Inducing Plant Defense Reactions in Tobacco Plants with Phenolic-Rich Extracts from Red Maple Leaves: A Characterization of Main Active Ingredients

Red maple leaf extracts (RME) were tested for their plant defence inducer (PDI) properties. Two extracts were obtained and compared by different approaches: RME1 using ethanol- 29 water (30-70%, v/v, 0.5% HCl 1N) and RME2 using pure water. Both extracts titrated at 1.9 30 g/L in polyphenols and infiltrated into tobacco leaves efficiently induced hypersensitive 31 reaction-like lesions and topical accumulation of auto-fluorescent compounds noted under UV 32 and scopoletin titration assays. The antimicrobial marker PR1 ,  -1,3-glucanase PR2 , chitinase 33 PR3, and osmotin PR5 target genes were all upregulated in tobacco leaves following RME1 34 treatment. The alkaline hydrolysis of RME1 and RME2 combined with HPLC titration of 35 gallic acid revealed that gallate functions were present in both extracts at levels comprised 36 between 185 and 318 mg.L -1 . HPLC-HR-MS analyses and glucose assay identified four 37 gallate derivatives consisting of a glucose core linked to 5, 6, 7 and 8 gallate groups. These 38 four galloyl glucoses possessed around 46% of total gallate functions. Their higher 39 concentration in RME suggested that they may contribute significantly to PDI activity. These 40 findings define the friendly galloyl glucose as a PDI and highlight a relevant methodology for 41 combining plant assays and chemistry process to their potential quantification in crude natural 42


Introduction 44
In the context of sustainable development, agriculture is incorporating more eco-45 friendly alternatives to limit the use of chemical pesticides and regulate pest management. 46 Increasing the natural resistance of plants is one favoured line of research, notably using 47 biological substances that can stimulate plant immunity [1,2]. A complex array of immune 48 response is triggered as early as plant detect pests [3,4]. The detection of pathogen-or plant-49 derived elicitors lead to the activation of numerous biochemical and molecular events in plant 50 cells which prevent pathogen development [5,6]. The reactive oxygen species (ROS) 51 production causes a hypersensitive reaction (HR) leading to topical cell death that restrict the 52 systemic spread of the pathogen [7,8]. Surrounding tissues will acquire local resistance 53 (named LAR) thanks to phytoalexin biosynthesis, cell wall and/or cuticle reinforcement with 54 phenylpropanoid compounds, callose deposition, defence enzymes and pathogenesis-related 55 (PR) proteins synthesis [9,10]. The whole plant will be mobilized with the systemic acquired 56 resistance (SAR) undertaken by salicylic acid which allows uninfected distal parts of the plant 57 to respond more effectively to subsequent infection [11,12]. 58 The non-host resistance strategy involved therefore the local and systemic production of 59 defence compounds with antimicrobial properties to counter pathogen development. Phenolic 60 compounds are plant secondary metabolites preformed (named phytoanticipins) or induced in 61 the plant after biotic attacks (named phytoalexins) and constitute inbuilt antibiotic chemical 62 barriers to a wide range of potential pests and pathogens [13][14][15][16]. Our group developed the 63 biotechnology concept consisting of extracting polyphenols (PPs) from biomass and 64 reapplying them to plants to intentionally protect them against pathogens. This way, we 65 showed that plant PP-rich extracts could trigger their own plant defence reactions. In 66 particular, the grape marc extracts enriched in PPs were first demonstrated as playing the role 67 of plant defence inducer (PDI) in tobacco [17][18][19][20]. Later on, we evidenced the elicitation 68 Purified RNAs were quantified by NanoDrop™ 2000 spectrophotometer (Thermo Fisher 140 Scientific) and the RNA concentration was measured using the Agilent 2200 Tape Station and 141 the RNA ScreenTape kit (Agilent Technologies). First-strand cDNA was synthetized from 1 142 g of total RNA with Euroscript Reverse Transcriptase (Eurogentec, France) according to the 143 manufacturer's instructions. PCR reactions were prepared using the qPCR kit manufacturer's 144 protocol. The cDNA concentration used produced a threshold value (C T ) of between 15 and 145 30 cycles. Amplification specificity was checked by melting-curve analysis. The relative 146 quantity (Q R ) of PR gene transcripts using EF-1 gene as internal standard was calculated 147 with the  mathematical model. QPCR data were expressed as the threshold cycle (Ct) 148 values normalized to EF-1 and calculated using the 2 −ΔΔCt method following standard 149 protocols [35]. For every PR gene analyzed, three independent biological replicates were run, 150 and every run was carried out at least in triplicate. Primers and amplicon sizes were given in 151 UV-vis spectra were recorded using a Varian Cary 3 spectrophotometer in a 1-cm quartz cell. 155 Analysis of RME1 and RME2 were performed with liquid chromatography (Alliance Waters 156 HPLC) using a Waters 2695 separation module and a Waters 2998 photodiode array detector. 157 HPLC-UV separation was conducted using a Phenomenex reversed phase column C 18 grafted 158 silica, (100 mm length, 2.1 mm i.d. 1.7 μm particle size) and a binary solvent system 159 composed of acetonitrile (solvent A) and water containing 0.1% orthophosphoric acid 160 (solvent B) at a flow rate of 0.2 ml min -1 . The initial composition 90% A and 10% B was 161 maintained for 4 min, then solvent B was linearly increased to 25% in 4 min, and to 40% in 162 22 min, to finish at 95% in 5 min. The identification of active constituents was performed 163 (Thermoscientific) and an ultra-high-performance liquid chromatography (UPLC) instrument, 165 the Ultimate 3000 RSLC (Thermoscientific). Analyses were carried out in both negative and 166 positive electrospray modes (ESI + and ESI -). UPLC separations were performed using the 167 same column and elution gradient as previously indicated. Identification of compounds was 168 based on structural elucidation of mass spectra and the use of accurate mass determination 169 was obtained with Orbitrap high resolution. MS-MS was done by the HCD technique (35 eV). Glucose measurements were recorded for both RME and h-RME (600 µL) after pH 185 readjustment to 7.8 as water negative control. The assay was calibrated with a set of glucose 186 concentrations. GOD-POD reagent (4 mL) was added to each sample, mixed by pipetting and 187 spectrophotometer. Glucose concentration was calculated using the calibration curve. We further investigated the RME1 ability to induce phytoalexin production and 230 defence-related gene expression. We monitored the formation of scopoletin, a phytoalexin 231 known to be involved in the activation of defence mechanism. The quantification of 232 scopoletin by HPLC reveal an over-accumulation in RME1-infiltrated tobacco leaves 233 reaching 307138 ng scopoletin/gFW. This was significantly higher at 3.5-fold (p-value < 234 0,001) than for the control leaves ( Figure 2). Control leaves were infiltrated with acidic water 235 and remained symptomless (data not shown). Additionally, RME1 did not show any natural 236 auto-fluorescence ( Figure 1B). were active in inducing plant defence reactions [19]. Based on these data, we focused on PPs 256 to further characterize the active ingredients responsible for these properties. 257

3.2.HPLC-UV fingerprints and UPLC-HR-MS analysis of RME1 and RME2 266
In order to identify the chemical compounds responsible for the PDI properties, we 267  Figure SI-1). 281 The five main components detected eluted after 21 min and were labelled G5-G8 (Figure 4). SI-4 and SI-5 C). In comparison with G5, compounds G6, G7 and G8 are likely hexa, hepta 307 and octagalloyl glucose derivatives, respectively. As glucose contains only 5 OH functions 308 and can only be linked to five gallic acids, the other gallic groups are evidently linked to OH 309

3.3.Quantification of gallate functions by alkaline hydrolysis 315
As the comparative HPLC analyses of water-and hydroalcoholic-RME revealed that 316 the organic solvent offered more extractable gallate derivatives and RME1 was more potent 317 than RME2 in the induction of HR-like reactions, we predicted that gallate derivatives were 318 involved in PDI activity. To titrate the gallate functions, we conducted alkaline hydrolysis of 319 RME1 and RME2 in order to convert gallate functions in gallic acid and ensure they were

3.4.Quantification of gallotanins in RME1 334
The amount of the gallotanin G5 (five gallate moieties linked to a glucose sugar) was 336 determined using the commercial pentagalloyl glucose as a reference. This was equal to 37.9 337 mg.L -1 in RME1 and to 12 mg.L -1 in RME2 at 0.19% in PPs. In G5-G8, the absorbing 338 where M gallic acid is the molecular mass of gallic acid. 357 basic hydrolysis of RME1. In the case of RME2, we found 85 mg.L -1 of gallate from the 359 same calculation, i.e. also to 46% of total gallate functions. 360 361

3.5.Suppression of topical symptoms induced by alkaline hydrolysed RME1 362
To investigate the involvement of gallotanins in RME1-PDI activity, we looked at the 363 comparative deployment of macroscopic symptoms on tobacco leaves at 4 dpi after 364 infiltration of RME1 before and after hydrolysis occurred (RME1 and h-RME1, respectively). 365 Tobacco leaves showed different levels of sensitivity to RME1 and h-RME1 ( Figure 5 A-D). 366 The h-RME1 provoked large and marked necrotic symptoms when infiltrated at the 0.19% 367 PPs and 4-and 8-fold diluted h-RME1-PP concentrations. No distinct chlorotic zones were 368 observed for lower h-RME1-PP concentrations ( Figure 5B). The h-RME1 also failed to 369 produce auto-fluorescent compounds within surrounding necrotic zones regardless of the h-370 RME1-PP concentrations ( Figure 5D). These data clearly show that h-RME did not display 371 PDI activity. We ascertain the symptomless action of gallic acid produced as a result of RME 372 hydrolysis ( Figure SI-6) and suggest that necrotic tissues observed after h-RME1 infiltration 373 should be the result of toxicity symptoms induced by the h-RME cocktail of molecules. 374 in tobacco leaves. Figure 2 shows the ratio of fluorescent scopoletin production in leaves 384 induced at 4 dpi in response to RME1 versus control (acidified water) and h-RME1 385 infiltrations. Since fluorescence never appeared within dead tissues, the experiment was 386 conducted with the 2-fold diluted RME1-PP concentration that induced restricted necrotic 387 zones. The h-RME1 infiltrated leaves produced 10551ng scopoletin/gFW that was 2.9 fold 388 lower than for the RME1-infiltrated conditions. The amount of scopoletin produced in 389 tobacco leaves after h-RME1 infiltration was similar to the amount produced in the control 390 leaves. These data clearly evidenced that h-RME1 was not able to induce local plant defence 391 reactions in tobacco leaves meaning that alkaline hydrolysis which suppress gallate functions 392 suppress PDI activity as well. 393 394

3.6.PDI activity of pentagalloyl glucose 395
The ability of the gallotanins to induce HR-like reactions was tested on tobacco leaves. 396 Since pentagalloyl glucose (G5) was the main RME1-gallate derivative and is readily 397 available commercially, it was infiltrated into tobacco leaves in the range 148 mg.L -1 -9.25 398 mg.L -1 , with the highest concentration corresponding to the amount of G5+G6+G6'+G7+G8 399 found in RME1. Figure

Conclusions 449
The paper describes an original, strong and reliable chemical methodology to detect 450 the galloyl-active ingredients from a complex mixture of biomolecules. Discovered here as 451 bioactive ingredients in RME and easily quantifiable by chemical methodology, these natural 452 molecules could offer a tremendous tool to screen plant or crude by-products extracts with 453 potential PDI activity. Future investigations will define the most suitable and abundant 454 galloyl bioproducts and the optimum efficiency for controlling the incidence of diseases in