Distinctive polymorphism at the HLA-C locus: implications for the expression of HLA-C.

The HLA-C locus remains an enigma. The serological polymorphism is poorly defined, HLA-C molecules are expressed at the cell surface at about 10% the levels of HLA-A and -B, and their importance for antigen presentation to either CD8-bearing T cells or natural killer cells is unclear. Our understanding of HLA-C polymorphism has also lagged behind that of HLA-A and -B. We have applied the polymerase chain reaction to the characterization of cDNA encoding HLA-C antigens. Combining the recent results with previously characterized HLA-C alleles gives a data base of 26 sequences, which was used to analyze the nature of HLA-C polymorphism and compare it to the variation seen in HLA-A and -B. The sequences form 10 families of alleles that correlate well with the patterns of serological crossreactivity, including the C blanks, and all major HLA-C allelic families appear to have been sampled. The families further divide into two groups of HLA-C alleles defined on the basis of linked substitutions in the 3' exons. In comparison with HLA-A and -B, HLA-C alleles are more closely related to each other, there being less variation in residues of the antigen recognition site and more variation at other positions. In particular, the helix of the alpha 1 domain of HLA-C molecules is unusually conserved. Despite the reduced diversity in the antigen recognition site, it is evident that HLA-C genes have been the target of past selection for polymorphism. Within the antigen recognition site, it is the alpha 1 domain that is most diagnostic of HLA-C, whereas the alpha 2 domain is similar to that of HLA-B, the locus to which HLA-C is most closely related. In particular, conserved motifs in the alpha 1 helix and the conserved glycine at the base of the B pocket (position 45) provide a combination of features that is uniquely found in HLA-C molecules. We hypothesize that these features restrict the peptides bound by HLA-C molecules and in this manner reduce the efficiency of HLA-C assembly and expression at the cell surface. The overall picture HLA-C polymorphism obtained from this sampling of HLA-C alleles is unlikely to change as further alleles are characterized.

Gorilla class I major histocompatibility complex alleles: comparison to human and chimpanzee class I.

14 gorilla class I major histocompatibility complex (MHC) alleles have been isolated, sequenced, and compared to their counterparts in humans and chimpanzees. Gorilla homologues of HLA-A, -B, and -C were readily identified, and four Gogo-A, four Gogo-B, and five Gogo-C alleles were defined. In addition, an unusual Gogo class I gene with features in common with HLA-A and its related pseudogene, HLA-H, is described. None of the gorilla alleles is identical or even closely related to known class I alleles and each encodes a unique antigen recognition site. However, the majority of polymorphic substitutions and sequence motifs of gorilla class I alleles are shared with the human or chimpanzee systems. In particular, elements shared with HLA-A2 and HLA-B27 are found in Gogo-A and -B alleles. Diversity at the Gogo-B locus is less than at the Gogo-A locus, a trend the opposite of that seen for HLA-A and -B. The Gogo-C locus also appears to have limited polymorphism compared to Gogo-A. Two basic Gogo-C motifs were found and they segregate with distinctive sets of HLA-C alleles. HLA-A allels are divided into five families derived from two ancient lineages. All chimpanzee A alleles derived from one of these lineages and all gorilla alleles derive from the other. Unlike chimpanzee Patr-A alleles, the Gogo-A alleles do not clearly partition with one of the HLA-A families but have similarities with two. Overall, gorilla class I diversity appears from this sampling to show more distinctions from class I HLA than found for chimpanzee class I.