Production of highly bioactive resveratrol analogues pterostilbene and piceatannol in metabolically engineered grapevine cell cultures

Abstract

Grapevine stilbenes, particularly trans-resveratrol, have a demonstrated pharmacological activity. Other natural stilbenes derived from resveratrol such as pterostilbene or piceatannol, display higher oral bioavailability and bioactivity than the parent compound, but are far less abundant in natural sources. Thus, to efficiently obtain these bioactive resveratrol derivatives, there is a need to develop new bioproduction systems. Grapevine cell cultures are able to produce large amounts of easily recoverable extracellular resveratrol when elicited with methylated cyclodextrins and methyl jasmonate. We devised this system as an interesting starting point of a metabolic engineering-based strategy to produce resveratrol derivatives using resveratrol-converting enzymes. Constitutive expression of either Vitis vinifera resveratrol O-methyltransferase (VvROMT) or human cytochrome P450 hydroxylase 1B1 (HsCYP1B1) led to pterostilbene or piceatannol, respectively, after the engineered cell cultures were treated with the aforementioned elicitors. Functionality of both gene products was first assessed in planta by Nicotiana benthamiana agroinfiltration assays, in which tobacco cells transiently expressed stilbene synthase and VvROMT or HsCYP1B1. Grapevine cell cultures transformed with VvROMT produced pterostilbene, which was detected in both intra- and extracellular compartments, at a level of micrograms per litre. Grapevine cell cultures transformed with HsCYP1B1 produced about 20 mg/L culture of piceatannol, displaying a sevenfold increase in relation to wild-type cultures, and reaching an extracellular distribution of up to 45% of total production. The results obtained demonstrate the feasibility of this novel system for the bioproduction of natural and more bioactive resveratrol derivatives and suggest new ways for the improvement of production yields.

Keywords: Vitis vinifera; grapevine cell culture; metabolic engineering; piceatannol; pterostilbene; resveratrol.

Figure 1

The biosynthetic pathway of stilbene compounds from sucrose as carbon source. The conversion of sucrose carbon into stilbenoids in grapevine cells involves several pathways from primary metabolism, that is glycolysis, pentose phosphate shunt and shikimate, to produce the amino acid phenylalanine as precursor of all polyphenolic compounds, and secondary metabolism, that is phenyl propanoid, to produce 4‐coumaroyl CoA as precursor of stilbenoids (Langcake and Pryce, 1977). The action of STS leads to the production of resveratrol as major stilbenoid, which is strongly induced (red dotted lines) under elicitation conditions with methylated β‐cyclodextrin (MBCD) and methyl jasmonate (MeJA). In transformed grapevine cell lines with either Vv ROMT or Hs CYP1B1 (green arrows), resveratrol is converted into methylated (pterostilbene) or hydroxylated (piceatannol) derivate, respectively. STS, stilbene synthase; Vv ROMT, resveratrol O‐methyltransferase from Vitis vinifera; Hs CYP1B1, human Cytochrome P450 hydroxylase 1B1.

Figure 2

Production of t‐Pn and t‐Pt in planta by agroinfiltration. ROMT and CYP1B1 activities were assessed in planta using Nicotiana benthamiana agro‐infiltration. The tobacco leaves were excised six days after agro‐infiltration and extracts analysed by HPLC‐ESI‐MS. Extracted ion chromatograms from resveratrol (XIC 229), piceatannol (XIC 245) and pterostilbene (XIC 257) are shown. (a) Standards. (b) Control, leaf extract agroinfiltrated with rolD‐EgfpER construct (Martínez‐Márquez et al., 2015). (c) Leaf extract agroinfiltrated with STS inducible by MBCD+MeJA. (d) Leaf extract agroinfiltrated with ROMT and STS inducible by MBCD+MeJA. (e) Leaf extract agroinfiltrated with CYP1B1 and STS inducible by MBCD+MeJA. Peaks: t‐Pc, trans‐piceid; t‐R, trans‐resveratrol; t‐Pn, trans‐piceatannol; t‐Pt, trans‐pterostilbene, small amounts of cis‐resveratrol (c‐R) and cis‐pterostilbene (c‐Pt) isomers are also present. Identity of peaks was confirmed by HPLC‐ESI‐MS/MS analysis (data not shown). Note that t‐Pc can be detected as a molecular ion [M+H⁺]=229 due to extensive in‐source fragmentation of this compound.

Figure 3

PCR amplification products from genomic DNA of transgenic callus lines of Vitis vinifera Monastrell and Gamay varieties. Amplification was carried out using P35S‐ROMT, CYP1B1 and virB specific primers (Table S3). (a,c) Amplification products of P35S‐ROMT (a) and CYP1B1 (c) from transgenic lines of V. vinifera Monastrell and Gamay varieties. (b, d) Amplification products of virB from transgenic lines of V. vinifera Monastrell and Gamay varieties. WT. wild‐type callus negative control using nontransgenic genomic DNA of Vitis cells, 1–7 randomly selected transgenic callus. PC: positive controls using plasmidic DNA as template.

Figure 4

Expression of ROMT and CYP1B1 recombinant protein in cell cultures of Vitis vinifera Monastrell and Gamay varieties. (a) Expression of HA‐tag fusion proteins was confirmed by Western blot analysis with anti‐HA‐tag antibody. (b) Expression of human liver CYP1B1 protein was confirmed by Western blot analysis with CYP1B1 antibody. WT, wild‐type callus negative control using protein extract of nontransformed Vitis cells, 1–7 randomly selected transgenic callus.

Figure 5

Conversion of t‐R into t‐Pt by cell extract of Vv ROMT‐transformed grapevine cell cultures. HPLC‐ESI‐MS analysis of reaction products following incubation of recombinant HA‐tagged ROMT with 500 μm resveratrol in the presence of SAM (25 mm). (a) Standards. Reaction was carried out with 400 μg total soluble protein of either wild (b) or transgenic callus cv. Monastrell (c) and was allowed to proceed for 24 h. A chromatographic profile is shown, and identity of peaks was confirmed by HPLC‐ESI‐MS/MS analysis. Peak t‐Pc, trans‐piceid; Peak t‐R, trans‐resveratrol; Peak t‐Pt, trans‐pterostilbene.

Figure 6

Production of t‐Pt by elicited VvROMT‐transformed grapevine cell cultures. Stilbenoid HPLC‐ESI‐MS analysis of transformed Vitis cv. Monastrell and cv. Gamay cell extracts and extracellular media. Stilbenes content was determined in extracts of transgenic and wild‐type cell cultures at 168 h. Extracted ion chromatograms from pterostilbene (XIC 257) and resveratrol (XIC 229) are shown, and identity of peaks was confirmed by MS/MS the spectra (See Figure S1 and S2 for MS/MS spectrum of pterostilbene in Vitis cv. Monastrell and cv. Gamay cell extracts and extracellular medium). Peaks: t‐Pc, trans‐piceid; t‐R, trans‐resveratrol; t‐Pt, trans‐pterostilbene.

Figure 7

Production of t‐Pn by elicited HsCYP1B1‐transformed grapevine cell cultures. HPLC‐ESI‐MS analysis of Vitis cv. Monastrell and cv. Gamay cell extracts and extracellular media. Content of stilbenes was determined in extracts of transgenic and wild‐type cell cultures at 72 h. Extracted ion chromatograms from piceatannol (XIC 245) and resveratrol (XIC 229) are shown, and identity of peaks was confirmed by the MS/MS spectra (See Figure S3 and S4 for MS/MS spectrum of piceatannol in Vitis cv. Monastrell and cv. Gamay cell extracts and extracellular medium). Peaks: t‐Pc, trans‐piceid; t‐R, trans‐resveratrol; t‐Pn, trans‐piceatannol.

Figure 8

Time‐course accumulation of t‐Pn in elicited HsCYP1B1‐transformed grapevine cell lines. Total piceatannol production in Vitis cv. Monastrell (a), cv. Gamay (b) and extracellular accumulation of resveratrol and piceatannol in cv. Monastrell (c) cell culture elicited with 50 mm MBCD and 0.1 mm MeJA between 24 and 168 h of incubation. Twenty grams of cells was suspended in sterile fresh standard medium containing elicitors brought to a final 100 mL. Total piceatannol amounts are calculated per 100 mL of cell culture. In (a) and (b), black bars: intracellular content; white bars: extracellular content. In (c), each Monastrell cell line is represented with a different colour; solid lines: piceatannol, scaled to the left axis; dotted lines: resveratrol, scaled to the right axis; wild type: solid squares; transformed lines: solid circles. M. Monastrell transformed lines; G: Gamay transformed lines. Data are the mean of three independent biological replicates ± SD.

Figure 9

Schematic diagram of T‐DNA region of binary plasmids used for transformation experiments. P35S, CaMV 35S promoter; nptII, neomycin phosphotransferase under the control of the P35S promoter; GFP, green fluorescent protein; T35S, CaMV 35S terminator; STS, stilbene synthase; ROMT, resveratrol O‐methyltransferase; CYP1B1, Cytochrome P450; LB, left border; RB, right border. (a) pJCV52‐ROMT. (b) pEarleyGate‐STS. (c) pK7WG2D‐CYP1B1.