Environmental factors affect the activity of biocontrol agents against ochratoxigenic Aspergillus carbonarius on wine grape

International Journal of Food Microbiology 159 (2012) 17–24

Contents lists available at SciVerse ScienceDirect

International Journal of Food Microbiology journal homepage: www.elsevier.com/locate/ijfoodmicro

Environmental factors affect the activity of biocontrol agents against ochratoxigenic Aspergillus carbonarius on wine grape F. De Curtis, D.V. de Felice, G. Ianiri, V. De Cicco, R. Castoria ⁎ Dipartimento di Agricoltura, Ambiente e Alimenti, Università degli Studi del Molise, Via F. De Sanctis snc, 86100 Campobasso, Italy

a r t i c l e

i n f o

Article history: Received 27 April 2012 Received in revised form 4 July 2012 Accepted 25 July 2012 Available online 31 July 2012 Keywords: Ochratoxin A Wine grape Biological control Aureobasidium pullulans Metschnikowia pulcherrima

a b s t r a c t The influence of temperature and relative humidity (RH) on the activity of three biocontrol agents—the yeast Metschnikowia pulcherrima LS16 and two strains of the yeast-like fungus Aureobasidium pullulans LS30 and AU34-2—against infection by A. carbonarius and ochratoxin A (OTA) accumulation in wine grape berries was investigated in lab-scale experiments. The presence of wounds on grape skin dramatically favored infection of berries by A. carbonarius strain A1102, since unwounded berries showed very low levels of infection at all conditions of RH and temperature tested. Artificially wounded berries pre-treated with the biocontrol agents were inoculated with the ochratoxigenic A. carbonarius strain A1102 and were incubated for 5 days at two levels of RH (60% and 100%) and three different temperatures (20, 25 and 30 °C). The three biocontrol agents were able to prevent infections at 60% RH and 20 °C. At 60% RH and 25 °C only strain AU34-2 achieved some protection on day 5, whereas at 30 °C a limited biocontrol efficacy was evident only up to day 2. At 100% RH, LS16, LS30 and AU34-2 showed effective protection of grape berries at 20 °C until the 5th day of incubation. The three biocontrol agents achieved significant protection at higher temperatures only until the 2nd day after the beginning of the experiment: all three strains at 25 °C, and only strain LS16 at 30 °C. After 5 days, the three biocontrol agents were able to significantly reduce the level of OTA in berries at all the conditions tested. This occurred even when protection from infection was not significant, except at 30 °C and 100% of RH for all the three strains, and at 25 °C and 100% of RH for strain LS16. The biocontrol agents displayed a higher rate of colonization on grape berries at 20 and 25 °C than at 30 °C. The higher value of RH (100%) appeared to increase the rate of colonization, in particular at 20 and 25 °C. Taken together, our results emphasize the significant influence of environmental factors on the effectiveness of biocontrol against A. carbonarius as well as on OTA contamination in wine grape berries, and the need for biocontrol agents that can cope with the environmental conditions that are conducive to attack by A. carbonarius. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Ochratoxin A (OTA) is one of the most toxic and widespread mycotoxins. It has been detected in food commodities such as cereals, coffee, cocoa beans, spices, dried fruits, beer and, since 1996, also in grape juice and wine (Zimmerli and Dick, 1996). Although cereals are considered the major source of OTA contamination, recent studies have highlighted the importance of red wines as a source of OTA intake in European populations (Commission of the European Communities, 2005). It has been estimated that in the EU about 10% of the total OTA intake originates from red wine (Miraglia and Brera, 2002). Ochratoxin A has been associated with Balkan endemic nephropathy and urinary tract tumors in humans (Hult et al., 1982; Pfohl-Leszkowicz et al., 2002), and an endemic porcine nephropathy in Northern Europe (Krogh et al., 1974). OTA has been reported to have nephrotoxic, teratogenic, immunosuppressive and neurotoxic effects (Bennett ⁎ Corresponding author. Tel.: +39 0874404698; fax: +39 0874404855. E-mail address: [email protected] (R. Castoria). 0168-1605/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijfoodmicro.2012.07.023

and Klich, 2003). The International Agency for Research on Cancer included this mycotoxin into Group 2B, because of its possibly carcinogenic activity in humans (IARC, 1993). Therefore, EU has set the maximum tolerable levels for OTA at 2 μg/kg in wine and grape juice and at 10.0 μg/kg in dried vine fruits (Commission of the European Communities, 2006). Although OTA contamination of cereals is generally associated with the presence of Aspergillus spp. in warm and tropical areas and Penicillium spp. in temperate and colder climatic zones (Magan and Aldred, 2005), recent evidence suggests that OTA contamination in wine grape is mainly due to Aspergillus carbonarius, a fungal species belonging to Aspergillus section Nigri (Battilani et al., 2006; Leong et al., 2007; Romero et al., 2005). The use of chemicals still is the most common practice to control plant diseases. Over the past decades, most of the treatments in vineyards consisted of applications of copper fungicides alone or in combination with other chemicals. Residues of fungicides can persist in the soil and in grape berries after harvest, and be transferred to the wine (Cabras et al., 1987, 1997; González-Rodríguez et al., 2009).

18

F. De Curtis et al. / International Journal of Food Microbiology 159 (2012) 17–24

Because of the potential harmful effects of fungicides on human health and the environment, the EU set maximum levels for their residues in food and feed (Commission of the European Communities, 2008). In addition, loss of effectiveness of fungicides due to development of resistant pathogen populations has fostered research on new approaches as alternatives to these chemicals. Beneficial microorganisms isolated from the phyllosphere and carposphere of different plant species can effectively antagonize and outcompete plant pathogens. The use of these microorganisms, named biocontrol agents (BCAs), represents a promising strategy to reduce the utilization of chemicals for the control of plant disease. Several studies have reported the activity and the efficacy of BCAs (Droby et al., 1989; Ippolito et al., 2004; Janisiewicz and Korsten, 2002; Lima and De Cicco, 2006), which were also found to be effective against mycotoxigenic fungi. The BCA, Rhodosporidium kratochvilovae LS11 (formerly designated Rhodotorula glutinis), effectively protected stored apples from attacks of the patulin-producing pathogen Penicillium expansum and reduced patulin contamination in these fruits (Castoria et al., 2005). Interestingly, R. kratochvilovae LS11 degraded patulin to the much less toxic desoxypatulinic acid in vitro (Castoria et al., 2011). Aureobasidium pullulans LS30 was able to reduce infections by A. carbonarius in the vineyard, thus reducing OTA contamination in grape berries. Both LS30 and the other strains of A. pullulans used in the same study (including strain AU34-2 used in this work, see below) degraded OTA to the less toxic ochratoxin α both in a synthetic medium and in fresh grape must (de Felice et al., 2008). Although some information is available about the effect of environmental factors on A. carbonarius' growth and its production of ochratoxin A (Bellí et al., 2007; Mitchell et al., 2004), to our knowledge only one report exists that discusses the influence of these same factors on the biocontrol activity of antagonistic microorganisms and on their capability to affect OTA accumulation in wine grape berries (Ponsone et al., 2011). In this study, we verified the influence of wounds on infection of wine grape berries by A. carbonarius at three temperatures (20, 25 and 30 °C) and at two values of relative humidity (60% and 100%), which were chosen on the basis of temperature and RH intervals at which growth of the fungal pathogen and OTA biosynthesis were reported (Bellí et al., 2007; Magan and Aldred, 2005; Oueslati et al., 2010). Subsequently, we assessed the wounded berries incubated in the same conditions of temperature and relative humidity for i) the biocontrol activity of strains AU34-2 and LS30 of the yeast-like fungus Aureobasidium pullulans and of strain LS16 of the yeast Metschnikowia pulcherrima against A. carbonarius, ii) the effect of these biocontrol agents on OTA accumulation, and iii) the population dynamics of the three biocontrol agents on wounded grape berries. 2. Materials and methods 2.1. Chemicals and reagents A commercial standard of ochratoxin A was purchased from Sigma-Aldrich (Milan, Italy). A stock solution of OTA was prepared by dissolving the toxin in toluene/acetic acid (99:1, v/v) and, immediately prior to use, working solutions at the desired concentration were prepared by drying aliquots of the stock solution under a nitrogen stream and by re-dissolving them in acetonitrile/water/acetic acid (99:99:2, v/v/v). Toluene, acetonitrile and acetic acid were of HPLC grade and were obtained from Sigma‐Aldrich (Milan, Italy). Nutrient broth, PDA (potato dextrose agar) and yeast extract powder were purchased from Oxoid Ltd (Oxoid Ltd, Basingstoke, Hampshire, England), and dextrose from Carlo Erba Reagenti S.p.a. (Milan, Italy). OchraTest immunoaffinity columns, glass microfiber filters GF/A and polyethylene glycol 8000 (PEG 8000) were purchased from Vicam (Watertown, MA, USA).

2.2. Microorganisms The biocontrol agents used in this study were the yeast Metschnikowia pulcherrima strain LS16 and the yeast-like fungus Aureobasidium pullulans strains AU34-2 and LS30, all belonging to the culture collection of the Dipartimento di Agricoltura, Ambiente e Alimenti, Università del Molise, Italy. Strains LS16 and AU34-2 were selectively isolated from grape berries (De Curtis et al., 1996, 2004), while strain LS30 was isolated from apple fruits cv. Annurca (Lima et al., 1999), as described by Wilson et al. (1993). Strains AU34-2 and LS16 were identified as A. pullulans and M. pulcherrima, respectively, on a morphological basis (colony color, shape and microscopic observations). Strain LS30 had previously been identified by CBS (Centraalbureau voor Schimmelcultures, P.O. Box 85167, 3508 AD Utrecht, The Netherlands) as A. pullulans (accession number CBS 110902). This strain had previously been characterized for its activity against different pathogens on different crops (de Felice et al., 2008; Lima et al., 1999, 2003), and was also shown to actively degrade ochratoxin A. Strain LS30 was also characterized for its modes of action (Castoria et al., 2001, 2003) and at a genomic level through fAFLP analysis (De Curtis et al., 2004) The ochratoxigenic strain A1102 of the pathogenic fungus A. carbonarius was kindly provided by Università Cattolica del Sacro Cuore, Piacenza, Italy. 2.3. Infection of Aspergillus carbonarius on wounded and unwounded berries Healthy wine grape bunches (cv. Montepulciano) were purchased from a local producer. Single berries were first separated from the stalk by cutting the stem at about 0.5 cm from each berry in order to prevent desiccation, then sterilized by dipping into a 1% sodium hypochlorite solution (w/v, pH 11.5) for 1 min, washed twice with distilled sterile water and dried externally on absorbent paper in a laminar flow bench until inoculation with the fungal pathogen. A conidial suspension of A. carbonarius A1102 was prepared as follows: the pathogen was grown on a Petri dish containing PDA for 15 days at 28 °C, scraped from the plate and transferred to a Falcon tube containing sterile distilled water. The tube was stirred on a vortex, filtered, and centrifuged at 3200 g for 10 min. The pellet was suspended in sterile distilled water and the obtained conidial suspension was adjusted to the desired concentration. A single wound was made by using a sterile needle on half of the berries, at an equidistant point at about midway between emergence of the grape stalk and the equator. Each single berry, both wounded and unwounded, was sprayed with 35 μl of the conidial suspension of A. carbonarius. All berries inoculated with A. carbonarius were placed on plastic grates presterilized in a 70% (v/v) ethanol solution. Grates were placed into disinfected plastic boxes and incubated in different growth chambers at three different temperatures (20, 25 and 30 °C), and at two levels of relative humidity (60 and 100% RH), for 5 days. During incubation, temperature and RH were constantly monitored. Infections on berries were assessed either directly or by means of a stereomicroscope. Percentages of infected berries were recorded on the 5th day of incubation. Each treatment consisted of 3 replicates each consisting of 20 berries. The experiment was performed three times. 2.4. Biocontrol activity of Metschnikowia pulcherrima LS16 and Aureobasidium pullulans AU34-2 and LS30 against A. carbonarius on wine grape berries Cell suspensions of the BCAs were prepared as reported previously (Lima et al., 1998). Briefly, cells of each BCA were scraped from a 48 h-old colony grown in potato dextrose agar (PDA, Oxoid Ltd, Basingstoke, Hampshire, UK) medium, added to 50 ml of nutrient yeast dextrose broth medium (NYDB, 8 g of nutrient broth, 5 g of yeast extract, and 10 g of dextrose in 1 l of water) (Oxoid Ltd,

F. De Curtis et al. / International Journal of Food Microbiology 159 (2012) 17–24

Basingstoke, Hampshire, UK) and incubated overnight on a rotary shaker at 150 rpm and 23 °C. Cell concentration was then adjusted to 1 × 10 5 CFU/ml, the cell suspensions were added to 100 ml of NYDB medium and incubated on a rotary shaker at 150 rpm and 23 °C. After 2 days, cells of the BCAs were washed twice with sterile distilled water and the suspensions were adjusted to a final concentration of 1 × 10 8 CFU/ml. Mature wine grape berries (cv. Montepulciano) to be used in the biocontrol experiments were wounded at three equidistant points as described above. Berries were then treated with the BCAs by shaking for 5 min at 80 rpm in the cell suspensions of the three antagonists LS16, AU34-2, and LS30. After 1 h, berries were sprayed with a suspension of A. carbonarius A1102 conidia (5 × 10 4 conidia/ml) and incubated in growth chambers as described above at three different temperatures (20, 25 and 30 °C) and two different values of relative humidity (60 and 100% of RH) for 5 days. Temperature and relative humidity were constantly monitored and the percentage of infected wounds was recorded as described above after 1, 2, 4 and 5 days. Control treatment consisted of wine grape berries not pre-treated with the BCAs and inoculated with A. carbonarius A1102. Each treatment consisted of 5 replicates of 20 berries each. Experiments were performed three times. 2.5. Determination of ochratoxin A in wine grape berries Extraction and analyses of OTA were carried out as described by Visconti et al. (1999) in wine grape berries from all the treatments of all the biocontrol assays. Berries were homogenized by means of an Ultra-Turrax blender (IKAWerke GmbH & Co. KG, Staufen, Germany) at 14,000 rpm. Sample of 5 g were diluted with a solution (10 ml) of 5% (w/v) sodium hydrogen carbonate and 1% (w/v) polyethylene glycol (PEG 8000), mixed for 3 min and centrifuged at 4000 rpm for 10 min. Supernatants were filtered on Whatman GF/A glass microfiber filters in order to obtain a clear juice and the extracts were loaded on a preconditioned immunoaffinity column containing antibodies specific for OTA (Vicam OchraTest, Vicam, Watertown, MA). The column was washed with 5 ml aqueous solution containing 2.5% (w/v) sodium chloride and 0.5% (w/v) sodium hydrogen carbonate, followed by water (5 ml). OTA was eluted from the column with methanol (2 ml) and evaporated to dryness under a nitrogen stream at 50 °C. All extracts were reconstituted in acetonitrile/water/acetic acid (99/99/2, v/v/v). Aliquots of 50 μl of the purified extracts were injected into the chromatographic apparatus that consisted of a Kontron HPLC Pump 422 (Kontron Instruments, Watford, United Kingdom), a Kontron HPLC 465 Variable Volume Autosampler, and a Kontron HPLC SFM25 fluorescence detector set at excitation and emission wavelengths of 333 and 460 nm, respectively. All samples were analyzed by an Agilent Zorbax SB (4.6 × 50 mm —1.8 μ) C18 column. The mobile phase was pumped at a flow rate of 1.0 ml/min and consisted of the following isocratic system: 49.5% acetonitrile, 49.5% water and 1% acetic acid (v/v). Quantification of OTA was achieved by comparing areas of peaks with the same retention time as the standard OTA with respective calibration curves. 2.6. Population dynamics of Metschnikowia pulcherrima LS16 and Aureobasidium pullulans AU34-2 and LS30 on grape berries at different environmental conditions The population dynamics of the three BCAs were assessed on single berries that were prepared, wounded and treated separately with the BCAs as described for the experiments of biocontrol activity, but no challenge with A. carbonarius was performed. Berries were incubated at three different temperatures (20, 25 and 30 °C) and two different values of relative humidity (60 and 100% of RH) for 5 days as described above. For each BCA, for each value of

19

temperature and RH and for each day of sampling, the treatment consisted of 5 replicates of 20 berries. The level of colonization of berries by the biocontrol agents was monitored 1, 2, 4 and 5 days after the treatment. For this purpose, at each time point all the berries of each replicate were dipped in 500 ml of sterile distilled water and kept on a rotary shaker at 150 rpm for 30 min and 23 °C. One hundred microliters of washing suspension was accurately diluted and dispensed in Petri dishes containing NYDA. Plates were incubated at 23 °C and colonies of the BCAs were counted and expressed as log10 CFU/cm2of berry surface. Experiments were performed three times. 2.7. Statistical analysis Percentage values of infections in wounded and unwounded berries artificially inoculated with A. carbonarius A1102 and in experiments on biocontrol activity were converted into Bliss angular values (arcsin √ %) prior to statistical analysis. Transformed data were submitted to factorial analysis of variance (ANOVA) by using the PROC general linear models (GLM) of SPSS (version 20.0 for Windows; SPSS Inc., Chicago, IL). Means were separated by the least significant difference (LSD) test (P b 0.01). Data from OTA analyses were submitted to factorial analysis of variance as described above and means were separated by the least significant difference (LSD) test (P b 0.001). 3. Results 3.1. Infection of wounded and unwounded berries by A. carbonarius After 5 days of incubation, very low levels of infection by A. carbonarius A1102 occurred on unwounded berries, either at 60% or 100% of RH and at all temperature values (20, 25 and 30 °C) tested. Percentages of unwounded berries that were infected by the fungal pathogen ranged from 0.5% at 20 °C and 60% RH to 4.4% at 30 °C and 100% RH. In the group of wounded berries, the percentage of infected berries was significantly higher than that in the unwounded group: at 60% of RH, 76.5%, 92.9% and 96.1% of infected berries were recorded at 20, 25 and 30 °C, respectively; at 100% of RH, 85.7%, 100% and 100% of berries were infected at the same temperatures as above (Fig. 1A and B). 3.2. Biocontrol of A. carbonarius strain 1102 by the three BCAs on grape berries at different environmental conditions Factorial analysis of percentages of infected wounds showed that the effects of RH (60 and 100%), temperature (20, 25, 30 °C), treatments (strains LS16, AU34-2, and LS30), and their interactions, were highly significant (P b 0.001), except than the interaction RH × treatment, was not statistically significant (P = 0.112) (Table 1). Fig. 2 shows the activity of the three BCAs LS16, LS30 and AU34-2 in the prevention of A. carbonarius A1102 infections on detached wine grape berries in laboratory-scale experiments. At 60% RH and 20 °C, 53.6% of the wounds were infected by A. carbonarius in the absence of biocontrol agents after 5 days from the beginning of the experiments. Conversely, pre-treatment of the artificially wounded grape berries with anyone of the three BCAs almost completely prevented infection of wounds: strains LS16, AU34-2 and LS30 lowered infections by 99.0%, 95.9% and 92.0%, respectively, compared to the untreated control (Fig. 2A). At 60% RH and 25 °C, only AU34-2 retained some significant biocontrol activity after 5 days of incubation, reducing infections of A. carbonarius (92.8% of infected wounds) by 35.8% (Fig. 2B). At 30 °C, none of the tested strains were able to control pathogen infections by day 4 (100% of infected wounds), whereas only limited biocontrol activity was expressed after 2 days of incubation (Fig. 2C).

20

F. De Curtis et al. / International Journal of Food Microbiology 159 (2012) 17–24

by 68.6% compared to the untreated control. No protection was achieved by the three BCAs by day 4. 3.3. HPLC quantification of ochratoxin A in grape berries

Fig. 1. Percentages of infected berries on artificially wounded ( ) and unwounded ( ) wine grape berries (cv. Montepulciano) inoculated with A. carbonarius A1102 after 5 days of incubation at three different temperatures (20, 25 and 30 °C) and at 60% (A) and 100% (B) of relative humidity. Values are means ± standard deviations of three different experiments (n = 9). Percentage values were converted into Bliss angular values (arcsin √ %) prior to statistical analysis.

At 100% RH, infections in wounded berries developed much faster than those observed at 60% RH at all temperatures tested. On day 4 at 20 °C, almost 93% of wounds were already infected by A. carbonarius in control berries not pre-treated with the biocontrol strains (Fig. 2D). Only at this temperature, strains LS16, AU34-2, and LS30 effectively protected grape berries until the fifth day of incubation. The three BCAs reduced A. carbonarius infections by 87.0% (strain LS16), 81.7% (strain AU34-2) and 69.1% (strain LS30) compared to the untreated control, in which all the wounds were infected (Fig. 2D). The biocontrol activity of all of the tested strains dramatically decreased at 25 °C and 30 °C. At 25 °C, protection of grape berries by the BCAs only lasted until day 2 (Fig. 2E). After 2 days at 30 °C (Fig. 2F), 100% of the wounds were already infected by A. carbonarius in the untreated control berries. At the same time point, only strain LS16 displayed a good biocontrol activity, significantly lowering the levels of infections

Table 1 Analysis of variance of the influence of relative humidity (RH, 60 and 100%), temperature (T, 20, 25, 30 °C) and treatments (strain LS16 of Metschnikowia pulcherrima, strains AU34-2, and LS30 of Aureobasidium pullulans) on the percentage of wounds of wine grape berries infected by Aspergillus carbonarius. Source of variability

df

Means of squares

F

P

Corrected model Intercept RH T Treatment RH × T RH × treatment T × treatment RH × T × treatment Error Total Corrected total

23 1 1 2 3 2 3 6 6 48 72 71

2,823.766 303,108.980 5,174.836 22,115.438 1,778.436 1,020.468 62.886 1,133.267 196.068 29.884

94.491 10,142.842 173.164 740.042 59.511 34.148 2.104 37.922 6.561

0.000 0.000 0.000 0.000 0.000 0.000 0.112 0.000 0.000

Factorial analysis of OTA accumulation data showed that the effects of RH (60 and 100%), temperature (20, 25, 30 °C), treatments (strains LS16, AU34-2, and LS30) and their interactions were highly significant (P b 0.001) (Table 2). Fig. 3 shows quantitative results of OTA analyses in the berries used for the biocontrol experiments, 5 days after inoculation with A. carbonarius. At 60% RH, the higher the temperature of incubation the higher the concentration of OTA detected in grape berries: levels of OTA contamination in berries not pre-treated with the BCAs were 2.8, 20.5 and 48.3 ng/g, at 20, 25 and 30 °C, respectively; at 20 °C, no OTA was detected in berries pre-treated with the biocontrol strains AU34-2 and LS16, whereas only 0.5 ng/g was recorded in berries pre-treated with LS30 (Fig. 3A); at 25 °C, pre-treatment of berries with LS16, AU34-2 and LS30 resulted in reductions of OTA contamination ranging from 77.5% (strain LS16) to 87.7% (strain LS30), as compared to the untreated control (Fig. 3A). At 30 °C, the concentrations of OTA in infected berries pre-treated with the BCAs were also significantly lower than in the control. Reductions of OTA accumulation ranged from 35.5% (strain AU34-2) to 70.8% (strain LS30) (Fig. 3A). At 100% RH and 20 °C, pre-treatments with LS16, AU34-2 and LS30 led to reductions of OTA contamination in grape berries ranging from 93.5% (strains LS16 and AU34-2) to 29.6% (strain LS30), as compared to untreated control (26.5 ng/g) (Fig. 3B). At 25 °C, AU34-2 was able to reduce OTA accumulation by 74.1%, as compared to the untreated control (49.0 ng/g). LS30 was less effective resulting in a 52.2% reduction of toxin concentration, while OTA reduction achieved by LS16 was not statistically different from the untreated control (Fig. 3B). Finally, none of the tested strains caused any reduction of OTA contamination at 30 °C, even if the amount of the mycotoxin produced by the pathogen on untreated berries was much lower (3.7 ng/g) than that produced at 20 and 25 °C. 3.4. Population dynamics of the three BCAs on grape berries at different environmental conditions The population dynamics of the BCAs on grape berries are shown in Fig. 4. The population dynamics appeared to have the same trend for the three microorganisms. The faster increases of berry surface colonization by the BCAs were achieved at 20 and 25 °C and 100% RH. On day 2, the values of log10 CFU/cm 2 ranged from 2.75 (strains AU34-2 and LS30 at 30 °C and 60% RH) to 4.20 (strain LS30 at 25 °C and 100%). The highest levels of colonization by the three BCAs were reached on day 5 at 20 °C and 100% RH: values of log10 CFU/cm2 ranged from 4.50 (strain LS16) to 4.55 (strain AU34-2). Lower levels of colonization were observed at 30 °C. In particular, a very slight increase was recorded at 60% RH on the days following treatment of berries with the BCAs, and the final values of log10 CFU/cm2 recorded on day 5 were much lower (more than an order of magnitude) than those observed at 20 and 25 °C and 100% RH. 4. Discussion The efficacy of different biocontrol agents (BCAs) in the control of phyllosphere, soil-borne and postharvest pathogens is well documented (Andrews, 1992; Andrews and Harris, 2000; Haas and Défago, 2005; Harman et al., 2004; Ippolito et al., 2004; Janisiewicz and Korsten, 2002; Lima and De Cicco, 2006; Sharma et al., 2009). This has made it possible to formulate and commercialize biofungicides based on these beneficial microorganisms (Castoria et al., 2008; Fravel,

F. De Curtis et al. / International Journal of Food Microbiology 159 (2012) 17–24

21

Fig. 2. Biocontrol activity of Metschnikowia pulcherrima strain LS16 ( ) and Aureobasidium pullulans strains LS30 ( ) and AU34-2 ( ) on artificially wounded wine grape berries (cv. Montepulciano) inoculated with Aspergillus carbonarius strain A1102. Biocontrol activity was expressed as the percentages of wounds infected after 1, 2, 4 and 5 days of incubation at 20, 25 and 30 °C and 60% of relative humidity (A, B, C, respectively), and at 20, 25 and 30 °C and 100% of relative humidity (D, E, F, respectively). The untreated control consisted of wounded berries not pre-treated with the BCAs prior to inoculation with A. carbonarius. Symbols at each time point represent mean values ± standard deviation of three different experiments (n = 9). Percentage values were converted into Bliss angular values (arcsin √ %) prior to statistical analysis.

2005; Paulitz and Bélanger, 2001). Several reports suggest that the utilization of BCAs is more promising in the storage phase of fruit, in which the different environmental parameters are under control (Janisiewicz and Korsten, 2002; Droby et al., 2009). Nevertheless, the preharvest application of postharvest BCAs fosters their colonization of carposphere and of fruit wounds (caused by harvest manipulations and representing the major site of infection of most postharvest pathogens), thus increasing their biocontrol efficacy during storage (Ippolito and Nigro, 2000; Nigro et al., 2006). Even more so, the application of BCAs in the field is necessary for the biocontrol of fungi that start colonization and pathogenic activity before harvest, which is the case of Aspergillus carbonarius in the vineyard, where environmental factors such as humidity and temperature cannot be controlled. Generally speaking, the biocontrol efficacy depends on the concentration of the BCAs, the amount of pathogen inoculum and the host species. In this regard, environmental factors play a key role, since they affect survival of the microbial species and the physiology/metabolism of all the players in the tritrophic interaction pathogen–BCA–host plant (El-Ghaouth et al., 2001; Lahlali et al., 2011). Metschnikowia pulcherrima strain LS16 and Aureobasidium pullulans strains AU34-2 and LS30 were previously reported to be effective in the biocontrol of different fungal pathogens, both in storage and in the field (De Curtis et al., 1996, 2012; de Felice et al., 2008; Lima et al., 1999, 2011). Strain LS30, in particular, was able to reduce infections by the ochratoxigenic fungal pathogen A. carbonarius in the

vineyard, thus reducing ochratoxin A (OTA) contamination in grape berries (de Felice et al., 2008). The goal of this study was to assess the influence of temperature and relative humidity (RH) on the biocontrol efficacy of the mentioned BCAs against A. carbonarius and the accumulation of OTA in wine grape berries. The biocontrol activity of these strains was assessed on wounded grape berries, since the presence of wounds on grape skin strongly favors infection of berries

Table 2 Analysis of variance of the influence of relative humidity (RH 60 and 100%), temperature (T, 20, 25, 30 °C), and treatments (strain LS16 of Metschnikowia pulcherrima, strains AU34-2, and LS30 of Aureobasidium pullulans) on ochratoxin A accumulation in wine grape berries. Source of variability

df

Means of squares

F

P

Corrected model Intercept RH T Treatment RH × T RH × treatment T × treatment RH × T × treatment Error Total Corrected total

23 1 1 2 3 2 3 6 6 48 72 71

683.652 14,317.140 356.000 1,102.752 986.520 3,170.502 228.609 222.565 306.784 10.384

65.839 1,378.802 34.284 106.200 95.006 305.333 22.016 21.434 29.545

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

22

F. De Curtis et al. / International Journal of Food Microbiology 159 (2012) 17–24

Fig. 3. Ochratoxin A (OTA) accumulation (ng/g) in wine grape berries (cv. Montepulciano) pre-treated with the biocontrol agents Metschnikowia pulcherrima strain LS16 ( ) and Aureobasidium pullulans strains LS30 ( ) and AU34-2 ( ) and challenged with Aspergillus carbonarius strain A1102. The untreated control consisted of berries that had not been pre-treated with the biocontrol agents prior to inoculation with the fungal pathogen. After pathogen inoculation, berries were incubated for 5 days at three temperatures (20, 25 and 30 °C) and at 60% (A) and 100% (B) relative humidity. Bars represent mean values±standard deviations of three different experiments.

by A. carbonarius, as shown by the percentage of infection that in wounded berries is dramatically higher than in unwounded ones, independent of the values of temperature and relative humidity. This result is in agreement with previous reports by other authors (Battilani et al., 2001; Bellí et al., 2007; Cozzi et al., 2006). The protection of wine grape berries from the challenging pathogen exerted by the three BCAs used in this study appears to decrease as RH and temperature increase, especially in the case of the two A. pullulans strains, AU34-2 and LS30. This is most likely due to the establishment of conditions that foster growth and infection of berries by A. carbonarius A1102. This is in agreement with previous reports showing that different strains of A. carbonarius, including A1102 used in this study, grow faster at higher temperatures and RH (or water activity) (Battilani et al., 2003; Mitchell et al., 2003, 2004). On the other hand, the BCAs probably do not cope with the same conditions that are favorable to A. carbonarius A1102. In particular, the three BCAs used in this study have their optimal growth temperatures at 23–24 °C, with poorer growth at higher temperatures, which is a trait required for registration of microbes to be used on food commodities. The lower fitness of the BCAs at higher temperature is supported by their lower colonization of berries that is recorded at 30 °C, which is only partially compensated at 100% RH. Therefore, the decrease of biocontrol activity at higher RH and temperature values examined in this work is likely due both to the higher aggressiveness of A. carbonarius A1102, which is witnessed by the much faster development of infections associated with the increase of temperature and RH (i.e. the faster achievement of 100% infection of wounds), and to the lower adaptability of the BCAs to these conditions. With regard to the OTA concentration measured after 5 days of incubation, the accumulation of this mycotoxin in control berries not

pre-treated with the BCAs does not completely mirror the aggressiveness of the pathogen, i.e. the percentage of infected wounds. This is in agreement with previous reports, which show that strains of A. carbonarius have optimal temperatures for growth that are higher than those for OTA biosynthesis (Abarca et al., 1994; Bellí et al., 2007; Esteban et al., 2004; Varga et al., 1996). In particular, strain A1102 of A. carbonarius used in this study appears to synthesize more OTA when temperature and RH increase, with the only exception at 100% of RH and 30 °C; in these conditions, the fungus synthesizes OTA at a level that is similar to the lowest one recorded at 60% of RH and 20 °C. Mitchell et al. (2004) reported an analogous behavior of this strain in vitro, where A1102 displayed augmented OTA synthesis when water activity was increased from 0.93 to 0.95, and a decrease of mycotoxin production when water activity was further increased to 0.98. With regard to berries pre-treated with the BCAs, a clear decrease of OTA concentration is recorded in all conditions of incubation, as compared to respective untreated controls, except for strain LS16 at 100% RH and 25 °C and for all the BCAs at 100% RH and 30 °C. Interestingly, the OTA decrease is recorded both in the presence of an effective protection of berries by the BCAs (as in the case of all the biocontrol treatments followed by incubation at 60% or 100% RH and 20 °C, and with the BCA AU34-2 at 60% RH and 25 °C) and in the absence of such protection (as recorded in all other treatments except for 30 °C and 100% of RH). Different hypotheses could explain the reduction of OTA accumulation recorded even when berries are not protected by the BCAs after 5 days, but it must be emphasized that all of these hypotheses need to be confirmed. BCAs could cause accelerated breakdown and/or lower synthesis of OTA by A. carbonarius A1102; OTA breakdown is known to be carried out by the ochratoxigenic fungi themselves (Abrunhosa et al., 2002; Bejaoui et al., 2006), and it has been reported that transcription of Aspergillus westerdijkiae genes involved in OTA biosynthesis is lower in the presence of the yeast Debaryomyces hansenii CYC 1244 (Gil-Serna et al., 2011). The BCAs could also cause a delay in the actual growth of A. carbonarius biomass associated with OTA synthesis. The BCAs could even degrade OTA, since two out of three strains used in this study (LS30 and AU34-2) and other biocontrol yeasts are able to degrade this mycotoxin to less toxic ochratoxin α in vitro and/or in grape must (de Felice et al., 2008; Patharajan et al., 2011). Whatever the mechanism of OTA reduction, the positive effect of BCAs on berry contamination is remarkable and in line with our previous results obtained in vineyard treatments (de Felice et al., 2008). Our results show that biocontrol treatment positively affects the incidence of infection by A. carbonarius on wine grape berries. A positive effect of biocontrol treatment is also recorded on the level of OTA contamination, even when protection from infection is not satisfactory. The positive effects of biocontrol are dependent on environmental factors. In particular, temperature appears to be more important than relative humidity for protection efficacy. Furthermore, protection efficacy by the biocontrol strains decreases or is absent when environmental conditions foster infection by the pathogen. This emphasizes the need for BCAs that are able to cope with the environmental conditions that are more conducive to colonization and infection of grape berries by A. carbonarius.

Acknowledgments We are grateful to Prof. Paola Battilani for the supply of strain A1102 of Aspergillus carbonarius. This research was funded by the Italian Ministry for University and Scientific Research (MIUR), through the project PRIN 2005071422 entitled “Biological and integrated control of Aspergillus carbonarius: effectiveness on ochratoxin A content and grape-wine chain”.

F. De Curtis et al. / International Journal of Food Microbiology 159 (2012) 17–24

5.00

A

4.50

4.50

4.00

4.00

3.50

3.50

3.00

3.00

2.50

2.50

2.00

2.00

5.00

Log CFU /cm2

5.00

20 oC - 60% RH

B

4.50

4.00

4.00

3.50

3.50

3.00

3.00

2.50

2.50

2.00

2.00

C

20 oC - 100% RH

E

25 oC - 100% RH

F

30 oC - 100% RH

5.00

25 oC - 60% RH

4.50

5.00

D

5.00

30 oC - 60% RH

4.50

4.50

4.00

4.00

3.50

3.50

3.00

3.00

2.50

2.50

23

2.00

2.00 0

1

2

3

4

0

5

1

2

3

4

5

Days Fig. 4. Population dynamics of the biocontrol agents Metschnikowia pulcherrima strain LS16 ( ) and Aureobasidium pullulans strains LS30 ( ) and AU34-2 ( ) on artificially wounded wine grape berries (cv. Montepulciano) after 1, 2, 4 and 5 days of incubation at 20, 25 and 30 °C and 60% relative humidity (A, B, C, respectively), and at 20, 25 and 30 °C and 100% relative humidity (D, E, F, respectively). Population dynamics was expressed as log10 CFU/cm2 of the berry surface. Symbols at each time point represent mean values ± standard deviations of three different experiments (n = 15).

References Abarca, M.L., Bragulat, M.R., Castella, G., Cabañes, F.J., 1994. Ochratoxin A production by strains of Aspergillus niger var. niger. Applied and Environmental Microbiology 60, 2650–2652. Abrunhosa, L., Serrs, R., Venancio, A., 2002. Biodegradation of ochratoxin A by fungi isolated from grapes. Journal of Agricultural and Food Chemistry 50 (25), 7493–7496. Andrews, J.H., 1992. Biological control in the phyllosphere. Annual Review of Phytopathology 30, 603–635. Andrews, J.H., Harris, R.F., 2000. The ecology and biogeography of microorganisms on plant surfaces. Annual Review of Phytopathology 38, 145–180. Battilani, P., Giorni, P., Bertuzzi, T., Pietri, A., 2001. Preliminary results on ochratoxin A production by fungi invading grape berries. Proceedings 11th Congress of Mediterranean Phytopathological Union and 3rd Congress of Sociedade Portuguesa de Fitopatologia, Evora, Portugal, pp. 120–122. Battilani, P., Pietri, A., Giorni, P., Bertuzzi, T., Barbano, C., 2003. Growth and ochratoxin A production of Aspergillus section Nigri isolates from Italian grapes. Aspects of Applied Biology 68, 175–180. Battilani, P., Giorni, P., Bertuzzi, T., Formenti, S., Pietri, A., 2006. Black aspergilla and ochratoxin A in grapes in Italy. International Journal of Food Microbiology 111, S53–S60. Bejaoui, H., Mathieu, F., Taillandier, P., Lebrihi, A., 2006. Biodegradation of ochratoxin A by Aspergillus section Nigri species isolated from French grapes: a potential means of ochratoxin A decontamination in grape juice sand musts. FEMS Microbiology Letters 255, 203–208. Bellí, N., Marín, S., Coronas, I., Sanchis, V., Ramos, A.J., 2007. Skin damage, high temperature and relative humidity as detrimental factors for Aspergillus carbonarius infection and ochratoxin A production in grapes. Food Control 18, 1343–1349. Bennett, J.W., Klich, M., 2003. Mycotoxins. Clinical Microbiology Reviews 16, 497–516. Cabras, P., Meloni, M., Pirisi, F.M., 1987. Pesticide fate from vine to wine. Reviews of Environmental Contamination and Toxicology 99, 83–117. Cabras, P., Angioni, A., Garau, V.L., Melis, M., Pirisi, F.M., Minelli, E.V., Cabitza, F., Cubeddu, M., 1997. Fate of some new fungicides (cyprodinil, fludioxonil, pyrimethanil, and tebuconazole) from vine to wine. Journal of Agricultural and Food Chemistry 45, 2708–2710. Castoria, R., De Curtis, F., Lima, G., Caputo, L., Pacifico, S., De Cicco, V., 2001. Aureobasidium pullulans (LS30) an antagonist of postharvest pathogens of fruits: study on its modes of action. Postharvest Biology and Technology 22, 7–17.

Castoria, R., Caputo, L., De Curtis, F., De Cicco, V., 2003. Resistance of postharvest biocontrol yeasts to oxidative stress: a possible new mechanism of action. Phytopathology 93, 564–572. Castoria, R., Morena, V., Caputo, L., Panfili, G., De Curtis, F., De Cicco, V., 2005. Effect of the biocontrol yeast Rhodotorula glutinis strain LS11 on patulin accumulation in stored apples. Phytopathology 95, 1271–1278. Castoria, R., Wright, S.A.I., Droby, S., 2008. Biological control of mycotoxigenic fungi in fruits. In: Barkai-Golan, R., Paster, N. (Eds.), Mycotoxins in Fruits and Vegetables. Elsevier, San Diego. ISBN: 978-012-374126-4, pp. 311–333. Castoria, R., Mannina, L., Duran-Patron, R., Maffei, F., Sobolev, A.P., de Felice, D.V., PinedoRivilla, C., Ritieni, A., Ferracane, R., Wright, S.A.I., 2011. Conversion of the mycotoxin patulin to the less toxic desoxypatulinic acid by the biocontrol yeast Rhodosporidium kratochvilovae strain LS11. Journal of Agricultural and Food Chemistry 59, 11571–11578. Commission of the European Communities, 2005. Commission regulation (EC) No 123/2005 of 26 January 2005 amending regulation (EC) No 466/2001 as regards ochratoxin A (Text with EEA relevance). Official Journal of the European Communities L25/3. Commission of the European Communities, 2006. Commission regulation (EC) No 1881/2006, setting maximum levels for certain contaminants in foodstuffs. Official Journal of the European Union L364/5. Commission of the European Communities, 2008. Commission regulation (EC) No. 149/ 2008 of 29 January 2008 amending regulation (EC) No. 396/2005 of the European parliament and of the council by establishing Annexes II, III and IV setting maximum residue levels for products covered by Annex I thereto. Official Journal of the European Union L58/1. Cozzi, G., Pascale, M., Perrone, G., Visconti, A., Logrieco, A., 2006. Effect of Lobesia botrana damages on black aspergilli rot and ochratoxin A content in grapes. International Journal of Food Microbiology 111S1, S88–S92. De Curtis, F., Torriani, S., Rossi, F. De, Cicco, V., 1996. Selection and use of Metschnikowia pulcherrima as a biological control agent for postharvest rots of peaches and table grapes. Annali di Microbiologia ed Enzimologia 46, 45–55. De Curtis, F., Caputo, L., Castoria, R., Lima, G., Stea, G., De Cicco, V., 2004. Use of fluorescent amplified fragment length polymorphism (fAFLP) for molecular characterization of the biocontrol agent Aureobasidium pullulans strain LS30. Postharvest Biology and Technology 34, 179–186. De Curtis, F., De Cicco, V., Lima, G., 2012. Efficacy of biocontrol yeasts combined with calcium silicate or sulphur for controlling durum wheat powdery mildew and increasing grain yield components. Field Crops Research 134, 36–46.

24

F. De Curtis et al. / International Journal of Food Microbiology 159 (2012) 17–24

de Felice, D.V., Solfrizzo, M., De Curtis, F., Lima, G., Visconti, A., Castoria, R., 2008. Strains of Aureobasidium pullulans can lower ochratoxin A contamination in wine grapes. Phytopathology 98 (12), 1261–1270. Droby, S., Chalutz, E., Wilson, C.L., Wisniewski, M.E., 1989. Characterization of the biocontrol activity of Debaryomyces hansenii in the control of Penicillium digitatum on grapefruit. Canadian Journal of Microbiology 35, 794–800. Droby, S., Wisniewski, M.E., Macarisin, D., Wilson, C., 2009. Twenty years of postharvest biocontrol research: it is time for a new paradigm? Postharvest Biology and Technology 52, 137–145. El-Ghaouth, A., Smilanick, J.L., Brown, G.E., Ippolito, A., Wilson, C.L., 2001. Control of decay of apple and citrus fruits in semicommercial tests with Candida saitoana and 2-deoxy-D-glucose. Biological Control 20 (2), 96–101. Esteban, A., Lourdes Abarca, M., Rosa Bragulat, M., Javier Cabañes, F., 2004. Effects of temperature and incubation time on production of ochratoxin A by black aspergilla. Research in Microbiology 155, 861–866. Fravel, D.R., 2005. Commercialization and implementation of biocontrol. Annual Review of Phytopathology 43, 337–359. Gil-Serna, J., Patiño, B., Cortés, L., González-Jaén, M.T., Vázquez, C., 2011. Mechanisms involved in reduction of ochratoxin A produced by Aspergillus westerdijkiae using Debaryomyces hansenii CYC 1244. International Journal of Food Microbiology 151, 113–118. González-Rodríguez, R.M., Cancho-Grande, B., Simal-Gándara, J., 2009. Multiresidue determination of 11 new fungicides in grapes and wines by liquid–liquid extraction/clean-up and programmable temperature vaporization injection with analyte protectants/gas chromatography/ion trap mass spectrometry. Journal of Chromatography. A 1216, 6033–6042. Haas, D., Défago, G., 2005. Biological control of soil-borne pathogens by fluorescent Pseudomonas. Nature Reviews. Microbiology 3, 307–319. Harman, G.E., Howell, C.R., Viterbo, A., Chet, I., Lorito, M., 2004. Trichoderma species— opportunistic, avirulent plant symbionts. Nature Reviews. Microbiology 2, 43–56. Hult, K., Plestina, R., Habazin Novak, V., 1982. Ochratoxin A in human blood and Balkan endemic nephropathy. Archives of Toxicology 51 (4), 313–321. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, 1993. Some Naturally Occurring Substances: Food Items and Constituents. : Heterocyclic Aromatic Amines and Mycotoxins, Vol. 56. International Agency for Research on Cancer, Lyon, France, pp. 489–521. Ippolito, A., Nigro, F., 2000. Impact of preharvest application of biological control agents on postharvest diseases of fresh fruits and vegetables. Crop Protection 19, 715–723. Ippolito, A., Nigro, F., Schena, L., 2004. Control of postharvest diseases of fresh fruits and vegetables by preharvest application of antagonistic microorganisms. In: Dris, R., Niskanen, R., Jain, S.M. (Eds.), Crop Management and Postharvest Handling of Horticultural Products. : Disease and Disorders of Fruits and Vegetables, 4. Science Publisher Inc., Enfield, NH, USA, pp. 1–30. Janisiewicz, W.J., Korsten, L., 2002. Biological control of postharvest diseases of fruits. Annual Review of Phytopathology 40, 411–441. Krogh, P., Axelsen, N.H., Elling, F., Gyrd-Hansen, N., Hald, B., Hyldgaard-Jensen, J., Larsen, A.E., Madsen, A., Mortensen, H.P., Moller, T., Petersen, O.K., Ravnskov, U., Rostgaard, M., Aalund, O., 1974. Experimental porcine nephropathy. Changes of renal function and structure induced by ochratoxin A-contaminated feed. Acta Pathologica, Microbiologica et Immunologica Scandinava Section A 246, 1–21. Lahlali, R., Hamadi, Y., El Guilli, M., Jijakli, M.H., 2011. Efficacy assessment of Pichia guilliermondii strain Z1, a new biocontrol agent, against citrus blue mould in Morocco under the influence of temperature and relative humidity. Biological Control 56, 217–224. Leong, S.L., Hocking, A.D., Scott, E.S., 2007. Aspergillus producing ochratoxin A: isolation from vineyard soils and infection of Semillon bunches in Australia. Journal of Applied Microbiology 102, 124–133. Lima, G., De Cicco, V., 2006. Integrated strategies to enhance biological control of postharvest diseases. In: Berkeblia, N., Shiomi, N. (Eds.), Advances in Postharvest Technologies for Horticultural Crops. Research Signpost, Kerala, India, pp. 173–194. Chapter 8.

Lima, G., De Curtis, F., Castoria, R., De Cicco, V., 1998. Activity of the yeasts Cryptococcus laurentii and Rhodotorula glutinis against postharvest rots on different fruits. Biocontrol Science and Technology 8, 269–279. Lima, G., Arru, S., De Curtis, F., Arras, G., 1999. Influence of antagonist, host fruit and pathogen on the biological control of postharvest fungal diseases by yeasts. Journal of Industrial Microbiology & Biotechnology 23, 223–229. Lima, G., De Curtis, F., Castoria, R., De Cicco, V., 2003. Integrated control of apple postharvest pathogens and survival of biocontrol yeasts in semi-commercial conditions. European Journal of Plant Pathology 109, 341–349. Lima, G., Castoria, R., De Curtis, F., Raiola, A., Ritieni, A., De Cicco, V., 2011. Integrated control of blue mould using new fungicides and biocontrol yeasts lowers levels of fungicide residues and patulin contamination in stored apples. Postharvest Biology and Technology 60, 164–172. Magan, N., Aldred, D., 2005. Conditions of formation of ochratoxin A in drying, transport and in different commodities. Food Additives and Contaminants 1, 10–16. Miraglia, M., Brera, C., 2002. Assessment of dietary intake of ochratoxin A by the population of EU Member States, January. http://europa.eu.int/comm/food/fs/scoop/ index_en.html. Mitchell, D., Aldred, D., Magan, N., 2003. Impact of ecological factors on growth and ochratoxin A production by Aspergillus carbonarius from different regions of Europe. Aspects of Applied Biology 68, 109–116. Mitchell, D., Parra, R., Aldred, D., Magan, N., 2004. Water and temperature relations of growth and ochratoxin A production by Aspergillus carbonarius strains from grapes in Europe and Israel. Journal of Applied Microbiology 97 (2), 439–445. Nigro, F., Schena, L., Ligorio, A., Pentimone, I., Ippolito, A., Salerno, M.G., 2006. Control of table grape storage rots by pre-harvest applications of salts. Postharvest Biology and Technology 42, 142–149. Oueslati, S., Lasram, S., Ramos, A.J., Marin, S., Mliki, A., Sanchis, V., Ghorbel, A., 2010. Alternating temperatures and photoperiod effects on fungal growth and ochratoxin A production by Aspergillus carbonarius isolated from Tunisian grapes. International Journal of Food Microbiology 139, 210–213. Patharajan, S., Reddy, K.R.N., Spadaro, D., Lore, A., Gullino, M.L., Garibaldi, A., Karthikeyan, V., 2011. Potential of yeast antagonists on in vitro biodegradation of ochratoxin A. Food Control 22, 290–296. Paulitz, T.C., Bélanger, R.R., 2001. Biological control in greenhouse systems. Annual Review of Phytopathology 39, 103–133. Pfohl-Leszkowicz, A., Petkova-Bocharova, T., Chernozemsky, I.N., Castegnaro, M., 2002. Balkan endemic nephropathy and associated urinary tract tumours: a review on aetiological causes and the potential role of mycotoxins. Food Additives and Contaminants 19 (3), 282–302. Ponsone, M.L., Chiotta, M.L., Combina, M., Dalcero, A., Chulze, S., 2011. Biocontrol as a strategy to reduce the impact of ochratoxin A and Aspergillus section Nigri in grapes. International Journal of Food Microbiology 151, 70–77. Romero, S.M., Comerio, R.M., Larumbe, G., Ritieni, A., Vaamonde, G., Fernández Pinto, V., 2005. Toxigenic fungi isolated from dried vine fruits in Argentina. International Journal of Food Microbiology 104 (1), 43–49. Sharma, R., Singh, D., Singh, R., 2009. Biological control of postharvest diseases of fruits and vegetables by microbial antagonists: a review. Biological Control 50, 205–221. Varga, J., Kevei, E., Rinyu, E., Teren, J., Kozakiewicz, Z., 1996. Ochratoxin production by Aspergillus species. Applied and Environmental Microbiology 62, 4451–4464. Visconti, A., Pascale, M., Centonze, G., 1999. Determination of ochratoxin A in wine by means of immune affinity column clean-up and high-performance liquid chromatography. Journal of Chromatography. A 864, 89–101. Wilson, C.L., Wisniewski, M.E., Droby, S., Chalutz, E., 1993. A selection strategy for microbial antagonists to control postharvest diseases of fruits and vegetables. Scientia Horticulturae 53, 183–189. Zimmerli, B., Dick, R., 1996. Ochratoxin A in table wine and grape-juice: occurrence and risk assessment. Food Additives and Contaminants 13, 655–668.