Induction of Extracellular Hydroxyl Radical Production by White-Rot Fungi through Quinone Redox Cycling

Abstract

A simple strategy for the induction of extracellular hydroxyl radical (OH) production by white-rot fungi is presented. It involves the incubation of mycelium with quinones and Fe³⁺-EDTA. Succinctly, it is based on the establishment of a quinone redox cycle catalyzed by cell-bound dehydrogenase activities and the ligninolytic enzymes (laccase and peroxidases). The semiquinone intermediate produced by the ligninolytic enzymes drives OH production by a Fenton reaction (H₂O₂ + Fe²⁺ → OH + OH⁻ + Fe³⁺). H₂O₂ production, Fe³⁺ reduction, and OH generation were initially demonstrated with two Pleurotus eryngii mycelia (one producing laccase and versatile peroxidase and the other producing just laccase) and four quinones, 1,4-benzoquinone (BQ), 2-methoxy-1,4-benzoquinone (MBQ), 2,6-dimethoxy-1,4-benzoquinone (DBQ), and 2-methyl-1,4-naphthoquinone (menadione [MD]). In all cases, OH radicals were linearly produced, with the highest rate obtained with MD, followed by DBQ, MBQ, and BQ. These rates correlated with both H₂O₂ levels and Fe³⁺ reduction rates observed with the four quinones. Between the two P. eryngii mycelia used, the best results were obtained with the one producing only laccase, showing higher OH production rates with added purified enzyme. The strategy was then validated in Bjerkandera adusta, Phanerochaete chrysosporium, Phlebia radiata, Pycnoporus cinnabarinus, and Trametes versicolor, also showing good correlation between OH production rates and the kinds and levels of the ligninolytic enzymes expressed by these fungi. We propose this strategy as a useful tool to study the effects of OH radicals on lignin and organopollutant degradation, as well as to improve the bioremediation potential of white-rot fungi.

White-rot fungi are unique in their ability to degrade a wide variety of organopollutants (36, 47), mainly due to the secretion of a low-specificity enzyme system whose natural function is the degradation of lignin (11). Components of this system include laccase and/or one or two types of peroxidase, such as lignin peroxidase (LiP), manganese peroxidase (MnP), and versatile peroxidase (VP) (31). Besides acting directly, the ligninolytic enzymes can bring about lignin and pollutant degradation through the generation of low-molecular-weight extracellular oxidants, including (i) Mn³⁺, (ii) free radicals from some fungal metabolites and lignin depolymerization products (7, 22), and (iii) oxygen free radicals, mainly hydroxyl radicals (OH) and lipid peroxidation radicals (21). Although OH radicals are the strongest oxidants found in cultures of white-rot fungi (1), studies of their involvement in pollutant degradation are scarce. One of the reasons is that the mechanisms proposed for OH production still await in vivo validation.

Several potential sources of extracellular OH based on the Fenton reaction (H₂O₂ + Fe²⁺ → OH + OH⁻ + Fe³⁺) have been postulated for white-rot fungi. In one case, an extracellular fungal glycopeptide has been shown to reduce O₂ and Fe³⁺ to H₂O₂ and Fe²⁺ (45). Enzymatic sources include cellobiose dehydrogenase, LiP, and laccase. Among these, only cellobiose dehydrogenase is able to directly catalyze the formation of Fenton's reagent (33). The ligninolytic enzymes, however, act as an indirect source of OH through the generation of Fe³⁺ and O₂ reductants, such as formate (CO₂⁻) and semiquinone (Q⁻) radicals. The first time evidence was provided that a ligninolytic enzyme was involved in OH production, oxalate was used to generate CO₂⁻ in a LiP reaction mediated by veratryl alcohol (4). The proposed mechanism consisted of the following cascade of reactions: production of veratryl alcohol cation radical (Valc⁺) by LiP, oxidation of oxalate to CO₂⁻ by Valc⁺, reduction of O₂ to O₂⁻ by CO₂⁻, and a superoxide-driven Fenton reaction (Haber-Weiss reaction) in which Fe³⁺ was reduced by O₂⁻. The OH production mechanism assisted by Q⁻ was inferred from the oxidation of 2-methoxy-1,4-benzohydroquinone (MBQH₂) and 2,6-dimethoxy-1,4-benzohydroquinone (DBQH₂) by Pleurotus eryngii laccase in the presence of Fe³⁺-EDTA. The ability of Q⁻ radicals to reduce both Fe³⁺ to Fe²⁺ and O₂ to O₂⁻, which dismutated to H₂O₂, was demonstrated (14). In this case, OH radicals were generated by a semiquinone-driven Fenton reaction, as Q⁻ radicals were the main agents accomplishing Fe³⁺ reduction. The first evidence of the likelihood of this OH production mechanism being operative in vivo had been obtained from incubations of P. eryngii with 2-methyl-1,4-naphthoquinone (menadione [MD]) and Fe³⁺-EDTA (15). Extracellular OH radicals were produced on a constant basis through quinone redox cycling, consisting of the reduction of MD by a cell-bound quinone reductase (QR) system, followed by the extracellular oxidation of the resulting hydroquinone (MDH₂) to its semiquinone radical (MD⁻). The production of extracellular O₂⁻ and H₂O₂ by P. eryngii via redox cycling involving laccase was subsequently confirmed using 1,4-benzoquinone (BQ), 2-methyl-1,4-benzoquinone, and 2,3,5,6-tetramethyl-1,4-benzoquinone (duroquinone), in addition to MD (16). However, the demonstration of OH production based on the redox cycling of quinones other than MD was still required.

In the present paper, we describe the induction of extracellular OH production by P. eryngii upon its incubation with BQ, 2-methoxy-1,4-benzoquinone (MBQ), 2,6-dimethoxy-1,4-benzoquinone (DBQ), and MD in the presence of Fe³⁺-EDTA. The three benzoquinones were selected because they are oxidation products of p-hydroxyphenyl, guaiacyl, and syringyl units of lignin (MD was included as a positive control). Along with laccase, the involvement of P. eryngii VP in the production of O₂⁻ and H₂O₂ from hydroquinone oxidation has also been reported (13). Since hydroquinones are substrates of all known ligninolytic enzymes, quinone redox cycling catalysis could involve any of them. Here, we demonstrate OH production by P. eryngii under two different culture conditions, leading to the production of laccase or laccase and VP. We also show that quinone redox cycling is widespread among white-rot fungi by using a series of well-studied species that produce different combinations of ligninolytic enzymes.

MATERIALS AND METHODS

Chemicals and enzymes.

H₂O₂ (Perhydrol 30%) was obtained from Merck. 2-Deoxyribose, 1,10-phenanthroline, 2-thiobarbituric acid (TBA), and bovine liver catalase (EC 1.11.1.6) were purchased from Sigma. 2,6-Dimethoxyphenol (DMP), 4-hydroxybenzoic acid, 3,4-dihydroxybenzoic (protocatechuic) acid, BQ, DBQ, MD, 1,4-benzohydroquinone (BQH₂), and MBQH₂ were from Aldrich. DBQH₂ was prepared from DBQ by reduction with sodium borohydride (2). MBQ was synthesized by oxidation of MBQH₂ with silver oxide (19). All other chemicals used were of analytical grade.

Laccase isoenzyme I (EC 1.10.3.2) and VP isoenzyme I from P. eryngii were produced and purified as previously described by Muñoz et al. (34) and Martínez et al. (32), respectively.

Organisms and culture conditions.

P. eryngii IJFM A169 (Fungal Culture Collection of the Centro de Investigaciones Biológicas) (= ATCC 90787 and CBS 613.91), Trametes versicolor IJFM A136, Phlebia radiata IJFM A588 (= CBS 184.83), Bjerkandera adusta IJFM A581 (= CBS 595.78), and Pycnoporus cinnabarinus IJFM A720 (= CECT 20448; Colección Española de Cultivos Tipo) were maintained at 4°C on 2% malt extract agar. Mycelial pellets were produced at 28°C in shaken (150 rpm) 250-ml conical flasks with 100 ml of a glucose-peptone (GP) medium containing 20 g glucose, 5 g peptone, 2 g yeast extract, 1 g KH₂PO₄, and 0.5 g MgSO₄·7 H₂O per liter (25). P. eryngii was also cultivated in the presence of 50 μM MnSO₄ (GPMn medium). Inocula were prepared by homogenizing 10-day-old mycelium. The dry weight of the inocula was 0.1 g per 100 ml of medium.

Enzyme activities.

Laccase activity was assayed in 100 mM sodium phosphate buffer, pH 5, using 10 mM DMP as a substrate and measuring the production of coerulignone (extinction coefficient at 469 nm [ɛ₄₆₉] = 27,500 M⁻¹ cm⁻¹, when referring to the DMP concentration) (32). VP and MnP activities were estimated by Mn³⁺-tartrate complex formation (ɛ₂₃₈ = 6,500 M⁻¹ cm⁻¹) in reaction mixtures containing 100 mM sodium tartrate buffer, pH 5, 100 μM MnSO₄, and 100 μM H₂O₂ (32). LiP activity was assayed in 100 mM tartrate buffer, pH 3, as the oxidation of veratryl alcohol (2 mM) to veratraldehyde (ɛ₃₁₀ = 9,300 M⁻¹ cm⁻¹) in the presence of 400 μM H₂O₂ (46). These enzymatic assays were performed at room temperature (22 to 25°C).

Mycelium washed with distilled water was used for the determination of cell-bound laccase, VP, and QR activities in P. eryngii. Laccase and VP activities were estimated as described above. To minimize underestimates of QR activity, BQ was selected, since laccase activity on BQH₂ has been shown to be quite low (34). QR activity was determined in 50 mM phosphate buffer, pH 5, using 500 μM BQ as a substrate and measuring the production of BQH₂ by high-performance liquid chromatography (HPLC). Samples (20 μl) were injected into a Pharmacia system equipped with a Spherisorb S50DS2 column (Hichrom) and a diode array detector. The analyses were carried out at 40°C with a flow rate of 1 ml min⁻¹ and 10 mM phosphoric acid-methanol (80/20) as an eluent. The UV detector operated at 280 nm, and BQH₂ levels were estimated using a standard calibration curve. For these cell-bound analyses of enzymatic activities, appropriate amounts of mycelium were incubated at room temperature with 20-ml substrate solutions in shaken 100-ml conical flasks (150 rpm). Samples were taken at 1-min intervals for 5 min. The mycelium was separated from the liquid by filtration. Absorbance was determined immediately after filtration for laccase and VP activity measurements. The pH of samples used for QR activity determination was lowered to 2 with phosphoric acid, and they were kept frozen at −20°C until they were analyzed. International units of enzyme activity (μmol min⁻¹) were used.

Incubation of fungi with quinones.

Ten-day-old mycelial pellets were collected from cultures by filtration, washed three times with distilled water, and resuspended in 50 ml of 20 mM phosphate buffer, pH 5, containing 500 μM BQ, MBQ, DBQ, or MD. The amount of mycelium used in these incubations was 202 ± 14 mg (dry weight). For Fe³⁺ reduction experiments, the complex 100 μM Fe³⁺-110 μM EDTA and 1.5 mM 1,10-phenanthroline were added to the quinone incubation solution. Iron salt (FeCl₃) solutions were made up fresh immediately before use. In OH production experiments, 1,10-phenanthroline was replaced by either 2.8 mM 2-deoxyribose or 1 mM 4-hydroxybenzoic acid, depending on the method used to estimate OH generation (see below). Incubations were performed in 50-ml volumes at 28°C and 150 rpm in 100 ml conical flasks. Samples were periodically removed from three replicate flasks, and the extracellular fluid was separated from the mycelium by filtration. In order to inactivate the ligninolytic enzymes that could be released to the extracellular solution during the experiments, samples were treated in different ways depending on the kind of analysis to be performed. For the analysis of quinone, hydroquinone, protocatechuic acid, and TBA-reactive substances (TBARS), the pH of samples was lowered to 2 with phosphoric acid. For H₂O₂ estimation, samples were heated at 80°C for 20 min (a treatment that does not affect H₂O₂ levels). Other analyses were carried out immediately after the samples were removed.

Analytical techniques.

The Somogyi-Nelson method for the determination of reducing sugars was used to estimate glucose concentrations in fungal cultures (41).

Levels of MBQ, DBQ, and their corresponding hydroquinones were determined by HPLC, using standard calibration curves for each compound (17). Samples (20 μl) were injected into a Pharmacia system equipped with a Spherisorb S50DS2 column (Hichrom). The analyses were carried out at 40°C with a flow rate of 1 ml min⁻¹ using 10 mM phosphoric acid-methanol (80/20) as the eluent. The UV detector operated at 254 and 280 nm.

H₂O₂ levels were estimated by measuring the production of O₂ with a Clark-type electrode after the addition of 100 U of catalase per ml (heat-denatured catalase was used in blanks) (17). The amount of H₂O₂ was calculated taking into account the stoichiometry of the catalase reaction (2 H₂O₂:1 O₂). The oxygen electrode was calibrated by the same procedure with known amounts of H₂O₂ from the standard commercial solution, which were calculated spectrophotometrically (ɛ₂₃₀ = 81 M⁻¹ cm⁻¹).

The production of ferrous ion was measured at 510 nm by the formation of its chelate with 1,10-phenanthroline (ɛ = 12,110 M⁻¹ cm⁻¹) (4).

Both TBARS production from 2-deoxyribose (20) and conversion of 4-hydroxybenzoic acid into protocatechuic acid (12) were used as procedures to estimate OH production. TBARS were determined as follows. A total of 0.5 ml of 2.8% (wt/vol) trichloroacetic acid and 0.5 ml of 1% (wt/vol) TBA in 50 mM NaOH was added to 1-ml samples and heated for 10 min at 100°C. After the mixture cooled, the absorbance was read at 532 nm against appropriate blanks (18). Protocatechuic acid production was estimated by HPLC as described above for quinones, except a linear 10 mM phosphoric acid/methanol gradient from 0 to 30% methanol in 20 min was used as the mobile phase. The absorbance of the eluate was monitored at 254 nm (14).

Statistical analysis.

All the results included in the text and shown in the figures are the means and standard deviations of three replicates (full biological experiments and technical analyses). The least-squares method was used for regression analysis of data.

RESULTS

Following our previous studies of quinone redox cycling (15, 16), incubations of P. eryngii with quinones were carried out in buffered solutions (pH 5) and 10-day-old washed pellets. The fungus was grown in GP medium with or without Mn²⁺ in order to obtain pellets producing, respectively, only laccase (Lac-mycelium) or laccase plus VP (LacVP-mycelium) (32). In both cases, maximum growth was reached after 8 to 12 days, coinciding with glucose depletion (10 days). The mycelial dry weights in 10-day-old cultures carried out in GP and GPMn media were 808 ± 54 mg (average of three replicate flasks ± standard deviation). The levels of extracellular and cell-bound laccase, VP, and QR activities in 10-day-old cultures are listed in Table 1. The presence of Mn²⁺ in the culture medium repressed the production of VP (38), increased laccase activity, and had no significant effect on cell-bound QR activity.

TABLE 1.

Extracellular and cell-bound laccase, VP, and quinone reductase activities present in 10 day-old cultures of P. eryngii grown in GP and GPMn media

Medium	Enzyme	Activity
		Culture liquid (mU ml−1)	Cell bound (mU mg−1)
GP	Lac	24 ± 2	8 ± 1
	VP	330 ± 13	4 ± 1
	QR	NDa	30 ± 3
GPMn	Lac	57 ± 5	11 ± 2
	VP	0	0
	QR	ND	28 ± 3

^aND, not determined.

As previously reported for BQ and MD (16), MBQ and DBQ were redox cycled when incubated with washed Lac-mycelium (Fig. 1). Reduction of MBQ and DBQ to their corresponding QH₂s was observed during the first 20 min. Then, QH₂/Q molar ratios remained constant until the end of the experiment: 6.8 and 2.6 (average of 30- to 90-min samples) for MBQ and DBQ, respectively. These ratios were the result of equilibrium between QR and laccase activities as confirmed in parallel experiments carried out with added laccase. It was found that when the extracellular laccase activity was raised to 100 and 200 mU ml⁻¹, QH₂/Q ratios decreased, respectively, to 2.1 and 0.6 for MBQ and to 0.7 and 0.3 for DBQ. The lower ratios observed in incubations with DBQ agreed with the higher efficiency of laccase oxidizing DBQH₂ than MBQH₂ (17). These results indicated that oxidation of hydroquinones in the absence of added laccase was the rate-limiting step of MBQ and DBQ redox cycles, and therefore, that the enzyme addition increased the rates of these cycles.

FIG. 1.

Time course of MBQ (A) and DBQ (B) reduction by P. eryngii in the presence of different amounts of laccase (filled symbols, MBQ and DBQ; open symbols, MBQH₂ and DBQH₂). Incubations were carried out with 10-day-old Lac-mycelium and 500 μM quinones in 50 ml 20 mM phosphate buffer, pH 5, in the absence and presence of 100 and 200 mU ml⁻¹ laccase. The error bars indicate standard deviations.

The production of H₂O₂ and reduction of Fe³⁺ by P. eryngii upon incubation with BQ, MBQ, DBQ, and MD was evaluated as a requisite for OH generation. Figure 2A and B shows the results for H₂O₂ production by LacVP-mycelium and Lac-mycelium, respectively. Regardless of the mycelium used, basal levels of extracellular H₂O₂ observed in 60- to 180-min samples remained around 2 to 3 μM with no quinone added. However, incubations carried out with quinones gradually increased H₂O₂ levels until steady values were attained (after 30 min). The average levels of 60- to 180-min samples increased in the order BQ < MBQ < DBQ < MD and were lower with LacVP-mycelium (6, 23, 61, and 179 μM, respectively) than with Lac-mycelium (8, 72, 192, and 308 μM, respectively). Figure 2 also shows that when P. eryngii mycelia were incubated with 500 μM H₂O₂, it nearly disappeared from the extracellular liquid in about 60 min. After this time, H₂O₂ levels were similar to those found without quinones (2 to 3 μM). Although this decrease in H₂O₂ levels was observed with both mycelia, indicating that they have one or more systems consuming it, the higher consumption rate observed with LacVP-mycelium suggested the involvement of VP. From these findings, it could be inferred that the steady H₂O₂ levels found in incubations with quinones were the result of equilibrium between the rates of production and consumption, and therefore, that quinone redox cycling was operative during the whole period studied.

FIG. 2.

Effects of quinones on the production of H₂O₂ by LacVP-mycelium (A) and Lac-mycelium (B) from P. eryngii. Mycelia were incubated in 50 ml 20 mM phosphate buffer, pH 5, in the absence (No Q) and presence of 500 μM BQ, MBQ, DBQ, or MD. Incubations with 500 μM H₂O₂ were also carried out in the absence of quinones. The error bars indicate standard deviations.

On the other hand, the effect of quinones on Fe³⁺ reduction by washed P. eryngii mycelia is shown in Fig. 3. Fe²⁺-phenanthroline complex from Fe³⁺-EDTA was linearly produced at different rates depending on the mycelium and quinone used. In the absence of quinones, basal iron reduction rates calculated from 0- to 60-min samples were 0.1 and 0.4 μM min⁻¹ with LacVP- and Lac-mycelium, respectively. This observation supports the existence of a cell-bound mechanism for iron reduction which did not imply quinone redox cycling. Incubations with BQ raised these rates to 0.4 μM min⁻¹ (LacVP-mycelium) and 0.9 μM min⁻¹ (Lac-mycelium). A much greater increase was obtained in incubations with MBQ, DBQ, and MD, although no significant differences were found among them: 4.0 μM min⁻¹ with LacVP-mycelium and 5.0 μM min⁻¹ with Lac-mycelium, as calculated from the linear increase observed in 0- to 12-min samples. The Fe³⁺ was completely reduced and scavenged by phenanthroline with these three quinones in about 25 min.

FIG. 3.

Effects of quinones on the reduction of chelated Fe³⁺ by LacVP-mycelium (A) and Lac-mycelium (B) from P. eryngii. Mycelia were incubated in 50 ml 20 mM phosphate buffer, pH 5, in the absence (No Q) and presence of 500 μM BQ, MBQ, DBQ, or MD, with 100 μM Fe³⁺-110 μM EDTA and 1.5 mM 1,10-phenanthroline. The error bars indicate standard deviations.

Following the demonstration of the formation of Fenton's reagent, production of OH was studied in incubations of washed mycelia with quinones and Fe³⁺-EDTA. Depending on the procedure used to detect OH, the incubation mixtures also contained either 2-deoxyribose or 4-hydroxybenzoic acid for the production of TBARS or protocatechuic acid, respectively. TBARS production was linear during the whole period studied with the four quinones and the two mycelia used (Fig. 4). The TBARS production rates were always higher with Lac-mycelium than with LacVP-mycelium. In both cases, the highest rate was obtained with MD, followed by DBQ, MBQ, and BQ. There existed, therefore, a positive correlation between TBARS levels and those of H₂O₂ shown in Fig. 2. Furthermore, taking into account that Fe²⁺ was produced at a rate similar to those of MBQ, DBQ, and MD (Fig. 3), it could be inferred that H₂O₂ was limiting the Fenton reaction during at least MBQ and DBQ redox cycling. TBARS production was also observed in the absence of quinones and the presence of Fe³⁺-EDTA, but at the lowest rates. This observation is in close agreement with the results in Fig. 3, showing the ability of P. eryngii to reduce Fe³⁺ with no Q. In addition to H₂O₂ basal levels generated by the fungus under these conditions (Fig. 2), OH production could be supported by that derived from ferrous iron autoxidation, followed by the dismutation of the resulting O₂⁻.

FIG. 4.

Hydroxyl radical production by LacVP-mycelium (A) and Lac-mycelium (B) from P. eryngii via quinone redox cycling, estimated as TBARS formation from 2-deoxyribose. The incubation mixtures were as described in the legend to Fig. 3, except that 1,10-phenanthroline was replaced by 2.8 mM 2-deoxyribose. The error bars indicate standard deviations.

In order to confirm OH production, the effect of OH scavengers, such as mannitol and dimethyl sulfoxide (DMSO), on TBARS formation was evaluated. Lac-mycelium was incubated with DBQ, Fe³⁺-EDTA, and 2-deoxyribose in the absence and presence of 5 mM mannitol or DMSO. The TBARS production rate during the first 120 min decreased from 14.8 mU A₅₃₂ min⁻¹ (incubation without scavengers) to 7.4 and 2.8 mU A₅₃₂ min⁻¹ with mannitol and DMSO, respectively. The production of OH was also confirmed by estimating the hydroxylation of 4-hydroxybenzoic acid during the redox cycling of the four quinones by Lac-mycelium. As shown in Fig. 5, protocatechuic acid was also produced on a constant basis throughout the experiment. Regression analyses of data obtained from 30- to 180-min samples produced the following rates of protocatechuic acid production: 64, 102, 199, 297, and 423 nM min⁻¹ in incubations with no Q, BQ, MBQ, DBQ, and MD, respectively. The results in Fig. 1 revealed that oxidation of MBQH₂ and DBQH₂ was the rate-limiting step of the MBQ and DBQ redox cycles. Based on these results, the effects of increasing extracellular laccase and VP activities on OH production were evaluated. Figure 6 shows TBARS production during the incubation of Lac-mycelium with DBQ, Fe³⁺-EDTA, 2-deoxyribose, and different amounts of laccase and VP. The addition of 50 and 100 mU ml⁻¹ laccase enhanced the TBARS production rate from 17 mU A₅₃₂ min⁻¹ (blank without added enzymes) to 40 and 50 mU A₅₃₂ min⁻¹, respectively, as calculated from the results obtained for 0- to 180-min samples. The addition of VP also caused an increase in the TBARS production rate, although much lower and inversely correlated with the amount of the enzyme: 23 and 20 mU A₅₃₂ min⁻¹ with 50 and 100 mU ml⁻¹ of added VP, respectively.

FIG. 5.

Hydroxyl radical production by Lac-mycelium from P. eryngii via quinone redox cycling, estimated as 4-hydroxybenzoic acid hydroxylation. The incubation mixtures were as described in the legend to Fig. 3, except 1,10-phenanthroline was replaced by 1 mM 4-hydroxybenzoic acid. The error bars indicate standard deviations.

FIG. 6.

Effects of laccase and VP on hydroxyl radical (TBARS) production by Lac-mycelium from P. eryngii via DBQ redox cycling. The incubations were carried out in 50 ml 20 mM phosphate buffer, pH 5, containing Lac-mycelium, 500 μM DBQ, 100 μM Fe³⁺-110 μM EDTA, and 2.8 mM 2-deoxyribose, in the absence (0 mU) and presence of 50 and 100 mU ml⁻¹ laccase (solid lines) and VP (dashed lines). The error bars indicate standard deviations.

A final series of experiments was carried out in order to test the validity of this inducible OH production mechanism in other white-rot fungi: B. adusta, P. chrysosporium, P. radiata, P. cinnabarinus, and T. versicolor. Quinone redox cycling experiments were carried out with DBQ and 10-day-old washed mycelia grown in GP medium. The ligninolytic enzymes produced by the fungi under these culture conditions were LiP and a Mn²⁺-oxidizing peroxidase (VP) by B. adusta (30); laccase by P. cinnabarinus, T. versicolor, and P. radiata; and none by P. chrysosporium (Table 2). These different enzyme production patterns, including several levels of laccase, allowed us to study the effect of ligninolytic enzymes on OH production from both qualitative and quantitative points of view. Incubations of the fungi with DBQ were carried out in the presence of Fe³⁺-EDTA and 2-deoxyribose, and samples were analyzed for TBARS production, as well as for DBQ and DBQH₂ levels (Fig. 7). Among the five fungi, P. chrysosporium produced TBARS at the lowest rate (2 mU A₅₃₂ min⁻¹ during a period from 0 to 120 min). Although the fungus showed a good ability to reduce DBQ, it is quite likely that the absence of ligninolytic enzymes limited the production of TBARS. The highest TBARS production rate, estimated during the same period, was obtained with P. radiata, followed by T. versicolor and P. cinnabarinus, i.e., 29, 21, and 10 mU A₅₃₂ min⁻¹, respectively (Fig. 7C). These results were positively correlated with laccase activity levels (Table 2), in agreement with the positive effect caused by the enzyme on TBARS production by P. eryngii (Fig. 6). Furthermore, the differences in laccase levels were also reflected in the DBQH₂/DBQ ratios found at system equilibrium (Fig. 7A and B), similar to the results obtained with P. eryngii (Fig. 1). Thus, whereas DBQH₂ was the prevalent compound in 60- to 180-min samples from the fungus producing the lowest laccase activity (P. cinnabarinus), DBQ was the prevalent compound with the best laccase producers (P. radiata and T. versicolor). A lower TBARS production rate was observed with B. adusta (8 mU A₅₃₂ min⁻¹), producing only ligninolytic peroxidases. However, it was four times higher than that observed with P. chrysosporium, indicating the involvement of these enzymes in DBQ redox cycling.

TABLE 2.

Extracellular laccase, MnP/VP, and LiP activities present in 10-day-old cultures of selected white-rot fungi grown in GP medium

Species	Enzyme activity (mU ml−1)
	Laccase	MnP/VP	LiP
B. adusta	0	38 ± 3	192 ± 15
P. chrysosporium	0	0	0
P. radiata	319 ± 22	0	0
P. cinnabarinus	19 ± 1	0	0
T. versicolor	217 ±14	0	0

FIG. 7.

Hydroxyl radical production by selected white-rot fungi via DBQ redox cycling. Levels of DBQ (A), DBQH₂ (B), and TBARS (C) are shown. The incubation mixtures contained 50 ml 20 mM phosphate buffer, pH 5; 200 ± 12 mg 10-day-old mycelia from B. adusta, P. chrysosporium, P. radiata, P. cinnabarinus, and T. versicolor; 500 μM DBQ; 100 μM Fe³⁺-110 μM EDTA; and 2.8 mM 2-deoxyribose. TBARS levels were corrected from those found in incubation blanks without DBQ. The error bars indicate standard deviations.

DISCUSSION

Our results provide in vivo validation of the Q⁻-assisted mechanism of reactive oxygen species (ROS) production previously demonstrated with purified laccases from P. eryngii (14, 17, 34) and Coriolopsis rigida (39), VP from P. eryngii, and MnP from Phanerochaete chrysosporium (13). They also support our earlier finding that quinone redox cycling is a mechanism for the production of extracellular O₂⁻ and H₂O₂ in P. eryngii (16). This mechanism can now be expanded to OH production, a greater number of quinones, and some widely studied white-rot fungi. The production of OH radicals has been inferred from the generation of Fe²⁺ and H₂O₂ (Fenton's reagent), as well as from the conversion of 2-deoxyribose into TBARS and the hydroxylation of 4-hydroxybenzoic acid, with the four parameters highly correlated. A definitive proof confirming OH identity would be required, such as electron spin resonance analysis. Overall, these results enable us to propose the scheme depicted in Fig. 8 as a model for extracellular OH production by white-rot fungi via quinone redox cycling. Although the scheme shows the main reactions involved in ROS production under the incubation conditions used in the present study, some steps of the process may present several alternatives (described below). As can be seen, two redox cycles are shown, one for the three benzoquinones (Fig. 8A) and the other for the naphthoquinone (Fig. 8B). Common reactions involved in the generation of Fenton's reagent in both cycles (reactions 4a, 5a to c, and 6) are shown only in Fig. 8A in order to avoid a crowded scheme. Figure 8B includes a distinctive reaction of MD redox cycling, i.e., propagation of MDH₂ oxidation by O₂⁻ (reaction 4b), which, as discussed below, was probably the factor contributing most to the production of the highest levels of H₂O₂ and OH radicals with this quinone (Fig. 2 and 4, respectively). With the exception of the enzymatic oxidation of QH₂, a similar quinone redox cycling process has been described in brown-rot fungi. These fungi produce several methoxyhydroquinones that can be chemically oxidized by Fe³⁺ (24). The mechanisms by which these fungi produce extracellular OH have been widely studied, since these radicals are considered the main agents causing the rapid cellulose depolymerization characteristic of brown rot (3). Thus, the involvement of QH₂ in ROS production by brown-rot fungi has been evidenced not only under defined liquid incubation conditions, but also on cellulose and wood (8, 44). Quinone redox cycling in other biological systems has been extensively studied and described as an intracellular process consisting of the enzymatic one-electron reduction of Q, followed by the autoxidation of the resulting Q⁻ (23). However, quinone redox cycling in white- and brown-rot fungi presents the distinctive characteristic of producing ROS in the extracellular environment due to the two-electron reduction of Q and the secretion of the resulting QH₂.

FIG. 8.

Scheme of the quinone redox cycling process in P. eryngii (see Discussion for an explanation). (A) Main reactions involved in ROS production through BQ, MBQ, and DBQ redox cycling in the absence and presence of Fe³⁺-EDTA (solid and dashed arrows, respectively). (B) MD redox cycling, showing hydroquinone propagation by O₂⁻. Reversible reactions are indicated by double arrows.

Quinone redox cycling in white-rot fungi is triggered by the actions of cell-bound enzymatic systems that reduce them to QH₂ (Fig. 8, reaction 1). Studies carried out mainly with P. chrysosporium have described the existence of two different systems reducing quinones, i.e., intracellular dehydrogenases using NAD(P)H as electron donors [Q + NAD(P)H + H⁺ → QH₂ + NAD(P)⁺] (6, 10) and a plasma membrane redox system (43). Two similar intracellular quinone reductases have been characterized in the brown-rot fungus Gloeophyllum trabeum (9). Our current research on the contributions of these systems to the reduction of quinones by P. eryngii has revealed that the plasma membrane redox system is not involved (data not shown). Following Q reduction, extracellular QH₂ oxidation presents several possibilities. Being phenolic compounds, QH₂s are susceptible to oxidation by any of the ligninolytic enzymes (Fig. 8, reaction 2). As inferred from the results shown in Fig. 1, 6, and 7, it is clear that oxidation of QH₂ by laccase (4 QH₂ + O₂ → 4 Q⁻ + 2 H₂O + 4 H⁺) and peroxidases (2 QH₂ + H₂O₂ → 2 Q⁻ + H₂O) increases the rates of quinone redox cycles and ROS production, provided Q reduction is the limiting reaction of the cycle. Chemical transformation of QH₂ into Q⁻ is also a possibility to consider, since other oxidants than enzymes are either present or produced in the course of redox cycling. First, QH₂ autoxidation (QH₂ + O₂ → Q⁻ + O₂⁻ + 2 H⁺), which is a spin-restricted reaction, has been documented to be catalyzed by transition metal ions (49). For instance, Fe³⁺, chelated or not with oxalate, has been shown to be the catalyst of QH₂ oxidation in the redox cycling process described in brown-rot fungi (24, 44, 48). In the present study, special emphasis was placed on the involvement of the ligninolytic enzymes in quinone redox cycling. This is the reason why an iron complex preventing QH₂ oxidation, such as Fe³⁺-EDTA (14), was used. The effect of replacing EDTA by oxalate on OH production by P. eryngii via DBQ redox cycling has already been determined and will be described in a separate publication from this study. Second, conversion of QH₂ into Q⁻ can be accomplished by comproportionation reaction (QH₂ + Q ⇆ 2 Q⁻ + 2 H⁺). This reaction has been demonstrated to trigger the “autoxidation” of several benzo- and naphthoquinones and is especially noticeable with QH₂ substituted with strong electron-donating groups, including DBQH₂ (37). The contribution of this reaction to the production of ROS could be evidenced in the absence of other enzymatic or chemical QH₂ oxidants. For instance, the production of OH by P. chrysosporium during the DBQ redox cycle without producing any ligninolytic enzyme (Fig. 7) could be explained on the basis of this comproportionation reaction taking place. Third, the O₂⁻ derived from Q⁻ autoxidation can propagate the oxidation of some QH₂: QH₂ + O₂⁻ → Q⁻ + H₂O₂ (Fig. 8, reaction 4b). The existence of this reaction with the QH₂ derived from the reduction of the four Qs studied here was previously determined by studying the effect of superoxide dismutase on either the rate of QH₂ “autoxidation” or oxidation by laccase (16, 17). A negative effect indicative of suppression of QH₂ propagation by O₂⁻ was observed only in the case of MDH₂, in agreement with other reported studies (35).

In the absence of Fe³⁺-EDTA, the pronounced increase exerted by quinones on the production of extracellular H₂O₂ by P. eryngii (Fig. 2) evidenced the transformation of Q⁻ into Q via autoxidation (Q⁻ + O₂ ⇆ Q + O₂⁻ [Fig. 8, reaction 3]), followed by O₂⁻ dismutation (O₂⁻ + HO₂ + H⁺ → O₂ + H₂O₂ [Fig. 8, reaction 4a]) and, in the case of MD, also by O₂⁻ reduction by MDH₂ (reaction 4b). As Q⁻ autoxidation is a reversible reaction, the continuous removal of both quinones by fungal reducing systems (Fig. 1) and of O₂⁻ by reactions 4a and b were among the factors contributing to the efficiency of quinone redox cycling as an H₂O₂ production mechanism. This was previously demonstrated in laccase reactions with MBQH₂ and DBQH₂ by the recycling of MBQ and DBQ with diaphorase (a reductase catalyzing the divalent reduction of quinones) (17). In this way, transformation of Q⁻ into Q via autoxidation was favored over two competing reactions (not shown in Fig. 8): Q⁻ dismutation (2 Q⁻ + 2 H⁺ ⇆ QH₂ + Q) and Q⁻ laccase oxidation (4 Q⁻ + O₂ + 4 H⁺ → 4 Q + 2 H₂O). With respect to the removal of O₂⁻ via spontaneous dismutation under the incubation conditions used here in redox cycling experiments (pH 5), it is quite likely that this was occurring around its optimum pH, as it implies the oxidation of O₂⁻ by the hydroperoxyl radical (HO₂) in equilibrium with O₂⁻, and the pK value of HO₂ is 4.8 (5). In the presence of Fe³⁺-EDTA, the high enhancement of the rates of Fe³⁺ reduction (Fig. 3), TBARS production (Fig. 4), and 4-hydroxybenzoic acid hydroxylation (Fig. 5) caused by Q evidenced the production of OH by P. eryngii via Fenton's reagent formation. Reactions involved in this process are shown in Fig. 8. Under these conditions, Q⁻ autoxidation is mainly catalyzed by Fe³⁺ (sum of reactions 5a, Q⁻ + Fe³⁺ → Q + Fe²⁺, and 5b, Fe²⁺ + O₂ ⇆ Fe³⁺ + O₂⁻), leading to the production of OH radicals by a Q⁻-driven Fenton reaction (reactions 5a and 6) (14). Two other pathways that must be considered to contribute to OH generation, but to a lesser extent, are the O₂⁻-driven Fenton reaction (reactions reverse 5b and 6), using the O₂⁻ produced by direct Q⁻ autoxidation (reaction 3), and that caused by the reduction of Fe³⁺ by the unknown cell-bound system whose existence was inferred above from the results shown in Fig. 3 to 5 with no Q (reactions 5c and 6). Possible cell-bound systems catalyzing Fe³⁺ reduction are the plasma membrane redox system described in P. chrysosporium (42) and the plasma membrane reductase (Fre1 protein) involved in iron uptake by several fungi (26). In this regard, a possible role in controlling iron levels available for Fenton chemistry has been suggested for two P. chrysosporium ferroxidase activities, one of which is extracellular (28) and the other a component of the cell-bound iron uptake complex Fet3-Ftr1 (27).

One interesting characteristic of quinone redox cycling is the low substrate specificity of the enzymes participating in the process. The results obtained in the present study with three of the most representative quinones produced during the degradation of both softwood and hardwood lignins enhances the likelihood of extracellular ROS production by white-rot fungi through this mechanism. Quinones have been also shown to be common intermediates during the degradation of aromatic pollutants by these fungi (36). In this regard, it is likely that some of these quinone intermediates can act as redox cycling agents contributing to ROS production and thus to the degradation of the pollutant from which they derive. This could be the case with MD being an oxidation product of 2-methylnaphthalene. The high correlation observed between the levels of H₂O₂ (Fig. 2) and OH radicals (Fig. 4 and 5) allows us to use the same rationales to explain the differences found in both cases with the different quinones. It is quite likely that O₂⁻ reduction by MDH₂ (Fig. 8, reaction 4b) was the main factor contributing to the production of the highest levels of H₂O₂ and OH during MD redox cycling for two reasons. First, it doubles the amount of H₂O₂ produced by O₂⁻ dismutation, which is the only reaction converting O₂⁻ into H₂O₂ with the three benzoquinones. Second, it increases the rate of QH₂ oxidation, which is the limiting reaction of MBQ and DBQ redox cycles, as inferred from the results shown in Fig. 1, and the BQ redox cycle, as previously reported by Guillén et al. (16). In fact, the latter study showed that the concerted action of laccase and O₂⁻ increased MDH₂ oxidation in such a way that MD reduction became the limiting reaction of its redox cycle. The differences observed in the extents of H₂O₂ and OH production during the redox cycles of the three benzoquinones (DBQ > MBQ > BQ) can be explained by considering the ability of Q⁻ radicals to be autoxidized. This ability, estimated previously as the production of H₂O₂ during the oxidation of DBQH₂, MBQH₂ and BQH₂ by purified laccase of P. eryngii, was shown to be DBQ⁻ > MBQ⁻ > BQ⁻ (16, 17). Furthermore, the higher the laccase and VP activities on QH₂, which have been reported to be DBQH₂ > MBQH₂ > BQH₂ (13, 17), the higher the rate of the quinone redox cycle.

The induction of TBARS production by DBQ and Fe³⁺-EDTA in B. adusta, P. chrysosporium, P. radiata, P. cinnabarinus, and T. versicolor (Fig. 7) led us to conclude that quinone redox cycling is a widespread mechanism for ROS production among white-rot fungi, including O₂⁻, H₂O₂, and OH radicals. This process can be operative in the absence of ligninolytic enzymes provided other QH₂ oxidants are present, as reported with brown-rot fungi (24, 48) and shown here in the case of P. chrysosporium. However, the significance of QH₂ enzymatic catalysis has been clearly evidenced by the following observations: the increase exerted by added laccase and VP on TBARS production by P. eryngii (Fig. 6) and the much higher TBARS production rates obtained with fungi expressing ligninolytic enzymes under the culture conditions used (Table 2 and Fig. 7). From the results shown in Fig. 4 and 7, it can be inferred that the enzyme playing a crucial role in terms of OH production was laccase. First, TBARS production by P. eryngii was found to be higher with Lac-mycelium than with LacVP-mycelium (Fig. 4). Second, TBARS production with the fungi expressing only laccase (P. radiata, P. cinnabarinus, and T. versicolor) was found to be higher than with the fungus expressing only peroxidases (B. adusta) (Fig. 7). This is not surprising, because oxidation of QH₂ by peroxidases implies the consumption of part of the H₂O₂ required for OH generation. The participation of laccase in the production of highly reactive oxidants deserves more attention, since it could explain the solubilization and mineralization of lignin observed with some ascomycetes producing only this ligninolytic enzyme (29, 40).

In summary, the results shown in the present study provide new information on the mechanisms used by white-rot fungi to activate O₂ in the extracellular environment. In addition, they supply the basis for a simple strategy that could be used in fundamental and practical studies, such as the determination of factors affecting OH production by quinone redox cycling, as well as the roles these radicals can play in lignin and organopollutant degradation by these fungi. In this regard, our current research in this field, carried out with P. eryngii and T. versicolor, is showing that the induction of extracellular OH radical production by quinones and iron occurs not only in incubations with washed mycelium, but also with whole fungal cultures during primary and secondary metabolism. This research is also revealing that the capability of these fungi to degrade pollutants is increased when quinone redox cycling is used as a strategy to induce the production of extracellular OH radicals (unpublished data).