A truncated form of the Carbon catabolite repressor 1 increases cellulase production in Trichoderma reesei

Background Rut-C30 is a cellulase-hyperproducing Trichoderma reesei strain and, consequently, became the ancestor of most industry strains used in the production of plant cell wall-degrading enzymes, in particular cellulases. Due to three rounds of undirected mutagenesis its genetic background differs from the wild-type QM6a in many ways, of which two are the lack of a 83 kb large sequence in scaffold 15 and the partial lack of the gene encoding the Carbon catabolite repressor 1 (CREI). However, it is still unclear, what exactly enhances cellulase production in Rut-C30. Results The investigation of the expression of two genes encoding cellulases (cbh1 and cbh2) and the gene encoding their main transactivator (xyr1) revealed that the presence of the truncated form of CREI (CREI-96) contributes more to the Rut-C30 phenotype than a general loss of CREI-mediated carbon catabolite repression (cre1 deletion strain) or the deletion of 29 genes encoded in the scaffold 15 (83 kb deletion strain). We found that the remaining cre1 in Rut-C30 (cre1-96) is transcribed into mRNA, that its putative gene product (Cre1-96) is still able to bind DNA, and that the CREI-binding sites in the upstream regulatory regions of the chosen CREI-target genes are still protected in Rut-C30. As it was previously reported that CREI acts on the nucleosome positioning, we also analyzed chromatin accessibility of the core promoters of CREI-target genes and found them open even on D-glucose in the presence of CREI-96. Conclusions The lack of the full version of CREI in Rut-C30 corresponds with a partial release from carbon catabolite repression but is not completely explained by the lack of CREI. In contrast, the truncated CREI-96 of Rut-C30 exerts a positive regulatory influence on the expression of target genes. Mechanistically this might be explained at least partially by a CREI-96-mediated opening of chromatin.


Background
The filamentous ascomycete Trichoderma reesei is a saprophyte known for its ability to efficiently degrade biomass material by plant cell wall (PCW)-degrading enzymes. A genome-wide analysis identified 10 cellulolytic and 16 hemicellulolytic enzyme-encoding genes in T. reesei [1], of which the two most prominent cellulosedegrading enzymes are the cellobiohydrolases CBHI and CBHII (EC 3.2.1.91) [2]. However, the defining feature of this fungus is the exceedingly high amount of secreted enzymes that provoked its industrial exploitation for their production. Next to the pulp and paper industry [3], the food and feed industry [4], and the textile industry [5], these enzymes are applied in the production of biofuels [4]. With regards to cellulosic ethanol, the production costs of the needed enzymes massively influence the price and the competitiveness of the end-product. As a result their efficient expression remains an important topic of research.
The ancestor of most current industry strains is Rut-C30 [6,7]. It was derived from the T. reesei wild-type isolate QM6a through three rounds of mutagenesis (ultraviolet (UV) light and N-nitroguanidine) followed by a screening for the release from carbon catabolite repression (CCR) and high cellulase activity [8][9][10]. Since then the study of the specific physiological and genetic changes in Rut-C30 has been of interest [7,11] and a number of properties have been identified, for example, a 83 kb large region located in scaffold 15, which encodes 29 genes is lacking in Rut-C30 [12,13]. Another important property of Rut-C30 is the lack of the full version of the Carbon catabolite repressor 1 (CREI, [14]) having left a cre1 sequence that would only encode for one of the two zinc finger regions of CREI (96 aa long) [15]. The native CREI is a C 2 H 2 -type zinc finger protein with the consensus sequence 5′-SYGGRG-3′ [14], and is orthologous to CreA from Aspergillus sp. [16][17][18] and MIG1 from Saccharomyces cerevisiae [19].
In T. reesei CREI is known to act on a number of regulatory levels. Firstly, it directly represses transcription of several genes encoding for PCW-degrading enzymes, for example the cbh1 gene [20], by binding to tandem and inverted repeats in their upstream regulatory regions (URRs) [20,21]. Secondly, in the presence of D-glucose it represses the expression of the main transactivator of PCW-degrading enzyme expression, the Xylanase regulator 1 (XYRI) [22,23]. XYRI is a Gal4-like Zn2Cys6 binuclear cluster protein, of which the expression can be induced by sophorose [6] or as above-mentioned, repressed by D-glucose. Notably, the extent of induction of xyr1 gene expression directly correlates with the induction of cbh1 and cbh2 gene expression, which is not the case for other genes in the XYRI regulon [6]. As a third regulatory level, CREI has been reported to play an essential role in correct nucleosome positioning, for example in the promoters of the cbh1 and cbh2 genes [24,25]. Altogether, CREI plays a major role in the regulation of 250 genes, both in a repressing and inducing way [26].
In this study we investigated to which extent the two mentioned striking genetic properties of Rut-C30, namely the lack of the 83 kb in scaffold 15 and the absence of a full CREI contribute to its cellulase hyper-producing phenotype. A special focus was studying the role of the remaining truncated version of CREI with regard to the transcriptional regulation of target gene expression, the ability of DNA binding, and the influence on the chromatin structure. The PCW-degrading enzyme-encoding genes cbh1 and cbh2, as well as xyr1 as the gene encoding their transactivator, were chosen as the CREI-target genes to be analyzed.

Results
The expression profile of cbh1, cbh2, and xyr1 in QM6a-CREI 96 is closest to Rut-C30 In order to find out which genetic properties of Rut-C30 contribute to its cellulase hyper-producing phenotype we compared its expression profile to three transgenic strains. Two of them bear Rut-C30-related mutations, namely Δscaff15, which is a QM6a-derived strain with an 83 kb deletion in scaffold 15, and QM6a-CREI 96 , which bears the truncated CREI version like Rut-C30 does. We also included a QM6a-derived cre1 deletion strain because the phenotype of Rut-C30 is sometimes associated with the lack of CREI. We investigated expression levels of cbh1, cbh2, and xyr1 under sophoroseinducing conditions. Interestingly, the transcript levels of all three genes were most similar in QM6a-CREI 96 compared to Rut-C30, while in the case of the other two strains (Δcre1 and Δscaff15) considerably lower levels were detected (Figure 1a, b, c). This result raised the consideration that regulatory molecular mechanisms are related to the truncated version of cre1 remaining in the Rut-C30 genome.
CREI sites in upstream regulatory regions of target genes are strongly protected from DNA methylation in Rut-C30 All above-mentioned analyzed genes (cbh1, cbh2, and xyr1) are under the regulatory influence of CREI [20,22,25]. To learn if the protection patterns against DNA methylation of the URRs bearing CREI sites of these genes differ between the wild-type QM6a and Rut-C30, we performed in vivo/ in vitro footprinting analyses. URRs of all three genes were investigated after both strains were pre-grown on glycerol and then incubated on D-glucose for 3 hours followed by dimethyl sulphate (DMS)-induced in vivo methylation. The footprinting pattern obtained for all genes showed the same or even stronger DNA occupancy in Rut-C30 compared to QM6a (Figure 2a, b, c). This finding prompted us to analyze if the truncated cre1 (cre1-96) in Rut-C30 is transcribed. Thus, after pre-growth both strains were transferred to media containing D-glucose as a repressing condition, sophorose as an inducing condition, and no carbon source as a control condition respectively, and incubated for 3 hours. While we detected a low basal transcript level (originating from the native cre1 gene) in the wild-type, we found increased levels in Rut-C30, whereupon the increase was more pronounced on D-glucose and sophorose than without carbon source ( Figure 3a). Since the cre1-96 mRNA could be detected in high amounts in Rut-C30 regardless of the applied condition (D-glucose or sophorose), we again performed in vivo/in vitro footprinting to investigate if the DNA protection pattern changes condition-dependent. We analyzed the same URRs of cbh1, cbh2, and xyr1 genes comparing the application of D-glucose and sophorose. As can be inferred from Figure 3b, c, d no conditionspecific differences could be detected even though strong DNA occupancy at the CREI sites was observed. This fits the expression results and suggests that CREI-96 may no longer act in a D-glucose specific manner.
The truncated CREI-96 protein of Rut-C30 can bind to DNA in vitro The results described above raise the question of whether the putative CREI-96 protein is still able to bind to DNA, especially because it lacks one of the two zinc fingers. To answer this question, we performed an electrophoretic mobility shift assay (EMSA) using the URR of cbh1 covering three functional CREI-binding sites as a probe (Figure 4a). The applied CREI-96 protein was heterologously expressed as GST-fusion protein and thrombincleaved before usage. The assay yielded two bands (Figure 4b), which represent most likely the binding of one or more CREI-96 proteins to the sites of the probe. Similar results were observed before reporting that shorter versions of CREI can still bind target DNA sequences yielding more than one band [28]. However, both complexes are specifically formed because the bands were diminished by adding a cold competitor, while they remained unchanged by adding a specifically mutated competitor ( Figure 4b). Altogether, this supports the working hypothesis that the truncated CREI-96, which is putatively formed in Rut-C30, is still able to bind its DNA target sequences. It should be noted that in filamentous fungi other regulatory proteins are known that are also able to bind their DNA with a single zinc finger, such as the Aspergillus nidulans AreA [29].

Higher gene expression corresponds with more open chromatin in QM6a-CREI 96
As we had strong indication that the truncated version of CREI-96 is present in Rut-C30 and still is able to bind the URRs of its target genes, we aimed to get more insights on its potential regulatory role. Consequently, QM6a-CREI 96 and the cre1 deletion strain were grown on D-glucose and samples were drawn after 36, 39, 42, and 45 hours of cultivation. Because CREI was reported to be involved in nucleosome positioning within the cbh1 and the cbh2 promoter of T. reesei [24,25], we investigated the chromatin packaging by applying chromatin accessibility real-time PCR (CHART-PCR) analysis of the core promoter regions of the cbh1, cbh2, and xyr1 genes. Second, we investigated the expression of these genes by RT-qPCR to see if there was a correlation of transcript levels with the chromatin accessibility. We could detect a higher expression of cbh1 and cbh2 genes in QM6a-CREI 96 compared to the Δcre1-strain, which corresponded with a more open chromatin in the core promoters of the two genes in QM6a-CREI 96 (Figure 5a, b). The higher transcript levels of cellulaseencoding genes in QM6a-CREI 96 were reflected by increased enzyme activity (45.8 ± 1.5 U/mg dry weight;) measured after 45 hours of cultivation on D-glucose compared to the Δcre1-strain (29.1 ± 1.8 U/mg dry weight). Interestingly, we found that in QM6a-CREI 96 the gene expression increased with a simultaneous opening of chromatin, in particular in xyr1 and cbh1 (Figure 5a). On the other hand, we could not observe a correlation of gene expression and chromatin accessibility in the cre1 deletion strain (Figure 5b). a b c Figure 1 Transcript analysis in Trichoderma reesei strains bearing Rut-C30-related mutations. Rut-C30 (yellow bars), QM6a-CREI 96 (red bars), the Δcre1-strain (blue bars), and the Δscaff15-strain (purple bars) were pre-grown on glycerol and then transferred to medium supplemented with sophorose and incubated for 3 hours. The transcript level analysis of the cbh1 (a), cbh2 (b), and xyr1 (c) genes was performed by qPCR using sar1 and act genes for data normalization. Levels refer to the wild-type strain QM6a incubated without carbon source. The values are means from measurements and biological experiments (cultivations) performed in triplicate.

Error bars indicate standard deviations.
To discover what this scenario looks like under inducing conditions, we applied the same experimental strategy on the two strains after pre-growth followed by incubation on sophorose for 30, 90, and 180 minutes. For cbh1 and cbh2 we observed in both strains an increase in transcript levels over time, which did not correlate in either of the two strains with a simultaneous opening of chromatin (Figure 6a, b). In the xyr1 gene, transcript levels did not increase however, the chromatin became more compact over time (Figure 6a, b). Comparing the chromatin status of the two strains, the chromatin was slightly more accessible in QM6a-CREI 96. (Figure 6a, b).

CREI-96 upregulates the expression of a helicase-like transcription factor
Due to the results obtained from the CHART-PCR analyses, we assumed that the truncated CREI-96 was involved in chromatin remodelling in both a direct and indirect a b c Figure 2 In vivo and in vitro footprinting analyses of URRs of CREI-target genes. The Trichoderma reesei wild-type strain QM6a and Rut-C30 were pre-grown on glycerol and then incubated on D-glucose for 3 hours followed by DMS-induced in vivo methylation. An URR bearing functional CREI sites (underlined in red) of the cbh1 (a), cbh2 (b), and xyr1 (c) genes was investigated, and methylated, naked DNA was used as the reference. Numbers indicate the position of the base upstream from ATG. Analysis of data and visualization was performed using ivFAST (in vivo footprinting analysis software tool) [27]. Only signals that are statistically different are considered. Protected bases are highlighted in red shades and hypersensitive bases are highlighted in blue shades. The three colour intensities each correspond to stronger differences between compared conditions; increasing colour intensity means more than 1.4-, 1.6-, and 1.8-fold difference in cbh1 and cbh2 (a, b), and more than 2.4-, 2.6-, and 2.8-fold difference in xyr1 (c). a b c d Figure 3 Analysis of the abundance of CREI-96 in Rut-C30. (a) The Trichoderma reesei wild-type strain QM6a (green bars) and Rut-C30 (yellow bars) were pre-grown on glycerol and then transferred to media supplemented with D-glucose (G), sophorose (SO) or without carbon source (NC), respectively, and incubated for 3 hours. The transcript level analysis of cre1-96 was performed by qPCR using sar1 and act genes for data normalization. Levels refer to the wild-type strain incubated without carbon source. The values are means from measurements in triplicates and three biological experiments (cultivations). Error bars indicate standard deviations. (b-d) In vivo and in vitro footprinting analysis of URRs of CREI-target genes in Rut-C30, which was pre-grown on glycerol and then incubated on D-glucose (G) or sophorose (SO) for 3 hours followed by DMS-induced in vivo methylation. An URR bearing functional CREI sites (underlined in red) of the cbh1 (b), cbh2 (c), and xyr1 (d) genes each was investigated, and methylated, naked DNA (ND) was used as the reference. Numbers indicate the position of the base upstream from ATG. Analysis of data and visualization was performed using ivFAST [27]. Colour codes are the same as in Figure 2.
manner. Therefore, we searched the T. reesei genome for chromatin-remodelling proteins. We selected seven proteins, of which the annotation pointed to an involvement in chromatin rearrangement, for further investigation (Table 1). To determine a potential regulatory influence by CREI-96 we investigated their transcript levels in QM6a-CREI 96 and the Δcre1-strain. Amongst the seven genes we observed one with different transcript levels, namely the gene encoding a helicase-like transcription factor (Protein ID 44747). Here we detected strong upregulation in QM6a-CREI 96 compared to the Δcre1-strain, regardless of whether the samples from the D-glucose growth experiment (Figure 7a) or the sophorose replacement ( Figure 7b) were investigated. We term the aforementioned gene in this study htf1 (helicase-like transcription factor 1). It should be noted that in previous reports this gene was called snf2, although it is not homologous to SNF2 of Saccharomyces cerevisiae [30]. However, the increased expression of htf1 in QM6a-CREI 96 might be one reason for the observed change in chromatin.

Discussion
For a long time its lack of the full version of CREI has been considered to be a positive genetic property of Rut-C30 because it leads to a partial carbon catabolite de-repression. However, Rut-C30 still possesses a short version of CREI, CREI-96. During this study it became clear that this protein still has a function and influences the expression of target genes, notably, in a positive manner. We found three ways in which CREI-96 mechanistically exerts its role as a regulatory protein. hours. The cbh1 (green), cbh2 (orange), and xyr1 (blue) genes were investigated. The gene expression analysis was performed by cDNA synthesis followed by qPCR, and transcript levels (log) are depicted on the x-axis. CHART-PCR was performed by Dnase I digestion followed by qPCR, and CAIs are depicted on the y-axis. In both cases sar1 and act genes were used for data normalization and QM6a-CREI 96 grown for 36 hours was the reference condition. All values are means from measurements in triplicates and three biological experiments (cultivations). Standard deviations were below 5%. Diagrams are identically scaled. (c) Spearman's rank correlation coefficients were calculated (close to −1 indicates negative correlation; close to 0 indicates no linear correlation; close to 1 indicates positive correlation). The P-values for each correlation coefficient were calculated to determine significance of data (how different from zero). P >0.05 was considered as indicative of no linear correlation.
binding sites (compare Figures 2 and 3). Obviously, contacting of CREI-binding sites in target genes by CREI-96 leads to a more open chromatin in the core promoter regions under repressing conditions (D-glucose). As expected, this effect was not observed in such a pronounced manner under inducing conditions (sophorose), as the core promoter there already has an open chromatin. If CREI-96 develops this de-regulating function by simply binding the DNA (possibly in high amounts) and thereby preventing nucleosome assembly, or if it is involved in a more complicated chromatin-DNA interaction mechanism, cannot yet be answered and this calls for further investigation.
Secondly, CREI-96 acts on the chromatin structure in an indirect way. The transcript levels of the htf1 gene encoding a chromatin-remodelling protein are upregulated in the presence of CREI-96. Notably, in the wild-type strain QM6a bearing the native CREI we found htf1 downregulated, specifically on D-glucose (data not shown). This is in good accordance with previous studies, in which this gene was reported to be repressed by CREI in the T. reesei strain QM9414 [26] and to be lowly expressed in QM6a on D-glucose [31]. It can be speculated that CREI-96 in this case reverses the antagonistic function of CREI on chromatin-remodelling proteins and even supports the opening of chromatin in an indirect way.
Thirdly, the loss of the auto-regulatory function of CREI is very likely. It should be considered that surprisingly high transcript levels of cre1-96 were detected in Rut-C30 (Figure 3a). In A. nidulans it was observed that a low steady-state level of creA mRNA can be increased within minutes by adding a repressing carbon source. While a prolonged incubation with the repressing carbon sources then led to auto-repression, the incubation on a derepressing carbon source maintained the high level [18]. In Rut-C30 the cre1-96 transcript levels on D-glucose are as high as on sophorose, indicating a loss of the mentioned native auto-repression. This might also be the reason for the generally higher abundance of transcript in Rut-C30 compared to the wild-type strain QM6a which we observed during this study. An alternative or additional explanation for the generally higher abundance of cre1-96 in Rut-C30 is mRNA stability. This may arise from less stability of the mRNA coding for the full-length protein, or a general increase in mRNA stability speculated to be caused by two mutations (protein ID 110423 and 66895) identified in the direct ancestor of Rut-C30 (NG14) [11]. However, it can be assumed that in Rut-C30 this results in high amounts of CREI-96 present in the cell, which is still able to bind DNA, thereby possibly changing the chromatin status in CREI-target genes.
Even if in previous studies truncated forms of CREI have been investigated, interestingly, none of them reported a positive effect of the shorter CREI version compared to a deletion of CREI. This is most likely due to a different experimental design applied. Nakari-Setälä et al. reported similar cellulase activities (volumetrically given) of strains bearing a deletion or a truncation of CREI when grown on lactose [32]. However, the investigated strains exhibited different growth behaviors (biomass formation and CO 2 -c b a Figure 6 Transcript and CHART analysis of Trichoderma reesei CREI-mutant strains during sophorose induction. QM6a-CREI 96 (a) and the Δcre1-strain (b) were pre-grown on glycerol and thereafter incubated on sophorose for 30 (dots), 90 (boxes), and 180 (triangles) minutes. The cbh1 (green), cbh2 (orange), and xyr1 (blue) genes were investigated. The gene expression analysis was performed by cDNA synthesis followed by qPCR, and transcript levels are depicted on the x-axis. CHART-PCR was performed by Dnase I digestion followed by qPCR, and CAIs are depicted on the y-axis. In both cases sar1 and act genes were used for data normalization and QM6a-CREI 96 incubated for 30 minutes was the reference condition. All values are means from measurements in triplicates and three biological experiments (cultivations). Standard deviations were below 5%. Diagrams are identically scaled. (c) Spearman's rank correlation coefficients were calculated (close to −1 indicates negative correlation; close to 0 indicates no linear correlation; close to 1 indicates positive correlation). The P-values for each correlation coefficient were calculated to determine significance of data (how different from zero). P >0.05 was considered as indicative of no linear correlation. levels) [32]. In order to exclude any growth effects we studied a resting cell system under inducing conditions (sophorose). Such a highly standardized transfer experiment allows determination of the mechanistic influence of an isolated phenomenon (presence of CREI-96 or lack of CREI).
We also investigated growth conditions on D-glucose as this is certainly industrially relevant (enzyme production on high glucose-containing inducers). In this case we also observed differences in transcript levels between the two CREI mutant strains, which is in accordance with the results of a Northern analysis of cbh1 on glucose by Nakari-Setälä et al. [32] and of the cbh1 transcript level analysis by Ries et al. [24].
We noticed that the increased transcript levels in strain QM6a-CREI 96 corresponded with an opening of chromatin on D-glucose, which was not found on sophorose. We propose that this different observation is likely due to the fact that the remodelling of chromatin usually becomes necessary under repressing conditions, while under inducing conditions certain regulatory factors and/ or mechanisms have already ensured the open chromatin status. Obviously, the differences between the two CREI mutant strains observed on sophorose are primarily related to induction, chromatin opening, and expression of XYRI. During early-stage induction (30 minutes) the presence of sophorose leads to a rapid and higher chromatin accessibility of xyr1 in QM6a-CREI 96 , which subsequently results in higher xyr1 transcript levels ( Figure 6). It was previously reported that the level of xyr1 expression is directly linked to the expression of cbh1 and cbh2 [6]. Accordingly, we detected higher cbh1 and cbh2 transcript levels in the QM6a-CREI 96 on sophorose.

Conclusions
From the result obtained during this study we conclude that the truncated version of CREI present in Rut-C30 (CREI-96) should be considered as a discrete transcription factor with different properties than CREI. It acts in a direct manner on target URRs, but also contributes indirectly to a more open chromatin status by regulating a chromatin remodeler. Altogether it should be considered that Rut-C30 not only lacks CREI and the thereby mediated CCR, but also gains the now positively acting regulatory protein CREI-96.
Orthologs of CREI regulate CCR in numerous filamentous fungi including those used in biofuel production. Unfortunately, a simple deletion of cre1 does not only lead to the desired release from CCR, but also to severe growth impairment. Consequently, molecular strain design can alternatively be based on the described CREI-96 truncation. This strategy does on one hand avoid growth deficiencies and the accompanying loss of productivity, and on the other hand additionally leads to a chromatin remodelling effect that results in increased expression of PCW-degrading enzymes.

Growth conditions
For carbon source replacement experiments mycelia were pre-cultured in 1 L Erlenmeyer flasks on a rotary shaker (180 rpm) at 30°C for 24 hours in 250 mL of Mandels-Andreotti (MA) medium [34] supplemented with 1% (w/v) glycerol as sole carbon source. A total of 10 9 conidia per litre (final concentration) were used as inoculum. Pre-grown mycelia were washed and equal amounts were resuspended in 20 ml MA media containing 1% (w/v) D-glucose or 2 mM sophorose (Serva Electrophoresis, Heidelberg, Germany) as sole carbon source, or no carbon source respectively, and incubated for 30 minutes to 3 hours. For direct cultivation experiments the conidia were incubated in 250 mL Erlenmeyer flasks on a rotary shaker (180 rpm) at 30°C for 45 hours in 50 mL of MA medium supplemented with 1% (w/v) D-glucose as sole carbon source. Samples were derived from three biological replicates and were pooled before RNA extraction and chromatin digestion.

Replacement of CREI in T. reesei QM6a
Transformation of T. reesei QM6aΔtmus53 [33] was performed using two overlapping 3 kb-DNA fragments. The first fragment consisted of the truncated cre1 of Rut-C30 amplified with primers RG186 and RG187 using genomic DNA as template and 3′-half of the expression cassette bearing the Escherichia coli hph marker gene amplified with primers RG188 and hph3′_fw using the plasmid pRLMex 30 [35] as template. The second fragment consisted of the 5′-half of the expression cassette bearing the E. coli hph marker gene amplified with primers hph5′_rev and RG189 using pRLMex 30 as template and the cre1 3′flank from Rut-C30 amplified with primers RG190 and RG191 using genomic DNA as template. Protoplast transformation of QM6a was performed as described in United States patent number 8,323,931 using 2.5 μg of each DNA fragment in a co-transformation. Genomic integration of the full construct (cre1-96::hph) into the cre1 locus was confirmed by southern blot analysis and DNA sequencing (Microsynth, Balgach, Switzerland).

In vivo footprinting
In vivo methylation using DMS followed by ligationmediated PCR was performed as described previously [27]. FAM (fluorescein amidite)-labelled fragments were analyzed by capillary gel electrophoresis (Microsynth) and results were analyzed using ivFAST [27].  Table 2. Cycling conditions and control reactions were performed as described previously [36]. Data normalization using sar1 and act as reference genes and calculations were performed as published previously [36].

Plasmid construction
A 307 bp fragment was amplified from Rut-C30 genomic DNA using primers RG182 and RG178 and was inserted into the expression vector pGEX-4 T-2 (GE Healthcare Life Sciences, Little Chalfont, Buckinghamshire, United Kingdom) via BamHI/XhoI digestion yielding the plasmid pGEX-cre1-RG182 for heterologous expression of GST (glutathione S-transferase)-fused CREI-96.  stored at −20°C overnight. GST-fusion protein (theoretical molecular weight 37 kDa) was purified from E. coli cell lysates using Glutathione-Superflow Resin (Qiagen, Hilden, Germany) according to the manufacturer's instructions.

Electrophoretic mobility shift assay (EMSA)
A 400-bp PCR product comprising the part of the 5′-URR region of cbh1, which contains 3 CREI-binding sites, was obtained with FAM-labelled primers RG89 and RG90 and used as probe ( Table 2). The protein-DNA binding assay and non-denaturing polyacrylamide gel electrophoresis were performed essentially as previously described [37]. ). This probe bears in all three CREI-binding sites a mutation from 5′-SYGGRG-3′ to 5′-SYTGRG-3′, which was previously shown to prevent the binding of CREI in vivo and in vitro [21].

Determination of cellulase activity
Cellulase activity in the culture supernatants was determined using AZCL HE-Cellulose (Megazyme International, Bray, Ireland) in 25 mM sodium acetate buffer pH 4.5 according to the manufacturer's instructions. To measure biomass (dry weight), the cultures were harvested by filtration, washed with an equal volume of 0.8% NaCl solution, dried at 80°C for 24 hours, and weighed. Samples from two biological replicates and two technical replicates were measured.

Chromatin accessibility real-time PCR (CHART-PCR)
DNase I digestion of chromatin and DNA extraction were carried out as described by Gonzalez and Scazzocchio [38] with minor modifications. Mycelia were harvested by filtration, pressed dry with filter paper, frozen in liquid nitrogen, and ground to a fine powder. Portions (100 mg) of the powder were suspended in 1 mL aliquots of nuclease digestion buffer (250 mM sucrose, 60 mM KCl, 15 mM NaCl, 0.05 mM CaCl 2 , 3 mM MgCl 2 , 0.5 mM DTT, and 15 mM Tris-HCl at pH 7.5), and 100-μL samples of the digestion mixture were incubated with 10 U of RQ1 RNase-free DNase I (Promega, Madison, Wisconsin, United States) for 2.5 minutes at 37°C. The reaction was stopped by adding 100 μL of 40 mM EDTA and 2% SDS, followed by two rounds of phenol-chloroform extraction and one round of chloroform extraction. Samples were then treated with 10 μg/mL of RNase A for 15 minutes at 37°C and precipitated with ethanol. DNA pellets were suspended in 100 μL of 5 mM Tris-HCl at pH 7.5. A control without DNase I was included to monitor endonuclease activity. qPCR analysis of the DNase I-treated samples was performed to measure the relative abundance of target regions. PCRs were performed in a Rotor-Gene Q system (Qiagen, Hilden, Germany). All reactions were performed in triplicate. The amplification mixture (final volume 20 μL) contained 10 μL 2 × iQ SYBR Green Mix (Bio-Rad, Hercules, USA), 200 nM forward and reverse primers and 10 ng of DNA. Primer sequences are provided in Table 2. Cycling conditions were as follows: 3 minutes initial denaturation at 95°C, followed by 40 cycles of 15 seconds at 95°C and 60 seconds at 60°C. The amount of intact input DNA of each sample was calculated by comparing the threshold values of the PCR amplification plots with a standard curve generated for each primer set using serial dilutions of genomic, undigested DNA. The chromatin accessibility index (CAI) was defined as: where Ds is the amount of intact DNA detected for each target region and Dc1 and Dc2 are the amounts of intact DNA detect for the promoter regions of sar1 and act respectively, used as reference genes for normalization. To access the relationship between the CAI and the transcript level, a two-tailed Spearman's rank correlation coefficient was determined for each gene analyzed. The P-value for each correlation coefficient was calculated to determine the significance of the data.

Selection of investigated genes encoding chromatin remodelling factors
Genes were selected by direct searching for SNF2, ISW1, ISW2, INO80, CDH1, RSC8 (the most prominent chromatin remodelling-related proteins characterized in yeast) in the T. reesei genome database [39]. Additional candidate genes were obtained by using BLASTp (basic local alignment search tool) search in the NCBI database employing respective S. cerevisiae sequences as baits to identify similar sequences in filamentous fungi. Subsequently, these sequences were used in a BLAST search in the T. reesei database.