|Home | About | Journals | Submit | Contact Us | Français|
This work has identified regulatory elements in the major fungal pathogen Candida albicans that enable response to nitrosative stress. Nitric oxide (NO) is generated by macrophages of the host immune system and commensal bacteria, and the ability to resist its toxicity is one adaptation that promotes survival of C. albicans inside the human body. Exposing C. albicans to NO induces upregulation of the flavohemoglobin Yhb1p. This protein confers protection by enzymatically converting NO to harmless nitrate, but it is unknown how C. albicans is able to detect NO in its environment and thus initiate this defense only as needed. We analyzed this problem by incrementally mutating the YHB1 regulatory region to identify a nitric oxide-responsive element (NORE) that is required for NO sensitivity. Five transcription factor candidates of the Zn(II)2-Cys6 family were then isolated from crude whole-cell extracts by using magnetic beads coated with this DNA element. Of the five, only deletion of the CTA4 gene prevented induction of YHB1 transcription during nitrosative stress and caused growth sensitivity to the NO donor dipropylenetriamine NONOate; Cta4p associates in vivo with NORE DNA from the YHB1 regulatory region. Deletion of CTA4 caused a small but significant decrease in virulence. A CTA4-dependent putative sulfite transporter encoded by SSU1 is also implicated in NO response, but C. albicans ssu1 mutants were not sensitive to NO, in contrast to findings in Saccharomyces cerevisiae. Cta4p is the first protein found to be necessary for initiating NO response in C. albicans.
This study analyzes a key defense mechanism that allows Candida albicans to tolerate the toxic effects of nitric oxide produced from numerous sources in its mammalian host. C. albicans is a dimorphic fungus responsible for a considerable proportion of fungal infections in humans. A normal inhabitant of the digestive and genitourinary systems, it can switch from an innocuous yeast form into an invasive mycelial one. In addition to superficial infections like oral and vaginal thrush that can occur in relatively healthy individuals, patients with compromised immune systems may succumb to entrenched infections that can become deadly.
One process occurring in host mammalian systems and their commensal bacteria that limits the growth of microflora like C. albicans is the discharge of nitric oxide (NO), a reactive nitrogen species that is toxic at high concentrations. Consequently, C. albicans must tolerate NO generated from dietary nitrates and nitrites by bacteria in the oral cavity (37) or in the gut (45) or from nitrite-treated food moving through the digestive system (18), as well as NO nonenzymatically generated in the stomach, in urine, or on the skin. As an invasive pathogen, it typically encounters low levels of NO synthesized by many mammalian cell types, as well as higher concentrations of NO released by immune cells in host defense.
Like many microorganisms that are able to infect mammalian tissues, C. albicans can counter various immune system attacks. We previously reported that the flavohemoglobin Yhb1p in C. albicans detoxifies NO (50), a finding that was subsequently confirmed by Hromatka et al. (19). This enzyme acts as a nitric oxide dioxygenase; it neutralizes NO toxicity by incorporating O2 to produce nitrate (NO3−). Flavohemoglobins are part of the ancient family of globin proteins, which also includes the better-known hemoglobins of vertebrate blood cells and the myoglobins of muscle. Flavohemoglobins are characteristically composed of a single polypeptide with a single heme group in an N-terminal domain homologous to vertebrate and plant hemoglobins and a C-terminal domain containing flavin adenine dinucleotide and NADPH binding sites (39).
Homologous flavohemoglobins are predicted to occur in many other microorganisms on the basis of genome sequencing and experiments using PCRs (31). Flavohemoglobin function in NO response has been reported in many bacterial species, such as Escherichia coli (38), Salmonella enterica serovar Typhimurium (11, 46), Yersinia pestis (44), Bacillus subtilis (41), and Erwinia chrysanthemi (5), as well as the protozoan social amoeba Dictyostelium discoideum (20) and the fungi Saccharomyces cerevisiae (27), Cryptococcus neoformans (10, 12, 33), and C. albicans (19, 50). Under hypoxic conditions, the mitochondria respiratory chain of S. cerevisiae will use endogenous nitrite instead of oxygen as an electron acceptor, thereby generating nitric oxide internally (9). The Yhb1p flavohemoglobin redistributes to the mitochondria under hypoxic conditions and consumes nitric oxide there (8). Thus, flavohemoglobins protect against nitric oxide coming from either exogenous or endogenous sources.
Flavohemoglobins are also a central mechanism for pathogenic fungi to counteract nitrosative stress during infection (32). For example, C. neoformans cells lacking the FHB1 flavohemoglobin show defects in NO consumption, increased inhibition of growth by NO, and decreased virulence (12). In C. albicans, similar results were observed in cells lacking the YHB1 flavohemoglobin (19, 50). Studies of mice lacking the inducible nitric oxide synthase (NOS) show that FHB1-dependent virulence of C. neoformans requires this source of NO (12) while YHB1-dependent virulence of C. albicans does not (19). One interpretation of this latter result is that the role of YHB1 in C. albicans virulence is independent of NO. However, the inducible NOS-dependent defense against systemic C. albicans infection via the tail vein route appears relatively small. Further examination of this question will require virulence assays in which C. albicans infection starts in mucosal tissues where NO levels are typically much higher.
Genome scale, inducible responses to NO in pathogenic fungi have been analyzed in C. neoformans (10, 33), Histoplasma capsulatum (35), and C. albicans (19). NO induces a large number of stress response genes in these fungi; the C. albicans study identified a gene set specifically induced by NO and not other stresses. Genes of this small set include (19) the YHB1 flavohemoglobin and SSU1, a sulfite transporter whose S. cerevisiae homolog is involved in the NO response, thus signifying that C. albicans has a means to sense NO in its environment; this signaling pathway has yet to be characterized. The use of NO as a signaling molecule is widespread, and many animals possess NOSs. To date, several mechanisms of NO sensing and response have been elucidated in other organisms. For example, in mammals soluble guanylyl cyclase acts as a NO receptor and stimulates production of cyclic GMP (6). A number of bacterial species utilize various simple pathways in which the NO sensing protein also functions as the transcription factor which induces the nitrosative stress defense mechanism. In E. coli, for example, the NorR protein binds NO at its N-terminal domain and regulates expression of a nitric oxide reductase gene by binding DNA with a helix-turn-helix motif in its C terminus (16). However, little is known about fungal nitrosative stress response mechanisms, and homologs of pathway components found in other organisms seem to be absent in fungi. Thus far, a single fungal transcription factor involved in NO sensing has been found—the C2H2 zinc finger factor Fzf1p in S. cerevisiae (42).
NO signaling in C. albicans, however, appears to be governed differently. We have focused on uncovering components of an NO regulatory pathway in C. albicans. In this work, we have not attempted to distinguish between effects from direct action by NO and indirect action by NO-derived products. This study identifies a putative transcription factor that is involved in nitrosative stress response in C. albicans but is not homologous to Fzf1p. Cta4p, a previously uncharacterized DNA-binding protein of the Zn(II)2-Cys6 family of fungal transcription factors, is necessary for induction of YHB1 transcription when C. albicans encounters nitrosative stress.
Strains used in this project are listed in Table S2 in the supplemental material. (Strains used in regulatory region analysis are derived from RM1000 ; deletion strains and the corresponding complementation strains were constructed from strain SN152 or SC5314.) C. albicans strains were routinely grown in either YPD (1% yeast extract, 2% peptone, 2% dextrose) medium for Ura+ strains or in YPD medium plus uridine (1% yeast extract 2% peptone, 2% dextrose, and 40 μg/ml uridine) for Ura− strains. YPD medium used with the NO donor dipropylenetriamine NONOate (DPTA NONOate) was adjusted to pH 7.4 and filter sterilized. C. albicans transformants were grown on selective synthetic medium (2% dextrose, 6.7% yeast nitrogen base with ammonium sulfate, and appropriate auxotrophic supplements).
The YHB1 regulatory region-open reading frame (ORF)-lacZ constructs were created from the pAU95 plasmid (49). Genomic DNA of C. albicans strain RM1000 was amplified by PCR using TripleMaster polymerase (Eppendorf) in HighFi buffer and the primers YHB1-5′-F and YHB1-ORF-R. Primers used in constructing linker scanning mutant strains are listed in Table S3 in the supplemental material. The reaction parameters were as follows: 94°C for 2 min; then 30 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 2 min; and finally 72°C for 10 min. The PCR product and the pAU95 plasmid were each digested with Acc65I and PstI, gel purified, and ligated together. The regulatory region of the new pAU95-YHB1-lacZ plasmid was sequenced to check sequence fidelity. Since C. albicans is diploid, each gene has two alleles. To ensure that the mutant regulatory regions were compared for the same allele (orf19.3707) each time, mutant regulatory regions were created by PCR amplification from pAU95-YHB1-lacZ. Substitution mutations were made in two consecutive rounds of PCR. First, two reactions were carried out in parallel to amplify the YHB1 regulatory region and ORF DNA on either side of the intended mutation, with the substituted nucleotides on a linker segment of the internal primers. Second, the two PCR products were fused and amplified in a third reaction. The reaction parameters were as follows: the fusion reaction was incubated at 94°C for 2 min, which was followed by 2 cycles of 94°C for 45 s, 40°C for 45 s, and 72°C for 2 min. Then, end primers YHB1-5′-F and YHB1-ORF-R were added, and the amplification reaction was carried out as follows: 94°C for 2 min; 30 cycles of 94°C for 45 s, 55°C for 45 s, and 72°C for 2 min; and finally 72°C for 10 min. The final PCR product and the vector pAU95-YHB1-lacZ were each digested with Acc65I and PstI, and the appropriate fragments were ligated together. Plasmids containing the substitution mutation were identified by a new restriction site (XhoI, PmeI, or MfeI) and confirmed by DNA sequencing. The plasmids were each transformed into strain RM1000 using a lithium acetate method (17). Each plasmid was targeted to the intergenic region downstream of the HWP1 gene by digesting the reporter gene at the PshAI site within the 3′ untranslated region (UTR) of HWP1 on the plasmid and then transforming the linearized plasmid. Insertion at the correct location in the genome was verified by PCR amplifying the chromosome-plasmid junctions using two pairs of primers: the pair Msub border F1 and Msub border R1 and the pair Msub border F2 and Msub border R2. Reaction parameters were as follows: 94°C for 2 min; then 35 cycles of 94°C for 30 s, 53°C for 30 s, and 72°C for 2 min; and finally 72°C for 8 min.
Procedures were modified from Kaiser et al. (22). Overnight cultures were diluted to an optical density at 600 nm (OD600) of 0.08 in 3.0 ml of YPD medium and grown for 3 h at 37°C in a tissue culture rotator. Sodium nitrite (NaNO2) dissolved in water was added to a final concentration of 10 mM, or an equal volume of water alone was added as a control. Cultures were incubated at 37°C for 1 h. Then, cells were pelleted by centrifugation, the supernatant was aspirated off, and 0.8 ml of Z buffer, 80 μl of 0.1% sodium dodecyl sulfate (SDS), and 120 μl of chloroform were added and mixed by vortexing. Samples were preincubated for 5 min at 37°C, and then 56 μl of a 1.355 mg/ml chlorophenol red-β-d-galactopyranoside solution in Z buffer (100 mM sodium phosphate, 10 mM KCl, 1 mM MgSO4, 38 mM β-mercaptoethanol, pH 7.0) was mixed in. When color changed from yellow to dark orange, the reaction was stopped with 0.4 ml of 1 M NaCO3. Cell debris was spun down by centrifugation, and β-galactosidase-catalyzed conversion of chlorophenol red-β-d-galactopyranoside to chlorophenol red was measured by absorbance at 578 nm in 1 ml of supernatant. Activity was expressed as β-galactosidase units, calculated as follows: OD578/[OD600 of assayed culture × volume assayed × time]. Each mutation was tested in at least three independent C. albicans transformant strains in three independent trials (minimum of nine data points). All trials were performed with wild-type controls, consisting of strains carrying the wild-type YHB1 regulatory region and ORF joined to the lacZ reporter gene. Data were normalized in each trial to the mean activity measurements of induced wild-type controls.
Procedures for isolating proteins using NO-responsive element (NORE) DNA-coated Dynabeads and their identification by tandem mass spectrometry are available in Data S4 in the supplemental material.
Homozygous gene deletion mutants of CTA4, STB5, WAR1, ZCF29, ZCF36, and YHB1 were constructed in strain SN152 using methods previously described (36). For primers used in constructing these deletions see Table S3 in the supplemental material. The ORF of each transcription factor gene was replaced with cassettes containing Candida maltosa LEU2 and Candida dubliniensis ARG4, whose 5′ UTRs and ORFs were amplified by PCR from plasmids pSN40 and pSN69. Replacement of the first gene copy with the C. maltosa LEU2 marker was confirmed by two PCRs using two pairs of primers: (gene) 5′ del check with Universal primer 5 and Universal primer 2 with (gene) 3′ del check, each amplifying a junction of the insert [where “(gene)” is the gene of interest]. Replacement of the second gene by CdARG4 was verified by PCR of each junction again, but this time with primers specific to the CdARG4 gene: (gene) 5′ del check with ARG4 3′ check-F (+1261) and ARG4 5′ check-R (+131) with (gene) 3′ del check. A negative test by PCR within the ORF DNA with the primers intracheck-F and intracheck-R specific for each gene verified that no copies of the genes remained in the genome.
The ssu1Δ/ssu1Δ mutant was constructed using the SAT1-flipper cassette (40). Briefly, a 300-bp fragment upstream of SSU1 was amplified by PCR from genomic DNA and cloned between KpnI and XhoI sites on pSFS1. Next, a 300-bp fragment downstream of SSU1 was amplified and cloned between the SacII and SacI sites on the same vector. The resulting plasmid was digested with KpnI and SacI to release the disruption cassette. The gel-purified disruption cassette was used to transform either SC5314 or the cta4Δ/cta4Δ mutant strain by electroporation; transformants were selected on YPD plates containing 100 μg/ml nourseothricin. Correct integration was confirmed by PCR. Nourseothricin-sensitive colonies were obtained following growth on yeast carbon base-bovine serum albumin medium to excise the cassette, as described previously (40), and confirmed by PCR. The second allele was disrupted using the same process.
Complementation plasmids were developed from the plasmid CIp20 (13) (GenBank accession no. DQ015894). CTA4 or YHB1 was amplified from SN152 genomic DNA using the primer pair CTA4 gene UP1 and CTA4 gene DN1 or the pair YHB1 gene UP1 and YHB1 gene DN1. For each gene the region amplified included the putative regulatory region, the ORF, and the 3′ UTR. For primers used in constructing complementation strains, see Table S3 in the supplemental material. The URA3 marker on the plasmid was replaced with either the CTA4 gene or the YHB1 gene by digesting the insert and vector with the restriction enzymes SacI and NheI, gel purifying the DNA fragments, and ligating them together. Complementation plasmids were transformed into the appropriate strain, cta4Δ/cta4Δ or yhb1Δ/yhb1Δ, and checked for integration into the desired location at the RPS1 locus by PCR using primers CIp20-upcheck-F with CIp20-upcheck-R for one junction and CIp20-downcheck-F with CIp20-downcheck-R for the other junction. The ssu1Δ/ssu1Δ mutants were complemented by replacing the 5′ fragment of SSU1 in the disruption plasmid with the complete ORF. This cassette was excised and used to transform the mutant strains. Correct integration into the SSU1 locus was confirmed by PCR.
Overnight cultures were diluted to an OD600 of 0.07 and grown at 30°C for 2 h in YPD medium buffered with 50 mM sodium phosphate, pH 7.4. Cultures were divided into 3-ml aliquots. Freshly prepared DPTA NONOate (a polyamine NO donor with a half-life of 3 h at 37°C at pH 7.0 to 7.4) solutions were dissolved in the same medium and added to each aliquot to final concentrations of 0, 0.1, 0.2, 0.3, 0.4, 0.5, or 1 mM. Cultures were incubated at 30°C for 3.5 h, and cell density was compared by OD600 measurements. Data were normalized by setting the cell density measurement in the 0 mM DPTA NONOate condition as equal to 1. Remaining cell density measurements (at OD600), recorded over the range of DPTA NONOate concentrations for each strain, were then expressed as a fraction of the maximum for that strain recorded during that particular trial. Three independent trials were performed on each set of strains.
Overnight cultures were diluted to an OD600 of 0.08 and grown in YPD medium for 4 h at 30°C. DPTA NONOate (0 mM or 0.1 mM) or sodium nitrite (0 mM or 10 mM) was added, and cultures were incubated at 30°C for 15 min. The cultures were chilled in ice water, and cells were harvested by centrifugation. Cell pellets were snap-frozen in dry ice and ethanol and stored at −80°C. Total RNA was isolated by hot phenol extraction (43) except that phenol-chloroform-isoamyl alcohol (125:24:1), pH 4.3, was used for all phenol solutions indicated above. RNA (25 μg per lane) was electrophoresed, blotted onto a Hybond-N+ nylon membrane (GE Healthcare), probed with YHB1 DNA labeled with [α-32P]dATP, and then stripped in boiling 0.1% SDS and reprobed with TEF1. Blots were exposed to a phosphor imaging plate. Images were scanned with a Fujifilm FLA-50001R imager, and contrast was adjusted in some scans across the entire image by using Adobe Photoshop, version 10.0. For sequences of the primers used in PCRs to make probes see Table S3 in the supplemental material.
The plasmid pMGC2 was constructed to generate a core module for targeted insertion of DNA encoding nine copies of the Myc epitope into the protein coding region of CTA4 and other C. albicans genes. (For sequences of the primers involved seeTable S3 in the supplemental material.) The ACT1 3′ UTR was digested from pAU36 (49) using BamHI and XbaI and ligated into pSN52 (CdHIS1) (36) digested with BamHI and SpeI, producing pMGC1. DNA encoding nine copies of the Myc tag was amplified by PCR from pYM6 (24) using 9×Myc-UP and 9×Myc-DN primers, digested with BglII and HindIII, and ligated into pMGC1 digested with BamHI and HindIII, producing pMGC2. The fusion PCR procedure used for COOH-terminal tagging of Cta4p with Myc9 is closely similar to that used by Noble and Johnson (36) for high-efficiency deletion of C. albicans genes by homologous recombination. The core module (B) containing Myc9/ACT1-3′ UTR/CdHIS1 was amplified by PCR from pMGC2 using the linking primers 2 and 5. One DNA segment (A) of ~300 bp within the protein coding region of CTA4 and just upstream of the CTA4 stop codon was amplified by PCR from genomic DNA using primers CTA4-1 and CTA4-3. Another DNA segment (C) of ~300 bp just downstream of the CTA4 stop codon and within the 3′ UTR of CTA4 was amplified by PCR using primers CTA4-4 and CTA4-6. DNA segments A, B, and C were fused and amplified by PCR in one reaction using primers CTA4-1b and CTA4-6b. These b primers have the same 5′ end as the CTA4-1 and CTA4-6 primers, respectively, but are 5 to 10 nucleotides longer, allowing greater yield and purity of the final targeting module (A-B-C fusion product). The targeting module was transformed into a CTA4/cta4Δ::LEU2 strain using a lithium acetate-based protocol (17), and successful targeting of the core module to CTA4 was confirmed in transformants by Western blotting and PCR amplification of the 3′ junction (linking primer 2 and CTA4 3′ del check) and 5′ junction (primers CTA4 myc upcheck F and CTA4 myc upcheck R). The Myc9-tagged CTA4 gene is fully functional because the DPTA-NONOate sensitivity of the heterozygous transformant was the same as the wild-type (SN152) strain while the growth of a cta4Δ/cta4Δ strain was inhibited when exposed to the NO-releasing drug.
Chromatin immunoprecipitation (ChIP) assays were performed as described by Aparicio et al. (3) with some modifications. Briefly, 50-ml cultures were grown in YPD medium to an OD600 of ~1.0 to 1.2. Next, 10 mM sodium nitrite was added to induce nitric oxide response in half of the cultures for 15 min. Cells were then cross-linked with 1% formaldehyde for 20 min, washed with Tris-buffered saline, and lysed by bead beating in 1 ml of lysis buffer (50 mM HEPES, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate, and 0.1% SDS) with 2 mM phenylmethylsulfonyl fluoride. Chromatin was sheared to an average size of 500 bp by sonication with a Misonix Sonicator 3000 with Cup Horn. The input sample (100 μl) was taken, and DNA was extracted as described below. The remaining sample (900 μl) was incubated for 1.5 h with 50 μl of protein G resin (GenScript) and 20 μl of c-Myc (9E10) antibody (SantaCruz). Beads were washed three times with lysis buffer and eluted in 100 μl of elution buffer (50 mM Tris, pH 7.5, 10 mM EDTA, 1% SDS) at 65°C for 10 min. Cross-linking was reversed by a 2-h incubation with 10 μl of 10 mg/ml proteinase K at 42°C and then for 6 h at 65°C. DNA was purified by phenol-chloroform extraction and ethanol precipitation. PCR was carried out in 50-μl reaction volume using 1/10 of the immunoprecipitated DNA and 1/15 of 10% input DNA. The PCR primers (TAGGTTAAAGCGCTCCGGCATTTTCC and GTTTAATGGTTTTCCCGCTGATGCTCC) were designed to flank the NORE of the YHB1 regulatory region. Samples were run through 26 PCR cycles and then separated on a 2.5% agarose gel. The gel was loaded with 22 μl of immunoprecipitation sample and 5 μl of input sample.
Overnight cultures were diluted to an OD600 of 0.05 and grown at 30°C for 1 h. DPTA NONOate was added to a concentration of 4 mM, and cultures were incubated for 5 h. Cells were collected by centrifugation and washed with ice-cold water. Cell pellets were frozen in dry ice and ethanol and stored at −80°C. Total RNA was isolated by hot phenol extraction (43). First-strand cDNA was synthesized according to the Invitrogen SuperScript II RT protocol. A 12-μl reaction mixture including 100 ng of RNA, 0.5 μg of oligo(dT), and 10 pmol of the deoxynucleoside triphosphates was heated at 65°C for 5 min, chilled on ice, and centrifuged. First-strand buffer and 0.1 M dithiothreitol were added, and the reaction mixture was incubated at 42°C for 2 min. SuperScriptII RT was added, and the reaction mixture was incubated at 42°C for 50 min and heat inactivated at 70°C for 15 min. The cDNA product was diluted 1:10 twice, and 2 μl of each dilution was used as a template in the PCR. Primers (for a list, see Table S3 in the supplemental material) were designed to amplify a ~300-bp internal fragment of SSU1 or ACT1 as a control.
Mouse virulence assays were performed as described previously (28). Adult, female, outbred ICR mice (21 to 25g) were obtained from Harlan. Cultures of C. albicans were grown in YPD medium to mid-exponential phase and collected by centrifugation. Cells were washed and resuspended at 1 × 107 cells/ml in phosphate-buffered saline, and mice were infected via tail vein injection with 106 C. albicans yeast-form cells. The group size was 10 mice per strain. Infected mice were subsequently monitored for signs of infection and euthanized when moribund, according to approved protocols. Survival data were analyzed with Prism 3 (Graphpad Software) using a log rank test. Statistical significance was defined as a P value of less than 0.05. All animal assays were conducted in accordance with protocols approved by the University of Texas Health Science Center Animal Welfare Committee.
NO exposure causes accumulation of YHB1 mRNA (19, 50), and we investigated this effect to find a DNA-binding transcription factor involved in this signaling pathway. In S. cerevisiae, the regulatory region of a gene is typically immediately upstream of the ORF within a kilobase of DNA from the ATG start codon. C. albicans belongs to the same fungal family and seems to be similarly configured with compact regulatory regions. We analyzed the regulatory region of the YHB1 gene to locate a cis-regulatory DNA element that could be a cognate binding site for a transcription factor.
Previously, it was found that the 1-kb region preceding the YHB1 ORF could confer nitrite inducibility on a reporter gene (50). Preliminary evidence with our β-galactosidase reporter system indicated that a 479-bp region upstream of YHB1 had similar activity levels as the 1-kb upstream sequence (data not shown). It was also found that substituting the YHB1 3′ flanking region for the HWP1 3′ UTR on the plasmid did not enhance the responsiveness of the YHB1-lacZ construct. Smaller 10-bp substitution mutations were developed to analyze the region −481 to −352 bp from the start codon, but these failed to show a loss of nitric oxide induction (see Fig. S5 in the supplemental material). The remaining 352 bp of the regulatory region was analyzed for potential cis-acting elements with the program ClusterBuster (15) (http://zlab.bu.edu/cluster-buster/), which compares selected or user-supplied transcription factor binding site matrices against a large background portion of DNA from the chosen organism. ClusterBuster finds possible binding sites based on the concept that regulatory regions are often patterned so that cis-acting elements are arranged in tight clusters. When we entered a 13-kb stretch (contig6-2060) of C. albicans DNA containing the YHB1 ORF and selected all prelisted matrices for transcription factors on the page, the program identified one 21-bp segment at −228 to −208 bp from the start codon in our 352-bp area of interest, shown underlined here among DNA in its vicinity: CAATTATTACCGTCGGGGAGCATCAGCGGGAAAAC CATTAAACATGTTTTTT.
Mutants were constructed to test the activity of this region in the β-galactosidase reporter assay. The crucial nucleotides turned out to contain a palindromic sequence at the upstream edge of the 21-bp sequence. Testing more linker scanning substitution mutations—Msub14 through Msub31—with the reporter assay refined the boundaries of the NORE sequence to the 12-bp sequence TATTACCGTCGG essential for nitrite induction, at −238 to −227 bp from the start codon. Results from the most informative mutations of this set are shown in Fig. Fig.11 (for the complete set, see Fig. S6 in the supplemental material). These mutations define the edges of the element and do not elucidate which internal base pairs were critical for induction in response to nitrosative stress. The analysis of the YHB1 regulatory region for NO-responsive sequences was not exhaustive in the 352 bp upstream and closest to the start codon. Therefore, it remains possible that another cis-acting element required for NO-responsiveness is present in the YHB1 regulatory region.
Isolation of candidate proteins became feasible upon identification of mutations in the regulatory region which prevented nitrosative stress-specific induction of YHB1 transcription. Transcription factor binding sites are typically 6 to 12 bp long in S. cerevisiae, so a 33-bp NORE sequence was conservatively chosen for purifying specific binding proteins by affinity. For comparison we also created another sequence, identical except for a 5-bp substitution mutation also used in the YHB1 regulatory region mutant Msub19, which had been shown to be uninducible by nitrite in the β-galactosidase reporter assay (Fig. (Fig.1).1). Each sequence was biotinylated at the 5′ end and affixed to streptavidin M-280 Dynabeads (Dynal Biotech).
Our strategy for isolating proteins that bind to the NORE was largely based on an approach described by Kroeger and Abraham (25). The major changes made were using crude whole-cell extracts rather than nuclear extracts, starting the affinity binding purification with a 10-fold higher protein concentration, and visually selecting proteins on an SDS-polyacrylamide gel for mass spectrometry analysis based on their binding specificity. Whole-cell extract containing 20 mg of protein from C. albicans strain RM1000 was incubated with the beads in the presence of 10 mM sodium nitrite (an NO donor) and 3 nmol of competitor DNA, identical to the 5-bp substitution mutant DNA used on one set of beads. The beads were washed with buffer, and proteins attached to the affinity beads were then eluted in a high-salt solution. The eluate contents were separated by SDS-polyacrylamide gel electrophoresis and detected by Coomassie staining. A similar silver-stained polyacrylamide gel is shown for illustration (Fig. (Fig.2)2) because the higher contrast allows greater band visibility.
We selected five protein bands that appeared more abundant in the eluate from the wild-type NORE than in the mutant NORE eluate; only three of the five chosen bands are clearly visible in Fig. Fig.2.2. Each of these five bands showed NORE selective enrichment in at least four independent preparations. We analyzed the protein bands by tandem mass spectrometry mass spectrometry using a linear quadrupole ion trap mass spectrometer (Thermo Electron).
Before mass spectrometry analysis, protein bands were first digested with trypsin, and high-performance liquid chromatography separated the peptide components within each band. Structural data were obtained by fragmenting the eight most intense peaks from each band using collision-induced dissociation. Spectra from a mass spectrometry scan of the resulting fragments were compared against predicted spectra generated by Sequest from sequences in the Candida Genome Database (http://www.candidagenome.org/). Five putative transcription factors were identified: Cta4p, Stb5p, War1p, Zcf29p, and Zcf36p. All five belong to the Zn(II)2-Cys6 zinc cluster family of fungal transcription factors, of which the best-studied member is S. cerevisiae Gal4p. The full list of proteins identified after DTASelect filtering of the mass spectrometry data is shown in Table S7 in the supplemental material. Because the eluates still contained a complex mixture of proteins and because one-dimensional electrophoresis cannot separate multiple proteins of similar sizes, it cannot be assumed that the proteins isolated and identified here have preferential binding to the NORE DNA.
To find out whether any of the candidate proteins are biologically relevant in NO response, deletion strains were constructed for each of the transcription factors identified by mass spectrometry using a gene deletion system designed by Noble and Johnson (36). Both copies of the gene ORFs were removed entirely from the auxotrophic strain SN152 (his1Δ/his1Δ, arg4Δ/arg4Δ, and leu2Δ/leu2Δ) and replaced by homologous recombination with selective markers, LEU2 and ARG4. A yhb1Δ/yhb1Δ strain was also created with the same markers for comparison. As a control, we created strain SLA1.1 from the SN152 reference strain by inserting the LEU2 and ARG4 marker cassettes at the site of the corresponding deleted native genes (i.e., the ARG4 cassette went into the former location of one copy of C. albicans ARG4).
We then investigated whether the homozygous deletion mutants exhibit sensitivity to NO treatment. Previous studies demonstrated that proliferation ability of a yhb1Δ/yhb1Δ mutant decreased as concentrations of various NO donors increased (19, 50). Removing a transcription factor necessary for inducible YHB1 expression is predicted to also cause hypersensitivity to NO, provided that the effect is not masked by functional redundancy. The deletion mutants were tested for their ability to grow in the presence of the NO donor DPTA NONOate. The growth rates of all five deletion mutants are similar to SN152 when grown in medium without DPTA NONOate. The five homozygous mutants were subjected to various concentrations of DPTA NONOate for 3.5 h, and cell density was measured by absorbance at 600 nm. For comparison, three other strains were tested simultaneously with the mutants: the parent strain SN152, matched wild-type SLA1.1, and yhb1Δ/yhb1Δ. Of all the putative transcription factor gene mutants, only the cta4Δ/cta4Δ mutant showed hypersensitivity to DPTA NONOate (Fig. (Fig.3A).3A). The cta4Δ/cta4Δ phenotype is not as strong as that of yhb1Δ/yhb1Δ until the DPTA NONOate concentration reaches about 0.5 mM. We tested whether reinsertion of a wild-type copy of CTA4 or YHB1 would reverse the NO hypersensitivity phenotype of the cta4Δ/cta4Δ or yhb1Δ/yhb1Δ mutants, respectively. One copy of each respective gene was inserted into plasmid CIp20 (13), which was then transformed into the genome at the RPS1 locus. The resistances of the complementation strains to DPTA NONOate appear similar to the wild-type level at the concentrations tested (Fig. (Fig.3B3B).
The role of SSU1 in NO resistance in C. albicans was not known. In S. cerevisiae, yhb1Δ mutants are sensitive to NO in rich, undefined (YPD) medium, but not in synthetic dextrose medium. Conversely, ssu1Δ strains are resistant in rich medium but sensitive in synthetic medium (42). To test whether SSU1 participates in NO detoxification in C. albicans, we constructed ssu1Δ and yhb1Δ ssu1Δ mutants. Surprisingly, deletion of SSU1 had no effect on NO sensitivity under growth conditions in either rich (Fig. (Fig.3C)3C) or synthetic (Fig. (Fig.3D)3D) medium, with growth rates essentially identical to the wild-type or yhb1Δ parent strains. The ssu1Δ mutant is sensitive to exogenous sulfite (see Fig. S8 in the supplemental material), a phenotype shared with the S. cerevisiae ssu1Δ strain. Sulfite resistance is not regulated by CTA4: the cta4Δ mutant strain behaves identically to wild type in this assay (data not shown). Thus, though we have confirmed that SSU1 is an NO-inducible gene and is regulated by CTA4, it does not have an obvious role in NO detoxification.
We tested whether loss of any of the putative transcription factors altered YHB1 mRNA induction by NO. Only cta4Δ/cta4Δ appears to have lost the ability to induce the YHB1 gene when nitrite or DPTA NONOate is present (Fig. 4A and B). The mutants for the other transcription factors have a nitrosative stress response similar to wild type. Reintroducing CTA4 or YHB1 to the appropriate deletion strain restored the induced increase in YHB1 mRNA (Fig. (Fig.4C).4C). Transcriptional activity of YHB1 was reproducibly weaker for both the CTA4 and YHB1 complementation strains relative to SN152, possibly because of having only one gene copy present or because of positional effects from inserting the genes at the nonnative locus within the RPS1 ORF.
The putative sulfite transporter SSU1 is induced by NO in both S. cerevisiae and C. albicans. To determine how broad the role of CTA4 was in the NO response, we examined expression of SSU1 in wild-type and cta4Δ/cta4Δ cells by reverse transcription-PCR. SSU1 was induced in the presence of 2,2′-(hydroxynitrosohydrazono)bis-ethanimine (DETA)-NONOate in the wild-type strain, but this induction was either abolished or reduced, depending on the strain background, in the absence of CTA4 (see Fig. S9 in the supplemental material). Thus, this gene is also under CTA4 control.
Zinc cluster transcription factors typically bind directly to promoter DNA sequences of the genes they regulate. Also, the CCGTCGG palindrome within the NORE of YHB1 is clearly similar to the three nucleotide repeats observed for binding sites of other zinc cluster proteins (47). To examine whether Cta4p binds to YHB1 in vivo, we first constructed a strain expressing a functional, Myc9 epitope-tagged Cta4p from the native CTA4 promoter. Binding of Cta4p to the NORE region of the YHB1 promoter in cells was then assessed using ChIP, as shown in Fig. Fig.5.5. To determine whether nitric oxide induces a recruitment of Cta4p to the YHB1 NORE region, we compared Cta4p binding to this DNA by ChIP assay in control cells versus cells treated with 10 mM sodium nitrite for 15 min. Exposure of C. albicans to nitrite as a source of nitric oxide promotes strong induction of YHB1 mRNA after 15 min (50). Cta4p was bound to YHB1 to approximately the same extent in control cells and nitrite-treated cells (Fig. (Fig.5),5), indicating that Cta4p is already present on this promoter in the absence of an inducing signal. This mode of regulation is found among several members of the zinc cluster transcription factor family, most notably S. cerevisiae Gal4p (48).
We tested the cta4Δ/cta4Δ and yhb1Δ/yhb1Δ deletion and complementation strains for virulence using a mouse tail vein injection model of systemic candidiasis. Survival rate of the mice injected with cta4Δ/cta4Δ mutant C. albicans was significantly higher (P = 0.015) than that of the mice injected with wild-type SC5314. The survival curves show that mice infected with the cta4Δ/cta4Δ mutant also lasted longer on average than those infected with the CTA4 complementation strain (P = 0.027; mean time to death: with SC5314, 4.7 days; with the cta4Δ/cta4Δ strain, 6.7 days; with the complemented strain, 5.0 days) (Fig. (Fig.6A).6A). Previous reports (19, 50) have shown yhb1Δ/yhb1Δ mutant strains to be mildly attenuated in virulence as well (mean time to death, 7.1 days); we tested a yhb1Δ/yhb1Δ strain in the SN152 background and found a reduction in virulence (Fig. (Fig.6B),6B), similar in magnitude to the cta4Δ/cta4Δ strain (despite the similar survival curves, the result with the yhb1Δ/yhb1Δ strain is not statistically significant as we used fewer mice to confirm this previously published result). There was no significant difference between results with yhb1Δ/yhb1Δ and cta4Δ/cta4Δ (P = 0.60).
Identifying a NORE involved in nitrosative stress response establishes a foundation for determining a mechanistic pathway of NO sensing in C. albicans. The 12-bp sequence TATTACCGTCGG, located about −230 bp from the YHB1 translation start site, is presently the only known cis-acting element involved in NO regulation in C. albicans. Mass spectrometry analysis identified five putative transcription factors that may have had specific affinity for the NORE sequence in vitro.
Several findings show that one of these five factors, CTA4, is required for NO signaling in C. albicans. When the five putative transcription factor genes were separately deleted, only the cta4Δ/cta4Δ homozygous deletion mutant manifested a decline in resistance to nitrosative stress. Exposing cta4Δ/cta4Δ cells to various NO sources decreased the growth rate and drastically reduced induction of YHB1 gene expression. ChIP results also indicated that the Cta4p protein localizes at the YHB1 regulatory region in vivo.
Data on CTA4 outside of this work are sparse. CTA4 was first isolated in a one-hybrid assay in S. cerevisiae during a screen for a C. albicans trans-activating regulator of pyruvate decarboxylase gene expression (21). Later, various microarray studies reported that transcription of CTA4 is upregulated by exposure to nitrosative stress (19) or growth in whole human blood (14) and downregulated upon adherence to polystyrene (30). The closest homolog for Cta4p in S. cerevisiae is Oaf1p, an oleate receptor which heterodimerizes with Pip2p to activate peroxisomal β-oxidation genes for breaking down fatty acids (4), but the C. albicans cta4Δ mutant is able to utilize oleic acid as the sole carbon source, indicating that this gene does not regulate β-oxidation (M. Ramirez and M. Lorenz, unpublished observations). The next closest homolog among S. cerevisiae proteins is Hap1p, a heme-responsive factor which binds DNA as a homodimer and is involved in oxygen regulation (52).
Cta4p belongs in the Zn(II)2-Cys6 transcription factor family, a group of proteins unique to fungi, whose members bind DNA by means of a binuclear cluster of six cysteine residues that coordinate two zinc atoms. The C. albicans genome is predicted to have genes for 77 of these factors (7). The prototype for this class of proteins is the well-studied transcription factor Gal4p in S. cerevisiae. The protein structure typically contains three domains: a DNA-binding domain almost always located at the N terminus, a central region believed to regulate the transcriptional activation activity of the factor, and an acidic activation domain at the C terminus (29). Zn(II)2-Cys6 transcription factors function as homodimers, heterodimers, or sometimes monomers. Their cognate DNA binding sites often have CGG pairs that may be oriented as direct, inverted, or everted repeats. The spacing between these repeats varies widely between different proteins and is often a critical factor in recognition by a specific transcription factor (47). The NORE sequence TATTACCGTCGG fits this archetype in that it contains an everted CGG repeat, but it also has the additional palindromic sequence TATTA. It may be comparable to Hap1p binding sites in S. cerevisiae, which have similar, conserved TA-rich sequences that interact with the N terminus of Hap1p between the CGG direct repeats (23).
Along with Cta4p, four other putative transcription factors from this zinc binuclear cluster family demonstrated in vitro affinity to NORE-coated magnetic particles: Zcf29p, Zcf36p, War1p, and Stb5p. Presently, we have no further evidence that these four transcription factors are biologically relevant to NO regulation. Their binding to the NORE could merely be an artifact of the artificial binding conditions, or they may have redundant functions in NO regulation that are not apparent when only one gene is eliminated at a time. The functions of the ZCF29 and ZCF36 genes, which are also the closest homologs of CTA4 in the C. albicans genome, are currently unknown. C. albicans WAR1 is known to provide resistance to the weak acid sorbate, a phenotype similar to its homolog in S. cerevisiae. Although the function of STB5 in C. albicans has not been investigated, its homolog in haploid S. cerevisiae (32.5% amino acid sequence identity) has a variety of functions, including roles in multidrug resistance (1), oxidative stress response (26), and low-temperature response (2).
The process of upregulation of the flavohemoglobin gene YHB1 during nitrosative stress in C. albicans appears to differ from that in S. cerevisiae. Liu et al., who first established the NO resistance function of YHB1 in S. cerevisiae, did not detect any increase in YHB1 enzymatic activity during nitrosative stress (27). Additionally, we previously reported constitutively activated YHB1 expression in S. cerevisiae, regardless of the presence of NO-generating compounds (50). However, Sarver and DeRisi documented induction of S. cerevisiae YHB1 transcription in a different strain specifically during nitrosative stress by transcription factor Fzf1p, a C2H2 zinc finger factor (42). The closest homolog of S. cerevisiae FZF1 in C. albicans has 19.3% amino acid sequence identity, and it is not known whether it can carry out similar functions. We have found that the analogous NO response in C. albicans requires a different factor, Cta4p.
Another difference between C. albicans and S. cerevisiae is the role of the sulfite transporter SSU1. This gene mediates NO resistance under specific environmental conditions in S. cerevisiae, notably in synthetic medium. We hypothesized that the C. albicans SSU1 may have a similar function and that the yhb1Δ ssu1Δ double mutant would have a more severe phenotype, both in the presence of NO in vitro and in vivo. However, SSU1 has no role in NO detoxification in C. albicans under the conditions tested here, though its induction by NO is at least partly CTA4 dependent. Why SSU1 is part of the NO regulon is not immediately apparent.
We also demonstrated that deletion of the CTA4 gene produces a statistically significant reduction in virulence in a mouse infection model, though the effect was small. Deleting the YHB1 gene produced a similar mild attenuation of virulence, but this was not statistically significant, possibly because of the smaller number of mice tested. These data, together with earlier reports (19, 50), indicate that increasing susceptibility to NO has a small effect on virulence in the mouse tail vein model of candidiasis.
The mildness of the virulence reduction could indicate that YHB1 and CTA4 aid C. albicans survival against other NO sources that are not tested by this experiment. Injecting C. albicans directly into the bloodstream may be an appropriate model for certain situations, such as infection by contaminated implants, catheters, or needles, but this delivery method circumvents several common infection routes. Since C. albicans most often resides as a commensal organism on the mucosal linings of the urogenital and digestive tracts, it usually must infiltrate these barriers first before it can reach the bloodstream and disseminate throughout the body.
The NO in these environments could affect how vital YHB1 and CTA4 are to the survival and spread of this pathogen inside a host organism. For example, gastric NO concentrations in the human small intestine following ingestion of dietary nitrates can rise to levels thousands of times higher than in tissues (51). Mouse models in which C. albicans is introduced orally may be more informative in testing this. Additionally, the major function of the CTA4 and YHB1 genes could be more subtle than direct opposition against the host immune system; they may instead be more relevant in providing C. albicans a survival advantage against other microorganisms vying for the same territory. Using more than one infection model will be important in establishing when and where these genes are critical for NO defense.
Our work takes the initial steps to decipher a pathway for NO signaling in C. albicans. Since this species is a major cause of fungal infections in humans, learning how it evades normal immune defenses and exploits weakened ones is essential for developing new strategies for treating infection.
We thank Alexander Johnson for supplying plasmids and strains, Aaron Mitchell for providing the cta4::UAU mutant, and Melissa Ramirez for her assistance with the virulence assays in mice.
This work was supported by the National Science Foundation under grants DGE-0114264, MCB-0091236, and MCB-0520873.
Published ahead of print on 14 December 2007.
†Supplemental material for this article may be found at http://ec.asm.org/.