Chemical process to improve natural grapevine-cane extract effectivity against powdery mildew and grey mould

Grapevine canes are vine growing byproducts studied for their antimicrobial activities. These properties are directly connected to the stilbene content; oligomeric stilbenes being the most active. In this study, we propose a chemical process, based on oxidative coupling, using metals to increase the oligostilbene rate and the biological effectivity of cane extract against grapevine pathogens. A total of ten compounds were obtained and identified by combining LCMS and NMR spectroscopies, including four newly reported compounds: trans-oxistilbenin C, trans-oxistilbenin D, and cisand trans-oxistilbenin E. The extract and the main stilbene formed were evaluated for their preventive effects on Plasmopara viticola and Botrytis cinerea growth. The processed extract was highly effective against both pathogens.


INTRODUCTION
Stilbenes are a group of polyphenols found in different plant species, especially Vitis vinifera (grapevine), and playing the role of phytoalexins (Langcake & Pryce, 1977;Pezet et al., 2004;Rivière et al., 2012). Several studies have already established a link between the attacks of grapevine pathogens and the increase in stilbene production in the plant (Pezet et al., 2003;Vrhovsek et al., 2012). High antimicrobial activity of some stilbenes, especially stilbenes with oligomeric structure (dimers, trimers and tetramers) has been noted (Gabaston et al., 2017;Schnee et al., 2013). Today, there is a growing interest in grapevine stilbenes, which can be sourced from vine growing byproducts, like grapevine canes and roots, for their antimicrobial potential and their likely use for fighting grapevine pathogens like Plasmopara viticola and Botrytis cinerea (Adrian et al., 1997;Billet et al., 2019;Gabaston et al., 2017;Richard et al., 2016;Schnee et al., 2013). The highest concentrations of resveratrol oligomers are mainly found in extracts of grapevine roots, while grapevine cane extracts show higher concentrations of monomers and dimers, mainly resveratrol and ε-viniferin (Gabaston et al., 2017). However, grapevine canes can be found in larger quantities in comparison with grapevine roots, due to the annual pruning of the plant.
Several studies showed that oxidative coupling of resveratrol was possible using metals (Sako et al., 2004;Snyder et al., 2011;Velu et al., 2008) leading to the formation of effective compounds against grapevine pathogens (El Khawand et al., 2020a). These reactions could take place in different media including wine (El Khawand et al., 2020b). Hence, the hemisynthesis of resveratrol oligomers using resveratrol and ɛ-viniferin found in large amounts in cane extracts could be an efficient and fast way to produce an enriched stilbene extract, with high antimicrobial activity.
The original purpose of this study was to use oxidative coupling to produce a more active extract from a natural source rich in resveratrol and ε-viniferin. Oxidative coupling using silver acetate was directly applied on a grapevine-cane fraction with high content in resveratrol and ɛ-viniferin. Main compounds of the reaction mixture were isolated and identified by spectroscopic analysis (UHPLC-MS, 1D-and 2D-NMR) including new reported active resveratrol oligomers.
Biological activities were evaluated in vitro on downy mildew (Plasmopara viticola) and grey mould diseases (Botrytis cinerea), two of the most common grapevine diseases.

Cane extract and CPC fractionation
Grapevine cane extract was kindly provided by Actichem (Montauban, France) produced using a mixture of Cabernet Sauvignon and Merlot cultivars grown in the Bordeaux region (France). The stilbene composition of the cane extract was previously described (Müller et al., 2009;Romain et al., 2014). To obtain a purified extract with high stilbene content, fractionation was performed using centrifugal partition chromatography (CPC) according to a previously published protocol . Briefly, the CPC was realized on a Kromaton FCPC® 1000 apparatus (Angers, France) equipped with an Iota S100 pump, a Flash 06 DAD 600 detector and a Spider mixing unit manufactured by Ecom (Prague, Czech Republic). Fractionation was achieved using the K Arizona biphasic solvent system composed of n-heptane/ethyl acetate/MeOH/H 2 O (1/2/1/2, v/v). The fraction richest in resveratrol and εviniferin was collected. This purified extract contained mainly resveratrol, εviniferin and rviniferin (320, 570, and 60 mg/g, respectively) .

Oxidative coupling reaction
Based on our previously published protocol (El Khawand et al., 2020a), regioselective oxidative coupling reaction was conducted on 1 g of the purified extract. The quantity of the silver acetate (AgOAc) metallic catalyst added to perform the reaction was calculated according to the quantities of resveratrol and εviniferin in the extract, to afford a stilbene:AgOAc ratio of 1:1.5 The purified extract and AgOAc were dissolved in MeOH (100 %, 100 mL). The reaction mixture turned to an ash colour as it was heated at 50 °C and stirred for 2 hours. The reaction was then stopped by cooling at 4°C and centrifuged. The MeOH fraction was collected and the solvent was removed under reduced pressure. Finally, the grapevine cane processed extract (GCPE) was lyophilized.

Isolation and identification of the reaction products
Stilbenes were purified from GCPE extract using preparative HPLC. The apparatus was a Gilson PLC 2050 system (Middleton, WI, USA) equipped with a UV-VIS detector. Elution was conducted on a Phenomenex Kinetex XB-C18 column (21.2 mm × 250 mm, 5 µm). The reaction extract powder was solubilized at 50 mg/mL in MeOH-H 2 O (50/50; v/v). The extract was then eluted with a flow rate of 20 mL/min using non-acidified ultrapure water (solvent A) and acetonitrile (solvent B), according to the following gradient: 33 % B (0-2 min), 33-50 % B (2-25 min), 50-100 % B (25-26 min) and 100 % B (26-31 min). The eluate was monitored at 280 and 306 nm and the fractions were automatically collected. Syringes were purchased from Millipore (Molsheim, France). The solvents were removed under reduced pressure and fractions were lyophilized.
The pure compounds were analyzed using UHPLC as previously described and their purity was estimated to be ≥ 90 %. Exact masses were determined by infusion on a Thermo Fisher Exactive Plus Orbitrap mass spectrometer (Waltham, Massachusetts, USA) in negative mode with the following parameters: full scan mode, scan range 100-800 m/z, resolution 280,000, automatic gain control (AGC) target value of 2E5, maximum ion injection time (IT) of 30 ms; H-ESI source parameters: spray voltage 2700 V, sheath gas flow rate 8, capillary temperature 320 °C. The structures of purified compounds were finally determined by 1D-and 2D-experiments including COSY, HSQC, HMBC and ROESY, using a Bruker Avance III 600 NMR spectrometer (Rheinstetten, Germany).

Plasmopara viticola development inhibition
Grapevine leaves (Vitis vinifera L., Cabernet-Sauvignon cultivar) were collected from the upper part of the shoots from plants produced by wood-cutting propagation and grown in the INRA nursery (Villenave d'Ornon, France), under controlled conditions of temperature (25/20 °C, day/night) and humidity (75 %) with a photoperiod of 16/8 hours (light/dark). The P. viticola isolate ANN-01 used in the assay was collected in 2015 from Ugni Blanc grapevine cultivar leaves in a commercial vineyard (Charente, France). To obtain the required quantity of pathogen material, an inoculum (sporangia suspension) of P. viticola was taken from spores stored at -20°C, and grown on Cabernet Sauvignon leaves for 7 days, before proceeding to the development inhibition assay. This multiplication step was followed by the inhibition assay. Leaf disks were cut from Cabernet Sauvignon fresh leaves, treated with the stilbene solutions prepared in sterile water with 1 % of ethanol, and then by an aqueous suspension of 20,000 sporangia/mL as described in a previous study (Gabastonxet al., 2017). According to previous works, and by monitoring the development of the pathogen on the control leaf disks treated with water/ethanol (99/1, v/v), it should be noted that one percent of ethanol in water does not affect zoospore mobility and disease development (Pezet et al., 2004).

Botrytis cinerea mycelium development inhibition
This in vitro assay was similar to that used by Adrian et al. (1997). The culture medium was a mixture of potato dextrose broth (24 g/L) and agar (15 g/L), autoclaved at 120 °C for 20 min. Ethanol solutions of the evaluated compounds were prepared at the concentrations of 0.5, 1, 2, 5, 10, 20 and 50 mg/mL. The compound solutions were then incorporated in the autoclaved culture medium (240 µL/12 mL, solution/medium) to obtain final concentrations of 5, 10, 20, 50, 100, 200 and 500 mg/L with 2 % ethanol. Dishes of 6 wells were used and each mixture was poured in 3 wells, 4 mL/well. A calibrated plug of mycelium of B. cinerea was then inoculated on the culture medium, and the dishes were incubated at 22 °C, with a photoperiod of 15/9 hours (light/dark). Measurements of the mycelium growth area were performed after 48 h and 60 h of incubation, leading to the calculation of the inhibition rate according to each concentration of the tested compounds and extracts, and hence determining IC 50 for each tested compound or extract.

Statistical Analyses
Three independent experiments of eight repetitions were carried out for each assay. Data are shown as means ± SEM. The statistical analysis was performed on the three extracts at each specific concentration. One-way ANOVA with post-hoc Tukey HSD tests was carried out. Significant differences between each extract were set at **p < 0.01 and *p < 0.05.

Production and analysis of the grapevine cane processed extract
The purpose of this study is to directly apply an oxidative coupling strategy directly on a stilbene enriched extract obtained from grapevine canes to increase the extract antimicrobial activity. Since cane extract contains several different compounds with no more than 45% of stilbenes (Müller et al., 2009;Romain et al., 2014), the extract was firstly fractionated by centrifugal partition chromatography (CPC) following a procedure previously published . The fraction containing the main stilbenes of the cane extract was collected to proceed to the oxidative coupling using silver acetate (AgOAc). The UPLC-DAD-MS chromatogram of this fraction is presented in Figure 1A. The fraction contained three main stilbenes (see Figure S1 for chemical structure): trans-resveratrol, trans-ε-viniferin, and r-viniferin (320, 570, and 60 mg/g, respectively). In addition, minor compounds could be present such as miyabenol C, ωviniferin and δviniferin .  The UPLC-DAD-MS chromatogram of the grapevine-cane extract (GCPE) obtained after oxidative coupling by AgOAc is presented in Figure  1B. The lyophilization of the reaction mixture led to a dark brown soft powder. The analysis of the chromatogram indicates the formation of at least eleven compounds, and shows that under stirring condition, 100 % of resveratrol and ɛ-viniferin reacted. The r-viniferin probably did not react but increased due to the oligomerization of ɛviniferin (Sako et al., 2004). The total transformation of resveratrol and ɛ-viniferin was achieved using a molar ratio 1:1.5 (stilbene:AcOAg).
The structures of the identified compounds were obtained by mass spectrometry analysis and NMR spectroscopy, including 2D experiments such as COSY, ROESY, HSQC, and HMBC.

Antimicrobial activity against Plasmopara viticola
Grapevine extract effectiveness against downy mildew has been demonstrated in vitro (Gabaston et al., 2017;Schnee et al., 2013) in greenhouse and vineyard (Richard et al., 2016). In this study, grapevine leaves were treated with the grapevine-cane extract obtained after oxidative coupling by AgOAc (GCPE extract). The results were compared to the effects of the initial grapevine-cane extract ( Figure 3A). GCPE extract gave a total inhibition for concentrations upper to 200 mg/L, whereas the same effect was observed for the CE extract from 500 mg/L. The chemically engineered extract was three times more active than the initial extract against downy mildew (IC 50 63 ± 9 mg/L and 197 ± 12 mg/L, respectively). This process induced the production of a more active solution than the initial grapevine extract. This solution could be used to reduce the application rates by maximizing the effectiveness of treatments.
In addition, the antimicrobial activities of the isolated compounds were evaluated. The IC 50 of the main stilbenes isolated and identified were reported in Table 3. Resveratrol was used as standard and the resulting IC 50 value agreed with literature data (Gabaston et al., 2017;Schnee et al., 2013). The highest activity against P. viticola growth, and thus the lowest IC 50 values, was obtained for rviniferin (14 μM) followed by a pool of de novo synthetized compounds including δ-viniferin (40 μM), trans-and cis-oxistilbenin E (31 and 34 μM, respectively), and trans-oxistilbenin B (44 μM).
As previously reported (Gabaston et al., 2017;Schnee et al., 2013), we observed a positive correlation between the degree of oligomerization and the inhibition of the development of pathogens.

Antimicrobial activity against Botrytis cinerea
Fungitoxicity of extracts and pure compounds were monitored using radial growth test. B. cinerea growth areas were measured to calculate the inhibition rate according to each concentration of the tested compounds and extracts, and hence determining IC 50 . Firstly, Efficacy of the GCPE extract was compared to that of initial grapevine-cane extract ( Figure 3B). The grapevine cane extract exhibited a moderate effect against the development of B. cinerea. This finding is in agreement with the literature (Schnee et al., 2013). Cane extract is less effective against the grey mould agent. In contrast, the GCPE extract was more effective reaching 100 % inhibition at 100 mg/L and exhibiting an IC 50 of 5 ± 9 mg/L.
The IC 50 of the pure constituents of the GCPE extract are shown in Table 3. As previously reported (Gabaston et al., 2017), resveratrol presented a moderate antifungal activity (IC 50 430 μM). The rviniferin was the most active inhibitor of B. cinerea development, presenting the lowest IC 50 9 μM, followed by δ-viniferin (13 μM) and resviniferin A (27 μM). The oxidative coupling induced the production of an active pool of stilbene oligomers against B. cinerea development. The process could be useful to increase the efficacy of canes to prevent grapevine diseases. The GCPE extract could maximize the effectiveness of treatments; reduce the application rates and the number of interventions.

CONCLUSION
First, at all, we have developed an original approach to produce an active product from a natural vine extract. Findings in this study confirmed that oligomeric stilbenes have strong antimicrobial activity against grapevine pathogens. This study confirmed a positive correlation between the degree of oligomerization and the inhibition of pathogen development. The protocol induced the hemisynthesis of a pool of oligomeric stilbenes including r-viniferin, δ-viniferin, and the newly reported trans-oxistilbenin C, trans-oxistilbenin D, cis-and trans-oxistilbenin E. The initial grapevine extract was more active on P. viticola than on B. cinerea development. Its low efficiency against grey mould agent precludes its use in the field due to the large amount of canes necessary to produce this active product. The GCPE extract produced by oxidative coupling had the highest antimicrobial activity against P. viticola and B. cinerea. Interestingly, this extract strongly inhibited both grapevine pathogens. The use of this process could be a solution to reduce the doses necessary to treat downy mildew and to prevent the effects of gray mold in grapevine. To confirm the interest of this innovative approach, greenhouse and field trials will have to be carried out.