Targeted Gene Disruption of the 14-α Sterol Demethylase (cyp51A) in Aspergillus fumigatus and Its Role in Azole Drug Susceptibility

Strains.

The strains used in this study were A. fumigatus strain CM-237, which was used for describing the sequence of cyp51A and cyp51B (7), and two clinical A. fumigatus strains, CNM-CM-1252 (AF-90) and CNM-CM-796 (filamentous fungus collection of the Spanish National Center for Microbiology), with elevated MICs to azole drugs (Table 1) and different Cyp51A amino acid substitutions (5, 8).

Molecular cloning and DNA sequencing.

The full coding sequence of cyp51A of A. fumigatus was PCR amplified as previously described (7) and cloned into the pGEM-T vector system (Promega, Madrid, Spain) to obtain plasmid pUM100. Restriction digestion of plasmid pID621 (kindly provided by D. W. Holden) was used to obtain the 1.4-kb SalI fragment of a hygromycin B (hph) resistance cassette (4) for construction of the disruption vector. The 1.4-kb hph cassette was inserted into the XhoI restriction site of pUM100 to create pUM102. A linear 3.0-kb DNA fragment obtained by SacI/SacII double digestion of pUM102 was used for A. fumigatus strain transformations (Fig. 1A).

• Open in new tab
• Download powerpoint

FIG. 1.

(A) Fragments of the A. fumigatus cyp51A gene and construct (plasmid pUM102) used for creating the cyp51A-deficient mutant strain. Unfilled boxes indicate the hygromycin resistance cassette (hph), hatched bars represent the cyp51A genomic sequence, and the striped arrow represents the full cyp51A coding sequence. Sites for restriction enzymes are as follows: S, SalI; X, XhoI; E, EcoRV. (B) Southern hybridization analysis of cyp51A single-mutant strain CM-A8, two transformants with ectopic integration (CM-A4 and CM-A5), and the wild-type strain (CM-237). Genomic DNAs were digested with EcoRV and hybridized using a 925-bp PCR fragment from the cyp51A gene as a probe (black box). (C) Southern hybridization analysis of the wild-type CM-237 strain, cyp51A mutant strain CM-A8, ITC-resistant parental strains (CM-1252 and CM-796), and the corresponding cyp51A mutant strains, CM-A83 and CM-A41. Genomic DNAs were digested with SalI and probed as before. Sizes of expected bands are indicated on the side in kb.

Aspergillus transformations.

A. fumigatus transformation experiments were achieved by electroporation using a protocol previously described (15) with subsequent modifications (5, 18). Hygromycin B (130 μg/ml; Sigma) was used for transformants selection. Mutants were named by a letter (e.g., A) followed by a number. Genomic DNAs from hygromycin-resistant transformants and the parental strain were digested with two different restriction enzymes (SalI and EcoRV; Amersham Biosciences, Madrid, Spain). Southern analysis was performed as previously described (7, 14).

Antifungal susceptibility testing.

Broth microdilution susceptibility testing was performed as described in NCCLS document M38-A (10), with modifications (3, 11, 13). Itraconazole (ITC), ketoconazole (KTC) (both from Janssen Pharmaceutical S.A., Madrid, Spain), voriconazole, fluconazole (FLC) (both from Pfizer S.A., Madrid, Spain), ravuconazole (Bristol-Myers Squibb, Madrid, Spain), and amphotericin B (AmB; Sigma Aldrich Quimica, S.A., Madrid, Spain) were tested. Susceptibility tests were performed at least three times with each strain on different days.

RNA extraction and LightCycler PCR.

RNA extraction from the A. fumigatus CM-237 strain and the derived CM-A8 mutant strain and reverse transcription reactions were performed (7). Amplification of cDNAs was carried out using the LightCycler PCR system (Roche Diagnostics, Madrid, Spain). Also, primers Tub1 (5′ AACCAAATTGGTGCCGC 3′) and Tub2 (5′ CACGGATCTTGGAGATC 3′) were used for amplification of the A. fumigatus β-tubulin housekeeping gene (Tub1) (GenBank accession number AY048754). LightCycler PCRs were set up with FastStart DNA Master SYBR Green (Roche Diagnostic). Each assay included duplicate reactions and was repeated three times on different days. The method described by Pfaffl (12) was employed for relative quantification. PCR efficiencies were calculated from the curve slopes given by LightCycler software (Roche Diagnostic).

Total ergosterol content.

Ergosterol was determined in strains CM-237 and CM-A8 using the protocol described by Arthington-Skaggs et al. (1). Ergosterol content was analyzed by high-performance liquid chromatography using a μBondapak C₁₈ column (Waters LC Module I plus; Waters Corporation, Madrid, Spain). The quantities of sterols were determined with Millenium³² and Millenium³² photodiode array detector (PDA) software (Waters Corporation). The experiments were repeated four times.

Pathogenicity test.

The pathogenicity of the cyp51A (CM-A8) mutant strain was assessed in neutropenic mice (ICR, SPF, 6 weeks old; CRIFFA, Barcelona, Spain) as previously described by Smith et al. (17) and compared with that of CM-237. The experiments were carried out with eight mice per group.

Data analysis.

The significance of the differences in MICs and ergosterol content was determined by Student′s t test (unpaired, unequal variance). A P value of <0.01 was considered significant. Kaplan-Meier survival analysis was used to determine differences in the pathogenicity test. Statistical analysis was done with the SPSS package (version 12.0; SPSS S.L., Madrid, Spain).

Results and discussion.

Three cyp51A gene knockout strains (CM-A8, CM-A41, and CM-A83) were produced (Fig. 1B and C). The cyp51A mutant strains demonstrated decreased MICs to all azoles, especially FLC and KTC, and also to ITC in ITC-resistant strains (Table 1).

No differences were found between cyp51A CM-A8 mutant and CM-237 parenteral strains regarding macroscopic and microscopic morphology and colony radial growth. The total ergosterol contents of both isolates were also similar, with an arithmetic mean ± standard deviation of 5.56 ± 0.34 μg/mg of dry mass for strain CM-237 and 5.98 ± 0.82 μg/mg for mutant CM-A8 (P > 0.01). No significant differences (P > 0.01) were found in the cyp51B expression levels between the cyp51A CM-A8 and CM-237 organisms. The mean of the three repetitions yielded a ratio of 1.01 with a standard deviation of 0.36. There were no significant differences in the onset of illness or any reduction in mortality between the CM-237 strain and the cyp51A CM-A8 mutant strain (Fig. 2) (P > 0.01).

• Open in new tab
• Download powerpoint

FIG. 2.

Survival of immunocompromised mice infected with the wild-type CM-237 strain (○) and the cyp51A-deficient CM-A8 strain (▴) of A. fumigatus. An inoculum of 10⁴ spores/mouse was used for each strain. A control group with mice immunocompromised but without inoculum was also included (⧫).

Although there are no reports of cyp51 functional analysis in any filamentous fungi, Erg11/Cyp51 function has been shown to be essential in Saccharomyces cerevisiae but not in other yeasts (2, 6, 9, 16). In our study, the lack of morphological defects, the unchanged ergosterol content, the absence of increased cyp51B expression in cyp51A-deficient mutants, and their pathogenicity indicate that the cyp51A gene in A. fumigatus is not essential for viability. The susceptibility testing results of mutants showed a decrease in the MICs of azoles, with the cyp51A mutants being 17 and 40 times more sensitive than wild-type organisms to FLC and KTC, respectively. The marked increased susceptibility to FLC and KTC is noteworthy since A. fumigatus is intrinsically resistant to both antifungals. This susceptibility pattern might suggest that the still active Cyp51B could be more susceptible to these antifungals.

An A. fumigatus cyp51A mutant could be used as an screen for identification of inhibitory compounds that specifically target Cyp51B activity. These findings could be confirmed by testing antifungal compounds in a murine model of invasive aspergillosis.