In article <37593E22.339D at post.queensu.ca>,
D. R. Forsdyke <forsdyke at post.queensu.ca> wrote:
>Kenneth Frauwirth wrote:
>>> intracellular self/non-self discrimination. However, having some
>> experience with the antigen processing field, my take on things is that
>> the biochemical evidence strongly suggests that non-self peptides are
>> *not* preferentially displayed. Peptides can be extracted from cells and
>> analyzed, either by mass spec/sequencing or by T cell recognition, and
>> these methods indicate that virally expressed antigens are not presented
>> in especially high copy number on MHC Class I;
>> Yes, it is easy to get hung-up on this point. Your logic is that, if
>there is intracellular S/NS discrimination (IC-[S/NS]-D)then there
>SHOULD be preferential display of NS with MHC. However, this does not
>follow. The initial IC-[S/NS]-D acts as an alarm (present even in
>unicellular organisms) with multiple secondary effects such as
>increasing MHC gene transcription and MHC protein expression (in
>multicellular organisms). The actual presentation with MHC of foreign
>peptide (derived from NS) is a bonus.
There is no doubt that pathogens initiate signals that result in such
effects as upregulation of MHC, TNF, IFN's, etc. - these are in fact the
original basis of the "danger" model. Bacterial cell wall components are
recognized by macrophage and B cell (and DC?) surface receptors; yeast
cell wall components can trigger the complement system; dsRNA from viruses
(which often transcribe both strands of their DNA) activates the NF-kB
pathway. However, none of these is usually referred to as "self/non-self"
discrimination (and perhaps that is a shortfall of the nomenclature); I
believe Janeway calls it "pattern recognition". It is certainly a form of
self/non-self discrimination, but relies on non-protein differences
between host and pathogen.
I suggest that transgene expression and tumorigenesis provide the
strongest arguments against intracellular self/non-self discrimination at
the protein level. Many tumors express mutant proteins, fusion proteins,
or other factors inappropriate for the original cell type, but the immune
system fails to reject them (and often becomes tolerant to those same
tumor antigens); likewise, many tissue-specific transgenes do not cause
autoimmunity in the transgenic tissue, even when T cells specific for the
transgenic antigen are present, although this isn't universal (in fact, I
find it difficult to see any pattern to the transgenes and expression
patterns which do or don't cause autoimmunity).
That is certainly not to say that it *doesn't* happen, of course; only
that there are other *known* mechanisms which potently trigger the same
effects, and phenomena (spontaneous autoimmunity and graft rejection)
which still cannot be explained by it (or, of course, by current models).
Ken Frauwirth
--
Ken Frauwirth (MiSTie #33025) kfrauwir at midway.uchicago.edu
Gwen Knapp Center for Lupus and Immunology Research
University of Chicago
http://www.ocf.berkeley.edu/~frauwirt
