[EV] proteomics uncovers energy metabolism, complement system, and [ER] stress response dysregulation postexercise in males with [ME/CFS], 2025,Glass+

[wigglethemouse]

Quite notable that study participation was concluded more than 5 years ago.

Dr Hanson was public a few years ago that the first 5 year NIH grant U54NS105541 2017-2021 did not have sufficient funding to process all samples at all time points. So I am glad when funding was renewed in 2023 & 2024 with NIH grant U54AI178855 that they could process the samples collected from the men. At least I think that is what must have happened.

 

[jnmaciuch]

Just got to say that mass spectroscopy and the technology around it is pretty close to magic. To look at a sample and identify nearly all the proteins in it - it's absolutely fantastic. The technology is there to understand ME/CFS. The results from this tiny study look important. But researchers should not have to be faffing around with tiny and old cohorts. If a solid bit of money was thrown at studies like this, I think the clear answers would come. The problem is important enough and the tools are available.

Haha I had the same exact reaction! “What do you mean you’re basically rocket launching a protein into a barrel of stuff to crash into and that somehow tells you what protein it is?” Obviously it still has its limitations and there will always be a place for more targeted wet lab experiments. But it is the technology that makes scientists look the most like wizards in my opinion.

 

[Hutan]

Differentially Abundant Proteins

We identified 31 DAPs at baseline, 39 at 15 min postexercise, and nine at 24 h postexercise. Most DAPs at all time points exhibited decreased levels of EVs from ME/CFS patients compared with controls, as indicated by the Log2 median fold change being below 0 (Figure 3A).

Bearing in mind that these differentially abundant proteins are in extracellular vesicles, so they are being sent somewhere, presumably for some reason.

Baseline
Lower: ANXA3, VIM, RAP1A, PRKAR2A, SCRN1, PCSK6, PPP1R14A, SYNCRIP, DARS1, MPST, DAPP1, PAFAH1B1, PGD, TXNDC5, FKBP3, ADH5, HNRNPK, RDX, KPNB1, CMPK1, RAB7A, AKR1A1, CNPY2, PTGES3, PSMA5,
Higher: BLMH, TBCA, GGCT, NSF, CD84, ACTN3

15 mins after exercise
Lower: DSG1, RAP1A, SERPIN89, PRKAR2A, SCP2, GAB, DARS1,TXNDC5,PTPN11, ADH5, PDCD6IP, CRLF3, HNRNPK, TAOK3, GSTP1, PGK1, ERP44, P1L4, IPO7, VBP1, FKBP3, UBE2L3, PSMA4, USO1, INPP4B, P4HB, YWHAE, CLTA, HYOU1, TPM3, PDIA4, RDX, RAB8B, PTGES3, CALR
Higher: BLMH, PRDX6, ADD1, TFRC

24 hours after exercise
Lower: VIM, SYNCRIP, PRKAR2A, ANXA3, ARCN1, DARS1, VBP1, TPT1
Higher: GGCT

Among these, PRKAR2A and DARS1 consistently showed decreased abundance across all three time points, suggesting persistent dysregulation in ME/CFS EVs, independent of exercise. PRKAR2A (Protein Kinase A Regulatory Subunit 2 Alpha) regulates the activity of protein kinase A (PKA) and is involved in various cellular processes including metabolism, cell division, and memory. DARS1 (Aspartyl-tRNA Synthetase 1) catalyzes the attachment of aspartic acid to tRNA, which is essential for protein synthesis.

Notably, ANXA3, VIM, GGCT, and SYNCRIP were differentially abundant at both 0 and 24 h, which suggests any potential changes due to exercise have stabilized at 24 h. ANXA3 is linked to membrane repair, VIM to cellular integrity, GGCT to oxidative stress management, and SYNCRIP to RNA processing, suggesting persistent cellular stress and dysfunction in ME/CFS.

Additionally, eight proteins, ADH5, BLMH, FKBP3, HNRNPK, PTGES3, RAP1A, RDX, and TXNDC5, were differentially abundant at 0 h and 15 min postexercise. ADH5 and BLMH are involved in detoxification, with BLMH showing increased levels in ME/CFS EVs compared with controls, suggesting a heightened stress response or protein degradation. FKBP3 is involved in protein folding and immunoregulation, HNRNPK in mRNA processing, PTGES3 in inflammation, RAP1A in cell adhesion, and RDX in cell structure. TXNDC5 is a member of the protein disulfide isomerase (PDI) family of endoplasmic reticulum proteins that can catalyze protein folding. The consistent dysregulation of these proteins before and after exercise suggests it is a baseline characteristic of ME/CFS, unaffected by immediate exercise responses, which again points to a chronic underlying pathology.

VHL binding protein 1 (VBP1) was the sole protein differentially abundant at both postexercise time points (15 min and 24 h), showing decreased levels in ME/CFS patients, indicative of exercise-induced dysregulation without recovery by 24 h. Significantly, 28 DAPs were unique to the 15 min postexercise time point, with 25 showing decreased levels in ME/CFS patients, highlighting a potential failure in rapid EV-mediated signalling responses to exercise in ME/CFS.

The first thing that occurs to me is the consistency of the results at the time points. Of course it might be noise - e.g did the ME/CFS participants dose up on coffee just before going to the lab? But, with all those lower levels, it really looks as though the ME/CFS participants are not mounting a normal response to exercise.

 

[Hutan]

VBP1
I just have to leap in here and comment on this protein, the only one that was found to be lower in EVs in the ME/CFS men 15 minutes and 24 hours after exercise - but not at baseline.

The VBP1 gene, which encodes the Von Hippel-Lindau binding protein 1, is involved in regulating the stability of proteins, including HIF-1α, which is crucial for cellular adaptation to low oxygen levels. Exercise can impact VBP1 expression, with some studies suggesting it may be influenced by exercise intensity and duration.
Here's a more detailed explanation:

• VBP1's Role:
  VBP1 is a chaperone protein that interacts with the Von Hippel-Lindau (pVHL) protein, a key player in the degradation of HIF-1α under normal oxygen conditions.
• Exercise and HIF-1α:
  Exercise, particularly intense exercise, can induce a transient increase in HIF-1α levels in skeletal muscle.
• VBP1 and HIF-1α Regulation:
  VBP1 has been shown to enhance pVHL's stability and facilitate the degradation of HIF-1α.
• Exercise-Related Changes:
  Some studies suggest that VBP1 expression may be influenced by exercise intensity, with higher intensity exercise potentially leading to changes in VBP1 expression and its interaction with HIF-1α.
• Mechanism of Action:
  VBP1 may negatively regulate CHIP, a ubiquitin ligase involved in the degradation of HIF-1α, potentially impacting its overall activity during and after exercise.

In summary: Exercise can influence the expression and activity of VBP1, which in turn affects HIF-1α levels. While the specific mechanisms and the extent of these changes are still being investigated, it is clear that VBP1 plays a role in the cellular response to exercise, particularly in regulating HIF-1α, a key regulator of cellular adaptation to low oxygen levels.

So, exercise temporarily increases levels of HIF-1a (Hypoxia Inducible Factor 1a). (That didn't show up in the EVs). VBP1 helps VHL protein degrade HIF-1a. So, if VBP1 isn't turning up to cells at around the same levels as in healthy people, that might mean that the HIF-1a isn't degraded as quickly as normal.

There are all sorts of interesting papers about VBP1 e.g.

VBP1 has been shown to cooperate with pVHL in regulating protein stability. For example, VBP1 regulated tubulin degradation together with pVHL

 

[Hutan]

Enriched pathways based on the differentially abundant proteins

To gain functional insight into these DAPs, we performed pathway analyses using Enrichr with the list of all DAPs at each time point as the input (Figure 3B, 3 < .05, Fisher's exact test followed by Benjamini and Hochberg false discovery rate (BH FDR) correction).

At baseline, the most significantly enriched pathways, featuring the largest number of DAPs, were metabolism (11 proteins), neutrophil degranulation (5 proteins), and biological oxidations (4 proteins). This suggests that even at rest, ME/CFS patients exhibit substantial disruptions in metabolic processes and immune functions, specifically those involving neutrophils, which are key players in the body's first line of defence.

At 15 min postexercise, three of the top 10 most significant pathways are related to the immune system and disease (12 proteins). This highlights a potential failure of the appropriate EV-mediated immune response in ME/CFS patients during the acute recovery phase after exercise. Another three of the top 10 pathways are involved in vesicle-mediated transport and biogenesis, suggesting that the large-scale dysregulation of EV proteomic cargo in response to exercise may be linked to defects in the production of EVs themselves.

The remaining top pathways with the highest number of proteins involved (8) are related to the cellular response to stress and stimuli. The reduction in the cellular response to stress/stimuli-related proteins in ME/CFS patients’ EVs 15 min postexercise is likely contributing to their exercise intolerance.

At 24 h postexercise, with only nine DAPs identified, there were no significantly enriched pathways that met our strict threshold.

 

[Hutan]

Enriched pathways based on all of the identified proteins

Protein set enrichment was assessed using four databases (KEGG, Reactome, Wikipathways, and PANTHER, multiGSEA R package,51 complete results in Table S5).

Our analysis revealed significant pathway alterations in ME/CFS patients compared with controls, particularly at 15 min postexercise, where 26 pathways were significantly enriched. In contrast, only one pathway was significantly dysregulated at baseline and three pathways at 24 h postexercise (Figure 3C, 3 < .05, Table S5). The top five most significant pathways, predominantly involving the complement and coagulation cascades, were consistently identified across three different databases.

Core contributors to the enrichment of these pathways included key components of the complement system, such as C3, CFH, C4BPA, and the classical complement pathway proteins C1qA/B/C. For the coagulation cascade, crucial members included fibrinogen chains FGA, FGB, and FGG, as well as coagulation factors XI and V, and plasminogen. For further details, refer to Table S5, which provides the leading-edge proteins for each set. The strong positive normalized enrichment scores (NES), ranging from 2.5 to 3, indicate elevated levels of EV proteins in male ME/CFS subjects relative to controls, with 10 to 35 proteins contributing to overactivation of these pathways.

We observed strong negative NES for many metabolism-related pathways at 15 min, suggesting a failure in male ME/CFS patients to mount an adequate metabolic signalling response to exercise in EVs. These dysregulated pathways encompass broad metabolic processes, including general metabolism (185 proteins), the citric acid (TCA) cycle and respiratory electron transport (26 proteins), valine, leucine, and isoleucine degradation (7 proteins), tryptophan metabolism (8 proteins), fatty acid metabolism (21 proteins), pyruvate metabolism and the TCA cycle (13 proteins), and lipid metabolism (46 proteins). The consistent downregulation in these pathways points to a potential energy production and utilization deficit in ME/CFS patients, which might underlie the profound fatigue and exercise intolerance characteristic of the disease.

Complement system and coagulation cascade, exercise-related metabolism. Pathways relevant to mitochondria are mentioned.

 

[jnmaciuch]

Interestingly, the only pathway significantly enriched at 0 h, the complement system (Wikipathways), remained elevated 15 min postexercise. This suggests that the complement system in ME/CFS patients is already primed for overactivation prior to exercise, and this overactivation is exacerbated during the immediate postexercise phase. This chronic activation of the complement system could be a contributing factor to the persistent inflammation and immune dysregulation observed in ME/CFS.

I’ll quickly flag what seems to be a misinterpretation of the data from the text. I may be wrong about this, but fairly certain that these findings can’t be used to determine complement activation, only the abundance of complement proteins. Complement activation involves various subunits breaking off and joining together and changing shape—things that would not be detectable in an LC-MS/MS experiment where you are fragmenting the proteins.

So you can’t conclude complement activity, but you can interpret this as a sign of other signaling which is known to upregulate the transcription of complement genes. Off the top of my head I think IL-6 and IL-1 do, probably more beyond those

 

[Hutan]

Within-subject fold changes

The protein with the largest |Log2 median FC| was CS (citrate synthase), which, on average was not altered by exercise in the controls but showed a reduction from 0 h to 15 min postexercise in most ME/CFS patients (Figure 4B). As the first enzyme in the TCA cycle, CS plays a critical role in cellular energy production, and its downregulation may point to impaired energy metabolism in ME/CFS patients.

For SERPINF1, EV levels increased in most ME/CFS patients during the rapid response to exercise and remained elevated 24 h postexercise, in contrast to the minimal changes observed in the controls. This sustained elevation, indicating dysregulation in the rapid response to exercise as well as impaired recovery, may be contributing to PEM experienced by ME/CFS patients.

Capping actin protein of muscle z-line alpha subunit 2 (CAPZA2) exhibited a different temporal pattern, with reduced levels at 24 h postexercise in most subjects from both groups, although the decrease was more pronounced in the control group compared with patients. CAPZA2 is crucial for regulating the dynamic response of actin filaments, which is essential for muscle contraction, and its dysregulation may reflect impaired muscle recovery or remodelling in the context of ME/CFS.

 

[Hutan]

I may be wrong about this, but fairly certain that these findings can’t be used to determine complement activation, only the abundance of complement proteins.

Something I haven't got my head around yet is in what direction the pathways may be affected. The authors mention in one place 'a strong negative NES' - I don't know what that means and haven't looked yet.

 

[jnmaciuch]

Something I haven't got my head around yet is in what direction the pathways may be affected. If most of the levels of proteins are lower, can that mean a pathway is activated?

It’s not any lack of understanding on your part—data like this, especially from the circulation, is especially vague precisely because many different scenarios can all lead to upregulation or downregulation.

In the specific context of EVs, it could be that the specific protein isn’t being produced enough in cells to get sent out in the first place, or it’s being really used up by the cells, or there’s some dynamics with vesicle packing that are affected.

It’s really a toss up, which is one of the biggest limitations of screening approaches like this. It’s sort of like examining the trash of dozens of people—is something appearing in the garbage less because they’re buying it less, or using it for longer, or because they’re disposing of it a different way? It’s anybody’s guess if you can only look at abundance in the trash.

At most, you can just conclude “there’s less/more of X protein in the EVs, we ought to do more specific experiments to sort out what’s actually going on.”

The authors mention in one place 'a strong negative NES' - I don't know what that means and haven't looked yet.

I assume it refers to normalized enrichment score in the context of GSEA. Enrichment score is just a metric for how much the analytes in a specific pathway are distributed towards the most upregulated or downregulated analytes between your comparison groups. Negative NES means the pathway contains many strongly downregulated proteins. Normalization in this context just means that the enrichment score is modified according to pathway size, so that pathways that happen to have hundreds of features don’t automatically generate higher scores than smaller pathways.

 

[forestglip]

At baseline, the most significantly enriched pathways, featuring the largest number of DAPs, were metabolism (11 proteins), neutrophil degranulation (5 proteins), and biological oxidations (4 proteins).

The proteins enriched in the neutrophil pathway at baseline were PSMA5, RAP1A, KPNB1, TXNDC5, RAB7A and at 15 minutes post-exercise were ERP44, RAP1A, GSTP1, DSG1, TXNDC5. All lower in ME/CFS EVs.

Neutrophils were a hot topic in the study from a few days ago:

Preprint Charting the Circulating Proteome in ME/CFS: Cross System Profiling and Mechanistic insights, 2025, Hoel, Fluge, Mella+

Among the proteins that showed the most comprehensive reduction were a group of histones (H1x, H2.1, and several H2B and H2A types) (Figure 1E-F). These are normally nuclear proteins but can be released by activated neutrophils as part of neutrophil extracellular traps (NETs)33.

For bone marrow proteins, proteins such as PADI4, BPI, and MPO, had a sharp reduction, which may indicate altered granulocyte/neutrophil cell function.

The reduced amount of granulocyte proteins was not associated with abnormally low neutrophil counts (the most abundant type of granulocytes) or other leukocyte types in the patients (DocumentS1, Table S1). Furthermore, comparing a list of proteins associated with neutrophil granules and stimulated neutrophil protein release 33 45, we found that about 40% or more of the proteins reported to be released by activated neutrophils showed lower serum concentrations in the ME/CFS group compared to the HC group, suggesting a suppressive effect on overall neutrophil activity.

 

[DMissa]

Only thing I’m always slightly concerned about is the amount of missingness and imputation. If they provide the data pre-imputation I’d like to manually check the how many of the top hits were from proteins with a lot of imputation.

per protein imputation frequency is in the supplementary files

 

[Hutan]

It’s really a toss up, which is one of the biggest limitations of screening approaches like this. It’s sort of like examining the trash of dozens of people—is something appearing in the garbage less because they’re buying it less, or using it for longer, or because they’re disposing of it a different way? It’s anybody’s guess if you can only look at abundance in the trash.

Ha. Even worse, I think it's like all the parcels that dozens of people put out for the courier to pick up got mixed up with their trash.

 

[Hutan]

There's truly something for everyone's pet theory here.

Vimentin - VIM, was the protein with the second biggest fold reduction compared to the controls at baseline and the biggest fold reduction at 24 hours after exercise.

This gene encodes a type III intermediate filament protein. Intermediate filaments, along with microtubules and actin microfilaments, make up the cytoskeleton. The encoded protein is responsible for maintaining cell shape and integrity of the cytoplasm, and stabilizing cytoskeletal interactions. This protein is involved in neuritogenesis and cholesterol transport and functions as an organizer of a number of other critical proteins involved in cell attachment, migration, and signaling. Bacterial and viral pathogens have been shown to attach to this protein on the host cell surface.

There's collagen and neurons, microtubules, pathogens.
There are interesting sounding papers about vimentin. Here's one
Extracellular vimentin as a modulator of the immune response and an important player during infectious diseases, 2024

Vimentin, an intermediate filament protein primarily recognized for its intracellular role in maintaining cellular structure, has recently garnered increased attention and emerged as a pivotal extracellular player in immune regulation and host–pathogen interactions. While the functions of extracellular vimentin were initially overshadowed by its cytoskeletal role, accumulating evidence now highlights its significance in diverse physiological and pathological events.

This review explores the multifaceted role of extracellular vimentin in modulating immune responses and orchestrating interactions between host cells and pathogens. It delves into the mechanisms underlying vimentin's release into the extracellular milieu, elucidating its unconventional secretion pathways and identifying critical molecular triggers. In addition, the future perspectives of using extracellular vimentin in diagnostics and as a target protein in the treatment of diseases are discussed.

"delves into the mechanisms underlying vimentin's release into the extracellular milieu, elucidating its unconventional secretion pathways and identifying critical molecular triggers"

Reading that paper, it is suggesting that all sorts of cellular stressors (pathogens, tissue damage, autoimmune disease) should increase extracellular vimentin, so it's rather surprising that the levels are lower than in healthy controls.

 

[forestglip]

Baseline
Lower: ANXA3, VIM, RAP1A, PRKAR2A, SCRN1, PCSK6, PPP1R14A, SYNCRIP, DARS1, MPST, DAPP1, PAFAH1B1, PGD, TXNDC5, FKBP3, ADH5, HNRNPK, RDX, KPNB1, CMPK1, RAB7A, AKR1A1, CNPY2, PTGES3, PSMA5,
Higher: BLMH, TBCA, GGCT, NSF, CD84, ACTN3

15 mins after exercise
Lower: DSG1, RAP1A, SERPIN89, PRKAR2A, SCP2, GAB, DARS1,TXNDC5,PTPN11, ADH5, PDCD6IP, CRLF3, HNRNPK, TAOK3, GSTP1, PGK1, ERP44, P1L4, IPO07, VBP1, FKBP3, UBE2L3, PSMA4, USO1, INPP4B, P4HB, YWHAE, CLTA, HYOU1, TPM3, PDIA4, RDX, RAB8B, PTGES3, CALR
Higher: BLMH, PRDX6, ADD1, TFRC

24 hours after exercise
Lower: VIM, SYNCRIP, PRKAR2A, ANXA3, ARCN1, DARS1, VBP1, TPT1
Higher: GGCT

Thank you for the color coding. What a good idea for quickly seeing matches. Adds some colorful decoration to the thread as well.

 

[wigglethemouse]

The proteins enriched in the neutrophil pathway at baseline were PSMA5, RAP1A, KPNB1, TXNDC5, RAB7A and at 15 minutes post-exercise were ERP44, RAP1A, GSTP1, DSG1, TXNDC5. All lower in ME/CFS EVs.

Neutrophils were a hot topic in the study from a few days ago:

Preprint Charting the Circulating Proteome in ME/CFS: Cross System Profiling and Mechanistic insights, 2025, Hoel, Fluge, Mella+

And neutrophils were also highlighted in the recent Swedish paper - see thread
[Transcriptome] in PBMCs of [LC] at a median follow-up of 28 months [...] reveals upregulation of JAK/STAT [...], 2025, Fineschi et al

Furthermore, we found an upregulation of the genes LTF (log2 fold change 1.6), ANXA3 (log2 fold change 1.0), and TCN1 (log2 fold change 1.3) that encode for the proteins Lactoferrin, Annexin3, and Transcobalamin1, respectively, all of which are associated with neutrophil degranulation and vesicular transport.

 

[DMissa]

At most, you can just conclude “there’s less/more of X protein in the EVs, we ought to do more specific experiments to sort out what’s actually going on.”

Jess is right here, we love to interpret and speculate but with omics data it's really a discovery process for downstream testing. I wouldn't get too hung up on specifics. Speculation is useful but we can only reasonably do so much of it with this type of data. Hopefully the changing landscape won't hamper Maureen's group from chasing up the next steps and delivering the answers to us

 

[Hutan]

And neutrophils were also highlighted in the recent Swedish paper - see thread
[Transcriptome] in PBMCs of [LC] at a median follow-up of 28 months [...] reveals upregulation of JAK/STAT [...], 2025, Fineschi et al

I think the Fineschi paper is saying that there was twice as much ANXA3 mRNA in Long Covid PBMCs compared to controls. Whereas this study found a reduction in ANXA3 in ME/CFS EVs. It's hard to know if that hangs together or not.

 

[DMissa]

I think the Fineschi paper is saying that there was twice as much ANXA3 mRNA in Long Covid PBMCs compared to controls. Whereas this study found a reduction in ANXA3 in ME/CFS EVs. It's hard to know if that hangs together or not.

the EVs could be coming from any tissue

 

[Jonathan Edwards]

I haven't had time to look at more than the abstract and members analyses. But to me there seems to be an over interpretation in the abstract presentation - which members have picked up on.

It talks of a adaptive response - which I think means a response with adverse outcomes. It also talks of EV-mediated processes. OK, there may be some evidence that EV are involved in signalling from one cell to another, but are any of the proteins measured here relevant to that? If EV are mostly a neat way to get IL-26 and TGF beta from one cell to another distant cell what does it matter if they contain garbage from an unhappy cell they have come from ? It seems that if anything the study is likely to be a way to pick up bad things in provider cells just by chance.

And maybe the most likely cells to be providing these EV are things like senescent neutrophils - which aren't very interesting and may be present in greater numbers in people with ME/CFS for uninteresting reasons.

I worry that people are measuring EV simply because they are rather easy to collect and measure. Garbage might be very important in ME/CFS but I am not sure there is justification for suggesting that somehow this is malfunctioning garbage.