Identification of small molecules that disrupt vacuolar function in the pathogen Candida albicans

Abstract

The fungal vacuole is a large acidified organelle that performs a variety of cellular functions. At least a sub-set of these functions are crucial for pathogenic species of fungi, such as Candida albicans, to survive within and invade mammalian tissue as mutants with severe defects in vacuolar biogenesis are avirulent. We therefore sought to identify chemical probes that disrupt the normal function and/or integrity of the fungal vacuole to provide tools for the functional analysis of this organelle as well as potential experimental therapeutics. A convenient indicator of vacuolar integrity based upon the intracellular accumulation of an endogenously produced pigment was adapted to identify Vacuole Disrupting chemical Agents (VDAs). Several chemical libraries were screened and a set of 29 compounds demonstrated to reproducibly cause loss of pigmentation, including 9 azole antifungals, a statin and 3 NSAIDs. Quantitative analysis of vacuolar morphology revealed that (excluding the azoles) a sub-set of 14 VDAs significantly alter vacuolar number, size and/or shape. Many C. albicans mutants with impaired vacuolar function are deficient in the formation of hyphal elements, a process essential for its pathogenicity. Accordingly, all 14 VDAs negatively impact C. albicans hyphal morphogenesis. Fungal selectivity was observed for approximately half of the VDA compounds identified, since they did not alter the morphology of the equivalent mammalian organelle, the lysosome. Collectively, these compounds comprise of a new collection of chemical probes that directly or indirectly perturb normal vacuolar function in C. albicans.

Fig 1. C. albicans ade2 mutant produces a fluorescent pigment that localizes within the fungal vacuole.

(A) ADE2^+/+ (SC5314), ade2^-/- (CAI8LUX1), ade2^-/-ypt72Δ/Δ (C8ypt72Δ/Δ) and ade2^-/-vps11Δ/Δ (C8vps11Δ/Δ) mutant strains were streaked to YPD agar and incubated for 48 hours prior to imaging. (B) ADE2^+/+ and ade2^-/- strains were grown in YNB + 5 μg/ml adenine overnight, stained with CMAC and examined by fluorescent microscopy using TRIT-C and DAPI filter sets. (C) ade2^-/- (WT), ade2^-/-ypt72Δ/Δ and ade2^-/-vps11Δ/Δ mutant strains were grown in YNB + 5 μg/ml adenine overnight and cells examined by fluorescent microscopy using TRIT-C filter sets.

Fig 2. Pigmentation of the ade2^-/- mutant can form the basis of a 96-well plate based screening assay of vacuolar integrity.

(A) The indicated strains were suspended at 1 x 10⁵ cells/ml of YNB + 5 μg/ml adenine medium, before 200 μl aliquots were dispensed into the wells of a 96-well plate. The wells were imaged after 48 hrs incubation at 30°C. (B) A spectral scan was performed on the wells of the 96-well plate described in part (A) above, with excitation at 488 ± 10 nm, and fluorescence emission read at the indicated wavelengths. At each wavelength, relative fluorescence is normalized for growth, as measured by OD_600nm. (C) The indicated strains (all ADE2⁺) were growth in YNB medium in the wells of a 96-well plate, stained with CMAC, and then CMAC fluorescence detected following excitation at 354 nm and emission at 469 nm. CMAC uptake is normalized for growth, as measured by OD_600nm, and is the mean of three independent experiments. * p < 0.0001 versus the WT control strain.

Fig 3. Several VDAs induce abnormal vacuolar morphologies.

(A) Bio-imaging cellomics analysis of vacuole number and size and cell count. The C. albicans strain GP100, expressing cytoplasmic mCherry and vacuolar membrane GFP-YPT72, was exposed to each VDA at 5 and 25 μM at 30°C. The numbers of cells and the number and size of the vacuoles were determined for each individual cell present in 16 different fields in each well, and that for 6 wells for each compound. The Spot detector BioApplication (HCS Studio^™ Cell Analysis Software) was optimized to identify each fungal cell via the mCherry channel, and the number and size of each vacuole, within each cell, via the GFP channel. Numbers in brackets indicate the fold variation relative to DMSO treated cells. Only significant decrease or increase is shown (p-value < 0.0001). (B) Confocal imaging of cells expressing GFP-YPT72 and treated with 0.5% DMSO, diflunisal at 50 μM, or monensin at 12.5 μM for 24h at 30°C. Images were acquired using a Zeiss LSM 710 laser scanning confocal microscope.

Fig 4. Hyphal growth assay in presence of the 3 NSAIDs in the VDA collection.

Hyphal growth was imaged after 24h of growth at 37°C in low glucose conditions using a strain expressing GFP-YPT72. Cells were challenged with 0.5% DMSO (A), 50 μM diflunisal (B), 50 μM tolfenamic acid (C), and 50 μM mefenamic acid (D). Images were acquired using a BioTek Synergy imaging reader at a 60X magnification for bright field imaging (left panels) and fluorescence (eGFP excitation filter of 485/20; right panels). The error bars represent 20 microns.

Fig 5. Lysosome cellomics analysis.

(A) Bio-imaging cellomics analysis of lysosome number and size, and HaCat cell count. Numbers in brackets indicate the fold variation relative to DMSO treated cells. Only significant decrease or increase are shown, and were obtained from 6 replicates (p-value < 0.0001). Compounds were tested at 5, 25 and 50 μM. The minimal concentration causing abnormal lysosome number and/or size is indicated for each of the active compounds. Fluconazole was used as a control, and the order of the compounds relates to the order in Fig 3. (B) Immunofluorescence was performed using DAPI to detect the cell nucleus and anti-LAMP1 antibody to detect lysosomes. Methyl benzethonium chloride was tested at 5 μM and monensin at 50 μM.