Preprint Initial findings from the DecodeME genome-wide association study of myalgic encephalomyelitis/chronic fatigue syndrome, 2025, DecodeMe Collaboration

[forestglip]

When I previously ran LDSC to test for genetic correlations between DecodeME and all the traits in the UK Biobank, the correlation with lupus didn't work, likely because the sample size in the Biobank was too small.

Recent discussions about lupus made me want to see if I could get it working with a different dataset. I found a larger SLE study, Bentham 2015 [1], and the summary statistics for this study are downloadable from a couple locations, like Open GWAS. This website says the study includes 5201 cases and 9066 controls*.

I ran LDSC between the DecodeME data and the SLE data, again using Bigagwas, and here are the results:

Heritability of phenotype 1 [ME/CFS]
---------------------------
Total Liability scale h2: 0.0885 (0.0058)
Lambda GC: 1.1019
Mean Chi^2: 1.1428
Intercept: 0.9089 (0.0075)
Ratio < 0 (usually indicates GC correction).

Heritability of phenotype 2/2 [Systemic lupus erythematosus]
-----------------------------
Total Liability scale h2: 0.1789 (0.022)
Lambda GC: 1.4673
Mean Chi^2: 1.3551
Intercept: 1.2021 (0.0093)
Ratio: 0.5692 (0.0262)

Genetic Covariance
------------------
Total Liability scale gencov: 0.0334 (0.0071)
Mean z1*z2: 0.0341
Intercept: -0.0212 (0.0062)

Genetic Correlation
-------------------
Genetic Correlation: 0.2658 (0.0535)
Z-score: 4.9637
P: 6.9154e-07

So genetic correlation is 0.27, which is small but not nothing, and the correlation is significant at p=6.9e-7.

I thought it might be interesting to look at some of the significant regions to see how they compare between the two studies.

Here is that RABGAP1L region. SLE is in blue and ME/CFS is in red.


chr6p22.2:


Not much else is very significant for SLE around the regions of the rest of the top 8 ME/CFS loci.

Spoiler: * Note about sample size

I think the numbers given for sample size on Open GWAS might be incorrect. When looking at the paper, the numbers given match with the sample size for the meta-analysis which included the main GWAS plus the Hom et al data.

But the p-values in the summary statistics seem to match the results from only the main GWAS when looking at Supplementary Table 3a. This would mean the sample size would be 4036 SLE cases and 6959 controls.

I used the larger sample size given by Open GWAS for LDSC, but this might not give exactly the right results if it should actually be the smaller sample size, which is about 25% smaller.

1. Bentham, James et al. “Genetic association analyses implicate aberrant regulation of innate and adaptive immunity genes in the pathogenesis of systemic lupus erythematosus.” Nature genetics vol. 47,12 (2015): 1457-1464. doi:10.1038/ng.3434 https://pmc.ncbi.nlm.nih.gov/articles/PMC4668589/

 

[Jonathan Edwards]

I would not necessarily expect any overlap in risk genes between ME/CFS and lupus, other than of course two doses of TLR7 and maybe other X genes.

The RABGAP1L patterns look mutually exclusive, which might be interesting?

 

[forestglip]

The RABGAP1L patterns look mutually exclusive, which might be interesting?

There's a lot of genes in the area. It might be that the loci relate to totally different genes and their proximity is a coincidence.

Though DNA is a very, very long thing, so it kind of seems unlikely to me to happen to be right next to each other by chance. But possible.

 

[Jonathan Edwards]

There's a lot of genes in the area. It might be that the loci relate to totally different genes and their proximity is a coincidence.

It might, or it might perhaps indicate two quite different influences from different linked SNP casettes on the same gene?

 

[forestglip]

It might, or it might perhaps indicate two quite different influences from different linked SNP casettes on the same gene?

It might, and it'd probably be very nice for both SLE and ME/CFS research to find a common pathway/gene.

 

[Jonathan Edwards]

Keep at it then!!

 

[forestglip]

The lupus locus on chromosome 1 above is right above TNFSF4 (AKA OX-40 ligand). There are a lot of papers talking about the connection between this gene and lupus, for example:

Polymorphism at the TNF superfamily gene TNFSF4 confers susceptibility to systemic lupus erythematosus (2013)

Spoiler: Abstract

Systemic lupus erythematosus (SLE) is a multisystem complex autoimmune disease of uncertain etiology (OMIM 152700). Over recent years a genetic component to SLE susceptibility has been established1–3. Recent successes with association studies in SLE have identified genes including IRF5 (refs. 4,5) and FCGR3B6.

Two tumor necrosis factor (TNF) superfamily members located within intervals showing genetic linkage with SLE are TNFSF4 (also known as OX40L; 1q25), which is expressed on activated antigen-presenting cells (APCs)7,8 and vascular endothelial cells9, and also its unique receptor, TNFRSF4 (also known as OX40; 1p36), which is primarily expressed on activated CD4+ T cells10.

TNFSF4 produces a potent co-stimulatory signal for activated CD4+ T cells after engagement of TNFRSF4 (ref. 11).

Using both a family-based and a case-control study design, we show that the upstream region of TNFSF4 contains a single risk haplotype for SLE, which is correlated with increased expression of both cell-surface TNFSF4 and the TNFSF4 transcript.

We hypothesize that increased expression of TNFSF4 predisposes to SLE either by quantitatively augmenting T cell–APC interaction or by influencing the functional consequences of T cell activation via TNFRSF4.

Web | PDF | Nature Genetics | Open Access

Blockade of OX40/OX40L signaling using anti-OX40L alleviates murine lupus nephritis (2024)

Spoiler: Abstract

Genetic variants of the OX40 ligand (OX40L) locus are associated with the risk of systemic lupus erythematosus (SLE), it is unclear how the OX40L blockade delays the lupus phenotype. Therefore, we examined the effects of an anti-OX40L antibody in MRL/Lpr mice. Next, we investigated the effect of anti-OX40L on immunosuppression in keyhole limpet hemocyanin-immunized C57BL/6J mice. In vitro treatment of anti-OX40L in CD4+ T and B220+ B cells was used to explore the role of OX40L in the pathogenesis of SLE.

Anti-OX40L alleviated murine lupus nephritis, accompanied by decreased production of anti-dsDNA and proteinuria, as well as lower frequencies of splenic T helper (Th) 1 and T-follicular helper cells (Tfh).

In keyhole limpet hemocyanin-immunized mice, decreased levels of immunoglobulins and plasmablasts were observed in the anti-OX40L group. Anti-OX40L reduced the number and area of germinal centers. Compared with the control IgG group, anti-OX40L downregulated CD4+ T-cell differentiation into Th1 and Tfh cells and upregulated CD4+ T-cell differentiation into regulatory T cells in vitro.

Furthermore, anti-OX40L inhibited toll-like receptor 7-mediated differentiation of antibody-secreting cells and antibody production through the regulation of the SPIB-BLIMP1-XBP1 axis in B cells.

These results suggest that OX40L is a promising therapeutic target for SLE.

Web | PDF | Eur J Immunol. | Open Access

Edit: Aha! It was mentioned in the DecodeME candidate genes document as a potential gene that this locus applies to (thanks Evergreen for posting about this):

TNFSF4 (Tier 1)

• Protein: Tumor necrosis factor ligand superfamily member 4, also known as CD252 or OX-40L. UniProt. GeneCards. The allele that increases the risk of ME/CFS is associated with decreasing TNFSF4 gene expression.​

​

• Molecular function: Cytokine that promotes T-cell proliferation, differentiation, and survival (12). Expressed on CD4+ T cells, CD8+ T cells and vascular endothelial cells.​

​

• Cellular function: Regulates T-cell-mediated immunity.​

​

• Link to disease: Region upstream of TNFSF4 is associated with systemic lupus erythematosus risk (13).​

​

• Potential relevance to ME/CFS: Potential dysregulation of T-cell mediated immunity.​

The DecodeME paper says that the ME/CFS variants here are associated with decreased expression of TNFSF4 in the lung, skin of sun exposed lower leg, and thyroid.

 

[ChronicallyOverIt]

Blockade of OX40/OX40L signaling using anti-OX40L alleviates murine lupus nephritis (2024)

A few OX40L monoclonals are coming to market soon:

Press Release: Sanofi’s amlitelimab met all primary and key secondary endpoints in the COAST 1 phase 3 study in adults and adolescents with atopic dermatitis

Sanofi’s amlitelimab met all primary and key secondary endpoints in the COAST 1 phase 3 study in adults and adolescents with atopic dermatitis Amlitelimab, dosed every four weeks or every 12 weeks,...

www.sanofi.com

"normalize the overactive immune system, without depleting T cells"

Originally for dermatitis, but I believe they think they will be good general immune T-cell suppression, possibly off-label

 

[jnmaciuch]

Thanks for this analysis @forestglip ! I think this is incredibly useful precisely because it's a comparison to what a GWAS would look like in an illness where we know TNF and BTNs are relevant in the pathophysiology (and can therefore be much more confident that the associations in the SLE cohort are driven by those genes).

It's interesting that you can see a similar indication of LD in the entire RABGAP1L region that we were discussing in another thread--the little island of blue pretty much confirms that the whole region tends to get inherited together. But there was an open question as to whether the little stretch right near DARS2 is being confounded by LD with RABGAP1L. I can't say this with complete confidence given the small cohort, but seeing the absence of an "LD island" around DARS2 in SLE makes me more confident that the relevant genes in that region for ME/CFS are DARS2 (or its immediately adjacent genes) and RABGAP1L separately (weakly).

Similarly, there was an open question of whether the hits on Chr6 were in relevant regulatory regions for BTNs. From the blue dots, we get a clear picture of exactly where the strongest regulatory regions are upstream of the BTNs, and it's not where the [edit: strongest] of the associated SNPs in ME/CFS are lying. And, even more interestingly, the region that is most highly associated with ME/CFS is almost entirely devoid of hits in the SLE cohort. Being cautious again about overinterpreting, I'd say this is one piece of evidence showing that it really is the histone proteins in that region that are relevant to ME/CFS.

 

[Jonathan Edwards]

Nice work.

 

[Andy]

Would it be worthwhile in trying to get those researchers who are already looking at the genetics of lupus (or other similar conditions) interested in the results of DecodeME?

 

[Jonathan Edwards]

Would it be worthwhile in trying to get those researchers who are already looking at the genetics of lupus (or other similar conditions) interested in the results of DecodeME?

Yes. The King's people are obviously interested. They also have an interest in pain/fibromyalgia genetics. I don't have contacts ith them but this should be discussed on Nov 6th.

 

[V.R.T.]

Blockade of OX40/OX40L signaling using anti-OX40L alleviates murine lupus nephritis (2024)
Genetic variants of the OX40 ligand (OX40L) locus are associated with the risk of systemic lupus erythematosus (SLE), it is unclear how the OX40L blockade delays the lupus phenotype. Therefore, we examined the effects of an anti-OX40L antibody in MRL/Lpr mice. Next, we investigated the effect of anti-OX40L on immunosuppression in keyhole limpet hemocyanin-immunized C57BL/6J mice. In vitro treatment of anti-OX40L in CD4+ T and B220+ B cells was used to explore the role of OX40L in the pathogenesis of SLE.

Anti-OX40L alleviated murine lupus nephritis, accompanied by decreased production of anti-dsDNA and proteinuria, as well as lower frequencies of splenic T helper (Th) 1 and T-follicular helper cells (Tfh).

In keyhole limpet hemocyanin-immunized mice, decreased levels of immunoglobulins and plasmablasts were observed in the anti-OX40L group. Anti-OX40L reduced the number and area of germinal centers. Compared with the control IgG group, anti-OX40L downregulated CD4+ T-cell differentiation into Th1 and Tfh cells and upregulated CD4+ T-cell differentiation into regulatory T cells in vitro.

Furthermore, anti-OX40L inhibited toll-like receptor 7-mediated differentiation of antibody-secreting cells and antibody production through the regulation of the SPIB-BLIMP1-XBP1 axis in B cells.

These results suggest that OX40L is a promising therapeutic target for SLE. Web | PDF | Eur J Immunol. | Open Access

Edit: Aha! It was mentioned in the DecodeME candidate genes document as a potential gene that this locus applies to (thanks Evergreen for posting about this):

In her recent talk Michelle James lists OX40 as a PET tracer that is being investigated in ME. In my email I asked if it was her or someone else studying it but I haven't heard back yet.

I know I've been a broken record about these tracers lately but it's an interesting connection!

 

[V.R.T.]

Would it be worthwhile in trying to get those researchers who are already looking at the genetics of lupus (or other similar conditions) interested in the results of DecodeME?

Yes if anyone knows of any other lupus researchers (than the kings people) who are any good, perhaps letting them know about the Nov. 6 event might be an idea.

 

[Turtle]

In her recent talk Michelle James lists OX40 as a PET tracer that is being investigated in ME. In my email I asked if it was her or someone else studying it but I haven't heard back yet.

I know I've been a broken record about these tracers lately but it's an interesting connection!

As long as there are interesting connections there is no broken record as far as I can see. Keep them coming! Others might make a connection, somehow, somewhere.

 

[wigglethemouse]

Impressive work @forestglip. Can you look to see if you can find any datasets for Sjogrens. It's a CNS disease with some overlap to ME/CFS such as small fiber neuropathy.

Here is that RABGAP1L region. SLE is in blue and ME/CFS is in red.

This Sjogrens GWAS meta-analysis abstract also mentions RABGAP1L.

Meta-Analysis of GWAS data from 10,003 Sjögren’s Disease Cases Identifies Thirteen Sjögren’s Risk Loci. - ACR Meeting Abstracts

Background/Purpose: Sjögren’s disease (SjD) is a systemic autoimmune condition with a complex genetic architecture. To date, 22 genome-wide significant (GWS) SjD risk loci have been identified, while >100 risk loci have been described in other related autoimmune diseases (Khatri et al. 2022)...

acrabstracts.org

 

[forestglip]

Can you look to see if you can find any datasets for Sjogrens.

This study looks promising and has downloadable data: https://www.ebi.ac.uk/gwas/studies/GCST012796

Though it's a relatively small sample (585 cases if using just European participants). But I'll try to test the correlation with ME/CFS using this.

 

[forestglip]

This study looks promising and has downloadable data: https://www.ebi.ac.uk/gwas/studies/GCST012796

Though it's a relatively small sample (585 cases if using just European participants). But I'll try to test the correlation with ME/CFS using this.

I'm running into an issue with the Sjogren's data because the dataset only has the effect allele for each variant and not the reference allele, but Bigagwas requires both. So I'm not sure I'll be able to test a genetic correlation with this dataset.

I did explore the plots, looking at the significant loci from each study to see if anything similar popped out. I used the data that included all ancestries combined (1405 cases and 4747 controls). I didn't really see anything very similar.

Unlike the plot for SLE, there's nothing going on in the TNFSF4/RABGAP1L area:


The chromosome 6 area is interesting because of how huge the locus is for Sjogren's compared to the little ME/CFS locus on the left. (I cut off a lot of gene names below the plot because there were too many to fit). This is with only 1405 Sjogren's cases. And the locus when looking at only the 585 European cases has nearly the same significance (p=~3e-34).




The closest I saw to a matching locus of the ones I looked at, but it's not genome-wide significant in either disease:


Link to paper.

 

[forestglip]

chr6p22.2:

From the blue dots, we get a clear picture of exactly where the strongest regulatory regions are upstream of the BTNs

Sorry, my plot was probably a bit misleading on this front since it was so zoomed in. It gives quite a different picture if I zoom out to include the full SLE locus (not all genes shown):

 

[V.R.T.]

A few OX40L monoclonals are coming to market soon:

Press Release: Sanofi’s amlitelimab met all primary and key secondary endpoints in the COAST 1 phase 3 study in adults and adolescents with atopic dermatitis

Sanofi’s amlitelimab met all primary and key secondary endpoints in the COAST 1 phase 3 study in adults and adolescents with atopic dermatitis Amlitelimab, dosed every four weeks or every 12 weeks,...

www.sanofi.com

"normalize the overactive immune system, without depleting T cells"

Originally for dermatitis, but I believe they think they will be good general immune T-cell suppression, possibly off-label

This particular monoclonal is made by Sanofi, who recently agreed to let Scheibenbogen trial their CD38 inhibitor in ME/CFS. So if there is a rationale for trying their OX40 monoclonal in ME they might well be amenable.