Ian A. York wrote:
>> In article <362FBCED.1958 at mail.tju.edu>,
> mark <mark.haynes at mail.tju.edu> wrote:
> >
> >I don't really think you need a system that can differentiate, just a
> >system
> >that deletes reactivity to peptides that find there way to the thymus and
> >presented via class i or class ii to cd8 or cd4 thymocytes respectively.
> >Everything else is non-self.
> If "everything else is non-self", then it's differentiating between
> self-and non-self. You've rephrased my point, is all.
Ian, my point relates to the degree of deletion of self. If every possible
'self' peptide that could be produced were presented and led to clone
deletion the cost would be higher than if a subset were processed. I think
this is an important distinction vis a vis tolerance and MHC restriction. I
think the self-nonself idea was a very helpful and reasonable scenario but
the definition of self can be cumbersome especially if most of the proteins
in existence are somehow related.
> >Me neither, but I would propose that it would just be shifted to APCs and
> >not
> >every cell.
>> For MHC class I, every cell *is* an APC. So, again, you've re-phrased my
> point.
I disagree with this definition. I would accept that all MHC class I+ cells
could be a target for an effector T cell, but I reserve the term APC for
activities related to stimulation of virgin or 'memory' t cells that express
the secondary signally machinery (cd80, 86). Can there be promiscuous
presentation--sure, but that probably relates to autoimmunity which might be
called "inappropriate relations".
markH
