Erg251 has complex and pleiotropic effects on sterol composition, azole susceptibility, filamentation, and stress response phenotypes

Abstract

Ergosterol is essential for fungal cell membrane integrity and growth, and numerous antifungal drugs target ergosterol. Inactivation or modification of ergosterol biosynthetic genes can lead to changes in antifungal drug susceptibility, filamentation and stress response. Here, we found that the ergosterol biosynthesis gene ERG251 is a hotspot for point mutations during adaptation to antifungal drug stress within two distinct genetic backgrounds of Candida albicans. Heterozygous point mutations led to single allele dysfunction of ERG251 and resulted in azole tolerance in both genetic backgrounds. This is the first known example of point mutations causing azole tolerance in C. albicans. Importantly, single allele dysfunction of ERG251 in combination with recurrent chromosome aneuploidies resulted in bona fide azole resistance. Homozygous deletions of ERG251 caused increased fitness in low concentrations of fluconazole and decreased fitness in rich medium, especially at low initial cell density. Homozygous deletions of ERG251 resulted in accumulation of ergosterol intermediates consistent with the fitness defect in rich medium. Dysfunction of ERG251, together with FLC exposure, resulted in decreased accumulation of the toxic sterol (14-ɑ-methylergosta-8,24(28)-dien-3β,6α-diol) and increased accumulation of non-toxic alternative sterols. The altered sterol composition of the ERG251 mutants had pleiotropic effects on transcription, filamentation, and stress responses including cell membrane, osmotic and oxidative stress. Interestingly, while dysfunction of ERG251 resulted in azole tolerance, it also led to transcriptional upregulation of ZRT2, a membrane-bound Zinc transporter, in the presence of FLC, and overexpression of ZRT2 is sufficient to increase azole tolerance in wild-type C. albicans. Finally, in a murine model of systemic infection, homozygous deletion of ERG251 resulted in decreased virulence while the heterozygous deletion mutants maintain their pathogenicity. Overall, this study demonstrates that single allele dysfunction of ERG251 is a recurrent and effective mechanism of acquired azole tolerance. We propose that altered sterol composition resulting from ERG251 dysfunction mediates azole tolerance as well as pleiotropic effects on stress response, filamentation and virulence.

Fig 1. The point mutation of ERG251 leads to the partial dysfunction of ERG251 causing acquisition of azole tolerance.

Liquid microbroth drug susceptibility assay. Fluconazole (FLC) resistance quantified as the MIC₅₀ at 24hr in increasing concentrations of FLC (left) and FLC tolerance quantified as the Supra-MIC growth at 48hr (SMG, right) which is the average growth above the MIC₅₀ for: A. the wild-type SC5314 (ERG251/ERG251), engineered heterozygous ERG251 point mutations strains in the SC5314 background, and both heterozygous deletion mutants of ERG251 in the SC5314 background; and B. the wild-type SC5314 (ERG251/ERG251), ERG251 overexpression strain, both heterozygous deletion mutants of ERG251 and their corresponding complementation strains, an ERG251 heterozygous deletion in the P75063 background and wild-type P75063 (P75063-ERG251/ERG251) as a control. C. Rhodamine 6G efflux kinetics of two heterozygous deletion mutants and the homozygous deletion of ERG251 in SC5314 with SC5314 (ERG251/ERG251) as the control in YPAD (left) and YPAD+1μg/ml FLC (right). Plots indicate average fluorescence intensity changes of Rhodamine 6G (R6G) from three biological replicates over 90 min. D. 24hr MIC (left, μg/ml) and 48hr SMG (right, tolerance) in FLC with or without radicicol (Hsp90 inhibitor) treatment for two heterozygous deletion mutants of ERG251 with SC5314 (ERG251/ERG251) and a Positive Control (a FLC resistant clinical isolate (C17/12-99)). A&B&D: For MIC values, each dot represents a single replicate and each bar represents the average of three biological replicates of a single strain; SMG values are mean ± SEM calculated from three biological replicates of a single strain.

Fig 2. Single allele dysfunction of ERG251 in combination with concurrent aneuploidy causes azole resistance.

A. Representative whole genome sequencing (WGS) data of the FLC-evolved strains 1.1/1.2, 2.1/2.2, and 3.1/3.2 that acquired heterozygous point mutations at ERG251 and Chr3 and Chr6 concurrent aneuploidy. B. WGS data of FLC-evolved strain 4.1 that had wild-type alleles of ERG251/ERG251 and Chr3 and Chr6 concurrent aneuploidy, plus two ERG251 heterozygous deletion mutants engineered in the Evolved 4.1 aneuploid background. A&B WGS data are plotted as the log2 ratio and converted to chromosome copy number (y-axis, 1–4 copies) as a function of chromosome position (x-axis, Chr1-ChrR). The baseline ploidy was determined by propidium iodide staining (S1 Table). Haplotypes relative to the reference genome SC5314 are indicated. C. 24hr MIC (left, μg/ml) and 48hr SMG (right, tolerance) in FLC for SC5314 (ERG251/ERG251), ERG251 heterozygous deletion mutant in the SC5314 background, FLC-evolved strain 4.1, and two ERG251 heterozygous deletion mutants engineered in the Evolved 4.1 aneuploid background (two independent transformants). MIC: each dot represents a single replicate and bar represents the average of three technical replicates of a single strain; SMG values are mean ± SEM calculated from three technical replicates of a single strain. D. Rhodamine 6G efflux kinetics of ERG251 heterozygous deletion mutant in evolved strain 4.1 background with evolved strain 4.1 and SC5314 (ERG251/ERG251) as the controls in YPAD (left) and YPAD+1μg/ml FLC (right). Plots indicate fluorescence intensity changes of Rhodamine 6G (R6G) over 90 min.

Fig 3. Homozygous deletion of ERG251 results in decreased fitness at low initial cell density and increased fitness in the presence of low concentrations of FLC (≤1μg/ml).

A. 48hr growth curve analysis of erg251Δ/Δ started at three different initial cell densities (OD₆₀₀ = 0.001, 0.005, or 0.01) with ERG251/ERG251 (SC5314, OD₆₀₀ = 0.001) as the control. Average slope and ±SEM for three technical replicates is indicated. B. Carrying capacity (K) and doubling time (Td, hrs), and lag phase (hrs) determined from growth curve analysis in Fig 2A. C&D. X-Y growth curve assay of (C) erg251Δ/Δ and (D) ERG251/ERG251 in the presence of increasing concentrations of FLC (X-axis, 0–256 μg/ml, 2-fold dilutions) and/or increasing concentrations of farnesol (FAR) (Y-axis, 0–1000 μM, 2-fold dilutions). Growth was estimated with the area under the curve (AUC heatmap) of the 48hr growth curve. E&F. Cell viability of (E) erg251Δ/Δ and (F) ERG251/ERG251 after 48 hr exposure to FLC or/and FAR. Cells from Fig 3B were plated on YPAD agar and imaged after 24hr incubation. G. Relative fitness calculated from head-to-head competitive assay for erg251Δ/ERG251, ERG251/erg251Δ, erg251Δ/Δ, erg251Δ/Δ+ERG251-A, and erg251Δ/Δ+ERG251-B compared to the fluorescent control strain (ERG251/ERG251). B&G: Values are mean ± SEM calculated from three technical replicates. Data were assessed for normality by Shapiro-Wilk, and significant differences between the ERG251/ERG251 and mutants were calculated using two-way ANOVA with Dunnett’s multiple comparisons test. ****p<0.0001, **p<0.01. A-G: At least three biological replicates were performed.

Fig 4. Homozygous deletion of ERG251 leads to increased sensitivity to cell membrane and osmotic stress but decreased sensitivity to oxidative stress.

A. Volcano plot for differentially expressed genes (log₂ fold change ≥ 1 or ≤-1 and adjusted p-value < 0.05) in the erg251Δ/Δ mutant compared to ERG251/ERG251 in YPAD. Genes that are significantly differentially expressed by both fold change and p-value cut-offs are in red. B. Gene Ontology (GO) terms for differentially expressed genes (log₂ fold change ≥ 1 or ≤ -1 and adjusted p-value < 0.05) in the erg251Δ/Δ mutant compared to ERG251/ERG251 in YPAD. Cell wall organization, biofilm formation, lipid metabolic process, filamentous growth and response to stress. GO terms are highlighted and differentially expressed genes contributing to the enrichment noted to the right. C. Relative growth (area under growth curve) of ERG251/ERG251, erg251Δ/ERG251, ERG251/erg251Δ, and erg251Δ/Δ in YPAD+NaCl (0 to 3.0M), YPAD+SDS (0 to 0.1%), and YPAD+H₂O₂ (0 to 25 mM) across different concentrations (Methods). Relative growth was calculated by normalizing to the growth of no drug control. Arrows indicate the minimum concentration that inhibits the growth (20%) of ERG251/ERG251 (gray) and erg251Δ/Δ (red) relative to no drug control. Dashed line indicates the cut-off for the 20% decreased growth. Data are presented as the mean ±SEM for three technical replicates. A-C: At least three biological replicates were performed.

Fig 5. Deletion of ERG251-A but not ERG251-B leads to decreased filamentation.

A. Representative filamentation images of wild-type ERG251/ERG251, erg251Δ/ERG251, erg251Δ/ERG251+ERG251-A, ERG251/erg251Δ, ERG251/erg251Δ+ERG251-B, erg251Δ/Δ, erg251Δ/Δ+ERG251-A, and erg251Δ/Δ+ERG251-B. Cells were induced in RPMI supplemented with 10% FBS for 4 hrs. Scale bar, 20 μm. B. Quantification of the yeast (<6μm), pseudohyphae (15–36 μm), and hyphae (>36 μm) from genotypes in Fig 5A. 150 to 500 cells were counted for each strain, and at least two biological replicates were performed. Values are mean ± SEM calculated from three biological replicates. Statistical significance for filamentation was compared to ERG251/ERG251 and assessed using two-way ANOVA with uncorrected Fisher’s LSD, ***P <0.001, **P <0.01, * P ≤ 0.05, ns: P >0.05. C. Principal component analysis of transcriptional data in YPAD and YPAD+FLC (1μg/ml) for ERG251/ERG251, erg251Δ/ERG251, ERG251/erg251Δ, and erg251Δ/Δ. D. Venn diagrams comparing the genes that are differentially expressed in erg251Δ/ERG251 and ERG251/erg251Δ (log2 fold change ≥ 0.5 or ≤-0.5 and adjusted p-value < 0.1) relative to ERG251/ERG251 in YPAD. E. The relative expression level (log2 fold change) of genes associated with filamentation in erg251Δ/ERG251, ERG251/erg251Δ, and erg251Δ/Δ compared to ERG251/ERG251 in YPAD.

Fig 6. Homozygous deletion of ERG251 leads to the decreased ergosterol accumulation in the absence of FLC and decreased production of toxic dienol in the presence of FLC.

A. Overview of the ergosterol biosynthetic pathway in C. albicans, including the mevalonate, late ergosterol, and alternate pathways [11,19,69,73]. Genes that were down-regulated (blue) and up-regulated (red) in the erg251Δ/Δ under no drug conditions relative to SC5314 (S4A Fig). B. Representative GC-MS profiling of ERG251/ERG251, erg251Δ/ERG251, ERG251/erg251Δ, and erg251Δ/Δ strains in absence of drug and in the presence of 1μg/ml FLC. The number above each peak represents the area of the peak based on the number of counts taken by the mass spectrometer detector at the point of retention. Labelled peaks indicate the input standard cholesterol, ergosterol, and unidentified sterols: sterol A, toxic dienol intermediates and alternative sterol. All unidentified sterols were compared with known standards: ergosterol, lanosterol, obtusifoliol, zymosterol, 4,4-dimethyl zymosterol, eburicol, episterol, and gramisterol (24-methylenelophenol). C. Abundance of sterols in tested strains from Fig 6B. Values are mean ± SEM calculated from three biological replicates. Statistical significance for filamentation was compared to ERG251/ERG251 from the same condition and assessed using two-way ANOVA followed by Dunnett’s multiple comparisons test", ****P <0.0001, ***p<0.001, **p< 0.01, *p<0.05, ns: P >0.05.

Fig 7. Dysfunction of ERG251 activates a Zinc transporter contributing to decreased azole susceptibility.

A. The relative expression level (log2 fold change) of CDR1, CDR2, MDR1 and ZRT2 in erg251Δ/ERG251, ERG251/erg251Δ, and erg251Δ/Δ compared to ERG251/ERG251 under YPAD+1μg/ml FLC condition. B. Venn diagrams comparing the genes that differentially expressed in erg251Δ/ERG251, ERG251/erg251Δ and erg251Δ/Δ relative to ERG251/ERG251 under YPAD+1μg/ml FLC condition. C. mRNA expression fold change (y-axis) of ZRT2 in ZRT2 overexpression strain (tetO-ZRT2-1) (Left, RT-qPCR) and in erg251Δ/ERG251 and ERG251/erg251Δ (Right, RNA-seq) under YPAD or YPAD+1μg/ml FLC condition relative to the wild-type control. Asterisk indicates the expression change is significant (adjusted p-value < 0.1). Dotted line indicates the SMG of erg251Δ/ERG251. At least three biological repeats were performed. D. 24hr MIC (left, μg/ml) and 48hr SMG (right, tolerance) in FLC for two ZRT2 overexpression strains (tetO-ZRT2-1 and tetO-ZRT2-2, independent transformants) in SC5314 background and two ZRT2 heterozygous deletion mutants in erg251Δ/ERG251 background (erg251Δ/ERG251: zrt2Δ/ZRT2-1 and erg251Δ/ERG251: zrt2Δ/ZRT2-2) together with SC5314 (ERG251/ERG251) and erg251Δ/ERG251 as the controls. MIC: each dot represents a single replicate and bar represents the average of three technical replicates of a single strain; SMG values are mean ± SEM calculated from three technical replicates of a single strain. At least three biological replicates were performed.

Fig 8. ERG251 heterozygous deletion mutants maintain their pathogenicity in a murine model.

A. ICR mice were injected via the tail vein with 5x10⁵ cells of ERG251/ERG251 (SC5314), erg251Δ/ERG251, ERG251/erg251Δ, and erg251Δ/Δ+ERG251-A and erg251Δ/Δ+ERG251-B and survival was presented over the time. The erg251Δ/Δ mutant survival curves were significantly attenuated from that of the ERG251/ERG251 (Log-rank (Mantel-Cox) test; **, p = 0.0015). Eight mice per strain were used.