Biochemical and molecular characterization of the diphenol oxidase of Cryptococcus neoformans: identification as a laccase

Melanin production is a major virulence factor for Cryptococcus neoformans, an organism causing life-threatening infections in an estimated 10% of AIDS patients. In order to characterize the events involved in melanin synthesis, an enzyme having diphenol oxidase activity was purified and its gene was cloned. The enzyme was purified as a glycosylated 75-kDa protein which migrated at 66 kDa on sodium dodecyl sulfate-polyacrylamide gel electrophoresis after deglycosylation by endoglycosidase F. Substrate specificity resembled that of a laccase in that it oxidized multiple diphenolic and diamino compounds. Dopamine was shown by mass spectroscopy to be oxidized to decarboxy dopachrome, an intermediate of melanin synthesis. The enzyme contained 4.1 +/- 0.1 mol of copper per mol. It resembled a laccase in its absorbance spectrum, containing a peak of 610 nm and the shoulder at 320 nm, corresponding to the absorbance of a type I and type III copper, respectively. The cloned gene of C. neoformans laccase (CNLAC1) contained a single open reading frame encoding a polypeptide 624 amino acids in length. The encoded polypeptide contained a presumptive leader sequence, on the basis of its relative hydrophobicity and by comparison of the sequence to that of the N-terminal sequence of the purified enzyme. CNLAC1 also contained 14 introns ranging from 52 to 340 bases long. Transcriptional activity of CNLAC1 was found to be derepressed in the absence of glucose and to correspond to an increase in enzymatic activity.

Cryptococcus neoformans is a pathogenic yeast-like fungus which causes disease preferentially in the brains of humans and experimental animals (see review in reference 17). Melanin production is thought to be a major virulence factor for C. neoformans in that melanin-negative mutants lose virulence and in that coreversion of the melanin phenotype and restoration of virulence have been observed (19). C. neoformans is the only member of the genus with demonstrable melanin production (3,39,47), a property that is used extensively for presumptive species identification. Melanin is thought to provide a protective role for the fungus by scavenging leukocytic antimicrobial oxidants, since melanin-negative mutants are more susceptible to oxidative damage (15). These observations make this system a potential target for inhibitors, similar to the melanin inhibitors tricyclazole, fthalide, and pyroquilon (41,42) used commercially to prevent rice blast disease caused by the plant pathogen Pyricularia oryzae.
The synthesis of melanin is a process within diverse biological systems in which mono-or polyphenolic substrates are oxidized to their respective quinones, which in turn undergo nonenzymatic polymerization to produce pigmented products (see review in reference 48). Production of a melanin-like pigment in C. neoformans was first described by Staib (43) and found to be dependent on the presence of o-orp-diphenols or diaminobenzenes (4). Characterization of a detergent-solubilized diphenol oxidase from C. neoformans by Polachek et al. (32) showed the enzyme to have a wide substrate specificity for the oxidation of polyaminobenzenes and polyphenolic compounds. This diphenol oxidase activity in C. neoformans has been found to be invariably absent in melanin-negative mutants (19,20) and present in strains producing melanin (4). Recently homogenized cells without detergent by Ikeda and Jacobson (14) using cells grown in acidic asparagine salts. However, the study of melanogenesis in C. neoformans has been hampered by an inability to purify the putative enzyme responsible for initial oxidation of the diphenolic substrate.
In order to characterize the oxidative enzyme associated with melanin synthesis in C. neoformans, the diphenol oxidase was purified from a pathogenic strain of the fungus. The enzyme was then assessed for its ability to oxidize diphenolic substrates and to produce decarboxy dopachrome, a putative intermediate in the production of melanin-like pigments. In addition, the gene encoding the enzyme was cloned and sequenced to obtain further information as to its classification as well as to investigate its transcriptional regulation.

MATERIALS AND METHODS
Strains. C. neoformans ATCC 34873 was a generous gift of K. J. Kwon-Chung, Escherichia coli SURE (Stratagene, La Jolla, Calif.) was the host strain used for screening the cDNA library after mass excision of the Uni-Zap cDNA library. E. coli XL1-Blue (Stratagene) was the recipient strain of the Bluescript phagemid following in vivo excision from the Uni-Zap XR vector (Stratagene) containing genomic clones. E. coli DH5a (GIBCO-BRL Life Technologies, Bethesda, Md.) was the host strain for the plasmid containing the cDNA clone coding for the small-subunit (18S) rRNA (10) which was kindly provided by A. Geber.
Enzyme assay. Laccase activity was measured with epinephrine as a substrate by a method similar to that of Polachek et al. (32). Enzyme solutions (10 ,ul) were incubated in 1 ml of 20 mM sodium phosphate, pH 6.5, with 10 mM epinephrine at 37°C for 30 min, and the A475 was recorded. One unit was defined as 0.001 absorbance unit in 30 min of assay. Enzyme activity for various substrates at 1 mM in 20 mM sodium phosphate, pH 6.5, was also assayed by measuring oxygen consumption (YSI 5300 oxygen monitor; Yellow Springs Instrument Co., Yellow Springs, Ohio). Oxygen measurements were calibrated with standard solutions of hydrogen peroxide and catalase according to the methods in reference 50. The latter method was also used to assess for the presence of hydrogen peroxide after completion of each assay by adding 500 U of catalase and looking for the rapid production of molecular oxygen.
Enzyme purification. C. neoformans ATCC 34873 was grown overnight in asparagine salts with glucose and then transferred to asparagine salts without glucose for 5.5 h to derepress the enzyme according to the method of Polachek et al. (32) as modified by Ikeda and Jacobson (14). In a typical preparation, 20 g of cells was harvested, washed in 0.02 M sodium citrate buffer, pH 3.8, containing 1 mM phenylmethanesulfonyl fluoride, and broken by mechanical shaking in a Braun homogenizer (B. Braun Biotech, Allentown, Pa.) with an equal volume of acid-washed glass beads (0.45-mm diameter) at 4°C. Unless otherwise noted, all further procedures were carried out at 4°C. The extract was adjusted to pH 3.8 with 1 N HCl and centrifuged at 5,000 x g for 15 min. The supernatant was then applied to an Affi-Blue gel column (2.5 by 5 cm; Bio-Rad Laboratories, Watford, Hertsfordshire, United Kingdom) equilibrated in 0.02 M sodium citrate buffer, pH 3.8. The enzyme solution was then washed with the pH 3.8 buffer and eluted with 0.02 M sodium citrate buffer, pH 5.0. Fractions containing enzyme activity were then pooled and applied to a PBE 94 chromatofocusing column (1 by 10 cm; Pharmacia-LKB, Uppsala, Sweden) which had been equilibrated in 0.025 M piperazine buffer, pH 5.0. The column was then washed with 120 ml of a 1:10 dilution of Polybuffer 74 (Pharmacia-LKB), pH 4.0. The enzyme was eluted with a solution of 0.12 M NaCl in 0.025 M piperazine buffer, pH 3.6. Fractions containing enzyme activity were dialyzed overnight against 0.020 sodium acetate, pH 6.0, and applied to a Q-300 high-pressure liquid chromatography (HPLC) ion-exchange column (0.46 by 20 cm; Synchrom, Lafayette, Ind.) previously equilibrated in 0.020 M sodium acetate, pH 6.0. The column was then eluted with a gradient of 0.020 to 0.35 M sodium acetate, pH 6.0, for 20 min and then in 0.35 to 0.70 M sodium acetate over 40 min. Fractions containing enzyme activity were pooled.
Purity of the final product was assessed by subjecting the indicated amount of protein to sodium dodecyl sulfate (SDS)-10% polyacrylamide gel electrophoresis (PAGE) by the procedure of Laemmli (21), as well as isoelectric focusing on Novex slab gels, pH 10 to 3, according to the manufacturer's directions. Protein was assayed by using the Bio-Rad protein assay according to the manufacturer's directions. In order to aid normalization of the copper content of laccase, quantitative amino acid analysis was performed using an ABI model 420 amino acid analyzer and norleucine as an internal standard. Analysis of laccase. N-terminal protein sequencing was performed as described by Roberts et al. (36). Internal protein sequencing was performed as described elsewhere (1). Neutral sugar content was determined by the method of Grollman et al. (12). Alternatively, sugar content was assessed by digestion of purified laccase by endoglucosidase F (Oxford Glycosystems, Oxford, United Kingdom) according to the manufacturer's directions after reduction and alkylation (6) and then subjected to SDS-PAGE. Absorbance spectra were determined on a DU-64 spectophotometer (Beckman Instruments, Columbia, Md.) with 0.2 mg of purified laccase or laccase from Rhus vernicifera (a generous gift from D. R. McMillin) previously dialyzed against 20 mM sodium phosphate, pH 6.5. Metal ion analysis of purified laccase was conducted according to the method of Gacheru et al. (9) using a Perkin-Elmer HGA 2100 analyzer with a D-2 background correction.
Mass spectroscopy of dopamine oxidation product. Laccase (1,000 U) was incubated in the presence of 1 mM dopamine in 1 mM sodium phosphate, pH 6.5, for 10 min. The mixture was then applied to a Synchropak RP-4 HPLC column (250 mm by 4.6 mm [internal diameter]) equilibrated with 0.5% acetic acid and eluted with a linear gradient of 0 to 5% acetonitrile in 0.5% acetic acid over 15 min. The eluted material was monitored for A254, and fractions were collected at a position corresponding to the retention time of standard decarboxy dopachrome prepared by the oxidation of dopamine according to the method of Aroca et al. (2). This material was then analyzed by a JEOL SX102 mass spectrometer fitted with an analytical electrospray source. Ions were sprayed from a 50:50 methanol-2% acetic acid solution and electrosprayed into the mass spectrometer.
RNA isolation, construction of the cDNA library, and Northern analysis. RNA for construction of the cDNA library was extracted from (i) cells grown in asparagine salts as described above for enzyme purification and (ii) cells grown in a liquid Guizotia abyssinica medium. One liter of G. abyssinica medium was made by homogenization of 200 g of the seeds in a Waring blender followed by extraction with 200 ml of boiling water. The extract was filtered through cheesecloth and added to 800 ml of distilled water with 2 g of glucose and 50 mg of chloramphenicol. Cells were grown in both media until laccase activity was first detected. RNA was then isolated according to the method of Chomczynski and Sacchi (5), except that lysis of stationary-phase organisms was achieved by a two-step procedure. Cells were first incubated in 10 mg of mureinase (United States Biochemicals, Cleveland, Ohio) per ml in 0.02 M sodium citrate, pH 5.8, and then washed in 4 M guanidinium thiocyanate-25 mM sodium citrate (pH 7)-0.5% sarcosyl-0.1 M 2-mercaptoethanol. Second, cells were vortexed with an equal volume of 0.45-mm-diameter glass beads for 3 min to release whole-cell RNA. Poly(A) RNA was isolated from whole-cell RNA by standard methods (38). The cDNA library was constructed in Uni-Zap according to the manufacturer's instructions (Stratagene). For mRNA derepression studies, whole-cell RNA was isolated according to the above-described method at the indicated times after resuspension of glucoserepressed cells at a density of 107 cells per ml in asparagine salts without glucose. Northern (RNA) blot analysis was performed by standard methods (38). The blots were probed with an 873-bp AccI fragment of the cDNA vector, pc2l, and total RNA was assessed with a radiolabeled 18S rDNA sequence (10) which exhibits 89% homology to the 18S rRNA sequence of C. neoformans (GenBank accession no. L05428).
Screening the cDNA library. The internal protein sequence EGDAFWLR was used to construct four degenerate oligonucleotides, CCARAANGCRTCXCCYTC, where X is A, T, C, or G for each of the oligonucleotides. End-labeled probes were constructed by standard techniques (38), except each reaction mixture contained only 50 ng of oligonucleotide. The labeled oligonucleotide was then used to probe four identical Southern blots according to the methods in reference 38, and the oligonucleotide giving the clearest band (in which X is A) was used to screen the cDNA library by standard techniques (38). pc2l, pc6l, and pc13 contained the laccase cDNA clone inserted into the EcoRI and XhoI sites of the Bluescript phagemid.
Genomic library, Southern analysis, and pulsed-field electrophoresis. The genomic library prepared from DNA of C. neoformans ATCC 34873 was the generous gift of Jeff Edman. Genomic clones were screened with a radiolabeled 873-bp AccI fragment of the cDNA vector, pc2l. The vectors pG21, pG81, and pG51 consisted of a 4.0-, 4.0-, and 6.0-kb partial Sau 3 DNA fragment, respectively, inserted into the Uni-Zap XR BamHI site. Genomic DNA was isolated by the method of Varma and Kwon-Chung (46). Southern blot hybridizations were performed according to standard methods (38). Pulsedfield electrophoreses and hybridization were performed as described previously (49). DNA sequencing and analysis. Plasmid DNA was sequenced by a dideoxy chain termination method according to the instructions of the Sequenase kit (United States Biochemicals). Sequence analysis was performed by using the Genetics Computer Group sequence analysis software package (7) on a Convex C240 computer maintained by the Division of Computer Research and Technology at the National Institutes of Health.
Nucleotide sequence accession number. The nucleotide sequence data reported in this paper have been submitted to GenBank and have been assigned accession number L22866.

RESULTS
Purification of laccase from C. neoformans. Laccase from C. neoformans was purified with a 25% yield by a combination of cibacron-blue chromatography, chromatofocusing, and ionexchange chromatography on Q-300 as shown in Table 1. Fig.  1 shows the purified enzyme to migrate on SDS-PAGE with an apparent molecular mass of 75 kDa. Treatment with endoglucosidase F results in an enzyme migrating with a molecular mass of 66 kDa. Neutral sugar assay confirmed the glycosylation of the enzyme, yielding 4   Substrate specificity of laccase. The enzyme from C. neoformans resembles other laccases (35) in its ability to oxidize a wide variety of aromatic diphenol and diamino groups in the ortho, meta, and para positions but not monophenolic groups such as in phenol, tyramine, or tyrosine ( Table 2). The spectrum of substrate specificity resembles that of the diphenol oxidase activity obtained from detergent-solubilized cells reported by Polachek et al. (32), although these studies did not show oxidation ofpara-substituted diphenols. However, previous studies used an impure preparation of enzyme and the development of a color reaction to quantitate product formation. In our work no significant hydrogen peroxide could be detected after oxidation of each substrate, again consistent with the oxidative pathway of intact laccases (35).
Mass spectrometry was used to assess for the potential of the cryptococcal laccase to produce a putative melanin intermediate, decarboxy dopachrome. The purified laccase was incubated with dopamine, and the product was purified by reversephase HPLC by collecting fractions comigrating with authentic decarboxy dopachrome, synthesized according to the method of Aroca et al. (2). Mass spectrometry of this fraction showed two peaks at 150 and 172 Da which were not present in identical HPLC fractions (data not shown) derived from dopamine solutions without enzyme. The two masses, 150 and 172 Da, correspond to the masses of the H and Na derivatives of decarboxy dopachrome.
Analysis of copper in laccase of C. neoformans. Two separate analyses of purified laccase yielded 4.2 and 4.0 mol of copper per mol and less than 0.7 mol of iron per mol, similar to other copper-containing laccases (35). To characterize the enzyme further, absorbance spectra were obtained on 0.2 mg (480,000 U/mg) of laccase in sodium phosphate, pH 6.5. As shown in Fig. 2, the enzyme shows a characteristicA610 similar to that of the laccase from R. vernicifera (35), which shows a prominent absorbance maximum in this region. This suggests that laccase contains a type I copper, similar to other laccases (35). Molar absorptivity for the cryptococcal laccase at 610 nm was calculated to be 950 M1cm -, much lower than that of the Rhus laccase (3,000 M`cm-1) under the same conditions. The ) or R vernicifera (---) was dialyzed against 0.02 M sodium phosphate, pH 6.5, prior to absorbance measurements. lower molar absorptivity for the C. neoformans enzyme may be due to a partially inactivated enzyme preparation or to differences in the copper environment. The latter explanation is favored since attempted reconstitution of laccase by dialysis against a fivefold excess of CuCl and CuS04 or oxidation of a possibly reduced Cu(I) with 10 mM ascorbate or potassium ferrocyanide did not result in an increase in specific activity or an increase in the A610 band (data not shown). The enzyme also shows an absorbance shoulder at approximately 320 nm, suggestive of a type III copper as described previously (13).
Isolation and sequence of the C. neoformans laccase (CNLZ4C) gene. In order to aid in the classification of the isolated enzyme from C. neoformans, 3 cDNA clones were isolated from approximately 120,000 screened by use of a degenerate oligonucleotide probe derived from the protein sequence of purified laccase as described in Materials and Methods. Restriction analysis of these clones with XbaI and XhoI showed two of the clones to have inserts of approximately 1.8 to 2.2 kb. Southern analysis of genomic DNA with pc2l as a probe (Fig. 3) indicated that a single gene encodes CNLACL. Chromosomes were separated by pulsed-field electrophoresis, and Southern blots localized the gene to chromosome 5-6 according to the numbering system of Wickes et al. (49). Both strands of one of the cDNA clones, pc2l, were sequenced in their entirety. The open reading frame of CNLAC1, its introns, and nearby flanking regions were sequenced by using both the sense and antisense strands of the genomic clones pG21 and pG51 and are shown in Fig. 4. A single open reading frame on the cDNA begins with the ATG start codon at nucleotide position + 1 and extends to a stop codon TGA at position 2587, encoding a putative polypeptide 624 amino acids in length. The first 20-amino-acid section of this sequence is most likely a leader sequence, based on its relative hydrophobicity and the fact that the N-terminal sequence of the native protein begins at a position corresponding to amino acid Asp-21. The putative leader sequence resembles those of the yeast Saccharomyces cerevisiae (23). It has a tripartite structure having an amino terminus with a positively charged residue (Lys-6), a larger hydrophobic region containing three phenylalanines, and a terminal hydrophilic chain of two serines in positions 19 and 20. The predicted molecular mass of the complete protein beginning at Asp-21 is 66 kDa, corresponding to the value determined from SDS-PAGE of the deglycosylated protein in Fig. 1. Three internal peptide sequences derived from the native protein appear in the derived amino acid sequence, as shown in Fig. 4, and correspond to the predicted amino acid sequences within the error expected of gas-phase protein sequencing techniques (26).
The genomic sequence is notable for 14 introns ranging from 52 to 340 bases long. Only 10 of 14 introns had the typical 5' splice junction sequence, GTNNGY (29), with 2 others showing variation in the fifth and sixth nucleotides. Introns I and II showed almost no homology to the consensus sequence. All 5' splice junctions were followed by a pyrimidine-rich stretch as in the URA5 (8) and DHFR (40) genes from C. neoformans. A YAG 3' splice junction could be located in 11 introns, but no consensus sequence was present at the end of introns I, II, and IX. A canonical TATA box was present at position -539, and a potential CAAT box was found at position -503. In the 3' flanking region no canonical polyadenylation signal (AATAAA) similar to that of S. cerevisiae could be found. A related sequence, AACAAA, which may represent a polyadenylation signal in C. neoformans was observed at position 2854. This sequence is also present in the DHFR (40) and TRP1 (30) genes but is not present in the URA5 gene (8) of C. neoformans. Beyond the poly(A) site, no terminal signal similar to that of S. cerevisiae (51) could be identified.
Comparison of the amino acid sequence of laccase from C. neoformans with that of other blue-copper proteins. While comparison of either the nucleotide or amino acid sequence of CNLACJ with other nucleotide or amino acid sequences in GenBank did not yield any striking overall homology, the strongly conserved copper-binding region of the blue-copper oxidases of eukaryotic species (28) was used to compare the potential for copper binding in laccase from C. neoformans. As shown in Fig. 5, the C. neofornans enzyme shows strong identity to amino acids involved in all three copper-binding sites with the exception of a His->Asn replacement at position 487. This was effected by a single substitution of adenine for cytidine at position 2066. Since this is an important replacement, the sequence was confirmed in this area by sequencing two genomic clones in both directions (pG21 and pG81) and all three cDNA clones (pc21, pc6l, and pc13) in both directions.
Derepression of the CNL4CI transcript by the absence of glucose. As shown in Fig. 6, growth in asparagine salts in the VOL. 176, 1994   (11), and Coriolus hirsutus (16); ascorbate oxidase from zucchini (A ox.) (25) updated using the Brookhaven protein data base; and human ceruloplasmin (H cp.) (37). Numbers on the left of each sequence represent the positions of the amino acid residues of the proteins. Identical amino acids are boxed. Potential coordination sites for the three different types of copper ions are indicated by 1*, 2*, and 3*. presence of glucose represses both enzymatic activity and the 2.2-kb CNLACJ transcript. Transfer of cells to asparagine salts in the absence of glucose resulted in a time-dependent increase in diphenol oxidase activity with a plateau in activity beginning at 3 h, preceded by a peak in the CNLACI transcript at 2 h as measured by densitometry of the probed bands (Fig. 6). Quantities of CNLAC1 transcripts were normalized for the amount of total RNA in each lane by using a radiolabeled 18S rDNA Candida albicans probe showing 89% homology with the corresponding sequence in C. neoformans.

DISCUSSION
The present paper describes the purification of the diphenol oxidase of C. neoformans and cloning of its gene, CNLAC1. The enzyme is capable of oxidizing a wide variety of aromatic diphenolic, aromatic diamino, and catecholamine substrates. This varied substrate specificity, along with an inability to oxidize monophenolic substrates such as tyrosine or tyramine, is consistent with the enzyme being a copper-containing laccase. Polachek et al. (32) first characterized the activity of a detergent-solubilized diphenol oxidase having a similar spectrum of activity but believed the enzyme to be an ironcontaining enzyme, on the basis of inhibitor studies. In the present study, copper but not iron was found in the purified enzyme. Absorbance measurements showed an A610 consistent with a type I copper and an absorbance shoulder at 320 nm, consistent with a type III copper, as previously described for other laccases (35). On the basis of the cDNA and genomic sequences of the cloned gene, the C. neoformans laccase was also found to contain four homologous regions common to multicopper oxidases, including laccases (11,16). Peculiar to the C. neoformans enzyme is the apparent substitution of Asn for His at position 487. This amino acid substitution was the product of a single nucleotide substitution and was found in all cDNA and genomic clones. Computer modeling of the structure of the blue oxidases has been based on ceruloplasminplastocyanin homology (37) and has been extended to cucumber ascorbate oxidase (28) and laccase (16). This work suggests that position 487 of the cryptococcal enzyme is a binding site of one of the type III coppers. Since the ,B-carboxamido group of Asn would not be expected to form a coordinate covalent bond to copper as tight as that of the ring nitrogen of His, it is expected that the function of the type III coppers in the laccase of C. neoformans would be compromised by such a substitution. However, the enzyme shows anA320 shoulder common to type III laccase coppers. In addition, hydrogen peroxide was not produced by the cryptococcal enzyme during enzymatic catalysis. Production of hydrogen peroxide is a characteristic of laccases with experimentally defective type III coppers. Thus, type III copper in cryptococcal laccase appears to function normally, without the participation of a His at position 487. X-ray crystallography data from another blue oxidase, zucchini ascorbate oxidase, also suggest that its corresponding His-452 (Fig. 5) is nonfunctional, since it is positioned outside the copper environment (25). While definitive proof of the replacement of Asn for His for the cryptococcal enzyme requires protein sequencing, it is less likely that alternate codon usage would produce a histidine at that position, since the nucleotide alteration is in the first position of the Asn codon and represents a transversion rather than an inversion.  6. Transcriptional derepression of CNLAC1. Northern blot hybridization of total RNA isolated from C. neoformans grown in asparagine salts with glucose (lane 0) or 2 h after transfer to asparagine salts without glucose (lane 2). Hybridization was with either radiolabeled laccase cDNA (A) or a radiolabeled 18S rDNA probe (B). (C) Cells were transferred from asparagine salts with glucose to identical medium without glucose at the indicated times. Ratios of laccase mRNA to 18S rRNA as measured by densitometery and expressed as percentages of maximal laccase mRNA (A) were calculated from Northern blots of total RNA. Cell extracts were assayed for enzyme activity and protein concentration, and results are expressed as specific activity (U) at the indicated times.
Analysis of the C. neoformans laccase shows the enzyme to be a glycosylated enzyme, both by direct neutral sugar measurements and by a shift in mobility to a lower apparent molecular mass on SDS-PAGE after endoglucosidase F treatment. This substantiates the findings of Ikeda and Jacobson (14), who found that the mobility of diphenol oxidase activity on native gels correlates with binding of such activity to concanavalin A-Sepharose. Such different mobilities in enzyme activity found by Ikeda and Jacobson may be wholly due to posttranslational processing such as glycosylation in strain ATCC 34873, as the present work could demonstrate only a single gene copy on Southern digests and a single size of RNA transcript during derepression of enzyme activity.
Melanin has been demonstrated in the walls of C. neoformans cells growing in mouse brain by Masson-Fontana stain (18); however, the exact chemical structure has yet to be elucidated. Catecholamines are thought to be precursor molecules in cryptococcal melanin formation in the brain in part because of their abundant supply within the dopaminergic pathways. In addition, catecholamines such as epinephrine protect against oxidants for diphenol oxidase-positive but not diphenol oxidase-negative strains of C. neoformans (33). The present study shows that laccase from C. neoformans is capable of oxidizing dopamine to a putative melanin intermediate, decarboxy dopachrome, by mass spectrometry. Such reactive intermediates have the ability to tautomerize to compounds such as 5,6-dihydroxyindole which can then nonenzymatically undergo polymerization to eumelanins (see review in reference 34).
Synthesis of CNLAC1 transcripts and laccase enzymatic activity in C. neoformans was found to be glucose repressed. This agrees with previous work showing that pigment production (27) and diphenol oxidase activity (32) were glucose repressed. Notable was the absence of a requirement for substrate induction in this serotype D organism. This may be a strain-related phenomenon, as substrates such as dopa have been noted to be required for induction of diphenol oxidase activity by some serotype C organisms (27).
CNLAC1 is the largest gene with the most numerous introns thus far cloned from C. neoformans. The large number of introns resembles the mushroom Coriolus hirsutus (16) and the Agaricus bisporus (31) laccase genes, which contain 10 and 14 introns, respectively. This similarity between C. neoformans and these two organisms has been reflected in the phylogenetic trees constructed on the basis of 18S rRNA gene sequences (44). In both trees, C. neofonnans showed a close relationship with mushrooms. An apparently distinguishing feature of cryptococcal genes (8,30,40) was also exhibited by CNLACl in that it does not contain typical polyadenylation or termination signals. This is in contrast to highly conserved polyadenylation and termination sequences of S. cerevisiae (51). In addition, much less conservation of sequence at the putative 5' and 3' splice junction was evident than is present in the ascomycete S. cerevisiae (22,24). Such a lack of conservation suggests promiscuity in the respective binding proteins. Other data suggesting promiscuity in the telomerase of this organism previously (45) suggest that this may be a common property of DNAbinding proteins of C. neoformans.