ABSTRACT
Candida albicans is an important cause of life-threatening systemic bloodstream infections in immunocompromised patients. In order to cause infections, C. albicans must be able to synthesize the essential metabolite inositol or acquire it from the host. Based on the similarity of C. albicans to Saccharomyces cerevisiae, it was predicted that C. albicans may generate inositol de novo, import it from the environment, or both. The C. albicans inositol synthesis gene INO1 (orf19.7585) and inositol transporter gene ITR1 (orf19.3526) were each disrupted. The ino1Δ/ino1Δ mutant was an inositol auxotroph, and the itr1Δ/itr1Δ mutant was unable to import inositol from the medium. Each of these mutants was fully virulent in a mouse model of systemic infection. It was not possible to generate an ino1Δ/ino1Δ itr1Δ/itr1Δ double mutant, suggesting that in the absence of these two genes, C. albicans could not acquire inositol and was nonviable. A conditional double mutant was created by replacing the remaining wild-type allele of ITR1 in an ino1Δ/ino1Δ itr1Δ/ITR1 strain with a conditionally expressed allele of ITR1 driven by the repressible MET3 promoter. The resulting ino1Δ/ino1Δ itr1Δ/PMET3::ITR1 strain was found to be nonviable in medium containing methionine and cysteine (which represses the PMET3 promoter), and it was avirulent in the mouse model of systemic candidiasis. These results suggest a model in which C. albicans has two equally effective mechanisms for obtaining inositol while in the host. It can either generate inositol de novo through Ino1p, or it can import it from the host through Itr1p.
Candida albicans is a dimorphic yeast that can exist in a human host either as a harmless commensal or as an opportunistic pathogen when the host's immune system is impaired (5). In patients that are neutropenic, Candida species are associated with severe and deadly systemic bloodstream infections with a high mortality rate (31.8%) (27). In fact, Candida species are the fourth most common cause of catheter-related bloodstream infections in hospitalized patients in intensive care units, and C. albicans is the species most commonly isolated from patients with these infections (4).
The ability of any pathogenic microbe to cause an infection depends on the organism's ability to acquire or generate essential nutrients during residence within the host. There are several examples of auxotrophies that compromise the virulence of C. albicans (16). The most notable example is the defect in uracil biosynthesis caused by a mutation in URA3. Homozygous URA3 mutations (ura3Δ/ura3Δ) in C. albicans block the ability to grow without uridine supplementation and lead to avirulence. In fact, the level of URA3-encoded orotidine 5′-monophosphate decarboxylase activity can be correlated with growth and virulence (20). Homozygous mutations in the ADE2 gene that block the ability of C. albicans to grow in minimal medium (MM) or serum in the absence of exogenous adenine also diminish virulence (10). In addition, homozygous mutations in the HEM3 gene leading to heme auxotrophy compromise virulence (16).
In contrast to these results, auxotrophies for a number of amino acids, including serine, lysine, leucine, histidine, and arginine, do not appear to compromise virulence (23, 32, 40). For example, Noble and Johnson (32) demonstrated that C. albicans his1Δ/his1Δ, his1Δ/his1Δ leu2Δ/leu2Δ, and his1Δ/his1Δ arg4Δ/arg4Δ auxotrophic mutants exhibit wild-type virulence in a mouse model of systemic candidiasis, and his1Δ/his1Δ leu2Δ/leu2Δ arg4Δ/arg4Δ triply auxotrophic strains are only mildly attenuated in virulence compared to the wild type. These results suggest that there are some metabolites, such as amino acids, that C. albicans can scavenge from the host during an infection and that there are other metabolites, such as uracil, adenine, and heme, that it must make de novo.
myo-inositol (referred to hereinafter as inositol) is essential for the growth of all eukaryotes and is needed by some bacteria (12). Inositol is involved in many intracellular processes, including growth regulation, membrane structure formation, osmotolerance, signal transduction, and the formation of glycosylphosphatidylinositol (GPI)-anchored proteins, which are themselves essential (3, 7, 33, 41). Inositol has been implicated in the pathogenicity of C. albicans because it is an essential precursor of phospholipomannan, a GPI-anchored glycolipid on the surfaces of Candida cells that is involved in pathogenicity (26).
Three potential sources of inositol available to C. albicans can be suggested based on work that has been done with the related yeast Saccharomyces cerevisiae: (i) de novo biosynthesis, in which glucose 6-phosphate is converted into inositol 1-phosphate by the inositol 1-phosphate synthase enzyme; (ii) inositol import from the extracellular environment by an inositol transporter; and (iii) the recycling of inositol from the dephosphorylation of inositol polyphosphate products (7, 30, 41).
In S. cerevisiae, de novo inositol biosynthesis and import have both been well studied. The S. cerevisiae inositol 1-phosphate synthase is encoded by the ScINO1 gene (9). The Ino1p enzyme converts glucose 6-phosphate into inositol 1-phosphate, and then a second enzyme encoded by the INM1 gene, inositol monophosphatase, dephosphorylates inositol 1-phosphate to create inositol (22). S. cerevisiae carries two distinct inositol transporter genes, S. cerevisiae ITR1 (ScITR1) and ScITR2, the products of which are both capable of importing inositol but at different efficiencies (31).
C. albicans carries a gene determined on the basis of sequence similarity to be a putative homolog of INO1; this gene (GenBank accession no. L22737) was identified and sequenced by Klig et al. (17, 18). C. albicans INO1 (CaINO1) shows 64% identity to ScINO1 at the amino acid level (18). In addition, it has been shown previously that C. albicans has potent inositol transporter activity (15). It was demonstrated that the inositol transporter in C. albicans has an apparent Km of 240 ± 15 μM, and the transport system appears to be active and energy dependent (15). C. albicans inositol transport activity differs in substrate and cotransporter specificities from the human inositol-Na+ transporter. However, the inositol transporter gene in C. albicans was not identified in this previous study.
Recently, the importance of inositol biosynthesis in some human infectious agents was demonstrated by studies of two unrelated pathogens, Mycobacterium tuberculosis and Trypanosoma brucei. In T. brucei, the INO1 homolog was shown to be necessary for inositol biosynthesis, efficient GPI synthesis, and growth in vitro (24, 25). It was found that the M. tuberculosis INO1 homolog is required for de novo inositol biosynthesis and full virulence in a mouse model of infection (29).
It is unknown what roles inositol import or biosynthesis play in the virulence of C. albicans. This fungus may be able both to synthesize and to import inositol. It became of interest to determine if C. albicans would behave like M. tuberculosis and T. brucei and require INO1 function for virulence and/or viability or if, in contrast to these pathogens, C. albicans would be able to utilize either de novo synthesis or inositol import to support virulence thereafter, the C. albicans INO1 and ITR1 homologs are referred to as INO1 and ITR1, respectively.
MATERIALS AND METHODS
Strains and growth media. C. albicans strains used in this study are shown in Table 1. The media used in this study include 1% yeast extract-2% peptone-2% glucose (YPD) (42), defined medium 199 (Invitrogen), yeast carbon base-bovine serum albumin (39), and MM (0.67% Difco yeast nitrogen base without amino acids, 2% glucose) (42). MM was supplemented with 75 μM inositol for the growth of the C. albicans ino1Δ/ino1Δ mutant and 2.5 mM (each) methionine and cysteine for MET3 promoter shutoff assays (6). The ino1Δ/ino1Δ itr1Δ PMET3::ITR2-NAT1 strain was selected on MM medium containing no methionine and cysteine, 75 μM inositol, and 1 mg/ml nourseothricin. YPD containing 250 μg/ml nourseothricin was used to select for other transformants. Inositol-free medium (42) was made for testing the phenotype of the ino1Δ/ino1Δ mutant. Agar plates were solidified with 2% agar (granulated; Fisher) for YPD and with 2% Bacto agar (which contains no residual inositol) for MM and inositol-free medium.
C. albicans strains
Strain construction.The INO1 gene was disrupted by using the CaNAT1-FLP cassette (39), whereas the ITR1 gene was disrupted by using the SAT1 flipper (35).
For the INO1 disruption construct, the 564-bp 5′ noncoding region (NCR) of INO1 (5′ INO1NCR) was amplified with primers JCO39 and JCO40 (Table 2), which introduced KpnI and ApaI restriction sites, and was cloned into pJK863 in the 5′ direction from the CaNAT1-FLP cassette (see Fig. 1A). The 583-bp 3′ INO1NCR was amplified with primers JCO41 and JCO42, which introduced SacII and SacI sites, and was cloned into pJK863 in the 3′ direction from the CaNAT1-FLP cassette (see Fig. 1A). This procedure created the INO1 knockout construct plasmid pYLC94 (Table 3; see Fig. 1A), which was cut with KpnI and SacI to release the disruption construct, and the wild-type SC5314 strain was transformed with the disruption construct by electroporation as described previously (8, 35). The disruption construct was used to sequentially disrupt both alleles of INO1 as previously described (39). The INO1 reconstitution construct was made by amplifying a 2.1-kb fragment containing the INO1 open reading frame (ORF) and the 5′ NCR from SC5314 genomic DNA by using primers (JCO39 and JCO47) that introduced KpnI and SalI sites. This fragment was ligated into the pRS316 vector, along with a 1.7-kb fragment containing the NAT1-3′ INO1NCR fragment amplified from plasmid pYLC94 by using primers JCO50 and JCO42, which introduced SalI and SacI sites. This procedure resulted in the INO1 reconstitution plasmid pYLC119 (Table 3; see Fig. 1B). The ino1Δ/ino1Δ mutant YLC113 was transformed with the 3.8-kb KpnI-SacI fragment from pYLC119 in order to create the reconstituted-INO1 (ino1Δ/ino1Δ::INO1) strain YLC120.
The INO1 gene is required for growth in the absence of exogenous inositol. (A) Structure of the INO1 disruption construct. Noncoding DNA sequences of approximately 500 bp each flanking the 5′ and 3′ ends of the INO1 gene (5′ and 3′ INO1NCR, respectively) were cloned onto either flank of the CaNAT1-FLP construct (39). The thick dark arrows represent the FRT sites of the FLP recombinase. The ball-and-stick symbol represents the ACT1 terminator (ACT1t), and the thinner arrow on a raised line represents the SAP2 promoter (PSAP2). (B) INO1-NAT1 construct used to reintegrate INO1 into the ino1Δ/ino1Δ mutant. (C) Southern blotting was used to confirm the INO1 disruptions. Lanes: 1, wild type; 2, ino1Δ::NAT1-FLP/INO1 strain; 3, ino1Δ/INO1 strain; 4, ino1Δ/ino1Δ::NAT1-FLP strain; 5, ino1Δ/ino1Δ strain; and 6, ino1Δ/ino1Δ::INO1 strain. (D) The cells were streaked onto medium containing either 0 or 75 μM inositol and grown for 2 days at 30°C.
PCR primers
Plasmids used in this study
A similar approach was used to knock out ITR1 with the SAT1 flipper plasmid pSFS2A (35). Approximately-500-bp 5′ and 3′ NCRs from ITR1 (5′ ITR1NCR and 3′ ITR1NCR) were amplified using the primer pairs JCO89-JCO90 and JCO93-JCO94, respectively, each of which introduced restriction sites into the corresponding fragments. The ApaI-XhoI 5′ ITR1NCR and NotI 3′ ITR1NCR fragments were ligated into pSFS2A, resulting in the ITR1 knockout construct plasmid pYLC164 (Table 3; see Fig. 2A). The 5-kb ApaI-SacII fragment from pYLC164 was introduced into SC5314 by electroporation and used to disrupt both copies of ITR1 by sequential disruption steps (28, 35). For the reconstituted-ITR1 construct, JCO89 and JCO111 primers, which added ApaI and EcoRI sites, were used to amplify a 2-kb fragment containing the 5′ ITR1NCR and the ITR1 ORF region, which was used to replace the ApaI-EcoRI fragment (containing the 5′ ITR1NCR-PMAL2-FLP fragment) in the pYLC164 plasmid, resulting in pYLC208 (Table 3; see Fig. 2B). The 5-kb ApaI-SacII fragment from pYLC208 was used to reconstitute ITR1 in the itr1Δ/itr1Δ mutant strain YLC196, creating the itr1Δ/itr1Δ::ITR1 strain YLC211.
The ITR1 gene in C. albicans was disrupted. (A) Structure of the ITR1 disruption construct. Noncoding DNA sequences of approximately 500 bp each flanking the 5′ and 3′ ends of the ITR1 gene (5′ and 3′ ITR1NCR, respectively) were cloned into the pSFS2A plasmid such that they flanked the SAT1 flipper cassette. The thick dark arrows represent the FRT sites of the FLP recombinase. The ball-and-stick symbol represents the ACT1 terminator (ACT1t), and the thinner arrow on a raised line represents the MAL2 promoter (PMAL2). (B) ITR1-SAT1 construct used to reintegrate ITR1 into the itr1Δ/itr1Δ mutant. (C) PMET3::ITR1 construct used to replace the ITR1 allele in the ino1Δ/ino1Δ itr1Δ/ITR1 strain to generate the ITR1 conditional allele. (D) Southern blotting was used to confirm the ITR1 disruptions. Lanes: 1, wild type; 2, itr1Δ::SAT1-FLP/ITR1 strain; 3, itr1Δ/ITR1 strain; 4, itr1Δ/itr1Δ::SAT1-FLP strain; 5, itr1Δ/itr1Δ strain; 6, itr1Δ/itr1Δ::ITR1-SAT1 strain; and 7, ino1Δ/ino1Δ itr1Δ/PMET3::ITR1 strain. The bands corresponding to itr1Δ::SAT1-FLP (the allele disrupted prior to the “flipping out” of the SAT1-FLP construct) and ITR1-SAT1 (reintegrated ITR1 marked with SAT1) ran together on the gel. (E) Northern blotting revealed that the expression of ITR1 was lost in the itr1Δ/itr1Δ mutant, but not the wild-type (WT), heterozygous mutant, and reconstituted-ITR1 (itr1Δ/itr1Δ::ITR1) strains. Strains were grown in defined medium 199 for 2 h at 37°C, and RNA was isolated, subjected to Northern blotting, and probed for ITR1 expression. ITR1 expression was normalized to CaACT1 expression on the same blot.
The ITR1 conditional mutant was made as follows. With SC5314 genomic DNA as the template, primers JCO118 and JCO119 (which introduced PstI and NotI sites) were used to amplify a 1,362-bp fragment containing the CaMET3 promoter (6), and primers JCO120 and JCO121 (which introduced NotI and SacII sites) were used to amplify the ITR1 ORF. These fragments were used to replace the PstI-SacII fragment in pYLC164, which deleted the 3′-end-flanking FLP recombination target (FRT) site, resulting in the conditional ITR1 expression construct plasmid pYLC229 (Table 3; see Fig. 2C). The ino1Δ/ino1Δ itr1Δ/ITR1 strain YLC184 was transformed with the 7.6-kb ApaI-SacII fragment from pYLC229 to create the ino1Δ/ino1Δ itr1Δ/PMET3::ITR1 conditional strains YLC261 and YLC266.
Northern blot analysis.Northern blotting for ITR1 expression was performed as described previously (36), with the following exceptions. Strains grown in liquid medium 199 at 37°C for 2 h were collected for total RNA extraction by the hot-phenol method (36). A PCR product containing bp 24 to 750 of the ITR1 ORF (amplified with primers JCO102 and JCO103) was used as a probe. Expression was normalized against the expression of the CaACT1 gene, probed for on the same membrane. The CaACT1 probe was generated with the primers JCO48 and JCO49.
Southern blot analysis.Hybridization conditions for the Southern blot analysis were similar to those for the Northern blot analysis, except that the Techne Hybrigene oven was set to 60°C for the incubation step and 42 and 60°C for the washing steps. The cells were grown in liquid YPD at 30°C overnight. The genomic DNA was extracted using the Winston-Hoffman method (13), and 20 μg of genomic DNA was subjected to Southern blotting. The genomic DNA of ino1 mutants was cut by AflII and SphI restriction enzymes, while the genomic DNA of itr1 mutants was cut by PstI. PCR products containing the ∼500-bp 3′ INO1NCR (amplified with primers JCO41 and JCO42) and the 3′ ITR1NCR (amplified with primers JCO93 and JCO94) were used as probes.
Inositol uptake assays.The inositol uptake assay protocol was adapted in part from a protocol of Jin and Seyfang (15). The wild-type, itr1Δ/ITR1, itr1Δ/itr1Δ, and itr1Δ/itr1Δ::ITR1 strains were grown in YPD liquid cultures overnight at 30°C. Cells were diluted in YPD to an optical density at 600 nm of 1, grown at 30°C, and collected at an optical density at 600 nm of 5 by centrifugation at 2,600 × g for 5 min. Cells were then washed twice with water at 4°C and resuspended in 2% glucose to a final concentration of 2 × 108 cells/ml as determined by a hemacytometer. From this time point on, cells were kept on ice until being used for the actual assay. For the uptake assay, the reaction mixture (250 μl) contained 2% glucose, 40 mM citric acid-KH2PO4 (pH 5.5), 0.15 μM [2-3H]inositol (1 μCi/μl; MP Biomedicals), and 200 μM unlabeled inositol (Alexis Biomedicals). Equal volumes of the reaction and cell mixtures (60 μl each) were warmed to 30°C and mixed for the uptake assay, which was performed for 10 min at 30°C. As negative controls, mixtures were kept at 0°C (on ice) during the 10-min incubation. Aliquots of 100 μl were removed and transferred onto prewetted Metricel filters on a vacuum manifold. The filters were washed four times each with 1 ml of ice-cold water. The washed filters were removed and added to liquid scintillation vials for measurements on a PerkinElmer TRI-CARB 2900TR scintillation counter. The uptake of radiolabeled inositol over 10 min was calculated and plotted as a function of the incubation temperature.
Mouse infection studies.Five- to 6-week-old male CD1 mice (18 to 20 g) from Charles River Laboratories were used in this study. Mice were housed in groups of five per cage. For infection, colonies from each C. albicans strain were inoculated into 20 ml of YPD or MM. Cultures were grown overnight and washed twice with 25 ml of sterile water, and cells were counted by using a hemacytometer and resuspended in sterile water at 107 cells per ml. The cells were then plated onto YPD to determine the viability. Mice were injected via the tail vein with 0.1 ml of the cell suspension (106 cells) (43), and the course of infection was monitored for up to 30 days. Survival was monitored twice daily, and moribund mice were euthanized. All experimental procedures were carried out according to the NIH guidelines for the ethical treatment of animals.
Statistics.The statistical analysis was done using Prism 4.0 software (GraphPad Software). For the mouse model of systemic infection, Kaplan-Meier survival curves were compared for significance by using the Mantel-Haenszel log rank test. The significance of differences in inositol uptake between strains was determined using the two-tailed unpaired t test. Statistical significance was set at a P value of <0.05.
RESULTS
The INO1 gene in C. albicans is required for growth in the absence of exogenous inositol.The C. albicans homolog of ScINO1 identified by Klig et al. (18) was disrupted in order to determine if it was required for inositol biosynthesis in C. albicans. The C. albicans INO1 homolog (orf19.7585) was disrupted by sequentially replacing both alleles of the gene with the NAT1-FLP cassette, which contains the nourseothricin resistance gene (39). The INO1 disruption construct is diagrammed in Fig. 1A. The INO1 gene was reintegrated into the INO1 locus of the homozygous mutant (ino1Δ/ino1Δ), by using the construct depicted in Fig. 1B, to verify the linkage of any resulting phenotypes with the genotype. Wild-type (INO1/INO1), heterozygous mutant (ino1Δ/INO1), homozygous mutant (ino1Δ/Δ), and reconstituted-INO1 (ino1Δ/ino1Δ::INO1) strains were analyzed by PCR (data not shown) and Southern blotting (Fig. 1C) to confirm the deletion and reintegration of the correct genes. The growth patterns of the wild-type, ino1Δ/INO1, ino1Δ/ino1Δ, and ino1Δ/ino1Δ::INO1 strains on medium lacking inositol (42) were then compared to determine if an ino1Δ/ino1Δ mutation would compromise the ability of the strain to grow in the absence of inositol. The ino1Δ/ino1Δ strain was unable to grow on inositol-free medium, while the wild-type, heterozygous, and reconstituted-INO1 strains grew equally well. As expected, the ino1Δ/ino1Δ strain was able to grow as well as the wild type on identical solid and liquid media with inositol added (Fig. 1D and data not shown, respectively).
The INO1 gene is not required for virulence in C. albicans.The INO1 gene does not appear to be required for virulence in a mouse model of disseminated candidiasis. The wild-type, ino1Δ/INO1, and ino1Δ/ino1Δ::INO1 strains and two separately derived ino1Δ/ino1Δ strains were tested for virulence in the mouse disseminated-infection model (43). These tests revealed no difference in virulence between the wild type and the ino1 mutant strains (see Fig. 5A). The ino1Δ/INO1 heterozygote behaved like the wild type, and results for this strain are not shown.
The ITR1 gene is required for inositol transport in C. albicans.The ability of the ino1Δ/ino1Δ mutant to grow on medium containing inositol supports results from previous studies that have shown that C. albicans harbors an inositol transporter (15). A C. albicans homolog of the ScITR1 and ScITR2 inositol transporter genes was identified by a BLAST analysis of the ScItr1p and ScItr2p protein sequences against the Candida Genome Database (CGD; http://www.candidagenome.org ). This homolog is orf19.3526 and is currently referred to as HGT15 in the CGD, but its alias in CGD is ITR2. This gene has never been characterized functionally. Based on our BLAST search of the CGD, C. albicans orf19.3526 is the closest homolog of ScITR1 and ScITR2, and based on the corresponding amino acid sequences, it is 51% identical to both S. cerevisiae transporter genes over its full length (data not shown). The protein product of C. albicans orf19.3526 also contains the critical inositol transporter motif (D/E)(R/K)φGR(R/K) (38). Based on the results described below, we will refer to orf19.3526 as ITR1 hereinafter.
The two alleles of the ITR1 gene were sequentially disrupted by replacing each ORF with the nourseothricin resistance cassette from the SAT1 flipper (35). The itr1Δ disruption construct is diagrammed in Fig. 2A. The ITR1 gene was reconstituted in the itr1Δ/itr1Δ strain by using the reconstitution construct depicted in Fig. 2B. Gene deletions and replacements were checked for accuracy by PCR (data not shown) and Southern blotting (Fig. 2D). As a further test, Northern blotting revealed that the ITR1 transcript could be detected only in wild-type (ITR1/ITR1), heterozygous mutant (itr1Δ/ITR1), and reconstituted-ITR1 (itr1Δ/itr1Δ::ITR1) strains but not in the itr1Δ/itr1Δ strain (Fig. 2E). The expression levels in the itr1Δ/ITR1 and itr1Δ/itr1Δ::ITR1 strains were lower than that in the wild type, presumably because these strains contained only one allele of the gene.
Analyses of inositol uptake by the wild-type, itr1Δ/ITR1, itr1Δ/itr1Δ, and itr1Δ/itr1Δ::ITR1 strains revealed that ITR1 is required for inositol uptake (Fig. 3). The level of inositol uptake at 0°C was very low (15) (Fig. 3); hence, this level was used as a negative control. The level of inositol uptake by the wild type at 30°C was 58-fold higher than that at 0°C. In contrast, the level of inositol uptake by the itr1Δ/itr1Δ mutant at 30°C was only fourfold higher than that at 0°C, but a two-tailed paired t test analysis revealed that there was no significant difference in uptake levels between these two temperatures (P = 0.34). The itr1Δ/ITR1 mutant showed decreased inositol import compared to the wild type, and the reconstituted-ITR1 strain had inositol import restored to a level similar to that seen in the wild type.
The ITR1 gene is required for inositol transport in C. albicans. The inositol uptake assay revealed that the ability of the itr1Δ/itr1Δ mutant to import [3H]myo-inositol was greatly reduced compared to that of the wild type (WT) and that, along with the wild type, the heterozygous mutant and reconstituted-ITR1 (itr1Δ/Δ::ITR1) strains exhibited the ability to import inositol. Each strain was assayed at 0°C (white bars) and 30°C (black bars).
The level of inositol uptake by the wild-type strain in 200 μM inositol was found to be 288 pmol per 5 × 107 cells during a 10-min uptake period. Jin and Seyfang showed a level of approximately 60 pmol per 5 × 107 cells during a single-minute uptake period (15). The difference between their data and ours is probably due to the saturation of the uptake system over a 10-min time course (15).
The ITR1 gene is not required for virulence in C. albicans.The itr1Δ/itr1Δ strain did not appear to be any less virulent than the wild type in a mouse model of systemic candidiasis. The wild-type, itr1Δ/ITR1, itr1Δ/itr1Δ, and itr1Δ/itr1Δ::ITR1 strains were compared in the mouse model of disseminated infection. There was not a significant difference between the virulence of the wild type and that of the itr1Δ/itr1Δ strains (P = 0.95) (see Fig. 5B). The similarity in virulence between wild-type and itr1Δ/itr1Δ strains was born out in two separate infection experiments. The itr1Δ/ITR1 mutant behaved like the wild type (see Fig. 5B) as well. In the group infected with the reintegrated-ITR1 strain (itr1Δ/itr1Δ::ITR1), only half of the mice succumbed to the infection. This result may be due to a technical error during the injection or an uncharacterized mutation within the strain. The fact that the wild-type, itr1Δ/ITR1, and itr1Δ/itr1Δ strains were all similarly virulent strongly indicates that a lack of ITR1 does not impair virulence.
The INO1 and ITR1 genes show synthetic defects in growth and virulence.Although neither ITR1 nor INO1 was required for virulence in a mouse model of systemic candidiasis, it was possible that each might compensate for the loss of the other during infection since it has been shown previously that rat serum contains 20 to 100 μM inositol (14, 34) and mouse liver contains approximately 100 μM inositol (2). In order to test this hypothesis, an attempt was made to disrupt both genes simultaneously. One allele of ITR1 was disrupted in the ino1Δ/ino1Δ strain. However, it was not possible to disrupt the other allele of ITR1, suggesting that a double mutation consisting of ino1Δ/ino1Δ and itr1Δ/itr1Δ was synthetically lethal. Therefore, in the ino1Δ/ino1Δ itr1Δ/ITR1 strain, the promoter of the remaining wild-type ITR1 allele was replaced with the CaMET3 conditional promoter (6) by using the construct depicted in Fig. 2C. When the CaMET3 promoter (PMET3) is used to replace the promoter of a target gene, it represses the transcription of that gene in the presence of methionine and cysteine in the medium. The correct insertion of the PMET3::ITR1 allele into the chromosome was confirmed by PCR analysis (data not shown) and Southern blotting (Fig. 2D, lane 7). The resulting ino1Δ/ino1Δ itr1Δ/PMET3::ITR1 strain was tested for growth in MM containing 75 μM inositol and either 0 or 2.5 mM (each) methionine and cysteine. It was found that unlike the wild-type and ino1Δ/ino1Δ itr1Δ/ITR1 strains, two separately derived ino1Δ/Δ itr1Δ/PMET3::ITR1 strains failed to grow in the presence of 2.5 mM (each) methionine and cysteine (Fig. 4).
The INO1 and ITR1 genes show synthetic growth defects on agar plates. The growth patterns of wild-type (WT), ino1Δ/ino1Δ itr1Δ/ITR1, and two separately derived ino1Δ/ino1Δ itr1Δ/PMET3::ITR1 strains were compared in fivefold serial dilutions on MM containing 75 μM inositol with or without 2.5 mM (each) methionine and cysteine. Numbers are cell populations determined by a hemacytometer.
The ino1Δ/ino1Δ itr1Δ/PMET3::ITR1 strain was tested in a mouse model of systemic candidiasis to determine whether its virulence was affected. Previous work had indicated that the presence of the PMET3 promoter on the essential gene CaFBA1 can compromise virulence (37). The methionine in the mouse bloodstream is presumably able to prevent the expression of the CaFBA1 gene sufficiently to compromise growth and virulence in the mouse. The wild-type and ino1Δ/ino1Δ itr1Δ/ITR1 strains and two ino1Δ/ino1Δ itr1Δ/PMET3::ITR1 strains were compared in the mouse model of systemic candidiasis. This experiment revealed that the ino1Δ/ino1Δ itr1Δ/ITR1 strain was attenuated in virulence and that the ino1Δ/ino1Δ itr1Δ/PMET3::ITR1 strains were avirulent (Fig. 5C).
The INO1 and ITR1 genes show synthetic defects in virulence. The survival of mice following intravenous challenge with 106C. albicans blastospores was monitored. (A) Mice were injected with the wild type (WT; n = 10) and the following INO1 mutant strains: the ino1Δ/ino1Δ strain YLC113 (n = 10), the ino1Δ/ino1Δ strain YLC126 (n = 11), and an ino1Δ/ino1Δ::INO1 strain (n = 10). (B) Mice were injected with the wild type (n = 5) and the following ITR1 mutant strains: an itr1Δ/ITR1 strain (n = 5), an itr1Δ/itr1Δ strain (n = 10), and an itr1Δ/itr1Δ::ITR1 strain (n = 6). (C) Mice were injected with the wild type (n = 10) and the following double mutant strains: the ino1Δ/ ino1Δ itr1Δ/ITR1 strain (n = 10), the ino1Δ/ino1Δ itr1Δ/PMET3::ITR1 strain YLC261 (n = 10), and the ino1Δ/ino1Δ itr1Δ/PMET3::ITR1 strain YLC266 (n = 10). Strains for the experiments described in the legends to panels A and B were pregrown in YPD before injection, while strains for the experiment described in the legend to panel C were pregrown in MM lacking methionine and cysteine but containing 75 μM inositol.
DISCUSSION
The mechanism by which C. albicans acquires the essential metabolite inositol during an infection has not been explored previously. We have found that C. albicans is able to generate inositol de novo via the INO1 gene product or import inositol from the environment via the ITR1 gene product with efficiencies that allow it to establish an infection regardless of which mechanism is employed. This conclusion applies only to bloodstream infections, as a bloodstream infection model was the only model tested. This result implies that the availability of inositol in mice is sufficient to support an infection even if C. albicans must acquire inositol solely by importing it. Although our search of the literature did not reveal the estimated inositol content of mouse serum, the inositol levels found in rats are 20 to 100 μM (14, 34), which may be comparable to those in mice and are similar to that found in humans (mean ± standard deviation, 61.0 ± 12.4 μM) (19).
The results found for C. albicans are in contrast to those obtained previously for two important human pathogens, M. tuberculosis and T. brucei, which require de novo inositol biosynthesis via INO1 homologs in order to be fully virulent in mouse models (M. tuberculosis) or to be viable (T. brucei) (25, 29). Both of these pathogenic microbes are capable of importing inositol, but import is not sufficient to allow for the viability of T. brucei or the virulence of M. tuberculosis. In the case of T. brucei, de novo-synthesized inositol is used to make GPI-anchored proteins, while imported inositol is used very inefficiently for this purpose (24). GPI-anchored proteins are required for the viability of T. brucei (21), so the disruption of INO1 compromises viability. In the case of M. tuberculosis, the inositol transporter is too inefficient to import inositol from the host in order to support infection (29).
Unlike either of these pathogens, C. albicans possesses an inositol transporter, encoded by ITR1, that is capable of transporting inositol efficiently enough to allow full virulence in a mouse model of systemic candidiasis even in the absence of de novo inositol biosynthesis. In the case of both M. tuberculosis and T. brucei, it has been suggested previously that the development of selective inhibitors of Ino1p homologs may be an effective way to generate antimicrobials (24, 29). The data reported here indicate that this strategy would not be effective for C. albicans, as it is fully virulent even in the absence of its Ino1 enzyme. As an alternative approach, it has been suggested previously that toxic inositol analogs that are selectively taken up by the C. albicans inositol transporter but not by the human inositol-Na2+ transporter may be effective drugs (15). This may be a possibility, although it needs to be determined which of these two mechanisms, de novo biosynthesis or import, is used by wild-type C. albicans during an infection. If import is used extensively, then this approach may work; however, since C. albicans is fully virulent in the absence of ITR1, the development of resistant mutants lacking the Itr1p transporter function may pose a problem. Nonetheless, a toxic analog may be useful in combination with drugs affecting other targets, such as azoles and polyenes (1).
Based on our results, it appears that Itr1p is the primary and perhaps sole inositol importer in C. albicans both in vitro and during infection. The disruption of ITR1 greatly inhibited inositol uptake in vitro (Fig. 3), and in the absence of INO1, a strain carrying only the PMET3::ITR1 allele of ITR1 could not grow in medium containing cysteine and methionine, even in the presence of 75 μM extracellular inositol (Fig. 4). The amounts of methionine and cysteine in mouse serum are apparently sufficient to decrease expression from the MET3 promoter as well, which is consistent with results obtained using a MET3-driven form of the CaFBA1 gene (37). Our conclusion that ITR1 encodes the primary or sole inositol transporter is consistent with the findings of a previous study of C. albicans inositol transport which concluded that there was one inositol transporter in C. albicans that was responsible for all, or at least the vast majority of, inositol transport activity (15).
The situation of C. albicans contrasts with that of S. cerevisiae, which has two inositol transporters that are expressed at widely different levels (31). In S. cerevisiae, the ScItr1p transporter is the more highly expressed of the two transporters and accounts for most of the transport activity. The residual transport activity in S. cerevisiae is carried out by ScItr2p, which is expressed at much lower levels. Based on findings from BLAST searches of the CGD using the ScITR1 or ScITR2 sequence as the query, the C. albicans ITR1 gene was the closest homolog to the S. cerevisiae inositol transporter genes, showing 51% identity over the whole sequence (based on the corresponding amino acids) to either ScITR1 or ScITR2. C. albicans carries at least one other homolog of the S. cerevisiae inositol transporter genes (orf19.5447, or HGT19) that is predicted to encode 12 transmembrane domains and the (D/E)(R/K)φGR(R/K) motif typical of inositol transporters. The predicted protein sequence corresponding to this gene bears 26 and 27% identity to ScItr1p and ScItr2p, respectively. It is possible that orf19.5447 encodes an inositol transporter, but our results suggest that if so, this other transporter is expressed at too low a level to transport inositol efficiently or it is expressed under different conditions from those tested in our experiments. Alternatively, it is a very low affinity transporter like that seen in M. tuberculosis (29).
Taken together, our results indicate that either Ino1p or Itr1p can supply the inositol requirement during an infection, but a number of questions remain to be answered. For wild-type C. albicans, it is not known whether de novo inositol biosynthesis or import is utilized during an infection or whether a combination of both is in operation. The answer may be variable, depending on the distribution of C. albicans cells within the host. In addition, it is not known whether de novo synthesis or import is favored during growth in host niches other than the bloodstream, such as the gut, the oral mucosa, and the vaginal tract. This answer again may be variable, depending on the nutrient conditions of the particular host niches and the locations of cells within those host sites. Further studies will be required to answer these questions.
ACKNOWLEDGMENTS
We gratefully acknowledge Jeffrey Becker, Michael Lorenz, and Pamela Small for their critical review of this work. We thank Julia Köhler and Joachim Morschhäuser for providing the CaNAT1-FLP cassette and the SAT1 flipper, respectively. We are grateful to Melinda Hauser and Li-Yin Huang for their assistance with the inositol uptake assay and animal studies, respectively. We also thank all members of the Reynolds, Kitazono, and Becker laboratories for many helpful discussions.
This work was funded in part by grant 1R03AI071863.
FOOTNOTES
- Received 14 November 2007.
- Returned for modification 19 December 2007.
- Accepted 28 January 2008.
- Copyright © 2008 American Society for Microbiology
