To assess the relationship between melanin production by Cryptococcus neoformans and virulence on a molecular basis, we asked: (a) is CNLAC1, the laccase structural gene of C. neoformans, expressed in vivo?; (b) can mouse virulence be restored to cnlac1 (Mel-) mutants by complementation with CNLAC1?; and (c) will targeted gene deletion of CNLAC1 decrease virulence for mice? Melanin is produced when cryptococcal laccase catalyzes the oxidation of certain aromatic compounds, including L-dopa, to quinones, which then polymerize to melanin. To assess CNLAC1 transcription, RNA was extracted from C. neoformans in cerebrospinal fluid of infected rabbits. Reverse transcriptase-polymerase chain reaction detected CNLAC1 transcript, indicating that laccase may be produced in the infected host. To assess the effect of CNLAC1 deletion on virulence, a Mel- mutant (10S) was obtained by disruption of the 5' end of the gene. After multiple backcrosses with a parental strain to remove unintended genetic defects introduced by the transformation process, a Mel- progeny was tested and found to be much less virulent for mice than a Mel+ progeny. Another Mel- strain (mel2), obtained from J.C. Edman (University of California at San Francisco, CA), produced CNLAC1 transcript but no detectable melanin. Characterization of this mutant revealed a base substitution in CNLAC1 that changed a histidine to tyrosine in a putative copper-binding site. When this base change was introduced into CNLAC1 by site-directed mutagenesis, it no longer transformed mel2 to Mel+, indicating the importance of this histidine in laccase activity. Complementation of a mel2-derived mutant with CNLAC1 restored the Mel+ phenotype and increased virulence. These results support the concept that the CNLAC1 gene product has a role in virulence.