Identification and Characterization of Four Azole

Institute of Life Science and School of Medicine, Swansea University, Swansea, SA2 8PP, Wales, United Kingdom,1 University Medical Center Göttingen, Institute for Medical Microbiology and German National Reference Center for Systemic Mycoses, Göttingen, Germany2

Institute of Life Science and School of Medicine, Swansea University, Swansea, SA2 8PP, Wales, United Kingdom,1 University Medical Center Göttingen, Institute for Medical Microbiology and German National Reference Center for Systemic Mycoses, Göttingen, Germany2

Institute of Life Science and School of Medicine, Swansea University, Swansea, SA2 8PP, Wales, United Kingdom,1 University Medical Center Göttingen, Institute for Medical Microbiology and German National Reference Center for Systemic Mycoses, Göttingen, Germany2

Institute of Life Science and School of Medicine, Swansea University, Swansea, SA2 8PP, Wales, United Kingdom,1 University Medical Center Göttingen, Institute for Medical Microbiology and German National Reference Center for Systemic Mycoses, Göttingen, Germany2

Institute of Life Science and School of Medicine, Swansea University, Swansea, SA2 8PP, Wales, United Kingdom,1 University Medical Center Göttingen, Institute for Medical Microbiology and German National Reference Center for Systemic Mycoses, Göttingen, Germany2

Institute of Life Science and School of Medicine, Swansea University, Swansea, SA2 8PP, Wales, United Kingdom,1 University Medical Center Göttingen, Institute for Medical Microbiology and German National Reference Center for Systemic Mycoses, Göttingen, Germany2

Institute of Life Science and School of Medicine, Swansea University, Swansea, SA2 8PP, Wales, United Kingdom,1 University Medical Center Göttingen, Institute for Medical Microbiology and German National Reference Center for Systemic Mycoses, Göttingen, Germany2

Institute of Life Science and School of Medicine, Swansea University, Swansea, SA2 8PP, Wales, United Kingdom,1 University Medical Center Göttingen, Institute for Medical Microbiology and German National Reference Center for Systemic Mycoses, Göttingen, Germany2

Institute of Life Science and School of Medicine, Swansea University, Swansea, SA2 8PP, Wales, United Kingdom,1 University Medical Center Göttingen, Institute for Medical Microbiology and German National Reference Center for Systemic Mycoses, Göttingen, Germany2

Sterol analysis identified four Candida albicans erg3 mutants in which ergosta 7,22-dienol, indicative of perturbations in sterol Δ5,6-desaturase (Erg3p) activity, comprised >5% of the total sterol fraction. The erg3 mutants (CA12, CA488, CA490, and CA1008) were all resistant to fluconazole, voriconazole, itraconazole, ketoconazole, and clotrimazole under standard CLSI assay conditions (MIC values, ≥256, 16, 16, 8, and 1 μg ml−1, respectively). Importantly, CA12 and CA1008 retained an azole-resistant phenotype even when assayed in the presence of FK506, a multidrug efflux inhibitor. Conversely, CA488, CA490, and three comparator isolates (CA6, CA14, and CA177, in which ergosterol comprised >80% of the total sterol fraction and ergosta 7,22-dienol was undetectable) all displayed azole-sensitive phenotypes under efflux-inhibited assay conditions. Owing to their ergosterol content, CA6, CA14, and CA177 were highly sensitive to amphotericin B (MIC values, Open in a separate windowFIG. 1.

Schematic representation of the ergosterol biosynthetic pathway in C. albicans. (A) Sterol intermediates that accumulate with perturbation of sterol Δ5,6-desaturase (Erg3p) function but without azole inhibition of 14α-demethylase (Erg11p) activity. (B) Sterol intermediates that accumulate with azole inhibition of Erg11p. Here, perturbation of Erg3p function circumvents build-up of the fungistatic sterol 14α-methylergosta-8,24(28)-dien-3β,6α-diol. Note that ordinarily, eburicol (not lanosterol) is the substrate for C. albicans Erg11p. Open arrow, sterol Δ5,6-desaturase (ERG3) step; broken arrows, multiple enzymatic steps; solid arrows, single enzymatic step.

Traditionally, polyene agents (e.g., amphotericin B [AMB]) that intercalate directly with ergosterol, resulting in membrane leakage and cell death (4, 25), have provided a treatment alternative for clinicians who encounter azole-resistant candidiasis. However, a molecular mechanism, first described in Saccharomyces cerevisiae (34, 35), that enables fungal cells to circumvent the inhibitory action of polyenes and azoles through erg3 mutation has also been reported in C. albicans (13, 14, 24). Briefly, ERG3 mutations that nullify the function of sterol Δ5,6-desaturase result in accumulation of 14α-methylfecosterol rather than 14α-methylergosta-8,24(28)-dien-3β,6α-diol following azole inhibition of Erg11p (11). This, together with a lack of membrane ergosterol, enables C. albicans erg3 mutants to withstand both azole and polyene treatment. Recently, studies have questioned the clinical relevance of mutations in ERG3 (21), with work in mouse models leading to the conclusion that loss of Erg3p activity attenuates virulence in C. albicans (3) and that this could explain why erg3 mutants are rarely seen in the clinic. However, it is arguably the undetected erg3 mutants that do persist in the clinical setting that warrant greater consideration.

Here, we identify four clinically isolated C. albicans erg3 mutants using gas chromatography-mass spectrometry (GC-MS) sterol analysis. Unlike research studies that focus on specific genes (e.g., ERG11) involved in ergosterol biosynthesis (Fig. ​(Fig.1),1), sterol analysis can identify perturbations in the functionality of many ergosterol biosynthetic proteins (here, we focus on sterol Δ5,6-desaturase [Erg3p]) that are expressed in clinical isolates in vivo. Our results indicate that strains harboring mutations in ERG3 may be more prevalent than currently supposed and that, in addition to mutants exhibiting total loss of Erg3p activity, leaky mutants should also be considered.

The isolates (CA6, CA12, CA14, CA177, CA488, CA490, and CA1008) described in this study were identified from a collection of >100 clinical specimens (O. Bader, M. Kuhns, C. M. Martel, J. E. Parker, K. Tintelnot, M. Seibold, D. Sanglard, D. E. Kelly, S. L. Kelly, M. Weig, and U. Gross, unpublished data) established as part of an EU FP6 research initiative (European Resistance Fungal Network [EURESFUN]). All were routinely maintained at 37°C on yeast extract-peptone-dextrose (YEPD) agar containing (wt/vol) 2% glucose, 2% Bacto peptone, 1% yeast extract, and 2% agar (Difco). RPMI 1640 medium (Sigma) buffered with 0.165 M MOPS (morpholinepropanesulfonic acid) (pH 7) was used to culture isolates for sterol analyses and antifungal susceptibility testing.

Single colonies were used to inoculate 15-ml volumes of RPMI to achieve overnight cultures (18 h; 37°C; 180 rpm) that all exhibited hyphal growth morphology. Cells were harvested by centrifugation and washed with sterile water prior to sterol extraction using a previously optimized methodology (19). In addition to the analysis of untreated RPMI-grown cultures, the sterol composition of all isolates, following treatment with final concentrations of fluconazole (FLC) and voriconazole (VOR) equivalent to half the minimum concentration required to inhibit growth (0.5× MIC), were also determined. Sterol extracts were evaporated to dryness using a centrifugal evaporator (Heto Maxi Dry Plus) and derivatized [N,O-bis(trimethylsiyl)trifluoroacetamide-trimethylchlorosilane (90:10) plus 50 μl anhydrous pyridine] at 70°C for 2 h. Trimethylsilyl (TMS)-derivatized sterols were analyzed using a 7890A GC system (Agilent Technologies) and identified with reference to retention times and fragmentation spectra for known standards. A DB-5MS fused silica column (30 m by 0.25 mm by 0.25-μm film thickness) was used for all GC separations (J&W Scientific). The initial oven temperature was held at 70°C for 4 min, followed by ramping (25°C/min) to a final temperature of 280°C; this temperature was held for 25 min. Samples were analyzed in splitless mode (1 μl injection volume) using helium carrier gas and electron impact ionization (ion source temperature, 150°C) and scanning from 40 to 850 atomic mass units. GC-MS data files were analyzed using Agilent software (MSD Enhanced ChemStation) to generate sterol profiles and for derivation of integrated peak areas.

The susceptibilities of CA6, CA12, CA14, CA177, CA488, CA490, and CA1008 to FLC, VOR, itraconazole (ITC), ketoconazole (KTC), clotrimazole (CLT), and AMB were determined using the standardized CLSI M27-A2 broth dilution method (23) and previous interpretive breakpoints (19): (i) FLC, ≤8 μg ml−1, sensitive, and ≥64 μg ml−1, resistant; (ii) VOR, ≤1 μg ml−1, sensitive, and ≥4 μg ml−1, resistant; (iii) ITC and KTC, ≤0.125 μg ml−1, sensitive, and ≥1 μg ml−1, resistant; (iv) CLT, ≥0.5 μg ml−1, resistant; and (v) AMB, ≤1 μg ml−1, sensitive. Assays for susceptibility to FLC and VOR were also performed in the presence of a fixed concentration (10 μM) of the putative multidrug efflux inhibitor FK506 (19). Azole MIC's were determined as the minimum drug concentration yielding at least 80% inhibition of growth compared with the growth of control wells. The MIC for AMB was defined as the lowest drug concentration at which growth was completely inhibited. All assays were performed in triplicate.

The full-length ERG3 and ERG11 genes were amplified from genomic DNA (single-colony extraction: 0.2% SDS; 90°C; 10 min) using gene-specific forward (F) and reverse (R) primers: ERG3F, 5′-GATCATAACTCAATATGG-3′; ERG3R, 5′-CTGAACACTGAATCG-3′; ERG11F, 5′-ATGGATATCGTACTAGAA-3′; ERG11R, 5′-TCATTGTTCAACATATTC-3′. DNA reads were translated into amino acid sequences and aligned with C. albicans ERG3 and ERG11 reference proteins (UniProtKB/Swiss-Prot accession no. {"type":"entrez-protein","attrs":{"text":"O93875","term_id":"51701379","term_text":"O93875"}}O93875 and {"type":"entrez-protein","attrs":{"text":"P10613","term_id":"1169073","term_text":"P10613"}}P10613, respectively).

The morphologies of all isolates were examined during growth on YEPD agar and 6 or 24 h following transfer of single colonies to liquid RPMI (37°C; 180 rpm).

Untreated CA6, CA14, and CA177 (Table ​(Table1,1, untreated) all exhibited normal sterol profiles in which ergosterol comprised >80% of the total cellular sterol fraction. No sterol intermediates indicative of perturbations in sterol Δ5,6-desaturase (Erg3p) function (Fig. ​(Fig.1A)1A) were detected in any of these isolates. Conversely, while ergosterol was also detected in untreated CA12, CA488, and CA490, episterol, ergosta 7-enol, and ergosta 7,22-dienol comprised over 40% of the total sterol fraction (Table ​(Table1,1, untreated). In CA1008, ergosta 7,22-dienol comprised >70% of the sterol fraction with ergosterol never comprising >2% of the total. Neither 14α-methyl fecosterol nor 14α-methylergosta-8,24(28)-dien-3β,6α-diol (here called 14α-methyl-3,6-diol) were detected in any sterol preparations from untreated isolates.

Typical sterol compositions of untreated and fluconazole- or voriconazole-treated C. albicans isolates

Typical GC-MS chromatograms for CA177 (A) and CA1008 (B) without treatment (boldface traces) or following azole treatment (thin traces). 1, ergosterol; 2, 14α-methylfecosterol; 3, 14α-methylergosta-8,24(28)-dien-3β,6α-diol; 4, lanosterol and/or obtusifoliol; 5, eburicol; 6, ergosta 8,22-dienol; 7, ergosta 7,22-dienol; 8, episterol; 9, ergosta 7-enol.

MIC data for study isolates

CA6, CA12, CA177, CA488, CA490, and CA1008 were resistant to azole antifungals under standard assay conditions (MICS, ≥256, 16, 16, 8, and 1 μg ml−1 for FLC, VOR, ITC, KTC, and CLT, respectively); CA14 was the only isolate found to be sensitive to all azoles (Table ​(Table2,2, standard azole MIC assay). When assayed in the presence of 10 μM FK506, CA6, CA14, CA177, CA488, and CA490 all reverted to an azole-sensitive phenotype. Conversely, CA12 and CA1008 retained an azole-resistant phenotype even under efflux-inhibited assay conditions (Table ​(Table2,2, FK506B). CA6, CA14, and CA177 were all equally susceptible to AMB (MICS, TABLE 3.

Amino acid substitutions identified in ERG3 and ERG11 protein translations for C. albicans study isolates

There were no clear differences between the cell morphologies of any of the study isolates following culture at 37°C (Fig. ​(Fig.3).3). All exhibited yeast-like growth on YEPD agar (Fig. ​(Fig.3A).3A). Following transfer to RPMI, pseudohyphae were observed in all cultures after 6 h (Fig. ​(Fig.3B),3B), with extensive hyphal growth seen after 24 h (Fig. ​(Fig.3C3C).

Typical growth morphologies of CA6 and CA1008 (in which ergosterol comprised >80% and 70% of the total sterol fraction) and low abundance of ergosterol (256 and >16 μg ml−1 of each azole, respectively), even following inhibition of multidrug efflux transporters (Table ​(Table2,2, 10 μM FK506), indicated the underlying importance of their erg3 mutant phenotype. It is noteworthy that a polyene-resistant C. albicans erg5 mutant in which cellular ergosterol was not detectable has recently been reported (19). The lack of cellular ergosterol in CA12 and CA1008 provides some explanation as to why both were able to tolerate relatively high concentrations of AMB (MICs, 0.5 and 1.0 μg ml−1, respectively).

It has been shown that leaky erg3 mutants of S. cerevisiae can be isolated as fluconazole-resistant strains (9). Sterol chromatograms for CA488 and CA490 (profiles not shown) indicate that they might be regarded as leaky erg3 mutants in which sterol Δ5,6-desaturase retains greater activity and thus ergosterol (and 14α-methyl-3,6-diol following azole inhibition of Erg11p) is detectable in higher abundance (Table ​(Table1,1, FLC and VOR treated). As in S. cerevisiae, it is possible that, owing to slight perturbation of Erg3p activity, accumulation of 14α-methylfecosterol enhances the ability of CA488 and CA490 to survive under azole treatment. However, the results from this study indicate that (like CA6 and CA177), drug efflux mechanisms (29, 31) also contribute to the azole resistance phenotypes of CA488 and CA490. Given its azole-sensitive phenotype and a lack of evidence for significant drug efflux in CA14 (Table ​(Table2,2, FK506), the results suggest that better understanding of the molecular bases of drug efflux transport in C. albicans could assist in the identification of new drug targets and thus in the development of novel antifungal agents.

Sequence data for the ERG3 and ERG11 genes in CA1008 and CA12 challenge the suggestion that mutations in ERG3 cannot cause resistance by themselves (21). While five amino acid substitutions were identified in CA1008 Erg3p, only a single-residue change (E266D) was recorded in the Erg11p sequence. This substitution is a conserved polymorphism in the G helix of C. albicans sterol 14α-demethylase (2) and was also seen in CA488 and CA490 (this study). E266D may exist in susceptible strains and is currently not understood to confer resistance (16, 17, 39). The fact that no amino acid substitutions were identified in the Erg11p sequence for CA12, the only isolate to harbor a single substitution (W332R) in Erg3p (Table ​(Table3),3), underscores this point. It is possible that the tryptophan residue at position 332 in CA12 is important for Erg3p function; modeling and site-directed mutagenesis studies are now required to enhance understanding of Erg3p structure and catalytic domains.

Sequence data for CA488 and CA490 indicate that both harbor multiple-residue changes in Erg3p, including A351V (also seen in CA1008 Erg3p). This variation has been reported previously in a C. albicans wild-type strain, B311 (22), and thus may not have a deleterious effect on Erg3p function. The additional Erg3p changes in CA488 and CA490 (H243N, T330A, and D147G) were not seen in any other isolates and now warrant further investigation. It should be noted that although the Erg3p substitutions observed in CA488 and CA490 do not result in classical erg3 mutant phenotypes, there is still value in indentifying both missense and nonsense mutations in ERG3, not least because of the potential for development of genetic screening tools that could help clinicians diagnose (and thus treat) problem C. albicans infections more readily. In this respect, it is also significant that several of the Erg11p substitutions observed in CA488 and CA490, e.g., E391G (32), V488I (17), and T229A (18), have been documented previously; however, their roles in resistance are not fully understood. Of the novel Erg11p amino acid substitutions reported here, F72S and N440S occur at positions where residue changes have also been seen previously: F72L (5) and N440K (33). The investigation of these substitutions and the previously undocumented R523G (CA490 Erg11p) constitutes a potential avenue for future research.

Aside from genotypic and sterol data, it is noteworthy that morphological examination of all of the isolates revealed no obvious differences in their abilities to form hyphae (Fig. ​(Fig.3).3). The ability to switch between yeast-like and hyphal growth is understood to be a potential virulence factor in C. albicans (30). Studies have reported that C. albicans erg3 mutants are impaired in the ability to form filaments (3, 21), possibly because they lack the ergosterol-rich rafts that are associated with the leading edges of developing hyphae (20). In this study, the extensive hyphal growth of all isolates (particularly CA1008) suggested that it is possible to maintain this pathogenicity factor while exhibiting defective Erg3p activity. The fact that the evolution of the ergosterol pathway conferred a selective advantage is reflected in the increased survival of wild-type strains over erg mutants when exposed to osmotic stress. The pathway of sterol biosynthesis most likely evolved with the addition of sequential steps, providing an evolutionary advantage to the ancestor of fungi. Hence, molecular properties, such as retention of a 14α-methyl group in the first sterol, lanosterol, is not necessarily inconsistent with growth (as for 14α-methylfecosterol), and this has been demonstrated in S. cerevisiae (7). The existence of erg3 mutants such as those reported here presumably does reflect selective pressure, most likely exposure to antifungal drugs.

Research has investigated how sterol modifications, specifically those in erg3 mutants, might mechanically influence the membrane properties and viability of fungal cells (1). Of additional interest are findings that suggest cells adjust their membrane compositions in response to sterol perturbations by preferentially changing the sphingolipid composition (8). Further studies of the membrane properties and sterol-lipid and drug-sterol interactions that might contribute to resistance in C. albicans must now be undertaken.

This research was supported by the EU FP6 project EURESFUN.

Analytical facilities were provided by the EPSRC National Mass Spectrometry Service Centre (Swansea University, Swansea, United Kingdom). Strains CA12 and CA1008 were obtained though the German National Reference Center for Systemic Mycoses. Strains CA488 and CA490 were kindly contributed by E. Mellado (ISCII, Madrid, Spain) to the EURESFUN strain collection.

▿Published ahead of print on 23 August 2010.