THE SYNONYMOUS SUBSTITUTION RATE OF THE MAJOR HISTOCOMPATIBILITY COMPLEX LOCI IN PRIMATES

Because the divergence of many allelic lineages at the major histocompatibility complex (MHC) loci predates species divergence, standard methods of calculating synonymous substitution rates are not applicable to this system. We used three alternative methods of rate estimation: one based on the minimum number of substitutions (D(m)), another on the nucleotide difference (D(xy)), and the third on the net nucleotide difference (D(n)). We applied these methods to the protein-encoding sequences of primate MHC class I (A, B, and C) and class II (DRB1) genes. To determine the reliability of the different estimates, we carried out computer simulation. The distribution of the estimates based on D(xy) or D(n) is generally much broader than that based on D(m). More importantly, the D(m)-based method nearly always has the highest probability of recovering true rates, provided that D(m) is not smaller than 5. Because of its desirable statistical properties, we used the D(m)-based method to estimate the rate of synonymous substitutions. The rate is 1.37 +/- 0.61 for A, 1.84 +/- 0.40 for B, 3.87 +/- 1.05 for C, and 1.18 +/- 0.36 for DRB1 loci, always per site per 10(9) years. Hence despite the extraordinary polymorphism, the mutation rate at the primate MHC loci is no higher than that of other loci.