Objective: To examine the relationship between SNP +39604 in SEEK1 and psoriatic arthritis (PsA) in two distinct Canadian populations.
Methods: 103 patients with PsA and 105 ethnically matched controls from Newfoundland and 202 patients with PsA and 100 controls from Ontario were studied. Patients and controls were genotyped for SNP +39604 of SEEK1 by time of flight mass spectrometry, using the Sequenom platform. Genomic DNA was amplified by the Dynal RELI SSO HLA- Cw* typing kit for HLA-C typing.
Results: The frequency of the minor SEEK1(T) allele in subjects with PsA and controls was 48.5% and 32.4%, respectively (odds ratio (OR) = 2.0; p = 0.017), in the Newfoundland population and 46.5% and 38.0%, respectively (OR = 1.4; p = 0.16), in the Ontario population. Although SEEK1 is associated with PsA, particularly in the Newfoundland population, multivariate analysis showed that SEEK1 does not seem to be a further susceptibility factor if the HLA-Cw*0602 status is already known. No association was noted between SEEK1(T) allele and onset of psoriasis, PsA, or arthritis pattern.
Conclusion:SEEK1 is associated with PsA in the Newfoundland founder population. This association is probably due to linkage disequlibrium between SEEK1 and HLA-Cw*0602 in this population.
- OR. odds ratio,
- PCR, polymerase chain reaction
- PsA, psoriatic arthritis
- SNP, single nucleotide polymorphism
- founder population
- psoriatic arthritis
Statistics from Altmetric.com
- OR. odds ratio,
- PCR, polymerase chain reaction
- PsA, psoriatic arthritis
- SNP, single nucleotide polymorphism
As psoriasis and psoriatic arthritis (PsA) are interrelated disorders, these entities, at times, share common genetic determinants.1 The association of the HLA-Cw*0602 allele with psoriasis and PsA is one such example. HLA-Cw*0602 has long been considered to be the major predisposing factor for psoriasis,1,2 and the prevalence of this allele is also increased among subjects with PsA.3 Furthermore, in both these entities, the presence of HLA-Cw*0602 predicts an earlier age of onset of psoriasis.4–6 However, whether the HLA-Cw*0602 allele actually represents the true psoriasis gene or is in linkage disequilibrium with the true gene has not been adequately elucidated.1,2SEEK1, which overlaps the 6p21-23 region and is in close proximity to the HLA-C locus, was recently noted to have a strong association with psoriasis in the Swedish population.7 In that study of 87 subjects with uncomplicated psoriasis, the +39604 polymorphism in exon 2 of SEEK1 was found in 66% of patients with psoriasis as compared with 22% of controls (p = 0.000001). This polymorphism exhibited the strongest association with psoriasis and its effect was independent of HLA-Cw*0602. Because of the striking association of single nucleotide polymorphism (SNP) +39604 in SEEK1 and psoriasis, we set out to determine the prevalence of this SEEK1 polymorphism in two distinct Canadian populations.
PATIENTS AND METHODS
This study was approved by the local ethics committees at Memorial University of Newfoundland and the University of Toronto. Informed consent was obtained from all patients. PsA was diagnosed as an inflammatory arthritis in patients with psoriasis, in the absence of other causes for inflammatory arthritis. Information was collected systematically and included age at onset of psoriasis, PsA, and disease pattern. The control subjects were obtained from their respective regions, and were all white and unrelated.
Blood samples were collected from volunteers in EDTA anticoagulant, and DNA was extracted from peripheral blood lymphocytes using the Wizard Genomic DNA Purification Kit from Promega (Madison, WI). Subjects with PsA and controls were genotyped for the SEEK1 polymorphism by time of flight mass spectrometry, using the Sequenom platform. The polymerase chain reaction (PCR) primers were designed using MassARRAY assay design software, version 1.3.4. Mass array assay design was as follows: PCR primer 1: ACGTTGGATGTGCAACAGAAACCATCACCC; PCR primer 2: ACGTTGGATGTATGTGCTGGGAACTCAGTG. The primer sequences were obtained from Integrated DNA Technologies (Coralville, IA).
For HLA genotyping, 200 ng of genomic DNA was amplified using the Dynal RELI SSO HLA-Cw* typing kit. PCR amplicons were identified by a reverse line assay using sequence-specific oligonucleotide probes. Assay results were interpreted using the pattern matching program provided by Dynal.
Logistic regression was used to study the relationship between genotyping information and case-control status. The results are summarised as odds ratios (ORs) and significance tests. The association between age at onset of psoriasis and age at onset of PsA was examined semiparametrically.8 A χ2 test was used to examine the relationship between the SEEK1(T) polymorphism and disease pattern.
One hundred and three patients with PsA (42% women) and 105 ethnically matched controls from Newfoundland were assessed. The mean (SD) age of the patients with PsA from Newfoundland was 49.6 (10.7) years, mean (SD) age of onset of psoriasis was 31.0 (14.1) years and of PsA 38.0 (10.8) years. Of the 103 patients with PsA cases who were genotyped for SNP +39604, there were 3 homozygotes for the mutant T allele (TT), 47 heterozgotes (TC), and 53 homozygotes for wild-type C allele (CC). Of the 105 Newfoundland controls genotyped, there were 3 homozygotes for the mutant T allele (TT), 31 heterozygotes (TC), and 71 homozygotes for the wild-type C allele (CC). The frequency for the minor (T) allele for the +39604 SNP was 48.5% in the Newfoundland patients with PsA compared with 32.4% in controls (OR = 2.0; p = 0.02).
Two hundred and two patients with PsA (41% women) and 100 ethnically matched controls from Ontario were genotyped. The mean (SD) age of the patients with PsA from Ontario was 50.5 (13.2) years, mean (SD) age of onset of psoriasis was 26.9 (12.2) years and of PsA 32.5 (10.8) years. Of the 202 patients with PsA who were genotyped for SNP +39604, there were 17 homozygotes for the mutant T allele (TT), 77 heterozgotes (TC), and 108 homozygotes for wild-type C allele (CC). Of the 100 Ontario controls genotyped, there were 6 homozygotes for the mutant T allele (TT), 32 heterozygotes (TC), and 62 homozygotes for the wild-type C allele (CC). The frequency for the minor (T) allele for the +39604 SNP was 46.5% in the Ontario patients with PsA compared with 38.0% in controls (OR 1.4; p = 0.16).
A combined analysis of the Newfoundland and Ontario data demonstrated an overall association between the SEEK1(T) allele and PsA (p = 0.009). Additionally, there was no evidence for a SEEK1(T) allele and geographic location interaction (p = 0.39). Thus it cannot be established that the role of the SEEK1(T) allele in the incidence of PsA differs in the two populations. In separate analyses, the effect of HLA-Cw*0602 is strong in the Newfoundland population (p = 0.0004) but is not noted in the Ontario population (p = 0.41). In contrast with the SEEK1 analysis, however, a multivariate analysis does provide evidence for an interaction between HLA-Cw*0602 and geographic location (p = 0.02), demonstrating that the effect of HLA-Cw*0602 on the incidence of PsA is significantly different in the two populations. Furthermore, when HLA-Cw*0602 and its interaction with location is included in a multivariate model, addition of the SEEK1(T) allele does not improve the model significantly (p = 0.26). Therefore, although SEEK1(T) is associated with PsA, SEEK1 does not appear to be a further susceptibility factor if the HLA-Cw*0602 status is already known. Table 1 presents a summary of the allele frequency for the minor SEEK1 allele (T) and the presence of the HLA-Cw*0602 allele.
Linkage disequilibrium between HLA-Cw*0602 and SEEK1(T) is reflected in the fact that in the Newfoundland population, 18/75 (24%) of control subjects without HLA-Cw*0602 have the SEEK1(T) allele, whereas 20/23 (87%) with HLA-Cw*0602 have the SEEK1(T) allele. The parallel numbers in the Ontario population are 38/107 (36%) versus 56/57 (98%). There is some evidence of a different level of disequilibrium in the two populations (p = 0.03). When we explored the genotype/phenotype correlations for the minor (T) allele compared with the corresponding wild-type variant for SEEK1, no differences in the age of onset of psoriasis, PsA, or pattern of arthritis were noted.
Presently debate exists about whether HLA-Cw*0602 is the causative allele in psoriasis or PsA as it is neither necessary nor sufficient to develop the disease. Thus it is prudent to investigate the role of neighbouring genes, such as SEEK1. At first glance, +39604 SNP in SEEK1 may not appear to be a strong candidate for PsA susceptibility as the function of SEEK1 is not yet known and the +39604 SNP variant does not code for any amino acid change. However a striking association was noted between SNP +39604 in exon 2 of SEEK1 and psoriasis in the Swedish population (p<0.000001).7 This association was independent of HLA-Cw*0602.
An efficient way to validate the importance of SEEK1 in PsA is to study multiple case-control populations, with the expectation that any important association will be seen in all populations. We examined two complementary populations. Newfoundland is a white founder population known to exhibit homogeneity comparable to the Hutterites.9 An advantage in examining this population is that it may allow the detection of small to modest genetic associations as a result of an increased “signal to noise ratio”. Meanwhile, the Ontario population is a well characterised heterogeneous PsA population that greatly enhances the generalisability of results.
In our study a statistically significant association was shown between SEEK1 and PsA in the Newfoundland population alone. There was a trend for increased prevalence of SEEK1 in the Ontario PsA population as compared with controls (46.5 v 38.0%) but this was not significant. However, there was also no evidence for a different relationship in the two separate populations, but when both populations were combined the evidence for an association was considerable (p = 0.009). Further analysis showed, however, that this association was not independent of HLA-Cw*0602 status. The association noted in the Newfoundland population can probably be attributed to the extended linkage disequilibrium that exists in this young founder population.
A puzzling finding in our study was the lack of association between HLA-Cw*0602 and PsA in the Ontario population, especially because previous studies from this cohort have confirmed this association.4,10 The allele frequency for the Ontario controls in this study (23.5%) was higher than previously reported for Ontario controls (9–15%).4,10 This group of controls and patients was different from groups previously reported and may highlight the heterogeneity within the Ontario population. A further limitation of our study is the inability to rule out the possibility that an association exists for other SNP variants in the SEEK1 gene, as only one SNP variant was studied. Finally, we are unable to decipher the independent contribution of psoriasis and inflammatory arthritis to SEEK1, as all our patients with PsA had both psoriasis and inflammatory arthritis.
In conclusion, data from our study does not support the premise that the +39604 SNP in SEEK1 is an independent genetic determinant leading to susceptibility to PsA in the Newfoundland and Ontario populations.
We thank the volunteers from Newfoundland and Ontario who participated in this study, the staff at St Clare’s Mercy Hospital and Toronto Western Hospital, and Newfound Genomics.
Grant support was received from the Canadian Institute of Health Research (CIHR) and the Arthritis Society of Canada (TAS).
PR is a CIHR New Investigator and TAS scholar.
Published Online First 11 February 2005