Comparative and Phylogenetic Analysis of Complete Chloroplast Genomes in Leymus (Triticodae, Poaceae)

Leymus is a perennial genus that belongs to the tribe Triticeae (Poaceae) which has an adaptive capacity to ecological conditions and strong resistance to cold, drought, and salinity. Most Leymus species are fine herbs that can be used for agriculture, conservation, and landscaping. Due to confusion taxonomy within genera, the complete chloroplast (cp) genome of 13 Leymus species was sequenced, assembled, and compared with those of three other previously published Leymus species (Leymus condensatus, Leymus angustus, and Leymus mollis) to clarify the issue. Overall, the whole cp genome size ranged between 135,057 (L. condensatus) and 136,906 bp (Leymus coreanus) and showed a typical quadripartite structure. All studied species had 129 genes, including 83 protein-coding genes, 38 transfer RNAs, and 8 ribosomal RNAs. In total, 800 tandem repeats and 707 SSR loci were detected, most of which were distributed in the large single-copy region, followed by the inverted repeat (IR) and small single-copy regions. The sequence identity of all sequences was highly similar, especially concerning the protein-coding and IR regions; in particular, the protein-coding regions were significantly similar to those in the IR regions, regardless of small sequence differences in the whole cp genome. Moreover, the coding regions were more conserved than the non-coding regions. Comparisons of the IR boundaries showed that IR contraction and expansion events were reflected in different locations of rpl22, rps19, ndhH, and psbA genes. The close phylogenetic relationship of Leymus and Psathyrostachys indicated that Psathyrostachys possibly is the donor of the Ns genome sequence identified in Leymus. Altogether, the complete cp genome sequence of Leymus will lay a solid foundation for future population genetics and phylogeography studies, as well as for the analysis of the evolution of economically valuable plants.


Introduction
Leymus Hochst is a significant perennial grass species of Triticeae (Poaceae) that is mainly distributed in Eurasia and North America [1,2] with strong adaptable characteristics to several environmental stressors such as drought, cold, and salinity [3]. Several species have been used for the genetic improvement of Triticeae cereal crops because of their larger spikes, higher grain yields, and better resistance to diseases and insects [4,5].
Nevertheless, its precise taxonomic status and the relationship among Leymus species are still debatable. Although morphological identification is relatively certain during the flowering period, it is difficult to perform morphological identification for roots, stems, and leaves in other growth stages, especially in the seedling stages. Initially, Nevski hypothesized that Leymus consisted of sect. Leymus and sect. Anisopyrum [6], but later on Tzvelev divided Leymus into four parts: sect. Leymus Hochst., sect. Anisopyrum (Griseb.) Tzvelev, sect. Aphanoneuron (Nevski) Tzvelev, and sect. Malacurus (Nevski) Tzvelev [7]. It has been confirmed that single-copy genes, the mitochondrial coxII intron, molecular markers containing AFLP, RAPD, and ISSR, as well as chloroplast intergenic spacers, are useful tools

Sampling, DNA Extraction, and Genome Sequencing
The cp genomes of Leymus arenarius (MK775256.1), Leymus duthiei (NC_058748), and Leymus komarovii (NC_058744) were obtained from GenBank (Table S1). Plant seeds whose accession numbers start with PI and W6 were obtained from the U.S. National Plant Germplasm System, and those that start with CF were collected by us. The cp genomes of 13 Leymus species were sequenced. Voucher specimens were deposited in the National Medium Term Genebank Forage Germplasm (Hohhot, China) (Supplementary Table S1). Total genomic DNA was isolated from fresh leaves using the CTAB method [25] and sequenced using an Illumina MiSeq platform with PE150 at Sangon Biotech (Shanghai, China).

Genome Assembly and Genome Annotation
Assembly of the cp genome sequences was performed using GetOrganelle and annotated using PGA [26,27] with manual corrections by Geneious v9.0.2 [28]. The circular cp genome map was visualized using Organellar GenomeDRAW [29]. The 13 assembled complete cp genomes were deposited in GenBank (Table S1).

Comparative Genome Analysis
All 13 Leymus cp genomes were aligned using MAFFT v7.313 [32] with default settings. The BRIG [33] and mVISTA [34] software were used to compare the variations among all available Leymus cp genomes, with L. chinensis being set as the reference sequence. In addition, IRScope was used to compare and visualize the junctions and borders of the IR regions [35]. The nucleotide diversity (Pi) rates of sequence divergence between Leymus species were calculated using DnaSP v6.12 [36].

Phylogenetic Analysis
The complete cp genomes and shared protein-coding genes of the 13 newly sequenced Leymus species were used for phylogenetic analysis, along with those published in NCBI, which included those of three previously sequenced Leymus species (Table S1) and the outgroups Brachypodium distachyon (NC_0110320) and Oryza sativa (NC_011032). All datasets were aligned using MAFFT v7.313 [32] with default settings. ModelFinder was employed to find the best model in PhyloSuite v1.2.2 [37]. The best substitution model based on the complete cp genomes was TVM + F + R2 for the maximum likelihood (ML) and GTR + F + I + G4 for the Bayesian inference (BI) analyses, and the best substitution model based on the shared protein-coding genes was GTR + F + R3 for the maximum likelihood (ML) and GTR + F + I + G4 for the Bayesian inference (BI) analyses, which were performed using RAxML v8.2.11 [38] with 1000 non-parametric bootstrap replicates and MrBayes v3.2.6 [39], respectively.

Characteristics of the Leymus Species Complete Chloroplast Genomes
The cp genomes of the Leymus species had the typical quadripartite structure, with the length of the LSC, SSC, and IR regions ranging between 80,149 bp in L. triticoides and 81,053 bp in L. paboanus, 12,708 bp in L. karelinii and 12,797 bp in L. duthiei, and 20,813 bp in L. angustus and 21,578 bp in L. komarovii, respectively (Table S1). The size of the 16 Leymus cp genomes ranged from 135,057 bp of L. condensatus and 136,906 bp of L. coreanus ( Figure 1, Table S1) and comprised 129 genes, including 83 protein-coding genes, 38 transfer RNA (tRNA) genes, and 8 ribosomal RNA (rRNA) genes (Tables 1 and S2). Among these genes, six protein-coding genes, four rRNAs, and nine tRNA were the same in the two identified IR regions; in other words, these 19 genes were duplicated in the two IR regions. The LSC region contained 62 protein-coding genes and 20 tRNA genes, whereas 11 protein-coding genes and one tRNA gene were located in the SSC regions. Among all genes, seven proteincoding genes and six tRNAs were within one intron; additionally, two protein-coding genes (rps12 and ycf3) were within two introns (Tables S2 and S3). In particular, rps12 was trans-spliced, with the 5 -end-exon being located in the LSC region, and two 3 -end-exons being both in the IRs. The GC content of the cp genomes ranged from 38.33-38.42%, which was almost identical to that of all other Leymus species (Table S2), with the IR regions having the highest GC content (43.89-44.01%), followed by the LSC (36.35-36.41%) and SSC regions (32.23-32.63%). Noteworthily, the two IR regions had the same GC content (Table S2).

Analysis of Repeat Sequences and SSRs
All 16 Leymus cp genomes comprised 800 repeats, which consisted of 518 forward (F), 269 palindromic (P), seven reverse (R), and six complementary (C) repeats ( Figure 2). The total number of repeats was consistent for each Leymus species, whereas the repeat type and distribution were different. L. condensatus and L. coreanus contained three repeat types, and L. karelinii and L. racemosus contained four repeat types, whereas all other species contained two repeat types. Among them, F repeats accounted for the biggest proportion (58-70%), followed by P repeats (30-38%), R repeats (L. condensatus, L. coreanus, L. karelinii, and L. racemosus had 4%, 2%, 4%, and 4%, respectively), and C repeats (6% in L. karelinii and L. racemosus) (Figure 2). The length of the repeats in all 16 Leymus cp genomes analyzed ranged from 30 to 286 bp, with most repeats being between 34 and 39 bp. The longest repeats were found in L. duthiei and L. komarovii.

Analysis of Repeat Sequences and SSRs.
All 16 Leymus cp genomes comprised 800 repeats, which consisted of 518 forward (F), 269 palindromic (P), seven reverse (R), and six complementary (C) repeats ( Figure 2). The total number of repeats was consistent for each Leymus species, whereas the repeat type and distribution were different. L. condensatus and L. coreanus contained three repeat types, and L. karelinii and L. racemosus contained four repeat types, whereas all other species contained two repeat types. Among them, F repeats accounted for the biggest proportion (58-70%), followed by P repeats (30-38%), R repeats (L. condensatus, L. coreanus, L. karelinii, and L. racemosus had 4%, 2%, 4%, and 4%, respectively), and C repeats (6% in L. karelinii and L. racemosus) ( Figure 2). The length of the repeats in all 16 Leymus cp genomes analyzed ranged from 30 to 286 bp, with most repeats being between 34 and 39 bp. The longest repeats were found in L. duthiei and L. komarovii. In total, 707 SSRs were identified in the 16 Leymus cp genomes ( Figure 3, Table 1). Mononucleotide repeats were the most abundant SSR among all Leymus species, accounting for 48.72-79.49% of all loci, followed by dinucleotide repeats (7.69-16.67%), trinucleotide repeats (4.88-7.69%), and tetranucleotide repeats (14.63-25.00%) (Table S4). Three pentanucleotide repeats were identified in L. cinereus and L. triticoides, two in L. coreanus, and one in L. angustus, L. condensatus, L. komarovii, and L. mollis. No hexanucleotide repeats were detected in the cp genomes. Most SSRs were composed of an A/T motif instead of a G/C motif. The pentanucleotide AAAAT/ATTTT and AAGAT/ATCTT repeats were identified in L. cinereus and L. triticoides, and the AATAT/ATATT repeat was found only in L. coreanus (Table S4). In total, 707 SSRs were identified in the 16 Leymus cp genomes ( Figure 3, Table 1). Mononucleotide repeats were the most abundant SSR among all Leymus species, accounting for 48.72-79.49% of all loci, followed by dinucleotide repeats (7.69-16.67%), trinucleotide repeats (4.88-7.69%), and tetranucleotide repeats (14.63-25.00%) (Table S4). Three pentanucleotide repeats were identified in L. cinereus and L. triticoides, two in L. coreanus, and one in L. angustus, L. condensatus, L. komarovii, and L. mollis. No hexanucleotide repeats were detected in the cp genomes. Most SSRs were composed of an A/T motif instead of a G/C motif. The pentanucleotide AAAAT/ATTTT and AAGAT/ATCTT repeats were identified in L. cinereus and L. triticoides, and the AATAT/ATATT repeat was found only in L. coreanus (Table S4).  The number of SSRs was not significantly different among the 16 Leymus cp genomes analyzed, ranging from 36 in L. duthiei to 50 in L. coreanus (Table 1). LSC was the region with more SSRs (mean 2.39 SSRs per kb), followed by the IR (mean 8.44 SSRs per kb) and SSC (mean 4.21 SSRs per kb) regions. Only two SSRs were located in the IRs (rrna4.5) of L. angustus, L. condensatus, and L. mollis, and the remaining species had 6-10 SSRs in the IR region ( Figure 3, Table 1). Within the cp genomes of the 16 Leymus species, 56.76% and 67.35% SSR loci were found in the intergenic areas of L. mollis and L. karelinii, respectively. Moreover, the gene-coding regions of L. karelinii and L. racemosus had 20.41% and 27.03% SSR loci, and 11.11% and 16.67% SSR loci were found in the intron areas of L. duthiei and L. chinensis. (Table S5). What is needed to pay attention to is that SSR located in IGS (trnR-UCU, rps14), IGS (atpB, rbcL), IGS (psbE, petL), and IGS (ccsA, ndhD) only remain in Leymus cinereus, Leymus arenarius, Leymus coreanus, and Leymus paboanus (Supplementary Table  S5)

Comparative Analysis of Genome Structure
All aligned cp genome sequences were found to be highly similar, especially in the protein-coding and IR regions (Figures 4 and 5). In this study, the average Pi values for the genome, coding DNA sequence (CDS), and intergenic spacers (IGS) among the 16 Leymus species were 0.0049, 0.0034, and 0.0081, respectively (Tables S6 and S7). Similar to other plants, coding regions were the most conserved among all Leymus species, and conservation of the IGS regions was secondary. In particular, the IR was the most conserved region, with merely a few hotspot regions, with the majority of the hotspot sections located in the LSC and SSC regions. In the 76 CDS regions, Pi% values ranged from 0 (ndhE, petG, petN, psbF, psbI, psbN, psbT, psbZ, rpl36, and rps8) to 0.0159 (rpl32), and the regions with large variations were rps16, ndhH, matK, rpl22, psbM, and rpl32 (Pi% > 0.009) ( Figure  6, Table S6). In the 121 IGS regions, Pi% values ranged from 0 (13 regions, Figure 6, Table  S7) to 0.0431 (trnT-trnL), and the regions with large variations were petG-trnW, rpl32-trnL, rpl22-rps19, and trnT-trnL (Pi% > 0.03). These results indicate that these variable regions can be used as new genetic markers in DNA barcoding and phylogenetic studies of Leymus species. The number of SSRs was not significantly different among the 16 Leymus cp genomes analyzed, ranging from 36 in L. duthiei to 50 in L. coreanus (Table 1). LSC was the region with more SSRs (mean 2.39 SSRs per kb), followed by the IR (mean 8.44 SSRs per kb) and SSC (mean 4.21 SSRs per kb) regions. Only two SSRs were located in the IRs (rrna4.5) of L. angustus, L. condensatus, and L. mollis, and the remaining species had 6-10 SSRs in the IR region ( Figure 3, Table 1). Within the cp genomes of the 16 Leymus species, 56.76% and 67.35% SSR loci were found in the intergenic areas of L. mollis and L. karelinii, respectively. Moreover, the gene-coding regions of L. karelinii and L. racemosus had 20.41% and 27.03% SSR loci, and 11.11% and 16.67% SSR loci were found in the intron areas of L. duthiei and L. chinensis. (Table S5). What is needed to pay attention to is that SSR located in IGS (trnR-UCU, rps14), IGS (atpB, rbcL), IGS (psbE, petL), and IGS (ccsA, ndhD) only remain in Leymus cinereus, Leymus arenarius, Leymus coreanus, and Leymus paboanus (Supplementary  Table S5)

Comparative Analysis of Genome Structure
All aligned cp genome sequences were found to be highly similar, especially in the protein-coding and IR regions (Figures 4 and 5). In this study, the average Pi values for the genome, coding DNA sequence (CDS), and intergenic spacers (IGS) among the 16 Leymus species were 0.0049, 0.0034, and 0.0081, respectively (Tables S6 and S7). Similar to other plants, coding regions were the most conserved among all Leymus species, and conservation of the IGS regions was secondary. In particular, the IR was the most conserved region, with merely a few hotspot regions, with the majority of the hotspot sections located in the LSC and SSC regions. In the 76 CDS regions, Pi% values ranged from 0 (ndhE, petG, petN, psbF, psbI, psbN, psbT, psbZ, rpl36, and rps8) to 0.0159 (rpl32), and the regions with large variations were rps16, ndhH, matK, rpl22, psbM, and rpl32 (Pi% > 0.009) ( Figure 6, Table S6). In the 121 IGS regions, Pi% values ranged from 0 (13 regions, Figure 6, Table S7) to 0.0431 (trnT-trnL), and the regions with large variations were petG-trnW, rpl32-trnL, rpl22-rps19, and trnT-trnL (Pi% > 0.03). These results indicate that these variable regions can be used as new genetic markers in DNA barcoding and phylogenetic studies of Leymus species. Genes 2022, 13, x FOR PEER REVIEW 7 of 15

Expansion and Contraction of the IR Region
The expansion or contraction of the IR region differs among plant species; therefore, LSC, SSC, and IR boundary structures were analyzed among the 16 Leymus species. Only minor differences in the junction positions were detected. In all species, the junction of LSC/IRb (JLB) was located between rp122 in LSC and rps19 in Irb (Figure 7): rp122 extended from 27 to 40 bp in length from LSC to Irb, and rps19 extended from 37 to 57 bp in length from Irb to LSC. Similarly, the JLAs (junction of Ira/LSC) were located between rps19 in Ira and psbA in LSC: rps19 extended from 37 to 57 bp in length from Ira to LSC, and psbA extended from 81 to 91 bp in length from LSC to Ira. The ndhF gene deviated from the junction of the Irb/SSC (JSB) region, ranging from 68 to 108 bp in length. The rps15 gene was located in the Irb, which ranged from 337 to 362 bp in length from the JSB border. Additionally, the ndhH gene of all Leymus species straddled the junction of the SSC/Ira (JSA) boundary regions, with 975 to 900 bp of the SSC region and 192 to 207 bp of the Ira region. Overall, the contraction and expansion events in Irs were reflected in the different locations of rpl22, rps19, ndhH, and psbA (Figure 7).     rps15 gene was located in the Irb, which ranged from 337 to 362 bp in length from the JSB border. Additionally, the ndhH gene of all Leymus species straddled the junction of the SSC/Ira (JSA) boundary regions, with 975 to 900 bp of the SSC region and 192 to 207 bp of the Ira region. Overall, the contraction and expansion events in Irs were reflected in the different locations of rpl22, rps19, ndhH, and psbA (Figure 7).

Phylogenetic Relationships between Leymus Species
Using Oryza sativa and Brachypodium distachyon as outgroups, cp complete genome and 74 shared protein-coding genes were used to perform phylogenetic analysis, respectively.
The topological structure of the phylogenetic tree showed that 16 Leymus species were clustered in different clades with other 27 species across 11 genera of Triticum. In addition to L. condensatus, L. angustus, and L. mollis, the 13 Leymus species herein analyzed clustered along with two Psathyrostachys species in two clades. Two North American Leymus species (L. triticoides and L. cinereus) along with two Central Asia Leymus species (L. coreanus and L. komarovii) clustered into one clade and showed a close phylogenetic relationship. The other clade consisted of five Central Asia Leymus species (L. racemosus, L. karelinii, L. multicaulis, L. ramosus, and L. paboanus), one Eurasia species (L. arenarius), one East Asia species (L. chinensis), two Qinghai-Tibet Plateau Leymus species (L. secalinus and L. duthiei), and the two Psathyrostachys species (P. juncea and P. huashanica) (Figure 8).

Phylogenetic Relationships between Leymus Species
Using Oryza sativa and Brachypodium distachyon as outgroups, cp complete genome and 74 shared protein-coding genes were used to perform phylogenetic analysis, respectively. The topological structure of the phylogenetic tree showed that 16 Leymus species were clustered in different clades with other 27 species across 11 genera of Triticum. In addition to L. condensatus, L. angustus, and L. mollis, the 13 Leymus species herein analyzed clustered along with two Psathyrostachys species in two clades. Two North American Leymus species (L. triticoides and L. cinereus) along with two Central Asia Leymus species (L. coreanus and L. komarovii) clustered into one clade and showed a close phylogenetic relationship. The other clade consisted of five Central Asia Leymus species (L. racemosus, L. karelinii, L. multicaulis, L. ramosus, and L. paboanus), one Eurasia species (L. arenarius), one East Asia species (L. chinensis), two Qinghai-Tibet Plateau Leymus species (L. secalinus and L. duthiei), and the two Psathyrostachys species (P. juncea and P. huashanica) (Figure 8).

Genome Organization and Genome Features
Leymus is a genus of the botanical tribe Triticeae (Poaceae) comprising approximately 30 species and 19 subspecies, almost all of which are polyploid and cross-pollinating [1,2]. Leymus can be used for agriculture and conservation, and as an important genetic resource for Triticeae cereal crop improvement. In this study, 13 Leymus species were sequenced and comprehensively analyzed in comparison with three previously reported Leymus sequences. The cp genomes of all 16 Leymus species exhibited the typical quadripartite structure, which consisted of LSC, SSC, and two inverted repeats (IRa and IRb), and the cp genome size ranged from 135,057 bp (L. condensatus) to 136,906 bp (L. coreanus) (Figure 1, Table S2), which was in agreement with previously published data of Triticeae [23,40,41]. The cp genomes of Leymus species were highly conserved, with the same genomic structure and gene content. In particular, accD, ycf1, and ycf2 genes coded in the Leymus cp genomes gradually degraded as compared to those of other plants, such as cattail (Typha orientalis Presl) and tobacco (Nicotiana tabacum L.) [42], and were similar to that of Triticeae

Genome Organization and Genome Features
Leymus is a genus of the botanical tribe Triticeae (Poaceae) comprising approximately 30 species and 19 subspecies, almost all of which are polyploid and cross-pollinating [1,2]. Leymus can be used for agriculture and conservation, and as an important genetic resource for Triticeae cereal crop improvement. In this study, 13 Leymus species were sequenced and comprehensively analyzed in comparison with three previously reported Leymus sequences. The cp genomes of all 16 Leymus species exhibited the typical quadripartite structure, which consisted of LSC, SSC, and two inverted repeats (IRa and IRb), and the cp genome size ranged from 135,057 bp (L. condensatus) to 136,906 bp (L. coreanus) (Figure 1, Table S2), which was in agreement with previously published data of Triticeae [23,40,41]. The cp genomes of Leymus species were highly conserved, with the same genomic structure and gene content. In particular, accD, ycf1, and ycf2 genes coded in the Leymus cp genomes gradually degraded as compared to those of other plants, such as cattail (Typha orientalis Presl) and tobacco (Nicotiana tabacum L.) [42], and were similar to that of Triticeae species [23,43]. This event may be explained by the insertion and deletion of the no-triple base in the pairing sequence which may have occurred many times, causing a frameshift mutation and gene loss [42]. Moreover, since 20 genes with introns were identified in Leymus species, contrasting with clpP and rpoC1 which had no introns, provides further evidence that the loss of clpP and rpoC1 introns occurs in most Poaceae species [23,43,44].
SSRs, which are highly diverse in copy number, are important molecular markers that play a significant role in plant population genetics and evolution investigations [45,46]. A total of 707 SSR loci were identified in the 16 Leymus cp genomes ( Figure 3, Table 1). Due to the rich A/T stretches of the repeats, the cp genome contained a high A/T content (Table S4). It is observed that the number of SSRs in the non-coding region was higher than that in the coding region (Table S5). These results are similar to those of other Triticeae species [23,47,48] and may be the reason why the mutation rate in IGS regions is higher than in the coding regions [49].

Evolution and Origin of IRs
The IR regions of the cp genome are considered to be the most conserved regions [55]. Similar to previous studies on Triticeae, all 16 Leymus species had the same gene arrangement, in which six protein-coding genes (rpl2, rpl23, rps12, ndhB, rps15, and rps19) were duplicated [23,43,47]. Moreover, we identified the smallest reverse IR region in L. angustus (20,813 bp) and the largest in L. komarovii (21,578 bp) (Table S3). It was possibly the expansion and contraction of the IR regions, along with the spacer of a single copy, that caused variations in the length of the complete cp genomes [56]. Some reports have shown that slight expansion of the IR region may be caused by gene transfer, with double-strand break repair leading to wider expansion [57]. The different locations of rpl22, rps19, ndhH, and psbA may have been induced by IR contraction and expansion events in Leymus species, leading to changes in the IR boundaries among them. All Leymus species showed minor variations in junction positions of JLB (~13 bp), JSB (~34 bp), and JLA (~8 bp) (Figure 7). In addition, ndhH was found to straddle the JSA, which is consistent with previous data on other Triticeae species, such as Agropyron, Hordeum, and Festuca [23,47,58,59]. The different position of ndhH in the JSA border suggests that these variations may be caused by intramolecular recombination during early evolution [60][61][62].

Phylogenetic Relationships
Leymus has been widely recognized based on morphological, cytogenetic, and molecular data after Hochstetter separated it from the traditional Elymus L. in 1848; however, there is currently a lack of understanding of the evolutionary relationships among Leymus species. In this study, the whole cp genome sequences were used to perform phylogenetic analysis. Overall, the 16 Leymus species analyzed were clustered in different clades along with other 27 species across 11 genera of Triticum. The evolutionary relationships herein identified were consistent with those of previous studies based on nuclear DNA markers and, to some extent, cp genome data [11,13,14,24,[51][52][53]63,64]. According to Sha's research, Psathyrostachys is the donor of the Ns genome in Leymus species [14]. What needs more attention is that the phylogenetic tree based on cp genome sequences existed with minor differences from the phylogenetic tree based on shared protein-coding genes. We inferred that the reason for this is barcode genes laid in the cp genome, and the result is consistent with Liu's and Guo's research [11,16]. What is more, the close relationship between Elymus and Leymus may also be related to their origin of genome [11]. Here, we provide further evidence based on plastome data and protein-coding genes. The results of our phylogenetic analysis indicate that the cp genome can be used as a useful tool to explore species relationships.

Conclusions
Based on 13 newly sequenced and three previously reported Leymus cp genomes, we have found that the plastome is highly conserved and similar concerning its overall structure within this genus. Our findings demonstrate that most SSRs are A/T rich and are more commonly located in non-coding regions. The cp genome sequence identity is highly similar among Leymus species, especially in the protein-coding and IR regions. The differences identified in the CDS and IGS regions of Leymus species revealed 10 highly variable regions that may be used as new genetic markers for DNA barcoding and phylogeny research. In addition, comparisons of the IR boundaries further showed that the IR contraction and expansion events in Leymus species lead to different genomic locations of the rpl22, rps19, ndhH, and psbA genes. The close relationship between Leymus and Psathyrostachys highlighted by a phylogenetic tree design based on plastome data further indicated that as early as in the 1980s, Psathyrostachys may be the donor of the Ns genome to Leymus species. Taken together, these findings provide useful information for the identification and phylogenetic analysis of Leymus species.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/genes13081425/s1, Table S1: The information of Leymus species in this study; Table S2: Gene composition in Leymus chloroplast genome; Table S3: Summary of chloroplast genome characteristics in Leymus species; Table S4: Numbers of identified SSRs motifs in the 16 Leymus chloroplast genomes; Table S5: SSRs location in the 16 Leymus chloroplast genomes; Table S6: Eta, Pi value, and PICs of Protein-coding regions(CDS) in Leymus species; Table S7: Eta, Pi value, and PICs of non-coding regions in Leymus species.
Author Contributions: Z.W. designed the experiment, carried out the bioinformatic analyses, and wrote the first manuscript. C.T. collected the plant materials and manuscript revision. Y.Y. participated in carrying out some experiments. Q.L. and Y.L. contributed to the result interpretation and manuscript revision. Z.L. and K.J. conceived the experiments and revised the manuscript. All authors have read and agreed to the published version of the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.