Genome-Wide Transcriptional Profiling of the Cyclic AMP-Dependent Signaling Pathway during Morphogenic Transitions of Candida albicans

C. albicans strains and media.The C. albicans strains used in this study are described in Table 1. For yeast and germ tube cultures, YNB medium (yeast nitrogen base minimal medium containing 50 mM glucose and 0.0002% biotin) and M199 (medium 199 containing 150 mM HEPES [pH 7.0]; Gibco-BRL) were used, respectively.

Growth conditions, total RNA isolation, and preparation of labeled cDNA.Yeasts for germ tube induction were grown to mid-logarithmic phase in YNB at 25°C to an optical density at 600 nm of 0.8 to 1.0. For germ tube induction, yeasts were briefly sonicated, washed, resuspended in phosphate-buffered saline (PBS), and inoculated at a concentration of 5 × 10⁶ cells/ml into prewarmed M199 and incubated at 37°C for 1.5 h. This time point was chosen to assess gene expression profiles that followed the transient peak of cAMP at 1 h as determined previously (6). At 1.5 h of incubation, the morphologies of the two mutants and the WT were markedly different, in that the pde2/pde2 mutant formed germ tubes that were longer than those of the WT strain, whereas the cap1/cap1 mutant exhibited budding yeast growth (Fig. 1). Indirect immunofluorescence using anti-Hwp1 antibodies showed abundant expression of Hwp1 on true germ tubes formed by the WT strain and the pde2/pde2 mutant (5). Parallel cultures of the pde2/pde2 mutant and WT strain as well as the cap1/cap1 mutant incubated at 25°C grew as yeasts.

• Open in new tab
• Download powerpoint

FIG. 1.

Morphologies of the WT, hyperfilamentous pde2/pde2 mutant, and budding cap1/cap1 mutant used to prepare RNA in profiling experiments. Cells were grown at 25°C in YNB to mid-logarithmic phase, reinoculated at a concentration of 5 × 10⁶ cells/ml into M199, and further incubated for 5 h at 25°C for yeast growth or at 37°C for germ tube induction. At each time point, cells were sampled and their morphologies were observed by microscopy. Total RNAs were isolated from a 1.5-h culture of each strain in M199 at 25 or 37°C (black-outlined boxes). True hyphal formation of the WT and pde2/pde2 mutant was confirmed by an indirect immunofluorescence assay with hypha wall protein Hwp1-specific antibodies as described before (5). The reference RNA was pooled from RNA prepared from each strain at 25°C (denoted as R). The proportion of cells with germ tubes after 1 h of incubation in M199 at 37°C was nearly 100% for the pde2/pde2 (Δpde2; strain BPS15) mutant, 20 to 30% for the WT strain (UnoPP-1), and 0% for the cap1/cap1 (Δcap1; CAC1-1A1E1) mutant.

To identify cAMP-dependent genes induced during morphogenesis, experimental RNA was prepared from each strain after incubation in germ-tube-inducing conditions for 1.5 h at 37°C. A single reference RNA preparation consisted of equal amounts of pooled RNA from the three cultures incubated in M199 for 1.5 h at 25°C. RNA was isolated using the hot-acid-phenol method (3), treated with DNase I (Promega) for 1 h at 37°C, and stored at −80°C.

For budding yeast growth conditions, cAMP-dependent gene expression profiles were generated using experimental RNA prepared from mid-logarithmic-phase yeast cultures of each strain. The single reference RNA preparation consisted of equal amounts of pooled RNA from the same three cultures.

For both the identification of cAMP-regulated genes during morphogenesis and unperturbed budding growth, cDNA from each of the three experimental RNAs (labeled with Cy5) and the single reference RNA (labeled with Cy3) was prepared. Experiment 1 consisted of three arrays, each incubated with a mixture of one of the experimental cDNAs and the reference cDNA. Experiment 2 was a technical replicate of experiment 1. A biological replicate was generated by growing new cultures from each strain to prepare cDNA for labeling and hybridizations in experiments 3 and 4.

C. albicans DNA microarray. C. albicans microarrays were spotted in the laboratory of Judith Berman (University of Minnesota; http://www.cbs.umn.edu//labs/berman/ ) from PCR products generated in the Berman laboratory and the laboratory of Lois Hoyer (University of Illinois at Urbana—Champagne). These C. albicans microarrays were constructed using PCR products consisting of 6,175 unique open reading frames (ORFs) (>100 amino acids). Each PCR product was spotted at least twice. Pre- and postprocessing of microarray slides, preparation of Cy3- or Cy5-labeled cDNA probes, and hybridization were performed as described before (18, 21). Array slides were scanned using an Affymetrix 428 scanner (Santa Clara, CA).

Data filtering and analysis.Image analysis of the scanned TIF file was performed using Scanalyze (version 2.51; Eisen Lab and Stanford University) and Gene Traffic (version 3.2-12; Iobion). Flagging of invalid spots occurred manually in Scanalyze and automatically in Gene Traffic. Names for the hybridization groups used in the analysis were chosen to be cap1 for the cap1/cap1 mutant, WT for the UnoPP-1 strain, and pde2 for the pde2/pde2 mutant. Differences in transcript levels between strains were termed fold changes. Data were normalized using the global intensity method and were filtered to exclude genes that did not have at least two valid spots in each hybridization group (cap1, WT, and pde2). The data from Gene Traffic were imported into Excel 2000 (Microsoft).

The spot average log₂ Cy5/Cy3 ratio was calculated for each gene. For the experiments in which germ tube growth conditions were used, the spot average log₂ Cy5/Cy3 ratio approximated the increase of a transcript in germ-tube-inducing conditions relative to that of yeast-inducing conditions and was termed fold induction. Genes that appear in Tables 2 and 3, as well as Tables S3 and S4 in the supplemental material, all were tested for significance using significance analysis of microarrays (SAM; version 2.23b; Stanford University) and were required to have SAM-generated q values of less than 0.5. A result of q < 0.5 indicated a false discovery rate of <0.5% for each gene. When significant duplicates of the same gene were present in a table, their log₂ ratios were averaged. The complete microarray data sets are available in Table S1 in the supplemental material. Genes appearing in each table also were required to meet the specific criteria outlined below.

Comparison of WT, pde2/pde2, and cap1/cap1 expression profiles.For the identification of genes up-regulated by the cAMP signaling pathway in germ-tube-inducing conditions (Table 2), criteria were applied as described below to identify positively regulated genes that had average log₂ Cy5/Cy3 ratios in the order pde2 ≥ WT > cap1. The difference in log₂ Cy5/Cy3 ratios between strains was termed fold change. The criteria (criteria set A) for positively regulated, cAMP-dependent genes following 1.5 h in germ-tube-inducing conditions were selected so that genes that are known to be cAMP regulated, including HWP1, ECE1, and RBT1 (8, 12, 73), were identified (Fig. 2). The difference between the average log₂ Cy5/Cy3 ratios of the WT and cap1 or pde2 and cap1 was required to be ≥0.6, which was equivalent to a minimum 1.5-fold change. The log₂ Cy5/Cy3 ratio for the WT was required to be greater than or equal to 0.6 (≥1.5-fold induction). These criteria guarantee that genes on this list are induced by WT levels of cAMP and are not induced by below-normal levels of cAMP. Genes fitting the pattern pde2 > WT > cap1 were kept as positively regulated genes and were subjected to SAM analysis. Additional criteria (criteria set B) were established to account for those genes that are maximally regulated with normal levels of cAMP but do not increase in expression above a threshold level of cAMP. This was accomplished by selecting genes that had similar Cy5/Cy3 ratios for the WT and the pde2/pde2 mutant, which were those genes that were found to have a covariance of ≤0.75 after taking spot averages for log₂ Cy5/Cy3 ratios from combined WT and pde2 hybrid groups. Genes for which the pde2 and WT Cy5/Cy3 ratios are equivalent (pde2 ≈ WT) and with averages (pde2 plus WT) of at least twofold (equal to 1 in terms of the log₂ ratio) greater than that of cap1 were kept as positively regulated genes in germ-tube-inducing conditions. Genes found to meet criteria set A or B (or both) were listed in Table 2.

• Open in new tab
• Download powerpoint

FIG. 2.

Expression patterns of known hypha-specific genes. (A) Microarray data showing the fold induction of ECE1, HWP1, and RBT1 in the WT (UnoPP-1), the pde2/pde2 (Δpde2; BPS15), and the cap1/cap1 (Δcap1; CAC1-1A1E1) strains under hypha-inducing (M199 at 37°C for 1.5 h) conditions relative to induction levels under yeast growth conditions (M199 at 25°C for 1.5 h), as described in Materials and Methods. (B) Northern blot showing ECE1 and HWP1 expression. Total RNA samples that also were used for the microarray experiments were separated by electrophoresis in a formaldehyde agarose gel, transferred to a nitrocellulose membrane, and probed with ³²P-labeled DNA probes for ECE1, HWP1, and 18S rRNA as a loading control. The membrane was exposed to X-ray film for 1 and 8 h for the detection of HWP1 and ECE1 mRNA, respectively, and for 4 h for the detection of 18S rRNA. exp., experiment. Y, yeast growth conditions; H, hyphal growth conditions.

For the identification of genes down-regulated by the cAMP signaling pathway in germ-tube-inducing conditions (Table 3), criteria were applied as described below to identify negatively regulated genes that had average log₂ Cy5/Cy3 ratios in the order cap1 > WT ≥ pde2. The criteria (criteria set C) for negatively regulated, cAMP-dependent genes following 1.5 h in germ-tube-inducing conditions were established by inversing the logic used to find positively regulated genes. The difference between the average log₂ Cy5/Cy3 ratios between WT and cap1 or between pde2 and cap1 was required to be less than 0.6. The log₂ Cy5/Cy3 ratio for the WT was required to be less than or equal to 0.6 (≤1.5-fold). These criteria guarantee that genes in Table 3 are not induced by WT or by above-normal levels of cAMP (which were found in the pde2/pde2 strain) but are possibly induced by below-normal levels of cAMP (which was found in the cap1/cap1 strain) when the list contains only genes that fit the order of cap1 > WT > pde2. Genes fitting the pattern cap1 > WT > pde2 were kept as negatively regulated genes and were subjected to SAM analysis. Additional criteria (criteria set D) were established to account for those genes that have minimal expression with normal or above-normal levels of cAMP. To select genes that had similar ratios for WT and pde2, those genes found to have a covariance of ≤0.75 after taking spot averages for log₂ Cy5/Cy3 ratios from combined WT and pde2 hybrid groups were identified. Genes with equivalent pde2 and WT log₂ Cy5/Cy3 ratios (pde2 ≈ WT) and with averages (pde2 plus WT) of at least twofold (equal to 1 in the log₂ ratio) less than that of cap1 were kept as negatively regulated genes in germ-tube-inducing conditions. Genes found to meet criteria set C or D (or both) were listed in Table 3.

Criteria and analysis methods for the identification of transcripts that were up-regulated or down-regulated by the cAMP signaling pathway during normal budding growth (mid-logarithmic phase) are described in the supplemental material.

Northern blot analysis.Northern blot analyses were performed as described before (6), using the same total RNA from each strain that was used to prepare labeled cDNA for the microarray experiments. Gene-specific probes were prepared by PCR using genomic DNA from strain UnoPP-1 as the template and the primer pairs shown in Table S2 in the supplemental material. The probe for HWP1 was the gel-purified 609-bp HWP1 insert of a phagemid clone (pBluescript+13) (73). To ensure that the lanes were loaded with equivalent amounts of RNA, a 687-bp PCR product representing the 18S rRNA gene of C. albicans SC5314 was generated and used as a probe (6).

Sensitivity to heat shock, high salt concentrations, and osmotic stress.To assess heat shock sensitivity, yeast extract-peptone-dextrose (YPD) plates spotted with cell suspensions (1 μl) of each strain, ranging in concentration from 10⁸ to 10⁴ cells/ml, were incubated at 55°C for 0, 10, 12, 15, and 20 min, followed by immediate transfer to an incubator set at 30°C. Plates were observed after 24 and 48 h of incubation at 30°C and photographed. To examine the sensitivity to high salt concentrations and osmotic stress, the serially diluted cell suspensions (10⁴ ∼ 10⁷ cells/ml) were spotted on YPD plates containing 1.5 M NaCl, 1.5 M KCl, and 2 M sorbitol, incubated at 30°C, and photographed.

Sensitivity to oxidative and nitrosative stress.To test responses to oxidative and nitrosative stress, H₂O₂ (Sigma), menadione (Sigma), sodium nitroprusside (SNP; Sigma), 3-morpholinosydnonimine (SIN-1; Calbiochem), and a peroxynitrite (Calbiochem) were used. YNB plates containing the indicated concentrations of each agent were spotted with 2 μl of 10⁶ cells/ml of each strain and were incubated at 30°C for 2 days prior to visual assessment of growth and photography. For quantitative measurements (81), cells were diluted to 10⁴ cells/ml with 1× PBS and were incubated at 37°C for 3 h with the indicated concentrations of the aforementioned agents. Fifty microliters of each sample then was cultured on YPD plates for 48 h at 30°C to measure the number of CFU. The percentage of survival of C. albicans strains in the presence of each agent at the indicated concentrations was calculated by the following formula: [(CFU of a strain incubated in PBS having the indicated concentration of an agent)/(CFU of the strain incubated in PBS only)] × 100.

Antifungal susceptibility test.The Etest (AB Biodisk) was performed to measure the fluconazole sensitivity of C. albicans cap1/cap1 and pde2/pde2 mutant strains by following the manufacturer's directions. Briefly, each yeast strain was grown at 30°C for 48 h in Sabouraud dextrose agar (Fisher Scientific) and resuspended in 0.85% NaCl to achieve a turbidity of 0.5 McFarland. Resuspended cells were evenly distributed onto solid RPMI 1640 medium (containing 2% glucose, morpholinepropanesulfonic acid [Sigma], 1.5% Bacto agar [Fisher Scientific]) by using a sterile swab and dried for at least 15 min. Fluconazole Etest strips (AB Biodisk) then were applied to the lawn of cells on the agar surface, and the plates were incubated at 35°C in a moist incubator until growth was clearly evident after 24 to 48 h. The MIC for each strain was determined using guidelines provided by the manufacturer.

Identification of genes modulated by the cAMP signaling pathway in germ-tube-inducing conditions.Modulation of cAMP levels by interfering with synthesis or breakdown as in the cap1/cap1 and pde2/pde2 mutants, respectively, affects many cellular characteristics, notably morphology, invasiveness, and resistance to nutritional starvation (5, 6). To identify genes with expression levels that correlated with cAMP levels by using the cap1/cap1 cAMP hypomorph and the cAMP pde2/pde2 hypermorph, we performed genome-wide transcriptional profiling experiments. Comparing the pde2/pde2 and cap1/cap1 mutant transcript profiles to that of WT is useful for identifying cAMP-regulated genes that influence cell metabolism and structure during the formation of nascent germ tubes. The transcriptional profiles of these two strains were compared to each other and to that of the WT.

To evaluate the reproducibility of cell cultures, RNA isolation, and sample processing, two arrays were hybridized with two separate batches of Cy3-labeled pooled RNA prepared from cultures growing as yeasts. Approximately 90% of the ORFs were within a twofold variation in normalized median intensities, as shown in the scatter plot in Fig. S1 in the supplemental material. To obtain another indication of reproducibility, the fold inductions of each gene in the two independent cultures were compared. Results of duplicate experiments in each independent culture were averaged. Approximately 90% of values differed by less than twofold. Overall, the control experiments indicated that the experimental conditions and sample preparations were reproducible.

The ability of the arrays to accurately reflect differences in the levels of cAMP-dependent transcripts under germ-tube-inducing conditions was provided by the signal intensity changes of known hypha-specific genes, such as HWP1, ECE1, and RBT1 (8, 12, 73). In particular, our previous study showed that the HWP1 transcript is markedly down-regulated in the cap1/cap1 mutant (5). HWP1, ECE1, and RBT1 were up-regulated by the activation of cAMP signaling in germ-tube-inducing conditions in the order pde2 > WT > cap1 in both microarray and Northern blot analyses (Fig. 2). These data indicate that our experimental scheme for microarray analysis was appropriate to perform a genome-wide search of cAMP-dependent genes modulated under germ-tube-inducing conditions.

Genes positively regulated by the cAMP signaling pathway in germ-tube-inducing conditions.Seventy-seven genes were differentially modulated by the cAMP signaling pathway in germ-tube-inducing conditions. The modulation of 37 genes was increased in both the WT and pde2/pde2 mutants relative to that of the cap1/cap1 mutant (Table 2). Conversely, 40 genes were found to be down-regulated (Table 3). The major portion of cAMP-dependent genes up-regulated during filamentous growth included cell surface proteins (18.9%) and those involved in glycolysis/gluconeogenesis (27.0%), ergosterol biosynthesis (10.8%), and fermentation (5.4%) (Table 2).

Six genes (HWP1, ALS3, RBT1, ECE99, PGA45, and PDC11) encoding known (12, 30, 73) or putative cell surface proteins were found to be up-regulated in hypha-inducing conditions in a cAMP-dependent manner (Table 2). With one exception, these genes exhibited reciprocal responses to hyper- or hypoactivation of the cAMP pathway. PDC11, which encodes pyruvate decarboxylase, was not elevated in the pde2/pde2 mutant above the level of the WT and appeared more similar to genes involved in glycolysis (see below). In contrast, the hypha-specific surface protein gene, HYR1 (7), was found to be hypha specific but cAMP independent, in that the fold induction in the WT (1.6) was approximately equivalent to that in the cap1/cap1 (1.7) and pde2/pde2 (1.6) mutants. In addition, several genes involved in ergosterol biosynthesis (ERG1, ERG11, and ERG251) and lipid metabolism (SUR2) were positively regulated by cAMP signaling during germ tube induction. Overall, these results indicated that during germ tube induction, pulses of cAMP change ergosterol and lipid contents in the cell membrane, in addition to remodeling the cell wall architecture with hyphal surface proteins.

Ten genes involved in gluconeogenesis or glycolysis (TDH3, TPI1, CDC19, HGT6, HGT7, HGT8, PFK1, PFK2, GPM1, and GPM2) (51, 57, 60, 69, 83) and two genes involved in fermentation (PDC12 and ADH2) were found to be up-regulated by the cAMP signaling pathway. Whereas these genes were decreased in the cap1/cap1 mutant, their expression levels in the pde2/pde2 mutant appeared similar to those of the WT, indicative of maximal expression in the presence of WT levels of cAMP. The role of the Ras/cAMP signaling pathway in glucose transport and the activation of glycolysis has been reported for S. cerevisiae (11, 67, 76). The upregulation of genes involved in glycolysis and fermentation is consistent with physiological studies performed three decades ago that demonstrated that C. albicans transitions from aerobic respiration to anaerobic glycolysis and fermentation when placed in germ-tube-inducing conditions (44).

Among the uncharacterized ORFs without significant homologies to any other genes, one ORF (YER067W; ORF6.1464) was named UCF1 (up-regulated by cAMP in filamentous growth). The expression pattern determined from the array was further proved by Northern blot analysis (Fig. 3). The UCF1 gene encodes a hypothetical protein of 201 amino acids without any conserved domains. Three ORFs (ORF6.6008, ORF6.4769, and ORF6.4770) homologous to the Tca2 retrotransposon were up-regulated. Although their roles in filamentous growth of C. albicans are unknown, it has been reported that transcription of retrotransposon-like elements is greatly up-regulated during a temperature shift from 25 to 37°C (16).

• Open in new tab
• Download powerpoint

FIG. 3.

Verification of expression patterns of genes modulated by the cAMP signaling pathway during bud-hypha transitions by Northern blot analysis. Northern blot analysis was performed with gene-specific PCR probes (genes positively regulated [A] and negatively regulated [B]) as described in the legend to Fig. 2B and Materials and Methods. Several replicated membranes were prepared using the same sample of total RNA and were used repeatedly for multiple probes after stripping the membrane as previously described (63). In general, the membrane was exposed to X-ray film for 1 to 2 days for detection of the mRNA of each probe, except for ERG1, ERG11, UCF1, and MCR1 probes, which required only 6 to 7 h of exposure, and for the XJP15 probe, which required only 1 h of exposure. The membrane probed with the control 18S rRNA was exposed for 4 h.

Genes negatively regulated by the cAMP signaling pathway in germ-tube-inducing conditions.Among the 40 genes that were down-regulated by the cAMP signaling pathway in germ-tube-inducing conditions (Table 3), two groups of genes were notable: those involved in amino acid/protein biosynthesis (35.0%) and those involved in the stress response (10.0%). Several genes encoding constituents of the ribosome (RPL10A, RPP1A, RPP0, RPL12, and RPL23A) and genes involved in protein or amino acid biosynthesis (ILV5, ACO1, ASN1, IDH1, HIP1, LYS22, and HOM6) were negatively regulated by the cAMP signaling pathway. These genes showed highest expression in the cap1/cap1 mutant, whereas the pde2/pde2 mutant appeared equivalent to the WT in terms of expression, with three exceptions (RPP0, ILV5, and HOM6). The negative regulation of the genes in this group is consistent with transitioning from respiration to aerobic glycolysis, which has been previously described to occur during the bud-hypha transition (44), and with the induction of filamentous growth during amino acid starvation through the interaction of CaGcn4 with the Ras/cAMP signaling pathway in C. albicans (78).

Of the four stress defense genes negatively regulated by cAMP (MCR1, SOD2, HSP12 [WHS11], and CCP1), all are known to be involved in the response to oxidative stress in S. cerevisiae (15, 23, 46, 59, 79). S. cerevisiae strains with mutations in MCR1, SOD2, and CCP1 genes show hypersensitivity to H₂O₂ and menadione, an O₂⁻ generator (15, 23, 46). C. albicans Sod2 regulates sensitivity to various stresses, including redox-cycling agents, high temperature, ethanol, and high concentrations of sodium and potassium (32). The negative regulation of C. albicans CCP1 by the cAMP pathway is consistent with that in S. cerevisiae in which CCP1 is regulated by POS9/SKN7 (15); POS9/SKN7 is negatively regulated by the cAMP pathway. In S. cerevisiae, HSP12 is involved in responses to heat shock, osmotic shock, and oxidative damage (59, 79).

TFS1 encodes a putative phosphatidylethanolamine-binding protein homologous to a human phosphatidylethanolamine-binding protein (Raf kinase inhibitor protein) that inhibits the Ras/cAMP signaling pathway by G-protein-mediated inhibition of AC (42). In S. cerevisiae, Tfs1 was discovered as a multicopy suppressor of the cdc25-1 mutation with a connection to the Ras/cAMP signaling pathway (61) and is a high-affinity inhibitor of carboxypeptidase Y (13). The role of Tfs1 may be conserved in fungal pathogens; recently, the C. neoformans TFS1 homologue, OVA1, was shown to be upregulated in the pka1 mutant by SAGE (serial analysis of gene expression) analysis and to negatively regulate the cAMP pathway (31).

cAMP-dependent genes modulated during unperturbed yeast growth.During logarithmic growth in yeast form, cAMP levels of the cap1/cap1 and pde2/pde2 mutants were lower and higher, respectively, than those of the WT (5). Having discovered that the cAMP signaling pathway controlled expression of hypha-specific genes such as HWP1, we wondered whether the other genes identified during hyphal growth were similar to HWP1 in having low expression levels during yeast growth or whether the cAMP pathway regulated additional genes during budding yeast growth. Among the 246 genes differentially modulated by the cAMP signaling pathway in unperturbed budding growth of C. albicans, 104 and 142 were positively or negatively regulated, respectively, by the cAMP signaling pathway (Tables S3 and S4 in the supplemental material).

Of the 104 genes positively regulated by cAMP signaling in budding growth, only 10 genes (HGT7, PFK1, GPM1, TPI1, PDC11, ERG251, PDC12, ADH2, PHO15, and UCF1) also were found to be positively regulated by cAMP signaling in germ-tube-inducing conditions as well (Table 2; also see Table S3 in the supplemental material). Of the 142 genes negatively regulated by the cAMP pathway in yeast growth, only seven (ORF6.2999, CCP1, ALS4, YET3, MAM33, HSP12, and SOD2) were negatively regulated by cAMP signaling in germ-tube-inducing conditions (Table 3; also see Table S4 in the supplemental material). Overall, glycolysis and fermentation were up-regulated by cAMP, whereas the stress response was down-regulated by cAMP, regardless of morphology. Respiration was down-regulated in hyphae but not in yeast by cAMP. Therefore, respiration and fermentation are more balanced in yeast growth than in hyphal growth, in which anaerobic fermentation prevails.

Northern blot analysis of cAMP-dependent genes regulated during bud-hypha transitions of C. albicans.Because of the importance of genes involved in ergosterol biosynthesis and the stress response in candidiasis, we wanted to determine if the profiling results were predictive of strain differences in response to azole drugs and stress-response-inducing agents. First, we verified the differential expression of these genes by Northern blot analysis (Fig. 3). In agreement with the microarray analysis, ERG1 and ERG11 mRNA levels of the pde2/pde2 mutant and the WT strain were increased in hyphal growth conditions compared to those of yeast growth conditions. Expression levels of ERG1 and ERG11 were very low in the cap1/cap1 mutant and were not increased in hyphal growth conditions (Fig. 3A). Expression of ERG251 also was induced in the cap1/cap1 mutant, but to a much lower level than that in the WT and pde2/pde2 strains.

The stress response transcripts identified in the microarray analysis (MCR1, SOD2, CCP1, and HSP12) also were confirmed to be highly up-regulated in the hypoactive cAMP signaling cap1/cap1 mutant compared to regulation of the WT and pde2/pde2 mutant strains (Fig. 3B). HSP12 was expressed only in the cap1/cap1 mutant.

Hyperactivation of the cAMP pathway confers decreased sensitivity to azoles.Because the expression of ERG11 has been reported to be increased in azole-resistant clinical isolates of C. albicans (49), we predicted that increased expression of ERG genes with potential membrane remodeling in the pde2/pde2 mutant would be accompanied by decreased sensitivity to azole drugs. To gain support for this hypothesis, antifungal susceptibility was determined for the WT, pde2/pde2, and cap1/cap1 strains. As predicted, the pde2/pde2 mutant with increased expression of ERG1 and ERG11 was eightfold more resistant (MIC = 1.0 μg/ml) to fluconazole than the WT (MIC = 0.125 μg/ml) (Fig. 4). In contrast, the cap1/cap1 mutant was slightly more sensitive to fluconazole (MIC = 0.094 μg/ml) than the WT. Reintegration of the WT CAP1 and PDE2 genes restored normal fluconazole susceptibility to both cap1/cap1 and pde2/pde2 mutants (MIC = 0.125 μg/ml). These important fluconazole susceptibility test results demonstrated for the first time for C. albicans that the expression of ERG genes, in addition to the expression of other genes that affect the composition of cell membranes, can be influenced by cAMP levels. We speculate that the presence of increased ergosterol levels leads to decreased sensitivity to the azole fluconazole.

• Open in new tab
• Download powerpoint

FIG. 4.

Increased azole resistance by hyperactivation of the cAMP pathway of C. albicans. Susceptibilities to fluconazole were determined by the Etest strip method for the following isogenic strains: WT (UnoPP-1), pde2/pde2 (Δpde2; BPS15), pde2/PDE2 reintegrants (Δpde2+PDE2; BPS9), cap1/cap1 (Δcap1; CAC1-1A1E1), and cap1/CAP1 reintegrants (Δcap1+CAP1; CACRE1). The plates were photographed, and the fluconazole MICs for each strain were determined after 48 h of incubation at 35°C.

Down-regulation of the cAMP signaling pathway increases resistance to heat shock, high salt concentrations, and oxidative and nitrosative stress.Based on results from the microarray and Northern blot analyses, we hypothesized that the activation of the cAMP signaling is inversely related to stress resistance. Therefore, we compared the viabilities of the WT strain to those of the cap1/cap1, pde2/pde2, and cap1/cap1 pde2/pde2 mutants in the presence of heat shock, high salt concentrations, osmotic shock, and oxidative/nitrosative damage. The cap1/cap1 mutant was more resistant to heat shock at 55°C than the WT and cap1/cap1 pde2/pde2 strains (Fig. 5A). In contrast, the pde2/pde2 mutant was more sensitive to heat shock than the WT and the cap1/cap1 pde2/pde2 strains (Fig. 5A), indicative of an inverse relationship between heat shock and cAMP levels. The pde2/pde2 mutant was more sensitive to sodium salt (1.5 M NaCl) than to potassium salt (1.5 M KCl) and sorbitol (2 M). Disruption of CAP1 restored normal resistance to high-sodium salt (Fig. 5B). The cap1/cap1, WT, and cap1/cap1 pde2/pde2 strains did not differ in response to the presence of NaCl or KCl or to high osmolarity.

• Open in new tab
• Download powerpoint

FIG. 5.

Effect of heat shock, high salt concentrations, and osmotic shock on the viability of the WT and the cAMP signaling mutants. The WT (Unopp-1) strain and the mutants cap1/cap1 (Δcap1; CAC1-1A1E1), pde2/pde2 (Δpde2; BPS15), and cap1/cap1 pde2/pde2 (Δcap1 Δpde2; BPS27) were tested for (A) heat shock sensitivity (55°C) or (B) high salt concentrations (1.5 M NaCl or KCl) and high osmolarity (2 M sorbitol).

To further investigate the role of the cAMP signaling pathway in response to oxidative/nitrosative stress that was predicted by the microarray analysis, cells were treated with H₂O₂ and menadione, an O₂⁻ generator. The pde2/pde2 and the cap1/cap1 mutants were more sensitive and resistant, respectively, to both H₂O₂ and menadione than the WT and cap1/cap1 pde2/pde2 strains (Fig. 6A). The heightened sensitivity of a C. albicans pde2/pde2 mutant to oxidative stress as well as to heat shock and high salt concentrations also was described in a recent study (84). The results showed that down-regulation of the cAMP signaling pathway promotes the resistance of C. albicans to reactive oxygen species (ROS)-mediated oxidative stress, whereas hyperactivation of the pathway has the opposite effect.

• Open in new tab
• Download powerpoint

FIG. 6.

Effect of hydrogen peroxide, superoxide, nitric oxide, and peroxynitrite on the viability of the WT and cAMP signaling mutants. (A) Volumes of 2 μl each of 10⁶ cells/ml of the C. albicans strains described in the legend to Fig. 5 were spotted on YNB plates containing the indicated concentrations of H₂O₂ or menadione, a O₂⁻ generator, were incubated for 2 days, and were photographed. (B) Each strain was incubated at 37°C for 3 h with the indicated concentrations of SNP, SIN-1 (an ONOO⁻ generator), or ONOO⁻ itself. Error bars indicate the standard deviations from two independent experiments. The significant difference in percent survival between the pde2/pde2 mutant and the other strains was denoted with an asterisk. The significant difference in percent survival between the cap1/cap1 mutant and the other strains was denoted with a pound sign (P < 0.05 using Bonferroni's multiple comparison test [Prism 2.0b; GraphPad Software]).

SNP, an NO generator, was used to test whether C. albicans employs the cAMP signaling pathway to respond to nitrosative stress. In agreement with the previous results of Vazquez-Torres et al., we found that NO itself does not have strong candidacidal activity (81). However, at a concentration of 10 mM, the pde2/pde2 mutant was slightly more susceptible to SNP than the cap1/cap1 mutant (Fig. 6B). The data indicate that C. albicans is resistant to NO, but elevated cAMP levels increase susceptibility to NO slightly (Fig. 6B).

Peroxynitrite (ONOO⁻), which results from the reaction of NO and O₂⁻, has more potent candidacidal activity than either NO or O₂⁻ alone (81). The strains described above were tested for sensitivity to SIN-1 (an ONOO⁻ generator) and to ONOO⁻. SIN-1 more effectively killed WT C. albicans than did SNP (Fig. 6B). Here, we found that the pde2/pde2 mutant was much more sensitive to SIN-1 than the WT and cap1/cap1 pde2/pde2 strains, whereas the cap1/cap1 mutant was more resistant. The concentration leading to approximately 50% killing of the WT strain after 3 h of incubation was 2 mM SIN-1, whereas a concentration of only 0.3 mM SIN-1 was required for approximately 70% killing of the pde2/pde2 mutant (Fig. 6B). In contrast, the cap1/cap1 mutant was highly resistant to SIN-1, in that survival was approximately 70% even at a concentration of 10 mM SIN-1 for the 3-h incubation period. ONOO⁻ killed these strains more effectively than SIN-1 (Fig. 6B). Approximately 80% of the WT strain was killed by ONOO⁻ at a concentration of 100 μM after 3 h of incubation, whereas approximately 6 mM of SIN-1 was required for killing a similar number of cells. As for the results using SIN-1, the pde2/pde2 and cap1/cap1 mutants showed much higher sensitivity and resistance, respectively, to ONOO⁻ than the WT and cap1/cap1 pde2/pde2 strains.

The steady-state mRNA levels of the stress response genes in response to peroxynitrite were determined by Northern blotting (Fig. 7). Expression of the MCR1, SOD2, and HSP12 genes generally was higher in the cap1/cap1 mutant than in the WT, but was lower in the pde2/pde2 mutant, after 1 h of incubation (Fig. 7). However, the expression kinetics of the MCR1, SOD2, and HSP12 genes were different from one another (Fig. 7). The kinetics of CCP1 expression exhibited a pattern of expression different from that of MCR1, SOD2, and HSP12. The basal expression level of CCP1 at time zero was lower in the pde2/pde2 strain, but higher in the cap1/cap1 strain, than that in the WT. In all strains, however, CCP1 expression peaked at 15 min after exposure to ONOO⁻ (Fig. 7). The expression level of CCP1 remained elevated until 60 min in the pde2/pde2 mutant, while it rapidly decreased after 15 min in the WT and cap1/cap1 mutant (Fig. 7). In general, the data indicate that the genes involved in oxidative stress defense identified through the microarray analysis also were differentially regulated in response to ONOO⁻. Taking these results together, the cAMP signaling pathway negatively modulates oxidative/nitrosative stress responses of C. albicans.

• Open in new tab
• Download powerpoint

FIG. 7.

Northern blot analysis of oxidative/nitrosative stress defense genes in response to peroxynitrite. Total RNAs isolated from C. albicans strains (WT [UnoPP-1], cap1/cap1 mutant [Δcap1; CAC1-1A1E1], and pde2/pde2 mutant [Δpde2; BPS15]) exposed to 50 μM of peroxynitrite for each indicated time of incubation were used for Northern blot analysis. A single membrane was stripped and repeatedly probed with stress-related genes (MCR1, CCP1, SOD2, and HSP12) as previously described. The membrane was exposed to X-ray film for 1 day (MCR1) or 2 days (SOD2, HSP12, and CCP1). The membrane probed with the control 18S rRNA was exposed for 4 h.