Cloning and functional characterization of three branch point oxidosqualene cyclases from Withania somnifera (L.) dunal

Abstract

Oxidosqualene cyclases (OSCs) positioned at a key metabolic subdividing junction execute indispensable enzymatic cyclization of 2,3-oxidosqualene for varied triterpenoid biosynthesis. Such branch points present favorable gene targets for redirecting metabolic flux toward specific secondary metabolites. However, detailed information regarding the candidate OSCs covering different branches and their regulation is necessary for the desired genetic manipulation. The aim of the present study, therefore, was to characterize members of OSC superfamily from Withania somnifera (Ws), a medicinal plant of immense repute known to synthesize a large array of biologically active steroidal lactone triterpenoids called withanolides. Three full-length OSC cDNAs, β-amyrin synthase (WsOSC/BS), lupeol synthase (WsOSC/LS), and cycloartenol synthase (WsOSC/CS), having open reading frames of 2289, 2268, and 2277 bp, were isolated. Heterologous expression in Schizosaccharomyces pombe, LC-MS analyses, and kinetic studies confirmed their monofunctionality. The three WsOSCs were found to be spatially regulated at transcriptional level with WsOSC/CS being maximally expressed in leaf tissue. Promoter analysis of three WsOSCs genes resulted in identification of distinct cis-regulatory elements. Further, transcript profiling under methyl jasmonate, gibberellic acid, and yeast extract elicitations displayed differential transcriptional regulation of each of the OSCs. Changes were also observed in mRNA levels under elicitations and further substantiated with protein expression levels by Western blotting. Negative regulation by yeast extract resulted in significant increase in withanolide content. Empirical evidence suggests that repression of competitive branch OSCs like WsOSC/BS and WsOSC/LS possibly leads to diversion of substrate pool toward WsOSC/CS for increased withanolide production.

Keywords: Gene Regulation; Isoprenoid; Mass Spectrometry (MS); Metabolic Engineering; Western Blot.

FIGURE 1.

Scheme of proposed withanolide biosynthesis pathway. The abbreviations of the pathway intermediates are as follows: GA-3P, glyceraldehyde-3-phosphate; DXS, 1-deoxy-d-xylulose 5-phosphate synthase; DOXP, 1-deoxy-d-xylulose 5-phosphate pathway; DXR, 1-deoxy-d-xylulose-5-phosphate reductoisomerase; MEP, 2-C-methyl-d-erythritol 4-phosphate; DMAPP, dimethylalyl pyrophosphate; IPP, isopentenyl pyrophosphate; IPI, isopentenyl diphosphate isomerase; HMG-CoA, 3-hydroxy-3-ethylglutaryl-coenzyme A; SQS, squalene synthase; SQE, squalene epoxidase; CS, cycloartenol synthase; CPR, cytochrome P450 reductase; BS, β-amyrin synthase; LS, lupeol synthase. Branch A indicates the branch leading to the biosynthesis of sterol/withanolide, and Branch B indicates the branch leading to the biosynthesis of β-amyrin and lupeol. Single dark arrows represent one step; multiple dark arrows represent multiple steps.

FIGURE 2.

Comparison of deduced amino acid sequences of WsOSC/BS, WsOSC/LS, and WsOSC/CS with other plant OSCs using ClustalW2 multiple alignment tool. OSCs used for the multiple alignment were from S. lycopersicum (SlBS/1–761, NCBI reference sequence: NP_001234604.1; SlLS/1–756, NCBI reference sequence: XP_004243674.1; SlCS/1–757, NCBI reference sequence: NP_001233784.1), Aralia elata (AeBS/1–763, NCBI reference sequence: ADK12003.1), V. vinifera (VvCS/1–766, NCBI reference sequence: XP_002264372.1), L. japonicas (LjLS/1–755, NCBI reference sequence: BAE53431.1), and W. somnifera (WsOSC/BS/1–762, NCBI reference sequence: JQ728553, WsOSC/LS/1–755; NCBI reference sequence: JQ728552; WsOSC/CS/1–758, NCBI reference sequence: HM037907). Motifs are indicated as follows: prenyltransferase and squalene oxidase repeat (Motif A), MWCYCR and MLCYCR motif (Motif B), catalytic Asp (Motif C), and terpene synthase signature (Motif D).

FIGURE 3.

Predicted three-dimensional models and ligand-binding sites for WsOSCs. A–C, ribbon model display of the three-dimensional structures of WsOSC/BS (A), WsOSC/LS (B), and WsOSC/CS (C) as predicted by Phyre² web server, using the crystal structure of human OSC (Protein Data Bank code 1w6ka) as a template for modeling of all the three proteins. D–E, predicted ligand-binding sites (highlighted in blue at the core of the structure) in WsOSC/BS (D), WsOSC/LS (E), and WsOSC/CS (F) as predicted by 3DLigandSite web server.

FIGURE 4.

Phylogenetic tree of WsOSC/BS, WsOSC/LS, and WsOSC/CS. Phylogenetic analysis was performed using the ClustalW program and MEGA 5 software based on the neighbor-joining method. OSCs grouped into three subgroups, namely OSC1 (cycloartenol synthase), OSC2 (lupeol synthase), and OSC3 (β-amyrin synthase). WsOSC/BS, WsOSC/LS, and WsOSC/CS clustered with their respective subgroups. 23 protein sequences used for analysis were from subsequent plant species: S. lycopersicum (S. lycopersicum BS, NCBI reference sequence: NP_001234604.1; S. lycopersicum LS, NCBI reference sequence: XP_004243674.1; S. lycopersicum CS, NCBI reference sequence: NP_001233784.1), L. japonicas (L. japonicas BS, NCBI reference sequence: BAE53429.1; L. japonicas LS, NCBI reference sequence: BAE53430.1), Euphorbia tirucalli (E. tirucalli BS, NCBI reference sequence: BAE43642.1), Artemisia annua (A. annua BS, NCBI reference sequence: ACA13386.1), B. kaoi (B. kaoi BS, NCBI reference sequence: AAS83468.1), V. vinifera (V. vinifera BS, NCBI reference sequence: XP_002270934.1; V. vinifera LS, NCBI reference sequence: XP_002269060.1; V. vinifera CS, NCBI reference sequence: XP_002264372.1), Taraxacum officinale (T. officinale LS, NCBI reference sequence: BAA86932.1), Glycyrrhiza uralensis (G. uralensis LS, NCBI reference sequence: BAL41371.1), O. europaea (O. europaea LS, NCBI reference sequence: BAA86930.1), A. thaliana (A. thaliana CS, NCBI reference sequence: AAC04931.1), Azadirachta indica (A. indica CS, NCBI reference sequence: AGC82085.1), Centella asiatica (C. asiatica CS, NCBI reference sequence: AAS01524.1), and Panax notoginseng (P. notoginseng CS, NCBI reference sequence: ABY60426.1).

FIGURE 5.

Identification of W. somnifera β-amyrin, lupeol, and cycloartenol synthases by use of the yeast heterologous expression system. A–D, extracted ion chromatograms of standards β-amyrin, lupeol, and cycloartenol and S. pombe cells transformed with expression constructs pDS472aB (A), pDS472aL (B), pDS472aC (C), and empty vector pDS472a (D). E–J, the MS data of pDS472Ab (E), β-amyrin standard (F), pDS472aL (G), lupeol standard (H), pDS472aC (I), and cycloartenol standard (J). K–P, fragmentation pattern of β-amyrin (K), pDS472aB (L), lupeol (M), pDS472aL (N), pDS472aC (O), and cycloartenol (P).

FIGURE 6.

SDS-PAGE profile of purified recombinant proteins. Shown are the results of SDS-PAGE (10%) of purified recombinant proteins from S. pombe transformed with pDS472aB, pDS472aL, and pDS472aC. First lane, purified recombinant GST-fused WsOSC/BS; second lane, purified recombinant GST fused WsOSC/LS; third lane, standard protein marker; fourth lane, purified recombinant GST-fused WsOSC/CS.

FIGURE 7.

Kinetic study of WsOSC/BS, WsOSC/LS, and WsOSC/CS. A–C, Michaelis-Menten plots of β-amyrin synthase (A, WsOSC/BS), lupeol synthase (B, WsOSC/LS), and cycloartenol synthase (C, WsOSC/CS) with 2,3-oxidosqualene. Kinetic parameters K_m and V_max were obtained by fitting the data in the Michaelis-Menten equation by nonlinear regression analysis using GraphPad Prism 5 software.

FIGURE 8.

Tissue-specific real time expression analysis. A–C, quantitative estimation of the expression of WsOSC/LS (A), WsOSC/BS (B), and WsOSC/CS (C) in leaf, roots, stalk, and berries of W. somnifera. The data were compared and analyzed with analysis of variance. The values are means with standard errors indicated by bars, representing three independent biological samples, each with three technical replicates. Differences were scored as statistical significance at p < 0.05 (*) and p < 0.01 (**) levels.

FIGURE 9.

Transcript profiles of WsOSCs in response to elicitor treatments. A, time courses of WsOSC/BS, WsOSC/LS, and WsOSC/CS expression in micropropagated W. somnifera elicited by MeJA (0.1 mm), GA₃ (0.1 mm), and YE (0.1% w/v). β-Actin was kept as endogenous control. B, densitometric quantification of WsOSC/BS, WsOSC/LS, and WsOSC/CS band intensities for the different treatments and controls (ethanol and water). Experiments were performed in triplicate with similar results; error bars indicate ± standard deviation of the mean. IOD, integrated optical density; A.U., arbitrary units.

FIGURE 10.

Western immunoblot of WsOSCs in response to elicitor treatments. A–C, time courses of WsOSC/BS, WsOSC/LS, and WsOSC/CS protein expression in micropropagated W. somnifera elicited by MeJA (A, 0.1 mm), GA₃ (B, 0.1 mm), and YE (C, 0.1% w/v). β-Actin was kept as endogenous control.

FIGURE 11.

Time course effect of elicitor treatments on accumulation of withanolides. A–C, withanolide accumulation in response to 0.1 mm MeJA (left panel), 0.1 mm GA₃ (middle panel), and 0.1% w/v YE (right panel) at different time courses. Variation in three key withanolides: WS-1, WS-2, and WS-3, was confirmed by HPLC analysis at 6, 12, 24, and 48 h. All values obtained were means of triplicates with standard errors. Time course accumulation of WS-1, WS-2, and WS-3 was statistically significant at the p < 0.01 level.

FIGURE 12.

Southern blot analysis of WsOSC/BS (A), WsOSC/LS (B), and WsOSC/CS (C). W. somnifera genomic DNA was digested with SpeI (non-cutter) and ScaI and EcoRI (single-cutter) for WsOSC/BS, with SalI (non-cutter) and NcoI and EcoRV (single-cutter) for WsOSC/LS, and with XbaI and XhoI (non-cutter) and HindIII and DraII (single cutter) for WsOSC/CS; separated on 0.8% agarose gel; blotted onto a nylon membrane; and hybridized with DIG-labeled ORF of WsOSC/BS, WsOSC/LS, and WsOSC/CS as probes.