RPD3 and UME6 are involved in the activation of PDR5 transcription and pleiotropic drug resistance in ρ0 cells of Saccharomyces cerevisiae

Background In Saccharomyces cerevisiae, the retrograde signalling pathway is activated in ρ0/− cells, which lack mitochondrial DNA. Within this pathway, the activation of the transcription factor Pdr3 induces transcription of the ATP-binding cassette (ABC) transporter gene, PDR5, and causes pleiotropic drug resistance (PDR). Although a histone deacetylase, Rpd3, is also required for cycloheximide resistance in ρ0/− cells, it is currently unknown whether Rpd3 and its DNA binding partners, Ume6 and Ash1, are involved in the activation of PDR5 transcription and PDR in ρ0/− cells. This study investigated the roles of RPD3, UME6, and ASH1 in the activation of PDR5 transcription and PDR by retrograde signalling in ρ0 cells. Results ρ0 cells in the rpd3∆ and ume6∆ strains, with the exception of the ash1∆ strain, were sensitive to fluconazole and cycloheximide. The PDR5 mRNA levels in ρ0 cells of the rpd3∆ and ume6∆ strains were significantly reduced compared to the wild-type and ash1∆ strain. Transcriptional expression of PDR5 was reduced in cycloheximide-exposed and unexposed ρ0 cells of the ume6∆ strain; the transcriptional positive response of PDR5 to cycloheximide exposure was also impaired in this strain. Conclusions RPD3 and UME6 are responsible for enhanced PDR5 mRNA levels and PDR by retrograde signalling in ρ0 cells of S. cerevisiae. Supplementary Information The online version contains supplementary material available at 10.1186/s12866-021-02373-1.


Background
In the yeast, Saccharomyces cerevisiae, multidrug resistance can result from the overexpression of plasma membrane-localized ABC transporters, such as Pdr5, Snq2, and Yor1. Pdr5 is a major efflux pump of functionally and structurally unrelated antifungal drugs or compounds, such as fluconazole and cycloheximide [1]. Expression of PDR5 can be induced by the paralogous Zn2Cys6 transcription factors, Pdr1 and Pdr3. Pdr1 and Pdr3 can form both homo-and heterodimers, and directly bind structurally diverse drugs and xenobiotics [2,3]. PDR5 has four perfect and degenerate pleiotropic drug response elements (PDREs) in its promoter region [4]. Pdr1 and Pdr3 are constitutively bound to PDRE in the PDR5 promoter [3]. Although PDR1 does not have a PDRE in its promoter region, PDR3 has two PDREs, and is thereby regulated by an autoregulatory loop involving Pdr3 [5]. Thus, PDR3 is transcriptionally regulated by both Pdr1 and Pdr3 via PDRE [5]. Gain-offunction mutations in Pdr1 and Pdr3 lead to hyperactive transcription of ABC transporter genes, such as PDR5, SNQ2, and YOR1, resulting in the induction of pleiotropic drug resistance (PDR) [6,7].
Although PDR1 and PDR3 have functionally redundant roles in PDR, PDR1 plays a predominant role in PDR to toxic agents and basal PDR5 expression [8,9]. Deletion of PDR1 increases susceptibility to cycloheximide compared to PDR3, whereas the disruption of PDR1 and PDR3 causes hypersensitivity to cycloheximide and oligomycin compared with a single disruption [9].
PDR3 plays a predominant role in the PDR of ρ 0/− cells without mitochondrial DNA in S. cerevisiae [10]. The retrograde signalling pathway is strongly activated in ρ 0/− cells of S. cerevisiae and Candida glabrata, and the expression of multidrug resistance genes, including PDR5 and CgCdr1 is induced. The induction of PDR5 in this retrograde signalling pathway requires Pdr3 but not Pdr1 [10]. In the retrograde signalling pathway, Pdr3 directly interacts with the Hsp70 chaperone Ssa1, and Pdr3 activity is inhibited [11]. Intriguingly, this Pdr3-Ssa1 association is decreased in ρ 0/− cells, and Pdr3 activity is strongly stimulated [11].
In S. cerevisiae, small (0.6 MDa) (Rpd3S) and large (1.2 MDa) (Rpd3L) corepressor complexes exist, sharing specific subunits such as a histone deacetylase, Rpd3, Ume1, and Sin3 [12]. They participate in chromatin remodelling and transcriptional repression [13,14]. Dep1 and Sds3 are specific subunits of the Rpd3L complex, whereas the Rpd3S complex contains two unique subunits, Rco1p and Eaf3p [12]. Rpd3 is a histone deacetylase, while Sds3 is involved in transcriptional silencing and sporulation. The Rpd3L complex contains the DNA binding transcription factors, Ume6 and Ash1, which are responsible for the targeted deacetylation of gene promoters [12]. For example, Ume6 binds the URS1 upstream regulatory sequence on the INO1 promoter and represses INO1 expression by recruiting Rpd3 via the corepressor Sin3 and chromatin remodelling factor Isw2 [14,15]. Although both Ume6 and Ash1 are bound to the INO1 and HO promoters, Ume6 specifically represses INO1 gene expression, and Ash1 specifically inhibits HO gene expression [12]. UME6 is known to repress carbon/ nitrogen metabolic and early meiotic gene expression while participating in gene activation [16].
Borecka-Melkusova et al. have shown that Rpd3 is required for basal PDR5 transcription and Pdr3-mediated PDR [17]. In addition, the sds3Δ, dep1Δ, and rpd3Δ strains are sensitive to drugs, indicating that the Rpd3L complex is involved in PDR [17]. In contrast to sds3Δ, dep1Δ, and rpd3Δ strains, the ume6Δ and ash1Δ strains displayed no sensitivity to cycloheximide at the minimum inhibitory concentration [17]. In addition, Yibmantasiri et al. have also reported that the ume6Δ strain does not confer sensitivity to a range of fungicides including cycloheximide, ketoconazole, fluconazole, oligomycin, and benomyl when tested in a spot dilution assay [18]. The authors also reported that deletion of UME6 does not reduce Pdr5 expression in western blot analysis [18]. Robbins et al. reported that the decreased azole resistance in the rpd3Δ strain of S. cerevisiae does not result from downregulation of PDR5 mRNA [19]. Rather, it results from diminished Hsp90-dependent antifungal drug resistance in Candida albicans and S. cerevisiae [19]. Furthermore, Jensen et al. reported that artemisinin sensitivity in the rpd3Δ strain of S. cerevisiae occurred due to the impaired endoplasmic reticulum (ER) to Golgi trafficking of Pdr5, and not from transcriptional downregulation of PDR5 [20].
As mentioned above, PDR5 expression or PDR in the rpd3Δ, ume6Δ, and ash1Δ mutant strains, has been examined mainly in ρ + cells with mitochondrial DNA, but not in ρ 0/− cells. However, Borecka-Melkusova et al. showed that ρ 0/− cells in the rpd3Δ strain also have significantly lower cycloheximide resistance than those in the wild-type [17]. Although PDR5 transcription and PDR are activated by retrograde signalling via Pdr3 in ρ 0/− cells, whether RPD3, UME6, and ASH1 are involved in the activation of PDR5 transcription and PDR in ρ 0/− cells has not yet been examined. Therefore, this study investigated the roles of RPD3, UME6, and ASH1 in the activation of PDR5 transcription and PDR by retrograde signalling in ρ 0 cells.

Susceptibility of ρ 0 cells in strain ume6∆ to fluconazole and cycloheximide
To investigate whether UME6 and ASH1 are involved in the PDR of ρ 0 cells, the sensitivity of ρ 0 cells in the ume6∆ and ash1∆ mutant strains to fluconazole and cycloheximide was examined using a spot dilution assay. ρ 0 cells in the ume6∆::bleMX6, pdr3∆::bleMX6, and rpd3∆::bleMX6 mutant strains were more sensitive to fluconazole and cycloheximide than those in the wildtype, ash1∆::bleMX6, and gat3∆::bleMX6 strains (Fig. 1). However, ρ 0 cells of the ume6∆ mutant strain were less susceptible to fluconazole and cycloheximide than those in the pdr3∆ and rpd3∆ mutant strains (Fig. 1). Similar results were also observed for ρ 0 cells in the wild-type, ume6∆::KanMX, pdr3∆::KanMX, and rpd3∆::KanMX, ash1∆::KanMX, and gat3∆::KanMX strains (data not shown). We also obtained similar results in ρ 0 cells of the ume6∆ mutant derived from the W303-1A strain (data not shown). Since fluconazole and cycloheximide are functionally and structurally unrelated compounds, these results suggest that UME6, but not ASH1, is responsible for the activation of PDR5 transcription and PDR in ρ 0 cells of S. cerevisiae. However, it cannot be ruled out that ASH1 may be responsible for resistance to other drugs in ρ 0 cells of S. cerevisiae.
In the spot dilution assay, the ρ 0 cells in the ume6∆ mutant displayed less susceptibility to fluconazole and cycloheximide than in the pdr3∆ and rpd3∆ mutants. Therefore, we further investigated whether UME6 is responsible for PDR in ρ 0 cells using a co-cultivation assay [21]. As ρ 0 cells of the ume6∆ mutant, but not of the ash1∆ and gat3∆ mutants, were more sensitive to fluconazole and cycloheximide than those in the wild-type strain, ash1∆ and gat3∆ mutants were used as controls for the ume6∆ mutant in the co-cultivation assay (Fig. 1). ρ 0 cells in two mutant strains, gat3∆::KanMX and ume6∆::bleMX6, were co-cultivated in the presence and absence of 100 μg/mL fluconazole. The number of viable cells of each mutant strain in the co-culture was estimated by spreading them onto yeast extract peptone dextrose (YPD) plates containing G418 or Zeocin. We found that ρ 0 cells in the ume6∆::bleMX6 strain were eliminated from the co-culture over time in the presence of fluconazole, but not in the absence of fluconazole (p < 0.05) ( Fig. 2A). Similar results were also observed when ρ 0 cells of the gat3∆::bleMX6 strain were co-cultivated with those of ume6∆::KanMX strain in the presence and absence of 100 μg/mL fluconazole, indicating that selection marker genes do not affect these changes in survival rate (data not shown). Furthermore, similar results were obtained when ρ 0 cells of the gat3∆ strain were co-cultivated with those of ume6∆ strain in the presence and absence of 0.5 μg/mL cycloheximide (data not shown). Rather than using the gat3∆ strain, ρ 0 cells of the ash1∆::bleMX6 strain were co-cultivated with those of ume6∆::KanMX in the presence and absence of 0.5 μg/mL cycloheximide. The survival rate of each strain in the co-culture was estimated in the same way. Consequently, ρ 0 cells in the ume6∆::KanMX strain were eliminated earlier from the co-culture over time in the presence of cycloheximide compared to that in the absence of cycloheximide (p < 0.05) (Fig. 2B). Furthermore, similar results were observed in the co-cultivation of ρ 0 cells in the ash1∆::KanMX and ume6∆::bleMX6 mutant strains in the presence of 100 μg/mL fluconazole (data not shown). These results suggest that UME6, but not ASH1, contributes to the activated PDR in ρ 0 cells of S. cerevisiae.

Deletion of RPD3 and UME6, but not ASH1, decreases the PDR5 mRNA level
Decreased PDR in ρ 0 cells of the rpd3∆ and ume6∆ strains suggested that enhanced transcriptional expression of PDR5 by retrograde signalling is suppressed in ρ 0 cells. Thus, we investigated the expression levels of PDR5 mRNA in ρ 0 and ρ + cells of the wild-type, rpd3∆, ume6∆, pdr3∆, ash1∆ mutant strains grown to the logarithmic phase by real-time RT-PCR.
PDR5 expression was more strongly induced in the wild-type strain of ρ 0 cells than in the wild-type strain of ρ + cells (p < 0.05) ( Fig. 3A and Table S1). Within ρ 0 cells, PDR5 expression of the pdr3∆ strain was significantly more suppressed compared to the wild-type strain (p < 0.05), while in case of ρ + cells, the PDR5 expression level in the pdr3∆ strain was not significantly different from the wild-type strain (p > 0.05) ( Fig. 3B and Table S2). These results were consistent with those of previous reports [9].
The PDR5 mRNA levels in ρ 0 cells of the rpd3∆, ume6∆, and pdr3∆ mutant strains were significantly lower than those in the wild-type and ash1∆ mutant strains (p < 0.05) ( Fig. 3A and Table S1). This suggests that activated PDR5 transcriptional expression by retrograde signalling was significantly reduced in ρ 0 cells of the rpd3∆, ume6∆, and pdr3∆ strains (Fig. 3A). Thus, the increased drug susceptibility of ρ 0 cells in the rpd3∆, ume6∆, and pdr3∆ strains shown in spot dilution and co-cultivation assays can be explained by the reduction in PDR5 mRNA levels. In contrast, PDR5 mRNA levels in ρ + cells of the ume6∆, rpd3∆, and pdr3∆ strains were similar to those in ρ + ash1Δ pdr3Δ rpd3Δ gat3Δ ume6Δ Fig. 1 UME6, but not ASH1, is responsible for the PDR of ρ 0 S. cerevisiae. Fluconazole or cycloheximide resistance of ρ 0 cells in the wild-type strain (FY1679-28C) and its derivative strains, ume6∆::bleMX6, ash1∆::bleMX6, pdr3∆::bleMX6, and rpd3∆::bleMX6, and gat3∆::bleMX6 was determined by the spot dilution assay cells of the wild-type and ash1∆ mutant strains (p > 0.05) ( Fig. 3B and Table S2). Therefore, UME6 and RPD3 are responsible for the enhanced transcriptional expression of PDR5 by Pdr3-mediated retrograde signalling in ρ 0 cells but not for basal expression of PDR5 in ρ + cells. In addition, ρ 0 cells, but not ρ + cells, in the ume6∆ mutant strain, had slightly more reduced PDR3 mRNA levels than those in the wild-type strain, suggesting minor involvement of UME6 in the activation of autoregulated transcriptional expression of PDR3 by retrograde signalling (Figs. 3A and 4).
Furthermore, this study investigated whether the activated transcriptional expression of PDR5 following cycloheximide exposure occurs in ρ 0 cells of the ume6∆ mutant strain using real-time RT-PCR. PDR3 and PDR5 mRNA levels in ρ 0 cells of the ume6∆ strain were lower than those in the wild-type strain, independent of the addition of cycloheximide. However, PDR5 mRNA levels in ρ 0 cells of the wild-type and ume6∆ strains increased 1.6-and 1.55 times, respectively, after exposure to 0.2 μg/mL cycloheximide for 45 min (Fig. 4 and Table S3). These increases in PDR5 mRNA levels in ρ 0 cells were statistically significant in the wild-type (p < 0.05), but not in the ume6∆ strains (p > 0.05). This suggests that UME6 is involved in the transcriptional expression of PDR5 in cycloheximide-exposed and unexposed ρ 0 cells and the intact induction of PDR5 transcription after drug exposure in ρ 0 cells.

Discussion
This study revealed that UME6 and RPD3, but not ASH1, are responsible for enhancing PDR5 expression and PDR by retrograde signalling in ρ 0 cells of S. cerevisiae. In addition, UME6 was involved in the transcriptional expression of PDR5 in cycloheximide-exposed and unexposed ρ 0 cells, and the enhancement of PDR5 transcription after cycloheximide exposure was also impaired in ρ 0 cells of the ume6∆ mutant strain. Reduced PDR5 mRNA levels in the presence and absence of cycloheximide were also reported in ρ + cells of the rpd3∆ strain by Borecka-Melkusova et al.; however, this report was invalidated later by Robbins et al. [17,19]. Histone deacetylation leads to transcriptional repression and activation [14,22]. Thus, Ume6 and Rpd3 may serve as enhancers of PDR5 expression by retrograde signalling in ρ 0 cells, different from their usual roles as repressors. It is currently unknown how Rpd3 and Ume6 enhance the transcriptional expression of PDR5 and PDR in ρ 0 cells and why they do not affect basal PDR5 expression in ρ + cells. As Ume6 binds to the PDR5 promoter region in ρ + cells, it may also be localised at the PDR5 promoter region in ρ 0 cells [23]. If this is true, Rpd3 [24]. L-Mediator (of the Mediator complexes) contains the Cdk8 subcomplex, which consists of the cyclin-dependent kinase Cdk8 (Med15/ Srb8), Med12 (Srb8), and Med13 (Srb9). Deletion of Med12 from the Cdk8 complex completely suppressed the induction of PDR5 expression in ρ 0 cells but not in ρ + cells, indicating a difference in the regulatory machinery of PDR5 transcription between ρ + and ρ 0 cells [25]. This difference may be associated with the difference in transcriptional regulation of PDR5 by Rpd3 and Ume6 between ρ 0 and ρ + cells. This study showed that ρ 0 cells in the ume6∆ mutant were less susceptible to fluconazole and cycloheximide than in the pdr3∆ and rpd3∆ mutants as assessed by the spot dilution assay (Fig. 1). This suggests that ρ 0 cells in the ume6∆ mutant, but not the rpd3∆ mutant, maintain Hsp90-dependent antifungal drug resistance and intact ER to Golgi trafficking of Pdr5 [19,20]. Furthermore, fewer multidrug resistance genes other than PDR5 are downregulated in ρ 0 cells of the ume6∆ mutant than in those of the pdr3∆ mutant [26]. ρ 0 cells, but not ρ + cells, in the ume6∆ mutant strain had slightly lower PDR3 mRNA levels than those in the wild-type strain (Figs. 3 and 4). Ume6 binds to the PDR3 promoter region in ρ + cells; therefore, it may also bind to the PDR3 promoter region in ρ 0 cells and may directly activate the transcriptionally auto regulated loop of PDR3 by chromatin remodelling and facilitating Pdr3 binding [23].
The emergence of multidrug-resistant fungi is a serious clinical concern [27]. Therefore, the efficacy of combined antifungal agents against multidrug-resistant fungi has been examined [28]. Furthermore, to treat multidrugresistant fungal infections, the efficacy of using histone deacetylase inhibitors or essential oils from plants in combination with the primary classes of antifungals has also been examined [29][30][31][32][33][34][35]. For example, a histone deacetylase inhibitor, trichostatin A, decreases the upregulation of CDR1, ERG1, and ERG11 by azole and enhances azole sensitivity in C. albicans [33]. Uracil-based histone deacetylase inhibitors 1c and 1d reduce acquired resistance to antifungals and trailing growth in C. albicans [34]. In addition, RPD3 is responsible for azole resistance and basal transcription of efflux genes such as CDR1, CDR2, and MDR1 in ρ + cells of pathogenic C. albicans [35]. Thus, Ume6 may also be responsible for multidrug resistance via transcriptional regulation of the efflux genes in ρ + and ρ 0 cells of pathogenic Candida species. Therefore, identifying specific inhibitors of Ume6 may lead to the development of drugs against multidrugresistant pathogenic Candida species.

Conclusions
PDR5 expression or PDR in the rpd3Δ, ume6Δ, or ash1Δ mutant strains has been examined primarily in ρ + cells, but not in ρ 0/− cells. In this study, we investigated the roles of RPD3, UME6, and ASH1 in the activation of PDR5 transcription and PDR by retrograde signalling in ρ 0 cells. Using spot dilution and co-cultivation assays, we have shown that RPD3 and UME6, but not ASH1, contribute to the PDR in ρ 0 cells of S. cerevisiae. In addition, using real-time PCR assay, we have shown that RPD3 and UME6, but not ASH1, are involved in the transcriptional expression of PDR5 in ρ 0 cells, and UME6 also contributes to PDR5 transcription and its enhancement in cycloheximide-exposed ρ 0 cells. This work provides useful knowledge on the genetic basis of yeast multidrug resistance via transcriptional regulation of efflux genes.

Spot dilution assay
A spot dilution assay was conducted to estimate the relative resistance of each yeast strain to fluconazole or cycloheximide [37,38]. Three independently derived ρ 0 cells from each yeast strain were aerobically grown to an OD 600 of 0.6-0.9 at 30 °C in YPD medium. Five microliters of 10-fold serial dilutions of the logarithmic phase cultures containing the same number of cells were spotted on YPD plates containing or not containing 20 μg/ mL fluconazole (Nacalai Tesque) (or 0.5 μg/mL cycloheximide (Wako)) and incubated at 30 °C for 7 days. Representative plate images of three replicates were captured after culturing at 30 °C for 7 days.

Co-cultivation of two gene deletion mutants replaced with KanMX or bleMX6 gene cassettes
The ume6∆::KanMX and ash1∆::bleMX6 mutant strains, or the ume6∆::bleMX6 and ash1∆::KanMX mutant strains were co-cultivated in YPD medium containing or not containing 100 μg/mL fluconazole (or 0.5 μg/ mL cycloheximide). The aliquots of the co-culture were recovered immediately before adding drugs and at various times after addition of the drugs, and spread on the YPD plates containing G418 (Wako) or Zeocin (Nacalai Tesque) [20]. The viability of each strain at each time point was estimated from the colony numbers on G418 and Zeocin plates [20]. The previous experiments were also performed in the ume6∆::KanMX and gat3∆::bleMX6 mutant strains, or the ume6∆::bleMX6 and gat3∆::KanMX mutant strains.

RNA extraction from ρ 0 and ρ + cells of each mutant strain grown to the logarithmic phase
Two independently derived ρ 0 and ρ + cells from each yeast strain were grown to an OD 600 of 7-9 in YPD, diluted to an OD 600 of 0.2, and grown for an additional 5 h in duplicate [39,40]. Aliquots of the duplicates were recovered. The cells in the aliquots above were pelleted, washed, frozen at − 80 °C, and used to extract total RNA [39,40]. Total RNA was isolated from yeast cells using Nucleospin RNA Plus (TaKaRa), according to the manufacturer's protocol.

RNA extraction from ρ 0 cells exposed to drug
Independently derived ρ 0 cells from each yeast strain were grown to an OD 600 of 7-9 in YPD, diluted to an OD 600 of 0.2, and grown for an additional 5 h in triplicate [39,40]. Aliquots of the triplicates were harvested just before the addition of cycloheximide to the medium. Cycloheximide (0.2 μg/mL) was added to one of the triplicates, and the triplicates were grown for 45 min and 90 min at 30 °C. Aliquots of the triplicates were recovered at 45 min and 90 min after adding cycloheximide to one of the triplicates. The cells in the aliquots above were pelleted, washed, frozen at − 80 °C, and used to extract total RNA [39,40]. Total RNA was isolated from yeast cells before and after exposure to cycloheximide using Nucleospin RNA Plus (TaKaRa), according to the manufacturer's protocol.

Real-time RT-PCR
Reverse transcription of total RNA was performed using the PrimeScript 1st strand cDNA Synthesis Kit (TaKaRa). SYBR Green qRT-PCR was performed using the TB Green ® Premix Ex Taq II (TaKaRa) in a Step One Real-time PCR system (Applied Biosystems). For quantitative PCR analysis, the housekeeping gene ACT1 was used as an endogenous control to normalise the expression level of each target gene [41]. Minus reverse transcriptase control was used as the negative control. qPCR for each sample was performed in duplicate or triplicate. Serial dilutions of the control cDNA from the wild-type strain were prepared to produce a standard curve for each primer pair. The primers used for PDR3 were: forward, 5′-TAC CGC AGA AGG AGG ATA GTT CCC A-3′ and reverse, 5′-GCT TAA TCG CAG TGT CCA GAT GCT GTAC-3′, yielding a PCR product of 117 bp. The qPCR for PDR5 was performed using primers 5′-CTC TGA GAG AAC CCT GAA CAA AGA TAT GCT A-3′ (forward) and 5′-ATA GCT TCA CGG CTT GCT TCA TCG T-3′ (reverse) to amplify a fragment of 165 bp. The primers used for ACT1 were as follows: forward, 5′-CAA ATT ATG TTT GAA ACT TTC AAC GTT CCA G-3′ and reverse, 5′-ACG TGA GTA ACA CCA TCA CCG GAA TC-3′, yielding a PCR product of 125 bp.

Statistical analysis
The survival rate of ume6∆ strain at each time point in Fig. 2 was normalized by that at 0 h. Paired t-test was used for statistical analysis in Figs. 2 and 4. Unpaired Student's t-test was used for statistical analysis in Fig. 3. p < 0.05 was considered significant.