Transcriptome analysis and identification of genes associated with ω-3 fatty acid biosynthesis in Perilla frutescens (L.) var. frutescens
- Hyun Uk Kim†1Email authorView ORCID ID profile,
- Kyeong-Ryeol Lee†2,
- Donghwan Shim3,
- Jeong Hee Lee4,
- Grace Q. Chen5 and
- Seongbin Hwang1
© The Author(s). 2016
Received: 3 February 2016
Accepted: 27 May 2016
Published: 24 June 2016
Perilla (Perilla frutescens (L.) var frutescens) produces high levels of α-linolenic acid (ALA), a ω-3 fatty acid important to health and development. To uncover key genes involved in fatty acid (FA) and triacylglycerol (TAG) synthesis in perilla, we conducted deep sequencing of cDNAs from developing seeds and leaves for understanding the mechanism underlying ALA and seed TAG biosynthesis.
Perilla cultivar Dayudeulkkae contains 66.0 and 56.2 % ALA in seeds and leaves, respectively. Using Illumina HiSeq 2000, we have generated a total of 392 megabases of raw sequences from four mRNA samples of seeds at different developmental stages and one mature leaf sample of Dayudeulkkae. De novo assembly of these sequences revealed 54,079 unique transcripts, of which 32,237 belong to previously annotated genes. Among the annotated genes, 66.5 % (21,429 out of 32,237) showed highest sequences homology with the genes from Mimulus guttatus, a species placed under the same Lamiales order as perilla. Using Arabidopsis acyl-lipid genes as queries, we searched the transcriptome and identified 540 unique perilla genes involved in all known pathways of acyl-lipid metabolism. We characterized the expression profiles of 43 genes involved in FA and TAG synthesis using quantitative PCR. Key genes were identified through sequence and gene expression analyses.
This work is the first report on building transcriptomes from perilla seeds. The work also provides the first comprehensive expression profiles for genes involved in seed oil biosynthesis. Bioinformatic analysis indicated that our sequence collection represented a major transcriptomic resource for perilla that added valuable genetic information in order Lamiales. Our results provide critical information not only for studies of the mechanisms involved in ALA synthesis, but also for biotechnological production of ALA in other oilseeds.
Perilla frutescens, commonly called perilla, is a cultivated crop of the mint family Lamiaceae. Two distinct varieties, P. frutescens var. frutescens, the oilseed crop for source of perilla oil, and P. frutescens var. crisp for the aromatic leafy herb, are cultivated in East Asia countries mainly in Korea, Japan and China . P. frutescens var. frutescens, hereafter called perilla, contains 35–45 % triacylglycerol (TAG) in seeds. It is a rich source of poly unsaturated fatty acids (FA) showing 54–64 % of ω-3 FA (ɑ-linolenic acid, ALA or 18:3) and 14 % ω-6 FA (linoleic acid, LA or 18:2) . Major oil seed crops (e.g., soybean, rapeseed, maize, peanut and sunflower) have relatively low ω-3 FA content (below 10 % in total FAs) in seed TAGs. The ω-3 and ω-6 FAs confer various health benefits for human . The recommended ω-6/ω-3 FA ratio in human diet is 2:1 or lower [4, 5]. However, a typical human diet has high ω-6/ω-3 FA ratio (approximately 15:1) which is considered as a major contributor to cardiovascular diseases . Perilla seed oils have an approximately 0.2:1 ratio of ω-6/ω-3 FAs. This extremely low ratio of ω-6/ω-3 FAs makes perilla a desirable dietary source of vegetable oils . Perilla oil also has many industrial uses, such as for drying oil in paint, varnish and ink manufacturing or as a substitute for linseed oil . Perilla seed cakes are used as animals and birds feed.
Most research for perilla has been focused on identification of metabolites and their biological activities for human health [7, 8]. Some of the genes involved in the biosynthesis of anthocyanins, flavones and monoterpenoids have been studied [9, 10]. Recent reports on the generation of transcriptome using high-throughput sequencing were primarily for identification of genes for anthocyanin pathways associated with red or green leaf varieties of perilla [11, 12]. In contrast, studies on the molecular basis of seed FA and TAG synthesis in perilla have been limited. A seed-specific omega-3 fatty acid desaturase cDNA has been cloned  and characterized in perilla . An oleosin promoter from perilla was found to have a seed-specific activity in transgenic Arabidopsis . Besides perilla, flax (Linum usitatissimum), sacha inchi (Plukenetia volubilis L.), and chai (Salvia hispanica L., a member of mint family Lamiacease) also contain high percentage of ALA in seed oil . Seed transcriptome data of Chai  and sacha inchi (Plukenetia volubilis L.)  have been published, but a few genes contributing to the accumulation of ω-3 FA have been characterized for their expression profiles during seed development.
In this study, we adopted Illumina HiSeq 2000 platform aiming at analyzing the seed transcriptome of perilla. A leaf transcriptome was also included which allows comparison and detection of differentially expressed gene (DEG) in developing seeds of perilla. We have identified 54,079 unique transcripts from a total of 392 mega-base raw sequences, including transcripts for the majority of enzymes involved in lipid biosynthesis and metabolism. We further characterize the expression profiles of 43 key genes involved in FA and TAG in developing seeds and leaf using quantitative PCR (qPCR) assays. To our knowledge this work describes the first seed transcriptome of perilla, and also the first spatial and temporal expression patterns of all known key genes for FA and TAG synthesis in perilla. Our results provide important information for understanding the mechanisms involved in ALA accumulation in perilla.
Results and discussion
Fatty acid profile in developing seeds and leaf
Transcriptome sequencing of perilla and de novo assembly
Summary of sequencing data of P. frutescens seeds and leaf transcriptomes
Total number of raw reads
Total number of clean reads
Functional annotation of perilla transcriptome
We validated and annotated the unique transcripts with BLASTx homology search in Phytozome database. Among total 32,237 annotated unique transcripts, 21,429 transcripts (66.5 %) are highly matched with proteins from Mimulus guttatus (Monkey flower), followed by 1709 (5.3 %), 1431 (4.4 %), and 977 (3.0 %) transcripts matched with proteins from Solanum tuberosum, Solanum lycopericum and Vitis vinifera, respectively. The remaining 6691 (21 %) transcripts matched protein sequences from 37 plant species (Additional file 2: Figure S2). It is not a surprise that most perilla transcripts have high sequence homology to M. guttatus , as both species are under the same Lamiales order. The results allow the translation of genomics and genetics research findings between M. guttatus and perilla.
Analysis of differentially expressed genes (DEG) in perilla developing seeds
To examined the difference in gene expression between seeds and leaves, we performed a DEG analysis using bowtie2 (v2.1.0) . The up- or down-regulated genes were determined by comparison with the level of corresponding genes in leaf. The number of transcripts with > 2-fold change with a false discovery rate (FDR) < 0.01 was presented in Additional file 2: Figure S3. In developing seeds at 1, 2 and 3 WAF, the numbers of up-regulated genes were about 28–48 % less than that of down-regulated genes, showing 1184, 1052 and 1032, respectively; whereas the number of down-regulated genes presented at 1640, 2027 and 2151, respectively (Additional file 2: Figure S3). When seeds reached to maturation at 4 WAF, the number of up- and down-regulated genes had almost identical numbers, 2059 or 2058 (Additional file 2: Figure S3). As we can see, the numbers (1032-1184 counts) of up-regulated genes were similar in seeds at the first three stages (1–3 WAF), and increased to 2-fold (2,059 counts) in 4 WAF. Whereas the numbers of down-regulated genes (2027–2059 counts) were similar in seeds at late three stages (2–4 WAF). The DEG detected in this study provides a global view of seed transcriptome which is important for further investigation of the molecular basis of seed development not only in perilla, but also in other oilseeds.
Clustering of DEGs
Hierarchical clustering was performed with the 6012 DEGs using Another Multidimensional Analysis Package (AMAP) library in R  to examine the similarity and diversity of expression profiles. Similarity of expression pattern of genes was estimated with pearson’s correlation. The results are displayed by Java Treeview (Additional file 2: Figure S4A). The normalized values are represented by different colors, with red representing positive values and green representing negative values. The analysis resulted in twelve clusters (Additional file 2: Figure S4B). Cluster 1 (374 DEGs) and 6 (602) had a similar declining pattern showing a higher level in seeds at 1 WAF, and decreased levels during the rest stages of the development (Additional file 2: Figure S4B). These DEGs may be important for early seed development. DEGs in Cluster 2 (1851) were down-regulated in seeds at all stages indicating that these genes were involved in cell metabolism in leaf. In contrast, DEGs in Cluster 3, 4 and 5 were all up-regulated with slightly different trends showing concave/flat, concave/rise and convex/flat, respectively. These DEGs were likely seed specific genes. Genes in Cluster 7–12 were less differentially expressed between leaf and seeds (Additional file 2: Figure S4B). Cluster 7 (51) had a convex/flat pattern with slightly higher expression levels in seeds at early (1 WAF) and late (4 WAF) stages. Cluster 8 (106) and 10 (118) had similar concave/flat expression patterns and both peaked in 2 WAF seeds. Cluster 9 (478) and 11 (131) were both flat/rise and peaked at 4 WAF. Cluster 12 (37) showed concaved/rise with a peaked expression at 3 WAF. The above variable temporal patterns indicate that multiple mechanisms were involved in regulating gene expression during perilla seed development. Similar temporal patterns of DEGs were also observed in other oilseeds [21–23].
Analysis of seed abundant DEGs in Cluster 3, 4, 5 and 10
Gene Ontology (GO) analysis was further used to classify functions of transcripts in cluster 3, 4, 5, 10 DEG. Using DAVID (http://david.abcc.ncifcrf.gov/tools.jsp) based on the Arabidopsis Information Resource Gene Ontology classification , a total of 2870 DEGs were categorized into 43 functional groups under main GO terms: cellular component, molecular function and biological process. DEGs in all four Cluster 3, 4, 5 and 10 showed similar functions. In the biological process, most transcripts were assigned to “nitrogen compound metabolic process (264 counts,)”, followed by “cellular metabolic process (235)”, “biosynthetic process (221) and “primary metabolic process (141)”. In the cellular components category, the majority of transcripts were associated with the terms “cell periphery (264)”, followed by “protein complex (57)”, and “organelle membrane (31)”. In the molecular function group, the majority of transcripts were related to the terms “ion binding (299)”, “transferase activity (281)” “hydrolase activity (249)”, “oxidoreductase activity (180)” and “transmembrane transporter activity (81)” (Additional file 1: Table S5). Additional file 1: Table S6 lists the top 50 DEGs from Cluster 3, 4, 5 and 10. Among them, the most abundant genes are seed storage proteins (cruciferin, cupins and late embryogenesis abundant (LEA) proteins) and lipids biosynthesis genes, including oleosins, hydroxysteroid dehydrogenase I for TAG biogenesis, acyl carrier protein and FAB2, FAD7/8 and FAD2 for FA synthesis.
Analysis of acyl-lipid genes in developing seeds
The most comprehensive database of plant acyl-lipid genes and pathways have been constructed for Arabidopsis (http://arabidopsisacyllipids.plantbiology.msu.edu/pathways/pathways) . To identify acyl-lipid genes involved in seed oil biosynthesis in perilla, we searched perilla assembled genes using Arabidopsis acyl-lipid genes as queries. Among 975 queries, a total of 540 unique transcripts were identified from perilla transcriptome (Additional file 1: Table S7), which is about 55 % matchup. A similar result (57 % match up) was obtained when searching lesquerella (Physaria fendleri) transcriptome using Arabidopsis acyl-lipid genes as queries . Considering lesquerella and Arabidopsis both belong to the same Brassicaceae, whereas perilla and Arabidopsis are from different order, our results indicate that acyl-lipid genes are conserved among different plant species. Furtherly, we have focused on 43 major genes whose functions are likely responsible for FA and TAG biosynthesis based on our knowledge from model Arabidopsis (Additional file 1: Table S8). Deduced amino acid sequences of perilla genes had varied sequence identities with those of Arabidopsis genes, showing a relatively higher range of 74–92 % for FA biosynthesis than 41–87 % for TAG assembly. Perilla oleosins involved in oil-body formation showed 51–69 % identity compared with those of Arabidopsis. Our data indicate that between perilla and Arabidopsis, genes for FA biosynthesis in plastid are more conserved than those for TAG assembly in ER. The high content of ALA in perilla seed TAG (Fig. 1b and c) is probably resulted from some the genes in ER modified through evolution and become favorable for generating ALA in seed oils.
Genes for FA biosynthesis in plastids
Stearoyl-ACP desaturase (SAD) catalyzes 18:0-ACP to 18:1-ACP in plastid (Fig. 2). Arabidopsis has seven SAD family genes included FAB2 (At2g43710), and FAB2 plays a major role in producing 18:1 . Perilla ortholog (Locus_13564) of Arabidopsis FAB2 was detected in the seed transcriptome, and a homologue of At1g43710, DES6 (Locus_9486), was detected in the leaf transcriptome. Detailed analysis of gene expression confirmed that indeed perilla FAB2 and DES6 were differentially expressed in seeds and leaf, respectively (Fig. 5b). 16:0-ACP, 18:0-ACP and 18:1-ACP are hydrolyzed to the acyl moiety from ACP by two fatty acid thioesterases. FATA and FATB are specific to 18:1-ACP and 16:0 or 18:0-ACPs, respectively. Two fatty acid thioesteases FATA (orthologous Locus_29919 of At3g25110) and FATB (orthologous Locus_6603 of At1g08510) were both detected in perilla seeds and leaf. However, the temporal expression of FATA and FATB were complementary to each other, showing a bell-shaped pattern with high levels at 2 and 3 WAF for FATA and inverted bell curve with high levels at 1 and 4 WAF for FATB (Fig. 5b). The higher expression of FATA at 2–3 WAF would suggest more 18:1 were terminated and released to ER, coinciding with the stages when seeds underwent rapid TAG synthesis. The highest transcript level of FATB detected in seeds at 1 WAF would suggest a swift demand of 16:0 and 18:0 for membrane biosynthesis at the onset of seed development, consisting with the higher levels of 16:0 and 18:0 detected in seeds at 1 WAF (Fig. 1c) . Long chain acyl-CoA synthase (LACS) is located membrane of plastid outer envelope and/or ER and catalyzes free fatty acid to add Coenzyme A (CoA) for producing fatty acyl-CoA. Two perilla LACSs, LACS8 (Locus_3838 ortholog of At2g04350) and LACS9 (Locus_23636 ortholog of At1g77590), were identified. Expression of the LACS9 exhibited a bell-shaped pattern with a maximum level at 2 WAF (Fig. 5b), which may associate with the increased demand of FA-CoA formation in cytosol  when developing seeds entering rapid growth phase. LACS9 was localized in plastid outer envelope . For the LACS8, more transcripts were detected in seeds at 3–4 WAF than 1–2 WAF (Fig. 5b), therefore, the ER-localized LACS8 might be involved in TAG synthesis .
Desaturases associated with ER
TAG biosynthesis in ER
Genes involved in Kennedy pathway and acyl editing reactions
Based on the putative Arabidopsis GPAT9 (At5g60620) sequence , a perilla GPAT9 (PfrGPAT9) transcript (Locus_10180) was found from the transcriptomes showing 81 % sequence identity to At5g60620 (Additional file 1: Table S8). PfrGPAT9 transcript levels were comparable among leaf and developing seeds at different stages, although a bell-shaped pattern peaked at 2 WAF, the overall changes were about 2-fold or less (Fig. 7b). Perilla LPAT2 (PfrLPAT2, Locus_6587), was identified using Arabidopsis LPAT2 (At3g57650) known to be involved in seed TAG biosynthesis . Perilla and Arabidopsis LPAT2s share 81 % sequence identity. (Additional file 1: Table S8). PfrLPAT2 expression showed a continuous increase from 1 to 4 WAF during seed development, and its expression is higher in leave than seeds (Fig. 7b). The spatial and temporal expression patterns of perilla GPAT9 and LPAT2 suggest their constitutive functions with house-keeping roles in both membrane lipid and TAG synthesis. DGAT is the last enzyme in Kennedy pathway and often thought to be the rate limiting step in determining synthesis of TAG . Perilla Locus_14696, Locus_12629 and Locus_1560 were revealed to encode PfrDGAT1, PfrDGAT2, and PfrDGAT3 and showed 79, 67 and 42 % sequence identity with Arabidopsis DGAT1 (At2g19450), DGAT2 (At3g51520) and DGAT3 (At1g48300), respectively, (Additional file 1: Table S8). PfrDGAT1 and PfrDGAT2 were expressed predominantly in seed, whereas DGAT3 expressed both in seeds and leaf at similar levels (Fig. 7b). PfrDGAT1 and PfrDGAT2 are probably involved in TAG biosynthesis in seeds, whereas PfrDGAT3 is a house-keeping enzyme.
As polyunsaturated FAs (PUFA) are major components in TAG of perilla seeds, the acyl editing mechanism [28, 64] would enrich acyl-CoA pool with PUFA-CoAs, facilitating the incorporation of PUFAs into TAGs. Although there are two Arabidopsis LPCATs, (LPCAT1, At1g12640 and LPCAT2, At1g63050) were reported [64, 65], we found only one perilla LPCAT (Locus_43749) in transcriptomes and it expressed both in leaf and developing seeds (Fig. 7b). The finding of PfrLPCAT would suggest acyl-editing through PfrLPCAT likely utilized in perilla.
Genes involved in PC-mediated pathways for TAG biosysnthesis
TAG can be synthesized directly between DAG and PC by Phospholipid:diacylglycerol acyltransferase (PDAT) through acyl-CoA independent pathway [28, 64]. PDAT transfers FA of sn-2 position in PC to sn-3 position of DAG and synthesize TAG (Fig. 7a) [55, 66]. This mechanism has been demonstrated well with a castor PDAT. Castor seed oil contains 90 % ricinoleic acid (18:1OH) which is synthesized on the sn-2 of PC [58, 67]. When a castor PDAT (RcPDAT) was introduced into Arabidopsis expressing a castor fatty acid hydroxylase gene (RcFAH12, [59, 60]), the transgenic Arabidopsis with dual RcFHA12 and RcPDAT enhanced 18:1OH level in TAG [68, 69]. Two perilla PDAT orthologs, PfrPDAT1 (Locus_7255) and PfrPDAT2 (Locus_29208), corresponding to Arabidopsis PDAT1 and PDAT2, respectively, were detected. PfrPDAT1 expressed in seeds and leaves, whereas PfrPDAT2 shows seed-specific expression (Fig. 7b). The spatial and temporal expression profiles of PfrPDATs are similar to that of Arabidopsis PDATs . Our data of PfrPDATs provide molecular basis for further investigation of the role of PfrPDATs in ALA-containing TAG synthesis.
Oil body protein, oleosins
In general, the expression profiles of genes involved in fatty acid and TAG biosynthesis detected by RNAseq analysis (Additional file 1: Table S8) and qPCR (Figs. 3, 4, 5, 6, 7, 8 and 9) are comparable, except for oleison genes. PfrOLN-15 showed a highest expression in developing seeds using qPCR whereas PfrOLN-19 and PfrOLN-19 were highest using RNAseq. The discrepancy of expression level between RNAseq data and qPCR data was likely caused by chimeric transcripts generated by assemble program, which is inevitable in a assemble process purely based on de novo transcriptome data.
Perilla frutescens (L.) var. frutescens, a valuable oilseed crop, contains high amount of ALA in seeds and leaves. Deep sequencing of cDNAs from developing perilla seeds and leaves was carried out to identify genes involved in the synthesis of seed TAG enriched with ALA. A total of 54,079 unique genes from 392 mega-base raw sequences were assembled. The majority (66 %, 21,429 out of 32,237) of the matched genes showed highest homology to Mimulus guttatus genes, confirming the close relationship between the two species. Genes involved in the synthesis of FA and TAG were identified and annotated by detailed sequence alignments. We have identified nearly all of the known genes for de novo FA biosynthesis in plastid, export from the plastid and TAG assembly in ER. In addition, we characterized the expression profiles of 43 key genes in TAG metabolism using quantitative PCR (qPCR). Two ω-3 fatty acid desaturase genes, PfrFAD3 and PfrFAD7/8 were identified as key genes for ALA synthesis in seeds and leaves, respectively. The identification of PfrDGATs, PfrPDATs, PfrPDCT and PfrCPTs provides additional key genes not only for future studies on the mechanisms of ALA-containing TAG synthesis in perilla, but also for use as targets in genetic engineering of other oilseeds to produce a high level of ALA.
Plant materials and RNA extraction
Seeds of Perilla frutescens (L.) var frutescens cultivar ‘Dayudeulkkae’ were obtained from the National Institute of Crop Science, Miryang Republic of Korea. Perilla plants were grown in the greenhouse at temperatures between 18 and 28 °C. After fertilization, developing seeds from 1, 2, 3, 4 weeks and mature leaves were collected, immediately frozen in liquid nitrogen and stored at −80 °C prior to RNA extraction. Total RNAs from developing seeds and leaves of three replicates were extracted using the Plant RNA Reagent (Invitrogen) and treated with DNase I (Takara) according to manufacturer’s instructions. RNA quality was examined using 1 % agarose gel and the concentration was determined using a Nanodrap spectrophotometer (Thermo). The RNA integrity number determined by Agilent 2100 Bioanalyzer was greater than 7.0 for all RNA samples to construct cDNA libraries.
Fatty acid content analysis
The fatty acid content of seeds and leaves were analyzed by gas chromatographic analysis with a known amount of 15:0 fatty acid as an internal standard. Samples were transmethylated at 90 °C for 90 min in 0.3 mL of toluene and 1 mL of 5 % H2SO4 (v/v methanol). After transmethylation, 1,5 mL of 0.9 % NaCl solution was added, and the fatty acid methyl esters (FAMEs) were transferred to a new tube for three sequential extraction with 1.5 mL of n-hexane. FAMEs were analyzed by gas chromatography using a GC-2010 plus instrument (Shimadzu, Japan) with 1 30 m × 0.25 um (inner diameter) HP-FFAP column (Agilent, USA), during which the oven temperature was increased from 170 to 180 °C at 1 °C/min.
cDNA library construction and massive parallel sequencing
RNA-Seq paired end libraries were prepared using the Illumina TruSeq RNA Sample Preparation Kit v2 (catalog #RS-122-2001, Illumina, San Diego, CA). Based on the instruction provided by the kit, mRNAs were purified from total RNA using poly (A) selection, and then chemically fragmented and converted into single-stranded cDNA. Using random hexamer priming, a second strand is generated to create double-stranded (ds) cDNAs. Library construction begins with generation of blunt-end cDNA fragments from ds-cDNAs. Then Adenine nucleotide (A)-base added to the blunt-end in order to make them ready for ligation of sequencing adapters. After the size selection of ligates, the ligated cDNA fragments which contain adapter sequences are enhanced via PCR using adapter specific primers. The library was quantified with KAPA library quantification kit (Kapa biosystems KK4854) following the manufacturer’s instructions. Each library was sequenced using Illumina Hiseq2000 platform, which created 100 bp paired-end sequencing reads.
De novo assembly and unique transcripts annotation
Raw sequencing data composed of 100 bp paired-end reads filtered by Phred quality score (Q ≥ 20) and read length (≥25 bp) with SolexaQA . We used all the sequence reads from different tissue samples to optimize the de novo assembly using the software tools Velvet (v1.2.07)  to assess k-mer sizes and assembled contigs. The contigs were joined into transcript isoforms using Oases (v0.2.08) . Velvet and Oases are based on the de Bruijn graph algorithm. We took several hash length into consideration to select the best de novo assembly. The unique transcripts of perilla were defined by merging the best de novo assembly and validated by direct comparison with gene sequences in the Phytozome (http://www.phytozome.net/) using BLASTx (e-value ≤ 1E-10). The proteins with the highest sequence similarity were retrieved for analysis.
Short read mapping and expression profiles in experimental samples
Reads for each sequence tag were mapped to the assembled unique transcripts using Bowtie software (v2.10) . The number of mapped clean reads for each unique transcript was counted and then normalized with DESeq package in . Only those representative transcripts with mapped reads counts of 1000 or above in at least one experimental sample were retained for further analysis. Fold change and binomial-Test were used to identify differentially expressed genes between each sample. FDR (false discovery rate) was applied to identify the threshold of the p-value in multiple tests and analysis and this value was calculated via DESeq. All correlation analysis, hierarchical clustering was performed using AMAP library in R .
Gene Ontology (GO) analysis was carried out via DAVID (http://david.abcc.ncifcrf.gov/tools.jsp) . The gene lists by annotated TAIR ID of transcripts of up- and down-regulated DEG were analyzed with counts ≥ 5 and FDR ≤ 0.01 of each GO terms.
Total RNA were reverse transcribed with the PrimeScrip™ 1st strand cDNA synthesis kit (Takara, Japan) according to manufacturer’s protocol. Real-time PCR was performed using the SYBR® Premix Ex Taq™ II (Takara, Japan) on the CFX96 Real-Time PCR system (Bio-Rad) with gene-specific primer pairs (Additional file 1: Table S5). Perilla ACTIN (AB002819.1) was used as the internal reference gene. The relative expression value was calculated via the ΔΔCt method.
Full-length cDNA cloning and sequence analysis
A cDNA containing full-length open reading frame (ORF) for FAD2. FAD3 and PDCT were amplified using KOD polymerase from total RNA of developing seeds or leaves samples using primers (Additional file 1: Table S9). PCR products were cloned into pCR-Blunt vector (Invitrogen) for Sanger sequencing. The amino acid sequence alignment of proteins was performed with CLUSTALW program of DNASTAR software with default parameters. Phylogenetic tree was built with the CLUSTALW method with DNASTAR MegAlign program.
This study was conducted with support from the National Institute of Agricultural Science (project no. PJ01007505) and the Cooperative Research Program for Agricultural Science & Technology Development (SSAC, Project No. PJ01108101) provided by the Rural Development Administration (RDA), Republic of Korea. USDA is an equal opportunity provider and employer. Mention of a specific product name by the United States Department of Agriculture does not constitute an endorsement and does not imply a recommendation over other suitable products.
Availability of supporting data
The sequence raw data from this study have been submitted to the NCBI Sequence Read Archive (SRA) (http://www.ncbi.nlm.nih.gov/sra) under the BioProject ID PRJNA287080.
HUK conceived and designed research. HUK and GQC wrote the article. DS and JHL analyzed de novo assembly and differentially expressed genes using Bioinformatics tools. HUK, KYL and SH conducted the experiments and contributed to the study design. All authors read and approved the final manuscript.
The authors declare that they have no competing interests.
Consent for publication
Ethics approval and consent to participate
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Nitta M, Lee JK, Ohnishi O. Asian perilla crops and their weedy forms: Their cultivation, utilization and genetic relationships. Econ Bot. 2003;57(2):245–53.View ArticleGoogle Scholar
- Asif M. Health effects of omega-3,6,9 fatty acids: Perilla frutescens is a good example of plant oils. Orient Pharm Exp Med. 2011;11(1):51–9.View ArticlePubMedPubMed CentralGoogle Scholar
- Simopoulos AP. Omega-3 fatty acids in health and disease and in growth and development. Am J Clin Nutr. 1991;54(3):438–63.PubMedGoogle Scholar
- Wijendran V, Hayes KC. Dietary n-6 and n-3 fatty acid balance and cardiovascular health. Annu Rev Nutr. 2004;24:597–615.View ArticlePubMedGoogle Scholar
- Simopoulos AP. Human requirement for N-3 polyunsaturated fatty acids. Poult Sci. 2000;79(7):961–70.View ArticlePubMedGoogle Scholar
- Brenner DM. Perilla, Botany, Uses and Genetic Resources. J Janick and JE 1993.Google Scholar
- Müller-Waldeck F, Sitzmann J, Schnitzler WH, Graßmann J. Determination of toxic perilla ketone, secondary plant metabolites and antioxidative capacity in five Perilla frutescens L. varieties. Food Chem Toxicol. 2010;48(1):264–70.View ArticlePubMedGoogle Scholar
- Yang SY, Hong CO, Lee H, Park SY, Park BG, Lee KW. Protective effect of extracts of Perilla frutescens treated with sucrose on tert-butyl hydroperoxide-induced oxidative hepatotoxicity in vitro and in vivo. Food Chem. 2012;133(2):337–43.View ArticlePubMedGoogle Scholar
- Yamazaki M, Shibata M, Nishiyama Y, Springob K, Kitayama M, Shimada N, Aoki T, Ayabe S, Saito K. Differential gene expression profiles of red and green forms of Perilla frutescens leading to comprehensive identification of anthocyanin biosynthetic genes. FEBS J. 2008;275(13):3494–502.View ArticlePubMedGoogle Scholar
- Mau CJ, Karp F, Ito M, Honda G, Croteau RB. A candidate cDNA clone for (−)-limonene-7-hydroxylase from Perilla frutescens. Phytochemistry. 2010;71(4):373–9.View ArticlePubMedGoogle Scholar
- Tong W, Kwon SJ, Lee J, Choi IY, Park YJ, Choi SH, Sa KJ, Kim BW, Lee JK. Gene set by de novo assembly of Perilla species and expression profiling between P. frutescens (L.) var. frutescens and var. crispa. Gene. 2015;559(2):155–63.View ArticlePubMedGoogle Scholar
- Fukushima A, Nakamura M, Suzuki H, Saito K, Yamazaki M. High-Throughput Sequencing and De Novo Assembly of Red and Green Forms of the Perilla frutescens var. crispa Transcriptome. PLoS One. 2015;10(6):e0129154.View ArticlePubMedPubMed CentralGoogle Scholar
- Chung CH, Kim JL, Lee YC, Choi YL. Cloning and characterization of a seed-specific omega-3 fatty acid desaturase cDNA from Perilla frutescens. Plant Cell Physiol. 1999;40(1):114–8.View ArticlePubMedGoogle Scholar
- Rao S, Abdel-Reheem M, Bhella R, McCracken C, Hildebrand D. Characteristics of high alpha-linolenic acid accumulation in seed oils. Lipids. 2008;43(8):749–55.View ArticlePubMedGoogle Scholar
- Chung KJ, Hwang SK, Hahn BS, Kim KH, Kim JB, Kim YH, Yang JS, Ha SH. Authentic seed-specific activity of the perilla oleosin 19 gene promoter in transgenic Arabidopsis. Plant Cell Rep. 2008;27(1):29–37.View ArticlePubMedGoogle Scholar
- Sreedhar RV, Kumari P, Rupwate SD, Rajasekharan R, Srinivasan M. Exploring triacylglycerol biosynthetic pathway in developing seeds of Chia (Salvia hispanica L.): a transcriptomic approach. PLoS One. 2015;10(4):e0123580.View ArticleGoogle Scholar
- Wang X, Xu R, Wang R, Liu A. Transcriptome analysis of Sacha Inchi (Plukenetia volubilis L) seeds at two developmental stages. BMC Genomics. 2012;13:716.View ArticlePubMedPubMed CentralGoogle Scholar
- Hellsten U, Wright KM, Jenkins J, Shu S, Yuan Y, Wessler SR, Schmutz J, Willis JH, Rokhsar DS. Fine-scale variation in meiotic recombination in Mimulus inferred from population shotgun sequencing. Proc Natl Acad Sci U S A. 2013;110(48):19478–82.View ArticlePubMedPubMed CentralGoogle Scholar
- Langmead B, Trapnell C, Pop M, Salzberg SL. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 2009;10(3):R25.View ArticlePubMedPubMed CentralGoogle Scholar
- Lucas A. amap: Another Multidimensional Analysis Package. R package version 0.8–12. ed. 2014.Google Scholar
- Wang X, Liu A. Expression of genes controlling unsaturated fatty acids biosynthesis and oil deposition in developing seeds of Sacha inchi (Plukenetia volubilis L.). Lipids. 2014;49(10):1019–31.View ArticlePubMedGoogle Scholar
- Zhang Y, Peng L, Wu Y, Shen Y, Wu X, Wang J. Analysis of global gene expression profiles to identify differentially expressed genes critical for embryo development in Brassica rapa. Plant Mol Biol. 2014;86(4–5):425–42.View ArticlePubMedGoogle Scholar
- Jiang H, Wu P, Zhang S, Song C, Chen Y, Li M, Jia Y, Fang X, Chen F, Wu G. Global analysis of gene expression profiles in developing physic nut (Jatropha curcas L.) seeds. PLoS One. 2012;7(5):e36522.View ArticlePubMedPubMed CentralGoogle Scholar
- da Huang W, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009;4(1):44–57.View ArticleGoogle Scholar
- Li-Beisson Y, Shorrosh B, Beisson F, Andersson MX, Arondel V, Bates PD, Baud S, Bird D, Debono A, Durrett TP, et al. Acyl-lipid metabolism. Arabidopsis Book. 2013;11:e0161.View ArticlePubMedPubMed CentralGoogle Scholar
- Kim HU, Chen GQ. Identification of hydroxy fatty acid and triacylglycerol metabolism-related genes in lesquerella through seed transcriptome analysis. BMC Genomics. 2015;16:230.View ArticlePubMedPubMed CentralGoogle Scholar
- Chapman KD, Ohlrogge JB. Compartmentation of triacylglycerol accumulation in plants. J Biol Chem. 2012;287(4):2288–94.View ArticlePubMedGoogle Scholar
- Bates PD, Stymne S, Ohlrogge J. Biochemical pathways in seed oil synthesis. Curr Opin Plant Biol. 2013;16(3):358–64.View ArticlePubMedGoogle Scholar
- Johnston ML, Luethy MH, Miernyk JA, Randall DD. Cloning and molecular analyses of the Arabidopsis thaliana plastid pyruvate dehydrogenase subunits. Biochim Biophys Acta. 1997;1321(3):200–6.View ArticlePubMedGoogle Scholar
- Konishi T, Shinohara K, Yamada K, Sasaki Y. Acetyl-CoA carboxylase in higher plants: most plants other than gramineae have both the prokaryotic and the eukaryotic forms of this enzyme. Plant Cell Physiol. 1996;37(2):117–22.View ArticlePubMedGoogle Scholar
- Ohlrogge J, Browse J. Lipid biosynthesis. Plant Cell. 1995;7(7):957–70.View ArticlePubMedPubMed CentralGoogle Scholar
- Brown AP, Affleck V, Fawcett T, Slabas AR. Tandem affinity purification tagging of fatty acid biosynthetic enzymes in Synechocystis sp. PCC6803 and Arabidopsis thaliana. J Exp Bot. 2006;57(7):1563–71.View ArticlePubMedGoogle Scholar
- To A, Joubes J, Barthole G, Lecureuil A, Scagnelli A, Jasinski S, Lepiniec L, Baud S. WRINKLED transcription factors orchestrate tissue-specific regulation of fatty acid biosynthesis in Arabidopsis. Plant Cell. 2012;24(12):5007–23.View ArticlePubMedPubMed CentralGoogle Scholar
- Shen B, Allen WB, Zheng P, Li C, Glassman K, Ranch J, Nubel D, Tarczynski MC. Expression of ZmLEC1 and ZmWRI1 increases seed oil production in maize. Plant Physiol. 2010;153(3):980–7.View ArticlePubMedPubMed CentralGoogle Scholar
- Liu J, Hua W, Zhan G, Wei F, Wang X, Liu G, Wang H. Increasing seed mass and oil content in transgenic Arabidopsis by the overexpression of wri1-like gene from Brassica napus. Plant Physiol Biochem. 2010;48(1):9–15.View ArticlePubMedGoogle Scholar
- Kachroo A, Shanklin J, Whittle E, Lapchyk L, Hildebrand D, Kachroo P. The Arabidopsis stearoyl-acyl carrier protein-desaturase family and the contribution of leaf isoforms to oleic acid synthesis. Plant Mol Biol. 2007;63(2):257–71.View ArticlePubMedGoogle Scholar
- Dussert S, Guerin C, Andersson M, Joet T, Tranbarger TJ, Pizot M, Sarah G, Omore A, Durand-Gasselin T, Morcillo F. Comparative transcriptome analysis of three oil palm fruit and seed tissues that differ in oil content and fatty acid composition. Plant Physiol. 2013;162(3):1337–58.View ArticlePubMedPubMed CentralGoogle Scholar
- Schnurr JA, Shockey JM, de Boer GJ, Browse JA. Fatty acid export from the chloroplast. Molecular characterization of a major plastidial acyl-coenzyme A synthetase from Arabidopsis. Plant Physiol. 2002;129(4):1700–9.View ArticlePubMedPubMed CentralGoogle Scholar
- Zhao L, Katavic V, Li F, Haughn GW, Kunst L. Insertional mutant analysis reveals that long-chain acyl-CoA synthetase 1 (LACS1), but not LACS8, functionally overlaps with LACS9 in Arabidopsis seed oil biosynthesis. Plant J. 2010;64(6):1048–58.View ArticlePubMedGoogle Scholar
- Sperling P, Heinz E. Isomeric sn-1-octadecenyl and sn-2-octadecenyl analogues of lysophosphatidylcholine as substrates for acylation and desaturation by plant microsomal membranes. Eur J Biochem. 1993;213(3):965–71.View ArticlePubMedGoogle Scholar
- Okuley J, Lightner J, Feldmann K, Yadav N, Lark E, Browse J. Arabidopsis FAD2 gene encodes the enzyme that is essential for polyunsaturated lipid synthesis. Plant Cell. 1994;6(1):147–58.View ArticlePubMedPubMed CentralGoogle Scholar
- Browse J, McConn M, James Jr D, Miquel M. Mutants of Arabidopsis deficient in the synthesis of alpha-linolenate. Biochemical and genetic characterization of the endoplasmic reticulum linoleoyl desaturase. J Biol Chem. 1993;268(22):16345–51.PubMedGoogle Scholar
- Lee KR, In Sohn S, Jung JH, Kim SH, Roh KH, Kim JB, Suh MC, Kim HU. Functional analysis and tissue-differential expression of four FAD2 genes in amphidiploid Brassica napus derived from Brassica rapa and Brassica oleracea. Gene. 2013;531(2):253–62.View ArticlePubMedGoogle Scholar
- Kang J, Snapp AR, Lu C. Identification of three genes encoding microsomal oleate desaturases (FAD2) from the oilseed crop Camelina sativa. Plant Physiol Biochem. 2011;49(2):223–9.View ArticlePubMedGoogle Scholar
- Yurchenko OP, Park S, Ilut DC, Inmon JJ, Millhollon JC, Liechty Z, Page JT, Jenks MA, Chapman KD, Udall JA, et al. Genome-wide analysis of the omega-3 fatty acid desaturase gene family in Gossypium. BMC Plant Biol. 2014;14:312.View ArticlePubMedPubMed CentralGoogle Scholar
- Khadake R, Khonde V, Mhaske V, Ranjekar P, Harsulkar A. Functional and bioinformatic characterisation of sequence variants of Fad3 gene from flax. J Sci Food Agric. 2011;91(14):2689–96.View ArticlePubMedGoogle Scholar
- Gibson S, Arondel V, Iba K, Somerville C. Cloning of a temperature-regulated gene encoding a chloroplast omega-3 desaturase from Arabidopsis thaliana. Plant Physiol. 1994;106(4):1615–21.View ArticlePubMedPubMed CentralGoogle Scholar
- McConn M, Hugly S, Browse J, Somerville C. A Mutation at the fad8 Locus of Arabidopsis Identifies a Second Chloroplast [omega]-3 Desaturase. Plant Physiol. 1994;106(4):1609–14.PubMedPubMed CentralGoogle Scholar
- Weiss SB, Kennedy EP, Kiyasu JY. The enzymatic synthesis of triglycerides. J Biol Chem. 1960;235:40–4.PubMedGoogle Scholar
- Weiss SB, Kennedy EP. The enzymatic synthesis of triglycerides. J Am Chem Soc. 1956;78(14):3550.View ArticleGoogle Scholar
- Lands WE. Lipid Metabolism. Annu Rev Biochem. 1965;34:313–46.View ArticlePubMedGoogle Scholar
- Bates PD, Browse J. The significance of different diacylgycerol synthesis pathways on plant oil composition and bioengineering. Front Plant Sci. 2012;3:147.View ArticlePubMedPubMed CentralGoogle Scholar
- Stymne S, Stobart AK. Evidence for the reversibility of the acyl-CoA:lysophosphatidylcholine acyltransferase in microsomal preparations from developing safflower (Carthamus tinctorius L.) cotyledons and rat liver. Biochem J. 1984;223(2):305–14.View ArticlePubMedPubMed CentralGoogle Scholar
- Stahl U, Stalberg K, Stymne S, Ronne H. A family of eukaryotic lysophospholipid acyltransferases with broad specificity. FEBS Lett. 2008;582(2):305–9.View ArticlePubMedGoogle Scholar
- Dahlqvist A, Stahl U, Lenman M, Banas A, Lee M, Sandager L, Ronne H, Stymne S. Phospholipid:diacylglycerol acyltransferase: an enzyme that catalyzes the acyl-CoA-independent formation of triacylglycerol in yeast and plants. Proc Natl Acad Sci U S A. 2000;97(12):6487–92.View ArticlePubMedPubMed CentralGoogle Scholar
- Hu Z, Ren Z, Lu C. The phosphatidylcholine diacylglycerol cholinephosphotransferase is required for efficient hydroxy fatty acid accumulation in transgenic Arabidopsis. Plant Physiol. 2012;158(4):1944–54.View ArticlePubMedPubMed CentralGoogle Scholar
- Lu C, Xin Z, Ren Z, Miquel M, Browse J. An enzyme regulating triacylglycerol composition is encoded by the ROD1 gene of Arabidopsis. Proc Natl Acad Sci U S A. 2009;106(44):18837–42.View ArticlePubMedPubMed CentralGoogle Scholar
- Bafor M, Smith MA, Jonsson L, Stobart K, Stymne S. Ricinoleic Acid Biosynthesis and Triacylglycerol Assembly in Microsomal Preparations from Developing Castor-Bean (Ricinus communis) Endosperm. Biochem J. 1991;280:507–14.View ArticlePubMedPubMed CentralGoogle Scholar
- van de Loo FJ, Broun P, Turner S, Somerville C. An oleate 12-hydroxylase from Ricinus communis L. is a fatty acyl desaturase homolog. Proc Natl Acad Sci U S A. 1995;92(15):6743–7.View ArticlePubMedPubMed CentralGoogle Scholar
- Lu C, Fulda M, Wallis JG, Browse J. A high-throughput screen for genes from castor that boost hydroxy fatty acid accumulation in seed oils of transgenic Arabidopsis. Plant J. 2006;45(5):847–56.View ArticlePubMedGoogle Scholar
- Gidda SK, Shockey JM, Rothstein SJ, Dyer JM, Mullen RT. Arabidopsis thaliana GPAT8 and GPAT9 are localized to the ER and possess distinct ER retrieval signals: functional divergence of the dilysine ER retrieval motif in plant cells. Plant Physiol Biochem. 2009;47(10):867–79.View ArticlePubMedGoogle Scholar
- Kim HU, Li Y, Huang AH. Ubiquitous and endoplasmic reticulum-located lysophosphatidyl acyltransferase, LPAT2, is essential for female but not male gametophyte development in Arabidopsis. Plant Cell. 2005;17(4):1073–89.View ArticlePubMedPubMed CentralGoogle Scholar
- Liu Q, Siloto RM, Lehner R, Stone SJ, Weselake RJ. Acyl-CoA:diacylglycerol acyltransferase: molecular biology, biochemistry and biotechnology. Prog Lipid Res. 2012;51(4):350–77.View ArticlePubMedGoogle Scholar
- Xu J, Carlsson AS, Francis T, Zhang M, Hoffman T, Giblin ME, Taylor DC. Triacylglycerol synthesis by PDAT1 in the absence of DGAT1 activity is dependent on re-acylation of LPC by LPCAT2. BMC Plant Biol. 2012;12:4.View ArticlePubMedPubMed CentralGoogle Scholar
- Bates PD, Fatihi A, Snapp AR, Carlsson AS, Browse J, Lu C. Acyl editing and headgroup exchange are the major mechanisms that direct polyunsaturated fatty acid flux into triacylglycerols. Plant Physiol. 2012;160(3):1530–9.View ArticlePubMedPubMed CentralGoogle Scholar
- Stahl U, Carlsson AS, Lenman M, Dahlqvist A, Huang B, Banas W, Banas A, Stymne S. Cloning and functional characterization of a phospholipid:diacylglycerol acyltransferase from Arabidopsis. Plant Physiol. 2004;135(3):1324–35.View ArticlePubMedPubMed CentralGoogle Scholar
- Galliard T, Stumpf PK. Fat metabolism in higher plants. 30. Enzymatic synthesis of ricinoleic acid by a microsomal preparation from developing Ricinus communis seeds. J Biol Chem. 1966;241(24):5806–12.PubMedGoogle Scholar
- Kim HU, Lee KR, Go YS, Jung JH, Suh MC, Kim JB. Endoplasmic reticulum-located PDAT1-2 from castor bean enhances hydroxy fatty acid accumulation in transgenic plants. Plant Cell Physiol. 2011;52(6):983–93.View ArticlePubMedGoogle Scholar
- van Erp H, Bates PD, Burgal J, Shockey J, Browse J. Castor phospholipid:diacylglycerol acyltransferase facilitates efficient metabolism of hydroxy fatty acids in transgenic Arabidopsis. Plant Physiol. 2011;155(2):683–93.View ArticlePubMedGoogle Scholar
- Kim HU, Hsieh K, Ratnayake C, Huang AH. A novel group of oleosins is present inside the pollen of Arabidopsis. J Biol Chem. 2002;277(25):22677–84.View ArticlePubMedGoogle Scholar
- Cox MP, Peterson DA, Biggs PJ. SolexaQA: At-a-glance quality assessment of Illumina second-generation sequencing data. BMC Bioinformatics. 2010;11:485.View ArticlePubMedPubMed CentralGoogle Scholar
- Zerbino DR, Birney E. Velvet: algorithms for de novo short read assembly using de Bruijn graphs. Genome Res. 2008;18(5):821–9.View ArticlePubMedPubMed CentralGoogle Scholar
- Schulz MH, Zerbino DR, Vingron M, Birney E. Oases: robust de novo RNA-seq assembly across the dynamic range of expression levels. Bioinformatics. 2012;28(8):1086–92.View ArticlePubMedPubMed CentralGoogle Scholar
- Anders S, Huber W. Differential expression analysis for sequence count data. Genome Biol. 2010;11(10):R106.View ArticlePubMedPubMed CentralGoogle Scholar