Influence of cis Element Arrangement on Promoter Strength in Trichoderma reesei

ABSTRACT Trichoderma reesei can produce up to 100 g/liter of extracellular proteins. The major and industrially relevant products are cellobiohydrolase I (CBHI) and the hemicellulase XYNI. The genes encoding both enzymes are transcriptionally activated by the regulatory protein Xyr1. The first 850 nucleotides of the cbh1 promoter contain 14 Xyr1-binding sites (XBS), and 8 XBS are present in the xyn1 promoter. Some of these XBS are arranged in tandem and others as inverted repeats. One such cis element, an inverted repeat, plays a crucial role in the inducibility of the xyn1 promoter. We investigated the impact of the properties of such cis elements by shuffling them by insertion, exchange, deletion, and rearrangement of cis elements in both the cbh1 and xyn1 promoter. A promoter-reporter assay using the Aspergillus niger goxA gene allowed us to measure changes in the promoter strength and inducibility. Most strikingly, we found that an inverted repeat of XBS causes an important increase in cbh1 promoter strength and allows induction by xylan or wheat straw. Furthermore, evidence is provided that the distances of cis elements to the transcription start site have important influence on promoter activity. Our results suggest that the arrangement and distances of cis elements have large impacts on the strength of the cbh1 promoter, whereas the sheer number of XBS has only secondary importance. Ultimately, the biotechnologically important cbh1 promoter can be improved by cis element rearrangement. IMPORTANCE In the present study, we demonstrate that the arrangement of cis elements has a major impact on promoter strength and inducibility. We discovered an influence on promoter activity by the distances of cis elements to the transcription start site. Furthermore, we found that the configuration of cis elements has a greater effect on promoter strength than does the sheer number of transactivator binding sites present in the promoter. Altogether, the arrangement of cis elements is an important factor that should not be overlooked when enhancement of gene expression is desired.

ment of cis elements on promoter strength has hardly been investigated so far. Therefore, during this study, we investigated the role of cis elements composed of Xyr1-binding sites (XBS) in the promoters of the cbh1 and xyn1 genes. Xyr1 is the main and essential transactivator of cbh1 and xyn1 gene expression (4) and has the binding consensus sequence 5=-GGC(A/T) 3 -3= (5). Corresponding schemes of the two promoters studied are depicted in Fig. 1.
The xyn1 gene is positively transcriptionally regulated by Xyr1 (4). The promoter region spanning from position Ϫ1 to Ϫ850 bp contains eight XBS (Fig. 1). Gene transcription is induced by xylan and its degradation product D-xylose. Notably, D-xylose induction is concentration dependent: low concentrations at around 0.5 mM induce gene transcription, but high concentrations, e.g., 1% (wt/vol) (66.6 mM), repress xyn1 transcription (6). On the other hand, Ace1 and Cre1 act as repressors of xyn1 transcription (7,8). A region between positions Ϫ321 and Ϫ538 bp in the xyn1 promoter was identified to be necessary for the induction of the gene (6), and later, it was reported that this fragment contains all cis elements for the regulation of the xyn1 gene (5). Rauscher and coworkers identified an important inverted repeat of XBS and observed that the deletion of Cre1-binding sites had no influence on the inducibility of xyn1 (5). These results encouraged us to investigate the impact of the identified inverted repeat of XBS (located between Ϫ405 and Ϫ426 bp) in greater detail. The corresponding sequence, 5=-GGCTAAATGCGACATCTTAGCC-3= (XBS are underlined), is termed cis element IR in this study.
The transcriptional activators for the cbh1 gene are Xyr1 and Ace2 (4,9). Cellulose and its degradation products induce gene transcription to a certain extent, while the strongest known inducer of cellulase expression is the transglycosilation product sophorose, a disaccharide formed by two ␤-1,2-linked glucose residues. As sophorose is a very expensive compound, the search for an equivalent replacement is still an important issue (10). Another inducing signal for cbh1 expression is lactose, which is also mediated by Xyr1, but not by Ace2 (11,12). The cbh1 core promoter from position Ϫ1 to Ϫ850 bp contains 14 XBS (Fig. 1). The regulatory proteins Cre1 and Ace1 are repressors of transcription of the cbh1 gene. Cre1 mediates carbon catabolite repression (CCR) by binding Cre1-binding sites (5=-SYGGRG-3=) (13) in the 5= region of the cbh1 core promoter in the presence of D-glucose. Ace1 binds DNA motifs similar to those bound by Xyr1. There are eight confirmed target sites present in the cbh1 promoter (14). It is hypothesized that the regulation is based on competition for these sites between Xyr1 and Ace1 (7). Liu and coworkers deleted Cre1-binding sites in the cbh1 promoter of the T. reesei Rut-C30 strain M3 and inserted the cbh1 promoter sequence spanning from Ϫ620 to Ϫ820 bp three more times into the 5= region (15). This approach yielded two-times-higher activity than the wild-type (WT) promoter construct using a glucuronidase reporter assay. Another approach was the exchange of Cre1-binding sites for Ace2-binding sites and binding sites for the Hap2/3/5 complex (16). This experiment was performed in the protease-deficient Rut-C30 strain RC30-8 and resulted in a 5.5-fold increase in enhanced green fluorescent protein (eGFP) expression on inducing medium by a single copy of the inserted reporter gene. The high abundance of XBS in the cbh1 makes these sites an interesting target for research (17). In the present study, we examined three different cis elements composed of XBS in the cbh1 promoter, including (i) a direct repeat of three XBS located at bp Ϫ173 of FIG 1 Schematic representation of the cbh1 and xyn1 promoters. Green arrows indicate XBS and their orientations, ‡ indicates the TATA signal, and ATG represents the start codon of the cbh1 or xyn1 gene. The blue box indicates the cis element Z (direct repeat of three XBS), the orange box indicates the cis element Y (an inverted repeat of XBS), the purple box indicates the cis element X (direct repeat of three XBS), and the brown box indicates the cis element IR (inverted repeat of XBS). Numbers on top indicate the distance from the ATG in base pairs. the cbh1 promoter, 5=-TTAGCCAAGAACAATAGCCGATAAAGATAGCC-3= (termed cis element X in this study), (ii) an inverted repeat of XBS located at Ϫ733 bp of the cbh1 promoter, 5=-GGCTAAACGTACCGTAATTTGCC-3= (termed cis element Y in this study), and (iii) three XBS located at Ϫ773 bp of the cbh1 promoter, 5=-GGCTAAAAGTACATA AGTTAATGCCTAAAGAAGTCATATACCAGCGGCTAA-3= (termed cis element Z in this study). The cis elements Y and Z are located on the same sequence stretch that was used in the study of Liu et al. (15). cis element X is a direct repeat of three XBS; however, it is located much closer to the TATA signal (Fig. 1).
There is only limited and rather general information about the impact of longer sequence stretches within the cbh1 and xyn1 core promoters, and there are no detailed investigations on the influence of cis elements on the inducibility and strength of these promoters. Therefore, we aimed to address the following questions during this study. (i) Is it possible to assign carbon source inducibility to certain cis elements and thereby modify promoter inducibility in a targeted manner by cis element rearrangement? (ii) What is the impact of the number of XBS on promoter inducibility and strength? (iii) Does the distance between cis elements and the transcription start site (TSS) have an influence on promoter activity? Certainly, the answers to these questions are important for improving the efficiency, cost-effectiveness, and productivity of target gene expression.

RESULTS
The cis element IR increases promoter strength. At the beginning of our studies, we were interested in whether the cis element X is responsible for sophorose induction in the cbh1 promoter on the one hand and, on the other hand, whether the cis element IR is responsible for D-xylose induction in the xyn1 promoter. Therefore, we constructed expression cassettes bearing the deletion of the respective cis element in each promoter or a motif exchange where X was exchanged for IR in the cbh1 promoter and termed cis element X-IR and IR was exchanged for X in the xyn1 promoter and termed cis element IR-X. Schemes of these promoters are depicted in Fig. 2A. The modified and native promoters were fused to the Aspergillus niger goxA gene, which is an established reporter system in Trichoderma species (18)(19)(20)(21), and subsequently transformed into T. reesei QM6aΔtmus53Δpyr4 (22). Using the reestablishment of pyr4 as a marker allowed the selection of uridine prototrophic strains. Correct genotypic modification, i.e., the insertion of one copy at the pyr4 locus, was tested by PCR and Southern blot hybridization (see Fig. S1A and B in the supplemental material). To validate the experimental design and as a first approach, we investigated these strains by a carbon source replacement experiment.
Pc_WT, the strain expressing goxA under the native cbh1 promoter, had basal GoxA activity on medium without any carbon source and on glycerol (Fig. 2B). While we observed CCR on D-glucose, we detected slight induction on a low concentration of D-xylose and strong induction on sophorose (Fig. 2B). The results of the GoxA assays reproduce already published quantitative PCR (qPCR) data on cbh1 from similar carbon source replacement experiments (2). Together, these results confirmed the suitability of this reporter assay for our research objectives. The deletion of cis element X (strain Pc_ΔX) released the cbh1 promoter slightly from CCR. With regard to inducing conditions, the deletion had no effect using D-xylose and led to a minor decrease of activity on sophorose (Fig. 2B). The cis element exchange in strain Pc_X-IR led to a general increase of cbh1 promoter activity regardless of which condition (i.e., carbon source) was used (Fig. 2B). In testing the xyn1 promoter, the native one in strain Px_WT was only induced on a low concentration of D-xylose (Fig. 2B). Induction on sophorose was not observed. Deletion of the cis element IR in strain Px_ΔIR led to a loss of inducibility on low concentrations of D-xylose (Fig. 2B). This loss could also be observed in strain Px_IR-X, bearing the motif exchange IR for X in the xyn1 promoter. Neither was induction by sophorose observed in this strain (Fig. 2B).
As mentioned above, the strain Pc_WT behaved as expected on all carbon sources. The deletion of the cis element X only led to minor changes, i.e., decreased GoxA expression under inducing conditions and a slight release from CCR under repressing conditions. Otherwise, the cis element exchange X-IR led generally to higher promoter activity; for example, 3.3-times-higher GoxA activity under inducing conditions than with the native promoter. The regulation of the xyn1 promoter is much tighter. Under several conditions, the GoxA activity was below quantification or detection limits (Fig.  2B). D-Xylose-mediated inducibility was lost when the cis element IR was absent, and the cis element X could not restore any inducibility. Therefore, we conclude that the cis element IR is necessary for inducibility of the xyn1 promoter on D-xylose, while the cis element X does not mediate sophorose inducibility in either of the two promoters. Interestingly, the presence of cis element IR in the cbh1 promoter led to a generally stronger promoter. These results raised the question of whether cis element modification allows the use of alternative carbon sources for the same or improved promoter activation as reached with known inducers.
The cis element IR mediates strong inducibility of the cbh1 promoter by wheat straw. The results showing that the cbh1 promoter strength can be increased by the insertion of the cis element IR prompted us to study the impact of this cis element in detail. We investigated four different promoter versions, which are depicted in Fig. 3A. All four approaches have in common that the cis element Z was exchanged for the cis element IR (this exchange is termed Z-IR). Besides this, an additional cis element IR was inserted either downstream (termed IR-DI) or upstream (termed IR-UI) from the cis element X (maintaining the nucleotide distances between the cis elements X and IR and the surrounding elements) or the cis element IR was used to replace the sequences downstream (IR-DR) or upstream (IR-UR) from cis element X (Fig. 3A). Since CBHI is the T. reesei enzyme that is most relevant for industry-scale production, we chose to perform cultivations, which are more closely related to biotechnological approaches than carbon source replacement experiments. Consequently, we cultivated the recombinant strains directly on medium containing glycerol (considered a neutral carbon source), xylan (a complex substrate that preferentially induces xylanase expression), carboxymethylcellulose (CMC; pure, soluble cellulose that preferentially induces cellulase expression), lactose (induces cellulase expression), or pretreated wheat straw (a very complex, renewable so-called "biowaste") as the carbon source.
GoxA activity was increased among all strains carrying a modified cbh1 promoter on all carbon sources compared to the GoxA activity in the strain bearing the native promoter (Fig. 3B). Surprisingly, the highest levels were observed on pretreated wheat straw, followed by CMC, xylan, and lactose as substrates (Fig. 3B). In addition to the promoter strength, the inducibility, i.e., the GoxA activity obtained on a certain carbon source compared to the GoxA activity on glycerol, was affected. The levels of inducibility of the native cbh1 promoter (strain Pc_WT) were 46.9 Ϯ 1.8 (mean Ϯ standard deviation), 14 Considering that the strong cbh1 promoter has a notably higher number of XBS than the weaker xyn1 promoter and that most of the strains investigated so far gained XBS by modifying the promoter, we decided to examine the question of whether the increase in the number of XBS alone causes the observed impacts on the inducibility and/or the strength of the cbh1 promoter.
Increasing distances of cis elements to TSS correlate positively with promoter inducibility. In order to investigate the influence of the number of XBS on the cbh1 promoter activity, we deleted the XBS from each cis element (termed ΔX, ΔY, and ΔZ, respectively) and all of them together (termed ΔXYZ). Figure 4A displays the modified promoters and provides the names of the corresponding strains that were investigated.
For all respective strains, the promoter activity decreased; however, it decreased to different extents depending on the particular cis element in which the XBS were deleted (Fig. 4B). What we observed was distinct losses of activity related to the lack of the native cis element in the following order: X, Y, Z (Fig. 4B). Notably, the pattern of induction (related to the carbon source used) remained the same. Finally, the strain that lacks the native cis element Z was only minimally inducible, and inducibility was completely lost in strain Pc_ΔXYZ, in which the XBS from all three cis elements are deleted (Fig. 4B). Obviously, the decrease in inducibility correlated directly with the distance from the cis element to the TSS of the reporter: the further upstream the cis element was located, the stronger was the decrease of promoter activity in the case of deletion of its XBS. Importantly, the strain lacking the native cis element Z had only a minimally higher GoxA activity than the strain lacking all three native elements, suggesting that just the number of XBS has a minor role compared to their position. Interestingly, the GoxA activity on the neutral carbon source glycerol is hardly affected by the deletion of XBS in any of these cis elements (Fig. 4B), which indicates a small impact of XBS in basal protein expression from this promoter. As we saw only a minor influence of the sheer number of XBS on promoter activity but a clear distance dependency, we decided to investigate whether the arrangement of cis elements can even overrule the number of XBS.
The configuration of cis elements is predominant over the sheer number of XBS. Concerning the impact of cis elements on promoter activity, we aimed to identify the exact hierarchy of the following properties: number of XBS, arrangement of XBS, and distance of XBS from TSS. Therefore, in strain Pc_ΔX, cis element Z was exchanged for cis element IR (resulting in strain Pc_ΔX_Z-IR), and in strain Pc_ΔZ, cis element X was replaced with cis element IR (resulting in strain Pc_X-IR_ΔZ) (Fig. 5A).
Comparing the GoxA activities of Pc_ΔX and Pc_ΔX_Z-IR with that of the strain bearing the native promoter, we observed a similar induction pattern (Fig. 5B). However, strain Pc_ΔX_Z-IR, which bears four fewer XBS than the native cbh1 promoter, had higher activity than strain Pc_WT. The highest increase in GoxA levels was observed on xylan (2.51-fold Ϯ 0.37-fold higher than in Pc_WT). Otherwise, strain Pc_ΔX, which has one more XBS than Pc_ΔX_Z-IR, could not reach the levels of Pc_WT (Fig. 5B). A similar observation was made in comparing Pc_ΔZ and Pc_X-IR_ΔZ: higher GoxA expression was observed when cis element X was exchanged for IR, even though the XBS number was lower. However, Pc_X-IR_ΔZ also did not reach the GoxA levels of Pc_WT. Finally, the best performing strain was Pc_ΔX_Z-IR, which has a rather low number of XBS. Interestingly, these XBS are located quite distant from the TSS. The promoter of strain Pc_X-IR_ΔZ, which bears the same number of XBS and the same cis elements but with the IR located close to the TSS, was much weaker. Besides this, GoxA expression on glycerol was again only slightly influenced by the promoter modifications (Fig. 5B).
These results suggest that the three factors arrangement of cis element, distance of cis element to TSS, and number of XBS clearly have a descending order in their influence on promoter inducibility. Finally, we aimed to get a more complete picture of the cbh1 promoter variants by combining all available promoter fragments (i.e., modifications in the 5= region and the 3= region).
The results for all possible cis element variations endorse the previous findings. To obtain a full picture of the cis element influence on the cbh1 promoter, we investigated cis element variations within the cbh1 promoter by combining all available 5= and 3= fragments (compare Table 2 below and Fig. 6). The respective recombinant strains were grown on xylan, lactose, and pretreated wheat straw.
On xylan, the GoxA expression of the majority of strains carrying a cbh1 promoter containing an additional cis element IR in the vicinity of cis element X (IR-UI, IR-UR, IR-DI, and IR-DR) was induced (Fig. 6A). In strains in which the cis element Z is exchanged for cis element IR (Z-IR), the induction was even stronger. The deletion of XBS in the cis elements led to a distance-dependent decrease in promoter activity (ΔX, ΔY, ΔZ, ΔYZ, and ΔXYZ). For the combinations of XBS deletions and the addition of cis element IR, we found a clear potential of the latter to compensate for the loss of XBS (Fig. 6A).
Similar observations were made on lactose (Fig. 6B). The addition of cis element IR and/or cis element exchanges increased the strength of the promoter, and the deletion of XBS from cis elements decreased the strength of the promoter in a manner dependent on the distance from the TSS. Overall, the effects were a bit less pronounced than on xylan.
Finally, strains were cultured on pretreated wheat straw and analyzed for cbh1 promoter strength (Fig. 6C). The addition of cis element IR and/or cis element exchanges increased the strength of the respective cbh1 promoter and led to very high levels of GoxA activity. Deletions of XBS from cis elements led to decreases in promoter activity. Again, the combinations of XBS deletions and the addition of cis element IR had a tendency to compensate for the loss of the native cis elements.
Summarizing the results from the latter detailed approach, which allowed us to investigate a greater number of variations of cis elements in the cbh1 promoter, this demonstrated the strong influence of the cis element IR on promoter strength. The comparison of the variations supported the finding that the number of XBS has only a secondary effect on promoter strength. All combinations of cis element deletions pointed to decreased promoter strength with increased distance from the TSS. Notably, the GoxA activity patterns of the strains carrying modified cbh1 promoter constructs are very constant when comparing different substrates.

DISCUSSION
In the past, relevant regions and motifs for the regulation of the xyn1 gene were identified (5,6). In this study, we present results demonstrating that the previously identified cis element IR, an inverted repeat of XBS, conveys transcriptional induction on xylan and wheat straw. We studied the role of this cis element by deleting only this specific motif and not by classic deletion analysis of the promoter. Phenotypic analysis was initially performed by carbon source replacement experiments because growth effects can be excluded and this allowed us to focus on the impact of a single carbon source on the promoter activity.
As Xyr1 is the essential transactivator not only of the xyn1 gene but also of the industrially more relevant cbh1 gene (4), we extended our study to investigate the cis element IR in both promoters. In the case of the cbh1 promoter, Ilmén and coworkers performed a first deletion mapping (23) and found that CCR was abolished when a sequence upstream from Ϫ500 bp was removed and that the modification of a DNA motif at Ϫ720 bp was enough to trigger derepression. Furthermore, they found that sophorose induction still worked when the promoter sequence upstream from position Ϫ161 bp was removed. Years later, two reports on the cbh1 promoter were published by Liu et al. (15) and Zou et al. (16): the first study inserted the sequence stretch from Ϫ620 to Ϫ820 bp of the cbh1 promoter multiple times in the 5= end of this promoter, leading to increased reporter activity. In the other study, Cre1-binding sites were exchanged with binding sites for Ace2 or the Hap2/3/5 complex, which also resulted in increased reporter gene expression. In both studies, the promoter-reporter constructs were randomly integrated into the genomes of Rut-C30 descendant strains. In the present study, we used targeted integration at the pyr4 locus of the wild-type T. reesei strain QM6a (22) in order to ensure genotypic comparability of all recombinant strains. Besides this, we specifically modified the cis elements instead of deleting or inserting longer sequence stretches. This strategy has the advantage that the core promoter does not become shorter and the effects of unknown regulatory sequences are not observed unintentionally.
In Saccharomyces cerevisiae, Gal4 binds inverted repeats of the sequence 5=-CGG (N 11 )CCG-3= as a dimer in the upstream activating sequence (UAS) of the gal1 gene (24). The Gal4 DNA-binding consensus is very conservative concerning the spacer between the inverted repeats. In the present study, we investigated binding sites that are contacted by Xyr1, a Gal4-like transcription factor. Xyr1 is also supposed to bind DNA as a dimer; however, the spacer between the target sequences does not seem to be restricted (25). As Ace1 contacts DNA sequences similar to those contacted by Xyr1, there could also be heterodimerization involved in the transcriptional regulation. Strikingly, in the present study, the exchange of a direct for an inverted repeat of XBS caused different substances to become strong inducers of gene expression (e.g., wheat straw in the case of the cbh1 gene). From a scientific point of view, it is an interesting consideration that genes whose expression is triggered by different inducing substances, although activated by the same transactivator, respond to another inducer as a result of another arrangement of the binding sites of the transactivator. And from an application-oriented point of view, this finding could support the reduction of operating costs by allowing cheaper, more easily available or more sustainable inducer substances to be used.
Murphy and coworkers investigated the impact of the position of cis elements by modifying the gal1 promoter in S. cerevisiae (26). They inserted one to three tetO 2 operators between the TATA signal and the TSS and analyzed the influence of the different promoters on yeast eGFP expression. This approach involved the addition of different concentrations of an inhibitor of the TetR repressor protein. When the tetO 2 operator was inserted near the TSS, the promoter was active even at low concentrations of the TetR inhibitor. Otherwise, when the tetO 2 operator was inserted near the TATA signal, the promoter became active only at higher concentrations of the inhibitor. Even if the strength of the promoter did not change significantly, the inducibility of the systems concerning dependence on dosage was modified. In our study, modifying the positions of relevant cis elements enhanced the general strength and the inducibility of the promoter. This might be due to the fact that we changed the positions of binding sites for a transactivator, while in S. cerevisiae, a repressor/inhibitor system was investigated. Blazeck and coworkers investigated hybrid promoters in S. cerevisiae (27). They used synthetic promoters composed of five different core promoters (of the cyc, gal1, gpd, leu3, and tef genes) and fused the UASs of the cit1, clb2, gal1, and tef genes to their 5= ends. The hybrid promoter combining the leu3 core promoter and three cbl2 UASs was 10 times stronger than the native leu3 promoter. They also complemented the leu3 promoter with inhibitory regions or Gal4p-binding sites. Altogether, they observed only active induction in the case of promoter-reporter strains containing the Gal4p-binding sites. From these and our findings, it can be assumed that for the aim of increasing promoter strength, the targeting of binding sites for the respective transactivator is the most straightforward strategy.
It should be noted that the reporter gene goxA, which was used in this study, originates from another filamentous fungus (namely, A. niger). Even if we cannot safely assume that any other heterologous target protein can be produced with the same efficiency as GoxA, the strategy of using the cbh1 promoter and T. reesei as the host organism suggests itself as a promising one. In particular, the option to use wheat straw as a cheap and sustainable substrate to provide high yields of a target protein adds attractiveness to this approach.
For carbon source replacement experiments, mycelia were precultured in 200 ml MA medium supplemented with 0.1% peptone and 1% (wt/vol) glycerol as the sole carbon source on a rotary shaker at 180 rpm at 30°C for 24 h. A total of 10 9 conidia per liter (final concentration) was used as the inoculum. Pregrown mycelia were washed, and equal amounts were resuspended in 20 ml MA medium containing either no carbon source or 1% (wt/vol) glycerol, 1% (wt/vol) D-glucose, 0.5 mM D-xylose, 1% (wt/vol) D-xylose, or 1.5 mM sophorose as the sole carbon source and were incubated for 8 h. Culture supernatants were derived from biological duplicates.
For direct cultivation experiments, mycelia were cultured in 20 ml MA medium supplemented with 0.1% peptone and either 1% (wt/vol) glycerol, 1% (wt/vol) lactose, 1% xylan (wt/vol) (Lenzing AG, Lenzing, Austria), 1% (wt/vol) carboxymethylcellulose, or 1% (wt/vol) pretreated wheat straw as the sole carbon source on a rotary shaker at 180 rpm at 30°C for 48 h if not indicated otherwise in the figure legends. The mycelia growing on pretreated wheat straw were incubated for 72 h. Samples were derived from biological duplicates.
Escherichia coli strain Top10 (Invitrogen, Life Technologies, Paisley, United Kingdom) was used for all cloning purposes throughout this study and maintained on LB at 37°C. If applicable, ampicillin or spectinomycin was added to a final concentration of 100 mg/ml or 100 mg/ml, respectively.
Plasmid construction. PCRs for cloning purposes were performed with Phusion high-fidelity DNA polymerase (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer's instructions. All primers used are listed in Table 1.
For the construction of pDK/Pcbh1_ΔX, two PCR products were created with primers pcbh1_fw_Cfr and pcbh1_delta_x_rv_(soe), as well as pcbh1_delta_x_fw_(soe) and pcbh1_rv_Bam-Nhe, using chromosomal DNA from T. reesei QM6a as the template. Both fragments were purified using an agarose gel extraction kit (Thermo Fisher Scientific). Thereafter, an SOE PCR was performed with primers pcbh1_fw_Cfr and pcbh1_rv_Bam-Nhe using the two fragments as the template. The resulting fragment   For determination of the biomass, the mycelium was separated from the culture supernatant by filtration using a predried piece of Miracloth and subsequently dried at 120°C for 24 h. Table S4 provides the biomass of strains grown on lactose. In the case of water-insoluble substrates, NaOH-soluble protein was determined and was used as an alternative reference to biomass.
NaOH-soluble protein extraction. Mycelia were separated from the culture supernatant by filtration (supernatant was used for glucose oxidase assays). Afterwards, each mycelium sample was washed using 30 ml deionized H 2 O and filtered again. The extraction of NaOH-soluble protein from the remaining biomass was performed as previously described (32). Twenty milliliters of 1 M NaOH was added to each mycelium sample and then fixed in a rotary shaker for 3 h. Samples were transferred into 1.5-ml tubes and centrifuged for 10 min at 20,000 ϫ g to separate the extracted protein from the remaining solids. The protein concentration of the extract was determined using the Bio-Rad protein assay (Bio-Rad) against a bovine serum albumin (BSA) standard. Tables S2, S3, and S5 provide the concentrations of NaOH-soluble protein for T. reesei QM6a and recombinant strains cultivated on water-insoluble substrates.
Accession number(s). The data sets generated and analyzed during the current study are available in the GenBank repository (https://www.ncbi.nlm.nih.gov/GenBank/) under accession numbers KY780443 through KY780480; further information can be found in Table 3.

SUPPLEMENTAL MATERIAL
Supplemental material for this article may be found at https://doi.org/10.1128/AEM .01742-17.