Abstract
Objective. Vitamin D (VitD) has immunomodulatory activity relevant to rheumatoid arthritis (RA) and acts by binding nuclear receptors that regulate gene transcription. VitD receptor polymorphisms have been variably associated with RA. Because North American Native (NAN) populations have a high prevalence of RA with a strong genetic contribution, we studied potential associations of the rs2228570 (Fok1) VitD receptor polymorphism in a Canadian NAN population.
Methods. The single-nucleotide polymorphism (SNP) Fok1 was tested by sequencing NAN patients with RA (n = 448) and unrelated NAN controls (n = 704). Associations were tested using genotypic, dominant, and recessive models.
Results. The minor allele frequency (F/C) in the NAN control population was 0.44 and lower than reported in white subjects of the same geographical area. The Fok1 VitD receptor SNP was significantly associated with RA. Comparing patients with RA to unaffected NAN controls, the Fok1 SNP was associated with RA using both genotypic [FF vs Ff vs ff: RA 20%, 54%, 26% vs control 22%, 44%, 34% (chi-square 13.35, p = 0.003)] and dominant models [FF/Ff vs ff: RA 74% vs 26% control 66% vs 34% (OR 1.5, 95% CI 1.16–1.96, p = 0.003)]. This association was strongest in shared-epitope-positive RA.
Conclusion. VitD receptor polymorphisms may contribute to the high prevalence of RA in NAN populations.
Vitamin D (VitD) has multiple effects on immune cell function, is important for normal host responses to infection, and influences bone metabolism and tumor proliferation. These mechanisms are all potentially important for the pathogenesis of inflammatory arthritis. The activity of 1,25 OH VitD is mediated through VitD nuclear receptors that, in complex with other nuclear proteins, regulate transcription of VitD-responsive genes1. VitD receptor binding sites are enriched in gene loci associated with autoimmunity and rheumatoid arthritis (RA)2. Both a deficiency of VitD and altered VitD receptor function have been variably linked to susceptibility to infection, malignancy, and autoimmune conditions including RA3,4,5. Several polymorphisms of the VitD receptor have been described. Only the Fok1 VitD receptor polymorphism has been associated with the development of RA in some populations6,7, while other polymorphisms may associate with early disease onset8 or activity9.
North American Native (NAN) populations have high RA prevalence rates, early onset of severe seropositive disease10, and a high frequency of HLA-DRB1 shared-epitope (SE) alleles in the background population11. Environmental factors associated with RA risk are also prevalent in this population, and population studies have documented low serum VitD levels in multiple North American populations including some NAN communities12,13.
We examined the association of the rs2228570 (Fok1) polymorphism of the VitD receptor gene with RA susceptibility in the NAN population of central Canada. The Fok1 F(C) allele was associated with RA in this NAN population, particularly in the setting of shared epitope.
MATERIALS AND METHODS
Subjects and procedures
Cree/Ojibway NAN patients with RA (n = 448) were recruited from clinic populations. All met American College of Rheumatology criteria for RA14. Cree/Ojibway controls (n = 704) were recruited from the same geographic areas. Controls were interviewed and examined by a rheumatologist (DBR or HSG) or trained research nurse and did not have clinical inflammatory arthritis or another autoimmune condition. All subjects had at least 3 of 4 grandparents of NAN origin (99% had 4 of 4) by self-report.
Blood samples were obtained at enrollment, stored at −80°C, and processed according to a standard procedure.
All study subjects provided informed consent. All aspects of the study were approved by the Research Ethics Board of the University of Manitoba and by the Band Councils of the study communities.
Rheumatoid factor (RF) and anticitrullinated protein antibodies (ACPA) testing
Blood samples were tested for RF by nephelometry or ELISA. Positive cutoffs were preset to be < 5% of a control white population (IgM RF > 50 IU/ml). Serum samples from RA patients and controls were tested for the presence of anticyclic citrullinated protein antibodies (ACPA) using a commercial anti-ACPA2 ELISA (Inova Diagnostics, San Diego, CA, USA).
HLA-DRB1 allele testing
HLA-DRB1 typing was performed by polymerase chain reaction using sequence-specific oligonucleotide primers in 353 patients with RA and 392 controls. The following DRB1 alleles were included as SE-bearing alleles: DRB1*0101, 0102, 0401, 0404, 0405, 0408, 0410, 1001, and 140211.
VitD receptor single-nucleotide polymorphism (SNP) testing
Genotyping was undertaken in the laboratory of Dr. Siminovitch. The multiplexed SNP assays were performed on the Sequenom Mass Array iPLEX platform (Sequenom, San Diego, CA, USA). Allele-specific extension products were plated onto a SpectroCHIP (Sequenom), subjected to mass spectrometric analysis, and the genotypes identified using SpectroCaller software (Sequenom). The VitD receptor gene is located on chromosome 12 (12q13.11). The Fok1 polymorphism is located at a translation initiation site in exon 2 at the 5’ gene region and is not in linkage disequilibrium with several other known VitD receptor polymorphisms located in noncoding introns and exon 9 at the 3’ gene region15.
Statistical analysis
Associations of VitD receptor gene polymorphisms and RA phenotype were analyzed using genotypic (FF vs Ff vs ff), allelic (F vs f), dominant (FF, Ff vs ff), and recessive (FF vs Ff, ff) models. Associations were tested by chi-square and regression methods using SPSS15. Attributable risk of Fok1 to RA was determined as the incidence of RA in the population with at least 1 dominant Fok1 allele minus the incidence of RA in the population without a dominant Fok1 allele16.
RESULTS
The clinical characteristics of subjects are shown in Table 1. As reported in this population, NAN have a high frequency of RF and ACPA-positive RA and there is a high frequency of RF and SE in the background population10,11. Patients with RA were older than the controls [mean 47 yrs (SD 15) vs 35 yrs (SD 12); p < 0.0001] and more likely to be female (82% vs 59%, respectively; p < 0.0001).
Clinical characteristics of North American Native patients with rheumatoid arthritis (RA) and Native controls. Results are reported as percentage positive, mean (SD), or median (interquartile range; IQR) as indicated. All comparisons between RA and controls, p < 0.0001.
VitD receptor Fok1 polymorphism is associated with RA in NAN
The minor allele frequency (F/C) in the NAN control population was 0.44, similar to that of previous reports of Fok1 in the Manitoba NAN population (0.43)17, and the allele frequencies were in Hardy-Weinberg equilibrium. The observed minor allele frequency in NAN was lower than that reported in the Manitoba white population (0.64) and in other ethnic groups18,19,20. Comparing RA patients to unaffected NAN controls, the Fok1 SNP distribution was associated with RA using both genotypic [FF vs Ff vs ff: RA 20%, 54%, 26% vs control 22%, 44%, 34% (chi-square 13.35, p = 0.003)] and dominant models [FF/Ff vs ff: RA 74%, 26% vs control 66%, 34% (OR 1.5, 95% CI 1.16–1.96, p = 0.003; Table 2]. The FF or Ff genotype was more common in NAN patients with RA than in controls. No associations were seen with the allelic or recessive models.
Vitamin D receptor single-nucleotide polymorphism rs2228570 (Fok1) is increased in North American Native patients with rheumatoid arthritis (RA) compared to unaffected Native controls. Values reported are n (% of RA or control); OR (95% CI).
Associations between RA and Fok1 were significant for SE-positive but not SE-negative subjects (Table 3), regardless of sex. This was seen for subjects with at least 1 copy of any SE allele in genotypic (FF vs Ff vs ff: chi-square 11.3, p = 0.003) and dominant models (FF or Ff vs ff: chi-square 9.8, p = 0.002), but not in recessive models. No additive or multiplicative interactions between SE and Fok1 were present in genotypic, dominant, or recessive models testing the association of Fok1 and RA. In multivariate binary logistic models controlling for age, sex, and SE, having at least 1 copy of the dominant Fok1 allele (FF or Ff) was independently associated with RA (OR 1.7, 95% CI 1.2–2.5, p = 0.002). The attributable risk of having at least 1 copy of the dominant Fok1 allele for RA in this NAN population was 0.098 (10%). In the NAN subset population with at least 1 copy of SE, the attributable risk was 0.084 (8%). In contrast, the attributable risk of at least 1 copy of SE to RA in this population was 0.22 (22%).
Vitamin D receptor single-nucleotide polymorphism rs2228570 (Fok1) and shared-epitope (SE) status in patients with rheumatoid arthritis (RA) and controls. Values reported are n (% of RA or control); OR (95% CI).
Small sample size reduced the power of the study to detect significant associations of SE alleles and Fok1. In this population (both controls and patients with RA), the presence of any SE allele was associated with the Fok1 allele in genotypic (ff/SE-negative 46/169, Ff/SE-negative 90/169, FF/SE-negative 33/169 vs ff/SE-positive 202/576, Ff/SE-positive 247/576, FF/SE-positive 127/576; chi-square 5.9, p = 0.052) but not dominant models (ff/SE-negative 46/169, Ff or FF/SE-negative 123/169 vs ff/SE-positive 202/576, FF or Ff/SE-positive 374/576; chi-square 3.6, p = 0.06). A modest association was found between having at least 1 copy of the HLA-DRB1*0101 SE allele and Fok1 receptor status (using the dominant model) in the entire study population (RA and controls: ff/HLA*0101-negative 195/662 vs FF or Ff/HLA*0101-positive 29/51; chi-square 4.2, p = 0.04, without correction for multiple comparisons), but not in the subset that had at least 1 copy of any SE allele (ff/HLA*0101-negative 159/503 vs FF or Ff/HLA* 0101-positive 29/51; chi-square 2.8, p = 0.09). The presence of 2 copies of DRB1*1402 was associated with the Fok1 allele using the dominant model (chi-square 6.1, p < 0.05, without correction for multiple comparisons; Table 4).
Associations between individual HLA-DRB1 shared-epitope (SE) alleles and Fok1 polymorphisms in the North American Native (NAN) population including patients and controls, and the subset of NAN with at least 1 HLA-DRB1 SE allele.
VitD receptor polymorphisms were not associated with RF or ACPA positivity in NAN patients with RA. However, in NAN controls with the SE, Fok1 was associated with RF: 14/18 (78%) FF, 18/31 (58%) Ff, and 23/61 (38%) ff were positive for RF (chi-square 10, p = 0.007) and 32/49 (65%) subjects with FF or Ff were positive for RF compared to 23/61 (38%) subjects with ff (chi-square 8.3, p = 0.004). No associations were seen with ACPA in NAN controls.
DISCUSSION
The VitD receptor Fok1 polymorphism was shown to be associated with RA in our NAN study population. Compared to the NAN control population, NAN patients with RA were more likely to be homozygous or heterozygous for the F(C) allele. Previous studies in white populations have also shown an association with the F(C) allele and RA, and that this association may be strongest for individuals with HLA-DRB1 SE alleles6. The frequency of the Fok1 F(C) allele in our NAN control population was lower than that reported in unaffected white subjects from the same geographical area (0.64)17 and other European21 and African American (0.78)22 populations.
Polymorphisms at the Fok1 site, located in a coding exon at the 5’ region of the VitD receptor gene, result in different protein structures. The Fok1 f(T) allele encodes a 427-amino acid receptor, and has been variably associated with an increased risk of several malignancies, especially in the context of environmental factors23,24. The F(C) allele encodes a smaller (424 amino acids) possibly more active protein product25 that appears to be associated with greater bone mineral density. However, NAN have lower site-specific bone mineral density26 and low serum VitD compared to whites27. The biological effects of the Fok1 VitD receptor polymorphisms may differ across VitD-responsive genes and cell types and may be influenced by serum VitD levels and other VitD receptor polymorphisms18,23.
Other VitD receptor polymorphisms have been identified in the promoter, exons 2–9, introns, and the 3’ region of the gene23 and were initially identified using restriction enzymes. Bsm, Apa1, EcoRV, and Tru9I restriction fragment length polymorphisms (RFLP) are located in noncoding introns at the 3’ region of the VitD receptor gene. Taq1 is located in exon 9, also at the 3’ region. These polymorphisms (Bsm1, Apa1, EcoRV, Tru91, and Taq1) are in linkage disequilibrium with each other, but not with Fok1. The Cdx2 polymorphism is located in the 5’ promotor region. To date, no functional consequences of these RFLP have been reported, although they may affect RNA stability and thus gene expression21.
The Bsm1, Taq1, and Apa1 VitD receptor RFLP have been studied in RA6,28. While no clear associations with disease prevalence have been demonstrated, there are suggestions that these RFLP may influence disease characteristics. In a Spanish population, although no differences in the distribution of RFLP between patients with RA and controls were observed, the combination of homozygous absence of Bsm1 and presence of Taq1 was associated with slightly earlier onset of RA, particularly for SE-positive women8. A separate study reported an association between Bsm1 RFLP and RA disease activity9. Multiple polymorphisms may influence disease features or susceptibility21. Further genotyping studies are planned to investigate the association of Fok1 and other VitD receptor polymorphisms in RA.
VitD is a key regulator of immune responses and has potent pro- and antiinflammatory effects on multiple immune cells1. Although the mechanisms for these effects are not fully understood, excess VitD receptor activity might evoke more robust innate immune responses to bacteria, resulting in a greater potential to break self-tolerance. Increased activity of this receptor may also alter the balance between immunostimulatory and immunoregulatory cells. RF titers have been shown to correlate modestly with immune responses to common pathogens in early inflammatory arthritis29, and in NAN controls with the SE, Fok1 appeared to associate with RF, possibly reflecting exposure to pathogens. In contrast, low serum levels of VitD have been associated with RA activity30 and with increased risk of developing RA in some but not all studies, and animal models suggest a functioning VitD receptor is important to limit synovial inflammation31.
Genome-wide association studies (GWAS) conducted primarily in whites of European descent have identified the HLA-DRB1 locus as the most significant genetic contributor to RA disease risk, although several other loci with lower contributions have been identified32,33. While no study has identified a specific association with the VitD receptor gene locus, an association with chromosome 12q13 and RA in the setting of HLA SE34,35 is interesting. The VitD receptor gene is located in this region36, as are the genes for small ubiquitin-like modifier 1-specific protease (SENP1; 12.13.1q), which influences expression of matrix metalloproteinase-1 in RA synoviocytes37 and collagen type 2 (COL2a; 12.13.1–13.2), although the latter was found not to be associated with RA in a small study38. Thus, the observed associations of Fok1 and RA may be due to linkage disequilibrium with other RA-associated genes.
Our finding of an association between the Fok1 F allele and disease risk primarily in SE-positive individuals, while potentially related to the low numbers of SE-negative subjects in this population, also raises the possibility that the autoimmune effects of the variant VitD receptor are enhanced against the background of SE positivity. This is supported by studies of RA family trios in which F(C) transmission was enhanced in the setting of the SE6. VitD receptor polymorphisms have also been shown to cosegregate with disease-specific HLA-DRB1 alleles in other autoimmune conditions such as multiple sclerosis39. In addition, binding sites for the VitD receptor have been found in gene regions associated with autoimmune disease2. Further studies are required to explore this hypothesis, to evaluate the extent to which the polymorphism correlates with VitD levels and function in both f/f homozygotes and F/f heterozygotes, and thereby to delineate the pathway linking this polymorphism to increased risk for RA in the NAN population12.
The association of Fok1 and RA observed in this population requires validation in a separate NAN population, preferably of Cree/Ojibway descent.
Acknowledgment
The authors thank the Assembly of Manitoba Chiefs and the Chiefs and Band Councils of the Norway House and St. Theresa Point communities for their role in facilitating this study. We also thank Donna Hart, RN, for assistance with the study and Drs. Kiem Oen (University of Manitoba, Winnipeg, Manitoba) and Janet Markland (University of Saskatchewan, Saskatoon, Saskatchewan) for assistance with subject recruitment.
Footnotes
-
Supported in part by funding from the Canadian Institutes of Health Research (MOP7770 and IIN 84040). Prof. Siminovitch is supported by a Canada Research Chair and holds the Sherman Family Chair in Genomic Medicine.
- Accepted for publication May 28, 2012.