Skewing of the B cell receptor repertoire in myalgic encephalomyelitis/chronic fatigue syndrome, 2021, Sato et al

[Andy]

Highlights

• ME/CFS is characterized by skewed B cell receptor gene usage.
• Upregulation of specific IGHV genes correlated to infection-related episodes at onset.
• Plasmablasts of ME/CFS patients upregulated interferon response genes.
• B cell receptor repertoire analysis can provide a useful diagnostic marker in ME/CFS.

Abstract
Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is a debilitating condition characterized by fatigue and post-exertional malaise, accompanied by various signs of neurological and autonomic dysfunction. ME/CFS is often triggered by an infectious episode and associated with an aberrant immune system.

Here we report that ME/CFS is a disorder characterized by skewed B cell receptor gene usage. By applying a next-generation sequencing to determine the clone-based IGHV/IGHD/IGHJ repertoires, we revealed a biased usage of several IGHV genes in peripheral blood B cells from ME/CFS patients. Results of receiver operating characteristic (ROC) analysis further indicated a possibility of distinguishing patients from healthy controls, based on the skewed B cell repertoire. Meanwhile, B cell clones using IGHV3-30 and IGHV3-30-3 genes were more frequent in patients with an obvious infection-related episode at onset, and correlated to expression levels of interferon response genes in plasmablasts. Collectively, these results imply that B cell responses in ME/CFS are directed against an infectious agents or priming antigens induced before disease onset.

Open access, https://www.sciencedirect.com/science/article/pii/S0889159121001537

 

[Jaybee00]

Hopefully @Jonathan Edwards could comment on this study.

 

[wigglethemouse]

I imagine this is of interest to @Chris Ponting (@Simon M ) as he has talked about having a researcher at Oxford look at B cell receptor sequencing in ME. He might find it interesting to compare results vs disease length and severity like in this study.

 

[Creekside]

I wonder whether this only applies to a subset of PWME. That would need a large study.

 

[Hutan]

A Japanese team

Wakiro Sato ab
Hirohiko Ono a
Takaji Matsutani c
Masakazu Nakamura a
Isu Shin d
Keiko Amano e
Ryuji Suzuki c
Takashi Yamamura ab

a Department of Immunology, National Institute of Neuroscience, National Center of Neurology and Psychiatry (NCNP), 4-1-1 Ogawahigashi, Kodaira, Tokyo 187-8502, Japan
b Multiple Sclerosis Center, National Center Hospital, NCNP, Tokyo, Japan
c Repertoire Genesis Incorporation, Ibaraki, Japan
d Sekimachi Medical Clinic, Tokyo, Japan
e Seifuso Hospital, Saitama, Japan

Study sponsorship or funding:
Health and Labor Sciences Research Grants for Comprehensive Research on Persons with Disabilities from the Japan Agency for Medical Research and development, AMED of Japan
Intramural Research Grant (28-5, 1-5) for Neurological and Psychiatric Disorders of National Center of Neurology and Psychiatry

 

[Hutan]

Each B cell expresses a single B cell receptor (BCR), and the diverse BCR expressed by an individual’s total B cell population is termed the “BCR repertoire.” BCR repertoire skewing is correlated with microbial interactions, vaccinations, or immune-mediated diseases, such as systemic lupus erythematosus (Hershberg and Prak, 2015) or multiple sclerosis (Baranzini et al., 1999). Next-generation sequencing (NGS) technology has enabled the comprehensive quantitative analysis of BCRs at the clonal level.

So, I think it's like the t-cell clonal expansion work. They looked at the B cell receptors in people with ME/CFS and healthy controls.

The ME/CFS people met Fukuda, CCC and ICC requirements.

Cohort 1 comprised patients recruited to NCNP hospital from October 2015 to December 2016. B cell subset analysis was performed using the peripheral blood of cohort 1 (ME/CFS n=61, HC n=23; Table 1). BCR repertoire analysis was performed using blood of randomly selected patients from cohort 1 (ME/CFS n=38, HC n=23; Table 2).

So the first analysis was on 38 people with ME/CFS and 23 healthy controls, recruited in 2016.

To validate the results of BCR repertoire analysis, new patients with ME/CFS were recruited from NCNP hospital from January 2017 to November 2019, which comprised cohort 2 (ME/CFS n=38; Table 3),

A validation sample was done later, with 38 more people with ME/CFS.

 

[Grigor]

The validation bit is very cool.

Are they saying that vaccination could somehow trigger ME?

 

[Jonathan Edwards]

A validation sample was done later, with 38 more people with ME/CFS.

Do we actually get to see the results?
I may be old fashioned but this sort of paper looks to me like presenting masses of information using obscure statistical methods to distract from the fact that nothing much was found.

I can only see one set of IGHV expression levels in figure 1. The panaormam is underwhelming with everything looking about the same. Some tiny spikes in the middle are statistically significant it seems but in immunology we tend to be interested in sore thumbs rather than statistics!

If IGHV30 was five times higher and a major player then yes, interesting, but the prominent signals here look pretty much the same in ME and controls. I have asked Jo Cambridge to look. She knows much more about the IGHV story having worked on IGHV expression for years.

 

[Hutan]

I haven't looked into the data yet, but I think the study deserves to be read.

From a quick skim read, they didn't find any evidence of differences in clonal expansion in terms of how many receptor types account for 50% of the variation. But they did find differences in the expression of some particular genes.

we showed the upregulation of several IGHV, IGHD, and IGHJ genes in the B cells of ME/CFS patients compared with those of healthy subjects. Furthermore, the skewing of the specific BCR repertoire was nicely replicated in cohort 2.

Previous reports suggested that IGHV3-30//3-30-3 genes are utilized in various infections including malaria, influenza and COVID-19 (Galson et al., 2020; Tan 340et al., 2018; Waltari et al., 2019). Because diverse infectious agents include vast kinds of antigens, it may not be a specific common antigenic motif derived from pathogens which stimulate B cell clones with IGHV3-30 or 3-30-3 genes. Various infections may trigger a situation where some common motif (possibly derived from self or from commensal bacteria) induce these B cell clones. It might be worthwhile to investigate what kind of antigens affect the B cells derived from patients with upregulated IGHV3-30/3-30-3.

The four differential IGH genes (IGHV1-3, 3-49, IGHD1-26, IGHJ6) did not show a relation to infection-related episode. The skewing of BCR repertoire occurs not only in various infections but also in autoimmune diseases (Hershberg and Prak, 2015). Therefore, it is possible that some common mechanisms may lead to the skewing of BCR repertoire in ME/CFS and in autoimmune diseases.

They noted that another researcher using another technique had identified the same genes as of interest in ME/CFS.

Recently, an association between IGHV3-23/30 and ME/CFS has been shown using a plasma proteomic approach (Milivojevic et al., 2020). Despite differences in methodology, the fact that the expression of the same IGHV region was significantly increased in ME/CFS patients provides further evidence of the importance of IGHV3-30.

They suggested expression of these genes could usefully distinguish ME/CFS from healthy controls.

ROC analysis of IGH gene expression that was conducted with two cohorts revealed that the AUC is ~0.9 with high sensitivity and specificity to separate ME/CFS patients from HCs. Although these results are still preliminary, this method has the potential to be developed as a diagnostic biomarker.

We have shown that the frequency of B cells in patients is significantly higher than those with HCs. Previous studies (Bradley et al., 2013; Ramos et al., 2015; Mensah et al. 2016) did not find significant differences between patients with ME/CFS and controls. However, Klimas et al. (1990) reported that B cells are increased in the peripheral blood of patients. These discrepancies might be due to different patient populations and/or different methodology.

This finding of increased B cells in the ME/CFS cohort is interesting (I haven't looked to see how robust the finding is), raising questions of whether this particular sample is different to other ME/CFS populations, and if this has affected the finding of gene expression.

Another potential HC would include those with an antecedent infection within a time frame similar to that of the patients with ME/CFS.

They note that what they are finding might just be a result of the ME/CFS patients having had an infection (the triggering one) and the healthy controls not having had such an infection. Maybe what they have found might possibly just be a result of ME/CFS patients not being able to control latent infections e.g. herpes simplex, that is, a downstream effect?

BCR repertoire analysis could be developed as a powerful tool to help diagnose ME/CFS. It may also be used to predict the efficacy of B cell-targeted therapy, which is effective in a subgroup of patients.

The company that did the BCR repertoire analysis is a private company. And I'm not sure where they have got the idea that B cell-target therapy is effective in a subgroup of patients from.

 

[Hutan]

Do we actually get to see the results?

Yeah, this was an issue. I'm assuming the data is in the supplementary tables. Perhaps that indicates a problem, when the data is not front and centre.

 

[Jonathan Edwards]

Huh, I just got an email from Jo C which says:

I reviewed it...got them to do major revision and now ok actually. I have been asked to write a commentary on it by the editor...

So we will see. She does not seem to be leaping out of her chair just yet!

 

[Grigor]

The paper also references to a paper by one of her PhD students Fane Mensah. He's Dutch.

https://solvecfs.org/jo-cambridge/

 

[Jonathan Edwards]

I haven't looked into the data yet, but I think the study deserves to be read.

Yes, I think this is one to take seriously even if it gives me a headache.

The company that did the BCR repertoire analysis is a private company. And I'm not sure where they have got the idea that B cell-target therapy is effective in a subgroup of patients from.

Even if there is a real prominence of IGHV30 I think the connection from there to any understanding of what might be going on is murky in the extreme. The most obvious idea is that it points to a particular inciting antigen but once one has gone through all the reasons why that might not be so easy to interpret in ME where there seem tone lots of inciting infections and remembered that at best these are statistical weightings for a few known infections it becomes hard work.

I suppose what puts me off papers like this most is the way the text seems to want to sell something and bring in every tiny bit of data that might support and forget all the caveats. In reality we can be pretty sure that B cells are not elevated in ME. Lots of people have looked. And if they think B cell targeted therapy works for a subset then either they really are selling something or they are just badly out of touch with the field. Either way is not encouraging.

 

[alex3619]

I suppose what puts me off papers like this most is the way the text seems to want to sell something and bring in every tiny bit of data that might support and forget all the caveats. In reality we can be pretty sure that B cells are not elevated in ME. Lots of people have looked. And if they think B cell targeted therapy works for a subset then either they really are selling something or they are just badly out of touch with the field. Either way is not encouraging.

We will have to wait and see. I agree that in science we have to be careful with caveats. More importantly they seem to be emphasising that finding a marker means its a significant causal problem and not an epiphenomena. It might be useful as a diagnostic tool, but as yet there is no solid evidence that it might indicate treatment options.

Five years ago or so I stopped counting statistically significant biological markers . . . we were up to over 2500. Most of these, if not all of them, will be secondary and not necessarily useful treatment targets. Many will be irrelevant, and just statistical anomalies. However we might be able to use them to indicate improvement or worsening of the underlying issues. This is still a nearly completely ill understood area though. These indicate possibilities not definitive treatment targets. Currently I see no evidence that this latest finding is any different. Nice possibilities, worth looking at further, but nothing solid yet.

 

[FMMM1]

I haven't looked into the data yet, but I think the study deserves to be read.

From a quick skim read, they didn't find any evidence of differences in clonal expansion in terms of how many receptor types account for 50% of the variation. But they did find differences in the expression of some particular genes.





They noted that another researcher using another technique had identified the same genes as of interest in ME/CFS.



They suggested expression of these genes could usefully distinguish ME/CFS from healthy controls.




This finding of increased B cells in the ME/CFS cohort is interesting (I haven't looked to see how robust the finding is), raising questions of whether this particular sample is different to other ME/CFS populations, and if this has affected the finding of gene expression.



They note that what they are finding might just be a result of the ME/CFS patients having had an infection (the triggering one) and the healthy controls not having had such an infection. Maybe what they have found might possibly just be a result of ME/CFS patients not being able to control latent infections e.g. herpes simplex, that is, a downstream effect?



The company that did the BCR repertoire analysis is a private company. And I'm not sure where they have got the idea that B cell-target therapy is effective in a subgroup of patients from.

This makes me think of the fact that Post Treatment Lyme disease and ME/CFS are clinically similar [https://pubmed.ncbi.nlm.nih.gov/21383843/]. If this paper has identified a mechanism, then could that explain Post Treatment Lyme disease/post infection ME?
Has this technique yielded results in understanding other diseases?

 

[Marky]

Question.. IGHV is genes coding for the function and structure of immunoglobulins?

I had a hard time googling it

 

[alex3619]

From a Google search -


"Immunoglobulin heavy chain variable region (IGHV) gene mutation status is a biomarker for the prognosis of chronic lymphocytic leukemia, whether it is associated with the diagnosis, staging, and prognosis of patients with mantle cell lymphoma (MCL) remains to be determined.

A high mutation rate of immunoglobulin heavy chain variable ...
https://journals.lww.com › md-journal › fulltext › a_high... "

 

[Hutan]

Has this technique yielded results in understanding other diseases?

That's a good question.

Each B cell expresses a single B cell receptor (BCR), and the diverse BCR expressed by an individual’s total B cell population is termed the “BCR repertoire.” BCR repertoire skewing is correlated with microbial interactions, vaccinations, or immune-mediated diseases, such as systemic lupus erythematosus (Hershberg and Prak, 2015) or multiple sclerosis (Baranzini et al., 1999). Next-generation sequencing (NGS) technology has enabled the comprehensive quantitative analysis of BCRs at the clonal level (Kitaura et al., 2017). For example, oligoclonal expansions of IGHV4–28/IGHJ4-carrying IgG-83expressing B cells were identified in patients with primary immune thrombocytopenia using NGS (Hirokawa et al., 2019).

The authors suggest that the B-cell receptor repertoire is affected by microbial interactions, vaccinations, or immune mediated diseases such as lupus or MS. They give an example of a study done by a Japanese researcher that found a biomarker for patients with primary immune thrombocytopenia.

 

[Jonathan Edwards]

Question.. IGHV is genes coding for the function and structure of immunoglobulins?

I had a hard time googling it

Each antibody is made from several genes cobbled together - a bit like taking a cone and then adding two different flavour scoops of ice-cream from a wide selection. The cone is the constant bit that gets the tag C. The flavours are variable so get the tag V. In fact the cone is made up of a long and a short part wrapped together like an almond twist, both getting variable bits added in. The long part is the heavy chain H and the short part the light chain L. So every segment in an antibody is either HC, HV, LC or LV.

So analysing all the HV (often traditionally called Vh) genes used by B cells means looking at the cobbled genes in each B cell and picking out the bit that encodes for the variable part of the heavy chain - which will have been moved in to the cobbled gene from the mini-gene selection (the row of flavour trays on the serving counter). If I remember rightly there are four main HV mini gene groups (used to be Vh1-Vh4) but there are sub variants within groups (chocolate, chocolate chip, white chocolate, rum and chocolate) so that there may be as many as 50 HV flavours (maybe more are now known).

So unlike in the ice-cream shop every antibody is just made at random from cobbling a light chain gene from some cone and some flavours and a heavy chain gene from the same and then making protein from those two genes and twisting them together. And the random ice-creams either prove unwanted and melt away or, if they fit the taste of a nice virus or bacteria they are made by the billion and exported into the bloodstream. It is the variable flavour regions that bind to antigens.

Now you might expect say flu to always fit antibodies using HV20.6 and LV 10.3 flavours and malaria to go for HV4.23 and LV2.00. But things are complicated because each microbe has several protein antigen each with maybe ten epitopes - maybe a hundred shapes to fit into and also the antibody genes go through a further process of variation that gives every single antibody a unique cobbled gene. So in practice the correlation between HV usage and immunity to specific infections is extremely weak and statistical.

I personally think that using HV repertoire to try to identify triggering infections is unlikely to show anything consistent. There is, however, a much more interesting story linked to HV gene usage. The HV4.34 gene (raspberry ripple) crops up on antibodies in lupus and RA. This is not because it relates to an infection but because this particular HV gene segment binds to innate ligands in such a way that any B cell using it has a different set of rules for survival. In simple terms it is allowed to survive outside lymphoid follicles and make a bit of antibody early on but in the normal system it is never allowed to turn into established plasma cells. So this is an HV gene that makes the antibody difference in a way a bit like the difference between IgM and IgG. Another HV gene (Vh3 family) codes for affinity of binding to the staphylococcus protein A that is a feature of all IgG antibodies. The binding site does not bind to antigen in the normal way - it is outside the antigen binding region - and the fact that staph protein A binds it is probably to do with a clever trick of staphylococcus and nothing to do with immunity to staph.

So it is all complicated but potentially relevant. I think Jo Cambridge's viewpoint would be that this was a very sensible study to try but the results may or may not lead us anywhere much.

 

[jaded]

I have just come across a YouTube video published on the 8th November entitled:

‘ME/CFS and Long COVID: A Discussion between Dr. Avindra Nath and Dr. Takashi Yamamura’



They discuss the earlier B cell receptor findings from the Japanese group.

Avi Nath seems to be impressed and thinks they’re on to something and has confirmed they also found immune dysregulation and mitochondrial problems as he’s previously mentioned in other videos in his NIH cohort of ME/CFS patients.

The Japanese group have mentioned starting a clinical trial aimed at B cells but didn’t allude to what.

Can anybody make out what autoantibodies Avi is referring to around the 19-20 minute mark?

The Japanese group also have a short summary of their findings in English here:

https://www.ncnp.go.jp/topics/2021/20210520eng.html