Wordsworth Group | Genetics of Ankylosing Spondylitis

Twenty years ago we initiated a global effort to identify these effects by studying the small genetic differences that exist between individuals with ankylosing spondylitis and the healthy general population.

More than 60 genetic influences have now been identified by the consortium.  Some of these discoveries have already helped us to develop new ideas about treating ankylosing spondylitis and led to successful clinical trials in humans.

It has long been recognised that there is a genetic component to ankylosing spondylitis (AS) as judged by the strong excess familial recurrence of the disease. This is particularly evident amongst first-degree relatives of those with AS, in whom the risk is approximately 10 per cent. 

In 1973 the first genetic association was described between AS and the immune response gene HLA-B27. The odds ratio in excess of 120 associated with HLA-B27 remains to this day the strongest genetic association with a polygenic disease.

For many years the strength of this association was such that that it was widely assumed that susceptibility to AS was in fact monogenic. Our AS twin studies at the end of the last century clearly showed that this was not the case (Brown et al., 1997). This and other subsequent studies suggest that HLA-B27 is probably responsible for no more than 25 to 30 per cent of the total genetic risk.

To identify the other genes involved we initially developed genetic linkage analysis in affected sib pairs with AS and published the first such studies 15 years ago. Although these confirmed the linkage with the major histocompatibility complex (containing HLA-B27) they were otherwise relatively disappointing.

We therefore rapidly moved on to genome-wide association studies (GWAS) and established a number of global consortia to recruit the necessary case samples, including cohorts from the National Ankylosing Spondylitis Society (UK), the Triple A (Australo-Anglo-American) Spondylitis Consortium (TASC) and IGAS (International Genetics of Ankylosing Spondylitis) Consortia.

Louise Appleton and Jon Lau in our group co-ordinate the recruitment of AS cases from the UK for these studies, supported by the NIHR Thames Valley Collaborative Research Network and Arthritis Research UK.

Louise plays a crucial role in the preparation, collation and curating of high quality DNA stocks for these collaborations. Approaching 20,000 cases have now been collected world-wide and the IGAS consortium has identified more than 60 independent genetic effects (reported at the IGAS Consortium meeting in Toronto 2015 by Professor Matt Brown). 

In 2007 we published, with others, the first limited GWAS which confirmed the strong association with HLA-B27 and additionally showed for the first time a strong association with ERAP1 (endoplasmic reticulum associated endopeptidase 1), a completely novel association that has opened an entirely new area of research into the causes and potential treatments of ankylosing spondylitis (Burton et al., 2007). 

Subsequently, the ERAP 1 association was demonstrated to be restricted to cases that were also HLA-B27 positive (Evans et al., 2011), the first time such a striking synergy between two genes had been described in susceptibility to a complex disease. 

Work by David Harvey in our group facilitated the definition of the crystal structure of ERAP1 in collaboration with Prof Udo Oppermann and his team in the Structural Genomics Consortium at the Botnar Research Centre (Harvey et al., 2009; Kochan et al., 2011).

We collaborate closely with Professor Bowness’ Immunology team at the Botnar Research Centre to investigate the mechanisms of the ERAP1 association and its potential utility as a treatment for AS.

We have previously demonstrated that loss of function mutations in ERAP1 protect against AS, which raises the possibility that small molecule inhibitors of ERAP1 could be useful drugs in this disorder.

Our work in 2007 (Burton et al., 2007) also demonstrated a strong association with the IL-23 receptor locus that has also been described in inflammatory bowel disease and psoriasis, two conditions which occur much more frequently in people with AS than the general population.  Tugce Karaderi, a DPhil student in our group, confirmed this association in a large meta-analysis (Karaderi et al., 2009). This association has subsequently been refined in further large studies (Evans et al., 2011; Cortes et al., 2013).

Furthermore, it is now apparent that there are two or more quite distinct genetic effects arising from the IL-23R/IL-12RB2 genetic regions. Currently, Dr Amity Roberts, a post-doctoral scientist in our group is evaluating these associations, particularly by looking at the functional effects that specific associated single nucleotide polymorphisms (SNP) have on transcription factor binding in this region. 

We perform various functional tests in our laboratory including the analysis of transcription factor binding, epigenetic marks and gene reporter assays to help describe the functional basis of genetic associations in AS. Prof Wordsworth was UK lead for the successful phase 2 trial in AS of secukinumab, a therapeutic monoclonal antibody targeting IL17A, the end product of the IL-23 driven pathways highlighted as a potential target for treatment by these genetic studies (Baeten et al., 2013).

We have also followed up several other “hits” from GWAS studies. Dr Matteo Vecellio, a post-doctoral scientist in the group, has identified particular SNPs upstream of the RUNX3 promotor that affect expression of this gene and investigated its effects on CD8 T-cell populations (Vecellio M et al., in press).

Since CD8 T-cell numbers are affected in AS this may provide fundamental insights into the pathogenesis of the disorder. Of the more than 60 genetic associations currently identified many do not have an obvious explanation. This is well exemplified by our studies on the ANTXR2 locus. The association has been independently confirmed (Karaderi et al., 2014) but it is far from clear whether this gene is the source of the primary association with AS or whether another locus may be primarily involved.

Our long-term goal is to produce a complete catologue of the genes involved in the aetiology of AS and the functional effects of the AS-associated SNPs. We are further supported in this work by Prof Julian Knight’s group at the Wellcome Trust Centre for Human Genetics in Oxford (Prof Wordsworth, Prof Bowness and Prof Opperman are co-applicants on Prof Knight’s current 5 year Arthritis Research UK Programme Grant).

Through our long-term collaboration with the Immunogenetics Group of Professor Matthew Brown at the University of Queensland we are currently part way through a further high-density GWAS that will study 20,000 cases at far greater depth than has so far been achieved.

We are confident that by the end of 2016 the number of genes credibly associated with AS will have risen to over 100, at least some of which will be potential targets for treatment. There are likely to be considerable opportunities for research positions within the group for those interested in functional genomics and bioinformatics.