|Home | About | Journals | Submit | Contact Us | Français|
CLEC16A, a putative immunoreceptor, was recently established as a susceptibility locus for type I diabetes and multiple sclerosis. Subsequently, associations between CLEC16A and rheumatoid arthritis (RA), Addison’s disease and Crohn’s disease have been reported. A large comprehensive and independent investigation of CLEC16A variation in RA was pursued. This study tested 251 CLEC16A single-nucleotide polymorphisms in 2542 RA cases (85% anti-cyclic citrullinated peptide (anti-CCP) positive) and 2210 controls (N = 4752). All individuals were of European ancestry, as determined by ancestry informative genetic markers. No evidence for significant association between CLEC16A variation and RA was observed. This is the first study to fully characterize common genetic variation in CLEC16A including assessment of haplotypes and gender-specific effects. The previously reported association between RA and rs6498169 was not replicated. Results show that CLEC16A does not have a prominent function in susceptibility to anti-CCP-positive RA.
Rheumatoid arthritis (RA) is the most common systemic autoimmune disease with a prevalence of 1%.1 This chronic inflammatory disease can cause substantial disability from the erosive and deforming processes in joints, and is associated with increased mortality.2 RA has a strong genetic component, as shown by twin and other family studies; however, the etiology is unknown.3 Major histocompatibility complex genes, particularly HLA class II, are strongly associated with risk of developing RA. However, major histocompatibility complex genes only account for a portion of the genetic risk. Several non-major histocompatibility complex genes have recently been associated with risk for RA, including PTPN22, STAT4 and TNFAIP3.4–6 Results from recent genome-wide association (GWA) studies underscore the overlap of replicated findings across complex diseases, including autoimmune conditions.7,8 Variants within some confirmed genetic risk loci for RA also confer risk for other autoimmune diseases. These include CTLA4 in type I diabetes (T1D), IL-2 in T1D and Celiac disease, PTPN22 in systemic lupus erythematosus (SLE), T1D and autoimmune thyroid disease, STAT4 in SLE and TNFAIP3 in SLE, T1D, Celiac disease and Crohn’s disease.5,9–19
The C-type lectin domain family 16, member A gene (CLEC16A, previously called KIAA0350) spans 237.7 kb and encodes a sugar-binding receptor that contains a putative immunoreceptor tyrosine-based activation motif.10 C-type lectin receptors can be expressed on dendritic cells to distinguish between self and non-self glycoproteins, and may be involved in immune activation and peripheral tolerance.20,21 These sugar-binding receptors have been shown to be important in multiple animal models for RA.22–25 For example, in rats, C-type lectin-like receptors are encoded by the antigen-presenting lectin-like receptor gene complex (APLEC), which have been shown to influence susceptibility to arthritis (oil-, collagen-, squalene- and pristine-induced), auto-immune phenotypes (autoantibody levels) and clinical phenotypes (day of disease onset, maximal severity, severity over time, body weight loss, arthritis symptoms).24 The effect of APLEC variation on susceptibility to arthritis and clinical phenotypes varied by gender.24
Recently, GWA studies have identified the sugar-binding receptor gene CLEC16A as a novel risk locus for T1D and MS, and this association has since been replicated in independent samples.10,26–31 CLEC16A is located on 16p13, a region that has been implicated in RA linkage studies.32 The purpose of this study was to perform a comprehensive haplotype-based investigation of CLEC16A as a candidate RA gene. This study sample consisted of 682 RA cases and 752 controls collected by the North American RA Consortium (RA1), 1860 RA cases collected by the Wellcome Trust Case Control Consortium (WTCCC) RA Group in the UK and 1458 controls collected by the WTCCC from the UK Blood Services (RA2) (total N = 4752) (Table 1).
We conducted allelic tests of association for 58 single-nucleotide polymorphisms (SNPs) and global haplotype tests (12 haplotype blocks encompassing 53 SNPs) in 682 anti-cyclic citrullinated peptide-positive (anti-CCP-positive) RA cases and 752 controls (N = 1434 (RA1)) (Figure 1). All results were negative after correcting for multiple testing (Figure 2, Supplementary Table 1). Next, we conducted allelic tests of 43 SNPs and global haplotype tests (7 haplotype blocks encompassing 37 SNPs) in the second RA data set composed of 1860 RA cases and 1458 controls (N = 3318 (RA2)). No evidence for association was present (Figure 2, Supplementary Table 1). Furthermore, allelic tests of 251 imputed SNPs within CLEC16A derived for the combined RA sample (2542 cases and 2210 controls, total N = 4752 (RA1 + RA2)) revealed no evidence for disease association (Figure 2, Supplementary Table 1).
The six CLEC16A SNPs shown to be associated with T1D and/or MS are intronic and were either genotyped or tagged (r2>0.95 based on the Caucasian HapMap population (CEU)) in this study. Similar to this study, candidate gene investigations of CLEC16A in Grave’s disease, Celiac disease and ulcerative colitis have been negative, but associations have been reported with Addison’s disease, Crohn’s disease and for RA in other data sets.10,29,33–36 A case–control study by Martinez et al.29 examined three CLEC16A SNPs and reported that rs6498169*G, a variant associated with MS, was over-represented in RA cases (38%) compared to controls (32%) (P = 8 × 10−3, odds ratio (OR) = 1.27, 95% confidence interval (CI) = 1.06–1.51). Although our study was well powered to detect such an effect size, with 80% power to detect an OR as low as 1.13, the association between RA and rs6498169 was not replicated. The rs6498169*G allele frequency did not differ between RA cases (33.6%) and healthy controls in this study (32.9%) (P = 0.45, OR = 1.03, 95% CI = 0.95–1.11).
It is also important to note that recent studies have revealed the presence of different major histocompatibility complex associations in anti-CCP-positive and anti-CCP-negative RA cases when considered separately.37–39 It is possible that this phenotypic difference may also be important for other RA genetic susceptibility loci. The well-established PTPN22 RA locus appears to be associated only with anti-CCP-positive RA, although some studies have reported association with both anti-CCP-positive and anti-CCP-negative RA.40–43 Anti-CCP autoantibodies and shared epitope alleles are also markers for increased RA severity, particularly when both are present.44 In this study, 85% of RA cases were anti-CCP positive, compared to only 50% in the Martinez et al. study. This difference may have contributed to the observed disparity between results. Indeed, Skinningsrud et al.36 have recently examined three CLEC16A SNPs and reported that the rs6498169*G variant was over-represented in anti-CCP-negative RA cases (44%) compared to anti-CCP-positive RA cases (37.7%) (P = 0.016, OR = 1.3, 95% CI = 1.05–1.61) and controls (35.9%) (P = 2 × 10−4, OR = 1.4, 95% CI = 1.18–1.68). Martinez et al. did not observe differences between cases and controls after stratifying for anti-CCP status or presence/absence of shared epitope alleles, but this may be due to a lack of statistical power. Although all of our RA1 cases were anti-CCP positive, only 80% of RA2 cases were anti-CCP positive and this information was not publicly available for the RA2 cases. Therefore, we were not able to stratify RA2 or RA1 + RA2 by anti-CCP status for analyses of CLEC16A SNPs.
Because animal models suggest that C-type lectin receptor genes may have gender-specific effects on autoimmunity, we conducted gender-stratified allelic tests and gender-adjusted global haplotype tests of CLEC16A within RA1 and RA2.24 The rs3960630 A variant was underrepresented in female RA1 cases (20%) compared to female controls (25%) (OR = 0.71, 95% CI = 0.59–0.86, P = 4 × 10−4). This intronic SNP was not present in or captured by RA2 data and therefore could not be tested in the larger combined data set. Given the number of multiple tests performed, these results should be interpreted with caution. Results did not differ when global haplotype tests were adjusted by gender (data not shown). Animal models of RA also indicate that it may be worthwhile to stratify cases by clinical phenotypes in future genetic studies of C-type lectin receptors and autoimmunity.24
Although rare variants in CLEC16A were not directly investigated here, for the first time all common genetic variation within CLEC16A was interrogated for a function in RA susceptibility. Even without imputed genotypes, the RA1 data set (N = 58 SNPs) captured 93%, RA2 (N = 43 SNPs) captured 80% and both data sets combined (N = 96 SNPs) captured 96% of the common variation based on CEU data from HapMap (see Figure 2 legend). The data used in this study were taken from GWA studies that did not identify CLEC16A as a risk locus for RA based on stringent genome-wide significance. A focused candidate gene study that captures a larger portion of genetic variation compared to initial GWA studies is a useful and complementary strategy.
In conclusion, this is the first candidate gene study of CLEC16A to fully characterize common genetic variation in CLEC16A, including assessment of haplotypes and gender-specific effects. We did not replicate the association between RA and rs6498169 reported by other studies. Results convincingly show that variation within CLEC16A does not have a prominent function in susceptibility to anti-CCP-positive RA.
We thank Farren BS Briggs, Benjamin A Goldstein, Alan Hubbard and Ira Tager for helpful discussion, as well as study participants. This work was supported by an Abbott Graduate Student Achievement Award (ACR REF), Grants R01 AI065841, R01 AI059829 and F31 AI075609 (NIH/NIAID), and Grants RO1 AR44422, NO1 AR22263, R01 AR050267 and K24 AR02175 (NIH/NIAMS). The contents of this paper are solely the responsibility of the authors and do not necessarily represent the official views of the NIH, NIAID or NIAMS. This study makes use of data generated by the WTCCC; a full list of the investigators who contributed to the generation of the data is available at www.wtccc.org.uk, and funding for the project was provided by the Wellcome Trust under award 076113. These studies were performed in part in the General Clinical Research Center, Moffitt Hospital, University of California, San Francisco, with funds provided by the National Center for Research Resources, 5 M01 RR-00079, US Public Health Service.