Preprint Dissecting the genetic complexity of myalgic encephalomyelitis/chronic fatigue syndrome via deep learning-powered genome analysis, 2025, Zhang+

[Jonathan Edwards]

HLA-B and C are usually in strong linkage disequilibrium as well

Indeed and that begs even more questions.

Wouldn't it be fun if working out the mechanism of ME/CFS had the spin off of explaining all the mysterious things we do not understand about other chronic immune disorders. The exception, by probing the rule, might show how it actually works.

 

[Utsikt]

Wouldn't it be fun if working out the mechanism of ME/CFS had the spin off of explaining all the mysterious things we do not understand about other chronic immune disorders.

If that happens, I hope it might teach one or two people that ignoring one disorder might be detrimental for many of them.

Although that should already have happened with ME/CFS due to the pandemic.

 

[V.R.T.]

I think the data may be telling us a lot about why we have not been finding anything so far in the effector biology and where we should be looking instead. If anything I think it confirms various thoughts people have been having on that.

Can you say more about this?

 

[Robert 1973]

My hunch is that they will be showing interest fairly soon. This is neurobiology and there are more neurobiologists in the UK than you have had hot dinners - literally.

Would you say that the data is starting to point towards neurologists as being the physicians who ought to be taking responsibility for running ME/CFS clinics?

[edited to make sense!]

 

[Kitty]

Wouldn't it be fun if working out the mechanism of ME/CFS had the spin off of explaining all the mysterious things we do not understand about other chronic immune disorders.

That's what I found interesting about there not being many disease associations.

Maybe there just aren't many obvious associations.

 

[Jonathan Edwards]

Can you say more about this?

There are all sorts of angles on this but in simple terms the data suggest that the problem may be with lymphocytes signalling to nerves (and maybe nerves signalling back to lymphocytes) without that being mediated by any inflammatory or damaging process outside lymphoid tissue itself. If the signalling is mostly mediated by cell surface ligand or local cytokine interaction nothing my be visible on standard tests.

It does looks if shifts in certain metabolic pathways and perhaps specifically muscle pathways may be susceptibility factors but if those were facilitating a nerve-mediated effect on muscle again there would be nothing obvious to see. Snow Leopard has been emphasising that muscle afferents might be malfunctioning and that might feed into efferent effects not just on power but on metabolic pathways perhaps.

One basic message is that beyond gross anatomical lesions neurology remains very poor at predicting clinical effects from knowledge about how networks of nerves actually process information and generate outputs. It is a bit like knowing where the main rush hour traffic spots are but having no idea where individual cars are going or why. Nerve function is just too complicated to build testable models around. Nobody knows why Stephane Grappelli was so good at playing the violin.

 

[Jonathan Edwards]

Would you say that the data is starting to point towards neurologists as being the physicians who ought to be taking responsibility for running ME/CFS clinics?

Not sure about that. Maybe we need some proper clinical immunologists who can cross systems.

 

[Sasha]

If the signalling is mostly mediated by cell surface ligand or local cytokine interaction nothing my be visible on standard tests.

It does looks if shifts in certain metabolic pathways and perhaps specifically muscle pathways may be susceptibility factors but if those were facilitating a nerve-mediated effect on muscle again there would be nothing obvious to see.

If that's the case, how could it ever be proved empirically?

 

[Kitty]

If that's the case, how could it ever be proved empirically?

If we've an idea what it is, there might be ways to show it (even if obliquely). We don't see it at the moment because we're not testing for it.

 

[forestglip]

With the synapse-related findings, it might be worth revisiting this other GWAS that found a nervous system related gene:

Genetic association study in myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) identifies several potential risk loci (Hajdarevic et al, 2022, Brain, Behavior, and Immunity)

Our most significant finding was with SNPs encompassing the TPPP region. TPPP SNPs showed association signals in both the Norwegian and the UK dataset. However, the strongest association signals were seen with different SNPs, i.e. a rare variant in the Norwegian cohort and a more common variant in the UK dataset. However, the LD analyses indicated clear haplotype patterns, and globally haplotype analyses showed association in both cohorts. This could potentially indicate that these SNPs pick up a common causal risk variant that had not been included in our current analyses. Notably, the region just centromeric of the association peak was poorly covered by SNPs in our combined meta-analysis.

Interestingly, the TPPP gene, encoding the tubulin polymerization promoting protein, is mainly expressed in the brain. The TPPP protein plays a pivotal role in the myelination of oligodendrocytes (Fu et al., 2019) and has been shown to correlate with shortened disease duration in multiple sclerosis (Höftberger et al., 2010). This may indicate a role for TPPP in myelin repair. Hence, changes in this gene may underlie neurological abnormalities and may be involved in pathologies like Alzheimer's disease.

They do also mention a gene SHANK3 which is involved in synapses:

In light of these observations, we investigated if our initial approach had overlooked other regions by performing a combined analysis of all SNPs across the genome available from the Norwegian and UK Biobank cohorts. This analysis revealed six novel associated regions (P < 1 × 10−5), three of these with several associated SNPs spanning EPHA7, SKAP1 and SHANK3 (Supplementary Table 8). The P-values for these SNPs were all < 0.0001 in the UK biobank, while most were non-significant in the Norwegian cohort, where the strongest P-values were seen for SNPs in SKAP1 (P = 0.02).

This gene is a member of the Shank gene family. Shank proteins are multidomain scaffold proteins of the postsynaptic density that connect neurotransmitter receptors, ion channels, and other membrane proteins to the actin cytoskeleton and G-protein-coupled signaling pathways. Shank proteins also play a role in synapse formation and dendritic spine maturation. Mutations in this gene are a cause of autism spectrum disorder (ASD), which is characterized by impairments in social interaction and communication, and restricted behavioral patterns and interests. Mutations in this gene also cause schizophrenia type 15, and are a major causative factor in the neurological symptoms of 22q13.3 deletion syndrome, which is also known as Phelan-McDermid syndrome. Additional isoforms have been described for this gene but they have not yet been experimentally verified. [provided by RefSeq, Mar 2012]

For the Zhang study, I searched the full gene list for the three SHANKs. Keeping in mind that there are 17,759 total genes in the table, SHANK1 ranked 131, SHANK3 ranked 327, and SHANK2 ranked 490. I don't know the actual probability, but I think it would be unlikely for all three to rank so high due to chance.

Though TPPP ranked 13,879. (The rankings of the other TPPP paralogs TPPP2 and TPPP3 weren't impressive here either: 9975 and 13,459 respectively)

 

[wigglethemouse]

If the synapses are highlighted, should researchers look into herpes viruses that infect and can travel in nerves. Could herpes virus be somehow disrupting the signalling?

Maybe a question for Jackie Cliff. Her HHV6B study seemed to track virus levels in saliva (or was it mucus?) with severity I think (ongoing work). Could HHV6 also be present in the sinuses, very close to the brain and olfactory nerves....

Google AI Overview for HHV6B and Synapse

HHV-6B, a human herpesvirus, can infect various brain cell types, including neurons and glial cells, and may affect synaptic function. HHV-6B can infect both glutamatergic (VGluT1-positive) and dopaminergic neurons, but not GABAergic neurons (GAD67-positive). While HHV-6B may have greater transmission potential than HHV-6A, HHV-6A induces more severe cytopathic effects in susceptible cells. Additionally, HHV-6B can impair synapse maturation and neurite outgrowth in primary human cortical neurons. Studies have also linked HHV-6 to neurodegenerative diseases like Alzheimer's, with potential involvement of synaptic dysfunction

The 1980's Scottish Ayrshire ME/CFS outbreak was tied to Coxsackie B4. This paper on that virus has some interesting highlights.
Coxsackievirus B4 infection and interneuronal spread in primary cultured neurons

* Coxsackievirus B4 infects and replicates in primary neurons.
* Coxsackievirus B4 neutralizing serum does not preserve primary neurons from viral cytopathic effect.
* Coxsackievirus B4 neutralizing serum does not enhance the viability of primary neurons.
* Coxsackievirus B4 is transmitted between neurons via synapses.

 

[Jonathan Edwards]

If that's the case, how could it ever be proved empirically?

Well, one possible option is the sort of Fluorospot test we have seen recently reported as showing an increase in gamma interferon production by CD8 cells when in contact with macrophages. The test shows signals as they are passed from one cell to the next at a microscopic level if you allow the cells to interact.

The other thing is that if you can infer indirectly what is likely to be happening you can try a therapeutic experiment - block the signal you think is working out of sight. That is what I did for RA. I had no way of seeing the submicroscopic event of TNF release in response to Fc receptor binding actually in the RA joint. But I worked out it must be going on so I tried taking away the antibodies that I thought were binding to the receptor. And it worked rather well.

 

[Jonathan Edwards]

With the synapse-related findings, it might be worth revisiting this other GWAS that found a nervous system related gene:

Ye, I think we are going to find that this all fits together. Synaptic and dendrite housekeeping looks to be central to what is going wrong.

 

[hotblack]

I am thinking more of Sancho Panza or Baldrick.

Whatever happened to, all the heroes?

 

[Jonathan Edwards]

If the synapses are highlighted, should researchers look into herpes viruses that infect and can travel in nerves. Could herpes virus be somehow disrupting the signalling?

One thing I keep coming back to is that if ME/CFS was due continued response to a persistent virus then we would almost certainly be talking about one unique virus. Polio produced a unique neurological illness, so does varicella zoster (shingles), so does measles (SSPE). Yet we are pretty sure that ME/CFS is triggered by all sorts of viruses and even bacteria (Q fever).

But there may be an answer to this paradox - that the triggering microbe may not be the one that mediates the long term illness. Maybe triggering by Tom Dick or Harry (maybe even Harry Lyme - zither tune plays in the background) sets up a persistent response to Jack virus, who has been sitting there in the nervous system of all of us since infancy. A Herpes virus would fit that scenario.

 

[hotblack]

HLA-B and C are usually in strong linkage disequilibrium as well

I’m going to try a definition since this has been coming up a lot recently…

Linkage disequilibrium is when at a population scale different segments of DNA, different genetic variations, occur together in different places more or less frequently than you would expect by chance.

About right?

 

[duncan]

Would you say that the data is starting to point towards neurologists as being the physicians who ought to be taking responsibility for running ME/CFS clinics?

I would have thought so. I sort of still do.

Then again, neurologists are in charge of channelopathies - God knows why - and you can't find half of a baker's dozen who know anything about anything when it comes to channelopathies.

To take it a bit further, if persistence is a factor, don't infectious disease doctors need to be involved? I suppose maybe not; by understanding the downstream mechanisms at work, do you bypass the need for what they may not even be able to bring to the table?

 

[duncan]

One thing I keep coming back to is that if ME/CFS was due continued response to a persistent virus then we would almost certainly be talking about one unique virus.

I'm not sure I follow this. ME/CFS appears in different forms. Couldn't that fact be due to discrete bugs?

Polio produced a unique neurological illness, so does varicella zoster (shingles), so does measles (SSPE). Yet we are pretty sure that ME/CFS is triggered by all sorts of viruses and even bacteria (Q fever).

There are ME/CFS schools of thought where there can be multiple agents/triggers. Perhaps pretty much any agent can set things into motion, but it's old entrenched latent viruses that emerge to do the heavy lifting? I'm unclear how accurate testing for the latter really is, so how would we know?

ut there may be an answer to this paradox - that the triggering microbe may not be the one that mediates the long term illness. Maybe triggering by Tom Dick or Harry (maybe even Harry Lyme - zither tune plays in the background) sets up a persistent response to Jack virus, who has been sitting there in the nervous system of all of us since infancy. A Herpes virus would fit that scenario.

Ha! Should have read your entire post before writing. Yes, this is the thinking among some others. The trick is proving it.

The genetic work that is being done may obviate the need to know which microbe(s) is responsible, and that would be fantastic. I'd still want to know.

 

[hotblack]

Back on the HLA-C talk. If I’m understanding this correctly it seems interesting that this is an area involved in the heavy chain given Audrey Ryback’s paper on BCR heavy chain repertoires?

The strong linkage disequilibrium may be throwing things off for HLA-C specifically though?

If this is an area involved in presenting stuff from inside the cell and particularly “presenting peptides derived from endoplasmic reticulum lumen” (from https://www.ncbi.nlm.nih.gov/gene/3107) could there be something going sideways in translation or protein folding? I seem to remember questions about this coming up before.

Or maybe this is all just linked to our susceptibility to particular infectious triggers and the ongoing process is elsewhere?

 

[Jonathan Edwards]

Linkage disequilibrium is when at a population scale different segments of DNA, different genetic variations, occur together in different places more or less frequently than you would be expected by chance.

Yes. Every generation DNA is supposed to be randomly chopped up and exchanged between sister chromosome threads. Which should mean that if a great-great-great-grandmother has HLA A3, B6,C2,DR2 genes in line on a chromosome that set will gradually get separated and re-assorted with each generation. The process is slow so in populations you see genes being passed down in sets for many generations but over millions of years the sets could be randomly re-sorted.

'Linkage disequilibrium' is sometimes used to mean slightly different concepts but here I understand it to mean that for some reason a set of genes doesn't get split up as often as expected, or, often assumed, that it may get split up but there is a significant survival difference in chromosomes with the original set and those with the set split up. So you may end up with a population where A3,B6,C2 and DR2 keep showing up as a set even over millions of years presumably because this is a really good set to have.

Up until now linkage disequilibrium has mostly been seen as a nuisance in genetic studies - it means that it may be difficult to pin down which gene is really related to a disease. But linkage disequilibrium is itself of biological interest. It should be telling us something but often nobody is quite sure what.