Genetic Risk Factors of ME/CFS: A Critical Review. Joshua J Dibble, Simon J McGrath, Chris P Ponting. 2020

https://academic.oup.com/hmg/article/doi/10.1093/hmg/ddaa169/5879704

 

Congrats Simon!

 

Is this an error, on p.6?

"Despite GWAS being expensive, sales of medications that have benefitted from this method
already exceed its costs (32), and these are declining rapidly."

Is it the costs that are declining?

 

Thanks! Full credit to the first author Joshua Dibble, Chris Ponting's PhD student, who is funded by Action for ME and Scotland's Chief Scientist Office, and to @Chris Ponting himself.

Yes- perhaps it should say, "and these costs are declining".

The concluding section might be worth sharing:

Expected outcomes of a ME/CFS GWAS

GWAS are proposed to have “substantially improved our understanding of the mechanisms responsible for many rare and common diseases and driven development of novel preventative and therapeutic strategies” (53). This suggests that large GWAS on ME/CFS are overdue. Replicated results from such studies would have four important implications.

Firstly, it would catalyse the gain of much-needed insight into genes, cellular processes and tissues or cell types that causally alter risk for ME/CFS. When combined with functional genomics and other technologies (53), a well-designed GWAS can pinpoint multiple chromosomal locations containing DNA variants that change the activity of genes – in specific cells or tissues – that thereby alter a person‟s risk of ME/CFS. If these genes are known to have an activity in common – such as a mitochondrial or neurological or immunological function – then this common feature prioritises cellular processes and molecular mechanisms that could be causally involved in disease. Framing such causal hypotheses has been aided considerably by the knowledgebase of gene function, including activity levels, molecular mechanism and cellular function, which have been growing substantially and rapidly over recent years as a result of novel and higher throughput technologies.

Secondly, a GWAS would enable detection of genetic signals that ME/CFS shares with other diseases or traits. Methods (e.g. (57)) that compare GWAS summary statistics for ME/CFS and other traits are available to calculate the genetic correlations between them. Genetic signals for ME/CFS could be shared with other diseases just as autoimmune diseases (for example, rheumatoid arthritis, type 1 diabetes, and autoimmune thyroid disease) share such signals and underlying mechanisms of disease (58).

Thirdly, a GWAS could aid stratification of ME/CFS sub-types. Despite their well-defined clinical diagnoses, complex diseases such as type 2 diabetes are caused by diverse molecular and cellular mechanisms (59) and this should also be expected of ME/CFS. Its underlying biological sub-types could eventually be detectable using methods that test for genetic effect heterogeneity (60).

Lastly, discovery of genetic factors for ME/CFS risk might be expected to improve how this disorder is perceived by health professionals and by society at large.


Future perspective

Genetics studies are the best way to understand the aetiology of ME/CFS, due to the causal nature of genetic associations. A large GWAS focused on discovering the biomolecular mechanisms of ME/CFS is urgently needed because no study on the genetics of ME/CFS yet has seen results repeated under replication. For an appropriately powered GWAS, at least 104 participants are required, and an equal or greater number of controls. A strict p < 5x10-8 or 1x10-8 statistical significance threshold must also be applied to reduce the numerous false positive associations seen from the meta-analyses presented here.

Although recruiting thousands of people with ME/CFS – particularly severely affected individuals who are house- or bed-bound – is a challenging task, it will be essential to perform a GWAS using their samples if we are to understand the mechanisms of the disease. With case criteria refined using genetic findings it may then be possible to begin stratifying the disease into distinct sub-types each with a different causal mechanism and potentially a specific treatment.

 

The one ornitine transporter gene identified seems interesting. It's related to the urea cycle, which other studies have found to be abnormal (I can't recall the details, I think at least one metabolomics study reported this). The urea cycle removes toxic ammonia from the body.

How many would describe the effects of overexertion as resembling some sort of poisoning?

Ammonia also affects the brain in particular.

Ammonia is formed when proteins are broke down. Again I recall that some studies reported that energy metabolism in ME/CFS was changed to rely more on amino acids as fuel source. One study by McGregor reported that PEM is associated with hypermetabolism, ie. the body is working hard.

What might be happening here is that the body is trying to compensate for some problem by using more amino acids for energy but runs into problems because ammonia is removed less efficiently from the body.

 

It's ambiguous. In speech the right intonation would get it across, but when written it's not so clear.

 

Joshua, Simon and I also wrote a slightly more accessible version which is available here: https://www.dropbox.com/s/w8lifiejqc6ovmv/Genetic Risk Factors of ME.pdf?dl=0

 

Your numbers have got a bit skewed, @Simon M - it should be 10 to the power of 4 rather than 104. There are probably some other examples.

 

Another possibility is a more indirect effect.

As removing ammonia is of utmost importance, I guess it would be tried to compensate, say by higher frequency by which transporters or ornithine molecules can get matched.

The arginase produces the ornithine (from arginine), so a higher activity of this enzyme may be a compensation for less transporters available.

The argininase is manganese dependent, so blood levels of manganese would be higher. I think, this could be an important factor.


EDIT: This atom is used surprisingly little, but in tasks which may be associated with wound healing (applying some longer associations, sometimes). There is a preview: Innate Immune Cells speak Mn. Its effects on and usage in the nervous system is interesting.

 

Not sure what I'm trying to say here --- could be that a higher incidence of people with ornithine transporter gene variant confirms the switch in metabolism to using amino acids --- Chris Armstrong, Fluge and Mella --- have all suggested a switch to using amino acids for energy production.

 

Joshua Dibble's thesis is here, Thesis: Investigating the Genetic and Immunological Aetiology of Myalgic Encephalomyelitis/Chronic Fatigue Syndrome 2022 Dibble

 

During the ZonMw conference on the Dutch ME/CFS research program yesterday, one researcher said that they estimated the heritability of ME/CFS in the Lifelines cohort at 43% [34-51] - findings that are still to be published.

 

Interesting. A caveat would be though whether the Lifelines cohort identifies ME/CFS accurately or not, and I believe there are doubts about that.

 

I wonder what that is based on. These things usually require something like monozygotic versus dizygotic twin studies.

If heritability was 40% I would have expected some risk loci to have shown up on the smaller screens done so far, although there might be reasons why not.

 

The researcher mentioned above was Martje Bos, who is a ME/CFS patient herself and is studying ME/CFS in relation to other functional syndromes such as IBS and fibromyalgia.

Another researcher, Cindy Boer said that she is leading a new major collaboration: ‘Genetic epidemiology of ME/CFS’ that has 30.000 people with ME/CFS and several hundred thousand without ME/CFS. [EDIT: I corrected this sentence and changed with to without: see the following post by MESci]

When someone asked if their study is not too similar to DecodeME she responded that there have been talks talked with the DecodeME team but that higher sample sizes around 40.000 will be needed and that therefore collaboration is necessary. She also said it is important to validate the results that will come out of DecodeME in other cohorts.

 

I think you've erroneously repeated 'ME/CFS' here. Should it be something else?

 

Apologies, the last one should have been 'without ME/CFS'.

 

The genetic Lifelines project description funded by ZonMW also states that "Getting ME/CFS is possibly 48-56% determined by a person's genetic (DNA) background. This has not been researched extensively until now, even though it is known that genetic research can be used to determine the cause of ME/CFS. The aim of this project is to investigate which genetic (DNA) variations are involved in ME/CFS in order to discover the genes and biological cause(s) of ME/CFS."

Does anybody have any idea where those numbers come from?

https://projecten.zonmw.nl/nl/project/mecvs-genetica-onderzoek-naar-de-biologische-oorzaak

 

So presumably they are doing another genome screen like DecodeME, but yes, the question is where the genetic component calculation comes from.

I thin there may have been a twin analysis in the US from a huge cohort about five years back but I heard no more about that and it seemed that the definition of the cohort might be pretty loose.

 

This is good news, and exactly what we need. With GWAS studies, size makes a big difference, and it’s fairly easy to combine the results of different studies to get more robust results. Separately, it’s also important to validate/replicate findings from one study in a separate, independent cohort.

As Chris Ponting said in the recent excellent piece on Channel 4, this kind of work is 20 years behind other diseases, but it’s good that things are getting going now.

That is very high, and as @Jonathan Edwards says, if it were this high, you’d expect some genetic findings by now.

I think planning for DecodeME estimated heritability at around 10%, based on previous studies, which I think included this one using data from Utah
https://pubmed.ncbi.nam.nih.gov/21619629/

When all the data for DecodeME is ready and has been analysed, it will be possible to an estimate for hereditability, which should then be the best estimate we have.