The Comparative Analyses of Six Complete Chloroplast Genomes of Morphologically Diverse Chenopodium album L. (Amaranthaceae) Collected in Korea.

Chenopodium album sensu stricto belonging to C. album aggregate is an annual cosmopolitan weed displaying the diversity of morphologies. We completed the six chloroplast genomes of C. album s. str. collected in Korea to understand the relationship between the diversity of chloroplast genomes and their morphological variations. All six C. album chloroplast genomes have a typical quadripartite structure with length ranging from 151,906 bp to 152,199 bp, similar to the previously sequenced C. album chloroplast genome (NC_034950). In total, 56 single nucleotide polymorphisms (SNPs) and 26 insertion and deletion (INDEL) regions (308 bp in total) were identified from the six chloroplast genomes, presenting a low level of intraspecific variations in comparison to the other angiosperm species. 376 normal simple sequence repeats were identified in all seven C. album chloroplast genomes. The phylogenetic analysis based on all available complete Amaranthaceae chloroplast genomes presents phylogenetic positions of six C. album samples as well as correlation with one of C. album morphological features. Our results provide the way to investigate intraspecific features of C. album chloroplast genomes and also the insights of understanding various intraspecific characteristics including morphological features.

1. Introduction

Chenopodium album aggregate, also known as Chenopodium album sensu lato, is one of the challenging groups for delimitating species boundary by phenotypes [1]. The main possibility of its phenotypic diversity is hybridization and polyploidization [2] because this group shows a diploid-polyploid complex ranging from diploids to decaploids with individual species which are very tightly correlated with a specific ploidy level and genome size [3-5]. Genetic diversity of C. album aggregate has been analyzed displaying that at least eight groups were identified, and their evolutionary lineages were supported well using flowering locus T-like (FTL) intron and chloroplast DNA maker sequences [2]. In addition, Central Asia presents the most diverse haplotypes, indicating that this area can be an origin of C. album aggregate [2].

C. album sensu stricto belonging to C. album aggregate is an annual cosmopolitan weed native to Eurasia [6]. It is one of the notorious weeds which reduce crop yield by exploiting resources such as light and nutrients from soil [6]. In contrast, C. album s. str. has also been cultivated as a crop in some countries: it was especially considered better nutritional crop species than wheat, barley, maize, and rice in the Himalaya area [7-9]. Besides it, C. album s. str. was also cultivated as a leafy vegetable [9, 10].

C. album s. str. has been recognized as a morphologically diverse species with difficulty in species identification [1, 11, 12]. This hexaploid species exhibited wide phenotypic plasticity covering morphological variations of other Chenopodium species, which shows stable morphologies under the greenhouse environment [1]. This high-level phenotypic variation of C. album s. str. can be explained by its allopolyploid originated from maternal tetraploids and paternal diploid parents [13]. Even though a few studies to understand the origin of C. album s. str. in the evolutionary aspect have been conducted [13, 14], there is no research to unravel the relationship between its morphological plasticity and the genetic diversity of organelle genomes.

Due to low sequencing cost caused by rapid development of next-generation sequencing (NGS) and third-generation sequencing (TGS) technologies [15-17] as well as the conserved structure of chloroplast genome reducing difficulties of de novo assembly, a huge number of chloroplast genomes have been massively sequenced. One of the evidences which reflects this situation is that the number of angiosperm complete chloroplast genomes deposited in NCBI is 9,323 covering 6,727 species as of 11 Jan 2021. These chloroplast genome sequences have been analyzed for the identification of phylogenetic positions [18-20] and for developing useful molecular markers [21, 22]. It presents that sequence variations of complete chloroplast genomes are enough information to conduct the studies for those purposes that complete chloroplast genomes can be also used to reveal the relationship between morphological plasticity and genetic diversity of C. album s. str.

Here, we completed the six chloroplast genomes of C. album s. str. of which both morphological features and geographical locations in the Korean Peninsula are distinct from each other. The C. album chloroplast genomes display a relatively low level of intraspecific variations (56 single nucleotide polymorphisms (SNPs) and 26 insertion and deletion (INDEL) regions across the seven C. album chloroplast genomes) with low nucleotide diversity compared to those of other angiosperm chloroplast genomes. 376 normal simple sequence repeats (SSRs), including two individual-specific normal SSRs, are detected among the seven chloroplast genomes, and the numbers of SSRs and their distribution on each C. album chloroplast genome are similar. Maximum likelihood and Bayesian inference phylogenetic trees based on all available complete Amaranthaceae chloroplast genomes present that seven C. album chloroplast genomes are clustered with short branch length and high supportive values. However, the correlation between their phylogenetic position and morphological features is not so strong. Hence, morphological plasticity of C. album which may not be explained by the genetic diversity of chloroplast genomes requires additional sequence data, such as nuclear marker sequences or specific genes which regulate leaf morphologies, to understand the origin of morphological plasticity.

2. Results

2.1. Six Complete Chloroplast Genomes and Their Morphological Features of Chenopodium album Collected in Korea

Six samples of C. album collected in Korea named CAGAP004, CAGOH01, CAJEJG05, CCANG01, CSJUK01, and CVHUP01 were selected based on their distinct morphological features and collected in different geographical positions in Korean Peninsula (Table 1 and Figure 1). The leaf shape of the six C. album samples presents a wide range which is from lanceolate to ovate. Their leaf margin shows the two types: (i) serrate (CAGAP004, CAGOH01, and CCANG01) and (ii) entire (CAJEJG05, CSJUK01, and CVHUP01; Figure 1 and Table 1). In addition, the thickness of leaves of the six samples is also divided into two types: (i) thick type (CAGAP004, CAJEJG05, and CVHUP01) and (ii) thin type (CAGOH01, CCANG01, and CSJUK01).

Table 1

List of six Chenopodium album samples used in this study.

Sample name	Voucher number∗	Morphological features			GPS coordinates
		Leaf shape	Leaf margin	Leaf thickness
CAGAP004	KYS130730	Lanceolate	Serrate	Thick	33°43′25.76″N 126°92′45.26″E
CAGOH01	sgu180626	Ovate	Serrate	Thin	34°29′30.09″N 127°21′19.84″E
CAJEJG05	sgu180521	Lance-ovate	Entire	Thick	33°10′16.00″N 126°15′57.73″E
CCANG01	sgu180601	Widely ovate	Serrate	Thin	37°35′35.75″N 126°30′53.71″E
CSJUK01	sgu180926	Lanceolate	Entire	Thin	37°58′19.00″N 128°45′51.00″E
CVHUP01	sgu180918	Lanceolate	Entire	Thick	35°32′54.26″N 126°40′27.30″E

∗All vouchers were deposited in Sung Kyun Kwan University Herbarium (SKKU) in Korea.

Figure 1

Geographical distribution of six Chenopodium album. The red arrows indicate the locations where six C. album samples were collected in South Korea. Pictures of voucher specimen were displayed on the left or right side of pictures habit in situ of the six samples.

To understand the relationship between morphological features with geographical distribution and its genetic background of chloroplast genomes, we completed chloroplast genomes of the six C. album samples (Table 1). The six chloroplast genomes have a typical quadripartite structure which has one large single-copy (LSC), one small single-copy, and two inverted repeat (IR) regions (Figure 2). The length of the six chloroplast genomes ranges from 151,906 bp (CCANG01) to 152,199 bp (CAGAP004; Table 2), presenting 293 bp differences. They are similar to that of one of the two previously sequenced C. album chloroplast genomes, of which GenBank accession is NC_034950 (152,167 bp) [23]. Interestingly, the other chloroplast genome of C. album, MF418659, displays much shorter (150,272 bp) than those of the remaining C. album chloroplast genomes [24]. Their overall GC contents are conserved as between 37.2% and 37.3%, which are slightly higher than that of MF418659 (37.0%; Table 2). The GC contents of LSC and SSC regions are identical in the seven C. album chloroplast genomes including the firstly sequenced chloroplast of C. album (NC_034950), which are 35.3% and 31.0%, respectively. Similarly, GC contents in the IR region of the seven chloroplast genomes are from 42.7% to 42.8% (that of MF418659 is 43.4%).

Figure 2

A circular gene map of six Chenopodium album chloroplast genomes. Genes shown outside are transcribed clockwise, and inside the circle are transcribed counterclockwise. Genes are color-coded to distinguish different functional groups. The dark grey and the light grey plots in the inner circle correspond to the GC content and AT content, respectively.

Table 2

List of six chloroplast genomes of Chenopodium album sequenced in this study.

Strain name	GenBank accession	Length (bp)				GC contents
		Whole	LSC	SSC	IR	Whole	LSC	SSC	IR
CCANG01	MW446241	151,906	83,681	17,969	25,128	37.3%	35.3%	31.0%	42.7%
CSJUK01	MW446242	152,197	83,679	18,132	25,193	37.2%	35.3%	31.0%	42.7%
CAJEJG05	MW446243	152,183	83,680	18,130	25,194	37.2%	35.3%	31.0%	42.7%
CVHUP01	MW446244	152,190	83,832	18,132	25,113	37.2%	35.3%	31.0%	42.8%
CAGOH01	MW446245	152,196	83,679	18,131	25,193	37.3%	35.3%	31.0%	42.7%
CAGAP004	MW446246	152,199	83,681	18,130	25,194	37.2%	35.3%	31.0%	42.7%

All seven C. album chloroplast genomes including NC_034950 contain 129 genes including 84 protein-coding genes (PGCs), 8 ribosomal RNAs (rRNAs), and 37 transfer RNAs (tRNAs; Table 3). Seventeen genes are duplicated in IR regions including 6 PCGs (rpl2, ycf2, ndhB, rps12, rps7, and ycf1), 4 rRNAs (rrn16, rrn23, rrn4.5, and rrn5), and 7 tRNAs (trnI-CAU, trnL-CAA, trnV-GAC, trnI-GAU, trnA-UGC, trnR-ACG, and trnN-GUU). The number of PCGs from the seven C. album chloroplast genomes is the same except MF418659 (Table 3), presenting that MF418659 has a quite different gene configuration. It has six additional PCGs, rpl23, ycf15, and ycf68 in the IR region, and loses psaJ PCG. In addition, it also misses one of the two rps12, displaying the same sequences of three exons of rps12 to the rest seven C. album chloroplast genomes; however, this missed rps12 should be added in the annotation of MF418659, resulting in 89 PCGs. After reannotation of two Chenopodium quinoa chloroplast genomes (KY635884 and MF805727), one ycf1 and four additional PCGs, ycf1, ycf2, and two rps12, are added, respectively, resulting in all chloroplast genomes of Chenopodium species having the same number of genes except MF418659 (Table 3).

Table 3

List of chloroplast genomes used for comparative analyses in this study.

Family	Species name	NCBI accession	Total length (bp)	# of PGCs	# of tRNAs	# of rRNAs	GC ratio (%)	Reference
Chenopodioideae	Chenopodium album	MW446241	151,906	84	37	8	37.3%	This study
	Chenopodium album	MW446242	152,197	84	37	8	37.2%	This study
	Chenopodium album	MW446243	152,183	84	37	8	37.2%	This study
	Chenopodium album	MW446244	152,190	84	37	8	37.2%	This study
	Chenopodium album	MW446245	152,196	84	37	8	37.3%	This study
	Chenopodium album	MW446246	152,199	84	37	8	37.2%	This study
	Chenopodium album	NC_034950	152,167	84	37	8	37.2%	[23]
	Chenopodium album ∗	MF418659	150,272	89 (88)∗∗	33	8	37.0%	[24]
	Chenopodium ficifolium	NC_041200	151,923	84	37	8	37.3%	[45]
	Chenopodium quinoa	KY635884	152,075	84 (83)∗∗	36	8	37.2%	[61]
	Chenopodium quinoa	MF805727	151,069	84 (80)∗∗	29	8	37.2%	[62]
	Chenopodium quinoa	NC_034949	152,099	84	36	8	37.2%	[23]
	Chenopodium quinoa cultivar Real Blanca	NSDK01003185.1	152,282	N/A	N/A	N/A	37.2%	[63]
	Atriplex centralasiatica	NC_045304	152,237	85	37	8	37.3%	[64]
	Dysphania ambrosioides	NC_041201	151,689	84	36	8	36.9%	[65]
	Dysphania botrys	NC_042166	152,055	83	37	8	36.8%	[66]
	Dysphania pumilio	MK541016	151,960	84	36	8	36.9%	[39]
	Dysphania pumilio	NC_041159	151,962	84	36	8	36.9%	[67]
	Oxybasis glauca	NC_047226	151,655	84	37	8	36.9%	[49]
	Spinacia oleracea	NC_002202	150,725	96	37	8	36.8%	[68]
Salicornioideae	Salicornia europaea	NC_027225	153,232	84	37	8	36.2%	Unpublished
	Salicornia bigelovii	NC_027226	153,076	83	37	8	36.3%	Unpublished
	Salicornia brachiata	NC_027224	153,324	84	37	8	36.2%	Unpublished
Suaedoideae	Suaeda japonica	NC_042675	152,109	83	37	8	36.4%	[46]
	Suaeda japonica	MK764271	152,112	80	37	8	36.4%	[38]
	Suaeda salsa	NC_045302	151,642	85	37	8	36.4%	[69]
	Suaeda glauca	NC_045303	149,807	85	37	8	36.5%	[70]
	Suaeda malacosperma	NC_039180	151,989	83	37	8	36.4%	[71]
	Bienertia sinuspersici	KU726550	153,472	86	36	8	37.8%	[72]
Salsoloideae	Haloxylon ammodendron	NC_027668	151,570	85	37	8	36.6%	[73]
	Haloxylon persicum	NC_027669	151,586	85	37	8	36.6%	[73]
Betoideae	Beta vulgaris	EF534108	149,696	N/A	N/A	N/A	35.4%	Unpublished
	Beta vulgaris subsp. vulgaris	KJ081864	149,635	85	37	8	37.0%	[74]
	Beta vulgaris subsp. vulgaris	KR230391	149,722	81	29	8	37.0%	[75]
Paronychieae	Gymnocarpos przewalskii	NC_036812	150,636	81	37	8	36.5%	[66]

∗ indicates that the species name should be reconsidered. ∗∗Numbers in parenthesis are the original number of PCGs based on the annotation, and numbers outside of parenthesis indicate the number of PCGs based on our reannotation results.

In the seven C. album chloroplast genomes, there are six ATP synthase subunit genes, 11 NADH dehydrogenase genes, four RNA polymerase genes, six genes encoding subunits of cytochrome b/f complex, and 15 photosystem subunit II genes. Five genes encoding photosystem subunit I are found in the seven C. album chloroplast genomes (Table 4), while MF418659 has only four genes with losing psaJ. There are 20 ribosomal proteins presented in the seven C. album chloroplast genomes consisting of 12 ribosomal proteins encoding small subunit and eight genes for large subunits (Table 4). Interestingly, an additional rpl23 gene is found only in the MF418659 chloroplast genome. Other remaining genes encode acetyl-CoA-carboxylase (accD), translation initiation factor (infA), protease (clpP), chloroplast envelope membrane protein (cemA), maturase K gene (matK), and cytochrome c biogenesis protein (ccsA; Table 4). The number of hypothetical proteins is four among the seven C. album chloroplast genomes, except for MF418659 which has six genes (Table 4), presenting the different gene configuration of MF418659.

Table 4

List of genes encoded by Chenopodium album chloroplast genomes.

Category	Group of gene	Name of gene
Self-replication	Ribosomal RNAs	rrn4.5, rrn5, rrn16, rrn23
	Transfer RNA genes	trnA-UGC, trnC-GCA, trnD-GUC, trnE-UUC, trnF-GAA, trnG-GCC, 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
	Small subunit of ribosome	rps2, rps3, rps4, rps7, rps8, rps11, rps12, rps14, rps15, rps16, rps18, rps19
	Large subunit of ribosome	rpl2, rpl14, rpl16, rpl20, rpl22, rpl32, rpl33, rpl36
	RNA polymerase	rpoA, rpoB, rpoC1, rpoC2
	Translation initiation factor	infA
Photosynthesis	ATP synthase	atpA, atpB, atpE, atpF, atpH, atpI
	NADH dehydrogenase subunit	ndhA, ndhB, ndhC, ndhD, ndhE, ndhF, ndhG, ndhH, ndhI, ndhJ, ndhk
	Cytochrome b/f complex subunit	petA, petB, petD, petG, petL, petN
	Photosystem subunit I subunit	psaA, psaB, psaC, psaI, psaJ
	Photosystem subunit II subunit	psbA, psbB, psbC, psbD, psbE, psbF, psbH, psbI, psbJ, psbK, psbL, psbM, psbN, psbT, psbZ
	Rubisco large subunit	rbcL
Other genes	Maturase	matK
	Protease	clpP
	Envelope membrane protein	cemA
	Subunit of acetyl-CoA-carboxylase	accD
	Cytochrome c-type biogenesis protein	ccsA
Genes of unknown function	Hypothetical reading frame	ycf1, ycf2, ycf3, ycf4

In the seven C. album chloroplast genomes, nine PCGs contain one intron (rps16, atpF, rpoC1, petB, petD, rpl16, rps12, ndhB, and ndhA) and only clpP and ycf3 have two introns, which are conserved across chloroplast genomes of the other Chenopodium species. MF418659 chloroplast also has the same intron structure as the remaining Chenopodium chloroplast genomes. Taken together with the different properties of the MF418659 chloroplast genome including its length, GC ratio, and a number of genes, we suspected that MF418659 chloroplast genome might not be C. album; hence, we will exclude MF418659 for further analyses conducted in this study.

2.2. Nucleotide Diversity and Intraspecific Variations Identified from the Seven Chenopodium album Chloroplast Genomes

To investigate nucleotide diversity (π) and intraspecific variations of C. album chloroplast genomes, the six C. album complete chloroplast genomes sequenced in this study are aligned against the previously sequenced C. album chloroplast genome (NC_034950). The average value of nucleotide diversity is 0.0000625 (Figure 3), and a total of 56 single nucleotide polymorphisms (SNPs) and 26 insertion and deletion (INDEL) regions (308 bp in total) are identified. The LSC region, where the average nucleotide diversity is the highest (π = 0.00102), contains 35 SNPs (62.5%) and 16 INDEL regions (47 bp in length; 59.3%). Fifteen SNPs (26.8%) and 5 INDEL regions (193 bp in length; 8.93%) are found in the SSC region, displaying that the number of SNPs in the LSC region is larger more than twice that of the SSC region; however, the total length of INDEL regions in the SSC region is about 4 times greater than that of the LSC region. The main reason for this phenomenon is the presence of the 162 bp INDEL located between rpl32 and trnL-UAG genes. An IR region covers three SNPs and three INDEL regions (68 bp in length), which corresponds to the lowest nucleotide diversity in the IR region (π = 0.0000146). The low level of sequence variations in the IR region is known as a general phenomenon in the chloroplast genomes [25-27].

Figure 3

Nucleotide diversity of the seven Chenopodium album chloroplast genomes. The X-axis presents coordination of chloroplast genome, and the Y-axis presents values of nucleotide diversity value of 500 bp window sliding by 200 bp step. Highest peaks of the nucleotide diversity graph display the name of genic regions. Green, pink, and grey bars below the X-axis indicate the LSC, IR, and SSC regions, respectively.

Twenty-six out of 56 SNPs (46.4%) are located in PCG regions, while seven SNPs (12.5%) are in intronic and the 24 SNPs (42.9%) are in intergenic regions. As one SNP classified as nonsynonymous SNP is also classified into intronic SNP because trnK-UUU contains two exons located before and after matK, the total number of SNPs mentioned above is 57. The numbers of SNP per one kilobase pairs (Kbp) of the PCG, intronic, and intergenic regions are 0.329 SNPs/Kbp, 0.305 SNPs/Kbp, and 0.552 SNPs/Kbp, respectively, presenting that the intergenic region shows the highest density of SNPs. Eleven PCGs (psbA, rpl16, clpP, infA, rps11, petA, ndhD, psbA, rpoC1, rps16, and atpB) contain only one SNP in their coding region, while four genes (matK, ccsA, ycf1, and ndhF) cover more than one SNP. In addition, three PCGs (clpP, petB, and ndhA) have one SNP in their intronic region, while rpl16 has two SNPs in their intronic region. In total, 15 nonsynonymous SNPs and 11 synonymous SNPs are identified in the 14 PCGs. Remarkably, a ratio of nonsynonymous to synonymous SNPs found on C. album chloroplast genomes is 15 : 11, which is different from the common phenomenon that the number of nonsynonymous SNPs is smaller than that of synonymous SNPs on the other plant chloroplast genomes [26, 28]. Each synonymous SNP is scattered in the eleven genes (psbA, rps16, rpoC1, atpB, petA, clpP, rps11, rpl16, ndhF, ccsA, and ndhD), while 15 nonsynonymous SNPs are concentrated only in the five genes (matK, infA, ycf1, ndhF, and ccsA). Ycf1, which is duplicated in the IR region and extends between IR and SSC regions, contains three nonsynonymous SNPs in each pair of the IR region, and one additional nonsynonymous SNP in the SSC region. The PCG containing the largest SNPs is ndhF, which contains five SNPs and is located mainly in the SSC region and partially extended over the IR region. Four out of the five SNPs identified in ndhF are nonsynonymous, among which one SNP found in the IR region is commonly shared with ycf1. Three PCGs (psbA, matK, and ycf1) have one INDEL region (11.5%) of which lengths are 6 bp, 6 bp, and 27 bp, respectively: none of these INDEL regions cause any frameshift mutation. Seven INDEL regions (26.9%) are identified in the intronic regions of rps16, atpF, ycf3, clpP, and rpl16. Most of them are less than 5 bp long; however, 66 bp INDEL is found in the intron of trnI-GAU in each IR region. The remaining 16 INDEL regions (61.5%) are intergenic. The longest INDEL region is 162 bp long, detected between rpl32 and trnL-UAG.

There are eight sites that have relatively higher π values (>0.0008) including two PCGs matK (π = 0.00114) and ccsA (π = 0.00163) and six intergenic regions (trnH-psbA, petN-psbM, rpl11-rpl36, rpl36-infA, ycf1-ndhF, and rpl32-trnL; Figure 3). CcsA contains two nonsynonymous SNPs and one synonymous SNP, and matK has two nonsynonymous SNPs and one 6 bp INDEL, presenting that ccsA displays the highest SNP density among the PCGs. The highest π value of the intergenic region is observed between trnH and psbA (π = 0.00128; Figure 3).

2.3. Comparative Analysis of Simple Sequence Repeat (SSR) Polymorphisms on Chloroplast Genomes of C. album

In the seven C. album chloroplast genomes, 376 normal SSRs are identified (Table 5; Supplementary Tables 1–7). In addition, we also identified 280 extended SSRs and 3,039 potential SSRs on the seven chloroplast genomes (see Materials and Methods; Supplementary Tables 1–7). We analyzed only normal SSRs hereafter because normal SSRs can be commonly recognized as SSRs in various studies (see Materials and Methods). The unit length of normal SSRs varies from 1 bp (monoSSR) to 5 bp (pentaSSR), and the numbers of normal SSRs in each chloroplast genome are from 53 to 55, displaying an almost similar manner: CAGAP004 contains 55 normal SSRs, the largest, and CAJEJG05, CCANG01, and CSJUK01 contains 54 normal SSRs, while CAGOH01, NC_034950, and CVHUP01 have 53 normal SSRs (Table 5). Interestingly, no hexaSSR is identified on the seven C. album chloroplast genomes. The majority of normal SSRs is monoSSR (60.6%), and pentaSSR (3.70%) is the least (Figure 4(a)). In monoSSR, only an A/T motif was detected in all seven chloroplast genomes.

Table 5

List of normal SSRs of six Chenopodium album.

SSR type	CAGAP004	CAGOH01	CAJEJG05	CCANG01	CSJUK01	CVHUP01	NC034950
MonoSSR	34	32	33	33	32	32	32
DiSSR	9	9	9	9	9	9	9
TriSSR	2	2	2	2	2	2	2
TetraSSR	9	9	9	9	10	9	9
PentaSSR	1	1	1	1	1	1	1

Figure 4

Number of normal SSRs identified on the seven Chenopodium album chloroplast genomes. (a) Displays the number of SSR types in each seven C. album chloroplast genome. The X-axis means the seven C. album samples, and the Y-axis indicates the number of normal SSRs. Five different colors indicate the five types of SSRs, monoSSRs, diSSRs, triSSRs, tetraSSRs, and pentaSSRs. The table below the X-axis presents numbers of SSRs along with samples and the five SSR types. (b) Shows the distribution of SSRs in noncoding, intron, intergenic, and coding regions along with the samples. The X-axis means the seven C. album samples, and the Y-axis indicates the number of normal SSRs. Four different colors correspond to the four different regions. The table below the X-axis shows the number of SSRs along with samples and regions. (c) Shows SSR density (# of SSRs/kb) of the LSC, SSC, and IR regions along with the samples. The X-axis means the seven C. album samples, and the Y-axis indicates the number of normal SSRs along with samples and the three regions. Three different colors in the bar graphs mean the three regions. The table below the X-axis shows SSR density along with samples and the three regions.

The overall distribution of normal SSRs on the seven C. album chloroplast genomes is similar to each other (Figure 4(b)). The intergenic region displays the largest number of normal SSRs, then coding, intron, and noncoding regions are in order (see Materials and Methods; Figure 4(b)). In all seven C. album chloroplast genomes, four normal SSRs are found in the noncoding region of rrn23, and 11 normal SSRs are identified in the coding regions of rpoC2, rpoB, atpB, rpoA, ycf1, and ndhB. In total, 30 to 33 normal SSRs were found in the intergenic regions on the seven chloroplast genomes. Most of the normal SSRs in the intergenic region are shared among the seven chloroplast genomes; however, CAJEJG05 and CAGAP004 chloroplast genomes have two distinct normal SSRs in their intergenic region of trnR-trnN and rpl32-trnL caused by one-bp INDEL and one SNP changing “T” to “A,” respectively. Also, CCANG01 and CSJUK01 chloroplast genomes have additional monoSSR between ndhC and rbcL by insertion of 1 bp nucleotide “T.” In contrast, the deletion of a single nucleotide of “A” between ndhF and rpl32 caused the removal of intergenic SSRs in CAGOH01, CCANG01, and CSJUK01 chloroplast genomes. All seven C. album chloroplast genomes have seven common normal SSRs located in the intronic regions of five PCGs, rps16, atpF, ycf3, rpl16, and ndhA. In the case of CAGOH01 and CCANG01 chloroplast genomes, one additional monoSSR is identified in the intronic region of trnK because one SNP changing “A” to “T” occurred in both chloroplast genomes. Besides, CSJUK01 has an extra unique tetraSSR “GTTT” in the intronic regions of ycf3 by 4 bp insertion. These differences of normal SSRs among the seven chloroplasts of C. album can be utilized as molecular markers to distinguish their origins inside the Korean Peninsula once more chloroplast genomes of C. album in Korea are available.

Due to the different lengths of the LSC and SSC regions, the density of SSRs per Kbp was calculated. Interestingly, three chloroplast genomes, CAGOH01, CCANG01, and CSJUK01, display similar density in both the LSC and SSC regions, while the remaining four chloroplast genomes present that the density in the SSC region is larger than that of the LSC region. Since in the SSC region of CAGAP004, CAJEJG05, CVHUP01, and NC_034950 there is an additional intergenic normal SSR located in between ndhF and rpl32 (Figure 4(c)). The densities of normal SSRs in IR regions of all seven chloroplast genomes are lowest (0.159 to 0.198 normal SSRs/Kbp; Figure 4(c)).

To understand conserved normal SSRs across the seven chloroplast genomes, we calculated SSR groups which contain normal SSRs of which left and right flanking sequences are similar to each other (see Materials and Methods). In total, 58 SSR groups and two singleton SSRs were identified, and 50 of 58 SSR groups (86.2%) contain seven normal SSRs originating from all seven chloroplast genomes, called the common SSR group. Eleven out of the 50 common SSR groups (22.0%) are located in the coding region, and 36 common SSR groups (72.0%) are in the intergenic region (Figure 4(b)), which is congruent to the analysis result of normal SSRs mentioned in the previous section. Five intergenic loci contain two common SSR groups, rpl32-trnL, atpH-atpI, ycf3-trnS, trnQ-psbK, and trnK-rps16 in each, and 31 intergenic loci contain one common SSR group. Two singleton SSRs were found in CSJUK01 and CCANG01 chloroplast genomes. These intraspecific variations of normal SSRs will provide insights into changes of SSRs inside the species, which can also be utilized to develop molecular markers of C. album efficiently.

2.4. Phylogenetic Analysis of Korean C. album Chloroplast Genome Sequence

Bootstrapped maximum likelihood (ML) and Bayesian inference (BI) phylogenetic trees of 34 Amaranthaceae chloroplast genomes including the six C. album chloroplasts sequenced in this study and one outgroup species, Gymnocarpos przewalskii, were constructed (see Materials and Methods). Phylogenetic trees present that six C. album chloroplast genomes are clustered with the previously sequenced C. album chloroplast genome (NC_034950) with high supportive values of ML and BI except the node containing NC_034950 in the ML tree (Figure 5(a)). The C. album s. str. clade is divided reciprocally into two clades in both trees. CAGAP004 and CAJEJG05 sharing the morphological feature of narrow leaves and collected in Jejudo island (Figure 1) only exhibit a correlation of geographical locations with high supportive values of ML and BI (Figure 5(b)). Leaf shape and margin of the six samples are not correlated to the two clades of C. album (Table 1 and Figure 5(b)); however, leaf thickness of the six C. album presents correlation to the clades: the clade containing CAGAP004, CAJEG05, and CVHUP01 shows thick leaves, called as a thick-leaf clade, and the clade consisting of CAGOH01, CCANG01, and CSJUK01 displays thin leaves, called as a thin-leaf clade (Table 1 and Figure 5(b)). NC_034950 clustered in the thick-leaf clade was not possible to be confirmed whether its leaves are thick or not. Taken together, the phylogenetic relationship of the six C. album chloroplast genomes seems not to be highly correlated with their morphological features and geographical locations, supporting that their high plasticity of morphology links to other factors such as nuclear markers, polyploidy, or any regulatory factor of leaf morphologies. With the additional chloroplast genomes as well as nuclear marker sequences of C. album collected in Korea, these relationships of morphological features and geographical locations will be more explicit.

Figure 5

Phylogenetic trees of 34 Amaranthaceae chloroplast genomes. (a) Maximum likelihood (ML) and Bayesian inference (BI) phylogenetic trees were constructed based on complete chloroplast genomes of 34 Amaranthaceae species and Gymnocarpos przewalskii as an outgroup species. The phylogenetic tree was drawn based on the ML tree. Numbers on branches in the phylogenetic tree indicate bootstrap values of ML and prior possibility of BI, respectively. The dotted red rectangle indicates the clade covering C. album and C. ficifolium. Chloroplast genomes sequenced in this study were presented as bold characters. (b) Partially enlarged phylogenetic tree of the clade of C. album and C. ficifolium was displayed. Numbers on branches in the phylogenetic tree indicate bootstrap values of ML and prior possibility of BI, respectively. Chloroplast genomes sequenced in this study were presented as bold characters.

3. Discussions

3.1. Species Incongruency of C. album Chloroplast Genomes

In this study, we sequenced six chloroplast genomes of C. album s. str. collected in Korea displaying various morphological features. One of the previously sequenced chloroplast genomes of C. album, MF418659, is quite different from the remaining C. album chloroplast genomes in the aspects of gene configuration of chloroplast genome (Table 3) as well as phylogenetic relationship (MF418659 chloroplast genome was located outside of the clade of Chenopodium and Atriplex in Figure 5). These differences indicate that MF418659 may neither be C. album nor genus Chenopodium.

It is partially supported by the fact that the collection site of MF418659 is the Himalayan area in India [24] where C. album has been mainly cultivated as crops [7-9]. Usually, species diversity of the Himalayan area is higher due to its wide variety of climates as well as various climatic perturbations that have been applied to different locations in the Himalayan area [29]. Considering the phylogenetic position of MF418659 (Figure 5), it is possible that MF418659 is misidentified or an unreported species which is very different from C. album.

This kind of incongruency problem of species has sometimes been found during comparative analyses in plant species. For instance, two Magnolia chloroplast genomes, Magnolia insignis (NC_035657) and Magnolia alba (NC_037005), were reported as examples of misidentification species based on the phylogenetic analysis based on complete chloroplast genomes [30]. This problem can occur due to not enough taxonomic coverage of whole chloroplast genomes or misidentification of the samples used in the studies because of difficulties in species identification based on morphologies. Therefore, the identification of MF418659 should be revised in some ways, such as species identification of the voucher used in the previous study or sequencing and analyzing more samples of C. album collected in the Himalayan area.

3.2. Possible Causes of C. album s. str. Morphological Variations at the Molecular Level

Based on cytogenic and nuclear molecular marker analysis of Chenopodium species, C. album is distinct to C. ficifolium (B genome diploid) and C. quinoa (A genome tetraploid) [14], and two major groups of C. album were identified based on the phylogenetic tree based on rrn5 and ITS sequences [14]. In comparison to the phylogenetic tree which displays that C. album and C. ficifolium were clustered in one clade with high supportive values (Figure 5), a maternal lineage of both species is nearer than that of biparental lineage. Several C. quinoa chloroplast genomes were clustered in the distinct clade to that of C. album and C. ficifolium (Figure 5), reflecting the different types of their genomes [14]. This phylogenetic tree based on complete chloroplast genomes (Figure 5) also indirectly supports that various intraspecific evolutionary events in several Chenopodium species, including C. album and C. quinoa, may have occurred, such as hybridization and polyploidization [2, 14].

Polyploidization and hybridization events can usually cause morphological plasticity and diversity: e.g., Nicotiana species display various flower colors based on events of polyploidization [31], and Centaurea stoebe, polyploidy species, shows that it causes various phenotypes to climate, resulting in boosting its invasion [32]. Similarly, morphological variations of C. album are not related to maternal lineage (Figures 1 and 5). It can be inferred that C. album presents various morphological differences because it is hexaploidy species. It can also be interpreted that these morphological variations are not fully genetically fixed but may be caused by nuclear genes related to leaf development, such as Class I KNOX genes, homeobox transcription factors which can regulate leaf shapes in Arabidopsis thaliana [33]. In this study, we found that the leaf morphology of six C. album has a weak correlation with their phylogenetic relation (Figures 1 and 5). If diverse leaf shapes of C. album are caused by these key regulators, we can deduce that the general trend of evolutionary process inferred from organelle genomes including chloroplast cannot explain this diversity because these regulators can display different evolutionary speeds and patterns from those of organelle genomes.

Several studies tried to delimitate species presenting different morphological features using whole chloroplast genome sequences. For instance, a phylogenetic tree constructed based on chloroplast genomes of Anemopaegma acutifolium supported that two leaf morphological trait types of A. acutifolium were caused by different maternal origins [34]. In the case of Triplostegia glandulifera and T. grandiflora, their chloroplast genomes were used for solving the boundary of the two species; however, they could not explain the high morphological plasticity of them [35]. Therefore, further analyses with more chloroplast genomes of C. album expressing various phenotypic characteristics will be necessary to understand the origin of its morphological plasticity.

3.3. Evaluation of Level of Intraspecific Variations on C. album Chloroplast Genomes

The intraspecific variations identified among the seven Chenopodium album chloroplast genomes (56 SNPs and 26 INDEL regions) are compared with the previous studies which investigated intraspecific variations on chloroplast genomes. Twenty cultivars and wild types of Ricinus communis (Castor bean) displayed 162 SNPs and 92 INDEL regions [28], which is three times more than those of C. album. Sixty-three chloroplast genomes of Macadamia integrifolia (Macadamia nut) are collected in eastern Australia, which is a smaller geographical range of C. album. Four hundred and seven SNPs [36] are detected from them, which is seven times more than the number of SNPs identified in this study. Comparing with that of our study, the numbers of intraspecific variations identified from C. album chloroplast genome are relatively lower. In the case of Dioscorea polystachya (Chinese yam), six chloroplast genomes collected in Northern and Southern China displayed 141 SNPs and 44 INDEL regions [37]. Its geographical coverage is larger than that of C. album, and climates of the six regions are quite different than those of C. album; the larger number of intraspecific variations in D. polystachya is reasonable.

To evaluate intraspecific variations on C. album chloroplast genomes considering its geographical distribution, various studies which identified intraspecific variations of organelle genomes from the plant species collected in Korea were surveyed. To compare intraspecific variations between two samples in the same species, we conducted a pairwise comparison of C. album chloroplast genomes, resulting in 0 to 33 SNPs and 7 to 36 INDEL regions being identified from the seven C. album chloroplast genomes. In the case of Suaeda japonica collected in Korea with different morphological features, only three SNPs and three INDEL regions were identified [38], which is mostly smaller than those identified in C. album. The number of intraspecific variations identified from Dysphania pumilio, another Amaranthaceae species, is 24 SNPs and one INDEL region [39], which is also in the range of those of C. album. Based on these previous results, intraspecific variations identified from the seven C. album chloroplast genomes are similar to those of Amaranthaceae species.

4. Materials and Methods

4.1. DNA Extraction of Natural Collection of Korean C. album

Six samples of C. album were collected in various places in the Korean Peninsula (Table 1 and Figure 1). All vouchers of the six samples were deposited to the Sung Kyun Kwan University Herbarium (SKKU; Table 1). Their total DNA was extracted from fresh leaves of the six samples using a DNeasy Plant Mini Kit (QIAGEN, Hilden, Germany).

4.2. Genome Sequencing and De Novo Assembly of the Natural Collection of Six C. album Chloroplast Genomes

Genome sequencing was performed using HiSeqX at Macrogen Inc., Korea, from the extracted DNA of the six C. album. De novo assembly with confirmation was accomplished with Velvet v1.2.10 [40] after filtering raw reads using Trimmomatic v0.33 [41]. After obtaining the first draft of the chloroplast genome sequences, gaps were filled with GapCloser v1.12 [42], and all bases from the assembled sequences were confirmed by checking each base in the alignment (tview mode in SAMtools v1.9 [43]) against the assembled chloroplast genome generated with BWA v0.7.17 [44]. All these processes were conducted under the environment of the Genome Information System (GeIS; http://geis.infoboss.co.kr/; Park et al., in preparation) like other Amaranthaceae chloroplast genomes assembled [38, 39, 45–49].

4.3. Chloroplast Genome Annotation

Geneious Prime® 2020.2.4 (Biomatters Ltd, Auckland, New Zealand) was used for chloroplast genome annotation based on the C. album chloroplast genome (NC_034950) [23] by transferring annotations while correcting exceptional cases, including missing start or stop codons. tRNA was predicted and confirmed based on the prediction of tRNAScan-SE v2.0.6 [50]. A circular map of C. album chloroplast was drawn by using the OGDRAW v1.3.1 [51].

4.4. Identification of Sequence Variations on the Complete Chloroplast Genomes of C. album

Single nucleotide polymorphisms (SNPs) and insertions and deletions (INDELs) were identified from the pair-wise alignments of two selected chloroplast genomes conducted by MAFFT v7.450 [52]. When the number of INDELs was calculated, continuous INDEL bases were considered one INDEL. In addition, we denote the four regions: (i) coding region is exon that encodes a protein, (ii) intron regions indicate the region which does not translate inside protein-coding genes, (iii) intergenic regions are the sequence between two genes, and (iv) noncoding region means the sequence located in tRNAs or rRNAs.

4.5. Identification of Simple Sequence Repeats (SSRs)

Simple sequence repeats (SSRs) were identified on the chloroplast genome sequence using the pipeline of the SSR database (SSRDB; http://ssrdb.infoboss.co.kr/; Park et al., in preparation). Based on the conventional definition of an SSR on the chloroplast genome, monoSSR (1 bp) to hexaSSR (6 bp), the total length of SSRs on the chloroplast genome exceeds 10 bp. Owing to the different criteria of SSRs on chloroplast genomes, we adopted the criteria used in chloroplast genomes of Dysphania [47] and Arabidopsis thaliana [53] and mitochondrial genome of Rosa rugosa [54] as follows: the monoSSR (unit sequence length of 1 bp) to hexaSSR (6 bp) are used as normal SSRs, and heptaSSR (7 bp) to decaSSR (10 bp) are defined as extended SSRs. Among the normal SSRs, pentaSSRs and hexaSSRs for which the repeat number of unit sequences is 2 are classified as potential SSRs. Classification of regions on chloroplast genome was conducted in the same way described in the above section.

4.6. Comparison of SSRs Identified from Seven C. album Chloroplast Genomes

SSRs identified from seven C. album chloroplast genomes were compared based on their flanking sequences under the environment of the SSRDB (http://ssrdb.infoboss.co.kr/; Park et al., in preparation). The pipeline of the SSR comparison implemented in the SSRDB used in various organelle genome studies [53, 55] was used with the following conditions: a cut-off e value of 1e − 10 and a maximum flanking sequence for the comparison of 60 bp.

4.7. Nucleotide Diversity Analysis

Nucleotide diversity was calculated using the method proposed by Nei and Li [56] based on the multiple sequence alignment of Chenopodium chloroplast genomes using a Perl script used in previous studies [47, 53, 57]. The window size was set to 500 bp, and the step size was 200 bp when using the sliding-window method. Genomic coordination of each window was compared to the gene annotation of the chloroplast genome under the GenomeArchive® (http://www.genomearchive.net/) [58] environment for further analyses.

4.8. Construction of Phylogenetic Trees

The whole 34 Amaranthaceae chloroplast genomes and one outgroup of Gymnocarpos przewalskii chloroplast genome were aligned by MAFFT v7.450 [52], and alignment quality was checked manually. The maximum likelihood (ML) tree was reconstructed in IQ-TREE v1.6.6 [59]. In the ML analysis, a heuristic search was used with nearest-neighbor interchange (NNI) branch swapping, TVM+F+R4 model, and uniform rates among sites. All other options used the default settings. Bootstrap analyses with 1,000 pseudoreplicates were conducted with the same options. The posterior probability of each node was estimated by Bayesian inference (BI) using the MrBayes v3.2.7a [60] plug-in implemented in Geneious Prime® 2020.2.4 (Biomatters Ltd, Auckland, New Zealand). The HKY85 model with gamma rates was used as a molecular model. A Markov chain Monte Carlo (MCMC) algorithm was employed for 1,100,000 generations, sampling trees every 200 generations, with four chains running simultaneously. Trees from the first 100,000 generations were discarded as burn-in.

5. Conclusions

We completed the six chloroplast genomes of Chenopodium album showing various morphological features. The structure and gene order of chloroplast are conserved among seven C. album including the previously sequenced chloroplast genome (NC_034950). The average nucleotide diversity calculated from the seven C. album chloroplast genomes is 0.0000625, and a total of 56 SNPs and 26 INDEL regions are found. In comparison to the other cases of chloroplast intraspecific variations, C. album chloroplasts present a low level of sequence variation. The number of normal SSR identified from the seven C. album chloroplast genomes ranges from 33 to 35 displaying similar distribution and density of SSRs. Interestingly, specific SNPs and INDEL regions in intronic and intergenic regions make SSR variation among the seven chloroplasts. All seven C. album chloroplast genomes are clustered in high supportive values of ML and BI trees with a short length of branches. In addition, one of the morphological characters of C. album s. str., the thickness of leaves, presented correlation with the phylogenetic position. Taking together the results in this study, our six chloroplast genomes of C. album s. str. will provide the way to investigate intraspecific features of chloroplast genomes, also the insights of intraspecific variations to understand various characteristics of one species including morphological features.