NoPv1: a synthetic antimicrobial peptide aptamer targeting the causal agents of grapevine downy mildew and potato late blight

Grapevine (Vitis vinifera L.) is a crop of major economic importance. However, grapevine yield is guaranteed by the massive use of pesticides to counteract pathogen infections. Under temperate-humid climate conditions, downy mildew is a primary threat for viticulture. Downy mildew is caused by the biotrophic oomycete Plasmopara viticola Berl. & de Toni, which can attack grapevine green tissues. In lack of treatments and with favourable weather conditions, downy mildew can devastate up to 75% of grape cultivation in one season and weaken newly born shoots, causing serious economic losses. Nevertheless, the repeated and massive use of some fungicides can lead to environmental pollution, negative impact on non-targeted organisms, development of resistance, residual toxicity and can foster human health concerns. In this manuscript, we provide an innovative approach to obtain specific pathogen protection for plants. By using the yeast two-hybrid approach and the P. viticola cellulose synthase 2 (PvCesA2), as target enzyme, we screened a combinatorial 8 amino acid peptide library with the aim to identify interacting peptides, potentially able to inhibit PvCesa2. Here, we demonstrate that the NoPv1 peptide aptamer prevents P. viticola germ tube formation and grapevine leaf infection without affecting the growth of non-target organisms and without being toxic for human cells. Furthermore, NoPv1 is also able to counteract Phytophthora infestans growth, the causal agent of late blight in potato and tomato, possibly as a consequence of the high amino acid sequence similarity between P. viticola and P. infestans cellulose synthase enzymes.


Scientific Reports |
(2020) 10:17574 | https://doi.org/10.1038/s41598-020-73027-x www.nature.com/scientificreports/ that have been associated with the massive use of chemical pesticides 12 . Residues of pesticides can be found in foods and beverages 13,14 , although in most of the cases the concentrations do not exceed the safe levels 15 . However, the simultaneous exposure to different compounds may have negative synergistic effects 16 17 . Taken these considerations together, it appears clear that new biotechnological sustainable solutions need to be explored to find safe and reliable alternatives to conventional pesticides that specifically inhibit the activity of pathogen key enzymes. In recent years, peptide aptamers-i.e. short synthetic peptides able to specifically bind and inhibit a protein target-have emerged as novel molecular tools that have attracted the attention of different research groups interested in developing antimicrobial compounds 18,19 , presumably with a better environmental fate, lower off-target effects and possibly a low-risk alternative to conventional pesticides, although this technology requires experimental validation.
Among crops, grapevine represents a great agricultural and economic value worldwide, with 7.534,00 Mha invested in viticulture 20 . However, the grapevine industry relies predominantly on Vitis vinifera, which is susceptible to a large spectrum of pathogens and requires a frequent application of chemical antimicrobials to avoid yield and quality loss 21 . In particular, under temperate-humid climate conditions downy mildew, a worldwide destructive disease caused by the biotrophic oomycete Plasmopara viticola Berl. & De Toni 22 , is considered of primary importance for viticulture 20 . In lack of treatments and with favourable weather conditions, downy mildew can devastate up to 75% of the crop in one season and weaken newly emerging shoots, causing serious economic losses 23,24 . P. viticola attacks grapevine green organs, such as leaves and young fruits, by releasing flagellate zoospores at high humidity and warm temperatures 25 . When the zoospores encounter a stoma, they shed the flagella and encyst. Subsequently, a germ tube emerges from each spore and reaches the substomatal cavity, where it dilates into a vesicle that initiates the infection 26 .
Data available for the oomycete Phytophthora infestans, the causal agent of late blight in potato and tomato 23 , provide useful information on the molecular mechanisms at the basis of oomycete plant infection. The main component of P. infestans cell wall is cellulose, accounting for the 33.6% over the total glucan composition (85.6%) 27 . Cellulose appears to be essential for appressorium formation and effective potato infection by P. infestans, since inhibition of its biosynthesis leads to a dramatic reduction in the number of normal germ tubes with appressoria, severe disruption of the cell wall in the pre-infection structures, and a complete loss of pathogenicity 28 . Furthermore, cellulose synthase gene expression is up-regulated during pre-and early-infection stages of P. infestans in potato 28 . Additionally, P. infestans and P. viticola cellulose synthase enzymes have been identified as the target of the fungicide mandipropamid, highlighting the relevance of cell wall biosynthesis in oomycete disease development 29,30 .
In this manuscript, we describe the identification of a novel peptide aptamer of 8 amino acid residues, named NoPv1 (No Plasmopora viticola 1), able to counteract P. viticola infection ex vivo (leaf disks) and in vivo (potted plants). In particular, the PvCesA2 (P. viticola cellulose synthase 2) enzyme was used as a bait in the yeast twohybrid assay to isolate the NoPv1 peptide, from a combinatorial peptide library 31 . NoPv1 prevents P. viticola germ tube formation and grapevine infection without affecting the growth of off-target organisms, and without being toxic for plant and human cells. Furthermore, NoPv1 is also able to arrest the growth of P. infestans probably due to the high amino acid sequence similarity of its cellulose synthase enzymes with the ones of P. viticola.

Results
Identification of peptide aptamers interacting with the Plasmopara viticola cellulose synthase 2, PvCesA2. The PvCesA2 cellulose synthase enzyme has been selected as target enzyme for our study since cellulose biosynthesis plays a pivotal role during pathogen infection, as demonstrated by Grenville-Briggs and co-workers 28 in the closely related oomycete species, P. infestans. Furthermore, phylogenetic analysis performed by comparing cellulose synthase amino acid sequences from different organisms (see Fig. 1, Table S1 and Data S1), highlighted that oomycete cellulose synthase proteins group together as a distinct clade, separate from bacterial, cyanobacterial, viral, fungal and plant cellulose synthases 28,30 . In addition, it is also interesting to observe that among the four Oomycota CesAs, CesA3 enzymes cluster in a slightly different clade from CesA1, CesA2 and CesA4, probably due to the lack of a Plekstrin Homology domain (PH domain), involved in regulation, trafficking and/or targeting of polysaccharide synthases to the correct site for cell wall biosynthesis 32 .
These notions, together with the fact that CesA2 is the most abundant among the PH domain-containing CesA enzymes 28 , and that cellulose synthase enzymes are absent in many beneficial organisms, such as Saccharomyces cerevisiae, and in animal cells, including human cells, make PvCesA2 the ideal target for developing antimicrobial active compounds.
Peptide aptamers able to specifically interact with PvCesA2 were identified by using the yeast two-hybrid strategy. The cytoplasmic soluble portion of PvCesA2 (from aa 331 to 790; see also Data S1), containing all typical signatures of most processive glycosyltransferases, including three aspartic acid residues and the QXXRW motif 28,33 , was fused to the GAL4 DNA Binding Domain (BD) and used as bait to screen a combinatorial library of linear peptide aptamers (8 a.a. in length; see also "Materials and methods" for further details). Ten small scale transformations were performed (⁓ 10 6 transformants obtained) and only 30 colonies were able to grow on selective media, lacking either adenine or histidine, and supplemented with 10 mM of 3-amino-1,2,4-triazole (3-AT), a competitive inhibitor for the HIS3 reporter gene. To identify the peptide aptamers interacting with PvCesA2, plasmids were purified and sequenced, and priority was given to peptides whose sequences appeared Phylogenetic analyses were conducted using MEGA version X 66 (see Data S1 for Clustal omega multiple sequence alignment). The generated dendrogram demonstrates that Oomycota CesAs group together in a separate clade from all the other considered cellulose synthase and cellulose synthase-like sequences. The analysis is based on complete amino acid sequences reported in Table S1 Table 1), were prepared via Microwave-assisted Solid Phase Peptide Synthesis (see "Materials and methods" for further details) and analysed for their capability to counteract P. viticola infection of grapevine leaf disks through co-inoculation assays. In details, leaf disks of V. vinifera cv. Pinot noir, susceptible to P. viticola infection, were treated with sporangia suspension in either absence or presence of 200 μM NoPv1, NoPv2 and NoPv3 (Fig. 2a-f). P. viticola sporulation was clearly visible on grapevine leaf disks in the absence of NoPv peptides (Fig. 2a,c,g; Control) at 5 and 7 days post-inoculation (dpi), and NoPv2 and NoPv3 peptide aptamers showed intermediate and no effect on P. viticola severity, respectively (Fig. 2e,f,g). On the contrary, 200 μM NoPv1 peptide was able to completely inhibit P. viticola severity on leaf disks (Fig. 2b,d,g), without causing any damage to the leaf tissues, as demonstrated by very similar values in the maximum quantum yield of photosystem II (F V /F M ) and effective quantum yield of photosystem II [Y (II) ] of young developing grapevine leaves, treated or not with 400 μM and 1 mM NoPv1, measured using the IMAGING-PAM (Walz, Germany) to evaluate their photosynthetic performance ( Fig. 3 and Table S2). Table 1. List of peptide aptamers identified by yeast two-hybrid assay using PvCesA2 as bait and mutated derivatives of NoPv1. Nucleotide and amino acid sequences (8 a.a.) are provided. Note that sequences of peptides identified at least three times among the 30 colonies isolated are provided.

Aptamer name
Nucleotide sequence Peptide sequence   Table 1, NoPv1 is a peptide of 8 l-amino acid residues with the following sequence: RLTAQCRL. Its molecular weight is of 960.16 Da, the isoelectric point is at pH 10.43, it has a net charge of + 1.9 at pH 7 and a good water solubility (pepcalc.com). Interestingly, the net positive charge has been proposed to be optimal for antimicrobial peptides, since cationic peptides are thought to undertake electrostatic interactions with the negatively charged phospholipid membranes of bacteria, fungi and other microorganisms [34][35][36][37] .
To verify the importance of the two Arginine (R) residues for the biological activity of NoPv1, three alanine point mutated derivatives were prepared: NoPv1-R1A (ALTAQCRL), NoPv1-R7A (RLTAQCAL), and NoPv1-R1A-R7A (ALTAQCAL) (see also Table 1). Interestingly, all three NoPv1 derivatives were able to interact physically with PvCesA2, as shown by yeast two-hybrid assay (Fig. 4a), but unable to counteract P. viticola infection (Fig. 4b), pointing to the fundamental role of the net positive charge of NoPv1 for its biological activity.
Antimicrobial activity and specificity of NoPv1 peptide aptamer. Properties of NoPv1 against P.
viticola were further investigated on leaf disks and on potted plants grown in greenhouse. In particular, NoPv1 displayed a good inhibitory activity in the co-inoculation assay on leaf disks at concentrations equal or higher than 100 μM, where the percentage of leaf disk area covered by P. viticola sporulation was far below 10% at 5 and 7 dpi (Fig. 5a). On the contrary, lower concentrations were able to contrast only partially P. viticola on leaf disks, as disease severity was almost half of the control sample at 5 and 7 dpi in the case of 50 μM NoPv1 and no anti-oomycete activity was observed at 20 μM NoPv1 (Fig. 5a). In addition, nebulization of 400 μM NoPv1 on leaf disks performed using the Potter Precision Spray Tower, recognised as the standard of reference for chemical spraying techniques in the laboratory 38 , indicated a reduction of P. viticola severity at 7 dpi (Fig. 5b). Similarly, a solution of 800 μM NoPv1 resulted to be as effective as the copper-based fungicide Kocide 2000 (1.42 g l −1 active compound; see also Fig. 5b). Preliminary nebulization tests were also conducted on leaf disks to assess both the preventive and curative action of NoPv1. In particular, 400 μM NoPv1 was the lowest concentration that maximises the efficacy against P. viticola (see Fig. 5b) and it was sprayed on leaf disks at different times before [from 7 days (− 7 d) to 2 hours (− 2 h)] and after [from 1 hour (+ 1 h) to 2 days (2 d)] pathogen inoculation (Fig. 6a). The treatment efficacy was evaluated as a measure of the difference of disease severity between control (43.2 ± 4.4%; average ± standard error values) and treated (400 μM NoPv1) samples (see "Materials and methods" for further details) for each time point (Table S3). NoPv1 was effective in suppressing P. viticola infection even when applied 7 days before inoculation (-7 d), whereas its efficacy decreased very rapidly when applied after P. viticola inoculation, i.e. in post-inoculation treatments (Fig. 6a).
On the basis of the ex vivo results, the activity of NoPv1 against P. viticola was tested in vivo on grapevine plants grown under greenhouse conditions. Different NoPv1 concentrations were sprayed on grapevine plants immediately before P. viticola inoculation and the disease severity of plants treated with 400 μM (27.3 ± 5.2%), 800 μM (26.2 ± 4.2%), 1 mM (20.8 ± 2.4%) and 2 mM (21.1 ± 4.2) NoPv1 was lower than that of control plants (58.7 ± 6.5%; P < 0.05, Kruskal-Wallis test). In particular, 1 mM NoPv1 was the lowest concentration showing a P. viticola severity comparable to the copper treatment (5.4 ± 3.9%; P > 0.05, Kruskal-Wallis test) and it was further used in the following experiments. Therefore, 1 mM NoPv1 was applied at 7 and 2 days before inoculation (− 7 d and − 2 d) and immediately before (0 h) P. viticola inoculation to assess the persistence on grapevine plants. As shown in Fig. 6b, the NoPv1 efficacy was always higher than 60% in plants treated with NoPv1 compared to control plants (Fig. 6b), which showed a mean disease severity of 58.15 ± 6.6% (see Table S4). . Photosynthetic performance of grapevine leaves exposed ex vivo to 400 µM and 1 mM NoPv1 for 7 days to evaluate NoPv1 cytotoxicity. The maximum quantum yield (F V /F M ) and the effective quantum yield [Y (II) ] of photosystem II were measured with an imaging Chl fluorometer (Imaging PAM) after 7 days exposure to 400 µM (P1) and 1 mM NoPv1 (P2). As positive control, the same leaves were treated with 0.1% v/v (BH1) and 0.2% v/v (BH2) of BASTA herbicide. Treatment with water (M1 and M2) was used as negative control. Representative images from 10 biological replicates are shown. Herbicide treated spots showed a clear reduction of photosynthetic performance, even though no clear symptoms were visible in bright-field leaf images, highlighting the high sensitivity of the assay. Furthermore, NoPv1-and water-treated spots showed comparable photosynthetic values (see Table S2), indicating that the peptide aptamer is not cytotoxic for plant cells. See Table S2 for average and standard deviation values of photosynthetic parameters. Scale bar = 1 cm. www.nature.com/scientificreports/ NoPv1 specificity was further tested to verify potential harmful effects on non-target organisms. Coherently with the very low homology of CesA enzymes from Escherichia coli, Agrobacterium tumefaciens and Bacillus amyloliquefaciens with PvCesA2 (see also Fig. 1 and Data S1), NoPv1 peptide aptamer, provided directly in the growth medium at 100 μM and 200 μM, was not able to inhibit the growth of these bacteria, as shown by the very similar growth curves (Fig. 7a-c). Furthermore, the co-inoculation of NoPv1 at 200 μM and 400 μM with Erysiphe necator, an Ascomycota responsible for grapevine powdery mildew, was not able to reduce the disease severity on leaves (Fig. 7d), in agreement with the fact that the cell wall of E. necator, as other Ascomycota, is mainly formed by chitin and other types of glucans 39 , rather than cellulose. On the contrary, comparison of the PvCesA2 cytoplasmic domain used in the yeast two-hybrid screening with the orthologous P. infestans CesA2 and co-inoculated with P. viticola sporangia suspension on grapevine leaf disks, as droplets, was monitored at 5 and 7 days post-inoculation (dpi), as percentage of leaf disk area covered by P. viticola sporulation, according to the protocol reported in Lazazzara et al. 74 . NoPv1 displayed a good anti-oomycete activity at concentrations equal or higher than 100 μM. (b) The anti-oomycete activity of NoPv1 was subsequently evaluated using a Potter Precision Spray Tower, employed to mimic in field spray conditions. Leaf disks were treated with NoPv1 at different concentrations, ranging from 100 to 800 μM, and its efficacy, measured as percentage of disease severity, was compared with a copper-based commercial fungicide, (Kocide 2000, copper hydroxide). Each experiment was carried out twice. For each treatment, average and standard error values of ten replicates from the two experiments are presented. Different lowercase and uppercase letters indicate significant differences among treatments at 5 and 7 dpi, respectively, according to a Kruskal-Wallis test (P ≤ 0.05). www.nature.com/scientificreports/ protein (PiCesA2) showed 97% of amino acid residue identity (Data S1). As it could be hypothesised, NoPv1 was able to inhibit P. infestans growth in vitro at the same concentrations used against P. viticola, i.e. 100 μM and 200 μM (Fig. 8), however further experiments are needed before concluding that NoPv1 interacts with PiCesA2 and it is able to inhibit cellulose biosynthesis.
Absence of potential cytotoxicity of NoPv1. To investigate further the potential cytotoxicity of NoPv1, human cells were used to perform an in vitro cell viability test, or MTT assay. Immortalized human cell cultures (HKC8) at different densities (1 K, 3 K and 6 K/100 μl) were grown in DMEM-F12 medium in the presence of 400 μM NoPv1 and their viability was measured after 24 and 48 h from the treatment. No significant differences between control and treated samples were observed after 24 h with respect to the level of absorbance at 570 nm, and cytotoxic effects could only be observed at 1 K cell density after 48 h, indicating that NoPv1 has no major deleterious effects on human cells (see Fig. 9).
NoPv1 blocks Plasmopara viticola germ tube formation. Leaf disks of V. vinifera cv. Pinot noir were inoculated with P. viticola zoospores in the absence (control, Fig. 10a-c) and in the presence (Fig. 10d-f) of 200 μM NoPv1. In control leaf disks, the biflagellate zoospores (n = 90, Fig. 10a) were able to reach the stomata localised on the abaxial leaf surface (Fig. 10b). There, they encysted either alone or in groups (up to 4 germinating cysts/stomata) and emitted a single germ tube (see arrow in Fig. 10b) for each zoospore that penetrated the stomata (Fig. 10b,c). In presence of 200 μM NoPv1, P. viticola zoospores were unable to develop any germ tube essential to penetrate the stomata (n = 160, Fig. 10d,e). In the very few cases (5 zoospores out of 160, 3.12%) where a germ tube primordium could be detected (see arrow in Fig. 10f), its orientation was opposite to the stomata localization (see asterisk in Fig. 10f), indicating that NoPv1 prevents germ tube formation and proper stomatal recognition (Fig. 10f).

Discussion
The need of alternatives to traditional pesticides. Plasmopara viticola is a relevant plant pathogen which requires control measures to avoid severe yield losses. Researchers have dedicated many energies to identify host resistance genes to the pathogen attacks, and a number of breeding programs have introgressed resist-   [42][43][44][45] . Because of that and due to the long time required for conventional breeding 46 , wine production is heavily dependent on the use of pesticides to control this disease. Nevertheless, strong selection pressure following repeated pesticide applications has led to the development of resistant pathogen populations, which has limited the success of chemical pesticides in grapevine and other crops [47][48][49][50] .
In this scenario, it is important to develop new tools that allow to easily and quickly isolate molecules, intrinsically less dangerous for consumers, farmers and the environment, able to counteract emerging pathogens. Here, we reported the development of an in vivo strategy aimed to isolate synthetic antimicrobial peptide aptamers from a combinatorial library. These synthetic peptide aptamers share several features, including size, level of hydrophobicity and net charge, with the antimicrobial peptides (AMPs), synthesised by the innate immune system of various species including human, animals and plants, which are the first-line defence against foreign attacks 51,52 . This strategy allowed the isolation of the 8 a.a. NoPv1 peptide aptamer able to counteract P. viticola and P. infestans infections, but it might have a much broader impact in the long term by being applicable to other species and type of crop pathogens, i.e. viruses and bacteria.
The mechanism of action of NoPv1. NoPv1 inhibits zoospore germ tube formation (see Fig. 10) that requires active cellulose biosynthesis, as reported in P. infestans where the cellulose synthases were shown to be localised and play a major role in the growing tip of the germ tube 28 , suggesting a possible inhibitory role of NoPv1 on PvCesA2. This notion seems to be supported by the capability of NoPv1 to affect severely P. infestans growth, given that its CesA2 enzyme shares high amino acid sequence identity (97%) with the PvCesA2 (see Fig. 8) and by the failure of NoPv1 to inhibit the growth of bacteria, such as E. coli, A. tumefaciens and B. amyloliquefaciens and to counteract the leaf infection caused by the Ascomycota E. necator (see Fig. 7). However, further interaction analyses and in vitro enzymatic assays are needed before concluding that NoPv1 is a specific inhibitor of PvCesA2 and PiCesA2 enzymatic activity. The selectivity of NoPv1 for certain microorganisms might suggest that NoPv1, with two positively charged "R" residues, does not behave as a typical broad spectrum www.nature.com/scientificreports/ cationic AMP able to alter membrane permeability by simply inserting into the lipid membranes and forming ion channels or pores that eventually result in leakage of cell contents and cell death [53][54][55][56] . However, the role of NoPv1 on cell membrane permeabilization remains to be investigated. The replacement of "R" residues with "A" in NoPv1-R1A, NoPv1-R7A and NoPv1-R1A-R7A (see Table 1) abolished the antimicrobial activity (Fig. 4b), but not the physical interaction with PvCesA2 (see Fig. 4a), suggesting that the two cationic residues might be essential for NoPv1 cellular uptake by P. viticola, thus reaching the PvCesA2 catalytically active portion, which localizes intracellularly. This is in agreement with the mechanism of action of some cationic AMPs, reported to penetrate cells and affect the cellular physiological processes, without altering the permeability of microbial membranes 57,58 . For instance, the natural echinocadin AMP family targets specifically 1,3 β glucan synthase, an enzyme essential for cell wall integrity of fungi 59 . Furthermore, chitin biosynthesis is blocked by nikkomycins which are the most widely studied peptidyl nucleoside inhibitors of chitin synthases 60,61 . Alternatively, NoPv1 biological activity could strictly rely on the presence of "R" residues without necessarily interacting with the PvCesA2 enzyme.
The safety and sustainability of NoPv1 potential application under field conditions. Beside target specificity, novel drugs must have a broad-spectrum activity, low toxicity and ideally no off-target organisms. These properties are generally used as guidelines in drug discovery and might give the impression that such molecules are unattainable. The selection of PvCesA2 as a target for peptide aptamers makes NoPv1 both specific for the target and, concomitantly, able to act on a broad spectrum of oomycete pathogens since their cellulose synthases group together in a clade separated from bacteria, viral and plant cellulose synthases (see Fig. 1). As a consequence, NoPv1 is able to inhibit the growth of P. infestans (see Fig. 8), an oomycete that causes the serious potato and tomato disease known as late blight or potato blight, but at the same time it does not affect any In representative images of P. infestans growth on Petri dishes at 7 dpi, the diameter of P. infestans mycelia is smaller in the presence of NoPv1 than under control conditions, indicating that NoPv1 is able to inhibit P. infestans growth. Each treatment was carried out on five petri-dishes (replicates), and experiments were carried out twice. Different lowercase, uppercase and italics letters indicate significant differences among treatments at 4, 5 and 7 dpi, respectively, according to a Kruskal-Wallis test (P ≤ 0.05).  Figs. 3 and 7). Furthermore, the absence of any major cytotoxicity on mammalian cells (see Fig. 9), likely due to the lack of cellulose synthase enzymes and the neutral net charge of cell membranes 35,62 confers to NoPv1 almost all the desirable features required for an antioomycete active compound. Nevertheless, the relatively high concentration of NoPv1 peptide, around 1 mM, shown to have efficient antimicrobial activity under greenhouse conditions seriously challenges its employment under field conditions, mainly as a consequence of the elevated costs of production that are certainly not competitive with the conventional pesticides. However, lower NoPv1 concentrations showed high efficacy on leaf disk assays (e.g. 100 µM and 400 µM in the co-inoculation and nebulization experiment, respectively), suggesting that appropriated formulations could improve NoPv1 stability under field conditions and possibly reduce the risk of degradation by UV light and/or leaf-associated microorganisms. Moreover, a lot can be learnt from the biomedical sector, where recent publications report the successful employment of antifungal peptides in combination with conventional antifungal drugs 63 . Similarly, NoPv1 could be sprayed in combination with conventional active compounds under field conditions, resulting in the reduction of peptide dose, i.e. reduced treatment costs, and in the prevention of drug resistance development as a direct consequence of the mechanistic polyfunctionality improvement of the active compounds. Furthermore, the design of novel delivery systems aimed to improve safety and bioavailability 64 , in combination with precision agriculture tools, may contribute to make the peptide-based pesticides sustainable for the agricultural sector in future.

Conclusions
We have described a novel approach to identify active compounds to counteract P. viticola infection and proposed that this strategy could be applicable to other crop pathogens. In this context, NoPv1 is reported as a highly effective, relatively safe peptide to be used as potential anti-oomycete agent. Despite these features, it must be observed that only a very limited number of antimicrobial peptide-based drugs, around 9, have been approved by the Food and Drug Administration, so far 56 . Based on that, it is reasonable to think that NoPv1 must be subjected to major improvements before its use for agricultural purposes. Obviously, the stability of NoPv1 must be verified at high temperature, in the presence of proteases or at different pH in order to be used www.nature.com/scientificreports/ under field conditions. However, unfavourable properties limiting peptide usage could be overcome by chemical optimization. For instance, synthetic peptides can be stabilised once chemically modified. Some of them have been reported to keep biological activity at high temperature, and some can resist to hydrolysis mediated by proteases, like tripsin and pepsin 54 . Moreover, the wide variety of structural and functional features identified for various natural antimicrobial peptides represent an invaluable source of ideas for adapting peptides to our needs, providing promising perspectives to the identification of safer alternatives to conventional pesticides. Finally, a cost-efficient peptide manufacturing technology needs to be set-up to compete with the production costs of conventional pesticides. Indeed, the peptide chemical synthesis is not applicable to agriculture purposes due to the high costs of production, however several companies in recent years offer cost-efficient fermentative bioprocesses, up to ton scales, for the production of recombinant peptides.

Materials and methods
Phylogenetic and amino acid sequence analysis. CesA and Csl amino acid sequences were obtained from UniprotKB database (uniprot.org), except for KJD55249.1 (BaCesA) which was obtained from the National Center for Biotechnology Information database (ncbi.nlm.nih.gov). Multiple sequence alignment was generated with Clustal omega (ebi.ac.uk/Tools/msa/clustalo) 65 . The dendrogram was constructed with a Minimum-Evolution algorithm and 1000 bootstrap repetitions using MEGA X (megasoftware.net) 66 . Transmembrane regions were predicted with SOSUI (harrier.nagahama-i-bio.ac.jp/sosui) 67 and domain prediction was performed with CD-BLAST (ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) 68 .

Plant material and phytopathogens.
Two year-old plants of the susceptible V. vinifera cv. Pinot noir were grown under greenhouse conditions at 25 ± 1 °C, with a photoperiod of 16 h light and a relative humidity of 70 ± 10%, as described by Perazzolli et al. 69 . A P. viticola population was collected in an untreated vineyard sited in the Trentino region (Northern Italy) and maintained by weekly inoculations on potted Pinot noir plants grown in greenhouse as described by Perazzolli et al. 69 . To obtain P. viticola inoculum, plants with disease symptoms were incubated overnight in the dark at 99-100% RH and 25 ± 1 °C to promote pathogen sporulation. Sporangia were collected by washing the abaxial surfaces bearing freshly sporulating lesions with cold (4 °C) distilled water and the concentration of the inoculum suspension was adjusted to 1 × 10 5 sporangia ml −1 by counting with a haemocytometer under a light microscope. The P. infestans isolate (kindly provided by M. Finckh and A. Butz, University of Kassel, Germany) was grown on pea agar medium (PAM, 12.5% w/v frozen All leaf disks were then fixed, covered with gold and observed with a LEO 1430 scanning electron microscope. NoPv1-treated zoospores were unable to develop a proper germ tube and to penetrate the stomatal cavity of grapevine leaf disks. In (f), the black arrow indicates a germ tube primordium that points in the opposite direction with respect to the nearest stomata (asterisk), suggesting that NoPv1-treated zoospores are also unable to sense stomata proximity. In total, about 250 zoospores with and without NoPv1 were observed. The analysis was carried out twice, using 12 independent leaf disks.
RNA extraction and cDNA synthesis. P. viticola sporangia were collected by washing the abaxial surfaces of grapevine leaves bearing freshly sporulating lesions with cold (4 °C) distilled water. After centrifugation at 4000×g and 4 °C, the pellet (0.05 g) was immediately frozen in N 2 -liquid and stored at − 80 °C. Total RNA was extracted using the Spectrum Plant total RNA kit (Sigma-Aldrich, St. Louis, MO), quantified using the Nanodrop 8000 (Thermo Fisher Scientific, Wilmington, DE) and its quality was checked by agarose gel electrophoresis. RNA was treated with DNase I (Invitrogen, Thermo Fisher Scientific), and first-strand cDNA was synthesised from 500 ng of total RNA using ImProm-II Reverse Transcription System Kit (Promega Corporation, Madison, WI, USA) with a combination of oligo (dT) primers and random hexamers. Construction of the peptide aptamer library. The peptide aptamer library was constructed according to Reverdatto et al. 31 with modifications. The oligonucleotide 5′-GGC AGA GTG GAT CCAA(NNK) 8 AAG GGG CCC CTT -3′, which contains complementary regions and flanking 5′-BamHI and 3′-ApaI restriction sites, was self-annealed and the 5′ and 3′ single-stranded ends were filled-in by using the Klenow Fragment (Thermo Fisher Scientific). Subsequently, the DNA fragment was digested with BamHI and ApaI restriction enzymes (Fastdigest, Thermo Fisher Scientific), dephosphorylated with FastAP Thermosensitive Alkaline Phosphatase (Thermo Fisher Scientific) and finally ligated into the plasmid pLib2 (kindly provided by Alexander Shekhtman). The library was then amplified in E. coli DH5α competent cells. About 2.5 × 10 6 colonies were collected with LB medium containing ampicillin and let grow for about 4 h. Plasmids were then purified by using the Qiagen Maxi prep kit.
Yeast two-hybrid (Y2H) library screening and peptide aptamer identification. The aptamer library was transformed into the yeast strain Saccharomyces cerevisiae AH109 (Clontech Laboratories, Palo Alto, CA, USA) harbouring the bait-containing plasmid pGBKT7-PvCesA2 71 . Transformants were selected for growth on selective media without tryptophan, leucine, and either adenine or histidine supplemented with 10 mM 3-AT (3-Amino-1,2,4-triazole, a histidine biosynthesis inhibitor). Yeast plasmid DNA was purified from the positive colonies, electroporated into E. coli cells, and the plasmids isolated and sequenced using the GrA_019 primer (5′-TCC AAG CTT TGC AAA GAT GG-3′).
Peptide synthesis and purification. The peptides identified by the Y2H strategy and the NoPv1 mutated derivatives were initially prepared by microwave assisted solid phase synthesis, based on Fmoc chemistry on pre-loaded Wang resin (0.4 meq/g substitution) using a fivefold molar excess of 0.2 M Fmoc-protected amino acids dissolved in NMP and using HOBT/HBTU/DIEA (5:5:10) as activators 72,73 ). Coupling reactions were performed for 5 min at 40 W with a maximum temperature of 75 °C. De-protection was performed in two stages using 20% v/v piperidine in DMF (5 min and 10 min each). The cleavage from the resin was performed using 10 ml of Reagent K (TFA/phenol/water/thioanisole/EDT; 82.5/5/5/5/2.5) for 180 min. After cleavage, peptides were precipitated out and washed using ice-cold anhydrous ethyl ether. All peptides were purified by RP-HPLC using a gradient elution of 5-70% v/v solvent B (solvent A: water/acetonitrile/TFA 95/5/0.1; solvent B: water/ acetonitrile/TFA 5/95/0.1) over 20 min at a flow rate of 20 ml/min. The purified peptides were freeze-dried and stored at 0 °C. The peptides purity was > 95% and was determined using analytical HPLC (95% v/v A for 5 min; 95-30% v/v A over 20 min) high resolution mass spectrometry (HRMS) and NMR (see Data S2). Peptide identity was also confirmed by electrospray ionization mass spectrometry (ESI-MS). NoPv1 peptide was later purchased at > 95% purity from Bio-Fab Research (Rome, Italy). All the lyophilized peptides and mocks were prepared using the same solvent, either distilled water or 5% v/v DMSO, according to the overall solubility of used peptides. www.nature.com/scientificreports/ Activity of peptide aptamers against Plasmopara viticola on leaf disks. Leaves (from the fourthsixth node) were collected from V. vinifera cv. Pinot noir plants grown under greenhouse conditions and surface sterilised by immerging them for 2 min in 1% v/v sodium hypochlorite solution. After that, they were rinsed three times with water and dried on paper. Leaf disks (18 mm diameter) were cut using a cork-borer and placed, abaxial side up, in 9 cm diameter plastic Petri dishes (five disks for each dish), containing four filter papers moistened with 4 ml of sterilised distilled water 42 . Each leaf disk was inoculated with five drops (10 μl) of a P. viticola sporangia suspension (1 × 10 5 sporangia ml −1 ) mixed with the respective peptide aptamer at the appropriate concentration. Dishes were sealed with a plastic film and incubated overnight (16 h) in a growth chamber at 22 ± 1° C in the dark, then dried with filter paper and incubated for 7 days in a growth chamber at 22 ± 1° C and 16 h photoperiod. Disease severity was assessed at 5 and 7 days post inoculation (dpi) as percentage of leaf disk area covered by P. viticola sporulation, calculated as sum of the disease severity of the five drops for each disk 74 . Each inoculum drop was scored as surface with no sporulation (0%), scarce sporulation (10%) or fully covered by sporulation (20%). Five replicates (dishes with five disks each) were assessed for each treatment and the experiment was carried out twice.
To analyse the peptide aptamer effects by spray nebulization, Petri dishes (with five leaf disks each) were sprayed using a Potter Precision Spray Tower (Burkard Scientific Co., Uxbridge, UK) with 1.67 ml of the peptide aptamer solution (corresponding to the standard dosage of 10 hl ha −1 in a vineyard with a 'Pergola trentina' training system) at a pressure of 55 kPa. As control treatments, Petri dishes were sprayed with water (mock) or with copper hydroxide (1.42 g l −1 Kocide 2000; Du Pont, Wilmington, DE, USA) as a reference fungicide. Petri dishes were incubated under a laminar flow hood for 5-10 min in order to dry the leaf disks, sprayed with a fresh P. viticola sporangia suspension using a small hand sprayer device (0.6 ml per dish). Afterwards, Petri dishes were incubated overnight in the dark at 22 ± 1 °C in a growth chamber, then dried under a laminar flow hood and incubated for 7 days in a growth chamber, as described above. Disease severity was assessed visually as percentage of leaf disk area covered by P. viticola sporulation at 5 and 7 dpi 75 . The disease reduction (efficacy) was calculated according to the following formula 76 : (disease severity of mock disks-disease severity of treated disks) / (disease severity of mock disks) × 100. Five replicates (Petri dishes with five leaf disks each) were assessed for each treatment and the experiment was carried out twice. To assess the persistence, as well as the preventive and curative action of the peptide aptamers, treatments were applied on the leaf disks at different times (from 7 days to 2 h) before pathogen inoculation. After each application, leaf disks were dried under a laminar flow hood for 5-10 min and then kept in the growth chamber. In the curative treatment the peptide aptamer was applied after pathogen inoculation (0, 1, 24, 48 h post infection).

Activity of peptide aptamers against Plasmopara viticola on greenhouse-grown plants.
Vitis vinifera cv. Pinot noir plants were grown in a greenhouse for 2 months as described by Perazzolli et al. 69 . Peptide aptamer was applied with a compressed-air hand sprayer to the abaxial and adaxial surfaces of all leaves (20 to 30 ml, depending on the number of leaves) at 7 days before inoculation, 2 and 0 (co-inoculation), whereas not treated plants were used as control. A fresh sporangia suspension (1 × 10 5 sporangia ml −1 ) was applied to the abaxial leaf surface using a compressed air hand sprayer (20-30 mL per plant). Inoculated plants were incubated overnight in the dark at 25 ± 1 °C with 99-100% RH. Six days after inoculation, plants were incubated overnight at 25 ± 1 °C with 99-100% RH to promote P. viticola sporulation and the disease severity was assessed visually as percentage of abaxial leaf area covered by sporulation in relation to the total leaf area according to EPPO standard guidelines 75 . Five replicates (plants) were assessed for each treatment and time point, and the experiment was carried out twice.
Antimicrobial activity of peptide aptamers on bacteria. The antibacterial activity of NoPv1 was evaluated in liquid culture using three different bacterial species, Escherichia coli, Agrobacterium tumefaciens and Bacillus amyloliquefaciens (subsp. plantarum) strain D 747. Overnight pre-cultures were diluted to OD 600 of 0.1 (E. coli; B. amyloliquefaciens) or 0.2 (A. tumefaciens) into sterile LB medium. NoPv1 (100 and 200 µM) was then added to the liquid medium of three biological replicates. Bacteria were grown at 28 °C (A. tumefaciens) or 37 °C (E. coli; B. amyloliquefaciens). The density of the cell population was measured spectrophotometrically at OD 600 every hour, for 5 h.
Activity of peptide aptamers against Phytophthora infestans. The activity of NoPv1 against P.
infestans was evaluated on Petri dishes with Pea Agar Medium (PAM) supplemented with 100 µM and 200 µM NoPv1. Mycelium plugs (7 mm in diameter) were cut from 7-day old colonies and placed upside down at the centre of the Petri dish. Petri dishes containing only PAM, without NoPv1, were used as control. Dishes were sealed with a plastic film and incubated in the dark at 20 ± 1 °C. The diameter of P. infestans colonies was assessed at 4, 5 and 7 dpi. Five replicates (petri dishes) were analysed for each treatment and the experiment was carried out twice.
Antifungal activity of peptide aptamers against Erysiphe necator. Evaluation of the ability of NoPv1 to protect grapevine plants against E. necator was carried out using the method described by Miclot et al. 77 with minor modifications. The second and third fully expanded leaves from the apex were collected from different V. vinifera cv. Pinot noir plants grown under greenhouse condition and surface sterilised by washing for 2 min in a 1% v/v sodium hypochlorite solution. Leaves were rinsed three times for 2 min with distilled water and briefly dried on paper. The base of the petiole was cut off and leaves were placed, adaxial side up, in square culture dishes (120 × 120 mm), containing 1% w/v agar covered with sterile filter paper (1 layer, moistened with 2 ml of sterile distilled water). A small cut was performed with a sterile scalpel on the paper and the petiole was