Prototype genetics app looking at PLA2-activation pathways in ME/CFS

[TamaraRC]

I have been working on a prototype genetics app based on the hypothesis that phospholipase A2 (PLA2) activation may be increased in at least some people with ME/CFS.

It is not intended to diagnose ME/CFS, and I don’t want to present it as clinically validated. It looks through a person’s genetic data for variants in pathways that could plausibly contribute to increased PLA2 activation. I worked on making the app several months ago, but then paused work on it because I was unsure how accurate the variant scoring would be, and I also became busy with other things. I am now considering posting it online as a work-in-progress, to gain feedback.

The app currently works by grouping variants into pathway categories rather than trying to identify a single “ME/CFS gene”. For example, it looks at possible genetic contributors to:

* immune/cytokine-driven PLA2 activation
* oxidised LDL or lipoprotein-related PLA2 activation
* histamine/mast-cell signalling that could feed into cPLA2 activation
* phosphatidylcholine synthesis and phospholipid buffering
* arachidonic acid handling and eicosanoid pathways

My concern is that although many of the pathways are biologically plausible, the evidence for individual SNPs is variable. Some variants have reasonably clear functional or association data, while others are more speculative or context-dependent. I therefore think it would be much safer and more useful to publish it openly as a hypothesis-generating tool and ask for feedback, rather than treating the output as any kind of medical or diagnostic result.

My aim is not to claim that PLA2 activation explains all ME/CFS, but to explore whether a subgroup of patients might show genetic patterns that make excessive phospholipid hydrolysis more likely.

I think there could be subsets- one subset of people could have increased PLA2 activation due to increased oxidation of LDL, and others may have increased PLA2 activation due to immune pathways, some could be a mixture of both.

 

[Jonathan Edwards]

Would you not expect 'Subsets' of people with ME/CFS to have all sorts of SNP patterns, just by chance? I discussed with Chris Ponting some time back the options for looking for correlations between SNP variants within the ME/CFS population and I think his view was that there are so many ways you can analyse that you end up with a statistically meaningless exercise.

 

[TamaraRC]

Would you not expect 'Subsets' of people with ME/CFS to have all sorts of SNP patterns, just by chance? I discussed with Chris Ponting some time back the options for looking for correlations between SNP variants within the ME/CFS population and I think his view was that there are so many ways you can analyse that you end up with a statistically meaningless exercise.

Yes that’s a fair point but I think that finding gene variants in a pathway is more meaningful if you are approaching it from a specific, pre-defined biological hypothesis, rather than freely searching for any SNP pattern that happens to appear in ME/CFS.

I think different pathways leading to a unifying dysregulation would be a good explanation for why people report different medications helping, for example there have been some reports that some people improve on JAK-STAT inhibitors- this could be because they have gene variants that cause high and prolonged IFN gamma, which increases sPLA2-IIA.

I have improved by taking supplements that reduce oxidative stress and statins. I have a lot of genetic variants related to LDL oxidation, reduced LDL receptor uptake and high homocysteine, so this fits with my recovery.

I also have genetic variants related to increased incorporation of arachidonic acid into PC which makes it more vulnerable to being oxidised in LDL, which makes it more vulnerable to hydrolysis by Lp-PLA2, and I also have a direct gene variant of Lp-PLA2 which increases its activity.



There is research that shows LDL in ME/CFS patients is more vulnerable to oxidation-

Antioxidant status and lipoprotein peroxidation in chronic fatigue syndrome

Obviously, a lot of people do not improve from taking statins, PC supplements and oxidative stress supplements, so this is clearly not happening in the majority but I think that most patients would likely have immune mediated PLA2 activation and a smaller subset would have oxLDL-related PLA2 activation.

 

[Creekside]

Could a result of "PWME are more likely to have this gene" have no present function to identify? Could a genetic factor only have its effect before the third trimester, or if the subject gets a cold virus before age 2, or has too much or too little of nutrient whichever at some stage of life?

Genetic studies can be important if the gene has a single, clear function at the present time, such as producing too much or too little of a specific protein. However, it's not the right tool for all diseases.

 

[Jonathan Edwards]

I think different pathways leading to a unifying dysregulation would be a good explanation for why people report different medications helping,

The problem there is that we have no reliable evidence for any of these medications actually working. I agree that focusing on a predetermined set of genes would help to get low p values but I think you still have to do that on all people with ME/CFS. As sson as you start subsetting you rapidly enter the zone of cherry-picking and statistics go out of the window.

The Precision Life people claim to have made legitimate use of subsetting but I don't know enough about the methodology to be convinced it holds water. For the common variant SNPs looked at so far the risk level is fairly marginal, with the total genetic effect being no more than 10-20%, so I see the hope of being able to relate genes to individual mechanistic subsets as pretty remote.