Figures
Abstract
Tung tree (Vernicia fordii) provides the sole source of tung oil widely used in industry. Lack of fatty acid composition and molecular markers hinders biochemical, genetic and breeding research. The objectives of this study were to determine fatty acid profiles and develop unigene-derived simple sequence repeat (SSR) markers in tung tree. Fatty acid profiles of 41 accessions showed that the ratio of α-eleostearic acid was increasing continuously with a parallel trend to the amount of tung oil accumulation while the ratios of other fatty acids were decreasing in different stages of the seeds and that α-eleostearic acid (18∶3) consisted of 77% of the total fatty acids in tung oil. Transcriptome sequencing identified 81,805 unigenes from tung cDNA library constructed using seed mRNA and discovered 6,366 SSRs in 5,404 unigenes. The di- and tri-nucleotide microsatellites accounted for 92% of the SSRs with AG/CT and AAG/CTT being the most abundant SSR motifs. Fifteen polymorphic genic-SSR markers were developed from 98 unigene loci tested in 41 cultivated tung accessions by agarose gel and capillary electrophoresis. Genbank database search identified 10 of them putatively coding for functional proteins. Quantitative PCR demonstrated that all 15 polymorphic SSR-associated unigenes were expressed in tung seeds and some of them were highly correlated with oil composition in the seeds. Dendrogram revealed that most of the 41 accessions were clustered according to the geographic region. These new polymorphic genic-SSR markers will facilitate future studies on genetic diversity, molecular fingerprinting, comparative genomics and genetic mapping in tung tree. The lipid profiles in the seeds of 41 tung accessions will be valuable for biochemical and breeding studies.
Citation: Zhang L, Jia B, Tan X, Thammina CS, Long H, Liu M, et al. (2014) Fatty Acid Profile and Unigene-Derived Simple Sequence Repeat Markers in Tung Tree (Vernicia fordii). PLoS ONE 9(8): e105298. https://doi.org/10.1371/journal.pone.0105298
Editor: Manoj Prasad, National Institute of Plant Genome Research, India
Received: January 20, 2014; Accepted: July 22, 2014; Published: August 28, 2014
This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Funding: This work was supported by the “948 Project” of China State Forestry Administration(2012-4-42), National Forestry Public Welfare Industry Research Project of China(201204403), the Key Project of Chinese Ministry of Education (212126), and USDA-ARS Quality and Utilization of Agricultural Products Research Program 306 through CRIS 6435-41000-102-00D and 6435-41000-102-10N. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Tung tree or tung oil tree (Vernicia fordii) is a native woody oil plant in subtropical areas of China. This important economical tree has been grown in China for the production of tung oil or ornamental garden for centuries [1]. Tung tree was introduced to the United States in 1904 [2] and grown mainly in the Southern regions of the United States [2], [3]. Tung seeds contain 50–60% oil with about 80 mole % α-eleostearic acid (9cis, 11trans, 13trans octadecatrienoic acid) [4]. Tung oil is oxidized easily due to the three conjugated double bonds in eleostearic acid. Dried tung oil possesses excellent characteristics such as insulation, acid and alkali resistance and anticorrosion. Unlike other drying oils, tung oil does not darken with age and it becomes a widely used drying ingredient in paints, varnishes, coatings and finishes [5], [6]. Tung oil has also been used as a raw material to produce biodiesel [7], polyurethane and wood flour composites [8], thermosetting polymer [9] and repairing agent for self-healing epoxy coatings [10].
Major efforts have been directed at understanding the genetic control of tung oil biosynthesis. Many tung oil biosynthetic genes have been identified, including those coding for diacylglycerol acyltransferases (DGAT) [11], [12], delta-12 oleic acid desaturase (FAD2) and delta-12 fatty acid conjugase (FADX) [13], acyl-CoA binding proteins [14] and oleosins [15], [16]. The expression of some tung genes has been studied by northern blotting [11]–[13], quantitative real-time PCR (qPCR) [12], [14], [16]–[18] and western blotting [12]. A few tung proteins have been expressed in heterologous systems including E. coli [14], [19], [20], fungi [20]–[22] and Arabidopsis [11], [14]. However, selection of target genes for genetic engineering of plant oils is difficult because oil is biosynthesized by at least 10 enzymatic steps and each step is catalyzed by multiple isozymes [11], [23], [24]. Furthermore, it has been difficult to study tung oil biosynthesis at the protein level because these enzymes are mostly hydrophobic and membrane-localized proteins [19], [20].
Understanding fatty acid composition and genetic diversity among tung tree germplasm resources is essential for tung tree breeding and clonal improvement. A series of elite V. fordii clones were released in China in the 1980s for cultivation on the basis of field survey, collection and evaluation data [1]. However, these economically important germplasm resources were severely damaged by human errors and environmental factors over the past 20 years [1]. In recent years, the importance of V. fordii germplasm resources has been more widely recognized. We initiated germplasm collection in 2007. Some superior germplasm were collected from main distribution areas of V. fordii in China and planted at the Central South University of Forestry and Technology Germplasm Repository.
Microsatellites, also known as simple sequence repeats (SSRs) or short tandem repeats, are repeating sequences of 2–6 base pairs of DNA [25]. They are widely used as molecular markers in genetics and used for studies of gene duplication or deletion, marker assisted selection and fingerprinting [25]–[29]. Therefore, SSR markers could be powerful tools for genetic diversity evaluation, molecular fingerprinting identification, comparative genomics analysis and genetic mapping in tung tree. Tung tree SSR markers have been analyzed in two studies. In one study, authors analyzed 2,407 expressed sequence tag (EST) sequences from the database and identified 22 V. fordii-specific EST-SSR markers [30]. In the other study, 40 polymorphic SSR markers were identified from the V. fordii genomic DNA by AFLP of Sequences Containing repeats protocol [31]. Clearly, there is a need for developing more SSR markers for tung tree improvement.
Great progress has been developed in high throughput sequencing technology, i.e. Next Generation Sequencing, utilizing the Roche/454 Genome Sequencer FLX Instrument, the ABI SOLiD System and the Illumina Genome Analyzer. These new sequencing technologies not only offer fast, cost-effective and reliable approaches for the generation of large expression-data sets in both model and non-model plants with large and complex genomes [32]–[34], but also provide an opportunity to identify and develop unigene-derived genic-SSR markers [35]–[37]. These new genic-SSR markers are considered better markers than genomic SSR markers because they potentially code for functional proteins and can increase the efficiency of marker-assisted selection [38].
The objectives of this study were to evaluate fatty acid profiles and develop unigene-derived SSR markers in 41 tung tree accessions collected from five Provinces in China. Gas-chromatography (GC) analyzed fatty acid profiles in the mature and developing seeds. We utilized Illumina platform-based transcriptome sequencing of cDNA library from developing tung seeds and characterized microsatellites from the transcriptome sequences and developed 15 new polymorphic genic-SSR markers. We also analyzed the expression levels of the identified polymorphic SSR-associated unigenes in developing tung tree seeds and correlated their expression levels with oil content and fatty acid composition in the seeds. The fatty acid composition profiles and novel genic-SSR markers will be useful for biochemical and genetic research and tung tree improvement.
Materials and Methods
Plant Materials
Tung trees (Vernicia fordii, a diploid plant) were collected from Henan (HEN), Hunan (HUN), Hubei (HB), Guizhou (GZ) and Shanxi (SX) Provinces in China. Collecting the samples did not require specific permits because the trees were public-owned and the field studies did not involve protected species. These tung trees were planted at Central South University of Forestry and Technology Germplasm Repository. Vouchers of the sampled accessions were deposited in the University’s Herbarium. Forty-one cultivated accessions at 4-year old were used in this study. The voucher numbers, original locations and geographical coordinates of these 41 tung tree accessions are described in Table 1.
Fatty Acid Analysis
Tung oil fatty acids were extracted from tung seeds and analyzed by GC using a similar method as described by Cao et al [12]. Briefly, tung seeds were dried in an oven (80–90°C), cracked, hulls were removed and the remaining seeds were made into fine powder with a grinder. Total seed oil was extracted with petroleum ether (approximately 10 ml/g), dried and weighted. Seed lipids in the oil extract were converted to methyl esters by KOH-methanol solution (10 mg oil extract in 0.5 ml of 1 M KOH and 40 ml methanol) and extracted with heptane. The organic phase containing lipids was transferred into a vial for GC analysis using a Gas Chromatograph (SHIMADZU GC-2014) equipped with a 60 m long capillary column (FUSED SILICA Capillary Column, SP 2340: 60 m×0.25 mm×0.2 µm film thickness-a non-bonded column highly effective for both high and low temperature separations of geometric isomers of fatty acid methyl esters, dioxins, carbohydrates and aromatic compounds) and a flame ionization detector (FID). The oven temperature was held initially at 50°C for 2 min. The oven temperature was increased from 50°C to 170°C at 10°C/min, held for 10 min, then increased from 170°C to 180°C at 2°C/min, held for 10 min and finally increased from 180°C to 220°C at 4°C/min, held for 10 min. The inlet and detector temperatures were held constant at 250°C and 300°C, respectively. The flow rate was 1 ml/min. The fatty acids in GC peaks were identified by retention times corresponding to those of the fatty acid methyl ester standards (Sigma, St. Louis, MO, USA).
Genomic DNA Isolation
Genomic DNA was isolated from young leaves of the 41 V. fordii cultivated accessions using a DNA Isolation Kit (Tiangen Biotech, Beijing, China).
RNA Isolation
Tung seeds from accession HUN42 were selected because its seeds contained the highest amount of seed oils among the 41 accessions and exhibited a typical lipid profile. The seeds were collected at lipid synthesis initiation phase (stage 1, 60 days after flowering, DAF), peak phase (stage 2, 120 DAF, equivalent to week 7 of the US collection [12]) and ending phase (stage 3, 165 DAF). Total RNA was extracted from the seeds using Micro-to-Midi Total RNA Purification System according to the manufacture’s protocols (Life Technologies Carlsbad, CA, USA). The quality and quantity of the purified RNA samples were characterized initially by agarose gel electrophoresis and NanoDrop ND1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA) and further assessed by RIN (RNA Integrity Number) and rRNA ratio using an Agilent 2100 Bioanalyzer (Santa Clara, CA, USA) as described [17].
cDNA Library Construction
Equal amounts of total RNA from each of the three seed stages were pooled together for better coverage of seed development. Poly-A containing mRNA was purified from 2 mg of total RNA using oligo (dT) magnetic beads and fragmented into 200–500 bp pieces using divalent cations at 94°C for 5 min. The cleaved mRNA fragments were reverse transcribed into first-strand cDNA using SuperScript II reverse transcriptase and random primers (Life Technologies). After double-stranded cDNA synthesis, fragments were end repaired and A-tailed. The final cDNA library was created by purifying and enriching the above products with polymerase chain reaction (PCR).
Unigene Assembly
The cDNA sequences were determined through a paired-end flow cell using an Illumina Solexa HiSeq 2000 Sequencing System at Beijing Genomics Institute (Shenzhen, China). The clean reads after DNA sequencing were de novo assembled using Trinity with default K-mers = 25 [39]. Contigs without ambiguous bases were obtained by conjoining the K-mers in an unambiguous path. The clean reads were mapped back to contigs using Trinity to construct unigenes with the paired-end information. This program detected contigs from the same transcript as well as the distances between these contigs. Finally, the contigs were connected with Trinity, and sequences that could not be extended on either end are defined as unigenes. The original sequencing data are available by contacting the authors.
Microsatellite Analysis
The microsatellites were detected from the assembled unigenes using the MIcroSAtellite tool [40], [41]. The search parameters were set for detection of perfect di-, tri-, tetra-, penta- and hexa-nucleotide SSR motifs with a minimum of six, five, five, four and four repeats, respectively. The numbers of SSR unit type were compiled from all detected di-, tri-, tetra-, penta- and hexa-nucleotide SSR motifs. The frequencies of SSR motifs were compiled according to specific di-, tri-, tetra-, penta- and hexa-nucleotide sequences.
Screening for Genic-SSR Markers
Genomic DNAs from leaves of three tung tree accessions (HUN42, GZ11 and HEN176) were used for testing PCR primers corresponding to 98 loci by agarose gel electrophoresis. PCR primer pairs were designed using Primer Premier 5.0. The parameters for primer design were set for primer length from 18 to 26 nucleotides, PCR product size from 100 to 400 bp and annealing temperature from 50°C to 60°C. The sequences of the forward and reverse primer pairs for 98 unigenes tested, the SSR repeated motifs and the amplicon sizes of PCR products are described (Table 2 and data not shown). PCR reactions were carried out in a total volume of 10 µL containing 10 ng of DNA template in 1×buffer, 2 mM of MgCl2, 200 µM of each dNTPs, 0.2 µM of each primer and 0.25 unit of Taq DNA polymerase (Takara, Japan). The PCR conditions were set at 95°C for 5 min, 35 cycles of 30 s at 94°C, 30 s at 56°C and 30 s at 72°C and a final extension of 10 min at 72°C on a DNA Engine thermal cycler (ABI9700, Applied Biosystems, Foster City, CA, USA). The amplification products were resolved on 2% agarose gels.
Development of Polymorphic Genic-SSR Markers
The loci that generated PCR products with expected sizes on agarose gel were assessed for polymorphisms by high-resolution capillary electrophoresis. PCR products were generated by Touchdown PCR with fluorescently labeled M13 (–21) (5′-TGTAAAACGACGGCCAGT-3′) sequence-tag method [42]. Touchdown PCR was carried out using the following program: 95°C for 5 min; 30 cycles of 30 s at 94°C, 45 s at 56°C and 45 s at 72°C; 10 cycles of 30 s at 94°C, 45 s at 53°C and 45 s at 72°C; and a final extension of 5 min at 72°C. Fluorescently labeled PCR products were initially evaluated by 2% agarose gel electrophoresis and then analyzed by capillary electrophoresis with the GeneScan-500 LIZ Size Standard on an ABI 3730XL sequencer and their sizes were determined with GeneMapper version 4.0 (Applied Biosystems).
Quantitative Real-Time PCR
The expression patterns of the 15 polymorphic SSR-associated unigenes in developing tung seeds were studied by quantitative real-time PCR (qPCR) using SYBR Green method essentially as described [12]. PCR primers in Table 2 were designed to identify polymorphism by amplifying DNA fragments from genomic DNA. Therefore, new sets of primers were designed to analyze the expression levels by amplifying cDNA corresponding to the identified 15 polymorphic genes in the seeds (Table 3). Tung tree EF1a gene was used as the reference gene [43]. The qPCR assay was carried out with three replicates in each reaction using the Bio-Rad CFX system (Bio-Rad). Unigene specific primers are listed in Table 3. PCR was performed in a 20 µL volume containing 2 µl diluted cDNA, 250 nM each primer and 1×SYBR Premix Ex Taq II (TaKaRa). The results were analyzed using the comparative Cq method which uses an arithmetic formula, 2−ΔΔCq, to obtain results for relative quantification [44].
Correlation Analysis
Gray correlation analysis software (V2.1) was used to generate correlation coefficient between gene expression levels and oil content or fatty acid composition [45]. The oil content/fatty acid composition was used as reference series and the mRNA levels of the 15 genes were used as comparison series. The higher correlation coefficient between the mRNA levels and oil content/fatty acid composition means the more positive effect of the gene product on oil content/fatty acid composition.
Genetic and Phylogenetic Analyses
The number of alleles was detected by capillary electrophoresis of the PCR-amplified products. Genetic parameters including the number of alleles (Na), effective number of alleles (Ne, the number of alleles that would be expected in a locus in each population), expected heterozygosity (He, the probability that any two alleles, chosen at random from the population, are different to each other at a single locus) and observed heterozygosity (Ho) were estimated based on the capillary electrophoresis data with POPGENE version 1.31 [46]. Polymorphism information content (PIC) values at each locus were calculated as described [47], [48]. Coefficients of genetic similarity for the 41 accessions were calculated using the SIMQUAL program of NTSYS-pc Version 2.10 (Exeter Software) [49]. The 15 genic SSR markers identified in this study were used initially for the phylogenetic analyses of the 41 accessions. In addition, polymorphism has been studied by other laboratories in tung tree [31]. To expand the phylogenetic analysis of polymorphic SSR-associated genes, we also analyzed polymorphism corresponding to genomic SSRs reported in the published paper [31]. Seventeen genes were confirmed with polymorphism. The names of loci, the sequences of PCR primers and SSR motifs for the confirmation studies are presented in Table S1. Phylogenetic analysis was therefore performed using the 32 polymorphic genes including 15 genes from current studies and 17 genes confirmed from previous studies. Unweighted Pair Group Method with Arithmetic Mean (UPGMA) dendrogram was constructed based on the genetic similarity matrix with the SHAN clustering program [50].
Results
Fatty Acid Composition and Accumulation in Tung Tree Seeds
Forty-one cultivated tung tree accessions used in this study were collected from five Chinese Provinces, planted at Central South University of Forestry and Technology Germplasm Repository and deposited in the University’s Herbarium (Table 1). The major economical value of tung tree is the unique α-eleostearic acid (9cis, 11trans, 13trans octadecatrienoic acid) in tung oil extracted from the seeds. We therefore analyzed the fatty acid profiles of mature seeds from all 41 tung tree accessions. GC typically identified 7 fatty acid peaks in tung oil corresponding to palmitic acid (16∶0), stearic acid (18∶0), oleic acid (18∶1), linoleic acid (18∶2), linolenic acid (18∶3), α-eleostearic acid (18∶3) and β-eleostearic acid (Figure 1A). Alpha-eleostearic acid consisted of the great majority of tung oil with an average of 77.2% of the total fatty acids in the seed oil from the 41 accessions (Figure 1B). The relative abundances of the other 6 fatty acids from the 41 accessions were linoleic acid (7.6%), oleic acid (5.9%), β-eleostearic acid (4.2%), palmitic acid (2.4%), stearic acid (2.3%) and linolenic acid (0.4%) (Figure 1B). The amount of linolenic acid was minimal and undetectable in oils from several accessions (Figure 1B). Tung oil and fatty acid profiles show that the initiation of α-eleostearic acid accumulation started at 60 DAF and peaked at 120 DAF; whereas the amount of other fatty acids declined during seed development (Table 4).
Tung oil was extracted from tung seeds by petroleum ether. Seed lipids were converted to methyl esters by KOH-methanol solution followed by separation and detection by GC-FID. (A) GC separation of fatty acids in tung oil from HUN4241 accessions. The first peak on the chromatogram was the solvent peak. (B) Fatty acid composition in mature tung seeds. The means and standard deviations of GC results from 41 accessions are marked in red color.
High Quality RNA Isolation from Tung Tree Seeds
As an initial step towards the goal of improving the agronomic traits of tung tree and oil contents in the seeds, we began to characterize DNA microsatellites and develop unigene-derived SSR markers. Tung seeds from accession HUN42 were selected for cDNA library construction because its seeds contained the highest amount of tung oil. Total RNA samples were isolated from three developmental stages of tung seeds. The amount and quality of RNA preparations were assessed by Agilent 2100 Bioanalyzer to be sure that high quality RNA was used for construction of cDNA library. These RNA preparations were extremely high quality as indicated by high RNA integrity number (RIN>8) and high 28S:18S rRNA ratio (close to 2.0) in the RNA preparations (Figure S1).
Unigene Identification from Tung Tree Seed Transcriptome
The pooled RNA from the three seed stages were used to construct cDNA library for better representation of the whole seed developmental stages. Sequences of the complete cDNA library were assembled into 81,805 unigenes with a mean length of 945 bp (Figure 2). These unigenes were used to identify microsatellites and develop SSR markers.
The cDNA sequences were determined by Illumina Solexa HiSeq 2000 Sequencing System and de novo assembled using Trinity program.
Types of Microsatellites in Tung Tree Unigenes
MIcroSAtellite tool was used to screen the types of microsatellites from the unigene dataset obtained from tung tree seeds. A total of 6,366 SSRs in 5,404 unigenes contained di-, tri-, tetra-, penta- or hexa-nucleotide repeats (Figure 3A). They represented 6.6% of the 81,805 unigenes in tung seeds with at least one of the considered SSR motifs. The maximum and minimum lengths of the SSR repeats were 179 and 12 nucleotides respectively, with an average length of 16 nucleotides. They were mostly di-nucleotide (47.8%) and tri-nucleotide (44.0%), and less tetra-nucleotide (2.4%), penta-nucleotide (3.3%) and hexa-nucleotide (2.5%) (Figure 3A). The complete list of 6,366 SSRs from 5,404 unigenes with di-, tri-, tetra-, penta- or hexa-nucleotide repeats is presented as “Supporting Information” (Table S2).
The search parameters were set for detection of perfect di-, tri-, tetra-, penta- and hexa-nucleotide SSR motifs with a minimum of six, five, five, four and four repeats, respectively. (A) Distribution of SSR unit type, (B) Frequency of classified SSR motifs.
Frequencies of Microsatellites in Tung Tree Unigenes
The most abundant SSR motif was (AG/CT), which accounted for 31.3% of the total SSR motif (1993 out of 6,366 potential SSRs) (Figure 3B). Other abundant SSR motifs included (AAG/CTT, 13.3%), (AT/AT, 12.0%), (AAT/ATT, 6.7%), (ATC/ATG, 6.5%), (ACC/GGT, 5.7%) and (AC/GT, 4.4%) (Figure 3B). Among the di-nucleotide repeats, the AG/CT motifs showed the most frequency (65.5%, 1993), followed by the AT/TA motifs (25.1%) and AC/GT (9.3%). Among the tri-nucleotide repeats, AAG/CTT motifs were the most common, accounting for 30.2% (847), followed by AAT/ATT (15.2%) and ATC/ATG (14.7%). Other motifs were identified in less significant numbers. The complete list of the frequency of identified SSR motifs is presented as “Supporting Information” (Table S3).
Screening for Genic-SSR Markers by Agarose Gel Electrophoresis
After eliminating undesirable unigenes (sequences were too short and contained unusual GC content and Tm for optimal primer design) and avoiding duplications of those published SSRs [30], [31], 98 loci were selected from the 5,404 unigenes in tung seeds for polymorphic genic-SSR development. PCR primer pairs corresponding to the 98 loci were designed using the criteria described in “Materials and Methods” (Table 2). These primers were used initially to amplify DNA fragments from genomic DNA of three tung tree accessions. Agarose gel shows that the PCR primer pairs for VfUg25262, VfUg31395 and VfUg77143 loci amplified DNA fragments with approximately 200, 150 and 350 bp, respectively, from the genomic DNA of tung tree HUN42, GZ11 and HEN176 accessions (Figure 4, left panels). Similar results from agarose gel electrophoresis revealed that 56 loci generated products of expected sizes (Table 2), whereas 27 loci yielded nonspecific PCR products and 15 loci yielded no PCR products (data not shown).
PCR primers for VfUg25262, VfUg3139 and VfUg77143 loci (unigenes) were used to amplify DNA fragments from genomic DNA of three tung tree accessions (HUN42, GZ11 and HEN176). The PCR products were separated by 2% agarose gel electrophoresis (left panels) and capillary electrophoresis (right panels). Vf and Ug in the locus name represent the abbreviation of tung tree (Vernicia fordii) and unigene. M represents the DNA size standards (DL600 DNA ladder: 100, 200, 300, 400, 500 and 600 bp). (A) VfUg25262 locus, (B) VfUg3139 locus, (C) VfUg77143 locus.
Development of Polymorphic Genic-SSRs by Capillary Electrophoresis
Capillary electrophoresis is more accurate to estimate the sizes of DNA molecules than agarose gel electrophoresis. The positively identified 56 loci by agarose gel electrophoresis were used for polymorphism analysis by capillary electrophoresis. Figure 4 (right panels) clearly shows that capillary electrophoresis separated each band shown on agarose gel (left panels) into two DNA fragments with minor size differences (right panels).
Figure 5 shows an example of using PCR primers for VfUg78868 locus to analyze the numbers and the sizes of this polymorphic SSR-associated gene in 4 tree accessions. PCR assay for VfUg78868 locus amplified a 168 bp fragment from accession GZ131, suggesting a homozygous gene in this accession (Figure 5A). Two different DNA fragments (heterozygous gene) from accession HEN176 (168 and 171 bp), accession HUN42 (168 and 174 bp) and accession HB60 (174 and 204 bp) were detected by this method (Figure 5B–D). The four sizes of PCR fragments separated by capillary electrophoresis (168, 171, 174 and 204 bp) indicated that there were four alleles of VfUg78868 locus in the four tree accessions (Figure 5). This method demonstrated that 41 out of the 56 loci exhibited monomorphism and 15 loci displayed polymorphism among the three tested accessions (Table 2). These 15 genic-SSR markers were validated by capillary electrophoresis using genomic DNA from all 41 V. fordii accessions (Table 5). The number and sizes of all alleles of the 15 loci detected among the 41 tree accessions by capillary electrophoresis are summarized in Table 5. The 15 unigene sequences have been deposited in the GenBank database under the accession numbers shown in Table 5.
The results from VfUg78868 (polymorphic gene) are shown as an example. PCR primers for VfUg78868 locus were used to amplify DNA fragments from genomic DNA of 41 tung tree accessions. The results from four accessions representing the complete set of 4 alleles are presented. The PCR products were separated by capillary electrophoresis. The length of the PCR product on the figure is 18 bp longer than the actual size in Table 3 because an 18 bp fluorescent primer M13 (–21) (5′-TGTAAAACGACGGCCAGT-3′) was used for labeling PCR products. (A) accession GZ131, (B) accession HEN176, (C) accession HUN42, (D) accession HB60.
Functional Annotation of Polymorphic SSR-associated Unigenes
GenBank database search was used to uncover the potential functions of the 15 polymorphic SSR-associated unigenes. The 15 unigene sequences were blasted against the GenBank nonredundant database using BLASTX with an E-value <1×10−5. Thirteen of the 15 sequences showed significant similarities to known genes (Table 5). Ten of the 15 loci putatively coded for a variety of proteins including RNA splicing protein mrs2, transcription factor VIP1-like protein, phosphate-induced protein, anthocyanidin reductase, V-type proton ATPase subunit H-like protein, 3′-N-debenzoyl-2′-deoxytaxol N-benzoyltransferase, plant cadmium resistance 10-like isoform 1, NifU-like protein 4, disease resistance protein RPM1 and protein binding protein (Table 5).
Polymorphic Evaluation of the Genic-SSRs
Genetic analysis estimated that the number of alleles (Na) per locus ranged from two to ten, the expected heterozygosity (He) per locus ranged from 0.140 to 0.692 and the polymorphism information content (PIC) per locus ranged from 0.134 to 0.624 (Table 5). Five of the 15 loci (VfUg25262, VfUg52875, VfUg78868, VfUg77143 and VfUg43685) were high polymorphic (PIC>0.5) and five (VfUg4197, VfUg15890, VfUg79257, VfUg16384 and VfUg15986) were moderate polymorphic (0.25<PIC<0.5) (Table 5).
Gene Expression and Correlation with Seed Oil Content and Fatty Acid Composition
Quantitative real-time PCR was used to study the expression of 15 polymorphic SSR-associated unigenes during tung seed development. Expression of these genes was experimentally confirmed by qPCR using RNA isolated from eight seed development stages (Figure 6). The expression levels of some genes were increased during seed development including VfUg4197, VfUg8413, VfUg15450, VfUg15890 and VfUg15986. The gray correlation analysis software evaluated the relevance between the mRNA levels of these genes and oil content/fatty acid composition (Figure 7). There was not significant correlation between the expression levels and oil content or α-eleostearic acid, the major component of tung oil (Figure 7). However, a strong correlation was obtained between mRNA levels of some genes (VfUg6285, VfUg15450, VfUg16384, VfUg25262, VfUg52875 and VfUg77143) and fatty acid composition (palmitic acid, stearic acid, oleic acid, linoleic acid and linolenic acid) (Figure 7).
The mRNA levels were quantified by qPCR using total RNA from eight seed developmental stages. The relative abundance of mRNA levels at 60 DAF was set at 1.0. qPCR was performed in triplicates by SYBR Green qPCR assay using EF1A gene as the reference gene. The mean and SD from triplicates are presented in the figure.
Gray correlation analysis was performed to generate correlation coefficient between gene expression levels and oil content and fatty acid composition. The higher correlation coefficient between the mRNA levels and oil content/fatty acid composition means the more positive effect of the gene product on oil content/fatty acid composition.
Phylogenetic Analysis of Tung Tree Accessions
Phylogenetic analysis was performed using 32 polymorphic SSR-associated genes including 15 genes identified above and 17 genes confirmed in this study based on a previous publication [31]. Phylogenetic relationships among the 41 V. fordii accessions were assessed by constructing an UPGMA dendrogram using similarity coefficients (Figure 8). The similarity values between the tung tree accessions ranged from 0.64 (between HB139 and GZ57) to 0.89 (between GZ123 and HEN132, GZ123 and HEN165, HB155 and HUN160) (data not shown). The dendrogram shows a mixed picture. Although most accessions from the same geographical location were clustered together, a number of exceptions were present in these 41 tung tree accessions (Figure 8). For instance, two accessions HUN42 and HUN160 collected from Hunan Province did not cluster together.
The dendrogram was generated using the Jaccard’s similarity coefficient based on 32 polymorphic SSR-associated genes including 15 new genes identified in this study and 17 genes confirmed based on a previous publication [31]. The boxed “HUN42” was used for cDNA library construction.
Discussion
Tung tree is an important oil woody plant due to the widely used tung oil from its seeds. In this report, we described 41 tung tree accessions collected from 5 Chinese Provinces and analyzed the lipid profiles of the seeds. We constructed a cDNA library using tung seed mRNA and sequenced them by Illumina platform-based transcriptome sequencing strategy. We discovered 6,366 SSR motifs with 2–6 nucleotide repeats from 5,404 SSR-containing unique putative transcripts among the 81,805 unigenes. We developed 15 new polymorphic genic-SSR markers in 41 cultivated tung tree accessions. Finally, we confirmed the expression of these 15 genes in developing tung seeds and correlated the expression levels with oil content and fatty acid composition in tung tree seeds.
The economical value of tung tree is due to the unique α-eleostearic acid in tung oil from the seeds. Fatty acid profiles of mature seeds from these tung tree accessions consisted of 7 fatty acids including palmitic acid, stearic acid, oleic acid, linoleic acid, linolenic acid, α-eleostearic acid and β-eleostearic acid. The major fatty acid in tung oil was α-eleostearic acid, which accounted for 77% of the total fatty acids in the seeds. This is in agreement with general observations [4], [12]. The relative abundances of the next 5 fatty acids were 2–8% including linoleic acid, oleic acid, β-eleostearic acid, palmitic acid and stearic acid. The amount of linolenic acid was less than 0.5% and undetectable in oils from several tung tree accessions. During tung tree seed development, the relative ratio of α-eleostearic acid was increasing continuously with a parallel trend to the amount of tung oil accumulation in different stages of the seeds while the ratios of other fatty acids were decreasing. These trends of fatty acid profiles reflect the fact that tung oil is the predominant storage component in tung tree seeds. However, the biological significance of tung oil accumulation in the seeds is not clear whether it is related to insect/pathogen resistance and/or affects seed germination.
MIcroSAtellite software discovered approximately 6.6% of the 81,805 unigenes in tung seeds contained at least one of the considered SSR motifs. This percentage is in agreement with previous studies using EST databases, which shows approximately 3–7% of expressed sequences containing putative SSR motifs [41], [51]. Most of the microsatellites in tung trees were di- and tri-nucleotide. Genomic SSRs identified in some plants such as C. pepo and C. moschata contained the same predominant di- and tri-nucleotide unit types [52], [53]. The most abundant SSR motifs in tung tree identified in this study were AG/CT and AAG/CTT. A similar bias towards AG and AAG and against CG repeats has been reported in EST-SSRs of many plants including V. fordii, C. pepo and A. hypogaea [30], [37], [52]. Gonzalez-Ibeas et al. proposed that this may be due to the tendency of CpG sequences to be methylated which might inhibit transcription [54].
We developed 15 new polymorphic genic-SSR markers in 41 cultivated tung tree accessions. All SSR motifs in the 15 SSR markers contained 20 or more nucleotides. These markers were different from those identified previously in V. fordii based on EST sequences and genomic DNA, although the genetic diversity parameters were within a similar range among these studies [30], [31]. The genetic similarity-based dendrogram revealed that most of the 41 accessions from the same geographic region were mainly in the same cluster. Our finding is in agreement with a previous report on genetic diversity of V. Montana [30] and V. fordii using ISSR markers [55]. One of the reasons for this phenomenon is that the accessions clustering together might have originated from the same geographic region and then were planted in different regions.
The polymorphic genic-SSR markers identified here differ from previously reported tung tree SSR markers in another important way because some of the new SSR markers potentially encoded functional genes. Genbank database search identified 10 of the 15 loci putatively coding for RNA splicing protein mrs2, transcription factor VIP1-like protein, phosphate-induced protein, anthocyanidin reductase, V-type proton ATPase subunit H-like protein, 3′-N-debenzoyl-2′-deoxytaxol N-benzoyltransferase, cadmium resistance 10-like isoform 1, NifU-like protein 4, disease resistance protein RPM1 and protein binding protein. These genes were expressed in developing tung seeds. The expression levels of some of the identified genes were well-correlated with fatty acid composition. However, these genes are not directly related to fatty acid biosynthesis in the seeds. Therefore, it was not surprising that there was a lack of positive correlation between the mRNA levels of these genes and tung oil content in the seeds. Nevertheless, these results demonstrate that genic-SSR markers have special features in comparison with genomic SSR markers, because genic-SSR markers are associated with functional genes and may increase the efficiency of marker-assisted selection [38].
Conclusions
We reported 41 accessions of tung tree (Vernicia fordii) collected from 5 Chinese Provinces and analyzed the lipid profiles of the seeds. A total of 81,805 unigenes were identified by transcriptome sequencing in developing seeds, of which 5,404 SSR-containing loci were identified. Out of 98 loci tested, 15 polymorphic genic-SSR markers were developed and characterized. These genes were expressed in developing tung tree seeds. Ten of the 15 loci putatively coded for functional proteins. These molecular markers increase current SSR marker resources and will greatly benefit future studies on genetic diversity, qualitative and quantitative trait mapping and marker-assisted selection studies in tung tree. The lipid profiles in the seeds of 41 tung tree accessions will be valuable for biochemical and breeding studies.
Supporting Information
Figure S1.
RNA quality assessment by Agilent 2100 Bioanalyzer. RNA isolated from seeds at 120 days after flowering (lipid synthesis peak phase) is shown. The quality of RNA isolated from 60 and 165 days after flowering (lipid synthesis initiation phase and ending phase, respectively) were similar (data not shown).
https://doi.org/10.1371/journal.pone.0105298.s001
(PDF)
Table S1.
PCR primers used to confirm 17 polymorphic SSR-associated genes in tung tree (V. fordii). (Microsoft Excel).
https://doi.org/10.1371/journal.pone.0105298.s002
(XLSX)
Table S2.
The complete list of 6,366 SSRs from 5,404 unigenes with di-, tri-, tetra-, penta- or hexa-nucleotide repeats in tung tree (V. fordii). (Microsoft Excel).
https://doi.org/10.1371/journal.pone.0105298.s003
(XLSX)
Table S3.
The complete list of the frequency of identified SSR motifs in tung tree (V. fordii). (Microsoft Excel).
https://doi.org/10.1371/journal.pone.0105298.s004
(XLSX)
Acknowledgments
The authors thank Drs. Jay E. Mellon, John M. Bland and Marina Naoumkina (USDA-ARS) for helpful comments on the manuscript. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture. USDA is an equal opportunity provider and employer.
Author Contributions
Conceived and designed the experiments: LZ XT HC. Performed the experiments: LZ BJ HL ML SW. Analyzed the data: HC LZ CST XT. Contributed reagents/materials/analysis tools: LZ BJ XT HC. Wrote the paper: HC LZ. Revised the manuscript: HC LZ XS XT.
References
- 1. Tan X, Jiang G, Tan F, Zhou W, Lu P, et al. (2011) Research report on industrialization development strategy of Vernicia fordii in China. Nonwood Forest Research 29: 1–7.
- 2. Potter GF (1968) The domestic tung industry today. J Am Oil Chem Soc 45: 281–284.
- 3. Abbott CE (1929) Fruit-bud development in tire tung-oil tree. J Agric Res 38: 679–695.
- 4.
Sonntag NOV (1979) Composition and characteristics of individual fats and oils. In: Swern D, editor. Bailey’s Industrial Oil and Fat Products. New York: John Wiley & Sons. 289–477.
- 5. Kopacz BM (1968) Tung oil research and development at Southern Regional Laboratory. J Am Oil Chem Soc 45: 285–288.
- 6. Pryde EH (1979) Vegetable oil raw materials. J Am Oil Chem Soc 56: 719A–725A.
- 7. Chen YH, Chen JH, Chang CY, Chang CC (2010) Biodiesel production from tung (Vernicia montana) oil and its blending properties in different fatty acid compositions. Bioresource Technology 101: 9521–9526.
- 8. Aranguren MI, Gonzalez JF, Mosiewicki MA (2012) Biodegradation of a vegetable oil based polyurethane and wood flour composites. Polymer Testing 31: 7–15.
- 9. Liu C, Dai Y, Wang C, Xie H, Zhou Y, et al. (2013) Phase-separation dominating mechanical properties of a novel tung-oil-based thermosetting polymer. Ind Crops Prod 43: 677–683.
- 10. Samadzadeh M, Boura SH, Peikari M, Ashrafi A, Kasiriha M (2011) Tung oil: An autonomous repairing agent for self-healing epoxy coatings. Progress in Organic Coatings 70: 383–387.
- 11. Shockey JM, Gidda SK, Chapital DC, Kuan JC, Dhanoa PK, et al. (2006) Tung tree DGAT1 and DGAT2 have nonredundant functions in triacylglycerol biosynthesis and are localized to different subdomains of the endoplasmic reticulum. Plant Cell 18: 2294–2313.
- 12. Cao H, Shockey JM, Klasson KT, Chapital DC, Mason CB, et al. (2013) Developmental regulation of diacylglycerol acyltransferase family gene expression in tung tree tissues. PLoS ONE 8: e76946
- 13. Dyer JM, Chapital DC, Kuan JC, Mullen RT, Turner C, et al. (2002) Molecular analysis of a bifunctional fatty acid conjugase/desaturase from tung. Implications for the evolution of plant fatty acid diversity. Plant Physiol 130: 2027–2038.
- 14. Pastor S, Sethumadhavan K, Ullah AHJ, Gidda S, Cao H, et al. (2013) Molecular properties of the class III subfamily of acyl-coenyzme A binding proteins from tung tree (Vernicia fordii). Plant Science 203–204: 79–88.
- 15. Long H, Tan X, Chen H, Zhang L, Hu J (2010) Cloning and sequence analysis of full-length cDNAs encoding oleosins from Vernicia fordii. Journal of Central South University of Forestry and Technology 30: 31–38.
- 16. Cao H, Zhang L, Tan X, Long H, Shockey JM (2014) Identification, classification and differential expression of oleosin genes in tung tree (Vernicia fordii). PLoS ONE 9: e88409.
- 17. Cao H, Shockey JM (2012) Comparison of TaqMan and SYBR Green qPCR methods for quantitative gene expression in tung tree tissues. J Agric Food Chem 60: 12296–12303 10.1021/jf304690e [doi].
- 18. Cao H, Cao F, Klasson KT (2013) Characterization of reference gene expression in tung tree (Vernicia fordii). Ind Crops Prod 50: 248–255.
- 19.
Cao H, Chapital DC, Shockey JM, Klasson KT (2011) Expression of tung tree diacylglycerol acyltransferase 1 in E. coli. BMC Biotechnol 11: 73. 1472-6750-11-73 [pii]; 10.1186/1472-6750-11-73 [doi].
- 20. Cao H, Chapital DC, Howard OD Jr, Deterding LJ, Mason CB, et al. (2012) Expression and purification of recombinant tung tree diacylglycerol acyltransferase 2. Appl Microbiol Biotechnol 96: 711–727 10.1007/s00253-012-3869-7 [doi].
- 21. Chen YC, Wang YD, Cui QQ, Zhan ZY (2012) FAD2-DGAT2 genes coexpressed in endophytic Aspergillus fumigatus derived from tung oilseeds. Scientific World Journal 2012: 390672 10.1100/2012/390672 [doi].
- 22. Dyer JM, Chapital DC, Kuan JCW, Shepherd HS, Tang FQ, et al. (2004) Production of linolenic acid in yeast cells expressing an omega-3 desaturase from tung (Aleurites fordii). J Am Oil Chem Soc 81: 647–651.
- 23. Dyer JM, Stymne S, Green AG, Carlsson AS (2008) High-value oils from plants. Plant J 54: 640–655.
- 24.
Cao H (2011) Structure-function analysis of diacylglycerol acyltransferase sequences from 70 organisms. BMC Res Notes 4: 249. 1756-0500-4-249 [pii]; 10.1186/1756-0500-4-249 [doi].
- 25.
Hearne CM, Ghosh S, Todd JA (1992) Microsatellites for linkage analysis of genetic traits. Trends Genet 8: 288–294. 0168-9525(92)90256-4 [pii].
- 26. Rafalski JA, Tingey SV (1993) Genetic diagnostics in plant breeding: RAPDs, microsatellites and machines. Trends Genet 9: 275–280.
- 27.
Schlotterer C (1998) Genome evolution: are microsatellites really simple sequences? Curr Biol 8: R132–R134. S0960-9822(98)70989-3 [pii].
- 28. Ellegren H (2004) Microsatellites: simple sequences with complex evolution. Nat Rev Genet 5: 435–445 10.1038/nrg1348 [doi]; nrg1348 [pii].
- 29. Li YC, Korol AB, Fahima T, Nevo E (2004) Microsatellites within genes: structure, function, and evolution. Mol Biol Evol 21: 991–1007 10.1093/molbev/msh073 [doi]; msh073 [pii].
- 30. Xu W, Yang Q, Huai H, Liu A (2012) Development of EST-SSR markers and investigation of genetic relatedness in tung tree. Tree Genetics & Genomes 8: 933–940.
- 31. Pan Y, Pan L, Chen L, Zhang L, Nevo E, et al. (2013) Development of microsatellite markers in the oil-producing species Vernicia fordii (Euphorbiaceae), a potential biodiesel feedstock. Applications in Plant Sciences 1: 1–4.
- 32.
Marioni JC, Mason CE, Mane SM, Stephens M, Gilad Y (2008) RNA-seq: an assessment of technical reproducibility and comparison with gene expression arrays. Genome Res 18: 1509–1517. gr.079558.108 [pii]; 10.1101/gr.079558.108 [doi].
- 33.
Mortazavi A, Williams BA, McCue K, Schaeffer L, Wold B (2008) Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat Methods 5: 621–628. nmeth.1226 [pii]; 10.1038/nmeth.1226 [doi].
- 34.
Nagalakshmi U, Wang Z, Waern K, Shou C, Raha D, et al.. (2008) The transcriptional landscape of the yeast genome defined by RNA sequencing. Science 320: 1344–1349. 1158441 [pii]; 10.1126/science.1158441 [doi].
- 35.
Gupta SK, Gopalakrishna T (2010) Development of unigene-derived SSR markers in cowpea (Vigna unguiculata) and their transferability to other Vigna species. Genome 53: 508–523. g10-028 [pii]; 10.1139/g10-028 [doi].
- 36. Dutta S, Kumawat G, Singh BP, Gupta DK, Singh S, et al. (2011) Development of genic-SSR markers by deep transcriptome sequencing in pigeonpea [Cajanus cajan (L.) Millspaugh]. BMC Plant Biol 11: 17 1471-2229-11-17 [pii];10.1186/1471-2229-11-17 [doi].
- 37.
Zhang J, Liang S, Duan J, Wang J, Chen S, et al.. (2012) De novo assembly and characterisation of the transcriptome during seed development, and generation of genic-SSR markers in peanut (Arachis hypogaea L.). BMC Genomics 13: 90. 1471-2164-13-90 [pii]; 10.1186/1471-2164-13-90 [doi].
- 38. Gupta PK, Rustgi S (2004) Molecular markers from the transcribed/expressed region of the genome in higher plants. Funct Integr Genomics 4: 139–162 10.1007/s10142-004-0107-0 [doi].
- 39.
Grabherr MG, Haas BJ, Yassour M, Levin JZ, Thompson DA, et al.. (2011) Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nat Biotechnol 29: 644–652. nbt.1883 [pii]; 10.1038/nbt.1883 [doi].
- 40.
Sharma PC, Grover A, Kahl G (2007) Mining microsatellites in eukaryotic genomes. Trends Biotechnol 25: 490–498. S0167-7799(07)00239-9 [pii]; 10.1016/j.tibtech.2007.07.013 [doi].
- 41. Thiel T, Michalek W, Varshney RK, Graner A (2003) Exploiting EST databases for the development and characterization of gene-derived SSR-markers in barley (Hordeum vulgare L.). Theor Appl Genet 106: 411–422 10.1007/s00122-002-1031-0 [doi].
- 42. Schuelke M (2000) An economic method for the fluorescent labeling of PCR fragments. Nat Biotechnol 18: 233–234 10.1038/72708 [doi].
- 43. Han X, Lu M, Chen Y, Zhan Z, Cui Q, et al. (2012) Selection of reliable reference genes for gene expression studies using real-time PCR in tung tree during seed development. PLoS One 7: e43084 10.1371/journal.pone.0043084 [doi]; PONE-D-12-13647 [pii].
- 44. Livak KJ, Schmittgen TD (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25: 402–408 10.1006/meth.2001.1262 [doi]; S1046-2023(01)91262-9 [pii].
- 45. Liu S, Cai H, Yang Y, Cao Y (2013) Advance in grey incidence analysis modeling. System Engineering-Theory & Practice 33: 2041–2046.
- 46.
Yeh FC, Yang R-C, Boyle T, Ye Z-H, Mao JX (1997) POPGENE: the user-friendly shareware for population genetic analysis. Molecular Biology and Biotechnology Centre, University of Albert, Canada.
- 47. Botstein D, White RL, Skolnick M, Davis RW (1980) Construction of a genetic linkage map in man using restriction fragment length polymorphisms. Am J Hum Genet 32: 314–331.
- 48.
Tian C, Liang X, Yang M, Zheng H, Dou Y, et al.. (2012) Isolation and characterization of novel genomic and EST-SSR markers in Coreoperca whiteheadi Boulenger and cross-species amplification. Int J Mol Sci 13: 13203–13211. ijms131013203 [pii]; 10.3390/ijms131013203 [doi].
- 49.
Rohlf FJ (2000) NTSYS–pc: Numerical Taxonomy and Multivariate Analysis System. New York: Exeter Software.
- 50. Michener CD, Sokal RR (1957) A quantitative approach to a problem in classification. Evolution 11: 130–162.
- 51.
Guo S, Zheng Y, Joung JG, Liu S, Zhang Z, et al.. (2010) Transcriptome sequencing and comparative analysis of cucumber flowers with different sex types. BMC Genomics 11: 384. 1471-2164-11-384 [pii]; 10.1186/1471-2164-11-384 [doi].
- 52.
Blanca J, Canizares J, Roig C, Ziarsolo P, Nuez F, et al.. (2011) Transcriptome characterization and high throughput SSRs and SNPs discovery in Cucurbita pepo (Cucurbitaceae). BMC Genomics 12: 104. 1471-2164-12-104 [pii]; 10.1186/1471-2164-12-104 [doi].
- 53. Gong L, Stift G, Kofler R, Pachner M, Lelley T (2008) Microsatellites for the genus Cucurbita and an SSR-based genetic linkage map of Cucurbita pepo L. Theor Appl Genet. 117: 37–48 10.1007/s00122-008-0750-2 [doi].
- 54.
Gonzalez-Ibeas D, Blanca J, Roig C, Gonzalez-To M, Pico B, et al.. (2007) MELOGEN: an EST database for melon functional genomics. BMC Genomics 8: 306. 1471-2164-8-306 [pii]; 10.1186/1471-2164-8-306 [doi].
- 55. Li P, Zhang XP, Chen YC, Lu GQ, Zhou GA, et al. (2008) Genetic diversity and germplasm resource research on tung tree (Vernicia fordii) cultivars, investigated by inter-simple sequence repeats. Afr J Biotechnol 7: 1054–1059.