MHC restriction

Ken Miyasaki kmiyasak at
Mon Mar 3 15:04:48 EST 1997

I just wanted to thank everyone for their thoughtful comments. I still
need to do some reading, but at least I have a feel for what
immunologists think, which is important to me. I have some responses,
below, and again, would appreciate any corrections or other feedback...

Derek Gray <derek.gray at>

Derek: Thanks for the comments regarding allograft rejection. Is there
an easily obtained citation for the observation that in allograft
rejection, the antigen carried by the MHC is irrelevant?  From a
theoretical perspective, it still seems rather contradictory that MHC
retriction (ie., lack of 'response' due to MHC pleomorphism) and
allograft rejection (ie., 'awesome response' due to MHC pleomorphism)
should be used to explain one-another.

Jeffrey Frelinger <jfrelin at

Jeff: Thanks for your comments about the TCR not seeing antigen at all.
I think you meant that it never sees soluble antigen in a measurable way
(and I never felt this to be the case) but the TCR, with it's variable
domains, must be somewhat interested in antigen otherwise, why not
design it with variability more like the killer inhibitory receptor
(KIR, NKAT gene products)? 

I know the TCR doesn't work like antibody. It's designed to work as part
of a cell, rather than as a soluble product. Thus, I find it totally
harmonious that you need 4000-8000 TCRs [out of 5000-40000 total TCRs]
downregulated to get a forward (proliferative) T-cell response and only
100 out of 100,000 MHC molecules loaded with antigen to get a forward
(proliferative) T-cell response. It must take advantage of it's rather
low affinity to accomplish 4000-8000 downregulations over time. By the
way, I do like pictures and I have cut that Garcia picture out and
scanned it already! I'm not quite certain if we are even arguing about
anything, but I just wanted to respond.

Ian York <iayork at

Ian: Thanks for your comments regarding crystal structures, I will
certainly check out the articles to which you referred (although I do
have that Garcia clip-out which includes Wilson as the senior author).
In that Garcia article, the authors conclude that CDR3 of both the alpha
chain and the beta chain of the TCR contact peptide alone. They note
that CDR3 is the most variable hypervariable domain of the TCR (and I
attach significance to that). The authors also conclude that CDR1 and
CDR2 of both chains contacts both peptide and MHC alpha helical domains.
As you know, the alpha helical regions of the MHC are rather limited in
their allelic variability (most pleomorphism occurs where the TCR should
not see it that easily -- the beta-sheet floor, especially near the
specificity pockets). So, my thought is this: how can allelic
variability (pleomorphism) lead to MHC restriction most easily? My
answer would be by determinant selection rather than MHC

"To summarize, I think you are confused about TCR recognition.  I'm not
sure what you mean by your statement, but it's wrong."

Joerg Kirberg <kirberg at>

Joerg: Clearly, I need to read those papers! Especially Matzinger...
Thanks!!!!! (Especially about how anything happens in the absence of
antigen... Since MHC molecules are always loaded with something at 37°
C, I need to see whether or not I believe those papers you cite).

More information about the Immuno mailing list