Recent plastid genome (plastome) studies of legumes (family Fabaceae) have shown that this family has undergone multiple atypical plastome evolutions from each of the major clades. The tribe Desmodieae belongs to the Phaseoloids, an important but systematically puzzling clade within Fabaceae. In this study, we investigated the plastome evolution of Desmodieae and analyzed its phylogenetic signaling. We sequenced six complete plastomes from representative members of Desmodieae and from its putative sister Phaseoloid genus Mucuna. Those genomes contain 128 genes and range in size from 148,450 to 153,826 bp. Analyses of gene and intron content revealed similar characters among the members of Desmodieae and Mucuna. However, there were also several distinct characters identified. The loss of the rpl2 intron was a feature shared between Desmodieae and Mucuna, whereas the loss of the rps12 intron was specific to Desmodieae. Likewise, gene loss of rps16 was observed in Mucuna but not in Desmodieae. Substantial sequence variation of ycf4 was detected from all the sequenced plastomes, but pseudogenization was restricted to the genus Desmodium. Comparative analysis of gene order revealed a distinct plastome conformation of Desmodieae compared with other Phaseoloid legumes, i.e., an inversion of an approximately 1.5-kb gene cluster (trnD-GUC, trnY-GUA, and trnE-UUC). The inversion breakpoint suggests that this event was mediated by the recombination of an 11-bp repeat motif. A phylogenetic analysis based on the plastome-scale data set found the tribe Desmodieae is a highly supported monophyletic group nested within the paraphyletic Phaseoleae, as has been found in previous phylogenetic studies. Two subtribes (Desmodiinae and Lespedezinae) of Desmodieae were also supported as monophyletic groups. Within the subtribe Lespedezinae, Lespedeza is closer to Kummerowia than Campylotropis.
Citation: Jin D-P, Choi I-S, Choi B-H (2019) Plastid genome evolution in tribe Desmodieae (Fabaceae: Papilionoideae). PLoS ONE 14(6): e0218743. https://doi.org/10.1371/journal.pone.0218743
Editor: William Oki Wong, Indiana University Bloomington, UNITED STATES
Received: January 22, 2019; Accepted: June 8, 2019; Published: June 24, 2019
Copyright: © 2019 Jin et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: We deposited our seven complete plastome sequences in the GenBank database (https://www.ncbi.nlm.nih.gov/) (accession number: MG867566–MG867572).
Funding: This study was supported by a scientific research fund from the Korea National Arboretum (KNA1-2-13, 14-2). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Fabaceae (Leguminosae), the third largest family within the angiosperms, includes 770 genera and over 19,500 species . Three well-known subfamilies–Caesalpinioideae, Mimosoideae, and Papilionoideae–have recently been reclassified into six subfamilies–Caesalpinioideae, Cercidoideae, Detarioideae, Dialioideae, Duparquetioideae, and Papilionoideae . The Phaseoloid clade is one lineage within Papilionoideae, which comprises the Phaseoleae sensu lato (s.l.) clade, Desmodieae, and Psoraleeae . Many species belonging to this clade are utilized as a source for food, ornamental foliage, and medicine. This clade displays a complex phylogenetic relationship among and within tribes [1, 4]. Tribe Desmodieae and Psoraleeae are monophyletic groups that are nested within the paraphyletic Phaseoleae s.l. group [1, 4] (Fig 1).
The tribe Desmodieae comprises 32 genera . Although these plants mainly grow in tropical and warm-temperate regions, some members are found in the cool-temperate and boreal regions of North America [5–6]. Species of this tribe usually occur in the form of herbs or shrubs, and rarely as trees . Their fruits are either loments (a loment consists of a single carpel that disarticulates into single-seeded segments when ripe) or legumes (fruits composed of a single article) . This tribe has traditionally been split into three subtribes: Bryinae, Desmodiinae, and Lespedezinae . Among these three, Bryinae has now been placed in Dalbergieae s.l. as a result of molecular phylogenetics studies [8–10]. The latter two subtribes are still considered as part of Desmodieae . Ohashi  used rbcL phylogeny  and morphological traits to place the groups DESMODIUM, PHYLLODIUM, and LESPEDEZA within the Desmodieae tribe. The DESMODIUM and PHYLLODIUM groups that belong to the subtribe Desmodiinae, include 17 (e.g. Desmodium Desv. and Hylodesmum H. Ohashi & R.R. Mill) and 12 (e.g. Ohwia H. Ohashi, Phyllodium Desv., and Ougeinia Benth.) genera, respectively. The LESPEDEZA group corresponds with subtribe Lespedezinae. This system has been confirmed via subsequent phylogenetic tree that used two chloroplast DNA regions (rbcL, psbA-trnH) , albeit genus Ougeinia was placed within DESMODIUM group. Within the Lespedezinae, the relationship among the three genera–Campylotropis Bunge, Kummerowia Schindl., and Lespedeza Michx.–has been debated based on their inflorescence traits [13–14] or floral structures . Data from phylogenetic studies based on chloroplast DNA and nuclear ribosomal DNA internal transcribed space (nrDNA ITS) of Lespedeza [16–17] have supported the opinion of Nemoto and Ohashi [14–15], i.e. Lespedeza is closer with Kummerowia than Campylotropis.
The plastid genome (plastome) is a source of numerous characters for phylogenetic as well as comparative genomics investigations . For the plastomes of seed plants, their shared conformations, i.e., a quadripartite structure [large single copy (LSC), small single copy (SSC), and pair of inverted repeats (IRs)], gene content, and gene order, have been accepted as their usual characters . Because the typical plastome has a conserved nature, the high number of copies and uni-parental inheritance have been acknowledged as excellent characters for use as molecular markers [12, 19]. Molecular phylogenies, based on complete plastome sequences, have become common and are now used to resolve many phylogenetically puzzling relationships. However, extensive studies have also found that atypical genome structures and gene contents frequently occur among several distantly related families . Among fully photosynthetic angiosperms, the variation in their genome structures is largely attributable to the expansion and contraction of IRs, the most extreme examples being the complete deletion of one copy (e.g., the IR-lacking clade of Fabaceae) and the expansion to an entire single-copy region shown from the genus Asarum L. within Aristolochiaceae [21–22]. Genome rearrangements via large inversions and reductions in gene numbers have been reported from Campanulaceae [23–24], Geraniaceae [25–27], and Oleaceae . Therefore, these atypical characters have provided unprecedented insight into plastome evolution and important traits that can help decipher phylogenetic relationships.
For legumes in the family Fabaceae, Doyle et al. [29–30] and Bailey et al.  have conducted studies with polymerase chain reactions (PCRs) and probe hybridization to reveal lineage-specific inversions (a 50-kb inversion for most of the papilionoid clade and a 78-kb inversion for Phaseoleae subtribe Phaseolinae) as well as gene or intron losses (e.g., the loss of the rpl2 intron and ycf4 in a core member of Desmodieae). Jansen et al.  revealed a rps12 intron loss in Desmodieae and most of IRLC (inverted repeat-lacking clade). More recently, complete plastome studies have re-validated those early findings and elucidated other interesting and atypical evolutions. These include large novel inversions from subfamily Cercidoideae [31–32] and Papilionoideae [33–36]; an accelerated substitution rate [37–38]; and multiple gene and intron losses [19, 32, 35–36]. Some of these characters are lineage-specific and can assist in resolving phylogenetic relationships at different ranking levels, such as the 24-kb inversion for Sophoreae . In contrast, some characters are shared among distantly related taxa, making them unreliable for circumscribing taxa. Those include the rps16 loss and 36-kb inversion for distantly related legume lineages [35–36].
Several molecular studies have surveyed the phylogenetic relationship of Desmodieae with related tribes (e.g. [8, 12, 19]). Those examinations, however, were conducted using restricted taxonomic sampling and/or individual genes, thereby limiting detailed discussions. Although more recent extensive plastome research on Fabaceae has greatly increased our understanding of its evolution, the available resources are still insufficient when considering the vast diversity within that family. In particular, the complete plastome of Desmodieae in the Phaseoloid clade has never been analyzed and the exact status of introns (rpl2 and rps12) and ycf4 losses from the tribe have not been surveyed by genome sequencing. Therefore, the purpose of our research described here was to provide important insight into legume plastome genomics and systematics by sequencing and analyzing the plastome of some species of Desmodieae. Our specific aims included: 1) sequencing of six complete plastomes for Desmodieae and one from its putative sister genus Mucuna Adans., 2) analyzing plastome evolution, and 3) presenting a discussion about the phylogenetics of Desmodieae among Phaseoloid legumes based on plastome sequences and characters.
Materials and methods
The plant species we sampled in Korea and in Japan are neither endangered nor protected. We did not collect any plant from protected areas.
We designed our sampling strategy to represent Desmodieae genera and the genus Mucuna, which is a sister group of this tribe [1–4]. It included six species of Desmodieae [Lespedeza maritima Nakai, Kummerowia striata (Thunb.) Schindl, Campylotropis macrocarpa (Bunge) Rehder, Desmodium heterocarpon (L.) DC., Hylodesmum podocarpum (DC.) H. Ohashi & R.R. Mill subsp. podocarpum, and Ohwia caudata (Thunb.) H. Ohashi], and M. macrocarpa Wall. These species were sampled in order to represent the main groups recognized in this tribe . Campylotropis, Kummerowia, and Lespedeza represented subtribe Lespedezinae and the LESPEDEZA group while the remaining three genera covered subtribe Desmodiinae. Desmodium and Hylodesmum were defined as the DESMODIUM group while Ohwia was representative of the PHYLLODIUM group. Leaves were collected and then preserved with silica gel. Sampling information is shown in S1 Table.
DNA sequencing and genome assembly
Genomic DNAs of the seven species were extracted from the silica-dried leaves with a Qiagen DNeasy Kit (QIAGEN, Seoul, Korea). The extracted genomic DNAs were visualized in 2% agarose gels by electrophoresis, and their quality and quantity were assessed with a NanoDrop ND-1000 (NanoDrop Technologies, Wilmington, DE, USA). The extracted DNA (Lespedeza maritima, 300 ng; all others, 150 ng) was fragmented to 500 bp with a Covaris S220 (Covaris, Woburn, Massachusetts, USA). A TruSeq Nano DNA Library Preparation Kit (Illumina, San Diego, California, USA) was used for library preparation after sequencing on the Illumina MiSeq platform at Life Is Art of Science (LAS; Gimpo, Korea; http://lascience.co.kr/).
For each species, we produced 2,982,620 to 10,758,454 paired-end reads (301 bp for each) (S1 Table). Low-quality reads were removed by Trimmomatic 0.32 . The plastome sequences were assembled following a process we reported previously . First, the paired-end reads of L. maritima were mapped onto the plastomes of the Phaseoloids: Apios americana Medik. (NC_025909), Glycine soja Siebold & Zucc. (NC_022868), Phaseolus vulgaris L. (EU196765), and Vigna radiata (L.) R. Wilczek (NC_013843) with “Medium-Low Sensitivity” settings in Geneious 7.1.8 (Biomatters Ltd., Auckland, NZ). Using those mapped reads, we then conducted the de novo assembly under Medium-Low Sensitivity in Geneious. Afterward, we confirmed all of the assemblies by read depths for the paired-end sequence data. Some regions showing low coverage of reads (<50) were re-checked by PCR, as were the plastome junctions, i.e., LSCs, SSCs, and IRs. The PCR products were treated with an MG PCR Purification Kit (MGmed, Seoul, Korea), and sequenced at Macrogen (Seoul, Korea). Subsequently, the paired-end reads of the six remaining species were mapped to the plastome of L. maritima, and the assembled plastomes were confirmed by the method described above. The process of read mapping and assembly was repeated using assembled contigs as references if the first assembly produce more than one plastome contigs. All primers were designed with Primer3 software .
Genome annotation, alignments, and visualization
The plastid genes for all seven tested species were annotated using Geneious 7.1.8, based on the annotation of Glycine max (L.) Merr. (NC_007942). This process was implemented when nucleotide sequences of the plastid genes for tested species showed over 90% of similarity with reference genome. Some protein-coding genes were manually identified by considering their start and stop codons. The tRNAs were confirmed by searching in tRNAscan-SE  and by comparing with reference data. Genome maps were drawn with OGDraw . We deposited our complete plastome sequences in the GenBank database (https://www.ncbi.nlm.nih.gov/) (MG867566–MG867572).
Analysis of gene/intron losses and genome rearrangement
Comparative analyses for intron, gene, and genome rearrangement included the seven completed plastomes of this study, and six completed plastome data of Phaseoleae legumes (S2 Table). Four genes (rpl2, rps12, rps16, and ycf4) were extracted from each plastome. Intron losses and pseudogenization for each sequence were aligned using the default parameters of MUSCLE , and manually edited. The entire plastome was aligned using progressive Mauve 2.3.1  with default settings to check for a genome rearrangement event. The order of genes in the cluster trnD-GUC, trnY-GUA, and trnE-UUC was compared between Desmodieae and the Phaseoloid legumes by sequence alignment as described above. Expansion or contraction of the IR region was investigated and visualized by IRscope .
Data from ten additional plastomes from GenBank were used to construct a phylogenetic tree for Desmodieae (S2 Table). Millettioids and Indigofereae plastomes were used as outgroups, because these two taxa were the sister groups to Phaseoloids in previous phylogenetic studies [1–2]. In all, we selected 67 plastid protein-coding genes that are conserved among those taxa (S3 Table). After combining gene sequences according to each taxon, we aligned them with MUSCLE under default parameters. Poorly aligned regions were either manually refined or deleted using Geneious 7.1.8. The final sequence alignment is available at “Supporting information” (S1 Fasta File). To construct the tree with a Maximum Likelihood (ML) analysis, we chose a nucleotide substitution model throughout jModelTest 2.1.6 . For this process, total 88 substitution models were compared along with a gamma distribution of site heterogeneity; followed by GTR + I + G being selected according to the Akaike Information criterion (AIC). The aligned data set was employed to construct an ML tree through RAxML  with 1,000 replicates.
Plastid genome sequences and contents
The plastid genome maps for six species from tribe Desmodieae and one from Mucuna are illustrated in Fig 2, S1 Fig and Table 1. Desmodium heterocarpon was the representative of Desmodieae. Plastome maps for the remaining species are shown in S1 Fig. The complete plastome of Desmodieae members ranges in length from 148,450 to 150,249 bp, and consists of an LSC (81,942–83,241 bp), an SSC (18,159–18,939 bp), and two IRs (each 23,720–24,264 bp). The plastome of Mucuna (153,826 bp) is larger than that of the Desmodieae species due to expansion by intergenic spacers (IGSs). Each genome harbors 128 genes, including 83 protein-coding genes, eight rRNA genes, and 37 tRNA genes (S4 Table). Approximately 49.7 to 52.5% of the total genome consists of protein-coding region, while the remaining 47.5 to 50.3% is composed of tRNA, rRNA, introns, and IGSs. The AT and GC contents are 64.8 to 65.1% and 34.9 to 35.2%, respectively. The plastome of M. macrocarpa is similar to that of the Desmodieae members based on genome features such as gene contents and length.
Genes on outside of outer circle are transcribed in clockwise direction; those on inside of outer circle are transcribed in counterclockwise direction. Colored rectangles indicate functional genes, with categories shown on bottom left. Gray scale in inner circle indicates GC content of plastid genome.
For these seven species, we observed the loss of infA and rpl22 that is typical of all legumes (S4 Table). The introns of rpl2 and rps12 were absent from all six Desmodieae species while Mucuna lost the rpl2 intron but retained the rps12 intron (S2 Fig). Although the sequence of ycf4 was highly variable (S3 Fig), the open reading frame (ORF) was intact for all but Desmodium. In particular, ycf4 from D. heterocarpon lost its start codon through substitution and showed abundant internal stop codons due to mutations in the nucleotide sequence. In M. macrocarpa, rps16 had a 70-bp deletion in exon 2 that resulted in a severe frame shift of amino acids (S4 Fig). Consequently, we deemed ycf4 of D. heterocarpon and rps16 of M. macrocarpa to be pseudogenes.
We examined structural changes in the Desmodieae and Mucuna plastomes by comparing them with other legumes. A 50-kb inversion in the LSC, common to most members of Papilionoideae, was shared among all the six Desmodieae species and M. macrocarpa (Fig 2 and S1 Fig). In addition, six plastomes exhibited a novel inversion (ca. 1.5-kb) of gene cluster trnE-UUC, trnY-GUA, and trnD-GUC specific to Desmodieae, located in the LSC (Fig 3). This inversion was situated between trnT-GGU and psbM. Among Phaseoloids, the typical arrangement of the five genes was trnT-GGU—trnE-UUC—trnY-GUA—trnD-GUC—psbM. However, members of Desmodieae had an inverted arrangement of trnT-GGU—trnD-GUC—trnY-GUA—trnE-UUC—psbM. The break point of this inversion coincided with a pair of 11-bp inverted repeats located on either side of trnE-Y-D (TATTGGATTTG and CAAATCCAATA) (Fig 3). However, this break point was absent from subtribe Lespedezinae (Campylotropis macrocarpa, Kummerowia striata, and Lespedeza maritima) because of a 500-bp deletion from the IGS of trnE-UUC and psbM.
(A) Comparisons of inversion for each tribe and subtribe. Nucleotide sequences were simplified in narrow rectangles; same-colored rectangles indicate similar sequence. The 11-bp repeats (blue bars) coincide with break points of inversion. (B) Aligned repeat motifs located upstream of trnT-GGU (Rt) and upstream of psbM (Rp). Rp sequences were reversely aligned.
No severe expansion or contraction of IRs occurred at the junctions for three of four regions, i.e., JLA (junction IRa/LSC), JLB (junction IRb/LSC), and JSB (junction IRb/SSC) (Fig 4). The JLA, JLB, and JSB junctions were conserved as the rps19-rpl2, rpl2-trnH, and ndhF regions, respectively. However, we noted a distinction from JSA (junction IRa/SSC) for Desmodieae subtribe Lespedezinae (Campylotropis, Kummerowia, and Lespedeza). These genera were characterized by an approximately 600-bp contraction of IR that resulted from the absence of a partial ycf1 copy (Table 1).
The cladistic (maximum likelihood) analyses of the aligned gene sequences revealed the phylogenetic relationships (Fig 5). The alignments included 50,301 nucleotide positions. All nodes of the ML tree were strongly supported by bootstrap values of 100%. On this ML tree, Apios americana is the sister group to the remaining Phaseoloid legumes. The remaining taxa were grouped into two clades. One clade consists of seven taxa from the tribe Desmodieae and its putative sister genus Mucuna, which was newly sequenced in the present study. The other clade is constructed of core-Phaseoleae members (except for A. americana and M. macrocarpa). In the former clade, two subtribes of Desmodieae (Desmodiinae and Lespedezinae) were also monophyletic groups. This topology of Desmodiinae coincided with the groups recognized within that subtribe: the DESMODIUM group (incl. Desmodium and Hylodesmum) and the PHYLLODIUM group (incl. Ohwia). In the case of Lespedezinae, Lespedeza and Kummerowia were more closely related to each other than to Campylotropis. In another clade, Cajanus cajan (L.) Huth (subtribe Cajaninae of Phaseoleae) was the earliest diverging taxon, while Phaseolus vulgaris and Vigna unguiculata (L.) Walp. (subtribe Phaseolinae of Phaseoleae) were reconstructed as a monophyletic group. In addition, Pediomelum argophyllum (Pursh) J.W. Grimes and Psoralidium tenuiflorum (Pursh) Rydb. (tribe Psoraleeae) were also a monophyletic group, nested within members of subtribe Glycininae of Phaseoleae [Glycine gracilis Skvortsov and Pachyrhizus erosus (L.) Urb.].
This tree shown is the one of 1,000 trees derived from ML analysis of 67 concatenated protein-coding gene sequences, log likelihood number = -146379.564971. All nodes were supported with absolute 100% bootstrap values. The fruits of the subtribes [Desmodiinae (multiple-seeded loments) and Lespedezinae (single-seeded legume)] are shown on the side. Asterisks, taxa sequenced here; red and green rectangles on bar, intron and gene losses, respectively, from plastid genome; black rectangles, genome rearrangement events; D, DESMODIUM group; L, LESPEDEZA group; P, PHYLLODIUM group; IRLC, inverted repeat-lacking clade.
We sequenced seven plastomes that included species from six representative genera from the tribe Desmodieae and one species of Mucuna, the sister genus of the tribe in the previous supertrees [1–2]. The sizes of those seven plastomes do not deviate significantly from the typical genome length of Phaseoloids (e.g. Glycine max) . One feature in common to all of these tested species is the absence of infA and rpl22 like other legume species. However, two genes show signs of pseudogenization: rps16 (in Mucuna) and ycf4 (in Desmodium). A multiple gene-loss event of rps16 has already been reported for various legume lineages [29, 35, 49]. Our comparative analysis demonstrated that the multiple loss event of rps16 was also found in Phaseoloid legumes (Fig 5). The cause of such events is assumed to be a consequence of dual targeting of the nuclear rps16 copy to the plastid as well as the mitochondria .
Results from an earlier study using slot blot hybridization  suggested a multiple loss of ycf4 from numerous legume tribes. Moreover, Bailey et al.  have argued that the lack of ycf4 can be used as molecular markers for the tribe Desmodieae, but they also note the extremely complex results from the subtribe Desmodiinae and Lespedezinae. In our results, Desmodieae plastomes showed substantial sequence substitutions and indels from each other (S3 Fig), which can produce complex signals from slot blot hybridization [8, 29]. However, sequence variations in most species are without internal stop codons or frameshift changes, with the exception of D. heterocarpon. Thus, the loss of ycf4 is not a shared character of Desmodieae in Phaseoloid legumes. However, Magee et al.  identified hypermutation of ycf4, and associated multiple losses of IRLC and several Phaseoloid species. Hence, the sequence divergence and subsequent loss of ycf4 in Desmodieae might not be caused by recent events in the tribe but may be related to ancient events that predated at least Phaseoloid, or the combination of IRLC and Phaseoloid.
Our complete plastome data shows the loss of an intron from rpl2 and rps12. The former has previously been reported from three distantly related genera (Bauhinia L., Soemmeringia Mart., and Mucuna) and the tribe Desmodieae [8, 29, 51]. Recently, the rpl2 intron loss from genus Bauhinia has been confirmed in a plastome analysis involving subfamily Cercidoideae . Here, we confirmed that an intron of rpl2 has been deleted from genus Mucuna and all six of the investigated Desmodieae genera.
The intron loss of rps12 among legumes was first investigated by Jansen et al.  who determined it occurred independently in some genera of Desmodieae (Desmodium, Kummerowia, Lespedeza) and in most of the members of IRLC. The exceptions were some early-diverging genera such as Callerya Endl. and Wisteria Nutt. . However, no examination had been made about whether the loss of the rps12 intron is shared by Desmodieae and Mucuna. Our data indicated that the loss of the rps12 intron was only limited to the six genera of Desmodieae, but not Mucuna. When considering our data, the rpl2 intron loss, shared by Desmodieae and Mucuna, appears to have happened prior to the rps12 intron loss since the tribe diverged from the genus.
Desmodieae-specific inversion of the trnE-Y-D region and IR contraction in the subtribe Lespedezinae
We found a novel plastome rearrangement specific to tribe Desmodieae (Fig 3) that features an inversion approximately 1.5 kb long and contains three tRNA genes (trnE, trnY, and trnD). This inversion is located between trnT-GGU and psbM. The gene order from trnT through trnD (trnT-GGU—trnE-UUC—trnY-GUA—trnD-GUC) is well-conserved among seed plants. Moreover, this region (also known as the trnD-trnT intergenic spacer) is part of the “Tier 1 region” selected by Shaw et al.  for phylogenetic analysis because of its high potential as an informative character. Therefore, this region is frequently used as a molecular marker in various seed plants phylogenies, e.g., Camassia Lindl.  and Pinus L. , and with Phaseoloid legumes such as Pueraria DC. . However, the disruption of this molecular marker (trnD-trnT) by inversion in Desmodieae means it is impossible to make a direct application of this marker to that tribe. As an alternative, future phylogenetic research on Phaseoloid legumes (including Desmodieae) would utilize only the conserved collinear block of trnD-trnE.
The plastome rearrangement can emerge through recombination [34, 56–57]. The Desmodieae-specific inversion discovered here coincides with a pair of 11-bp inverted repeats, “TATTGGATTTG” and “CAAATCCAATA” (Fig 3). If we consider this match of inversion breakpoints and inverted repeats, then we should regard the inversion of Desmodieae as a consequence of microhomology-driven recombination events via 11-bp inverted repeats. On the other hand, this inversion could be additional genetic evidence for the monophyly of Desmodieae. We found it interesting that one of the 11-bp repeat motifs is deleted from three genera of Desmodieae subtribe Lespedezinae (Fig 3). This was caused by a 500-bp deletion from the IGS between trnT-GGU and psbM. Variations in IGS length according to subtribes could serve as genetic markers for identifying the two subtribes of Desmodieae.
We detected an IR contraction from the SSC region in members of tribe Desmodieae subtribe Lespedezinae (Campylotropis, Kummerowia, and Lespedeza). Such variations in IR junctions from the LSC and SSC due to IR expansion or contraction have often been described from various types of land plants (e.g. [58–60]). In the plastomes of legumes, examples include notable IR variations such as a deletion in the IRLC  or expansion in the inverted repeat-expanding clade [62–63]. The contraction of IR shown from Lespedezinae is relatively small (ca. 600 bp) when comparing to other changes in the family . Though, the removal of a partial copy of ycf1 from the IR is noteworthy because the partial or entire duplication of ycf1 by IRa is a well-conserved character among legumes [48, 64]. Therefore, the absence of a partial ycf1 is a distinct feature of the plastomes for Desmodieae subtribe Lespedezinae.
Phylogeny of the tribe Desmodieae within Phaseoloid legumes
The Maximum Likelihood (ML) analysis of 15 Phaseoloids legumes also used the related taxa Millettia pinnata (L.) Panigrahi and Indigofera tinctoria L. as outgroups. Based upon 67 conserved protein coding-genes, a matK phylogeny analysis revealed these outgroups as sister groups . Earlier plastome-scale phylogenetic analyses of Fabaceae members had also been conducted by our research group as well as others (e.g., [35–36]). However, the work presented here is the first to focus on Phaseoloids and to cover tribe Desmodieae and Mucuna. In doing so, we compared the plastome phylogeny of Phaseoloids with previous phylogenies based on partial DNAs (e.g., [1–3, 10, 12, 55]) and found that our results are consistent with the earlier findings, even though our sampling was not as complete as the earlier studies. As shown by our plastome phylogenetic tree (Fig 5), Apios is earliest branching while the remaining Phaseoloid legumes form two clades: core-Phaseoleae, a group that includes three subtribes (Cajaninae, Glycininae, and Phaseolinae) and the tribe Psoraleeae , and Desmodieae/Mucuna. In the former, Phaseolinae is confirmed as a monophyletic group, and Psoraleeae is also a monophyletic group that is nested within Glycininae, which is supported by previous phylogenetic studies [2, 55]. Members of Desmodieae are gathered as a monophyletic group and as a sister to Mucuna. Consequently, Desmodieae and Psoraleeae are found to monophyletic while simultaneously being included within Phaseoleae. Given this, we might reconsider the taxonomic rank of Desmodieae and Psoraleeae as a subtribe of Phaseoleae or we might change the other subtribes of Phaseoleae to a tribal level . This particular topology may also explain the loss of genes/introns among Phaseoloid legumes. As shown in our plastome maps (Fig 2 and S1 Fig), the rpl2 intron loss is shared in Desmodieae and Mucuna [8, 29] while the rps12 intron loss is detected only in Desmodieae . Taking into account this plastome phylogeny, it is likely that the rpl2 intron loss preceded the rps12 intron loss within the Phaseoloids. Furthermore, we might infer that the rps16 gene loss occurred independently at least three times in those Phaseoloids, i.e., two losses when Apios and Mucuna were separated from their recent ancestors and a third loss when subtribe Phaseolinae diverged.
The two subtribes of Desmodieae (Desmodiinae and Lespedezinae) are well-supported as a monophyletic group, an outcome similar to that already reported from phylogenetic studies of that tribe (e.g., ). This finding is also in accordance with circumstances based on morphological traits . Desmodiinae plants bear multiple-seeded loments, standard petal without auricles at the base, and frequent stipels whereas Lespedezinae has single-seeded fruits, a standard with auricles at the base, and no stipels . These novel genome rearrangements that were determined for Lespedezinae in the present study (i.e., 500-bp deletion in the IGS at trnE-psbM, and an IR contraction; Figs 3 and 4) support the monophyly of Lespedezinae. In addition, our phylogenetic tree recognized the three groups (DESMODIUM, PHYLLODIUM, and LESPEDEZA) of this tribe that are suggested by rbcL phylogeny and morphological traits . One taxonomic study based on rbcL phylogeny and morphological characters placed Hylodesmum within DESMODIUM ; however, matK phylogenetic tree placed Hylodesmum at early branching group of Desmodiinae . Our ML tree shows that H. podocarpum is grouped with Desmodium heterocarpon (included in DESMODIUM group), similar to another cpDNA (rbcL, psbA-trnH) phylogeny . Because of controversy regarding the relationship among genera of Lespedezinae based on morphological characters, several phylogenetic studies have been conducted (e.g., [16–17]). Our plastome tree indicated that Campylotropis is a sister to the group containing Kummerowia and Lespedeza, and this result coincides with those from other analyses. Therefore, we recognize that Lespedeza is more closely related to Kummerowia.
S1 Fig. Maps of plastid genome of tribe Desmodieae and genus Mucuna.
(A) Hylodesmum podocarpum subsp. podocarpum. (B) Ohwia caudata. (C) Campylotropis macrocarpa. (D) Kummerowia striata. (E) Lespedeza maritima. (F) Mucuna macrocarpa. Genes on outside of outer circle are transcribed in clockwise direction; those on inside of outer circle are transcribed in counterclockwise direction. Colored rectangles indicate functional genes, with categories shown on bottom left. Gray scale in inner circle indicates GC content of plastid genome.
S2 Fig. Comparison of rpl2 and rps12 introns in Phaseoloid legumes.
Red-shaded rectangles indicate absence of introns. Parts of introns were omitted. (A) rpl2 sequence. (B) rps12 sequence.
S3 Fig. Alignment of ycf4 gene sequences in Phaseoloid legumes.
Red-shaded rectangles indicate severe nucleotide variations that resulted in frameshift or missing start codon. This gene from Desmodium heterocarpon is considered to be pseudogenes.
S4 Fig. Alignment of rps16 gene sequences in Phaseoloid legumes.
Red-shaded rectangles indicate severe nucleotide variations that resulted in frameshift or missing start codon. Parts of introns were omitted. rps16 genes from Apios americana, Phaseolus vulgaris, and Vigna radiata are considered to be pseudogenes.
S1 Table. Sampling and sequencing information for taxa from tribe Desmodieae and Mucuna.
S2 Table. GenBank accession numbers of taxon used in plastome phylogeny.
S3 Table. Gene list employed in plastome phylogeny.
S4 Table. Gene list of taxa analyzed from tribe Desmodieae and Mucuna.
The authors thank the National Institute of Biological Resources (NIBR) of Korea for providing DNA samples of Ohwia caudata (voucher no. NIBR378625). We are also grateful to colleagues JS Park and JW Park at the Plant Systematics Laboratory of Inha University for their help in editing this manuscript.
- 1. Legume Phylogeny Working Group. Legume phylogeny and classification in the 21st century: progress, prospects and lessons for other species-rich clades. Taxon. 2013;62: 217–248.
- 2. Legume Phylogeny Working Group. A new subfamily classification of the Leguminosae based on a taxonomically comprehensive phylogeny. Taxon. 2017;66: 44–77.
- 3. Stefanović S, Pfeil BE, Palmer JD, Doyle JJ. Relationships among phaseoloid legumes based on sequences from eight chloroplast regions. Syst Bot. 2009;34: 115–128.
- 4. Schrire BD. Phaseoleae. In: Lewis G, Schrire B, Mackinder B, Lock M, editors. Legumes of the World. Richmond: Royal Botanic Gardens, Kew; 2005. pp. 392–431.
- 5. Ohashi H. Desmodieae. In: Lewis G, Schrire B, Mackinder B, Lock M, editors. Legumes of the World. Richmond: Royal Botanic Gardens, Kew; 2005. pp. 432–445.
- 6. Ohashi H, Ohashi K. Ototropis, a genus separated from Desmodium (Leguminosae). J Jap Bot. 2012;87: 108–118.
- 7. Ohashi H, Polhill RH, Schubert BG. Desmodieae. In: Polhill RM, Raven PH, editors. Advances in Legume Systematics. Part 1. Richmond: Royal Botanic Gardens, Kew; 1981. pp. 292–300.
- 8. Bailey CD, Doyle JJ, Kajita T, Nemoto T, Ohashi H. The chloroplast rpl2 intron and ORF184 as phylogenetic markers in the legume tribe Desmodieae. Syst Bot. 1997;22: 133–138.
- 9. Doyle JJ, Chappill JA, Bailey CD, Kajita T. Towards a comprehensive phylogeny of legumes: evidence from rbcL sequences and non-molecular data. In: Herendeen PS, Bruneau A, editors. Advances in Legume Systematics. Part 9. Richmond: Royal Botanic Gardens, Kew; 2000. pp. 1–20.
- 10. Lavin M, Pennington RT, Klitgaard BB, Sprent JI, Lima HC, Gasson PE. The dalbergioid legumes (Fabaceae): delimitation of a monophyletic pantropical clade. Am J Bot. 2001;88: 503–533. pmid:11250829
- 11. Kajita T, Ohashi H, Tateishi Y, Bailey CD, Doyle JJ. rbcL and legume phylogeny, with particular reference to Phaseoleae, Millettieae, and allies. Syst Bot. 2001;26: 515–536.
- 12. Jabbour F, Gaudeul M, Lambourdière J, Ramstein G, Hassanin A, Labat JN, et al. Phylogeny, biogeography and character evolution in the tribe Desmodieae (Fabaceae: Papilionoideae), with special emphasis on the New Caledonian endemic genera. Mol Phylogenet Evol. 2018;118: 108–121. pmid:28966123
- 13. Akiyama S, Ohba H. The branching of inflorescence and vegetative shoot and taxonomy of the genus Kummerowia (Leguminosae). Bot Mag Tokyo. 1985;78: 137–150.
- 14. Nemoto T, Ohashi H. Organographic and ontogenetic studies on the inflorescence of Lespedeza cuneata (Dum. Cours.) G. Don (Leguminosae). Bot Mag Tokyo. 1990;103: 217–231.
- 15. Nemoto T, Ohashi H. Floral nectaries in Lespedeza, Kummerowia and Campylotropis (Leguminosae). J Jap Bot. 1988;63: 112–126.
- 16. Han JE, Chung KH, Nemoto T, Choi BH. Phylogenetic analysis of eastern Asian and eastern North American disjunct Lespedeza (Fabaceae) inferred from nuclear ribosomal ITS and plastid region sequences. Bot J Linn Soc. 2010;164: 221–235.
- 17. Xu B, Wu N, Gao XF, Zhang LB. Analysis of DNA sequences of six chloroplast and nuclear genes suggests incongruence, introgression, and incomplete lineage sorting in the evolution of Lespedeza (Fabaceae). Mol Phylogenet Evol. 2012;62: 346–358. pmid:22032991
- 18. Wolf PG, Roper JM, Duffy AM. The evolution of chloroplast genome structure in ferns. Genome. 2010;53: 731–738. pmid:20924422
- 19. Jansen RK, Wojciechowski MF, Sanniyasi E, Lee SB, Daniell H. Complete plastid genome sequence of the chickpea (Cicer arietinum) and the phylogenetic distribution of rps12 and clpP intron losses among legumes (Leguminosae). Mol Phylogenet Evol. 2008;48: 1204–1217. pmid:18638561
- 20. Jansen RK, Ruhlman TA. Plastid genomes of seed plants. In: Bock R, Knoop V, editors. Genomics of Chloroplasts and Mitochondria. Dordrecht, Netherlands: Springer; 2012. pp. 103–126.
- 21. Lim CE, Lee SC, So S, Han SM, Choi JE, Lee BY. The complete chloroplast genome sequence of Asarum sieboldii Miq. (Aristolochiaceae), a medicinal plant in Korea. Mitochondrial DNA B Resour. 2018;3: 118–119.
- 22. Sinn BT, Sedmak DD, Kelly LM, Freudenstein JV. Total duplication of the small single copy region in the angiosperm plastome: rearrangement and inverted repeat instability in Asarum. Am J Bot. 2018;105: 71–84. pmid:29532923
- 23. Cosner ME, Raubeson LA, Jansen RK. Chloroplast DNA rearrangements in Campanulaceae: phylogenetic utility of highly rearranged genomes. BMC Evol Biol. 2004;4: 27. pmid:15324459
- 24. Haberle RC, Fourcade HM, Boore JL, Jansen RK. Extensive rearrangements in the chloroplast genome of Trachelium caeruleum are associated with repeats and tRNA genes. J Mol Evol. 2008;66: 350–361. pmid:18330485
- 25. Chumley TW, Palmer JD, Mower JP, Fourcade HM, Calie PJ, Boore JL, et al. The complete chloroplast genome sequence of Pelargonium x hortorum: organization and evolution of the largest and most highly rearranged chloroplast genome of land plants. Mol Biol Evol. 2006;23: 2175–2190. pmid:16916942
- 26. Weng ML, Blazier JC, Govindu M, Jansen RK. Reconstruction of the ancestral plastid genome in Geraniaceae reveals a correlation between genome rearrangements, repeats, and nucleotide substitution rates. Mol Biol Evol. 2014;31: 645–659. pmid:24336877
- 27. Röschenbleck J, Wicke S, Weinl S, Kudla J, Müller KF. Genus-wide screening reveals four distinct types of structural plastid genome organization in Pelargonium (Geraniaceae). Genome Biol Evol. 2016;9: 64–76. pmid:28172771
- 28. Lee HL, Jansen RK, Chumley TW, Kim KJ. Gene relocations within chloroplast genomes of Jasminum and Menodora (Oleaceae) are due to multiple, overlapping inversions. Mol Biol Evol. 2007;24: 1161–1180. pmid:17329229
- 29. Doyle JJ, Doyle JL, Palmer JD. Multiple independent losses of two genes and one intron from legume chloroplast genome. Syst Bot. 1995;20: 272–294.
- 30. Doyle JJ, Doyle JL, Ballenger JA, Palmer JD. The distribution and phylogenetic significance of a 50-kb chloroplast DNA inversion in the flowering plant family Leguminosae. Mol Phylogenet Evol. 1996;5: 429–438. pmid:8728401
- 31. Kim Y, Cullis C. A novel inversion in the chloroplast genome of marama (Tylosema esculentum). J Exp Bot. 2017;68: 2065–2072. pmid:28158587
- 32. Wang YH, Wicke S, Wang H, Jin JJ, Chen SY, Zhang SD, et al. Plastid genome evolution in the early-diverging legume subfamily Cercidoideae (Fabaceae). Front Plant Sci. 2018;9: 138. pmid:29479365
- 33. Kazakoff SH, Imelfort M, Edwards D, Koehorst J, Biswas B, Batley J, et al. Capturing the biofuel wellhead and powerhouse: the chloroplast and mitochondrial genomes of the leguminous feedstock tree Pongamia pinnata. PLoS One. 2012;7: e51687. pmid:23272141
- 34. Martin GE, Rousseau-Gueutin M, Cordonnier S, Lima O, Michon-Coudouel S.; Naquin D, et al. The first complete chloroplast genome of the Genistoid legume Lupinus luteus: evidence for a novel major lineage-specific rearrangement and new insights regarding plastome evolution in the legume family. Ann Bot. 2014;113: 1197–1210. pmid:24769537
- 35. Schwarz EN, Ruhlman TA, Sabir JSM, Hajarah NH, Alharbi NS, Al-Malki AL, et al. Plastid genome sequences of legumes reveal parallel inversions and multiple losses of rps16 in papilionoids. J Syst Evol. 2015;53: 458–468.
- 36. Choi IS, Choi BH. The distinct plastid genome structure of Maackia fauriei (Fabaceae: Papilionoideae) and its systematic implications for genistoids and tribe Sophoreae. PLoS One. 2017;12: e0173766. pmid:28399123
- 37. Magee AM, Aspinall S, Rice DW, Cusack BP, Semon M, Perry AS, et al. Localized hypermutation and associated gene losses in legume chloroplast genomes. Genome Res. 2010;20: 1700–1710. pmid:20978141
- 38. Schwarz EN, Ruhlman TA, Weng ML, Khiyami MA, Sabir JSM, Hajarah NH, et al. Plastome-wide nucleotide substitution rates reveal accelerated rates in Papilionoideae and correlations with genome features across legume subfamilies. J Mol Evol. 2017;84: 187–203. pmid:28397003
- 39. Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30: 2114–2120. pmid:24695404
- 40. Untergasser A, Cutcutache I, Koressaar T, Ye J, Faircloth BC, Remm M, et al. Primer3—new capabilities and interfaces. Nucleic Acids Res. 2012;40: e115. pmid:22730293
- 41. Lowe TM, Chan PP. tRNAscan-SE On-line: integrating search and context for analysis of transfer RNA genes. Nucleic Acids Res. 2016;44: W54–W57. pmid:27174935
- 42. Lohse M, Drechsel O, Kahlau S, Bock R. OrganellarGenomeDRAW—a suite of tools for generating physical maps of plastid and mitochondrial genomes and visualizing expression data sets. Nucleic Acids Res. 2013;41: W575–W581, pmid:23609545
- 43. Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 2004;32: 1792–1797. pmid:15034147
- 44. Darling AC, Mau B, Blattner FR, Perna NT. Mauve: multiple alignment of conserved genomic sequence with rearrangements. Genome res. 2004;14: 1394–1403. pmid:15231754
- 45. Amiryousefi A, Hyvönen J, Poczai P. IRscope: an online program to visualize the junction sites of chloroplast genomes. Bioinformatics 2018;34: 3030–3031. pmid:29659705
- 46. Darriba D, Taboada GL, Doallo R, Posada D. jModelTest 2: more models, new heuristics and parallel computing. Nat Methods. 2012;9: 772. pmid:22847109
- 47. Stamatakis A. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics. 2014;30: 1312–1313. pmid:24451623
- 48. Saski C, Lee SB, Daniell H, Wood TC, Tomkins J, Kim HG, et al. Complete chloroplast genome sequence of Glycine max and comparative analyses with other legume genomes. Plant Mol Biol. 2005;59: 309–322. pmid:16247559
- 49. Keller J, Rousseau-Gueutin M, Martin GE, Morice J, Boutte J, Coissac E, et al. The evolutionary fate of the chloroplast and nuclear rps16 genes as revealed through the sequencing and comparative analyses of four novel legume chloroplast genomes from Lupinus. Dna Res. 2017;24: 343–358. pmid:28338826
- 50. Ueda M, Nishikawa T, Fujimoto M, Takanashi H, Arimura SI, Tsutsumi N, et al. Substitution of the gene for chloroplast RPS16 was assisted by generation of a dual targeting signal. Mol Biol Evol. 2008;25: 1566–1575. pmid:18453549
- 51. Lai M, Sceppa J, Ballenger JA, Doyle JJ, Wunderlin RP. Polymorphism for the presence of the rpl2 intron in chloroplast genomes of Bauhinia (Leguminosae). Syst Bot. 1997;22: 519–528.
- 52. Shaw J, Lickey EB, Beck JT, Farmer SB, Liu WS, Miller J, et al. The tortoise and the hare II: relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis. Am J Bot. 2005;92: 142–166. pmid:21652394
- 53. Fishbein M, Kephart SR, Wilder M, Halpin KM, Datwyler SL. Phylogeny of Camassia (Agavaceae) inferred from plastid rpl16 intron and trnD–trnY–trnE–trnT intergenic spacer DNA sequences: implications for species delimitation. Syst Bot. 2010;35: 77–85.
- 54. Hernández-León S, Gernandt DS, de la Rosa JAP, Jardón-Barbolla L. Phylogenetic relationships and species delimitation in Pinus section Trifoliae inferred from plastid DNA. PLoS One. 2013;8: e70501. pmid:23936218
- 55. Egan AN, Vatanparast M, Cagle W. Parsing polyphyletic Pueraria: delimiting distinct evolutionary lineages through phylogeny. Mol Phylogenet Evol. 2016;104: 44–59. pmid:27495827
- 56. Rogalski M, Ruf S, Bock R. Tobacco plastid ribosomal protein S18 is essential for cell survival. Nucleic Acids Res. 2006;34: 4537–4545. pmid:16945948
- 57. Ruhlman TA, Zhang J, Blazier JC, Sabir JSM, Jansen RK. Recombination-dependent replication and gene conversion homogenize repeat sequences and diversify plastid genome structure. Am J Bot. 2017;104: 559–572, pmid:28400415
- 58. Wang RJ, Cheng CL, Chang CC, Wu CL, Su TM, Chaw SM. Dynamics and evolution of the inverted repeat-large single copy junctions in the chloroplast genomes of monocots. BMC Evol Biol. 2008;8: 36. pmid:18237435
- 59. Wu CS, Wang YN, Hsu CY, Lin CP, Chaw SM. Loss of different inverted repeat copies from the chloroplast genomes of Pinaceae and cupressophytes and influence of heterotachy on the evaluation of gymnosperm phylogeny. Genome Biol Evol. 2011;3: 1284–1295. pmid:21933779
- 60. Sanderson MJ, Copetti D, Búrquez A, Bustamante E, Charboneau JL, Eguiarte LE, et al. Exceptional reduction of the plastid genome of saguaro cactus (Carnegiea gigantea): Loss of the ndh gene suite and inverted repeat. Am J Bot. 2015;102: 1115–1127. pmid:26199368
- 61. Wojciechowski MF, Sanderson MJ, Steele KP, Liston A. Molecular phylogeny of the ‘‘temperate herbaceous tribes” of papilionoid legumes: a supertree approach. In: Herendeen PS, Bruneau A, editors. Advances in Legume Systematics. Part 9. Richmond, UK: Royal Botanic Gardens, Kew; 2000. pp. 277–298.
- 62. Dugas DV, Hernandez D, Koenen EJM, Schwarz E, Straub S, Hughes CE, et al. Mimosoid legume plastome evolution: IR expansion, tandem repeat expansions, and accelerated rate of evolution in clpP. Sci Rep. 2015;5: 16958. pmid:26592928
- 63. Wang YH, Qu XJ, Chen SY, Li DZ, Yi TS. Plastomes of Mimosoideae: structural and size variation, sequence divergence, and phylogenetic implication. Tree Genet Genomes. 2017;13: 41.
- 64. Williams AV, Boykin LM, Howell KA, Nevill PG, Small I. The complete sequence of the Acacia ligulata chloroplast genome reveals a highly divergent clpP1 gene. PLoS One. 2015;10: e0125768. pmid:25955637