Jonathan Edwards
"Gibberish"
- Messages
- 5,256
I always find your answers very interesting, Prof, Edwards, but they always make me want to ask more questions...how can autoantibodies cause disease in some cases but not in others? You said in the other thread that there is not a direct correlation to the level of antibodies, so is it a binding thing?
One starts off as a medical student with the simple idea that any antibody to self will cause harm. But then you have to think of the effector mechanism. And it turns out there about a dozen major pathways. And it also turns out that none of them will necessarily get engaged, or even if they do, cause any trouble. Just an antibody sticking to something is just an antibody stuck to something. To cause trouble there would need to be cytokine triggering or complement cascade activation or blocking of function or apoptosis or whatever. The autoantibodies I know most about are rheumatoid factors and when I realised I did not really know what to expect them to do I built CADCAM (computer assisted 3D printing) models from crystal structures and started joining them up. It became clear that it all depended on where the antibody bound to its antigen, how many binding sites were available, where complement might engage, where Fc receptors might engage, and on to more questions than one could hope to answer. And you cannot even just answer for one antibody because effects are likely to be dependent on different complementary antibodies binding to different parts of the same antigen. The immune system makes a lot of use of rules about how many sites you need to bind to activate pathways. This is probably a way of stopping individual rogue antibodies causing havoc.
So the simple idea that any antibody that binds to self will cause disease is way off target.