APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Oct. 1997, p. 3804–3809 0099-2240/97/$04.0010 Copyright © 1997, American Society for Microbiology
Vol. 63, No. 10
Cellobiose Dehydrogenase, an Active Agent in Cellulose Depolymerization SHAWN D. MANSFIELD, ED
JOHN N. SADDLER*
Department of Wood Science, Faculty of Forestry, University of British Columbia, Vancouver, British Columbia, Canada Received 26 June 1997/Accepted 6 August 1997
The ability of cellobiose dehydrogenase purified from Phanerochaete chrysosporium to modify a Douglas fir kraft pulp was assessed. Although the addition of cellobiose dehydrogenase alone had little effect, supplementation with cellobiose and iron resulted in a substantial reduction in the degree of polymerization of the pulp cellulose. When the reaction was monitored over time, a progressive depolymerization of the cellulose was apparent with the concomitant production of cellobiono-1,5-lactone. Analysis of the reaction filtrates indicated that glucose and arabinose were the only neutral sugars generated. These sugars are derived from the degradation of the cellobiose rather than resulting from modifications of the pulp. These results suggest that the action of cellobiose dehydrogenase results in the generation of hydroxyl radicals via Fenton’s chemistry which subsequently results in the depolymerization of cellulose. This appears to be the mechanism whereby a substantial reduction in the degree of polymerization of the cellulose can be achieved without a significant release of sugar.
have been determined by small-angle X-ray scattering (28), while the cDNA of Phanerochaete chrysosporium CDH has been successfully cloned and sequenced (34). It has also been shown that the proteolysis of CDH both in vitro (20) and by the action of isolated proteases (13, 16) results in the generation of a second active enzyme, CBQ (45). CBQ is a flavoenzyme which, in the presence of cellobiose, can reduce compounds such as quinones and phenoxy radicals (1). Although these enzymes were originally isolated and characterized more than 20 years ago (41, 42), their role in wood decomposition has yet to be clearly defined. Several putative functional roles for this enzyme have been suggested: CDH acting as a vehicle to generate radicals as antibacterial agents (43); the reduction of quinones as a defense mechanism against toxins (32); CDH acting as a regulatory enzyme preventing the repolymerization of lignin radicals generated by other oxidative enzymes (1, 14); and the generation of highly active hydroxyl radicals which participate in Fenton’s reactions to degrade cellulose, xylan, and lignin (18, 46). However, as Morpeth so elegantly put it, “cellobiose oxidoreductases are enzymes in search of a function” (32). Although CDH has been shown to both enhance the action of cellulases on crystalline cellulose (7) and degrade model wood components such as carboxymethylcellulose, xylan, and synthetic lignin (18), there have been few studies to date on its direct role in cellulose modification in substrates such as pulp. In this report, we describe how we isolated CDH from the white rot fungus P. chrysosporium and assessed its ability to modify the degree of polymerization (DP) of a Douglas fir kraft pulp in the absence or presence of various cofactors.
Microorganisms are known to degrade the cellulose, hemicellulose, and lignin components of wood through the action of extracellular enzymes. Hydrolytic enzymes such as cellulases, xylanases, and mannanases contribute to the degradation of the carbohydrate moieties, while oxidative enzymes such as laccases, lignin peroxidase, and manganese peroxidase, in combination with low-molecular-weight mediators, have been shown to be involved in lignin biodegradation (10). Although many of the wood-degrading fungi can be grouped into categories such as white rot, brown rot, or soft rot fungi, this delineation does not readily carry over into the specific enzymes produced by these organisms involved in the degradation of the different wood components. While there has been a great deal of progress made in characterizing many of the enzymes involved in the degradation of the native wood constituents, there is still the apprehension that some enzymes may be multifunctional and cause modifications to more than one or all of the wood components. One such group of enzymes are the extracellular cellobioseoxidizing enzymes cellobiose dehydrogenase (CDH) and cellobiose:quinone oxidoreductase (CBQ), which have been shown to be produced by a number of basidiomycete fungi, including both white rot fungi (4, 22, 36, 41) and a brown rot fungus (37). This class of enzymes has also been isolated from the culture filtrates of nonligninolytic, cellulolytic microorganisms such as Monales sitophila (12), Chaetomium cellulolyticium (15), Sporotrichium thermophile (8), and bacterial strains (29). CDH is an enzyme consisting of two prosthetic groups, a heme and a flavin adenine dinucleotide moiety (hemoflavoenzyme). The latter domain is directly involved in both the oxidative and reductive half-reactions, while the former stimulates the reduction of one-electron acceptors such as cytochrome c and Fe31 (19, 20). Recently, the size and shape of this enzyme
MATERIALS AND METHODS Growth conditions. P. chrysosporium BKM F-1767 was maintained on agar plates containing 5 g of glucose and 3.5 g of malt extract per liter. Plates were inoculated, and organisms were grown at 27°C. The growth medium was a modification of the medium described by Bao et al. (6), containing 2.28 g of (NH4)2HPO4, 0.5 g of MgSO4 z 7H2O, 0.74 g of CaCl2, 0.01 g of FeCl3, 0.0158 g of NaNO3, 6.6 mg of ZnSO4 z 7H2O, 3.8 mg of MnSO4 z 1H2O, 1 of mg CoCl2 z 6H2O, 0.1 mg of thiamine-HCl, 6.75 g of succinate disodium salt hexahydrate, and 10 g of unquilted cotton cosmetic pads (Safeway) per liter. P. chrysosporium was pregrown in 20 ml of medium containing 2% glucose and 0.5% yeast extract
* Corresponding author. Mailing address: Faculty of Forestry, 2702357 Main Mall, University of British Columbia, Vancouver, B.C., Canada V6T 1Z4. Phone: (604) 222-3200. Fax: (604) 222-3267. E-mail: [email protected]
VOL. 63, 1997 in 250-ml Erlenmeyer flasks at 27°C for 2 days at 150 rpm. The cultures were then homogenized in a Waring blender for 8 s and used to inoculate 250 ml of the succinate medium in 500-ml Erlenmeyer flasks. The cultures were grown for 30 days at 27°C. Protein purification. Culture filtrates were collected by vacuum filtration, and the extracellular protein present in the supernatant was precipitated with 90% (NH4)2SO4 at 0°C. Precipitated protein was collected by centrifugation at 8,000 rpm for 2 h. Between each purification step, the protein solution was washed with 10 mM ammonium acetate (pH 4.5) and concentrated with an Amicon system, using a Diaflo YM10 membrane with a 10-kDa cutoff. The concentrated protein was loaded on a DEAE-Sepharose CL-6B column (Pharmacia Biotech Inc.) equilibrated with 10 mM ammonium acetate (pH 4.5) and eluted with 1.5 liters of a linear gradient of 10 to 250 mM ammonium acetate (pH 4.5). Only fractions with an A420/A280 ratio above 0.03 were pooled, washed, and concentrated. Subsequently, concentrated protein was loaded on a Phenyl-Superose HR 5/5 column (Pharmacia) equilibrated with 0.6 M (NH4)2SO4 in 50 mM ammonium acetate (pH 4.0) and eluted with 30 ml of a linear gradient of 0.6 to 0 M (NH4)2SO4 in 50 mM ammonium acetate (pH 4.0). Fractions with an A420/A280 ratio above 0.1 were pooled, washed, concentrated, and applied on a Superose 6 HR 10/30 column (Pharmacia) equilibrated with 50 mM ammonium acetate (pH 5.0). After gel filtration, CDH-containing fractions were pooled, concentrated, and stored at 280°C before use. Enzyme assays. Cellobiose dehydrogenase activity was assayed by the reduction of cytochrome c at 550 nm (ε 5 28 mM21 z cm21). The assay mixture contained 3 mM cellobiose, 20 mM succinate (pH 4.5), 12.5 mM cytochrome c, and various amounts of enzyme preparations to a total of 1 ml. 2,6-Dichlorophenolindophenol (DCPIP) was used to measure the combined CDH and CBQ activity at 515 nm (ε 5 6.8 mM21 z cm21). The DCPIP assay mixture contained 3 mM cellobiose, 20 mM succinate (pH 4.5), 7.5 mM DCPIP, and various amounts of enzyme preparations to a total of 1 ml. All assays were performed at 23°C. Enzyme activity, expressed in international units (IU), was equivalent to the reduction of 1 mmol of DCPIP per min. Cytochrome c activity of CDH was found to be equivalent to 3.4 U of the DCPIP activity. Pulp chelation and characterization. Unbleached kraft pulp derived from Douglas fir (Pseudotsuga menziesii) was obtained from the Crofton mill (Fletcher Challenge), British Columbia, Canada. The pulp was fractionated in a BauerMcNett fiber length classifier to collect the 14R fraction (pulp fiber entrapped by sieve openings of 1.18 mm). Thus, only the longest fiber length fraction was retained (TAPPI Test Method T233 cm-82), and this was subsequently used as the substrate for the enzymatic treatments. For the chelation of metals, pulp samples were maintained at a 2% consistency and 50°C, while the pH was adjusted to 5 with H2SO4. The sodium salt of EDTA was applied as a concentrated 38% solution at a charge of 0.6% on the pulp (oven dry basis), and the chelation was allowed to continue for 30 min at 50°C. The pulp was then thoroughly washed with distilled water. The pulp was then analyzed for metal solids by inductively coupled plasma-optical emission spectrophotometry (Analytical Service Laboratory Ltd., Vancouver, B.C., Canada). These results indicated that the chelation step had significantly reduced the solid metal content of the pulp. Specifically, copper, iron, calcium, magnesium, and manganese had been reduced from 228 to 7 ppm, 67 to 21 ppm, 796 to 169 ppm, 79 to 42 ppm, and 5 to ,2 ppm, respectively. Pulp treatment conditions. Douglas fir kraft pulp (10 mg/ml) was treated with CDH (0.1 IU/ml) alone in 50 mM sodium acetate buffer (pH 4.5) at 30°C for 18 h. CDH was also supplemented with (final concentrations) (i) 20 mM cellobiose, (ii) 1.8 mM hydrogen peroxide, (iii) 0.2 mM ferric chloride (FeCl3), (iv) 20 mM cellobiose–1.8 mM hydrogen peroxide, (v) 20 mM cellobiose–0.2 mM ferric chloride, or (vi) 20 mM cellobiose–1.8 mM hydrogen peroxide–0.2 mM ferric chloride. We also investigated controls of the above combinations without added CDH as well as with pulp samples which had previously undergone a chelation step. A time course experiment was conducted with CDH (0.1 IU/ml) in 50 mM sodium acetate buffer (pH 4.5) at 30°C, supplemented with 20 mM cellobiose–0.2 mM ferric chloride. The reactions were terminated at 1, 2, 6, 12, and 18 h, and the reaction mixtures were inactivated by boiling for 15 min. Reaction filtrates were then removed; a portion was used directly for monosaccharide and oligosaccharide determination, while the remaining portion was freeze-dried and acid hydrolyzed for determination of wood sugars solubilized by the enzymatic treatments. The DP of the pulp was determined after tricarbanylation. Statistical analysis. All reactions were done in duplicate, while the specific quantifications were averages of replicate evaluations. Statistical analysis was conducted with SYSTAT 6.0 for Windows. Analysis of variance showed that there were significant differences between all reaction scenarios at the 95% confidence level. Pairwise comparisons between treatments were done by using a Bonferroni adjusted t test (95% confidence level). Statistical analysis clearly indicated that all control reactions were not significantly different from the pulp alone, with the exception of the reaction scenario containing hydrogen peroxide and iron. Furthermore, all samples containing CDH, cellobiose, and iron were significantly different from the controls at the 95% confidence interval. Carbohydrate determination. The lignin and sugar compositions of the pulp were determined by using replicated sulfuric acid hydrolysates (TAPPI Test Method T249 cm-85). Each hydrolysate was filtered through a sintered-glass
CELLULOSE DEPOLYMERIZATION BY CDH
filter of medium coarseness for the gravimetric determination of Klason lignin (acid-insoluble lignin), and its absorbance at 205 nm was measured for the quantification of acid-soluble lignin (TAPPI Useful Method UM250, 1991). Both the carbohydrates found within the pulp samples and those solubilized during the enzymatic treatments were also measured by high-performance liquid chromatography (HPLC). The oligosaccharide constituents and monomeric wood sugars (before and after secondary acid hydrolysis) were quantified by high-performance anion-exchange chromatography. Monomeric sugars were analyzed as described previously (11). Oligomeric sugars and cellobionic acid, which was generated by the base hydrolysis (150 mM NaOH conditions) of the cellobiono1,5-lactone during HPLC analysis, were separated on a CarboPac PA-1 column, using a Dionex DX-500 HPLC system (Dionex, Sunnyvale, Calif.) controlled by Peaknet 4.30 software. Calcium cellobionate (ICN Pharmaceuticals, Cleveland, Ohio) was used as a standard for cellobionic acid. The column was equilibrated with 150 mM NaOH and 50 mM sodium acetate and regenerated with 300 mM NaOH. After injection of 20 ml of sample via a SpectraSYSTEM AS3500 autoinjector (Spectra-Physics, Fremont, Calif.), the oligosaccharides were eluted with a 50 to 200 mM gradient of sodium acetate (over 20 min) at a flow rate of 1 ml/min. The oligosaccharides were monitored with a Dionex ED40 electrochemical detector (gold electrode), with parameters set for pulsed amperometric detection of sugars as recommended by the manufacturer. DP. The molecular weight distributions of both the control and enzymetreated pulps were obtained by gel permeation chromatography (GPC) analyses of their tricarbanyl derivatives. Carbanylation of the cellulose was carried out as described previously (38). The cellulose tricarbanylate was recovered by evaporation of the reaction solvents (44), which was subsequently treated with isooctane, evaporated to dryness, and solubilized in tetrahydrofuran (THF) at concentrations of approximately 0.2 mg/ml. The GPC of the tricarbanyl derivatives was carried out on a Waters 625 liquid chromatography system (Millipore Corp., Milford, Mass.). The cellulose tricarbanylate samples were filtered through a Teflon membrane (0.45 mm) and analyzed by using a series of four TSK-GEL columns (Varian [Sunnyvale, Calif.] types G1000 HXL, G3000 HXL, G4000 HXL, and G6000 HXL with molecular weight cutoffs of 103, 6 3 104, 4 3 105, and 4 3 107, respectively). THF was used as the eluting solvent at a flow rate of 1 ml/min. The samples in the eluent were detected by a Waters 486 UV spectrophotometer (Millipore) at a wavelength of 254 nm. The GPC calibration curve was generated from the elution profile of polystyrene standards with narrow molecular weight distributions. Using the MarkHouwink coefficients previously reported for polystyrene in THF, Kp 5 1.18 3 1024 and ap 5 0.74, and for cellulose tricarbanylate in THF, Kc 5 2.01 3 1025 and ac 5 0.92, the molecular weight of the tricarbanylated cellulose was obtained (39). The DP of cellulose was obtained by dividing the molecular weight of the tricarbanylated polymer (MW) by the corresponding molecular weight of the tricarbanylated derivative of anhydroglucose (DP 5 MW/519).
RESULTS As we had previously used Douglas fir kraft pulps to assess the actions of various cellulases and xylanases (30, 31), we thought that this would be an effective substrate to use to evaluate what changes, if any, CDH caused to the carbohydrate constituents of the pulp. In an attempt to create a homogeneous substrate, the pulp was fractionated and only the longest fiber length fraction (14R) was used. This eliminated some of the variability that can arise in treating whole pulps. The 14R fiber length fraction accounted for 55% of the total pulp and was composed of approximately 75% glucose, 5.5% xylose, 6.3% mannose, 0.5% galactose, 0.4% arabinose, and 4.1% total lignin. Previous work has shown that CDH initiates a two-electron oxidation of cellodextrins, generating the corresponding lactone (21). Cellobiose was found to be the substrate of choice, with cellodextrins of increasing DP demonstrating reduced activity. Meanwhile limited activity was observed when glucose was used as the substrate (19). The reductive half-reactions (see reaction scheme below) include a preferential single-electron reduction of Fe(III) to Fe(II) (9), while in situations of limited Fe(III), CDH can reduce O2 to generate H2O2 directly (40, 41) or via the production of superoxide which can subsequently react with Fe(II) (46). Furthermore, autooxidation of Fe(II) can result in the generation of hydrogen peroxide (46). Together hydrogen peroxide and reduced iron undergo Fenton’s chemistry, generating hydroxyl radicals, which then can
MANSFIELD ET AL.
APPL. ENVIRON. MICROBIOL.
FIG. 1. DP of CDH-treated Douglas fir kraft pulp. Reactions were conducted in 50 mM sodium acetate (pH 4.5) for 18 h at 30°C. The pulp (10 mg/ml) was incubated with combinations of 0.1 IU of CDH per ml, 20 mM cellobiose (C), 1.8 mM H2O2 (H), and 0.2 mM FeCl3 (F).
actively be involved in the attack on localized substrates (5, 17, 24, 25): CDHox 1 cellobiose 3 CDHred 1 cellobiono-1,5-lactone 2CDHred 1 2Fe31 3 2CDHox 1 2Fe21 CDHred 1 O2 3 CDHox 1 H2O2 2Fe21 1 O2 1 2H1 3 2Fe31 1 H2O2 Fe21 1 H2O2 3 Fe31 1 OH2 1 OHz
(1) (2) (3) (4) (5)
The potential role of these various components encouraged us to assess the effect of different combinations of CDH, cellobiose, iron, and hydrogen peroxide on the molecular weight distribution of the 14R fraction of the Douglas fir kraft pulp (Fig. 1; Table 1). All control combinations without CDH (data not shown) gave profiles similar to that of the original sample in buffer (DPN 5 341; DPW 5 2,337) with the exception of the reaction mixtures containing both hydrogen peroxide and ferric chloride (Fig. 1, line 2). This combination generates Fenton’s reagent and results in a reduction in the DP of the pulp cellulose even in the absence of CDH. The addition of CDH alone to the pulp (line 3) resulted in little change in the DP
TABLE 1. Average DPN and DPW values of CDH-treated Douglas fir kraft pulp Avg (SD)
C C1F1H C1H C 1 CDH (0.1 IU/ml, 18 h) F 1 H 1 CDH (0.1 IU/ml, 18 h) C 1 H 1 CDH (0.1 IU/ml, 18 h) C 1 F 1 CDH (0.1 IU/ml, 1 h) C 1 F 1 CDH (0.1 IU/ml, 2 h) C 1 F 1 CDH (0.1 IU/ml, 6 h) C 1 F 1 CDH (0.1 IU/ml, 12 h) C 1 F 1 CDH (0.1 IU/ml, 18 h) C 1 F 1 CDH (0.2 IU/ml, 18 h) C 1 F 1 H 1 CDH (0.1 IU/ml, 18 h)
341 (4) 301 (4) 334 (6) 295 (5) 293 (4) 290 (5) 326 (3) 307 (4) 293 (3) 282 (6) 262 (9) 193 (8) 299 (12)
2,337 (11) 2,093 (4) 2,279 (7) 2,069 (8) 2,047 (6) 1,987 (7) 2,307 (5) 2,243 (8) 2,190 (10) 2,086 (6) 1,832 (15) 1,588 (21) 2,201 (14)
a Reaction conditions included combinations of CDH (0.1 or 0.2 IU/ml), 20 mM cellobiose (C), 1.8 mM H2O2 (H), and 0.2 mM FeCl3 (F).
FIG. 2. Progressive changes in the DP of CDH-treated Douglas fir kraft pulp over time. Pulps (10 mg/ml) were incubated at 30°C in 50 mM sodium acetate (pH 4.5) supplemented with 0.1 and 0.2 IU of CDH per ml, 20 mM cellobiose, and 0.2 mM FeCl3.
(DPN 5 334; DPW 5 2,279). However, with the addition of excess cellobiose, the DP of the pulp was reduced substantially. It was apparent that the combination of CDH and cellobiose (line 4) could depolymerize the cellulose marginally more than the reaction mixture lacking CDH but capable of generating Fenton’s chemistry (line 2). The addition of CDH to the reaction mixture containing the individual components of Fenton’s reagent in the absence of cellobiose (line 5) caused a reduction in the DP of the cellulose even greater than that exhibited by the ferric chloride and hydrogen peroxide alone (line 2). This result confirmed that the CDH could modify the substrate without the addition of easily oxidizable substrates such as cellobiose (26). However, when the pulp samples were first subjected to chelation and then treated with CDH without the supplementation of iron, the ability of CDH to modify the DP of the pulp cellulose was substantially reduced (DPN 5 331; DPW 5 2,254). The greatest change in the DP of the cellulose occurred when the cellobiose and ferric chloride were added to the CDH. The mixture containing CDH, cellobiose, iron, and hydrogen peroxide demonstrated only slightly more depolymerization than the corresponding control lacking the enzyme. In all cases, the attack seems to be directed at the highermolecular-mass (DPs of 2,000 to 5,000) cellulose component of the pulp, as indicated by the absolute values of DPN and DPW for the different reaction scenarios (Table 1). These values clearly indicate that the depolymerization of pulp cellulose is substantially enhanced by the addition of CDH. Having established that the CDH supplemented with cellobiose and ferric chloride resulted in the greatest changes in cellulose morphology, the nature of the reaction was followed by monitoring the changes in DP through a time-controlled experiment (Fig. 2; Table 1). As the reaction proceeded, the enzyme complex continued to react with the cellulosic material of higher DP (5,000), generating molecules of a lower DP and producing a new maximal peak at approximately a DP of 1,000 to 2,000 (Fig. 2; Table 1). A subsequent treatment with twice as much CDH enzyme resulted in a further reduction in the average DP (DPN 5 193; DPW 5 1,558), as indicated by the shift in the shoulder of material found between DPs of 200 and 800 (Fig. 2; Table 1). The supplementation of additional cellobiose, by a 20 mM cellobiose spike at hour 9 of an 18-h reaction, had no effect on the DP (data not shown). High-performance anion-exchange chromatography was subsequently used to determine if any oligosaccharides were
VOL. 63, 1997
CELLULOSE DEPOLYMERIZATION BY CDH
TABLE 2. Polysaccharides liberated by CDH after 18 h of incubation Avg concn (mg/ml)b Reaction conditionsa
C C C C C C C
1 Fe 1F1H 1 CDH 1 H 1 CDH 1 F 1 CDH 1F1H1 CDH
0.022 (0.002) 0.023 (0.002) 0.114 (0.001) 0.157 (0.006) 0.225 (0.010) 0.171 (0.005) 0.175 (0.003)
0.000 0.000 0.006 (0.001) 0.008 (0.001) 0.013 (0.002) 0.008 (0.001) 0.010 (0.001)
6.799 (0.020) 6.811 (0.013) 6.772 (0.019) 6.046 (0.009) 6.188 (0.022) 5.252 (0.016) 6.389 (0.011)
0.000 0.000 0.000 1.285 (0.009) 0.893 (0.017) 1.999 (0.024) 0.463 (0.008)
a Reaction conditions included combinations of CDH (0.1 IU/ml), 20 mM cellobiose (C), 1.8 mM H2O2 (H), and 0.2 mM FeCl3 (F). b Values in parentheses indicate standard deviations.
liberated into the filtrates during the reactions (Table 2). For all of the controls, no sugars other than those containing added cellobiose were detected. However, in the reactions containing CDH and additional components, cellobionic acid was detected after 18 h of incubation. Cellobionic acid is generated during HPLC analysis by base hydrolysis of the cellobionolactone found within the actual reaction filtrates. Maximum amount of cellobionic acid (cellobionolactone) was generated when both cellobiose and ferric chloride accompanied the CDH (Table 2). Cellobionic acid (cellobionolactone) was also produced when only CDH and cellobiose were present, but not to the same extent as was observed when iron was also present. Supplementation with hydrogen peroxide demonstrated a reduced level of cellobionic acid (cellobionolactone) generation, which was even more pronounced when both hydrogen peroxide and iron were present. In conjunction with the generation of cellobionolactone from cellobiose by these different reaction scenarios, the generation of glucose and to a lesser extent arabinose was also observed (Table 2). No other neutral wood sugars (galactose, xylose, or mannose) were produced. Small amounts of glucose were also present in the reaction filtrates supplemented with only cellobiose, indicating a minor contamination of the cellobiose with glucose. When the optimal reaction mixture of CDH supplemented with only cellobiose and ferric chloride was used, it was apparent that as the reaction proceeded over an 18-h incubation period, both glucose and cellobionolactone were generated while the cellobiose was slowly degraded (Table 3). Similarly, using twice as much CDH resulted in an approximate proportional twofold increase in the amount of glucose and lactone generated while depleting the supplemented cellobiose. Both arabinose and glucose were generated as the reaction proceeded, and the addition of twice as much CDH also generated approximately twice as much of these two sugars (Table 3). To examine whether the action of CDH liberated any sugars from the pulp, the reaction filtrates were acid hydrolyzed and subsequently analyzed for the possible presence of wood sugars. Anion-exchange chromatography indicated that no neutral wood sugars were liberated. DISCUSSION To date, much of the work related to CDHs has focused on their interaction with ligninolytic enzymes and their action on lignin-related compounds (2, 36). It has been suggested that CDH may be one of the components of the fungal enzymatic machinery involved in the depolymerization-repolymerization
of lignin-related compounds by a variety of fungi (1). This is most likely directly related to the oxidoreductive nature of this enzyme. However, other workers (7) have also shown that CDH enhanced the cellulolytic degradation of crystalline cellulose. These authors indicated that the supplementation of Trichoderma cellulases with CDH increased the substrate hydrolysis by approximately 20% over the non-CDH-supplemented reaction. More recently it has been demonstrated that CDH can degrade or modify cellulose and xylan derivatives, as well as synthetic lignin (18). Previous studies using 1H nuclear magnetic resonance spectroscopy indicated that CDH does selectively convert b-D-cellobiose to its cellobiono-1,5-lactone derivative (21). Our present study using HPLC methods confirmed these findings, indicating the rapid production of cellobionolactone. Prolonged incubation times resulted in increased cellobionolactone generation, with the simultaneous generation of smaller amounts of glucose. Similarly, the addition of CDH and cellobiose alone to the pulp without either hydrogen peroxide or iron generated substantial amounts of the cellobionolactone, indicating that the iron content of the pulp (67 ppm) or other transition metals such as cobalt or copper (4 or 228 ppm, respectively) was high enough for the putative reaction to proceed. Furthermore, treating the pulp first with an EDTA chelation step effectively reduced the depolymerizing action of CDH. It is also possible that other reducible substrates such as quinones are present. In the situations when the reaction mixture was supplemented with hydrogen peroxide, reduced amounts of lactone were generated. These results are contrary to those found by Henriksson et al. (18), who reported increased degradation of both xylan and cellulose when CDH was incubated in the presence of cellobiose, iron, and hydrogen peroxide. However, our reaction conditions differed not in iron concentration but in source. These previous workers (18) used ferric cyanide, while we used ferric acetate. It has been shown that ferric acetate participates much more readily in Fenton’s chemistry than does ferric cyanide (46), ultimately generating hydroxyl radicals. These radicals have a high and indiscriminate reactivity and a very short half-life. It has been postulated that enzyme damage can occur when the Fenton’s reaction takes place in close proximity to the enzyme (46). This is likely the case when additional hydrogen peroxide was added to this reaction mixture, while in the reaction mixtures not supplemented with hydrogen peroxide, this compound can readily be generated by the action of CDH and the supplemented iron acetate (see reaction scheme in Results). Our results clearly indicate that CDH can depolymerize wood-derived cellulose. In vivo, an even more efficient depolymerization may be possible, as an accumulation of hydrogen peroxide may result from the autooxidization of iron complexes such as Fe(II)
TABLE 3. Polysaccharides liberated by CDH (0.1 or 0.2 IU/ml) supplemented with 20 mM cellobiose and 0.2 mM iron over time Avg concn (mg/ml)a Reaction conditions
1h 2h 6h 12 h 18 h 18 h (0.2 IU/ml)
0.028 (0.001) 0.041 (0.003) 0.091 (0.004) 0.143 (0.008) 0.173 (0.003) 0.273 (0.010)
0.000 0.000 0.002 (0.001) 0.005 (0.001) 0.009 (0.001) 0.018 (0.002)
6.454 (0.011) 6.194 (0.019) 6.001 (0.023) 5.655 (0.011) 5.191 (0.017) 3.108 (0.044)
0.401 (0.009) 0.532 (0.022) 1.105 (0.019) 1.663 (0.031) 2.081 (0.028) 3.995 (0.016)
Values in parentheses indicate standard deviations.
MANSFIELD ET AL.
oxalate (46) and/or the slow continuous generation of small amounts of hydrogen peroxide by extracellular oxidases (10). The attack on the natural woody substrate seems to be directed at the cellulose of higher DP, generating molecules of cellulosic material of smaller size. This is indicated by the appearance of the shoulders in the lower DP ranges of the chromatographs. Recently, Ander et al. (3) found, through the use of polarized light micrographs, that the attack by CDH on pine holocellulose fiber seemed to be directed at specific sites of compression. These “nodes,” whose origin are still under investigation, appear approximately every 100 to 200 mm. It would be interesting to see whether these nodes are enriched in iron or other transition metals, as concentrated deposits in these regions would support the localized attack previously observed (3). The appearance of arabinose in the acid hydrolysates of the reaction filtrates was rather surprising, as no xylose was detected. Since the majority of the arabinose found in pulp is directly associated with the xylan backbone, and no xylose was liberated, this observation suggested that the arabinose did not result from the selective liberation of a neutral wood sugar from the pulp but rather was a degradation product of the supplemented cellobiose. This assumption was confirmed when arabinose was detected in reaction filtrates which mimicked the pulp treatments but lacked pulp as a substrate. Maximal arabinose was detected in samples which had been supplemented with hydrogen peroxide, suggesting that Fenton’s chemistry was directly involved in arabinose generation. Previously it had been reported that the products of cellobiose degradation under Fenton’s chemistry include both glucose and arabinose as well as other organic acids (23). Thus, it is highly likely that the observed cellulose depolymerization was directly related to the action of hydroxyl radicals generated by this enzyme. The breakdown products of cellobiose were also observed when only cellobiose and CDH were added to pulp, suggesting that the concentration of transition metals within the pulp was enough to initiate Fenton’s-type reactions. However, this did not occur to the extent that was exhibited when iron was added to the reaction mixture and was reduced substantially when the pulps were first subjected to a chelation step. This observation was further supported by the fact that CDH and cellobiose alone could depolymerize cellulose to a greater extent than was observed with just the addition of iron and hydrogen peroxide. Most of the past work on CDH has primarily investigated its action on lignin or lignin-related compounds, implying that this enzyme may have a primary role in lignin degradation. However, the fact that this enzyme contains a cellulose binding domain (35), is produced during primary metabolism with cellulose as a substrate (33), and demonstrates the ability to degrade both native cellulose and cellulose and xylan derivatives suggests that this enzyme may actually be more closely associated with the cellulases than the ligninases, while it functions as a dehydrogenase. Recent work by Ander et al. (3) supports this suggestion, as these workers found that the attack by CDH on pine holocellulose was more pronounced when the reaction mixture was supplemented with exoglucanases compared to endoglucanases. As the natural iron content of the pulp (20 to 100 ppm) (27) seems to support enzyme activity and the generation of cellobiose occurs by the action of accompanying cellulases (exoglucanases), it is probable that all of the required cofactors will be present in the natural environment. Further evidence for its close association with the cellulases is that CDH has been found in nonligninolytic fungi such as Monilia sp. (12), Sporotrichium thermophile (8), and Chaetomium cellulolyticum (15) and in bacteria (29), which all
APPL. ENVIRON. MICROBIOL.
possess a complete cellulase system and no ligninolytic enzymes. It is also known that cellobionolactone, which is generated by CDH, can act as an inducer of cellulolytic enzymes while inhibiting b-glucosidases. By actively competing for the available cellobiose with the b-glucosidases, this may limit the actual amount of degradation that occurs naturally. It is worth noting that all of the white rot fungi and the single strain of brown rot fungus (Coniophora puteana) from which CDH has been purified to date contain a complete cellulase system. Other brown rot fungi, which are not members of the family Coniophoraceae, do not possess a complete cellulase system, as they lack exoglucanases, and these fungi have yet to demonstrate the existence of CDH as part of their enzymatic machinery. Therefore, the generation of cellobiose by exoglucanases and the natural iron content in woody material complete the requirements for active CDH enzyme, which could then act in concert with the cellulolytic enzymes to degrade the carbohydrate moieties while also causing modifications to the lignin. ACKNOWLEDGMENTS We thank NSERC, the Science Council of British Columbia, Canada, and Weyerhaeuser for a scholarship held by S. D. Mansfield. REFERENCES 1. Ander, P. 1994. The cellobiose-oxidizing enzymes CBQ and CbO as related to lignin and cellulose degradation—a review. FEMS Microbiol. Rev. 13: 297–312. 2. Ander, P. 1996. Effects of cellobiose dehydrogenase on guaiacyl and guaiacyl/ syringyl lignins in the presence of iron and hydrogen peroxide. Holzforschung 50:413–419. 3. Ander, P., G. Daniel, B. Pettersson, and U. Westermark. 1996. Possible applications of cellobiose oxidizing and other flavine adenine nucleotide enzymes in the pulp and paper industry. ACS Symp. Ser. 655:297–307. 4. Ander, P., and K.-E. Eriksson. 1977. Selective degradation of wood components by white-rot fungi. Physiol. Plant. 41:239–248. 5. Backa, S., J. Gierer, T. Reitberger, and T. Nilsson. 1993. Hydroxyl radical activity associated with the growth of white-rot fungi. Holzforschung 37:181– 187. 6. Bao, W., E. Lymar, and V. Renganathan. 1994. Optimization of cellobiose dehydrogenase and b-glucosidase production by cellulose-degrading cultures of Phanerochaete chrysosporium. Appl. Microbiol. Biotechnol. 42:642–646. 7. Bao, W., and V. Renganathan. 1992. Cellobiose oxidase of Phanerochaete chrysosporium enhances crystalline cellulose degradation by cellulases. FEBS Lett. 302:77–80. 8. Canevascini, G., P. Borer, and J.-L. Dreyer. 1991. Cellobiose dehydrogenase of Sporotrichum (Chrysosporium) thermophile. Eur. J. Biochem. 198:43–52. 9. 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. 10. de Jong, E., J. A. Field, and J. A. M. De Bont. 1994. Aryl alcohols in the physiology of ligninolytic fungi. FEMS Microbiol. Rev. 13:153–188. 11. de Jong, E., K. K. Y. Wong, and J. N. Saddler. 1997. The mechanism of xylanase prebleaching of kraft pulp: an examination using model pulps prepared by redepositing lignin and xylan on cellulose fibers. Holzforschung 51:19–26. 12. Dekker, R. F. H. 1980. Induction and characterization of a cellobiose dehydrogenase produced by a species of Monilia. J. Gen. Microbiol. 120:309–316. 13. Eggert, C., N. Habu, U. Temp, and K.-E. L. Eriksson. 1996. Cleavage of Phanerochaete chrysosporium cellobiose dehydrogenase (CDH) by three endogenous proteases, p. 551–554. In E. Srebotnik and K. Messner (ed.), Biotechnology in the pulp and paper industry. Recent advances in applied and fundamental research. Facultas-Universita¨tsverlag, Vienna, Austria. 14. Eriksson, K.-E. L., N. Habu, and M. Samejima. 1993. Recent advances in fungal cellobiose oxidoreductases. Enzyme Microb. Technol. 15:1002–1008. 15. 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. 16. Habu, N., M. Samejuma, J. Dean, and K.-E. Eriksson. 1993. Release of the FAD domain from cellobiose oxidase by proteases from cellulolytic cultures of Phanerochaete chrysosporium. FEBS Lett. 327:161–164. 17. Halliwell, G. 1965. Catalytic decomposition of cellulose under biological conditions. Biochem. J. 95:35–40. 18. Henriksson, G., P. Ander, B. Pettersson, and G. Pettersson. 1995. Cellobiose dehydrogenase (cellobiose oxidase) from Phanerochaete chrysosporium as a wood-degrading enzyme. Studies on cellulose, xylan and synthetic lignin.
VOL. 63, 1997 Appl. Microbiol. Biotechnol. 42:790–796. 19. Henriksson, G., G. Johansson, and G. Pettersson. 1993. Is cellobiose oxidase from Phanerochaete chrysosporium a one-electron reductase? Biochim. Biophys. Acta 1144:184–190. 20. Henriksson, G., G. Pettersson, G. Johansson, A. Ruiz, and E. Uzcategui. 1991. Cellobiose oxidase from Phanerochaete chrysosporium can be cleaved by papain into two domains. Eur. J. Biochem. 196:101–106. 21. Higham, C. W., D. Gordon-Smith, C. E. Dempsey, and P. M. Wood. 1994. Direct 1H-NMR evidence for conversion of b-D-cellobiose to cellobionolactone by cellobiose dehydrogenase from Phanerochaete chrysosporium. FEBS Lett. 351:128–132. 22. Hu ¨ttermann, A., and A. Noelle. 1982. Characterization and regulation of cellobiose dehydrogenase in Fomes annosus. Holzforschung 36:283–286. 23. Kane, R. W., and J. D. Timpa. 1992. High-performance liquid chromatography study of D-cellobiose degradation under Fenton conditions. J. Carbohydr. Chem. 11:779–797. 24. Koenigs, J. W. 1975. Hydrogen peroxide and iron: a microbial cellulolytic system. Biotechnol. Bioeng. Symp. 5:151–159. 25. Koenigs, J. W. 1972. Production of extracellular hydrogen peroxide and peroxidase by wood-rotting fungi. Phytopathology 62:103–110. 26. 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. 27. Lapierre, L., J. Bouchard, R. M. Berry, and B. van Lierop. 1995. Chelation prior to hydrogen peroxide bleaching of kraft pulps. J. Pulp Paper Sci. 21:J268–J273. 28. Lehner, D., P. Zipper, G. Henriksson, and G. Pettersson. 1996. Small-angle X-ray scattering studies on cellobiose dehydrogenase from Phanerochaete chrysosporium. Biochim. Biophys. Acta 1293:161–169. 29. Li, X. Z., Y. Z. Huang, D. G. Xu, D. P. Xiao, F. G. Jin, and P. J. Gao. 1996. Cellobiose-oxidizing enzyme from a newly isolated cellulolytic bacterium Cytophaga sp LX-7. Biotechnol. Lett. 18:205–210. 30. Mansfield, S. D., K. K. Y. Wong, E. de Jong, and J. N. Saddler. 1996. Modification of Douglas-fir mechanical and kraft pulps by enzyme treatment. Tappi J. 79:125–132. 31. Mansfield, S. D., K. K. Y. Wong, E. de Jong, and J. N. Saddler. 1996. Xylanase prebleaching of different fiber length fractions of Douglas-fir kraft pulp. Appl. Microbiol. Biotechnol. 46:319–326. 32. Morpeth, F. F. 1991. Cellobiose oxidoreductases, p. 337–348. In F. Mu ¨ller (ed.), Chemistry and biochemistry of flavoproteins, vol. 1. CRC Press, Boca Raton, Fla. 33. Morpeth, F. F. 1987. Intracellular oxygen-metabolizing enzymes of Phanero-
CELLULOSE DEPOLYMERIZATION BY CDH
chaete chrysosporium. J. Gen. Microbiol. 133:3521–3525. 34. Raices, M., E. Paifer, J. Cremata, R. Montesino, J. Ståhlberg, C. Divne, I. J. Szabo ´, G. Henriksson, G. Johansson, and G. Pettersson. 1995. Cloning and characterization of a cDNA encoding a cellobiose dehydrogenase from the white rot fungus Phanerochaete chrysosporium. FEBS Lett. 369:233–238. 35. Renganathan, V., S. N. Usha, and F. Lindenburg. 1990. Cellobiose-oxidizing enzymes from the lignocellulose-degrading basidiomycete Phanerochaete chrysosporium: interaction with microcrystalline cellulose. Appl. Microbiol. Biotechnol. 32:609–613. 36. Roy, B. P., T. Dumonceaux, A. A. Koukoulas, and F. S. Archibald. 1996. Purification and characterization of cellobiose dehydrogenases from the white rot fungus Trametes versicolor. Appl. Environ. Microbiol. 62:4417– 4427. 37. Schmidhalter, D. R., and G. Canevascini. 1992. Characterization of the cellulolytic enzyme system from the brown-rot fungus Coniophora puteana. Appl. Microbiol. Biotechnol. 37:431–436. 38. Schroeder, L. R., and F. C. Haigh. 1979. Cellulose and wood pulp polysaccharides. Gel permeation chromatography analysis. Tappi 62(10):103–105. 39. Valtasaari, L., and K. Saarela. 1975. Determination of chain length distribution of cellulose by gel permeation chromatography using the tricarbanilate derivative. Paperi ja Puu—Papper och Tra¨. 57(1):5–10. 40. Westermark, U., and K.-E. L. Eriksson. 1974. Carbohydrate-dependent enzymic quinone reduction during lignin degradation. Acta Chem. Scand. Ser. B 28:204–208. 41. Westermark, U., and K.-E. L. Eriksson. 1974. Cellobiose:quinone oxidoreductase, a new wood-degrading enzyme from white-rot fungi. Acta Chem. Scand. Ser. B 28:209–214. 42. Westermark, U., and K.-E. L. Eriksson. 1975. Purification and properties of cellobiose:quinone oxidoreductase from Sporotrichum pulverulentum. Acta Chem. Scand. Ser. B 29:419–424. 43. Wilson, M. T., N. Hogg, and G. D. Jones. 1990. Reactions of reduced cellobiose oxidase with oxygen. Is cellobiose oxidase primarily an oxidase? Biochem. J. 270:265–267. 44. Wood, B. F., A. H. Conner, and C. G. Hill, Jr. 1986. The effect of precipitation on the molecular weight distribution of cellulose tricarbanilate. J. Appl. Polym. Sci. 32:3703–3712. 45. Wood, J. D., and P. M. Wood. 1992. Evidence that cellobiose:quinone oxidoreductase from Phanerochaete chrysosporium is a breakdown product of cellobiose oxidase. Biochim. Biophys. Acta 1119:90–96. 46. Wood, P. M. 1994. Pathways of production of Fenton’s reagent by woodrotting fungi. FEMS Microbiol. Rev. 13:313–320.