Shusheng Fang, Liemei Zhang,Jianmin Qi, Liwu Zhang,c,*

aKey Laboratory of Ministry of Education for Genetics, Breeding and Multiple Utilization of Crops/Fujian Provincial Key Laboratory of Crop Breeding by Design,College of Agriculture,Fujian Agriculture and Forestry University, Fuzhou 350002,Fujian,China

bExperiment Station of Agriculture and Rural Affairs of Ministry for Jute and Kenaf in Southeast China/Fujian Public Platform for Germplasm Resources of Bast Fiber Crops/Fujian International Science and Technology Cooperation Base for Genetics, Breeding and Multiple Utilization Development of Southern Economic Crops,Fujian Agriculture and Forestry University, Fuzhou 350002,Fujian,China

cCenter for Genomics and Biotechnology, Haixia Institute of Science and Technology, Fujian Agriculture and Forestry University, Fuzhou 350002,Fujian,China

Keywords:Jute Corchorus Genome Chloroplast Insertion or deletion

ABSTRACT Jute (Corchorus spp.) is a member of the Malvaceae family, which comprises more than 100 species. The systematic positions of jute species have remained unsettled. Chloroplasts are maternally inherited and their genomes are widely used for plant phylogenetic studies.In the present study, the chloroplast genomes of Corchorus capsularis and C. olitorius were assembled,with sizes of respectively 161,088 and 161,766 bp.Both genomes contained 112 unique genes(78 protein-coding,four rRNA,and 30 tRNA genes).Four regions with high variation between the two species were located in single-copy rather than inverted-repeat regions. A total of 66 simple sequence repeats(SSRs)were identified in the C.capsularis chloroplast genome and 56 in that of C.olitorius. Comparison of the two chloroplast genome sequences permitted the evaluation of nucleotide variation including 2417 single-nucleotide polymorphisms sites and 294 insertion or deletion sites, of which one marker (cpInDel 205) could discriminate the two jute species.Comparison of the C.capsularis and C.olitorius chloroplast genomes with those of other species in the Malvaceae revealed breakpoints in the accD locus,which is involved in fatty acid synthesis,in C.capsularis and C.olitorius.This finding suggests that genes from the chloroplast genome might have been transferred to the nuclear genome in some Corchorus species. This hypothesis was supported by synteny analysis of the accD region among the nuclear, chloroplast, and mitochondrial genomes.To our knowledge,this is the first report of the assembled chloroplast genome sequences of C.capsularis and C.olitorius.C.capsularis and C.olitorius are closely related to Gossypium species and there are abundant microstructure variations between these two genera.These results will expand our understanding of the systematics of species in the Malvaceae.

1.Introduction

Jute(Corchorus spp.)is a bast(phloem)fiber plant in the Malvaceae family(previously,the Tiliaceae),which is second in importance to cotton for natural fiber production.Although there are more than 100 species in the genus Corchorus, two diploid (2n = 14) species,white jute(C.capsularis)and dark jute(C.olitorius),are cultivated as crops.The two cultivated jute species produce an environmentally friendly, biodegradable, and renewable natural fiber [1-3]. In general,the two cultivated species differ in growth habit,disease resistance, and characteristics associated with flowering and silique shape.Their phylogenetic origins are debated[4].

Using morphological traits,Xiong et al.[3]suggested that the center of origin of C.capsularis is in Indo-Burma and that of dark jute in Africa. However, using nuclear and chloroplast simple sequence repeats (SSRs) or microsatellites, Kundu et al. [5]proposed that the two cultivated species of jute originated in Africa.Based on specific chloroplast DNA regions(MatK,IGS1,and IGS2), internal transcribed spacer (ITS) sequences, and the nuclear-encoded xyloglucan endotransglucosylase/hydrolase(XTH)gene,Tanmoy et al.[4]proposed that these two cultivated Corchorus species are related via maternal inheritance and presented archeological support for their dispersal routes.However,all of these previous studies identified variation using universal primers for intergenic regions rather than Corchorusspecific chloroplast genomic markers. To investigate evolutionary lineages and phylogenetic relationships, it is desirable to generate assembled chloroplast genome sequences.

The plant chloroplast genome is more conserved than the nuclear genome.The maternal inheritance and unique evolutionary features of chloroplasts have led to the wide use of their genome sequences in plant phylogenetic studies. Complete chloroplast genome sequences have revealed many mutation events, including InDels (insertions or deletions), substitutions,and inversions [6]. At the subfamily level, a 22-kb DNA inversion event was used to confirm that the Barnadesioideae comprise the most basal lineage in the Asteraceae [7]. Three DNA inversion events were used as a nested set of phylogenetic characters to clarify the close relationship between the Poaceae and Joinvilleaceae[8].In rice,the within-genus polymorphism rates of InDels and SNPs between Oryza sativa and O.nivara was estimated at 0.02%[9],and that between O.sativa indica and O.sativa japonica at 0.07% [10]. These results suggest the range of variation among chloroplast genomes at and above the species level.

Although the draft genomes of C. capsularis and C. olitorius have recently been reported by Islam et al. [11], the chloroplast genome assemblies were not completely resolved. The objectives of the present study were to assemble chloroplast genome sequences for both species, identify variable regions in these genomes,and establish the systematic positions of C.capsularis and C. olitorius in the Malvaceae.

2.Materials and methods

2.1. DNA extraction and sequencing

Fresh leaves of‘Huangma 179'(C.capsularis)and ‘Kuanyechangguo'(C. olitorius), the most widely grown jute cultivars in China, were collected from 30-day-old seedlings growing on the farm of Fujian Agriculture and Forestry University, Fuzhou, China on May 30, 2016. Genomic DNA was extracted from these jute leaves using the CTAB method [12,13]. Both genomes were sequenced following Dong[14],and 138 paired specific primers from that study were used to bridge gaps in the chloroplast sequences.

2.2. Chloroplast genome assembly and annotation

Paired-end sequence reads of 150 nt (PE150) from the Illumina HiSeq 4000 platform were assembled into contigs using SPAdes 3.6.1[15]and SOAPdenovo2[16],respectively.All the contigs were aligned to the reference chloroplast genome(Arabidopsis thaliana(NC_000932))by BLASTN(Basic Local Alignment Search Tool)[17].These contigs were proofread and assembled using Sequencher 4.10 (http://www.genecodes.com). All genes encoding proteins,ribosomal RNAs (rRNAs), and transfer RNAs (tRNAs) were annotated using Dual Organellar Genome Annotator (DOGMA)software[18].The chloroplast genome maps of C.capsularis and C.olitorius were drawn with Organellar Genome DRAW (http://ogdraw.mpimp-golm.mpg.de/index.shtml)[19].

2.3. Comparative and phylogenetic analysis of Corchorus chloroplast genomes

The complete chloroplast genome sequences of C. capsularis and C. olitorius were compared with those of other species(Gossypium raimondii (NC_016668), G. arboretum (HQ325740), G.hirsutum (DQ345959), G.barbadense(AP009123),Hibiscus syriacus(KP688069), and Arabidopsis thaliana (NC_000932)), which were retrieved from NCBI (https://www.ncbi.nlm.nih.gov/), to determine their phylogenetic relationships. Multiple alignments of protein sequences were performed with MUSCLE [20]and a phylogenetic tree was constructed with MEGA 6.0[21].

The aligned chloroplast sequences of the Malvaceae species C. capsularis, C. olitorius, G. hirsutum, G. raimondii, and Theobroma cacao (NC_014676) were used to perform comparative genome analysis using the Shuffle-LAGAN mode of the mVISTA program [22]. Microstructural mutation events were counted and located in these chloroplast sequences. The chloroplast genome sequence of A. thaliana was used as the reference for detecting microstructural mutation events.

2.4. Evaluation of nucleotide variation in the chloroplast genomes

The assembled chloroplast sequences were aligned and manually adjusted using MAFFT software[23].A sliding window scan was used to evaluate nucleotide variation (Pi) across the chloroplast genomes using DnaSP version 5 software[24].The window length was set to 800 bp and the step size was set to 50 bp.

2.5. Microstructural mutation event analysis

To identify microstructural mutations between the C.capsularis and C. olitorius chloroplast genomes, MISA software(Microsatellite identification; http://pgrc.ipk-gatersleben.de/misa/) was used to identify simple sequence repeat markers(SSRs), with the minimum repeat unit limited to 10 for mononucleotides, five for dinucleotides, and four for tri-,tetra-, penta-, and hexanucleotides. The chloroplast genome sequence of C. capsularis was used as a reference for calling InDels and SNPs and identifying transition (Ts) or transversion(Tv) events in SNPs.

2.6. Analysis of genetic diversity in jute germplasm using cpInDel markers

Based on alignment of C. capsularis and C. olitorius chloroplast genomes, cpInDel markers were developed using Primer premiere version 5.0 and that with high polymorphism were chosen to analyze the genetic diversity of jute.A dendrogram was constructed using NTSYSpc v2.21c software by unweighted pair group method with arithmetic averages.

3. Results

3.1. Size, gene content, and organization of C. capsularis and C.olitorius chloroplasts

Fig.1-Gene map of Corchorus capsularis and Corchorus olitorius plastomes.The annotation of the genome was performed with DOGMA.Genes drawn outside of the circle are transcribed clockwise and those inside counterclockwise.The darker gray in the inner circle indicates GC content of the chloroplast genomes of C.capsularis and the lighter gray,AT content.Small single copy(SSC),large single copy(LSC),and inverted repeats(IRa,IRb) are indicated.

The two reference cultivars yielded respectively 2.6 and 1.2 Gb of clean chloroplast sequence for C. capsularis and C. olitorius after removal of low-quality reads. The chloroplast genome assembly of C. capsularis (deposited in GenBank under accession ID MK251464), with a complete length of 161,088 bp, was 678 bp shorter than that of C. olitorius(161,766 bp; accession ID MK251465) (Fig. 1). Both of these genomes are larger than that of the G. raimondii chloroplast(160,161 bp; accession IDNC_016668) [25]. The C. capsularis chloroplast genome contains a pair of inverted repeats(IRs)of 26,063 bp, separated by a large single-copy (LSC) region of 88,615 bp and a small single-copy (SSC) region of 20,347 bp.The C. olitorius chloroplast genome contains a pair of IRs of 25,845 bp,separated by an LSC region of 89,661 bp and an SSC region of 20,415 bp (Table 1). The G + C content of the C.capsularis chloroplast genome is 61% and that of C. olitorius is 62%. Both of the genomes contain 112 different functional genes, including 78 protein-coding, 30 tRNA, and four rRNA genes (Table 2). The gene map is shown in Fig. 1 Among the functional genes,15 genes contain one intron in both species and two genes(ycf3 and clpP) contain two introns.

There were 2417 bp involved in transitions and transversions between the chloroplast genomes of the two Corchorus species, indicating relatively large differences between the genomes(Fig.2,Table S1).Four highly variable loci,including the rps12-clpP-clpP intron,rpl23-trnI,ycf1a,and ycf1b,were precisely located (Fig. 3), and showed much higher Pi values than other regions (Pi > 0.1). All of these loci were located in single-copy rather than in IR regions.

3.2. Phylogenetic analysis of C. capsularis and C. olitorius chloroplast genomes with those of other Malvaceae

The phylogenetic tree for the selected Malvaceae species is displayed in Fig. 4. In the tree, C. capsularis and C. olitorius cluster with Malvaceae.

The overall sequence identity of the chloroplast genomes of the five species C.capsularis,C.olitorius,T.cacao,G.raimondii,and G. hirsutum, is shown in Fig. 5. As expected, lower sequence divergence was observed between the two Corchorus species than between Corchorus, Theobroma, and Gossypium,indicating that the differences between the two Corchorus species were smaller than those among the genera in the Malvaceae. Several intervals, including 4-9 kb, 28-35 kb,44-53 kb, 68-78 kb, 82-90 kb, 116-122 kb, and 130-134 kb,were particularly highly variable among these five members of the Malvaceae. The psbL, petN, psbM, rpl33, trnQ-UUG, trnTGGU, trnS-GGA, trnL-UAA, trnW-CAA, and trnL-UAG genes are found in these polymorphic regions(Fig.5).

Comparison of the C. capsularis and C. olitorius chloroplast sequences with those of T. cacao, G. raimondii, and G. hirsutum revealed breakpoints in the accD locus,which is involved in fatty acid synthesis in C. capsularis and C. olitorius (Fig. 6). The breakpoints suggested that this part of the jute chloroplast genome was transferred to the nuclear genome.Synteny analysis of the accD locus among the nuclear,chloroplast,and mitochondrial(GenBank accessions KT894204.1 and KT894205.1)genomes of C.capsularis(Tables S2)and C.olitorius(Tables S3)supported this hypothesis.

?

3.3. Repeat features in the C. capsularis and C. olitorius chloroplast genomes

Fig.2- Patterns of nucleotide substitutions among Corchorus capsularis and C. olitorius plastomes.The patterns were divided into six types as indicated by the six non-strand-specific base-substitution types(i.e.,numbers of considered G to A and C to T sites for each respective set of associated mutation types).The plastome of C. capsularis was used as a reference.

Fig.3- Evaluation of nucleotide variation of the plastomes of Corchorus capsularis and C.olitorius with C. capsularis as a reference.Window length,800 bp;Step size,50 bp.X-axis,position of the midpoint of a window;Y-axis,nucleotide diversity of each window.

Fig.4- Phylogenetic relationship of Corchorus capsularis and C. olitorius with species of the Malvaceae family based on the chloroplast genome sequences.Arabidopsis thaliana is used as outgroup.The maximum-likelihood phylogenetic tree was constructed by MEGA 6.0 with 1000 bootstrap replicates. Numbers at nodes are bootstrap support values.

Fig.5- Identity plot comparing the chloroplast genomes of Corchorus capsularis,C. olitorius,Gossypium hirsutum,G.raimondii,and Theobroma cacao using Arabidopsis thaliana as a reference sequence.The vertical scale indicates the percentage of identity(50%to 100%),using a 50%identity cutoff.The horizontal axis indicates the coordinates in the chloroplast genome.Genome regions are color-coded as protein-coding,rRNA, tRNA,intron,and conserved non-coding sequences (CNS).

A set of 66 SSRs(CccpSSR001 through 066)specific to the C.capsularis chloroplast genome sequence with repeat lengths of at least 10 nucleotides were identified(Fig.7,Table S4),and 56 SSRs (CocpSSR001 through 056) specific to the C. olitorius chloroplast genome sequence (Table S5) were identified. For example,SSR markers in C.olitorius included repeat sequences with 21 mono-,eight di-,two tri-,11 tetra-,10 penta-,and four hexanucleotide repeats. Among these, the mononucleotide repeats were abundant, with frequencies of 37.5% and 24.2%in C. olitorius and C. capsularis, respectively. Tetranucleotide repeats occurred at lower frequencies.

3.4. Microstructural variants between chloroplast genomes of C. capsularis and C. olitorius

InDels and SNPs in the chloroplast genomes of C.capsularis and C. olitorius are described in Fig. 2 and Tables S1 and S6).SNP markers were the most abundant type of mutation in Corchorus species(Fig.2,Table S1).The 2053 SNPs included 990 transitions and 1032 transversions. Among the transitions,430 were between A and G or T and C and 560 were between C and T or G and A.Among the transversions,296 were between A and C or T and G,299 were between A and T or T and A,308 were between C and A or G and T,and 129 were between C and G or G and C.

A total of 294 InDels (JcpID001 through 294), including 138 deletions, 148 insertions, five repeat contractions, and three repeat expansions were detected(Table S6).The largest InDel(110 bp) was found in the intergenic sequence between petN and psbM in the LSC region.Another large InDel was found in rpl16(38 bp),which encodes a ribosomal protein.

3.5. Classification of accessions using cpInDel markers

Fig.6-Identity plot comparing the region from rbcL to psaI in the chloroplast genomes of Theobroma cacao,Gossypium raimondii,G.hirsutum,Corchorus capsularis,and C.olitorius,using Arabidopsis thaliana as a reference sequence.The vertical scale indicates the percentage of identity(50%to 100%),using a 70%identity cutoff.The horizontal axis indicates the coordinates in the chloroplast genome.Genome regions are color-coded as gene,exon,and conserved non-coding sequences(CNS).

A set of 30 cpInDel markers were selected to characterize the genetic diversity of 47 jute accessions, including 30 of C.capsularis and 17 of C.olitorius(Tables S7,S8).Cluster analysis indicated that the similarity coefficients of the germplasm ranged from 0.43 to 0.96. When the similarity coefficient was at 0.57, the germplasm was divided into two categories, one containing 30 accessions of C.capsularis and aone of C.olitorius and the other containing 16 accessions of C. olitorius (Fig. 8).The clustering result indicated that those InDel markers could be used to identify the genotypes of subspecies in jute. One marker (cpInDel 205) could distinguish C. olitorius and C.capsularis unambiguously, and consisted of a forward primer(5′-GATGAATCATTTCTTCGGAC-A-3′) and a reverse primer(5′-GGTTCTTTACTTCGACAGGGT-3′).

4.Discussion

Molecular markers such as InDels, SNPs, SSRs that have been used in previous studies for species identification and analysis of population structure in the Malvaceae and other plant families also have high potential for phylogenetic analysis. Xu et al. [25]identified variants among 13 Gossypium chloroplast genomes and found divergence among the chloroplast genomes of Gossypium allotetraploids of approximately 0.00159 to 0.00454(Pi).Huang et al. [26]identified 15 molecular markers with greater than 1.5%sequence divergence among five Camellia (Theacea) chloroplast genomes, which will be useful for further phylogenetic analysis and species identification in Camellia. Middleton et al. [27]identified InDel and SNP markers in the chloroplast genomes of 12 Triticeae species and estimated that barley, rye, and wheat diverged around 8.5 million years ago(MYA).These results show that molecular markers based on different chloroplast genome sequences can be useful in evolutionary and systematic studies.The results of the present study represent the first assembly of the complete chloroplast genomes of two cultivated Corchorus species and identification of molecular markers based on microstructural variations, and will expand our understanding of phylogenetic relationships in the Malvaceae.

Although Xiong[3]proposed dispersal routes for C.capsularis and C.olitorius using analyses of morphological traits,molecular markers,and archeological references,the relationship between these species remains under debate.The microstructural events identified in this study provide potential markers for evolutionary studies at both intra- and interspecific levels in jute. First,the 66 and 56 SSR loci identified in the C.capsularis and C.olitorius chloroplast genome sequences will be useful for studies of population structure and evolution. Second, the 294 InDels between the C. capsularis and C. olitorius chloroplast genome sequences indicate differences in micro-inversion events between C.species.Third,2053 SNP markers were detected in the C. capsularis and C. olitorius chloroplast genome sequences.Similar work previously identified 159 SNP sites between the chloroplast genomes of O.sativa and O.nivara[9],591 SNP sites between chloroplasts of Solanum tuberosum and S.bulbocastanum[28], and 330 SNP sites between the chloroplast genomes of Citrus sinensis and Corchorus aurantiifolia[29].Our results suggest that the nucleotide substitution events between the chloroplast genomes of Corchorus species are more abundant than those in rice,potato[28],and orange[29],suggesting that the chloroplast genomes of C.capsularis and C.olitorius differ above the species level.

Fig.7-Statistics of simple sequence repeats in the chloroplast genomes of Corchorus capsularis and C.olitorius plastomes.The X-axis represents different types of simple sequence repeats and the Y-axis represents the number of simple sequence repeats.Black bars represent C. olitorius and the gray bars C. capsularis.

Among molecular markers based on microstructural variations,294 InDel markers were identified for the C.capsularis and C. olitorius chloroplast genomes, among which a marker labeled cpInDel 205 could distinguish the two jute subspecies.The largest 110-bp InDel located in the intergenic sequence between petN and psbM and the large InDel in rpl16 indicate that most of the large InDel events between these species occurred in single-copy regions rather than in IR regions, in agreement with results of previous studies. Song et al. [29]identified 65 InDel markers between the chloroplast genomes of Machilus balansae and M. yunnanensis and found the three largest InDels within the intergenic sequences between ndhF-rpl32(94 bp),trnH-psbA(22 bp)and psbM-trnD(21 bp).Shaw et al. [30]reported that nine regions, including rpl32-trnL(UAG),trnQ(UUG)-5′rps16, 3′trnV(UAC)-ndhC, ndhF-rpl32, psbD-trnT(GGU), psbJ-petA, 3′rps16-5′trnK(UUU), atpI-atpH, and petL-psbE,showed substantial variation and could be appropriate choices for taxonomic analysis. Nashima et al. [31]identified large InDels in pineapple (Ananas comosus) in intergenic spacers such as ndhF-rpl32, rpoB-trnC, trnE-trnT, rpl32-trnL,trnQ-rps16, and protein coding genes such as accD, rpl20, ycf1,and ycf15.

Comparison of the chloroplast genomes of C.capsularis and C. olitorius revealed four highly variable loci including the rps12-clpP-clpP intron, rpl23-trnI, ycf1a, and ycf1b. Because previous studies [32-34]also suggested that the loci clpP,rpl32-trnL,and ycf1 could be hotspots for chloroplast sequence variation in seed plants, these loci have been widely used for phylogenetic studies [26,30]. Thus, the four microstructure variation events observed in Corchorus chloroplast genome sequences might also be considered non-random hotspots[30,35].However,variants in none of these highly variable loci(clpP,rpl32-trnL,and ycf1)identified among other species were identified in Corchorus chloroplast genome sequences in the present study, although two rarely reported highly variable loci (rps8-rpl14 and trnQ-psbI) have been found in Machilus chloroplast genome sequences [29]. In particular, the breakpoints of the accD locus from the C. capsularis and C.olitorius chloroplast genomes were determined by comparison with the chloroplast genomes of other species in the Malvaceae. Although the breakpoints of accD have been illustrated by the synteny analysis of the locus among the nuclear,chloroplast,and mitochondrial genomes of Corchorus,further experimental evidence is needed. Development of greater numbers of specific highly variable markers from Corchorus chloroplast genomes will require a larger number of Corchorus species for chloroplast genome sequence analysis.In contrast, specific primers developed in the four highly variable regions in the Corchorus chloroplast genomes will be better than the universal primers from intergenic regions used previously at evolutionary or phylogenetic analysis[1,3,4]. Developing specific primers based on the four highly variable loci identified in our study could thus improve assessments of phylogenetic relationships among Corchorus species.

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2020.05.010.

Declaration of competing interest

Authors declare that there are no conflicts of interest.

Fig.8- Dendrogram of 47 jute accessions. The dendrogram was obtained with NTSYSpc v2.21c using the unweighted pair group method with arithmetic mean.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (31771369) and the China Agriculture Research System for Crops of Bast and Leaf Fiber, China(nycytx-19-E06).

Author contributions

Liwu Zhang conceived and designed the experiments. Liwu Zhang and Limei Zhang helped to carry out the data analysis and participated in drafting the manuscript.Jianmin Qi helped to prepare the reagents and materials. Shusheng Fang and Liemei Zhang performed data analysis and wrote the paper.All authors read and approved the final manuscript.