The Complete Nucleotide Sequence of a Snake (Dinodon semicarinatus) Mitochondrial Genome With Two Identical Control Regions

Corresponding author: Yoshinori Kumazawa, Department of Earth and Planetary Sciences, Graduate School of Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8602, Japan., h44858a{at}nucc.cc.nagoya-u.ac.jp (E-mail).

Communicating editor: N. TAKAHATA

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

The 17,191-bp mitochondrial DNA (mtDNA) of a Japanese colubrid snake, akamata (Dinodon semicarinatus), was cloned and sequenced. The snake mtDNA has some peculiar features that were found in our previous study using polymerase chain reaction: duplicate control regions that have completely identical sequences over 1 kbp, translocation of tRNA^Leu(UUR) gene, shortened TC arm for most tRNA genes, and a pseudogene for tRNA^Pro. Phylogenetic analysis of amino acid sequences of protein genes suggested an unusually high rate of molecular evolution in the snake compared to other vertebrates. Southern hybridization experiments using mtDNAs purified from multiple akamata individuals showed that the duplicate state of the control region is not a transient or unstable feature found in a particular individual, but that it stably occurs in mitochondrial genomes of the species. This may, therefore, be regarded as an unprecedented example of stable functional redundancy in animal mtDNA. However, some of the examined individuals contain a rather scanty proportion of heteroplasmic mtDNAs with an organization of genes distinct from that of the major mtDNA. The gene organization of the minor mtDNA is in agreement with one of models that we present to account for the concerted evolution of duplicate control regions.

SINCE the early 1980s, complete mitochondrial DNA (mtDNA) sequences have been determined for a number of vertebrates (ANDERSON et al. 1981 ; WOLSTENHOLME 1992 ; LEE and KOCHER 1995 ; JANKE and ARNASON 1997 ; ZARDOYA and MEYER 1997 and references therein). However, these examples are extremely biased toward mammals and fish, with the exception of the chicken (DESJARDINS and MORAIS 1990 ), ostrich (HARLID et al. 1997 ), African clawed toad (ROE et al. 1985 ), and alligator (JANKE and ARNASON 1997 ). The vertebrate mtDNAs characterized to date are double-stranded, circular DNAs of 16–17 kbp, they encode genes for 13 proteins, two rRNAs, and 22 tRNAs, and they have a major noncoding or control region that contains signals for replicating the heavy strand of mtDNA and for transcription (e.g., see CLAYTON 1992 ; WOLSTENHOLME 1992 ). These genomes are, therefore, conserved with respect to gene content. No repetition or deletion of any of the genes or the control region has been shown.

In the 1980s, the gene order of the vertebrate mtDNAs was also considered to be conserved. This, however, no longer holds true in light of recent findings of distinct mtDNA gene orders for marsupials (PAABO et al. 1991 ; JANKE et al. 1994 ), birds (e.g., see DESJARDINS and MORAIS 1990 ; QUINN and WILSON 1993 ; HARLID et al. 1997 ), crocodilians (SEUTIN et al. 1994 ; KUMAZAWA and NISHIDA 1995 ; QUINN and MINDELL 1996 ; JANKE and ARNASON 1997 ), agamas (MACEY et al. 1997 ), the Texas blind snake (KUMAZAWA and NISHIDA 1995 ), the bullfrog (YONEYAMA 1987 ), the rice frog (MACEY et al. 1997 ), and sea lamprey (LEE and KOCHER 1995 ). In these examples of gene rearrangements, there is no evidence of duplication or deletion of specific genes or the control region element, but rather, simply the reordering of the same 37 genes and the control region. Although some tRNA genes are tandemly duplicated in amphisbaenian mtDNA, one of the redundant gene copies may be a pseudogene (MACEY et al. 1998 ).

In our recent work (KUMAZAWA et al. 1996 ), however, it was found that mtDNAs of several snakes (the viperids himehabu and western rattlesnake, the colubrid akamata, and the boids ball python and boa constrictor) have duplicate control regions. This finding was based on polymerase chain reaction (PCR) amplification and sequencing of two separate mtDNA regions: one surrounded by cytochrome b (cytb) and 12S rRNA genes, and the other surrounded by NADH dehydrogenase subunits 1 and 2 (ND1 and ND2, respectively) genes. Notably, in the former three species, the duplicate control regions were found to share nearly the same nucleotide sequence.

This finding raised profound evolutionary questions. We considered that the original duplication of the control region took place in rather ancient lineages of snakes because it seems unlikely that the insertion of a control region into a distant and unique location occurred in parallel in multiple snake lineages. Rather, it is more likely that the control region had already been duplicated in mtDNAs of the last common ancestor of the above-mentioned five snakes, which existed at least 70 mya according to fossil records for the minimum divergence time between the Boidae and the Viperidae-Colubridae clade (RAGE 1987 ). If so, it is mysterious that the duplicate state of the control region may have been maintained for such a long period of time in snake mitochondrial genomes because any redundant genes or sequences in mtDNAs, including one of the duplicate control regions, would most likely be deleted rapidly (see KUMAZAWA et al. 1996 for the reasoning).

Another mysterious point is that the control region sequences were nearly identical between the duplicate copies within the species while they were otherwise divergent to a large extent between the species. If the paralogous control regions within an mtDNA molecule had evolved independently since the presumed original insertion event (>70 mya), a much more substantial sequence difference would be expected. It was thus proposed that the duplicate control regions have evolved in a highly concerted fashion (KUMAZAWA et al. 1996 ). However, unlike the commonness of the concerted evolution in nuclear DNAs, examples of such concerted evolution in mtDNA have not been reported from animals other than the snakes. Currently available information is too scanty to reveal the mechanism of this bizarre phenomenon, making it intriguing to further characterize the concerted evolution in snake mtDNAs.

In this communication, we report an entire mtDNA nucleotide sequence of a Japanese colubrid, akamata. On the basis of the complete gene organization of snake mtDNA, we present models for the concerted evolution of snake control regions. These models are assessed by Southern hybridization experiments that help to detect minor mtDNA components, and by cloning and sequencing of such minor molecules. The Southern hybridization experiments also indicated that the duplicate state of the control region is stable in character and shared by multiple individuals of the species. This may, therefore, be regarded as an unprecedented example of stable functional redundancy in animal mtDNA. Determination of the entire sequence of snake mtDNA also led to an unexpected finding that molecular evolutionary rates of protein genes encoded in snake mtDNA are unusually high.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

Akamata (Dinodon semicarinatus) is a nonvenomous colubrid snake that is endemic to the central Ryukyus of Japan. mtDNAs were analyzed for four individuals from Okinawajima Island (designated Okinawa 1–4), two individuals from Tokunoshima Island (Toku 1–2), and two individuals from Amamioshima Island (Amami 1–2). These islands have long been isolated from each other, and it thus seems that there has recently been no gene flow between populations of these islands in nature (OTA 1986 ). mtDNA was isolated from fresh liver essentially as described by TAPPER et al. 1983 . Briefly, the liver (2–10 g, depending on body size) was homogenized in 250 mM sucrose, 100 mM Tris-HCl, and 10 mM EDTA (pH 7.6), and the mitochondrial fraction was pelleted by the differential centrifugation, followed by lysis with SDS, phenol extraction, and ethanol precipitation. Closed circular mtDNA was recovered by ethidium bromide-CsCl equilibrium density gradient centrifugation. After removal of ethidium bromide, the separate fractions were desalted and concentrated with a Centricon-30 centrifugal microconcentrator (Amicon, Beverly, MA).

Akamata mtDNA was cut with restriction enzymes and cloned into Escherichia coli pUC19 or pBR322 vector. Figure 1A shows cloned mtDNA fragments and their arrangements. Clones for three kinds of XbaI-digested fragments (3.9, 1.3, and 1.3 kbp) were obtained. Two kinds of XbaI/EcoRI-digested fragments (4.4 and 1.7 kbp) were also cloned. As described in detail in RESULTS AND DISCUSSION, the remaining portion corresponding to a 4.6-kbp EcoRI-cut fragment was refractory to the cloning. Instead, a 5.9-kbp fragment resulting from partial digestion of akamata mtDNA with EcoRI and BglII was cloned successfully. Both strands of these cloned mtDNA fragments were sequenced with a DNA sequencer (model 373A; Applied Biosystems, Foster City, CA) using the primer walking strategy. All overlaps between the clones were confirmed by direct sequencing of PCR products amplified from the purified mtDNA template.

View larger version (34K): In this window In a new window Download PPT slide   Figure 1. (A) Gene organization and restriction map of the akamata mtDNA. A 17,191-bp circular mtDNA is depicted as a ring that is divided into individual genes and control regions. Transfer RNA genes are shown as single letters standing for the corresponding amino acid, and those whose sense strand is the light and heavy strand are shown outside and inside the ring, respectively. L1, L2, S1, S2, and P* represent genes for tRNALeu(UUR), tRNALeu(CUN), tRNASer(AGY), tRNASer(UCN), and a pseudogene for tRNAPro, respectively. Abbreviations for the other genes are as follows: ND1-6, NADH dehydrogenase subunits 1–6; CO I–III, cytochrome oxidase subunits I–III; ATP6, ATPase subunit 6; ATP8, ATPase subunit 8; and cytb, cytochrome b. OH and OL stand for an origin of heavy- and light-strand replication, respectively. Sites for restriction enzymes XbaI, EcoRI, BglII, and PstI are shown, and restriction fragments that were cloned and sequenced are indicated by bold lines. The regions corresponding to the hybridization probes A and B are also shown outside the ring. (B) Gene organization of the 4.6-kbp fragment with EcoRI termini, which was unexpectedly cloned from the Okinawa 2 individual and may have been derived from a minor mtDNA component (see text for details).

Southern hybridization was done with a chemiluminescence detection system (ECL direct nucleic acid labeling and detection systems; Amersham, Arlington Heights, IL) according to the manufacturer's instructions. Akamata mtDNA (50–150 ng) was cut with appropriate restriction enzymes, electrophoresed on a 0.4% agarose gel, transferred to a nylon membrane (Pall Biodyne A transfer membranes, 0.2-µm pore), and hybridized with a peroxidase-attached DNA probe. After hybridization at 42° overnight in the manufacturer's hybridization buffer (ECL gold buffer), to which 0.5 M NaCl and 5% blocking agent were added, the filter was washed twice at 42° for 20 min with the primary wash buffer containing 6 M urea, 0.4% SDS, and 0.3x SSC, and then washed twice at room temperature for 5 min with 2x SSC. In this study, two kinds of probes were prepared from heat-denatured PCR products: probe A, which corresponds to a sequence containing the tRNA^Val gene and parts of flanking rRNA genes (positions 880–1078 in Figure 2), and probe B, which corresponds to a sequence inside the control region (positions 4057–4225 and 16,628–16,796). The chemiluminescent signal was detected with Kodak XAR-5 film. After hybridization with probe A, the blot was stored at room temperature for a few days to allow the chemiluminescent signal to decay completely. The blot was then washed with 2x SSC and reprobed with probe B.

View larger version (346K): In this window In a new window Download PPT slide   Figure 2. Complete nucleotide sequence of the akamata mitochondrial genome. The light-strand sequence of the 17,191-bp akamata mtDNA is presented and numbered from the 5' end of tRNAPhe gene. The starting and ending positions predicted for each gene are shown with () or (), together with an arrow that indicates the polarity of the gene. Protein-coding genes are translated into amino acid sequences by the mammalian genetic code. Note that stop codons for ND1, ND3, CO II, CO III, and cytb genes appear by polyadenylation. Anticodons of tRNA genes and recognition sites by major restriction enzymes XbaI, EcoRI, BglII, and PstI are underlined. Also underlined are several structural features that exist within the control regions and the origin for light-strand replication. Note that the sequences of two control regions corresponding to 3602–4619 and 16,173–17,190 are completely identical. A guanosine at position 3750, a thymidine at 16,223, and a guanosine at 16,321 were sequenced to be an adenosine, a cytidine, and an adenosine, respectively, for a different individual collected from the same locality of Okinawajima Island (KUMAZAWA et al. 1996 ). Note that CSB II was not identified clearly in our previous study (KUMAZAWA et al. 1996 ).

Genes in the determined mtDNA sequences were identified in light of their sequence similarity with the corresponding genes from other vertebrates, as well as structural features of mitochondrial tRNA genes (e.g., KUMAZAWA and NISHIDA 1993 ). The entire nucleotide sequence of the akamata mtDNA will appear in the DDBJ, EMBL, and GenBank nucleotide sequence databases under the accession no. AB008539.

	RESULTS AND DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

Gene organization:
Figure 1A shows the gene organization of akamata mtDNA derived from its complete nucleotide sequence (Figure 2). Akamata mtDNA is 17,191 bp long and contains 13 protein-coding genes, two rRNA genes, and 22 tRNA genes that are commonly encoded in the mtDNAs from other vertebrates. It also contains two control regions that, in other vertebrates, occur in a single conservative location surrounded by cytb and 12S rRNA genes. The snake genome conserves the control region in this location (control region I) and, in addition, possesses another control region (control region II) within a tRNA gene cluster where genes for tRNA^Ile, tRNA^Gln, and tRNA^Met are typically arrayed in other vertebrates (IQM region). The nucleotide sequences of the two control regions of akamata mtDNA are completely identical over 1 kbp (Figure 2). Because our previous study (KUMAZAWA et al. 1996 ) used PCR to show the existence of two control regions, we were anxious that we may have analyzed something like PCR artefacts, especially resulting from the jumping PCR phenomenon (PAABO et al. 1990 ). However, the conventional cloning-and-sequencing strategy was used in the present study, which unambiguously proves that the snake mtDNA has two control regions within a single molecule. This conclusion was additionally confirmed by Southern hybridization experiments (see below).

Besides this anomaly in the control region, akamata mtDNA has two distinctive features in gene organization: the tRNA^Leu(UUR) gene, which occurs between genes for 16S rRNA and ND1 in other vertebrates, is translocated to the 3' vicinity of control region II, and a pseudogene for tRNA^Pro exists in the 5' vicinity of control region II (Figure 1A). Thus, the alteration in gene organization is centered on the IQM region, where insertion of an array of tRNA^Pro pseudogene, control region, and tRNA^Leu(UUR) gene occurs between genes for tRNA^Ile and tRNA^Gln. No other distinctive features from the typical vertebrate organization for placental mammals, the African clawed toad, and bony fish are found. A stem-and-loop structure for the putative origin for replicating the light strand (WONG and CLAYTON 1985 ) appears to be present between genes for tRNA^Asn and tRNA^Cys (Figure 2).

The size of akamata mtDNA (17,191 bp) is nearly equal to or slightly larger than those of mtDNAs from other vertebrates reported to date, and the sizes of individual structural genes are mostly common between akamata and other vertebrates (data not shown). The base composition of akamata mtDNA (Table 1) is skewed similarly to those of other metazoan mtDNAs, i.e., generally more A-T base pairs than G-C base pairs, and more A and C contents (accordingly, fewer G and T contents) in the light strand than in the heavy strand (e.g., see ASAKAWA et al. 1991 ).

Control region:
Each of the two control regions found in akamata mtDNA is 1018 bp long. This is the typical length for the control region of vertebrate mtDNAs known to date. Conserved sequence blocks (CSBs) I, II, and III have been known as conserved sequence elements among mammalian control regions (WALBERG and CLAYTON 1981 ). All or partial members of the CSBs are also identifiable for control regions of vertebrates other than mammals (e.g., ROE et al. 1985 ; QUINN and WILSON 1993 ; LEE et al. 1995 ; ZARDOYA and MEYER 1997 ). The akamata control region contains sequences that show an appreciable level of similarity with CSBs I, II, and III of other vertebrates (Figure 2). However, unlike other vertebrates, CSBs II and III appear to be tightly connected with each other. In light of experimental evidence obtained for mammals (CLAYTON 1992 and references therein), the displacement loop (D-loop) region of akamata mtDNA may start from 3' downstream of CSB III and end in the vicinity of termination-associated sequences (TASs; DODA et al. 1981 ), two copies of which are identifiable at the 5' end of the snake control region (Figure 2). As is often the case with other vertebrates, these TASs are situated in a region where hairpin-like secondary structures and repetitive sequences occur (Figure 2).

5' upstream of the hairpins and repeats, there exists a C-rich sequence that commonly occurs among several snakes so far examined (Figure 2; KUMAZAWA et al. 1996 ). Sequencing reactions for both strands tend to stall in the C-rich region (data not shown), suggesting that this is a structural barrier for DNA polymerization. The C-rich sequence may facilitate formation of the D-loop by decelerating the extension of heavy-strand synthesis at this location. The C-rich sequence is not found clearly in the corresponding locations of mammals, frog, and fish. Birds have similar sequences, but they have been interpreted to be involved in a hairpin structure (QUINN and WILSON 1993 ). As described below, snake mtDNAs possess, just 5' upstream of the control region, rather stable hairpin-like structures derived from the tRNA^Pro gene or pseudogene or an intergenic spacer. This hairpin may also serve as a structural barrier in DNA polymerization and play a role in facilitating the formation of the D-loop concertedly with TASs, the C-rich sequence, and other hairpins at the 5' end of the control region.

Replication and transcription:
In mammals, the control region contains all the template information required for the initiation of nascent heavy-strand synthesis, as well as the essential major cis-acting elements necessary for transcriptional initiation (CLAYTON 1992 ; and references therein). Because the nucleotide sequences of control regions I and II are completely identical (Figure 2), it is reasonable to infer that heavy-strand synthesis and transcription for both strands can be initiated from two separate locations in the snake mtDNA, whereas light-strand synthesis preceded by heavy-strand synthesis begins from a single location (Figure 1A). This apparently serves to elevate the efficiency of DNA replication and gene expression in the snake mitochondria.

Mammalian mtDNAs maintain a transcriptional attenuator-like, 13-nucleotide sequence at the 5' end of the tRNA^Leu(UUR) gene (CHRISTIANSON and CLAYTON 1988 ). Light-strand RNA transcripts terminate at the end of the 16S rRNA gene much more frequently than they extend beyond this site to make a polycistronic RNA precursor for all the light-strand gene products (GELFAND and ATTARDI 1981 ; OJALA et al. 1981 ). This attenuator is thought to regulate expression levels of 12S and 16S rRNAs, which should be much higher than those of tRNAs and mRNAs. Because the snake tRNA^Leu (UUR) gene (Figure 3D) is translocated to the 3' end of control region II, the attenuation mechanism at the 3' end of the 16S rRNA gene may not be operative in snake mitochondria. It is currently unknown how the high expression levels of rRNAs are maintained in the snake mitochondria. One possibility is that light-strand transcripts may not be able to extend beyond control regions because of the presence of a transcriptional terminator within this region, and that light-strand RNA synthesis from control region I may be regulated to be more efficient than that from control region II.

View larger version (21K): In this window In a new window Download PPT slide   Figure 3. Several tRNA genes and relevant structural features found in akamata mtDNA. (A) tRNAPro gene (positions 16,111–16,172 in Figure 2). (B) tRNAPro pseudogene (3566–3601). (C) An alternative secondary structure for the tRNAPro pseudogene (3566–3601). (D) tRNALeu(UUR) gene (4620–4692). (E) tRNAGlu gene (14,866–14,927). (F) An antisense sequence of the tRNAGlu gene (14,864–14,931). (G) tRNAPhe gene (1–65). (H) An alternative secondary structure for the tRNAPhe gene (1–70). Bars in stems represent Watson-Crick basepairs, and dots stand for wobble G-U pairs. Note that the guanosine indicated by an arrow in B and C, which is the only difference from the 5' half portion of the tRNAPro gene shown in A, was not found for a different individual from the same locality of Okinawajima Island (KUMAZAWA et al. 1996 ).

Transfer RNA genes:
Akamata mtDNA encodes 22 tRNA genes and a pseudogene of tRNA^Pro. The anticodon triplet sequences of the 22 tRNA genes (Figure 2) are exactly the same as their counterparts for mammals, chicken, and frog. It is therefore likely, though not conclusive, that the snake mitochondria use the same genetic code as mammalian mitochondria (ANDERSON et al. 1981 ), and that this set of tRNAs is capable of decoding all 60 sense codons of the mitochondrial genetic code without importing nuclear-cytosolic tRNAs.

Considerable truncation of the TC arm (T arm), previously described for snake mitochondrial tRNAs (KUMAZAWA et al. 1996 ), is also notable in akamata mitochondrial tRNA genes, most of which have <5 bp in the T stem. For instance, genes for tRNA^Pro (Figure 3A), tRNA^Glu (Figure 3E), and tRNA^Phe (Figure 3G) have only 3 bp in the T stem. Together with small numbers of nucleotides assigned for the T loop region, this results in considerably truncated T arm regions for the snake tRNAs (Figure 4). For 16 tRNAs from tRNA^Pro to tRNA^Ile in Figure 4, the number of nucleotides in the T arm is smaller in the snake than in other vertebrates and starfish. For the four remaining tRNAs, i.e., leucine tRNAs, tRNA^Asn, and tRNA^Gln, the figure may be constrained to be 17 because of the need for stabilization of the dihydrouridine loop (D loop)/T loop interaction via conserved sequence elements, such as GG [or AG for tRNA^Leu(CUN)] in the D loop and TCR [or GCA for tRNA^Leu(CUN)] in the T loop (Figure 2). These results strengthen a previous notion (KUMAZAWA et al. 1996 ) that snake mitochondrial tRNA genes may have evolved to be somewhat simpler in the T arm structure, which is associated with the loss of the canonical tertiary interactions between the D and T loops.

View larger version (26K): In this window In a new window Download PPT slide   Figure 4. Simplification of the T arm in snake mitochondrial tRNA genes. The number of nucleotides in the T arm of mitochondrial tRNA genes was compared among the mouse (BIBB et al. 1981 ), chicken (DESJARDINS and MORAIS 1990 ), African clawed toad (ROE et al. 1985 ), akamata (this study), and starfish (ASAKAWA et al. 1995 ). The comparison was made for all tRNA genes, except for two serine isoacceptors in which a common secondary structure model cannot be defined among vertebrates and starfish (e.g., see KUMAZAWA and NISHIDA 1993 ). The assignment of the T arm region (T stem plus T loop) was done according to the method of KUMAZAWA and NISHIDA 1993 . Refer to Figure 1 for the abbreviations of the tRNA genes.

Because the snake mitochondrial tRNAs do not exactly conform to the canonical secondary structure model, there remains an ambiguity in identifying the tRNA^Phe gene. The tRNA^Phe gene located in the 3' vicinity of control region I (Figure 2) can be folded into two alternative secondary structures (Figure 3G and Figure H), both of which have considerable mismatches in the acceptor stem region. Although a few mismatched pairs may be tolerated in the acceptor stem of mitochondrial tRNAs (KUMAZAWA and NISHIDA 1993 ), mismatches in these cases occur at the acceptor end, in which the structural destabilization is crucial to the function of tRNAs. Another possible candidate for the tRNA^Phe gene exists as an antisense sequence of the tRNA^Glu gene (Figure 3F). The antisense sequence appears to include a proper anticodon loop for tRNA^Phe with more stable acceptor stem structures than the putative tRNA^Phe genes described above. One shortcoming in the antisense gene is the shortness of the variable loop (only two nucleotides), which may cause difficulty in tertiary interactions involving nucleotides in the variable loop. We thus consider that the most plausible tRNA^Phe gene is the one shown in Figure 3G, under the assumption that RNA editing corrects mismatched base pairs in the acceptor end. RNA editing at the acceptor stem is known to occur for mitochondrial tRNAs of pulmonate gastropods and mammals (YOKOBORI and PAABO 1995A , YOKOBORI and PAABO 1995B ).

Pseudogene for tRNA^Pro:
The tRNA^Pro pseudogene (Figure 3B) consists of a 5' half portion of the tRNA^Pro gene that exists immediately 5' upstream of control region I (Figure 3A). The insertion of a guanosine into the anticodon loop, as indicated by the arrow in Figure 3B, is the only difference between these two paralogous sequences. It seems improbable that a functional tRNA^Pro molecule is expressed from the half-sized gene sequence. Instead, it can be folded into a rather stable hairpin-like structure, as shown in Figure 3C. As discussed above, this structure may facilitate pausing of heavy-strand replication for the formation of the D-loop within control region. In a viperid snake, himehabu, which was also shown to possess duplicate control regions (KUMAZAWA et al. 1996 ), a functional tRNA^Pro gene is found 5' upstream of control region II, and a similar tRNA^Pro pseudogene corresponding to its 5' half portion exists 5' upstream of control region I. In the rattlesnake, instead of the half-sized tRNA^Pro pseudogene, a long intergenic spacer exists in the 5' vicinity of control region I (KUMAZAWA et al. 1996 ). This spacer contains the sequence TTTATAGTAAATACTGTAAA, which may form a stable hairpin structure by use of the complementarity between the underlined parts, and this may substitute for the possible role of the tRNA^Pro pseudogene sequence.

Protein genes:
Translated amino acid sequences for 12 light-strand protein genes were concatenated and used for phylogenetic analyses among akamata and other vertebrates. A maximum likelihood tree (Figure 5) was obtained by the quartet puzzling (QP) algorithm (STRIMMER and VON HAESELER 1996 , PUZZLE version 4.0), assuming the mtREV24 substitution model (ADACHI and HASEGAWA 1996 ) and the gamma-distributed rates of substitution over sites (OTA and NEI 1994 ). The tree is in agreement with previous studies (e.g., ADACHI and HASEGAWA 1996 ; JANKE and ARNASON 1997 ; ZARDOYA and MEYER 1997 ) with respect to topology and branch length relating to taxa other than akamata. Akamata clusters with archosaurian taxa (i.e., birds and crocodiles) relative to mammals, which is supported with relatively high QP reliability values (>84%) that have the same practical meaning as bootstrap values (STRIMMER and VON HAESELER 1996 ). This phylogenetic relationship is congruent with a traditional one in which snakes are affiliated to Reptilia, Squamata (RAGE 1987 ).

View larger version (13K): In this window In a new window Download PPT slide   Figure 5. Maximum likelihood tree constructed from mitochondrial protein sequences. Aligned amino acid sequences for 12 light-strand protein genes from which gap sites were deleted were concatenated (3444 sites in total) and analyzed by PUZZLE version 4.0 (STRIMMER and VON HAESELER 1996 ). Shown here is the maximum likelihood tree by the QP algorithm (STRIMMER and VON HAESELER 1996 ) assuming the mtREV24 substitution model (ADACHI and HASEGAWA 1996 ) and the gamma-distributed rates of substitution over sites (OTA and NEI 1994 ). The shape parameter of the gamma distribution (0.38) was calculated from the data set. Values along internal branches are QP reliability values in percent obtained from 1000 puzzling steps (STRIMMER and VON HAESELER 1996 ). A scale bar indicating 0.2 substitutions per site is shown. Data source: loach, TZENG et al. 1992 ; trout, ZARDOYA et al. 1995 ; coelacanth, ZARDOYA and MEYER 1997 ; lungfish, ZARDOYA and MEYER 1996 ; Xenopus, ROE et al. 1985 ; mouse, BIBB et al. 1981 ; rat, GADALETA et al. 1989 ; human, ANDERSON et al. 1981 ; cow, ANDERSON et al. 1982 ; fin whale, ARNASON et al. 1991 ; opossum, JANKE et al. 1994 ; platypus, JANKE et al. 1996 ; akamata, this study; alligator, JANKE and ARNASON 1997 ; chicken, DESJARDINS and MORAIS 1990 ; and ostrich, HARLID et al. 1997 .

A striking feature in Figure 5 is that a terminal branch leading to akamata is rather long. The branch length measured from the latest common ancestor between snakes and birds is roughly three times larger for akamata than for birds, as averaged between chicken and ostrich. Moreover, pairwise distances among the sequence data computed with various substitution models were substantially larger between akamata and other taxa than among nonserpentine animals (data not shown). This tendency was seen not only for the concatenated sequences, but for individual protein sequences (data not shown). These observations suggest considerably higher rates of molecular evolution in the snake lineage. It has recently been shown that crocodilian mtDNAs evolve faster than mtDNAs of most other vertebrates (KUMAZAWA and NISHIDA 1995 ; JANKE and ARNASON 1997 ), but the rate in snakes appears to be even higher than that in crocodilians (Figure 5). Snakes and crocodilians are cold-blooded reptiles and, thus, the cold bloodedness may not necessarily lead to deceleration of molecular evolution despite suggestions by some earlier studies (THOMAS and BECKENBACH 1989 ; ADACHI et al. 1993 ; MARTIN and PALUMBI 1993 ). The mechanisms of the rate acceleration in snake mtDNA remain unclear. One possibility is that a replicational mechanism using two control regions (see above) may not be very accurate.

The tree topology shown in Figure 5 was not obtained when the same data set was analyzed on the assumption of uniform rate over sites or two rates (i.e., mixture of variable sites with uniform rate and invariable sites). In these cases, akamata clustered with the alligator relative to birds (data not shown), which most likely represents a false phylogeny resulting from considerable rate fluctuation among lineages, as described above. The Archosauria clade has been supported by various morphological and molecular studies (e.g., CARROLL 1988 ; HEDGES 1994 ; KUMAZAWA and NISHIDA 1995 ; JANKE and ARNASON 1997 ), and thus the inclusion of snakes in this clade is very unlikely. It seems reasonable that the gamma-distributed rates were used for analyzing the mitochondrial protein data in Figure 5 because an estimated shape parameter of the gamma distribution for the data set (0.38) points to considerable rate heterogeneity over sites. However, we consider that future phylogenetic studies using akamata mtDNA sequences should be done very carefully because of the unusually high rates of sequence evolution. Simulation studies (NEI 1991 ) pointed to the difficulty of reconstructing the correct phylogeny by currently used tree-building methods when extreme rate variations exist among lineages.

Cloning of an unexpected fragment:
During the course of cloning mtDNA fragments as described in MATERIALS AND METHODS, a 4.6-kbp EcoRI-cut fragment covering from ND5 to 12S rRNA genes (Figure 1A) was refractory to cloning using EcoRI-cut E. coli vectors. Uncloneable mtDNA fragments are known to exist (e.g., see DROUIN 1980 ; MITA et al. 1988 ; ASAKAWA et al. 1995 ), and suspected reasons for this are inhibitory effects of cloned mitochondrial genes on the growth of E. coli as well as profound structural contexts near restriction sites. Repeated rounds of the cloning trial actually generated clones having a 4.6-kbp insert, albeit in a low efficiency (five clones in total after three rounds of the cloning trial using an electroporation apparatus). These clones were initially taken for the wanted ones owing to the reasonable size of the inserts, and one of the clones was sequenced completely.

It turned out that the 4.6-kbp insert consists of an entirely unexpected sequence starting from an EcoRI site inside the 12S rRNA gene and continuing to genes for tRNA^Val, 16S rRNA, ND1, tRNA^Ile, a pseudogene for tRNA^Pro, control region, tRNA^Phe, and the remaining 5' portion of the 12S rRNA gene (Figure 1B). This sequence corresponds exactly with a concatenated sequence between positions 275–4619 and positions 17191–274 in Figure 2, except for a single base difference: a nucleotide corresponding to a cytidine at position 266 is an adenosine for the 4.6-kbp insert. The control region sequence in this insert is thus surrounded by ND1 and 12S rRNA genes, and this organization is distinct from those around control regions I and II (see Figure 1A and Figure B). Sequences of the other four clones determined only from EcoRI termini were identical to that of the first clone. Moreover, inserts of these five clones were not cut with PstI, suggesting no relation to the 4.6-kbp EcoRI fragment ranging from ND5 to 12S rRNA genes.

The recovery of multiple clones having an unexpected sequence necessitated us to confirm the gene organization of akamata mtDNA more carefully, as well as to examine heterogeneity of mtDNA sequences both between and within individuals. These problems were addressed by Southern hybridization experiments, and the results are presented in the next two sections. The origin of the unexpected 4.6-kbp fragment is discussed in the section titled Minor mtDNA components (see below).

Confirmation of the gene organization and its stability within species by Southern hybridization:
In the Southern hybridization experiments shown in Figure 6, purified mtDNAs from eight akamata individuals (Okinawa 1–4, Amami 1–2, and Toku 1–2) were used, and one of them (Okinawa 2) corresponds to an individual from which the complete mtDNA sequence (Figure 2) was determined. In Figure 6A, mtDNA purified from Okinawa 4 was cut with various restriction enzymes and hybridized with two kinds of probes, as described in MATERIALS AND METHODS. The sizes of mtDNA fragments generated by the enzyme digestions, as well as the profile of hybridized bands with the probes A and B, corresponded well with expectations from the physical map (Figure 1A) based on the complete mtDNA sequence (see legend of Figure 6). In addition, this digestion pattern for mtDNA of Okinawa 4 was in agreement with the pattern for mtDNA of Okinawa 2 (data not shown) that had been obtained before its cloning and sequencing. Because probe B matches a part of the control region sequence, its hybridization to two distinct bands is direct evidence for the existence of duplicate control regions within the snake mtDNA.

View larger version (126K): In this window In a new window Download PPT slide   Figure 6. Southern hybridization of restriction fragments of akamata mtDNA. Samples analyzed are akamata mtDNAs from Okinawa 4, which were digested with various restriction enzymes (A), and akamata mtDNAs from eight individuals, which were uncut (B), digested with PstI (C), and digested with EcoRI (D). Hybridization experiments were done as described in MATERIALS AND METHODS. Electrophoresed gels were stained with ethidium bromide (a), and their blots were hybridized with probe A (b and d) and then reprobed with probe B (c). The corresponding locations for probes A (complementary to positions 880–1078 in Figure 2) and probe B (positions 4057–4225 and 16,628–16,796) are also indicated in Figure 1. The time in parentheses indicates exposure time to a Kodak film. The arrowheads indicate faint signals discussed in the text. Size markers used are DNA digested with HindIII (HindIII) and circularized DNA digested with BamHI (BamHI). The approximate positions corresponding to several marker bands are indicated on the right and left of the figure. Sizes of observed restriction fragments in A-a are as follows: PstI (17.2 kbp), XbaI (10.7, 3.9, and a doublet of 1.3 kbp), EcoRI (12.6 and 4.6 kbp), BglII (10.3 and 6.9 kbp), EcoRI/XbaI (a mixture band of 4.6 and 4.4, 3.9, 1.7, and a doublet of 1.3 kbp), and EcoRI/BglII (10.3, 4.6, 1.3, and 1.0 kbp). Among these restriction fragments, probe A hybridized with the 17.2-kbp fragment produced by PstI, 10.7 kbp by XbaI, 12.6 kbp by EcoRI, 6.9 kbp by BglII, 4.6 and/or 4.4 kbp by EcoRI/XbaI, and 1.3 kbp by EcoRI/BglII (A-b), whereas probe B hybridized with 17.2 kbp by PstI, 10.7 kbp by XbaI, 12.6 and 4.6 kbp by EcoRI, 10.3 and 6.9 kbp by BglII, 4.6 and/or 4.4 kbp by EcoRI/XbaI, and 10.3 and 4.6 kbp by EcoRI/BglII (A-c).

When purified mtDNAs from eight individuals were electrophoresed on 0.4% agarose gel (Figure 6B-a), they were separated into a closed circular form with a higher mobility on the gel and an open circular form with a lower mobility. The proportion of the two bands varied from sample to sample, presumably because of the variation in intactness of mtDNA samples stored at -30° for different periods. Heterogeneous DNAs with the lowest mobility were apparent in some samples (e.g., lane 11 of Figure 6B-a); however, they do not represent heterogeneity of mtDNA because neither probe A nor probe B hybridize with them (Figure 5B and Figure B-C). They just represent contaminating nuclear DNAs.

All the mtDNAs from eight individuals were cut with a unique PstI site, giving rise to a ~17-kbp fragment (Figure 6C-a). When digested with EcoRI, mtDNAs from Okinawa 1–4 were cut into 12.6- and 4.6-kbp fragments (Figure 6D-a). mtDNAs from individuals of the Amami Islands (Toku 1–2 and Amami 1–2) were also cut with EcoRI into two major fragments, but of distinct sizes (~9.5 and ~7 kbp) from those of Okinawajima Island (Okinawa 1–4). To account for the difference in size of restriction fragments between the two mtDNA haplotypes by base substitutions, at least two substitutions (a loss and a gain of EcoRI site) need to be assumed. This suggests a considerably smaller genetic distance between akamata populations of the two Amami Islands than that between these populations and the Okinawajima population, which is in qualitative agreement with the deduced geohistorical relationship among the three islands (OTA 1986 ). As shown in Figure 6D, probe B hybridizes with both of the EcoRI-cut fragments for all eight individuals, irrespective of their geographic origin. This indicates that the duplicate state of the control region is not a transient or unstable feature found in a particular individual, but that it stably occurs in the mitochondrial genomes of the species.

These results validate the gene organization of akamata mtDNA presented in Figure 1A. No distinct length polymorphism was apparent in major mtDNA molecules among the eight individuals tested, although extensive length polymorphism has been known to occur in mtDNAs of some lizards, amphibians, and fish at both the intra- and interindividual levels (MORITZ and BROWN 1986 , MORITZ and BROWN 1987 ; WALLIS 1987 ; ZEVERING et al. 1991 ; STANTON et al. 1994 ; GACH and BROWN 1997 ). The duplicate state of the control region was previously shown to be maintained among representative individuals of several snake species (KUMAZAWA et al. 1996 ). On the basis of this solid evidence, the duplication of control regions in snake mtDNAs may be regarded as an unprecedented example of stable functional redundancy in animal mtDNA.

Minor mtDNA components:
As outlined earlier, we unexpectedly cloned a 4.6-kbp EcoRI-cut fragment that contains the control region sequence surrounded by ND1 and 12S rRNA genes (Figure 1B). However, the two control regions shown in Figure 1A are surrounded either by cytb and 12S rRNA genes or by ND1 and ND2 genes. If the gene organization of Figure 1A is unambiguous, from where did the 4.6-kbp fragment originate? It is evident that even though the 4.6-kbp fragment is present in mtDNA samples from Okinawa 2, it is not a part of the major 17.2-kbp mtDNA molecule because probe A does not hybridize with a major 4.6-kbp fragment generated by EcoRI-digestion of the Okinawa 2 mtDNA (see lane 10 of Figure 6D). This 4.6-kbp fragment actually corresponds to a sequence from ND5 to 12S rRNA genes, as judged from its further digestion into smaller pieces (4.0 and 0.6 kbp) with PstI (data not shown).

However, much longer exposure of the same filter to an X-ray film gave rise to a faint but discrete signal of ~4.6 kbp in size, as indicated by the arrowhead in Figure 6D. This signal was clearly detected in Okinawa 1–2, only weakly detected in Okinawa 3, and not detectable in Okinawa 4, Amami 1–2, and Toku 1–2. This signal is not caused by nonspecific hybridization of probe A to the 4.6-kbp major fragment ranging from ND5 to 12S rRNA genes because the size of the faintly hybridized signal (~4.6 kbp) was unchanged, even when mtDNAs cut with EcoRI and PstI, with which the 4.6-kbp major fragment is divided into 4.0- and 0.6-kbp fragments, were used for the hybridization (data not shown). Furthermore, no hybridization with an excessive amount of marker DNAs in Figure 6 supports that nonspecific hybridization is generally negligible under the conditions used. It is, therefore, suggested that mtDNA samples from some individuals of Okinawajima Island origin contain, as a heteroplasmic state, a minor component that gave rise to the unexpected 4.6-kbp fragment cloned into EcoRI-cut plasmids.

A much longer exposure time was also applied to filters for uncut (Figure 6B-D) and PstI-cut (Figure 6C-D) mtDNA samples. In lanes 10 (Okinawa 2) and 11 (Okinawa 1) of Figure 6B-D, major signals for both closed and open circular bands are accompanied by faint signals with slightly decreased mobilities, as indicated by arrowheads. These faint signals also appeared when a filter hybridized with probe B (Figure 6B-C) was exposed for an extended period (data not shown). In lanes 10 and 11 of Figure 6C-D, a major signal corresponding to the 17.2-kbp linear form of mtDNA is followed by a faint signal with a lower mobility, which again showed up in prolonged exposure of the filter shown in Figure 6C (data not shown). Because the individuals from which faint signals appeared or did not appear are well correlated among Figure 5B-D, Figure C-D, and Figure D, it is reasonable to assume that the heteroplasmic 4.6-kbp EcoRI-cut fragment is derived from the larger mtDNA component. Unfortunately, the content of the heteroplasmic mtDNA relative to the major mtDNA molecules is too low (close to the detection limit by the hybridization) for the larger mtDNA molecules to be fully characterized. In this regard, the heteroplasmy in akamata mtDNAs is substantially different from previous examples of heteroplasmic mtDNA length polymorphism with a much higher content of larger mtDNA molecules (e.g., see MORITZ and BROWN 1987 ; GACH and BROWN 1997 ).

Molecular mechanisms of the concerted evolution:
To account for the mechanism of the concerted evolution of control regions, two possible models are presented in this study (Figure 7). One is a tandem duplication model that operates during replication (Figure 7A). A region surrounded by two control regions is tandemly duplicated via slipped-strand mispairing (MORITZ and BROWN 1986 , MORITZ and BROWN 1987 ) or other mechanisms using two identical control region sequences, giving rise to either a larger mtDNA with three control regions or a normal-sized mtDNA with two homogeneous control region sequences (see legend of Figure 7). In the other model, sequences of two control regions can be homogenized by a mechanism such as frequent gene conversion (Figure 7B). The crossing over of nicked strands between two control regions within a mtDNA molecule leads to formation of a Holliday structure, and a sequence of one control region (and possibly of flanking tRNA genes as well) may be replaced by that of the other via repair of heteroduplex DNA intermediates (Figure 7B). For such a mechanism to occur, DNA recombination activities are necessary in snake mitochondria. Homologous DNA recombination activity, which has been thought to be inactive in animal mitochondria, is now known to exist in mammalian (THYAGARAJAN et al. 1996 ) and nematode (LUNT and HYMAN 1997 ) mitochondria.

View larger version (35K): In this window In a new window Download PPT slide   Figure 7. Models for the concerted evolution or sequence homogenization between two control regions. (A) Tandem duplication model. (B) Gene conversion model. Numerals 1 and 2 stand for control regions I and II, respectively. The model in A is based on the well-elucidated asymmetric replication mechanisms of mammalian mtDNAs (CLAYTON 1992 and references therein) with consideration to peculiar features of the snake mtDNA. Although initiation of heavy-strand replication and subsequent D-loop formation may be possible at both control regions I and II (see text for discussions), only the replication starting from control region II is considered in this model. A newly synthesized heavy strand pausing at control region I with an extended D-loop structure is unwound to reanneal to control region II for de novo initiation of heavy-strand synthesis. Light-strand synthesis in the next round of replication could produce either a larger mtDNA with three control regions or a normal-sized mtDNA with two homogeneous control region sequences. The model in B is based on a well-known mechanism for homologous DNA recombination (e.g., see LEWIN 1994 ). Only parts corresponding to two control regions are shown in B. Note, however, that the two control regions have intramolecular connections within a mtDNA molecule.

These two models were assessed by Southern hybridization experiments (Figure 6). If the second model were to operate, two minicircle mtDNAs with a single control region (12.6 and 4.6 kbp, respectively) might be generated as alternative products of the homologous DNA recombination (Figure 7B). As shown in Figure 6B-D, however, there is no hybridization signal for the 4.6-kbp minicircle in the reasonable position of a gel; a closed circular DNA being 4.6 kbp in size should appear between two marker bands of 4.4 and 2.3 kbp. This indicates that the unexpectedly cloned 4.6-kbp fragment with EcoRI termini did not originate from its circular form. On the other hand, the first model predicts the heteroplasmic existence of larger mtDNA molecules generated by tandem duplication (Figure 7A). As outlined earlier, a scanty proportion of larger mtDNA molecules was actually detected, and the unexpectedly cloned 4.6-kbp fragment was thought to be derived from it (Figure 5, B–D). The 4.6-kbp fragment contained a control region sequence that was surrounded by genes for ND1 and 12S rRNA (Figure 1B), an organization that is predicted to arise by tandem duplication of a region between control regions I and II (see Figure 1A).

These observations are in agreement with the tandem duplication model (Figure 7A), although it is uncertain, and may be difficult to prove because of its extremely low content, that the larger mtDNA molecule detected in Figure 6B-D corresponds with that depicted in Figure 7A or its derivatives. At this point in time, the first model (Figure 7A) is favored by the Southern hybridization experiments, but the second model (Figure 7B) is not excluded. Questions still remain as to the frequency of sequence homogenization between two paralogous control regions and whether the replicational error mechanism that operates for only a minor proportion of mtDNAs in a cell can reasonably account for it (Figure 7A). Biochemical studies for mechanisms of replication and transcription in snake mitochondria, as well as further genetic studies on evolution of snake control region sequences, are expected to provide clues to further unravel the mystery in the concerted evolution.

	ACKNOWLEDGMENTS

We thank H. SATO and S. IWANAGA for helping to access part of the akamata samples. We also thank M. YAMAGUCHI for kind assistance in analyzing a part of the mtDNA sequences. Gratitude is extended to S. YOKOBORI for providing comments to the original version of manuscript. This work was supported by grants from the Ministry of Education, Science, Sports and Culture of Japan to Y.K. (07740583 and 09214102), H.O. (C-09839024), and M.N. (10660189).

Manuscript received December 25, 1997; Accepted for publication June 9, 1998.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

ADACHI, J. and M. HASEGAWA, 1996  Model of amino acid substitution in proteins encoded by mitochondrial DNA. J. Mol. Evol. 42:459-468[Medline].

ADACHI, J., Y. CAO, and M. HASEGAWA, 1993  Tempo and mode of mitochondrial DNA evolution in vertebrates at the amino acid sequence level: rapid evolution in warm-blooded vertebrates. J. Mol. Evol. 36:270-281[Medline].

ANDERSON, S., A. T. BANKIER, B. G. BARRELL, M. H. L. DE BRUIJN, and A. R. COULSON et al., 1981  Sequence and organization of the human mitochondrial genome. Nature 290:457-465[Medline].

ANDERSON, S., M. H. L. DE BRUIJN, A. R. COULSON, I. C. EPERON, and F. SANGER et al., 1982  Complete sequence of bovine mitochondrial DNA. Conserved features of the mammalian mitochondrial genome. J. Mol. Biol. 156:683-717[Medline].

ARNASON, U., A. GULLBERG, and B. WIDEGREN, 1991  The complete nucleotide sequence of the mitochondrial DNA of the fin whale, Balaenoptera physalus. J. Mol. Evol. 33:556-568[Medline].

ASAKAWA, S., Y. KUMAZAWA, T. ARAKI, H. HIMENO, and K. MIURA et al., 1991  Strand-specific nucleotide composition bias in echinoderm and vertebrate mitochondrial genomes. J. Mol. Evol. 32:511-520[Medline].

ASAKAWA, S., H. HIMENO, K. MIURA, and K. WATANABE, 1995  Nucleotide sequence and gene organization of the starfish Asterina pectinifera mitochondrial genome. Genetics 140:1047-1060[Abstract].

BIBB, M. J., R. A. VAN ETTEN, C. T. WRIGHT, M. W. WALBERG, and D. A. CLAYTON, 1981  Sequence and gene organization of mouse mitochondrial DNA. Cell 26:167-180[Medline].

CARROLL, R. L., 1988 Vertebrate paleontology and evolution. W. H. Freeman, New York.

CHRISTIANSON, T. W. and D. A. CLAYTON, 1988  A tridecamer DNA sequence supports human mitochondrial RNA 3'-end formation in vitro. Mol. Cell. Biol. 8:4502-4509[Abstract/Free Full Text].

CLAYTON, D. A., 1992  Transcription and replication of animal mitochondrial DNA. Int. Rev. Cyt. 141:217-232.

DESJARDINS, P. and R. MORAIS, 1990  Sequence and gene organization of the chicken mitochondrial genome: a novel gene order in higher vertebrates. J. Mol. Biol. 212:599-634[Medline].

DODA, J. N., C. T. WRIGHT, and D. A. CLAYTON, 1981  Elongation of displacement-loop strands in human and mouse mitochondrial DNA is arrested near specific template sequences. Proc. Natl. Acad. Sci. USA 78:6116-6120[Abstract/Free Full Text].

DROUIN, J., 1980  Cloning of human mitochondrial DNA in Escherichia coli. J. Mol. Biol. 140:15-34[Medline].

GACH, M. H. and W. M. BROWN, 1997  Characteristics and distribution of large tandem duplications in brook stickleback (Culaea inconstans) mitochondrial DNA. Genetics 145:383-394[Abstract].

GADALETA, G., G. PEPE, G. DE CANDIA, C. QUAGLIARIELLO, and E. SBISÀ et al., 1989  The complete nucleotide sequence of the Rattus norvegicus mitochondrial genome: cryptic signals revealed by comparative analysis between vertebrates. J. Mol. Evol. 28:497-516[Medline].

GELFAND, R. and G. ATTARDI, 1981  Synthesis and turnover of mitochondrial ribonucleic acid in Hela cells: the mature ribosomal and messenger ribonucleic acid species are metabolically unstable. Mol. Cell. Biol. 1:497-511[Abstract/Free Full Text].

HÄRLID, A., A. JANKE, and U. ARNASON, 1997  The mtDNA sequence of the ostrich and the divergence between paleognathous and neognathous birds. Mol. Biol. Evol. 14:754-761[Abstract].

HEDGES, S. B., 1994  Molecular evidence for the origin of birds. Proc. Natl. Acad. Sci. USA 91:2621-2624[Abstract/Free Full Text].

JANKE, A. and U. ARNASON, 1997  The complete mitochondrial genome of Alligator mississippiensis and the separation between recent Archosauria (birds and crocodiles). Mol. Biol. Evol. 14:1266-1272[Abstract].

JANKE, A., G. FELDMAIER-FUCHS, W. K. THOMAS, A. VON HAESELER, and S. PÄÄBO, 1994  The marsupial mitochondrial genome and the evolution of placental mammals. Genetics 137:243-256[Abstract].

JANKE, A., N. J. GEMMELL, G. FELDMAIER-FUCHS, A. VON HAESELER, and S. PÄÄBO, 1996  The mitochondrial genome of a monotreme—the platypus (Ornithorhynchus anatinus). J. Mol. Evol. 42:153-159[Medline].

KUMAZAWA, Y. and M. NISHIDA, 1993  Sequence evolution of mitochondrial tRNA genes and deep-branch animal phylogenetics. J. Mol. Evol. 37:380-398[Medline].

KUMAZAWA, Y. and M. NISHIDA, 1995  Variations in mitochondrial tRNA gene organization of reptiles as phylogenetic markers. Mol. Biol. Evol. 12:759-772[Abstract].

KUMAZAWA, Y., H. OTA, M. NISHIDA, and T. OZAWA, 1996  Gene rearrangements in snake mitochondrial genomes: highly concerted evolution of control-region-like sequences duplicated and inserted into a tRNA gene cluster. Mol. Biol. Evol. 13:1242-1254[Abstract].

LEE, W.-J. and T. D. KOCHER, 1995  Complete sequence of a sea lamprey (Petromyzon marinus) mitochondrial genome: early establishment of the vertebrate genome organization. Genetics 139:873-887[Abstract].

LEE, W.-J., J. CONROY, W. H. HOWELL, and T. D. KOCHER, 1995  Structure and evolution of teleost mitochondrial control regions. J. Mol. Evol. 41:54-66[Medline].

LEWIN, B., 1994 Genes V. Oxford University Press, New York.

LUNT, D. H. and B. C. HYMAN, 1997  Animal mitochondrial DNA recombination. Nature 387:247[Medline].

MACEY, J. R., A. LARSON, N. B. ANANJEVA, Z. FANG, and T. J. PAPENFUSS, 1997  Two novel gene orders and the role of light-strand replication in rearrangement of the vertebrate mitochondrial genome. Mol. Biol. Evol. 14:91-104[Abstract].

MACEY, J. R., J. A. SCHULTE, II, A. LARSON, and T. J. PAPENFUSS, 1998  Tandem duplication via light-strand synthesis may provide a precursor for mitochondrial genomic rearrangement. Mol. Biol. Evol. 15:71-75[Abstract].

MARTIN, A. P. and S. R. PALUMBI, 1993  Body size, metabolic rate, generation time, and the molecular clock. Proc. Natl. Acad. Sci. USA 90:4087-4091[Abstract/Free Full Text].

MITA, S., R. J. MONNAT, JR., and L. A. LOEB, 1988  Direct selection of mutations in the human mitochondrial tRNA^Thr gene: reversion of an `uncloneable' phenotype. Mutat. Res. 199:183-190[Medline].

MORITZ, C. and W. M. BROWN, 1986  Tandem duplications of D-loop and ribosomal RNA sequences in lizard mitochondrial DNA. Science 233:1425-1427[Abstract/Free Full Text].

MORITZ, C. and W. M. BROWN, 1987  Tandem duplications in animal mitochondrial DNAs: variation in incidence and gene content among lizards. Proc. Natl. Acad. Sci. USA 84:7183-7187[Abstract/Free Full Text].

NEI, M., 1991 Relative efficiencies of different tree-making methods for molecular data, pp. 90–128 in Phylogenetic Analysis of DNA Sequences, edited by M. M. MIYAMOTO and J. CRACRAFT. Oxford University Press, New York.

OJALA, D., J. MONTOYA, and G. ATTARDI, 1981  tRNA punctuation model of RNA processing in human mitochondria. Nature 290:470-474[Medline].

OTA, H., 1986  A review of reptiles and amphibians of the Amami Group, Ryukyu Archipelago. Mem. Fac. Sci., Kyoto Univ. 11:57-71.

OTA, T. and M. NEI, 1994  Estimation of the number of amino acid substitutions per site when the substitution rate varies among sites. J. Mol. Evol. 38:642-643.

PÄÄBO, S., D. M. IRWIN, and A. C. WILSON, 1990  DNA damage promotes jumping between templates during enzymatic amplification. J. Biol. Chem. 265:4718-4721[Abstract/Free Full Text].

PÄÄBO, S., W. K. THOMAS, K. M. WHITFIELD, Y. KUMAZAWA, and A. C. WILSON, 1991  Rearrangements of mitochondrial transfer RNA genes in marsupials. J. Mol. Evol. 33:426-430[Medline].

QUINN, T. W. and D. P. MINDELL, 1996  Mitochondrial gene order adjacent to the control region in crocodile, turtle, and tuatara. Mol. Phylogenet. Evol. 5:344-351[Medline].

QUINN, T. W. and A. C. WILSON, 1993  Sequence evolution in and around the mitochondrial control region in birds. J. Mol. Evol. 37:417-425[Medline].

RAGE, J.-C., 1987 Fossil history, pp. 51–76 in Snakes: Ecology and Evolutionary Biology, edited by R. A. SEIGEL, J. T. COLLINS and S. S. NOVAK. Macmillan, New York.

ROE, B. A., D.-P. MA, R. K. WILSON, and J. F.-H. WONG, 1985  The complete nucleotide sequence of the Xenopus laevis mitochondrial genome. J. Biol. Chem. 260:9759-9774[Abstract/Free Full Text].

SEUTIN, G., B. F. LANG, D. P. MINDELL, and R. MORAIS, 1994  Evolution of the WANCY region in amniote mitochondrial DNA. Mol. Biol. Evol. 11:329-340[Abstract].

STANTON, D. J., L. L. DAEHLER, C. C. MORITZ, and W. M. BROWN, 1994  Sequences with the potential to form stem-and-loop structures are associated with coding-region duplications in animal mitochondrial DNA. Genetics 137:233-241[Abstract].

STRIMMER, K. and A. VON HAESELER, 1996  Quartet puzzling: a quartet maximum-likelihood method for reconstructing tree topologies. Mol. Biol. Evol. 13:964-969.

TAPPER, D. P., R. A. VAN ETTEN, and D. A. CLAYTON, 1983  Isolation of mammalian mitochondrial DNA and RNA and cloning of the mitochondrial genome. Methods Enzymol. 97:426-434[Medline].

THOMAS, W. K. and A. T. BECKENBACH, 1989  Variation in salmonid mitochondrial DNA: evolutionary constraints and mechanisms of substitution. J. Mol. Evol. 29:233-245[Medline].

THYAGARAJAN, B., R. A. PADUA, and C. CAMPBELL, 1996  Mammalian mitochondria possess homologous DNA recombination activity. J. Biol. Chem. 271:27536-27543[Abstract/Free Full Text].

TZENG, C.-S., C.-F. HUI, S.-C. SHEN, and P. C. HUANG, 1992  The complete nucleotide sequence of the Crossostoma lacustre mitochondrial genome: conservation and variations among vertebrates. Nucleic Acids Res. 20:4853-4858[Abstract/Free Full Text].

WALBERG, M. W. and D. A. CLAYTON, 1981  Sequence and properties of the human KB cell and mouse L cell D-loop regions of mitochondrial DNA. Nucleic Acids Res. 9:5411-5421[Abstract/Free Full Text].

WALLIS, G. P., 1987  Mitochondrial DNA insertion polymorphism and germ line heteroplasmy in the Triturus cristatus complex. Heredity 58:229-238.

WOLSTENHOLME, D. R., 1992  Animal mitochondrial DNA: structure and evolution. Int. Rev. Cyt. 141:173-216.

WONG, T. W. and D. A. CLAYTON, 1985  In vitro replication of human mitochondrial DNA: accurate initiation at the origin of light-strand synthesis. Cell 42:951-958[Medline].

YOKOBORI, S. and S. PÄÄBO, 1995a  Transfer RNA editing in land snail mitochondria. Proc. Natl. Acad. Sci. USA 92:10432-10435[Abstract/Free Full Text].

YOKOBORI, S. and S. PÄÄBO, 1995b  tRNA editing in metazoans. Nature 377:490[Medline].

YONEYAMA, Y., 1987  The nucleotide sequences of the heavy and light strand replication origins of the Rana catesbeiana mitochondrial genome. J. Nippon Med. Sch. 54:429-440. (in Japanese)..

ZARDOYA, R. and A. MEYER, 1996  The complete nucleotide sequence of the mitochondrial genome of the lungfish (Protopterus dolloi) supports its phylogenetic position as a close relative of land vertebrates. Genetics 142:1249-1263[Abstract].

ZARDOYA, R. and A. MEYER, 1997  The complete DNA sequence of the mitochondrial genome of a "living fossil," the coelacanth (Latimeria chalumnae). Genetics 146:995-1010[Abstract].

ZARDOYA, R., A. GARRIDO-PERTIERRA, and J. M. BAUTISTA, 1995  The complete nucleotide sequence of the mitochondrial DNA genome of the rainbow trout, Oncorhynchus mykiss. J. Mol. Evol. 41:942-951[Medline].

ZEVERING, C. E., C. MORITZ, A. HEIDEMAN, and R. A. STURM, 1991  Parallel origins of duplications and the formation of pseudogenes in mitochondrial DNA from parthenogenetic lizards (Heteronotia binoei; Gekkonidae). J. Mol. Evol. 33:431-441[Medline].

This article has been cited by other articles:

				K. Suga, D. B. Mark Welch, Y. Tanaka, Y. Sakakura, and A. Hagiwara Two Circular Chromosomes of Unequal Copy Number Make Up the Mitochondrial Genome of the Rotifer Brachionus plicatilis Mol. Biol. Evol., June 1, 2008; 25(6): 1129 - 1137. [Abstract] [Full Text] [PDF]

				A. Kurabayashi, M. Sumida, H. Yonekawa, F. Glaw, M. Vences, and M. Hasegawa Phylogeny, Recombination, and Mechanisms of Stepwise Mitochondrial Genome Reorganization in Mantellid Frogs from Madagascar Mol. Biol. Evol., May 1, 2008; 25(5): 874 - 891. [Abstract] [Full Text] [PDF]

				A. Tatarenkov and J. C Avise Rapid concerted evolution in animal mitochondrial DNA Proc R Soc B, July 22, 2007; 274(1619): 1795 - 1798. [Abstract] [Full Text] [PDF]

				D. San Mauro, D. J. Gower, R. Zardoya, and M. Wilkinson A Hotspot of Gene Order Rearrangement by Tandem Duplication and Random Loss in the Vertebrate Mitochondrial Genome Mol. Biol. Evol., January 1, 2006; 23(1): 227 - 234. [Abstract] [Full Text] [PDF]

				R. L. Mueller and J. L. Boore Molecular Mechanisms of Extensive Mitochondrial Gene Rearrangement in Plethodontid Salamanders Mol. Biol. Evol., October 1, 2005; 22(10): 2104 - 2112. [Abstract] [Full Text] [PDF]

				R. Shao, S. C. Barker, H. Mitani, Y. Aoki, and M. Fukunaga Evolution of Duplicate Control Regions in the Mitochondrial Genomes of Metazoa: A Case Study with Australasian Ixodes Ticks Mol. Biol. Evol., March 1, 2005; 22(3): 620 - 629. [Abstract] [Full Text] [PDF]

				S.-i. Yokobori, N. Fukuda, M. Nakamura, T. Aoyama, and T. Oshima Long-Term Conservation of Six Duplicated Structural Genes in Cephalopod Mitochondrial Genomes Mol. Biol. Evol., November 1, 2004; 21(11): 2034 - 2046. [Abstract] [Full Text] [PDF]

				L. Cao, E. Kenchington, E. Zouros, and G. C. Rodakis Evidence That the Large Noncoding Sequence Is the Main Control Region of Maternally and Paternally Transmitted Mitochondrial Genomes of the Marine Mussel (Mytilus spp.) Genetics, June 1, 2004; 167(2): 835 - 850. [Abstract] [Full Text] [PDF]

				J. G. Inoue, M. Miya, K. Tsukamoto, and M. Nishida Evolution of the Deep-Sea Gulper Eel Mitochondrial Genomes: Large-Scale Gene Rearrangements Originated Within the Eels Mol. Biol. Evol., November 1, 2003; 20(11): 1917 - 1924. [Abstract] [Full Text] [PDF]

				J. M. Serb and C. Lydeard Complete mtDNA Sequence of the North American Freshwater Mussel, Lampsilis ornata (Unionidae): An Examination of the Evolution and Phylogenetic Utility of Mitochondrial Genome Organization in Bivalvia (Mollusca) Mol. Biol. Evol., November 1, 2003; 20(11): 1854 - 1866. [Abstract] [Full Text] [PDF]

				R. Shao, M. Dowton, A. Murrell, and S. C. Barker Rates of Gene Rearrangement and Nucleotide Substitution Are Correlated in the Mitochondrial Genomes of Insects Mol. Biol. Evol., October 1, 2003; 20(10): 1612 - 1619. [Abstract] [Full Text]

				R. Shao and S. C. Barker The Highly Rearranged Mitochondrial Genome of the Plague Thrips, Thrips imaginis (Insecta: Thysanoptera): Convergence of Two Novel Gene Boundaries and an Extraordinary Arrangement of rRNA Genes Mol. Biol. Evol., March 1, 2003; 20(3): 362 - 370. [Abstract] [Full Text] [PDF]

				X.-M. Lu, Y.-X. Fu, and Y.-P. Zhang Evolution of Mitochondrial Cytochrome b Pseudogene in Genus Nycticebus Mol. Biol. Evol., December 1, 2002; 19(12): 2337 - 2341. [Full Text] [PDF]

				D. V. Lavrov, J. L. Boore, and W. M. Brown Complete mtDNA Sequences of Two Millipedes Suggest a New Model for Mitochondrial Gene Rearrangements: Duplication and Nonrandom Loss Mol. Biol. Evol., February 1, 2002; 19(2): 163 - 169. [Abstract] [Full Text] [PDF]

				J. R. Eberhard, T. F. Wright, and E. Bermingham Duplication and Concerted Evolution of the Mitochondrial Control Region in the Parrot Genus Amazona Mol. Biol. Evol., July 1, 2001; 18(7): 1330 - 1342. [Abstract] [Full Text] [PDF]

				R. Zardoya and A. Meyer On the origin of and phylogenetic relationships among living amphibians PNAS, May 30, 2001; (2001) 111455498. [Abstract] [Full Text] [PDF]

				R. Zardoya and A. Meyer On the origin of and phylogenetic relationships among living amphibians PNAS, June 19, 2001; 98(13): 7380 - 7383. [Abstract] [Full Text] [PDF]

				R. E. Broughton, J. E. Milam, and B. A. Roe The Complete Sequence of the Zebrafish (Danio rerio) Mitochondrial Genome and Evolutionary Patterns in Vertebrate Mitochondrial DNA Genome Res., November 1, 2001; 11(11): 1958 - 1967. [Abstract] [Full Text] [PDF]