Hunting down the cause of ME/CFS & other challenging disorders - Lipkin in London
In a talk to patients in London on 3rd September, Dr. W. Ian Lipkin described the extraordinary lengths he and his team are prepared to go to in order to track down the source of an illness, with examples ranging from autism to the strange case of Kawasaki disease.
Discuss the article on the Forums.

Invest in ME/Prof Jonathan Edwards statement on UK Rituximab trial, 30 July

Discussion in 'General ME/CFS News' started by Sasha, Jul 30, 2013.

  1. Jonathan Edwards

    Jonathan Edwards Board Member

    Messages:
    2,066
    Likes:
    8,751
    EBV has been known to mess up B cells to a different state for at least forty years. Once you have had an EBV infection your B cells are never the same again. But the problem is that this happens to more or less all of us (95%+ I think) so EBV messes up your B cells to a state that everyone else has and is completely normal for modern human beings. So this messing up cannot really be 'the cause' of anything. The more subtle issue is whether or not certain diseases specifically require EBV to have been around. That we know to be the case. Certain forms of Burkitt lymphoma and nasopharyngeal carcinoma need EBV. Lymphomas can be cured by removing EBV. The same sort of thing goes for Helicobacter, which most of us have. Certain lymphomas disappear if you remove the Helicobacter.

    I think a lot of people working in this field think that the presence of EBV may 'permit' certain types of ME from occurring, but something else must explain why some people have this sort of ME and most people do not.
     
    Gemini, Countrygirl, Ritto and 7 others like this.
  2. aimossy

    aimossy Senior Member

    Messages:
    3,779
    Likes:
    5,598
    I thought this epigenetics study- although very small adds support to other research. Pathways in epigenetics showing the immune system and metabolic and kinase (protein?) are not working properly? Well that is my take on it, but again its a small study.
     
  3. natasa778

    natasa778 Senior Member

    Messages:
    1,578
    Likes:
    1,575
    London UK
    NK17 likes this.
  4. Jonathan Edwards

    Jonathan Edwards Board Member

    Messages:
    2,066
    Likes:
    8,751
    I think this is difficult to interpret. Taking hospital populations as a basis for epidemiological studies is problematic for various reasons. It also seems to me that they have not excluded a genetic reason - both MS and expression of HIV as AIDS are linked to HLA genes and maybe mutually exclusively. And as they point out it is very difficult to disentangle the potential causal role of HIV itself and of treatment. All in all the idea that the HIV treatment was the reason for a lower incidence of MS looks to me pretty speculative and not the most plausible. I don't think the finding would change my way of thinking about viruses and MS.
     
    MeSci and Valentijn like this.
  5. Bob

    Bob

    Messages:
    10,014
    Likes:
    17,414
    South of England
    Last edited: Aug 31, 2014
    aimossy, rosie26, catly and 1 other person like this.
  6. Sasha

    Sasha Fine, thank you

    Messages:
    10,562
    Likes:
    11,914
    UK
    You've saved me the job of linking to it, so thanks, Bob! I thought it deserved a new thread, given that it's news that even people not following this 1,000-post thread would want to see. No reason not to discuss it here, of course.
     
    aimossy, Simon and Bob like this.
  7. Hip

    Hip Senior Member

    Messages:
    4,249
    Likes:
    4,485
    @Jonathan Edwards

    Could you please expand a little bit further on this.

    I am not familiar with antibody signaling, nor the need for at least two antibodies to bind to an antigen in order for signaling to commence.


    Also, can I ask, is it possible that, even with just one antibody binding to the serotonin molecule, this might disable the molecule, in the sense that this serotonin molecule + antibody might no longer be able to bind to a serotonin receptor and activate the receptor?

    If so, then might these anti-serotonin antibodies effectively cause a serotonin deficit in the brain?
     
  8. Jonathan Edwards

    Jonathan Edwards Board Member

    Messages:
    2,066
    Likes:
    8,751
    In order for an antibody to generate inflammation when it finds, and binds to, its antigen, it has to signal to inflammatory cells through receptors that bind the 'back end' or 'Fc portion' of the antibody. In order to signal when bound to antigen but not when on its own the antibody has to make use of subtle steric/thermodynamic changes rather than just non-covalent chemical binding to a receptor because binding to antigen does not actually change the chemical structure of the back end that binds receptors.

    In practice the antibody takes advantage of a phenomenon known as cross-linking. What that means is that if two antibodies are bound to different faces on an antigen molecule, or groups of antibody molecules form a complex meshwork with antigen molecules (every antigen binding at least two antibodies and and the antibodies binding two antigens), then you get a steric arrangement of antibody back ends (Fcs) all close together.

    The Fc portion signals by binding to one of two systems. If it is outside the circulation (but not in brain) it can generate inflammation through binding a huge multimeric molecule called complement C1q. C1q has six arms and it at least three bind an antibody Fc it changes shape and activates the complement cascade. This can result in binding and activating complement receptors on macrophages.

    The Fc can also bind directly to macrophages through Fc receptors. But again, nothing happens unless several receptors are bound at the same place because of cross-linking of antibodies to an antigen. If three or more Fc receptors (possibly two) are bound by Fcs at the same place a phosphorylating kinase wins out over counteracting phosphatases and intracellular signalling occurs.

    So the interesting thing is that for antigens with only 'one face' or one 'epitope' binding to antibody cannot cause inflammation. As to what is required to have two epitopes, that involves unbelievably complicated stereochemical and thermodynamic issues, but there are good reasons for thinking that you cannot get stable binding of two antibodies to a single molecule that is smaller than 500 Daltons and normally not for less than several kiloDaltons - quite a large peptide. So you cannot get cross linking of C1q or Fc receptors and inflammation.

    What is not clear from the abstract is whether or not in the paper quoted the suggestion is that the autoantibodies are causing the inflammation. The alternative might be that the inflammation is causing autoantibody formation but I don't know of any reason why that should occur.

    One of the problems of clinical immunology is that very few clinicians, or even lab people think through the inflammatory effector pathways. When I moved from working on macrophage physiology to immunology per se I could not find anyone who could explain to me exactly how immune complexes worked. Later I discovered that Keith Peters understood but nobody else had taken much notice. So we had to work out the different inflammatory mechanisms in RA and lupus for ourselves.

    The bottom line is that antibody binding to antigen in itself does nothing. Each antibody has a different shape to its antigen binding site. If it bound 5HT all you would get is a slightly different shape again (the 5HT would appear to be part of the overall antibody shape). That new shape might be recognised but the binding would be so local and the antibodies involved so huge that they would almost certainly pull apart immediately under thermal energy.

    On the other issue. Antibody can mop up a mediator and cause a deficit of the mediator. That is how the anti-TNF antibody infliximab works. However, neurotransmitters are secreted in very tiny synaptic clefts in brain tissue where antibody levels are very low and only stay in the cleft for a few microseconds. Antibodies can certainly block transmission by binding to transmitter receptors but I doubt there would ever be enough antibody for long enough to mop up a transmitter itself. I may be wrong but I doubt it. And there would be no inflammation.
     
    Hip, Valentijn, snowathlete and 2 others like this.
  9. Snow Leopard

    Snow Leopard Senior Member

    Messages:
    2,833
    Likes:
    3,066
    Australia
    I was under the impression that Several kDa is still rather small (less than 30 amino acids), with most secreted peptides being more than ten times this size. I was thinking about this from a 'silent autoimmunity' perspective. The most common examples of secreted small peptides of <5 kDa are neuropeptides, but these all tend to have quite short half lives, on the order of minutes.

    What kind of size restrictions (for multiple antibody binding) are there for the extracellular domain of membrane bound molecules, eg. glycosphingolipids in lipid rafts, such as gangliosides or small receptor proteins of say, 20-30 kDa?
     
  10. Jonathan Edwards

    Jonathan Edwards Board Member

    Messages:
    2,066
    Likes:
    8,751
    You are probably right that to have two functional epitopes you normally need something bigger still but I was erring on the side of generosity. In reality I am not sure that any cast iron rules can be predicted. Normally an antibody binds to an antigen via several amino acids in an extended 'patch'. But I have models of a rheumatoid factor antibody that more or less binds just to one arginine residue.
     
    MeSci likes this.
  11. Hip

    Hip Senior Member

    Messages:
    4,249
    Likes:
    4,485
    @Jonathan Edwards
    Thank you for that detailed explanation of antibody signaling.
     
  12. Jonathan Edwards

    Jonathan Edwards Board Member

    Messages:
    2,066
    Likes:
    8,751
    I realise that one sentence is a bit misleading. It is the neurotransmitters that only stay in the clefts for microseconds, not antibodies (although the result is the same).
     
    MeSci and Hip like this.
  13. Snow Leopard

    Snow Leopard Senior Member

    Messages:
    2,833
    Likes:
    3,066
    Australia
    In terms of the chemistry, I'm sure you could have an antibody bind to a single arginine, but I doubt it would stick on for long when considering the binding kinetics.

    But all else equal, for say, a 20 kDa membrane bound receptor, there should be no problems binding more than one antibody?

    Or am I asking the wrong questions? I know receptors internalise upon binding and some receptors, eg scavenger receptors may internalise more easily since they bind a wider range of ligands. As an example, CD36 is an interesting candidate interesting because it is also involved with fatty acid metabolism, along with regulation of some leukocyte functions, is involved with regulating NO, ROS in some circumstances... The question is, if there were anti-CD36 antibodies, then why wouldn't we see the attack of platelets, muscle cells etc which express CD36?

    Lastly, when you get some time, what do you think about the ideas presented in the following paper? (even if you don't like the ideas, it presents some of the findings of CD36 with respect to antibody production as an example)
    http://www.ncbi.nlm.nih.gov/pubmed/21333356
     
    voner likes this.
  14. Jonathan Edwards

    Jonathan Edwards Board Member

    Messages:
    2,066
    Likes:
    8,751
    My friend Brian Sutton crystallised the complex of the Fab portion of a rheumatoid factor stuck to the Fc portion of an IgG. We then got the separate protein fragments modelled using 3D printing in resin and slotted the models together in the position predicted by the crystal structure. Amazingly there was a hole in the middle of the Fab binding domain into which a single arginine on the Fc popped like a popper catch on a cupboard door. There did not seem to be much scope for further contact. Brian was amazed (steric interactions are his speciality). I think this must be unusual although arginines are probably especially important for interactions because they stick out and are highly charged.

    We need to be careful what we say is binding to what and where. Membrane bound receptors will tend to be present in the membrane in flocks, that can be coralled together into closely packed rafts or 'synapses'. I suspect that the part of a monomeric receptor sticking out of the membrane will tend only to present one epitope to potential anti-receptor antibodies because it would be hard for two antibodies to get their Fab regions into the hemisphere domain available. It might be possible but I have not heard of people talking in those terms. However, if cell A is expressing receptor R then cell B expressing Fc receptors can have many of its receptors find an anti-R IgG stuck to each of many R molecules on A. Whether this would functionally cross link the Fc receptors I am not sure.

    Fc receptors are all monomeric as far as I know, so each only binds one IgG Fc (FcRn may be different but it does not signal as far as I know).

    It is all extremely complicated and difficult to describe without models - which is why I persuaded Brian to get our models made because he could visualise it without but I could not.

    I think that scavenger receptors may well be important for autoimmunity. There has been a longstanding idea that Toll like receptors that can pick up bacterial DNA or nucleic acid binding proteins can provide an extra signal that might overcome self-tolerance thresholds in lupus leading to anti-DNA antibodies etc. We think it is quite likely that the anti-citrullinated peptide antibodies in RA may also be influenced by citrulline acting as an innate immune signal. These theories are hard to prove but I think they are plausible.
     
    catly, Snow Leopard and Valentijn like this.

See more popular forum discussions.

Share This Page