Candida albicans Hap43 is a repressor induced under low-iron conditions and is essential for iron-responsive transcriptional regulation and virulence

Abstract

Candida albicans is an opportunistic fungal pathogen that exists as normal flora in healthy human bodies but causes life-threatening infections in immunocompromised patients. In addition to innate and adaptive immunities, hosts also resist microbial infections by developing a mechanism of "natural resistance" that maintains a low level of free iron to restrict the growth of invading pathogens. C. albicans must overcome this iron-deprived environment to cause infections. There are three types of iron-responsive transcriptional regulators in fungi; Aft1/Aft2 activators in yeast, GATA-type repressors in many fungi, and HapX/Php4 in Schizosaccharomyces pombe and Aspergillus species. In this study, we characterized the iron-responsive regulator Hap43, which is the C. albicans homolog of HapX/Php4 and is repressed by the GATA-type repressor Sfu1 under iron-sufficient conditions. We provide evidence that Hap43 is essential for the growth of C. albicans under low-iron conditions and for C. albicans virulence in a mouse model of infection. Hap43 was not required for iron acquisition under low-iron conditions. Instead, it was responsible for repression of genes that encode iron-dependent proteins involved in mitochondrial respiration and iron-sulfur cluster assembly. We also demonstrated that Hap43 executes its function by becoming a transcriptional repressor and accumulating in the nucleus in response to iron deprivation. Finally, we found a connection between Hap43 and the global corepressor Tup1 in low-iron-induced flavinogenesis. Taken together, our data suggest a complex interplay among Hap43, Sfu1, and Tup1 to coordinately regulate iron acquisition, iron utilization, and other iron-responsive metabolic activities.

Fig. 1.

Construction of hap43-null mutants. (A) The amino acid sequence of C. albicans Hap43 was aligned with those of Aspergillus species HapX, N. crassa HapX, S. cerevisiae Yap5, S. pombe Php4, and S. cerevisiae Hap4 using Clustal W (50). The highly conserved N-terminal regions are boxed. The black bar indicates the putative CBC-interacting domain. The hatched bar indicates the basic domain of the putative bZip structure. The sources for the protein sequences were as follows: NfHapX, NFIA_038200 (Aspergillus Comparative Database); AfHapX, XP_747952 (NCBI); AoHapX, BAE61614 (EMBLCDS); AnHapX, AN8251.3 (Aspergillus Comparative Database); NcHapX, NCU08891.3 (Neurospora crassa Database); CaHap43, ORF19.681 (CGD); ScYap5, YIR018W (SGD); SpPhp4, SPBC16E9.01c (Schizosaccharomyces pombe GeneDB); ScHap4, YKL109W (SGD). The amino acid residues are coded as follows: red, small hydrophobic residues; blue, acidic residues; magenta, basic residues; green, hydroxyl and amine basic residues. (B) The HAP43 alleles of C. albicans were knocked out by the homologous-recombination-based SAT1 flipper method. The structure of the deletion cassette from pSFS1A-35fCaHap43 (and pSFS2A-35fCaHap43) is shown. The HAP43 coding region is entirely replaced by the SAT1 flipper cassette. The HAP43 coding region is represented by the white arrow and the upstream and downstream sequences by the hatched bars. (C) Southern blot analysis of XbaI-digested genomic DNA of the derivative strains of the hap43 mutant and the parental strain, SC5314 (HAP43/HAP43), with the specific DNA probe (black bar). The fragments corresponding to HAP43 (9.1 kb), hap43Δ::SAT1-FLIP or hap43Δ::FRT (6 kb), and HAP43-SAT1-FLIP or HAP43-FRT (7.9 kb) alleles are indicated.

Fig. 2.

Deletion of C. albicans HAP43 causes growth defects under low-iron conditions. (A) Cells of C. albicans wild-type, hap43 heterozygous and homozygous deletion, and HAP43 reconstituted strains were serially diluted and spotted onto NIM-based iron agar plates with 10 μM (LIM) or 100 μM (HIM) ferrous ammonium sulfate (Fe²⁺). Two independent lineages of isogenic mutant strains from SC5314 were examined for some of the constructs. (B) Cells from the same strains were spotted onto YPD-based agar plates. To restrict free iron, the plates were supplemented with 0, 100, 200, or 400 μM BPS. Cell numbers are shown at the top of each panel. Deletion of both HAP43 alleles led to growth defects under iron-deficient conditions (≥200 μM BPS).

Fig. 3.

Deletion of HAP43 attenuates C. albicans virulence. (A) Ten female BALB/c mice were injected via the tail vein with 5 × 10⁶ C. albicans cells, including SC5314 (orange), a heterozygous hap43 deletion mutant (black), a homozygous hap43 deletion mutant (blue), and a HAP43 reconstituted strain (red). The number of surviving mice was plotted against time (in days). Notably, one mouse in the group infected with the hap43-null mutant died soon after injection, possibly because of operational error or some unknown shock. A representative result is shown (log rank test, wild type versus HAP43/hap43Δ, P = 0.647; wild type versus hap43Δ/hap43Δ, P = 1.25E−05; reconstituted HAP43 versus hap43Δ/hap43Δ, P = 3E−04; reconstituted HAP43 versus HAP43/hap43Δ, P = 0.0581; reconstituted HAP43 versus wild type, P = 0.00527). The hap43-null mutant had the lowest virulence in comparison with the wild-type, HAP43/hap43Δ, and HAP43 reconstituted strains. (B) Kidney section of hap43Δ strain-infected mice showing no fungal colonization (×100 magnification). (C) Multiple sites (arrows) of fungal colonization were observed in wild-type-infected kidneys (×100 magnification). (D) The boxed area in panel C enlarged (×400 magnification) to show the filamentous forms of wild-type C. albicans. The cells of C. albicans were visualized by PAS staining. The cell walls of C. albicans were stained purple, whereas the kidney tissues were stained blue, as shown.

Fig. 4.

The hap43Δ strain is not defective in iron acquisition under iron-depleted conditions. (A) The iron uptake activities of C. albicans were evaluated in a 30-min uptake assay in fresh YPD medium. Normal and iron-starved stationary-phase cells were used. The wild type (WT) was taken as a normal control, whereas the sfu1Δ and ftr1Δ strains were used as a control of iron overload strain and defective-uptake strain, respectively. Intracellular iron as measured by the colorimetric assay is represented as the mean ± standard deviation (SD) in AU (left). All data were collected from at least three independent experiments. (B) C. albicans colony spots were overlaid with agarose containing TTC and incubated at 30°C. A red color on the colony spot indicates increased cell surface reductase activity. Antimycin A is a mitochondrial electron transport inhibitor. (C) All C. albicans cultures (initial cell density, 0.5 OD₆₀₀/ml) were incubated at 30°C for 5 h in high-iron (YPD) or low-iron (YPD plus 200 μM BPS) medium. Iron starvation before the growth assay increased final intracellular iron in all strains (right), in contrast to the control samples without pre-iron starvation (left). The inset is an enlarged illustration indicating the level of intracellular iron from iron-starved cells grown in the low-iron medium. The iron-starved ftr1Δ mutant showed no growth in the low-iron medium, and thus its intracellular iron content was not measured. The hap43Δ cells accumulated higher levels of intracellular iron, as did sfu1Δ cells, after 5 h of growth in the high-iron medium compared with the wild type, and the hap43Δ cells contained a level of iron similar to those of the wild type and the sfu1Δ mutant when grown in the low-iron medium. (D) Cells were serially diluted and spotted onto YPD plates containing the iron-dependent free-radical generator phleomycin and incubated at 30°C for 1 day. Cell numbers are shown at the top of each image. Only the sfu1Δ mutant was hypersensitive to phleomycin, and addition of 100 μM BPS partially inhibited this effect.

Fig. 5.

Hap43 represses the transcription of genes encoding iron-dependent proteins under iron-deprived conditions. Quantitative real-time PCR was performed for selected iron-responsive genes. Cells were inoculated into high-iron (YPD) or low-iron (YPD plus 400 μM BPS) medium, incubated at 30°C for 5 h, and used for RNA isolation. The threshold cycle (C_T) value of each gene was derived from the average of three technical repeats in each experiment. The ΔC_T value was determined by subtracting the average C_T of endogenous ACT1 from the average C_T of target genes. The ΔΔC_T of each target gene was calculated by subtracting the ΔC_T value of the corresponding calibration value (from a wild-type sample grown under high-iron conditions). The average ΔΔC_T and SD were determined from at least triplicate experiments. The relative fold change of each gene is shown as 2^−ΔΔCT. The de-repression of each gene in the hap43Δ mutant in response to iron depletion (YPD plus 400 μM BPS) is highlighted by an asterisk.

Fig. 6.

One-hybrid analysis demonstrated that Hap43 is an activator under iron-rich conditions and a repressor under iron-deficient conditions. LexA-Hap43 binds to the LexA operator (LexAOP) upstream of the basal promoter of the lacZ reporter gene and modulates the expression of lacZ. LacZ activity was measured by a liquid β-galactosidase assay and was also shown with the X-Gal overlay assay. The known activator Gcn4 and the repressor Nrg1 fused with LexA were used as controls for the assays. The strains expressing LexA protein only and strains without the LexA operator upstream of the lacZ reporter gene generated a basal expression level of lacZ. (A) C. albicans Hap43 activity was assayed by the β-galactosidase method. In the mid-log phase cells grown in YPD (i.e., iron-rich medium), Hap43 displayed activation activity in cells carrying LexAOP. (B) Dilutions of YPD overnight cultures (5 μl; 5.0 OD₆₀₀/ml) were spotted on YPD agar plates supplemented with 50 μM ferrous ammonium sulfate (Fe²⁺) or 200 μM BPS. The plates were incubated at 30°C overnight. An X-Gal overlay assay was performed on the agar plates, and the plates were incubated at 30 or 37°C for 21 h. The basal strains developed a light-blue color because of the basal level of lacZ expression; activator-expressing strains generated a darker blue color. Repressor-expressing strains produced completely white colonies. (C) The iron-dependent conversion of Hap43 activity was determined with the β-galactosidase assay using cells grown in YPD under iron restriction by adding 0, 25, 50, 100, 200, 400, or 800 μM BPS. (D) The repression activity of Hap43 under iron-deprived conditions was assayed using reporter strains that included (GCRE)₅ (85) between the LexA operator and the basal promoter of lacZ. These strains possess higher levels of basal expression and were used to assay the repression activity of Hap43 in response to iron depletion. The strains expressing LexA only (without Hap43 fusion) were used as basal controls. Two independent clones of each strain were tested. (E) The LacZ activities of the LexAOP-(GCRE)₅ strain were also displayed by X-Gal overlay assay.

Fig. 7.

Iron deficiency induces nuclear accumulation of Hap43. The C. albicans CAI4 strain was transformed with a GFP-fused Hap43 under the control of the ACT1 promoter. Cells expressing GFP-Hap43 were grown in YPD overnight and inoculated into iron-depleted YPD medium (400 μM BPS) or iron-sufficient medium (YPD without BPS). (A) Mid-log-phase cells were fixed, washed, and examined with differential interference contrast optics. Nuclei were visualized by DAPI staining. DAPI (red) and GFP (green) signals were viewed by laser scanning confocal microscopy, and the images were superimposed. GFP-Hap43 was dispersed throughout the cell when iron was sufficient and accumulated in the nucleus in response to iron depletion. (B) For protein extractions, samples were prepared as for panel A but without fixation. Whole-cell extracts (80 μg each) were resolved by SDS-PAGE, and GFP-Hap43 was detected with anti-GFP. Protein molecular mass standards are indicated on the left (kDa). The CAF2-1 prototroph was used as the empty control. Two independent GFP-HAP43 clones are shown. GFP-Hap43 was expressed under both high-iron (no BPS) and low-iron (400 μM BPS) conditions. The black arrow indicates the predicted molecular mass of GFP-Hap43, whereas the white arrow indicates the exact migration of GFP-Hap43. These Western blots were repeated several times with anti-GFP from different manufacturers; only one set of data is presented.

Fig. 8.

Hap43 is essential for the flavinogenesis induced by iron starvation. (A) The absorbance at 446.3 nm of each supernatant from the stationary-phase culture in NIM (Fe⁻) or SC medium (Fe⁺) is displayed. This measurement is directly correlated with the levels of flavin in the supernatants. Quantitative data were calculated from at least three independent experiments. The error bars indicate SD. (B) Supernatants prepared as indicated for panel A in NIM or SC medium were collected and photographed under UV light. Flavin molecules in the supernatants emitted fluorescence by UV excitation. The result from one of the replicates is displayed. The absorbance and fluorescence values of supernatants from the wild-type, hap43 deletion, and HAP43 reconstituted strains and sfu1- and tup1-null mutants were compared.

Fig. 9.

A simple model of Hap43-mediated iron metabolism is proposed. When cells encounter a shift from iron sufficiency to iron deficiency, the expression of HAP43 is released from repression by Sfu1. In turn, Hap43 is induced to repress the expression of SFU1, leading to depression of many iron uptake genes and elevated iron assimilation. In contrast, Hap43 is also responsible for the attenuation of excess iron consumption by repressing the expression of genes encoding iron-dependent proteins. The function of Hap43 depends on its low-iron-induced nuclear accumulation and controls the normal growth, virulence, and flavinogenesis in iron-limiting states. The Tup1 corepressor acts as a coregulator of the expression of some iron uptake genes and flavinogenesis, possibly cooperating with Hap43 and Sfu1. The CCAAT-binding complex (Hap2/3/5) may take part in this complicated iron metabolism, together with Hap43, but its precise role remains to be elucidated.