Functional Characterization of a New Bifunctional Terpene Synthase LpNES1 from a Medicinal Plant Laggera pterodonta.

Laggera pterodonta, known in China as 'Choulingdan' for its stimulous odor, has long been used as traditional herbal medicine. The essential oil of L. pterodonta, which exhibits various pharmacological activities, is a rich resource of monoterpenes and sesquiterpenes. To date, however, the terpene synthases responsible for their production remain unknown. In present study, a new terpene synthase gene (LpNES1) was identified from L. pterodonta, transcript level of which was significantly upregulated in response to methyl jasmonate treatment. Recombinant LpNES1 could synthesize (E)-nerolidol and minor β-farnesene from farnesyl diphosphate and linalool from geranyl diphosphate in vitro. Whereas, only sesquiterpenes including (E)-nerolidol and minor β-farnesene were released when LpNES1 was reconstituted in yeast, even coexpressed with a geranyl diphosphate synthase (ERG20WW). Combined with subcellular localization experiment, the result indicated that the cytosol-targeted LpNES1 was responsible for (E)-nerolidol biosynthesis exclusively in L. pterodonta. Additionally, the expression level of LpNES1 gene was more prominent in floral buds than that in other tissues. LpNES1 characterized in present study not only lays the molecular foundation for sesquiterpene biosynthesis of L. pterodonta, but provides a key element for further biosynthesis of bioactive compound in microbes.


Introduction
Laggera pterodonta, belonging to Asteraceae family, is widely distributed in the sub-Saharan Africa and in the southwestern China. As a traditional herbal medicine, L. pterodonta has long been used as ethnomedicine to treat bronchitis, pneumonia, and epidemic influenza, etc. 1,2 . Additionally, the essential oils of L. pterodonta are reported to exhibit antiviral 3 , antioxidant 4 , and antifungal properties 5 . Due to these important pharmacological activities and practical applications, up to now, many studies have focused on characterization of the chemical compositions of the essential oils and/or extracts of L. pterodonta, which lead to the identification of 42 monoterpenoids, and 151 sesquiterpenoids 1, 6 .
In higher plants, terpenoids are biosynthesized from two universal building blocks, isopentenyl diphosphate IPP and dimethylallyl diphosphate DMAPP , which are pro-duced by two distinct pathways, the plastidic methylerythritol 4-phosphate MEP pathway and the cytosolic mevalonate MVA pathway. Further condensation of DMAPP and 1-3 IPP s gives rise to geranyl diphosphate GPP , farnesyl diphosphate FPP , and geranylgeranyl diphosphate GGPP by corresponding prenyltransferases. Subsequently, these prenyldiphosphate precursors are converted into a wide range of structurally diverse mono-, sesqui-, and diterpenes by the action of terpene synthases TPSs 7,8 . In general, mono-and di-terpenes are produced compartmentally in plastids, whereas sesquiterpenes are produced in the cytosol 9 . A few studies, however, reveal that monoand diterpenes can be formed in the cytosol and sesquiterpenes can be formed in plastids 10 12 , suggesting that there is a cross-talk between subcellular compartments for their substrates formation in plants.
To date, many terpenoids have been identified from L.
2.2 Identification of terpene synthases and sequence analysis Transcripts were annotated by using as a query for BLASTX-searches by UniProtKB/Swiss-Prot25 and KEGG https://www.kegg.jp/kegg/kegg1.html . Phylogenetic analyses were performed using MEGAX to generate a neighborjoining tree using 1000 bootstrap replicates. Multiple amino acid sequence alignments of LpNES1 and other TPSs belonging to TPS-g subfamily were aligned using ClustalW 2.1 16 .

Plasmids construction
Plasmids used and constructed in this study are listed in Table S1. All the primers are listed in Table S2. Full-length ORF sequence of LpNES1 gene was amplified using primers LpNES1-F/R and subcloned into the pMD20-T vector to get 20T-LpNES1. For expression in Escherichia coli, PCR-amplified LpNES1 from 20T-LpNES1 was inserted into the BamH I/Hind III site of plasmid pCold-ProS2 to generate expression plasmid ProS2-LpNES1; For overexpression of LpNES1 in yeast, LpNES1 gene was codonoptimized and synthesized by GenScript Inc China , design a t e d a s O P -L p N E S 1 G e n e b a n k A c c e s s i o n N O : MZ190022 . The auxotrophic marker gene LEU2 was amplified from vector pRS605 stored in our lab , upstream and downstream homologous region of the GAL80 gene indicted as 80up and 80down respectively , and terminator tTEF1, tCYC1 were amplified from the genomic DNA of Saccharomyces cerevisiae BY4742. The 80up-LEU-tCYC1 and tTEF1-80down fragments were obtained through overlap extension PCR OE-PCR and then sequentially ligated to vector pESC-URA harboring the GAL1,10 bidirectional promoter by Xho I/BamH I and Not I/Sac I, respectively, to get backbone plasmid pZK10. The OP-Lp-NES1 was then ligated to plasmid pZK10 to get pZK11 finished by GenScript . The mutant ERG20 WW F96W-N127W gene was obtained by OE-PCR according to Ignea, C. et al. 13 , first cloned to pMD-20T and then subcloned to plasmid pZK10/11 at BamH I site using infusion ligase Biomed Beijing China to get pZK12/13, respectively. Final fragments 80up-leu-GAL1-ERG20 WW -80down and 80upleu-GAL1,10-ERG20 WW , OP-LpNES1-80down, could be obtained through Xho I/Sac I digestion from plasmid ZK-12 and ZK-13, respectively Fig S1 ; For transient expression of LpNES1 in Nicotiana benthamiana, full-length of LpNES1 without stop codon , fused with GFP gene, driven by 35S promoter and terminator tNOS1 were subcloned into plasmid pCAMBIA1301 using infusion ligase to obtain plasmid 1301-LpNES1-GFP.

Expression and purification of the recombinant Lp-NES1
ProS2-LpNES1 was chemically transformed into Rosetta DE3 for expression. The expression and purification procedures for recombinant LpNES1 were performed as previously described 17 with some modifications: 25 μg/mL ampicillin and 17 μg/mL chloramphenicol were used during cultivation and the expression was induced at 15 , 170 rpm/min for 20 hours with 0.8 mM IPTG. The Ni-NTA resin purified protein was concentrated, desalted, and exchanged with PBS buffer 2 mM KH 2 PO 4 , 8 mM Na 2 HPO 4 , 136 mM NaCl, 2.6 mM KCl, pH 7.4 . The ProS2 tag was removed by HRV 3C Protease TaKaRa Bio, Beijing, China , and reaction was performed according to the instructions. Purified LpNES1 was detected by 12 SDS-PAGE and the protein concentration was determined using the BCA Protein Assay Kit TaKaRa Bio, Beijing, China .

Functional characterization of LpNES1 in vitro
To assay catalytic activity of the purified LpNES1, the reaction was conducted in 20 mL headspace bottle with 25-50 μg purified protein in assay buffer 25 mM HEPES pH 7.4, 100 mM KCl, 7.5 mM MgCl 2 , 1 mM MnCl 2 , 5 v/v glycerol, 5 mM DTT , and initiated by addition of 6 μg of three prenyldiphosphate precursors, respectively, at 30 for 2 h. Terpene products were collected and analyzed by solid-phase micro extraction SPME and gas chromatography-mass spectrometry GC-MS . Heat-inactivated at 95 for 15 min protein was used as a negative control. Products were identified by comparison of retention time and mass spectra with those of authentic standards.

Functional characterization of LpNES1 in yeast
Gel-purified 80up-leu-GAL1-ERG20 WW -80down and 80up-leu-GAL1,10-ERG20 WW , OP-LpNES1-80down fragments were transformed into S. cerevisiae BY4742 to obtain strain 4742-01 control and 4742-02, respectively, by the LiAc/PEG/ssDNA method 18 . Transformants were selected on SD-leu solid medium 20 g/L glucose, 15 g/L agar, 6.7 g/L yeast nitrogen base, 2 g/L amino acids without leucine and further validated by primers Δ80-F and Δ80-R Table S1 . For solid fermentation, verified colonies were inoculated on yeast extract-peptone-dextrose YPD plates at 28 , for 6 days. Pellets were transferred to a 20 mL headspace bottle for SPME-GC-MS analysis. Liquid fermentation was performed by transferring the overnight culture to 20 mL of fresh YPD medium with an initial OD 600 of 0.2, terpenes were captured by overlaying dodecane 10 v/v . The mixture was then incubated at 28 , 200 rpm for 6 days. Then, the dodecane layer was harvested, dehydrated with anhydrous Na 2 SO 4 , and detected by GC-MS.

Quantitative real-time polymerase chain reaction
qRT-PCR The total RNA was extracted using the plant RNA extraction kit Omega Guangzhou, China . Reverse transcription was performed using PrimeScript TM RT reagent Kit TaKaRa Bio, Beijing, China and the final cDNAs were diluted 4-fold before use. 18S ribosomal RNA was used as the internal reference gene, and the primers were designed using NCBI Primer software https://www.ncbi. nlm.nih. gov/tools/primer-blast/, Table S1 . qRT-PCR was performed by Applied Biosystems AB 7500 Real-time PCR system and AB power SYBR green PCR master mix Life Technologies . The expression levels of LpNES1 in the four various tissues were calculated using the 2 CT method. All the experiments were performed by three biological replicates and three technical replicates.

GC-MS analysis
Samples were injected at 1 μL volume in splitless mode on Agilent 7890A GC coupled with 5977B inert XL MS detector Agilent Technologies . The separation was achieved by using a HP-5MS column 30 m 0.25 mm 0.25 μm, Agilent Technologies following the program: 50 for 1 min, ramped at 8 min 1 to 300 for 5 min. MSD Chem-Station Data Analysis program was used for data processing Agilent Technologies . Details of GC-MS analysis were described in the supporting information.

Identi cation and phylogenetic analysis of LpNES1
Based on comparative transcriptome analysis of L. pterodonta, one transcript named LpNES1, which was upregulated with 8-fold increase after MeJA treatment, was screened out. The full-length LpNES1 Genebank Accession NO: MW701393 contained an open reading frame ORF of 1749 bp encoding a protein of 493 amino acids with theoretical isoelectric point pI and molecular weight of 5.69 and 56.9 kDa, respectively http://web.expasy.org/ compute_pi/ . According to the multiple alignments Fig.  1 , despite from different origins, these terpene synthases shared the highly conserved motifs, including the absence of the RRx8W motif in the N-terminal domain which is essential for cyclic monoterpenes biosynthesis 19 , and two highly conserved aspartate-rich motifs, consistently occurring as DDTFD and DDLGS C AKD N E among these enzymes. This two conserved motifs were demonstrated to be located oppositely in the entrance of active cave for fixing the phosphate group of prenyldiphosphate precursors via binding Mg 2 or Mn 2 cofactors 20 22 . To define the evolutionary relationship of LpNES1, the phylogenetic tree was constructed based on the representative TPSs from other species Table S1 , which were downloaded from NCBI database. Inspection of the phylogenetic tree clearly indicated the LpNES1 clustered into the terpenoid synthase TPS subfamily TPS-g Fig. 2 .

Cytosol-targeted LpNES1
According to the software, LpNES1 gene was predicted to be localized in the cytosol, as described in the supporting information. To experimentally determine the subcellular localization of LpNES1 gene, the agroinfiltrated leaves were analyzed for transient GFP expression using confocal laser scanning microscopy. It was obvious that GFP fluorescence was observed in the cytosol of protoplasts ex- pressing LpNES1-GFP Fig. 3 , indicating LpNES1 gene was targeted to the cytosol.

In vitro biochemical characterization of recombinant
LpNES1 revealed its bifunction It is difficult to accurately predict the products profiles of a candidate TPS enzyme only by protein sequence analysis 23 . Therefore, to determine the catalytic function of LpNES1, enzymatic assay experiment was conducted. The purified LpNES1 protein was detected in SDS-PAGE Fig.  S2 . GPP, FPP, and GGPP were used as substrates for enzymatic assay in vitro, and products were analyzed by SPME-GC-MS. When incubated with GPP, LpNES1 protein could form the linalool as its main compound Fig. 4A ; when incubated with FPP, LpNES1 protein could form the E -nerolidol as its principal product, accompanied with minor β -farnesene Fig. 4C . LpNES1 protein did not accept GGPP as a substrate for no terpenoids were generated Fig. S3 , and no terpenoids were detected in control assays. The major products of the recombinant LpNES1 exhibited identical retention time and mass spectrum to the linalool and nerolidol standard Figs. 4B and 4D .

Formation of sesquiterpenes exclusively by LpNES1
in yeast To further investigate whether LpNES1 could produce  Table S2 .
the same terpene products in eukaryote, we introduced the OP-LpES1 codon-optimized LpNES1 gene into S. cerevisiae BY4742. Given the Gal80p was a negative regulatory factor that could repress the GAL promoter 24 , we disrupted the GAL80 gene by integrating the expression cassettes of OP-LpNES1 at this loci. Initially, linalool was not detected by SPME-GC-MS data not shown . This might be attributed to the fact that the pool of GPP in yeast is too small, for GPP was an intermediate compound for FPP synthesis 25 . To improve the pool of GPP, the mutant ERG20 WW F96W-N127W gene, the encoding protein of which showed more efficient in biosynthesis of GPP 26 was coexpressed with OP-LpNES1 under the control of Gal1,10 bidirectional promoter in strain 4742-02. Surprisingly, compare to strain 4742-01 control , only sesquiterpenes including E -nerolidol, and minor β-farnesene were released from strain 4742-02, and still no monoterpenes were detected Fig. 5A . In order to avoid the errors caused by volatility of linalool, we further detected the products captured by dodecane from liquid fermentation, which got the same results Fig. S4 .

Expression analysis of LpNES1 gene
Biosynthesis and emission of volatile terpenoids from specific plant tissues often correlate with the spatial expression of their terpene synthases 27 . In order to investigate the spatial regulation patterns of LpNES1 expression in L. pterodonta, relative gene expression analysis was carried out in various tissues. The results showed that the expression level of LpNES1 was significantly predominant in floral buds and leaves, approximately 300-and 180-fold higher than that in roots respectively Fig. 5B .  . Reaction products were achieved by incubation of LpNES1 and control heat-inactive LpNES1 with GPP A and FPP C as substrate. Peaks were identified and confirmed by the mass spectra reference library and authentic standard including linalool and E -nerolidol. The mass spectrums of peak 1 and peak 2 were identical to that of the standard linalool B and E -nerolidol D respectively; peak 3 was β-farnesene which was identified by mass spectra reference library. TIC, total ion chromatogram; RT, retention time; m/z, mass-to-charge ratio.

Discussion
LpNES1, characterized in this study, belongs to TPS-g clade proteins. One prevalent property of these proteins is their acyclic products such as linalool, E -nerolidol, and geranyllinalool, which is attributed to lack of RRx8W motif shared by them 28 . Moreover, the members from TPS-g clade are more promiscuous than conventional terpene synthases in regard of substrates, for two or more of prenyldiphosphate precursors including GPP, FPP, and/or GGPP can be converted into linalool, nerolidol, and/or geranyllinalool 29,30 . This characteristic is similar to that of an aromatic prenyltransferase AtaPT from Aspergillus terreus, which could accept three prenyldiphosphate precursors DMAPP, GPP, FPP . The study also revealed the residue G326, the short hydrophobic side chain of which contributes to enlarge substrate binding pocket, was key for AtaPT to accept prenyl diphosphates with longer chain such as GPP and FPP 31 . We, therefore, speculated that enzymes of TPS-g family might possess a spacious activesite cavity that allows for accommodating substrates of varying size. To shed light on the catalytic mechanism of linalool/nerolidol synthases with different substrates and the determinants of substrate specificity, further work with protein crystal structure will be badly needed.
Generally, FPP biosynthesis through MVA pathway is known to happen in the cytosolic compartment in plants 9 . However, it has also been revealed that a small pool of GPP exists in the cytoplasm of plants 11,12 . Similar to that of plant cell, in the cytosol of yeast, FPP is the major component synthesized by ERG20 with a minor component of GPP released through the same pathway 32 . Thus, whether a small pool of cytosolic GPP can be encountered by linalool/nerolidol synthases like LpNES1 and converted into lin-alool in vivo is interesting to explore. Several studies demonstrate that cytosolic linalool/nerolidol synthase are only responsible for biosynthesis of E -nerolidol when expression in recombinant yeast or in plants 33 35 . The same result was obtained in our study that only FPP was applied by LpNES1 for E -nerolidol generation in yeast. In fact, previous study demonstrated that S. cerevisiae cell harboured enough free GPP that could be catalytically converted into monoterpenes through introducing monoterpene synthases 36 . Moreover, ERG20 WW was overexpressed under the bidirectional GAL1,10 promoter to further improve the pool of GPP in yeast. Thus, we can say, LpNES1 exhibited preference to FPP than GPP. Combined with cytosolic location, LpNES1 was suggested to be responsible for E -nerolidol synthesis exclusively in L. pterodonta.
MeJA is often used experimentally as mimic of herbivory or abiotic stress to elicit the expression of many defense genes and the production of defense compounds 37,38 . The LpNES1 gene was obviously up-regulated by the MeJA treatment, indicating its significant role in plant defense. E -nerolidol is ubiquitous compound in floral aroma of flowers, involved in attracting pollinators 39 . Together with 4,8-dimethylnona-1,3,7-triene DMNT , a degraded derivative from E -nerolidol, they are emitted from plant leaves in response to stimuli and exert defensive activities in a direct or indirect way 40, 41 . Thus, LpNES1 might be involved in ecological interactions against abiotic stress or biotic factors through regulating the biosynthesis of E -nerolidol. Fig. 5 A GC-MS analysis of the terpene products produced by BY4742-02 expressing OP-LpNES1 and ERG20 WW and control BY4742-01 . Peak 4 was phenylethyl alcohol; B qRT-PCR analysis of LpNES1 in four different tissues including the root, stem, leaf, and floral bud. Amplification of 18S ribosomal RNA gene was applied as an internal control. The value represents the means of the results, which were generated via six samples: three biological replicates and three technical replicates and standard error were shown.

Conclusion
In our study, a new terpene synthase LpNES1 was first identified from L. pterodonta Asteraceae . Functional characterization in vitro and demonstrated that LpNES1 was a bifunctional linalool/nerolidol synthase. Further investigation in yeast indicated that LpNES1 possessed a robust activity for E -nerolidol generation whereas no linalool formation. Gene expression patterns showed that LpNES1 was expressed prominently in floral buds and leaves. The LpNES1 gene identified in our study can be applied for further E -nerolidol production in microbes through metabolic engineering strategy.