Sequence analysis of target genes
The full amino acid sequences translated from ORFs (primers are provided in table S1) were used for multiple sequence alignent performing on an ESPript (http://espript.ibcp.fr/ESPript/ESPript/) as described previously (Robert et al. 2014). Phylogenetic trees were constructed by MEGA7 using the Neighbor-Joining method, and Interactive Tree of Life (iTOL ) was used for visual editing (http://itol2.embl.de/) (Letunicet al. 2007). Analysis of signal peptides in TPSs, and prediction localization of candidate proteins were performed on Phyre2 (http://www.sbg.bio.ic.ac.uk/phyre2) and signal P (http://www.cbs.dtu.dk/services/SignalP-4.1/) as described previously (Kelley et al.2015). Protein secondary structure was predicted using I-TASSER (http://zhanglab.ccmb.med.umich.edu/I-TASSER).
Determination oftissue-specific gene expression in A. thaliana
Approximately 1.5 kb of a promoter for each gene was isolated by FPNI-PCR from genomic DNA using three specific gene primers and three arbitrary primers and PCR settings described by Wang et al. (2011). Resulting sequences were analysed on PlantCARE (bioinformatics.psb.ugent.be/webtools/plantcare/html) (Rombauts et al. 1999) and then integrated into pCXGUS-P vector harbouring a cloning cassette upstream of GUS(uidA ). After the constructs were confirmed via sequencing, they were introduced into GV3101Agrobacterium , which was used to transform A. thaliana using a floral dip method described previously (Clough et al. 1998). Transformed lines were selected on MS plates containing 50 µg·mL-1hygromycin, and were then confirmed via PCR. Enzymatic assay using 5-bromo-4-chloro-3-indolyl-β-D-glucuronide (GUS) was performed according to the method of Jefferson et al. (1987) to assess candidate genes tissue-specific expression.
Subcellular localization of the LaTPS7, LaTPS8, and LaCYP71D582 in N. benthamiana
GV3101 with eGFP expression vectors controlled by a mannopine synthase promoter (MAS), constructed using a Trelief™ SoSoo Cloning Kit (TSINGKE, China), were used for transient transformation of 4-week-oldN. benthamiana plant as described by Jin et al. (2015). After 3 days of infiltration, N. benthamiana leaves were excised, mounted on slides, and analyzed using a confocal laser-scanning microscope equipped with a standard filter set (Leica TCS SP5). Images were processed by Image J (https://imagej.nih.gov/ij).
Heterologous expression of TPSs in Escherichia coli and enzymatic assayin vitro
TPS sequences of truncated predicted signal peptide were inserted into a pDE2 vector with a recombinant C-terminal and poly-histidine tag using primers shown in table S1 and a pDE2 Directional Expression Kit Ver.2 (TSINGKE, China). The final constructs was transformed into E. coli BL21 (DE3); recombinant proteins were subjected to isopropyl-β-D-thiogalactopyranoside (IPTG) induction at the concentration of 1.0 mM until OD600 reached 0.6~0.8, and then to purification using a His-Tagged Gravity Column (Merck Millipore). The purified proteins were assayed by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) to ensure that the proteins had been extracted successfully.
The in-vitro enzymatic assay for TPS activity was performed according to the method of Chen et al. (2011) in a final volume of 500 μL buffer [25mM HEPES, pH 7.3; 10 mM MgCl2; 10 mM MgCl2; 10% glycerol; 10 mM DTT], approximately 20 μL pure protein, and 10 µg of either FPP, NPP, or GPP (Sigma-Aldrich). The mixtures were vortexed, incubated at 30°C for 2 h, after which 250 μL hexane was added to each mixture and vortexed for 1 min. The upper layers were centrifuged at 1200 g and 4°C for 30 min, and then transferred into 2-mL glass vials for analysis via GC-MS. Heat-inactivated recombinant protein was used as negative control.