Azole Drug Import into the Pathogenic Fungus Aspergillus fumigatus

Abstract

The fungal pathogen Aspergillus fumigatus causes serious illness and often death when it invades tissues, especially in immunocompromised individuals. The azole class of drugs is the most commonly prescribed treatment for many fungal infections and acts on the ergosterol biosynthesis pathway. One common mechanism of acquired azole drug resistance in fungi is the prevention of drug accumulation to toxic levels in the cell. While drug efflux is a well-known resistance strategy, reduced azole import would be another strategy to maintain low intracellular azole levels. Recently, azole uptake in Candida albicans and other yeasts was analyzed using [³H]fluconazole. Defective drug import was suggested to be a potential mechanism of drug resistance in several pathogenic fungi, including Cryptococcus neoformans, Candida krusei, and Saccharomyces cerevisiae. We have adapted and developed an assay to measure azole accumulation in A. fumigatus using radioactively labeled azole drugs, based on previous work done with C. albicans. We used this assay to study the differences in azole uptake in A. fumigatus isolates under a variety of drug treatment conditions, with different morphologies and with a select mutant strain with deficiencies in the sterol uptake and biosynthesis pathway. We conclude that azole drugs are specifically selected and imported into the fungal cell by a pH- and ATP-independent facilitated diffusion mechanism, not by passive diffusion. This method of drug transport is likely to be conserved across most fungal species.

INTRODUCTION

The fungal pathogen Aspergillus fumigatus is one of the most common and ubiquitous environmental molds. Aspergillus infections represent a significant human health burden. Aspergillus causes serious illnesses, ranging from sinus infections to invasive or chronic aspergillosis in immunocompromised individuals (1–3). Infection can be fatal when Aspergillus colonizes or invades tissues, such as the lungs and blood vessels (2). The patients most at risk for these infections are those with prior lung conditions such as asthma, tuberculosis, chronic obstructive pulmonary disease (COPD), or cystic fibrosis, as well as bone marrow transplant patients and people living with HIV, AIDS, or other immune deficiencies (1–4). Aspergillus infections greatly affect the quality of life, cause dramatically lengthened hospital stays, and cost the United States more than 1 billion dollars each year (3, 5). In recent years, there has been a significant rise in the number of fungal infections due to the growing subpopulation of individuals with weakened immune health (3).

If not swiftly and appropriately treated, these fungal infections can progress to serious illness and rapidly become fatal (2, 6). The azole class of drugs is still the most commonly prescribed treatment for many fungal infections because the drugs are relatively cheap to produce, are generally nontoxic to humans, and are usually more effective for controlling an infection than other classes of antifungals (5, 7). Although A. fumigatus has an intrinsic reduced susceptibility to the most common azole drug, fluconazole (FLC), other azoles such as voriconazole, itraconazole (ITR), and posaconazole (POS) are common drugs of choice for prevention and treatment of aspergillosis (5, 7).

In a susceptible cell, azole drugs target the ergosterol biosynthesis pathway, which is unique to fungi but similar to the human cholesterol biosynthesis pathway (7). Specifically, azoles enter the fungal cell and inhibit the fungal cytochrome P450-dependent enzyme, lanosterol 14α-demethylase, encoded by the cyp51A and cyp51B genes in A. fumigatus (7, 8). Disruption of the ergosterol biosynthesis pathway leads to increased membrane permeability and instability in the fungal cell, which is deleterious to cell growth and replication (6, 9). The intrinsic resistance of A. fumigatus to FLC is possibly due to the differential FLC binding affinity to the Cyp51A and Cyp51B proteins (8, 10).

In response to reduced sterol availability, such as that resulting from azole treatment, the fungal cell may activate transcription factors, such as SrbA in A. fumigatus or Upc2 in C. albicans, which regulate the ergosterol biosynthesis and uptake pathways in an attempt to maintain sterol homeostasis (11, 12). In A. fumigatus, a deletion of the SrbA transcription factor leads to a mutant strain with FLC susceptibility, showing critical roles for sterol biosynthesis regulators such as SrbA in response to azole treatment (11, 12). Further, given the presumed intracellular localization of the azole target enzyme, the FLC-susceptible phenotype is highly indicative of cellular entry of FLC into A. fumigatus.

Azole drugs are frequently used prophylactically, sometimes for long periods of time, to prevent Aspergillus infections, as well as to treat these infections (1, 13). In addition, agricultural azoles are routinely used to control plant fungal pathogens (5). All of these factors have led to an emergence of fungal strains that have acquired the molecular mechanisms of azole drug resistance (5, 14–16). Common mechanisms of acquired drug resistance in fungi include mutations in or overexpression of the azole drug target enzyme lanosterol 14α-demethylase, other mutations in genes encoding enzymes in the sterol biosynthesis pathway, and prevention of accumulation of drugs to toxic levels in the cell (9, 17, 18).

The prevention of toxic levels of drug accumulation in the cell frequently involves the action of membrane efflux pumps. Increased expression and activity of cell membrane transporters that export antifungal drugs, such as the ATP binding cassette (ABC) and major facilitator superfamily (MFS) efflux pumps, are well-known mechanisms of drug resistance and have been extensively characterized in many fungal species. However, reduced azole import into the cell may be another mechanism to prevent drug accumulation (5, 19, 20). Mechanisms by which the drug enters the fungal cell or mechanisms that prevent the drug from entering have not been well characterized. Mansfield et al. (19) analyzed azole uptake in Candida albicans extensively using radioactively labeled azoles to measure how drug accumulation in the fungal cell changes under a variety of in vitro conditions. Some of the conditions examined included changes in temperature, pH, and oxygen availability, as well as different C. albicans cellular morphologies and C. albicans clinical isolates (19). The results of the C. albicans experiments using [³H]FLC show that azoles are not taken up solely by passive diffusion in C. albicans and suggest that deficient drug import might be a potential mechanism of drug resistance in C. albicans.

For Aspergillus, there has been an increase in azole-resistant A. fumigatus isolates (17, 21, 22). While many of these azole-resistant isolates have mutations in cyp51A or increased efflux, a significant fraction of clinical isolates have unknown resistance mechanisms that are independent of Cyp51A and efflux pumps (15, 23). Some specific isolates have been characterized as having an efflux-independent reduction in cellular drug accumulation (24, 25). Thus, in the interest of public health, there is an urgent need to better characterize biological processes in fungi, especially those that can lead to drug resistance, as well as to continue tracking and predicting future azole resistance mechanisms (22). This knowledge will be used to improve the existing therapeutic strategies or facilitate the development of new approaches for more effective treatments and preventions of fungal infections (5).

In this study, we developed an assay to directly measure azole drug accumulation in A. fumigatus under a variety of conditions that were previously examined for C. albicans and in a variety of A. fumigatus morphologies and strains. Consistent with the Candida work, our import experiments with A. fumigatus have shown that azole drugs do not accumulate in the fungal cell solely by passive diffusion. The evidence for azole import other than passive diffusion includes (i) dramatically reduced azole accumulation in heat-inactivated cells compared with that in living cells, (ii) decreased drug accumulation at higher temperatures, (iii) competition for import by other compounds, (iv) import specificity for certain chemical moieties present only on azoles and azole-like compounds, (v) differences in drug accumulation between different morphologies, including conidia, germlings, and mycelia, and (vi) differences in drug accumulation in exponentially growing and stationary-phase cells. We also found a significant decrease in the drug accumulation in the A. fumigatus srbA deletion mutant compared to that in the wild type. We conclude that azole drugs are specifically selected and imported into the fungal cell by a pH- and ATP-independent facilitated diffusion mechanism. This method of drug transport is likely to be conserved across most fungal species.

MATERIALS AND METHODS

Strains, media, materials, and strain preparation.

The A. fumigatus wild-type, sequenced strain CEA10 (CBS 144.89) was used for all of the azole import experiments unless otherwise noted. The srbA null mutant strain SDW1 (ΔsrbA::A. parasiticus pyrG pyrG1) and the complement strain SDW2 (ΔsrbA::A. parasiticus pyrG1 + srbA) were used in the mutant strain experiment. SDW1 and SDW2 were created in a CEA17 (pyrG1) background strain, a uracil auxotroph of wild-type strain CEA10. All strains were provided by Robb Cramer (Dartmouth University). Strains can be found in Table S1 in the supplemental material.

The strains were grown either in liquid or agar complete synthetic medium (CSM) (0.75 g of CSM [Bio 101, Vista, CA], 1.7 g of yeast nitrogen base without amino acids or ammonium sulfate, 5 g of ammonium sulfate, and 20 g of dextrose per liter) at 37°C. Conidia were harvested from 5- to 7-day-old agar plates by pipetting 7.0 ml of 0.01% Tween 20–water directly onto sporulating plates and harvesting conidia using a sterilized glass spreader. The dark green suspension was allowed to settle in a 14-ml round-bottom tube for 10 to 15 min, and then 500 μl of supernatant was aliquoted into cryotubes containing 500 μl of 60% glycerol and vortexed for a final glycerol concentration of 30%. The conidia concentration was determined by a hemocytometer, containing approximately 1 to 5 × 10⁸ conidia/ml. Stock solutions were stored at −80°C.

Liquid cultures were started from 20 μl of the glycerol conidia stock solution, inoculated into 5 ml, and grown in 50-ml conical tubes in a 37°C shaking incubator at 180 rpm for 24 h unless noted otherwise, at which time they formed mycelial masses or fungal balls, approximately 3 mm in diameter.

Medium components and plasticware were obtained from Fisher Scientific (Pittsburgh, PA) or Bio 101. General chemicals and unlabeled drugs used for competition were obtained from Fisher Scientific or Sigma-Aldrich (St. Louis, MO).

Radioactively labeled azole import by A. fumigatus.

Radioactive drugs included [³H]FLC (481 GBa/mmol, 13 Ci/mmol, 1 μCi/μl; 77 μM FLC; custom synthesis by Amersham Biosciences, United Kingdom), [³H]ketoconazole (KTC) (370 GBa/mmol, 10 Ci/mmol; American Radiolabeled Chemicals), and [³H]clotrimazole (CLT) (740 GBa/mmol, 20 Ci/mmol; American Radiolabeled Chemicals). Drug concentrations used during the import assay were well below (10-fold below) the MICs for the strain.

All experiments were performed as biological triplicates unless noted otherwise. To determine azole import in A. fumigatus isolates, we used [³H]FLC, [³H]KTC, or [³H]CLT in our drug uptake assay. Conidia were grown overnight in liquid CSM with 2% glucose at 37°C and 180 rpm shaking for 20 to 24 h at which point they were mycelial masses or fungal balls, unless otherwise noted. Every triplicate biological sample was tested separately. The fungal balls were harvested from each 50-ml tube and transferred to a 2-ml microcentrifuge tube and washed by centrifugation and resuspension three times with yeast nitrogen base (YNB) complete (1.7 g of yeast nitrogen base without amino acids or ammonium sulfate, 5 g of ammonium sulfate per liter [pH 5.0]) without glucose, unless otherwise noted. After the washing, the fungal ball pellet was transferred to 14-ml round-bottom tubes containing 1 ml of YNB for a 2-h glucose-deprived (starvation) deenergizing period. The glucose starvation was done to keep the cells in a deenergized state. The deenergized cells showed no further growth for the extent of the import assay, as determined by dry weight at the conclusion of the assay. The cells were still viable after the assay, as shown by agar plating (with the exception of the heat-killed samples). After the glucose starvation period, reaction mixes consisting of 1 ml of YNB containing no glucose, with fungal balls and 25 μl of diluted [³H]FLC (freshly diluted × from stock) were made. The resulting final [³H]FLC concentration is 19.25 nM (5.89 ng/ml), which is >50-fold less than the MIC value of the ΔsrbA mutant (MIC of 1 μg/ml) that is susceptible to FLC, and the wild-type strain (CEA10), which has an MIC of ≥50 μg/ml to FLC. Thus, the azole concentration used for the import assay was not expected to have any effect on cell viability.

After 24 h of incubation or at other specified times, a 5-ml stop solution (YNB + 20 mM [6 mg/liter] unlabeled FLC) was added to each 14-ml round-bottom tube sample. The tube was filtered by vacuum over preweighed and wetted glass fiber filters (24-mm GF/C; Whatman; Kent, United Kingdom). After filtering, another 5 ml of stop solution was used to wash each sample again. The filters with fungal balls were either allowed to dry for 24 to 48 h or were baked in a drying oven for 15 min at 95°C. Filters with attached fungal balls were then weighed to obtain the dry mass of each fungal sample. The weighed filters were transferred to 5-ml scintillation vials. Then 3 ml of scintillation cocktail (Ecoscint XR; National Diagnostics, Atlanta, GA) was added, and the radioactivity associated with the filter was measured in a liquid scintillation analyzer (LS 6500 multipurpose scintillation counter; Beckman Coulter). Results were calculated as counts per minute per milligram of mycelial mass. While absolute counts per minute values varied between experiments, relative import differences between samples remained consistent.

Characterization of conditions that affect FLC import.

Further studies were done to determine the effect of changes in the growth or incubation conditions on FLC import.

(i) Heat-killed cells.

Uptake of [³H]FLC was measured in cells inactivated (killed) by heat (95°C for 30 to 40 min). This killing method decreased CFU by >100-fold compared to CFU for the non-heat-treated culture. Samples were processed according to the above protocol with the exception that the heat-killing step was performed during the final 45 min of the glucose starvation period. Heat-killed samples were treated with [³H]FLC and analyzed identically to the live cell samples. The results were compared with live A. fumigatus data. Heat-killed samples were used as a control for baseline drug accumulation and nonspecific cell surface binding in all testing conditions unless otherwise noted.

Other methods of cell inactivation, including UV treatment (data not shown), amphotericin B treatment, and sodium azide treatment also each reduced azole import significantly (see Fig. S1 in the supplemental material). Heat killing was found to be the most reliable method for reducing viable cell counts to <1%. The reduced import due to amphotericin B treatment suggests that membrane permeability allows the drug to pass through the membrane in both directions. Caspofungin treatment at concentrations 4-fold above the MIC did not affect drug import, suggesting that cell wall disruption does not influence azole import.

(ii) Energy depletion.

To determine whether [³H]FLC import was energy dependent, cells were deenergized either by glucose starvation for 2 h in glucose-depleted medium or by treatment with the glycolysis inhibitor 2-deoxyglucose (5 mM) for 2 h. Following the deenergization period, the cells were then treated with [³H]FLC for 24 h in glucose-depleted medium and processed as described above.

(iii) Temperature.

Cells were grown overnight at 37°C as previously discussed and incubated with [³H]FLC at either 4°C, room temperature (20°C), 30°C, 37°C, or 42°C for 24 h. Samples were then processed as described above. As an alternative temperature characterization, cells were also grown overnight at different temperatures (20°C, 30°C, 37°C, and 42°C) and incubated with [³H]FLC at room temperature (20°C) for 24 h per standard treatment and processing.

(iv) pH.

Cells were incubated with [³H]FLC in YNB without supplementation, adjusted to pH 5, 7, or 9 with potassium hydroxide (to increase pH) or hydrochloric acid (to decrease pH) and buffered with 100 mM morpholinepropanesulfonic acid (MOPS), Tris, or HEPES, respectively. All samples were processed after 24 h as described above.

(v) Exponential- versus stationary-phase uptake of [³H]FLC.

Conidia were grown at 37°C in a shaking incubator at 180 rpm in CSM with glucose for either the standard 24 h (exponentially growing) or for 48 h or 72 h (stationary phase). These 1-, 2-, or 3-day-old fungal balls were then treated with [³H]FLC as described previously.

(vi) Competition for azole import.

To determine if compounds compete for azole uptake in A. fumigatus cells, we processed samples as described above for our [³H]FLC assay, and, in addition, treated the samples simultaneously with potential competitive inhibitors at 1.95 μM (100× molar excess of the labeled FLC). [³H]FLC uptake was measured as usual after 24 h of incubation with [³H]FLC and the competitor. Decreased [³H]FLC uptake in the presence of an unlabeled competitor suggests that both drugs use the same transporter.

(vii) [³H]FLC uptake among distinct A. fumigatus morphologies.

To compare [³H]FLC uptakes between the different morphological stages of A. fumigatus, conidia were harvested from agar plates as described previously. For conidia samples, conidia were washed, starved, and treated immediately. For germling and mycelial samples, the harvested conidia were allowed to germinate in liquid CSM with glucose and shaking at 180 rpm at 37°C for 4.5 h (clumping of conidia and apical extension visible), 7.5 h (germ tubes are long and distinct but not branching), or 24 h (balls of mycelial mass). Once the cells reached the desired time point, they were washed and starved in YNB with no glucose and incubated with [³H]FLC for 24 h as described previously. Because there was no carbon source during the [³H]FLC treatment, the morphologies were maintained, and there was no further growth of the A. fumigatus. [³H]FLC uptake was calculated as counts per minute per milligram of dry mass for each sample.

(viii) [³H]FLC import in A. fumigatus srbA deletion strain.

[³H]FLC accumulation in the A. fumigatus wild-type strain CEA10 was compared with accumulation in the SDW1 and the SDW2 strains. Heat-killed controls for each strain were included. Overnight growth and [³H]FLC treatment and processing were done as described previously. The mutant strains grew similarly to the wild type in shaking liquid cultures.

(ix) Efflux kinetics.

Fungal balls were preloaded with [³H]FLC by treating them at 19.5 nM for 24 h per the standard assay. The cells were then washed and diluted into YNB, and the amount of labeled drug associated with the cells was determined as a function of time at 4, 8, and 24 h. Efflux was evaluated in both glucose-energized (2% glucose) and deenergized (glucose-starved) cells (see above).

(x) Statistical analysis.

Differences between sets of samples were evaluated by an unpaired two-tailed Student's t test. A P value of <0.05 was considered significant.

RESULTS

Recently, radioactively labeled azole drugs were used to analyze [³H]FLC uptake in C. albicans (19). In that study, FLC was shown to be taken up by a facilitated diffusion mechanism, and import was affected by several in vitro environmental conditions (19). We have adapted the assay used in the previous study to characterize azole uptake in A. fumigatus. The basic protocol is described in Materials and Methods and outlined in Fig. 1A. As cell numbers are difficult to calculate with A. fumigatus, all values are expressed as drug accumulation in counts normalized to the dry weight of the sample. For our analysis, A. fumigatus was treated with ³H-labeled azoles, including FLC, KTC, and CLT (Fig. 1B), under a variety of conditions, and drug accumulation in the fungal cells was measured.

FIG 1.

FLC import in A. fumigatus. (A) Schematic diagram of the adapted import assay using radioactively labeled azoles to directly measure drug accumulation in A. fumigatus cells under a variety of conditions. (B) Comparison of ³H-labeled azole uptake (FLC, KTC, and CLT) in live cells versus that in heat-inactivated (HK) cells, used as a control. Error bars represent standard deviations of the means from biological triplicates for each condition. *, statistical significance of P < 0.05. (C) FLC accumulation in A. fumigatus wild-type strain CEA10 over a 72-h time course.

Azole uptake in A. fumigatus.

Figure 1B shows that all three radioactively labeled azoles (FLC, KTC, and CLT) are taken up by living A. fumigatus cells, while the heat-killed (HK) control cells showed significantly reduced drug accumulation. Heat-inactivated cells were subsequently used in all experiments as a baseline/background drug uptake control. The heat-killed cells treated with labeled KTC and CLT showed higher baseline counts than those treated with labeled FLC.

A. fumigatus and many other filamentous fungal species and molds are intrinsically resistant to FLC (26). However, due to the limited supply and expense of radioactively labeled KTC and CLT and the evidence that all three azoles are indeed taken up by A. fumigatus cells (Fig. 1B), subsequent experiments were performed using readily available [³H]FLC.

Figure 1C shows that [³H]FLC accumulation over a period of 72 h steadily increased up to 24 h at which maximum accumulation was reached. For this reason, unless stated otherwise, treatment assays were stopped at 24 h, and azole accumulation was measured at this time.

Import of [³H]FLC with varied in vitro treatment conditions. (i) Availability of energy.

To examine the effect of energy-requiring efflux mechanisms and to determine whether the accumulation of [³H]FLC in A. fumigatus cells requires energy, the effect of glucose was examined in the assay. Cells were deenergized by glucose starvation in glucose-depleted medium or by treatment with the glycolysis inhibitor 2-deoxyglucose. [³H]FLC accumulation of the deenergized samples was compared with [³H]FLC accumulation of energized samples grown in 2% glucose-replete medium (Fig. 2A). [³H]FLC uptake was observed in both deenergized and energized cells. However, the energized cells showed significantly reduced intracellular [³H]FLC concentrations compared to those in deenergized cells. The reduced azole accumulation in the energized cells is most likely the result of glucose activation of the efflux pumps. For subsequent experiments, we used deenergized (glucose-starved) cells to focus solely on drug uptake and eliminate efflux mechanisms.

FIG 2.

Effects of glucose, pH, temperature, and cell phase on FLC import. (A) Effect of glucose. Cells were deenergized either by glucose-depleted medium (−GLC) or by 5 mM treatment with the glycolysis inhibitor 2-deoxyglucose (2-Deoxy-Glc) compared to cells in the presence of glucose (+GLC) and measured for FLC accumulation. FLC import was not statistically different between the −GLC and 2-Deoxy-Glc cells. (B) Effect of pH. [³H]FLC was imported into A. fumigatus cells at pH 5, 7, and 9 using 100 mM MOPS, Tris, and HEPES, respectively, and adjustment with KOH and HCl. There were no statistically significant differences between import at the three conditions. (C) Effect of temperature. Cells grown overnight at the standard 37°C were treated with [³H]FLC and then incubated at 4°C, room temperature (20°C), 30°C, 37°C, or 42°C. The difference in import between cells treated at 20 and 30°C is not statistically significant. All other temperatures showed statistically significant import differences compared to that at 20°C. (D) Effect of stationary phase. Cells were grown in shaking liquid medium for 24 h, 48 h, or 72 h and then treated with [³H]FLC for 24 h. For each panel, error bars represent standard deviations of the means from biological triplicates for each condition. *, statistical significance of P < 0.05 between two conditions. Statistical differences or lack of differences from heat-killed controls (HK) are not shown.

(ii) pH.

To determine if [³H]FLC import is pH dependent or affected by a proton gradient (Fig. 2B), we measured drug accumulation after [³H]FLC treatment at pH 5, 7, and 9 using the buffers MOPS, Tris, and HEPES, respectively. There was [³H]FLC accumulation at all pHs tested, but there was no statistically significant difference between the different pHs, indicating that [³H]FLC import is not pH dependent. There was a trend toward alkaline sensitivity for drug uptake as seen by a decrease in [³H]FLC uptake in samples with pH 9 medium. However, deficiencies in cell growth and robustness are observed in A. fumigatus cells at pH 9, so import at this pH may be affected by other cellular factors directly or indirectly related to pH and proton gradients.

(iii) Temperature.

Figure 2C shows that [³H]FLC accumulation is affected by incubation temperature during treatment. Samples treated at 4°C and 42°C showed baseline drug accumulation. Samples treated at room temperature (20°C) showed significantly higher [³H]FLC accumulation than samples at the other temperatures, while samples at 30°C and 37°C showed an intermediate amount of accumulation. Cells were viable at all temperatures (data not shown). The decrease in drug accumulation at the higher temperatures argues against passive diffusion of [³H]FLC in A. fumigatus cells.

(iv) Stationary- versus exponential-phase cells.

Figure 2D shows that [³H]FLC accumulation is affected by the stage of hyphal growth of the A. fumigatus cells before treatment. Samples grown for the standard 24 h before drug treatment were compared to samples grown for 48 or 72 h before drug treatment. After [³H]FLC treatment, the exponentially growing (24 h) samples accumulated significantly more [³H]FLC than the stationary-phase (48 and 72 h) samples.

Competition for [³H]FLC import in A. fumigatus with azoles and other compounds.

To determine whether all azoles use the same transporter or family of transporters in A. fumigatus cells, non-radioactively labeled azoles were tested for competition against labeled FLC (Fig. 3; see also Tables S2 and S3 in the supplemental material). The concentration of all competitors was 1.95 μM (100× molar excess to [³H]FLC). All of the nonradiolabeled azoles that were tested (fluconazole [FLC], clotrimazole [CLT], itraconazole [ITR], ketoconazole [KTC], metconazole [MET], miconazole [MCZ], posaconazole [POS], prochloraz [PCZ], propiconazole [PROP], prothioconazole [PROT], and tebuconazole [TEB]) competed for import with [³H]FLC, as indicated by a reduction in the [³H]FLC accumulation to baseline levels. These azoles include medically important azoles (FLC, CLT, ITR, KTC, MCZ, POS) as well as agriculturally important azoles (MET, PCZ, PROP, PROT, TEB). 1-(Triphenylmethyl)-imidazole (1-TRI), a CLT analog, also competed with FLC for import. The results of these competition experiments are consistent with azole import being mediated by a facilitated diffusion carrier(s) and suggest that these azoles use the same transport system to enter A. fumigatus cells.

FIG 3.

Competition for [³H]FLC import in A. fumigatus. Compounds were tested for competition at 1.95 μM (100× molar excess to [³H]FLC) during simultaneous treatment with [³H]FLC. [³H]FLC accumulation was measured after 24 h of incubation with competitors. Error bars represent standard deviations of the means from biological triplicates for each condition. Compounds clustered to the left side of the figure compete with [³H]FLC for import and are statistically different from the no-competitor sample. Compounds clustered to the right side of the figure do not compete with [³H]FLC and are not statistically different from the no-competitor sample with the exception of 1-benzylimidazole, which had statistically higher FLC import. HK, heat-killed control; FLC, fluconazole; CLT, clotrimazole; ITR, itraconazole; KTC, ketoconazole; MET, metconazole; MCZ, miconazole; POS, posaconazole; PCZ, prochloraz; PROP, propiconazole; PROT, prothioconazole; TEB, tebuconazole; 1-TRI, 1-(triphenylmethyl)-imidazole; AMP, amphotericin B; AZO, azoxystrobin; 1-BENZ, 1-benzylimidazole; CFG, caspofungin; 5-FC, 5-flucytosine; HEM, hematin; 4-IMID, 4-(imidazole-1-yl) phenol; NIC, nicotinamide; R-6G, rhodamine 6-G; RUF, rufinamide; TRB, terbinafine; VIN, vinclozolin.

Other compounds were also tested for competition against [³H]FLC (Fig. 3; see also Tables S2 and S3 in the supplemental material), including azole-like compounds and other common antifungals (azoxystrobin [AZO], 1-benzylimidazole [1-BENZ], caspofungin [CFG], 5-flucytosine [5-FC], hematin [HEM], 4-(imidazole-1-yl) phenol [4-IMID], nicotinamide [NIC], rhodamine 6-G [R-6G], rufinamide [RUF], terbinafine [TRB], and vinclozolin [VIN]). Most of the nonazole compounds did not compete with [³H]FLC for import into A. fumigatus cells. These included medically important antifungals (CFG, 5-FC, TRB), as well as agricultural antifungals (AZO, VIN) and molecules with some similarity to the medical azoles (1-BENZ, NIC, RUF). The results of these competition experiments indicate transport specificity to certain chemical structures and suggest that the compounds do not use the same transporter(s) to enter A. fumigatus cells.

The structures of FLC and the compounds tested in this assay are shown in Tables S2 and S3 in the supplemental material. FLC has two 5-membered triazole rings containing 2 nitrogens and a 6-member halogenated benzene ring. Previous screening of moieties important for import in C. albicans are consistent with this result. Analyses in both C. albicans and A. fumigatus suggest that to compete for FLC import, a compound requires a 5-membered ring with two (imidazole) or three (triazole) nitrogens, in addition to a halogenated 6-membered ring with the halogen in position 1 or 3 but not necessarily both (19). The only exception to this has been the competition of 1-(triphenylmethyl) imidazole, which has a 5-membered imidazole ring as well as three 6-membered rings but is not halogenated.

Import of [³H]FLC in different A. fumigatus morphological forms.

A. fumigatus undergoes several stages of growth from conidia through germlings to mature hyphae. Each is important for different aspects of survival and pathogenesis. We determined azole accumulation during several stages of growth (Fig. 4). The different stages analyzed are pictured in Fig. S2 in the supplemental material. The amount of radioactivity inside the cells was measured per sample biomass and compared. Conidia showed baseline [³H]FLC accumulation, while the young germlings (4.5 h of germination), older germlings (7.5 h of germination), and the standard mature mycelia (24-h fungal balls) accumulated significantly more [³H]FLC at each stage with 24 h being maximal. The mycelial fungal balls accumulated the greatest amount of [³H]FLC compared to those for the other morphological forms. Thus, the 24-h mycelial form was used for other characterizations. The difference in import between the 4.5- and 7.5-h germlings was not statistically significant. Curiously, the HK 7.5-h germlings showed reduced baseline counts.

FIG 4.

A. fumigatus fungal balls take up more [³H]FLC than conidia or germlings. The different growth stages of A. fumigatus (see Materials and Methods and Fig. S2 in the supplemental material) were incubated with 19.25 nM [³H]FLC for 24 h, and drug uptake was measured. Error bars represent standard deviations of the means from biological triplicates for each condition. *, statistical significance of P < 0.05 between two conditions.

Import of [³H]FLC in mutant A. fumigatus strains.

SrbA is a sterol regulatory element binding protein in Aspergillus that primarily regulates ergosterol biosynthesis and also regulates many genes important for basal, as well as stressed, cell activity. This includes transcriptional activation of a variety of transporter or putative transporter genes, as well as genes important for membrane stability and function (11, 12). Import analyses on mutant Aspergillus strains showed differences in drug accumulation when a gene for the major transcription factor SrbA, SDW1, was deleted compared to that for the wild-type strain and the complemented strain SDW2 (Fig. 5). SDW1 shows significantly reduced [³H]FLC accumulation compared to that for the wild-type strain and SDW2. This suggests that the SrbA transcription factor is important for azole import, whether by direct or indirect effects. SDW2 recovered the wild-type drug susceptibility and morphological phenotype (11, 12). However, the drug import was not restored to the CEA10 wild-type levels; instead there was an intermediate level of drug accumulation in this strain, lower than the wild type and higher than the deletion mutant (Fig. 5). This might be due to differences in transcriptional activity in the ectopic reinsertion of the srbA gene, additional point mutations in this strain, or the CEA17 mutant background strain.

FIG 5.

srbA deletion affects [³H]FLC accumulation. [³H]FLC uptake in the wild-type A. fumigatus strain CEA10 was compared to those in the SDW1 and SDW2 strains. SDW1 accumulated significantly less [³H]FLC than the wild-type CEA10 and SDW2 strains, while the SDW2 strain showed intermediate drug uptake. Error bars represent standard deviations of the means from biological triplicates for each condition. *, statistical significance of P < 0.05 between two conditions.

Efflux of [³H]FLC from preloaded cells.

Given the import of radiolabeled azoles over 24 h, it was important to study how the drug is exported after import. Figure 6 shows cells that were preloaded with [³H]FLC at 19.5 nM for 24 h. The cells were diluted into a 50-fold excess volume of YNB medium, and the amount of labeled drug associated with the cells was determined as a function of time. Efflux was evaluated in both glucose-energized (gray line with squares) and deenergized (glucose-starved) (black line with circles) cells. [³H]FLC efflux was slow and dependent on energy, consistent with the activation of membrane efflux transporters to pump the azole out of the cell. By 24 h, most of the [³H]FLC was exported from the cells in both energized and deenergized conditions. Those samples incubated in the presence of glucose show faster export of labeled drug, consistent with the idea that membrane efflux pumps require energy and that glucose starvation limited the energy and thus the efflux pump activity.

FIG 6.

Efflux of [³H]FLC from preloaded cells. Cells were preloaded with 19.25 nM [³H]FLC for 24 h. The cells were washed with YNB medium and placed in either deenergized (glucose-free [−GLC]) (circles) or a 2% glucose-energized (+GLC) (squares) medium for 4, 8, and 24 h to measure the efflux of [³H]FLC in A. fumigatus. Error bars represent standard deviations of the means from biological triplicates for each condition. Some error bars are hidden by the symbols.

DISCUSSION

In this work, we have analyzed a potential molecular mechanism of azole drug resistance in A. fumigatus, by characterizing azole import into the fungal cells under a variety of environmental conditions. Azole drug uptake is required so that the drug can inhibit the intracellular Cyp51 target enzyme. Therefore, reduced or modified drug import may help to explain why some pathogenic fungi are more resistant to azoles than others. For A. fumigatus, there is an increasing number of cases of azole-resistant isolates in which the resistance mechanism is unknown (15). Many of these isolates do not have a mutation in the azole target enzyme Cyp51, nor do they show active efflux of the drug from the cells. These A. fumigatus isolates may well have alterations in drug import. Our assay will further be used to compare drug import in clinical isolates and resistant strains of A. fumigatus and other filamentous fungi.

The study of drug import in A. fumigatus cells using radioactively labeled azoles has not been done previously. This approach includes novel experiments that may identify differences in substrate specificities of medically relevant compounds. Our experiments thus far, in agreement with the work of Mansfield et al. work in C. albicans (19), have demonstrated that azoles are not passively diffused into the A. fumigatus cell nor are they actively transported in an energy-dependent fashion. Instead, azoles most likely enter the cell by facilitated diffusion via a membrane protein carrier that recognizes a specific moiety found in azole drugs.

The labeled azoles do not simply bind to the cell surface as demonstrated by baseline drug accumulation in heat-inactivated cells compared to that in living cells. Drug uptake can be measured over time in live cells, reaching a maximum accumulation after 24 h of treatment (Fig. 1C), while heat-inactivated cells maintain a constant drug concentration.

Import is not diffusion limited, since the internal drug concentration is much lower than that of the external medium (data not shown). The drug concentrations used in our import analysis were in the nanomolar range. The possibility that normal diffusion or nonspecific carrier transport occurs at higher drug concentrations cannot be ruled out. It is also not known whether the import transporters also act as efflux transporters for azoles or other molecules.

Drug uptake by this carrier is energy independent, as seen by drug accumulation measured in deenergized cells. Azoles were imported in both energy-replete and glucose-containing media. Even cells treated with 2-deoxyglucose, a glycolytic inhibitor, showed drug uptake (Fig. 2A). However, the medium that contained glucose showed reduced final drug accumulation levels, presumably due to activation of glucose-dependent efflux pumps. It is possible that cells are better able to metabolize and degrade FLC in energized conditions. However, the radioactive label would still be detected regardless of cleavage or FLC degradation. There has been no evidence to date that FLC is degraded by fungal cells. Our evidence indicates that efflux of azoles is dependent on energy (Fig. 6), suggesting distinct transporters for influx and efflux of azoles. Import of azoles also did not require a proton gradient as shown by no change in uptake over a range of buffered pHs (Fig. 2B).

Changes in treatment temperatures had a significant effect on azole drug uptake in A. fumigatus (Fig. 2C). A. fumigatus has a remarkable ability to thrive at a range of environmental temperatures, including 4°C to 42°C. However, drug accumulation was near baseline levels for both 4°C and 42°C, with the highest accumulation at 20°C to 22°C (Fig. 2C). These dramatic differences in temperature-dependent accumulation are being further analyzed but might be due to changes in gene expression, membrane fluidity, drug stability, or protein folding among other things.

A comparison of drug import in exponentially versus postexponentially growing cells (Fig. 2D) showed a dramatic decrease in accumulation in the older cells. This is in agreement with postexponential cells being less metabolically active than exponentially growing cells. A cessation of cell division, a decrease in macromolecular production, and other changes in cellular regulation might be responsible for this decrease in azole uptake in postexponential cell growth.

Evidence for a saturable protein carrier is shown with competitive inhibition of [³H]FLC uptake by other azoles (Fig. 3). [³H]FLC import was significantly inhibited by simultaneous treatment with an excess of unlabeled azoles. However, other antifungal drugs did not compete for import into the cell. This indicates substrate specificity for moieties found in the azole structure. The structures of competitive inhibitors and drugs that did not compete for import are shown in Tables S2 and S3, respectively, in the supplemental material.

Differences in import between the different Aspergillus morphologies (Fig. 4) is interesting but expected since there are many differentially expressed genes and cellular changes during the different growth phases of this organism. The inactive conidia accumulated only baseline levels of azole per weight, while the germlings had increased azole accumulation. The mature hyphae forming mycelial masses imported the largest amounts of drug per mass. Images of the cellular morphologies used can be seen in Fig. S1 in the supplemental material. A major factor to consider is whether comparing drug import per mass is appropriate for each morphology. The weight per volume ratio is very likely different between the cellular morphologies. Conidia and germlings are perhaps more dense than actively growing cells due to nutrient and protein reserves. In addition to these considerations, many cell wall changes take place during germination and hyphal growth, including variations in surface-expressed proteins.

Import analyses for mutant Aspergillus strains showed differences in drug accumulation when the gene for a major transcription factor, SrbA, was deleted (Fig. 5). SDW1 showed significantly reduced azole import compared with the wild-type CEA10 strain. It is possible that this mutant has a disrupted cell membrane, altering membrane transport. Many genes regulated by the SrbA transcription factor, such as the azole drug targets cyp51A/B, have altered expression in this strain and may be responsible for differences in drug accumulation data observed in the mutant. It is also possible that SrbA regulates a drug transporter involved in bringing azoles into the cell so that when this transcription factor is deleted, the transporter is downregulated. Future work should be done to determine if the azole importer protein is regulated by the SrbA transcription factor. It is curious that reinsertion of the srbA gene in SDW2 did not fully restore azole import to wild-type levels. However, this might be the result of the ectopic reintroduction of the gene or additional point mutations in the mutant background strain.

While we have performed a preliminary investigation of drug transport across the cell membrane, the next steps are to identify the proteins that are involved in drug import. Identification of the specific channel or transport protein by which this occurs is still in progress and is a major goal of our research. Determining the proteins involved in transporting azoles into the fungal cell would provide insight into how our current drug treatments are working. Novel import proteins might potentially provide new targets for drug treatment. The long-term goals of this research are to discover and to better characterize the cellular mechanisms of drug resistance in A. fumigatus that can be targeted by new and better-designed drugs for more effective treatments of fungal infections.