Peer, for what it’s worth (with the very limited examples at hand) the differential immunodominance associated with route of infection hasn’t involved viral genes that are particularly differentially expressed (though that’s also a field that could be much tidier than it is). As a more general statement, though, I agree that antigen availability is a likely cause, but I lean more toward differential antigen processing in different tissues (since some of the proteases involved in epitope generation do have different expression in different tissues). But there’s really little or no data to support that yet.

But I think that as a general statement we don’t understand relative importance of T cell vs. APC (and other factors) in establishing immunodominance hierarchies.

Assuming T-cells (and maybe B-cells, too) are isolated in your tissues to prevent too much diversity of TCRs in your body, and assuming this is a mechanism that prevents autoimmune disease, wouldn’t we expect to see some cases of physical damage to a tissue leading to autoimmune disease? (I guess this would work sort-of like damage to an eyeball leading to autoimmune attack on the other eyeball, only in this case, the injury and the autoimmune attack would be for different kinds of tissue.)