Evolution of the primate cytochrome c oxidase subunit II gene

We examined the nucleotide and amino acid sequence variation of the cytochrome c oxidase subunit II (COII) gene from 25 primates (4 hominoids, 8 Old World monkeys, 2 New World monkeys, 2 tarsiers, 7 lemuriforms, 2 lorisiforms). Marginal support was found for three phylogenetic conclusions: (1) sister-group relationship between tarsiers and a monkey/ape clade, (2) placement of the aye-aye (Daubentonia) sister to all other strepsirhine primates, and (3) rejection of a sister-group relationship of dwarf lemurs (i.e., Cheirogaleus) with lorisiform primates. Stronger support was found for a sister-group relationship between the ring-tail lemur (Lemur catta) and the gentle lemurs (Hapalemur). In congruence with previous studies on COII, we found that the monkeys and apes have undergone a nearly two-fold increase in the rate of amino acid replacement relative to other primates. Although functionally important amino acids are generally conserved among all primates, the acceleration in amino acid replacements in higher primates is associated with increased variation in the amino terminal end of the protein. Additionally, the replacement of two carboxyl-bearing residues (glutamate and aspartate) at positions 114 and 115 may provide a partial explanation for the poor enzyme kinetics in cross-reactions between the cytochromes c and cytochrome c oxidases of higher primates and other mammals. Correspondence to: R.L. Honeycutt     

Cloning of Paracoccus cytochrome c oxidase subunit II

Cytochrome c oxidase from Paracoccus denitrificans is composed of two subunits, yet is active in both electron transport and proton translocation. A cloning approach and immunologic screening protocol is described for the isolation of the subunit II gene expressed in E. coli. DNA sequencing should establish the extent of homology to eukaryotic oxidase.     

Evolution of eutherian cytochrome c oxidase subunit II: heterogeneous rates of protein evolution and altered interaction with cytochrome c

Cytochrome c oxidase subunit II (COII), encoded by the mitochondrial genome, exhibits one of the most heterogeneous rates of amino acid replacement among placental mammals. Moreover, it has been demonstrated that cytochrome c oxidase has undergone a structural change in higher primates which has altered its physical interaction with cytochrome c. We collected a large data set of COII sequences from several orders of mammals with emphasis on primates, rodents, and artiodactyls. Using phylogenetic hypotheses based on data independent of the COII gene, we demonstrated that an increased number of amino acid replacements are concentrated among higher primates. Incorporating approximate divergence dates derived from the fossil record, we find that most of the change occurred independently along the New World monkey lineage and in a rapid burst before apes and Old World monkeys diverged. There is some evidence that Old World monkeys have undergone a faster rate of nonsynonymous substitution than have apes. Rates of substitution at four-fold degenerate sites in primates are relatively homogeneous, indicating that the rate heterogeneity is restricted to nondegenerate sites. Excluding the rate acceleration mentioned above, primates, rodents, and artiodactyls have remarkably similar nonsynonymous replacement rates. A different pattern is observed for transversions at four-fold degenerate sites, for which rodents exhibit a higher rate of replacement than do primates and artiodactyls. Finally, we hypothesize specific amino acid replacements which may account for much of the structural difference in cytochrome c oxidase between higher primates and other mammals.      

Rapid evolution of cytochrome c oxidase subunit II in camelids (Tylopoda, Camelidae)

Within cetartiodactyl species, both New and Old World camelids are uniquely adapted to the extremely hot and dry climates of African-Asian territories and to the high altitude cold and hypoxic environment of the whole Andean area. In order to investigate the potential association between these particular adaptations and mitochondrial aerobic energy production, we examined the camelid genes of cytochrome c oxidase subunits I, II, and III and the replacement of amino acids inferred. We found that all subunits had undergone a number of replacements in sites otherwise conserved in other cetartiodactyls. Changes of COXI and COXIII were mainly located in the transmembrane helices of proteins. For COXII, although most of the changes did not occur in sites directly involved in electron transfer, a shift of D by T at 115 position of Old World camelid might modify electrostatic interactions with cytochrome c. COXII also showed an increased relative evolutionary rate respect to other cetartiodactyls compared.     

Structure and evolution of cytochrome oxidase

The structural features of cytochrome oxidases are reviewed in light of their evolution. The substrate specificity (quinol vs. cytochromec) is reflected in the presence of a unique copper centre (Cu_A) in cytochromec oxidases. In several lines of evolution, quinol oxidases have independently lost this copper. Also, the most primitive cytochromec oxidases do not contain this copper, and electron entry takes place viac-type haems. These enzymes, exemplified by the rhizobial FixN complex, probably remind the first oxidases. They are related to the denitrification enzyme nitric oxide reductase.     

The sequence of the gene for cytochrome c oxidase subunit I, a frameshift containing gene for cytochrome c oxidase subunit II and seven unassigned reading frames in Trypanosoma brucei mitochrondrial maxi-circle DNA.

A 9.2 kb segment of the maxi-circle of Trypanosoma brucei mitochondrial DNA contains the genes for cytochrome c oxidase subunits I and II (coxI and coxII) and seven Unassigned Reading Frames ("URFs"). The genes for coxI and coxII display considerable homology at the aminoacid level (38 and 25%, respectively) to the corresponding genes in fungal and mammalian mtDNA, the only striking point of divergence being an unusually high cysteine content (about 4.5%). The reading frame coding for cytochrome c oxidase subunit II is discontinuous: the C-terminal portion of about 40 aminoacids, is present in the DNA-sequence in a -1 reading frame with respect to the N-terminal moiety. URF5, 8 and 10, show a low but distinct homology (about 20%) to mammalian mitochondrial URF-1, 4 and 5, respectively. In URF5, the first AUG is found at codon 145, whereas extensive homology to mammalian URF-1 sequences occurs upstream of this position. The possibility exists that UUG can serve as an initiator codon. URF7 and URF9 have a highly unusual aminoacid composition and do not possess AUG or UUG initiator codons. These URFs probably do not have a protein-coding function. The segment does not contain conventional tRNA genes.     

Paracoccus denitrificans mutants deleted in the gene for subunit II of cytochrome c oxidase also lack subunit I.

As a prerequisite to site-directed mutagenesis on cytochrome c oxidase, two different mutants are constructed by inactivating the cta gene locus encoding subunits II and III (ctaC and ctaE) of the Paracoccus denitrificans oxidase. Either a short fragment encoding part of the putative copper binding site near the C terminus of subunit II, or a substantial fragment, comprising parts of the coding region for both subunits and all of the intervening three open reading frames, are removed and replaced by the kanamycin resistance gene. Each construct, ligated into a suicide vector, is mated into Paracoccus, and mutants originating from double homologous recombination events are selected. We observe complete loss of alpha-type heme and of oxidase subunits, as well as a substantial decrease in the cytochrome c oxidase activity. Upon complementation with the ctaC gene (plus various lengths of downstream sequence extending into the operon), subunit II gets expressed in all cases. Wild-type phenotype, however, is only restored with the whole operon. Using smaller fragments for complementation gives interesting clues on roles of the open reading frames for the assembly process of the oxidase complex; two of the open reading frame genes most likely code for two independent assembly factors. Since homologous genes have been described not only for other bacterial oxidases, but their gene products shown to participate also in the assembly of the yeast enzyme, they seem to constitute a group of evolutionary conserved proteins.     

Genetic and physical analysis of the mitochondrial gene for subunit II of yeast cytochrome c oxidase.

The mitochondrial genetic locus oxi 1 contains the structural gene for subunit II of Cytochrome c oxidase. In this study, the oxi 1 locus, or at least a major portion of it, has been localized to a 2·4 kb2 HpaII fragment of mitochondrial DNA, by examining the mtDNA of oxi 1 mutants, and rho^? yeast strains that selectively retained in amplified form, this region of the mitochondria) genome. The 2·4 kb fragment is missing from the mtDNA of an oxi 1 locus deletion mutant, but is present in the mtDNAs retained by two rho^? strains that genetically recombine with all 16 oxi 1 mutants tested, to produce respiring progeny. Two other rho^? strains, that retained different but overlapping portions of the oxi 1 locus as determined genetically, contained mtDNAs consisting of “cloned” segments derived from within the 2·4 kb fragment: these rho^? mtDNAs hybridized only to the 2·4 kb HpaII fragment of wild-type mtDNA and could not be cleaved with HpaII. Furthermore, these two rho^? mtDNAs were found to correspond to sequences from opposite sides of the 2·4 kb fragment that overlap for 100 to 300 base-pairs near the middle of the fragment. Thus, five oxi 1 mutations that recombine with both of these rho^? strains could be further localized to this relatively short region of overlap. One such mutation, of particular interest because it produces an altered form of subunit II, was shown to lie on a 75-base-pair fragment that maps in this region of the overlap. The 75-base-pair fragment from the mutant migrates slightly faster during electrophoresis than the corresponding wild-type fragment. In contrast, the mobility of the fragment from a spontaneous revertant was indistinguishable from wild type.     

Molecular evolution of cytochrome c oxidase: rate variation among subunit VIa isoforms

Cytochrome c oxidase (COX) consists of 13 subunits, 3 encoded in the mitochondrial genome and 10 in the nucleus. Little is known of the role of the nuclear-encoded subunits, some of which exhibit tissue-specific isoforms. Subunit VIa is unique in having tissue-specific isoforms in all mammalian species examined. We examined relative evolutionary rates for the COX6A heart (H) and liver (L) isoform genes along the length of the molecule, specifically in relation to the tissue-specific function(s) of the two isoforms. Nonsynonymous (amino acid replacement) substitutions in the COX6AH gene occurred more frequently than in the ubiquitously expressed COX6AL gene. Maximum-parsimony analysis and sequence divergences from reconstructed ancestral sequences revealed that after the ancestral COX6A gene duplicated to yield the genes for the H and L isoforms, the sequences encoding the mitochondrial matrix region of the COX VIa protein experienced an elevated rate of nonsynonymous substitutions relative to synonymous substitutions. This is expected for relaxed selective constraints after gene duplication followed by purifying selection to preserve the replacements with tissue-specific functions.      

Some features on RNA folding structures of cytochrome c oxidase subunit II and cytochrome P450

We have defined the folding index of the RNA secondary structureand compared the value of this index obtained from a given RNAmolecule with one from the random sequences with the same A,U, G, C ratio. The mRNA of the cytochrome c oxidase subunitII and cytochrome P450 from various species was analyzed. Wefound that the mRNA of these genes were more folded than therandom sequences and the coding regions of these mRNA were asequally folded as the random sequences. This result suggeststhat the coding region of these mRNA has no more important secondarystructure than have the random sequences with the same A, U,G, C ratio.