Chlorination and Cleavage of Lignin Structures by Fungal Chloroperoxidases

Two fungal chloroperoxidases (CPOs), the heme enzyme from Caldariomycesfumago and the vanadium enzyme from Curvularia inaequalis, chlorinated1-(4-ethoxy-3-methoxyphenyl)-2-(2-methoxyphenoxy)-1,3-dihydroxypropane,a dimeric model compound that represents the major nonphenolicstructure in lignin. Both enzymes also cleaved this dimer togive 1-chloro-4-ethoxy-3-methoxybenzene and 1,2-dichloro-4-ethoxy-5-methoxybenzene,and they depolymerized a synthetic guaiacyl lignin. Since fungalCPOs occur in soils and the fungi that produce them are commoninhabitants of plant debris, CPOs may have roles in the naturalproduction of high-molecular-weight chloroaromatics and in ligninbreakdown.

INTRODUCTION

Top

Introduction

Natural organochlorine compounds are widely distributed in theenvironment, but the processes that generate them are not completelyunderstood. It is clear that many natural organochlorines aremicrobial secondary metabolites, whereas others are generatedabiotically (22), but little attention has been given to thepossibility that biological chlorinations of lignin, the mostabundant terrestrial aromatic substance, may produce chloroaromaticcompounds. If microbial mechanisms exist to oxidize Cl^- in plantmaterial, the resulting electrophilic chlorine species are likelyto react with the electron-rich aromatic rings of lignin (4).This chemistry could explain, in part, why field samples ofsoil, litter, and decayed wood have all been shown to containhigh-molecular-weight chloroaromatics (5, 14).

Certain plant-pathogenic ascomycetes that are cosmopolitan inhabitantsof plant debris provide a possible route for Cl^- oxidation inthe vicinity of lignin (6, 8, 20). These fungi produce extracellularchloroperoxidases (CPOs), heme- or vanadium-containing enzymesthat oxidize Cl^- to hypochlorous acid (HOCl) or a similarlyreactive chlorine electrophile (18, 21). CPOs chlorinate a varietyof aromatic substrates (2, 15, 19, 23), but their reactivitywith lignin is unknown. We report here that the major structuresin lignin are not only chlorinated but also cleaved by the hemeCPO of Caldariomyces fumago and the vanadium CPO of Curvulariainaequalis.

Chlorination and cleavage of a lignin model dimer.

Top

Chlorination and Cleavage of...

Compound I, threo-1-(4-ethoxy-3-methoxy-ring-[¹⁴C]phenyl)-2-(2-methoxyphenoxy)-1,3-dihydroxypropane(1.0 mCi/mmol), a model of the major nonphenolic arylglycerol-ß-arylether structure in lignin (Fig. 1A), was prepared as describedpreviously (9, 12). The model was treated with either commerciallyavailable Caldariomyces fumago heme CPO (Sigma) or partiallypurified Curvularia inaequalis vanadium CPO. The latter enzymewas obtained by DEAE-Sephacel chromatography of the extracellularmedium from Curvularia inaequalis cultures (ATCC 10713) as describedearlier (17). The reaction mixtures (1.0 ml) contained compoundI (3.4 x 10⁵ dpm, 160 µM), KCl (20 mM), and buffer (25mM). The buffers were potassium phosphate, pH 3.0, for hemeCPO reactions and sodium acetate, pH 4.0, for vanadium CPO reactions.Enzyme was added (35 monochlorodimedone units of heme CPO or1 monochlorodimedone unit of vanadium CPO) (6), followed byH₂O₂ (1.7 mM), and the reaction mixtures were stirred overnightat ambient temperature.

View larger version (24K):
[in this window]
[in a new window]
FIG. 1. (A) Structures of model I and of chlorinated products produced from it by CPO. A standard of compound VII was prepared by treating 4-chloroguaiacol (Aldrich) with ethyl iodide. ¹H nuclear magnetic resonance: 1.36 (t), J = 7.0 Hz; 3.83 (s); 4.02 (q), J = 7.0 Hz; 6.86 (dd), J = 8.4 Hz, 2.2 Hz; 6.91 (d), J = 8.4 Hz; 6.95 (d), J = 2.2 Hz. A standard of compound VIII was prepared by treating compound VII with one equivalent of Cl₂. ¹H nuclear magnetic resonance: 1.39 (t), J = 7.0 Hz; 3.86 (s); 4.08 (q), J = 7.0 Hz; 7.08 (s); 7.10 (s). (B) Diagnostic ions from the fragmentation of diacetylated compounds I to VI in the mass spectrometer. The distribution and number of chlorine substituents in each molecule were calculated from the relative frequencies of ³⁵Cl- and ³⁷Cl-containing ions (13). See Table 1 for the mass spectra.

The completed reaction mixtures were stirred for 1 h with 0.25ml of acetonitrile, and portions (0.50 ml) were analyzed byreversed-phase high-performance liquid chromatography (HPLC)on a Luna 5-µm-pore-size C₁₈ column (Phenomenex). Thecolumn was eluted at 1.0 ml/min and ambient temperature witha linear gradient of acetonitrile ranging from 25 to 50% inwater over 60 min. Fractions (0.5 ml) were collected and assayedfor ¹⁴C by scintillation counting. The results showed that theCPOs generated numerous products from model I when H₂O₂ andCl^- were both present. No reaction occurred when either H₂O₂or Cl^- was omitted. Typical data for the vanadium CPO are shownin Fig. 2 and Table 1. Experiments with heme CPO or HOCl gavethe same products (data not shown).

View larger version (19K):
[in this window]
[in a new window]
FIG. 2. HPLC radiochromatogram of products obtained from the treatment of model I with vanadium CPO. See Fig. 1A for the structures of compounds I to VIII.

View this table:
[in this window]
[in a new window]
TABLE 1. Products generated in the reaction of model I with vanadium CPO

The fractionated products from a scaled-up reaction were extracted,acetylated, and analyzed by gas chromatography-electron impactmass spectrometry. The results showed that most of the products(II to VI) were uncleaved, chlorinated derivatives of modelI (Fig. 1; Table 1). However, two of the minor products, VIIand VIII, resulted from C-aryl cleavage of model I. HPLC andgas chromatography analyses showed that these products cochromatographedwith standards of 1-chloro-4-ethoxy-3-methoxybenzene and 1,2-dichloro-4-ethoxy-5-methoxybenzene,respectively. Mass spectra of the products and standards confirmedthese identifications (Table 1). Product VII probably resultedfrom attack by an electrophilic chlorine species at C-1 of theA ring in model I, followed by elimination of the model's aliphaticside chain (4). Product VIII presumably resulted from reactionsbetween chlorine electrophiles and products (e.g., II or VII)that already carried one chlorine on the A ring. These resultssuggested that CPOs might cleave lignin when they chlorinateit.

Depolymerization of synthetic lignins.

Top

Depolymerization of synthetic...

A high-molecular-weight synthetic ß-¹⁴C-labeled guaiacyllignin (0.01 mCi/mmol of phenylpropane substructures) and apermethylated (nonphenolic) preparation of the same lignin wereprepared as described previously (7, 10) and treated with CPOs.The reaction mixtures (final volume, 40.0 ml; magnetically stirredat ambient temperature) contained [¹⁴C]lignin (2.6 x 10⁴ dpm;30 µM in phenylpropane substructures), Tween 20 (0.25%),KCl (20 mM), and buffer (25 mM) as described above. The reactionswere run by pumping in separate solutions of H₂O₂ (final concentration,0.6 mM) and CPO (300 U for heme CPO and 20 U for vanadium CPO)over an 8-h period.

N,N-Dimethylformamide (DMF; 10 ml) was added to the completedreaction mixtures, which were then concentrated to less than5 ml by rotary vacuum evaporation and clarified by centrifugationto give DMF-soluble [¹⁴C]lignin samples in a 75 to 95% yield.The samples were subjected to gel permeation chromatography(GPC) on a 1.8- by 33-cm column of Sephacryl S-100 in DMF thatcontained 0.1 M LiCl, and fractions (2.0 ml) were collectedfor quantitation of ¹⁴C by scintillation counting.

The GPC column was calibrated with polystyrene standards thatspanned the molecular weight range of 500 to 35,000 and alsowith two lignin model compounds: a tetramer with a molecularweight of 1,062 (compound 197 in the U.S. Dairy Forage ResearchCenter NMR Database of Lignin and Cell Wall Model Compounds[http://www.dfrc.ars.usda.gov/software.html]) and a monomerwith a molecular weight of 166 (3,4-dimethoxybenzaldehyde).An iterative least-squares procedure was used to determine therelationship between the elution volume and the molecular weightfor the standards. The weight-average molecular weight (M_w)and number-average molecular weight (M_n) for each lignin samplewere then calculated by using the standard formulas (24).

The heme CPO had no discernible effect on the lignin in controlreaction mixtures that contained Cl^- without H₂O₂, whereas itpolymerized the lignin in reaction mixtures that contained H₂O₂without Cl^- (Fig. 3A). This result reflects the fact that hemeCPO in the absence of Cl^- exhibits a general peroxidase activity(3), which is expected to catalyze coupling reactions betweenphenolic structures in lignin. By contrast, depolymerizationwas the predominant fate of the lignin in heme CPO-catalyzedreaction mixtures that contained both H₂O₂ and Cl^-. Since Cl^-is the preferred reductant for CPO, its ability to reverse theenzyme's polymerizing activity suggests that an oxidized chlorinespecies was the ligninolytic agent in this experiment. Reactionsconducted with vanadium CPO or HOCl showed that they also depolymerizedlignin (data not shown).

View larger version (15K):
[in this window]
[in a new window]
FIG. 3. GPC radiochromatograms showing the molecular weight distributions of synthetic lignins treated with heme CPO. (A) Results for phenolic lignin. Dashed line, reaction without H₂O₂ (M_w = 8,329; M_n = 3,673); dotted line, reaction without Cl^- (M_w = 11,838; M_n = 4,441); solid line, complete reaction (M_w = 7,662; M_n = 1,190). (B) Results for nonphenolic lignin. Dashed line, reaction without H₂O₂ (M_w = 9,321; M_n = 4,316); solid line, complete reaction (M_w = 8,437; M_n = 1,700). Elution positions for the tetrameric and monomeric lignin model compound standards are indicated.

The heme CPO also cleaved synthetic lignin in which the phenolicstructures had been blocked as methyl ethers (Fig. 3B). Thisresult is significant because the chlorine electrophiles generatedby CPOs are expected to be less reactive with aryl ethers, whichcomprise most of the linkages in lignin, than they are withphenols, which are minor structures in the polymer (4). Thatis, heme CPO can act as a functional, if inefficient, analogueof lignin peroxidase (7). Our supplies of vanadium CPO and permethylatedlignin were too small to assess their reactivity together.

Conclusion.

Top

Conclusion.

Our results show that fungal CPOs are potential chlorinatorsof lignin in plant debris and soil and may thus account forsome of the high-molecular-weight organochlorine residues thatoccur naturally. Because they are ligninolytic, CPOs may alsocontribute to the slow breakdown of lignin in soils. CPO activity(1) and H₂O₂ (11, 16) both occur in soils, and Cl^- is of courseubiquitous in this environment. CPO-catalyzed ligninolysis providesan interesting biological parallel to the well-known industrialuse of chlorine for wood pulp delignification (4).

ACKNOWLEDGMENTS

We thank Jennie Hunter-Cevera and Ron Wever for valuable adviceon CPO production. Important contributions to this work weremade by Diane Dietrich, Kolby Hirth, Carl Houtman, Ken Jensen,Jr., and Sally Ralph.

This work was supported by grants from the U.S. Department ofEnergy (DE-FG02-94ER20140) and the U.S. Department of Agriculture(98-35103-6583) to K.E.H.

FOOTNOTES

* Corresponding author. Mailing address: Institute for Microbial and Biochemical Technology, USDA Forest Products Laboratory, One Gifford Pinchot Dr., Madison, WI 53726. Phone: (608) 231-9528. Fax: (608) 231-9262. E-mail: kehammel{at}wisc.edu.

REFERENCES

Top