ME Research article, 26 February 2016: Genome-wide associations
In human beings, a very small number of DNA gene sequences differ between individuals, and these consist largely of single nucleotide polymorphisms (SNPs, pronounced “snips”). While most SNPs are silent, others have important consequences; for instance, an HLA-DRB1 gene SNP is associated with a 3–6 fold higher risk of multiple sclerosis, and a SNP mutation in the APOE gene with an increased risk of Alzheimer’s disease. At present, scientists in many countries are working to identify particular SNPs and link them with particular diseases.
Only a handful of studies have tried to locate SNPs associated with specific genes or pathways in ME/CFS patients; one of them was an ME Research UK-funded investigation at St George’s University of London targetting SNPs for particular “susceptibility genes” (read more). Rapid technological developments, however, have meant that ‘genome-wide’ scanning can now be undertaken quickly and at low cost (see a review), and large-scale ‘genome-wide association studies’ (GWAS) have been published for a range of diseases. In GWAS, a complete set of each participant’s DNA (their genome) is obtained from a tissue sample, set on tiny chips and scanned in the laboratory using machine automation. This allows most recognised SNPs in the genome to be rapidly surveyed at one time – a tremendous advance, particularly for the investigation of chronic illnesses like ME/CFS where large numbers of SNP variants are likely to be working in concert.
The most comprehensive GWAS of an ME/CFS cohort yet conducted has just been published by a consortium of researchers from Nevada, USA, Hungary and Russia (read more) . They recruited 42 people with a confirmed diagnosis of ME/CFS (Fukuda 1994 and ICC 2011) and 38 healthy controls; in each, genomic DNA was extracted from white blood cells and loaded onto a Genome-Wide SNP Array (‘SNP chip’) representing over 906,600 known SNPs. In total, 659,094 SNPs passed quality control, and 442 (407 autosomal and 35 on the X chromosome) were found to be associated with ME/CFS, though most of these were in non-coding regions of the genome and believed (at the present time) to have no functional significance.
Overall, 23 SNPs had a strong statistical association with ME/CFS, and the most significant SNP was in an intragenic region of the RECK gene which is thought to have a role in slowing or preventing secondary tumours. Two of the 23 SNPs were in locations associated with T-cell receptors, and another was observed in the area of GRIK3, a gene involved in neurotransmission and also identified in a previous SNP study of ME/CFS (read more).
Twelve SNPs were located within the coding region of a gene, so could ‘linked’ to a particular gene. Two of these were in the immunoglobulin lambda locus, involved in antibody production, and another was in the CLEC4M gene locus involved in the recognition of pathogens. In addition, there were three regions with multiple statistically significant SNPs in proximity to specific genes (the MAP7 gene on chromosome 6; the CCDC7 gene on chromosome 10; and the T-cell receptor loci). Very little is known about the first two, but T-cell receptors are critical components of adaptive immunity, and many believe ME/CFS to have an an autoimmune component.
Like all genome-wide association studies, this was essentially a fishing expedition – a dense and comprehensive survey of the entire genome to find SNPs that differ between ME/CFS patients and healthy people. The authors identified 23 interesting candidates – which are passed on from generation to generation, like all SNPs – but whether they predispose or cause ME/CFS remains unknown. Essentially, the importance of this study lies in showing other researchers where to concentrate their efforts, while recognising that a genetic predisposition to ME/CFS, if one exists at all, is most likely to involve multiple genes acting in concert and the influence of environmental factors.