Advertisement

Planta

, Volume 249, Issue 1, pp 271–290 | Cite as

RNA-Seq in the discovery of a sparsely expressed scent-determining monoterpene synthase in lavender (Lavandula)

  • Ayelign M. Adal
  • Lukman S. Sarker
  • Radesh P. N. Malli
  • Ping Liang
  • Soheil S. MahmoudEmail author
Original Article
Part of the following topical collections:
  1. Terpenes and Isoprenoids

Abstract

Main conclusion

Using RNA-Seq, we cloned and characterized a unique monoterpene synthase responsible for the formation of a scent-determining S-linalool constituent of lavender oils from Lavandula × intermedia.

Several species of Lavandula produce essential oils (EOs) consisting mainly of monoterpenes including linalool, one of the most abundant and scent-determining oil constituents. Although R-linalool dominates the EOs of lavenders, varying amounts (depending on the species) of the S-linalool enantiomer can also be found in these plants. Despite its relatively low abundance, S-linalool contributes a sweet, pleasant scent and is an important constituent of lavender EOs. While several terpene synthase genes including R-linalool synthase have been cloned from lavenders many important terpene synthases including S-linalool synthase have not been described from these plants. In this study, we employed RNA-Seq and other complementary sequencing data to clone and functionally characterize the sparsely expressed S-linalool synthase cDNA (LiS-LINS) from Lavandula × intermedia. Recombinant LiS-LINS catalyzed the conversion of the universal monoterpene precursor geranyl diphosphate to S-linalool as the sole product. Intriguingly, LiS-LINS exhibited very low (~ 30%) sequence similarity to other Lavandula terpene synthases, including R-linalool synthase. However, the predicted 3D structure of this protein, including the composition and arrangement of amino acids at the active site, is highly homologous to known terpene synthase proteins. LiS-LINS transcripts were detected in flowers, but were much less abundant than those corresponding to LiR-LINS, paralleling enantiomeric composition of linalool in L. × intermedia oils. These data indicate that production of S-linalool is at least partially controlled at the level of transcription from LiS-LINS. The cloned LiS-LINS cDNA may be used to enhance oil composition in lavenders and other plants through metabolic engineering.

Keywords

Essential oils Lavandula S-Linalool S-Linalool synthase RNA-Seq Terpene synthase 

Abbreviations

EOs

Essential oils

GPP

Geranyl pyrophosphate

mTPS

Monoterpene synthase

LiS-LINS

L. × intermedia S-linalool synthase

LiR-LINS

L. × intermedia R-linalool synthase

NPP

Neryl pyrophosphate,

qPCR

Quantitative real-time PCR

TPS

Terpene synthase

Introduction

The genus Lavandula (Lamiaceae) includes over 39 species and numerous hybrids of woody aromatic shrubs, among which Lavandula angustifolia, Lavandula latifolia, and their natural hybrid Lavandula × intermedia are economically important and cultivated worldwide for the production of their essential oils (EOs). The EOs of these plants are composed mainly of monoterpenes, although a few sesquiterpenes can also be detected in the oil. Though the exact function for most of these metabolites is unknown, they are believed to have various physiological and ecological roles, for example, in allelopathy, plant defense, and pollinator attraction (Erland and Mahmoud 2015; Aprotosoaie et al. 2017). Lavender EOs are of substantial economic value as they are extensively used in cosmetic, personal care products, and alternative medicines, among others (Cavanagh and Wilkinson 2002; Denner 2009; Erland and Mahmoud 2015). Of the various monoterpenes in lavender EOs, linalool, linalyl acetate, borneol, camphor, and 1,8-cineole are the most abundant, and determine the quality of the oil (Lis-Balchin 2002; Aprotosoaie et al. 2017). While linalool and linalyl acetate are desired constituents of the oil, camphor and to a lesser extend borneol and 1,8-cineole contribute undesired odor. In this context, EOs with a high linalool/linalyl acetate-to-camphor ratio, for example, EO from L. angustifolia, are considered to be of high quality, and are used in fine perfumes and cosmetic products, and in aromatherapy (Aprotosoaie et al. 2014, 2017).

Linalool—one of the two top desirable monoterpenes found in Lavandula oils—exists in two isomers, R and S types. R-linalool that has a woody lavender-like odor is the dominant monoterpene in Lavandula oils, while S-linalool (sweet and pleasant odor) occurs in a lesser amount (Table 1) (Renaud et al. 2001). Depending on the species, the previous studies showed that the enantiomeric distribution of R- and S-linalool varies from 80 to 93 and 7 to 20% in L. angustifolia oils, and 87 to 97 and 3 to 13% in L. × intermedia oils, respectively (Renaud et al. 2001; Özek et al. 2010). Although R-linalool is an abundant and a major scent-determining monoterpene in Lavandula oils, there is an interest of enhancing S-linalool composition in oils through metabolic engineering, for example, overexpression of the Clarkia berweri S-linalool synthase cDNA in spike lavender (Mendoza-Poudereux et al. 2014).
Table 1

Content of linalool and its enantiomer compositions in previously reported flower tissue-derived oil from the two major commercial lavender species, L. angustifolia and L. × intermedia (Renaud et al. 2001)

Commercial lavender species/cultivar

Linalool in flower oil (%)

Linalool enantiomers (%)

R-Linalool

S-Linalool

L. angustifolia

2.8–53.4

80.9–92.7

7.3–19.1

L. × intermedia

27.9–38.5

87.2–97.3

2.7–12.8

L. × intermedia cv. Grosso

27.9

89.8

10.2

The biosynthesis of EO constituents in lavenders takes place in the secretory cells of the glandular trichomes that cover the surfaces of all above-ground plant parts (Turner et al. 2000; Iriti et al. 2006; Zuzarte et al. 2010). The process begins with the formation of isopentenyl diphosphate (IPP) and its allylic isomer, dimethylallyl diphosphate (DMAPP) through the cytosolic mevalonate (MVA), and the plastidial 2-C-methyl-d-erythritol 4-phosphate (MEP) pathways (Mendoza-Poudereux et al. 2015; Liao et al. 2016; Rehman et al. 2016). IPP and DMAPP are then condensed to form geranyl diphosphate (GPP) in plastids and farnesyl diphosphate (FPP) in the cytosol (Vranová et al. 2012). Subsequently, GPP and FPP are converted to mono- and sesquiterpenes, respectively, through the action of specific terpene synthase enzymes (TPSs) (Christianson 2006; Degenhardt et al. 2009; Demissie et al. 2011, 2012; Sarker et al. 2013; Adal et al. 2017).

Using various molecular tools, several TPSs including R-linalool synthase from different lavender species have been identified and functionally characterized (Landmann et al. 2007; Demissie et al. 2011, 2012; Sarker et al. 2013; Jullien et al. 2014; Benabdelkader et al. 2015; Adal et al. 2017). One of the common approaches that have been widely used to identify highly expressed TPSs from Lavandula is mining of Expressed Sequence Tags (EST) databases derived from EO-producing tissues. Previously, we have developed a database of over 22,000 ESTs, derived from whole leaf and flower tissues, as well as isolated floral glandular trichomes to identify genes involved in isoprenoid metabolism (Lane et al. 2010; Demissie et al. 2012). Although this database has been and will likely continue to be instrumental in cloning TPSs and other genes (Demissie et al. 2011, 2012; Sarker et al. 2012, 2013; Demissie et al. 2013; Adal et al. 2017), it does not contain sequence information for genes for which transcripts are not highly abundant, including the S-linalool synthase. On the other hand, RNA-Seq has been widely used as a cost-effective and efficient approach for obtaining deep sequence information for plant transcriptomes (Wang et al. 2009; Grabherr et al. 2011), which could enable to identify genes that have low expression. Thus, using a homology-based cloning approach and sequence information derived from flower and leaf tissues of Lavandula species via RNA-Seq, this study aimed to clone and functionally characterize S-linalool synthase cDNA from L. × intermedia. Here, we report the expression and functional characterization of LiS-LINS cDNA in E. coli. The recombinant LiS-LINS catalyzed the conversion of GPP to the sole product, S-linalool. LiS-LINS protein exhibited a unique amino acid sequence, lacked the well-conserved RR(x)8W motif, and was grouped under TPS-g subfamily of terpene synthase genes. We also used homology-based protein modeling to study the 3D structure of the protein, and investigated the subcellular localization and spatial expression of LiS-LINS in vivo.

Materials and methods

Plant materials and chemicals

The three economically important lavender species L. angustifolia cv. Lady, L. latifolia, and their natural breed L. × intermedia cv. Grosso were used in this study. L. × intermedia and L. angustifolia were grown at a field site at the University of British Columbia, Okanagan campus (Kelowna, BC, Canada). L. latifolia flower and leaf tissues were kindly provided by Dr. Tim Upson (Cambridge University, UK). Tissue samples were flash frozen in liquid nitrogen and used for total RNA isolation. Authentic standards such as racemic linalool (R/S-linalool) and S-linalool (75% linalool extracted from orange) for assay product identification were purchased from Sigma (http://www.sigmaaldrich.com/) and from TREATT, Lakeland, Florida, USA, respectively. Geranyl diphosphate (GPP), neryl diphosphate (NPP), farnesyl diphosphate (FPP), isopernyl diphosphate (IPP), and dimethylallyl diphosphate (DMAPP) substrates were purchased from Echelon Biosciences (http://www.echelon-inc.com/). Other chemical reagents used in this study were obtained from Sigma unless otherwise noted.

RNA-Seq, de novo assembly, and annotation

Library preparation and transcriptome sequencing using the Illumina HiSeq 2000 platform (Illumina Inc., San Diego, CA, USA) was performed at Plant Biosis (Lethbridge, Alberta, Canada) using a standard protocol. After removing low quality and short reads (length ≤ 20 bp) and adaptor contamination, Illumina reads from leaf and flower tissues of three lavender species (L. angustifolia, L. × intermedia and L. latifolia) along with previously developed EST sequences (L. angustifolia flower and leaf and, L. × intermedia cv Grosso glandular trichomes from floral tissues) (Lane et al. 2010; Sarker et al. 2013) were assembled by CLC Genomics Workbench version 8.03 (QIAGEN) to generate nine transcriptomes using various combinations of word (30–60) and bubble (50–600) sizes. Based on the presence of known terpene-related genes in a given assembly, five assemblies were selected and combined and cleaned up (by removing redundant sequences using CD-HIT-EST (Li and Godzik 2006) with the threshold of 0.90 identity) to generate a final transcriptome database. The transcriptome was validated by mapping the pre-processed Illumina sequence reads back to the assembled transcriptome using CLC Genomics Workbench.

The assembled sequences were annotated using BLASTx and BLASTp against UniProKB, NCBI non-redundant (nr) protein, Arabidopsis protein and Phytozome v10 databases (Goodstein et al. 2012). We also annotated the transcriptome using Trinotate v3.0.0 along with transdecoder v2.1.0 (Haas et al. 2013) using Compute Canada’s facilities (westgrid.ca). BLAST hits with significant matches (e value ≤ 1e−5) were employed for additional inferences about gene function using plant-specific GO terms obtained from Arabidopsis TAIR database and validated using PANTHER version 11 database (Mi et al. 2017). Interproscan analysis (Jones et al. 2014) was also conducted to further classify protein families.

TPS sequence analysis and cloning of LiS-LINS cDNA

TPS candidates were identified from Lavandula transcriptome database based on their sequence similarity to known TPSs genes. Briefly, cDNA sequences for known monoterpene synthases (mTPSs) from lavender species belonging to the genus Lavandula and S-linalool synthases from different plant species were used to BLAST against the database, followed by manual analysis targeting the common mTPS conserved motifs. The open reading frames (ORF) for selected candidate were further analyzed for the presence of potential transit peptide using different bioinformatics tools, including TargetP 1.1 (http://www.cbs.dtu.dk/services/TargetP/), ChloroP 1.1 (http://www.cbs.dtu.dk/services/ChloroP/), PWoLF PSORT (https://wolfpsort.hgc.jp/), SignalP 4.1 (http://www.cbs.dtu.dk/services/SignalP/), and iPSORT (http://ipsort.hgc.jp/). Gene-specific cloning primers were designed for heterologous expression in E. coli, such as LiS-LINS forward 5′-CTGGCATATGAAGCAGTGTCGTATCAAC-3′ and reverse 5′-ATATAGAATTCAAGCTCTGAATCAGGTTTGGC-3′. The coding sequence for LiS-LINS excluding the transit peptide and the stop codon was then amplified from L. × intermedia flower cDNA and cloned into the NdeI/EcoRI sites (underlined on the above primer sequences) of pET41b(+) by sticky-end cloning (Zeng 1998).

Genomic structure and number of copies of S- and R-LINS were determined from the draft genome of L. angustifolia (Malli et al., unpublished) using the corresponding cDNAs. The coding sequences (CDS) or exons, introns, and upper/downstream sequences of the detected sequences were then determined using GSDS—gene structure display server2.0 (Hu et al. 2015).

Multiple sequence alignments of deduced amino acids of LiS-LINS and other closely related TPSs (Table S1) were generated using web-based the CLUSTAL O (1.2.4) (Sievers et al. 2011). Furthermore, to infer the evolutionary history of LiS-LINS, non-rooted neighbor-joining phylogenetic trees were constructed based on deduced amino acids of TPSs identified from the genus Lavandula, and other selected plant TPS-g subfamilies available in public databases with 1000 bootstrap replicates in MEGA7 (Kumar et al. 2016).

3D modeling of linalool synthases

Homology-based 3D protein modeling and prediction of substrate-binding residues were performed for LiS-LINS and R-linalool synthase from L. angustifolia (LaR-LINS) (Landmann et al. 2007) against S. officinalis bornyl pyrophosphate synthase (SoBPPS) (Whittington et al. 2002) using the template-based protein structure modeling web server, RaptorX (Källberg et al. 2012). The model quality was then assessed using ProSA-web (Wiederstein and Sippl 2007). To assess the model similarity between LiS-LINS and LaR-LINS, pairwise structural alignment was made using RaptorX Structural Alignment server (Wang et al. 2013). Virtual geranyl diphosphate (GPP) was then docked into the active site pocket of LiS-LINS and LaR-LINS 3D structure using 1-Click Docking program (https://mcule.com/apps/1-click-docking/). Best docked GPP orientation was chosen based on the best energetic fit in the designated active site pocket. The 3D structure and docking results were visualized and further analyzed using PyMOL v1.1 (The PyMOL Molecular Graphics System, Schrödinger, LLC).

Localization studies

To experimentally confirm the subcellular localization of LiS-LINS (GenBank: MG870571), the first 50 amino acids of LiS-LINS, designated hereafter as LiS-LINS-TP(50aa), were fused upstream of GFP in the cloning sites XbaI and SalI of the p326::SGFP vector harboring CaMV 35S promoter (a gift from Dr I. Hwang, Pohang University of Science and Technology, Korea). R-Linalool synthase from our L. × intermedia (LiR-LINS, GenBank.: MH375883) ESTs database exhibited about 95% identity at amino acid level with previously cloned R-linalool synthase from L. angustifolia (Landmann et al. 2007). Thus, the first 50 amino acids of R-linalool synthase from L. × intermedia (LiR-LINS-TP(50aa)) was also fused upstream of GFP in the cloning sites XbaI and SalI of the p326::SGFP. Primer sets used for cloning of the transit peptides into p326::SGFP vector were: LiS-LINS-TP(50aa) forward 5′-ATCTCTAGATGGCACTTCCATGCAATATC-3′ and reverse 5′-AATAGTCGACGTTGATACGACACTGCT-3′; and LiR-LINS-TP(50aa) forward 5′-GCCGTCTAGATGTCGATCAATATCAACA-3′ and reverse 5′-GCCAGTCGACCTTATACTGAGAATTGAG-3′. The restriction sites for XbaI (forward primer) and SalI (reverse primer) were underlined. The integrity of the resulting constructs was then confirmed by Sanger sequencing prior to transfecting to N. benthamiana protoplasts. N. benthamiana protoplasts were isolated and transformed following a modified procedure described by Yoo et al. (2007). Ten micrograms of DNA of each construct was used for PEG-mediated transformation of 300 µl of ice-cold protoplasts. Transient expression of GFP fusion proteins was observed 24 h post-PEG-mediated transfection using inverted fluorescence microscope (Zeiss Axiovert 200M).

E. coli expression, protein purification, and enzymatic assays of LiS-LINS

The ORF of LiS-LINS excluding the transit peptide was expressed in E. coli RosettaTM (DE3)plysS cells (EMD Chemicals, Darmstadt, Germany). E. coli cells were grown in LB media supplemented with 30 mg/l kanamycin and 34 mg/l chloramphenicol at 37 °C to an OD600 of 0.5–0.6, followed by induction with 0.4 mM IPTG, and then grown at 20 °C overnight. After harvesting, the pelleted cells were resuspended in lysis buffer (50 mM NaH2PO4, 300 mM NaCl, and 10 mM imidazol, pH 8.0) and were further disrupted by sonication. The crude-soluble protein was purified using PerfectPro Ni–NTA-agarose chromatography (5Prime Gmbh, Germany), in which Ni–NTA agarose bound to His-tag recombinant proteins according to the manufacturer’s instructional manual. Partially purified recombinant protein was then quantified through a Bradford assay prior to using it for further enzymatic assays.

The activity of partially purified recombinant LiS-LINS protein was assayed with each of GPP, NPP, FPP, DMAPP, and a combination of IPP and DMAPP as substrates. The assay reactions were prepared in a final volume of 500 µl, containing an enzyme reaction buffer [2 mM dithioerythreitol, 12.5% (v/v) glycerol, 1 mM MgCl2, and 1 mM MnCl2], MOPS (pH 6.5) buffer, purified enzyme, and substrates, and were overlaid with 400 µl of pentane and incubated at 30 °C for 30 min. The assays were frozen at − 80 °C, and then, the products were extracted with pentane according to the previous reports (Demissie et al. 2011, 2012; Adal et al. 2017). As control, heat-denatured LiS-LINS enzyme and purified proteins from E. coli lysate cells harboring the pET41b(+) lacking LiS-LINS were assayed with substrates with similar conditions. Assay product was identified using gas chromatography–mass spectrometry (GC–MS) following methods described in Demissie et al. (2012). Furthermore, linalool enantiomers were analyzed by GC–flame ionization detector (GC–FID) using a 30 m × 0.25 mm fused silica capillary column coated with Chirasil-DEX (cyclodextrin directly bonded to dimethylpolysiloxane) (Agilent Technologies Canada Inc, Mississauga, ON, CAN) according to Galata et al. (2014). Authentic standards of linalool isomers were also run on GC–MS and GC–FID under similar conditions to identify the assay products.

Relative expression of linalool synthases

Transcript levels for LiS-LINS (GenBank: MG870571) and LiR-LINS (GenBank: MH375883) were quantified by qPCR using the StepOne Plus Real-Time PCR system (Applied Bioscience). Briefly, total RNA was extracted from each of flower, leaf, and root sample of L. × intermedia, cv Grosso using RNeasy Plant Mini Kit (Qiagen) according to the manufacturer’s protocol, and was then reverse transcribed using iScript cDNA synthesis Kit (Bio-Rad). β-Actin was used as an internal reference. qPCR was performed in a final reaction volume of 10 µl that consisted of 5 µl SYBR Select premix, 0.6 µM of each primer, and approximately 150 ng of cDNA. The qPCR primers were: LiS-LINS forward 5′-CAATCCCATCAACTCCCTAAA-3′ and reverse: 5′-GCTCTACTTCGTCTCTGAATTT-3′; LiR-LINS forward 5′-ACACGCACGACAATTTGCCA-3′ and reverse: 5′-AGCCCTCCAATGAAGTGGGAT-3′ and β-actin reference gene forward 5′-TGTGGATTGCCAAGGCAGAGT-3′ and reverse: 5′-AATGAGCAGGCAGCAACAGCA-3′. The reactions were then amplified using the following program: (1) holding stage at 50 °C for 2 min, and 95 °C for 2 min, followed by (2) amplification stage of 50 cycles at 95 °C for 3 s and 60 °C for 30 s, as well as (3) a final melting curve stage at 95 °C for 15 s and 60 °C for 1 min. Then, melting dissociation was performed from 60 to 95 °C at 0.3 °C/s melt rates with a smooth curve setting. The PCR efficiency was calculated using LinRegPCR for all the primers used in this experiment, and updated in the SteponePlus data analysis software (Ruijter et al. 2009). The generated data were analyzed using the Livak method (Livak and Schmittgen 2001), expressed as 2−ΔΔCT. For both genes, three biological and three technical replicates were used in qPCR assays.

Results

RNA sequencing, de novo assembly, and annotation

Total RNA extracted from flower and leaf tissues of L. angustifolia, L. latifolia, and L. × intermedia plants was sequenced via RNA-Seq using the Illumina platform to generate approximately total 30 million raw reads (29 bp single read). Removal of the adaptor sequences, ambiguous nucleotides, and low-quality reads yielded a total of ~ 28 million high-quality reads. These reads, along with previously generated 22,592 EST sequences (Lane et al. 2010; Demissie et al. 2012), were assembled (together) using CLC Workbench 8.0.3 (QIAGEN). All the assembled files were blasted against the known lavender TPS genes to initially verify the quality of the assembly. Best assembled transcripts were then combined, and redundant sequences were removed using CD-HIT-EST (Li and Godzik 2006) to generate 101,618 unique contigs (unigenes). The assembly results were validated by mapping the reads back to the assemblies resulting in 97% of the reads uniquely mapped back with over 92% of which were properly paired. An overview of the sequencing and assembly results is outlined in Table 2. The N50 value used to evaluate de novo assembly was 831. The unigene pool contained over 100,000 sequences of > 200 bp long, including 3081 unigenes > 2000 bp, 15,211 unigenes between 1001 and 2000 bp, and 35,044 unigenes between 501 and 1000 bp (Fig. 1). Unigene sequences were blasted against different public databases including NCBI non-redundant proteins (nr), Swiss-Prot proteins, Arabidopsis proteins, and phytozome plant proteins (Table 3). We annotated the unigenes further against the public databases using Trinotate along with transdecoder annotation (Haas et al. 2013), and the results (Table 3) were compared with these blast outputs. The phytozome annotation showed a better coverage, particularly in pfam and GO, and yielded over 75,400 annotated sequences accounting for 74.6% of the total unigenes (Appendix S1). Among the 26,139 unannotated unigenes, 20,152 (77.1%) were < 500 bp indicating the importance of the unigene length during the annotation. Within the annotated pool, the majority of the top hits corresponded to protein sequences from S. lycopersicum, V. vinifera, T. cacao, and P. trichocarpa (Table S2). In addition, annotation by GO terms was performed to classify the unigenes (Appendix S1). A total of 40,412 unigenes were mapped to at least one GO term from either Arabidopsis or PANTHER database. Specifically, 44.27% of the unigenes were assigned to “biological process”, 32.4% to “molecular function”, and 23.33% to “cellular component” categories. Over 36% of the unigenes corresponding to biological process were grouped in the “metabolic process” subcategory, while about 31% were classified under the “cellular process” subcategory. Of the unigenes in the “molecular function” category, 45.5% were grouped under “catalytic activity”, and 26.5% were grouped under “binding” subcategories. Under cellular components, a high proportion of unigenes was grouped to “cell part (41.1%)” and “organelle (26.4%)” subcategories (Fig. 2).
Table 2

Summary of the Lavandula transcriptome assembly

Total number of the raw reads

29,008,569

Total number of clean reads

28,830,705

GC content (%)

45.92

Total number of unigenes

101,618

Mean length of unigenes (bp)

692.64

Min length of unigenes (bp)

201

Max length of unigenes (bp)

12,223

N50 value (bp)

831

Approximately 30 million reads were assembled into 101,618 unigenes, with length of unigenes ranging from 201 to 12,223 bp

Fig. 1

Length distribution of unigenes generated from Lavandula transcriptome assembly

Table 3

Summary of annotations of unigenes against public databases

Public databases

Number of annotated unigenes

Percent of annotated unigenes (%)

NCBI (nr)

72,081

70.93

Swiss-prot

52,749

51.90

ARTH

76,607

75.38

Phytozome annotation

75,479

75.48

 pfam

60,237

59.27

 KOG

27,382

26.94

 KO

25,480

25.07

 GO

40,412

39.76

Trinotate annotation

 pfam

36,172

35.59

 Eggnog

17,224

16.94

 GO

37,637

37.03

 sprot_blastx

55,755

54.86

 sprot_blastp

39,225

38.60

 signalP

2500

2.46

 TmHMM

9319

9.17

Out of the total unigenes, over 75,000 were annotated against different public databases, with a minimum (25.07%) and maximum (75.6%) number of unigenes against KO and phytozyme databases, respectively

Fig. 2

Functional classification of unigenes expressed in lavenders. Gene ontology (GO) terms are summarized in three main categories of a biological process, b cellular component, and c molecular function. Of the total GO term annotated unigenes (40,412), nearly, 45% of which were grouped under eleven biological processes, 23.33% unigenes under six cellular components, and 32.4% unigenes under eight molecular functions

To identify protein families, functional domains, and conserved regions, the Lavandula transcriptome was blasted against other databases using the InterProscan server (Jones et al. 2014). The top 50 InterPro entries obtained are presented in Table S3. Here, we found protein kinases and nucleoside triphosphate hydrolase as the most abundant classes of enzymes. Cluster of orthologous group was generated using the EggNOG database, and over 16,670 (16.5%) unigenes with top hit of serine/threonine protein kinase (11.8%) were identified (Table S2). In addition, the unigenes were annotated using the KEGG orthology database (KO) to identify those that are involved in biological pathways. Over 25,480 unigenes were annotated with KEGG pathways (Appendix S2). Out of the pathway genes identified, the unigenes corresponding to secondary metabolic pathways including phenylpropanoids, flavonoids, terpenoids, and steroids were manually screened and summarized in Table 4. Over 60 annotated unigenes were assigned to phenylpropanoid biosynthesis, with 20 unigenes particularly assigned to peroxidases. Other pathways were represented with a total of 10 (steroids) to 39 (terpenoids) unigenes, with one-to-seven sequences per specific enzyme. More specifically, we detected all TPS genes previously cloned from the three lavender species in the assembled sequences, including trans-α-bergamotene synthase, τ-cadinol synthase, germacrene synthase, limonene synthase, β-caryophyllene synthase, bornyl diphosphate synthase, caryophyllene synthase, R-linalool synthase, 3-carene synthase, 1,8-cineole synthase, β-phellandrene synthase, borneol dehydrogenase, and three monoterpene acetyltransferases (Landmann et al. 2007; Demissie et al. 2011, 2012; Sarker et al. 2012, 2013; Jullien et al. 2014; Sarker and Mahmoud 2015; Adal et al. 2017; Despinasse et al. 2017). The presence of the full complement of previously reported genes involved in essential oil metabolism is an indicative of the quality and depth of the assembled sequences.
Table 4

KEGG pathway related to the biosynthesis of secondary metabolites found in the Lavandula transcriptome database

KEGG pathway

EC number

Enzyme name

# of sequences

Phenylpropanoid biosynthesis

   
 

ec:1.1.1.7

Peroxidases

20

 

ec:2.1.1.104

Caffeoyl-CoA O-methyltransferase

7

 

ec:3.2.1.21

Beta-glucosidase

9

 

ec:2.1.1.68

Caffeate O-methyltransferase

2

 

ec:1.14.13.11

Trans-cinnamate-CoA ligase

1

 

ec:6.2.1.12

4-Coumarate-CoA ligase

6

 

ec:1.1.1.195

Cinnamyl alcohol dehydrogenase

8

 

ec:2.3.1.133

Shikimate O-hydroxycinnamoyltransferase

1

 

ec:4.3.1.24

Phenylalanine aminomutase

1

 

ec:6.2.1.12

4-Coumarate-CoA ligase

6

Flavonoid biosynthesis

   
 

ec:1.1.1.219

Dihydroflavonol 4-reductase

2

 

ec:1.14.11.22

Flavone synthase

1

 

ec:1.14.11.23

Flavanone 3-hydroxylase

1

 

ec:1.14.11.9

Flavonol synthase

2

 

ec:1.14.13.11

Trans-cinnamate 4-monooxygenase

1

 

ec:1.14.13.21

Flavonoid 3′-monooxygenase

1

 

ec:1.14.13.88

Flavonoid 3′,5′-hydroxylase

1

 

ec:2.1.1.104

Caffeoyl-CoA O-methyltransferase

7

 

ec:2.3.1.133

Shikimate O-hydroxycinnamoyltransferase

1

 

ec:2.3.1.74

Chalcone synthase

2

 

ec:5.5.1.6

Chalcone–flavonone isomerase 

2

Terpenoid biosynthesis

   
 

ec:1.1.1.216

Farnesol dehydrogenase

2

 

ec:1.1.1.34

3-Hydroxy-3-methylglutaryl-coenzyme A reductase

2

 

ec:1.1.1.354

Farnesol dehydrogenase

1

 

ec:1.1.1.88

3-Hydroxy-3-methylglutaryl-coenzyme A reductase

2

 

ec:1.17.7.1

4-Hydroxy-3-methylbut-2-en-1-yl diphosphate synthase

1

 

ec:1.17.7.3

4-Hydroxy-3-methylbut-2-en-1-yl diphosphate synthase

1

 

ec:1.17.7.4

4-Hydroxy-3-methylbut-2-enyl diphosphate reductase

2

 

ec:2.2.1.7

1-Deoxy-d-xylulose-5-phosphate synthase

1

 

ec:2.3.1.9

Acetyl-CoA acetyltransferase

2

 

ec:2.3.3.10

Hydroxymethylglutaryl-CoA synthase

1

 

ec:2.5.1.10

Geranylgeranyl pyrophosphate synthase

3

 

ec:2.5.1.1

Dimethylallyltranstransferase

3

 

ec:2.5.1.20

Cis-Prenyltransferase

1

 

ec:2.5.1.29

Geranylgeranyl pyrophosphate synthase

2

 

ec:2.5.1.68

(2Z,6E)-Farnesyl diphosphate synthase

1

 

ec:2.5.1.92

(2Z,6Z)-Farnesyl diphosphate synthase

1

 

ec:2.7.1.148

4-Diphosphocytidyl-2-C-methyl-d-erythritol kinase

2

 

ec:2.7.1.185

Mevalonate-3-kinase

1

 

ec:2.7.1.186

Mevalonate-3-phosphate-5-kinase

1

 

ec:2.7.1.36

Mevalonate kinase

1

 

ec:2.7.4.26

Isopentenyl phosphate kinase

1

 

ec:2.7.4.2

Phosphomevalonate kinase

1

 

ec:2.7.7.60

2-C-methyl-d-erythritol 4-phosphate cytidylyltransferase

1

 

ec:3.1.7.6

(E, E)-farnesol synthase

1

 

ec:4.1.1.33

Diphosphomevalonate decarboxylase

1

 

ec:4.1.1.99

Phosphomevalonate decarboxylase

1

 

ec:4.2.3.27

Isoprene synthase

1

 

ec:4.6.1.12

2-C-methyl-d-erythritol 2,4-cyclodiphosphate synthase

1

 

ec:5.3.3.2

IPP isomerase

2

Steroid biosynthesis

   
 

ec:1.1.1.145

3 Beta-hydroxysteroid dehydrogenase

1

 

ec:1.14.14.17

Squalene monooxygenase

1

 

ec:1.14.14.1

Cytochrome P450

1

 

ec:1.14.21.6

Lathosterol oxidase

1

 

ec:1.3.99.5

3-Oxo-5-alpha-steroid 4-dehydrogenase

1

 

ec:2.1.1.41

Sterol 24-C-methyltransferase

1

 

ec:2.1.1.6

Catechol O-methyltransferase

2

 

ec:2.5.1.21

Squalene synthase

1

 

ec:5.3.3.5

Cholestenol Delta-isomerase

1

High number of annotated unigenes was assigned to phenylpropanoid biosynthesis (61 unigenes), followed by terpenoid biosynthesis (39 unigenes). Other pathways were represented by small number of unigenes, ranging from 10 (steroids) to 21 unigenes (flavonoids)

Isolation and sequence analysis of S-linalool synthase

Several TPS genes have been cloned and functionally characterized from various Lavandula species. In addition, several partial TPS-like sequences exist in our previously described EST databases (Lane et al. 2010; Demissie et al. 2012), and in our recently sequenced draft L. angustifolia genome (Malli et al., unpublished). Using a sequence similarity-based approach, we screened the assembled transcriptomes reported here for TPS-related sequences. The search demonstrated that the database contained sequences for all TPS genes previously cloned from the three lavender species, and several unique TPS-like sequences that have not yet been characterized. Among them, was a sequence corresponding to a unique TPS-like gene with six exons (unlike to the previously reported Lavandula TPSs with seven exons) we had identified in the L. angustifolia draft genome (Malli et al. unpublished). This gene, which was later shown to be S-linalool synthase (LiS-LINS), is distantly related to other Lavandula TPSs and lacks one of the common conserved motifs (Fig. 3). Transcripts for this gene were not present in any of our previously described EST databases including one derived from floral-based glandular trichomes of L. × intermedia (Demissie et al. 2012), indicating that it is not very strongly expressed, or that its mRNA is not very stable. The complete ORF of LiS-LINS (GenBank: MG870571) encoded a protein of 571 amino acids with a predicted molecular mass of 66.14 kDa and calculated pI of 5.16. Comparison of the deduced amino acids to orthologous linalool synthases revealed a high degree of sequence similarity within the conserved motifs. As with most plant TPSs, the key conserved motifs including DDxxD and (N,D)D(L,I,V)x(S,T)x3E) that are crucial for divalent metal ion binding were well conserved in LiS-LINS (Fig. 3). However, like other S-linalool synthases from different plants, LiS-LINS lacked the N-terminal RR(x)8W motif that is important for cyclization and protein stability (Fig. 3).
Fig. 3

Multiple sequence alignment of LiS-LINS (MG870571) with AmNES/LINS1 (EF433761) and AmNES/LINS2 (EF433762) that are the closest TPS genes in plants. LiS-LINS shared over 50% amino acid sequences with both AmNES/LINS1 and AmNES/LINS2. As most angiosperms, LiS-LINS also consisted of the two conserved aspartate-rich motifs, but lacked RR(x)8W motif. The predicted transit peptide is shown by broken line, and the main TPS conserved motifs are underlined

Phylogenetic analysis indicated that LiS-LINS belongs to TPS-g subfamily that lacks RR(x)8W (Fig. 4), although it exhibited a relatively low amino acids identity with TPSs within this group. The highest amino acid identity of LiS-LINS was 53.27 and 50.57% with nerolidol synthase (AmNES/LINS-2) and S-linalool synthase (AmNES/LINS-1) of Antirrhinum majus, respectively (Figs. 3, 4b). LiS-LINS has even less sequence similarity to the rest of known TPSs of the genus Lavandula, with maximum amino acid sequence identities with germacrene D synthase (28.74%) and R-linalool synthase (28.47%) of L. angustifolia (Fig. 4a). In addition, this gene is distantly related to known S-linalool synthases of Coriandrum sativum (29.14% identity) (TPS-b) and C. breweri (22.64% identity) (TPS-f).
Fig. 4

Unrooted neighbor-joining tree based on amino acid sequences of LiS-LINS and selected plant TPS sequences. Comparison of LiS-LINS amino acid sequences with a most of previously known Lavandula TPSs, and with b selected plant TPS-g subfamily. The tree showed that LiS-LINS was clustered to TPS-g, which is different from the TPS-a and TPS-b subfamilies that include all of previously reported Lavandula sesqui-TPS and mono-TPS, respectively. The tree was generated using Phylogeny Analysis MEGA7.0 program (Kumar et al. 2016). The resulting trees were bootstrap analyzed with 1000 replicates, and bootstrap values > 50% are mentioned on the trees. The taxon in rectangular box indicates the L. × intermedia S-LINS sequence identified in this study. The clades of TPS subfamilies were designated as TPS-a, TPS-b, and TPS-g according to Bohlmann et al. (1998) and Chen et al. (2011). The scale bar shows the number of amino acid substitution per site. The Genbank accession numbers of TPSs used in the phylogenetic analysis are mentioned in Table S1

To identify the copy number and genomic organization of the genes encoding R- and S-linalool synthases, we explored the draft genome of L. angustifolia (Malli et al., unpublished) using cDNA sequences of S-linalool synthase from L. × intermedia (this study) and previously reported R-linalool synthase from L. angustifolia (Landmann et al. 2007) as probes. The results revealed the presence of a single full-length sequence for each of R-linalool synthase gene and S-linalool synthase gene in L. angustifolia draft genome. The full-length S-linalool synthase identified in the genome-encoded 550 amino acids containing 21 amino acids deletion in the last exon compared to transcriptome-derived S-linalool synthase (571 amino acids). The genomic sequence of S-linalool synthase included six exons and five introns, while that of R-linalool synthase contained all exons (seven) and introns (six) previously reported for this gene (Fig. 5).
Fig. 5

Genome organization of genes encoding linalool enantiomer synthases from a draft genome of L. angustifolia.aLaS-LINS exhibits a genome structure with six exons and five introns and carries 63 bp in-frame deletion in the last exon. bLaR-LINS contains a complete genome structure with seven exons and six introns. Exons are represented by coding sequences (CDS) that are designated as kilo base pairs

3D modeling and analysis of active site

A homology-based approach was used to build the 3D structural model of LiS-LINS (this study, GenBank: MG870571) (Fig. 6a) and R-linalool synthases from L. angustifolia (LaR-LINS, GenBank: ABB73045.1) (Landmann et al. 2007) (Fig. 6d) against experimentally determined 3D structure of bornyl diphosphate synthase (SoBPPS) (PDB:1N20), a closely related monoterpene synthase from S. officinalis (Whittington et al. 2002). We also estimated the overall model quality (Z-score) of both linalool isomer synthase homology models along with SoBPPS 3D structure. The qualities of both homology models (Z-score: − 8.75 for LiS-LINS and − 11 for LaR-LINS) were in close range with SoBPPS protein structure (Z-score = − 10.75). Pairwise alignment of LiS-LINS and LaR-LINS 3D structures with length of alignment (LaLi): 516, root mean square deviation (RMSD): 1.62 and TMscore: 0.929 showed that both protein structures very likely share a similar fold. The models predicted the active site pocket, which contains substrate (GPP) docking site as well as catalytic residues (Fig. 6). Despite the very low overall sequence similarity (~ 30% amino acids identity) between LiS-LINS and LaR-LINS, the active sites of these proteins exhibited over 70% identity to one another (Fig. 6b, e). However, there are a few amino acid residues in the active site that are different between LiS-LINS and LaR-LINS (Fig. 6). These residues are 324Y, 428S, 479R, and 552Q in LiS-LINS, and 317T, 421I, 480K, and 547N in LaR-LINS. Three out of the four residues [324Y, 428S and 552Q in LiS-LINS, and 317T, 421I and 547N in LaR-LINS] are positioned together around the carbon backbone of the docked GPP.
Fig. 6

3D structure and active site pockets of LiS-LINS and LaR-LINS protein. The first column is for LiS-LINS displaying a 3D protein structure with active site pocket, b the active site pocket with predicted GPP binding residues and c LiS-LINS specific binding residues. The second column represents LaR-LINS d 3D protein with its active site pocket, e active site pocket with GPP and binding residues and f unique binding residues. A model of the substrate (GPP) bound in the active site residues is shown at the center of the active site in both LiS-LINS and LaR-LINS 3D proteins. Several common and unique binding residues in the active site pocket are indicated in both 3D proteins. L. angustifolia R-LINS (LaR-LINS) amino acid sequences was reported previously (Landmann et al. 2007). Prediction of 3D protein structure and binding residues was carried out using online RaptorX server (Källberg et al. 2012), and docking of GPP into active sites of 3D protein was performed using 1-Click Docking online software (https://mcule.com/apps/1-click-docking/). The best docking score was − 6.8 for LiS-LINS and − 7.0 for LiR-LINS. The more negative score shows the better ligand orientation

Subcellular localization of LiS-LINS protein

Different bioinformatics tools were used to predict the likelihood of targeting transit peptides in N terminus of LiS-LINS protein sequence. Among all tested prediction tools (detailed in “Materials and methods” section), three programs detected plastid-targeting N terminus transit peptide in LiS-LINS protein. Among these ‘ChloroP’ and ‘TargetP’ tools predicted very short plastid-targeting N terminus transit peptides with length of four amino acids, while ‘iPSORT’ detected a 30 amino acid long targeting sequence at N terminus. To experimentally determine the subcellular localization of LiS-LINS protein, we fused the transit peptide—the first 50 amino acids—of this protein to the N terminus of GFP reporter gene, and expressed the fusion protein transiently in N. benthamiana protoplasts. We analyzed transient expression of GFP-fused constructs in protoplasts by fluorescent microscopy (Fig. 7). Protoplasts expressing LiS-LINS-TP(50 aa) produced GFP fluorescence in discrete patterns associated with chloroplast (Fig. 7d–f). The pattern of the LiS-LINS-TP(50 aa)::GFP accumulation in chloroplasts resembled to the signals of GFP fused to TP(50 aa) of R-linalool synthase (transit peptide), a plastidial monoterpene synthase (Fig. 7g–i), and was entirely different from empty vector (GFP) that was localized in cytosol (Fig. 7a–c). These results clearly showed that the LiS-LINS protein contains a transit peptide that targets it to the chloroplast.
Fig. 7

Transiently expressed LiS-LINS-TP(50 aa)::GFP fusion in N. benthamiana protoplasts analyzed by inverted fluorescent microscopy. TP transit peptide, 50 aa the first 50 amino acids of the full ORF. The “Green” column shows GFP fluorescence detected in the green channel; the “Red” column shows chlorophyll autofluorescence detected in red channel, and the “Merged” column shows combined green and red channels. p326::GFP (cytosolic) and LiR-LINS-TP(50 aa)::GFP (plastidial targeting transit peptide) are used as control markers. The figure showed the plastidial localization of LiS-LINS-TP(50 aa) ::GFP similar to the positive control LiR-LINS-TP(50 aa)::GFP, confirming its chlorophyll targeting enzyme

Expression and assay of the recombinant LiS-LINS

We expressed the LiS-LINS ORF, excluding the transit peptide, in E. coli Rosetta™ (DE3) pLysS strain using the pET41b(+) expression vector, and purified the recombinant protein (62.23 kDa) using Ni–NTA-agarose affinity chromatography. The partially purified recombinant protein was assayed for activity with GPP, NPP, FPP, DMAPP as well as a combined IPP and DMAPP as substrate. The products of each of the reaction were analyzed along with the authentic standards by GC–MS. Linalool was detected as a single product from both assays of partially purified LiS-LINS with GPP and NPP (Fig. 8a–f). However, the recombinant LiS-LINS protein was catalytically inactive when supplied with FPP, DMAPP, and a mix of IPP and DMAPP. As a control, affinity-purified extracts from empty vector transformed E. coli, or heat-denatured purified recombinant LiS-LINS protein was assayed with GPP or NPP. None of the control reactions produced detectable linalool.
Fig. 8

GC–MS analysis of monoterpene standards and reaction products generated by the recombinant L. × intermedia S-linalool synthase from GPP as a substrate. Gas chromatogram and Mass spectrum of standard S-linalool (a, b), racemic mixture (50/50%) of S- and R-linalool (c, d), and LiS-LINS assay product (e, f)

The assay products of the recombinant LiS-LINS protein supplemented with each of GPP and NPP were further analyzed with GC–FID using a chiral column. The results revealed the presence of enantiomerically pure S-linalool from GPP or NPP supplemented assay reactions (Fig. 9b–d).
Fig. 9

Linalool enantiomers biosynthesis by recombinant linalool synthases from Lavandula, and chiral analysis using GC–FID. a biosynthesis of linalool enantiomers from GPP/NPP; b Racemic mixture (50/50%) of R- and S-linalool standard; cS-linalool from assay product; and dS-linalool standard. Previously reported recombinant LaLINS catalyzed the conversion of GPP to R-linalool from L. angustifolia (Landmann et al. 2007). The recombinant LiS-LINS assay produced S-linalool as a single product from GPP/NPP (this study)

Relative expression of the two linalool synthases

To examine the spatial expression patterns of the two linalool isomer synthases, we quantified the levels of LiS-LINS and LiR-LINS transcripts in L. × intermedia flower, leaf, and root tissues using qPCR (Fig. 10). The results showed that although both genes are expressed in floral tissue, they have distinct expression patterns. In flowers, LiR-LINS transcripts were much more abundant than those of LiS-LINS. On the other hand, transcript levels for LiS-LINS were substantially (~ 7 fold) more abundant than those of LiR-LINS in leaves. In this context, transcript levels for LiR-LINS and LiS-LINS paralleled the concentration of the corresponding linalool enantiomer (Table 1).
Fig. 10

qPCR analysis for the spatial expression of L. × intermedia linalool synthases (LiS-LINS and LiR-LINS) and the corresponding products, S- and R- linalool. a Transcript levels of LiS-LINS and LiR-LINS in flower, leaf and root tissues of L. × intermedia, cv Grosso. The transcripts were normalized to β-actin gene. The control (in this case root tissues) was assigned with the arbitrarily value of 1.0. Bars represent mean values of biological replications ± standard errors (n = 3)

Discussion

Traditional EST databases derived from cDNA libraries have been tremendously important in gene discovery. However, this approach suffers from certain limitations. For example, compare to RNA-Seq, building an EST databases is tedious, costly, and low throughput. In addition, EST databases are often not deep enough to enable discovery of genes for which transcripts are not highly abundant (Wang et al. 2009). These bottlenecks have now been resolved through RNA-Seq, which is widely applied to transcriptome profiling studies in various plants. For example, this approach has been very effective in the discovery of biosynthetic genes from leaves and rhizomes of C. pictus, and C. longa, respectively (Annadurai et al. 2012), and numerous other plants. It has also been useful in identifying several genes involved in resistance to bacterial wilt in mango ginger, in different flower developmental stages in H. coronarium (Prasath et al. 2014; Yue et al. 2015), and in terpenoid biosynthesis in C. sativum (Galata et al. 2014). Applications of RNA-Seq are well-beyond gene discovery, and are summarized in recent reviews (Wang et al. 2009; Han et al. 2015; Hrdlickova et al. 2017; Lowe et al. 2017).

In this study, over 28 million high-quality reads generated from EO-producing tissues of three economically important lavender species were assembled into 101,618 unigenes (N50 = 831 bp). Over 75% of the unigenes were annotated to produce a genomic resource that facilitates the discovery of novel genes. The annotation proportion (ca. 75%) falls well within the range reported for other plants, for example, Curcuma amada (73.17%) and Zingiber officinale (76.55%) (Prasath et al. 2014). Annotated sequences contained most of the genes involved in isoprenoid metabolism including all genes of the MVA and MEP pathways, and genes encoding all known major prenyltransferases. The assembly also included sequences corresponding to all of the previously reported TPSs from various lavender species, and for several new putative (uncharacterized) TPS genes that can potentially involved in isoprenoid metabolism in lavenders. Altogether, our RNA-Seq is of high quality with respect to the number and type of sequences in contains as well as the proportion of annotated sequences, and provides a good resource for the identification of isoprenoid related genes in lavenders.

TPS subfamily and identification of LiS-LINS

The genomes of most studied plants contain a wide range of TPSs (20–150 genes), which have been classified into seven families—TPS-a to TPS-h—based on the structure and function of the encoded enzymes (Bohlmann et al. 1998; Chen et al. 2011). While all of the TPS genes reported previously from Lavandula have been described in one of the above classification, scarcely expressed genes have not been reported before. In this study, we cloned and characterized one of these genes (S-LINS) (GenBank: MG870571). The putative function of this gene as a TPS was initially predicted based on the presence of TPS conserved motifs as well as structural homology to known TPSs from Lavandula and other plants. For example, LiS-LINS protein contained both DDxxD and (N,D)D(L,I,V)x(S,T)x3E) motifs that are responsible for substrate binding and coordination of divalent metal ion cofactors (Degenhardt et al. 2009). However, unlike most TPSs, LiS-LINS lacked RR(x)8W—a motif required for product cyclization in Class III TPS proteins (Whittington et al. 2002; Hyatt et al. 2007)—suggesting that this gene potentially encoded acyclic TPS. Based on phylogenetic analysis using deduced amino acids, LiS-LINS was placed in the TPS-g subfamily. TPS-g subfamily is defined as a group of mono-TPSs that are responsible for the synthesis of acyclic monoterpenes, including myrcene and ocimene in A. majus, linalool in A. thaliana and V. vinifera (Dudareva et al. 2003; Chen et al. 2003; Martin et al. 2010). Within TPS-g subfamily, LiS-LINS was further clustered together with the common bifunctional linalool/nerolidol synthase of A. majus (Nagegowda et al. 2008) and Fragaria spp (Aharoni et al. 2004). It is worth noting that unlike LiS-LINS, most of previously known Lavandula TPSs are grouped to TPS-a and -b subfamilies (Demissie et al. 2011, 2012; Sarker et al. 2013; Jullien et al. 2014; Adal et al. 2017).

Gene structural diversity within a gene family—one of the common evolutionary mechanisms that enhance variability—can be generated by several mechanisms, including loss or gain of exon/intron in plants (Li et al. 2009, 2014; Xu et al. 2012). In this report, S-LINS identified from L. angustifolia (half parental line of L. × intermedia) genome (Malli et al. unpublished) (LaS-LINS) consisted of six exons and five introns, with 63 bp in-frame deletion in the last exon compared to those TPSs clustered under Class III that encompass seven exons and six introns (Trapp and Croteau 2001; Chen et al. 2011). The first five exons of this gene were organized similar to most monoterpene synthases found within Class III including previously reported Lavandula TPSs (Landmann et al. 2007; Demissie et al. 2012; Adal et al. 2017), but exon-6 is composed of the equivalent of the last two exons of Class III genes. The presence of different number of exons and introns between LaS-LINS and Class III TPSs likely occurred through loss of an intron during a structural evolution of LaS-LINS in L. angustifolia. The structural variation of this gene in a very closely related species could contribute to gene structure diversification, and eventually functional diversity and divergence of TPSs in Lavandula.

3D modeling and potential binding residues in Lavandula linalool synthases

The 3D structure of enzymes can be explored by performing X-ray crystallography on the recombinant protein (Whittington et al. 2002). Where a crystal structure is not available, comparative homology-based modeling may be done based on the crystal structures of closely relate proteins. We performed a 3D homology modeling of S-linalool synthase from L. × intermedia (LiS-LINS) and R-linalool synthases from L. angustifolia (LaR-LINS) (Landmann et al. 2007) against the known SoBPPS (Whittington et al. 2002). LiS-LINS exhibited a high (> 70%) degree of homology (in terms of the 3D protein structure and the amino acid composition of the active site pocket) to LaR-LINS (Fig. 6). However, four residues, including 324Y, 428S, 479R and 552Q in LiS-LINS, and 317T, 421I, 480K and 547N in LaR-LINS, were found to be protein-specific. Residues 324Y, 428S, and 552Q in LiS-LINS, and 317T, 421I and 547N in LaR-LINS were positioned together around the carbon backbone of the docked GPP. This indicated that these residues are likely involved in the specific orientation of the docked substrate (GPP) in the active site, and are likely important for the enantio-specificity of the two linalool synthases.

Functional analysis of the recombinant LiS-LINS

Several Lavandula EO-related genes and the related encoding enzymes have been previously described (Landmann et al. 2007; Demissie et al. 2011, 2012, 2013; Sarker et al. 2012, 2013; Jullien et al. 2014; Benabdelkader et al. 2015; Adal et al. 2017; Despinasse et al. 2017). The R-linalool synthase was first cloned from L. angustifolia, and its recombinant protein catalyzed the conversion of GPP into R-linalool (Fig. 9a) (Landmann et al. 2007). However, the gene responsible for the production of S-linalool in Lavandula has not been previously reported. In this study, we cloned the S-linalool synthase from L. × intermedia (LiS-LINS) through deep transcriptome sequencing. The recombinant LiS-LINS converted GPP—the universal precursor of regular monoterpenes—into a single product, S-linalool (Fig. 9a). Although most TPSs are multi-products, mono-product TPSs are also common in plants. For example, S-linalool synthase of A. thaliana (Chen et al. 2003) and R-linalool synthase of L. angustifolia (Landmann et al. 2007) are both uni-product enzymes. The recombinant LiS-LINS also accepted NPP (cis isomer of GPP) as a substrate and transformed it to S-linalool. The acceptance of an alternative but closely related substrate (in this case NPP) is not surprising, and has been reported for other plant TPSs before (Demissie et al. 2011, 2012; Adal et al. 2017).

LiS-LINS is differentially regulated

qPCR analysis showed that LiS-LINS is strongly expressed in flower—major EO-producing tissue—compared to other tissues examined. However, transcripts for this gene were much less abundant than those of LiR-LINS. Given that most of the linalool found in lavender flowers are of the R-type, this finding was not surprising. In this context, the abundances of R- and S-LINS transcripts paralleled to the concentrations of the corresponding products [R- and S-linalool, respectively] in floral tissue of L. × intermedia cv. grosso reported previously (Renaud et al. 2001). The low expression of LiS-LINS is probably due to the negative regulation of the gene at transcription and/or translation levels. Furthermore, expression levels of LiS-LINS would be associated with certain floral developmental stages, as different stages of floral development strongly influence emission of S- and R-linalool, and transcript levels of linalool synthase gene in Jasminum grandiflorum (Pragadheesh et al. 2017).

Implication of S-linalool biosynthesis in Lavandula oil quality

Although the EOs of lavenders contain several mono- and sesquiterpenes, linalool and linalyl acetate are considered to be the most important aroma-determining EO constituents in these plants. S-Linalool is of particular importance as it imparts a sweet scent to these oils (Renaud et al. 2001; Flores et al. 2005; Özek et al. 2010). In lavenders, the natural enantiomeric distribution of R-linalool and S-linalool varies from 80–93 to 5–20% in L. angustifolia, and from 87–97 to 3–13% in L. × intermedia oils, respectively (Renaud et al. 2001; Baser et al. 2005; Özek et al. 2010; Satyal and Pappas 2016). The possible addition of exogenous S-linalool to enhance the scent and market value of lavender oils has been reported (Renaud et al. 2001; Flores et al. 2005; Aprotosoaie et al. 2014; Do et al. 2015). In this context, some of the commercially marketed lavender EOs contain up to 32% S-linalool (Renaud et al. 2001).

There is an interest in enhancing S-linalool content in lavenders through metabolic engineering. For example, a recent study reported enhancing S-linalool production by overexpressing the C. berweri S-linalool synthase gene in spike lavender (Mendoza-Poudereux et al. 2014). The cloning of S-LINS from L. × intermedia provides a new tool for such studies. The overexpression of the homologous S-LINS may result in enhanced S-linalool production in lavenders. However, increased S-linalool composition in lavender oils may need further fine-tuning since the chirality of linalool influences sensory profile and biological activity of lavender oils (Aprotosoaie et al. 2014).

Conclusions

Using RNA-Seq derived from EO-producing tissues of Lavandula, we cloned and functionally characterized one of the scarcely expressed terpene synthases (LiS-LINS), which is responsible for the production of S-linalool in lavenders. LiS-LINS protein exhibits a strong sequence similarity to Class III TPSs, and unlike most other monoterpene synthases, it is clustered under TPS-g subfamily that lacks the RR(x)8W conserved motif. The protein contains transit peptide mediating chloroplast targeting, and is mainly expressed in flowers. The recombinant LiS-LINS protein catalyzes GPP into a sole product, S-linalool. The cloned LiS-LINS would be useful in studies aimed at producing recombinant S-LINS in vitro or in microbial systems, and in improving essential oil composition and scent in lavenders and other plants through metabolic engineering.

Author contribution statement

AMA and SSM conceived and designed the research. AMA conducted experiments related to S-linalool synthase. LSS conducted RNA-Seq assembly and contributed on cloning of LiS-LINS. RPNM and PL analyzed a genome sequence of LaS-LINS. AMA led manuscript preparation. All authors read and approved the manuscript.

Notes

Acknowledgements

This work was supported through Grants and/or in-kind contributions to SSM by UBC, and Natural Sciences and Engineering Research Council of Canada. We thank Dr I. Hwang, Pohang University of Science and Technology, Korea for kindly providing p326::sGFP construct DNA; Aaron Johnstone, UBC, Dr. Phillip Barker’s lab for assisting on inverted fluorescent microscopy and Joan Chisholm, Dr. Helene Sanfacon’s Lab, Summerland Agriculture and Agri-Food Canada for assisting on N. benthamiana protoplasts isolation and transfection.

Compliance with ethical standards

Conflict of interest

The authors declare no conflicts of interest.

Supplementary material

425_2018_2935_MOESM1_ESM.xlsx (4.9 mb)
Appendix S1: Annotated transcriptome sequences from Lavandula (XLSX 5066 kb)
425_2018_2935_MOESM2_ESM.xlsx (19 kb)
Appendix S2: KEGG pathway distribution using KO annotation (XLSX 19 kb)
425_2018_2935_MOESM3_ESM.docx (14 kb)
Table S1. TPSs clustered under TPS-g subfamily, and previously known from Lavandula, with the corresponding GenBank/ UniProt accession numbers that are incorporated in the phylogenetic analysis of LiS-LINS (DOCX 13 kb)
425_2018_2935_MOESM4_ESM.docx (13 kb)
Table S2. Species distribution of the top BLAST hits of Lavandula transcriptome unigenes. The transcriptomic sequences from Lavandula were highly homologous to Solanum lycopersicum (23.59 %) and Vitis vinifera (15.45 %) compared to transcriptome sequences from any of other plant species (< 8 %) (DOCX 12 kb)
425_2018_2935_MOESM5_ESM.docx (16 kb)
Table S3. Summary of the most common protein families and domains found in the Lavandula transcriptome database. Of the total protein classes and domains detected from Lavandula transcriptome sequences, protein kinases, nucleoside triphosphate hydrolase and Leucine-rich repeat domain are the most abundant containing over 1000 copy numbers (DOCX 16 kb)

References

  1. Adal AM, Sarker LS, Lemke AD, Mahmoud SS (2017) Isolation and functional characterization of a methyl jasmonate-responsive 3-carene synthase from Lavandula × intermedia. Plant Mol Biol.  https://doi.org/10.1007/s11103-017-0588-6 CrossRefPubMedGoogle Scholar
  2. Aharoni A, Giri AP, Verstappen FWA, Bertea CM, Sevenier R (2004) Gain and loss of fruit flavor compounds produced by wild and cultivated strawberry species. Plant Cell 16:3110–3131.  https://doi.org/10.1105/tpc.104.023895.native CrossRefPubMedPubMedCentralGoogle Scholar
  3. Annadurai RS, Jayakumar V, Mugasimangalam RC, Katta MA, Anand S, Gopinathan S, Sarma SP, Fernandes SJ, Mullapudi N, Murugesan S, Rao SN (2012) Next generation sequencing and de novo transcriptome analysis of Costus pictus D. Don, a non-model plant with potent anti-diabetic properties. BMC Bioinf 13:1–15.  https://doi.org/10.1186/1471-2164-13-663 CrossRefGoogle Scholar
  4. Aprotosoaie AC, Hǎncianu M, Costache II, Miron A (2014) Linalool: a review on a key odorant molecule with valuable biological properties. Flavour Fragr J 29:193–219.  https://doi.org/10.1002/ffj.3197 CrossRefGoogle Scholar
  5. Aprotosoaie AC, Gille E, Trifan A, Luca VS, Miron A (2017) Essential oils of Lavandula genus: a systematic review of their chemistry. Phytochem Rev 16:761–799.  https://doi.org/10.1007/s11101-017-9517-1 CrossRefGoogle Scholar
  6. Baser KHC, Özek T, Konakchiev A (2005) Enantiomeric distribution of linalool, linalyl acetate and camphor in bulgarian lavender oil. J Essent Oil Res 17:135–136.  https://doi.org/10.1080/10412905.2005.9698855 CrossRefGoogle Scholar
  7. Benabdelkader T, Guitton Y, Pasquier B, Magnard JL, Jullien F, Kameli A, Legendre L (2015) Functional characterization of terpene synthases and chemotypic variation in three lavender species of section Stoechas. Physiol Plant 153:43–57.  https://doi.org/10.1111/ppl.12241 CrossRefPubMedGoogle Scholar
  8. Bohlmann J, Meyer-Gauen G, Croteau R (1998) Plant terpenoid synthases: molecular biology and phylogenetic analysis. Proc Natl Acad Sci USA 95:4126–4133.  https://doi.org/10.1073/pnas.95.8.4126 CrossRefPubMedGoogle Scholar
  9. Cavanagh HMA, Wilkinson JM (2002) Biological activities of lavender essential oil. Phytother Res 16:301–308.  https://doi.org/10.1002/ptr.1103 CrossRefPubMedGoogle Scholar
  10. Chen F, Tholl D, D’Auria JC, Farooq A, Pichersky E, Gershenzonb J (2003) Biosynthesis and emission of terpenoid volatiles from arabidopsis flowers. Plant Cell 15:481–494.  https://doi.org/10.1105/tpc.007989.2 CrossRefPubMedPubMedCentralGoogle Scholar
  11. Chen F, Tholl D, Bohlmann J, Pichersky E (2011) The family of terpene synthases in plants: a mid-size family of genes for specialized metabolism that is highly diversified throughout the kingdom. Plant J 66:212–229.  https://doi.org/10.1111/j.1365-313X.2011.04520.x CrossRefPubMedGoogle Scholar
  12. Christianson DW (2006) Structural biology and chemistry of the terpenoid cyclases. Chem Rev.  https://doi.org/10.1021/cr050286w CrossRefPubMedGoogle Scholar
  13. Degenhardt J, Köllner TG, Gershenzon J (2009) Monoterpene and sesquiterpene synthases and the origin of terpene skeletal diversity in plants. Phytochemistry 70:1621–1637.  https://doi.org/10.1016/j.phytochem.2009.07.030 CrossRefPubMedGoogle Scholar
  14. Demissie ZA, Sarker LS, Mahmoud SS (2011) Cloning and functional characterization of β-phellandrene synthase from Lavandula angustifolia. Planta 233:685–696.  https://doi.org/10.1007/s00425-010-1332-5 CrossRefPubMedGoogle Scholar
  15. Demissie ZA, Cella MA, Sarker LS, Thompson TJ, Rheault MR, Mahmoud SS (2012) Cloning, functional characterization and genomic organization of 1,8-cineole synthases from Lavandula. Plant Mol Biol 79:393–411.  https://doi.org/10.1007/s11103-012-9920-3 CrossRefPubMedGoogle Scholar
  16. Demissie ZA, Erland LAE, Rheault MR, Mahmoud SS (2013) The biosynthetic origin of irregular monoterpenes in Lavandula: isolation and biochemical characterization of a novel cis-prenyl diphosphate synthase gene, lavandulyl diphosphate synthase. J Biol Chem 288:6333–6341.  https://doi.org/10.1074/jbc.M112.431171 CrossRefPubMedPubMedCentralGoogle Scholar
  17. Denner SS (2009) Lavandula angustifolia Miller: english lavender. Holist Nurs Pr 23:57–64.  https://doi.org/10.1097/01.HNP.0000343210.56710.fc CrossRefGoogle Scholar
  18. Despinasse Y, Fiorucci S, Antonczak S, Moja S, Bony A, Nicolè F, Baudino S, Magnard JL, Jullien F (2017) Bornyl-diphosphate synthase from Lavandula angustifolia: a major monoterpene synthase involved in essential oil quality. Phytochemistry 137:24–33.  https://doi.org/10.1016/j.phytochem.2017.01.015 CrossRefPubMedGoogle Scholar
  19. Do TKT, Hadji-Minaglou F, Antoniotti S, Fernandez X (2015) Authenticity of essential oils. Trends Anal Chem 66:146–157.  https://doi.org/10.1016/j.trac.2014.10.007 CrossRefGoogle Scholar
  20. Dudareva N, Martin D, Kish CM, Kolosova N, Gorenstein N, Fäldt J, Miller B, Bohlmann J (2003) (E)-Beta-ocimene and myrcene synthase genes of floral scent biosynthesis in snapdragon: function and expression of three terpene synthase genes of a new terpene synthase subfamily. Plant Cell 15:1227–1241.  https://doi.org/10.1105/tpc.011015 CrossRefPubMedPubMedCentralGoogle Scholar
  21. Erland LAE, Mahmoud SS (2015) Lavender (Lavandula angustifolia) oils. Essent Oils Food Preserv Flavor Saf.  https://doi.org/10.1016/B978-0-12-416641-7.00057-2 CrossRefGoogle Scholar
  22. Flores G, Blanch GP, del Castillo MLR, Herraiz M (2005) Enantiomeric composition studies in Lavandula species using supercritical fluids. J Sep Sci 28:2333–2338.  https://doi.org/10.1002/jssc.200500124 CrossRefPubMedGoogle Scholar
  23. Galata M, Sarker LS, Mahmoud SS (2014) Transcriptome profiling, and cloning and characterization of the main monoterpene synthases of Coriandrum sativum L. Phytochemistry 102:64–73.  https://doi.org/10.1016/j.phytochem.2014.02.016 CrossRefPubMedGoogle Scholar
  24. Goodstein DM, Shu S, Howson R, Neupane R, Hayes RD, Fazo J, Mitros T, Dirks W, Hellsten U, Putnam N, Rokhsar DS (2012) Phytozome: a comparative platform for green plant genomics. Nucleic Acids Res 40:1178–1186.  https://doi.org/10.1093/nar/gkr944 CrossRefGoogle Scholar
  25. Grabherr MG, Haas BJ, Yassour M, Levin JZ, Thompson DA, Amit I, Adiconis X, Fan L, Raychowdhury R, Zeng Q, Chen Z, Mauceli E, Hacohen N, Gnirke A, Rhind N, di Palma F, Birren BW, Nusbaum C, Lindblad-Toh K, Friedman N, Regev A (2011) Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nat Biotechnol 29:644–652.  https://doi.org/10.1038/nbt.1883 CrossRefPubMedPubMedCentralGoogle Scholar
  26. Haas BJ, Papanicolaou A, Yassour M, Grabherr M, Blood PD, Bowden J, Couger MB, Eccles D, Li B, Lieber M, MacManes MD, Ott M, Orvis J, Pochet N, Strozzi F, Weeks N, Westerman R, William T, Dewey CN, Henschel R, LeDuc RD, Friedman N, Regev A (2013) De novo transcript sequence reconstruction from RNA-seq using the Trinity platform for reference generation and analysis. Nat Protoc.  https://doi.org/10.1038/nprot.2013.084 CrossRefPubMedPubMedCentralGoogle Scholar
  27. Han Y, Gao S, Muegge K, Zhang W, Zhou B (2015) Advanced applications of RNA sequencing and challenges. Bioinform Biol Insights 9:29–46.  https://doi.org/10.4137/BBI.S28991 CrossRefPubMedPubMedCentralGoogle Scholar
  28. Hrdlickova R, Toloue M, Tian B (2017) RNA-Seq methods for transcriptome analysis. Wiley Interdiscip Rev RNA.  https://doi.org/10.1002/wrna.1364 CrossRefPubMedGoogle Scholar
  29. Hu B, Jin J, Guo AY, Zhang H, Luo J, Gao G (2015) GSDS 2.0: an upgraded gene feature visualization server. Bioinformatics 31:1296–1297.  https://doi.org/10.1093/bioinformatics/btu817 CrossRefGoogle Scholar
  30. Hyatt DC, Youn B, Zhao Y, Santhamma B, Coates RM, Croteau RB, Kang C (2007) Structure of limonene synthase, a simple model for terpenoid cyclase catalysis. Proc Natl Acad Sci USA 104:5360–5365.  https://doi.org/10.1073/pnas.0700915104 CrossRefPubMedGoogle Scholar
  31. Iriti M, Colnaghi G, Chemat F, Smadja J, Faoro F, Visinoni FA (2006) Histo-cytochemistry and scanning electron microscopy of lavander glandular trichomes following conventional and microwave-assisted hydrodistillation of essential oils: a comparative study. Flavour Fragr J 21:704–712.  https://doi.org/10.1002/ffj.1692 CrossRefGoogle Scholar
  32. Jones P, Binns D, Chang HY, Fraser M, Li W, McAnulla C, McWilliam H, Maslen J, Mitchell A, Nuka G, Pesseat S, Quinn AF, Sangrador-Vegas A, Scheremetjew M, Yong SY, Lopez R, Hunter S (2014) InterProScan 5: genome-scale protein function classification. Bioinformatics 30:1236–1240.  https://doi.org/10.1093/bioinformatics/btu031 CrossRefPubMedPubMedCentralGoogle Scholar
  33. Jullien F, Moja S, Bony A, Legrand S, Petit C, Benabdelkader T, Poirot K, Fiorucci S, Guitton Y, Nicolè F, Baudino S, Magnard JL (2014) Isolation and functional characterization of a τ-cadinol synthase, a new sesquiterpene synthase from Lavandula angustifolia. Plant Mol Biol 84:227–241.  https://doi.org/10.1007/s11103-013-0131-3 CrossRefPubMedGoogle Scholar
  34. Källberg M, Wang H, Wang S, Peng J, Wang Z, Lu H, Xu J (2012) Template-based protein structure modeling using the RaptorX web server. Nat Protoc 7:1511–1522.  https://doi.org/10.1038/nprot.2012.085 CrossRefPubMedPubMedCentralGoogle Scholar
  35. Kumar S, Stecher G, Tamura K (2016) MEGA7: molecular evolutionary genetics analysis version 7.0 for bigger datasets. Mol Biol Evol 33:msw054.  https://doi.org/10.1093/molbev/msw054 CrossRefGoogle Scholar
  36. Landmann C, Fink B, Festner M, Dregus M, Engel KH, Schwab W (2007) Cloning and functional characterization of three terpene synthases from lavender (Lavandula angustifolia). Arch Biochem Biophys 465:417–429.  https://doi.org/10.1016/j.abb.2007.06.011 CrossRefPubMedGoogle Scholar
  37. Lane A, Boecklemann A, Woronuk GN, Sarker L, Mahmoud SS (2010) A genomics resource for investigating regulation of essential oil production in Lavandula angustifolia. Planta 231:835–845.  https://doi.org/10.1007/s00425-009-1090-4 CrossRefPubMedGoogle Scholar
  38. Li W, Godzik A (2006) Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics 22:1658–1659.  https://doi.org/10.1093/bioinformatics/btl158 CrossRefPubMedGoogle Scholar
  39. Li W, Liu B, Yu L, Feng D, Wang H, Wang J (2009) Phylogenetic analysis, structural evolution and functional divergence of the 12-oxo-phytodienoate acid reductase gene family in plants. BMC Evol Biol 9:1–19.  https://doi.org/10.1186/1471-2148-9-90 CrossRefGoogle Scholar
  40. Li W-Y, Wang X, Li R, Li W-Q, Chen K-M (2014) Genome-wide analysis of the NADK gene family in plants. PLoS ONE 9(6):e101051CrossRefPubMedPubMedCentralGoogle Scholar
  41. Liao P, Hemmerlin A, Bach TJ, Chye ML (2016) The potential of the mevalonate pathway for enhanced isoprenoid production. Biotechnol Adv 34:697–713.  https://doi.org/10.1016/j.biotechadv.2016.03.005 CrossRefPubMedGoogle Scholar
  42. Lis-Balchin M (2002) Chemical composition of essential oils from different species, hybrids and cultivars of Lavandula. Lavender: the genus Lavandula. Taylor & Francis, London, pp 251–262Google Scholar
  43. Livak KJ, Schmittgen TD (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2(−ΔΔCT) method. Methods 25:402–408.  https://doi.org/10.1006/meth.2001.1262 CrossRefGoogle Scholar
  44. Lowe R, Shirley N, Bleackley M, Dolan S, Shafee T (2017) Transcriptomics technologies. PLoS Comput Biol 13:1–23.  https://doi.org/10.1371/journal.pcbi.1005457 CrossRefGoogle Scholar
  45. Martin DM, Aubourg S, Schouwey MB, Daviet L, Schalk M, Toub O, Lund ST, Bohlmann J (2010) Functional annotation, genome organization and phylogeny of the grapevine (Vitis vinifera) terpene synthase gene family based on genome assembly, FLcDNA cloning, and enzyme assays. BMC Plant Biol 10:226.  https://doi.org/10.1186/1471-2229-10-226 CrossRefPubMedPubMedCentralGoogle Scholar
  46. Mendoza-Poudereux I, Muñoz-Bertomeu J, Navarro A, Arrillaga I, Segura J (2014) Enhanced levels of S-linalool by metabolic engineering of the terpenoid pathway in spike lavender leaves. Metab Eng 23:136–144.  https://doi.org/10.1016/j.ymben.2014.03.003 CrossRefPubMedGoogle Scholar
  47. Mendoza-Poudereux I, Kutzner E, Huber C, Segura J, Eisenreich W, Arrillaga I (2015) Metabolic cross-talk between pathways of terpenoid backbone biosynthesis in spike lavender. Plant Physiol Biochem 95:113–120.  https://doi.org/10.1016/j.plaphy.2015.07.029 CrossRefPubMedGoogle Scholar
  48. Mi H, Huang X, Muruganujan A, Tang H, Mills C, Kang D, Thomas PD (2017) PANTHER version 11: expanded annotation data from Gene Ontology and Reactome pathways, and data analysis tool enhancements. Nucleic Acids Res 45:D183–D189.  https://doi.org/10.1093/nar/gkw1138 CrossRefPubMedGoogle Scholar
  49. Nagegowda DA, Gutensohn M, Wilkerson CG, Dudareva N (2008) Two nearly identical terpene synthases catalyze the formation of nerolidol and linalool in snapdragon flowers. Plant J 55:224–239.  https://doi.org/10.1111/j.1365-313x.2008.03496.x CrossRefPubMedGoogle Scholar
  50. Özek T, Tabanca N, Demirci F, Wedge DF, Hüsnü Can Başer K (2010) Enantiomeric distribution of some linalool containing essential oils and their biological activities. Rec Nat Prod 4:180–192Google Scholar
  51. Pragadheesh VS, Chanotiya CS, Rastogi S, Shasany AK (2017) Scent from Jasminum grandiflorum flowers: investigation of the change in linalool enantiomers at various developmental stages using chemical and molecular methods. Phytochemistry 140:83–94.  https://doi.org/10.1016/j.phytochem.2017.04.018 CrossRefPubMedGoogle Scholar
  52. Prasath D, Karthika R, Habeeba NT, Suraby EJ, Rosana OB, Shaji A, Eapen SJ, Deshpande U, Anandaraj M (2014) Comparison of the transcriptomes of ginger (Zingiber officinale Rosc.) and mango ginger (Curcuma amada Roxb.) in response to the bacterial wilt infection. PLoS One.  https://doi.org/10.1371/journal.pone.0099731 CrossRefPubMedPubMedCentralGoogle Scholar
  53. Rehman R, Hanif MA, Mustaq Z, Al-Sadi AM (2016) Biosynthesis of essential oils in aromatic plants: a review. Food Rev Int 32:117–160CrossRefGoogle Scholar
  54. Renaud ENC, Charles DJ, Simon JE (2001) Essential oil quantity and composition from 10 cultivars of organically grown lavender and lavandin. J Essent Oil Res 13:269–273.  https://doi.org/10.1080/10412905.2001.9699691 CrossRefGoogle Scholar
  55. Ruijter JM, Ramakers C, Hoogaars WMH, Karlen Y, Bakker O, Hoff MJB, Den Van, Moorman AFM (2009) Amplification efficiency: linking baseline and bias in the analysis of quantitative PCR data. Nucleic Acids Res.  https://doi.org/10.1093/nar/gkp045 CrossRefPubMedPubMedCentralGoogle Scholar
  56. Sarker LS, Mahmoud SS (2015) Cloning and functional characterization of two monoterpene acetyltransferases from glandular trichomes of L. × intermedia. Planta 242:709–719.  https://doi.org/10.1007/s00425-015-2325-1 CrossRefPubMedGoogle Scholar
  57. Sarker LS, Galata M, Demissie ZA, Mahmoud SS (2012) Molecular cloning and functional characterization of borneol dehydrogenase from the glandular trichomes of Lavandula × intermedia. Arch Biochem Biophys 528:163–170.  https://doi.org/10.1016/j.abb.2012.09.013 CrossRefPubMedGoogle Scholar
  58. Sarker LS, Demissie ZA, Mahmoud SS (2013) Cloning of a sesquiterpene synthase from Lavandula × intermedia glandular trichomes. Planta 238:983–989.  https://doi.org/10.1007/s00425-013-1937-6 CrossRefPubMedGoogle Scholar
  59. Satyal P, Pappas RS (2016) Antique lavender essential oil from 1945, its chemical composition and enantiomeric distribution. Nat Volat Essent Oils 3:20–25Google Scholar
  60. Sievers F, Wilm A, Dineen D (2011) Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol Syst Biol.  https://doi.org/10.1038/msb.2011.75 CrossRefPubMedPubMedCentralGoogle Scholar
  61. Trapp SC, Croteau RB (2001) Genomic organization of plant terpene synthases and molecular evolutionary implications. Genetics 158:811–832PubMedPubMedCentralGoogle Scholar
  62. Turner GW, Gershenzon J, Croteau RB (2000) Development of peltate glandular trichomes of peppermint. Plant Physiol 124:665–679.  https://doi.org/10.1104/pp.124.2.665 CrossRefPubMedPubMedCentralGoogle Scholar
  63. Vranová E, Coman D, Gruissem W (2012) Structure and dynamics of the isoprenoid pathway network. Mol Plant 5:318–333.  https://doi.org/10.1093/mp/sss015 CrossRefPubMedGoogle Scholar
  64. Wang Z, Gerstein M, Snyder M (2009) RNA-Seq: a revolutionary tool for transcriptomics. Nat Rev Genet 10:57–63CrossRefPubMedPubMedCentralGoogle Scholar
  65. Wang S, Ma J, Peng J, Xu J (2013) Protein structure alignment beyond spatial proximity. Sci Rep.  https://doi.org/10.1038/srep01448 CrossRefPubMedPubMedCentralGoogle Scholar
  66. Whittington DA, Wise ML, Urbansky M, Coates RM, Croteau RB, Christianson DW (2002) Bornyl diphosphate synthase: structure and strategy for carbocation manipulation by a terpenoid cyclase. Proc Natl Acad Sci USA 99:15375–15380.  https://doi.org/10.1073/pnas.232591099 CrossRefPubMedGoogle Scholar
  67. Wiederstein M, Sippl MJ (2007) ProSA-web: interactive web service for the recognition of errors in three-dimensional structures of proteins. Nucleic Acids Res 35:407–410.  https://doi.org/10.1093/nar/gkm290 CrossRefGoogle Scholar
  68. Xu G, Guo C, Shan H, Kong H (2012) Divergence of duplicate genes in exon-intron structure. Proc Natl Acad Sci 109:1187–1192.  https://doi.org/10.1073/pnas.1109047109 CrossRefPubMedGoogle Scholar
  69. Yoo SD, Cho YH, Sheen J (2007) Arabidopsis mesophyll protoplasts: a versatile cell system for transient gene expression analysis. Nat Protoc 2:1565–1572.  https://doi.org/10.1038/nprot.2007.199 CrossRefGoogle Scholar
  70. Yue Y, Yu R, Fan Y (2015) Transcriptome profiling provides new insights into the formation of floral scent in Hedychium coronarium. BMC Genomics 16:470.  https://doi.org/10.1186/s12864-015-1653-7 CrossRefPubMedPubMedCentralGoogle Scholar
  71. Zeng G (1998) Sticky-end PCR: new method for subcloning. Biotechniques 25:206–208CrossRefPubMedGoogle Scholar
  72. Zuzarte MR, Dinis AM, Cavaleiro C, Salgueiro LR, Canhoto JM (2010) Trichomes, essential oils and in vitro propagation of Lavandula pedunculata (Lamiaceae). Ind Crops Prod 32:580–587.  https://doi.org/10.1016/j.indcrop.2010.07.010 CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  • Ayelign M. Adal
    • 1
  • Lukman S. Sarker
    • 1
  • Radesh P. N. Malli
    • 2
  • Ping Liang
    • 2
  • Soheil S. Mahmoud
    • 1
    Email author
  1. 1.Department of BiologyUniversity of British ColumbiaKelownaCanada
  2. 2.Department of Biological SciencesBrock UniversitySt. CatharinesCanada

Personalised recommendations