TYPE Original Research PUBLISHED 16 December 2022 DOI 10.3389/fpls.2022.1075231 Grapevine response to a OPEN ACCESS Dittrichia viscosa extract and EDITED BY Maria Paz Diago, University of La Rioja, Spain a Bacillus velezensis strain REVIEWED BY Xiaochi Ma, Mélina Ramos 1,2, Núria Daranas 1, Mercè Llugany 2, Jiangsu Academy of Agricultural Sciences (JAAS), China Roser Tolrà 2, Emilio Montesinos 1 and Esther Badosa 1* Jen-Tsung Chen, National University of Kaohsiung, 1 Institute of Food and Agricultural Technology-CIDSAV-XaRTA, University of Girona, Girona, Spain, Taiwan 2 Plant Physiology (BABVE), Faculty of Biosciences, Universitat Autònoma de Barcelona, Muhammad Uzair, Bellaterra, Spain National Institute for Genomics and Advanced Biotechnology (NIGAB), Pakistan Marta R. M. Lima, Virginia Tech, United States The present study aims to evaluate the response of the three Mediterranean *CORRESPONDENCE local grapevines ‘Garnacha Blanca’, ‘Garnacha Tinta’, and ‘Macabeo’ to Esther Badosa treatments with biocontrol products, namely a botanical extract (Akivi, esther.badosa@udg.edu Dittrichia viscosa extract) and a beneficial microorganism (Bacillus UdG, SPECIALTY SECTION Bacillus velezensis). A combination of transcriptomics and metabolomics This article was submitted to approaches were chosen in order to study grapevine gene expression and Crop and Product Physiology, a section of the journal to identify gene marker candidates, as well as, to determine differentially Frontiers in Plant Science concentrated grapevine metabolites in response to biocontrol product RECEIVED 24 October 2022 treatments. Grapevine plants were cultivated in greenhouse under ACCEPTED 28 November 2022 controlled conditions and submitted to the treatments. Thereafter, leaves PUBLISHED 16 December 2022 were sampled 24h after treatment to carry out the gene expression study by CITATION RT-qPCR for the three cultivars and by RNA-sequencing for ‘Garnacha Ramos M, Daranas N, Llugany M, Tolrà R, Montesinos E and Badosa E Blanca’. Differentially expressed genes (DEGs) were investigated for both (2022) Grapevine response to a treatments and highly influenced DEGs were selected to be tested in the Dittrichia viscosa extract and a Bacillus velezensis strain. three cultivars as treatment gene markers. In addition, the extraction of leaf Front. Plant Sci. 13:1075231. components was performed to quantify metabolites, such as doi: 10.3389/fpls.2022.1075231 phytohormones, organic acids, and phenols. Considering the upregulated COPYRIGHT and downregulated genes and the enhanced metabolites concentrations, the © 2022 Ramos, Daranas, Llugany, Tolrà, Montesinos and Badosa. This is an treatments had an effect on jasmonic acid, ethylene, and phenylpropanoids open-access article distributed under defense pathways. In addition, several DEG markers were identified the terms of the Creative Commons presenting a stable overexpression after the treatments in the three Attribution License (CC BY). The use, distribution or reproduction in other grapevine cultivars. These gene markers could be used to monitor the forums is permitted, provided the activity of the products in field treatments. Further research will be original author(s) and the copyright necessary to confirm these primary results under field conditions. owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted KEYWORDS academic practice. No use, distribution or reproduction is biocontrol products, transcriptomics, RT-qPCR, gene markers, metabolites, grapevine permitted which does not comply with these terms. Frontiers in Plant Science 01 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 1 Introduction (Ongena and Jacques, 2008; Wang et al., 2013; Mora et al., 2015; Vilà et al., 2016 ). Moreover, laminarin (algae extract) and chito- The European Union is the main world producer, consumer, oligosaccharides associated with oligogalacturonides (COS- and exporter of grapevine for wine-making (Vitis vinifera), and OGA) are already used in vineyards as plant defense the production is mainly concentrated in three countries: Italy stimulators, protecting grapevine against downy mildew and (29.7%), Spain (27.1%), and France (24.2%) (European powdery mildew (Van Aubel et al., 2014; Bodin et al., 2020). Commission, 2021). Vineyards are threatened by several Some beneficial microorganisms are already authorized and diseases, including powdery mildew and gray mold caused by used in vineyards (Otoguro and Suzuki, 2018). It is reported the fungal pathogens Erysiphe necator and Botrytis cinerea, Bacillus subtilis strains that show antagonism activity against respectively, and downy mildew caused by the oomycete gray mold (Maachia et al., 2015) and a Saccharomyces cerevisiae Plasmopara viticola (Avenard et al., 2003; Gessler et al., 2011; cell wall derivatives-based product that induce resistance against Reynier, 2011; Beris et al., 2021). These causal agents are able to downy mildew, gray mold and powdery mildew (De Miccolis infect several grapevine tissues starting from flowers and leaves Angelini et al., 2019). Biocontrol products with a combination (E. necator), from leaves (P. viticola), and from berries (B. between the two types of mechanisms are described as well cinerea). If the first infections are not controlled, the diseases (Krzyzaniak et al., 2018; Esmaeel et al., 2020; Zhou et al., 2020). spread quickly in the vineyard and mildews can infect berries as Plant defense response to biotic stresses relies on different well. These diseases can cause severe crop losses depending on levels of recognition. After pathogen infection, molecular the season and the cultivation area, reducing the harvest quality patterns or effectors of the pathogen are recognized leading to and yield, plant vigor and photosynthesis (Calonnec et al., 2006; (pathogen-associated molecular patterns) PAMP-triggered Leroy et al., 2013; Kunova et al., 2021). immunity (PTI) or effector-triggered immunity (ETI). Both The main grapevine cultivars are susceptible to these PTI and ETI stimulate plant systemic acquired resistance diseases and vineyard protection requires intensive treatments (SAR) (Abdul Malik et al., 2020). Beneficial microorganisms with plant protection products (PPPs), such as chemical recognition can also trigger plant defense response called fungicides from bud burst until ripening (Boubakri et al., induced systemic resistance (ISR). SAR and ISR responses 2013). The frequency average of the applied fungicide involve phytohormones, such as salicylic acid (SA), jasmonic treatments is around ten treatments per year, which can rise acid (JA), and ethylene (ET), being SA more specific to SAR and up to 20 treatments under the most critical conditions (Butault JA/ET pathway to ISR (Pieterse et al., 2014). It is reported that et al., 2010; Leroy et al., 2013; Pertot et al., 2017). This intensive SA is involved in the defense against biotrophic pathogens, use of PPPs can affect the treated crops, the environment, and including P. viticola and E. necator, whereas JA/ET pathway the consumer health as well (Alavanja et al., 2004; Boubakri against necrotrophic pathogens, such as B. cinerea. However, the et al., 2012; Boubakri et al., 2013; Krzyzaniak et al., 2018; two pathways can be activated simultaneously (Burdziej et al., Zambito Marsala et al., 2020). To prevent the negative impact 2021). Direct application of phytohormones or analogues are of the intensive use of synthetic PPPs, more environmentally able to trigger defense response in grapevine against downy friendly compounds, such as biocontrol products are promoted mildew (Bodin et al., 2020; Burdziej et al., 2021). Despite by European governments (European Parliament, Council of the biocontrol products modes of action are not always well- European Union, 2009a; European Parliament, Council of the understood in several plant species and cultivars, it is European Union, 2009b). Among the different types of important to assess that they have no impact on the treated biocontrol products, there are natural substances derived from plants or the environment. plant, animal, or mineral extracts and beneficial microorganisms This study aims to evaluate the response of three appreciated able to protect the plant from pests and diseases. and autochthonous grapevine cultivars of the Mediterranean Natural substances as well as beneficial microorganisms used zone, concretely of Catalonia (Spain): Garnacha Blanca, as biocontrol products present modes of action mainly relying Garnacha Tinta, and Macabeo to biocontrol product on (i) direct action against the pathogen (Bonaterra et al., 2012; treatments. Two biocontrol products from different origins Persaud et al., 2019) or (ii) indirect action by stimulating plant were investigated: a botanical extract (Akivi, Dittrichia viscosa defense (Perazzolli et al., 2011; Perazzolli et al., 2012; Pieterse extract) and a beneficial bacterial strain (Bacillus UdG, Bacillus et al., 2014; Rienth et al., 2019; Nishad et al., 2020; Burdziej et al., velezensis living bacteria). Both products are still in development 2021). It has been reported that a plant extract from Vitis can and the combination of a whole genome transcriptomics present direct activity in grapevine against downy mildew approach and a targeted metabolomics approach was chosen (Schnee et al., 2013). Some beneficial microorganisms are able to elucidate the grapevine response to the treatments. The to compete against pathogens for space and nutrient supplies objectives of this work are: (i) to study grapevine gene (Bonaterra et al., 2012) or to show antagonism activity against expression response after biocontrol product treatment using pathogens through antimicrobial or lytic enzyme production transcriptomics; (ii) to identify robust gene marker candidates Frontiers in Plant Science 02 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 presenting stable differential expression after treatment within rootstock 110R, were obtained from commercial nurseries the three grapevine cultivars; and (iii) to determine grapevine (Agromillora Iberica and Viveros Villanueva Vides, Spain). metabolites variations after biocontrol product treatment using One-year-old bench-grafted grapevine rootlings were planted targeted metabolomics. in a 2 L pot with 80% of the growing media (Prodeasa BV35, Buré s Profesional, Spain), 20% of perlite (A-13, Agroteibe, Spain), and 4 g of the fertilizer (Osmocote® Exact Mini 3-4M, 2 Materials and methods ICL Specialty Fertilizers, France). Bench-grafted grapevines were grown in a greenhouse at 25 ± 2°C, 60 ± 10% relative humidity 2.1 Bacillus UdG production and and a 16:8 h light:dark photoperiod. Young stocks with at least plant extract about 4 to 6 expanded leaves were used for the experiments. The treatments consisted of Akivi at 0.521 g L-1 (Aki), and Bacillus velezensis UdG strain was isolated from a wild plant Bacillus UdG at 108 CFU mL-1 lyophilized (BL) and fresh (BF). collected during a sample screening as reported by Mora et al. Water was used as the solvent to prepare each treatment. The BF (2011). B. velezensis UdG was routinely cultivated on a Luria- treatment was only used in the experiment with cv. Garnacha Bertani agar and incubated at 28°C for 24h. For the assays, two Blanca. A non-treated control (NTC) using water was included different products consisting of lyophilized and fresh cells in all the experiments. The products were sprayed on adaxial and were prepared. abaxial leaf surfaces using an airbrush until near run-off. For lyophilized Bacillus (BL), a fermentation process was done in The experimental design for cv. Garnacha Blanca stocks a pilot-scale bioreactor (Biostat® C, Sartorius, Germany) with a included 4 randomized blocks corresponding to the different working volume of 30 L of production medium for 48 h at 28°C, treatment modalities (Aki, BL, BF, and NTC), while for cvs. pH7 and agitation ramp from 50 to 500 rpm. The production Garnacha Tinta and Macabeo included 3 blocks (Aki, BL, and medium consisted of a modification of the original recipe of Walker NTC). Each block was composed of 4 biological replicates of and Abraham (1970). Specifically, the following modifications were 5 plants. considered: 7 g L-1 instead of 1 g L-1 of KH2PO4, 1 g L-1 instead of 4 g L-1 of L-monosodium glutamate, 5 g L-1 of molasses and 1 g L-1 of soy flour instead of 342 g L-1 of saccharose, 1 mL L-1 instead of 5 mL 2.3 Sampling plant material and L-1 of ferric citrate solution, and 1 mL L-1 of oligoelement solution at RNA isolation 0.1 mg mL-1 instead of at 0.1 mg L-1. After fermentation, the cells were harvested by centrifugation (SA-1-02-175, GEA Westfalia, Sampling was carried out 24 h after spraying plants with the Granollers, Spain) at 10,000 rpm and the concentrated cell products. At that time transcriptomics response, as well as, suspension was mixed with skimmed milk (15% final phytormone signalization and metabolomics responses can be concentration). The bacterial suspension was frozen at −70°C and evaluated. Four biological replicates were sampled for each lyophilized in a laboratory scale freeze-dryer (Unitop HL, VirTis, treatment for RNA-sequencing analysis (RNA-seq), and three Gardiner, NY). Dried samples were stored in vacuum sealed plastic- biological replicates for reverse transcription quantitative PCR coated aluminum bags. (RT-qPCR) analysis. Two leaves from each plant (5 plants per For fresh Bacillus (BF), a fermentation process was carried out in biological replicate) were harvested, grounded, and soaked in a 2-L Erlenmeyer flask for 48 h at 28°C and shaking at 150 rpm with liquid nitrogen. Each ground leaf sample was added to 2 mL 500 mL of the original recipe of production medium (modification: tubes containing two borosilicate glass beads in order to obtain a oligoelement solution was used at 0.1 mg mL-1 instead of at 0.1 mg L- fine powder using Tissuelyzer II system (Qiagen, USA) for 1 min 1 ). After fermentation, the cells were harvested by centrifugation at at 30 Hz. 13,200 g for 10 min (Centrifuge 5810R, Eppendorf) and For total RNA isolation from grapevine leaves, the concentrated 10X with the corresponding volume of supernatant. commercial kit Spectrum™ Plant Total RNA Kit (Sigma- The plant extract Akivi was provided by S.A.S. AkiNaO Aldrich, USA) was used (Supplementary Table 1) following (France). It is a formulated botanical extract prototype from manufacturer’s instructions. Residual DNA was removed using Dittrichia viscosa composed of a high content of polyphenols Invitrogen™ TURBO DNA-free™ Kit (Applied and terpenes (Tamm et al., 2017). Biosystems, USA). The concentration and purity of RNA was assessed by spectrophotometric measurements using NanoDrop ND-1000 2.2 Plant material, treatments, and Spectrophotometer (Thermo Fisher Scientific, USA). RNA experimental design quality was evaluated using electrophoresis on 1.0% agarose gels. Prior to RNA-seq analysis, a R.I.N. measurement was carried Three grapevine cultivars (Vitis vinifera L.), namely out using an Agilent 2100 Bioanalyzer (Agilent technologies, Garnacha Blanca, Garnacha Tinta and Macabeo, grafted on USA) to check RNA integrity from cv. Garnacha Blanca samples Frontiers in Plant Science 03 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 and RNA extracted in each sample was quantified by using the biological function category, different GO terms clusters Qubit 2.0 Fluorometer (Invitrogen, USA). (representative groups) that presented semantic similarity were obtained. The affected DEGs corresponding to all the GO terms of each cluster were added. In addition, GO terms that presented 2.4 RNA-sequencing and reads mapping a background number over 1000 genes (BG-Item) were discarded since they are general GO terms. Clusters that The plant response to treatments using transcriptomics was showed less than 10 DEGs were joined under the term “other” studied on cv. Garnacha Blanca grapevine leaves after spray considering the total number of genes. In addition, metabolic application with Aki, BL, BF or water (NTC). A total of 16 pathways influenced by the treatments were defined using Kyoto samples were used for the library construction. Encyclopedia of Genes and Genomes (KEGG) annotation The RNA-seq transcriptome library was prepared using the (Kanehisa et al., 2016). KEGG pathways with a corrected p- TruSeq Stranded mRNA Sample Prep kit (Illumina, USA) Value < 0.05 were considered significantly influenced by following the manufacturer’s instructions using 1-2 μg of good the treatments. quality RNA (R.I.N. > 7) as input. The RNA was fragmented by 3 minutes at 94°C and each purification step was performed by using 0.81X Agencourt AMPure XP beads. Final libraries were 2.5 Screening of differentially quantified by using the Qubit 2.0 Fluorometer (Invitrogen, USA) expressed genes and quality tested by Agilent 2100 Bioanalyzer RNA Nano assay (Agilent technologies, USA). Libraries were then processed with Screening of DEGs was carried out for each treatment Illumina cBot for cluster generation on the flowcell, following modality (Aki, BL and BF) in comparison with the NTC. The the manufacturer’s instructions and sequenced on paired-end two Bacillus modalities were studied together in order to identify (2x150 bp, 30M reads per sample) at the multiplexing level common genes exclusively due to the bacterial activity, requested on NovaSeq6000 (Illumina). The CASAVA 1.8.2 eliminating the effect of the freeze-drying. version of the Illumina pipeline was used to process raw data Gene expression levels were assessed on the basis of unique for both format conversion and de-multiplexing. mapped genes and were calculated using the FPKM method. Raw sequence files were first subjected to quality control FPKM values were used to analyze the differences in gene analysis by using FastQC v0.10.1 (https://www.bioinformatics. expression between treatments (Aki, BL, and BF) and NTC, by babraham.ac.uk/projects/fastqc/) before trimming and removal calculating a Fold-Change (FC) value. of adapters with BBDuk (https://jgi.doe.gov/data-and-tools/ Due to the high biological variability, the DEGs screening bbtools/) setting a minimum base quality of 25 and a was conducted on the three biological replicates that presented minimum read length of 35 bp. Reads were then mapped less variability between each other, in order to avoid hiding a against the V. vinifera L. genome (V. vinifera cv. Pinot noir part of the treatment impact on the plant. DEGs exclusively var. PN40024) (version 12X Ensembl) with STAR v2.6 (https:// altered by each treatment were targeted. The criteria of selection www.ncbi.nlm.nih.gov/pmc/articles/PMC3530905/). during the screening were based on DEGs presenting high FeatureCounts v1.6.1 (https://academic.oup.com/ differential expression value, specifically Log2(FC) > |1.4| and bioinformatics/article/30/7/923/232889) was then used to good repeatability among the three biological replicates. obtain raw expression counts for each annotated gene using only uniquely mapping reads (MAPQ>=30). The differential gene expression (DGE) analysis was conducted with the R 2.6 Validation of DEGs by RT-qPCR package edgeR (https://www.ncbi.nlm.nih.gov/pmc/articles/ PMC2796818/) using the Trimmed mean of M-values (TMM) To confirm the transcriptome data obtained by RNA-seq normalization method and considering as significant the genes analysis, 27 DEGs were selected (Log2(FC) > |1.4| and good with a False Discovery Rate (FDR) ≤ 0.05. Fragments Per repeatability among the three biological replicates) and their Kilobase Million (FPKM) were obtained with edgeR. Gene expression level was validated by RT-qPCR (Supplementary Ontology (GO) Enrichment Analysis was performed using in- Table 2). The UBQ gene, coding for the Ubiquitin-conjugating house scripts based on the AgriGO publication (https:// enzyme, was used in this study as the endogenous gene for data academic.oup.com/nar/article/45/W1/W122/3796337). GO normalization. This endogenous gene was previously selected enrichment analysis was carried out using a threshold value according to the method described by Silver et al. (2006) (p-Value) < 0.05. The main biological functions were selected (Supplementary Figure 1). considering the Gene Ontology (GO) terms that showed at least Standard curves for DEGs and the endogenous gene were 4 affected DEGs. Then the selected GO terms were analyzed obtained using decimal dilutions of extracted recombinant using REVIGO web platform (http://revigo.irb.hr/) in order to plasmid DNA (target sequences were cloned into a vector summarize GO terms by removing redundancies. For each pSpark® in Escherichia coli DH5a cells) corresponding to Frontiers in Plant Science 04 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 copy numbers ranging between 102 and 107. Ct values in each For phytohormone extraction, 250 mg of fresh grapevine dilution were measured in triplicate and a negative non-template leaves were grounded in an ice-cold mortar with 750 mL of control was included in each run. Real-time PCR reactions extraction solution (methanol:isopropanol:acetic acid; 20:79:1 by included 10 mL SYBR® Green PCR Master Mix (Applied vol.) (Llugany et al., 2013). Then, the supernatant was collected Biosystems), 6 mL RNase-free water, 1 mL of each forward and after centrifugation at 1000 g for 5 min at 4°C. These steps were reverse primer (Supplementary Table 2) at the corresponding repeated two more times and pooled supernatants were concentration, and 2 mL DNA in a final volume of 20 mL. The lyophilized. Finally, samples were dissolved in 250 mL pure optimal primer concentration (100, 300 or 600 nM) was methanol and filtered with a Spin-X centrifuge tube filter of previously defined. The thermal cycling conditions were as 0.22 mm cellulose acetate (Costar, Corning Incorporated, USA). follows: 10 min at 95˚C for initial denaturation; 40 cycles of Phytohormone quantification was done using a standard 15 s at 95˚C, and 1 min at 60˚C; and a final melting curve addition calibration curve spiking control plant samples with program of 60 to 95˚C with a heating rate of 0.5˚C s-1. Ct values the standard solutions of gibberellin A1 (GA1), gibberellin A4 were plotted against the logarithm of their initial template copy (GA4), methyl jasmonate (MeJA), salicylic acid (SA), numbers and each standard curve was generated by a linear (±)-jasmonic acid (JA), (+)-cis, trans-abscisic acid (ABA) and regression of the plotted points. The efficiency of each standard 1-aminocyclopropane-1-carboxylic acid (ACC) ranging from 5 curve was calculated using the formula E = (10(-1/a) -1) x100, to 250 ppb and extracting as described above. Deuterated where “a” is the slope of the curve. hormones jasmonic acid-d5 (JA-d5) and salicylic acid-d6 (SA- For RT-qPCR, total RNA was extracted from leaf samples of d6) at 30 ppb and 300 ppb, respectively, were used as internal treated plants using Spectrum™ Plant Total RNA Kit (Sigma- standards in all the samples and standards measurements. Aldrich) as explained above. First-strand of cDNA was Standards were purchased from Merk (Germany). synthetized from RNA using the High-Capacity cDNA Reverse Plant hormones were analyzed by LC-ESI-MS/MS system in Transcription Kit (Applied Biosystems) according to the multiple reaction monitoring mode (MRM) according to Segarra manufacturer’s instructions. The absence of chromosomal et al. (2006). Phytohormones were separated using HPLC DNA contamination was confirmed by minus-reverse Acquity (Waters, USA) on a Luna Omega C18 column 1.6 μm transcriptase control in qPCR. Quantitative PCR was carried 100 Å 50 x 2.1 mm (Phenomenex, USA) at 50°C at a constant out in a QuantStudio™ 5 Real-Time PCR System (Applied flow rate of 0.8 mL min-1 and 10 μl injected volume. The elution Biosystems) to assess the transcriptional level of 27 DEGs. All gradient was carried out with a binary solvent system consisting the information of the selected genes and primers designed by of 0,1% of formic acid in methanol (solvent A) and 0,1% formic Primer-BLAST tool from the Nacional Centre for Biotechnology acid in milli-Q H2O (solvent B) with the following proportions Information (NCBI) are shown in Supplementary Table 2. (v/v) of solvent A (t (min), %A): (0, 2) (0.2, 2), (1.6, 100), (2, Optimized qPCR reactions and the thermal cycling conditions 100), (2.1, 2) and (3, 2). MS/MS experiments were performed on were described above. Each qPCR assay included duplicates of an ABI 4000 Qtrap mass spectrometer (Sciex). All the analyses each cDNA sample, no-template and RNA controls to check for were performed using the Turbo Ionspray source in negative ion contamination. Ct values from three biological replicates were mode except for MeJA and ACC. averaged, and UBQ gene was used for data normalization. Quantification was made by injection of extracted and The comparative critical threshold (DDCt) method was used spiked samples in multiple reaction monitoring (MRM) mode. to assess the relative quantification of gene expression. Similar Identification of phytohormones was based on retention time amplification efficiencies of all gene primer pairs were checked and presence of peak in the MRM trace compared with those of (Supplementary Table 3) making the DDCt method appropriate the standards. to calculate the Fold-Change (FC). The DCt of the NTC leaf Organic acids (OA) were extracted with a classical extraction samples was used as the calibrating condition to calculate the FC. protocol. Briefly, 250 mg of fresh grapevine leave powder was Genes were considered to be up- or downregulated if their FC grounded in an ice-cold mortar with 2 mL of hydrochloric acid were at least two-fold (FC = 21 or 2−1) higher or less than the (0.025N). Then, the supernatant was collected after calibrator condition (Livak and Schmittgen, 2001) and showed centrifugation at 1000 g for 15 min at 4°C. Meanwhile, Sep- statistically significant differences with the NTC. Pack C18 cartridges (Waters, USA) were activated with (i) 1.4 mL of methanol, (ii) 0.7 mL of milli-Q water, and (iii) 1.4 mL of hydrochloric acid (0,025M). Supernatant (1.4 mL) were passed 2.7 Metabolite analysis through the cartridge to recover 0.7 mL of clean extract. Finally, samples were filtered at 0.22 mm just prior to injection into an Metabolite extractions were carried out from powdered HPLC system. samples of grapevine leaves of cvs. Garnacha Blanca, Garnacha Organic acids were analyzed by HPLC-UV system (Shimazu, Tinta, and Macabeo obtained 24 h after spraying them with Aki, Japan) in the following conditions (Tolrà et al., 2005): YMC- BL, or water (NTC) as explained above. Pack ODS-A HPLC column 5μm 120Å 250 x 4.6 mm (YMC, Frontiers in Plant Science 05 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 Germany) at a constant flow rate of 0.8 mL min-1 and 10 μl treatment modality with the NTC. The statistical analysis of injected volume. The injection method ran 15 min with an the RT-qPCR data was done using REST2009 Software (Pfaffl isocratic flow of 50 nM Potassium dihydrogen phosphate et al., 2002). DEGs standard curves for gene expression (KH 2 PO 4 ) adjusted at a pH of 2.8 using Phosphoric quantification were made by linear regression on Excel. acid (H3PO4). Validation of the DEGs was performed by a correlation study The following standards were used for OA measurements: between the gene expression measured by RNA-seq and RT- acetic acid, cis-aconitic acid, trans-aconitic acid, ascorbic acid, qPCR techniques. Pearson correlation analysis was applied to citric acid, isocitric acid, formic acid, fumaric acid, galacturonic the data for each treatment modality using R software (R acid, gluconic acid, glucuronic acid, glutamic acid, glycine, version 3.5.2). glycolic acid, glyoxylic acid, lactic acid, maleic acid, malic acid, For metabolite measurements, to identify significant malonic acid, oxalic acid, oxoglutaric acid, pyruvic acid, quinic differences between treated (Aki and BL) and NTC leaves, acid, succinic acid, tannic acid, and tartaric acid. several statistical tests were performed. All tests were performed Four peaks corresponding to OA were detected on samples on R software (R version 3.5.2) with a significant level of p-Value HPLC-UV chromatograms. Then, these peaks’ retention times < 0.05. First, each of the metabolite datasets were tested (Shapiro- were compared with the retention times of 26 standards injected Wilk and Bartlett tests) to determine the suitability of parametric in the same conditions, and the identification was confirmed by or non-parametric tests. For parametric tests, one-way analysis of standard enrichment injection within the grapevine samples. variance (ANOVA) was carried out followed by a Tukey’s The four OA were identified (oxalic acid, tartaric acid, malic multiple comparisons test. For non-parametric tests, Kruskal- acid, and oxoglutaric acid) and quantified thanks to calibration Wallis test followed by Dunn test were carried out. curves. Calibration curves: malic acid (y=1.2967x+7.0154, R2 = 0.9967), oxalic acid (y=0.2891x+4.7116, R2 = 0.9993), oxoglutaric acid (y=1.4261x+17.324, R2 = 0.9972), tartaric acid (y=2.6801x 3 Results +2.4512, R2 = 0.9998). Putative identification was carried out by comparing the 3.1 Quality assessment of RNA-seq data retention time of the standards with the peaks obtained in the and gene expression estimation grapevine leaf samples. The standard addition to the samples was done to check that the standard matches the targeted peak in The 16 sequencing samples produced around 39 million of leaf matrix conditions. Calibration curves were done at an total sequencing raw reads for NTC and Aki treatments, while appropriate range for each putatively identified organic acid around 36 million for BL and BF treatments (Supplementary and R² must be above >0.99 to allow quantification. Table 4). Following the filtering and trimming process, around Quantification was made within the samples using the 33 (NTC and Aki) and 31 (BL and BF) millions of cleaned reads calibration curves. were obtained (85 and 86%, respectively, of the total Phenolic compounds were extracted according to Solecka sequencing reads). et al. (1999) with modifications (Kidd et al., 2001). Briefly, leaves When the reads were paired and aligned to the reference V. were extracted with 70% methanol and after centrifugation vinifera L. PN40024 genome, around 14.7 (BL and BF) and 15.8 (10 min, 5000 x g) the supernatant was re-extracted three (NTC and Aki) million reads from each treatment could be times with ethyl ether to eliminate ether soluble lipids. The mapped, (94.9 and 94.6%, respectively, of the input paired reads) remaining water phase was treated with 2 M HCl for acid (Supplementary Table 5). Moreover, between 85.6 (NTC and hydrolysis of soluble conjugated phenolic compounds. After Aki) and 86.7% (BL and BF) of the input paired reads were extraction with ethyl acetate and drying, the residue was re- assigned to genes. dissolved in 50% methanol. Total phenolic compounds levels The overall quality of the experiment was evaluated were determined by spectrophotometry (Shimadzu UV-2450, considering the consistency between the biological replicates Duisburg, Germany) following the method of Folin-Ciocalteau using the normalized gene expression values (normalization of (Singleton et al., 1999), using gallic acid (Sigma, Steinheim, the FPKM) from each treatment. The PCA analysis revealed that Germany) as the standard with detection at 765 nm. The one out of four biological replicates of each treatment (Aki_R1, results were expressed in Gallic Acid Equivalents (GAE). BL_R1, BF_R1, NTC_R3) did not cluster as expected from the experimental design (Supplementary Figure 2). This variability among replicates could hide some of the treatment effect on gene 2.8 Statistical analysis expression, thus, this replicate was not included in further analysis. Principal Component Analysis (PCA) was applied to the PCA on normalized gene expression using the three retained RNA-seq data comparing the biological replicates of each biological replicates showed that the two first principal Frontiers in Plant Science 06 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 components explained 83.57% (Aki), 84.21% (BL) and 85.43% replicates but differed between treatments in comparison with (BF) of the variance. In addition, the PC1 explained 63.33% the NTC. After Aki and BL treatments, the number of genes that (Aki), 60.5% (BL) and 49.38% (BF) of the variability in gene were over-expressed (red) and down-expressed (green) in expression between each treatment and the NTC. comparison with the NTC were equivalent. However, after BF The RNA-seq raw transcriptomic data were submitted to the treatment a higher number of genes were over-expressed (red) in GEO repository of the National Center for Biotechnology comparison with the NTC. Information (NCBI) (GSE211268). As shown in Venn diagrams (Figure 1), 793 genes were upregulated and 652 genes were downregulated (log2(FC)>|1.4|) within the different treatments (Aki, BL and BF) in grapevine 3.2 Analysis of the differential expression leaves after the treatments. Bacillus treatments (BL and BF) of genes after the treatments altered the expression level of a higher number of genes than the botanical extract Akivi treatment (Aki). BL and BF treatments Gene transcription in cv. Garnacha Blanca grapevine showed 438 and 396 upregulated DEGs, respectively, and 481 leaves was triggered by Aki, BL and BF treatments to and 313 downregulated DEGs, respectively, whereas Aki varying degrees. The volcano plots show the degree of treatment showed a total of 278 upregulated and 225 variation of the Differential Expression of Genes (DEGs) downregulated DEGs. In addition, the plant response towards based on red and green dots (Supplementary Figure 3). The Bacillus (both BL and BF) and Akivi (Aki) treatments was fairly relationship between the fold-change (Log2 (FC)) and the specific since only 31 upregulated and 68 downregulated genes statistical significance of the differential expression test were common to all three treatments. (-Log10(FDR)) is displayed. However, the Bacillus treatments, both lyophilized and fresh, Plot similarities within Bacillus treatments (BL and BF) were shared a high number of up- (43.1%) and downregulated observed since the most of genes were distributed between Log2 (41.3%) genes. These genes were altered by the Bacillus (FC) values of -4 and 4 and with significance values (-Log10 treatments, independently of being the product lyophilized or (FDR)) up to 75 (downregulated genes) and 50 (upregulated not. Therefore, these shared genes were used for the following genes). However, Akivi plot differed from Bacillus ones since the validation of RNA-seq results by RT-qPCR. From the 583 main of genes were distributed between Log2(FC) values of -3 upregulated genes after either BL or BF treatments, 251 genes and 3 and with lower significance values of 20 (downregulated were shared. From the rest of genes, 187 and 145 were only genes) and 60 (upregulated genes). upregulated after BL and BF treatments, respectively. Whereas Additionally, heatmaps of these DEGs for each treatment from the 562 downregulated genes after BL or BF treatments, effect are shown in Supplementary Figure 4. The expression 232 genes were shared. From the remaining genes, 249 and 81 patterns of DEGs were consistent within the three biological were only upregulated after BL and BF, respectively. A B FIGURE 1 Venn diagrams showing the relationship between upregulated (A) and downregulated (B) differentially expressed genes (DEGs) identified in leaves of cv. Garnacha Blanca grapevine. Data correspond to 24h after treatments with Akivi (Aki), lyophilized (BL) and fresh (BF) Bacillus UdG, compared to the non-treated control (NTC). Frontiers in Plant Science 07 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 3.3 Functional analysis of DEGs in catabolic process”, “organelle organization”, “protein folding”, grapevine after treatments and “developmental process”, “RNA modification”, and “protein refolding” were related to by BL and/or BF treatments. 3.3.1 GO analysis of DEGs Four biological processes associated with downregulated GO enrichment analysis was carried out to evaluate the genes, namely “metabolic process”, “microtubule-based major biological functions of DEGs influenced by the Aki, BL, movement”, “stress response” and “transmembrane transport” and BF treatments. The biological functions are classified into were shared by the three treatments. “Catabolic process”, “aerial three categories: biological process (BP), cellular component part development”, and “cell wall biogenesis” were exclusively (CC), and molecular function (MF). Upregulated GO terms reduced by Aki treatment. Whereas “carbohydrate metabolic were identified in 34.4, 25.8 and 25.0% of DEGs after the Aki, BL process” and “mitotic cell cycle” were reduced by both Aki and and BF treatments, respectively, whereas 35.3, 32.9 and 41.4% BF treatments, “organelle organization”, “photosynthesis”, and were downregulated (Supplementary Table 6). “biosynthetic process” were related to BF and/or BL treatments. Figures 2, 3 show the upregulated and downregulated GO Some of the upregulated GO terms from BP category that were term clusters obtained by REVIGO analysis. Three biological arranged in two well-defined clusters are related to plant defense processes associated with upregulated genes, namely response, namely “stress response” and “regulation of defense “transmembrane transport”, “stress response”, and “regulation response” (Figure 4). These two clusters include 30 GO terms of defense response” were shared by the three treatments. The (Table 1). In general, only 6 out of 30 GO terms were shared by Aki GO term clusters “Phosphorylation”, “biosynthetic process”, and Bacillus (BF and/or BL) treatments. Five GO terms were shared “cell differentiation”, and “recognition of pollen” were by the two Bacillus treatments (BF and BL), while eight, three, and exclusively enriched by Aki treatment, whereas “protein eight GO terms were unique for Aki, BF and BL, respectively. A C B FIGURE 2 Upregulated genes according to Gene Ontology (GO) enrichment and REVIGO analysis in cv. Garnacha Blanca grapevine after treatment with Akivi (Aki), lyophilized (BL) and fresh (BF) Bacillus UdG, compared to the non-treated control (NTC). (A, B) Bar graphs show the number of upregulated DEGs in each GO term cluster. Clusters that showed less than 10 DEGs were included under the term “other”, indicating in parenthesis the number of clusters that represent. (C) Venn diagrams show the total upregulated GO term clusters. Categories of processes: biological (BP), cellular component (CC), and molecular function (MF). Frontiers in Plant Science 08 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 A C B FIGURE 3 Downregulated genes according to Gene Ontology (GO) enrichment and REVIGO analysis in cv. Garnacha Blanca grapevine after treatment with Akivi (Aki), lyophilized (BL) and fresh (BF) Bacillus UdG, compared to the non-treated control (NTC). (A, B) Bar graphs show the number of downregulated genes in each GO term cluster. Clusters that showed less than 10 DEGs (BP and CC for Aki, BF and BL and MF for Aki) or 20 DEGs (MF for BF and BL) were included under the term “other”, indicating in parenthesis the number of clusters that represent. (C) Venn diagrams show the total downregulated GO term clusters. Categories of processes: biological (BP), cellular component (CC), and molecular function (MF). From the cluster named “regulation of defense response”, Some of the downregulated GO terms from BP category the GO term regulation of jasmonic acid mediated signalling were arranged in a cluster related to plant defense response, pathway was shared by all treatments. Whereas the GO term namely “stress-related response” (Figure 5). This cluster include regulation of defense response was shared by Aki and BF, 18 GO terms related to regulation of cellular cycle and cell regulation of systemic acquired resistance was unique for Aki. population proliferation, plant development, metabolic From the cluster named “stress response”, the GO terms processes and their regulation, stress response, defense and response to osmotic stress, response to karrikin, and response response to stimuli and signal transduction (Table 3). The GO to hydrogen peroxide were unique for Bacillus (BF and BL), terms related to cellular cycle, stress and stimuli response, while the GO terms immune response and plant-type metabolic processes regulation and signal transduction were hypersensitive response, and response to wounding, biotic shared by the three treatments. However, GO terms related to stimulus and ethylene were unique for Aki. plant development, defense response and metabolic processes The upregulated genes (Log2(FC) > 1.4) related to the GO were unique for Bacillus treatments. terms included in “regulation of defense response” and “stress The downregulated genes (Log2(FC) < 1.4) related to the GO response” clusters are shown in Table 2 (Supplementary terms included in “stress-related response” cluster are shown in Tables 7, 8). Interestingly, some upregulated DEGs were also Table 4 (Supplementary Table 9). Interestingly, after both Aki unique for each treatment (Aki and Bacillus). After Bacillus and Bacillus treatments, some DEGs related to cellular cycle treatment, one gene related to the regulation of jasmonic acid, were downregulated. In particular, six cyclin proteins and one and several genes related to transcription factors, chaperones, annexin protein were downregulated for Aki treatment. While enzymes as catalase, PR protein with antimicrobial activity, four cyclin proteins were downregulated for Bacillus treatments. abscisic acid receptor, and cold induced protein were Moreover, genes related to plant growth and development, upregulated. However, after Akivi treatment, two genes related such as transcriptional factors and zinc finger proteins, DNA to the regulation of SAR, one defense response related gene, replication, and two lipid transfer protein (LTP) that intervene three genes related to the response to chitin, and one gene related in systemic acquired resistance SAR were downregulated after to PR protein with antimicrobial activity were upregulated. Bacillus treatments. Frontiers in Plant Science 09 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 The following pathways “Glutathione metabolism”, “ Linoleic a c id meta bolism ”, “alpha-Linolenic acid metabolism”, “Terpenoid backbone biosynthesis”, “beta- Alanine metabolism”, and “Fatty acid degradation” were triggered after Aki treatment. Whereas “Starch and sucrose metabolism” was triggered by both BL and BF treatments, “Ribosome biogenesis in eukaryotes” was exclusively triggered by BF treatment. The pathway “Porphyrin and chlorophyll metabolism” was reduced after Aki treatment, while “DNA replication” was reduced after both BL and BF treatments. “Cysteine and methionine metabolism” was reduced after BL treatment; while “Fructose and mannose metabolism”, “Phenylpropanoid biosynthesis”, and “Starch and sucrose metabolism” were reduced after BF treatment. 3.4 Gene marker candidates on grapevine 3.4.1 Selection of DEGs A total of 27 DEGs were selected since their expression level was modified due to Aki, BL and BF treatments according to the results of RNA-seq analysis (Table 5). From the 12 DEGs highly triggered by Aki treatment, eight genes are related to defense response (A1, A3, A4, A5, A6, A7, A11, and A12). Specifically, two genes are involved in detoxification of reactive oxidative species (A5 and A6); two genes are related to hormone signalling pathway (A3, and A4), one gene is involved in biosynthesis of secondary metabolites (A7), and one gene is a marker of SAR response (A11). From the 15 DEGs highly triggered by BL and BF treatments, four genes are involved in defense response (B1, B6, B8, and B14). FIGURE 4 REVIGO graphs of upregulated GO term clusters (regulation of defense response and stress response) in leaves of cv. Garnacha 3.4.2 Validation of selected DEGs by RT-qPCR Blanca grapevine included in biological process category after Standard curves of the 27 DEGs showed R-squared values treatments with Akivi (Aki), lyophilized (BL) and fresh (BF) Bacillus above 0.99 and, in general, amplification efficiencies above 90%, UdG. ID: identification of GO terms associated with Table 1. except for three DEGs (A1, A3 and A4) that showed slightly lower efficiencies above 80% (Supplementary Table 3). The expression levels of the 27 DEGs within the NTC samples on Considering different stress responses, after Aki treatment the ‘Garnacha Blanca’ experiment were stable showing FC values two genes connected with receptor like kinases that intervene in close to 1 (Table 6). The selected DEGs were upregulated after plant innate immunity were downregulated, while transcription Aki (12) and BL treatments (14) with significant differences in factors to several stresses and abiotic stresses were comparison with the NTC, with the exception of B5 gene that downregulated after Bacillus treatments. was downregulated. Moreover, the relative expression levels of the 27 DEGs on 3.3.2 KEGG pathway analysis of DEGs cv. Garnacha Blanca obtained by RT-qPCR and RNA-seq KEGG pathway analysis was performed to evaluate the analysis were highly consistent for both Aki and BL treatments biological mechanisms influenced by the Aki, BL, and BF (Figure 6). That was confirmed by Pearson correlation test that treatments. Few pathways were associated with DEGs affected showed high correlation coefficient values, 0.729 and 0.938 for by the treatments and none was shared between Akivi and Aki and BL, respectively, and statistical significances with p- Bacillus treatments (Supplementary Table 10). values<0.05 (Supplementary Figure 5). Therefore, the 27 DEGs Frontiers in Plant Science 10 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 TABLE 1 Representative groups (clusters) of upregulated GO terms of biological processes obtained with REVIGO and associated to plant defense responses, after treatments of cv. Garnacha Blanca grapevine leaves with Akivi (Aki), lyophilized (BL) and fresh (BF) Bacillus UdG. Uniqueness* ID Representative group GO ID: GO Term Aki BF BL Regulation of defense response 9 GO:2000022: regulation of jasmonic acid mediated signalling pathway 13 GO:0031347: regulation of defense response 26 GO:0051096: positive regulation of helicase activity 14 GO:0010112: regulation of systemic acquired resistance 10 GO:0045454: cell redox homeostasis 12 GO:0010469: regulation of signalling receptor activity 30 GO:0006879: cellular iron ion homeostasis Stress response 1 GO:0010200: response to chitin 8 GO:0061408: positive regulation of transcription from RNA polymerase II promoter in response to heat stress 5 GO:0034605: cellular response to heat 11 GO:0006355: regulation of transcription, DNA-templated 19 GO:0000165: MAPK cascade 25 GO:0006970: response to osmotic stress 20 GO:0080167: response to karrikin 24 GO:0042542: response to hydrogen peroxide 2 GO:0006955: immune response 3 GO:0009611: response to wounding 4 GO:0009607: response to biotic stimulus 6 GO:0009723: response to ethylene 7 GO:0009626: plant-type hypersensitive response 27 GO:0046686: response to cadmium ion 28 GO:0009739: response to gibberellin 29 GO:0009651: response to salt stress 15 GO:0010039: response to iron ion 16 GO:0070413: trehalose metabolism in response to stress 17 GO:0035556: intracellular signal transduction 18 GO:0009617: response to bacterium 21 GO:0006073: cellular glucan metabolic process 22 GO:0009738: abscisic acid-activated signalling pathway 23 GO:0010167: response to nitrate Akivi (Aki), lyophilized (BL) and fresh (BF) Bacillus UdG treatments ID: GO term assigned identifier White space means not GO term *Smaller values denote higher uniqueness. Red (0.7-0.8), orange (0.8-0.9), yellow (0.9-1.0) that were previously selected by RNA-seq analysis were validated A2, A3, A4, A9, and A12) in ‘Garnacha Tinta’ and ‘Macabeo’, by RT-qPCR on grapevine cv. Garnacha Blanca. respectively, showed differential expression levels with statistical significance compared to the NTC (regardless of the FC value) 3.4.3 Expression of validated DEGs in the three (Table 6). After Akivi treatment, the A1, A4, and A12 genes were grapevine cultivars upregulated on the three grapevine cultivars (Figure 7). The expression levels of the 27 DEGs were subjected to RT- However, the A1 gene showed a FC value of 4.86 without qPCR using samples from experiments performed with two significant differences with the NTC on ‘Garnacha Tinta’. other grapevine cultivars, namely Garnacha Tinta and Whereas the A9 gene was upregulated on ‘Garnacha Blanca’ Macabeo. Within the NTC samples of the ‘Garnacha Tinta’ and ‘Garnacha Tinta’, this gene was downregulated on and ‘Macabeo’ experiments, the expression levels of the 27 DEGs ‘Macabeo’. In the case of A8 gene, despite it was upregulated were stable showing fold change values close to 1 (Table 6). on ‘Garnacha Blanca’ and ‘Garnacha Tinta’, its gene expression Concerning the 12 selected DEGs by Akivi treatment, nine was not affected on ‘Macabeo’. Seven genes, namely, A2, A3, A5, (A2, A3, A4, A5, A7, A8, A9, A10, and A12) and six genes (A1, A6, A7, A10 and A11, were only upregulated on ‘Garnacha Frontiers in Plant Science 11 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 TABLE 2 Upregulated genes included in the GO terms that belong to regulation and stress response groups after treatments of cv. Garnacha Blanca grapevine leaves with Akivi (Aki), lyophilized (BL) and fresh (BF) Bacillus UdG. Description Gene ID GO ID Aki BF BL Enzymes Leucoanthocyanidin dioxygenase VIT_13s0067g01020 9 Trehalose 6-phosphate synthase VIT_17s0000g08010; VIT_01s0026g00280 16 Trehalose-phosphatase VIT_12s0028g01670 16 Xyloglucan endotransglucosylase/hydrolase VIT_11s0052g01280; VIT_05s0062g00250; VIT_01s0026g00200 21 Catalase VIT_00s0698g00010 27 Proteins that mediate the attachment of integral membrane proteins to the cytoskeleton Ankyrin repeat VIT_14s0081g00370 13 Ankyrin repeat VIT_05s0165g00010; VIT_14s0081g00360 13 Transcriptional regulators/Transcriptional factors Jasmonate ZIM domain-containing protein 8 VIT_10s0003g03790 9, 13 Cold induced protein VIT_17s0000g08010 27 Zinc finger (C2H2 type) family VIT_13s0019g00480 1 Myb domain protein 14 VIT_05s0049g01020 1, 27, 29 Salt tolerance homolog2 VIT_03s0038g00340 1, 20 WRKY DNA-binding protein 33 VIT_08s0058g00690 1, 5, 25, 29 Heat shock transcription factor C1 VIT_11s0016g03940 5, 8 Modulators and regulators of related defense responses and cell death program NIM1 VIT_07s0005g02070; VIT_01s0011g03430 14 NSL1 (necrotic spotted lesions 1) VIT_01s0011g05950 2, 7 Abscisic acid receptor PYL1 RCAR12 VIT_13s0067g01940 22 Plant peptide growth factors. Phytosulfokines PSK1 VIT_08s0007g03870 12 DNA replication DNA mismatch repair protein MSH3 VIT_00s0388g00030 26 Iron storage and transport proteins Ferritin VIT_08s0058g00440, VIT_08s0058g00430, VIT_08s0058g00410 15, 18, 24, 30 Metal-nicotianamine transporter YSL1 VIT_02s0025g02510 15 Chaperones (HSP) Heat shock protein 18.2 kDa class II VIT_12s0035g01910 24, 29 Heat shock protein 17.6 kDa class I VIT_13s0019g03160 24, 29 HSP (HSP26.5-P) 26.5 kDa class P VIT_00s0992g00020 24, 29 Pathogenesis related proteins Pathogenesis protein 10 VIT_05s0077g01570 22 Pathogenesis protein 10 VIT_05s0077g01600 4 Unknown unknown VIT_09s0002g03340 27 Go ID: GO term assigned identifier (Table 1) Blanca’, while the expression pattern of these genes was not In relation to the 15 selected DEGs by BL treatment, twelve affected on ‘Garnacha Tinta’ and ‘Macabeo’. Therefore, the genes showed differential expression levels with statistical expression pattern after Akivi treatment was quite similar in significance compared to the NTC within ‘Garnacha Tinta’ the ‘Garnacha Blanca’ and ‘Garnacha Tinta’ (5 out of 12 genes (B1, B2, B3, B5, B6, B7, B9, B10, B12, B13, B14, and B15) and were upregulated). However, the expression pattern obtained on ‘Macabeo’ (B1, B2, B3, B5, B6, B7, B8, B9, B11, B12, B13, and ‘Macabeo’ differed from ‘Garnacha Blanca’ and ‘Garnacha Tinta’ B14) (regardless of the FC value) (Table 6). After BL treatment, since only 3 out of 12 (A1, A4, A12) genes were upregulated on B1, B2, B3, B9, B12, B13, and B14 genes were upregulated on the all cultivars tested. In particular, three genes (A9, A10 and A11) three grapevine cultivars (Figure 7). The only gene that was showed FC below 1 only on ‘Macabeo’, being A9 downregulated. downregulated on ‘Garnacha Blanca’ was B5, which was Frontiers in Plant Science 12 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 were not affected on ‘Macabeo’. Therefore, the expression pattern after BL treatment was quite similar in the ‘Garnacha Blanca’ and ‘Garnacha Tinta’ (9 out of 15 genes were upregulated). Similar results were also observed comparing expression patterns in ‘Garnacha Blanca’ and ‘Macabeo’ (9 out of 15 genes were upregulated) despite B5 gene was clearly upregulated on ‘Macabeo’ and downregulated on ‘Garnacha Blanca’. However, the expression pattern obtained on ‘Macabeo’ differed from ‘Garnacha Tinta’ since a smaller number of genes (7 out of 15) shared the same upregulation transcriptional pattern. 3.5 Metabolite concentrations Foliar Aki and BL treatments slightly influenced some of the mineral nutrient concentrations in the grapevine leaves, but not enough to affect plant development in any of the three cultivars (Supplementary Table 11). Phytohormones, organic acids (OA) and total phenolic compounds concentrations were compared for each treatment (Aki and BL) with the NTC (Table 7). Regarding Aki treatment, no phytohormones were significantly influenced in the same way among the three grapevine cultivars. The SA tended to present higher levels after Aki treatment in ‘Garnacha Blanca’ and ‘Macabeo’ and was significantly enhanced in ‘Garnacha Tinta’. The GAs showed an opposite pattern being their levels significantly increased in ‘Garnacha Blanca’, but reduced in ‘Macabeo’. The MeJA in ‘Garnacha Blanca’ was significantly enhanced after Aki treatment. Although all the studied phytohormones, with the exception of JA, tended to be enhanced after Aki treatment in ‘Garnacha Blanca’; JA, ACC, and ABA were not significantly influenced by Aki treatment in any of the three cultivars. FIGURE 5 BL treatment significantly enhanced JA in ‘Garnacha Tinta’ REVIGO graphs of downregulated GO term clusters (regulation and MeJA in ‘Garnacha Blanca’. Oppositely, GAs were of defense response and stress response) in leaves of cv. Garnacha Blanca grapevine included in biological process significantly reduced after BL treatment in ‘Macabeo’. category after treatments with Akivi (Aki), lyophilized (BL) and However, neither SA, ACC, nor ABA were significantly fresh (BF) Bacillus UdG. ID: identification of GO terms associated with Table 3. influenced by BL treatment in any of the three cultivars. After BL or Aki treatment, phytohormone content changed without a clear pattern and the establishment of a defense signalling triggering mechanism was not possible to infer from unaltered and upregulated on ‘Garnacha Tinta’ and ‘Macabeo’, our data. respectively. Three genes, namely B4, B6, and B7, were only It is worth to mention that ABA global values detected in upregulated on ‘Garnacha Blanca’, while they were unaltered ‘Macabeo’ are higher than the values detected in the two (FC between 0.5-2) on ‘Garnacha Tinta’ and ‘Macabeo’. In the ‘Garnacha’ varieties. case of B8 and B11 genes, their expression levels were Four OA were identified in the leaves of the three grapevine upregulated on both ‘Garnacha Blanca’ and ‘Macabeo’, while cultivars: oxalic, tartaric, malic, and oxoglutaric (Table 7). Aki their expression levels were not affected on ‘Garnacha Tinta’. and BL treatments caused a significant reduction in oxoglutaric Two genes, namely B10 and B15, were upregulated on both acid in ‘Macabeo’, but BL significantly increased the amount of ‘Garnacha Blanca’ and ‘Garnacha Tinta’, while their expressions this organic acid in ‘Garnacha Blanca’. BL treatment also Frontiers in Plant Science 13 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 TABLE 3 Representative groups (clusters) of downregulated GO terms of biological processes obtained with REVIGO and associated to plant defense responses, after treatments of cv. Garnacha Blanca grapevine leaves with Akivi (Aki), lyophilized (BL) and fresh (BF) Bacillus UdG. Uniqueness* Representative group ID GO ID: GO Term Aki BF BL Stress-related response 3 GO:0045787: positive regulation of cell cycle 4 GO:0006355: regulation of transcription 8 GO:0009414: response to water deprivation 7 GO:0008284: positive regulation of cell population proliferation 1 GO:0009734: auxin-activated signaling pathway 10 GO:0010017: red or far-red light signaling pathway 13 GO:0009744: response to sucrose 11 GO:0045910: negative regulation of DNA recombination 2 GO:0043086: negative regulation of catalytic activity 5 GO:0007178: transmembrane receptor protein serine/threonine kinase signaling pathway 6 GO:0071249: cellular response to nitrate 9 GO:0009909: regulation of flower development 12 GO:0043085: positive regulation of catalytic activity 14 GO:0046686: response to cadmium ion 15 GO:0010112: regulation of systemic acquired resistance 16 GO:0009627: systemic acquired resistance 17 GO:0000076: DNA replication checkpoint signaling 18 GO:0045893: positive regulation of transcription Akivi (Aki), lyophilized (BL) and fresh (BF) Bacillus UdG treatments ID: GO term assigned identifier White space means not GO term * Smaller values denote higher uniqueness. Purple (0.7-0.8), blue (0.8-0.9), clear blue (0.9-1.0) reduced the level of tartaric acid in ‘Garnacha Tinta’. The rest of influenced by the treatments. Interestingly, the two Bacillus organic acids were not altered. UdG treatments, both BF and BL triggered the same pathways. Total phenolic compounds concentration was significantly However, BF and BL did not always trigger the same genes of the enhanced after BL treatment in ‘Macabeo’ and in ‘Garnacha above-mentioned pathways. These results underline the Blanca’. Aki treatment only tend to increase the level of total importance of the product formulation since it may determine phenolic compounds. its efficacy and mode of action. Only Aki and BL treatments will be discussed hereafter. Gene transcription related to JA biosynthesis was slightly 4 Discussion influenced by either Aki and BL treatments since lipoxygenases (LOX) related genes did not show overexpression with Log2(FC) The response of grapevines to foliar treatments with the values higher than 1.4. However, phytohormone concentrations botanical extract Akivi (Aki) and the beneficial microorganism related to JA defense pathway were affected in cvs. Garnacha Bacillus UdG (fresh, BF or lyophilized, BL) was investigated. Blanca and Tinta. Specifically, MeJA concentration was Among the environmentally friendly compounds considered as significantly higher after Aki and BL treatments in ‘Garnacha PPPs by the European Union legislation, beneficial Blanca’ and JA concentration was doubled after BL treatment in microorganisms and natural substances (i.e. plant extracts) are ‘Garnacha Tinta’. In agreement with metabolite contents, several included. Bacillus UdG and Akivi were used in this study genes regulated by JA pathway were upregulated by the representing each modality. Previous studies have reported the treatments. Namely, the expression level of genes related to features and potential of both biocontrol products (Mora et al., non-inducible immunity 1 (NIM1) were upregulated by Aki 2011; Mora et al., 2015; Ramos et al., 2022). Figure 8 shows a treatment and the expression level of a gene related to enhanced scheme as a summary of the genes related to the main plant disease susceptibility (EDS1) and to nonexpressor of defense response pathways (Jasmonic Acid, JA; Salicylic Acid, pathogenesis-related genes 1 (NPR1) was slightly upregulated SA; Ethylene, ET; Abscisic Acid, ABA; phenylpropanoids by both treatments. Moreover, Aki and BL treatments pathway; and mitogen activated protein kinases and Ca2+ upregulated the expression of several genes involved in signalling induction, MAPKs) whose expression levels were different transcription factors with WRKYs domain (WRKYs Frontiers in Plant Science 14 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 TABLE 4 Downregulated genes included in the GO terms that belong to stress-related response groups after treatments of cv. Garnacha Blanca grapevine leaves with Akivi (Aki), lyophilized (BL) and fresh (BF) Bacillus UdG. Gene description Gene ID GO ID Aki BF BL Transcription factor related to auxin signalling pathway IAA31 VIT_05s0020g01070 1 Proteins that control the cell cycle by activating cyclin-dependent kinase (CDK)/Cycle regulators Cyclin delta-3 (CYCD3_1) VIT_18s0001g09920 Cyclin D3_2 VIT_03s0180g00040 2, 3, 6, 7 Cyclin CYCB1_2 VIT_06s0009g02090 2, 3, 6, 7 Cyclin B-type VIT_08s0040g00930 2, 3, 6, 7 Cyclin 1b (CYC1b) VIT_13s0067g01420 2, 3, 6, 7 Cyclin-dependent protein kinase regulator CYCB2_4 VIT_18s0001g14170 2, 3, 6, 7 Cyclin B2;4 VIT_03s0038g02800 3, 7 Cyclin delta-2 VIT_03s0091g01060 3, 7 Cyclin A1 VIT_18s0001g02060 3, 7 Cyclin-dependent protein kinase CYCB3 VIT_19s0085g00690 3, 7 Annexin ANN4 VIT_00s0131g00080 3,8 Protein kinase WEE1 VIT_07s0104g01740 17 Proteins that join DNA to form nucleosomes Histone H4 VIT_06s0004g04370; VIT_13s0019g00780; VIT_13s0019g00800 8 Histone H1 VIT_07s0005g01060; VIT_07s0141g00730; VIT_14s0081g00500 11 Receptors like-Kinases (RLK) Proline extensin-like receptor kinase 1 (PERK1) VIT_01s0127g00670 3 Receptor protein kinase VIT_05s0020g01690 3 DNA replication and repair ATP-dependent DNA helicase RecQ VIT_01s0010g02590 3, 7 Origin recognition complex subunit 5 VIT_01s0011g04400 13 Origin recognition complex subunit 4 VIT_17s0000g01960 13 DNA mismatch repair protein VIT_01s0011g03440 15 B ZipDNA binding proteins/Transcription factors/Zinc finger proteins BZIP protein HY5 (HY5) VIT_04s0008g05210 10 BZIP protein HY5 (HY5) VIT_05s0020g01090 10 BZIP transcription factor BZIP6 VIT_00s0541g00020 18 AP2/ERF domain containing protein VIT_08s0007g08150 18 NAC Secondary wall thickening promoting factor1 VIT_02s0025g02710 18 Late meristem identity1 HB51/LMI1 VIT_08s0007g04200 18 Homeodomain leucine zipper protein HB-1 VIT_01s0026g01550 18 Homeobox-leucine zipper protein HB-7 VIT_15s0048g02870 18 Constans 2 (COL2) VIT_14s0083g00640 9 Zinc knucle VIT_01s0010g01670 17 Lipid Transfer Proteins (LTP) DIR1 (defective IN induced resistance 1) VIT_00s0333g00050 16 Protease inhibitor/seed storage/lipid transfer protein (LTP) VIT_08s0007g01370 16 Unknown unknown VIT_04s0008g04200 5 unknown VIT_04s0023g03760 5 unknown VIT_07s0129g00200 3, 7 unknown VIT_13s0067g02560 18 Go ID: GO term assigned identifier (Table 3) Frontiers in Plant Science 15 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 TABLE 5 Selected Differentially Expressed Genes (DEGs) on cv. Garnacha Blanca grapevine leaves after treatment with Akivi (Aki) and lyophilized Bacillus UdG (BL). Code Gene ID Log2 (FC) FC FDR vCOST Description Akivi A1 VIT_12s0059g02600 4.96 31.18 5.01E-15 Receptor protein kinase RK20-1 A2 VIT_06s0004g03350 3.46 11.03 9.18E-24 Lateral organ boundaries protein 1 A3 VIT_05s0077g00520 3.17 8.99 7.76E-11 Gibberellin 2-oxidase A4 VIT_17s0000g00200 3.23 9.40 1.58E-17 Ethylene-responsive transcription factor ERF114 A5 VIT_08s0058g00970 2.39 5.23 1.67E-17 Cationic peroxidase A6 VIT_12s0055g01010 3.04 8.23 1.38E-28 Peroxidase A7 VIT_00s0372g00040 2.74 6.68 1.68E-08 1,8-cineole synthase, chloroplast A8 VIT_04s0023g02240 2.83 7.11 7.56E-56 S-adenosyl-L-methionine:salicylic acid carboxyl methyltransferase A9 VIT_12s0034g01140 2.07 4.21 5.28E-21 Plastocyanin domain-containing protein A10 VIT_19s0090g00660 2.01 4.03 1.03E-32 Lipase GDSL A11 VIT_03s0088g00810 1.88 3.67 3.94E-16 Pathogenesis-related protein 1 precursor (PRP 1) A12 VIT_07s0005g06090 1.67 3.19 9.06E-19 Pore-forming toxin-like protein Hfr-2 Bacillus B1 VIT_16s0022g00860 5.23 37.41 1.25E-28 Invertase/pectin methylesterase inhibitor B2 VIT_06s0004g07210 5.45 43.58 4.78E-65 CCT motif constans-like B3 VIT_16s0100g00740 4.26 19.16 2.49E-15 unknown B4 VIT_14s0068g01160 2.91 7.53 5.22E-12 Cytokinin-repressed protein CR9 B5 VIT_00s1490g00010 -2.46 0.18 9.93E-38 5’-adenylylsulfate reductase (APR1) B6 VIT_13s0064g01370 3.08 8.43 4.30E-07 Polygalacturonase inhibiting protein 1 PGIP1 B7 VIT_09s0002g04280 3.14 8.81 2.33E-47 Dynein light chain LC6, flagellar outer arm B8 VIT_03s0091g00310 2.96 7.80 6.23E-16 Indole-3-acetic acid-amido synthetase GH3.8 B9 VIT_01s0011g01980 2.47 5.52 3.91E-22 fasciclin arabinogalactan-protein (FLA21) B10 VIT_01s0026g02740 2.64 6.25 9.54E-28 unknown B11 VIT_08s0058g00430 1.82 3.52 1.32E-02 ferritin B12 VIT_10s0116g00530 1.96 3.89 1.07E-30 Thiazole biosynthetic enzyme, chloroplast (ARA6) B13 VIT_00s0480g00060 1.49 2.81 7.91E-20 Polyphenol oxidase [Vitis vinifera] B14 VIT_07s0031g02610 2.98 7.92 3.48E-14 NAC domain containing protein 2 B15 VIT_13s0067g02130 2.50 5.64 6.67E-10 Dehydration-induced protein (ERD15) FC, fold change FDR, false discovery rate TFs) and pathogenesis related proteins (PR). Interestingly, one treated with NTC or BL on all three cultivars, especially in of the PR related genes (VIT_03s0088g00810) was highly ‘Garnacha Tinta’ in which significant differences were observed. influenced by Aki treatment. Therefore, JA defensive pathway It could be explained by the upregulation of the expression of seemed to be triggered by both treatments. This is in agreement some genes related to EDS1, NIM1, and NPR1 already with the essential role of JA as a phytohormone involved in the commented above in JA paragraph. Actually the mentioned regulation of defense gene expression (Rienth et al., 2019), such genes are described as modulators that intervene in SA as EDS1, NPR1, or NIM1 related genes (Ochsenbein et al., 2006; accumulation and they are produced in crosstalk between SA Pajerowska-Mukhtar et al., 2013; Backer et al., 2019). These and JA pathways (Rusté rucci et al., 2001; Zhu et al., 2011; Chen results enlightened the link between PR upregulated genes and et al., 2021). Interestingly, the expression level of the WRKYs TFs but they did not underline the bond with glutathione-S-transferase (GST) related gene was also transcriptional regulators jasmonate-zim domain (JAZ) upregulated by Aki treatment. It is reported that SA is able to intermediate related genes as previously described (Kazan and regulate several genes from GST family that are upregulated Manners, 2012; Guerreiro et al., 2016). through SA pathway in treated plants with beneficial The expression level of genes related to SA pathway were not microorganisms resulting in ISR priming (Gullner et al., 2018). clearly affected by neither Aki nor BL treatments. However, the GST family enzymes are involved in detoxifying cytotoxic measured concentration of SA phytohormone in grapevine compounds and the process implies transmembrane transport leaves treated with Aki tended to be higher than in leaves (Burdziej et al., 2021), which is in agreement with these results Frontiers in Plant Science 16 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 TABLE 6 Expression levels of the selected DEGs influenced by treatments of cvs. Garnacha Blanca, Garnacha Tinta, and Macabeo with Akivi (Aki) and lyophilized Bacillus UdG (BL). ‘Garnacha Blanca’ ‘Garnacha Tinta’ ‘Macabeo’ DEGs NTC Aki NTC Aki NTC Aki A1 1.29 14.37 ± 3.96 * 1.29 4.86 ± 4.91 1.11 4.46 ± 1.73 * A2 1.08 10.12 ± 0.79 * 1.00 1.88 ± 0.71 * 1.02 1.58 ± 0.50 * A3 1.05 5.64 ± 0.57 * 1.03 1.43 ± 0.25 * 1.04 1.74 ± 0.81 * A4 1.12 3.27 ± 0.38 * 1.02 2.22 ± 0.72 * 1.02 3.20 ± 1.19 * A5 1.05 8.01 ± 1.07 * 1.02 1.55 ± 0.32 * 0.99 1.68 ± 0.30 A6 1.05 9.11 ± 0.44 * 1.09 1.39 ± 0.49 1.05 1.16 ± 0.80 A7 1.06 3.39 ± 0.61 * 1.06 1.65 ± 0.59 * 1.10 1.10 ± 0.60 A8 1.08 4.71 ± 1.71 * 1.05 2.94 ± 1.90 * 1.16 1.78 ± 1.27 A9 1.06 4.56 ± 1.14 * 1.01 2.79 ± 2.32 * 1.01 0.44 ± 0.21 * A10 1.01 3.15 ± 0.72 * 1.00 1.69 ± 0.53 * 1.01 0.92 ± 0.29 A11 1.01 3.24 ± 0.96 * 1.12 1.44 ± 0.38 1.02 0.69 ± 0.40 A12 1.10 3.54 ± 0.40 * 1.01 9.96 ± 3.41 * 1.01 2.23 ± 0.68 * DEGs NTC BL NTC BL NTC BL B1 1.16 58.79 ± 23.12 * 1.06 14.51 ± 3.64 * 1.01 4.07 ± 1.64 * B2 1.05 22.43 ± 3.57 * 1.03 3.19 ± 0.34 * 1.02 4.05 ± 1.17 * B3 1.06 7.98 ± 2.92 * 1.00 2.21 ± 0.17 * 1.00 2.02 ± 0.19 * B4 1.15 5.49 ± 1.84 * 1.09 0.77 ± 0.20 0.90 1.63 ± 0.13 B5 1.04 0.22 ± 0.07 * 1.02 0.50 ± 0.18 * 1.09 3.03 ± 0.58 * B6 1.01 5.17 ± 1.71 * 1.01 1.48 ± 0.15 * 1.04 0.54 ± 0.08 * B7 1.29 6.86 ± 1.57 * 1.04 1.39 ± 0.25 * 1.07 0.71 ± 0.12 * B8 1.03 4.10 ± 0.67 * 1.01 1.38 ± 0.24 1.08 2.43 ± 0.21 * B9 1.09 8.30 ± 3.71 * 1.01 5.07 ± 1.22 * 1.06 2.76 ± 0.59 * B10 1.07 7.66 ± 1.43 * 1.01 3.17 ± 0.73 * 1.01 1.26 ± 0.38 B11 1.15 4.71 ± 1.45 * 1.02 1.16 ± 0.44 1.07 7.18 ± 1.72 * B12 1.01 4.68 ± 0.90 * 1.01 3.76 ± 0.54 * 1.03 8.59 ± 1.48 * B13 1.00 4.32 ± 0.62 * 1.02 2.84 ± 0.28 * 1.02 2.07 ± 0.51 * B14 1.12 5.96 ± 1.32 * 1.04 3.15 ± 0.76 * 1.22 2.50 ± 0.22 * B15 1.03 6.26 ± 1.58 * 1.04 2.66 ± 0.58 * 1.03 0.92 ± 0.27 DEG functions are indicated in Supplementary Table 2. Data correspond to RT-qPCR. The relative expression level of each gene was calculated by the comparative critical threshold (DDCt) method using the non-treated control samples (NTC) as the calibrator and UQB gene as internal control for data normalization. Data mean the fold change (2-DDCt) ± confidence interval and significant differences according to REST2009 Software are represented by *. since “transmembrane transport” GO term was influenced by response to biotic and abiotic stresses, activating PR genes, such Aki and BL treatments. as osmotins (PR-5), chitinases (PR-3) and b-1,3-glucanases (PR- The biosynthesis of ET seemed to be triggered by both 2) (Mizoi et al., 2012; Bahieldin et al., 2016; Rienth et al., 2019), treatments through the upregulation of the expression level of which were indeed upregulated after both Aki and 1-aminocyclopropane-1-carboxylate oxidase (ACO) related BL treatments. genes. Moreover, the concentration of the ET precursor 1- The ABA biosynthesis was triggered by BL treatment aminocyclopropane-1-carboxylic acid (ACC) tended to present through the upregulation of 9-cis-epoxycarotenoid higher levels in ‘Garnacha Blanca’ leaves after Aki and BL dioxygenase (NCED) related gene expression. In fact, ABA treatments. These results were in agreement with the biosynthesis starts with carotenoids and involves NCED upregulation of expression of genes related to response factors enzyme that is strongly upregulated by stress (Xiong and Zhu, regulated by ET (ERF TF, (AP2)/ERF TF, AP2 TF) by both 2003). ABA is involved in the response to water stress and treatments. In addition, the expression level of one of the genes particularly intervenes in stomatal closure (Catacchio et al., related to ERF TF was highly influenced by Aki treatment 2019; Postiglione and Muday, 2020). It is expected variability (VIT_17s0000g00200). Therefore, ET defensive pathway in water stress response between grapevine cultivars because the seemed to be triggered by both treatments. ET response two ‘Garnacha’ are more resistant to drought than ‘Macabeo’ factors are key regulators of JA, ET, and ABA pathways in (Mirá s-Avalos and Araujo, 2021). These results are consistent Frontiers in Plant Science 17 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 A B FIGURE 6 Expression levels of twenty-seven genes selected for validation of the RNA-Seq data by RT-qPCR. The gene expression was analysed after treatments with Akivi (A) and lyophilized Bacillus UdG (B). RNA-seq (stripped bars) and RT-qPCR (black bars) analysis. Gene functions are indicated in Supplementary Table 2. RT-qPCR data are shown as the mean of Log2(FC)of three biological replicates, where FC is the fold-change value and was calculated as 2-DDCt using non- treated control (NTC) samples as the calibrator and UBQ gene for data normalization. Error bars mean confidence interval of three biological replicates. with ABA measured concentrations that were twice or three also triggered by both treatments through the upregulation of times higher in ‘Macabeo’ than in the two ‘Garnacha’. Actually, Stilbene Synthase (STS) and Myb TF related gene expression. ‘Macabeo’, which is less resistant to drought, is more likely to The transcriptomic results were in accordance with the total trigger water stress response involving ABA signaling. ABA is phenolic concentration in leaves, which tended to be higher after also involved in pathogen response signaling pathway and linked Aki and BL treatments in ‘Garnacha Blanca’ and statistically with SA, JA, and ET related genes regulation (Nishad et al., higher in ‘Macabeo’ after BL treatment. It is worth to mention 2020). For instance, ABA biosynthesis induction by laminarin that the phytohormones JA, MeJA, SA, ET, and ABA positively treatment triggered JA production in grapevine (Balestrini et al., regulate stilbene biosynthesis (Dubrovina and Kiselev, 2017). In 2020). However, ABA relation with JA-dependent related genes agreement with these results, JA and ET strongly trigger are closely linked with MYCs TF (Pieterse et al., 2014) but were phenylpropanoids pathway, notably stilbene biosynthesis not influenced by any of the treatments in the present study. (Belhadj et al., 2008; Rienth et al., 2019). This study also underlined that the expression of a The expression level of several genes related to Mitogen- Phenylalanine Ammonia Lyase (PAL) and a Chalcone Activated Protein Kinases (MAPKs) and Calcium ion (Ca2+) Synthase (CHS) related genes were upregulated after BL signaling pathways were slightly upregulated after Aki treatment treatment, whereas the expression of one Leucoanthocyanidin and some of them after BL treatment as well, such as Ca2+ dioxygenase (LAR) related gene was upregulated after both Aki Dependent Kinases (CDPKs), Calmodulin (CaM), Respiratory and BL treatments. PAL, CHS, LAR, and flavonol synthase (FS) Burst Oxydase Protein (RBOHF) and Heat Shock Transcription are key enzymes for biosynthesis of several secondary Factors (HS TFs) with an upregulation lower than Log2(FC) > 1.4. metabolites, such as phenylpropanoids, or phytoalexins In addition, one CaM and several peroxidases (PO) related genes isoflavonoids (Campos et al., 2003; Yonekura-Sakakibara et al., were clearly upregulated after Aki treatment, and two of them 2019). These enzymes are related to SA biosynthesis sharing were highly upregulated by Aki treatment (VIT_12s0055g01010; PAL enzyme as showed in the results. Stilbene biosynthesis was VIT_08s0058g00970) and involved in hypersensitive response Frontiers in Plant Science 18 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 TABLE 7 Phytohormone, organic acids, and total phenolic contents in leaves of grapevine cvs. Garnacha Blanca, Garnacha Tinta, and Macabeo treated with Akivi (Aki) and lyophilized Bacillus UdG (BL) and water (NTC). ‘Garnacha Blanca’ ‘Garnacha Tinta’ ‘Macabeo’ NTC Aki BL NTC Aki BL NTC Aki BL Phytohormone JA 4.39 ± 1.65 4.83 ± 1.21 6.15 ± 1.24 4.18 ± 1.74 5.54 ± 1.48 10.68 ± 0.72* 5.75 ± 0.73 6.95 ± 1.70 5.85 ± 2.25 MeJA 3.71 ± 0.26 4.95 ± 0.24* 5.12 ± 0.55* 4.98 ± 0.23 4.76 ± 0.16 4.31 ± 0.13 4.41 ± 0.34 4.13 ± 0.22 4.00 ± 0.33 SA 375 ± 78 553 ± 75 321 ± 68 215 ± 52 654* ± 147 375 ± 67 131 ± 17 218 ± 26 159 ± 34 ACC 8.86 ± 0.28 10.44 ± 0.48 10.65 ± 0.65 11.06 ± 0.65 10.23 ± 0.38 11.36 ± 0.33 12.69 ± 0.67 11.54 ± 1.13 12.13 ± 0.55 ABA1 1.51 ± 0.20 0.94 1.54 6.15 ± 0.99 7.37 ± 0.39 5.71 ± 0.60 21.25 ± 3.48 21.55 ± 1.05 17.96 ± 0.71 GA1&4 8.39 ± 1.07 31.05 ± 5.96* 18.29 ± 2,77 9.49 ± 2.47 6.02 ± 4.34 6.70 ± 0.29 27.34 ± 5.57 4.13* ± 1.60 7.85* ± 5.43 Organic acid Oxalic 4.47 ± 0.50 3.75 ± 0.95 3.28 ± 0.66 3.38 ± 0.32 3.54 ± 0.90 5.35 ± 1.59 2.33 ± 0.05 1.76 ± 0.42 2.80 ± 0.38 Tartaric 15.51 ± 0.65 15.23 ± 0.61 16.30 ± 0.52 19.45 ± 0.40 17.48 ± 0.20 17.20* ± 1.20 17.13 ± 1.20 16.43 ± 0.30 15.02 ± 0.55 Malic 1.48 ± 0.24 1.87 ± 0.38 1.78 ± 0.13 1.34 ± 0.20 2.02 ± 0.24 1.39 ± 0.14 0.92 ± 0.16 0.88 ± 0.18 1.18 ± 0.29 Oxoglutaric 346 ± 63 413 ± 96 631* ± 21 667 ± 21 674 ± 30 671 ± 21 155 ± 33 88* ±4 78* ± 11 Total phenolic 200 ± 25 359 ± 18 395* ± 61 876 ± 35 912 ± 51 808 ± 19 656 ± 52 679 ± 79 984* ± 46 JA: jasmonic acid; MeJA: methyl jasmonate; SA: salicylic acid; ACC: ethylene precursor; ABA: abscisic acid; and GA1&4: Gibberellins A1 and A4. 1 ABA concentration values in ‘Garnacha Blanca’ was at the limit of detection and only one value was detected for the treatment modalities (Aki and BL), thus, they were not included in statistical analysis. The following concentrations correspond to phytohormone (ng/g FW), organic acid (mg/g FW, except for oxoglutaric acid in μg/g FW), and total phenolic (μg gallic acid equivalent/g PF). Results are means ± standard deviation (n=3 biological replicates). Significant differences according to the Tukey test (parametric tests) or the Dunn test (non-parametric tests) between treatment (Aki or BL) and NTC are represented by asterisks (*). (HR). As no phytotoxicity was observed after the treatments, Aki In addition, Aki and BL treatments had an effect on other treatment may prime HR to be faster in case of pathogen metabolites, including the Oxoglutaric acid (2-OG) that showed infection. A crosstalk is described between MAPKs, JA, SA, and higher concentration in ‘Garnacha Blanca’ leaves after BL ET pathways (Rasmussen et al., 2012; Jagodzik et al., 2018; Nishad treatment and lower concentration in ‘Macabeo’ leaves after et al., 2020) and it was confirmed in this study since all these Aki and BL treatments. The 2-OG is involved in gibberellin pathways were upregulated by the treatments. (GA), alkaloid and flavonoid biosynthesis (Puhl et al., 2008; A B FIGURE 7 Transcriptional pattern of DEGs after treatments of grapevine cvs. Garnacha Blanca, Garnacha Tinta, and Macabeo with Akivi (A) or lyophilized Bacillus UdG (B). The fold change was assessed by the DDCt method. The UBQ gene was used as the internal control for data normalization. The DCt of the non-treated control (NTC) samples was defined as the calibrator. Three independent biological replicated were performed. Gene functions are indicated in Supplementary Table 2. Frontiers in Plant Science 19 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 Araú jo et al., 2012). Indeed, it was reported that treating the induction of some genes immediately after the treatment and grapevine with a structural mimic of 2-OG (prohexadione-Ca) the reinforcement of the modulation of defense response inhibit the enzyme and alter flavonoid biosynthesis (high through other genes after pathogen inoculation. The pathogen amount of unusual flavonoids) (Puhl et al., 2008). This is in infection may trigger biocontrol product activity and different agreement with the obtained results since phenylpropanoids grapevine response as it was observed using transcriptomics in pathway was triggered by both Aki and BL treatments. watermelon (Citrullus lanatus) roots treated with the beneficial Moreover, the GA content in leaves was also affected in the microorganism candidate B. velezensis against Fusarium present study, being GA1 and GA4 concentrations higher after oxysporum fungal pathogen (Jiang et al., 2019). More insights the treatments (Aki or BL) in ‘Garnacha Blanca’ and lower in on Aki and BL possible modes of action could be revealed ‘Macabeo’. However, the link between 2-OG and GA through another investigation introducing pathogen inoculation concentrations’ variations was not clear, that reinforce the in the study, such as P. viticola, E. necator, or B. cinerea and hypothesis linking the 2-OG with flavonoid biosynthesis. analyzing grapevine response to the treatment after pathogen The concentration of Tartaric acid in grapevine leaves was infection. This new research could be more accurate by doing a also affected since lower concentrations were detected in sampling kinetics to study the plant response to both treatments ‘Garnacha Tinta’ leaves treated with BL. Grapevine presents a and pathogen inoculation along time by transcriptomic and high concentration on tartaric acid and its biosynthesis occurs in metabolomics approaches (Jiang et al., 2019). leaves and berries. Tartaric acid was shown to be involved in From the twelve DEGs selected for Aki treatment eight are various processes and abrupt changes in its biosynthesis were related to plant defense (A1, A3, A4, A5, A6, A7, A11, and A12). linked with oxidative burst as well as ascorbate/glutathione Some of them (A4, A5, A6, and A11) are involved in the main redox state in berries (Burbidge et al., 2021). More insight on pathways related to plant defense response (Figure 8) and were this matter could give interesting results like a kinetics study of discussed above. From the fifteen DEGs selected for BL tartaric acid after BL treatment. treatment, four of them are related to plant defense (B1, B6, Grapevine response to Aki and BL treatments at transcripts B8, and B14). In addition, another gene could be related to and metabolic level indicate the ability of these products to trigger defense response (B12-VIT_10s0116g00530) as it is involved in plant defense response. In fact, many transcripts related to defense thiazole biosynthesis. Thiazole is a precursor of thiamine that responses were detected by the RNA-seq analysis of leaves. Some has been showed to be able to stimulate defense response of the transcripts did not present differential expression, but others (Boubakri et al., 2012; Boubakri et al., 2013). It is reported that were highly affected by Aki treatment (VIT_17s0000g00200, thiamine is able to induce resistance to downy mildew defense VIT_08s0058g00970, VIT_12s0055g01010, and response elicitation in leaves of ‘Chardonnay’ cultivated in VIT_03s0088g00810) and were selected as DEGs markers greenhouse-controlled conditions. The elicited defense candidates (A4, A5, A6, and A11, respectively). Considering all response included accumulation of hydrogen peroxide (H2O2), the upregulated transcripts in JA, ET, SA, and ABA pathways callose deposition in stomata cells, phenylpropanoid compounds (Figure 8) and the higher concentrations of some phytohormones; accumulation (stilbenes, phenolic compounds, flavonoids and the application of Aki and BL treatments to grapevine can lignin) and hypersensitive response. Thiamine triggered several stimulate several processes related to plant defense immune genes involved in defense response like PR genes (glucanase, system like SAR. Particularly, the treatments upregulated JA, chitinases, serine protease inhibitor, glutathione-S-transferase) ET, and phenylpropanoid pathways. Moreover, Aki treatment and lipoxygenases pathway involved in JA biosynthesis seemed to trigger several genes involved with HR. Further (Boubakri et al., 2012; Boubakri et al., 2013). The high rate of investigations are necessary to identify the mode of action of the DEGs highly impacted by the treatments and related to defense two biocontrol product candidates (Aki and BL). response is consistent with the transcripts analysis previously These results also indicate that the treatments with Aki and mentioned, and with the hypothesis that Aki and BL treatments the BL might prime a defense response through ISR. However, could be able to induce resistance on grapevine. the study was designed to investigate the interaction between the Several DEGs markers presented stable overexpression after biocontrol products and grapevine without pathogen infection. Aki (A1, A4 and A12 genes) and BL (B1, B2, B3, B9, B12, B13, If the mode of action is priming ISR, the effect could be seen only and B14 genes) treatments in the three grapevine cultivars. with the presence of the pathogen attack (Van Wees et al., 2008; Therefore, they could be considered as appropriate markers of Pieterse et al., 2014; Esmaeel et al., 2020). Actually, a complex Akivi and Bacillus treatments and could be used to test different effect acting in two steps was observed on various biocontrol doses and formulations of the biocontrol products in products, such as the Rheum palmatum plant extract (Godard greenhouse-controlled conditions. Actually, defining treatment et al., 2009), Trichoderma harzianum T39 (Perazzolli et al., dose and formulation are crucial steps in product development 2011), and sulphated laminarin (Trouvelot et al., 2008), being that highly impact its efficacy. As observed in the present study, this last one already used in vineyards against downy mildew. grapevine response is variable according to the studied cultivar These products show plant defense stimulation activity through as described other studies (Bota et al., 2016; Catacchio et al., Frontiers in Plant Science 20 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 FIGURE 8 Scheme of main pathways related to plant defense response: Jasmonic Acid (JA); Salicylic Acid (SA); Ethylene (ET); Abscisic Acid (ABA); phenylpropanoids pathway; and mitogen activated protein kinases, Ca2+ signalling induction (MAPKs). DEGs results are presented from RNA-seq analysis of grapevine leaves treated with the botanical extract (Aki, blue) and the microbial product (BF, yellow; or BL, orange). Complete DEGs transcript codes are written when the differential expression is above Log (FC) > 1.4; only VIT_ is written otherwise. DEGs highly impacted by one of the treatments are underlined. Gene groups from the different pathways are indicated, the box is white coloured when transcripts related to the genes’ groups were found, the box is grey coloured otherwise. JA and ET interactions with other pathways are represented with arrows. Black arrow represents JA and SA crosstalk. LOX, LipOXygenase; AOS, Allene Oxide Synthases; AOC, Allene Oxide Cyclase; OPR, OPDA Reductase; ACX, Acetyl-CoA oXidase; EDS1/NPR1, Enhanced Disease Susceptibility/Non-expressor of Pathogen Related genes 1; NIM1, Non- Inducible Immunity 1; SAR, Systemic Acquired Resistance; JAZ, JAsmonate-Zim domain; PR, Pathogenesis Related proteins; BHLH TFs, Helix Loop Helix TFs, WRKY TFs, Transcription Factors with domain WRKY. ICS/SID2, IsoChorismate Synthase; PAL, Phenylalanine Ammonia Lyase; PAD4, PhytoAlexin Deficient 4; FS, Flavonoid Synthase; LAR, LeucoAnthocyanicin Dioxygenase; GSTs, Glutatione-S-Tranferase; ISR, Induced Systemic Resistance; STS, STilbene Synthase; MyB TF, MyB Transcription Factors. ACS; 1-Aminocyclopropane-1-Carboxylate Synthase; ACO, 1- Aminocyclopropane-1-Carboxylate Oxidase; ERF TF, (AP2)/ERF TF/AP2TF; Ethylene Response Factors Transcription Factors; PR, Pathogenesis Related proteins. NCED, 9-Cis-Epoxycarotenoid Dioxygenase. MAPKs, Mitogen Activated Protein Kinases; CDPKs, Ca2+ DePendent Kinases; CaM, CalModulin; RBOHF, Respiratory Burst Oxidase Homologue protein F; PO, PerOxidases; HS TF, Heat Shock Transcription Factors. 2019; Fasoli et al., 2019; Balestrini et al., 2020). Therefore, the complemented with metabolic analysis (phytohormones, identified markers are only robust for the three tested cultivars phenols, and organic acids). Considering all the upregulated and should be tested on other cultivars to extend their use. The transcripts and enhanced metabolites concentrations related to markers could also be tested in field conditions, but it could be JA, ET, and phenylpropanoids pathways, strong indication was difficult to detect an impact on transcriptome in field conditions found of grapevine defense induction by the treatments. due to vineyard biological variability (Balestrini et al., 2020). However, further studies are necessary to confirm these first results and a kinetics study could be interesting. In addition, several DEGs markers were identified presenting a stable 5 Conclusion overexpression after the treatments (Aki or BL) in the three grapevine cultivars. They could be used as markers of activity of Grapevine response to the Aki and BL treatments at the products for further investigations. transcripts and metabolites levels gave insights on modes of action of these biocontrol products that are under development. RNA sequencing analysis showed different gene expression Data availability statement patterns after foliar treatments with the biocontrol products compared with the NTC in ‘Garnacha Blanca’. Furthermore, The original contributions presented in the study are RT-qPCR enabled the quantification of several selected genes publicly available. This data can be found here: https://www. (DEG) in three different cultivars. This information was ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE211268. Frontiers in Plant Science 21 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 Author contributions and Rosa Padilla for her help in the laboratory. We are grateful to Anna Bonaterra for her help and expertise in cultivating Bacillus ML, EM, and EB obtained the financial support. MR, ND, UdG strain and to S.A.S. AkiNaO for providing Akivi biocontrol ML, EM, and EB conceived and designed the research. MR, ND, product. We are also thankful to Anna Bonaterra and Luis Moll ML, RT and EB conducted and performed the experiments. MR, for their help during the grapevine leaves samplings. ML, and EB analyzed the data. MR and ND wrote the first draft of the manuscript. ML, EM, and EB revised and edited the manuscript. All authors contributed to the article and approved Conflict of interest the submitted version. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could Funding be construed as a potential conflict of interest. This work was supported by PALVIP (EFA182/16) project that has been 65% co-financed by the European Regional Publisher’s note Development Fund (ERDF) through the Interreg V-A Spain France Andorra programme (POCTEFA 2014-2020). All claims expressed in this article are solely those of the POCTEFA aims to reinforce the economic and social authors and do not necessarily represent those of their affiliated integration of the French-Spanish-Andorran borders. Its organizations, or those of the publisher, the editors and the support is focused on developing economic, social and reviewers. Any product that may be evaluated in this article, or environmental cross-border activities through joint strategies claim that may be made by its manufacturer, is not guaranteed favouring sustainable territorial development. or endorsed by the publisher. Acknowledgments Supplementary material The authors would like to thank Gemma Roselló , Héctor The Supplementary Material for this article can be found Naú n Saravia, and Josep Pereda for the technical support in the online at: https://www.frontiersin.org/articles/10.3389/ greenhouse experiments and for growing up the grapevine stocks, fpls.2022.1075231/full#supplementary-material References Abdul Malik, N. A., Kumar, I. S., and Nadarajah, K. (2020). Elicitor and receptor of PR proteins, stilbene and anthocyanin accumulation in grapevine cell cultures. molecules: Orchestrators of plant defense and immunity. Int. J. Mol. Sci. 21, 963. Plant Physiol. Biochem. 46, 493–499. doi: 10.1016/j.plaphy.2007.12.001 doi: 10.3390/ijms21030963 Beris, E., Totkas, M., and Stathaki, M. (2021). Signaling pathways of conidial Alavanja, M. C. R., Hoppin, J. A., and Kamel, F. (2004). Health effects of chronic germination and growth of Botrytis cinerea: Host detection, pathogenesis on Vitis pesticide exposure: Cancer and neurotoxicity. Annu. Rev. Public Health 25, 155– vinifera and preference for wine grapes. Plant Prot. 5, 49–57. doi: 10.33804/ 197. doi: 10.1146/annurev.publhealth.25.101802.123020 pp.005.01.3582 Araú jo, W. L., Tohge, T., Osorio, S., Lohse, M., Balbo, I., Krahnert, I., et al. Bodin, E., Bellé e, A., Dufour, M.-C., André , O., and Corio-Costet, M.-F. (2020). (2012). Antisense inhibition of the 2-oxoglutarate dehydrogenase complex in Grapevine stimulation: A multidisciplinary approach to investigate the effects of tomato demonstrates its importance for plant respiration and during leaf biostimulants and a plant defense stimulator. J. Agric. Food. Chem. 68, 15085– senescence and fruit maturation. Plant Cell 24, 2328–2351. doi: 10.1105/ 15096. doi: 10.1021/acs.jafc.0c05849 tpc.112.099002 Bonaterra, A., Badosa, E., Cabrefiga, J., Francé s, J., and Montesinos, E. (2012). Avenard, J. C., Bernos, L., Grand, O., and Samir, B. (2003). Manuel De Prospects and limitations of microbial pesticides for control of bacterial and fungal production inté grée en viticulture (Bordeaux: É ditions Fé ret). pomefruit tree diseases. Trees (Berl West) 26, 215–226. doi: 10.1007/s00468-011- Backer, R., Naidoo, S., and van den Berg, N. (2019). The nonexpressor of 0626-y pathogenesis-related genes 1 (NPR1) and related family: Mechanistic insights in Bota, J., Tomàs, M., Flexas, J., Medrano, H., and Escalona, J. (2016). Differences plant disease resistance. Front. Plant Sci. 10. doi: 10.3389/fpls.2019.00102 among grapevine cultivars in their stomatal behaviour and water use efficiency Bahieldin, A., Atef, A., Edris, S., Gadalla, N. O., Ali, H. M., Hassan, S. M., et al. under progressive water stress. Agric. Water Manage 164, 91–99. doi: 10.1016/ (2016). Ethylene responsive transcription factor ERF109 retards PCD and j.agwat.2015.07.016 improves salt tolerance in plant. BMC Plant Biol. 16, 216. doi: 10.1186/s12870- Boubakri, H., Poutaraud, A., Wahab, M. A., Clayeux, C., Baltenweck-Guyot, R., 016-0908-z Steyer, D., et al. (2013). Thiamine modulates metabolism of the phenylpropanoid Balestrini, R., Ghignone, S., Quiroga, G., Fiorilli, V., Romano, I., and Gambino, pathway leading to enhanced resistance to Plasmopara viticola in grapevine. BMC G. (2020). Long-term impact of chemical and alternative fungicides applied to Plant Biol. 13, 31. doi: 10.1186/1471-2229-13-31 grapevine cv nebbiolo on berry transcriptome. Int. J. Mol. Sci. 2020, 21, 6067. Boubakri, H., Wahab, M. A., Chong, J., Bertsch, C., Mliki, A., and Soustre- doi: 10.3390/ijms21176067 Gacougnolle, I. (2012). Thiamine induced resistance to Plasmopara viticola in Belhadj, A., Telef, N., Saigne, C., Cluzet, S., Barrieu, F., Hamdi, S., et al. (2008). grapevine and elicited host–defense responses, including HR like-cell death. Plant Effect of methyl jasmonate in combination with carbohydrates on gene expression Physiol. Biochem. 57, 120–133. doi: 10.1016/j.plaphy.2012.05.016 Frontiers in Plant Science 22 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 Burbidge, C. A., Ford, C. M., Melino, V. J., Wong, D. C. J., Jia, Y., Jenkins, C. L. Kazan, K., and Manners, J. M. (2012). JAZ repressors and the orchestration of D., et al. (2021). Biosynthesis and cellular functions of tartaric acid in grapevines. phytohormone crosstalk. Trends Plant Sci. 17, 22–31. doi: 10.1016/ Front. Plant Sci. 12. doi: 10.3389/fpls.2021.643024 j.tplants.2011.10.006 Burdziej, A., Bellé e, A., Bodin, E., Valls Fonayet, J., Magnin, N., Szakiel, A., et al. Kidd, P. S., Llugany, M., Poschenrieder, C., Gunsé , B., and Barceló , J. (2001). The (2021). Three types of elicitors induce grapevine resistance against downy mildew role of root exudates in aluminium resistance and silicon-induced amelioration of via common and specific immune responses. J. Agric. Food Chem. 69, 1781–1795. aluminium toxicity in three varieties of maize (Zea mays l.). J. Exp. Bot. 52, 1339– doi: 10.1021/acs.jafc.0c06103 1352. doi: 10.1093/jexbot/52.359.1339 Butault, J. P., Dedryver, C. A., Gary, C., Guichard, L., Jacquet, F., Meynard, J. M., Krzyzaniak, Y., Trouvelot, S., Negrel, J., Cluzet, S., Valls, J., Richard, T., et al. et al. (2010). É cophyto R&D: quelles voies pour réduire l’usage des pesticides? (2018). A plant extract acts both as a resistance inducer and an oomycide against synthèse du rapport de l'é tude (Paris: INRA Editions). doi: 10.15454/r7ae-b824 grapevine downy mildew. Front. Plant Sci. 9. doi: 10.3389/fpls.2018.01085 Calonnec, A., Cartolaro, P., Delière, L., and Chadoeuf, J. (2006). Powdery Kunova, A., Pizzatti, C., Saracchi, M., Pasquali, M., and Cortesi, P. (2021). mildew on grapevine: the date of primary contamination affects disease Grapevine powdery mildew: Fungicides for its management and advances in development on leaves and damage on grape. IOBC WPRS Bull. 29, 67–73. molecular detection of markers associated with resistance. Microorganisms 9, Campos, Â .D., Ferreira, A. G., Hampe, M. M. V., Antunes, I. F., Brancão, N., 1541. doi: 10.3390/microorganisms9071541 Silveira, E. P., et al. (2003). Induction of chalcone synthase and phenylalanine Leroy, P., Smits, N., Cartolaro, P., Deliere, L., Goutouly, J.-P., Raynal, M., et al. ammonia-lyase by salicylic acid and Colletotrichum lindemuthianum in common (2013). A bioeconomic model of downy mildew damage on grapevine for bean. Braz. J. Plant Physiol. 15, 129–134. doi: 10.1590/S1677-04202003000300001 evaluation of control strategies. Crop Prot. 53, 58–71. doi: 10.1016/ Catacchio, C. R., Alagna, F., Perniola, R., Bergamini, C., Rotunno, S., Calabrese, j.cropro.2013.05.024 F. M., et al. (2019). Transcriptomic and genomic structural variation analyses on Livak, K. J., and Schmittgen, T. D. (2001). Analysis of relative gene expression grape cultivars reveal new insights into the genotype-dependent responses to water data using real-time quantitative PCR and the 2–DDCT method. Methods 25, 402– stress. Sci. Rep. 9, 2809. doi: 10.1038/s41598-019-39010-x 408. doi: 10.1006/meth.2001.1262 Chen, H., Li, M., Qi, G., Zhao, M., Liu, L., Zhang, J., et al. (2021). Two interacting Llugany, M., Martin, S. R., Barceló , J., and Poschenrieder, C. (2013). Endogenous transcriptional coactivators cooperatively control plant immune responses. Sci. jasmonic and salicylic acids levels in the cd-hyperaccumulator noccaea (Thlaspi) Adv. 7, 45. doi: 10.1126/sciadv.abl7173 praecox exposed to fungal infection and/or mechanical stress. Plant Cell Rep. 32, De Miccolis Angelini, R. M., Rotolo, C., Gerin, D., Abate, D., Pollastro, S., and 1243–1249. doi: 10.1007/s00299-013-1427-0 Faretra, F. (2019). Global transcriptome analysis and differentially expressed genes Maachia, B., Rafik, E., Ché rif, M., Nandal, P., Mohapatra, T., and Bernard, P. in grapevine after application of the yeast-derived defense inducer cerevisane. Pest (2015). Biological control of the grapevine diseases “grey mold” and “powdery Manage. Sci. 75, 2020–2033. doi: 10.1002/ps.5317 mildew” by Bacillus B27 and B29 strains. Indian J. Exp. Biol. 53, 109–115. Dubrovina, A. S., and Kiselev, K. V. (2017). Regulation of stilbene biosynthesis in Mirá s-Avalos, J. M., and Araujo, E. S. (2021). Optimization of vineyard water plants. Planta 246, 597–623. doi: 10.1007/s00425-017-2730-8 management: Challenges, strategies, and perspectives. Water 13, 746. doi: 10.3390/ Esmaeel, Q., Jacquard, C., Sanchez, L., Clé ment, C., and Ait Barka, E. (2020). The w13060746 mode of action of plant associated Burkholderia against grey mould disease in Mizoi, J., Shinozaki, K., and Yamaguchi-Shinozaki, K. (2012). AP2/ERF family grapevine revealed through traits and genomic analyses. Sci. Rep. 10, 19393. transcription factors in plant abiotic stress responses. Biochim. Biophys. Acta Gene doi: 10.1038/s41598-020-76483-7 Regul. Mech. 1819, 86–96. doi: 10.1016/j.bbagrm.2011.08.004 European Commission (2021) Agricultural production – crops. Available at: Mora, I., Cabrefiga, J., and Montesinos, E. (2011). Antimicrobial peptide genes in https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Agricultural_ Bacillus strains from plant environments. Int. Microbiol. 14, 213–223. doi: 10.2436/ production_-_crops (Accessed July 12, 2021). 20.1501.01.151 European Parliament, Council of the European Union (2009a). Directive 2009/ Mora, I., Cabrefiga, J., and Montesinos, E. (2015). Cyclic lipopeptide 128/EC of the European parliament and of the council of 21 October 2009 biosynthetic genes and products, and inhibitory activity of plant-associated establishing a framework for community action to achieve the sustainable use of Bacillus against phytopathogenic bacteria. PloS One 10, e0127738. doi: 10.1371/ pesticides. Off. J. Eur. Union 309, 71–86. journal.pone.0127738 European Parliament, Council of the European Union (2009b). Regulation (EC) Nishad, R., Ahmed, T., Rahman, V. J., and Kareem, A. (2020). Modulation of no 1107/2009 of the European parliament and of the council of 21 October 2009 plant defense system in response to microbial interactions. Front. Microbiol. 11. concerning the placing of plant protection products on the market and repealing doi: 10.3389/fmicb.2020.01298 council directives 79/117/EEC and 91/414/EEC. Off. J. Eur. Union 309, 1–50. Ochsenbein, C., Przybyla, D., Danon, A., Landgraf, F., Göbel, C., Imboden, A., Fasoli, M., Dell’Anna, R., Amato, A., Balestrini, R., Dal Santo, S., Monti, F., et al. et al. (2006). The role of EDS1 (enhanced disease susceptibility) during singlet (2019). Active rearrangements in the cell wall follow polymer concentration during oxygen-mediated stress responses of Arabidopsis. Plant J. 47, 445–456. postharvest withering in the berry skin of Vitis vinifera cv. Corvina. Plant Physiol. doi: 10.1111/j.1365-313X.2006.02793.x Biochem. 135, 411–422. doi: 10.1016/j.plaphy.2018.11.020 Ongena, M., and Jacques, P. (2008). Bacillus lipopeptides: versatile weapons for Gessler, C., Pertot, I., and Perazzolli, M. (2011). Plasmopara viticola: a review of plant disease biocontrol. Trends Microbiol. 16, 115–125. doi: 10.1016/ knowledge on downy mildew of grapevine and effective disease management. j.tim.2007.12.009 Phytopathol. Mediterr. 50, 3–44. doi: 10.14601/Phytopathol_Mediterr-9360 Otoguro, M., and Suzuki, S. (2018). Status and future of disease protection and Godard, S., Slacanin, I., Viret, O., and Gindro, K. (2009). Induction of defence grape berry quality alteration by micro-organisms in viticulture. Lett. Appl. mechanisms in grapevine leaves by emodin- and anthraquinone-rich plant extracts Microbiol. 67, 106–112. doi: 10.1111/lam.13033 and their conferred resistance to downy mildew. Plant Physiol. Biochem. 47, 827– Pajerowska-Mukhtar, K. M., Emerine, D. K., and Mukhtar, M. S. (2013). Tell me 837. doi: 10.1016/j.plaphy.2009.04.003 more: roles of NPRs in plant immunity. Trends Plant Sci. 18, 402–411. doi: 10.1016/ Guerreiro, A., Figueiredo, J., Sousa Silva, M., and Figueiredo, A. (2016). Linking j.tplants.2013.04.004 jasmonic acid to grapevine resistance against the biotrophic oomycete Plasmopara Perazzolli, M., Moretto, M., Fontana, P., Ferrarini, A., Velasco, R., Moser, C., viticola. front. Plant Sci. 7. doi: 10.3389/fpls.2016.00565 et al. (2012). Downy mildew resistance induced by Trichoderma harzianum T39 in Gullner, G., Komives, T., Kirá ly, L., and Schröder, P. (2018). Glutathione s- susceptible grapevines partially mimics transcriptional changes of resistant transferase enzymes in plant-pathogen interactions. Front. Plant Sci. 9. genotypes. BMC Genomics 13, 660. doi: 10.1186/1471-2164-13-660 doi: 10.3389/fpls.2018.01836 Perazzolli, M., Roatti, B., Bozza, E., and Pertot, I. (2011). Trichoderma Jagodzik, P., Tajdel-Zielinska, M., Ciesla, A., Marczak, M., and Ludwikow, A. harzianum T39 induces resistance against downy mildew by priming for defense (2018). Mitogen-activated protein kinase cascades in plant hormone signaling. without costs for grapevine. Biol. Control. 58, 74–82. doi: 10.1016/ Front. Plant Sci. 9. doi: 10.3389/fpls.2018.01387 j.biocontrol.2011.04.006 Jiang, C.-H., Yao, X.-F., Mi, D.-D., Li, Z.-J., Yang, B.-Y., Zheng, Y., et al. (2019). Persaud, R., Khan, A., Isaac, W.-A., Ganpat, W., and Saravanakumar, D. (2019). Comparative transcriptome analysis reveals the biocontrol mechanism of Bacillus Plant extracts, bioagents and new generation fungicides in the control of rice sheath velezensis F21 against Fusarium wilt on watermelon. Front. Microbiol. 10. blight in Guyana. Crop Prot. 119, 30–37. doi: 10.1016/j.cropro.2019.01.008 doi: 10.3389/fmicb.2019.00652 Pertot, I., Caffi, T., Rossi, V., Mugnai, L., Hoffmann, C., Grando, M. S., et al. Kanehisa, M., Sato, Y., Kawashima, M., Furumichi, M., and Tanabe, M. (2016). (2017). A critical review of plant protection tools for reducing pesticide use on KEGG as a reference resource for gene and protein annotation. Nucleic Acids Res. grapevine and new perspectives for the implementation of IPM in viticulture. Crop 44, D457–D462. doi: 10.1093/nar/gkv1070 Prot. 97, 70–84. doi: 10.1016/j.cropro.2016.11.025 Frontiers in Plant Science 23 frontiersin.org Ramos et al. 10.3389/fpls.2022.1075231 Pfaffl, M. W., Horgan, G. W., and Dempfle, L. (2002). Relative expression leaves. Plant Physiol. Biochem. 37, 491–496. doi: 10.1016/S0981-9428(99) software tool (REST©) for group-wise comparison and statistical analysis of 80054-0 relative expression results in real-time PCR. Nucleic Acids Res. 30, e36. Tamm, L., Schaerer, H. J., Levert, A., Andreu, V., and Bertrand, C. (2017). doi: 10.1093/nar/30.9.e36 “Agent antifongique, procé dé et composition,” in France Patent no FR 3046907A1. Pieterse, C. M. J., Zamioudis, C., Berendsen, R. L., Weller, D. M., Van Wees, S. C. (Courbevoie: Intitut National de la Propriété industrialle). M., and Bakker, P. A. H. M. (2014). Induced systemic resistance by beneficial microbes. Annu. Rev. Phytopathol. 52, 347–375. doi: 10.1146/annurev-phyto- Tolrà, R. P., Poschenrieder, C., Luppi, B., and Barceló , J. (2005). Aluminium- 082712-102340 induced changes in the profiles of both organic acids and phenolic substances underlie Al tolerance in Rumex acetosa l. Environ. Exp. Bot. 54, 231–238. Postiglione, A. E., and Muday, G. K. (2020). The role of ROS homeostasis in doi: 10.1016/j.envexpbot.2004.07.006 ABA-induced guard cell signaling. Front. Plant Sci. 11. doi: 10.3389/ fpls.2020.00968 Trouvelot, S., Varnier, A.-L., Allègre, M., Mercier, L., Baillieul, F., Arnould, C., et al. (2008). A beta-1,3 glucan sulfate induces resistance in grapevine against Puhl, I., Stadler, F., and Treutter, D. (2008). Alterations of flavonoid biosynthesis Plasmopara viticola through priming of defense responses, including HR-like in young grapevine (Vitis vinifera l.) leaves, flowers, and berries induced by the cell death. Mol. Plant Microbe Interact. 21, 232–243. doi: 10.1094/MPMI-21- dioxygenase inhibitor prohexadione-Ca. J. Agric. Food Chem. 56, 2498–2504. 2-0232 doi: 10.1021/jf0727645 Van Aubel, G., Buonatesta, R., and Van Cutsem, P. (2014). COS-OGA: A novel Ramos, M., Ghosson, H., Raviglione, D., Bertrand, C., and Salvia, M. V. (2022). oligosaccharidic. Crop Protection. 65, 129–137. doi: 10.1016/j.cropro.2014.07.015 Untargeted metabolomics as a tool to monitor biocontrol product residues' fate on field-treated Prunus persica. sci. Total Environ. 807 (Pt 1), 150717. doi: 10.1016/ Van Wees, S. C. M., van der Ent, S., and Pieterse, C. M. J. (2008). Plant immune j.scitotenv.2021.150717 responses triggered by beneficial microbes. Curr. Opin. Plant Biol. 11, 443–448. doi: 10.1016/j.pbi.2008.05.005 Rasmussen, M., Roux, M., Petersen, M., and Mundy, J. (2012). MAP kinase cascades in arabidopsis innate immunity. Front. Plant Sci. 3. doi: 10.3389/ Vilà, S., Badosa, E., Montesinos, E., Planas, M., and Feliu, L. (2016). Synthetic fpls.2012.00169 cyclolipopeptides selective against microbial, plant and animal cell targets by incorporation of d-amino acids or histidine. PloS One 11, e0151639. Reynier, A. (2011). Manuel De viticulture (11 ed.) (Paris: Lavoisier). doi: 10.1371/journal.pone.0151639 Rienth, M., Crovadore, J., Ghaffari, S., and Lefort, F. (2019). Oregano essential oil Walker, J. E., and Abraham, E. P. (1970). Isolation of bacilysin and a new amino vapour prevents Plasmopara viticola infection in grapevine (Vitis vinifera) and acid from culture filtrates of bacillus subtilis. Biochem. J. 118, 557–561. doi: primes plant immunity mechanisms. PloS One 14, e0222854. doi: 10.1371/ 10.1042/bj1180557 journal.pone.0222854 Wang, B., Yuan, J., Zhang, J., Shen, Z., Zhang, M., Li, R., et al. (2013). Effects of Rusté rucci, C., Aviv, D. H., Holt, B. F., Dangl, J. L., and Parker, J. E. (2001). The novel bioorganic fertilizer produced by Bacillus amyloliquefaciens W19 on disease resistance signaling components EDS1 and PAD4 are essential regulators of antagonism of Fusarium wilt of banana. Biol. Fertil. Soils. 49, 435–446. the cell death pathway controlled by LSD1 in Arabidopsis. Plant Cell 13, 2211–2224. doi: 10.1007/s00374-012-0739-5 doi: 10.1105/tpc.010085 Xiong, L., and Zhu, J.-K. (2003). Regulation of abscisic acid biosynthesis. Plant Schnee, S., Queiroz, E. F., Voinesco, F., Marcourt, L., Dubuis, P.-H., Wolfender, Physiol. 133, 29–36. doi: 10.1104/pp.103.025395 J.-L., et al. (2013). Vitis vinifera canes, a new source of antifungal compounds against Plasmopara viticola, Erysiphe necator, and Botrytis cinerea. J. Agric. Food Yonekura-Sakakibara, K., Higashi, Y., and Nakabayashi, R. (2019). The origin Chem. 61, 5459–5467. doi: 10.1021/jf4010252 and evolution of plant flavonoid metabolism. Front. Plant Sci. 10. doi: 10.3389/ fpls.2019.00943 Segarra, G., Já uregui, O., Casanova, E., and Trillas, I. (2006). Simultaneous quantitative LC-ESI-MS/MS analyses of salicylic acid and jasmonic acid in crude Zambito Marsala, R., Capri, E., Russo, E., Bisagni, M., Colla, R., Lucini, L., et al. extracts of Cucumis sativus under biotic stress. Phytochemistry. 67, 395–401. (2020). First evaluation of pesticides occurrence in groundwater of tidone valley, an doi: 10.1016/j.phytochem.2005.11.017 area with intensive viticulture. Sci. Total. Environ. 736, 139730. doi: 10.1016/ Silver, N., Best, S., Jiang, J., and Thein, S. L. (2006). Selection of housekeeping j.scitotenv.2020.139730 genes for gene expression studies in human reticulocytes using real-time PCR. Zhou, Q., Fu, M., Xu, M., Chen, X., Qiu, J., Wang, F., et al. (2020). Application BMC Mol. Biol. 7, 33. doi: 10.1186/1471-2199-7-33 of antagonist Bacillus amyloliquefaciens NCPSJ7 against Botrytis cinerea in Singleton, V. L., Orthofer, R., and Lamuela-Raventó s, R. (1999). Analysis of total postharvest red globe grapes. Food Sci. Nutr. 8, 1499–1508. doi: 10.1002/ phenols and other oxidation substrates and antioxidants by means of folin- fsn3.1434 ciocalteu reagent. Methods Enzymology 299, 152–178. doi: 10.1016/S0076-6879 Zhu, S., Jeong, R.-D., Venugopal, S. C., Lapchyk, L., Navarre, D., Kachroo, A., (99)99017-1 et al. (2011). SAG101 forms a ternary complex with EDS1 and PAD4 and is Solecka, D., Boudet, A.-M., and Kacperska, A. (1999). Phenylpropanoid required for resistance signaling against turnip crinkle virus. PloS Pathog. 7, and anthocyanin changes in low-temperature treated winter oilseed rape e1002318. doi: 10.1371/journal.ppat.1002318 Frontiers in Plant Science 24 frontiersin.org