The yeast Saccharomyces cerevisiaePDR16 (Pleiotrophic Drug Resistance 16) gene was originally identified in a genome-wide microarray analysis of genes that are regulated by the multiple drug resistance transcription regulators Pdr1p and Pdr3p (DeRisi et al., 2000) (do Valle Matta et al., 2001). The PDR16 gene promoter contains pleiotropic drug response elements (PDREs), sequences responsible for activation by the Pdr1p and Pdr3p transcription factors. Using several methods (microarray analysis, heterologous β-galactosidase assays and northern blots), it has been shown that PDR16 gene transcription is highly induced in yeast strains containing gain-of-function alleles of Pdr1p and Pdr3p (DeRisi et al., 2000; do Valle Matta et al., 2001). These transcription factor homologues, Pdr1p (Balzi et al., 1987) and Pdr3p (Delaveau et al., 1994), are master regulators of the yeast PDR network (Prasad and Goffeau, 2012). Similar to most prokaryotes and eukaryotes, yeasts have the ability to adapt to a broad range of chemically and functionally unrelated cytotoxic compounds. One major general mechanism of adaptation is drug export. Yeasts have developed a large family of exporters involved in pleiotropic drug resistance, the cellular efflux of a wide variety of drugs. Typically, transcription of these exporters is under the control of the Pdr1p and Pdr3p transcription factors. However, not all proteins under the control of Pdr1p and Pdr3p are efflux pumps (DeRisi et al., 2000). Phenotypically, S. cerevisiae pdr16Δ cells are hypersusceptible to the azole antifungals miconazole, ketoconazole (van den Hazel et al., 1999) and fluconazole (Gulshan et al., 2010). Interestingly, pdr16Δ cells do not display increased susceptibility towards other drugs tested, such as nystatin, cycloheximide, rhodamine-6G, oligomycin, 4-nitroquinoline-N-oxide, antimycin A, ethidium bromide and crystal violet (van den Hazel et al., 1999).
Pdr16p, also called Sfh3p, belongs to the family of phosphatidylinositol (PI) transfer proteins (PITP) (Griac, 2007; Li et al., 2000). The founding member of this protein family in yeast, Sec14p, facilitates the transfer of PI and phosphatidylcholine (PC) between donor and acceptor membranes in in vitro assays (Bankaitis et al., 1989; Daum and Paltauf, 1984). Pdr16p is considered to be a non-classical PITP, since it can bind and transfer PI but not PC (Li et al., 2000). As for its subcellular localization, Pdr16p was found to be present in lipid particles, microsomes and at the cell periphery (Schnabl et al., 2003). Intensive structural studies of the yeast Pdr16p (named Sfh3p in the corresponding paper) started by (Ren et al., 2011). Pdr16p was purified to homogeneity and crystals for structural analysis were obtained. Subsequently, diffraction data were successfully collected. Initial failure to phase the Pdr16p diffraction data to the members of the Sec14p family in yeast with known structures, namely Sec14p and Sfh1p, indicate that the conformation of Pdr16p is sufficiently different from either Sec14p or Sfh1p (Ren et al., 2011). Indeed, the recent paper from the Niu and Teng laboratories (Yuan et al., 2013) reported the dimeric structure for Pdr16p (named Sfh3p in this paper), which differs from the Sec14 group of protein structures, all of which were monomeric. The authors propose an interesting model that explans the dimer–monomer state change of the Pdr16p, depending on PI binding that may destabilize the Pdr16p dimer.
Candida albicans is a major fungal pathogen of humans, causing serious medical problems, including life-threatening systemic infections in immunocompromised patients. Azole derivatives are often used to treat these infections. It was shown that many clinical isolates of C. albicans azole-resistant strains overexpress the CaPDR16 gene in addition to overexpression of multidrug transporters (Saidane et al., 2006). Further experiments showed that increased dosage of CaPdr16p confers increased azole resistance, even in the absence of additional molecular alterations, indicating that overexpression of the CaPDR16 gene plays a role in C. albicans azole resistance (Saidane et al., 2006). Deletion of CaPDR16 in azole-resistant cells decreased their azole resistance, demonstrating CaPDR16 to be a factor of clinical azole resistance in C. albicans (Saidane et al., 2006; Znaidi et al., 2007).
In S. cerevisiae, the PDR16 deletion increases susceptibility to azole antifungals and reduces the evolutionary potential to develop fluconazole resistance (Anderson et al., 2009). Thus, the effect of Pdr16p goes beyond simply affecting the resistance to azoles. Based on these findings, the authors suggested that PDR16 should be considered an additional target for prevention of azole resistance.
The current study is aimed at a better understanding of Pdr16p function, especially in relation to azole resistance in yeast. We have identified changes in ergosterol biosynthesis in the pdr16Δ strain when the yeast cells were challenged with azoles. However, these changes were not caused by the increased accumulation of azoles. Based on complementation studies of the azole-susceptibility phenomenon, we propose a hypothesis that Pdr16p could assist in intracellular shuttling of sterols or their intermediates between membranes or, alternatively, between biosynthetic enzymes or complexes.
Materials and methods
Materials and drugs
Media components were obtained from Becton-Dickinson (USA) or BioLife (Italy). Fluconazole, paclobutrazol and azaconazole were obtained from Sigma-Aldrich (USA), ketoconazole and miconazole from MP Biomedicals (USA) and voriconazole from Discovery Fine Chemicals (UK). Fine chemicals were mostly from Sigma-Aldrich or MP Biomedicals. [3H]-labelled fluconazole for uptake studies into de-energized cells was custom-synthesized by Amersham Biosciences, UK (specific activity 740 GBq/mm) (Mansfield et al., 2010). [3H]-labelled fluconazole for other studies was obtained from Moravek Biochemicals, USA (9.25 MBq/250 µl, specific activity 103.6 GBq/mm). [14C] acetic acid was from MP Biomedicals (9.25 MBq/250 µl, specific activity 2.07 GBq/mm).
Strains and culture conditions
Wild-type strain FY1679-28c and its derived pdr16Δ, originally from A. Goffeau's laboratory (Catholic University Louvain, Belgium) (van den Hazel et al., 1999), were kindly provided by G. Daum (Technical University, Graz, Austria). BY4741 wt and its derived pdr16Δ strain were from the S. cerevisiae haploid gene-deletion library (Research Genetics, Huntsville, AL, USA). The AD1-8u– strain, in which seven major azole efflux pumps were deleted (Decottignies et al., 1998), was a kind gift from A. Goffeau (Catholic University Louvain, Belgium). AD1-8u–pdr16Δ strain was constructed by a one-step disruption method (Wach et al., 1994), using a kanMX module flanked by long homology regions of PDR16 from plasmid pYORC-YNL231c (EUROFAN, Germany). Wild-type and mutant strains used have matched auxotrophies. The URA3 gene used to create the pdr16Δ mutant in the FY1679-28c genetic background was 'looped-out' using recombination between hisG sequences flanking the URA3 gene (van den Hazel et al., 1999). Episomal plasmids containing SEC14 or its homologues, transcribed from their own promoters, were constructed on the basis of the 2 µ plasmid YEplac181. Details of their construction are described in Schnabl et al. (2003).
Yeast strains were grown on yeast extract/peptone/dextrose (YEPD; 2% glucose) medium, unless otherwise stated. Yeast strains containing episomal plasmids were maintained and pre-grown on standard synthetic minimal medium (0.67% YNB without amino acids, 2% glucose) supplemented with essential amino acids and bases as required for plasmid maintenance. For the fluconazole uptake experiment under de-energized conditions, yeast cells were pre-grown in CSM complete medium (Bio 101, Vista, CA, USA).
Drug susceptibility testing
Drug susceptibility was determined by a spot assay. Yeast cultures were grown overnight in YEPD or YNB – LEU medium (for strains containing episomal plasmids), diluted and spotted as 10-fold dilutions onto YEPD medium containing indicated azoles. The growth was scored after 2 days of incubation at 28 °C.
Pulse-labelling of neutral lipids using [14C] acetic acid
The wt strain FY1679-28c and its corresponding pdr16Δ strain were pre-grown overnight in YEPD medium and diluted to 5 × 106 cells/ml into fresh YEPD medium; 4 h later, 37 kBq [14C] acetic acid/ml medium was added. The cells were then incubated at 28 °C with shaking. At 30 and 120 min after the addition of radioactive acetic acid, the equivalent of 5 × 107 cells was collected, washed twice with water and combined with 1 × 109 non-radioactive carrier yeast cells. Sub-inhibitory concentrations of either miconazole (5 ng/ml) or fluconazole (2 µg/ml) were added 2 h before addition of radioactive acetic acid when pulse-labelling of lipids was done in the presence of azoles.
Lipid extraction and analysis
After [14C] acetic acid labelling, lipids were extracted for TLC analysis of neutral lipids. This was performed according to Bligh and Dyer (1959), with minor modifications. Combined radioactive and carrier yeast cells, prepared as described above, were disrupted by vortexing with glass beads six times for 1 min each, with cooling on ice. Lipids were extracted from the cell homogenate by hot methanol (15 min at 70 °C), followed by two-fold extraction in chloroform:methanol:H2O (1:2:0.8) at room temperature. The organic phase containing lipids was separated and evaporated. The dry lipid residue was dissolved in chloroform:methanol (2:1) and applied to silica gel plates (Merck, Germany), using a Linomat 5 automatic sampler (Camag, Switzerland). Neutral lipids were separated by ascending two-step thin-layer chromatography (first step, petroleum ether:diethyl ether:acetic acid 70:30:2; second step, petroleum ether:diethyl ether 49:1). Silica gel plates with resolved lipids were exposed for 7–10 days to Kodak storage phosphor screens. Storage screens were subsequently scanned using a Bio-Rad FX Phosphorimager and data were analysed using Quantity One software (Bio-Rad, USA). Finally, lipids on TLC plates were visualized by charring with sulphuric acid and individual lipids were identified by comparison with known standards run on the same TLC plate.
For sterol analysis by HPLC, non-saponifiable lipids were isolated by the modified procedure of Breivik and Owades (1957). Briefly, 1 × 109 cells were broken by homogenization with glass beads and incubated in 3 ml 60% w/v KOH in 50% methanol (v/v) for 2 h at 70 °C. Non-saponifiable lipids were extracted twice with 3 ml n-hexane and the combined extracts were dried under N2. The lipid residue was dissolved in acetone and analysed by reversed-phase HPLC on an Agilent 1100 HPLC equipped with Eclipse XDB-C8 column (Agilent Technologies, USA), diode array detector (Agilent Technologies) and Corona charged aerosol detector (ESA Inc., USA). Sterols were eluted at 40 °C with 95% methanol at a flow rate of 1 ml/min. The peak identity was determined from the retention times of standards [ergosterol, lanosterol (Serva, Germany) and squalene (Sigma-Aldrich, USA)] and from their characteristic spectra. The quantity of the sterols was calculated from calibration curves constructed for individual standards.
Fluconazole accumulation by S. cerevisiae cells
Two types of fluconazole accumulation experiments were performed. The first was to study the uptake of fluconazole into yeast cells under de-energized conditions, according to Mansfield et al. (2010). Briefly, cells were grown overnight in CSM complete medium, harvested by centrifugation and washed three times with YNB complete medium (1.7 g yeast nitrogen base without amino acids or ammonium sulphate, 5 g/l ammonium sulphate, pH 5.0) without glucose. Cells were resuspended at OD600 = 75 in YNB for 2–3 h for glucose starvation. Reaction mixes consisted of 250 µl YNB, 200 µl cells (75 OD) and 50 µl 1/100 dilution of stock solution of [3H]-labelled fluconazole (specific activity 740 GBq/mm). Chemically, the resulting [3H] fluconazole concentration was 15 ng/ml, well below the MIC for all strains tested. 100 µl samples were removed at 24 h and placed into 5 ml stop solution (YNB + 20 µm fluconazole), filtered, washed with the stop solution and the radioactivity associated with the cells was counted with a liquid scintillation analyser. Data were normalized to CPM/1 × 108 cells.
The second type of fluconazole accumulation experiment measured the accumulation of radioactive fluconazole in the yeast cells under the conditions in which energy-driven importers and efflux pumps were operational. Yeast cells were pre-grown in YEPD medium to the exponential phase of growth (2 × 107 cells/ml). These cultures were washed twice and concentrated in YNB medium containing 2% glucose to the concentration 1 × 109 cells/ml. 10 µl 10 times-diluted stock solution (37 kBq) of [3H] labelled fluconazole (specific activity 103.6 GBq/mm) were added per ml cell suspension. Chemically, the final concentration of [3H] fluconazole, 109 ng/ml, was significantly below the MIC for both tested strains. At the indicated time points, 500 µl cell suspension was removed, filtered and washed three times with a wash solution containing 10 µg/ml fluconazole and 1% NP40. The radioactivity associated with the cells was counted using a liquid scintillation analyser (Beckman LS6000SE). Data were normalized to DPM/1 × 108 cells.
pdr16Δ strain showed increased susceptibility towards all tested azoles
van den Hazel et al. (1999) reported that deletion of PDR16 resulted in hypersusceptibility to azole inhibitors of ergosterol biosynthesis, namely miconazole and ketoconazole. Later, Gulshan et al. (2010) showed increased susceptibility of the pdr16Δ strain to fluconazole. We extended these observations. In addition to miconazole, ketoconazole and fluconazole, three other azoles were tested: a triazole antifungal drug used in human medicine, voriconazole, and two agricultural fungicides, azaconazole and paclobutrazole. The yeast strain deleted for PDR16 exhibited an increased susceptibility to all azoles tested compared to its parental strain FY1679-28c (Figure 1). In agreement with the previous observation, an increased dosage of the PDR16 gene under its own promoter on a multicopy plasmid caused a slight increase in resistance towards miconazole (data not shown).
Sterol analysis of the pdr16Δ strain showed enhanced lipid alteration compared to wt cells in the presence of sub-inhibitory concentrations of azoles
In the paper describing the role of Pdr16p and Pdr17p in lipid biosynthesis, van den Hazel et al. (1999) reported some changes in sterol composition of the purified plasma membrane of the pdr16Δ strain when compared to its parental strain, FY1679-28c. We measured individual sterols in the wt strain FY1679-28c and in the pdr16Δ mutant in cells grown in YEPD without azoles and with a sub-inhibitory concentration of miconazole. In the absence of azoles, we were not able to identify any changes in cellular sterol composition of the pdr16Δ strain compared to its parental strain, FY1679-28c, using HPLC analysis (data not shown). If there were changes in sterol biosynthesis, these changes could be more likely identified by pulse-labelling and separation of neutral lipids using radiolabelled acetate (Figure 2).
The results showed that, in the absence of azoles (Figure 2A), there are no changes in the dynamics of sterol biosynthesis between the pdr16Δ strain and its corresponding parental strain. Thus, it is very unlikely that any changes in sterol composition or the dynamics of sterol biosynthesis between the pdr16Δ strain and its corresponding parental strain are present under the standard growth conditions without the presence of azole antifungals. However, in the presence of sub-inhibitory concentrations of miconazole there was a dramatic change in sterol biosynthesis between the pdr16Δ strain and its corresponding parental strain, FY1679-28c (Figure 2B). Neutral lipids in the pdr16Δ strain and its parental strain were labelled for 30 and 120 min with radiolabelled acetate and separated by TLC. The results indicated an increased accumulation of lanosterol and squalene at the expense of the final product of the yeast sterol biosynthetic pathway, ergosterol, especially after 120 min pulse in the pdr16Δ strain. We repeated the experiment using fluconazole, with the same result. In the presence of sub-inhibitory concentration of fluconazole, we have found increased accumulation of lanosterol and squalene and decrease of ergosterol after 120 min pulse in the pdr16Δ strain compared to its parental strain (data not shown). Azole antifungals affect ergosterol biosynthesis mostly by inhibiting the ERG11 gene product, lanosterol 14α-demethylase (Lupetti et al., 2002; MacCallum et al., 2010). As a result, increased accumulation of lanosterol, the substrate for lanosterol 14α-demethylase and also squalene can be seen even in the wt strain at the expense of the final product of the yeast sterol biosynthetic pathway, ergosterol. The pdr16Δ mutant, when grown in the presence of miconazole or fluconazole, accumulates significantly more lanosterol and squalene, together with a considerable decrease in ergosterol compared to the wt. Thus enhanced inhibition of sterol biosynthesis can explain the increased azole susceptibility of the pdr16Δ strain.
Pdr16p does not influence the known major azole efflux pumps – genetic evidence
One possible explanation for the increased azole susceptibility of the pdr16Δ strain and the enhanced sterol alteration in the presence of azoles could be that the intracellular levels of azoles are increased in the pdr16Δ strain compared to wt. This potential increase of intracellular levels of azoles could occur because of enhanced drug uptake or because of defective drug efflux in the pdr16Δ strain. The yeast plasma membrane contains a number of protein pumps which can extrude drugs from the cell (Prasad and Goffeau, 2012). Pdr16p encodes a protein of 351 amino acids that do not resemble an efflux pump. The Pdr16p protein is not predicted to contain transmembrane domains (Reynolds et al., 2008) or nucleotide-binding sites. It is nevertheless possible that deletion of the PDR16 gene alters the activity of efflux pumps important for extrusion of azole molecules from the cells. To test this possibility, we performed a genetic experiment using the yeast strain AD1-8u–, in which seven major ABC transporters have been deleted (Decottignies et al., 1998). We deleted PDR16 in addition to the seven major yeast efflux pumps, Yor1p, Snq2p, Pdr5p, Pdr10p, Pdr11p, Ycf1p and Pdr15p, already deleted in the AD1-8u– strain. We named this newly constructed strain AD1-8u–pdr16Δ and compared the susceptibility of AD1-8u– and AD1-8u–pdr16Δ strains on plates containing miconazole (Figure 3). If the major mechanism of action of Pdr16p is to directly or indirectly influence known drug efflux pumps or their activity, additional deletion of the PDR16 gene in the strain in which all known major azole efflux pumps are already deleted, AD1-8u–, should not further increase susceptibility of the AD1-8u–pdr16Δ to the azole antifungal miconazole. The results depicted in Figure 3 demonstrate that deletion of PDR16 in the AD1-8u– background increased the susceptibility of yeast cells to miconazole. This genetic experiment suggests that Pdr16p mediates azole susceptibility by a mechanism independent of azole efflux pumps deleted in the AD1-8u– strain. However, this experiment does not rule out the possibility that other transporters, such as the major facilitator Flr1p, or other uncharacterized ABC transporters, could be involved in the efflux of azoles regulated by Pdr16p. To better exclude the possibility that the susceptibility to azole drugs in the pdr16Δ mutant is not due to an increased intracellular concentration of azoles in the pdr16Δ mutant, we performed the following drug uptake and accumulation studies.
Fluconazole uptake and accumulation into pdr16Δ and wt cells are the same
We compared radioactive fluconazole uptake into wt and pdr16Δ cells under de-energizing conditions in two different genetic backgrounds, using the method described in Mansfield et al. (2010) (for details, see Materials and methods). The data clearly demonstrate that deletion of the PDR16 gene did not alter the import of [3H] fluconazole into yeast cells (Figure 4).
Using [3H]-labelled fluconazole, we then measured azole accumulation into wt cells and cells deleted for PDR16 under conditions in which energy-driven importers and efflux pumps were operational (Figure 5). Briefly, yeast cells were pre-grown in YEPD medium to the exponential phase of growth (2 × 107 cells/ml). Next, the cultures were washed and concentrated in YNB medium containing 2% glucose to a concentration of 1 × 109 cells/ml. Subsequently, accumulation of radioactive fluconazole into yeast cells was measured using the cell filtration technique (for details, see Materials and methods). Chemically, the final concentration of [3H]-fluconazole was significantly below the MIC for both tested strains. The results show the identical accumulation of radiolabelled fluconazole into wt cells and into cells of the pdr16Δ mutant over 3 h. Heat-killed cells were used as a control for non-specific binding of fluconazole to the cells (Mansfield et al., 2010). Only a very small percentage (< 5%) of added radiolabelled fluconazole remains bound to heat-killed cells, indicating that the rest of the radioactivity associated with the cells represents intracellular fluconazole. The results show that both strains, pdr16Δ and wt, accumulate the same amount of radioactive fluconazole.
Overexpression of PDR17 complements the increased susceptibility of pdr16Δ cells to miconazole
Pdr16p (Sfh3p) belongs to the family of yeast phosphatidylinositol transfer proteins (Griac, 2007; Li et al., 2000). Pdr16p displays a limited homology to the founding member of this family, Sec14p, with 18% of overall sequence identity and 35% overall sequence similarity (Griac, 2007). Pdr16p has a close homologue among S. cerevisiae PITPs, Pdr17p. Pdr16p and Pdr17p are 49% identical and 75% similar (van den Hazel et al., 1999). In the following experiment, we asked whether any of the yeast Sec14-like proteins can substitute for the function of Pdr16p, as measured by the susceptibility of yeast cells toward the azole antifungal miconazole (Figure 6). In S. cerevisiae, Pdr17p is the closest homologue to Pdr16p. When Pdr17p was over-expressed from the multicopy plasmid under its own promoter, it was able to complement the most pronounced phenotype of pdr16Δ, increased susceptibility to azole inhibitors of sterol biosynthesis. To some degree, Sec14p, when over-expressed from its own promoter, was also able to complement this phenotype. Interestingly, overproduction of Pdr16p from a multicopy plasmid was not able to fulfil the function of Pdr17p as an essential part of a complex with Psd2p, which converts phosphatidylserine (PS) to phosphatidylethanolamine (PE) in endosomes (Gulshan et al., 2010; Routt et al., 2005; Wu et al., 2000). Our own experiments confirmed this observation (data not shown).
To characterize the mechanism by which overexpression of PDR17 complements the azole susceptibility of the pdr16Δ strain, we measured the relative amounts of ergosterol and lanosterol in the following strains when challenged by sub-inhibitory concentration of miconazole: (a) wt; (b) the pdr16Δ strain containing an empty cloning vector; (c) the pdr16Δ strain containing the PDR16 gene on a multicopy plasmid; and (d) the pdr16Δ strain containing the PDR17 gene on a multicopy plasmid. Both PDR16 and PDR17 genes were under the control of their own promoters. The results show that, in the absence of miconazole, the relative amounts of ergosterol and lanosterol were the same in all four strains (Figure 7A). In the presence of sub-inhibitory concentrations of miconazole, the pdr16Δ cells displayed increased accumulation of lanosterol at the expense of the final product of the sterol biosynthetic pathway, ergosterol (Figure 7B). Introduction of the PDR17 gene on a multicopy plasmid into the pdr16Δ cells resulted in reversion of the sterol profile of the pdr16Δ strain to that of a parental wt (Figure 7B).
Pdr16p is required for resistance of yeast cells to azole antifungals. This fact has been shown for S. cerevisiae (van den Hazel et al., 1999) as well as for the major human fungal pathogen C. albicans (Saidane et al., 2006). Azoles affect ergosterol biosynthesis mostly at the level of the ERG11 gene product, lanosterol 14α-demethylase, a cytochrome P450 enzyme (Lamb et al., 1999; Lupetti et al., 2002). The principal mechanisms of resistance to azoles include reduced intracellular accumulation of azoles, mostly by upregulation of multidrug transporters, and the modification of the target enzyme, lanosterol 14α-demethylase (Lupetti et al., 2002; MacCallum et al., 2010; Prasad and Goffeau, 2012). The mechanism(s) by which Pdr16p contributes to the azole resistance remains highly controversial. van den Hazel et al. (1999) suggested that the activity of enzymes of ergosterol biosynthesis are affected in the pdr16Δ strain, making it more susceptible to inhibition by azoles. This hypothesis was based mostly on the subtle changes of sterol composition of the isolated plasma membranes of the pdr16Δ strain. Ren et al. (2011) considered the role of Pdr16p in sterol biosynthesis rather speculative and predicted its role in phosphoinositide signalling (Bankaitis et al., 2010). It was not known whether loss of Pdr16p alters the function of relevant efflux pumps directly or indirectly (e.g. via alterations of membrane characteristics). Neither was it known whether influx rates of azoles are increased in the pdr16Δ strain or whether membranes are more susceptible to changes in sterol content caused by the action of azoles. In the current paper we have addressed some of these questions.
First, using sensitive radiolabelling pulse experiments in the absence of azoles, we did not identify any changes in ergosterol biosynthesis between the pdr16Δ strain and its corresponding wt (Figure 2A). Thus, it is highly unlikely that, in the absence of azoles, the activities of enzymes that play a role in ergosterol biosynthesis are affected in the pdr16Δ strain. The situation changes rather dramatically when the wt and pdr16Δ strains are challenged by sub-inhibitory concentrations of miconazole or fluconazole. In this case, the pdr16Δ strain accumulates more lanosterol and squalene at the expense of ergosterol, compared to its parental strain.
The simplest explanation of these results would be that, in the pdr16Δ strain, the intracellular levels of azoles are increased due to the changes in either drug uptake or its efflux. Using three independent experiments we show that intracellular concentrations of azoles are the same in the pdr16Δ strain and in its corresponding parental strain. First, uptake of radioactive fluconazole into de-energized cells, using an established method (Mansfield et al., 2010) (Figure 4), is the same in the pdr16Δ strain and in its corresponding parental wt strain. Second, overall intracellular fluconazole balance is the same in the pdr16Δ strain and in its corresponding parental wt strain (Figure 5). A supporting genetic experiment (Figure 3) indicates that Pdr16p mediates azole susceptibility by a mechanism independent of the major known azole efflux pumps, Yor1p, Snq2p, Pdr5p, Pdr10p, Pdr11p, Ycf1p and Pdr15p. This genetic experiment, however, does not formally rule out the indirect impact of Pdr16p on the major facilitator Flr1p or some unknown azole transporters. An experiment demonstrating the same uptake of radioactive fluconazole in the pdr16Δ strain and in its corresponding parental wt (Figure 4), together with the experiment demonstrating that the overall intracellular fluconazole balance is the same in the pdr16Δ strain and in its corresponding parental wt strain (Figure 5), provide evidence that differences in intracellular concentrations of azoles are not the cause of increased susceptibility of the pdr16Δ strain to azoles.
Pdr16p belongs to the group of phosphatidylinositol transfer proteins in yeast (Griac, 2007; Li et al., 2000). The founding member of this family is Sec14p, which facilitates the transfer of PI and PC between donor and acceptor membranes in in vitro assays, using either radiolabelled or fluorescent lipids (Bloj and Zilversmit, 1977; van Paridon et al., 1988). Sec14p is essential for yeast viability and it is required for secretion from the yeast Golgi apparatus (Bankaitis et al., 1990). It regulates: (a) PC biosynthesis via the CDP–choline pathway (Skinner et al., 1995); (b) phospholipase D-mediated PC turnover (Sreenivas et al., 1998); and (c) the levels of PI(4)P (Hama et al., 1999). It is also essential for proper localization of lipid raft proteins to the plasma membrane (Curwin et al., 2013). The mechanism of Sec14p action is still a matter of some controversy, but recent data indicate that its major function is to present PI to PI-kinases (Bankaitis et al., 2012). In contrast to Sec14p, all other members of this group, including Pdr16p, are unable to transfer PC in in vitro assays (Griac, 2007; Li et al., 2000). Notably, even the transfer of PI, facilitated by Pdr16p (or any other Sec14p homologue) in in vitro assays, is much slower and requires much more of the protein to proceed compared to Sec14p (Li et al., 2000). It seems that the Sec14p protein group in yeast is a diverse group of proteins with distinct subcellular localizations (Schnabl et al., 2003) and varied functions related to lipid metabolism, membrane trafficking and phosphoinositide-mediated signalling (Ghosh and Bankaitis, 2011). To identify the function of Pdr16p, we determined whether any of the other Sec14p group members could substitute for Pdr16p in regulating azole resistance in yeast (Figure 6). Because different Sec14p family members are expressed from a multicopy plasmid under the control of their own promoters, the expression levels of individual proteins can vary. Thus, it is difficult to conclude the extent of pdr16Δ complementation by various Sec14 homologues. We have concentrated on the pdr16Δ azole susceptibility complementation by Pdr17p. Ghaemmaghami et al. (2003) measured absolute levels of proteins in the S. cerevisiae cell through immunodetection. Based on their data, under standard conditions, a yeast cell contains similar amounts of Sec14 homologues, except for Sec14p itself, which is expressed at about 10 times higher levels. Our results show (Figure 6) that the closest homologue to Pdr16p, Pdr17p (also called Sfh4p or PstB2p), does substitute remarkably well for Pdr16p in relation to miconazole susceptibility when expressed from the multicopy plasmid using the PDR17 native promoter.
In S. cerevisiae, two PS decarboxylases can convert PS to PE (reviewed in Birner and Daum, 2003). Mitochondrially localized Psd1p is the primary route of de novo production of PE (Trotter et al., 1993). However, yeast cells lacking Psd1p can grow in the absence of exogenous ethanolamine, due to the presence of the second PS decarboxylase, Psd2p (Trotter et al., 1995; Trotter and Voelker, 1995). In eukaryotic cells, PS synthesis and PS decarboxylation occur in different subcellular locations. In the case of PS-to-PE conversion via Psd2p, Pdr17p (in the respective papers designated PstB2p) was found to be an essential component of the transport machinery for PS delivery to the subcellular location of the second yeast PS decarboxylase, Psd2p (Wu and Voelker, 2001, 2004). Other components of this docking/transport complex are phosphatidylinositol-4-kinase Stt4p (Trotter et al., 1998) and the C2 domain of Psd2p itself (Kitamura et al., 2002). A model was proposed in which a macromolecular complex between donor and acceptor membranes generates zones of close contact between membranes, through which phospholipids are transported (Voelker, 2005).
The fact that Pdr17p, with a proposed function at inter-organellar junctions, can complement for the role of Pdr16p in azole resistance leads to stimulating questions. How could a protein, Pdr17p, which is an essential part of an inter-membrane contact site for PS transport, influence sterol biosynthesis when yeast cells are challenged with azoles? Could Pdr16p play a role similar to that of Pdr17p in helping to bring membrane compartments to close contact? Does Pdr16p play a role in shuttling sterols or their intermediates via inter-membrane contact sites? Clearly, these are important questions to be addressed to understand the role of Pdr16p in azole resistance and sterol metabolism in general and require future experimentation.
One particularly interesting question is the role of Pdr16p in lipid particles. It has been shown that Pdr16p localizes to lipid particles, microsomes and to the cell periphery (Schnabl et al., 2003). Sterol biosynthesis is a complex process. Some sterol biosynthetic enzymes display dual localization to the endoplasmic reticulum and to lipid particles (Ott et al., 2005). Thus, it is tempting to speculate that a role of Pdr16p could be to assist in the migration of sterols or sterol precursors between these two organelles. However, our preliminary results (not shown) do not support this hypothesis. We compared two pdr16Δ mutants, one prepared in the yeast strain devoid of lipid particles (Sandager et al., 2002), the second prepared in its parental strain having lipid particles. Both of these pdr16Δ-containing strains were equally susceptible to azole antifungals, indicating the role of Pdr16p outside of lipid particles.
In conclusion, we show that the absence of Pdr16p increased the susceptibility of the yeast S. cerevisiae to all azoles tested. Ergosterol biosynthesis seems not to be affected in the pdr16Δ strain grown under standard conditions without azoles. However, when yeast cells are challenged with azoles, considerable differences in the inhibition of sterol biosynthetic pathways were observed between the pdr16Δ strain and its corresponding wt strain. In the presence of sub-inhibitory concentrations of azoles, the pdr16Δ strain accumulated more lanosterol and squalene at the expense of the final product, ergosterol. We further show that the increased susceptibility of the pdr16Δ strain to azoles and the enhanced changes in sterol biosynthesis upon exposure to azoles are not due to the increased intracellular concentrations of azoles in the pdr16Δ cells compared to the parental wt strain. Finally, our results suggest a hypothesis that Pdr16p could play a role in shuttling sterols or their intermediates between membranes or, alternatively, between biosynthetic enzymes or complexes. Whether the role of Pdr16p in these processes is direct, or mediated via some signalling role of Pdr16p, remains to be established.
We thank A. Goffeau (Catholic University Louvain, Belgium) and G. Daum (Technical University, Graz, Austria) for kindly providing yeast strains used in this study. This work was supported by Scientific grant agency of the Ministry of Education of the Slovak Republic and of Slovak Academy of Sciences (Grants Nos 2/0077/10 and 2/0058/11).