Identification of Mycoparasitism-Related Genes against the Phytopathogen Botrytis cinerea via Transcriptome Analysis of Trichoderma harzianum T4

Trichoderma harzianum is a well-known biological control agent (BCA) that is effective against a variety of plant pathogens. In previous studies, we found that T. harzianum T4 could effectively control the gray mold in tomatoes caused by Botrytis cinerea. However, the research on its biocontrol mechanism is not comprehensive, particularly regarding the mechanism of mycoparasitism. In this study, in order to further investigate the mycoparasitism mechanism of T. harzianum T4, transcriptomic sequencing and real-time fluorescence quantitative PCR (RT-qPCR) were used to identify the differentially expressed genes (DEGs) of T. harzianum T4 at 12, 24, 48 and 72 h of growth in the cell wall of B. cinerea (BCCW) or a sucrose medium. A total of 2871 DEGs and 2148 novel genes were detected using transcriptome sequencing. Through GO and KEGG enrichment analysis, we identified genes associated with mycoparasitism at specific time periods, such as encoding kinases, signal transduction proteins, carbohydrate active enzymes, hydrolytic enzymes, transporters, antioxidant enzymes, secondary metabolite synthesis, resistance proteins, detoxification genes and genes associated with extended hyphal longevity. To validate the transcriptome data, RT-qCPR was performed on the transcriptome samples. The RT-qPCR results show that the expression trend of the genes was consistent with the RNA-Seq data. In order to validate the screened genes associated with mycoparasitism, we performed a dual-culture antagonism test on T. harzianum and B. cinerea. The results of the dual-culture RT-qPCR showed that 15 of the 24 genes were upregulated during and after contact between T. harzianum T4 and B. cinerea (the same as BCCW), which further confirmed that these genes were involved in the mycoparasitism of T. harzianum T4. In conclusion, the transcriptome data provided in this study will not only improve the annotation information of gene models in T. harzianum T4 genome, but also provide important transcriptome information regarding the process of mycoparasitism at specific time periods, which can help us to further understand the mechanism of mycoparasitism, thus providing a potential molecular target for T. harzianum T4 as a biological control agent.


Introduction
Botrytis cinerea (Ascomycota) is a necrotrophic plant pathogen with a wide range of hosts (>200 species) [1]. Botrytis cinerea can cause serious loss of crops during production and post-harvesting, the value of which can reach up to USD 1 billion per year [2,3], and for this reason, it is listed as the second most destructive pathogenic fungus in the world [4]. Today, chemical prevention is the main strategy used worldwide to control fungal diseases [5]. However, long-term use of chemical pesticides has led to human and animal health hazards, severe pesticide residues, environmental pollution and microbial resistance [6,7]. Therefore, it is essential to develop novel control methods that are environmentally friendly.
Biocontrol agents are an environmentally friendly alternative to chemical pesticides and have numerous advantages over the use of fungicides, such as cost-effectiveness, safety and environmental protection [8,9]. Strains of Trichoderma are excellent biological control agents and represent some of the most widely used fungi, playing a vital role in agriculture [10,11].
In a previous study, our research team isolated and purified from soil a biocontrol strain of T. harzianum T4 with a high application potential. This strain can effectively control the plant disease caused by B. cinerea [12], promote plant growth and increase crop yield [13,14]. In our previous study, we optimized the culture conditions for T. harzianum T4 and improved the yield of the T. harzianum biocontrol agent [15]. However, the research on its biocontrol mechanism is not comprehensive, particularly the mechanism of mycoparasitism. It has been shown that the mycoparasitism process mainly includes decomposition of the host cell wall by the secretion of hydrolytic enzymes; these processes are regulated by signal transduction proteins (heterotrimeric G protein and mitogen-activated protein kinase, MAPK) [16][17][18]. However, the mechanism of mycoparasitism is affected by the types of pathogen and Trichoderma isolates [19,20]. Therefore, it is necessary to study different strains to increase our understanding of the mycoparasitism of T. harzianum.
In previous studies, high-pressure induction of inactivated hyphae was used to simulate the presence of a fungal host in order to study the genes associated with fungal mycoparasitism [19,[21][22][23]. In this study, using RNA-Seq, we identified the DEGs in the T. harzianum T4 strain after 12, 24, 48 and 72 h of culturing in a high-pressure-induced liquid medium containing inactivated B. cinerea cell wall (BCCW), and used the Czapek-Dox broth (CDB) medium inoculated only with T. harzianum as a control. This study helps to determine a large number of the DEGs involved in specific stage functions at different time periods and enhances understanding of the mycoparasitism of T. harzianum T4. We identified a large number of genes that are strongly associated with mycoparasitism. These genes are significantly upregulated in a specific time period and have the functions of coding for signal transduction (G protein and MAPK kinase), carbohydrate-active enzymes (especially β-glucosidase), transporters, antioxidant proteins, resistance to stress, detoxification, delaying hyphal senescence and toxic secondary metabolites. Finally, quantitative RT-PCR was used to verify the differential expression of candidate genes for T. harzianum T4 antagonism against B. cinerea via mycoparasitism in both transcriptome and dual culture.

Source of Fungal Strains
The wild-type T. harzianum T4 strain was stored in our laboratory (East China University of Science and Technology, Shanghai, China). Botrytis cinerea was obtained from the China General Microbial Species Preservation and Management Center (CGMCC, Beijing, China).

Culture Conditions
Trichoderma harzianum T4 was inoculated on a potato dextrose agar (PDA) plate and cultured for 3-5 days in a constant-temperature incubator at 28 • C. When the conidia had spread all over the plate, the conidia collected from the culture were washed repeatedly with 5 mL of sterile water and the spore suspension was diluted to 1 × 10 7 for later use. Next, 1 mL of 1 × 10 7 T. harzianum T4 conidia suspension was added to a fresh potato dextrose broth (PDB) liquid culture medium and the conidia were germinated in a rotating shaker at 28 • C for 12 h at 180 rpm. The grown mycelium was centrifuged at 4000 rpm for 10 min, collected and washed three times with sterile distilled water for subsequent induction experiments. The grown mycelium was transferred to the Czapek-Dox broth (CDB) induction medium containing 5 g sucrose, 3 g NaNO 3 , 1 g K 2 HPO 4 , 0.5 g MgSO 4 , 0.5 g KCl, 10 mg FeSO 4 and supplemented with 1% inactivated cell wall of B. cinerea (previously autoclaved at 120 • C for 20 min) (BCCW). The CDB induction medium inoculated only with T. harzianum-grown mycelium without BCCW was used as the control. All cultures were incubated at 180 rpm at 28 • C for 12, 24, 48 and 72 h and the mycelium was collected. Samples for each incubation period were collected three times, including all treatments. The mycelium was immediately frozen using liquid nitrogen for 20-30 min and stored at −80 • C until the RNA was extracted. All collected RNA samples were used for the following transcriptome sequencing and real-time fluorescence quantitative PCR experiments (RT-qPCR).

Total RNA Extraction, cDNA Synthesis and Sequencing
Total RNA was extracted from 24 samples using the TransZol up Plus RNA kit (Trans-Gen Biotech, Beijing, China) according to the manufacturer's instructions. RNA was quantified and verified for integrity using the Nano assay in an Agilent BioAnalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA). A total of 1 µg RNA per sample was used as input material for RNA sample preparation. Sequencing libraries were generated using the NEBNext UltraTM RNA Library Prep Kit for Illumina (NEB, Ipswich, MA, USA). Briefly, mRNA was purified from total RNA using poly-T oligo-attached magnetic beads. Fragmentation was carried out using divalent cations under elevated temperature in NEB Next First Strand Synthesis Reaction Buffer (5X). First-strand cDNA was synthesized using random hexamer primer and M-MuLV reverse transcriptase. Second-strand cDNA synthesis was subsequently performed using DNA polymerase I and RNase H. Remaining overhangs were converted into blunt ends via exonuclease/polymerase activities. The AMPure XP system (Beckman Coulter, Beverly, CA, USA) was used to purify fragments and amplify them by PCR to prepare cDNA libraries. All cDNA libraries were sequenced on the Illumina platform (Illumina, San Diego, CA, USA) to generate paired-end reads. SOAPnuke v2.1.0 software (BGI, Shenzhen, China) [24] was used to filter and splice the original readings to remove low-quality and unmeasurable bases (denoted by N), so as to obtain clean readings. The genome of T. harzianum T4 (separately cultured without B. cinerea cell wall) was annotated (NCBI: PRJNA715015) and used as a reference genome in this study.

Analysis of DEGs
DESeq2 [25] was used to analyze the significance of differential gene expression in T. harzianum T4 between the treated group and the control group at different growth stages (12, 24, 48 and 72 h). The threshold condition for DEGs was set as p-adjust (q-value) ≤ 0.05 and |Log 2 FC| ≥ 1. GO and KEGG databases were used for the enrichment and annotation analyses of the DEGs. The enrichment analysis method was hypergeometric distribution, with a q-value ≤ 0.05 taken as the threshold screening condition for significant enrichment of the GO term and KEGG pathway.

RT-qPCR
Twelve DEGs were randomly selected for RT-qPCR validation using the Bio-Rad CFX 96 quantitative fluorescence PCR instrument (Bio-Rad, Hercules, CA, USA). Primer5 v5.00 software (Premier Biosoft International) was used to design the verification primers (Table 1). RNA was reverse transcribed into cDNA using a TUREscript 1st Stand cDNA SYNTHESIS Kit. Each 20 µL PCR reaction consisted of 10 µL of 2 × SuperNovaPCRMix (Dye), 0.5 µL of forward primer, 0.5 µL of reverse primer, 1 µL of cDNA template and 8 µL of nuclease-free water. The operating cycle conditions were 95 • C for 4 min (1 cycle), 95 • C for 15 s, then 60 • C for 30 s (39 cycles), and for the melt curve analysis the temperature was increased from 65 • C to 95 • C in increments of 0.5 • C for 5 s (1 cycle). The gene expression level was calculated according to the threshold period (CT) 2 −∆∆CT method and the 18s rRNA and α-tubulin were used as internal references to normalize the amount of total RNA present in each reaction. Dual culture is the most effective way to verify the expression of genes associated with mycoparasitism in a specific culture time period. RT-qPCR was used to analyze the expression of 24 genes possibly related to mycoparasitism of T. harzianum T4 (Table 1). Using a hole punch, a disc with a diameter of 5 mm was cut from the T. harzianum T4 and B. cinerea plates as a dual-culture inoculum. Trichoderma harzianum T4 and B. cinerea were inoculated 8 cm apart on opposite sides of PDA plate medium covered with cellophane. As a control, dual-culture confrontation assays were conducted following the same procedure described above, except that T. harzianum was challenged with itself. The antagonistic plate was cultured in a constant-temperature incubator at 28 • C for 5 days and T. harzianum T4 mycelium samples (3 replicates of each sample) were harvested at 3 growth stages (before contact (BC), during contact (C) and after contact (AC)) of the dual culture and these were then used for RT-qPCR. The experiment was conducted with three repetitions for each sample and results were statistical analyzed by one-way analysis of variance (ANOVA) with Duncan tests conducted by SPSS statistics software (V. 21.0, SPSS Inc., Chicago, IL, USA). All results at p < 0.05 were considered statistically significant.

Data Analysis of the Transcriptome
In this experiment, we used RNA-Seq to screen the differentially expressed genes of T. harzianum T4 at different growth stages (12, 24, 48 and 72 h) in the presence of BCCW. As shown by the Venn diagram in Figure 1A

GO Enrichment Analysis of DEGs
GO enrichment analysis of DEGs showed that at the different growth stages (12, 24, 48 and 72 h) of T. harzianum in the presence of BCCW, GO terms mainly focused on biological processes (metabolism and single-organ processes); cell components (cell membrane, cell membrane components, cells and cell components); and molecular functions (catalytic activity, molecular binding and transporter activity) ( Figure 2). In order to further analyze the genes related to mycoparasitism, GO enrichment analysis was performed on the upregulated DEGs ( Figure 3). The results showed that in the biological process, carbohydrate metabolism was significantly enriched at 12, 24 and 48 h. The oxidationreduction process and single organism metabolic process were significantly enriched at 72 h. In addition, kinase activity, phosphorylation, catalytic activity, hydrolase activity, transporter activity, cofactor binding, transmembrane transport activity, metal ion binding and oxidoreductase activity involved in the molecular process were all significantly enriched in different culture time periods of T. harzianum grown in the presence of BCCW.

GO Enrichment Analysis of DEGs
GO enrichment analysis of DEGs showed that at the different growth stages (12,24,48 and 72 h) of T. harzianum in the presence of BCCW, GO terms mainly focused on biological processes (metabolism and single-organ processes); cell components (cell membrane, cell membrane components, cells and cell components); and molecular functions (catalytic activity, molecular binding and transporter activity) ( Figure 2). In order to further analyze the genes related to mycoparasitism, GO enrichment analysis was performed on the upregulated DEGs ( Figure 3). The results showed that in the biological process, carbohydrate metabolism was significantly enriched at 12, 24 and 48 h. The oxidation-reduction process and single organism metabolic process were significantly enriched at 72 h. In addition, kinase activity, phosphorylation, catalytic activity, hydrolase activity, transporter activity, cofactor binding, transmembrane transport activity, metal ion binding and oxidoreductase activity involved in the molecular process were all significantly enriched in different culture time periods of T. harzianum grown in the presence of BCCW.

KEGG Enrichment Analysis of DEGs
The KEGG enrichment analysis of the DEGs in T. harzianum grown for 12, 24, 48 and 72 h in the presence of BCCW showed that 25 KEGG pathways (q ≤ 0.05) were significantly enriched ( Table 2). These pathways are mainly involved the metabolism of carbohydrates, lipids and amino acids, as well as the biodegradation and metabolism of exogenous substances and the metabolism of terpenoids and polyketides. To further analyze the genes associated with mycoparasitism, a KEGG enrichment analysis was performed on the upregulated DEGs. It showed the top 20 enrichment pathways of T. harzianum when grown for 12, 24, 48 and 72 h; of these pathways, 8 KEGG pathways were significantly enriched (q ≤ 0.05), involving carbohydrate metabolism and amino acid biosynthesis and degradation ( Figure 4). Of these, starch and sucrose metabolism (ko00500) and amino sugar and nucleotide sugar metabolism (ko00520) were enriched at 12 h, 24 h and 48 h. The ordinate represents the GO classification description, and the abscissa rich factor represents the ratio of differentially expressed genes of this term to all genes that were annotated in the term.

KEGG Enrichment Analysis of DEGs
The KEGG enrichment analysis of the DEGs in T. harzianum grown for 12, 24, 48 and 72 h in the presence of BCCW showed that 25 KEGG pathways (q ≤ 0.05) were significantly enriched ( Table 2). These pathways are mainly involved the metabolism of carbohydrates, lipids and amino acids, as well as the biodegradation and metabolism of exogenous substances and the metabolism of terpenoids and polyketides. To further analyze the genes associated with mycoparasitism, a KEGG enrichment analysis was performed on the upregulated DEGs. It showed the top 20 enrichment pathways of T. harzianum when grown for 12, 24, 48 and 72 h; of these pathways, 8 KEGG pathways were significantly enriched (q ≤ 0.05), involving carbohydrate metabolism and amino acid biosynthesis and degradation ( Figure 4). Of these, starch and sucrose metabolism (ko00500) and amino sugar and nucleotide sugar metabolism (ko00520) were enriched at 12 h, 24 h and 48 h.  The ordinate represents the GO classification description, and the abscissa rich factor represents the ratio of differentially expressed genes of this term to all genes that were annotated in the term.   In this experiment, we found that kinase activity-related and signal transducer activityrelated genes were significantly upregulated at 12 h ( Figure 5A). This mainly involves genes related to serine/threonine kinase, G protein signaling and MAPK kinase activity. We focused on 10 genes, including 7 kinase activity genes (log 2 FC > 2), of which 4 were kinase activity genes involved in glycolysis (scaffold 14.g95, scaffold 39.g6, scaffold 13.g198, scaffold 9.g267), and 3 were involved in protein serine/threonine kinase activities (scaffold 17. g109, scaffold 4.g70, novel.6707). The other three genes were related to the signal transduction activity: the guanine nucleotide-binding protein gamma subunit 1 gng-1 (scaffold 6.g186) and the guanine nucleus-binding protein alpha-2 subunit gna-2 (scaffold 6.g263) positively regulated the G protein, and the phosphate intermediate protein in YPD1 (scaffold 28.g81) regulated the MAPK signaling pathway.

Kinase Activity and Signal Transducer Activity
In this experiment, we found that kinase activity-related and signal transducer activity-related genes were significantly upregulated at 12 h ( Figure 5A). This mainly involves genes related to serine/threonine kinase, G protein signaling and MAPK kinase activity. We focused on 10 genes, including 7 kinase activity genes (log2FC > 2), of which 4 were kinase activity genes involved in glycolysis (scaffold 14.g95, scaffold 39.g6, scaffold 13.g198, scaffold 9.g267), and 3 were involved in protein serine/threonine kinase activities (scaffold 17. g109, scaffold 4.g70, novel.6707). The other three genes were related to the signal transduction activity: the guanine nucleotide-binding protein gamma subunit 1 gng-1 (scaffold 6.g186) and the guanine nucleus-binding protein alpha-2 subunit gna-2 (scaffold 6.g263) positively regulated the G protein, and the phosphate intermediate protein in YPD1 (scaffold 28.g81) regulated the MAPK signaling pathway.  Figure 6A). There were 140 CAZyme genes with significant differential expression ( Figure 6A). Large numbers of CAZyme-related genes were significantly upregulated at 12, 24 and 48 h, but not at 72 h. At 12 h, there were 62 differentially expressed genes, 42 of which were significantly upregulated. At 24 and 48 h, 24 and 26 CAZyme genes were upregulated, respectively. In this experiment, a total of 82 GH glycoside hydrolase differential genes, 23 coenzyme activity genes, 17 glycosyltransferase and carbohydrate esterase genes and 1 carbohydrate-binding module gene were significantly differentially expressed ( Figure 6B).

Carbohydrate Active Enzymes (CAZymes)
Based on KEGG and GO enrichment analysis, carbohydrate metabolism was significantly enriched at 12, 24 and 48 h. A large number of carbohydrate active enzymes, in particular, GO trem associated with hydrolase, were significantly enriched at 24 h and 48 h in response to the presence of BCCM. We statistically analyzed the DEGs of carbohydrate active enzymes (CAZymes) in T. harzianum T4 at 12, 24, 48 and 72 h ( Figure 6A). There were 140 CAZyme genes with significant differential expression ( Figure 6A). Large numbers of CAZyme-related genes were significantly upregulated at 12, 24 and 48 h, but not at 72 h. At 12 h, there were 62 differentially expressed genes, 42 of which were significantly upregulated. At 24 and 48 h, 24 and 26 CAZyme genes were upregulated, respectively. In this experiment, a total of 82 GH glycoside hydrolase differential genes, 23 coenzyme activity genes, 17 glycosyltransferase and carbohydrate esterase genes and 1 carbohydrate-binding module gene were significantly differentially expressed ( Figure 6B).

Transmembrane Transport
Transmembrane transport plays an essential role in fungal growth and development, as well as in mycoparasitism, and is responsible for the transport of carbon sources, nitrogen sources, metal ions and toxic substances. GO enrichment analysis showed that in T. harzianum in the presence of BCCW, the upregulated gene was significantly enriched in the GO term of the transmembrane transport activity at 48 h. Of the 22 significantly upregulated transporters ( Figure 5C), 14 belonged to the major facilitator superfamily (MFS) of transporters, including 7 sugar transporter family genes (scaffold 37.g3, scaffold 22.g139, scaffold 17.g32, scaffold 16.g31, scaffold 12.g146, scaffold 10.g185 and scaffold 10. g155), 2 proton-dependent oligopeptide transporter (POT/PTR) genes (scaffold 81.g2 and scaffold 24.g29), 2 monocarboxylate transporter family genes (scaffold 9.g58 and scaffold 29.g70), and 1 iron transporter (scaffold 7.g253). In summary, MFS transporters play an influential role in Trichoderma mycoparasitism. Three amino acid polyamine organic (APC) family genes (scaffold 4.g305, scaffold 6.g129 and scaffold 11.g154) are responsible for the transport of amino acids and choline. The ammonium transporter MEP1 (scaffold 27.g12) is closely related to the nitrogen source transport and absorption of Trichoderma in the process of mycoparasitism. In general, transporter-related genes are involved in a series of material transports, such as carbon source, nitrogen source, amino acid, drug efflux protein and metal ion transport, during the growth and development of T. harzianum T4 in the presence of BCCW. Although the upregulated DEGs were significantly enriched in hydrolase-related genes at 24 and 48 h, the gene expression heat maps show that some hydrolase genes had an upregulated trend at 12 h. At 72 h, a large number of hydrolase-related genes were no longer upregulated or even significantly downregulated. This result suggests that hydrolase-related genes are unnecessary in the late growth phase of T. harzianum in the presence of BCCW. A large number of cellulase-related genes, such as beta glucosidase (scaffold 30.g33), endo-1,3 (4) beta glucosanase (scaffold 23.g102), exo-beta-1,3-glucanase (scaffold 11.g130), cellulase (scaffold 32.g55), and endochitinase 42 (scaffold 19.g145), were significantly upregulated at 12, 24 and 48 h. In addition, trehalose 6-phosphate synthase (scaffold 37.g2) genes were significantly upregulated at 12, 24 and 48 h and acetoxylan esterase (scaffold 32.g11) genes were significantly upregulated at 12 h and 48 h ( Figure 5B).

Transmembrane Transport
Transmembrane transport plays an essential role in fungal growth and development, as well as in mycoparasitism, and is responsible for the transport of carbon sources, nitrogen sources, metal ions and toxic substances. GO enrichment analysis showed that in T. harzianum in the presence of BCCW, the upregulated gene was significantly enriched in the GO term of the transmembrane transport activity at 48 h. Of the 22 significantly upregulated transporters ( Figure 5C), 14 belonged to the major facilitator superfamily (MFS) of transporters, including 7 sugar transporter family genes (scaffold 37.g3, scaffold 22.g139, scaffold 17.g32, scaffold 16.g31, scaffold 12.g146, scaffold 10.g185 and scaffold 10. g155), 2 proton-dependent oligopeptide transporter (POT/PTR) genes (scaffold 81.g2 and scaffold 24.g29), 2 monocarboxylate transporter family genes (scaffold 9.g58 and scaffold 29.g70), and 1 iron transporter (scaffold 7.g253). In summary, MFS transporters play an influential role in Trichoderma mycoparasitism. Three amino acid polyamine organic (APC) family genes (scaffold 4.g305, scaffold 6.g129 and scaffold 11.g154) are responsible for the transport of amino acids and choline. The ammonium transporter MEP1 (scaffold 27.g12) is closely related to the nitrogen source transport and absorption of Trichoderma in the process of mycoparasitism. In general, transporter-related genes are involved in a series of material transports, such as carbon source, nitrogen source, amino acid, drug efflux protein and metal ion transport, during the growth and development of T. harzianum T4 in the presence of BCCW.

Antioxidant Enzymes
Transcriptome sequencing of the T. harzianum T4 strain grown for 12, 24, 48 and 72 h in the presence of BCCM showed that the genes encoding oxidoreductase activity were significantly enriched in several culture time periods. Based on differential expression multiples, 22 differentially expressed genes that were significantly upregulated were identified ( Figure 5D). The differentially expressed genes include genes involved in cellular detoxification, oxidase, peroxidase, osmotic pressure regulation and amino acid biosynthesis and metabolism. Of these, NADPH oxidase Nox (scaffold 2.g181, scaffold 3.g249), which is responsible for ROS production, was significantly upregulated at 12 h. PRX1 (scaffold 13.g225), a bifunctional enzyme that can protect the body from oxidative damage, was significantly upregulated at 12 h. AzaB (scaffold 3. g68), which belongs to the polyketide synthase (PKS) family, was significantly upregulated at 24 h. Glt1 (scaffold14 14.g81), a glutamate synthase precursor-related gene, was significantly upregulated at several culture times, which attracted our attention and is further elaborated on in the discussion below.

Dual-Culture of T. harzianum T4 and B. cinerea to Verify the Genes Related to Mycoparasitism
In order to further verify the genes in T. harzianum related to mycoparasitism, we analyzed the expression patterns of these genes using a double-culture direct confrontation experiment [26,27]. RT-qPCR was performed using the total RNA of samples of the T. harzianum and B. cinerea dual culture before (BC), during (C) and after (AC) physical

Discussion
Some studies have confirmed that liquid media containing fungal host cell walls can be used as a model to identify the genes related to mycoparasitism [21][22][23][28][29][30]. The mechanism of mycoparasitism is a complex process, which is regulated by multiple factors such as different isolated strains and pathogens. In this study, we used transcriptome methods to explore the genes related to the mycoparasitism of T. harzianum T4 at the 4 growth stages of 12, 24, 48 and 72 h in the presence of BBCW or sucrose. A total of 10,266 predicted genes were identified from our transcriptome data, including 2871 DEGs and 2148 novel genes, greatly enriching the available genome annotation information for T. harzianum T4. Through GO and KEGG enrichment analysis of DEGs, we found that carbohydrate metabolism was significantly enriched at 12, 24 and 48 h. At 24 and 48 h, the T. harzianum T4 strain upregulated most of the genes related to carbohydrate active enzymes (in particular, hydrolase genes) in response to the presence of BCCW. By enriching the upregulated DEGs at different culture times, we further analyzed the genes that were significantly enriched at specific epochs, such as kinases and signal transduction genes, CAZymes (in particular, hydrolase), transmembrane transporters and oxidoreductase. This discovery confirmed that T. harzianum could maintain its own growth and eliminate the host through a series of processes, such as upregulating the expression of carbohydrate activity enzyme-related genes, enhancing the activity of hydrolase to hydrolyze the host cell wall and regulating the method of nutrition acquisition.
The mycoparasitism process of T. harzianum involves a series of activities such as recognition, adherence, attachment, persistence, hydrolysis and parasitism of the host hyphae at different growth stages [31]. In this regard, MAPK and G protein have been proven to be involved in the recognition and attachment stages of mycoparasitism [18,32]. Recent studies have shown that MAPK participates in the formation of cellulase and regulates its expression at the transcription level [33][34][35]. From the transcriptome of T. harzianum T4, we found that two genes encoding G protein and three MAPK protein kinases were significantly upregulated under BCCM induction compared with the control. In addition, an opportunistic 2-component system, phosphorelay intermediate protein YPD1 (mpr1), was significantly upregulated in the presence of BCCM at 12 h, which could positively regulate the MAPK cascade [36]. YPD1 also plays a key role in cell functions such as cell wall synthesis and spore formation, osmotic and oxidative stress, and toxicity [37,38].
CAZymes are closely related to the nutritional patterns of fungi and participate in the degradation of the cell walls of pathogens and stored compounds [39,40]. When the T. harzianum T4 strain was grown in an induction medium with BCCW for 24 and 48 h, a significant number of hydrolase genes, such as chitinase, cellulase and glucanase, were significantly upregulated. In addition to hydrolase genes, glycosyltransferase and protease genes were significantly upregulated, and these play a key role in mycoparasitism [41,42]. Cellulase can hydrolyze the mycelia of fungal pathogens during the mycoparasitism process. Some studies have shown that the greater the cellulase activity, the greater the biological control potential [43][44][45]. A new gene encoding endo-1,3 (4) beta glucanase (bgn13.1) was significantly upregulated in multiple culture time periods and this was directly involved in the mycoparasitism between T. harzianum and B. cinerea [46,47]. The role of these upregulated genes in the mycoparasitism of T. harzianum T4 involves the penetration and degradation of the cell wall, which inhibits the spore germination and mycelial growth of the pathogen.
Trichoderma harzianum T4 adapted to the presence of the host cell walls by upregulating genes associated with CAZymes in carbohydrate metabolism, particularly glycosidic hydrolases. After hydrolyzing carbohydrate macromolecules, transport and absorption also play a role in mycoparasitism. The upregulation of the glucose transporter HXT at 12 h was closely related to improvements in the sexual reproduction, antioxidant capacity, pathogenicity and toxicity of T. harzianum T4 [48][49][50]. The POT/PTR family protein PTR2 was upregulated at 24 and 48 h, is involved in the transport of small peptides and is closely related to the internal transport of nutrient stores and the absorption of host surface decom-position products during mycoparasitism [51]. MEP1, an ammonium transporter that can absorb low concentrations of ammonium under nitrogen restriction to maintain cell growth, was upregulated at 12, 24, 48 and 72 h [52]. SIT1, which mediates iron absorption, was upregulated at 48 h and is essential to T. harzianum mycoparasitism [53][54][55]. The upregulation of choline transporter HNM1 at 48 h and 72 h is critical for conidial germination, sprout tube elongation and fungal toxicity [56]. Therefore, transmembrane transporters participate in the transport of a range of materials, such as carbon and nitrogen sources, metal ions and toxic substances, that play an important role in the process of mycoparasitism.
We found that T. harzianum responded to the oxidative damage by upregulating antioxidant enzymes or metabolites during stress conditions (in the presence of the host or at the late phase of cultivation) [57]. Transcriptome data revealed that T. harzianum responded to BCCW by upregulating NADPH oxidase NOxA to produce ROS while upregulating peroxidase Prx1 to protect cells from oxidative damage [58,59]. In addition, MtlD and P5CR were significantly upregulated; these genes participate in the biosynthesis of mannose and proline and play a key role in regulating cellular osmolarity and oxidative stress [60,61]. Of these, mannitol is a powerful quencher of reactive oxygen species (ROS) in fungi [62]. T. harzianum directly attacks the host during mycoparasitism and produces toxic secondary metabolites to kill the pathogen. Of these, nonribosomal peptide synthetases (NRPSs) and lovastatin nonaketotide syntheses AzaB (a polyketone compound) were significantly upregulated. They are important secondary metabolites of T. harzianum and these genes can be combined with hydrolase genes to further inhibit the growth of pathogens and enhance the mycoparasitism of T. harzianum T4 [63,64].
When faced with host and secondary metabolites, T. harzianum upregulated resistance proteins and detoxification-related genes to increase its ability to parasitize by enhancing its adaptation to stress environments. RP-L3e is a Trichoderma-resistant protein gene which plays an important role in amino acid tRNA binding, peptidyltransferase activity and drug resistance [65]. GST is a unique multifunctional cell membrane enzyme that is involved in the detoxification of toxic compounds and protection from oxidative damage [66][67][68][69][70]. An aging-related NAD-dependent histone deacetylase, SIR2, regulated by S-glutathione, extends lifespan by increasing lipid droplet and trehalose content and plays a key role in fungal parasites [71,72]. We also found that the expression of lipid droplet-associated hydrolase was downregulated at 12 h. TPS involved in trehalose synthesis was upregulated at 12, 24 and 48 h. Trehalose formation is essential for the maintenance of cell homeostasis and formation of appressoria [73,74].
In addition, a gene encoding a WSC domain-containing protein attracted our attention. The gene was significantly upregulated at 24, 48 and 72 h of T. harzianum growth in the presence of BCCM. Some studies have shown that the WSC domain-containing protein can activate the MAPK cascade [75], which is related to the integrity of cell walls, oxidation, hypertonic and metal ion binding. Studies have also shown that this protein is related to lectin, although this has not been definitively proved [76]. Two new genes encoding transketolase, TKL1 and TKL2, were significantly upregulated during mycoparasitism. TKL1 is mainly responsible for the response to osmotic stress and promotion of biotrophic growth and its overexpression can increase lipid content. TKL2 is mainly responsible for the response to oxidative stress [77]. These two transketolase enzymes play unique and complementary roles in coping with different environmental pressures [78]. However, the specific functions of these novel genes in the mycoparasitism of T. harzianum require verification by further experiments.
Among the DEGs in T. harzianum T4, we identified many unknown functional or hypothetical protein genes. These DEGs are highly induced by BCCW but are not, or are rarely, expressed when T. harzianum T4 is grown with itself, which may be significant in mycoparasitism. In our forthcoming work, we will further verify the specific functions of mycoparasitism-related genes screened by the transcriptome via gene knockout, overexpression and double-culture antagonism experiments.

Conclusions
In conclusion, we used the transcriptome sequencing technology to identify DEGs related to mycoparasitism of T. harzianum T4 at 12, 24, 48 and 72 h of growth in the presence of cell walls of B. cinerea (BCCW). Through GO and KEGG enrichment analysis of DEGs, we identified genes associated with mycoparasitism at specific time periods, such as genes encoding kinases, signal transduction proteins, carbohydrate active enzymes, hydrolytic enzymes, transporters, antioxidant enzymes, secondary metabolite synthesis, resistance proteins, detoxification and genes associated with the extension of hyphal longevity. The transcriptome data provided in this study will not only improve the annotation information of gene models in the T. harzianum T4 genome, but also provide important transcriptome information involved in the process of mycoparasitism at specific time periods, thus providing a potential molecular target for other genes that use T. harzianum T4 as a biological control agent.
Author Contributions: Conceptualization, formal analysis, investigation, data curation, writingoriginal draft, writing-review and editing, and visualization, Y.W.; investigation and data curation, X.Z.; methodology, data curation and original draft modification, J.W.; investigation, C.S.; conceptualization, methodology, validation, supervision, project administration and funding acquisition, W.W. All authors have read and agreed to the published version of the manuscript.
Funding: This work was supported by the Project of Prospering Agriculture through Science and Technology of Shanghai, China (grant number 2022-02-08-00-12-F01163).