Arthritis & Rheumatism, Volume 62,
November 2010 Abstract Supplement

Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Atlanta, Georgia November 6-11, 2010.


Genome-Wide Association Study of Ankylosing Spondylitis Identifies New Loci, a Tag SNP for HLA-B27, and an Interaction between HLA-B27 and Variants in ERAP1.

A. Brown,  Matthew, Behalf of the Wellcome Trust Case-Control Consortium 2,  on, Spondyloarthritis Consortium,  Australo-Anglo-American

In order to identify genetic variants predisposing to risk of Ankylosing Spondylitis (AS), we performed a gwas of 1782 British and Australian cases fulfilling modified New York Criteria, and 5167 historical controls from the Wellcome Trust Case Control Consortium 2 (WTCCC2). After imputing to Hapmap, the study was combined with existing results from the Australo-Anglo-American AS Consortium (TASC) involving a non-overlapping cohort of Australian, British and North American cases and ethnically matched controls, using inverse variance meta-analysis. Replication was then performed in 2109 cases and 4410 controls from Australia, Britain, and the Canadian SPARCC consortium.

As well as confirming known associations at HLA, IL23R, ERAP1, KIF21B, 2p15 and 21q22, we identified and subsequently replicated risk predisposing variants in RUNX3 (combined p=3.3 × 10-12), IL12B (p=1.8 × 10-8), and LTBR (p=4.5 × 10-10), and found suggestive association at PTGER4 (p=7.8 × 10-8), and TBKBP1-TBX21 (p=5.9 × 10-8), as well as providing further support for previously reported loci with suggestive association with AS, including ANTXR2 (p=3.8 × 10-7), CARD9 (p=1.2 × 10-6), and TRADD (p=4.1 × 10-6).

We also identified a single SNP, rs4349859 near the gene MICA, which tagged HLA-B27 with near perfect sensitivity (98%) and specificity (99%) in 531 cases and 729 controls of Australian and British origin. These findings were confirmed in independent sets of Sardinian and Azorean cases, and show that whilst rs4349859 tagged the non-AS associated HLA-B*2709 subtype, it did not tag the AS-associated HLA-B*2707 subtype, indicating that rs4349859 is not AS causative, and further reducing the likelihood that a B27-linked gene is responsible for the association of B27 with AS.

In most common diseases the proposed genetic model involves interaction between loci, but to date no convincing examples of such interaction have been demonstrated. In the current study however, we identified an interaction between HLA-B27 and variants within ERAP1 in the WTCCC2 (p=0.008), TASC (p=0.004) and replication datasets (p=0.004). Specifically, risk variants in ERAP1 increased odds of disease in HLA-B27-positive, but not B27-negative, cases (combined interaction p=1.4 × 10-6). In contrast, IL23R was associated with disease in both B27+ve and –ve cases. This result indicates that B27+ and B27-ve forms of disease have substantially different but overlapping aetiologies. The findings support mechanisms of association of B27 and ERAP1 with AS that involve peptide presentation, and that ERAP1 contributes to disease risk through its action in trimming peptides prior to loading into nascent HLA class I molecules, rather than by cleaving pro-inflammatory cytokine receptors on the cell membrane.

In summary, we have identified several new loci that affect risk of AS, bringing the total number of confirmed AS genetic susceptibility loci to 14, report a single SNP variant that tags HLA-B27 which may be used in the future as a cheap alternative to expensive B27 typing, and have discovered one of the first convincingly replicated examples of a genetic interaction affecting risk of a complex disease.

To cite this abstract, please use the following information:
A. Brown, Matthew, Behalf of the Wellcome Trust Case-Control Consortium 2, on, Spondyloarthritis Consortium, Australo-Anglo-American; Genome-Wide Association Study of Ankylosing Spondylitis Identifies New Loci, a Tag SNP for HLA-B27, and an Interaction between HLA-B27 and Variants in ERAP1. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :1355
DOI: 10.1002/art.29121

Abstract Supplement

Meeting Menu

2010 ACR/ARHP