The complete chloroplast genome sequences of eight Orostachys species: Comparative analysis and assessment of phylogenetic relationships

We analyzed the complete chloroplast genomes of eight Orostachys species and compared the sequences to those of published chloroplast genomes of the congeneric and closely related genera, Meterostachys and Hylotelephium. The total chloroplast genome length of thirteen species, including the eight species analyzed in this study and the five species analyzed in previous studies, ranged from 149,860 (M. sikokianus) to 151,707 bp (H. verticillatum). The overall GC contents of the genomes were almost identical (37.6 to 37.8%). The thirteen chloroplast genomes each contained 113 unique genes comprising 79 protein-coding genes, 30 tRNA genes, and four rRNA genes. Among the annotated genes, sixteen genes contained one or two introns. Although the genome structures of all Orostachys and Hylotelephium species were identical, Meterostachys differed in structure due to a relatively large gene block (trnS-GCU-trnS-GGA) inversion. The nucleotide diversity among the subsect. Orostachys chloroplast genomes was extremely low in all regions, and among the subsect. Appendiculatae, genus Orostachys, and all thirteen chloroplast genomes showed high values of Pi (>0.03) in one, five, or three regions. The phylogenetic analysis showed that Orostachys formed polyphyly, and subsect. Orostachys and Appendiculatae were clustered with Hylotelephium and Meterostachys, respectively, supporting the conclusion that each subsection should be considered as an independent genus. Furthermore, the data supported the taxonomic position of O. margaritifolia and O. iwarenge f. magnus, which were treated as synonyms for O. iwarenge in a previous study, as independent taxa. Our results suggested that O. ramosa and O. japonica f. polycephala were individual variations of O. malacophylla and O. japonica, respectively. The exact taxonomic position of O. latielliptica and the phylogenetic relationship among the three species, O. chongsunensis, O. malacophylla and O. ramosa, should be a topic of future study.


Introduction
The family Crassulaceae DC., belonging to Rosales Bercht. & J.Presl, includes succulent herbaceous plants. Approximately 1500 species in 35 genera are known and they are mainly distributed throughout the Northern Hemisphere [1,2].
Among these, the genus Orostachys Fisch. includes approximately 20 to 25 taxa that are distributed from the Ural Mountains to Japan [3][4][5]. This genus has traditionally been used as an ornamental and a medicinal plant, and some species have recently been shown to be effective in antioxidant and anticancer treatments, thus becoming recognized as a very important plant resource [6,7].
The taxa belonging this genus are mostly biennial herbaceous plants that are usually succulent. The morphological characteristics of this genus are as follows: the roots are fibrous and there is no rhizome. The leaves are linear to ovate, often with dull purple dots; the apex is usually cuspidate with a white and cartilaginous appendage that is softly obtuse or acuminate. In the first year, the leaves stand together in solitary, basal, and dense rosettes. The flowering stem arises from the center of the rosette in the second year. The inflorescence is a dense raceme or thyrse, narrowly pyramidal to cylindrical, and contains many flowers and foliagelike bracts. The flowers are bisexual, subsessile or pedicellate, and pentamerous. The sepals are usually shorter than the petals. The petals are subconnate at the base and are white, yellowgreen, or pinkish to reddish [8,9].
Due to differences in the growth habits among closely related taxa, the genus Orostachys was first described as a genus independent of Coyledon L. by Fischer [10]. However, Steudel [11] and de Candolle [12] proposed the classification of Orostachys within the genera Sedum and Umbilicus DC., respectively. Since then, various studies [1,13,14] have been conducted to support the proposal to treat Orostachys as an independent genus; as a result, it is currently recognized as a genus.
The genus comprised two sections, Orostachys Ohba and Schoenlandia Ohba, that differ with respect to the shape of their leaves, number of stamens, and type of inflorescence [15]. Later sect. Schoenlandia was classified in a new genus, Kungia K.T.Fu [16]. Currently, therefore, the genus Orostachys has only one section, Orostachys, and it is split into two subsections, Appendiculatae (Boriss.) Ohba and Orostachys (Boriss.) Ohba, depending on the absence or presence of an appendage at the leaf apex [9,17].
Recently molecular phylogenetic studies [18][19][20][21] revealed a large phylogenetic distance between the two subsections of Orostachys; subsect. Orostachys was in the Hylotelephium Ohba clade, and subsect. Appendiculatae formed a clade with Meterostachys Nakai. Therefore, more research is necessary to clarify the proper classification of Orostachys. The external morphology characteristics among species of Orostachys are very similar. Within each taxon, there is wide variation in external morphological characteristics. For these reasons, the classification of these plants is known to be very difficult. Although several taxonomic studies [5,6,[18][19][20][21][22] have been conducted, the phylogenetic relationships among the species and the taxonomic position of many taxa remain unclear.
We obtained the whole chloroplast genome sequences of eight  Ohba. The main goal of this study was to evaluate the phylogenomic relationships among the two subsections of Orostachys and its closely related genera. To address the fact that taxonomists differ in their opinions regarding the classification some species, it was our goal to clarify the taxonomic position of several taxa.

Materials and methods
Taxon sampling, DNA extraction, sequencing, assembly and annotation The eight Orostachys taxa examined in this study were not classified as endangered or protected. We did not collect materials from any privately owned or protected areas requiring permission for collection. The plant materials for this study were collected from the native habitats of each taxon, and the voucher specimens were deposited in the Sangji University Herbarium (SJUH) (S1 Table). Total DNA was extracted from approximately 100 mg of fresh leaves using a DNeasy Plant Mini Kit (Qiagen Inc., Valencia, CA, USA), and sequenced using the Illumina MiSeq and NovaSeq 6000 platforms (Illumina Inc., San Diego, CA, USA) at Labgenomics (Seongnam, Korea). The DNA of the Orostachys taxa was sequenced to produce 385,646-21,445,885 raw reads with lengths of 301 bp and 150 bp (S1 Table). Low-quality sequences (Phred score < 20) were trimmed using CLC Genomics Workbench (version 6.04; CLC Inc., Arhus, Denmark). Then, reads were assembled using a Geneious assembler with a medium sensitivity option via Geneious Prime v.2022.1.1 (Biomatters Ltd., Auckland, New Zealand). The draft genome contigs were merged into a single contig by joining the overlapping terminal sequences of each contig. The protein-coding genes, transfer RNAs (tRNAs), and ribosomal RNAs (rRNAs) in the chloroplast genome were predicted and annotated using Geneious Prime v.2022.1.1 and manually edited by comparison with the published chloroplast genome sequences of Orostachys. The tRNAs were confirmed using tRNAscan-SE [23]. A circular chloroplast genome map was drawn using the OGDRAW program [24].

Comparative genome analyses in Orostachys and allied genera
The newly complete chloroplast genome sequences of eight Orostachys taxa were used along with the following chloroplast genome sequences from GenBank of NCBI for comparative analysis: two published Orostachys, O. japonica (accession no. MN794320) and O. minuta (accession no. OK094425) sequences, one Meterostachys, M. sikokianus (Makino) Nakai (accession no. MZ365442), and two Hylotelephium, H. erythrostictum (accession no. MZ519882) and H. verticillatum (accession no. MT558730) sequences.
The program mVISTA was used to compare similarities among the thirteen species using the shuffle-LAGAN mode [25]. The annotated O. malacophylla chloroplast genome was used as a reference. Additionally, the genome structure of the thirteen species were compared using the MAUVE program [26]. The large single copy/inverted repeat (LSC/IR) and inverted repeat/small single copy (IR/SSC) boundaries of these species were also compared and analyzed.

Nucleotide diversity analysis
To assess the nucleotide diversity (Pi) among the thirteen chloroplast genomes, including ten Orostachys, one Meterostachys and two Hylotelephium, the complete chloroplast genome sequences were aligned using the MAFFT [27] aligner tool and manually adjusted with BioEdit [28]. We then performed sliding window analysis to calculate the nucleotide variability (Pi) values using DnaSP 6 [29] with a window length of 600 bp and a step size of 200 bp [30].

Chloroplast genome features of Orostachys and related genera
The chloroplast genomes of eight new Orostachys species have been submitted to GenBank of the National Center for Biotechnology Information (NCBI) ( Table 1). The total length of the chloroplast genomes of the thirteen species, i.e., the eight species analyzed in this study and the species analyzed in previous studies ( 1). All thirteen cp genomes exhibited the typical quadripartite structure, consisting of a pair of IR regions (25,285-25,854 bp), separated by an LSC region (82,293-83,070 bp), and an SSC region (16,839-17,018 bp). Their overall GC contents were almost identical (37.6-37.8%). The chloroplast genomes of the thirteen species contained 113 unique genes comprising 79 protein-coding genes, 30 tRNA genes, and four rRNA genes (Table 1). Among the annotated genes, fourteen genes (atpF, ndhA, ndhB, petB, petD, rpl2, rps12, rpl16, rpoC1, trnA-UGC, trnI-GAU, trnK-UUU, trnL-UAA, and trnV-UAC) contained one intron, and two genes (clpP and ycf3) contained two introns.
The pairwise cp genomic alignment among the thirteen species (ten Orostachys, one Meterostachys and two Hylotelephium species) were all similar, with the exception of that of Meterostachys. The LSC and SSC regions were more variable than the IR regions (Fig 2). In the chloroplast genome of Meterostachys, the sequence similarity was very low in the relatively large gene block (trnS-GCU-trnS-GGA, approximately 37,000 bp) of the LSC region, which was confirmed to be caused by an inversion (S1 Fig). A comparison of the LSC/IR and IR/SSC boundaries in the thirteen species is shown in

Nucleotide diversity among the Orostachys and related genera
The average nucleotide diversity (Pi) among the five subsect.  (Fig 4). Meanwhile, the Pi values of rbcL and matK, which are corebarcode regions, were very low, 0.005 and 0.021, respectively.

Phylogenetic analyses of the Telephium clade of Crassulaceae
The two ML trees constructed based on the two data sets, whole cp genome sequences and 79 protein-coding genes, were well supported at the genus level, except for those of Orostachys.

PLOS ONE
The complete chloroplast genome sequences of eight Orostachys species The two ML trees were divided into two subclades. The first clade consisted of Rhodiola L. and Phedimus Raf., and both genera were well supported monophyly.

Comparison of the chloroplast genomes of Orostachys and related genera
Many recent studies have been carried out to clarify the taxonomy of related taxa using complete cp genome sequences. The cp genome is known to be highly conserved in most land plants, but structural changes in chloroplast genomes, such as gene duplication and deletion and inversion due to occasional rearrangements, provide important taxonomic data [34][35][36][37][38][39][40][41][42]. This study showed that the genome structures of Orostachys and Hylotelephium species were identical, and the sequence identities were also very similar among species in most of the chloroplast regions (Fig 2 and S1 Fig). The LSC/IRs/SSC boundaries were also very similar except for a very slight difference in sequence length (Fig 3). Therefore, these results indicated that the chloroplast genomes of Orostachys and Hylotelephium were very conservative, and that the two genera were closely related. However, the cp genome of Meterostachys was different in structure from those of the other taxa in this study due to the inversion of a relatively large gene block (trnS-GCU-trnS-GGA); this characteristic supports the classification of this group as unique genus.

Phylogenetic relationships of Orostachys and related genera
In many phylogenetic studies [18][19][20][21], the genus Orostachys, first described by Fischer [10], was confirmed to be not monophyletic, and the two subsections, subsect. Orostachys and Appendiculatae, showed close relationships with Hylotelephium and Meterostachys, respectively. These results cast doubt on the classification of Orostachys to a single genus. In our study, we found that the genus was not monophyletic; we confirmed that each subsection was monophyletic and closely related to the two genera mentioned above (Fig 5). The two subsections of Orostachys were clearly distinguished morphologically by the absence (subsect. Orostachys) (Fig 6A) or presence (subsect. Appediculatae) (Fig 6B) of appendage such as a thorn at the leaf apex [9,17]. Two genera, which were closely clustered with each subsection, were identified that shared this characteristic.
Meterostachys, which formed a clade with subsect. Appendiculatae, is a monotypic genus containing only one species, M. sikokianus. It was initially described as Cotyledon sikokianus Makino [43]. Later, it was segregated into a new genus [44], but a relatively recent phylogenetic study [45] suggested that it should be classified in the genus Orostachys because it was situated within a subsect. Appendiculatae clade and formed a clade with Orostachys thyrsiflora Fisch. Meterostachys is clearly distinguished from Orostachys by inflorescence morphology (thyrsoid-paniculate to paniculate in Orostachys and bracteates cymose in Meterostachys) (Fig 6C and 6D) [15]. We found that Meterostachys had a unique chloroplast genome structure and formed an independent clade in this study (Figs 2 and 5). Therefore, we think that the taxonomic position of Meterostachys as an independent genus is supported. Based on these results, we strongly agree with Gontcharova et al. [19] that subsect. Appediculatae should be recognized as a distinct genus.
Meanwhile, the phylogenetic relationships between subsect. Orostachys and Hylotelephium were not clear in previous studies [18][19][20][21]. The phylogenetic relationships examined in this study were very clear (Fig 5), but only two Hylotelephium species were assessed. Further studies including more diverse taxa are needed.

Phylogenetic relationships and taxonomic position of Orostachys species
The phylogenetic relationships and taxonomic position of many taxa in Orostachys remain contested. Species, such as O. iwarenge f. magnus, O. ramosa, O. latielliptica, O. chongsunensis, and O. margaritifolia, first described by Lee and Lee [8,46], are particularly problematic in terms of their taxonomic position because Ohba [17]  In all ML trees obtained in this study (Fig 5), O. margaritifolia (Fig 6E), which was treated as synonym for O. iwarenge (Fig 6F) by Ohba [17], was not clustered within the subsect.
Orostachys clade containing O. iwarenge, but in subsect. Appeidiculatae, characterized by the presence of an appendage at the leaf apex. This species was clearly distinguished from other taxa of this subsection due to morphological characteristics such as obovate leaves and purple anthers. Additionally, O. iwarenge f. magnus (Fig 6G), which has also been treated as a synonym for O. iwarenge (Fig 6F), was independently clustered, although the topology was different in each tree in this study. This species is distributed only on Ulleung-do Island, an oceanic island in Korea, and is reproductively completely isolated. Additionally, it is morphologically very similar to O. iwarenge, except for differences in leaf shape (oval in O. iwarenge f. magnus and oblong to spatulate in O. iwarenge) and stamen color (orange in O. iwarenge f. magnus, and yellow in O. iwarenge). Therefore, we strongly agree with Kim and Park [6] that these two species should be treated as independent taxa.
O. latielliptica (Fig 6H) was described as a new species [8] because it has appendages at the leaf apex, as well as glaucous ovate leaves, and one to four aggregated flowers on its pedicel. The latter two characteristics are unique and distinguish them from all other species. Therefore, the taxonomic position of this species as an independent taxon is supported based on these characteristics. However, the exact phylogenetic relationship could not be confirmed because the topology of this species was different in the two ML trees of this study (Fig 5). Furthermore, this species showed a very close relationship to O. japonica, which was treated as synonym [17], in the ML tree based on 79 protein coding gene sequences in this study. To investigate the exact taxonomic position of this species, further in-depth studies are necessary.
O. chongsunensis (Fig 6I and 6J), which was treated as a synonym for O. japonica by Ohba [17], was clustered within the subsect. Orostachys clade. This result was due to the absence of appendages at the leaf apex, indicating that O. chongsunensis is not the same species as O. japonica. In the ML trees in this study, O. chongsunensis showed a close relationship with O. ramosa ( Fig 5A) and O. malacophylla (Fig 5B). Morphologically, this species was very similar to O. malacophylla except for purplish variegate leaves. Additionally, O. malacophylla was very similar to O. ramosa (Fig 6K) except that it does not branch at the base of the stem. In our experience, branching at the base of the stem is an occasional mutation in O. malacophylla ( Fig 6N). Therefore, we concluded that the three species mentioned above are the same species; the purplish variegate leaves of O. chongsunensis are thought to be due to the growth environment and, specifically, factors such as exposure to limestone and the light intensity (Fig 6I and  6J). However, the exact phylogenetic relationship of O. ramosa could not be confirmed in this study because the topology in the two ML trees was different (Fig 5).
Meanwhile, since individuals with branching at the base of the stem in O. japonica are also relatively common in the natural population of O. japonica (Fig 6M), we concluded that O. japonica f. polycephala (Fig 6L) is an individual variation of O. japonica. Additionally, O. japonica f. polycephala showed the closest relationship to O. japonica in all ML trees in this study (Fig 5), which well supported our opinion.

Conclusion
In this study, we assembled the chloroplast genomes of eight Orostachys, which had a total length ranging from 150,464 bp to 151,462 bp. The cp genomes of Orostachys and Hylotelephium had identical structures and were highly conserved. However, the structure of Meterostachys was different due to the relatively large gene block (trnS-GCU-trnS-GGA) inversion, which is considered important information supporting its the taxonomic position as an independent genus. The results of phylogenetic analyses suggested that the two subsections of Orostachys, subsect. Orostachys and Appendiculatae, were independent genera. In addition, the results supported the taxonomic