© 2004 American Thoracic Society
Sarcoidosis and Human Leukocyte Antigen Class I and II GenesIt Takes Two to Tango?
a Henry Ford Health System Detroit, Michigan Classical human leukocyte antigen (HLA) genes come in two flavors: class I and class II. Class I molecules, expressed in virtually all cells, present cytosolic-derived peptides of 8 to 10 amino acids long to CD8+ cytotoxic T cells. Class II molecules, expressed in antigen-presenting cells, such as dendritic cells, present endocytically derived peptides of 11 to 17 amino acids long to CD4+ helper T cells. Proteins coded in the class III gene region, a 1.5 megabase DNA segment between the class I and class II regions, have immunomodulatory effects but are not directly involved in antigen presentation like class I and II proteins. Most HLA genes have multiple alleles and, because of their proximity, distinct HLA allelic combinations are often inherited as a multigene haplotype. HLA association studies in sarcoidosis are not new. The earliest reported association between sarcoidosis and a HLA gene, HLA-A7, is now over 30 years old (1). In the year following that first report, a recurrent frustration in sarcoidosis HLA studies began when independent studies were unable to replicate the initial reported HLA-A7 association (2, 3). To the body of literature accumulated over the past 30 years is added a report in this issue of the Journal (pp. 696–702) by Grunewald and coworkers (4). They demonstrate HLA-B and HLA-DRB1 allelic associations with both persistent and resolving sarcoidosis. At first glance, sarcoidosis associations with HLA-B and HLA-DRB1 would not appear novel, but upon closer inspection one finds that this report may change the present paradigm regarding sarcoidosis and HLA associated risk. In 166 clinically well characterized sarcoidosis patients and 210 healthy controls, Grunewald and coworkers (4) genotyped class I HLA-A, B, and C alleles and the class II HLA-DRB1 and DQB1 alleles. Patients were divided into two phenotypic groups based on persistence or resolution of disease after two years of follow-up. The strongest associations were between the class II HLA-DRB1*03 and disease resolution (p < 0.0001) and HLA-DRB1*15 alleles and disease persistence (p = 0.02). This, however, was not the main story the authors decided to tell. Independent of these two HLA-DRB1 associations, the authors found a twofold increased risk of disease (both resolved and persistent) associated with the class I HLA-B*07 allele. The HLA-B*08 allele had a similar, albeit statistically insignificant, effect on disease risk. Brewerton and coworkers (5) first reported an HLA-B8 association with acute sarcoidosis that was later confirmed by other groups (6–8). Some of these later studies also noted that the HLA-B8 allele was inherited as a HLA-B8/DR3 haplotype in white individuals. Subsequent studies (9, 10), including that of Grunewald and coworkers (4), have replicated the HLA-B8/DR3 haplotype association with acute sarcoidosis. Two possible explanations exist why HLA class I and class II alleles are both associated with sarcoidosis. Class I alleles might independently confer risk suggestive of a more complex immunopathogenic pathway in sarcoidosis. For instance, recent reports indicate that CD8+ T cells functioning in the context of the HLA class I molecules may play a critical role in processing intracellular mycobacterial infection (11). The other explanation is that HLA class I allelic associations are derived from linkage disequilibrium with HLA class II alleles associated with sarcoidosis. Hence the HLA association conundrum: which linked HLA genes predispose to disease? An extreme example of HLA associations confounded by linkage disequilibrium is the cosegregation of HLA-DRB1 and HLA-DQB1 allelic pairs in white individuals. Grunewald and coworkers (4) show that in their study population the associations of HLA-DRB1*03 with resolved disease and HLA-DRB1*15 with persistent disease are synonymous with the HLA-DQB1*0201–resolved disease and HLA-DQB1*0602–persistent disease associations. The latter results replicate HLA-DQB1 sarcoidosis disease outcome associations reported in a recent study performed in a British and Dutch population (12). Consequently, determining whether HLA-DRB1 or HLA-DQB1 is of primary importance in sarcoidosis may be an intractable problem in white individuals, particularly for studies of modest sample size. Linkage disequilibrium in white populations also confounds disease-allele associations involving both HLA class II and class I genes, but to a lesser degree than genes so closely linked that recombination rarely occurs between them, such as HLA-DRB1 and HLA-DQB1. Grunewald and coworkers (4) show that the confounding influence of linkage disequilibrium from DRB1-related associations with sarcoidosis can be observed in part by the change between the unadjusted and adjusted odds ratios of the HLA-A and B alleles (see Table 4 in Reference 4). The true effect of both class I and class II HLA alleles on disease risk can only be estimated in a population where complete linkage equilibrium exists between the class I and class II regions. In this regard, populations of African ancestry may provide a better resource for HLA association studies (13). Despite the residual linkage disequilibrium that exists in the HLA region in white individuals, Grunewald and coworkers (4) have been able to tease apart sarcoidosis risk effects for both HLA class I and II alleles. Clearly, the report of Grunewald and coworkers (4) supports the assertion from other recent notable studies (14, 15) that the class II HLA-DRB1 (or -DQB1) gene is a key player in sarcoidosis susceptibility, but they also show a significant effect for the HLA-B*07 allele after taking the effect of HLA-DRB1 into account. Furthermore, Grunewald and coworkers (4) show that any increased risk the alleles HLA-A*03 and -B*07 exert toward persistent sarcoidosis requires the presence of the HLA class II DRB1*15 allele. Therefore, what is novel, and most important, to take away from the report of Grunewald and coworkers (4) is that class I HLA together with class II may play complementary roles in the promotion of the sarcoidosis disease process. Grunewald and coworkers (4) have provided us a new way to think about HLA class I and II genes with regard to sarcoidosis risk and progression. The full extent of HLA gene involvement in sarcoidosis susceptibility and progression will ultimately require large study samples drawn from populations with little or no linkage disequilibrium across the HLA region. FOOTNOTES Conflict of Interest Statement: B.A.R. and M.C.I. have no declared conflict of interest. REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
