TAP1,2: bare lymphocyte syndrome* (association with rheumatoid arthritis, vitiligo) BRD2: juvenile chronic epilepsy
Figure 12.4 MHC associations with disease. Examples of specific genes, alleles, or haplotypes that have been associated with disease susceptibility through linkage analysis or association are indicated. Where a causal relationship has been established an asterix is shown.
HLA-DRB1 alleles were found to be most significant in determining disease susceptibility. These include DRB1*0401, *0404, and *0101 (Stastny 1976). Studies of different disease associated DRB1 alleles across a range of populations implicated a common short sequence of amino acids that was part of the peptide binding groove of the DRP1 chain, leading to the 'shared epitope' hypothesis (Gregersen et al. 1987).
The molecular mechanism underlying the DRB1 association remains unresolved. It appears that despite a very strong and reproducible genetic association, possessing a shared epitope DRB1 allele is neither necessary nor sufficient for disease to occur (Box 12.3). The situation is complex as the major disease associated alleles vary between ethnic groups and there is a hierarchy of strength of association, with some DRB1 alleles actually associated with protection from disease. The effect has low penetrance as about one-third of the UK population carry DRB1*04. Surprisingly one, rather than two, copies of particular alleles appear associated with more severe disease. Nor can this be the whole story as one-third of patients with rheumatoid arthritis do not carry an allele encoding the shared epitope motif. Intriguingly, there is also evidence that the shared epitope alleles are associated with the development of pathogenic autoantibodies (to anticitrunillated protein)
and only show association with disease in patients with these antibodies (van der Helm-van Mil et al. 2006).
Are there more strongly associated loci or indeed causative variants in linkage disequilibrium with DRB1? Certain haplotypes, for example DRB1*0401-DQB1*0301, have been associated with disease severity but remain unresolved (Cranney et al. 1999). Susceptibility loci in the neighbouring class III region are proposed, for example at NFKBIL1 (nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor-like 1) in a Japanese population (Okamoto et al. 2003) and the TNF gene cluster where specific haplotypes have been implicated but causative variants not resolved (Newton et al. 2003). There is some recent evidence that diversity involving a nearby gene, AIF1, encoding allograft inflammatory factor-1 may be involved (Harney et al. 2008). Analyses have not been restricted to disease risk; pharmacogenomic studies in rheumatoid arthritis have shown association with the MHC, for example between TNF promoter polymorphisms and disease activity following treatment with anti-TNF monoclonal antibodies (Mugnier et al. 2003).
Why should it prove so hard to localize disease associations? Some reasons are common to studying genetic factors in any multifactorial disease trait such as infectious or autoimmune diseases where genetic, epigenetic,
Box 12.3 Non-MHC disease associations with rheumatoid arthritis
Although the MHC is the major genetic risk locus for rheumatoid arthritis, a number of other loci have been implicated (Coenen and Gregersen 2009). These include variants of STAT4 on chromosome 2q32 encoding the transcription factor 'signal transducer and activator of transcription 4' in a region originally highlighted by linkage analysis (Remmers et al. 2007), and of chromosome 1p13 where evidence from linkage analysis and a candidate gene approach resolved a specific nonsynonymous SNP of the intracellular phosphatase gene PTPN22, rs2476601 (c.1858C>T, p.R620W) (Jawaheer et al. 2003; Begovich et al. 2004). Intriguingly this variant also shows association with susceptibility to a number of other autoimmune diseases. PTPN22 was highlighted in recent genome-wide association studies of rheumatoid arthritis (section 9.3.2) that have also demonstrated significant association with chromosome 9q33-q34, a region which includes tumor necrosis factor receptor-associated factor 1 (TRAF1) and complement component 5 (C5) (Plenge et al. 2007b). A clearer picture is beginning to emerge with other recently reported disease associations including chromosome 6q23 near TNFAIP3 (encoding tumor necrosis factor-alpha-induced protein 3) (Plenge et al. 2007a; Thomson et al. 2007), CD40 (Raychaudhuri et al. 2008) and REL (encoding a member of the NF-kB transcription factor family) (Gregersen et al. 2009) which suggest a key role for the CD40 signalling pathway in the pathogenesis of rheumatoid arthritis.
and environmental factors may be involved. Genetic susceptibility loci in such diseases often involve several independent genetic regions, constitute individually a modest magnitude of effect, and may be confounded by population stratification and other risk factors (Section 2.5). Linkage disequilibrium or coinheritance between genetic polymorphism (Box 2.8) is particularly pronounced in the MHC. This, combined with the level and complexity of polymorphism, the heterogeneity and density of genes in this region, and a paucity of knowledge about the biological function of individual genes and the functional consequences of genetic diversity, have all contributed to a difficult task. In many ways the MHC has provided a paradigm for complex disease genetics: as we will see in the remainder of this chapter there have been success stories but further levels of complexity arise as we look deeper into the story.
Was this article helpful?