Selection of reference genes in Thermothelomyces fergusii: insights from genomic and expression analysis in response to different carbon sources

Abstract

Background

RNA expression analysis using reverse transcription qPCR requires reference genes for accurate data normalization. However, some reference genes can show significant variability in specific biological samples, raising concerns about the reliability of this method. This study assessed the expression levels of 12 reference genes in Thermothelomyces fergusii during spore and mycelium cultivation under different carbon source media (avicel, cellobiose, and glucose) using RT-qPCR. T. fergusii is a thermophilic fungus known for its abundance of cellulase and industrial applications. Quantifying fungal gene expression using the RT-qPCR method is essential for investigating the molecular biology of T. fergusii.

Results

The stability of 12 reference genes in T. fergusii was evaluated using RefFinder, geNorm, NormFinder, BestKeeper, and delta Ct techniques. The robustness of the chosen reference gene was verified using the LPMO AA9-4 gene. Analysis of the different methods showed variations in the results of reference genes under different sugar treatments. Consistently, EF1, GAPDH, and 18 S-3 were identified as reliable reference genes by the geometric mean of ranking values from Pearson correlation coefficient (r) analysis of pooled data, ranking values of RefFinder in pooled data, and NormFinder of pooled data. However, BT-3, ACT-3, and 18 S-2 exhibited inconsistency as reference genes.

Conclusion

These results emphasize the importance of verifying reference genes in specific experimental models and provide valuable recommendations for selecting reliable reference genes in RT-qPCR. These findings have broad applicability for gene expression investigations involving T. fergusii.

Background

Thermothelomyces fergusii is a thermophilic fungus with significant industrial applications, particularly in cellulose breakdown and the biofuel industry [1]. Although T. fergusii, formerly Corynascus thermophilus, is a thermophilic fungus belonging to the Chaetomiaceae family [2]. There is still a lack of comprehensive knowledge regarding the molecular mechanisms underlying gene expression during the development of various carbon sources. Gene can be either upregulated or downregulated during the development of this fungus under different experimental conditions. However, there is limited knowledge regarding the expression of reference genes in this specific thermophilic fungus.

Reverse transcription-quantitative PCR (RT-qPCR) is a widely used method for accurately measuring changes in gene expression in microbiological research [3]. However, one of its major concerns is the erroneous normalization of expression data [4]. To ensure accurate gene expression analysis, using reference genes is currently the most reliable method for normalizing mRNA fractions of interest [5]. However, recent research has shown that the expression levels of reference genes can vary depending on the specific gene, cell type, and experimental conditions [6, 7]. Therefore, it is crucial to validate the use of reference genes in each experimental condition to avoid detrimental fluctuations in gene expression [8, 9]. Unfortunately, there is limited research on suitable internal control genes (ICGs) for filamentous fungi, with current studies mainly focusing on relative expression rather than precise expression levels [10–12].

It is well-established that no single gene within a different cell can consistently and uniformly express itself under all conditions [10]. Traditionally, widely recognized reference genes such as β-actin (ACTB), β-tubulin (β-TUB), glyceraldehyde-3-phosphate dehydrogenase gene (GAPDH), and 18 S rRNA (18 S) have been used as internal reference genes in previous research [5, 11, 12]. However, recent evidence indicates that these conventional reference genes exhibit unstable expression and significant variability, especially with the widespread use of high-throughput sequencing techniques [13]. The expression levels of GAPDH and ACT have been found to vary across different species of organisms. For example, variation in gene expression was observed when using these reference genes in asthmatic disease tissues in humans, as well as different developmental stages, and environmental conditions in Lentinus polychrous, Ulocladium, and Trichoderma atroviride species of fungi [10, 14–17].

Several fungal species, such as Volvariella volvacea, Pandora neoaphidis, Colletotrichum camelliae, and Amylostereum areolatum, have undergone validation of suitable reference genes for normalization [5, 12, 18, 19]. In a study involving 90 experiments, twenty reference genes were analyzed, with beta-tubulin being the most frequently used, followed by ACT, 18 S, and GAPDH [20]. However, even with stable expression across different environments, microorganisms can still exhibit variations in gene expression due to differences in organizational structures and the specific genes being expressed. The main challenge with these internal controls is the circular dilemma in evaluating the stability of a proposed normalization gene. To address this issue, Vandesompele et al. [21] introduced a method called geNorm over a decade ago, which allows for assessing and ranking candidate reference genes based on their expression stability.

Inaccurate gene expression can occur when an inappropriate reference gene with fluctuating expression is selected, leading to erroneous assumptions about the function of these target genes. To ensure reliable transcript expression analysis, it is crucial to identify suitable reference genes. Several statistical algorithms have been developed to identify internal controls with minimal variation in expression. These algorithms utilize RT-qPCR data from a specific set of potential genes and evaluate the stability of their expression. The five main methods commonly used for assessing the suitability of reference genes are RefFinder [22, 23], geNorm [21], NormFinder [24], Best-Keeper [25], and Delta Ct [26]. The availability of reference genes for RT-qPCR analysis in T. fergusii is currently limited. However, the current study focused on the screening of reference genes in T. fergusii under diverse experimental conditions. This study aimed to evaluate the stability of twelve candidate genes to determine the most appropriate reference genes for transcript normalization during spore germination and mycelial growth on different carbon source media (avicel, cellobiose, and glucose) in T. fergusii. Additionally, we examined the expression of the target gene Lytic polysaccharide monooxygenases (LPMOs) an auxiliary activity 9 (AA9), which plays a role in cellulose degradation in T. fergusii, to assess the effectiveness of specific reference genes. To the best of our knowledge, this study represents the initial investigation into the expression stability of suitable reference genes in T. fergusii.

Results

Screening and primer efficiency analysis of candidate reference genes

The twelve (12) internal control genes were identified from the genome of T. fergusii by conducting a reference gene analysis similar to that performed in Puccinia helianthi. The reference genes analysis from the transcript of P. helianthi was performed with four algorithms, including geNorm, NormFinder, BestKeeper, and delta Ct [8]. The specific characteristics, including the gene’s original ID, the gene’s ID renaming, product size, and melting points, are given in a table (Table 2). The primer efficiencies were calculated by employing a five-fold dilution of cDNA for all twelve reference gene primers, resulting in an r2 value ≥ 0.98 (Table 2; Fig. 1A). All 12 pairs of primers were used to amplify the candidate’s genes to confirm the primer’s efficiency (Fig. 1B, Fig. S1). The melting curves of the amplified products, generated using 12 pairs of primers, indicated the successful amplification of a single product of the predicted size. This demonstrated the absence of contamination in the reagents, as there was no amplification of other genomic DNA and non-specific amplification of other complementary DNA. The maximum CT value recorded was 34.89, specifically in the EF-2 genes (Fig. 1A).

[IMAGE OMITTED: SEE PDF]

Gene expression spectra of candidate reference genes

To determine the reliability of a reference gene, it needs to maintain a consistent level of expression throughout different experimental conditions. Therefore, we proceeded to assess the expression of these potential reference genes by examining their Cq values. RNA was isolated from the mycelium developed from fresh spores of T. fergusii that were cultivated on a liquid seed medium for 24 h. In addition, a liquid fermentation medium containing three different carbon sources, namely avicel, cellobiose, and glucose, each at a concentration of 1.5%, was utilized to cultivate T. fergusii for 5 days. The whole RNA was converted into complementary DNA (cDNA) using reverse transcription, and subsequently employed for RT-qPCR analysis. Five serial dilutions were performed for cDNA on 1% glucose to quantify the expression analysis (Fig. 2). The average Cq values across these dilutions for the 12 putative reference genes exhibited a relatively narrow range. Among them, ACT-3, BT-3, and 18 S-2 showed higher Cq values, 28.65, 27.07, and 30.02, respectively, indicating lower expression levels. In contrast, EF-1, GAPDH, and 18 S-3 displayed lower Cq values of 19.75, 21.24, and 25.31, respectively, suggesting comparatively higher expression levels (Fig. 2). Each reference gene exhibited diverse expression across all analyzed samples.

[IMAGE OMITTED: SEE PDF]

Expression stability of reference genes by BestKeeper and NormFinder

The BestKeeper algorithm was applied to evaluate the expression stability of 12 candidate reference genes by calculating the standard deviation (SD) and coefficient of variation (CV). Genes with SD values less than 1 were considered to exhibit stable expression and were deemed suitable for use as reference genes. The analysis was conducted across four conditions: control (1% glucose), and media supplemented with 1.5% avicel, 1.5% cellobiose, and 1.5% glucose (Table S1). Under control conditions, GAPDH and EF-1 were among the top-ranked genes, displaying the lowest SD and CV values, indicating high expression stability. In contrast, ACT-3 and 18 S-2 were among the least stable genes, showing high variability in expression (Fig. 3A, Table S1). In the avicel-treated group, GAPDH sustained high stability, ranking within the top two genes, while EF-1 and 18 S-3 maintained moderate stability. However, ACT-3, BT-3, and 18 S-2 showed increased variability, consistently ranking among the least stable genes under this condition (Fig. 3B, Table S1). Similarly, in the 1.5% cellobiose treatment group, EF-1 emerged as the most stable gene, followed closely by 18 S-3 and GAPDH. These genes exhibited low SD and CV values, reinforcing their reliability as reference genes. On the other hand, ACT-3, 18 S-2, and BT-3 were once again ranked among the least stable, showing inconsistent expression across different treatments (Fig. 3C, Table S1). In the 1.5% glucose treatment, the pattern remained consistent, with EF-1, GAPDH, and 18 S-3 ranking within the top five most stable genes. Conversely, ACT-3, BT-3, and 18 S-2 were placed among the lowest ranks, reflecting poor expression stability under this condition as well (Fig. 3D, Table S1).

[IMAGE OMITTED: SEE PDF]

NormFinder analysis further supported these findings. It consistently ranked EF-1, GAPDH, and 18 S-3 among the six most stable genes across all treatments, whereas ACT-3, BT-3, and 18 S-2 showed substantial variability. For instance, EF-1 and GAPDH were among the top four ranked genes in the control, cellobiose, and glucose groups, while 18 S-3 showed high stability, particularly under control, avicel, and cellobiose treatment conditions (Fig. 4A–D, Table S1). Overall, both BestKeeper and NormFinder analyses confirmed that EF-1, GAPDH, and 18 S-3 were the most suitable reference genes for gene expression studies in T. fergusii under different lignocellulosic carbon sources. In contrast, ACT-3, BT-3, and 18 S-2 displayed inconsistent expression patterns and are thus unsuitable for normalization purposes.

[IMAGE OMITTED: SEE PDF]

Expression stability of reference genes by comprehensive analysis

To determine the overall expression stability of candidate reference genes in T. fergusii, a comprehensive ranking was performed using RefFinder, which integrates four established algorithms: geNorm, NormFinder, BestKeeper, and the ΔCt method. Geometric mean (geomean) ranking values were calculated to evaluate the stability ranking across different approaches. In the control condition (1% glucose), GAPDH, EF-1, and 18 S-3 were consistently ranked among the five most stable reference genes, whereas 18 S-2, BT-3, and ACT-3 were classified among the six least stable genes (Fig. 5A). Under 1.5% avicel treatment, GAPDH, EF-1, and BT-3 were included in the six most stable genes. In contrast, ACT-3, and 18 S-2 appeared among the least stable reference genes (Fig. 5B), highlighting the context-dependent variability of BT-3. In samples treated with 1.5% cellobiose, GAPDH, EF-1, and 18 S-3 consistently ranked among the top four most stable genes, whereas BT-3, 18 S-2, and ACT-3 remained among the least stable (Fig. 5C). Similarly, under 1.5% glucose treatment, EF-1, GAPDH, and 18 S-3 were again classified as the top three stable genes, in contrary, BT-3, 18 S-2, and ACT-3 were positioned among the bottom seven least stable genes (Fig. 5D).

[IMAGE OMITTED: SEE PDF]

To validate and strengthen these findings, RefFinder was used to assess the pooled dataset combining all four carbon source conditions. The overall geomean ranking revealed that EF-1, GAPDH, and 18 S-3 were the most stable reference genes across all conditions, conversely, ACT-3, 18 S-2, and BT-3 were consistently ranked among the least stable (Fig. 6A). Pearson correlation coefficient analysis further supported these results, placing 18 S-3, EF-1, and GAPDH among the top four most stable genes, while, BT-3, ACT-3, and 18 S-2 showing the lowest stability (Fig. 6B). Bestkeeper analysis revealed that BT-3, GAPDH, EF-1, and 18 S-3 were identified in the top six stable reference genes, meanwhile, 18 S-2 and ACT-3 were evaluated in the six most least stable reference genes in T. fergusii (Table 1, Fig. 6C). NormFinder analysis also indicated that 18 S-3, EF-1, and GAPDH as the most stable genes, while BT-3, ACT-3, and 18 S-2 as the least stable reference genes (Table 1, Fig. 6D). The ΔCt method yielded similar results, ranking EF-1, 18 S-3, and GAPDH as the top three stable reference genes. According to geNorm analysis, EF-1 and GAPDH were identified as the most stable reference genes, whereas 18 S-3 was ranked seventh, indicating slight variation in gene stability rankings among the different evaluation methods (Table 1). Predominantly, BT-3, ACT-3, and 18 S-2 consistently emerged as the least stable genes across all analytical tools.

[IMAGE OMITTED: SEE PDF]

[IMAGE OMITTED: SEE PDF]

Discrepancies observed in the rankings of 18 S-3 and BT-3 across certain conditions may be attributed to differential gene expression responses to specific carbon sources and possible fluctuations in fungal growth phases, underscoring the importance of validating reference genes under each experimental condition. In conclusion, EF-1, GAPDH, and 18 S-3 demonstrated superior expression stability across all treatments and are recommended as the most reliable reference genes for qPCR normalization in T. fergusii under lignocellulosic substrate conditions. In contrast, BT-3, ACT-3, and 18 S-2 should be avoided due to their inconsistent expression and poor stability.

Impact of specific reference genes in gene normalization analysis

The influence of reference gene selection on the estimation of gene expression was evaluated using the expression of the auxiliary activity 9 (AA9-4) family of genes (gene ID_3234). Three highly stable and three least stable genes were used to normalize the expression of the target gene AA9-4 under different carbon source treatments (Fig. 7). When utilizing the least stable genes BT-3, ACT-3, and 18 S-2, the expression pattern of AA9-4 showed a significant difference for each of 1.5% AV, 1.5% C, 1.5% G, and the average expression (pooled) of AA9-4 under the three sugars (Fig. 7B). Nevertheless, the relative expression levels of AA9-4 were not significantly different when the three stable reference genes EF-1, 18 S-3, and GAPDH were used for data normalization on 1.5% cellobiose, 1.5% glucose, and the pooled data of the three sugars (Fig. 7A). Interestingly, significant variation was observed in the expression level of the AA9-4 gene when EF-1, 18 S-3, and GAPDH, the three most stable genes, were used for data normalization on 1.5% AV. This variation suggests that the crystalline nature of avicel substantially deviated the expression of AA9-4 when different reference genes were used for data normalization, underscoring the impact of substrate type on the stability of reference genes. The expression level of the AA9-4 gene varied across different carbon sources, reflecting their structural variability, ranging from simple sugars (glucose) to complex sugars (avicel). However, statistical analysis demonstrated that EF-1, 18 S-3, and GAPDH were the most reliable genes for data normalization, while ACT-3, BT-3, and 18 S-2 were the least stable genes (Fig. 7).

[IMAGE OMITTED: SEE PDF]

Discussions

T. fergusii has been used in gene expression studies; however, selecting suitable reference genes for this species remains unknown. The specificity of internal reference genes becomes crucial when considering T. fergusii for industrial applications in the biofuel industry. T. fergusii is known for its exceptional cellulose-degrading abilities, as filamentous fungi produce cellulases and other beneficial enzymes [1]. Industrial strains of Trichoderma are capable of producing significant quantities of external enzymes that play an important role in cellulose degradation [27]. Plant cell walls primarily consist of lignin, pectin, cellulose, and hemicellulose, with cellulases being a group of enzymes responsible for breaking down cellulose into glucose. In this study, we used various carbon sources such as avicel, cellobiose, and glucose, which play a significant role in gene expression. Under the influence of glucose, cellobiose, and avicel, the expression of the cellulase gene in Streptomyces halstedii JM8 was repressed by glucose but induced by cellobiose and avicel [28]. Myceliophthora thermophila strain JG207 was found to decrease malic acid production in the presence of cellobiose but significantly increase the expression of three intracellular β-glucosidase genes under avicel [29]. Previous research demonstrated that cellobiose or avicel enhances the production of lignocellulolytic genes in the model organism Neurospora crassa [30]. These carbon sources were therefore selected to evaluate the expression stability of reference genes in T. fergusii, with the subsequent goal of identifying suitable reference genes for RT-qPCR analysis under diverse growth conditions.

RT-qPCR is widely employed to investigate gene expression variations in different growth conditions and developmental stages. Accurate gene expression analysis requires appropriate reference genes, which are necessary to exhibit minimal variability across experimental treatments [6]. The reliability of gene expression analysis relies on selecting appropriate and stable reference genes, which should demonstrate consistent expression patterns over time and across different experimental conditions [31]. In this study, we evaluated the expression stability of twelve commonly used reference genes in T. fergusii. The amplification profiles revealed that average Cq values of five sequential dilutions on 1% glucose ranged from 19.75 to 30.02, with most genes maintaining Cq values below 35 even at high dilutions. Predominantly, EF-1 and GAPDH displayed consistently lower Cq values (≈ 21), indicating relatively higher transcript abundance (Fig. 2). The GAPDH Ct values were consistent with Flammulina filiformis at different developmental stages. In contrast, the EF-1 Ct values matched with Macrophomina phaseolina in various experimental conditions [32].

Utilizing a single Cq value approach is inadequate for screening reference genes. To address this limitation, we employed multiple software programs, including RefFinder, geNorm, NormFinder, BestKeeper, and Delta Ct, to evaluate the stability of candidate reference gene expression. Our findings demonstrate that different algorithms do not identify a universal reference gene suitable for normalizing target genes across various methods. Therefore, the identification of reference genes in T. fergusii must be determined for each specific experimental condition. In our results under control conditions (1% glucose), EF-1 and GAPDH consistently ranked among the top four most stable reference genes. A similar trend was observed under 1.5% cellobiose and 1.5% glucose treatments. Although minor variations in stability rankings were observed across different carbon sources, EF-1 and GAPDH were reliably classified among the most stable genes by all algorithms. 18 S-3 also demonstrated high expression stability, particularly under glucose and cellobiose treatments, however, its stability was less consistent under avicel treatment. Nevertheless, we consider that GAPDH, EF-1, and 18 S-3 exhibit excellent performance for gene normalization analysis. Similar variations in reference gene expressions have also been observed in studies on Macrophomina phaseolina during its interaction with a host and in other experimental studies [18, 32]. In contrast, ACT-3, BT-3, and 18 S-2 were repeatedly identified as the least stable reference genes under all tested conditions, according to each analytical method applied. Their high variability in expression patterns makes them unsuitable for normalization in T. fergusii, especially under fluctuating carbon source environments.

With advances in RT-qPCR technology, more studies have been conducted to validate suitable reference genes. Many studies use multiple reference genes and calculate the geometric mean of their relative expression levels for normalization purposes. This approach helps prevent any misinterpretation of gene expression data [33]. RefFinder analysis further supported these results by ranking EF-1, GAPDH, and 18 S-3 among the top five most stable genes under control, glucose, and cellobiose treatments (Fig. 5). In comparison, ACT-3, BT-3, and 18 S-2 were consistently placed among the least stable genes under the same experimental conditions. The use of geometric mean to integrate rankings across multiple algorithms confirmed the high stability of EF-1, GAPDH, and 18 S-3, while consistently identifying ACT-3, BT-3, and 18 S-2 as the unstable reference genes. These findings were further validated by RefFinder, Pearson correlation analysis, NormFinder, and the comparative ΔCt method (Fig. 6A, B, D; Table 1). Similarly, geNorm analysis identified EF-1 and GAPDH as the most stable gene pair, with 18 S-3 ranking slightly lower but remaining within acceptable stability limits. Collectively, the integrated analyses identify EF-1, GAPDH, and 18 S-3 as reliable reference genes for accurate normalization in gene expression studies of T. fergusii under varying carbon source conditions.

Many studies have identified various reference genes showing differing stability across diverse fungal species, growth stages, and experimental conditions, highlighting the necessity of validating reference genes for accurate RT-qPCR normalization in each specific condition. EF1a was identified as the most stable reference gene for gene expression analysis in Fusarium graminearum across various culture conditions [34]. Similarly, EF1 was chosen as the most stable reference gene for both roots and leaves of tomato plants colonized by Rhizophagus irregularis [35]. In Alternaria sp. MG1, EF1 also exhibited the highest expression stability for accurate RT-qPCR normalization, while 18 S rRNA was found to be the least stable [36]. Additionally, a comparative study among various filamentous fungi reported that β-tubulin (BT) and ACT did not show consistent stability and were therefore not recommended as reference genes [11]. In contrast, GAPDH was determined to be the most stable gene in Sclerotium rolfsii across various experimental treatments [37]. Other studies have found that GAPDH and 18 S were not reliable reference genes because highly expressed genes do not effectively represent the expression variations among genes with low expression levels [38, 39]. Similarly, the EF2 reference gene was stable in Puccinia helianthi Schw [8], and EF1 reference genes was stable during soybean root infection caused by Macrophomina phaseolina [32]. Our comprehensive analysis revealed that EF1, GAPDH, and 18 S-3 are the most stable reference genes for RT-qPCR data normalization in T. fergusii.

The expression of LPMOS AA9-4 genes was examined to evaluate the performance of selected reference genes. Our analysis demonstrated that normalization of AA9-4 expression using the least stable reference genes BT-3, ACT-3, and 18 S-2 led to substantial deviations from the expression levels obtained with more stable reference genes across different samples. Therefore, the normalization results derived from these genes were unreliable for assessing the expression level of AA9-4. However, when the expression levels of AA9-4 were normalized using stable reference genes EF1, GAPDH, and 18 S-3, their normalized expression remained stable across samples, even under varying experimental conditions. Thus, these three genes are suitable as reference genes for investigating gene expression in T. fergusii under our experimental conditions.

Conclusion

This study aims to identify suitable reference genes for precise normalization of RT-qPCR during spore and mycelial cultivation stages in T. fergusii under different carbon sources, including avicel, cellobiose, and glucose. A total of twelve reference genes from the T. fergusii genome were assessed. The genes EF1, GAPDH, and 18 S-3 demonstrated consistent expression levels and stability across various experimental conditions. This was determined using various computer algorithms, including RefFinder, geNorm, BestKeeper, NormFinder, and the ΔCT method. In contrast, BT-3, ACT-3, and 18 S-2 were found to be unreliable reference genes due to their instability. These findings provide important insights for accurately normalizing RT-qPCR data in gene expression studies involving T. fergusii.

Materials and methods

Cultivation condition for fungal strain

The thermophilic fungus Thermothelomyces fergusii CBS 405.69 was acquired from the Westerdijk Fungal Biodiversity Institute in the Netherlands for extensive study and expression analysis. The carbon sources, including avicel (AV), cellobiose (C), and glucose (G), were obtained from Sangon Biotech Co., Ltd. (Shanghai, China). Initially, T. fergusii was cultivated on PDA at 45°C for 3 days. Subsequently, the spores of the culture strain were filtered through a sterilized filter paper, and their concentration was adjusted to 7 × 107 spores/mL as previously described [40]. Then, the spores were cultured in a seed medium for 24 h at 45°C. The seed medium (total volume: 1 L) contained 100 mL Mendel’s nutrient salts [(NH₄)₂SO₄, urea, KH₂PO₄, MgSO₄·7H₂O], 1 mL trace elements (FeSO₄·7H₂O, ZnSO₄·7H₂O, MnSO₄·H₂O/MnCl₂), citrate buffer 50 mL (1 M, pH 4.5), tryptone (1 g), Tween-80 (1.5 mL), CaCl₂·2 H₂O (0.4 g), and glucose (20 g) [41]. Except for CaCl2.2H2O and glucose, all the reagents were mixed in distilled water on a magnetic shaker plate, and the volume was made up to 800 ml. CaCl2.2H2O and glucose were mixed in 200 mL distilled water with other reagents after sterilization at 115°C for 15 min. Five hundred microliters spores were used to inoculate the 50 ml of seed medium and incubated at 45°C with shaking at 180 rpm for one day. The liquid fermentation broth was prepared by adding Mendel’s nutrient salts (100 mL), citrate buffer (50 mL), tryptone (20 g), yeast extract (5 g), Tween-80 (0.5 mL), and CaCl₂·2 H₂O (0.4 g) [41]. After 24 h, 5 ml of grown mycelial solution was transferred to 50 ml of fermentation medium (without glucose) containing the 1.5% concentrations of each carbon source sugar (AV, C, and G) and incubated at 45°C with shaking at 150 rpm for five days. After five days, mycelia were collected by pouring the culture medium through 0.6 μm filter paper for RNA extraction. Mycelia obtained from seed media (24 h) before the induction was considered a control (reference) sample for RTqPCR analysis.

RNA extraction and cDNA synthesis

RNA extraction from the collected mycelia was performed using the TransZol Up technique, following the instructions provided by the manufacturer. RNA extraction was performed using approximately 100 mg of fresh mycelia of T. fergusii. The mycelia were initially cryopreserved in liquid nitrogen and pulverized using a pestle and mortar. The RNA was subsequently isolated using TransZol Up plus RNA kit (Cate. No. ER501, Quanshi Jinzhu Technology, Co., Ltd Beijing, China). RNA extraction protocols are freely available on the manufacturer’s website (https://www.transgenbiotech.com/rna/286.html). For RNA quality assessment, 1% agarose gel was prepared, and RNA samples were run on gel electrophoresis for 25 min in a freshly cleaned gel tank containing 1% TAE buffer. The gel was examined for the presence of distinct bands to verify the integrity and purity of the RNA. RNA concentrations and purity were further evaluated using a Nanodrop ND-1000 spectrophotometer (Nanodrop Technologies, Wilmington, DE, USA). The 260/280 ratio (2.13–2.15) was assessed to determine potential protein contamination, while the 260/230 (1.92–2.11) ratio was analyzed to check for residual ethanol or other chemical impurities. cDNA synthesis was performed using HiScript III SuperMix cDNA Synthesis Kit (Nanjing Vazyme Biotech Co., Ltd, China). A total cDNA reaction volume of 20 µL was prepared by combining 1 µL (1.0 µg) of total RNA, 4 µL of gDNA wiper mix (4×), and 11 µL of ddH2O in an RNA-free tube. The mixture was incubated at 42°C for 2 min in a PCR machine. Following this, 4 µL of HiScript III qRT SuperMix (5×) was added to the mixture, and the reaction was carried out at 37°C for 15 min, followed by a 5-second incubation at 85°C to terminate the reaction.

Protocol for primer selection and RT-qPCR analysis

Twelve reference genes were chosen as reference genes for real-time qPCR in the genome of T. fergusii (https://mycocosm.jgi.doe.gov/Thefe1/Thefe1.home.html). All the reference genes were selected based on the previous studies [5, 11, 12]. The reference genes included in this study were actin−1 (ACT-1), actin−2 (ACT−2), actin−3 (ACT-3), Glyceraldehyde-3-phosphate dehydrogenase (GAPDH), beta tubulin-1 (BT-1), beta tubulin−2 (BT-2), beta tubulin−3 (BT-3), elongation factor-1 (EF-1), elongation factor-2 (EF-2), elongation factor−3 (EF3), 18 S rRNA−1 (18 S-1), 18 S rRNA−2 (18 S-2), and 18 S rRNA−3 (18 S-3). Reference gene’s original IDs, ID renames, product sizes, primer sequence, and melting points are given here (Table 2).

All primers used were designed employing Integrated DNA Technologies (IDT: https://sg.idtdna.com/pages/products) and analyzed by the IDT program. The primer designing criteria included the following specifications: primer lengths ranging from 19 to 22 base pairs, GC contents between 50 and 60%, melting temperature (Tm) within the range of 55–60°C, and amplicon lengths between 100 and 150 bp. To ensure the primer specificity for T. fergusii, we employed conventional PCR to assess their amplification activity. The amplification accuracy for each pair of primers was assessed using the standard curve approach, employing five consecutive dilutions of cDNA on 1% glucose. The Relative Expression Software Tool (REST) was utilized to perform computations. All primer pairs exhibited amplification efficiency with primer efficiency r2 values ≥ 0.98, which meets the demands for RT-qPCR (Fig. 1). Moreover, the slope generated from the Cq values of the dilution was used to quantify the % primer efficiency (E) [42]. Most of the primer efficiencies were around 90–110%, which revealed the primer accuracy (Table 2).

[IMAGE OMITTED: SEE PDF]

The RT-qPCR was conducted in 96-well plates using a total reaction volume of 20 µL for each gene. The procedure was carried out as follows: The sample contained 10.5 µL Hieff UNICON® Universel Blue qPCR SYBR Green Master Mix (Yeasen Biotechnology Shanghai Co., Ltd, China), 0.4 µL of each forward and reverse primer at a final concentration of 2.5 µM, 1.2 µL of diluted cDNA (10 ng), and 7.5 µL of nuclease-free H2O. RT-qPCR was conducted on qTOWER3 G (Jena, Germany) employing SYBR® Green Pro-Taq-HS (AG, China). The experimental procedure comprised a 2-minute denaturation step at 95°C, followed by 40 amplification cycles consisting of a 10-second denaturation at 95°C, a 30-second annealing at 60°C, and a 1-minute extension at 60°C. The melting curve analysis was conducted within the temperature range of 60 to 95°C to confirm the specificity of the primers. Each sample was subjected to three biological replicates.

Statistical analysis and validation of reference genes

The expression of the twelve candidate genes was initially assessed based on the quantification cycle or cycle threshold (Cq/Ct) value [26]. The expression stability of the reference genes was evaluated using the RefFinder methods [22, 23], which included various statistical algorithms. The geNorm method was employed to calculate the stability of candidate reference genes based on pairwise variation and to assess their correlation [21]. The NormFinder method was used to evaluate the stability of reference genes by considering both intra- and inter-group variation, providing a stability value for each gene [24]. The Best-Keeper method was utilized to compute the standard deviation and coefficient of variation, ranking the reference genes based on their stability as determined from Ct values [25]. Finally, the Delta Ct method was used to assess the stability of reference genes by comparing the ΔCt for each candidate gene across all samples [26]. The Cq values obtained for each of the investigated genes were utilized to assess their stability using these five approaches, thus identifying the most suitable reference genes for data normalization in RT-qPCR studies. RefFinder was then employed to accurately evaluate and select the reference genes based on experimental data. The original Cq data (Supplementary Table S2) will be deposited in the RGeasy database after the manuscript publication [43]. The 2ˉ△△CT technique was applied to validate and quantify the expression of the target auxiliary activity 9 (AA9-4) gene (gene ID_3234) after the fungus was cultivated on 1.5% avicel, cellobiose, and glucose. The expression of AA9-4 under 1% glucose was used as a control for data analysis. Statistical analyses were conducted using a one-way analysis of variance (ANOVA), followed by Tukey’s Honest Significant Difference (HSD) test to assess pairwise differences among samples. Data analysis was conducted using Statistics V10 software, and graphical representation was finalized using GraphPad Prism.

Data availability

All data generated or analyzed during this study are included in this published article [and its supplementary information file S2]. The sequences of reference genes were assessed from the genome of T. fergusii, which is freely available in a public repository (https://mycocosm.jgi.doe.gov/Thefe1/Thefe1.home.html). However, the original Cq values of gene expression are available in a supplementary Table S2.

References

Rosenberg S. Cellulose and lignocellulose degradation by thermophilic and thermotolerant fungi. Mycologia. 1978;70(1):1–13.

Morgenstern I, Powlowski J, Ishmael N, Darmond C, Marqueteau S, Moisan M-C, Quenneville G, Tsang A. A molecular phylogeny of thermophilic fungi. Fungal Biology. 2012;116(4):489–502.

Chen Z, Halford NG, Liu C. Real-time quantitative PCR: primer design, reference gene selection, calculations and statistics. Metabolites. 2023;13(7):806.

Hendriks-Balk MC, Michel MC, Alewijnse AE. Pitfalls in the normalization of real-time polymerase chain reaction data. Basic Res Cardiol. 2007;102:195–7.

Chen C, Xie T, Ye S, Jensen AB. Selection of reference genes for expression analysis in the entomophthoralean fungus Pandora neoaphidis. Brazilian J Microbiol. 2016;47:259–65.

Thellin O, Zorzi W, Lakaye B, De Borman B, Coumans B, Hennen G, Grisar T, Igout A, Heinen E. Housekeeping genes as internal standards: use and limits. J Biotechnol. 1999;75(2–3):291–5.

Panina Y, Germond A, Masui S, Watanabe TM. Validation of common housekeeping genes as reference for qPCR gene expression analysis during iPS reprogramming process. Sci Rep. 2018;8(1):8716.

Song Y, Wang Y, Guo D, Jing L. Selection of reference genes for quantitative real-time PCR normalization in the plant pathogen Puccinia Helianthi Schw. BMC Plant Biol. 2019;19:1–10.

Sun Z-B, Li S-D, Sun M-H. Selection of reliable reference genes for gene expression studies in clonostachys rosea 67– 1 under sclerotial induction. J Microbiol Methods. 2015;114:62–5.

Yang Y, Xu X, Jing Z, Ye J, Li H, Li X, Shi L, Chen M, Wang T, Xie B. Genome-wide screening and stability verification of the robust internal control genes for RT-qPCR in filamentous fungi. J Fungi. 2022;8(9):952.

Llanos A, François JM, Parrou J-L. Tracking the best reference genes for RT-qPCR data normalization in filamentous fungi. BMC Genomics. 2015;16(1):1–18.

Tao Y, van Peer AF, Huang Q, Shao Y, Zhang L, Xie B, Jiang Y, Zhu J, Xie B. Identification of novel and robust internal control genes from Volvariella Volvacea that are suitable for RT-qPCR in filamentous fungi. Sci Rep. 2016;6(1):29236.

Lu Y-P, Liao J-H, Guo Z-J, Zheng H-Q, Lu L-F, Cai Z-X, Zeng Z-H, Ying Z-H, Chen M-Y. Selection and evaluation of potential reference genes for quantitative real-time PCR in agaricus Blazei based on transcriptome sequencing data. Biomed Res Int. 2021;2021:1–13.

Glare EM, Divjak M, Bailey M, Walters EH. β-Actin and GAPDH housekeeping gene expression in asthmatic airways is variable and not suitable for normalising mRNA levels. Thorax. 2002;57(9):765–70.

Ma L-G, Geng Y. Determination of the reference genes for qRT-PCR normalization and expression levels of MAT genes under various conditions in Ulocladium. PeerJ. 2020;8:e10379.

Flatschacher D, Eschlböck A, Zeilinger S. Identification and evaluation of suitable reference genes for RT-qPCR analyses in Trichoderma atroviride under varying light conditions. Fungal Biology Biotechnol. 2023;10(1):20.

Bisht A, Bhalla S, Kumar A, Kaur J, Garg N. Gene expression analysis for selection and validation of suitable housekeeping gene (s) in cadmium exposed Pigeonpea plants inoculated with arbuscular mycorrhizae. Plant Physiol Biochem. 2021;162:592–602.

Fu N, Li J, Wang M, Ren L, Zong S, Luo Y. Identification and validation of reference genes for gene expression analysis in different development stages of Amylostereum areolatum. Front Microbiol. 2022;12:827241.

He S, An T, Liu S. Validation of reliable reference genes for RT-qPCR studies of target gene expression in Colletotrichum camelliae during spore germination and mycelial growth and interaction with host plants. Front Microbiol. 2019;10:2055.

Archer M, Xu J. Current practices for reference gene selection in RT-qPCR of Aspergillus: outlook and recommendations for the future. Genes. 2021;12(7):960.

Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A, Speleman F. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. 2002;3(7):1–12.

Xie F, Xiao P, Chen D, Xu L, Zhang B. MiRDeepFinder: a MiRNA analysis tool for deep sequencing of plant small RNAs. Plant Mol Biol. 2012;80:75–84.

Xie F, Wang J, Zhang B. RefFinder: a web-based tool for comprehensively analyzing and identifying reference genes. Funct Integr Genom. 2023;23(2):125.

Andersen CL, Jensen JL, Ørntoft TF. Normalization of real-time quantitative reverse transcription-PCR data: a model-based variance Estimation approach to identify genes suited for normalization, applied to bladder and colon cancer data sets. Cancer Res. 2004;64(15):5245–50.

Pfaffl MW, Tichopad A, Prgomet C, Neuvians TP. Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper–Excel-based tool using pair-wise correlations. Biotechnol Lett. 2004;26:509–15.

Silver N, Best S, Jiang J, Thein SL. Selection of housekeeping genes for gene expression studies in human reticulocytes using real-time PCR. BMC Mol Biol. 2006;7(1):1–9.

Tiwari P, Misra B, Sangwan NS. β-Glucosidases from the fungus Trichoderma: an efficient cellulase machinery in biotechnological applications. BioMed Res Int. 2013;2013(1):203735.

Fernández-Abalos JM, Ruiz-Arribas A, Garda AL, Santamaría RI. Effect of carbon source on the expression of celA1, a cellulase-encoding gene from Streptomyces halstedii JM8. FEMS Microbiol Lett. 1997;153(1):97–103.

Li J, Gu S, Zhao Z, Chen B, Liu Q, Sun T, Sun W, Tian C. Dissecting cellobiose metabolic pathway and its application in biorefinery through consolidated bioprocessing in Myceliophthora thermophila. Fungal Biology Biotechnol. 2019;6:1–12.

Znameroski EA, Coradetti ST, Roche CM, Tsai JC, Iavarone AT, Cate JH, Glass NL. Induction of lignocellulose-degrading enzymes in Neurospora crassa by cellodextrins. Proc Natl Acad Sci. 2012;109(16):6012–7.

Teste M-A, Duquenne M, François JM, Parrou J-L. Validation of reference genes for quantitative expression analysis by real-time RT-PCR in Saccharomyces cerevisiae. BMC Mol Biol. 2009;10(1):1–15.

Orrego A, Gavilán MC, Arévalos A, Ortíz B, Gaete Humada B, Pineda-Fretez A, Romero-Rodríguez MC, Flores Giubi ME, Kohli MM, Iehisa JC. Identification of reference genes and their validation for gene expression analysis in phytopathogenic fungus Macrophomina Phaseolina. PLoS ONE. 2022;17(8):e0272603.

Nailis H, Coenye T, Van Nieuwerburgh F, Deforce D, Nelis HJ. Development and evaluation of different normalization strategies for gene expression studies in Candida albicans biofilms by real-time PCR. BMC Mol Biol. 2006;7(1):1–9.

Kim H-K, Yun S-H. Evaluation of potential reference genes for quantitative RT-PCR analysis in Fusarium graminearum under different culture conditions. Plant Pathol J. 2011;27(4):301–9.

Fuentes A, Ortiz J, Saavedra N, Salazar LA, Meneses C, Arriagada C. Reference gene selection for quantitative real-time PCR in Solanum lycopersicum L. inoculated with the mycorrhizal fungus Rhizophagus irregularis. Plant Physiol Biochem. 2016;101:124–31.

Che J-x, Shi J-l, Lu Y, Liu Y-l. Validation of reference genes for normalization of gene expression by qRT-PCR in a resveratrol-producing entophytic fungus (Alternaria sp. MG1). Amb Express. 2016;6:1–10.

Jiang C, Zhou L, Zhao Q, Wang M, Shen S, Zhao T, Cui K, He L. Selection and validation of reference genes for reverse-transcription quantitative PCR analysis in Sclerotium rolfsii. Int J Mol Sci. 2023;24(20):15198.

Xu Z, Xu J, Ji A, Zhu Y, Zhang X, Hu Y, Song J, Chen SJG. Genome-wide selection of superior reference genes for expression studies in Ganoderma lucidum. Genes. 2015;574(2):352–8.

Huggett J, Dheda K, Bustin S, Zumla AJG. Immunity: Real-time RT-PCR normalisation; strategies and considerations. Genes Immun. 2005;6(4):279–84.

Li X, Han C, Li W, Chen G, Wang L. Insights into the cellulose degradation mechanism of the thermophilic fungus Chaetomium thermophilum based on integrated functional omics. Biotechnol Biofuels. 2020;13:1–18.

Waheed A, Chen Y, Su Y, Yan Y, Liu G. Harnessing Xylanase potential in thermothelomyces fergusii: insights from computational and functional analysis. J Fungi. 2025;11(4):250.

Pfaffl MW. A new mathematical model for relative quantification in real-time RT–PCR. Nucleic Acids Res. 2001;29(9):e45–45.

de Souza MR, Araújo IP, da Costa Arruda W, Lima AA, Ságio SA, Chalfun-Junior A, Barreto HG. RGeasy: a reference gene analysis tool for gene expression studies via RT-qPCR. BMC Genomics. 2024;25(1):907.