Genetics of multiple sclerosis in the Northern Isles of Scotland
Barnes, Catriona Louise Kerr
Multiple Sclerosis affects around 2 million people worldwide (Kantarci and Wingerchuk, 2006; Dutta and Trapp, 2011). The disease is typified by the destruction of the central nervous system neurons’ myelin sheaths, caused by the individual’s own immune system (Hauser and Oksenberg, 2006). This destruction results in the inflammation and chronic degeneration of the CNS, causing varying symptoms including pain, fatigue, cognitive impairment and paralysis (Costelloe et al., 2008). Not only is the life expectancy of individuals with MS 10 years below that of the age-matched general population (Ragonese et al., 2008), but life quality is often severely affected from the start of disease onset (typically around 30 years of age (Hauser and Oksenberg, 2006)). There are several treatments available to aid in the relief of specific symptoms; however, the treatment is lifelong which places a burden on healthcare services. Current research looks to expand the available knowledge on the causes of MS, to improve preventative measures (such as lifestyle changes to accommodate environmental conditions) and targeted treatments (for example, focusing on the product of an MS-associated gene variant). Of particular interest to MS research are the two population isolates of Orkney and Shetland, which together make up the Northern Isles of Scotland. Shetland has 295 MS cases per 100,000 individuals, while Orkney has the highest global prevalence of MS at 402 cases per 100,000 individuals (Visser et al., 2012). Orkney, at a lower latitude than Shetland, has a significantly higher prevalence than what would be expected. Multiple theories behind the excess of MS cases in Orkney have been investigated, including vitamin D deficiency and homozygosity: neither were found to cause the high prevalence of MS. It is possible that this excess prevalence may be explained through unique genetics. This thesis sought to better understand these high rates of MS, with the aim of passing this knowledge on to the island residents of Orkney and Shetland and to contribute the findings to the wider understanding of MS. Analyses were conducted using the ORCADES and VIKING datasets. ORCADES contained 2215 individuals from the Orkney islands, including 97 MS cases (some recruited because they were cases); VIKING contained 2015 individuals from the Shetland islands, including 15 cases. First, a heritability study was conducted using GCTA to determine the SNP heritability of MS in both Orkney and Shetland and how it compared to published estimates of heritability. The SNP heritability of MS in Orkney was estimated at 0.31 (95% CI 0.13, 0.49). An estimate of SNP heritability for MS in Shetland could not be determined due to low case numbers. Second, a genome-wide association study was conducted using a combined ORCADES/VIKING dataset containing 112 cases and 4223 controls. The aim of this study was to determine if unique common MS risk variants existed in the Northern Isles. Here, 89 SNPs were identified to suggestive significance, mostly within six key regions of the genome. Within the literature, only one of these (chromosome 6 SNP rs9268154) was associated with Multiple Sclerosis. Four of the five other regions had possible functions within the immune or nervous system. However, as these did not reach genome-wide significance it is likely these results were due to chance; further investigation is needed to clarify this. Finally, a polygenic risk score study looked at the contribution of common risk variants to MS. The 127 most strongly associated MS SNPs were used to calculate risk scores in mainland Scotland, Orkney and Shetland. These risk scores were compared between controls in all three populations to determine if the Northern Isles, by chance, had higher frequencies of common risk variants and if this contributed to the excess of cases. These common risk variants explained 3% of variance in MS risk, and had an AUC score of 0.69 (95% CI 0.65, 0.74). However, no difference existed between common risk variants in the three populations, aside from one variant: rs9271069, a tag SNP for HLA-DRB1*1501. This SNP was found to have a significantly higher frequency in Orkney (RAF = 0.23, p-value = 8 x 10-13) and Shetland (RAF = 0.21, p-value = 2.3 x 10-6) than mainland Scotland (RAF = 0.17). This SNP accounted for 6 cases (95% CI 3, 8) out of 150 observed excess cases per 100,000 individuals in Shetland and 9 cases (95% CI 8, 11) of the observed 257 excess cases per 100,000 individuals in Orkney. The question of why the Northern Isles have such high rates of MS remains open. This thesis explains a small proportion of this excess. It is hoped that the findings and discussions found here will encourage dialogue within the Northern Isles and bring awareness to the genetic, environmental and lifestyle factors that contribute to MS within the islands.