Abstract

The cloverleaf secondary structure of transfer RNA (tRNA) is highly conserved across all forms of life. Here, we provide sequence data and inferred secondary structures for all tRNA genes from 8 new arachnid mitochondrial genomes, including representatives from 6 orders. These data show remarkable reductions in tRNA gene sequences, indicating that T-arms are missing from many of the 22 tRNAs in the genomes of 4 out of 7 orders of arachnids. Additionally, all opisthothele spiders possess some tRNA genes that lack sequences that could form well-paired aminoacyl acceptor stems. We trace the evolution of T-arm loss onto phylogenies of arachnids and show that a genome-wide propensity to lose sequences that encode canonical cloverleaf structures likely evolved multiple times within arachnids. Mapping of structural characters also shows that certain tRNA genes appear more evolutionarily prone to lose the sequence coding for the T-arm and that once a T-arm is lost, it is not regained. We use tRNA structural data to construct a phylogeny of arachnids and find high bootstrap support for a clade that is not supported in phylogenies that are based on more traditional morphological characters. Together, our data demonstrate variability in structural evolution among different tRNAs as well as evidence for parallel evolution of the loss of sequence coding for tRNA arms within an ancient and diverse group of animals.

Introduction

The cloverleaf-shaped secondary structure of transfer RNA (tRNA) is one of the most highly conserved features known among living organisms. Because tRNA plays an essential role in protein synthesis, natural selection presumably maintains tRNA structure so that tRNA molecules can interact effectively with other components of the translation system. Mitochondrial (mt) tRNAs of metazoans are typically reduced in size relative to those of their free-living bacterial relatives and to their counterpart tRNAs encoded by the nucleus (Wolstenholme 1992; Lynch 1996), but most mt tRNAs maintain a canonical cloverleaf secondary structure.

Some metazoans are known to possess one to a few mt tRNA genes that lack the sequence for either the T-arm or the D-arm of the cloverleaf structure (Hatzoglou et al. 1995; Terrett et al. 1996; Boore and Brown 2000; Kurabayashi and Ueshima 2000; Helfenbein et al. 2001; Yokobori et al. 2005; Waeschenbach et al. 2006). However, most of the 22 tRNA genes encoded within these metazoan mt genomes are inferred to possess cloverleaf secondary structures. One exception shared among metazoans is the gene coding for tRNASer(AGN), which apparently lost sequence coding for the D-arm prior to metazoan diversification (Wolstenholme 1992).

The most extensive loss of the cloverleaf structure from mt tRNAs has occurred in nematodes. Members of 1 nematode lineage possess mt genomes in which all the tRNA genes (except those coding for tRNASer) lack the sequence that encodes the T-arm and in its place have sequence for a TV-replacement loop (Wolstenholme et al. 1987; Okimoto et al. 1992; Keddie et al. 1998; He et al. 2005; Montiel et al. 2006). The apparent T-arm losses evident from DNA sequencing have been confirmed in the mature tRNAs of 1 of these nematodes, in which it has also been shown that these mt tRNAs possess typical L-shaped structures and therefore are likely functional (Watanabe et al. 1994). Within nematodes, taxa in the class Secernentea are delimited by T-arm loss, whereas the non-secernentean Trichinella nematode possesses some cloverleaf-shaped tRNAs (Lavrov and Brown 2001).

Although it was thought that the extensive evolutionary loss of T-arms from most tRNA genes in the mt genome was unique to nematodes, we subsequently found that a jumping spider (Araneae; Araneomorphae; Habronattus oregonensis) mt genome possesses many extremely truncated tRNA genes that apparently encode neither a T-arm nor the 3′ aminoacyl acceptor stem (Masta and Boore 2004). Because most of this spider's 22 mt tRNA genes are truncated, it suggests that spider mitochondria are able to function with most or all of their tRNA genes drastically shorter than those of other metazoans. Subsequent analyses of a distantly related spider (Araneae; Mygalomorphae) revealed that it, too, possesses many similarly truncated tRNA genes (Qiu et al. 2005).

These findings in spiders prompted us to ask whether other members of the Class Arachnida possess numerous truncated tRNA genes and lack well-paired acceptor stems. In general, non-arachnid arthropods primarily possess typical mt tRNA genes whose sequences can be folded into cloverleaf structures with well-paired stems. One exception is the centipede Lithobius forticatus, which has been found to possess some tRNA genes with mispaired aminoacyl acceptor stem sequences that are posttranscriptionally edited (Lavrov et al. 2000). The origin of the aberrant tRNA genes we found in spiders is unknown.

To understand the evolutionary history of loss of structural features from mt tRNAs, we need knowledge of the evolutionary relationships among the taxa under study. Multiple phylogenies have been published for arachnids, with varying degrees of support for relationships among the 11 orders. Phylogenetic trees that include analysis of morphological data typically group Araneae, Amblypygi, Thelyphonida, and Schizomida in a clade (referred to as the Tetrapulmonata or 4-lunged arachnids), although inferred relationships among these orders vary (Weygoldt and Paulus 1979; Shultz 1990; Wheeler and Hayashi 1998; Giribet et al. 2002; Shultz 2007). Likewise, Scorpiones, Opiliones, Pseudoscorpiones, and Solifugae are generally agreed to be more closely related (in a clade called Dromopoda) than they are to the Tetrapulmonata, but researchers disagree about their relationships with other arachnid orders (Shultz 2007).

Although there are 11 orders of arachnids, most mt genome sequencing efforts have focused on the order Acari, the mites and ticks. This order is composed of 2 divergent lineages, the Acariformes and the Parasitiformes. Of the 15 published acarid genomes, 12 are from the parasitiform lineage. Some of the 12 parasitiform mt genomes exhibit tRNA genes with reduced T-arm sequence, but in general, the inferred structures do not differ from those of most arthropods. The 3 published mt genomes from the acariform lineage, all from the genus Leptotrombidium, possess many tRNA genes that lack either D-arm or T-arm sequence (Shao et al. 2005, 2006).

Besides spiders (Araneae) and mites and ticks (Acari), scorpions (Scorpiones) are the only arachnid order for which mt genome data have been published. The 2 published scorpion mt genomes, both from the family Buthidae, also possess some tRNA genes that lack T-arm sequence (Davila et al. 2005; Gantenbein et al. 2005). Scorpions, spiders, and mites and ticks are not thought to be closely related, but instead, represent some of the most deeply diverged lineages within Arachnida. It is not known whether the evolutionary loss of T-arm sequence from mt tRNA genes occurred early in the evolution of arachnids and is a synapomorphy shared by most arachnids or whether this loss evolved independently in multiple lineages. We aimed to address this question by better determining the evolutionary extent of tRNA gene truncation in arachnid mt genomes. To do this, we sequenced representative taxa from 6 of the 11 orders of arachnids. We included multiple taxa from divergent lineages within orders thought to have diversified long ago and therefore to harbor significant genetic diversity.

In this study, we present inferred secondary structure data for all tRNA genes from 8 newly sequenced mt genomes. Four taxa are from the Tetrapulmonata and 4 are from the Dromopoda. Two taxa are spiders, which together with previously published data span the highly diverse order Araneae. Two taxa are scorpions, selected to span the deepest divergences within the order. One taxon is from the Solifugae, or camel spiders. Three taxa are representatives of arachnid orders for which there is no prior mt genome data: Amblypygi (whip-spiders), Thelyphonida or Uropygi (vinegaroons), and Opiliones (harvestmen). Together, these taxa span the diversity of the Arachnida, an ancient and diverse group.

We use 2 approaches to examine the evolution of mt tRNA genes within arachnids. First, we code inferred tRNA structural data as a phylogenetic character and trace the evolution of tRNA structure onto published phylogenies of arachnids. Second, we use our data sets on inferred tRNA structure to construct phylogenetic trees, which we then contrast with trees constructed from more traditional morphological characters. Rare genomic changes such as loss of conserved structural features may be used as phylogenetic markers to delimit clades of organisms that possess such features (reviewed in Rokas and Holland 2000). Together, these approaches allow us to assess how often arachnid mt genomes may have evolved the propensity to lose sequences that encode the T-arm of tRNAs and whether some tRNA genes have been more evolutionarily labile in their loss of sequence that encodes a region of the canonical cloverleaf structure of tRNA.

Materials and Methods

Taxon Sampling and mt Genome Amplification and Sequencing

We sequenced the mt genomes of 8 taxa from 6 orders of arachnids. We also included published data from another 8 chelicerate mt genomes (see table 1). Together, these genomes include at least 1 representative taxon from 7 of the 11 orders of arachnids.

Table 1

Taxa Sampled in This Study

OrderSuborderFamilyGenusSpeciesGenBank NumberReferences
AcariParasitiformesIxodidaeRhipicephalussanguineusAF081829Black and Roehrdanz (1998)
ParasitiformesIxodidaeIxodeshexagonusAF081828Black and Roehrdanz (1998)
ParasitiformesVarroidaeVarroadestructorAJ493124Navajas et al. (2002)
AcariformesTrombiculidaeLeptotrombidiumpallidumAB180098Shao et al. (2005)
AmblypygiPhrynidaePhyrnussp.EU520641This study
AraneaeMesothelaeHeptathelidaeHeptathelakimuraiAY309258Qui et al. (2005)
MygalomorphaeTheraphosidaeOrnithoctonushuwenaAY309259Qui et al. (2005)
MygalomorphaeTheraphosidaeAphonopelmasp.EU523754This study
AraneomorphaeHypochilidaeHypochilusthorelliEU523753This study
AraneomorphaeSalticidaeHabronattusoregonensisAY571145Masta and Boore (2004)
OpilionesPalpatoresPhalangiidaePhalangiumopilioEU523757This study
ScorpionesButhidaeButhusoccitanusEU523755This study
ChactidaeaUroctonusmordaxEU523756This study
SolifugaeEremobatidaeEremobatescfr. palpisetulosusEU520642This study
ThelyphonidaThelyphonidaeMastigoproctusgiganteusEU520643This study
XiphosuraXiphosuridaLimulidaeLimuluspolyphemusAF216203Lavrov et al. (2000)
OrderSuborderFamilyGenusSpeciesGenBank NumberReferences
AcariParasitiformesIxodidaeRhipicephalussanguineusAF081829Black and Roehrdanz (1998)
ParasitiformesIxodidaeIxodeshexagonusAF081828Black and Roehrdanz (1998)
ParasitiformesVarroidaeVarroadestructorAJ493124Navajas et al. (2002)
AcariformesTrombiculidaeLeptotrombidiumpallidumAB180098Shao et al. (2005)
AmblypygiPhrynidaePhyrnussp.EU520641This study
AraneaeMesothelaeHeptathelidaeHeptathelakimuraiAY309258Qui et al. (2005)
MygalomorphaeTheraphosidaeOrnithoctonushuwenaAY309259Qui et al. (2005)
MygalomorphaeTheraphosidaeAphonopelmasp.EU523754This study
AraneomorphaeHypochilidaeHypochilusthorelliEU523753This study
AraneomorphaeSalticidaeHabronattusoregonensisAY571145Masta and Boore (2004)
OpilionesPalpatoresPhalangiidaePhalangiumopilioEU523757This study
ScorpionesButhidaeButhusoccitanusEU523755This study
ChactidaeaUroctonusmordaxEU523756This study
SolifugaeEremobatidaeEremobatescfr. palpisetulosusEU520642This study
ThelyphonidaThelyphonidaeMastigoproctusgiganteusEU520643This study
XiphosuraXiphosuridaLimulidaeLimuluspolyphemusAF216203Lavrov et al. (2000)
a

Uroctonus has recently been moved from the superfamily Vaejovoidea, family Vaejovidae, and placed into the superfamily Chactoidea, family Chactidae (Soleglad and Fet 2003). We do not endorse one view over the other, but for convenience in this paper, we use the later placement of Uroctonus in Chactidae.

Table 1

Taxa Sampled in This Study

OrderSuborderFamilyGenusSpeciesGenBank NumberReferences
AcariParasitiformesIxodidaeRhipicephalussanguineusAF081829Black and Roehrdanz (1998)
ParasitiformesIxodidaeIxodeshexagonusAF081828Black and Roehrdanz (1998)
ParasitiformesVarroidaeVarroadestructorAJ493124Navajas et al. (2002)
AcariformesTrombiculidaeLeptotrombidiumpallidumAB180098Shao et al. (2005)
AmblypygiPhrynidaePhyrnussp.EU520641This study
AraneaeMesothelaeHeptathelidaeHeptathelakimuraiAY309258Qui et al. (2005)
MygalomorphaeTheraphosidaeOrnithoctonushuwenaAY309259Qui et al. (2005)
MygalomorphaeTheraphosidaeAphonopelmasp.EU523754This study
AraneomorphaeHypochilidaeHypochilusthorelliEU523753This study
AraneomorphaeSalticidaeHabronattusoregonensisAY571145Masta and Boore (2004)
OpilionesPalpatoresPhalangiidaePhalangiumopilioEU523757This study
ScorpionesButhidaeButhusoccitanusEU523755This study
ChactidaeaUroctonusmordaxEU523756This study
SolifugaeEremobatidaeEremobatescfr. palpisetulosusEU520642This study
ThelyphonidaThelyphonidaeMastigoproctusgiganteusEU520643This study
XiphosuraXiphosuridaLimulidaeLimuluspolyphemusAF216203Lavrov et al. (2000)
OrderSuborderFamilyGenusSpeciesGenBank NumberReferences
AcariParasitiformesIxodidaeRhipicephalussanguineusAF081829Black and Roehrdanz (1998)
ParasitiformesIxodidaeIxodeshexagonusAF081828Black and Roehrdanz (1998)
ParasitiformesVarroidaeVarroadestructorAJ493124Navajas et al. (2002)
AcariformesTrombiculidaeLeptotrombidiumpallidumAB180098Shao et al. (2005)
AmblypygiPhrynidaePhyrnussp.EU520641This study
AraneaeMesothelaeHeptathelidaeHeptathelakimuraiAY309258Qui et al. (2005)
MygalomorphaeTheraphosidaeOrnithoctonushuwenaAY309259Qui et al. (2005)
MygalomorphaeTheraphosidaeAphonopelmasp.EU523754This study
AraneomorphaeHypochilidaeHypochilusthorelliEU523753This study
AraneomorphaeSalticidaeHabronattusoregonensisAY571145Masta and Boore (2004)
OpilionesPalpatoresPhalangiidaePhalangiumopilioEU523757This study
ScorpionesButhidaeButhusoccitanusEU523755This study
ChactidaeaUroctonusmordaxEU523756This study
SolifugaeEremobatidaeEremobatescfr. palpisetulosusEU520642This study
ThelyphonidaThelyphonidaeMastigoproctusgiganteusEU520643This study
XiphosuraXiphosuridaLimulidaeLimuluspolyphemusAF216203Lavrov et al. (2000)
a

Uroctonus has recently been moved from the superfamily Vaejovoidea, family Vaejovidae, and placed into the superfamily Chactoidea, family Chactidae (Soleglad and Fet 2003). We do not endorse one view over the other, but for convenience in this paper, we use the later placement of Uroctonus in Chactidae.

Total genomic DNA was extracted from one individual of each species using the Qiagen DNAeasy kit. Conserved primers were used to amplify and then sequence the CO1 and Cytb genes (see Boore and Brown 2000). Based on these sequences, taxon-specific primers were designed such that they faced outward from the known sequence (primer sequences available upon request). Using these new primers, mt genomes were amplified in 1 or 2 fragments via long–polymerase chain reaction (PCR). Long-PCR amplification was performed using the Takara LA Taq DNA polymerase kit. Reaction conditions were similar to the manufacturers’ suggestions, with final concentrations of 0.16 mM of each deoxynucleoside triphosphate, 0.4 mM of each primer, 1× Takara polymerase buffer, 1 μl of genomic DNA (concentration not determined), and 2.5 U of Takara polymerase per 100 μl reaction. Reactions were cycled at 92 °C for 30 s, 50–58 °C for 25 s (depending on the primer's Tm), and 68 °C for 12 min, for 37 cycles, followed by a final extension at 72 °C for 15 min.

PCR products were processed for DNA sequencing by the production facility of the DOE Joint Genome Institute. DNA was mechanically sheared into fragments of about 1.5–2 kb, end repaired, and gel purified. These fragments were ligated into pUC18 and transformed into Escherichia coli to create plasmid libraries using standard techniques (Sambrook and Russell 2001). Automated colony pickers were used to select colonies into 384-well plates. For PCR products that spanned the entire mt genome, colonies were picked to fill an entire 384-well plate, whereas with partial genome amplification, half plates were used. After overnight incubation, a small aliquot was processed robotically through amplification of plasmids, sequencing reactions, reaction clean-up, and electrophoretic separation on automated DNA sequencer machines to produce a sequencing read from each end of each plasmid. A full 384-well plate of colonies therefore yielded twice that many sequences, which typically gave about 15–30× coverage for a mt genome, meaning that any given region was sequenced an average of 15–30×. Sequences were quality scored and trimmed using Phred, assembled using Phrap, and electropherograms and assemblies were viewed and verified for accuracy using Sequencher (GeneCodes, Ann Arbor, MI) before consensus sequences were made from all the assembled sequences.

Gene Annotation and Secondary Structure Determination

Protein-coding genes were identified by similarity of inferred amino acid sequences to those of other arthropod mtDNAs. Ribosomal RNA gene locations were inferred based on sequence similarity to other chelicerates and by inferring regions of conserved secondary structures of the small ribosomal RNA and large ribosomal RNA. Exact determination of the 5′ and 3′ ends was not possible, so rRNA gene boundaries were inferred to be located at the 5′ or 3′ ends of the upstream or downstream genes, or we noted that gene boundaries were ambiguous with adjacent tRNA genes.

We used a multipronged approach for finding tRNA genes and inferring their secondary structure. After all protein-coding and ribosomal gene sequences were determined, the remaining regions were searched for tRNAs with the program tRNAscan-SE 1.21 (Lowe and Eddy 1997). For tRNA searches, we allowed approximately 30 nt of overlap with annotated genes. For some taxa, tRNAscan was able to identify all 22 tRNAs. For other taxa, no tRNAs were found using this program's default parameters for mtDNA. In these cases, the search was modified to find secondary structures that had very low Cove scores (<0.1) and to find tRNAs that lacked the T-arm, as in nematode mtDNA. For some taxa, few potential tRNAs were identified, even with these low-stringency searches. In these cases, we searched both DNA strands for tRNAs by comparing unknown regions with alignments of all chelicerate tRNA sequences, 1 alignment for each tRNA. In each of the 22 alignment files, we searched for anticodon arm motifs that were conserved among chelicerate taxa. Gene boundaries were most difficult to determine for the tRNA genes from spiders, due to lack of well-paired aminoacyl acceptor stems. For these tRNAs, we recorded where potential overlap with the adjacent genes encoded on the same strand occurred.

Secondary structures were drawn manually, using both the alignment and the ability to form paired stems as criteria for accepting both the tRNA gene and its structure. We accepted regions with as few as 2 nt per one side of a stem as D- or T-arm stems, provided that there were 3 nt included in the loop of the arm, and a variable loop of a size similar to other arthropods. We used these criteria as a conservative estimate for determining whether a tRNA gene lacked a T-arm, as previous studies of nematode mt tRNAs have revealed that the shortest stems that have been found consist of 3 paired nucleotide, and the smallest loops consist of 3 nt (e.g., Ohtsuki et al. 2002). Therefore, our conservative criteria ensured that we would not overestimate—but might underestimate—the number of tRNAs in arachnids that lack a T-arm.

Character Evolution Analyses

To map the evolution of inferred tRNA secondary structure, we used 2 published phylogenetic trees depicting arachnid relationships. The first tree (fig. 5 in Shultz 2007) is based on morphological characters and is therefore completely independent of the tRNA structural characters used in this study. The second tree (fig. 9 in Wheeler and Hayashi 1998) was also constructed with characters (18S rDNA, 28S rDNA, and morphology) that are independent of those we examined. The Shultz (2007) and Wheeler and Hayashi (1998) trees are generally similar, but differ in the relationships within Tetrapulmonata and Dromopoda.

These 2 trees were simplified by including only the orders represented in our data set. This was done by drawing the trees in the program MacClade 4.0 (Maddison DR and Maddison WP 2000) and then substituting our taxa or GenBank taxa for those used in the previous studies. Both the Shultz (2007) and Wheeler and Hayashi (1998) analyses focused on recovering relationships among the arachnid orders and do not provide detailed data on intraordinal relationships. Therefore, we further modified their trees to include well-accepted intraordinal data for Acari and Araneae, after Coddington et al. (2004) (see their figs. 2 and 3). It is accepted that Parasitiformes and Acariformes are each distinct lineages of mites and ticks, although there is debate as to whether Acari is monophyletic. We followed Coddington et al. (2004) in depicting Acari as a clade. Arachnologists broadly agree that the mesothele spiders are the sister lineage to opisthothele spiders (Mygalomorphae plus Araneomorphae) and that mygalomorph and araneomorph spiders are each monophyletic groups (Coddington et al. 2004). Therefore, we added these findings within Araneae to both the modified Shultz (2007) and Wheeler and Hayashi (1998) phylogenetic trees before mapping tRNA structural characters.

We coded inferred tRNA secondary structures as characters in 2 different ways. In 1 matrix, each of the 22 tRNAs was coded as possessing 1 of 3 character states: cloverleaf structure, lacking T-arm, or lacking D-arm. In another matrix, we encoded only presence or absence of a T-arm for each tRNA because we found that most structural variation is due to T-arm loss. MacClade 4.0 (Maddison DR and Maddison WP 2000) was used to trace all changes on the stylized Shultz (2007) and Wheeler and Hayashi (1998) trees. For each matrix, we analyzed character change on the trees in 2 ways: 1) with T-arm loss as an unordered character, as would be the case if T-arms may be lost and then regained, and 2) with T-arm loss as an irreversible character, as would be the case if a T-arm could not be regained once lost from a lineage.

Phylogenetic Tree Reconstruction

We next used our 2 data matrices of tRNA structural characters to perform parsimony analyses of arachnid relationships. For each matrix, characters were coded as unordered and “gaps” due to loss of particular tRNA genes in some genomes were treated as missing data. Parsimony branch-and-bound searches were performed for both data sets, using the program PAUP* version 4.0 (Swofford 2000). For both data sets, we ran 1000 parsimony branch-and-bound bootstrap replicates, with characters coded as unordered. Trees were rooted with structures from the horseshoe crab, Limulus polyphemus, a member of the sister group to arachnids.

Results and Discussion

Genome-wide Losses of T-arm Sequence

We find mt tRNA gene sequences that can be inferred to code for 4 different types of tRNA structures: 1) those with a typical cloverleaf shape, 2) those that lack a T-arm, 3) those that lack a D-arm, and 4) those that lack sequence that could encode a paired acceptor stem and a T-arm (fig. 1). The many tRNA genes that lacked T-arm sequence had well-conserved D-arm and anticodon arm sequences that could form well-paired stems, indicating that these were not pseudogenes.

Examples of the 4 types of inferred tRNA secondary structures found in arachnids. (A) Cloverleaf structure, as found in tRNAAsn in the amblypygid Phrynus. (B) Structure lacking a D-arm, as found in tRNASer(AGN) in the amblypygid Phrynus. (C) Structure lacking a T-arm, as found in tRNAAsn in the scorpion Buthus. (D) Structure lacking both a T-arm and a well-paired aminoacyl acceptor stem, as found in tRNAAsn in the spider Aphonopelma. Note that the sequence in gray in (D) encodes the downstream gene (in this case, tRNAAla).
FIG. 1.—

Examples of the 4 types of inferred tRNA secondary structures found in arachnids. (A) Cloverleaf structure, as found in tRNAAsn in the amblypygid Phrynus. (B) Structure lacking a D-arm, as found in tRNASer(AGN) in the amblypygid Phrynus. (C) Structure lacking a T-arm, as found in tRNAAsn in the scorpion Buthus. (D) Structure lacking both a T-arm and a well-paired aminoacyl acceptor stem, as found in tRNAAsn in the spider Aphonopelma. Note that the sequence in gray in (D) encodes the downstream gene (in this case, tRNAAla).

Taxa from 4 orders of arachnids (Araneae, Scorpiones, Thelyphonida, and acariform Acari) lack sequences that could encode T-arms from numerous tRNA genes in their mt genomes. In contrast, taxa from other orders (Amblypygi, Opiliones, Solifugae, and parasitiform Acari) possess mt tRNA genes that are typical for metazoans. Inferred structural changes in the tRNAs for each taxon are summarized in table 2, where we indicate how many of the 22 tRNAs show the canonical cloverleaf shape, how many possess a TV-loop in lieu of a T-arm, and how many possess a D-loop in lieu of a D-arm. We classify genomes as possessing “typical” tRNAs if 3 or fewer of their 21 inferred tRNA secondary structures lack T-arms or D-arms (we excluded tRNASer(AGN) because the D-arm is not present in metazoans). Figures of inferred secondary structures for all 174 tRNA genes in the mt genomes from these 8 taxa are presented in supplementary figures S1–S8 (Supplementary Material online).

Table 2

Summary of Inferred tRNA Structures in the 8 New Arachnid mt Genomes We Sequenced and in 8 Previously Published mt Genomes

OrderTaxonCanonicalTV-loopD-loopReferenceNotes
Typical tRNA genes
    XiphosuraLimulus polyphemus2101Lavrov et al. (2000)
    SolifugaeEremobates cfr. palipsetulosus2101This study
    OpilionesPhalangium opilio2001This studyLacks tRNALeu(UUR)
    AmblypygiPhrynus sp.2002This study
    Scorpiones (Chactidae)Uroctonus mordax1831This study
    Acari (Parasitiformes)Ixodes hexagonus2101Black and Roehrdanz (1998)
    Acari (Parasitiformes)Rhipicephalus sanguineus2002Black and Roehrdanz (1998)
    Acari (Parasitiformes)Varroa destructor2002Navajas et al. (2002)
Many tRNAs lack T-arm
    Scorpiones (Buthidae)Buthus occitanus1281This studyLacks tRNAAsp
    ThelyphonidaMastigoproctus giganteus5143This study
    AraneaeHabronattus oregonensis11101Masta and Boore (2004)
    AraneaeHypochilus thorelli4171This study
    AraneaeAphonopelma sp.4153This study
    AraneaeOrnithoctonus huwena2182Qiu et al. (2005)
AraneaeHeptathela hangzhouensis1381Qiu et al. (2005)
    Acari (Acariformes),Leptotrombidium pallidum679Shao et al. (2005)
Many tRNAs lack paired acceptor stems
    AraneaeHabronattus oregonensisMasta and Boore (2004)
    AraneaeHypochilus thorelliThis study
    AraneaeAphonopelma sp.This study
    AraneaeOrnithoctonus huwenaQiu et al. (2005)
OrderTaxonCanonicalTV-loopD-loopReferenceNotes
Typical tRNA genes
    XiphosuraLimulus polyphemus2101Lavrov et al. (2000)
    SolifugaeEremobates cfr. palipsetulosus2101This study
    OpilionesPhalangium opilio2001This studyLacks tRNALeu(UUR)
    AmblypygiPhrynus sp.2002This study
    Scorpiones (Chactidae)Uroctonus mordax1831This study
    Acari (Parasitiformes)Ixodes hexagonus2101Black and Roehrdanz (1998)
    Acari (Parasitiformes)Rhipicephalus sanguineus2002Black and Roehrdanz (1998)
    Acari (Parasitiformes)Varroa destructor2002Navajas et al. (2002)
Many tRNAs lack T-arm
    Scorpiones (Buthidae)Buthus occitanus1281This studyLacks tRNAAsp
    ThelyphonidaMastigoproctus giganteus5143This study
    AraneaeHabronattus oregonensis11101Masta and Boore (2004)
    AraneaeHypochilus thorelli4171This study
    AraneaeAphonopelma sp.4153This study
    AraneaeOrnithoctonus huwena2182Qiu et al. (2005)
AraneaeHeptathela hangzhouensis1381Qiu et al. (2005)
    Acari (Acariformes),Leptotrombidium pallidum679Shao et al. (2005)
Many tRNAs lack paired acceptor stems
    AraneaeHabronattus oregonensisMasta and Boore (2004)
    AraneaeHypochilus thorelliThis study
    AraneaeAphonopelma sp.This study
    AraneaeOrnithoctonus huwenaQiu et al. (2005)
Table 2

Summary of Inferred tRNA Structures in the 8 New Arachnid mt Genomes We Sequenced and in 8 Previously Published mt Genomes

OrderTaxonCanonicalTV-loopD-loopReferenceNotes
Typical tRNA genes
    XiphosuraLimulus polyphemus2101Lavrov et al. (2000)
    SolifugaeEremobates cfr. palipsetulosus2101This study
    OpilionesPhalangium opilio2001This studyLacks tRNALeu(UUR)
    AmblypygiPhrynus sp.2002This study
    Scorpiones (Chactidae)Uroctonus mordax1831This study
    Acari (Parasitiformes)Ixodes hexagonus2101Black and Roehrdanz (1998)
    Acari (Parasitiformes)Rhipicephalus sanguineus2002Black and Roehrdanz (1998)
    Acari (Parasitiformes)Varroa destructor2002Navajas et al. (2002)
Many tRNAs lack T-arm
    Scorpiones (Buthidae)Buthus occitanus1281This studyLacks tRNAAsp
    ThelyphonidaMastigoproctus giganteus5143This study
    AraneaeHabronattus oregonensis11101Masta and Boore (2004)
    AraneaeHypochilus thorelli4171This study
    AraneaeAphonopelma sp.4153This study
    AraneaeOrnithoctonus huwena2182Qiu et al. (2005)
AraneaeHeptathela hangzhouensis1381Qiu et al. (2005)
    Acari (Acariformes),Leptotrombidium pallidum679Shao et al. (2005)
Many tRNAs lack paired acceptor stems
    AraneaeHabronattus oregonensisMasta and Boore (2004)
    AraneaeHypochilus thorelliThis study
    AraneaeAphonopelma sp.This study
    AraneaeOrnithoctonus huwenaQiu et al. (2005)
OrderTaxonCanonicalTV-loopD-loopReferenceNotes
Typical tRNA genes
    XiphosuraLimulus polyphemus2101Lavrov et al. (2000)
    SolifugaeEremobates cfr. palipsetulosus2101This study
    OpilionesPhalangium opilio2001This studyLacks tRNALeu(UUR)
    AmblypygiPhrynus sp.2002This study
    Scorpiones (Chactidae)Uroctonus mordax1831This study
    Acari (Parasitiformes)Ixodes hexagonus2101Black and Roehrdanz (1998)
    Acari (Parasitiformes)Rhipicephalus sanguineus2002Black and Roehrdanz (1998)
    Acari (Parasitiformes)Varroa destructor2002Navajas et al. (2002)
Many tRNAs lack T-arm
    Scorpiones (Buthidae)Buthus occitanus1281This studyLacks tRNAAsp
    ThelyphonidaMastigoproctus giganteus5143This study
    AraneaeHabronattus oregonensis11101Masta and Boore (2004)
    AraneaeHypochilus thorelli4171This study
    AraneaeAphonopelma sp.4153This study
    AraneaeOrnithoctonus huwena2182Qiu et al. (2005)
AraneaeHeptathela hangzhouensis1381Qiu et al. (2005)
    Acari (Acariformes),Leptotrombidium pallidum679Shao et al. (2005)
Many tRNAs lack paired acceptor stems
    AraneaeHabronattus oregonensisMasta and Boore (2004)
    AraneaeHypochilus thorelliThis study
    AraneaeAphonopelma sp.This study
    AraneaeOrnithoctonus huwenaQiu et al. (2005)

In the mt genomes of all 4 opisthothele spiders and the vinegaroon, the majority of tRNA genes lack sequence to encode a T-arm. All scorpions possess some tRNA genes that lack sequence for the T-arm, but there is variation among scorpion families in the number of such tRNA genes. Buthid scorpions possess many truncated mt tRNA genes, whereas the chactid scorpion possesses few tRNA genes that lack T-arm sequences.

Although it is common for metazoan mt genomes to possess one to a few tRNA genes that lack either D-arm or T-arm sequence, it is exceptionally rare to find these losses in many to most of the tRNA genes. However, we find that vinegaroons, acariform mites, buthid scorpions, and all spiders share large-scale losses of T-arm sequence from their mt tRNA genes. The extensive losses shown by these arachnids are not shared by other arthropods.

Because other buthid scorpions have many mt tRNA genes that lack T-arm sequence (Davila et al. 2005; Gantenbein et al. 2005), a tendency for genome-wide loss of T-arms may be a synapomorphy for the scorpion family Buthidae, if not for all scorpions. Likewise, all spiders share the loss of T-arm sequence from 5 particular tRNA genes, and all opisthothele spiders share the loss of T-arm sequence from an additional 4 tRNA genes. Whether all vinegaroons share a loss of T-arm sequence with Mastigoproctus will need to await further sampling. The acariform chigger mite possesses many tRNA genes that lack either T-arm or D-arm sequence (Shao et al. 2005), but insufficient data exist to determine whether this loss is shared with divergent acariform mites. Together, these genome-wide losses suggest that each of these groups have an evolutionary predisposition to lose T-arm sequence from their mt tRNA genes.

How could each of these orders come to lose these sequences from so many tRNA genes? One potential scenario involves a multistep evolutionary process. First, changes may have occurred in the mt ribosome that allowed the loss of arms from tRNAs to be tolerated, as previously suggested (Okimoto and Wolstenholme 1990; Masta 2000). Alternatively, changes could have occurred in the elongation factors (EF-Tus) that bind to the T-arm and carry these tRNAs to the ribosome. Such changes have been found in the EF-Tus in nematodes (Ohtsuki et al. 2001; Arita et al. 2006). Either scenario would lead to what we term a “propensity to lose T-arms.” Following such changes, substitutions could accrue in regions of tRNA genes that code for D- or T-arms, eventually resulting in the loss of paired stems of these arms. Such mutations could accumulate via genetic drift if selection is no longer acting to maintain stem pairs in these regions. Finally, selection may act to remove these regions if there is a selective advantage for small genome size (for instance, if smaller genomes replicate faster). Regions of tRNA genes that formerly coded for stems and loops of D- or T-arms could thereby be deleted from the genome (see figs. 1B and C).

Such a scenario may have played out just once in the history of arachnids or independently in each order that possesses truncated tRNA genes. The timing and order of divergence of arachnid orders is not fully understood. We do know that arachnids are an ancient group, with a fossil record for spiders, amblypygids, acariform mites, and harvestmen from the Devonian, about 376–400 MYA (Shear et al. 1984; Selden et al. 1991; Dunlop et al. 2003, 2004). Moreover, an acariform mite fossil exists from the Ordovician (Bernini et al. 2002), suggesting that Acariformes had diverged much earlier than the Devonian. If the truncated tRNA genes found in the mt genome of the chigger mite are representative of acariform mite mt tRNA genes, it implies that changes allowing truncated tRNAs to maintain function occurred early in the history of arachnids.

Genome-Wide Losses of Paired Acceptor Stem Sequence

All opisthothele spiders sampled possess some tRNA genes that lack sequences to encode a 3′ aminoacyl acceptor stem. Instead, the sequence in this region often shows little potential to form a well-paired stem with the 5′ aminoacyl acceptor stem sequence and frequently encodes another adjacent gene (see fig. 1D; and supplementary figs. S1 and S2, Supplementary Material online). A tRNA that lacks an acceptor stem should not be recognized by aminoacyl tRNA synthetase and therefore should not be able to function in protein synthesis. The gene encoding such a nonfunctioning tRNA would therefore be expected to become a pseudogene and to gradually lose its similarity to a functional gene.

However, this does not appear to be the case with these spider mt tRNA genes as they show conservation of the anticodon and D-arms, both of which possess well-paired stems. We have previously suggested (Masta 2000; Masta and Boore 2004) that a posttranscriptional editing mechanism likely edits spider mt tRNA acceptor stems to enable them to function. The data we present here suggests that if such an editing mechanism exists, it arose very early in the diversification of spiders. Fossil araneomorph (Selden et al. 1999) and mygalomorph (Selden and Gall 1992) spiders exist from the Triassic, suggesting that spiders had diverged into these 2 major lineages at least 250 MYA. Fossil mesothele spiders from the Late Carboniferous (Selden 1996) suggest that opisthothele and mesothele spiders had diverged by almost 300 MYA. Given the long evolutionary history and great species richness of spiders (over 37,000 described species [Platnick 2006]), neither lack of well-paired mt tRNA acceptor stems nor lack of T-arms has seemed to constrain spiders’ evolutionary success.

Phylogenetic Implications of tRNA Structure Mapped onto Arachnid Phylogenies

When we mapped T-arm absence as an unordered character onto phylogenetic trees depicting arachnid relationships, we found that the most parsimonious reconstructions required parallel loss of T-arms in multiple lineages. Figure 2 shows one example of character evolution, using tRNAPro traced over the Shultz (2007) tree and the Wheeler and Hayashi (1998) tree. Reconstructing the evolution of secondary structure for this tRNA required 5 steps on each tree, indicating that T-arm sequences were lost at least 5 separate times in the history of these arachnids. Although we would expect such rare genomic changes to be excellent phylogenetic characters for reconstructing ancient divergences, in this data set, losses of T arm sequences do not always demarcate orders. As an extreme example of homoplasy, for tRNAPro, reconstruction on the trees requires independent evolutionary losses within Araneae, Acari, and Scorpiones.

Reconstruction of the loss of the T-arm (coded as an unordered character) from inferred structure from tRNAPro, traced over the phylogenetic trees of (A) Shultz (2007) and (B) Wheeler and Hayashi (1998). The reconstruction requires 5 steps on each tree; gray shading indicates loss of a T-arm. Bootstrap values for the clades Araneae, Tetrapulmonata, Acari, and Dromopoda found by Shultz (2007) are indicated along the branches of the tree in (A). Bootstrap values were not given by Wheeler and Hayashi (1998).
FIG. 2.—

Reconstruction of the loss of the T-arm (coded as an unordered character) from inferred structure from tRNAPro, traced over the phylogenetic trees of (A) Shultz (2007) and (B) Wheeler and Hayashi (1998). The reconstruction requires 5 steps on each tree; gray shading indicates loss of a T-arm. Bootstrap values for the clades Araneae, Tetrapulmonata, Acari, and Dromopoda found by Shultz (2007) are indicated along the branches of the tree in (A). Bootstrap values were not given by Wheeler and Hayashi (1998).

Ideally, to understand evolutionary patterns by tracing characters onto a tree, one would want a well-resolved tree with well-supported nodes. Support is low for the nodes of the basal relationships among arachnids for these trees, but some higher-level relationships have better support. The Tetrapulmonata (Araneae, Thelyphonida, Schizomida, and Amblypygi) has a bootstrap support value of 93% in the Shultz (2007) tree (see fig. 2A). However, studies have not agreed upon relationships among the orders that comprise Tetrapulmonata. Whereas Shultz (2007) found strong bootstrap support (100%) for Amblypygi being the sister group to Thelyphonida + Schizomida, these relationships were not recovered by Wheeler and Hayashi (1998), Weygoldt and Paulus (1979), or van der Hammen (1985).

Support for the Dromopoda and for the order Acari is low (10% and 33%, respectively) in the Shultz analysis, although these groups were also recovered by Wheeler and Hayashi (1998). Overall, some interordinal relationships are still debated, but aside from Acari, arachnologists broadly agree that each arachnid order is monophyletic, with supporting suites of synapomorphies. For example, in the Shultz (2007) analysis, the spider clade and the scorpion clade each have 100% bootstrap support. Therefore, the fact that we observe a lack of T-arm sequences in only some members of each of these orders suggests that T-arm sequences were lost independently within each order following their diversification.

Phylogenetic Inference Based on tRNA Structural Characters

Most phylogenies of arachnids to date have been based on anatomical characters, some of which show high degrees of homoplasy. This homoplasy may be the result of natural selection acting in parallel on traits that help organisms adapt to particular diets or niches. Ideally, to construct an accurate phylogeny of an ancient group, we want to select characters that evolve very slowly and are not subject to strong natural selection. Structural changes in RNAs are rare evolutionary events, and changes in rRNA structure have been used to help infer phylogenetic relatedness among arthropods (Billoud et al. 2000; Swain and Taylor 2003). Therefore, it seems reasonable that changes in structure in tRNA genes, which are far more conserved than rRNA genes, would be ideal characters to help infer relationships among arachnid orders. In our data set, we have 22 such characters that show varying levels of homoplasy (see next section). Because these could prove useful in inferring arachnid relationships, we used our tRNA structural data matrices to construct phylogenetic trees.

Searches based on the data matrix in which tRNA structure was coded with 3 unordered character states yielded 30 most parsimonious trees. Searches based on the data matrix in which T-arms were coded as present or absent yielded 3 most parsimonious trees. The strict consensus trees from these 2 searches are very similar. The bootstrap majority-rule consensus tree illustrating T-arm presence or absence is shown in figure 3. Bootstrap support values for the presence/absence analyses are shown above the branches, and bootstrap values for the 3 character-state analyses are shown below the branches. These 2 analyses differed in that the presence/absence data showed the 4 opisthothele spiders as a clade with 60% bootstrap support, whereas the 3 character-state analysis excluded Habronattus from this clade. Taxa that possess typical mt tRNA genes form a basal polytomy, as their tRNA structures do not differ significantly from those of Limulus, the outgroup taxon.

The bootstrap majority-rule consensus tree from an analysis based on the tRNA T-arm presence/absence matrix. Bootstrap values are based on 1000 parsimony branch-and-bound replicate searches. Numbers given above the branches are bootstrap values based on the T-arm presence/absence matrix; numbers below the branches are bootstrap values based on the 3 character-state–encoded tRNA structure data matrix. Bootstrap values ≥60% are shown.
FIG. 3.—

The bootstrap majority-rule consensus tree from an analysis based on the tRNA T-arm presence/absence matrix. Bootstrap values are based on 1000 parsimony branch-and-bound replicate searches. Numbers given above the branches are bootstrap values based on the T-arm presence/absence matrix; numbers below the branches are bootstrap values based on the 3 character-state–encoded tRNA structure data matrix. Bootstrap values ≥60% are shown.

Our phylogenetic tree based on tRNA structure places Araneae, Thelyphonida, Scorpiones, and Acariformes together in a well-supported clade. Our data suggest that these taxa share a propensity to lose T-arm sequences throughout their mt genomes. Prior to this, the only group known to show such a propensity for T-arm sequence loss is nematodes. The topology of our consensus trees indicates that the propensity for widespread T-arm sequence loss arose only once. However, our topology differs from traditional hypotheses in several ways.

First, Amblypygi is missing from this clade. Morphological analyses have traditionally placed this order in the Tetrapulmonata with Araneae and Thelyphonida.

Second, scorpions are also in this well-supported clade. Many morphological analyses place scorpions as well diverged from spiders and vinegaroons (Thelyphonida) (van der Hammen 1985; Shultz 1990; Wheeler and Hayashi 1998; Giribet et al. 2002; Shultz 2007). However, some sequence-based analyses have recovered scorpions as sister to spiders (Giribet et al. 2001). In conjunction with our data, these analyses raise the possibility that scorpions are more closely related to spiders and vinegaroons than previously thought. Indeed, scorpions share a morphological feature (possessing 4 lungs) in common with the members of Tetrapulmonata. This trait was assumed to have evolved convergently in scorpions, but our data instead suggest that the possession of 4 lungs could have arisen earlier in the history of arachnids and may unite scorpions with tetrapulmonates. Our data also show relatively high bootstrap support for a grouping that suggests that scorpions may not be monophyletic. However, this seems unlikely given the number of other morphological characters that support monophyly of this order.

Third, our tRNA-structure-based phylogeny indicates that the 2 major lineages of Acari (Acariformes and Parasitiformes) are not closely related. Although some researchers consider it possible that the Acari are not monophyletic, phenotypic characters do not support the placement of acariform mites with spiders, scorpions, and vinegaroons. Acariformes are not tetrapulmonates but instead respire across their cuticle. Thus, it seems unlikely that our tree depicts the actual placement of the 2 major acarid lineages.

Our tree differs from traditional morphological trees in some respects. The tRNA sequence loss provides an informative phylogenetic character for the reasons stated above. However, the extent and nature of differences between our tree and previous trees suggests that tRNA characters have likely undergone some degree of parallel evolution. For instance, we recover a clade of opisthothele spiders that arachnologists widely accept but do not find strong support for the monophyly of scorpions. This suggests that tRNA structural characters may display homoplasy only in certain groups. Taken as a whole, our data indicate parallel evolution, both within and among arachnid orders. To pinpoint exactly how much homoplasy exists, we will need further sampling across this species-rich group.

Variation Among tRNA Genes in T-Arm Sequence Loss

When we map T-arm absence as an unordered character onto arachnid phylogenies, we find that for 17 of the 22 mt tRNAs, loss of the T-arm sequence requires at least 2 steps (fig. 4), indicating parallel evolution. This is true when characters are traced over either the Shultz (2007) tree or the Wheeler and Hayashi (1998) tree. Coding the characters as irreversible changes the number of steps only for tRNAHis, which increases from 5 to 6 steps over each tree.

The number of tRNAs showing each of several numbers of steps (T-arm losses) when mapped over the Shultz (2007) phylogenetic tree of chelicerates. Light gray shading indicates that the characters were reconstructed without homoplasy; dark gray shading indicates that loss of the T-arm required 2 or more steps. One-letter amino acid abbreviations are used to indicate tRNA identity, with numerals used to designate the anticodons: L1 = CUN, L2 = UUR, S1 = AGN, and S2 = UCN.
FIG. 4.—

The number of tRNAs showing each of several numbers of steps (T-arm losses) when mapped over the Shultz (2007) phylogenetic tree of chelicerates. Light gray shading indicates that the characters were reconstructed without homoplasy; dark gray shading indicates that loss of the T-arm required 2 or more steps. One-letter amino acid abbreviations are used to indicate tRNA identity, with numerals used to designate the anticodons: L1 = CUN, L2 = UUR, S1 = AGN, and S2 = UCN.

We previously inferred (fig. 3 in Masta and Boore 2004) the presence of a T-arm in multiple tRNAs (tRNATyr, tRNALeu(UUR), tRNAAsn, tRNAHis, and tRNAVal) in the jumping spider Habronattus, even though these showed weak 2-bp stems, unusually small variable loops, or unclear 3′ gene boundaries. In light of subsequent data from additional spider genomes that more clearly lack T-arm sequences, our earlier judgments appear overly conservative, and it now seems likely that Habronattus also lacks these sequences. If we reinterpret these Habronattus tRNAs as instead possessing TV-loops, then all 4 opisthothele spiders are fixed for TV-loops in these 5 tRNAs as well as in the 9 tRNAs discussed above. Recoding Habronattus decreases the number of inferred steps for these tRNA structures when they are reconstructed on the phylogenetic trees and increases to 14 the number of synapomorphies for opisthothele spiders. A parsimony search run with this data set yields trees with the same consensus topology as for the other data sets but with increased (87%) bootstrap support for the clade of opisthothele spiders. However, the reconstruction of T-arm loss from these tRNAs still requires 2 or more steps on the Shultz (2007) and Wheeler and Hayashi (1998) trees, indicating that losses evolved in parallel.

Our data indicate that tRNA genes vary in how frequently they have lost their T-arm sequences during the diversification of arachnids (fig. 4, table 3). Only the structure of tRNASer(AGN) has remained unchanged, consistent with a loss of the D-arm early in the evolution of metazoans. Loss of the T-arm from tRNAArg, tRNALys, and tRNAMet occurred only once, and each loss is a synapomorphy for opisthothele spiders. The T-arm was lost from tRNA Ser(UCN) only in the mesothele spider Heptathela. All other losses of T-arms are inferred to have occurred multiple times during the evolution of arachnids. The genes coding for tRNAPro, tRNAHis, tRNAAla, tRNACys, tRNAGln, and tRNATyr have experienced the greatest number of T-arm losses in the arachnids we surveyed. We examined whether the amino acids that these are charged with had any similar characteristics, but we saw no pattern with regard to their size or polarity. Therefore, it seems unlikely that tRNAs arm loss is associated with certain characteristics of amino acids.

Table 3

The tRNA Structure Coded as Presence or Absence of the T-Arm for the Chelicerates Sampled in This Study

TaxaARNDCEQGHIL1L2KMFPS1S2TWYV
Araneae Habronattus0101100100101111000100
Araneae Hypochilus1111110111111110000111
Araneae Aphonopelma0111100111111110000111
Araneae Ornithoctonus1111111111111111000111
Araneae Heptathela1001100110000010010100
Amblypygi Phrynus0000000000000000000000
Thelyphonida Mastigoproctus1001111110100011001111
Scorpiones Buthus001101110000001001010
Scorpiones Uroctonus0000100110000000000000
Opiliones Phalangium000000000000000000000
Solifugae Eremobates0000000000000000000000
Parasitiformes Ixodes0000000000000000000000
Parasitiformes Rhipicephalus0000000000000000000000
Parasitiformes Varroa0000000000000000000000
Acariformes Leptotrombidium0000101010010011001100
Xiphosura Limulus0000000000000000000000
TaxaARNDCEQGHIL1L2KMFPS1S2TWYV
Araneae Habronattus0101100100101111000100
Araneae Hypochilus1111110111111110000111
Araneae Aphonopelma0111100111111110000111
Araneae Ornithoctonus1111111111111111000111
Araneae Heptathela1001100110000010010100
Amblypygi Phrynus0000000000000000000000
Thelyphonida Mastigoproctus1001111110100011001111
Scorpiones Buthus001101110000001001010
Scorpiones Uroctonus0000100110000000000000
Opiliones Phalangium000000000000000000000
Solifugae Eremobates0000000000000000000000
Parasitiformes Ixodes0000000000000000000000
Parasitiformes Rhipicephalus0000000000000000000000
Parasitiformes Varroa0000000000000000000000
Acariformes Leptotrombidium0000101010010011001100
Xiphosura Limulus0000000000000000000000

NOTE.—The tRNA identity is abbreviated with the standard one-letter code, where L1 = CUN, L2 = UUR, S1 = AGN, and S2 = UCN anticodons. 0 indicates that a T-arm is present. 1 indicates that a T-arm is absent. An em-dash indicates that the tRNA gene was not present in the genome.

Table 3

The tRNA Structure Coded as Presence or Absence of the T-Arm for the Chelicerates Sampled in This Study

TaxaARNDCEQGHIL1L2KMFPS1S2TWYV
Araneae Habronattus0101100100101111000100
Araneae Hypochilus1111110111111110000111
Araneae Aphonopelma0111100111111110000111
Araneae Ornithoctonus1111111111111111000111
Araneae Heptathela1001100110000010010100
Amblypygi Phrynus0000000000000000000000
Thelyphonida Mastigoproctus1001111110100011001111
Scorpiones Buthus001101110000001001010
Scorpiones Uroctonus0000100110000000000000
Opiliones Phalangium000000000000000000000
Solifugae Eremobates0000000000000000000000
Parasitiformes Ixodes0000000000000000000000
Parasitiformes Rhipicephalus0000000000000000000000
Parasitiformes Varroa0000000000000000000000
Acariformes Leptotrombidium0000101010010011001100
Xiphosura Limulus0000000000000000000000
TaxaARNDCEQGHIL1L2KMFPS1S2TWYV
Araneae Habronattus0101100100101111000100
Araneae Hypochilus1111110111111110000111
Araneae Aphonopelma0111100111111110000111
Araneae Ornithoctonus1111111111111111000111
Araneae Heptathela1001100110000010010100
Amblypygi Phrynus0000000000000000000000
Thelyphonida Mastigoproctus1001111110100011001111
Scorpiones Buthus001101110000001001010
Scorpiones Uroctonus0000100110000000000000
Opiliones Phalangium000000000000000000000
Solifugae Eremobates0000000000000000000000
Parasitiformes Ixodes0000000000000000000000
Parasitiformes Rhipicephalus0000000000000000000000
Parasitiformes Varroa0000000000000000000000
Acariformes Leptotrombidium0000101010010011001100
Xiphosura Limulus0000000000000000000000

NOTE.—The tRNA identity is abbreviated with the standard one-letter code, where L1 = CUN, L2 = UUR, S1 = AGN, and S2 = UCN anticodons. 0 indicates that a T-arm is present. 1 indicates that a T-arm is absent. An em-dash indicates that the tRNA gene was not present in the genome.

Losses versus Gains of tRNA Arms

Once a T-arm is lost, it would seem unlikely ever to be regained. If a tRNA can function without a T-arm, then little selective pressure would exist to create a new one de novo. We examined this by tracing the evolution of tRNA secondary structure on the phylogenetic trees shown in figure 2. When characters were coded as unordered, there was only one instance of an apparent gain of a T-arm (in tRNAHis), which appeared to be lost early in the evolution of spiders and then regained in Habronattus. However, once we recoded the inferred secondary structure of tRNAHis in H. oregonensis (because it yielded a more plausible structure in light of our new data; see above), all reconstructed changes were losses. Therefore, our data is consistent with the irreversible loss of T-arms from tRNA genes.

Hypotheses to Explain Patterns of tRNA Evolution

Given the otherwise extreme conservation of tRNA structure across all of life, we are led to ask why independent reductions in mt tRNA sequences have occurred within arachnids but apparently not in other groups of arthropods or widely in metazoans. One hypothesis is that the common ancestor of arachnids underwent some modifications that allowed these rare changes to occur. Such a modification could involve relaxation of the structural interactions with the mt ribosome with which these tRNAs interact during protein synthesis or changes in the elongation factor genes that carry the tRNAs to the ribosome.

One general hypothesis to explain mt tRNA sequence reduction is that we expect to find a gradual accumulation of deleterious mutations, given that metazoan mitochondria are typically unable to undergo recombination. Indeed, metazoan mt tRNA genes appear to have accumulated deleterious mutations compared with their nuclear counterparts (Lynch 1996). However, the accumulation of deleterious changes alone is not sufficient to explain why arachnids appear more prone to lose arms from their mt tRNAs than other metazoans. Instead, it seems likely that arachnids must first have evolved a compensatory mechanism to allow truncated tRNAs to function. Such a compensatory mechanism would either permit the tRNAs to function in a truncated state or allow posttranscriptional editing to repair losses of arms. In either case, what appear to be deleterious mutations at the DNA level may not actually exert effects on the fitness of the mt tRNA.

The mt tRNAs from nematodes lack either the T-arm or D-arm yet appear able to function. Analyses suggest that the tertiary L structure required for interaction of tRNA with the translation system can be achieved when either the D-arm or T-arm is absent (Watanabe et al. 1994; Wolstenholme et al. 1994; Ohtsuki et al. 1996). A study of the mt tRNAPhe (which lacks the T-arm) of the nematode Caenorhabditis elegans showed that its tertiary structure differs from the canonical tRNA tertiary structure, in that its aminoacyl acceptor stem and anticodon interstem angles are wider than in canonical tRNAs (Frazer-Abel and Hagerman 2004). Steinberg and Cedergren (1994) predicted that the interstem angle between the anticodon and aminoacyl acceptor stem is less important for correct interaction of tRNA with the ribosome than the distance between the anticodon and the end of the tRNA. Our findings of mt tRNAs in spiders that are so truncated as to appear to end at the anticodon stem raises intriguing questions as to how these spider tRNAs can function.

The highly unusual changes in arachnid tRNA genes that we document appear to have arisen multiple times de novo. If true, then we must infer that there are unique evolutionary pressures present in arachnids that can cause parallel evolution in secondary structure among taxa in the same helices of the tRNA genes. With no knowledge of the nuclear genome or of what compensatory role nuclear genes may potentially confer on mt functioning, we can only speculate on what such pressures may have been. We do know that the lineages that possess these rare molecular changes diversified millions of years ago and that they include some of the most species-rich groups of metazoans. Clearly, given the right molecular context, organisms can tolerate major structural changes in their tRNA genes.

Most specimens for this study were provided by colleagues: Marshal Hedin (Hypochilus, Mastigoproctus, and Phrynus), Darrell Ubick (Phalangium), Benjamin Gantenbein (Buthus), and Warren Savary (Uroctonus and Eremobates). Jennifer Kuehl assisted with library creation and clone selection, and Amy Clawser and Sarah Martha assisted with figures of secondary structure. This work was funded by National Science Foundation award number DEB-0416628 to S.E.M. and J.L.B. We thank Franz Lang and 2 anonymous reviewers for suggestions that improved the manuscript.

References

Arita
M
Suematsu
T
Osanai
A
Inaba
T
Kamiya
H
Kita
K
Sisido
M
Watanabe
Y-i
Ohtsuki
T
,
An evolutionary ‘intermediate state’ of mitochondrial translation systems found in Trichinella species of parasitic nematodes: co-evolution of tRNA and EF-Tu
Nucleic Acids Res
,
2006
, vol.
34
(pg.
5291
-
5299
)
Bernini
F
Carnevale
G
Bagnoli
G
Stouge
S
Bernini
F
Nannelli
R
Nuzzaci
G
de Lillo
E
,
An early Ordovician oribatid mite (Acari: Oribatida) from the island of Öland, Sweden
Acarid phylogeny and evolution: adaptations in mites and ticks
,
2002
The Hague
Kluwer
(pg.
45
-
48
)
Billoud
B
Guerrucci
M-A
Masselot
M
Deutsch
JS
,
Cirripede phylogeny using a novel approach: molecular morphometrics
Mol Biol Evol
,
2000
, vol.
17
(pg.
1435
-
1445
)
Black
RL
4th
Roehrdanz
WC
,
Mitochondrial gene order is not conserved in arthropods: prostriate and metastriate tick mitochondrial genomes
Mol Biol Evol
,
1998
, vol.
15
(pg.
1772
-
1785
)
Boore
JL
Brown
WM
,
Mitochondrial genomes of Galathealinum, Helobdella, and Platynereis: sequence and gene arrangement comparisons indicate that Pogonophora is not a phylum and Annelida and Arthropoda are not sister taxa
Mol Biol Evol
,
2000
, vol.
17
(pg.
87
-
106
)
Coddington
JA
Giribet
G
Harvey
MS
Prendini
L
Walter
DE
Cracraft
J
Donoghue
MJ
,
Arachnida
Assembling the tree of life
,
2004
New York
Oxford University Press
(pg.
296
-
318
)
Davila
S
Pinero
D
Bustos
P
Cevallos
MA
Davila
G
,
The mitochondrial genome sequence of the scorpion Centruroides limpidus (Karsch 1879) (Chelicerata; Arachnida)
Gene
,
2005
, vol.
360
(pg.
92
-
102
)
Dunlop
JA
Anderson
LI
Kerp
H
Hass
H
,
Preserved organs of Devonian harvestmen
Nature
,
2003
, vol.
425
pg.
916
Dunlop
JA
Anderson
LI
Kerp
H
Hass
H
,
A harvestman (Arachnida: Opiliones) from the Early Devonian Rhynie cherts, Aberdeenshire, Scotland
Trans R Soc Edinb Earth Sci
,
2004
, vol.
94
(pg.
341
-
354
)
Frazer-Abel
AA
Hagerman
PJ
,
Variation of the acceptor-anticodon interstem angles among mitochondrial and non-mitochondrial tRNAs
J Mol Biol
,
2004
, vol.
343
(pg.
313
-
325
)
Gantenbein
B
Fet
V
Gantenbein-Rittter
IA
Balloux
F
,
Evidence for recombination in scorpion mitochondrial DNA (Scorpiones: Buthidae)
Proc R Soc Lond B Biol Sci
,
2005
, vol.
272
(pg.
697
-
704
)
Giribet
G
Edgecombe
GD
Wheeler
WC
,
Arthropod phylogeny based on eight molecular loci and morphology
Nature
,
2001
, vol.
413
(pg.
157
-
161
)
Giribet
G
Edgecombe
GD
Wheeler
WC
Babbitt
C
,
Phylogeny and systematic position of Opiliones: a combined analysis of chelicerate relationships using morphological and molecular data
Cladistics
,
2002
, vol.
18
(pg.
5
-
70
)
Hatzoglou
E
Rodakis
GC
Lecanidou
R
,
Complete sequence and gene organization of the mitochondrial genome of the land snail Albinaria coerulea
Genetics
,
1995
, vol.
140
(pg.
1353
-
1366
)
He
Y
Jones
J
Armstrong
M
Lamberti
F
Moens
M
,
The mitochondrial genome of Xiphinema americanum sensu stricto (Nematoda: Enoplea): considerable economization in the length and structural features of encoded genes
J Mol Evol
,
2005
, vol.
61
(pg.
819
-
833
)
Helfenbein
KG
Brown
WM
Boore
JL
,
The complete mitochondrial genome of the articulate brachiopod Terebratalia transversa
Mol Biol Evol
,
2001
, vol.
18
(pg.
1734
-
1744
)
Keddie
EM
Higazi
T
Unnasch
TR
,
The mitochondrial genome of Onchocerca volvulus: sequence, structure and phylogenetic analysis
Mol Biochem Parasitol
,
1998
, vol.
95
(pg.
111
-
127
)
Kurabayashi
A
Ueshima
R
,
Complete sequence of the mitochondrial DNA of the primitive opistobranch gastropod Pupa strigosa: systematic implication of the genome organization
Mol Biol Evol
,
2000
, vol.
17
(pg.
266
-
277
)
Lavrov
DV
Boore
JL
Brown
WM
,
The complete mitochondrial DNA sequence of the horseshoe crab Limulus polyphemus
Mol Biol Evol
,
2000
, vol.
17
(pg.
813
-
824
)
Lavrov
DV
Brown
WM
,
Trichinella spiralis mtDNA: a nematode mitochondrial genome that encodes a putative ATP8 and normally structured tRNAs and has a gene arrangement relatable to those of coelomate metazoans
Genetics
,
2001
, vol.
157
(pg.
621
-
637
)
Lavrov
DV
Brown
WM
Boore
JL
,
A novel type of RNA editing occurs in the mitochondrial tRNAs of the centipede Lithobius forficatus
Proc Natl Acad Sci USA
,
2000
, vol.
97
(pg.
13738
-
13742
)
Lowe
TM
Eddy
SR
,
tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence
Nucleic Acids Res
,
1997
, vol.
25
(pg.
955
-
964
)
Lynch
M
,
Mutation accumulation in transfer RNAs: molecular evidence for Muller's ratchet in mitochondrial genomes
Mol Biol Evol
,
1996
, vol.
13
(pg.
209
-
220
)
Maddison
DR
Maddison
WP
MacClade 4: analysis of phylogeny and character evolution
,
2000
Sunderland (MA)
Sinauer Associates, Inc
Masta
SE
,
Mitochondrial sequence evolution in spiders: intraspecific variation in tRNAs lacking the TΨC arm
Mol Biol Evol
,
2000
, vol.
17
(pg.
1091
-
1100
)
Masta
SE
Boore
JL
,
The complete mitochondrial genome sequence of the spider Habronattus oregonensis reveals rearranged and extremely truncated tRNAs
Mol Biol Evol
,
2004
, vol.
21
(pg.
893
-
902
)
Montiel
R
Lucena
MA
Medeiros
J
Simoes
N
,
The complete mitochondrial genome of the entomopathogenic nematode Steinernema carpocapsae: insights into nematode mitochondrial DNA evolution and phylogeny
J Mol Evol
,
2006
, vol.
62
(pg.
211
-
U215
)
Navajas
M
Le Conte
Y
Solignac
M
Cros-Arteil
S
Cornuet
JM
,
The complete sequence of the mitochondrial genome of the honeybee ectoparasite mite Varroa destructor (Acari: Mesostigmata)
Mol Biol Evol
,
2002
, vol.
19
(pg.
2313
-
2317
)
Ohtsuki
T
Kawai
G
Watanabe
K
,
The minimal tRNA: unique structure of Ascaris suum mitochondrial tRNASerUCU having a short T arm and lacking the entire D arm
FEBS Lett
,
2002
, vol.
514
(pg.
37
-
43
)
Ohtsuki
T
Kawai
G
Watanabe
Y-i
Kita
K
Nishikawa
K
Watanabe
K
,
Preparation of biologically active Ascaris suum mitochondrial tRNAMet with a TV-replacement loop by ligation of chemically synthesized RNA fragments
Nucleic Acids Res
,
1996
, vol.
24
(pg.
662
-
667
)
Ohtsuki
T
Watanabe
Y-i
Takemoto
C
Kawai
G
Ueda
T
Kita
K
Somei
K
Kaziro
Y
Nyborg
J
Watanabe
K
,
An “elongated” translation elongation factor Tu for truncated tRNAs in nematode mitochondria
J Biol Chem
,
2001
, vol.
276
(pg.
21571
-
21577
)
Okimoto
R
Macfarlane
JL
Clary
DO
Wolstenholme
DR
,
The mitochondrial genomes of two nematodes, Caenorhabditis elegans and Ascaris suum
Genetics
,
1992
, vol.
130
(pg.
471
-
498
)
Okimoto
R
Wolstenholme
DR
,
A set of tRNAs that lack either the TψC arm or the dihydrouridine arm: towards a minimal tRNA adaptor
EMBO J
,
1990
, vol.
9
(pg.
3405
-
3411
)
Platnick
NI
The world spider catalog, version 7.0
,
2006
American Museum of Natural History
 
Qiu
Y
Song
D
Zhou
K
Sun
H
,
The mitochondrial sequences of Heptathela hangzhouensis and Ornithoctonus huwena reveal unique gene arrangements and atypical tRNAs
J Mol Evol
,
2005
, vol.
60
(pg.
57
-
71
)
Rokas
A
Holland
PWH
,
Rare genomic changes as a tool for phylogenetics
Trends Ecol Evol
,
2000
, vol.
15
(pg.
454
-
459
)
Sambrook
J
Russell
DW
,
Molecular cloning: a laboratory manual
,
2001
Cold Spring Harbor (NY)
Cold Springs Harbor Laboratory Press
Selden
PA
,
Fossil mesothele spiders
Nature
,
1996
, vol.
379
(pg.
498
-
499
)
Selden
PA
Anderson
JM
Anderson
HM
Fraser
NC
,
Fossil araneomorph spiders from the Triassic of South Africa and Virginia
J Arachnol
,
1999
, vol.
27
(pg.
401
-
414
)
Selden
PA
Gall
JC
,
A Triassic mygalomorph spider from the northern Vosges, France
Palaeontology
,
1992
, vol.
35
(pg.
211
-
235
)
Selden
PA
Shear
WA
Bonamo
PM
,
A spider and other arachnids from the Devonian of New York, and reinterpretations of Devonian Araneae
Palaeontology
,
1991
, vol.
34
(pg.
241
-
281
)
Shao
R
Mitani
H
Barker
SC
Takahashi
M
Fukunaga
M
,
Novel mitochondrial gene content and gene arrangement indicate illegitimate inter-mtDNA recombination in the chigger mite, Leptotrombidium pallidum
J Mol Evol
,
2005
, vol.
60
(pg.
764
-
773
)
Shao
RF
Barker
SC
Mitani
H
Takahashi
M
Fukunaga
M
,
Molecular mechanisms for the variation of mitochondrial gene content and gene arrangement among chigger mites of the genus Leptotrombidium (Acari: Acariformes)
J Mol Evol
,
2006
, vol.
63
(pg.
251
-
261
)
Shear
WA
Bonamo
PM
Grierson
JD
Rolfe
WDI
Smith
EL
Norton
RA
,
Early land animals in North America: evidence from Devonian age arthropods from Gilboa, New York
Science
,
1984
, vol.
224
(pg.
492
-
494
)
Shultz
JW
,
Evolutionary morphology and phylogeny of arachnida
Cladistics
,
1990
, vol.
6
(pg.
1
-
38
)
Shultz
JW
,
A phylogenetic analysis of the arachnid orders based on morphological characters
Zool J Linn Soc
,
2007
, vol.
150
(pg.
221
-
265
)
Soleglad
ME
Fet
V
,
High-level systematics and phylogeny of the extant scorpions (Scorpiones: Orthosterni)
Euscorpius
,
2003
, vol.
11
(pg.
1
-
175
)
Steinberg
S
Cedergren
RJ
,
Structural compensation in atypical mitochondrial tRNAs
Nat Struct Biol
,
1994
, vol.
1
(pg.
507
-
510
)
Swain
TD
Taylor
DJ
,
Structural rRNA characters support monophyly of raptorial limbs and paraphyly of limb specialization in water fleas
Proc R Soc B Biol Sci
,
2003
, vol.
270
(pg.
887
-
896
)
Swofford
DL
PAUP*: phylogenetic analysis using parsimony (*and other methods). Version 4
,
2000
Sunderland (MA)
Sinauer Associates
Terrett
JA
Miles
S
Thomas
RH
,
Complete DNA sequence of the mitochondrial genome of Cepaea nemoralis (Gastropoda: pulmonata)
J Mol Evol
,
1996
, vol.
42
(pg.
160
-
168
)
van der Hammen
L
,
Functional morphology and affinities of extant Chelicerata in evolutionary perspective
Trans R Soc Edinb
,
1985
, vol.
76
(pg.
137
-
146
)
Waeschenbach
A
Telford
MJ
Porter
JS
Littlewood
DTJ
,
The complete mitochondrial genome of Flustrellidra hispida and the phylogenetic position of Bryozoa among the Metazoa
Mol Phylogenet Evol
,
2006
, vol.
40
(pg.
195
-
207
)
Watanabe
Y-i
Tsurui
H
Ueda
T
Furushima
R
Takamiya
S
Kita
K
Nishikawa
K
Watanabe
K
,
Primary and higher order structures of nematode (Ascaris suum) mitochondrial tRNAs lacking either the T or D stem
J Biol Chem
,
1994
, vol.
269
(pg.
22902
-
22906
)
Weygoldt
P
Paulus
HF
,
Unteruschungen zur morphologie, taxonomie und phylogenie der Chelicerata. II. Cladogramme und die enfaltung der Chelicerata
Z Zool Syst Evol
,
1979
, vol.
17
(pg.
117
-
200
)
Wheeler
WC
Hayashi
CY
,
The phylogeny of the extant chelicerate orders
Cladistics
,
1998
, vol.
14
(pg.
173
-
192
)
Wolstenholme
DR
Wolstenholme
DR
Jeon
KW
,
Animal mitochondrial DNA: structure and evolution
Mitochondrial genomes
,
1992
New York
Academic Press
(pg.
173
-
216
)
Wolstenholme
DR
Macfarlane
JL
Okimoto
R
Clary
DO
Wahleithner
JA
,
Bizarre tRNAs inferred from DNA sequences of mitochondrial genomes of nematode worms
Proc Natl Acad Sci USA
,
1987
, vol.
84
(pg.
1324
-
1328
)
Wolstenholme
DR
Okimoto
R
Macfarlane
JL
,
Nucleotide correlations that suggest tertiary interactions in the TV-replacement loop-containing mitochondrial transfer-RNAs of the nematodes, Caenorhabditis elegans and Ascaris suum
Nucleic Acids Res
,
1994
, vol.
22
(pg.
4300
-
4306
)
Yokobori
S-i
Oshima
T
Wada
H
,
Complete nucleotide sequence of the mitochondrial genome of Doliolum nationalis with implications for evolution of urochordates
Mol Phylogenet Evol
,
2005
, vol.
34
(pg.
273
-
283
)

Author notes

Franz Lang, Associate Editor

Supplementary data