High-level diterpene production by transient expression in Nicotiana benthamiana

Abstract

Background: Characterization of plant terpene synthases is typically done by production of recombinant enzymes in Escherichia coli. This is often difficult due to solubility and codon usage issues. Furthermore, plant terpene synthases which are targeted to the plastids, such as diterpene synthases, have to be shortened in a more or less empirical approach to improve expression. We report here an optimized Agrobacterium-mediated transient expression assay in Nicotiana benthamiana for plant diterpene synthase expression and product analysis.

Results: Agrobacterium-mediated transient expression of plant diterpene synthases in N. benthamiana led to the accumulation of diterpenes within 3 days of infiltration and with a maximum at 5 days. Over 50% of the products were exported onto the leaf surface, thus considerably facilitating the analysis by reducing the complexity of the extracts. The robustness of the method was tested by expressing three different plant enzymes, cembratrien-ol synthase from Nicotiana sylvestris, casbene synthase from Ricinus communis and levopimaradiene synthase from Gingko biloba. Furthermore, co-expression of a 1-deoxy-D-xylulose-5-phosphate synthase from tomato and a geranylgeranyl diphosphate synthase from tobacco led to a 3.5-fold increase in the amount of cembratrien-ol produced, with maximum yields reaching 2500 ng/cm2.

Conclusion: With this optimized method for diterpene synthase expression and product analysis, a single infiltrated leaf of N. benthamiana would be sufficient to produce quantities required for the structure elucidation of unknown diterpenes. The method will also be of general use for gene function discovery, pathway reconstitution and metabolic engineering of diterpenoid biosynthesis in plants.

Figure 1

Production of CBT-ol. (A) Total ion chromatograms of hexane washes from agro-infiltrated N. benthamiana leaves co-expressing p19 and NsCBTS2a or the empty T-DNA vector (pL1F-1) as a control. The CBT-ol peak is framed by vertical dashed lines. The α- and β-stereoisomers of CBT-ol could not be separated under standard GC-MS conditions. (B) Total ion chromatograms of total extracts from agro-infiltrated N. benthamiana leaves. The extracts were purified over SPE column prior to GC-MS analysis. (C) CBT-ol content in agro-infiltrated leaves harvested three and five days after transformation of N. benthamiana plants with NsCBTS2a and p19. Hexane washes from the surface of treated leaves (black bars) and hexane extracts from the rest of the leaf material (total extracts, grey bars) were analyzed by GC-MS. No CBT-ol was detected in leaves transformed with the empty T-DNA vector (see Additional file 3). Samples were taken from two N. benthamiana plants, in each case from two infiltrated leaves. Mean CBT-ol values ± standard error of the mean (SEM) are presented (n = 8). There is a statistically significant difference (P < 0.001, two-way-ANOVA) in the mean CBT-ol values regarding the timepoint of sampling post-infiltration. ISTD – internal standard sclareol.

Figure 2

Production of casbene and of levopimaradiene. (A) Total ion chromatograms of hexane washes from N. benthamiana leaves agro-infiltrated with p19 and either the empty T-DNA vector (control) or RcCS or GbLS. Peaks for casbene, levopimaradiene and other novel products (a – e) are indicated. (B) Casbene and levopimaradiene content in hexane washes from leaves agro-infiltrated with p19 and RcCS or GbLS, respectively. Samples were taken five days post-infiltration from two individual N. benthamiana plants, each of them having two completely infiltrated leaves. Mean casbene and levopimaradiene values ± SEM are presented (n = 16, two independent experiments). Sclareol was used as internal standard (ISTD) for quantification.

Figure 3

Increasing expression levels of isoprenoid precursor pathway genes DXS and GGPPS lead to increased CBT-ol production in agro-infiltrated N. benthamiana leaves. (A) CBT-ol content in hexane washes from N. benthamiana leaves co-expressing p19 and NsCBTS2a with SlDXS2 and/or NtGGPPS2 in different combinations. The appropriate Agrobacterium cultures were infiltrated into separate parts of the same N. benthamiana leaf and in two leaves per individual plants. Leaves were harvested five days post-infiltration. Mean CBT-ol values ± SEM are presented (n = 8). Groups indicated by different letters differ significantly from each other regarding their CBT-ol values (P < 0.05, Student’s t-test and one-way ANOVA). ISTD – internal standard sclareol; (B) Full total ion chromatograms demonstrating high CBT-ol peak (dashed lines) when NsCBTS2a was co-expressed with both SlDXS2 and NtGGPPS2. No CBT-ol was detected in leaves agro-infiltrated with the empty T-DNA vector as a control.