Catalytic Properties and Classification of Cellobiose Dehydrogenases ...

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Mar. 2011, p. 1804–1815 0099-2240/11/$12.00 doi:10.1128/AEM.02052-10 Copyright © 2011, American Society for Microbiology. All Rights Reserved.

Vol. 77, No. 5

Catalytic Properties and Classification of Cellobiose Dehydrogenases from Ascomycetes䌤† Wolfgang Harreither,1‡ Christoph Sygmund,1‡ Manfred Augustin,1 Melanie Narciso,1 Mikhail L. Rabinovich,2 Lo Gorton,3 Dietmar Haltrich,1 and Roland Ludwig1* Department of Food Sciences and Technology, Food Biotechnology Laboratory, BOKU—University of Natural Resources and Life Sciences, Muthgasse 18, A-1190 Vienna, Austria1; A. N. Bach Institute of Biochemistry, Russian Academy of Sciences, 33 Leninsky Prospect, 119071 Moscow, Russia2; and Department of Analytical Chemistry/Biochemistry, Lund University, P.O. Box 124, SE-22100 Lund, Sweden3 Received 31 August 2010/Accepted 23 December 2010

Putative cellobiose dehydrogenase (CDH) genes are frequently discovered in various fungi by genome sequencing projects. The expression of CDH, an extracellular flavocytochrome, is well studied in white rot basidiomycetes and is attributed to extracellular lignocellulose degradation. CDH has also been reported for plant-pathogenic or saprotrophic ascomycetes, but the molecular and catalytic properties of these enzymes are currently less investigated. This study links various ascomycetous cdh genes with the molecular and catalytic characteristics of the mature proteins and suggests a differentiation of ascomycete class II CDHs into two subclasses, namely, class IIA and class IIB, in addition to the recently introduced class III of hypothetical ascomycete CDHs. This new classification is based on sequence and biochemical data obtained from sequenced fungal genomes and a screening of 40 ascomycetes. Thirteen strains showed CDH activity when they were grown on cellulose-based media, and Chaetomium atrobrunneum, Corynascus thermophilus, Dichomera saubinetii, Hypoxylon haematostroma, Neurospora crassa, and Stachybotrys bisbyi were selected for detailed studies. In these strains, one or two cdh-encoding genes were found that stem either from class IIA and contain a C-terminal carbohydrate-binding module or from class IIB without such a module. In several strains, both genes were found. Regarding substrate specificity, class IIB CDHs show a less pronounced substrate specificity for cellobiose than class IIA enzymes. A pH-dependent pattern of the intramolecular electron transfer was also observed, and the CDHs were classified into three groups featuring acidic, intermediate, or alkaline pH optima. The pH optimum, however, does not correlate with the CDH subclasses and is most likely a species-dependent adaptation to different habitats. ascomycetes or microfungi are the most important lignocellulose degraders (37). The exact biological role of CDH has not been fully elucidated yet, but our current knowledge points to its participation in the degradation and modification of polymers such as cellulose, hemicellulose, and lignin by generating hydroxyl radicals via the Fenton reaction (16, 19). CDH thus might also be an important component of the extracellular lignocellulose-degrading enzyme system of ascomycetes. Although CDH was already reported in a soft rot Monilia species as early as 1980 (6), little attention has been paid to the formation and biochemical properties of ascomycetous CDHs. CDH activity was reported in cultures of the ascomycetes Thielavia heterothallica (synonyms, Myceliophthora thermophila and Sporotrichum thermophile) (3, 4), Chaetomium sp. INBI 2-26(⫺) (38), Chaetomium cellulolyticum (10), Neurospora crassa (11), Humicola insolens (32), and Myriococcum thermophilum (14). While some of these enzymes were characterized biochemically to various extents, only the cdh genes from Sporotrichum thermophile (34), H. insolens (40), and Myriococcum thermophilum (44) were cloned. Phylogenetic analysis of all known cdh genes before 2006 showed a separation into two classes: class I, representing only basidiomycetous CDHs, and class II, exclusively comprising the longer and more complex ascomycetous CDHs, which sometimes contain a C-terminal type 1 carbohydrate-binding module (CBM) (42, 43). As more cdh sequences from genome projects became available, a third class of hypothetical CDHs found in the ascomycetes was in-

Cellobiose dehydrogenase (CDH; EC 1.1.99.18; cellobiose: [acceptor] 1-oxidoreductase) is an extracellular flavocytochrome secreted by white and brown rot and plant-pathogenic as well as composting fungi from the dikaryotic phyla Basidiomycota and Ascomycota under cellulolytic culture conditions (43). A number of important lignocellulose-degrading species are described from both phyla, but knowledge of the lignocellulose-degrading system of ascomycetes is less advanced. Although ascomycetes can attack all cell wall constituents, cellulose and hemicelluloses are more preferred targets than lignin (39) and the efficiency of their lignocellulose-degrading enzymatic systems is, in general, not comparable to that of basidiomycetes (20, 33). While ascomycetes are less capable of rapid wood decay than basidiomycetes, they can degrade wood and other lignocellulosic material under extreme environmental conditions which are unsuited for most basidiomycetes, e.g., under conditions of high humidity or under conditions of elevated pH and temperature, such as those occurring during the composting process. Under these conditions (thermophilic), * Corresponding author. Mailing address: Department fu ¨r Lebensmittelwissenschaften und -technologie, Universita¨t fu ¨r Bodenkultur, Muthgasse 18/2, A-1190 Vienna, Austria. Phone: 431 47654 6149. Fax: 431 47654 6199. E-mail: [email protected]. ‡ W.H. and C.S. contributed equally. † Supplemental material for this article may be found at http://aem .asm.org/. 䌤 Published ahead of print on 7 January 2011. 1804

VOL. 77, 2011

CELLOBIOSE DEHYDROGENASES FROM ASCOMYCETES

troduced in 2008 (44). To our best knowledge, no CDH from class III has been characterized or actively expressed so far. In general, CDHs are monomeric proteins with both FAD and heme b as prosthetic groups located in two distinct domains, which are connected by a mobile linker (13). Typical class I basidiomycete CDHs show a strong preference for cellobiose and cello-oligosaccharides, which are oxidized at the anomeric carbon atom, while glucose and other monosaccharides are very poor substrates. On the basis of the few data available, class II ascomycete CDHs also prefer cellobiose as a substrate, but glucose turnover is less discriminated than it is with their basidiomycetous counterparts and lower Km values toward glucose, lactose, and maltose are found (43). In the ensuing oxidative half-reaction of the catalytic cycle of CDH, the two electrons obtained from the sugar substrate and stored in FAD are transferred to two-electron acceptors, e.g., quinones, or one-electron acceptors, such as complexed metal ions, e.g., Fe(III), Mn(II), or the in vitro substrate cytochrome c (cyt c). One-electron acceptors can be reduced at CDH’s cytochrome domain after an intramolecular electron transfer (IET) from the FAD to the heme b. CDH generates ferrous iron species via its cytochrome domain by reduction of functional iron complexes such as ferric oxalate. The subsequent reaction between these reduced iron species and H2O2, which can be produced by CDH itself (it also shows a weak oxidase activity) or by other fungal extracellular flavin-dependent oxidases, generates reactive hydroxyl radicals (19, 25, 26). The pH optima for electron acceptors of class II CDHs (14, 17, 32) seem to be less acidic than those reported for basidiomycetous CDHs (1), but because of the currently limited availability of data on class II CDHs, it cannot be concluded whether this is a general feature of class II CDHs or is found only in some species. Likewise, the pH optima of IET measured with cyt c vary between different enzymes, e.g., pH 6.0 for Chaetomium sp. INBI 2-26(⫺) CDH and pH 7.5 for H. insolens CDH (17, 32). As mentioned above, only limited information on ascomycetous CDHs is available at present, and the information is too scarce for investigators to be able to draw more general and definite conclusions, e.g., (i) whether all ascomycetous class II CDHs form one distinct group that is clearly different from basidiomycetous CDHs (besides the not yet expressed class III CDHs, which are therefore not available for biochemical characterization), (ii) whether class II CDHs can be further classified into subclasses (whether they show a broader diversity in their molecular and catalytic properties than class I CDHs), and (iii) whether certain catalytic properties of class II CDHs are typical for these subclasses or are more related to the natural habitat of each fungal species. This knowledge can be of significant importance, especially for the application of CDH in biosensors and biofuel cells, in which significant interest has recently been raised (23). Hence, the aim of this article is to provide detailed knowledge on the molecular and catalytic properties of ascomycetous CDHs, to indicate their role in lignocellulose degradation by ascomycetes, and to investigate possible applications of ascomycetous CDHs in glucose biosensors and implantable, glucose-powered biofuel cells.

1805

MATERIALS AND METHODS Organisms and culture conditions. The complete list of all 40 fungal strains that were screened for CDH activity is given in Table S1 in the supplemental material); these strains were obtained from the Centraalbureau voor Schimmelcultures (CBS; Utrecht, Netherlands), the Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ; Braunschweig, Germany), and the Institute of Molecular Biotechnology (IMBT; Graz, Austria). The cultures were periodically subcultured on potato dextrose agar (PDA) plates. Freshly inoculated agar plates were incubated at 30°C until colonies reached a diameter of ⬃5 cm, and the colonies were then used to inoculate shaking flasks. The medium used for submerged cultures contained 20 g liter⫺1 of ␣-cellulose, 5 g liter⫺1 of meat peptone, 0.3 ml liter⫺1 of a trace element solution (31), and H3PO4 to titrate the pH to 5.0 prior to sterilization (after autoclaving of the medium, the pH was 5.4 to 5.6). To study the impacts of cellulose and the nitrogen concentration on selected CDH-producing fungi, the concentrations of the carbon and nitrogen sources were varied in the medium, which always contained the trace element solution at its standard concentration. ClowNlow medium contained 5 g liter⫺1 ␣-cellulose and 5 g liter⫺1 peptone, ChighNlow medium contained 20 g liter⫺1 ␣-cellulose and 5 g liter⫺1 peptone, ClowNhigh medium contained 5 g liter⫺1 ␣-cellulose and 20 g liter⫺1 peptone, and ChighNhigh medium contained 20 g liter⫺1 ␣-cellulose and 20 g liter⫺1 peptone (see Table S2 in the supplemental material). For the cultivation in shaken flasks 1-liter Erlenmeyer flasks were filled with 0.3 liter medium, sealed with cotton plugs, autoclaved (121°C, 20 min), inoculated with ⬃3 cm2 of finely cut, actively growing mycelium from PDA plates, and incubated in a rotary shaker (110 rpm; eccentricity, 1.25 cm) at 25°C or 40°C; the latter cultivation temperature was chosen for thermophilic fungi. Samples were taken daily, and the extracellular protein concentration, pH value, and CDH activity in the supernatant were measured. Activity assays and protein measurement. CDH activity was specifically determined by following the reduction of 20 ␮M cyt c (ε550 ⫽ 19.6 mM⫺1 cm⫺1) in McIlvaine (citrate-phosphate) buffer (27) containing 30 mM lactose at the indicated pH values and at 30°C (this temperature was chosen for the comparison of different enzymes under the same conditions and does not represent the temperature optimum of the enzymes isolated from either mesophilic or thermophilic ascomycetes). Alternatively, the 2,6-dichloro-indophenol (DCIP) assay, measuring the activities of both the intact holoenzyme and the flavodehydrogenase domain, was performed by measuring the time-dependent reduction of 300 ␮M DCIP at 520 nm in McIlvaine buffer containing 30 mM lactose at the given pH values and 30°C. When crude culture samples were used, sodium fluoride (final concentration in the assay, 4 mM) was added to inhibit potentially present laccase activity (2). The absorption coefficient for DCIP is pH dependent; however, its absorbance at 520 nm differs by only about 3% within the pH range of 3.0 to 8.0. The absorption coefficient was determined experimentally to be 6.8 mM⫺1 cm⫺1 (17). One unit of CDH activity was defined as the amount of enzyme that oxidizes 1 ␮mol of lactose per min under assay conditions. No activity was found in assays lacking lactose; thus, it can be excluded that the cello-oligosaccharides formed during the cultivation were used as an alternative substrate or influenced the measurement. The protein concentration was determined by the method of Bradford (2a) using a prefabricated assay, the protocol from Bio-Rad Laboratories (Hercules, CA), and bovine serum albumin as the standard. Protein purification. Hyphae and cellulose particles were removed from the culture broth by centrifugation (6,000 ⫻ g for 10 min). The clear supernatant was concentrated and diafiltered with water using a polyethersulfone hollow-fiber cross-flow module with a 10-kDa cutoff (Microza UF module SLP-1053; Pall Corporation) until a conductivity of 2 to 3 mS cm⫺1 was obtained. The concentrated enzyme solution was loaded onto a DEAE-Sepharose fast-flow column (XK50/70; bed volume, 1 liter; all chromatography equipment was from GE Healthcare Biosciences) that had been equilibrated with 50 mM sodium acetate buffer, pH 5.5. Proteins were eluted by increasing the amount of elution buffer (50 mM sodium acetate buffer, pH 5.5, containing 500 mM NaCl) linearly from 0 to 100% in 10 column volumes. Fractions containing CDH activity were pooled and ammonium sulfate was added to 20% saturation. The sample was loaded onto a Phenyl-Source fast-flow column (HR26/20) equilibrated with 100 mM sodium acetate buffer, pH 5.5, containing 0.2 M NaCl and ammonium sulfate (20% saturation). Proteins were eluted by a linear gradient of elution buffer (20 mM sodium acetate buffer, pH 5.5) in 10 column volumes. Fractions with CDH activity were pooled again. After diafiltration with a polyethersulfone flat-stack cross-flow module having a 10-kDa cutoff (Vivaflow 50; Sartorius, Go ¨ttingen, Germany) to a conductivity of 5 mS cm⫺1 in 20 mM sodium acetate buffer, pH 5.5, the CDH preparation was aliquoted and frozen at ⫺70°C for further use. Electrophoretic analysis. SDS-PAGE was carried out on a Hoefer Mighty Small SE 250 vertical electrophoresis unit (electrophoretic equipment and con-

1806

HARREITHER ET AL.

APPL. ENVIRON. MICROBIOL. TABLE 1. Purification of ascomycete cellobiose dehydrogenases

Source

C. atrobrunneum C. thermophilus D. saubinetii H. haematostroma N. crassa S. bisbyi

Expressed protein

CDH CDH CDH CDH CDH CDH

Sp acta at pH 4.0/sp act at pH opt. (pH)

RZ value

Purification (fold)

% yield

7.1/11.5 (6.0) 8.9/17.8 (5.0) 12.3/13.2 (5.5) 7.4/9.0 (5.0) 5.7/10.3 (5.5) 2.2/2.7 (4.5)

0.39 0.60 0.52 0.37 0.20 0.12

293 7.3 26.3 21.2 48.1 14.8

130 58 34 41 57 34

IIA IIB IIA IIB IIA IIA2

a The specific activity (U mg⫺1) is calculated from the volumetric activity (measured by the DCIP assay at pH 4.0), before the detailed characterization of the enzymes and afterwards at their (indicated) pH optima (pH opt.; values in parentheses), and the protein concentration was measured by the Bradford method.

regression by fitting the observed data to the Michaelis-Menten equation (Sigma Plot 11; Systat Software, San Jose, CA). The catalytic rate (kcat) obtained for homogeneous Corynascus thermophilus CDH (CtCDH) was calculated from the measured protein concentration and the molecular mass determined from the protein sequence. In order to compare kcat values and catalytic efficiencies (kcat/Km), despite the various purities of the CDH preparations, a purity factor was introduced by dividing an RZ value (calculated from the absorbance ratio A420/A280) of 0.6 for a hypothetically homogeneous ascomycete CDH preparation by the experimentally measured RZ values of each preparation (Table 1). These factors were 1.54 for Chaetomium atrobrunneum CDH (CaCDH), 1.0 for CtCDH, 1.15 for Dichomera saubinetii CDH (DsCDH), 1.62 for Hypoxylon haematostroma CDH (HhCDH), 3.0 for Neurospora crassa CDH (NcCDH), and 5.0 for Stachybotrys bisbyi CDH (SbCDH) and were used to multiply the kcat values based on the protein concentration determined with the Bradford assay and molecular masses of the proteins calculated from their deduced amino acid sequences (Table 2). Isolation of CDH-encoding genomic DNA and cDNA. The known amino acid sequences of ascomycetous CDHs with and without a carbohydrate-binding module were used to generate conserved sequence blocks using the program Block Maker (http://blocks.fhcrc.org/blocks/make_blocks.html). These blocks then provided the basis for the design of degenerated primers using the program CODEHOP (http://blocks.fhcrc.org/blocks/codehop.html). These primers (see Table S3 in the supplemental material) were used for the amplification of genomic DNA fragments of various lengths from six of the CDH-expressing ascomycetes found in this study (Chaetomium atrobrunneum, Corynascus thermophilus, Dichomera saubinetii, Hypoxylon haematostroma, Neurospora crassa, and Stachybotrys bisbyi). Samples for nucleic acid isolation were taken from freshly grown shaking-flask cultures on cellulose-based medium to induce CDH formation. Mycelia were squeezed dry between filter papers and then frozen in liquid nitrogen. Portions of approximately 100 mg of mycelium were used for DNA extraction (21). Total RNA was isolated using the TriFast reagent (PEQlab, Erlangen, Germany). cDNA first-strand synthesis was performed with a RevertAid first-strand cDNA synthesis kit (Fermentas, St. Leon-Rot, Germany) and the anchor primer (5⬘-GGCCACGCGTCGACTAGTACTTTTTTT

sumables were from GE Healthcare Biosciences). Gels (10.5 by 10 cm; 10% total acrylamide-bisacrylamide monomer concentration [T] and 2.7% cross-linker concentration [C]) were cast and run according to the manufacturer’s modifications of the Laemmli procedure. Proteins were visualized by Coomassie brilliant blue staining. Isoelectric focusing (IEF) in the range of pH 2.5 to 6.5 was performed on a Multiphor II system using precast, dry gels (Clean gel) rehydrated with carrier ampholytes. A broad-range pI marker protein kit (pH 3 to 10) was used to determine the isoelectric points. CDH bands were visualized by active staining. To that purpose, the gel was first soaked in 3 mM DCIP solution and then in 100 mM sodium citrate buffer, pH 5.0, before a 300 mM lactose solution was added over the surface. Colorless bands in a purple background appeared due to DCIP reduction by CDH and were marked. Subsequently, the gel was transferred into 20% acetic acid for fixation, and silver staining was performed. UV/visible spectra. The spectra of purified CDHs were recorded from 700 to 250 nm in both the oxidized and the reduced states using an Hitachi U-3000 spectrophotometer (Tokyo, Japan). CDH, which was in the oxidized state after purification, was diluted with 50 mM sodium citrate buffer, pH 4.0, to an absorbance at 280 nm of ⬃1, and the spectrum was then recorded. The spectrum of the reduced enzyme was obtained after a 1,000-fold molar excess of lactose was added to the cuvette. To assess the purity of the measured samples, the RZ (from the German “Reinheitszahl,” meaning purity index) values of the oxidized enzymes were calculated as the ratio of the absorbance at 420 nm to the absorbance at 280 nm. Kinetic measurements. Initial rates for the determination of pH/activity profiles and kinetic constants were measured at 30°C in McIlvaine buffer (27) at pHs ranging from 3.0 to 8.0. Stock solutions of carbohydrates used for kinetic measurements were prepared in the respective buffer and allowed to stand overnight for mutarotation, while stock solutions of electron acceptors were prepared in water and immediately used. Kinetic constants for electron acceptors were determined at their pH optima using 30 mM lactose as the electron donor. The reaction stoichiometry is 1 for the two-electron acceptor DCIP (1 mol of DCIP reduced per mole of lactose oxidized), but it is 2 for the one-electron acceptor cytochrome c. All kinetic constants were calculated using nonlinear least-squares

TABLE 2. Properties of nucleotide and protein sequences of various ascomycetous CDHs Organism (CDH gene, GenBank accession no.)

Full gene lengtha (bp)

Protein lengthb (aa)

CBM (aa)c

Massd (Da)

Masse (kDa)

pId

pIf

C. atrobrunneum (cdhIIA, HQ116815) C. thermophilus (cdhIIB, HQ116814) D. saubinetii (cdhIIA, HQ116816) H. haematostroma (cdhIIA, HQ116817) H. haematostroma (cdhIIB, HQ116818) N. crassa (cdhIIA, XM_951498) N. crassa (cdhIIB, XM_953141) S. bisbyi (cdhIIA1, HQ116819) S. bisbyi (cdhIIA2, HQ116820)

2,496 2,364 2,514 2,538 2,409 2,490 2,487 2,490 2,499

809 765 815 816 780 806 805 808 811

36 —g 38 36 — 36 — 36 36

85,942 81,840 86,721 87,414 83,901 83,587 86,630 86,108 86,688

90 85 115 NAh 85 90 NA NA 100

5.0 4.6 5.0 6.4 5.7 6.7 7.9 6.4 5.1

4.1 3.8 4.2 NA 4.3 4.4 NA NA 4.3

a

Full length of the gene, including the leader sequence. Number of amino acids (aa) of the mature, secreted protein, including the carbohydrate-binding module, if present. Number of amino acids (aa) of the carbohydrate-binding module, if present. d Theoretical value calculated from amino acid sequence. e Experimental data from SDS-PAGE. f Experimental data from IEF. g —, no carbohydrate-binding module present. h NA, not applicable. Protein was not expressed under the selected culture conditions. b c

VOL. 77, 2011


TTTTTTT-3⬘). PCRs were done either with GoTaq polymerase (Promega, Mannheim, Germany) or, for amplification of the full-length cDNA, with Phusion high-fidelity DNA polymerase (New England BioLabs, Frankfurt am Main, Germany), a deoxynucleoside triphosphate mixture (Fermentas), oligonucleotide primers (VBC Biotech, Vienna, Austria), and a C-1000 thermocycler (BioRad). For the amplification of the 5⬘-flanking regions, a DNA Walking SpeedUp premix kit (Seegene, Seoul, South Korea) was used. Target-specific reverse primers (see Table S4 in the supplemental material) were designed and used together with the DNA walking-annealing control primer and the universal primer (5⬘-GGCCACGCGTCGACTAGTAC-3⬘), both of which are part of the kit, according to the manufacturer’s guidelines. To obtain full-length cDNA clones, a nested PCR with two forward primers specific for a sequence upstream of the putative start codon and two reverse primers, one complementary to the 5⬘ region of the anchor promoter and a second reverse primer specific for a sequence shortly downstream of the stop codon, was done. PCR fragments were cloned into the CloneJET vector (Fermentas), according to the instructions in the manual, and sequenced by a commercial service provider (AGOWA, Berlin, Germany). Peptide sequencing. Purified CDH samples were subjected to SDS-PAGE, destained, carbamidomethylated, digested with sequencing-grade trypsin, and extracted from gel pieces as described previously (8). The extracts were dried in a SpeedVac apparatus and reconstituted with water containing 0.1% formic acid before subsequent liquid chromatography (LC)-electrospray ionization (ESI) tandem mass spectrometry analysis using a Dionex Ultimate 3000 Capillary LC system and a Waters (Milford, MA) Q-TOF Ultima equipped with its standard ESI source. The tryptic digests were separated on a BioBasic C18 column (300 Å; 5 ␮m; 150 by 0.32 mm; Thermo Scientific) by developing a linear gradient from 5% to 85% acetonitrile in 0.3% formic acid, pH 3.0, at a flow rate of 6 ␮l min⫺1 over 40 min. The capillary voltage of the mass spectrometer instrument was set at 3.2 kV, the cone voltage was set to 80 V, the source temperature was set to 100°C, and the desolvation temperature was set to 120°C. Positive ions were detected in the m/z range of from 50 to 2,000. De novo sequencing and pBLAST searches were performed manually using MassLynx (version 4.0) software (Waters). Phylogenetic analysis. All translated amino acid sequences of the newly cloned ascomycete CDH-encoding cDNA were submitted to GenBank. Other sequences used were from basidiomycetous CDHs, ascomycetous CDHs, and the CBM-carrying proteins given in Table S5 in the supplemental material. Multiplesequence alignments were done with the Clustal X program (see Table S4 in the supplemental material). Evolutionary relationships were inferred with the minimum-evolution method (30) by using MEGA (version 4) software (35). The initial tree was generated by the neighbor-joining method (with the complete deletion of gaps and missing data). The minimum-evolution tree was searched using close-neighbor interchange. The bootstrap consensus tree was inferred from 1,000 replicates, and evolutionary distances were computed using the Jones-Taylor-Thornton (JTT) matrix and are given by the number of amino acid substitutions per site. As a second method, the maximum-likelihood method using the JTT matrix to compute evolutionary distances was used (MEGA, version 5, beta release). The bootstrap consensus tree is based on 1,000 replicates. Nucleotide sequence accession numbers. All translated amino acid sequences of the newly cloned ascomycete CDH-encoding cDNA were submitted to GenBank, and the accession numbers are given in Table 2.

RESULTS Occurrence of CDH activity in ascomycete fungi. A total of 40 ascomycetes were screened for their ability to form CDH activity when they were grown on a cellulose-based medium to induce CDH formation. Enzymatic activity in the supernatant was detected by the CDH-specific cyt c assay (detection limit, 5 U liter⫺1; see Table S1 in the supplemental material). Twelve strains showed significant activity, namely, Acremonium strictum DSM 3567, Chaetomium atrobrunneum CBS 238.71, Corynascus thermophilus strains CBS 405.69 and 174.70, Dichomera saubinetii CBS 990.70, Hypoxylon bipapillatum CBS 375.86, Hypoxylon haematostroma CBS 255.63, Melanocarpus albomyces CBS 638.94, Monilia brunnea CBS 240.33, Neurospora crassa CBS 232.56, Scytalidium thermophilum CBS 619.91, and

1807

Stachybotrys bisbyi DSM 63042. Myriococcum thermophilum CBS 208.89, which is a known CDH producer (14), was chosen as a positive control for the screening. CDH activity varied greatly, starting from 9 U liter⫺1 for H. bipapillatum to more than 4,000 U liter⫺1 in the case of C. thermophilus. The maximum enzymatic activity was typically measured after 14 to 19 days of cultivation, with the exception of that of C. thermophilus (5 days). Cyt c activity was measured at pH 4.0 and 6.0 to differentiate between acidic and neutral CDHs. The pH of the initially acidic culture liquid (⬃5.5) increased during the cultivation of the CDH-producing strains to neutral or alkaline values. Only Sclerotium cepivorum acidified the culture medium during growth. Six strains did not produce detectable extracellular protein or cyt c activity under the chosen culture conditions, even though growth was visible (Capnodium salicinum, Cenococcum geophilum, Cheilymenia pulcherrima, Paecilomyces variotii, Rhizosphaera kalkhoffii, and Talaromyces thermophilus). On the contrary, large amounts of extracellular protein (ⱖ100 mg liter⫺1) but also no detectable CDH activity were found in cultures of Ceramothyrium linneae, Chaetomium thermophilum var. coprophilum, Chaetomium thermophilum var. thermophilum, Lamprospora wrightii, Malbranchea cinnamomea, Melanocarpus thermophilus, Stigmina compacta, Talaromyces emersonii, and Thermoascus thermophilus. Effect of carbon and nitrogen source concentrations on CDH production. The cyt c CDH activity in the culture supernatant was subsequently investigated in more detail for nine selected strains using media with various cellulose and peptone concentrations. The availability of cellulose during growth of the organism had a considerable impact on CDH formation (see Table S2 in the supplemental material). In general, high cellulose concentrations in the medium stimulated CDH production. The highest CDH production was observed in ChighNlow medium in the cultures of A. strictum, C. thermophilus, D. saubinetii, H. haematostroma, N. crassa, and S. bisbyi. On the contrary, the cultures of C. atrobrunneum, M. brunnea, and Sc. thermophilum showed the highest CDH activity in ChighNhigh medium. The increased amount of peptone in ClowNhigh medium did not improve CDH production over that in ClowNlow medium or even had a negative effect, as in the case of C. atrobrunneum, M. brunnea, and Sc. thermophilum. The extracellular protein concentration was always higher in media containing the higher cellulose concentration. A higher peptone concentration increased the amount of extracellular protein only in cultures of C. atrobrunneum, N. crassa, and Sc. thermophilum, had no significant influence on A. strictum, C. thermophilus, H. haematostroma, and M. brunnea, and decreased the amount of secreted protein strongly in cultures of D. saubinetii and S. bisbyi. In cultures of C. atrobrunneum, N. crassa, and M. brunnea, the maximum CDH activity was measured before the extracellular protein maximum was reached. D. saubinetii showed the opposite behavior, while for all other strains, the appearance of maximum CDH activity and protein concentration coincided. Most of the investigated strains showed CDH activity that was stable in the cultures at least for several days after it reached the maximal value, with the exception of H. haematostroma, for which the harvesting time was critical due to a fast decrease of CDH activity within 24 h after it reached the maximum.

1808

HARREITHER ET AL.

APPL. ENVIRON. MICROBIOL.

FIG. 1. Spectra of purified ascomycetous CDHs in both their oxidized (black line) and reduced (gray line) forms. Lactose was used to reduce the enzymes.

Purification. CDH was purified from the culture supernatants of six selected strains using an identical purification protocol. Throughout the purification procedure, CDH was detected by its absorbance at 280 and 420 nm and the DCIP assay, and only the purest fractions were pooled. Results are summarized in Table 1. In general, the CDH preparations thus obtained showed high specific activities (⬎9 U mg⫺1), with the exception of S. bisbyi CDH, which could not be purified satisfactorily with the applied procedure and still contained a brownish red chromophore produced by the fungus. The purification yields are only moderate because of strict pooling of only the purest fractions. All CDHs were stable throughout the purification, which was performed at room temperature. Starting from a low initial specific activity, C. atrobrunneum was purified 293-fold, whereas in cultures of C. thermophilus, the enzyme was already a major component of the extracellular proteins, and hence, only a low extent of purification was attained. The UV/visible spectra of all the purified enzymes (Fig. 1) show typical characteristics of a flavocytochrome, with the Soret band of the heme b cofactor being at about 420 nm and a broad shoulder occurring at between 450 and 500 nm, which is attributed to the FAD. The typical Soret band shift is observed upon reduction with lactose, while the ␣ and ␤ peaks of the heme b appear. Concomitantly, the absorbance in the region between 450 and 500 nm decreases due to the reduction of the FAD. To assess the purity of the various CDH preparations, the RZ values, defined by the ratio A420/A280, were calculated for each CDH in its oxidized state (Table 1). All six purified enzymes (CaCDH, CtCDH, DsCDH, HhCDH, NcCDH, and SbCDH) were further analyzed by SDS-PAGE (see Fig. S1 and S2 in the supplemental material) and IEF (see Fig. S3 in the supplemental material). The previously studied CDHs from Myriococcum thermophilum, Sclerotium rolfsii, and Trametes villosa were used as internal stan-

dards in IEF and gave pI values of 3.8, 4.2, and 4.1 and 4.3 (double band), respectively, which is in agreement with published results (2, 24). Since some of the CDH preparations showed several protein bands on IEF, the exact position of the CDH bands was determined by activity staining with DCIP. Only CtCDH was purified to electrophoretic homogeneity, resulting in a single band on both SDS-polyacrylamide and IEF gels. DsCDH shows a single band on SDS-PAGE, but additional faint bands on IEF. CaCDH and HhCDH show minor impurities on SDS-PAGE but significant impurities (⬃50%) on IEF. NcCDH and SbCDH were impure, with CDH contents of 30 and 15%, respectively; this is also indicated by the rather low RZ values for these two CDH preparations (0.20 and 0.12, respectively). The calculated molecular masses and isoelectric points of the various ascomycetous CDHs are summarized in Table 2. pH profiles. The pH profile for each of the six ascomycetous CDHs was determined both with DCIP and with cyt c as electron acceptors. The optimal pH values with DCIP were found to be in the range of pH 4.5 to 6.0, but distinct differences between different CDHs were observed (Fig. 2). Significant activity at pH 3.0 was found for CaCDH (34% of the maximum activity), DsCDH (36%), HhCDH (42%), and SbCDH (50%), while a very low substrate turnover was observed for CtCDH below a pH value of 4.0. CtCDH, however, showed considerable activity with DCIP at pH 7.5 (40% of maximum activity), together with SbCDH (50%) and HhCDH (20%). A very narrow pH profile was measured for NcCDH, while rather broad pH/activity peaks were obtained for SbCDH, DsCDH, and HhCDH. Whereas DCIP as a substrate interacts with the flavodomain of CDH, cyt c is exclusively reduced at the cytochrome domain of CDH, where it accepts electrons via the enzyme’s b-type heme. Hence, when cyt c is used to determine the pH profile of

VOL. 77, 2011


1809

FIG. 2. pH profiles of CDH activity from various ascomycetes showing relative activities for the two-electron acceptor 2,6-dichloro-indophenol (gray line, circles) and the one-electron acceptor cytochrome c (black line, squares), using lactose as electron donor.

CDH, this pH profile reflects the IET between the flavin and the heme domains, i.e., the transfer of electrons from FAD to cyt c. Typically, the pH optima for DCIP and cyt c of the various ascomycetous CDH activities investigated did not differ significantly, and pH maxima were found in the region of pH 5.0 and 6.0, for all CDHs with the exception of CtCDH. With the latter enzyme, the pH optimum determined was 5.0 when DCIP was the electron acceptor, whereas it was 7.5 with cyt c, indicating that IET is considerably favored at neutral pH. The activities of all CDHs were very low at pH 3.0 when cyt c was the substrate (⬍5% relative activity); however, some of these enzymes show considerable activity with cyt c at pH 7.5. Especially, the values for the relative activity at pH 7.5 for HhCDH (70%), NcCDH (49%), and SbCDH (60%) are remarkably high. As mentioned above, CtCDH even has optimum activity with cyt c at pH 7.5. Substrate specificity. Steady-state kinetic constants for the ␤-1,4-linked disaccharides cellobiose and lactose, the ␣-1,4linked disaccharide maltose, and the monosaccharide glucose were measured for all six enzymes using the two-electron acceptor DCIP (Table 3). Isoelectric focusing showed single bands for all CDH preparations, confirming that only one enzyme (CDH) with activity toward DCIP was present in each preparation. The kcat values given for these enzyme preparations should hence be considered approximations, which, however, are very helpful for reasons of comparison. The Michaelis-Menten constants for the ␤-1,4-linked substrates cellobiose and lactose are in the micromolar range, whereas the Km values for maltose and glucose are in the low and high millimolar ranges, respectively. Three of the six ascomycetous CDHs have a higher catalytic efficiency for cellobiose (cel) than for lactose (lac) [(kcat/Km)cel/(kcat/Km)lac]: 1.8fold for CaCDH, 1.7-fold for DsCDH, and 2.1-fold for NcCDH. For the others, lactose is an almost equal or even preferred substrate (CtCDH, 0.75-fold; HhCDH, 1.3-fold;

SbCDH, 1.1-fold). Catalytic efficiencies for maltose and glucose are generally much lower than those for cellobiose but differ widely between the species. It is worth pointing out that the kcat values for glucose (glc) are always in a range similar to that for cellobiose; for CtCDH this kcat,glc value, is however, almost twice the value measured for the presumed in vivo substrate cellobiose. The lower catalytic efficiencies for glucose, hence, always result from the unfavorably high Km values. Substrate inhibition was typically not observed for any enzyme with the exception of CaCDH (Ki,cel ⫽ 170 mM, Ki,lac ⫽ 175 mM) and DsCDH (Ki,cel ⫽ 475 mM, Ki,lac ⫽ 435 mM), which was, however, very weak. Analysis of gene and protein sequences. In addition to the two already known cdh genes from N. crassa (the sequences of the genes that we cloned from N. crassa CBS 232.56 are identical to the published sequences, except for a single nucleotide exchange in cdhIIa, which, however, is a silent mutation), seven cdh genes were cloned from five novel CDH-producing ascomycete strains (Table 2). In accordance with this, N. crassa, H. haematostroma, and S. bisbyi carry at least two cdh genes, whereas only one cdh gene could be identified in C. atrobrunneum, C. thermophilus, and D. saubinetii, despite great efforts to identify potentially existing additional genes (this does not exclude the presence of more cdh genes). The lengths of the genes coding for the N-terminal leader sequence and the secreted protein chains (765 to 816 amino acids) varies from 2,364 to 2,538 bp (Table 2). A CBM of 36 amino acids was found in most sequences, with the exceptions being those of CtCDH, HhCDH, and NcCDH IIB. In order to correlate the expressed cdh genes to the corresponding enzyme formed under the selected growth conditions, the purified CDHs were subjected to partial protein sequencing using mass spectrometry. In the case of CaCDH IIA (cdhIIa gene), CtCDH IIB (cdhIIb), and DsCDH IIA (cdhIIa), only one gene was identified in the genomic DNA of each strain and the

1810

HARREITHER ET AL.

APPL. ENVIRON. MICROBIOL.

TABLE 3. Apparent kinetic constants for carbohydrates (electron donors) for various ascomycetous CDHs Vmax (U mg⫺1)

kcat (s⫺1)

kcat/Km (M⫺1 s⫺1)

72 217 15,500 87,000

9.1 15.3 1.81 7.6

20.1 33.7 4.0 16.8

2.79 ⫻ 105 1.55 ⫻ 105 260 195

Cellobiose Lactose Maltose Glucose

140 200 2,200 120,000

9.5 18.1 1.32 18.8

13.0 24.7 1.80 25.6

0.93 ⫻ 105 1.24 ⫻ 105 820 215

5.0


125 215 8,000 79,000

16.7 17.4 0.83 14.2

27.8 28.9 1.38 23.6

2.22 ⫻ 105 1.34 ⫻ 105 175 300

HhCDH IIB

5.0


160 230 18,000 71,000

9.6 11.0 0.64 8.1

21.8 25.0 1.45 18.4

1.36 ⫻ 105 1.09 ⫻ 105 80 260

NcCDH IIA

5.5


160 325 120,000 4,000,000

12.7 12.4 0.14 9.5

53.1 51.8 0.60 37.7

3.32 ⫻ 105 1.59 ⫻ 105 5 9

SbCDH IIA2

5.5


170 185 50,000 1,700,000

19.9 20.2 0.51 24.2

1.17 ⫻ 105 1.09 ⫻ 105 10 14

CDH

Assay pHa

Electron donor

CaCDH IIA

6.0


CtCDH IIB

5.0

DsCDH IIA

a

Km (␮M)

2.75 2.80 0.070 3.35

Activity was measured at the indicated pH with 2,6-dichloro-indophenol as the saturating substrate.

peptide sequences determined showed 100% identity to the genes (see Table S5 in the supplemental material). H. haematostroma, N. crassa, and S. bisbii each carry two separate cdh genes, but apparently, only one of these two genes was expressed under the chosen culture conditions, as is also suggested by the presence of only a single active protein band identified by activity staining with DCIP on IEF. Mass spectrometry results corroborate the exclusive presence of only one CDH protein, namely, HhCDH IIB (encoded by cdhIIb), NcCDH IIA (cdhIIa), and SbCDH IIA2 (cdhIIa2) (see Table S5 in the supplemental material). By analyzing the sequences of the genes, the most obvious difference is that the cdhIIa genes contain a sequence encoding a CBM at their 3⬘ (downstream) ends, while the cdhIIb genes do not. It is, furthermore, exceptional that both genes found in S. bisbyi (cdhIIa1 and cdhIIa2) contain a CBM sequence. DISCUSSION Expression of CDH in ascomycete fungi. Overall, 35% of the screened ascomycetes produced detectable levels of CDH activity under the chosen culture conditions. This is a very high percentage of positives in a screening, and one additionally has to take into account the fact that the cultivation conditions in the screening probably might not have induced CDH activity in some of the screened ascomycete strains selected from the classes Sordariomycetes, Leotiomycetes, and Dothideomycetes. Most of the screened, positive strains belong to the Sordariomycetes, with the exceptions being D. saubinetii (Dothideomycetes) and M. brunnea (Leotiomycetes). Eleven out of

16 screened Sordariomycetes strains produced CDH activity; these were mainly from the order Sordariales (and, additionally, Hypocreales [A. strictum] and Xylariales [Hypoxylon spp.]). The previously published CDH-producing fungi Chaetomium sp. INBI 2-26(⫺) and Thielavia heterothallica also belong to this order, which consists of many wood- and dung-inhabiting species (45). This underlines the importance of CDH for lignocellulose degradation, and we propose that CDH is a widely distributed enzyme in ascomycetes, much more so than was previously anticipated. A recent study (18) investigated the occurrence of gene transcripts encoding lignocellulolytic enzymes in forest soil and also found four cdh genes; however, all were from basidiomycetes. This study stresses the importance of CDH for lignocellulose degradation, but it also highlights that the detection of ascomycete cdh gene transcripts is not easy and fails with nonoptimized primers. The mass percentage of CDH in the culture supernatants was calculated by dividing the CDH protein concentration in shaking flasks (calculated from the specific activity of the pure CDH and the measured volumetric activity) by the total extracellular protein concentration. The values obtained, which were between 0.2 and 3.3%, are in good agreement with published data on other fungal CDH producers. Among the screened ascomycetes, C. thermophilus is outstanding because of its ability to secrete up to 220 mg of CDH protein per liter medium, which is by far the highest value reported for an ascomycetous producer so far. However, the most impressive fact is that CDH accounts for 13 to 14% of the total extracellular protein secreted by C. thermophilus under these culture conditions, which emphasizes the

VOL. 77, 2011


importance of CDH for this fungus. We also report for the first time the production and characterization of wild-type CDH from the plant-pathogenic model organism N. crassa. Two cdh genes are already known from the genome-sequencing project of N. crassa (12). CDH formation was, without exception, increased in media containing high cellulose concentrations, which supports the proposed role of cellulose/cellobiose as an inducer of CDH synthesis. A high peptone concentration increased the formation only in C. atrobrunneum, N. crassa, and Sc. thermophilum. This is in contrast to basidiomycete producers, for which increased CDH production was demonstrated in media containing high concentrations of a complex nitrogen source (22, 24). During growth of most of the strains, the culture pH increased to neutral or even slightly alkaline. This corresponds to the reported pH conditions in their natural habitats, e.g., soil (pH 5.5 to 7.0) or compost, where the pH is raised from pH 5.0 to 7.5 (37). In contrast to white rot fungi, which acidify the culture liquid to pH values below 4.5 and do not grow well above pH 7.5, the screened ascomycetes were growing in the neutral or alkaline pH region, which certainly exerts an effect on the working pH range of the secreted CDHs. Catalytic properties of ascomycetous CDHs. The molecular properties of ascomycetous CDHs are, with the exception of the presence of CBM, very similar, with isoelectric points being between 3.8 and 4.4, molecular masses ranging from 81,840 to 87,414 Da, and the number of amino acids without the CBM ranging from 770 to 780 for class IIA and IIB CDHs. In regard to their catalytic properties, the CDHs characterized in this study can be classified into three groups, according to their pH profiles with cyt c, which gives a good approximation of the pH profile of the intramolecular electron transfer between the FAD-containing dehydrogenase and the cytochrome domain. The acidic group is composed of ascomycetous CDHs with a pH optimum of about pH 5.0. Additionally, acidic CDHs show a very narrow optimal working range. CaCDH and DsCDH, along with the previously studied CDH from M. thermophilum (14), are examples of acidic CDHs. Members of the intermediate group have acidic pH optima as well but have much broader pH-activity curves and thus exhibit notably high activities under alkaline pH conditions. Examples for intermediate CDHs are HhCDH, NcCDH, and SbCDH, which have pH optima of 5.0 to 6.0 while still showing significant activity at pH 7.5, with relative activities being 70%, 50%, and 60% of the maximum value, respectively. A previously described CDH with such a characteristic is from Chaetomium sp. INBI 2-26(⫺) (17). The last group comprises alkaline CDHs, which have their pH optima with cytochrome c in a neutral or alkaline pH region. A representative of this alkaline group is CtCDH, with a pH optimum at pH 7.5 and a broad activity peak over the range of pH 6 to 9. CDH from Humicola insolens is another published example (32, 40) for a CDH showing similar catalytic behavior. Interestingly, some of the new CDHs exhibited a pH optimum for cyt c that is more alkaline than that for DCIP. This is most distinct in the case of CtCDH, where the pH optimum for cyt c is found at pH 7.5, whereas the optimum for DCIP is at pH 5.0. These results differ completely from the pH profiles of basidiomycetous CDHs, where the pH optima are typically between pH 3.5 and 4.5 and the activity with cyt c above pH 6.0 is very low (⬍5%). The different pH

1811

profiles (acidic, intermediate, neutral/alkaline) show, however, no phylogenetic correlation. We propose that these variations reflect the adaptation to different, mostly slightly acidic or neutral environments, whereas basidiomycetous CDHs are adapted to the acidic milieu created by wood-degrading or phytopathogenic fungi. In general, the substrate specificity of ascomycete class II CDHs differs from that of basidiomycete class I CDHs by a less pronounced discrimination against the substrates lactose, maltose, and glucose, while class I CDHs clearly prefer cellobiose as their sugar substrate. There is no significant difference between the Km values (ranging from 50 to 200 ␮M) and kcat values (10 to 30 s⫺1) for cellobiose of class I CDHs and that of the class II CDHs of this study (44). The Km values for lactose in class I CDHs are 10- to 20-fold higher than those for cellobiose, while these differences are far less pronounced in class II CDHs, which have 1.1- to 3-fold higher Km values for lactose. On the basis of the catalytic efficiencies, three of the novel ascomycetous CDHs of this study even show no preference for cellobiose over lactose (CtCDH IIB, HhCDH IIB, SbCDH IIA2). The importance of several amino acid residues in the substrate-binding subsite B of CDH from Phanerochaete chrysosporium, which accommodates the nonreducing sugar moiety of disaccharide substrates of CDH, was reported in a recent study (7). Especially, residue Glu279 of PcCDH, which forms hydrogen bonds with the hydroxyl groups at C-2 and C-3 of the nonreducing sugar in subsite B, seems to be important for reactivity with lactose, since its exchange with Gln (E279Q variant of PcCDH) resulted in the complete loss of activity with lactose, while the variant was still active with cellobiose. Glu279 is strictly conserved in class I CDHs (43), while an equivalent position is taken up by a conserved Asn in all class II CDHs. Apparently, the active-site geometry differs significantly between class I and class II CDHs. In addition, the discrimination against maltose and glucose is far less pronounced in CDHs of class II than in class I CDHs, as is evident from the substrate selectivity values, the ratio of the catalytic efficiencies (kcat/Km) for two substrates (28). This ratio for the substrate pair cellobiose and maltose [(kcat/Km)cel/ (kcat/Km)mal] ranges from 115 to 1,700 for CaCDH, DsCDH, CtCDH, and HhCDH, while the substrate selectivity for cellobiose/glucose [(kcat/Km)cel/(kcat/Km)glc] ranges from 435 to 1,430 for these four class II CDHs. In contrast, the substrate selectivity values for the substrate pairs cellobiose/maltose and cellobiose/glucose for a typical class I CDH, e.g., from P. chrysosporium, are ⬃30,000 and 86,500, respectively, indicating the very strong discrimination against this ␣-linked disaccharide or against monosaccharides. The lowest values for the substrate selectivity of the substrate pair cellobiose/glucose were found for the class IIB CDHs from C. thermophilus and H. haematostroma: 435 and 525, respectively. Molecular and phylogenetic properties of CDHs. The phylogenetic relationship of 56 mature protein sequences was inferred by minimum evolution (Fig. 3; see also Fig. S5 in the supplemental material [sequences without CBM]) or maximum likelihood (see Fig. S6 in the supplemental material). All three phylogenetic trees are very similar: the four selected class I CDH sequences from the basidiomycete lignocellulolytic model organisms Phanerochaete chrysosporium, Trametes versicolor, and Ceriporiopsis subvermispora and the hemicellulase-

FIG. 3. Bootstrapped phylogenetic tree of 56 CDH mature protein sequences inferred by the minimum-evolution algorithm. Basidiomycete sequences cluster in class I, and ascomycete CDHs are separated into two clades: class II and class III. Within the class II clade, all sequences of CBM carrying CDHs are indicated by cbm⫹, which indicates that they are class IIA CDHs. Only CDHs isolated and characterized so far are additionally annotated by IIA when they carry a CBM or by IIB when they lack it. CDH has been isolated from species marked in gray shading. 1812

VOL. 77, 2011


1813

FIG. 4. Multiple-sequence alignment of cellulose binding modules of the ascomycete CDHs and three related CBM1 sequences from P. chrysosporium cellobiohydrolase I (PcCBHI), ␤-glucosidase (PcBGL), and cytochrome b562 (PcCytb562) (A). Alignment of the proposed cellulosebinding sequences from basidiomycetous CDHs (CsCDH, CDH from C. subvermispora; PcCDH, CDH from P. chrysosporium; TvCDH, CDH from T. versicolor) and the corresponding positions of 12 ascomycete CDHs (B). Proposed binding residues are marked by arrows. MtCDH, CDH from M. thermophilum; ThCDH, CDH from T. heterothallica.

producing plant pathogen Sclerotium rolfsii formed a well-defined class, as was supposed for this outgroup. For ascomycete cdh sequences, the previously described separation into class II and class III was also observed (44). Putative ascomycete CDHs from genome sequencing projects (including species from Aspergillus, Botryotinia, Gibberrella, Glomerella, Nectria, Neosartorya, Phaeosphaeria, Pyrenophora, Sclerotinia, and Verticillium) are found in class III, which is a strongly supported group, but also in class II. All the putative CDHs in class III lack a CBM, whereas all sequences with a CBM can be found in class II (including Botryotinia fuckeliana and Pyrenophora spp.). Additionally, putative cdh sequences from Aspergillus spp. and Neosartorya fischeri lacking a CBM form their own very well defined group in class II. The other sequences found in class II come from the class of Sordariomycetes, mostly the order Sordariales, but also of other orders, like Hypocreales, Magnaporthales, Phyllachorales, and Xylariales. All sequences of the most numerously presented Sordariales separate into two subclasses: class IIA, with a CBM, and class IIB, without a CBM. This separation pattern is also observed for the cdh genes of other Sordariomycetes from the genomes of Neurospora crassa (12), Magnaporthe oryzae (5), Podospora anserina (9), Chaetomium globosum, and Sordaria macrospora (29), but because of order-specific differences, the separation into two clearly distinct phylogenetic groups is not observed. The identities of such cdh genes with CBM versus genes without CBM are, however, quite low and support the division into subclasses IIA and IIB (H. haematostroma strains with GenBank acces-

sion number HQ116818 versus those with GenBank accession number HQ116817, 60% identity; M. oryzae, 55% identity; Ch. globosum 54% identity). The cdhIIa genes are characterized by a C-terminal family 1-type carbohydrate-binding module (CBM1), while the cdhIIb genes lack CBM1. However, the evolutionary difference between class IIA and class IIB was not found only in the presence or absence of the CBM, which is a minor (⬃5%) part of the sequence. The same phylogenetic pattern is also observed in Fig. S5 in the supplemental material when the CBM-encoding sequence was removed from the data set. In Sordariales, sequence identities of the mature proteins are approximately 10% lower between class IIA and class IIB CDHs within the same fungus (NcCDH IIA versus NcCDH IIB, 55% identity; HhCDH class IIA and HhCDH class IIB, 58% identity, M. oryzae CDH class IIA and M. oryzae CDH class IIB, 54% identity; P. anserina CDH class IIA and P. anserina CDH class IIB, 58% identity) than within the same subclass (NcCDH IIA versus P. anserina CDH IIA, 71% identity; HhCDH IIA versus M. oryzae CDH IIA, 64% identity; NcCDH IIB versus P. anserina CDH IIB, 68% identity; HhCDH IIB versus NcCDH IIB, 72% identity). A sequence comparison of the carbohydrate-binding modules from various cdh genes with the CBMs of ␤-glucosidase, cellobiohydrolase I, and cytochrome b562 from P. chrysosporium is shown in Fig. 4A. The central elements of the CBM, which enable class IIA CDHs to bind to cellulose, seem to be well conserved. The binding interactions between these CBMs

1814

HARREITHER ET AL.

and cellulose are attributed to conserved aromatic residues and four structurally important cysteines. While the CBMs of ␤-glucosidase, cellobiohydrolase I, and cytochrome b562 typically contain no or only a single charged amino acid residue, the CBMs of the ascomycete CDHs feature 4 to 6 positively and negatively charged residues, which might have a function in pH-dependent absorption/desorption of CDH onto cellulose. The binding of class IIA CBMs might not be restricted to cellulose, as is the case for class I CDHs, which have a very specific cellulose-binding site at the dehydrogenase domain. It is likely that class IIA CBMs also interact with xylan, mannan, arabinan, or chitin, as was shown for the CBM of cytochrome b562 from the basidiomycete P. chrysosporium (41). Interestingly, the cellulose-binding site on class I dehydrogenase domains, which is attributed to nine aromatic amino acids (15), is not observed in any class II CDH (Fig. 4B). This strongly suggests that class IIB CDHs lacking the CBM cannot bind to cellulose or hemicelluloses and have a different physiological role. Results obtained in an expression study of N. crassa, showing a much higher mRNA induction level of cdhIIa over that of cdhIIb in microcrystalline cellulose , (Avicel)-containing medium, also point in this direction (36). Conclusions. As shown by this study, cellobiose dehydrogenase is widely distributed in ascomycete fungi and appears commonly in wood-degrading/composting sordariomycetes, as well as other classes. Although cdh genes with and without CBM were previously known, this study demonstrates the frequent and often simultaneous presence of cdh genes with and without a 3⬘ cbm region in ascomycete strains. These concurrently occurring cdh genes belong to two distinct subclasses of class II CDHs. Both cdhIIa and cdhIIb genes might frequently occur together in an organism, as suggested by the genomes of fully sequenced ascomycetes, which is certainly not a coincidence (the cellulose/hemicellulose-binding properties of a CDH protein could be more elegantly and efficiently regulated by alternative splicing). This suggests different functions for class IIA and class IIB CDHs, which are not limited to their cellulose-binding properties but most likely comprise other properties as well, e.g., substrate specificity. The presence of a CBM together with the high preference for cellobiose in class IIA CDHs seems logical and indicates a role similar to that for class I CDHs, i.e., an early attack on cellulose and other recalcitrant material. The less tight substrate specificity of class IIB CDHs, which might also include the oxidation of mono-, di-, or oligomeric hemicelluloses and their presumably noncellulose-associated localization, suggests a different role for class IIB CDHs. ACKNOWLEDGMENTS We thank the following agencies for financial support: the Austrian Science Fund (translational project FWF L395-B11), the European Commission (project 3D, Nanobiodevice NMP4-SL-2009-229255), the Swedish Research Council (project 621-2007-4124), and the Austrian Academy of Sciences (APART project 11322). Roland Ludwig is recipient of an APART fellowship from the Austrian Academy of Sciences at the Department of Food Sciences and Technology, BOKU—University of Natural Resources and Life Sciences, Vienna, Austria. REFERENCES 1. Baldrian, P., and V. Valaskova. 2008. Degradation of cellulose by basidiomycetous fungi. FEMS Microbiol. Rev. 32:501–521.

APPL. ENVIRON. MICROBIOL. 2. Baminger, U., B. Nidetzky, K. D. Kulbe, and D. Haltrich. 1999. A simple assay for measuring cellobiose dehydrogenase activity in the presence of laccase. J. Microbiol. Methods 35:253–259. 2a.Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248–254. 3. Canevascini, G., P. Borer, and J. L. Dreyer. 1991. Cellobiose dehydrogenases of Sporotrichum (Chrysosporium) thermophile. Eur. J. Biochem. 198:43–52. 4. Coudray, M. R., G. Canevascini, and H. Meier. 1982. Characterization of a cellobiose dehydrogenase in the cellulolytic fungus Sporotrichum (Chrysosporium) thermophile. Biochem. J. 203:277–284. 5. Dean, R. A., et al. 2005. The genome sequence of the rice blast fungus Magnaporthe grisea. Nature 434:980–986. 6. Dekker, R. F. H. 1980. Induction and characterization of a cellobiose dehydrogenase produced by a species of Monilia. J. Gen. Microbiol. 120:309–316. 7. Desriani, S. Ferri, and K. Sode. 2010. Functional expression of Phanerochaete chrysosporium cellobiose dehydrogenase flavin domain in Escherichia coli. Biotechnol. Lett. 32:855–859. 8. Dragosits, M., et al. 2009. The effect of temperature on the proteome of recombinant Pichia pastoris. J. Proteome Res. 8:1380–1392. 9. Espagne, E., et al. 2008. The genome sequence of the model ascomycete fungus Podospora anserina. Genome Biol. 9:R77. 10. Fa ¨hnrich, P., and K. Irrgang. 1982. Conversion of cellulose to sugars and cellobionic acid by the extracellular enzyme system of Chaetomium cellulolyticum. Biotechnol. Lett. 4:775–780. 11. Fang, J., Y. Qu, and P. Gao. 1997. Wide distribution of cellobiose-oxidizing enzymes in wood-rot fungus indicates a physiological importance in lignocellulosics degradation. Biotechnol. Tech. 11:195–197. 12. Galagan, J. E., et al. 2003. The genome sequence of the filamentous fungus Neurospora crassa. Nature 422:859–868. 13. Hallberg, B. M., G. Henriksson, G. Pettersson, and C. Divne. 2002. Crystal structure of the flavoprotein domain of the extracellular flavocytochrome cellobiose dehydrogenase. J. Mol. Biol. 315:421–434. 14. Harreither, W., V. Coman, R. Ludwig, D. Haltrich, and L. Gorton. 2007. Investigation of graphite electrodes modified with cellobiose dehydrogenase from the ascomycete Myriococcum thermophilum. Electroanalysis 19:172– 180. 15. Henriksson, G., A. Salumets, C. Divne, and G. Petersson. 1997. Studies of cellulose binding by cellobiose dehydrogenase and a comparison with cellobiohydrolase 1. Biochem. J. 324:833–838. 16. Hyde, S. M., and P. M. Wood. 1997. A mechanism for production of hydroxyl radicals by the brown-rot fungus Coniophora puteana: Fe(III) reduction by cellobiose dehydrogenase and Fe(II) oxidation at a distance from the hyphae. Microbiology 143:259–266. 17. Karapetyan, K. N., et al. 2006. Properties of neutral cellobiose dehydrogenase from the ascomycete Chaetomium sp. INBI 2-26(⫺) and comparison with basidiomycetous cellobiose dehydrogenases. J. Biotechnol. 121:34–48. 18. Kellner, H., and M. Vandenbol. 2010. Fungi unearthed: transcripts encoding lignocellulolytic and chitinolytic enzymes in forest soil. PLoS One 5:e10971. 19. Kremer, S. M., and P. M. Wood. 1992. Production of Fenton’s reagent by cellobiose oxidase from cellulolytic cultures of Phanerochaete chrysosporium. Eur. J. Biochem. 208:807–814. 20. Liers, C., R. Ullrich, K. T. Steffen, A. Hatakka, and M. Hofrichter. 2006. Mineralization of 14C-labelled synthetic lignin and extracellular enzyme activities of the wood-colonizing ascomycetes Xylaria hypoxylon and Xylaria polymorpha. Appl. Microbiol. Biotechnol. 69:573–579. 21. Liu, D., S. Coloe, R. Baird, and J. Pederson. 2000. Rapid mini-preparation of fungal DNA for PCR. J. Clin. Microbiol. 38:471. 22. Ludwig, R., and D. Haltrich. 2003. Optimisation of cellobiose dehydrogenase production by the fungus Sclerotium (Athelia) rolfsii. Appl. Microbiol. Biotechnol. 61:32–39. 23. Ludwig, R., W. Harreither, F. Tasca, and L. Gorton. 2010. Cellobiose dehydrogenase: a versatile catalyst for electrochemical applications. ChemPhysChem 11:2674–2697. 24. Ludwig, R., et al. 2004. Characterisation of cellobiose dehydrogenases from the white-rot fungi Trametes pubescens and Trametes villosa. Appl. Microbiol. Biotechnol. 64:213–222. 25. Mansfield, S. D., E. De Jong, and J. N. Saddler. 1997. Cellobiose dehydrogenase, an active agent in cellulose depolymerization. Appl. Environ. Microbiol. 63:3804–3809. 26. Mason, M. G., M. T. Wilson, A. Ball, and P. Nicholls. 2002. Oxygen reduction by cellobiose oxidoreductase: the role of the haem group. FEBS Lett. 518:29–32. 27. McIlvaine, T. C. 1921. A buffer solution for colorimetric comparison. J. Biol. Chem. 49:183–186. 28. Morley, K. L., and R. J. Kazlauskas. 2005. Improving enzyme properties: when are closer mutations better? Trends Biotechnol. 23:231–237. 29. Nowrousian, M., et al. 2010. De novo assembly of a 40 Mb eukaryotic genome from short sequence reads: Sordaria macrospora, a model organism for fungal morphogenesis. PLoS Genet. 6:e1000891. 30. Rzhetsky, A., and M. Nei. 1992. A simple method for estimating and testing minimum evolution trees. Mol. Biol. Evol. 9:945–967.

VOL. 77, 2011


31. Sachslehner, A. 1997. Production of hemicellulose- and cellulose-degrading enzymes by various strains of Sclerotium rolfsii. Appl. Biochem. Biotechnol. 63-65:189–201. 32. Schou, C., M. H. Christensen, and M. Schu ¨lein. 1998. Characterization of a cellobiose dehydrogenase from Humicola insolens. Biochem. J. 330:565–571. 33. Shary, S., S. A. Ralph, and K. E. Hammel. 2007. New insights into the ligninolytic capability of a wood decay ascomycete. Appl. Environ. Microbiol. 73:6691–6694. 34. Subramaniam, S. S., S. R. Nagalla, and V. Renganathan. 1999. Cloning and characterization of a thermostable cellobiose dehydrogenase from Sporotrichum thermophile. Arch. Biochem. Biophys. 365:223–230. 35. Tamura, K., J. Dudley, M. Nei, and S. Kumar. 2007. MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24:1596–1599. 36. Tian, C., et al. 2009. Systems analysis of plant cell wall degradation by the model filamentous fungus Neurospora crassa. Proc. Natl. Acad. Sci. U. S. A. 106:22157–22162. 37. Tuomela, M., M. Vikman, A. Hatakka, and M. Ita ¨vaara. 2000. Biodegradation of lignin in a compost environment: a review. Biores. Technol. 72:169– 183. 38. Vasil’chenko, L. G., V. V. Khromonygina, K. N. Karapetyan, O. V. Vasilenko,

39. 40.

41.

42.

43.

44.

45.

1815

and M. L. Rabinovich. 2005. Cellobiose dehydrogenase formation by filamentous fungus Chaetomium sp. INBI 2-26(⫺). J. Biotechnol. 119:44–59. Worrall, J. J., S. E. Anagnost, and R. A. Zabel. 1997. Comparison of wood decay among diverse lignicolous fungi. Mycologia 89:199–219. Xu, F., et al. 2001. Humicola insolens cellobiose dehydrogenase: cloning, redox chemistry, and “logic gate”-like dual functionality. Enzyme Microb. Technol. 28:744–753. Yoshida, M., et al. 2005. Characterization of carbohydrate-binding cytochrome b-562 from the white-rot fungus Phanerochaete chrysosporium. Appl. Environ. Microbiol. 71:4548–4555. Zamocky, M., B. M. Hallberg, R. Ludwig, C. Divne, and D. Haltrich. 2004. Ancestral gene fusion in cellobiose dehydrogenases reflects a specific evolution of GMC oxidoreductases in fungi. Gene 338:1–14. Zamocky, M., et al. 2006. Cellobiose dehydrogenase—a flavocytochrome from wood-degrading, phytopathogenic and saprotropic fungi. Curr. Protoc. Pept. Sci. 7:255–280. Zamocky, M., et al. 2008. Cloning, sequence analysis and heterologous expression in Pichia pastoris of a gene encoding a thermostable cellobiose dehydrogenase from Myriococcum thermophilum. Prot. Expr. Purif. 59:258– 265. Zhang, N., et al. 2006. An overview of the systematics of the Sordariomycetes based on a four-gene phylogeny. Mycologia 98:1076–1087.