Complete mitochondrial genomes of four deep-sea echinoids: conserved mitogenome organization and new insights into the phylogeny and evolution of Echinoidea

Introduction

The sea urchin (Echinoidea) is one of the most iconic lineages of marine animals, representing a numerically relatively small class (slightly over 1,000) of marine invertebrates (Kroh & Mooi, 2022). The small number of extant echinoid species is opposed by its outstanding fossil record, with more than 10,000 nominal fossil species (Kier & Lawson, 1978; Kroh, 2010). Echinoids are an important component in benthic marine environments, and they occur at all depths from the shallow-water hard substrates to abyssal depths (Emlet, 2002; Furman & Heck, 2009; Tuya et al., 2004; Schultz, 2015). All extant 15 echinoid orders, except Stomopneustoida, have deep-sea representatives including 350 species (Glover, Higgs & Horton, 2022). The sea urchins have a long independent evolutionary history, which split from their closest relatives (the sea cucumbers) at Ordovician, more than 450 million years ago (Mya) (Smith, 1984; Smith & Savill, 2001; Thompson et al., 2017).

Researchers have attempted to explore the phylogeny of Echinoidea based on morphological and molecular data (summarized in Fig. 1) (Smith et al., 2006; Kroh & Smith, 2010; Mongiardino Koch et al., 2018; Lin et al., 2020; Mongiardino Koch & Thompson, 2020). Phylogenetic incongruences between morphological and molecular data have been found, which are mainly concentrated on the positions of orders Echinothurioida and Clypeasteroida. The morphological evidences placed Echinothurioida as the sister clade to all other Euechinoidea (Smith et al., 2006; Kroh & Smith, 2010). The molecular evidences from transcriptomes, however, revealed a derived position of Echinothurioida, nested in a group that contains orders Diadematoida and Pedinoida (Mongiardino Koch et al., 2018; Mongiardino Koch & Thompson, 2020). The phylogenetic placement of Clypeasteroida and Echinolampadoida is another important issue in echinoid phylogeny. The suborders Clypeasterina and Scutellina were once placed in the order Clypeasteroida based on the morphological characters (Smith et al., 2006; Kroh & Smith, 2010). Based on the morphological evidences, Clypeasteroida was a monophyletic sister to Cassiduloida (Kroh & Smith, 2010) or Cassiduloida/Echinolampadoida lineages (Smith et al., 2006). Mongiardino Koch et al. (2018) modified Scutellina and Echinolampadoida to suborder level and placed them in the order Echinolampadacea based on the transcriptomic data. In their study, Scutellina and Clypeasterina were renamed to Scutelloida and Clypeasteroida, respectively. The transcriptomic study further suggested a close relationship between Clypeasteroida and Echinolampadacea (Mongiardino Koch et al., 2018; Mongiardino Koch & Thompson, 2020). The recent phylomitogenomic analyses failed to sample the taxa necessary to test these hypothesis (Lin et al., 2020). Therefore, the origin and evolution research of sea urchin is now limited by the topological uncertainty.

Few attempts have been made to address the relationship between the shallow-water and deep-sea lineages of Echinoids (Smith & Stockley, 2005). Deep sea (greater than 200 m) occupies about 66% of the bottom of global ocean, representing the largest and most remote biome of the world (Woolley et al., 2016). It has special environment factors, such as low temperatures, low oxygen level, high hydrostatic pressure, limited food and constant darkness (Naganoa & Nagahama, 2012). Deep sea plays an important part in generating the current patterns of marine biodiversity (Carney, 2005; Brown & Thatje, 2014). Explaining the origin of deep-sea organisms is still a central priority for marine biogeographic research (McClain & Hardy, 2010). Evidences now suggest that most faunal groups may have originated in shallow water (Strugnell et al., 2008; Raupach et al., 2009; Yang et al., 2015; Sun, Sha & Wang, 2019). Despite this, it is assumed that evolutionary innovations in stylasterid corals are generated in deep sea (Lindner, Cairns & Cunningham, 2008). The evolution of echinoderms seems more complicated. The phylogenomic analysis of ophiuroids suggested that most of the oldest lineages showed a deep-sea origin, however, the multiple colonization events indicated that the evolution of brittle stars were neither a simple onshore-offshore pattern, nor the opposite hypothesis (Thuy, 2013; Bribiesca-Contreras et al., 2017). The phylomitogenetic relationship of the sea cucumbers showed that the deep-sea species formed the basal clades (Sun, Sha & Xiao, 2021). Thus, the role of the deep sea in biogeographic processes cannot be ignored.

The mitochondrial genome (mitogenome) is characterized by several advantages, such as maternal inheritance, small genome size, fast evolutionary rates and low recombination (Boore, 1999; Curole & Kocher, 1999). Furthermore, the mitogenome can provide more phylogenetic signals than single mtDNA markers, therefore, it is widely used for phylogenetic and evolutionary analysis at different taxonomic level of the Echinoderma (Perseke et al., 2010; Reich et al., 2015; Láruson, 2017; Bronstein & Kroh, 2019; Galaska et al., 2019). Prior to our study, only thirty-nine mitogenomes from Echinoidea are available, which distributed within seven orders (Cidaroida, Diadematoida, Arbacioida, Camarodonta, Stomopneustoida, Clypeasteroida, Spatangoida), with nine orders still no representatives. Compared to their diversity and abundance, the mitogenomes available are limited in the class Echinoidea, and only four mitogenomes of deep-sea species have been reported (Kober & Bernardi, 2013; Direct submission for MW354512). Inadequate taxon sampling and taxon biases can lead to topological distortions owing to the artefactual sources of error (Timm & Bracken-Grissom, 2015). Thus, improving the representation of mitogenomes for Echinoidea is indispensable for understanding the phylogenetic and evolution of sea urchin.

We newly sequenced four complete mitogenomes of the deep-sea sea urchins (Aspidodiadema arcitum Mortensen, 1939, Caenopedina pulchella (Agassiz & Clark, 1907), Phormosoma bursarium Agassiz, 1881, Araeosoma owstoni Mortensen, 1904, which were the first representatives of the orders Aspidodiadematoida, Pedinoida and Echinothurioida, respectively. In this study, we aim to improve our understanding of Echinoidea phylogeny and evolution of deep-sea sea urchins by: (1) comparing the organization and composition of deep-sea echinoid mitogenomes with those of shallow water ones; (2) conducting a phylogenetic investigation on Echinoidea with expanded mitogenome data and taxon sampling; (3) exploring the macro-evolutionary processes of deep-sea and shallow water echinoids.

Materials and Methods

Results

Mitogenome organization

The sequencing output data for the four deep-sea sea urchins were summarized in Table 1. The mitogenomes of the four echinoids were all circularized from the reads, with 15,704 (A. owstoni) to 15,870 bp (A. arcitum) in length. The average coverage were 26–104x for these four mitogenomes. These new mitogenomes contained 13 protein-coding genes (PCGs), 2 ribosomal RNA genes (rRNAs), 22 transfer RNA genes (tRNAs) and a control region (CR) (Fig. 2, Table S4). Among the 37 genes, six (nad6, tRNA-Ser^UCN, tRNA-Gln, tRNA-Ala, tRNA-Val, tRNA-Asp) located on the heavy-strand (H-strand), with the remaining ones on the light-strand (L-strand). The gene arrangement of the four new mitogenomes were identical to that of other echinoids.

All the tRNA genes can be folded into canonical clover-leaf secondary structures. Unlike most of the echinoids, the tRNA-Ser^AGY gene showed complete DHU arm (hereafter D-arm) and loop (hereafter D-loop). In order to understand the evolution of tRNA-Ser^AGY gene among the echinoids, we mapped the structures of tRNA-Ser^AGY on the phylogenetic tree of echinoids constructed in this study (Fig. 3). Under this framework, we found the normal cloverleaf pattern was the most basal and was maintained in only several lineages, while the D-arm lost in most species.

The 12S and 16S rRNA genes exhibited similar size among the four new genomes (895–906 bp and 1,415–1,520 bp, respectively). Compared with the other regions, the control regions (CRs) of the four mitogenomes exhibited more variation in length, ranging from 101 bp in A. owstoni to 281 bp in A. arcitum. Obviously, the length variation among them were congruent with the differences of their mitogenome sizes. The echinoid CRs harbored a poly-G stretch, which participated in the formation of the stem-loop structures (Fig. 2).

Base composition and codon usage

The overall A + T content of the four new mitogenomes ranged from 59.94% in C. pulchella to 65.68% in P. bursarium (Fig. 4A, Table S5). The values of the AT-skews were mostly positive, indicating more As than Ts. And the GC-skews for all echinoid mitogenomes were positive, suggesting Gs were more abundant than Cs (Fig. 4B, Table S5). tRNA and rRNA showed the highest (60.60% and 60.58%, respectively) A + T content, while the CR showed the lowest (53.48%) A + T content. The nucleotide composition analysis of the PCGs suggested that nad2, nad4l and nad6 possessed the highest A + T content (more than 60%), while that of the atp8 gene was the lowest. Furthermore, the third codon position of the PCGs had higher A + T content (64.25%) than that of the first and second codon positions (54.19% and 60.03%, respectively) (Fig. 4C, Table S5).

Significantly, among the reported echinoid mitogenomes, the deep-sea species P.bursarium showed the highest A + T content in mtDNA, PCG, tRNA, and rRNA. In order to investigate if the base composition between the deep-sea and shallow-water species was different, the A + T content at complete mtDNA, PCG, tRNA, rRNA genes and CR were compared. However, statistical t-tests proved that only the difference of A + T content in tRNA between deep-sea and shallow water echinoids was significant (p = 0.047).

In the four newly sequenced mitogenomes, almost all PCGs (except atp8) started with the typical ATN. The typical stop codon TAA was used more frequently than TAG, while the incomplete stop codons T (nad4 in P. bursarium and A. owstoni) and TA (cytb in P. bursarium) were also used as the stop codon. The codon usage analysis of the proteins (Fig. 4D, Table S5) indicated that Leu₁, Phe, Met, Ser₂, Thr and Ala were the six most common amino acids, accounting for about a half of the total. The relative synonymous codon usage (RSCU) further revealed a base composition bias in echinoid mitogenomes. Usually, the RSCU values of NNU and NNA were higher than 1 (Table S5).

Phylogenetic relationships of Echinoidea

Saturation tests were performed for each of the PCG genes, the dataset PCG1 (the first codon position), PCG2 (the second position) and PCG3 (the third position). Substitution saturation was detected in atp8 gene as reflected from the linear correlation of the number of transitions (s) and transversions (v) plotted against the F84 genetic distance, as well as from a significantly higher value of I_ss in comparison to I_ss.c (Fig. 5, Table S6, Fig. S1). The dataset PCG3 exhibited some saturation (Fig. S1), although I_ss was significantly lower than I_ss.c (Table S6).

Maximum Likelihood (ML) and Bayesian inference (BI) trees based on the nucleotide sequences (NT) (Fig. 5) and amino acid sequences (AA) (Fig. 6) datasets were mostly congruent. The tree topologies resulting from the BI and ML analysis were identical. The phylogenetic trees supported the monophyly of all the echinoid orders except Spatangoida. Cystechinus giganteus (Holasteroida) appeared as subgroup within, rather than a sister group to Spatangoida. In both topologies, Echinoidea was divided into four groups (Clades A–D in Fig. 5). The order Cidaroida (Clade D) was placed as a basal sister group to the clades of the remaining echinoids. Arbacioida and the irregular sea urchins (Spatangoida + Holasteroida + Echinolampadacea + Clypeasteroida) clustered together (Clade A). Camarodonta showed the closest relationship with Stomopneustoida (Clade B), and then clustered together with Clade A. Aspidodiadematoida, Pedinoida, Echinothurioida and Diadematoida formed a sister clade (Clade C), which was supported as the sister clade to Clade A + Clade B.

Habitat origin of the echinoids

The ancestral habitat reconstruction provides new insights into the origin of echinoids (Fig. 7). Our result did not support a consistent offshore to onshore or onshore to offshore pattern across the echinoid phylogeny, but multiple transitions between deep sea and shallow water lineages supporting a mixed origin. Although the deep-sea echinoid species were slightly underrepresented in this study, they were recovered as older lineages, indicating that the deep-sea habitats might be an important site for origination.

Discussion

In this study, we successfully sequenced and analyzed four complete mitochondrial genomes of deep-sea Echinoidea species. They are the first representatives of the orders Aspidodiadematoida, Pedinoida and Echinothurioida. Compared with the lengths of other previously sequenced echinoid sequences (15,650 bp in Strongylocentrotus purpuratus to 16,358 bp in Pygmaeocidaris prionigera), the sequence of A. arcitum (15,870 bp) obtained here is the second largest one. The minor length differences of all the PCGs (<20 bp, and cox2, cox3, cytb and nad4l are of the exact same length) indicates that the PCGs have relatively conserved features. The location and length of tRNAs are highly conserved. Compared with PCGs and tRNAs, the length of rRNAs show relatively high variations (2,402 bp in Temnopleurus reevesii to 2,504 bp in Stylocidaris reini). The differences in mitogenome size of echinoids are mainly attributed by the variations of the CR (101 bp in A. owstoni to 677 bp in P. prionigera).

No gene rearrangements are found in Echinoidea. Indeed, the gene order and orientation of the mitogenomes of the echinoids are identical to the echinoderm ground pattern (Shen et al., 2009). Among the Echinodermata, the genome architecture of Echinoidea is most similar to that of the vertebrate basal genome differing only in the transpositions of the 16S rRNA and the nad4l gene (Boore, 1999).

Among the 22 putative tRNA secondary structures in metazoans, tRNA-Ser^AGY usually lacks a DHU stem and therefore forms a simple loop structure (Lavrov, Brown & Boore, 2000; Ruiz-Pesini & Wallace, 2006). In contrast, the cloverleaf structure of tRNA-Ser^AGY has been found in ten echinoids, and most of the species with the structure locate in the basal lineages. We speculate that the tRNA-Ser^AGY with normal cloverleaf pattern represents an ancestral state. The derived D-loop structures of tRNA-Ser^AGY occurs in younger lineages. Previous studies have suggested that tRNA isomerisms really exist and are functional (Sakurai et al., 2006), and however whether the sporadically observed tRNA isomerism is related to adaptive processes awaits further investigation.

The calculation of A + T content in the echinoid mitogenomes reveals a A + T biased nucleotide composition. Among the partitions of mitogenome, the CR shows the lowest A + T content. This is contrast to the widely accepted phenomenon that the CR (also called AT-rich region) contains the highest AT content of the complete mitogenome (Zhang, Szymura & Hewitt, 1995). Several studies have found that in other echinoderms, e.g., holothuroids (Sun, Sha & Xiao, 2021) and asteroids (Sun, Xiao & Sha, 2022), the A + T content of mtDNA, PCG, tRNA and and rRNA of deep-sea groups are significantly higher than those of the shallow-water species. However, this phenomenon has not been detected in the echinoids.

The RSCU values of echinoid mitogenomes indicate a strong A+T-bias in the third codon position of PCGs. This result is also consistent with the hypothesis that the codon usage bias may be positively correlated with the A + T-bias of the third codon position in mitogenomes (Salvato et al., 2008; Kim et al., 2009; Chai, Du & Zhai, 2012; Hao et al., 2012). However, unlike holothuroids (Sun, Sha & Xiao, 2021) and asteroids (Sun, Xiao & Sha, 2022), no difference has been found in the codon usage between the deep-sea and shallow water groups of echinoid.

In this study, a total of 12 representative orders of the Echinoidea have been selected for phylogenetic analysis to understand the genetic relationship among each order. The topology based on mitogenomes in this study is inconsistent with previous morphological and molecular studies (Smith et al., 2006; Kroh & Smith, 2010; Mongiardino Koch et al., 2018; Lin et al., 2020; Mongiardino Koch & Thompson, 2020). The closest relationship of Holasteroida and Spatangoida has been proved with morphological and transcriptomic dataset (Smith et al., 2006; Kroh & Smith, 2010; Mongiardino Koch et al., 2018; Mongiardino Koch & Thompson, 2020). However, the monophyly of Spatangoida is not supported in our result, with the lineage of Holasteroida nested as a subgroup, instead of the sister group of Spatangoida. Recent transcriptomic studies have highlighted the incongruence between molecular trees and morphological trees related to the position of Scutelloida and Clypeasteroida (Mongiardino Koch et al., 2018; Mongiardino Koch & Thompson, 2020). Our mitogenomic phylogenetic work confirms the closest relationship between Scutelloida and Echinolampadoida, and supports the hypothesis of Mongiardino Koch et al. (2018) that Scutelloida and Echinolampadoida should be placed in order Echinolampadacea. The closest relationship of Arbacioida and Stomopneustoida has been recovered by the morphological and molecular evidences (Smith et al., 2006; Kroh & Smith, 2010; Lin et al., 2020; Mongiardino Koch & Thompson, 2020). Our investigation expands the representatives of Arbacioida and Stomopneustoida. We sheds new light on the position of Arbacioida, which indicates a novel sister relationship between Arbacioida and the irregular sea urchins. Another noteworthy result of our analyses is that we have first positioned Aspidodiadematoida in the phylogeny of Echinoidea using mitogenomic data. The relationship (((Aspidodiadematoida + Pedinoida) + Echinothurioida) + Diadematoida) discredits the hypothesis based on morphological evidences (Smith et al., 2006; Kroh & Smith, 2010). Inconsistency between molecular and morphological data has been found in many extant echinoderms (Janies, 2001; Miller et al., 2017). The incongruences demonstrate that further taxonomic revision is required for these groups.

There have been two contrasting hypothesises about the origin of deep-sea organisms. Some researchers have regarded them to be ancient (Wilson, 1999), while otherssupport their recent origin (Jacobs & Lindberg, 1998). The explanation of ancient origin has been evidenced by the discovery of relict species from the deep sea (Gage & Tyler, 1999), indicating that some modern deep-sea groups experienced situ diversification over evolutionary time (Lindner, Cairns & Cunningham, 2008). Furthermore, the geological history has recorded the extreme changes in oxygen, temperature and circulation, which may result in mass extinction of the deep-sea fauna (Jacobs & Lindberg, 1998; McClain & Hardy, 2010) followed by reinvasions of shallow-water organisms, e.g., molluscans and crustaceans (Strugnell et al., 2008; Raupach et al., 2009; Sun, Sha & Wang, 2019). It seems that increasing evidences point to the important role of the shallow water in macro-evolutionary processes of marine faunas and that of the deep sea has been discounted or ignored. Our data does not show a consistent offshore to onshore or an opposite pattern across the echinoid phylogeny, as there have been multiple transitions back and forth between deep-sea and shallow-water lineages. However, the older lineages are suggested to have inhabited deep sea, indicating the important role of the deep sea in the diversification of Echinoidea. It has been reported that the aspidodiadematoids and echinothurioids sea urchins first appeared in Jurassic shallow-water environments, then disappeared from the fossil record, and survived in deep-sea until today (Smith & Stockley, 2005). In addition, the similar evolutionary pattern has also been found in other echinoderms, e.g., holothuroids (Sun, Sha & Xiao, 2021), asteroids (Sun, Xiao & Sha, 2022) and ophiuroids (Bribiesca-Contreras et al., 2017). Perhaps this is not surprising. Previous studies have demonstrated that oceanic anoxic events did not have a strong effect on echinoderm deep-sea diversity (Smith & Stockley, 2005; Thuy et al., 2012). Diversification of primary producers and increasing nitrification of the deep sea environment may increase the diversity of deep-sea sea urchin (Smith, 2013) so they can escape from deep-sea hypoxia events and persist in the deep sea up to date. Based on this hypothesis, we can interpret why older echinoid clades have a deep-sea origin.

Conclusion

In this study, we contributed four complete mitogenomes of deep-sea sea urchins from three orders, enhancing the taxonomic coverage of Echinoidea mitogenomic data. We did the first comparative analysis of mitogenome base composition, phylogenetic relationships and evolutionary history in Echinoidea. The genome architecture of Echinoidea was highly conserved. We identified the cloverleaf structure of tRNA-Ser^AGY, previously considered missing DHU arm, and speculated that tRNA-Ser^AGY with DHU arm may represent the ancestral status. The base composition and codon usage showed no difference between deep-sea and shallow water groups. The phylogenetic relationships based on PCGs supported the monophyly of all the echinoid orders except Spatangoida. Cidaroida was placed as a basal sister group to all the others. We shed new light on the positions of Arbacioida and Aspidodiadematoida. Although the echinoid phylogeny showed no consistent offshore-onshore or onshore-offshore pattern, the deep-sea lineages positioned in the older clades revealed the important role of the deep sea in the diversification of Echinoidea.

Supplemental Information