The amino acid sequences of the mouse class I antigens are shown. Residues which are conserved between all clas~ I molecules are specified only on the H-2K d reference sequence at the top of each page, and are left as blank spaces elsewhere. The amino acid sequences of the characterized class I mutants are shown only in those domains bearing amino acid substitutions with respect to their parental alleles. Dashes are used to indicate the absence of codon(s) in certain genes relative to the others to facilitate sequence alignment. Sequences are listed from N-to C-termini starting with the c~l domain. Macintosh discs containing the sequences will be provided upon request.

A large number of H-2K and H-2D cDNA clones from a C3HfB/HeN spleen cDNA library were extensively characterized. All H-2Dk cDNAs were shown to exhibit the short form of exon 8, consistent with the presence of a single lariat branchpoint site within intron 7. Twenty-five H-2Kkm2 cDNAs were found to bear a short exon 8, whereas only two clones were shown to carry the longer form of this exon. In one of the H-2Kkm2 cDNAs, a novel pattern of H-2 splicing was identified, in which an extra 15 nucleotides, derived from the 3' end of intron 5, were inserted between the intact and unaltered exon 5 and exon 6 sequences. Resulting from the apparent use of a cryptic splice acceptor site in place of the canonical intron 5 site, this insertion is predicted to generate an in-frame insertion of five nonpolar amino acid residues within a highly polar region of the intracytoplasmic domain of the H-2K polypeptide. The features of this novel splice form served as the basis for predicting additional rare, alternative H-2 pre-mRNA splicing events that might produce functionally relevant microheterogeneity in the encoded H-2 gene products.