Molecular mechanisms underlying the emergence of polygenetic antifungal drug resistance in msh2 mismatch repair mutants of Cryptococcus

Abstract

Background: Fungal infections are common life-threatening diseases amongst immunodeficient individuals. Invasive fungal disease is commonly treated with an azole antifungal agent, resulting in selection pressure and the emergence of drug resistance. Antifungal resistance is associated with higher mortality rates and treatment failure, making the current clinical management of fungal disease very challenging. Clinical isolates from a variety of fungi have been shown to contain mutations in the MSH2 gene, encoding a component of the DNA mismatch repair pathway. Mutation of MSH2 results in an elevated mutation rate that can increase the opportunity for selectively advantageous mutations to occur, accelerating the development of antifungal resistance.

Objectives: To characterize the molecular mechanisms causing the microevolutionary emergence of antifungal resistance in msh2 mismatch repair mutants of Cryptococcus neoformans.

Methods: The mechanisms resulting in the emergence of antifungal resistance were investigated using WGS, characterization of deletion mutants and measuring ploidy changes.

Results: The genomes of resistant strains did not possess mutations in ERG11 or other genes of the ergosterol biosynthesis pathway. Antifungal resistance was due to small contributions from mutations in many genes. MSH2 does not directly affect ploidy changes.

Conclusions: This study provides evidence that resistance to fluconazole can evolve independently of ERG11 mutations. A common microevolutionary route to the emergence of antifungal resistance involves the accumulation of mutations that alter stress signalling, cellular efflux, membrane trafficking, epigenetic modification and aneuploidy. This complex pattern of microevolution highlights the significant challenges posed both to diagnosis and treatment of drug-resistant fungal pathogens.

Figure 1.

The frequent emergence of antifungal drug resistance in the msh2Δ mutant enables rapid isolation of antifungal drug-resistant strains. (a) Compared with the WT (MSH2⁺) and msh2Δ MSH2⁺ controls, the msh2Δ exhibits an elevated frequency of fluconazole-resistant and amphotericin B-resistant colonies on media containing 72 mg/L fluconazole (16× MIC) or 4.8 mg/L amphotericin B (32× MIC). Asterisks indicate statistical significance using a two-tailed Student’s t-test; *P < 0.05, **P < 0.005. (b) An example of one of the independent cultures of msh2Δ grown in liquid culture to allow mutations to accumulate and plated on plates containing fluconazole (12× MIC) or (c) amphotericin B (32× MIC). WT (MSH2⁺) and msh2Δ MSH2⁺ controls are included for comparison.

Figure 2.

Mutations accumulate in the msh2Δ antifungal drug-resistant strains in response to exposure to antifungal drugs. (a) The position of mutations in msh2Δ antifungal drug-resistant strains not present in the original msh2Δ before antifungal selection across the 14 C. neoformans chromosomes. (b) The mutational spectrum of sequence variants in the genomes of msh2Δ antifungal drug-resistant strains: red, transitions; orange, transversions; medium blue, single base indels at homopolymers; dark blue, single base indels; light blue, larger indels at microsatellites and homopolymers; light green, large indels; dark green, other variants.

Figure 3.

Genes encoding components of membrane trafficking pathways mutated in the msh2Δ antifungal drug-resistant strains. Schematic of the membrane trafficking pathways in a fungal cell (adapted from Feyder et al.) showing the role of proteins encoded by genes mutated in the msh2Δ antifungal drug-resistant strains (black text). Transporting newly synthesized proteins from the endoplasmic reticulum (ER) to the Golgi requires formation of COPII vesicles (involves Sec23) and the fusion of vesicles tethered to the Golgi membrane that is dependent on the GTPase Ypt1 and Trs130, a component of the transport protein particle (TRAPP) complex, which acts as a multimeric guanine nucleotide exchange factor for Ypt1 (brown arrow). Snx3 is a late-Golgi sorting nexin and Kes1 is required for negative regulation of Golgi secretory functions. Secretion from the Golgi to the plasma membrane (pink arrow) requires the exocyst complex, containing Sec8, to tether post-Golgi secretory vesicles to sites of exocytosis. There are two pathways from the Golgi to the vacuole: the direct ALP pathway (dark blue arrow) and the indirect VPS pathway via endosomes (light blue arrow). In the ALP pathway (dark blue arrow), ALP is packaged into vesicles through the AP-3 adapter complex, containing Apl6 and Apl3, which also recruits clathrin. In the VPS pathway (light blue arrow), soluble carboxypeptidase Y pro-protease (CPY) binds to its receptor Vps10 in the Golgi lumen and is transported from the trans-Golgi network via AP-1 adapter and clathrin-coated vesicles to endosomes. The Rab GTPase Vps21 and guanine exchange factor for Rab GTPases Vps9 are required for Golgi-endosome trafficking. Vps16 and Vps39 are part of the HOPS complex essential for docking and fusion of vesicles. Fusion of the late endosome (multivesicular body, MVB) to the vacuole also requires the HOPS complex. In the late endosome, proteins are sorted into vesicles that bud into the lumen in a process that requires the ESCRT complexes containing Hse1 (ESCRT-0) and Snf7 (ESCRT-III). The endocytic pathway (red) and retrograde transport (green dashed arrows) are also indicated.

Figure 4.

Two of the msh2Δ fluconazole-resistant strains are aneuploids. Log₂ fold changes in read coverage (every 100 bp) across the 14 chromosomes in drug-resistant strains isolated on fluconazole: KBCN0137 (a), KBCN0138 (b), SACN00B1 (c), SACN00B2 (d); on amphotericin B: KBCN0140 (e) and KBCN0142 (f); and on fluconazole and amphotericin B: KBCN0134 (g) and KBCN0135 (h). Two of the eight strains, SACN00B1 (c) and SACN00B2 (d), are aneuploids, possessing duplications of chr1 (shown in red) and chr4 (shown in dark blue).

Figure 5.

Deletion of MSH2 does not directly affect heteroresistance. Populations of cells from the WT haploid control strain KBCN001 (Haploid), a diploid control strain KBCN105 (Diploid), the msh2Δ mutant strain AISVCN195 (msh2Δ) and the fluconazole-resistant aneuploid strains SACN00B1 (msh2Δ B1) and SACN00B2 (msh2Δ B2) were analysed by flow cytometry to assess changes in DNA content as the ploidy changes in response to the absence of fluconazole (No fluc), the presence of fluconazole (+fluc) and 1 day after subsequent removal of fluconazole (−fluc). The diploid control has twice the DNA content of the haploid and msh2Δ in the absence of fluconazole. The chromosome aneuploidy of strain msh2Δ B1 (chr1 and chr4) can be observed as an increase in DNA content compared with the haploid control; however, the increased aneuploidy of strain msh2Δ B2 (chr1) cannot be detected. In response to fluconazole, the ploidy of the haploid, diploid, msh2Δ and msh2Δ B1 increases. Some cells of the diploid do not increase ploidy resulting in two fluorescent peaks. No increase in ploidy observed in the msh2Δ B2 strain indicating it cannot undergo heteroresistance. When the fluconazole is removed, the haploid, diploid, msh2Δ and msh2Δ B1 strains revert back to their original ploidy.

Figure 6.

The msh2Δ fluconazole-resistant aneuploids show reduced relative fitness compared with WT in the absence of fluconazole. Competition assays with WT and the nourseothricin-resistant control (NAT⁺), msh2Δ::NAT, SACN00B1 or SACN00B2 cells showing the percentage of colonies derived from each original strain. Aneuploid strains SACN00B1 or SACN00B2 show decreased growth in competition with WT. Asterisks indicate statistical significance using a two-tailed Student’s t-test; *P < 0.05, **P < 0.005.