Genome-Wide Analysis of Experimentally Evolved Candida auris Reveals Multiple Novel Mechanisms of Multidrug Resistance

Abstract

Candida auris is globally recognized as an opportunistic fungal pathogen of high concern, due to its extensive multidrug resistance (MDR). Still, molecular mechanisms of MDR are largely unexplored. This is the first account of genome-wide evolution of MDR in C. auris obtained through serial in vitro exposure to azoles, polyenes, and echinocandins. We show the stepwise accumulation of copy number variations and novel mutations in genes both known and unknown in antifungal drug resistance. Echinocandin resistance was accompanied by a codon deletion in FKS1 hot spot 1 and a substitution in FKS1 "novel" hot spot 3. Mutations in ERG3 and CIS2 further increased the echinocandin MIC. Decreased azole susceptibility was linked to a mutation in transcription factor TAC1b and overexpression of the drug efflux pump Cdr1, a segmental duplication of chromosome 1 containing ERG11, and a whole chromosome 5 duplication, which contains TAC1b The latter was associated with increased expression of ERG11, TAC1b, and CDR2 but not CDR1 The simultaneous emergence of nonsense mutations in ERG3 and ERG11 was shown to decrease amphotericin B susceptibility, accompanied with fluconazole cross-resistance. A mutation in MEC3, a gene mainly known for its role in DNA damage homeostasis, further increased the polyene MIC. Overall, this study shows the alarming potential for and diversity of MDR development in C. auris, even in a clade until now not associated with MDR (clade II), stressing its clinical importance and the urge for future research.IMPORTANCECandida auris is a recently discovered human fungal pathogen and has shown an alarming potential for developing multi- and pan-resistance toward all classes of antifungals most commonly used in the clinic. Currently, C. auris has been globally recognized as a nosocomial pathogen of high concern due to this evolutionary potential. So far, this is the first study in which the stepwise progression of multidrug resistance (MDR) in C. auris is monitored in vitro Multiple novel mutations in known resistance genes and genes previously not or vaguely associated with drug resistance reveal rapid MDR evolution in a C. auris clade II isolate. Additionally, this study shows that in vitro experimental evolution can be a powerful tool to discover new drug resistance mechanisms, although it has its limitations.

Keywords: Candida auris; amphotericin B; antifungal agents; caspofungin; drug resistance evolution; experimental evolution; fluconazole; genome analysis; microevolution; multidrug resistance; whole-genome sequencing.

FIG 1

Schematic overview of the in vitro experimental evolution assay. (A) A single colony is cultured in RPMI-MOPS medium (2% glucose) for 24 h at 37°C after which a standardized inoculum (10⁶ cells) is resuspended in medium containing no drug (control), the drug at a concentration of 2× MIC₅₀, 1× MIC₅₀, and 0.5× MIC₅₀ (shown here) of the particular starting strain. Daily, the culture is rediluted (1/10) in fresh RPMI-MOPS medium (2% glucose) with a concentration of drug based on the OD₆₀₀ of the control culture (see Materials and Methods). All strains were evolved in triplicate. Daily aliquots of evolving populations were stored in RPMI-MOPS medium containing 25% glycerol at −80°C for later analysis. (B) Ancestry of the five evolved strains that were sequenced. WGS was performed on a single colony. The name of each strain represents the experimental treatment (letter) and day of isolation (number), respectively.

FIG 2

Resistance profiles of endpoint and intermediate evolved strains. (A) Summary of MIC₅₀ values and associated mutations/CNVs for each endpoint strain and divergent intermediate strain. (B to D) Growth profiles of evolved strains relative to the wild-type strain (wt) in a broth dilution assay (BDA) of fluconazole (B), caspofungin (C), and amphotericin B (D). The percentage of growth was calculated from growth without drug and based on OD₆₀₀ measurements after 48 h of incubation at 37°C. Each data point and its standard deviation is calculated from 3 biological repeats, each represented by the mean of 2 technical repeats. Pdup: partial duplication, dup: duplication. Resistance profiles of endpoint evolved strains for all drugs are found in supplemental material Fig. S1.

FIG 3

Coverage plot of whole-genome sequencing of endpoint evolved strains. The coverage displayed is calculated by normalizing the average coverage depth per 5-kb window. Each color represents 1 chromosome (from left to right, chromosomes 1 to 7). Indicated are the significant duplication in chromosome 1 (Chr1) in strain F30 and FC17 and chromosome 5 (Chr5) in strain CF16.

FIG 4

Hot spot (HS) region mutations of the FKS genes that confer echinocandin resistance. Amino acid sequence of hot spots 1, 2, and 3 (HS1 to -3, respectively) of C. auris and other fungi are aligned along with all mutations found to decrease echinocandin susceptibility as described in the literature (references are given between brackets). Species-specific polymorphisms of HS are indicated in gray, and the mutations found to confer echinocandin resistance in this study are indicated by a grid. Δ, deletion; *, nonsense mutation; a, mutations R647G and P649L were exclusively heterozygous; b, FKS2 and FKS1 are functionally redundant in C. glabrata and both mutated in echinocandin-resistant isolates; c, the naturally occurring alanine at position 660 allows intrinsic reduced echinocandin susceptibility in C. parapsilosis.

FIG 5

Growth curves of endpoint evolved strains. Growth curves were plotted based on culture density (spectrophotometric quantification of OD₆₀₀; see Materials and Methods) over 72 h of incubation in RPMI-MOPS medium containing (A) 0.2% glucose and (B) 2% glucose at 37°C. Data points are average values of three biological repeats, each represented by the average of two technical repeats.

FIG 6

Relative expression of various genes of interest among evolved strains. Fold change of expression levels for CDR1, CDR2, ERG11, and TAC1b for the wild type (wt), endpoint evolved strains (A29, F30, FC17, C20, CF16) and intermediate strains F12 and F13 (for CDR1 and ERG11). Bars represent log₂-transformed means with standard deviation accounting for data obtained from 3 biological repeats, each represented by the mean of 2 technical repeats. Asterisks indicate significant overexpression; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001.