Sclerotinia sclerotiorum strain NJC09 was isolated from soybean plants showing Sclerotinia stem rot symptoms in Xuzhou, China (117.30° E, 34.28° N). Isolation and identification of the pathogen are described in Supplementary Figure S1. The isolation of Valsa pyri strain Vp297, Colletotrichum brevisporum strain GM1, and Botryosphaeria dothidea isolate SBL303 was previously reported by our research group (Chen et al., 2020; Shi et al., 2020; Song et al., 2020). Fungal strains were cultured on potato–dextrose–agar (PDA) medium (200 g potato, 20 g dextrose, and 15 g agar in 1 L water) at 28°C.

Fungal plant pathogens B. dothidea, C. brevisporum, S. sclerotiorum, and V. pyri were used in the antifungal assay. KA was purchased from Macklin (China). The antifungal activity of KA was evaluated in PDA medium containing 5, 10, and 15 mM KA (0.71, 1.42, and 2.13 mg/ml KA), respectively. The control experiments were performed in the absence of KA. The cultures were incubated under darkness at 28°C for 3 days. The colony diameter was measured using a ruler (Tremarin et al., 2015), and the antifungal effect was determined according to the mycelial growth. Experiments were repeated five times.

To calculate the half maximal effective concentrations (EC₅₀), KA, carbendazim, and prochloraz were used. S. sclerotiorum was grown on PDA plates containing 0.5, 1, 2, 3, 5, 10, and 15 mM of each fungicide (KA: 0.071, 0.14, 0.28, 0.43, 0.71, 1.42, and 2.13 mg/ml; carbendazim: 0.096, 0.19, 0.38, 0.57, 0.96, 1.91, and 2.87 mg/ml; prochloraz: 0.19, 0.38, 0.75, 1.13, 1.88, 3.77, and 5.65 mg/ml). The negative control experiment was performed by culturing S. sclerotiorum on PDA in the absence of fungicides. The dishes were incubated at 28°C for 3 days. Experiments were repeated five times. Prism 7.0 (GraphPad Software, United States) was used to calculate the EC₅₀ values.

Sclerotinia sclerotiorum was cultured on PDA medium under darkness at 28°C for 5 days. Three plugs (7 mm diameter) of mycelium were inoculated into 40 ml honey–barley–tryptone medium (600 mg honey, 100 mg barley, and 200 mg tryptone in 40 ml water) and placed at 28°C and 200 rpm for 12 h. After centrifugation at 8,000 rpm and 4°C, the produced mycelial fragments were washed twice with 20 ml sterilized distilled deionized water (ddH₂O). Harvested cells were suspended in 200 μl yeast extract–peptone–dextrose medium (YEPD; 0.15 g yeast extract, 0.5 g peptone, and 1 g glucose in 50 ml water; 1 × 10⁶ mycelial fragments/ml) containing 0 and 10 mM KA (0 and 1.42 mg/ml KA). The number of mycelial fragments was calculated using a Leica DM2500 microscope (Germany) at ×40 magnifications and adjusted using sterilized ddH₂O.

The effect of pH was calculated by adjusting the YEPD medium to pH values 3, 4, 5, 6, 7, and 8. The effect of iron(II), nickel(II), zinc(II), cobalt(II), and copper(II) was examined by adding the corresponding metal chlorides (1 mM final concentration) into the YEPD medium (pH 4). After 12 h of culture at 28°C and 200 rpm, fungal growth was detected using a Leica DM2500 microscope at ×20 magnifications. The antifungal activity was calculated according to the number of sclerotial aggregates per microliter. Experiments were repeated three times.

Sclerotinia sclerotiorum mycelial fragments were cultured in YEPD medium with 0 and 10 mM KA (0 and 1.42 mg/ml KA) as described in “Effect of pH and Metals on the Inhibitory Activity of KA” section. After 12 h at 28°C and 200 rpm, scanning electron microscope (SEM Gemini 300 Instrument, United States) and fluorescent images (Olympus BX51, Japan) were acquired. For the preparation of SEM samples, fungal cells were centrifuged and immobilized in 4% paraformaldehyde. The samples were dehydrated with 65, 75, 85, 95, and 100% (twice) ethanol at room temperature (each dehydration condition was carried out for 10 min). Then, the samples were stored in a desiccator for 6 h. Dried samples were placed on the sample holder, and coated with gold. Images were obtained with an acceleration voltage of 20 kV. In the analysis using the fluorescent microscope, 10 μl YEPD medium was stained with 10 μl 4,6-diamino-2-phenylindole (DAPI, 10 μg/ml solution, Solarbio, China). The resulting suspension was kept in darkness during 5 min before the analysis. Images were collected using a DAPI-selective filter (excitation wavelength: 377 nm; emission wavelength: 447) at ×40 magnifications.

Two 7-mm-diameter PDA plugs containing S. sclerotiorum mycelium were inoculated into 50 ml YEPD medium with 0, 2, 5, and 10 mM KA (0, 0.28, 0.71, and 1.42 mg/ml KA). After 15 days at 28°C and 200 rpm, DHN was extracted following the standard procedure (Kogej et al., 2004) and DHN content was determined by analytical HPLC-MS using a QTRAP 5500 system (AB SCIEX, United States), equipped with an Agilent C18 column (250 × 4.6 mm, United States). The system was maintained at 30°C. The mobile phase was 0.1% formic acid aqueous solution/acetonitrile 30:70. The injection volume was 5 μl and the flow rate was 0.2 ml/min. The peak corresponding to DHN was detected at 1.26 min. The concentration of DHN was calculated according to the peak area in MS using the transition from 161.0 to 143.3 [(M + H)⁺ calcd. For C₁₀H₉O₂, 161.0603; found, 161.0]. Experiments were repeated three times.

Freshly harvested cells prepared as described above were resuspended in YEPD medium (8 ml, 1 × 10⁶ mycelial fragments/ml) with 0 and 10 mM KA (0 and 1.42 mg/ml KA), and shake at 28°C and 200 rpm for 24 h. After centrifugation at 8,000 rpm and 4°C, the formed aggregates were collected. Total RNA was extracted using TRIzol reagent (Ambion, United States). cDNA was obtained by reverse transcription using the Transcript All-in-One First-Strand DNA Synthesis SuperMix for qPCR (OneStep gDNA Removal) Kit (Tsingke, China). qRT-PCR was performed using a set of two PCR primers with SYBR Green I Real-Time PCR (Solarbio, China) using a 7500 Real-time PCR system (Applied Biosystems, United States). Genes related to cell wall formation (CHS1, CHS2, CHS3, and GSH) and melanin synthesis (PKS12 and PKS13) were examined. Actin transcript was used as the internal reference gene (Melo et al., 2019). Primers are shown in Supplementary Table S1. The relative gene expression was calculated by the 2^−ΔΔCt method. Experiments were repeated five times.

After suspension of S. sclerotiorum mycelial fragments in 50 ml YEPD medium (1 × 10⁶ mycelial fragments/mL) containing 0 and 10 mM KA (0 and 1.42 mg/ml KA), the cells were shaken at 28°C and 200 rpm for 0, 1, 2, and 3. Chitin content was determined using the procedure described by Song et al. (2020). Briefly, after centrifugation (8,000 rpm, 4°C, 10 min) of the cells contained in the 50 ml YEPD medium, the fungal cells were hydrolyzed in 1 ml 6 M HCl at 100°C for 17 h. After evaporation of the solvent using a freeze-drier (FreeZone 2.5 Plus, LABCONCO, United States), the samples were dissolved in 1 ml water. Then, 0.75 M Na₂CO₃ solution (100 μl) was added to 100 μl sample. The mixture was incubated at 100°C for 20 min. Then, 95% ethanol (0.7 ml) and solution A (100 μl; solution A: 1.6 g p-dimethylaminobenzaldehyde in 30 ml concentrated HCl and 30 ml ethanol) were added. The absorbance was measured at 420 nm and compared with the standard curve from 0.1 to 20 mg/ml glucosamine. Experiments were repeated three times.

Two 7-mm-diameter PDA plugs containing S. sclerotiorum mycelium were inoculated into 50 ml YEPD medium with 0, 2, 5, and 10 mM KA (0, 0.28, 0.71, and 1.42 mg/ml KA). Samples were collected after 0, 1, 2, and 3 days. The content of oxalic acid was measured using the Oxalic Acid Content Detection (Visible Spectrophotometry) Kit (Solarbio, China). Briefly, 50 μl of culture medium were mixed with reagents 1, 2, and 3 following the manufacturer’s instructions, and the resulting solution was kept at room temperature for 20 min. The concentration of oxalic acid was calculated according to the absorbance at 510 nm. Experiments were repeated three times.

Sclerotinia sclerotiorum was cultured on PDA medium at 28°C for 5 days, and the mycelium was divided into 7-mm-diameter plugs. For the preventive assay, 2-mm-diameter wounds were made on the surface of soybean pods directly detached from soybean plants. After spraying 50 ml aqueous solutions (pH 5) containing 0, 10, 20, and 50 mM KA (0, 1.42, 2.84, and 7.11 mg/ml KA), the soybean pods were dried, and a mycelial plug was kept in contact with the wound. Carbendazim and prochloraz at 20 and 50 mM (carbendazim: 3.82 and 9.56 mg/ml; prochloraz: 7.53, and 18.84 mg/ml) were applied, instead of KA, as positive controls (4 kg/ha carbendazim and 0.45 kg/ha prochloraz are the recommended rates in fields trials). After 3 days at 28°C and 70% humidity, the preventive efficacy was measured according to the lesion length (necrotic area) caused by S. sclerotiorum.

For the curative assay, 2-mm-diameter wounds were made on the soybean pods. After inoculation using a mycelial plug, the inoculated soybean pods were incubated at 28°C and 70% humidity for 1 day and the mycelial plugs were removed. Then, 50 ml aqueous solutions containing KA, prochloraz, or carbendazim at 20 and 50 mM concentration were sprayed. Sterilized ddH₂O (50 ml) was sprayed in the control experiment. Treatments with combinations of fungicides were carried out using a 50 ml solution containing the fungicides at 10 mM concentration (KA: 1.42 mg/ml; carbendazim: 1.91 mg/ml; and prochloraz: 3.77 mg/ml).

Twenty soybean pods were used for each treatment condition. Experiments were repeated three times. The brownish-black color of the lesions caused by S. sclerotiorum, covered in some cases by white hyphae, was easily observed with the naked eye. All the lesion length measurements were pooled to calculate the mean value and deviation.

The statistical analyses were performed using SPSS (Statistical Package, Version 20.0). The variables were subjected to student’s t-test and were tested for significance at p < 0.05 (*), p < 0.01 (**), p < 0.001 (***), and p < 0.0001 (****) levels (ns = no significance). The SD, which was calculated using Microsoft Excel 2010, was used to quantify the dispersion.

Although KA at 15 mM slightly reduced the mycelial growth of B. dothidea, C. brevisporum, and V. pyri, this concentration of KA inhibited completely the growth of S. sclerotiorum (Figure 1A), whereas 5 and 10 mM KA reduced S. sclerotiorum growth by 29 and 56%, respectively (Supplementary Figure S2). These results indicated that S. sclerotiorum is more sensitive to KA in comparison to the other fungal species used in the assay. After 15 days of incubation, no mycelial growth was detected when treating S. sclerotiorum with 15 mM KA. When comparing the antifungal activity of KA with that of commercial fungicides carbendazim and prochloraz, it was found that the commercial fungicides showed higher antifungal activity in vitro than KA. The EC₅₀ value for KA was 6.3 ± 0.7 mM (0.89 ± 0.09 mg/ml), while the EC₅₀ values for carbendazim and prochloraz were 2.8 ± 0.3 (0.51 ± 0.06 mg/ml) and 0.7 ± 0.2 mM (0.26 ± 0.07 mg/ml), respectively.

Fluorescent microscope observations revealed that, in the absence of KA, S. sclerotiorum mycelial fragments formed sclerotial aggregates (number of observed sclerotia: 64), and only a few non-attached hyphae were detected (number of observed hyphae: 26; Figure 1B; Supplementary Figure S3). In some cases, hyphal structures were observed growing from the sclerotial aggregates (number of observed sclerotia with hyphae: 53; number of sclerotia without hyphae: 11). In contrast, after treatment with 10 mM KA, most S. sclerotiorum mycelial fragments remained in the original state (number of observed mycelial fragments: 112), and the number of hyphae increased in comparison with that observed in the absence of KA (number of observed hyphae: 53). After treatment with KA, no sclerotial aggregates were observed, indicating that KA was able to inhibit the formation of sclerotia and, subsequently, to temporarily halt S. sclerotiorum cell cycle.

The antifungal activity of KA was screened at different pH values, from 3 to 8. Although S. sclerotiorum could grow at all pH values, the formation of sclerotial aggregates was higher in an acidic environment than under alkaline conditions. The number of aggregates per microliter was 270 at pH 3, whereas the number of aggregates decreased to 15 at pH 8. In all tested pH values, the inhibitory activity of KA against S. sclerotiorum varied from 40 to 45%, revealing that the antifungal activity of KA was not obviously affected by pH (Figure 1C).

The presence of metals inhibited the formation of sclerotial aggregates (Figure 1D). In this sense, nickel(II) and copper(II) reduced the number of aggregates by 53 and 32%, respectively, in comparison to the control experiment. When KA treatment was combined with metal ions, the antifungal effect of KA was significantly reduced, with cobalt(II) and zinc(II) producing the greatest effect. KA, in the presence of cobalt(II) and zinc(II), reduced the number of aggregates per μl by 28 and 33%, respectively; whereas KA, without metal addition, reduced the number of aggregates by 45%.

Cultivation of S. sclerotiorum on PDA plates allowed the formation of sclerotia after 15 days (96 ± 19 sclerotia per plate), which could be easily observed due to the dark color produced by melanin (Figure 2A; Sousa Melo et al., 2019). Instead, the presence of 10 mM KA inhibited the formation of sclerotia (6 ± 3 sclerotia per plate), and cultures were completely white showing only non-melanized structures.

In order to confirm that KA was inhibiting the biosynthesis of melanin in S. sclerotiorum, the content of DHN, a kind of melanin and a key intermediate involved in biosynthesis of other melanin compounds in S. sclerotiorum (Supplementary Figure S4; Butler et al., 2009), was examined after application of different concentrations of KA. In agreement with the lack of dark sclerotia in the plates, 2, 5, and 10 mM KA reduced DHN content by 12% (no significant difference), 32% (no significant difference), and 62% (significant difference), respectively (Figure 2B); confirming that KA is inhibiting the biosynthesis of melanin. Consistently, the expression of polyketide synthase PKS12, which is involved in melanin biosynthesis, was found 4.8-fold downregulated in the presence of 10 mM KA, whereas no significant difference was observed in the mRNA level of polyketide synthase PKS13 after KA treatment (Figure 2C).

The antifungal activity assay indicated that KA can not only inhibit the formation of sclerotia, but can also inhibit S. sclerotiorum mycelial growth. SEM observations revealed that KA was altering cell wall integrity (Figure 3A). KA-treated cells (10 mM KA) showed numerous holes and irregularities in the cell wall that were not observed in the non-treated cells.

In order to understand the metabolic alterations caused by KA in S. sclerotiorum cell wall, the mRNA levels of three chitin synthase genes (CHS1, CHS2, and CHS3) and one β-1,3-glucan synthase gene (GSH) were measured after KA treatment (10 mM KA; Figure 3B). qRT-PCR analysis revealed that the gene expression of these genes was significantly reduced under KA treatment. The gene expression of GSH was 9.3-fold lower in KA-treated cells compared to that detected without KA treatment, while the gene expression of CHS1, CHS2, and CHS3 decreased by 73, 85, and 87%, respectively. These results suggested that KA is not only inhibiting melanin biosynthesis but also inhibiting the biosynthesis of chitin and β-1,3-glucans.

To confirm that KA was able to inhibit chitin biosynthesis, the concentration of chitin in the KA-treated cells was measured after 0, 1, 2, and 3 days (Figure 3C). Although chitin content increased steadily in non-treated cells, 10 mM KA significantly decreased chitin content, which is consistent with the mRNA levels of the chitin synthase genes after treatment with KA. The concentration of chitin decreased by 75, 45, and 63% in the KA-treated cells after 1, 2, and 3 days, respectively.

The concentration of virulence factor oxalic acid when culturing S. sclerotiorum in the absence of KA was 3.13 ± 0.76, 3.38 ± 0.67, and 4.05 ± 0.73 μM after 1, 2, and 3 days, respectively (Figure 4). The concentration of oxalic acid was reduced by 12% (no significant difference) and 91% after application of 2 and 5 mM KA, respectively. When treating S. sclerotiorum cells with 10 mM KA, no oxalic acid was observed. These results demonstrated that KA is able to inhibit oxalic acid production in S. sclerotiorum.

Kojic acid showed both curative and preventive abilities, reducing the symptoms of S. sclerotiorum in soybean (Figures 5A,B). In preventive applications (Table 1), the lesion length inhibition was proportional to the concentration of KA, observing complete inhibition with 50 mM KA. This, carbendazim and prochloraz at 20 mM provided 64, 65, and 53% lesion length inhibition, respectively, indicating that KA shows similar antifungal activity in preventive applications in comparison with commonly used fungicides.

In curative applications, the lesion length inhibition of 20 mM KA was slightly higher than those of carbendazim and prochloraz at the same concentration (KA: 22%; carbendazim: 20%; and prochloraz: 17% lesion length inhibition), and the same trend was also observed when using 50 mM concentrations (KA: 66%; carbendazim: 61%; prochloraz: 64% lesion length inhibition). Comparison between the curative and preventive abilities revealed that the three antifungal agents used in this study showed higher preventive ability in comparison with curative ability on S. sclerotiorum. After combining 10 mM KA with 10 mM carbendazim or 10 mM KA with 10 mM prochloraz, the inhibitory effect was higher than that detected with the treatment with single fungicides at 20 mM concentration (KA + carbendazim: 62%; KA + prochloraz: 54% lesion length inhibition). When 10 mM KA was applied together with 10 mM carbendazim and 10 mM prochloraz, the inhibitory effect was enhanced up to 80% lesion length inhibition.