Azole resistance by loss of function of the sterol Δ⁵,⁶-desaturase gene (ERG3) in Candida albicans does not necessarily decrease virulence

Abstract

The inactivation of ERG3, a gene encoding sterol Δ⁵,⁶-desaturase (essential for ergosterol biosynthesis), is a known mechanism of in vitro resistance to azole antifungal drugs in the human pathogen Candida albicans. ERG3 inactivation typically results in loss of filamentation and attenuated virulence in animal models of disseminated candidiasis. In this work, we identified a C. albicans clinical isolate (VSY2) with high-level resistance to azole drugs in vitro and an absence of ergosterol but normal filamentation. Sequencing of ERG3 in VSY2 revealed a double base deletion leading to a premature stop codon and thus a nonfunctional enzyme. The reversion of the double base deletion in the mutant allele (erg3-1) restored ergosterol biosynthesis and full fluconazole susceptibility in VSY2, confirming that ERG3 inactivation was the mechanism of azole resistance. Additionally, the replacement of both ERG3 alleles by erg3-1 in the wild-type strain SC5314 led to the absence of ergosterol and to fluconazole resistance without affecting filamentation. In a mouse model of disseminated candidiasis, the clinical ERG3 mutant VSY2 produced kidney fungal burdens and mouse survival comparable to those obtained with the wild-type control. Interestingly, while VSY2 was resistant to fluconazole both in vitro and in vivo, the ERG3-derived mutant of SC5314 was resistant only in vitro and was less virulent than the wild type. This suggests that VSY2 compensated for the in vivo fitness defect of ERG3 inactivation by a still unknown mechanism(s). Taken together, our results provide evidence that contrary to previous reports inactivation of ERG3 does not necessarily affect filamentation and virulence.

Fig 1

Spotting susceptibility assays of C. albicans strains on YEPD agar plates. Strains were spotted onto agar plates containing fluphenazine or fluconazole in the presence and absence of cyclosporine (A) or containing fluphenazine, cycloheximide, or hygromycin B (B) and were incubated for 2 days at 35°C.

Fig 2

(A) Analysis of efflux pump activity by comparison of R6G efflux, in relative fluorescence units (RFU), after the addition of d-glucose (filled symbols; +) or without d-glucose (open symbols; −). (B) Northern blot analysis of CDR1, CDR2, and MDR1 expression (in comparison to that of the ACT1 internal control) in C. albicans strains, with (+) or without (−) exposure to fluphenazine (for upregulation of CDR1 and CDR2) or benomyl (for upregulation of MDR1). Data points are averages for two independent measurements (error bars represent standard deviations).

Fig 3

Sterol biosynthesis pathway in C. albicans. The pathway includes major sterols [ergosta-7-enol, ergosta-7,22,24(28)-trienol, and ergosta-7,22-dienol] identified in the erg3 strains VSY2 and VSY18. Dotted lines denote multiple enzymatic steps.

Fig 4

Yeast-to-hypha conversion in C. albicans strains CAF2-1, DSY1751, and VSY2 after 3-h incubations in RPMI 1640 without serum or supplemented with 10% (vol/vol) FBS.

Fig 5

C. albicans kidney burdens in a murine model of disseminated infection. Immunocompetent mice (n = 2 × 6) were infected with 1 × 10⁵ cells of each strain and treated (+F) with intraperitoneal injections of fluconazole at 200 mg/kg/day. Mice were sacrificed at day 3 postinfection and both kidneys collected. Results represent CFU counts per gram of kidney, with geometric means indicated by horizontal bars. Strains are identified by the relevant ERG3 genotype. Circles indicate the SC5314 strain background, and squares indicate the VSY2 strain background, while closed and empty symbols indicate untreated and fluconazole-treated animals, respectively. Statistical analyses were performed using the Kruskal-Wallis nonparametric test (P < 0.0001), followed by pairwise comparisons in PASW Statistics 18.

Fig 6

Survival curves for mice infected with the tested C. albicans strains. Immunocompetent mice (n = 7) were infected with 5 × 10⁵ cells of each strain and then monitored daily until day 10. Strains are identified by the relevant ERG3 genotype, while the strain backgrounds SC5314 and VSY2 are indicated by filled and empty symbols, respectively. Statistical analyses were performed using the log rank test to compare survival curves in GraphPad Prism 5.0.