Philip Fung <phfung at NOSPAM4ME!owlnet.rice.edu> wrote in message
news:7vvlbb$n18$1 at joe.rice.edu...
> I may be wrong on this one, but I'm pretty sure that all (or nearly all)
> nucleated
> cells have MHC Is, right?
I agree on that. All nucleated cells express class I MHC. Therefore all
nucleated cells are potentially vulnerable to killing by _activated_ CD8+
T-cells. But in order to become _activated_, _naive_ T-cells need
co-stimulation by B7.1 or B7.2, which is present only on professional APCs.
It is here that the problem lies: In order to raise activated CTLs an APC
with viral antigen on class I molecules is required. Traditionally class I
MHC was thought only to present endogenous peptides (i.e produced in the
cell). If this was true a virus would have been able to avoid raising CTLs
by avoiding to infect APCs.
In the article refered to by Ian York earlier in this thread it is suggested
that uninfected APCs are able to take up viral antigen and present on class
I molecules via and _exogenous_ pathway. This effectively raises activated
CTLs. (and answers my original question)
>> Activation doesn't have to happen in the lymphoid tissue, either.
Probably not, but if there weren't these central meeting points were naive
T-cells can scan antigens, I think there would be too little a chance that a
particular T-cell receptor would find its corresponding antigen.
/Morten
