Electron transfer chain reaction of the extracellular flavocytochrome cellobiose dehydrogenase from the basidiomycete Phanerochaete chrysosporium

Authors


K. Igarashi, Department of Biomaterials Sciences, Graduate School of Agricultural and Life Sciences, The University of Tokyo, Bunkyo-ku, Tokyo 113-8657, Japan
Fax: +81 3 5841 5273
Tel: +81 3 5841 5258
E-mail: aquarius@mail.ecc.u-tokyo.ac.jp

Abstract

Cellobiose dehydrogenase (CDH) is an extracellular flavocytochrome containing flavin and b-type heme, and plays a key role in cellulose degradation by filamentous fungi. To investigate intermolecular electron transfer from CDH to cytochrome c, Phe166, which is located in the cytochrome domain and approaches one of propionates of heme, was mutated to Tyr, and the thermodynamic and kinetic properties of the mutant (F166Y) were compared with those of the wild-type (WT) enzyme. The mid-point potential of heme in F166Y was measured by cyclic voltammetry, and was estimated to be 25 mV lower than that of WT at pH 4.0. Although presteady-state reduction of flavin was not affected by the mutation, the rate of subsequent electron transfer from flavin to heme was halved in F166Y. When WT or F166Y was reduced with cellobiose and then mixed with cytochrome c, heme re-oxidation and cytochrome c reduction occurred synchronously, suggesting that the initial electron is transferred from reduced heme to cytochrome c. Moreover, in both enzymes the observed rate of the initial phase of cytochrome c reduction was concentration dependent, whereas the second phase of cytochrome c reduction was dependent on the rate of electron transfer from flavin to heme, but not on the cytochrome c concentration. In addition, the electron transfer rate from flavin to heme was identical to the steady-state reduction rate of cytochrome c in both WT and F166Y. These results clearly indicate that the first and second electrons of two-electron-reduced CDH are both transferred via heme, and that the redox reaction of CDH involves an electron-transfer chain mechanism in cytochrome c reduction.

Abbreviations
CDH

cellobiose dehydrogenase

F166Y

Phe166Tyr mutant CDH

NHE

normal hydrogen electrode

WT

wild-type CDH

Cellulose is the most abundant natural polymer on earth, and its degradation is thus an important component of the carbon cycle. Although cellulose is often referred to as a β-linked glucose polymer, cellobiose, a β-1,4-linked glucose dimer, should strictly be regarded as the repeating unit of cellulose, because adjacent glucoses show opposing faces to each other in the cellulose chain [1,2]. Many microorganisms recognize this repeating unit and hydrolyze cellulose to cellobiose as an initial step in the metabolism [3]. In filamentous fungi, cellulose degradation had been thought to proceed via two-step hydrolysis, i.e. cellulose is hydrolyzed to cellobiose by various cellulases and the product is further hydrolyzed to glucose by β-glucosidase. However, recent cytochemical, kinetic, and transcriptional studies [4–6] have supported another hypothesis concerning the contribution of cellobiose dehydrogenase (CDH; EC 1.1.99.18) to the extracellular cellulose metabolism of the white-rot fungus Phanerochaete chrysosporium, and the importance of a combination of hydrolytic and oxidative reactions in cellulose-degrading fungi as discussed previously [7–13].

CDH is the only extracellular flavocytochrome known to be secreted by filamentous fungi during cellulose degradation [11–13]. This enzyme carries flavin and a b-type heme in different domains, and the flavin domain catalyzes the dehydrogenation of cellobiose and cello-oligosaccharides to the corresponding δ-lactones [8,9,14,15]. Although this enzyme was initially characterized as an oxidase (cellobiose oxidase; EC 1.1.3.25) [8], its higher affinity for quinones and ferric compounds than for oxygen [16–18] and the low-spin character of the heme in both the ferric and ferrous states [19] indicate that the electron acceptor of this enzyme is not molecular oxygen. Although many candidates have been proposed for the electron acceptor of CDH, its natural electron acceptor and the physiological function of the oxidative half reaction of this enzyme are still uncertain.

In a previous kinetic study, we showed the presteady-state two-electron reduction of flavin, followed by interdomain one-electron transfer from flavin to heme, resulting in the formation of the flavin semiquinone radical and reduced heme [20]. In order to achieve full understanding of the redox reaction of CDH, the reduction mechanism of the electron acceptor should be clarified. However, several groups have proposed two possible mechanisms for this reaction; electron transfer chain and electron sink mechanisms [21–23]. In the putative electron transfer chain reaction, ferric compounds such as cytochrome c are reduced by heme after one-electron transfer from flavin, whereas the flavin radical or fully reduced flavin is an electron donor in the electron sink mechanism. In this study, Phe166, which approaches heme propionate, as shown in Fig. 1, was mutated to Tyr, and the slow interdomain electron transfer mutant F166Y was produced using the heterologous expression system of Pichia pastoris. Presteady-state reduction of cytochrome c and re-oxidation of heme in the recombinant wild-type (WT) and mutant enzymes were observed by sequential mixing in a three-syringe stopped-flow spectrophotometer to clarify the redox mechanism of CDH.

Figure 1.

Stereo representation of the heme in CDH and the amino acids within 4 Å of the heme. Heme and Phe166 are indicated in color. The model was generated from PDB 1D7C [35] using the software pymol (DeLano Scientific LLC, San Francisco, CA, USA).

Results

Redox properties of WT and F166Y CDH

The redox potentials of WT and F166Y were compared at various pH values, as shown in Fig. 2A. The potential of F166Y was lower than that of WT at all pH values, but the difference was larger at lower pH than at neutral pH. The potentials of WT and F166Y were estimated to be 176 and 151 mV vs. normal hydrogen electrode (NHE) at pH 4.0. Although the potential of F166Y is 25 mV lower than that of WT, it is still high enough to receive the electron from reduced flavin, because addition of cellobiose causes spectral changes in F166Y from the oxidized to the reduced form (Fig. 2B). Reduced F166Y has typical absorption maxima of CDH at 562, 533 and 428 nm due to α-, β- and γ-(Soret) bands, respectively, and the absorption decreased at around 480 nm mainly because of flavin reduction. These results are essentially the same as those for the WT enzyme [24] and suggest that both flavin and heme are active in the mutant enzyme.

Figure 2.

(A) pH dependence of the midpoint potential of heme in WT (□) and F166Y (•). (B) Absorption spectra of oxidized (solid line) and reduced (dotted line) forms of F166Y. Midpoint potentials of heme in CDH were measured by cyclic voltammetry as described in Experimental procedures. For the absorption spectra, 2.0 µm F166Y was scanned with or without 50 µm cellobiose in 50 mm sodium acetate buffer, pH 4.0.

Presteady-state and steady-state kinetics of F166Y

The presteady-state reduction of flavin and the subsequent electron transfer from flavin to heme in WT and F166Y were observed by stopped-flow spectrophotometry (Fig. 3). The addition of cellobiose caused rapid reduction of flavin, and the absorption at the isosbestic point of heme (449.0 nm) decreased similarly in both WT and F166Y (Fig. 3A). The observed rates (kobs) of WT and F166Y reduction were 53.0 ± 1.0 and 50.0 ± 1.3 s−1, respectively. In contrast, an apparent difference in heme reduction was observed between WT and F166Y, as shown in Fig. 3B. In WT CDH, ≈ 90% of heme was reduced within 0.1 s, whereas in F166Y, 13% of heme was still in the oxidized form at 0.2 s after mixing with the substrate. The kobs values for heme reduction of WT and F166Y were 30.2 ± 1.2 and 12.5 ± 0.9 s−1, respectively. These results suggest that only the electron transfer step from flavin to heme was halved in F166Y, whereas the initial flavin reduction was not affected by the mutation. Presteady-state kinetic experiments for flavin and heme reduction were carried out at various substrate concentrations, as shown in Fig. 4. The kobs values for flavin reduction were identical for WT and F166Y (Fig. 4A), whereas those of heme reduction in F166Y was almost half of that of WT at all substrate concentrations tested (Fig. 4B). The dissociation constants of WT (inline image) and F166Y (inline image) obtained from the plots were 107 ± 6.6 and 111 ± 6.5 µm, respectively. As shown in a previous presteady-state kinetic study of WT CDH, heme reduction was inhibited at high substrate concentrations. In this study, this phenomenon was also observed in F166Y, and the substrate inhibition constant of F166Y (inline image = 1130 ± 130 µm) was similar to that of WT (inline image = 1230 ± 180 µm), reflecting the similar Kd values of the two enzymes. The limiting rates of heme reduction for WT (inline image) and F166Y (inline image) were 46.9 ± 6.8 and 18.8 ± 1.0 s−1, respectively.

Figure 3.

Presteady-state reduction of WT and F166Y by cellobiose. (A) Time courses of absorption at 449 nm for monitoring flavin reduction. (B) Time courses of absorption at 562 nm for monitoring heme reduction. Solid line, F166Y; dashed line, WT. CDH and cellobiose (final concentrations of 5 and 100 µm, respectively) were mixed in 50 mm sodium acetate buffer (pH 4.0), and the absorption changes were monitored by stopped-flow photometry at 30 °C.

Figure 4.

Cellobiose concentration dependence of the observed rate (kobs) for flavin (A) and heme (B) reduction in WT (○) and F166Y (bsl00001). kobs values for both prosthetic groups were obtained under the same conditions as in Fig. 3, using 25–500 µm cellobiose as a substrate. The fitting of the data was performed as described in Experimental procedures.

The steady-state kinetic parameters are summarized in Table 1. As expected from the presteady-state experiment, there is no difference in these parameters between WT and F166Y using ubiquinone as an electron acceptor, whereas the mutation affected the kinetic parameters when the redox reaction was monitored in terms of cytochrome c reduction. The kcat values of cellobiose oxidation monitored in terms of cytochrome c reduction were quite similar to the klim values of heme reduction in both WT and F166Y. Interestingly, an increase in cytochrome c concentration inhibited its reduction only in the case of F166Y.

Table 1.  Steady-state kinetic constants for WT and F166Y. All measurements were carried out at 30 °C in 50 mm sodium acetate buffer, pH 4.0. Cellobiose oxidation was monitored by following the reduction of 1 mm ubiquinone or 50 µm cytochrome c as described in Experimental procedures. ND, no significant substrate inhibition was observed.
 Cellobiose oxidation with electron acceptorsUbiquinone reductionCytochrome c reduction
UbiquinoneCytochrome c
Km
m)
kcat
(s−1)
Km
m)
kcat
(s−1)
Ki
m)
Km
m)
kcat
(s−1)
Km
m)
kcat
(s−1)
Ki
m)
WT57.9 ± 7.140.1 ± 1.228.6 ± 1.643.5 ± 0.92510 ± 380326 ± 3544.2 ± 1.41.46 ± 0.1237.2 ± 0.3ND
F166Y56.0 ± 5.536.8 ± 0.917.9 ± 3.019.2 ± 0.93540 ± 310293 ± 1646.4 ± 0.80.67 ± 0.1119.0 ± 0.5233 ± 42

Sequential mixing experiment of WT and F166Y

To monitor the redox state of heme in CDH and cytochrome c independently in the same reaction mixture, the oxidized and reduced spectra of CDH and cytochrome c were compared, as shown in Fig. 5. In each hemoprotein, there are four isosbestic points in the 500–600 nm region, where the α- and β-bands of heme absorb. We selected 549.0 and 556.7 nm to monitor cytochrome c and heme in CDH, respectively, because these gave the maximum absorption difference.

Figure 5.

Absorption spectra of heme in CDH (A) and cytochrome c (B) for comparison of the isosbestic points. The spectra of the oxidized (solid line) and reduced (dashed line) forms are compared in the range of 450–650 nm. Dotted lines at 549.0 (left) and 556.7 (right) show the isosbestic points of heme in CDH and cytochrome c, respectively.

After the initial mixing of WT or F166Y with cellobiose, cytochrome c was added to the reaction mixture and the changes in absorption at 549.0 and 556.7 nm were monitored (Fig. 6). The reduction of cytochrome c and re-oxidation of heme in CDH were observed synchronously, and the kobs values for cytochrome c reduction by WT and F166Y were estimated as 662 ± 17 and 643 ± 10 s−1, respectively (Fig. 6A,C) i.e. almost the same value for the two enzymes. The kobs values for the secondary phase, however, were 27.7 ± 2.1 (WT) and 13.3 ± 4.1 (F166Y). These values are quite similar to those of heme reduction in both enzymes at the same cellobiose concentration. As shown in Fig. 6B,D, heme in WT and F166Y remained oxidized during this phase, but was re-reduced with reduction of cytochrome c. The kobs of cytochrome c reduction depended on the concentration of cytochrome c in the region tested (5–20 µm, data not shown), and was almost identical for WT and F166Y with limiting values of 1460 ± 140 and 1380 ± 80 s−1, respectively.

Figure 6.

Time courses of redox state of heme in WT (A, B) and F166Y (C, D) after sequential mixing with cellobiose and cytochrome c. Absorptions at 556.7 (solid line) and 549.0 (dashed line) nm were used for monitoring the heme in CDH and cytochrome c, respectively. WT or F166Y (5 µm) was mixed with 100 µm cellobiose, and 20 µm cytochrome c was then added and mixed after 0.1 s (WT) or 0.2 s (F166Y) using a sequential mixing stopped-flow apparatus. The absorption changes were monitored after mixing with cytochrome c, and the initial (0–0.020 s) and secondary (0–0.2 s) phase are seen in left (A, C) and right (B, D) panels, respectively. Conditions: 50 mm sodium acetate buffer (pH 4.0) at 30 °C.

Discussion

Two mechanisms, electron transfer chain and electron sink, have been proposed for the redox reaction of CDH, and several previous kinetic studies have attempted to clarify the overall reaction of this enzyme [21–23]. However, uncertainty remains, possibly because of the special features of this enzyme. The optimum pH values of flavin reduction by cellobiose (pH 4.5–5.0) and of electron transfer from flavin to heme (pH 3.5–4.0) differ from each other, and the rate-limiting step of the reaction thus depends on the pH of the reaction mixture [20,25]. Moreover, a higher concentration of substrate (cellobiose) inhibits presteady-state heme reduction, but not flavin reduction [20,26], suggesting that binding of substrate to the active site of the flavin domain inhibits electron transfer from flavin to heme. This phenomenon makes it difficult to solve the redox mechanism of this enzyme using kinetic results obtained with only WT CDH. In this study, therefore, we compared the thermodynamic and kinetic features of recombinant WT and the slow electron transfer mutant F166Y.

The presteady-state electron transfer from flavin to heme was halved in F166Y compared with WT. This was expected from the thermodynamic result that the redox potential of heme in F166Y was lower than that of WT. However, when the electron transfer rate k (s−1) and the driving force ΔG (eV) were analyzed in terms of electron transfer theory, which was recently developed by Dutton's group [27], the edge-to-edge distance R and the reorganization energy λ of F166Y were higher than those of WT when one of these parameters was fixed and used for the calculation (data not shown). This indicates that the halved electron transfer rate of F166Y is due not only to thermodynamic factors, but also involves kinetic changes in this mutant; for example, the change of Phe to Tyr may change the charge of the protein surface, resulting in a change in the interaction between the two domains. This would produce a higher R or λ value for F166Y compared with the WT enzyme. That only F166Y shows significant substrate (cytochrome c) inhibition might be because of the weaker domain interaction of this mutant enzyme. The F166Y mutant is, however, useful for monitoring the electron transfer step of CDH, because the mutation affects only the electron transfer step from flavin to heme, but not initial flavin reduction or cytochrome c reduction.

WT or F166Y was first mixed with cellobiose, and then the cellobiose–CDH mixture was mixed with cytochrome c after 0.1 s (WT) or 0.2 s (F166Y). At the second mixing time, ≈ 90% of heme is in the ferrous state and the flavin forms a semiquinone, as confirmed previously by EPR [20], suggesting that most of the enzyme is in the two-electron reduced form. In previous studies, prereduced CDH was often used to observe the electron transfer from heme to cytochrome c with cellobiose or ascorbate as an electron donor [22,23,28]. Without a sequential mixing technique, however, it is difficult to monitor cytochrome c reduction, because premixing with the electron donors produces one- (ascorbate) or three- (cellobiose) electron-reduced CDH, but not the two-electron-reduced form with flavin radical and reduced heme. Soon after the two-electron-reduced CDH was mixed with cytochrome c, synchronous cytochrome c reduction and re-oxidation of heme were clearly observed in WT and F166Y with similar biomolecular rate constants (6 × 106 m−1·s−1). This indicates that, at the initial phase, the electron is transferred from heme to cytochrome c. In contrast, the second phase of cytochrome c reduction was not concentration dependent, but depended on electron transfer from flavin to heme. Considering that the kcat value for cytochrome c reduction was almost identical to the klim value for heme reduction in both WT and F166Y, the two electrons in reduced CDH are sequentially transferred from reduced flavin to cytochrome c via heme. Cameron and Aust [22] recently demonstrated that the reduction rate of cytochrome c by fully (three-electron) reduced CDH was lower than the rate under steady-state turnover, and suggested that the flavin radical reduces cytochrome c. In the same report, moreover, they proposed that the heme in CDH acts as an electron sink, because Rogers and co-workers reported that heme is oxidized during steady-state cytochrome c reduction [28]. Although this phenomenon was also observed in the second phase of presteady-state cytochrome c reduction, it is because of the significant difference between the electron transfer rate from flavin to heme (30 s−1) and that from heme to cytochrome c (≈ 1500 s−1). After the electron is loaded from flavin to heme, it is transferred to cytochrome c without any significant time lag. Consequently, all the results obtained in this study apparently indicate that the overall redox reaction of CDH occurs through the electron transfer chain mechanism, as shown in Scheme 1.

Figure Scheme 1. .

Proposed catalytic cycle of CDH using cellobiose and cytochrome c as electron donor and electron acceptor, respectively. CB, cellobiose; CBL, cellobionolactone; F, flavin; H, heme; ox, oxidized form; red; reduced form; sq: semiquinone form.

Although this study clearly demonstrates an electron transfer chain mechanism of CDH when cytochrome c is used as an electron acceptor, it is too early to conclude that the mechanism is also used during cellulose degradation by the fungus because its natural electron acceptor is still uncertain. Considering the kinetic efficiency of cytochrome c for CDH is quite high compared with other ferric compounds, it is possible that the filamentous fungi produce a cytochrome c-like hemoprotein extracellularly and utilize it as an electron acceptor of CDH. Indeed, there are several hypothetical proteins with a secretion signal and a cytochrome c-binding motif encoded in the total genome sequence of P. chrysosporium[29]. To clarify the true role of CDH, therefore, it is important to consider the interaction with other extracellular redox proteins. As demonstrated in our kinetic studies, including this work, the redox reaction of CDH is regulated by cellobiose concentration and pH at the interdomain electron transfer step. This might provide a clue to identify the redox system of CDH in vivo.

Experimental procedures

Materials

d-Cellobiose was purchased from ICN Biomedicals (Irvine, CA, USA). Ubiquinone (2,3-dimethoxy-5-methyl-1,4-benzoquinone) and bovine heart cytochrome c were purchased from Wako Pure Chemical Industries (Osaka, Japan). To assess the pH dependence of the mid-point potential of heme, 50 mm buffers were used as described previously [20,30].

Steady-state enzyme assays

To obtain the steady-state kinetic parameters of cellobiose oxidation, the reduction rate of 1 mm ubiquinone or 50 µm cytochrome c was plotted against concentration of cellobiose (0–1 mm). Reduction rates for ubiquinone and cytochrome c were also measured at various concentrations (0–1 mm for ubiquinone and 0–50 µm for cytochrome c) using 500 µm cellobiose as a substrate. The reductions of ubiquinone and cytochrome c were monitored photometrically at 406 nm (Δε406 = 0.745 mm−1 cm−1) and 550 nm (Δε550 = 17.5 mm−1 cm−1), respectively. Because apparent substrate inhibition was observed when cytochrome c was used as an electron acceptor, the obtained substrate dependence plots were fitted to the Michaelis–Menten equation with a substrate inhibition constant (Ki). Unless otherwise noted, steady-state kinetic parameters (Km and kcat) were estimated by nonlinear fitting of the data to the Michaelis–Menten equation using deltagraph v. 5.5 (SPSS Inc., Cary, NC, USA and Redrock Software, Inc., Salt Lake City, UT, USA) and kaleidagraph v. 3.0.8 (Synergy Software, Reading, PA, USA).

Preparation of wild-type CDH and F166Y mutant

Recombinant wild-type P. chrysosporium CDH was heterologously expressed in the methylotropic yeast Pichia pastoris and purified as described previously [24]. Site-directed mutagenesis was carried out based on the overlap extension and nested primers methods, as described elsewhere [31–34]. Synthetic oligonucleotides, F166Y-F, 5′-CACACCGACTACGGCTTCTT-3′ (mutated nucleotides are underlined); AP1-EcoRI-F, 5′-TTTTCAGCGTTCTCGGAATTCCAGAGTGCCTCACAGTTTACCGAC-3′; AP1-F, 5′-TCAGCGTTCTCGGAATTC-3′; AP2-Xba-R, 5′-TTTTACAGTAATATAAAGAATTTCGCTCTAGATCAAGGACCTCCCGCAAGCGCGAG-3′; and AP2-R, 5′-TTACAGTAATATAAAGAATTTCGCTCTAGA-3′, were used to obtain the nucleotide fragment of F166Y (f166y). The fragment was subcloned into the pCR®4Blunt-TOPO vector (Invitrogen, Carlsbad, CA, USA) as described in the manufacturer's instructions, and the vector pCR4®Blunt-TOPO/f166y was digested with EcoRI and XbaI (TaKaRa Bio, Shiga, Japan) and ligated into the pPICZα-A vector (Invitrogen) at the same restriction sites. The vector pPICZα-A/f166y was then linearized with Bpu1102I (TaKaRa Bio) and transformed into Pichia pastoris KM-71H using a MicroPulser electroporation device (Bio-Rad Laboratories, Hercules, CA, USA). The Zeocin-resistant transformant was cultivated in a growth medium (1% yeast extract, 2% polypeptone, 1% glycerol; w/v) for 24 h at 30 °C, followed by the induction medium (1% yeast extract, 2% polypeptone, 1% methanol; w/v) for 48 h at 26.5 °C, and F166Y was purified from the culture filtrate with the same protocol as described previously [24]. The purity of wild-type CDH and F166Y was confirmed by SDS/PAGE and by the absorption spectrum.

Measurement of midpoint potential of heme in CDH

A direct electrochemical technique was used to measure the mid-point potential according to our previous report [30]. Glassy carbon, platinum, and Ag/AgCl were used as the working, counter, and reference (+205 mV vs. NHE at 25 °C) electrodes, respectively. Cyclic voltammetry was performed in the presence of 50 mm MgCl2 using an ALS Electrochemical Analyzer 624A, and the potential was determined by averaging the anodic and cathodic peak potentials.

Presteady-state kinetic studies

Presteady-state reduction of flavin and heme was monitored at the appropriate isosbestic point using an Applied Photophysics SX-18MV kinetic spectrophotometer and the observed rate (kobs) was estimated by fitting to the double exponential curve (flavin reduction) or single exponential curve with lag (heme reduction) according to our previous report [20]. Because apparent substrate inhibition was observed in heme reduction, the substrate inhibition constant (Ki) was used to estimate the limiting rates (klim) of heme reduction as described previously [20]. Rapid reduction of cytochrome c by reduced CDH and re-oxidation of heme in CDH were monitored with the same equipment, but using a sequential mixing mode as follows. Wild-type CDH and F166Y were first mixed with cellobiose, and the solution was then mixed with cytochrome c at 0.1 s (WT) or 0.2 s (F166Y) after the initial mixing, when almost 90% of heme was reduced. Final concentrations of WT, F166Y, cellobiose and cytochrome c were 5.0, 5.0, 100 and 5–20 µm, respectively. To monitor the reduction of heme in CDH and cytochrome c independently, absorption at the isosbestic points, 549.0 nm for cytochrome c (Δε549.0 = 16.8 mm−1·cm−1) and 556.7 nm for heme in CDH (Δε556.7 = 9.29 mm−1·cm−1), was monitored. All presteady-state measurements were carried out at least three times in 50 mm sodium acetate buffer pH 4.0 at 30 °C, and the data were analyzed as described previously [20].

Acknowledgements

This research was supported by Grants-in-Aid for Scientific Research to KI (No. 15780206), MS (No. 14360094), and TN (No. 16205021), and by Grant-in-Aid for Scientific Research on Priority Area to TN (No. 12147208) from the Ministry of Education, Culture, Sports, Science and Technology, and a Research Fellowship to MY (No. 08446) from the Japan Society for the Promotion of Science.

Ancillary