BY MARIA BURKE
A new large-scale study has found that myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is partly caused by genes related to both the immune and nervous systems. The findings should advance understanding of the illness as well as help future development of diagnostic tools.
ME/CFS is a debilitating, incapacitating disease that often starts after an infection and affects more women than men. Sufferers may struggle to clear infections and experience ongoing symptoms of pain, fatigue and illness from which other people recover. Little is known about the biological mechanisms that cause ME/CFS, despite much research to uncover them, and it has no effective treatments.
The new study, Decode ME, is a genome-wide association study (or GWAS) run by the MRC Human Genetics Unit at the University of Edinburgh, UK. A GWAS is a relatively new technique that focuses on small differences in DNA between people. As DNA remains unchanged by disease, any DNA differences linked to the disease must be a cause of disease, not an effect. Identifying differences between ME/CFS sufferers and healthy people would indicate what is going wrong in people with ME/CFS at a biological level. Specifically, it should help identify genes, biological molecules and types of cells that probably play a part in causing ME/CFS.
Using UK Biobank data, Decode ME compared the DNA of 15,579 people with ME/CFS with the DNA of 259,909 people without ME/CFS, all of European descent. It found that people with ME/CFS are more likely to carry certain DNA differences in eight regions of their genome (University of Edinburgh Research Explorer). In total, it pinpointed 43 protein-coding genes, of which 29 looked especially promising.
However, these differences in DNA are also often found in people without ME/CFS which makes it hard to distinguish who is at risk and who is not. Most of these regions contain several genes. The methods the researchers used did not allow them to conclusively locate the ones most relevant to ME/CFS in each region, but they could pick out the most likely ones.
Three of the most likely genes produce proteins that respond to an infection, the team reports. One such gene, OLFM4, codes for a protein called olfactomedin-4 that is involved in the body’s antimicrobial responses. Another, ZNFX1, is associated with responses to RNA viruses. A third highlighted gene, CA10, has been linked to chronic pain reinforcing neurological contributions to ME/CFS. A fourth, BXL4 – crucial for keeping mitochondria (the batteries of cells) functioning correctly – is identified as being under-expressed in some people with ME/CFS. These signals align with how people with ME/CFS describe their illness.
‘If Decode ME had found a single genetic signal in one gene that clearly caused ME, then this signal would, in time, become diagnostic, and that would be like genetic variants in breast cancer, for example,’ says study lead Chris Ponting of the University of Edinburgh. ‘But ME, we know now, has lots of genetic signals, and each of them is small and not of large effect, so they do not diagnose individuals.’
The chances of getting ME are affected by four factors, he explains. There are the genetic signals just discovered; genetic signals that might be discovered in the future; the non-genetic influences that exist in the environment such as exposure to viruses or bacteria; and how genes and the environment interact. If someone doesn’t carry these genes, it doesn’t mean that they won’t get ME/CFS because of the three other factors involved.
The study also found no genes related to depression or anxiety. In addition, it could not shed light on the overlap between ME/CFS and Long COVID.
‘It’s very clear that the symptomology between Long COVID and ME is highly similar, but not for everyone,’ Ponting says. ‘As a geneticist, the key question is whether there are overlapping genetic factors, and we haven’t found that in DeCode ME with the methods that we’ve employed.’
However, the identification of 43 protein-coding genes is a major advance in debunking the claims of a psychosocial basis of ME/CFS, he concludes. The team hopes that the study will stimulate substantial functional research to translate these findings into new treatments for ME/CFS.
‘This study is, by some way, the largest ever conducted on the genetics of ME/CFS,’ says Alan Carson, a neuropsychiatrist at the University of Edinburgh. ‘To date, there is reasonable evidence of familial clustering which suggests, but does not confirm, a likely genetic contribution to the condition. Most clinicians in the field believe a genetic contribution is likely but one would say that of the majority of disorders.’
‘This is a well-designed study with large numbers and careful case selection,’ Amy Mason, a cardiovascular epidemiologist at the University of Cambridge, comments. ‘It is particularly notable for how well it has recruited ME/CFS patients, who are usually hard to identify in existing national biobanks.’
Mason says it is interesting that the team found no evidence that depression and ME/CFS have shared genetic links, but that they did find evidence of both pain and the immune system being involved. ‘This fits with what patients often report and helps shift the narrative; ME/CFS is not psychosomatic but linked to measurable differences in genes affecting pain and immunity. The study is important because it identifies some key potential areas for future study. It lays the groundwork for other researchers and pharmaceutical companies to follow, by identifying the areas to look at both for understanding the causes of ME/CFS and for developing new drugs to treat it.’
‘It is very interesting to me as an immunologist to see that genes involved in the immune response have been identified: this will guide us to understand better the role of infection and the mechanism of disease-susceptibility,’ says Jackie Cliff of Brunel University of London, UK. ‘Another fascinating finding is the identification of genes related to mitochondrial function: mitochondria are vital for producing cellular energy, and mitochondrial dysfunction has been implicated previously in ME/CFS. I would also like to see analysis of the X and Y chromosome data, to see if they yield insight into the sex-differences observed in ME/CFS incidence, although the role of hormones may be more important than genetics. It would also be of interest to see if people with different disease severity and presentation have different genetic sequences, and potentially different pathogenesis mechanisms.’
But the study does have some limitations.
Carson stresses that replicating the results will be key and points out that diagnosis was done by questionnaire, which has a significant error rate. ‘The National Institute of Health in the US has suggested diagnostic error rates of around one third. Whilst much is made of patients having post-exertional malaise it should be remembered that this does not have a definition, and it is found across multiple disorders, not just ME. Finally, there were significant co-morbidities within the sample with disorders known to have a genetic contribution like depression, chronic pain and irritable bowel syndrome.’
Mason also notes that the study does not show strong replication of these results in other biobanks, although there was some suggestive evidence when using more specific definitions of ME/CFS. This may reflect poor or inconsistent diagnosis data in those other datasets, she says, rather than flaws in the Decode ME findings themselves.
She also points out that the study includes people of European ancestry only because there is a global lack of more diverse genetic data sets, which is a common problem with GWAS studies. ‘However, the results are still likely to be relevant more broadly, as humans are overwhelmingly genetically similar. Another limitation is that they have not yet looked at the X and Y chromosomes, which may be why their results cannot help explain why ME/CFS diagnoses are more common in women. This doesn’t affect the validity of the signals they have found.’
As with any genomic study, the work has limitations because of its inability to detect the sex-bias, notes geneticist Alena Pance of the University of Hertfordshire, UK. ‘This is due to the difficulties presented by this particular disease, mainly that diagnosis is not always clear due to the lack of precise tools and the fact that it is a multigenic and multifactorial disease meaning that is it determined by several genes and environmental factors. These characteristics of the disease make it difficult to select and discriminate patients and healthy donors, which will have an impact on the analysis of the data.’
Much more research is needed to continue this opening path, she says, particularly in the search of ‘expression quantitative trait loci’ for the variants found to associate the genes pinpointed in the study with differences in their expression levels in patients. ‘This is fundamental because ultimately it is the genes affected that will produce more or less of the protein they encode, which will affect cell function. But this is totally dependent on the type of tissue examined because genes are expressed differently in distinct tissues. While the study used the most complete data available for tissue-specific gene expression, it is not exhaustive.’
Visit Blog
