Article Text

Extended report
Association of STAT3 and TNFRSF1A with ankylosing spondylitis in Han Chinese
  1. Stuart I Davidson1,
  2. Yu Liu2,
  3. Patrick A Danoy1,
  4. Xin Wu2,
  5. Gethin P Thomas1,
  6. Lei Jiang2,
  7. Linyun Sun3,
  8. Niansong Wang4,
  9. Jun Han5,
  10. Huanxing Han6,
  11. Australo-Anglo-American Spondyloarthritis Consortium,
  12. Peter M Visscher7,
  13. Matthew A Brown1,
  14. Huji Xu2
  1. 1The University of Queensland Diamantina Institute, University of Queensland, Princess Alexandra Hospital, Brisbane, Australia
  2. 2Department of Rheumatology and Immunology, Shanghai Changzheng Hospital, The Second Military Medical University Hospital, Shanghai, China
  3. 3Department of Rheumatology and Immunology, Nanjing Gulou Hospital, Nanjing, China
  4. 4Department of Nephrology, Shanghai No. 6 Hospital, Shanghai, China
  5. 5Modern Research Centre for traditional Chinese Medicine, The Second Military Medical University Hospital, Shanghai, China
  6. 6Department of Clinical Immunology, Shanghai Changzheng Hospital, The Second Military Medical University Hospital, Shanghai, China
  7. 7Queensland Institute of Medical Research, Herston, Queensland, Australia
  1. Correspondence to Professor Matthew A Brown, University of Queensland Diamantina Institute, University of Queensland, Brisbane, Queensland 4102, Australia; matt.brown{at} or Professor Huji Xu, Department of Rheumatology and Immunology, Shanghai Changzheng Hospital, The Second Military Medical University Hospital, Shanghai 200003, China; xuhuji{at}


Objectives Recent association studies by the Australo-Anglo-American Spondyloarthritis Consortium (TASC) in Caucasian European populations from Australia, North America and the UK have identified a number of genes as being associated with ankylosing spondylitis (AS). A candidate gene study in a Han Chinese population was performed based on these findings to identify associated genes in this population.

Methods A case-control study was performed in a Han Chinese population of patients with AS (n=775) and controls (n=1587) from Shanghai and Nanjing. All patients met the modified New York criteria for AS. The cases and controls were genotyped for 115 single nucleotide polymorphisms (SNPs) tagging IL23R, ERAP1, STAT3, JAK2, TNFRSF1A and TRADD, as well as other confirmation SNPs from the TASC study, using the Sequenom iPlex and the ABI OpenArray platforms. Statistical analysis of genotyped SNPs was performed using the Cochran–Armitage test for trend and meta-analysis was performed using METAL. SNPs in AS-associated genes in this study were then imputed using MaCH, and association with AS tested by logistic regression.

Results SNPs in TNFRSF1A (rs4149577, p=8.2×10−4), STAT3 (rs2293152, p=0.0015; rs1053005, p=0.017) and ERAP1 (rs27038, p=0.0091; rs27037, p=0.0092) were significantly associated with AS in Han Chinese. Association was also observed between AS and the intergenic region 2p15 (rs10865331, p=0.023). The lack of association between AS and IL23R in Han Chinese was confirmed (all SNPs p>0.1).

Conclusions The study results demonstrate for the first time that genetic polymorphisms in STAT3, TNFRSF1A and 2p15 are associated with AS in Han Chinese, suggesting common pathogenic mechanisms for the disease in Chinese and Caucasian European populations. Furthermore, previous findings demonstrating that ERAP1, but not IL23R, is associated with AS in Chinese patients were confirmed.

Statistics from


Ankylosing spondylitis (AS) is a common inflammatory arthritis with a prevalence of 0.2% to 0.54% in Chinese populations.1 AS is characterised by axial skeletal inflammation and enthesitis causing pain and stiffness, eventually leading to increased bone formation and progressive joint ankylosis. AS is strongly associated with the class I major histocompatibility complex (MHC) gene HLA-B27 (B27), which is found in >80% of Chinese patients with AS compared with approximately 3% to 5% of healthy Chinese.2 However, only 1% to 5% of B27 carriers develop AS, suggesting that while B27 plays a large role in disease susceptibility, there are other genetic factors responsible for modifying disease penetrance.3 4 Other genes that have been shown to be associated with AS in Chinese populations include HLA-B60,5 IL1A/B,6 and ERAP1.7

Recent genome-wide association studies in populations of Caucasian European ancestry have identified ERAP1 (the gene for endoplasmic reticulum aminopeptidase 1) and IL23R (the gene for interleukin 23 receptor (IL-23R)) as having strong association with AS.8 We have previously demonstrated association of ERAP1, but not the single nucleotide polymorphisms (SNPs) in IL23R associated with AS in Caucasian Europeans, with AS in a Han Chinese population.7 Thus, although the disease has similar genetic causes in different populations, there are associations that differ between populations.

We have recently also reported identification of further AS-associated genetic loci in a genome-wide association study of Caucasian Europeans from Australia, North America and Britain.9 Based on the preliminary results of this work, we designed a candidate gene study in order to confirm any associations in a Han Chinese population. The six genes investigated included IL23R and ERAP1, as an extension of our previous confirmation study, as well as four new genes: STAT3 (the gene for signal transducer and activator of transcription factor 3), JAK2 (the gene for Janus kinase 2), TNFR1 and TRADD. The four new genes were selected due to association observed in the Australo-Anglo-American Spondyloarthritis Consortium (TASC) study, as well as their functional and potential pathogenic roles. STAT3 and JAK2 are closely linked through their involvement together in signal transduction pathways leading to DNA transcription. STAT3 and JAK2 are both involved in the regulation of the Th17 subset of CD4 T cells, and have been associated with inflammatory bowel disease (IBD), a condition often found in association with AS. Engagement of IL-23R is a key regulatory step required for the differentiation and maintenance of T helper 17 (Th17) T lymphocyte populations.10,,12 The association of IL23R with AS suggests that Th17 cells may play a role in disease pathogenesis, and this is supported by evidence of elevated levels of circulating Th17 cells in the blood of patients with AS.13 JAK2 and STAT3 are found immediately downstream of IL-23R in the IL-23 signalling cascade, suggesting that polymorphisms in either of these genes could possibly contribute to the aberrant Th17 cell function thought to contribute to AS susceptibility.

The final two genes selected were TNFRSF1A and TRADD. These genes encode the adjacent proteins in the tumour necrosis factor (TNF) signalling pathway tumour necrosis factor receptor superfamily 1A (TNFRSF1A) and TNFR 1-associated death domain (TRADD). Increased numbers of TNFα expressing macrophages have been observed in patients with AS,14 and anti-TNFα blocking agents are highly effective in the treatment of AS. Recent studies have also suggested that TRADD is associated with AS in patients in the UK,15 and suggestive association of TNFRSF1A was seen in the TASC genome-wide association study (strongest associated SNP rs1800693, p=6.9×10−5).9

We describe here the results of a confirmation study that was undertaken to further investigate the genetic basis of AS. Cases and controls were genotyped across six genes to determine whether polymorphisms in these genes contribute to AS susceptibility in a Han Chinese population.

Patients and methods


Unrelated patients with AS (n=775) were recruited from outpatient clinics at several hospitals in Shanghai and Nanjing. Diagnosis of AS was confirmed by a qualified rheumatologist, with all cases meeting the modified New York criteria for AS.16 Unrelated healthy, ethnically matched blood donors (n=1587) recruited from the Shanghai Blood Bank were included as controls. All patients gave informed, written consent, and the study was approved by the relevant ethics committee.

Single nucleotide polymorphism selection

Haplotype data on Chinese and Japanese subjects was obtained from the International HapMap Project (release 23, National Centre for Biotechnology Information (NCBI) B36 assembly, dbSNP b36) for ERAP1, IL23R, STAT3, JAK2, TNFRSF1A and TRADD and their surrounding areas to select the SNPs for this study. The program ‘Tagger’ within Haploview V.4.017 was used to select tag SNPs covering these genomic regions to give 100% coverage of these genes (r2≥0.8), with the forced inclusion of the most associated SNPs from the TASC study. In total 115 SNPs were genotyped, comprising 18 for IL23R, 34 for ERAP1, 20 for JAK2, 10 for TNFRSF1A, 7 for TRADD, 7 for STAT3 and a further 19 TASC confirmation SNPs. In ERAP1, although only 22 SNPs were required to cover the known variation in Chinese and Japanese, an additional 12 SNPs were typed covering variation in Europeans as well.


Genotyping the first cohort of samples was performed using the MassArray platform (Sequenom, San Diego, California, USA), which uses chip-based matrix-assisted laser desorption ionisation–time-of-flight mass spectrometry technology. PCR and iPlex extension reactions were designed using MassArray Assay Design V.3.1 in multiplex format. Genotyping was then performed according to the manufacturer's standard protocols. MassArray Typer V.4.0 was used to read the extended mass and genotype calls. The second round of genotyping was performed using the OpenArray platform (Applied Biosystems, Foster City, California, USA), which uses chip-based TaqMan genotyping technology. Genotyping was performed in accordance with the manufacturer's standard protocols, and genotype calls were made using OpenArray SNP Genotyping Analysis Software V.1.0.3.

Statistical analysis

Association analysis was performed using the Cochran–Armitage test for trend in Plink V.1.03 (available online at∼purcell/plink/).18 SNPs with missingness rates >0.1 and minor allele frequency (MAF) <1% were excluded, and an exact test for Hardy–Weinberg equilibrium (HWE) was performed in controls; markers with p<0.0005 were excluded. The SNP marker map used for analysis was NCBI dbSNP b129 (July 2008). Meta-analysis of the two genotyping rounds was performed using METAL (; all findings of genotyped SNPs are from this analysis. SNP imputation was performed using MaCH V.1.0 (, with phased data from Chinese and Japanese individuals from the International HapMap Project (release 22) serving as the reference set of haplotypes. Association analysis for imputed data was performed using logistic regression of the probability scores for the imputed genotypes for SNPs within a 50 kb region 5' and 3' of the respective gene. SNPs were excluded if confidence regarding imputation accuracy was low (r2<0.3 with flanking SNPs; quality score <0.95). ORs and 95% CIs were calculated. Study power was determined using the Genetic Power Calculator (∼purcell/gpc/). Physical distances in the study are according to the NCBI dbSNP genome build 128 (October 2007).


A total of 775 AS cases and 1587 healthy controls were genotyped for 115 polymorphisms using the Sequenom MassArray and Applied Biosystems OpenArray platforms. Samples with low genotyping rates (genotyping <90% completed) were excluded (n=132), leaving 773 cases and 1457 healthy controls in the final analysis. A total of 14 markers were excluded because of low MAF, low completion rates or for not being in HWE, leaving 99 markers in the final dataset.

Based on the assumption of a population prevalence of disease of 0.2%, α=0.05, minor allele frequencies of 0.1–0.5, and linkage disequilibrium of r2=0.8, the study had an 80% power to detect an additive association with an additive OR of 1.24 to 1.36.

A quantile–quantile plot of observed versus expected association findings shows substantially greater levels of association observed than would be expected by chance (figure 1), reflected by the genomic inflation factor (λ) for the study of 1.43.

Figure 1

Quantile–quantile plot of χ2 values.

Significant SNP association findings are presented in table 1, along with the corresponding association values from the TASC study.9 Results of all genotyped SNPs are presented in supplementary table 1.

Table 1

Significant (p<0.05) genetic associations observed in genotyped SNPs, with comparison p values from the TASC study

Peak association with AS was observed with the marker rs4149577 in TNFRSF1A (OR 0.81, 95% CI 0.70 to 0.92; p=8.2×10−4). Association with AS was observed in STAT3 with the marker rs2293152 (OR 0.81, 95% CI 0.72 to 0.92; p=0.0016), with moderate association also observed in STAT3 with the marker rs1053005 (OR 1.18, 95% CI 1.03 to 1.35; p=0.017). For two markers in ERAP1, association with AS was observed with p values of <0.01, and a further four markers with p values of <0.05. Peak association was observed for the marker rs27038 (OR 1.18, 95% CI 1.04 to 1.34; p=0.0091), an SNP also associated with AS in the TASC study.9 Association with AS was also observed for markers at 2p15 (rs10865331, OR 1.16, 95% CI 1.02 to 1.32; p=0.023) and 6p22.1 (rs7740197, OR 0.82, 95% CI 0.67 to 1; p=0.049). No association was observed for genotyped markers in IL23R, JAK2 or TRADD (for all SNPs p≥0.09).

Imputation was used to assess association at SNPs not typed in TNFRSF1A, STAT3 and ERAP1, as these three genes were identified as having association with AS. The combined genotyped and imputed SNP association findings for these genes are presented in supplementary figure 1A–C. The strongest imputed SNP association observed at TNFRSF1A was for rs1800693 (p=0.0045). At STAT3, the strongest association observed between an imputed SNP and AS was for rs12601982 (p=0.0069). Association (p<0.05) was observed across this locus for markers extending 16 kb, from 37 709 372 to 37 725 506 bp (markers rs9906989 to rs3809758). The strongest imputed SNP association observed at TNFRSF1A was for rs1800693 (p=0.0045). At ERAP1, the strongest imputed SNP association was observed for rs27583 (p=0.0042), with association (p<0.05) observed across this locus for markers extending 57 kb, from 96 106 224 to 96 163 596 bp (markers rs26507 to rs26495).


We have undertaken a case-control genotyping study to investigate the association between AS and polymorphisms in ERAP1, IL23R, STAT3, JAK2, TNFRSF1A and TRADD, first observed in a Caucasian European cohort, in the Han Chinese population. Our results indicate that ERAP1, STAT3 and TNFRSF1A are associated with AS in Han Chinese populations in addition to Caucasian Europeans. This is the first study to show that STAT3 and TNFRSF1A are associated with AS in a non-Caucasian European population, suggesting a common pathogenic mechanism for the disease in Chinese and Caucasian Europeans.

The association of STAT3 with AS is an exciting finding, particularly in the context of the lack of association observed with IL23R. STAT3 is found directly downstream of IL-23R in the IL-23 signalling cascade and is a critical transcription factor in the differentiation of Th17 cell populations.19 20 Loss of function mutations in STAT3 cause Job syndrome (OMIM 147060;, which is associated with recurrent, severe infections with extracellular bacteria and fungi. The association of STAT3 and not IL23R with Chinese AS indicates that it is likely that there are different mechanisms of disease pathogenesis that have the same effect on Th17 cells. STAT3 has also been shown to be associated with both forms of IBD, Crohn's disease21 and ulcerative colitis,22 and both are clinically associated with AS. There is strong evidence linking IBD and AS. Around 68% of patients with spondyloarthritis have gut inflammation resembling Crohn's disease,23 and spondyloarthritis is a common complication of IBD.24 Strong cofamiliality has been demonstrated between the two conditions, suggesting shared aetiological factors.25 The shared association with STAT3 for AS and IBD suggests that the diseases have some pathogenic mechanisms in common, of which STAT3 is one shared component. STAT3 is activated by phosphorylation by JAK2. JAK2 is associated with Crohn's disease, but was not associated with AS in the current study.

The observed association between TNFRSF1A and AS is also very interesting. As with STAT3, TNFRSF1A has been previously shown to be associated with Crohn's disease and ulcerative colitis,26 27 which supports the association with AS. Furthermore, studies in mice have shown that overexpression of TNF leads to IBD, and sacroiliitis resembling AS, in a manner dependent on the expression of TNFRSF1A,28 supporting a role for TNFRSF1A in AS pathogenesis. SNPs in TNFRSF1A were moderately associated with AS in the TASC genome-wide association study (strongest associated SNP rs1800693, p=6.9×10−5); the same SNP was AS associated in the current study (p=0.045). This finding in Han Chinese is supportive of TNFRSF1A being a true AS susceptibility gene.

The association with TNFRSF1A may also help to elucidate the role of ERAP1 in AS aetiopathogenesis. Association of ERAP1 with AS in Han Chinese has been observed at multiple markers, validating our previous results.7 Cleavage of TNFR1 (and IL-1R2 and IL-6Rα) from the cell surface is one of the previously identified functions of ERAP1.29 ERAP1 is also involved in trimming peptides to optimal length within the endoplasmic reticulum for MHC class I presentation.30 31 AS in Chinese, as with AS in Caucasian Europeans, is a class I MHC mediated disease, with >80% of Chinese patients with AS carrying the HLA-B27 allele.2 Further studies will be required in order to determine whether the association of ERAP1 relates to its role in peptide presentation or its role in receptor shedding, but it may help to also explain the association of HLA-B27 and TNFRSF1A with AS.

This study also suggests a lack of association between IL23R and AS in Chinese, consistent with our previous results but in an expanded dataset. The polymorphism rs11209026, widely believed to be the causative SNP in Caucasian European AS, as well as Crohn's disease and psoriasis, was again shown to be not polymorphic in Chinese. This, along with the lack of association with other common SNPs found to be significantly associated in Caucasian European populations, suggests that IL23R may be an ethnicity-specific association.

The results of this confirmation study have shown the value of exploring the genetic diversity between different ethnic groups for gene mapping. As well as confirming shared associations between populations, our findings also suggest that there are significant differences in susceptibility genes between Han Chinese and Caucasian European populations. These differences may possibly be used to distinguish true associations from artefact effects. Furthermore, identification of new susceptibility genes in other ethnic groups may provide further information on the pathogenesis of AS in Caucasian Europeans. This will make further gene mapping studies in East Asian populations very valuable in improving our understanding of the genetic basis of disease.


The authors would like to thank all participating AS cases and healthy individuals who provided the DNA and clinical information necessary for this study. The assiduous DNA extraction work by their colleagues Drs Zhongwei Wang, Ting Li, Yiping Lin and Chao Wu, is also greatly appreciated. HX was funded by the National Natural Science Foundation of China grant 30972339 and the Science and Technology Commission of Shanghai Municipality grant 08XD1400400. MAB and PVM are supported by fellowships from the National Health and Medical Research Council (NHMRC; Australia). SD is supported by an NHMRC PhD studentship. This project was supported by the Rebecca Cooper Medical Research Foundation. TASC is funded by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) grants P01-052915, R01-AR046208, and the intramural Research Program, NIAMS. This study was also funded by Arthritis Australia.


Supplementary materials


  • MAB and HX are equal senior authors.

  • SD and YL contributed equally to this study.

  • Patient consent Obtained.

  • Ethics approval This study was conducted with the approval of the Shanghai Changzheng Hospital, The Second Military Medical University Hospital, Shanghai 200003, China.

  • Provenance and peer review Not commissioned; externally peer reviewed.

    The Australo-Anglo-American Spondyloarthritis Consortium (TASC) are: John D Reveille, Rheumatology and Clinical Immunogenetics, University of Texas Health Science Center at Houston, Houston, Texas, USA; David M Evans, MRC Centre for Causal Analyses in Translational Epidemiology, Department of Social Medicine, University of Bristol, Bristol, UK; Laurie Savage, The Spondylitis Association of America, Sherman, Oaks, California, USA; Michael M Ward, National Institute of Arthritis and Musculoskeletal and Skin Diseases, NIH, Bethesda, Maryland, USA; Michael H Weisman, Department of Medicine/Rheumatology, Cedars-Sinai Medical Centre, Los Angeles, California, USA; B Paul Wordsworth, Botnar Research Centre, Nuffield Department of Orthopaedics, Rheumatology, and Musculoskeletal Sciences, University of Oxford, Oxford, UK; Matthew A Brown, The University of Queensland Diamantina Institute, University of Queensland, Princess Alexandra Hospital, Brisbane, Australia.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.