Broad antifungal resistance mediated by RNAi-dependent epimutation in the basal human fungal pathogen Mucor circinelloides

Abstract

Mucormycosis-an emergent, deadly fungal infection-is difficult to treat, in part because the causative species demonstrate broad clinical antifungal resistance. However, the mechanisms underlying drug resistance in these infections remain poorly understood. Our previous work demonstrated that one major agent of mucormycosis, Mucor circinelloides, can develop resistance to the antifungal agents FK506 and rapamycin through a novel, transient RNA interference-dependent mechanism known as epimutation. Epimutations silence the drug target gene and are selected by drug exposure; the target gene is re-expressed and sensitivity is restored following passage without drug. This silencing process involves generation of small RNA (sRNA) against the target gene via core RNAi pathway proteins. To further elucidate the role of epimutation in the broad antifungal resistance of Mucor, epimutants were isolated that confer resistance to another antifungal agent, 5-fluoroorotic acid (5-FOA). We identified epimutant strains that exhibit resistance to 5-FOA without mutations in PyrF or PyrG, enzymes which convert 5-FOA into the active toxic form. Using sRNA hybridization as well as sRNA library analysis, we demonstrate that these epimutants harbor sRNA against either pyrF or pyrG, and further show that this sRNA is lost after reversion to drug sensitivity. We conclude that epimutation is a mechanism capable of targeting multiple genes, enabling Mucor to develop resistance to a variety of antifungal agents. Elucidation of the role of RNAi in epimutation affords a fuller understanding of mucormycosis. Furthermore, it improves our understanding of fungal pathogenesis and adaptation to stresses, including the evolution of drug resistance.

Fig 1. sRNA hybridization and phenotypic analysis of 5-FOA-resistant epimutants.

(A) sRNA hybridization of epimutants and revertants from an rdrp3 mutant background, before (R–resistant) and after 5 (P5) or 10 (P10) passages without selection. P, rdrp3Δ parental strain (MU439). Blots were hybridized with antisense-specific probes against pyrF, pyrG, or 5S rRNA (loading control). (B) sRNA blot of epimutants and revertants from an rdrp1 mutant background, before (R–resistant) and after 5 passages without selection (P5). P, rdrp1Δ parental strain (MU419). Blots were hybridized with antisense-specific probes against pyrF, pyrG or 5S rRNA (loading control). (C) Phenotypic analysis of one representative epimutant, before and after reversion. A pyrF epimutant (E2) is shown before and after 5 (P5) and 10 (P10) passages without selection, grown on MMC media, MMC supplemented with uridine and uracil, and MMC supplemented with 5-FOA, uridine, and uracil. pyrG^-, a known mutant of pyrG, served as a negative control; P, rdrp3Δ parental strain (MU439). (D) A pyrG epimutant (E4) is shown before and after 5 (P5) passages without selection, grown on MMC, MMC supplemented with uridine and uracil, and MMC supplemented with 5-FOA, uridine, and uracil. P, rdrp1Δ parental strain (MU419).

Fig 2. 5-FOA resistance is associated with increased levels of sRNAs against either the pyrF or pyrG locus.

(A) Representative diagram of sRNAs mapped across the pyrF locus showing accumulation of both sense (- values) and antisense sRNA (+ values) in epimutant E1 (in red). Expression levels are greatly decreased in the revertant after 5 passages without selection (in blue). No increase in sRNA levels is seen in the surrounding regions. (B) Representative diagram of sRNAs mapped across the pyrG locus showing accumulation of both sense (+ values) and antisense sRNA (- values) in epimutant E4 (in red), with greatly decreased sRNA levels in the revertant after 5 passages without selection (in blue). No increase of sRNA levels is observed in the surrounding regions.

Fig 3. Length and terminal nucleotide analysis of sRNAs in epimutants.

Antisense sRNAs from epimutant strains and their corresponding revertants that map against the pyrF and pyrG loci. (A) Analysis of the 5’ nucleotide of antisense sRNAs that map to the pyrF locus, isolated from the pyrF epimutant strain E1 and revertant. (B) Analysis of the 5’ nucleotide of antisense sRNAs that map to the pyrG locus, isolated from the pyrG epimutant strain E4 and revertant. (C) Analysis of the size of antisense sRNAs that map to pyrF in strain E1 and revertant. (D) Analysis of the size of antisense sRNAs that map to pyrG in strain E4 and revertant.

Fig 4. Epimutation decreases expression of pyrF and pyrG mRNA.

(A) Expression of pyrF mRNA in pyrF epimutant and revertant strains (Passage 5, P5) as determined through qRT-PCR, with actin expression used for the reference gene. Gene expression levels were normalized relative to the rdrp3Δ parental strain (PS), using actin as the reference gene via the comparative ΔΔCt method. N = 3 experimental replicates. Significance determined via one-way ANOVA (P = 0.0005, F = 13.37, 4 degrees of freedom) with post-hoc Tukey’s Multiple Comparison test. (B) Expression of pyrG mRNA in pyrF epimutant and revertant strains. N = 1. (C) Expression of pyrF mRNA in a pyrG epimutant and revertant strain as determined through qRT-PCR, with actin expression used for the reference gene. Percent expression was normalized relative to rdrp1Δ parental strain (PS). N = 3 experimental replicates. Significance determined via one-way ANOVA (P = 0.51, F = 0.74, 2 degrees of freedom). (D) Expression of pyrG mRNA in pyrG epimutant and revertant strain. N = 3 experimental replicates. Significance determined via one-way ANOVA (P = 0.059, F = 4.7, 2 degrees of freedom).