Synthetic Rhamnolipid Bolaforms trigger an innate immune response in Arabidopsis thaliana

Stimulation of plant innate immunity by natural and synthetic elicitors is a promising alternative to conventional pesticides for a more sustainable agriculture. Sugar-based bolaamphiphiles are known for their biocompatibility, biodegradability and low toxicity. In this work, we show that Synthetic Rhamnolipid Bolaforms (SRBs) that have been synthesized by green chemistry trigger Arabidopsis innate immunity. Using structure-function analysis, we demonstrate that SRBs, depending on the acyl chain length, differentially activate early and late immunity-related plant defense responses and provide local increase in resistance to plant pathogenic bacteria. Our biophysical data suggest that SRBs can interact with plant biomimetic plasma membrane and open the possibility of a lipid driven process for plant-triggered immunity by SRBs.

In the last few years, attention to amphiphilic molecules has increased because of their multiple applications in different areas including bioremediation, pharmacology, medical devices sanitization and agriculture [20][21][22][23][24] . Recent researches have been focused on D-xyloside-based and L-rhamnoside-based bolaamphiphiles surfactants characterized by their biocompatibility, biodegradability or low toxicity and targeted for the development of efficient and low cost lipid-based drug delivery systems [25][26][27][28][29] . The configuration of the bolaamphiphile surfactants consists of a long hydrophobic spacer connecting two hydrophilic entities; the molecules are more water soluble than the average surfactant and their properties make them extremely suitable for applications in nanotechnology, electronics, and gene and drug delivery 30 . Synthetic xylolipid bolaforms (SXBs) and rhamnolipid bolaforms (SRBs) containing a C 18 acyl chain have recently shown to interact with mammalian-based biomimetic systems of plasma membranes 31 .
In the present work, we investigated the potential of SRBs as new synthetic elicitors in plants. We show that SRBs are perceived by Arabidopsis and trigger an atypical plant immune response. We also show that the increase in resistance to the hemibiotrophic pathogen Pst depends on the fatty acid chain length of SRBs. Moreover, our results suggest that direct interaction of SRBs with the lipid fraction of plasma membrane could participate in their perception by plants.

Results
Synthesis of SRBs. The efficient synthesis of symmetric bolaamphiphiles derived from L-rhamnose SRBs ( Fig. 1A) has been realized using green chemistry principle (principle 5 i.e. "Safer Solvents and Auxiliaries", principle 7 i.e. "Use of Renewable Feedstocks" and principle 9 i.e. "Catalysis") 29,32 . Glycosidations were performed without solvent because the alcohols can play this role; excess of alcohol could be then removed and recycled for another reaction batch. Moreover, this method led to rhamnosides with a shorter reaction time and high yields (75-85%). Finally, the metathesis steps were performed in the presence of Grubbs I catalyst, without protecting steps of the OH functions, in methylene chloride alone without addition of methanol. The unsaturated SRBs could be then hydrogenated through classical Pd-catalyzed reactions and led to saturated SRBs surfactants (Fig. 1B). As lipid elicitors including lipopeptides and plant or microorganism-derived fatty acids have been shown to induce immune responses that depend on the chain length, unsaturation degree and position of the double bond in the fatty acid chain 8,[33][34][35][36] , we selected six SRBs with a fatty acid chain length of C 10 , C 14 or C 18 , saturated (SRB 10 , SRB 14 , SRB 18 , respectively) or unsaturated (SRB 10i , SRB 14i , SRB 18i , respectively) for the following experiments.
SRBs are perceived by Arabidopsis and display unconventional signalling-related immune responses. In order to investigate whether SRBs are perceived by plants and could induce an immune response, leaves or petioles from Arabidopsis were challenged with the different SRBs and monitored for production of extracellular ROS, a widely used marker of plant immunity 37 . Although ROS production gave similar profiles on both organs, the test was more robust and sensitive with petioles that were therefore used in the following ROS experiments ( Supplementary Fig. S1). All unsaturated SRBs induced a sustained ROS production when applied at 350 µM, the minimal concentration necessary to induce local plant protection using natural rhamnolipids 9 , suggesting that all these molecules could be perceived by Arabidopsis (Fig. 2B,D,F). For all molecules, the response lasted for several hours and tended to return to basal levels 12 hours after treatment. By contrast, the canonical elicitor flg22 induced a rapid and transient ROS production immediately after plant treatment in our experimental conditions ( Supplementary Fig. S2) 38 . SRB 10i and SRB 14i also stimulated a long lasting ROS burst at 100 µM in contrast to SRB 18i that was inactive at this concentration ( Fig. 2A,C,E). Moreover, SRB 14i induced the earliest and highest response when compared to other SRBs ( Fig. 2; Supplementary Fig. S3). Different levels of Figure 2. Extracellular ROS production upon elicitation of Arabidopsis with unsaturated SRBs. Petioles of 6-weeks-old wild type Arabidopsis plants were placed in a 96-wells plate and incubated in water overnight prior SRB elicitation. For ROS monitoring, a luminol-peroxidase solution containing 100 µM, or 350 µM of the corresponding SRB or solvent (0.5% ethanol) was added to each well. All SRBs contain the same amount of ethanol. The luminescence was read immediately after elicitation every 2 min with a Tecan Infinity F200 PRO for 720 min. Data presented are means of at least triplicate experiments ± standard error of the mean (SEM) with n = 6 for each experiment. (A,C,E) SRB 10i , SRB 14i and SRB 18i at 100 µM, respectively. (B,D,F) SRB 10i , SRB 14i and SRB 18i at 350 µM, respectively. (G) Control (ethanol).
Dose response experiments on SRB 14i showed that the minimal concentration necessary for this molecule to induce a robust ROS response in Arabidopsis was 50 µM (Fig. 4A). SRB 14i -triggered ROS production was fully dependent on the membrane bound NADPH oxidase RBOHD 39 as no ROS production could be detected in rbohD mutant plants (Fig. 4B). Receptor-like kinases (RLKs) are key components required for activation of the immune response following IP perception 40 . We monitored SRB 14i -mediated ROS production in the mutants bak1-5-bkk1-1 41 and bik1-pbl1 42 which are essential nodes involved in IP-triggered immunity. In addition, we used dorn1-1 mutant to investigate the sensing of extracellular ATP. ATP is among the molecules that are released by cell damage, and recent evidence suggests that ATP can serve as damage-associated molecular patterns (DAMPs) 43,44 . Compared to wild type (wt) plants, none of these mutants displayed a significant decrease in ROS production ( Fig. 5) after SRB 14i perception, suggesting that signal transduction following perception of the synthetic elicitor does not involve these RLKs.
MAP kinases are also important signalling components involved in immune related signalling 45 . More specifically, MAPK3 and MAPK6 phosphorylation is a key process related to immunity signalling upon perception of canonical IP elicitors 2 . None of the MAP kinases were activated after SRB 10i , SRB 14i or SRB 18i perception at 100 µM or 350 µM when compared to control 15, 60 or even 180 minutes post-treatment, the latest time point corresponding to the peak of ROS production (Fig. 6, Supplementary Fig. S4).

SRBs differentially induce immunity-related gene markers and electrolyte leakage in Arabidopsis.
To assess the capacity of SRBs to induce transcriptomic changes related to Arabidopsis immunity, we monitored the expression of PDF1.2, NPR1 and CYP71A12 genes. PDF1.2 is a well-known plant defensin gene activated concomitantly by the JA and ET pathways 46,47 while NPR1 is an important regulator of PR proteins linked to the SA pathway 48,49 . CYP71A12 encodes a cytochrome P450, which catalyzes the conversion of indole-3-acetaldoxime to indole-3-acetonitrile during biosynthesis of the phytoalexin camalexin 50 . RT-qPCR was performed 9 hours after elicitation on seedlings elicited with SRB 10i , SRB 14i or SRB 18i . All SRBs were able to significantly induce CYP71A12 whereas only SRB 10i and SRB 14i stimulated PDF1.2 expression (Fig. 7A,B). Interestingly, the SA dependent NPR1 gene was only activated after SRB 14i challenge (Fig. 7C).   We also performed conductivity measurements on Arabidopsis leaf discs to assess if SRBs induced changes in plant plasma membrane permeability ( Fig. 8). At 350 µM, SRB 14i triggered a strong electrolyte leakage in Arabidopsis leaves within 24 hours post-treatment. SRB 10i and SRB 18i treatment also increased medium conductivity, SRB 18i being the most active, but at a lower extent than SRB 14i (Fig. 8). At 100 µM, only SRB 18i provoked a slight electrolyte leakage response. Despite the electrolyte leakage responses, we did not observe significant phytotoxicity effects (chlorosis or necrotic spots) on Arabidopsis plants infiltrated with SRBs even at the highest concentration ( Supplementary Fig. S5). SRB 14i triggers local increase in resistance to Pst but not to Botrytis cinerea. In order to investigate whether SRBs are able to trigger local induced resistance to pathogens, we performed protection experiments using two different lifestyle microorganisms, the hemibiotrophic bacteria Pst and the necrotrophic fungus B. cinerea 51,52 . For the protection test against Pst, Arabidopsis plants were sprayed with SRBs and spray-inoculated with Pst two days later. Three days after inoculation leaf samples were collected for Pst CFU titration. SRB 14i but not SRB 10i or SRB 18i , induced significant disease reduction against Pst (Fig. 9A). SRB i did not show any bacteriostatic or bactericidal effect on Pst when applied at 350 µM ( Supplementary Fig. S6), suggesting that the increased resistance was mediated by plant defense responses. To monitor B. cinerea infection, leaf discs pre-treated with SRBs were inoculated two days later with a conidia suspension (10 5 conidia/mL) and four days after inoculation, the necrotized leaf area was measured. As shown in Fig. 9B, none of the SRBs induced plant protection against the necrotrophic fungus.
To investigate the hormone signalling pathways involved in SRB 14i -driven increase in resistance to Pst, we performed protection experiments in sid2 53,54 and jar1 55 mutants plants impaired in SA and JA signalling, respectively 9 . Enhance protection to Pst observed in wt plants was lost in the sid2 mutant, demonstrating that the SA signalling pathway is involved in SRB 14i -triggered immunity to the pathogen. JA signalling pathway is however not involved in the process as the increase in resistance to Pst was conserved in jar1 mutant (Fig. 9C). SRB 14i interacts with plant biomimetic plasma membrane. The atypical immune signature of SRBs raised the hypothesis of a direct interaction of the molecules with lipids from the plasma membrane instead of a receptor-dependent perception, as it has already been proposed for the amphiphilic bacterial elicitor surfactin 33 . In order to assess the affinity of SRB 14i for plasma membrane lipids, its partitioning into PLPC/sitosterol (80: 20) vesicles and its insertion into PLPC or sitosterol monolayers were determined by ITC and Langmuir monolayer experiments. ITC raw data (Fig. 10A) displayed a gradual decrease of the positive heat flow signal over the course of the successive LUV injections. This profile is typical of a binding phenomenon. The thermodynamic parameters obtained from the fitting of the cumulative heat vs lipid concentration plot (Fig. 10B) indicated that the binding reaction of SRB 14i with vesicles is spontaneous (ΔG < 0), endothermic (ΔH > 0) and leads to a positive change of the entropy (ΔS > 0) (Fig. 10B). The absolute value of entropy change is much higher than the absolute values of enthalpy change indicating that the binding is notably driven by hydrophobic interactions 56 . ITC results showed that SRB 14i could bind to liposomes and thus suggest an interaction of SRB 14i with the lipid phase of the plant plasma membrane. As observed in Fig. 10C, SRB 14i was preferably inserted into PLPC than into sitosterol monolayers (MIP PLPC > MIP sitosterol ) but both systems are compatible with the SRB 14i membrane insertion. In addition, SRB 14i also induced 20-30% vesicle permeabilization, when applied at 10 µM (Fig. 10D), suggesting a transient perturbation of the bilayer.

Discussion
In this study, we show for the first time that the synthetic bolaamphiphilic glycolipids SRBs are perceived by Arabidopsis cells and induce an immune response characterized by unconventional signalling events, defence gene activation and enhanced resistance to the hemibiotrophic pathogen Pst. In the last years, the advancement in combinatorial and organic chemistry has boosted the interest on synthetic compounds with plant eliciting capacities since they have been demonstrated to be effective in control of plant pathogens and to induce plant immunity as efficiently as natural elicitors 10 . A recent publication reported a large screening identifying 114 synthetic elicitors that activate expression of the pathogen-responsive CaBP22−333::GUS reporter gene in Arabidopsis; 33 of which being [(phenylimino)methyl]phenol (PMP) derivatives or PMP-related compounds 17 . Their chemical preparation is relatively difficult and costly due to the starting materials and the synthetic procedures involving the use of environmental unfriendly compounds and solvents. As previously described, the synthesis of SRBs follows the principles of green chemistry and uses biosourced materials. Interestingly, most of the synthetic elicitors described so far are aromatic compounds exhibiting at least a benzene ring 10 . All these compounds induce disease resistance to pathogens. However, their mode of action can strongly differ depending on the structures of the molecules. For instance, DCA transiently induces defence reactions to Hpa and Pst and DCA-induced resistance to Hpa is partially dependent on the NPR1 pathway 16 . This is in contrast with SA analogues like INA or BTH that fully involve NPR1 to induce a broad range resistance to biotrophs 12,57,58 . BHTC induces Arabidopsis disease resistance against bacterial, oomycete, and fungal pathogens. BHTC-triggered protection to Hpa is independent from SA pathway but requires NPR1 15 . DPMP mode of action is distinct from that of DCA and similar to BHTC, since its ability to induce immunity against Hpa is completely blocked in npr1 mutant plants 17 . Sulfonamides like sulfamethoxazole are potent inducers of plant immunity against Pst that do not require NPR1-dependent canonical SA defence pathway 59,60 . Apart from aromatic-derived elicitors, a synthetic cationic lipid diC 14 with amphiphilic properties, and therefore more related to SRBs, has recently been described as a potent elicitor 19 . Like SRB 14i , diC 14 enhances plant protection to Pst but not to B. cinerea and induced resistance to Pst is SA-dependent but JA-independent. Moreover, the amidine headgroup and chain length were important for its activity.
Interestingly, all the synthetic elicitors characterized so far are acting at relatively high concentrations (generally from 100 µM to 1 mM) compared to canonical MAMP/IPs like flagellin, active at micromolar or even nanomolar ranges (supplementary Fig. S2) 5,10 . This is also the case for SRBs, inducing defence reactions like ROS production at 50-100 µM. This could be explained by a hormone-like mode of action or a different perception mechanism not involving plant receptors. To our knowledge, little data exist on the signalling events involved in synthetic elicitor-triggered immunity. Here, we show that SRB 14i induces early and sustained ROS production. ROS production is a characteristic immune-related signalling event following IP perception 37 . In Arabidopsis, ROS production during the immune response is mainly achieved by the NADPH oxidase RBOHD 39 . However, a recent report on LPS perception also suggested that long lasting ROS production could originate from chloroplasts 61 . We found that sustained ROS production triggered by unsaturated SRBs was fully dependent on RBOHD demonstrating that a long lasting oxidative burst could also be generated by the membrane bound NADPH oxidase. These results also demonstrate that SRB-driven ROS accumulation is related to an active process and not a direct result of potential release after cell damages. MAPK3 and 6 are usually involved in MAMP-triggered immunity 2 . Interestingly, although ROS production was triggered by SRBs, we did not observed activation of these canonic MAPKs in our conditions.
Regarding the cell perception of SRBs, we have demonstrated that the chemical structure of SRBs, in particular the length and saturation of the acyl chain, is important for the plant perception and induction of plant defences. We observed that the most active SRB possessed a C 14 acyl chain with an unsaturation (SRB 14i ). When the SRB 14i was assessed in ITC experiments, we observed that it effectively bounds to the model membrane via hydrophobic interactions. The importance of the saturation and chain length has been proposed for arachidonic acid and derivatives in potato tubers elicitation 34 . Acyl chain of 20 carbons resulted in effective elicitation by contrast to C 16 , C 18 or C 22 , which were inactive or showed a very low eliciting activity. By contrast, synthetic ultrashort cationic lipopeptides were more active with chains containing 16 carbons 18 . Cambiagno et al. noticed that diC 14 was more efficient in inducing plant resistance than diC 16 19 . Lipopeptides such as orfamide or surfactin, both amphiphilic elicitors, display an immune response that depends on the concentration and/or the length of the fatty acid chain 8,33,62,63 . Surfactin with 14 or 15 carbons were more efficient than the C 12 and C 13 counterparts 8,33 . Orfamide Plasma membrane perturbation and lipid signalling play an important role in the adaptation to biotic and abiotic challenges, especially by modifying the compartmentalization, distribution, abundance and type of lipids and proteins present in the membrane upon stress perception 65,66 . Our results on ITC experiments suggest that SRB 14i has the ability to insert into plasma membrane and this insertion could result in plasma membrane reorganization. This membrane perturbation attested by the biophysical permeabilization experiments could in turn result in electrolyte leakage and plant defense activation. Savchenko and colleagues 35 demonstrated that even minor modifications of the plant cell lipid composition with exogenous fatty acids derived from animals or pathogens induced plant defence activation. Some MAMPs have also been reported to alter the plasma membrane state. Tobacco and Arabidopsis cells treated with flg22, oligogalacturonides or cryptogein induced an increase of lipid self-association in specific domains of the plasma membrane upon elicitation coinciding with the onset of plant defence signalling 67,68 . The same reports state that cryptogein enhances ROS production by recruiting plant sterols from the plasma membrane increasing its fluidity. Similarly, it has been observed that surfactin favourably interacts with phospholipids, suggesting that surfactin insertion into the plasma membrane participates in the elicitation mechanism 33 . Synthetic rhamnolipids with modified carboxylated fatty acid chains were recently shown to differentially induce ROS production 69 . Alkylated rhamnolipids induced a higher response than carboxylated ones. Surprisingly, alkylated rhamnolipids were more favourably inserted into model membrane suggesting that the differences in the biological activity could be derived from a stronger interaction of alkylated rhamnolipids with the domain boundary regions within the membrane where the elicitor is inserted 69 . It is known that the effect of fengycin, orfamide or surfactin lipopeptides on the cell plasma membrane varies from transient permeabilization to solubilisation [70][71][72] . We observed that SRBs induced changes in plant plasma membrane permeability depending on the concentration and the chain length. This process is unlikely to lead to strong membrane damages, as we did not observe phytotoxic effects on the plants even at high concentration over a long period of time (several days). Interestingly, the laminarin sulfate elicitor PS3 also induced electrolyte leakage in tobacco leaves without inducing cell death and was proposed to interact with plasma membranes 73 . Yeast elicitors consisting in a mixture of glucan, mannan, and chitin, were shown to exhibit pore-forming properties resulting in ion fluxes 74 . This pore-forming activity could explain the electrolyte leakage caused by SRBs.
In the present work, we show that the synthetic elicitor SRB 14i is able to induce early immune responses, defence gene activation and enhanced plant protection against Pst that dependent on the SA pathway. This fact opens the door to a wide array of new synthetic elicitors by generating amphiphilic compounds that could participate in broadening the variety of crop protection products available in the agriculture market but also to better understand the molecular mechanisms involved in plant immunity. Moreover, when compared to the difficulty and high cost that purification of microbial elicitors involves, chemical synthesis proved to be easy to scale-up at industrial level with affordable costs and provides the additional advantage of making possible structural modification, in order to adapt the final product to new uses.
General procedure for the preparation of the rhamnoside based bolaamphiphiles by metathesis. The rhamnoside (10 mmol, 1 eq) was diluted in CH 2 Cl 2 (40 mL) in a Schlenk tube under argon and the Grubbs' I catalyst (823 mg, 1 mmol, 0.1 eq) was added in three portions over 3 h. After 8 h of reaction at 45 °C, the solvent is evaporated under reduced pressure and the residue is purified using flash chromatography (eluting mixture: CH 2 Cl 2 /MeOH 9:1). Bolaamphiphiles were obtained with good yields (65-75%).
General procedure for palladium catalyzed hydrogenation. SRBs (10 mmol) were dissolved in 25 mL of ethanol under Ar atmosphere. After 10 min, 80 mg of palladium on activated charcoal (Pd/C, 10% w/w) were added and the solution was stirred another 10 min under Ar atmosphere before being submitted to H 2 flow until completion (24 hours at room temperature). Once the reaction was completed, the reaction mixture was filtered through Celite. The obtained solution was then evaporated under reduced pressure. Saturated SRBs were obtained with a quantitative yield.
Plant material and elicitation. Arabidopsis thaliana ecotype Col-0 was used as WT parent for all experiments. Seeds from rbohD, bak1-5-bkk1-1, bik1-pbl1 were provided by C. Zipfel. dorn1-1 seeds were obtained from NASC stock (SALK_042209). All mutants, including sid2 and jar1 9 are in the Col-0 background. Arabidopsis were grown in soil (Gramoflor, Germany) at 21 °C with 60% relative humidity and a 12 h light/12 h dark cycle (light intensity 150 µM/m 2 .s). For RT-qPCR analysis, Arabidopsis seedlings were germinated in solid Murashige an Skoog (MS) basal medium with vitamins (pH 5.7) and transferred to 12-well sterile plate containing 1 mL of liquid MS (5 seedling per well) 4 days later. When seedlings were 10-days-old, medium was changed with 1 mL of fresh medium prior elicitation. Purified SRBs were dissolved in 100% ethanol and were used at the concentrations mentioned in the text (from 10 to 300 µM). Final concentration of ethanol for the experiments did not exceed 0.5%, a concentration that was not stressful under our experimental conditions. Control plants were treated with the corresponding final ethanol concentration of the SRB solutions. For infiltration assays, four weeks-old Arabidopsis were syringe infiltrated with corresponding elicitor concentration or ethanol for control. Photographs were taken with a canon Powershot G12 camera 48 h after infiltration.
Protection assays. For the protection tests against B. cinerea (strain B05.10) 9 , 6-weeks-old plants were leaf spray-elicited with 350 µM SRB solution. Two days after elicitation, 7 mm leaf discs were cut and placed in Petri dishes onto water-wet Whatman filter paper. B. cinerea was initiated from a silica gel crystal stock and cultivated in PDA plates at 22 °C for 2 weeks. Conidia were collected by adding 4 mL of conidia suspension (KH 2 PO 4 1.75 g/L, MgSO 4 0.75 g/L, Glc 4 g/L, peptone 4 g/L, Tween 20 0.02% [v/v]) and filtered with cheesecloth to separate hyphae from conidia. The conidia suspension was adjusted to 10 5 conidia/mL, incubated for 9 h at 22 °C for germination and Arabidopsis leaf discs were inoculated with one drop of the conidia suspension. Petri dishes containing the inoculated leaf discs were parafilm-sealed and incubated at 22 °C for 4 days before quantifying the necrotic area with the ImageJ software 75 .
For protection assays against Pseudomonas syringae pv. tomato DC3000 (Pst) 76 , 15 Arabidopsis seeds were sawn in order to form a clump. When the clump was 3 to 5-weeks-old, plants were spray-elicited 2 days before pathogen inoculation. Pst was cultured overnight at 28 °C in liquid King's B medium, supplemented with rifampicin (50 µg/mL) and kanamycin (50 µg/mL). Subsequently, bacterial cells were collected by centrifugation at 3000 g for 5 min and resuspended in 10 mM MgCl 2 -Silwet L-77 0.025% to a final optical density of 0.01 (OD 600nm ). For plant inoculation, 3 mL of the bacterial solutions were sprayed on plant leaves (Arabidopsis clumps). Mock plants were treated with the same solution free of bacteria. Inoculated plants were incubated in sealed boxes under the same conditions of Arabidopsis culture. Three days after inoculation, clump leaves were harvested, grinded in 10 mL of 10 mM MgCl 2 solution and different dilutions were used to determine the CFU/mg fresh weight on solid KB medium dishes supplemented with antibiotics.
Bacteriostatic and bactericide assay. After growth in King's B medium, Pst cells were collected by centrifugation at 3000 g for 5 min and resuspended to a final optical density of 0.01 (OD 600nm ). Then corresponding elicitor or ethanol was added to this solution, which was subsequently distributed in 96 wells plate. OD 600nm was monitored every hour during 48 h with a TECAN F200 pro. The bactericide effect of elicitor was monitored by counting CFU 24 h post-treatment. To this end, King's B medium containing Pst supplemented with elicitor or ethanol was serial diluted in 10 mM MgCl 2 and CFU/mL was counted on solid KB medium dishes supplemented with antibiotics. Extracellular ROS production. Five mm leaf disks or petioles of 6-weeks-old Arabidopsis plants were incubated overnight in 96-wells plates containing 150 µL of distilled water at room temperature. Samples were then rinsed with distilled water and challenged in darkness with a luminol-horseradish peroxidase solution 65 containing SRBs at the concentrations mentioned in the figure's legends. ROS production was monitored with a Tecan Infinite F200 PRO (Tecan) every 2 min. For each time point, measurement is integrated over a 2 s period and results were expressed in Relative Light Units (RLUs).
Electrolyte leakage. The assay was performed as previously described 65 with slight modification. Eight leaf discs of 6-mm-diameter were incubated in distilled water overnight. One disc was transferred into 1.5 mL tube containing fresh distilled water and the corresponding elicitor concentration or ethanol for the control. Conductivity measurements (three to four replicates for each treatment) were then conducted over time using a B-771 LaquaTwin (Horiba) conductivity meter.

Isothermal titration calorimetry (ITC).
Isothermal titration calorimetry measurements were performed on a VP-ITC Microcalorimeter (Microcal). SRB 14i titrations were carried out by injecting 10 µL aliquots of large unilamellar vesicles (LUVs) made with PLPC/sitosterol (80/20-5 mM) into the calorimeter cell (V cell = 1.4565 mL) containing SRB 14i at 100 µM at constant time intervals of 5 min and at 26 °C. The solution in the sample cell was stirred at a speed of 305 rpm. The reference cell was filled with milliQ water. Prior each analysis, all solutions were degassed using sonicator bath. LUVs dispersion and SRB 14i solution were prepared in Tris 10 mM buffer at pH 7.4. Initially, SRB 14i stock solution (100 mM) was prepared in DMSO. A 1000× dilution of the SRB 14i stock solution was performed in the buffer. An appropriate amount of DMSO was added in the LUV dispersion in order to avoid artifact due to the presence of DMSO.
Adsorption into a lipid monolayer. An automated LB system (KSV minitrough, KSV instruments Ltd. 75 × 160mm 2 -volume of 80 cm³) equipped with a home-made injection setup was used as described previously 77 . During the entire duration of the experiment, the subphase was stirred using two cylindrical micromagnetic rods (8 × 1.5 mm 2 ) and two electronic stirrer heads located beneath the trough (model 300, Rank Brothers). An autoreversing mode with slow acceleration and a stirring speed of 100 rpm was selected. PLPC or sitosterol monolayers were prepared at the interface of the Langmuir trough filled with 10 mM Tris buffer at pH 7.4 and 22 ± 1 °C. The defined initial surface pressure of these monolayers was obtained by spreading a precise volume of PLPC or sitosterol solutions prepared in chloroform/methanol (2:1 v/v). As soon as the initial surface pressure was stabilized (∼20 min), SRB 14i solubilized in DMSO was injected into the subphase to a final concentration of 4.5 µM. After the injection of SRB 14i , the increase in surface pressure was recorded. Pure DMSO injections into the subphase did not modify the initial surface pressure of the lipid monolayers.
Permeability assays on liposomes. Membrane permeabilization was followed as described previously 78 . Release of entrapped calcein at self-quenching concentrations from LUV composed by PLPC/sitosterol (80/20-16 µM) can be monitored by the fluorescence increase upon dilution following their leakage from the vesicles. LUVs dispersion was prepared in Tris 10 mM buffer at pH 7.4 as previously described 79 .
SRB 14i was added from a stock solution in DMSO and fluorescence intensities were immediately recorded. The percentage of calcein released was defined as [(Ft-Fcontr)/(Ftot-Fcontr)]/100, where Ft is the fluorescence signal measured after 15 min in the presence of SRB 14i , Fcontr is the fluorescence signal measured at the same time for control liposomes, and Ftot is the total fluorescence signal obtained after complete disruption of the liposomes by 0.1% Triton X-100. All fluorescence determinations were performed at room temperature on a Perkin Elmer LS-50B Fluorescence Spectrophotometer (Perkin-Elmer) using λ exc of 450 nm and a λ em of 512 nm. Data Availability. The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.