Understanding the Role of Trichoderma reesei Vib1 in Gene Expression during Cellulose Degradation

Vib1, a member of the Ndt80/PhoG-like transcription factor family, has been shown to be essential for cellulase production of Trichoderma reesei. Here, we combined transcriptomic and genetic analyses to gain mechanistic insights into the roles of Vib1 during cellulose degradation. Our transcriptome analysis showed that the vib1 deletion caused 586 genes with decreased expression and 431 genes with increased expression on cellulose. The downregulated genes were enriched for Gene Ontology terms associated with carbohydrate metabolism, transmembrane transport, oxidoreductase activity, and transcription factor activity. Of the 258 genes induced by cellulose, 229 showed no or decreased expression in Δvib1 on cellulose, including almost all (hemi)cellulase genes, crucial sugar transporter genes (IDs:69957, 3405), and the genes encoding main transcriptional activators Xyr1 and Ace3. Additionally, Vib1 also regulated the expression of genes involved in secondary metabolism. Further comparison of the transcriptomes of Δvib1 and Δxyr1 in cellulose revealed that the genes regulated by Vib1 had much overlap with Xyr1 targets especially for the gene set induced by cellulose, presumably whose expression requires the cooperativity between Vib1 and Xyr1. Genetic evidence indicated that Vib1 regulates cellulase gene expression partially via Xyr1. Our results will provide new clues for strain improvement.


Introduction
Lignocellulosic biomass is the most abundant source readily available that presents an enormous potential for the production of biofuels and other bio-based products [1]. Bioconversion of cellulose substrate into soluble sugars through cellulolytic enzymes represents a green, sustainable strategy for biorefinery [2,3]. The filamentous fungus Trichoderma reesei is one of the most extraordinary producers of cellulases and hemicellulases, and it has become a paradigm in breakdown of cellulosic biomass [4]. However, since the recalcitrance of lignocellulosic biomass to enzymatic hydrolysis leads to a requirement for large amounts of lignocellulolytic enzymes, the high cost of enzyme, which ranges from $0.23 to $0.78 per gallon of ethanol [5], is still a major bottleneck in lignocellulosic biofuel production [6,7]. Genetic engineering of T. reesei for enhanced enzyme production is an efficient approach to lower the contribution of enzymes to biofuel production costs. However, the progress is very slow due to our incomplete knowledge of the molecular mechanisms underlying cellulase expression. A better understanding of the regulatory network controlling cellulase production in T. reesei is imperative and would assist to develop new biotechnologies for cheaper production of cellulases.
Vib1, a member of the Ndt80/PhoG-like transcription factor family that participates in the regulation of various metabolism processes, including meiosis [21], biofilm formation [22], and the response to nutrient stress [23], has recently been identified as a crucial regulator of cellulase production in T. reesei by two research groups with different means [17,18]. Vib-1, an orthologue of Vib1 in Neurospora crassa, is not only required for the expression of genes necessary for programmed cell death [24], but also essential for cellulose utilization [25]. It was proposed that the effect of N. crassa Vib-1 on cellulase production is achieved by inhibiting Cre1, and Col26, the BglR homologue of N. crassa, and inducing the positive master Clr-2 [25]. However, it is not clear which genes are subjected to regulation of the T. reesei Vib1 during cellulose degradation and the extent of overlap between the Vib1 regulon and the Xyr1 targets. In addition, it remains unknown whether Vib1 exerts its functions through Xyr1.
Here, to determine whether the requirement for vib1 varies with the respective inducers, we investigated the cellulase production of the ∆vib1 mutant under different carbon sources. To define the extent of the Vib1 regulon, we performed RNA-seq to assess the genome-wide gene expression differences between parent strain and ∆vib1 grown on cellulose. In addition, to determine the genes regulated by Vib1 and Xyr1, we compared the transcriptomes of ∆vib1 and ∆xyr1 and mainly focused on analyzing the expression level of the Avicel regulon, the gene set that had higher expression on cellulose than under either glucose or no carbon source conditions. Moreover, we introduced an additional copy of constitutive xyr1 expression cassette into ∆vib1 to test whether Vib1 exerts its functions through Xyr1.
For replacement experiments [26], T. reesei strains were precultured in MM with glucose as sole carbon source for 30 h. Pre-grown mycelia were collected by gauze filtration and washed twice with carbon-free MM. Equal amounts of mycelia were transferred to flasks containing the appropriate carbon source (Avicel cellulose, lactose, cellobiose, and glucose) or no carbon source added and continued cultivation. At indicated time points, the cultures were sampled and centrifuged at 12,000 rpm, for 10 min, at 4 • C. The culture supernatants were then used for cellulase activity and protein concentration assays, whereas the harvested mycelia were used for biomass determination or total RNA isolation.

Enzyme Activity and Protein Concentration Assays
The filter paper hydrolyzing activity (FPase) and endo-beta-1,4-glucanase (CMCase) activity in culture supernatants were measured according to the International Union of Pure and Applied Chemistry (IUPAC) standard [27]. The extracellular protein concentration was determined by the Bradford method (Sangon Biotech Co., Ltd., Shanghai, China).

RNA Isolation, Library Preparation, RNA Sequencing
Mycelia were sampled at 8 h after shifting to MM containing Avicel cellulose, glucose, or no carbon source. Total RNA was extracted with TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA), following the manufacturer's instructions. RNA integrity and purity were monitored on 1% agarose gels. Additional quality assessments were performed with the RNA Nano 6000 Assay Kit of the Bioanalyzer 2100 system (Agilent Technologies, Santa Clara, CA, USA) and the NanoPhotometer ® spectrophotometer (IM-PLEN, Los Angeles, CA, USA). RNA concentration was measured by using a Qubit ® RNA Assay Kit in a Qubit ® 2.0 Fluorometer (Life Technologies, Carlsbad, CA, USA).
The libraries were generated by using the NEBNext ® Ultra™ RNA Library Prep Kit for Illumina ® , following the manufacturer's recommendations, and subjected to sequencing on an Illumina HiSeq platform with paired-end reads. The RNA-seq raw data are available at the SRA web site (https://www.ncbi.nlm.nih.gov/sra, accessed on 28 March 2021), under accession number SRP312496.

RNA-Seq Data Analysis
Clean data were obtained by removing reads containing adapter, poly-N and lowquality reads from raw data through in-house Perl scripts. Then the clean reads were mapped to the transcripts from the reference genome (https://genome.jgi.doe.gov/Trire2 /Trire2.home.html, accessed on 26 January 2018), using Hisat2 v2.0.4. HTSeq v0.9.1 was used to count the reads numbers mapped to each gene and calculate FPKM (expected number of fragments per kilobase of transcript sequence per millions base pairs sequenced) [28]. Differential expression analysis of two conditions or groups was performed with the DESeq R package (1.18.0), with read counts as inputs. Genes with an adjusted p-value < 0.05 and |log 2 fold change| ≥ 1 were assigned as differentially expressed. Gene Ontology (GO) enrichment analysis of differentially expressed genes was implemented by the GOseq R package, in which gene length bias was corrected. GO terms with corrected p-value less than 0.05 were considered significantly enriched by differentially expressed genes.

Real-Time Quantitative PCR (RT-qPCR)
DNase I-treated total RNA was used to synthesize first-strand cDNA according to the Invitrogen Superscript III first-strand synthesis kit. Except for rpl6e, which was used as a reference gene [29], RT-qPCR assays were performed as described by Chen et al. [30].

Construction of the T. reesei Mutants
For the recyclable use of the marker gene pyr4 (ID 74020), a 3.7 kb pyr4 blaster cassette with a loopout fragment was constructed as described [31]. In brief, the primer pairs Fpyr4-DR/Rpyr4-DR and Fpyr4/Rpyr4 were designed to amplify direct repeats (1.0 kb upstream of the pyr4 start codon) and 2.3 kb of the pyr4 expression cassette that consisted of direct repeats, pyr4 coding regions, and pyr4 terminators, respectively. The two purified PCR products were digested with HindIII/NdeI and NdeI/BamHI and ligated with HindIII/BamH1-digested pBluescript SK-plus (Stratagene), resulting in the pPYR4 plasmid. The 3.7 kb pyr4 blaster cassette was released through the digestion of plasmid pPYR4 with HindIII/BamHI and used for subsequent plasmid construction.
The ∆vib1 mutant was generated by electroporating the uridine auxotrophic strain TU6_∆tku70 with the vib1 gene deletion cassette, which contains 1.8 kb of 5 and 3 flanking regions of the open reading frame of vib1 and a 3.7 kb of pyr4 blaster cassette. The ∆xyr1 mutant was obtained in the same manner, except that the 5 and 3 flanking regions of the xyr1 gene were employed.
The vib1ce strain, in which vib1 was constitutively expressed in ∆vib1, was constructed by transforming the uridine auxotrophy strain ∆vib1 with a DNA fragment containing A. nidulans glyceraldehyde-3-phosphate dehydrogenase (GenBank: M33539.1) promoter, the open reading frame and terminator region of the vib1 gene, and 2.7 kb pyr4 expression cassette and 1.8 kb of sequence downstream of the pyr4 stop codon used as homologous arms for targeted integration in the pyr4 locus. The xyr1 expression cassette targeting the pyr4 locus was constructed in the same way, except that the coding and terminator regions of xyr1 gene were used. The xyr1 expression cassette was transformed into ∆vib1 to obtain ∆vib1::xyr1 strain.
The protocol for electroporation of T. reesei was performed according to the method described by Schuster et al. [29], except that a BIO-RAD Gene Pulser Xcell electroporation system was used in our study. The PCR method was used to confirm targeted integration of gene deletion or expression cassettes [32]. The copy number of the cassette integrated in the genome was determined by quantitative real-time PCR on the genomic DNA, as previously described [33].
All primers used in this study are listed in Supplementary Materials Table S1.

Vib1 Is Required for Cellulolytic Enzyme Production Independent of Carbon Sources
The vib1 deletion mutant (∆vib1) failed to provoke cellulase formation under cultivation conditions supplemented with a mixture of Solka-Floc cellulose and lactose as carbon sources [17]. We wondered whether the vib1 deletion causes differential responses to the respective inducers; thus, we investigated the cellulase production of the ∆vib1 mutant under liquid MM with crystalline cellulose (Avicel PH-101, Sigma-Aldrich, St. Louis, MI, USA), lactose, or cellobiose as the sole carbon source. Irrespective of inducer strengths, ∆vib1 failed to produce extracellular protein on Avicel, lactose, or cellobiose ( Figure 1), which is consistent with the results that almost no cellulase activity indicated by filter paper activity and CMCase activity was detected in strains lacking vib1 ( Figure 1). The ∆vib1 mutant showed defect growth on Avicel but not on lactose and cellobiose (Supplementary Materials Figure S1). The ectopic integration of a copy of vib1 expression cassette driven by glyceraldehyde-3-phosphate dehydrogenase promoter from Aspergillus nidulans completely restored the growth of the ∆vib1 mutant on Avicel, extracellular protein production, and cellulase activity.

Comparative Transcriptome Analysis of Parent Strain and the ∆vib1 Mutant
To comprehensively understand the exact role of Vib1 in cellulose degradation, we utilized next-generation RNA sequencing to profile genome-wide mRNA abundances when the ∆vib1 mutant and parent strain were shifted to Avicel, glucose, or carbon-free medium for 8 h, following 30 h of pre-cultivation in glucose. RNA samples of three biological replicates from each condition were used for library preparation and sequencing, resulting in 15 sets of RNA-seq data. The data from three replicates of each condition showed a high Pearson correlation (Supplementary Materials Data S1).
For the parent strain, a comparative transcriptional profiling analysis between Avicel and no carbon source showed that 707 genes were differentially expressed, including 371 genes upregulated and 336 genes downregulated upon exposure to Avicel (Supplementary Materials Data S2). Only in the 371 upregulated gene set, there were enriched GO terms associated with hydrolase activity (corrected p-value: 1.33 × 10 −17 ), carbohydrate metabolism (corrected p-value: 1.15 × 10 −15 ), cellulose binding (corrected p-value: 1.14 × 10 −7 ), and polysaccharide binding (corrected p-value: 7.48 × 10 −7 ) (Supplementary Materials Figure S2). Of the 371 genes highly expressed on Avicel, 258 genes accumulated much more transcripts on Avicel than on glucose or no carbon (Supplementary Materials Figure S3), which were assigned as the "Avicel regulon" (Supplementary Materials Data S2). The Avicel regulon contained 18 genes involved in cellulose degradation, including major cellulase genes and nonenzymatic cellulose attacking protein encoding genes; 23 characterized or predicted hemicellulase genes; 14 genes encoding predicted proteins with signal peptides. Additionally, included in the Avicel regulon were 22 genes encoding sugar transporters, including recently identified lactose permease Ctr1 (ID 3405) [32,34], mannose/cellobiose/xylose transporter (ID 69957) [35]. Additionally, the genes especially induced by cellulose contained 16 transcription factors, including Xyr1, Ace3, AmyR, and N. crassa cellulase regulator Clr-2 orthologue (IDs: 122208, 77513, 55105 and 26163). Genes involved in protein folding and modification were also included in the Avicel regulon. According to our conservative differential expression analysis (adjusted p-value < 0.05 and |Log 2 fold change| ≥ 1), vib1(ID 54675) was not grouped into the Avicel regulon, but the mRNA level of vib1 was relatively higher on Avicel than either on no-carbon or glucose conditions.

Comparative Transcriptome Analysis of Parent Strain and the Δvib1 Mutant
To comprehensively understand the exact role of Vib1 in cellulose degradation, we utilized next-generation RNA sequencing to profile genome-wide mRNA abundances when the Δvib1 mutant and parent strain were shifted to Avicel, glucose, or carbon-free medium for 8 h, following 30 h of pre-cultivation in glucose. RNA samples of three biological replicates from each condition were used for library preparation and sequencing, resulting in 15 sets of RNA-seq data. The data from three replicates of each condition showed a high Pearson correlation (Supplementary Materials Data S1).
For the parent strain, a comparative transcriptional profiling analysis between Avicel and no carbon source showed that 707 genes were differentially expressed, including 371 genes upregulated and 336 genes downregulated upon exposure to Avicel (Supplementary Materials Data S2). Only in the 371 upregulated gene set, there were enriched GO terms associated with hydrolase activity (corrected p-value: 1.33 × 10 −17 ), carbohydrate metabolism (corrected p-value: 1.15 × 10 −15 ), cellulose binding (corrected p-value: 1.14 × 10 −7 ), and polysaccharide binding (corrected p-value: 7.48 × 10 −7 ) (Supplementary Materials Figure S2). Of the 371 genes highly expressed on Avicel, 258 genes accumulated much more transcripts on Avicel than on glucose or no carbon (Supplementary Materials Figure S3), Figure 1. The effect of Vib1 on extracellular protein production and cellulase activity under different carbon sources. Parent strain TU6 and the vib1 deletion mutant (∆vib1), together with the vib1 constitutive expression strain (vib1ce), were precultured in glucose-containing MM for 30 h; thereafter, the pre-grown mycelia were equally shifted to MM with Avicel cellulose, lactose, or cellobiose as the sole carbon source and continued cultivation for 120 h. The supernatants were collected for measuring extracellular protein content (A) and the volumetric cellulase activity (B-D). The values were means from three biological replicates. Error bars denoted standard deviations.
To gain insights into the extent to which Vib1 affects gene expression on Avicel, we compared the transcriptomes of the vib1 deletion and parent strains under Avicel cultivation condition, and the expression of the abovementioned 707 differentially expressed genes in ∆vib1 on Avicel vs. on no carbon. The absence of vib1 caused downregulation of 586 genes and upregulation of 431 genes in relative to parent strain on Avicel (Figure 2 and Supplementary Materials Data S3). The downregulated 586 genes were named as the "Vib1dependent gene set", which was enriched in functional categories, such as carbohydrate metabolism (corrected p-value: 1.36 × 10 −23 ), transmembrane transport (corrected p-value: 5.58 × 10 −19 ), hydrolase activity acting on glycosyl bonds (corrected p-value: 6.30 × 10 −29 ), oxidoreductase activity (corrected p-value: 4.27 × 10 −16 ), and RNA polymerase II transcription factor activity (corrected p-value: 7.15 × 10 −9 ) (Supplementary Materials Figure S4A). For the upregulated genes, the most enriched functional categories were single-organism metabolic process (corrected p-value: 1.47 × 10 −5 ), oxidation-reduction process (corrected p-value: 4.96 × 10 −10 ), and transmembrane transport (corrected p-value: 2.86 × 10 −5 ) (Supplementary Materials Figure S4B). For the 707 genes differentially expressed on Avicel and no carbon in the parent strain, the ∆vib1 mutant on Avicel showed almost the same expression profile as that on the no-carbon source (Figure 2 and Supplementary Materials Data S2), indicating that the ∆vib1 mutant failed to respond to the presence of cellulose in the environment, which supports the notion that sensing and responding to nutritional status is one of the ancestral roles for the Ndt80 family members [23].
and Supplementary Materials Data S3). The downregulated 586 genes were named as the "Vib1-dependent gene set", which was enriched in functional categories, such as carbohydrate metabolism (corrected p-value: 1.36 × 10 −23 ), transmembrane transport (corrected p-value: 5.58 × 10 −19 ), hydrolase activity acting on glycosyl bonds (corrected p-value: 6.30 × 10 −29 ), oxidoreductase activity (corrected p-value: 4.27 × 10 −16 ), and RNA polymerase II transcription factor activity (corrected p-value: 7.15 × 10 −9 ) (Supplementary Materials Figure S4A). For the upregulated genes, the most enriched functional categories were singleorganism metabolic process (corrected p-value: 1.47 × 10 −5 ), oxidation-reduction process (corrected p-value: 4.96 × 10 −10 ), and transmembrane transport (corrected p-value: 2.86 × 10 −5 ) (Supplementary Materials Figure S4B). For the 707 genes differentially expressed on Avicel and no carbon in the parent strain, the Δvib1 mutant on Avicel showed almost the same expression profile as that on the no-carbon source (Figure 2 and Supplementary Materials Data S2), indicating that the Δvib1 mutant failed to respond to the presence of cellulose in the environment, which supports the notion that sensing and responding to nutritional status is one of the ancestral roles for the Ndt80 family members [23].

Analyses of the Vib1-Regulated Genes
Of the 586 Vib1-dependent genes, about 89% (229 genes) of the Avicel regulon showed no or reduced gene expression on Avicel in ∆vib1 when compared to parent strain ( Figure 2 and Supplementary Materials Data S3), which comprise almost all (hemi)cellulosedegrading genes except cel3e within the Avicel regulon and 18 sugar transporter genes, including ID 69957 and ID 3405. Additionally, 13 out of 16 transcription-factor-encoding genes in the Avicel regulon were significantly downregulated in the ∆vib1 mutant, including xyr1, ace3, amyr, and the N. crassa clr-2 orthologue (Figure 2). Vib1 also had positive effects on the transcriptional levels of genes involved in protein folding and modification (IDs: 122415, 122920, 60085, and 73678). The remaining 29 genes within the Avicel regulon mainly encode hypothetical proteins. These data indicated that Vib1 is a crucial regulator of the Avicel regulon in T. reesei. Besides the Avicel regulon, Vib1 also influenced the expression levels of genes encoding the transcription factors AreA (ID: 76817) and the N. crassa Clr-1 orthologue (ID: 27600). It shall be noted that reduced expression of cre1 (log 2 fold change = −0.91818, p = 0.023) was observed in ∆vib1.

The Vib1-Dependent Gene Set Considerably Overlaps with Xyr1 Targets
For cellulase and extracellular protein production on cellulose, Vib1 demonstrated almost the same output as the cellulase transcriptional regulator Xyr1, suggesting that Vib1 may share regulatory targets with Xyr1. To test this hypothesis, we also performed transcriptome sequencing of the ∆xyr1 mutant on cellulose in parallel with the ∆vib1 mutant and parent strain. The deletion of xyr1 caused 2064 genes differentially expressed when compared with parent strain on Avicel, including 1218 downregulated genes and 846 upregulated genes (Supplementary Materials Data S4). The Vib1-dependent 586-gene set overlapped with Xyr1 targets by 371 genes. Of the 229 genes within the Avicel regulon which were downregulated in ∆vib1, 211 genes showed decreased or no expression in ∆xyr1 on Avicel (Figure 3).
The overlapped 211 genes included almost all (hemi)cellulose-degrading genes, except for cel3e, within the Avicel regulon; 14 sugar transporter genes, including ID 69957 and ID 3405; and 11 (putative) transcription factor-encoding genes, including ace3 and theclr−2 homologue (Table 2). Genes involved in posttranslational modification, intracellular trafficking, and secretion (e.g., ER chaperone Bip1, protein disulfide isomerase Pdi1, DnaJ superfamily molecular chaperone, COPII vesicle protein, ER-resident chaperone calnexin, alpha -1,2-mannosidase, alpha-mannosyltransferase, glucosidase II catalytic (alpha) sub-unit GII, and secretory pathway protein Ysy6) (Supplementary Materials Table S3) were also regulated in common by Vib1 and Xyr1. Of note, over 50% of 211 genes were affected by Vib1 and Xyr1 to a similar degree, suggesting that the cooperativity between Vib1 and Xyr1 might be required for the expression of these genes.

The Vib1-Dependent Gene Set Considerably Overlaps with Xyr1 Targets
For cellulase and extracellular protein production on cellulose, Vib1 demonstrated almost the same output as the cellulase transcriptional regulator Xyr1, suggesting that Vib1 may share regulatory targets with Xyr1. To test this hypothesis, we also performed transcriptome sequencing of the Δxyr1 mutant on cellulose in parallel with the Δvib1 mutant and parent strain. The deletion of xyr1 caused 2064 genes differentially expressed when compared with parent strain on Avicel, including 1218 downregulated genes and 846 upregulated genes (Supplementary Materials Data S4). The Vib1-dependent 586-gene set overlapped with Xyr1 targets by 371 genes. Of the 229 genes within the Avicel regulon which were downregulated in Δvib1, 211 genes showed decreased or no expression in Δxyr1 on Avicel (Figure 3). The overlapped 211 genes included almost all (hemi)cellulose-degrading genes, except for cel3e, within the Avicel regulon; 14 sugar transporter genes, including ID 69957 and ID 3405; and 11 (putative) transcription factor-encoding genes, including ace3 and theclr−2 homologue (Table 2). Genes involved in posttranslational modification, intracellular trafficking, and secretion (e.g., ER chaperone Bip1, protein disulfide isomerase Pdi1, DnaJ superfamily molecular chaperone, COPII vesicle protein, ER-resident chaperone calnexin, alpha -1,2-mannosidase, alpha-mannosyltransferase, glucosidase II catalytic (alpha) subunit GII, and secretory pathway protein Ysy6) (Supplementary Materials Table  S3) were also regulated in common by Vib1 and Xyr1. Of note, over 50% of 211 genes were affected by Vib1 and Xyr1 to a similar degree, suggesting that the cooperativity between Vib1 and Xyr1 might be required for the expression of these genes.    The result that the vib1 deletion significantly reduced the transcriptional level of xyr1 on cellulose prompted us to ask whether Vib1 acts through Xyr1 to regulate cellulase gene expression. If Xyr1 is subjected to the direct control of Vib1, the constitutive expression of xyr1 gene theoretically should bypass the requirement of Vib1. To test this hypothesis, we constructed strain ∆vib1::xyr1, in which an additional copy of the xyr1 constitutive expression cassette was integrated into the genome of ∆vib1. As shown in Figure 4, constitutive expression of xyr1 only partially restored the cellulolytic capacity of ∆vib1; however, the mRNA level of xyr1 in the ∆vib1::xyr1 mutant was higher than that in parent strain. These results suggested that Vib1 acts at least in part via Xyr1 to regulate cellulase biosynthesis; on the other hand, the exertion of the Xyr1 function likely depends on the involvement of Vib1 or other additional factors.
on cellulose prompted us to ask whether Vib1 acts through Xyr1 to regulate cellulase gene expression. If Xyr1 is subjected to the direct control of Vib1, the constitutive expression of xyr1 gene theoretically should bypass the requirement of Vib1. To test this hypothesis, we constructed strain Δvib1::xyr1, in which an additional copy of the xyr1 constitutive expression cassette was integrated into the genome of Δvib1. As shown in Figure 4, constitutive expression of xyr1 only partially restored the cellulolytic capacity of Δvib1; however, the mRNA level of xyr1 in the Δvib1::xyr1 mutant was higher than that in parent strain. These results suggested that Vib1 acts at least in part via Xyr1 to regulate cellulase biosynthesis; on the other hand, the exertion of the Xyr1 function likely depends on the involvement of Vib1 or other additional factors.

Discussion
In the present study, we demonstrated that the Δvib1 mutant abolished cellulase production irrespective of inducer strength and displayed defect growth on Avicel but not on lactose and cellobiose. By RNA-seq, we analyzed the genes regulated by Vib1 under Avicel cultivation conditions and identified the core gene set commonly regulated by Vib1

Discussion
In the present study, we demonstrated that the ∆vib1 mutant abolished cellulase production irrespective of inducer strength and displayed defect growth on Avicel but not on lactose and cellobiose. By RNA-seq, we analyzed the genes regulated by Vib1 under Avicel cultivation conditions and identified the core gene set commonly regulated by Vib1 and Xyr1. By genetic analysis, we showed that Vib1 controls cellulase expression partially via Xyr1.
Our transcriptome comparison between parent strain and ∆vib1 showed that the vib1 deletion significantly reduced the expression level of xyr1, ace3, cre1, and the N. crassa clr-1 and clr-2 homologues, but exerted no effect on the mRNA level of bglR on Avicel. Similarly, the expression of bglR on lactose was also not affected by Vib1 [17]. This is different from the case in N. crassa, where functional loss of N. crassa Vib1 resulted in significantly reduced mRNA levels of the transcription factor Clr2 essential for cellulase induction [42] but increased transcriptional levels of the genes encoding carbon catabolite repressor Cre1 and the T. reesei BglR orthologue Col-26 critical for glucose sensing/metabolism [25]. The reason for this phenomenon is the difference of transcription machinery between T. reesei and N. crassa [43]. Therefore, despite functional conservation in cellulase production, the regulatory circuit of Vib1 was rewired in T. reesei.
Previous studies revealed that the transcription factor Ace3 is essential for the expression of cellulase genes [14], which governs cellulase activity through Xyr1 and the cellulose response transporter Crt1 [34]. Here, we showed that the deletion of vib1 or xyr1 significantly reduced the transcript accumulation of ace3. Meanwhile, Vib1 and Xyr1 positively regulated each other irrespective of the extent. Additionally, by analyzing the gene expression pattern of the Avicel regulon in ∆vib1 and ∆xyr1 on cellulose, we found that there was considerable overlap between the genes regulated by Vib1 and Xyr1. Taken together, it is possible that Vib1, Xyr1, and Ace3 constitute a core regulatory circuit and may depend on each other to achieve individual functions. Alternatively, the effect of Vib1 on cellulase expression may be achieved through activating Ace3, thus facilitating the interaction of Ace3 with Xyr1 or other factors.
There are several lines of evidence pointing to the crosstalk between plant cell wall deconstruction enzyme expression and secondary metabolism biosynthesis in T. reesei [37,[44][45][46][47][48], which involves methyltransferase Lae1, YPK1-type kinase Usk1, catalytic subunit of protein kinase A (PKAc1), and transcription factors (Vel1, Cre1, Xpp1, and Ypr2). It has been demonstrated that Cre1, Usk1, PKAc1, and Xpp1 control the expression of the SOR cluster, including the transcription factors Ypr1 and Ypr2 [37,45,47,48]. Our results indicated that, besides cellulase biosynthesis, Vib1 controls the expression of the SOR gene cluster; however, Vib1 had no effect on the mRNA levels of Usk1, PKAc1, and Xpp1, indicating that Vib1 might act downstream of the three regulators. This requires further investigation. In addition to the SOR cluster, we found that VIB1 also regulated the expression of three PKS genes, four NRPS genes, one NRPS/PKS fusion gene, and the global secondary metabolism regulator Vel1. These findings suggest that Vib1 might exert a broad influence on secondary metabolism in T. reesei.

Conclusions
This study showed that Vib1 was involved in cellulose degradation through the control of multiple gene expression beyond carbohydrate metabolism. The set of 586 Vib1dependent genes overlapped with Xyr1 targets by 371 genes, including 211 genes within the Avicel regulon. The overlapped 211 genes included almost all (hemi)cellulose-degrading genes, except for cel3e; and characterized crucial sugar transporter-and transcription factorencoding genes, which may constitute the backbone of the whole cellulose degradation system. We propose that their expression may be dependent on the cooperativity between Vib1 and Xyr1. The result that constitutive expression of xyr1 only partially restored the cellulase production ability of the ∆vib1 mutant, to some extent, showed that the implementation of the Xyr1 function required the involvement. Further research is needed to test this hypothesis. Taken together, the present work has revealed new aspects of cellulase expression regulation in T. reesei. Such knowledge has important implications for the improvement of cellulase production by T. reesei.

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/ 10.3390/jof7080613/s1. Figure S1: Mycelial biomass accumulation of the vib1 deletion strain TU6_∆tku70∆vib1 (∆vib1) relative to the parent strain TU6_∆tku70 (TU6) and the vib1 constitutive expression strain vib1ce. Figure S2: Genome-wide analysis of transcriptional induction by Avicel cellulose in T. reesei. Figure S3: Hierarchical cluster of expression levels in TU6 on glucose, no-carbon, and Avicel cellulose for 707 genes differentially expressed on Avicel cellulose and no carbon 8 h after transfer. Figure S4. Gene Ontology (GO) enrichment analysis of genes differentially expressed in ∆vib1 on Avicel as compared to TU6 on Avicel 8 h after transfer from glucose. The most enriched GO terms for downregulated genes (A) and upregulated genes (B) in the ∆vib1 mutant. The GO items with adjusted p-value < 0.05 are indicated with the symbol "*". Data S1: A summary of transcriptome data. Data S2: The 707 differentially expressed genes in TU6 exposed to Avicel in relative to no carbon and their expression level and fold change in ∆vib1. Data S3: The genes regulated by VIB1. Data S4: Gene list regulated by XYR1 on Avicel. Table S1: Oligonucleotides used in this study.