Comparative genomic and transcriptomic analyses unveil novel features of azole resistance and adaptation to the human host in Candida glabrata

Comparative genomic and transcriptomic analyses unveil novel features of azole resistance and... Abstract The frequent emergence of azole resistance among Candida glabrata strains contributes to increase the incidence of infections caused by this species. Whole-genome sequencing of a fluconazole and voriconazole-resistant clinical isolate (FFUL887) and subsequent comparison with the genome of the susceptible strain CBS138 revealed prominent differences in several genes documented to promote azole resistance in C. glabrata. Among these was the transcriptional regulator CgPdr1. The CgPdr1 FFUL887 allele included a K274Q modification not documented in other azole-resistant strains. Transcriptomic profiling evidenced the upregulation of 92 documented targets of CgPdr1 in the FFUL887 strain, supporting the idea that the K274Q substitution originates a CgPdr1 gain-of-function mutant. The expression of CgPDR1K274Q in the FFUL887 background sensitised the cells against high concentrations of organic acids at a low pH (4.5), but had no detectable effect in tolerance towards other environmental stressors. Comparison of the genome of FFUL887 and CBS138 also revealed prominent differences in the sequence of adhesin-encoding genes, while comparison of the transcriptome of the two strains showed a significant remodelling of the expression of genes involved in metabolism of carbohydrates, nitrogen and sulphur in the FFUL887 strain; these responses likely reflecting adaptive responses evolved by the clinical strain during colonisation of the host. antifungal resistance, Candida glabrata, comparative genomics and comparative transcriptomics, CgPdr1 INTRODUCTION An alarming increase in the incidence of infections caused by Candida glabrata has been reported in the last years (Fidel, Vazquez and Sobel 1999; Khatib et al.2016). As such, in most epidemiological surveys this species ranks as the second major causative agent of invasive fungal infections worldwide, behind C. albicans and in some cases it is even the most frequently isolated species (Fidel, Vazquez and Sobel 1999; Pfaller et al.2010). This increase in the incidence of infections caused by C. glabrata is believed to result from its naturally high resilience to azoles, the frontline antifungal therapy used to treat candidiasis (Fidel, Vazquez and Sobel 1999; Pfaller et al.2010). The remarkably high rate at which C. glabrata strains acquire resistance to azoles, higher than the one registered for any other Candida spp., is another key factor contributing for the emergence of resistant strains (Borst et al.2005; Pfaller et al.2013; Kolaczkowska and Kolaczkowski 2016). Mutations in Erg11, the enzyme targeted by azoles, is not a primary mechanism of resistance in C. glabrata, in contrast with what is observed to occur for other Candida species (Sanguinetti et al.2005; Tsai et al.2010; Kolaczkowska and Kolaczkowski 2016). This observation suggests that C. glabrata has evolved other mechanisms to cope with azole stress, in particular through the activity of multidrug resistance (MDR) transporters (Sanguinetti et al.2005; Kolaczkowska and Kolaczkowski 2016). Specifically, several reports have underlined the important role of the ABC drug-efflux pumps CgCDR1 and CgPDH1 in contributing to azole resistance both in both laboratory strains and in resistant clinical isolates (Sanglard et al.1999; Vermitsky and Edlind 2004; Sanguinetti et al.2005; Torelli et al.2008; Kolaczkowska and Kolaczkowski 2016). More recently, the involvement of MDR transporters belonging to the major facilitator superfamily in azole resistance of C. glabrata has also been described (Costa et al.2014). The transcriptional regulation of drug-efflux pumps in C. glabrata, as in other yeasts, relies on the activity of the well-organised regulatory pleiotropic-drug resistance (PDR) network (Kolaczkowska and Kolaczkowski 2016). In C. glabrata, the key regulator of the PDR network is the transcription factor CgPdr1 and, concomitantly, this has been demonstrated to play an essential role in conferring tolerance to azoles (Vermitsky and Edlind 2004; Vermitsky et al.2006), including in resistant clinical isolates (Sanglard et al.1999; Sanguinetti et al.2005; Tsai et al.2006, 2010; Ferrari et al.2009; Caudle et al.2011). CgPdr1 has been implicated in the regulation of the drug-efflux pumps CgCDR1, CgPDH1, CgQDR2 and CgYOR1 (Kolaczkowska and Kolaczkowski 2016). Other azole-responsive genes regulated by CgPdr1 have biological functions related to stress response, metabolism of fatty acids and sterols, transcriptional regulation and adhesion (Vermitsky et al.2006; Ferrari et al.2009, 2011; Tsai et al.2010; Caudle et al.2011). Analysis of the coding sequence of the CgPDR1 gene in susceptible and resistant clinical isolates identified a panoply of gain-of-function (GOF) mutations that are believed to constitutively activate the transcription factor resulting in enhanced C. glabrata azole resistance (Vermitsky and Edlind 2004; Vermitsky et al.2006; Torelli et al.2008; Ferrari et al.2009; Caudle et al.2011). In this work, the genome and transcriptome of the reference strain C. glabrata CBS138 and of a clinical isolate (named FFUL887) recovered along the course of an epidemiological survey undertaken in hospitals of the Lisbon area herein demonstrated to be resistant to voriconazole and fluconazole (while the CBS138 strain was found to be sensitive) were compared. Besides providing clues to the mechanisms underlying acquisition of azole resistance inside the host, the results may also contribute for a better understanding of the different responses evolved by C. glabrata in human colonisation. METHODS Strains and growth media The strains used in this work are listed in Table 1. In addition to the laboratory strains listed, a cohort of 58 Candida glabrata clinical isolates recovered from patients attending three major Hospitals of the Lisbon area between 2000 and 2008 was also used (Table S1, Supporting Information). The strains were cultivated in rich growth media Yeast Peptone Dextrose (YPD), in RPMI (Roswell Park Memorial Institute Medium) or in minimal medium (MM). YPD contains, per litre, 20 g glucose (Merck Millipore), 10 g yeast extract (HiMedia Laboratories, Mumbai, India) and 20 g Peptone (HiMedia Laboratories). RPMI contains, per litre, 20.8 g RPMI-1640 synthetic medium (Sigma), 36 g glucose (Merck Millipore), 0.3g of L-glutamine (Sigma) and 0.165 mol/L of MOPS (3-(N-morpholino) propanesulfonic acid, Sigma). MM contains, per litre, 20 g glucose (Merck Millipore), 2.65 g (NH4)2 SO4 (Merck Millipore) and 1.7 g of yeast nitrogen base without amino acids and without ammonium sulphate (Difco). YPD and MM medium were sterilised by autoclaving for 15 min at 121°C and 1 atm, while RPMI medium was filtered with a 0.22-μm pore size filter and preserved at 4°C until further use. Solid media was prepared supplementing the corresponding liquid media with 2% agar. Table 1. List of strains used in this study. Strain  Parent  Description  Reference  KUE100  2001H  Parent strain, histidine auxotroph, the recipient enable high efficient gene targeting in which yku80 is repressed with a SAT1 flipper  Ueno et al. (2011)  KUE100_ΔCgPdr1  KUE100  ΔCgPDR1 strain, CgPDR1 (CAGL0A00451g) was replaced with the CgHIS3 marker  This study  CBS138  –  Reference strain  CBS-KNAW Fungal Biodiversity Centre  Strain  Parent  Description  Reference  KUE100  2001H  Parent strain, histidine auxotroph, the recipient enable high efficient gene targeting in which yku80 is repressed with a SAT1 flipper  Ueno et al. (2011)  KUE100_ΔCgPdr1  KUE100  ΔCgPDR1 strain, CgPDR1 (CAGL0A00451g) was replaced with the CgHIS3 marker  This study  CBS138  –  Reference strain  CBS-KNAW Fungal Biodiversity Centre  View Large Table 1. List of strains used in this study. Strain  Parent  Description  Reference  KUE100  2001H  Parent strain, histidine auxotroph, the recipient enable high efficient gene targeting in which yku80 is repressed with a SAT1 flipper  Ueno et al. (2011)  KUE100_ΔCgPdr1  KUE100  ΔCgPDR1 strain, CgPDR1 (CAGL0A00451g) was replaced with the CgHIS3 marker  This study  CBS138  –  Reference strain  CBS-KNAW Fungal Biodiversity Centre  Strain  Parent  Description  Reference  KUE100  2001H  Parent strain, histidine auxotroph, the recipient enable high efficient gene targeting in which yku80 is repressed with a SAT1 flipper  Ueno et al. (2011)  KUE100_ΔCgPdr1  KUE100  ΔCgPDR1 strain, CgPDR1 (CAGL0A00451g) was replaced with the CgHIS3 marker  This study  CBS138  –  Reference strain  CBS-KNAW Fungal Biodiversity Centre  View Large Assessment of resistance to antifungals To assess the resistance of the C. glabrata clinical isolates and of the reference strain CBS138 to voriconazole, fluconazole, anidulafungin and caspofungin, the MIC50 (minimum inhibitory concentration) of each of the antifungals was estimated using the microdilution method recommended by EUCAST (Subcommittee on Antifungal Susceptibility Testing (AFST) of the ESCMID European Committee for Antimicrobial Susceptibility Testing (EUCAST) 2003). The concentrations of fluconazole tested ranged from 0.125 to 64 mg/L and for the remaining antifungals ranged from 0.015 to 8 mg/L. MIC50 value was taken as being the first concentration of antifungal that reduced growth of the strains to half that registered in drug-free medium, as defined by EUCAST (2003). The MIC50 of each antifungal determined for the strains was compared with the clinical breakpoints recommended by EUCAST (32 mg/L for fluconazole, 0.06 mg/L for anidulafungin) to classify the strains as resistant (the MIC50 determined is above the defined breakpoint), susceptible (MIC50 equal or below the breakpoint) or intermediate (MIC50 shows represent a dose-dependent sensitivity according to EUCAST). Since a breakpoint value is not yet established for voriconazole in C. glabrata, the epidemiological cut-off value (1 mg/L) was used to detect non-WT isolates as recommended (Pfaller et al.2011). The stock solutions of the antifungals were prepared from the powder and using DMSO (dimethyl sulfoxide, Sigma) as the solvent. Fluconazole was purchased from Sigma, voriconazole and anidulafungin were kindly provided by Pfizer and caspofungin was provided by Merck Sharp Dohme. Growth curves of the C. glabrata reference strain CBS138 and of the FFUL887 isolate in the presence of fluconazole and voriconazole were performed using essentially the same experimental setup described above for the microdilution method with the difference that instead of measuring OD of the culture only after 24 h, this was measured every 30 min during 42 h at OD590nm. FFUL887 genomic DNA extraction and whole-genome sequencing FFUL887 cells were cultivated in YPD growth medium (up to an OD600nm of ∼3.0) and then centrifuged at 5000 rpm for 5 min at 4°C. The pellet was resuspended in 1 ml solution A (1 M Sorbitol (Sigma); 0.1 M EDTA (tetrasodium salt dehydrate) at pH 7.5). Afterwards, 10 mg/mL zymolyase (Zymo research) was added to the cellular suspension and the solution was incubated at 37˚C until protoplast formation. The suspension was centrifuged at 5000 rpm for 5 min and the pellet was resuspended in 1 mL solution B (50 mM Tris-HCL at pH 7.4 (Sigma-Aldrich), 20 mM EDTA). After this step, 30 μL of SDS (10%) was added to the mixture and this was left for 30 min at 65°C. Potassium acetate (250 μL, 5 M; Merck) was subsequently added and the mixture was left for 1 h on ice. The suspension was clarified by centrifugation (10 000 rpm, 10 min), and the supernatant was transferred to two fresh microfuge tubes. One volume of cold isopropanol was used to precipitate the pellet followed by centrifugation at 5000 rpm for 15 min. Supernatant was discarded, and the resulting pellet was incubated in 1 mL ethanol 70% during 5 min, and washed with ethanol 70% twice. The pellet was then dried in speed vacuum and resuspended in 200 μL TE (pH 7.4). 0.5 μL of RNase (10 mg/ml) was added followed by 1-h incubation at 37°C. Mixture was centrifuged at 10 000 rpm for 15 min, and the supernatant was preserved at 4°C till further use. FFUL887 genome sequencing was based on Ion Torrent and was performed by Stab Vida (Portugal) as a paid service. Two rounds of paired-end sequencing were performed which resulted in ∼6 million reads with an average size of 199 bp. The reads were trimmed based on quality and for SNP calling the trimmed reads were mapped against the reference genome of CBS138 (available on Candida Genome Database) using CLC Genomics Workbench. To increase confidence on the results obtained variant detection was performed from the mapped reads using both the probabilistic and quality-based variant detection tools embedded on CLC. To annotate the genome of the FFUL887 strain, the reads were de novo assembled into 799 contigs yielding a total of assembled bases of 12.29 Mb and a genome coverage of around ×96 Ab initio gene detection was performed using different algorithms followed by manual curation to select the more appropriate gene models. The sequence and annotation of the genome of the FFUL887 clinical strain have been deposited in ENA (http://www.ebi.ac.uk/ena/data/view/FWDN01000001-FWDN01000799). CgPDR1 gene disruption The deletion of the C. glabrata PDR1 (CAGL0A00451g) in strain KUE100 (parental to the ΔCgPDR1 mutant used in spot assays) was carried out using the method described by Ueno et al. (2011). The target gene CgPDR1 was replaced by a DNA cassette including the CgHIS3 gene through homologous recombination. The replacement cassette was prepared by PCR (Table 2). Recombination locus and gene deletion were verified by PCR. Deletion of the same gene in the background of the clinical isolate strain FFUL887 was carried out using a little modified method from the one above described for the KUE100 strain. In specific, CgPDR1 was replaced by a DNA cassette containing the zeocin-resistant marker by homologous recombination. Transformants were isolated on YPD supplemented with 100 μg/ml zeocin. Correct insertion of the zeocine resistance cassette and corresponding deletion of CgPDR1 were verified by PCR and by confirming the absence of CgPDR1 expression in the mutant strains by real-time RT-PCR (primers available in Table 2). Table 2. List of primers used in this study. Primer identification  Primer sequence  CgRDN5.8  Forward 5΄-AACAACTCACCGGCCGAAT-3΄    Reverse 5΄-CTTGGTTCTCGCATCGATGA-3΄  CgPDR1  Forward 5΄-CGATTGCCAACCCGTTAGA-3΄    Reverse 5΄-GACGACCTTGGTGTAGGAGTCAT-3΄  CgCDR1  Forward 5΄-GCTTGCCCGCACATTGA-3΄    5΄-CCTCAGGCAGAGTGTGTTCTTTC-3΄  CgPDH1  Forward 5΄-GCCATGGTACCTGCATCGAT-3΄    5΄-CCGAGGAATAGCAAAACCAGTATAC-3΄  CgQDR2  Forward 5΄-TCACTGCATAGTTTCATATCGGACTA-3΄    Reverse 5΄-TGCCGATATGTTCCCAAGTGA-3΄  Primer identification  Primer sequence  CgRDN5.8  Forward 5΄-AACAACTCACCGGCCGAAT-3΄    Reverse 5΄-CTTGGTTCTCGCATCGATGA-3΄  CgPDR1  Forward 5΄-CGATTGCCAACCCGTTAGA-3΄    Reverse 5΄-GACGACCTTGGTGTAGGAGTCAT-3΄  CgCDR1  Forward 5΄-GCTTGCCCGCACATTGA-3΄    5΄-CCTCAGGCAGAGTGTGTTCTTTC-3΄  CgPDH1  Forward 5΄-GCCATGGTACCTGCATCGAT-3΄    5΄-CCGAGGAATAGCAAAACCAGTATAC-3΄  CgQDR2  Forward 5΄-TCACTGCATAGTTTCATATCGGACTA-3΄    Reverse 5΄-TGCCGATATGTTCCCAAGTGA-3΄  View Large Table 2. List of primers used in this study. Primer identification  Primer sequence  CgRDN5.8  Forward 5΄-AACAACTCACCGGCCGAAT-3΄    Reverse 5΄-CTTGGTTCTCGCATCGATGA-3΄  CgPDR1  Forward 5΄-CGATTGCCAACCCGTTAGA-3΄    Reverse 5΄-GACGACCTTGGTGTAGGAGTCAT-3΄  CgCDR1  Forward 5΄-GCTTGCCCGCACATTGA-3΄    5΄-CCTCAGGCAGAGTGTGTTCTTTC-3΄  CgPDH1  Forward 5΄-GCCATGGTACCTGCATCGAT-3΄    5΄-CCGAGGAATAGCAAAACCAGTATAC-3΄  CgQDR2  Forward 5΄-TCACTGCATAGTTTCATATCGGACTA-3΄    Reverse 5΄-TGCCGATATGTTCCCAAGTGA-3΄  Primer identification  Primer sequence  CgRDN5.8  Forward 5΄-AACAACTCACCGGCCGAAT-3΄    Reverse 5΄-CTTGGTTCTCGCATCGATGA-3΄  CgPDR1  Forward 5΄-CGATTGCCAACCCGTTAGA-3΄    Reverse 5΄-GACGACCTTGGTGTAGGAGTCAT-3΄  CgCDR1  Forward 5΄-GCTTGCCCGCACATTGA-3΄    5΄-CCTCAGGCAGAGTGTGTTCTTTC-3΄  CgPDH1  Forward 5΄-GCCATGGTACCTGCATCGAT-3΄    5΄-CCGAGGAATAGCAAAACCAGTATAC-3΄  CgQDR2  Forward 5΄-TCACTGCATAGTTTCATATCGGACTA-3΄    Reverse 5΄-TGCCGATATGTTCCCAAGTGA-3΄  View Large Transcriptomic analysis The transcriptomes of FFUL887 and CBS138 strains were compared using DNA microarrays specifically designed for C. glabrata (design ID 064590) (Rossignol et al.2007). Both strains were cultivated overnight in 25 mL of YPD at 30°C with orbital agitation (250 rpm) and then re-inoculated (at an OD600nm of 0.2 ± 0.05) in 150 mL of fresh RPMI. The cells were harvested by centrifugation (8000 × g, 7 min, 4°C—Beckman J2.21 Centrifuge, rotor JA.10) in mid-exponential phase (OD600nm∼2 ± 0.05) and immediately frozen at –80°C until further use. RNA extraction was performed as described in Bernardo et al. (2017) and in Rossignol et al. (2007). Comparison of gene transcript levels based on real time RT-PCR The expression of CgCDR1, CgPDH1, CgQDR2 and CgPDR1 genes was compared in CBS138, in FFUL887 and in FFUL_ΔCgpdr1 strains by real-time RT-PCR. Cells of the different strains were cultivated in identical conditions to those used for the microarray analysis. Conversion of total RNA into cDNA was performed using 1 μg of RNA. The reverse-transcription step was performed in a C1000 Thermal Cycler (Bio-Rad, Hercules, USA). The subsequent quantitative PCR step was performed using 2.5 μL of the cDNA. The sequence of the primers used is available on request. Gene expression was calculated using CgRDN5.8 as an internal control. Comparison of susceptibility to environmental stressors based on spot assays Comparison of susceptibility of CBS138, KUE100, KUE100_ΔCgPDR1, FFUL887 and FFUL887_ΔCgPDR1 cells to inhibitory concentrations of H2O2 and of the organic acids acetic, propionic and butyric acids was based on spot assays. Cells of the different strains were cultivated in MM growth medium until mid-exponential phase (OD600nm∼0.8) and then diluted in 1 mL of sterile water to obtain a cell suspension having an OD600nm of 0.05. Four microlitres of this cell suspension and of two subsequent dilutions (1:5 and 1:10) were applied onto the surface of MM agarised plates supplemented or not with inhibitory concentrations of H2O2 (5–18 mM), of acetic acid (50–60 mM), of propionic acid (17–20 mM) and of butyric acids (15–17 mM). The plates were incubated at 30°C for 2 to 3 days depending on the severity of growth inhibition. The same experimental setup was used to assess susceptibility of the strains to heat stress with the difference that instead of incubating the inoculated plates at 30°C, these were incubated at 37°C, 40°C or 42°C. RESULTS Identification of FFUL887 as a fluconazole- and voriconazole-resistant strain To characterise the incidence of resistance to fluconazole, voriconazole, caspofungin and anidulafungin in a cohort of 58 Candida glabrata clinical isolates, the concentration of each of these antifungals leading to a 50% growth inhibition (generally designated as MIC50), comparing with growth registered in drug-free medium, was determined. This phenotypic screening was performed using the highly standardised microdilution method recommended by EUCAST. As a control we have also included the reference strain CBS138 in the screening. The MIC value obtained for each drug and each isolate is shown in Fig. S1 (Supporting Information), and the distribution of MIC values across all isolates is shown in Fig. 1. The MIC values obtained were compared with the clinical resistance breakpoints defined by EUCAST (32 mg/L for fluconazole and 0.06 mg/L for anidulafungin) to classify the strains as resistant, intermediate or susceptible. To identify voriconazole-resistant strains, the MIC value obtained was compared with the ECV value (1 mg/mL) which can be used to distinguish wild-type from non-wild-type isolates (Pfaller et al.2011). For caspofungin, no breakpoint has been defined by EUCAST and therefore the strains were not classified. Under the experimental conditions used, the MIC value of the reference strain CBS138 was 16 mg/L for fluconazole and 0.25 mg/L for voriconazole, indicating that the reference strain is susceptible to these two azoles. Seven isolates (FFUL412, FFUL443, FFUL674, FFUL830, FFUL866, FFUL878, FFUL887) were resistant to fluconazole and voriconazole, two only resistant to fluconazole (FFUL98 and FFUL4012) and one only resistant to voriconazole (FFUL677) (Fig. S1). Notably, three of the cross-resistant isolates (FFUL412, FFUL443 and FFUL674) were retrieved from patients undergoing fluconazole-based therapy (Table S1). Despite the small number of isolates examined in this study, the percentage of resistance obtained for fluconazole and voriconazole (16% for fluconazole and 14% for voriconazole) are close to the values reported (10%–15%) in antifungal surveillance tests undertaken with much larger cohorts of strains (e.g. Pfaller et al.2010). None of the C. glabrata isolates tested could be considered susceptible to fluconazole as the MIC values were always above 0.002 mg/L (Fig. S1), consistent with the described increased resilience of C. glabrata to this azole drug (Fidel, Vazquez and Sobel 1999; Kolaczkowska and Kolaczkowski 2016). All the isolates tested exhibited high susceptibility to anidulafungin, none of them exhibiting growth when cultivated in the presence of 0.06 mg/L, the defined resistance breakpoint (Fig. 1B; Fig. S1). For caspofungin, only one isolate, FFUL887, exhibited a MIC value of 0.25, while for the remaining isolates this MIC value was of 0.125 or below (Fig. 1B; Fig. S1). Figure 1. View largeDownload slide Distribution of MIC50 values of fluconazole, voriconazole (A) caspofungin and anidulafungin (B) obtained for the cohort of C. glabrata clinical isolates tested in this work. The dashed line indicates the resistance breakpoints (in the case of voriconazole of the ECV value) defined by EUCAST (as detailed in materials and methods section). These results were obtained based on the assessment of MIC50 value by the microdilution method recommended by EUCAST and that gave rise to the results shown in Fig. S1. Figure 1. View largeDownload slide Distribution of MIC50 values of fluconazole, voriconazole (A) caspofungin and anidulafungin (B) obtained for the cohort of C. glabrata clinical isolates tested in this work. The dashed line indicates the resistance breakpoints (in the case of voriconazole of the ECV value) defined by EUCAST (as detailed in materials and methods section). These results were obtained based on the assessment of MIC50 value by the microdilution method recommended by EUCAST and that gave rise to the results shown in Fig. S1. The FFUL887 strain was selected for further analysis since it was demonstrated to be resistant to fluconazole and voriconazole and also exhibited higher resilience to caspofungin compared to the other clinical isolates in this study. To assess how the presence of the fluconazole and voriconazole affected growth kinetics of the FFUL887 and CBS138 strains, growth curves in liquid medium were performed using the same experimental setup that was used for estimation of MIC50 (Fig. S2, Supporting Information). Three concentrations of voriconazole and fluconazole were tested: one corresponding to the resistance breakpoint, one below and one above that value (Fig. S2). The results obtained show that the two strains exhibited a similar fitness when cultivated in drug-free medium, with only a slight decrease in the final biomass produced by the FFUL887 strain (Fig. S2). Supplementation of the RPMI medium with the two azole drugs led to a drastic growth inhibition of the CBS138 strain, whereas growth of the FFUL887 strain was almost identical to the control conditions, with only a small detectable decrease of the growth rate (e.g. 0.033 h−1 in the presence of 64 mg/L fluconazole and 0.044 h−1 in control conditions) (Fig. S2) and a slight increase in the lag phase that was observed upon inoculation in the drug-supplemented medium (Fig. S2). To assess if the resistant phenotype exhibited by the FFUL887 strain towards fluconazole and voriconazole was generalised for azoles or was limited to azoles of the triazole family (including fluconazole or voriconazole), growth of this strain in the presence of the imidazoles ketoconazole and clotrimazole was examined (Fig. S3, Supporting Information). Cells were cultivated for 24 h in 96-multiwell plates containing RPMI or in the same growth medium supplemented with 4 mg/L ketoconazole and 1 mg/mL of clotrimazole, the defined resistance breakpoints for these two drugs. Figure S3 shows that FFUL887 cells are more tolerant to the tested concentrations of ketoconazole and clotrimazole than the CBS138 strain; however, the FFUL887 strain is still considered susceptible to the two imidazoles since the concentrations tested reduced growth by more than 50% the one registered in control conditions (Fig. S3). FFUL887 genome sequencing and annotation The fluconazole and voriconazole resistance phenotype exhibited by the FFUL887 strain prompted us to obtain the genome sequence of this isolate. The assembled contig sequences were used for automatic ab initio gene detection that was subsequently manually curated to select the more appropriate gene models. This analysis allowed us to predict that the ORFeome of the FFUL887 strain includes 5079 genes, which corresponds to 96% of the total number of genes annotated for the CBS138 strain (Dujon et al.2004). The number of ORFs obtained for the FFUL887 is also in line with those recently reported for other C. glabrata clinical isolates (in the range of 5300 CDSs) (Havelsrud and Gaustad 2017; Vale-Silva et al.2017). The vast majority (5039) of the protein pairs present in CBS138 and FFUL887 shared more than 90% identity at the amino acid level indicating that the proteins encoded by the two strains are fairly similar. For the identification of SNPs that could underlie the observed resistance to fluconazole and voriconazole of the FFUL887 strain, the reads obtained were mapped against the genome sequence of the CBS138 strain, as detailed in materials and methods section. This comparison yielded 77 749 SNPs between the genomes of FFUL887 and CBS138 (Fig. S4, Supporting Information). A similar high number of SNPs was also reported in a recent comparative genomic analysis between two C. glabrata clinical isolates and the reference strain CBS138 (Vale-Silva et al.2017). About 45% of the SNPs identified between CBS138 and FFUL887 were located in coding regions, affecting 3194 of the gene sequences predicted for the FFUL887 strain (Fig. S4). The percentage of genes harbouring non-synonymous SNPs in the FFUL887 strain was similar throughout the nine C. glabrata nuclear chromosomes (∼60%) but considerably smaller (∼18%) in the mitochondrial chromosome (Fig. S4). On average, FFUL887 and CBS138 orthologous genes harboured five non-synonymous SNPs; however, in some cases this number increased up to more than 30 non-synonymous SNPs including in CAGL0K12078g (>50 non-synonymous SNPs), encoding a putative transcription factor similar to ScNrg1 transcription factor; CAGL0C00231g (42 non-synonymous SNPs), encoding a presumed plasma membrane nucleobase transporter, and the adhesin CgPWP4 (52 non-synonymous SNPs) (Fig. 2). Other adhesin-encoding genes were also observed to harbour a high number of non-synonymous SNPs between FFUL887 and CBS138 including CgEPA8, CgPWP5, CAGL0C03575g and CAGL0L10092g (Fig. 2). Figure 2. View largeDownload slide Number of non-synonymous SNPs present in FFUL887 predicted proteins, when compared with their CBS138 counterpartners. The reads obtained after whole-genome sequencing of the FFUL887 strain were mapped against the genome of the reference strain CBS138, as detailed in materials and methods section. Those genes exhibiting a higher number of non-synonymous SNPs in the FFUL887 strain are highlighted in the figure. Adhesin-encoding genes are evidenced in light blue. Figure 2. View largeDownload slide Number of non-synonymous SNPs present in FFUL887 predicted proteins, when compared with their CBS138 counterpartners. The reads obtained after whole-genome sequencing of the FFUL887 strain were mapped against the genome of the reference strain CBS138, as detailed in materials and methods section. Those genes exhibiting a higher number of non-synonymous SNPs in the FFUL887 strain are highlighted in the figure. Adhesin-encoding genes are evidenced in light blue. Notably, no SNPs were found in the sequence of ERG11 gene encoded by FFUL887, consistent with the idea that azole resistance of these strains is not driven by alterations in the drug target as shown for most C. glabrata azole-resistant isolates (Sanguinetti et al.2005; Tsai et al.2010; Kolaczkowska and Kolaczkowski 2016). No SNPs were also found in FFUL887 Msh2, a DNA repair protein whose mutations had been linked to the development of azole resistance in C. glabrata-resistant clinical strains (Healey et al.2016). Concerning the CgFKS1 and CgFKS2, the two enzymes targeted by echinocandins, the FFUL887 alleles harbour one non-synonymous SNP each (Gly14Ser in CgFks1 and Thr926Pro in CgFks2), compared to CBS138 orthologues; however, these polymorphisms are outside of the hotspot regions commonly found to be altered in echinocandin-resistant isolates (Garcia-Effron et al.2009). It thus remains to be examined whether the higher tolerance of the FFUL887 strain to caspofungin comes from these polymorphisms or if it results from other genetic traits. Comparative genomic analysis between FFUL887 and CBS138 focused on azole-resistance genes: emphasis on CgPdr1 To gather insights into the differential levels of resistance to fluconazole and voriconazole of the CBS138 and FFUL887 strains, the set of proteins found to harbour non-synonymous SNPs in the resistant strain were compared with a comprehensive list of 78 genes previously implicated in C. glabrata resistance to fluconazole and voriconazole (listed in Table S2, Supporting Information). Fifty-three genes associated with resistance to these two azoles differed in FFUL887 and in CBS138, a subset of these being shown in Table 3. Table 3. Subset of proteins previously described to be involved in fluconazole and/or voriconazole resistance in C. glabrata and that were found to harbour non-synonymous SNPs in the FFUL887 strain, when compared with their CBS138 counterpartners. Gene or ORF name  Function  Amino acid modification found in FFUL887  CgPDR1  Zinc finger transcription factor, activator of drug resistance genes  Val91Ile; Leu98Ser; Asp243Asn; Lys274Gln  CgGAL11ACgGAL11B  Component of the transcriptional Mediator complex that provides interfaces between RNA polymerase II and upstream activator proteins; essential for CgPdr1-dependent activation of azole-resistance genes  Ser134Asn; Ser965Gly; Ser1084AsnGln203Pro; His322Asn; Pro372Ser; Ile615Asn; *839Gly  CgPDH1  Multidrug transporter of the ATP-binding cassette superfamily  Lys438Gln; Glu839Asp  CAGL0L04400g  Zinc finger transcription factor involved in transcriptional regulation of MDR genes. Orthologue of S. cerevisiae ScYRR1  Cys24Gly; Ala58Val; Ile137Leu; Asp229Glu; Ile346Val; Glu574Lys; Ile593Leu; Glu710Asp; Ala933Val  CgUPC2A  Zinc finger transcription factor required for transcriptional regulation of genes involved in uptake and biosynthesis of ergosterol  Arg92Lys; Asn304Ser; Glu822Val  CgSIN3  Component of the Rpd3S and Rpd3L histone deacetylase complexes  Asn50Lys; Lys288Thr  CAGL0L03377g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Gly134Asp; Gly172Ser; Lys252Arg; Ile347Met; Ala695Gly; Lys813Arg; Asp819Glu; Thr867Met  CAGL0L09383g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Ser116Asn; Ile185Val; Met202Val  CgHST1  Histone deacetylase that regulates gene expression in niacin-limiting conditions  Met70Lys  Gene or ORF name  Function  Amino acid modification found in FFUL887  CgPDR1  Zinc finger transcription factor, activator of drug resistance genes  Val91Ile; Leu98Ser; Asp243Asn; Lys274Gln  CgGAL11ACgGAL11B  Component of the transcriptional Mediator complex that provides interfaces between RNA polymerase II and upstream activator proteins; essential for CgPdr1-dependent activation of azole-resistance genes  Ser134Asn; Ser965Gly; Ser1084AsnGln203Pro; His322Asn; Pro372Ser; Ile615Asn; *839Gly  CgPDH1  Multidrug transporter of the ATP-binding cassette superfamily  Lys438Gln; Glu839Asp  CAGL0L04400g  Zinc finger transcription factor involved in transcriptional regulation of MDR genes. Orthologue of S. cerevisiae ScYRR1  Cys24Gly; Ala58Val; Ile137Leu; Asp229Glu; Ile346Val; Glu574Lys; Ile593Leu; Glu710Asp; Ala933Val  CgUPC2A  Zinc finger transcription factor required for transcriptional regulation of genes involved in uptake and biosynthesis of ergosterol  Arg92Lys; Asn304Ser; Glu822Val  CgSIN3  Component of the Rpd3S and Rpd3L histone deacetylase complexes  Asn50Lys; Lys288Thr  CAGL0L03377g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Gly134Asp; Gly172Ser; Lys252Arg; Ile347Met; Ala695Gly; Lys813Arg; Asp819Glu; Thr867Met  CAGL0L09383g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Ser116Asn; Ile185Val; Met202Val  CgHST1  Histone deacetylase that regulates gene expression in niacin-limiting conditions  Met70Lys  View Large Table 3. Subset of proteins previously described to be involved in fluconazole and/or voriconazole resistance in C. glabrata and that were found to harbour non-synonymous SNPs in the FFUL887 strain, when compared with their CBS138 counterpartners. Gene or ORF name  Function  Amino acid modification found in FFUL887  CgPDR1  Zinc finger transcription factor, activator of drug resistance genes  Val91Ile; Leu98Ser; Asp243Asn; Lys274Gln  CgGAL11ACgGAL11B  Component of the transcriptional Mediator complex that provides interfaces between RNA polymerase II and upstream activator proteins; essential for CgPdr1-dependent activation of azole-resistance genes  Ser134Asn; Ser965Gly; Ser1084AsnGln203Pro; His322Asn; Pro372Ser; Ile615Asn; *839Gly  CgPDH1  Multidrug transporter of the ATP-binding cassette superfamily  Lys438Gln; Glu839Asp  CAGL0L04400g  Zinc finger transcription factor involved in transcriptional regulation of MDR genes. Orthologue of S. cerevisiae ScYRR1  Cys24Gly; Ala58Val; Ile137Leu; Asp229Glu; Ile346Val; Glu574Lys; Ile593Leu; Glu710Asp; Ala933Val  CgUPC2A  Zinc finger transcription factor required for transcriptional regulation of genes involved in uptake and biosynthesis of ergosterol  Arg92Lys; Asn304Ser; Glu822Val  CgSIN3  Component of the Rpd3S and Rpd3L histone deacetylase complexes  Asn50Lys; Lys288Thr  CAGL0L03377g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Gly134Asp; Gly172Ser; Lys252Arg; Ile347Met; Ala695Gly; Lys813Arg; Asp819Glu; Thr867Met  CAGL0L09383g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Ser116Asn; Ile185Val; Met202Val  CgHST1  Histone deacetylase that regulates gene expression in niacin-limiting conditions  Met70Lys  Gene or ORF name  Function  Amino acid modification found in FFUL887  CgPDR1  Zinc finger transcription factor, activator of drug resistance genes  Val91Ile; Leu98Ser; Asp243Asn; Lys274Gln  CgGAL11ACgGAL11B  Component of the transcriptional Mediator complex that provides interfaces between RNA polymerase II and upstream activator proteins; essential for CgPdr1-dependent activation of azole-resistance genes  Ser134Asn; Ser965Gly; Ser1084AsnGln203Pro; His322Asn; Pro372Ser; Ile615Asn; *839Gly  CgPDH1  Multidrug transporter of the ATP-binding cassette superfamily  Lys438Gln; Glu839Asp  CAGL0L04400g  Zinc finger transcription factor involved in transcriptional regulation of MDR genes. Orthologue of S. cerevisiae ScYRR1  Cys24Gly; Ala58Val; Ile137Leu; Asp229Glu; Ile346Val; Glu574Lys; Ile593Leu; Glu710Asp; Ala933Val  CgUPC2A  Zinc finger transcription factor required for transcriptional regulation of genes involved in uptake and biosynthesis of ergosterol  Arg92Lys; Asn304Ser; Glu822Val  CgSIN3  Component of the Rpd3S and Rpd3L histone deacetylase complexes  Asn50Lys; Lys288Thr  CAGL0L03377g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Gly134Asp; Gly172Ser; Lys252Arg; Ile347Met; Ala695Gly; Lys813Arg; Asp819Glu; Thr867Met  CAGL0L09383g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Ser116Asn; Ile185Val; Met202Val  CgHST1  Histone deacetylase that regulates gene expression in niacin-limiting conditions  Met70Lys  View Large One of the proteins that differed between FFUL887 and CBS138 is CgPdr1, which exhibited four non-synonymous SNPs (Val91Ile, Leu98Ser, Asp243Asn and Lys274Gln), when compared with the corresponding CBS138 orthologue. Three of these SNPs (Val91Ile; Leu98Ser and Asp243Asn) were found to be present simultaneously in the CgPDR1 alleles encoded by isolates resistant and susceptible to azoles (Ferrari et al.2009), while the Lys274Gln was not previously described. Based on this observation, it was hypothesised that the substitution K274Q could represent a GOF substitution of C. glabrata CgPdr1. To test the hypothesis that K274Q mutation leads to hyperactivation of CgPdr1, the transcriptomes of the FFUL887 and CBS138 strains were compared in drug-free RPMI medium using species-specific DNA microarrays. Out of the 409 genes found to be overexpressed in the FFUL887 strain (above 1.5-fold, P-value below 0.001; listed in Table S3, Supporting Information), 89 genes are documented targets of CgPdr1, according to the information available in the PathoYeastract database (Monteiro et al.2017) (highlighted in grey in the Table S3). Among these were the well-characterised CgPdr1 targets CgPDH1 and CgCDR1, as well as CgPDR1 itself (Table S3). Deletion of CgPDR1K274Q abrogated the increase in transcription of CgPDH1, CgCDR1, CgPDR1 and CgQDR2 genes registered in the FFUL887 strain (Fig. 3A). As expected, deletion of the CgPDR1K274Q allele also resulted in sensitisation of the FFUL887 strain to fluconazole and voriconazole (Fig. 3B). Overall, the results of the transcriptomic profiling of the CBS138 and FFUL887 strains strongly support the idea that FFUL887 encodes a CgPdr1 GOF allele, resulting from the K274Q SNP. Figure 3. View largeDownload slide (A) MIC for fluconazole and voriconazole obtained for the CBS138, FFUL887 and the FFUL887_ΔCgPDR1 strains, as determined by the microdilution method recommended by EUCAST. For the statistical analysis, the results obtained for the mutant strain devoid of CgPDR1 gene were compared with those gathered for the wild-type FFUL887 strain. ****P-value below 0.0001; (B) comparison of the transcript levels of CgPDR1, CgCDR1, CgPDH1 and CgQDR2 genes in CBS138, FFUL887 and the FFUL887_ΔCgPDR1 strains. Cells of the different strains were cultivated in RPMI growth medium until mid-exponential phase after which the expression of CgPDR1, CgCDR1, CgPDH1 and CgQDR2 genes was compared by qRT-PCR. The values represented for the FFUL887 and FFUL887_ΔCgPDR1 strains are relative to the value obtained for the CBS138 strain, which was considered to be equal to 1. For the statistical analysis, the results obtained for the FFUL887 strain were compared with those gathered for CBS138, while the results obtained for the FFUL887_ΔCgPDR1 mutant were compared with those obtained for FFUL887. *P-value below 0.05, ****P-value below 0.0001. Figure 3. View largeDownload slide (A) MIC for fluconazole and voriconazole obtained for the CBS138, FFUL887 and the FFUL887_ΔCgPDR1 strains, as determined by the microdilution method recommended by EUCAST. For the statistical analysis, the results obtained for the mutant strain devoid of CgPDR1 gene were compared with those gathered for the wild-type FFUL887 strain. ****P-value below 0.0001; (B) comparison of the transcript levels of CgPDR1, CgCDR1, CgPDH1 and CgQDR2 genes in CBS138, FFUL887 and the FFUL887_ΔCgPDR1 strains. Cells of the different strains were cultivated in RPMI growth medium until mid-exponential phase after which the expression of CgPDR1, CgCDR1, CgPDH1 and CgQDR2 genes was compared by qRT-PCR. The values represented for the FFUL887 and FFUL887_ΔCgPDR1 strains are relative to the value obtained for the CBS138 strain, which was considered to be equal to 1. For the statistical analysis, the results obtained for the FFUL887 strain were compared with those gathered for CBS138, while the results obtained for the FFUL887_ΔCgPDR1 mutant were compared with those obtained for FFUL887. *P-value below 0.05, ****P-value below 0.0001. The deletion of CgPDR1K274Q allele affects tolerance to environmental stress of the FFUL887 isolate Previous studies have shown that deletion of GOF CgPdr1 alleles results in altered stress resilience of azole-resistant C. glabrata strains (Vermitsky et al.2006). We therefore compared growth of CBS138, FFUL887 and FFUL887_ΔCgPDR1K274Q in the presence of various environmental stressors including H2O2, acetic acid, propionic acid, butyric acid and at different temperatures (30°C, 37°C, 40°C or 42°C). Under the experimental conditions used, the deletion of CgPDR1 in the FFUL887 background led to a mild decrease in growth of the strains when cultivated at all the temperatures tested (Fig. 4). No significant differences were observed upon CgPDR1 deletion in the FFUL887 background concerning tolerance to H2O2 but, surprisingly, during cultivation in the presence of inhibitory concentrations of the organic acids acetic, propionic and butyric acids (at pH 4.5) the deletion of CgPDR1K274Q was beneficial (Fig. 4). These experiments were also performed in the genetic background of the laboratory strain KUE100 that encodes a wild-type CgPdr1 allele. The results obtained confirmed a slight protective effect exerted by CgPDR1 expression against high temperatures, while in the presence of acetic, propionic or butyric acids there was no significant differences in growth of the wild-type or of the ΔCgPDR1 mutant (Fig. 4). Figure 4. View largeDownload slide Comparison of the susceptibility of CBS138, FFUL887 and the FFUL887_ΔCgPDR1 strains to environmental stressors based on spot assays. Mid-exponential phase cells of the different strains were cultivated in solid MM growth medium or in this same medium supplemented with inhibitory concentrations of H2O2 or of the fatty acids acetic acid, propionic acid and butyric acid. Lanes b and c correspond to 1:5 and 1:10 dilutions of the cell suspension used in lane a. Growth was compared after 2 to 3 days of incubation at 30°C, depending on the severity of growth inhibition. The same experimental setup was used to compare tolerance of the strains to different temperatures. Figure 4. View largeDownload slide Comparison of the susceptibility of CBS138, FFUL887 and the FFUL887_ΔCgPDR1 strains to environmental stressors based on spot assays. Mid-exponential phase cells of the different strains were cultivated in solid MM growth medium or in this same medium supplemented with inhibitory concentrations of H2O2 or of the fatty acids acetic acid, propionic acid and butyric acid. Lanes b and c correspond to 1:5 and 1:10 dilutions of the cell suspension used in lane a. Growth was compared after 2 to 3 days of incubation at 30°C, depending on the severity of growth inhibition. The same experimental setup was used to compare tolerance of the strains to different temperatures. Comparative transcriptomic analysis between FFUL887 and CBS138 shows dramatic alterations in the expression of genes involved in carbohydrate, nitrogen and sulphur metabolism We have further explored our results concerning the comparison of the transcriptome of the FFUL887 and CBS138 strains during growth in RPMI medium aiming to gain insights into the responses evolved by C. glabrata during colonisation of the human urinary tract considering that this was the niche where FFUL887 isolate was retrieved from. The CBS138 strain has an intestinal origin; however, its extensive utilisation in the laboratory as likely resulted in its domestication leading to large phenotypic differences compared to those observed in C. glabrata clinical isolates including those of intestinal origin (Gregori et al.2007; Cunha et al.2017). The genes overexpressed (above 1.5-fold and having a P-value below 0.001) in the FFUL887 isolate were clustered according to their biological function using the MIPS functional catalogue (Fig. S5, Supporting Information). Results revealed a significant enrichment (P-value below 0.001) of genes involved in ‘metabolism of amino acids’, ‘metabolism of carbohydrates’, ‘nitrogen, sulphur and selenium metabolism’, ‘lipid, fatty acid and isoprenoid metabolism’, ‘generation of energy’, ‘vacuolar protein degradation’, ‘transport’ and ‘oxidative stress response’ (Fig. S5 and Table S3). Similarly, the set of genes upregulated in CBS138 (and consequently downregulated in FFUL887) was enriched (P-value below 0.001) in genes related with ‘protein synthesis’ (Fig. S5 and Table S3). The genes upregulated in the FFUL887 strain related with carbohydrate and lipid metabolism included enzymes involved in fatty-acid β-oxidation, in catabolism of acetate, of propionate and of glycogen, as well as genes encoding neoglucogenic and Krebs cycle enzymes (Fig. S6, Supporting Information). This observation was surprising considering that at the time point where FFUL887 and CBS138 cells were harvested for the microarray analysis (after 6 h of cultivation in the rich RPMI medium supplemented with 20 g/L glucose) there was still a considerable amount of glucose present in the culture supernatant of the two cultures (∼18.5 g/L, based on HPLC analysis of the supernatants). The genes upregulated in the FFUL887 strain related with amino acid and sulphur metabolism classes were essentially those involved in metabolism of various amino acids and genes of the trans-sulfuration pathway that allows transport and incorporation of sulphate in methione and cystheine (Fig. S6). Consistently, several transporters involved in the uptake of amino acids, small peptides and inorganic sulphur were also found to be upregulated in the FFUL887 strain (Fig. S6). DISCUSSION In this work, we have disclosed the genome sequence of a Candida glabrata clinical isolate, FFUL887, resistant to voriconazole and fluconazole and also exhibiting enhanced tolerance to caspofungin. The higher resistance of the FFUL887 strain to these two types of antifungals is striking considering that they have different modes of action. Resistance to voriconazole and fluconazole in FFUL887 was largely dependent on the expression of CgPdr1; however, this is not likely to underlie the higher tolerance of this strain to caspofungin since CgPdr1 expression is dispensable for C. glabrata tolerance to echinocandinds (Schwarzmuller et al.2014). A very high number of SNPs were obtained when comparing the genomic sequences of CBS138 and FFUL887, probably reflecting the different genetic background of these two strains. Nevertheless, the vast majority of the proteins encoded by the two strains were still very similar with >90% of the proteins encoded by the two strains sharing a degree of homology above 90%. Recent comparative genomic analysis between C. glabrata clinical isolates and the CBS138 strain also revealed very prominent differences, within the range of those reported in our study (Havelsrud and Gaustad 2017; Vale-Silva et al.2017), while a strain used for carboxylic acids production was much more similar to CBS138 (Xu et al.2016). Interestingly, even the comparison of cohorts of related clinical isolates shows very prominent differences, similar to those that are observed when the isolates are compared with the CBS138 strain (Vale-Silva et al.2017). These observations reflect the described genomic plasticity of C. glabrata species which gives rise to a large genetic and phenotypic diversity among isolated strains (Carreté et al.2017). While in our study we have focused on the comparison between the transcriptome and genome of an azole-resistant strain with the susceptible CBS138 strain, others have performed similar analyses but using related isolates (for example, strains retrieved from patients obtained before and after application of azole-therapy (Vermitsky and Edlind 2004; Vermitsky et al.2006; Vale-Silva et al.2017)). In our case, a similar approach was not possible since we could not identify among the sensitive strains tested one that could be related with FFUL887. Necessarily, the option of comparing the genomes of CBS138 vs FFUL887 is difficult for the establishment of genotype–phenotype associations; however, it has the advantage of allowing the identification of new SNPs that could be relevant for azole resistance and that would not be detected when comparing two already adapted clinical strains such as modifications occurring early during the process of colonisation of the human host. Besides CgPdr1, several other well-characterised determinants of C. glabrata resistance to azoles were found to harbour SNPs in the resistant strain FFUL887 including the MDR transporters CgPhd1, CgTpo1_1 and CgTpo1_2 and the transcriptional regulators CgUpc2A, CgYrr1, CgStb5 and CgGal11A (Table S2). It is difficult to understand whether these polymorphisms contribute to the higher resistance exhibited by FFUL887 cells towards fluconazole and voriconazole because the biochemical activity of these proteins is not well studied, and therefore it is hard to predict the consequences for protein activity of the identified SNPs. Nevertheless, these indicatives deserve further exploration since azole-resistance genes are surely under selective pressure, as occurs with CgPdr1. The extensive upregulation of about 90 documented targets of CgPdr1 that was observed in the FFUL887 strain during cultivation in drug-free growth medium strongly supports the idea that the K274Q substitution is, indeed, a new GOF mutation of this protein. Interestingly, a K274N substitution has been previously reported, however, while this results in mild increase in fluconazole tolerance (MIC of 16 mg/L)(Caudle et al.2011), the herein reported K274Q substitution results in a much higher resistance (MIC of 64 mg/L). The K274 residue lies within a region of CgPdr1 where several other mutations have been described (as detailed in Fig. S7, Supporting Information) and is located near a predicted inhibitory regulatory domain of CgPdr1 (residues 322–465) (Fig. S7). In Saccharomyces cerevisiae, this regulatory domain inhibits the activity of ScPdr1 (Kolaczkowska et al.2002), for which it can be hypothesised that the K274Q modification could compromise the function of the inhibitory domain resulting in an hyperactivation of CgPdr1. Further studies are required to better understand how the K274Q and other GOF mutations modulate the activity of CgPdr1. The genes that are under regulation of different CgPdr1 GOF mutants have a modest overlap (Ferrari et al.2009, 2011; Tsai et al.2010; Caudle et al.2011). In order to determine the effect of the CgPdr1 K274Q substitution in C. glabrata genomic expression, the 92 genes upregulated in the FFUL887 strain harbouring a PDRE motif (TCCRYGSR) in their promoter (presumed to be the direct targets of CgPdr1) were compared with the set of genes regulated by three other GOF alleles: P927L and L946S and K274N (Fig. 5). Only five genes were in common in the three datasets: CgCDR1, CgYOR1, CgPDR1, CgPUP1 and CAGL0M09713g (Fig. 5). Consistently, CgCDR1 and CgPUP1 genes were recently shown to be upregulated among a cohort of CgPdr1 GOF mutants different from those used to build (Fig. 5) (Ferrari et al.2011). The pattern of expression of other drug-efflux pumps varied according to the GOF mutation: while K274Q, L94S and P927L led to the upregulation of CgQDR2 and CgPHD1, K274N was the only mutation causing upregulation of CgTPO1–1 (Fig. 5). The expression of adhesin-encoding genes was also found to vary according to the CgPdr1 GOF mutation (Fig 5). This observation is particularly interesting in light of the described effect of CgPdr1 in contributing for C. glabrata adhesion to epithelial cells (Vale-Silva et al.2013). Surprisingly, the overlap between the genes regulated by the CgPdr1 GOF mutants K274N and K274Q was very limited (Fig. 5) demonstrating that even polymorphisms in the same CgPdr1 residue have a very different impact on the control of gene expression. One of the mechanisms that has been hypothesised to explain this divergence in the set of genes regulated by different CgPdr1 mutants is that they might be differently activated thereby resulting in a different interaction with the transcriptional machinery (Paul, Schmidt and Moye-Rowley 2011). The different genetic background of the strains used in the different transcriptomic profilings may also contribute for some of the observed divergences. Figure 5. View largeDownload slide Venn diagram comparing the set of genes regulated by the CgPdr1 GOF mutants K274Q, K274N, P927L and L946S, as revealed by transcriptomic analyses. The set of genes herein identified as being upregulated in the FFUL887 isolate and harbouring in their promoter region a PDRE motif (TCCRYGSR) was compared with the set of genes previously described to be under the regulation of K274N, P927L and L946S CgPdr1 GOF mutations (Vermitsky et al.2006; Tsai et al.2010). Figure 5. View largeDownload slide Venn diagram comparing the set of genes regulated by the CgPdr1 GOF mutants K274Q, K274N, P927L and L946S, as revealed by transcriptomic analyses. The set of genes herein identified as being upregulated in the FFUL887 isolate and harbouring in their promoter region a PDRE motif (TCCRYGSR) was compared with the set of genes previously described to be under the regulation of K274N, P927L and L946S CgPdr1 GOF mutations (Vermitsky et al.2006; Tsai et al.2010). In addition to contributing to maximal resistance to voriconazole and fluconazole, we also showed that the CgPDR1K274Q allele is detrimental for growth of FFUL887 cells when cultivated in the presence of organic acids at a low pH. Similarly, cells expressing a CgPdr1 P927L GOF allele were also found to be susceptible to organic acids at a low pH (Vermitsky et al.2006). On the background of the KUE100 strain (derived from CBS138) which encodes a wild-type CgPdr1 allele, this phenotype towards organic acids was not observed indicating that it could be a feature of CgPdr1 GOF mutants, or at least of a subset of them. It is not possible with the data available until so far to clarify the reasons why the presence of organic acids seems to sensitise FFUL887 cells, although this is certainly a feature that deserves further exploration as it could be used to improve treatment of infections caused by isolates harbouring CgPdr1 GOF alleles. Besides contributing to better understand the acquisition of azole resistance, the comparative analyses of the genome and transcriptome of the CBS138 and FFUL887 strains also had the potential to elucidate some aspects underlying C. glabrata colonisation of the human urinary tract (the site where FFUL887 was retrieved from). In this sense, one of the observations that emerged from the comparative genomic analysis performed was the identification of several genes encoding adhesins as among those that had the higher number of non-synonymous SNPs in the FFUL887 strain. Adhesion is a fundamental step for C. glabrata ability to successfully colonise infection sites and as such adhesin-encoding genes are subjected to a tight selective pressure demonstrated to occur both at the transcriptional and genomic level (Halliwell et al.2012; Carreté et al.2017; Vale-Silva et al.2017). The microarray analysis performed revealed only two adhesin-encoding genes, CAGL0K10164g and CAGL0H08844g, as being upregulated in the FFUL887 strain; however, this analysis was performed using planktonic cells and thereby adhesion was not being favoured. The herein observed prominent differences in the primary sequence of the adhesins encoded by FFUL887 and CBS138, with emphasis for CgPwp4 which was the protein that differed the most in the two strains (Fig. 2), shows that these genes are also subjected to a strong selective pressure probably to select those variants contributing more to improve adherence to the available surfaces. Another observation that emerged from the comparative transcriptomic analysis performed was the significant upregulation in the FFUL887 strain of genes involved in metabolism of amino acids and of sulphur as well as a large number of genes involved in metabolisation of fatty acids, glycogen and other carbon sources. Since the RPMI growth medium where the strains were cultivated for the transcriptomic analysis contains glucose (20 g/L) and sulphate (50 mg/L), the higher expression of these genes in the FFUL887 strain is more likely to reflect a higher basal level of expression compared to the one observed in the CBS138 strain. In C. glabrata, the presence of glucose in the medium does not appear to repress metabolisation of other carbon sources nor of the genes involved in those processes, as observed in S. cerevisiae (Bernardo et al.2017; Cunha et al.2017). A similar alleviation of glucose repression over metabolism of alternative carbon sources was observed in C. albicans (Childers et al.2016). We have searched the genome sequence of the CBS138 and of the FFUL887 strains for genes homologous to those that mediate glucose repression in S. cerevisiae (Table S4, Supporting Information). With the exception of the ScMTH1 gene, all the other genes mediating glucose repression in S. cerevisiae have robust homologues in CBS138 and in FFUL887; however, in some cases, there were marked differences between the S. cerevisae and the C. glabrata proteins, those more prominent being the Mig1, Mig2 and Mig3 transcriptional regulators (Table S4). Further studies are required to understand the molecular players underlying the alleviation of glucose repression in C. glabrata and how their activity is modified by selective pressure during colonisation. The increased resilience of C. glabrata to antifungal therapy and the persistent emergence of resistant strains is highly problematic considering the high rates of morbidity and mortality associated with infections caused by this pathogenic yeast. The results presented in this study provide a further contribution for a better understanding of the key players contributing for the acquisition of resistance in the host, with special emphasis on CgPdr1 transcription factor, a knowledge that can be used to guide the development of more efficient therapeutical strategies. Specifically, it was shown for the first time that the K274Q substitution results in a GOF CgPdr1 mutant and, consequently, in enhanced azole resistance. An observation of remark from our study and others was that the expression of the K274Q and of P927L CgPdr1 allele increases susceptibility to organic acids at a low pH, suggesting that these molecules could be used to sensitise azole-resistant strains dependent on CgPdr1 GOF alleles. The herein reported whole-genome analysis of the FFUL887 strain and subsequent comparison with CBS138 strain reinforced the extreme genetic diversity among C. glabrata strains providing clues into the adaptive responses evolved during colonisation of the human host and advancing current knowledge on the biology and physiology of this yeast species. SUPPLEMENTARY DATA Supplementary data are available at FEMSYR online. FUNDING Funding received by iBB-Institute for Bioengineering and Biosciences from FCT-Portuguese Foundation for Science and Technology (UID/BIO/04565/2013) and from Programa Operacional Regional de Lisboa 2020 (Project N. 007317) is acknowledged. GB acknowledges Ireland science foundation (grant number 12IA1343) for financial support. Conflict of interest. None declared. REFERENCES Bernardo RT, Cunha DV, Wang C et al.   The CgHaa1-regulon mediates response and tolerance to acetic acid stress in the human pathogen Candida glabrata. G3  2017; 7: 1– 18. Google Scholar CrossRef Search ADS PubMed  Borst A, Raimer MT, Warnock DW et al.   Rapid acquisition of stable azole resistance by Candida glabrata isolates obtained before the clinical introduction of fluconazole. Antimicrob Agents Ch  2005; 49: 783– 7. Google Scholar CrossRef Search ADS   Carreté L, Ksiezopolska E, Pequeroles C et al.   Patterns of genomic variation in the opportunistic pathogen Candida glabrata suggest the existence of mating and a secondary association to the human host. 2017. Caudle KE, Barker KS, Wiederhold NP et al.   Genomewide expression profile analysis of the Candida glabrata Pdr1 regulon. Eukaryot Cell  2011; 10: 373– 83. Google Scholar CrossRef Search ADS PubMed  Childers DS, Raziunaite I, Mol Avelar G et al.   The rewiring of ubiquitination targets in a pathogenic yeast promotes metabolic flexibility, host colonization and virulence. PLoS Pathog  2016; 12: e1005566. Google Scholar CrossRef Search ADS PubMed  Costa C, Dias PJ, Sa-Correia I et al.   MFS multidrug transporters in pathogenic fungi: do they have real clinical impact? Front Physiol  2014; 5: 197. Google Scholar PubMed  Cunha DV, Salazar SB, Lopes MM et al.   Mechanistic insights underlying tolerance to acetic acid stress in vaginal Candida glabrata clinical isolates. Front Microbiol  2017; 8: 1– 13. Google Scholar CrossRef Search ADS PubMed  Dujon B, Sherman D, Fischer G et al.   Genome evolution in yeasts. Nature  2004; 430: 35– 44. Google Scholar CrossRef Search ADS PubMed  Ferrari S, Ischer F, Calabrese D et al.   Gain of function mutations in CgPDR1 of Candida glabrata not only mediate antifungal resistance but also enhance virulence. PLoS Pathog  2009; 5: e1000268. Google Scholar CrossRef Search ADS PubMed  Ferrari S, Sanguinetti M, Torelli R et al.   Contribution of CgPDR1-regulated genes in enhanced virulence of azole-resistant Candida glabrata. PLoS One  2011; 6: e17589. Google Scholar CrossRef Search ADS PubMed  Fidel PL Jr, Vazquez JA, Sobel JD. Candida glabrata: review of epidemiology, pathogenesis, and clinical disease with comparison to C. albicans. Clin Microbiol Rev  1999; 12: 80– 96. Google Scholar PubMed  Garcia-Effron G, Lee S, Park S et al.   Effect of Candida glabrata FKS1 and FKS2 mutations on echinocandin sensitivity and kinetics of 1,3-beta-D-glucan synthase: implication for the existing susceptibility breakpoint. Antimicrob Agents Ch  2009; 53: 3690– 9. Google Scholar CrossRef Search ADS   Gregori C, Schuller C, Roetzer A et al.   The high-osmolarity glycerol response pathway in the human fungal pathogen Candida glabrata strain ATCC 2001 lacks a signaling branch that operates in baker's yeast. Eukaryot Cell  2007; 6: 1635– 45. Google Scholar CrossRef Search ADS PubMed  Halliwell SC, Smith MC, Muston P et al.   Heterogeneous expression of the virulence-related adhesin Epa1 between individual cells and strains of the pathogen Candida glabrata. Eukaryot Cell  2012; 11: 141– 50. Google Scholar CrossRef Search ADS PubMed  Havelsrud OE, Gaustad P. Draft genome sequences of Candida glabrata isolates 1A, 1B, 2A, 2B, 3A, and 3B. Genome Announc  2017; 5: e00328-16. Google Scholar CrossRef Search ADS PubMed  Healey KR, Zhao Y, Perez WB et al.   Prevalent mutator genotype identified in fungal pathogen Candida glabrata promotes multi-drug resistance. Nat Commun  2016; 7: 11128. Google Scholar CrossRef Search ADS PubMed  Khatib R, Johnson LB, Fakih MG et al.   Current trends in candidemia and species distribution among adults: Candida glabrata surpasses C. albicans in diabetic patients and abdominal sources. Mycoses  2016; 59: 781– 6. Google Scholar CrossRef Search ADS   Kolaczkowska A, Kolaczkowski M. Drug resistance mechanisms and their regulation in non-albicans Candida species. J Antimicrob Chemoth  2016; 71: 1438– 50. Google Scholar CrossRef Search ADS   Kolaczkowska A, Kolaczkowski M, Delahodde A et al.   Functional dissection of Pdr1p, a regulator of multidrug resistance in Saccharomyces cerevisiae. Mol Genet Genomics  2002; 267: 96– 106. Google Scholar CrossRef Search ADS PubMed  Monteiro PT, Pais P, Costa C et al.   The PathoYeastract database: an information system for the analysis of gene and genomic transcription regulation in pathogenic yeasts. Nucleic Acids Res  2017; 45: D597– 603. Google Scholar CrossRef Search ADS PubMed  Paul S, Schmidt JA, Moye-Rowley WS. Regulation of the CgPdr1 transcription factor from the pathogen Candida glabrata. Eukaryot Cell  2011; 10: 187– 97. Google Scholar CrossRef Search ADS PubMed  Pfaller MA, Andes D, Arendrup MC et al.   Clinical breakpoints for voriconazole and Candida spp. revisited: review of microbiologic, molecular, pharmacodynamic, and clinical data as they pertain to the development of species-specific interpretive criteria. Diagn Micr Infec Dis  2011; 70: 330– 43. Google Scholar CrossRef Search ADS   Pfaller MA, Diekema DJ, Gibbs DL et al.   Results from the ARTEMIS DISK Global Antifungal Surveillance Study, 1997 to 2007:a 10.5-year analysis of susceptibilities of Candida Species to fluconazole and voriconazole as determined by CLSI standardized disk diffusion. J Clin Microbiol  2010; 48: 1366– 77. Google Scholar CrossRef Search ADS PubMed  Pfaller MA, Messer SA, Woosley LN et al.   Echinocandin and triazole antifungal susceptibility profiles for clinical opportunistic yeast and mold isolates collected from 2010 to 2011:application of new CLSI clinical breakpoints and epidemiological cutoff values for characterization of geographic and temporal trends of antifungal resistance. J Clin Microbiol  2013; 51: 2571– 81. Google Scholar CrossRef Search ADS PubMed  Rossignol T, Logue ME, Reynolds K et al.   Transcriptional response of Candida parapsilosis following exposure to farnesol. Antimicrob Agents Ch  2007; 51: 2304– 12. Google Scholar CrossRef Search ADS   Sanglard D, Ischer F, Calabrese D et al.   The ATP binding cassette transporter gene CgCDR1 from Candida glabrata is involved in the resistance of clinical isolates to azole antifungal agents. Antimicrob Agents Ch  1999; 43: 2753– 65. Sanguinetti M, Posteraro B, Fiori B et al.   Mechanisms of azole resistance in clinical isolates of Candida glabrata collected during a hospital survey of antifungal resistance. Antimicrob Agents Ch  2005; 49: 668– 79. Google Scholar CrossRef Search ADS   Schwarzmuller T, Ma B, Hiller E et al.   Systematic phenotyping of a large-scale Candida glabrata deletion collection reveals novel antifungal tolerance genes. PLoS Pathog  2014; 10: e1004211. Google Scholar CrossRef Search ADS PubMed  Subcommittee on Antifungal Susceptibility Testing (AFST) of the ESCMID European Committee for Antimicrobial Susceptibility Testing (EUCAST) Rodriguez-Tudela JL, Barchiest F, Bille J et al.   EUCAST discussion documente E.Dis 7.1: Method for the determination of minimum inhibitory concentration (MIC) by broth dilution of fermentative yeasts. Clin Microbiol Infect  2003; 9: 1– 8. Google Scholar CrossRef Search ADS PubMed  Torelli R, Posteraro B, Ferrari S et al.   The ATP-binding cassette transporter-encoding gene CgSNQ2 is contributing to the CgPDR1-dependent azole resistance of Candida glabrata. Mol Microbiol  2008; 68: 186– 201. Google Scholar CrossRef Search ADS PubMed  Tsai HF, Krol AA, Sarti KE et al.   Candida glabrata PDR1, a transcriptional regulator of a pleiotropic drug resistance network, mediates azole resistance in clinical isolates and petite mutants. Antimicrob Agents Ch  2006; 50: 1384– 92. Google Scholar CrossRef Search ADS   Tsai HF, Sammons LR, Zhang X et al.   Microarray and molecular analyses of the azole resistance mechanism in Candida glabrata oropharyngeal isolates. Antimicrob Agents Ch  2010; 54: 3308– 17. Google Scholar CrossRef Search ADS   Ueno K, Matsumoto Y, Uno J et al.   Intestinal resident yeast Candida glabrata requires Cyb2p-mediated lactate assimilation to adapt in mouse intestine. PLoS One  2011; 6: e24759. Google Scholar CrossRef Search ADS PubMed  Vale-Silva L, Beaudoing E, Tran VDT et al.   Comparative genomics of two sequential Candida glabrata clinical isolates. G3  2017; 7: 2413– 26. Google Scholar CrossRef Search ADS PubMed  Vale-Silva L, Ischer F, Leibundgut-Landmann S et al.   Gain-of-function mutations in PDR1, a regulator of antifungal drug resistance in Candida glabrata, control adherence to host cells. Infect Immun  2013; 81: 1709– 20. Google Scholar CrossRef Search ADS PubMed  Vermitsky JP, Earhart KD, Smith WL et al.   Pdr1 regulates multidrug resistance in Candida glabrata: gene disruption and genome-wide expression studies. Mol Microbiol  2006; 61: 704– 22. Google Scholar CrossRef Search ADS PubMed  Vermitsky JP, Edlind TD. Azole resistance in Candida glabrata: coordinate upregulation of multidrug transporters and evidence for a Pdr1-like transcription factor. Antimicrob Agents Ch  2004; 48: 3773– 81. Google Scholar CrossRef Search ADS   Xu N, Ye C, Chen X et al.   Genome sequencing of the pyruvate-producing strain Candida glabrata CCTCC M202019 and genomic comparison with strain CBS138. Sci Rep  2016; 6: 34893. Google Scholar CrossRef Search ADS PubMed  © FEMS 2017. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png FEMS Yeast Research Oxford University Press

Comparative genomic and transcriptomic analyses unveil novel features of azole resistance and adaptation to the human host in Candida glabrata

Loading next page...
 
/lp/ou_press/comparative-genomic-and-transcriptomic-analyses-unveil-novel-features-eIK0BokoZl
Publisher
Blackwell
Copyright
© FEMS 2017. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
1567-1356
eISSN
1567-1364
D.O.I.
10.1093/femsyr/fox079
Publisher site
See Article on Publisher Site

Abstract

Abstract The frequent emergence of azole resistance among Candida glabrata strains contributes to increase the incidence of infections caused by this species. Whole-genome sequencing of a fluconazole and voriconazole-resistant clinical isolate (FFUL887) and subsequent comparison with the genome of the susceptible strain CBS138 revealed prominent differences in several genes documented to promote azole resistance in C. glabrata. Among these was the transcriptional regulator CgPdr1. The CgPdr1 FFUL887 allele included a K274Q modification not documented in other azole-resistant strains. Transcriptomic profiling evidenced the upregulation of 92 documented targets of CgPdr1 in the FFUL887 strain, supporting the idea that the K274Q substitution originates a CgPdr1 gain-of-function mutant. The expression of CgPDR1K274Q in the FFUL887 background sensitised the cells against high concentrations of organic acids at a low pH (4.5), but had no detectable effect in tolerance towards other environmental stressors. Comparison of the genome of FFUL887 and CBS138 also revealed prominent differences in the sequence of adhesin-encoding genes, while comparison of the transcriptome of the two strains showed a significant remodelling of the expression of genes involved in metabolism of carbohydrates, nitrogen and sulphur in the FFUL887 strain; these responses likely reflecting adaptive responses evolved by the clinical strain during colonisation of the host. antifungal resistance, Candida glabrata, comparative genomics and comparative transcriptomics, CgPdr1 INTRODUCTION An alarming increase in the incidence of infections caused by Candida glabrata has been reported in the last years (Fidel, Vazquez and Sobel 1999; Khatib et al.2016). As such, in most epidemiological surveys this species ranks as the second major causative agent of invasive fungal infections worldwide, behind C. albicans and in some cases it is even the most frequently isolated species (Fidel, Vazquez and Sobel 1999; Pfaller et al.2010). This increase in the incidence of infections caused by C. glabrata is believed to result from its naturally high resilience to azoles, the frontline antifungal therapy used to treat candidiasis (Fidel, Vazquez and Sobel 1999; Pfaller et al.2010). The remarkably high rate at which C. glabrata strains acquire resistance to azoles, higher than the one registered for any other Candida spp., is another key factor contributing for the emergence of resistant strains (Borst et al.2005; Pfaller et al.2013; Kolaczkowska and Kolaczkowski 2016). Mutations in Erg11, the enzyme targeted by azoles, is not a primary mechanism of resistance in C. glabrata, in contrast with what is observed to occur for other Candida species (Sanguinetti et al.2005; Tsai et al.2010; Kolaczkowska and Kolaczkowski 2016). This observation suggests that C. glabrata has evolved other mechanisms to cope with azole stress, in particular through the activity of multidrug resistance (MDR) transporters (Sanguinetti et al.2005; Kolaczkowska and Kolaczkowski 2016). Specifically, several reports have underlined the important role of the ABC drug-efflux pumps CgCDR1 and CgPDH1 in contributing to azole resistance both in both laboratory strains and in resistant clinical isolates (Sanglard et al.1999; Vermitsky and Edlind 2004; Sanguinetti et al.2005; Torelli et al.2008; Kolaczkowska and Kolaczkowski 2016). More recently, the involvement of MDR transporters belonging to the major facilitator superfamily in azole resistance of C. glabrata has also been described (Costa et al.2014). The transcriptional regulation of drug-efflux pumps in C. glabrata, as in other yeasts, relies on the activity of the well-organised regulatory pleiotropic-drug resistance (PDR) network (Kolaczkowska and Kolaczkowski 2016). In C. glabrata, the key regulator of the PDR network is the transcription factor CgPdr1 and, concomitantly, this has been demonstrated to play an essential role in conferring tolerance to azoles (Vermitsky and Edlind 2004; Vermitsky et al.2006), including in resistant clinical isolates (Sanglard et al.1999; Sanguinetti et al.2005; Tsai et al.2006, 2010; Ferrari et al.2009; Caudle et al.2011). CgPdr1 has been implicated in the regulation of the drug-efflux pumps CgCDR1, CgPDH1, CgQDR2 and CgYOR1 (Kolaczkowska and Kolaczkowski 2016). Other azole-responsive genes regulated by CgPdr1 have biological functions related to stress response, metabolism of fatty acids and sterols, transcriptional regulation and adhesion (Vermitsky et al.2006; Ferrari et al.2009, 2011; Tsai et al.2010; Caudle et al.2011). Analysis of the coding sequence of the CgPDR1 gene in susceptible and resistant clinical isolates identified a panoply of gain-of-function (GOF) mutations that are believed to constitutively activate the transcription factor resulting in enhanced C. glabrata azole resistance (Vermitsky and Edlind 2004; Vermitsky et al.2006; Torelli et al.2008; Ferrari et al.2009; Caudle et al.2011). In this work, the genome and transcriptome of the reference strain C. glabrata CBS138 and of a clinical isolate (named FFUL887) recovered along the course of an epidemiological survey undertaken in hospitals of the Lisbon area herein demonstrated to be resistant to voriconazole and fluconazole (while the CBS138 strain was found to be sensitive) were compared. Besides providing clues to the mechanisms underlying acquisition of azole resistance inside the host, the results may also contribute for a better understanding of the different responses evolved by C. glabrata in human colonisation. METHODS Strains and growth media The strains used in this work are listed in Table 1. In addition to the laboratory strains listed, a cohort of 58 Candida glabrata clinical isolates recovered from patients attending three major Hospitals of the Lisbon area between 2000 and 2008 was also used (Table S1, Supporting Information). The strains were cultivated in rich growth media Yeast Peptone Dextrose (YPD), in RPMI (Roswell Park Memorial Institute Medium) or in minimal medium (MM). YPD contains, per litre, 20 g glucose (Merck Millipore), 10 g yeast extract (HiMedia Laboratories, Mumbai, India) and 20 g Peptone (HiMedia Laboratories). RPMI contains, per litre, 20.8 g RPMI-1640 synthetic medium (Sigma), 36 g glucose (Merck Millipore), 0.3g of L-glutamine (Sigma) and 0.165 mol/L of MOPS (3-(N-morpholino) propanesulfonic acid, Sigma). MM contains, per litre, 20 g glucose (Merck Millipore), 2.65 g (NH4)2 SO4 (Merck Millipore) and 1.7 g of yeast nitrogen base without amino acids and without ammonium sulphate (Difco). YPD and MM medium were sterilised by autoclaving for 15 min at 121°C and 1 atm, while RPMI medium was filtered with a 0.22-μm pore size filter and preserved at 4°C until further use. Solid media was prepared supplementing the corresponding liquid media with 2% agar. Table 1. List of strains used in this study. Strain  Parent  Description  Reference  KUE100  2001H  Parent strain, histidine auxotroph, the recipient enable high efficient gene targeting in which yku80 is repressed with a SAT1 flipper  Ueno et al. (2011)  KUE100_ΔCgPdr1  KUE100  ΔCgPDR1 strain, CgPDR1 (CAGL0A00451g) was replaced with the CgHIS3 marker  This study  CBS138  –  Reference strain  CBS-KNAW Fungal Biodiversity Centre  Strain  Parent  Description  Reference  KUE100  2001H  Parent strain, histidine auxotroph, the recipient enable high efficient gene targeting in which yku80 is repressed with a SAT1 flipper  Ueno et al. (2011)  KUE100_ΔCgPdr1  KUE100  ΔCgPDR1 strain, CgPDR1 (CAGL0A00451g) was replaced with the CgHIS3 marker  This study  CBS138  –  Reference strain  CBS-KNAW Fungal Biodiversity Centre  View Large Table 1. List of strains used in this study. Strain  Parent  Description  Reference  KUE100  2001H  Parent strain, histidine auxotroph, the recipient enable high efficient gene targeting in which yku80 is repressed with a SAT1 flipper  Ueno et al. (2011)  KUE100_ΔCgPdr1  KUE100  ΔCgPDR1 strain, CgPDR1 (CAGL0A00451g) was replaced with the CgHIS3 marker  This study  CBS138  –  Reference strain  CBS-KNAW Fungal Biodiversity Centre  Strain  Parent  Description  Reference  KUE100  2001H  Parent strain, histidine auxotroph, the recipient enable high efficient gene targeting in which yku80 is repressed with a SAT1 flipper  Ueno et al. (2011)  KUE100_ΔCgPdr1  KUE100  ΔCgPDR1 strain, CgPDR1 (CAGL0A00451g) was replaced with the CgHIS3 marker  This study  CBS138  –  Reference strain  CBS-KNAW Fungal Biodiversity Centre  View Large Assessment of resistance to antifungals To assess the resistance of the C. glabrata clinical isolates and of the reference strain CBS138 to voriconazole, fluconazole, anidulafungin and caspofungin, the MIC50 (minimum inhibitory concentration) of each of the antifungals was estimated using the microdilution method recommended by EUCAST (Subcommittee on Antifungal Susceptibility Testing (AFST) of the ESCMID European Committee for Antimicrobial Susceptibility Testing (EUCAST) 2003). The concentrations of fluconazole tested ranged from 0.125 to 64 mg/L and for the remaining antifungals ranged from 0.015 to 8 mg/L. MIC50 value was taken as being the first concentration of antifungal that reduced growth of the strains to half that registered in drug-free medium, as defined by EUCAST (2003). The MIC50 of each antifungal determined for the strains was compared with the clinical breakpoints recommended by EUCAST (32 mg/L for fluconazole, 0.06 mg/L for anidulafungin) to classify the strains as resistant (the MIC50 determined is above the defined breakpoint), susceptible (MIC50 equal or below the breakpoint) or intermediate (MIC50 shows represent a dose-dependent sensitivity according to EUCAST). Since a breakpoint value is not yet established for voriconazole in C. glabrata, the epidemiological cut-off value (1 mg/L) was used to detect non-WT isolates as recommended (Pfaller et al.2011). The stock solutions of the antifungals were prepared from the powder and using DMSO (dimethyl sulfoxide, Sigma) as the solvent. Fluconazole was purchased from Sigma, voriconazole and anidulafungin were kindly provided by Pfizer and caspofungin was provided by Merck Sharp Dohme. Growth curves of the C. glabrata reference strain CBS138 and of the FFUL887 isolate in the presence of fluconazole and voriconazole were performed using essentially the same experimental setup described above for the microdilution method with the difference that instead of measuring OD of the culture only after 24 h, this was measured every 30 min during 42 h at OD590nm. FFUL887 genomic DNA extraction and whole-genome sequencing FFUL887 cells were cultivated in YPD growth medium (up to an OD600nm of ∼3.0) and then centrifuged at 5000 rpm for 5 min at 4°C. The pellet was resuspended in 1 ml solution A (1 M Sorbitol (Sigma); 0.1 M EDTA (tetrasodium salt dehydrate) at pH 7.5). Afterwards, 10 mg/mL zymolyase (Zymo research) was added to the cellular suspension and the solution was incubated at 37˚C until protoplast formation. The suspension was centrifuged at 5000 rpm for 5 min and the pellet was resuspended in 1 mL solution B (50 mM Tris-HCL at pH 7.4 (Sigma-Aldrich), 20 mM EDTA). After this step, 30 μL of SDS (10%) was added to the mixture and this was left for 30 min at 65°C. Potassium acetate (250 μL, 5 M; Merck) was subsequently added and the mixture was left for 1 h on ice. The suspension was clarified by centrifugation (10 000 rpm, 10 min), and the supernatant was transferred to two fresh microfuge tubes. One volume of cold isopropanol was used to precipitate the pellet followed by centrifugation at 5000 rpm for 15 min. Supernatant was discarded, and the resulting pellet was incubated in 1 mL ethanol 70% during 5 min, and washed with ethanol 70% twice. The pellet was then dried in speed vacuum and resuspended in 200 μL TE (pH 7.4). 0.5 μL of RNase (10 mg/ml) was added followed by 1-h incubation at 37°C. Mixture was centrifuged at 10 000 rpm for 15 min, and the supernatant was preserved at 4°C till further use. FFUL887 genome sequencing was based on Ion Torrent and was performed by Stab Vida (Portugal) as a paid service. Two rounds of paired-end sequencing were performed which resulted in ∼6 million reads with an average size of 199 bp. The reads were trimmed based on quality and for SNP calling the trimmed reads were mapped against the reference genome of CBS138 (available on Candida Genome Database) using CLC Genomics Workbench. To increase confidence on the results obtained variant detection was performed from the mapped reads using both the probabilistic and quality-based variant detection tools embedded on CLC. To annotate the genome of the FFUL887 strain, the reads were de novo assembled into 799 contigs yielding a total of assembled bases of 12.29 Mb and a genome coverage of around ×96 Ab initio gene detection was performed using different algorithms followed by manual curation to select the more appropriate gene models. The sequence and annotation of the genome of the FFUL887 clinical strain have been deposited in ENA (http://www.ebi.ac.uk/ena/data/view/FWDN01000001-FWDN01000799). CgPDR1 gene disruption The deletion of the C. glabrata PDR1 (CAGL0A00451g) in strain KUE100 (parental to the ΔCgPDR1 mutant used in spot assays) was carried out using the method described by Ueno et al. (2011). The target gene CgPDR1 was replaced by a DNA cassette including the CgHIS3 gene through homologous recombination. The replacement cassette was prepared by PCR (Table 2). Recombination locus and gene deletion were verified by PCR. Deletion of the same gene in the background of the clinical isolate strain FFUL887 was carried out using a little modified method from the one above described for the KUE100 strain. In specific, CgPDR1 was replaced by a DNA cassette containing the zeocin-resistant marker by homologous recombination. Transformants were isolated on YPD supplemented with 100 μg/ml zeocin. Correct insertion of the zeocine resistance cassette and corresponding deletion of CgPDR1 were verified by PCR and by confirming the absence of CgPDR1 expression in the mutant strains by real-time RT-PCR (primers available in Table 2). Table 2. List of primers used in this study. Primer identification  Primer sequence  CgRDN5.8  Forward 5΄-AACAACTCACCGGCCGAAT-3΄    Reverse 5΄-CTTGGTTCTCGCATCGATGA-3΄  CgPDR1  Forward 5΄-CGATTGCCAACCCGTTAGA-3΄    Reverse 5΄-GACGACCTTGGTGTAGGAGTCAT-3΄  CgCDR1  Forward 5΄-GCTTGCCCGCACATTGA-3΄    5΄-CCTCAGGCAGAGTGTGTTCTTTC-3΄  CgPDH1  Forward 5΄-GCCATGGTACCTGCATCGAT-3΄    5΄-CCGAGGAATAGCAAAACCAGTATAC-3΄  CgQDR2  Forward 5΄-TCACTGCATAGTTTCATATCGGACTA-3΄    Reverse 5΄-TGCCGATATGTTCCCAAGTGA-3΄  Primer identification  Primer sequence  CgRDN5.8  Forward 5΄-AACAACTCACCGGCCGAAT-3΄    Reverse 5΄-CTTGGTTCTCGCATCGATGA-3΄  CgPDR1  Forward 5΄-CGATTGCCAACCCGTTAGA-3΄    Reverse 5΄-GACGACCTTGGTGTAGGAGTCAT-3΄  CgCDR1  Forward 5΄-GCTTGCCCGCACATTGA-3΄    5΄-CCTCAGGCAGAGTGTGTTCTTTC-3΄  CgPDH1  Forward 5΄-GCCATGGTACCTGCATCGAT-3΄    5΄-CCGAGGAATAGCAAAACCAGTATAC-3΄  CgQDR2  Forward 5΄-TCACTGCATAGTTTCATATCGGACTA-3΄    Reverse 5΄-TGCCGATATGTTCCCAAGTGA-3΄  View Large Table 2. List of primers used in this study. Primer identification  Primer sequence  CgRDN5.8  Forward 5΄-AACAACTCACCGGCCGAAT-3΄    Reverse 5΄-CTTGGTTCTCGCATCGATGA-3΄  CgPDR1  Forward 5΄-CGATTGCCAACCCGTTAGA-3΄    Reverse 5΄-GACGACCTTGGTGTAGGAGTCAT-3΄  CgCDR1  Forward 5΄-GCTTGCCCGCACATTGA-3΄    5΄-CCTCAGGCAGAGTGTGTTCTTTC-3΄  CgPDH1  Forward 5΄-GCCATGGTACCTGCATCGAT-3΄    5΄-CCGAGGAATAGCAAAACCAGTATAC-3΄  CgQDR2  Forward 5΄-TCACTGCATAGTTTCATATCGGACTA-3΄    Reverse 5΄-TGCCGATATGTTCCCAAGTGA-3΄  Primer identification  Primer sequence  CgRDN5.8  Forward 5΄-AACAACTCACCGGCCGAAT-3΄    Reverse 5΄-CTTGGTTCTCGCATCGATGA-3΄  CgPDR1  Forward 5΄-CGATTGCCAACCCGTTAGA-3΄    Reverse 5΄-GACGACCTTGGTGTAGGAGTCAT-3΄  CgCDR1  Forward 5΄-GCTTGCCCGCACATTGA-3΄    5΄-CCTCAGGCAGAGTGTGTTCTTTC-3΄  CgPDH1  Forward 5΄-GCCATGGTACCTGCATCGAT-3΄    5΄-CCGAGGAATAGCAAAACCAGTATAC-3΄  CgQDR2  Forward 5΄-TCACTGCATAGTTTCATATCGGACTA-3΄    Reverse 5΄-TGCCGATATGTTCCCAAGTGA-3΄  View Large Transcriptomic analysis The transcriptomes of FFUL887 and CBS138 strains were compared using DNA microarrays specifically designed for C. glabrata (design ID 064590) (Rossignol et al.2007). Both strains were cultivated overnight in 25 mL of YPD at 30°C with orbital agitation (250 rpm) and then re-inoculated (at an OD600nm of 0.2 ± 0.05) in 150 mL of fresh RPMI. The cells were harvested by centrifugation (8000 × g, 7 min, 4°C—Beckman J2.21 Centrifuge, rotor JA.10) in mid-exponential phase (OD600nm∼2 ± 0.05) and immediately frozen at –80°C until further use. RNA extraction was performed as described in Bernardo et al. (2017) and in Rossignol et al. (2007). Comparison of gene transcript levels based on real time RT-PCR The expression of CgCDR1, CgPDH1, CgQDR2 and CgPDR1 genes was compared in CBS138, in FFUL887 and in FFUL_ΔCgpdr1 strains by real-time RT-PCR. Cells of the different strains were cultivated in identical conditions to those used for the microarray analysis. Conversion of total RNA into cDNA was performed using 1 μg of RNA. The reverse-transcription step was performed in a C1000 Thermal Cycler (Bio-Rad, Hercules, USA). The subsequent quantitative PCR step was performed using 2.5 μL of the cDNA. The sequence of the primers used is available on request. Gene expression was calculated using CgRDN5.8 as an internal control. Comparison of susceptibility to environmental stressors based on spot assays Comparison of susceptibility of CBS138, KUE100, KUE100_ΔCgPDR1, FFUL887 and FFUL887_ΔCgPDR1 cells to inhibitory concentrations of H2O2 and of the organic acids acetic, propionic and butyric acids was based on spot assays. Cells of the different strains were cultivated in MM growth medium until mid-exponential phase (OD600nm∼0.8) and then diluted in 1 mL of sterile water to obtain a cell suspension having an OD600nm of 0.05. Four microlitres of this cell suspension and of two subsequent dilutions (1:5 and 1:10) were applied onto the surface of MM agarised plates supplemented or not with inhibitory concentrations of H2O2 (5–18 mM), of acetic acid (50–60 mM), of propionic acid (17–20 mM) and of butyric acids (15–17 mM). The plates were incubated at 30°C for 2 to 3 days depending on the severity of growth inhibition. The same experimental setup was used to assess susceptibility of the strains to heat stress with the difference that instead of incubating the inoculated plates at 30°C, these were incubated at 37°C, 40°C or 42°C. RESULTS Identification of FFUL887 as a fluconazole- and voriconazole-resistant strain To characterise the incidence of resistance to fluconazole, voriconazole, caspofungin and anidulafungin in a cohort of 58 Candida glabrata clinical isolates, the concentration of each of these antifungals leading to a 50% growth inhibition (generally designated as MIC50), comparing with growth registered in drug-free medium, was determined. This phenotypic screening was performed using the highly standardised microdilution method recommended by EUCAST. As a control we have also included the reference strain CBS138 in the screening. The MIC value obtained for each drug and each isolate is shown in Fig. S1 (Supporting Information), and the distribution of MIC values across all isolates is shown in Fig. 1. The MIC values obtained were compared with the clinical resistance breakpoints defined by EUCAST (32 mg/L for fluconazole and 0.06 mg/L for anidulafungin) to classify the strains as resistant, intermediate or susceptible. To identify voriconazole-resistant strains, the MIC value obtained was compared with the ECV value (1 mg/mL) which can be used to distinguish wild-type from non-wild-type isolates (Pfaller et al.2011). For caspofungin, no breakpoint has been defined by EUCAST and therefore the strains were not classified. Under the experimental conditions used, the MIC value of the reference strain CBS138 was 16 mg/L for fluconazole and 0.25 mg/L for voriconazole, indicating that the reference strain is susceptible to these two azoles. Seven isolates (FFUL412, FFUL443, FFUL674, FFUL830, FFUL866, FFUL878, FFUL887) were resistant to fluconazole and voriconazole, two only resistant to fluconazole (FFUL98 and FFUL4012) and one only resistant to voriconazole (FFUL677) (Fig. S1). Notably, three of the cross-resistant isolates (FFUL412, FFUL443 and FFUL674) were retrieved from patients undergoing fluconazole-based therapy (Table S1). Despite the small number of isolates examined in this study, the percentage of resistance obtained for fluconazole and voriconazole (16% for fluconazole and 14% for voriconazole) are close to the values reported (10%–15%) in antifungal surveillance tests undertaken with much larger cohorts of strains (e.g. Pfaller et al.2010). None of the C. glabrata isolates tested could be considered susceptible to fluconazole as the MIC values were always above 0.002 mg/L (Fig. S1), consistent with the described increased resilience of C. glabrata to this azole drug (Fidel, Vazquez and Sobel 1999; Kolaczkowska and Kolaczkowski 2016). All the isolates tested exhibited high susceptibility to anidulafungin, none of them exhibiting growth when cultivated in the presence of 0.06 mg/L, the defined resistance breakpoint (Fig. 1B; Fig. S1). For caspofungin, only one isolate, FFUL887, exhibited a MIC value of 0.25, while for the remaining isolates this MIC value was of 0.125 or below (Fig. 1B; Fig. S1). Figure 1. View largeDownload slide Distribution of MIC50 values of fluconazole, voriconazole (A) caspofungin and anidulafungin (B) obtained for the cohort of C. glabrata clinical isolates tested in this work. The dashed line indicates the resistance breakpoints (in the case of voriconazole of the ECV value) defined by EUCAST (as detailed in materials and methods section). These results were obtained based on the assessment of MIC50 value by the microdilution method recommended by EUCAST and that gave rise to the results shown in Fig. S1. Figure 1. View largeDownload slide Distribution of MIC50 values of fluconazole, voriconazole (A) caspofungin and anidulafungin (B) obtained for the cohort of C. glabrata clinical isolates tested in this work. The dashed line indicates the resistance breakpoints (in the case of voriconazole of the ECV value) defined by EUCAST (as detailed in materials and methods section). These results were obtained based on the assessment of MIC50 value by the microdilution method recommended by EUCAST and that gave rise to the results shown in Fig. S1. The FFUL887 strain was selected for further analysis since it was demonstrated to be resistant to fluconazole and voriconazole and also exhibited higher resilience to caspofungin compared to the other clinical isolates in this study. To assess how the presence of the fluconazole and voriconazole affected growth kinetics of the FFUL887 and CBS138 strains, growth curves in liquid medium were performed using the same experimental setup that was used for estimation of MIC50 (Fig. S2, Supporting Information). Three concentrations of voriconazole and fluconazole were tested: one corresponding to the resistance breakpoint, one below and one above that value (Fig. S2). The results obtained show that the two strains exhibited a similar fitness when cultivated in drug-free medium, with only a slight decrease in the final biomass produced by the FFUL887 strain (Fig. S2). Supplementation of the RPMI medium with the two azole drugs led to a drastic growth inhibition of the CBS138 strain, whereas growth of the FFUL887 strain was almost identical to the control conditions, with only a small detectable decrease of the growth rate (e.g. 0.033 h−1 in the presence of 64 mg/L fluconazole and 0.044 h−1 in control conditions) (Fig. S2) and a slight increase in the lag phase that was observed upon inoculation in the drug-supplemented medium (Fig. S2). To assess if the resistant phenotype exhibited by the FFUL887 strain towards fluconazole and voriconazole was generalised for azoles or was limited to azoles of the triazole family (including fluconazole or voriconazole), growth of this strain in the presence of the imidazoles ketoconazole and clotrimazole was examined (Fig. S3, Supporting Information). Cells were cultivated for 24 h in 96-multiwell plates containing RPMI or in the same growth medium supplemented with 4 mg/L ketoconazole and 1 mg/mL of clotrimazole, the defined resistance breakpoints for these two drugs. Figure S3 shows that FFUL887 cells are more tolerant to the tested concentrations of ketoconazole and clotrimazole than the CBS138 strain; however, the FFUL887 strain is still considered susceptible to the two imidazoles since the concentrations tested reduced growth by more than 50% the one registered in control conditions (Fig. S3). FFUL887 genome sequencing and annotation The fluconazole and voriconazole resistance phenotype exhibited by the FFUL887 strain prompted us to obtain the genome sequence of this isolate. The assembled contig sequences were used for automatic ab initio gene detection that was subsequently manually curated to select the more appropriate gene models. This analysis allowed us to predict that the ORFeome of the FFUL887 strain includes 5079 genes, which corresponds to 96% of the total number of genes annotated for the CBS138 strain (Dujon et al.2004). The number of ORFs obtained for the FFUL887 is also in line with those recently reported for other C. glabrata clinical isolates (in the range of 5300 CDSs) (Havelsrud and Gaustad 2017; Vale-Silva et al.2017). The vast majority (5039) of the protein pairs present in CBS138 and FFUL887 shared more than 90% identity at the amino acid level indicating that the proteins encoded by the two strains are fairly similar. For the identification of SNPs that could underlie the observed resistance to fluconazole and voriconazole of the FFUL887 strain, the reads obtained were mapped against the genome sequence of the CBS138 strain, as detailed in materials and methods section. This comparison yielded 77 749 SNPs between the genomes of FFUL887 and CBS138 (Fig. S4, Supporting Information). A similar high number of SNPs was also reported in a recent comparative genomic analysis between two C. glabrata clinical isolates and the reference strain CBS138 (Vale-Silva et al.2017). About 45% of the SNPs identified between CBS138 and FFUL887 were located in coding regions, affecting 3194 of the gene sequences predicted for the FFUL887 strain (Fig. S4). The percentage of genes harbouring non-synonymous SNPs in the FFUL887 strain was similar throughout the nine C. glabrata nuclear chromosomes (∼60%) but considerably smaller (∼18%) in the mitochondrial chromosome (Fig. S4). On average, FFUL887 and CBS138 orthologous genes harboured five non-synonymous SNPs; however, in some cases this number increased up to more than 30 non-synonymous SNPs including in CAGL0K12078g (>50 non-synonymous SNPs), encoding a putative transcription factor similar to ScNrg1 transcription factor; CAGL0C00231g (42 non-synonymous SNPs), encoding a presumed plasma membrane nucleobase transporter, and the adhesin CgPWP4 (52 non-synonymous SNPs) (Fig. 2). Other adhesin-encoding genes were also observed to harbour a high number of non-synonymous SNPs between FFUL887 and CBS138 including CgEPA8, CgPWP5, CAGL0C03575g and CAGL0L10092g (Fig. 2). Figure 2. View largeDownload slide Number of non-synonymous SNPs present in FFUL887 predicted proteins, when compared with their CBS138 counterpartners. The reads obtained after whole-genome sequencing of the FFUL887 strain were mapped against the genome of the reference strain CBS138, as detailed in materials and methods section. Those genes exhibiting a higher number of non-synonymous SNPs in the FFUL887 strain are highlighted in the figure. Adhesin-encoding genes are evidenced in light blue. Figure 2. View largeDownload slide Number of non-synonymous SNPs present in FFUL887 predicted proteins, when compared with their CBS138 counterpartners. The reads obtained after whole-genome sequencing of the FFUL887 strain were mapped against the genome of the reference strain CBS138, as detailed in materials and methods section. Those genes exhibiting a higher number of non-synonymous SNPs in the FFUL887 strain are highlighted in the figure. Adhesin-encoding genes are evidenced in light blue. Notably, no SNPs were found in the sequence of ERG11 gene encoded by FFUL887, consistent with the idea that azole resistance of these strains is not driven by alterations in the drug target as shown for most C. glabrata azole-resistant isolates (Sanguinetti et al.2005; Tsai et al.2010; Kolaczkowska and Kolaczkowski 2016). No SNPs were also found in FFUL887 Msh2, a DNA repair protein whose mutations had been linked to the development of azole resistance in C. glabrata-resistant clinical strains (Healey et al.2016). Concerning the CgFKS1 and CgFKS2, the two enzymes targeted by echinocandins, the FFUL887 alleles harbour one non-synonymous SNP each (Gly14Ser in CgFks1 and Thr926Pro in CgFks2), compared to CBS138 orthologues; however, these polymorphisms are outside of the hotspot regions commonly found to be altered in echinocandin-resistant isolates (Garcia-Effron et al.2009). It thus remains to be examined whether the higher tolerance of the FFUL887 strain to caspofungin comes from these polymorphisms or if it results from other genetic traits. Comparative genomic analysis between FFUL887 and CBS138 focused on azole-resistance genes: emphasis on CgPdr1 To gather insights into the differential levels of resistance to fluconazole and voriconazole of the CBS138 and FFUL887 strains, the set of proteins found to harbour non-synonymous SNPs in the resistant strain were compared with a comprehensive list of 78 genes previously implicated in C. glabrata resistance to fluconazole and voriconazole (listed in Table S2, Supporting Information). Fifty-three genes associated with resistance to these two azoles differed in FFUL887 and in CBS138, a subset of these being shown in Table 3. Table 3. Subset of proteins previously described to be involved in fluconazole and/or voriconazole resistance in C. glabrata and that were found to harbour non-synonymous SNPs in the FFUL887 strain, when compared with their CBS138 counterpartners. Gene or ORF name  Function  Amino acid modification found in FFUL887  CgPDR1  Zinc finger transcription factor, activator of drug resistance genes  Val91Ile; Leu98Ser; Asp243Asn; Lys274Gln  CgGAL11ACgGAL11B  Component of the transcriptional Mediator complex that provides interfaces between RNA polymerase II and upstream activator proteins; essential for CgPdr1-dependent activation of azole-resistance genes  Ser134Asn; Ser965Gly; Ser1084AsnGln203Pro; His322Asn; Pro372Ser; Ile615Asn; *839Gly  CgPDH1  Multidrug transporter of the ATP-binding cassette superfamily  Lys438Gln; Glu839Asp  CAGL0L04400g  Zinc finger transcription factor involved in transcriptional regulation of MDR genes. Orthologue of S. cerevisiae ScYRR1  Cys24Gly; Ala58Val; Ile137Leu; Asp229Glu; Ile346Val; Glu574Lys; Ile593Leu; Glu710Asp; Ala933Val  CgUPC2A  Zinc finger transcription factor required for transcriptional regulation of genes involved in uptake and biosynthesis of ergosterol  Arg92Lys; Asn304Ser; Glu822Val  CgSIN3  Component of the Rpd3S and Rpd3L histone deacetylase complexes  Asn50Lys; Lys288Thr  CAGL0L03377g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Gly134Asp; Gly172Ser; Lys252Arg; Ile347Met; Ala695Gly; Lys813Arg; Asp819Glu; Thr867Met  CAGL0L09383g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Ser116Asn; Ile185Val; Met202Val  CgHST1  Histone deacetylase that regulates gene expression in niacin-limiting conditions  Met70Lys  Gene or ORF name  Function  Amino acid modification found in FFUL887  CgPDR1  Zinc finger transcription factor, activator of drug resistance genes  Val91Ile; Leu98Ser; Asp243Asn; Lys274Gln  CgGAL11ACgGAL11B  Component of the transcriptional Mediator complex that provides interfaces between RNA polymerase II and upstream activator proteins; essential for CgPdr1-dependent activation of azole-resistance genes  Ser134Asn; Ser965Gly; Ser1084AsnGln203Pro; His322Asn; Pro372Ser; Ile615Asn; *839Gly  CgPDH1  Multidrug transporter of the ATP-binding cassette superfamily  Lys438Gln; Glu839Asp  CAGL0L04400g  Zinc finger transcription factor involved in transcriptional regulation of MDR genes. Orthologue of S. cerevisiae ScYRR1  Cys24Gly; Ala58Val; Ile137Leu; Asp229Glu; Ile346Val; Glu574Lys; Ile593Leu; Glu710Asp; Ala933Val  CgUPC2A  Zinc finger transcription factor required for transcriptional regulation of genes involved in uptake and biosynthesis of ergosterol  Arg92Lys; Asn304Ser; Glu822Val  CgSIN3  Component of the Rpd3S and Rpd3L histone deacetylase complexes  Asn50Lys; Lys288Thr  CAGL0L03377g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Gly134Asp; Gly172Ser; Lys252Arg; Ile347Met; Ala695Gly; Lys813Arg; Asp819Glu; Thr867Met  CAGL0L09383g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Ser116Asn; Ile185Val; Met202Val  CgHST1  Histone deacetylase that regulates gene expression in niacin-limiting conditions  Met70Lys  View Large Table 3. Subset of proteins previously described to be involved in fluconazole and/or voriconazole resistance in C. glabrata and that were found to harbour non-synonymous SNPs in the FFUL887 strain, when compared with their CBS138 counterpartners. Gene or ORF name  Function  Amino acid modification found in FFUL887  CgPDR1  Zinc finger transcription factor, activator of drug resistance genes  Val91Ile; Leu98Ser; Asp243Asn; Lys274Gln  CgGAL11ACgGAL11B  Component of the transcriptional Mediator complex that provides interfaces between RNA polymerase II and upstream activator proteins; essential for CgPdr1-dependent activation of azole-resistance genes  Ser134Asn; Ser965Gly; Ser1084AsnGln203Pro; His322Asn; Pro372Ser; Ile615Asn; *839Gly  CgPDH1  Multidrug transporter of the ATP-binding cassette superfamily  Lys438Gln; Glu839Asp  CAGL0L04400g  Zinc finger transcription factor involved in transcriptional regulation of MDR genes. Orthologue of S. cerevisiae ScYRR1  Cys24Gly; Ala58Val; Ile137Leu; Asp229Glu; Ile346Val; Glu574Lys; Ile593Leu; Glu710Asp; Ala933Val  CgUPC2A  Zinc finger transcription factor required for transcriptional regulation of genes involved in uptake and biosynthesis of ergosterol  Arg92Lys; Asn304Ser; Glu822Val  CgSIN3  Component of the Rpd3S and Rpd3L histone deacetylase complexes  Asn50Lys; Lys288Thr  CAGL0L03377g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Gly134Asp; Gly172Ser; Lys252Arg; Ile347Met; Ala695Gly; Lys813Arg; Asp819Glu; Thr867Met  CAGL0L09383g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Ser116Asn; Ile185Val; Met202Val  CgHST1  Histone deacetylase that regulates gene expression in niacin-limiting conditions  Met70Lys  Gene or ORF name  Function  Amino acid modification found in FFUL887  CgPDR1  Zinc finger transcription factor, activator of drug resistance genes  Val91Ile; Leu98Ser; Asp243Asn; Lys274Gln  CgGAL11ACgGAL11B  Component of the transcriptional Mediator complex that provides interfaces between RNA polymerase II and upstream activator proteins; essential for CgPdr1-dependent activation of azole-resistance genes  Ser134Asn; Ser965Gly; Ser1084AsnGln203Pro; His322Asn; Pro372Ser; Ile615Asn; *839Gly  CgPDH1  Multidrug transporter of the ATP-binding cassette superfamily  Lys438Gln; Glu839Asp  CAGL0L04400g  Zinc finger transcription factor involved in transcriptional regulation of MDR genes. Orthologue of S. cerevisiae ScYRR1  Cys24Gly; Ala58Val; Ile137Leu; Asp229Glu; Ile346Val; Glu574Lys; Ile593Leu; Glu710Asp; Ala933Val  CgUPC2A  Zinc finger transcription factor required for transcriptional regulation of genes involved in uptake and biosynthesis of ergosterol  Arg92Lys; Asn304Ser; Glu822Val  CgSIN3  Component of the Rpd3S and Rpd3L histone deacetylase complexes  Asn50Lys; Lys288Thr  CAGL0L03377g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Gly134Asp; Gly172Ser; Lys252Arg; Ile347Met; Ala695Gly; Lys813Arg; Asp819Glu; Thr867Met  CAGL0L09383g  Predicted zinc finger transcription factor; required for resistance to voriconazole and fluconazole  Ser116Asn; Ile185Val; Met202Val  CgHST1  Histone deacetylase that regulates gene expression in niacin-limiting conditions  Met70Lys  View Large One of the proteins that differed between FFUL887 and CBS138 is CgPdr1, which exhibited four non-synonymous SNPs (Val91Ile, Leu98Ser, Asp243Asn and Lys274Gln), when compared with the corresponding CBS138 orthologue. Three of these SNPs (Val91Ile; Leu98Ser and Asp243Asn) were found to be present simultaneously in the CgPDR1 alleles encoded by isolates resistant and susceptible to azoles (Ferrari et al.2009), while the Lys274Gln was not previously described. Based on this observation, it was hypothesised that the substitution K274Q could represent a GOF substitution of C. glabrata CgPdr1. To test the hypothesis that K274Q mutation leads to hyperactivation of CgPdr1, the transcriptomes of the FFUL887 and CBS138 strains were compared in drug-free RPMI medium using species-specific DNA microarrays. Out of the 409 genes found to be overexpressed in the FFUL887 strain (above 1.5-fold, P-value below 0.001; listed in Table S3, Supporting Information), 89 genes are documented targets of CgPdr1, according to the information available in the PathoYeastract database (Monteiro et al.2017) (highlighted in grey in the Table S3). Among these were the well-characterised CgPdr1 targets CgPDH1 and CgCDR1, as well as CgPDR1 itself (Table S3). Deletion of CgPDR1K274Q abrogated the increase in transcription of CgPDH1, CgCDR1, CgPDR1 and CgQDR2 genes registered in the FFUL887 strain (Fig. 3A). As expected, deletion of the CgPDR1K274Q allele also resulted in sensitisation of the FFUL887 strain to fluconazole and voriconazole (Fig. 3B). Overall, the results of the transcriptomic profiling of the CBS138 and FFUL887 strains strongly support the idea that FFUL887 encodes a CgPdr1 GOF allele, resulting from the K274Q SNP. Figure 3. View largeDownload slide (A) MIC for fluconazole and voriconazole obtained for the CBS138, FFUL887 and the FFUL887_ΔCgPDR1 strains, as determined by the microdilution method recommended by EUCAST. For the statistical analysis, the results obtained for the mutant strain devoid of CgPDR1 gene were compared with those gathered for the wild-type FFUL887 strain. ****P-value below 0.0001; (B) comparison of the transcript levels of CgPDR1, CgCDR1, CgPDH1 and CgQDR2 genes in CBS138, FFUL887 and the FFUL887_ΔCgPDR1 strains. Cells of the different strains were cultivated in RPMI growth medium until mid-exponential phase after which the expression of CgPDR1, CgCDR1, CgPDH1 and CgQDR2 genes was compared by qRT-PCR. The values represented for the FFUL887 and FFUL887_ΔCgPDR1 strains are relative to the value obtained for the CBS138 strain, which was considered to be equal to 1. For the statistical analysis, the results obtained for the FFUL887 strain were compared with those gathered for CBS138, while the results obtained for the FFUL887_ΔCgPDR1 mutant were compared with those obtained for FFUL887. *P-value below 0.05, ****P-value below 0.0001. Figure 3. View largeDownload slide (A) MIC for fluconazole and voriconazole obtained for the CBS138, FFUL887 and the FFUL887_ΔCgPDR1 strains, as determined by the microdilution method recommended by EUCAST. For the statistical analysis, the results obtained for the mutant strain devoid of CgPDR1 gene were compared with those gathered for the wild-type FFUL887 strain. ****P-value below 0.0001; (B) comparison of the transcript levels of CgPDR1, CgCDR1, CgPDH1 and CgQDR2 genes in CBS138, FFUL887 and the FFUL887_ΔCgPDR1 strains. Cells of the different strains were cultivated in RPMI growth medium until mid-exponential phase after which the expression of CgPDR1, CgCDR1, CgPDH1 and CgQDR2 genes was compared by qRT-PCR. The values represented for the FFUL887 and FFUL887_ΔCgPDR1 strains are relative to the value obtained for the CBS138 strain, which was considered to be equal to 1. For the statistical analysis, the results obtained for the FFUL887 strain were compared with those gathered for CBS138, while the results obtained for the FFUL887_ΔCgPDR1 mutant were compared with those obtained for FFUL887. *P-value below 0.05, ****P-value below 0.0001. The deletion of CgPDR1K274Q allele affects tolerance to environmental stress of the FFUL887 isolate Previous studies have shown that deletion of GOF CgPdr1 alleles results in altered stress resilience of azole-resistant C. glabrata strains (Vermitsky et al.2006). We therefore compared growth of CBS138, FFUL887 and FFUL887_ΔCgPDR1K274Q in the presence of various environmental stressors including H2O2, acetic acid, propionic acid, butyric acid and at different temperatures (30°C, 37°C, 40°C or 42°C). Under the experimental conditions used, the deletion of CgPDR1 in the FFUL887 background led to a mild decrease in growth of the strains when cultivated at all the temperatures tested (Fig. 4). No significant differences were observed upon CgPDR1 deletion in the FFUL887 background concerning tolerance to H2O2 but, surprisingly, during cultivation in the presence of inhibitory concentrations of the organic acids acetic, propionic and butyric acids (at pH 4.5) the deletion of CgPDR1K274Q was beneficial (Fig. 4). These experiments were also performed in the genetic background of the laboratory strain KUE100 that encodes a wild-type CgPdr1 allele. The results obtained confirmed a slight protective effect exerted by CgPDR1 expression against high temperatures, while in the presence of acetic, propionic or butyric acids there was no significant differences in growth of the wild-type or of the ΔCgPDR1 mutant (Fig. 4). Figure 4. View largeDownload slide Comparison of the susceptibility of CBS138, FFUL887 and the FFUL887_ΔCgPDR1 strains to environmental stressors based on spot assays. Mid-exponential phase cells of the different strains were cultivated in solid MM growth medium or in this same medium supplemented with inhibitory concentrations of H2O2 or of the fatty acids acetic acid, propionic acid and butyric acid. Lanes b and c correspond to 1:5 and 1:10 dilutions of the cell suspension used in lane a. Growth was compared after 2 to 3 days of incubation at 30°C, depending on the severity of growth inhibition. The same experimental setup was used to compare tolerance of the strains to different temperatures. Figure 4. View largeDownload slide Comparison of the susceptibility of CBS138, FFUL887 and the FFUL887_ΔCgPDR1 strains to environmental stressors based on spot assays. Mid-exponential phase cells of the different strains were cultivated in solid MM growth medium or in this same medium supplemented with inhibitory concentrations of H2O2 or of the fatty acids acetic acid, propionic acid and butyric acid. Lanes b and c correspond to 1:5 and 1:10 dilutions of the cell suspension used in lane a. Growth was compared after 2 to 3 days of incubation at 30°C, depending on the severity of growth inhibition. The same experimental setup was used to compare tolerance of the strains to different temperatures. Comparative transcriptomic analysis between FFUL887 and CBS138 shows dramatic alterations in the expression of genes involved in carbohydrate, nitrogen and sulphur metabolism We have further explored our results concerning the comparison of the transcriptome of the FFUL887 and CBS138 strains during growth in RPMI medium aiming to gain insights into the responses evolved by C. glabrata during colonisation of the human urinary tract considering that this was the niche where FFUL887 isolate was retrieved from. The CBS138 strain has an intestinal origin; however, its extensive utilisation in the laboratory as likely resulted in its domestication leading to large phenotypic differences compared to those observed in C. glabrata clinical isolates including those of intestinal origin (Gregori et al.2007; Cunha et al.2017). The genes overexpressed (above 1.5-fold and having a P-value below 0.001) in the FFUL887 isolate were clustered according to their biological function using the MIPS functional catalogue (Fig. S5, Supporting Information). Results revealed a significant enrichment (P-value below 0.001) of genes involved in ‘metabolism of amino acids’, ‘metabolism of carbohydrates’, ‘nitrogen, sulphur and selenium metabolism’, ‘lipid, fatty acid and isoprenoid metabolism’, ‘generation of energy’, ‘vacuolar protein degradation’, ‘transport’ and ‘oxidative stress response’ (Fig. S5 and Table S3). Similarly, the set of genes upregulated in CBS138 (and consequently downregulated in FFUL887) was enriched (P-value below 0.001) in genes related with ‘protein synthesis’ (Fig. S5 and Table S3). The genes upregulated in the FFUL887 strain related with carbohydrate and lipid metabolism included enzymes involved in fatty-acid β-oxidation, in catabolism of acetate, of propionate and of glycogen, as well as genes encoding neoglucogenic and Krebs cycle enzymes (Fig. S6, Supporting Information). This observation was surprising considering that at the time point where FFUL887 and CBS138 cells were harvested for the microarray analysis (after 6 h of cultivation in the rich RPMI medium supplemented with 20 g/L glucose) there was still a considerable amount of glucose present in the culture supernatant of the two cultures (∼18.5 g/L, based on HPLC analysis of the supernatants). The genes upregulated in the FFUL887 strain related with amino acid and sulphur metabolism classes were essentially those involved in metabolism of various amino acids and genes of the trans-sulfuration pathway that allows transport and incorporation of sulphate in methione and cystheine (Fig. S6). Consistently, several transporters involved in the uptake of amino acids, small peptides and inorganic sulphur were also found to be upregulated in the FFUL887 strain (Fig. S6). DISCUSSION In this work, we have disclosed the genome sequence of a Candida glabrata clinical isolate, FFUL887, resistant to voriconazole and fluconazole and also exhibiting enhanced tolerance to caspofungin. The higher resistance of the FFUL887 strain to these two types of antifungals is striking considering that they have different modes of action. Resistance to voriconazole and fluconazole in FFUL887 was largely dependent on the expression of CgPdr1; however, this is not likely to underlie the higher tolerance of this strain to caspofungin since CgPdr1 expression is dispensable for C. glabrata tolerance to echinocandinds (Schwarzmuller et al.2014). A very high number of SNPs were obtained when comparing the genomic sequences of CBS138 and FFUL887, probably reflecting the different genetic background of these two strains. Nevertheless, the vast majority of the proteins encoded by the two strains were still very similar with >90% of the proteins encoded by the two strains sharing a degree of homology above 90%. Recent comparative genomic analysis between C. glabrata clinical isolates and the CBS138 strain also revealed very prominent differences, within the range of those reported in our study (Havelsrud and Gaustad 2017; Vale-Silva et al.2017), while a strain used for carboxylic acids production was much more similar to CBS138 (Xu et al.2016). Interestingly, even the comparison of cohorts of related clinical isolates shows very prominent differences, similar to those that are observed when the isolates are compared with the CBS138 strain (Vale-Silva et al.2017). These observations reflect the described genomic plasticity of C. glabrata species which gives rise to a large genetic and phenotypic diversity among isolated strains (Carreté et al.2017). While in our study we have focused on the comparison between the transcriptome and genome of an azole-resistant strain with the susceptible CBS138 strain, others have performed similar analyses but using related isolates (for example, strains retrieved from patients obtained before and after application of azole-therapy (Vermitsky and Edlind 2004; Vermitsky et al.2006; Vale-Silva et al.2017)). In our case, a similar approach was not possible since we could not identify among the sensitive strains tested one that could be related with FFUL887. Necessarily, the option of comparing the genomes of CBS138 vs FFUL887 is difficult for the establishment of genotype–phenotype associations; however, it has the advantage of allowing the identification of new SNPs that could be relevant for azole resistance and that would not be detected when comparing two already adapted clinical strains such as modifications occurring early during the process of colonisation of the human host. Besides CgPdr1, several other well-characterised determinants of C. glabrata resistance to azoles were found to harbour SNPs in the resistant strain FFUL887 including the MDR transporters CgPhd1, CgTpo1_1 and CgTpo1_2 and the transcriptional regulators CgUpc2A, CgYrr1, CgStb5 and CgGal11A (Table S2). It is difficult to understand whether these polymorphisms contribute to the higher resistance exhibited by FFUL887 cells towards fluconazole and voriconazole because the biochemical activity of these proteins is not well studied, and therefore it is hard to predict the consequences for protein activity of the identified SNPs. Nevertheless, these indicatives deserve further exploration since azole-resistance genes are surely under selective pressure, as occurs with CgPdr1. The extensive upregulation of about 90 documented targets of CgPdr1 that was observed in the FFUL887 strain during cultivation in drug-free growth medium strongly supports the idea that the K274Q substitution is, indeed, a new GOF mutation of this protein. Interestingly, a K274N substitution has been previously reported, however, while this results in mild increase in fluconazole tolerance (MIC of 16 mg/L)(Caudle et al.2011), the herein reported K274Q substitution results in a much higher resistance (MIC of 64 mg/L). The K274 residue lies within a region of CgPdr1 where several other mutations have been described (as detailed in Fig. S7, Supporting Information) and is located near a predicted inhibitory regulatory domain of CgPdr1 (residues 322–465) (Fig. S7). In Saccharomyces cerevisiae, this regulatory domain inhibits the activity of ScPdr1 (Kolaczkowska et al.2002), for which it can be hypothesised that the K274Q modification could compromise the function of the inhibitory domain resulting in an hyperactivation of CgPdr1. Further studies are required to better understand how the K274Q and other GOF mutations modulate the activity of CgPdr1. The genes that are under regulation of different CgPdr1 GOF mutants have a modest overlap (Ferrari et al.2009, 2011; Tsai et al.2010; Caudle et al.2011). In order to determine the effect of the CgPdr1 K274Q substitution in C. glabrata genomic expression, the 92 genes upregulated in the FFUL887 strain harbouring a PDRE motif (TCCRYGSR) in their promoter (presumed to be the direct targets of CgPdr1) were compared with the set of genes regulated by three other GOF alleles: P927L and L946S and K274N (Fig. 5). Only five genes were in common in the three datasets: CgCDR1, CgYOR1, CgPDR1, CgPUP1 and CAGL0M09713g (Fig. 5). Consistently, CgCDR1 and CgPUP1 genes were recently shown to be upregulated among a cohort of CgPdr1 GOF mutants different from those used to build (Fig. 5) (Ferrari et al.2011). The pattern of expression of other drug-efflux pumps varied according to the GOF mutation: while K274Q, L94S and P927L led to the upregulation of CgQDR2 and CgPHD1, K274N was the only mutation causing upregulation of CgTPO1–1 (Fig. 5). The expression of adhesin-encoding genes was also found to vary according to the CgPdr1 GOF mutation (Fig 5). This observation is particularly interesting in light of the described effect of CgPdr1 in contributing for C. glabrata adhesion to epithelial cells (Vale-Silva et al.2013). Surprisingly, the overlap between the genes regulated by the CgPdr1 GOF mutants K274N and K274Q was very limited (Fig. 5) demonstrating that even polymorphisms in the same CgPdr1 residue have a very different impact on the control of gene expression. One of the mechanisms that has been hypothesised to explain this divergence in the set of genes regulated by different CgPdr1 mutants is that they might be differently activated thereby resulting in a different interaction with the transcriptional machinery (Paul, Schmidt and Moye-Rowley 2011). The different genetic background of the strains used in the different transcriptomic profilings may also contribute for some of the observed divergences. Figure 5. View largeDownload slide Venn diagram comparing the set of genes regulated by the CgPdr1 GOF mutants K274Q, K274N, P927L and L946S, as revealed by transcriptomic analyses. The set of genes herein identified as being upregulated in the FFUL887 isolate and harbouring in their promoter region a PDRE motif (TCCRYGSR) was compared with the set of genes previously described to be under the regulation of K274N, P927L and L946S CgPdr1 GOF mutations (Vermitsky et al.2006; Tsai et al.2010). Figure 5. View largeDownload slide Venn diagram comparing the set of genes regulated by the CgPdr1 GOF mutants K274Q, K274N, P927L and L946S, as revealed by transcriptomic analyses. The set of genes herein identified as being upregulated in the FFUL887 isolate and harbouring in their promoter region a PDRE motif (TCCRYGSR) was compared with the set of genes previously described to be under the regulation of K274N, P927L and L946S CgPdr1 GOF mutations (Vermitsky et al.2006; Tsai et al.2010). In addition to contributing to maximal resistance to voriconazole and fluconazole, we also showed that the CgPDR1K274Q allele is detrimental for growth of FFUL887 cells when cultivated in the presence of organic acids at a low pH. Similarly, cells expressing a CgPdr1 P927L GOF allele were also found to be susceptible to organic acids at a low pH (Vermitsky et al.2006). On the background of the KUE100 strain (derived from CBS138) which encodes a wild-type CgPdr1 allele, this phenotype towards organic acids was not observed indicating that it could be a feature of CgPdr1 GOF mutants, or at least of a subset of them. It is not possible with the data available until so far to clarify the reasons why the presence of organic acids seems to sensitise FFUL887 cells, although this is certainly a feature that deserves further exploration as it could be used to improve treatment of infections caused by isolates harbouring CgPdr1 GOF alleles. Besides contributing to better understand the acquisition of azole resistance, the comparative analyses of the genome and transcriptome of the CBS138 and FFUL887 strains also had the potential to elucidate some aspects underlying C. glabrata colonisation of the human urinary tract (the site where FFUL887 was retrieved from). In this sense, one of the observations that emerged from the comparative genomic analysis performed was the identification of several genes encoding adhesins as among those that had the higher number of non-synonymous SNPs in the FFUL887 strain. Adhesion is a fundamental step for C. glabrata ability to successfully colonise infection sites and as such adhesin-encoding genes are subjected to a tight selective pressure demonstrated to occur both at the transcriptional and genomic level (Halliwell et al.2012; Carreté et al.2017; Vale-Silva et al.2017). The microarray analysis performed revealed only two adhesin-encoding genes, CAGL0K10164g and CAGL0H08844g, as being upregulated in the FFUL887 strain; however, this analysis was performed using planktonic cells and thereby adhesion was not being favoured. The herein observed prominent differences in the primary sequence of the adhesins encoded by FFUL887 and CBS138, with emphasis for CgPwp4 which was the protein that differed the most in the two strains (Fig. 2), shows that these genes are also subjected to a strong selective pressure probably to select those variants contributing more to improve adherence to the available surfaces. Another observation that emerged from the comparative transcriptomic analysis performed was the significant upregulation in the FFUL887 strain of genes involved in metabolism of amino acids and of sulphur as well as a large number of genes involved in metabolisation of fatty acids, glycogen and other carbon sources. Since the RPMI growth medium where the strains were cultivated for the transcriptomic analysis contains glucose (20 g/L) and sulphate (50 mg/L), the higher expression of these genes in the FFUL887 strain is more likely to reflect a higher basal level of expression compared to the one observed in the CBS138 strain. In C. glabrata, the presence of glucose in the medium does not appear to repress metabolisation of other carbon sources nor of the genes involved in those processes, as observed in S. cerevisiae (Bernardo et al.2017; Cunha et al.2017). A similar alleviation of glucose repression over metabolism of alternative carbon sources was observed in C. albicans (Childers et al.2016). We have searched the genome sequence of the CBS138 and of the FFUL887 strains for genes homologous to those that mediate glucose repression in S. cerevisiae (Table S4, Supporting Information). With the exception of the ScMTH1 gene, all the other genes mediating glucose repression in S. cerevisiae have robust homologues in CBS138 and in FFUL887; however, in some cases, there were marked differences between the S. cerevisae and the C. glabrata proteins, those more prominent being the Mig1, Mig2 and Mig3 transcriptional regulators (Table S4). Further studies are required to understand the molecular players underlying the alleviation of glucose repression in C. glabrata and how their activity is modified by selective pressure during colonisation. The increased resilience of C. glabrata to antifungal therapy and the persistent emergence of resistant strains is highly problematic considering the high rates of morbidity and mortality associated with infections caused by this pathogenic yeast. The results presented in this study provide a further contribution for a better understanding of the key players contributing for the acquisition of resistance in the host, with special emphasis on CgPdr1 transcription factor, a knowledge that can be used to guide the development of more efficient therapeutical strategies. Specifically, it was shown for the first time that the K274Q substitution results in a GOF CgPdr1 mutant and, consequently, in enhanced azole resistance. An observation of remark from our study and others was that the expression of the K274Q and of P927L CgPdr1 allele increases susceptibility to organic acids at a low pH, suggesting that these molecules could be used to sensitise azole-resistant strains dependent on CgPdr1 GOF alleles. The herein reported whole-genome analysis of the FFUL887 strain and subsequent comparison with CBS138 strain reinforced the extreme genetic diversity among C. glabrata strains providing clues into the adaptive responses evolved during colonisation of the human host and advancing current knowledge on the biology and physiology of this yeast species. SUPPLEMENTARY DATA Supplementary data are available at FEMSYR online. FUNDING Funding received by iBB-Institute for Bioengineering and Biosciences from FCT-Portuguese Foundation for Science and Technology (UID/BIO/04565/2013) and from Programa Operacional Regional de Lisboa 2020 (Project N. 007317) is acknowledged. GB acknowledges Ireland science foundation (grant number 12IA1343) for financial support. Conflict of interest. None declared. REFERENCES Bernardo RT, Cunha DV, Wang C et al.   The CgHaa1-regulon mediates response and tolerance to acetic acid stress in the human pathogen Candida glabrata. G3  2017; 7: 1– 18. Google Scholar CrossRef Search ADS PubMed  Borst A, Raimer MT, Warnock DW et al.   Rapid acquisition of stable azole resistance by Candida glabrata isolates obtained before the clinical introduction of fluconazole. Antimicrob Agents Ch  2005; 49: 783– 7. Google Scholar CrossRef Search ADS   Carreté L, Ksiezopolska E, Pequeroles C et al.   Patterns of genomic variation in the opportunistic pathogen Candida glabrata suggest the existence of mating and a secondary association to the human host. 2017. Caudle KE, Barker KS, Wiederhold NP et al.   Genomewide expression profile analysis of the Candida glabrata Pdr1 regulon. Eukaryot Cell  2011; 10: 373– 83. Google Scholar CrossRef Search ADS PubMed  Childers DS, Raziunaite I, Mol Avelar G et al.   The rewiring of ubiquitination targets in a pathogenic yeast promotes metabolic flexibility, host colonization and virulence. PLoS Pathog  2016; 12: e1005566. Google Scholar CrossRef Search ADS PubMed  Costa C, Dias PJ, Sa-Correia I et al.   MFS multidrug transporters in pathogenic fungi: do they have real clinical impact? Front Physiol  2014; 5: 197. Google Scholar PubMed  Cunha DV, Salazar SB, Lopes MM et al.   Mechanistic insights underlying tolerance to acetic acid stress in vaginal Candida glabrata clinical isolates. Front Microbiol  2017; 8: 1– 13. Google Scholar CrossRef Search ADS PubMed  Dujon B, Sherman D, Fischer G et al.   Genome evolution in yeasts. Nature  2004; 430: 35– 44. Google Scholar CrossRef Search ADS PubMed  Ferrari S, Ischer F, Calabrese D et al.   Gain of function mutations in CgPDR1 of Candida glabrata not only mediate antifungal resistance but also enhance virulence. PLoS Pathog  2009; 5: e1000268. Google Scholar CrossRef Search ADS PubMed  Ferrari S, Sanguinetti M, Torelli R et al.   Contribution of CgPDR1-regulated genes in enhanced virulence of azole-resistant Candida glabrata. PLoS One  2011; 6: e17589. Google Scholar CrossRef Search ADS PubMed  Fidel PL Jr, Vazquez JA, Sobel JD. Candida glabrata: review of epidemiology, pathogenesis, and clinical disease with comparison to C. albicans. Clin Microbiol Rev  1999; 12: 80– 96. Google Scholar PubMed  Garcia-Effron G, Lee S, Park S et al.   Effect of Candida glabrata FKS1 and FKS2 mutations on echinocandin sensitivity and kinetics of 1,3-beta-D-glucan synthase: implication for the existing susceptibility breakpoint. Antimicrob Agents Ch  2009; 53: 3690– 9. Google Scholar CrossRef Search ADS   Gregori C, Schuller C, Roetzer A et al.   The high-osmolarity glycerol response pathway in the human fungal pathogen Candida glabrata strain ATCC 2001 lacks a signaling branch that operates in baker's yeast. Eukaryot Cell  2007; 6: 1635– 45. Google Scholar CrossRef Search ADS PubMed  Halliwell SC, Smith MC, Muston P et al.   Heterogeneous expression of the virulence-related adhesin Epa1 between individual cells and strains of the pathogen Candida glabrata. Eukaryot Cell  2012; 11: 141– 50. Google Scholar CrossRef Search ADS PubMed  Havelsrud OE, Gaustad P. Draft genome sequences of Candida glabrata isolates 1A, 1B, 2A, 2B, 3A, and 3B. Genome Announc  2017; 5: e00328-16. Google Scholar CrossRef Search ADS PubMed  Healey KR, Zhao Y, Perez WB et al.   Prevalent mutator genotype identified in fungal pathogen Candida glabrata promotes multi-drug resistance. Nat Commun  2016; 7: 11128. Google Scholar CrossRef Search ADS PubMed  Khatib R, Johnson LB, Fakih MG et al.   Current trends in candidemia and species distribution among adults: Candida glabrata surpasses C. albicans in diabetic patients and abdominal sources. Mycoses  2016; 59: 781– 6. Google Scholar CrossRef Search ADS   Kolaczkowska A, Kolaczkowski M. Drug resistance mechanisms and their regulation in non-albicans Candida species. J Antimicrob Chemoth  2016; 71: 1438– 50. Google Scholar CrossRef Search ADS   Kolaczkowska A, Kolaczkowski M, Delahodde A et al.   Functional dissection of Pdr1p, a regulator of multidrug resistance in Saccharomyces cerevisiae. Mol Genet Genomics  2002; 267: 96– 106. Google Scholar CrossRef Search ADS PubMed  Monteiro PT, Pais P, Costa C et al.   The PathoYeastract database: an information system for the analysis of gene and genomic transcription regulation in pathogenic yeasts. Nucleic Acids Res  2017; 45: D597– 603. Google Scholar CrossRef Search ADS PubMed  Paul S, Schmidt JA, Moye-Rowley WS. Regulation of the CgPdr1 transcription factor from the pathogen Candida glabrata. Eukaryot Cell  2011; 10: 187– 97. Google Scholar CrossRef Search ADS PubMed  Pfaller MA, Andes D, Arendrup MC et al.   Clinical breakpoints for voriconazole and Candida spp. revisited: review of microbiologic, molecular, pharmacodynamic, and clinical data as they pertain to the development of species-specific interpretive criteria. Diagn Micr Infec Dis  2011; 70: 330– 43. Google Scholar CrossRef Search ADS   Pfaller MA, Diekema DJ, Gibbs DL et al.   Results from the ARTEMIS DISK Global Antifungal Surveillance Study, 1997 to 2007:a 10.5-year analysis of susceptibilities of Candida Species to fluconazole and voriconazole as determined by CLSI standardized disk diffusion. J Clin Microbiol  2010; 48: 1366– 77. Google Scholar CrossRef Search ADS PubMed  Pfaller MA, Messer SA, Woosley LN et al.   Echinocandin and triazole antifungal susceptibility profiles for clinical opportunistic yeast and mold isolates collected from 2010 to 2011:application of new CLSI clinical breakpoints and epidemiological cutoff values for characterization of geographic and temporal trends of antifungal resistance. J Clin Microbiol  2013; 51: 2571– 81. Google Scholar CrossRef Search ADS PubMed  Rossignol T, Logue ME, Reynolds K et al.   Transcriptional response of Candida parapsilosis following exposure to farnesol. Antimicrob Agents Ch  2007; 51: 2304– 12. Google Scholar CrossRef Search ADS   Sanglard D, Ischer F, Calabrese D et al.   The ATP binding cassette transporter gene CgCDR1 from Candida glabrata is involved in the resistance of clinical isolates to azole antifungal agents. Antimicrob Agents Ch  1999; 43: 2753– 65. Sanguinetti M, Posteraro B, Fiori B et al.   Mechanisms of azole resistance in clinical isolates of Candida glabrata collected during a hospital survey of antifungal resistance. Antimicrob Agents Ch  2005; 49: 668– 79. Google Scholar CrossRef Search ADS   Schwarzmuller T, Ma B, Hiller E et al.   Systematic phenotyping of a large-scale Candida glabrata deletion collection reveals novel antifungal tolerance genes. PLoS Pathog  2014; 10: e1004211. Google Scholar CrossRef Search ADS PubMed  Subcommittee on Antifungal Susceptibility Testing (AFST) of the ESCMID European Committee for Antimicrobial Susceptibility Testing (EUCAST) Rodriguez-Tudela JL, Barchiest F, Bille J et al.   EUCAST discussion documente E.Dis 7.1: Method for the determination of minimum inhibitory concentration (MIC) by broth dilution of fermentative yeasts. Clin Microbiol Infect  2003; 9: 1– 8. Google Scholar CrossRef Search ADS PubMed  Torelli R, Posteraro B, Ferrari S et al.   The ATP-binding cassette transporter-encoding gene CgSNQ2 is contributing to the CgPDR1-dependent azole resistance of Candida glabrata. Mol Microbiol  2008; 68: 186– 201. Google Scholar CrossRef Search ADS PubMed  Tsai HF, Krol AA, Sarti KE et al.   Candida glabrata PDR1, a transcriptional regulator of a pleiotropic drug resistance network, mediates azole resistance in clinical isolates and petite mutants. Antimicrob Agents Ch  2006; 50: 1384– 92. Google Scholar CrossRef Search ADS   Tsai HF, Sammons LR, Zhang X et al.   Microarray and molecular analyses of the azole resistance mechanism in Candida glabrata oropharyngeal isolates. Antimicrob Agents Ch  2010; 54: 3308– 17. Google Scholar CrossRef Search ADS   Ueno K, Matsumoto Y, Uno J et al.   Intestinal resident yeast Candida glabrata requires Cyb2p-mediated lactate assimilation to adapt in mouse intestine. PLoS One  2011; 6: e24759. Google Scholar CrossRef Search ADS PubMed  Vale-Silva L, Beaudoing E, Tran VDT et al.   Comparative genomics of two sequential Candida glabrata clinical isolates. G3  2017; 7: 2413– 26. Google Scholar CrossRef Search ADS PubMed  Vale-Silva L, Ischer F, Leibundgut-Landmann S et al.   Gain-of-function mutations in PDR1, a regulator of antifungal drug resistance in Candida glabrata, control adherence to host cells. Infect Immun  2013; 81: 1709– 20. Google Scholar CrossRef Search ADS PubMed  Vermitsky JP, Earhart KD, Smith WL et al.   Pdr1 regulates multidrug resistance in Candida glabrata: gene disruption and genome-wide expression studies. Mol Microbiol  2006; 61: 704– 22. Google Scholar CrossRef Search ADS PubMed  Vermitsky JP, Edlind TD. Azole resistance in Candida glabrata: coordinate upregulation of multidrug transporters and evidence for a Pdr1-like transcription factor. Antimicrob Agents Ch  2004; 48: 3773– 81. Google Scholar CrossRef Search ADS   Xu N, Ye C, Chen X et al.   Genome sequencing of the pyruvate-producing strain Candida glabrata CCTCC M202019 and genomic comparison with strain CBS138. Sci Rep  2016; 6: 34893. Google Scholar CrossRef Search ADS PubMed  © FEMS 2017. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

FEMS Yeast ResearchOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off