Cordyceps cicadae strain 2-2, stored at Zhejiang University, was cultured in YES medium (Supplementary Table S8) for 10 days at 25°C. The vector pKO1B for knockout and pKD5 for complementation was stored in our laboratory. Escherichia coli DH5α (Sangon Biotech, Shanghai) for plasmid preservation and Agrobacterium tumefaciens AGL1 (Tsingke Biotechnology, Beijing) for Agrobacterium-mediated transformation (ATMT) were cultured in LB medium at 37°C and 28°C, respectively. The preserved strains and bacteria were stored at −80°C with the addition of 20% glycerol.

Using the principle of homologous recombination, we utilized the pKO1B vector, including both upstream and downstream homology arms, to the target gene as well as HPH fragments for knockout vector construction. To get a complementation vector, we also constructed the pKD5 vector including the target gene. U1/U2 and D1/D2 primers were used to amplify the 1.0–1.5 kb upstream and downstream fragments of the target gene from the WT genome. Validation primers KOCX/MRBF and Trptrcx/cmcx2 were used to verify the vectors, respectively. The transformation of bacteria was achieved by the freeze-thaw method. The fungal transformation procedure was conducted in accordance with the methodology described by Li et al. (2022a). The knockout mutants were confirmed as a single copy via qPCR using the 2^–∆∆CT method and were repeated three times. The primers used in the above experiments are listed in Supplementary Table S1.

Because cyanidin is not stable for detection and transfers to cyanidin-3-O-glucoside (C3G) quickly in cells, we detected the amount of C3G, which is a stable and downstream substance of cyanidin (Bu et al., 2020).

The extraction of C3G: the mycelium was introduced into the CM liquid medium (Supplementary Table S9) for incubation in a shaker (25°C, 150 rpm) over a period of 2 days. A measurement of 1.0 g of freeze-dried mycelia samples was collected and mixed with 10 mL of a 1% hydrochloric acid-methanol solution. The samples were subjected to 24 h ultrasonic cleaning for fixing extraction. The resulting filtrate was obtained by filtering through three layers of filter paper and transferring 1.0 mL into a 25 mL brown volumetric bottle. Prior to sample testing, the filtrate was shaken with methanol and filtered through a 0.25 μm microporous filter membrane.

The determination of C3G: a series of cyanidin-3-o-glucoside control solutions were prepared at concentrations ranging from 0.5 to 25.0 μg/mL and filtered through a 0.25 μm microporous filter membrane before testing. Phenomenex Synergi Hydro-RP column (250 mm × 4.60 mm, 5 μm) was used. The mobile phase consisted of 0.5% phosphoric acid solution as phase A and water acetonitrile as phase B (50:50, V/V). Gradient elution was carried out with the following program: 0 min (18%B), 23 min (50%B), 28 min (18%B), 33 min (18%B), and 40 min (stop); flow rate: 0.8 mL/min; detection wavelength: 520 nm; column temperature: 30°C. The standard solution of each volume concentration was repeated three times to obtain the corresponding peak area (Supplementary Figure S1B). A standard curve was constructed using C3G concentration as the abscissa and its corresponding peak area as the ordinate, with a linear regression equation calculated. The resulting regression equation between C3G concentration and peak area was y = 20.559x − 2.3947 (R² = 0.9998), demonstrating a strong linear correlation within the range of 0.5–25 μg/mL (Supplementary Figure S1C).

The content of the active ingredient: polysaccharide: the standard curve was drawn using the glucose standard, the polysaccharide was extracted by the phenol sulfuric acid method, and the OD value was determined at 490 nm. Adenosine: the standard curve was generated through HPLC analysis of adenosine standards. Phenomenex Synergi Hydro-RP column (250 mm × 4.60 mm, 5 μm) was used. The mobile phase consisted of 0.5% phosphoric acid solution as phase A and methanol as phase B. Gradient elution: 0 min (100%A), 13 min (85%A), 30 min (40%A), and 40 min (stop); flow rate: 1 mL/min; detection wavelength: 260 nm; sample size: 20 μL; column temperature: 25°C. The standard curve of adenosine was obtained by plotting the volume concentration of adenosine on the horizontal axis and its corresponding peak area on the vertical axis, resulting in a linear equation y = 122.8x − 7.0017 (R² = 0.9999). The adenosine was extracted by sonication and methanol. Chitinase and Anthocyanin: Chitinase and anthocyanin were analyzed using the kit (Boxbio, Beijing) and microplate reader.

The rate of growth: the 5 mm agar disks with mycelium were inoculated on a 7 cm YES plate medium and incubated at 28°C under alternating conditions of 16 h of light and 8 h of darkness. After 8 days, the colony diameter was measured.

The yield of spores: the 5 μL of spore suspension was inoculated on a 9 cm plate PDA medium and cultured for 9 days. A measurement of 4 mL of ddH₂O was added to the plate, and the spores were washed with a spreader before being filtered through three layers of filter paper to obtain the spore solution. The yield of spores was determined by observing and counting with a hemocytometer and microscope.

The mycelial growth under chemical stress: the 5 mm agar disks with mycelium were inoculated on the YES media supplemented with different chemicals, including 100 μM menadione (VK₃), 1 mM hydrogen peroxide (H₂O₂), rose Bengal (RB), 0.6 g/mL Congo red (CR), 0.4 M sodium chloride (NaCl), 0.4 M potassium chloride (KCl), and 0.5 M Sorbitol. The colony diameters were also measured at 8 days post incubation at 25°C in the dark. Each group was repeated three times with five replicates.

The formation of fruiting body: the strains were inoculated onto a sterilized medium of wheat seeds. The fruiting body formation was observed after dark culture for 7 days and light culture for 2 weeks at 25°C.

The bacteriostatic ability: Pseudomonas solanacearum, Xanthomonas oryzae, Dickeya dadantii NCPPB 898, and Escherichia coli were inoculated on NA solid medium. The crude extracts of wild type and mutant, each with a volume of 50 μL, were added into the 5 mm hole of the medium. The diameter of the inhibition zone was recorded after incubation at 28°C overnight.

The genes related to the synthesis pathway of cyanidin were identified by the local blast (Supplementary Table S1), and the expression levels were analyzed by RT-qPCR.

In this study, the non-targeted metabolomics analysis of secondary metabolites was conducted on WT and mutants ΔCcDFR and ΔCcOMT9 using liquid chromatography-mass spectrometry (LC–MS) technology (Bioincloud, Shenzhen), and the specific analytical method was done according to the methodology described by Shan et al. (2023).

Mycelium was harvested from YES plate medium after being cultured for 10 days at 25°C (16 h light and 8 h darkness) and total RNA was extracted. The cDNA and RNA libraries were sequenced according to previous research (Chen et al., 2018). Ribosomal reads were removed by bowtie2, and the remaining total reads were used for subsequent transcriptome analysis. Total Reads (-rRNA) were aligned with the genome of C. cicadae using HISAT2 software. DESeq2 (Love et al., 2014) software was utilized for differential expression analysis between the control and treatment groups. Differentially expressed genes (DEGs) were identified based on FDR < 0.05 and |log2FC| > log2(2), with up-regulated and down-regulated DEGs being obtained. MeRIP (Methylated RNA Immunoprecipitation), based on the principle of m6A-specific antibody recognition and binding to m6A-modified mRNA, is a technique used for enriching methylated RNA followed by high-throughput sequencing analysis to study regions of m6A methylation modification across the entire transcriptome. MACS2 (version: 2.1.2) software was used to identify read-enriched regions from MeRIP-seq data (Zhang et al., 2008). Dynamic Poisson distribution was used to calculate the p-value of the specific region based on the unique mapped reads. The region would be defined as a peak when the q-value <0.05. The DiffBind (version 2.8) software was used to analyze the RNA methylation rate between groups (Stark and Brown, 2011). The relative methylation rate of each peak was calculated using MeRIP data and input data. We identified peaks with log2|FC| ≥ 1 and FDR ≤ 0.05 in a comparison as significant differential peaks for subsequent peak-related genes GO and KEGG enrichment analysis. GO and KEGG databases were employed to significantly enrich all DEGs and differential peaks in the WT and mutant ΔCcOMT9, thereby identifying major metabolic and signal transduction pathways.

The C. cicadae strain 2-2 can secrete red pigment to medium when inoculated on PDA. We found that C3G, the important anthocyanin, could be detected by HPLC (Supplementary Figure S1A). In the process of investigation on the functional genes related to growth, development, and metabolism, we found two genes in strain 2-2 genome, marked as FUN_1459 and FUN_9127 in our Lab, involved in red color pigment biosynthesis, for the mutants of the two genes, changed color when inoculated in PDB after gene knock-out. Gene FUN_1459 contains 1,142 bp, including 1 intron, and encodes 353 amino acids (accession No. C_AA042572.1). The protein sequences with high similarity and the sequences of model organisms were collected from NCBI and analyzed by the neighbor-joining method (NJ) to establish the phylogenetic tree (Supplementary Figure S2). FUN_1459 has the highest homology to Cordyceps fumosorosea ARSEF 2679 dihydroflavonol-4-reductase (accession No. XP_018706651.1), followed by dihydroflavonal-4-reductase in Cordyceps javanic (accession No. TQW08783.1), Beauveria brongniartii RCEF 3172 (accession No. OAA49002.1), and Cordyceps militaris CM01 (accession No. XP 006670355.1). However, it is hard to say whether FUN_1459 is homologous to plants like Arabidopsis thaliana, Pyrus communis, and Dianthus caryophyllus. When blast analysis was performed in the genome of Aspergillus sydowii H-1, we revealed that FUN_1459 gene was homologous with DFR_1; the key gene encoded dihydroflavonol-4-reductase in the anthocyanin biosynthesis pathway (Bu et al., 2020), and we named FUN_1459 as CcDFR (accession No. C_AA042572.1). Gene FUN_9127 contains 1,521 bp, including 1 intron, and encodes 351 aa. Blastp showed that the homologous protein of FUN_9127 is all putative methyltransferase and actin-binding protein in fungi. As shown in the phylogenetic tree, FUN_9127 has the highest homology to methyltransferase (Trm140) in Beauveria bassiana (accession No. PMB67945.1), followed by acting-binding protein in Trichoderma simmonsii (accession No. QYT05915.1). It seems there was no obvious homology to Saccharomyces cerevisiae S288C (accession no. Q08641.3) and Schizosaccharomyces pombe 972 h- (accession no. Q9P7L6.2). There is no study on B. bassiana about Trm140. However, in consideration of Trm140 as a methyltransferase responsible for 3-methylcytidine modification in the tRNA anti-codon loop in model organism S. cerevisiae (accession no. Q08641.3), we named FUN_9127 as CcOMT9 (accession no. C_AA042573.1).

Knockout vectors and their complementary vectors with corresponding resistance were obtained based on the schematic diagram in Figure 1 and transformed by the ATMT method. The mutants exhibiting hygromycin B resistance generated with ATMT were validated following PCR (Figures 1A,B). The single copy of the mutant strains was determined by RT-qPCR. The complementation mutant was confirmed by obtaining a band size identical to that of the WT (Figures 1C–H).

During the process of liquid fermentation, notable distinctions in color were observed between the WT and ΔCcDFR mutants (Figure 2A). Upon knockout of methyltransferase-related genes, we observed significant differences between the ΔCcOMT9 mutant and WT, with the former having a more obvious red pigment production (Figure 2A).

To further substantiate that CcDFR and CcOMT9 are influencing C3G yield and resulting in color change, we conducted quantitative analysis via HPLC (Shirley et al., 1995). The C3G content of the ΔCcDFR mutant was 0.7194 ± 0.17 mg/g and the ΔCcOMT9 mutant was 5.7748 ± 0.50 mg/g, while the WT was 3.1016 ± 0.31 mg/g (Figures 2B,C).

Previous research demonstrated the antibacterial ability of C. cicadae, cyanidin, and C3G (Gopu and Shetty, 2016; Zhang et al., 2017; Cen et al., 2021; Li et al., 2022b). The inhibitory zone method was employed to assess the antibacterial ability, and the findings indicated that the inhibitory zone of ΔCcDFR mutant was significantly reduced compared to that of WT. C. cicadae exhibits inhibitory effects against P. solanacearum, X. oryzae, D. dadantii NCPPB 898, and E. coli, and knockout of CcDFR significantly reduces the inhibitory zone size. Specifically, the inhibitory zone against P. solanacearum was 27.23 ± 0.03 mm, while that of WT was 31.33 ± 0.05 mm. The inhibitory zone of ΔCcDFR was 28.28 ± 0.05 mm for X. oryzae and 30.53 ± 0.06 mm for WT, indicating that CcDFR plays a crucial role in regulating the antibacterial activity of C. cicadae (Supplementary Table S2).

ΔCcOMT9 exhibited a significantly larger inhibitory zone compared to WT treatment. Knockout of CcOMT9 significantly increased the inhibitory zone size against P. solanacearum, X. oryzae, D. dadantii NCPPB 898, and E. coli, indicating that CcOMT9 affects the antibacterial activity of C. cicadae, while knockout of CcDFR significantly decreased the inhibitory zone size (Supplementary Table S2).

Compared to the WT, the ΔCcDFR and ΔCcOMT9 mutant showed significant differences in polysaccharide content, while the former was down-regulated measuring 4.02 ± 0.04% and the latter was up-regulated measuring 10.36 ± 0.67%, which was a decrease of 46.33% and an increase of 38.32% from the WT’s value of 7.49 ± 0.17%, respectively (Figure 3A). The effect of CcDFR on adenosine production was analyzed by HPLC. Following ultrasonic extraction of adenosine, deletion of CcDFR resulted in a significant reduction in adenosine content for ΔCcDFR mutant compared to WT, with the former yielding 124.51 ± 4.44 μg/mL and the latter yielding 180.86 ± 0.52 μg/mL, representing a decrease of 31.16%. Deletion of CcOMT9 resulted in a significant increase in adenosine content for ΔCcOMT9 mutant compared to WT, with the former yielding 263.2 ± 18.09 μg/mL and the latter yielding 180.86 ± 0.52 μg/mL, representing an increase of 45.65% (Figure 3B). Compared to the WT chitinase activity of 10.52 ± 0.75 U/g, the chitinase activity of ΔCcDFR mutant was significantly reduced by 43.06% to only 5.99 ± 0.63 U/g, and ΔCcOMT9 mutant was significantly reduced by 25.10% to only 7.88 ± 0.26 U/g (Figure 3C). The content of anthocyanins in the ΔCcDFR mutant was significantly reduced by the pH-differential method (Figure 3D). The ΔCcDFR mutant exhibited a significantly lower content of anthocyanins at 0.15 ± 0.0056 μmol/g, representing a reduction of 65.12% compared to 0.43 ± 0.0043 μmol/g of WT, while the ΔCcOMT9 mutant exhibited a significantly higher content of anthocyanins at 0.61 ± 0.04 μmol/g, representing an increase of 41.86% compared to 0.43 ± 0.0043 μmol/g of WT.

Compared to the WT and complementation strain, the ΔCcDFR mutant exhibited a significant reduction in colony diameter after 10 days of growth on YES medium at 25°C, with a decrease in growth rate of 25.4% compared to the WT strain. Additionally, thinner aerial mycelium was observed and the sporulation yield decreased by 78.26% relative to the WT (Figures 4A,B).

During the fruiting body stage, it is observed that the ΔCcOMT9 mutant exhibits precociousness, shorter synnemata length, lighter color, and earlier “flowering” compared to the WT, whereas the ΔCcDFR mutant fails to produce fruiting bodies (Figure 4C), indicating that CcDFR plays a crucial role in regulating spore production and growth rate as well as promoting fruiting body formation during C. cicadae development.

In the process of growth, differentiation, reproduction, and invasion, C. cicadae will be stressed by the external environment. Compared with the WT (54.27 ± 2.5%), Sorbitol has more significant inhibition on the ΔCcDFR mutant (76.74 ± 2.2%). On the YES + RB condition, the growth inhibition rate of the mutant (48.17 ± 1.7%) was observed to be significantly higher than that of the WT (34.05 ± 4.1%) (Figure 5A). The inhibitory rate of KCl for WT was 40.13 ± 0.92% while the ΔCcOMT9 mutant was 68.30 ± 0.30%. Moreover, on the YES + VK3 plate, the growth inhibition rate of the ΔCcOMT9 mutant (45.21 ± 0.26%) was also markedly greater than that of the WT (22.29 ± 0.34%) (Figure 5B). These results indicated that CcDFR and CcOMT9 play an important role in the resistance to cell wall stress.

To investigate the impact of CcDFR and CcOMT9 on genes related to the cyanidin synthesis pathway in C. cicadae, we extracted RNA at different growth stages and assessed expression levels of cyanidin-related genes using RT-qPCR. On the second day of culture, compared to the WT, there was a varying degree of down-regulation in genes related to cyanidin synthesis of ΔCcDFR mutant (around 0.05, 0.46, 0.44, 0.01, 0.13, and 0.40-fold, respectively), indicating that deletion of CcDFR could reduce the expression of other genes in the cyanidin synthesis pathway and affect normal cyanidin synthesis. Compared with day 2, the expression levels of genes related to cyanidin synthesis were down-regulated to varying degrees in the WT cultured on day 6 (around 0.07, 0.56, 0.74, 1.49, 0.34, and 0.51-fold, respectively). This suggests that when cyanidin accumulation reaches a certain threshold, the genes involved in its synthesis may be affected by its content and undergo feedback inhibition. The genes associated with cyanidin synthesis are up-regulated to varying degrees in the ΔCcOMT9 mutant on both day 2 and day 6 of culture compared to the WT (around 1.12, 1.54, 1.34, 1.06, 2.24, and 1.16-fold for day 2 and 2.26, 2.24, 1.08, 1.29, 1.12, and 1.15-fold for day 6, respectively) (Table 1). On day 6, the expression levels of genes related to cyanidin biosynthesis exhibited varying degrees of reduction in ΔCcOMT9 mutant compared to day 2, except FUN_008622 (around 0.14, 0.81, 0.59, 1.78, 0.17, and 0.50-fold, respectively), potentially suggesting that excessive cyanidin production exerts a feedback inhibition effect on the expression of relevant synthetic genes.

The PCA results revealed a distinct separation trend among the WT, ΔCcDFR, and ΔCcOMT9 mutants on the two-dimensional plot, with all strains exhibiting aggregation within this group (Figure 6A). The results indicate significant differences among the WT, ΔCcDFR, and ΔCcOMT9 mutants, with high stability observed. The results of the cluster heat map analysis revealed distinct differences among the WT, ΔCcDFR, and ΔCcOMT9 mutants, with a consistent and highly aggregated trend in metabolite accumulation observed within each group, and the phenylalanine content was significantly up-regulated in CcOMT9 mutants (Figure 6B). After comparing the differential metabolites among the WT and ΔCcDFR and ΔCcOMT9 mutants, significant differences were observed in flavonoids, including phenylalanine (a precursor to cyanidin synthesis) and cyanidin. The C3G (kuromanin) content of the ΔCcDFR mutant was found to be significantly lower than that of the WT and ΔCcOMT9 mutant, while the phenylalanine content of the ΔCcOMT9 mutant was significantly higher than that of the WT and ΔCcDFR mutant (Figure 6C).

We conducted transcriptome sequencing to further compare the difference between WT and ΔCcOMT9 mutants. Among them, 541 were up-regulated and 756 were down-regulated among a total of 1,297 differential genes between the WT and ΔCcOMT9 mutant (Supplementary Figures S3A,B). KEGG analysis showed that 105 DEGs were enriched between the WT and ΔCcOMT9 mutant, and the metabolism pathway mainly involved in the cell cycle, flavonoid biosynthesis, etc. was analyzed. The top 20 metabolic pathways with significant enrichment (p < 0.05) were selected for display in an enrichment bubble map, and four of these pathways were found to be related to anthocyanin biosynthesis (Supplementary Figure S3C). The GO (Gene Ontology) enrichment analysis was carried out on the DEGs of the two treatment groups (Supplementary Figure S3D). Among biological processes, DEGs are primarily concentrated in cellular processes, metabolic processes, biological regulation, regulation of biological processes, response to stimulus, and so on. In terms of cell components, DEGs are mainly found in the cellular anatomical entity, protein-containing complex, ribonucleoprotein component, and so on. In molecular function, DEGs are mainly enriched in binding, catalytic activity, transporter activity, transcription regulator activity, and ATP-dependent activity. The down-regulation of 9 out of 10 genes involved in the phenylalanine metabolism pathway suggests a blockage in the ΔCcOMT9 mutant. This blockage leads to the accumulation of phenylalanine and an increase in cyanidin biosynthesis ability. This conclusion is supported by the metabolomics analysis, which shows the accumulation of phenylalanine in the ΔCcOMT9 mutant.

To validate the precision of RNA-seq sequencing analysis outcomes, we picked 15 genes from the DEGs across the four metabolic pathways (Supplementary Table S3) and tested the expression of these genes after different treatments by RT-qPCR. The results of RT-qPCR detection were basically consistent with those of RNA-seq, which confirmed the reliability of transcriptome data (Supplementary Figure S4).

The differences in mRNA methylation between WT and ΔCcOMT9 mutants were analyzed to check if CcOMT9 can be responsible for mRNA methylation in C. cicadae. Analysis of the relative methylation rate showed there were 401 up-regulated and 243 down-regulated differential peaks between WT and ΔCcOMT9 mutant (Supplementary Figures S5A,B). Functional enrichment analysis of genes related to differential peaks was conducted using GO and KEGG analysis. GO analysis clustered into the cellular process and metabolic process in biological process, cellular anatomical entity and protein-containing complex in cellular component and binding and catalytic in molecular function activity (Supplementary Figure S5D). KEGG pathway enrichment analysis clustered to phosphatidylinositol signaling system, MAPK signaling pathway-yeast, and aminoacyl-tRNA biosynthesis (Supplementary Figure S5C).