Evolution and phylogeny of hominoids inferred from mitochondrial DNA sequences

This dissertation addresses the 4.9 kb (kilobases) nucleotide sequences of mitochondrial (mt) DNAs from five hominoid species (common and pygmychimpanzees, gorilla, orangutan and simang), and presents their detailed analyses, together with the known human whole sequence, to assess the tempo and mode of hominoid mtDNA evolution. Particular attention was paid to the rate of synonymous substitutions in protein coding region as well as of silent substitutions in other regions. This work was further extended to the whole mitochondrial genomes of four hominoid species (human, common chimpanzee,′ gorilla and orangutan) with additionally determined l0 to 12 kb mtDNAs from common chimpanzee, goriIIa and orangutan. These hominoid mtDNAs revealed several functionally and evolutionarily characteristic features and provided useful information on the history of hominoid species. Most significant observations drawn from the present data are summarized as follows. First, comparsion of the base compositions in any specified region of hominoid mtDNAs showed a strong base composition bias, as observed in other vertebrate mtDNAs. The L-stand of hominoid mtDNAs is rich in A (adenine) and C (cytosine) contents, but low in G (guanine) content. Base composition biases are strongest at the third codon positions and are evident along the whole genome,independent of the genomic regions. Both codon usage and amino acid preference of mitochondrial protein genes are in agreement with the base composition biases. These observations suggested that there is a biased mutation pressure in mtDNA. A possible cause may be differential diaminations of C residues owing to the asymmetric replication of both L- and H-strands of mtDNA. It is possible that diffferential deamination has resulted in the reduced number of C residues in the H-strand,although there has been no clear evidence for this possibility in hominoid mtDNAs. Second, there exist functionally important nucleotide sites over the genome.Together with information on tertiary structures of proteins, as Well as on secondary structures of transfer (t) RNAs, ribosomal (r) RNA genes and noncoding regions, the distributjon of variable sites among hominoid mtDNAs suggested that some nucleotide sites have been playing important roles in peptide folding, assembly of proteins, or interaction to some other proteins and regulatory elements. Noteworthy are two functionally distinct regions in the maior noncoding region (D-loop). One is concerned with promoter sequences for transcripdon and the other is with three conserved blocks. Oranguan mtDNA sequence revealed unusual substitutions at both of these regions. This suggested that the replication and transcription machinery in orangutan mtDNA may differ from that of other hominoid mtDNAs. Third, comparsion of nucleotide differences observed among closely related hominoids revealed a remarkably biased mode of changes. Between human and chimpanzee, 70% of the observed nuculeotide differences are silent changes that occur mostly in the small noncoding regions or at the third codon positions of protein genes. Extensive deletions and additions are observed, but they are found only in the noncoding regions. Such observations suggested a conserved mode of the evolution of hominoid mtDNA genomes. There is also a strong preference to transitions over transversions. Out of 852 variable third positions of codons between the human and common chimpanzee mtDNAs, 93% account for transitions of which 66% are TC transitions (in the L-strand). Within the remaming 7% transversions, CA differences are most frequent while GT are least. These substitution biases correlate well with biased base compositions, particularly the low G content of the L-strand. Fourth, owing to the outnumbered transitions and strong biases in the base compositions, synonymous substitutions reach rapidly a rather low saturation level. AG transitions attain a saturation level lower than TC transitions (in the L-strand), and such a low ceiling is observed even between the human and chimpanzee pair that diverged around five million years ago. At present,it seems inevitable to select appropriate regions that have experienced theoretically tractable numbers of substitutions.In the case of hominoid mtDNAs, candidates are all types of changes in the tRNA and rRNA regions, transversions in the noncoding regions, and nonsynonymous changes and synonymous transversions in the protein coding regions. Fifth, rapidly evolving mtDNAs are potentially useful for addressing classical issues in taxonomy, provided that each nucletide site has not undergone extensive multiple-hit substitutions. From the Whole 16209 sites of mtDNAS compared among the four hominoid specles, it appears that 12137 such sites are suitable to phylogenetic use. The analysis strengthened the pattern and dating in hominoid diversifjcation infened from the Previous analysis of 4.9 kb reglon in six homjnoid species(among African apes,gorilla diverged first about 7.7 million years ago and then chimpanzee and human became distinct about 4.7 million years ago). Finally, the synonymous and nonsynonymous substitution rates were examined under the assumption of the gorilla divergence being 7.7 miIIion yearsago. The extent of the compositional biases differs from gene to gene. Such differences in base compositions, even if small, can bring about considerablevariations in observed synonymous differences, and may result in the region-dependent estimate of the synonymous substitution rate. A care should be taken for heterogeneous transition and base composition biases as Well as different saturation levels of transition changes. The synonymous substitution rateestimated with this caution showed the uniformity over genes (2.37 ± 0.11 x 10-8 per site per year) and the high transition rate, about 17 times faster than the transversion rate. These synonymous and transition rates are comparable to the silent substitution rate in the noncoding segments dispersed between genes. On the other hand, the rate of nonsynonymous substitutions differs considerably from gene to gene as expected under the neutral theory of molecular evolution. The average differences in the gorilla - human and gorilla - chimpanzee comparisons indicated that the lowest rate is 0.7 x 10-9 per site per year for COI and that the highest rate is 5.7 x 10-9 for ATPase 8. The degree of functional constraints (measured by the ratio of the nonsynonymous to the synonymous substitution rate) is 0.03 for COI and 0.24 for ATPase 8. tRNA genes also showed variability in the base content and thus in the extent of nucleotide differences as well. The substitution rate averaged over 22 tRNAS is 5.6 x 10-9 per site per year. The rate for 12S rRNA and 16S rRNA is 4.1 x 10-9 and 6.9 x 10-9 per site per year. respectively. All of these observations strongly suggested that mutations themselves occur more or less with the same rate and compositional biases