Fluconazole resistance and CDR1 expression in Candida albicans mediated by the hyperactive Tac1-5 transcriptional activator requires Tlo proteins

Abstract

Candida albicans is an opportunistic fungal pathogen associated with superficial and systemic infections in humans. Azole antifungal resistance in C. albicans is of clinical concern, and both oral and systemic Candida infections can be difficult to treat due to the lack of alternative antifungal drugs. Expression of a hyperactive form of the transcription factor Tac1 is a major contributor to azole resistance in C. albicans isolates resulting in the increased expression of the azole efflux pump Cdr1. In this study, we investigated whether the Mediator tail component Med2, encoded by the expanded (n=14) TLO gene family of C. albicans, was required for Tac1 activity. A homozygous TAC1-5 gain-of-function point mutation was introduced into WT, tloΔ and med3Δ strains of C. albicans which enables them to express hyperactive Tac1. qRT-PCR analysis revealed that tloΔ-TAC1-5 had reduced basal and fluphenazine-induced CDR1 expression relative to WT-TAC1-5 strains and exhibited reduced levels of resistance to fluconazole and terbinafine. Individual copies of representatives from each of the alpha, beta and gamma TLO clades were reintroduced into tloΔ-TAC1-5 to investigate their ability to restore Tac1-activated resistance. These studies show that alpha and beta TLO genes could restore fluconazole resistance in the tloΔ-TAC1-5 background, whereas gamma clade genes did not result in any detectable phenotypic complementation. Transcript profiling showed that reintroduction of TLOα1 led to increased expression of TAC1-5-activated genes such as CDR1. Further analysis using ChIP-qPCR revealed that Tloα1 localizes to the drug response element which is the site where Tac1 binds to the CDR1 promoter. These data have identified that the TLO gene family is required for the expression of Tac1-mediated fluconazole resistance. However, this effect is confined to members of the alpha and beta, but not the gamma, TLO clades.

Data Availability

All RNAseq data are available from the NCBI Sequence Read Archive, submission number SUB15226071, under the accession numbers SRX28236589, SRX28236588, SRX28236587, SRX28236586, SRX28236585, SRX28236584, SRX28236583, SRX28236582, SRX28236581, SRX28236580, SRX28236579 and SRX28236578.

Introduction

Candida albicans is an opportunistic pathogenic yeast species and is the major cause of superficial candidiasis and candidemia globally [1,2]. The World Health Organization has recently designated C. albicans as a fungal priority pathogen [3,4]. Surveillance reports suggest that resistance to the commonly used azole antifungal drugs, including fluconazole, continues to be a concern when treating C. albicans infections [5,7]. Multiple factors have been identified that may contribute to the increased incidence of azole-resistant C. albicans, such as the wide and long-term usage and the fungistatic nature of azole drugs [1,2].

The most common mechanisms of azole resistance in C. albicans are the alteration of the drug target enzyme (Erg11), the overexpression of Erg11 and the upregulation of plasma membrane efflux pumps [8]. Most azole-resistant isolates of C. albicans exhibit increased expression of multidrug transporters encoded by either the major facilitator efflux pump CaMDR1 (multidrug resistance 1) or the ATP-binding cassette (ABC) transporters CDR1 (Candida drug resistance) and CDR2 [9]. Furthermore, gain-of-function (GOF) mutations in the transcriptional activator Tac1 are commonly responsible for the overexpression of efflux pumps in these isolates [9,10]. Tac1 is a classical zinc cluster transcription factor that possesses a highly conserved Zn₂Cys₆ motif [11]. Tac1 has three domains, an N-terminal DNA-binding domain (DBD; 1–1,320 nucleotides), a C-terminal acidic activation domain (AAD; 1,354–2,946 nucleotides) and a middle homology region (MHR; 1,321–1,653 nucleotides) [11,12]. The Tac1-DBD binds to the drug response element (DRE) in the promoter regions of CDR1 and CDR2 [11]. Tac1-MHR is known to be a xenobiotic-binding domain that interacts with xenobiotics including fluphenazine, resulting in Tac1 hyperactivity and upregulation of CDR1 and CDR2 expression [11]. The Tac1-AAD was shown to interact with the TATA-binding protein (TBP) that plays a crucial role in transcription [11]. The well-studied TAC1^GOF mutation identified in a C. albicans azole-resistant isolate, known as TAC1-5, is a nonsynonymous substitution of Asn to Asp in Tac1p [9]. It is located within the Tac1-AAD, and it has been suggested that this substitution enhances the interaction between Tac1-AAD and TBP, thus increasing the transcription of CDR1 and CDR2 [12].

The Mediator multi-polypeptide complex is an important component in transcription initiation and has been shown to regulate the transcription of various antifungal resistance-related mechanisms. In C. albicans, Mediator tail subunits Med3 and Med15 have both been shown to be required for Tac1-activated azole resistance and xenobiotic induction of CDR1 expression [13]. In Candida glabrata, the Mediator tail components CgMed15A and CgMed2 are required for the regulation of expression of the ABC transporter-encoding genes CgCDR1 and CgCDR2 (orthologues of C. albicans CDR1/2) via interaction with the transcription factor CgPdr1 [14,15]. The Mediator complex in C. albicans differs from most other species due to the large expansion of genes encoding the Med2 tail component, encoded by a 10–15 member gene family, referred to as the TLO family [16,17]. Phylogenetic analysis divides the gene family into three distinct clades, and previous studies suggest that the TLO genes/clades play different roles in controlling the expression of a wide range of cellular functions, including virulence and tolerance to azole drugs [18,22]. Given the role of the Mediator tail components Med3 and Med15 in Tac1-activated CDR1 expression and azole resistance in C. albicans, we wanted to investigate if members of the expanded TLO gene family (which encode the Med2 Mediator tail component) are required for basal CDR1 expression and Tac1-activated CDR1 expression in C. albicans by introducing the homozygous TAC1-5 allele (N977D substitution) into a tloΔ mutant (in which the entire TLO family had been deleted previously).

Methods

Candida strains and growth conditions

All C. albicans strains used in this study, including CRISPR-Cas9-edited strains, are described in Table S1 (available in the online Supplementary Material). All C. albicans strains were routinely grown on yeast extract peptone dextrose (YEPD) agar at 30 or 37 °C, as indicated, for 24–48 h in a static incubator (Gallenkamp, Leicester, UK). For liquid culture, C. albicans was routinely grown in YEPD broth in a shaking incubator (New Brunswick Scientific, Edison, NJ, USA) set at 200 r.p.m., 30 or 37 °C overnight. Nourseothricin-resistant strains were selected and maintained on YEPD medium containing 200 µg ml⁻¹ of the antibiotic (CloNAT, Werner Bioagents, Germany).

Antifungal susceptibility testing

Fluconazole spot plate assays were performed on YEPD plates with or without fluconazole (10 or 15 µg ml⁻¹). Plates were inoculated with 5 µl spots of serial dilutions of overnight cultures (2×10⁶ to 2×10² cells ml⁻¹). Plates were incubated in a static incubator for 48 h at 30 ˚C. Fluconazole disc diffusion assays and Etests were performed on solid RPMI-1640 medium, prepared from liquid RPMI-1640 medium (Merck) as described by O’Connor-Moneley et al. [21]. Overnight YEPD broth cultures were washed with sterile PBS and resuspended to OD₆₀₀=0.07. Sterile cotton swabs were dipped in the suspension and used to lawn solid RPMI-1640 agar plates. A 25 µg fluconazole susceptibility disc (OXOID), or fluconazole Etest (bioMérieux), was placed at the centre of each RPMI-1640 agar plate and incubated at 30 ˚C for 48 h. The plates were imaged on a ‘FLASH and GO’ plate imager (IUL Instruments) using autoexposure settings.

Design and construction of the CRISPR-Cas9 oligonucleotides

The CRISPR-Cas9 LEUpOUT system designed for C. albicans [23] was used to introduce a homozygous TAC1-5 SNP (an A to G substitution at nucleotide position 2929) into the C. albicans strain MAY1244 (LEU2/leu2Δ) and its tloΔ and med3Δ derivatives. A unique 20 bp gRNA (Table S2) was designed to cut TAC1 within the TAD, and a repair template (94 bp length, generated from overlapping oligonucleotides; Table S2) containing the A to G substitution at nucleotide position 2929 was introduced. The gRNA-expressing cassette was generated by a fusion PCR using oligonucleotides and plasmids previously described [23]. Transformation into C. albicans was performed using the electroporation method [24]. C. albicans transformants with homozygous TAC1-5 mutations were identified via Sanger sequencing following amplification of the region with the primers TAC1ampF and TAC1-5 R (Table S2) yielding a 440 bp DNA product. The recycling of the Cas9 cassette was carried out on the same day as PCR streaking colonies on YNB plates without amino acids or ammonium sulphate (leucine-negative media) and incubation at 30 °C for 2–3 days. Colonies that grew on plates without leucine and exhibited nourseothricin sensitivity were considered to have excised the CRISPR-Cas9 components from the LEU2 locus.

RNA extraction and qRT-PCR analysis

YEPD broth cultures were inoculated from overnight cultures to an OD₆₀₀=0.1 and incubated at 37 °C with 200 r.p.m. shaking until an OD₆₀₀=0.8 was reached. RNA was then extracted from cells using the RNeasy extraction kit (Qiagen) as per the manufacturer’s instructions. mRNA sequencing was performed with strand-specific libraries and sequenced on the Illumina NovaSeq 6000 Sequencing System using paired-end 150 bp reads. Each experiment generated a minimum of >20 million read pairs per sample with a Q30 score of ≥85%. Raw reads were aligned to the C. albicans SC5314 Assembly 21 genome (downloaded from CGD) in the Strand NGS 4.0 software package using the default settings. Reads were quantified and normalized in Strand NGS using DeSeq2 [25], and statistical analysis of differential expression was carried out with post hoc Benjamini–Hochberg testing (FDR q<0.05). Further analysis on lists of differentially expressed genes was performed via GO analysis on the Candida Genome Database and Gene Set Enrichment Analysis (GSEA) [26]. For GSEA, a custom database of C. albicans genes associated with GO Terms and previously published RNA sequencing experiments was queried (Supplementary data file). Sequence data are available for download from the NCBI Sequence Read Archive, accession PRJNA1245622.

For qRT-PCR analysis, Candida cultures were established as described above in 100 ml YEPD (OD₆₀₀ 0.1) and grown to exponential phase (OD₆₀₀ 0.6–0.8) in a shaking incubator set at 200 r.p.m., 37 °C. An 80 ml volume of mid-log phase C. albicans culture was then split equally into two conical flasks, with one being exposed to fluphenazine at a final concentration of 20 µg ml⁻¹ and the other as a control (no drugs added). A 10 ml volume of each culture was then collected at given time points (0 and 30 min) and was used for RNA extraction as above. The Superscript IV Reverse Transcriptase kit (Invitrogen) was used to convert RNA into cDNA. For qRT-PCR, reactions consisted of 7.5 µl SYBR Green, 0.375 µl of each forward and reverse primer (10 mM) and 5.75 µl of molecular-grade water. A 14 µl volume of this mixture was combined with 1 µl cDNA (diluted 1 : 10 in water) as the template. All qRT-PCR primer sequences are listed in Table S2. qRT-PCR reactions were performed in an AB7500 Real Time PCR machine (Applied Biosystems) with each biological replicate split into three technical replicates. The ACT1 housekeeping gene was used as an endogenous control to determine relative expression using the comparative C_T method and graphically represented using GraphPrism v10 (San Diego, CA, USA).

Reintroduction of TLO genes using pNIM1

Individual HA-tagged alpha, beta and gamma TLO genes were reintroduced into the ADH1 locus of the tloΔ-TAC1-5 mutants using the pNIM1-TLO cassettes developed by Fletcher et al. [19]. The plasmid was linearized for transformation purposes using KpnI and SacII restriction enzymes (New England Biolabs). Integration of the pNIM1 cassette containing the specific TLO gene was confirmed using the ADH1 locus-specific primer AAK29 and the pNIM1-specific primer AAK31. The presence of the reintroduced TLO gene was confirmed with the TLO primer ‘Pan-Tlo’ and the HA-tag reverse primer, which yielded products of different size for each TLO gene visualized by gel electrophoresis (Table S2). The expression of each reintroduced TLO gene in the tloΔ-TAC1-5 background was measured by qRT-PCR using gene-specific primers listed in Table S2. Further verification was performed using Western blotting against HA-protein to visualize HA-tagged Tlo proteins.

ChIP-qPCR

C. albicans cultures were grown with shaking at 200 r.p.m. and 37 °C in 50 ml YEPD. After reaching OD₆₀₀ 1.0, the cells were washed in PBS and resuspended in 36 ml PBS and 1 ml of 37% (v/v) formaldehyde. Cultures were incubated at 30 °C with shaking at 200 r.p.m. for 30 min, with subsequent quenching with 2 ml sterile 2.5 M glycine for 10 min. Cells were then pelleted and flash-frozen in liquid nitrogen, followed by storage at −80 °C for future use. The pellet containing formaldehyde-fixed cells was defrosted on ice, and 400 µl formaldehyde-assisted (FA) lysis buffer (50 mM HEPES pH 7.5, 140 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate) was added followed by 4 µl PI (1 : 100), 3.2 µl PMSF (2 mM) and an equal volume of zirconia beads (0.5 mm in size, ~500 µl). The resuspended mixture was processed for seven cycles (30 s each) using a FastPrep Homogeniser 24 (4 m/s) and chilled for 1 min on ice between each cycle. The lysate was mixed with 1 ml FA lysis buffer and spun at low speed (8,000 r.p.m.) for 30 s to avoid pellet formation. An 80 µl ‘pre-sonication’ sample was removed after this centrifugation. The remaining centrifuged lysate was gently resuspended and sonicated 14 times (10 s pulses at 8 microns, Sanyo Soniprep 150) with 1 min on ice between each sonication. After sonication, samples were centrifuged at 13,200 r.p.m. for 30 min at 4 °C. An 80 µl aliquot of supernatant was processed for DNA fragment size verification, and the remaining supernatant was then divided into 100 µl aliquots and stored at −80 °C for future use. DNA fragment size was verified by agarose gel electrophoresis.

Prior to analysis, the samples of sonicated lysate were thawed and brought to a volume of 500 µl with FA lysis buffer containing protease inhibitors (PIs) (15 ml FA lysis buffer+150 µl 100X PI, made fresh). The solution was mixed, and 20 µl of lysate was transferred to a PCR tube as the input (IN) sample (Input). Pronase treatment of IN sample was then carried out with the addition of 100 µl ChIP elution buffer, 60 µl TE (pH7.5), 20 µl Pronase (20 mg ml⁻¹) and 1 µl 1 M CaCl₂ with incubation in a thermocycler (42 °C for 2 h, 65 °C for 6 h) before storage at −20 °C. To the remaining 480 µl lysate, 2 µl anti-HA antibody (Abcam ab9110) was added and placed on a rotating wheel at 4 °C overnight at 16 r.p.m. Antibody–DNA complexes were purified using a mix containing 20 µl Protein A and 20 µl Protein G beads (Invitrogen Dynabeads) according to the manufacturer’s instructions. Purified bead supernatants (250 µl) from the IP pulldown samples were subjected to treatment with 50 µl Pronase (20 mg ml⁻¹) and 1.5 µl 1M CaCl₂ with incubation in a thermocycler (42 °C for 2 h, 65 °C for 6 h).

All IN samples and IP pulldown samples were purified with the Qiagen PCR purification kit. qPCR was carried out using the GoTaq^® qPCR kit from Promega using IP pulldown samples (diluted 1 in 5) and IN samples (diluted 1 in 50) as DNA template. A serial tenfold dilution (neat to 10⁻⁴) of DNA (200 ng µl⁻¹) isolated from C. albicans MAY1244-TAC1-5 was used to generate a DNA standard curve. A 0.12 µl volume of primer mix containing forward and reverse primer (50 µM), 2× SYBR master mix and 9.88 µl water was mixed with 2 µl diluted DNA for each GoTaq qRT-PCR reaction. Reactions for each biological replicate experiment were run in technical triplicate, for 40 cycles of 95 °C for 15 s and 60 °C for 1 min. Target enrichment in the IP sample (IP/IN) was normalized to the IP/IN signal detected at an internal negative control site (CDR1 Up, a region ∼1.4 kb upstream of the CDR1 ORF) and expressed as ‘Relative occupancy’.

Results

Introduction of TAC1-5 hyperactive mutation in C. albicans Mediator tail mutants

CRISPR-Cas9-mediated mutagenesis was used to introduce a homozygous TAC1-5 point mutation in a tloΔ mutant in which the entire TLO gene family (n=14) had previously been deleted [19], and for comparative purposes, in a med3Δ mutant in which the (single-copy) gene encoding the Med3 mediator tail component had been deleted. An identical mutation was also introduced in the parental WT C. albicans strain MAY1244 to generate strain WT-TAC1-5. Confirmation of the single base pair substitution from A to G at position 2929 bp in both alleles of the TAC1 gene was confirmed by Sanger sequencing. At the cellular level, the TAC1-5 derivatives of tloΔ (tloΔ-TAC1-5) and med3Δ (med3Δ-TAC1-5) were indistinguishable from the parental mutants, exhibiting a pseudohyphal morphology in YEPD medium and exhibiting an inability to form true hyphae in 10% serum. The WT-TAC1-5 derivative was morphologically identical to its parental WT isolate (i.e. it grows as yeast in YEPD and forms hyphae in 10% serum).

As expected, the introduction of the hyperactive Tac1 mutation into the WT parental strain (WT-TAC1-5) resulted in resistance to fluconazole (Fig. 1a), with the mutant being able to grow on YEPD spot plates supplemented with 10 µg ml⁻¹ and 15 µg ml⁻¹ fluconazole. Introduction of the mutation into the tloΔ and med3Δ mutants also resulted in reduced susceptibility to fluconazole; however, these strains only grew on plates with 10 µg ml⁻¹ but not 15 µg ml⁻¹ fluconazole (Fig. 1a). Fluconazole susceptibility was further investigated using a disc diffusion assay (Fig. 1b) in the presence and absence of cyclosporine A (1 µg ml⁻¹), to suppress fluconazole tolerance. The introduction of the TAC1-5 allele conferred increased fluconazole resistance to both WT and tloΔ mutants as they both exhibited smaller zones of inhibition compared to their parental strains (Fig. 1b and Table S3). The WT-TAC1-5 strain was less susceptible to fluconazole compared to the tloΔ-TAC1-5 strain, which exhibited a smaller zone of inhibition (3.2 mm reduction, Table S3). As expected from our previous studies, in the absence of cyclosporine, the tloΔ mutant exhibited a high degree of fluconazole tolerance indicated by growth within the zone of inhibition (Fig. 1b). Interestingly, even though the TAC1-5 mutation conferred increased fluconazole resistance in the tloΔ background, it was found to slightly reduce the fluconazole tolerance of tloΔ, indicated by the clear zone of inhibition (Fig. 1b). Fluconazole Etest assays were also carried out to obtain precise fluconazole MIC values. This analysis showed that the three strains without the TAC1-5 mutation (MAY1244, tloΔ and med3Δ mutants) expressed similar MIC values between 1 and 2 µg ml⁻¹ (Table 1). Confirming the spot plate and disc diffusion assays, WT-TAC1-5 showed the highest fluconazole MIC value (16 µg ml⁻¹), while tloΔ-TAC1-5 and med3Δ-TAC1-5 strains exhibited a twofold lower MIC (8 µg ml⁻¹).

Fig. 1. Fluconazole susceptibility of C. albicans WT (MAY1244), tloΔ and med3Δ null mutants and derivatives expressing the TAC1-5 GOF mutation. (a) Fluconazole spot plate assays were performed by spotting 5 µl of serial dilutions of overnight cultures (2×10⁶ to 2×10² cells ml⁻¹) on YEPD plates with or without fluconazole (10 or 15 µg ml⁻¹) and incubated for 48 h at 30 °C. (b) Disc diffusion assays on RPMI-1640 plates containing fluconazole (25 µg ml⁻¹) with and without cyclosporine (1 µg ml⁻¹) at 30 °C for 48 h. (c) Terbinafine spot plate assays were carried out as in (a) on YEPD plates containing terbinafine (2.5 and 5 µg ml⁻¹).

Table 1. Morphological characteristics and fluconazole Etest assay MIC values of C. albicans strains and derivatives expressing the TAC1-5 mutation.

Strains		Cellular morphology in YEPD (30 °C)	Morphology in 10% serum	MIC values (Etest)
WT (MAY1244)		Yeast morphology	True hyphae	1.5
WT-TAC1-5		Yeast morphology	True hyphae	16
med3Δ		Pseudohyphae	Pseudohyphae	1
med3Δ-TAC1-5		Pseudohyphae	Pseudohyphae	8
tloΔ		Pseudohyphae	Pseudohyphae	1.5
tloΔ-TAC1-5		Pseudohyphae	Pseudohyphae	8
tloΔ-TAC1-5+TLOα1	Alpha clade	Yeast morphology	True hyphae	16
tloΔ-TAC1-5+TLOα3	Alpha clade	Yeast morphology	True hyphae	12
tloΔ-TAC1-5+TLOα34	Alpha clade	Yeast morphology	True hyphae	16
tloΔ-TAC1-5+TLOβ2	Beta clade	Pseudohyphae	Pseudohyphae	6
tloΔ-TAC1-5+pENO-TLOβ2	Beta clade	Yeasts/hyphae	True hyphae	16
tloΔ TAC1-5+TLOγ5	Gamma clade	Pseudohyphae	Pseudohyphae	6
tloΔ TAC1-5+TLOγ7	Gamma clade	Pseudohyphae	Pseudohyphae/hyphae	8
tloΔ TAC1-5+TLOγ11	Gamma clade	Pseudohyphae	Pseudohyphae	8

Sensitivity to the antifungal agent terbinafine, which is also a substrate of the Cdr1 efflux pump, was also assessed in the C. albicans TAC1-5 mutants on YEPD plates containing 2.5 µg ml⁻¹ and 5 µg ml⁻¹ terbinafine (Fig. 1c). Both MAY1244 and its derivative WT-TAC1-5 were capable of growth on 5 µg ml⁻¹ terbinafine plates. Surprisingly, although the Δmed3 mutant was able to grow in the presence of 2.5 µg ml⁻¹ terbinafine, the Δtlo mutant was not. However, the introduction of the TAC1-5 mutation resulted in similar terbinafine resistance profiles as they both exhibited strong growth on the 2.5 µg ml⁻¹ terbinafine, but not on 5 µg ml⁻¹ terbinafine (Fig. 1c).

Expression of CDR1 and CDR2 in Mediator tail and TAC1-5 mutants

In order to investigate the expression of CDR1 and CDR2, the WT C. albicans and Mediator tail mutant strains were grown in the absence and presence of the xenobiotic fluphenazine for 30 min (Fig. 2a). qRT-PCR revealed that both tloΔ and med3Δ mutants exhibit significantly lower basal CDR1 expression and lower basal CDR2 expression compared to WT C. albicans. Exposure to fluphenazine for 30 min resulted in high-level induction of CDR1 and CDR2 mRNA in WT C. albicans but only weakly induced the expression of CDR1 and CDR2 mRNA in tloΔ and med3Δ mutants. As per previous studies, CDR1 expression was significantly higher than that of CDR2 in all strains tested [9,10].

Fig. 2. CDR1 and CDR2 mRNA expression in C. albicans WT and Mediator tail null mutants. (a) CDR1 expression measured via qRT-PCR in YEPD alone (−FNZ) or in the presence of fluphenazine (+FNZ, 20 µg ml⁻¹) for 30 min. (b) CDR1 expression in TAC1-5 derivatives of WT MAY1244, tloΔ and med3Δ. The expression of CDR1 was normalized relative to the expression of ACT1 and was expressed as a log₂ fold difference. Error bars were calculated based on three biological replicates, and ordinary one-way ANOVA was carried out to detect any significant differences between strains (* indicates P value<0.05, ** indicates P value<0.01 and **** indicates P value<0.0001).

As expected, the presence of the homozygous TAC1-5 mutation significantly increased the expression of CDR1 in the WT (~4-fold log2; P<0.01) strain (Fig. 2b). However, no significant increase in CDR1 expression was observed in either of the Mediator tail mutants (tloΔ and med3Δ; Fig. 2b).

Reintroduction of TLOα1, TLOβ2 and TLOγ11 in tloΔ-TAC1-5

In order to investigate if individual TLO genes can complement the drug response phenotypes of the Δtlo mutant, we reintroduced representatives from each of the three TLO clades (TLOα1, TLOβ2 and TLOγ11) into the tloΔ-TAC1-5 mutant under the control of the pTET promoter. Expression levels of the pTET-regulated TLOα1 mRNA and the Tloα1 protein in the tloΔ-TAC1-5 strain in doxycycline-free medium were similar to that previously described by Fletcher et al. [19] when the same construct was expressed in the parental tloΔ background. Phenotypically, reintroduction of TLOα1 also reversed the pseudohyphal morphology of the tloΔ-TAC1-5 strain and restored true hypha formation in 10% serum (Fig. S1). Expression of TLOβ2 mRNA from the same pTET promoter in tloΔ-TAC1-5 was significantly lower than TLOα1 and also lower than the previously reported levels of expression of TLOβ2 when the same construct was expressed in the parental tloΔ background [19] (Fig. 3). Western blotting could not detect expression of HA-tagged Tloβ2 protein in the tloΔ-TAC1-5 mutant, even following doxycycline induction, and the construct could not complement the cell morphology or hyphal growth defects in the tloΔ-TAC1-5 background (Fig. S1). Only when TLOβ2 was reintroduced under the strong, constitutive enolase promoter (pENO) were we able to detect Tloβ2 and observe phenotypic complementation of the cell morphology and hyphal growth defects of the mutant strain, indicating functional Tloβ2 expression (Fig. S1).

Fig. 3. qRT-PCR and Western blot analysis to detect HA-tagged Tlo expression in tloΔ and tloΔTAC1-5 mutants expressing individual TLOs (TLOα1, TLOβ2 and TLOγ11). (a) TLO gene expression was measured by qRT-PCR in mid-exponential YEPD broth cultures. The expression of TLO genes was normalized relative to the expression of ACT1 and was expressed as log₂ fold difference. (b) Crude protein was extracted from exponential phase tloΔ and tloΔ-TAC1-5 mutants expressing individual TLOs (TLOα1, TLOβ2 or TLOγ11) grown in YEPD media at 37 ˚C. Anti-HA antibody (ab9110, Abcam) was used to detect HA-tagged Tlo proteins.

Expression levels of pTET-TLOγ11 mRNA in the tloΔ-TAC1-5 background were similar to that described by Fletcher et al. [19]. Despite the presence of detectable levels of mRNA, we could not detect expression of the Tloγ11 protein in the tloΔ-TAC1-5 background, which concurs with the findings of Fletcher et al. [19] who were unable to detect the Tloγ11 protein, even when expressed from the strong pENO promoter (Fig. 3b).

Antifungal susceptibility of tloΔ-TAC1-5 mutants complemented with TLOα1, TLOβ2 or TLOγ11

We next investigated the effect of reintroducing the TLO genes on fluconazole susceptibility in the tloΔ and tloΔ-TAC1-5 backgrounds. Reintroduction of TLOα1, TLOβ2 or TLOγ11 to the tloΔ mutant without the TAC1-5 allele did not impact significantly the fluconazole MIC (all 1 µg ml⁻¹) and did not enable growth on 15 µg ml⁻¹ fluconazole spot plates (Fig. 4a). In the tloΔ-TAC1-5 background, introduction of pTET-TLOα1 restored strong growth in the presence of 15 µg ml⁻¹ fluconazole, whereas the weakly expressed pTET-TLOβ2 and pTET-TLOγ11 genes exhibited limited growth (Fig. 4a). As previously noted, expression of TLOβ2 from the strong pENO promoter resulted in phenotypic complementation of the morphological defects of the tloΔ-TAC1-5 mutant. Analysis of fluconazole susceptibility showed that the pENO-TLOβ2 gene could restore growth on 15 µg ml⁻¹ fluconazole in the tloΔ-TAC1-5 mutant to a similar level as pTET-TLOα1 (Fig. 4b).

Fig. 4. Fluconazole and terbinafine susceptibility of tloΔ and tloΔ-TAC1-5 mutants complemented with TLOα1, TLOβ2 or TLOγ11. (a) Spot plate assays were inoculated with serial dilutions of overnight cultures (2×10⁶ to 2×10² cells ml⁻¹) on YEPD plates with or without fluconazole or terbinafine. Plates were incubated in a static incubator for 48 h at 30 °C. (b) Spot plate assays comparing the susceptibility of tloΔ-TAC1-5 derivatives expressing either pTET-TLOβ2 or pENO-TLOβ2. (c) Disc diffusion assays on RPMI-1640 plates containing fluconazole (25 µg ml⁻¹) with and without cyclosporine (1 µg ml⁻¹) incubated at 30 °C for 48 h.

Disc diffusion assays confirmed the results of the fluconazole spot plate experiments. In the presence of cyclosporine, the tloΔ-TAC1-5 mutant expressing TLOα1 exhibited the highest level of fluconazole resistance with the smallest inhibition zone, while the weakly expressed pTET-TLOβ2 and -TLOγ11 genes exhibited intermediate inhibition zone sizes (Fig. 4c and Table S4). In the absence of cyclosporine, fluconazole tolerance was still apparent in each of the transformants (Fig. 4c). Fluconazole Etest assays were also carried out to measure fluconazole MIC values for tloΔ-TAC1-5 mutants expressing individual TLOs (Table 1). The tloΔ-TAC1-5 mutant expressing TLOα1 exhibited an MIC value identical to WT-TAC1-5 (16 µg ml⁻¹). The tloΔ-TAC1-5 mutants expressing pTET-TLOβ2 and pTET-TLOγ11 had MIC values of 6–8 µg ml⁻¹ which were similar to the parental tloΔ-TAC1-5 mutant. However, expression of TLOβ2 from the strong pENO1 promoter in the tloΔ-TAC1-5 background resulted in a similar MIC to the pTET-TLOα1 strain of 16 µg ml⁻¹.

Terbinafine susceptibility was also examined using spot plate assays (Fig. 4). Reintroduction of TLOα1 in tloΔ-TAC1-5 significantly enhanced growth of the strain on 5 µg ml⁻¹ terbinafine plates. Restoration of TLOβ2 under the control of the pENO promoter also restored growth on 5 µg ml⁻¹ terbinafine plates to a similar level as TLOα1, whereas TLOγ11 had no impact on terbinafine susceptibility.

We next tested whether additional representatives of the TLO alpha and gamma clades conferred the same phenotypes as TLOα1 and TLOγ11 in the tloΔ-TAC1-5 background. Similar to pTET-TLOα1, the alpha clade genes pTET-TLOα3 and pTET-TLOα34 restored WT cellular morphology and hyphal growth in 10% serum in the tloΔ-TAC1-5 background (Fig S1). TLOα34 was similar to TLOα1 in its ability to restore fluconazole resistance in the tloΔ-TAC1-5 background (Etest MIC 16 µg ml⁻¹ and Fig. S2a). However, pTET-TLOα3-expressing derivatives exhibited weaker growth on fluconazole spot plates (Fig. S2b) and exhibited a slightly lower fluconazole MIC of 12 µg ml⁻¹ in the tloΔ-TAC1-5 background. We also tested two additional TLO gamma clade members, pTET-TLOγ5 and pTET-TLOγ7. Neither gene restored cellular morphology or hyphal growth (Fig. S1) and had no effect on fluconazole resistance (Etest MIC 6 µg ml⁻¹ and Fig. S2). TLOα34 was also able to restore comparable levels of growth as TLOα1 on plates containing 5 µg ml⁻¹ terbinafine, whereas TLOα3 expression resulted in an intermediate phenotype (Fig. S3). The strongly expressed pENO-TLOβ2 also restored growth on 5 µg ml⁻¹ terbinafine, whereas the gamma genes pTET-TLOγ5 and pTET-TLOγ7 did not alter terbinafine susceptibility in the tloΔ-TAC1-5 background (Fig. S3).

Global gene expression in TLO-complemented tloΔ-TAC1-5 strains

In order to understand the impact of the TAC1-5 allele on gene expression and to investigate the functionality of the TLO genes in this background, we carried out RNAseq analysis on mid-exponential YEPD cultures (Fig. 5 and Supplementary data file). We first compared the transcriptomes of the tloΔ mutant and the tloΔ-TAC1-5 derivative to investigate the impact of the TAC1-5 allele on the tloΔ transcriptome. This analysis showed increased expression of CDR1 and CDR2 in the TAC1-5 derivative and also identified increased expression of other Tac1 target genes such as RTA3 and PDR16, as well as the TAC1 gene itself. Unexpectedly, we observed decreased expression of many hypha-specific genes that are normally overexpressed in the tloΔ mutant background, most significantly those regulated by EFG1 including ECE1, HWP1, IHD1 and SOD5 (Fig. 5a). Introduction of TLOα1 further enhanced expression of the Tac1-activated genes including CDR1, CDR2 and RTA3 (Fig. 5b). Comparison of the TAC1-5-activated gene set with the TLOα1-induced gene set showed that 28% of TAC1-5-activated genes (157 of 563 genes) exhibited further increases in expression in the presence of TLOα1 (Fig. S4). As expected, based on our previous studies, TLOα1 also further reduced the expression of EFG1-regulated genes and opaque-phase gene expression and restored glycolytic gene expression to WT levels [21].

Fig. 5. RNAseq analysis of the tloΔ and tloΔ-TAC1-5 mutants and tloΔ-TAC1-5 mutant complemented with TLOα1. (a) The left panel shows a volcano plot of significant (Padj <0.05, log₂FC>1.0) changes in gene expression in tloΔ-TAC1-5 compared to tloΔ. The right panel shows a plot of representative, non-redundant categories of differentially expressed genes identified by GSEA. Positive NES=increased expression; negative NES=decreased expression. Gene set descriptions can be found in the Supplementary data file. (b) Comparison of tloΔ-TAC1-5+TLOα1 versus tloΔ-TAC1-5 as described in (a).

In contrast, the introduction of the weakly expressed pTET-TLOβ2 gene in the tloΔ-TAC1-5 background resulted in very few changes in gene expression, most likely due to the lower level of expression of Tloβ2 protein (Fig. S5). We observed increased expression of TLOβ2 itself (log₂ FC 1.56) and a significant reduction in ADH1 expression, corresponding to the integration locus of the pTET-TLOβ2 construct.

Levels of CDR1 mRNA in the tloΔ-TAC1-5 mutant and TLO-expressing derivatives were confirmed using qRT-PCR (Fig. 6a). The reintroduction of TLOα1 in the tloΔ-TAC1-5 mutant restored TAC1-5-activated CDR1 expression to a similar level as that observed in WT-TAC1-5. The reintroduction of the weakly expressed pTET-TLOγ11 only partially restored the expression of CDR1 activation by TAC1-5 to levels that were significantly lower than those observed in the TLOα1-complemented tloΔ-TAC1-5 and WT-TAC1-5 (Fig. 6a).

Fig. 6. CDR1 mRNA expression and Tlo protein occupancy across the CDR1 promoter region in tloΔ-TAC1-5 mutants complemented with TLOα1 and TLOγ11. (a) CDR1 expression measured via qRT-PCR in TAC1-5 derivatives of WT (MAY1244), tloΔ and med3Δ and tloΔ-TAC1-5 mutants complemented with TLOα1 or TLOγ11. The expression of CDR1 was normalized relative to ACT1 and is expressed as log₂ fold difference. Data represent three biological replicates analysed using one-way ANOVA (***** indicates P value<0.0001). (b) Tlo occupancy at CDR1 promoter sites (CDR1 1–2, CDR1 1–3 and CDR1 DRE) in tloΔ-TAC1-5 complemented with TLOα1 or TLOγ11 determined by qPCR following anti-HA antibody ChIP of HA-tagged Tlo proteins. HA-tagged Tlo protein signals (IP/Input) at the three sites were normalized to an internal negative control region (IP/IN at CDR1 Up) to show relative occupancy; a ratio of 1 was therefore set as the cut-off value to determine positive occupancy (positive=>1). Data represent three biological replicates analysed using one-way ANOVA (*** indicates P value<0.001, and **** indicates P value<0.0001).

Tloα1 localizes to the DRE in the CDR1 promoter

ChIP-qPCR was performed for HA-tagged Tlo to detect occupancy across the CDR1 promoter region (Fig. 6b). Three upstream CDR1 promoter sites were tested; CDR1 1–2, CDR1 1–3 and CDR1 DRE (Fig. 6b). An upstream CDR1 promoter site (CDR1 Up), previously shown to lack any Mediator binding sites [13], was used as an internal negative control. Extracts from WT-TAC1-5 were chosen as a negative ChIP control since they did not express any HA-tag. Occupancy at each site was expressed as a ratio to CDR1 Up levels and ratios higher than 1 were considered positive enrichment. ChIP-qPCR analysis revealed that Tloα1 from tloΔ-TAC1-5+TLOα1 showed significant occupancy at the promoter site CDR1 1–3 and CDR1 DRE as it showed significantly higher ratios than WT-TAC1-5 (~1). Tloγ11, which we could not detect by Western blot analysis, did not exhibit any significant occupancy at the CDR1 1–3 and CDR1 DRE sites compared to WT-TAC1-5. None of the Tlo proteins exhibited significant occupancy at promoter site CDR1 1–2, with Tloα1 showing a slightly increased ratio relative to CDR1 Up.

Discussion

We have previously shown that the massively expanded TLO gene family, which encodes the Med2 component of the Mediator tail module [27], is involved in a wide range of cellular processes in C. albicans including responses to antifungal drugs [19,21]. The current study aimed to specifically investigate the role of the TLO gene family members in Tac1-activated expression of the CDR1 gene, which encodes an azole antifungal efflux pump and represents one of the most common mechanisms of azole resistance in C. albicans. Previous studies have shown that CDR1 mRNA expression is highly dependent on the Mediator tail components Med3 and Med15 [13]. The role of the Tlo component of the Mediator tail in this Tac1-mediated regulation of CDR1 expression is of particular interest due to the expansion of the TLO gene family (i.e. there are 14 TLO genes, divided into three distinct clades, in C. albicans SC5314). We and others have previously shown that there are functional differences between the three TLO gene family clades [18,19], suggesting that individual Tlo proteins may interact differently with Tac1 or Mediator under specific conditions or perhaps with different affinities.

As expected from our previous studies, the tloΔ, a null mutant in which all 14 copies of TLO have been deleted, phenocopies the med3Δ Mediator tail mutant [13]. Mutation of TAC1 using CRISPR-Cas9 mutagenesis to generate a homozygous hyperactive TAC1-5 allele in the tloΔ and med3Δ mutant backgrounds resulted in reduced fluconazole susceptibility compared to WT strains expressing TAC1-5 (Fig. 1). Reduced levels of CDR1 mRNA expression in C. albicans WT MAY1244 and the two Mediator tail null mutants, tloΔ and med3Δ, confirm that basal CDR1 mRNA expression requires a functional Mediator tail. Deletion of tloΔ and med3Δ also prevented Tac1-mediated CDR1 induction by fluphenazine, indicating that TLO genes likely play a major role in the control of CDR1 expression. Neither of the Mediator tail null mutants we tested significantly affected the basal CDR2 mRNA expression, and fluphenazine was unable to activate strong CDR2 mRNA expression (Fig. 2), supporting recent studies showing that CDR2 expression is highly dependent on chromatin remodelling activity (Swi/Snf) rather than the Mediator tail module [28]. These findings also support the hypothesis that Tac1^GOF mutants and xenobiotic-induced WT Tac1 activate CDR1 transcription through a similar Mediator-dependent mechanism, supporting a potential antifungal strategy to block Tac1 hyperactivation via small molecule inhibitors [13,29].

Introduction of the hyperactive TAC1-5 allele in the tloΔ background provided some additional insights as our data indicate that even in the absence of TLO genes, the hyperactive TAC1-5 allele still exhibited significant functionality. Introduction of TAC1-5 into the tloΔ mutant increased CDR1 and CDR2 expression, as seen by qRT-PCR and RNAseq analysis, albeit to lower levels than in a WT TAC1-5 strain. Despite the relatively low levels of CDR1 and CDR2 mRNA, the tloΔ-TAC1-5 strain exhibited a significant increase in fluconazole MIC (~8-fold), which may not be explained solely by CDR1 and CDR2 mRNA levels. RNAseq analysis of the tloΔ-TAC1-5 strain indicates that the TAC1-5 allele has wide-ranging impacts on gene expression in the tloΔ background, including a negative impact on the expression of genes that require UPC2 for expression (UPC2_DN category). Unexpectedly, expression of TAC1-5 in the tloΔ mutant also reduced fluconazole tolerance (Fig. 1b). O’Connor-Moneley et al. [21] recently characterized fluconazole tolerance in the tloΔ mutant and linked this phenotype with reduced expression of ERG genes in the tloΔ mutant versus WT strains [21]. The impact of the TAC1-5 allele on fluconazole susceptibility clearly goes beyond regulation of CDR1 expression and may involve multiple pathways and targets.

In order to investigate if there is functional diversity within the Tlo family with regards to their ability to facilitate hyperactive Tac1 activity, we used the pTET promoter to express representative TLO genes from the alpha (TLOα1, TLOα3 and TLOα34), beta (TLOβ2) and gamma (TLOγ5, TLOγ7 and TLOγ11) clades in the tloΔ-TAC1-5 strain. We included representatives from the TLO gamma clade even though previous studies showed that these genes (i.e. TLOγ5 and TLOγ11) do not appear to be translated, even when expressed from the strong pENO promoter, and did not confer any detectable phenotypic changes in the tloΔ mutant [19]. This study supports the apparent lack of functionality of gamma Tlos, as none of the genes examined here could restore WT cellular morphology or had a significant impact on fluconazole susceptibility. It is possible that gamma TLO genes have regulatory functions distinct from alpha and beta TLO genes (e.g. production of non-coding regulatory RNAs [30]), and studies to generate specific TLO gamma mutants are underway to investigate this.

In contrast to the gamma clade, expression of the single member of the beta clade, TLOβ2, in the tloΔ-TAC1-5 background could restore filamentous growth in serum and increased resistance to fluconazole in the tloΔ mutant. However, in order to detect these phenotypes, it was necessary to express TLOβ2 from the strong pENO1 promoter. The expression of pTET-TLOβ2 mRNA in the tloΔ-TAC1-5 strain was much lower than that previously observed in the parental tloΔ background and could be related to reduced mRNA stability.

All of the alpha TLO genes examined in our study (i.e. TLOα1, TLOα3 and TLOα34) restored WT cellular morphology, hyphal growth in 10% serum and enhanced resistance to fluconazole in the tloΔ mutant. Expression of these alpha TLO genes resulted in an increase in terbinafine and fluconazole resistance, although TLOα3 induced an intermediate fluconazole resistance phenotype (MIC 12 µg ml⁻¹ compared to 16 µg ml⁻¹). The weaker TLOα3 phenotypes could be expression-related, as was observed with TLOβ2.

We examined Tloα1 activity in detail and showed that it was strongly expressed, restored WT gene expression and increased CDR1 mRNA levels, as well as increased fluconazole resistance, similar to the levels observed in the WT-TAC1-5 strain. ChIP analysis of Tloα1 showed that it localizes to the DRE which is the site of Tac1 binding at the CDR1 promoter. The CDR1 1–3 and CDR1 DRE regions of the CDR1 promoter regions exhibited the greatest enrichment in Tloα1 levels, whereas CDR1 1–2 (the furthest upstream promoter site examined here) had very little enrichment in all Tlo pulldowns. The CDR1 1–3 promoter site includes the SRE2 element, a cis element that has been characterized as a steroid-responding region that can be induced by progesterone and β-estradiol [31]. The DRE promoter site has been previously proposed as the binding site of Tac1 to initiate the induction of CDR1 expression [10,13]. It is currently not clear whether the Tloα1 protein binds directly to the CDR1 DRE, or if the localization is indirect as it is possible that Tloα1 binds indirectly to the DRE via other Mediator components. Future ChIP experiments will be required to clearly show the direct/indirect nature of Tloα1 binding at the CDR1 DRE and CDR1 1–3 promoter sites and could be used to determine which Mediator components can be recruited by Tloα1 at the CDR1 DRE promoter site.

To summarize, we have shown that individual alpha and beta TLO genes can restore fluconazole resistance in a hyperactive Tac1-5 strain where all TLO genes have been deleted. This supports the hypothesis that Tlo protein is an important component of the Mediator tail and, for the first time, directly links Tlo as a requirement for the full activity of a transcriptional regulator. This study also suggests functional diversity in the ability of the Tlo family to activate transcription, providing further evidence that the function of the Tlo gamma sub-family has diverged from encoding canonical Mediator tail polypeptide components. Future studies should determine whether there is diversity in Tac1 interactions within the alpha and beta families, which could impact on the development of antifungal resistance.

Abbreviations

AAD

acidic activation domain

ABC

ATP-binding cassette

DBD

DNA-binding domain

DRE

drug response element

FA

formaldehyde-assisted

GOF

gain-of-function

GOF

gain-of-function

GSEA

Gene Set Enrichment Analysis

IN

input

MHR

middle homology region

PIs

protease inhibitors

TBP

TATA-binding protein

YEPD

yeast extract peptone dextrose