During the period from 2021 to 2022, we collected fruiting bodies (basidiomata) of G. sichuanense displaying symptoms of green mold disease from three farms situated in Zhejiang, Hubei, and Jilin Province, China. The incidence of the disease ranged from 3 to 15%, significantly impacting the growth and development of G. sichuanense.

To conduct a comprehensive investigation of the disease, we isolated the fungus from the infected fruiting bodies using the tissue-isolation method. This involved carefully excising small pieces (0.3 cm) from the edges of the lesions on the diseased fruiting bodies using a sterile scalpel. The excised tissues were then subjected to surface sterilization by treating them with 75% ethanol (vol/vol) for 30 s, followed by 1% NaOCl (wt/vol) for 10 s. Subsequently, the tissues underwent three rinses with sterilized distilled water.

The tissues were placed onto dried and sterilized potato dextrose agar (PDA) plates and incubated in darkness at 25°C for three to 5 days. Regular inspections were carried out to monitor any fungal growth. Colonies that developed from the infected tissues were transferred to new PDA plates using the hyphal tip culture method to obtain pure cultures. All purified isolates were further subcultured on PDA medium for 3 days and preserved on PDA slants at 4°C.

To evaluate the characteristics of the isolates, mycelia plug with a diameter of 5.0 mm were obtained from the edges of actively growing cultures aged 5 days. These plugs were then placed at the center of agar plates containing potato dextrose agar (PDA), cornmeal dextrose agar (CMD), and synthetic low-nutrient agar (SNA). The plates were incubated at 25°C with a 12-h light/dark photoperiod for a duration of 5–7 days.

During the incubation period, careful observations and recordings were made on various colony characteristics, including color, shape, radial growth, and texture. The colony diameters were measured in two perpendicular directions. The daily growth rate was determined by calculating the average mean daily growth (mm/day).

For further analysis, one-week-old colonies cultivated on SNA plates were utilized to examine the conidia and conidiophores following the methods outlined by Chaverri et al. (2015). The shape and color of the conidia were observed, and the sizes of 20 randomly selected conidia from each isolate were measured under a Zeiss Axio lab. A1 microscope equipped with a differential interference contrast (DIC) optics camera (Carl Zeiss Microscopy GmbH, Germany), utilizing 1,000× magnification.

To obtain DNA for analysis, mycelia were collected from colonies cultivated on potato dextrose agar (PDA) for 3–5 days. DNA extraction was performed using the NuClean Plant Genomic DNA Kit (Cowin Biotech Co., Ltd., Taizhou, China).

For amplification of the target genes, specific primer pairs were used. The primer pair fRPB2-5f and fRPB2-7cr (Liu et al., 1999) amplified a 1 kb fragment of the RNA polymerase II second largest subunit (RPB2) gene. Additionally, the primer pair EF1-728F and TEF1LLErev (Chaverri et al., 2003; Jaklitsch et al., 2005) amplified a 1.3 kb fragment of the translation elongation factor 1-alpha (TEF1-a) gene. PCR amplification was conducted in a 30 μL reaction system comprising 15 μL of 10× PCR mix, 1.5 μL of each primer, 1.5 μL of template DNA, and 10.5 μL of ddH₂O. For both RPB2 and TEF1-a genes, PCR conditions included an initial denaturation step at 95°C for 5 min, followed by 30 cycles of denaturation at 95°C for 1 min, annealing at 59°C for RPB2 or 55°C for TEF1-a for 90 s, extension at 72°C for 90 s, and a final extension at 72°C for 10 min. The PCR products were purified using the PCR Product Purification Kit, and gel electrophoresis was performed to confirm successful amplification.

Sequencing of the PCR products was carried out bidirectionally using the fRPB2-5f/fRPB2-7cr and TEF1/TEF2 primers (Jaklitsch, 2009) at Comate Biosciences Co. Ltd (Changchun, Jilin, China). The obtained sequences were assembled using CAP3 software (Huang and Madan, 1999) to generate consensus sequences. BioEdit software (version 7.0.0) was used to remove 20 to 30 bp from the terminal ends. Basic Local Alignment Search Tool (BLAST) analysis ¹ was conducted for each gene locus to confirm the identity of the isolates. The consensus sequences were deposited in GenBank (Table 1). ²

After performing a BLAST search using the obtained ITS, RPB2, and TEF1-a sequences in the NCBI GenBank database, sequences that met specific criteria: ≥ 99% similarity for RPB2, ≥ 97% for TEF1-a, and ≥ 76% for ITS, were utilized to verify the identity of Trichoderma species in our phylogenetic analysis (Cai and Druzhinina, 2021). We retrieved homologous RPB2, TEF1-a, and ITS gene sequences of the isolates from GenBank. These sequences were aligned using the MUSCLE program (Edgar, 2004), and the resulting alignment was further refined using BioEdit 7.2.5 (Hall, 1999; Hall, 2011). Finally, we concatenated the gene sequences using Phylosuit V1.2.2 (Zhang et al., 2020). For the phylogenetic analysis, we employed the Maximum-Likelihood (ML) method using PhyML 3.0 (Guindon et al., 2010). The best substitution model was determined with PartitionFinder v2.1.1 (Lanfear et al., 2017). To assess statistical support, we conducted bootstrapping with 1,000 replicates (ML). Detailed lists of the fungal isolates used in this study can be found in Table 1 and Supplementary Table S1. The resulting ML tree was visualized using Figtree v1.4.4, ³ providing a clear representation of the phylogenetic relationships among the isolates.

Pathogenicity experiments were conducted following Koch’s postulates, with each experiment replicated twice to ensure accuracy. Fully colonized substrate bags containing G. sichuanense were sourced from the Panshi Mushroom Base in Jilin Province, China. These bags were placed in a growth room with controlled conditions, including a temperature range of 25–30°C and humidity levels set between 80 and 90%, to promote fruiting. Once the fruiting bodies were formed, the bottom surface of the cap and the stipe were meticulously damaged using a sterilized needle. Subsequently, they were inoculated with a spore suspension of the isolates at a concentration of 1 × 10⁵ spores per milliliter. As a comparison, the control group was inoculated with sterilized distilled water.

For each strain, six bags of G. sichuanense were inoculated. The development of symptoms was monitored daily for a period of 14 days. To confirm the causative agents of green mold disease, the pathogens were re-isolated from the inoculated G. sichuanense showing green mold symptoms. Identification was performed using the aforementioned morphological and molecular methods, considering strains that matched the original inoculum as the causative agents of green mold disease.

To assess the aggressiveness of the isolates, a subset of nine isolates representing nine different species was selected from the total of 47 isolates. The experiments were performed with three replicates, following the procedure outlined below. Mycelial agar plugs with a diameter of 8 mm were obtained from the advancing edge of 10-day-old G. sichuanense colonies. These plugs were then inoculated onto potato dextrose agar (PDA) plates, positioned 1 cm from the edge of Petri plates with a diameter of 9 cm. After 7 days, mycelial plugs from Trichoderma cultures were inoculated in the same manner, but on the opposite side of the plate, 1 cm away from the edge. The growth of Trichoderma species in confrontation with G. sichuanense mycelia was carefully observed and recorded.

To evaluate the efficacy of fungicides against green mold in mushrooms, a preliminary screening of six fungicides (mancozeb, chlorothalonil, fludioxonil, carbendazim, prochloraz, and prochloraz-Mn) was conducted. Stock solutions of each fungicide at a concentration of 100 mg/mL were prepared by dissolving them in sterilized distilled water. The growth inhibition rate of the fungi was assessed through mycelial growth assays.

PDA medium plates with different concentrations of each fungicide were prepared by adding the appropriate volume of the stock solution to sterilized distilled water. Mycelial plugs with a diameter of 7 mm were obtained from the edges of 3-day-old colonies grown on PDA and placed at the center of the PDA plates containing varying fungicide concentrations. All plates were then incubated at 25°C for 3 days. The growth inhibition rate of the mycelia was calculated using the formula i = a 1 − a 2 a 1 × 100 , where “i” represents the growth inhibition rate, “a1” is the hyphae area of the untreated pathogen, and “a2” is the hyphae area of the treated pathogen (Etebarian et al., 2005).

Each fungicide treatment and the control were replicated on three plates, and the experiment was repeated twice. Based on the preliminary screening results of the six fungicides using the nine isolates, a suitable fungicide was selected. The sensitivity of G. sichuanense to these fungicides was further tested using the same method described above.

The sensitivity of the fungi to fungicides was determined by measuring the fungicide concentration that inhibited fungal development by 50% [half maximal effective concentration (EC50)] (Wong and Midland, 2007; Kim et al., 2020). The relative growth (RG) of the fungi at a specific fungicide concentration was calculated as a percentage of fungal growth compared to the control plates. The EC50 value was obtained by performing linear regression analysis on the probit-transformed relative inhibition values (1 - RG) at log10-transformed fungicide concentrations. The EC50 value for each isolate was calculated as the average of three experiments. The correlation coefficients (r) among EC50 values for different fungicides were determined using statistical algorithms provided by SAS software (version 9.4 for Windows; SAS Institute, Cary, NC, U.S.A.).

The fruiting bodies of G. sichuanense exhibited symptoms of green mold disease, which were visually distinct. Infected basidiomata displayed a layer of green mycelia, leading to decay and withering of the affected fruiting bodies (Figure 1). The severity of the disease was evident, as it progressed rapidly, particularly after watering flushes of the fruiting bodies. The development of symptoms followed a specific pattern: initially, white spots and mycelium appeared on the infected fruiting bodies. Under hot weather conditions or high humidity, there was a significant proliferation of green conidia within a short time, gradually covering the entire surface of the fruiting bodies. Subsequently, the spores dispersed through various means, such as water flow, human movement, or wind, resulting in the demise of G. sichuanense fruiting bodies and the loss of their ability to produce spores.

To investigate the pathogens responsible for green mold disease, we conducted fungal isolation from the infected fruiting bodies. A total of 47 pathogens were isolated and identified during the study (Tables 1 2; Supplementary Table S1). Among the isolated pathogens, we identified one strain of T. harzianum in Zhejiang Province and three strains in Hubei Province, namely T. koningiopsis, T. paratrovide, and T. virens. Interestingly, in Jilin Province, we observed a diverse range of strains, with a total of nine different species identified (Table 2). These pathogens were characterized by their ability to induce the distinctive symptoms associated with green mold disease on G. sichuanense.

Using the classification methods proposed by Bissett (1984), Gams and Bissett (1998), and Park et al. (2006), we conducted a meticulous examination of colony shape, conidia, conidiophore size, chlamydospores, and pigmentation (Figures 2, 3) to identify nine Trichoderma species. The isolated species include T. ganodermatiderum (Figures 2A–G), T. citrinoviride (Figures 2H–L), T. hamatum (Figures 2M–P), T. asperellum (Figures 2Q–U), T. guizhouense (Figures 2V–Z), T. harzianum (Figures 3A–C), T. virens (Figures 3D–F), T. paratroviride (Figures 3G–J), and T. koningiopsis (Figures 3K–O).

After 1 day of cultivation at 25°C, all strains displayed white villous colonies on PDA, SNA, and CMD media. By the fifth day, light green to dark green sporulation bands emerged on all media, gradually extending toward the center. CMD and SNA media supported the growth of relatively thin colonies. By the seventh day, green spores were dispersed throughout the entire plate, exhibiting a grayish green or chartreuse color (Figure 4). T. citrinoviride exhibited the production of a yellow pigment at a later stage (Figure 4B), and some exhibited concentric rings (Figure 4).

Microscopic analysis unveiled notable distinctions in the morphology of conidiophores, phialides, and conidia among the Trichoderma species. T. ganodermatedrum (Figures 2A–G), T. asperellum (Figures 2Q–U), T. harzianum (Figures 3A–C), and T. koningiopsis (Figures 3K–O) exhibited dendriform branches and green spherical or ellipsoidal spores. While T. ganodermatedrum (Figures 2A–G) and T. harzianum (Figures 3A–C) shared similar spore sizes, the former displayed densely distributed conidiophores, whereas the latter had sparser conidiophore clusters. T. asperellum (Figures 2Q–U) and T. koningiopsis (Figures 3K–O) exhibited similar spore sizes, but there were significant differences in the sizes of their phialides. Specifically, T. koningiopsis had phialides measuring 5.0–7.5 × 3.0–4.8 μm (Figures 3K–M), while T. asperellum had phialides measuring 7–11 × 2–4 μm (Figures 2Q–S).

T. virens presented irregular branches at the top of its conidiophores, often accompanied by 3–6 closely arranged phialides, resulting in a more complex structure (Figures 3D–F). T. hamatum displayed highly branched conidiophores, primarily with opposite lateral branches and a few solitary branches (Figures 2M–P). The phialides of T. hamatum were densely packed, short, and round, measuring 5–7.5 × 3.0–4.4 μm (Figures 2M–N). In the case of T. citrinoviride, its conidiophores appeared either opposite or alternate, and it possessed small spores measuring 2.9–4.0 × 1.8–2.2 μm (Figures 2H–L). While T. guizouense (Figures 2V–Z) and T. paratroviride (Figures 3G–J) featured nearly spherical spores, the former exhibited conidiophores in pairs or whorls, with phialides typically arranged in groups of 2–4. Conversely, T. guizouense predominantly displayed conidiophores in 2–3 whorls, occasionally occurring solitary, and its phialides were symmetrically distributed (Figures 2V–X). Notably, T. hamatum (Figure 2P), T. asperellum (Figure 2U), T. guizouense (Figure 2Z), and T. koningiopsis (Figure 3O) exhibited abundant chlamydospores in later stages. For further details regarding the specific characteristics of each Trichoderma isolate, please consult Table 3.

The TEF1-a and RPB2 gene sequences of all Trichoderma isolates were compared to the NCBI database using BLAST analysis. Matches exhibiting a high similarity level (≥90%) were chosen for subsequent analysis. A phylogenetic analysis was conducted using the concatenated sequences of the TEF1-a and RPB2 genes from all Trichoderma isolates. The analysis revealed that the Trichoderma isolates could be classified into nine distinct clades: T. asperellum, T. citrinoviride, T. ganodermatiderum, T. guizhouense, T. hamatum, T. harzianum, T. koningiopsis, T. paratroviride, and T. virens (Figure 5). The phylogenetic trees were constructed using a dataset consisting of 19 sequences derived from two gene loci (TEF1-a and RPB2) obtained from a total of 47 samples. Among these sequences, 38 were newly generated, including 19 TEF1-a sequences and 19 RPB2 sequences. For more detailed information on the specific characteristics of each Trichoderma isolate, please refer to Table 1 and Supplementary Table S1.

In the pathogenicity test, mechanical damage was induced on the fruiting bodies followed by in vitro inoculation using a spore suspension. Two weeks after inoculation, all Trichoderma species showed similar green mold symptoms, as observed in Figure 6. Initially, small oval spots with white to pale green centers surrounded by a chlorotic area appeared on the G. sichuanense fruiting bodies 7 days post-inoculation. Over time, these lesions progressively increased in size and merged together. In severe cases, the infected fruiting bodies were completely covered by green spores. These symptoms observed under greenhouse conditions were consistent with the field symptoms of G. sichuanense. No symptoms were observed in the control group (Figure 6A).

Additionally, all Trichoderma species were consistently re-isolated and confirmed using morphological and molecular methods, while no Trichoderma isolates were obtained from the control group, satisfying Koch’s postulates. The pathogenicity study demonstrated that all Trichoderma isolates induced green mold disease in G. sichuanense fruiting bodies upon inoculation. Among the isolates, T. harzianum exhibited the highest virulence (Figure 6J), followed by T. citrinoviride (Figure 6C), T. paratroviride (Figure 6H), T. guizhouense (Figure 6E), T. ganodermatiderum (Figure 6B), T. asperellum (Figure 6D), T. virens (Figure 6G), T. koningiopsis (Figure 6I), and T. hamatum (Figure 6F).

To assess the impact of different Trichoderma species on G. sichuanense mycelia, we conducted plate dual culture experiments. The results revealed that all Trichoderma species inhibited the growth of G. sichuanense mycelia and produced antagonistic lines. However, there were variations in the interactions between the nine Trichoderma species and G. sichuanense mycelia (Supplementary Figure S2). Notably, T. ganodermatiderum (Supplementary Figure S2A), T. citrinoviride (Supplementary Figure S2B), T. asperellum (Supplementary Figure S2D), and T. paratroviride (Supplementary Figure S2H) exhibited significant inhibition on G. sichuanense mycelial growth, while T. hamatum (Supplementary Figure S2C), T. guizhouense (Supplementary Figure S2E), T. harzianum (Supplementary Figure S2F), T. virens (Supplementary Figure S2G) and T. koningiopsis (Supplementary Figure S2I) showed relatively milder inhibition.

In terms of mycelial morphology, Trichoderma mycelia demonstrated the ability to overgrow and spread on G. sichuanense mycelia, leading to the formation of irregular conidial clusters (Supplementary Figure S2). This resulted in the gradual withering of G. sichuanense mycelia. In some cases, certain Trichoderma strains completely covered the G. sichuanense mycelium with their spores. Additionally, we observed various pigments and antagonistic streaks on the back of the culture medium (Supplementary Figure S2).

In order to assess the effectiveness of fungicides against green mold disease, we conducted tests using six different fungicides in this study. Initially, we selected nine Trichoderma isolates to evaluate their sensitivity to these fungicides. The results showed that the inhibitory effect of the fungicides on Trichoderma growth varied, with stronger inhibition observed at higher fungicide concentrations. Table 4 presents the results, highlighting that prochloraz-manganese exhibited the highest inhibitory effect among the tested fungicides, as indicated by its minimum EC50 value, while Mancozeb showed the weakest inhibition with the highest EC50 value.

Furthermore, we evaluated the inhibitory effects of the fungicides on the growth of G. sichuanense mycelium through extensive tests. Significant variations in the effects of the six fungicides on the growth of G. sichuanense mycelium were observed, which generally aligned with the effects observed on Trichoderma strains. Notably, prochloraz-manganese had the least impact on the growth of G. sichuanense mycelium, displaying the highest EC50 value while exhibiting the strongest inhibitory effect on Trichoderma mycelium (Table 4). These findings suggest that low concentrations of prochloraz-manganese can be effective in controlling Trichoderma. Additionally, prochloraz and carbendazim demonstrated good inhibitory effects on all Trichoderma strains (Supplementary Figures S3–S11).