Striking similarities between antigen receptor J pieces and sequence in the second chain of the murine CD8 antigen

Abstract

The CD8 antigen is a marker for T-lymphocyte subsets that is absent from helper T cells but expressed on cytotoxic T cells which recognize foreign determinants in association with class I major histocompatibility complex (MHC) antigens¹. It has been suggested that CD8 plays some part in recognition by CD8⁺ cytotoxic T cells since anti-CD8 antibodies can block their functions² and the human CD8 antigen contains a domain with clear similarities to immunoglobulin and T-cell receptor (TCR) variable-region (V) domains^3,4. Human CD8 antigen is thought to be a homodimer⁵ but in the mouse and rat the equivalent antigens (alternatively called Lyt2,3 and OX8) are heterodimeric^1,6. Rat CD8 contains two chains of relative molecular mass 32,000 (32K) and 37K: the 32K chain is the rat homologue of human CD8 and mouse Lyt2 (ref. 6). We describe here the molecular cloning of the rat 37K chain using an oligonucleotide probe predicted from peptide sequence. The full protein sequence is derived from the complementary DNA and matches limited peptide sequence for mouse Lyt3. The new sequence is more like immunoglobulin and T-cell receptor V domains than other T-cell antigens and includes a patch that is almost identical to some joining (J) piece sequences. This suggests that the CD8 and receptor heterodimers may have evolved directly from a common ancestor.