Transcriptomic analyses of plant pathogens can contribute to our understanding of genetic mechanisms related to pathogenicity, disease, and biological functions (DiGuistini et al., 2011; Keeling et al., 2012; Toledo-Arana et al., 2009) by examining gene expression at different stages of infection and disease development. Expressed genes vary depending on cell/tissue type, environmental factors, developmental stage, and many other interacting factors and signals. For these reasons, transcriptomic analyses are essential for characterizing the genetic bases of an organism’s functions under variable environments. Transcriptomic analyses of fungal pathogens that cause tree diseases can also provide baseline information for examining host-pathogen and host-vector interactions, pathogenesis, general metabolism, development, environmental adaptation, and ecological functions. Information on basic physiological mechanisms of D. quercus-mongolicae is currently quite limited. Consequently, the objective of this study was to analyze the functional annotations of expressed transcripts from D. quercus-mongolicae grown in vitro to provide baseline information toward understanding biological functions of this fungus.

The D. quercus-mongolicae isolate (KACC44405) used in this study was obtained from the Korean Agricultural Culture Collection (National Agrobiodiversity Center, Jeonju, South Korea). This isolate was selected for our transcriptomic study because of the available genomic sequence information (Jeon et al., 2017), and it also serves as the type specimen for describing D. quercus-mongolicae (as R. quercus-mongolicae) (Kim et al., 2009). The fungus was grown under in vitro conditions on three culture media that represented a high-carbon source condition (potato dextrose agar; PDA), very low carbon source condition (water agar; WA), and partially resembling host condition (sawdust-amended water agar; SWA). The SWA medium contained sawdust collected from Q. mongolica using a wood-chip grinding machine (FJ Tech, Gimpo, South Korea) that was added to the WA medium before autoclaving. Pieces containing (3 × 3 × 3 mm) fungal inoculum were excised from an actively growing culture and placed onto a sterile, nylon-membrane filter (Millipore Sigma, Burlington, MA, USA) that was overlayed on each of the three culture media, PDA (39 g/L, BD-Difco, Sparks, MD, USA), SWA (100 g/L sawdust with 1.5% WA, BD-Difco), and WA (1.5% agar, BD-Difco), before incubating at 25 ± 1°C. After 7 days, each mycelial sample (100 mg wet weight) was gently scraped off from the nylon membranes overlaying each of the three culture media and immediately placed into pre-chilled, 2.0-mL microtubes containing two stainless steel beads (5-mm diameter). The tubes containing mycelia from each respective medium were stored at −80°C for 5 min, and then mycelial samples were ground three times for 3 min at 40 Hz using TissueLyser (TissueLyser LT, Qiagen, Hilden, Germany). After grinding the tissue, tubes were stored again at −80°C for 3 min before adding 800 μL RNA lysis buffer and again disrupting the mycelia for 1 min at 50 Hz. Total RNA was extracted using the Quick-RNA Fungal/Bacterial Mini-Prep Kit (R2014, Zymo Research, Tustin, CA, USA) following the manufacturer’s protocol. The resulting RNA concentration was measured using a NanoDrop 2000 (Thermo Fisher Scientific, Waltham, MA, USA), and RNA integrity number was checked using TapeStation (Agilent Technologies, Santa Clara, CA, USA).

The mRNA was isolated from total RNA using an mRNA purification kit (Zymo Research), and an Illumina cDNA library was generated from isolated mRNA using TruSeq RNA Sample Prep Kit (Illumina Inc., San Diego, CA, USA). The cDNA library was sequenced at Macrogen Korea (Seoul, South Korea) using Illumina HiSeq 2000 (Illumina Inc.), producing a read length of 101 bp. The resulting sequence reads were checked for quality using FastQC v0.12.1 (https://www.bioinformatics.babraham.ac.uk). The software Trimmomatic v0.38 (Bolger et al., 2014) was used to remove adapters and remove low-quality sequences, if any. Resulting reads were aligned to the D. quercus-mongolicae genome (accession no. NIPS00000000) (Jeon et al., 2017) using HISAT2 version 2.2.1 (Kim et al., 2019), setting the maximum intron length to 6,000 bp. The aligned reads were assembled, and their FPKM values (fragments per kilobase of transcript per million fragments mapped) were estimated with using Cufflinks (version 2.2.1) (Trapnell et al., 2010). This Transcriptome Shotgun Assembly project has been deposited to the Sequence Read Archive (BioProject: PRJNA513385) in National Center for Biotechnology Information (accession no. SRR32758122 for PDA; SRR32757902 for SWA; and SRR32757901 for WA). All transcripts were annotated using the UniProtKB database using MMseqs2 (Release 14-7e284) (Steinegger and Söding, 2017) to determine the predicted functions of proteins encoded by each transcript. The longest open reading frames were predicted using Orfipy (Singh and Wurtele, 2021). These protein-encoding sequences were analyzed with DeepLoc version 2.0 (Thumuluri et al., 2022) to predict the subcellular localization of putative proteins. Any putative proteins that were localized in the ‘Extracellular’ region were subjected to analysis by EffectorP version 3.0 (Sperschneider and Dodds, 2021) as secreted proteins. Carbohydrate-active enzymes (CAZymes) were predicted using dbCAN3 version 4.0 (Zheng et al., 2023).

The transcriptome libraries from all mycelial samples of D. quercus-mongolicae generated a total of 94,951,536 reads. From the PDA sample, a total of 31,183,399 reads were generated (GC content: 57%), and the ratio of aligned reads was 97.40%, while from the SWA sample, a total of 34,052,359 reads were generated (GC content: 57%), and the ratio of aligned reads was 97.49%. Lastly, the reads from the WA sample generated 29,715,778 reads (GC content: 56%), with an alignment ratio of 97.13% (Table 1). The number of transcripts generated from PDA, SWA, and WA samples were 6,739, 6,900, and 6,855 transcripts, respectively. A total of 6,679 transcripts were commonly expressed across all three media types (Supplementary Fig. 1). Among all assembled transcripts, 4,368 transcripts were successfully annotated with putative protein and gene names, Gene Ontology (GO) terms, organism, Kyoto Encyclopedia of Genes and Genomes, and UniPathway by searching the Uniprot database using the MMseq2 program.

Among the 4,368 annotated transcripts, Table 2 shows the top 20 most highly expressed transcripts that encoded putative proteins based on the matched sequences. Mean values derived from the three culture media were applied to select the top 20 highly expressed transcripts. The 40S ribosomal protein S25, ceramide very long chain fatty acid hydroxylase, Epl1 protein, ADP/ATP translocase, and dihydrofolate synthetase fol3 were determined to be highly expressed in D. quercus-mongolicae (Table 2). Ribosomes play a well-known crucial role in protein production and are vital for a wide variety of organisms, from bacteria to animals (Karsi et al., 2002). The eukaryotic 80S ribosome consists of a small subunit (40S) and a large subunit (60S), with the 40S subunit containing the 18S rRNA and 32 ribosomal proteins (Wool, 1979). Ribosomal protein genes are responsible for coding the structural elements of ribosomes, which are the cellular machinery for the protein synthesis process (Mullis et al., 2020). Because ribosomal proteins are essential for fungal growth and survival, it is expected that the 40S ribosomal protein S25 would be highly expressed in D. quercus-mongolicae. Very long chain fatty acids (VLCFAs) are fatty acids that have chain lengths more than 18 carbon atoms (Batsale et al., 2023). VLCFAs serve as essential components for creating more complex molecules and play important roles in various cellular functions, such as energy storage, biochemical signaling, and membrane structure as various types of lipids, such as triacylglycerols, sphingolipids, and phosphoglycerolipids (Haslam and Kunst, 2013). In numerous organisms, the VLCFAs are also fundamental in forming an extracellular structure known as the cuticle, which covers the outside of these organisms and acts as the main physical barrier against the environment (Raffaele et al., 2009). Although VLCFAs compose a small fraction of the total fatty acids in cells, their importance is exemplified by the fact that plants cannot survive without VLCFAs (Beaudoin et al., 2009); however, further investigation is warranted to understand the precise role of VLCFAs in D. quercus-mongolicae. Ceramide hydroxylase plays a role in the hydroxylation process of VLCFAs that are associated with sphingolipids (Chen et al., 2008). Higher expression levels of ceramide VLCFA hydroxylase-encoded transcripts indicates that this protein is likely important for survival and growth of this fungus. Eliciting plant response (Epl) 1 protein-encoded transcripts were highly expressed in D. quercus-mongolicae. The Epl1 protein, which is reported as a plant-resistance elicitor, is crucial for fungal adaptation to environmental conditions, particularly in Trichoderma and other interactions among fungal pathogens and plants (Gomes et al., 2015). In addition, ADP/ATP translocase, dihydrofolate synthetase, cytochrome c oxidase, woronin body major protein, An1-type zinc finger protein, and ABM domain-containing protein-encoded transcripts were also greatly enriched in transcriptome of D. quercus-mongolicae (Table 2).

Gene annotation based on Gene Ontology Functional Classification System (GO terms) was applied to functional annotation of transcripts from D. quercus-mongolicae, and the results were shown in Fig. 1. The three GO aspects provided by the GO database comprised biological process (BP), molecular function (MF), and cellular component (CC). Results from this study indicated that 39.4% (1,722 out of 4,368 transcripts), 61.2% (2,675 out of 4,368 transcripts), and 43.3% (1,890 out of 4,368 transcripts) were successfully annotated to BP, MF, and CC aspects, respectively. The expressed transcripts within the BP aspect were annotated to fatty acid biosynthetic process (GO:0006633), mitochondrial ADP/ATP transmembrane transport (GO:0140021, GO:1990544), one-carbon metabolic process (GO:0006730), positive regulation of translational elongation/termination (GO:0045901, GO:0045905), and translation (GO:0006412) (Fig. 1). Meanwhile, the MF aspect comprised expressed transcripts encoding fatty acid α-hydroxylase activity (GO:0080132), ATP:ADP antiporter activity (GO:0005471), ATP binding (GO:0005524), ATP hydrolysis activity (GO:0016887), and ribosome binding (GO:0043022). Within the transcriptome of D. quercus-mongolicae, genes with higher expression that were annotated to the CC aspect included ribonucleoprotein complex (GO:1990904), endoplasmic reticulum membrane (GO:0005789), extracellular region (GO:0005576), mitochondrial inner membrane (GO:0005743), and mitochondrial respiratory chain complex IV (GO:0005751) (Fig. 1).

Based on the results of pathway annotation, three different pathways were annotated, which included sphingolipid metabolism (lipid metabolism), L-tryptophan biosynthesis (amino acid biosynthesis), and glycolysis (carbohydrate degradation) (Table 3). Sphingolipids are critical components of the fungal plasma membrane, contributing to the stability and integrity of the cellular structure (Fabri et al., 2020), and these components play an essential role in regulating membrane fluidity, which enables fungi to adapt altering environmental conditions (Usmani et al., 2023). Furthermore, sphingolipids have been shown to be important for the modulation of signal transduction pathways correlated with cell growth and stress responses (Dickson et al., 2006). Moreover, the sphingolipids also contribute to fungal colonization and infection processes (Usmani et al., 2023). Based on our results, sphingolipid metabolism is also likely important for growth of D. quercus-mongolicae. The important roles of fungi-specific tryptophan synthases gene (TRP1 gene) in fungal survival through the tryptophan biosynthesis are well demonstrated in previous studies (Ireland et al., 2008; Sen et al., 2016; Staats et al., 2004). In our study, the transcripts for the tryptophan synthase pathway and its biosynthesis metabolism were highly expressed in D. quercus-mongolicae, which suggests a critical role for this metabolic pathway.

Many fungi produce various enzymes to degrade plant cell-wall components, such as cellulose, hemicellulose, and pectin, or plant carbohydrates, such as starch, sucrose, and mannose (Ibarra Caballero et al., 2019). These enzyme families, which are known as CAZymes, comprise several groups, including glycoside hydrolases (GHs), glycosyltransferases, polysaccharide lyases, carbohydrate esterases, auxiliary activities, and carbohydrate-binding modules (Drula et al., 2022). Among all of the expressed transcripts in D. quercus-mongolicae, 10 transcripts were annotated as CAZymes (Table 4) including heat shock protein 30 and D-lactate dehydrogenase with cellulose as the likely substrate. Annotated CAZymes also included α-1-rhamnosidase and GH family 32 proteins that likely use pectin and sucrose as substrates, respectively. Furthermore, annotated transcripts included a class 3 chitinase 2.

The function of a protein is highly dependent on its localization and prediction of its subcellular location can provide key information about its functionality and possible interactions with other proteins (Gillani and Pollastri, 2024). In this study, we attempted to localize encoded proteins. After identifying the locations of each protein with the DeepLoc model, a total of 515 proteins were localized in the extracellular region, and these proteins were considered as putative secreted proteins from D. quercus-mongolicae. Among the 515 proteins, 100 (19.4%) were identified as predicted effector proteins, and 415 (80.6%) were identified as non-effector proteins. Of the predicted 100 effector proteins, 57 (11.1%) were predicted cytoplasmic effectors and 43 (8.3%) were predicted apoplastic effectors. Subsequently, a 0.75 cut-off value of prediction score was implemented to select proteins with a higher probability of being an effector protein (Table 5). Of these higher probability effector proteins, 13% (13 out of 100) and 15% (15 out of 100) of effectors were localized to the apoplast and cytoplast, respectively. Because apoplastic effectors directly interact with host plant tissues on the outside of plant cell, including the cell wall and intercellular spaces (Rocafort et al., 2020), we focused on these apoplastic effectors. In this study, apoplastic effectors identified in D. quercus-mongolicae included non-hemolytic phospholipase C, chitinase, DNAse1 protein, amine oxidase, GH, ribosomal protein S17, and uncharacterized protein. Phospholipids and proteins constitute the primary chemical components of the plant cell membrane (Colin and Jaillais, 2020). Consequently, enzymes that can hydrolyze these plant cell membrane components, such as phospholipases and proteinases, are likely to play a significant role in the processes of membrane disruption that occur during the infection and invasion of plant host cells (Schmiel and Miller, 1999). The action of phospholipases, which cleave phospholipids, leads to membrane destabilization and ultimately results in host cell lysis (Ghannoum, 2000). Furthermore, fungal chitinases may serve a fundamental role in the plasticization of the cell wall, or they may function more specifically in processes such as cell separation, the acquisition of nutritional chitin, or competitive interactions with other fungal species (Langner and Göhre, 2016). A fungal GH, which is also a CAZyme class, was also identified as an apoplastic effector in the D. quercus-mongolicae. This enzyme can hydrolyze glycosidic bonds between carbohydrate molecules of fungi (Valadares et al., 2016). Additionally, previous studies provide substantial evidence for the significant role various GH protein families play in the virulence of phytopathogenic fungi (Bradley et al., 2022; Vu et al., 2012). As such, various apoplastic effectors, non-hemolytic phospholipase C, chitinase, and GH, can be considered to have potential roles in the virulence of D. quercus-mongolicae.

In summary, this study has identified 4,368 significant transcripts of D. quercus-mongolicae, many of which encode proteins essential for the fungal survival and growth. Within in vitro grown D. quercus-mongolicae, these highly expressed proteins include 40S ribosomal protein S25, ceramide very long chain fatty acid hydroxylase, Epl1 protein, and ADP/ATP translocase. These proteins are particularly noteworthy due to their critical roles in the metabolism and environmental adaptation of fungi. Key metabolic pathways, including sphingolipid metabolism, L-tryptophan biosynthesis, and glycolysis, which have important implications for the physiological functions of D. quercus-mongolicae. Furthermore, the phospholipase C, chitinase, and GH were identified apoplastic effectors, which could be associated with virulence of D. quercus-mongolicae. As a consequence, this baseline transcriptomic study provides useful background information for future studies aimed at better understanding the physiological mechanisms and biological functions of D. quercus-mongolicae.