Biotech Advs Vol. 2, pp 183-199, 1984

BIOCHEMISTRY OF THE OXIDATION OF LIGNIN BY PHANEROCHAETE CHRYSOSPORIUM

T. KENT KIRK, MING TIEN* and BRENDLYN D. FAISON†

Forest Products Laboratory, USDA Forest Service, P.O Box 5130, Madison, Wisconsin 53705, USA

ABSTRACT

The objective of this research was to identify the biochemical agents responsible for the oxilative degradation of lignin by the white-rot fungus Phanerochaete chrysosporium. We examined the hypothesis that activated oxygen species are involved, and we also sought the agent in ligninolytic cultures responsible for a specific oxidative degradative reaction in substructure model compounds. Results of studies of the production of activated oxygen species by cultures, of the effect of their removal on ligninolytic activity, and of their action on substructure model compounds support a role for hydrogen peroxide (H2O2) and possibly superoxide (O2 · ) in lignin degradation. Involvement of hydroxyl radical (·OH) or singlet oxygen (1O2) is not supported by our data. The actual biochemical. agent responsible for one important oxidative C-Cbond cleavage reaction in non-phenolic lignin substructure model compounds, and in lignin itself, was found to be an enzyme. The enzyme is extracellular, has a molecular weight of 42,000 daltons, is azide-sensitive, and requires H2O2 for activity.

* On assignment from North Carolina State University Department of Wood and Paper Science, Raleigh, NC 27607, USA.

† Department of Bacteriology, University of Wisconsin, Madison 53706, USA.

183

184 T.K. KIRK et al.

INTRODUCTION

Potential applications of the microbial systems that degrade lignin are of substantial import to the pulp and paper industry. Possibilities include pulping wood, bleaching pulps, modifying fiber surfaces of mechanical pulps, cleaning up wastewaters, and converting byproduct lignins to useful chemicals. The recognition of these potentials within the last decade has greatly accelerated research on lignin biodegradation. Even though the details of the biodegradative process are not clearly understood, scientists have explored certain applications: bio-mechanical pulping, which has been studied especially by Eriksson and coworkers at the Swedish Forest Products Research Laboratory (9); and wastewater cleanup, which has been investigated in some detail by our laboratory in collaboration with North Carolina State University (3). These studies employed intact living fungi. Application of the isolated ligninolytic system has not been explored, simply because basic research has neither defined the system nor disclosed how to separate it from the living cells. Applications such as bleaching chemical pulps and modifying the surface of fibers might be more readily accomplished with the isolated system.

The purpose here is to review our recent progress toward characterizing and isolating the biochemical ligninolytic system. The research has been limited to fungi which cause the white-rot type of wood decay and play a major role in lignin degradation. These fungi probably have the most potential for practial exploitation. In particular, we have studied Phanerochaete chrysosporium Burds., which has several characteristics (20)--including fast growth and a high temperature optimum--that make it especially suitable for study and application. Our presentation here deals primarily with experimental results and conclusions. In so far as possible, original publications and papers in press have been referenced; they can be consulted for details of methodology.

BACKGROUND

The biochemical system that degrades the lignin polymer is 'oxidative, relatively nonspecific, and extracellular (17,28). That the ligninolytic system is extracellular is apparent in the fact that it degrades a

Polymeric substrate. That the major reactions are oxidative has been

OXIDATION OF LIGNIN BY P. Chrysosporium 185

determined by characterizations of fungal degradation products of lignin, both polymeric (5,6,19) and low molecular weight (4). That the fungal attack is nonspecific is apparent from a consideration of the heterogeneous nature of lignin, including the fact that the several types of intermonomer linkages lack stereoregularity (1); from the fact that industrially modified lignins are still readily degraded by the fungi (26); and from the absence of stereospecificity in fungal metabolism of a lignin model compound(29).

It has been speculated recently that "diffusible" activated oxygen species (14) or "chemical agents" (36), rather than oxidative enzymes, are directly responsible for the degradation of lignin. This suggestion is attractive for two reasons: (a) Extracellular nonspecific oxidizing enzymes which catalyze the kinds of reactions comprising lignin degradation have not been described for any other system; i.e., there is no. precedent for the existence of such enzymes; and (b) nonenzymatic oxidations have been implicated in a number of other biochemical systems (30).

Our approach to elucidating the biochemical mechanism of lignin degradation has been to determine the nature of the fungal system that catalyzes each individual reaction of lignin breakdown. Although straightforward, this approach involved considerable background work including learning how to achieve ligninolytic activity in cultures and identifying individual degradative reactions in model compounds in the cultures. Several years of research have now taught us how to grow and manipulate the selected fungus P. chrysosporium so that it reproducibly develops maximal ligninolytic activity (17), defined as the oxidation of synthetic 14C-lignins to 14CO2 (23). By studying the pathways of the degradation of lignin substructure model compounds in such cultures, we have now defined several individual degradative reactions (18). Gold and coworkers (e.g. 7,8) and Higuchi and coworkers (15) have also described the degradative pathways for model compounds in the optimized cultures of P. chrysosporium. This background work has enabled us to evaluate the ability of biochemical preparations from ligninolytic cultures, and of artificially generated active oxygen species, to mimic the degradation of the model compounds seen in intact cultures. In addition, we have examined the optimized cultures for production of activated oxygen species. As reported here, our results do not support an involvement of free (nonenzyme-bound) highly reactive activated oxygen species. Instead, we have found that an extracellular

186 T.K. KIRK et al.

oxidative enzyme is responsible for at least one important degradative reaction.

PRODUCTION OF ACTIVATED OXYGEN SPECIES

BY LIGNINOLYTIC CULTURES

OXIDATION OF LIGNIN BY P. chrysosporium 187

Using whole cultures of P. chrysosporium, we have just completed (10) a study which had the following objectives: (a) to determine the kinetics of the production of activated oxygen species, (b) to describe the effect of high dioxygen (O2) concentration on biosynthesis of activated oxygen species, and (c) to assess the effect of removal of activated oxygen species on ligninolytic activity in cultures. The work focused on hydrogen peroxide (H2O2), superoxide (O2 · ), hydroxyl radical (·OH) and singlet oxygen (1O2), all of which have been assigned roles in other biological systems.

188 T.K. KIRK et al.

H2O2. Our assay for this species was based on the H2O2-dependent oxidation of o-dianisidine by horseradish peroxidase (10,32). We found H2O2 to be produced by ligninolytic cultures, as reported recently also by Forney et al. (12,13). The kinetics of its production paralleled the development of ligninolytic activity (Fig. 1), which was also reported by the other workers (12). Production of H2O2 was markedly stimulated by growing the cultures under O2 instead of air (Fig. 1), which also greatly stimulates the synthesis of the ligninolytic system in this fungus (17). Removal of H2O2 from ligninolytic cultures strongly suppressed ligninolytic activity (Table 1). Importantly, our work demonstrates for the first time a role for extracellular H2O2 in lignin degradation.

We were unable to develop a satisfactory quantitative assay for in the intact cultures (10). However, its removal, like removal

of H2O2, suppressed ligninolytic activity (Table 1).

OXIDATION OF LIGNIN BY P. chrysosporium 189

·OH. The kinetics of production of ·OH, which was assessed by the decarboxylation of benzoate (31), did not parallel ligninolytic activity, and an oxygen atmosphere did not stimulate its production during ligninolysis (Fig. 2). The scavengers of ·OH, however, inhibited ligninolytic activity (Table 1).

1O2. This species was assaved by the bleaching of the UV chromophore of AES, a water-soluble derivative of anthracene (2,28). The kinetics of its production and the effect of O2 paralleled ligninolytic activity (Fig. 3). The anthracene derivative strongly suppressed ligninolytic activity (Table 1).

190 T.K. KIRK et al.

Interpretation. We have concluded (10) that these results support a role , in lignin degradation. Other work has

already shown that chemically generated 1O2 does not (22) cause one of the reactions of lignin degradation which we had previously (28) ascribed to it (see next section). It is not clear whether 1O2 is actually produced by ligninolytic cultures; i.e. it is possible that the AES assay is not specific. It is clear, however, that AES interferes with ligninolysis (10,28). In any event, the kinds of reactions usually attributed to 1O2 (35) are probably not of primary importance in lignin degradation. A role for ·OH is not supported by the kinetic data or by the effect of O2 on the titer of · OH; the ·OH scavengers, although inhibiting lignin degradation, cannot be considered to be specific for ·OH. This interpretation is at odds with recent conclusions of Forney et al. (12) and of Kutsuki and Gold (24) who used other assays for ·OH in cultures or culture extracts of P. chrysosporium.

for H2O2, and possibly O2

REACTIONS OF ACTIVATED OXYGEN SPECIES

Under physiological conditions, H2O2 and O2 per se are not reactive enough to be responsible for reactions of lignin degradation. Both, however, are precursors of much more reactive species. The

has been proposed to generate low levels of 1O2 (16). and nonenzymatic dismutation of O2

In the presence of transition metals, O2 H2O2 can be converted to more reactive species, vis. ·OH and transition metal-oxygencomplexes (11). We have compared the effects of Fenton's reagent (Fe2+ + H2O2), which generates ·OH (34) and probably iron-oxvgencomplexes (see (11,34)), and of a mixture of HOCl and H2O2, which generates 1O2 (27), with the effects of ligninolytic cultures on substrucutre model compounds. The β-1 and β-O-4 model compounds (I and VII) that were used, and their fates in ligninolytic cultures of P. chrysosporium, are shown in Fig. 4.

Fenton's reagent. Products formed from β-1 model compound I by the Fenton reagent are given in Table 2. Products II, III, and IV, which are formed also in cultures (Fig. 4A), were detected, but were not the major products as they are in cultures (29). [It should he noted that the cultures rapidly reduce aldehyde IV, which is the initial product, to alcohol V,

which is the major product detected (Fig. 4A, (21))]. Major products of

OXIDATION OF LIGNIN BY P. chrysosporium 191

the action of the Fenton reagent on compound I were water-solubles(24%), and unidentified organic-extractable compounds, some of which were more polar than compound I. Phenolic materials (detected by color tests on thin layer chromatography plates) were prominent among the latter. We did not pursue the identities of these products. Interestingly, when the Fenton

reagent was made up with H2 18O2 (generated by reduction of 18O2 with glucose oxidase + glucose) and incubated with compound I, 18O was incorporated into the benzyl hydroxyl group of phenylglycol product II. Enrichment by 18O was approximately 70%, which is the same as observed whencompoundI is degraded in ligninolytic cultures under 18O2, (29). The incorporation of 18O from 18O2 into product II in cultures is illustrated in Fig. 4.

192 T.K. KIRK et al. Results of the action of Fenton's reagent on β-O-4 model compound VII (labeled uniformly with 14C in the A-ring (25)) are given in Table 3. The major initial products formed in cultures, compound VIII (rapidly reduced to IX), and compound X (Fig. 4B), were only minor products in the Fenton system. Water-solubles (not extracted into the organic solvent chloroform/acetone, 1:1 by vol.) and polar organic solubles were the major products. Phenolic materials comprised a major portion of the polar organic-solubles.

OXIDATION OF LIGNIN BY P. chrysosporium 193

These results suggest that the fungus uses an oxidizing species that may be similar to that generated by the Fenton reagent, but that the biochemical system exerts considerably more control over the reactions than does the Fenton reagent.

HOCl + H2O2. This system was examined only with β-1 model compound I. None of the products formed in cultures (Fig. 4A) were detected, although the model compound was degraded. As alreadymentioned, we have concluded that 1O2 is not responsible for the degradation of the model compound in the cultures (22).

THE ACTUAL BIOCHEMICAL SYSTEM

We have now used the Cα-Cβ cleavage reactions of compounds I and VII

shown in Fig. 4 to identify the responsible biochemical system in

194 T.K. KIRK et al.

ligninolytic cultures (33). We surmised that the reactive agent is extracellular and that H2O2 and/or dioxygen are required. Consequently, we examined the effects of the extracellular fluid on model compound I in the presence of O2 and various concentrations of added H2O2. At low concentrations of H2O2 (<0.2 mM), small amounts of aldehyde IV, formed by the cleavage of I (Fig. 4A) were detected in these reaction mixtures. There was no cleavage reaction in the absence of H2O2, or at concentrations greater that 0.5 mM H2O2. Subsequent work showed that the active agent in the culture fluid could be concentrated by ultrafiltration--i.e. that it is macromolecular. It was later found to have a molecular weight of 42,000 daltons (gel filtration, SDS electrophoresis). The cleavage reaction exhibits saturation kinetics. Thus the "agent" is an extracellular enzyme.

The concentrated extracellular fluid + H2O2 not only cleaves compound I, but it also cleaves the phenylglycol product (II) to form IV (+ an unidentified C1 fragment), and oxidizes the phenylglycol product to ketol V (Fig. 4A). Compound I is completely cleaved, despite the fact that is is a mixture of four stereoisomers.

The reconstituted system (concentrated culture fluid + H2O2) not only cleaves β-1 models, however; it also cleaves non-phenolic β-O-4 models. Compound VII, for example, is cleaved with formation of aldehyde product VIII and Cα-oxidation product X (Fig. 4B). [We have not yet identified the cleavage product from the Cβ part of the model, but we assume that an unstable hemiacetal 'is formed as illustrated in Fig. 4B.]

Because the two types of model compounds (I and VII) represent over 50% of the substructures in lignin (1), we considered it likely that enzymatic cleavage might be observable with lignin itself. To assess this possibility, we used spruce lignin, purified from an aqueous acetone extract of the wood of Picea engelmanii Parry, and milled wood lignin of birch (Betula verrucosa L.). Free phenolic hydroxyl groups were methylated with 14CH3I to facilitate detection and product identification. The highest molecular weight portions (~> 1,500 daltons) of the methylated lignins, fractionated with Sephadex LH-20, were employed. Incubation with the reconstituted system resulted in formation of vanillin 14C-methyl ether (IV) from both the spruce and birch lignins, and of syringaldehyde 14C-methyl ether (XI) as well from the latter. Based on the results with

OXIDATION OF LIGNIN BY P. chrysosporium 195 the model compounds, we concluded (33) that these two aldehydes (IV and XI) were formed by cleavages between Cα and Cβ in end groups as illustrated in Fig. 5. Aldehyde IV accounted for 4.5% of the original 14C in the spruce lignin, and aldehydes IV and XI contained 0.6 and 0.4% of the 14C from the birch lignin. These results are in accord with the known chemistry of spruce and birch lignins: The latter is a copolymer of guaiacyl (monomethoxyphenyl) and syringyl (dimethoxyphenyl) units, which gave rise to products IV and XI, whereas spruce lignin is comprised only of guaiacyl units (1).

196 T.K. KIRK et al.

The reconsituted system also partially depolymerized the lignins. As determined by LH-20 column chromatography, depolymerization products accounted for approximately 22% and 6% of the original 14C in the spruce

and birch lignins, respectively. Figure 6 illustrates the partial

OXIDATION OF LIGNIN BY P. chrysosporim 197 depolymerization of the latter. Cleavage of internal Cα-Cβ bonds probably contributed to the depolymerization.

Activity against the lignins, model I and model VII resides in a single enzyme, as indicated by polyacrylamide gel electrophoresis and gel permeation chromatography (33). Surprisingly, it is the major extracellular protein in the cultures, and indications are that it will be relatively easy to isolate for further characterization. Its sensitivity to azide (33) suggests that it is a metalloenzyme.

This work (33) has provided the first proof of a lignin-degrading enzyme. The reaction it catalyzes, Cα-Cβ cleavage (Fig. 4A,B), is known to be an important one in lignin degradation (4,5,15,17,19). Good evidence indicates, however, that other reactions are also important, including methoxyl demethylation, aromatic hydroxylation, and ring cleavage (15,17). Work underway is designed to describe the biochemical agents responsible for these other oxidative reactions, as well as to characterize the enzyme that has been discovered.

IMPLICATIONS FOR THE PULP AND INDUSTRY

Identification of one lignin polymer-degrading enzyme suggests that other ligninolytic reactions will also be found to be enzyme-catalyzed. Whether this is the case should soon be known, making it possible to evaluate the potentials of enzyme technology in such applications as biopulping, biobleaching, fiber modification, lignin bioconversion, and wastewater cleanup.

Identification of specific gene products will facilitate genetic approaches to improving ligninolytic activity in the fungi, and will also make it possible to move the encoding DNA from one organism to another via genetic engineeringtechniques.

Perhaps equally important is that understanding the enzyme-catalyzed reactions of lignin degradation could point to new chemical approaches to pulping, bleaching, and other processes dependent on lignin degradation. The enzyme described here catalyzes a reaction (Cα-Cβ cleavage), which,

if thorough, should give extensive depolymerization of lignin. BY

198 T.K. KIRK et al.

understanding the chemical mechanisms of its catalysis, we may be in a position to construct a biomimetic catalyst with considerable potential in the pulp and paper industry.

REFERENCES

OXIDATIONS OF LIGNIN BY P. chrysosporium 199

NOTE ADDED IN PROOF

Since this Conference, the ligninase has been purified and characterized [M. Tien and T.K. Kirk, Proc. Nat. Acad. Sci. USA 81, 2280 (1984)). It is a hemoprotein, containing about 15% carbohydrate. In the presence of H2O2 and O2, it catalyzes several seemingly disparate oxidations, including those illustrated in Fig. 4A and 4B, and it partially depolymerizes lignin. It does not function simply by generating free hydroxyl radical [T.K. Kirk, M.D. Mozuch and M. Tien, Biochem. J., in press], but it does seem to employ a radical mechanism (unpublished results). M.H. Gold and co-workers have also purified an enzyme from P. chrysosporium which is apparently the same [FEBS Lett. 169, 247 (1984)].