Cloning, sequence analysis, and expression of ligninolytic phenoloxidase genes of the white-rot basidiomycete Coriolus hirsutus.

Two cDNAs and two genomic DNAs coding for the allelic forms of the ligninolytic phenoloxidase were isolated from the white-rot fungus Coriolus hirsutus. The cloned genes were identified in genetic libraries by hybridization screening using four deoxyoligonucleotide probes which corresponded to the partial amino acid sequence of the purified enzyme. Each cDNA encoded the full-length of the phenoloxidase, a protein consisting of 499 amino acid residues, and its putative signal peptide of 21 amino acid residues. The nucleotide sequences of the two alleles differed by 18 single base changes within the open reading frames resulting in one amino acid substitution. Ten small introns interrupted both genomic DNAs as indicated by direct comparison with the corresponding cDNAs. Putative eukaryotic regulatory sequences, "CAAT" and "TATA," were observed in the 5'-flanking region of both genomic DNAs. Each of the phenoloxidase cDNAs was successfully expressed in an active form in Saccharomyces cerevisiae using the useful yeast expression vector YEp51.

Lignin is a structurally complex aromatic biopolymer and a major component of woody plants. Degradation of lignin is an important step in the pulp and paper making process and for other future applications of the biomass from woody plants. Since lignin is extremely resistant to attack by most microorganisms, chemical degradation of lignin has been the only applicable method for the pulping process.
In the last decade biological degradation of lignin has attracted considerable interest. Extracellular enzymes of some white-rot fungi have been shown to be highly effective agents (Kirk et al., 1987). Lignin peroxidase has been intensively studied in this regard and the gene for this enzyme has been cloned (Tien et al., 1987;Reddy et al., 1987;Cullen et al., 1988Cullen et al., , 1989Kuwahara et al., 1988). This hydrogen peroxide-dependent enzyme has been implicated in lignin degradation because of its capability of degrading lignin model compounds in uitro. For complete degradation of native lignin, however, the action of other enzymes, such as laccases, should be required in addition to that of lignin peroxidase.
We have studied the lignin-degrading white-rot fungus Coriolus hirsutus and demonstrated that this organism produces * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) 505562. and secretes several potent' laccases and lignin peroxidases (Sugiura et al., 1987). One of the laccases, a ligninolytic phenoloxidase, appears to be a key enzyme involved in the biological degradation of lignin (Sugiura et al., 1987). This enzyme cleaves the C-C bond and C-O bond of some phenolic lignin model compounds (Higuchi et al., 1988;Sugiura et al., 1987). In order to investigate the regulatory mechanisms controlling production of the ligninolytic phenoloxidase of C. hirsutus, we have cloned and determined the nucleotide sequence of phenoloxidase genes from this fungus. In addition, we have expressed the cloned genes in yeast. at 28 "C for 7 days in a glucose-peptone medium (glucose 30 g, peptone 10 g, KH,PO, 1.5 g, MgS0,.7H20 0.5 g, thiamine-HC12.0 mg, CuSO, 16 mg, and distilled water 1,000 ml) on a rotary shaker as described previously by Sugiura et al. (1987). Total RNA was extracted from the frozen phenoloxidase-producing mycelia as previously described (Broda et al., 1985). Poly(A)-containing RNAs were purified by oligo(dT)-cellulose chromatography twice according to the method of Aviv et al. (1972 Corp.) and T, polynucleotide kinase (Maniatis et al., 1983). Each filter replicate of the X-phage cDNA library and genomic cosmid library was screened by plaque and colony hybridizations, respectively, with these "'P-labeled oligonucleotide probes. Subclonin,g and Sequencing-Restriction DNA fragments were isolated corresponding to the cDNA inserts of phage DNAs from hybridization-positive plaques and the genomic DNA inserts of cosmids from hybridization-positive colonies. These DNA fragments were subcloned into appropriately digested pUC19 vectors (Messing et al., 1985). For nucleotide sequencing, a series of deletions of the subcloned DNA fragments were prepared using a "Deletion kit for kilo-sequencing" (Takara shuzo Co. Ltd. Japan). DNA sequencing was carried out in both directions by the dideoxy-chain termination method (Sanger et al., 1977) using alkali-denatured plasmid templates (Hattori et al., 1986). Isolation of Basidiospores  (Ito et al., 1983). S. cerevisiae harboring the constructed expression plasmids was cultured at 28 "C for 3-7 days in phenoloxidase assay medium, SD or SG media (0.67% yeast nitrogen base supplemented with amino acids and either 2% glucose or galactose as previously described (Broach et al., 1983)), containing 0.1 IIIM CuS04 and 0.1 ml/liter guaiacol.
Since inosine can form a base pair with any of the four natural nucleotides, 5 inosine residues were introduced at the highly redundant positions into probe 2, and 1 inosine residue into probe 1, in order to lower the overall redundancy of the oligonucleotide mixture (Ohtsuka et al., 1985).
Cloning of cDNAs for the Phenoloxidase of C. him&us-Polyadenylated RNAs were prepared from phenoloxidaseproducing mycelia of C. hirsutus and used for the cDNA synthesis. The cDNA library was constructed with Xgtll as a vector and screened by plaque hybridization with the four 32Plabeled oligonucleotides as hybridization probes. Five hybrid- ization-positive clones were obtained from 20,000 recombinant phages.
Five EcoRI restriction fragments inserted at the cloning site were in the size range of about 1.8-1.9 kilobase pairs and were subcloned into a pUCl9 vector. In a restriction endonuclease analysis, these clones showed similar but not identical restriction patterns (data not shown). These clones could be classified into two groups: one set with a EcoRV site (designated POl, Fig. 2B), the other without a EcoRV site (designated P02, Fig. 2B). Numbers refer to the nucleotide sequence and the predicted amino acid sequence of genomic clone 21. Numbering starts at the translational initiation site of the phenoloxidase. The 5' terminus of the cloned cDNAs is indicated by a triangle. Polyadenylated sites are marked with an arrow. Ten introns are shown in small letters. Putative promoter elements, "CAAT" and "TATA" box-like sequences, are underlined.
The cleavage site of the signal peptide is indicated by a solid triangle. idases. The cDNA clones, PO1 and P02, correspond to the genomic clones, 21 and 2A, respectively, which are further described below. Both cDNA inserts contain polyadenylated sequences which are connected to the T of position 2314 (marked with an arrow); however, the polyadenylated sequences are not depicted in the figure. Both cDNA inserts contain a single large open reading frame of 1560 nucleotides which starts from ATG at position +l. Both amino acid sequences predicted by these regions contain 520 amino acid residues (designated as 21aa and 2Aaa in Fig. 3) and should represent a full-length phenoloxidase. This result is completely consistent with the partial amino acid sequence (122 amino acid residues) of the phenoloxidase of C. hirsutus (Fig. lA) as determined by Edman degradation. The two open reading frames differed by 18 single base changes but 17 of these base changes occurred in the third position of the codon. Therefore, only one amino acid substitution, Ala-411 to Pro, was indicated. The calculated molecular weights of the two mature polypeptides were estimated at 53,516 (POl) and 53,542 (PO2), respectively.
A comparison of the amino acid sequence deduced from the DNA sequence with the amino acid sequence of the NH, terminus of the purified mature phenoloxidase (Fig. IA) indicated that the cleavage site of the signal peptide exists between Ala-21 and Ala-22 of the polypeptide chain 21aa (Fig.  3). These signal peptides, consisting of 21 amino acid residues, were typical signal peptides. A basic amino acid residue, Arg-3, was located in the NHp-terminal region and many hydrophobic amino acid residues were found in the central region. A short side chain amino acid residue, Ala-21, was located at the cleavage site in the COOH-terminal region of the signal peptide.
Cloning of Genomic Genes for the Phenoloxidase of C. hirsutus-The genomic cosmid library constructed with cosmid vector pHC79 was screened by colony hybridization with the same four 32p-labeled oligonucleotide probes (Fig. 1B) used in the cDNA cloning. Nine hybridization-positive clones were obtained from 40,000 transformants. Their cosmid DNAs were analyzed by digestion with several restriction endonucleases, followed by Southern transfer (Southern, 1975) and hybridization with the four 32P-labeled oligonucleotide probes.
Each of these cosmids contained an inserted 4.6-kb EcoRI DNA fragment which hybridized with all four probes. Two different types of genomic DNA clones were recognized according to the existence of an EcoRV site within the 4.6-kb EcoRI DNA fragment. This result is consistent with the existence of the two allelic types of cDNAs cloned. EcoRI-SmaI DNA fragments (extending from the EcoRI site on the left side of ATG to the SmaI site on the right side of TAG, Fig. 2A) were subcloned into appropriately digested pUC19 vectors. A typical representative of both types of these genomic DNA clones was then sequenced: 21, one of six positive clones having an EcoRV site, and 2A, one of three positive clones lacking an EcoRV site. The results are shown in Fig.  3.
A comparison of the two nucleotide sequences of the genomic phenoloxidase genes with those of the two cDNAs (Fig.  2) revealed that the EcoRI-SmaI DNA fragments of clones 21 and 2A contained the whole nucleotide sequences of cDNA clones PO1 and P02. In addition both structural genes for the phenoloxidases contained 10 introns consistently located at the same positions.
Genomic Southern Hybridization-Two extremely homologous phenoloxidase genes were isolated from dikaryotic genomes of C. hirsutus. In order to determine the multiplicity of the phenoloxidase genes existing in the genome of C. hirsutus, genomic Southern hybridizations were carried out (Fig. 4). The dikaryotic genomic DNA fragments digested with EcoRI and EcoRV were used for the analysis. A 32Plabeled 4.6-kb EcoRI DNA fragment containing the phenoloxidase gene was used as a probe. It was observed that three bands (4.6, 2.6 and 2.0 kb) hybridized with this probe (Fig. 4, nomic clone 21. respectively. Southern hvhridization of EcoRl and EroRV-digested ze-.
"omit DNAs'of the four types of monokaryotic mycelia (AlBl,AlBZ, A"B1, and .4&j ofC. hirsutr~.swasalsocarried out with the same ."P-labeled probe as described above. As indicated in Fig, 4 (Inne Z-51, each monokaryotic genome apparently has only one of the two t.ypvpes of phenoloxidase gl?"eS.

Exprr.wion
o/ the f'hewluxidose cDNA in S. cerwisine-Construction of the phenoloxidase expression plasmids, pYK28 and pYK?S, is depicted in Fig. 5. The two cloned phenoloxidase cDNAs were ligated to plasmid pYK25 which was produced by the insertion of a AnmHI linker into the yeast expression vector YEpSl. With these plasmids, erpression of the cloned phenoloxidase genes is controlled hy the yeast Gall0 promotor.
Expression of the yeast GAL10 gene coding for uridine diphosphoglucose 4.epimerase (EC 5.1.3.2) is strictly regw lated in S. cerevisinp according to the existence of specific carbon s"urces in the medium. Thus. the expression of this gene was shown t" be induced in the presence of galactose and repressed by glucose (Oshima. 1982;Johnston, 1987).
Transformants were cultured on not occur (Fig. FL These results indicated that the GAL10 rxomoter is activated in the presence of galactose. and expression of the phenoloxidase cDNAs is induced. Moreover. phenoloxidase produced in S. cerucisirr~ is secreted into the medium and catalyzes the "xidative polymerization of guaiacol to form reddish-brown zones in the medium in the presence of 0,. craw--The nucle"tide sequence of the lnccase geene from N. cmsso. which encodes abuut 600 amino acid residues. has been reported ILerch P, a/.. 1988t. In order t" analyze the structural relationships between the phenoloxidase from C. hirsutus and the lactase from N. crass", the nucleotide sequences and the amino acid sequences mere compared by dot-matrix analysis.  (Lerch et al., 1966) and human ceruloplasmin (Hcp, Takahasi et al., 1983). Numbers on the left of each sequence represent the positions of the amino acid residues of these enzymes. Identical amino acid residues are boxed. Potential coordination sites for the three different types of copper ions are indicated by *I, *2, and *3.
shown). However, the amino acid sequences exhibited weak homology between the entire sequence of the phenoloxidase of C. hirsutus and about 500 amino acid residues at the COOH-terminal region of the N. crassa lactase (Fig. 7). Homologous Region of Copper Enzymes-A strongly conserved region of some multicopper oxidases of eukaryotic species, including fungal and human, has been reported (Lerch et al., 1986). This region is considered as a potential coordination site for four copper ions to form a redox center. A search for this conserved region in the amino acid sequence of the phenoloxidase from C. hirsutus revealed a region of high homology within the COOH-terminal region of the enzyme (Fig. 8).

DISCUSSION
In this paper, we have described the cloning, sequence analysis, and expression in S. cereuisiue, of two allelic types of phenoloxidase genes isolated from the white-rot basidiomycete C. hirsutw.
We observed that the two cDNAs, and similarly the two genomic cosmid clones, differed according to the existence of an EcoRV site. Furthermore the two alleles were shown to differ by a number of nucleotide base changes as determined by sequencing the two cDNAs (PO1 and PO2) and the two genomic clones (21 and 2A). Many of the base changes were located in the 3'-noncoding regions or the introns. Seventeen of the 18 base changes within the open reading frame of the phenoloxidase genes occurred at the third position of the triplet codon. These 17 changes do not result in amino acid substitution, therefore, only one amino acid difference was observed between the two allelic forms of phenoloxidase.
A comparison of the nucleotide sequences of the two genomic phenoloxidase genes with those of the two cDNAs ( Fig.  2) revealed that 1) the EcoRI-SmaI DNA fragments of clones 21 and 2A contained the whole nucleotide sequences of the cDNA clones (PO1 and PO2), and 2) both structural genes for the phenoloxidase contained 10 introns consistently located at the same positions. The 10 introns were homologous and small, ranging in size between 50 and 62 base pairs. Two consensus sequences "GTa/ga/tGt/c" at the 5' splicing site and "c/tAG" at the 3' splicing site of each intron are similar to the intron consensus sequences of higher eukaryotes.
The sequence ACC . ATG which signals initiation of translation of the phenoloxidase genes is consistent with the consensus sequence, ACC. ATG or GCC-ATG, for initiation of translation in vertebrates found by Kozak (1987). As shown in Fig. 3, a "CAAT" box-like sequence, CAATCT, and "TATA" box-like sequence, TATAAA, are located at 199 and 99 base pairs upstream from the translation initiation codon in the 5'-flanking region of the clone 21. This result is consistent with the finding of similar putative eukaryotic regulatory sequences by Corden et al. (1980). These sequences should have a functional role in transcription of the phenoloxidase gene.
A comparison of the amino acid sequences of the phenoloxidase from C. hirsutus and those of the lactase from N. crassa, revealed weak homology between the entire sequence of the phenoloxidase from C. hirsutus and 500 amino acid residues of the COOH-terminal region of the lactase from N. crassa ( Fig. 7). This result suggests that the active site and conformation of the ligninolytic phenoloxidase might be similar to those of other laccases.
The C. hirsutus phenoloxidase was found to contain a region homologous to a strongly conserved region of some multicopper oxidases of eukaryotic species. This region is considered to contain a potential coordination site for four copper ions of three different types which form a redox center. Comparison of the homologous region of the phenoloxidase from C. hirsutus and other enzymes (Fig. 8) suggests that the coordination site for the three types of copper ions is as follows. A type-l copper ion (blue copper) coordinates at His-416 and Cys-474. A type-2 copper ion coordinates at His-475. Two type-3 copper ions coordinate at His-419, His-421, His-423, and His-473. Briving et al. (1975) reported that fungal lactase B from Polyporus versicolor (Coriolus versicolor) consists of a polypeptide chain of approximately 545 amino acid residues with two internal disulfide bridges and one sulfhydryl group. Nyman et al. (1980) determined the primary structure around the sulfhydryl group and located one of the disulfide bridges of the lactase. Judging from these results, we can conclude that the primary structure of the phenoloxidase of C. hirsutus is similar to that of lactase B. Both proteins contain 5 cysteine residues at similar positions and two disulfide bridges of the phenoloxidase could similarly be proposed to exist between Cys-106 and Cys-509, and between Cys-138 and Cys-226, respectively.
It is consistent that the remaining sulfhydryl group, i.e. Cys-474 of the peptide -Leu-His-Cys-His-is the cysteine which coordinates a blue copper as described above. Analysis of secreted active phenoloxidases produced by S.
cereuisiae indicated that disulfide bridges were constructed appropriately and four copper ligands occupied the correct coordination sites.' Lerch et al. (1988) reported that laccases isolated from many different sources consist of a single polypeptide chain with a length of approximately 500 amino acids residues. Similarly, the two phenoloxidases from C. hirsutus consist of 499 amino acids residues. The calculated molecular weights, 53,516 (POl) and 53,542 (PO2), for the two phenoloxidases are not identical to the molecular weight of the purified enzyme, 63,000, as determined by sodium dodecyl sulfatepolyacrylamide gel electrophoresis. However, this difference could be explained if glycosylation of the phenoloxidase occurs at any of the seven potential N-glycosylation sites (Asn-X-Ser/Thr) observed in their amino acid sequences. Indeed, this hypothesis was confirmed by detection a carbohydrate content of about 15s.' The carbohydrate chains may not be essential for enzyme activity and extracellular secretion of the phenoloxidase. The carbohydrate chain of lactase III produced by C. uersicolor is not essential for enzyme activity (Morohoshi et al., 1983). Furthermore, it seems likely that the two cloned phenoloxidases of C. hirsutus produced in S. cerevisiae were glycosylated with a yeast-specific carbohydrate, i.e. mannose, and yet such enzymes were active and secreted into the medium. It should also be pointed out that the original signal peptides of the phenoloxidase of C. hirsutus were processed correctly during secretion of the enzyme in yeast cells.
The results of Southern hybridization of the genomic DNA of monokaryotic mycelium of C. hirsutus suggested that the allelic forms of the phenoloxidase genes exist in separate nuclei of dikaryotic cells. This observation can be explained as follows. First, two nuclei never fuse in the dikaryotic basidiomycete cell, except during sporulation in the fruiting body, and the period of two distinguished nuclei in the dikaryotic cell is very long. Second, restriction site polymorphism between the two genome equivalents in the dikaryotic basidiomycete mycelium was observed and a similar observation was reported by Broda et al. (1986). Since all mycelia prepared from isolated basidiospores of C. hirsutus produced phenoloxidase, apparently both allelic forms of the phenoloxidase gene are active and thus neither can be a pseudogene. The proof is the active expression and secretion of the two cloned phenoloxidases in yeast.