EBV/HHV-6A dUTPases contribute to ME/CFS pathophysiology by enhancing TFH cell differentiation and extrafollicular activities, 2022, Cox et al

[Andy]

Full title: EBV/HHV-6A dUTPases contribute to Myalgic Encephalomyelitis/Chronic-Fatigue-Syndrome pathophysiology by enhancing TFH cell differentiation and extrafollicular activities

Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) is a chronic, debilitating multisystem illness of unknown etiology for which there is no cure and no diagnostic tests available. Despite increasing evidence implicating EBV and human herpesvirus-6A (HHV-6A) as potential causative infectious agents in a subset of ME/CFS patients, there are few mechanistic studies to address a causal relationship.

In this study we examined a large ME/CFS cohort (n=351) and 77 controls and demonstrate a significant increase in activin A and IL-21serum levels, which correlated with seropositivity for antibodies to the EBV and HHV-6 protein deoxyuridine-triphosphate nucleotidohydrolase (dUTPase), but not CXCL13. These cytokines are critical for T follicular helper (TFH) cell differentiation, generation of high-affinity antibodies and long-lived plasma cells. Notably, ME/CFS serum was sufficient to drive TFH cell differentiation via an activin A-dependent mechanism. The lack of simultaneous CXCL13 increase with IL-21 indicates impaired TFH-function in ME/CFS. In vitro studies revealed that virus-dUTPases strongly induced activin A secretion while in vivo, EBV-dUTPase induced the formation of splenic marginal zone B and invariant NKTFH cells.

Altogether, our data indicate abnormal germinal center (GC) activity in ME/CFS subjects and highlight a mechanism by which EBV and HHV6-dUTPases may alter GC and extrafollicular Ab responses.

Open access, https://insight.jci.org/articles/view/158193

 

[Hutan]

This looks interesting, looks like quite a few of us missed it. Brandon Cox is presenting at the 2022 IACFS/ME conference.

 

[Hutan]

identified activin A, an early regulator of human follicular helper T cell (TFH) differentiation, as well as IL-21, a regulator of germinal center (GC) TFH cell survival and plasma cell differentiation, to be markedly increased in ME/CFS sera. The lack of simultaneous increase of CXCL13 with IL-21 indicates impaired TFH function in ME/CFS.

we measured serum activin A levels in 351 individuals with ME/CFS, 54 veterans with GWI, and 77 control individuals

Figure 1 - measures of activin A, IL-21 and CXCL13 in the sera of ME/CFS, GWI and healthy controls



Those results look very interesting. Clearly, the increases in Activin A and IL-21 aren't happening in every person diagnosed with ME/CFS and GWI, but there does seem to be a significant subset. Note the y axis scales - these levels are very definitely different.

And the sample size is very good. So, is the result something that can be explained by different lifestyles or is it a real clue? And would some of the other ME/CFS and GWI people show similar increases if the blood sample was taken on a bad day?

 

[Hutan]

So, who has high activin A levels?
Serum activin A and B levels predict outcome in patients with acute respiratory failure: a prospective cohort study 2013
These authors thought high activin A levels were predictive of bad health outcomes and suggested they indicate inflammatory challenges:

30 day mortality in patients with Acute Respiratory Failure (ARF) is approximately 30%, defined as patients requiring ventilator support for more than 6 hours.
Serum activin A and B were significantly elevated in most patients and in most of the diagnostic groups. Patients who had activin A and/or B concentrations above the reference maximum were significantly more likely to die in the 12 months following admission
Given the existing data from animal studies linking high activin A levels to significant inflammatory challenges, the results from this study suggest that approaches to modulate activin A and B bioactivity should be explored as potential therapeutic agents.

Serum immunoreactive activin A levels in normal subjects and patients with various diseases 1996
This 1996 paper suggests normal levels of activin A are marginally higher in men than in women.

Serum ir-activin A level in healthy adults was 1.27 +/- 0.03 micrograms/L (mean +/- SEM, n = 180); being 1.38 +/- 0.05 micrograms/L (n = 90) in male, and 1.16 +/- 0.05 micrograms/L (n = 90) in female subjects, with a tendency to increase with age.

But, there were a lot more women in the ME/CFS (83%) and the GWI (93%) cohorts than in the healthy control cohort (51%) in the 2022 study, so gender would suggest that the disease samples should have had lower, not higher levels.

The 1996 paper also suggests that activin A levels increase with age. But, the average ages of the disease cohorts (47 and 44 years) were a lot lower than the average of the healthy control cohort (67 years). So age would suggest the disease cohorts should have had lower levels of activin A.

The 1996 paper reports increases in pregnancy and in some diagnosable health conditions, but these seem unlikely to account for differences in the 2022 paper findings.

serum ir-activin A level was elevated in patients with hyperthyroidism (1.91 +/- 0.37 micrograms/L, n = 31), liver cirrhosis (2.03 +/- 0.71 micrograms/L, n = 10), chronic renal failure (3.41 +/- 0.34 micrograms/L, n = 41), and advanced solid cancer (2.24 +/- 0.52 micrograms/L, n = 67).

The units in the 2022 paper are pg/mL (picograms per millilitre), so I think we need to divide by 1000 to get to micrograms per litre. The mean for the controls in the 2022 paper was around 500, so 0.5 micrograms/L whereas the means for ME/CFS and GWI were around 10,000, so 10 micrograms/L. I don't know if the methods are comparable, but even with different methods, differences between healthy and unhealthy levels should be comparable. The 1996 paper was suggesting that unhealthy levels are about 2 to 3 times the healthy levels. But here, we have a lot of people in the disease cohorts having activin A levels that are 20 or more times higher than those of the healthy controls.

Given people with chronic renal failure and advanced solid cancer are probably not very active, that would seem to rule out the possibility that a sedentary lifestyle is somehow causing the higher activin levels.

So, I think the idea of higher activin A levels is interesting.

 

[Hutan]

One problem is that another study of activin A found normal levels in ME/CFS cases:

Our findings concerning serum activin A levels are in conflict with the study by Lidbury et al. (32), who reported a significant increase in activin B but not activin A. The reasons for these differences are unknown but may reflect differences in the genetic background of these individuals and/or variations in the case definition criteria used. However, a recent study by the same group using a different cohort of patients with ME/CFS (n = 134) reported no significant differences in activin B plasma levels between ME/CFS cases and controls (n = 54) (33). Thus, using activin B levels as a biomarker for ME/CFS remains controversial.

We discussed that Lidbury paper here: Activin B is a novel biomarker for chronic fatigue syndrome/myalgic encephalomyelitis (CFS/ME) diagnosis, 2017, Lidbury et al
and here:Rethinking ME/CFS Diagnostic Reference Intervals via Machine Learning & Utility of Activin B for Defining Symptom Severity (2019) Lidbury et al.
and here: Potential of Activin B as a Clinical Biomarker in Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS), 2021, Gravelsina et al

Maybe the Lidbury group had technical issues? (Or maybe this group did...)

 

[Hutan]

So, what about IL-21, and the subset this 2022 study found with the stunningly high values? (around 100 pg/mL for the healthy cohort; with the disease subsets with around 20,000 pg/mL).


We have looked at this paper with a POTS cohort:
Inflammatory Biomarkers in [POTS] with Elevated G-Protein-Coupled Receptor Autoantibodies, 2021, Gunning, Grubb et al

Those values reported for the POTS study are also in pg/mL and seem to fit. Less than 200 pg/mL is reported as normal, and in this case, levels in the POTS cases were around 10 times higher at around 2000 pg/ml.



Serum IL-21 levels are elevated in atopic dermatitis patients with acute skin lesions 2016
This 2016 study found as the title suggests for this allergic skin disease, with higher IL-21 levels being associated with higher severity of acute skin lesions.

The AD patients' serum IL-21 levels (median: 454 pg/mL, n = 79)were significantly higher than those of the healthy control subjects(median: 50.6 pg/mL, n = 17) (P < 0.05).

So those levels also seem to fit with a normal value of less than 200 pg/mL (although I start to wonder if some of the levels in the 2022 ME/CFS +GWI study are possibly unbelievably high).
The atopic dermatitis study authors say:

Several studies have shown that IL-21 contributes to the pathogenesis of allergic diseases. Serum IL-21 levels increase during the acute exacerbation of asthma, and they fall again after treatment.

. They also say that IL-21 levels are raised with mechanical injuries to the skin. They note that serum levels of IL-21 were found to be lower in children with atopic dermatitis, unlike the adults in the 2016 study. So, I don't know. Maybe measuring these substances accurately is difficult? Maybe the levels of them change with the time of day or with storage?

common allergens, house dust mites were shown to induce IL-21 expression by T follicular helper cells in the lymph nodes of allergic model mice.

Plasma interleukin-21 levels and genetic variants are associated with susceptibility to rheumatoid arthritis 2021

The plasma level of IL-21 was significantly higher in subjects with rheumatoid arthritis relative to healthy controls (p < 0.0001). A positive correlation was observed between IL-21 level and DAS28 score, indicating the association of the cytokine with the worsening of the disease (Spearman r = 0.61, p < 0.0001).

Further, in the experimental arthritis model, the blockade of IL-21/IL-21 receptor pathways significantly improved disease severity [16], suggesting an essential role of IL-21 in disease pathogenesis.

RA patients (19.6 ± 0.79 ng/mL) displayed significantly higher levels of plasma IL-21 compared to healthy controls (2.12 ± 0.08 ng/mL) (p < 0.0001).

There seems to be a substantial body of findings linking IL-21 to RA. The unit here is nanograms/mL - to convert that to the picograms/mL used in the 2022 ME/CFS+GWI study, we multiply by 1000. So, they are suggesting the healthy controls had 2000 pg/mL and the RA patients had around 20,000 pg/mL. That healthy control level in the RA study is higher than what the 2022 study found, but it's still in the ball park, as the subset of the disease cohorts had IL-21 levels of around 20,000 pg/mL. The RA study found an exceptionally good relationship between RA severity (DAS 28 score) and IL-21 levels.

The authors mention a number of studies in different countries that have found similar results in RA. But they also note some studies that didn't find the relationship.

So, it looks to me as though the finding of increased IL-21 could tell us something about ME/CFS and GWI. But there is variability in studies of IL-21 in other diseases, which makes me wonder if there are technical difficulties. A study of Kawasaki disease noted

Only one other published study, ..., has included IL-21 levels in there [sic] published data [17]. Unfortunately, the cytokine bead array used was comparably much less sensitive than the published ELISA-based data [15], making direct comparison difficult.

Therefore, it is encouraging to note that this 2022 study used an ELISA method:

Serum levels of activin A, IL-21, and CXCL13 in patients with ME/CFS were determined by ELISA following the manufacturer recommendations.

 

[Hutan]

To me, it looks as though this team have some results in the Activin A and IL-21 levels well worth investigating further. But I'm not convinced by their suggestion that seropositivity of cases to antibodies against herpes virus dUTPases is related to action A and IL-21 levels:

As shown in Figure 1D, the levels of serum activin A were significantly higher in ME/CFS cases seropositive for Abs against the virus dUTPases (5156 pg/mL ± 585.1) than in seronegative cases (3809 pg/mL ± 500.8, P = 0.0353). Likewise, a significant difference was observed in IL-21 levels in ME/CFS cases seropositive for virus dUTPases’ Abs compared with virus-seronegative cases (7818 ± 1481 versus 2672 ± 356.5, P = 0.0014) (Figure 1E).

 

[Hutan]

The next bit of the study - testing to see what the ME/CFS sera can do!

we next tested whether serum from patients with ME/CFS could induce TFH cell differentiation of naive CD4+ T cells in vitro.

It took me a while to work this out. I didn't find the text very clear, but the Figure 2 caption helps a lot.

So they took human naive CD4+ T cells and stimulated them with anti-CD3/CD28–coated beads
1. with nothing else [negative control]
2. + control sera
3. + ME/CFS sera
4. + control sera +IL-12
5. +ME/CFS sera + IL-12
6. + Activin A
7. + IL-12
8. + Activin A +IL-12 [positive control]




The chart shows the percentage of PD-1+CXCR5+ cells after 3 days. (Yes, that's PD-1: Programmed cell death protein, recently discussed elsewhere on the forum.)

That's a pretty remarkable result in Figure 2b. ME/CFS sera, and especially ME/CFS sera helped along with IL-12, produces the same sort of percentages of T cells expressing PD-1 and CXCR5 as a combination of Activin and IL-12 do.

What does this mean; is it important? I don't know yet, I'll have to read on. But one thing it tells us is, if this study is right, the ME/CFS sera is not like the control sera under these particular conditions. Could this finding be wrong? They have done quite a lot of replications (as evidenced by the dots).

 

[Hutan]

And the ME/CFS serum isn't just good at making PD-1 + CXCR5+ cells. The authors found that it was considerably better at making T cells expressing CD4+BCL6+ and BCL6+CXCR5+ than the control serum. The results are very nicely clustered, and there's a decent number of replicates.

 

[Hutan]

The next part of the study was to begin to find out 'what's in the ME/CFS serum that makes it different'? They looked at whether activin A was part of the story by adding follistatin-315 (FST), an activin inhibitor. Figure 3 B shows that the addition of FST did reduce the percentage of those T cells expressing PD-1 and CXCR5 produced by the ME/CFS serum by a lot, as it did for the positive control (Activin A + IL-12).



Figure 3c looks at how much IL-21 the cells incubated for 3 days were secreting.


The cells exposed to 3 days of control sera weren't secreting much IL-21. Relatively speaking, the cells exposed to 3 days of ME/CFS sera were secreting a lot more IL-21. Adding the FST (the activin blocker) to the ME/CFS sera reduced IL-21 secretion down to nearly control sera levels.

Look at the relatively huge amount of IL-21 secreted by the T cells after three days of ME/CFS sera and IL-12. Again, the activin blocker knocks IL-21 levels down.

 

[Hutan]

Gosh, this paper has a lot in it.

Next up - can EBV and HHV-6A dUTPase proteins induce the secretion of regulators of the T cell differentiation?
Well, yes, it seems they can.

At this point, I think I need to know more about viral dUTPase. Brandon Cox and some of the other co-authors of this paper have co-authored
Viral dUTPases: Modulators of Innate Immunity 2022 Ariza at al
It's mostly a survey of the various sorts of viral dUTPases - I think this is the key part for our current purposes:

It has been demonstrated fairly conclusively that the human herpesviruses dUT- Pase proteins possess non-canonical immune and neurological functions that may contribute to the pathologies observed in diseases associated with these viruses. The human herpesviruses dUTPase proteins act as Pathogen-Associated Molecular Pattern (PAMP) molecules and, through their ligation with Toll-like receptor (TLR)2 homodimers in the case of the EBV dUTPase or TLR2/1 heterodimers in the case of the HSV-2, HHV-6, HHV-8 and VZV dUTPases, differentially activate NF-κB, with subsequent modulation of downstream genes involved in chronic inflammation, innate and adaptive immune modulation and neurotransmitter function

In addition to increasing the expression/secretion of the proinflammatory cytokines IL-1β, IL-6, IL-8, IL-12, TNF-α, and IL-10 and IFN-γ from human dendritic cells, macrophages and PBMCs, studies with the EBV dUTPase protein have demonstrated that it is released in extracellular vesicles [111], modulates the differentiation and proliferation of follicular helper T cells, a cell type critical for germinal center development [112], and alters the expression of in vitro and in vivo of targets with central roles in blood–brain barrier integrity/function, fatigue, pain, synapse structure and function, as well as tryptophan, dopamine, and serotonin metabolism [114].

The idea of viral dUTPases in ME/CFS is covered in a recent paper discussed on the forum:
Article: Myalgic Encephalomyelitis/Chronic Fatigue Syndrome: The Human Herpesviruses Are Back! 2021, Ariza

I'm not sure that things are so 'demonstrated fairly conclusively' as this team suggests, but their work has been funded by the NIH for some time.

I'll call it a night for now, but I hope I or someone else can sift through the rest of the paper at some point.

 

[NelliePledge]

I looked up IL 21 on Wikipedia. As a non scientist what jumped out to me was the use in immunotherapy for cancer and flu like symptoms being a side effect https://en.m.wikipedia.org/wiki/Interleukin_21

in all the time skimming through trying to get the gist at my low level of understanding over the last few years the results plots rarely if ever? have differentiated between controls and PWME with clear blue water

 

[Hutan]

Next up - can EBV and HHV-6A dUTPase proteins induce the secretion of regulators of the T cell differentiation?
Well, yes, it seems they can.

Carrying on with this - Figure 4



Monocyte-derived human dendritic cells were stimulated with three different dUTPases for a day:
1. human dUTPase
2. HHV6-dUTPase
3. EBV-dUTPase
and then the culture supernatant was assessed for activin A.
The viral dUTPase clearly made the cells produce activin A in a way that human dUTPase did not.

They note that the herpesviruses’ dUTPase proteins also substantially increased the expression of B cell–activating factor (BAFF) and pentraxin-3 - figures are given to support the claim. They say

These molecules are important for TFH cell differentiation (19), for the proliferation and differentiation of B cells (20), and as adjuvants for marginal zone (MZ) B cells (21).

It looks convincing.

 

[Hutan]

Figure 4B


This team have been busy.

So this experiment was back to the naive CD4+ T cells, cultured with anti-CD3/anti-CD28-coated beads (why those beads? I'm not sure at the moment).

Then various things were added:
1. nothing - control
2. media that has had dendritic cells incubating in it (DCM) with nothing (plus or minus IL-12)
3. DCM +HHV-6A dUTPase (plus or minus IL-12)
4. DCM+ EBV dUTPase (plus or minus IL-12)

After 3 days, the T cells were examined for T cell follicular helper markers - PD-1 +ICOS+ CXCR5+
The viral dUTPases clearly increases the conversion of naive T cells to TFH cells, with the addition of IL-12 increasing the response further.

 

[Hutan]

So, the media that has had human dendritic cells incubating in it (DCM) with viral dUTPase can make T cells differentiate to TFH cells. But do these TFH cells do anything much?
Figure 5



Yes, they do. They produce a lot more IL-21.

If they are re-stimulated with phorbol 12-myristate 13-acetate plus ionomycin (PMA/I), then they produce even more.

(What's PMA/I? It's not explained in the results section of this paper.

Phorbol 12-myristate 13-acetate (PMA) is a diester of phorbol and activates the signal transduction enzyme protein kinase C (PKC). It is used as a tumor promoter and to stimulate division of B-cells. With ionomycin, PMA induces the activation of many cell types to produce cytokines.

So, I'll just accept it as an inducer of cells to produce cytokines for now.)

These cells also produced a lot more IFN-y (interferon gamma), CXCL13 and IL-10.

 

[Snow Leopard]

Activin A secretion is associated with other infections as well, it is not specific to EBV.

Warning speculation about hypothetical pathophysological role of Activin:

Activin A has been known to sensitise afferent nerves (in conjunction with nerve growth factor) and is also involved in neuroregeneration.

https://www.sciencedirect.com/science/article/pii/S0012160604002465
https://www.sciencedirect.com/science/article/abs/pii/S0306452207011827
https://www.sciencedirect.com/science/article/pii/S0012160606010943

https://academic.oup.com/stmcls/article/27/6/1330/6402444
https://www.sciencedirect.com/science/article/abs/pii/S0065242306430043

 

[Hutan]

Abstract of that last paper that @Snow Leopard linked above.

Activin A is a growth factor composed of two βA subunits belonging to the transforming growth factor β (TGF‐β) superfamily of dimeric proteins. The biological activity of activin A is mediated by two different types of receptors, the type I (ARI and ARIB) and the type II receptors (ARII and ARIIB), and by two activin‐binding proteins, follistatin and follistatin‐related gene. These factors bind to activin A and thereby inhibit its biological effects.

Activin A, its receptors, and binding proteins are widely distributed throughout the brain. Studies employing models of acute brain injury strongly implicate enhanced activin A expression as a common response to acute neuronal damage of various origins. Hypoxic/ischemic injury, mechanical irritation, and chemical damage of brain evoke a strong upregulation of activin A. Subsequent experimental studies have shown that activin A has a beneficial role to neuronal recovery and that, by activating different pathways, activin A has robust neuroprotective activities.

Because activin A induction occurs early after brain injury, its measurement may provide a potential biochemical index of the presence, location, and extent of brain injury. This approach may also facilitate the diagnosis of subclinical lesions at stages when monitoring procedures are unable to detect brain lesion and furthermore establish a prognosis.