Geoff’s Narration

The GIST

 

Finally, we have a study that’s worthy of this disease. This is a disease, after all, that affects millions, is one of the most functionally disabling diseases known to man, and mostly strikes people in the prime of their lives. It’s a pretty big bore disease, and it needs big bore studies to match it.

It finally got one. Everything about this DecodeME study, “Initial findings from the DecodeME genome-wide association study of myalgic encephalomyelitis/chronic fatigue syndrome,” was big. Ten years of work culminated in a study which included almost 16,000 ME/CFS participants and over 250,000 controls. Over 50 authors (plus over 50 contributors) produced the paper. (The next biggest ME/CFS genetic study had about 2,500 participants).

The University of Edinburgh, from which the study emanated, called it a “landmark” study (it is), and a “milestone in ME/CFS research”. A milestone is “an action or event marking a significant change or stage in development”.

Putting together a 15,000 person genetic study (they have samples from 27,000 people with ME/CFS), getting the funding (convincing both the NIH and MRC to fund it), recruiting the participants, building the databases (the databases, by the way, are open to other researchers), doing the complex statistical analyses – i.e. being able to produce a study of this size and complexity is indeed in milestone in ME/CFS research. It should be sending out the alert that the ME/CFS research community is up to big things.

Big Stakes

The stakes were not small. A null result would have cast doubt on ME/CFS as a disease entity – something that some would have welcomed. In a disease that appears to be triggered by virtually every pathogen under the sun, and which is not always associated with an infectious onset – maybe a null genetic finding was not an impossibility.

At the recent Keystone Long COVID Conference, David Putrino noted how complex the triggering factors for diseases like ME/CFS can be. In some cases, there are no clear triggering factors. In other cases, the trigger (infection, trauma, toxin exposure) is clear. In other cases, a careful patient history will reveal a sequence of triggering factors; i.e., there’s an earlier hit, then another hit, and then maybe another hit which finally tips things upside down.

ME/CFS isn’t HIV/AIDS, multiple sclerosis, or even long COVID: it’s on an order of complexity higher than them.

When Ponting and company put the big, heterogeneous, symptom-diagnosed, “waste-basket”, “yuppie-flu” (fill in whatever negative connotations you want to), etc., disease to the test, the stakes were pretty high. If ME/CFS is as messy as some people think and isn’t a disease at all, they could have found nothing! Alternatively, they could have found a bunch of gobbledygook that didn’t make sense with what we know, suggesting that ME/CFS is indecipherable at a basic genetic level.

Thankfully, neither of those happened. What they found: a) not only made sense with what we know about ME/CFS; but b) provided insights for further research. Dr. Ponting told David Tuller that he was “exhausted and relieved that the study found genetic signals”.

Note that the results pertain to the broad ME/CFS population – not to any one individual. The study was designed to illuminate regions of the genome that shed light on what’s happening in ME/CFS.

Big Genetics

Note that this paper is a preprint and may undergo some changes once the reviewers review it. The authors, though, are well-published and were confident enough in their results to release the paper on the University of Edinburgh’s site.

The DecodeME study has samples from about 27,000 people. For this study, the researchers chose to assess samples from about 16,000 people who were from common genetic (European) backgrounds.

For a change, the size of the study was good. Genome-wide association (GWAS) studies for common diseases typically contain several to tens of thousands of participants, and can run into the millions. Because the more complex the disease is, the larger the sample size that is desired, larger studies – which are on their way – can only help decipher this decidedly complex disease. Since at least 10,000 samples are needed for common diseases, and this study had about 16,000, the size was acceptable.

The study wasn’t just big participant-wise – it was big genetically. In its attempt to find vulnerabilities that could help explain ME/CFS, this GWAS study assessed millions of common genetic variants. Because these variants are baked in (i.e., they are present when we are born), DecodeME had the potential to uncover core biological issues that helped to set the stage for ME/CFS.

Note that complex chronic diseases like ME/CFS are the result of many factors, of which genes are only one, and that this GWAS is only one of several kinds of genetic studies that can be done. This first stage of the DecodeME, for instance, was not designed to pick up rare or novel genes that a “whole genome study” (WGS), which covers the entire DNA of each individual (all 3 billion base pairs), would pick up. Nor was it designed to pick up epigenetic changes which alter the expression of our genes over time.

We wouldn’t expect DecodeME, therefore, to uncover rare or novel genetic changes associated with ME/CFS. By highlighting specific regions of the genome, though, GWAS studies help to “map the genetic architecture of complex diseases”; i.e. the study is looking for somewhat broad regions of our genome that may predispose us to ME/CFS. By doing so, it can help researchers determine which biological pathways to explore, and can point to therapeutic targets.

Results

The authors asserted the study put ME/CFS on a firm biological foundation.

“DecodeME’s results, grounded in the principles of statistical genetics, now place ME/CFS research on a firm biological foundation. They begin to explain the disease’s heritable component, they improve the likelihood of finding effective drugs for ME/CFS, and they place this long-neglected disease on more equal terms with other common genetic conditions”, the authors.

One of the most important findings of the study is simply that it found regions of the genome that were disturbed across a very wide array of ME/CFS patients. That finding places, as the authors said, ME/CFS research on a “firm biological foundation”.

The study performed six GWAS analyses, of which the first was most prominent and the second was included. DecodeME cases (The others stratified the ME/CFS group by sex and infectious onset; and two comparison analyses were done). For a variety of possible factors, including less stringent ME/CFS diagnostic criteria and smaller sizes, the two comparative analyses did not replicate the DecodeME findings.

Reliability

These six GWAS showed that their statistical results closely matched what would be expected if there were no bias or confounding in the study. This supports the reliability of these findings and indicates the results are unlikely to be skewed by subtle differences in ancestry, or technical issues.

Heritability

The heritability of ME/CFS, based on the common genetic polymorphisms found, was a modest 9.5%. This suggests that common genetic variants contribute about 10% to the risk of developing ME/CFS.

This puts ME/CFS at the lower end of heritability for common chronic diseases (Crohn’s disease – 24.6%; multiple sclerosis: –  45%; heart attack – 38%) but interestingly enough, in line with other diseases, it’s been associated with those such as migraine (11%), irritable bowel syndrome (10-13%), endometriosis (10-12%), asthma (11%) and long COVID (@12%).

In David Tuller’s interview, Ponting noted that a relatively small effect size in a study can translate into an large effect with a drug; that is – this study could point to enormously effective drugs.

The modest heritability suggests that rare or novel gene variants not picked up by the GWAS analysis, and/or that environmental influences (infection, toxin exposure, etc.) play a larger role than the common variants of the genes we are born with.

Some ME/CFS researchers, such as Xiao and Liz Worthey, believe that rare and/or novel gene variants may play a large role. DecodeME has apparently saved half the genetic material it’s gathered for whole genome studies (WGS) to flesh out the role rarer gene variants might play. (Whether it has the funding to do that is unclear.) As to pathogen exposure, no studies that I can remember suggest that people with ME/CFS have been exposed to more pathogens than the general population.

The Genetic Architecture of ME/CFS?

The primary analysis compared 15,579 DecodeME cases with 259,909 UKB controls and yielded six “genome-wide significant loci” and the second GWAS analysis added two more loci. The infectious subset analysis yielded three significant loci or regions, two of which overlapped with the first GWAS analysis – a nice, consistent finding.

The term “genome-wide significant locus” refers to a stretch of DNA containing a “lead” genetic polymorphism (variation) that is closely accompanied by other genetic polymorphisms that appear to contribute to the disease; i.e., these are rather broad regions of the genome that appear dysregulated in ME/CFS.  Regions must meet strict statistical criteria to be considered “genome-wide significant loci”. These loci could be said to represent the “genetic architecture” of the disease.

In an interview with David Tuller, Dr. Ponting said he believed that the study was so rigorously done that he believed the results were “causal”; i.e. they directly pointed to causes of ME/CFS.

The number of loci found in this first study is similar to that found in the first studies of other illnesses (5-15) and would be expected to grow over time as bigger studies are done. Very large studies have plucked out dozens to 50 or more loci.

The loci are described in part below.

The Tier I Genes 

The study identified variants in 43 protein-coding genes found in the “genome significant loci” that could be affecting how much protein is being produced. Because these kinds of genes directly affect functioning, they are prime candidates in any disease. Each of these genes also needed to be present in one of the tissues the genetic analysis suggested were most affected in ME/CFS (see below).

The authors prioritized 29 genes they called Tier I genes for which there is “strong evidence” that the gene contributes to the risk of having ME/CFS. Some of the genetic loci and genes believed to contribute most to ME/CFS are below.

chr1q25.1 – boasting no less than 11 Top Tier genes, the chr1q25.1 locus was the clear winner in the significant genome-wide ME/CFS loci stakes. This locus plays a role in brain development and immune signaling.

The RABGAP1L  gene popped out big time in this locus when an analysis suggested this gene variant may be operative in a remarkably high number of the tissues of concern in ME/CFS. Given that the RABGAP1L gene plays an important role in inhibiting viral replication and immune and nervous system health, it’s clearly a top candidate for further study.

chr6p22.2 – The chr6p22.2 locus found in ME/CFS “spans a segment rich in functionally important genes” that contribute to neurodevelopmental and immune conditions. The authors proposed that the gene variants found in this locus could “impair T-cell responses and worsen autoimmune disease.” No less than seven Tier one genes were found in this locus.

chr12q24.23 – The chr12q24.23 ME/CFS-associated locus contained 3 Tier 1 genes – Large studies suggest that this locus affects energy balance, metabolism, susceptibility to allergic disorders, and cardiovascular health.

One of the genes associated with this locus (SUDS3) is a negative regulator of microglial inflammation. Suppression of the SUDS3 gene results in microglia pumping our more inflammatory cytokines and factors. The MAGMA analysis suggested this gene could be operative in almost half the tissues of concern in ME/CFS.

chr17q22 – Only one top tier gene – CA10 – was found in the chr17q22 interval but it was a doozy. CA10 plays a fundamental role in nervous system circuits involved in pain sensation and neuroplasticity. In fact, multiple genome-wide association studies (GWAS) of chronic musculoskeletal pain, multisite chronic pain, and back pain have implicated CA10 in these diseases.

Altered CA10 expression, interestingly, could modify the signals coming from the dorsal horn of the spinal cord – the main hub for sensory signals entering the central nervous system – and an item of interest in both ME/CFS and fibromyalgia.

chr20q13.13 contained seven top-tier genes. The role the CSEIL gene plays in macrophages could affect the inflammation found in ME/CFS. Plus, this gene regulates the secretion of an enzyme that could disrupt the blood-brain barrier.

Note that the themes in the top-tier genes – brain development, neurodevelopment, viral replication, energy balance, metabolism, inflammation – are the kinds of themes we would expect to show in a genetic analysis of ME/CFS; a sign that DecodeME is on the right track, and is uncovering some of the biological roots of this disease.

They are also unique! Dr. Ponting said they don’t fit genes found in diseases like arthritis, multiple sclerosis, or depression, which makes sense given what a red herring ME/CFS has been – and underscores the fact that ME/CFS is a unique biological entity.

The Whole Gene Assessment

Next, instead of looking for single polymorphisms in genes, the authors assessed all the variations found in each gene to see if whole genes might be associated with ME/CFS. Of the 18,637 genes, they found 13 genes that were mostly involved in familiar themes: immune system regulation, the neuro-immune interface, and metabolic and detox pathways.

Noting that AI is great at finding patterns and connections, and not so great at determining if they are valid or not, check out a model AI Perplexity suggested when asked what effect having these problematic genes could be having.

It proposed a process where immune dysregulation comes first, which in turn produces neuroinflammation, which impacts metabolic issues (steroids, fatty acids, and inflammation). The metabolic issues then amplify the other issues in a feedback loop, leaving the illness intact. AI Perplexity concluded, “This cyclical connectivity could help explain the persistence and complexity of ME/CFS symptoms.”

While it’s clear ME/CFS is much more than the genes we inherited, it’s fascinating to see three tenets of ME/CFS (immune dysregulation, neuroinflammation, metabolism) possibly replicated using just these 13 genes – another sign that DecodeME is on the right track.

Check out a pathway map created by AI Perplexity

Visual Pathway Map — ME/CFS Candidate Gene Network

┌────────────────────────┐
│   IMMUNE DYSREGULATION  │
│   (T-cell / innate)     │
└────────────────────────┘
▲
│
┌──────────────────┼────────────────────┐
│                  │                    │
TNFSF4           TNFSF18               IKZF2
(OX40L)         (GITRL)           (Treg transcription factor)
│ Activates     │ Balances Treg   │ Maintains immune tolerance
│ T-cells       │ & effector T’s  │ Prevents overactivation
│               │                 │
TRIM38          RABGAP1L          CSE1L
│ Controls      │ Immune receptor │ Cytokine release
│ innate        │ trafficking     │ BBB permeability
│ inflammation  │ & antiviral     │
BTN2A2          └─────────────────┘
│ Mucosal
│ immunity
└──────────────────────────────────────────────────────────────────────────
│
▼
┌───────────────────────────────────────┐
│ NEURO–IMMUNE COMMUNICATION             │
│ (Neuroinflammation, synaptic function) │
└───────────────────────────────────────┘
▲
│
CA10 – Synapse formation, pain circuitry
ARFGEF2 – Receptor trafficking (GABA, TNFR1)
SUDS3 – Microglial inflammation brake
CSE1L – Links immune signals & CNS barriers
└──────────────────────────────────────────────────────────────────────────
│
▼
┌─────────────────────────────────────────┐
│ METABOLIC CONTROL OF INFLAMMATION       │
│ (Bioactive lipid & xenobiotic metabolism)│
└─────────────────────────────────────────┘
▲
│
CYP2C8 – Fatty acid mediator metabolism
CYP2C9 – Eicosanoid / drug metabolism
CYP2C19 – Similar to above, affects hormone &
neuroactive compound metabolism

(based on DecodeME GWAS MAGMA analysis)

A Brain Disease?

This next section almost got my hair standing on end. Next, the researchers asked what tissues in the body these 13 dysregulated genes were particularly active in; in other words, what parts of the body were likely being most impacted by them.

There was no contest. Of the 54 tissues assessed, the 13 tissues that met statistical significance were all found in the brain. Immune and muscle tissues did not make the cut. The brain regions – many of which may be familiar – include:

• frontal cortex (executive functioning, attention problems),
• basal ganglia (nucleus accumbens) / basal ganglia (dopamine, problems with movement, reward and motivation/fatigue),
• amygdala and hippocampus (emotional reactivity, anxiety),
• cerebellar hemisphere (poor coordination),
• hypothalamus and cerebellum (autonomic, hormonal and metabolic problems),
• basal ganglia (putamen),
• spinal cord (movement and sensory problems),
• substantia nigra (connected to basal ganglia).

Note that Jarred Younger’s preliminary neuroinflammation findings highlighted similar brain regions, including the frontal cortex, amygdala, hippocampus, hypothalamus, and basal ganglia. As noted in a recent blog on Jarred Younger’s recent ME/CFS PET scan findings, dysregulations in these brain regions seem made to order to keep a person in bed and keep them from participating in life.

When asked what symptoms issues in these brain areas might produce, ChatGPT suggested it could cause “a combination of motor control problems, cognitive impairment, emotional dysregulation, reward-processing issues, and hormonal/autonomic disturbances”; i.e., something very similar to what we see in ME/CFS.

Again, it’s remarkable to see that even though the common gene variants appear to contribute only about 10% to the risk of developing ME/CFS, they still seem to recapitulate what we know about this disease.

These are signs that ME/CFS research field is building a biological foundation of this disease. What we want to see are similar findings being generated from different methodologies, and DecodeME is providing that.

Misses

With far more females than males having ME/CFS, ME/CFS is clearly a gender-based disease, yet the GWAS provided no clues as to why this is so. The answer to that rather important question may lie in further tests, or in larger, genome-wide studies which assess novel or rare variants.

The authors noted they have yet to test for the association to chromosome X (or Y) variants. (For some reason, most “GWAS” studies do not assess the X and Y chromosomes – and so are not actually “genome-wide” studies (!). The poor little, steadily disappearing Y chromosome found on males is hardly ever assessed. Further studies of the chromosomes would assess whether genetic differences found on them contribute to gender disparities found in ME/CFS.

Moving Forward

“We’re providing the X’s that mark a treasure. What’s now required is a bunch of scientists to go and dig, and find those treasures chests and bust open the treasure”  Chris Ponting talking to David Tuller

This DecodeME study is the first really full-throated shot at the genetics underlying ME/CFS – and it appears it’s just the beginning. It will winnow down its top gene targets to provide more precise targeting, it will study the X chromosome, dig into its HLA findings, compare the genetic findings in ME/FS with other diseases, and expand its analyses to the approximately 10,000 ME/CFS patients in its database with diverse genetic backgrounds.

Plus, it will embark on a 5,000-person US-based GWAS study involving Solve M.E., and I believe, Ian Lipkin, and then merge those results with the UK cohort to increase the predictive power of the study. Liz Worthey is working on an assessment of the rare/novel genes in ME/CFS, and hopefully, a DecodeME genome-wide study is in the works.

Lastly, the ME/CFS Roadmap initiative at the NIH reportedly spurred the creation of an ME/CFS genetics group. Decode ME’s willingness to give outside researchers access to its data provides another boatload of data for enterprising researchers to assess.

Conclusion

The first really big genetic test of ME/CFS (“is it a real disease or not?”) was a success. While the heritability was “modest”, it appeared to fit other diseases that co-occur with ME/CFS. The genes we are born with did help set the stage for this disease.

Importantly, the “genetic architecture” of ME/CFS DecodeME uncovered comports with what we know about the disease. DecodeME’s ability to identify the same brain regions at risk, genetically, that brain imaging studies have indicated are disturbed in ME/CFS, was striking.

The 29 top-tier genes DecodeME identified, which may play a causal role in ME/CFS, provide researchers with new avenues of study.

More is to come. Ponting said the results “begin to explain the disease’s heritable component”. Larger studies will amplify this study’s findings, plus this study’s strong findings make genome-wide studies that track rare and novel variants an even bigger necessity. We will undoubtedly learn much from them.

Check out David Tuller’s Interview with lead author Chris Ponting

Check out Jarred Younger’s early take on the study.

 Keep the Information Flowing!

Support Health Rising

: HEALTH RISING IS NOT A 501 (c) 3 NON-PROFIT