Antibodies to ß adrenergic and muscarinic cholinergic receptors in patients with CFS

[adreno]

Alpha-2 adrenoceptors are also heavily involved in cognition - agonists are used for ADHD - so if they are screwed up it could have a huge impact. It would be nice to know.

 

[Sidereal]

All this discussion is over my head but Alan Light is giving a Solve webinar in a couple of weeks - would it be helpful to email him with any relevant questions before the talk?

No because he wasn't involved in the study being discussed on this thread so he wouldn't be able to answer.

 

[Sasha]

No because he wasn't involved in the study being discussed on this thread so he wouldn't be able to answer.

I realise he couldn't answer questions about this study but wondered if he might have comments on the relevance to his own work - but maybe I'm not making sense because I've just jumped into this thread at the end.

 

[MeSci]

I don't want to go too far off-topic but for the measurement to be valid it has to be standing still, no moving or even fidgeting. I was surprised by how hard it was the first time I tried it. Only lasted a few minutes before getting woozy. I wouldn't recommend doing this alone because the suddenness of the reaction may take you by surprise and you could have a syncopal episode or just get unsteady and fall and injure yourself before you've had a chance to sit/lie down.

I have stood in queues for 30 mins at times and not felt woozy. It's just my legs that get tired. I'm quite active. It's only recently I've been getting a bit woozy/dizzy (apart from my hyponatraemia caused by ACE inhibitors), and I think that is due to bouts of migraine.

 

[sdmcvicar]

How exactly would you identify the relevant binding regions? I mean in a typical aptamer experiment, the aptamers would stick anywhere they felt like (okay, based on stochastic effects) on the target ligand.

Sorry for the slow reply Snow Leopard. Didn't even hop on yesterday. Thanks for the more complete overview of agonist and antagonist types.

My impression after talking to BasePair Bio, which is close by me in Houston, is that they can do a kind of multiplexed
SELEX, where they would counter-select anything that binds to control antibodies, and can also select for or counterselect against variants which interfere with the autoantibody binding to a receptor. This would remove a lot of the aptamer binding outside of the variable region, and help isolate epitope regions. I know they've had projects that only take a few weeks, for example, that build aptamers specific for phosphorylation sites on proteins that don't have crystal structures, so I think it's technically possible. I may just be integrating several techniques in my head, though.

On a related note for both yourself and alex3619, if any of the receptors can be produced in micelles, or if the soluble portion of these receptors can be recombinantly expressed, the binding constant of the autoantibodies to receptors (also aptamers to either) can be determined using microscale thermophoresis, which is used by BasePair for aptamer validation and on a contract basis. I had a chance to demo this technology a couple of years ago when analysing affinity of a peptide to isolated cell membrane and to antibodies raised to the peptide, and it actually worked really well.

 

[Simon]

I’m still struggling to get to grips with this paper – thoughts so far. As is my wont, I've focused on the things that don't quite make sense to me. Answers/explanations appreciated

Some comments/highlights (1/2)​

1. Evidence for a causal role for autoantibodies, from Rituximab patients...
This seems to be the stand out finding:

Remarkably, elevated autoantibody levels had normalized in the majority of clinical responder post-treatment. In responder posttreatment levels for all four antibodies were significantly lower compared to pre-treatment [vs most autoantibody levels unchanged in non-responders]....This finding provides further evidence for a possible pathogenic role of these autoantibodies in CFS

It is conceivable that various symptoms of CFS including cognitive deficits, autonomic dysregulation...could be partly mediated by autoantibodies against these receptors in a subset of patients

Other evidence that would support such a causal role for autoantibodies is correlation between symptom levels and autoantibody levels. And the authors looked for just this in the much larger sample of Berlin patients (n=268) - but found none:

...but no correlation of autoantibodies with symptoms
"The only association we observed was of M1 AChR antibodies with dizziness (p=0.05)" – which wouldn’t survive an appropriate statistical correction for so many comparisons (lots of symptoms v lots of antibodies, greatly increasing the chance of a false positive).

Also note

Our observation of a decrease of M and ß receptor autoantibodies in patients responding to rituximab, in whom levels pretreatment were within the normal range of control subjects, suggests that we may miss functionally pathogenic antibodies by just assessing quantitative levels by ELISA [standard test for antibodies].

2. Overall, higher levels of autoantibodies in patients than controls

We found significantly elevated levels of antibodies against ß2 AdR, M3 AChR and M4 AChR in a subset [edit: 29.5%] of patients with CFS.

Some controls had elevated antibodies as well:

Similar to our findings, in most other diseases autoantibodies are found in only a subset of patients with a wide overlap of levels in patients and controls.

This could be in part, as they argue, that it depends on the nature of the autoantibodies – do they have a biological effect or not. But it seems these findings for mecfs patients are less conclusive than those for patients with known autoimmne diseases:

Is this true for all conditions? My impression was certain autoantibodies were very common in some conditions.

I think it is relative - and I think they are stretching the analogy a bit. As you say, in a lot of autoimmune conditions we have antibody tests that will pick up 60-95% of patients. There are some more muddly ones like vasculitis, but not so many.

Maybe this a subset effect, with these mecfs patients actually including several different diseases, weakening the signal, but it does muddy the water.

continued...​

 

[Simon]

...continued

Some comments/highlights (2/2)​

3. Evidence for activation of the immune system
The authors looked for evidence linking the autoantibodies to three different measures of immune activation:

1. Overall antibody (immunogloblulin) levels (higher levels indicating immune activation)
   
2. T cell activation
3. Increased Anti-nuclear antibodies (ANA)
   
4. Antibodies to thyroglobulin and thyroperoxidase (TPO/TG)

Immunoglobuline levels

We observed a remarkably strong correlation of higher autoantibodies for all M and Beta subtypes with higher IgG1-3 sublcasses, but not IgG4

Which looks pretty interesting, but there are a couple of caveats:

• The correlation was typically 0.25 (0-1.0 scale), which is moderate, not high (elsewhere they found correlations of over 0.7, which are very high).
  
• Also, while 27% of patients had high levels of Ig, another 20% had low levels, which complicates things; "The most prominent finding was a single or concomitant IgG3 (10.1%) deficiency in CFS patients". We don't have a comparison figure for controls. They counted levels as elevated if ANY of 6 (maybe 7) Ig classes/subclasses were elevated so it wasn't necessarily a hard threshold to reach.

So I'm not sure how much importance to give to this finding of elevated antibody levels (that's total antibody, to anything, not specific autoantibodies measured in this study)

Other types of immune activation
Patients with high had significantly more activation according to the three other measures (activated T cells, elevated ANA and TPO/TG antibodies) - as well as the immunoglobulin correlation above.

There was also an association between M1 AChR antibodies and both ANA and elevated TPO/TG antibodies. However, the overall differences between patients and controls was not significant for M1, only for type M3 and M4 receptors, which didn't correlate with these other measures of immune activation.

So the best evidence seems to be for a link between ß2 antibodies and immune activation, with less impressive evidence for the other antibodies.

we observed a significant correlation of levels of ß AdR and M AChR antibodies with immunoglobulin levels, T cell activation, and elevated ANA and TPO antibodies. This fits well with findings in other autoimmune diseases. T cell activation is frequently present in autoimmune disease, as well as elevated ANA titres and TPO antibodies... Thus a direct effect of autoantibodies on the observed skewed immune parameters is conceivable.

How antibodies could activate the immune system
It turns out that B cells express ß2 Adrenergic receptors, which can play a role in regulation:

The effect of adrenergic stimulation on immune cells is complex depending on the activation status of the immune cell and type of stimulant and both immune stimulation and suppression were reported... Thus, the association of autoantibodies with enhanced IgG levels we observed suggests agonistic effects of ß2 AdR antibodies on B cell receptors.

crucially, this is testable:

In vitro functional assays in CFS patients are required to clarify the direct effect of these autoantibodies on immune cell function.

Hopefully this will be the next step for the authors. If the confirm a direct effect of these autoantibodies on immune function in mecfs, it would be a breakthrough.


4. Authors' conclusion
Worth listing in full

In conclusion, there is evidence for elevated autoantibodies against ß2 AdR and M AChR in a subset of patients with CFS.

Although the function of these antibodies in CFS at present is unclear, the association of ß2 AdR and M AChR antibodies with immune activation markers and their decline in CFS patients responding to B-cell depletion may support a pathogenic role and warrants their testing as potential biomarkers in clinical trials of B-cell/ antibody depleting therapy.

It is conceivable that various symptoms of CFS including cognitive deficits, autonomic dysregulation and immune activation could be partly mediated by autoantibodies against these receptors in a subset of patients.

 

[Jonathan Edwards]

Autoantibody levels do not tend to correlate with clinical features in any autoimmune disease. This is presumably because only a subset of antibodies to an autoantigen are pathogenic and these are not reliably represented by binding in an in vitro assay.

 

[Sidereal]

We found significantly elevated levels of antibodies against ß2 AdR, M3 AChR and M4 AChR in a subset of patients with CFS. [they don’t give the size of the subset, oddly]

The figure appears in the abstract

We provide evidence that 29.5% of patients with CFS had elevated antibodies against one or more M acetylcholine and beta adrenergic receptors which are potential biomarkers for response to B-cell depleting therapy.

but not in the results section.

 

[Simon]

Autoantibody levels do not tend to correlate with clinical features in any autoimmune disease. This is presumably because only a subset of antibodies to an autoantigen are pathogenic and these are not reliably represented by binding in an in vitro assay.

Interesting. I take the point about the importance of pathogenic antibodies as opposed to autoantibodies in general. But there are generally higher levels of autoantibodies in autoimmune diseases, as you mentioned before (and in this study too) so overall levels seem to play a role too. Could you explain a bit more about this?

 

[BurnA]

Are the elevated autoantibodies found in this study the same as those found in the 2014 POTS study http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3259269/

Are there any similarities ?

 

[BurnA]

Oops meant this study http://m.jaha.ahajournals.org/content/3/1/e000755.full

 

[jimells]

A 3-month lead time would probably be enough to generate an appropriate screening assay, and a month would be plenty of time to run patient samples in the thousands, but grant approval, protocol approval, patient enrollment and curation of patient clinical data for each sample would likely add two years as an optimistic estimate.

It is immensely frustrating to think that critical research is subject to so much delay by bureaucratic gatekeepers more interested in the rules and protecting their careers than in the results.

I noticed how all those barriers rapidly dissolved when it was time to disprove XMRV, but immediately re-appeared once that job was finished. Funny how that works.

 

[alex3619]

1. Evidence for a causal role for autoantibodies, from Rituximab patients...

This is actually very weak evidence. Its association not causation. Even presuming we have an autoantibody issue, there remains a range of autoantibodies that have yet to be found. The actual causative autoantibody, presuming there is one, could be one of the ones not found yet. Which might explain why the antibodies found in this study were only found in less than half of the responders.

So the current antibody findings might explain disease causation in a subset, or be contributory, or be irrelevant. While these antibodies are high in nonresponders, we have to ask the question as to what other antibodies, not discovered yet, are also high in nonresponders. How can we be sure which specific antibodies are causative, if any?

That is why its so important to continue studying this.

if any of the receptors can be produced in micelles, or if the soluble portion of these receptors can be recombinantly expressed, the binding constant of the autoantibodies to receptors (also aptamers to either) can be determined

Yes, and this would be good to know. However we also need a test platform to see what is happening to the activity of these receptors. I could even make a case for genetically engineering bacteria to express these kinds of receptors, though I think human tissue could probably be grown in enough quantity and would be experimentally superior. However a recombinant or stimulated human tissue that overexpresses these receptors might have some limited value.

We might have a need for tissue cultures in which the appropriate receptors are active to use as a test bed.

 

[alex3619]

I noticed how all those barriers rapidly dissolved when it was time to disprove XMRV, but immediately re-appeared once that job was finished. Funny how that works.

I think many of us noticed this. Furthermore they were willing to accept low quality research against the XMRV finding, rather than critically appraise it. XMRV eventually had very high quality research against it, but that took time. There seems to be not double but many layers of standards involved here.

Yet, somehow, extremely bad studies in psychiatry go unchallenged. This is way beyond double standards.

 

[anciendaze]

Please note that this paper did not stratify their cohort w.r.t. time, e.g. according to time since onset. If these autoantibodies change over time, the way cytokines do, it is entirely possible that lumping patients at different stages together will cause statistical correlations to vanish. What little evidence we have shows that immune response does change over time with immune exhaustion setting in after a period of several years. This is long enough for antibody response to decline as well. Cytokine levels also change in response to exercise, and we have no idea where in such a process these samples were taken. We really need to get some measures of the altered dynamics of patient immune systems.

Frankly, I am not impressed with the state of the art in the treatment of known autoimmune diseases. Too many patients have told me the treatment they experienced would be worse than a disease which stops short of putting me in the hospital. They "tough it out" until this becomes impossible. Except for some MS patients, or RA patients with joint damage, most do not experience setbacks lasting several days after exercise. This is a very distinctive characteristic of this illness, and it is clearly a dynamic change.

 

[BurnA]

This is actually very weak evidence. Its association not causation. Even presuming we have an autoantibody issue, there remains a range of autoantibodies that have yet to be found. The actual causative autoantibody, presuming there is one, could be one of the ones not found yet. Which might explain why the antibodies found in this study were only found in less than half of the responders.

So the current antibody findings might explain disease causation in a subset, or be contributory, or be irrelevant. While these antibodies are high in nonresponders, we have to ask the question as to what other antibodies, not discovered yet, are also high in nonresponders. How can we be sure which specific antibodies are causative, if any?

That is why its so important to continue studying this.

Does this paper tell us anything worthwhile ?

 

[jimells]

There is always a reason. "No reason" is just shorthand for saying "I don't know but I do not want to admit ignorance". No known reason would be a better claim.

An honest response might be, "There's not much interest in finding out, 'cause repeat business is easy and profitable. Cures, not so much."

 

[BurnA]

Yet, somehow, extremely bad studies in psychiatry go unchallenged. This is way beyond double standards.

How about a rotten tomatoes type website for scientific studies where members of the scientific community can rank or critique academic papers based on certain criteria?

I'm surprised it doesn't already exist.!

 

[snowathlete]

The rituximab trials suggest more than 29.5% of patients benefit from it, so that woudl seem to suggest there are more auto antibodies to be identified, which could represent further subsets, or even all responders. How do researchers go about finding these? I am guessing there are other known auto antibodies that we can test against with existing tests, and lets hope we find something more with those, but presumably there may be some novel auto antibodies that need to be found. How do we do that?