Oxidation of a Tetrameric Non-phenolic Lignin Model Compound by Lignin Peroxidase†

The present study maps the active site of lignin peroxidase in respect to substrate size using either fungal or recombinant wild type as well as mutated recombinant lignin peroxidases. A nonphenolic tetrameric lignin model was synthesized which contains b -O-4 linkages. The fungal and recombinant wild type lignin peroxidase both oxidized the tetrameric model forming four products. The four products were identified by mass spectral analyses and compared to synthetic standards. They were identified as tetrameric, trimeric, dimeric and monomeric carbonyl compounds. All four of these products were also formed from single turnover experiments. This indicates that lignin peroxidase is able to attack any of the C a -C b linkages in the tetrameric compound and that the substrate-binding site is well exposed. Mutation of the recombinant lignin peroxidase (isozyme H8) in the heme access channel, which is relatively restricted and was previously proposed to be the veratryl alcohol binding site (E146S), had little effect on the oxidation of the tetramer. In contrast, mutation of a Trp residue (W171S) in the alternate proposed substrate-binding site completely inhibited the oxidation of the tetrameric model. These results are consistent with lignin peroxidase having an exposed active site capable of directly interacting with the lignin polymer without the advent of low molecular weight mediators.


INTRODUCTION
located at the enzyme surface and would only allow for long-range electron transfer (20). We have designed a tetrameric nonphenolic lignin model compound, which does not fit into the heme access channel (Fig. 1) and determined the initial products of its oxidation. The product profile indicates that the site of electron transfer of LP is very exposed. Site-directed mutagenesis studies where Glu146 and Trp171 were altered, indicate that the site of electron transfer is Trp171.

MATERIALS AND METHODS
Chemicals. Ammonium cerium(IV) nitrate and 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) were purchased from Aldrich. The oligonucleotides for site directed mutagenesis were synthesized by Integrated DNA Technologies, Inc. (Coralville, IA). H 2 O 2 was purchased from VWR and prepared fresh daily; the concentration was determined spectrophotometrically at 240 nm using an extinction coefficient of 39.4 M -1 cm -1 (21). The tetrameric lignin model and its carbonyl degradation products (Fig. 2) were synthesized and purified as described by (22). The tetrameric compound is a mixture of two diasteromeric compounds (ratio 1:1) because of the presence of R and S configurations in the terminal α-position. All other chemicals were commercially available and used without further purification.  peroxidases were refolded and purified as previously described (24).
Wild type fungal LP isozymes H2 and H8 were purified as previously described (25). The concentrations of the peroxidases were determined spectrophotometrically at 409 nm using an extinction coefficient of 169 mM -1 cm -1 and 168 mM -1 cm -1 for LP isozymes H2 and H8, respectively (26). The purified preparations exhibited an RZ ratio (A409/A280) of least 3.

Identification of oxidation products. Incubation of the tetramer with fungal LP isozyme H2
and H 2 O 2 resulted in cleavage of the tetramer without the addition of the putative mediator, veratryl alcohol. Previous work (10) had shown that the predominant products from LP-catalyzed oxidation of lignin models are that of C α -C β cleavage. The predicted products from such cleavage are shown in Fig. 2. Incubation under multiple turnover conditions resulted in formation of four products as detected by HPLC (Fig. 3A), which were identified by LC/MS. The mass spectra and the corresponding retention times of the oxidation products were compared to those of the authentic standards ( Table 1). The proposed products from C α -C β cleavage are also intermediates in the synthesis of the tetramer (22) and consequently were available as known standards. As shown in Table 1, three of the four products were identical to the standards and were identified as the trimeric, dimeric and monomeric aldehydes. The fourth predicted product, labeled 2 in Fig. 3 did not readily ionize and thus was not easily identified by LS/MS. Previous work had also shown that oxidation of Cα alcohols to ketones was also a predominant reaction of LP catalyzed reactions (10).
We thus suspected that unknown 2 is a products where one of the Cα alcohols was oxidized to the corresponding carbonyls. To test this possibility, we incubated the tetramer (100 µM) with a chemical oxidant specific for oxidizing Cα alcohols to the ketone/aldehyde, DDQ (23). The oxidation was carried out with tenfold less DDQ relative to the tetramer to ensure that only one product was formed. Analysis by HPLC (Fig. 3B) indicated that the product generated by DDQ had an identical retention time as that of the unknown degradation product. This is consistent with unknown 2 as one of the four possible tetrameric carbonyl compounds. The results also suggest that the oxidation products are substrates for further LP catalysis.
We thus determined whether the trimer, dimer and monomeric (aldehyde) models were substrates for LP. These models were incubated under the same conditions as used in the experiment of tetramer oxidations. The products were again identified and quantified by HPLC ( Table 2). The trimeric and dimeric aldehydes were further oxidized, while the monomer was not a substrate for isozyme LPH2. The oxidation of the trimeric compound resulted in formation of both dimeric and monomeric products, whereas the dimer yielded only the monomeric aldehyde.
Effect of veratryl alcohol. Several previous studies reported the stimulatory effect of veratryl alcohol on the oxidation of a variety of chemicals with ranging from aromatic monomers to polymeric lignin. Therefore, we also tested the effect of veratryl alcohol on the oxidation of this tetrameric lignin model. Fig. 5A shows the decrease of tetramer in the absence and presence of veratryl alcohol over a four minute time period. The addition of veratryl alcohol enhanced the rate of tetramer degradation. Approximately 70% of the tetramer was degraded in the presence of veratryl alcohol during the first 30 seconds, whereas only 50% of the tetramer disappeared without veratryl alcohol in the same interval. The product profile was similar with or without added veratryl alcohol. The fate of veratryl alcohol was also followed by HPLC during this experiment.
All of the veratryl alcohol was oxidized to veratraldehyde within 2 min. (Fig. 5B).
Products formed from single turnover. The time course experiment shown in Fig. 4 suggested that LP equally oxidized all four possible cleavage sites. If LP acts only as an "exo" lignin-degrading enzyme, then the time course would first show the formation of the trimer, concomitant with formation of the monomer. This would then be followed by formation of dimers.
To further investigate whether LP could act as an "endo" enzyme yielding also dimers as the initial product, single turnover experiments were performed. Single turnover conditions were attained in a three-syringe stopped flow apparatus as described previously (27). Slightly less than one equivalent of H 2 O 2 was pre-incubated with LP isozyme H2; this ensured only single turnover. The reaction mixture then aged for 1.5 second to permit complete formation of compound I. The second push of the stop flow then mixed the freshly prepared compound I with the tetramer. This mixture was collected and immediately analyzed by HPLC. As observed under steady state conditions, all four oxidation products were detected ( Table 3). The products were roughly similar in concentration.
Oxidation by Ce 4+ . Detection of all four possible oxidation products in single turnover experiments would indicate that no specificity existed with LP-catalyzed reactions. To corroborate these results, we also determined the product formed from oxidation of the tetramer with a chemical oxidant. As shown in Table 3, oxidation of the tetramer with sub-stoichiometric amount of Ce 4+ , a single electron oxidant, resulted in the formation of all four possible products.
Characterization of mutants of LP. Two substrate binding sites have been suggested for LP, the so-called heme access channel and a more surface exposed site with Trp 171 (8,28). We investigated the nature of the substrate-binding site using the tetrameric model as the substrate.
Mutant E146S (heme access channel) and W171S (surface residue) were generated in the recombinant mutant LP (isozyme H8). Their activity with the tetramer was then examined and compared to wild type LP. The product profile from oxidation of the tetramer by the fungal H8 and the recombinant H8 (rH8) was similar to that of fungal isozyme H2. Mutant E146S, located at the heme-access channel was also able to oxidize the tetramer. The product profile was similar to the wild type enzyme (Table 4). In contrast, oxidation of the tetramer was completely inhibited by the mutation of W171S (Table 4). However, as reported by Doyle et al. (28), the oxidation of ABTS was not inhibited by this mutation (data not shown).

DISCUSSION
Results from the present study help resolves the issue of whether LP can degrade polymeric lignin without redox mediators and has also identified the site of electron transfer for polymeric substrates. These results have relevance on the role of LP in lignin degradation, a question still debated after the discovery of LP 17 years ago. This debate is due, in part, to the finding that many lignin-degrading fungi apparently do not contain LP (29) and to the relatively slow rate of this enzyme with the substrate lignin (30). Because LP is proficient in oxidizing low molecular weight models of lignin, some have proposed that the role of LP in lignin degradation is solely for the oxidation of smaller lignin fragments (30), generated by MnP catalysis. The ability of LP to bind to its putative substrate, lignin, has also been questioned. Research from a number of laboratories has shown that the pores of the wood matrix are not large enough to allow access to enzyme (12)(13)(14) To gain insight into the nature of the binding site, we have synthesized a larger (tetrameric) model of lignin to map the size of the LP site of electron transfer. We utilized a tetrameric substrate, which is actually larger than a tetramer since a benzyl group rather than a methyl group is attached to the terminal phenolic oxygen. The tetramer was designed such that if the site of electron transfer is the heme access channel, it could only gain access from an "end-on" manner. This is clearly illustrated in Fig. 1 where the size of the tetramer is compared to that of LP and its heme access channel. Our studies clearly indicate, through time course studies and through single turnover studies, that LP is able to oxidize the tetramer to yield the dimer as the first product. In all incubations with the tetramer, the monomeric, dimeric, trimeric, tetrameric aldehydes were detected. Of significance here is the finding that dimeric products are formed from single turnover studies. This clearly maps the site of electron transfer to be relatively unhindered and that LP can oxidize the lignin polymer directly. In accord with this interpretation is our results from the chemical oxidation system. The single-electron oxidant, Ce 4+ , yielded the same profile of products. Thus, LP can act as an "endo" lignin-degrading enzyme. This is consistent with LP being able to interact with lignin directly and catalyze C α -C β cleavage anywhere along the polymer.
Indirect evidence for interaction between LP and lignin (synthetic polymer DHP) was recently obtained with a resonant mirror biosensor Johjima et al. (33). The interaction between the enzyme and DHP was only specific for LP; none was observed with laccase, MnP or horseradish peroxidase. In addition to this spectroscopic evidence, these workers also obtained rate constants between LP compound I and the DHP. However, this study did not reveal the nature of the products or reveal the location of the substrate-binding site. Terminal phenolic groups may have served as the reducing agent for both compounds I and II in this study.
Despite the finding that LP can directly interact with a large polymeric lignin substrate, we also investigated whether veratryl alcohol could enhance the oxidation. As found in the oxidation of many other substrates for LP, veratryl alcohol stimulated the oxidation of the tetrameric model.
The stimulation was measured to be approximately 20%, a result, which neither supported nor refuted the mediation hypothesis. Past work has shown that facilitation of oxidation of secondary substrates can be explained by several mechanisms. Although, our experiments were not designed to address the mediation issue, they are more consistent with veratryl alcohol not acting as a mediator. In incubations where veratryl alcohol acts as a mediator, a lag phase in its oxidation has been observed (18). As shown in Fig. 5, a lag phase of veratraldehyde formation was not observed.
On the contrary, both the tetramer and veratryl alcohol were oxidized simultaneously. Thus, it most probably acted to protect LP from inactivation. In the case of poor LP substrates, veratryl alcohol has been demonstrated to promote the completion of the enzymatic catalytic cycle, decreasing the inactivation of LP, and consequently increase the oxidation rate of the other substrate. The mechanism of stimulation was not further investigated. Even if some mediation does occur with veratryl alcohol, these results do not refute our finding that LP is able to directly oxidize large lignin molecules.
Our studies have also identified the location of the electron transfer for polymeric substrates. We have made mutations among the amino acids in the heme excess channel (24).
Mutation of E146 at this site was further characterized in the present study. This amino acid residue was proposed to participate in veratryl alcohol binding and involved in pH dependency (8).
The enzyme mutated at the heme access channel was able to oxidize the tetrameric lignin model resulting in the same products as those by the wild type. Recently W171 of the surface exposed potential binding site was mutated (28). The involvement of this amino acid in the oxidation of veratryl alcohol was proposed. The W171S mutant, in contrast, was not able to oxidize the tetramer, no products were found. These findings illustrate the significance of this site for electron transfer with veratryl alcohol and the polymeric lignin. The involvement of W171 in veratryl alcohol oxidation is supported by other studies (34,35). The significance of W171 is also suggested from the results on bi-functional MnPs. Certain MnPs from Bjerkandera spp. and Pleurotus oxidize veratryl alcohol (36,37). These enzymes also contain a Trp at the homologous site (37). Moreover, when MnP H4 was mutated to contain a Trp (168W) (34,35), it was able to oxidize veratryl alcohol.
In conclusion, our results support the hypothesis that LP can directly oxidize lignin.
Therefore, LP may have important role in the degradation of phenolic and nonphenolic lignin.
Site-directed mutagenesis studies suggest that the substrate-binding site is surface exposed and that this site may accommodate both lignin and veratryl alcohol. However, other compounds such as ABTS and DFAD may bind at an alternate site. The presence of more than one binding site is supported by the fact that the mutation of W171 in LP did not affect the oxidation of certain dyes such as ABTS and DFAD (28).   47.1 0.0 0.0 0.0 0.0 a Incubation conditions were the same as those described in Table 2.  The arrows drawn in the structure show location of cleavage. Arrows leading away from these smaller arrows point to products formed from cleavage at that site.  (v), tetrameric carbonyl; (q), trimeric aldehyde; (∆), dimeric aldehyde; (O), monomeric aldehyde. Products were identified by LC/MS and quantitated by HPLC using standard curves generated from synthetic standards.