The biocontrol yeast R. mucilaginosa strain 3617 (preserved in the China General Microbiological Culture Collection Center, Accession number 3617) was isolated from the surface of peach blossoms in a chemically untreated orchard in China. The biocontrol yeast R. kratochvilovae LS11 was obtained from the culture collection of the Department of Agricultural, Environmental and Food Sciences, at the University of Molise, Italy. Both BCAs were routinely maintained at 4°C on nutrient yeast dextrose agar [NYDA: 8 g nutrient broth, 5 g yeast extract, 10 g glucose, and 20 g agar (Sangon Biotech, Shanghai, China)] in 1 L distilled water. Liquid cultures of the yeasts were prepared in 250-mL Erlenmeyer flasks containing 50 mL of nutrient yeast dextrose broth (NYDB) that had been inoculated with cells withdrawn with a loop from the agar medium mentioned above. Cellular suspensions of R. mucilaginosa 3617 or R. kratochvilovae LS11 were incubated for 20 h on a rotary shaker at 180 rpm, at 28 and 23°C, respectively. Following incubation, the cultures were centrifuged at 7000 ×g (TGL-16M Centrifuge, Xiangyi Co., Changsha, China) for 10 min and washed twice with sterile distilled water. Yeast cell pellets were suspended in sterile distilled water and their concentrations were adjusted with a hemocytometer to 1 × 10⁸ cells mL^-1, as required for the subsequent experiments.

Penicillium expansum strain PY (preserved in the China General Microbiological Culture Collection Center, accession number 3703) was isolated from infected apples in China, while P. expansum strain FS7 was obtained from the Culture Collection Center of the Department of Agricultural, Environmental and Food Sciences, at the University of Molise, Italy. The fungal cultures were maintained on potato dextrose agar (PDA: 200 g extract of boiled potatoes, 20 g glucose, 20 g agar, and 1 L distilled water) at 4°C. Fresh cultures were grown on PDA plates at 25°C before use. Spore concentrations were determined with a hemocytometer, using sterile distilled water for the adjustments.

Apples (Malus domestica Borkh, cv. Fuji) were harvested at commercial maturity from an orchard in Yantai, Shandong province, China, and selected based on their uniformity of size, ripeness, and absence of apparent injury or infection. Fruits were surface disinfected with 0.1% (w/v) sodium hypochlorite for 1 min, washed with tap water and allowed to dry at room temperature.

Apples were wounded with a sterile cork borer (approximately 3-mm-diameter and 3-mm-deep). In each wound, 30 μL of the following suspensions were alternatively applied: (1) R. mucilaginosa 3617 (1 × 10⁸ cells mL^-1), (2) R. kratochvilovae LS11 (1 × 10⁸ cells mL^-1), (3) sterile distilled water as a control. Two hours later, 30 μL of P. expansum (PY or FS7) suspensions (5 × 10⁴ spores mL^-1) were inoculated into each wound. After air drying, the samples were stored in enclosed plastic trays to maintain a high relative humidity (about 95%) and incubated at 20°C in an incubator (Radford Technology Co., Ltd., Ningbo, China). The disease incidence and symptoms, expressed as the percentage of infected fruit wounds and diameter of lesions in mm, respectively, were recorded at 3, 5, 7, and 9 days after inoculation. The experiment was conducted twice. Each experiment consisted of three replicates per time point, and each replicate consisted of nine apples, each with three wounds. Data from the two experiments were similar as resulting from ANOVA analysis and were pooled. Bars in the figure represent mean values ± standard deviations.

Patulin production by strains PY and FS7 of P. expansum in rotting apple tissues was assessed as previously described (Castoria et al., 2011; Yang et al., 2015) with slight modifications. Samples for the analysis of PAT were collected and extracted on days 5, 7, and 9 after inoculation. In order to consistently represent the expected progression of the fungal disease over time (i.e., in terms of increase of lesion size and of fungal biomass), the samples (infected fruit wounds) that were chosen for withdrawal on days 7 and 9 were the ones that at the respective previous time points (day 5 for day 7, and day 7 for day 9) had similar lesion diameters to those of samples that were withdrawn on days 5 and 7, respectively. After withdrawing the rotten tissue and 1 cm of surrounding healthy tissue using a sterile cork borer from infected wounds, the latter were pooled and homogenized by using a homogenizer (TTL-260, Beijing TongTaiLian Technology co., Ltd., Beijing, China) with rotor speed set at 15000 rpm. Afterward, each sample was weighed and was split into samples of 20g each, one used for PAT extraction and analysis, and the other 20 g was used for DNA quantification (see below). The sample for PAT determination was mixed with 25 mL of sterilized distilled water and transferred into a 50 mL conical flask. The samples were then treated overnight at room temperature with 100 μg mL^-1 of pectinase (500 units/mg, Sangon Biotech). Afterward, the same volume of ethyl acetate was added to each sample and shaken vigorously for 5 min. The upper layer was then transferred into a separatory funnel. This process was repeated twice and the organic phases of each sample were pooled. Ten milliliters of 14 g L^-1 sodium carbonate solution were then added to the organic phases and shaken for 10 s. The phases were allowed to settle, then they were separated and the aqueous phase was immediately extracted with 10 mL of ethyl acetate by shaking for 1 min. Subsequently, five drops of glacial acetic acid were added to the organic phases, which were then evaporated to dryness in a water bath at 40°C. The resulting residue was immediately dissolved in 1 mL of acetonitrile/water (1:9 v/v), the mobile phase used for HPLC analyses, and filtered through a 0.22 μm filter (WondaDisc NY organic filter, SHIMADZU-GL Sciences, Shanghai, China). Finally, 20 μL of each filtered solution was injected into a high-performance liquid chromatography (HPLC) apparatus to determine the PAT content. Agilent 1260 series system equipped with a quaternary pump and variable wavelength detector (Agilent Technologies, Santa Clara, CA, United States) was used with a Zorbax^® analytical column (SB-C₁₈ 250 mm × 4.6 mm, 5 μm, Agilent Technologies). The mobile phase was acetonitrile/water (1:9 v/v) with a flow rate of 1 mL min^-1. Detection was performed by measuring the absorbance of UV light at 276 nm.

Each experiment consisted of three replicates per time point, and each replicate consisted of three apples, each with three wounds. Rotting wounds collected from each replicate were pooled before extraction. The experiment was conducted twice. Data from the two experiments, which were expressed as μg patulin/g of decayed apple tissue, were similar as resulting from ANOVA analysis and were pooled. Bars in the figure represent mean values ± standard deviations.

The yeasts were cultivated on NYDA and a single colony was picked out and transferred to NYDB. After 48 h of growth at 28 and 23°C (for R. mucilaginosa 3617 and R. kratochvilovae LS11, respectively) with agitation at 180 rpm, the yeast cells were harvested for DNA extraction. Strains FS7 and PY of P. expansum were freshly grown on PDA for 7 days at 25°C. Afterward, the spores were withdrawn, their concentration was adjusted to 10⁵ spores mL^-1, and they were added to a 250-mL Erlenmeyer flask containing 100 mL of PDB that was maintained under constant agitation at 180 rpm for 4 days at 25°C. Genomic DNA of the yeast strains, P. expansum strains, and healthy apple fruits were extracted using the method described by Ihrmark et al. (2012). The yeast cells, the fungal mycelia and the apples were collected and freeze-dried. Two hundred milligrams of fungal mycelia or yeast cells were ground into powder with liquid nitrogen. The resulting powder was collected into micro-centrifuge tubes to which 700 μL of lysis buffer CTAB [2% hexadecyltrimethylammonium bromide (w/v), 1.4 M NaCl, 20 mM ethylenediaminetetraacetic acid pH 8, 100 mM Tris-HCl pH 8] was added. The lysis mixtures were incubated at 65°C for 1 h and then cooled on ice for 1 h. Afterward, the samples were treated with 500 μL of phenol: chloroform (1:1 v/v) and vortexed for 1 min and the supernatant was centrifuged at 13,000 rpm for 15 min at 4°C. Genomic DNA was precipitated with an equal volume of ice-cold isopropanol by incubation overnight at -20°C. After incubation, the samples were centrifuged at 13,000 rpm for 10 min at 4°C. The pellets were thoroughly rinsed with 70% ethanol, air-dried and re-suspended in 50 μL of sterile distilled water. The DNA purity ratios and concentrations were measured using a NanoDrop 2000 Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, United States). Extraction of genomic DNA from yeasts, fungi and healthy apples was performed to assess the specificity of the primers chosen for the quantification of P. expansum biomass through qPCR.

For the assessment of the development of the pathogenic and mycotoxigenic P. expansum strains PY and FS7 in apple fruits, we used a primer pair specific for P. expansum PatF (GenBank Accession No. AIG62137), a gene involved in PAT biosynthesis: patF_F (ATGAAATCCTCCCTGTGGGTTAGT) and patF_R (GAAGGATAATTTCCGGGGTAGTCATT), designed by Tannous et al. (2015). The primers were synthesized by Sangon Biotech and their specificity was tested by carrying out polymerase chain reaction (PCR) experiments with genomic DNA from the P. expansum strains, the BCAs, and healthy Fuji apples as templates. All PCR amplifications were performed in a total reaction volume of 25 μL consisting of 2 μL DNA template (∼50 μg), 2.5 μL 10 × PCR buffer (containing Mg²⁺), 2.5 μL dNTPs (2.5 μM each), 2.5 μL of each primer (0.5 μM each), 0.2 μL Taq DNA polymerase (LA taq TAKARA, Japan), and H₂O up to 25 μL. The PCR conditions involved five steps: 94°C for 5 min, 94°C for 30 s, 60°C for 30 s, 72°C for 10 s, 72°C for 10 min. A total of 40 cycles of amplification were performed from steps 2–4. The PCR results were analyzed by agarose gel (1% w/v) electrophoresis.

DNA extractions from apples infected by P. expansum strains, both treated and untreated with the BCAs were performed at days 5, 7, and 9 after the inoculation of the fruits. The extractions of genomic DNA for qPCR analysis were performed from 20 g portions of the same samples as those used for PAT extraction and analysis. Extractions were performed according to the protocol described by Rodriguez et al. (2011) with slight modifications. Fifty milliliters of Tris-HCl (pH 8) were added to 20 g of homogenized sample and then vortexed for homogenization. The sample was filtered through gauze paper and then centrifuged at 10,000 rpm for 5 min. The resulting pellet was suspended in 100 μL of pre-heated sterile distilled water and then cooled on ice for 10 min. Five hundred microliters of lysis buffer (2% hexadecyltrimethylammonium bromide, 1.4 M NaCl, 20 mM ethylenediaminetetraacetic acid EDTA pH 8, 100 mM Tris-HCl pH 8) and 20 μL of proteinase K (Sangon Biotech) were added. The sample was incubated at 65°C for 1 h and then centrifuged at 10,000 rpm for 5 min. The supernatant was transferred to a new centrifuge tube containing 500 μL phenol/chloroform (1:1 v/v). The solution was thoroughly mixed and then centrifuged at 13,000 rpm for 10 min at 4°C. The upper layer was carefully transferred to a new centrifuge tube and 10 μL of RNase solution (Sangon Biotech) were added before incubation at 37°C for 1 h. Afterward, an equal volume of chloroform was added, the mix was vortexed and centrifuged at 13,000 rpm for 5 min. The top layer was transferred to a new microcentrifuge tube, DNA was precipitated with an equal volume of ice-cold isopropanol and incubated at 20°C. The obtained pellet was rinsed with 70% ethanol, air-dried and suspended in 30 μL sterile distilled water. The DNA purity ratio and concentration were measured as described above. Since the experiments for the quantification of fungal DNA were performed with aliquots of the same samples as those collected for PAT quantification, they consisted of the same number of replicates, apples and wounds. Fungal biomass was expressed as ng DNA/μg decayed apple tissue. All the genomic DNA sample used in this work were also tested for their suitability for PCR amplification using primers actin-F and actin-R (actin-F: TCCTTCGTCTTGACCTTGCT, actin-R: ACTTCATGATGGAGTTGTAGGTAGT) (Sangon Biotech) on the basis of the actin gene sequence of Malus pumila (Hightower and Meagher, 1986).

qPCR assays with DNA samples extracted from infected apples were performed to determine fungal biomass using SYBR Premix Ex Taq TM II (Tli RNaseH Plus) (TAKARA, Japan) with the Bio-Rad CFX96 (Bio-Rad, Hercules, CA, United States) and the primer pair patF_F and/patF_R. Two negative controls were also performed (one without primers and fungal DNA and the other one without fungal DNA) to rule out any possible matrix effect. The PCR conditions were as follows: 94°C for 30 s, 40 cycles of 94°C for 20 s, 60°C for 20 s, and 72°C for 20 s. The dissociation curve analysis followed the same trend as the amplification cycle and was constructed by continuously measuring the fluorescence when increasing the temperature from 65 to 95°C, at the rate of 0.5°C/s. The threshold cycle (CT) values were automatically determined by the CFX Manager^TM software (Bio-Rad Laboratories). For each DNA sample, three technical replicates were analyzed. Fungal biomass was expressed as ng DNA/μg decayed apple tissue.

The specific rates of PAT biosynthesis (specific mycotoxigenic activity) by P. expansum strains PY and FS7 were calculated using the mean concentration of PAT and P. expansum DNA in decayed apple tissue as measured in the two experiments and expressed as ng patulin/μg fungal DNA.

The percentages of apple wounds infected with P. expansum were converted to Bliss angular values (arcsine square root) before analysis of variance. Data were subjected to analysis of variance (ANOVA, SPSS release 17.0 for Windows; SPSS Inc., Chicago, IL, United States). All the means were compared by using Duncan’s multiple range test (P < 0.05). All the experiments were performed twice.

The two biocontrol yeasts significantly reduced the percentage of infected wounds in apples inoculated with P. expansum strain PY (Figure 1A). At days 3, 5, 7, and 9 after inoculation, yeast strain LS11 reduced the disease incidence caused by P. expansum strain PY by 59.7, 50.5, 41.1, and 15.7%, respectively. In apples treated with yeast strain 3617 the disease incidence was reduced by 55.8, 41.9, 30.9, and 11.1% at the same time intervals (Figure 1A). When the BCAs were challenged by P. expansum strain FS7, no decrease of disease incidence was recorded after day 5 in the apples pretreated with R. mucilaginosa 3617 or after day 7 in apples pretreated with R. kratochvilovae LS11. Strain 3617 reduced the disease incidence by 53.9 and 38.4% at days 3 and 5, respectively, while strain LS11 reduced the disease incidence 75.2, 45.2, and 14.9% after 3, 5, and 7 days, respectively (Figure 1B).

Furthermore, both BCAs caused significant reductions in the mean diameters of the lesions caused by both strains of P. expansum at all tested time intervals, although this effect was less pronounced at days 5 and 7, especially in the apples pretreated with R. mucilaginosa 3617 (Figures 1C,D). In apples infected by P. expansum strain PY, yeast strain 3617 reduced the lesion diameters by 47.7, 35.4, 22.0, and 11.4% on days 3, 5, 7, and 9, respectively. On the same days, yeast LS11 reduced the lesion diameters by 32.3, 38.5, 38.1, and 20.5%, respectively, as compared to untreated control apples (Figure 1C). When the BCAs were challenged by P. expansum strain FS7, yeast strain 3617 decreased the lesion diameters by 37.4, 21.0, 14.3, and 9.0% on days 3, 5, 7, and 9, respectively. Yeast strain LS11 reduced lesion diameters by 48.8, 30.9, 20.3, and 13.4% at the same time intervals, respectively, as compared to the untreated control apples (Figure 1D).

The concentration of PAT that accumulated in rotted apple tissue infected with P. expansum PY and FS7 are shown in Figures 2A,B, respectively.

The PAT accumulation due to P. expansum FS7 was more than twice as high as that due to P. expansum PY at almost all of the tested time intervals. In apples infected with P. expansum FS7, a progressive increase of PAT contamination was observed throughout the experiment, ranging from 3.8 μg patulin/g infected apple tissue on day 5 to 8.7 μg/g on day 9 (Figure 2B). Conversely, the lower PAT accumulation that was recorded in apples infected with P. expansum PY was similar on days 5 (2.4 μg/g) and 7 (2.6 μg/g), and reached its highest value (3.2 μg/g) on day 9 (Figure 2A). Significant reductions of PAT contamination were evident at all tested time points in the infected apples pre-treated with both yeast strains (Figures 2A,B). The BCA LS11 decreased PAT accumulation due to P. expansum PY by 16.7, 19.2, and 9.4% on days 5, 7, and 9, respectively. Those decreases in PAT accumulation were significantly higher than those recorded in apples pretreated with strain 3617 on days 5 and 7, whereas on day 9 there was no significant difference between the apples pre-treated with the two BCAs (Figure 2A). The percentage decreases of PAT accumulation yielded by the two BCAs were more pronounced in apples inoculated with P. expansum FS7, the higher PAT producer, at all tested time points. In addition, the BCA strain LS11 yielded significantly greater decreases in PAT accumulation than strain 3617 at almost all tested time points. Strain LS11 yielded decreases in PAT accumulation of 34.2, 35.2, and 39.1% at days 5, 7, and 9, respectively, versus decreases of 52.6, 22.5, and 19.5% obtained by strain 3617 at the same time intervals, as compared to untreated control apples (Figure 2B).

The data shown in Figure 3A confirmed that the primers patF-F and patF-R are specific to the genomic DNA of P. expansum. The primers annealed to the genomic DNA of strains PY and FS7, yielding the expected PCR product of 250 bp in each case. Conversely, no PCR products were obtained when genomic DNA samples from the two biocontrol yeast strains (lanes 5–6) and the Fuji apple fruits used in this study (lane 6) were used as templates (Figure 3A). Figure 3B shows the PCR positive control, in which the primers actin-F and actin-R were used with genomic DNA samples from the two P. expansum strains, the two biocontrol yeasts, and Fuji apple fruits. The DNA bands were of the expected size (330 bp), thus confirming that the quality of the genomic DNA samples used in this study was sufficient for PCR.

After confirming the specificity of the primers patF-F and patF-R (see Figures 3A,B), the analytical sensitivity of the qPCR assay was determined using serial 10-fold dilutions, ranging from 1 μg to 1 pg of pure genomic DNA from P. expansum strains PY and FS7 of. For each sample dilution, the fluorescence emission was measured and plotted against the log of Ct values (Figures 4A,B). The detection limit of the assay for both P. expansum strains was 1 pg of DNA. The dissociation curve analyses performed after the last amplification step detected a single PCR product with a specific melting temperature of 85°C (Figures 4C,D), since no primer dimers were generated during the qPCR amplification. The R² values of the linear correlations were 0.9979 and 0.9992 for strains PY and FS7, respectively (Figures 4E,F), and the slopes of the linear regression curves were -3.1538 and -3.337, respectively.

In infected apples that were not pre-treated with the BCAs, the biomass of both P. expansum strains, as measured in terms of DNA content appeared to steadily increase from day 5 to day 9 following the inoculation of fruits. The calculated biomass of P. expansum ranged from 0.47 to 1.28 μg/mg of infected apple tissue for PY and from 0.54 to 1.08 μg/mg for FS7 (Figures 5A,B).

In infected apples pre-treated with the BCAs LS11 and 3617, the biomass of both P. expansum strains increased at the same time intervals as in the non-pretreated apples, but at significantly lower rates (Figures 5A,B). Furthermore, the yeast strains LS11 and 3617 appeared to cause similar decreases in the DNA contents of P. expansum strains PY and FS7, although LS11 yielded greater reductions at almost all tested time points. The reductions of P. expansum strain PY DNA content in infected of apple tissue achieved by LS11 were 46.8, 46.7, and 36.2% at 5, 7, and 9 days after inoculation, respectively. In comparison, yeast strain 3617 yielded reductions in DNA content of 23.4, 35.6, and 25.8% at the same time intervals, respectively (Figure 5A). In fruits challenged with P. expansum FS7, the DNA content of this strain was reduced by 46.3, 45.0, and 47.3%, by LS11 on days 5, 7, and 9, respectively, whereas3617 yielded reductions in DNA content of 38.9, 23.8, and 37.0% on the same days (Figure 5B).

Figures 6A,B show the results for the pretreatment of apples with the BCAs on the specific rates of PAT biosynthesis (specific mycotoxigenic activity) by P. expansum strains PY and FS7 in the infected fruits. The rationale for these measurements was to assess whether the BCA enhance PAT biosynthesis owing to the competition and stress they exert on P. expansum in the infected wounds. The specific mycotoxigenic activity by P. expansum PY decreased from day 5 (5.1 ng PAT/μg DNA) to days 7 and 9 (2.9 and 2.5 ng PAT/μg DNA, respectively) (Figure 6A). In comparison, an approximately constant rate of PAT biosynthesis was observed for strain FS7 (7.0, 8.0, and 8.1 ng PAT/μg DNA at days 5, 7, and 9, respectively) (Figure 6B). Both strains 3617 and LS11 significantly increased PAT biosynthesis by P. expansum at all tested time points (Figures 6A,B), except for P. expansum strain FS7 in the presence of 3617 at days 5 and 7, when the biosynthetic rates were lower than and similar to the untreated controls, respectively. Increases in the biosynthetic rate of 58.8, 51.7, and 44.0% were observed at days 5, 7, and 9, respectively, for P. expansum PY in the presence of the yeast strain LS11, and 11.8, 34.5, and 24% at the same time points in the presence of the yeast strain 3617. In apples infected with P. expansum FS7, increases in the biosynthetic rate of 22.9, 15.6, and 13.6% were detected at days 5, 7, and 9, respectively, in the presence of the yeast strain LS11, and of 27.2% at day 9 for P. expansum FS7 in the presence of strain 3617.