A novel transcription factor specifically regulates GH11 xylanase genes in Trichoderma reesei

Background The filamentous fungus Trichoderma reesei is widely utilized in industry for cellulase production, but its xylanase activity must be improved to enhance the accessibility of lignocellulose to cellulases. Several transcription factors play important roles in this progress; however, nearly all the reported transcription factors typically target both cellulase and hemi-cellulase genes. Specific xylanase transcription factor would be useful to regulate xylanase activity directly. Results In this study, a novel zinc binuclear cluster transcription factor (jgi|Trire2|123881) was found to repress xylanase activity, but not cellulase activity, and was designated as SxlR (specialized xylanase regulator). Further investigations using real-time PCR and an electrophoretic mobility shift assay demonstrated that SxlR might bind the promoters of GH11 xylanase genes (xyn1, xyn2, and xyn5), but not those of GH10 (xyn3) and GH30 (xyn4) xylanase genes, and thus regulate their transcription and expression directly. We also identified the binding consensus sequence of SxlR as 5′- CATCSGSWCWMSA-3′. The deletion of SxlR in T. reesei RUT-C30 to generate the mutant ∆sxlr strain resulted in higher xylanase activity as well as higher hydrolytic efficiency on pretreated rice straw. Conclusions Our study characterizes a novel specific transcriptional repressor of GH11 xylanase genes, which adds to our understanding of the regulatory system for the synthesis and secretion of cellulase and hemi-cellulase in T. reesei. The deletion of SxlR may also help to improve the hydrolytic efficiency of T. reesei for lignocellulose degradation by increasing the xylanase-to-cellulase ratio. Electronic supplementary material The online version of this article (doi:10.1186/s13068-017-0878-x) contains supplementary material, which is available to authorized users.


Background
Lignocellulosic biomass, consisting mostly of cellulose, hemi-cellulose, and lignin, is the most abundant and renewable energy source on earth [1]. Degradation of lignocellulosic biomass and continuation of the carbon cycle in nature is maintained mainly by microbial action, including different fungal species, such as Trichoderma, Aspergillus, and Penicillium. The biomass-degrading enzymes produced by these organisms also have applications in various fields of industry including food, fodder, paper, and textile industries. Trichoderma reesei is a wellknown efficient producer of cellulase and hemi-cellulase, and is therefore widely employed by the enzyme industry for production of its own endogenous enzymes as well as production of heterogeneous proteins [2].
Cellulosic biofuel production continues to increase world-wide every year with a consequent rapidly increasing requirement for cellulase production from T. reesei. However, the production and optimization of enzyme formulations for lignocellulose degradation are still the major barriers to its extensive application. It is necessary to further enhance the hydrolytic efficiency of T. reesei enzyme preparations for lignocellulose and lower their relatively high cost. Due to the complex constitution and structure of native lignocellulose, the enzyme preparations produced by T. reesei must be supplemented with several types of exogenous enzymes to achieve effective degradation of natural complex lignocellulosic materials. For example, exogenous hemi-cellulase and other auxiliary enzymes are added to commercial cellulase complexes from Danisco or Novozymes [3]. This indicates that increasing the production of the most prominent hemi-cellulase (endo-β-1,4-xylanase), which catalyzes the hydrolysis of 1,4-β-d-xylosidic linkages in xylan to short xylooligosaccharides of varying length in T. reesei, is a good way to strengthen its hydrolysis activity [4].
Expression of genes encoding cellulase and hemi-cellulase in T. reesei is tightly controlled at the transcriptional level. Therefore, deleting and/or over-expressing transcription factors (TFs) that specifically regulate xylanase gene expression in T. reesei is a straightforward way to perform knowledge-based strain design. However, most known TFs regulate expression of both cellulase and xylanase genes in the same way. The most extensively studied TF is the negative TF CRE1 [5], which mediates carbon catabolite repression (CCR). In the cellulase hyperproduction strain T. reesei RUT-C30, CRE1 is truncated, which renders this strain carbon catabolite depressed [6]. The global transcriptional activator Xyr1 is obligatory for expression of most cellulase and hemi-cellulase genes [7,8]. Other recognized TFs are the positively acting factor ACE2 [9,10] and HAP2/3/5 complex [11], and the negatively acting factor ACE1 [12]. In addition, BglR [13] was identified as a new TF that upregulates expression of specific genes encoding β-glucosidases in T. reesei, and a putative methyltransferase, LAE1 [14,15], was found to be essential for cellulase and hemi-cellulase production in T. reesei, although the mechanism is still unclear. ACE3 has been newly characterized to be indispensable for cellulase and xylanase activity in T. reesei [16]. All the above TFs have effects on the expression of both cellulase and hemi-cellulase genes.
Recently, a study found that some segmentally aneuploid (SAN) T. reesei strains exhibited enhanced growth in xylan-based media and produced higher levels of xylanase [17]. Further analysis confirmed that D segment duplication, not L segment deletion, in the genomes of these SAN strains, was responsible for the growth advantage in xylan-based media, but did not affect the cellulase activity. In fact, the Xpp1, TF, was identified by a pull-down assay based on the xyn2 promoter and was confirmed as a negative regulator of the xyn1 and xyn2 xylanase genes and the bxl2 β-xylosidase gene [18].
In this study, using bioinformatics analysis and gene deletion with the CRISPR/Cas9 system [19], we identified a gene encoding a protein designated as specialized xylanase regulator (SxlR). Deletion of the sxlr gene resulted in increased xylanase activity while not affecting cellulase activity. This indicated that SxlR might be a novel TF that regulates xylanase expression. According to the results of real-time PCR (qPCR) and electrophoretic mobility shift assay (EMSA) analyses, we demonstrated that SxlR plays a critical role in the inhibition of the expression of xylanases belonging to the glycoside hydrolase 11 family (GH11) in T. reesei through binding to the promoter regions of target genes directly. Finally, we identified the binding consensus sequence of SxlR as 5′-CATCSGSWCWMSA-3′.

Screening the putative xylanase-specific TFs
Based on bioinformatics analysis, we chose seven putative TFs as candidates (Additional file 1) for screening. All these genes were located in five different chromosomes [20,21], and six of candidates were located in the D segment duplication resulting in higher xylanase activities as reported by Chen et al. [17]. We overexpressed these genes in T. reesei RUT-C30. After creating monoconidial cultures for genetic stability, we measured the xylanase and cellulase activities of the transformants. The xylanase activity, but not the cellulase activity, of the strain overexpressing the sxlr gene (jgi|TrireRUTC30_1|26638, Osxlr) decreased significantly (t test, P < 0.05) (Fig. 1a, b), and its extracellular protein concentration was decreased (Fig. 1c), which might indicate that this gene affects the xylanase activity alone. In contrast, no changes in xylanase activity, cellulase activity, or protein concentration were detected between RUT-C30 and the other six overexpression strains (Fig. 1).
To investigate how the sxlr gene is involved in regulation of xylanase activity in T. reesei, we deleted sxlr gene from the RUT-C30 strain to obtain Δsxlr transformants ( Fig. 2a, b). In contrast to the Osxlr strain, the xylanase activity of the Δsxlr strains was increased significantly (Fig. 2c). However, the cellulase activity of the Δsxlr strains was almost the same as the parent RUT-C30 strain. We also deleted the homologous sxlr gene of T. reesei Qm6a (jgi|Trire2|123881, Δ6a-sxlr) using the CRISPR/Cas9 system to verify the common regulation of xylanase activity in T. reesei strains. As expected, the xylanase activity of Δ6a-sxlr transformants was also dramatically increased, similar to the Δsxlr strain (Fig. 2c), while its cellulase activity was not affected. Taken together, these results suggested that sxlr encoded a TF that negatively regulates xylanase activity.

The deletion of SxlR in T. reesei RUT-C30 results in higher xylanase activity and higher reducing sugar yield
We selected the cellulase hyperproduction strain RUT-C30 to explore the potential regulatory mechanism of SxlR in T. reesei. In addition to the sxlr deletion strain (Δsxlr, transformant 1) and the sxlr overexpression strain (Osxlr), we also constructed the in situ re-complementation strain (Rsxlr) based on the Δsxlr strain (Additional file 2). The xylanase activity of the Δsxlr strain increased relative to that of RUT-C30 by 0.7-fold and 1.4-fold after 3 and 7 days, respectively, of incubation in inducing medium containing wheat bran and Avicel (Fig. 3a). The xylanase activity was also examined when xylan or lactose was used as the inducer (the sole carbon source in the inducing medium). The xylanase activity of Δsxlr was 1.3-, 0.9-, and 0.7-fold higher than RUT-C30 after 1, 2, and 3 days, respectively, of cultivation with xylan as the inducer (Fig. 3b), and 14.2-, 4.7-, and 5.2-fold higher, respectively, with lactose as the inducer (Fig. 3c). The xylanase activity of the sxlr re-complementation control strain Rsxlr reverted to the same level as that of RUT-C30. In contrast, the sxlr overexpression strain Osxlr demonstrated weaker xylanase activity than RUT-C30 ( Fig. 3a-c). However, no significant difference in cellulase activity was detected between the four strains (Fig. 3d).
Degradation of lignocellulose is a growth-associated process. The growth rate affects the secretion of enzymes directly [22]. We observed the growth of RUT-C30 and its derivative strains on potato dextrose agar (PDA) and minimal medium (MM) containing 1% xylose, xylan, glucose, lactose, or Avicel as carbon sources, respectively (Fig. 4a). RUT-C30 and its derivative strains did not show significantly different growths and sporulations on PDA and MM containing 1% glucose, lactose or Avicel (Fig. 4a). However, the Δsxlr strain demonstrated rapid growth and the Osxlr strain demonstrated reduced growth on MM containing xylose or xylan when compared to RUT-C30. These results could be explained by differences in xylanase activities.
Enzyme activities are also influenced by the concentration and composition of lignocellulose preparations. The extracellular protein concentrations of the Δsxlr, Osxlr, and RUT-C30 strains were 15.275 ± 0.096, 10.943 ± 0.399, and 13.378 ± 0.240 mg/mL, respectively. The Δsxlr strain had the highest extracellular protein concentration, and Osxlr had the lowest (Fig. 4b).
According to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), the band corresponding to ~20 kDa, which was identified as the xylanase protein XYN2 by MALDI-TOF/TOF, was enhanced dramatically in the Δsxlr strain but decreased in the Osxlr strain ( Fig. 4c). This indicates that SxlR affects the amount of secreted XYN2 in the lignocellulose preparations produced by T. reesei.
To determine whether the deletion of SxlR in T. reesei RUT-C30 could improve the strain's efficiency to hydrolyze pretreated lignocellulose, steam-exploded rice straw and steam-exploded rice straw mixed with corn straw were used as saccharification substrates. The crude enzyme complex of the mutant Δsxlr strain produced more reducing sugar than that of RUT-C30 (Fig. 4d). The straw hydrolysis increased to 21% by the supernatant of the Δsxlr strain, while 14.1% by the supernatant of RUT-C30 strain after 3 days of hydrolysis from the pretreated rice and corn straw. Similarly, the straw hydrolysis increased to 11.9% by the supernatant of the Δsxlr strain, while 7.6% by the supernatant of RUT-C30 strain after 3 days of hydrolysis from the pretreated rice straw.

SxlR regulates the GH11 xylanase genes
To further identify the regulation mechanism of SxlR on xylanase activity, we examined the expression levels of five xylanase-encoding genes using qPCR; the xylanases encoded by these genes included three members of GH11 family (XYN1, XYN2, and newly discovered XYN5), a GH10 family member XYN3 and a GH30 family member XYN4 [23]. The transcription levels of the genes encoding the three GH11 members showed significant differences in the Δsxlr strain (upregulated) and the Osxlr strain (downregulated) at all sampling time points using wheat bran and Avicel were used as inducers (Fig. 5a).
The transcription level of the genes encoding XYN3 and XYN4 were also downregulated in the Osxlr strain at all sampling time points (Fig. 5a). However, their transcription levels in the Δsxlr strain changed differently (Fig. 5a).
In the Δsxlr strain, the relative expression level of xyn3 was upregulated compared to the wild-type stain in the first 8 h of incubation in the inducing medium. The relative expression level of xyn4 was upregulated after 4 h of induction, but downregulated after 8 and 12 h of induction (Fig. 5a). When using xylan as the inducer, the transcription level variation tendency of the encoding genes of XYN1, XYN2, XYN5, and XYN3 was similar to that using wheat bran and Avicel as the inducer (Fig. 5b). In contrast, the relative expression level of xyn4 was upregulated in the Osxlr strain after 4 h of induction (Fig. 5b).
To further confirm the qRT-PCR results, qRT-PCR experiments were re-carried out using rpl6e (a ribosomal protein encoding gene) [24] and sar1 (a small GTPase encoding gene) [25] as the reference gene. The transcription level variation tendency of the five xylanase-encoding genes was similar to that using β-actin gene as the reference gene (Additional file 3).
Additionally, we examined the relative expression levels of other genes coding (hemi-) cellulases or their regulators, including three major cellulase genes (cbh1, cbh2, egl1), the key transcriptional activator (xyr1), and two hemi-cellulase genes (β-mannanase, man1 and α-l-arabinofuranosidase, abf1). There were no significant differences in the relative expression levels of these genes between RUT-C30 strain and Δsxlr strain (Additional file 3). These results provide the first experimental evidence that SxlR is a negative and specific regulator of GH11 family xylanases.
Xyr1, a major transcription activator in T. reesei, regulates expression of most cellulase and hemi-cellulase genes directly, including xylanases [26,27]. To investigate the relationship between Xyr1 and SxlR, we determined the relative expression levels of sxlr in the Δxyr1 strain (an xyr1-deletion strain derived from RUT-C30) and in the Oxyr1 strain (an xyr1-overexpression strain derived from RUT-C30) as well as the relative expression levels of xyr1 in Δsxlr strain and Osxlr strain (Fig. 5c). Comparing with xyr1, the transcription level of sxlr was quite low in RUT-C30. The xyr1 transcription level decreased significantly in the Osxlr strain at all the sampling time points. In contrast, its transcription level in the Δsxlr strain significantly increased at 4-h induction, and then decreased from 8-h induction. It seemed that SxlR might repress xyr1 expression somehow. The transcription level of sxlr in the Oxyr1 strain was nearly the same as that in RUT-C30 strain, and its transcription level in the Δxyr1 strain was lower or similar to that in RUT-C30 strain. The variation tendency of the sxlr transcription level was quite different from that of the downstream genes of Xyr1, of which the transcription were sharply repressed after the deletion of Xyr1 [27].

SxlR binds the promoters of GH11 xylanase genes
According to functional domain analysis, SxlR contains two distinct conserved domains, one GAL4-like Zn 2 Cys 6 binuclear cluster DNA binding domain at the N-terminus (cd00067: residues 282-317) and one fungal transcription factor regulatory middle homology region at the C-terminus (cd12148: residues 436-853).
To verify the regulation of the GH11 xylanase genes by SxlR, the GAL4-like Zn 2 Cys 6 binuclear cluster DNA binding domain was expressed in vitro for EMSAs. The nearly 1500-bp upstream regions (nucleotide position −1500 to −1) assumed to be the promoter regions of the xylanase-encoding genes were divided into six parts (for example, xyn2-P1, 268 bp, nucleotide position −268 to Based on the EMSA gel shifts, SxlR could bind the xyn1-P5, xyn2-P4, xyn2-P5, and xyn5-P5 promoter regions of the three GH11 xylanases (Fig. 6a). No specific gel shift was observed for the xyn3 and xyn4 promoters, which belonged to the GH10 and GH30 families, respectively ( Fig. 5b; Additional file 4). It means that SxlR plays a critical role in the inhibition of GH11 family genes through binding to their promoter regions directly.
To identify the binding motif of SxlR, three promoter fragments (xyn2-P4, xyn1-P5, and xyn5-P5) were each divided into two 144-bp segments (Additional file 5A). SxlR bound to xyn2-P4-1, xyn1-P5-2, and xyn5-P5-2 (Additional file 5B). Based on MEME Suite (http:// meme-suite.org/tools/meme) analysis, three candidate consensus motifs were predicted (Additional file 5C, D). However, the deletion of these three motifs did not affect the binding of SxlR to these DNA fragments (Additional file 5E). The 144 -bp fragments were further shortened for identification of a consensus motif, and two motifs were predicted (Fig. 7a). Based on the disappearance of the SxlR-DNA complex with the deletion of motif 5 (Fig. 7b), the consensus motif of SxlR was determined as 5′-CATCSGSWCWMSA-3′ (Fig. 7c, d). In this motif, one guanine and one cytosine are present in the core region and the consistent nucleotides are located in the flanks. The potential SxlR binding motif were not detected in the promoter regions of Xyr1 and xylanases XYN3 and XYN4. It suggested that SxlR might not regulate the transcription of Xyr1 and the GH10 family member XYN3 as well as the GH30 family member XYN4 directly.

Discussion
The aim of this study was to gain a comprehensive understanding of how viable SAN progeny produces higher levels of xylanases [17]. We focused on candidate TFs exhibiting differential expression levels between SAN and euploid progeny (six TF-encoding genes: ID in jgi|Trire2: 106677, 65854, 111446, 68930, 111515, 36913), as well as another candidate TF, SxlR. We confirmed that SxlR is a negative regulator of GH11 family xylanases. However, overexpression of the other six tested TFs had no effect on the xylanase activity. It seems that the segmental aneuploidy did not result in a change in sxlr gene copy number; however, it is still unknown whether segmental aneuploidy affects the expression of sxlr in SAN progeny.
Recently, another negative regulator of hemi-cellulase, xylanase promoter-binding protein 1 (Xpp1) was reported in T. reesei [18]. Xpp1 regulates transcription of hemi-cellulase genes only at later stages of cultivation. There was no significant difference in xylanase activity between an xpp1-disrupted strain and its parent strain before 72 h [18]. In this study, SxlR regulated expression of GH11 xylanase genes during the entire induction period, and the xylanase activity of the deletion strain was significantly higher than that of the parent strain (Fig. 3a). It has been suggested that SxlR may play an important role in regulating xylanase expression in T. reesei. Xpp1 repressed the expression of xyn1, xyn2, and bxl2 (encoding a putative β-xylosidase) but like SxlR, did not affect cellulase genes [18]. SxlR bound the xyn1, xyn2, and xyn5 promoter regions in vitro (Fig. 6a) and was demonstrated to be a GH11-specific TF. This indicated that the regulation mechanisms of SxlR and Xpp1 are not identical. Additionally, in inducing medium containing xylan, the xpp1-disrupted strain showed nearly 2.2-fold increases in transcription levels of xyn2 at 72 h, while the Δsxlr strain showed 9.9-, 2.1-, and 4.6-fold increases in transcription levels of xyn2 at 4, 8, and 12 h, respectively. Both the intensity and timing of these two TFs differed.
The binding consensus sequence of SxlR (5′-CATC-SGSWCWMSA-3′) differs from that of the hemicellulase regulator Xpp1 (a hexameric palindrome 5′-WCTAGW-3′ together with an inverted AGAArepeat [18]). In the xyn1 promoter region, the binding consensus sequence of Xpp1 is located in the xyn1-P3 fragment, while SxlR binds to xyn1-P5. In the xyn2 promoter region, the binding consensus sequence of Xpp1 is located in xyn2-P1 fragment, while SxlR binds to xyn2-P4 and xyn2-P5. In the xyn5 promoter region, only the SxlR binding consensus sequence was found. Meanwhile, in the bxl2 promoter region, only the Xpp1 binding consensus sequence was found. Therefore, SxlR has a regulon that differs from that of Xpp1.
Most xylanases (endo-β-1, 4-xylanase, EC 3. 2. 1. 8) belong to the GH10 and GH11 families. Compared to GH10, the GH11 family is much more xylan-specific [28]. In T. reesei, XYN2 belonging to GH11 family was the dominant extracellular xylanase. The Δsxlr strain grew Fig. 6 EMSAs of SxlR binding to the promoter regions of xylanase genes. a DNA binding of SxlR to the promoter regions of xyn1, xyn2, and xyn5. The amounts of purified SxlR binding domain (SxlR-B, μM) used were as indicated; ~10 ng of Cy5-labeled probe was added to each reaction. The shifts were verified to be specific by adding 100-fold excess of unlabeled specific (S) and non-specific (NS) competitor DNA. The SxlR-DNA complex is indicated by the arrow. b DNA binding of SxlR to the xyn3 and xyn4 promoter regions. We used three concentrations of SxlR-B: 0, 0.5, and 1 μM; ~10 ng of Cy5-labeled probe was added to each reaction more vigorously than the wild-type on MM with xylan or xylose as the sole carbon source. This indicates that the deletion of sxlr results in the high ability to utilize xylan or xylose. According to our hypothesis, the role of SxlR is the downregulation of the main xylanases to save energy when cellulose or glucose is present in the habitat of Fig. 7 Identification of the SxlR binding consensus sequence. a Sequence motifs of a putative SxlR binding consensus sequence derived by MEME from Dxyn2-P4-1, Dxyn1-P5-2 and Dxyn5-P5-2. Two putative SxlR binding consensus sequences were obtained. b EMSA results of SxlR binding to Dxyn2 P4-1, D4 (Motif 4 deletion), and D5 (Motif 5 deletion); the SxlR-DNA complex is indicated by the arrow. The amounts of purified SxlR binding domain (SxlR-B, μM) used were as indicated; ~10 ng of Cy5-labeled probe was added to each reaction. c Alignment of SxlR binding consensus sequence on sense (+) or antisense (−) strand in the upstream regions of xyn1, xyn2, and xyn5. The numbers following the gene name indicated the point of the 5′ starting nucleotide relative to the translation start point and the same nucleotide was indicated by asterisk. d The location of motif 4 and motif 5 in Dxyn2 P4-1 probe. Motif 4 was indicated by underline while motif 5 was indicated by pane T. reesei. Hexose is, after all, the preferred carbon resource of these organisms [29]. The novel TF SxlR is indicative of the sophisticated regulatory network evolved by T. reesei to adapt to complex environments.
Unlike the previously reported global regulators of carbohydrate-active enzymes (CAZymes), SxlR, as a specific regulator of GH11 family xylanases, might provide new leads for strain engineering to enhance hemi-cellulase activity in T. reesei. In the straw hydrolysis assay, more reducing sugar was produced by the Δsxlr culture supernatant than by the parent strain. Compared with steam-exploded rice straw, higher straw hydrolysis was achieved by steam-exploded rice and corn straw mixture (Fig. 4d). This was probably because the hemi-cellulose content of corn straw was much higher than that of rice straw (41% vs 22%) [30,31]. It has been suggested that the utilization of hemi-cellulose-rich substrate was significantly improved by the modified enzyme preparation. Because most TFs regulate cellulase and hemi-cellulase genes concurrently [32], the ratio of cellulase and hemicellulase is difficult to optimize by genetic manipulation of TFs. Our results shown that the appropriate proportion of the enzyme preparation can be conveniently optimized via engineering of sxlr. The derived strain would have a higher xylanase activity, while the cellulase activity would not be affected.

Conclusions
SxlR appears to be a major xylanase regulator in T. reesei that represses the expression of xylanases belonging to the GH11 family and does not affect cellulase activity. Deletion of sxlr dramatically increases xylanase activity. SxlR is a good target candidate for T. reesei strain modification, especially for hydrolyzing substrates rich in hemi-cellulose.

Strains and culture conditions
Trichoderma reesei strains including QM6a (ATCC 13631) and RUT-C30 (ATCC 56765) were maintained on potato dextrose agar plate (PDA) at 28 °C for 7 days for spore collection. Escherichia coli DH5α used as cloning host was culture at 37 °C in Luria-Bertani (LB) medium. Agrobacterium tumefaciens AGL1 was used to transform the gene to T. reesei strains. To induce enzyme production, the conidial suspension (0.5 mL, 1 × 10 7 conidia/mL) was inoculated into a 50-mL Erlenmeyer flask containing 10 mL of Sabouraud Dextrose Broth (SDB) and incubated for 40 h on an orbital shaker at 200 rpm at 28 °C. The culture was then transferred into a flask containing 10 mL of inducing fermentation medium at 10% inoculum ratio (v/v). The flasks were incubated on an orbital shaker at 200 rpm at 28

Deletion, overexpression, and re-complementation of sxlr in T. reesei RUT-C30
To delete the sxlr gene (jgi|TrireRUTC30_1|26638) of T. reesei, the 2.8-kb sxlr coding region was replaced by the hph (hygromycin phosphotransferase) gene. This was performed by amplifying 1.5 kb from upstream and 1.7 kb from downstream of sxlr from genomic DNA of T. reesei using the primer pairs listed in Additional file 7. Then, the two resulting PCR fragments were ligated into the HindIII (upstream) and XhoI (downstream) sites of the linearized pXBthg vector [30] using the ClonExpress II One Step Cloning Kit (Vazyme, Nanjing, China).
For overexpression of sxlr under a strong constitutive promoter in T. reesei, we fused the T. reesei translationelongation factor 1α (tef1) promoter, the sxlr coding region and the Ttrpc terminator from A. nidulans and inserted this fragment into the HindIII and XbaI sites of pXBthg using the ClonExpress II One Step Cloning Kit. Besides, we overexpressed 76601, 83920, 89588, 133726, 26508, and 43491 (gene ID in jgi|TrireRUTC30_1) with the same method.
For re-complementation of sxlr in the Δsxlr strain, the promoter, coding region, and terminator of sxlr were inserted into the HindIII and XbaI sites of the pXBt vector [the sequence is similar to pXBthg, except the hph marker gene was replaced by the bleomycin resistance (ble) gene] and the sequence downstream of sxlr was inserted into the XhoI site by the same method mentioned above.
All vectors constructed were verified by sequencing. Agrobacterium-mediated transformation was conducted as described previously by Ma et al. [34].

Biochemical assays
FPAase activities were determined using cellulose filter paper as described previously [35]. Xylanase activities were tested using 2% xylan from beechwood (Sigma, St. Louis, USA) as described previously [36]. Protein concentrations in the culture supernatant were determined by RC-DC protein assay (Bio-Rad, Hercules, CA, USA) (1976). Biomass was tested by the diphenylamine colorimetric method [37].
For straw hydrolysis, different pretreated biomasses such as rice straw and a corn stover plus rice straw mixture (3:7 ratio) [38] were used as substrates. The standard hydrolysis assay was carried out in a volume of 1 mL of 50 mM sodium acetate, pH 4.8, and 5% (w/v) pretreated straw as substrate in a 2-mL FastPrep tube (MP Biomedicals, Santa Ana, CA, USA). The enzyme dosage was 20 U of cellulase activity/g substrate. Incubation was typically at 50 °C with shaking at 200 rpm for 3 days, followed by centrifugation at 12,000 rpm for 10 min to obtain the supernatant to determine the reducing sugar with DNS. Each sample was examined in triplicate.

Growth of sxlr transformants on plates
Spores of T. reesei RUT-C30, Δsxlr, Osxlr, and Rsxlr were first prepared by growth on PDA plates at 28 °C and harvesting spores after 7 days. Spores were counted using a hemocytometer and 10 5 spores of each strain were inoculated onto PDA plates and MM plates containing 1% xylose, xylan, glucose, lactose, or Avicel. Double-layer Avicel plates were prepared by first casting an MM agarose bottom layer containing no carbon source, followed by an MM agarose top layer containing 1% Avicel; each strain was grown in triplicates. These plates were incubated at 28 °C before phenotypic examination.

Protein identification with MALDI-TOF/TOF analysis
After 7 days' culture with wheat bran and Avicel medium, the supernatant of RUT-C30, Δsxlr, Osxlr, and Rsxlr strain was collected. The samples were loaded into 12.5% PAGE gel, 70 min at 120 V were preferred for proteins separation. Coomassie brilliant blue R250 (Sangon, Shanghai, China) was used to color the gel for 1 h. Destainer (distilled water: ethanol: acetic acid, 5:4:1, v/v) was added for another 2-4 h. Finally, the protein band was excised from the gel and send to Shanghai Applied Protein Technology Co. Ltd, protein analysis was conducted with a MALDI-TOF/TOF mass spectrometer 5800 Proteomics Analyzer (Applied Biosystems, Framingham, MA, USA), NCBI Trichoderma (59929 protein sequence) was selected as database.

RNA extraction and real-time PCR
Mycelia were harvested after induction by wheat bran and Avicel for 4, 8, and 12 h, then transferred into TRIzol reagent (Invitrogen, USA) with standard protocol. Three biological replicates were prepared in the process. Real-time PCR was performed with primers listed in Additional file 7 with standard method using β-actin gene, rpl6e (a ribosomal protein encoding gene) and sar1 (a gene encoding a small GTPase) as the reference genes, respectively.

Expression and purification of the DNA binding domain of SxlR
The DNA binding domain (residues 252-347) of SxlR (SxlR-B) was expressed by the pGEX system according to the manufacturer's guidelines. First-strand cDNA was used as a template to amplify the fragment encompassing the DNA binding domain of SxlR using the primers indicated in Additional file 7. The fragment was then ligated into plasmid pGEX-4T-1 via BamHI and XhoI double digestion to produce pGEX-4T-SxlR-B, then subsequently introduced into E. coli BL21 (DE3) for protein production. The recombinant protein was purified using a GST-Bind resin column (Merck, Darmstadt, Germany) according to the supplier's recommendations. SDS-PAGE was used to verify the purified protein.

Electrophoretic mobility shift assays
For EMSA, the labeling of probes containing the promoter regions of specific genes was performed via PCR using the primers listed in Additional file 7. The nearly 1500-bp upstream regions (nucleotide position −1500 to −1) assumed to be the promoter regions of the xylanaseencoding genes from T. reesei RUT-C30 were divided