ABSTRACT
Although the collection of completely sequenced mitochondrial genomes is expanding rapidly, only recently has a phylogenetically broad representation of mtDNA sequences from protists (mostly unicellular eukaryotes) become available. This review surveys the 23 complete protist mtDNA sequences that have been determined to date, commenting on such aspects as mitochondrial genome structure, gene content, ribosomal RNA, introns, transfer RNAs and the genetic code and phylogenetic implications. We also illustrate the utility of a comparative genomics approach to gene identification by providing evidence that orfB in plant and protist mtDNAs is the homolog of atp8, the gene in animal and fungal mtDNA that encodes subunit 8 of the F0 portion of mitochondrial ATP synthase. Although several protist mtDNAs, like those of animals and most fungi, are seen to be highly derived, others appear to be have retained a number of features of the ancestral, proto-mitochondrial genome. Some of these ancestral features are also shared with plant mtDNA, although the latter have evidently expanded considerably in size, if not in gene content, in the course of evolution. Comparative analysis of protist mtDNAs is providing a new perspective on mtDNA evolution: how the original mitochondrial genome was organized, what genes it contained, and in what ways it must have changed in different eukaryotic phyla.
Mitochondrial DNA (mtDNA) is extraordinarily diverse in size, gene content and genome organization (1-5) and it is a daunting task to attempt to elucidate the mechanisms and reconstruct the pathways by which this evolutionary diversification has occurred. The preferred approach to answering such evolutionary questions is through comparative analysis of complete mtDNA sequences, which provides a genome-level perspective on such issues as what genes are present, how they are arranged, whether there are introns (and, if so, what types), how spacer sequences are distributed and how large they are, whether segments of the genome are repeated and other relevant information. Currently, 63 complete mtDNA sequences are available through public domain databases; however, the phylogenetic range that these sequences represent is both narrow and biased: 47 (75%) are from animal species (31 vertebrate, 16 invertebrate); five (8%) are from fungi; two (3%) are from plants; only nine (14%) are from protists, in spite of the fact that the latter group of organisms (mostly unicellular) comprises the bulk of the biological diversity of the eukaryotic lineage (6). This limited and highly non-representative data set has made it difficult to draw meaningful conclusions about the ancestral form of the mitochondrial genome, a necessary starting point for inferences about subsequent mitochondrial genome evolution.
To redress this imbalance, the Organelle Genome Megasequencing Program (OGMP) was established in 1992, having as a specific aim the systematic and comprehensive determination of complete protist mtDNA sequences. [Brief descriptions of the OGMP and two allied databases, the Protist Image Database (PID) and the Organelle Genome Database Project (GOBASE), appear at the end of this review]. At that time only three complete protist mitochondrial genome sequences had been published: the 6 kb mtDNA sequences of the apicomplexans Plasmodium yoelii (a rodent parasite) (7) and Plasmodium falciparum (the human malaria parasite) (8) and the 40 kb mtDNA sequence of the ciliate protozoan Paramecium aurelia (9). Partial but extensive mtDNA sequence information was also available for another ciliate protozoan, Tetrahymena pyriformis, several trypanosomatid protozoa (in the genera Trypanosoma, Leishmania and Crithidia) and the green alga (chlorophyte) Chlamydomonas reinhardtii. These limited data suggested that protist mtDNAs might be even more structurally variable than their counterparts in the multicellular eukaryotic lineages (1).
In the ensuing 5 years, a larger selection of complete protist mtDNA sequences has become available through the efforts of the OGMP, a complementary Fungal Mitochondrial Genome Project (FMGP) (5) and other research groups. This review summarizes and comments upon various aspects of protist mitochondrial genome structure, particularly gene content, that have emerged from these new sequences. In recent years comprehensive reviews of animal (10), fungal (5,11) and plant (12,13) mtDNAs have been published, but reviews of protist mtDNAs have been limited to specific groups, e.g. ciliates (14), trypanosomatids (15) and apicomplexans (16). Because protists encompass most of the phylogenetic breadth of the eukaryotic lineage and, by definition, contain a number of clades whose evolutionary depth exceeds that of the traditional animal, plant and fungal kingdoms, it is important to sample widely within this disparate assemblage to obtain a clear perspective on the range of mtDNA structural diversity in protists, in comparison with the more widely studied mitochondrial genomes from other eukaryotes. The data assembled here emphasize that most non-protist mtDNAs, particularly those of animals, are substantially derived relative to most of their protist counterparts, having lost many genes that are commonly still found in protist mitochondrial genomes. The compilation provided here better defines the properties of a typical ancestral (i.e. minimally diverged) protist mtDNA and allows us to suggest with greater confidence what genes were likely contained in the proto-mitochondrial genome (i.e. the last common ancestor of contemporary mitochondrial genomes).
Table 1 identifies the 23 complete protist mtDNA sequences that to our knowledge have been determined to date. These sequences encompass a reasonably broad selection of protist taxa, although they still represent only a fraction of recognized protist lineages (6). Nine of these sequences are in the public domain; the remainder are unpublished ones determined by the OGMP (eight), the FMGP (two) or other research groups (four). As well, we include complete mtDNA sequences from representative non-protists for purposes of comparison. Figure 1 displays the relative phylogenetic positions (to the extent that these can be inferred or proposed at present) of the protists listed in Table 1, together with other protist species, including future candidates selected by the OGMP for complete mtDNA sequencing.
In the case of complete mtDNA sequences published by other groups and deposited in the public domain we have used the standardized and corrected versions available in GOBASE (17; see below). Importantly, annotations accompanying these sequences have been unified with respect to gene and product nomenclature. These particular sequences have also been re-analyzed by us using informatics tools developed in-house and described below.
With the exception of BLAST (used for remote database searches) (18), FASTA (used for detailed sequence comparison) (19) and NIP (the Staden nucleotide sequence analysis package) (20), all of the informatics tools employed for this compilation have been developed by the OGMP Sequencing Unit. Many of the programs make use of the OGMP `masterfile' (mf) concept, an ASCII-based sequence file format that integrates nucleotide sequence, gene annotations and technical notes.
The sequence retrieval and analysis tools developed by the OGMP have for the most part been written in the Perl programing language. These tools include: BBLAST [batch mode BLAST search of the National Center for Biotechnology Information (NCBI) GenBank database]; BOB (BLAST output browser); FERRET, BADGER and CLEVER, retrieval tools used in conjunction with the NCBI Entrez database; GOBASE2MF [a program for converting from sequence records stored in Sybase tables of GOBASE (17) into mf format]; CLEANMF (used to verify sequence files in mf format as to annotation syntax and logic); PEPPER (for translation of protein coding sequences and extraction of non-coding regions); ONIP (command line interface to the Staden NIP program, used in the creation of codon usage tables of various gene classes); CN (sequence counter and checker). For compiling the body of data presented in Table 2, a number of wrapper scripts were written in the Bourne shell script language; these programs call upon the above tools and produce output files of appropriate layout. Scripts that use genome sequence files in mf format as input include: CODAT (calculation of A+T content of coding and non-coding regions); COTAB [creation of codon usage tables of three types of protein coding regions: genes, intronic open reading frames (ORFs) and unique ORFs]; BFASTA (batch FASTA search, used in comparing the protein sequences of two library files); TRNLIST (which creates a list of tRNA genes present in a genome). Further information about these programs is available at the OGMP website (see below).
Complete sequence analysis has provided evidence of both circular mapping and linear mapping protist mtDNAs, with circular mapping genomes predominating (Table 2). Among the protist mitochondrial genomes characterized as linear, no common end structures have been identified (see Table 2 for details).
The protist mtDNAs listed in Table 2 have a median size of ~40 kb, ranging from 6 kb in the three apicomplexan species (the smallest known mtDNAs) to 77 kb in the choanoflagellate Monosiga brevicollis. The majority of protist mtDNAs are compact, gene-rich genomes, with few or no large non-coding regions. Intergenic spacers are generally small and sparse, accounting in nine cases for <10% of the mtDNA, with coding regions sometimes overlapping. In Acanthamoeba castellanii, Dictyostelium discoideum, M.brevicollis, Chlamydomonas eugametos and Pedinomonas minor all genes are transcribed from the same strand of the mtDNA; otherwise, more than one potential transcription unit is present in protist mitochondrial genomes.
The overall A+T content is high (>70% in 15 cases) in protist mtDNAs and is usually elevated in non-coding intergenic regions compared with coding regions (up to 1.2-fold higher in M.brevicollis mtDNA). The numbers in Table 2 suggest that, in general, protist mtDNAs have evolved in the direction of higher A+T content.
In animals, as exemplified by Homo sapiens and Metridium senile in Table 2, the evolutionary trend has clearly been toward a further compaction of the mitochondrial genome, both by loss of genes and by virtual elimination of intergenic spacers. Conversely, in plants (e.g. Marchantia polymorpha) the trend has been in the opposite direction, with the mtDNA tending to increase in size, primarily by acquisition of a large amount of apparently non-coding DNA of currently unknown origin and function (Table 2). In the recently sequenced 366 924 bp mitochondrial genome of the angiosperm Arabidopsis thaliana (21), fewer genes are encoded than are found in M.polymorpha mtDNA, which is half the size (Table 2); overall <10% of the A.thaliana mtDNA has an assigned coding function. A key question is how and why evolution has produced such divergent mitochondrial genome patterns in different eukaryotic lines.
Table 1
In vertebrate animals, e.g. H.sapiens (Hsa), the mitochondrial genome contains genes for 13 inner mitochondrial membrane proteins involved in electron transport and coupled oxidative phosphorylation (nad1-6 and 4L, cob, cox1-3 and atp6 and 8) (Table 3), as well as genes for large subunit (LSU) and small subunit (SSU) rRNAs (rnl and rns respectively; Table 4). This `standard set' of mtDNA-encoded genes (plus atp9) is also found in fungal (e.g. Allomyces macrogynus, Ama) mtDNAs, except that certain ascomycete fungi (e.g. Schizosaccharomyces pombe, Spo) lack all nad genes. Animal and fungal mtDNAs do not encode a 5S rRNA (Table 4) nor, with the exception of rps3 in A.macrogynus mtDNA (22), do they carry any ribosomal protein genes (Table 5). In land plant mtDNAs a few extra respiratory chain protein genes are found (e.g. nad9 and atp1 in M.polymorpha; Table 3); however, the most notable departure from animal and fungal mtDNAs is the presence in plant mtDNA of a set of ribosomal protein genes (Table 5) as well as a gene for 5S rRNA (rrn5; Table 4). In the case of M.polymorpha mtDNA several homologs of known mitochondrial genes (e.g. sdh3,4 and yejR,U,V; Tables 3 and 6) were initially considered to be unique ORFs (23).
With only a few exceptions, protist mtDNAs encode LSU and SSU rRNAs whose potential secondary structures deviate minimally from their eubacterial counterparts (OGMP, unpublished results). This corresponds to what has been observed with plant mitochondrial rRNAs, but stands in marked contrast to most fungal but particularly animal mitochondrial rRNAs (40,41). Clearly recognizable in most protist mitochondrial LSU rRNAs are the 5'- and 3'-terminal regions corresponding to the `5.8S' and `4.5S' domains of a eubacterial counterpart such as Escherichia coli 23S rRNA. These terminal regions have largely been eliminated from animal mitochondrial LSU rRNAs (41). These observations reinforce the emerging view that the most ancestral (minimally derived) mitochondrial genomes will be found among the protists.
A minority of protist mtDNAs encode rRNA genes whose structure and/or the structure of their products is very unusual. The 9S (SSU) and 12S (LSU) mitochondrial rRNAs of trypanosomatid protozoa (e.g. Leishmania tarentolae and Trypanosoma brucei) are among the smallest and structurally most divergent of known rRNAs, having potential secondary structures in which only a few of the expected conserved structural elements are identifiable (40,41). Also unusual are the mitochondrial rnl genes of Paramecium aurelia (42,43), Tetrahymena pyriformis (43) and Pedinomonas minor (OGMP, unpublished results), which are split into two pieces that are separated in the genome and interspersed with other genes (Table 4). The Pedinomonas situation is particularly intriguing because a more extreme case of rnl fragmentation and scrambling is seen in the mtDNA of a phylogenetically later branching green algal genus, Chlamydomonas (44-46). Fragmented and dispersed rRNA gene elements, encoded on both strands of the mtDNA, have also been found in the small apicomplexan mtDNAs (8,47). Because most protist mtDNAs encode conventional, 16S-like and 23S-like rRNAs (the ancestral state), these deviant examples must represent derived patterns of mitochondrial rRNA gene structure and organization within the specific lineages in which they occur.
Like animal and fungal mtDNAs, most protist mtDNAs lack a 5S rRNA gene, the current exceptions (Table 4) being the chlorophyte algae Prototheca wickerhamii (48) and Nephroselmis olivacea (Nol) (M.Turmel, C.Otis and C.Lemieux, unpublished results), the red alga C.crispus (see Table 4, footnote g) and the jakobid flagellate R.americana (49). As in the case of sdh genes noted above, the sporadic phylogenetic distribution of mitochondrial rrn5 suggests that this gene was lost from mtDNA a number of times.
Complete sequencing of an organelle genome is the only way to determine unequivocally whether that genome encodes all of the tRNA species necessary to support organellar protein synthesis. Several protist mtDNAs (those of M.brevicollis, P.wickerhamii, R.salina and Malawimonas jakobiformis in Table 1) do appear to encode the minimal required tRNA set, if one allows that a single tRNA is able to decode the four-codon family specifying a given amino acid (see Table 7). However, in most cases, tRNAs recognizing one or more codons are evidently absent from the mitochondrial genome, and tRNA import from the cytosol is usually invoked as the mechanism for making up the deficit. Import of nuclear DNA-encoded cytosolic tRNAs into mitochondria is clearly required in the case of A.castellanii, D.discoideum, P.aurelia, T.pyriformis, Chlamydomonas spp. and P.minor, whose mtDNAs encode substantially fewer than the minimal required set (Table 7); in fact, import of tRNA into Tetrahymena mitochondria, long inferred on the basis of tRNA population studies (50), has recently been documented experimentally (51). No tRNA genes have been found in the mitochondrial genomes of apicomplexan or trypanosomatid protists, where import of a full set of tRNAs from the cytoplasm is assumed (52,53). The data in Table 7 indicate that mitochondrial tRNA import is not only likely to be widespread among protists [as it is also in plants (54) and several chytridiomycete fungi (5)], but that it emerged early in the evolution of the mitochondrial translation system, probably a number of times independently. Genes for certain tRNAs (e.g. Met and Trp) are encoded by the mitochondrial genomes of virtually all protists, whereas genes for other tRNAs (notably Thr) are found infrequently among protist mtDNAs (Table 7).
Table 7
aSee Table 1 for complete organism names. n gene present; m gene absent. Aminoacylation specificity (a.a.) is indicated by the standard one letter symbols for amino acids (Me, elongator methionine; Mf, initiator methionine). The predicted anticodon of each tRNA is shown in lower case letters, with the predicted codon(s) that would be recognized shown in upper case letters (N = any nucleotide; R = A or G; Y = C or U). Expanded wobble base pairing is assumed, such that anticodons beginning with uridine are considered to recognize all codons in a four-codon family.
Several protist mitochondrial genomes, as well as that of M.polymorpha, lack only one or two of the minimal required set of tRNA genes. Again, in these cases it is generally held that import of cytosolic tRNAs makes up the deficit. Indeed, import into M.polymorpha mitochondria has recently been documented in the case of nucleus-encoded tRNAIle(aau) (55) and tRNAThr(agu) (56), genes for which have not been identified in M.polymorpha mtDNA (23). However, an alternative possibility that should be considered is that the anticodon sequence in a single mtDNA-encoded tRNA might be subject to partial editing, such that the unedited and edited versions accept different amino acids and pair with codons corresponding to these amino acids. Partial C -> U editing of a tRNA`Gly'(gcc) to generate a tRNAAsp(guc) in opossum mitochondria (57) serves as a precedent for this possibility.
In A.castellanii, sequencing of the mtDNA has provided evidence of a novel type of tRNA editing that affects most of the mtDNA-encoded tRNAs (57-57; D.H.Price and M.W.Gray, unpublished results). This editing is confined to one or more of the first three positions at the 5'-end of the tRNA (62). Except for the mismatching in the acceptor stem that is corrected by this editing, the secondary structures of Acanthamoeba mitochondrial tRNAs are quite conventional (58-62). What appears to be the same type of mitochondrial tRNA editing has recently been documented in the chytridiomycete fungus Spizellomyces punctatus (63) and several other primitive fungi (B.F.Lang, unpublished results); moreover, in the case of tRNAs encoded by D.discoideum mtDNA secondary structure modeling strongly suggests that several of these undergo a similar type of editing. Orthodox cloverleaf secondary structures are the rule for mitochondrial tRNAs throughout the protists, one notable variant being an unusual tRNAMet in Tetrahymena mitochondria (64). The structurally aberrant tRNAs characteristic of animal mitochondria (65,66) are therefore exceptional, representing a highly derived form of mitochondrial tRNA which, nevertheless, is able to assume the required L-shaped tertiary structure (67).
In almost half of the protists listed in Table 7 we infer, on the basis of codon usage and the presence of a tRNATrp having a CCA anticodon, that the mitochondrial translation system uses the standard genetic code, as is the case in land plants. In the remaining protists UGA appears to be decoded as tryptophan rather than as stop (Table 7), being the preferred Trp codon in all but P.aurelia; in fact, UGA is used almost exlusively to encode Trp in M.brevicollis and T.pyriformis mitochondria. From the phylogenetic distribution of this code variation it is evident that the change in UGA coding must have occurred on more than one occasion.
Compared with plant mtDNA, protist mtDNAs seem to have remarkably few introns (Table 8). At least half of these genomes entirely lack group I and group II introns. So far, among the 23 completely sequenced protist mtDNAs listed in Table 1, group I introns have only been found (and then only in small numbers) in the amoeboid protozoa A.castellanii and D.discoideum, the green algae P.wickerhamii, N.olivacea and C.eugametos and the choanoflagellate M.brevicollis. Prototheca wickerhamii and M.polymorpha mtDNAs share with one another (and with fungal mtDNA) positionally equivalent and structurally homologous cox1 introns, suggesting that these introns have been inherited vertically from a mitochondrial ancestor of fungi, green algae and plants (68). On the other hand, horizontal transfer of other group I introns is suggested by the fact that in the rnl gene of A.castellanii mtDNA and in the chloroplast DNA of certain Chlamydomonas species, several mobile group I introns are not only positionally identical, but have homologous intron core structures and intron ORFs (69).
Table 8
aFull organism names are listed in Table 1.
Very few group II introns have been found in protist mtDNAs (a total of seven such introns in five of 23 completely sequenced protist mtDNAs). Again, we have some evidence suggesting acquisition of certain of these introns by horizontal transfer (OGMP, unpublished results), as appears also to be the case for certain group II introns found in the rnl gene of the brown alga Pylaiella littoralis (70). In our view the paucity of group II introns in protist mtDNAs coupled with their sporadic distribution and evidence of horizontal transfer makes it quite unlikely that there was a wholesale acquisition of group II introns by the eukaryotic cell via the [alpha]-proteobacteria-like proto-mitochondrial endosymbiont.
Accumulating sequence data are aiding in the identification of some of the unassigned ORFs that have been uncovered in the course of sequencing mitochondrial genomes. As an example we provide evidence here that orfB, a conserved gene of unknown function originally identified in plant mtDNA (see Table 3, footnote i), is the homolog of atp8, which encodes subunit 8 of the F0 portion of the ATP synthase. The latter gene has been found in a number of animal and fungal mtDNAs, but up to now has not been identified in plant or protist mitochondrial genomes. Conversely, orfB is found in almost all plant and protist mtDNAs, but not in those of animals or fungi. Both Atp8 and OrfB proteins are characterized by the same block of three identical amino acids at the N-terminus, followed by an otherwise quite variable sequence (Fig. 2). The known OrfB proteins of plants differ from Atp8 essentially in their increased length. Because there is also much length variation among OrfB homologs in some protist mtDNAs, we were prompted to assess the possibility that atp8 and orfB are homologous genes.
The N-terminal functional domain (71) of ATP synthase subunit 8 is well conserved in different fungi compared with the central hydrophobic domain (72) and the C-terminal domain (73). The latter domain contains a region enriched in positively charged amino acid residues (73), which are thought to play an important role in assembly of the F0 complex (see below). If OrfB is indeed homologous to Atp8, we should find similar amino acid signatures in a multiple alignment of a phylogenetically diverse collection of both types of sequences. Such a collection has recently become available through the sequencing efforts of the OGMP and FMGP.
As shown in Figure 2, the highly conserved N-terminal domain provides the best evidence for homology between orfB and atp8. Further evidence supporting this inference is the presence of perfectly aligned central hydrophobic and positively charged domains. Based on the alignment of the first 57 amino acids shown in Figure 2, we suggest that there is little basis for a distinction between the `Atp8' and `OrfB' classes of protein. With two notable exceptions, this sequence compilation further demonstrates that a long C-terminal extension (position 78 and beyond in Fig. 2) is only found among plants and protists. In the stramenopiles Cafeteria roenbergensis and Ochromonas danica the mtDNA codes for a shorter protein, about as long as the longest fungal sequences. This feature is not clade specific because in another stramenopile, Phytophthora infestans, the mitochondrial genome specifies an Atp8 protein that is rather typical in size for protists. The C-terminal extension is not only quite variable in size, but indeed is so divergent in sequence that it can only be reasonably well aligned among very closely related species (e.g. land plants). Thus the presence or absence of a C-terminal extension also does not distinguish between `Atp8' and `OrfB' classes.
Conserved sequence motifs within the hydrophobic and C-terminal domains of the Atp8/OrfB protein are restricted to the boundaries between these domains, the `LP motif' (71), which is immediately followed by a region with one or several positively charged amino acids. Previous studies in fungi have shown that thesepositively charged amino acids play an important role in assembly of subunits 6, 8 and 9 (73).
In summary, plant and protist mitochondrial OrfB proteins contain all of the conserved sequence elements characteristic of animal and fungal Atp8 proteins. Thus the orfB gene represents the best candidate for the previously `missing' atp8 homolog in plant and protist mtDNAs.
The mitochondrial gene content and genome organization data being generated by the OGMP and other groups are serving to further clarify our views about the origin and evolution of the mitochondrial genome. One example involves the relationship between land plant and Chlamydomonas mtDNAs, which are so different in structure, organization and mode of expression that they show little evidence of having a common evolutionary origin (1,2,74). In the absence of a phylogenetically broad database of comparative information we at one time entertained the possibility that the plant mitochondrial genome might have had a different, more recent evolutionary ancestry than Chlamydomonas and other mitochondrial genomes (75). However, sequencing of P.wickerhamii (48) and other (24,34,61) protist mtDNAs has clearly demonstrated that plant mtDNA has retained an ancestral pattern that has evidently been lost in the more rapidly evolving and highly derived Chlamydomonas mtDNA (74). It is worth emphasizing that the majority of the protist mtDNAs sequenced to date by the OGMP, particularly those from more obscure protists selected from the wild on the basis of ultrastructural or other phylogenetic considerations, retain a more or less ancestral pattern of gene content and organization. In contrast, most of the mtDNAs that had been sequenced prior to the inception of the OGMP (those from animals, most fungi, chlamydomonadalean green algae, ciliates and trypanosomatid protozoa) are highly derived. It is curious that the majority of the protists that have been selected as models for biochemical, genetic and molecular biological research happen to have mtDNAs that are the least representative of the ancestral form.
Organelle Genome Megasequencing Program (OGMP) (http://megasun.bch.umontreal.ca/ogmp/). The OGMP was initiated as a multi-disciplinary and inter-university consortium of Canadian investigators interested in organelle genome evolution and eukaryotic phylogeny. As currently constituted it consists of a Team (B.F.Lang, administrative coordinator; M.W.Gray, scientific coordinator; G.Burger, C.Lemieux and M.Turmel) and an Advisory Board (R.Cedergren, G.B.Golding, D.Sankoff, T.G.Littlejohn and C.J.O'Kelly), with external collaborators on some individual projects. The experimental arm of the OGMP, the Sequencing Unit (directed by G.Burger), is located in the Département de Biochimie, Université de Montréal. The Sequencing Unit comprises two divisions: Molecular Biology (I.Plante, D.Saint-Louis and Y.Zhu), which constructs clone libraries, performs the actual sequencing and works out improved cloning and sequencing methods; Informatics (N.Brossard and P.Rioux), which develops and implements tools required for project management, data handling, sequence analysis and annotation. As the data production arm of the OGMP, the Sequencing Unit delivers analyzed and fully annotated mitochondrial genome sequences for submission to public domain databases. The OGMP website (URL given above) contains additional information about the program, as well as data summaries and gene maps for the individual OGMP sequencing projects completed to date (Table 1).
We thank J.Chesnick, R.W.Lee, P.J.Myler, K.Stuart, Y.Tanaka and D.R.Wolstenholme for providing unpublished sequence data and other information in advance of publication. We are also grateful to C.J.O'Kelly and T.Nerad for provision of organisms and for invaluable advice on phylogeny, taxonomy and culture conditions. The OGMP has been supported by a Special Project grant (SP-34) from the Medical Research Council of Canada and by a grant (G0-12323) from the Canadian Genome Analysis and Technology Program (CGAT), which has also provided funding to GOBASE (GO-12984). Generous grants of equipment from Sun Microsystems Inc. and LI-COR Inc. are gratefully acknowledged, as is salary and interaction support from the Canadian Institute for Advanced Research.
Nucleic Acids Research
Pages
Introduction
Scope Of The Review
Methodology
Data collection and analysis
Results And Discussion
Mitochondrial genome structure
Gene content
Ribosomal RNA
Transfer RNAs and the genetic code
Introns
A comparative genomics approach to gene identification: the case of orfB and atp8
Phylogenetic implications
Descriptions
Acknowledgements
References
bDuplicate identical genes.
cDuplicate non-identical genes.
dTriplicate genes, two of which are identical, the third differing by a single T -> C transition.
eGenome specifies a single trnM(cau).
fC in the first position of the anticodon presumed to be modified to lysidine, which converts the tRNA to an AUA-decoding isoleucine acceptor (97).
gA in first the position of the anticodon presumed to be modified to inosine, with the resulting tRNA able to pair with codons ending in C, U and A, and perhaps also G (see 98).
htrnK(cuu), the corresponding tRNA of which would be expected to recognize AAG but not AAA (61).
iOnly UGG Trp codons appear in conserved protein coding genes in S.pombe mtDNA, however, several UGA codons occur in rps3 and intron ORFs (92).
jBoth UGG and UGA are decoded as Trp in A.castellanii mitochondria (61), whereas the tRNA specified by trnW(cca) would be expected to recognize only UGG.
kIncludes a trnL(aag) not listed in the table.
lIncludes a presumptive trnE pseudogene, unrelated in sequence to authentic trnE.
mIncludes a trnI(uau) not listed in the table.
nTranscripts of most Aca mitochondrial tRNA genes (12 of 15) undergo substitutional RNA editing at one or more of the first three positions of the acceptor stem (61,64; D.H.Price and M.W.Gray, unpublished results). Transcripts of at least half of the Ddi mitochondrial tRNA genes are predicted to undergo a similar type of editing.°Includes a trnX(uuua) pseudogene (D.H.Price and M.W.Gray, unpublished results), the transcript of which is predicted to have an 8 nt anticodon loop (61).
pIncludes an unusual tRNA-like element whose anticodon sequence would pair with UAA and UAG (99), which are normally termination codons.
qIncludes a trnI(aau) not listed in the table.
rIncludes a trnX(cua), the corresponding tRNA of which would be expected to recognize UAG (normally a termination codon).
sIncludes a trnL(gag) not listed.
bA group I intron in nad5 contains nad1 and nad3 genes (100).
REFERENCES
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 6 Feb 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. J. Heidel and G. Glockner Mitochondrial Genome Evolution in the Social Amoebae Mol. Biol. Evol., July 1, 2008; 25(7): 1440 - 1450. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Lynch, B. Koskella, and S. Schaack Mutation pressure and the evolution of organelle genomic architecture. Science, March 24, 2006; 311(5768): 1727 - 1730. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Hauth, U. G. Maier, B. F. Lang, and G. Burger The Rhodomonas salina mitochondrial genome: bacteria-like operons, compact gene arrangement and complex repeat region Nucleic Acids Res., August 5, 2005; 33(14): 4433 - 4442. [Abstract] [Full Text] [PDF]
|
|
|
|