The genome sequence of A. phaeospermum was downloaded from NCBI resources (Li et al., 2019) (NCBI GenBank accession number QYRS00000000). The SDR protein sequences of Saccharomyces cerevisiae, Fusarium oxysporum, Magnaporthe oryzae, and Rosellinia necatrix were downloaded from NCBI¹ and were used for BLASTP search and phylogenetic tree construction. Using BLASTP, the 23 amino acid sequence of the ScSDRs was used as a query sequence to identify homologous genes (e < 1 × 10^–5). The genome set of predicted proteins was selected by HMMER3² using three Pfam HMMs: PF00106, PF01370, and PF01073. The NCBI CDD³ and SMART⁴ were used to analyze ApSDRs protein sequences in order to predict protein structures, remove redundant sequences, hand-select candidate SDR-encoding genes and perform functional annotation.

The protein sequences of 180 ApSDRs, 23 ScSDRs, 21 FoSDRs, 6 MoSDRs, and 36 RnSDRs were analyzed in a multi-sequence comparison by ClustalW algorithm (Larkin et al., 2007). Using MEGA 11.0⁵, a phylogenetic tree with 1,000 bootstrap repetitions was constructed using the neighbor-joining method.

Based on the structural characteristics of the sequences, the SDR superfamily is usually classified into seven categories: Classical, Extended, Intermediate, Complex, Divergent, Unassigned, and Atypical (Labesse et al., 1994; Persson and Kallberg, 2013). Nowadays, SDRs segmentation and nomenclature are frequently based on Hidden Markov Models (HMMs), with each family having a name SDR followed by a number and a letter to indicate the type (Persson et al., 2009). All ApSDRs were categorized and named according to the HMMs and phylogenetic tree.

Conserved motifs were analyzed using MEME⁶ (Bailey et al., 2009). The maximum number of motifs was set to 10, and the length of motifs was set to 10∼50 amino acids. NCBI CDD was used to predict the conserved structural domains of proteins (see text footnote 3). The exon-intron organization was predicted by comparing the coding sequence with the corresponding DNA sequence using Gene Structure Display Server 2.0 (GSDS 2.0).⁷ The results were expressed using TBtools software (Chen et al., 2020).

The chromosomal location of ApSDRs was analyzed using genome annotation files. MG2C was used to map the ApSDRs’ precise location on the chromosome.⁸

Homologous gene pairs of the SDR gene family in A. phaeospermum, F. oxysporum, Apiospora arundinis, and Arthrinium hydei were identified using OrthoMCL software,⁹ with parameters set to allow a minimum protein length of 10 and a maximum percentage of stop codons of 20%. Utilizing the default parameters, gene pair covariance was determined using MCScanX software.¹⁰

Based on the whole genome sequencing and annotation results of A. phaeospermum, the genes of ApSDRs were used as foreground genes, and all the genes of A. phaeospermum were used as background genes by using the OmicShare cloud platform.¹¹ GO and KEGG analysis and mapping of SDR gene family in A. phaeospermum.

The protein interaction network database STRING¹² was used to predict protein interactions within gene families for all SDRs in A. phaeospermum to obtain protein interaction networks. The results were exported to Cytoscape software for further annotation and mapping.

Signal peptide was predicted using SignalP-6.0.¹³ Transmembrane structure was predicted using TMHMM-2.0.¹⁴ Protein hydropathicity/hydrophobicity was mapped using ProtScale.¹⁵ Protein tertiary structure was predicted using SWISS-MODEL.¹⁶

Transcriptome sequencing data of A. phaeospermum infesting B. pervariabilis × D. grandis at 0, 7, 14, and 21 days were used to collect the expression data for all ApSDRs (Fang et al., 2022) (the original data were published in the NCBI Sequence Read Archive (SRA) under the accession number SAMN19312317). After selecting the top 30 ApSDRs with the highest expression, we analyzed their data at different times of pathogen infestation in comparison with the control (0 days). Then, we used Heml software after taking the log2 value, analyzed the differences in the expression levels of the SDR family genes during the infestation of A. phaeospermum, and screened out the possible key pathogenicity factors that involved in the invasion of the hybrid bamboos.

A PDA medium coated with 200 μL of sterile hybrid bamboo tissue fluid (T1) and 200 μL of sterile deionized water (CK) was used to inoculate A. phaeospermum. The mycelium was harvested after the strains were incubated for 7, 14, and 21 days at 25°C (Fang et al., 2021). Three duplicates of each sample were made.

A. phaeospermum was treated with 2 mg/mL Congo red, 2 mol/L sodium chloride (NaCl) and 40 mmol/mL hydrogen peroxide (H₂O₂), respectively, and an equal volume of water was used as the control. The samples were taken at 25°C for 10 days. Each sample was replicated 3 times.

Using the Tubulin of A. phaeospermum as the internal reference gene, primers for six candidate genes and Tubulin gene were designed using Primer Premier 5.0 software (Supplementary Table S1). The 2^–ΔΔCt method was used to evaluate the data (Livak and Schmittgen, 2001).

When the accumulation of geneticin is 100 μg/mL or hygromycin reaches 350 μg/mL, A. phaeospermu’s mycelium stops growing completely (Fang et al., 2021). Therefore, hph and KanMx genes were selected as screening markers for knockout and backfill. The strategy employed in this study for the knockout of the SDR gene is shown in Supplementary Figure S1. The homology arms of ApSDR53C2 and ApSDR548U2 were designed with a length of 1,500 bp. For knockout backfill, the ApSDR53C2 and ApSDR548U2 genes were linked to the kanMx gene by complementary base pairing and fusion PCR. Then, the fusion fragments were ligated to the upstream and downstream homology arms of the target genes. Primers were displayed in Supplementary Table S2.

Enzymatic digestion: A. phaeospermum was inoculated on PDA medium and incubated at 25°C for 3 days. 5 agar blocks were scratched from the edge of the colony inoculated in PDB at 180 rpm and incubated at 25°C for 3 days. Then filtered through a single layer of mirocloth to collect the hyphae, and rinsed with 1.2 M KCl 5 times to obtain clean hyphae for spare. Another 0.1 g of lysing enzymes and 0.5 g of driselease enzymes were in a small beaker, added 20 mL of external 1.2 M KCl, added magnetic beads stirred for 10–15 min, transferred to a 50 mL centrifuge tube, rotated horizontally at 4°C, 3,500 rpm, up and down acceleration of 9, centrifuged for 10 min. Pouring back the supernatant into a small beaker, and filtered through a bacterial filter. The enzyme solution was sterilized and set aside. Transfer the mycelium into the enzyme solution, seal it, and put it into the shaker at 30°C, 70 rpm for 8 h. Microscopic examination: take out the cultured protoplasts, aspirate a drop on the slide, and examine it microscopically. A double layer of mirocloth was used to filter it. 25 mL of 1.2 M KCl stabilized osmotic solution rinsed to collect the filtrate, and centrifuged for 10 min at 4°C, 3,000 rpm, with an up and down acceleration of 5. Suction beating was used to mix the protoplasts, and the supernatant was disposed of. To clean the tube walls, 15 mL of KCl stabilizing osmotic solution was then added. After centrifuging the tubes once more for 10 min, 1 mL of STC buffer was added. Microscopic examination, according to the concentration of protoplasts will be divided into 2–5 centrifuge tube spares.

In order to investigate the differences in the sensitivity of the wild-type (WT) strain and the knockout and backfill transformant strains to cell wall stress, hyperosmotic stress, endoplasmic reticulum stress, and extracellular oxidative stress, the WT strain and the knockout and backfill transformant strains were inoculated with PDA medium containing 2 mg/mL of Congo red, 0.01% sodium dodecyl sulfate (SDS), 2 mol/L NaCl, 5 mmol/L dithiothreitol (DTT), and 40 mmol/L H₂O₂ on PDA medium.

To investigate the differences in the responses of the WT strain and the transformant strains to metal ions, different concentrations of MgSO₄, MnSO₄, and ZnSO₄ were added to the PDA medium, and the concentrations of MgSO₄ were set as follows: 0, 1, 2, 3, and 4 mM; the concentrations of MnSO₄ were set as follows: 0, 5, 10, 15, and 20 mM; the concentrations of ZnSO₄ were set to 0, 5, 10, 15, and 20 mM.

To compare the differences between the WT strains and the transformed strains in terms of resistance to different fungicides, dimethyl sulfoxide was used as a solvent to solubilize the agents and added into the PDA medium to adjust the medium to different concentrations for the determination of the resistance of the strains. Tricyclazole and phthalide were added to the medium to study the differences in the response of WT and transformed strains to osmotic-resistant fungicides, and the concentrations were set as follows: 0, 25, 50, 100, 200 μg/mL and 0, 12.5, 25, 50, 100 μg/mL. Azoxystrobin was added to the medium to study the differences in the response of WT and transformed strains to Qi inhibitors, and the concentrations of azoxystrobin were set as follows. 0, 12.5, 25, 50, 100 μg/mL.

The above experimental groups were incubated at 25°C for 5 days and then observed and photographed, and the colony diameters were measured and analyzed. To measure particular differences between pairs of means, all assays were conducted three times, and all data were analyzed using SPSS 16.0. Statistical significance was defined as a p-value of less than 0.05.

Forty uniformly growing 2-year-old B. pervariabilis × D. grandis plants were chosen. Five plants were chosen at random, and the WT strain’s mycelial suspension was sprayed on each plant’s upper shoots. The remainder were incubated with the mycelial suspension of mutants ΔApSDR53C2, ΔApSDR548U2, ΔApSDR53C2: ApSDR548U2, ΔApSDR53C2+, ΔApSDR548U2+, ΔApSDR53C2: ApSDR548U2+, and sterile water. The inoculated branches were bagged moisturized and sprayed every 12 h. After 25 days, the incidence was examined in order to determine the disease index.

The disease grading standard is shown in Supplementary Table S3. Disease index = (Σ (disease grade × number of diseased branches)/(total number of branches × highest disease grade) × 100.

The WT strains and transformant strains were inoculated onto PDA mediums covered with sterile cellulose membrane, and were incubated at 25°C with alternating light and darkness for 4, 7, and 14 days, and then the cellulose membrane was removed and incubated for 4, 7, and 14 days, and the phenotypic changes were observed. Using a stereomicroscope (Nikon P-PS32), the colonies on the 14 days of cellulose membrane induced culture were observed, and the magnification of 2x and 5x objective lenses were selected. Photographs were taken to record the observations. After cellulose membrane induction for four, seven, and fourteen days, mycelium samples were prepared. RNA was extracted, and cDNA was obtained by reverse transcription. Primers of melanin synthesis-related gene and internal reference gene (Tubulin) for qPCR were displayed in Supplementary Table S4. cDNA was utilized as the template. The 2^{–Δ Δ Ct} method was used to evaluate the data.

PDA mediums with tricyclazole concentrations of 0, 12.5, 25, and 50 μg/mL were prepared, respectively, and inoculated with WT and transformed strains after covering with sterile cellulose membrane. Then inverted at 25°C for 7 days. The expression levels of melanin-related genes were analyzed and recorded under the different conditions of treatment.

Using blast and HMMER3, 213 ApSDRs were found in A. phaeospermum. The structural domains of the genes that encoded these ApSDRs were then annotated, and the genes lacking the adh_short domain were eliminated, leaving 180 ApSDRs.

According to the Hidden Markov Model and phylogenetic tree (Figure 1), all SDR genes in A. phaeospermum were classified, renamed, and categorized into three major groups (Classical, Extended, and Unassigned) and eight subgroups, among which the Classical type has the most ApSDRs with 141, the Extended type has 2, and Unassigned type has 37. In addition, ApSDRs in the classic type were further categorized into 5 subgroups (clusters C1, 2, 3, 4, and 5) containing 10, 15, 24, 29, and 63 members, respectively. Two subgroups (clusters U1 and U2) with 8 and 29 members were further divided from ApSDRs in the unassigned phenotype.

We successfully annotated 180 ApSDRs into GO categories (Figure 2A). These ApSDRs were enriched in “metabolic process” (GO:0008152), “single-organism process” (GO:0044699), and “catalytic activity” (GO:0003824). As shown in Figure 2B, analysis of 180 ApSDRs KEGG-enriched pathways revealed that ApSDRs were mainly enriched in KEGG-enriched pathways such as global and overview maps, lipid metabolism, carbohydrate metabolism, and metabolism of cofactors and vitamins.

As shown in Supplementary Figure S2, ten conserved motifs ranging in size from 11 to 35 amino acid residues, numbered 1 through 10, were obtained and displayed in colored boxes.

Among the conserved motifs of ApSDR proteins, Motif 1, Motif 3, and Motif 4 encode conserved structural domains in the SDR gene family: the NAD (P)-binding Rossmann-fold. Motif 1 or Motif 3 can be observed in all ApSDRs, suggesting that all of them belong to the short-chain dehydrogenase/reductase (SDR) superfamily. Other conserved motifs (Motif 2, Motif 4–10) are also present in ApSDR proteins, ranging in number from 1 to 9.

Analysis of the conserved structural domains of ApSDR members showed that all members have NADB_Rossmann superfamily structural domains. Analysis of the exon-intron structures of 180 ApSDRs showed that the number of exons and introns in the same group was relatively conserved. The ApSDRs belonging to the classical SDR fold, on the other hand, had significantly more exons and introns than those belonging to the extended SDR fold, and the unassigned ApSDRs, suggesting that the functions of ApSDRs belonging to the classical SDR fold may be different from those of the other two groups.

It was discovered that, despite the low levels of amino acid sequence similarity, the ApSDRs were still highly conserved in the important amino acid sites. Nearly all of them contained the conserved motif TGxxxGxG at the N-terminal, which defines the cofactor binding site, and is a relatively conserved glycine-rich region (Oppermann et al., 2003). In this motif, amino acids 1, 2, 6, and 8 were also found to be highly conserved, as valine (T) and glycine (G), respectively (Supplementary Figure S3).

These 180 ApSDRs, widely dispersed on 18 of the 19 A. phaeospermum chromosomes, ranged from 1 to 22 ApSDR genes per chromosome, as Figure 3A illustrates. With 22 ApSDRs, chromosome Chr99 has the greatest distribution of ApSDRs. There were no ApSDR found on Chr354.

The orthologous relationship of SDR genes between A. phaeospermum and F. oxysporum, A. arundinis, or A. hydei was examined (Figure 3B). The results revealed that there were 9 pairs of homologous genes identified between A. phaeospermum and F. oxysporum, 123 pairs between A. phaeospermum and A. arundinis, and 141 pairs between A. phaeospermum and A. hydei. There is low synteny and fewer homologous SDR in F. oxysporum species because they are in different genera and families, whereas there is high synteny in A. phaeospermum and A. hydei species because they belong to the same genus (Li et al., 2019). The degree to which homologous genes persist on corresponding chromosomes (synteny) and in conserved orders (collinearity) throughout evolution can be determined by comparing the genomes of related eukaryotic organisms. The more homologous gene pairs a gene has between species, the more important the gene is likely to be in the phylogeny of that gene family (Coghlan et al., 2005). Eight ApSDRs (ApSDR566C1, ApSDR86C5, ApSDR124C1, ApSDR49C2, ApSDR76C1, ApSDR126C1, ApSDR51C1, and ApSDR53C4) are related to at least three homologous gene pairs, indicating that the phylogeny of the SDR gene family may be significantly influenced by these genes.

We built a PPI network of the SDR gene family in A. phaeospermum with 29 nodes and 172 edges (Figure 4A). The larger diameter of the protein node icon represents the larger degree, and the thicker interaction lines between the two proteins represent the higher interaction score. Proteins ApSDR53C2 and ApSDR548U2 were identified as the two interaction centers of the two interaction networks.

The cytoHubba plugin of Cytoscape was used to identify the top 10 hub genes of network node degree in the PPI network in accordance with the Degree method (Figure 4B). The degree of a protein node represents the number of proteins interacting with it. The higher the degree, the higher the importance of the protein in the organism, which is essential to maintain its life activities, and also essential to maintain the morphology of the whole network. The analysis yielded five proteins in the constructed network with a degree higher than 10. The highest one was ApSDR548U2 with a degree of 13, followed by the protein ApSDR53C2 with a degree of 12. Among the top 10 hub genes with network node degree, seven genes belonged to the Classical class, two genes belonged to the Unassigned class and one belonged to the Extended class.

The expression data of all ApSDRs were collected from the transcriptome sequencing data of A. phaeospermum infesting B. pervariabilis × D. grandis at 0 (S1 period: uninfested phase), 7 (S2 period: contact phase), 14 (S3 period: penetration phase), and 21days (S4 period: incubation phase). The previous study showed that A. phaeospermum infested hybrid bamboos in the contact and penetration phase (S2 and S3), which caused only PAMP-triggered immunity (PTI) response in hybrid bamboos, while the effector-triggered immunity (ETI) response of hybrid bamboos was activated in the incubation phase (S4) (Fang et al., 2022). Considering the traits of hybrid bamboos, like their thick cuticle and high fiber content in their branches, it was hypothesized that the genes which were significantly up-regulated during S2 and S3 infestation and down-regulated during S4 may be the key pathogenicity factors for the invasion of hybrid bamboos. As shown in Figure 5A, among the top 30 ApSDRs in expression, ApSDR53C2, ApSDR58C1, ApSDR86C6, ApSDR548U2, ApSDR472C1, ApSDR118C3 were significantly up-regulated during S2 and S3 infestation and down-regulated during S4.

Both transmembrane proteins and secreted proteins have signal peptides. In addition, transmembrane proteins should have transmembrane structures and hydrophobic regions. As shown in Supplementary Figure S4A, in the signal peptide analysis by signalP, only proteins ApSDR53C2 and ApSDR58C1 were present with the signal peptides. In the transmembrane region analysis by TMHMM, proteins ApSDR53C2, ApSDR58C1, and ApSDR86C6 were present in the transmembrane regions (Supplementary Figure S4B). The hydrophobicity of amino acids reflects the folding of proteins, with hydrophobic regions occurring in potential transmembrane regions and playing an important role in maintaining the tertiary structure of proteins, which includes biological membrane structure. Hydropathicity/hydrophobicity mapping of proteins can provide a reference for the identification of transmembrane regions of proteins. In general, the transmembrane region of a protein is hydrophobic, while the portion outside the membrane is hydrophilic. Hydrophobic proteins have a disproportionately high number of hydrophobic amino acid residues throughout the polypeptide chain compared to hydrophilic ones, and hydrophilic proteins have the opposite pattern. As shown in Supplementary Figure S4C, the protein hydropathicity/hydrophobicity maps of six candidate ApSDRs were plotted by ProtScale. The other four proteins were hydrophilic proteins, while proteins ApSDR53C2 and ApSDR58C1 were hydrophobic proteins, and their predicted results of the hydrophobic region matched the predicted results of the transmembrane region. It is hypothesized that proteins ApSDR53C2 and ApSDR58C1 are transmembrane proteins.

Six putative ApSDRs’ protein tertiary structures were estimated using SWISS-MODEL (Supplementary Figure S4D). According to the prediction results, all six proteins have the conserved ‘Rossmann-fold’ three-dimensional structure of the SDR family, which includes the catalytic and cofactor-binding sites (Kavanagh et al., 2008). The “Rossmann-fold” is composed of three α-helices on each side and a center β-fold with 6–7 strands.

As can be shown in Figures 5B–G, after inoculating A. phaeospermum into PDA mediums containing hybrid bamboo tissue fluid, the expression of all six genes dramatically dropped at day 21, and it exceeded that of the control at days 7 and 14. Among these, ApSDR58C1 and ApSDR472C1 gene expression was considerably higher on day 7 compared to the other two periods and exhibited a declining trend from days 7 to 21. Between days 7 and 14, the expression of the genes ApSDR53C2, ApSDR548U2, ApSDR86C6, and ApSDR118C3 increased at day 14 and then decreased at day 21. The expression of ApSDR53C2 and ApSDR548U2 was substantially higher than that of the other genes. This appears to suggest that the six SDR genes’ expression rose when A. phaeospermum was infested with B. pervariabilis × D. grandis. These genes may be crucial to the infestation process, with ApSDR53C2 and ApSDR548U2 having a more substantial impact. The transcriptome sequencing of A. phaeospermum infesting B. pervariabilis × D. grandis revealed the same trends in the expression of these six genes.

As shown in Figures 5H–M, in the PDA medium with the addition of 2 mol/L NaCl, the expression levels of the other five genes were considerably greater than that in the unstressed control, except for gene ApSDR472C1. With the exception of ApSDR86C6, the expression levels of the other five genes were considerably higher under the condition of 2 mg/mL Congo red than in the unstressed control. Under the treatment of 40 mmol/mL H₂O₂, only ApSDR53C2 and ApSDR548U2 genes were appreciably highly expressed compared to those in normal PDA medium, while the other four genes showed no significant changes. Nearly all of these six genes’ expression levels increased significantly under different stressors, particularly in Congo red stress and NaCl stress.

Out of the six genes, ApSDR53C2 and ApSDR548U2 exhibited the greatest variation in expression from the control under various stress scenarios, potentially indicating a crucial function in the pathogenicity exhibited by the pathogens.

The knockout vectors for ApSDR53C2 and ApSDR548U2 were constructed, which were successfully assayed by an improved isolation marker method (Supplementary Figure S5). A large number of transparent rounded protoplasts of A. phaeospermum were obtained after centrifugal washing (Supplementary Figure S6). Recombinant vectors were transformed by protoplasts separately to 350 μg/mL of hygromycin in TB3. As shown in Supplementary Figure S7A, colonies grown under 350 μg/mL of hygromycin were inoculated onto PDA supplemented with 400 μg/mL of hygromycin, and the colonies grown under 400 μg/mL of hygromycin were positive knock-out transformants. The method of obtaining backfill strains is the same as above. The positive complemented transformants were shown in Supplementary Figure S7B. Then, “hph,” “ApSDR53C2,” “ApSDR548U2,” “ApSDR53C2-5-hph,” “ApSDR53C2-3-hph,” “ApSDR548U2-5-hph,” “kanMx” fragments of positive transformant strains were identified by PCR, as Supplementary Figure S8 illustrates. Knock-out mutants ΔApSDR53C2, ΔApSDR548U2, and co-knockout mutant ΔApSDR53C2:ApSDR548U2, were effectively obtained for both genes and back-complemented with three deletion mutants.

As shown in Figures 6A,B, compared to the WT and backfilled strain, the deletion mutants’ phenotypes and relative inhibition rate were noticeably different. For the cell wall stress sensitivity assay, the relative inhibition rate of ΔApSDR53C2, ΔApSDR548U2, and ΔApSDR53C2:ApSDR548U2 colonies after the addition of 2 mg/mL Congo red was 71.9, 70.3, and 75.9%, respectively, and all three mutant strains were more particularly responsive to Congo red stress than the WT strain (50.4%). The relative inhibition of ΔApSDR53C2, ΔApSDR548U2, and ΔApSDR53C2:ApSDR548U2 colonies after the addition of 0.01% SDS was 53.9, 52.7, and 55.7%, respectively, and all three mutant strains were more particularly responsive to SDS stress than the WT strain (40.0%). Comparing the difference in sensitivity to osmotic stress between the transformant strains and the WT strain, the relative inhibition of ΔApSDR53C2, ΔApSDR548U2, and ΔApSDR53C2:ApSDR548U2 colonies was considerably greater than WT strain (49.0%) under 2 mol/L NaCl treatment, and the relative growth inhibition of the three deletion mutants was 76.0, 75.4, and 81.3%, respectively. In addition, in the oxidative stress sensitivity assay, the three mutants showed significantly lower tolerance to H₂O₂. The growth area of ΔApSDR53C2, ΔApSDR548U2, and ΔApSDR53C2:ApSDR548U2 colonies was significantly reduced by 56.2, 59.3, and 65.8%, respectively, at a concentration of 40 mmol/L H₂O₂. Nevertheless, none of the three knockout mutants’ colony growth was profoundly impacted by the 5 mmol/L DTT stress. The phenotypes of the WT and complemented strains did not differ significantly. These results suggest that ApSDR53C2 and ApSDR548U2 have important roles in cell wall inhibitor stress response, osmotic stress response, and oxidative stress response of A. phaeospermum, without participating in the endoplasmic reticulum stress of its cells themselves.

Comparing the heavy metal stress of the respective knockout transformant strains of the ApSDR53C2 and ApSDR548U2 genes, their co-knockout transformant strain, and the backfill strains with the WT strains (Figures 6C–E), the relative inhibition of ΔApSDR53C2, ΔApSDR548U2, and ΔApSDR53C2:ApSDR548U2 colonies on PDA mediums with added Mg²⁺ and Mn²⁺ was significantly increased compared to that of the WT strain, and the knockout transformants showed a greater degree of increase in the relative inhibition with the addition of Mg²⁺. The differences in the relative inhibition of ΔApSDR53C2, ΔApSDR548U2, and ΔApSDR53C2:ApSDR548U2 versus the WT strain were not significant with Zn²⁺ added. Under the three heavy metal stresses, the relative inhibition of the backfilled strains and the WT strain did not differ significantly. It suggests that the deletion of these two genes resulted in a decrease in resistance to Mg²⁺ and Mn²⁺ in A. phaeospermum, and thus Mg²⁺ and Mn²⁺ may promote the enzymatic activities of the proteins encoded by ApSDR53C2 and ApSDR548U2.

Comparing the fungicide stress of the transformant strains with the WT strains (Figures 6F–H), ΔApSDR53C2, ΔApSDR548U2, and ΔApSDR53C2:ApSDR548U2 strains showed a considerable decrease in their fungicide stresses under the anti-osmotic fungicide tricyclazole and phthalide, the relative inhibition rate was considerably decreased compared to the WT strain, where the difference in relative inhibition rate was more significant under the stress of tricyclazole. Under azoxystrobin treatment, ΔApSDR53C2 and ΔApSDR53C2:ApSDR548U2 strains showed a significant difference in relative inhibition rate compared to the WT strain, but the ΔApSDR548U2 strain did not. Under the three fungicide treatments, the relative inhibition of the backfilled strains and the WT strain did not differ significantly.

Mycelial suspension from WT and the corresponding mutants were prepared and inoculated on the shoots and leaves of B. pervariabilis × D. grandis separately to perform a pathogenicity test. The WT strain was shown to generate much more severe symptoms than the ΔApSDR53C2, ΔApSDR548U2, and ΔApSDR53C2:ApSDR548U2 mutants (Figure 7A).

Figure 7B displays the quantitative measurements of the disease indices of hybrid bamboo inoculated with various strains. The disease indices of hybrid bamboos rose with increasing inoculation time, regardless of whether they were inoculated with the WT, ΔApSDR53C2, ΔApSDR548U2, or ΔApSDR53C2:ApSDR548U2 mutant. After 25 days, the disease indices of them reached 83.53, 53, 55, and 50%, while that of hybrid bamboo inoculated with sterile water was 0. We also found that the disease index of hybrid bamboos inoculated with WT at the same time was significantly higher than that of all three deletion mutants, whereas that of the ΔApSDR53C2 mutant was lower than the ΔApSDR548U2 mutant, the ΔApSDR53C2:ApSDR548U2 mutant was lower than that of the ΔApSDR53C2 and ΔApSDR548U2 mutants. There was no discernible difference between the disease indices of hybrid bamboos inoculated with all backfill mutants and the WT. This result suggests that genes ApSDR53C2 and ApSDR548U2 play important roles in A. phaeospermum virulence, and the gene ApSDR53C2 has a stronger virulence role.

As shown in Figure 8A, after cellulose membrane treatment of the strains to be analyzed for 7 days, the WT strains were able to penetrate the cellulose membrane, However, ΔApSDR53C2 was not able to pass through the cellulose membrane, and no obvious hyphae were observed in the PDA mediums after uncovering the membranes. The level of ΔApSDR53C2+ cellulose membrane penetration was close to that of the WT.

After 14 days of cellulose membrane induction, the WT, ΔApSDR53C2+ colonies were obviously blackened, while the ΔApSDR53C2 colonies were lighter in color (Figure 8B), which indicated that ApSDR53C2 has the function of regulating melanin synthesis. The expression of melanin synthesis-related genes was all considerably lower in the ΔApSDR53C2 than in the WT strain, whereas the expression in the gene-function backfilling strains was close to that of the WT (Figure 8C). Prolonging the cellulose membrane induction time, the expression levels of melanin synthesis-related genes in the ΔApSDR53C2 still could not be induced to rise to the WT level. The aforementioned findings suggest that ApSDR53C2 influences the level of melanin synthesis during the penetration of cellulose membranes, and the process of cellulose membrane penetration induces the production of melanin in A. phaeospermum, suggesting that ApSDR53C2 promotes mycelial penetration by influencing melanin synthesis.

Two ApSDR53C2 complemented mutants with widely varying ApSDR53C2 gene expression were chosen, with the aim of exploring the changes in melanin synthesis-related genes’ expression level with different ApSDR53C2 gene expression and further demonstrating the correlation between the ApSDR53C2 gene and melanin biosynthesis. The ApSDR53C2 expression in ΔApSDR53C2 complemented mutants was analyzed under cellulose membrane-induced conditions. The results of qPCR demonstrated that ApSDR53C2 expression levels in ΔApSDR53C2+ 1 and ΔApSDR53C2+ 2 were 1.30-fold and 5.67-fold higher than those in the WT (Figure 8D). Detecting the melanin synthesis-related genes’ expression (Figure 8E), the expression levels in ΔApSDR53C2+ 2 were significantly higher than those of ΔApSDR53C2+ 1. It indicated that the ApSDR53C2 gene level was positively correlated with melanin synthesis-related genes, and positively promoted the biosynthesis of melanin in A. phaeospermum.

The effect of melanin on the penetration of cellulose membranes by A. phaeospermum was analyzed using the melanin synthesis inhibitor tricyclazole (Figure 9A). The findings demonstrated that tricyclazole could effectively prevent A. phaeospermum from penetrating the cellulose membrane, and that when the concentration of tricyclazole increased, the expression of genes linked to melanin formation in the treated group declined dramatically (Figure 9B). This suggests that melanin is crucial for A. phaeospermum’s ability to penetrate the cellulose membrane.