https://orcid.org/0000-0002-2655-5464
Figures
Abstract
It is critical to gather biological information about rare and endangered plants to incorporate into conservation efforts. The secondary metabolism of Pityopsis ruthii, an endangered flowering plant that only occurs along limited sections of two rivers (Ocoee and Hiwassee) in Tennessee, USA was studied. Our long-term goal is to understand the mechanisms behind P. ruthii’s adaptation to restricted areas in Tennessee. Here, we profiled the secondary metabolites, specifically in flowers, with a focus on terpenes, aiming to uncover the genomic and molecular basis of terpene biosynthesis in P. ruthii flowers using transcriptomic and biochemical approaches. By comparative profiling of the nonpolar portion of metabolites from various tissues, P. ruthii flowers were rich in terpenes, which included 4 monoterpenes and 10 sesquiterpenes. These terpenes were emitted from flowers as volatiles with monoterpenes and sesquiterpenes accounting for almost 68% and 32% of total emission of terpenes, respectively. These findings suggested that floral terpenes play important roles for the biology and adaptation of P. ruthii to its limited range. To investigate the biosynthesis of floral terpenes, transcriptome data for flowers were produced and analyzed. Genes involved in the terpene biosynthetic pathway were identified and their relative expressions determined. Using this approach, 67 putative terpene synthase (TPS) contigs were detected. TPSs in general are critical for terpene biosynthesis. Seven full-length TPS genes encoding putative monoterpene and sesquiterpene synthases were cloned and functionally characterized. Three catalyzed the biosynthesis of sesquiterpenes and four catalyzed the biosynthesis of monoterpenes. In conclusion, P. ruthii plants employ multiple TPS genes for the biosynthesis of a mixture of floral monoterpenes and sesquiterpenes, which probably play roles in chemical defense and attracting insect pollinators alike.
Citation: Chen X, Nowicki M, Wadl PA, Zhang C, Köllner TG, Payá‐Milans M, et al. (2023) Chemical profile and analysis of biosynthetic pathways and genes of volatile terpenes in Pityopsis ruthii, a rare and endangered flowering plant. PLoS ONE 18(6): e0287524. https://doi.org/10.1371/journal.pone.0287524
Editor: Waqas Khan Kayani, University of Kotli, PAKISTAN
Received: March 7, 2023; Accepted: June 5, 2023; Published: June 23, 2023
This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Data Availability: All relevant data are within the paper and its Supporting Information files. The sequences of 5 functionally characterized terpene synthases reported in this paper have been deposited in the GenBank (https://www.ncbi.nlm.nih.gov/genbank/) database (accession numbers ON166544 - ON166550). The 3 transcriptome archives are available under NCBI PRJNA778727 (https://www.ncbi.nlm.nih.gov/bioproject/?term=PRJNA778727).
Funding: This work was supported by the U.S. Department of Agriculture (USDA/MOA number 58-6404-1-637) awarded to RNT. Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation by the U.S. Department of Agriculture or the authors. USDA is an equal opportunity provider and employer. The publication costs are covered by the University of Tennessee Open Publishing Fund, awarded to RNT. The funding agencies played no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Conserving rare and endangered plant species is a daunting task. The passage of the Endangered Species Act in 1973 in the USA provided the legal mandate for collecting and analyzing biological information for such plants [1]. This study focused on Pityopsis ruthii Small, a narrowly distributed endangered plant that occurs only along small sections of the Hiwassee (~ 4 km) and Ocoee (~ 2.5 km) Rivers in Polk County, Tennessee, USA. This small, fall-flowering, herbaceous species belongs to the Asteraceae and typically grows up to 30 cm in height within cracks of massive phyllite boulders situated between the river channel and adjacent forested slopes. P. ruthii is a small, rounded perennial with erect green to brown-colored stems that are covered with silvery hairs and yellow daisy-like flowers [2].
Although P. ruthii can persist in shaded habitats, flowering, seed set, and establishment of plants is the most successful in open areas that receive full sun for a substantial portion of the day [3, 4]. White [4] examined the effects of light intensity, drought, rock surface, soil, leaf, and air temperatures on the growth of P. ruthii and of competing species and demonstrated the ability of P. ruthii to survive the extremes of the harsh environment. P. ruthii is tolerant of prolonged drought [3] and of inundating high flow events (A. J. Dattilo, unpublished), but the plant is a very poor competitor against other species when grown in areas of deeper soils that build-up on the boulder complex or outside of the exposed boulder complexes [5, 6]. Regardless, the species has a very narrow range in a habitat that is subjected to extreme conditions. Although P. ruthii is well adapted to the harsh environment, the range and combination of possible adaptations that permit it to survive the intermittent, but often prolonged, drought and submergence are not known. Although extreme drought and flood events are increasing globally, certain ecological niches have historically undergone regular, radical shifts, and a few plant species, such as P. ruthii, have adapted to survive such events.
P. ruthii is self-incompatible and produces scant amounts of pollen per flower. It is dependent on insects for cross-pollination. Honeybees, bumblebees, and -to a lesser extent- native bees are the primary pollinators for the species [3, 7]. For successful cross-pollination, these bees must find small populations in nature, and therefore, a chemical attractant may provide signals for insects [8]. Considering the peculiar ecological niche and the biology of P. ruthii, it is important to understand the mechanisms that make it a successful plant in this environment.
Many such mechanisms contribute to the adaptation of plants to unique niches. One of the possible mechanisms is the production of a myriad of secondary metabolites, which are characterized by their vast chemical diversity. Terpenes are the largest family of secondary metabolites produced by plants and can be further categorized as monoterpenes (C10), sesquiterpenes (C15), and diterpenes (C20) [9]. The biological functions of terpenes include tolerance of abiotic stresses such as flooding and drought, defense against insects and microbial pathogens, habitat acclimation, and attraction of beneficial organisms such as pollinators [8–14]. Mechanistically, terpenes can regulate the fluidity of cell membranes, maintain membrane integrity, and stabilize photosynthetic machinery under stress conditions [12].
Our long-term general goal was to understand the mechanisms underlying the adaptation of P. ruthii to very limited areas in Tennessee. In this study, we focused on the analyses of secondary metabolites from various tissues, particularly flowers. With terpenes identified as the main constituents of secondary nonpolar metabolites from flowers of P. ruthii, it was also our objective to use transcriptomic and biochemical approaches to understand the genomic and molecular basis of terpene biosynthesis in P. ruthii flowers.
Materials and methods
Plant material
Stem cuttings of P. ruthii plants collected from populations growing along the Hiwassee and Ocoee Rivers, Tennessee, were rooted and grown in pots under natural light [15]. Plants were collected under Tennessee Valley Authority Permit # TE117405-2 and U.S. Fish and Wildlife Service Permit # TE134817-1. Roots, stems, undamaged leaves, and fully open flowers were harvested in September 2015 from three-year-old plants growing in Pro-Mix BX (Premier Tech Horticulture, Quakertown, PA, USA) maintained outdoors at the University of Tennessee. Leaves were cut longitudinally into five sections using a sterile scalpel for physical wounding. Fully open flowers were used to collect volatile terpenes.
Chemical profiling
Leaves, stems, roots, and rhizomes were collected from plants and ground in liquid nitrogen to a fine powder with mortar and pestle. Methylene chloride containing 0.003% (v/v) of 1-octanol as an internal standard was added to 100 mg of each tissue type at a ratio of 10:1 (v/w). Materials were extracted for 4 h with continuous shaking at 100 rpm at room temperature. After centrifugation at 13000 × g for 30 min, 5 μL of extract from each preparation was analyzed using Shimadzu QP5050A Gas chromatography-Mass spectrometry (GCMS). Organic extraction of each tissue type was analyzed in triplicate with each replicate combining plant materials collected from three individual plants. For GCMS analyses, a splitless injection port with injection temperature 250°C was set and the column temperature was set at 60°C with 6 min holding, then increased to 280°C at a rate of 5°C per min. Separation was carried out on a Rxi®-5Sil MS column (30 m × 0.25 mm i.d. × 0.25 μm thickness, Restek, Bellefonte, PA, USA). Helium as carrier gas was set at flow rate of 1 mL min-1. Terpenes were identified using various available MS libraries (NIST, WILEY, and Adams) and by comparison to the mass spectra of authentic standard compounds (https://www.sigmaaldrich.com). Quantification was performed by peak area comparison to that of the internal standard.
Volatiles from P. ruthii were collected using an open headspace sampling system (Analytical Research System, Gainesville, FL) as previously described [16]. Respective plant materials (about 1 to 2 grams per sample) were placed in a 40-mL glass beaker filled with 10 mL of sterile distilled water. The glass beaker with plant samples was placed in a glass chamber with an air inlet and outlet. With the flow of air, volatiles emitted from plant samples were trapped on the volatile collection trap (VCT) containing 25 mg Porapak-QTM (http://www.volatilecollectiontrap.com/) placed at the air outlet. After 4 h of collection at room temperature, volatiles were eluted using 100 μL of methylene chloride containing 1-octanol (0.003% v/v) as internal standard and then directly injected into GC-MS for analysis, as described above. Each headspace collection was replicated three times with each replicate combining appropriate plant materials collected from three individual plants. Terpenes were identified and quantified as described.
Transcriptome sequencing and analysis
To gain access to the P. ruthii transcriptome, total RNA was isolated from freshly collected flowers that were immediately flash frozen in liquid nitrogen. Three independent accessions maintained at the University of Tennessee were sampled, and the florets removed, so that only involucre tissues were used. About 100 mg of tissue per accession were subjected to RNA isolation using the Ribospin II kit (GeneAll Biotechnology, Seoul, Korea) following the manufacturer’s protocol. Total RNA quality was assessed using BioAnalyzer RNA chip (Agilent, Santa Clara, CA, USA), after which the RNA was sequenced using HiSeq Illumina (150 bp PE; GeneWiz Inc., South Plainfield, NJ, USA).
Rcorrector [17] was applied to raw reads with default parameters. Rcorrector is a kmer-based error correction method that uses a De Bruijn graph to represent trusted k-mers. This method was similar to that used on de novo assembly and helped improve the quality of assemblies. Corrected reads were trimmed to remove the Illumina adapter sequences using Skewer v0.2.2 [18], using a minimum read length cutoff of 30 bp. FastQC v0.11.4 (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/) was used for quality control of reads. Reads matching ribosomal, plastid, and mitochondrial DNA were removed from the analysis using Bowtie2 [19].
Cleaned reads were used to assemble transcripts with Trans-ABySS [20] on a multi-kmer assembly, and used kmers of 25, 45, and 65 for individual assemblies with minimum contig length of 200 bp, and transabyss-merge to combine them. Substantial removal of assumed isoforms was carried-out afterwards with RapClust [21], which groups transcripts using information from multi-mapper paired-ended reads. Read mapping was performed with Salmon v0.8.2 [22], a fast quasi-mapping tool. The clustering information yielded by RapClust was used to obtain a reduced transcriptome after the selection of the longest transcript per cluster.
Basic assembly metrics were obtained with in-house script. Completeness of the de novo transcriptome assembly was assessed with BUSCO v2.0 [23]. A custom protein database of 54 enzymes involved in terpene biosynthesis was obtained from their EC codes with KEGGREST [24]. A custom database for the ACT, EF1-A, GAPDH, and COX1 housekeeping genes from UniProtKB in eudicots was used to find putative sequences in P. ruthii. Read counts per contig for the extended P. ruthii TPS candidate list were normalized taking into account the respective contig lengths (RPKMi [25]).
Identification of terpene synthase genes in Pityopsis ruthii
Protein coding regions of the flower transcriptome were predicted using Transdecoder 5.5.0 [26]. HMM files with N-domain (PF01397) and C-domain (PF03936) downloaded from (http://pfam.xfam.org/) were used as a query to identify TPSs using HMMER v3.3.1 [27] with an E-value of 1e-5. Multiple sequence alignment of putative PrTPS genes and TPS family of Arabidopsis thaliana [28] was performed using MAFFT v7.480 [29] with 1,000 iterate improvement. Maximum likelihood tree was built using MEGA v11.0 [30] with the JTT model and 1,000 bootstraps. Only bootstrap support values higher than 50% were shown in the tree.
TPS genes cloning and terpene synthase activity assays
Total RNA was isolated from fully open flowers using the RNeasy Plant Mini Kit (https://www.qiagen.com/) according to the manufacturer’s protocol. cDNA was synthesized using a First Strand cDNA Synthesis Kit (https://www.cytivalifesciences.com/) according to the manufacturer’s protocol. Primer Notl-(dt)18 (5’-AACTGGAAGAATTCGCGGCCGCAGGAA TTTTTTTTTTTTTTTTT-3’) was used for cDNA synthesis. Signal peptides were predicted using TargetP [31], and were cleaved before the conserved RRWx8W motif to create truncated recombinants, which typically improves activity of the soluble enzymes [32]. Full length cDNAs of PrTPS1, PrTPS2, PrTPS3, PrTPS4, PrTPS5, PrTPS6, and PrTPS7 obtained from the transcriptome analyses were amplified using gene specific primers extended with restriction enzyme sites (Table 1) and cloned into pGEM®-T Easy Vector (www.promega.com). After sequencing confirmation, they were cut with restriction enzymes and cloned into pET32a. The plasmids containing full length cDNAs were transformed into E. coli BL21 codon plus (DE3) for heterologous expression. E. coli harboring the expression plasmids were cultured at 37°C to an OD600 of 0.6, then expression was induced through addition of 1 m isopropyl-1-thio-d-galactopyranoside (IPTG) to liquid LB cultures until final concentration of 1 mM. The cells were harvested by centrifugation at 4000 × g after 16 h of culture at 22°C. Cell pellets were resuspended in chilled extraction buffer (50 mM MOPSO, pH 7.0, 5 mM MgCl2, 5 mM DTT, 5 mM Na-Ascorbate, 0.5 mM PMSF, 10% (v/v) glycerol) and disrupted by ultrasonication 6 × 30 s using cell disruptor (Misonix, Framingdale, NY). Supernatant obtained after centrifugation at 13000 × g to remove cell debris was desalted into assay buffer (10 mM MOPSO, pH 7.0, 1 mM dithiothreitol, 10% (v/v) glycerol) by passage through a PD-10 desalting column (http://www.cytivalifesciences.com/). Enzyme assays were performed in Teflon-sealed screw capped 2 mL glass vials containing 50 μL assay buffer (20 mM MOPSO, pH 7.0, 10 mM MgCl2, 1 mM DTT, 0.2 mM NaWO4, 0.1 mM NaF, 0.05 mM MnCl2), 50 μL desalted crude protein extract, and 4 μM FPP or GPP, respectively. Volatiles produced by each respective reaction were collected by solid-phase microextraction (SPME, https://www.sigmaaldrich.com/) for 30 min at 22°C and analyzed using GC-MS. Terpenes were identified as described previously [33]. As a negative control, the empty vector pET32a was introduced into E. coli Bl21 and the crude protein extracts were isolated and expressed with the substrates GPP, FPP, respectively, and analyzed as described above.
3-D modelling
The predicted translated amino acid (AA) sequences of PrTPS6 and PrTPS7 were analyzed in 3-dimensional models, owing to their very close similarity (Fig 3; S1 Fig). The sequences were aligned and projected using the Swiss Model server and their internal implemented software [34]. The display of 3-D structure of both coding sequences simultaneously was arranged to highlight the differences (red hue) vs. the AA identity (green hue) in the Swiss Model server over a web browser. Highlights of the identified conserved motifs were marked in MS PowerPoint.
Statistical analyses
All experiments were performed in at least three technical replicates, unless stated otherwise. Significance was calculated in one- or two-way ANOVA, respectively, including the analyses of factorial interaction. Post-hoc test of Fisher’s Honestly Significant Difference was carried-out at α = 0.05. All these were computed using R v4.2.2 and the package agricolae v1.3–5 [35].
Results
Terpene profiles in leaves, stems, flowers, and roots of Pityopsis ruthii plants
The volatile terpene chemistry of roots and rhizomes was assessed using organic extraction and the terpenoid compounds were identified using GC-MS. Two sesquiterpenes, β-elemene and allo-aromadendrene were detected from root tissues with concentrations of 14.14 μg × g FW-1 and 0.28 μg × g FW-1, respectively (Table 2). For the above-ground parts, stems of P. ruthii produced allo-aromadendrene at relatively low concentration of 0.80 μg × g FW-1 (Table 2). Six terpenes were detected in leaves and included two monoterpenes and four sesquiterpenes. Despite having a higher number of individual terpenes, the observed sum concentration of all terpenes in leaves was higher than that in stems (Table 2). Twelve terpenes including 4 monoterpenes and 8 sesquiterpenes were identified in the flower extracts (Table 2). Among them, the most abundant terpenes were two sesquiterpenes β-elemene (14.05 μg × g FW-1) and allo-aromadendrene (16.42 μg × g FW-1) that together accounted for 54.20% of total terpenes detected in the P. ruthii flowers extracts. Among four monoterpenes, the concentrations of myrcene (5.13 μg × g FW-1) and limonene (7.09 μg × g FW-1) were higher than those of α-pinene and β-pinene, accounting for 9.12% and 12.61% of total terpenes detected in the P. ruthii flower extracts, respectively. Except for allo-aromandendrene (16.42 μg × g FW-1) and β -elemene (14.05 μg × g FW-1), all other remaining sesquiterpenes were detected in low concentrations (< 5 μg × g FW-1). Flowers of P. ruthii were the primary organs that produced and stored volatile terpenes.
Flowers of Pityopsis ruthii emit a bouquet of volatile terpenes
Fully open flowers of P. ruthii are moderately fragrant to the human nose. To determine the chemical composition of their fragrance, volatiles emitted from the open flowers were collected using headspace collection and analyzed using GC-MS. Four monoterpenes, ten sesquiterpenes, and two diterpenes were detected from the flowers (Fig 1). The collective emission rate of volatile terpenes was 1632.74 ng × h-1 × g FW-1. The four major monoterpenes accounted for 68% (w/w) of all terpenes emitted; comparatively, the sesquiterpenes were emitted in lower levels of less than 50.00 ng × h-1 × g FW-1 (Table 3). α-Pinene was the most abundant volatile terpene emitted by flowers and accounted for almost one-half of the monoterpenes emitted but was undetected in leaves, stems, and roots. The second most abundant terpene was β-pinene. Similar to the extracts from leaves, sesquiterpenes were common in the flower volatiles. The most abundant sesquiterpenes were (E)-α-bergamotene, guaia-1(10), 11-diene, α-selinene, aromadendrene, allo- aromadendrene, and (E)-α-bergamotene. Emission rates of the remaining sesquiterpenes were ≤ 10.3 ng × h-1 × g FW-1-1 and varied from about 0.96 to 10.29 ng × h-1 × g FW-1.
A. Compounds collected from open flowers using headspace. 1. α-pinene; 2. β-pinene; 3. myrcene; 4. limonene; 5 internal standard (1-octanol); 6. β-elemene; 7. (E)-α-bergamotene; 8. unidentified sesquiterpene 1; 9. guaia-1(10),11-diene; 10. unidentified sesquiterpene 2; 11. aromadendrene; 12. germacrene D; 13. allo-aromadendrene; 14. α-selinene; 15. pogostol; 16. unidentified diterpene 3; 17. kaur-16-ene. B. chemical structures of known compounds identified from open flowers in Pityopsis ruthii.
Transcriptomic analyses of Pityopsis ruthii flowers
Sequencing of P. ruthii flower samples from three accessions yielded 191 million paired-end 150-bp Illumina reads (NCBI PRJNA778727). Trimming retained 99.8% of total reads with an average length of 145.6 bp. After removal of ribosomal reads, 98.8% of total reads were used for de novo assembly. Initial assembly from Trans-ABySS yielded a transcriptome consisting of 702,209 contigs of average length 528 bp and N50 of 630. Considering the putative introduced redundancy incorporated from merging the results of individual assembly runs, and to greatly reduce the number of similar sequences, RapClust was used. The reduced assembly consisted of 129,317 contigs of average length of 934 bp and N50 of 1,151. Despite the reduced transcriptome representing 18.4% of the initial one, quality expressed as BUSCO completeness was only slightly affected; from 1,440 BUSCOs searched, complete BUSCOs were reduced from 1,243 to 1,215 after clustering, and fragmented BUSCOs increased from 79 to 85. These results underscore the high representation of conserved orthologs sequences in the P. ruthii transcriptome, suggesting its high quality. In our reduced transcriptome, 300 sequences displayed homology with any of the 54 enzymes searched that are involved in terpene biosynthesis (S1 Table). The respective normalized read counts for those putative TPSs varied in the transcriptome (S2 Table).
Similar to other plants, two pathways are naturally involved in terpene synthesis in P.ruthii: the mevalonate (MVA) pathway located in the cytosol and methyl-erythritol phosphate (MEP/DOXP) pathway located in the plastids; both pathways produce two C5 isoprene precursors isopentenyl pyrophosphate (IPP) and its isomer dimethylallyl pyrophosphate (DMAPP). Geranyl pyrophosphate (GPP), farnesyl pyrophosphate (FPP) and geranylgeranyl diphosphate (GGPP), substrates for monoterpenes, sesquiterpenes and diterpenes, respectively, are synthesized by the condensation of IPP and DMAPP by Isoprenyl diphosphate synthases (IDS) (Jia and Chen, 2016) [36]. We investigated transcriptome-covered terpene pathways of P. ruthii (Fig 2) and assessed the relative expression of genes in both the MVA and the MEP/DOXP pathways from P. ruthii during flowering using the transcriptome analyses. Copy numbers (assumed isoforms) of AACT, HNGS, HMGR, MK, PMK, and PMD in the MVA pathway were 8, 6, 5, 2, 2, and 3, respectively. Whereas, copy numbers of DXS, DXR, CMS, CMK, MDS, HDS, and HDR in the MEP/DOXP pathway were 6, 2, 1, 2, 2, 4, and 3, respectively. The transcriptional levels of the initial steps in terpene backbone biosynthesis were relatively the highest: AACT, HMGS, and HMGR in the MVA pathway, and DXS and DXR in the MEP/DOXP pathway (Fig 2).
Heatmap of the data from flower transcriptome analyses performed in triplicate along with representative gene indicated expression level, which was calculated with fragments per kilobase of transcripts per million mapped fragments (FPKM). Abbreviations of genes: AACT, acetoacetyl-CoA thiolase; HMGS, hydroxylmethylglutaryl-CoA synthase; HMGR, hydroxymethylglutaryl-CoA reductase; MVK, mevalonate kinase; PMK, 5-phospho-mevalonate kinase, PMD, mevalonate diphosphate decarboxylase; FPPS, farnesyl pyrophosphate synthase; DXS, 1-deoxy-d-xylulose-5-phosphate synthase; DXR, 1-deoxy-d-xylulose-5-phosphate reductoisomerase; CMS, 2-C-methyl-d-erythritol 4-phosphate cytidylyltransferase; CMK, 4-diphosphocytidyl-2-C-methyl-d-erythritol kinase; MDS, 2-C-methyl-D-erythritol 2,4-cyclodiphosphate synthase; HDS, (E)-4-hydroxy-3-methylbut-2- enyl diphosphate synthase; HDR, 4-hydroxy-3-methylbut-2-enyl diphosphate reductase; IDI, isopentenyl-diphosphate delta-isomerase; GPPS, geranyl diphosphate synthase. Compound abbreviations: HMG-CoA, 3-hydroxy-3-methylglutaryl-CoA; MVA, mevalonate; DXP, 1-Deoxy-D-xylulose 5-phosphate; MEP, 2-C-Methyl-D-erythritol 4-phosphate; CDP-ME, 2-C-Methyl-d-erythritol-2,4-cyclodiphosphate; CDP-MEP, 2-Phospho-4-(cytidine 5’-diphospho)-2-C-methyl-D-erythritol; ME-CPP, 2-C-methyl-d-erythritol-2,4-cyclodiphosphate; HMBPP, 1-hydroxy-2-methyl-2-(E)-butenyl-4-diphosphate; FPP, (E,E)-farnesyl pyrophosphate; IPP, isopentenyl pyrophosphate; DMAPP, dimethyallyl pyrophosphate; GPP, geranyl pyrophosphate; GGPP, geranylgeranyl pyrophosphate.
Identifying putative terpene synthase genes in Pityopsis ruthii
The functional analyses of genes associated with terpene metabolism of P. ruthii flowers also leveraged the generated transcriptome data. The HMM files of Terpene_synth_N (PF01397) and Terpene_synth_C (PF03936) were used to search for the putative terpene synthases (TPSs) in the transcriptomes. In total, 130 unigenes were identified as putative TPS. After removal of the repeat sequences, 67 genes were identified as unique TPSs (S1 Table). Lengths of the translated proteins for these genes varied from 101 to 793 AA, and most of them were deemed partial open reading frames (ORFs) by comparison with confirmed TPS proteins from other plants that showed sequences longer than 540 AA. Other features supporting the claim of their full lengths were the presence of start and stop codons, and detection of α and β domains in TPSs-a, -b, and -g, or α, β, and γ domains in TPSs-c and TPSs-e/f. Whereas, partial ORFs contained only one of these domains. There were seven TPS genes that appeared to be full length and were designated PrTPS1 to PrTPS7 (GenBank accession numbers: ON166544 to ON166550). PrTPS1, PrtPS2, PrTPS3, PrTPS4, PrTPS5, PrTPS6, and PrTPS7 encode proteins of 592, 593,597, 587, 569, 550, and 548 AA in length, respectively. These seven PrTPS genes were analyzed phylogenetically with the TPS gene family of Arabidopsis. Based on this analysis, PrTPS1 to -3 were classified into the TPS-b subfamily, PrTPS5 to -7 into the TPS-a subfamily, and PrTPS4 into the TPS-g subfamily, respectively (Fig 3). At the N-terminal position of PrTPS1 to -3, a conserved motif RRX8W was identified, typical for the TPS-b subfamily, whereas no such motif was detected in PrTPS4 from the TPS-g subfamily. At the C-terminal position of PrTPS1 to -7, a conserved motif RDR was present, except PrTPS4 displaying a variant RDQ. Also, the Asp-rich DDxxD motif was detected in PrTPS1 to -7. The NSE/DTE motif was detected in PrTPS1 to -7 with a variant DDxxGxxxE in PrTPS6 and -7, and DDxxSxxxE in PrTPS4 (S1 Fig).
Phylogenetic tree was reconstructed by maximum likelihood method based on JTT model. The classification of TPS subfamilies was determined as previously reported [9].
Biochemical characterization of TPSs from Pityopsis ruthii
Based on the phylogenetic analysis, PrTPS1, PtTPS2, and PtTPS3 of the TPS-b subfamily and PrTPS4 of the TPS-g subfamily were predicted to encode monoterpene synthases based on previous analysis of the TPS gene family [9] Recombinant proteins expressed in E. coli for each of these four PrTPS genes were tested for monoterpene synthase activities using GPP as substrate. Both PrTPS1 (Fig 4A) and PrTPS3 (Fig 4B) produced α-pinene as a major product and β-pinene as a minor one using GPP as substrate. Neither PrTPS2 nor PrTPS4 showed any activity under the conditions tested.
Monoterpene activity of PrTPS1 (A) and PrTPS3 (B). Crude protein extracts from heterologous expression in E. coli catalyzed the conversion of the substrate GPP into monoterpenes. PrTPS genes are identified. Products were identified by GC-MS: 1. α-pinene; 2. β-pinene. Peak numbers are consistent with those reported in Fig 1.
PrTPS5, PrTPS6, and PrTPS7 of the TPS-a subfamily (Fig 3) were predicted to encode sesquiterpene synthases. These three genes were also expressed in E. coli to produce recombinant proteins, which were subject to sesquiterpene synthase activity assays. Recombinant PrTPS5 catalyzed the conversion of FPP into a single terpene, (E)-α-bergamotene (Fig 5A), which was one of the predominant sesquiterpenes identified in flowers (Fig 1A). PrTPS6 (Fig 5B) converted (E,E)-FPP to multiple sesquiterpenes with β-ylangene, γ-elemene, and germacrene D as predominant products and δ-elemene, (+) -cycloisosativene, and δ-cadinene as the minor ones. In contrast, PrTPS7 (Fig 5C) produced germacrene D as a major product and β-ylangene as a minor product. Besides accepting (E,E)-FPP, PrTPS5 also converted GPP to multiple monoterpenes: α-pinene, α-terpinolene, and myrcene as predominantly detected volatile monoterpenes, and limonene, α-phellandrene, (Z)-β-ocimene, and (E)-β-ocimene as the minor ones (S2 Fig).
Sesquiterpene synthase activity of PrTPS5 (A), PrTPS6 (B), and PrTPS7 (C). Crude proteins extracted from heterologous expression in E. coli catalyzed the conversion of the substrate FPP into sesquiterpenes. Products identified by GC-MS: 6. β-elemene; 7. (E)-α-bergamotene; 12. germacrene D; 19.δ-elemene; 20. (+)-cycloisosativene; 21. β- ylangene; 22. γ-elemene; 23. δ-cadinene. Peak numbers are consistent with those reported in Fig 1.
3-D modelling
To gain insights into functional variations of PrTPSs, analyses of the 3-D models of two very resemblant PrTPSs, PrTPS6 and PrTPS7, were accomplished in the Swiss Model. The AA identity between these two translated TPSs was 89.1% in the 548 residues overlap and gap frequency of 0.2%. The next-best similarity score among the cloned PrTPSs was between PrTPS1 and PrTPS3, with 54.3% identity in 549 residues overlap and the gap frequency of 2.4%. From the 3 regions identified as polymorphic between PrTPS6 and PrTPS7 (Fig 6), the most likely region affecting the substrate specificity and activities of either PrTPSs is the stretch of AA placed very close to the conserved regions RRx8W, NSE/DTE, RDR, and DDxxD (Fig 6; right edge). Data from the deduced AA sequences, phylogenetic analyses, and the 3-D models imply the changes in the protein’s active site architecture as underlying the observed differences in the biochemical assays (see above).
The models were computed by Swiss Model [34] using default parameters. Identical amino acid residues are shown in green; polymorphic amino acid chains are shown in red. The conserved motifs were identified using the Swiss Model server over web browser and marked per each respective coding frame.
Discussion
In this study, we analyzed the terpene profiles in tissues and emitted volatiles of P. ruthii, an endangered Asteraceae plant. We further supported that data with the transcriptomic analyses of terpene biosynthetic pathways, followed by cloning and functional analyses of seven PrTPS genes with enzyme activities in terpene biosynthesis confirmed for five of them. This information was leveraged towards the assessment of how the terpene profiles may help the plant to survive in its peculiar, highly variable ecological niche.
Terpenes, as predominant volatiles in flower scents, play important roles in plant reproductive biology by attracting pollinators or deterring florivores [37–41]. Floral terpenes, including linalool and geraniol, emit distinct aromas that attract pollinators such as bees and butterflies, thus enhancing pollination efficiency [42]. They can contribute to insect pollinator specificity by emitting specific blends of volatile compounds that attract and interact with specific pollinator species [43]. In combination with flower color, terpenes can coordinate with scent compounds to attract specific pollinators, promoting effective pollination and reproductive success [44]. Contrastingly, terpenes such as β-caryophyllene and α-pinene have been shown to deter florivores, such as herbivorous insects, by acting as repellents or causing toxicity [45].
Another important example of physiologically relevant terpenes is the abscisic acid, a terpenoid plant hormone, that plays a critical role in regulating plant responses to drought stress. It acts as a signaling molecule and helps the plants close their stomata, thereby reducing the transpiration-based water loss. Terpenes can also enhance plant water-use efficiency and improve drought tolerance by modulating plant physiology and metabolism [46]. Beyond helping the plants withstand the drought stress, terpenes have been shown to enhance plant tolerance to cold and freezing temperatures. They act as cryoprotectants by reducing the freezing point of cellular fluids, preventing ice crystal formation, and maintaining membrane stability. Additionally, terpenes can regulate the expression of genes involved in cold acclimation and promote the synthesis of protective proteins and enzymes. Monoterpenes, such as α-pinene and limonene, have been found to enhance the cold tolerance of plants by regulating the expression of cold-responsive genes and protecting the photosynthetic machinery from cold-induced damage [47]. Sesquiterpenes β-caryophyllene and α-humulene have been found to enhance freezing tolerance by reducing ice nucleation and promoting ice formation at higher subzero temperatures, protecting plant tissues from freeze-induced damage [48]. Terpenes, including isoprene and monoterpenes, have been implicated in the process of cold acclimation, enhancing the overall cold tolerance of plants by modulation of various physiological and biochemical responses [49]. Beyond freezing tolerance, terpenes are imparting salt and osmotic stress tolerance to plants: They help regulate ion transport and osmotic balance, reducing the toxic effects of high salt concentrations on plant cells [50].
Our results are consistent with the general observation that terpenes are major constituents of floral scents in many plant species [7, 42, 51–53]; this may also be the function of the floral terpenes produced by P. ruthii [3, 7]. α-Pinene is the predominant monoterpene detected from P. ruthii flowers and probably acts as an important chemical cue to attract pollinators, and similar to observations in Eucalyptus polybractea [54], or the moths Helicoverpa armigera reacting to this terpene [55]. Contrastingly, α-pinene was identified as a repellent of bee pollinators in melon flowers [56]. β-Pinene has shown potential in enhancing plant resistance against various stresses, including pathogen attack and oxidative stress [57]. Besides the pinenes, limonene was another monoterpene detected from P. ruthii flowers; it was previously identified as an attractant for bee pollinators [56]. Both limonene and myrcene attracted bumblebee pollinators [58].
Most terpenes, including some identified from P. ruthii in this study, also serve as plant chemical defense molecules [59]. They may be toxic to microbial pathogens and/or insect pests [60–62]. The variation in terpene species and concentrations in organs of P. ruthii plants may provide specific defenses to the various types of natural enemies they encounter, such as herbivorous insects and pathogenic fungi. For instance, during a previous reintroduction effort, we found mealybugs in many of the collected seeds samples (pers. obs.; [3]); terpenes emitted from the desiccating flower heads may have been of suboptimal composition or concentration to repel the pest that foraged on the seeds and further endangered the plant occurrence. When plants are attacked by pests, blends of terpenes are emitted from various tissues[40, 41, 63, 64]. α-Bergamotene has been involved in plant defenses against herbivorous insects and possesses antimicrobial properties [65]. β-Elemene could be important in the establishment of mycorrhizae [66], which may aid in drought tolerance and more efficient absorption of various mineral compounds (e.g., phosphorus) from nutritionally poor sites. Furthermore, phytoalexins derived from the terpenoids in Zea mays roots may be associated with drought tolerance [67] or defense against biotic soilborne pathogens, including nematodes [68, 69].
Sequenced genomes in Asteraceae contain large families of terpene synthases. Many full length TPS genes were predicted in Helianthus annuus (n = 99 [70]), Chrysanthemum nankingense (n = 59 [53]), C. seticuspe (n = 66 [53]), and Artemisia annua (n = 88 [71]), respectively. Consistent with the TPS gene abundance in other Asteraceae, 67 TPS unigenes were identified in our study of P. ruthii. The expansion of the TPS gene family in many plants is the generally accepted underlying mechanism behind the diversity of the terpenoids produced on the one hand, and the neofunctionalization and spatio-temporal variability of expression of TPSs on the other hand [9, 38, 40, 72]. For the PrTPSs, we observed a possible duplication of PrTPS6 and PrTPS7 due to their unusually high sequence identity, close phylogenetic placement, and the striking overlap of their 3-dimensional models. Thus, the AA polymorphisms in the 3-D model can be underlying the detected differences in the proteins’ activity, as the sequence mutation clearly affected the enzyme active site architecture. Due to lack of the high-quality genomic resources for P. ruthii, the claim of expansion and neofunctionalization awaits verification in future research.
Of the seven cloned putative PrTPSs, activity detection failed for two cloned candidates. One plausible reason could be the missing parts of ORFs, inherent to the transcriptome based ORF finding and cloning. Other reasons may be related to their expression in insoluble form or improper folding, or to the improper reaction composition including the buffering agent or pH, species and/or concentration of the metal divalent ions, or species of the substrate used [73–77]. In our study, the most abundant volatile terpenoids detected are monoterpenes α-pinene, β-pinene produced by PrTPS1 and PrTPS3, myrcene and limonene, which are the common in the floral volatile blend of genus Chrysanthemum [53]. Only one sesquiterpene, (E)-α-bergamotene produced by PrTPS5 was detected in vivo, and no products of the heterologously expressed PrTPS6 and PrTPS7 were detected neither in the extracts nor in the emitted volatiles. Their enzymatic products probably get quickly converted to other nonvolatile products; for example, volatile sesquiterpenes β-bisabolene and β-macrocarpene get readily converted to nonvolatile zealexins (β-bisabolene derivatives, β-macrocarpene derivative) by cytochrome P450 [78]. Similar was observed for many other plants including Asteraceae species: the volatile terpenes did not accumulate or get emitted, but instead were converted to more polar terpenoids. One interesting observation was that of the promiscuity of PrTPS5 that accepted both tested substrates. Such a mechanism of terpene diversity was recorded for several plants as one of the ways towards the range of products formed [74, 76]. Overall, several of the PrTPSs identified here may be of interest for heterologous expression and accretion of rarely occurring terpene species or whose synthesis has been thus far otherwise hampered [37, 73, 77, 79].
Isopentenyl pyrophosphate (IPP) and its isomer dimethyallyl pyrophosphate (DMAPP) are produced by MVA and MEP/DOXP pathways in plants. Compared with C. nankingense and C. seticuspe, the copy number of some of the enzymes involved in both pathways of P. ruthii are dramatically different. For example, AACT and HMGS had eight and six assumed isoforms, respectively, whereas only two assumed isoforms were detected each in both C. nankingense and C. seticuspe [53]. This suggested that P. ruthii requires more copies of AACT and HMGS to ensure its supply of the basic five-carbon units for terpene biosynthesis. Furthermore, HMGR is a known rate-limiting step for the MVA pathway and the last of the initial reactions with multiple isoforms [53]. Similarities in the higher number of assumed isoforms and in the biochemical background of the reactions that involve NAPDH suggests the same rate-limiting role of DXR in the MEP/DOXP pathway for P. ruthii. Multiple assumed isoforms detected in the flower transcriptomes suggested a comparatively higher turnover rates in the terpene biosynthetic machinery than in the other analyzed organs that were documented in our volatile profiling. This feature also points to the metabolic flexibility and tissue/organ specific expression patterns [63], to be analyzed in follow-up studies. Finally, richness in the terpene species we detected may be related to this group of compounds playing roles in thermal and oxidative stresses. This feature may render P. ruthii uniquely suited to the particularly harsh environments it is most often found in ([3]; A.J. Dattilo, unpublished), with barely any other plants competing for the habitat.
In summary, we have documented and characterized volatile terpene chemistry in P. ruthii. The insights provided in our study will lay the foundation for determining the mechanisms underlying some aspects of adaptation of P. ruthii to a very harsh habitat that is subject to abiotic and biotic stresses. Considering the known biological functions of terpenes, the diverse and tissue- or development-specific production of monoterpenes and sesquiterpenes in P. ruthii suggests that they contribute to its adaptation to its niche environment. This emerging hypothesis is undergoing assessment for this species as well as for other related endangered Asteraceae plant systems.
Supporting information
S1 Table. Pityopsis ruthii transcriptomic contigs with homology to any of the 54 KEGG enzymes involved in terpene biosynthesis.
Reads for each RNA sample (s.13; s.15; s.16) and sums thereof are presented before (SEQ) and after RapClust (CLUS), respectively.
https://doi.org/10.1371/journal.pone.0287524.s001
(TXT)
S2 Table. Pityopsis ruthii transcriptomic normalized read counts for putative TPSs.
Reads for each RNA sample (s.13; s.15; s.16) and sums thereof are presented before (SEQ) and after RapClust (CLUS), respectively. Those are then normalized using a given contig length [bp] and expressed for before (RPKMi) and after RapClust (RapClust RPKMi), respectively.
https://doi.org/10.1371/journal.pone.0287524.s002
(TXT)
S3 Table. Raw data underlying the Tables 2 and 3.
https://doi.org/10.1371/journal.pone.0287524.s003
(CSV)
S1 Fig. Alignment of translated protein sequences of 7 terpene synthases from Pityopsis ruthii cloned in this study.
Identical amino acids are highlighted with increasingly darker backgrounds. Motifs typically found in TPSs are boxed with red boarders, and the respective motif names are identified below.
https://doi.org/10.1371/journal.pone.0287524.s004
(TIF)
S2 Fig. Monoterpene synthase activity of PrTPS5.
Crude proteins extracted from heterologous expression in E. coli catalyzed the conversion of the substrate GPP into monoterpenes. Products identified by GC-MS: 1. α-pinene; 2. Myrcene; 3. α-phellandrene; 4. Limonene; 5. (Z)-β-ocimene; 6. (E)-β-ocimene; 7. α-terpinolene; 8. unidentified monoterpene.
https://doi.org/10.1371/journal.pone.0287524.s005
(TIF)
Acknowledgments
Plants were collected under Tennessee Valley Authority Permit # TE117405-2 and U.S. Fish and Wildlife Service Permit # TE134817-1.
References
- 1. Congress U. Endangered species act. US Code. 1973;16:1534–44.
- 2.
New York and Oxford: efloras.org; 2006 [cited 2023 2023-03-06].
- 3. Moore PA, Wadl PA, Skinner JA, Trigiano RN, Bernard EC. Current knowledge, threats, and future efforts to sustain populations of Pityopsis ruthii (Asteraceae), an endangered southern Appalachian species. The Journal of the Torrey Botanical Society. 2016;143(2):117–34.
- 4.
White AJ. An autecological study of the endangered species, Heterotheca ruthii (Small) Harms: University of Tennessee, Knoxville; 1977.
- 5. Ruth’s Golden Aster (Pityopsis ruthii). 5-year review: Summary and evaluation: U.S. Fish and Wildlife Service; 2018 [cited 2023 2023-03-06]. Available from: https://ecos.fws.gov/docs/five_year_review/doc5953.pdf.
- 6. Cruzan M, Beaty P. Population biology or Ruth’s Golden Aster (Pityopsis ruthii): Final report. ID-96-05937-6-00. Tennessee Department of Environment and Conservation …, 1998.
- 7. Moore PA, Klingeman WE, Wadl PA, Trigiano RN, Skinner JA. Seed production and floral visitors to Pityopsis ruthii (Asteraceae: Asterales), an endangered Aster native to the Southern Appalachians. Journal of the Kansas Entomological Society. 2021;93(4):327–48.
- 8. Dudareva N, Negre F, Nagegowda DA, Orlova I. Plant volatiles: recent advances and future perspectives. Critical reviews in plant sciences. 2006;25(5):417–40.
- 9. Chen F, Tholl D, Bohlmann J, Pichersky E. The family of terpene synthases in plants: a mid‐size family of genes for specialized metabolism that is highly diversified throughout the kingdom. The Plant Journal. 2011;66(1):212–29. pmid:21443633
- 10. Gershenzon J, Dudareva N. The function of terpene natural products in the natural world. Nature chemical biology. 2007;3(7):408–14. pmid:17576428
- 11. Pichersky E, Raguso RA. Why do plants produce so many terpenoid compounds? New Phytologist. 2018;220(3):692–702. pmid:27604856
- 12. Sharkey TD, Wiberley AE, Donohue AR. Isoprene emission from plants: why and how. Annals of botany. 2008;101(1):5–18. pmid:17921528
- 13. Xu M, Jiang Y, Chen S, Chen F, Chen F. Herbivory-induced emission of volatile terpenes in Chrysanthemum morifolium functions as an indirect defense against Spodoptera litura larvae by attracting natural enemies. Journal of Agricultural and Food Chemistry. 2021;69(34):9743–53.
- 14. Zhang K, Jiang Y, Zhao H, Köllner TG, Chen S, Chen F, et al. Diverse terpenoids and their associated antifungal properties from roots of different cultivars of Chrysanthemum morifolium Ramat. Molecules. 2020;25(9):2083. pmid:32365690
- 15. Wadl PA, Rinehart TA, Dattilo AJ, Pistrang M, Vito LM, Milstead R, et al. Propagation for the conservation of Pityopsis ruthii, an endangered species from the southeastern United States. HortScience. 2014;49(2):194–200.
- 16. Chen F, Al-Ahmad H, Joyce B, Zhao N, Köllner TG, Degenhardt J, et al. Within-plant distribution and emission of sesquiterpenes from Copaifera officinalis. Plant Physiology and Biochemistry. 2009;47(11–12):1017–23.
- 17. Song L, Florea L. Rcorrector: efficient and accurate error correction for Illumina RNA-seq reads. GigaScience. 2015;4(1):s13742–015-0089-y. pmid:26500767
- 18. Jiang H, Lei R, Ding S-W, Zhu S. Skewer: a fast and accurate adapter trimmer for next-generation sequencing paired-end reads. BMC bioinformatics. 2014;15:1–12.
- 19. Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nature methods. 2012;9(4):357–9. pmid:22388286
- 20. Robertson G, Schein J, Chiu R, Corbett R, Field M, Jackman SD, et al. De novo assembly and analysis of RNA-seq data. Nature methods. 2010;7(11):909–12.
- 21. Trapnell C, Hendrickson DG, Sauvageau M, Goff L, Rinn JL, Pachter L. Differential analysis of gene regulation at transcript resolution with RNA-seq. Nature biotechnology. 2013;31(1):46–53. pmid:23222703
- 22. Patro R, Duggal G, Love MI, Irizarry RA, Kingsford C. Salmon provides fast and bias-aware quantification of transcript expression. Nature methods. 2017;14(4):417–9. pmid:28263959
- 23. Simão FA, Waterhouse RM, Ioannidis P, Kriventseva EV, Zdobnov EM. BUSCO: assessing genome assembly and annotation completeness with single-copy orthologs. Bioinformatics. 2015;31(19):3210–2. pmid:26059717
- 24. Tenenbaum D, RUnit S, Maintainer MBP, Carlson M, biocViews Annotation P, ThirdPartyClient K. Package ‘keggrest’. R Foundation for Statistical Computing: Vienna, Austria. 2019.
- 25. Cheplyaka R. RNA-Seq normalization explained 2016 [cited 2023 2023-03-06]. Available from: https://ro-che.info/articles/2016-11-28-rna-seq-normalization.
- 26. Haas BJ, Papanicolaou A, Yassour M, Grabherr M, Blood PD, Bowden J, et al. De novo transcript sequence reconstruction from RNA-seq using the Trinity platform for reference generation and analysis. Nature protocols. 2013;8(8):1494–512. pmid:23845962
- 27. Finn RD, Clements J, Eddy SR. HMMER web server: interactive sequence similarity searching. Nucleic acids research. 2011;39(suppl_2):W29–W37. pmid:21593126
- 28. Aubourg S, Lecharny A, Bohlmann J. Genomic analysis of the terpenoid synthase (AtTPS) gene family of Arabidopsis thaliana. Molecular Genetics and Genomics. 2002;267:730–45. pmid:12207221
- 29. Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Molecular biology and evolution. 2013;30(4):772–80. pmid:23329690
- 30. Tamura K, Stecher G, Kumar S. MEGA11: molecular evolutionary genetics analysis version 11. Molecular biology and evolution. 2021;38(7):3022–7. pmid:33892491
- 31. Almagro Armenteros JJ, Salvatore M, Emanuelsson O, Winther O, von Heijne G, Elofsson A, et al. Detecting novel sequence signals in targeting peptides using deep learning. bioRxiv; 2019.
- 32. Lee S, Chappell J. Biochemical and genomic characterization of terpene synthases in Magnolia grandiflora. Plant Physiology. 2008;147(3):1017–33.
- 33. Zhuang X, Köllner TG, Zhao N, Li G, Jiang Y, Zhu L, et al. Dynamic evolution of herbivore‐induced sesquiterpene biosynthesis in sorghum and related grass crops. The Plant Journal. 2012;69(1):70–80. pmid:21880075
- 34. Waterhouse A, Bertoni M, Bienert S, Studer G, Tauriello G, Gumienny R, et al. SWISS-MODEL: homology modelling of protein structures and complexes. Nucleic acids research. 2018;46(W1):W296–W303. pmid:29788355
- 35. de Mendiburu F, de Mendiburu MF. Package ‘agricolae’. R Package, version. 2019;1(3).
- 36. Jia Q. and Chen F., 2016. Catalytic functions of the isoprenyl diphosphate synthase superfamily in plants: a growing repertoire. Molecular plant, 9(2), pp.189–191. pmid:26751961
- 37. Barreto IC, de Almeida AS, Sena Filho JG. Taxonomic insights and its type cyclization correlation of volatile sesquiterpenes in Vitex species and potential source insecticidal compounds: A review. Molecules. 2021;26(21):6405. pmid:34770814
- 38. Boncan DAT, Tsang SS, Li C, Lee IH, Lam H-M, Chan T-F, et al. Terpenes and terpenoids in plants: Interactions with environment and insects. International Journal of Molecular Sciences. 2020;21(19):7382. pmid:33036280
- 39. Dudareva N, Pichersky E. Biochemical and molecular genetic aspects of floral scents. Plant physiology. 2000;122(3):627–34. pmid:10712525
- 40. Huang X, Xiao Y, Köllner TG, Zhang W, Wu J, Wu J, et al. Identification and characterization of (E)-β-caryophyllene synthase and α/β-pinene synthase potentially involved in constitutive and herbivore-induced terpene formation in cotton. Plant Physiology and Biochemistry. 2013;73:302–8.
- 41. Zhang W, Jiang Y, Chen F, Guan Z, Wei G, Chen X, et al. Dynamic regulation of volatile terpenoid production and emission from Chrysanthemum morifolium capitula. Plant Physiology and Biochemistry. 2022;182:11–21.
- 42. Knudsen JT, Gershenzon J. The chemical diversity of floral scent. Biology of plant volatiles. 2020:57–78.
- 43. Schiestl FP. The evolution of floral scent and insect chemical communication. Ecology letters. 2010;13(5):643–56. pmid:20337694
- 44. Junker RR, Blüthgen N. Floral scents repel potentially nectar-thieving ants. Evolutionary Ecology Research. 2008;10(2):295–308.
- 45. Dicke M, Baldwin IT. The evolutionary context for herbivore-induced plant volatiles: beyond the ‘cry for help’. Trends in plant science. 2010;15(3):167–75.
- 46. Cutler SR, Rodriguez PL, Finkelstein RR, Abrams SR. Abscisic acid: emergence of a core signaling network. Annual review of plant biology. 2010;61:651–79. pmid:20192755
- 47. Wu J, Zhang Y, Yin L, Qu J, Lu J. Linkage of cold acclimation and disease resistance through plant–pathogen interaction pathway in Vitis amurensis grapevine. Functional & integrative genomics. 2014;14:741–55. pmid:25154381
- 48. Holopainen JK, Gershenzon J. Multiple stress factors and the emission of plant VOCs. Trends in plant science. 2010;15(3):176–84. pmid:20144557
- 49. Sharkey TD, Monson RK. Isoprene research–60 years later, the biology is still enigmatic. Plant, cell & environment. 2017;40(9):1671–8.
- 50. Hasanuzzaman M, Nahar K, Fujita M. Plant response to salt stress and role of exogenous protectants to mitigate salt-induced damages. Ecophysiology and responses of plants under salt stress. 2013:25–87.
- 51. Jiang Y, Chen X, Lin H, Wang F, Chen F. Floral scent in Wisteria: chemical composition, emission pattern, and regulation. Journal of the American Society for Horticultural Science. 2011;136(5):307–14.
- 52. Jiang Y, Qian R, Zhang W, Wei G, Ma X, Zheng J, et al. Composition and biosynthesis of scent compounds from sterile flowers of an ornamental plant Clematis florida cv.‘Kaiser’. Molecules. 2020;25(7):1711. pmid:32276485
- 53. Jiang Y, Zhang W, Chen X, Wang W, Köllner TG, Chen S, et al. Diversity and biosynthesis of volatile terpenoid secondary metabolites in the Chrysanthemum genus. Critical Reviews in Plant Sciences. 2021;40(5):422–45.
- 54. Goodger JQ, Sargent D, Humphries J, Woodrow IE. Monoterpene synthases responsible for the terpene profile of anther glands in Eucalyptus polybractea RT Baker (Myrtaceae). Tree Physiology. 2021;41(5):849–64.
- 55. Cunningham JP, Moore CJ, Zalucki MP, West SA. Learning, odour preference and flower foraging in moths. Journal of Experimental Biology. 2004;207(1):87–94. pmid:14638836
- 56. Fernandes N, Silva FAN, de Aragão FA, Zocolo GJ, Freitas BM. Volatile organic compounds role in selective pollinator visits to commercial melon types. 2019.
- 57. Aoki Y, Van Trung N, Suzuki S. Impact of Piper betle leaf extract on grape downy mildew: effects of combining 4-allylpyrocatechol with eugenol, α-pinene or β-pinene. Plant Protection Science. 2018;55(1):23–30.
- 58. Byers KJ, Vela JP, Peng F, Riffell JA, Bradshaw HD Jr. Floral volatile alleles can contribute to pollinator‐mediated reproductive isolation in monkeyflowers (Mimulus). The Plant Journal. 2014;80(6):1031–42. pmid:25319242
- 59. Aharoni A, Jongsma MA, Bouwmeester HJ. Volatile science? Metabolic engineering of terpenoids in plants. Trends in plant science. 2005;10(12):594–602. pmid:16290212
- 60. Birkett MA, Bruce TJ, Pickett JA. Repellent activity of Nepeta grandiflora and Nepeta clarkei (Lamiaceae) against the cereal aphid, Sitobion avenae (Homoptera: Aphididae). Phytochemistry Letters. 2010;3(3):139–42.
- 61. Bruce TJ, Wadhams LJ, Woodcock CM. Insect host location: a volatile situation. Trends in plant science. 2005;10(6):269–74. pmid:15949760
- 62. Prisic S, Xu M, Wilderman PR, Peters RJ. Rice contains two disparate ent-copalyl diphosphate synthases with distinct metabolic functions. Plant physiology. 2004;136(4):4228–36. pmid:15542489
- 63. Irmisch S, Krause ST, Kunert G, Gershenzon J, Degenhardt J, Köllner TG. The organ-specific expression of terpene synthase genes contributes to the terpene hydrocarbon composition of chamomile essential oils. BMC Plant Biology. 2012;12(1):1–13.
- 64. Zhan X, Qian Y, Mao B. Metabolic profiling of terpene diversity and the response of prenylsynthase-terpene synthase genes during biotic and abiotic stresses in Dendrobium catenatum. International Journal of Molecular Sciences. 2022;23(12):6398.
- 65. Zhang P-J, Zheng S-J, van Loon JJ, Boland W, David A, Mumm R, et al. Whiteflies interfere with indirect plant defense against spider mites in Lima bean. Proceedings of the National Academy of Sciences. 2009;106(50):21202–7. pmid:19965373
- 66. Gualandi RJ, Augé RM, Kopsell DA, Ownley BH, Chen F, Toler HD, et al. Fungal mutualists enhance growth and phytochemical content in Echinacea purpurea. Symbiosis. 2014;63:111–21.
- 67. Vaughan MM, Christensen S, Schmelz EA, Huffaker A, Mcauslane HJ, Alborn HT, et al. Accumulation of terpenoid phytoalexins in maize roots is associated with drought tolerance. Plant, cell & environment. 2015;38(11):2195–207. pmid:25392907
- 68. Lin J, Wang D, Chen X, Köllner TG, Mazarei M, Guo H, et al. An (E, E)‐α‐farnesene synthase gene of soybean has a role in defence against nematodes and is involved in synthesizing insect‐induced volatiles. Plant Biotechnology Journal. 2017;15(4):510–9.
- 69. Lorimer SD, Perry NB, Foster LM, Burgess EJ, Douch PG, Hamilton MC, et al. A nematode larval motility inhibition assay for screening plant extracts and natural products. Journal of Agricultural and Food Chemistry. 1996;44(9):2842–5.
- 70. Goodstein DM, Shu S, Howson R, Neupane R, Hayes RD, Fazo J, et al. Phytozome: a comparative platform for green plant genomics. Nucleic acids research. 2012;40(D1):D1178–D86. pmid:22110026
- 71. Shen Q, Zhang L, Liao Z, Wang S, Yan T, Shi P, et al. The genome of Artemisia annua provides insight into the evolution of Asteraceae family and artemisinin biosynthesis. Molecular plant. 2018;11(6):776–88.
- 72. Karunanithi PS, Zerbe P. Terpene synthases as metabolic gatekeepers in the evolution of plant terpenoid chemical diversity. Frontiers in plant science. 2019;10:1166. pmid:31632418
- 73. Degenhardt J, Köllner TG, Gershenzon J. Monoterpene and sesquiterpene synthases and the origin of terpene skeletal diversity in plants. Phytochemistry. 2009;70(15–16):1621–37. pmid:19793600
- 74. Johnson SR, Bhat WW, Sadre R, Miller GP, Garcia AS, Hamberger B. Promiscuous terpene synthases from Prunella vulgaris highlight the importance of substrate and compartment switching in terpene synthase evolution. New Phytologist. 2019;223(1):323–35.
- 75. Lau KH, Bhat WW, Hamilton JP, Wood JC, Vaillancourt B, Wiegert-Rininger K, et al. Genome assembly of Chiococca alba uncovers key enzymes involved in the biosynthesis of unusual terpenoids. DNA Research. 2020;27(3):dsaa013.
- 76. Li D-S, Hua J, Luo S-H, Liu Y-C, Chen Y-G, Ling Y, et al. An extremely promiscuous terpenoid synthase from the Lamiaceae plant Colquhounia coccinea var. mollis catalyzes the formation of sester-/di-/sesqui-/mono-terpenoids. Plant Communications. 2021;2(5):100233. pmid:34746763
- 77. Rinaldi MA, Ferraz CA, Scrutton NS. Alternative metabolic pathways and strategies to high-titre terpenoid production in Escherichia coli. Natural Product Reports. 2022;39(1):90–118.
- 78. Ding Y, Weckwerth PR, Poretsky E, Murphy KM, Sims J, Saldivar E, et al. Genetic elucidation of interconnected antibiotic pathways mediating maize innate immunity. Nature plants. 2020;6(11):1375–88. pmid:33106639
- 79. Kirby J, Keasling JD. Biosynthesis of plant isoprenoids: perspectives for microbial engineering. Annual review of plant biology. 2009;60:335–55. pmid:19575586