Differential Involvement of β-Glucosidases from Hypocrea jecorina in Rapid Induction of Cellulase Genes by Cellulose and Cellobiose

Abstract

Appropriate perception of cellulose outside the cell by transforming it into an intracellular signal ensures the rapid production of cellulases by cellulolytic Hypocrea jecorina. The major extracellular β-glucosidase BglI (CEL3a) has been shown to contribute to the efficient induction of cellulase genes. Multiple β-glucosidases belonging to glycosyl hydrolase (GH) family 3 and 1, however, exist in H. jecorina. Here we demonstrated that CEL1b, like CEL1a, was an intracellular β-glucosidase displaying in vitro transglycosylation activity. We then found evidence that these two major intracellular β-glucosidases were involved in the rapid induction of cellulase genes by insoluble cellulose. Deletion of cel1a and cel1b significantly compromised the efficient gene expression of the major cellulase gene, cbh1. Simultaneous absence of BglI, CEL1a, and CEL1b caused the induction of the cellulase gene by cellulose to further deteriorate. The induction defect, however, was not observed with cellobiose. The absence of the three β-glucosidases, rather, facilitated the induced synthesis of cellulase on cellobiose. Furthermore, addition of cellobiose restored the productive induction on cellulose in the deletion strains. The results indicate that the three β-glucosidases may not participate in transforming cellobiose beyond hydrolysis to provoke cellulase formation in H. jecorina. They may otherwise contribute to the accumulation of cellobiose from cellulose as inducing signals.

INTRODUCTION

Cellulose is a linear polymer of β-1,4-linked glucose molecules piled up into highly ordered fibrillar structures. As the most abundant biomass from the plant cell wall, its microbial decomposition not only plays a key role in the carbon cycle in nature but also provides great potential for a number of applications, most notably biofuel production (17, 18). Since its initial isolation as a decomposer of cellulosic materials, Hypocrea jecorina (anamorph Trichoderma reesei) has been developed into one of the most prolific cellulase producers in industry. The cellulase mixture of H. jecorina has been shown to consist of at least three types of enzymes that act upon the insoluble substrate to achieve the efficient conversion of native crystalline cellulose to glucose (18). Among others, it has been widely accepted that endoglucanases (EC 3.2.1.4) and exoglucanases (EC 3.2.1.91) act synergistically on cellulose to produce mainly cellobiose. This disaccharide and other cellooligosaccharides are further hydrolyzed to glucose by β-glucosidases (EC 3.2.1.21).

Although the cellulases have been extensively characterized for H. jecorina, the regulation of cellulase production in H. jecorina is still insufficiently understood. For rapid launching of the cellulase machinery, H. jecorina has to either sense the presence of the insoluble cellulose outside the cell or detect the substrate by uptake of the degradation products (14–16, 20). Although the precise nature of the true “inducer” has been elusive, several lines of evidence pointing to a role for β-glucosidases in the rapid induction of the cellulase genes have been presented. First, slow feeding of cellobiose as the sole carbon source or inhibition of extracellular hydrolysis of cellobiose by β-glucosidase leads to cellulase formation (7, 15, 35). Second, it has been shown that the absence of the extracellular β-glucosidase (BglI) results in a delay in induction of the cellulase genes (6), while recombinant strains bearing multiple copies of the bgl1 gene display enhanced cellulase induction not only by cellulose but also by sophorose, a potent inducer potentially formed by the plasma membrane-bound β-glucosidase activity from cellulose degradation products via transglycosylation reactions (20, 33). Both extracellular and intracellular β-glucosidases have been reported to exist in H. jecorina (3, 10, 20). While BglI, belonging to glycosyl hydrolase family 3 (GH3), has been considered to account for the majority of extracellular and cell-wall-bound activities, BglII (CEL1a), belonging to GH family 1 (GH1), has been shown to be intracellularly localized (27). Moreover, five other β-glucosidase sequences have been identified (5). Like bgl1, cel1a and cel1b are among the transcripts highly induced upon growth on cellulose or sophorose (5). However, their significance for cellulase gene regulation has not yet been investigated.

We report here some enzymatic properties and cellular localization of a second GH1 β-glucosidase (CEL1b). We further report disruptions of the major extracellular and intracellular β-glucosidase gene loci in H. jecorina. The corresponding knockout strains were used to investigate the contribution of these β-glucosidase activities to the induction of cellulase genes by cellulose and cellobiose.

MATERIALS AND METHODS

Strains, plasmids, and medium.

Hypocrea jecorina QM9414 (ATCC 26921) and its uridine-auxotrophic derivative TU-6 with a mutant pyr4 (ATCC MYA-256 [8]) were maintained on malt extract agar (Sigma) supplemented with 10 mM uridine when necessary. Strains were grown in 1-liter Erlenmeyer flasks on a rotary shaker (200 rpm) at 30°C in the medium as described by Mandels and Andreotti (21). Carbon sources were used at a final concentration of 10 g liter⁻¹. Escherichia coli DH5α was used for routine gene cloning and vector construction.

For expression of CEL1b and its mutant derivatives in Escherichia coli, the coding sequence of cel1b was amplified from the cDNA of H. jecorina with primers harboring EcoRI and HindIII sites and ligated into pET32a(+) after it was digested with the same enzymes to obtain pET32acel1b. All the relative mutants were obtained by oligonucleotide-mediated mutagenesis of the cel1b gene using a two-step fusion PCR with pET32acel1b as the template. The mutated sites were verified by sequencing before being subcloned into pET32a. For expression of enhanced green fluorescent protein (EGFP)-tagged CEL1b (CEL1b-EGFP) in H. jecorina to determine the subcellular localization of CEL1b, the cel1b gene was inserted into the NcoI site of pIG1783 and fused in frame with the egfp coding sequence to obtain pIGcel1b (26). Oligonucleotides, including gene-specific primers used in this study for plasmid constructions, gene deletion, or probe preparation, are listed in Table S1 in the supplemental material.

For the transcript and secreted protein analysis, strains were pregrown on glycerol (1% vol/vol) for 48 h. Mycelia were harvested by filtration and washed twice with medium without a carbon source. Equal amounts of mycelia were transferred to a fresh medium containing the respective carbon sources, including Avicel cellulose or cellobiose without peptone, and incubation was continued for the indicated time period. For resting cell cultivations, H. jecorina was pregrown on glycerol medium and then washed extensively with the minimal medium lacking a carbon source and resuspended in the replacement medium lacking nitrogen (and therefore enabling no growth) as previously described, except that sophorose was used at a final concentration of 1 mM (29).

Production of recombinant CEL1b in E. coli.

For purification of CEL1b, an E. coli strain with the cel1b expression construct was grown at 37°C until the optical density at 600 nm (OD₆₀₀) reached 0.5 to 0.6. Isopropyl-β-d-thiogalactopyranoside (IPTG) at a final concentration of 100 μM was added, and the incubation was continued for 16 h at 20°C. The CEl1b protein was purified with Ni-nitrilotriacetic acid-agarose (Qiagen) essentially according to the instructions of the manufacturer (Qiagen).

Disruption of the cel1a, cel1b, and bgl1 genes of H. jecorina.

The 2.7-kb pyr4 fragment released from pFGI with either XbaI/SalI or SpeI/SalI was ligated into the same sites of pUC19 to obtain pUCpyr4 or pUCpyr4-1. The two 2-kb fragments upstream of the ATG codon and downstream of the stop codon of cel1a or cel1b, respectively, were amplified from H. jecorina chromosomal DNA and ligated into the corresponding sites to yield the disruption vector pUCcel1apyr4 or pUCcel1bpyr4.

The 6.7-kb fragments containing the complete pyr4 gene plus the 5′ and 3′ regulatory sequences of cel1a or cel1b were released from pUCcel1apyr4 or pUCcel1bpyr4 with EcoRI/HindIII or XbaI/HindIII and were used to transform H. jecorina TU6.

To construct the Δcel1a Δcel1b deletion strain, a 3.0-kb amdS fragment released from pALK424 by SpeI and SalI digestion was introduced into pUC19 to generate pUCamdS (12). The 2.0-kb 5′ and 3′ regulatory fragments of the cel1b gene were inserted into pUCamdS, resulting in the cel1b deletion vector pUCcel1bamdS. The deletion cassette was further released from pUCcel1bamdS by XbaI and HindIII digestion and used to transform the Δcel1a strain in the same way as for TU6 except that the transformants were selected on plates containing acetamide as the sole nitrogen source. Simultaneous deletions of the BglI-encoding gene cel3a were made by further disrupting cel3a in the Δcel1a Δcel1b strain. A hygromycin resistance cassette containing the gpd (encoding glyceraldehyde-3-phosphate dehydrogenase) promoter from H. jecorina and hygromycin resistance gene from pRLMex30 (19) was ligated into the NotI and SpeI sites of pUC19 to obtain pUChph. The 2.0 kb of 5′ and 3′ flanking sequences of the cel3a gene were successively inserted into pUChph, resulting in pUCcel3ahph. After linearization with SacI, the disruption vector was used to transform the Δcel1a Δcel1b strain. Transformants were selected on minimal medium containing 100 μg/ml hygromycin. To complement the Δcel1a Δcel1b strain with CEL1b (I174C), the cel1b disruption vector was digested with SpeI and StuI to remove the pyr4 gene, followed by ligation with the hygromycin resistance cassette digested with the same enzymes to create pUCcel1bhph. This plasmid was digested with SpeI, dephosphorylated using shrimp alkaline phosphatase (TaKaRa), and further ligated with the coding region for Cel1b (I174C) amplified from pET32acel1b (I174C). After linearization with HindIII, the fragment was transformed into the Δcel1a Δcel1b strain.

Transformation of H. jecorina was carried out essentially as described by Penttila (25). Transformants were selected on minimal medium either for uridine prototrophs or for resistance to hygromycin and acetamide.

Fluorescence microscopy.

For visualization of CEL1b-EGFP, spores of recombinant strains harboring chromosome-integrated pIGcel1b were inoculated into minimal medium and grown for 20 h at 30°C. Mycelia were used directly for microscopic observation. To simultaneously stain the nuclei, 100 μg ml⁻¹ of 4′, 6-diamidino-2-phenylindole (DAPI) dihydrochloride solution in 50% glycerol was added. Fluorescence was detected with a Nikon Eclipse 80i fluorescence microscope, and images were captured and processed with the NIS-ELEMENTS AR software program.

Nucleic acid isolation and hybridization.

Fungal mycelia were harvested by filtration, washed with tap water, and frozen in liquid nitrogen. Fungal genomic DNA was isolated according to the instructions of the E.Z.N.A. fungal DNA miniprep kit (Omega Biotech, Doraville, USA). Total RNA was isolated using the TRIzol reagent (Invitrogen) according to the manufacturer's protocol. Southern hybridization and Northern analysis were performed with the digoxigenin nonradioactive system from Roche Applied Science as described previously (9). Relative transcription levels were analyzed semiquantitatively by densitometry using the software program ImageJ (http:///rsb.info.nih.gov/ij). The values were normalized by densitometry of the cbh1 signal to that of the 18S RNA control.

Quantitative reverse transcription-PCR (RT-PCR).

The total RNA was further purified using a Turbo DNA-free kit (Ambion) according to the manufacturer's instructions. Reverse transcription was carried out using the PrimeScript RT reagent kit (TaKaRa) according to instructions. Quantitative PCRs were performed using a Bio-Rad myIQ2 thermocycler (Bio-Rad) and the SYBR green supermix (TaKaRa). Reactions were carried out in triplicate with a total reaction volume of 20 μl, including 300 nM (each) forward and reverse primers and 100 ng template cDNA. Data analysis was performed using the relative quantitation/comparative threshold cycle (C_T) (ΔΔC_T) method, and results were normalized to those for the endogenous control actin with expression on glycerol as the reference sample.

Enzyme activity measurements.

β-Glucosidase activity was determined by measuring the amount of p-nitrophenol released from p-nitrophenyl-β-d-glucopyranoside (pNPG) (Sigma), used as the substrates. The transglycosylation activity of CEL1b was measured as described previously (27). Cellobiohydrolase (CBH) activities in the culture supernatants were measured using p-nitrophenol-d-cellobioside (pNPC) (Sigma) as a substrate (4). Cellular extracts used for assay of intracellular β-glucosidase activity were prepared as follows: H. jecorina strains were grown on Mandels-Andreotti medium with Avicel or glycerol as the carbon source. The mycelia were harvested and washed twice with 0.9% NaCl. Lysates were prepared by grinding the mycelia into fine powder under liquid nitrogen, which was then suspended in 50 mM Na-phosphate buffer (pH 7.0) with protease inhibitors. Cell debris was removed by centrifugation at 14,000 × g for 10 min. Determination of total Avicel hydrolysis activity was performed at 50°C with 1% Avicel in 50 mM sodium acetate (pH 5.0) with equal amounts of extracellular culture filtrates relative to biomass. Culture broth supernatant was buffer exchanged using a 10-kDa-molecular-mass-cutoff centrifugal filter to remove any soluble sugars prior to initiating hydrolysis experiments. Sugars released were determined by the dinitrosalicylic acid (DNS) method of Miller et al. with d-glucose as a standard (23).

HPLC analysis.

Analysis of in vitro transglycosylation activity by high-performance liquid chromatography (HPLC) was performed as follows. Samples were desalted by mixing with bed resin TMD-8 (Sigma) and vortexing for 1 min. Salt-free supernatants were applied on a Bio-Rad Aminex HPX-42A carbohydrate column and analyzed by using an LC-10AD HPLC (Shimadzu, Japan), equipped with a RID-10A refractive index detector. The column was maintained at 75°C and eluted with double-distilled water at a flow rate of 0.4 ml/min.

Protein analysis.

SDS-PAGE and Western blotting were performed according to standard protocols (28). Total secreted and intracellular proteins were determined using the method of the Bradford protein assay. Detection of the cellobiohydrolase CBH1 was performed by immunoblotting using a polyclonal antibody raised against amino acids (426 to 446) of CBH1 (1).

RESULTS

CEL1b is an intracellular β-glucosidase with transglycosylation activities.

CEL1a and CEL1b are members of glycosyl hydrolase family 1, and both of the corresponding genes are among the highly induced β-glucosidase genes in H. jecorina growing on cellulose or on sophorose (5). CEL1a has been shown to be an intracellular enzyme displaying not only hydrolytic properties but also transglycosylation activity in vitro (27). Amino acid sequence comparison of CEL1b with CEL1a and two other characterized GH1 β-glucosidases from Phanerochaete chrysosporium revealed relatively high sequence identity (53% between CEL1a and CEL1b) (see Fig. S1 in the supplemental material). Both CEL1a and CEL1b, as well as BglI, also showed significant homology to the recently identified GH1-1, GH3-3, and GH3-4 proteins of Neurospora crassa, respectively (see Fig. S3 and S4). The modeled structure of CEL1b, with the determined structure of CEL1a as the template, showed that the overall structure of CEL1b was very similar to that of CEL1a, forming a classical (β/α)₈ barrel with the active site being located at the bottom of the pocket (see Fig. S2) (11). Specifically, while the surrounding residues at the glycone binding site are highly conserved, residues around the aglycone binding site, which have been proposed to be the basis of substrate specificity (24), are highly divergent. Residues with smaller side chains at the aglycone site have also been observed in BGL1B from P. chrysosporium, which is more efficient than BGL1A in hydrolysis of cellobiose (32). In order to characterize the role of amino acids around the aglycone binding site in enzymatic activity of CEL1b, several mutants with a directed change of these divergent amino acids either to residues with smaller side chains (I174 to C and H265 to A) or to the corresponding residues in CEL1a or BGL1B (Y187 to L, D242 to H, and S442 to A) were made and purified from soluble extracts of E. coli Origami B (DE3) (Fig. 1A). All of the purified proteins exhibited hydrolytic activity toward p-nitrophenyl-β-d-pyranoside (pNPG), though there were apparent differences in the kinetic parameters, as summarized in Table 1. Among others, change of Ile 174 to Cys significantly increased the hydrolytic efficiency, probably due to a wider entrance of the catalytic site that resulted from the smaller side chain of cysteine. These results suggest that residues around the entrance of the catalytic site of CEL1b may not only be involved in determining the substrate specificity but also play a role in affecting the overall hydrolytic efficiency. Similar to CEL1a, recombinant CEL1b exhibited significant transglycosylation activity when incubated with relatively high concentrations of glucose and cellobiose (Fig. 1B and C). A maximal yield of 189.6 mg liter⁻¹ was achieved for cellotriose when CEL1b was incubated with 20% cellobiose. None of the mutants displayed higher activity of transglycosylation than the wild type (WT) (data not shown). Note that we cannot exclude the possibility of the existence of sophorose in the transglycosylation products. To determine the subcellular localization of CEL1b, the CEL1b-EGFP fusion protein was expressed from a constitutive gpd promoter in the H. jecorina QM9414 strain. CEL1b-EGFP fluorescence was readily observed to be dispersed throughout the cytosol (Fig. 1D). Taken together, these results indicate that CEL1b is mainly an intracellular β-glucosidase with in vitro transglycosylation activities.

Fig 1.

Enzymatic characterization and subcellular localization of CEL1b. (A) SDS-PAGE analysis of purified CEL1b and its mutants produced in E. coli. M, molecular mass standard; lane WT, WT CEL1b; lane 1, CEL1b (I174C); lane 2, CEL1b (Y178L); lane 3, CEL1b (D242H); lane 4, CEL1b (H265A); lane 5, CEL1b (S442A). (B and C) Recombinant CEL1b exhibited transglycosylation activity. Purified CEL1b was incubated with glucose (20% and 40% [wt/vol]) or cellobiose (10% and 20% [wt/vol]). Reaction products were analyzed by HPLC. G1, glucose; G2, disaccharide; G3, trisaccharide; G4, tetrasaccharide. (D) CEL1b was mainly localized in the cytoplasm. EGFP was used as a reporter to visualize subcellular localization of CEL1b. Panels, from left to right, show phase-contrast image of mycelia, a fluorescent image of GFP-fused CEL1b, and nuclei visualized by DAPI staining.

Table 1.

Kinetic parameters of different β-glucosidases and CEL1b mutants^a

Enzyme	Vmax (nmol/[liters · min])	Km (mmol/liter)	Kcat (s−1)	Kcat/Km (s−1 mmol−1 liter)	Reference
CEL1b	35.5	0.44	0.57	1.31	This study
CEL1b (I174C)	2,903.73	1.88	31.55	16.82	This study
CEL1b (Y178L)	28.29	0.14	0.38	2.65	This study
CEL1b (D242H)	106.27	0.59	1.57	2.68	This study
CEL1b (H265A)	13.84	0.10	0.20	1.91	This study
CEL1b (S442A)	22.59	0.23	0.46	1.98	This study
CEL1a	NA	2.22	34.8	NA	31
CEL3a (BglI)	NA	0.38	87.9	NA	2

^aNA, not available.

Disruption of cel1a and cel1b compromises induced production of cellulases.

To probe further into the in vivo function of CEL1a and CEL1b, mutants of H. jecorina lacking the coding sequences of cel1a or cel1b were obtained by targeted gene replacement (Fig. 2A). A Δcel1a Δcel1b strain that lacks cel1a and cel1b was also constructed. Anchored PCR and Southern blot analysis confirmed that each deletion event had occurred as predicted, integrating only once within the H. jecorina genome, resulting in the removal of the expected coding sequences while leaving the flanking sequences intact (Fig. 2B). Analysis of the intracellular β-glucosidase activity demonstrated that while the parent and Δcel1b strains exhibited only a slight difference in pNPG hydrolytic activity, the absence of CEL1a resulted in a dramatic decrease in the intracellular β-glucosidase activities in Δcel1a and Δcel1a Δcel1b strains to about 25% that of the wild-type strain when cultured on cellulose (Fig. 3A). The simultaneous absence of CEL1a and CEL1b also led to a significant upregulation of the extracellular β-glucosidase activities over a longer period of induction (12 h). To further test the effect of the absence of intracellular CEL1a and CEL1b on cell growth on different carbon sources, WT and mutant strains were incubated on plates containing glucose or cellobiose. While there was hardly any difference in growth between the parent and mutant strains on glucose, disruption of cel1a and cel1b apparently affected the growth of H. jecorina on cellobiose (Fig. 3B). Total extracellular protein production and cellobiohydrolase (CBH) activity were further measured with the parent and Δcel1a Δcel1b strains with Avicel as the carbon source. Similar to results observed for bgl1-deleted strain (6), the initial rate of both extracellular protein production and exoglucanase activity was lower for the mutant strain than for the parental strain (Fig. 3C and D). Together, these results indicate that intracellular CEL1a and CEL1b not only participate in the metabolism of cellulose degradation products, including cellobiose, but may also be involved in efficient induction of the cellulases.

Fig 2.

Targeted disruptions of cel1a, cel1b, and cel3a. (A) Schematic illustration of the homologous integration of the H. jecorina pyrG gene at the cel1a or cel1b gene locus, resulting in the deletion of the coding sequences. (B) Southern blot analysis of the Δcel1a, Δcel1b, Δcel1a Δcel1b, and ΔtriβG strains. Genomic DNA was digested with SacI (Δcel1a), PstI (Δcel1b andΔcel1a Δcel1b), and NcoI (ΔtriβG) prior to electrophoresis and probed with a 2-kb DNA fragment upstream of the ATG codons of cel1a, cel1b, and cel3a, respectively. Bands of 3.1 kb (lane 2), 3.5 kb (lane 4), 4.1 kb (lane 6), and 3.6 kb (lane 8), corresponding to the sizes expected for the disrupted genes in Δcel1a, Δcel1b, Δcel1a Δcel1b, and ΔtriβG, replaced the parental 6.9 kb (lane 1), 6.6 kb (lane 3 and lane 5), and 2.6 kb (lane 7), respectively.

Fig 3.

Disruption of cel1a, cel1b, and cel3a affected growth of H. jecorina on cellobiose as well as cellulase production on Avicel. (A) Intracellular and extracellular β-glucosidase activities in the parental and disruption strains. H. jecorina strains were precultured on glycerol for 48 h and then transferred to Avicel. Intracellular and extracellular β-glucosidase activities were measured using soluble cellular extract and extracellular proteins of culture filtrates, respectively, with pNPG as the substrate after 6 h and 12 h of induction. Data shown are the means for three independent experiments. (B) Disruption of cel1a, cel1b, or cel3a resulted in slower growth on cellobiose. Strains were incubated on plates containing either glucose or cellobiose as the carbon source. (C and D) Extracellular protein production and exoglucanase activity from culture supernatant of WT and mutant H. jecorina strains grown in the presence of 1% (wt/vol) Avicel. Data shown are the means for three independent experiments.

The absence of CEL1a and CEL1b results in a delay in the induction of cellulase gene transcription by cellulose.

To test whether CEL1a and CEL1b exert their effect on cellulase production at the transcriptional level, we examined the endogenous cbh1 mRNA by Northern blotting (Fig. 4). This analysis showed that in comparison to the rapid induction in the WT strain, which occurred as early as 3 h upon induction by cellulose, productive activation of transcription of cbh1 was delayed by about 3 h and 21 h in the Δcel1b and Δcel1a strains, respectively (Fig. 4A to C). This lag in gene expression was further extended to 36 h in the Δcel1a Δcel1b strain, although the final level of transcription was almost the same (Fig. 4D). Correspondingly, secretion of CBH1 into the culture supernatant was also delayed in strains with deletions of cel1a and cel1b as assayed by Western blotting (Fig. 4E). Complementation of the Δcel1a Δcel1b strains with CEL1b (I174C) significantly increased the intracellular pNPG hydrolytic activity and partially improved the induction kinetics to a degree more rapid than that of the Δcel1a strain (Fig. 4F). Notably, the observed delay in transcription induction cannot be attributed to the differential growth on cellulose caused by the potentially compromised metabolism of cellulose degradation products in mutants, because a similar delay in gene expression was observed in a resting-cell inducing system (Fig. 5A and B). Surprisingly, analysis of cbh1 transcription on induction with lactose revealed that a significant lag in cbh1 transcription also existed in the absence of cel1a compared with that for the wild-type strain (Fig. 5C and D).

Fig 4.

The absence of ce11a and cel1b resulted in delayed transcription of the cbh1 gene on cellulose induction. H. jecorina strains, including the WT (A), Δcel1a (B), Δcel1b (C), and Δcel1a Δcel1b (D) strains, were precultured on glycerol for 48 h and then transferred to the same medium containing 1% (wt/vol) Avicel instead of glycerol. One microgram of total RNA was electrophoresed and blotted onto Hybond N⁺ nylon membranes (Amersham). cbh1 mRNA and 18S RNA was probed at different times after Avicel induction. The values below the panels indicate the ratio of the intensity of the cbh1 signal as measured by densitometry to that of the 18S RNA control. (E) Western blot analysis of CBHI secreted into the culture supernatant of WT and deletion mutant strains on 1% Avicel. Equal amounts of culture supernatant relative to biomass were loaded for all strains. (F) Northern blot analysis of cbh1 mRNA and 18S RNA from the Δcel1a Δcel1b strain complemented with CEL1b (I174C) on 1% (wt/vol) Avicel. Quantitation was performed as for panels A to D.

Fig 5.

Productive transcription was compromised in a resting-cell inducing system but was restored by sophorose. Northern blot analysis of cbh1 mRNA and 18S RNA from the WT (A) or Δcel1a Δcel1b (B) strain was performed in a resting-cell inducing system with Avicel (1% [wt/vol]) as the carbon source. Northern blot analysis of cbh1 mRNA and 18S RNA was also performed for the WT (C) or Δcel1a (D) strain after the precultured mycelia were transferred to 1% (wt/vol) lactose. (E and F) Northern blot analysis of cbh1 mRNA from the WT and Δcel1a Δcel1b strains. H. jecorina strains were precultured on glycerol for 48 h and then washed extensively with minimal medium lacking a carbon source and resuspended in the same medium with 1 mM sophorose; incubation was continued for the indicated time period before RNA was extracted for analysis. The values below the panels indicate the ratio of the intensity of the cbh1 signal as quantitated in Fig. 4.

The disaccharide sophorose has been assumed to be a putative natural inducer for cellulose-mediated induction which has been detected in culture fluids of H. jecorina (13, 34). To examine whether intracellular CEL1a and CEL1b would otherwise participate in the formation of the inducer, we investigated the effect of sophorose on the efficacy of cbh1 induction (Fig. 5E and F). Similarly efficient cbh1 transcription occurred in both the wild-type and the Δcel1a Δcel1b strains in response to sophorose. The findings suggest that intracellular CEL1a and CEL1b contribute to the induction of cellulase genes, probably through participating in the formation of cellulase inducer. It has been reported that extracellular β-glucosidase BglI may also be involved in the formation of such an inducer (6, 20) (Table S2).

Deletion of bgl1 in a cel1a and cel1b mutant strain causes induction on cellulose to deteriorate further.

To probe further into the relationship between BglI and the intracellular β-glucosidases in modulating cellulase induction, we deleted bgl1 in the Δcel1a Δcel1b strain (ΔtriβG) on the assumption that the absence of these three β-glucosidases would not only significantly slow down the hydrolysis of cellobiose but also eliminate any putatively associated activities transforming cellobiose. Analysis of β-glucosidase activity demonstrated that the ΔtriβG strain displayed a dramatic decrease in extracellular β-glucosidase activities compared with those of the wild-type and Δcel1a Δcel1b strains, though its growth on cellobiose was retarded only to a level similar to that of the cel1a and cel1b mutants (Fig. 3A and B). Analysis of secreted CBH1 in the presence of cellulose revealed that productive formation of CBH1 was further compromised in the ΔtriβG strain compared to the Δcel1a Δcel1b strain, with a slower kinetics and a lower level of CBH1 secretion (Fig. 6A). Quantitative RT-PCR analysis of cbh1 mRNA showed that a similar lag in cbh1 expression occurred in the ΔtriβG strain (Fig. 6B). This result was further corroborated by both the significantly lower cellobiohydrolase activities as measured by pNPC hydrolysis and the smaller amount of secreted protein in the ΔtriβG cultures over the period of induction by Avicel (Fig. 3C and D). These results indicate that the major extracellular and intracellular β-glucosidases act together to ensure the efficient production of cellulases on Avicel.

Fig 6.

Simultaneous absence of BglI further compromised CBH1 expression, and cellobiose efficiently induced cellulase production, in β-glucosidase deletion strains. (A) CBH1 in culture filtrates of WT, Δcel1a Δcel1b, and ΔtriβG strains after induction with Avicel (1% [wt/vol]) was analyzed by Western blotting. Right panel, CBH1 was quantitated by scanning densitometry of the developed membranes. (B) Gene expression of cbh1 as analyzed by quantitative RT-PCR after induction with Avicel for different time periods. Relative gene expression levels were normalized to 1 when incubation with glycerol was done. Expression levels of actin were used as an endogenous control in all samples. Error bars indicate 1 SD. (C) Western blot analysis of CBHI secreted in culture filtrates of the WT, Δcel1a Δcel1b, and ΔtriβG strains after induction with cellobiose (0.25% [wt/vol]). Equal amounts of culture filtrates relative to biomass were loaded for all strains. Right panel, CBH1 was quantitated by scanning densitometry of the developed membranes.

Cellulase gene induction by cellobiose is not compromised in the absence of β-glucosidases.

As the major end product of cellulose hydrolysis by the synergistic action of endoglucanases and cellobiohydrolases, cellobiose has been shown to be able to induce cellulase production. This has been achieved by simultaneous addition of inhibitors of β-glucosidase or by slow feeding of cellobiose (7, 30). It has been assumed that cellobiose is transformed to sophorose by the membrane-bound β-glucosidase and endoglucanases via transglycosylation (27, 33, 34). To further test whether the induction defect as observed in the β-glucosidase-deleted strains was due to their inability to transform cellobiose, we investigated the effect of absence of the major β-glucosidases on the efficacy of cbh1 induction by cellobiose (Fig. 6C). While cellobiose plus δ-gluconolactone was usually used to provoke cellulase induction in H. jecorina, cellobiose alone was found to be capable of inducing cellulase formation in both WT and mutant strains, though the level of secreted proteins was lower than those induced by Avicel in the WT (data not shown). However, in comparison with the WT, induced production of cellulases as represented by CBH1 occurred more efficiently in the presence of 0.25% cellobiose in the cel1a and cel1b deletion strain (Fig. 6C). Similar to results for the Δcel1a Δcel1b strain, the simultaneous absence of BglI was also capable of provoking efficient formation of cellulases at relatively lower concentrations of cellobiose than was the case with the WT. Therefore, in accordance with previous results, lowering the degree of cellobiose hydrolysis may facilitate the stimulation of cellulase synthesis. These results also suggest that further metabolism of cellobiose beyond hydrolysis by major extracellular and intracellular β-glucosidases may not be involved in the efficient induction of cellulase genes.

Addition of cellobiose restores the efficient induction on Avicel in β-glucosidase-disrupted strains.

The above results imply that the productive induction defect as observed in the absence of the major β-glucosidases may to a larger extent be caused by the insufficient cellobiose initially available for triggering the induction cascade. We therefore asked whether the addition of cellobiose would rescue the induction defect on Avicel as displayed in Δcel1a Δcel1b and ΔtriβG strains. After induction with 1% Avicel plus 0.25% cellobiose, there was a slight decrease in the amount of secreted protein and Avicel hydrolysis activities in the wild-type strain (Fig. 7A and C). In contrast, the Δcel1a Δcel1b and ΔtriβG cultures produced an amount of protein, including CBH1, comparable to that produced by the WT during the early induction on Avicel (Fig. 7A and B). In addition, the induced Δcel1a Δcel1b and ΔtriβG cultures showed a significant increase in hydrolytic activity toward Avicel over this same period of induction (Fig. 7C). These data indicate that β-glucosidases may influence the productive transcription of cellulase genes on Avicel by ensuring an appropriate level of cellobiose available for triggering the induction cascade.

Fig 7.

Cellobiose rescued the induction defect by Avicel in β-glucosidase deletion strains. (A) Western blot analysis of CBHI secreted in culture filtrates of the WT, Δcel1a Δcel1b, and ΔtriβG strains after induction with 1% Avicel plus cellobiose (0.25% [wt/vol]). (B and C) Extracellular protein production (B) or cellulase activity of culture supernatant (C) of the WT, Δcel1a Δcel1b, and ΔtriβG strains from panel A after 24 h of induction toward Avicel. Data shown are the means for three independent experiments.

DISCUSSION

Successful perception of the existence of insoluble cellulose outside the cell is critical for cellulolytic H. jecorina to initiate the rapid production of the enzymatic machinery needed to sustain its growth on the breakdown products. In this study, we demonstrate that the absence of the intracellular β-glucosidases CEL1a and CEL1b significantly delays cbh1 gene expression on crystalline cellulose. Also, we show that simultaneous absence of the major extracellular β-glucosidase BglI builds on this defect by displaying a further delay in initiating the induction process. However, we further demonstrate that deletion of these major β-glucosidases has no effect on the productive induction of cellulase genes by cellobiose and that addition of cellobiose rescues the induction defect of the mutant strains on Avicel.

It is reasonably believed that one mode of perception of the presence of cellulose outside the cell would be sensing its degradation products (14, 16, 22). Several lines of evidence that cellobiose or its close relatives are capable of efficiently inducting the cellulases in H. jecorina have indeed been obtained (7, 30). In all cases, β-glucosidase capable of both hydrolyzing and transglycosylating cellobiose has been thought to play an important role within the cascade regulating cellulase formation. A balance regarding the β-glucosidase-mediated metabolism of cellobiose which significantly influences cellobiose's role in provoking cellulase formation has thus been suggested to exist (7, 20). In this work, our initial observations that the absence of three major β-glucosidases compromised the efficient induction of cellulases and that the productive transcription was largely restored by sophorose indicate a potential involvement of these intracellular β-glucosidases in the rapid induction of cellulases by forming a cellulase inducer. However, our further result, that the efficacy of induction on cellobiose was not compromised in these mutants, argues against the previous surmise that the induction defect on Avicel results from the inability to process cellobiose into a cellulase inducer. To the contrary, the absence of these major β-glucosidases rather seemed to facilitate the induction on cellobiose by displaying a more productive response. The response of β-glucosidase-deleted H. jecorina strains to cellobiose described here is consistent with that recently reported for Neurospora crassa (36) and could be ascribed to sparing more cellobiose for induction while in the meantime unmasking the inducing effect from the glucose-mediated catabolite repression. Together with the findings that transglycosylation activities of β-glucosidases are only observed in vitro with extremely high concentrations of sugars (27, 33), the above results imply that participation of transglycosylation activities from these β-glucosidases in forming cellulase inducers, if any, is not relevant and therefore may not constitute a major part of the mechanism for their involvement in efficient induction of cellulases. Note that we cannot at present draw the conclusion that cellobiose itself is the physiological inducer, since the possibility exists that proteins other than the β-glucosidases under study may still participate in modifying cellobiose to another true inducer. The physiological role of transglycosylation activity associated with β-glucosidase in cellulase induction still remains to be established.

Given the comparable induction of a filter paper hydrolyzing activity to that induced by cellulose when cellobiose was fed at a continuous low level (35), we predict that the absence of the bulk of β-glucosidase in H. jecorina would also make cellobiose alone an easily manipulated potent inducer for enzyme production and regulation, as reported recently for N. crassa (36). In this respect, a possible interpretation for the defective induction in the β-glucosidase-deleted strains on Avicel would be that the initial release of cellobiose from Avicel may be less efficient, probably due to a transient feedback inhibition of exoglucanases by cellobiose, whose hydrolysis is slowed in the absence of β-glucosidases. The insufficiently released cellobiose thus failed to efficiently initiate the induction cascade when the mutants were incubated with Avicel. The possibility that cellooligosaccharides released from cellulose may in the absence of β-glucosidases not be cleaved at a rate sufficient to build up the required concentration of cellobiose for induction also exists. Our results that increasing the intracellular β-glucosidase activity by expressing CEL1b (I174C) or addition of an appropriate amount of cellobiose improved the induction defect in the Δcel1a Δcel1b and ΔtriβG strains do point to such a possibility. Therefore, we hypothesize that a subtle balance between cellobiose production and metabolism exists during Avicel hydrolysis, which is tightly controlled by extracellular and intracellular β-glucosidases in influencing their release from cellulose for signal induction (Fig. 8). Our results thus indicate that the requirement for the three major β-glucosidases in the rapid induction of cellulase genes on Avicel may lie in their ability to ensure an appropriate amount of cellobiose from Avicel for efficiently initiating the signal cascade in H. jecorina.

Fig 8.

Model of β-glucosidases' role in cellulase induction with cellulose versus cellobiose in H. jecorina. Upon induction with cellulose, cellobiose is released from cellulose by the synergistic action of exoglucanases and endoglucanases. Released cellobiose is then either transported intracellularly to initiate the transcriptional induction or hydrolyzed by extracellular and intracellular β-glucosidases (solid line). The absence of β-glucosidases, however, compromises the initial release of cellobiose from Avicel, and thus the efficient initiation of the following signaling cascade though hydrolysis of cellobiose may also be slowed down (dashed line). Upon incubation with cellobiose, the downstream pathway from cellobiose for transcriptional induction is strengthened while that for repression is weakened in the β-glucosidase deletion strains. The question mark denotes putative cellobiose derivatives.