A STE12 Homolog Is Required for Mating but Dispensable ... - Genetics

Copyright  2000 by the Genetics Society of America

A STE12 Homolog Is Required for Mating but Dispensable for Filamentation in Candida lusitaniae Laura Y. Young, Michael C. Lorenz1 and Joseph Heitman Departments of Genetics, Pharmacology and Cancer Biology, Microbiology, and Medicine, Howard Hughes Medical Institute, Duke University Medical Center, Durham, North Carolina 27710 Manuscript received October 27, 1999 Accepted for publication January 12, 2000 ABSTRACT Candida lusitaniae is a dimorphic yeast that is emerging as an opportunistic fungal pathogen. In contrast to Candida albicans, which is diploid and asexual, C. lusitaniae has been reported to have a sexual cycle. We have employed genetic approaches to demonstrate that C. lusitaniae is haploid and has a sexual cycle involving mating between MATa and MAT␣ cells under nutrient deprivation conditions. By degenerate PCR, we identified a C. lusitaniae homolog (Cls12) of the Ste12 transcription factor that regulates mating, filamentation, and virulence in Saccharomyces cerevisiae, C. albicans, and Cryptococcus neoformans. Comparison of the CLS12 DNA and protein sequences to other STE12 homologs and transformation experiments with selectable markers from S. cerevisiae (URA3, KanMX, HphMX ) and C. albicans (CaURA3) provide evidence that the CUG codon encodes serine instead of leucine in C. lusitaniae, as is also the case in C. albicans. The C. lusitaniae CLS12 gene was disrupted by biolistic transformation and homologous recombination. C. lusitaniae cls12 mutant strains were sterile but had no defect in filamentous growth. Our findings reveal both conserved and divergent roles for the C. lusitaniae STE12 homolog in regulating differentiation of this emerging fungal pathogen.

C

ANDIDA species are the most frequently encountered human fungal pathogens (Hazen 1997). These fungi commonly cause infections in immunocompromised patients that are either localized on mucosal membrane surfaces or widely disseminated and involving multiple organs. Bloodstream infection by these fungi, candidemia, is also a major complication in patients with indwelling intravascular catheters. The incidence of Candida infections is increasing as a result of AIDS, organ transplantation, cytotoxic chemotherapy, and other conditions that suppress the immune system. C. albicans causes a majority (60%) of all Candida infections (Kwon-Chung and Bennett 1992). Infections attributable to other non-albicans Candida species, such as C. lusitaniae, are increasing. In the early 1960s, only five Candida species were known to cause disease: C. albicans, C. tropicalis, C. stellatoidea, C. parapsilosis, and C. guillermondii (Hazen 1997). At least 17 additional species have since been identified as emerging opportunistic pathogens (Hazen 1997; Coleman et al. 1998). In a recent survey of fungemia, 7% of Candida infections were attributable to C. glabrata, 2% to C. hansenula, and 1% to C. krusei (Hazen 1997). C. lusitaniae was first described by van Uden and

Corresponding author: Joseph Heitman, 322 CARL Building, Box 3546, Research Dr., Duke University Medical Center, Durham, NC 27710. E-mail: [email protected] 1 Present address: Whitehead Institute for Biomedical Research, 9 Cambridge Center, Cambridge, MA 02142. Genetics 155: 17–29 (May 2000)

do Carmo-Sousa, who isolated the organism from the gastrointestinal tracts of warm-blooded animals (van Uden and Buckley 1970). By 1989, C. lusitaniae had been recovered from a variety of clinical specimens in a small percentage (0.64%) of Candida infections (Blinkhorn et al. 1989). By 1996, ⵑ1% of all candidemia infections were attributable to C. lusitaniae (Nguyen et al. 1996). C. lusitaniae has been recovered from multiple sites in the human body, including the kidney, peritoneum, bloodstream, bladder and urine, vagina, skin, and respiratory tract (Baker et al. 1984; Merz 1984; Hadfield et al. 1987; Sanchez et al. 1992; Yinnon et al. 1992). Prior to 1985, nearly all reported cases of C. lusitaniae infection were fatal, largely as a consequence of intrinsic or acquired amphotericin B resistance (Blinkhorn et al. 1989). The introduction of azoles improved therapy for C. lusitaniae fungemia, but amphotericin B resistance is an ongoing clinical problem (Pappagianis et al. 1979; Ahearn and McGlohn 1984; Yoon et al. 1999). Because the frequency of infection by C. lusitaniae is increasing, especially in immunocompromised patients, this organism has been classified as an emerging opportunistic pathogen (Hazen 1997). Candida species are polymorphic and can form colonies containing budding yeasts, pseudohyphae, and true hyphae. The ability to filament and produce either pseudohyphae, true hyphae, or both is thought to enable Candida species to colonize and disseminate within infected hosts (Braun and Johnson 1997; Lo et al. 1997). The related model yeast, Saccharomyces cerevisiae, also undergoes filamentous differentiation, and the signal-

18

L. Y. Young, M. C. Lorenz and J. Heitman

ing pathways regulating filamentation are conserved with those in C. albicans (Liu et al. 1994; Clark et al. 1995; Kohler and Fink 1996; Leberer et al. 1996; Lo et al. 1997; Csank et al. 1998). In S. cerevisiae, the pheromone-responsive mitogen-activated protein kinase cascade regulates both filamentation and mating (Liu et al. 1993). Interestingly, the C. albicans homolog of the S. cerevisiae Ste12 transcription factor, Cph1, is also involved in filamentation (Liu et al. 1994; Lo et al. 1997). Specifically, a mutant strain in which both alleles of the CPH1 gene have been deleted exhibits decreased filamentation, and when combined with mutations in the EFG1 gene, the resulting cph1 efg1 double mutant strains are nonfilamentous and avirulent (Lo et al. 1997; Stoldt et al. 1997). C. albicans has no known sexual cycle and thus classical genetic approaches to understand the life cycle and mechanisms of pathogenicity are difficult, if not impossible (Scherer and Magee 1990). C. lusitaniae is one of the few Candida species thought to have a sexual cycle, and MAT␣ and MATa isolates have been reported to occur in a ratio of 6:1, respectively (Gargeya et al. 1990). The sexual or teleomorphic form of this organism has been designated Clavispora lusitaniae. We refer here to both Clavispora lusitaniae and Candida lusitaniae as C. lusitaniae. Because C. lusitaniae is thought to be haploid, is pathogenic, and may have a sexual cycle, this organism offers an attractive model in which genetic methods could be utilized to elucidate mechanisms involved in mating and filamentation in Candida species in general. Here we have developed a mating assay for C. lusitaniae that utilizes auxotrophic and drug-resistant mutant strains, cloned the C. lusitaniae gene encoding a homolog of the Ste12 transcription factor, developed a transformation system, and disrupted the CLS12 gene by homologous recombination. We show that the Ste12 homolog Cls12 is required for mating but not for filamentation of C. lusitaniae. Our studies provide a genetic and molecular foundation to dissect pathogenesis in a haploid Candida species with a sexual cycle.

MATERIALS AND METHODS Media and strains: YPD medium, yeast nitrogen base (YNB) medium, and synthetic dextrose medium lacking uracil (SDUra), methionine (SD-met), and isoleucine (SD-ile) were prepared as described (Sherman 1991). Medium containing 5-fluoroorotic acid (5-FOA) and filamentation agar were prepared as described (Boeke et al. 1987; Wickes et al. 1996). Medium containing cycloheximide was prepared by adding sterile cycloheximide to autoclaved YPD, at a final concentration of 10 ␮g/ml. Potato dextrose agar medium (PDA) was prepared by adding 1.81% of potato dextrose agar (Difco, Detroit) to 1 liter of distilled water. Synthetic low (SLAD), medium (SMAD), and high (SHAD) ammonium dextrose medium contain 0.17% yeast nitrogen base without amino acids or ammonium sulfate (Difco), 2% dextrose, and 50 ␮m (SLAD), 500 ␮m (SMAD), or 5000 ␮m (SHAD) ammonium

sulfate, respectively. V8 medium was prepared with 20% V8 juice (Campbell’s Soup Co.), 0.05% KH2PO4, and adjusted to pH 7.0 with 5 m KOH. C. lusitaniae strains used in this study (Table 1) were obtained from the American Type Culture Collection (ATCC): 24009 (CL1, MATa), 38533 ⫽ CBS 6936 (CL2, MATa), and 42720 (CL3, MAT␣) and from the strain collection maintained at the Duke University Mycology Research Unit by Wiley Schell: 111.92 ⫽ ATCC 64125 (urine culture, CL5, MAT ?) 100.97 (urine culture, CL6, MATa), and 117.96 (abdominal biopsy culture, CL7, MAT␣). The S. cerevisiae strain used for gap repair was PJ69-4A ( James et al. 1996). Spontaneous and induced mutagenesis: Spontaneous 5-FOA (ura⫺) and cycloheximide-resistant (chxr) strains of C. lusitaniae were isolated. C. lusitaniae strains were grown in 5-ml liquid YPD medium cultures, washed in water, and plated on YPD medium to quantify the number of cells and ⵑ1 ⫻ 109 cells were plated on 5-FOA medium or YPD medium containing cycloheximide (10 ␮g/ml) to isolate mutants and incubated at 30⬚ for 24 hr. Auxotrophic mutants were also isolated by UV mutagenesis. Cultures of the MATa strain CL2 and the MAT␣ strain CL3 were grown in liquid YPD, diluted, and plated to YPD medium. C. lusitaniae cells (ⵑ106) were irradiated in a UV Stratalinker 2400 (Stratagene, La Jolla, CA) at 5000 or 7500 J and incubated in the dark for 24 hr at 30⬚ until individual colonies formed. Colonies were replica-plated to YNB medium, incubated for 24 hr at 30⬚, and colonies were isolated that were viable on YPD medium, but inviable on YNB medium. Two auxotrophic mutants were isolated and tested on synthetic dextrose medium lacking individual amino acids. One mutant (from strain CL3) was auxotrophic for methionine, and the second (from strain CL2) was auxotrophic for isoleucine. Mating assays: A 5-FOA-resistant, MAT␣ ura3 strain and a 5-FOA-resistant, MATa ura1 strain of C. lusitaniae were grown on PDA, SLAD, or V8 medium. Mating assays consisted of cospotting ⵑ107 cells each of the two strains to be tested, flanked by spots containing only one parent strain or the other. Mating mixtures and controls were incubated at 30⬚ for 72 to 120 hr, replica-plated to synthetic medium lacking uracil or YNB medium, and incubated at 30⬚ for 24 to 72 hr. The same assay was conducted with 5-FOA-resistant strains and cycloheximide-resistant strains of opposite mating types, and the mating mixtures were replica-plated to medium containing 5-FOA and cycloheximide (10 ␮g/ml). Quantitative matings were performed by growing cells to saturation in 5 ml YPD medium, washing once with sterile water, resuspending cells in 5 ml water, and cospotting 107 cells of each strain in a 20-␮l spot on SLAD medium, or spreading 109 cells of each strain on an entire 65 mm SLAD plate. Matings were incubated for 96 hr at 30⬚, replica-plated to 5-FOA plus cycloheximide (10 ␮g/ml) medium, and incubated for 48 to 96 hr. The number of recombinant colonies was counted, multiplied by four (since only one-fourth of recombinant colonies are ura⫺ chxr), and divided by the number of cells plated (107 or 109) to yield the frequency of mating. Cloning of the C. lusitaniae STE12 homolog CLS12: The wild-type CLS12 gene was cloned from C. lusitaniae by degenerate PCR. Primers were designed against regions of identity in the S. cerevisiae STE12 gene, the Kluyveromyces lactis STE12 gene, and the C. albicans STE12 homolog, CPH1: 5⬘-AAYTGGCARG ARAAYCA and 5⬘-TTYTTYGTYTTYCAIAARAARACCAA (Y is T or C, R is G or A, and I is inosine). PCR conditions were 5 min at 94⬚, 35 cycles of 30 sec at 94⬚, 30 sec at 45⬚, and 1.5 min at 72⬚, and a final 5-min extension step at 72⬚. The resulting 315-bp fragment was cloned in the TA System (Invitrogen, San Diego) and sequenced, revealing identity to STE12 homologs. The C. lusitaniae CLS12 gene fragment was

Mating in Candida lusitaniae

19

TABLE 1 Strain list Strain

Genotype

Original designation

CL1 CL2 CL3 CL5 CL6 CL7 CL8 CL16 CL17 CL24 CL27 CL45 CL49 CL59 CL60 CL62 CL70 CL71 CL73 CL92 CL93 CL117 CL118 CL119 CL120 CL121

MATa MATa MAT␣ MAT ? (sterile?) MATa MAT␣ MATa ura1 (CL1) MATa ura3 (CL2) MATa ura3 (CL2) MAT␣ ura3 (CL3) MAT␣ ura1 (CL3) MATa ura3 (CL6) MAT␣ ura3 (CL7) MATa chxr (CL2) MAT␣ chxr (CL3) MAT␣ chxr (CL3) MAT␣ met1 (CL3) MATa ile1 (CL2) MAT ? ura3 chxr (CL17 ⫻ CL62) MAT␣ cls12::CaURA3 ura3 (CL24) MAT␣ cls12::CaURA3 ura3 (CL24) MAT ? cls12::CaURA3 chxr ura3 (CL73) MAT␣ ectopic CaURA3 ura3 (CL24) MAT ? ectopic CaURA3 ura3 (CL73) MAT␣ cls12::CaURA3 chxr ura3 (CL92) MAT␣ cls12::CaURA3 chxr ura3 (CL92)

(ATCC 24009) (ATCC 38533/CBSa 6936) (ATCC 42720) (ATCC 64125/DUMCb 111.92) (DUMC 100.97) (DUMC 117.96)

a b

Centralbureau voor Schimmelcultures, The Netherlands. Duke University Medical Center.

labeled with [32P]dCTP using the Boehringer-Mannheim (Indianapolis) random-primer DNA labeling kit and used as a probe for Southern blot. Genomic DNA from the MAT␣ strain CL3 was isolated using a protocol previously described for S. cerevisiae (Hoffman and Winston 1987) and digested with HindIII. Cleaved DNA was gel-electrophoresed, transferred to nitrocellulose, and probed (Sambrook et al. 1989). To isolate the CLS12 gene by size selection, CL3 genomic DNA was digested with HindIII, gel-electrophoresed, and 4- to 4.5-kb fragments containing the CLS12 gene were gel-purified using the Qiaex protocol (QIAGEN, Chatsworth, CA). Plasmid pUC18 was digested with HindIII, dephosphorylated with shrimp alkaline phosphatase (SAP), and ligated to the purified genomic DNA fragments. This library was electroporated into DH5␣ Escherichia coli cells, and colony lifts and secondary screens were performed (Sambrook et al. 1989). The 315-bp PCR product, described above, was used as the probe, and four colonies were identified that contained the pUC18/CLS12 plasmid (pUCL12). The entire 4.2-kb DNA fragment was sequenced on both strands at the Duke Sequencing Facility using synthetic primers. Construction of the cls12⌬::CaURA3 disruption allele: The CLS12 gene was disrupted with the C. albicans URA3 gene by a gap repair recombination approach in S. cerevisiae. First, plasmid pUCL12, a pUC18 derivative containing the CLS12 gene, was digested with PstI and EcoRI, releasing a 1.9-kb fragment of the CLS12 gene. Next, the E. coli–S. cerevisiae shuttle plasmid pGAD424 was digested with PstI and EcoRI and ligated with the 1.9-kb PstI-EcoRI CLS12 gene fragment to yield plasmid pGAS12. Plasmid pGAS12 was linearized with

EagI and cotransformed into S. cerevisiae with a 1.6-kb PCR product consisting of the C. albicans URA3MX4 gene flanked by 40 bp of homology to DNA flanking the EagI site in the CLS12 gene on plasmid pGAS12. Gap repair results in the insertion of the C. albicans URA3 gene into the CLS12 gene by homologous recombination (Orr-Weaver et al. 1981). The PCR product was obtained with two primers, each with 40 bp of homology to DNA flanking an EagI site (to result in the deletion of 191 bp spanning the EagI site, ⌬139–329) in the CLS12 gene at the 5⬘ end and with 19–22 bp of homology to the sequence flanking the C. albicans URA3 gene (5⬘-AGTCATCA GAAGATATTACTTGAACAACGATGAGGGTTTCCAGCTG AAGCTTCGTACGC and 5⬘-CTGAGTGCGGAGACAAGAGT TGTTGTAGAGGAACTCGAGAGCATAGGCCACTAGTGGA TCTG). PCR reaction conditions [with the CaURA3 gene on plasmid pAG60 as template, Goldstein et al. (1999)] were 5 min at 94⬚, 35 cycles of 30 sec at 94⬚, 30 sec at 55⬚, and 2 min at 72⬚, and a final 5-min extension step at 72⬚. The resulting plasmid now bearing the cls12⌬::CaURA3 disruption allele, pUS12, was rescued from yeast and amplified in E. coli. A 2.5kb KpnI-XbaI fragment containing the cls12⌬::CaURA3 allele was cloned into the corresponding sites of pBluescript to create plasmid pBUS12. The pBUS12 disruption construct was transformed into strains CL24 (CL3 background, MAT␣ ura3) and CL73 (CL17 ⫻ CL62, MAT ? chxr ura3) as circular DNA, linear DNA (digested with XbaI and KpnI), or linear DNA that had been treated with ddATP and terminal deoxytransferase (TdT) to reduce nonhomologous recombination events (Shah-Mahoney et al. 1997). The biolistic transformation protocol developed for C. neoformans was used (Toffaletti

20


et al. 1993). Ura⫹ isolates were selected on synthetic dextrose medium lacking uracil and containing 1 m sorbitol and screened by PCR and Southern blot to identify cls12⌬::CaURA3 transformants. Flow cytometry, 5ⴕ and 3ⴕ rapid amplification of cDNA ends (RACE) analysis, and Southern blot: Cells were processed for flow cytometry as described (Haase and Lew 1997). Mating assays were performed on SLAD medium, and after 36 hr of incubation at 30⬚, samples were removed from the mating mixture and control parental strains, resuspended in YPD, and grown to mid-log phase. DNA was stained with propidium iodide and analyzed on the FL1 channel on a Becton-Dickinson FACScan. A total of 2 to 5 ␮g of mRNA extracted from strain CL3 (RNeasy kit, QIAGEN) was used for both a 5⬘ and a 3⬘ RACE (GIBCO BRL, Gaithersburg, MD). The resulting RACE products were cloned into the TopoTA cloning vector (Invitrogen) and analyzed by DNA sequencing at the Duke Sequencing Facility. Genomic DNA was isolated from presumptive mutant and wild-type strains, digested with HindIII, gel-electrophoresed, transferred to nitrocellulose, and probed, following a protocol previously described (Sambrook et al. 1989). The Southern blot was probed with the 4.2-kb HindIII DNA fragment from plasmid pUCL12 containing the CLS12 gene.

RESULTS

Isolation of auxotrophic and cycloheximide-resistant mutants of C. lusitaniae: C. lusitaniae has been reported to have a sexual cycle, as detected by the production of asci and ascospores following incubation of MATa and MAT␣ strains on potato dextrose agar (Gargeya et al. 1990). We sought to address whether a C. lusitaniae sexual cycle could be analyzed by genetic approaches through the development of a mating assay for this organism. To this end, five clinical and one environmental isolate of C. lusitaniae were obtained from the ATCC and the Duke University Mycology Research Unit strain collections. These isolates include one from the first reported case of opportunistic infection (ATCC 42720, blood culture from a leukemia patient) and also the type culture (ATCC 38533/CBS 6936) isolated from citrus peel juice in Israel. All six isolates grew on YNB minimal media and are therefore prototrophic. The mating type of strain CL2 (ATCC 38533/CBS 6936) was previously reported as MATa and that of strain CL3 (ATCC 42720) as MAT␣ (Gargeya et al. 1990). The mating types of the other four isolates were unknown, but through the isolation of auxotrophic mutations in all of the strains, and mating assays with C. lusitaniae strains of known mating types, strains CL1 and CL6 were found to be MATa, and strain CL7 was MAT␣. We could not mate strain CL5, which may therefore be sterile. To isolate genetically marked C. lusitaniae strains for crosses, mutants were isolated on 5-FOA medium or medium containing cycloheximide (see materials and methods). Both 5-FOA and cycloheximide-resistant mutants were readily obtained. The frequency at which C. lusitaniae 5-FOA-resistant mutants were isolated was ⵑ1 ⫻ 10⫺7; by comparison, no 5-FOA-resistant mutants were obtained in the diploid organism C. albicans (strain

SC5314, frequency ⬍2 ⫻ 10⫺9, at least 50-fold lower than C. lusitaniae). Cycloheximide-resistant mutants occurred at a frequency of ⵑ1 ⫻ 10⫺8 in C. lusitaniae, whereas C. albicans strain SC5314 was inherently resistant. Because recessive mutations confer 5-FOA and cycloheximide resistance in other organisms, these findings suggest C. lusitaniae is either haploid or heterozygous for preexisting mutations at these loci. As expected, 5-FOA-resistant mutants were all auxotrophic for uracil. As described below, the C. albicans URA3 gene complemented the 5-FOA ura⫺ mutation present in one class of uracil auxotrophic mutant strains, which were designated ura3 mutants. A second class of 5-FOA-resistant uracil auxotrophic mutants was not complemented by the C. albicans URA3 gene and in genetic crosses complemented the ura3 mutation. These have been designated ura1 mutations. C. lusitaniae mutants auxotrophic for methionine or isoleucine were isolated following UV mutagenesis. C. lusitaniae has a sexual cycle: We next developed a mating assay that consists of coincubating presumptive MATa and MAT␣ uracil auxotrophic mutant strains on potato dextrose agar medium. When the mating patches were replica-plated to synthetic medium lacking uracil, certain pairs of MATa and MAT␣ strains mated to produce Ura⫹ progeny (Figure 1A). In contrast, few or no Ura⫹ isolates were obtained when either parental strain was incubated in the absence of the mating partner. These observations demonstrate that different uracil auxotrophic mutations are present in the 5-FOA-resistant mutant strains and that mating results in the production of recombinant prototrophic progeny that could include heterokaryons, diploids, or haploid meiotic recombinants. We also conducted mating crosses between cycloheximide-resistant mutants and uracil auxotrophic strains of opposite mating type. When replica-plated to 5-FOA medium containing cycloheximide, mating mixtures yielded 5-FOA and cycloheximide-resistant recombinant strains, whereas neither parental strain yielded isolates resistant to both 5-FOA and cycloheximide (Figure 1B). Given that recessive mutations confer resistance to 5-FOA and cycloheximide in other organisms, the isolation of recombinant strains resistant to both agents following mating in C. lusitaniae suggests that conjugation is quickly followed by meiosis and sporulation to produce haploid meiotic recombinants, as is known to be the case in Schizosaccharomyces pombe and C. neoformans in which the diploid state is transient. To further address the nature of the recombinant progeny resulting from mating in C. lusitaniae, we analyzed the segregation of the MAT locus in the progeny of a genetic cross. For this purpose, we crossed the MATa ura1 strain CL8 and the MAT␣ chx r strain CL60 and isolated ura1 chx r recombinants on 5-FOA medium containing cycloheximide. The mating type of these recombinant progeny was then scored by crosses to two mating-type tester strains with a complementing ura3


21

Figure 1.—Mating assay for C. lusitaniae. (A) 107 cells of the MATa ura1 strain CL8, the MATa ura3 strain CL16, and the MAT␣ ura3 strain CL24 were incubated alone (right and left spots) or together (center spots containing CL8 cells mixed with CL16 and CL8 cells mixed with CL24 cells) on SLAD medium for 96 hr at 30⬚ and then replica-plated to YNB synthetic medium and incubated for 48 hr. (B) A total of 107 cells of the MAT␣ ura3 strain CL24, the MATa chxr strain CL59, and the MAT␣ chxr strain CL62 were incubated alone (right and left spots) or together (center spots containing CL24 cells mixed with CL59 cells and CL24 cells mixed with CL62 cells) on SLAD medium for 96 hr at 30⬚ and then replicaplated to 5-FOA medium plus cycloheximide and incubated at 30⬚ for 72 hr. (C) The MATa ura1 and MAT␣ ura3 strains CL8 and CL24 were coincubated on PDA, PDA with 2% glucose, PDA with 5000 ␮m ammonium sulfate, SLAD, SMAD, and SHAD media for 96 hr at 30⬚ and replica-plated to YNB medium.

mutation: MATa ura3 strain CL16 and the MAT␣ ura3 strain CL24. Of the 20 haploid recombinant progeny, 19 mated with either the MATa or the MAT␣ tester strain, and the ratio was 10 MATa:9 MAT␣:1 sterile isolate. These findings provide evidence that recombinant progeny are produced by mating and meiosis in C. lusitaniae, that mating type is regulated by a bipolar allelic MAT locus in this heterothallic yeast, and that C. lusitaniae is a haploid yeast. FACS analysis of the total amount of DNA within the wild-type parental strains and in cells isolated from a mating mixture of MATa and MAT␣

cells revealed an increase in DNA ploidy in mating cells (Figure 2). The additional 4N peak in the coculture of the MAT␣ ura1 strain CL27 and the MATa ura3 strain CL16 confirms that diploid cells or heterokaryons are present in a mating mixture. The 4N peak is absent in the FACS analysis of the single parents (CL27 and CL16) and a wild-type strain (CL3) analysis (Figure 2), which contained only 1N (G1) and 2N (G2) peaks of DNA. Mating occurs under nitrogen starvation conditions in C. lusitaniae: In S. pombe and C. neoformans, mating has been shown to occur in response to adverse nutrient

22


Figure 2.—Measurement of cellular DNA content in parental strains and mating cultures. The FACS analysis shows a haploid DNA content (1N, 2N) in strains CL16, CL27, and CL3 (A, B, and D). The M3 peak in the coincubation mating culture of the MATa and MAT␣ strains CL16 and CL27 represents diploid cells (2N, 4N) resulting from mating prior to meiosis (C).

conditions, such as low nitrogen (Alspaugh et al. 1998; Dong and Courchesne 1998; Schweingruber et al. 1998). We tested whether C. lusitaniae also undergoes mating during nutrient deprivation. To establish nutritional conditions required for mating, assays were conducted on a variety of different media, including YPD, YNB, SLAD, and PDA media, and V8 juice agar (see materials and methods). Mating mixtures were incubated on these media and then replica-plated to synthetic medium lacking uracil. This analysis revealed that PDA, SLAD, and V8 minimal media all supported mating, whereas YPD and YNB media did not (Figure 1C). To further establish nutritional requirements for mating, assays were conducted on SLAD medium containing a constant level of carbon source (2% glucose) and different levels of nitrogen source: 50, 500, or 5000 ␮m (NH4)2SO4, or PDA medium alone or with 2% glucose or 5 mm (NH4)2SO4 added (see materials and methods). Increasing the nitrogen concentration in SLAD medium from 50 to either 500 or 5000 ␮m ammonium sulfate blocked mating. Similarly, supplementing PDA medium with 5 mm ammonium sulfate blocked mating, whereas the addition of 2% glucose did not. These findings indicate that, similar to C. neoformans and S. pombe, mating occurs in C. lusitaniae when nutrients are limiting (Figure 1C). Isolation of the C. lusitaniae STE12 homolog CLS12: In S. cerevisiae, the STE12 gene plays an important role

in regulating mating and filamentation. The STE12 homolog in C. albicans, CPH1, also plays a central role in filamentation and virulence (Liu et al. 1994; Lo et al. 1997). We have addressed whether a STE12 gene homolog plays a similar role in mating and filamentation in C. lusitaniae. The C. lusitaniae STE12 homolog was isolated by degenerate PCR amplification using primers designed to areas of sequence identity shared between the S. cerevisiae, C. albicans, and K. lactis STE12 homologs. Degenerate PCR under low stringency conditions yielded a 315bp PCR product from genomic DNA of both the C. lusitaniae MATa strain CL2 (ATCC 38533) and the MAT␣ strain CL3 (ATCC 42740). Both PCR products were cloned into the TA cloning vector (Invitrogen). DNA sequence analysis revealed a unique open reading frame with marked identity to known STE12 genes, and the 145-bp sequences derived from the homeodomain in MATa and MAT␣ C. lusitaniae strains, CL2 and CL3, were identical, indicating that these nonisogenic isolates are closely related. Southern analysis further confirmed that this sequence was derived from C. lusitaniae and hybridized to a single 4.2-kb HindIII fragment present in both MATa and MAT␣ strains. A size-selected library was constructed from genomic DNA of the MAT␣ strain CL3, and the 4.2-kb HindIII fragment containing the CLS12 gene was isolated by colony hybridization. Sequence analysis revealed a 1.6-kb open reading frame, and 5⬘ and 3⬘ RACE analyses established the 3⬘ terminus and polyadenylation site of the CLS12 mRNA (see materials and methods) and suggested that the ATG codon indicated in Figure 3 is the likely start codon of the CLS12 open reading frame. The open reading frame of the CLS12 gene spans 1431 bp and encodes a 477-amino-acid protein (Figure 3). The overall size of the predicted Cls12 protein is somewhat shorter than other Ste12 homologs. The N-terminal region and homeodomain of the Cls12 protein shares sequence identity with the S. cerevisiae, K. lactis, and C. albicans STE12 homologs, with 67.2, 71.0, and 77.6% identity, respectively (Figure 4). In addition, the six-amino-acid proposed binding site for the Dig1 and Dig2 proteins is also conserved in the C. lusitaniae Ste12 homolog, Cls12 (Figure 4). CUG encodes serine and not leucine in C. lusitaniae: In C. albicans and several other Candida species, the universal genetic code is violated such that CUG codons encode serine rather than leucine (Santos et al. 1993, 1997). Previous studies have provided conflicting evidence whether this is also the case in C. lusitaniae. For example, based on phylogenetic trees derived from rRNA sequences, C. lusitaniae is quite divergent from Candida species that use CUG to encode serine (Pesole et al. 1995). Moreover, a serine tRNA specific for CUG codons that has been reported to be from C. lusitaniae is identical to a tRNA sequence reported to be derived from C. melibiosica (Pesole et al. 1995). On the other


hand, Ueda et al. (1994) used in vitro translation reactions to demonstrate that CUG codons are translated as serine in extracts derived from C. lusitaniae (Ueda et al. 1994), providing evidence that this unusual feature of C. albicans and related Candida species is shared with C. lusitaniae. Our findings provide additional evidence that in C. lusitaniae CUG codons are translated to serine and not leucine, similar to C. albicans. First, we compared the sequence of the C. lusitaniae CLS12 gene with homolo-

23

gous genes from S. cerevisiae, C. albicans, and K. lactis. Two CUG codons are present in the highly conserved homeodomain in the C. lusitaniae CLS12 gene (Figure 4). At the corresponding positions in the Ste12 homologs of the other organisms, the amino acid at both of these positions is serine, suggesting that the two CUG codons in the C. lusitaniae CLS12 gene likely encode conserved serine residues at these positions. There are three other CUG codons in the C. lusitaniae CLS12 gene in less conserved regions of the protein, which we predict also encode serines. A second independent line of evidence that CUG codons are translated as serine residues is based on transformation experiments in C. lusitaniae. For example, no Ura⫹ or drug-resistant isolates were obtained in more than a dozen different experiments in which ura3 C. lusitaniae mutant strains were transformed with the S. cerevisiae URA3 gene or the KanMX and HphMX gene cassettes (Goldstein and McCusker 1999) using several different approaches (LiAc treated cells, electroporation, or biolistic DNA delivery). In contrast, Ura⫹ transformants were readily obtained following biolistic transformation of C. lusitaniae ura3 mutant strains, but not ura1 mutant strains, with the C. albicans URA3 gene (CaURA3). Because the S. cerevisiae URA3 gene contains a CUG codon that encodes an essential leucine residue (Suzuki et al. 1997), and the KanMX gene contains six CUG codons, these transformation experiments provide additional evidence that CUG encodes serine in C. lusitaniae. Disruption of the C. lusitaniae CLS12 gene by homologous recombination: The STE12 gene is not necessary for growth but is necessary for mating and filamentation in S. cerevisiae. Thus, our hypothesis was that C. lusitaniae cls12 mutants would be viable but have defects in mating or filamentation. To test this hypothesis, the CLS12 gene was disrupted with the C. albicans URA3 gene. The cls12⌬::CaURA3 disruption deletion allele was created by inserting the C. albicans URA3MX4 gene cassette (Goldstein et al. 1999) into the homeodomain of the CLS12 gene using a yeast gap repair method (see materials and methods). The resulting cls12::CaURA3 disruption allele was introduced into C. lusitaniae by biolistic transformation of circular DNA, the linearized cls12::URA3 DNA fragment, or a linearized DNA fragment with dideoxyadenosine triphosphates added to the ends to inhibit nonhomologous recombination (see materials and methods). Strain CL24, a MAT␣ ura3

Figure 3.—The nucleotide and encoded amino acid sequence of the CLS12 gene. The 1431-bp and 477-amino-acid open reading frame of the CLS12 gene, with 5⬘ and 3⬘ untranslated regions in lowercase letters. The TGA stop codon is in boldface type, and olignucleotide primers used for degenerate PCR and the 5⬘ and 3⬘ RACE products are underlined. The GenBank accession number is AF175524.

24


Figure 4.—C. lusitaniae Cls12 homeodomain is highly conserved. The Cls12 protein was aligned with the S. cerevisiae Ste12, C. albicans Cph1, and K. lactis Ste12 proteins. The proposed binding site for the Dig1 and Dig2 proteins is underlined, and the CUG encoded serines are marked with asterisks.


Figure 5.—Disruption of the CLS12 gene and Southern analysis of cls12⌬::URA3 mutants. (A) The C. albicans URA3 gene was used to replace 191 bp in the homeodomain of the CLS12 open reading frame and the resulting cls12::CaURA3 allele was introduced by biolistic transformation and homologous recombination. (B) Genomic DNA was prepared from the CLS12 wild-type parental strains (CL3, CL24, and CL73) and from the cls12::CaURA3 mutant strains (CL92, CL93, CL117, CL120, and CL121), digested with HindIII, electrophoresed, transferred to nitrocellulose, and probed with the CLS12 open reading frame.

mutant derived from the clinical isolate strain CL3, and strain CL73, a ura3 chx r mutant derived from a cross of strain CL17 with strain CL62, served as recipient strains for transformation. Uracil prototrophic transformants were selected on synthetic medium lacking uracil and containing 1 m sorbitol. Of 160 Ura⫹ prototrophic transformants, 36 were colony-purified, and PCR and Southern blot analysis confirmed that the wild-type CLS12 locus had been replaced by the cls12::CaURA3 disruption construct through homologous recombination in 3 of 36 transformants (8.3%; Figure 5). Two cls12 mutants were isolated following transformation with linear DNA, one cls12 mutant was obtained following transformation of linear DNA in which the ends had been blocked with ddATP, and none were obtained with circular DNA. These observations suggest that the efficiency of gene replacement may be higher with linear compared to circular DNA, as is the case in S. cerevisiae. In contrast to C. albicans, which is diploid and requires two independent transformation events to achieve gene disruption, the wild-type CLS12 locus in C. lusitaniae was readily disrupted in a single transformation event, providing

25

additional molecular evidence that C. lusitaniae is haploid. C. lusitaniae cls12 mutant strains are sterile: We first tested the effects of the cls12 mutation on mating. For this purpose, cycloheximide-resistant mutants of the cls12::CaURA3 mutant strains were isolated to provide a genetic marker for selection. Two independent cycloheximide-resistant isolates were obtained (strains CL120 and CL121), and mating assays between the MAT␣ cls12::URA3 chx r strains, CL120 and CL121, and the MATa ura1 strain CL8 were conducted on SLAD medium. As controls for mating, the MAT␣ ura3 strain CL24 was crossed with the MATa chx r strain CL59, and the MAT␣ chx r strain CL62 was crossed with strain CL8 (MATa ura1), CL16 (MATa ura3), and CL17 (MATa ura3). Mating assays were replica-plated to medium containing 5-FOA and cycloheximide. Whereas the control matings readily yielded abundant recombinant progeny, there were either no or only very rare 5-FOA and cycloheximide-resistant progeny from the crosses involving one cls12 mutant parent and one CLS12 wild-type parent strain (Figure 6). Similar observations were obtained with a third cls12::URA3 mutant strain, CL117. In quantitative mating assays, the frequency of ura3 chx r recombinants produced by the control matings of the CLS12 wild-type strains CL24 and CL59 was 1.2 ⫻ 10⫺4, 5.1 ⫻ 10⫺5, and 3.8 ⫻ 10⫺5 (ave. ⵑ7 ⫻ 10⫺5) in three different experiments, and 2.8 ⫻ 10⫺5 with the control CLS12 wild-type strains CL17 and CL62. In contrast, the frequency of ura3 chx r recombinants when the cls12 mutant strains CL120 or CL121 were crossed to the CLS12 wild-type strain CL8 was ⬍4 ⫻ 10⫺7, ⬍1 ⫻ 10⫺9, and 4 ⫻ 10⫺9, indicating that the frequency of mating is reduced by ⬎10,000-fold in the cls12 mutant strains. Taken together, these observations demonstrate that the CLS12 gene is necessary for mating in C. lusitaniae. C. lusitaniae cls12 mutant strains have no defect in filamentation: C. lusitaniae can form pseudohyphae and hyphae on certain nutritional deprivation media, including V8 medium, filamentation agar, and SLAD medium. To test if the Ste12 homolog Cls12 plays a role in filamentous growth of C. lusitaniae, the isogenic CLS12 prototrophic wild-type strain CL3 and the cls12::CaURA3 mutant strains CL92 and CL93 were grown on V8 and filamentation agar. We observed no difference in the ability of the cls12 mutant strains to filament on either medium compared to the isogenic CLS12 wild-type parental strain following either short or long incubation periods (Figure 7 and data not shown). Moreover, a control strain in which the C. albicans URA3 gene had been ectopically integrated into the genome of the ura3 strain CL24 filamented to the same extent as the CLS12 wild-type and cls12 mutant strains on filamentation and V8 agar. C. lusitaniae does not filament on spider medium, which is commonly employed to analyze filamentous growth of C. albicans (not shown). In addition, while C. albicans forms germ tubes when cultured in

26


Figure 6.—cls12 mutant strains are sterile. A total of 107 cells of the MAT␣ chxr cls12 mutant strains (CL120, CL121) were coincubated with 107 cells of the MATa ura1 strain CL8 (top) or the MATa ura3 strain CL16 (bottom), and control matings consisting of CLS12 wild-type strains CL24 (MAT␣ ura3) with CL59 (MATa chxr) (top and bottom), CL62 (MAT␣ chxr) with CL8 (MATa ura1) (top), and CL62 (MAT␣ chxr) with CL16 (MATa ura3) (bottom) were coincubated on SLAD medium for 96 hr at 30⬚, replica-plated to 5-FOA medium containing cycloheximide (10 ␮g/ml), and incubated at 30⬚ for 96 hr.

YPD medium containing 10% fetal calf serum, C. lusitaniae does not (data not shown). We conclude that the Ste12 homolog Cls12 is not required for filamentation in C. lusitaniae.

DISCUSSION

We developed a mating assay to analyze the sexual cycle of C. lusitaniae. We cloned and sequenced the CLS12 gene encoding a homolog of the Ste12 transcription factor. To analyze the functions of Cls12, a transformation system was developed and the CLS12 gene was disrupted by homologous recombination. cls12 mutant strains were sterile but had no defect in filamentous growth. DNA sequence comparison and transformation experiments provided evidence that the CUG codon encodes serine and not leucine in C. lusitaniae, as in C. albicans. In summary, these studies illustrate the utility of genetic and molecular studies in a haploid pathogenic yeast with a sexual cycle that is evolutionarily related to the most common fungal pathogen, C. albicans. Relationship of Cls12 to known roles of Ste12 homologs in other organisms: We have identified the C. lusitaniae CLS12 gene, which encodes a homolog of the Ste12 homeodomain transcription factor that regulates mating, filamentation, invasive growth, and virulence in other fungi. By gene disruption, we find that the CLS12

gene is required for mating but dispensable for filamentation in C. lusitaniae. These findings are somewhat reminiscent of the functions of Ste12 in the model yeast S. cerevisiae, in which ste12 mutations confer an absolute mating defect (Hartwell 1980). However, ste12 mutations in S. cerevisiae do confer a demonstrable filamentous growth defect (Liu et al. 1993). In the human pathogen C. albicans, mutations in the Ste12 homolog Cph1 also reduce filamentation in response to some environmental signals but not others (Liu et al. 1994). Finally, in the pathogenic basidiomycete C. neoformans, mutants lacking the Ste12 homolog have little or no defect in mating but have an absolute defect in haploid fruiting, which is a differentiation cascade triggered in MAT␣ strains in response to nitrogen starvation (Yue et al. 1999; Chang et al. 2000). Taken together, our findings suggest that Ste12 homologs play divergent functions in the regulation of mating and filamentation in these diverse organisms. The C. lusitaniae Cls12 protein plays a function in mating that is shared with the Ste12 homolog in S. cerevisiae but not with the Ste12 homolog in C. neoformans. In contrast, the Ste12 and Cph1 proteins regulate filamentation in S. cerevisiae, C. albicans, and C. neoformans, but the Cls12 homolog is dispensable for filamentation in C. lusitaniae. These studies illustrate how functions of a conserved and ubiquitous transcrip-


Figure 7.—cls12 mutant strains have no defect in filamentation. The CLS12 wild-type strain (CL3), cls12::CaURA3 mutant strains (CL92 and CL93), and a ura3 strain with an ectopic integration of the CaURA3 gene were grown on filamentation and V8 agar at 30⬚ for 48 hr. The edges of the cell patches were photographed at ⫻10 magnification.

tion factor can be conserved and diverged during evolution, possibly in response to different evolutionary pressures and the ability of other transcription factors (Phd1, Efg1) to serve partially overlapping functions. Evolutionary relationship of C. lusitaniae to C. albicans: The genus Candida consists of many different yeast species that are typically asexual. However, the subsequent discovery of sexual forms of a variety of different Candida spp. has led to confusion because these organisms are now classified into two different genera based on their asexual and sexual forms (Hazen 1997). The relationship of these diverse Candida species to the most common human pathogenic form, C. albicans, has also been unclear. The majority of Candida infections are caused by C. albicans and several additional Candida species, including C. tropicalis and C. parapsilosis. Based on comparisons of 18S rRNA genes, these Candida species are evolutionarily closely related to C. albicans (Barns et al. 1991) and, in general, these organisms are diploid, have no known sexual cycle, and employ CUG codons to encode serine (Santos et al. 1997). By com-

27

parison, C. glabrata occurs relatively frequently in patients but is a haploid yeast with no known sexual cycle in which CUG codons encode leucine and that is more closely related to S. cerevisiae than to C. albicans. One other relatively frequent pathogen is C. krusei, which has a sexual teleomorph called Issatchenkia orientalis; this organism may also be more closely related to S. cerevisiae than to C. albicans. Two other less commonly encountered Candida pathogens are C. lusitaniae and C. guillermondii. By comparison of 18S rRNA genes, these organisms are more closely related to C. albicans than to S. cerevisiae (Barns et al. 1991). Moreover, both C. guillermondii (Pesole et al. 1995) and C. lusitaniae (our studies and those of Ueda et al. 1994) use CUG codons to encode serine. Thus, C. lusitaniae and C. guillermondii belong to a clade of related Candida species that infect humans and employ a divergent genetic code. Finally, our studies have identified a C. lusitaniae Ste12 homolog that is more closely related to the C. albicans Cph1 Ste12-homolog (78%) than to the Ste12 proteins from S. cerevisiae (67%) or K. lactis (71%). Because C. lusitaniae is haploid and has a sexual cycle, and classic mycological studies have also identified a sexual cycle in C. guillermondii, our findings suggest that the common ancestor to this group of pathogenic Candida spp. was sexual and that some descendants (C. albicans and others) have evolved to be asexual whereas others, such as C. lusitaniae, have maintained the sexual cycle. These considerations suggest that studies of the sexual cycle of C. lusitaniae may provide insights into a possible cryptic sexual cycle in C. albicans, which has thus far eluded discovery but may be responsible for genetic recombination in nature (Graser et al. 1996). Moreover, they indicate that the evolution of virulence and the loss of sexual proficiency are correlated and could be causally linked. C. lusitaniae as a model for fungal pathogenesis: C. lusitaniae is isolated from ⵑ1% of patients with candidemia and is now classified as an emerging fungal pathogen. Our studies illustrate the potential utility of this system for experimental studies on the sexual cycle and to elucidate the role of signal transduction cascades in the regulation of mating and filamentation. The utility of this system would be considerably extended by the development of animal models for studies of virulence. We have infected mice by tail vein injection with several of the wild-type clinical C. lusitaniae isolates studied here but have found no evidence for fatal infections in wildtype immunocompetent mice (L. Young, G. Cox and J. Heitman, unpublished results). These observations suggest that an immunocompromised host may be necessary to develop an animal model system for C. lusitaniae. Once such a system becomes available, we plan to conduct studies comparing the virulence of wild-type and cls12 mutant strains of C. lusitaniae to explore the possible role of the Cls12 homolog of Ste12 in virulence. Additional technical advances that would further en-

28


hance the utility of C. lusitaniae as a model system would be the development of congenic MAT␣ and MATa strains, episomal plasmids, additional markers for transformation and gene disruption, and the identification of regulatable promoters. The definition of conditions under which the diploid could be isolated, sporulated, and dissected would be useful in the analysis of essential genes. Finally, because C. lusitaniae is evolutionarily related to the most common fungal pathogen, C. albicans, further studies in this genetically tractable model system may provide insight into a past or cryptic sexual cycle in C. albicans and provide a facile system to explore the genetic regulation of virulence in a haploid organism. We thank Maria Cardenas, Rey Sia, Ping Wang, Jenifer Go¨rlach, Rob Davidson, and Cristina Cruz for valuable scientific advice and assistance, Miguel Arevalo-Rodriguez for assistance preparing figures, Wiley Schell for C. lusitaniae strains, and John McCusker and Alan Goldstein for advice, discussion, and plasmids. These studies were supported in part by National Institutes of Allergy and Infectious Disease R01 grant AI42159 and P01 grant AI44975 to the Duke University Mycology Research Center. Joseph Heitman is a Burroughs Wellcome Scholar in Molecular Pathogenic Mycology and an associate investigator of the Howard Hughes Medical Institute.

LITERATURE CITED Ahearn, D. G., and M. S. McGlohn, 1984 In vitro susceptibilities of sucrose-negative Candida tropicalis, Candida lusitaniae, and Candida norvegensis to Amphotericin B, 5-Fluorocytosine, Miconazole, and Ketoconazole. Clin. Microbiol. 19: 412–416. Alspaugh, J. A., J. R. Perfect and J. Heitman, 1998 Signal transduction pathways regulating differentiation and pathogenicity of Cryptococcus neoformans. Fungal Genet. Biol. 25: 1–14. Baker, J. G., H. L. Nadler, P. Forgacs and S. R. Kurtz, 1984 Candida lusitaniae: a new opportunistic pathogen of the urinary tract. Diagn. Microbiol. Infect. Dis. 2: 145–149. Barns, S. M., D. J. Lane, M. L. Sogin, C. Bibeau and W. G. Weisburg, 1991 Evolutionary relationships among pathogenic Candida species and relatives. J. Bacteriol. 173: 2250–2255. Blinkhorn, R. J., D. Adelstein and P. J. Spagnuolo, 1989 Emergence of a new opportunistic pathogen, Candida lusitaniae. J. Clin. Microbiol. 27: 236–240. 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. Braun, B. R., and A. D. Johnson, 1997 Control of filament formation in Candida albicans by the transcriptional repressor TUP1. Science 277: 105–109. Chang, Y. C., B. L. Wickes, G. F. Miller, L. A. Penoyer and K. J. Kwon-Chung, 2000 Cryptococcus neoformans STE12␣ regulates virulence but is not essential for mating. J. Exp. Med. 19: 871–881. Clark, K. L., P. J. Feldmann, D. Dignard, R. Larocque, A. J. Brown et al., 1995 Constitutive activation of the Saccharomyces cerevisiae mating response pathway by a MAP kinase kinase from Candida albicans. Mol. Gen. Genet. 249: 609–621. Coleman, D. C., M. G. Rinaldi, K. A. Haynes, J. H. Rex, R. C. Summerbell et al., 1998 Importance of Candida species other than Candida albicans as opportunistic pathogens. Med. Mycol. 36: 156–165. Csank, C., K. Schro¨ppel, E. Leberer, D. Harcus, O. Mohamed et al., 1998 Roles of the Candida albicans mitogen-activated protein kinase homolog, Cek1p, in hyphal development and systemic candidiasis. Infect. Immun. 66: 2713–2721. Dong, H., and W. Courchesne, 1998 A novel quantitative mating assay for the fungal pathogen Cryptococcus neoformans provides insight into signalling pathways responding to nutrients and temperature. Microbiology 144: 1691–1697.

Gargeya, I. B., W. R. Pruitt, R. B. Simmons, S. A. Meyer and D. G. Ahearn, 1990 Occurrence of Clavispora lusitaniae, the teleomorph of Candida lusitaniae, among clinical isolates. J. Clin. Microbiol. 28: 2224–2227. Goldstein, A. L., and J. H. McCusker, 1999 Three new dominant drug resistance cassettes for gene disruption in Saccharomyces cerevisiae. Yeast 15: 1541–1553. Goldstein, A. L., X. Pan and J. H. McCusker, 1999 Heterologous URA3MX cassettes for gene replacement in Saccharomyces cerevisiae. Yeast 15: 507–511. Graser, Y., M. Volovsek, J. Arrington, G. Schonian, W. Presber et al., 1996 Molecular markers reveal that population structure of the human pathogen Candida albicans exhibits both clonality and recombination. Proc. Natl. Acad. Sci. USA 93: 12473–12477. Haase, S. B., and D. J. Lew, 1997 Flow cytometric analysis of DNA content in budding yeast. Methods Enzymol. 283: 322–332. Hadfield, T. L., M. B. Smith, R. E. Winn, M. G. Rinaldi and C. Guerra, 1987 Mycoses caused by Candida lusitaniae. Rev. Infect. Dis. 9: 1006–1012. Hartwell, L. H., 1980 Mutants of Saccharomyces cerevisiae unresponsive to cell division control by polypeptide mating hormone. J. Cell Biol. 85: 811–822. Hazen, K. C., 1997 New and emerging yeast pathogens. Clin. Microbiol. Rev. 8: 462–478. Hoffman, C. S., and F. Winston, 1987 A ten-minute DNA preparation from yeast efficiently releases autonomous plasmids for transformation of Escherichia coli. Gene 57: 267–272. James, P., J. Halladay and E. A. Craig, 1996 Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast. Genetics 144: 1425–1436. Kohler, J. R., and G. R. Fink, 1996 Candida albicans strains heterozygous and homozygous for mutations in mitogen-activated protein kinase signalling components have defects in hyphal development. Proc. Nat. Acad. Sci. USA 93: 13223–13228. Kwon-Chung, K. J., and J. E. Bennett, 1992 Cryptococcosis, pp. 397–446 in Medical Mycology. Lea & Febiger, Malvern, PA. Leberer, E., D. Harcus, I. D. Broadbent, K. L. Clark, D. Dignard et al., 1996 Signal transduction through homologs of the Ste20p and Ste7p protein kinases can trigger hyphal formation in the pathogenic fungus Candida albicans. Proc. Natl. Acad. Sci. USA 93: 13217–13222. Liu, H., C. A. Styles and G. R. Fink, 1993 Elements of the yeast pheromone response pathway required for filamentous growth of diploids. Science 262: 1741–1744. Liu, H., J. Ko¨hler and G. R. Fink, 1994 Suppression of hyphal formation in Candida albicans by mutation of a STE12 homolog. Science 266: 1723–1726. Lo, H.-J., J. R. Kohler, B. DiDomenico, D. Loebenberg, A. Cacciapuoti et al., 1997 Nonfilamentous C. albicans mutants are avirulent. Cell 90: 939–949. Merz, W. G., 1984 Candida lusitaniae : frequency of recovery, colonization, infection, and amphotericin B resistance. J. Clin. Microbiol. 20: 1194–1195. Nguyen, M. H., A. J. Morris, M. E. Dobson, D. R. Snydman, J. J. E. Peacock et al., 1996 Candida lusitaniae: an important emerging cause of Candidemia. Infect. Dis. Clin. Pract. 5: 273–278. Orr-Weaver, T. L., J. W. Szostak and R. J. Rothstein, 1981 Yeast transformation: a model system for the study of recombination. Proc. Natl. Acad. Sci. USA 78: 6354–6358. Pappagianis, D., M. S. Collins, R. Hector and J. Remington, 1979 Development of resistance to amphotericin B in Candida lusitaniae infecting a human. Antimicrob. Agents Chemother. 16: 123–126. Pesole, G., M. Lotti, L. Alberghina and C. Saccone, 1995 Evolutionary origin of nonuniversal CUGSer codon in some Candida species as inferred from a molecular phylogeny. Genetics 141: 903–907. Sambrook, J., E. F. Fritsch and T. Maniatis, 1989 Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sanchez, V., J. A. Vazquez, D. Barth-Jones, L. Dembry, J. D. Sobel et al., 1992 Epidemiology of nosocomial acquisition of Candida lusitaniae. Clin. Microbiol. 30: 3005–3008. Santos, M. A. S., G. Keith and M. F. Tuite, 1993 Non-standard translational events in Candida albicans mediated by an unusual

Mating in Candida lusitaniae seryl-tRNA with a 5⬘-CAG-3⬘ (leucine) anticodon. EMBO J. 12: 607–616. Santos, M. A. S., T. Ueda, K. Watanabe and M. F. Tuite, 1997 The non-standard genetic code of Candida spp.: an evolving genetic code or a novel mechanism for adaptation? Mol. Microbiol. 26: 423–431. Scherer, S., and P. T. Magee, 1990 Genetics of Candida albicians. Microbiol. Rev. 54: 226–241. Schweingruber, A. M., N. Hilti, E. Edenharter and M. E. Schweingruber, 1998 Methionine induces sexual development in the fission yeast Schizosaccharomyces pombe via an ste11-dependent signalling pathway. J. Bacteriol. 180: 6338–6341. Shah-Mahoney, N., T. Hampton, R. Vidaver and D. Ratner, 1997 Blocking the ends of transforming DNA enhances gene targeting in Dictyostelium. Gene 203: 33–41. Sherman, F., 1991 Getting started with yeast. Methods Enzymol. 194: 3–21. Stoldt, V. R., A. Sonneborn, C. E. Leuker and J. F. Ernst, 1997 Efg1, an essential regulator of morphogenesis of the human pathogen Candida albicans, is a member of a conserved class of bHLH proteins regulating morphogenetic processes in fungi. EMBO J. 16: 1982–1991. Suzuki, T., T. Ueda and K. Watanabe, 1997 The ‘polysemous’ codon—a codon with multiple amino acid assignment caused by dual specificity of tRNA identity. EMBO J. 16: 1122–1134. Toffaletti, D. L., T. H. Rude, S. A. Johnston, D. T. Durack and

29

J. R. Perfect, 1993 Gene transfer in Cryptococcus neoformans by use of biolistic delivery of DNA. J. Bacteriol. 175: 1405–1411. Ueda, T., R. Suzuki, T. Yokogawa, K. Nishikawa and K. Watanabe, 1994 Unique structure of new serine tRNAs responsible for decoding leucine codon CUG in various Candida species and their putative ancestral tRNA genes. Biochimie 76: 1217–1222. van Uden, N., and H. Buckley, 1970 Candida Berkhout, pp. 893– 1087 in The Yeasts: A Taxonomic Study, edited by J. Lodder. North Holland, Amsterdam. Wickes, B. L., M. E. Mayorga, U. Edman and J. C. Edman, 1996 Dimorphism and haploid fruiting in Cryptococcus neoformans : association with the ␣-mating type. Proc. Natl. Acad. Sci. USA 93: 7327– 7331. Yinnon, A. M., K. A. Woodin and K. R. Powell, 1992 Candida lusitaniae infection in the newborn: case report and review of the literature. Pediatr. Infect. Dis. 11: 878–880. Yoon, S. A., J. A. Vazquez, P. E. Steffan, J. D. Sobel and R. A. Akins, 1999 High-frequency, in vitro reversible switching of Candida lusitaniae clinical isolates from amphotericin B susceptibility to resistance. Antimicrob. Agents Chemother. 43: 836–845. Yue, C., L. M. Cavallo, J. A. Alspaugh, P. Wang, G. M. Cox et al., 1999 The STE12␣ homolog is required for haploid filamentation but largely dispensable for mating and virulence in Cryptococcus neoformans. Genetics 153: 1601–1615. Communicating editor: A. P. Mitchell