Calcium- and Calcineurin-Independent Roles for Calmodulin in Cryptococcus neoformans Morphogenesis and High-Temperature Growth

Abstract

The function of calcium as a signaling molecule is conserved in eukaryotes from fungi to humans. Previous studies have identified the calcium-activated phosphatase calcineurin as a critical factor in governing growth of the human pathogenic fungus Cryptococcus neoformans at mammalian body temperature. Here, we employed insertional mutagenesis to identify new genes required for growth at 37°C. One insertion mutant, cam1-ts, that displayed a growth defect at 37°C and hypersensitivity to the calcineurin inhibitor FK506 at 25°C was isolated. Both phenotypes were linked to the dominant marker in genetic crosses, and molecular analysis revealed that the insertion occurred in the 3′ untranslated region of the gene encoding the calcineurin activator calmodulin (CAM1) and impairs growth at 37°C by significantly reducing calmodulin mRNA abundance. The CAM1 gene was demonstrated to be essential using genetic analysis of a CAM1/cam1Δ diploid strain. In the absence of calcineurin function, the cam1-ts mutant displayed a severe morphological defect with impaired bud formation. Expression of a calmodulin-independent calcineurin mutant did not suppress the growth defect of the cam1-ts mutant at 37°C, indicating that calmodulin promotes growth at high temperature via calcineurin-dependent and -independent pathways. In addition, a Ca²⁺-binding-defective allele of CAM1 complemented the 37°C growth defect, FK506 hypersensitivity, and morphogenesis defect of the cam1-ts mutant. Our findings reveal that calmodulin performs Ca²⁺- and calcineurin-independent and -dependent roles in controlling C. neoformans morphogenesis and high-temperature growth.

Calcium is a ubiquitous second messenger that functions in signal transduction pathways in eukaryotic organisms. Ca²⁺ signaling has been implicated in governing myriad biological processes as broad as fertilization and development, exocytosis and muscle contraction, and transcription and chromatin remodeling in multicellular eukaryotes (5). In the fungal kingdom, key features of the Ca²⁺ signaling machinery are conserved with multicellular eukaryotes, and Ca²⁺ signaling underlies diverse fungal physiological processes. In the budding yeast Saccharomyces cerevisiae, Ca²⁺ signaling is an essential component of cell cycle regulation, mating, and stress responses (11). In filamentous fungi, the role of Ca²⁺ extends into the regulation of hyphal morphogenesis, including hyphal tip growth events involving branching and orientation (22). The recent availability of fungal genome sequences has allowed a comparative analysis of fungal Ca²⁺ signaling components, and it is clear that fungi share key regulators of Ca²⁺ signaling (45). These include Ca²⁺-permeable channels, pumps, and transporters; calmodulin; and calmodulin-regulated proteins including calmodulin-dependent kinases and the protein phosphatase calcineurin.

Calmodulin is a small Ca²⁺-binding protein that is conserved from fungi to humans and contributes to the regulation of mitosis, transcription, cytoskeletal rearrangements, and stress responses (8). Calmodulin functions as a critical Ca²⁺ sensor, and yet calmodulin has both Ca²⁺-dependent and Ca²⁺-independent binding partners, highlighting the versatility of calmodulin as a signaling molecule. Calmodulin acts as a Ca²⁺ sensor by binding Ca²⁺ ions via four EF hands, which each contain a Ca²⁺-binding loop in which conserved aspartate and glutamate residues bind the Ca²⁺ ions (4). Ca²⁺ binding induces a conformational change in calmodulin that results in the release of free energy, which is the basis of its ability to act as a Ca²⁺ sensor (28). Studies of genetically tractable organisms have shown that calmodulin is essential and that its essential functions require the ability to bind Ca²⁺ in most organisms (13, 30, 42). Saccharomyces cerevisiae is the sole known exception, where only the Ca²⁺-independent functions of calmodulin are essential for cell viability (21).

One well-defined target of calmodulin is calcineurin, a serine/threonine-specific protein phosphatase whose mechanisms of activation and inhibition are conserved throughout eukaryotic organisms (3, 24). Calcineurin exists as a heterodimer consisting of a catalytic A subunit and a regulatory B subunit. Association of the two subunits is necessary but not sufficient for calcineurin function. Activation of calcineurin occurs when the Ca²⁺/calmodulin complex binds to the calmodulin-binding domain in the C-terminal regulatory region of the A subunit, eliciting conformational changes that free the calcineurin active site from occlusion by an autoinhibitory domain (43). Calcineurin function is inhibited by the immunosuppressive antifungal drugs cyclosporine A and FK506 (tacrolimus) in complex with the peptidyl-prolyl isomerases cyclophilin A and FKBP12, respectively. The drug-protein complexes inhibit calcineurin function by binding to the hydrophobic interface between the A and B subunits and, by steric hindrance, preventing large substrates from docking into the active site.

In the yeast S. cerevisiae calcineurin serves at least three functional roles: regulating stress-activated transcription, Ca²⁺ homeostasis, and morphogenesis (11, 17). The role of calcineurin in stress responses is conserved between model and pathogenic fungi, but the specific functions of calcineurin are unique (32). Calcineurin is critical for virulence in two pathogenic fungi that infect humans, Cryptococcus neoformans and Candida albicans, yet the precise roles of calcineurin in promoting virulence differ between the two. Calcineurin is required for growth at mammalian body temperature of C. neoformans but not of C. albicans, whereas calcineurin is required for C. albicans to survive in serum and disseminate in the host (6, 38). The ability to grow at 37°C is a prerequisite for pathogenesis, and C. neoformans strains that lack either the calcineurin A or B subunit are avirulent in animal models of cryptococcosis (10, 16).

Here, we have analyzed the role of calmodulin as a Ca²⁺ sensor that activates Ca²⁺ signaling in C. neoformans. By using gene disruption approaches and diploid strains, calmodulin was found to be essential, as in other fungi. A unique temperature-sensitive allele of the CAM1 gene encoding calmodulin was isolated via insertional mutagenesis and found to confer temperature-sensitive growth by reducing but not abolishing calmodulin expression. The cam1-ts allele, a calcium-independent calmodulin mutant in which the four Ca²⁺-binding EF hands are mutated (cam1-4DA), and a truncated CNA1 allele encoding a calmodulin-independent calcineurin A allele (CNA1-AIΔ) were employed to probe Ca²⁺, calmodulin, and calcineurin signaling. Our studies reveal that calmodulin plays Ca²⁺-independent and -dependent roles, and also calcineurin-dependent and independent roles, to govern morphogenesis and growth at 37°C in C. neoformans.

MATERIALS AND METHODS

Strains and media.

Strains used are listed in Table 1. Strains were grown in the standard yeast growth media YPD, SD, and YNB with appropriate supplements (2). For growth with FK506, strains were grown on YPD medium or synthetic proline (SP) medium (2% dextrose, 1 mM proline, and YNB without amino acids). Genetic crosses were conducted on V8 medium (19). Where indicated, FK506 was added to growth medium at a concentration of 1 μg/ml.

TABLE 1.

C. neoformans strains

Strain	Genotype	Source
JEC20	MATa	36
JEC21	MATα	36
JEC43	MATα ura5	36
JEC171	MATaade2 lys2	36
MCC16	MATα cna1Δ::ADE2 ade2 ura5	9
MCC10	MATacna1Δ::ADE2 ade2	9
RAS009	MATa/MATα ADE2/ade2 LYS2/lys2 ura5/ura5	41
CSB56	MATacam1-ts	This study
CSB57	MATα cam1-ts	This study
PK7	MATα ura5 cam1-ts	This study
PK14	MATa/MATα ADE2/ade2 LYS2/lys2 ura5/ura5 CAM1/cam1Δ	This study
PK15	PK7 [pPM8]	This study
PK42	MCC16 [pPM8]	This study
PK43	MCC16 [pPM8-CNA1-AIΔ]	This study
PK45	PK7 [pPM8-CNA1-AIΔ]	This study
PK47	PK7 [pPM8-CAM1]	This study
PK50	PK7 [pPM8-cam1-4DA]	This study

Insertional mutagenesis.

The serotype D strain JEC43 (MATα ura5) was transformed by biolistic transformation with plasmid pCH233, which contains the nourseothricin (NAT) dominant marker. The NAT dominant marker consists of the Streptomyces noursei nourseothricin acetyltransferase gene flanked by the C. neoformans serotype A ACT1 promoter and TRP1 terminator from strain H99 (27, 34). The plasmid was transformed as circular DNA or after linearization with XmnI. Transformed cells were allowed to recover for 4 to 6 h on YPD medium containing 1 M sorbitol and were then transferred to YPD medium containing 100 μg/ml nourseothricin. Transformants were purified on nourseothricin plates and then tested for stability of nourseothricin resistance. Transformants were grown in liquid YPD medium lacking nourseothricin for 4 to 5 days, transferred to YPD solid medium using a 48-prong transfer device, replica plated onto YPD medium plus nourseothricin, and incubated for 24 h. Transformants containing a stably integrated NAT marker exhibited robust growth in this assay.

Genetic analyses.

Linkage of the NAT dominant marker to the temperature-sensitive phenotype was determined by crossing insertion mutants to strain JEC171 (MATa ade2 lys2). Strains were coincubated on V8 medium for 7 to 10 days in the dark. For random spore analysis, meiotic progeny were isolated by plating a portion of the mating mixture on SD medium containing lysine and lacking both adenine and uridine to select for recombinant progeny. Progeny were then scored for the presence of the lys2 mutation, the NAT dominant marker, mating type, and the growth defect at 37°C. For dissections, individual basidiospores were isolated using a micromanipulator and allowed to germinate on YPD medium. Germinated spores were then scored for auxotrophy, nourseothricin resistance, and growth at 37°C.

To determine if the CAM1 gene is essential, a cam1Δ allele was constructed using the neomycin (NEO) dominant marker as previously described (12, 19). The diploid strain RAS009 (41) was transformed with the cam1Δ::NEO allele, and G418-resistant transformants were isolated. Heterozygous CAM1/cam1Δ::NEO transformants with one allele of the CAM1 gene disrupted were identified by Southern analysis and diagnostic PCR. Transformants were incubated on V8 medium for 7 days to obtain haploid basidiospores. Meiotic progeny were isolated either by plating on YPD medium at 25°C and choosing adenine auxotrophs (based on red colony color) that were no longer self-filamentous and therefore a or α but not a/α or by micromanipulation as described above. All progeny were scored for resistance to G418.

Identification of the marker insertion site.

The insertion site of the NAT dominant marker was determined using a modified version of the vectorette PCR approach (40). Genomic DNA from insertion mutant 3E7 was digested with EcoRV, and anchor-bubble linkers were ligated onto the blunt ends. PCR primers specific for the NAT dominant marker and the complementary strand of the anchor-bubble were used to amplify a DNA fragment containing the genomic DNA that flanks the marker insertion site. Conditions used were as follows: 94°C for 20 s, 60°C for 30 s, and 72°C for 3 min. The anchor-bubble primer is designed such that it can prime synthesis only on newly synthesized DNA primed by the NAT dominant marker primer, conferring specificity upon the PCR. PCR products were cloned and sequenced, and the C. neoformans genomic DNA portion of the PCR product was used for BLAST analysis of the C. neoformans JEC21 genome and the nonredundant NCBI database.

Northern analysis, RT-PCR, RACE, and real-time PCR analysis.

Northern analyses were performed as previously described (31). Probes for the CAM1 and ACT1 genes were generated by PCR and labeled with [α-³²P]dCTP using the Redi-Prime kit (Amersham Biosciences). Reverse transcription (RT)-PCR and rapid amplification of cDNA ends (RACE) analysis were performed using the SMART Race cDNA kit (Clontech). First-strand synthesis was primed with oligo(dT), and amplification of calmodulin cDNA was accomplished using primers that can prime synthesis only if introns have been spliced from the template to allow discrimination of the desired PCR products from those that resulted from spurious amplification of contaminating genomic DNA (JOHE 9363 plus JOHE 9364 [see Table S1 in the supplemental material]). cDNA from wild-type and cam1-ts strains (JEC21 and PK7, respectively) was used as a template for quantitative real-time PCR using iQ SYBR Green Supermix (Bio-Rad) and a Bio-Rad iCycler iQ Multicolor real-time detection system. The iCycler iQ Multicolor real-time detection system was used as the fluorescence detector under the following PCR conditions: an initial denaturing cycle of 95°C for 3 min, 40 cycles of denaturation at 95°C for 20 s, and annealing/extension at 50°C for 45 s. A standard melt curve from 50°C to 90°C with fluorescent monitoring each 0.5°C was included. Reactions were performed in duplicate.

Construction of CNA1-AIΔ and cam1-4DA alleles.

The CNA1-AIΔ allele was constructed using an overlap PCR approach. The allele contains ∼1 kb of 5′ and 3′ flanking DNA (984 and 991 bp, respectively) from the 5′ and 3′ untranslated regions (UTRs) of the CNA1 gene and the genomic copy of the CNA1 gene with a 678-bp deletion resulting in the removal of codons 486 to 639, which contain the calmodulin-binding site and the autoinhibitory domains of calcineurin A (38). The CNA1-AIΔ allele was confirmed by sequencing and then subcloned into plasmid pPM8 (35). The cam1-4DA allele was constructed in a similar fashion using overlap PCR. The allele contains ∼1 kb of 5′ and 3′ flanking DNA (744 and 873 bp, respectively) from the 5′ and 3′ untranslated regions of the CAM1 gene and the genomic version of the CAM1 gene with four substitutions. Codons 21, 57, 94, and 130 were mutated to encode alanine instead of aspartate. These are highly conserved residues in the Ca²⁺-binding loop of the calmodulin EF hand domains, and aspartate-to-alanine substitutions in the three EF hands of S. cerevisiae calmodulin abolish Ca²⁺ binding (21) (a detailed description of the construction of these alleles is found in Table S1 and Fig. S1, which is included as supporting information, in the supplemental material).

Expression of Cam1 and Cam1-4DA.

To examine the Ca²⁺-binding activity of the calmodulin quadruple mutant, cDNA versions of the wild-type calmodulin gene and the cam1-4DA mutant were subcloned into the Escherichia coli expression plasmid pRSET-A (Invitrogen, Carlsbad, CA) (a detailed description of the construction of these alleles is found in Table S1 and Fig. S1, which is included as supporting information, in the supplemental material). Expression of the fusion proteins was induced by addition of 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside) to rapidly growing E. coli cultures harboring the expression plasmids. Total proteins were separated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride filters by electroblotting. ⁴⁵Ca²⁺ overlay assays were performed as previously described, with modifications (33). Briefly, polyvinylidene difluoride membranes were washed five times for 10 min in wash buffer (10 mM imidazole, pH 6.8, 70 mM KCl, 1 mM MgCl₂) and incubated with 2 μM ⁴⁵CaCl₂ (18.53 mCi/mg; Perkin-Elmer) in wash buffer for 1 h at room temperature. Membranes were then washed twice for 2 min in distilled water and once for 2 min in 50% ethanol, dried, and exposed to autoradiography film for 48 h.

Microscopy.

Differential interference microscopy and fluorescent images were captured with a Zeiss Axioskop 2 Plus fluorescent microscope equipped with an AxioCam MRM digital camera. Cells were grown in SP medium in the presence or absence of 1 μg/ml FK506 for 24 h and fixed for 30 min in phosphate-buffered saline (PBS) containing 10% formaldehyde. To visualize actin and DNA, cells were permeabilized with 1% Triton X-100 in PBS for 10 min, washed three times in PBS, and stained with 8 μg/ml 4,6-diamidino-2-phenylindole (DAPI) (lot no. D-21490; Molecular Probes) to visualize DNA or 10 μg/ml rhodamine-conjugated phalloidin (lot no. P-195; Sigma-Aldrich) to visualize actin.

RESULTS

An insertional mutagenesis screen for high-temperature growth-defective mutants.

To identify other components of the calcineurin signaling cascade that governs growth at 37°C, we implemented an insertional mutagenesis approach. Approximately 3,000 nourseothricin-resistant transformants were obtained after biolistic transformation of either circular or linear DNA of plasmid pCH233 into the C. neoformans serotype D strain JEC43 (MATα ura5). Transformants were tested for stability of the NAT dominant marker insertion, and 430 transformants exhibited stable nourseothricin resistance after passage on nonselective medium (14.3% stable). Stable transformants were tested for a growth defect at 37°C, and 12 isolates were unable to grow or grew poorly at 37°C compared to the parental strain JEC43.

All of the temperature-sensitive isolates were crossed with strain JEC171 (MATa ade2 lys2) to determine if the 37°C growth defect was linked to the NAT insertion. Crosses were analyzed by both using random spore analysis and isolating individual basidiospores by micromanipulation and determining the phenotypes of meiotic progeny produced by germination. For 1 of the 12 isolates, 3E7, all meiotic segregants that displayed a growth defect at 37°C were also nourseothricin resistant, and nourseothricin-sensitive segregants grew normally at 37°C, indicating that the NAT dominant marker and the growth defect are genetically linked. In addition to the growth defect at 37°C, isolate 3E7 displayed hypersensitivity to the calcineurin inhibitor FK506 at the permissive growth temperature of 25°C (Fig. 1A). This phenotype was the result of calcineurin inhibition by the FK506-FKBP12 drug-protein complex, as an frr1-3 cam1-ts mutant isolated following a genetic cross, in which FKBP12 contains an active-site point mutation, was resistant to FK506 (data not shown). Thus, calcineurin becomes essential for growth at 25°C in the 3E7 mutant, whereas calcineurin is required for growth only at 37°C in wild-type backgrounds.

FIG. 1.

An insertion in the calmodulin gene CAM1 confers temperature-sensitive growth. (A). Southern analysis of genomic DNA from insertion mutant 3E7 and parental strain JEC43. Genomic DNA was digested with EcoRV or HindIII. The probe used is indicated by the black bar. Abbreviations: WT, wild type; E, EcoRV; H, HindIII. (B). Schematic diagram of the CAM1 locus in the 3E7 mutant strain. The inserted NAT dominant marker lies 49 bp downstream of the CAM1 stop codon in the 3′ untranslated region. (C). Fivefold serial dilutions of cultures of the insertion mutant 3E7 and the parental strain JEC43 were inoculated on YPD medium and grown at 25°C or 37°C for 3 days or on YPD medium containing 1 μg/ml FK506 and incubated at 25°C for 3 days.

Insertion in CAM1 reduces CAM1 mRNA abundance.

Recovery of the genomic DNA adjacent to the NAT dominant marker insertion in isolate 3E7 using a modified version of the vectorette PCR technique revealed an insertion in the CAM1 gene encoding calmodulin (40). Sequence analysis of the genomic DNA adjacent to the NAT dominant marker insertion in isolate 3E7 revealed that the insertion occurred in the 3′ UTR of the CAM1 gene. The insertion occurred 49 bp downstream from the stop codon, and its location was confirmed by Southern analysis of genomic DNA and RACE (Fig. 1A and data not shown). Introduction of the wild-type CAM1 gene on an episomal plasmid restored growth at 37°C (see Fig. 4C). Based on the insertion location, the phenotype of the 3E7 isolate, and complementation by the wild-type CAM1 gene, the mutant allele was designated cam1-ts.

FIG. 4.

Calcineurin and calmodulin contribute a shared function in promoting high-temperature growth. (A) Schematic diagram of the calcineurin A subunit and the truncated, constitutively active calmodulin-independent Cna1-AIΔ mutant. (B) ⁴⁵Ca²⁺ overlay blot of E. coli-expressed Cam1 and Cam1-4DA, which contains mutations in the Ca²⁺-binding loops of the four calmodulin EF hand domains. Western blot analysis using antibodies directed against the Xpress epitope (Invitrogen) contained in the expression plasmid confirmed that both the Cam1 and the Cam1-4DA proteins were expressed. (C) Fivefold serial dilutions of strains with the indicated genotype were spotted onto YNB medium and incubated at the indicated temperature for 3 days. Two independent transformants for each strain were tested. (D) Fivefold serial dilutions of the indicated strains were inoculated onto SP medium or SP medium containing 1 μg/ml FK506 and incubated for 3 days at 25°C.

The NAT dominant marker insertion does not disrupt the CAM1 coding region, yet the cam1-ts allele confers strong growth defects both at 37°C and in the absence of calcineurin function. We hypothesized that the insertion in the CAM1 3′ UTR might disrupt mRNA stability and reduce CAM1 message abundance and thereby confer a temperature-sensitive growth defect. Therefore, calmodulin mRNA abundance was assessed by Northern analysis, RT-PCR, and real-time PCR. In the cam1-ts mutant, the mRNA level was significantly reduced compared to that of the wild-type strain (Fig. 2A). Furthermore, when Ca²⁺ or Na⁺ ions were added to the culture medium, the levels of calmodulin mRNA increased in the wild-type strain, whereas any message in the cam1-ts mutant was still below the limit of detection. RT-PCR was then used to determine if any calmodulin mRNA is present in the cam1-ts cells. Calmodulin cDNA was not detected after 20 PCR cycles using untreated or Ca²⁺-treated cam1-ts cells; however, it was detected after 25 cycles (data not shown). Real-time PCR analysis of wild-type and cam1-ts strains revealed an approximately 60-fold reduction in the calmodulin transcript in the cam1-ts strain compared to that of the wild type (less than 2% of the wild type), whereas the abundance of the GPD1 control transcript was the same in wild-type and cam1-ts mutant cells. (Fig. 2B).

FIG. 2.

Reduction in CAM1 mRNA abundance in the cam1-ts mutant. (A) Northern analysis of total RNA from the wild-type and cam1-ts mutant strains. Cells were untreated or treated with either 50 mM CaCl₂ or 0.4 M NaCl for 3 hours prior to RNA extraction. (B) Real-time PCR analysis of wild-type (JEC21 [solid lines]) and cam1-ts (PK7 [dashed lines]) strains. ΔC_t (where C_t denotes threshold cycle) values shown represent the difference in average cycle threshold values in two independent experiments determined at a value of 25 relative fluorescence units (RFU). One ΔC_t corresponds to an approximately twofold difference in template abundance. The GPD1 gene was included as a standard.

Calmodulin transcripts in the cam1-ts mutant were also apparent during RACE analysis. Based on RACE analysis, the 3′ untranslated region of the cam1-ts mRNA is 162 nucleotides (nt) long [49 nt from the CAM1 gene and 113 nt from the NAT plasmid, excluding the poly(A) tail], while the 3′ untranslated region of CAM1 mRNA is 147 nt. Because the size of the message is similar, we attribute the decrease in the calmodulin cDNA level of cam1-ts cells compared to that of the wild type to a decrease in mRNA abundance rather than a decrease in first-strand synthesis efficiency due to a large mRNA size difference. This decrease in mRNA abundance may be due to decreased message stability caused by the insertion or a disruption of 3′ regulation of expression.

CAM1 is essential.

Calmodulin is encoded by essential genes in Drosophila melanogaster, Aspergillus nidulans, Saccharomyces cerevisiae, and Schizosaccharomyces pombe (13, 23, 30, 42). Given that the cam1-ts mutant cells express less than 2% of the wild-type level of calmodulin mRNA, we sought to test if CAM1 is an essential gene and if this reduced level is sufficient to maintain cell viability or alternatively whether the CAM1 gene is dispensable for viability in C. neoformans at 25°C but required at 37°C. To distinguish between these models, a cam1Δ::URA5 deletion allele was constructed using PCR overlap. A total of 125 Ura⁺ transformants were isolated at 25°C following biolistic transformation of the ura5 mutant strain JEC43 with the cam1Δ::URA5 allele. PCR screening and Southern analysis of genomic DNA revealed that none of the 125 transformants contained a disrupted CAM1 gene. These findings provided presumptive evidence that the CAM1 gene is essential.

To definitively establish whether or not the CAM1 gene is essential, a heterozygous cam1Δ/CAM1 diploid strain was constructed and analyzed. The diploid strain RAS009 that is heterozygous for the ade2 marker (41) was transformed with the cam1Δ::NEO allele produced by overlap PCR, and G418-resistant colonies were isolated. A total of 52 G418-resistant transformants were recovered, and the presence of the cam1Δ::NEO allele in the correct genomic location was determined by PCR using an external genomic DNA primer and a NEO-specific primer. Eight transformants produced the PCR product expected if at least one copy of CAM1 was replaced with the cam1Δ::NEO allele, and a single insertion of the cam1Δ::NEO allele was confirmed by Southern analysis.

The eight CAM1/cam1Δ diploid isolates were incubated on V8 mating medium at 25°C to induce sporulation, and haploid strains were isolated based on the appearance of the red colony color indicative of adenine auxotrophic strains. Sixty-four haploid strains were isolated from each of the eight sporulated diploid strains and tested for G418 resistance. Forty-five of the 512 strains isolated were G418 resistant and therefore represented putative cam1 deletion mutants. However, the CAM1 open reading frame was found to still be present in all 45 of these strains based on PCR analysis, indicating that none are haploid cam1 deletion mutants and instead are all aneuploid for at least the CAM1/cam1Δ locus. The G418-resistant segregants from one transformant (transformant 25) were analyzed by PCR using primers that amplify the CAM1 open reading frame, the cam1Δ::NEO allele, and the a and α alleles of the MAT locus. All six ade⁻ G418-resistant segregants from this transformant contained both the CAM1 wild type and the cam1Δ::NEO allele, but each segregant inherited only one of the two MAT alleles (Fig. 3). The segregation of markers in transformant 25 was also analyzed by dissection of basidiospores. Twenty-six spores were germinated following micromanipulation and transfer to YPD medium; however, none of these meiotic segregants were resistant to G418. In contrast, the adenine auxotrophy segregated in a 1:1 ratio (14 Ade⁺ and 12 ade⁻). Taken together, these findings indicate that the CAM1 gene is essential in C. neoformans.

FIG. 3.

Calmodulin is essential in C. neoformans. PCR analysis of segregants from the heterozygous CAM1/cam1Δ diploid strain was performed. Primers specific for the CAM1 open reading frame, the cam1Δ allele, and the a and α mating-type-specific genes STE20a and STE20α were used in PCRs using genomic DNA from the indicated strains as the template.

CAM1 promotes high-temperature growth in a calcineurin-dependent and -independent manner.

Calcineurin is essential for C. neoformans high-temperature growth (16, 38). The cam1-ts mutant grows very slowly at 37°C, and this defect may result from the inability to activate calcineurin, the loss of a different function of calmodulin, or both. To determine this, we constructed a calcineurin A truncated allele (CNA1-AIΔ) that lacks the coding region for the C-terminal calmodulin-binding site and the autoinhibitory domain. Calcineurin lacking these domains is catalytically active and constitutive as the requirement for Ca²⁺/calmodulin binding for activation has been relieved (26, 44). The CNA1-AIΔ allele was subcloned into the multicopy plasmid pPM8 and transformed into cna1 and cam1-ts mutants. Growth of the cna1 mutant strain harboring the CNA1-AIΔ allele was restored at 37°C, while the cna1 mutant harboring only the control plasmid failed to grow at 37°C (Fig. 4C). In contrast, the cam1-ts mutant strain was not capable of growing at 37°C when transformed with the calmodulin-independent CNA1-AIΔ allele (Fig. 4C), demonstrating that calmodulin has at least a second function in promoting high-temperature growth in addition to its well-established role in activating calcineurin.

CAM1 promotes high-temperature growth independently of calcium ions.

Calmodulin has both Ca²⁺-dependent and Ca²⁺-independent functions. To test if the calcineurin-independent function of calmodulin in promoting high-temperature growth depends on Ca²⁺ binding, a calmodulin mutant allele (cam1-4DA) that contains a substitution in each of the four Ca²⁺ binding loops of the EF hand domains was constructed. Aspartic-acid-to-alanine substitutions in the EF hands of S. cerevisiae calmodulin completely abolish Ca²⁺ binding ability (21), and the corresponding mutations were engineered in codons 21, 57, 94, and 130 of C. neoformans calmodulin.

cDNA versions of the CAM1 gene and the cam1-4DA allele were subcloned into the E. coli expression plasmid pRSET-A, and the Ca²⁺ binding ability of the expressed proteins was assessed by a ⁴⁵Ca²⁺ overlay assay. The wild-type Cam1 protein bound ⁴⁵Ca²⁺, while no ⁴⁵Ca²⁺ binding to the Cam1-4DA mutant protein was detected, indicating that the engineered mutations in the EF hands of the Cam1-4DA protein abolish Ca²⁺-binding activity (Fig. 4B). The cam1-4DA allele and wild-type CAM1 were then subcloned into plasmid pPM8 and transformed into the cam1-ts strain. Growth of the cam1-ts strain at 37°C was restored by either the cam1-4DA allele or wild-type CAM1 (Fig. 4C). These observations indicate that Ca²⁺ binding is not required for the calcineurin-independent function of calmodulin in promoting high-temperature growth.

The ability of the cam1-4DA allele to complement the cam1-ts growth defect when calcineurin function is impaired was assessed by a growth assay on medium containing FK506. Wild-type CAM1 complemented the FK506 hypersensitivity of the cam1-ts mutant at 25°C, while the cam1-4DA allele only partially complemented this defect (Fig. 4D). These findings indicate that calmodulin promotes C. neoformans viability and high-temperature growth in a calcineurin-dependent and -independent manner and that both Ca²⁺-bound and Ca²⁺-free calmodulin share critical functions with calcineurin in C. neoformans growth control.

CAM1 promotes mating filament formation.

In addition to its role in growth at 37°C, calcineurin is essential for filamentous growth during C. neoformans mating (9). To examine if calmodulin plays a similar role, the ability of the cam1-ts mutant to produce mating filaments was assessed by a mating assay. Mating filament production was reduced but not abolished in a bilateral cross in which both parents harbored the cam1-ts mutation compared to a cross between wild-type strains (Fig. 5A and D). Mating filament production was restored to the wild-type level in crosses when one strain contained either the wild-type CAM1 gene or the cam1-4DA allele (Fig. 5E and F). In contrast, crosses in which both parents contained calcineurin mutations failed to produce mating filaments, as expected (Fig. 5G). The mating defect of calcineurin mutant cells was complemented by the constitutive, calmodulin-independent CNA1-AIΔ allele (Fig. 5I). However, the CNA1-AIΔ allele failed to complement the mating defect of the cam1-ts mutant cells (Fig. 5C). These observations provide evidence that calmodulin plays a Ca²⁺/calcineurin-independent role in mating in addition to its known role in activating calcineurin in a calcium-dependent fashion.

FIG. 5.

Calmodulin promotes mating filament formation. Strains with the indicated genotype were cocultured on V8 mating medium and photographed after incubation for 5 days at 25°C. Strain PK7 (MATα cam1-ts ura5) was used as the parental strain for the cam1-ts mutants harboring pPM8-derived plasmids which contain the wild-type URA5 gene. Crosses performed were as follows: (A) CSB57 (MATα cam1-ts) × CSB56 (MATa cam1-ts); (B) PK15 (PK7 [pPM8]) × CSB56; (C) PK45 (PK7 [pPM8-CNA1-AIΔ]) × CSB56; (D) JEC20 (MATa) × JEC21 (MATα); (E) PK47 (PK7 [pPM8-CAM1]) × CSB56; (F) PK50 (PK7 [pPM8-cam1-4DA]) × CSB56; (G) MCC16 (MATα cna1Δ::ADE2 ade2 ura5) × MCC10 (MATa cna1Δ::ADE2 ade2); (H) PK42 (MCC16 [pPM8]) × MCC10; (I) PK43 (MCC16 [pPM8-CNA1-AIΔ]) × MCC10. Scale bar, 100 μM.

CAM1 and calcineurin regulate cellular morphology.

Calcineurin is a known downstream target of Ca²⁺-bound calmodulin, and in previous studies, we established that calcineurin A indeed binds calmodulin in C. neoformans (38). In the cam1-ts mutant, calcineurin becomes essential at 25°C as cam1-ts mutant cells were hypersensitive to the calcineurin inhibitor FK506 (Fig. 1C) and cam1-ts cna1Δ double mutants could not be recovered following a genetic cross (data not shown). Because the phenotype of the cam1-ts cna1Δ double mutant is more severe than the cna1Δ single mutant, this synthetic lethal genetic interaction indicates that calmodulin and calcineurin share a function that is independent of the activation of calcineurin by calmodulin. To gain insight into this shared function, the cam1-ts mutant was examined by microscopy after treatment with FK506 to inactivate calcineurin. After 24 h of FK506 exposure at 25°C, cam1-ts mutant cells uniformly exhibited an aberrant bud morphology, forming elongated cells, whereas wild-type cells exposed to FK506 exhibited a normal budding yeast morphology (Fig. 6). Nuclear staining revealed the presence of multiple nuclei, indicating that FK506-treated cam1-ts cells are capable of continued nuclear division in the absence of cell division. Actin filaments were concentrated at the tip of the hyphal structures in FK506-treated cam1-ts mutant cells, indicating that these cells are still capable of directing polarized growth via the actin cytoskeleton. Introduction of wild-type CAM1 or the cam1-4DA allele restored the ability of cam1-ts cells to produce a normal bud. These results indicate that calcineurin and Ca²⁺-free calmodulin contribute a shared function in the regulation of cellular morphology.

FIG. 6.

Calmodulin and calcineurin control cellular morphology. The indicated strains were grown in SP medium, diluted, and shifted to SP medium plus 1 μg/ml FK506 for 24 h. Cells were fixed and stained with DAPI and rhodamine-conjugated phalloidin to visualize nuclei and actin, respectively. Cells were photographed at ×1,000 magnification. Scale bar, 20 μM. DIC, differential interference microscopy.

DISCUSSION

The goal of this study was to determine the role of Ca²⁺ signaling in promoting C. neoformans high-temperature growth, a crucial virulence attribute. Here, we isolated a mutant containing an insertion in the 3′ UTR of the gene encoding calmodulin, CAM1, and utilized the mutant strain to probe calmodulin and calcineurin functions in high-temperature growth and morphogenesis. Similar to calmodulin in other genetically tractable organisms, calmodulin was found to be essential for viability in C. neoformans. In addition to activating calcineurin, we found that calmodulin plays calcineurin-independent roles in promoting C. neoformans high-temperature growth and morphogenesis during mating and bud formation. These functions are also independent of Ca²⁺ binding. Taken together, our findings provide evidence for a bifurcated pathway whereby high-temperature growth and cellular morphogenesis are controlled by Ca²⁺-free calmodulin in one branch and Ca²⁺/calmodulin-activated calcineurin in another branch (Fig. 7).

FIG. 7.

Control of differentiation and 37°C growth by a bifurcated Ca²⁺ signaling pathway. Both the Ca²⁺-bound and the Ca²⁺-free forms of calmodulin govern cellular morphology and 37°C growth in C. neoformans. Ca²⁺-bound calmodulin functions via the activation of calcineurin, and Ca²⁺-free calmodulin functions via targets that remain to be determined, which may include Myo2. Abbreviation: CaM, calmodulin.

The cam1-ts allele was generated using insertional mutagenesis, capitalizing upon the observation that transforming DNA often integrates into the genome via nonhomologous recombination in C. neoformans (12, 37). The disadvantages of this procedure are the relatively low frequency of integration (10 to 15% of transformants) compared to the maintenance of transforming DNA as an extrachromosomal episome. In addition, multiple integrations are possible, complicating the ability to demonstrate linkage between the inserted marker and associated phenotypes by a genetic cross. The finding that a low percentage of mutants isolated displayed linkage of the phenotype and the inserted marker was unexpected and may be a result of the low number of transformants analyzed, the possible mutagenic effects of the biolistic transformation procedure, or epigenetic events that limit growth at elevated temperatures. Importantly, a major advantage of using random insertional mutagenesis as a tool for genetic analysis is the ability to recover hypomorphic alleles in addition to null alleles, extending the analysis to include essential as well as nonessential genes. We note that some of the disadvantages encountered in this study have been addressed by the development of Agrobacterium-mediated transformation as a transkingdom DNA delivery vehicle for mutagenesis of C. neoformans (27).

The discovery of the hypomorphic cam1-ts allele of the essential CAM1 gene in C. neoformans provided a unique opportunity to probe the mechanics of Ca²⁺ signaling in this ubiquitous human fungal pathogen. The insertion of the marker into the 3′ UTR of the CAM1 gene significantly reduces calmodulin mRNA abundance; however, enough calmodulin gene expression remains to sustain cell viability at 25°C but not 37°C. We initially suspected that the high-temperature growth defect of the cam1-ts mutant might be attributable to an increased requirement for calmodulin to activate calcineurin, which itself is required for 37°C growth. However, the Ca²⁺-binding-defective calmodulin mutant (Cam1-4DA) was sufficient to restore growth of the cam1-ts mutant at 37°C, illustrating a role independent of Ca²⁺ binding required for promoting 37°C growth beyond the known Ca²⁺-dependent role of calmodulin in activating calcineurin. The calmodulin function that remains in the cam1-ts mutant is sufficient to activate calcineurin, as evidenced by the strong growth defect at 25°C when calcineurin is inhibited by FK506. These findings further indicate that there is a synthetic lethal interaction between calcineurin inhibition (via FK506) and compromised calmodulin function (via the cam1-ts mutation), a conclusion that was reinforced by our inability to recover cam1-ts cna1 double mutants following genetic crosses (data not shown).

The results reported here indicate that calmodulin and calcineurin share an essential function that governs cellular morphogenesis. S. cerevisiae calmodulin has diverse essential functions as revealed by a study of intragenic complementation of conditional mutants, and the defects observed in three of the complementation groups (cmd1A, cmd1B, and cmd1D) indicate that calmodulin may function at multiple steps of the pathway controlling morphogenesis in yeast (39). The regulation of actin organization is a common theme revealed by these previous studies in yeast; however, recent reports have also identified a role for calcium signaling in microtubule dynamics in other fungi (1, 15).

Here, we found that the execution of the shared function of calcineurin and calmodulin occurs via the Ca²⁺-free form of calmodulin. Calmodulin has a panoply of Ca²⁺-dependent and -independent binding partners, which are best understood in S. cerevisiae (11). Microscopic evaluation of cam1-ts cells in which calcineurin function has been inhibited suggests that a shared essential function of Ca²⁺-free calmodulin and calcineurin is involved in cellular morphogenesis. Attractive candidates for the binding partner(s) of calmodulin that contribute to this essential cellular function include Myo2, which is required for polarized growth in S. cerevisiae and binds the Ca²⁺-free form of calmodulin (7). In our studies, the Ca²⁺-binding-defective allele of calmodulin complemented the morphological defect of FK506-treated cam1-ts mutant cells. Interestingly, the C. neoformans MYO2 gene lies in the MAT locus, is essential, and is a member of the most ancient class of MAT genes, which is indicative of a potential functional role in C. neoformans mating (18). Furthermore, calmodulin controls the actin cytoskeleton by regulating phosphatidylinositol (4, 5) bisphosphate synthesis via Mss4, although it has not yet been determined if this interaction is Ca²⁺ dependent or independent (14). An important consideration is that the phenotype of FK506-treated cam1-ts cells appears to be a hyperpolarization of growth rather than a loss of the ability to undergo polarized growth, as has been observed in S. cerevisiae calmodulin and myosin mutants (25, 29). Genetic interactions between calcineurin and type 2 myosin have been observed in S. pombe, suggesting that calcineurin functions in actin contractile ring formation and septum formation (20). The observation that calcineurin is required for the production of C. neoformans mating filaments implicates calcineurin in the regulation of polarized growth; however, a direct role for calcineurin in promoting actin polymerization or other processes important for cellular morphogenesis remains to be established in C. neoformans.