Tolypocladium guangdongense strain CCTCCM206051 was used in this study and deposited in the Agricultural Culture Collection of China. Several mycelial plugs were inoculated into 200ml improved CYM medium (20g dextrose, 2g yeast extract, 2g peptone, 1g KH₂PO₄, 1g NaNO₃, 0.5g K₂HPO₄, 0.46g MgSO₄•7H₂O, and 0.1g vitamin B1 in 1l ddH₂O), and cultured for 10days in an incubator shaker at 140 rotations/min (rpm) and 23°C. Next, mycelial pellets were transferred to PDA (potato dextrose agar) medium with or without 10μM H₂O₂ and 2mg/ml CR and incubated for 14days at 23°C. Mycelia cultured at 23°C for 4days were subjected to 14days 30°C or 14°C treatment as heat treatment (HT) and cold treatment (CT), respectively. In addition, the samples of mycelium (My), P, and young fruiting body (FB) were collected as described previously (Wang et al., 2020a). All the samples were immediately frozen in liquid nitrogen and then were preserved at −80°C until metabolite, and RNA isolation were performed.

Samples (0.1g) from different experimental conditions were homogenized in 1ml of a methanol/water (50:50, v/v) mixture with 1.0ng/g [²H₄] MLT as an internal standard. After vortex-induced vibration for 15min, the mixture was centrifuged at 12,000rpm for 5min at 4°C, and the supernatant was collected. This procedure was repeated once. Both supernatants were merged and evaporated under a mild nitrogen stream. Then, the evaporated samples were re-dissolved in 100μl methanol/water (50:50, v/v) mixture. After 5min vortex-induced vibration, the mixture was centrifuged at 12,000rpm for 5min and filtered using a 0.22μm filter membrane. Intracellular melatonin levels were quantified by Wuhan Metware Biotechnology Co., Ltd. (Wuhan, China) using an liquid chromatography–tandem mass spectrometry (LC–MS/MS) platform (Waters Iclass-AB Sciex QTRAP 6500+) platform as described by Ye et al. (2017).

The indolic compounds tryptophan, tryptamine, 5HTryp, NAT, serotonin, NAS, and 5MT were purchased from Sigma-Aldrich (Madrid, Spain) and Shanghai Yuanye Bio-Technology Co., Ltd (China). These indolic compounds were extracted using the method described in our previous study (Wang et al., 2020a). Briefly, sample tissues (0.5g) from the CR and control (CK) groups was finely ground, followed by extraction in an 1ml buffer containing 50% methanol and 0.1% formic acid for 15min in a shaker. After ultrasound treatment for 30min in an ice bath, the mixtures were incubated at 4°C for 30min and then centrifuged for 20min at 12,000rpm. The extracts were filtered through a 0.3μm filter membrane and were analyzed using LC–MS/MS platform (Waters Iclass-AB Sciex 6500) platform at Beijing Genomics Institute (BGI)-Shenzhen (Wuhan, China) as described by Sara et al. (2019).

Genes related to melatonin biosynthesis in T. guangdongense were predicted by comparing the amino acid sequences with those of Oryza sativa and Mus musculus in T. guangdongense genome (BioProject: PRJNA399600). Tissue samples collected from My, P, and FB stages and the different periods of CR stress (6h, 12h, 1days, 2days, 4days, 8days, and 14days) and were used for RNA isolation using the OminiPlant RNA Kit (DNase I; CWBIO, Beijing, China). The mRNA was reverse transcribed into cDNA and used for quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) analysis of genes related to melatonin biosynthesis. The gene-specific primers used for gene quantification were designed using Primer Primer5.0, and their sequences were shown in Table_S1. Histone H4 and the vacuolar protein sorting gene VPS were used as the internal control genes. The qRT-PCR reaction was performed using an Applied Biosystems ABI 7500 instrument (ABI, Foster City, California, United States) and three biological and technical repeats. PCR reaction was performed with an initial denaturation for 3min at 95°C, followed by 40cycles of 10s at 95°C and annealing at 60°C for 20s. The relative expression levels of the target genes were calculated using the 2^−△△Ct method (Wang et al., 2020b).

Ten micrometer H₂O₂, 2mg/ml CR, and 10μM melatonin were added to PDA medium at approximately 55°C. Mycelial pellets activated for 10days were inoculated into PDA medium and cultured at 23°C. For the HT and CT groups, the PDA plates containing 10μM melatonin were placed in an incubator at 30°C and 14°C. After 24days of culture, the diameters of the fungal colonies were measured to assess the effects of melatonin on T. guangdongense mycelia under different experimental conditions. In addition, the groups containing 10μM melatonin were transferred into the cultivation room at a temperature of approximately 23°C and relative humidity of 60–70%. Meanwhile, the mycelial pellets were subjected to treatment of 10μM melatonin, and then were inoculated in rice medium. After 15days culture, they were transferred into the cultivation room at a temperature of approximately 23°C and relative humidity of 60–70%. The mycelia were illuminated (500lx) for 10h each day to explore the effect of melatonin on primordial formation.

Significant differences were determined using the Statistical Package for the Social Sciences program (SPSS 22.0, SPSS Inc., Chicago, IL, United States). Error bars represent as the standard deviation from the mean of 10 independent replications in every group. Lowercase letters and asterisks were used to refer to significant differences between each group and the control group (CK).

The levels of endogenous melatonin are regulated by multiple environmental factors. For example, endogenous melatonin synthesis was promoted by salt stress and a high temperature in sunflower and Arabidopsis seedlings (Mukherjee et al., 2014; Shi et al., 2015). In S. cerevisiae, a low temperature and low sugar level delayed the peaking of intracellular melatonin levels, whereas melatonin synthesis production was significantly promoted at 12°C (Wang et al., 2016; Morcillo-Parra et al., 2020a). To determine whether the intracellular melatonin levels of T. guangdongense markedly differ under different experimental conditions, T. guangdongense samples from three different developmental stages and four different abiotic stresses were collected, and melatonin contents were measured by LC–MS/MS.

As shown in Figure 1A, three types of abiotic stresses significantly promoted intracellular melatonin production, whereas H₂O₂ stress tended to a decrease trend in intracellular melatonin synthesis. Compared to the intracellular melatonin content of the CK, CR exhibited the highest increase level (approximately 55.6-fold), flowed by CT (approximately 10.6-fold), and HT (approximately 5.13-fold), which was similar to the increase of melatonin levels in V. volvacea under Cd stress and A. bisporus during low temperature storage (Aghdam et al., 2019; Gao et al., 2020). Compared to My stage, intracellular melatonin levels had approximate 16-fold and 28-fold increase at the P and FB stages, respectively (Figure 1B). These results suggested that intracellular melatonin may be involved in the primordial formation, fruiting body growth and the response of T. guangdongense to abiotic stresses, which need to be confirmed by further study.

In S. cerevisiae, Muñiz-Calvo et al. (2019) deciphered the melatonin biosynthesis and metabolism based on the bioconversion of related metabolites. According to their results and metabolome analysis conducted in our laboratory, the intermediates related to melatonin biosynthesis were detected in the mycelia and fruiting bodies of T. guangdongense (unpublished data). To more accurately quantify their dynamic changes and identify the melatonin-synthesis pathway, biosynthetic intermediate levels were quantified in fungal samples collected after 14days CR stress and during different developmental stages. The amino acid sequences of proteins related to melatonin biosynthesis in O. sativa and M. musculus were characterized by BLAST analysis against the genome of T. guagndongense. Proteins with an E-value of <0.001 and >35% amino acid similarity were treated as the homologous proteins. In the genome of T. guangdongense, nine genes were predicted to be the homologous genes involved in melatonin biosynthesis (Supplementary Table S1).

Our LC–MS/MS analysis revealed that the abundance of most intermediates markedly changed under CR condition compared to the corresponding levels found under normal culture condition (Figure 2A). The levels of tryptophan, the precursor of melatonin, were significantly increased in the CR medium. Tryptamine levels were not significantly different in the CR and CK media; however, the levels of four other intermediates exhibited a remarkable decease under CR condition. This phenomenon was inconsistent with the significant increase of melatonin content detected after being subjected to 14days of CR stress. Therefore, the relative levels of these genes related to melatonin synthesis were analyzed during different periods of CR stress. As shown in Figure 2B, six genes were markedly upregulated in the early period but significantly downregulated in the late period, whereas the relative expression level of two genes showed an obvious down-regulation on days 8 and 14. Specifically, five genes (TDC-1, TPH, T5H, SNAT-1, and ASMT-2) were significantly upregulated during the first 8days of CR stress, but markedly downregulated on day 14. The relative expression levels of AADC and ASMT-1 decreased significantly on days 8 and 14. TDC-2 was markedly upregulated on day 8 but downregulated after 12h and on day 1. SNAT-2 was significantly downregulated on days 1, 2, and 8, whereas was markedly upregulted after 6h and on day 4. From the perspective of gene expression pattern at different periods of CR stress, we speculated that the intermediates were synthesized and began to accumulate on day 8 or during an earlier period of CR stress, resulting in the accumulation and increase of melatonin at 14 d of CR stress.

LC–MS/MS analysis showed that compared to My stage, the levels of five metabolites (tryptophan, tryptamine, 5HTryp, serotonin, and NAS) were significantly increased at the P and FB stages, whereas the levels of 5MT were markedly reduced (Figure 3A). In terms of the expression profile of genes related to melatonin synthesis, all the genes were significantly upregulated at P and FB stages compared to My stage (Figure 3B), being consistent to the higher levels of the intermediates at the P and FB stages. Comparing gene expression levels at the P and FB stages, the relative levels of six genes (AADC, TDC-2, TPH, T5H, SNAT-1, and ASMT-2) showed a significant upregulation at the FB stage, and two genes (TDC-1 and SNAT-2) were markedly upregulated at the P stage, in agreement with the quantitative analysis of the intermediate levels.

The synthesis pathway of melatonin was speculated in T. guangdongense, based on the above results and the related results from yeast and plants (Muñiz-Calvo et al., 2019; Nawaz et al., 2021). As shown in Figure 4, the putative melatonin-biosynthesis pathway in T. guangdongense was relatively complex compared to that in S. cerevisiae. In the first step of the proposed pathway, tryptophan was converted into tryptamine and 5HTryp by TDC/AADC and TPH, respectively. Later, this tryptamine was converted to NAT or serotonin by acetyltation (SNAT) and hydroxylation (T5H), respectively, whereas 5HTryp was converted to serotonin via 5-hydroxytryptophan carboxylation (TDC/AADC). Subsequently, serotonin was converted to NAS and 5MT by N-acetylation (SNAT) and O-methylation (ASMT), respectively, whereas NAT was hydroxylated by T5H to form NAS. Finally, T. guangdongense can covert NAS and five MT into melatonin by O-methylation (ASMT) and N-acetylation (SNAT), respectively. According to the above results, it was suggested that serotonin and NAS were the main melatonin-synthesis intermediates, and the melatonin-biosynthesis pathway in T. guangdongense was markedly similar to that in plants (Nawaz et al., 2021).

By interacting with other signaling molecules, such as reactive oxygen species and nitric oxide, melatonin participates in various physiological reactions and enhances the resistance of plants to abiotic stress (Yu et al., 2018; Qiao et al., 2019). In fungi, exogenous melatonin enhanced the resistance of S. cerevisiae and V. volvacea against oxidative stress and cadmium stress, respectively (Vázquez et al., 2017; Gao et al., 2020). Therefore, colony diameters under different experimental conditions were measured to evaluate the effect of exogenous melatonin on T. guangdongense mycelia.

Previously, it was found that treatment with 1μM and 1M melatonin obviously inhibited the mycelial growth of T. guangdongense, whereas no significant difference was found after treatment with 10μM or 100μM melatonin (Supplementary Figure S1). In addition, the mean of colony diameters in the 10μM melatonin group was the largest (4.23cm). Therefore, 10μM melatonin was selected for this study. In terms of the four abiotic stresses applied, 2mg/ml CR showed an evident promotion of the mycelial growth, whereas the other three stresses had a significantly negative effect on the mycelial growth of T. guangdongense (Figure 5). The stimulatory effect of 2mg/ml CR was reduced in the presence of exogenous melatonin (Figure 5A). Nevertheless, exogenous melatonin promoted the mycelial growth of T. guangdongense and enhanced its resistance to heat, cold, and H₂O₂ stresses (Figures 5B-D), which was consistent with previous results (Wei et al., 2014; Yu et al., 2018; Nawaz et al., 2021). These results suggested that melatonin plays important roles in the response of T. guangdongense to abiotic stresses.

As a growth-stimulating compound, melatonin is involved in several biological functions in plants such as the growth of roots and coleoptiles, seed germination, seedling growth, and senescence (Hernández-Ruiz et al., 2005; Wang et al., 2013; Arnao and Hernández-Ruiz, 2019). In this study, it was found that intracellular melatonin levels obviously increased at the P and FB stages (Figure 1B). Therefore, exogenous melatonin was added to PDA and rice medium to evaluate the effect of melatonin on primordial formation. Significantly more primordia were observed in the melatonin treatment group than those in the control group (Figure 6A). Additionally, the primordial density had an approximately one-fold increase in the melatonin treatment group (Figure 6B), indicating an important regulatory role in primordial formation in T. guangdongense. These results suggested that exogenous melatonin has the good potential for enhancing the cultivation of edible-medicinal mushroom.