Researchers have now probably pinpointed all the genes in the MHC genomic region that are risk factors in type 1 diabetes. As the MHC is unusually rich in genes involved in immunity, this is truly exciting.
If there were 'Guinness World Records' for genomic regions, the major histocompatibility complex (MHC) would have several entries. It is the most gene-dense region in mammals, and variations in the human MHC are linked to susceptibility to more common diseases than any other region in the human genome. Moreover, MHC-associated diseases include almost all autoimmune diseases. But because of high gene density in the MHC region and the localization of these genes very close to each other, determining which gene, or group of genes, is associated with a specific disease is akin to looking for a needle in a haystack. For example, some variants of the MHC-localized genes that encode HLA proteins are known to determine different susceptibilities to the autoimmune disorder type 1 diabetes — ranging from increased vulnerability to increased protection. But this seems not to be the whole story, and the identities of other HLA genes suspected to be associated with type 1 diabetes remain elusive. Nejentsev et al.1 (page 887 of this issue) now zoom in on the MHC to identify these missing links.
In individuals genetically susceptible to type 1 diabetes2, misregulation of the immune system results in immunological intolerance towards insulin-producing β-cells, which are part of cell populations in the pancreas called the islets of Langerhans. Type 1 diabetes is characterized by the inflammation of islets of Langerhans and, subsequently, the selective destruction of the β-cells. Consequently, patients become permanently dependent on insulin injections.
The highly variable HLA genes are clustered in two loci, or positions, in the MHC region on chromosome 6, separated by one genomic map unit (1 centimorgan). They encode two classes of proteins: the HLA class I and the HLA class II molecules. These proteins are expressed on cell surfaces, where they present both self and foreign peptide antigens to different types of T cells.
Which MHC-localized genes are risk factors for autoimmune diseases is extensively debated. For example, although gene variants (alleles) of HLA class II genes are known to impose the greatest predisposition to type 1 diabetes, the MHC region has long been suspected of harbouring other genes contributing to the risk of this disease3. Many MHC genes have immunological functions, however, making them suspects in contributing to autoimmune responses, and the unusually high density of genes in the MHC offers an almost endless list of potential candidates4. What is more, because of their strong genetic linkage (linkage disequilibrium), most of the MHC genes are inherited together4, which makes it difficult to separate their effects. Thus, several attempts at pinpointing additional culprits in the pathology of type 1 diabetes have been unsuccessful. Consequently, the entire MHC region has been deemed a 'super-locus', consisting of a cornucopia of discrete risk loci for type 1 diabetes5.
In the latest attempt at disentangling risk factors for type 1 diabetes, Nejentsev et al.1 have mapped all remaining MHC genes known to be associated with an increased risk of the disease to several alleles of the HLA class I HLA-A and HLA-B genes. This provides evidence that HLA class I genes (as well as, and independently of, HLA class II genes) make a strong contribution to the pathology of this disease. Although the authors leave room for other minor MHC contributors, their work suggests that classic antigen-presenting HLA class I and II molecules, and nothing else, account for MHC-associated genetic susceptibility to this disease. All other possible associations are probably attributable to their strong genetic linkage with HLA class I and II genes.
The association of HLA class I genes with type 1 diabetes has been reported previously6. But Nejentsev and colleagues' findings stand out from earlier observations in two respects. First, their data are based on an experimental approach that is sufficiently powerful to overcome the confounding effect of HLA class II genes — they used a large number of patients and control subjects and performed a comprehensive scan of the entire MHC region.
Second, to rule out the influence of HLA class II genes, the authors used the statistical method of recursive partitioning and regression, which avoids unwarranted assumptions, thereby forming a tree of known, disease-risk-associated HLA class II genes in which secondary associations were tested. They then meticulously validated the contribution of these HLA class II genes to confirm that no residual effects, other than the detected HLA class I contribution, was responsible for the observed associations.
Why do certain HLA genes affect disease susceptibility? Although evidence for a link between the disease association of these genes and the function of their protein products is scarce, it is tempting to speculate on such a link. Variation in the HLA genes leads to a large variety of peptide-binding molecules that generate specific immune responses. For example, HLA proteins play an essential part in shaping the developing T-cell repertoire in the thymus, limiting autoreactivity and guiding antigen-specific T-cell responses.
This feature might explain why certain alleles of HLA class I and II genes confer an increased risk of type 1 diabetes. Both pro-inflammatory, autoreactive T helper cells and cytotoxic T cells can recognize β-cell-derived peptides presented by class II and class I HLA proteins, respectively7,8(Fig. 1). These responses are reflected in the inflammation of islets of Langerhans in type 1 diabetes, and in the β-cell destruction that occurs after transplantation of islets into patients with diabetes. Moreover, HLA class I proteins bind to KIR receptors on natural killer cells. In response, these cells of the innate, or nonspecific, immune system contribute to the regulation of adaptive (specific) immune responses. Thus, HLA class I proteins link innate and adaptive immunity, and their misregulation in disease can lead to attack on the individual's own β-cells.
The innovative analytical approach of Nejentsev et al.1 underscores the outstanding importance of HLA class I genes not just in the pathogenesis of type 1 diabetes but, conceivably, in other HLA-associated autoimmune diseases. It also suggests that, with sufficiently powerful experimental tools and appropriate study design, specific HLA genes can be implicated in specific autoimmune diseases. MHC genes other than HLA class II genes have also been linked to other autoimmune diseases such as rheumatoid arthritis, coeliac disease, systemic lupus erythematosus and myasthenia gravis. So it is crucial to define whether these associations can be attributed to class I alleles and to elucidate the involvement of this class of HLA molecules in the general mechanisms of autoimmune diseases. For type 1 diabetes, it is now beyond doubt that the contribution of HLA to the process of β-cell autoimmune destruction is more than just guilt by genetic association.