Postharvest biocontrol ability of Pseudomonas synxantha against Monilinia fructicola and Monilinia fructigena on stone fruit

The biocontrol properties of the endophyte Pseudomonas synxantha DLS65 were tested in vitro and in vivo against Monilinia fructicola and Monilinia fructigena, causal agents of postharvest brown rot of stone fruit. P. synxantha cells significantly reduced the mycelial growth of both pathogens on Potato Dextrose Agar (PDA), and strongly inhibited the Monilinia fructicola growth on Peach Extract Agar (PEA). Cell-free culture filtrates inhibited the pathogens on PDA and PEA to lesser extent. The production of volatile organic compounds (VOCs), with in vitro inhibitory effects on mycelial growth, was also observed. P. synxantha significantly reduced brown rot incidence and severity on peach fruit artificially inoculated with M. fructicola after 5 d at 25 °C. Moreover, P. synxantha more significantly reduced incidence and severity after 10 d at 10 °C and after 20 d in cold storage at 0 °C in comparison to control fruit, even if its activity was never comparable to that of the synthetic fungicide Scholar (fludioxonil). Similarly, P. synxantha exhibited an excellent antagonistic activity against M. fructigena on fruit at 10 and at 0 °C, and a weak biocontrol activity at 25 °C. Competition for nutrients and space, production of diffusible toxic metabolites and VOCs may play a role in the antagonism of P. synxantha toward M. fructicola and M. fructigena, especially at the lowest temperatures of storage. For that reason, this strain of P. synxantha could be suggested as active ingredient for the setting up of bioformulates against Monilinia species representing a limiting factor for stone fruit production.


Introduction
Peach and nectarine represent important crops widely cultivated in Europe with 259.286 ha of area harvested and a production of 4,373,494 tons in 2016 (FAOSTAT, 2016). Approximately 33% of this production is mainly concentrated in Italy, one among the most important country in European area. Unfortunately, the cultivation of stone fruit species including peach (Prunus armeniaca) is heavily threatened by brown rot disease caused by Monilinia species. In detail, three species are considered 'key' postharvest pathogens for stone fruit i.e. Monilinia fructicola (Winter) Honey, Monilinia fructigena (Aderhold and Ruhland) and Monilinia laxa (Aderhold and Rulhland) Honey. These fungal species infect blooms, twigs and fruit in the field, but prevalent damages occur in the postharvest stage during storage, shipping and marketing inducing up to 80% yield losses under favorable conditions to disease development (Usall et al., 2015).
Up to recent days, fungicide applications are needed in orchards but different factors i.e. the arising of fungicide resistance and relative in field breakdown efficacy, latent infections, and the fungal capacity to develop at low temperatures could strongly limit their performances (Miessner and Stammler, 2010). Moreover, the recent EuropeanDirective on "Sustainable Use of Pesticide" encouraged developing of alternative control methods (Lopez-Reyes et al., 2013;Sisquella et al., 2014). Therefore, numerous scientific papers and world-wide programs focused on isolation of biological control agents (BCAs) effective against several postharvest fungal pathogens, including species belonging to Botrytis, Monilinia and Penicillium genera (Mari et al., 2012;Panebianco et al., 2015;Parafati et al., 2015;Parafati et al., 2017;Platania et al., 2012;Restuccia et al., 2006;Scuderi et al., 2009;Zhang et al., 2010a, b) on numerous foodstuffs. For example, many bacteria have also shown a good biocontrol potential, including Rahnella aqualitis (Calvo et al., 2007), Bacillus amyloliquefaciens (Calvo et al., 2017), Pseudomonas cepacia (Janisiewicz and Roitman, 1988), P. syringae (Cirvilleri et al., 2005;Panebianco et al., 2015), Burkholderia gladioli (Scuderi et al., 2009), and P. fluorescens (Wallace et al., 2017). In detail, the history of biological control in postharvest began when Pusey and Wilson (1984) successfully proposed a bacterium recovered from soil to manage brown rot on stone fruit caused by Monilinia fructicola. Since then, more papers were produced about potential use of alternatives to control postharvest diseases of stone fruit although only few of them are currently applied under commercial conditions (Usall et al., 2015).
Currently, no bioformulate containing BCAs is registered for Monilinia postharvest decays of peach whereas only an active ingredient (a.i.) -fludioxonil (Scholar ® ) -is used in Italy for managing postharvest Monilinia decay. This molecule is a derivative of pyrrolnitrin, an antibiotic initially isolated from Pseudomonas species with high antagonistic activity versus postharvest fungal decays (Arima et al., 1964;Janisiewicz and Roitman, 1988). Pseudomonas species are widely distributed in the rhizosphere, and are able to promote the plant growth and to counteract plant diseases. Some of these antagonistic bacterial species have been often isolated from natural environment in which they compete with pathogenic epiphytic microorganisms (Tontou et al., 2016a, b). Several characteristics of Pseudomonas spp. can explain the dominance and persistence of this genus in a large range of environments, such as frequent introduction into the food production environment and its ability to survive cleaning and disinfection and to grow even with low levels of nutrients and at low temperatures (Moore et al., 2006). Among Pseudomonas species, P. synxantha produces a bioactive compound, a long chain aliphatic hydrocarbon with a terminal double bond and intermediate electronegative atom with activity similar to bio-surfactant molecule, which is effective against several strains of Mycobacteria (Mukherjee et al., 2014). In addition, P. synxantha strain DLS65, isolated from vascular tissues of Actinidia spp. plants, showed strong in vitro inhibition activity against Gram + and Gram -phytopatogenic bacteria (Tontou et al., 2016a). Therefore the aims of the present study were i) to determine antagonistic activity of DLS65 strain against Monilia fructicola and Monilia fructigena in vitro and ii) to evaluate the potential biocontrol activity of brown rot caused by these pathogens in postharvest on peach fruit by comparison with a standard fungicide.

Pathogens
Two strains of Monilinia spp., i.e. M. fructicola and M. fructigena were employed for this study. Monilinia fructicola was isolated from rotten peach fruit in South Italy (Sicily), preliminarily identified and maintained in collection belonging to Dipartimento di Agricoltura, Alimentazione e Ambiente (Di3A, University of Catania) while M. fructigena (strain number 2138) was provided by Bank of Plant Pathogens (Environmental Protection Institute in Poznan, Poland) (Grzegorczyk et al., 2017). To maintain their virulence both pathogenic strains were routinely recovered from inoculated peach fruit, and the monoconidial isolates of both fungi were used in in vitro and in vivo experiments. Fungal strains were stored on Potato Dextrose Agar (PDA, Oxoid) slants at 4°C.
2.3. In vitro antagonistic activity of P. synxantha cells P. synxantha DLS65 was tested for its inhibitory activity against M. fructicola and M. fructigena on PDA and on Peach Extract Agar (PEA) (100 g L -1 of peach pulp extract; 20 g L -1 of agar). Fungi to be tested were grown on PDA for 10 d at 25°C. An aliquot (1 mL) of P. synxantha DLS65 suspension, obtained from 2-day-old culture on NA (approximately 1 × 10 9 CFU mL -1 on phosphate buffer), was added to 14 mL of fluid PDA/PEA media maintained at 45°C. After mixing and agar solidification, an agar plug (5 mm diameter) from actively growing margins of M. fructicola and M. fructigena colonies was placed at the center of PDA/PEA plates 1 h (co-inoculation) and 24 h (differed inoculation) after DLS65 application. The control samples consisted of PDA/PEA Petri plates with only fungal inoculums. After incubation at 25°C for 6 d, the diameter of the fungal mycelium was measured. Each experiment was conducted twice and included four replicates per treatment.
2.4. In vitro inhibitory activity of P. synxantha cell-free culture filtrates P. synxantha DLS65 culture filtrates were prepared as reported by Cirvilleri et al. (2005) partially modified. P. synxantha DLS65 was grown overnight in NA and 1 mL of bacterial suspension (approximately 1 × 10 9 CFU mL -1 ) was added to 100 mL of Nutrient Broth (NB, Oxoid). After 2 d of incubation at 25°C in still culture and centrifugation (9000 × g for 20 min) of bacterial suspensions, the supernatant was passed through a 0.22 μm Millipore filter (Millipore, Billerica, MA, USA) to obtain cell-free culture filtrates. Aliquots (1 mL) of culture filtrates were added to 14 mL of fluid PDA/PEA maintained at 45°C. After mixing and agar solidification, agar plugs of M. fructicola and M. fructigena were placed at the center of plates 1 h (co-inoculation) and 24 h (differed inoculation) after DLS65 culture filtrate application. After incubation at 25°C for 6 d, the diameter of the fungal mycelium was measured. Each experiment was conducted twice and included four replicates per treatment.

In vitro effect evaluation of volatile organic compounds
To test the activity of volatile organic compounds (VOCs) produced by bacterial strain in reducing mycelial growth of both fungi a dual culture technique was employed according to recent studies (Grzegorczyk et al., 2017;Parafati et al., 2017). Amounts of 100 μL of bacterial suspension (1 × 10 9 CFU mL -1 ) were cultured on PDA plates and incubated at 25°C for 48 h. Subsequently, agar plugs (5-mm in diameter) from active growing mycelium of M. fructicola and M. fructigena were put on the center of plates. The PDA dishes with pathogens were individually covered face to face under plates containing 48 h bacterial strain cultures. The control consisted of bacteria-unseeded PDA plates. The two plates were wrapped together with Parafilm around the edges to prevent air leakage, and incubated at 25°C. Diameter of the fungal mycelium and radial growth reduction was calculated after 6 d of incubation as previously described. All experiments were repeated twice and included four replicates per treatment.

In vivo biocontrol activity of P. synxantha DLS65
To assess the in vivo efficacy of P. synxantha DLS65, experiments hereinafter referred as experiment I were carried out, and the method described by Panebianco et al. (2015) and Grzegorczyk et al. (2017) with modifications was used. Peach fruit [Prunus persica (L.) Batsch.] 'Leonforteʼ (Sicilian variety awarded by the European Community with label "Protected Geographical Indication) used in this study were recovered from commercial orchard in Sicily (Italy), and they have not received any in field (pre-harvest) fungicide treatment. Homogenous and healthy fruit were selected and randomly assigned to different treatments. Prior to treatment and inoculation, peach fruit were washed, disinfected and wounded as it was recently reported by Grzegorczyk et al. (2017). Bacterial strain was grown in NB medium for 48 h at 25°C. Resulting suspension was centrifuged at 9000 g for 10 min, suspended in phosphate buffer and concentration was adjusted to 10 8 CFU mL -1 . Each wound was inoculated with 20 μl of P. synxantha DLS65 and allowed to dry. Fruit were put onto perforated aluminum support and placed into plastic trays containing wet paper towels. The plastic containers were put into plastic bags, hermetically sealed to avoid air dispersal, and incubated at 25°C. Following 24 h, treated wounded sites were inoculated with mycelia disks (5-mm in diameter) of active growing mycelium of M. fructicola and M. fructigena. Fruit were then incubated for 5 and 10 d at 25°C and 10°C, respectively. Chemical control was represented by fruit treated with a suspension containing 130 mL/100 L of fludioxonil (Scholar ® , Syngenta; 23.0% a.i.). Peach fruit treated with sterile distilled water (SDW) and only with fungal inocula were included as controls. Three replicates of 9 peach fruit were used for each treatment (27 fruit/treatment). The experiments were performed twice. Data concerning disease incidence (DI) (percentage of decayed fruit) and disease severity (DS) (diameter of infected area around the wounds) were calculated as the average of each replicate. Data concerning the disease incidence (DI) were transformed into arcsine square root values to normalize the distribution before performing the analysis of variance.

Effect of low storage temperature on biocontrol efficacy
To determine the effect of low storage temperatures on biocontrol efficacy, experiments hereinafter referred as experiments II were carried out. To this aim, the artificially wounded peach fruit were treated with 20 μL of 1 × 10 8 CFU mL -1 of P. synxantha DLS65. After 24 h, mycelial disks (5-mm square plugs) of actively growing fungal mycelium of M. fructicola and M. fructigena were inoculated as described above. Chemical control was represented by fruit treated with a suspension containing 130 mL/100 L of Scholar. Fruit treated with SDW and only with fungal inocula were included as control. Treated fruit were stored at 0°C for 20 d followed by 4 d at 25°C. The incidence and severity of disease was calculated as previously described. Three replicates of 9 peach fruit were used for each treatment (27 fruit/treatment). The experiment was performed twice.

Statistical analysis
Data from in vitro and in vivo experiments were analyzed separately by using the factorial ANOVA module of Statistica 10 software package (StatSoft Inc, Tulsa, OK, USA) to determine significant differences among the tested treatments in vitro and in vivo performances against M. fructicola and M. fructigena. Initial analyses were conducted by calculating F and P values associated for all experiments to evaluate any significant treatment × species, treatment × temperature, treatment × species × temperature interactions. Thus, in all repeated experiments, arithmetic means of mycelial growth, disease incidence (DI) and disease severity (DS) were calculated, averaging the values determined for the single replicates of each treatment. Percentage data concerning DI were previously transformed using the arcsine transformation (sin -1 square root x). Post-hoc comparisons among different treatments were achieved by means of the Fisher's least significant difference test.

Results
3.1. In vitro antagonistic activity of P. synxantha cells P. synxantha DLS65 was able to inhibit the mycelial growth of M. fructicola and M. fructigena with variable efficacy in a substrate-dependent manner (Table 1; Figs. 1 and 2). A complete fungal growth inhibition (100%) of both pathogens was obtained on PDA, both in coinoculation and in differed-inoculation treatments. Otherwise, on PEA medium the growth of M. fructicola and M. fructigena was partially reduced, with M. fructicola inhibited more efficiently (49.8% fungal growth reduction) than M. fructigena (24.9% fungal growth reduction) in differed inoculation treatment (significant data). The worst results were obtained on PEA in co-inoculation treatments, with 42.5% and 0% of fungal growth reduction, respectively, for M. fructicola and M. fructigena (Figs. 1 and 2).

In vitro inhibitory activity of P. synxantha cell-free culture filtrates
Cell free culture filtrates of P. synxantha showed variable inhibition activity against M. fructicola and M. fructigena ( Table 2). The effectiveness of culture filtrates was significantly higher when they were simultaneously applied (co-inoculation) with M. fructicola (21.8% growth reduction in PDA; 26.7% growth reduction in PEA), whereas reduction data were significantly lower (15.7% growth reduction in PDA) or similar to the control (20.1% growth reduction in PEA) when they were applied 24 h prior to pathogens (differed inoculation). Culture filtrates inhibited less efficiently M. fructigena, both on PDA and PEA, in co-and differed inoculation, with reduction data similar to the control. Overall, P. synxantha culture filtrate never showed higher inhibitory activity if compared with bacterial cells.

In vitro production of volatile organic compounds (VOCs)
As shown in Table 3, there were no treatment × fungal species interactions (F = 2.3, P = 0.17), thus indicating that VOCs produced by P. synxantha showed similar performances against both targeted species. For this reason, the data referring to M. fructicola and M. fructigena were combined and analyzed together. In detail, VOCs were always able to reduce significantly the mycelial growth of M. fructicola and M. fructigena if compared with relative controls (Table 3). Averagely, these data show a lesser activity of VOCs than those of bacterial cells.

In vivo biocontrol activity of P. synxantha DLS65
Significant effects were detected in experiment I for two interactions (treatment × temperature and treatment × species × temperature on disease severity) (Table 4).
For this reason, the trials were analyzed separately for targeted species and temperature and post-hoc comparisons among different treatments were achieved by means of the Fisher's least significant difference test (Table 5).
Treatments with P. synxantha DLS65 reduced the incidence and severity of brown rot in wounded fruit of peaches inoculated with M. fructicola and M. fructigena and stored at 25°C and at 10°C for 5 d and for 10 d, respectively (Table 5; Figs. 3 and 4).
After storage at 25°C for 5 d, P. synxantha significantly reduced the disease incidence and severity caused by M. fructicola on peach fruit by 50% and 76% respectively, whereas it was less effective against M. fructigena reducing significantly only severity by 70%.
Following storage at 10°C, P. synxantha showed the best performances being always able to significantly reduce disease incidence and severity caused by both pathogens. Moreover, the antagonist was more effective against M. fructicola (68% DI reduction; 92% DS reduction) than against M. fructigena (55% DI reduction; 90% DS reduction). However, the fungicide (Scholar) applications always showed the best performances in reducing (100%) incidence and severity of brown decay caused by M. fructicola and M. fructigena at two different temperatures.

Effect of low storage temperature on biocontrol efficacy
Since the interactions (treatment × species) detected in experiment II (0°C for 20 d followed by 25°C for 4 d) were not significant (Table 4), the data obtained for each pathogen were combined ( Table 6). As clearly shown in this table, a good efficacy was detected with antagonist application at 0°C for 20 d followed by a 4-day storage period at 25°C, since P. synxantha significantly reduced disease incidence and severity caused by both fungi by about 63.9% and 78.5%, respectively. Also under these conditions (0°C) fludioxonil application showed the best performances in reducing decay amount (significant data; Fig. 5).

Discussion
A possible and safe alternative to manage Monilinia fruit decays is biological control using antagonistic microorganisms often isolated from food sources and from natural environment in which they compete with pathogenic epiphytic microorganisms (Grzegorczyk et al., 2017;Tontou et al., 2016a). Also bacterial endophytes, which provide important benefits to plants, might be deserved for developing innovative and sustainable biocontrol strategies (Compant et al., 2005). To this regard, the present study evaluated the antagonistic activity of P. synxantha DLS65, isolated from vascular tissues of Actinidia spp. plants, against M. fructigena and M. fructicola.
In in vitro experiments, P. synxantha strain totally inhibited mycelial  growth of M. fructicola e M. fructigena both in co-and differed inoculation treatments on standard medium. Also on a more acidic medium resembling the fruit tissue composition, the strain significantly inhibited both pathogens when applied 24 h before, with the higher activity detected against M. fructicola. Unlikely, Grzegorczyk et al. (2017) reported that M. fructigena is more efficiently inhibited by killer yeasts in acidic medium, whereas M. fructicola was more efficiently inhibited in media at pH 6.0. Diversely to living cells, the greatest inhibition by culture filtrates was averagely observed in co-inoculation trials, and M. fructicola was more efficiently inhibited by culture filtrates if compared to M. fructigena. The lower activity showed by bacterial filtrates after 24 h could be probably due to a reduction of activity of produced antifungal compounds over time. The best performances showed by bacterial cells compared to filtrates and VOCs clearly suggested that P. synxantha acts, besides by antibiosis and VOCs, also by other mechanisms involving the presence of living cells (e.g. competition for the space and nutrients, etc.).
Comprehensively, P. synxantha in vivo applications effectively managed Monilinia peach decays under all storage temperatures. However, the best performances of P. synxantha in controlling brown rot were recorded at the lowest storage temperatures considered. At 10°C, P. synxantha had a slightly higher activity against M. fructicola than against M. fructigena, whereas at 0°C antagonistic bacterium performances were similar toward both phytopathogenic fungi. Otherwise, at 25°C storage conditions, P. synxantha applications revealed the least performances in managing Monilinia rots and in one case the bacterium failed in reducing the number of peach infections caused by M. fructigena. P. synxantha is a fluorescent bacterium placed within Pseudomonas fluorescens group (Wechter et al., 2002). Past researches demonstrated the efficacy of this bacterium as a biological control agent against mycobacteria (Mukherjee et al., 2014), and its activity as nematicidal agent (Wechter et al., 2002). Accordingly, our data showed for the first time the antagonistic effects of living cells, diffusible compounds (cell-free culture filtrates) and volatile organic compounds (VOCs) of P. synxantha against the postharvest fungal pathogens M. fructicola and M. fructigena, although a certain activity against Botrytis cinerea was preliminarily reported on lettuce (Card, 2005 -PhD Thesis).
Several strains of Pseudomonas genus have been studied as potential BCAs against a wide range of fungal pathogens and for their production of anti-fungal compounds (Lugtenberg and Kamilova, 2009;Rojas-Solís et al., 2018;Sang and Kim, 2014;Smilanick et al., 1993). Some of them are able to produce hydrolytic enzymes, such as proteases, cellulases, chitinase and ß-glucosidase, as well as hydrogen cyanide (HCN), pyoluteorin, phenazines, siderophores, cyclic lipopeptides, 2,4-diacetylphloroglucinol (DAPG), and pyrrolnitrin (Calderón et al., 2015;   a F test of fixed effects and P value associated to F; *, ** = significant at 0.001 < p < 0.01 and p < 0.001, respectively; ns = not significant. b Each value represents the mean of 4 replicates, each formed by 3 Petri dishes with fungal pathogen. Data followed by different letters within each column are significantly different according to Fisher's least significant difference test (α = 0.05).

Table 4
In vivo effects of treatments and treatments × target species interactions on disease incidence and severity of fruit decays caused by target Monilinia fructicola and Monilinia fructigena at T = 25°C and at T = 10°C in the experiment I and at T = 0°C in experiment II, respectively.   Raaijmakers et al., 2010). In addition, Pseudomonas spp. are able to induce systemic resistance in plants, and to promote plant growth by excreting phytohormones (e.g. indole-3-acetic acid, IAA) and by emitting VOCs (de Vleesschauwer et al., 2008;Lo Cantore et al., 2015;Raza et al., 2016;Rojas-Solís et al., 2018). Previous studies were performed to understand the molecular basis of the biocontrol properties of P. synxantha; in detail an acyl-homoserine lactone acylase gene (pvdQ), a glucose-6-phosphate dehydrogenase gene (zwf) and an mbtH-like gene were found to be involved directly or indirectly in Non Ribosomal Peptides (NRPs) synthesis (Tontou et al., 2016b). These authors suggested that a molecule with antibiotic properties, produced by NRP synthetases (NRPSs), contributes to the antagonistic activity of this bacterium. Non Ribosomal Peptides (NRPs) have antibiotic properties and simultaneously they induce plant defence (Rosenblueth and Martinez-Romero, 2006;Compant et al., 2010). Moreover, Mukherjee et al. (2014) identified a bioactive compound produced by P. synxantha, i.e. a long chain aliphatic hydrocarbon with a terminal double bond and intermediate electronegative atom with activity similar to bio-surfactant molecule. About side effects of P. synxantha application, our data preliminarily showed no effects for peach fruit; moreover, Tontou et al. (data not published) demonstrated that this bacterium is an endophyte able to colonise, move and persist inside the vascular tissue of kiwifruit plants without side effects on this host. Related to human and animal effects of this bactetirum, the available data showed that both P. synxantha and its metabolites could induce the hemolysis of human blood and no effects for growth and survival of prawns (Van Hai et al., 2009;Mukherjee et al., 2014).

Table 6
Post-hoc analyses of treatment effects on disease incidence (DI) and severity (DS) of fruit decays caused by target Monilinia fructicola and Monilinia fructigena in the experiment II at T = 0°C. Data followed by different letters within each column are significantly different according to Fisher's least significant difference test (α = 0.05). However, further studies should be performed to evaluate the specific compounds involved in antagonistic activity and the important implications for human and environment.

Conclusions
Even if the findings of the present study underlined that fludioxonil is the most effective against Monilinia decays, it is clear that P. synxantha could represent a promising alternative for a sustainable management of brown rots of peach caused by M. fructicola and M. fructigena. Although additional studies should be performed prior to promote and develop a new commercial bio-formulate for large scale applications, the good efficacy of P. synxantha in preventative treatments against M. fructicola and M. fructigena, together with their ability to maintain or increase its antifungal activity at cool temperatures, could make this antagonist suitable for postharvest application in peach production.