Single-cell transcriptomics of the immune system in ME/CFS at baseline and following symptom provocation, 2022, Ahmed, Hanson et al.

Preprint

https://www.biorxiv.org/content/10.1101/2022.10.13.512091v1

SUMMARY
ME/CFS is a serious and poorly understood disease. To understand immune dysregulation in ME/CFS, we used single-cell RNA-seq (scRNA-seq) to examine immune cells in cohorts of patients and controls. Post-exertional malaise (PEM), an exacerbation of symptoms following strenuous exercise, is a characteristic symptom of ME/CFS. Thus, to detect changes coincident with PEM, we also performed scRNA-seq on the same cohorts following exercise. At baseline, ME/CFS patients displayed dysregulation of classical monocytes suggestive of inappropriate differentiation and migration to tissue. We were able to identify both diseased and more normal monocytes within patients, and the fraction of diseased cells correlated with metrics of disease severity. Comparing the transcriptome at baseline and post-exercise challenge, we discovered patterns indicative of improper platelet activation in patients, with minimal changes elsewhere in the immune system. Taken together, these data identify immunological defects present at baseline in patients and an additional layer of dysregulation following exercise.

Highlights ME/CFS is a debilitating disease with unknown causes. Here, we provide, for the first time, an extensive single cell resolution dataset detailing the gene expression programs of circulating immune cells of ME/CFS cases at baseline and after symptom provocation. We were able to detect robust dysregulation in certain immune cells from patients, with dysregulation of classical monocytes manifesting the strongest signal. Indeed, the fraction of aberrant monocytes in ME/CFS patients correlated with the degree of disease severity. Surprisingly, platelet transcriptomes were also altered in ME/CFS, and they were the only component of the immune system that showed large-scale changes following symptom provocation.

 

https://twitter.com/user/status/1585079174704033792

 

While I am concerned that I have not read details yet, and any such study will need follow up studies, I was actually clapping as I read the summary above. Its beyond time this kind of thing was done, though this says nothing about the quality of the study. I still have to read it. It also does not show, so far as I can tell, causation nor really how this fits with everything else. Its just an important step.

 

While I do not understand the technicalities of the work, the main findings in Figure 2 (panels A, B, E and F) do not seem to be suggestive of major differences between controls and patients, whether at baseline or post-CPET, especially with such a small sample size.


Caption:

 

To me the presentation of the paper is very wordy and speculative and I find it hard to get a grasp of what was really found.

My main gripe is that I do not know what they mean by dysregulated. Maybe they should show us clearly that something is different and mention briefly what they think that means in the discussion.

 

What you are seeing is data for Differential Expressed (DE) genes between patient and control per immune cell type cluster. Figure 2A is showing ~800 genes different at baseline (BL) and post CPET (PC) in a subset of monocytes between patients and controls.

That topic is covered in the results section - see the sections on "Monocyte dysregulation is heterogenous within and between ME/CFS patients" and "An abnormal platelet state is coincident with PEM in ME/CFS". I really do think this paper is worthy of your time @Jonathan Edwards and I welcome some informed opinion from you, and maybe speculation of how this could tie into the endothelial dysfunction findings by multiple groups.

I believe the scope of the project was a "fishing expedition" by using single cell RNA-seq of immune cells as a whole. Now they have identified two subsets (monocytes and platelets) they can do a deeper dive to get a better understanding of these and things like macrophages that will have matured from monocytes.

 

No it isn't. Why is this 'dysregulation' rather than just different numbers - which might indicate a very normal response to some different stimulus?

I haven't ever seen anything very interesting arise from this sort of shotgun fishing approach. In the old days we looked to see if there was a single factor different enough and interesting enough to provide a useful lead. Bunches of things going this way or that could just be an effect of sleep schedule or activity pattern or painkiller usage. I also find the way things are presented in highly filtered ways confusing. I find it much easier to lookout raw results and then make the relevant comparisons.

 

Comments

1. Overview

This is a small sample of only 30 patients (I'm pretty sure they took blood samples from 90 people with ME).

Two findings stand out: for monocytes and platelets

At baseline, they found big differences between patients and controls for monocytes. (They found smaller changes for some adaptive immunity cells: some B, T subtypes and even NK cells)

After exercise, the only change seen relative to baseline were for platelets, though even that was modest. And surprisingly, after exercise patient platelets looked ‘normal,’ i.e. much more like control platelets. But BEFORE exercise patient platelets had low levels of activation. It appeared that exercise ‘restored’ the normal status of patient platelet populations.

It's hard to square this 'normalisation' of the platelet population with platelets playing a causal role in PEM.

No changes in proportions of cell types

They found that, unlike with other studies, there were no significant (and meaningful) changes in the relative proportions of different immune cell types in patients relative to controls or after exercise.

One general change: ME immune cells are less ready to make proteins

One interesting finding across cell types was the expression of translational machinery (the bits that make proteins) is reduced in ME relative to controls for many different cell types. The authors suggested this meant the ME immune cells were ’quiescent’, or less active.

to follow if migraines permit:
2. Monocytes
3. Platelets

 

2. The big story: gene expression indicates "primed" monocytes in ME

Monocytes are large immune cells, and their role is to migrate to where they are needed and become macrophages. Macrophages engulf and neutralise invading pathogens like bacteria, viruses and even yeast. They can also Hoover up debris like dead cells and bits of cells. This applies to classical monocytes that make up 80%-95% of all monocytes.

https://www.youtube.com/watch?v=jTs-q76XQuU

A bias to activation?

The stand-out finding of the paper is abnormalities in classical monocytes – with a notable increase (relative to controls) of genes associated with:
1. Signaling (mostly involving chemokines ( molecules that act a bit like scents used to guide migration of monocytes), but also cytokines
2. Migration (to inflamed tissues).
3. Activation to become macrophages

The authors suggest this shows that patient monocytes are biased towards migration to tissues and maturation there into macrophages.

@Jonathan Edwards Dysfynction might not be the right word to describe these changes, but were you saying these changes are not meaningful?

Broadly speaking, patients had a mixture of 2 types of monocyte, 'normal' and 'diseased'.

The degree of monocyte abnormality correlates with several disease severity measures
For most patients, predicted-Diseased cells made up 25%-75% of classical monocytes, with the balance as predicted-Normal. The proportion of predicated-Diseased monocytes correlated with MFI (fatigue) scores, but also SF – 36 physical components score and PEM severity.

 

3. Platelets: slightly off at baseline, normal after exercise

Platelets are plate-shaped cells whose primary role is to plug blood vessels that are punctured to promote clot formation.

The authors looked at the impact of exercise on the expression of individual genes (their main method), and they found no difference. They then switched to Gene Set Enrichment Analysis, GSEA. This looks at changes in the expression of related sets of genes. This can pick up more subtle differences.

With GSEA they found exercise changed expression levels for platelets (and a few other cell types to a lower level).

At baseline, gene sets related to platelet activation were underexpressed. In other words, they were on the quiet side. After exercise, they looked like control platelets (exercise made no difference to control platelets).

Note that the platelet population measured at baseline are probably different (some platelets dying and being replaced) from that after exercise.

I find it hard to relate this 'normalising' of the platelet population after exercise to post-exertional malaise.

A link between monocytes and platelets? Signalling and inflammation

This section is interesting but more speculative.

The study tried to understand what signals were flying around by analysing the expression of monocyte-related signalling molecules (and receptor) genes using a program called CellChat.

They found increased monocyte signalling to certain T and Natural Killer cells. They also found increased signalling to platelets. The authors noted that monocyte-platelet interactions had been demonstrated to a strong marker of inflammation. (Note this is an inference from analysing gene expression - the study did not show that there was monocyte-platelet interaction leading to inflammation.)

 

I find it very difficult to assess results like this.
The results suggest maturation changes. It is very hard to know what that means for monocytes in blood. Monocytes in blood are basically cells not doing anything but going around ready to be called into use. In the normal situation a few are probably going through maturation from low level signals over time and then when they are a bit too mature getting cleared away by spleen or liver. Changes in maturation markers could easily occur with different levels of exercise or other irrelevant factors.

If these differences genuinely reflect monocyte activation in blood then I would expect some standard markers to show something - CD16 expression maybe, or even just cytoplasm to nucleus volume ratio. I would lookout things like that first rather than using complicated mic techniques if I wanted to look at maturation or activation. If the mics showed something other than that - something not just suggesting maturation or activation but engagement of a specific pathway suggesting a specific signal from an unexpected cytokine then that would be very interesting.

Its a bit like doing detailed biochemical analysis on feathers to show that you have blue tits rathe than robins - there are more obvious ways of doing that!

I am afraid I get very irritated by histograms of 'numbers of genes' for that reason. We know enough about gene products to make an intelligent analysis of specific ones. Numbers is not interesting to me.

 

Thanks

To be fair to the authors, they used an omics approach as a broad sweep (given how little we know), as a way of identifying areas to worth pursuing. The discussion section says:
"our analysis discovered upregulation of chemokine/cytokine pathway genes in patientderived
monocytes as well as a clear correlation between the proportion of predicted-diseased
monocytes and multiple metrics of disease severity. Future studies investigating macrophages
from tissue biopsies isolated from ME/CFS patients will be important"
Let's hope this happens.

 

It is unclear to me what tissues could be systematically biopsied. Minimum skin biopsies leave marks (I still have them thirty year later). Muscle biopsies are probably worse. And when tissues have been looked at no pathology was found. If studies of macrophage gene expression in biopsies with no structural pathology showed some differences I think we would have to conclude that the differences didn't do much.

I fear that the approach here is 'oh maybe it's inflammation so lets look for some inflammation (even though we already know there isn't any). My experience of this sort of research is that you need to look for something specific that you did not expect to find. From that point of view doing omics has the advantage of being able to pick out specificities amongst things that do not change but I don't see any interpretation n those terms - just that they found some stuff consistent with inflammation patterns or maybe just maturation.

 

The study has now been published in Cell Reports Medicine as
Single-cell transcriptomics of the immune system in ME/CFS at baseline and following symptom provocation.