OUP user menu

Genomewide screening for genes involved in biofilm formation and miconazole susceptibility in Saccharomyces cerevisiae

Davy Vandenbosch, Evelien De Canck, Inne Dhondt, Petra Rigole, Hans J. Nelis, Tom Coenye
DOI: http://dx.doi.org/10.1111/1567-1364.12071 720-730 First published online: 1 December 2013

Abstract

Infections related to fungal biofilms are difficult to treat due to the reduced susceptibility of sessile cells to most antifungal agents. Previous research has shown that 1–10% of sessile Candida cells survive treatment with high doses of miconazole (a fungicidal imidazole). The aim of this study was to identify genes involved in fungal biofilm formation and to unravel the mechanisms of resistance of these biofilms to miconazole. To this end, a screening of a Saccharomyces cerevisiae deletion mutant bank was carried out. Our results revealed that genes involved in peroxisomal transport and the biogenesis of the respiratory chain complex IV play an essential role in biofilm formation. On the other hand, genes involved in transcription and peroxisomal and mitochondrial organization seem to highly influence the susceptibility to miconazole of yeast biofilms. Additionally, our data confirm previous findings on genes involved in biofilm formation and in general stress responses. Our data suggest the involvement of peroxisomes in biofilm formation and miconazole resistance in fungal biofilms.

Keywords
  • yeast
  • resistance
  • peroxisome

Introduction

Biofilms are microbial communities of cells attached to a surface and embedded in an extracellular polymeric matrix (Donlan & Costerton, 2002). Biofilms can be formed on both host tissues and artificial surfaces, including medical devices such as catheters and prostheses (Kojic & Darouiche, 2004; Mukherjee et al., 2005; Ramage et al., 2006). Fungal biofilm-related infections have increased in frequency and result in increased morbidity and mortality in immunocompromised patients (Wilson et al., 2003; Menzin et al., 2003). One of the major fungal human pathogens is Candida albicans. This opportunistic human commensal causes infections ranging from superficial mucous membrane infections to life-threatening systemic diseases. The increased resistance to antifungals of sessile compared with planktonic cells often prevents a successful therapy (Ramage et al., 2012). Production of extracellular matrix, increased cell density, upregulation of efflux pumps, decreased growth rate, overexpression of drug targets, and the presence of persister cells are known to play a role in the resistance of sessile cells (White, 1997; Baillie & Douglas, 1998; Mukherjee et al., 2003; Sanglard et al., 2003; Al-Fattani & Douglas, 2004; Mukherjee & Chandra, 2004; Perumal et al., 2007; Lewis, 2008, 2010; Ramage et al., 2012).

Azoles are widely used to treat fungal infections. These antifungal compounds decrease the production of ergosterol by interacting with cytochrome P450 and inhibiting the 14α-demethylation of lanosterol. As ergosterol is an important constituent of the cytoplasmatic membrane, treatment with azole antifungals leads to growth inhibition (Yoshida, 1988). Besides this fungistatic mechanism of action, recent data indicate a fungicidal effect for miconazole (an imidazole) against Candida spp. cells in suspension and in young and mature biofilms (Lamfon et al., 2004; Vandenbosch et al., 2010). Accumulation of reactive oxygen species (ROS) appears to be involved in this process, although it is likely that other mechanisms also account for the fungicidal activity (Kobayashi et al., 2002; François et al., 2006; Thevissen et al., 2007). Despite the observed fungicidal effect of miconazole on biofilms, 1–10% of the sessile C. albicans cells survive exposure to high levels of this antifungal agent (Vandenbosch et al., 2010).

The aim of the present study was to identify genes involved in fungal biofilm formation and to unravel the mechanisms of resistance of these biofilms to miconazole. To this end, we screened a deletion mutant bank of Saccharomyces cerevisiae for biofilm formation and miconazole susceptibility. Previous work indicated that S. cerevisiae forms biofilms and could be used as a model for fungal biofilm formation (Reynolds & Fink, 2001). Furthermore, the staining procedure used in this study has been optimized previously and showed to be a reliable method to measure biofilm formation (Peeters et al., 2008; Vandenbosch et al., 2012).

Materials and methods

Strains

The strains used in this study are S. cerevisiae BY4741, the BY4741-derived haploid set of deletion mutants in nonessential genes from the EUROSCARF collection (n = 4961), C. albicans SC5314, C. albicans Δpex4pex4 (Noble et al., 2010), and C. albicans Δpex8pex8 (Noble et al., 2010). Stock cultures of these strains were kept at −80 °C. Strains were cultured on Sabouraud dextrose agar (Oxoid, Hampshire, UK) at 37 °C for at least 48 h.

Screening

Suspensions of S. cerevisiae BY4741 and BY4741-derived deletion mutants, containing c. 107 cells mL−1, were prepared in yeast–peptone–dextrose medium (BD, Franklin Lakes, NJ). One hundred microlitre of these cell suspensions was added to the wells (12 replicates per strain) of a U-bottomed 96-well microtiter plate (SPL Lifesciences, Pocheon, Korea) to initiate biofilm formation. After 1-h incubation at 37 °C, the supernatant was removed, and the wells were rinsed with 100 μL of physiological saline (0.9% NaCl) to remove unattached cells. The microtiter plates were further incubated for 24 h at 37 °C after addition of 100 μL yeast–peptone–dextrose medium to each well. Subsequently, the supernatant was removed, and the mature biofilms were rinsed with 100 μL physiological saline before treatment with miconazole (1000 μg mL−1; Certa, Braine-l'Alleud, Belgium). For this purpose, 100 μL of a miconazole suspension in phosphate-buffered saline containing 2% DMSO (Sigma-Aldrich, St Louis, MO) was added to six biofilms of each strain and 100 μL of phosphate-buffered saline containing 2% DMSO to the other six biofilms (control). After 24-h incubation at 37 °C, the supernatant was removed, and 120 μL of a diluted resazurin solution (CellTiter-Blue 1:6 in physiological saline; Promega, Leiden, the Netherlands) was added to each well. Fluorescence was measured after 2-h incubation in the dark at 37 °C using an Envision microtiter plate reader (Perkin-Elmer; λex 535 nm; λem 590 nm). As six mutants (HTL1,CDC26,ACB1,YDJ41,CDC40, and SAM37) were not able to grow at 37 °C, the incubation temperature was lowered to 25 °C for these strains. Two mutants (ADH1 and SDS23) failed to grow under the conditions of our study and were consequently not included.

Mutants showing a significant difference in either biofilm formation or susceptibility to miconazole in the first screening were retested at least once, and the average result of all tests was calculated.

Calculations and statistical analysis

For each deletion mutant, relative values for biofilm formation and susceptibility to miconazole were calculated. Biofilm formation expressed as the average fluorescence of untreated biofilms (corrected for the blank) was compared between the wild type (WT) and each mutant. The WT strain was included on every plate, and results obtained with a particular mutant were always compared with the results obtained with the WT included on the same plate.

Values lower than 1 indicate decreased biofilm formation, and values higher than 1 indicate increased biofilm formation compared with the WT. The susceptibility to miconazole expressed as the ratio of the average fluorescence of miconazole-treated biofilms to untreated biofilms (both corrected for the blank) was compared between the WT and each mutant. Values lower than 1 indicate an increased susceptibility to miconazole, and values higher than 1 indicate a decreased susceptibility to miconazole compared with the WT. Statistical analysis was performed using the nonparametric Mann–Whitney U-test (spss Statistics 17.0 software). Results were considered significantly different when P < 0.01. While this may increase the number of false-positive results, we did not apply a correction for multiple testing, as we preferred to minimize the number of false-negative results.

Data processing

For a general overview of the data, mutants with significant differences in biofilm formation or susceptibility to miconazole were grouped using the Gene Ontology Slim Mapper (SGD project; http://www.yeastgenome.org; 22/01/2013) according to the biological processes in which the deleted genes are involved.

For a more in-depth analysis, genes knocked out in mutants that showed a significant difference in at least one phenotype were categorized using the Gene Ontology Term Finder (SGD project; http://www.yeastgenome.org; 22/01/2013), searching for significantly shared gene ontology (SSGO) terms (P < 0.01). The frequency of SSGO terms in our dataset was compared with that in the total genome of S. cerevisiae S288C.

Detection of ROS

Biofilms of S. cerevisiae and C. albicans were grown as described above. ROS accumulation was measured in a fluorometric assay using 2′,7′-dichlorofluorescein diacetate (DCFHDA; Invitrogen, Carlsbad, CA; Kobayashi et al., 2002). For this purpose, biofilms were incubated with 10 µM DCFHDA, simultaneously with the antifungal treatment. Fluorescence was measured after 0- and 6-h incubation using an Envision microtiter plate reader (Perkin-Elmer; λex 485 nm; λem 535 nm). Values obtained were corrected for background fluorescence (measured in the absence of cells) and compared with those obtained with untreated WT biofilms. ROS levels were quantified in triplicate on six biofilms (n = 18) for each strain.

Gene expression

Candida albicans biofilms were grown and treated as described above. Sessile cells were collected, and cell disruption, RNA purification, and DNase treatment were performed according to the manufacturers' instructions (RiboPure-Yeast kit; Applied Biosystems, Carlsbad, CA). The iScript cDNA Synthesis Kit (Bio-Rad, Hercules, CA) was used for the reverse transcriptase reaction. To this end, 1 μL reverse transcriptase and 4 μL reaction mix were added to each tube (5 min at 25 °C, 30 min at 42 °C, and 5 min at 85 °C). After development of a forward (FW) and a reverse (RV) primer for the genes PEX4 (FW: TTGTTAGACCAACCCGAGCCAGAC; RV: TTTGCTGCATCGATGTTCAACGGC), PEX8 (FW: AGCTTGGGTCCTCAAGGTAGAGC; RV: ATTTGGGGTGCCCAGCAAGG), LSC2 (FW: CGTCAACATCTTTGGTGGTATTGT; RV: TTGGTGGCAGCAATTAAACCT), and RPP2B (FW: TGCTTACTTATTGTTAGTTCAAGGTGGTA; RV: CAACACCAACGGATTCCAATAAA) and testing their specificity, real-time PCR (CFX96 Real-Time System, Bio-Rad) was performed using the Sso Advanced SYBR Green Supermix (Bio-Rad). The expression levels of the genes of interest were normalized using two stably expressed reference genes (LSC2 and RPP2B). Experiments were performed in triplicate and analyzed with the Bio-Rad CFX Manager software (Bio-Rad).

Results and discussion

General overview

A total of 4961 haploid S. cerevisiae deletion mutants were screened for biofilm formation and susceptibility to miconazole (Fig. 1 and Supporting Information, Table S1). Three hundred and forty-one mutants (6.9%) showed significant differences (P < 0.01) in biofilm formation compared with the WT. The majority (242/341) formed less biofilm as opposed to 99 mutants showing significantly increased biofilm formation. Three hundred and eighty-seven mutants (7.8%) exhibited significantly different (P < 0.01) susceptibility to miconazole compared with the WT, 136 mutants being more susceptible and 251 mutants being more resistant compared with the WT. Eighty-four mutants showed both decreased biofilm formation and decreased susceptibility, 37 mutants showed decreased biofilm formation and increased susceptibility, seven mutants showed increased biofilm formation and decreased susceptibility, and 18 mutants showed increased biofilm formation and increased susceptibility (Fig. @@@).

1

Overview of screening of Saccharomyces cerevisiae deletion mutant bank for biofilm formation and susceptibility to miconazole.

A decrease in biofilm formation was observed in the mutants, of which the deleted genes are involved in mitochondrial organization, protein complex biogenesis, and protein targeting. An increase in biofilm formation was observed mainly in the mutants, of which the deleted genes are involved in transcription, cell wall organization, and mitotic cell cycle (Table S2). Deletion of several genes involved in protein targeting, transcription, and response to stress increased the susceptibility to miconazole, while deletion of several genes involved in mitochondrial organization and protein complex biogenesis led to increased resistance to miconazole. The deletion of several genes involved in lipid metabolism also affected miconazole susceptibility, with about half of the mutants showing increased susceptibility and about half showing decreased susceptibility (Table S2).

Validation of screening

A total of 917 mutants (18.5%) were retested following the first screening because of a significant difference (P < 0.01) in biofilm formation or miconazole susceptibility compared with the WT. The results for 819 of these 917 retested mutants (89.3%) confirmed the previously observed significant differences (including 237 mutants that showed a significantly different phenotype at a lower level of significance; P < 0.05). Overall, these data indicate a good repeatability of our method and confirm the validity of our screening.

As a second measure of the validity of our screening, we compared our results with results previously obtained with selected C. albicans mutants. Deletion of SUN4, encoding a cell wall protein related to glucanases, led to decreased biofilm formation in S. cerevisiae. It was previously demonstrated that its ortholog in C. albicans,SUN41, is required for biofilm formation (Hiller et al., @@@; Norice et al., @@@). Similarly, our data show that SUV3 is required for biofilm development in S. cerevisiae, as was previously also reported for C. albicans (Richard et al., @@@).

Deletion of LCB4, a gene involved in sphingolipid biosynthesis, increased susceptibility to miconazole. We have previously shown that a heterozygous LCB4/lcb4 C. albicans deletion mutant was also hypersusceptible to miconazole (Vandenbosch et al., @@@).

In addition, our experiments identified several genes involved in general and azole-specific resistance in C. albicans biofilms. Deletion of ergosterol biosynthesis genes (ERG2,ERG4,ERG24, and ERG28) increased the susceptibility to miconazole, probably due to a decreased level of ergosterol, an important constituent of the cytoplasmatic membrane. The failure of adapting the sterol composition of the cytoplasmatic membrane as a mechanism to inhibit the penetration of miconazole in these mutants may also contribute to the observed phenotype. Induction of ROS production has been observed in planktonic and sessile Candida cells after miconazole treatment (François et al., @@@; Vandenbosch et al., @@@). This increased ROS production was preceded by changes in the actin cytoskeleton (Thevissen et al., @@@) and may contribute to the fungicidal action of miconazole. We identified three genes (SIT4,VPS1, and END3), which were involved in the organization of the actin cytoskeleton and which appear to play a role in the resistance to miconazole. Furthermore, superoxide dismutases are important for ROS detoxification, and their protective effect against miconazole-induced cell death in C. albicans biofilms was observed previously (Bink et al., @@@). Also, in the present study, a deletion of SOD1, encoding a zinc/copper superoxide dismutase, led to increased susceptibility to miconazole. Deletion of MXR1 resulted in a similar phenotype. The latter gene encodes methionine sulfoxide reductase and is known for its antioxidative capacities leading to increased survival of cells (Koc et al., @@@; Moskovitz, @@@). Trehalose, a disaccharide, is also known for its protective effect against ROS. Deletion of TPS2, encoding trehalose-6-phosphate phosphatase, leads to a decrease in trehalose levels and the accumulation of trehalose-6-phosphate to toxic levels (Alvarez-Peral et al., @@@; Van Dijck et al., @@@). Data from the present study suggest that TPS2 is also involved in resistance of yeast biofilms to miconazole.

Mutants showing decreased biofilm formation

The SSGO terms of the mutants with decreased biofilm formation could be divided in 10 categories (Table 1). The growth rate of several mutants within these categories was determined (Table 1). As only a minority of mutants showed a significant increase in doubling time (2 of 15 mutants tested), we can conclude that a defect in biofilm formation is not always linked to a lower growth rate and that a lower metabolic activity in the mutants is not directly involved in decreased biofilm formation.

View this table:
1

Frequency of SSGO terms in the results database compared with the overall frequency in the Saccharomyces cerevisiae genome for mutants with decreased biofilm formation, increased susceptibility, decreased susceptibility, increased biofilm formation and increased susceptibility, decreased biofilm formation and decreased susceptibility, and decreased biofilm formation and increased susceptibility

Frequency (%)
Results databaseS. cerevisiae genome
Decreased biofilm formation
Mitochondrial organization16.14.7
Mitochondrial translation9.51.9
Organelle organization30.617.0
Protein complex biogenesis10.33.3
Mitochondrial respiratory chain complex IV biogenesis3.30.4
Post-transcriptional regulation of gene expression6.21.5
Cellular component organization or biogenesis41.728.3
Cellular protein complex assembly7.92.4
Peroxisomal transport2.90.3
Single-organism cellular process59.545.8
Increased susceptibility
Vacuolar transport10.32.0
Single-organism process69.147.8
Organic substance transport21.38.2
Regulation of transcription by glucose3.70.2
Response to stress22.19.0
Biological regulation36.019.2
Peroxisome organization6.60.9
Cellular component organization40.423.4
Macromolecule localization19.98.1
Response to stimulus27.213.2
Protein transport14.04.5
Regulation of biological quality14.75.0
Decreased susceptibility
Mitochondrial respiratory chain complex assembly3.60.5
Post-transcriptional regulation of gene expression6.01.5
Increased biofilm formation and increased susceptibility
tRNA wobble base modification16.70.4
Decreased biofilm formation and decreased susceptibility
Mitochondrial organization20.24.7
Post-transcriptional regulation of gene expression10.71.5
Decreased biofilm formation and increased susceptibility
Proteasome assembly13.50.4
Ubiquinone biosynthetic process10.80.2
View this table:
2

Doubling times of mutants with decreased biofilm formation

Doubling time (min)STDEV (min)
WT 13722
∆ atg11 12526
∆ coq1 1547
∆ coq4 14212
∆ coq5 163*9
∆ csg2 1104
∆ fis1 13117
∆ pex1 18172
∆ pex10 17048
∆ pex12 15927
∆ pex13 201*70
∆ pex5 15743
∆ pex6 16451
∆ slt2 203113
∆ vps15 13416
∆ ydc1 1134
  • Significantly (P < 0.05) increased doubling times compared to the WT are marked with an asterisk.

Compared with the overall prevalence in the genome, genes involved in peroxisomal transport were overrepresented among the mutants showing decreased biofilm formation (10-fold more compared with the total genome). Pex1, Pex5, Pex6, Pex10, Pex12, and Pex13 are involved in the peroxisomal matrix protein import, and deletion of the corresponding genes led to decreased biofilm formation. Pex5 is the receptor, which binds to the peroxisomal proteins in the cytosol. A set of peroxins, including the ones mentioned before, is anchored in the peroxisomal membrane and is involved in the binding of the receptor–protein complex, its dissociation, the uptake of the peroxisomal protein, and the release of the Pex5 receptor into the cytosol (Wolf et al., @@@). Furthermore, Pex3, of which the deficiency also led to decreased biofilm formation in S. cerevisiae, is required for the proper localization of peroxisomal membrane proteins (Hettema et al., @@@). Peroxisomes are organelles with a single membrane in which β-oxidation of fatty acids takes place (Tabak et al., @@@; Trotter, @@@). The gene CAT2, encoding the enzyme for transport of acetyl units (products of the β-oxidation), is known to play a role in biofilm formation (Strijbis et al., @@@). Peroxisomes also contain antioxidative systems to neutralize ROS produced during metabolism (del Rio et al., @@@). As β-oxidation in yeast cells exclusively takes place in peroxisomes (Tabak et al., @@@; Trotter, @@@), it is not surprising that mutations in peroxins may lead to a disordered balance in lipid composition and may therefore change the composition of cellular membranes. Differences in the distribution of lipids between planktonic and sessile cells have been described previously. Furthermore, lipids are important for adhesion and play a critical role in the formation of biofilms (Ghannoum et al., @@@; Lattif et al., @@@).

Many mutants in which genes involved in mitochondrial organization are deleted showed significantly decreased biofilm formation. Particularly, genes involved in the biogenesis of the respiratory chain complex IV (PET54,PET100,PET122,COX12,COX14,COX16,COX20, and MSS51) were overrepresented in this group (eightfold more prevalent than in the total genome). Complex IV, or cytochrome c oxidase, is the final enzyme in the electron transport chain, creating the proton gradient necessary for ATP production. Mutants in which PET54,PET100, or PET122 is deleted are known to form so-called ‘petite’ colonies. Petite mutants often arise spontaneously during growth. The partial or complete loss of mitochondrial DNA leads to a lower growth rate and the formation of small colonies (Baruffini et al., @@@). It is likely that the reduced growth rate of these mutants is responsible for the decreased biofilm formation. COX12 plays a role in the assembly of complex IV, but does not seem to be required for its function (Vogtle et al., @@@). For this reason, the mechanism by which COX12 influences the biofilm formation remains unclear. The deletion of COX14, COX16, and MSS51 leads to a defect in respiratory growth; the function of COX16 is unknown (Barrientos et al., @@@). The decreased ATP production in these mutants is likely to result in a lack of energy for the formation of a dense biofilm.

In S. cerevisiae, the flocculin gene family, including FLO1,FLO5,FLO9,FLO10, and FLO11, encodes cell wall proteins, which are important for cell–cell adhesion. The latter gene is also involved in cell surface adhesion. The first group of FLO proteins (encoded by FLO1,FLO5, and FLO9) acts as lectins, while Flo10 and Flo11 confer adhesion by increased cell surface hydrophobicity (Verstrepen & Klis, @@@; Van Mulders et al., @@@; Bruckner & Mosch, @@@). S. cerevisiae strain BY4741 used in the present study, a derivative of S288c, expresses low levels of FLO11 and therefore has reduced biofilm-forming capacity (Purevdorj-Gage et al., @@@). Nevertheless, confocal laser scanning microscopy images of the biofilms in our study showed a biofilm-like morphology, and washing with PS did not affect this structure (Supporting Information, Fig. S1). The flocculin mutants did not show a change in biofilm formation compared with the WT under the conditions of our study (Fig. S1), supporting the idea that one FLO gene can compensate for the absence of another (Guo et al., @@@).

The transcriptional network controlling biofilm formation is well characterized in C. albicans and consists of six regulators: Tec1, Efg1, Ndt80, Rob1, Brg1, and Bcr1 (Nobile et al., @@@). However, there are differences in the controlled target genes (Tec1 and Efg1) or differences in the function (Ndt80) between the orthologs in C. albicans and S. cerevisiae. For Rob1 and Brg1, the regulatory function is only detectable in species closely related to C. albicans. Finally, Bcr1 orthologs have not been found in S. cerevisiae (Nobile et al., @@@). None of the S. cerevisiae orthologs of these C. albicans genes seems to be involved in biofilm formation.

Mutants showing increased biofilm formation

Only 99 mutants showed a significantly increased biofilm formation, and no SSGO terms could be detected within this category, indicating the involvement of a broad variation of biological processes. Furthermore, 25% of the deleted genes were associated with unknown biological processes. We hypothesize that the deletion of genes leading to increased biofilm formation disturbs the global metabolic balance and that other pathways compensate as reaction. This idea is also supported by the presence of a relatively high number of genes involved in transcription within this group of 99 mutants. Their deletion influences a variety of other cellular metabolic pathways, and probably, a combination of factors is responsible for the observed phenotype.

Mutants showing increased susceptibility to miconazole

The SSGO terms of the mutants with increased susceptibility to miconazole could be divided in 12 categories (Table 1). Compared with the overall prevalence in the genome, we observed that genes involved in regulation of transcription by glucose were overrepresented among the mutants showing increased susceptibility to miconazole (18-fold more compared with the total genome). Although only five genes are clustered in this group (NRG2,TUP1,VPS36,SNF8, and GCR1), they seem to be highly involved in the resistance to miconazole. As all of these genes regulate the expression of several other genes involved in various pathways, it is likely that a combination of factors account for the increased susceptibility to miconazole. Tup1, a general repressor forming a complex with Ssn6, regulates the expression of more than 300 genes involved in metabolic processes, transport, meiosis, cell wall organization, stress responses, and transcription (Smith & Johnson, @@@; Green & Johnson, @@@; Malave & Dent, @@@). Gcr1 is a transcriptional activator that coordinates the expression of several glycolytic enzymes (Holland et al., @@@; Lopez & Baker, @@@). Nrg2 is known to repress a large number of stress responsive genes. However, Nrg1 (a paralog of Nrg2) is more important for this regulation. Furthermore, genes involved in mitochondrial organization, carbon and nitrogen signaling, cell wall organization, mating, and transcription are also known to be Nrg-repressed (Vyas et al., @@@). Finally, SNF8 and VPS36 derepress the transcription of SUC2, encoding a sucrose-hydrolyzing enzyme (Kamura et al., @@@). Vps36 and Snf8 are both components of the ESCRT-II (endosomal sorting complex required for transport) complex, which is involved in protein sorting and the biogenesis of multivesicular bodies. This complex also regulates the formation of ESCRT-III (Teo et al., @@@; Hurley, @@@; Wollert & Hurley, @@@). Interestingly, defects in the ESCRT machinery lead to the accumulation of receptors and transporters in the cytoplasmatic membrane (Bugnicourt et al., @@@), which possibly results in an enhanced uptake of miconazole and a consequent increase in susceptibility.

A second cluster of genes overrepresented among hypersusceptible mutants (sevenfold more prevalent than in the total genome) was involved in peroxisomal organization: PEX2,PEX4,PEX8,PEX34,VPS1,VPS15,SLT2,FIS1, and ATG11. Three of these genes (PEX2,PEX4, and PEX8) have a function in the peroxisomal matrix protein import machinery (Wolf et al., @@@), which was also found to be involved in biofilm formation. VPS15,SLT2, and ATG11 are involved in the degradation of peroxisomes. Their corresponding mutants are deficient in pexophagy (Yorimitsu & Klionsky, 2003; Manjithaya et al., @@@). In contrast, VPS1,FIS1, and PEX34 are involved in determining the number of peroxisomes per cell (Kuravi et al., @@@; Tower et al., @@@). It seems contradictory that genes involved, on the one hand, in the degradation of peroxisomes and, on the other hand, in the development of peroxisomes play a role in the resistance to miconazole. However, this probably indicates that a balanced level of peroxisomes is crucial for sessile cells. A low number of peroxisomes may disturb the lipid housekeeping and the antioxidative capacities of the cells. A decrease in the degradation of peroxisomes is known to induce intracellular protein aggregation, and this affects the catalase activity and consequently increases ROS levels (Aksam et al., @@@; Saraya et al., @@@). As the induction of ROS has previously been linked to the fungicidal activity of miconazole (François et al., 2006; Vandenbosch et al., 2010), this may contribute to the hypersusceptibility of these peroxisomal mutants. Additional experiments confirmed this hypothesis, as all S. cerevisiae mutants within this category showed increased ROS levels upon miconazole treatment compared with the WT, ranging from a 1.5- to 5.0-fold increase (Fig. 2). Only minor changes were observed in the ROS levels of untreated biofilms of the S. cerevisiae mutants compared with the WT (1.0- to 3.0-fold increase).

2

ROS production in Saccharomyces cerevisiae biofilms with and without miconazole treatment. The results were relatively expressed compared with the untreated WT biofilm (100%). Experiments were performed in triplicate on six biofilms. Error bars represent the standard error of the mean.

Mutants showing decreased susceptibility to miconazole

Within the large group of mutants (250) with decreased susceptibility to miconazole, only two categories with SSGO terms could be distinguished; that is, genes involved in mitochondrial respiratory chain complex assembly and genes involved in post-transcriptional regulation of gene expression, respectively (Table 1). This indicates that the majority of the genes in this group play a role in a broad range of different biological processes. The genes involved in the assembly of the mitochondrial respiratory chain complex (CBP6,MZM1,EMI1,COX12,COX14,COX16,COX19,COA6, and CBP3) were seven times more prevalent among the mutants showing decreased miconazole susceptibility than in the total S. cerevisiae genome. Genes belonging to this category play a role in the production of ATP under aerobic conditions. Their deletion may lead to decreased ATP levels and as a result a dormant state of the cells. Probably, this mechanism is important for making these mutants more resistant to miconazole, as observed in previous research (LaFleur et al., 2006; Lewis, 2007, 2010, 2012). The second category of genes, of which deletion led to decreased susceptibility to miconazole, is involved in post-transcriptional regulation and therefore affecting the expression of many other genes. Two main groups may be distinguished: genes related to mitochondrial organization (CBP6,CBS2,ICP55,COX14, and ATP25) and genes related to ribosomal organization (RPS9B,RPL31A, and ASC1). As they have a global regulatory effect, it is not clear which mechanisms exactly contribute to the observed decrease in susceptibility to miconazole.

Mutants affected in biofilm formation and miconazole susceptibility

One hundred and forty-six mutants were affected in both biofilm formation and miconazole susceptibility. SSGO terms could not be defined for all mutants within each combination (Table 1). Most mutants affected in both phenotypes were found to exhibit decreased biofilm formation and increased resistance (84 mutants). This is not completely unexpected, as genes involved in mitochondrial organization in this group were highly represented (fourfold more prevalent than in the total genome), and it is likely that a lower ATP level both decreases biofilm formation and increases resistance by inducing dormancy. A second significantly overrepresented group of mutants with decreased biofilm formation and increased resistance to miconazole have defects in post-transcriptional regulation of gene expression. Our results also reveal that genes involved in peroxisomal organization play an important role both in the formation of biofilms and the resistance to miconazole, although they do not have significantly shared genes for this biological process: biofilm formation is regulated mainly by genes related to peroxisomal transport, while the resistance of sessile cells to miconazole is particularly induced by genes involved in the general peroxisomal organization.

Within the group of 18 mutants with increased biofilm formation and increased susceptibility to miconazole, transfer RNA (tRNA) wobble uridine modification is a SSGO term represented by SIT4,IKI3, and TRM9. These modifications play an important role in the folding and stability of tRNA. Furthermore, they are necessary for an accurate and efficient translation, and they have recently been linked to the control of gene expression in response to stress (Huang et al., @@@; El Yacoubi et al., @@@).

Finally, we found two SSGO terms for mutants with decreased biofilm formation and increased susceptibility, that is, the biosynthesis of ubiquinone, a component of the electron transport chain important for ATP production by respiration, and the assembly of proteasomes, protein complexes responsible for the degradation of unneeded or damaged proteins. As many other genes involved in mitochondrial organization showed decreased biofilm formation, probably due to a dormant state induced by lower ATP production, it is very likely that a decrease in ubiquinone may also contribute to this phenomenon. In contrast to our previous hypothesis that this dormant state leads to increased resistance, deletion of four genes (COQ1,COQ4,COQ5, and COQ8) in the ubiquinone biosynthetic pathway increases the susceptibility to miconazole. The antioxidative property of the reduced form of ubiquinone (Do et al., @@@; Soballe & Poole, @@@, @@@) may be important for the latter phenomenon as a decrease in ubiquinone may diminish the protection against miconazole-induced ROS. Also, proteasomes are involved in the response to oxidative stress (Shang & Taylor, @@@). Deletion of the genes involved in proteasome assembly (UMP1,PRE9,NAS2,IRC25, and POC4) may impair the proteasome activity in the mutants and may consequently contribute to the hypersusceptibility to miconazole. Inactivation of proteasomes has previously been shown to decrease the growth rate of C. albicans and to inhibit its biofilm formation (Evensen & Braun, @@@).

The involvement of peroxisomes in C. albicans biofilms

Results of the screening of the S. cerevisiae deletion mutant bank indicated that peroxisomes are important for both biofilm formation and miconazole resistance. For this reason, we performed additional experiments to investigate the role of these organelles in the human pathogen C. albicans.

Similar to S. cerevisiae, mutant biofilms of C. albicans Δpex4pex4 and Δpex8pex8 were hypersusceptible to miconazole compared to the WT (relative values of 0.65 and 0.48, respectively). In contrast, no decrease in the MIC for miconazole was observed for both strains compared with the WT, suggesting that the hypersusceptibility of both strains is biofilm specific. Biofilms of both mutants also showed an increased ROS level upon miconazole treatment, which was more pronounced for C. albicans Δpex8pex8 (threefold increase compared to WT) than for C. albicans Δpex4pex4 (twofold increase compared to WT). Miconazole treatment of C. albicans WT biofilms led to a twofold overexpression of the genes PEX4 and PEX8 compared to untreated biofilms. The susceptibility to fluconazole (included as negative control) of the biofilms of both C. albicans mutants was similar to that of the C. albicans WT (relative values of 1.02 for C. albicans Δpex4pex4 and 1.15 for C. albicans Δpex8pex8), and fluconazole treatment did not induce ROS. Taken together, these additional experiments confirm the biofilm specific importance of peroxisomes in miconazole resistance of S. cerevisiae and C. albicans biofilms.

Conclusion

A large number of genes were found to be involved in biofilm formation and drug resistance in S. cerevisiae, indicating the complexity of both processes. The validity of the screening was confirmed by the identification of genes previously observed to be involved in biofilm formation and drug resistance in C. albicans. Peroxisomal transport and mitochondrial organization appear to be important for yeast biofilm formation. Additionally, genes involved in transcription and peroxisomal and mitochondrial organization influence the susceptibility to miconazole. Peroxisomes were also found to be important for miconazole resistance in the human pathogen C. albicans, and this may offer perspectives for the treatment of fungal biofilm-related infections. Still, a considerable number of genes identified in this study are associated with unknown biological processes, requiring further research.

Supporting Information

Additional Supporting Information may be found in the online version of this article:

Fig. S1. Confocal laser scanning microscopy images of S. cerevisiae biofilms of untreated WT, untreated FLO11 mutant, miconazole treated FLO11 mutant and miconazole treated PEX8 mutant.

Table S1. Phenotype of 4961 haploid S. cerevisiae deletion mutants when screened for biofilm formation and susceptibility to miconazole.

Table S2. Frequency of genes, categorized according their biological processes involved.

Acknowledgements

This research has been funded by ‘Fonds Wetenschappelijk Onderzoek – Vlaanderen’ (FWO) and by the Interuniversity Attraction Poles Programme initiated by the Belgian Science Policy Office. We would like to thank Bruno Cammue and Karin Thevissen of the Centre of Microbial and Plant Genetics (CMPG) of KU Leuven for providing the S. cerevisiae BY4741 deletion mutant bank, C. albicans Δpex4pex4, and C. albicans Δpex8pex8. We would like to thank Jolien De Sadeleer for excellent technical assistance.

Footnotes

  • Editor: Richard Calderone

References

View Abstract