Abstract
Nucleotide sequencing adjacent to the right inverted terminal repetition of the vaccinia virus genome revealed two genes encoding polypeptides that are structurally related to members of the plasma serine protease inhibitor superfamily (SPI). Inclusion in the superfamily is based on extensive amino acid sequence similarities as well as a consensus sequence adjacent to the active-site region near the carboxyl ends of the proteins. The genes designated SPI-1 and SPI-2 are located 10,000 and 17,000 base pairs from the right end of the genome, respectively. The predicted SPI-1 polypeptide is 11 amino acids longer than that of SPI-2, and the deduced masses are 40,471 and 38,125 daltons, respectively. Similarities between SPI-1 and SPI-2 are indicated by the percentage of identical amino acids (44%) and corresponding hydrophobicity plots. The maximum amino acid sequence diversity occurs precisely in the putative active-site region, suggesting that SPI-1 and SPI-2 may inhibit different proteases. SPI-2 is homologous to a previously described cowpox virus gene (D. J. Pickup, B. S. Ink, W. Hu, C. A. Ray, and W. K. Joklik, Proc. Natl. Acad. Sci. USA 83:7698-7702, 1986). Evidence for a cowpox virus homolog of SPI-1 was obtained by DNA hybridization. Thus, the presence of two genes that belong to the plasma serine protease inhibitor superfamily may be characteristic of orthopoxviruses.