SPECIES SPECIFICITY IN THE INTERACTION OF CD8 WITH THE ALPHA-3 DOMAIN OF MHC CLASS-I MOLECULES

The alpha-1 and alpha-2 domains of the class I MHC molecule constitute the putative binding site for processed peptides and the TCR, although the alpha-3 domain has been implicated as a binding site for the CD8 molecule. Species specificity in the binding of CD8 to the alpha-3 domain has been suggested as an explanation for the low xenogeneic T cell response to class I molecules, but results on this point have been conflicting and controversial. We have addressed this issue using CTL lines from HLA-A2.1 transgenic mice that specifically recognize and lyse A2.1-expressing cells infected with influenza A/PR/8 or pulsed with influenza matrix peptide M1(57-68). Species specificity was examined using transfectants that expressed hybrid molecules containing the alpha-1 and alpha-2 domains from HLA-A2.1 and the alpha-3 domain from a murine class I molecule. Lower levels of M1(57-68) peptide were required to sensitize L cell transfectants expressing a chimera that contained an H-2D(d) alpha-3 domain than targets expressing the intact A2.1 molecule. However, at high doses of peptide, lysis of these two targets was similar. However, no reproducible difference in sensitization was observed using EL4 or Jurkat transfectants expressing A2.1 or A2.1 chimeric molecules that contained an H-2K(b) alpha-3 domain. In all cases, however, lysis of peptide-pulsed A2.1 expressing targets was more sensitive to inhibition with anti-CD8 mAb than lysis of cells expressing these chimeric molecules. Thus, under suboptimal conditions such as low Ag density or in the presence of anti-CD8 mAb, these CTL preferentially recognize class I molecules with a murine alpha-3 domain. This suggests that there is some species specificity in the interaction of CD8 with the alpha-3 domain of the class I molecule. However, CTL recognition was inhibited by point mutations in the alpha-3 domain of HLA-A2.1 that have been shown to inhibit binding of human CD8 and recognition by human CTL, suggesting that murine CD8 interacts to some degree with human alpha-3 domains, and that similar alpha-3 domain residues may be important for murine and human CD8 binding. The relevance of these results to an understanding of low xenogeneic responses is discussed.