Candida glabrata Med3 Plays a Role in Altering Cell Size and Budding Index To Coordinate Cell Growth

This study demonstrates that CgMed3AB can regulate cell growth in C. glabrata by coordinating the homeostasis of cellular acetyl-CoA metabolism and the cell cycle cyclin CgCln3. Specifically, we report that CgMed3AB regulates the cellular acetyl-CoA level, which induces the transcription of Cgcln3, finally resulting in alterations to the cell size and budding index. In conclusion, we report that CgMed3AB functions as a wheel responsible for driving cellular acetyl-CoA metabolism, indirectly inducing the transcription of Cgcln3 and coordinating cell growth. We propose that Mediator subunits may represent a vital regulatory target modulating cell growth in C. glabrata.

ABSTRACT

Candida glabrata is a promising microorganism for the production of organic acids. Here, we report deletion and quantitative-expression approaches to elucidate the role of C. glabrata Med3AB (CgMed3AB), a subunit of the mediator transcriptional coactivator, in regulating cell growth. Deletion of CgMed3AB caused an 8.6% decrease in final biomass based on growth curve plots and 10.5% lower cell viability. Based on transcriptomics data, the reason for this growth defect was attributable to changes in expression of genes involved in pyruvate and acetyl-coenzyme A (CoA)-related metabolism in a Cgmed3abΔ strain. Furthermore, the mRNA level of acetyl-CoA synthetase was downregulated after deleting Cgmed3ab, resulting in 22.8% and 21% lower activity of acetyl-CoA synthetase and cellular acetyl-CoA, respectively. Additionally, the mRNA level of CgCln3, whose expression depends on acetyl-CoA, was 34% lower in this strain. As a consequence, the cell size and budding index in the Cgmed3abΔ strain were both reduced. Conversely, overexpression of Cgmed3ab led to 16.8% more acetyl-CoA and 120% higher CgCln3 mRNA levels, as well as 19.1% larger cell size and a 13.3% higher budding index than in wild-type cells. Taken together, these results suggest that CgMed3AB regulates cell growth in C. glabrata by coordinating homeostasis between cellular acetyl-CoA and CgCln3.

IMPORTANCE This study demonstrates that CgMed3AB can regulate cell growth in C. glabrata by coordinating the homeostasis of cellular acetyl-CoA metabolism and the cell cycle cyclin CgCln3. Specifically, we report that CgMed3AB regulates the cellular acetyl-CoA level, which induces the transcription of Cgcln3, finally resulting in alterations to the cell size and budding index. In conclusion, we report that CgMed3AB functions as a wheel responsible for driving cellular acetyl-CoA metabolism, indirectly inducing the transcription of Cgcln3 and coordinating cell growth. We propose that Mediator subunits may represent a vital regulatory target modulating cell growth in C. glabrata.

INTRODUCTION

Mediator, a multisubunit transcriptional coactivator, functions as a physical bridge between transcription factors and RNA polymerase II and facilitates the assembly of the preinitiation complex on core promoters (1–4). It comprises 25 subunits in Saccharomyces cerevisiae, divided into four modules: head, middle, tail, and kinase (5). Each module has specific functions: the head and middle modules mainly contact RNA polymerase II, whereas the tail and kinase modules integrate the transcriptional-regulatory signals from sequence-specific transcription factors with RNA polymerase II (6). Mediator is a global regulator of the transcriptional process (7), which involves chromatin remodeling (5), transcription elongation (8), and RNA processing (9). Therefore, mediator is responsible for several physiological processes, such as cell growth-related regulation (10), stress response (11–14), and morphological development (15, 16).

The Mediator complex is an essential part of the machinery regulating gene expression, which is necessary for an organism to survive (3). Single deletions of 10 of 16 subunits in the head and middle modules are lethal (4), whereas 9 subunits of the tail and kinase modules are required for diverse cell growth-related processes (17). Deletion of Med18 in the head module causes slower growth than wild-type cells (18), and lack of Med19 and Med20 weakens cell viability as a result of downregulating the transcriptional level of Cyc1 (19). Loss of the middle module subunit Med7 alters the ribosomal regulon, glycolysis-related genes, and the filamentous growth regulator family (20). The tail module comprises five subunits: Med2, Med3, Med5, Med15, and Med16. Med2 in Candida spp. not only regulates cell viability under diverse stress conditions, but also facilitates filamentous growth (21–23). Med5Δ and Med16Δ commonly possess several phenotypes, including enhanced growth on nonfermentable carbon sources, increased citrate synthase activity, and increased oxygen consumption (24). Med15 leads to amyloid-like protein formation under H₂O₂ stress conditions, and overexpression of the Med15 poly(Q) domain causes a slow-growth phenotype under osmotic stress conditions (12). The kinase module contains Srb8, Srb9, Ssn3, and Ssn8. Deletion of Ssn3 has been shown to cause a wrinkled colony morphology in the presence of PYO (15) and an abnormal conidium morphology in Fusarium graminearum (16). The Med3 subunit of the tail module, which is essential for transcriptional activation of SAGA-dependent genes, is phosphorylated by the kinase activity of the CDK8 module (25). In S. cerevisiae, ScMed3 enables cell growth under different stress conditions (12, 26) and expression of carbohydrate metabolism-related genes (27). In Candida albicans, CaMed3 combines with TLO to influence hyphal morphology formation and filamentous growth, both of which are essential for fungal pathogenicity (21, 28, 29).

Candida glabrata is a major industrial microorganism that is used for the production of organic acids, such as fumaric acid (30), malic acid (31), and α-ketoglutaric acid (32). In addition, C. glabrata is the only microorganism used for the commercial fermentation of pyruvate, which has diverse applications in the pharmaceutical and agrochemical fields (33). Moreover, homeostasis between cell growth and product formation remains a challenge for industrial strains. In our previous study, we uncovered an orthologue of CaMed3 in C. glabrata, CgMed3AB, and found that it could modulate cell membrane lipid composition in response to acid stress (13). Early studies of Med3 have focused mainly on its involvement in the stress response, whereas the role of CgMed3AB in modulating cell growth has remained unclear. Here, we evaluated the role of CgMed3AB in regulating cell growth and whether increasing its expression could potentiate growth ability (Fig. 1).

FIG 1.

Schematic diagram of the study. CgMed3AB functions as a wheel that drives cellular acetyl-CoA metabolism and then regulates transcription of Cgcln3.

RESULTS

CgMed3AB changes C. glabrata growth performance.

To investigate whether CgMed3AB affected C. glabrata growth performance, we first carried out spot assays of wild-type, Cgmed3abΔ, and Cgmed3abΔ/(Cgmed3ab)_OE (Cgmed3ab overexpressed by plasmid pY26) strains grown on YNB (0.67% yeast nitrogen base without amino acid, 2% glucose, pH 6.0). We found that the Cgmed3abΔ strain exhibited a weak growth defect, whereas overexpression (OE) of Cgmed3ab resulted in improved growth (Fig. 2A). Next, growth curves were plotted for all three strains. As shown in Fig. 2B, the final biomass of the Cgmed3abΔ strain decreased by 8.6% compared with the wild type, whereas that of the Cgmed3abΔ/(Cgmed3ab)_OE strain showed a 12.1% increase. Furthermore, cells were incubated in YNB for 12 h, and the viability of CgMed3AB mutants was assayed. The Cgmed3abΔ strain exhibited a 10.2% decrease in viability compared to the wild type, whereas overexpression of Cgmed3ab resulted in a 1.5-fold increase (Fig. 2C). These results indicate that CgMed3AB could be relevant to the growth performance of C. glabrata.

FIG 2.

CgMed3AB changes C. glabrata growth performance. (A) Wild-type, Cgmed3abΔ, and Cgmed3abΔ/(Cgmed3ab)_OE strains were spotted on YNB plates. (B) Growth curves for the wild-type, Cgmed3abΔ, and Cgmed3abΔ/(Cgmed3ab)_OE strains. (C) Cell viability in the wild-type, Cgmed3abΔ, and Cgmed3abΔ/(Cgmed3ab)_OE strains. The data refer to biological repeats; the error bars represent standard deviations (SD). *, P < 0.05; **, P < 0.01.

Deletion of CgMed3AB decreases acetyl-coenzyme A (CoA) synthetase (ACS) activity.

To further clarify the role of CgMed3AB in regulating cell growth performance, transcriptome data for the wild type and the Cgmed3abΔ strain grown at pH 6.0 were analyzed. Restrictive thresholds (greater than or equal to 1.5 or less than or equal to −1.5-fold change; P ≤ 0.05) of differentially expressed genes (DEGs) were used to screen the genes. Transcriptome sequencing (RNA-Seq) revealed 1,053 genes whose expression changed significantly at pH 6.0 between the wild type and the Cgmed3abΔ strain, including 633 upregulated genes and 420 downregulated genes (see Data Set S1 in the supplemental material). The affected genes were involved mainly in carbohydrate metabolism, amino acid metabolism, nucleotide metabolism, and cell growth and death (as annotated in the Kyoto Encyclopedia of Genes and Genomes [KEGG] database). Specifically, carbohydrate metabolism was the module exhibiting the largest variation, accounting for 25.0% of all hits (Fig. 3A; see Data Set S2 in the supplemental material). The above-mentioned data indicated that the cell growth defect of the Cgmed3abΔ strain could be related to carbohydrate metabolism.

FIG 3.

Transcriptome analysis of the wild type and the Cgmed3abΔ strain at pH 6.0. (A) Comparison of differentially expressed genes between the wild type and the Cgmed3abΔ strain as annotated using the KEGG database. (B) Heat map of the carbohydrate metabolism module at pH 6.0 following comparison between the wild type and the Cgmed3abΔ strain.

Therefore, we focused on comparing the expression of genes related to carbohydrate metabolism between the wild type and the Cgmed3abΔ strain at pH 6.0. Figure 3B shows significant enrichment among genes involved in pyruvate metabolism, the citrate cycle, and glycolysis/gluconeogenesis. Further analysis revealed that the most altered genes were those of the acetyl-CoA and pyruvate metabolic pathways (see Data Set S3 in the supplemental material). Furthermore, the mRNA levels of genes involved in acetyl-CoA and pyruvate-related metabolism were analyzed by real-time PCR (RT-PCR). In the acetyl-CoA biodegradation pathways, mRNA levels were 1.8-fold higher for acetyl-CoA carboxylase (acc1), 1.2-fold higher for malate synthase, and 1.9-fold lower for citrate synthase (cit1) in the Cgmed3abΔ strain than in the wild type. In the acetyl-CoA biosynthesis pathways, the values were 1.2-fold, 1.3-fold, 1.2-fold, and 1.5-fold higher for pda1, lat1, pot1, and erg10, respectively, and 3.8-fold lower for acs2. In addition, mRNA levels in pyruvate-related metabolism were 1.8-fold, 1.9-fold, and 2.1-fold lower for pyk1, thi3, and pdc3, respectively (Fig. 4). These data indicated that deletion of Cgmed3ab strongly downregulated transcription of acs2, encoding ACS. Meanwhile, ACS activity in the Cgmed3abΔ strain was 22.8% lower than in the wild type, whereas the activity in the overexpression strain exhibited a 24.4% increase (Fig. 5A). This raised the question of how CgMed3AB regulated the expression of acs2.

FIG 4.

Deletion of Cgmed3ab affects gene expression of cellular acetyl-CoA and pyruvate-related metabolism. Pyruvate dehydrogenase (PDH), ACS, and acetyl-CoA acyltransferase (pot1) are the main pathways in cellular acetyl-CoA biosynthesis, whereas acetyl-CoA carboxylase (ACC), citrate synthase (CS), and malate synthase (MS) are the main pathways in cellular acetyl-CoA biodegradation. Pyruvate kinase (PYK) and pyruvate decarboxylase (PDC) are the main pathways in pyruvate-related metabolism. TCA, tricarboxylic acid cycle. Each experiment was followed by three biological repeats; the error bars represent SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

FIG 5.

CgMed3AB affects acetyl-CoA synthetase activity by activating the transcription of Cgacs2. (A) Acetyl-CoA synthetase activity in wild-type, Cgmed3abΔ, and Cgmed3abΔ/(med3ab)_OE strains. (B) Cgino4 mRNA levels in the wild type and the Cgmed3abΔ strain. (C) Cgacs2 mRNA levels in Cgmed3abΔ, Cgino4Δ, and Cgmed3abΔ Cgino4Δ strains. (D) Yeast two-hybrid assays confirmed the interaction between CgMed3AB and CgIno4. P53-simian virus 40 (SV40) large T antigen (Clontech) protein interaction was used as a positive control. (E) Coimmunoprecipitation assays were used to detect the interaction between CgMed3AB and CgIno4 in vivo. (F) EMSAs of CgIno4 protein with upstream promoter regions of Cgacs2. The DNA probe (10 nM) was incubated with a protein concentration gradient (0, 0.5, 1, and 1.5 μM). EMSAs with a 200-fold excess of unlabeled specific probe and nonspecific competitor DNA (salmon sperm DNA) were conducted as controls. (G) Association of CgIno4 with the core promoter of Cgacs2 as determined by ChIP analysis and RT-PCR to measure occupancy. Signals were normalized to the input DNA, ChrV was used as a negative control, and the promoter was the core region of the Cgacs2 promoter. All experiments were repeated three times; the error bars represent SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

To investigate the role of CgMed3AB in regulating expression of acs2, we first analyzed the transcription level of CgIno4, a transcription activator of acs2 (34). The mRNA level of CgIno4 in the Cgmed3abΔ strain decreased by 16% compared with the wild type (Fig. 5B). Seemingly, the mRNA level of Cgacs2 in the Cgmed3abΔ Cgino4Δ double mutant was 40.7% and 17.8% lower than in the respective Cgmed3abΔ and Cgino4Δ single mutants (Fig. 5C). Therefore, we speculated that CgMed3AB might interact with CgIno4 and then regulate the transcription process of Cgacs2. To validate this hypothesis, yeast two-hybrid assays and coimmunoprecipitation assays were performed, and the results indicated that CgMed3AB interacted with CgIno4 (Fig. 5D and E). Furthermore, to investigate whether the occupancy of CgIno4 binding to the acs2 promoter in the Cgmed3abΔ strain was changed, electrophoretic mobility shift assays (EMSAs) were used to analyze the binding in CgIno4 and acs2. The results revealed a direct binding interaction between CgIno4 and Cgacs2, as the DNA probe containing the promoter region of Cgacs2 clearly shifted following incubation with purified recombinant CgIno4 (Fig. 5F), suggesting that CgIno4 bound to the promoter of Cgacs2. Finally, chromatin immunoprecipitation (ChIP) combined with RT-PCR was performed to analyze the occupancy of CgIno4 binding to the Cgacs2 promoter. The ChIP signal in the Cgmed3abΔ strain was 16.3% lower than in the wild type (Fig. 5G). Taken together, these data suggest that CgMed3AB could interact with CgIno4 to coordinate transcription of CgAcs2.

Acetyl-CoA synthetase activity affects the content of acetyl-CoA.

Based on the above-described results, we further investigated whether cellular acetyl-CoA levels changed as a result of lower ACS activity. Cellular acetyl-CoA levels were measured in the wild-type, Cgmed3abΔ, and Cgmed3abΔ/(med3ab)_OE strains. The acetyl-CoA level was 27.4% lower in Cgmed3abΔ than in wild-type cells but increased by 32.8% in overexpressing cells (Fig. 6A).

FIG 6.

CgMed3AB affects transcription by regulating the cellular acetyl-CoA level. (A) Cellular acetyl-CoA levels in wild-type, Cgmed3abΔ, and Cgmed3abΔ/(Cgmed3ab)_OE strains. (B) Cells were cultivated in YNB medium for 48 h, washed three times with PBS, and added to prewarmed medium containing only 10 mM sodium acetate. The graph shows Cgcln3 mRNA levels measured every 5 min by RT-PCR and normalized to act1. (C) Cgcln3 mRNA levels in the Cgacs1Δ Cgacs2-ts strain at 25°C and 37°C following repletion with 1 mM acetate (cells were collected and added to 1 mM acetate and then grown at 30°C). (D) Cgcln3 mRNA levels in the Cgicl1Δ strain. (E) Median cell sizes (femtoliters [fl]) and percentages of budding were measured in the wild type and the CAGL0M11990gΔ strain. (F) Cgcln3 mRNA levels following in vitro regulation of the cellular acetyl-CoA level by adding 5 mM, 10 mM, and 15 mM acetate. (G) Cgmed3ab mRNA levels following in vivo regulation of the cellular acetyl-CoA level. The histograms represent cellular acetyl-CoA, and the circles represent the mRNA levels of Cgcln3. All experiments were repeated three times; the error bars represent SD. *, P < 0.05; **, P < 0.01.

Acetyl-CoA content affects the CAGL0M11990g transcriptional level.

The above-mentioned transcriptome data revealed that the cell growth and death module was one of the most notably differentially regulated pathways. We report that CAGL0M11990g, CAGL0I00286g, and CAGL0D02662g were significantly decreased (less than −3.0-fold change; P ≤ 0.05) in this module (see Data Set S3 in the supplemental material). Based on the above-mentioned results, we wondered whether acetyl-CoA activated transcription of the three genes. To investigate this hypothesis, exogenous acetate was added, and the transcript levels of the three genes were measured. The mRNA level of CAGL0M11990g rapidly increased after 10 min, whereas no significant changes were observed for CAGL0I00286g and CAGL0D02662g (Fig. 6B). Then, to confirm that acetate had been transformed to acetyl-CoA in vivo, the acs1Δ acs2-ts and iclΔ strains were constructed. In the acs1Δ acs2-ts strain, the CAGL0M11990g mRNA level was 48% lower at 37°C than at 25°C (Fig. 6C), suggesting that exogenous acetate had been transformed into endogenous acetyl-CoA. Meanwhile, the mRNA level of CAGL0M11990g remained unaltered in the icl1Δ strain (Fig. 6D), indicating that acetate had not been transformed into other metabolites by the glyoxylate cycle and gluconeogenesis. Therefore, we inferred that acetyl-CoA could induce transcription of CAGL0M11990g. Next, to determine the function of CAGL0M11990g, homologous alignment was applied to screen for similar sequences in the S. cerevisiae genome. CAGL0M11990g displayed 47.7% identity with an E value of 7.00E−99 to YAL040C, suggesting similarity between CAGL0M11990g and S. cerevisiae Cln3, a G₁ cyclin involved in cell cycle progression. To identify the function of CAGL0M11990g, the coding region of CAGL0M11990g was deleted, resulting in even larger cell size and decreased cell budding relative to the wild type (Fig. 6E). Given a function similar to that of ScCln3 (35), we identified CAGL0M11990g as CgCln3. Taking the data together, we concluded that CAGL0M11990g performed the same function as ScCln3 and that transcription of CgCln3 was induced by acetyl-CoA.

Finally, to determine whether the level of Cgcln3 mRNA in the Cgmed3abΔ strain was induced by the cellular acetyl-CoA level, in vitro and in vivo methods to regulate cellular acetyl-CoA were applied. First, 5 mM, 10 mM, and 15 mM acetate were added to boost endogenous acetyl-CoA levels. A higher acetate concentration led to higher Cgcln3 mRNA levels, with 15 mM acetate causing the highest expression (Fig. 6F). To validate the above-mentioned observation in vivo, expression of Cgmed3ab was quantified according to three strengths (strong, middle, and low) using a green fluorescent protein (GFP) reporter (see Fig. S1 in the supplemental material). Strong expression of Cgmed3ab coincided with 14% and 32% higher levels of acetyl-CoA than in middle or low expression, respectively. Moreover, the mRNA level of CgCln3 was 0.9-fold and 1.8-fold higher in the strong expression strain than in the middle and low expression strains, respectively (Fig. 6G). Taking the data together, we suggest that transcription of CgCln3 is regulated by the cellular acetyl-CoA content.

CgCln3 changes cell size and budding of C. glabrata.

Based on the above-described results, we analyzed the CgCln3 mRNA level in the Cgmed3abΔ strain and found a 34% decrease compared with the wild type (Fig. 7A). Furthermore, to investigate whether the decreased mRNA level of CgCln3 was the key reason for the observed cell growth defect in the Cgmed3abΔ strain, we assessed the cell size and cell budding index in the wild-type, Cgmed3abΔ, and Cgmed3abΔ/(Cgmed3ab)_OE strains. As shown in Fig. 7B, cell size in the Cgmed3abΔ strain was 15.1% lower than in the wild type, whereas that of the Cgmed3abΔ/(Cgmed3ab)_OE strain exhibited a 19.1% increase. Similarly, the budding index was 16.7% lower in the Cgmed3abΔ strain but 13.3% higher in the overexpression strain than in the wild type.

FIG 7.

CgCln3 changes cell size and budding. (A) Cgcln3 mRNA levels and cell sizes in the wild type and the Cgmed3abΔ strain. (B) Cell sizes were measured in the wild-type, Cgmed3abΔ, and Cgmed3abΔ/(Cgmed3ab)_OE strains. (C) Cells were treated with centrifugal elutriation and hydroxyurea, and the cell cycle was synchronized in G₁ phase. After release into fresh YNB medium, the budding index was determined by microscopic observation; the cell number was at least 200 cells. (D) Specific growth rates in the wild-type, Cgmed3abΔ, and Cgmed3abΔ/(Cgmed3ab)_OE strains. Each experiment was repeated three times; the error bars represent SD. *, P < 0.05; ***, P < 0.001.

Accordingly, we speculated that decreased cell size and cell budding index were the main factors responsible for the cell growth defect in the Cgmed3abΔ strain. The duration of the lag phase and the specific growth rate were analyzed based on the data shown in Fig. 2B. We found that the specific growth rate of the Cgmed3abΔ strain was 18.3% lower than that of the wild type, whereas that of the overexpression strain was 29.9% higher (Fig. 7D). Additionally, the duration of the lag phase was longer in the Cgmed3abΔ strain than in the wild type, whereas it was shorter in the overexpression strain (Fig. 7D). Therefore, we suggest that decreased CgCln3 gene expression could extend the lag phase and slow the specific growth rate in the Cgmed3abΔ strain and ultimately result in reduced cell growth.

DISCUSSION

Cellular metabolism is fundamental as a source of energy, building blocks, and biomass. In this study, we show that CgMed3AB regulates cellular acetyl-CoA metabolism by altering the occupancy of CgIno4 on the Cgacs2 promoter, the mRNA level of Cgacs2, and the activity of acetyl-CoA synthetase. Consequently, acetyl-CoA induces transcription of Cgcln3, as indicated by a decrease in the mRNA level of Cgcln3 and a cell growth defect following CgMed3AB deletion. These data indicate that CgMed3AB could control the cellular acetyl-CoA level, then coordinate the homeostasis between acetyl-CoA metabolism and transcription of Cgcln3, and finally regulate cell growth.

Med3 proteins in S. cerevisiae, C. albicans, and C. glabrata have various physiological functions (12, 21, 26). ScMed3 is related to various stress responses, including H₂O₂ (12), rapamycin (12), and osmotic stress (26), whereas CaMed3 determines the nuclear localization of TLOs and correlates with virulence of C. albicans (21). In addition, CgMed3 helps cells tolerate acid stress (13). Although ScMed3, CaMed3, and CgMed3 are orthologues, the amino acid sequence of CgMed3 shares only 35.7% and 30.7% similarity with those of S. cerevisiae and C. albicans, respectively. ScMed3 responds to stress by modulating subunit composition (12) and stress-related gene expression. CgMed3 regulates the acid stress response by changing the membrane lipid composition (13). The difference between ScMed3 and CgMed3 stress response mechanisms may be attributed to a discrepancy between their amino acid sequences. In addition, CaMed3 could facilitate filamentous growth and then regulate C. albicans pathogenicity. Together, these different roles of Med3 suggest that it is a multifunction protein, affecting diverse physiological processes. In this study, deletion of Cgmed3ab led to a decrease in biomass and viability under various conditions, whereas its overexpression increased cell growth (Fig. 2).

Specifically, we demonstrate that CgMed3 can regulate acetyl-CoA synthetase activity, thus altering cellular acetyl-CoA levels. Deletion of Cgmed3ab resulted in the differential expression of genes involved mostly in carbohydrate metabolism. These data are consistent with previous findings, where lack of Med3 modulated expression of carbohydrate metabolism-related genes (27). Furthermore, the present results suggest that genes participating in acetyl-CoA-related metabolism were mainly downregulated, and RT-PCR analysis showed acs2 mRNA was the most downregulated. This reduction was accompanied by a decrease in acetyl-CoA synthetase activity. In the presence of glucose, ACS1 is repressed and ACS2 takes over the essential function of acetyl-CoA synthetase (36). The present results are consistent with a published study showing that inactivation of Acs2p limited acetyl-CoA synthesis and caused histone deacetylation (37). Acetyl-CoA synthetase is an essential source of cellular acetyl-CoA (38), and its activity maintains cellular acetyl-CoA metabolism homeostasis in bacteria (39), yeast (40), and mammals (41). In our study, we saw both a decrease in acetyl-CoA synthetase activity and a reduction in cellular acetyl-CoA. Although there is no direct evidence for which intracellular pool of acetyl-CoA is changed, the reduced transcription of acs2, which encodes the nuclear acetyl-CoA synthetase (37), implies that the nucleocytosolic pool of acetyl-CoA might be the most reduced. Several studies have revealed that cellular acetyl-CoA metabolism could be controlled by transcriptional regulators, such as the lysine acetyltransferase NuA4 (42), the nitrogen response regulator GlnR (39), and the stress response regulator Msn2/Msn4 (43). Here, we report that the cellular acetyl-CoA level was regulated by subunits of the mediator complex CgMed3.

Deletion of Cgmed3ab reduced the cellular acetyl-CoA level, whereas CgMed3AB overexpression increased it. Acetyl-CoA is the acetyl donor for protein acetylation and provides a link between cellular acetyl-CoA metabolism and histone acetylation (44, 45). Several studies have proposed that acetyl-CoA could induce expression of growth-related genes by promoting histone acetylation (46, 47). In this study, we identified the orthologue of ScCln3 in C. glabrata, CgCln3. Exogenous acetate induction assays and endogenous quantitative-expression assays demonstrated that acetyl-CoA could induce the transcriptional process of CgCln3. Acetate is the key factor in exogenous acetate induction, whereby it is converted into acetyl-CoA through acetyl-CoA synthetase. This transient effect was used to determine whether acetyl-CoA induced the transcription of CgCln3. As lack of CgMed3AB decreased the cellular acetyl-CoA level, we suggest that CgMed3AB can indirectly induce transcription of CgCln3 by regulating the cellular acetyl-CoA level. Thus, the Cgcln3 mRNA level decreased following a decline in cellular acetyl-CoA in the Cgmed3abΔ strain. Acs2 was the main factor responsible for inducing transcription of Cgcln3, as indicated by its mRNA level being lower in the acs2Δ than in the acs1Δ mutant (Fig. 6D). To further validate the relationship between Cgacs2 and cell growth, we constructed a Cgacs2 overexpression strain and determined whether it could recover the growth defect of the Cgmed3abΔ strain. After overexpressing Cgacs2, the activities of acetyl-CoA synthetase and cellular acetyl-CoA were 76.3% (Fig. 5A) and 208.5% (Fig. 6A) higher, respectively, than those in the Cgmed3abΔ strain, and the mRNA level of Cgcln3 was 2.9-fold higher (Fig. 7A). The cell size and budding index were both rescued after overexpression compared to those of the Cgmed3abΔ strain (see Fig. S2 in the supplemental material). Notably, the cell growth defect of the Cgmed3abΔ strain was rescued by overexpressing Cgacs2 (Fig. 2). The G₁ cyclin CLN3 is related to the START checkpoint in the cell cycle, which responds to the attainment of a critical cell size (48). In our study, a decline in Cgcln3 mRNA caused a smaller cell size. In budding yeast, the critical cell size is an essential requirement for G₁/S transition, and smaller cells spend a longer time in G₁ than larger ones (49, 50). In addition, in the present study, the budding index was also decreased in the Cgmed3abΔ strain, a phenomenon previously linked to absence of ScCln3 (51) and cell cycle progression delay in G₁ phase (52). Fluorescence-activated cell sorting (FACS) analysis showed no significant difference in cell cycle progression (see Fig. S3 in the supplemental material), although it should also be noted that it might have been unable to distinguish early S phase from G₁ phase (53). In this study, we found that a decreased CgCln3 mRNA level caused cell budding delay, which could be the main reason for the observed cell growth defect in the Cgmed3abΔ strain. Single cells divide by budding, which leads to the formation of a colony. Our data indicate that decreased cell size and cell budding could cause a lower specific growth rate and longer duration of log phase (Fig. 7D). The reduction in the Cgcln3 mRNA level weakens cell growth, in line with previous findings reporting that CaCln3 could control the transition between proliferation and hypha formation (54). In conclusion, our study suggests that CgMed3AB can regulate cell growth by coordinating the homeostasis between acetyl-CoA metabolism and CgCln3, making CgMed3 a novel possible target for cell growth regulation.

MATERIALS AND METHODS

Strains and media.

The strains used in this study are listed in Table 1, and all the strains were manipulated by standard genetic techniques under a C. glabrata ATCC 55 background. The temperature-sensitive strain acs2-ts was constructed following a previously described protocol (56), and the other knockout strains were constructed by homologous recombination (13). The gene expression strengths of Cgmed3ab were divided into three levels: strong in the TEF promoter of plasmid pY26, middle in the GPD promoter of plasmid pY26, and low in the native promoter of Cgmed3ab.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Relevant characteristics	Reference or source
Strains
C. glabrata ATCC 55	his3Δ trp1Δ ura3Δ	55
ATCC 55 Cgmed3abΔ	his3Δ trp1Δ ura3Δ Cgmed3a::His3 Cgmed3b::Trp1	13
Cgmed3ab/overexpression-strong level	his3Δ trp1Δ ura3Δ pY26-PGPD/Cgmed3a Cgmed3b	This study
Cgmed3ab-middle level	his3Δ trp1Δ ura3Δ pY26-PTEF/Cgmed3a Cgmed3b	This study
Cgmed3ab-low level	ATCC 55Cgmed3abΔYEplac112-Pmed3a/med3a Pmed3b/med3b	This study
Cgacs1Δacs2-ts	ATCC 55 Cgacs1::His3 Cgacs2::Trp1 acs2-ts::Ura3	This study
Cgicl1Δ	his3Δ trp1Δ ura3Δ Cgicl1::His3	This study
CgCAGL0M11990gΔ	his3Δ trp1Δ ura3Δ CgCAGL0M11990gΔ::His3	This study
ATCC 55 pY26	his3Δ trp1Δ ura3Δ pY26	This study
ATCC 55 Cgmed3abΔ CgACS2OE	his3Δ trp1Δ ura3Δ Cgmed3a::His3 Cgmed3b::Trp1 pY26-PGPD/CgACS2	This study
ATCC 55 pY26-GPD-GFP	his3Δ trp1Δ ura3Δ pY26-PGPD/GFP	This study
ATCC 55 pY26-TEF-GFP	his3Δ trp1Δ ura3Δ pY26-PTEF/GFP	This study
ATCC 55 Yeplac112 -GFP	his3Δ trp1Δ ura3ΔYEplac112-Pnative-GFP	This study
AH109	trp1Δ leu2 ura3Δhis3Δ gal4Δ gal80Δ lys2::GAL1UAS-GAL1TATA-HIS3 GAL2UAS-GAL2TATA-ADE2 URA3::MEL1UAS- MEL1TATA-LacZ MEL1	Clontech
Rosetta-DE3	For enzyme expression and purification
Plasmids
pY26	2μm Amp URA3 PGPD PTEF	Turbo
pY26(PGPD)-CgACS2	2μm Amp URA3 PGPD-CgACS2 PTEF
pY26(PGPD)-GFP	2μm Amp URA3 PGPD-GFP PTEF	This study
pY26(PTEF)-GFP	2μm Amp URA3 PGPD PTEF-GFP	This study
YEplac112(Pmed3a)-GFP	2μm Amp TRP1 Pmed3a-GFP	This study
pY26(PGPD)-MED3AB	2μm Amp URA3 PGPD-med3a med3b PTEF	This study
pY26(PTEF)-MED3AB	2μm Amp URA3 PGPD PTEF-med3a med3b	This study
YEplac112	2μm Amp TRP1	YouBio
YEplac112(Pmed3a/3b)-MEd3AB	2μm Amp TRP1 Pmed3a-med3a Pmed3b-med3b	This study
pKL187	Amp URA3 R-DHFR(heat-inducible degron)	EuroScarf
pGBKT7	Kan TRP1 GAL4 DNA-BD fusion	Clontech
pGADT7	Amp LEU2 GAL4 DNA-BD fusion	Clontech
pET28a	Kan pBR322 ori PT7 6His T7(Ter)

Yeast cells were cultivated in YPD medium (1% yeast extract, 2% tryptone, 2% dextrose) and YNB medium.

Spot assays and viability assays.

Cells were cultivated to log phase and diluted to an absorbance at 660 nm (A₆₆₀) of 1.0. Then, 10-fold serial dilutions of the cells were spotted on YNB medium under different stress conditions. Finally, the cells were incubated for 4 days at 30°C. The method to analyze cell viability followed that in our previous study (13).

Growth assay.

To test cell growth curves, cells were collected in log phase and then released in fresh YNB medium with an initial A₆₆₀ of 0.1. We tested the optical density of the cells every 2 h. The A₆₆₀ was converted to the cell dry weight (CDW) according to the following formula: A₆₆₀/CDW = 1/0.23 g/liter.

Transcriptomic analysis by RNA-Seq.

The wild-type, Cgmed3aΔ, Cgmed3bΔ, and Cgmed3abΔ strains were grown to early log phase and washed three times with diethylpyrocarbonate (DEPC)-treated water. Then, the cells were collected by centrifugation at 5,000 × g for 5 min at 4°C. The total RNA was isolated using a MiniBest universal RNA extraction kit (TaKaRa Bio, Shiga, Japan). The global gene analysis was performed based on a method described previously (57). The RNA-Seq data are available at https://www.ncbi.nlm.nih.gov/sra/SRX1528065[accn]. Differential expression analysis (DEGs) was performed using the cutoffs of an adjusted P value of ≤0.05 and a 1.5-fold change. Annotation and Gene Ontology (GO) information were based on the Candida Genome Database (CGD) and the KEGG database, respectively.

Quantitative real-time PCR analysis.

Yeast cells were grown at log phase, and the cells were collected by centrifugation at 5,000 × g for 5 min at 4°C. Total RNA extraction has been described using the method of transcriptomic analysis by RNA-Seq. cDNA was synthesized using a PrimeScript II first-strand cDNA synthesis kit (6210A; TaKaRa Bio). Finally, quantitation of the mRNA level was performed using SYBR Premix Ex Taq (RR420A; TaKaRa Bio). ACT1 was used as a standard control to normalize gene expression. Each experiment was repeated three times.

Yeast cell synchronization.

To synchronize the yeast cells in G₁ phase, we obtained small G₁ cells by centrifugal elutriation using a method based on that of Rosebrock (58). The cells were released in prewarmed YPD medium at a concentration of about 1 × 10⁷ cells ml⁻¹.

Cell size and cell budding index measurements.

Cells were grown asynchronously to early log phase in YNB medium. In order to obtain an accurate cell size distribution, the culture was briefly sonicated in a moderate way, and the cells were analyzed with a Coulter Z2 particle cell analyzer (Beckman-Coulter). For cell budding experiments, the budding of at least 200 yeast cells was observed with a microscope.

Acetate induction assay.

For the Cln3 transcriptional-activation assay, cells were cultured for 48 h to stationary phase and washed two times with phosphate-buffered saline (PBS). We then added the cells to prewarmed medium, which contained only 10 mM sodium acetate, and the mRNA level of Cgcln3 was measured every 5 min. The acetyl-CoA level in vivo was simulated by adding 5 mM, 10 mM, and 15 mM acetate. mRNA extraction and the quantitation of mRNA levels were as described above.

Yeast two-hybrid assays.

The yeast two-hybrid assays were performed using a Matchmaker library construction and screening kit (Clontech). The bite-BD (pGBKT7-Cgmed3a and pGBKT7-Cgmed3b) and prey-AD (pGADT7-Cgino4) fusion protein plasmids were constructed by standard genetic techniques (Table 2). The combinations of bite-BD and prey-AD fusion protein plasmids were cotransformed into the yeast strain AH109 (Clontech). Then, potential positive transformants were plated on synthetic dropout (SD)/Leu⁻ Trp⁻ His⁻ Ade⁻ medium, and colony PCRs were used to select the positive clones. Finally, the positive clones were plated on SD/Leu⁻ Trp⁻ His⁻ Ade⁻ medium containing X-α-Gal (5-bromo-4-chloro-3-indolyl α-d-galactopyranoside).

TABLE 2.

Primers used for plasmid construction in this study

^aRestriction sites are underlined.

Coimmunoprecipitation assays.

One milligram of total proteins was extracted from wild-type cells with CgMed3A-HA/CgIno4-Myc (cells with CgMed3A containing a hemagglutinin tag and CgIno4 containing a c-Myc tag) and CgMed3B/CgIno4 (cells with CgMed3B containing a hemagglutinin tag and CgIno4 containing a c-Myc tag), respectively; the protein extraction buffer contained 50 mM Tris·HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 10% (vol/vol) glycerol, 0.1% Tween 20, 1 mM phenylmethylsulfonyl fluoride (PMSF), and 1× complete protease inhibitor mixture (Sangon Biotech). The extracts were incubated with 25 μl anti-hemagglutinin (HA)-conjugated magnetic beads (Bio-Rad) for 6 h at 4°C. Then, the magnetic beads were washed three times with protein extraction buffer. Next, the precipitates were eluted into 100 mM glycine (pH 2.5) and 100 mM NaCl and immediately neutralized with 2 M Tris-HCl (pH 9.0) and 100 mM NaCl, and the immunoblot analysis was performed.

Electrophoretic mobility shift assay.

The EMSA method was as described previously (59) with slight modifications. First, the CgMed3AB protein was extracted and purified as described previously (60). The core promoter fragments of interest were cloned from the genome of C. glabrata ATCC 2001 by PCR and the probe DNA was decorated with Cy3 at the 3′ end. The PCR fragments were purified by electrophoresis in a 1% (wt/vol) agarose gel. CgMed3AB protein and promoter fragments were incubated in 10× binding buffer (100 mM Tris [pH 7.5 at 20°C], 10 mM EDTA, 1 M KCl, 1 mM dithiothreitol [DTT], 50% [vol/vol] glycerol, 0.10 mg/ml bovine serum albumin [BSA]) at 20°C for 30 min. At the end of the equilibration period, 30-μl samples were loaded and resolved in a 6% (wt/vol) native polyacrylamide (75:1 acrylamide-bisacrylamide) gel. The electrophoresis was conducted at a constant power of 100 V in 1× Tris-acetate-EDTA buffer (10× Tris-acetate-EDTA electrophoresis buffer; 400 mM Tris, 25 mM EDTA, brought to pH 7.8 with acetic acid).

Chromatin immunoprecipitation.

The ChIP assays were performed as described previously (61). Briefly, cultures were grown in YNB medium to an optical density at 660 nm (OD₆₆₀) of 1, and 1% formaldehyde was added to the culture to cross-link chromatin and protein. Cells were collected by centrifugation, washed three times, resuspended in lysis buffer (50 mM HEPES-KOH [pH 7.5], 150 mM NaCl, 1 mM EDTA, 0.1% sodium deoxycholate, 0.1% SDS, 1% Triton X-100, 1 mM PMSF), and lysed with a bead beater. In order to generate chromatin fragments with a mean size of 300 bp, a sample was sonicated. Magnetic beads (161-4821; Bio-Rad) conjugated with anti-HA (ab9110; Abcam) were added to the sample and incubated at 4°C for 3 h. The magnetic beads were washed twice with lysis buffer, twice with wash buffer (10 mM Tris-HCl [pH 8.0], 0.25 M LiCl, 1 mM EDTA, 0.5% NP-40, 0.5% sodium deoxycholate), and once with TE solution (10 mM Tris-HCl [pH 8.0], 1 mM EDTA); elution buffer was added to the washed beads, and they were eluted at 65°C overnight. Finally, DNA was phenol extracted and resuspended in TE solution. The occupancy was analyzed by quantitative real-time PCR (CFX96 real-time system; Bio-Rad). The specific primers are listed in Table 3.

TABLE 3.

Primers used for RT-PCR and ChIP in this study

Primer	Sequence (5′–3′)	Description
Cln3-F	GCCTCTTCAGCATTCTCTT	RT-PCR
Cln3-R	TTAGCAGTGTGATTAGCATTATTG	RT-PCR
Acs1-F	ATTGATTGTCTTGGTTGA	RT-PCR
Acs1-R	CAGGATTGGATAAGGTAG	RT-PCR
Acs2-F	TATGAAGCAGACGACGAGAA	RT-PCR
Acs2-R	AGATACCAGCAACCTTAGATACT	RT-PCR
Erg10-F	GTGAACAACAAGATGACTT	RT-PCR
Erg10-R	ATGGTAACTGGAACGATT	RT-PCR
Pot1-F	GGACGATGTTGTTATTGTT	RT-PCR
Pot1-R	AGCAGATAGTCAGTGTTG	RT-PCR
Pda1-F	GAACTACTCAGCCACTAAG	RT-PCR
Pda1-R	GCCATCTCCATTCTTCTG	RT-PCR
Pdb1-F	TCAAGAAGACTAACCATT	RT-PCR
Pdb1-R	CCTCAGATTCCATAACTT	RT-PCR
Lat1-F	TACAGCCAAGGACCAATA	RT-PCR
Lat1-R	TGAAGCCAATAAGAGTTAGC	RT-PCR
Acc1-F	TCTTCTGATGCCTTGATAG	RT-PCR
Acc1-R	ATGTTCGTGTGCTTCTAA	RT-PCR
Cit1-F	AACTACTACTGGAACAAG	RT-PCR
Cit1-R	TCTACCTCTGAATCTAATAC	RT-PCR
Mls1-F	TGGAGTTCATTGTCTTGTTG	RT-PCR
Mls1-R	AAGTTGCCGCTGTCTAAT	RT-PCR
Acs2-Core-F	GAAGCGTTATTGCCGATAT	ChIP
Acs2-Core-R	AACCAGTGACGAAGAAGT	ChIP
ChrV-Up	GGCTGTCAGAATATGGGGCCGTAGTA	ChIP
ChrV-Down	CACCCCGAAGCTTTCACAATAC	ChIP

Preparation of cell extracts and enzyme activity assay.

Cells were harvested in log phase, and cell extracts were prepared for the determination of enzyme activity (62). The activity of ACS was analyzed by measurement of the absorbance at 340 nm as described previously (39). The reaction mixture contained 100 mM Tris-HCl (pH 7.7), 50 mM l-malate, 20 mM ATP, 50 mM MgCl₂, 2 mM CoA, 60 mM NAD⁺, 50 U/ml malate dehydrogenase, and 25 U/ml citrate synthase. The reaction was started by adding 100 mM potassium acetate. One unit of ACS activity was determined to correspond to 1 mmol of NADH formed per minute, and the final data were normalized by CDW.

Fluorescence intensity determination.

Cells were grown to log phase and diluted to the same A₆₆₀. Then, cells were harvested, washed, and suspended in PBS, and the fluorescence intensity was measured as described previously (63).

Cellular acetyl-CoA level measurement.

For cellular acetyl-CoA measurement, cells were grown to early log phase. To quench the cell metabolism, 2 ml of cells was injected rapidly into 5 ml prechilled (−80°C) methanol, and the cells were collected by centrifugation at 4,000 × g for 5 min at 4°C. Then, 2 ml boiling ethanol was added to the cell pellets, and the mixture was boiled for another 15 min to release the intercellular metabolite. Next, the mixture was centrifuged at 12,000 × g for 10 min at 4°C, and the supernatant was retained. Finally, after vacuum drying, 200 μl double-distilled H₂O (ddH₂O) was added to resuspend the metabolite. The analysis of acetyl-CoA was performed by liquid chromatography-tandem mass spectrometry (LC–MS-MS) (64).