Mitochondrial Gene Arrangement of the Horseshoe Crab Limulus polyphemus L.: Conservation of Major Features Among Arthropod Classes Joseph L. Staton,’ L. Lynne Daehler, and Wesley M. Brown Department of Biology, University of Michigan Numerous complete mitochondrial DNA sequences have been determined for species within two arthropod groups, insects and crustaceans, but there are none for a third, the chelicerates. Most mitochondrial gene arrangements reported for crustacean and insect species are identical or nearly identical to that of Drosophila yakuba. Sequences across 36 of the gene boundaries in the mitochondrial DNA (mtDNA) of a representative chelicerate, Limulus polyphemus L., also reveal an arrangement like that of Drosophila yakuba. Only the position of the tRNALeu(UUR) gene differs; in Limulus it is between the genes for tRNA Leu(CUN) and NDl. This positioning is also found in onychophorans, mollusks, and annelids, but not in insects and crustaceans, and indicates that tRNALeU(CUN)-t- RNALeU(UUR)-ND1 was the ancestral gene arrangement for these groups, as suggested earlier. There are no differences in the relative arrangements of protein-coding and ribosomal RNA genes between Limulus and Drosophila, and none have been observed within arthropods. The high degree of similarity of mitochondrial gene arrangements within arthropods is striking, since some taxa last shared a common ancestor before the Cambrian, and contrasts with the extensive mtDNA rearrangements occasionally observed within some other metazoan phyla (e.g., mollusks and nematodes). Introduction The mitochondrial genome of metazoans is typi- cally a circular DNA molecule of ca. 16 kb which is highly variable in sequence (Brown, George, and Wilson 1979). A much lower amount of variation is found in the relative arrangement of the 37 genes usually encoded in it, of which there are one each for the small and large ribosomal subunit RNAs (s-i-RNA and l-i-RNA), 13 for protein subunits (cytochrome oxidase I-III [COl-31, ATP synthase 6 and 8 [A6 and A8], NADH dehydro- genase l-6 and 4L [NDl-6, ND4L], and cytochrome b apoenzyme [Cytb]), and 22 for tRNAs. One or more regions of variable length that appear not to code for structural genes are also typical. These and other fea- tures of metazoan mtDNA have been summarized else- where (Brown 1985; Wolstenholme 1992; Boore and Brown 1995). The arrangement of the genes in metazoan mito- chondrial DNA (mtDNA) has been hypothesized to con- tain phylogenetic information of use in determining the order of ancient divergences based on the low frequency of gene rearrangements observed (Brown 1985; Moritz, Dowling, and Brown 1987; Jacobs et al. 1988; Smith et al. 1993; Boore and Brown 1994b, 1995; Macey et al. 1997). For example, one arrangement predominates among vertebrates, although variant arrangements found in some taxa can be derived from it by one or two re- arrangement events (see Macey et al. [1997] for an ex- cellent summary). In this case, it is clear that the pre- dominant arrangement is also ancestral, since it is shared among several vertebrate classes and since a species rep- resentative of the’ sister taxon to vertebrates, the ceph- l Present address: Department of Biology, University of Califor- nia, Los Angeles. Key words: mtDNA, gene arrangement, Limulus, chelicerate, ar- thropod. Address for correspondence and reprints: Joseph L. Staton, De- partment of Biology, University of California, 405 Hilgard Avenue, Los Angeles, California 900951606. E-mail: [email protected]. Mol. Biol. Evol. 14(8):867-874. 1997 0 1997 by the Society for Molecular Biology and Evolution. ISSN: 0737-4038 alochordate Brunchiostoma Joridue (unpublished data), has an arrangement identical to it for all the genes whose positions have been found to vary among vertebrates. An ancestral gene arrangement is also discernible for the four extant echinoderm classes so far examined, al- beit with somewhat more variation than has been ob- served within vertebrates or arthropods: echinoids and holothuroids differ from asteroids and ophiuroids by an inversion of one large genomic segment, and several tRNA positions differ among the four lineages (Smith et al. 1989, 1993). However, in mollusks (Hoffmann, Boore, and Brown 1992; Lecanidou, Doris, and Rodakis 1994; Ter- rett, Miles, and Thomas 1994; Hatzoglou, Rodakis, and Lecanidou 1995; Boore and Brown 1994a, 1994b; un- published data) and nematodes (Okimoto et al. 1991, 1992), the pattern of a single phylum-characteristic gene arrangement is broken, and radically different intragroup arrangements exist. All insect and crustacean mitochondrial gene ar- rangements studied conform to one basic pattern, typi- fied by Drosophila mtDNA (Clary and Wolstenholme 1985), within which some variation in tRNA gene po- sition is seen (Crozier and Crozier 1993; Mitchell, Cockburn, and Seawright 1993; Valverde et al. 1994; Flook, Rowell, and Gellissen 1995). However, no com- plete mitochondrial gene arrangement is known from a representative of the chelicerates, a third major arthro- pod group. We fill that gap by reporting here the gene arrangement of the horseshoe crab Limulus polyphemus L. (Merostomata: Xiphosurida), which we determined by sequencing, in the aggregate, 8.5 kb of mtDNA from those regions that identify and link its 37 mitochondrial genes and A+T-rich region. Materials and Methods Purified Limulus mtDNA of type 1 (Saunders, Kes- sler, and Avise 1986) was a gift from John Avise (Uni- versity of Georgia, Athens). Limulus mtDNA was di- gested with Xba I, and the two resulting fragments were 867 Downloaded from https://academic.oup.com/mbe/article/14/8/867/982323 by guest on 21 November 2021
Transcript
Mitochondrial Gene Arrangement of the Horseshoe Crab Limulus polyphemus L.: Conservation of Major Features Among Arthropod Classes
Joseph L. Staton,’ L. Lynne Daehler, and Wesley M. Brown Department of Biology, University of Michigan
Numerous complete mitochondrial DNA sequences have been determined for species within two arthropod groups, insects and crustaceans, but there are none for a third, the chelicerates. Most mitochondrial gene arrangements reported for crustacean and insect species are identical or nearly identical to that of Drosophila yakuba. Sequences across 36 of the gene boundaries in the mitochondrial DNA (mtDNA) of a representative chelicerate, Limulus polyphemus L., also reveal an arrangement like that of Drosophila yakuba. Only the position of the tRNALeu(UUR) gene differs; in Limulus it is between the genes for tRNA Leu(CUN) and NDl. This positioning is also found in onychophorans, mollusks, and annelids, but not in insects and crustaceans, and indicates that tRNALeU(CUN)-t- RNALeU(UUR)-ND1 was the ancestral gene arrangement for these groups, as suggested earlier. There are no differences in the relative arrangements of protein-coding and ribosomal RNA genes between Limulus and Drosophila, and none have been observed within arthropods. The high degree of similarity of mitochondrial gene arrangements within arthropods is striking, since some taxa last shared a common ancestor before the Cambrian, and contrasts with the extensive mtDNA rearrangements occasionally observed within some other metazoan phyla (e.g., mollusks and nematodes).
Introduction
The mitochondrial genome of metazoans is typi- cally a circular DNA molecule of ca. 16 kb which is highly variable in sequence (Brown, George, and Wilson 1979). A much lower amount of variation is found in the relative arrangement of the 37 genes usually encoded in it, of which there are one each for the small and large ribosomal subunit RNAs (s-i-RNA and l-i-RNA), 13 for protein subunits (cytochrome oxidase I-III [COl-31, ATP synthase 6 and 8 [A6 and A8], NADH dehydro- genase l-6 and 4L [NDl-6, ND4L], and cytochrome b apoenzyme [Cytb]), and 22 for tRNAs. One or more regions of variable length that appear not to code for structural genes are also typical. These and other fea- tures of metazoan mtDNA have been summarized else- where (Brown 1985; Wolstenholme 1992; Boore and Brown 1995).
The arrangement of the genes in metazoan mito- chondrial DNA (mtDNA) has been hypothesized to con- tain phylogenetic information of use in determining the order of ancient divergences based on the low frequency of gene rearrangements observed (Brown 1985; Moritz, Dowling, and Brown 1987; Jacobs et al. 1988; Smith et al. 1993; Boore and Brown 1994b, 1995; Macey et al. 1997). For example, one arrangement predominates among vertebrates, although variant arrangements found in some taxa can be derived from it by one or two re- arrangement events (see Macey et al. [1997] for an ex- cellent summary). In this case, it is clear that the pre- dominant arrangement is also ancestral, since it is shared among several vertebrate classes and since a species rep- resentative of the’ sister taxon to vertebrates, the ceph-
l Present address: Department of Biology, University of Califor- nia, Los Angeles.
Address for correspondence and reprints: Joseph L. Staton, De- partment of Biology, University of California, 405 Hilgard Avenue, Los Angeles, California 900951606. E-mail: [email protected].
Mol. Biol. Evol. 14(8):867-874. 1997 0 1997 by the Society for Molecular Biology and Evolution. ISSN: 0737-4038
alochordate Brunchiostoma Joridue (unpublished data), has an arrangement identical to it for all the genes whose positions have been found to vary among vertebrates. An ancestral gene arrangement is also discernible for the four extant echinoderm classes so far examined, al- beit with somewhat more variation than has been ob- served within vertebrates or arthropods: echinoids and holothuroids differ from asteroids and ophiuroids by an inversion of one large genomic segment, and several tRNA positions differ among the four lineages (Smith et al. 1989, 1993).
However, in mollusks (Hoffmann, Boore, and Brown 1992; Lecanidou, Doris, and Rodakis 1994; Ter- rett, Miles, and Thomas 1994; Hatzoglou, Rodakis, and Lecanidou 1995; Boore and Brown 1994a, 1994b; un- published data) and nematodes (Okimoto et al. 1991, 1992), the pattern of a single phylum-characteristic gene arrangement is broken, and radically different intragroup arrangements exist.
All insect and crustacean mitochondrial gene ar- rangements studied conform to one basic pattern, typi- fied by Drosophila mtDNA (Clary and Wolstenholme 1985), within which some variation in tRNA gene po- sition is seen (Crozier and Crozier 1993; Mitchell, Cockburn, and Seawright 1993; Valverde et al. 1994; Flook, Rowell, and Gellissen 1995). However, no com- plete mitochondrial gene arrangement is known from a representative of the chelicerates, a third major arthro- pod group. We fill that gap by reporting here the gene arrangement of the horseshoe crab Limulus polyphemus L. (Merostomata: Xiphosurida), which we determined by sequencing, in the aggregate, 8.5 kb of mtDNA from those regions that identify and link its 37 mitochondrial genes and A+T-rich region.
Materials and Methods
Purified Limulus mtDNA of type 1 (Saunders, Kes- sler, and Avise 1986) was a gift from John Avise (Uni- versity of Georgia, Athens). Limulus mtDNA was di- gested with Xba I, and the two resulting fragments were
867
Dow
nloaded from https://academ
ic.oup.com/m
be/article/14/8/867/982323 by guest on 21 Novem
ber 2021
868 Staton et al.
cloned into Bluescript@@ vectors (Stratagene); these clones were used for subsequent subcloning and se- quencing.
Dideoxy-sequencing employed T3 and T7 primers, modified T7 polymerase (Sequenasem), and standard protocols. Products were labeled with [35S]dATPc& (1,000 Ci/mmol; New England Nuclear) and separated on 60-cm 6% acrylamide gels. Custom primers were employed to obtain sequence information from internal regions. Some sequences were confirmed by cycle se- quencing, using ABI PRISM@’ dye terminator chemistry and analysis on ABI 373 or 377 automated sequencers (ABI/Perkin-Elmer).
Mitochondrial protein and rRNA gene sequences were identified and aligned with the corresponding Dro- sophila yakuba sequences (Clary and Wolstenholme 1985). Eugene software on a Sun Microsystems work- station, BLAST searches (Altschul et al. 1990), and Clustal W version 1.4 (Thompson, Higgins, and Gibson 1994) were employed. As a class, tRNA genes were identified as sequences between protein and/or r-RNA genes with the potential to form tRNA-like secondary str..tctures; specific tRNA gene identities were assigned according to their anticodons, which are highly con- served among metazoans. Contiguous sequences from one or both strands were obtained through 36 of the 38 expected gene and A+T-region boundaries. These se- quences are deposited in GenBank under accession numbers AFOO2644-AFO02653 (except for U29708, previously reported in Boore et al. 1995).
Results and Discussion
The size of Limulus mtDNA, estimated by sum- mation of restriction fragments (Saunders, Kessler, and Avise 1986), is ca. 16 kb. We have determined 8.5 kb of nonoverlapping sequence, an amount sufficient to ac- count for the positions and orientations of 13 protein- coding genes, 2 x-RNA genes, 22 tRNA genes, and a noncoding region that is positionally homologous to the A+T-rich region in Drosophila yakuba mtDNA. The se- quence orientations are shown in greatly abbreviated form in figure 1. The positions and orientations of all protein and rRNA genes and of the A+T-rich region are identical to those of D. yakuba (Clary and Wolsten- holme 1985) and Artemia franciscana (Valverde et al. 1994) (fig. 2). Those and many other studies of insect and crustacean mtDNAs provide a picture of overall mtDNA stability as regards gene arrangement, but with some variation in tRNA gene position. This report ex- tends that conclusion to a third arthropod class, the Chelicerata.
Base Composition and Gene Content
For the 8,459 bases of Limulus mtDNA sequenced, the nucleotide (nt) composition of the strand encoding most (9/13) of the protein genes is A = 3,095, T = 2,638, G = 855, C = 1,871. The G+C-content (32.2%) differs significantly from that of the corresponding se- quence in Drosophila yakuba (21.4%), despite the sim- ilarity of these two mtDNAs in other respects. Limulus
mtDNA contains 13 open reading frames (ORFs) that correspond unambiguously to the 13 protein-subunit genes typical of metazoan mtDNAs: COl-3, A6 and A8, NDl-6, ND4L, and Cytb. Both l-x-RNA and s-rRNA genes are present, along with the full complement of 22 tRNA genes typical of metazoan mtDNAs (fig. 1).
Unassigned DNA
A large region (ca. 1.3 kb), corresponding in po- sition to the A+T-rich region of Drosophila mtDNA, lies between the s-rRNA and tRNAne genes (figs. 1 and 2). The 161-nt portion we sequenced from this region is rich in A+T (39.1% A and 37.3% T). However, the A+T content of this sequence is only slightly higher than those of nearby regions, and is less than the cor- responding portions of the A+T-rich region of Dro- sophila (A+T > 90%; Clary and Wolstenholme 1985). Thus, because of the small number of nt sampled, the slightly increased A+T content over that of other regions may not be significant, although an elevated A+T content is a general feature of noncoding regions of metazoan mtDNAs.
Gene Arrangement
The mitochondrial gene arrangement of Limulus, nearly identical to those of Drosophila and Artemiu, dif- fers from each by an inverted translocation of the t- RNALeu(uuR) e g ne which, in Drosophila and Artemia, is located between and transcribed from the same strand as the CO1 and CO2 genes (Clary and Wolstenholme 1985) but, in Limulus, is between the ND1 and t- RNAhu(cuN) genes and is transcribed from the opposite strand of that coding for CO1 and C02. This arrange- ment, with its attendant phylogenetic implications, was discussed in an earlier study (Boore et al. 1995). An additional difference between Limulus and Artemia, which corresponds to a translocation of the adjacent gene pair tRNAne-tRNAGin and an inversion of tRNA1te, is inferred to result from events specific to the Artemia lineage, since the position and relative orientations of this gene pair are identical not only in Limulus and Dro- sophila, but also in three other crustaceans (Duph- nia-Van Raay and Crease 1994; Homarus and Daph- nia-Boore et al. 1995; Penaeus-Garcia-Machado et al. 1996). These differences notwithstanding, the con- servation of the mitochondrial gene arrangement is note- worthy and remains a general feature of arthropod mtDNAs. One or more representatives from several ar- thropod classes have now been surveyed and, although a few differences in tRNA gene positions have been noted, thus far no variation in the relative arrangement or polarity of protein or ribosomal RNA genes has been observed.
Gene pairs whose sequences overlap are common in metazoan mtDNAs (see Wolstenholme 1992), and Li- mulus mtDNA has at least seven of these. In five (t- RNA-Q’-CO1 , tRNAG’“-tRNAPhe, tRNA”“-tRNAG’“, t- RNAoln_tRNAMet and tRNATv-tRNACys; fig. l), the overlapping genes are encoded on opposite mtDNA strands and, thus, the transcripts do not overlap. How- ever, in two (A8-A6 and tRNAArg-tRNAAsn), the genes
Dow
nloaded from https://academ
ic.oup.com/m
be/article/14/8/867/982323 by guest on 21 Novem
ber 2021
Mitochondrial Gene Arrangement of the Horseshoe Crab 869
I L T L P Met F T L L Tm A/60/TCTAATCCTAACTCTTCCC{lOOO?TTTACCTTATTAT~GAT/5a/TCTT~GCC/54/~GCCTTGATA/5l/GCCACT
h cvs Tvr
FIG. 1 .-Schematic representation, emphasizing the sequences across gene boundaries of Limulus mtDNA. The circular mtDNA is depicted as linear, with the 5’ end of the CO1 gene at top left and the immediately adjacent portion of the gene for tRNATyr at bottom right. Each gene is labeled (see text for abbreviations) and its extent is depicted by a horizontal bar; in several cases, genes are contiguous with sequence shown on the preceding or succeeding line. Genes shown above the DNA sequence are transcribed from left to right; those below are transcribed from right to left. The inferred amino acids at the amino and carboxy termini of protein genes are given, using their single-letter abbreviations. The number of nucleotides intervening between the sequence portions shown is displayed between slashes (I/) when known, and between brackets ({ }) when estimated from combined sequence and cleavage mapping data. Nucleotides participating in termination codons are underscored with carets (A). Question marks within brackets [??] between the adjacent A6K03, ND4/ND4L, and srRNA/NCR indicate that contiguous sequence across those junctions was not obtained.
are encoded on the same strand, and two functional tran- scripts cannot be resolved by cleavage. AS-A6 overlaps are common, and there is evidence that the two proteins are translated from a single bicistronic message by ini- tiation at a 5’ terminal start site for A8 and at an internal start site for A6 (reviewed in Wolstenholme 1992). The case of tRNAArs-tRNAAsn is more difficult; these genes overlap by 2 nt, and processing of a transcript having the same sequence as the genes will result in a 2-nt truncation of one or a 1-nt truncation of both tRNAs. Two-nucleotide overlaps between tRNA gene pairs have also been seen in mollusk and annelid mtDNAs, and alternate processing, posttranscriptional editing, and re- laxed constraints on tRNA structure have been invoked as possible mechanisms for resolving the two products (Boore and Brown 1994a, 1995).
Translation Initiation and Termination Signals
Sequences across 22 of the 26 termini of the 13 Limulus protein genes were obtained (fig. 1). Termini were inferred by their adjacency to flanking genes and by the similarity of their amino acid sequences to those reported for termini of the corresponding genes in other metazoans. All termini are shown in figure 1 and are summarized in table 1.
Members of the four-codon family ATN initiate translation in 10 of the 11 genes whose 5’ ends could be identified with confidence. The start codon for the eleventh, COl, could not be identified. Neither an in-frame ATN codon nor an alternate initiation codon (GTG, TTG, GTT and ATAA; Boore and Brown 1994b; Wolstenholme 1992) occurs within 40 nt of the expected CO1 initiation site in Limulus. We have designated an
Dow
nloaded from https://academ
ic.oup.com/m
be/article/14/8/867/982323 by guest on 21 Novem
ber 2021
870 Staton et al.
Sac/ Xba EC@/ EcoRl Xba Sad
K G ANE H T SOJCN) Lo Y IM WY I I
co2 2 A6 co3 g k
Limulus co1 ND5 - ND4 2 g CYTB ND1 16s 12s $ ND2 -- II
FIG. 2.-Comparison of gene arrangements in the mtDNAs of Limulus polyphemus, Drosophila yakuba, and Artemia salina. The circular mtDNAs are depicted as linear, with the 5’ ends of CO1 on the left. Unless underlined, genes are transcribed from left to right. Lines connect genes that differ in position among the species; arrows circling those lines indicate sequence inversions. The Drosophila and Artemia arrange- ments are from Clary and Wolstenholme (1985) and Valverde et al. (1994). respectively.
in-frame TTA codon as the initiation site in figure 1, since a similar codon, TTG, initiates several genes in nematode mtDNAs (Okimoto et al. 1992). This is purely conjectural, however, and identification of the actual site must await more definitive data (e.g., the mRNA or pro- tein sequence).
In Limulus mtDNA, 5 of the 11 genes for which 3’ sequences were determined end with a complete ter- mination codon (TAA or TAG); the remaining 6 end in T or TA, which is presumably converted to UAA by polyadenylation after cleavage of the presumptive po- lycistronic transcript (Ojala, Montoya, and Attardi 1981, reviewed in Wolstenholme 1992).
Transfer RNA Genes
Twenty-two Limulus mtDNA regions could be folded into tRNA-like structures with acceptable anti- codons (fig. 3). Of these, 18 have a T at position 8, which corresponds to the first turn at the base of the acceptor arm; the remaining 4 have an A in this position, an alternative that is often seen in metazoan mitochon-
drial tRNAs. All anticodons are immediately preceded by (C/T)-T and followed by (G/A). The most mis- matched stem among the 22 tRNAs is in the T\CIC arm of tRNAG’“.
The positions and polarities of 21 of the 22 tRNA genes are identical to those in Drosophila yakuba mtDNA (fig. 2), the single exception being the t- mALeu(UUR) gene. The 22 tRNA genes correspond to the standard set found in metazoan mtDNAs. No other tRNA-like sequences were identified in Limulus.
The 22 anticodons employed are sufficient for translating the genetic code found in most invertebrate mitochondrial systems. Excluding that of tRNAMet, 12 of the remaining 21 anticodons have T in the “wobble” (i.e., first anticodon) position, 8 have G, and one (t- RNALys) has C. Among the 9 tRNAs with anticodons that recognize four-fold degenerate codons, 8 have T in the wobble position, and the ninth, tRNASercAGN), has G (fig. 2, table 2). Of the 12 anticodons that recognize two- fold degenerate codons, 7 have G in the first position, and 1 (tRNALYs) has C. Comparisons of anticodons with
Dow
nloaded from https://academ
ic.oup.com/m
be/article/14/8/867/982323 by guest on 21 Novem
ber 2021
Mitochondrial Gene Arrangement of the Horseshoe Crab 871
Table 1 Amino Acid Sequence Comparisons of Portions of 13 Proteins Encoded by Limulus polyhemus mtDNA with Those of Drosophila yakuba
No. of Residues
Identity in start stop Gene (%) AA N/C D. yakuba Codon Codon
NOTE.-AA is the total number of amino acid residues compared, and N/C lists the number compared at amino (N) and carboxy (C) termini. Sequences were aligned using Clustal W version 1.4 (Thompson, Higgins, and Gibson 1994) and, if necessary, edited manually. The Drosophila yakuba sequence and associated data are from Clary and Wolstenholme (1985).
a Codon not sequenced. b Complete gene sequence obtained. c Incomplete termination codon.
codon usage (table 2) indicate that the 7 anticodons with G in the first position bind predominantly to codons end- ing in T rather than C, even though C would pair more stably. The situation for tRNALys is similar; the codon AAA is used over five times as frequently as the more stably pairing AAG. Although seemingly paradoxical, this situation is similar to those in Drosophila (Clary and Wolstenholme 1985) and Artemia (Valverde et al. 1994), except that in Artemia, AAA and AAG are used with equal frequency. The usage could simply result from the strong A +T bias found in these mtDNAs. Many similar cases of C/A (anticodon/codon) mispair- ing at the wobble position are known, but the mecha- nism for tolerating this mispairing remains unexplained (see Wolstenholme [ 19921 for a summary and discus- sion) .
All Limulus anticodons are identical. with their counterparts in Drosophila yakuba. Although most metazoan mitochondrial systems utilize tRNAMetcAUG) and tRNALYS(UUU) variation in these anticodon sequenc- es has been obierved. For example, the variant t- RNAMet@tJA) h as been found among mollusks (Hoff- mann, Boore, and Brown 1992, Boore and Brown 1994a, 1994b), and the variant tRNALyS(UUG) has been found in isolated taxa among arthropods (Limulus, Dro- sophila [Clary and Wolstenholme 19851 and Artemia [Valverde et al. 19941) and echinoderms (Cantatore et al. 1989). The mosaic pattern of this variation in the relatively small number of species whose mt-tRNA genes have been characterized suggests that the wobble position in at least some metazoan mt-tRNAs is under less stringent selection than it is in cytoplasmic tRNAs, possibly as a result of compensating base modifications in the tRNAs, a “two out of three” recognition pattern
for some codons/anticodons, or relaxed constraints on codon/anticodon recognition in metazoan mitochondrial translation systems.
Protein and Ribosomal RNA Genes
With the exception of that for A8, the amino acid sequences translated from the ORFs found in Limulus mtDNA were readily alignable to those of protein-cod- ing genes in Drosophila yakuba mtDNA, and conserved sequence elements and secondary structures allowed positive identification of portions of the s- and l-r-RNA genes. Although the incomplete sequences of those genes cannot provide some parameters (e.g., gene sizes), after alignment with sequences from other taxa they can be compared and used to make provisional estimates of others, such as percent amino acid identity (table 1) and relative codon usage (table 2).
The A8 gene could not be unambiguously identi- fied by sequence comparisons alone, and two additional criteria, gene size and hydrophilicity profile, were used. The latter, calculated by the method of Kyte and Doo- little (1982), had a broad, bimodally negative peak over the 5’ half of the gene and a positive peak at the 3’ end, a pattern similar to those of A8 genes in other metazoan taxa.
Provisional alignments of the sequenced portions of the Limulus s-t-RNA and l-x-RNA genes with the cor- responding Drosophila sequences showed a high level of identity (73.2% for 541 nt and 68.5% for 646 nt, respectively; data not shown).
Phylogenetic Implications of Gene Arrangement
The gene arrangement (fig. 1) in Limulus polyphe- mus mtDNA is similar to those in other arthropods (fig. 2), thus extending the observation that mitochondrial gene arrangement is conserved to chelicerates. Although some variation in tRNA gene position occurs, the rela- tive arrangement of protein and rRNA genes is identical in all arthropods so far investigated. This result parallels a similarly high degree of conservation found among chordates (Macey et al. 1997), but contrasts with the differences observed among (and in one case within) some molluscan classes (Hoffmann, Boore, and Brown 1992; Boore and Brown 1994a, 1994b; Lecanidou, Dor- is, and Rodakis 1994; Terrett, Miles, and Thomas 1994; Hatzoglou, Rodakis, and Lecanidou 1995; unpublished data) and within nematodes (Okimoto et al. 1991, 1992).
The single difference between Limulus mtDNA or- ganization and that of Drosophila is in the position of tRNALeU(UUR) (fig. 2). That discovery and its far-ranging phylogenetic implications were reported in detail by Boore et al. (1995), who concluded that the relative po- sition of the two tRNALeu genes in Limulus, 5’ to and encoded on the same strand as the ND1 gene, is con- sistent with the hypothesized ancestral state for verte- brates, mollusks, annelids and other arthropods. As also pointed out in that study, the failure of Limulus to share the same derived position for the mt-tRNALeUcUUR) gene as insect and crustacean species supports the hypothesis that insects and crustaceans form a derived, monophy-
Dow
nloaded from https://academ
ic.oup.com/m
be/article/14/8/867/982323 by guest on 21 Novem
ber 2021
872 Staton et al.
Alanine A
A-T G-C G-C
Arginine C
A-T A-T A-T A-T T-A A-T G-C
A A fWIAA A T
Asparagine S:i A-T A-T T-A A-T G-C
A Aspartate A-T
A-T Cysteine .T
E-Z A-T C-G A-T T-G G-C T-G A-T T-A
AA T A A Ir,II/; A
ACCG III T T T
A AGA
GGC C
T-AAGA
TAAAAT A T
T-A AG A
C-G C-G T-A T-A A-T A-T C-G A-A A-T A-T C-G A-T
C A T A T A T GTT*
T A T A GTC GCA
T-A A-T A-T G-C
T A TCW”A
ATTT ATTGGA~ C A T T-A A~
A A
T
CA A C
T-A cC
T-A c-c T-A A-T A-T
T A T G
TGC
C-G A-T G-C
T C T A TCG
Glutamine T-GA A-T A-T T-A T-A T-A T-A
T
ATGTG G yt7yAA;
C *III GGCAC
ZATGGTT G G
T-A AA T-A G-C A-T A-T
T T x TTG
A Leucine A-T
(WR) T-A
:I; G-T G-T
A G-C A
AT T TAG]leG
T ~~~!~A;T
T T
AG A A
A-T TT
Glycine G*; C-G T-A T-G A-T A-T G-C
T TG A - I GC T T A G" T-A T-A G-C G-T G-C T T T G GTG
Glutamate$;i T-A
:I; T-A A-T
T GGG
GGT T;"lfr,eG GT*
A T A T-A AA A-T A-T T-A A-T
T T f TTC
TTTGAA . T
AGTTTA AATA A - I A AGT
Lysine c-AG G-C C-G T-A
G"-: G-C
T
AAG~c~ G TC
AG
CTt AAC
Methionine A
A-T G-C A-T A-T A A
;I:
Leucine G
A-T
(Cum * X-i A-T
AG T A AG A GG P-A
ITAAT TTA
T CT: T_‘P; A-T G-C A-T
T A T A TAG
A-T A-T
T E TAA
Proline C-f4 Serine G-Z Serine G-: A-T (AGN) A-T
A-T (UCN) G-C
;I; Threonine :I%
T-A T-A
A-T A-T T-A T-A A-T CA
A-T A-T A-T TA A-T
G-C A-T
TT TT A
T
GAA TTGAj&;:
. . h* A TTTG
tAGGAG*
TG
"ITA
AAGC 'TTF'fFTTAA TATS ATTwAI~f~~ATTA A~TTTFAT
T-G G A-T
cf:cGGTTA " AIIIT A>
????"'i GGGTcA
ACT
CTA 'cAAAACc A;A
T-A C-G
T-A A-T ;I;
G-C T-A G-C G-C
A-T G-C G-C .
Z-E G-C T-A T-A
T G C A T : Fi
T G T A T i TGG GCT TGA TGT
Tryptophan ;I; G-C A-T T-A TT
A A A G - Cc T-A A-T A- T C-G C A T A
Tyrosine G-t G-T T-A A-T A-T A-T
T-A A-T G-C A-T
T A T GTA
A
Ualine A C-G A-T A-T A-T G.T T-A A-T
T-A
;I; T C T A TAC
ATC TAA
I I I TAG ,” TG TGAG~~
T TTAGd
FIG. 3.-Twenty-two putative tRNA genes from Limulus polyphemus mtDNA folded into conventional cloverleaf structures. Lines denote standard Watson-Crick pairs; dots denote G.T pairs.
letic group from which chelicerates are excluded (see also Friedrich and Tautz 1995).
The Limulus data indicate that the basic arrange- ment of genes in arthropod mtDNA was established pri- or to and has been relatively stable over at least 530 Myr. The chelicerate lineage to which Limulus belongs, Xiphosurida, was well established by the mid-Silurian, 420 MYA (Fisher 1984), and the divergence of cheli-
cerates and crustaceans had taken place prior to Middle Cambrian, 530 MYA, because a definitive chelicerate (Sunctucaris; Briggs and Collins 1988) and a represen- tative crustacean (Cunaduspis; Briggs 1978) are both present in the Burgess shale.
Conservation over this long period could suggest either strong positive selection for this particular gene arrangement or a strong mechanistic barrier to changes
Dow
nloaded from https://academ
ic.oup.com/m
be/article/14/8/867/982323 by guest on 21 Novem
ber 2021
Mitochondrial Gene Arrangement of the Horseshoe Crab 873
Table 2 Codon Usage in 13 Protein-Coding Genes of the Limdus Mitochondrial Genome
Amino Acid Codon N %
Amino Acid Codon N %
Amino Acid Codon N %
Amino Acid Codon N %
Phe TTT (GAA) TTC Leu TTA (UAA) TTG
Leu CTT (UAG) CTC
CTA CTG
Ile (GAU) Met (CAU)
Val (UAC)
ATT ATC ATA ATG
GTT GTC GTA GTG
105 5.4 50 2.6
119 6.2 34 1.8
57 2.9 35 1.8 64 3.3
5 0.3
139 7.2 64 3.3 90 4.7 20 1.0
44 2.3 4 0.2
43 2.2 14 0.7
Ser (UGA)
Pro (UGG)
Thr (UGU)
Ala (UGC)
TCT 50 2.6 TCC 40 2.1 TCA 62 3.2 TCG 5 0.3
CCT ccc CCA CCG
ACT ACC ACA ACG
GCT GCC GCA GCG
36 1.9 His 15 0.8 (GUG) 26 1.3 Gln
2 0.1 (UUG)
30 1.6 Asn 20 1.0 (GUU) 45 2.3 Lys
1 0.1 (CUU)
25 1.3 Asp 20 1.0 (GUC) 29 1.5 Glu
1 0.1 (UUC)
TYr GUN TER
TAT TAC TAA TAG
CAT CAC CAA CAG
AAT AAC AAA AAG
GAT GAC GAA GAG
49 2.5 15 0.8 3 0.1 2 0.1
25 1.3 21 1.1 30 1.5
7 0.4
61 3.1 28 1.4 39 2.0
7 0.4
21 1.1 7 0.4
38 2.0 12 0.6
CYS GCA) Trp WW A% WCG)
Ser (GCU)
GAY WCC)
TGT TGC TGA TGG
CGT CGC CGA CGG
17 0.9 6 0.3
48 2.5 9 0.5
7 0.4 4 0.2
16 0.8 5 0.2
AGT 11 0.6 AGC 8 0.4 AGA 37 1.9 AGG 6 0.3
GGT 16 0.8 GGC 14 0.7 GGA 45 2.3 GGG 25 1.3
NOTE.-N = the number of occurrences of each codon; % = the percentage of total codon usage over all genes based on 1,933 codons. Only complete codons were included in the analysis. Anticodons for each corresponding tRNA are shown in parentheses beneath their respective amino acid designations.
stop
in gene arrangement in general. Strong positive selection on mtDNA arrangement seems unlikely, given the rich- ness of the variation seen in this character among meta- zoans as a whole, and given the abruptness and global nature of some of the within-group rearrangements ob- served.
Conservation due to a strong mechanistic barrier appears more likely. The genes in metazoan mtDNA are usually closely adjacent, with few or no nucleotides sep- arating them (fig. 1 is typical). Given this highly com- pact arrangement, the probability of maintaining a func- tionally active mtDNA by any rearrangement process involving random breakage and rejoining would be low, since only a tiny proportion of such events will occur at or near enough to a gene boundary to allow continued function of both genes, and since a functional rearrange- ment requires two such events. Furthermore, there is ev- idence that most enzymes capable of mediating DNA rearrangements (e.g., those involved in recombination- based repair) are absent from mammalian mitochondrial systems. Studies of mtDNA inactivation kinetics and cell fusion studies using mtDNA markers both suggest that DNA repair processes, if present, are inefficient, and that recombination-based repair is probably absent (re- viewed in Brown 1983; Wolstenholme 1992). Given that many correlated characteristics are shared by metazoan mtDNAs in general, it is reasonable to hypothesize that this lack is general among metazoan mitochondrial sys- tems. Thus, mtDNA rearrangements appear to result from very rare accidents in which mitochondrial en- zymes (e.g., topoisomerase, transcript processing en- zymes, DNA polymerase) might participate, or on equal- ly rare or even rarer ones in which nuclear enzymes that promote recombination are accidentally targeted to mi- tochondria. The involvement of stem-loop structures and duplication-deletion cycles in the rearrangement process has been inferred (Mortiz and Brown 1987; Moritz, Dowling, and Brown 1987; Stanton et al. 1994), and
Macey et al. (1997) have recently proposed an important role for the origin of lagging strand replication in this process.
Acknowledgments
This work was supported by NSF grants DEB- 9303301 (to J.L.S. and W.M.B.) and DEB-9220640 (to W.M.B.), and by NASA Exobiology Program grant NAGW 4223. We thank J. C. Avise for the Limulus mtDNA, J. L. Boore and N. Campbell for many helpful comments on an earlier manuscript draft, and D. Stanton and E. Weinstein for early sequencing efforts.
LITERATURE CITED
ALTSCHUL, S. E, W. GISH, W. MILLER, E. W. MYERS, and D. J. LIPMAN. 1990. Basic local alignment search tool. J. Mol. Biol. 215:403-410.
BOORE, J. L., and W. M. BROWN. 1994~. Complete DNA se- quence of the mitochondrial genome of the black chiton, Katharina tunicata. Genetics 138:423-443.
- 1994b. Mitochondrial genomes and the phylogeny of mol&sks. Nautilus lOS(Supp1. 2):61-78.
-. 1995. Complete sequence of the mitochondrial DNA of the annelid worm Lumbricus terrestris. Genetics 141: 305-3 19.
BOORE, J. L., T. M. COLLINS, D. STANTON, L. L. DAEHLER, and W. M. BROWN. 1995. Deducing the pattern of arthropod phylogeny from mitochondrial DNA rearrangements. Na- ture 376: 163-165.
BRIGGS, D. E. G. 1978. The morphology, mode of life, and affinities of Cunadaspis perfectu (Crustacea: Phyllocarida), Middle Cambrian, Burgess Shale, British Columbia. Phils. Trans. R. Sot. Lond. B 281:439-487.
BRIGGS, D. E. G., and D. COLLINS. 1988. A Middle Cambrian chelicerate from Mount Stephen, British Columbia. Pale- ontology 31:779-798.
BROWN, W. M. 1983. Evolution of animal mitochondrial DNA. Pp. 62-88 in M. NEI and R. KOEHN, eds. Evolution of genes and proteins. Sinauer, Sunderland, Mass.
Dow
nloaded from https://academ
ic.oup.com/m
be/article/14/8/867/982323 by guest on 21 Novem
ber 2021
874 Staton et al.
-. 1985. The mitochondrial genome of animals. Pp. ~ 95-130 in R. J. MACINTYRE, ed. Molecular evolutionary genetics. Plenum Press, New York.
BROWN, W. M., M. GEORGE, JR., and A. C. WILSON. 1979. Rapid evolution of mitochondrial DNA. Proc. Natl. Acad. Sci. USA 76:1967-1971.
CANTATORE, I?, M. ROBERTI, G. RAINALDI, M. N. GADALETA, and C. SACCONE. 1989. The complete nucleotide sequence, gene order and genetic code of the mitochondrial genome of Purucentrotus Zividus. J. Biol. Chem. 264: 10965-10975.
CLARY, D. O., and D. R. WOLSTENHOLME. 1985. The mito- chondrial DNA molecule of Drosophila yakuba: nucleotide sequence, gene organization, and genetic code. J. Mol. Evol. 22:252-27 1.
CROZIER, R. H., and Y. C. CROZIER. 1993. The mitochondrial genome of the honeybee Apis metliferu-complete se- quence and genome organization. Genetics 133: 97- 117.
FLOOK, I!, C. H. E ROWELL, and G. GELLISSEN. 1995. The sequence, organization, and evolution of the Locustu mig- rutoriu mitochondrial genome. J. Mol. Evol. 41:928-941.
FISHER, D. C. 1984. The Xiphosurida: archetypes of bradytely? Pp. 196-213 in N. ELDREDGE and S. M. STANLEY, eds. Living fossils. Springer Verlag, New York.
FRIEDRICH, M., and D. TAUTZ. 1995. Rihosomal phylogeny of the major extant arthropod classes and the evolution of myr- iapods. Nature 376: 165-167.
GARCIA-MACHADO, E., N. DENNEBOUY, M. 0. SUAREZ, J.-C. MOUNOLOU, and M. MONNEROT. 1996. Partial sequence of the shrimp Penueus notiulis mitochondrial genome. C. R. Acad. Sci. Paris 319:473-486.
HATZOGLOU, E., G. C. RODAKIS, and R. LECANIDOU. 1995. Complete sequence and gene organization of the mitochon- drial genome of the land snail Albinuriu coeruleu. Genetics 140: 1353-1366.
HOFFMANN, R. J., J. L. BOORE, and W. M. BROWN. 1992. A novel mitochondrial genome organization for the blue mus- sel, Mytilus edulis. Genetics 131:397-412.
JACOBS, H. T., I? BALFE, B. L. COHEN, A. FARQUHARSON, and L. COMITO. 1988. Phylogenetic implications of genome re- arrangement and sequence evolution in echinoderm mito- chondrial DNA. Pp. 121-137 in C. R. C. PAUL and A. B. SMITH, eds. Echinoderm phylogeny and evolutionary biol- ogy. Clarendon Press, Oxford, England.
KYTE, J., and R. E DOOLI-ITLE. 1982. A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 157: 105-132.
LECANIDOU, R., V. DORIS, and G. C. RODAKIS. 1994. Novel features of metazoan mtDNA revealed from sequence anal- ysis of three mitochondrial DNA segments of the land snail Albinuriu turritu (Gastropoda: Clausiliidae). J. Mol. Evol. 38:369-382.
MACEY, J. R., A. LARSON, N. B. ANAJEVA, Z. FANG, and T. J. PAPENFUSS. 1997. Two novel gene orders and the role of light-strand replication in the rearrangement of the verte- brate mitochondrial genome. Mol. Biol. Evol. 14:91-104.
MITCHELL, S. E., A. E COCKBURN, and J. A. SEAWRIGHT. 1993. The mitochondrial genome of Anopheles quudrimuculutus
species A-complete nucleotide sequence and gene orga- nization. Genome 36: 1058-1073.
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.
MORITZ, C., T. E. DOWLING, and W. M. BROWN. 1987. Evo- lution of animal mitochondrial DNA: relevance for popu- lation biology and systematics. Annu. Rev. Ecol. Syst. 18: 269-292.
OJALA, D., J. MONTOYA, and G. A+ITARDI. 1981. tRNA punc- tuation model of RNA processing in human mitochondria. Nature 290:470-474.
OKIMOTO, R., H. M. CHAMBERLIN, J. L. MACFARLANE, and D. R. WOLSTENHOLME. 1991. Repeated sequence sets in mi- tochondrial DNA molecules of root knot nematodes (Me- Zoidogyne): nucleotide sequences, genome location and po- tential for host race identification. Nucleic Acids Res. 19: 1619-1626.
OKIMOTO, R., J. L. MACFARLANE, D. 0. CLARY, and D. R. WOLSTENHOLME. 1992. The mitochondrial genomes of two nematodes, Cuenorhubditis eleguns and Ascuris suum. Ge- netics 130:471-498.
SAUNDERS, N. C., L. G. KESSLER, and J. C. AVISE. 1986. Ge- netic variation and geographic differentiation in mitochon- drial DNA of the horseshoe crab, Limulus polyphemus. Ge- netics 112:613-627.
SMITH, M. J., A. ARNDT, S. GORSKI, and E. FAJBER. 1993. The phylogeny of echinoderm classes based on mitochondrial gene rearrangements. J. Mol. Evol. 36:545-554.
SMITH, M. J., D. K. BANFIELD, K. DOTEVAL, S. FORSKI, and D. J. KOWBEL. 1989. Gene rearrangement in sea star mi- tochondrial DNA demonstrates a major inversion event dur- ing echinoderm evolution. Gene 76: 18 l-l 85.
STANTON, D. J., L. L. DAEHLER, C. C. MORITZ, and W. M. BROWN. 1994. Potential stem-and-loop structures are asso- ciated with coding-region duplications in animal mitochon- drial DNA. Genetics 137:233-241.
TERRETT, J., S. MILES, and R. THOMAS. 1994. The mitochon- drial genome of Cepueu nemorulis (Gasteropoda, Stylom- matophora): gene order, base composition and heteroplas- my. The Nautilus lOS(Supp1. 2):79-84.
THOMPSON, J. D., D. G. HIGGINS, and T J. GIBSON. 1994. Clus- tal W-improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position- specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673-4680.
VALVERDE, J. R., B. BATUECAS, C. MORATILLA, R. MARCO, and R. GARESSE. 1994. The complete mitochondrial DNA sequence of the crustacean Artemiu franciscana. J. Mol. Evol. 39:400-408.
VAN RAAY, T. J., and T. J. CREASE. 1994. Partial mitochondrial DNA sequence of the crustacean Duphniu pulex. CWT. Gen- et. 2566-72.
WOLSTENHOLME, D. R. 1992. Animal mitochondrial DNA: structure and evolution. Int. Rev. Cytol. 141:173-216.
