INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

-Mannanase enzymes (endo-1,4--D-mannanase, mannan endo-1,4--mannosidase; EC 3.2.1.78) are endohydrolases that catalyze the random hydrolysis of -1,4-mannosidic linkages in the main chain of galactomannan, glucomannan, galactoglucomannan, and mannan (20). These enzymes are of particular interest in ligninolytic fungi, where there is a complex interplay between the processes of degradation of lignin, cellulose, and hemicellulose (including the mannans). There are possible biotechnological applications. -Mannanase can be used to bleach pulp, to reduce the viscosity of instant coffee, and for the clarification of fruit juices and wines (35). At the same time, the role of hemicellulases in wood decay and the breakdown of leaf litter remains poorly understood. The cultivated mushroom Agaricus bisporus, a ligninolytic leaf litter degrader (8), is the source of the mannanase described in this paper.

The deduced CEL4 amino acid sequence shows that the encoded protein has a modular structure (40). There is a signal peptide at the N terminus, typical of proteins that are secreted into the medium, a catalytic domain, and a linker rich in serine, proline, and threonine that separates a cellulose-binding domain from the catalytic domain. The catalytic domain of CEL4 had the most amino acid sequence similarity with ascomycete mannanases from Aspergillus aculeatus (5) and Trichoderma reesei (28), which belong to glycosyl hydrolase family 5 (43 and 42%, respectively). Therefore, based on amino acid similarities of the catalytic domains, CEL4 also belongs to family 5 (11). Both CEL4 and the mannanase from T. reesei contain a cellulose-binding domain but at opposite ends (N proximal and C terminal, respectively, of the protein). cel4 expression is regulated at the transcriptional level (40).

The preferred substrate for growth of A. bisporus is compost. Utilization of the compost requires the production of a number of extracellular enzymes that degrade lignocellulosic components. Studies of axenic compost cultures during growth and fruiting of A. bisporus found large changes in laccase, cellulase, and xylanase levels during fruit body development (34, 38).

Our objectives in this study were (i) to determine the genomic sequence of cel4, (ii) to express CEL4 in Saccharomyces cerevisiae and Pichia pastoris to determine its enzymatic activity, (iii) to determine the effect of the cellulose-binding domain on CEL4 activity, and (iv) to determine if CEL4 is regulated during fungal development. This work may lead to the utilization of CEL4 in the food or pulp industries and to its manipulation for crop improvement.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Strains and culture conditions. We isolated cel4 from a genomic library of the carboxin-resistant A. bisporus strain C54-carb8 (18, 26). cel4 cDNA from A. bisporus strain D649 (40) was used for the yeast heterologous expression experiments.

Recombinant-DNA techniques and enzymes. cel4 cDNA (pEYc1200) was previously isolated from a library in ZAP (40). Standard DNA manipulations were carried out essentially as described by Sambrook et al. (25). Restriction enzymes and other enzymes used for DNA manipulations were purchased from Promega (Madison, Wis.) or New England Biolabs (Beverly, Mass.).

Sequencing of cel4 genomic sequence. Screening of cel4 from A. bisporus strain C54-carb8 (26) was according to the method of Yagüe et al. (39). A ³²P-random primer-labeled HindIII-NotI 1.3-kb cel4 cDNA fragment from pEYc1200 consisting of only the catalytic domain was used as a probe. The cel4 genomic sequence was obtained by directly sequencing from the cosmid (ABI sequencer 377; University of Durham, Durham, United Kingdom). Cosmid clone 31C1 was the only clone that hybridized to the cel4 cDNA probe. The sequence from 553 bp upstream of the start codon to 308 bp downstream of the stop codon (2,853 bp) was sequenced (both strands) by using specific primers, with the following exception. The sequence from 974 to 2853 bp was subcloned into pBluescript KS(+) (Stratagene) as a 1.8-kb XbaI fragment, which enabled standard primers (T7 and T3) to be used for sequencing of the end regions of this fragment.

Prediction of O-glycosylation sites. Putative O-glycosylation sites in the predicted CEL4 amino acid sequence were found by using the program NetOGly 2.0 (10), which was obtained from the internet site http://www.cbs.dtu.dk/databases/OGLYCBASE/.

Production of CEL4 polyclonal antibody. We also expressed cel4 in an E. coli expression vector, pT79 (a derivative of pMW172 [33]), so that the polypeptide could be used as an immunogen. A 1.1-kb cel4 cDNA fragment specifying only the catalytic domain (252 to 1,342 bp) was synthesized by PCR. The cellulose-binding domain and linker region were not included since exclusion of this part of the protein, which is very similar to CEL1, CEL2, and CEL3, was expected to elicit an antiserum that was more specific to CEL4. The primers were the Sfi-cel4 primer, 5'-CGATCGGCCGACGTGGCCGTGTCGACCGGATTT-3', and the Not-cel4 primer, 5'-CGTTCGCGGCCGCCGTGATCAACAATA-3'. The restriction sites are underlined, and the nucleotides (in bold) are complementary to the cel4 catalytic domain sequence. The recombinant protein was produced in E. coli BL21(DE3) as inclusion bodies that allowed a one-step purification by step gradient centrifugation (2). Antibodies against CEL4 were raised in a New Zealand White rabbit by immunization with 600 µg (500 µl) of the purified inclusion body proteins, which were emulsified with an equal volume of Speecol adjuvant (Id-dlo; Institute for Animal Science and Health, Lelystad, The Netherlands). Boosters of 150 µg of protein were given approximately every 2 weeks. The maximum titer (by enzyme-linked immunosorbent assay) was found at 13 weeks. A high dilution of the antiserum was required in the enzyme-linked immunosorbent assay, which indicates that the antiserum has a high affinity for the antigen (CEL4 secreted in P. pastoris). Serum collected prior to immunization was used as a control.

Construction of the CEL4 yeast recombinants. Splicing by overlap extension PCR (13) was used to splice an S. cerevisiae triose phosphate isomerase promoter and SUC2 invertase secretion peptide to cel4. The yeast promoter and secretion peptide were subcloned into pBluescript (4). These clones include the full-length sequence of cel4, except that the sequence up to and including most of the secretion peptide was removed. The flanking primers used were T3 (Stratagene) and CM5 5'-CGTGATCATCTAGACTAATTCAAGCCCGG-3', which incorporates an XbaI restriction site (underlined) and is cel4 specific (in bold). The overlap primers used were CM1, 5'-GCAGCCAAAATAGCCGATGTTCCAGTCTGG-3', and CM2, 3'-CGTCGGTTTTATAGACGGCTACAAGGTCAGACC-5' (the cel4 sequence is highlighted in bold). The T3 and CM2 primers were used to amplify the yeast promoter and secretion peptide fragment (1 kb), and the CM1 and CM5 primers were used to amplify the cel4 sequence (1.3 kb, nucleotides 63 to 1341). The PCR conditions used were 95°C for 20 s, 52°C for 20 s, and 74°C for 90 s for 20 cycles; 10 ng of cDNA template was used in each reaction mixture. The splicing by overlap extension PCR conditions used were 95°C for 1 min, 58°C for 30 min, and 74°C for 30 s for 20 cycles; 100 ng of each of the two PCR products was used with the flanking primers. An XhoI site upstream of the yeast promoter in the multiple cloning site of pBluescript and the XbaI restriction site (incorporated into the primer) were used to directionally clone the spliced construct into the pYES expression vector. For the cloning of cel4 into the P. pastoris shuttle vector, the cel4 cDNA (nucleotides 63 to 1320) was amplified with the following primers: XhPcel4F (5'-GGCGCGCTCGAGAAGAGAGATGTTCCAGTCTGGGG-3') and XhPcel4R (5'-GGAGGGCGTTCTAGAGCCCGGTTTTTCATTGCAG-3') (cel4 sequence is in bold). Primer XhPcel4F introduces an XhoI site (underlined) upstream of the cel4 sequence and primer XhPcel4R introduces an XbaI site (underlined) downstream of cel4. P. pastoris was the host used to express truncated cel4 (i.e., without the cellulose-binding domain). The 1.1-kb fragment (nucleotides 171 to 1320) encoding the linker and catalytic domain was amplified using the following primers: CEL4FCBD (5'-GGCGCGCTCGAGAAGAGACCTGGATCAACAACT-3') (the XhoI restriction site is underlined, and the cel4 sequence is in bold) and the XhPcel4R primer.

Activity measurements and Western analysis. Qualitative activity determinations by Congo Red staining were according to the method of Pentilä et al. (22), and the hydrolysis halos were quantified with an image analyzer (EASY Plus enhanced analysis system; Herolab, Wiesloch, Germany) running EASY Plus version 4.16 software. Culture filtrates were dialyzed against 50 mM sodium acetate, pH 5, overnight prior to enzymatic assays. The Nelson-Somogyi assay was used to make the quantitative measurements (as increases in reducing groups) essentially according to the methods of Nelson (21) and Somogyi (27). CEL4 activity was assayed with the following substrates: 1% (wt/vol) locust bean gum (LBG) (Sigma, St. Louis, Mo.), 1% (wt/vol) xylan (Sigma), 1% (wt/vol) carboxymethyl cellulose (CMC) (BDH, Toronto, Canada), and 0.1% (wt/vol) barley -1,3-glucan (Sigma). Protein concentrations were determined using the modified Lowry assay (12) and the Bradford assay according to the instructions of the manufacturer (Bio-Rad, Hercules, Calif.). Protein samples were analyzed by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis using 10% (wt/vol) polyacrylamide gels containing 0.1% (wt/vol) SDS according to the method of Laemmli (16). Either the gels were stained directly using Coomassie brilliant blue R250 or separated proteins were electroblotted onto nitrocellulose (Hybond-C; Amersham, Little Chalfont, Buckinghamshire, United Kingdom) as described by Zamanian and Mason (42). Standard protein molecular mass markers were obtained from Novex (San Diego, Calif.). Western blots were probed with anti-CEL4 antiserum and horseradish peroxidase-labeled anti-rabbit F(ab')₂ goat secondary antibody (Sigma). CEL4 was visualized with a horseradish peroxidase conjugate substrate detection kit as described by the supplier (Bio-Rad). The CEL4 antiserum was used to detect CEL4 protein (i) secreted by S. cerevisiae and P. pastoris recombinants, (ii) in the compost extracts, and (iii) in the mycelium and culture filtrates of A. bisporus grown in minimal medium containing fructose or Solka floc. The intensities of the bands from the Western blots from the compost extracts and minimal medium experiments were quantified by using an image analyzer.

Nucleotide sequence accession number. The sequence reported in this paper has been deposited in the EMBLNEW database under accession number AJ271862.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Isolation and sequencing of the cel4 gene. We isolated one recombinant clone from the cosmid genomic DNA library from A. bisporus after screening with a cel4 cDNA low-stringency probe. This clone, when digested with HindIII, gave a single positive band of 1.8 kb, suggesting that A. bisporus contains a single copy of cel4 and no other mannanases with similar sequences. We designed primers to sequence out from the flanking regions of the cel4 cDNA. Primers were then designed based on those sequences but reading inwards. For the sequence reported, both strands were analyzed. We found 12 short (46- to 65-bp) introns (Fig. 1). These introns had the consensus GTNN(G/T/A)-5' and (A/C/T)AG-3' sequences that are commonly found in fungi (32). The fifth and sixth introns delimited an exon of only 12 bp. A putative TATA box sequence, TATAAAA, was found 52 bp upstream from the start codon. The TATA box region is not GC rich. There are five putative CAAT boxes in both orientations in the 550 bp upstream of the start codon of cel4 286 and in the 382 bp (sense strand) and 206, 291, and 532 bp (antisense strand) upstream. Putative CAAT boxes are present in all A. bisporus genes sequenced to date. No sequence resembling the consensus eukaryotic polyadenylation sites was found (14), but the transcribed 3' noncoding region may extend beyond the 308 bp sequenced from this region.

View larger version (6K): [in this window] [in a new window]   FIG. 1.   Genomic organization of cel4. The predicted signal peptide (SS) is at Met-1 to Ala-19. The putative cellulose-binding domain (CBD) is at Asp-20 to Leu-55, and the putative catalytic domain is at Gly-86 to Ala-439 (thick lines). The domains are separated by a proline/serine/threonine-rich linker (L). M represents the start codon. The presence of 12 introns (triangles) was deduced by comparison of cDNA and genomic sequences.

Heterologous expression of cel4 in S. cerevisiae and P. pastoris. S. cerevisiae containing the cel4 gene produced clear halos on LBG plates, indicating that -mannanase was secreted and was enzymatically active. The parent vector (control) did not produce a clear halo, and no halos were seen on plates containing 1% (wt/vol) xylan or 1% (wt/vol) CMC in place of LBG. Therefore, no xylanase or endoglucanase activities were detected. With P. pastoris, the host-vector controls secreted significant endoglucanase activity, such that any minor activity of the CEL4 protein would have been masked. The P. pastoris strains transformed with cel4 produced clear halos on the LBG plate, however, and the parent vector (control) did not. We made reducing sugar assays of the mannanase activity of CEL4 in the P. pastoris and S. cerevisiae culture filtrates and pellets (Table 1). CEL4 from which the cellulose-binding domain had been removed was expressed in P. pastoris. Its activity was less than for the unmodified protein (Table 1).

Regulation of CEL4 in fruiting cultures of A. bisporus compost extracts. Western blots of the various compost extract stages were probed with CEL4 antiserum, with laccase antiserum used as a control (equal volumes of extracts containing approximately 0.3 µg of protein per µl were used). Mannanase activities from Nelson-Somogyi assays and Congo Red assays of the compost extracts from different developmental stages were also measured (Table 2). The activities from the Congo Red plate assays were determined from the areas of the hydrolysis halos using an image analyzer (Table 2).

Regulation of CEL4 secretion in A. bisporus laboratory cultures. We compared cellulase and mannanase activities from the laboratory liquid culture filtrates of A. bisporus grown in the presence of Solka floc (Fig. 2). These data were compiled from the averages of four (cellulase) and five (mannanase) measurements made on separate occasions. Time zero was the point at which mycelium was transferred from fructose-containing medium. The levels of cellulase and mannanase activity rose to a peak and subsequently declined. In the case of cellulase, the activity rose steadily to a maximum at 42 days and then dropped steadily, which is consistent with previously published data (19). Mannanase activity peaked at 21 days and then dropped more gradually. The highest level of cellulase activity is about five times more than the highest level of mannanase activity.

View larger version (35K): [in this window] [in a new window]   FIG. 2.   Levels of CEL4 protein and mannanase activity in liquid cultures of A. bisporus (D649 strain) containing 0.05% (wt/vol) Solka floc grown at 25°C. Nelson-Somogyi assays were used to measure cellulase (, CMC as the substrate) and mannanase (, LBG as the substrate) activities. The levels of CEL4 in Solka floc-bound mycelium () and in culture filtrates () were detected by CEL4 antiserum in Western blots. Error bars indicate standard errors of activities measured on five and four separate occasions for mannanase and cellulase, respectively. Western blots were repeated four and seven times for mycelium and culture filtrates, respectively. The amount of CEL4 protein from culture fluid was approximately 10-fold more than that found in the mycelium- and cellulose-bound fraction in the 3-week-old sample.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The cel4 genomic DNA and the cel4 cDNA (accession number Z50095) differ at 16 of 1,432 nucleotides. These differences change seven amino acid residues, of which two were not changes in amino acids with functional differences. These differences might be because different alleles were cloned but may also represent differences at more than one locus. cel4 has 12 introns ranging from 46 to 65 bp. The other hemicellulase gene sequenced from A. bisporus, xylanase (accession number X78330 [7]), also has a large number of introns (11), while the cellulase genes do not (4, 24). The Congo Red plate assays and reducing sugar assays showed that CEL4 protein is a specific mannanase that, when expressed in S. cerevisiae, has no detectable activity against CMC, xylan, or barley -glucan. The specific activity of CEL4 secreted by P. pastoris was approximately three times that secreted by S. cerevisiae. About 2.5 times more of the mannanase activity was found in the S. cerevisiae recombinant cell extracts than in the culture fluid. In P. pastoris, about six times more activity was found in the culture filtrate than in the cell extracts. Therefore, P. pastoris is much better at secreting CEL4 into the culture medium, presumably because P. pastoris has a more powerful inducible promoter and secretion peptide (-factor pre-pro secretion signal from S. cerevisiae) than does S. cerevisiae (6).

The molecular mass of mature CEL4 deduced from the nucleotide sequence is 49 kDa; however, the apparent molecular mass of CEL4 of both the S. cerevisiae and P. pastoris cultures from the Western blot was approximately 60 kDa, probably due to glycosylation (results not shown). Hyperglycosylation has been observed with CEL3 of A. bisporus and two endoglucanases of T. reesei when they were expressed in S. cerevisiae (4, 22).

There is no evidence that CEL4 is a low-level- or mixed-specificity enzyme, and we do not know why some hemicellulases, e.g., CEL4, contain specific cellulose-binding domains and others, e.g., xylanase from A. bisporus (7), do not. Removal of the cellulose-binding domain of CEL4 reduced the specific activity significantly relative to that of full-length CEL4. Removal of the cellulose-binding domain has little influence on the activities of cellulases towards soluble substrates but decreases their activities towards insoluble cellulose (17).

A. bisporus is cultivated on composted wheat straw and can degrade lignin, cellulose, hemicellulose, protein, and microbial cell wall polymers (1, 3, 4, 7, 8, 9, 19, 23, 34, 41). Laccase accumulates in compost during its colonization by mycelium of A. bisporus, and the level of laccase rapidly decreases at the time of fruiting (30). Cellulase (31) increases during fruiting. The large increase in mannanase activity in both postharvest and senescent cultures contrasts with the changes observed in other extracellular enzymes and has no obvious explanation. Mannanase may be required to replenish stored carbohydrate levels in the mycelium. Mannanase activity is detectable at all stages, however, so mannan hydrolysis may be a source of carbon and energy even during colonization, although it is unlikely that this activity is due to CEL4 (Fig. 2). During fruit body development, CEL4 follows a profile similar to that of cellulase activity (38) and endoxylanase activity (7, 34).

In laboratory liquid cultures, the amount of CEL4 detected in the culture filtrate does not increase over time after an initial rapid increase (Fig. 2). We expected that CEL4 would remain bound to the mycelium until the cellulose was used up. The mycelium- and cellulose-bound fraction declines in later samples, but CEL4 in the medium does not increase correspondingly. Although the CEL4 activity drops after 21 days, the amount of protein remains fairly constant, consistent with CEL4 inactivation and similar to the behavior of both laccase (36) and cellulase (19). Around 80% of CEL4 remains in the culture filtrate throughout the time course.

Our results show large changes in the level of CEL4 activity during fruiting-body development, suggesting that hemicellulose degradation contributes to the large carbon and energy demand during sporophore formation and that hemicellulose degradation parallels the mobilization of cellulose. The cellulose-binding domain in CEL4 might function in some manner in the coordination of these functions. The expression of CEL4 in P. pastoris provides a baseline for future studies of compost degradation and the extracellular enzymes that fungi deploy in this process.