Invest in ME Conference 12: First Class in Every Way
OverTheHills wraps up our series of articles on this year's 12th Invest in ME International Conference (IIMEC12) in London with some reflections on her experience as a patient attending the conference for the first time.
Discuss the article on the Forums.

The role of parvovirus B19 in the pathogenesis of autoimmunity and autoimmune disease (J Kerr)

Discussion in 'Other Health News and Research' started by Dolphin, Dec 12, 2015.

  1. Dolphin

    Dolphin Senior Member

    Messages:
    10,557
    Likes:
    27,561
    Jonathan Kerr was a formerly prominent ME/CFS researcher. From what I can make out, he initially became interested/involved from following a cohort of people who had been infected with parvovirus B19, with some going on to develop a CFS-type syndrome

    http://jcp.bmj.com/content/early/2015/12/07/jclinpath-2015-203455.abstract

     
    Last edited: Dec 12, 2015
    snowathlete, Snow Leopard and SOC like this.
  2. halcyon

    halcyon Senior Member

    Messages:
    2,254
    Likes:
    5,278
    @Jonathan Edwards I'm curious what you think about these "documented mechanisms"?
     
  3. Jonathan Edwards

    Jonathan Edwards Senior Member

    Messages:
    5,020
    Likes:
    29,129
    They are not documented mechanisms. All antibodies can be shown to cross react a bit with all antigens. Demonstrating that a cross reaction actually causes a disease is a completely different matter and over a period of fifty years in almost every case it has got nowhere.

    The list of 'cross reactions' is drearily familiar. People have been banging on about collagen II since the 1970s because it was suggested to be a mechanism in rheumatoid arthritis. What the theorists did not bother to realise is that RA is a disease of synovium, not cartilage - the two just happen to be close together in certain joints (but not in tendon sheaths where there is synovium and RA or the nose where there is cartilage and no RA). There are still researchers using collagen II arthritis in mice as a model of RA. Keratin pops up because it is very easy to get non-specific reactivity to it. Myelin basic protein is what everybody wanted the antigen in MS to be but the antibodies turned out mostly not to bind to it. Cardiolipin is a mush of antigens that shows low grade cross reactivity in syphilis (but nobody thinks the cross reactivity has anything to do with the disease) and turns out to contain autoantigens recognised in lupus and anti-phospholipid syndrome. It is one of the most notorious tests for getting false positives.

    Sorry, but to me this is classic immunobabble - a pot-pourri of all the same unproductive ideas that have been going round and round in immunology since I was a student.

    Even in rheumatoid arthritis nobody has actually proven that rheumatoid factors cause the disease and by and large post parvo arthritis is RF negative - so there would have to be two causes in parallel. It just makes no sense.
     
    MeSci, voner, Scarecrow and 3 others like this.
  4. msf

    msf Senior Member

    Messages:
    3,150
    Likes:
    4,341
    Prof. Edwards, I´ve been meaning to ask you, what do you make of the claim that those cross-reactions are much more common than would be expected if they hadn´t evolved to cross-react?

    Is it true? If so, is it just one of God´s jokes?
     
  5. msf

    msf Senior Member

    Messages:
    3,150
    Likes:
    4,341
    To my way of thinking, it would be incredibly surprising if nature had only evolved mimicry on a macroscopic level. Of course, that doesn´t necessarily mean that it is necessarily a significant factor in human disease, but I think the possibility is definitely there.
     
  6. Jonathan Edwards

    Jonathan Edwards Senior Member

    Messages:
    5,020
    Likes:
    29,129
    I think you are confusing antibody cross reactivity with microbes mimicking human receptors or ligands. The latter goes on all the time and is an evolutionary strategy. But mimicry by antibodies is just a random chance fact of stereochemistry. It would be entirely disadvantageous if it did occur and as far as we know it hardly ever does. Remember that a cross reacting antibody is not something encoded in genes - it arises from a random shuffling of heavy and light chain variable genes - with something like ten to the power fourteen possible combinations.
     
    MeSci likes this.
  7. msf

    msf Senior Member

    Messages:
    3,150
    Likes:
    4,341
    But who would it be disadvantageous for? And if microbes mimic human receptors and ligands, wouldn´t it be possible for antibodies to the former to affect the latter (because they have a high avidity for the former but a low avidity for the latter, or by some other possible?) I know the generation of antibodies is random, but the clonal expansion isn´t, is it?
     
  8. Jonathan Edwards

    Jonathan Edwards Senior Member

    Messages:
    5,020
    Likes:
    29,129
    It would be disadvantageous for the host.

    It is possible for antibodies to microbes to bind to host proteins but the whole point about clonal selection is to avoid that and as far as we know the system avoids high affinity cross reactivity very well. And only high affinity cross reactivity is relevant because low affinity binding does not activate effector mechanism like complement and Fc receptors. There might be an exception to this with FcRI that I have mentioned but that is pure speculation.

    So we would expect hosts to continue to evolve in the direction of clean clonal selection. I am unclear that any of this has much interest to the microbe. When post-infective illnesses do occur, because of microbes confusing innate immune signals - like in rheumatic fever - it is unclear that thes help the microbe. Giving someone chronic heart disease is of no great interest to a streptococcus that reproduces by spreading through coughs and sneezes. If anything killing off hosts seems bad tactics.

    I am not quite sure what you are driving at to be honest.
     
    MeSci likes this.
  9. msf

    msf Senior Member

    Messages:
    3,150
    Likes:
    4,341
    Rather topically: http://www.ncbi.nlm.nih.gov/pubmed/23463689

    Here is one possible model: Pathogen proteins evolve to mimic some host proteins to evade the immune response. Host proteins evolve to counter this. Result: antibodies to the former have a low avidity to the latter.

    Here the autoimmunity is a result of the push-pull of evolution.
     
    Last edited: Dec 13, 2015
  10. Jonathan Edwards

    Jonathan Edwards Senior Member

    Messages:
    5,020
    Likes:
    29,129
    But we haven't got any evidence of any autoimmunity based on cross reactivity.
    I cannot see the connection. Antibodies recognise shapes generated by a few amino acids in a particular twisted conformation. For any protein there are hundreds of bumps and dents and ridges that could be recognised in different overlapping ways. If one protein in a microbe has a small functional patch that resembles a small patch on a receptor that tells us nothing about the likelihood of antibodies cross reacting with all the other bumps and dents. It just not work at that scale. Antibodies have no preference for a 'business end' of a molecule. If a molecule is the shape of a teapot with a business end spout you can still have antibodies binding to the handle or the knob on the lid or the base. Statistically I cannot see any connection between your first two proposals and the third I am afraid.
     
  11. msf

    msf Senior Member

    Messages:
    3,150
    Likes:
    4,341
    I wasn´t talking about evidence, I was talking about the possibility.

    Do people only talk about cross-reactivity in reference to the ´business end´ of a molecule? I don´t see how it is necessary for the antibody to be targeting the business end for the model I am suggesting to work. Perhaps I´m being a little dense, but your last post seems to be a bit beside the point.
     
  12. msf

    msf Senior Member

    Messages:
    3,150
    Likes:
    4,341
    As for statistics, it seems to me that the model I am proposing (I´m sure someone has proposed it before) could explain the unexpectedly high number of pathogen epitopes that mimic host epitopes.
     
  13. Jonathan Edwards

    Jonathan Edwards Senior Member

    Messages:
    5,020
    Likes:
    29,129
    That was my point. Antibodies can bind to any part of the molecule. So although your second point:

    is perfectly valid I don't think it has anything to do with antibodies or cross-reactivity. Except that what you are referring to are local conformations and not epitopes because an epitope is by definition something bound to by an antibody and we have no evidence for an unexpectedly high number of pathogen epitopes mimicking host epitopes.
     
  14. msf

    msf Senior Member

    Messages:
    3,150
    Likes:
    4,341
    I still don´t see the relevance of your first point for my argument. According to the Wikipedia article on molecular mimicry, it is not possible to search for mimicry of conformational epitopes because of their 3D structure, so if that is the case it´s not very surprising that we don´t have evidence for an unexpectedly high number of pathogen conformational epitopes mimicking host epitopes. I realise that I have a model that explains something that there is no evidence for, but I was arguing more for the possibilty of molecular mimicry autoimmunity than for its existence in humans.
     
  15. SOC

    SOC

    Messages:
    7,802
    Likes:
    16,454
    Absence of evidence is not evidence of absence.
     
  16. Jonathan Edwards

    Jonathan Edwards Senior Member

    Messages:
    5,020
    Likes:
    29,129
    It is perfectly possible to search for mimicry of conformational epitopes - it is called cross reactivity of sera. To pin down specific epitopes is hard but involves cross-competition assays.

    I think you have got the wrong end of the stick. Molecular mimicry is not a theory about pathogens mimicking host proteins. The 'mimicry' in the theory is sort of metaphorical. If anything it is the host protein mimicking the pathogen that has come in and stimulated antibody. But since it does not actually do that, because the antibody producing clones are deleted (because the host protein also binds) there is no mimicry. The presence or absence of mimicry is not a function of similarity of pathogens to host but of the efficacy of the clonal deletion system. Lots of antibody species will bind very tightly to almost anything - the level of similarity can be trivial. But these antibodies normally get deleted straight away in bone marrow.
     
  17. msf

    msf Senior Member

    Messages:
    3,150
    Likes:
    4,341
    No, I meant it´s not possible to search a database for them, so it´s hard to get an accurate idea of their relative frequency in nature.

    How do we know that the clonal deletion system gets rid of low-affinity antibodies (which according to the article I posted a link for above, can still sometimes actívate complement)? Do we have any evidence for how it does this? I.e. is there any evidence that antibodies that bind to self antigens are deleted because they actívate complement? When I looked it up, there seem to be several ways in which tolerance to self is produced - why would there need to be many different ways if the clonal deletion worked so well?
     
  18. M Paine

    M Paine Senior Member

    Messages:
    296
    Likes:
    816
    Auckland, New Zealand
    Theoretically, it is currently possible to do computational searches based on DNA/RNA databases, it's just not practical.

    The basic rules of protein folding are known, the 3D structure and electrostatic charge of each atom of each amino acid is known, the 3D structure of proteins can be predicted from a DNA or RNA sequence. So it's possible that someone could look at this in silico.

    Anyone who has studied your average viral genome will be quick to point at things like irregular codon usage, protease activity or other post translational activity which might alter the shape or size of the protein, such as other proteins marshaling the folding process.

    In any case, it's possible (to a limited degree) to do a purely computer generated search for cross-reactivity, but it's not practical due to the gargantuan computational requirements exceeding today's capabilities, and not currently having a fit for purpose protein structure prediction algorithm.

    Interesting thought.
     
  19. msf

    msf Senior Member

    Messages:
    3,150
    Likes:
    4,341
    Ok, not not possible, not practical.
     
  20. Jonathan Edwards

    Jonathan Edwards Senior Member

    Messages:
    5,020
    Likes:
    29,129

    I may be wrong but I think there is a continuing misunderstanding in this conversation that cross reactivity is a property of a microbial protein and a host protein. It is not. It is a property of an antibody species that happens to bind to both - the two proteins bound to do not actually need to be similar in shape, they just need to have a complementary charge pattern to the antibody. An antibody with two arginines sticking out in a certain relation to some lysines and methionines or whatever might happen to lock electrostatically to two proteins with positive and negative charges in similar relations but the actual contour of the protein can be different. Brian Sutton and I played around with 3D printed models of rheumatoid factor antibodies and their antigens years ago. We could show how the models interlocked, but we had no way of predicting actual dissociation constant. Using protein structure software you can see if two proteins are similar in contour in particular places but there are so many ways of turning the protein around that you what be at it for weeks just for two proteins. And although in theory you can pick out likely charge complementarities the would mediate binding I very much duoubt even now that anyone could predict dissociation constant.
     
    voner and Scarecrow like this.

See more popular forum discussions.

Share This Page