SMPDL3B a novel biomarker and therapeutic target in myalgic encephalomyelitis, 2025, Moreau, Fluge, Mella et al

[DMissa]

I just went through 7 different gene expression datasets that I have for different cell types from people with ME/CFS and nothing for these two genes stood out as clearly altered or consistent in direction of change from a quick scan. I was hoping to see something because they could fit well into the story I am trying to fit together. At least much of the theme is consistent. And probably most importantly this is yet again more evidence for messed up lipid homeostasis in immune cell populations. I cannot recall a null or even tepid result in this space.

 

[Hutan]

Sadly, I don't think the reported difference in SMPDL3B means much.

Figure 2a - Canadian cohorts. They say that the controls were matched, but the percentage of males is enormously different between the ME/CFS (41/249) and healthy controls (30/63). It looks as though much of the difference is females on contraceptive pills.

 

[Hutan]

This is for the females in the Canadian cohorts (the ME/CFS and healthy controls are lumped together). Look at the very big difference between the group not on contraception and the group that is. It's a much stronger difference than between the ME/CFS and HC individuals in the Canadian sample.



Damn. I'm a bit surprised that Fluge and Mella are associated with this.

 

[Hutan]

Controls were frequency-matched to the ME patient group based on age and sex distribution at the group level.

The results section says that controls were frequency-matched.

Clinical and demographic data of study participants
The clinical and demographic characteristics of the Canadian and Norwegian cohorts are summarized in Table 1. The study design is illustrated in Fig. 1. The sex distribution in both cohorts reflects the well-documented higher prevalence of ME in women, with females comprising 83% of the Canadian cohort and 84% of the Norwegian cohort. Given that sex is a crucial biological variable influencing immune responses and disease susceptibility, its consideration in ME is essential [15].

And they say that the high percentage of females in the ME/CFS cohorts reflects the higher prevalence of ME/CFS in women. Which is fine. But 33/63 controls is a female percentage of 52%. That's a problem, I think it's quite a major one. But, they don't even mention the percentage of females in the control cohort in the 'Clinical and demographic data of study participants'.

 

[Murph]

Sadly, I don't think the reported difference in SMPDL3B means much.

Figure 2a - Canadian cohorts. They say that the controls were matched, but the percentage of males is enormously different between the ME/CFS (41/249) and healthy controls (30/63). It looks as though much of the difference is females on contraceptive pills.

View attachment 26903

This chart should be data on the same cohort. A large majority of those on contraception scored over 100 and therefore must be in the ME group. If I'm reading this right, almost everyone they got in their study who was on contraception also had ME. (plenty of people with ME and not on contraception but few people on contraception in the control group.

Hard to be absolutely sure what's going on since the labels are here are screwy but I don't like it.

 

[Hoopoe]

SMPDL3b is related to CDP-Choline, which has been associated with ME/CFS and peroxisomal dysfunction.

 

[Hutan]

I thought I remembered something about this. Here's a link to a presentation on SMDPDL3B by Moreau in 2023.

 

[Jonathan Edwards]

The data do look problematic, particularly if the sex ratios are matched for 'frequency' in group rather than just matched.

Reading the abstract makes it sound important and it made me think of other ways female sex and oestrogen might be relevant. In the light of the data this may be a red herring but I wonder if the high female sex ratio in ME/CFS could reflect a stereochemical pattern in lipid rafts that is oestrogen dependent. Maybe there is a shift in a sphingolipid used both by oestrogen receptor docking and immune recognition docking that does not normally shift immune function between sexes but might just happen to cross-talk with something like an HLA ligand in the context of a TLR or something.

Maybe it is relevant to neural involvement too. We keep seeing hints that immune and nerve synapses may make use of similar gene products. Maybe in order to have the sort of sophisticated signalling that is dependent on cofactors that occurs for both systems you need accessory molecules like lipid raft components which have doubled up in function, having been 'borrowed' during evolution to serve a second system.

But maybe all that all our comments show is how easy it is to find an apparent link between things even if the prompt was a red herring.

I don't see this as having anything to do with rituximab or daratumumab.
I am also unimpressed that anyone should want to use drugs to alter levels of a protein without having some idea why the protein level is changed. The drug might make things much worse.

 

[Midnattsol]

Sphingomyelin concentrations in the blood (and cell membrane) can also be affected by diet, which could I guess also have downstream effects on the enzymes that can modify these lipids.

*Other lipids from the diet is inserted into the cell membrane up to a point, I don't remember a sphingomyelins and cell membrane study.

If there is a connection with sex hormones I think it will be difficult to replicate findings in female subjects

 

[Sasha]

Interesting! I don't know what to make of it, but will be interested to read what the resident immunologists/hamburger fans think.

The buns can't cope with this!

 

[forestglip]

It seems like some charts are missing from figure 2. The caption talks about fig 2a through 2n, but I only see 2a through 2g. And 2b through 2g in the image seem to correspond to 2i through 2n in the caption, with 2b through 2h missing.

Alain Moreau responded to an email:

Thank you very much for your careful review and for bringing these issues to our attention.

You are absolutely right—Figure 2 itself is correct, but the error lies in the legend, which unfortunately was not updated as instructed during the eProofing stage. I have already contacted the Editor and the publishing team at Springer to request the necessary corrections.

For your information, the details related to the panels removed from Figure 2 are now provided in Supplementary Table S1, as suggested by the reviewers.

We sincerely appreciate your attention to detail and your kind assistance in this matter. Please find attached the corrected legend for Figure 2.

Corrected legend:

Figure 2. Influence of covariates on plasma SMPDL3B levels and symptom severity in ME patients. (a) Plasma levels of SMPDL3B in ME patients (n = 249) and healthy controls (n = 63) from the Canadian cohort. (b) Plasma levels of SMPDL3B in different groups of patients with ME from the Norwegian cohort grouped by symptom severity (n = 16 for mild, n = 40 for mild/moderate, n = 41 for moderate, n = 22 moderate/severe and n = 21 severe). Disease severity was determined based on clinical assessments incorporating standardized and trial-specific questionnaires. (c) Plasma SMPDL3B levels in ME Canadian, ME Norwegian and healthy control female participants (n = 208, n = 119 and n = 33 respectively) and males (n = 41, n = 22 and n = 30 respectively). (d) Plasma concentrations of SMPDL3B in Canadian female ME participants across different age groups (n = 16 for 18–30 years, n = 97 for 31–50 for years, n = 95 for 51 years and up). (e) Plasma concentrations of SMPDL3B in Norwegian female ME participants across different age groups (n = 37 for 18–30 years, n = 65 31–50 for years, n = 16 for 51 years and up). (f) Plasma levels of SMPDL3B in Canadian female participants with or without oral contraceptive use (n = 176 and n = 32 respectively). (g) Plasma levels of SMPDL3B in Norwegian female participants with or without oral contraceptive use (n = 88 and n = 31 respectively). An unpaired T test was performed when comparing two groups. For comparisons between two groups, the Mann–Whitney U test was used. For comparisons involving more than two groups, the Kruskal–Wallis test was performed, followed by Dunn’s post hoc test with multiple comparison correction where appropriate. Differences were found to be significant at *P < 0.05, **P < 0.01, *** P-value < 0.001 and ****P < 0.0001

Edit: Note that I think he meant Supplementary Table S3. It has most of the items from the original legend, but not all, though.