The North American Rheumatoid Arthritis Consortium (NARAC) recently reported on two new RA linkage regions at chromosomes 2q33 and 11p12 with logarithm of odds (LOD) scores of 3.52 and 3.09, respectively (20
). Dense SNP mapping of the 2q RA linkage peak led to identification of a new susceptibility gene, STAT4, for RA (16
). This association was replicated in several independent White RA case and control populations (16
). In the current study, we have now confirmed that STAT4 is associated with RA in a large Korean population dataset, with the same common haplotype, which is more common in Koreans than in North American populations, but nevertheless confers a similar degree of risk.
A significant source of variability in the RA genetics literature has been the inability to replicate genetic findings across the major racial groups, particularly Whites and Asians. An interesting example of this is the association of the intra-cellular phosphatase, PTPN22, with RA and other autoimmune diseases. These disease associations have been widely replicated in White populations(7
), but the PTPN22 risk allele (R620W) is exceedingly rare in Asian populations(15
). Furthermore, attempts to identify other risk variants of PTPN22 that might be associated with RA in Asians have been unrevealing (9,H-S Lee unpublished). This has raised the possibility that there is true locus heterogeneity for RA among these major racial groups.
Several other examples have arisen in which associations are observed in Asian populations, but not in Whites. The most robust of these examples is the association of PADI4 with rheumatoid arthritis. PADI4 is a compelling candidate gene, because it encodes one of the enzymes responsible for citrullination of endogenous proteins, and an antibody response to citrullinated peptides is highly specific for RA(21
). Numerous studies in Asian populations have demonstrated the association of PADI4 with RA (10
), but these associations are either absent or very weak in populations of European ancestry (13
). It is possible that this difference reflects an interaction of PADI4 genetic susceptibility with environmental factors, because citrullination may also be related to smoking or other environmental exposures (1
The current report is the first clear demonstration of non-MHC related susceptibility gene for RA that confers a similar degree of risk among both White and Asian populations. Furthermore, it appears that the risk haplotype is likely to be identical in the two racial groups, suggesting that the responsible functional variant is ancient in origin. Indeed, the same haploytpe also is found in African populations (TTCG, 0.14 in the Yoruba people of Ibadan, Nigeria, www.hapmap.org
) and it will be of great interest to see if the STAT4 associations with RA and lupus are also present in this population group . The associated haplotype is located primarily in the third intron of the STAT4 gene, and the actual functional allele(s) remain to be identified. A full resequencing of the STAT4 gene is in progress, and this will help to direct future studies of splice variation and/or expression differences that may explain the disease associated haplotype.
encodes a transcription factor that lies in the signaling pathway of several important cytokines, including IL-12 and type I interferons, as well as IL-23 (27
). STAT4 is present in the cytosol and upon cytokine signaling it becomes phosphoryated and translocates to the nucleus. The target genes for STAT4 include γIFN and therefore it plays a key role in the IL12 induced differentiation of T cells into the Th1 pathway. In addition, STAT4 may also be involved in the production of IL17 by Th17 cells, in response to IL23 (28
). At the same time, γIFN production tends to inhibit differentiation toward the Th17 pathway(29
). Thus, while genetic differences in STAT4 dependent signaling may be involved in regulating the balance of Th1 vs. Th17 responses, the expected effects of increased vs. decreased STAT4 activity are not obvious. Furthermore, compared with the mouse, the production of IL17 is not as clearly restricted to an easily definable Th17 subset in humans (30
), and recent work suggests that some subsets of CD4 cells can produce both IL17 and γIFN (29
). Therefore, it will be important to try to relate the major risk haplotype of STAT4 to phenotypic differences in these various T cell subsets.
Given the current evidence of Th17 involvement in chronic inflammation in RA (29
), it would be expected that the STAT4 risk alleles would generally enhance a Th17 response. Several studies in animal models strongly suggest a key regulatory role of STAT4 in experimental arthritis (32
). There is evidence that STAT4 may play a role both during initiation of disease as well as in the maintenance of the inflammatory process, leading to the idea that STAT4 may be a useful therapeutic target as has been demonstrated in murine collagen induced arthritis (32
). In this context, it may be relevant that STAT4 also is involved in signaling responses to type 1 interferons in activated monocytes, macrophages, and mature dendritic cells. The dependence of type 1 interferon signaling on STAT4 is of particular interest in view of our recent observation that in addition to RA, STAT4 has a strong association with systemic lupus (16
), a disease in which dysregulation of interferon pathways is prominent.
The finding of a common risk haplotype for RA among Asian and White populations shows that replication studies across racial boundaries may be useful for confirming at least some risk loci. In view of the large number of potential risk alleles that are coming out of whole genome scans (33
), our data suggest that comparison across racial groups can be a reasonable approach to gene identification, and we hope to be able to carry out a genome-wide study of the Korean RA population in the near future.