Comparative Analysis of Six Chloroplast Genomes in Chenopodium and Its Related Genera (Amaranthaceae): New Insights into Phylogenetic Relationships and the Development of Species-Specific Molecular Markers

Abstract

Species within the genus Chenopodium hold significant research interest due to their nutritional richness and salt tolerance. However, the morphological similarities among closely related species and a dearth of genomic resources have impeded their comprehensive study and utilization. In the present research, we conduct the sequencing and assembly of chloroplast (cp) genomes from six Chenopodium and related species, five of which were sequenced for the first time. These genomes ranged in length from 151,850 to 152,215 base pairs, showcased typical quadripartite structures, and encoded 85 protein-coding genes (PCGs), 1 pseudogene, 37 tRNA genes, and 8 rRNA genes. Compared with the previously published sequences of related species, these cp genomes are relatively conservative, but there are also some interspecific differences, such as inversion and IR region contraction. We discerned 929 simple sequence repeats (SSRs) and a series of highly variable regions across 16 related species, predominantly situated in the intergenic spacer (IGS) region and introns. The phylogenetic evaluations revealed that Chenopodium is more closely related to genera such as Atriplex, Beta, Dysphania, and Oxybase than to other members of the Amaranthaceae family. These lineages shared a common ancestor approximately 60.80 million years ago, after which they diverged into distinct genera. Based on InDels and SNPs between species, we designed 12 pairs of primers for species identification, and experiments confirmed that they could completely distinguish 10 related species.

1. Introduction

In the most recent APG IV (Angiosperm Phylogeny Group) plant classification system, the genus Chenopodium is categorized under the Amaranthaceae family, which encompasses approximately 250 species globally [1,2]. Chenopodium plants are mostly annual or perennial herbs, rarely semi-shrubs, with cystic hairs (powder) or cylindrical hairs, a small quantity of glandular hairs or completely hairless, and very low odor. They generally grow in arid or semi-arid areas and have tolerance to multiple stresses, including a strong salt tolerance. These characteristics make Chenopodium plants have extremely high scientific research value [3,4,5,6]. Chenopodium plants also have a high nutritional value. They can not only be grown as pseudocereals but also eaten as leafy vegetables [7]. Among them, the leaves of C. album are an important source of vitamins and trace elements in certain areas [8]. C. quinoa, known as a “superfood”, is not only rich in fatty acids, vitamins, minerals, dietary fiber, and proteins, containing many amino acids, but also rich in carotenoids, vitamin C, and phenolic compounds. These compounds contribute to the prevention of conditions such as allergies and cardiovascular diseases [9,10,11,12,13]. In some arid areas, they are also used as medicinal plants, fodder, and dye raw materials [14].

In addition, some plants of the related genera of Chenopodium also have various utilization values. For example, some plants from Atriplex can revegetate arsenic-contaminated saline–alkali soils, while effectively accumulating and preventing the spread of arsenic and reducing the risk of human exposure to arsenic [15]. Some compounds extracted from Atriplex can protect the chromosome damage caused by gamma radiation and can be used to develop new radioprotective drugs [16]. D. ambrosioides exhibits high antimicrobial and antioxidant activities and can enhance the antibacterial activity with the synergistic effect of antibiotics. At the same time, some compounds contained in it can achieve the purpose of killing mosquitoes by destroying the nervous system of insects and inhibiting the activity of key enzymes [17,18,19].

However, due to the high similarity and polymorphism of leaf shape, floral structure, and seed morphology and the existence of possible hybrid species, the classification and identification of Chenopodium and its related genera have become extremely complex and controversial, which severely hinders the development of related scientific research and the efficient utilization of Chenopodium plants [20,21,22,23].

Previous investigations in Chenopodium primarily relied on conventional molecular markers for phylogenetic analysis and species identification. These markers include nuclear DNA amplification fragment length polymorphism, ribosomal DNA (rDNA) internally transcribed spacer sequence (ITS) variation, and chloroplast DNA (cpDNA) fragments [24,25,26,27,28]. However, these research methods have limitations due to the restricted information they provide regarding mutation sites. Consequently, these approaches may not fully address the requirements of related research [29].

Several scholars have suggested employing the complete cp genome as a prospective solution to address this issue [30,31,32,33]. The cp genome of closely related species is highly conserved in structure, yet it exhibits subtle differences. Furthermore, the longer sequences provide a wealth of mutation site information, thereby enhancing its capability for species identification [34,35,36,37]. Simultaneously, the cp genome of angiosperms is mostly maternally inherited, reducing gene pollution and ensuring the accurate transmission and storage of genetic information [38,39]. Some researchers have used the cp genome to mine mutation sites to distinguish closely related species (especially some weed relatives), or have used the cp genome to elaborate new perspectives on plant phylogenetic relationships and taxonomy. The importance and efficiency of chloroplast genomes in related research have been confirmed by more and more studies [40,41].

In this study, we obtain the complete cp genomes of six closely related species of Chenopodium and describe their basic characteristics in detail. By comparing with the sequence information of other species that have been published, we studied their phylogenetic relationship and divergence time, which provided valuable clues to understand the evolutionary history of the genus Chenopodium. In addition, we also developed a set of highly efficient molecular marker primers that can be used to accurately identify these closely related species; these molecular markers could be widely applied to assist in screening and identifying Chenopodium plants with specific genomic features, thereby accelerating the breeding process for new varieties of crops, such as C. quinoa.

2. Materials and Methods

2.1. Plant Material and DNA Isolation

We collected the young leaves of 6 plant species (A. patens, A. prostrata, C. album, C. berlandieri, C. bryoniifolium, and L. franchetii) from Yantai, Shandong Province, China (121°44′ E, 37°51’ N, 264,003). The identification of samples was based on the classification and external morphological diagnosis of the samples according to the Chinese Virtual Herbarium (CVH) and Flora of China [42]. The specimen voucher materials were preserved in the College of Life Sciences, Yantai University. The total DNA was obtained from plants through the modified CTAB method [43]. The quality of DNA was assessed using a NanoDrop spectrophotometer 2000 (Thermo Scientific, Waltham, MA, USA), and DNA integrity was verified through 1% gel electrophoresis.

2.2. DNA Sequencing, Assembly, and Annotation

To create libraries, 350 bp fragments were generated using paired-end sequencing on the Illumina HiSeq 2500 platform (Illumina Inc., San Diego, CA, USA).

To obtain clean data, Fastp 0.23.1 was used to remove ambiguous bases or low-quality bases from raw sequencing reads [44]. GetOrganelle v1.7.1 was utilized for the assembly of the cp genome, with the k-mer length parameter configured as follows: −k 21, 45, 65, 85, and 105 [45]. Bandage v0.81 and Gepard v1.40 were used to examine the cp gene composition loop and collinearity, respectively [46,47].

The cp genome was annotated by employing CPGAVAS2 and GeSeq. Subsequently, Apollo was used for proofreading [48,49,50]. To locate and annotate tRNA genes, tRNAscan-SE v2.0.7 was employed for analysis, and CPgview was employed to check the annotation of the rps12 gene [51,52]. Finally, the complete cp genome map was generated through Chloroplot [53]. The annotated cp genome sequence was stored in GenBank with the accession number OR374019-OR374024.

2.3. Genome Structure and Comparison Analysis

Combined with the published sequence information in the NCBI database, a sum of 16 complete cp genomes of Chenopodium and related species were used for comparative analysis. The analysis of SSRs was conducted using MISA v2.1 [54]. The parameters were set as follows: mononucleotides ≥ 10; dinucleotides ≥ 5; trinucleotides ≥ 4; and tetranucleotides, pentanucleotides, and hexanucleotides ≥ 3. REPuter was used to detect the four types of interspersed repeats, including forward repeats, palindromic repeats, reverse repeats, and complementary repeats [55]. The minimum repeat fragment size was set to 30 bp, and the Hamming distance was set to 3. The relative synonymous codon usage (RSCU) was calculated by CodonW v1.4.2, and the results were subsequently visualized by Tbtools v1.113 [56,57].

The Perl script CPJSdraw v1.0 was employed to compare and visualize the inverted repeat (IR) region boundary characteristics between the cp genomes of the 16 species of Chenopodium and its related genera [58]. Using A. centralasiatica as the reference genome, in the Shuffle-LAGAN model of mVISTA, the cp genomes were compared to understand the sequence variations among closely related species [59,60]. DnaSP v5.10 was employed to analyze the nucleotide diversity of 111 gene regions and 135 intergenic regions (including introns) shared by all closely related species to obtain high information content sites [61].

2.4. Phylogenomic Analysis and Divergence Time Estimation

For the purpose of acquiring an in-depth understanding of the phylogenetic relationship among the species of Chenopodium and its related genera, data were downloaded from other species under the Amaranthaceae from the NCBI database as a supplement. In total, 155 complete cp genome sequences from 69 species in 24 genera were obtained, and their sequence starting points and small single copy (SSC) region directions were corrected and unified. Among them, five species of Plumbaginaceae were employed as outgroups. The alignment of all cp genome sequences was performed using MAFFT v7.50 and then trimmed using trimAl v1.41; the phylogenetic relationships were inferred based on the maximum likelihood (ML) and Bayesian inference (BI) [62,63]. The best nucleic acid replacement model was selected using Model Finder, and the ML tree was reconstructed using IQ-TREE 2.03 under the TVM + F + R3 model [64,65]. The ML branch support was evaluated by 1000 ultrafast guided repetitions.

The BI tree was reconstructed using MrBayes 3.2.7 [66]. A total of 20 million generations were iterated with the GTR + F + I + G4 model, with one cold chain and three heated chains, and the initial 25% of the sampled data was discarded as burn-in. Finally, the average standard deviation of split frequencies (ASDSF) < 0.01, the effective sample size was >200, and the potential scale reduction factor (PSRF) was close to 1, reaching convergence.

Based on the ML tree, the divergence time of the 16 Chenopodium related species (each species retains only one sequence) in Clade I was calculated using the Markov chain Monte Carlo (MCMC) tree program in PAML [67]. According to previous studies, five fossil calibration points were used to limit each node, of which 81.3–97.0 Ma was used as the age of the root node [68]. The independent rates that follow the lognormal distribution were selected as the clock model. We selected HKY85 as the nucleic acid replacement model, in which the alpha for gamma rates at the sites was set to 0.5. The birth–death process was used to generate uniform node age priors in the tree, using the default parameters (l = 1, m = 1, and s = 0.1). We ran the Markov chain Monte Carlo program for 200 million generations, sampling every 100 cycles. The first 25% of the samples was discarded as burn-in. Finally, Tracer v1.72 was used to check the output file to confirm the convergence based on ESS > 200.

2.5. Development and Validation of InDel and dCAPS Markers

Based on the variable regions in the cp genomes of Chenopodium and its related species, a series of molecular markers were developed to distinguish a total of 10 related species. Primer3 4.1.0 and dCAPS Finder 2.0 were used to design primers for the amplification of InDel markers and SNP-based dCAPS markers in the conserved regions near the variant sites [69,70].

Polymerase chain reaction (PCR) was performed in a total volume of 10 μg. It contained 5 μg 2 × Taq Master Mix (Vazyme, Nanjing, China), 1 μg DNA template, 0.5 μg of upstream and downstream primers, and 3 μg of ddH2O. The PCR was executed with a T100TM thermal cycler (Bio-Rad, Hercules, CA, USA). The settings were established as: pre-denaturation at 95 °C for 5 min, followed by 35 cycles of denaturation at 95 °C for 15 s, annealing at 55–58 °C for 15 s, extension at 72 °C for 1–3 min, and finally extension at 72 °C for 7 min. The PCR products were electrophoresed in 2–3% agarose gel, and the PCR products of the dCAPS marker were first reacted with the restriction enzyme (BsrI) at 65 °C for 1 h. Finally, Gel DocTM XR + gel imager (Bio-Rad, Hercules, CA, USA) was used for visualization.

3. Results

3.1. Genome Structure

We sequenced and assembled the cp genomes of six Chenopodium species and related genera, and then compared these to the existing cp genomes of the associated genera in the NCBI database. The lengths of these newly sequenced cp genomes varied from 151,850 bp (A. patens) to 152,215 bp (A. prostrata). Both Chenopodium and Atriplex cp genomes presented a typical quadripartite structure (Figure 1). The IR region (varying from 25,161 bp (A. patens) to 25,192 bp (C. bryoniifolium)) were separated by a large single copy (LSC) region (ranging from 83,675 bp (C. bryoniifolium) to 83,887 bp (A. prostrata)) and an SSC region (ranging from 17,811 bp (A. patens) to 18,131 bp (C. berlandieri)). The overall GC content in the cp genomes of Chenopodium and Atriplex was found to range between 37.25% and 37.3%. Nevertheless, there was a non-uniform distribution of the GC content in distinct genomic regions. Precisely, the GC content within the IR region (42.72–42.81%) exceeded that of both the LSC (35.30–35.40%) and the SSC regions (30.82–31.02%), which is consistent with other Chenopodium and related species previously reported [71,72]. Several studies proposed that the elevated GC content observed in the IR region possibly stems from the higher prevalence of GC-rich genes, such as rRNA and tRNA, within these genomic sections [73,74,75].

The cp genomes of Chenopodium and Atriplex encode 131 genes (totaling 113 unique genes), including 85 protein-coding genes (PCGs), 1 pseudogene (a shorter incomplete copy of ycf1), 37 transfer RNA (tRNA) genes, and 8 ribosomal RNA (rRNA) genes. The order and orientation of these genes were the same in all species, which is in accordance with the outcomes documented in the related genera, such as Beta, Dysphania, and Oxybasis (

Table 1. Summary of the cp genomes for the 16 Chenopodium and related species.

Accession	Species	Total	LSC	IR	SSC	Total	LSC	IR	SSC	Total	Protein-Coding	rRNA	tRNA
Number		Length (bp)	(bp)	(bp)	(bp)	GC Content (%)	GC Content (%)	GC Content (%)	GC Content (%)	Genes	Genes	Genes	Genes
NC_045304	A. centralasiatica	152,237	83,721	25,209	18,095	37.27	35.38	42.72	30.85	131	86	8	37
NC_059062	A. gmelinii	151,852	83,716	25,161	17,811	37.29	35.35	42.81	30.86	131	86	8	37
OR374023	A. patens *	151,850	83,714	25,161	17,811	37.29	35.35	42.81	30.86	131	86	8	37
OR374024	A. prostrata *	152,215	83,887	25,186	17,953	37.3	35.4	42.77	30.82	131	86	8	37
ON641300	B. vulgaris	149,724	83,107	24,409	17,796	37	34.97	42.84	30.44	131	86	8	37
NC_054154	C. acuminatum	152,200	83,683	25,192	18,130	37.25	35.3	42.72	31.01	131	86	8	37
OR374020	C. album *	152,166	83,675	25,192	18,104	37.25	35.3	42.72	30.99	131	86	8	37
OR374022	C. berlandieri *	152,194	83,676	25,192	18,131	37.25	35.3	42.72	31.02	131	86	8	37
OR374021	C. bryoniifolium *	152,166	83,675	25,192	18,104	37.25	35.3	42.72	30.99	131	86	8	37
NC_041200	C. ficifolium	151,823	83,668	25,108	17,936	37.26	35.31	42.75	31	131	86	8	37
OQ957163	C. petiolare	152,064	83,520	25,217	18,107	37.24	35.28	42.76	30.92	131	86	8	37
NC_034949	C. quinoa	152,099	83,582	25,204	18,106	37.24	35.29	42.74	30.96	131	86	8	37
NC_041201	D. ambrosioides	151,689	83,421	25,102	18,061	36.92	34.86	42.74	30.29	131	86	8	37
NC_042166	D. botrys	152,055	83,769	25,184	17,915	36.84	34.74	42.72	30.14	131	86	8	37
NC_041159	D. pumilio	151,962	83,758	25,230	17,741	36.95	34.84	42.74	30.44	131	86	8	37
NC_047226	O. glauca	151,655	83,579	25,080	17,913	36.87	34.73	42.77	30.35	131	86	8	37

* Represents the newly obtained chloroplast genome data in this study.

). Among the PCGs, 48 genes were associated with photosynthesis, while 75 genes were involved in transcription and translation processes (

Table 2. Gene composition of the cp genome in Chenopodium and its related species.

Category	Group of Genes	Genes
Photosynthesis	Photosystem I	psaA, psaB, psaC, psaI, psaJ
Related genes	Photosystem II	psbA, psbB, psbC, psbD, psbE, psbF, psbH, psbI, psbJ, psbK,
		psbL, psbM, psbN, psbT, psbZ
	NADH oxidoreductase	ndhA *, ndhB *#, ndhC, ndhD, ndhE, ndhF, ndhG, ndhH,
		ndhI, ndhJ, ndhK
	Cytochrome b6/f complex	petA, petB *, petD *, petG, petL, petN
	ATP synthase	atpA, atpB, atpE, atpF *, atpH, atpI
	Rubisco	rbcL
	Assembly/stability	ycf3 **, ycf4
	Cytochrome c synthesis	ccsA
Transcription and	Large subunit ribosomal proteins	rpl14, rpl16 *, rpl2 #, rpl20, rpl22, rpl23 #, rpl32, rpl33, rpl36
Translation-related genes	Small subunit ribosomal proteins	rps11, rps12 **#, rps14, rps15, rps16 *, rps18, rps19, rps2, rps3,
		rps4, rps7 #, rps8
	Transcription	rpoA, rpoB, rpoC1 *, rpoC2
	Translation initiation factor	infA
	Ribosomal RNAs	rrn16 #, rrn23 #, rrn4.5 #, rrn5 #
	Transfer RNAs	trnA-UGC *#, trnC-GCA, trnD-GUC, trnE-UUC, trnF-GAA,
		trnG-GCC, trnG-UCC *, trnH-GUG, trnI-CAU #, trnI-GAU *#,
		trnK-UUU *, trnL-CAA #, trnL-UAA *, trnL-UAG, trnM-CAU,
		trnN-GUU #, trnP-UGG, trnQ-UUG, trnR-ACG #, trnR-UCU,
		trnS-GCU, trnS-GGA, trnS-UGA, trnT-GGU, trnT-UGU,
		trnV-GAC #, trnV-UAC *, trnW-CCA, trnY-GUA, trnfM-CAU
Other genes	RNA processing	matK
	Proteolysis	clpP **
	Carbon metabolism	cemA
	Fatty acid synthesis	accD
	Proteins with unknown functions	ycf1 #, ycf2 #

Gene *: Gene containing one intron; Gene **: Gene containing two introns; Gene #: Two gene copies in the IRs (ycf1 located at the boundary of SSC and IRb is a pseudogene).

). In total, 17 genes contained one intron (rps16, atpF, rpoC1, petB, petD, rpl16, ndhB (×2), ndhA, trnK-UUU, trnG-UCC, trnL-UAA, trnI-GAU (×2), trnA-UGC (×2), and trnV-UAC) and 3 genes contained two introns (ycf3, clpP, and rps12). Additionally, rps12 underwent trans-splicing, involving the concatenation of two exons. The 5′ exon is situated within the LSC region, while the 3′ exon is positioned in the IR region. A total of 17 genes, including 6 PCGs (rpl2, rpl23, ycf2, ndhB, rps7, and rps12), 7 tRNA genes (trnI, trnL, trnV, trnI, trnA, trnR, and trnN), and all 4 rRNA genes (rrn16, rrn23, rrn4.5, and rrn5), were found to be duplicated in the inverted repeat (IR) region.

3.2. Repeat Sequence Analysis

SSRs with 1–6 bp repeat nucleotide units are widely distributed in plant cp genomes [76]. The number of SSRs in the 16 species of Chenopodium and related genera ranged from 44 (C. quinoa and C. petiolare) to 74 (A. centralasiatica). Except for pentanucleotide SSRs, which were not found in A. prostrata, A. patens, and A. gmelinii, all species contain mononucleotide, dinucleotide, trinucleotide, tetranucleotide, and pentanucleotide SSRs. However, hexanucleotide SSRs only exist in A. centralasiatica. Among all SSRs, A/T and AT/AT were the most abundant, accounting for 65.30% and 7.53% respectively. Research has demonstrated that a high AT content in the cp genome of higher plants is a prevalent phenomenon [77,78,79]. SSRs were predominantly located within the intergenic spacer (IGS) of the LSC and SSC regions, with a smaller fraction being distributed within the coding regions of ycf1, ycf2, and rrn23 in the IR region (Figure 2).

Within the cp genomes of 16 species, the count of interspersed repeats varied between 36 (D.ambrosioides) and 87 (O.glauca). Forward and palindrome repeats were identified across all species, whereas reverse and complementary repeats were exclusively observed in C. quinoa and D. ambrosioides, respectively. The majority of the interspersed repeats were forward repeats (n = 14–41) and palindrome repeats (n = 21–46), accounting for 47.62% and 52.15%, respectively, and the size was mainly concentrated in 30–39 bp (Figure 3).

3.3. Analysis of Codon Usage Bias

The gene coding sequences in the 16 closely related Chenopodium species consist of 64 codons, with 61 encoding 20 amino acids, while the remaining 3 (UAA, UAG, and UGA) serve as stop codons. Leucine exhibits the highest frequency, while cysteine demonstrates the lowest frequency among all amino acids. The RSCU indicates the codon usage frequency relative to the expected frequency [80]. In this study, 30 codons exhibited RSCU values exceeding 1, indicating values higher than expected, whereas 32 codons showed RSCU values below 1, signifying values lower than expected. Furthermore, the RSCU values of the remaining two codons, AUG and UGG, equaled 1. In addition, the bases of the high-frequency codons in the third position showed a strong A/T preference (Figure 4).

3.4. Genome Comparison and Nucleotide Diversity

In the field of plant evolution research, the dynamic alteration of the IR region within the cp genome stands as a pivotal event. It holds the potential to induce variations in both the size and genetic composition of the cp genome [81]. The comparative analysis of seven species of Chenopodium and nine species of related genera showed that, except for the obvious contraction of the IR region of Beta, the cp genome structure and gene sequence were substantially conserved within and among the genera, except for a subtle variation in the IR/SC region junctions. The LSC/IRb (JLB) junctions of all 16 species are located within rps19, but the part of rps19 located in the IRb region is less in Chenopodium (79 bp), while all other closely related species, except for D. botrys (79 bp), are more (142–156 bp). The IRb/SSC (JSB) junctions are located within ndhF and range from 2198 to 2229 bp, except for D. ambrosioides (811 bp from ndhF). The SSC/IRa (JSA) junction is located at 1443–1481 bp inside ycf1. The IRa/LSC (JLA) junction is located between rpl2 and trnH and is 0–4 bp away from trnH (Figure 5).

Taking A. centralasiatica as a reference, the cp genomes of the 16 species of Chenopodium and related genera were compared and analyzed (Figure 6). The cp genome is extremely conserved within the same plant genus, yet intergeneric differences persist. On the whole, the variation level of IGS and intron regions is higher than that of exon regions. A nucleotide diversity (Pi) analysis was further performed on 42 cp genomes of the above 16 species (Figure 7). The results show that the IGS region (average Pi = 0.04430) had higher polymorphism than the gene region (average Pi = 0.01810). The most variable gene regions included: matK, accD, psaI, rps11, rps8, rpl22, ycf1, ndhF, ccsA, and ndhD (PI > 0.035), six of which were situated in the LSC region, while four were positioned in the SSC region. The IGS regions with a large variation include: trnH-GUG-psbA, psbI-trnS-GCU, trnS-GCU-trnG-UCC, psbM-trnD-GUC, rps4-trnT-UGU, ycf4-cemA, rpl36-infA, rpl22-rps19, ndhF-rpl32, rpl32-trnL-UAG, ccsA-ndhD, and ndhG-ndhI (Pi > 0.080). Among them, eight were situated in the LSC region and four in the SSC region.

3.5. Phylogenomic Analysis and Divergence Time Estimation

The cp genome plays a pivotal role in revealing the evolutionary relationships among plants [82]. Taking several species of Limonium as the outgroup, we reconstructed the phylogenetic trees of 63 species of Amaranthaceae, including Chenopodium (Table S2). The maximum likelihood (ML) trees and Bayesian inference (BI) trees had almost the same topological structure, with a strong bootstrap support (BS) and posterior probabilities (Figure 8). In the ingroup, all species were classified into three main clades (clade I, clade II, and clade III). Based on morphological studies (Flora of China), all species in clade I belong to Subfam. Amaranthaceae, while all species in clade II and clade III belong to Subfam. Chenopodiaceae. Among them, all the species in clade II are distributed in three tribes (Trib. Atripliceae, Trib. Beteae, and Trib. Chenopodieae), while the species in clade III are distributed in four other tribes (Trib. Camphorosmeae, Trib. Salicornieae, Trib. Salsoleae, and Trib. Suaedeae). In addition, all the species of Chenopodium and its four related genera (Atriplex, Beta, Dysphania, and Oxybasis) in clade II were clustered together as genera, forming five monophyletic groups. This result further supports the inference of other studies that Chenopodium is a monophyletic group [7].

Based on the ML tree, we estimated the divergence time of the 16 species in clade I (Figure 9). The outcomes demonstrate that the 16 species of Chenopodium and related genera had a common ancestor in the Paleocene period at 60.80 Mya (95% highest posterior densities (HPD): 56.27–64.52 Mya). At the same time, Beta also split from other genera. After that, Dysphania and Oxybasis split from other species at 51.47 Mya (95% HPD: 49.91–52.51 Mya) and 47.63 Mya (95% HPD: 46.54–49.17 Mya), respectively. Atriplex was the closest relative to Chenopodium in all related genera; they were sister clades and split at 30.28 Mya (95% HPD: 28.22–33.62 Mya). The seven species of Chenopodium began to split in the Miocene period at about 9.13 Mya (95% HPD: 4.61–14.89 Mya) and separated into two branches represented by C. quinoa and C. album. C. album and C. bryoniifolium are two species that were extremely difficult to distinguish in morphological taxonomy. Their split occurred 20,000 years ago (95% HPD: 0–0.06 Mya) and is also the most recent.

3.6. Development and Validation of the InDel and dCAPS Markers

Based on the interspecific analysis, we designed a series of primers on the conserved region near the variant sites to amplify fragments with genetic marker information (Table S3). The results show that 12 pairs of primers could successfully amplify the expected polymorphic sequences and could completely distinguish 10 species of Chenopodium and related genera (Figure 10). Most of the primers were designed based on InDel markers. Among them, primers 1 and 2 can be used to distinguish Atriplex from other genus species, and primers 3 and 4 can further distinguish A. patens and A. prostrata. Primer 5 and primer 6 can be used to distinguish Beta species. Primer 7 can distinguish Chenopodium species, except for distinguishing C. quinoa from other species. Primer 8 can be used to distinguish C. acuminatum from other Chenopodium species, and primer 8 and primer 9 can further distinguish C. berlandieri. In addition to directly distinguishing D. ambrosioides, primers 10 and 11 can also distinguish C. ficifolium from other Chenopodium species. At the same time, the combination of primers 7 can further distinguish C. quinoa. Primer 12 is the only pair of primers based on dCAPs markers, which are more cumbersome than InDel markers. Because there are only two SNPs in the cp genome of C. album and C. bryoniifolium, we assembled and annotated their ITS sequences, but there was no difference after comparison.

4. Discussion

After sequencing and assembling, we obtained the complete cp genomes of six closely related Chenopodium species, five of which are published in this paper for the first time, which greatly enriches the genome database of Chenopodium species and facilitates their classification, evolutionary relationship inference, and species identification.

Similar to previous studies, their lengths ranged from 151,850 to 152,215 bp and exhibited typical quadripartite structures [83,84]. The cp genomes of most terrestrial plants exhibit a nearly collinear sequence and genetic information, but there are also other occurrences, such as sequence inversion, gene loss, and IR region boundary contraction or expansion [85,86,87]. In the comparison of the cp genome of Chenopodium with other related genera, we found that there was an inversion of almost 6 kb in the LSC region of Beta, Dysphania, and Oxybasis (Figure 11). Studies have shown that this phenomenon is caused by tRNA activity or a high G + C content [88,89]. There was a trnV gene on the flanking of the region where the inversion occurs, but no significant change in the GC content was found.

The primary drivers behind cp genome size alterations are the contraction or expansion of the IR boundary [90,91,92]. The IR boundaries of Chenopodium and its related species are relatively conserved on most species, except for the obvious contraction on Beta, which also leads to the cp genome of B. vulgaris, which is about 2000 bp smaller than that of the other related species.

Repeat sequences exert a pivotal influence on the cp genome rearrangement and stabilization and population genetic diversity [93,94]. In this study, interspersed repeats and SSRs were widely distributed in all related species, and the proportion of repeat sequences in the IGS region was significantly greater than that in the CDS region, which also confirms the previous research results.

Previous studies on the phylogeny of Chenopodium based on cpDNA fragments (trnL-trnF, matK-trnK, atpB, atpB-rbcL, and rbcL) and ITS in rDNA presented some taxa conclusions that are different to our results, and the relationship was relatively vague [2,95,96]. Consequently, we reconstructed the phylogenetic tree utilizing the complete cp genome sequence. The results show that Chenopodium, Atriplex, Beta, Dysphania, and Oxybasis were clustered into a clade, and each of them became a monophyletic group. This also confirmed that the genetic distance between Chenopodium and Beta was closer than that of Suaeda and Haloxylon. In addition, the traditional plant classification method based on morphological characteristics included Chenopodium in the same tribe as Dysphania and Oxybasis, while Atriplex was classified as Trib. Atripliceae, which is obviously contrary to the phylogenetic outcomes derived from the complete cp genome, because Chenopodium and Atriplex have a closer genetic distance.

We used Limonium franchetii, which is also Caryophyllales, as the outgroup to estimate the evolutionary time of Chenopodium and its related species. The findings indicate that the divergence time between Amaranthaceae and the other families of Caryophyllaceae was 90.81 Mya, and the divergence time between Chenopodium and other genera was 30.28 Mya, which was different from the results of previous studies.

Previously, rbcL, trnH-psbA, and matK were regarded as critical barcodes for plant identification, but their resolution at the species level was inadequate [97]. The complete cp genome has more variable regions and has a higher resolution when used as a DNA barcode. Dependent on the complete cp genome, we developed InDel and dCAPS that can completely distinguish the 10 related species of Chenopodium. At the same time, a series of primers were designed based on these markers and verified by PCR experiments. These primers have a good repeatability and can be efficiently used for species identification and germplasm evaluation among the related species of Chenopodium. The ITS sequences of Chenopodium species were also assembled and we attempted to utilize these sequences for molecular marker development, but unfortunately, there are less variable sequences than the complete cp genome. For example, the ITS sequences of C. album and C. bryoniifolium are exactly the same and cannot be used for species discrimination.

In conclusion, this study introduced a significantly more efficient means for the identification of species within the Chenopodium genus and their closely related species.

5. Conclusions

In this investigation, the complete cp genomes of six species were assembled and annotated, and compared with the published sequences of related species. The outcomes reveal that the cp genomes of Chenopodium and its related species were highly conserved in structure and gene content, but there were also subtle variances. We identified a series of SSRs and highly variable regions from all Chenopodium and related species, and found that most of them were distributed in IGS regions and introns. The phylogenetic analysis showed that Chenopodium had the closest relationship with Atriplex, Beta, Dysphania, and Oxybasis and belonged to a monophyletic group compared with other Amaranthaceae species. They had a common ancestor 60.80 Mya, and since then, it gradually divided into different genera. Based on the interspecific variation sites, we designed 12 pairs of primers for species identification, and the experiments confirmed that they could completely distinguish 10 related species.