Isolation and Characterization of the Cryptococcus neoformans MATa Pheromone Gene

Corresponding author: Brian L. Wickes, Mail Code 7758, University of Texas Health Science Center, 7703 Floyd Curl Dr., San Antonio, TX 78229-3900., wickes{at}uthscsa.edu (E-mail)

Communicating editor: J. RINE

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Cryptococcus neoformans is a heterothallic basidiomycete with two mating types, MATa and MAT. The mating pathway of this fungus has a number of conserved genes, including a MAT-specific pheromone (MF1). A modified differential display strategy was used to identify a gene encoding the MATa pheromone. The gene, designated MFa1, is 42 amino acids in length and contains a conserved farnesylation motif. MFa1 is present in three linked copies that span a 20-kb fragment of MATa-specific DNA and maps to the MAT-containing chromosome. Transformation studies showed that MFa1 induced filament formation only in MAT cells, demonstrating that MFa1 is functionally conserved. Sequence analysis of the predicted Mfa1 and Mf1 proteins revealed that, in contrast to other fungi such as Saccharomyces cerevisiae, the C. neoformans pheromone genes are structurally and functionally conserved. However, unlike the MF1 gene, which is found in MAT strains of both varieties of C. neoformans, MFa1 is specific for the neoformans variety of C. neoformans.

CRYPTOCOCCUS neoformans is an important human fungal pathogen that can cause serious and sometimes life-threatening infections in humans and animals. Isolates of the fungus can be divided into two varieties, neoformans (serotypes A and D) and gattii (serotypes B and C), although there is increasing evidence that serotype A may, in fact, constitute a separate variety (var. grubii; FRANZOT et al. 1999 ). This classification is determined by differences in polysaccharide structure, which can be distinguished by serotyping using rabbit sera or a commercially available kit (WILSON et al. 1968 ). The varietal status was established after it was found that the teleomorphs of C. neoformans (Filobasidiella neoformans) and C. gattii (F. bacillispora) were interfertile and intermediate in DNA relatedness (55–63% reassociation; AULAKH et al. 1981 ). For all four serotypes it is assumed that infection occurs by inhalation and initiates in the lungs. From the lungs, the organism can spread to a variety of different body sites, with the preferred site being the brain, where it causes meningitis. Cryptococcal meningitis is especially dangerous for immunosuppressed patients, particularly those with AIDS. The incidence of cryptococcosis in these patients was estimated to be almost 10% at the peak of the AIDS epidemic (CHUCK and SANDE 1989 ). Although the incidence has decreased with a reduction in AIDS cases, the disease is incurable and requires life-long antifungal prophylaxis (MCNEIL and KAN 1995 ). More importantly, an increasing frequency of both HIV drug resistance (MATSUSHITA 2000 ) and new infections (PIOT et al. 2001 ) suggests that morbidity and mortality due to cryptococcosis could increase again in the near future.

C. neoformans is one of the best models for studying major systemic mycotic agents. It is normally haploid and has a single locus, bipolar mating system with two mating types, MATa and MAT. The ability to genetically manipulate this fungus has contributed to the characterization of a number of virulence factors. These include the ability to grow at 37° (KWON-CHUNG et al. 1982B ), capsule expression (KWON-CHUNG and RHODES 1986 ), melanin production (KWON-CHUNG and RHODES 1986 ), mannitol production (CHATURVEDI et al. 1996 ), and phospholipase activity (S. C. CHEN et al. 1997 ). One of the most complex characteristics associated with virulence is mating type. Analysis of congenic MAT and MATa strains revealed -cells to be more virulent than a-cells (KWON-CHUNG et al. 1992 ). Additionally, clinical isolates are almost all MAT in mating type (KWON-CHUNG and BENNETT 1978 ). Environmental isolates also show a severe bias of - over a-cell types (KWON-CHUNG and BENNETT 1978 ), which may suggest that frequency of exposure plays a role in the mating-type bias. Interestingly, in contrast to the other three serotypes for which both mating types have been recovered, a true haploid MATa serotype A isolate that is fertile has yet to be conclusively identified. Recent evidence suggests that these isolates may exist (LENGELER et al. 2000 ); however, their ecological niche has yet to be identified. The rarity of serotype A MATa isolates is significant because serotype A isolates cause >90% of all human infections (MITCHELL and PERFECT 1995 ).

Preliminary studies indicate that several genes of the Saccharomyces cerevisiae pheromone and pseudohyphal response pathways are conserved in C. neoformans. These genes include homologs of GPA1 (ALSPAUGH et al. 1997 ), STE12 (WICKES et al. 1997 ), GPB1 (WANG et al. 2000 ), STE11 (CLARKE et al. 2001 ), STE20 (LENGELER et al. 2000 ), FUS3 (DAVIDSON et al. 2000 ), and the MAT pheromone gene (MOORE and EDMAN 1993 ). However, a number of the pheromone response homologs, including STE11, STE12, and STE20, are mating-type specific with both a MATa and MAT allele and are physically located within their respective mating-type loci (WICKES et al. 1997 ; LENGELER et al. 2000 ; CHANG et al. 2001 ; CLARKE et al. 2001 ). These alleles display conserved domains found in other fungal homologs, such as homeodomains or catalytic domains, but differ by having divergent regulatory regions. In spite of the existence of a-specific alleles, less is known about the MATa mating type. The lack of information about this mating type reflects the reduced role of a-cells in virulence and the difficulty in isolating this cell type from nature. Recent studies that have included this cell type suggest that a-cells may signal -cells, which in turn are capable of responding with a hyphal phenotype (CHANG et al. 2000 ; WANG et al. 2000 ; CLARKE et al. 2001 ). This phenotype, called monokaryotic fruiting (WICKES et al. 1996 ), is characterized by the production of hyphae by haploid cells. These hyphae display clamp connections typical of dikaryotic hyphae, which are produced during mating, but the connections are unfused. The hyphae also display a single nucleus per hyphal compartment (hence they are monokaryotic) in contrast to the paired nuclei found in dikaryotic hyphae. Most importantly, monokaryotic hyphae can produce basidia and spores (fruiting) that are indistinguishable from sexually produced spores, except that they are all the same mating type. Monokaryotic fruiting is induced by starvation and is analogous to the S. cerevisiae pseudohyphal response. It also can be induced by coculturing a-cells in close proximity (CHANG et al. 2000 ; WANG et al. 2000 ; CLARKE et al. 2001 ), which suggests the presence of a signaling molecule. An obvious candidate for a signaling molecule is a pheromone, which -cells have already been shown to produce (MOORE and EDMAN 1993 ).

The MAT pheromone gene, designated MF1, was previously identified by MOORE and EDMAN 1993 and shown to encode a small 38-amino-acid peptide that terminates with the amino acid sequence CVIA. The presence of the CVIA sequence and small size of MF1 suggested that the C. neoformans -pheromone was similar to other fungal lipopeptide mating factors (KAMIYA et al. 1978 ; SAKAGAMI et al. 1979 ; BRAKE et al. 1985 ; BOLKER et al. 1992 ; DAVEY 1992 ; WENDLAND et al. 1995 ; ZHANG et al. 1998 ; ANDERSON et al. 1999 ; SHEN et al. 1999 ). These pheromones are characteristically short peptides that terminate with a CAAX motif, where A is an aliphatic amino acid and X can be cysteine, serine, methionine, glutamine, or alanine (CALDWELL et al. 1995 ). This motif is found in a number of membrane-anchored proteins, most notably the RAS superfamily of GTP-binding proteins, where it serves as a signal for prenylation and carboxy methyl esterification (SCHAFER and RINE 1992 ). For S. cerevisiae, two similar and functionally redundant genes, MFA1 and MFA2, encode a-factor (BRAKE et al. 1985 ). MFA1 and MFA2 encode precursor peptides of 36 and 38 amino acids long, respectively, which are eventually processed to a mature form of a-factor (12 amino acids long) that is both farnesylated and carboxylmethylated (ANDEREGG et al. 1988 ) and then secreted via a nonclassical export mechanism (KUCHLER et al. 1989 ; MICHAELIS 1993 ).

In this study we report the isolation of the C. neoformans MATa pheromone using a modified differential display PCR (DD-PCR) strategy. Using this approach, a small open reading frame (ORF) was identified that was predicted to encode a 42-amino-acid peptide. The gene, designated MFa1, was found to be MATa specific and present in multiple copies. Similar to MF1, MFa1 was expressed only under starvation conditions or during mating and is linked to the chromosome that contains the mating loci. MFa1 induced filament formation in -cells but not in a-cells, which is consistent and complementary to previous studies of the MAT pheromone gene (MOORE and EDMAN 1993 ; WICKES and EDMAN 1995 ). However, although structurally and functionally conserved when compared to MF1, MFa1 was found to be specific for the neoformans variety of C. neoformans.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Media, strains, and transformations:
YPD agar consisted of 1% yeast extract, 2% peptone, 2% dextrose, and 2% agar. Synthetic dextrose medium contained 0.67% yeast nitrogen base without amino acids (Difco, Detroit), 2% dextrose, and 2% agar and was supplemented with adenine or uracil as needed. 5-Fluoroorotic acid (5-FOA; U.S. Biologicals, Swampscott, MA) agar was prepared as previously described (BOEKE et al. 1987 ). V8 agar (pH 7.2) contained 5% V8 juice, 0.5 g/liter KH₂PO₄, 4% agar, 0.75% dextrose or galactose as required, and any nutritional supplements (KWON-CHUNG et al. 1982A ).

C. neoformans strains WSA-20 (MATa), WSA-21 (MAT), WSA-2 (MAT ade2), WSA-43 (MAT ura5), and WSA-34 (MATa ura5) have been previously characterized (EDMAN and KWON-CHUNG 1990 ; CLARKE et al. 2001 ). Strain WSA-258 is WSA-43 transformed with plasmid pC5. WSA-285 is WSA-43 transformed with pMFa-1. Strains WSA-279 and WSA-282 are WSA-34 transformed with pC5 and pMFa-1, respectively. Strain WSA-442 (MAT ade2 ura5::GAL7_p::MFa1::GAL7_t-ADE2) is WSA-2 transformed with the insert of pUMFa-GA, which is integrated at the URA5 locus.

Transformations of C. neoformans were performed as previously described (WICKES and EDMAN 1994 ). Bacterial transformations were performed in Escherichia coli XL1-Blue cells (Stratagene, La Jolla, CA) using a BTX electroporater (Genetronics, San Diego) according to the manufacturer's instructions.

Plasmids:
Plasmids pC5 and pA11 are the C. neoformans URA5 and ADE2 genes, respectively, in pBluescript SK⁺ (Stratagene). pMFa-3' is a cDNA clone of the 3' noncoding region of the MFa1 pheromone gene isolated by TA cloning of the MFa3-HT₉G DD-PCR product into pCR2.1 (Invitrogen, Carlsbad, CA). pRT-67 is the 5' rapid amplification of cDNA ends (RACE) product from MFa1 cloned into pCR2.1. pMFa-RA2 is a complete cDNA clone of MFa1 in pCR2.1. pMFa-1 is the genomic clone of MFa1 recovered as a 6.5-kb fragment from a Tsp 509I partial digest genomic DNA library cloned into the EcoRI site of pC5. pMFa-NB was constructed by PCR amplification of the MFa1 coding region and ligation into pCR2.1. The ATG start codon was modified to an NdeI site with the 5' oligonucleotide (5'-AACATATGGACGCCTTCACTGCTATCTTC-3'), and a BglII site was engineered immediately 3' to the stop codon with the 3' oligonucleotide (5'-AAAGATCTTTAAGCAATAACGCAAGAGTAAGTCGG-3'). The NdeI-BglII fragment containing the MFa1 coding sequence (CDS) from pMFa-NB was cloned into pBS1 between the C. neoformans GAL7 promoter and GAL7 terminator (WICKES and EDMAN 1995 ) using the same enzymes to create pMFa-G in a pBluescript SK+ backbone. pMFa-GA was constructed by ligating the 984-bp ClaI-KpnI fragment of pMFa-G, which contains the GAL7 promoter-MFa1-GAL7 terminator cassette, into the same vector sites of pA11. The entire cassette (GAL7p::MFa1::GAL7_t-ADE2) was then used to disrupt URA5 in pC5 by removing the cassette from pMFa-GA as an EcoRI-KpnI fragment, blunt ending with T4 DNA polymerase (New England Biolabs, Beverly, MA), and ligating into pC5 digested with SacI (blunt ended) and StuI. The plasmid, pUMFa-GA, contained a 194-bp deletion of the URA5 gene replaced with the C. neoformans MFa1 coding region fused between the GAL7 promoter and terminator, with ADE2 as the selectable marker.

Preparation and analysis of nucleic acids:
A bead-beating method was used for rapid isolation of DNA from small amounts of C. neoformans cells as described (CLARKE et al. 2001 ). Large amounts of DNA were prepared using the spheroplasting method as described (CLARKE et al. 2001 ). To prepare DNA from single progeny basidiospores for PCR screening, a boiling miniprep method was utilized. Briefly, cells were grown for 16 hr at 30° and 275 rpm in 1 ml of YPD broth. The culture was transferred to a microfuge tube, washed twice in distilled water, boiled for 15 min, and then pelleted by centrifugation. The supernatant was removed to a new tube and stored at -20° until use. PCR reactions were performed with 2 µl of supernatant in a 50-µl reaction. Total RNA was isolated and prepared as described (WICKES and EDMAN 1995 ). The QIAGEN (Valencia, CA) poly(A)⁺ RNA isolation kit was used to isolate poly(A)⁺ RNA according to the manufacturer's instructions. Blotting was performed with a positively charged nylon membrane (Hybond-N⁺, Amersham, Piscataway, NJ) and hybridizations were performed in Rapid Hyb buffer (Amersham) according to the manufacturer's instructions. Nucleic acid probes were labeled by random priming (High Prime, Roche Diagnostics, Indianapolis) with [-³²P]dCTP (NEN Life Science Products, Boston). MFa1 and MF1 probes were prepared from PCR products as described in the section on linkage analysis (below). A 1.3-kb actin probe was amplified from WSA-21 using (forward) 5'-ATGGAAGAAGAAGGTACGTTC-3' and (reverse) 5'-TTAGAAACACTTTCGGTGGAC-3' as primers, an annealing temperature of 56°, and an extension time of 60 sec under standard thermocycler conditions (see below). An ADE2 probe was prepared using 5'-AAGGTCTTTGTGAAGTCCAGCCCAG-3 as a forward primer and 5'-AGCACCAGGAACAGTGAGAGCATTG-3' as a reverse primer to amplify an 800-bp fragment using an annealing temperature of 60° and an extension time of 1 min under standard conditions (see below). Pheromone copy number was determined using various restriction enzymes to digest 5.0 µg of genomic DNA in single enzyme digests, which were electrophoresed in an 0.8% gel and then blotted and hybridized as described above.

PCR protocols:
Standard PCR reactions were performed in 50 µl with 2.5 units Taq DNA polymerase (GIBCO/BRL, Grand Island, NY) using a thermocycler (MJ Research, Waltham, MA) with the following program: an initial cycle of 94° for 2 min, primer-specific annealing temperature for 30 sec, 72° extension temperature for 1 min/kb of target length, followed by 29 identical cycles (except the 94° cycle was reduced to 20 sec), and a final 5-min extension at 72°. Reactions were held at 4° until analysis by agarose gel electrophoresis. The Expand High Fidelity PCR system (Roche Diagnostics) was used according to the manufacturer's instructions to amplify templates for sequencing. The Expand Long Template PCR system (Roche Diagnostics) was used to amplify templates longer than 6 kb.

Identification of MFa genes:
A series of oligonucleotide primers (Table 1) were designed for use in a modified DD-PCR reaction (LIANG and PARDEE 1992 ). The upstream primers were degenerate 11-mers of the amino acid sequence CVIA, with only the first two bases of the alanine codon included. Six primers were designed with a single degenerate base instead of a single primer with multiple degenerate bases due to the short primer length and necessarily low annealing temperature. The downstream primers consisted of a HindIII site followed by T₉ ending in an A, G, or C. The design of this primer was based on the one-base anchor primer described by LIANG et al. 1994 . The single additional base (A, G, or C) anchors the primer and eliminates random binding within a potentially long poly(A) tract. A HindIII site is added to contribute 2Ts to the poly(T) region, while also serving as a sequencing landmark. This primer design strategy kept the number of products per reaction low.

Template cDNA was prepared from total RNA (2.5 µg) using Superscript II reverse transcriptase (GIBCO/BRL) according to the manufacturer's instructions and each of the three downstream anchor primers in separate reactions. The cDNAs were normalized to 10 ng/µl after reverse transcription. Ten microliters were then used in a PCR reaction with the corresponding anchor primer and one of the six 11-mer primers. Standard PCR conditions were employed with an annealing temperature of 37° and an extension time of 30 sec, except that the annealing time was increased to 1 min/cycle and the amount of Taq polymerase was increased to 5.0 units per reaction. PCR products were electrophoresed on a 2.5% NuSieve agarose gel (FMC, Rockland, ME) and bands <300 bp in size that appeared to be differentially expressed in WSA-34 grown on V8 agar and in the WSA-2/WSA-34 cross, purified, and subcloned into pCR2.1 for sequence analysis. The sequence, which represented the 3' untranslated region (UTR) of the candidate MFa1 gene, was used to recover the upstream region of the gene by RACE. The RACE protocol was employed as described (ZHANG and FROHMAN 1997 ) using 5'-TGGTGTCCTCATTCTCTCTCTCTG-3' as the nested gene-specific primer. The MFa1 genomic clone was recovered from a genomic library of WSA-20 DNA prepared in pBluescript SK+. Five clones were partially sequenced and all were found to be identical, indicating that all were MFa1. Therefore, alternate methods were used to recover MFa2 and MFa3.

MFa2 was cloned using a modified random amplification of genomic ends (RAGE) protocol (MIZOBUCHI and FROHMAN 1993 ). Five-microgram aliquots of WSA-20 genomic DNA were digested to completion with EcoRV, SphI, ClaI, SspI, PstI, NsiI, KpnI, or FspI (New England Biolabs), purified by glassmilk (Elu-Quick, Schleicher & Schuell, Keene, NH), and poly(C) tailed with terminal transferase (GIBCO/BRL). Ten microliters (100 ng) were used as template for high-fidelity PCR with an MFa-specific primer 5'-AGATGGACTTCGCCAAAGATGTGC-3' and one of four RAGE primers containing the sequence NG₁₂HD, where N is a NotI site, H is A, T, or C, and D is T, G, or A. Primer RAGE-1 (HD = AT) was used with EcoRV, SphI, ClaI, and SspI. RAGE-2 (HD = TG) was used with PstI and NsiI. RAGE-3 (HD = TA) and RAGE-4 (HD = CA) were used with KpnI and FspI, respectively. An annealing temperature of 60° and an extension time of 3 min were used for all PCR reactions. Amplicons were TA cloned into pCR2.1 and inserts were sequenced for comparison to MFa1. A 616-bp clone from the FspI digest was identified that contained a divergent 5' upstream region, but a coding region identical to MFa1. Therefore, this sequence represented the second copy of the MATa pheromone and was designated MFa2. The unique 5' upstream sequence was used to design an MFa2-specific primer (5'-GTAAAGTATTCTCGCCGGCATTC-3') that was used to clone the 3' downstream region by the same method.

Attempts to recover MFa3 using RAGE yielded only MFa1 and MFa2 clones. Therefore, high-fidelity PCR was used with a single primer (5'-GAAGATAGCAGTGAAGGCGTCC-3') to amplify the 8.0-kb region between MFa1 and MFa3. The PCR product was blunt ended with T7 polymerase (New England Biolabs), digested with XbaI, and cloned as a 2.6-kb fragment into pBluescript SK⁺ digested with XbaI and EcoRV. XbaI was arbitrarily selected because it cleaved the 8.0-kb PCR product into easily distinguishable, unequally sized fragments, both of which were larger than 2.0 kb. Sequencing from the XbaI end revealed a divergent 5' region that differed from MFa1 and MFa2. A 1.5-kb XbaI-SphI fragment representing this region was then used to screen the MATa genomic library to obtain MFa3. MFa3 was identified by sequencing the 5' end of multiple clones to identify a clone that diverged from both MFa1 and MFa2. Continued sequencing of the clone provided 3' flanking sequence of MFa3. The three MATa pheromone gene sequences were deposited in GenBank under accession nos. AF305931 (MFa1), AF339448 (MFa2), and AF339449 (MFa3).

Physical mapping of MFa1, MFa2, and MFa3:
The distance between the three pheromone genes, as well as their orientation, was determined by long-distance PCR. Oligonucleotide primers 5'-CCCGACTTACTCTTGCGTTATTGC-3' (+) and 5'-GAAGGTAGCAGTGAAGGCGTCC-3' (-) were designed from the MFa coding sequence (CDS) as forward (+) or reverse (-) primers. All possible primer combinations (+/+, +/-, and -/-) were used in the Expand Long Template PCR system with WSA-20 genomic DNA (100 ng) as template, a 60° annealing temperature, and 12-min extension time. PCR products were separated in a 0.6% agarose gel. To confirm the results of this experiment, the long-distance PCR reaction was repeated using gene-specific primers. The primers were MFa1 reverse 5'-AGTAGTAGGAGGGTGACAGAAGC-3' and MFa3 reverse 5'-GGGAAATCAGGGGCATGTGAAC-3'. This product spans the gap between MFa1 and MFa3. The second reaction primers were MFa3 forward 5'-GGTGAAAGAGGTGGTTTTGCCTG-3' and MFa2 reverse 5'-ATTATTTGTTAACGGGTACCAGTACAGTATC-3'. This product spans the gap between MFa2 and MFa3. PCR conditions were the same as for the (+) and (-) primers.

Linkage analysis:
Linkage of MFa1 to the MATa mating type was confirmed by several methods. First, single basidiospores from a cross between WSA-2 and WSA-34 were isolated by micromanipulation and tested for mating type by backcrossing to each parent in a qualitative mating assay (CLARKE et al. 2001 ). DNA from 10 MATa and 10 MAT progeny was recovered by boiling miniprep and screened for MFa1 by PCR with primers 5'-AACACCAACAACCCGCTACAATGG-3' and 5'-AGATGGACTTCGCCAAAGATGTGC-3', which amplify a 212-bp MATa-specific fragment. As controls, progeny were screened with MF1 primers (5'-AGCAACCAAGGATCGCTACTCCAC-3' and 5'-ACGCTTCGTTACCATCTGTCCGAC-3'), which amplify a 327-bp fragment from MAT cells and URA5 primers. The URA5 primers, URA5.F (5'-TCGAACATGGCGTGCTTCTTTTC-3') and URA5.R (5'-TGACCTCTTGCAGCTCCTTTTCCC), amplify a 689-bp fragment from both mating types. All PCR reactions were carried out with an annealing temperature of 60° and an extension time of 1 min. PCR products were separated on either a 1.5% (MFa1 and MF1) or a 0.8% (URA5) agarose gel. A second method used Southern hybridization of MFa1 to blots of WSA-20 and WSA-21 karyotypes prepared as previously described (WICKES et al. 1994 ). This method was used to determine if MFa1 was mating-type specific and linked to the chromosome containing the MAT locus. The third linkage method compared the hybridization patterns of MFa1 and MF1 to a panel of isolates representing all four serotypes of C. neoformans.

Filament assays:
A filament induction assay was used as a functional screen for pheromone gene activity as previously described for MF1 (WICKES and EDMAN 1995 ). Strains containing episomal pMFa-1 (confirmed by rapid 5-FOA-induced segregation) were streaked on V8 medium, incubated at 25° for 24–48 hr, and examined for the presence of filamentous projections as described (WICKES and EDMAN 1995 ). Strains containing pC5 alone were used as negative controls. Strain WSA-442 was used to assay for galactose-regulated pheromone production on V8 medium containing 0.75% galactose or glucose and supplemented with uracil.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Identification of the MATa pheromone gene 3' untranslated region:
To identify the MATa pheromone gene, we took advantage of a number of factors that could be employed as screening criteria. First, a CAAX-box motif is found at the carboxy terminus of many fungal pheromone genes (Table 2). In S. cerevisiae, the pheromones fall into two separate classes with only one (a-factor) using the CAAX-box motif. Since C. neoformans MATa cells differentiate into a hyphal phenotype when transformed with the CAAX-box-containing MF1 gene (MOORE and EDMAN 1993 ; WICKES and EDMAN 1995 ), it seemed plausible that the MATa pheromone could also fall into this class. Second, a functional pheromone-associated CAAX box could be distinguished from a random CAAX sequence at the DNA level by a stop codon, which should immediately follow the X residue. Additionally, a PCR product that spanned the CAAX box and poly(A)⁺ site would be small if, as predicted, the CAAX motif was located at the end of the gene. Third, unlike S. cerevisiae, C. neoformans does not mate in rich media. Since starvation conditions are required to induce mating, many mating genes, including a putative pheromone, should be inducible and therefore differentially expressed. This difference could be exploited in the primary PCR screen. Fourth, since C. neoformans is heterothallic and does not switch mating types, it seemed highly likely that a MATa pheromone gene would be mating-type specific. Since a congenic pair of isolates has been established (KWON-CHUNG et al. 1992 ), candidate sequences could be tested easily for MATa specificity. Fifth, since virtually all fungal pheromones in the CAAX-box class described to date are <500 bp in size, this characteristic could be used as an additional PCR screening criterion. Finally, since the MAT pheromone had already been identified, our screening strategy could be pilot tested on this gene prior to performing a full MATa pheromone screen.

Pilot cDNA templates for MF1 analysis were prepared as described in MATERIALS AND METHODS. Primer MFa5 in combination with anchor primer HT₉C (Table 1) produced a band 150–200 bp in size from MAT cDNA prepared from cells grown on V8 agar and from the cross (Fig 1A). This band was not present in MATa-specific cDNA from either growth condition, nor from MAT cells grown on YPD. Sequence analysis of the MFa5-HT₉C product showed that this band was identical to the 3' flanking region of the genomic MF1 gene with the exception of a 56-bp intron not previously identified in this gene (Fig 1B). The results of this experiment suggested that this technique would work if, indeed, the MATa pheromone sequence contained CVIA as the CAAX motif. cDNAs prepared in the above experiment were next amplified in all combinations using the primers in Table 1. The combination of five RNA types, three anchor primers, and six degenerate primers required 90 reactions, which were separated on agarose gels immediately after amplification. After excluding bands >300 bp, which would likely be too large to represent a 3' UTR in C. neoformans, six bands <300 bp in length were observed only in lanes that used MATa cDNA as template DNA. Cloning and sequencing of these bands showed a stop codon immediately following the primer-encoded CVIA motif only from the band produced by primers MFa3 and HT₉G (Fig 2). The small size of the band, MATa strain and V8 agar specificity, and predicted DNA sequence were preliminary evidence of a MATa pheromone candidate.

View larger version (51K): In this window In a new window Download PPT slide   Figure 1. Pilot PCR of MF1 3' untranslated region. (A) cDNA templates MAT, MATa, and a cross of both mating types (X) grown on either YPD agar (Y) or V8 agar (V8) were amplified with upstream CVIA primer MFa5 and downstream anchor primer HT9C. The 172-bp MAT-specific V8 bands (indicated by the arrow) were TA cloned and sequenced. (B) Comparison of cDNA to genomic DNA sequence obtained from the above clones. Immediately following the CVIA primer sequence is a stop codon, which would be predicted to occur immediately after a CAAX-class pheromone. An intron was also identified (lowercase letters) based on alignment to the genomic sequence and consensus 5' (GTNNGY) and 3' (YAG) splice sites.

View larger version (38K): In this window In a new window Download PPT slide   Figure 2. Identification of a candidate MATa pheromone sequence by DD-PCR. (A) The bands at 182 bp produced by primers MFa3 and HT9G (indicated by the arrow) were specific for MATa cDNAs under conditions predicted to induce a putative MATa pheromone gene. The MATa-V8 and X-V8 bands, as well as MATa-specific bands from other reactions, were TA cloned and sequenced. (B) Sequence analysis revealed that only the MFa3-HT9G PCR product contained a stop codon immediately following the CVIA sequence.

Isolation of the complete MFa1 gene:
Using 5' RACE, a 264-bp cDNA clone was isolated from MATa RNA prepared from V8 agar-grown cells. The small size of this clone was again consistent with a candidate pheromone gene. The terminal 5' sequence of this clone and the 3' terminal sequence of the initial 3' UTR clone were used to design primers that were used to amplify, clone, and sequence two independent cDNAs. Analysis of the 334-bp cDNA sequences identified an ORF predicted to be 42 amino acids in length with a molecular weight of 4.29 kD. The putative pheromone terminated with an in-frame CVIA motif that matched the sequence of primer MFa3 and was immediately followed by an in-frame stop codon. Sequence alignment of the cDNA with the genomic clone revealed a 97-bp intron located in the 3' UTR region, similar in location to the intron found in MF1 (Fig 3A). Five 10-bp repeats of TTTTGTTCTT were found in the promoter region of the genomic clone. These small repeats are similar to sequences found in MF1 and in the pheromone genes of Ustilago maydis (BOLKER et al. 1992 ). For U. maydis, these elements were shown to be pheromone response elements, which are involved in transcriptional regulation of genes that respond to the opposite mating-type pheromone (URBAN et al. 1996 ). Similar response elements have also been found in the promoters of pheromone-inducible genes in S. cerevisiae (SENGUPTA and COCHRAN 1990 ). On the basis of the preliminary sequence, the putative C. neoformans pheromone gene was named MFa1.

View larger version (62K): In this window In a new window Download PPT slide   Figure 3. Identification of MFa1. (A) MFa1 sequence. The 10-bp repeat is shown in boldface type, TATA is underlined, and predicted transcriptional start site is double underlined. (B) Comparison of MFa1p and MF1p. Boldface letters are conserved amino acids. The box indicates conserved regions that correspond to predicted processing sites.

Alignment of the coding sequences of MFa1 and MF1 showed a number of similarities. The first 8 amino acids and 23 out of 24 nucleotides of the two genes were identical (Fig 3B). On the basis of the predicted protein sequence, post-translational processing of the Mfa1p pheromone precursor is likely to be similar to other fungal pheromones in this class. For S. cerevisiae a-factor, modification of the carboxy terminus proceeds as a well-defined series of events that are required for pheromone maturation. These modifications include farnesylation of the cysteine residue followed by proteolysis of the terminal VIA residues and, finally, methylation of the newly exposed carboxy-terminal cysteine (P. CHEN et al. 1997 ). The predicted Mfa1 protein contains an asparagine residue at position 25, which is 14 amino acids from the carboxy-terminal cysteine. All of the pheromones listed in Table 2 contain an asparagine within 17 amino acids of the carboxy-terminal cysteine. In S. cerevisiae, this asparagine is the cleavage site for removal of the amino terminus portion of the immature pheromone by Axl1p (ADAMES et al. 1995 ). Cleavage of Mfa1p at this position would give a predicted mature pheromone of 14 amino acids. Structurally, therefore, Mfa1p appears to be highly conserved when compared to other fungal pheromones, including the C. neoformans -pheromone.

MFa1 is linked to the MATa mating type:
Linkage of MFa1 to the MATa mating type was confirmed by several methods. First, progeny from a cross of a- and -cells were recovered, scored for mating type, and tested for the presence of MFa1 by PCR (Fig 4A). All progeny spores gave a product with the URA5 primers, regardless of mating type. All MATa progeny were found to give an MFa1 PCR product, indicating that the MFa1 sequence is linked to the MATa locus with a calculated distance of <5.0 MU. Conversely, as expected all MAT progeny gave a product with the MF1-specific primers, confirming that MF1 is linked to the MAT phenotype. No progeny gave products when tested with PCR primers of the opposite mating-type pheromone. Second, hybridization of MFa1 to karyotype blots of the congenic strains WSA-20 and WSA-21 demonstrated that the gene was MATa specific and located on the MAT-containing chromosome (Fig 4B). The MATa-specific hybridization pattern was also confirmed by traditional Southern hybridization of restriction digests of genomic DNA (see below). Third, slot blots were used to determine if the MATa linkage was conserved throughout the species or was serotype specific. An a and strain were included for serotypes B, C, and D; however, since a serotype A MATa strain has yet to be confirmed, two unrelated isolates were used. The results of the MFa1 hybridization differed from the MF1 hybridization. While MF1 hybridized to MAT cells of all serotypes, MFa1 hybridized only to the serotype D MATa strain in this panel (Fig 5), as well as other MATa serotype D strains in our collection (data not shown).

View larger version (52K): In this window In a new window Download PPT slide   Figure 4. Linkage of MFa1 to the MATa mating type. (A) Single basidiospore (Sb) progeny of a cross were tested for mating type by crossing to MATa and MAT tester strains. A total of 10 MATa and 10 MAT were then screened for the presence of MFa1 and MF1 by PCR with pheromone-specific primers. URA5-specific primers were used as controls. (B) Linkage of MFa1 and MF1 in congenic strains to the same 2.5-Mb chromosome that contains the mating-type loci. Actin (ACT) was used as a control hybridization probe.

View larger version (96K): In this window In a new window Download PPT slide   Figure 5. Variety-specific hybridization pattern of MFa1 by slot blot. DNA was prepared from each serotype (A, D, B, C). Both mating types (a or ) were included for each serotype except serotype A, where two independent MAT strains were tested because a serotype A MATa strain has yet to be conclusively identified. Probes were actin (ACT), MFa1, and MF1. MF1 hybridizes to DNA from MAT strains from all four serotypes. MFa1 hybridizes only to MATa DNA from serotype D.

The MATa pheromone gene is present in multiple copies and induced by starvation:
Genomic DNA from WSA-20 and WSA-21 was digested with various restriction enzymes and probed with MFa1 (Fig 6A). The hybridization pattern indicated that there were three MATa-specific copies of the gene per haploid genome. This organization is similar to other fungal pheromones such as S. cerevisiae, where two copies of the a-factor gene are found; Schizosaccharomyces pombe, which contains three copies of the M-factor gene (BRAKE et al. 1985 ; DAVEY 1992 ; KJAERULFF et al. 1994 ); and the basidiomycetous yeast R. toluroides, which also contains three copies of the rhodotorurine A gene (AKADA et al. 1989 ). This pattern is consistent with MF1, which is present in multiple copies (DAVIDSON et al. 2000 ; B. WICKES, unpublished data).

View larger version (31K): In this window In a new window Download PPT slide   Figure 6. Copy number and expression pattern of MFa genes. (A) DNA from WSA-21 () and WSA-20 (a) was digested with BamHI, XbaI, or HindIII, and probed with ADE2 (control) or MFa1. The hybridization pattern is consistent with the presence of three copies of pheromone. (B) Poly(A)+ RNA was hybridized with Actin (control) or MFa1. Results show that MFa1 is induced on V8 agar, which is consistent with conditions that support mating.

To verify the expression pattern of MFa1, poly(A)⁺ RNA was isolated for Northern hybridizations from cells grown on rich and starvation media, as well as from a cross. Mating in C. neoformans occurs in a nitrogen-starved environment, but occurs poorly if at all on rich media (KWON-CHUNG 1976 ). Fig 6B shows clearly that MFa1 is expressed only under starvation conditions in haploid cells or during a mating reaction, which matches the pattern observed in the initial PCR cloning of the MFa1 3' untranslated region. The transcript size was also found to be between 200 and 400 bp in size (data not shown), consistent with the size of the cDNA and small size of other fungal pheromones. The expression pattern of this gene is therefore inducible and is greatest under starvation conditions, which fits the optimum mating conditions for C. neoformans.

The physical distance between the three MFa copies within the MAT locus, as well as the orientation of each copy, was determined using long-distance PCR (Fig 7A). It was found that two copies of the pheromone were in the same orientation, 10.5 kb apart, and that the final copy was 8 kb away and in the opposite orientation (Fig 7B). In this experiment, the primers were derived from the coding region and were used in PCR reactions as a single forward (+) or reverse (-) primer or both (+/-). Single bands were observed in the (-) and (+/-) PCR reactions, but not in the (+) reaction. The results of the (+/-) reaction theoretically should have given two bands, the (+/-) band and a band derived from just the (-) primer since the single (-) primer reaction gave a band. However, only a single band was observed. Since it was larger than the band in the (-/-) reaction, it was concluded that this band was derived from the (+/-) combination. Because the result was fortuitous in that only a single product was recovered in a reaction that could have yielded two products, it became necessary to verify that the orientation was indeed correct. Therefore, the experiment was repeated using primers specific for each MFa sequence. The results of this experiment confirmed the orientation and distances of the previous PCR reactions using the (+) and (-) primer combinations (Fig 7C).

View larger version (79K): In this window In a new window Download PPT slide   Figure 7. Linkage mapping of MFa1, MFa2, and MFa3. (A) Predicted orientation and distance of each pheromone gene with respect to each other. Distances are coding sequence to coding sequence and primers were designed based on the coding sequence, with (+) being a forward primer at the 3' end and (-) being a reverse primer at the 5' end. (B) Long-distance PCR using primers derived from the coding sequence. (+) is a primer in the forward direction and (-) is a primer in the reverse direction. L is a 1.0-kb ladder. Only reactions that gave products (+/-) and (-/-) are shown. The PCR reaction containing the (+) and (-) primers gave a 10.5-kb product while the reaction containing only the (-) primer gave a 8.0-kb product. (C) Confirmation of pheromone orientation and distance using pheromone-specific primers. The lane designated (2/3) contained an MFa3-specific forward primer located in the promoter and an MFa2-specific reverse primer located in the 3' flank of MFa2. The lane designated (3/1) contains an MFa1-specific reverse primer located in the 3' flank and an MFa3-specific reverse primer located in the MFa3 flank. The product size difference from the (+/-) reactions reflects the location of the primers in the flank of each gene, rather than in the coding sequence, and confirms the predicted orientation.

The results of the Southern hybridization and the long-distance PCR indicated that there were three copies of the MATa pheromone. In addition to MFa1, the two remaining copies were cloned and sequenced. MFa2 and MFa3, which were designated to indicate the order in which they were isolated, were aligned to MFa1. The three genes are virtually identical over 902 bp. They diverge 200 bp upstream from the initial ATG, with MFa2 being more divergent than the other two pheromone copies, MFa1 and MFa3, which eventually diverge 633 bp upstream from the ATG site. In contrast to MFa1 and MFa3, MFa2 contains only four copies of TTTTGTTCTT. This 10-bp repeat is found in two 56-bp direct repeats in the promoter of MFa2. The MFa1 and MFa3 promoters, on the other hand, contain only a single copy of the 56-bp repeat found in MFa2. In the 3' flanking region, the three genes diverge 573 bp from the stop codon, with MFa1 and MFa3 remaining identical for at least another 186 bp. The 3' flanking sequence of these two genes was not extended so it is not known how far the similarity of the 3' flanking region extends.

MFa1 induces filament formation in MAT cells:
MOORE and EDMAN 1993 originally identified the MAT pheromone gene by transforming MATa strains with a series of constructs containing various fragments of the MAT locus. The filament-inducing plasmids were identified by plating transformants onto V8 agar and then screening microscopically for the presence of hyphal filaments. A similar method was employed to test the functionality of MFa1. This strategy was necessary because while the predicted phenotype of an mfa strain would be sterile, producing a triple disruptant in C. neoformans would be extremely laborious due to the poor homologous integration frequency.

To test for the hyphal phenotype, MATa and MAT strains were transformed with either the pMFa-1 genomic clone or the vector alone. Ura⁺ transformants were purified and then plated onto V8 agar to look for the production of filaments (Fig 8A). Filaments were visible in MAT cells transformed with pMFa-1 within 24–48 hr. MATa cells transformed with pMFa-1 did not produce filaments. Neither mating type produced filaments when transformed with the vector alone. These results are complementary to the original MF1 experiment (MOORE and EDMAN 1993 ) and showed that MFa1 is capable of inducing filament formation only in cells of the opposite mating type.

View larger version (34K): In this window In a new window Download PPT slide   Figure 8. Induction of filaments by MFa1. (A) MFa1 on an episomal URA5 vector (pC5-MFa) was transformed into MAT (WSA-43) and MATa (WSA-34) strains. Transformants were plated onto V8 agar and screened for filaments. Only the MAT pC5-MFa transformant produced filaments. Neither the pC5 nor MATa control transformants produced filaments. (B) Filament production under galactose regulation. Strain WSA-442, which contained an integrated MFa1 gene under galactose regulation was plated onto V8 agar with either 0.75% galactose or glucose. Cells were visualized and photographed at x25. Bars, 20 µm.

To rule out the possibility of media-induced haploid filament formation, which is manifested as monokaryotic fruiting in -cells (WICKES et al. 1996 ) but has not been observed in a-cells, a strain containing a GAL7-MFa1 fusion was tested for filament formation on media containing galactose or dextrose. This approach allows precise regulation of MFa1 expression by galactose since the promoter is induced by galactose but repressed by glucose (WICKES and EDMAN 1995 ). Fig 8B shows strain WSA-442 plated onto V8 agar containing either galactose or glucose. Filaments were observed within 24 hr on galactose-V8 agar while no filaments were observed from cells plated onto glucose-V8 agar, demonstrating that MFa1 induces the hyphal phenotype.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The C. neoformans MATa pheromone gene has a number of characteristics in common with other fungal pheromones. It is small, multicopy, and predicted to be extensively processed post-translationally. MFa1 appears to be structurally similar to a-factor. In particular, Mfa1p contains a consensus carboxy-terminal CAAX motif that would permit the cysteine, based on the S. cerevisiae a-factor model as well as on other fungal pheromone models (CALDWELL et al. 1995 ), to be farnesylated and then carboxylmethylated. The CAAX motif is a defining characteristic for this class of pheromones, which is noteworthy because the pheromones in this class are derived from a taxonomically diverse group of organisms spanning two phyla (ascomycetes and basidiomycetes). The C. neoformans pheromones, however, have a distinct difference from the S. cerevisiae pheromones. Both mating-type-specific pheromones in the C. neoformans model would be predicted, based on amino acid sequence, to mature via similar steps.

Comparison of the amino acid sequences of MFa1 and MF1 reveals that the two peptides have three conserved regions. The first 8 N-terminal amino acids are identical, which may suggest that this region is acted upon by a common processing mechanism. In S. cerevisiae, the first amino terminus proteolytic processing event is mediated by Ste24p (P. CHEN et al. 1997 ), which removes a fragment similar in size (7 amino acids) to the eight C. neoformans N-terminal amino acids found in both Mfa1p and Mf1p pheromone precursors. The plausibility of this argument is strengthened by the existence, based on sequence homology, of a C. neoformans homolog of the S. cerevisiae STE24 gene that can be found in the C. neoformans genomic database (www-sequence.stanford.edu/group/C.neoformans/). A second processing site defined by an asparagine residue located 11 amino acids from the carboxy-terminal cysteine in S. cerevisiae (ADAMES et al. 1995 ) is found in both C. neoformans pheromones. In C. neoformans, Mfa1p contains a single asparagine, which is located at the carboxy terminus 13 amino acids from the cysteine. In Mf1p the single asparagine is 10 amino acids from the carboxy-terminal cysteine. In S. cerevisiae, this carboxy-terminal asparagine is a proteolytic processing site whose cleavage is mediated by Axl1p/Ste23p (ADAMES et al. 1995 ). In virtually all fungal pheromones in the CAAX class, an asparagine residue can be found in close proximity to the carboxy-terminal cysteine (Table 2). The asparagine in Mfa1p is preceded by three additional amino acids to form a sequence of APRN, which is also present in Mf1p. The conserved carboxy-terminal position of the asparagine and the presence of the same 4 amino acids in Mfa1p and Mf1p suggest the possibility of a conserved function, which could be a conserved cleavage site. In fact, we have isolated a homolog of the S. cerevisiae AXL1 gene from C. neoformans that is induced under mating conditions (our unpublished data). Finally, both Mfa1p and Mf1p contain the carboxy-terminal CVIA residues, which suggests that there could be common farnesylation-carboxylmethylation maturation steps in both mating types, which are similar to other fungal pheromones in the CAAX class. Therefore, in the case of the C. neoformans pheromones, it is reasonable to conclude that MFa1 and MF1 are structurally as well as functionally conserved. In fact, expression of each pheromone in the opposite cell type results in morphological changes that lead to the production of hyphal filaments (this study and WICKES and EDMAN 1995 ). These phenotypes can also be induced in untransformed cells by placing a tester strain in close proximity (CHANG et al. 2000 ; WANG et al. 2000 ; CLARKE et al. 2001 ). Under these conditions the filaments will orient toward the opposite cell type and are likely responding to a pheromone gradient. Finally, synthetic MF1p has recently been prepared (DAVIDSON et al. 2000 ) and shown to induce MATa cells to respond with the filamentous phenotype. The structure of the synthetic version of Mf1p follows the mature farnesylated-carboxylmethylated S. cerevisiae Mfa1p pheromone as a model. Since each predicted site of post-translational modification is conserved in Mfa1p and Mf1p, and the asparagine-mediated Axl1p/Ste23p site and CAAX motif are widely conserved in other fungi, it is reasonable to expect Mfa1p to be processed in a manner similar to Mf1p and the S. cerevisiae a-factor. We have investigated preparing synthetic Mfa1p; however, we have not succeeded in recovering enough pheromone to assay activity. These problems have been due to the sequence of Mfa1p. The pheromone consists of a 14-amino-acid peptide (EEAYGSGQGPTYSC) in which the terminal cysteine is farnesylated and carboxylmethylated. Unfortunately, during the methylation reaction, the –COOH groups on the two amino-terminal glutamate residues must be blocked or they will be methylated as well, which in our experience has drastically decreased the yield of mature peptide. Two alternative strategies are presently being considered: to prepare a synthetic Mfa1p pheromone backbone with and without the glutamate residues and to farnesylate and carboxylmethylate without modifying the glutamate residues in the backbone.

In spite of the numerous similarities between MFa1 and MF1, there is an important difference between the two genes. MF1 hybridizes to MAT DNA of all four serotypes of C. neoformans, whereas MFa1 hybridizes only to DNA from strains that are variety neoformans (serotype D, and possibly A). These results suggest that, at least for the MATa pheromone, there may be serotype or variety specificity. For the MAT pheromone, the specificity does not appear to extend beyond mating type since MF1 hybridizes to MAT cells from all four serotypes. The difference in hybridization pattern may provide clues as to the evolution of mating types in C. neoformans. Although the pheromone genes will need to be recovered and analyzed from both mating types and all four serotypes, on the basis of the existing data it seems possible that the MATa pheromone gene evolved after the MAT pheromone gene. In fact, given the multicopy nature of the pheromone genes, which would allow variation in one sequence without loss of fertility, and the multiple conserved regions within MF1 and MFa1, it seems plausible that at least for serotype D cells, MFa1 could have evolved from MF1.

The primary reason for attempting to clone the MATa pheromone was to obtain a MATa-specific marker for the MATa mating-type locus. Previous studies have shown that the MAT locus, at 50 kb, is among the largest of all fungi that utilize a single-locus two-allele mating system (KAROS et al. 2000 ). After cloning MFa1, hybridization studies revealed the multicopy nature of the gene. The remaining two pheromone genes were cloned and then mapped with respect to MFa1. The mapping results revealed that the distance spanning the three MATa pheromones is 20 kb. Two other MATa-specific genes, STE12a and STE20a, have recently been described (LENGELER et al. 2000 ; CHANG et al. 2001 ) and we have isolated a MATa homolog of STE11 (our unpublished data). On the basis of these observations, we anticipate that the MATa locus will contain most, if not all, of the mating-type-specific homologs of genes already identified in MAT cells and therefore will likely be as large as its MAT counterpart.

A second reason for recovering the MATa pheromone was to establish a MATa-specific sequence that could be used in PCR or hybridization reactions to distinguish the two mating types. Previously, MATa strains were identified solely on the basis of the lack of PCR product or hybridization signal using the MAT pheromone sequence as a target. As shown in this study, isolates can now be typed on the basis of both the presence and absence of the two pheromones. Importantly, we have recently begun investigating diploid C. neoformans isolates and have found them to be extremely common and even recoverable among clinical isolates (COGLIATI et al. 2001 ). The use of the pheromone genes in addition to other mating-type-specific genes will allow more detailed genetic studies of these isolates and perhaps a better understanding of the way in which they originated.

A third reason for recovering the MATa pheromone gene in C. neoformans is the role that mating type plays in virulence. MAT cells are more virulent than MATa cells (KWON-CHUNG et al. 1992 ), and disruption of a number of previously cloned STE genes has been shown to drastically reduce virulence (ALSPAUGH et al. 1997 ; CHANG et al. 2000 ; LENGELER et al. 2000 ). While the pheromone is unlikely to be a virulence factor since mating of C. neoformans has never been described in vivo, there may an indirect role for the pheromone in virulence by virtue of its effect on cell morphology. Earlier studies of monokaryotic fruiting in C. neoformans revealed that in addition to producing hyphae, monokaryotic fruiting can also result in the production of basidiospores. Although the yeast cell has been hypothesized to be the infectious particle in cryptococcosis (TACKER et al. 1972 ), spores are the appropriate size for maximum penetration into the lungs, whereas encapsulated yeast cells are not (WICKES et al. 1996 ). MATa cells, through the pheromone response pathway, may mate with -cells and yield spores in the traditional manner or MATa cells may secrete pheromone into the environment at concentrations high enough to induce monokaryotic fruiting of haploid -cells, which would again lead to the production of spores. In spite of the fact that C. neoformans isolates readily mate in the laboratory, the true role of mating in the life cycle is still unsettled. Some researchers have suggested that mating is dispensable to C. neoformans because their life cycle is primarily clonal (CASADEVALL and PERFECT 1998 ). In fact, while isolates can be readily made to mate under laboratory conditions, a substantial proportion of isolates do not mate, and mating of clinical isolates is normally difficult under most circumstances. Additionally, in spite of the fact that C. neoformans can reach potentially massive numbers in nature (5 x 10⁷ cfu/g material; EMMONS 1955 ), outbreaks due to exposure to contaminated material have never been reported. This absence may reflect high levels of natural resistance in healthy individuals or the production of low numbers of actual infectious particles (i.e., spores). Therefore, the nature of the infectious particle and the mechanism by which it is produced are still unresolved. By characterizing both mating types in detail, a clearer understanding of how people become infected and the role of mating type and mating-associated genes in virulence may be obtained.

	ACKNOWLEDGMENTS

The authors thank David Clarke, K. J. Kwon-Chung, and Yun Chang for helpful discussions. C. neoformans genomic databases were searched at the C. neoformans Genome Project, Stanford Genome Technology Center (http://www-sequence.stanford.edu), which is funded by the National Institutes of Health under cooperative agreement AI47087. B.L.W. is a Burroughs-Wellcome New Investigator in Molecular Pathogenic Mycology and is supported by U.S. Public Health Service grant R29AI43522 from the National Institutes of Health, U.S. Public Health Service grant P30 CA 54174 to the San Antonio Cancer Institute from the National Institutes of Health, and an award to the UTHSCSA for the Research Resources Program of Medical Schools of the Howard Hughes Medical Institute.

Manuscript received September 20, 2001; Accepted for publication January 9, 2002.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ADAMES, N., K. BLUNDELL, M. N. ASHBY, and C. BOONE, 1995  Role of yeast insulin-degrading enzyme homologs in propheromone processing and bud site selection. Science 270:464-467[Abstract/Free Full Text].

AKADA, R., K. MINOMI, J. KAI, I. YAMASHITA, and T. MIYAKAWA et al., 1989  Multiple genes coding for precursors of rhodotorucine A, a farnesyl peptide mating pheromone of the basidiomycetous yeast Rhodosporidium toruloides.. Mol. Cell. Biol. 9:3491-3498[Abstract/Free Full Text].

ALSPAUGH, J. A., J. R. PERFECT, and J. HEITMAN, 1997  Cryptococcus neoformans mating and virulence are regulated by the G-protein alpha subunit GPA1 and cAMP. Genes Dev. 11:3206-3217[Abstract/Free Full Text].

ANDEREGG, R. J., R. BETZ, S. A. CARR, J. W. CRABB, and W. DUNTZE, 1988  Structure of Saccharomyces cerevisiae mating hormone a-factor. Identification of S-farnesyl cysteine as a structural component. J. Biol. Chem. 263:18236-18240[Abstract/Free Full Text].

ANDERSON, C. M., D. A. WILLITS, P. J. KOSTED, E. J. FORD, and A. D. MARTINEZ-ESPINOZA et al., 1999  Molecular analysis of the pheromone and pheromone receptor genes of Ustilago hordei.. Gene 240:89-97[Medline].

AULAKH, H. S., S. E. STRAUS, and K. J. KWON-CHUNG, 1981  Genetic relatedness of Filobasidiella neoformans (Cryptococcus neoformans) and Filobasidiella bacillospora (Cryptococcus bacillisporus) as determined by DNA base composition and sequence homology studies. Int. J. Bacteriol. 31:97-103.

BOEKE, J. D., J. TRUEHEART, G. NATSOULIS, and G. R. FINK, 1987  5-Fluoroorotic acid as a selective agent in yeast molecular genetics. Methods Enzymol. 154:164-175[Medline].

BOLKER, M., M. URBAN, and R. KAHMANN, 1992  The a mating type locus of U. maydis specifies cell signaling components. Cell 68:441-450[Medline].

BRAKE, A. J., C. BRENNER, R. NAJARIAN, P. LAYBOURN and J. MERRYWEATHER, 1985 Structure of genes encoding precursors of the yeast peptide mating pheromone a-factor, pp. 103–108 in Protein Transport and Secretion, edited by M. J. GETHING. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

CALDWELL, G. A., F. NAIDER, and J. M. BECKER, 1995  Fungal lipopeptide mating pheromones: a model system for the study of protein prenylation. Microbiol. Rev. 59:406-422[Abstract/Free Full Text].

CASADEVALL, A., and J. R. PERFECT, 1998 Ecology of Cryptococcus neoformans, pp. 41–70 in Cryptococcus neoformans. ASM Press, Washington, DC.

CHANG, Y. C., B. L. WICKES, G. F. MILLER, L. A. PENOYER, and K. J. KWON-CHUNG, 2000  Cryptococcus neoformans STE12alpha regulates virulence but is not essential for mating. J. Exp. Med. 191:871-882[Abstract/Free Full Text].

CHANG, Y. C., L. A. PENOYER, and K. J. KWON-CHUNG, 2001  The second STE12 homologue of Cryptococcus neoformans is MATa-specific and plays an impor