A Novel Mitochondrial Gene Order in the Crinoid Echinoderm Florometra
serratissima
Andrea Scouras and Michael J. Smith
Department of Molecular Biology and Biochemistry, Simon Fraser University, Burnaby, British Columbia, Canada
The complete nucleotide sequence of the mitochondrial genome of the crinoid Florometra serratissima has been
determined. It is a circular DNA molecule, 16,005 bp in length, containing the genes for 13 proteins, small and
large ribosomal RNAs, and 22 transfer RNAs (tRNAs). Three regions of unassigned sequence (UAS) greater than
73 bp have been located. The largest, UAS I, is 432 bp long and exhibits sequence similarity to the putative
mitochondrial control regions seen in other animals. UAS II (77 bp) and UAS III (73 bp) are located between the
59 ends of coding sequences and may play roles as bidirectional promoters. Analyses of nucleotide composition
revealed that the major peptide-encoding strand is high in T and low in C. This bias is reflected in a specific pattern
of codon usage. Molecular phylogenetic analyses based on cytochrome c oxidase (COI, COII, and COIII) amino
acid and nucleotide sequences did not resolve all the relationships between echinoderm classes. The overall animal
mitochondrial gene content has been maintained in the crinoid, but there is extensive rearrangement with respect
to both the echinoid and the asteroid mtDNA gene maps. Florometra serratissima has a novel genome organization
in a segment containing most of the tRNA genes, large and small rRNA genes, and the NADH dehydrogenase
subunit 1 and 2 genes. Potential pathways and mechanisms for gene rearrangements between mitochondrial gene
maps of echinoderm classes and vertebrates are discussed as indicators of early deuterostome phylogeny.
Introduction
The complete DNA sequences of mitochondrial genomes are now available from a variety of different animals. Most of these genomes are circular DNA molecules that range in size from 14 to 17 kb. Linear mitochondrial genomes are found in some cnidarians (the
cubozoans, scyphozoans, and hydrozoans) (Bridge et al.
1992). There is a notable conservation of gene content
in animal mitochondrial genomes but gene maps vary
between taxa (Cantatore et al. 1989; Desjardins and Morais 1990; Pääbo et al. 1991; Okimoto et al. 1992; Wolstenholme 1992; Boore and Brown 1994a, 1994b, 1995;
Yamazaki et al. 1997). The animal mitochondrial genome contains genes for 13 proteins, the oxidative phosphorylation system cytochrome c oxidase subunits I, II,
and III (COI, COII, and COIII) and cytochrome b (Cyt
b), the ATPase complex subunits 6 and 8 (ATPase 6 and
ATPase 8), and the respiratory chain NADH dehydrogenase subunits 1–6 and 4L (ND1–ND6 and ND4L); 22
transfer RNA (tRNA) genes; and small and large ribosomal RNA genes (srRNA and lrRNA). In a limited
number of cases, sequences associated with the initiation
of replication and transcription of the mitochondrial
DNA (mtDNA) have been demonstrated (Walberg and
Clayton 1981; Clayton 1982, 1984, 1991; Montoya et
al. 1982; Montoya, Gaines, and Attardi 1983; Brown et
al. 1986).
Among vertebrates, the mtDNA gene maps are relatively stable, although rearrangements have been reported for a sea lamprey (Lee and Kocher 1995), frogs
(Yoneyama 1987; Macey et al. 1997), reptiles (Seutin et
al. 1994; Kumazawa and Nishida 1995; Quinn and MinKey words: Florometra, crinoid, mtDNA, mitochondrial gene rearrangements, nucleotide composition bias, echinoderm phylogeny.
Address for correspondence and reprints: Michael J. Smith, Department of Molecular Biology and Biochemistry, Simon Fraser University, 8888 University Boulevard, Burnaby, British Columbia, Canada V5A 1S6. E-mail: [email protected].
Mol. Biol. Evol. 18(1):61–73. 2001
q 2001 by the Society for Molecular Biology and Evolution. ISSN: 0737-4038
dell 1996; Macey et al. 1997), birds (Desjardins and
Morais 1990, 1991; Quinn and Wilson 1993; Härlid,
Janke, and Arnason 1997, 1998; Mindell, Sorenson, and
Dimcheff 1998), and marsupials (Pääbo et al. 1991; Janke et al. 1994; Janke, Xu, and Arnason 1997). In the
phylum Echinodermata, complete mitochondrial genome sequences are available for the echinoids (order
Camarodonta) Strongylocentrotus purpuratus (Jacobs et
al. 1988b), Paracentrotus lividus (Cantatore et al. 1989),
and (order Stirodonta) Arbacia lixula (De Giorgi et al.
1996), as well as the asteroid (order Valvatida) Asterina
pectinifera (Asakawa et al. 1995). Partial sequence data
from a second asteroid, Pisaster ochraceus (order Forcipulatida), is also available (Jacobs et al. 1989; Smith
et al. 1990).
The analysis of the complete DNA sequence of the
S. purpuratus mitochondrial genome (Jacobs et al.
1988b) demonstrated a novel gene map for echinoderms
in comparison to vertebrates and Drosophila. Considering only protein and rRNA genes, the echinoid mitochondrial map can be related to the vertebrate pattern
with only two rearrangements, the transpositions of
lrRNA and ND4L (Jacobs et al. 1988b; Cantatore et al.
1987b, 1989). Among echinoderms, a major inversion
event has been described in sea stars (class Asteroidea)
in comparison to sea urchins (class Echinoidea). The
asteroid mtDNA inversion contains a 4.6-kb segment
encompassing a 13-tRNA-gene cluster, ND1, ND2, and
lrRNA (Jacobs et al. 1989; Smith et al. 1989, 1990;
Asakawa et al. 1991, 1995). This inversion was seen in
two asteroid orders (Forcipulatida and Valvatida). PCR
and partial sequence data indicated that the sea cucumbers (class Holothuroidea) maintained the echinoid mitochondrial gene order, but the brittle stars (class Ophiuroidea) demonstrated the basic asteroid pattern (Smith
et al. 1993). This 4.6-kb inversion supports phylogenies
that have two major echinoderm lineages: an echinoid/
holothuroid lineage and an asteroid/ophiuroid one. Subsequent investigations have demonstrated the presence
61
62
Scouras and Smith
of a novel tRNA duplication event within the mtDNA
of the holothuroid genus Cucumaria with respect to other holothuroids (Arndt and Smith 1998) and multiple
inversion events succeeding the original 4.6-kb inversion in the ophiuroids (unpublished data).
This study reports the mtDNA sequence and genome map for a member of the echinoderm class Crinoidea, or feather stars, that is generally considered to
represent a primitive echinoderm lineage. Crinoidea is
the only extant class of the stalked echinoderm subphylum Pelmatozoa; the Asteroidea, Echinoidea, Holothuroidea, and Ophiuroidea classes belong to the free-living
subphylum Eleutherozoa. The overall mitochondrial
gene content has been maintained in the crinoid Florometra serratissima, but the mitochondrial genome demonstrates a unique gene order with respect to both the
echinoid and the asteroid mtDNA gene maps. The complete crinoid mtDNA sequence allows the examination
of echinoderm phylogenies by comparison of gene sequences and maps between echinoderm classes. With
knowledge of the mtDNA gene order for five echinoderm classes, both potential pathways and mechanisms
of echinoderm mitochondrial genome rearrangements
may be inferred.
Materials and Methods
mtDNA Isolation
Florometra serratissima specimens were obtained
by SCUBA diving at Bowen Island, B.C., or were supplied by the Bamfield Marine Station, Bamfield, B.C.
Florometra mtDNA was extracted by two methods. To
generate a partial mtDNA genomic library, mtDNA was
isolated according to Barriga Sosa et al. (1995), followed by CsCl isopycnic centrifugation. The purified
mtDNA was restriction-digested with HindIII according
to the supplier’s specifications (Bethesda Research Labs)
and cloned into the plasmid vector pUC19. mtDNA was
also obtained for direct PCR amplification from isolated
Florometra eggs. The eggs were digested in a final volume of 10 ml Protease K buffer (100 mM Tris-Cl [pH
8], 100 mM NaCl, 50 mM EDTA, 1% SDS, 200 mg/ml
Protease K) for 2 h at 658C. The digests were extracted
with an equal volume of 1 3 TE equilibrated phenol
(pH 8.0) until the interface was clean, and then they
were extracted with an equal volume of 1:1 phenol/sevag (24:1 chloroform/isoamyl alcohol). The phases were
separated by centrifugation in a Beckman JS-13.1 rotor
at 208C for 10 min at 8,000 rpm. The aqueous layer was
extracted with CTAB/NaCl (10% [w/v] cetyl trimethyl
ammonium bromide in 0.7M NaCl); a one-sixth volume
of 5 M NaCl was added, followed by the addition of a
one-eighth (original) volume of CTAB/NaCl. The sample was incubated at 658C for 35 min, mixed occasionally, and extracted with 0.5 volumes of sevag and centrifuged as before. The aqueous phase was precipitated
with 2.5 volumes of 95% ethanol at 2208C overnight.
The DNA was pelleted using a Beckman JA-13.1 rotor
at 13,000 rpm, 48C, for 20 min, rinsed in 70% ethanol,
and recentrifuged for 5 min. The dry pellet was taken
up in 1 ml of 1 3 TE.
PCR and DNA Sequencing
Florometra-specific oligonucleotide primers that
span gene junction regions were designed from sequence obtained from partial F. serratissima mtDNA
genomic HindIII clones. Internal gene primers were designed from consensus echinoderm sequences, from the
mtDNA sequence of the holothuroid Cucumaria miniata
(A. Arndt, personal communication), or from published
sequences. A complete list of the primer sequences and
relative positions used in this analysis may be obtained
from the authors on request. Circularity of the mtDNA
genome was indicated by the successful PCR amplification and sequencing across all genes and gene
junctions.
Florometra mtDNA was amplified in a Genetic
Thermal Cycler (GTC-2) (GL Applied Research Inc.)
under the following conditions: an initial cycle of denaturation at 948C for 90 s, annealing between 508C and
588C (depending on the primer utilized) for 30 s, and
extension at 728C for 1–3 min (approximately 45 s/kb
amplified product), followed by 29 cycles of 948C for
30 s, 50–588C for 30 s, and 728C for 1–3 min, with the
final cycle extension time being lengthened to 10 min.
The amplification products were electrophoresed in
0.7% agarose gels and stained with ethidium bromide.
The agarose gel slices containing the product band were
excised under long wavelength UV (360 nm), and the
DNA was purified from the agarose plug by sedimentation through silanized glass wool plugs for 20 min at
6,000 rpm in a desktop microcentrifuge, followed by an
ethanol precipitation. Amplified products were either sequenced directly or cloned using one of two methods:
the TA cloning method of Marchuk et al. (1990) utilizing the plasmid vector pUC19 or pBluescript II KS(1),
or use of the pCR-Script Amp SK(1) cloning kit (Stratagene) according to the manufacturer’s protocol.
All DNA sequences were obtained by the chain termination method of Sanger, Nicklen, and Coulson
(1977) using the Sequenase version 2.0 DNA sequencing kit, the Sequenase PCR product sequencing kit, or
the Thermo Sequenase radiolabeled terminator cycle sequencing kit (all from U.S. Biochemicals/Amersham).
Alternatively, sequences were obtained by automated
ABI (Applied Biosystems) sequencing protocols. Sequenase version 2.0 sequencing reaction mixtures with
recombinant plasmid template included 10% DMSO (T.
Snutch, personal communication). Standard plasmid
primers or specific PCR amplification primers were used
for DNA sequencing reactions. All reported sequences
resulted from two or more sequence determinations of
overlapping independent PCR amplifications, cloned
PCR fragments, or mtDNA genomic clones. Equivocal
segments were resequenced from novel PCR sites either
in the same strand or in the opposite strand.
Phylogenetic Methods and Data Analyses
Ribosomal RNA and tRNA genes were identified
by sequence comparison to known echinoderm genes.
In addition, tRNA genes were folded to verify their secondary structures (Sprinzl et al. 1989; Steinberg and
mtDNA Sequence of the Crinoid
Cedergren 1994; Steinberg, Gautheret, and Cedergren
1994). Unassigned sequence (UAS) regions were examined for stable stem-loop structures using the
PCFOLD program (Zuker and Stiegler 1981).
Florometra DNA sequence encoding mtDNA proteins was translated using the echinoderm mtDNA genetic code (Himeno et al. 1987; Jacobs et al. 1988b;
Cantatore et al. 1989; Asakawa et al. 1995; De Giorgi
et al. 1996). Since complete sequence data were not
available for all of the protein genes in every echinoderm class, the phylogenetic analyses were restricted to
the genes for cytochrome c oxidase subunits I, II, and
III. COI, COII, and COIII are all transcribed from the
same DNA strand. Florometra nucleotide and predicted
amino acid sequences were aligned with homologous
genes from the following deuterostome species (GenBank accession numbers in parentheses): the sea lamprey Petromyzon marinus (U11880) (Lee and Kocher
1995); the amphioxus Branchiostoma lanceolatum
(Y16474) (Spruyt et al. 1998); the hemichordate Balanoglossus carnosus (AF051097) (Castresana et al.
1998); the echinoids Strongylocentrotus purpuratus
(X12631) (Jacobs et al. 1988b), Paracentrotus lividus
(J04815) (Cantatore et al. 1989), and Arbacia lixula
(X80396) (De Giorgi et al. 1996); the asteroids Asterina
pectinifera (D16387) (Asakawa et al. 1995) and Pisaster
ochraceus (X55514) (Smith et al. 1990); the holothuroid
Cucumaria miniata; and the ophiuroid Ophiopholis aculeata (unpublished data). In addition, the three genes
(COI, COII, and COIII) were combined to form concatenated nucleotide (2,982 nt) and predicted amino acid
(994 aa) sequences. Alignments were constructed as follows: the individual translated COI, COII, and COIII
genes (start and stop codons removed) were aligned
manually using ESEE, version 3.2 (Cabot and Beckenbach 1989), and truncated at the 59 and 39 ends to the
first conserved amino acid. The three truncated amino
acid sequences were joined to form the concatenated
COI-COII-COIII amino acid segments. The individual
and concatenated nucleotide sequences were adjusted to
reflect these amino acid alignments. Protein and nucleotide maximum-likelihood trees were prepared using the
PROTML and NUCML programs in the MOLPHY
package, version 2.3 (Adachi and Hasegawa 1996a), and
with PUZZLE, version 4.0.2 (Strimmer and von Haeseler 1996). The mtREV24 model for mtDNA-encoded
proteins (Adachi and Hasegawa 1996b) was employed
for amino acid sequences, and the HKY85 model (Hasegawa, Kishino, and Yano 1985) was used for nucleotide sequences, excluding third codon positions. Differing tree topologies were compared using the KishinoHasegawa test (Kishino and Hasegawa 1989) at a 95%
confidence interval. Nucleotide sequences were analyzed in two ways: first and second nucleotide positions
were considered separately and their log likelihood values totaled, and first and second nucleotide positions
were analyzed together. The nucleotide sequences of the
individual and combined cytochrome c oxidase genes
were also analyzed using a LogDet paralinear distance
(Lockhart et al. 1994) implemented in the PAUP* pro-
63
FIG. 1.—The mitochondrial genome map of the crinoid Florometra serratissima. Arrows indicate the transcriptional polarity of each
gene. Names of protein-coding and ribosomal RNA genes are abbreviated as in the text. tRNA genes are represented by their single-letter
codes, and the arrowheads indicate their transcriptional polarity. Serine
and leucine tRNA genes are also identified by codon family (in parentheses). Unassigned sequence regions UAS I, UAS II, and UAS III are
shown as crosshatched areas. The lengths (bp) and positions of the
intergenic UAS regions are indicated inside the circular map.
gram, version 4.0b2a (Swofford 1998), excluding the
third nucleotide position of each codon.
Mitochondrial gene order rearrangements were analyzed directly by visually comparing proposed rearrangement steps between each mtDNA genome and by
utilizing the rearrangement program DERANGE II
(Blanchette, Kunisawa, and Sankoff 1996) to explore the
spectrum of potential rearrangements. The mitochondrial gene orders for all five echinoderm classes and the
consensus nonavian vertebrates were compared.
Results and Discussion
General Features of the F. serratissima Mitochondrial
Genome
The complete mtDNA sequence has been determined for F. serratissima (GenBank accession number
AF049132). A genetic map of the crinoid mitochondrial
genome is shown in figure 1. It is a circular DNA molecule, 16,005 bp in length, and it encodes the complete
sequences of 13 protein genes, 22 tRNA genes, and 2
ribosomal RNA genes as found in other animal mtDNA
(Jacobs et al. 1988b; Cantatore et al. 1989; Asakawa et
al. 1995; De Giorgi et al. 1996; reviewed in Wolstenholme 1992). There are three regions of UAS greater
than 73 bp and 14 short intergenic UAS segments (1–
16 bp in length: see fig. 1 for map locations).
All of the Florometra mtDNA protein genes begin
with ATG, except for ND4L, which appears to begin
with GTG. This use of GTG as an initiation codon for
64
Scouras and Smith
Table 1
Nucleotide Composition Percentages of the 13 Mitochondrial Protein Genes from the Major and Minor Sense Strands
of Representative Echinoderms
PERCENTAGE
SPECIES
Florometra serratissima . . . . . . .
Asterina pectinifera . . . . . . . . . .
Ophiopholis aculeata . . . . . . . . .
Arbacia lixula . . . . . . . . . . . . . . .
CLASS
A
T
C
G
G1T
Crinoidea
Asteroidea
Ophiuroidea
Echinoidea
26.9
27.4
28.0
27.0
45.6
32.5
35.3
35.4
12.1
23.5
19.3
20.9
15.4
16.5
17.5
16.7
61.0
49.0
52.8
52.1
Florometra ND4L is unique among echinoderm ND4L
genes. The ND4L genes of three echinoids and two asteroids utilize the initiation codons ATC or ATT (Jacobs
et al. 1988b; Cantatore et al. 1989; Smith et al. 1990;
Asakawa et al. 1995; De Giorgi et al. 1996), whereas
the ophiuroid uses the standard ATG (unpublished data).
GTG initiation codons have been reported in prokaryotes (Stormo, Schneider, and Gold 1982; reviewed in
Osawa et al. 1992), as well as in the mtDNA protein
genes of other echinoderms, vertebrates, Drosophila (reviewed in Wolstenholme 1992), and Mytilus edulis
(Hoffmann, Boore, and Brown 1992).
The Florometra mtDNA protein genes all contain
full termination codons, except COI, which terminates
with TA, and ND1 and ND2, which end with a T. In all
three cases, the 39 ends of the genes are punctuated with
a tRNA gene (fig. 1). A complete termination codon
may be created through posttranscriptional polyadenylation, as suggested for other organisms (Anderson et al.
1981; Ojala, Montoya, and Attardi 1981; Clayton 1991).
Animal mtDNA protein-coding genes with incomplete
stop codons TA or T are common (Attardi 1985; Wolstenholme 1992).
Apart from the small numbers of base pairs between identified genes, three significant regions of unassigned sequence (UAS I, UAS II, and UAS III) have
been found within the F. serratissima mtDNA genome
(fig. 1). The largest of these (UAS I) is 432 bp and is
located between tRNAAsp and tRNAThr (fig. 1). UAS II
(77 bp) is located between the 59 ends of ND1 and COI
(fig. 1), and UAS III (73 bp) is located between the 59
termini of Cyt b and ND6 (fig. 1). In echinoids, the
origin of replication has been shown to occur between
tRNAThr and tRNAPro (Jacobs, Herbert, and Rankine
1989). In asteroids, a putative control region is located
between tRNAThr and lrRNA (Smith et al. 1990; Asakawa et al. 1995). The Florometra UAS I region contains
sequence similarities to the putative control regions in
these other echinoderms. Adjacent to tRNAThr, there is
a high pyrimidine containing segment that is followed
by a G-rich region (13 out of 17 nucleotides are G) and
an A1T-rich region. UAS I can be folded into a complex stem-loop structure of high thermal stability (282.7
kcal/mol) if G-T pairing is allowed.
UAS III contains only A and T nucleotides, including a TTATATATAA motif that is frequently conserved in other intergenic regions of echinoderm
mtDNAs (Jacobs et al. 1989). In S. purpuratus, P. lividus, A. lixula, A. pectinifera, and A. amurensis, variations of this sequence motif occur between the 59 ends
of oppositely transcribed Cyt b and ND6. Since mitochondrial genes are transcribed from both DNA strands
in crinoids, these A1T-rich segments may function as
bidirectional promoters (Elliot and Jacobs 1989; Jacobs
et al. 1989).
The significance of the Florometra UAS II region
is still unclear. No obvious TTATATATAA-like motif
(Jacobs et al. 1989) is seen in UAS II, although the
region is A1T-rich (84.4% A1T). Florometra COI and
ND1 flank the UAS II region and are transcribed in opposite directions (fig. 1). The location and A1T content
of UAS II suggest that it may also act as a bidirectional
promoter, such as has been proposed between the oppositely transcribed COI and tRNAPro genes in asteroids
(Jacobs et al. 1989; Smith et al. 1990; Asakawa et al.
1995).
Nucleotide Bias and Codon Usage
Many animal mitochondrial genomes deviate from
a random usage of nucleotides, particularly in the third
positions of synonymous codons. These deviations are
said to result from directional mutation pressure (Sueoka
1962; Asakawa et al. 1991; Jermiin et al. 1994, 1996;
Jermiin, Graur, and Crozier 1995). In F. serratissima,
the mtDNA major sense strand includes 10 protein and
10 tRNA genes, whereas the minor sense strand includes
the remaining three protein (ND1, ND2, and ND6), 12
tRNA, and the 2 ribosomal RNA genes. The nucleotide
composition of the major sense strand is 26.4% A,
46.4% T, 15.6% G, and 11.6% C. When compared with
other echinoderm mtDNA sequences, a clear positive
bias for T with an apparent loss of C is seen in Florometra. This T-bias is also seen in the nucleotide composition of the protein genes. The nucleotide compositions of all 13 protein genes (from both the major and
minor sense strands) from representative echinoderms
are reported in table 1. The observed Florometra protein
gene T-bias is substantially higher than the reported Tbias (with a decrease in C) for the echinoid A. lixula
(De Giorgi et al. 1996) and for the other representative
echinoderms (table 1). This T over C bias is particularly
exaggerated in the third codon positions of the 10 F.
serratissima protein genes encoded on the major sense
strand (62.5% T vs. 1.9% C).
Several potential causes for nucleotide biases have
been suggested, including reduced pools for specific nucleotides, a biased preference of the mitochondrial gamma DNA polymerase for specific nucleotides, and a propensity for a specific mutational direction for relatively
mtDNA Sequence of the Crinoid
65
Table 2
Relative Synonymous Codon Usage in Four-Codon Families
MAJOR SENSE STRAND
MINOR SENSE STRAND
FOUR-CODON FAMILY
S.p.
A.p.
O.a.
F.s.
S.p.
A.p.
O.a.
F.s.
Ala . . . . . . .
0.94
1.54
1.27
0.24
0.34
0.51
2.42
0.73
0.57
0.86
1.76
0.81
1.19
0.80
1.59
0.42
1.17
1.29
1.29
0.25
1.31
1.36
1.12
0.21
0.56
0.84
1.79
0.81
1.02
1.24
1.55
0.20
1.41
0.69
1.25
0.65
0.69
1.52
1.49
0.29
0.21
0.28
2.83
0.69
0.52
0.64
2.36
0.49
1.18
0.78
1.79
0.25
1.17
1.61
1.14
0.08
0.99
1.55
1.21
0.25
0.47
1.09
2.08
0.36
0.97
1.69
1.12
0.22
0.94
0.97
1.80
0.29
1.24
1.24
1.22
0.30
1.00
0.50
1.75
0.75
0.34
0.74
2.66
0.26
1.11
0.86
1.89
0.14
1.30
1.23
1.44
0.04
1.21
0.94
1.68
0.17
0.68
0.80
2.20
0.32
0.97
1.20
1.66
0.17
0.99
0.71
2.01
0.28
3.35
0.00
0.59
0.06
2.69
0.15
0.58
0.58
2.47
0.12
0.70
0.72
3.76
0.07
0.17
0.00
3.40
0.04
0.50
0.07
3.42
0.08
0.41
0.08
2.04
0.18
1.19
0.60
3.15
0.10
0.65
0.10
2.94
0.02
0.86
0.18
2.40
0.80
0.80
0.00
1.33
0.00
1.33
1.33
0.80
0.53
1.60
1.07
2.67
0.44
0.44
0.44
2.00
1.33
0.67
0.00
1.54
0.62
0.92
0.92
0.44
0.44
1.78
1.33
1.00
1.00
2.00
0.00
1.63
0.89
1.19
0.30
2.33
0.47
0.65
0.56
1.33
0.27
1.33
1.07
1.00
0.35
0.94
1.71
2.13
0.17
0.68
1.02
2.13
0.40
0.80
0.67
2.87
0.27
0.20
0.67
1.27
0.20
1.37
1.17
2.50
0.67
0.50
0.33
2.04
0.16
0.76
1.04
2.22
0.43
0.92
0.43
1.45
0.36
1.45
0.73
1.53
0.26
0.72
1.49
2.92
0.38
0.54
0.16
2.58
0.27
0.61
0.54
2.39
0.32
0.97
0.32
2.16
0.16
0.72
0.96
2.00
0.25
1.13
0.63
2.18
0.34
0.74
0.74
1.16
0.53
2.22
0.09
0.00
0.25
3.50
0.25
1.13
0.21
2.56
0.10
0.72
0.00
2.51
0.19
0.97
0.12
2.91
0.00
1.21
0.29
2.50
0.00
0.43
0.14
3.29
0.14
0.77
0.43
2.81
0.00
1.00
0.40
2.60
0.00
Arg . . . . . .
Gly . . . . . .
Leu(CUN) . . .
Pro . . . . . . .
Ser(UCN) . . .
Ser(AGN) . . .
Thr . . . . . . .
Val . . . . . . .
GCU
GCC
GCA
GCG
CGU
CGC
CGA
CGG
GGU
GGC
GGA
GGG
CUU
CUC
CUA
CUG
CCU
CCC
CCA
CCG
UCU
UCC
UCA
UCG
AGU
AGC
AGA
AGG
ACU
ACC
ACA
ACG
GUU
GUC
GUA
GUG
NOTE.—Boldface type indicates the highest relative synonymous codon usage (Sharp, Tuohy, and Mosurski 1986) value for each four-codon family. The major
sense strands of Florometra serratissima (F.s.) and Asterina pectinifera (A.p.) contain the 10 protein genes Cyt b, COI, COII, COIII, ATPase 6, ATPase 8, ND3,
ND4, ND4L, and ND5, and the minor sense strands contain the remaining 3 protein genes, ND1, ND2, and ND6. The major sense strand of Ophiopholis aculeata
(O.a.) encodes COI, COII, COIII, ATPase 6, ATPase 8, ND3, ND4, ND4L, and ND5, and the minor sense strand encodes Cyt b, ND1, ND2, and ND6. The major
sense strand of Strongylocentrotus purpuratus (S.p.) includes all of the protein genes except ND6, which is on the minor strand.
long term exposed H-strands during replication (discussed in Asakawa et al. 1991). The immediate significance of these observed nucleotide biases is that extreme caution must be exercised in the use of conventional models for ascertaining molecular phylogenies
that assume no bias in nucleotide replacements.
A corollary of nucleotide bias is clearly reflected
in codon usage. In table 2, we compare the four-codon
family relative synonymous codon usage (RSCU)
(Sharp, Tuohy, and Mosurski 1986) values for S. purpuratus, A. pectinifera, O. aculeata, and F. serratissima
major and minor sense strands. RSCU is the observed
frequency of a codon divided by the expected frequency
if there is uniform usage of all codons for that amino
acid. In the major protein-encoding strand of F. serratissima, T is used most frequently in the third codon
position in every four-codon family. In the minor sense
strand, encoding ND1, ND2, and ND6, A is the most
recurrent third codon position nucleotide. The other
echinoderms do not exhibit this same type of codon bias
(table 2). Asterina pectinifera, S. purpuratus, and O.
aculeata demonstrate a codon usage pattern that appears
to be opposite to that of F. serratissima. Their major
protein-encoding strands generally follow an L-strand
pattern of codon usage (A or C in the third position)
adopted by most animal major sense strands. However,
the major protein-encoding strand of F. serratissima
seems to have a codon usage pattern commonly adopted
by H-strand proteins (T or G in the third position). The
codon usage pattern of the echinoid A. lixula is more
like that of F. serratissima, as there is similar usage of
A and T in both strands (De Giorgi et al. 1996).
Molecular Phylogenies
Traditionally, echinoderm phylogenies have been
based on larval and/or adult morphological characters
and fossil evidence. More recently, DNA and protein
sequence data have become central to the establishment
of phylogenies. We analyzed both nucleotide and pre-
66
Scouras and Smith
Table 3
Maximum-Likelihood Comparison of Proposed
Echinoderm Phylogenetic Trees
Tree
ln L
Dln L
SE
TBL
RELL-BP
Fa . . . .
Ga . . . .
Ha . . . .
Ib . . . . .
210,195.9
210,207.2
210,206.6
210,212.3
0.0
211.2
210.7
216.3
—
67.2
67.4
67.7
ME
1.4
1.3
1.9
0.9025
0.0403
0.0539
0.0033
NOTE.—These four alternative tree topologies were tested using the combined COI, COII, and COIII amino acid data by the Kishino-Hasegawa method
(Kishino and Hasegawa 1989). Respective log likelihood (ln L) values plus difference in log likelihood values (Dln L) from the preferred topology (tree F) are
shown with their corresponding standard errors (SE). TBL 5 total branch length;
ME 5 tree with the least total minimum edge length. Bootstrap probabilities
(RELL-BP) were determined by the RELL method (Kishino, Miyata, and Hasegawa 1990) with 104 replications.
a Tree topologies as shown in figure 2F–H.
b Tree topology: ((((((S.p., P.l.), A.l.), O.a.), ((P.o., A.p.), C.m.)), F.s.), B.c.,
(B.l., P.m.)). Species abbreviations are given in figure 2.
FIG. 2.—Alternative maximum-likelihood trees for echinoderm
phylogeny. Combined COI, COII, and COIII amino acid sequences
were analyzed using the mtREV24 model of amino acid substitution.
Petromyzon marinus (P.m.) and Branchiostoma lanceolatum (B.l.)
were used as outgroups. A–E, The five top scoring trees estimated by
an exhaustive search using PROTML in the MOLPHY package. F–H,
User trees constructed to reflect proposed echinoderm phylogenies described by Littlewood et al. (1997). Comparative log likelihood values
were generated with the Kishino-Hasegawa user specified tree option
(Kishino and Hasegawa 1989). The lowest log likelihood value (ln L)
is shown for tree A; the differences in log likelihood values (Dln L)
from tree A and the corresponding standard errors (SE) are shown
below trees B–H. A.l. 5 Arbacia lixula; A.p. 5 Asterina pectinifera;
B.c. 5 Balanoglossus carnosus; C.m. 5 Cucumaria miniata; F.s. 5
Florometra serratissima; O.a. 5 Ophiopholis aculeata; P.o. 5 Pisaster ochraceus; S.p. 5 Strongylocentrotus purpuratus.
dicted amino acid sequence alignments of COI, COII,
and COIII from representative echinoderm, hemichordate, and chordate species (see Materials and Methods).
The maximum-likelihood and LogDet approaches were
used primarily because of their resistance to artifacts
resulting from nucleotide compositional biases. Parsimony analyses were consistent with both LogDet and
maximum-likelihood results (data not shown).
The five highest scoring phylogenetic trees constructed from a PROTML maximum-likelihood analysis
utilizing the combined cytochrome c oxidase amino acid
sequences are shown in figure 2A–E. The phylogenetic
tree with the highest likelihood topology placed the crinoid as a sister group to clades containing echinoids,
asteroids, and the holothuroid; however, the ophiuroid
consistently coupled an extremely long branch length
with a position closer to the chordates than the crinoid
(fig. 2A). This same tree topology was resolved by quartet puzzling using the PUZZLE program for both amino
acid and nucleotide (first and second positions only) sequence analyses. Figure 2A is also the ‘‘best tree’’ generated in a NUCML maximum-likelihood analysis with
the combined cytochrome c oxidase nucleotide sequences (first and second nucleotide positions only) and from
a LogDet paralinear distance analysis on the same nucleotide data set. Of the remaining four top scoring trees
(fig. 2B–E), none have a likelihood value that is significantly different than that of figure 2A. The fourth tree
(fig. 2D) places Florometra as a sister group to the
eleutherozoan clades, but the holothuroid position is inconsistent with proposed echinoderm phylogenies (reviewed in Littlewood et al. 1997).
The log likelihood values of several tree topologies
that reflect a range of proposed echinoderm phylogenetic relationships were compared with that of the best
tree. The three most prominent echinoderm tree topologies, proposed by Littlewood et al. (1997), are shown
in figure 2F–H. In all of the proposed phylogenies, the
crinoids are positioned closer to the hemichordate/chordate clade than any of the eleutherozoan classes, including the ophiuroid. Tree topologies that place the crinoid
as a sister group to other echinoderms reflect the general
opinion that the crinoids and the eleutherozoans diverged very early in echinoderm evolution. All three
alternative trees have log likelihood values that are significantly greater than that of the ‘‘best tree’’ (fig. 2A).
Among trees constructed to compare the alternative
proposed echinoderm phylogenies (fig. 2F–H), that with
the lowest likelihood value places the holothuroid with
the echinoids and the ophiuroid as a sister group to the
rest of the eleutherozoans (fig. 2F). Forcing the ophiuroid into the echinoid/holothuroid clade (fig. 2G) or into
a clade with the asteroids (fig. 2H) does not result in
significantly different likelihood values (table 3). However, the likelihood value is significantly different when
the ophiuroid is grouped with the echinoids and the holothuroid is grouped with the asteroids (tree I in table 3).
The extremely long branch length of the ophiuroids
may have distorted the phylogenetic analysis. Alterna-
mtDNA Sequence of the Crinoid
tive phylogenetic trees were compared excluding the
ophiuroid data set. Without any imposed restrictions, the
best tree with the lowest likelihood value placed the crinoid as a sister group to the eleutherozoans, with the
holothuroid diverging prior to the echinoid/asteroid node
(fig. 3A). An alternative tree that groups the holothuroid
with the echinoids, separate from the asteroid clade (fig.
3B), is not significantly different from figure 3A. However, placing the holothuroid with the asteroids (fig. 3C)
results in a tree with a significantly different likelihood
value.
Several conclusions may be drawn from these types
of amino acid and nucleotide analyses. Although the
data resolve the monophyly of the echinoids and the
asteroids, they do not definitively reconcile the relationships between the echinoderm classes. Phylogenetic proposals that place the holothuroid with the echinoids rather than with the asteroids (reviewed in Littlewood et al.
1997) are not rejected by our data. However, the placement of the ophiuroid branch with respect to the other
echinoderms is still unresolved. Long branch lengths observed for the crinoid and especially the ophiuroid, resulting from either extreme nucleotide bias or differential evolutionary rates, may result in incorrect inferences
of phylogenetic relationships. A detailed report on the
mitochondrial genome and affinities of the ophiuroid
will be discussed elsewhere (unpublished data).
Mitochondrial Gene Rearrangements in Echinoderms
and Vertebrates
Recently, Boore and Brown (1998) noted three reasons why incorrect phylogenies may result from morphological or molecular data: (1) a lack of a satisfactory
number of shared features that support each branch, (2)
errors in assessing homology among features of various
organisms, and (3) convergent evolutionary changes being misinterpreted to support false relationships. To that
list should be added long branch effects due to base
composition biases or differential evolutionary rates
such as may be the case with Florometra and Ophiopholis. Mitochondrial gene arrangement comparisons
have been suggested as a more promising tool to resolve
deeper phylogenies (Brown 1985; Jacobs et al. 1988a;
Smith et al. 1993; Boore and Brown 1994a, 1995, 1998;
Macey et al. 1997). It has been argued that the large
number of possible gene arrangements would make it
unlikely that the same gene order arose from convergent
evolution (reviewed in Boore and Brown 1998). Therefore, shared gene arrangements would likely indicate a
common ancestry. Recently, however, the utility of mitochondrial gene order as a phylogenetic tool has been
questioned by Mindell, Sorenson, and Dimcheff (1998)
and Curole and Kocher (1999). These authors point out
both the potential for convergent evolution of gene order
and the limitations to forming phylogenetic conclusions
with insufficient knowledge of the rate and mechanisms
of gene rearrangement.
The mechanisms of gene rearrangement within animal mitochondrial genomes are restricted. There is no
direct evidence that intermolecular recombination be-
67
FIG. 3.—Maximum-likelihood tree comparisons excluding the
ophiuroid data set. Tree topologies were generated by an exhaustive
PROTML maximum-likelihood analysis (the mtREV24 model of amino acid substitution) with the combined COI, COII, and COIII genes.
A, The exhaustive search tree with the lowest log likelihood value (ln
L). B, Alternative tree topology that places the holothuroid as a sister
group to the echinoids. C, Alternative tree topology that places the
holothuroid as a sister group to the asteroids. Comparative log likelihood values were generated with the Kishino-Hasegawa user-specified
tree option (Kishino and Hasegawa 1989). The differences in log likelihood values (Dln L) and corresponding standard errors (SE) for trees
B and C versus tree A are given. Species are abbreviated as described
in figure 2.
tween animal mtDNA molecules occurs; however, intramitochondrial recombination has not been ruled out (Jacobs et al. 1988a; Boore and Brown 1998). Recently,
Lunt and Hyman (1997) presented evidence for the deletion of minicircles from the mtDNA genome of nematodes that is consistent with an excision-based model
of recombination/transposition. Many protein and rRNA
gene rearrangements between mtDNA genomes are
thought to be the result of intramolecular inversions
(Smith et al. 1993) or intramolecular duplications with
subsequent deletions of gene duplicates (Moritz and
Brown 1987). Since single tRNA genes may move more
frequently and perhaps by mechanisms distinct from
those of structural genes (Cantatore et al. 1987a; Jacobs
et al. 1989; Voytas and Boeke 1993), they are generally
excluded when calculating the number of rearrangement
steps required to obtain the different structural gene orders found between mitochondrial genomes. However,
pairs or groups of tRNA genes may be displaced by
inversions or transpositions, as is seen in both the ophiuroids (Smith et al. 1993; unpublished data) and the crinoids, or by duplication in some holothuroids (Arndt
and Smith 1998).
Clearly, there has been a complicated pattern of
rearrangements between the mitochondrial genomes in
the lineages leading to extant echinoderm classes and
the vertebrates. The crinoid mitochondrial gene order
encompassing COI to ND6 (fig. 1) is conserved across
all five echinoderm classes, with the notable exception
of a segment duplication in the holothuroid genus Cuc-
68
Scouras and Smith
FIG. 4.—Mitochondrial gene order arrangements between 59-Cyt
b and COI-39 of the basic echinoid, asteroid, and crinoid patterns. The
putative origin of replication is denoted in the echinoid as a displacement loop (dl) where the origin of replication has been demonstrated
(Jacobs, Herbert, and Rankine 1989), and as unassigned sequence
(UAS) in the asteroid and the crinoid. The tRNA genes are represented
by their single letter abbreviations. The leucine tRNA gene is further
identified by codon family (Lc 5 tRNALeu(CUN) and Lu 5 tRNALeu(UUR)).
Echinoids and asteroids contain three conserved blocks, boxed and
labeled A, B, and C for clarity. Arrows indicate the relative orientation
of the conserved blocks with respect to the echinoid map, and not the
transcriptional polarity of the individual genes. In the crinoid, the A
segment is maintained, but the B segment (tRNA cluster) is fragmented
into B9, containing 10 tRNA genes, B0, containing 2 tRNA genes, and
tRNATyr (designated *), inserted between lrRNA and ND2 (C9 segment).
umaria (Arndt and Smith 1998). Moreover, the genome
from COI to Cyt b is conserved in four of five echinoderm classes, the exception being the Ophiuroidea. This
region of the mtDNA genome (COI to Cyt b) is relatively stable when compared with the area around the
putative control region. The more variable mitochondrial genome segments from Cyt b to COI (encompassing
the putative control region) for the echinoid, asteroid,
and crinoid are shown in figure 4. Although most mitochondrial gene order rearrangements are found in this
region, particular blocks of genes are conserved between
echinoderms. The three complete echinoid mtDNA
maps, from species representing two orders, all exhibit
identical mtDNA gene arrangements (Jacobs et al.
1988b; Cantatore et al. 1989; De Giorgi et al. 1996).
These two echinoid orders were reportedly distinct by
150–200 MYA (Smith, Lafay, and Christen 1992) or
155 MYA (De Giorgi et al. 1996), supporting mtDNA
gene order stability within the class Echinoidea. Jacobs
et al. (1989), Smith et al. (1989, 1990), and Asakawa et
al. (1991, 1995) reported that the asteroids contained a
4.6-kb inversion in comparison with the mtDNA of the
echinoids. This inversion includes a 13-tRNA-gene cluster, ND1, ND2, and lrRNA (segments B and C in fig. 4).
The inversion is present in asteroids from the orders
Forcipulatida and Valvatida (Jacobs et al. 1989; Smith
et al. 1989, 1990, 1993; Asakawa et al. 1991, 1995) that
have been distinct for at least 225 Myr (Blake 1987).
PCR amplification of mitochondrial gene junctions indicated that the holothuroid Parastichopus californicus
has the basic echinoid mitochondrial gene orientation
(Smith et al. 1993). Similar studies with the ophiuroids
demonstrated that they displayed the basic asteroid pattern with respect to the 4.6-kb inversion event (Smith et
al. 1993). However, multiple inversion events occurred
in the ophiuroid subsequent to the characteristic 4.6-kb
inversion (unpublished data). In addition, the ophiuroids
(Smith et al. 1993; unpublished data), as well as holothuroid species from the genus Cucumaria (Arndt and
Smith 1998), exhibit altered tRNA clusters that contain
less than the 13 tRNA genes found in echinoids and
asteroids.
Although there are highly modified tRNA gene arrangements among the echinoderms, the original 4.6-kb
inversion supports an affinity of the echinoids with the
holothuroids as distinct from the asteroids. The Asteroidea and Ophiuroidea classes have been distinct since
the Ordovician (490–530 MYA), as have the Echinoidea
and Holothuroidea classes (450–530 MYA) (Smith
1988). Fossil evidence suggests that the class Crinoidea
was distinct from the rest of the echinoderms probably
before the end of the Lower Cambrian (550–560 MYA)
(Smith 1988). Since fossil evidence indicates a very early divergence of the crinoids from the eleutherozoans,
knowledge of the F. serratissima mtDNA gene map
might provide added insights into the structure of the
ancestral echinoderm mtDNA genome arrangement.
The original 4.6-kb inversion has a conserved gene
order and encompasses the region spanning from tRNA Pro (of the basic echinoid/holothuroid/asteroid
tRNA cluster) to lrRNA inclusive (segments B plus C
in fig. 4). In the echinoid gene order, the tRNA gene
cluster follows the putative control region (fig. 4). In the
asteroid mtDNA pattern, the polarity of the 4.6-kb segment is inverted, which joins the reversed tRNA gene
cluster to the 59 end of COI and an inverted lrRNA abutting the putative control region (fig. 4). In the crinoid,
the integrity of the 4.6-kb segment is not maintained.
Only part of the 4.6-kb segment, ND1 to lrRNA, is inverted in the Florometra mtDNA genome (segment C9
in fig. 4). The majority of the crinoid modified tRNA
cluster (10 out of 13 tRNA genes) is not inverted (segment B9 in fig. 4). The terminal three tRNA genes seen
in the 13-tRNA-gene cluster after tRNAAsp (59-tRNATyrtRNAGly-tRNALeu(UUR)-39) are displaced, with tRNAGlytRNALeu(UUR) probably as a single event between the ribosomal RNA genes (segment B0 in fig. 4). tRNATyr is
uniquely positioned, in comparison with other echinoderms, between the 39 end of ND2 and the 59 end of
lrRNA (59-ND2-tRNATyr-lrRNA-39) (figs. 1 and 4).
In the echinoids, the transcriptional order of the
ribososmal RNA genes is srRNA preceding lrRNA (fig.
4). In vertebrates, srRNA and lrRNA are also transcribed
in the same polarity, srRNA preceding lrRNA with an
intervening tRNAVal. In the crinoid, the ribosomal RNA
genes are transcribed from the same strand, but lrRNA
precedes srRNA (figs. 1 and 4). The presence of tRNAPhe
at the 59 end of srRNA is a conserved feature throughout
the echinoderms and vertebrates.
In addition to this novel ribosomal RNA gene arrangement, there is a unique organization of Cyt b, the
remaining Florometra tRNA cluster, and the putative
control region (fig. 4). The Florometra Cyt b is directly
adjacent to its modified tRNA cluster, and the putative
Florometra control region is situated on the opposite
side of the cluster. In echinoids, srRNA and the putative
control region separate the original tRNA cluster from
mtDNA Sequence of the Crinoid
FIG. 5.—Minimal step models relating the crinoid mtDNA map
to that of the echinoid and the asteroid patterns using protein-coding
and ribosomal RNA genes only. All gene names are abbreviated as in
the text. Gene orders start at COI and are arranged 59–39. Dark underlines indicate genes transcribed on the opposite strand. Shaded
curved arrows indicate inversions, and shaded straight arrows indicate
transversions. A, Outline of the minimum number of steps required to
correlate the echinoid or the asteroid gene order with that of the crinoid. After the echinoid↔crinoid gene order transversion, the crinoid
map was reoriented to a map starting with 59-COI-39. The crinoid↔asteroid rearrangement requires one inversion. B, An alternative
pathway to reach the crinoid gene order from that of the echinoid,
requiring two inversions.
Cyt b (fig. 4). The unique gene order between Cyt b and
the rRNA genes, including the modified tRNA cluster
and the putative control region, supports the hypothesis
that the 59-ND1-ND2-lrRNA-39 Florometra inversion occurred independent of the Cyt b to srRNA gene order
modifications. The class Crinoidea is the first of the extant echinoderms to appear in the fossil record (Paul and
Smith 1984; Smith 1988). However, the variations in
gene order seen between the echinoderm classes suggest
that the mitochondrial gene arrangement in crinoids is
most probably derivative of an ancestral pattern.
Phylogenetic Interpretations of Gene Orders
Several approaches, or ‘‘models,’’ may be used to
establish phylogenies based on gene orders. One could
disregard potential rearrangement mechanisms and simply score the minimum number of steps required to derive each gene order. Alternatively, the mechanism of
rearrangement could be considered and weighted for the
type, i.e., inversion, transposition, or transversion (transposition plus inversion), as suggested in the rearrangement model of Blanchette, Kunisawa, and Sankoff
(1996). We have no data on rates, frequency, or costs
of various rearrangement mechanisms, so weighting is
arbitrary. In addition, simply scoring the presence or
absence of gene junctions is not strictly a valid approach
because observed changes are not independent
characters.
There is only one change, the 4.6-kb inversion, between the basic echinoid/holothuroid and asteroid
mtDNA genome patterns. The ancestral polarity of the
69
FIG. 6.—Rearrangement paths required to achieve the various mitochondrial gene orders based on protein-coding and ribosomal RNA
genes and the tRNA gene cluster (tclus). Gene names are abbreviated
as in the text. Gene orders are arranged 59–39. Dark underlines indicate
genes transcribed on the opposite strand. Transpositions are indicated
by white straight arrows, transversions by shaded straight arrows, and
inversions by shaded curved arrows. A, Conversion steps between the
consensus nonavian vertebrate and echinoid mtDNA gene orders, disregarding the tRNA gene cluster. B, Conversion pathways between the
echinoid↔crinoid↔asteroid mitochondrial gene maps involving only
the variable region (including the tRNA cluster and the ND1, ND2,
srRNA, and lrRNA genes). C, Alternative conversion pathways between echinoid↔asteroid↔crinoid mitochondrial gene orders involving the variable region. In B and C, the Cyt b gene is included in the
partial gene orders to provide a focal point for their orientations.
inversion could be either orientation. Considering protein-coding and ribosomal RNA genes only, we identify
a minimum of one movement required to obtain the crinoid pattern from either the echinoid or the asteroid
gene order. The crinoid mtDNA gene order can be
reached from the echinoid order by means of the transversion of approximately two thirds of the echinoid genome (COI through Cyt b segment) to a position between srRNA and ND1 (fig. 5A). Conversely, the echinoid gene order can be reached by the opposite transversion of the crinoid COI-Cyt b segment between
srRNA and lrRNA. The crinoid gene order and that of
the asteroid can be related through the inversion of
srRNA (fig. 5A). An alternative path to correlate the crinoid gene order with that of the echinoid is with two
rearrangements: the inversion of srRNA and the inversion of 59-ND1-ND2-lrRNA-39 (fig. 5B).
In order to establish a polarity in gene rearrangements for echinoderms, an outgroup is required. A likely
candidate for such an outgroup could be the consensus
nonavian vertebrate gene map. Jacobs et al. (1988b),
excluding the tRNA genes, noted that the echinoid pattern and the vertebrate pattern could be related with one
protein (ND4L) and one rRNA (lrRNA) gene transposition (fig. 6A). Converting the asteroid gene order to that
of the vertebrate requires a minimum of two steps as
well: the ND4L transposition and the transversion of the
large COI-srRNA asteroid genome segment between
lrRNA and ND2. Correlating the crinoid gene order with
that of the vertebrates requires a minimum of three
steps: the ND4L transposition as in the echinoids and
asteroids, the transversion of the large COI-srRNA frag-
70
Scouras and Smith
ment as in the asteroids, and the inversion of srRNA.
Such COI-srRNA transversion events would require the
unlikely excision and insertion of a segment consisting
of approximately three fourths of the mitochondrial genome to the junction between the lrRNA and ND2 genes.
Alternatively, the asteroid pattern can be converted to
that of the vertebrate with three moves: the ND4L transposition and two independent inversions (lrRNA and
ND1-ND2). The crinoid↔vertebrate map conversion
can also be achieved in four moves: the transposition of
ND4L and three independent inversions (lrRNA, ND1ND2, and srRNA).
Therefore, if we exclude the tRNA genes and extremely large segment transversions, there are fewer
steps needed to convert the echinoid gene order to the
vertebrate gene order than from any of the other echinoderms. This implies that the ancestral polarity of the
4.6-kb inversion is the echinoid pattern. The asteroid
and crinoid patterns would then be derivative.
Superficially, these types of gene order analyses
could also imply a closer relationship between the crinoid and the asteroid rather than between the crinoid
and the echinoid. However, the 13-tRNA-gene cluster
should be included in the analysis, as it is maintained
in the echinoid, basic holothuroid, and asteroid genomes
and is lacking only the three terminal tRNA genes,
tRNATyr, tRNAGly, and tRNALeu(UUR), in the crinoid. Once
included, the numbers and mechanisms of the rearrangement steps necessary to convert the crinoid pattern to
that of either the echinoid or the asteroid are similar.
Both require two moves: one transversion and one inversion to reach the crinoid pattern (fig. 6B and C). Other models that adopt an inversion only mechanism of
movement and include the tRNA gene cluster also result
in an equal number of events required to correlate the
echinoid↔crinoid or the asteroid↔crinoid mtDNA
maps. Therefore, inclusion of the gene cluster does not
support a closer relationship of the crinoids to either the
echinoids or the asteroids.
Conclusions
The molecular data and mitochondrial map analyses, excluding the ophiuroid set, result in similar phylogenetic conclusions. First, both approaches support an
echinoid/holothuroid clade; the echinoid and the basic
holothuroid mitochondrial gene orders are identical.
Second, the position of the asteroids as a sister clade to
the echinoid/holothuroid clade is also reinforced by the
conserved block of genes in the 4.6-kb segments.
We propose that the original 4.6-kb inversion that
supports the separation of the echinoid/holothuroid and
asteroid lineages must have occurred after the split of
the Pelmatozoa, to which the crinoids belong, and the
Eleutherozoa, which includes the other four classes. The
relationship of the echinoid mtDNA pattern to that of
the vertebrates and crinoids suggests that the echinoderm ancestor to these groups may have had an mtDNA
gene order very similar to that of present day echinoids.
Although extant crinoids represent the class of echinoderms that appeared first in the fossil record (Paul and
Smith 1984; Smith 1988), all extant echinoderm classes
are equally distant from the common echinoderm ancestor. Considering the position of the modified tRNA
gene cluster in the crinoid (fig. 4), the apparent inversion
event in the Florometra mtDNA genome from ND1 to
lrRNA most likely happened in that lineage, independent
of the inversion seen in the asteroids. Multiple independent origins of a novel mitochondrial gene order have
also been demonstrated in birds (Mindell, Sorenson, and
Dimcheff 1998). Furthermore, it is unclear if this
mtDNA arrangement within F. serratissima is common
to all crinoids or if there is variation within the class
Crinoidea. Obtaining the mtDNA gene order from different species would establish whether the mtDNA gene
order within the crinoids is stable. The variability in
known echinoderm mitochondrial gene arrangements
emphasizes the limitations of gene order as a phylogenetic tool. As noted by Curole and Kocher (1999), until
a more complete sample of genomes is available and
some measure of the probabilities of various gene rearrangement mechanisms is known, phylogenetic conclusions based on gene order are tenuous.
Acknowledgments
We would like to thank S. Gorski for her work in
isolating some of the F. serratissima mitochondrial genomic clones, and K. Beckenbach for her technical assistance. We would also like to thank A. Arndt and B.
Hartwick for collecting most of our animals. This work
was supported by a National Science and Engineering
Research Council grant to M.J.S.
LITERATURE CITED
ADACHI, J., and M. HASEGAWA. 1996a. MOLPHY version 2.3:
programs for molecular phylogenetics based on maximum
likelihood. Comput. Sci. Monogr. 28:1–150.
———. 1996b. Model of amino acid substitution in proteins
encoded by mitochondrial DNA. J. Mol. Evol. 42:459–468.
ANDERSON, S., A. T. BANKIER, B. G. BARRELL et al. (14 coauthors). 1981. Sequence and organization of the human
mitochondrial genome. Nature 290:457–465.
ARNDT, A., and M. J. SMITH. 1998. Mitochondrial gene rearrangement in the sea cucumber genus, Cucumaria. Mol.
Biol. Evol. 15:1009–1016.
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.
ASAKAWA, S., Y. KUMAZAWA, T. ARAKI, H. HIMENO, K. MIURA, and K. WATANABE. 1991. Strand-specific nucleotide
composition bias in echinoderm and vertebrate mitochondrial genomes. J. Mol. Evol. 32:511–520.
ATTARDI, G. 1985. Animal mitochondrial DNA: an extreme
example of genetic economy. Int. Rev. Cytol. 93:93–145.
BARRIGA SOSA, I. A., K. BECKENBACH, B. HARTWICK, and M.
J. SMITH. 1995. The molecular phylogeny of five eastern
North Pacific octopus species. Mol. Phylogenet. Evol. 4:
163–174.
BLAKE, D. B. 1987. A classification and phylogeny of postPalaeozoic sea stars (Asteroidea: Echinodermata). J. Nat.
Hist. 21:481–528.
mtDNA Sequence of the Crinoid
BLANCHETTE, M., T. KUNISAWA, and D. SANKOFF. 1996. Parametric genome arrangement. Gene 172:GC11–GC17.
BOORE, J. L., and W. M. BROWN. 1994a. Mitochondrial genomes and the phylogeny of mollusks. Nautilus 108(Suppl.
2):61–78.
———. 1994b. Complete DNA sequence of the mitochondrial
genome of the black chiton, Katharina tunicata. Genetics
138:423–443.
———. 1995. Complete sequence of the mitochondrial DNA
of the annelid worm Lumbricus terrestris. Genetics 141:
305–319.
———. 1998. Big trees from little genomes: mitochondrial
gene order as a phylogenetic tool. Curr. Opin. Genet. Dev.
8:668–674.
BRIDGE, D., C. W. CUNNINGHAM, B. SCHIERWATER, R. DESALLE, and L. W. BUSS. 1992. Class-level relationships in
the phylum Cnidaria: evidence from mitochondrial genome
structure. Proc. Natl. Acad. Sci. USA 89:8750–8753.
BROWN, G. G., G. GADALETA, G. PEPE, C. SACCONE, and E.
SBISÀ. 1986. Structural conservation and variation in the Dloop-containing region of vertebrate mitochondrial DNA. J.
Mol. Evol. 192:503–511.
BROWN, W. M. 1985. The mitochondrial genome of animals.
Pp. 95–130 in R. J. MACINTYRE, ed. Molecular evolutionary genetics. Plenum Press, New York.
CABOT, E. L., and A. T. BECKENBACH. 1989. Simultaneous
editing of multiple nucleic acid and protein sequences with
ESEE, version 3.2. Comput. Appl. Biosci. 5:233–234.
CANTATORE, P., M. N. GADALETA, M. ROBERTI, C. SACCONE,
and A. C. WILSON. 1987a. Duplication and remolding of
tRNA genes during the evolutionary rearrangement of mitochondrial genomes. Nature 329:853–855.
CANTATORE, P., M. ROBERTI, P. MORISCO, G. RAINALDI, M. N.
GADALETA, and C. SACCONE. 1987b. A novel gene order in
the Paracentrotus lividus mitochondrial genome. Gene 53:
41–54.
CANTATORE, P., M. ROBERTI, G. RAINALDI, M. N. GADALETA,
and C. SACCONE. 1989. The complete nucleotide sequence,
gene organization, and genetic code of the mitochondrial
genome of Paracentrotus lividus. J. Biol. Chem. 264:
10965–10975.
CASTRESANA, J., G. FELDMAIER-FUCHS, S. YOKOBORI, N. SATOH, and S. PÄÄBO. 1998. The mitochondrial genome of the
hemichordate Balanoglossus carnosus and the evolution of
deuterostome mitochondria. Genetics 150:1115–1123.
CLAYTON, D. A. 1982. Replication of animal mitochondrial
DNA. Cell 28:693–705.
———. 1984. Transcription of the mammalian mitochondrial
genome. Annu. Rev. Biochem. 53:573–594.
———. 1991. Replication and transcription of vertebrate mitochondrial DNA. Annu. Rev. Cell Biol. 7:453–478.
CUROLE, J. P., and T. D. KOCHER. 1999. Mitogenomics: digging
deeper with complete mitochondrial genomes. Trends Ecol.
Evol. 14:394–398.
DE GIORGI, C., A. MARTIRADONNA, C. LANAVE, and C. SACCONE. 1996. Complete sequence of the mitochondrial DNA
in the sea urchin Arbacia lixula: conserved features of the
echinoid mitochondrial genome. Mol. Phylogenet. Evol. 5:
323–332.
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.
71
———. 1991. Nucleotide sequence and evolution of coding
and noncoding regions of a quail mitochondrial genome. J.
Mol. Evol. 32:153–161.
ELLIOT, D. J., and H. T. JACOBS. 1989. Mutually exclusive
synthetic pathways for sea urchin mitochondrial rRNA and
mRNA. Mol. Cell. Biol. 9:1069–1082.
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.
———. 1998. The complete mitochondrial genome of Rhea
americana and early avian divergences. J. Mol. Evol. 46:
669–679.
HASEGAWA, M., H. KISHINO, and T. YANO. 1985. Dating of
the human-ape splitting by a molecular clock of mitochondrial DNA. J. Mol. Evol. 22:160–172.
HIMENO, H., H. MASAKI, T. KAWAI, T. OHTA, I. KUMAGAI, K.
MIURA, and K. WATANABE. 1987. Unusual genetic codes
and a novel gene structure for tRNASer(AGY) in starfish mitochondrial DNA. Gene 56:219–230.
HOFFMANN, R. J., J. L. BOORE, and W. M. BROWN. 1992. A
novel mitochondrial genome organization for the blue mussel, Mytilus edulis. Genetics 131:397–412.
JACOBS, H. T., S. ASAKAWA, T. ARAKI, K. MIURA, M. J. SMITH,
and K. WATANABE. 1989. Conserved tRNA gene cluster in
starfish mitochondrial DNA. Curr. Genet. 15:193–206.
JACOBS, H. T., P. BALFE, B. L. COHEN, A. FARQUHARSON, and
L. COMITO. 1988a. Phylogenetic implications of genome
rearrangement and sequence evolution in echinoderm mitochondrial DNA. Pp. 121–137 in C. R. C. PAUL and A. B.
SMITH, eds. Echinoderm phylogeny and evolutionary biology. Oxford University Press, Oxford, England.
JACOBS, H. T., D. J. ELLIOT, V. B. MATH, and A. FARQUHARSON. 1988b. Nucleotide sequence and gene organization of
sea urchin mitochondrial DNA. J. Mol. Biol. 202:185–217.
JACOBS, H. T., E. R. HERBERT, and J. RANKINE. 1989. Sea
urchin egg mitochondrial DNA contains a short displacement loop (D-loop) in the replication origin region. Nucleic
Acids Res. 17:8949–8965.
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.
JANKE, A., X. XU, and U. ARNASON. 1997. The complete mitochondrial genome of the wallaroo (Macropus robustus)
and the phylogenetic relationship among Monotremata,
Marsupialia, and Eutheria. Proc. Natl. Acad. Sci. USA 94:
1276–1281.
JERMIIN, L. S., P. G. FOSTER, D. GRAUR, R. M. LOWE, and R.
H. CROZIER. 1996. Unbiased estimation of symmetrical directional mutation pressure from protein-coding DNA. J.
Mol. Evol. 42:476–480.
JERMIIN, L. S., D. GRAUR, and R. H. CROZIER. 1995. Evidence
from analyses of intergenic regions for strand-specific directional mutation pressure in metazoan mitochondrial
DNA. Mol. Biol. Evol. 12:558–563.
JERMIIN, L. S., D. GRAUR, R. M. LOWE, and R. H. CROZIER.
1994. Analysis of directional mutation pressure and nucleotide content in mitochondrial cytochrome b genes. J. Mol.
Evol. 39:160–173.
KISHINO, H., and M. HASEGAWA. 1989. Evaluation of the maximum likelihood estimate of the evolutionary tree topolo-
72
Scouras and Smith
gies from DNA sequence data, and the branching order in
Hominoidea. J. Mol. Evol. 29:170–179.
KISHINO, H., T. MIYATA, and M. HASEGAWA. 1990. Maximum
likelihood inference of protein phylogeny and the origin of
chloroplasts. J. Mol. Evol. 31:151–160.
KUMAZAWA, Y., and M. NISHIDA. 1995. Variations in mitochondrial tRNA gene organization of reptiles as phylogenetic markers. Mol. Biol. Evol. 12:759–772.
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.
LITTLEWOOD, D. T. J., A. B. SMITH, K. A. CLOUGH, and R. H.
EMSON. 1997. The interrelationships of the echinoderm
classes: morphological and molecular evidence. Biol. J.
Linn. Soc. 61:409–438.
LOCKHART, P. J., M. A. STEEL, M. D. HENDY, and D. PENNY.
1994. Recovering evolutionary trees under a more realistic
model of sequence evolution. Mol. Biol. Evol. 11:605–612.
LUNT, D. H., and B. C. HYMAN. 1997. Animal mitochondrial
DNA recombination. Nature 387:247.
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.
MARCHUK, D., M. DRUMM, A. SAULINO, and F. S. COLLINS.
1990. Construction of T-vectors, a rapid and general system
for direct cloning of unmodified PCR products. Nucleic Acids Res. 19:1154.
MINDELL, D. P., M. D. SORENSON, and D. E. DIMCHEFF. 1998.
Multiple independent origins of mitochondrial gene order
in birds. Proc. Natl. Acad. Sci. USA 95:10693–10697.
MONTOYA, J., T. CHRISTIANSON, D. LEVENS, M. RABINOWITZ,
and G. ATTARDI. 1982. Identification of initiation sites for
heavy-strand and light-strand transcription in human mitochondrial DNA. Proc. Natl. Acad. Sci. USA 79:7195–7199.
MONTOYA, J., G. L. GAINES, and G. ATTARDI. 1983. The pattern of transcription of the human mitochondrial rRNA
genes reveals two overlapping transcriptional units. Cell 34:
151–159.
MORITZ, C., and W. M. BROWN. 1987. Tandem duplication in
animal mitochondrial DNAs: variation in incidence and
gene content among lizards. Proc. Natl. Acad. Sci. USA 84:
7183–7187.
OJALA, D., J. MONTOYA, and G. ATTARDI. 1981. tRNA punctuation model of RNA processing in human mitochondria.
Nature 290:470–474.
OKIMOTO, R., J. L. MACFARLANE, D. O. CLARY, and D. R.
WOLSTENHOLME. 1992. The mitochondrial genomes of two
nematodes, Caenorhabditis elegans and Ascaris suum. Genetics 130:471–498.
OSAWA, S., T. H. JUKES, K. WATANABE, and A. MUTO. 1992.
Recent evidence for evolution of the genetic code. Microbiol. Rev. 56:229–264.
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.
PAUL, C. R. C., and A. B. SMITH. 1984. The early radiation
and phylogeny of echinoderms. Biol. Rev. 59:443–481.
QUINN, T. H., and D. P. MINDELL. 1996. Mitochondrial gene
order adjacent to the control region in crocodile, turtle, and
tuatara. Mol. Phylogenet. Evol. 5:344–351.
QUINN, T. H., and A. C. WILSON. 1993. Sequence evolution in
and around the mitochondrial control region in birds. J.
Mol. Evol. 37:417–425.
SANGER, F., S. NICKLEN, and A. R. COULSON. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl.
Acad. Sci. USA 74:5463–5467.
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.
SHARP, P. M., T. M. F. TUOHY, and K. R. MOSURSKI. 1986.
Codon usage in yeast: cluster analysis clearly differentiates
highly and lowly expressed genes. Nucleic Acids Res. 14:
5125–5143.
SMITH, A. B. 1988. Fossil evidence for the relationships of
extant echinoderm classes and their times of divergence. Pp.
85–97 in C. R. C. PAUL and A. B. SMITH, eds. Echinoderm
phylogeny and evolutionary biology. Oxford University
Press, Oxford, England.
SMITH, A. B., B. LAFAY, and R. CHRISTEN. 1992. Comparative
variation of morphological and molecular evolution through
geological time: 28S ribosomal RNA versus morphology in
echinoids. Philos. Trans. R. Soc. Lond. B Biol. Sci. 338:
365–382.
SMITH, M. J., A. ARNDT, S. GORSKI, and E. FAJBER. 1993. The
phylogeny of echinoderm classes based on mitochondrial
gene arrangements. J. Mol. Evol. 36:545–554.
SMITH, M. J., D. K. BANFIELD, K. DOTEVAL, S. GORSKI, and
D. J. KOWBEL. 1989. Gene arrangement in sea star mitochondrial DNA demonstrates a major inversion event during echinoderm evolution. Gene 76:181–185.
———. 1990. Nucleotide sequence of nine protein-coding
genes and 22 tRNAs in the mitochondrial DNA of the sea
star Pisaster ochraceus. J. Mol. Evol. 31:195–204.
SPRINZL, M., T. HARTMANN, J. WEBER, J. BLANK, and R. ZEIDLER. 1989. Compilation of tRNA sequences and sequences
of tRNA genes. Nucleic Acids Res. 17:r1–r172.
SPRUYT, N., C. DELARBRE, G. GACHELIN, and V. LAUDET.
1998. Complete sequence of the amphioxus (Branchiostoma
lanceolatum) mitochondrial genome: relations to vertebrates. Nucleic Acids Res. 26:3279–3285.
STEINBERG, S., and R. CEDERGREN. 1994. Structural compensation in atypical mitochondrial tRNAs. Struct. Biol. 1:507–
510.
STEINBERG, S., D. GAUTHERET, and R. CEDERGREN. 1994. Fitting the structurally diverse animal mitochondrial tRNAsSer
to common three-dimensional constraints. J. Mol. Biol. 236:
982–989.
STORMO, G. D., T. D. SCHNEIDER, and L. M. GOLD. 1982.
Characterization of translational initiation sites in E. coli.
Nucleic Acids Res. 10:2971–2996.
STRIMMER, K., and A. VON HAESELER. 1996. Quartet puzzling:
a quartet maximum likelihood method for reconstructing
tree topologies. Mol. Biol. Evol. 13:964–969.
SUEOKA, N. 1962. On the genetic basis of variation and heterogeneity of DNA base composition. Proc. Natl. Acad. Sci.
USA 48:582–592.
SWOFFORD, D. L. 1998. PAUP*: phylogenetic analysis using
parsimony (*and other methods). Version 4.0b2a. Sinauer,
Sunderland, Mass.
VOYTAS, D. F., and J. D. BOEKE. 1993. Yeast retrotransposons
and tRNAs. Trends Genet. 9:421–427.
WALBERG, M. W., and D. A. CLAYTON. 1981. Sequence and
properties of the human KB cell and mouse L cell D-loop
mtDNA Sequence of the Crinoid
regions of mitochondrial DNA. Nucleic Acids Res. 9:5411–
5421.
WOLSTENHOLME, D. R. 1992. Animal mitochondrial DNA:
structure and evolution. Int. Rev. Cytol. 141:173–216.
YAMAZAKI, N., R. UESHIMA, J. A. TERRETT et al. (12 co-authors). 1997. Evolution of pulmonate gastropod mitochondrial genomes: comparisons of gene organizations of Euhadra, Cepaea and Albinaria and implications of unusual
tRNA secondary structures. Genetics 145:749–758.
YONEYAMA, Y. 1987. The nucleotide sequences of the heavy
and light strand replication origins of the Rana catesbeiana
73
mitochondrial genome. J. Nippon Med. Sch. 54:429–440
[in Japanese].
ZUKER, M., and P. STEIGLER. 1981. Optimal computer folding
of large RNA sequences using thermodynamics and auxiliary information. Nucleic Acids Res. 9:133–148.
STEPHEN PALUMBI, reviewing editor
Accepted September 29, 2000