Analysis of the Contribution of cyp51 Genes to Azole Resistance in Aspergillus Section Nigri with the CRISPR-Cas9 Technique

Cyp51 contribution to azole resistance has been broadly studied and characterized in Aspergillus fumigatus, whereas it remains poorly investigated in other clinically relevant species of the genus, such as those of section Nigri. In this work, we aimed to analyze the impact of cyp51 genes (cyp51A and cyp51B) on the voriconazole (VRC) response and resistance of Aspergillus niger and Aspergillus tubingensis.

ABSTRACT

Cyp51 contribution to azole resistance has been broadly studied and characterized in Aspergillus fumigatus, whereas it remains poorly investigated in other clinically relevant species of the genus, such as those of section Nigri. In this work, we aimed to analyze the impact of cyp51 genes (cyp51A and cyp51B) on the voriconazole (VRC) response and resistance of Aspergillus niger and Aspergillus tubingensis. We generated CRISPR-Cas9 cyp51A and cyp51B knockout mutants from strains with different genetic backgrounds and diverse patterns of azole susceptibility. Single-gene deletions of cyp51 genes resulted in 2- to 16-fold decreases of the VRC MIC values, which were below the VRC epidemiological cutoff value (ECV) established by the Clinical and Laboratory Standards Institute (CLSI), irrespective of their parental strains’ susceptibilities. Gene expression studies in the tested species confirmed that cyp51A participates more actively than cyp51B in the transcriptional response of azole stress. However, ergosterol quantification revealed that both enzymes comparably impact the total ergosterol content within the cell, as basal- and VRC-induced changes to ergosterol content were similar in all cases. These data contribute to our understanding of Aspergillus azole resistance, especially in non-A. fumigatus species.

INTRODUCTION

Sterol 14α-demethylase (Cyp51) enzymes are included in the cytochrome P450 protein superfamily, which is present in all biological kingdoms. In fungi, these enzymes display essential functions within the ergosterol biosynthetic pathway since they catalyze the oxidative removal of 14α-methyl groups from sterol precursors to be transformed into ergosterol. In turn, ergosterol constitutes an essential component for the permeability and fluidity of fungal membranes, and it is involved in several fungal regulatory and developmental processes (1, 2). Inhibition of Cyp51 enzymes, such as the effect exerted by azole antifungals, blocks ergosterol synthesis by impairing the demethylation of its precursors. This results in an accumulation of methylated sterols within the cell that affects the fluidity of the fungal membrane and leads to its disruption and the inhibition of cell proliferation (3–6).

In Aspergillus, Cyp51 enzymes are encoded by the cyp51 genes, and, interestingly, different species of the genus display various numbers of cyp51 paralogs in their genomes. For example, Aspergillus fumigatus and Aspergillus niger contain two cyp51 copies (cyp51A and cyp51B), while Aspergillus flavus displays three different paralogs (cyp51A, cyp51B, and cyp51C) (6, 7). In A. fumigatus, this genetic redundancy is translated into two Cyp51 proteins that carry equivalent enzymatic functions and share the same substrates. However, while Cyp51A seems to encode the predominant enzymatic activity and displays a higher influence on azole stress response, Cyp51B has been considered a redundant gene that might carry out alternative functions (8, 9). Azole resistance in this species has been broadly linked to overexpression or nonsynonymous mutations of the cyp51A gene (10, 11), which contrasts to the unclear role of cyp51B in azole response (12, 13).

Despite the extensive data obtained from azole resistance analyses in A. fumigatus, the potential contribution of every Cyp51 paralog to azole response and/or resistance remains poorly explored and obscure in non-A. fumigatus species. This is the case of clinically important species of Aspergillus section Nigri (i.e., A. niger and Aspergillus tubingensis), which constitute the second most frequent agents causing invasive aspergillosis (IA) (7, 14, 15).

In previous work (16), we explored the role of Aspergillus section Nigri Cyp51 enzymes in azole resistance, which could not be clearly explained by Cyp51A protein alteration or by cyp51 gene upregulation, contrary to what has been described in A. fumigatus (10, 11). Therefore, the aim of the present study was to elucidate the role of A. niger and A. tubingensis cyp51 paralogs in azole response and resistance. With this purpose, CRISPR-Cas9 knockout (KO) mutants from strains with different genetic backgrounds and diverse azole susceptibility were generated. In addition, KOs were characterized in terms of susceptibility, gene expression, and ergosterol biosynthesis under standard conditions and azole stress.

RESULTS

Mutant generation with CRISPR-Cas9.

A total of 10 cyp51 KO (Δcyp51A or Δcyp51B) strains from five parental isolates with diverse voriconazole (VRC) susceptibility (FMRs 15441, 14712, 14635, 11906, and 15388) (16) were generated using a CRISPR-Cas9-driven gene-targeting technique (17). After selection with hygromycin (Hyg), mutant confirmation was carried out with three different genomic DNA (gDNA) PCRs (Fig. 1) combining four primers for every species and gene as follows: Fw scr-Rev scr, Fw scr-Rev intern, and Fw intern-Rev scr.

FIG 1.

cyp51 deletion strategy using a CRISPR-Cas9 gene-targeting technique and a repair template with microhomology regions to the target genes. Mutant confirmation was carried out by three different gDNA PCRs using four primers.

For the A. niger Δcyp51A confirmation PCRs, the resulting band sizes were 4,614 bp, 1,814 bp, and 2,800 bp, whereas, in the Δcyp51B case, band sizes were 4,710 bp, 1,844 bp, and 2,866 bp, respectively. In the case of A. tubingensis Δcyp51A PCRs, bands showed sizes of 4,557 bp, 1,844 bp, and 2,713 bp, while Δcyp51B PCR fragments showed 4,502 bp, 1,777 bp, and 2,725 bp, respectively.

Antifungal susceptibility.

The VRC, posaconazole (PSC), and itraconazole (ITC) MICs determined in the 10 KO strains (Δcyp51A or Δcyp51B) generated from section Nigri are shown in Table 1. In general, the two single-gene deletions impacted VRC susceptibility by notably decreasing its MICs. Moreover, both single-gene deletions resulted in MICs below the VRC ECV established by the CLSI (2 μg/ml) (18) regardless of the MICs in the parental strains. Specifically, deletion of the cyp51A gene showed a 16-fold decrease in the VRC MIC value for the strain initially displaying the highest VRC MIC (FMR 15441), while a 2-fold decrease in the VRC MIC was observed for the FMR 15388 strain, which originally showed the lowest MIC of the set. The other cyp51A KO mutants also exhibited an increase in VRC susceptibility, being 4 times (FMRs 14712 and 11906) and 8 times (FMR 14635) more susceptible than their parental strains. Regarding the effects of the cyp51B deletion on VRC susceptibility, MICs were reduced in all strains but FMR 15388, in which VRC susceptibility was not altered in comparison to its parental strain. Again, the strain initially displaying the highest VRC MIC (FMR 15441) suffered the most marked MIC reduction, with an 8-fold decrease in the VRC MIC. With respect to the other cyp51B KOs, VRC MICs were 2-fold (FMRs 14712 and 11906) and 4-fold (FMR 14635) lower than the MIC values of the parental strains. In relation to the other two azoles tested, similar effects were noted after the single-gene deletions, with exceptions to this tendency for strains FMRs 14635, 14712, and 11906 in the case of PSC and FMRs 15441 and 15388 in the case of ITC. In these cases, deletions did not affect azole MICs.

TABLE 1.

Antifungal susceptibility testing results^a

Species	Parental strain	MIC (μg/ml) of:
		VRC	Δcyp51A and VRC	Δcyp51B and VRC	PSC	Δcyp51A and PSC	Δcyp51B and PSC	ITC	Δcyp51A and ITC	Δcyp51B and ITC
A. niger	FMR 15441	8	0.5	1	0.5	0.25	0.5	1	1	1
A. tubingensis	FMR 14635	4	0.5	1	0.12	0.12	0.12	4	1	2
A. tubingensis	FMR 14712	4	1	2	0.25	0.25	0.25	16	1	4
A. tubingensis	FMR 11906	2	0.5	1	0.12	0.12	0.12	2	0.5	2
A. niger	FMR 15388	1	0.5	1	0.5	0.25	0.5	1	1	1

^aParental strains MICs were determined in a previous study (16). Bold represents MICs above the ECVs established by the CLSI.

Gene expression analysis of cyp51A and cyp51B in azole-inducing conditions.

Transcription levels of cyp51A and cyp51B were determined in the 10 KO mutant strains by quantitative reverse transcription-PCR.

Results from cyp51A gene expression (in Δcyp51B strains) showed fluctuations in the cyp51A transcription among strains, ranging from 2- to 9-fold with respect to the housekeeping gene (Fig. 2A). VRC exposure resulted in a noticeable and statistically significant cyp51A expression induction (P = 0.0002) of this gene in strains FMRs 15441 ΔB (4-fold increase), 14712 ΔB (15-fold increase), 11906 ΔB (3-fold increase), and 15388 ΔB (3-fold increase).

FIG 2.

Gene expression analysis results of the cyp51A gene in Δcyp51B strains (A) and the cyp51B gene in Δcyp51A strains (B). White bars, results of the control condition; black bars, results after exposure to 4 μg/ml VRC for 8 h. The actA gene was used in all cases as the housekeeping gene to normalize transcription levels. Statistical significance is marked by asterisks (*, P ≤ 0.05; ***, P ≤ 0.001).

Regarding results on cyp51B expression (in Δcyp51A strains), transcription levels ranged from 0.9- to 3-fold with respect to the housekeeping gene (Fig. 2B). In contrast to cyp51A observations, transcription of cyp51B upon VRC exposure was slightly affected, with a statistically significant increase in strains FMR 15441 ΔA (P = 0.0002) and FMR 14635 ΔA (P = 0.0379).

Ergosterol content.

Percentages of ergosterol and the intermediate sterol 24(28) dehydroergosterol (DHE) dry weight were determined for all Δcyp51A- and Δcyp51B-generated mutants in standard conditions and upon VRC exposure stress (Fig. 3). In general, basal ergosterol content in Δcyp51A mutants showed minimal differences since it ranged from 0.031 to 0.035% (dry weight). In the case of Δcyp51B, the ergosterol level range was slightly broader, ranging from 0.026 to 0.040% (dry weight). Comparison of Δcyp51A and Δcyp51B ergosterol basal quantification revealed no general differences, with only one (FMR 11906 ΔA and 11906 ΔB) significant exception (P = 0.0286). In addition, ergosterol was significantly depleted in all cases after VRC exposure, ranging from 0.009 to 0.016% dry weight within all strains (P ≤ 0.0286).

FIG 3.

Quantification of ergosterol (A) and the intermediate sterol 24(28)DHE (B) expressed as percentage of the mycelia dry weight. White bars, results of the control condition; black bars, results after exposure to 4 μg/ml VRC for 8 h. Statistical significance (P ≤ 0.05) is marked by an asterisk.

Regarding 24(28)DHE quantification, basal levels were very similar among all Δcyp51A (0.022 to 0.026% dry weight) and all Δcyp51B (0.019 to 0.022%) mutants. Interestingly, exposure to VRC resulted in very small variations compared to the control condition, except for the reduction observed for strains FMRs 15541 ΔA and 11906 ΔA or the increase displayed by FMR 14712 ΔB.

DISCUSSION

In the current guidelines for aspergillosis management, azoles and, more specifically, VRC, are established as the first-line therapeutic and prophylactic treatments for most infections caused by Aspergillus (19). Nevertheless, azole resistance is continuously rising due to the wide use of azole fungicides in the environment (for plant and animal protection as well as food production) and in long-term human therapies (20, 21). This phenomenon has shown an important impact in the clinical management of Aspergillus infections, especially impairing IA patient outcomes. Mortality rates linked to therapeutic failure in IA caused by resistant A. fumigatus isolates have been estimated to be around 90%, while estimation of death rates in IA caused by strains with wild-type susceptibilities is significantly lower (30 to 50%) (22).

Although A. fumigatus is the most commonly isolated species from IA patients, other species, such as section Nigri members, can also cause disseminated and invasive infections (7, 14). Interestingly, A. niger and related species have been described to show lower azole susceptibility than A. fumigatus (23), which, together with azole resistance mechanisms, might potentially impact clinical management and patient outcomes.

In a previous study, we analyzed azole resistance in a set of section Nigri strains with different azole susceptibility patterns. Investigation on cyp51 expression and cyp51A nonsynonymous mutations revealed no clear association to azole resistance within these species (16).

On this basis, the current work is focused on the generation, for the first time to our knowledge, of single defective cyp51A and cyp51B mutants in section Nigri strains with different genetic backgrounds and diverse VRC susceptibility to assess the real contribution of each cyp51 paralog into azole response and resistance since the deletion of both genes is lethal for the fungus. With that purpose, a CRISPR-Cas9-based methodology recently developed for A. fumigatus (17) was successfully used.

Results with the cyp51 KOs surprisingly show that VRC susceptibility is similarly contributed by the two cyp51 paralogs (cyp51A and cyp51B) in A. niger and A. tubingensis since the two single deletions reduced VRC MICs below the ECV established by the CLSI (18), irrespective of the parental strains’ susceptibility. Interestingly, VRC susceptibility in Δcyp51A strains was slightly higher than in the Δcyp51B strains. These results are in line with those obtained from A. fumigatus in a very recent study, where deletion of either cyp51 paralog resulted in increased susceptibility, with a greater effect on VRC MICs after deletion of cyp51A (24). Curiously, previous A. fumigatus analyses showed that cyp51A deletion or silencing by small interfering RNA (siRNA) increased azole susceptibility (9, 25), while deletion of cyp51B had no effects (8, 26). Further studies to elucidate the role of every paralog will be needed to resolve the conflicting results on this species.

In addition, our data are in agreement with that obtained from A. flavus since azole response seems to rely on cyp51A and cyp51B in that species (27), as it has been described to occur in Fusarium oxysporum (28). In the latter, cyp51A deletion clearly reduced azole MICs, while cyp51B deletion increased azole susceptibility to a lesser extent. However, it is worth noting that these two species contain a third cyp51 paralog (cyp51C), and its contribution to azole response in both cases seems negligible (27). In other filamentous fungi, such as Rhizopus oryzae (also containing the cyp51B paralog), azole resistance was similarly attributed to cyp51A (29).

Altogether, these data suggest that every cyp51 paralog has different levels of involvement in azole response within the Aspergillus genus and other filamentous fungi. In this sense, assumption of analogous cyp51 contribution toward azole response within the same genus or group of fungi is not supported by the current evidence.

Furthermore, regarding expression results, no mechanism of transcriptional compensation appeared to occur when deleting either of the cyp51 genes in the tested species of section Nigri. This is similar to what occurs in A. fumigatus since transcriptional response of cyp51B after cyp51A deletion in this species remains unaltered (9). However, higher expression of cyp51A (in Δcyp51B strains) was generally found than cyp51B expression (in Δcyp51A strains) in the KO mutants. Moreover, minor expression differences were observed among cyp51 basal expression of all strains tested here, as it had been described in wild-type strains (16, 30), which reinforces the idea of no cyp51 transcriptional compensation in section Nigri, even after deleting one of the two paralogs.

Although cyp51 expression of the parental strains was analyzed and discussed in a previous study (16), VRC exposure in the KOs generated here resulted in a clear upregulation of cyp51A expression, while that of cyp51B was mildly affected by VRC. This represented a very similar tendency to the expression fluctuations observed in the wild-type strains (16). A greater impact on azole stress response of cyp51A than cyp51B has been suggested to occur in A. fumigatus by others (4), which could be supported to occur in section Nigri by our expression results, from both these data and those previously obtained with the parental strains (16). However, the basal ergosterol content within the cells did not differ between section Nigri Δcyp51A and Δcyp51B mutants. This suggests a similar contribution of both enzymes to the ergosterol biosynthetic pathway in these species. Ergosterol was significantly and similarly depleted upon VRC exposure in all cases. Additionally, the intermediate sterol 24(28)DHE content was analyzed in the KO mutants since this sterol can replace ergosterol due to their structural and chemical resemblance (31, 32). As similar amounts of this sterol were found in all strains in addition to its small alteration upon VRC exposure, it seems that no evident compensatory effect of this sterol exists in section Nigri, even after deleting one of the two cyp51 paralogs.

In summary, we have generated, for the first time, defective cyp51 mutants in section Nigri strains with different genetic and VRC susceptibility backgrounds. This approach has allowed the investigation of azole response in a realistic and reliable manner, and thus, the present study might be useful to clearly decipher the cyp51 role in azole stress, especially in Aspergillus section Nigri.

This study contributes new evidence to the understanding of azole resistance in non-A. fumigatus species of Aspergillus. In this context, we could not clearly correlate cyp51 expression, inducibility by VRC, or ergosterol content observations to initial VRC MICs displayed by the isolates tested, which reinforces the multifactorial nature and the molecular complexity of azole resistance in section Nigri. Therefore, further studies of the ergosterol biosynthetic pathway and the characterization of other potential resistance mechanisms in Aspergillus are essential to really establish its molecular basis.

MATERIALS AND METHODS

Strains, media, and growth conditions.

Five strains belonging to species A. niger and A. tubingensis of Aspergillus section Nigri (FMR 15441, FMR 14712, FMR 14635, FMR 11906, and FMR 15388) previously characterized and with different VRC susceptibilities (16) were used in this study. They were retrieved from the Facultat de Medicina de Reus (FMR) fungal collection and grown on potato dextrose agar (PDA) media (Conda-Pronadisa) for 3 to 5 days at 35°C twice before use.

For gene expression analyses, 1 × 10⁷ conidia/ml suspensions were germinated in YG medium (0.5% yeast extract, 2% glucose) at 37°C and 180 rpm for 12 h. After this, mycelia were filtered and aseptically transferred to new YG medium (for the control condition) or new YG medium supplemented with VRC (Pfizer, Inc.) at 4 μg/ml. All cultures were further incubated at 37°C and 180 rpm for 8 h, and mycelia were harvested by filtration, washed with sterile distilled water, frozen in liquid nitrogen, and stored at −80°C until use.

For sterol extraction, the same growth conditions were followed, with a final step consisting of lyophilization using the AdVantage freeze dryer (The VirTis Company, Inc.) and mechanical grinding.

Targeted gene deletion by CRISPR-Cas9.

Generation of cyp51A or cyp51B single-gene KOs was carried out following a previously described CRISPR-Cas9 protocol (17). Gene deletion was achieved by replacing the complete cyp51A or cyp51B open reading frames (ORFs) with a repair template (Hyg^r) amplified from the pUCGH plasmid with primers introducing flanking microhomology regions of 40 bp adjacent to the upstream and downstream regions of either gene in every species (Table 2). Protoplast generation was performed as reported elsewhere (33), and fungal transformations were carried out by using dual Cas9 ribonucleoprotein (RNP) complexes as follows: two CRISPR RNA (crRNA) sequences were designed to direct the Cas9 cleavage within the 5′ untranscribed region (UTR) and the 3′ UTR of each gene (Table 3) and were used, together with the trans-activating RNA (tracrRNA), for the generation of the guide RNA (gRNA) complexes. These complexes were then combined with commercial Cas9 enzyme and Cas9 working buffer (20 mM HEPES, 150 mM KCl, pH 7.5) to build the RNP complexes, which were used for protoplast transformation. The obtained transformants were phenotypically and molecularly analyzed to verify correct integration and recombination by means of Hyg resistance evaluation and by gDNA PCRs with the appropriate primer pair (i.e., Fw scr-Rev scr, Fw scr-Rev intern, and Fw intern-Rev scr) for every species and gene (Table 4 and Fig. 1).

TABLE 2.

Oligonucleotides used for cyp51 gene deletion^a

Primer	Sequence (5′–3′)	Species
Hyg Fw tail Ancyp51A	CTTGCTCTTGACTTCCTCTTTACAACAATCTTTCTCATCAagcttgcatgcctgcaggtc	A. niger
Hyg Rv tail Ancyp51A	AGGTACTAGTTGGGAAAAGACTGGAGACGAAAGGTAACCTGCcatcgatgatatcagatc	A. niger
Hyg Fw tail Ancyp51B	CCCAAAGTGCACTTTCCTGCGCCGTTACAAGCAATTTACAagcttgcatgcctgcaggtc	A. niger
Hyg Rv tail Ancyp51B	ATTTCATCGTATTACATGAGAGGTTTAAGTACAATCACCGAAcatcgatgatatcagatc	A. niger
Hyg Fw tail Atcyp51A	TCCCCATCAATTGGTCCTTTTTCCTGCCTACGGTCGCTTCagcttgcatgcctgcaggtc	A. tubingensis
Hyg Rv tail Atcyp51A	CTAGTTGGGAAAGACTGGAGATGAAAGATAGCTTGCAGAGAAcatcgatgatatcagatc	A. tubingensis
Hyg Fw tail Atcyp51B	CTTCCTTGTGAACACAACCAACTCACTCCAATTGTCTGTTagcttgcatgcctgcaggtc	A. tubingensis
Hyg Rv tail Atcyp51B	GGTTTGAGAACAATCACAAAAGGCGGTTTCTATGCAGTAACAcatcgatgatatcagatc	A. tubingensis

^aCapital letters represent homologous region to the DNA template. Lowercase represent homologous region to the Hyg^r repair template.

TABLE 3.

crRNA sequences used for genetic transformation

crRNA	Sequence (5′–3′)	Species
Ancyp51A 5′ gRNA	CAACAATCTTTCTCATCAAC	A. niger
Ancyp51A 3′ gRNA	GAATTAGGCCACCTTCTTGC	A. niger
Ancyp51B 5′ gRNA	CCGTTACAAGCAATTTACAC	A. niger
Ancyp51B 3′ gRNA	TACTGTATAGAAACCGCGTT	A. niger
Atcyp51A 5′ gRNA	TCCTGCCTACGGTCGCTTCC	A. tubingensis
Atcyp51A 3′ gRNA	AAAGATAGCTTGCAGAGAAG	A. tubingensis
Atcyp51B 5′ gRNA	CTCACTCCAATTGTCTGTTC	A. tubingensis
Atcyp51B 3′ gRNA	CGGTTTCTATGCAGTAAGAG	A. tubingensis

TABLE 4.

Oligonucleotides used as PCR primers for cyp51A and cyp51B knockout screening

Primer	Sequence (5′–3′)	Species
Ancyp51A Fw scr	GAACGATGCAGCTCAGTGCCA	A. niger
Ancyp51A Rev scr	ACACGAGGGATCTCACGTGC	A. niger
Ancyp51B Fw scr	CCTGCTACTAGGAGAGCCGC	A. niger
Ancyp51B Rev scr	GCCTGTATGGGACAGTCGCT	A. niger
Atcyp51A Fw scr	CCGATGTAGCTCAGTGCCAGG	A. tubingensis
Atcyp51A Rev scr	ACTGGGCAGCGGTCATATCTC	A. tubingensis
Atcyp51B Fw scr	GTGCACTTTCCTGCGCTGTTG	A. tubingensis
Atcyp51B Rev scr	CCGGCCATTAAAGCCCCAAG	A. tubingensis
Hyg Fw intern	CGATTCCTTGCGGTCCGAAT	All
Hyg Rev intern	ATCGGCACTTTGCATCGGC	All

Antifungal susceptibility testing.

For the in vitro susceptibility testing, the M38 protocol for broth microdilution developed by the Clinical and Laboratory Standards Institute (CLSI) (34) was followed. Briefly, stock solutions of VRC (Pfizer, Inc.) were diluted in RPMI 1640 medium and dispensed in 96-well microdilution trays. Then, conidial suspensions of every strain were adjusted by hemocytometer count and inoculated onto every well. Trays were then incubated at 35°C for 48 h without light or agitation. VRC MICs, corresponding to the lowest VRC concentration that completely inhibited fungal growth, were determined by direct visualization with an inverted mirror. A. flavus ATCC 204304 was used as the quality control strain.

RNA extraction and gene expression analysis.

Mycelia were ground in liquid nitrogen with mortar and pestle, and total RNA was extracted with the TRIzol reagent (Thermo Fisher) as previously described (35). The NucleoSpin RNA kit with an on-column DNase digestion (Macherey-Nagel) was used to purify the RNA. Total RNA quality and quantity were determined by spectrophotometric analysis in a NanoDrop 2000 spectrophotometer (Thermo Fisher) and running aliquots in RedSafe-stained agarose gels.

For gene expression analysis, RNA was reverse transcribed into first-strand complementary DNA (cDNA) using the iScript cDNA synthesis kit (Bio-Rad) following the manufacturer’s instructions. Real-time PCRs were carried in a StepOne Plus real-time PCR system (Applied Biosystems) using 7.5 μl of TB Green Premix Ex Taq (Tli RNase H Plus) (TaKaRa Bio), 6.6 μl of cDNA template, 0.3 μl of the reference dye ROX, and 300 nM of each gene-specific primer (16) in a final reaction volume of 15 μl. Thermal cycling conditions were as follows: 95°C (30 s) and 45 cycles of 95°C (5 s), 60°C (30 s), and 72°C (30 s). For all reactions, a melting curve was obtained after PCR completion to check amplification specificity under the following conditions: 95°C (15 s), 60°C (15 s), and 95°C (15 s).

Expression levels were calculated after normalization of cyp51A and cyp51B cycle thresholds (C_T) with those of the housekeeping gene actA (36).

Sterol extraction and quantification.

Total sterol extraction was carried out as previously published (37) with some modifications. In brief, 60 mg (dry weight) of every sample were placed into borosilicate glass screw-cap tubes with 6 ml of 25% alcoholic potassium hydroxide solution and vortex mixed for 1 min. After incubating the tubes at 85°C for 1 h in a water bath, 2 ml of sterile distilled water and 6 ml of n-heptane were added, tubes were vortex mixed, and the upper layer (n-heptane) was transferred to a new tube and stored at −20°C for a maximum of 20 h.

Sterol quantification was carried out as described elsewhere (31) by spectrophotometric scan between 240 and 300 nm with a NanoDrop 2000 spectrophotometer. The presence of ergosterol and the intermediate sterol 24(28)DHE in the extracted samples resulted in four-peaked curves. Both ergosterol and 24(28)DHE absorb at 281.5 nm, but 24(28)DHE shows intense absorption at 230 nm, so ergosterol quantification can be determined by subtracting the 24(28)DHE content from the total sterol amount following the equations previously established by Arthington-Skaggs et al. (37).

Statistical analysis.

The nonparametric Mann-Whitney test was used to compare relative gene expression values and ergosterol contents. All statistical analyses were performed with GraphPad Prism 6.0 for Windows. P values of ≤ 0.05 were considered statistically significant.