The nearly neutral theory proposes that interaction of selection and drift is important in molecular evolution. The theory predicts that evolutionary rate is negatively correlated with the species population size, for those genes whose function has been fixed long time ago. The prediction is tested through analysis of sequence data of 17 mammalian genes by estimating divergence among genes separately for synonymous substitutions and nonsynonymous substitutions. Star phylogenies composed of rodentia, artiodactyla and primates are examined. The generation-time effect is found to be more conspicuous for synonymous substitutions than for nonsynonymous substitutions. This result supports the nearly neutral theory.

View full abstract

Available genetic data reveals that the human population is more variable than the chimpanzee population at the protein level, whereas the opposite is the case at the DNA level. The lower level of silent polymorphism in the human population suggests that its long-term breeding size is smaller than the chimpanzee's. The neutral theory suggests that natural selection has been relaxed in the human population under the improved environment. The possibility that the relaxation began with the emergence of Homo sapiens is examined, because it is known that H. habilis underwent for the first time dramatic changes in brain size, way of life, and culture, and that the childhood of H. erectus was already twice as long as that of chimpanzee. The relaxation hypothesis predicts that, relative to chimpanzee, some 20% of deleterious mutations became harmless under the changed environment throughout the Pleistocene. More extensive study of genetic variation in non-human primates is necessary not only to confirm the hypothesis, but also to better understand the human genome itself.

View full abstract

Repeated sequences are found ubiquitously in the eukaryotic genome. Population genetic studies on the evolution of such repeated sequences are reviewed while paying special attention to those sequences found in the non-coding regions of the genome. Specifically, the evolution of dispersed repeated sequences by the transposition as well as the evolution of short tandemly repeated sequences due to either replication slippage or unequal sister chromatid exchange are considered. The approach of combining both model and data analyses which has been successfully employed in the development of the neutral theory is also considered to be useful in better understanding the evolution and biological meaning of these sequences.

View full abstract

A large amount of genetic variation can be maintained in natural populations. In order to understand the mechanism maintaining genetic variation, we must first estimate the amount of genetic variation. There are two measures for estimating the amount of DNA polymorphism, i.e., the average number of pairwise nucleotide differences and the number of segregating sites among a sample of DNA sequences. Using these two measures, we can test the neutral mutation-random drift hypothesis (the neutral theory). The expectation of the amount of DNA polymorphism has been studied under several models, including population subdivision, change in population size, and natural selection. When a population is subdivided, a large amount of DNA polymorphism can be maintained in the population if the migration rates among subpopulations are small. In this case the amount of DNA polymorphism in the subpopulation with lower migration rate is expected to be smaller than that of higher migration rate. When the population size changes, the number of segregating sites changes more rapidly than does the average number of nucleotide differences. When purifying selection is operating, the number of segregating sites is more strongly affected by the existence of deleterious mutants than is the average number of nucleotide differences. On the other hand, when balancing selection is operating, the effect of the selection on the average number of nucleotide differences is larger than that on the number of segregating sites. A mutant under natural selection affects the amount of DNA polymorphism at linked sites (hitchhiking effect). DNA sequences are not random sequences and there may be conservative and variable regions in them. A statistical method for determining the window size and for finding nonrandom regions in the sequence is also presented.

View full abstract

The ORF100 region within Pst-12 fragment of chloroplast DNA (cpDNA) in rice (Oryza sativa) were amplified by polymerase chain reaction in 137 cultivars and 117 strains of wild rice from various countries. Two cpDNA types, with and without a deletion at the region, were found. In O. sativa, 63 and 74 cultivars had the deletion and the non-deletion cpDNAs, respectively. It corresponded to indica and japonica classified by morphological and physiological characters. Plant groups of 45 strains having the deletion and 72 strains with the non-deletion in spDNA were found in wild rice. Wild species that distantly related to O. sativa had mostly cpDNAs without the deletion. The deletion is likely to have differentiated in O. rufipogon. Annual types of O. rufipogon tend to have the deletion, while perennial types tend to loose it. The results of this study supported a view that the indica and japonica types were differentiated in wild species before domestication.

View full abstract

Spontaneous mutations were accumulated for a total of 1,678,388 allelegenerations in several hundred replicate second chromosome lines. These lines were composed of a Cy chromosome and one of three different lethal bearing chromosomes from natural populations. Electrophoretic mobility and / or activity change was screened for seven enzyme loci. Forty-four null mutations were detected, but no band-morph mutations were observed. There is significant variation in null mutation rate among enzyme loci and among lethal chromosomes. This may be attributable to differences in the distribution of transposable elements among the lines. The band-morph mutation rate is pooled with previous results and becomes 7.48 × 10^-7 with a 95% confidence limit of 2.04 × 10^-7 to 1.91 × 10^-6 per locus per generation. Similarly, the pooled null mutation rate becomes 1.30 × 10 ^-5 with 95% confidence limit of 1.15 × 10 ^-5 to 1.52 × 10^-5 per locus per generation.

View full abstract

The retinoblastoma 1 gene was assigned to mouse chromosome 14 at band D3 and rat chromosome 15 at band q12, respectively, by fluorescence in situ hybridization (FISH). These results suggest the syntenic association between human 13q14, mouse 14D3 and rat 15q12.

View full abstract