Abstract
Objective. Genetic components play important roles in the incidence and development of ankylosing spondylitis (AS). Aminopeptidase regulator of tumor necrosis factor receptor shedding 1 (ERAP1) was recently found to be associated with AS in North American and British cohorts. We evaluated whether ERAP1 is associated with AS in a Chinese Han population.
Methods. A sample of 50 patients and 50 healthy controls was recruited for preliminary screening for informative single-nucleotide polymorphisms (SNP). Then 6 SNP of suggestive significance in the initial screening were followed up in a large sample of 471 patients with AS and 456 ethnically matched controls. Diagnosis of AS followed the 1984 modified New York criteria. Linkage disequilibrium coefficient (D’ and r2) and haplotypes were estimated by Haploview.
Result. Two SNP (rs27434, p = 0.00039, and rs27529, p = 0.0083) in ERAP1 other than that reported previously were found to be significantly associated with AS. Haplotype analysis using 5 SNP within 1 linkage disequilibrium block identified 2 risk haplotypes (GATGT and GACGT) and 1 protective haplotype (GGTGT) for AS.
Conclusion. Our study demonstrated that 2 novel SNP in ERAP1 were associated with AS in the Han Chinese population, suggesting that ERAP1 might confer genetic risk for AS in Han Chinese through the common mechanism shared by different populations, although the AS-associated SNP in ERAP1 might be population-specific.
- ERAP1
- ANKYLOSING SPONDYLITIS
- CHINA
Ankylosing spondylitis (AS) is a progressive chronic disease characterized by inflammatory low back pain, sometimes accompanied by peripheral arthritis, enthesis, iritis, spinal deformity, and ankylosis. The prevalence of AS ranges from 0.1% to 1.8% in Europe and is 0.24% in China1,2. Although genetic components have been proposed as important contributors to the pathogenesis of AS3, the responsible genetic molecules and related mechanisms remain unclear.
HLA-B27 is considered to be associated with the familial aggregation of AS4, but it was estimated that it contributed to no more than 50% of the overall genetic risk for AS5,6. A recent genome-wide scan using 14,500 customized nonsynonymous single-nucleotide polymorphisms (SNP) revealed that ERAP1 and IL23R were associated with AS in a group of Europeans, and specifically, the associations at rs27044, rs17482078, rs10050860, rs30187, and rs2287987 in ERAP1 were reported7. Most recently, the association of ERAP1 with AS was validated in a Korean group8. ERAP1 encodes an endoplasmic reticulum-associated aminopeptidase (also known as ARTS1). This protein facilitates HLA class I presentation by processing peptides to optimal length9,10. It also binds and cleaves several cytokine receptors including interleukin 6 receptor (IL-6R), tumor necrosis factor R1 (TNFR1), and IL-1RII from cell surface and thus promotes receptor shedding11,12,13,14.
We carried out a case-control association study to determine whether ERAP1 is also associated with the incidence of AS in a Chinese population and whether it is correlated with clinical features.
MATERIALS AND METHODS
Subjects
All participants signed informed consent for their blood samples to be taken and used in this study, which was approved by the Ethics Committee of the Third Affiliated Hospital of Sun Yat-sen University. Fifty patients and 50 healthy controls were recruited for preliminary screening for informative SNP. The sample for validation consisted of 471 patients with AS and 456 ethnically matched controls. Diagnosis of AS followed the 1984 modified New York criteria15, and patients were assessed by at least 2 qualified rheumatologists. The clinical data included age, sex, family history, onset age, dactylitis, peripheral arthritis, hip joint involvement, iritis, enthesitis, erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP).
Genotyping
Genomic DNA was obtained from peripheral blood samples using the phenol/chloroform extraction method16. To identify the candidate SNP in the promoter region (3000 base pairs upstream of the first exon), exonic region, and exon-intron junctions of the ERAP1 gene, the fragments were amplified, respectively, using polymerase chain reaction (PCR) with specific primers (Table 1), and then sequenced with an ABI 3700 automated sequencer after purification of PCR products. In the validation stage, PCR-restriction fragment–length polymorphism (RFLP) was carried out to genotype 6 SNP. Briefly, each fragment flanking the SNP was PCR-amplified and then digested by the nucleases. Primers and nucleases designed for the PCR-RFLP analysis are shown in Table 1.
Statistical analysis
Hardy-Weinberg equilibrium by chi-squared test was used to evaluate the genotyping accuracy of each SNP. In the initial stage, allelic associations between cases and controls were performed using Fisher’s exact test to generate significance, regarding the matching of sex and age between cases and controls. The magnitude of association was expressed as OR for each addition of 1 minor allele, with a 95% CI. In the followup stage, since age and sex were not matched between cases and controls (both p < 0.0001), multivariate logistic regression analysis was also performed using SAS 8.0 to test the correlations between the SNP and AS, adjusting for age and sex. To adjust for multiple testing on the 6 candidate SNP, p values were corrected by Bonferroni method17. Linkage disequilibrium (LD) coefficient (D’ and r2) and haplotypes were estimated by Haploview 4.118. Population-attributable risk (PAR) was calculated as follows: PAR% = (It – I0)/It × 100%, where It is the incidence in the general population and I0 is the incidence in a nonexposed group.
Finally, relationships among a list of clinical measurements (age, sex, family history, onset site, dactylitis, peripheral arthritis, hip joint involvement, iritis, enthesitis, ESR, and CRP) and SNP were evaluated by SAS 8.0 and multivariate logistic regression with age and sex as covariates.
RESULTS
Clinical data
In the initial screening sample of 50 patients with AS and 50 controls, for patients, the mean age was 28.2 ± 6.7 years and the mean disease duration was 8.1 ± 7.1 years; the male to female ratio was 5.25. Most patients with AS (92%) were HLA-B27-positive, of which 20% had a family history of the disease. For controls, the mean age was 27.2 ± 7.1 years; male to female ratio was 5. Age and sex were not significantly different between AS cases and controls. Among those cases selected for validation, the mean age and mean disease duration were 28.7 ± 8.7 years and 7.07 ± 7.78 years, respectively. The male to female ratio was 5.73. HLA-B27 positivity was found in 86.7% (n = 408) of patients with AS, of which 21% had a family history of the disease. Among controls selected for validation, the mean age was 26.08 ± 8.88 years; male to female ratio was 1.65.
Initial screening for suggestive significant SNP
By using PCR-based direct sequencing, a total of 41 SNP were found in the promoter and exon regions of ERAP1. After stringent quality control filtering including deviation from Hardy-Weinberg disequilibrium, minor allele frequency, and homogeneity, 26 SNP were retained for analysis. The SNP that were associated with AS7,8 (rs27044, rs17482078, rs10050860, rs30187, and rs2287987) were not included. Six SNP (rs26653, rs27434, rs27640, rs27529, rs26510, and rs27582) showed significant association (p < 0.05) in the initial sample, while the others did not, with p values ranging from 0.087 to 1.00.
Validation of SNP in a larger population
Six SNP (rs26653, rs27434, rs27640, rs27529, rs26510, and rs27582) that showed significance (p < 0.05) in the initial screening stage were subjected for validation in a sample of 471 AS cases and 456 healthy controls using PCR-RFLP assays. All 6 SNP were in Hardy-Weinberg equilibrium in the control group.
Single-marker association analysis for SNP in ERAP1 (Table 2) showed that 2 SNP, rs27434 and rs27529, were significantly associated with AS. The frequency of risk allele A of rs27434 is higher in cases of AS (51%) than in controls (43%) (OR 1.38, 95% CI 1.09 ∼ 1.74, p = 0.00039; after Bonferroni correction, pc = 0.002), by adjustment for age and sex. The allele A of SNP rs27434 contributed to 7.4% of the overall genetic risk in AS. The frequency of risk allele G of rs27529 was higher in cases (60%) than in controls (55%) (OR 1.23, 95% CI 1.02 ∼ 1.49, p = 0.0083, pc = 0.049) adjusted for age and sex (Table 2).
Identification of 6 haplotypes associated with AS in a Chinese Han group
Tests revealed strong LD (Figure 1) between any 2 of the 5 SNP (rs27653, rs27434, rs27640, rs27529, rs27510), while rs27582 shared poor LD with any of these 5 SNP (Figure 1). Therefore, haplotype analysis concerning 5 SNP was carried out (rs27653, rs27434, rs27640, rs27529, and rs27510). The result showed that the proportions of haplotypes GATGT (OR 1.68, p = 0.002) and GACGT (p = 0.0004) were higher in AS cases, while the haplotype GGTGT (OR 0.18, p = 3 × 10−7) was more frequent in controls (Table 3). These findings suggested that haplotypes GATGT and GACGT contributed significant risk effects, while GGTGT is a protective haplotype.
Correlations between 6 SNP and clinical measurements
Using multivariate logistic regression with age and sex as covariates, none of the 6 SNP showed significant association with the clinical measurements of AS (age, sex, family history, onset site, dactylitis, peripheral arthritis, hip joint involvement, iritis, enthesitis, ESR, and CRP) except for the SNP rs27510, which was found to be significantly correlated with onset age in patients with AS (p = 0.0083).
DISCUSSION
It is well accepted that AS is a hereditary disease involving many genes19. Although HLA-B27 is the most consistent locus associated with AS, it contributes no more than 50% of the overall genetic risk for AS. Recently, 5 SNP within ERAP1 locus were found to be associated with AS in US and British cohorts, and 2 of these SNP were replicated in a Korean cohort8. However, further validation in other populations is necessary, to determine whether ERAP1 is the common risk locus for AS.
We determined the SNP in the promoter and exon regions of ERAP1 in a screening sample of Han Chinese by direct sequencing, and performed association tests among the SNP. Six SNP showing suggestive significance were subjected for validation in the same ethnic Han Chinese population of a larger sample size. Two novel SNP within the exon regions of ERAP1, rs27434 and rs27529, revealed significant associations with AS, with OR of 1.38 and 1.2, respectively, suggesting that ERAP1 might be associated with AS in a Han Chinese population. Similarly, another 2 SNP in ERAP1 had been revealed to be associated with AS in a Han Chinese population (rs27980, p = 0.0048; rs7711564, p = 0.0081). These findings were consistent with the previous report that ERAP1 or ARTS-1 was associated with AS in mainland US and British cohorts, indicating that the association of ERAP1 with AS was replicated in different ethnic populations. On the other hand, the SNP of top significance were different among studies of different populations, suggesting the existence of allelic heterogeneity for ERAP1 among populations with different ethnic backgrounds. In the Chinese population, risk allele A of SNP rs27434 contributed to 7.4% of the overall genetic risk in AS, which would be a promising target for further study of AS pathogenesis.
Our results indicated that the SNP rs27510 in ERAP1 was correlated with age at onset in patients with AS, while none of the other 5 associated SNP was correlated with peripheral symptoms or disease activity indices or other measurements. This might suggest the SNP rs27510 in ERAP1 was associated with age at onset in patients with AS.
However, the power to detect association with an OR of 1.5 is less than 80%. Although this will not affect our conclusions of positive association, the lack-of-association results regarding clinical measurements need to be interpreted with caution.
Single-nucleotide variants in ERAP1 might have direct functional implications in the molecular pathogenesis of AS. The ERAP1 gene is located on chromosome 5 and encodes a protein required for shedding of type II interleukin 1 decoy receptor (IL-1RII) and IL-6 receptor7. IL-1 is a proinflammatory cytokine that plays an important role in inflammation, host defense, and immunity as well as the pathogenesis of AS20. Proteolytic cleavage of the extracellular domain of the IL-1RII generates soluble IL-1-binding proteins that prevent excessive bioactivity by binding free IL-17. Moreover, IL-6 exerts functions through its receptor complex and could induce the expression of an IL-1 receptor antagonist; and stimulates shedding of TNFR1 to a soluble TNF-binding protein13. Therefore, the involvement of ERAP1, IL-6 signaling, and TNF may be another way that ERAP1 participates in the inflammation process in AS, since TNF-α is a crucial inflammation mediator and has been considered as an inflammatory factor of pathogenesis in AS21.
Our study provided evidence that ERAP1 is associated with AS in a Chinese population. With the evidence of previous reports, the study suggested that ERAP1 is the common susceptibility gene conferring risk of AS, and the pathogenetic mechanism involved might be similar in a general population. However, as indicated by the previous studies, the associated SNP are likely tagging the causative variant(s) that lie within the ERAP1 locus. Further study is needed to determine the causative variant of ERAP1 that contributes to the susceptibility of AS. Although discovery of ERAP1 susceptibility in AS has been a novel breakthrough in genetic studies of AS since the 1970s, when HLA-B27 was identified as a risk locus, further investigations are needed to reveal the underlying molecular mechanisms leading to AS.
Acknowledgment
We thank all the patients with AS and the healthy subjects for their cooperation.
Footnotes
-
Supported by grants (2005A30801005) to J. Gu from the Foundation of Guangdong Province of China, clinical subject (2007) of the Ministry of Public Health of China and 5010 Subject (2009–2010) of Sun Yat-sen University.
- Accepted for publication September 17, 2010.
REFERENCES
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
- 17.
- 18.
- 19.
- 20.
- 21.