Adenylyl Cyclase-Associated Protein Aca1 Regulates Virulence and Differentiation of Cryptococcus neoformans via the Cyclic AMP-Protein Kinase A Cascade^†

Strains and media.

The strains used in the present study are listed in Table 1. Yeast extract-peptone-dextrose (YPD) and synthetic (SD) media, V8 mating medium (pH 5.0), Niger seed and l-DOPA media for melanin production and Dulbecco modified Eagle (DME) medium for capsule production were as described previously (2, 20, 21, 46). Agar-based DME medium for capsule production was prepared by combining filter-sterilized 2× DME liquid medium (pH 7.2) with autoclaved 2% agar solution. Escherichia coli TOP10 cells (Invitrogen) served as the host strain for transformation and propagation of all plasmids used in the present study.

Identification of the 5′ and 3′ regions of the ACA1 gene.

Strain H99 was incubated overnight at 30°C in YPD medium, pelleted, lyophilized, and total RNA was isolated with TRIzol (Gibco-BRL) according to the manufacturer's instruction. 5′ and 3′ rapid amplification of cDNA ends (RACE) was performed with the GeneRacer kit (Invitrogen) according to the manufacturer's instructions. Each RACE product was cloned into the pCR2.1-TOPO vector (Invitrogen) and sequenced.

Yeast two-hybrid assay.

To construct plasmids expressing the Gal4 DNA binding domain (BD) or activation domain (AD) fused to the full-length ACA1 open reading frame (ORF), the ACA1 gene was amplified by reverse transcription-PCR (RT-PCR) with the primers 13086 and 13087 (13086/13087; see Table S1 in the supplemental material for each primer sequence) and cloned into two-hybrid vectors pGBT9 and pGAD424, generating pGBT-ACA1 and pGAD-ACA1. Similarly, the C-terminal region (2126 to 2260 amino acids [aa]) of CAC1 was amplified by using the primers 12873/12874 and cloned into pGBT9 and pGAD424, generating pGBT-CAC1_2126-2260 and pGAD-CAC1_2126-2260, respectively. The full-length GPA1 and RAS1 ORFs were also generated by RT-PCR by using primers 10961/10962 for GPA1 and primers 12875/12876 for RAS1 and cloned into pGBT9, generating pGBT-GPA1 and pGBT-RAS1, respectively. The reporter yeast strain PJ69-4A was cotransformed, and at least three independent Leu⁺ Trp⁺ transformants grown on SD medium lacking leucine and tryptophan (SD−Leu−Trp) were further analyzed for growth ability on SD−Leu−Trp−His and SD−Leu−Trp−His−Ade media.

Disruption of the ACA1 gene.

The aca1-null mutant (aca1Δ) was generated in the congenic C. neoformans serotype A MATα (H99) and MATa (KN99) strain background (35) by PCR overlap as previously described (10). In the first round of PCR, the 5′ and 3′ regions of the ACA1 gene were generated by PCR (ExTaq; Takara Co.) by using H99 or KN99 genomic DNA and primers (Fig. 2): 10341/10342 for the 5′ end and 10343/10344 for the 3′ end. Nat^r or Neo^r dominant selectable markers were also generated by PCR with M13 forward (M13F) and reverse (M13R) primers by using plasmid pNATSTM#43 with a unique signature tag (kindly provided by Jennifer K. Lodge, St. Louis University School of Medicine; tag sequences available upon request) or pJAF1 (17) as templates, respectively. ACA1 disruption fragments were generated by PCR overlap with the primers 10341/10344. A 4.3-kb ACA1 disruption construct was gel extracted and precipitated onto 600-μg gold microcarrier beads (0.8 μm; Bioworld, Inc.) and biolistically transformed into strain H99 or KN99 as described previously (11). Stable transformants were selected on YPD medium containing nourseothricin (100 mg/liter) or G418 (200 mg/liter).

To screen aca1Δ mutant strains, diagnostic PCR was performed by analyzing the 5′ junction of disrupted aca1Δ alleles with primers 10345 and 8994. Screened transformants were further confirmed by Southern blot analysis with a PstI-digested genomic DNA and an ACA1-specific probe made by PCR with primers 10409 and 10410. Uracil auxotrophic aca1Δ mutants (YSB108 and YSB109) (Table 1) were generated by inducing spontaneous ura5 mutations in strains YSB6 and YSB58, respectively, on SD medium containing 5-fluoroorotic acid (5-FOA).

To construct the aca1Δ+ACA1 reconstituted strains, H99 genomic DNA containing the entire ACA1 gene was isolated from a C. neoformans H99 bacterial artificial chromosome (BAC) library using the ACA1 specific probe described above. The 3-kb SalI-HindIII fragment containing the ACA1 gene (Fig. 1) was cloned into plasmid pJAF12 containing the Neo^r marker, generating pNEOSRV2. Plasmids pNATSRV2 and pURASRV2 were further constructed by inserting the 3-kb XhoI-HindIII fragment of pNEOSRV2 into pJAF13 or pJAF7 containing a Nat^r or URA5 selectable marker, respectively. SbfI-linearized plasmids pNEOSRV2, pNATSRV2, or pURASRV2 were biolistically transformed into strains YSB6, YSB58, or ura5 YSB108/YSB109 (Table 1), respectively.

RT-PCR confirmation of ACA1 disruption and reconstitution.

First-strand cDNA was generated by using 5 μg of DNase I-treated total RNA according to the manufacturer's instructions (SuperScript III First-Strand Synthesis System for RT-PCR; Invitrogen). To detect the presence or absence of ACA1 messages, a portion of the ACA1 gene (733 to 1,150 bp of a 1,521-bp coding sequence) was PCR amplified (418 bp) from the first-strand cDNA by using ACA1-specific primers 11669/11974. As a control, a portion of the CAC1 gene (53 to 1,275 bp of a 6,813-bp coding sequence) was PCR amplified (1,223 bp) from the same first-strand cDNA by using CAC1-specific primers 11614/11615.

Disruption of the GPA1, CAC1, RAS1, GPB1, PKA1, and PKA2 genes in the H99 and KN99 strain backgrounds.

Although the GPA1, CAC1, RAS1, GPB1, PKA1, and PKA2 genes have been previously disrupted and characterized in serotype A C. neoformans (1-3, 50), these mutants were isolated in mutagenized auxotrophic H99 derivatives that may contain other unintended mutations. Therefore, these genes were disrupted in the congenic prototrophic H99 and KN99 genetic background by using methods equivalent to those described above. The 5′ regions of each gene were generated with the following primers: 10913/11736 for GPA1, 11608/11609 for CAC1, 11600/11601 for RAS1, 11592/11593 for GPB1, 12911/12912 for PKA1, and 12916/12917 for PKA2. The 3′ regions for each gene were generated with the following primer pairs: 11737/10957 for GPA1, 11610/11611 for CAC1, 11602/11603 for RAS1, 11594/11595 for GPB1, 12913/12936 for PKA1, and 12918/12919 for PKA2. The Neo^r marker was generated as described above. Nat^r markers were similarly generated with the following plasmids as templates: pNATSTM#146 for GPB1, pNATSTM#150 for RAS1, pNATSTM#5 for GPA1, pNATSTM#159 for CAC1, pNATSTM#191 for PKA1, and pNATSTM#205 for PKA2. Overlap PCR and biolistic transformation were performed as described above and the genotype of each mutant was confirmed by diagnostic PCR and Southern blot (not shown).

Construction of serotype A aca1Δ gpa1Δ, aca1Δ cac1Δ, aca1Δ ras1Δ, ras1Δ cac1Δ, pka1Δ pka2Δ, and crg1Δ aca1Δ double-mutant strains.

For epistasis analysis, the GPA1, CAC1, and RAS1 genes were further disrupted in MATα aca1Δ (YSB6) and MATa aca1Δ (YSB58) mutant strains by using the same disruption cassette described above, generating MATα and MATa aca1Δ gpa1Δ, aca1Δ cac1Δ, and aca1Δ ras1Δ double-mutant strains (Table 1). The RAS1 gene was also disrupted in MATα cac1Δ (YSB42) and MATa cac1Δ (YSB79) mutant strains, creating ras1Δ cac1Δ double mutants (Table 1). The PKA2 gene was further disrupted in the MATα pka1Δ mutant (YSB188), generating pka1Δ pka2Δ double mutants (Table 1). Stable double-mutant strains (Nat^r Neo^r) were selected on YPD medium containing both nourseothricin and G418. Positive transformants were further screened by diagnostic PCR and confirmed by Southern blot analysis as described. The MATa crg1Δ aca1Δ mutant (YSB96) is an F₁ progeny from a meiotic cross between strains JKH43 (MATa pde2Δ::NAT crg1Δ::URA5, unpublished) and YSB58. Strain JKH43 is an F₁ progeny from a meiotic cross between strains JKH33 (MATα pde2Δ::NAT [unpublished data]) and PPW196.

Mating, fusion, and confrontation assays.

Mating experiments were performed by coincubating cells of opposite mating type on V8 mating medium with or without 1 or 10 mM cAMP in the dark at room temperature for 1 to 4 weeks.

Cell fusion efficiency was measured as previously described (21) with minor modifications. A total of 10⁷ cells of each MATα and MATa strain bearing Nat^r or Neo^r markers, respectively, per ml was mixed in an equal volume, and 5 μl of this cell mixture was spotted onto V8 medium, followed by incubation for 24 h at room temperature in the dark. The cells were scraped and resuspended in 1 ml of distilled H₂O, and 20 μl of cell suspension was plated onto YPD medium containing nourseothricin and G418. The number of colonies on each plate was determined after 4 days of incubation at room temperature.

For the confrontation assay to monitor pheromone production and response of cells, α cells were streaked in confrontation with a cells on filament agar, followed by incubation for 5 days at room temperature in the dark. Images of mating and confrontation assays were captured with a Nikon Eclipse E400 microscope equipped with a Nikon CoolPix digital camera.

Assay for capsule production.

Each strain (a single colony from solid YPD medium) was incubated for 16 h at 30°C in liquid YPD medium, spotted (3 × 10⁵ cells) onto agar-based DME medium, and further incubated for 24 h at 37°C. After incubation, the capsule was stained by India ink and observed microscopically. Differential interference microscopy for cell and capsule images was performed with a Zeiss Axioskop 2 equipped with an AxioCam MRM digital camera. Quantitative measurement of capsule size was performed as previously described (57) by microscopically measuring the diameters of the capsule and the cell by using AxioVision 3.1 software (Zeiss). The relative capsule diameter, calculated as (Dw − Dc × 100)/Dw, where Dw and Dc indicate the diameters of the whole-cell body (cell plus capsule) and cell body only, respectively, was statistically compared between each mutant and wild-type strains by using the Bonferroni multiple comparison test performed with Prism 4.0 (GraphPad Software) (57).

Assay for melanin production.

Each strain was incubated for 16 h at 30°C in YPD medium, spotted (3 × 10⁵ cells) onto Niger seed medium containing 0.1% glucose, and incubated for 3 days at 30°C or 37°C. Melanin production was monitored daily and photographed. Quantitative measurements of laccase activity were performed as previously described (21). Briefly, 10⁸ cells of each strain were grown in 25 ml of l-DOPA medium at 30°C for 16 h. The cultures were shifted to 25°C and further incubated for 6 and 24 h; then, 1 ml of culture was centrifuged, and the optical density at 475 nm (OD₄₇₅) of the supernatant was measured with a spectrophotometer.

cAMP assays.

Cells from a single colony grown on solid YPD medium were incubated at 30°C for 20 h in YPD medium. Cells were collected by centrifugation and washed twice with water and once with MES buffer (10 mM morpholineethanesulfonic acid [pH 6.0], 0.1 mM EDTA). Cells were resuspended in MES buffer at an OD₆₀₀ of 2.0 and further incubated at 30°C with shaking for glucose starvation. After 2 h of incubation, glucose was added to a final concentration of 2%. At various time points (0, 0.5, 1, and 3 min), 1 ml of cell suspension was sampled, filtered through 0.45-μm-pore-size membrane filters (Millipore), immediately extracted with butanol-saturated formic acid (4.1% [vol/vol]), and lyophilized. The intracellular cAMP concentration was determined by using a cAMP [¹²⁵I] Direct Biotrak Scintillation Proximity Assay System (Amersham Biosciences) and normalized to the wet weight of cells.

Virulence assays.

Yeast strains were grown at 30°C in YPD medium overnight with shaking (250 rpm), collected by centrifugation, and washed twice with sterile phosphate-buffered saline (PBS), and the final concentration was adjusted to 2 × 10⁶ CFU/ml with sterile PBS. Female A/Jcr mice (NCI/Charles River Laboratories) mice (20 to 24 g each) were infected with wild-type (H99, six mice), aca1Δ (YSB6, nine mice), aca1Δ+ACA1 reconstituted (YSB117, ten mice), and cac1Δ (YSB42, 10 mice) mutant strains by intranasal inhalation. Each mouse was inoculated with 10⁵ CFU, in a volume of 50 μl, via nasal inhalation as previously described (9). Mice that appeared moribund (i.e., lethargic, rapid weight loss, or in pain) were sacrificed by using CO₂ inhalation. Survival data from the murine experiments were statistically analyzed between paired groups by using the log-rank test with the Prism program 4.0. The animal protocol used for these experiments was approved by the Duke University Animal Use Committee.

Nucleotide sequence accession number.

The sequence for the ACA1 gene from strain H99 has been assigned GenBank accession number AY629559.

Disruption of the ACA1 gene and recapitulation of gpb1, ras1, gpa1, cac1, pka1, and pka2 mutations in the congenic serotype A MATα and MATa strains.

To characterize the function of Aca1 in known signaling pathways of C. neoformans, the ACA1 gene was disrupted. aca1Δ::NAT and aca1Δ::NEO disruption alleles were introduced into strains H99 and KN99, respectively, by biolistic transformation, deleting an internal 1,880-bp (bp 1 to 1880 from the start) of the 1,908-bp ORF (Fig. 2A). The genotypes of two independent MATα aca1Δ mutants (YSB6 and YSB7) and three independent MATa aca1Δ mutants (YSB58, YSB59, and YSB60) were verified by diagnostic PCR (not shown) and Southern blot analysis (Fig. 2B). The wild-type ACA1 gene was reintroduced into the aca1Δ mutants by targeting to the ACA1 locus with linearized ACA1-NEO or ACA1-NAT constructs (Fig. 2B). Successful disruption and reintegration of ACA1 were further confirmed by RT-PCR analysis, demonstrating the loss and recovery of ACA1 expression in the aca1Δ mutant and aca1Δ+ACA1 reconstituted strains, respectively (Fig. 2C). All of the independent MATα or MATa aca1Δ mutant strains showed identical in vitro phenotypes, and representative results from strains YSB6 (MATα aca1Δ) and YSB58 (MATa aca1Δ) (Table 1) are shown here.

For comparative phenotypic analysis of aca1Δ mutants with other cAMP-dependent or -independent mutants, we then redisrupted the GPB1, RAS1, GPA1, CAC1, PKA1, and PKA2 genes in the congenic serotype A MATα H99 and MATa KN99 strains, which were recently generated by Nielsen et al. (35). We regenerated these mutant strains for the following reasons. First, the previous mutants had been isolated in different mutagenized auxotrophic parental strain backgrounds (ade2 or ura5), which complicates comparative phenotypic analysis between those mutants and mutants generated directly in H99 or KN99 by using dominant selectable markers. Second, utilization of auxotrophic strains as parental strains for gene disruption might cause misinterpretation of phenotypes through direct or indirect effects of auxotrophic mutations or unexpected background mutations introduced during their generation by UV or gamma irradiation. Finally, the GPB1, RAS1, GPA1, CAC1, PKA1, and PKA2 genes had not been disrupted previously in the KN99 MATa genetic background. Disruption of these genes in the congenic MATa background makes it possible to investigate serotype A α versus a bilateral mating efficiency, instead of by using heterologous serotype D mating-type tester strains.

We regenerated all mutants by using the Nat^r and Neo^r dominant selectable markers in the congenic H99 and KN99 backgrounds as described in Materials and Methods. In general, newly constructed MATα and MATa gpb1Δ (YSB49 and YSB76), ras1Δ (YSB51 and YSB73), gpa1Δ (YSB83 and YSB85), cac1Δ (YSB42 and YSB79), pka1Δ (YSB188 and YSB191), and pka2Δ (YSB194 and YSB198) mutants (Table 1) displayed phenotypes comparable to mutants generated previously (1-3, 50), in terms of defective mating (gpb1Δ, ras1Δ, gpa1Δ, cac1Δ, and pka1Δ mutants), reduced capsule and melanin production (gpa1Δ, cac1Δ, and pka1Δ mutants), or inability to grow at high temperature (ras1Δ mutants) (data not shown). The genotypes for each mutant strain were confirmed by Southern blot analyses and at least two independent mutants for each gene exhibited identical phenotypes.

Aca1 is not essential for growth at high temperature.

C. neoformans ras1Δ mutants exhibit a temperature-sensitive growth defect (1). Based on the functional connection between Ras1 and CAP/Srv2 in S. cerevisiae, we tested growth of aca1Δ mutants at high temperature to address the question of whether Aca1 is involved in Ras1 signaling. ras1Δ mutants newly constructed in the congenic H99 and KN99 background (YSB51 and YSB73) showed severe growth defects at 37 to 39°C (not shown), in accord with previous studies (1, 52). In contrast, aca1Δ mutants had no growth defect at 30, 37, or 39°C (not shown). MATα and MATa wild-type (H99 and KN99), aca1Δ mutants (YSB6 and YSB58), and aca1Δ+ACA1 reconstituted strains (YSB117 and YSB118) showed similar growth rates in YPD medium at either 30 or 37°C. When the RAS1 gene was disrupted in the aca1Δ (aca1Δ ras1Δ mutants YSB174 and YSB175) or cac1Δ (ras1Δ cac1Δ mutants YSB182 and YSB185) mutant background, the temperature-sensitive growth defect conferred by the ras1Δ mutant was readily apparent at 37° and 39°C. We conclude that Aca1 is not involved in the Ras1-specific signaling pathway regulating growth at high temperature.

Aca1 promotes mating via the cAMP-signaling pathway in parallel with Gpa1.

In previous studies, Cac1 and Gpa1 were found to be involved in mating (2, 3). We hypothesized that Aca1 should play a similar role in mating if it is also a component of the cAMP signaling pathway. In unilateral mating crosses with the wild-type strains (H99 or KN99) of opposite mating type, both MATα and MATa aca1Δ mutants mated less efficiently on V8 medium than did the wild type (Fig. 3A). The defect in unilateral mating was more pronounced at earlier time points, and basidiospores were observed at later time points (Fig. 3A). aca1Δ mutants displayed more dramatic defects in bilateral mating crosses (aca1 × aca1) in which only very short filaments were produced within the mating patches but no basidiospores were observed, even after 4 weeks of incubation (Fig. 3A).

To test whether the mating defects observed with aca1Δ mutants result from defects in cell fusion or mating filamentation, cell fusion efficiency was determined with Nat^r or Neo^r marked aca1Δ mutants (YSB6 and YSB58) and normalized to the fusion of control strains (YSB119 and YSB121, aca1Δ+ACA1 reconstituted strains). aca1Δ mutants were less efficient in cell fusion compared than the wild-type strain (Fig. 3B) but still yielded fusion products (11.9% ± 1.4%), indicating that the aca1 disruption results in both defective mating filamentation and reduced cell fusion frequency. In support of this conclusion, all aca1/aca1 diploid strains isolated from cell fusion assays exhibited less prolific filamentous growth compared to wild-type diploid control strains (Fig. 3C).

Defects in filamentous growth conferred by the aca1 mutation were further assayed by using confrontation assays. For the present study, we used pheromone hypersensitive crg1Δ mutants, which lack an RGS protein that normally desensitizes cells to pheromone exposure (35, 47). When MATα crg1Δ mutants were confronted with MATa crg1Δ mutants, both cell types formed filaments toward the mating partner in response to pheromone (Fig. 3D). Disruption of ACA1 in the MATa crg1Δ mutant (aca1Δ crg1Δ double mutants; Table 1) still resulted in normal conjugation tube formation from confronting α crg1Δ cells but defective filamentous growth from the a aca1Δ crg1Δ mutant (Fig. 3D), indicating that Aca1 is not required for pheromone production but is involved in filamentation during response to pheromone.

To further determine whether Aca1 regulates mating in a cAMP-dependent manner, the mating defect of aca1Δ mutants was compared to those of gpb1Δ and ras1Δ (cAMP-independent) versus gpa1Δ and cac1Δ (cAMP-dependent) mutants. We found that gpb1Δ, ras1Δ, gpa1Δ, and cac1Δ mutants all showed more severe unilateral mating defects than aca1Δ mutants (Fig. 4A). The addition of 1 or 10 mM exogenous cAMP, however, greatly enhanced unilateral mating and bilateral mating, respectively, of gpa1Δ, cac1Δ, and aca1Δ mutant cells to levels comparable to wild type (Fig. 4A). In contrast, unilateral or bilateral mating defects of ras1Δ and gpb1Δ mutants were not rescued by 1 or 10 mM cAMP (Fig. 4B), further confirming that Ras1 and Gpb1 transmit mating signals through cAMP-independent signaling pathways.

Disruption of the GPA1 gene resulted in reduced cell fusion, although this was less severe compared to aca1Δ mutants (Fig. 3B). cac1Δ mutants showed even more severe cell fusion defects than either gpa1Δ or aca1Δ mutants (Fig. 3B). In contrast to the less severe cell fusion defect observed in gpa1Δ mutants compared to aca1Δ mutants, gpa1/gpa1 diploid strains were completely blocked in filamentous growth, as were cac1/cac1 diploid strains (Fig. 3C). These results suggest that Gpa1 and Aca1 independently contribute to activate the Cac1-cAMP pathway during the mating process. Compared to these cAMP pathway mutants, disruption of the RAS1 or GPB1 genes conferred a complete lack of cell fusion even in unilateral mating, which contrasts with the reduced but apparent cell fusion events in cells lacking components of the cAMP pathway (Fig. 3B). Therefore, the data indicate that the Cac1-cAMP pathway controls mating, is independently activated by either (or both) Aca1 or Gpa1, and functions in a pathway distinct from Ras1 and Gpb1.

Previous studies identified the two PKA catalytic subunits Pka1 and Pka2, and Pka1 plays the major role in mating and filamentous growth in serotype A (14, 21). In accord with these results, pka1Δ mutant strains (YSB188 and YSB191) displayed a severe mating defect (particularly in bilateral matings), whereas pka2Δ mutant strains (YSB194 and YSB198) had no mating defect (Fig. 4B). However, the finding that the pka1Δ mutant showed less severe unilateral mating defects than the cac1Δ mutant suggested that Pka2 might play a limited role in mating in pka1Δ mutant cells (Fig. 4). In fact, pka1Δ pka2Δ double-mutant strains showed a more severe unilateral mating defect than pka1Δ mutant cells, similar to the cac1Δ mutant, suggesting that the mating signal from Cac1 is bifurcated into Pka1 and Pka2, with Pka1 serving as the principal signaling element of the cAMP pathway.

Aca1 is required for capsule production via the Cac1-cAMP-Pka1-dependent signaling pathway.

Previous studies showed that the Gpa1-Cac1-cAMP-Pka1 signaling pathway is the major cascade for capsule regulation (2, 3), whereas Ras1 is not required (1). Consistent with our assignment of Aca1 as a component of the cAMP-signaling pathway, Aca1 was required for capsule production (Fig. 5). aca1Δ mutants were highly defective in capsule production in response to a variety of capsule-inducing signals, including agar-based DME (Fig. 5), low iron, and 10% serum (not shown). These mutants were hypocapsular and not acapsular, and a residual level of capsule was still observed, similar to other cAMP pathway mutants (i.e., gpa1Δ, cac1Δ, or pka1Δ) (Fig. 5). Based on quantitative measurements of relative capsule sizes, the defect in capsule production of aca1Δ mutants was comparable to those of gpa1Δ, cac1Δ and pka1Δ (Fig. 5B). Addition of 10 mM cAMP to agar-based DME media completely restored capsule production in aca1Δ, gpa1Δ, and cac1Δ mutants but not of pka1Δ mutants, demonstrating that Aca1 functions upstream of Pka1 in the cAMP pathway.

To test by epistasis analysis whether Aca1 and Cac1 function in a linear fashion in the cAMP-PKA pathway, aca1Δ cac1Δ double-mutant strains (YSB42 and YSB79) were generated (Table 1). aca1Δ cac1Δ double mutants did not display any additive defects in capsule production relative to aca1Δ or cac1Δ single-mutant strains and still produced minimal capsules compared to a complete lack of capsule in the acapsular strain cap59Δ (Fig. 5). The data suggest then that Aca1 signals in a linear fashion with Cac1 and Pka1. As reported (1), Ras1 is not involved in capsule production, and aca1Δ ras1Δ double mutants exhibited the same level of capsule defect observed with aca1Δ single mutants (Fig. 5). Also in accord with a previous study (21), Pka2 was found to be completely dispensable for capsule production. pka2Δ mutants produced wild-type levels of capsule, and pka1Δ pka2Δ double mutants (YSB200) did not exhibit any additive defects in capsule production compared to pka1Δ single-mutant strains (data not shown). In conclusion, Aca1 transmits capsule inducing signals through the Cac1-cAMP-PKA-dependent signaling pathway.

Aca1 and Gpa1 individually contribute to melanin production via the Cac1-cAMP-PKA-dependent pathway.

Melanin plays a role as an antioxidant during host infection and serves as a virulence factor regulated by the Gpa1-Cac1-cAMP-PKA signaling pathway. Therefore, we investigated the role of Aca1 in melanin production, both by visual inspection of melanin accumulation in cells on Niger seed medium and by quantitative measurements of laccase activity. Disruption of the ACA1 gene attenuated melanin production, and this defect was more apparent at 37°C than at 30°C (Fig. 6A). The defect of the aca1Δ mutant was comparable to gpa1Δ and cac1Δ mutants at 37°C. Melanin synthesis in the aca1Δ, gpa1Δ, and cac1Δ mutants was restored by exogenous cAMP (10 mM), further confirming that Aca1, Gpa1, and Cac1 function in a cAMP-dependent manner. Interestingly, we found that gpa1Δ and aca1Δ mutants exhibited only minor defects in melanin biosynthesis at 30°C compared to a more pronounced defect of cac1Δ mutants (Fig. 6A). Quantitative measurement of laccase activity in l-DOPA medium further confirmed that aca1Δ and gpa1Δ mutants are deficient in melanin production compared to wild-type, but neither is as defective as cac1Δ mutants (Fig. 6B). These data suggest that Cac1 requires multiple inputs to activate the synthesis of melanin.

We hypothesized that Gpa1 and Aca1 provide parallel inputs to Cac1 for melanin production. gpa1Δ aca1Δ double-mutant strains showed a more severe melanin defect than either gpa1Δ or aca1Δ single mutant, and this synergistic defect was comparable to cac1Δ mutants on both Niger seed and l-DOPA media (Fig. 6). The melanin defect of gpa1Δ aca1Δ mutants was completely rescued by addition of exogenous cAMP (10 mM) (Fig. 6A). In contrast, introducing an aca1 mutation into the cac1Δ mutant background did not exacerbate the melanin defect of cac1Δ mutant cells (Fig. 6). These findings support a model in which Gpa1 and Aca1 are two parallel upstream components activating the Cac1-cAMP-PKA signaling pathway for production of melanin.

As previously reported (21), pka1Δ mutants displayed defects in melanin synthesis similar to other cAMP cascade mutants at 37°C on Niger seed medium, which could not be rescued by exogenous cAMP (Fig. 6A). We note however that pka1Δ mutants showed a less severe defect in melanin accumulation at 30°C on Niger seed medium than cac1Δ mutants, implicating the potential role of the other catalytic subunit Pka2 in this pathway (Fig. 6A). Supporting this hypothesis, pka1Δ pka2Δ double mutants were found to be more melanin defective in both 30 and 37°C compared to pka1Δ mutants, similar to cac1Δ or aca1Δ gpa1Δ double mutants. Similarly, pka1Δ pka2 double mutants largely failed to produce melanin in l-DOPA medium similar to cac1Δ mutants (Fig. 6B). The pka2Δ single-mutant strain exhibited wild-type melanin production (Fig. 6), as reported previously (21). These data demonstrate that Pka1 plays the predominant role in cAMP signaling to control melanin synthesis but, in its absence, Pka2 can fulfill a limited signaling capacity.

Aca1 is required for glucose-induced but not basal cAMP levels.

The role of Aca1 in controlling cAMP signaling in conjunction with Gpa1 and Cac1 was confirmed by measuring cAMP levels during glucose sensing. In accord with a previous report (3), cAMP levels in the wild type rapidly increased after glucose readdition to glucose-starved cells (Fig. 7). In contrast, the gpa1Δ and cac1Δ mutant exhibited significantly lower cAMP concentrations at all time points and lacked any detectable cAMP pulse (Fig. 7), and thus both Gpa1 and Cac1 are required for both glucose-induced cAMP signaling and maintenance of basal cAMP levels. The aca1Δ mutant was defective in induced cAMP production but maintained basal cAMP levels equivalent to the wild-type at the zero time point (Fig. 7). cAMP levels in the aca1Δ mutant were significantly higher than those in the gpa1Δ and cac1Δ mutants, indicating that Aca1 plays a role in glucose-induced but not basal cAMP levels. Reintroduction of the ACA1 gene restored a wild-type cAMP signaling pattern in the aca1Δ mutant (Fig. 7). The basal cAMP levels in the aca1Δ mutant were further reduced to the levels of the cac1Δ mutant after disruption of the GPA1 gene (aca1Δ gpa1Δ) (Fig. 7), a finding consistent with the identical in vitro phenotypes observed in aca1Δ gpa1Δ and cac1Δ mutants. Taken together, these findings confirm that Aca1 controls cAMP signaling through Cac1 in a manner distinct from Gpa1.

Aca1 is required for virulence of C. neoformans.

Because capsule and melanin are important virulence factors, we tested whether Aca1 is required for virulence of C. neoformans. To test this hypothesis, we used the murine inhalation model of systemic cryptococcosis. A/Jcr mice were infected with wild-type (H99), aca1Δ (YSB6), aca1Δ+ACA1 reconstituted (YSB117), or cac1Δ (YSB42) mutant strains by intranasal inhalation, which mimics the natural route of human infection by C. neoformans. Inhaled cells first infect the lung and then disseminate to the brain, causing meningoencephalitis. Survival of infected mice was monitored for 40 days. All mice infected with the wild-type strain became moribund between days 19 and 24 postinfection (median survival, 23 days) (Fig. 8). In contrast, all mice infected with the aca1Δ mutant strains survived throughout the course of the experiment without showing any signs of illness (P < 0.0001 compared to the wild-type or reconstituted strains) (Fig. 8). The aca1Δ mutants were as avirulent as cac1Δ mutants and reintroduction of the ACA1 gene into the aca1Δ mutants completely restored virulence (Fig. 8). Animals infected with the aca1Δ+ACA1 reconstituted strain became moribund between days 23 and 25 postinfection (median survival of 23 days; P = 0.2027 compared to wild-type strains). These findings provide evidence that Aca1 regulates virulence of C. neoformans via the Cac1-cAMP pathway.

Aca1 controls virulence via cAMP signaling in C. neoformans.

We identified and characterized here the gene encoding the adenylyl cyclase-associated protein (CAP/SRV2) in the basidiomycetous fungus C. neoformans. Several lines of evidence indicate that Aca1 is a homolog of the CAP/Srv2 protein found in mammalian cells and ascomycetous fungi. First, Aca1 shares structural features conserved with CAP/Srv2 proteins in other organisms. Second, Aca1 interacted with the C-terminal heptad repeat-containing domain (bp 2126 to 2260) of Cac1 in the yeast two-hybrid system. Aca1-Aca1 interactions were observed, as also seen with the S. cerevisiae CAP/Srv2 protein (55, 58). Although multimerization of CAP appears to be essential for its localization but not cAMP signaling in S. cerevisiae, the role of Aca1 multimerization in C. neoformans should be addressed in future studies. Third, as with CAP/Srv2 found in other organisms, Aca1 also modulates cAMP signaling. aca1 mutations caused defects in mating, capsule and melanin production, and virulence of C. neoformans, similar to mutations in other cAMP signaling components, including Gpa1, Cac1, and Pka1. Furthermore, all of the in vitro phenotypes of the aca1Δ mutant were restored to wild-type by exogenous cAMP. Finally, Aca1 controls intracellular cAMP levels in response to glucose readdition. The aca1Δ mutant maintained a basal level of cAMP but failed to induce cAMP production after the readdition of glucose. Similar to the aca1Δ mutant, the C. albicans cap1Δ mutant was found to be defective in induced cAMP levels, but maintained a normal basal level during the bud-hypha transition (4). Therefore, Aca1 encodes a member of CAP/Srv2 protein family regulating cAMP signaling in C. neoformans.

Aca1 was found to be an important virulence regulator for C. neoformans. The complete absence of virulence of aca1Δ mutant strains is likely attributable to lack of capsule and melanin production rather than to defects in mating and filamentation because other mutants showing only defects in mating and filamentous growth (e.g., ste11αΔ, ste7Δ, and cpk1Δ mutants) are virulent (12). Previously, the adenylyl cyclase-associated protein (Cap1) was found to be a virulence regulator for the ascomycetous pathogenic fungus C. albicans (4). Interestingly, avirulence of the C. albicans cap1Δ mutant appears to result from its inability to switch between yeast and hyphal growth (4), an essential virulence factor for C. albicans (30). Therefore, the adenylyl cyclase-associated proteins Aca1 and Cap1 serve as important virulence regulators for C. neoformans and C. albicans via conserved cAMP-signaling pathways that drive unique developmental cascades required for the acquisition of virulence.

Construction of ras1, gpb1, and cAMP cascade mutants by using the congenic serotype A C. neoformans with dominant selectable markers.

Another major contribution of the present study is the recreation of cAMP-independent (gpb1Δ and ras1Δ) and cAMP-dependent (gpa1Δ, cac1Δ, pka1Δ, and pka2Δ) mutants by using dominant selectable markers and the congenic serotype A C. neoformans MATα and MATa strains, H99 and KN99, as parental strains. In previous studies, the functions of Gpb1, Ras1, Gpa1, Cac1, and Pka1/2 were analyzed by using mutants created in uracil or adenine auxotrophic parental strain backgrounds (1-3, 50). Here, however, we suggest that use of auxotrophic strains as parents for mutant construction should be avoided, if possible, because earlier auxotrophic strains generated by UV or gamma irradiation may contain undesired background mutations and ura5 mutants have recently been found to be temperature sensitive (23). We found few, if any, examples of problems based on these concerns, but it seems prudent to advocate advancing the state of the art in the field at this time.

New insights into cAMP-dependent and -independent signaling pathways regulating virulence and differentiation of C. neoformans.

Our studies further expand existing knowledge about signal transduction systems regulating virulence factors (melanin and capsule) and morphological differentiation (mating and filamentation), which are summarized in Fig. 9. In this model, Aca1 functions as an upstream component of the Cac1-cAMP signaling pathway in parallel with Gpa1. First, aca1Δ mutant strains showed subtle phenotypic differences compared to gpa1Δ mutant strains in cell fusion efficiency and filamentation during mating and in melanin synthesis. Second, disruption of both GPA1 and ACA1 conferred additive defects in melanin production, resulting in defects similar to those observed in cac1Δ mutants, providing evidence that Aca1 and Gpa1 are two necessary and sufficient upstream regulators for Cac1. aca1Δ cac1Δ double-mutant strains were comparable to cac1Δ single mutants, showing that Aca1 signals in a linear fashion with Cac1. We speculate that Aca1 and Gpa1 constitute two upstream components of adenylyl cyclase in C. neoformans and serve roles similar to Ras1/2 and Gpa2 in S. cerevisiae.

A key downstream cAMP-signaling element in C. neoformans is PKA. PKA has two catalytic subunits, Pka1 and Pka2, and a single regulatory subunit, Pkr1. Recently, it has been demonstrated that Pka1 plays a major role in regulation of capsule and melanin production in serotype A C. neoformans, whereas Pka2 performs this role in serotype D strain (21). pka1Δ and pka2Δ mutants newly constructed in the present study showed phenotypes consistent with those of the previous study; however, we did uncover a minor role for Pka2 in mating and melanin production through generation of pka1Δ pka2Δ double-mutant strains. pka1Δ pka2Δ double-mutant strains displayed a more severe defect in unilateral mating crosses and melanin biosynthesis than pka1Δ mutants, similar to the cac1Δ mutant or aca1Δ gpa1Δ mutant strains. Thus, Pka1 plays the major and Pka2 a minor role, and the two are necessary and sufficient downstream elements of the Cac1-cAMP signaling pathway for complete mating and melanin production.

We found that Ras1 also has a minor role in melanin production, independent of cAMP signaling. Although ras1Δ single-mutant strains did not show significant defects in melanin production in accord with a previous report (1), aca1Δ ras1Δ and cac1Δ ras1Δ double-mutant strains exhibited a more severe melanin defect than aca1Δ and cac1Δ single mutants, respectively (data not shown). Therefore, Ras1 mediates melanin-inducing signals independent of the cAMP pathway. Previously, it has been shown that overexpression or disruption of STE12 results in induction or reduction of the LAC1 gene, respectively, for melanin production (8, 53) and Ste12 is in the Ras1-mediated signaling pathway (56). Therefore, it is possible that Ras1 could contribute to melanin production through activation of pathways impinging on Ste12.

In conclusion, the present study identifies a new cAMP-signaling component, Aca1, and provides insights into the role of the cAMP pathway in regulating virulence and differentiation of C. neoformans. The isogenic set of MATα and MATa mutants described here represents an excellent platform from which to identify and characterize additional upstream or downstream components of the cAMP-dependent and other signaling pathways. Using the set of strains constructed in the present study, genome-wide transcriptional profiling studies can also be performed to identify the downstream targets of the cAMP pathway.