Amino acid substitutions at the major insertion loop of Candida albicans sterol 14alpha-demethylase are involved in fluconazole resistance

Abstract

Background: In the fungal pathogen Candida albicans, amino acid substitutions of 14alpha-demethylase (CaErg11p, CaCYP51) are associated with azole antifungals resistance. This is an area of research which is very dynamic, since the stakes concern the screening of new antifungals which circumvent resistance. The impact of amino acid substitutions on azole interaction has been postulated by homology modeling in comparison to the crystal structure of Mycobacterium tuberculosis (MT-CYP51). Modeling of amino acid residues situated between positions 428 to 459 remains difficult to explain to date, because they are in a major insertion loop specifically present in fungal species.

Methodology/principal finding: Fluconazole resistance of clinical isolates displaying Y447H and V456I novel CaErg11p substitutions confirmed in vivo in a murine model of disseminated candidiasis. Y447H and V456I implication into fluconazole resistance was then studied by site-directed mutagenesis of wild-type CaErg11p and by heterogeneously expression into the Pichia pastoris model. CLSI modified tests showed that V447H and V456I are responsible for an 8-fold increase in fluconazole MICs of P. pastoris mutants compared to the wild-type controls. Moreover, mutants showed a sustained capacity for producing ergosterol, even in the presence of fluconazole. Based on these biological results, we are the first to propose a hybrid homology structure-function model of Ca-CYP51 using 3 different homology modeling programs. The variable position of the protein insertion loop, using different liganded or non-liganded templates of recently solved CYP51 structures, suggests its inherent flexibility. Mapping of recognized azole-resistant substitutions indicated that the flexibility of this region is probably enhanced by the relatively high glycine content of the consensus.

Conclusions/significance: The results highlight the potential role of the insertion loop in azole resistance in the human pathogen C. albicans. This new data should be taken into consideration for future studies aimed at designing new antifungal agents, which circumvent azole resistance.

Figure 1. Expression of recombinant Pichia pastoris transformants.

(A): SDS-polyacrylamide gel electrophoresis analysis. P. pastoris CaErg11p transformants were induced for protein expression on BMMY medium for 72 h as described in Material and Methods. Soluble cytosolic proteins were loaded into a 10% SDS-PAGE gel and stained with Coomassie brilliant blue. (B): Western blot analysis of control and mutated CaErg11p proteins produced in P. pastoris. Cytosolic proteins were transferred to PVDF membranes and incubated with a 1∶100 dilution of a polyclonal rabbit anti-yeast Erg11p and a 1∶2000 dilution of goat-anti rabbit-HRP. Signals were visualized using supersignal west pico substrate detection reagent.

Figure 2. In vitro antifungal activity of FLC against P. pastoris transformants.

(A): P. pastoris cells transformed with WT (□), K143R (▪), Y447H (○) and V456I (•) mutants of CaErg11p were tested according to the CLSI method with some modifications. MIC values were determined as the lowest antifungal concentration giving a 50% or less reduction in the optical density at 450 nm compared to the OD of the corresponding drug-free incubation medium. (B): Susceptibilities of CaErg11p P. pastoris transformants to azole fluconazole using spot assay. Serial dilutions of each control and mutant clones were spotted onto BMMY agar plates containing different concentrations of fluconazole and incubated for 72 h at 30°C. Untreated conditions (a), 2 µg/ml FLC (b), 4 µg/ml FLC (c) and 8 µg/ml FLC (d).

Figure 3. CaErg11p activity of P. pastoris transformants in the presence of FLC. P. pastoris CaErg11p methanol-induced transformants were treated with FLC in BMMY medium for 24 h at 30°C.

Non-saponifiable lipids (sterols) were extracted as described in Material and Methods. Sterol identification was done in reference to the relative retention times and mass spectra previously reported , . Activity results were expressed as the ratio of ergosterol biosynthesis compared to the lanosterol accumulation (E/L). (A): Untreated P. pastoris clones, (B): 4 µg/ml FLC and (C): 8 µg/ml FLC (n = 4).

Figure 4. Localization of the major long insertion sequence of CaErg11p and azole-resistance substitutions mapping on CaErg11p structures obtained by homology modeling.

(A): ModWeb model with human CYP51 ketoconazole liganded (3ld6). (B): YASARA model with human CYP51 liganded with econazole (3jus) with protoporphyrin IX and econazole represented as a stick respectively in orange and yellow. (C): SwissModel with M. tuberculosis 4-phenylimidazole liganded structure as template (1e9x). (D): ModWeb model zebrafish prostacyclin synthase CYP450 8a1 free (3b98). All the models are aligned, based on their secondary structure with the N-terminal part of the beta-strand rich domain left and the alpha-helices rich domain right. The secondary structure of the models are color-coded according to their type, red: beta-stands, blue =  alpha-helices, green =  turn, cyan  =  coil, magenta  =  the long fungi specific amino acid sequence insertion. Amino acid positions that have been proven to be responsible for azole-resistance are indicated with the amino acid name of the WT CaCYP51 with yellow for substrate binding site hot-spots and magenta in the insertion fungi specific sequence. N- and C-terminal are labeled by single letters N and C.