1879 Journal o f Food Protection, Vol. 76, No. 11, 2013, Pages 1879-1886 doi:10.4315/0362-028X.JFP-13-072 C o pyright © , Internatio nal A sso cia tio n fo r Food Protection
Control of Postharvest Fungal Pathogens by Antifungal Compounds from Penicillium expansum WAFA ROUTSSI,1 LUISA UGOLINI,2 CAMILLA MARTINI,1 LUCA LAZZERI,2 a n d MARTA MARI'' 1Criof’, University o f Bologna, Via Gandolfi 19,40057 Cadriano, Bologna, Italy; and 2Consiglio per la Ricerca e la Sperimentazione in Agricoltura—Centro di Ricerca per le Colture Industriali (CRA-CIN), Via di Corticella 133, 40128 Bologna, Italy MS 13-072: Received 24 February 2013/Accepted 4 June 2013
ABSTRACT The fungicidal effects of secondary metabolites produced by a strain of Penicillium expansum (R82) in culture filtrate and in a double petri dish assay were tested against one isolate each of Botrytis cinerea, Colletotrichum acutatum, and Monilinia laxa and six isolates of P. expansum, revealing inhibitory activity against every pathogen tested. The characterization of volatile organic compounds released by the R82 strain was performed by solid-phase microextraction-gas chromatographic techniques, and several compounds were detected, one of them identified as phenethyl alcohol (PEA). Synthetic PEA, tested in vitro on fungal pathogens, showed strong inhibition at a concentration of 1,230 ug/ml of airspace, and mycelium appeared more sensitive than conidia; nevertheless, at the concentration naturally emitted by the fungus (0.726 + 0.16 ug/ml), commercial PEA did not show any antifungal activity. Therefore, a combined effect between different volatile organic compounds produced collectively by R82 can be hypothesized. This aspect suggests further investigation into the possibility of exploiting R82 as a nonchemical alternative in the control of some plant pathogenic fungi.
After harvest, fruit are cold stored in order to prevent decay, but economically important losses caused by fungal pathogens can occur during storage. Gray mold (Botrytis cinerea Pers.: Fr.) is a very common pathogen of pome, stone, and kiwifruit (6), brown rot (Monilinia spp.) is the most important disease of stone fruit (3), and blue mold rot (.Penicillium expansion Link) is responsible for the losses in stored pear and apple (12). Colletotrichum acutatum J.H. Simmonds causes anthracnose on strawberry (21) and is frequently associated with bitter rot in apple (13). It was recently detected for the first time on Italian apple (22). In the past, the postharvest application of synthetic fungicides has been an important strategy in reducing postharvest losses, although the possibility of controlling these patho gens with nonsynthetic chemical means, such as microbial antagonists, natural bioactive compounds, and physico chemical methods, is now being assiduously investigated and reviewed (24, 35). According to Hegde et al. (9), the emphasis is on the exploitation of unusual and previously ignored ecosystems (microorganisms and plant extracts), such as secondary metabolites produced by microorganisms, which could be one of the important alternative manage ment strategies. Secondary metabolites can be produced by almost all types of living organisms, including bacteria and fungi. In the domain of Bacteria, Bacillus and Pseudomonas species are the most frequent producers (2). Several studies have reported the activities of their secondary metabolites Author for correspondence. Tel: +39051766563; Fax: +39051765049; E-mail: [email protected].
against the growth of postharvest fungal pathogens in citrus (34), in peach (30), and in apple and pear (11). Among the microscopic fungi, the ability of imperfect fungi, ascomycetes, and several other filamentous and endophytic fungal species to produce bioactive products is the most significant. The total number of bioactive fungal compounds is approximately 8,600, representing 38% of all microbial products: 950, 900, and 350 compounds have been isolated from the most common ascomycetes, such as Aspergillus, Penicillium, and Fusarium species, respectively (2), and some of them are bioactive against some phytopathogens (29). The production of such a diversified array of active secondary metabolites by Penicillium species is well documented. Culture filtrates of isolates belonging to the species Penicillium canescens and Penicillium janczewskii showed inhibitory activities against Rhizoctonia solani (28), while citrine in the filtrate of P. citrinum inhibited Sclerotinia minor growth (26). A new antifungal compound isolated from P. expansum reduced the growth of Lasiodiplodia theobromae in fungicide disk assays (8). In a previous work, the P. expansum strain R82 showed inhibitory activity in vitro against some postharvest pathogens (31). The aims of the present study were (i) to evaluate the antifungal activities of secondary metabolites produced by the R82 strain in culture filtrate and in the atmosphere (volatile organic compounds [VOCs]) against B. cinerea, C. acutatum, M. laxa, and P. expansum, (ii) to investigate the compound responsible for the observed antifungal activity by the solid-phase microextraction (SPME)-gas chromatographic technique, (iii) to quantify
1880
ROUISSI ET AL.
the most-representative VOC identified, and (iv) to test its antifungal effect in vitro on the previously reported fungal pathogens. MATERIALS AND METHODS P. expansum stra in R82. P. expansum strain R82, derived from the C riof collection, was cultured on malt extract agar (MEA) for 2 days at 20°C in order to obtain a conidial suspension. The conidial suspension was prepared by washing the surface o f the culture with 5 ml of sterile distilled water amended with Tween 80 (0.05%, vol/vol) and rubbing with a glass rod. Spores were collected in 5 ml o f distilled water, counted using a hematocytometer, and adjusted to 103 conidia per ml. For fungitoxicity in vitro assays and to establish the phenethyl alcohol (PEA) production kinetics, conidial suspensions o f R82 strain were prepared at concentrations of 105 and 106 conidia per ml, respectively, chosen after preliminary trials. Pathogens. B. cinerea, C. acutatum, M. laxa, and six P. expansum strains used in the experiments were isolated individ ually from infected fruit tissue, identified by PCR analysis to sequence rDNA internal transcribed spacer regions, and kept at 4°C on MEA as single spore cultures until use. In order to obtain a conidial suspension, each pathogen was grown under the following conditions: B. cinerea was grown on oatmeal agar (60 g of oatmeal, 10 g o f sodium nitrate, 30 g of saccharose, and 12 g of agar per 1,000 ml of distilled water) and the cultures were incubated at 25°C under UV (350 to 420 nm) light for 12 h daily to favor sporulation; C. acutatum was cultured on potato dextrose agar at 25°C for 10 days; M. laxa was grown on V8 agar (250 ml of V8 vegetable juice and 40 g of agar per 1,000 ml of distilled water) and incubated at 25°C with 12-h-dark/l2-h-light cycles for 10 days; and P. expansum was cultured on MEA for 3 days at 20°C.
In vitro fungitoxicity assays. The activities o f the secondary metabolites produced by the R82 strain against target pathogens were studied using the liquid medium test to determine the percentage of mycelium growth inhibition and the double petri dish assay in order to evaluate the production of VOCs. The mycelium growth inhibition assay was carried out in flasks containing 20 ml of malt extract broth (MEA without agar) that were inoculated with the R82 conidial suspension (106 conidia per ml) and incubated at 20°C. After 10 days, the mycelium was separated from the medium by centrifugation at 4,800 x g for 20 min, and the supernatant was fdtered through a sterile filter (0.45 pm). Aliquots o f 20 ml from the culture filtrate o f R82 were added to sterile 50-ml flasks, and then the flasks were inoculated with 100 pi of the single conidial suspension (103 conidia per ml) o f B. cinerea, C. acutatum, M. laxa, or P. expansum and incubated at 20°C for 7 days. The contents o f the flasks were then filtered using a Whatman filter number 1 (preconditioned overnight at 80°C) and dried in an oven at 80°C until reaching a constant weight in order to determine the mycelium dry weight. Flasks containing 20 ml of malt extract broth were inoculated with target pathogens and used as the control. The sample unit was represented by three replications (flask) for each pathogen. The trials were repeated twice. In the double petri dish assay, the antifungal activities of VOCs produced by the R82 strain were tested against the mycelial growth and conidial germination of B. cinerea, C. acutatum, M. laxa, and P. expansum by a dual-culture technique. For this purpose, a 6-mm agar plug was removed from the edge of an actively growing colony of each pathogen cited above and
J. Food Prot., Vol. 76, No. 11
inoculated in a MEA dish. Subsequently, the lid o f the plate was replaced with a base plate o f MEA inoculated with R82, and the plate was incubated at 20°C for 2 days. The two base plates were sealed immediately, using a double layer of Parafilm to make a closed chamber of almost 180 cm3 in volume. For the inhibition of conidial germination, aliquots (100 pi per dish) o f a conidial suspension (103 conidia per ml) of each pathogen were spread on MEA plates. The lids o f the MEA plates were removed and replaced with base plates containing previously inoculated R82 as described above. The two base plates were sealed immediately, using a double layer of Parafilm as described above. For both assays, the control treatment was inoculating dishes with pathogens only and sealing them with Parafilm. The percentages of inhibition o f growth diameter and o f conidial germination were assessed after 7 and 3 days of incubation, respectively, at 20°C. The percentages o f inhibition o f growth diameter and of conidial germination were calculated according to the formula of Trivedi et al. (38), as follows: [(Tl — T2)/T1] x 100, where T1 is the diameter of growth or CFU o f the target pathogen not exposed to R82 (control) and T2 is the diameter of growth or CFU o f the target pathogen exposed to R82. The assay was conducted in five replicates (dishes) and repeated twice.
Identification of volatile compounds. Headspace volatiles from R82 were qualitatively evaluated and identified by SPME (36) coupled with gas chromatography-mass spectrometry (GCMS) (GCD System, model G1800A, Agilent-Hewlett-Packard, Santa Clara, CA). Fresh cultures of the R82 strain were prepared following the same protocol used for the double petri dish assay. The needle o f the SPME device, containing the extraction fiber coated with 85-gm polyacrylate film or 100-pm polydimethylsiloxane (Supelco, Bellefonte, PA), was inserted into each plate through a small hole, and the fiber was exposed to the gas phase for 20 min at 20°C. The SPME device was then removed from the petri dish and inserted into the gas chromatograph injector port. Thermal desorption o f compounds extracted from the fiber was performed at 250°C for 2 min, and the subsequent separation of compounds was achieved through an HP-5MS Varian capillary column (30-m length by 0.25-mm inner diameter and 0.25-pm film thickness) at a flow rate of 1 ml/min with helium as the carrier gas. The column temperature was set at 40°C for 3 min and then programmed to rise from 40 to 300°C at 20°C/min. The temperatures of the injection port and ion source were set at 250 and 280°C, respectively; splitless injection mode and electron impact ionization (70 eV) were established. The VOCs were identified by considering their mass spectra and their retention times in comparison with those o f reference substances in the National Institute of Standards and Technology (NIST) library, 1992. Blank samples, represented by plates prepared as described above but without the R82 inoculum, were analyzed following the same method in order to exclude interfering substances coming from the medium. PEA (International Union o f Pure and Applied Chemistry name, 2-phenylethanol) was one of the volatiles observed, and its identification was achieved by injecting pure PEA standard (Chemical Abstracts Service no. 60-12-8, Sigma Aldrich, St. Louis, MO) using the same chromatographic conditions and confirming the mass spectrum and retention time. Effects of pure PEA on mycelium growth and conidial germination of fungal pathogens. The antifungal activities of pure PEA, the most-representative VOC produced by the R82 strain, identified by SPME-GC-MS analysis as described above, against the mycelium growth and conidial germination of target pathogens were assayed. A plug (6-mm diameter) from an actively
J. Food Prot., Vol. 76, No. 11
growing pathogen culture or 100 pi of a conidial suspension were placed or spread, respectively, in the center of MEA plates. In each case, different aliquots of pure PEA corresponding to 77, 148, 308, 615, and 1,230 pg/ml of headspace were placed, using a microsyringe, on a paper filter (Whatman no. 1, 90-mm diameter) positioned inside the cover. The dishes were quickly closed, sealed with Parafilm, and incubated at 20°C. The dishes were opened and evaluated after 7 days for mycelium growth and after 3 days for conidial germination. Petri dishes inoculated with pathogens but treated with distilled water instead of PEA were used as the control. Five dishes were prepared for each dose and pathogen, and the experiment was repeated twice. Mycelium growth (mm) was gauged with a tape measure. Conidial germination was determined by counting the CFU developed on MEA. PEA production kinetics. A method was defined to quantitatively establish the production kinetics of PEA released by R82. One hundred microliters of a conidial suspension of R82 (105 conidia per ml) was spread on MEA plates and incubated at 20°C. The fungal PEA production was followed by headspace SPME coupled with GC-flame ionization detector (FID) analysis. Sampling was performed for 100 h at time intervals of 48, 52, 65, 72, 76, 89, 96, and 100 h after inoculation. The same assay was performed on CADRP28, a P. expansum strain chosen randomly from strains of the CRIOF-UNIBO collection. Strain CADRP28 is a producer of the same major volatiles emitted by R82 as confirmed through SPME-GC-MS (data not shown). The SPME sampling was performed in the same way as described above. An Agilent 7820A GC-FID and a Varian HP-5 capillary column (30-m length by 0.25-mm inside diameter, 0.25pm film thickness) were used for the chromatographic analysis. The instrument settings were as follows: the injector and detector temperatures were 250 and 300°C, respectively, the oven program started at 40°C for 3 min and was raised to 300°C at a rate of 20°C/ min, the flow rate of the earner gas (He) was 1 ml/min, and the splitless injection mode was established. Solutions of an individual standard synthetic PEA were prepared in the laboratory to correctly identify natural PEA by GC-FID. The production kinetics of the naturally released PEA was established by determining the mean value of five areas measured at each sampling time. Three replications were performed for each trial. PEA quantification. A calibration curve for quantification of the naturally produced PEA was established by SPME-GC-FID analysis using synthetic PEA as the standard. For this aim, 5 pi of the synthetic PEA water solutions at different concentrations were injected into closed MEA petri dishes through holes made just before injection with a gastight syringe. The final PEA headspace concentrations used in the trial were 0.252, 0.611, and 1.268 pg/ml. The plates were then incubated for 10 min at 20°C before SPME sampling, until headspace equilibration was achieved. Five replications were performed for each concentration in order to reduce the variability. The chromatographic data were collected, stored, and processed with Excel, and a calibration curve was defined by plotting GC-FID peak areas versus PEA concentrations. Effects of pure PEA on conidial germination and mycelium growth at the concentration naturally produced by the R82 strain. In vitro trials were performed to study the effects of PEA at the real concentration naturally produced by R82, calculated previously from the calibration curve (0.726 pg/ml of headspace), and at 2 times the natural concentration (1.452 pg/ml of headspace) on B. cinerea, C. acutatum, M. laxa, and P.
ANTIFUNGAL COMPOUNDS FROM P. EXPANSUM
1881
expansum. PEA solutions were obtained by diluting 95 or 190 pi of pure PEA in 10 ml of sterile distilled water and vortexing for 3 min to homogenize the solution. Two-compartment petri dishes were used; in one compartment, a 5-pl drop of PEA solution was put in the center, while the second compartment hosted the target pathogen, allowing its exposure to the volatile PEA. The mycelium growth assessment and the conidial germination of the target pathogens were carried out as previously reported. The controls were the target pathogens not exposed to PEA. All Petri dishes were wrapped with two layers of Parafilm. C. acutatum, M. laxa, and P. expansum growth was determined after 3 days and, for B. cinerea, after 24 h of exposure by measuring the diameters of the colonies from two orthogonal diameter measurements or by counting the CFU. Five replicate plates were used for each pathogen, and the experiment was repeated twice. Statistical analysis. All data were subjected to a one-way analysis of variance using the statistical package Statistica for Windows (StatSoft, Inc., Tulsa, OK). Separation of means was performed by using the least significant difference test at a P value of
Control of Postharvest Fungal Pathogens by Antifungal Compounds from Penicillium expansum WAFA ROUTSSI,1 LUISA UGOLINI,2 CAMILLA MARTINI,1 LUCA LAZZERI,2 a n d MARTA MARI'' 1Criof’, University o f Bologna, Via Gandolfi 19,40057 Cadriano, Bologna, Italy; and 2Consiglio per la Ricerca e la Sperimentazione in Agricoltura—Centro di Ricerca per le Colture Industriali (CRA-CIN), Via di Corticella 133, 40128 Bologna, Italy MS 13-072: Received 24 February 2013/Accepted 4 June 2013
ABSTRACT The fungicidal effects of secondary metabolites produced by a strain of Penicillium expansum (R82) in culture filtrate and in a double petri dish assay were tested against one isolate each of Botrytis cinerea, Colletotrichum acutatum, and Monilinia laxa and six isolates of P. expansum, revealing inhibitory activity against every pathogen tested. The characterization of volatile organic compounds released by the R82 strain was performed by solid-phase microextraction-gas chromatographic techniques, and several compounds were detected, one of them identified as phenethyl alcohol (PEA). Synthetic PEA, tested in vitro on fungal pathogens, showed strong inhibition at a concentration of 1,230 ug/ml of airspace, and mycelium appeared more sensitive than conidia; nevertheless, at the concentration naturally emitted by the fungus (0.726 + 0.16 ug/ml), commercial PEA did not show any antifungal activity. Therefore, a combined effect between different volatile organic compounds produced collectively by R82 can be hypothesized. This aspect suggests further investigation into the possibility of exploiting R82 as a nonchemical alternative in the control of some plant pathogenic fungi.
After harvest, fruit are cold stored in order to prevent decay, but economically important losses caused by fungal pathogens can occur during storage. Gray mold (Botrytis cinerea Pers.: Fr.) is a very common pathogen of pome, stone, and kiwifruit (6), brown rot (Monilinia spp.) is the most important disease of stone fruit (3), and blue mold rot (.Penicillium expansion Link) is responsible for the losses in stored pear and apple (12). Colletotrichum acutatum J.H. Simmonds causes anthracnose on strawberry (21) and is frequently associated with bitter rot in apple (13). It was recently detected for the first time on Italian apple (22). In the past, the postharvest application of synthetic fungicides has been an important strategy in reducing postharvest losses, although the possibility of controlling these patho gens with nonsynthetic chemical means, such as microbial antagonists, natural bioactive compounds, and physico chemical methods, is now being assiduously investigated and reviewed (24, 35). According to Hegde et al. (9), the emphasis is on the exploitation of unusual and previously ignored ecosystems (microorganisms and plant extracts), such as secondary metabolites produced by microorganisms, which could be one of the important alternative manage ment strategies. Secondary metabolites can be produced by almost all types of living organisms, including bacteria and fungi. In the domain of Bacteria, Bacillus and Pseudomonas species are the most frequent producers (2). Several studies have reported the activities of their secondary metabolites Author for correspondence. Tel: +39051766563; Fax: +39051765049; E-mail: [email protected].
against the growth of postharvest fungal pathogens in citrus (34), in peach (30), and in apple and pear (11). Among the microscopic fungi, the ability of imperfect fungi, ascomycetes, and several other filamentous and endophytic fungal species to produce bioactive products is the most significant. The total number of bioactive fungal compounds is approximately 8,600, representing 38% of all microbial products: 950, 900, and 350 compounds have been isolated from the most common ascomycetes, such as Aspergillus, Penicillium, and Fusarium species, respectively (2), and some of them are bioactive against some phytopathogens (29). The production of such a diversified array of active secondary metabolites by Penicillium species is well documented. Culture filtrates of isolates belonging to the species Penicillium canescens and Penicillium janczewskii showed inhibitory activities against Rhizoctonia solani (28), while citrine in the filtrate of P. citrinum inhibited Sclerotinia minor growth (26). A new antifungal compound isolated from P. expansum reduced the growth of Lasiodiplodia theobromae in fungicide disk assays (8). In a previous work, the P. expansum strain R82 showed inhibitory activity in vitro against some postharvest pathogens (31). The aims of the present study were (i) to evaluate the antifungal activities of secondary metabolites produced by the R82 strain in culture filtrate and in the atmosphere (volatile organic compounds [VOCs]) against B. cinerea, C. acutatum, M. laxa, and P. expansum, (ii) to investigate the compound responsible for the observed antifungal activity by the solid-phase microextraction (SPME)-gas chromatographic technique, (iii) to quantify
1880
ROUISSI ET AL.
the most-representative VOC identified, and (iv) to test its antifungal effect in vitro on the previously reported fungal pathogens. MATERIALS AND METHODS P. expansum stra in R82. P. expansum strain R82, derived from the C riof collection, was cultured on malt extract agar (MEA) for 2 days at 20°C in order to obtain a conidial suspension. The conidial suspension was prepared by washing the surface o f the culture with 5 ml of sterile distilled water amended with Tween 80 (0.05%, vol/vol) and rubbing with a glass rod. Spores were collected in 5 ml o f distilled water, counted using a hematocytometer, and adjusted to 103 conidia per ml. For fungitoxicity in vitro assays and to establish the phenethyl alcohol (PEA) production kinetics, conidial suspensions o f R82 strain were prepared at concentrations of 105 and 106 conidia per ml, respectively, chosen after preliminary trials. Pathogens. B. cinerea, C. acutatum, M. laxa, and six P. expansum strains used in the experiments were isolated individ ually from infected fruit tissue, identified by PCR analysis to sequence rDNA internal transcribed spacer regions, and kept at 4°C on MEA as single spore cultures until use. In order to obtain a conidial suspension, each pathogen was grown under the following conditions: B. cinerea was grown on oatmeal agar (60 g of oatmeal, 10 g o f sodium nitrate, 30 g of saccharose, and 12 g of agar per 1,000 ml of distilled water) and the cultures were incubated at 25°C under UV (350 to 420 nm) light for 12 h daily to favor sporulation; C. acutatum was cultured on potato dextrose agar at 25°C for 10 days; M. laxa was grown on V8 agar (250 ml of V8 vegetable juice and 40 g of agar per 1,000 ml of distilled water) and incubated at 25°C with 12-h-dark/l2-h-light cycles for 10 days; and P. expansum was cultured on MEA for 3 days at 20°C.
In vitro fungitoxicity assays. The activities o f the secondary metabolites produced by the R82 strain against target pathogens were studied using the liquid medium test to determine the percentage of mycelium growth inhibition and the double petri dish assay in order to evaluate the production of VOCs. The mycelium growth inhibition assay was carried out in flasks containing 20 ml of malt extract broth (MEA without agar) that were inoculated with the R82 conidial suspension (106 conidia per ml) and incubated at 20°C. After 10 days, the mycelium was separated from the medium by centrifugation at 4,800 x g for 20 min, and the supernatant was fdtered through a sterile filter (0.45 pm). Aliquots o f 20 ml from the culture filtrate o f R82 were added to sterile 50-ml flasks, and then the flasks were inoculated with 100 pi of the single conidial suspension (103 conidia per ml) o f B. cinerea, C. acutatum, M. laxa, or P. expansum and incubated at 20°C for 7 days. The contents o f the flasks were then filtered using a Whatman filter number 1 (preconditioned overnight at 80°C) and dried in an oven at 80°C until reaching a constant weight in order to determine the mycelium dry weight. Flasks containing 20 ml of malt extract broth were inoculated with target pathogens and used as the control. The sample unit was represented by three replications (flask) for each pathogen. The trials were repeated twice. In the double petri dish assay, the antifungal activities of VOCs produced by the R82 strain were tested against the mycelial growth and conidial germination of B. cinerea, C. acutatum, M. laxa, and P. expansum by a dual-culture technique. For this purpose, a 6-mm agar plug was removed from the edge of an actively growing colony of each pathogen cited above and
J. Food Prot., Vol. 76, No. 11
inoculated in a MEA dish. Subsequently, the lid o f the plate was replaced with a base plate o f MEA inoculated with R82, and the plate was incubated at 20°C for 2 days. The two base plates were sealed immediately, using a double layer of Parafilm to make a closed chamber of almost 180 cm3 in volume. For the inhibition of conidial germination, aliquots (100 pi per dish) o f a conidial suspension (103 conidia per ml) of each pathogen were spread on MEA plates. The lids o f the MEA plates were removed and replaced with base plates containing previously inoculated R82 as described above. The two base plates were sealed immediately, using a double layer of Parafilm as described above. For both assays, the control treatment was inoculating dishes with pathogens only and sealing them with Parafilm. The percentages of inhibition o f growth diameter and o f conidial germination were assessed after 7 and 3 days of incubation, respectively, at 20°C. The percentages o f inhibition o f growth diameter and of conidial germination were calculated according to the formula of Trivedi et al. (38), as follows: [(Tl — T2)/T1] x 100, where T1 is the diameter of growth or CFU o f the target pathogen not exposed to R82 (control) and T2 is the diameter of growth or CFU o f the target pathogen exposed to R82. The assay was conducted in five replicates (dishes) and repeated twice.
Identification of volatile compounds. Headspace volatiles from R82 were qualitatively evaluated and identified by SPME (36) coupled with gas chromatography-mass spectrometry (GCMS) (GCD System, model G1800A, Agilent-Hewlett-Packard, Santa Clara, CA). Fresh cultures of the R82 strain were prepared following the same protocol used for the double petri dish assay. The needle o f the SPME device, containing the extraction fiber coated with 85-gm polyacrylate film or 100-pm polydimethylsiloxane (Supelco, Bellefonte, PA), was inserted into each plate through a small hole, and the fiber was exposed to the gas phase for 20 min at 20°C. The SPME device was then removed from the petri dish and inserted into the gas chromatograph injector port. Thermal desorption o f compounds extracted from the fiber was performed at 250°C for 2 min, and the subsequent separation of compounds was achieved through an HP-5MS Varian capillary column (30-m length by 0.25-mm inner diameter and 0.25-pm film thickness) at a flow rate of 1 ml/min with helium as the carrier gas. The column temperature was set at 40°C for 3 min and then programmed to rise from 40 to 300°C at 20°C/min. The temperatures of the injection port and ion source were set at 250 and 280°C, respectively; splitless injection mode and electron impact ionization (70 eV) were established. The VOCs were identified by considering their mass spectra and their retention times in comparison with those o f reference substances in the National Institute of Standards and Technology (NIST) library, 1992. Blank samples, represented by plates prepared as described above but without the R82 inoculum, were analyzed following the same method in order to exclude interfering substances coming from the medium. PEA (International Union o f Pure and Applied Chemistry name, 2-phenylethanol) was one of the volatiles observed, and its identification was achieved by injecting pure PEA standard (Chemical Abstracts Service no. 60-12-8, Sigma Aldrich, St. Louis, MO) using the same chromatographic conditions and confirming the mass spectrum and retention time. Effects of pure PEA on mycelium growth and conidial germination of fungal pathogens. The antifungal activities of pure PEA, the most-representative VOC produced by the R82 strain, identified by SPME-GC-MS analysis as described above, against the mycelium growth and conidial germination of target pathogens were assayed. A plug (6-mm diameter) from an actively
J. Food Prot., Vol. 76, No. 11
growing pathogen culture or 100 pi of a conidial suspension were placed or spread, respectively, in the center of MEA plates. In each case, different aliquots of pure PEA corresponding to 77, 148, 308, 615, and 1,230 pg/ml of headspace were placed, using a microsyringe, on a paper filter (Whatman no. 1, 90-mm diameter) positioned inside the cover. The dishes were quickly closed, sealed with Parafilm, and incubated at 20°C. The dishes were opened and evaluated after 7 days for mycelium growth and after 3 days for conidial germination. Petri dishes inoculated with pathogens but treated with distilled water instead of PEA were used as the control. Five dishes were prepared for each dose and pathogen, and the experiment was repeated twice. Mycelium growth (mm) was gauged with a tape measure. Conidial germination was determined by counting the CFU developed on MEA. PEA production kinetics. A method was defined to quantitatively establish the production kinetics of PEA released by R82. One hundred microliters of a conidial suspension of R82 (105 conidia per ml) was spread on MEA plates and incubated at 20°C. The fungal PEA production was followed by headspace SPME coupled with GC-flame ionization detector (FID) analysis. Sampling was performed for 100 h at time intervals of 48, 52, 65, 72, 76, 89, 96, and 100 h after inoculation. The same assay was performed on CADRP28, a P. expansum strain chosen randomly from strains of the CRIOF-UNIBO collection. Strain CADRP28 is a producer of the same major volatiles emitted by R82 as confirmed through SPME-GC-MS (data not shown). The SPME sampling was performed in the same way as described above. An Agilent 7820A GC-FID and a Varian HP-5 capillary column (30-m length by 0.25-mm inside diameter, 0.25pm film thickness) were used for the chromatographic analysis. The instrument settings were as follows: the injector and detector temperatures were 250 and 300°C, respectively, the oven program started at 40°C for 3 min and was raised to 300°C at a rate of 20°C/ min, the flow rate of the earner gas (He) was 1 ml/min, and the splitless injection mode was established. Solutions of an individual standard synthetic PEA were prepared in the laboratory to correctly identify natural PEA by GC-FID. The production kinetics of the naturally released PEA was established by determining the mean value of five areas measured at each sampling time. Three replications were performed for each trial. PEA quantification. A calibration curve for quantification of the naturally produced PEA was established by SPME-GC-FID analysis using synthetic PEA as the standard. For this aim, 5 pi of the synthetic PEA water solutions at different concentrations were injected into closed MEA petri dishes through holes made just before injection with a gastight syringe. The final PEA headspace concentrations used in the trial were 0.252, 0.611, and 1.268 pg/ml. The plates were then incubated for 10 min at 20°C before SPME sampling, until headspace equilibration was achieved. Five replications were performed for each concentration in order to reduce the variability. The chromatographic data were collected, stored, and processed with Excel, and a calibration curve was defined by plotting GC-FID peak areas versus PEA concentrations. Effects of pure PEA on conidial germination and mycelium growth at the concentration naturally produced by the R82 strain. In vitro trials were performed to study the effects of PEA at the real concentration naturally produced by R82, calculated previously from the calibration curve (0.726 pg/ml of headspace), and at 2 times the natural concentration (1.452 pg/ml of headspace) on B. cinerea, C. acutatum, M. laxa, and P.
ANTIFUNGAL COMPOUNDS FROM P. EXPANSUM
1881
expansum. PEA solutions were obtained by diluting 95 or 190 pi of pure PEA in 10 ml of sterile distilled water and vortexing for 3 min to homogenize the solution. Two-compartment petri dishes were used; in one compartment, a 5-pl drop of PEA solution was put in the center, while the second compartment hosted the target pathogen, allowing its exposure to the volatile PEA. The mycelium growth assessment and the conidial germination of the target pathogens were carried out as previously reported. The controls were the target pathogens not exposed to PEA. All Petri dishes were wrapped with two layers of Parafilm. C. acutatum, M. laxa, and P. expansum growth was determined after 3 days and, for B. cinerea, after 24 h of exposure by measuring the diameters of the colonies from two orthogonal diameter measurements or by counting the CFU. Five replicate plates were used for each pathogen, and the experiment was repeated twice. Statistical analysis. All data were subjected to a one-way analysis of variance using the statistical package Statistica for Windows (StatSoft, Inc., Tulsa, OK). Separation of means was performed by using the least significant difference test at a P value of
