similar to those of the control, but child-onset patients had high frequencies of the HLA-DR3,-DR4,-DR9, DQA1*0301, DQA1*0501 and DQB1*0201 genotypes. In conclusion, Korean adult-onset T1DM patients had a lower prevalence of GADA, which was comparable to that found in Caucasian patients. The detection of GADA might help to predict the insulin dependency of adult-onset diabetes. Difference in the frequencies of diabetes associated with HLA type suggests that there might be a heterogeneity in the pathogenesis of diabetes according to the age of onset.

The list of alleles in the HLA-DRB, HLA-DQA, and HLA-DQB gene loci has grown enormously since the last listing in this journal 8 years ago. Crystal structure determination of several human and mouse HLA class II alleles, representative of two gene loci in each species, enables a direct comparison of ortholog and paralog loci. A new numbering system is suggested, extending earlier suggestions by [Fremont et al. in Immunity 8:305-317, (1998)], which will bring in line all the structural features of various gene loci, regardless of animal species. This system allows for structural equivalence of residues from different gene loci. The listing also highlights all amino acid residues participating in the various functions of these molecules, from antigenic peptide binding to homodimer formation, CD4 binding, membrane anchoring, and cytoplasmic signal transduction, indicative of the variety of functions of these molecules. It is remarkable that despite the enormous number of unique alleles listed thus far (DQA = 22, DQB = 54, DRA = 2, and DRB = 409), there is invariance at many specific positions in man, but slightly less so in mouse or rat, despite their much lower number of alleles at each gene locus in the latter two species. Certain key polymorphisms (from substitutions to an eight-residue insertion in the cytoplasmic tail of certain DQB alleles) that have thus far gone unnoticed are highly suggestive of differences or diversities in function and thus call for further investigation into the properties of these specific alleles. This listing is amenable to supplementation by future additions of new alleles and the highlighting of new functions to be discovered, providing thus a unifying platform of reference in all animal species for the MHC class II allelic counterparts, aiding research in the field and furthering our understanding of the functions of these molecules.

There are many millions of dogs worldwide, and these dogs have many different functions. The most obvious use is providing companionship, but there are also many working dogs, including guide dogs for the blind, hearing dogs, guard dogs and farm dogs, to mention a few. The health and welfare of these dogs is of great concern to dog owners, dog breeders and to those who use dogs in their work. Dogs spontaneously develop many diseases that are very similar to their human counterparts. Dogs may, therefore, provide exceptional animal models for such diseases. Identifying genetic markers in the dog may be easier than in humans, and may then provide useful information about genes that can be transferred to humans. This study looked for associations between DLA and two autoimmune diseases of the dog, diabetes and hypothyroidism. DLA associations were found for both of these diseases.

Immune-mediated haemolytic anaemia (IMHA) is the commonest immune-mediated disease of the dog, representing a major health concern to this species. The aim of this investigation was to determine whether genetic susceptibility to IMHA is associated with genes of the canine major histocompatibility complex (MHC; dog leucocyte antigen system, DLA). Samples were collected from 108 dogs with primary idiopathic, Coombs' positive IMHA. This diseased population was subdivided on the basis of Coombs' test results into two groups: 1) dogs with dominant warm-reactive immunoglobulin (Ig) G haemagglutinins and (2) dogs with an additional or dominant cold-reactive IgM haemagglutinin. The DLA class II alleles and haplotypes of the diseased population were characterised, and these data were compared with those derived from a breed-matched control cohort and a much larger group of DLA-typed dogs. Two haplotypes were increased in the patient group: DLA-DRB1*00601/DQA1*005011/DQB1*00701 (in the group with warm-reactive IgG haemagglutinins only) and DLA-DRB1*015/DQA1*00601/DQB1*00301 (in both groups, but more so in the group with cold-reactive IgM haemagglutinins). One haplotype, DLA-DRB1*001/DQA1*00101/DQB1*00201, was decreased in the total patient group, but this decrease was limited to the warm-reactive IgG haemagglutinins group, and it was actually increased in the cold-reactive IgM haemagglutinins group. A second haplotype, DLA-DRB1*015/DQA1*00601/DQB1*02301, was also decreased in the total patient group, and this decrease was found in both subgroups. In addition, all haplotypes carrying DLA-DRB1*001 were significantly increased in the cold-reactive IgM haemagglutinins group. When the overall patient group was divided on the basis of individual breeds with more than six animals represented, each of the haplotypes could be shown to be implicated in one of the breeds. Thus, it was apparent that different breeds had different MHC associations with canine IMHA, which is similar to the observation that different human ethnic groups can have different HLA associations with the same immune-mediated disease.

Diabetes mellitus occurs spontaneously in dogs, which is believed to have an autoimmune component and to be a model of human latent autoimmune diabetes of adults (LADA). Some dog breeds (e.g. Samoyed) are particularly predisposed, whereas others (e.g. Boxer) are highly resistant. With the completion of the Dog Genome Assembly, comparative genomic studies of complex diseases in dogs, including diabetes, could provide an important investigative approach into such disorders. Type 1 diabetes in humans is strongly associated with major histocompatibility complex (MHC) class II polymorphisms. We have investigated whether canine dog leucocyte antigen (DLA) class II haplotypes are associated with diabetes. DNA from 460 cases and 1047 controls were genotyped for DLA-DRB1, DLA-DQA1 and DLA-DQB1 using sequence-based typing. Three DLA haplotypes, DRB1*009/DQA1*001/DQB1*008, DRB1*015/DQA1*0061/DQB1*023 and DRB1*002/DQA1*009/DQB1*001, were found at significantly increased frequency in cases with diabetes compared with controls. One DLA-DQ haplotype, DQA1*004/DQB1*013, was significantly reduced in cases with diabetes. Further analysis showed that DQA1 alleles carrying arginine at codon 55 of DQA1 were increased in dogs with diabetes. To our knowledge, this is the first report of a comparative study of MHC and diabetes in a non-rodent species. Since no laboratory model of LADA exists and dogs and humans share similar environments, further research into canine diabetes is warranted.

Islet autoimmunity is made evident by the appearance of islet-cell antibodies directed against insulin (IAA), glutamic acid decarboxylase (GADA), protein tyrosine phosphatase IA-2 (IA-2Ab) and other autoantigens. IAA and IA-2Ab are predominantly detected in childhood type 1 diabetes mellitus (T1DM), while frequency of GADA is not affected by age. In adult-onset T1DM patients, GADA is the immune marker of higher diagnostic sensitivity. In adult diabetic patients not requiring insulin treatment for at least 6 months after diagnosis, GADA identifies the so-called latent autoimmune diabetes in adults (LADA). In over 80% of cases, LADA patients develop insulin dependency within a few years after the diagnosis and have an increased risk for the development of other organ-specific autoimmune diseases. High GADA titers identify a subgroup of LADA patients with low body mass index (BMI), low C-peptide levels and increased frequency of T1DM-related HLA class II haplotypes. GADA assay should be offered to every diabetic patient, and in cases of positivity screening for other autoimmune diseases should be carried out.

Type I diabetes is a complex disease in which multiple susceptibility loci have been implicated by whole genome scans. IDDM8, a susceptibility locus, is located on chromosome 6q27, however the specific susceptibility gene has yet to be identified. We have examined five potential candidate genes using 36 genetic markers, spanning 360kb located near the chromosome 6q27 terminus in 478 families for diabetes association. No associations with type I diabetes susceptibility were detected with the strength previously observed for IDDM1 or IDDM2. However, a novel CAG/CAA polymorphism was detected in exon 3 of the TATA box-binding protein gene, which shows preliminary evidence of association with diabetes susceptibility (p<0.05).

The IDDM5 gene, which is identified by whole-genome searches, is located on chromosome 6q25. TAB2 (MAP3K7IP2 [mitogen-activating protein kinase kinase kinase 7 interacting protein 2]) is a potential candidate gene for type 1 diabetes because it is located on chromosome 6q25 and is involved in nuclear factor (NF)-kappaB regulation. We have conducted familial association studies using 478 families and demonstrate that a type 1 diabetes susceptibility gene resides within a 212-kb region containing the TAB2 gene (Tsp = 1.0 x 10(-2) to 4.0 x 10(-4)). No amino acid polymorphisms were detected in TAB2; however, multiple single nucleotide polymorphisms (SNPs) found within 5' untranslated, 3' untranslated, and intron regions were associated with type 1 diabetes susceptibility. Two additional genes, LOC340152, a predicted gene with currently unknown function, and SMT3, which has homology to SUMO (small ubiquitin-related modifier) were found within the 212-kb region and were associated with type 1 diabetes susceptibility. Functional studies of the three genes will be required to determine their biological relevance to type 1 diabetes. However, both TAB2 and SUMO are involved in NF-kappaB activation and may thus be involved in type 1 diabetes through apoptosis in pancreatic beta-cells.

Most cases of type 1 diabetes (T1DM) are due to an immune-mediated destruction of the pancreatic beta cells, a process that is conditioned by multiple genes and environmental factors. The main susceptibility genes are represented by the class II HLA-DRB1 and DQB1 alleles. The aim of our study was to reconfirm the contribution of HLA-DQB1 polymorphisms to T1DM genetic susceptibility for the Romanian population. For this, 219 Romanian T1DM families were genotyped at high resolution for HLA DQB1 using the PCR-SSOP method (Polymerase Chain Reaction - Sequence Specific Oligonucleotide Probes). Allele transmission to diabetics and unaffected siblings was studied using the Transmission Disequilibrium Test (TDT). We found an increased transmission of DQB1*02 (77.94% transmission, p(TDT) = 7.18 x 10(-11)) and DQB1*0302 (80.95% transmission, p(TDT) = 2.25 x 10(-10)) alleles to diabetics, indicating the diabetogenic effect of these alleles. Conversely, DQB1*0301, DQB1*0603, DQB1*0602, DQB1*0601 and DQB1*05 alleles are protective, being significantly less transmitted to diabetics. In conclusion, our results confirmed the strong effect of HLA-DQB1 alleles on diabetes risk in Romania, with some characteristics which can contribute to the low incidence of T1DM in this country.

Diabetes encompasses a heterogeneous group of diseases, each with a substantial genetic component. We review the division of diabetes into different subtypes based on clinical phenotype, the fruitful pursuit of genes underlying monogenic forms of the disease, the successes and drawbacks of whole-genome linkage scans in type 1 and type 2 diabetes, and the recent identification of several diabetes genes by large association studies. We use the lessons learned from this extensive body of evidence to illustrate general implications for the genetic analysis of complex traits.

Family and twin studies indicate that a substantial fraction of susceptibility to type 1 diabetes is attributable to genetic factors. These and other epidemiologic studies also implicate environmental factors as important triggers. Although the specific environmental factors that contribute to immune-mediated diabetes remain unknown, several of the relevant genetic factors have been identified using two main approaches: genome-wide linkage analysis and candidate gene association studies. This article reviews the epidemiology of type 1 diabetes, the relative merits of linkage and association studies, and the results achieved so far using these two approaches. Prospects for the future of type 1 diabetes genetics research are considered.

The major histocompatibility complex Class II alleles, HLA-DQ, and the related HLA-DR, are the chief genetic elements of human type 1 diabetes. These genes code for polymorphic heterodimeric proteins, whose chief function is to trap peptide antigens in the endosome and present them on the surface of antigen-presenting cells (dendritic cells, B lymphocytes, monocytes/macrophages) to CD4(+) T helper cells. A systematic investigation of the molecular properties of HLA-DQ alleles linked to susceptibility or resistance to type 1 diabetes has shown that these properties segregate along lines of susceptibility or resistance. A correlation of these features with the function of each particular segment of the HLA-DQ molecule yields interesting insights into the possible pathways leading to type 1 diabetes. There remain, however, areas to be clarified, including mechanisms by which dominant protection is conferred by certain alleles, the interplay between HLA-DQ and the related locus HLA-DR, that also shows autoantigen-specific reactivity, and the cross-Class help delivered to CD8(+) T cells, the final effectors in pancreatic beta-cell destruction. Clarification of these issues may lead to ways to prevent diabetes in predisposed individuals already exhibiting the genetic and immunological characteristics, and perhaps a cure in those with the disease, by means of transplantation, and measures for prevention of disease recurrence.

Vitiligo is a depigmenting disorder characterised by the loss of melanocytes from the cutaneous epidermis. Although the exact cause of the condition remains to be established, an autoimmune aetiology has been suggested and several observations support this theory. These will be the topic of discussion in this review. In brief, the disease is frequently associated with other disorders which have an autoimmune origin such as autoimmune thyroiditis and insulin-dependent diabetes mellitus. Furthermore, circulating antibodies and T lymphocytes which react against melanocyte antigens are present in the sera of a significant proportion of vitiligo patients compared with healthy individuals. Immunosuppressive therapies which are reasonably effective in treating the condition, well-studied animal models of the disease as well as the association of vitiligo with MHC antigens, all add credence to the hypothesis that immune mechanisms play a role in the development of vitiligo.

To assess the effect of Asian-specific HLA haplotypes on susceptibility to type 1 diabetes, we investigated the association of genotypic combinations of DRB1-DQB1 haplotypes with susceptibility to type 1 diabetes. We studied 132 Japanese patients with type 1 diabetes and 157 control subjects, along with 67 Korean patients and 109 control subjects. DRB1*0405-DQB1*0401 and DRB1*0901-DQB1*0303 were confirmed to be two major susceptible HLA haplotypes in the Japanese population. The frequencies of heterozygotes and homozygotes with DRB1*0405-DQB1*0401 were similarly higher in patients than in control subjects (homozygotes, 5.3% vs. 3.8%; heterozygotes, 48.5% vs. 26.1%). In contrast, homozygotes, but not heterozygotes, with DRB1*0901-DQB1*0303 were more frequent in patients with type 1 diabetes than in control subjects (homozygotes, 12.9% vs. 0.6%; heterozygotes, 22.0% vs. 24.8%). A similar tendency was also observed in the Korean population. In multiple logistic regression analysis, DRB1*0405-DQB1*0401 fitted a dominant model and DRB1*0901-DQB1*0303 fitted a recessive model. These data, which indicate that the contribution of HLA haplotypes to the genetic susceptibility to type 1 diabetes differs depending on the genotypic combination of HLA haplotypes, suggest the importance of extensive analysis of genotypes in studies on HLA and disease association in general.

Several studies of the association between HLA and type 1 diabetes have been carried out revealing differences between ethnic groups. Our study, as part of the studies that should be performed about this association in the rest of the word, aims at elucidating the HLA DRB1, DQB1 polymorphism in Tunisian type 1 diabetes. This study includes 43 unrelated type 1 diabetes patients, and their mean age at onset is less than 15 years. Analysis of the frequency of alleles and haplotypes in these subjects, compared to a reference group (n = 101) led to the following results. 1) The Tunisian insulin-dependent diabetics present similarities as well as differences with other ethnic groups (Caucasians, North Africans). 2) The haplotype DRB1*04 DQ*0302 and DRB1*03 DQB1*0201 is positively associated to type 1 diabetes. 3) The heterozygotic genotype DRB1*04 DQB1*0302 / DRB1*03 DQB1*0201 is strongly associated to type 1 diabetes. 4) The haplotypes DRB1*01501 DQB1*0602 and DRB1*11 DQB1*0301 proved to be protective. In addition, the study of the subtypes DRB1*04 showed that alleles DRB1*0405 predispose to type 1 diabetes, whereas the allele DRB1*0403, which is in linkage disequilibrium with the DQB1*0402 in the Tunisian population, has a protective effect.

Nuclear factor kappa B (NFkappaB) is an important transcription factor that is involved in the response to oxidative stress and inflammation. Recent studies suggest that it may be involved in the development of diabetic microvascular complications. A highly polymorphic (CA) dinucleotide repeat microsatellite has been identified in the regulatory region of the NFkappaB gene. The aim of this study was to investigate whether this polymorphic region was associated with susceptibility to type 1 diabetes, or its late complications. Genomic DNA was extracted from the peripheral blood of 217 patients with type 1 diabetes mellitus (T1DM) and 111 normal healthy controls. In our population 18 alleles (A1-A18) were identified. There was a highly significant decrease in the frequency of allele 146 bp (A14) in type 1 diabetes (0.03) compared with the normal controls (0.28) (chi(2) = 79.8, Pc = 0.00001). In contrast, the frequency of the allele 138 bp (A10) was significantly increased in patients with type 1 diabetes (0.17) compared with the normal controls (0.02) (chi(2) = 32.8, P < 0.00000). These results demonstrate that the NFkappaB gene may play a role in the susceptibility to type 1 diabetes: individuals with the A10 allele may be more likely to develop diabetes compared with the A14 allele.

The CTLA4 (cytotoxic T lymphocyte associated antigen-4) gene encodes the T cell receptor involved in the control of T cell proliferation and mediates T cell apoptosis. C-T polymorphism is present at position -318 from the ATG start codon in the promoter region of the gene. We report a study on the polymorphism in 347 unrelated children with type 1 diabetes mellitus (DM) (age at diagnosis 7.2+/-3.8 years) and their 260 healthy siblings as controls. Genotype C/C conferred a risk of type 1 DM (RR = 2.02, 95% CI 1.32-3.10, pc = 0.0033). The gene frequency of the C allele was higher in patients (RR = 1.91, 95% CI 1.28-2.84, pc = 0.0026). The gene frequency and phenotype frequency of the T allele were negatively associated with type 1 DM (RR = 0.52, 95% CI 0.35-0.78, pc = 0.0026 and RR = 0.49, 95% CI 0.32-0.76, pc = 0.0022, respectively). The frequency of genotype C/T was lower in patients (RR = 0.50, 95% CI 0.32-0.78, pc = 0.0051). This study demonstrates that nucleotide -318 C-T polymorphism of the CTLA4 gene is associated with type 1 DM. The promoter allele -318 C confers a risk of type 1 DM but allele -318 T confers protection against this disease.

Previous studies have suggested that the human leukocyte antigen (HLA) class I region may be involved in determining the age at onset and clinical severity of type 1 diabetes. We have investigated the frequency of polymorphisms of the nonclassical HLA class I gene, HLA-E, in 199 British Caucasian patients with type 1 diabetes and 82 healthy controls. A highly significant increase in the frequency of the HLA-E 0101 genotype was found in the patients compared to controls (chi(2) = 15.3, p < 0.00009). The frequency of the HLA-E 0101 genotype was increased in those patients diagnosed after 10 years of age, while the frequency of the 0101, 0103 genotype was significantly increased in those subjects diagnosed before 10 years of age (chi(2) = 26.0 p < 0.000003 and chi(2) = 13.0 p < 0.0003, respectively). No obvious interaction between the HLA-E locus and the class II DQB1*0201, 0302, and 0501 susceptibility alleles was found. This is the first report of an association between the HLA-E locus and susceptibility to an autoimmune disease.

The CTLA4 (cytotoxic T lymphocyte associated antigen-4) gene encodes the T cell receptor involved in the control of T cell proliferation and mediates T cell apoptosis. Thus it is a strong candidate gene for T cell-mediated autoimmune disease. There is polymorphism at position 49 in exon 1 of the CTLA4 gene, providing a A-G exchange. This polymorphism is reportedly associated with type 1 diabetes in Caucasians but not in a small data set of Chinese. We wished to test this polymorphism in a larger and more homogeneous data set of Chinese children with type 1 diabetes and normal adult controls.

A population-based case-control study of a CTLA4 gene 49 A-G polymorphism was performed to look for an association with type 1 diabetes in Chinese children.

We analysed this polymorphism in 253 unrelated children (128 boys) with type 1 diabetes (age at diagnosis 7.1 +/- 3.7 years) and 91 randomly selected normal adults. All individuals were Han Chinese.

The genotype and gene frequencies of children with type 1 diabetes differed significantly from those of adult controls (P = 0.0091 and P = 0.0051, respectively). Genotype CTLA4 49 G/G and G allele conferred a risk of type 1 diabetes (RR = 2.13, 95% CI = 1.31-3.46, P = 0.0022; RR = 1.68, 95% CI = 1.17-2.43, P = 0.0051, respectively).

This study demonstrates that CTLA4 49 A-G polymorphism is associated with type 1 diabetes in Han Chinese children. The CTLA4 49 G allele confers an increased risk of type 1 diabetes.

The prevalence of human leukocyte antigen (HLA) DQB1 and DQA1 alleles has been determined in 78 Kuwaiti Arab children with insulin-dependent diabetes mellitus (IDDM) and in 57 normal healthy controls with similar ethnic background. The typing of HLA-DQ alleles was carried out using an allele-specific DNA-based polymerase chain reaction (PCR) SSP method. DR typing was also performed in 212 control subjects using PCR-SSP (sequence specific primer) method. A significantly higher frequency of DQB1*0201 allele was found in IDDM cases compared to the controls (p<0.001). There was no significant difference in the prevalence of DQB1 alleles *0302, *0501, and *0602 between IDDM cases and the controls. In contrast, DQB1 alleles *0301, *0402, *0502, *0602, and *0603 were represented at a somewhat higher frequency in controls compared to the IDDM cohort. The frequency of DQA1 allele *0301, which encode for an Arg at codon 52, was significantly higher in the IDDM patients compared to the controls (p<0.001). The frequency of DQA1 allele *0302 was also higher in IDDM cases than controls (p = 0.034) but the difference was less pronounced than DQA1*0301. Amongst the Arg52 alleles, no significant difference was detected in the frequency of *0401 between IDDM cases and the controls and the allele *0501 was detected only in controls. For non-Arg52 alleles *0103, *0104, and *0201, the differences in the two groups were not significant, with the exception of allele *0104 (p = 0.024). DR3 was the most common type in the Kuwaiti general population (28%) and DRB1*0301 was detected in 41% of the individuals with DR3 specificity. Analysis of HLA-DQBI/DQA1 haplotypes from IDDM cases and controls revealed a significantly high frequency of haplotype DQA1*0301/DQB1*0201 between Kuwaiti IDDM cases (49/78, 63%) and the controls (8/57, 14%).

Type 1 (insulin-dependent) diabetes mellitus is associated with HLA DR and DQ factors, but the primary risk alleles are difficult to identify because recombination events are rare in the DQ-DR region. The risk of HLA genotypes for type 1 diabetes was therefore studied in more than 420 incident new onset, population-based type 1 diabetes children and 340 age, sex and geographically matched controls from Sweden. A stepwise approach was used to analyse risk by relative and absolute risks, stratification analysis and the predispositional allele test. The strongest relative and absolute risks were observed for DQB1*02-DQA1*0501/DQB1*0302-DQA1*0301 heterozygotes (AR 1/46, P < 0.001) or the simultaneous presence of both DRB1*03 and DQB1*0302 (AR 1/52, P < 0.001). Stratification analysis showed that DQB1*0302 was more frequent among DRB1*04 patients than DRB1*04 controls (P < 0.001), while DRB1*03 was more frequent among both DQA1*0501 (P < 0.001) and DQB1*02 (P < 0.001) patients than respective controls. The predispositional allele test indicated that DRB1*03 (P < 0.001) would be the predominant risk factor on the DRB1*03-DQA1*0501-DQB1*02 haplotype. In contrast, although DQB1*0302 (P < 0.001) would be the predominant risk factor on the DRB1*04-DQA1*0301-DQB1*0302 haplotype, the predispositional allele test also showed that DRB1*0401, but no other DRB1*04 subtype, had an additive risk to that of DQB1*0302 (P < 0.002). It is concluded that the association between type 1 diabetes and HLA is due to a complex interaction between DR and DQ since (1) DRB1*03 was more strongly associated with the disease than DQA1*0501-DQB1*02 and (2) DRB1*0401 had an additive effect to DQB1*0302. The data from this population-based investigation suggest an independent role of DR in the risk of developing type 1 diabetes, perhaps by providing diseases-promoting transcomplementation molecules.

The association between HLA-DR and DQ and insulin dependent diabetes mellitus (IDDM) was analyzed in 47 patients and 76 controls of Yemenite Jewish origin. The IDDM susceptibility alleles DRB1*03011, DQA1*0501, DQB1*02 and DRB1*0402, DQA1*0301, DQB1*0302 found in Caucasians had a very strong predisposing effect also in the Yemenite IDDM group. The DRB1*07, DQA1*0201 and DQB1*02 alleles were found to have a strong negative association with IDDM. None of the patients carried DRB1*07 and DQA1*0201 compared with healthy controls (43.7%). Our analysis revealed that the DRB1*03011 DQA1*0501 DQB1*02/DRB1*04 DQA1*03 DQB1*0302 heterozygous genotype confers the highest susceptibility (59.6% in patients vs. 0% in controls). The homozygous DRB1*03 and DRB1*04 genotypes were also found to be positively associated with the disease. 81% of the patients compared to 1.3% of controls carried the susceptibility alleles on both haplotypes. In conclusion, the development of IDDM in Yemenite Jews is strongly dependent on the presence of the susceptibility HLA alleles and on the absence of the DRB1*07 haplotype. The Yemenite Jewish group is uniquely homogenous with regard to genetic susceptibility factors involved in the process of IDDM, and may thus be an ideal model for further genetic studies.

Diabetes-prone DP-BB rats spontaneously develop insulin-dependent diabetes mellitus resembling type 1 diabetes mellitus in man. Expression of T cell lymphopenia and presence of at least one class II major histocompatibility complex (MHC) RT1u haplotype are required for development of diabetes. Diabetes segregation was studied in lymphopenic backcross (BC) offspring from a cross between male DP-BB/HRI and female BN/Mol rats. Diabetes occurred in 75% of BC rats with genotype RT1u/u and in 18% of those being RT1n/u in genotype. The latter developed diabetes significantly later than MHC homozygotes and parental DP-BBs. Our data further point to the existence of additional genes of minor importance for development of IDDM. One of these seemed to be positioned on the X chromosome. The recently published linkage to chromosome 18 could not be confirmed however. Finally, the BN-derived non-albino allele of the C gene was associated with higher diabetes incidence. This points to the existence of minor susceptibility genes in other strains of rats.

Many studies have shown an association of IDDM with polymorphisms in the HLA region on chromosome 6p21. Previously our case-control study in the Belgian population showed significant association between IDDM and certain HLA class II alleles, in particular Lys71+, encoding DRB1 alleles. In the present study, 81 Danish multiplex IDDM families and 82 healthy Danish controls were examined for polymorphisms in the HLA-DRB genes and 54 of the 81 families for polymorphisms in HLA-B, -DQA1, -DQB1, -TNFA, and -TNFB genes. The results confirm our previous studies in the Belgian population and show that DRB1Lya71+/+ homozygotes have a relative risk (RR) of 103.5. Linkage between IDDM and DRB1 alleles that encode Lys71+ was shown by affected zib pair analysis which showed strong linkage (p < 1 x 10(-6). By family based association studies, the DRB1Lys71+ was identified as the allale which increased susceptibility to develop IDDM most in the HLA region (haplotype relative risk = 8.38). Haplotype analysis confirmed the increased risk contributed by DRB1Lys71+ alleles and in addition showed that DRB1Lys71- provides protection against IDDM even in the presence of DQB1Aep47-. These results indicate that DRB1Lys71+ screening is a powerful test compared to full HLA typing to determine the risk for a random person to develop IDDM in the Danish population, with an even higher probability than shown previously for the Belgians.

The structural features of HLA-DQ alleles which are susceptible and resistant to insulin-dependent diabetes mellitus (IDDM) have been examined using a model of their three-dimensional structure obtained by energy minimisation, based on the published structure of HLA-DR1. The model shows DQ molecules to have an overall shape nearly identical to that of DR molecules, but with significant differences in the fine structure: 1) the antigen-binding groove of DQ molecules has a polymorphic first pocket; this pocket can be either amphiphilic or hydrophilic, 2) The beta 49-56 dimerisation domain of DQ is polymorphic: hydrophobic, or amphiphilic, or hydrophilic and positively charged, leading to spontaneous or T-cell receptor-induced homodimer formation, or T-cell receptor-induced homodimer formation, or difficulty of the formation of such dimers, respectively; 3) a prominent Arg-Gly-Asp loop is formed by some DQ alleles (beta 167-169) and probably functions in cell adhesion. There are also small differences in the residues and sequences implicated in CD4 binding (mostly in DQ beta 134-148) but the significance of these differences cannot be evaluated at present. All seven DQ alleles which confer susceptibility to IDDM possess a hydrophilic first pocket in the antigen-binding groove, a hydrophobic or amphiphilic beta 49-56 dimerisation patch that allows for spontaneous or T-cell receptor-induced dimerisation, and the Arg-Gly-Asp loop. By contrast, in the protective alleles at least one of these three features is absent. This segregation of phenotypes according to susceptibility or resistance can well explain the model of tighter autoantigen binding by the protective alleles compared to the susceptible alleles, previously proposed for the pathogenesis of IDDM.

The (MHC) class II association with insulin-dependent diabetes mellitus (IDDM) is well documented. However, it is likely that genes within the MHC class III and the class I region also play a role in determining susceptibility to IDDM. In this study we have used a novel molecular probe to investigate the class I P3A and P3B loci of 179 patients with IDDM and 142 normal control subjects. A highly significant increase in the frequency of the class I P3 4.0;1.5 kilobase (kb) and 4.0;1.8;1.5 kb genotypes was found in patients compared to the control subjects (chi 2 46.8, 6 df, p < 0.0001). The association with the P3B 1.5 kb allele was strongly associated with the age at onset of diabetes, being present in 96.2% of subjects who developed diabetes between the age of 10-20 years compared to 55.0 and 74.6% who developed diabetes before 10 years or after 20 years, respectively (chi 2 31.4, p < 0.0001). There was no evidence for linkage disequilibrium between the DQA1 and DQB1 loci and P3B suggesting that this is an independent association. In conclusion, these results suggest that genes in both the MHC class I and II regions confer susceptibility to IDDM and are related to the age at onset of the disease.

Insulin-dependent diabetes mellitus is strongly associated with certain HLA types and the presence of islet cell-specific autoantibodies. The pathogenesis is a specific loss of pancreatic beta cells. The dissection of IDDM genes is complicated by the low recurrence rate of the disease among first-degree relatives. HLA-DQ2 and 8 are closest to IDDM with a marked synergistic effect of DQ2/8 heterozygotes. The associations with other HLA genes are often explained by linkage disequilibrium. Genetic factors on other chromosomes which influence the pathogenesis are still to be fully identified but candidates are on chromosomes 11 (insulin gene polymorphisms) and 7 (TCR gene polymorphisms). The autoreactivity against the GAD65 isoform is pronounced both before and at the clinical onset of IDDM. GAD65 autoantibodies show the highest predictive value and may represent an initiating autoantigen. Autoantibodies to numerous other beta cell autoantigens are detected at the clinical onset but may represent a secondary response and antigen spreading during a sustained autoimmune attack on the beta cells. The role of T cells in human IDDM is yet to be defined. GAD65 and other islet autoantibodies have a low positive predictive value for IDDM and further investigations are needed to clarify ways to predict IDDM in the general population.

The transmission of HLA-DR and DQ was compared between 46 families with at least one child affected by insulin dependent diabetes mellitis (IDDM) and 43 healthy control families. In the patient families, there was an increased transmission of DR4 (p < 0.025) and DQB1*0302 (p < 0.01) from both parents to the index patient. There was an increased transmission of DQB1*0302 (p < 0.03) from the mothers only. The non-inherited maternal haplotypes showed a significantly decreased frequency (p < 0.01) of positively associated haplotypes (DR4-DQA1* 0301-DQB1*0302, DR3-DQA1*0501-DQB1*0201) compared to all parental haplotypes in the control families. In the control families neither transmission rates nor frequencies of non-inherited haplotypes differed from those expected in the control families. In conclusion, the observed reduction of IDDM-positively associated haplotypes in patient non-inherited maternal haplotypes, but not in non-inherited paternal haplotypes, suggests that tolerance during fetal life to maternal non-inherited HLA molecules may be important to diabetes development.

We have developed a rapid and simple method to detect the relation between HLA-DQ beta 57 Asp and Chinese IDDM patients. The method involved the selective amplification of a DNA fragment from the HLA-DQ B1 gene by using the mutagenic primers. After PCR, if the HLA-DQ beta 57 was Asp, then there was an artificially created restriction enzyme cutting site. We then can accurately obtain the results by enzyme digestion and electrophoresis. Sixty-nine IDDM patients and 30 nondiabetic control subjects were analyzed using this method. Twenty-two (42%) IDDM patients had non-Asp 57 homozygous, 31/45%) were Asp/non-Asp 57 heterozygous, and 9 (13%) had Asp-57 homozygous. Of the 30 control subjects, the number of cases for these three types were 6 (20%), 18 (60%), and 6 (20%), respectively. The relative risk of homozygous DQ beta 57 non-Asp in our group was 2.9 and the p value was greater than 0.05. Using this kind of approach, we were able to provide a simple, rapid, and non-radioactive method to detect whether the HLA DQ beta 57 was Asp or not.

The clinical onset of insulin-dependent diabetes is associated with several autoimmune phenomena including islet cell antibodies, glutamic acid decarboxylase (the GAD65 isoform) autoantibodies (GAD65Ab) as well as insulin autoantibodies. The molecular cloning of these autoantigens has permitted the development of precise and reproducible antibody immunoassays to identify marker-positive patients and control subjects. Among patients with new-onset diabetes about 70% were GAD65Ab positive compared to 1.5% among control subjects while 46% of patients had IAA compared to 1% among control subjects. The autoreactive sites or epitopes of GAD65 and insulin remain to be determined. The disease association with HLA on chromosome 6 may help to define the epitope specificity of the autoimmune reaction. Recent data suggest that 95% of new-onset IDDM children (0-15 years of age) are positive for either DQ2, DQ8 or both compared to about 50% of healthy control subjects. HLA-DQ6 is negatively associated with the disease. Both HLA-DQ2 and DQ8 therefore seem to be necessary, but not sufficient for diabetes. Molecular modelling suggests comparable physicochemical properties of DQ2 and DQ8 but are widely different from DQ6. In 1984, the conclusion was that molecular cloning of the genes for the autoantigens, antibodies, T-cell receptors, as well as HLA class I and II molecules associated with diabetes are essential for analysing the components which control the development of pancreatic beta-cell autoimmunity. In 1994, autoantigens and HLA molecules have been cloned and recombinant reagents developed to be used in experiments aimed at testing whether it will be possible to predict IDDM.(ABSTRACT TRUNCATED AT 250 WORDS)

Fifty juvenile insulin dependent diabetes mellitus (JIDDM) patients of Tamil Nadu (South India) were typed for HLA-A, -B, -C, -DR, and -DQ, ESD, GLOI, C3 and HP polymorphisms. The frequencies of B8, DR3, DR4, DR53 and DQ2 antigens of the HLA system were significantly higher in the patients than in controls (relative risk, RR = 4.81; 5.14; 3.98; 3.36 and 2.53, respectively). However HLA-DR2, -DR5 and -DQ1, observed less frequently in the patient group, appear to play a role of protection against the disease (RR = 0.32; 0.30 and 0.20 respectively). HLA haplotype analysis demonstrated very high relative risk associated with two hitherto unreported haplotypes namely A3,DR1 and Cw3,DR4 (RR = 27.30 and 20.00, respectively) and also scanty distribution of the haplotypes A1,B17 and DR2,DQ1 (RR = 0.39 and 0.36, respectively) in the patient group. Among other genetic markers tested, GLOI is informative with its phenotype GLOI 2-1 showing positive association with JIDDM (RR = 4.06).

A major aim of HLA and disease association studies is to identify the causative HLA factor truly responsible for the association. This is usually difficult due to the pronounced linkage disequilibrium between most HLA determinants. The causative factor must show the strongest association compared to all other factors. Here we describe a simple analysis which can be used to identify which of two factors, say A and B, shows the strongest association. The basic data for the analysis are the entries of the two-by-four table giving the four phenotypic combinations of A and B in patients and controls, respectively. These data are analyzed in various two-by-two tables involving stratification of each of the two factors against the other. A stronger increase of factor A is established if A is significantly associated with the condition both in B-positives and in B-negatives, when this is not true for B in A-positives and A-negatives. Using simulation with control data, it is demonstrated how linkage disequilibrium may influence secondary associations. The analysis may also be used to investigate interaction between HLA factors, but linkage disequilibrium complicates the interpretation in such cases. The method is exemplified using various published data. Finally, some statistical recommendations are given. Thus, we advise that phenotype (marker) frequencies are generally used instead of gene (i.e. allele, or haplotype) frequencies. The importance of correcting p-values, the levels of significance, and the power of Fisher's exact test are discussed.