Abstract
The budding yeast, Saccharomyces cerevisiae is one of the most studied organisms used for the synthesis of products, to explore the human diseases and eukaryotic gene expression mechanisms. The yeast cells with flocculation property are in high demand for industrial applications. However, the pathogenic yeast becomes drug-resistant due to flocculation/biofilm phenotype. The flocculation property of yeast depends on the expression of specific FLO genes. Genetic and epigenetic factors have been suggested to induce the expression of FLOs and flocculation, an evolutionarily conserved process. The present study was undertaken to identify a molecular link between stress caused by genetic and epigenetic factors and expression of FLOs. We utilized flocculating yeast strains to study the regulation of FLO genes and flocculation phenotype. We found rough surface morphology and constitutive activation of Slt2 in flocculating cells. The external cell wall stress factors as well as specific mutations in Sen1 and histone proteins strongly correlated with the induction of FLO genes whereas deletion of SLT2/RLM1, suppressed the expression and flocculation phenotype. We detected constitutive binding of Rlm1 and eviction of Tup1 from the promoters of FLO1 and FLO5 genes in flocculating cells. Thus we provide evidence for the CWI pathway dependent flocculation of yeast, regulated by Sen1 mediated interplay between Tup1 and Rlm1.
1. Introduction
Gene expression mechanisms in eukaryotes are tightly regulated by complex interplay between regulatory factors (Smith and Shilatifard 2010; Smith and Workman 2012). Often regulatory factors compete for their binding to genes. The cooperation among the elements of transcriptional machinery maintains cellular homeostasis. For survival, it is essential to generate an appropriate transcriptional outcome in response to extrinsic and intrinsic factors (Mcmurray and Tainer 2003; Thorsenet al. 2006; Vihervaaraet al. 2018). Living organisms are constantly exposed to a wide variety of agents which can induce genetic and epigenetic changes (Baccarelli and Bollati 2009). Mutations in activators or repressors of transcription apparatus have been linked with a variety of diseases; including cancer and developmental disorders (Lee and Young 2013). The survival in response to such perturbations is dependent on coordination among factors of specific molecular signaling pathways and transcriptional mechanisms (Taymaz-nikerelet al. 2016). For example, upon exposure to the cell wall (Garciaet al. 2004; Levin 2005) and DNA damaging agents (O’Connor 2015), specific cell wall sensor proteins transmit the signal to nucleus through ‘Cell Wall Integrity’ (CWI) (Jung and Levin 1999) and genomic integrity signaling pathways which subsequently triggers the de-repression of cell wall and DNA repair genes respectively. In the absence of DNA damage, transcription of repair genes is turned off by the recruitment of repressor complex leading to the formation of transcription resistant chromatin structure (Smith and Johnson 2000).
Yeast mitogen-activated protein kinase (MAPK) signaling pathways transduce external signals via Slt2/Mpk1 and Hog1 kinases, generates appropriate cellular responses (Chen and Thorner 2007; Gustinet al. 1998) which are required for survival under cell wall and osmotic stress respectively (Bermejoet al. 2008; De Nobelet al. 2000; Levin 2011). Zymolyase treatment induces the phosphorylation of Slt2 and Hog1 to activate the expression of the CWI pathway genes (Garciaet al. 2009). Under cell wall stress conditions, sensor proteins activate a kinase cascade; Rom2-Rho1-Pkc1 which integrates into the MAP Kinase module which composed of Bck1, Mkk1, and Mkk2 proteins. The active form of Bck1 activates two kinases; Mkk1 and Mkk2, which further phosphorylates a downstream MAPK, Slt2 (Kimet al. 2010). The phosphorylated form of Slt2 subsequently enters the nucleus and activates two transcription factors; Rlm1 and SBF (Garciaet al. 2016; Maddenet al. 1997). For example; the transcriptional induction of one of the CWI genes occurs in response to cell wall damage through the Slt2 activation-dependent recruitment of Rlm1, SWI2/SNF2 and SAGA complexes (Sanzet al. 2012; Sanzet al. 2016). Under non-inducing conditions, transcription of CWI genes is terminated by the binding of Nrd1-Nab3-Sen1 (NNS) complex. However, under cell wall stress conditions, Mpk1 mediates the up-regulation of CWI genes by blocking recruitment of NNS transcription termination complex (Kim and Levin 2011). Furthermore, the significance of MAPK pathway has also been implicated in the filamentous growth of pathogenic yeast due to up-regulation of cell-cell adhesion/flocculation genes (Chavelet al. 2014; Chavelet al. 2010). Although above studies suggest that the expression of CWI and flocculation genes are probably regulated through a common MAP kinase signaling dependent mechanism, however, a molecular link between these two is yet to be established.
The genetic, epigenetic and environmental factors have been shown to induce flocculation phenotypes (Kim and Rose 2015; Liuet al. 2015; Soares 2011) such as mutations in certain genes, nutritional stress, temperature, pH, oxygen supply and alteration in cell wall composition. The flocculation or aggregation of cells is also a property of yeast and bacteria to form biofilm and become resistant to drugs (Chandraet al. 2001). Yeast also forms multi-cellular flocs/clumps under adverse growth or particular physiological conditions such as lipid unsaturation and limited oxygen supply (Degreifet al. 2017). Induction of flocculation protects cells from such stressful environmental situations and allows them to survive for long (Hopeet al. 2017). There are six major FLO genes in yeast; FLO1, FLO5, FLO8, FLO9, FLO10, and FLO11, which are responsible for flocculation and biofilm formation. The role of Flo1 is quite well established in flocculation, bio-film like phenotype of yeast (Goossenset al. 2011; Simet al. 2013; Smukallaet al. 2008). The Flo8p is considered as a transcription factor to drive the expression of FLO11 (Fichtneret al. 2007), and significance of FLO11 has been implicated in the mating process of yeast (Guoet al. 2000) and invasive growth.
A transcription factor, Mss11 is also known to regulate the expression of FLO10 and FLO11 (Bester et al. 2012) essential for cellular adhesion phenotype. The expression of FLO11 is linked with mutations in certain genes leading to flocculation of yeast. The FLO11-associated phenotypes such as biofilm formation and adhesion to polystyrene are inhibited by L-Histidine through altering the chitin and glycan content on the cell wall of flor yeasts (BOU ZEIDAN et al. 2014). Similarly, the deletion of the ribosomal RPL32 gene in fission yeast (Liuet al. 2015) and deletion mutants of the COMPASS complex (Histone H3 methyltransferase) in budding yeast induces the flocculation phenotype (Dietvorst and Brandt 2008).
Sen1p is a member of Nrd1-Nab3-Sen1 (NNS) complex, primarily involved in termination and processing of non-coding transcripts through the process of nuclear exosome complex mediated degradation of RNA (Chinchillaet al. 2012; Finkelet al. 2010). Role of Sen1p has also been implicated in genomic integrity by preventing co-transcriptional R-loop formation (Cohenet al. 2018; Grunseichet al. 2018). Defects in Sen1p are associated with down-regulation of DNA repair genes and sensitivity to DNA damaging agents (Gollaet al. 2013). Furthermore, significant alteration in the expression of lipid homeostasis pathway genes is reported (Sarikiet al. 2016) in Sen1 mutants. Moreover, many mutations in Senataxin, the human homolog of yeast Sen1 are correlated with genetic disorders, ALS4, and AOA2 (Grohet al. 2017). However, the underlying mechanism is yet to be elucidated for the development of disease pathogenesis. All the fungal cell wall adhesins/flocculins, as well as membrane adhesion proteins of higher eukaryotes, are Gpi-anchored (Sahaet al. 2016; Verstrepen and Klis 2006). In fungus, these glycosyl-phosphatidylinositol (GPI) anchored glycoproteins are essential for invasive growth, cell-cell adhesion, and mating (Guoet al. 2000). The conserved amyloid-like amino acid sequences are also found in yeast cell adhesins; Flo1p and Flo11p (Lipkeet al. 2012; Rameauet al. 2016; Ramsooket al. 2010) which suggest that cellular aggregation by amyloid sequence containing proteins is the probable reason for neurological diseases in humans, and their localization on the yeast cell wall contributes in cell-cell aggregation, leading to biofilm formation.
Recently, upregulation of Flo1p has been observed in Rpb11 and Sen1 mutant yeast cells (Chenet al. 2017) supporting our earlier observations. However, the molecular mechanism has not been explored. The deletion of TUP1 in budding yeast induces strong flocculation phenotype (Lipke and Hull-Pillsbury 1984). The yeast FLO1 and DNA repair genes are repressed by a global repressor Cyc-Tup1 complex along with histone deacetylases; Hda1 and Rpd3 (Churchet al. 2017; Zhang and Reese 2004). During DNA damaging conditions or in the absence of Cyc-Tup1 repressor, the occupancy of Hda1 and Rpd3 decreases whereas occupancy of Swi/Snf increases (Fleminget al. 2014; Fleming and Pennings 2001). Thus the genes of DNA repair and flocculation are regulated by the antagonistic actions of chromatin remodeling and repressor complexes (Fleming and Pennings 2001).
We have earlier reported strong flocculation phenotype of a few Sen1 mutant strains of Saccharomyces cerevisiae (Singhet al. 2015) as well as down-regulation of SOD1 (Sarikiet al. 2016). Another study has shown higher chitin content in the cell wall of sod1∆ cells (Liuet al. 2010) indicating the role of Sen1 in cell wall maintenance. However, why do mutations in Sen1 result in flocculation phenotype is not clear. In the present study through our extensive investigations, we have established a Sen1 regulated novel link between the CWI pathway and flocculation. We utilized flocculating mutant strains of Sen1 and histones as well as tup1∆ cells for the studies. We found that flocculating cells exhibit rough surface morphology, enlarged cell size, higher cell wall chitin content, sensitivity to cell wall perturbing agents and constitutive activation (phosphorylation) of Slt2. Our careful analysis of upstream promoter sequences of FLO genes suggested a binding site for Rlm1. Indeed through chromatin immunoprecipitation assays; we revealed constitutive binding of Rlm1, TBP, Pol-II and the loss of Tup1 occupancy at the FLO1 and 5 genes in flocculating Sen1 mutant cells. The growth of flocculating cells was impaired in the presence of cell wall damaging agents probably due to alteration in cell wall chitin content. The phosphorylation of Slt2 and de-repression of FLO genes was also found in wild-type strain in response to stress conditions such as temperature and cell wall damaging agent. Moreover, the drastic decrease in basal level expression of FLO1, 5, 9, 10 genes in deletion mutants of CWI pathway was observed which further suggest that CWI pathway is required for the regulation of FLO genes and flocculation phenotype. The mutations in Sen1p and histones, and exposure to external stress factors, resulted in transcriptional induction of FLOs. Interestingly the expression of FLO genes was significantly reduced after the deletion of SLT2 and RLM1, leading to suppression of flocculation phenotype. Thus we provide evidence that the CWI pathway plays an essential role in the regulation of yeast flocculation. Altogether for the first time, we established the role of ‘Cell Wall Integrity’ signaling pathway in yeast Flocculation, regulated by Sen1 mediated interplay between Tup1 and Rlm1.
2. Materials and Methods
2.1. Strains, chemicals, growth media and conditions
The Saccharomyces cerevisiae strains used in this study were in the BY4741 background and had been listed in Table S1. The library of histone mutants was purchased from Dharmacon, Cat# YSC5106. The diploid gene deletion library mutants were purchased from Open Biosystems, Cat# YSC1056. Gene tagging and deletions were performed by polymerase chain reaction (PCR) based methods (Jankeet al. 2004; Wach 1996) using primers listed in Table S2. For the selection of positive colonies, transformants were plated on YPD medium containing G418 sulfate (300μg/ml), and the deletions were verified by PCR using ORF-specific primers. All gene deletions and tagging were confirmed by either PCR or western blotting. SLT2 deletion was confirmed by western blotting using Anti-Mpk1 antibody and PCR with SLT2 ORF primer. RLM1 deletion was confirmed by PCR using RLM1 primer. Western blot was conducted using an Anti-Myc antibody to verify the genomic Myc tagged Rlm1.
Chemicals used in the experiments were mostly purchased from Sigma Aldrich unless otherwise mentioned. Yeast strains were grown in SC (synthetic complete) media. For making SC media, all amino acids, yeast nitrogen base (YNB), ammonium sulfate and glucose were mixed as per the standard laboratory protocol (Amberget al. 2005). All Yeast strains used in this study were grown at 30°C unless otherwise stated.
2.2. Growth sensitivity Assay
Spot assays were performed to investigate the effect of cell wall perturbing agents on the growth of WT and mutants. Mid-log phase cultures of wild-type and mutant yeast cells were 10 fold serially diluted. 3 μl of serially diluted cultures were spotted onto solid Synthetic Complete (SC) Agar plates with or without cell wall perturbing agents; Calcofluor White (CFW), Congo Red (CR) and Caffeine. All plates were incubated at 30°C and growth of the yeast strains was recorded at 72 h by scanning the plates using an HP scanner. For growth curve analysis, exponentially growing cells were seeded in a 96-well cell culture plate in triplicates and incubated in liquid growth media at 30°C with different concentrations of CFW. The OD600 at regular time intervals using an automatic plate reader (Eon Microplate
Spectrophotometer) was recorded, and the data were analyzed using GraphPad Prism 5.0 software.
2.3. Colony Forming Units
The growth of mutant strains of NNS, TRAMP/Exosome and Histones were determined by CFU assay in the presence of Cell Wall damaging agents was measured by counting CFUs and plotted as plating efficiencies. Saturated cultures of the above-mentioned cells (mutants and respective wild-type cells) were serially diluted, and aliquots containing 1,000 cells were plated on SC Agar plates with and without varying concentrations of Cell Wall damaging agents; CFW (50μg/ml), CR (10μg/ml) and Caffeine (6mM) and incubated at 30°C. Percent cellular viability was determined as the portion of cells that formed viable colonies after 3 days of incubation. This assay was conducted three times. The plating efficiencies were calculated as % of colonies viable from all plated cells.
2.4. Microscopy
Images of cells were captured by using an Apotome Carl Zeiss fluorescence microscope through Apochromat 63× objective lens. For the analysis of the cell size of wild-type and mutant strains, cells were grown at 30°C in synthetic complete liquid medium till exponential phase, harvested, washed twice with PBS, visualized and images were acquired using DIC filter.
2.5. Flow Cytometry
The intracellular chitin levels and cell size were measured using BD-FACS Aria III equipped with CELL QUEST software. To measure chitin content, WT and mutant strains were grown at 30°C in SC medium till exponential phase and subsequently stained with 50μg/ml CFW for 30 min at 30°C and thereafter cells were collected by centrifugation, washed twice with PBS, re-suspended again in PBS, and fluorescence intensity was measured using DAPI filter. The relative cell size was estimated by plotting the areas of forward (FSC) versus side (SSC) scattering signal parameters.
2.6. Scanning Electron Microscopy (SEM)
The SEM analysis of WT and mutants (sen1-1, sen1 ∆N, and sen1-K128E) by following a standard protocol was performed with slight modifications (Tanget al. 2014). In brief, 1 ml of overnight cultures were collected by centrifugation and washed twice with dH2O before prefixing the cells with 4% Paraformaldehyde (PFA) for 5 min at RT. After PFA fixation, cells were washed twice with 1X PBS and dehydrated serially in different concentrations of ethanol ranging from 30% - 100%, for 10 min in each at room temperature. After final resuspension in 100% ethanol, the cells were sonicated to remove the clumps. Subsequently, cells were diluted 10 times in 100% ethanol and sonicated again. 20 μl of the cells were placed on stubs using glass coverslip and kept for drying in a CO2 desiccator for overnight. After the overnight drying, the stubs containing cells were gold coated using sputter coater and visualized in HR FESEM (High-resolution field emission scanning electron microscope) with a resolution of 1.0 nm at 15 kV.
2.7. Immunoblotting Analysis
The 5 ml cultures of WT and different mutants of yeast were grown at 24°C until 1.5 OD600, harvested by centrifugation at low speed, washed once with 20% TCA and stored at −80°C freezer until they were used. The whole cell protein lysates were prepared from the frozen cells by 20% Trichloroacetic acid (TCA), and immunoblotting was conducted. In brief, cell pellets were re-suspended in 200 μl of 20% TCA and lysed in the presence of glass beads by vortexing for 20 min at room temperature, centrifuged for 10 min at 7K rpm, and pellets were collected and washed once with 1.0 ml 0.5M Tris-HCl, pH7.5. The washed pellets were re-suspended in 200 μl of 0.5M Tris-HCl, pH7.5; SDS loading dye was added, boiled for 5 min and analyzed by SDS-PAGE. Proteins were transferred onto Nitrocellulose membrane for western blotting. Blots were probed with following primary antibodies were used: Anti-Slt2-P/phospho-p42/44 MAPK (catalog No. 4370; Cell Signaling), Anti-Mpk1 (yC-20-catalog No. sc-6803; Santa Cruz), and Anti-Myc (9E10). The IR-Dye 800CW Anti-Rabbit IgG, Anti-Goat, and Anti-Mouse were used as secondary antibodies for respective primary antibodies. Polyclonal antibodies against recombinant TBP was raised in rabbit. Western blots were scanned by using an Odyssey infrared imager (LI-COR Biosciences).
2.8. Observation and analysis of flocculation
Overnight grown yeast cells were re-suspended to an equal cell density in SC media. Flocculation was observed using a method mentioned earlier (Churchet al. 2017) with some modifications. In brief, 3 ml of cell cultures of 2 OD600 were placed into a tissue culture plate. Cells were subsequently agitated by shaking and then left undisturbed. Images of plates were subsequently recorded by HP scanner at time 0, 15 and 30 minutes. The % Flocculation was measured by using a method described earlier (Bonyet al. 1998). Briefly, yeast cells were deflocculated by two washes with wash buffer (50mM sodium acetate, pH 4.5 and 5mM EDTA) and twice with distilled water. Cells were re-suspended in flocculation buffer (50mM sodium acetate, pH 4.5 and 5mM calcium chloride) and incubated for 60 min at 100 rpm in the incubator shaker at RT and then 6 mL of cell suspension was taken into culture tubes and kept vertically undisturbed for 10 min to allow cells for settling. The following equation determined percentage flocculation ability (F): F= (A-B/A) × 100% as described previously (Singhet al. 2015). ‘A’ indicates initial OD before shaking at time=0, and ‘B’ indicates final OD after 70 min of incubation (60 min in shaking + 10 min settling time). The percentage of flocculation is represented as mean of two independent biological replicates.
2.9. Isolation of total RNA and RT-qPCR
Total RNAs were extracted using a standard method (Schmittet al. 1990). The cDNA preparations were performed by using an iScript cDNA synthesis kit procured from BioRad. For PCR amplification, 2X SYBR green master mix was used. The conditions for PCR amplification were as follows: initial denaturation for 3 min at 95°C, PCR amplification for 40 cycles with denaturation for the 20s at 95°C, annealing for 20s at 58°C and elongation at 72°C for 20s in an ABI 7300 Real-Time PCR machine (Applied Biosystems). The relative expression levels (mRNAs) of each gene in wild-type and mutants were normalized by subtracting the β-actin threshold cycle (CT) values and the fold change (increase or decrease) was calculated through the 2-ΔΔCT method (Livak and Schmittgen 2001). Three independent experiments were performed, and each sample was run in triplicate. Primers used in this study are listed in Supplementary Table S2.
2.10. Chromatin Immunoprecipitation
The ChIP assays were performed as described earlier (Tomaret al. 2008). Cells were grown in 100 ml of SC growth media until 1.2 to 1.4 OD600. Subsequently formaldehyde was added (1% final concentration) for crosslinking by incubation for 10 min, and then 6ml of 2.5M glycine was added to stop the crosslinking reaction by incubation for 5 min at 25°C, cells were harvested by centrifugation, washed by water and stored in −80°C freezer until they were used. For preparation of soluble chromatin extracts, frozen cell pellets were lysed by vortexing for 30 min at 4°C in presence of glass beads in 1.2 ml of FA lysis buffer (50 mM HEPES, pH7.5, 150 mM NaCl, 2mM EDTA, pH8.0, 1% Triton X-100, 0.1% Sodium Deoxycholate and 0.1% SDS), glass beads were removed by centrifugation and final volume was made up to 1.8 ml by adding FA lysis buffer. The lysed cells were then sonicated to shear the chromatin into fragments averaging 200 - 400 bp using a Bioruptor (Diagenode) sonicator. The sonication conditions were as follows; high power, 8 cycles with the 30s ON and OFF in ice-chilled cold water. The soluble chromatin of sonicated cell lysates was collected by centrifugation at 14K RPM for 30 min at 4°C. For ChIP assay, 100 μl of chromatin extracts were incubated with 1μl of Anti-TBP, Anti-Tup1, Anti-Pol-II (ab817, 8WG16) and Anti-Myc (Rlm1) antibodies for overnight. The Anti-TBP and Anti-Tup1 antibodies were generous gifts from Dr. Joseph Reese. Multiple dilutions of input DNA were used to find out the amount of template DNA for PCR amplification to be in linear range. The immune-precipitated DNAs were amplified by RT-qPCR using SYBR Green master mix and ABI 7300 PCR machine. The following primer sets, specific for upstream regions of FLO1; FLO1P1 (−215/+14), FLO1P2 (−431/-312), FLO1P3 (−612/-457) and FLO1P4 (−817/-659) and FLO5; FLO5P1 (−139/+20), FLO5P2 (−413/-264), FLO5P3 (−621/-434) and FLO5P4 (−804/-595) were used. The STE6 amplicons were used for normalization, as an internal control. The relative fold change was calculated by using the 2-ΔΔCT protocol. The results presented here are the averages of three independent biological repeats.
2.11. Statistical analysis
Statistical analyses of the data were performed using GraphPad Prism 5.0 software of three independent repeats. Results of statistical analyses were presented as means ± standard deviations (SD). Differences between groups were tested using Student’s t-test, with P values of <0.05 were considered as statistically significant. Asterisks used in the figures represent the following significance values: *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001.
3. Results
3.1. Mutations in SEN1 induces enlarged cell size, rough surface morphology, and abnormal cell wall chitin content
We have earlier shown de-repression of FLO genes in mutants of Sen1, Nab3 and Rnt1 cells of yeast which leads to calcium-dependent flocculation phenotype (Singhet al. 2015). Recently one study reported the upregulation of FLO1 in a Sen1 mutant strain suggesting the contribution of Sen1 in the regulation of flocculation phenotype (Chenet al. 2017). Another study has shown a correlation between higher cell wall chitin content and sensitivity to cell wall damaging agents (Liuet al. 2010). However, a molecular mechanism connecting the CWI pathway to flocculation needs to be elucidated. The cell wall of yeast remains in direct contact with the surrounding environment and regulates; cell size, cell shape, cell-cell, and cell-surface interactions. Many of the studies have suggested that cell-cell interactions are majorly facilitated by cell wall adhesin proteins, also known as flocculins which are responsible for flocculation phenotype (Baruaet al. 2016; Besteret al. 2012). To elucidate the role of Sen1 in the mechanism of flocculation, we performed a series of experiments to find whether or not flocculation phenotype is connected with cell wall composition, cellular morphology and growth response in the presence of cell wall damaging agents. First; we performed microscopy and SEM experiments for morphological analysis. Second; the FACS experiment was conducted to measure cell wall chitin content using fluorescent CFW stain and cell size by plotting values of forward versus side scattering of wild-type, non-flocculating and flocculating strains. Third; the effect of cell wall damaging agents (CongoRed, CFW and caffeine) on the growth of cells was analyzed by spot assay and counting CFUs. Our careful analysis of microscopy and FACS results suggested that the flocculating mutant cells as compared to wild-type possess enlarged cell size and distinct rough surface morphology (Figure 1A and B) indicating defect in the cell wall. Strikingly we observed a higher level of chitin in the flocculating Sen1 mutant strains as compared to wild-type and non-flocculating mutant cells (Figure 1C). In spot assays, we found that most of the flocculating Sen1 mutants (sen1-1, sen1 ∆N, and sen1-K128E) which display abnormal cell size and surface morphology, showed growth defect, suggesting that Sen1 plays an essential role in cell wall maintenance (Figure S1 and S2). The early studies revealed that Sen1 is one of the main components of a protein complex known as NNS (Nrd1, Nab3, and Sen1), possess RNA/DNA helicase activity (Martin-Tumasz and Brow 2015) and functionally linked with factors of RNA processing machinery; Rrp6, TRAMP/Exosome and Rnt1 (Faskenet al. 2015; Finkelet al. 2010; Foxet al. 2015). We therefore, extended our investigations with available mutants of Nrd1 (nrd1-102), Nab3 (nab3-11), deletion mutants of TRAMP (air1∆, air2∆, trf4∆), rrp6∆ and rnt1∆ cells. Strikingly we detected higher chitin content in rnt1∆ and nab3-11 mutants as well. However, the cell size and chitin content of TRAMP and Nrd1 mutants was found similar to wild-type cells indicating that the role of Sen1 in regulation cell wall maintenance is not dependent TRAMP and Nrd1 (Figure S3A and S3B). Furthermore, as Sen1 has been shown to interact with Rnt1 (Finkelet al. 2010; Ursicet al. 2004) and Nab3 as part of NNS complex (Vasiljevaet al. 2008), we believe that role of Sen1 in cell wall integrity is probably dependent on Nab3 and Rnt1. Our comparative analysis of flocculating and non-flocculating cells indicate that only strongly flocculating mutant strains showed enlarged cell size, rough surface morphology, and abnormal chitin content. These studies demonstrate that yeast Sen1 in association with Nab3 and Rnt1 plays an essential role in the regulation of flocculation phenotype by cell wall maintenance.
Cell size measurement of wild-type and Sen1 mutants (sen1-1, sen1 ΔN, sen1-K128E and sen1-R302W) by plotting values of SSC vs. FSC obtained from FACS analysis by using BD-FACS Area III instrument. Cells at 0.8 OD600 were harvested, washed and re-suspended for FACS analysis. Equal numbers of cells (30000 events) were gated, and scattering was recorded. Plots of forward and side scattering were obtained and analyzed. P3 gate indicates the percentage of large cells. Representative DIC images of cells as indicated in figure acquired by using a Carl Zeiss motorized Apotome fluorescence microscope showing the size of each cell (scale bar 2 μm) are presented on the right side of respective plots.
The saturated cultures of cells (WT, sen1-1, sen1 ∆N and sen1-K128E) were harvested, washed, fixed with 4% Paraformaldehyde for 5 min, washed twice with PBS and dehydrated sequentially with increasing concentration of ethanol (30-100%). Cells were subsequently pulse sonicated and diluted 10 times in 100% ethanol. 20 μl of diluted cells were placed on stubs for O/N drying in a CO2 desiccator; gold coating was done by sputter coater and images were captured by using Zeiss High-Resolution Field Emission Scanning Electronic Microscope (HR-FESEM).
An equal number of exponentially growing cells (1 ml culture) as indicated in the figure were stained with CFW (50μg/ml) for 30 min, harvested, washed, re-suspended in PBS and analyzed by FACS. Equal numbers of cells (30000 events) were used for fluorescence using DAPI filter. P2 gate indicates the percentage of cells having increased cell wall chitin.
3.2. Mutations in SEN1, NAB3, and deletion of RNT1 induces constitutive activation of CWI pathway
In budding yeast, the Cell Wall Integrity (CWI) signaling pathway which composed of an evolutionarily conserved kinase cascade is majorly responsible for maintenance of cell wall homeostasis (Jimenez-Sanchezet al. 2007; Levin 2005; Levin 2011). The CWI pathway regulates the transcription of enzymes, required for cell wall synthesis under a variety of stress conditions including unfolded protein response (Scrimaleet al. 2009) and lipid homeostasis. Many studies have shown that during cell wall stress, different cell membrane sensor proteins interacts with the guanine nucleotide exchange factor (Rom2) leading to the activation of small GTPase (Rho1) which activates protein kinase C (Pkc1) as shown in the schematic (Figure 2A). Pkc1 transmits the signal to a MAPK module, MAPKKK (Bck1), MAPKK (Mkk1 and Mkk2), and the MAPK (Slt2/Mpk1). The activated Slt2 is known to trigger the cell wall damage-specific transcriptional response by activating Rlm1, a transcription factor (Bermejoet al. 2008; Sanzet al. 2012). The Mpk1/Slt2 has also been shown to prevent the recruitment of Sen1-Nrd1-Nab3 termination complex to regulate the expression of CWI pathway genes (Kim and Levin 2011). The cell wall chitin through covalent binding with glucan plays an essential role in controlling the morphogenesis and fungal cell wall remodeling which is necessary to counterbalance the cell wall stress (Arroyoet al. 2016; Eneet al. 2015). Since flocculating mutant strains showed higher chitin content, enlarged cell size and sensitivity to cell wall damaging agents, we first decided to measure the phosphorylation of Slt2 by quantitative western blotting using an antibody specific to the phosphorylated form of Slt2. Interestingly we found constitutive phosphorylation of Slt2 (8-10 fold increase) in all flocculating Sen1 mutant strains; sen1-1, sen1 ∆N, sen1-K128E (Figure 2B and 2C). However, the phosphorylation of Slt2 was also induced in nab3-11, rnt1∆ and nrd1-102 but lesser than flocculating Sen1 mutants in comparison to wild-type and a non-flocculating Sen1 mutant (sen1-R302W), TRAMP mutants (air1∆, air2∆, trf4∆) and rrp6∆ (Figure S4A and S4B). Western blot experiment to measure Slt2 phosphorylation was repeated 3-4 times, only one of the representative images is presented here. Western blot results of two biological repeats to measure Slt2 phosphorylation in Sen1 mutant strains are available in the supplementary file (Figure S4C). Previously, we reported the flocculation phenotype of few of the mutants of Sen1, Nrd1, Nab3, and rnt1∆, but the connection with CWI pathway was not established. These results suggest that the activation of the CWI pathway is functionally linked with flocculation of yeast cells.
Schematic representation of the Cell Wall Integrity (CWI) pathway. Under cell wall stress conditions, sensor proteins transmit the signal to Rom1/2 G-protein which in turn activates Pkc1 leading to the activation of MAP Kinase cascade consisting of Bck1, Mkk1/2, and Slt2. Activated Slt2 enters into the nucleus to activate Rlm1, a transcription factor which binds to responsive genes to induce the expression.
Exponentially growing cultures of wild-type, flocculating (sen1-1, sen1 ∆N and sen1-K128E) and non-flocculating (sen1-R302W) mutants of Sen1 were harvested, washed and whole-cell extracts were prepared by 20% TCA as described in material and methods. Cell extracts were separated on 10% SDS-PAGE and transferred onto nitrocellulose membrane. Blots were probed with primary antibodies as indicated for checking levels of phosphorylated and non-phosphorylated forms of Slt2. For protein loading controls, blots were re-probed with anti-TBP antibody and stained with Ponceau S as indicated.
Quantification of phosphorylated and non-phosphorylated forms of Slt2 relative to the wild-type and normalized by TBP using ImageJ software.
Bars in this data represent the difference in Slt2 phosphorylation in each mutant compared to that of wild-type cells. Error bars represent SD. Data shown here are the averages of three independent experiments.
3.3. The phosphorylation of Slt2 correlates with de-repression of FLO genes and flocculation phenotype
It is now clear that flocculating Sen1 mutant cells (sen1-1, sen1 ∆N, and sen1-K128E) exhibit constitutive increase in phosphorylation of Slt2 and alteration in cell wall integrity pathway as they are sensitive to cell wall damaging agents (Figure S1A) and enlarged cell size. Thus we decided to understand the underlying mechanism further. Under cell wall stress conditions, Slt2 is phosphorylated to induce the expression of responsive genes. Therefore, first, we measured the Slt2 phosphorylation in wild-type cells upon cell wall damage. We treated the cells with cell wall damaging agent (CFW) and exposed to a higher temperature to artificially induce the cell wall damage. The growth of wild-type cells was not affected at a higher temperature as well as in the presence of CFW (Figure S5A, S5B and Figure 3A, 3B), whereas, the phosphorylation of Slt2 was significantly induced in both these conditions (Figure 3C). Next, to investigate whether phosphorylation of Slt2 has any correlation with the expression of FLO genes, we performed the RT-qPCR to measure the mRNA expression of FLO1, 5 and 9. Interestingly we observed 3-5 fold increase in expression of FLO1, FLO5 and FLO9 (Figure 3D) than untreated cells, leading to the induction of flocculation/aggregation of cells upon stress induced by temperature and CFW treatments (Figure 3E). To further understand the role of CWI pathway in expression of FLOs (1, 5, 9 and 10), we measured the basal level mRNA expression by RT-qPCR of these genes in deletion mutant of CWI pathway; wsc1∆/∆, wsc2∆/∆, wsc3∆/∆, mid2∆/∆, rom2∆/∆, bck1∆/∆, mlp1∆/∆, slt2∆/∆, rlm1∆/∆ and double deletion mutant, slt2∆/∆ mlp1∆/∆. We detected about 2-3 fold decrease in basal expression of FLO genes in almost all the mutants except mlp1∆/∆ cells which further suggest that CWI signaling pathway regulates the expression of FLO genes (Figure 3F). The double deletion mutant (slt2∆/∆ mlp1∆/∆) was utilized to test whether the signal is transmitted through Mlp1 or Slt2. Interestingly we noticed a decrease in transcription of FLO genes in double deletion mutant than single deletion (mlp1∆/∆ indicating that transcription is regulated through Slt2-Rlm1, not via Mlp1-Swi4/6 branch. A schematic for the entire CWI signaling pathway has been shown in Figure 2A. Thus our results suggest that the activation of CWI pathway is essential for eliciting the expression of FLO genes and flocculation phenotype which provides a suitable microenvironment required for cell survival under stress conditions (Goossenset al. 2015).
Growth curve analysis of wild-type cells was measured in liquid growth media in absence and presence of CFW concentrations as indicated in an automatic spectrophotometer plate reader by taking OD600 at regular intervals of 30 min for 36 hours.
Spot assay was performed to test the growth of wild-type cells in the presence of CFW and at 37°C. Cells were grown O/N, diluted to 1 OD600 in 1 ml of water and diluted further 4 times (10 fold serial dilution). 3 μl of cells from each dilution were spotted on SC-Agar plates with and without CFW, incubated at 30°C. For temperature stress, cells were spotted on SC-Agar plates and incubated at 37°C. Growth was recorded by scanning after 72 hours of incubation.
Wild-type cells were grown until 0.8 OD600 and grown further for 3 hours in the presence of CFW as well as at 37°C as indicated to induce cell wall stress. Whole cell extracts were prepared by 20% TCA and phosphorylation of Slt2 were measured by quantitative western blotting.
Quantification of mRNAs of FLO1, 5, and 9 by RT-qPCR. The 5 ml cultures of exponentially growing wild-type cells were grown for 3 hours at 30°C, 37°C, 40°C and in the presence of CFW concentrations as indicated at 30°C for 3 hours. The cells were subsequently harvested, total RNAs were extracted by ‘heat phenol freeze method’ as described in materials and methods and cDNA was prepared by using an iScript cDNA synthesis kit purchased from Biorad. The gene-specific primers were used for amplification of mRNAs by RT-qPCR. ACT1 mRNAs were used as an internal control for quantification of FLO mRNAs.
Images showing flocculation of wild-type yeast cells grown O/N at 30°C, 37°C, 40°C and in the presence of CFW. O/N grown cultures were diluted to 2 OD600 in 3 ml volume of fresh SC growth media and poured into 35 mm Petri dishes. The cells in plates were agitated for uniform distribution by shaking and then left uninterrupted; subsequently, pictures were recorded at time 0, 15 and 30 min.
Quantification of mRNAs of FLO1, 5, 9 and 10 by RT-qPCR. Exponentially growing cultures at 30°C of wild-type and deletion mutant cells of the CWI pathway as indicated were harvested, total RNAs were extracted, and cDNA was prepared for RT-qPCR to measure the expression of FLO genes. ACT1 mRNAs were used as internal control for quantification.
Bars in this data represent the difference in fold change of FLO gene expression in WT cells compared to that of untreated and treated cells. Error bars represent SD. Data shown here are the averages of three independent experiments. Statistical analysis was carried out with a two-tailed, unpaired, Student’s t-test to analyze differences between the untreated to the Temp at 40°C and CFW treated (500μg/ml) of wild-type cells:**P ≤ 0.01; ***P ≤ 0.001.
3.4. Deletion of SLT2 suppresses the expression of FLO genes and flocculation
We have so far observed a strong correlation between Slt2 phosphorylation and expression of FLO genes. To further validate this observation, we deleted the SLT2 from two of the flocculating Sen1 strains (sen1-1 and sen1-K128E) to create double mutants which were confirmed by western blotting using an antibody specific to Slt2 protein (Figure 4A). The wild-type and mutant strains (sen1-1 and sen1-K128E) with and without SLT2 were utilized to measure the mRNA expression by RT-qPCR of FLO1, 5, 9 and 10 genes. As expected we observed a drastic decrease in expression of FLO genes in the absence of SLT2 measured by RT-qPCR (Figure 4B) leading to the suppression of flocculation phenotype, examined by recording cell aggregation in Petri dishes at indicated time points (Figure 4C). Furthermore, flocculation of cells was quantified as % flocculation which was also decreased in the absence of SLT2 in sen1-1 and sen1-K128E (Figure 4D). Moreover, the growth of the wild-type and Sen1 mutant strains with and without SLT2 was tested by spot assay in the presence of cell wall damaging agents; CongoRed, CFW and Caffeine. Interestingly, sen1-1 slt2∆ and sen1-K128E slt2∆ cells were found more sensitive to the cell wall damaging agents in comparison to wild-type, sen1-1 and sen1-K128E cells (Figure S6). However, the slt2∆ cells were found more sensitive than wild-type, sen1-1 and sen1-K128E but less than sen1-1 slt2∆ in the presence of cell wall damaging agents. These observations suggest that Slt2 dependent regulation of FLO genes is required for survival under cell wall stress conditions.
Confirmation of SLT2 deletion in WT, sen1-1 and sen1-K128E cells by western blotting. TBP western was performed for loading control.
The mRNA quantification of FLO1, 5, 9 and 10 by RT-qPCR. Cells were grown at 30°C, total RNAs, and cDNAs were prepared. ACT1 mRNAs were used as internal control for quantification of FLO mRNAs.
Flocculation of wild-type and Sen1 mutant cells (sen1-1 and sen1-K128E) with and without SLT2 was recorded by taking pictures at indicated time points as described earlier.
The % flocculation of wild-type and Sen1 mutant cells (sen1-1 and sen1-K128E) with and without SLT2 was calculated.
Bars in this data represent the difference in fold change of FLO gene expression (C) and % flocculation (D) in each mutant compared to that of wild-type cells. Error bars represent SD. Data shown here are the averages of three independent experiments. Statistical analysis was carried out with a two-tailed, unpaired, Student’s t-test to analyze differences between the sen1-1 and sen1-K128E with and without SLT2 deletion:*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001.
3.5. The binding of Rlm1 correlates with upregulation of FLO genes and flocculation phenotype
The activation of the CWI pathway occurs in cell wall damage conditions. A few years ago the role of Slt2-dependent CWI pathway was suggested in cell wall remodeling and filamentous growth of fungus (Birkayaet al. 2009) which further supports our observations. Phosphorylation of Slt2 is the key event in the activation of the CWI signaling pathway leading to the recruitment of Rlm1 and SWI2/SNF2 to regulate the expression of CWI responsive genes (Sanzet al. 2012). As our results reveal constitutive phosphorylation of Slt2 in flocculating strains, we hypothesized that it might lead to the recruitment of Rlm1 at the promoters of FLO genes. First, we searched for the Rlm1 binding sites at the FLO genes through sequence analysis. A consensus nucleotide sequence, TA(A/T)4TAG was identified as the Rlm1 binding sequence by site selection experiments using random oligonucleotide sequences (Dodou and Treisman 1997; Jung and Levin 1999). Interestingly our sequence analysis revealed similar recognition sequences at the upstream regions of FLO1, FLO5, FLO9 and FLO10 promoters (Figure 5A) suggesting that Rlm1 probably binds at the promoters of FLO genes to activate the transcription. To investigate whether Rlm1 is recruited to the FLO promoters or not, we performed chromatin IP experiments. Chromatin extracts were prepared from formaldehyde crosslinked Rlm1 Myc tagged strains; wild-type, sen1-1 (flocculating strain) and a non-flocculating mutant, sen1-R302W. The Anti-Myc antibody was used for the immunoprecipitation. Four primer sets for FLO1 upstream sequences (ranging from +14 to −817) and FLO5 (ranging from +20 to −804) were used for PCR amplification of Rlm1 ChIP DNA (Figure 5B). Our analysis also suggests that Rlm1 binding sites overlap with the Tup1 binding site (Fleminget al. 2014) indicating that these two proteins probably compete for binding to regulate the expression of FLO genes. We found around 2.5 fold enrichment of Rlm1 at the FLO1, and FLO5 analyzed by RT-qPCR using Rlm1 ChIP DNA and primer sets for FLO1 and FLO5 as indicated in the schematic. The occupancy of Rlm1 was relatively found more in sen1-1 flocculating cells than wild-type and non-flocculating Sen1 mutant, suggesting that Rlm1 is constitutively recruited at the FLO genes (Figure 5C). This result provides strong support to our hypothesis that Rlm1 physically binds at the promoters of FLO genes to regulate their expression. We observed peak PCR amplification of Rlm1 chip DNA with P2 and P3 primer sets, suggesting that recruitment of Rlm1 majorly occurs at around −300 to - 600 bp of FLO1 and FLO5. Primers for the STE6 gene were used as internal control for the calculation of fold change. Further to study the role of Rlm1 in the expression of FLO genes and flocculation phenotype, we first deleted the RLM1 gene from wild-type, flocculating Sen1 mutant strains (sen1-1, sen1 ∆N, and sen1-K128E) and non-flocculating Sen1 mutant, sen1-R302W. Interestingly upon deletion of RLM1, we detected a noticeable decrease in flocculation/aggregation of cells measured by taking pictures at indicated time points of cell suspensions placed in Petri-dishes (Figure 5D). The % flocculation calculated by taking OD600 at different time points during the settling of cells in glass tubes was also reduced after deletion of RLM1 (Figure 5E). Subsequently, RT-qPCR was performed to quantify the expression of FLO genes in all these strains in the presence and absence of RLM1. As expected, we found good reduction (~50%) in the expression of mRNAs of FLO genes in sen1-1, sen1 ∆N, and sen1-K128E strains after deletion of RLM1 indicating that Rlm1 acts as a transcription factor for the expression of FLO genes (Figure 5F). Furthermore, the effect of cell wall damaging agents on the growth of cells after RLM1 deletion as well as phosphorylation of Slt2 was examined. To this end the inhibitory effect of cell wall damaging agents on the growth was suppressed after deletion of RLM1 in contrast to deletion of SLT2, suggesting that CWI pathway dependent expression of FLO genes is essential for survival under stress conditions (Figure S7A). Since Slt2 is located at the upstream of Rlm1 in the signaling pathway, we did not detect any significant reduction in phosphorylation of Slt2 after RLM1 deletion (Figure S7B and S7C). A slight decrease in Slt2 protein level was observed because Rlm1 also acts as the transcription factor for expression of Slt2 through a regulatory feedback mechanism (Garciaet al. 2016). A molecular mechanism for the recruitment of Rlm1 and regulation of FLO genes has been investigated further as described from now onward in this manuscript.
Identification of Rlm1 binding sites at the upstream regions of FLO genes (FLO1, 5, 9 and 10).
Schematic representation of FLO1 and FLO5 promoters showing primer sets designed for upstream regions including Rlm1 binding site used for Chromatin Immunoprecipitation (ChIP) analysis. Tup1 binding is also indicated in the schematic, adapted from (Fleminget al. 2014).
ChIP assay was performed as described in materials and methods with chromatin extracts isolated from Rlm1 Myc tagged wild-type, sen1-1, and sen1-R302W cells. Anti-Myc antibody was used for immunoprecipitation. ChIP DNAs were analyzed by RT-qPCR. Primers of STE6 gene were used as internal control for quantification of Rlm1 occupancy at P1, P2, P3 and P4 upstream locations of FLO1 and FLO5 genes.
Flocculation of wild-type and Sen1 mutant cells (sen1-1, sen1 ∆N, sen1-K128E and sen1-R302W) with and without RLM1 was recorded by taking pictures at indicated time points.
The % flocculation of wild-type and Sen1 mutant cells (sen1-1, sen1 ∆N, sen1-K128E, and sen1-R302W with and without RLM1 were calculated.
The mRNA quantification by RT-qPCR of FLO1, 5, 9 and 10 genes in wild-type and Sen1 mutant cells (sen1-1, sen1 ∆N, sen1-K128E and sen1-R302W) with and without RLM1. Cells were grown at 30°C, total RNAs, and cDNAs were prepared. ACT1 mRNAs were used as internal control for quantification of FLO mRNAs.
Bars in this data represent the difference in fold change of Rlm1 binding (C), % flocculation (E) and FLO gene expression (F) in each mutant compared to that of wild-type cells. Error bars represent SD. Data shown here are the averages of three independent experiments. Statistical analysis was carried out with a two-tailed, unpaired, Student’s t-test to analyze differences between the indicated mutants and the wild-type strain (C) and between mutants with and without RLM1 deletion (E and F): *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001.
3.6. Binding of Rlm1 reduces the occupancy of Tup1 leading to the recruitment of TBP and Pol-II at the promoters of FLO genes
Previous studies have shown that reduced occupancy/deletion of TUP1 results in strong de-repression of FLO genes (Fleminget al. 2014). It was also suggested that acetylation of core histones, nucleosome disruption, recruitment of Pol-II and Swi2 at the FLO1 promoter robustly induces the transcription. However, the contribution of Rlm1 in transcription of FLO genes has not been explored before. Our results suggest that transcriptional induction of FLO genes is dependent on activation of Slt2 and Rlm1. Through ChIP assay; we detected binding of Rlm1 over FLO1 and FLO5 (Figure 5C) promoters in flocculating sen1-1 strain as compared to wild-type cells and sen1-R302W, a non-flocculating mutant. Therefore, we hypothesized that binding of Tup1 and Rlm1 at the promoter of FLO genes is oppositely regulated. To validate the relationship between Tup1 and Rlm1 in the regulation of FLO genes, we first performed ChIP assays using TBP, Pol-II, and Tup1 specific antibodies to examine occupancy at the FLO1 and FLO5 genes in a flocculating mutant of Sen1 (sen1-1). The wild-type and non-flocculating Sen1 mutant (sen1-R302W) cells were used as a control. The ChIP DNAs were amplified by using primer sets for upstream regions of FLO1 and FLO5 genes through RT-qPCR for analyzing the occupancy of TBP, Pol-II, and Tup1. Remarkably we found constitutive recruitment of TBP, Pol-II and a significant decrease in occupancy of Tup1 at the promoters of FLO1 (Figure 6A) and FLO5 (Figure 6B) in sen1-1 cells than wild-type and non-flocculating mutant. These results suggest that signaling through the CWI pathway is critical for Rlm1 recruitment at the FLO genes. The low occupancy of Pol-II at the promoters of FLO genes in sen1-R302W cells could be because this substitution is known to disrupt the physical interaction between Sen1 and Rpb1 (Chinchillaet al. 2012). Further experiments would be required to explore the role of Rlm1, Sen1, and Tup1 in transcriptional regulation of FLO genes. It is quite possible that stress conditions induced by genetic or epigenetic factors trigger changes in local chromatin structure which allows Sen1 mediated binding of Rlm1 resulting in the eviction of Tup1 to induce the expression of FLO genes. Alternatively, the ‘loss of Sen1 function’ may prevent Tup1 to bind or recruit Rlm1 to evict Tup1. We believe this mechanism has been evolved to develop stress-tolerant phenotype.
ChIP assays were performed as described in materials and methods with chromatin extracts isolated from wild-type, sen1-1, and sen1-R302W cells. ChIP DNAs were analyzed by RT-qPCR. ChIP assays were performed with the following antibodies: Anti-TBP, Anti-Pol-II, and Anti-Tup1 antibodies.
Schematic representation indicating locations of primers of the FLO1 gene used for amplification of ChIP DNAs is shown at the top. Chip assays were performed to check the relative occupancy of TBP, Pol-II, and Tup1 at the promoter of FLO1.
Schematic representation indicating locations of primers of the FLO5 gene used for amplification of ChIP DNAs is shown at the top. Chip assays were performed to check the relative occupancy of TBP, Pol-II, and Tup1 at the promoter of FLO5.
Primers of STE6 gene were used as internal control for quantification of TBP, Pol-II and Tup1 occupancy. Bars in this data represent the difference in the fold change of relative occupancy of TBP, POLII, and Tup1 at FLO1 and FLO5 gene promoters in sen1-1 and sen1-R302W mutants compared to that of wild-type cells. Error bars represent SD. Data shown here are the averages of three independent experiments. Statistical analysis was carried out with a two-tailed, unpaired, Student’s t-test to analyze differences between the sen1-1 mutant and the wild-type cells: *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001.
3.7 Slt2 is constitutively phosphorylated in flocculating histone H3 and H4 mutants leading to the Rlm1 dependent expression of FLO genes
Until now our results strongly suggest that the CWI pathway plays an essential role in the regulation of FLO genes and flocculation phenotype. However, to study whether CWI pathway dependent regulation of FLO genes is a general mechanism applicable to all flocculating strains or is it specific to only certain strains (for example, Sen1 mutants) needs to be elucidated. Therefore, we decided to extend our investigations by utilizing other flocculating strains. To this end, we first looked for the additional flocculating strains. It has been earlier reported that mutants defective in COMPASS (Complex Proteins Associated with Set1) exhibit flocculation property (Dietvorst and Brandt 2008). Since subunits of COMPASS possess Histone methyltransferase activity, we decided to screen tail mutants of histone H3 and H4 for flocculation phenotype. Interestingly we were successful in finding two of the mutants; H3R63A and H4R3K with better flocculation property. First, we decided to examine the phosphorylation of Slt2 in these two mutants along with respective wild-type cells. As compared to wild-type cells, we observed more phosphorylation of Slt2 in both the mutants indicating that the CWI pathway is involved in flocculation of the histone mutants as well (Figure 7A). We then went ahead to identify the role of Rlm1 on mRNA expression of FLO genes, phosphorylation of Slt2, flocculation/aggregation phenotype as well as sensitivity to cell wall damaging agents in flocculating histone mutants. For this purpose, we deleted the RLM1 gene from the mutants and wild-type cells. Relative to wild-type cells, we observed constitutive upregulation of FLO genes in both the flocculating histone mutants (Figure 7B and C). However, deletion of RLM1 significantly reduced the mRNA expression of FLO genes, leading to a decrease in flocculation phenotype (Figure 7D and E) in both the mutants. Our careful analysis revealed low levels of Slt2 phosphorylation, reduction in expression of FLO genes and flocculation efficiency in the H4R3K mutant in comparison to H3R63A. Additionally, H4R3K cells showed less sensitivity to cell wall damaging agents relative to H3R63A mutant which correlates with differences in Slt2 phosphorylation (Figure S8A and S8B) (Liuet al. 2010). However, unlike H3R63A, we did not observe resistance to CR and CFW, and sensitivity to Caffeine in H4R3K after deletion of RLM1. Above results once again strongly suggest that the defect in cell wall induces the expression of FLO genes mediated by activation of the CWI pathway. Thus our results further suggest that the CWI pathway dependent activation of Slt2 and Rlm1 is required for the de-repression of FLO genes.
Exponentially growing cultures of flocculating histone mutants (H3R63A and H4R3K) and respective wild-type cells with and without RLM1 deletion were harvested; washed and whole-cell extracts were prepared by 20% TCA. Cell extracts were separated on 10% SDS-PAGE and transferred onto nitrocellulose membrane. Blots were probed with primary antibodies as indicated for checking levels of phosphorylated and non-phosphorylated forms of Slt2 by quantitative western blotting using IR Dye-labeled secondary antibodies. Protein extracts of gas1∆/∆ and slt2∆ strains were used as positive and negative control respectively. Blots were re-probed with anti-TBP antibody for protein loading control.
B. and C. The mRNA quantification of FLO1, 5, 9 and 10 in flocculating histone H3R63A (B) and H4R3K (C) mutants with respective wild-type cells, with and without RLM1 deletion by RT-qPCR. Cells were grown at 30°C, total RNAs, and cDNAs were prepared. ACT1 mRNAs were used as internal control for quantification of FLO mRNAs.
Flocculation of flocculating histone mutants (H3R63A and H4R3K) and respective wild-type cells with and without RLM1 deletion was recorded by taking pictures at indicated time points.
The % flocculation of flocculating histone mutants (H3R63A and H4R3K) and respective wild-type cells with and without RLM1 deletion was calculated.
Bars in this data represent the difference in the fold change of FLO gene expression (B and C) and % flocculation (E) in each mutant compared to that of wild-type cells. Error bars represent SD. Data shown here are the averages of three independent experiments. Statistical analysis was carried out with a two-tailed, unpaired, Student’s t-test to analyze differences between the indicated mutants and the wild-type strain (B-C) and between mutants with and without RLM1 deletion (E):: *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001.
3.8. De-repression of FLO genes in the absence of TUP1 is not dependent on Slt2 and Rlm1
It is known from few studies that Cyc8-Tup1 repressor complex robustly suppresses the expression of FLO genes and deletion of CYC8 or TUP1 results in constitutive expression of FLO genes (Fleminget al. 2014). To extend our studies further, we decided to identify the role of the CWI pathway in the expression of FLO genes upon deletion of TUP1 as well. To our surprise, we did not find the activation of Slt2 in strongly flocculating deletion mutants of tup1 and cyc8 (Figure 8A) unlike the flocculating strains of Sen1 and Histones. Furthermore, we deleted the SLT2 and RLM1 in tup1∆ mutant and measured the expression of FLO genes. However, we did not find any difference in the expression of FLO genes in the tup1∆ mutant with and without deletion of SLT2 or RLM1 (Figure 8B).
Exponentially growing cultures of tup1∆, tup1∆/∆, and cyc8∆/∆ with respective wild-type were harvested, washed and whole-cell extracts were prepared by 20% TCA. The cell extracts were separated on 10% SDS-PAGE and transferred onto nitrocellulose membrane. Blots were probed with primary antibodies as indicated for checking levels of phosphorylated and non-phosphorylated forms of Slt2. For protein loading control, blots were re-probed with an anti-TBP antibody. Extracts from slt2∆ and gas1∆/∆ cells were taken as –ve and +ve control respectively for Slt2 phosphorylation.
The mRNA quantification of FLO1, 5, 9 and 10 in the tup1∆ mutant with and without SLT2 and RLM1 deletion along with wild-type cells by RT-qPCR. Cells were grown at 30°C, total RNAs, and cDNAs were prepared. The ACT1 mRNAs were used as internal control for quantification of FLO mRNAs.
Bars in this data represent the difference in fold change of FLO gene expression (B) in each mutant compared to that of wild-type cells. Error bars represent SD. Data shown here are the averages of three independent experiments. Statistical analysis was carried out with a two-tailed, unpaired, Student’s t-test to analyze differences between the tup1∆ mutant with double mutants, tup1∆ slt2∆ and tup1∆ rlm1∆ (B): *P ≤ 0.05; **P ≤ 0.01.
4. Discussion
Many factors are required for the regulation of gene expression in eukaryotes. The epigenetic and genetic factors induce structural and chemical changes to the chromatin, impacting the expression of responsive genes. Upon cell wall stress, S. cerevisiae activates an adaptive transcriptional response to counterbalance the stress, which is mediated by a Slt2-dependent MAPK signaling pathway. Flocculation of yeast cells is another mechanism which provides adaptation to several types of stress conditions including cell wall stress. This study was conducted to identify the role of CWI signaling pathway in the expression of FLO genes.
4.1. The yeast flocculation is regulated through a conserved MAP kinase pathway
Yeast cells are exposed to a variety of environmental stress forms and respond through a conserved MAPK pathway which induces transcription of stress-responsive genes. The specific signaling pathways through the cascade of reactions integrate into the chromatin to induce or repress the transcription of responsive genes required for survival under a variety of pathophysiological stress conditions. The environmental cell wall stress conditions result in the activation of the CWI signaling pathway leading to the upregulation of responsive genes required for maintenance of cell wall. The transcriptional program triggered by cell wall stress conditions is mediated by Slt2 and Rlm1 to induce the expression of responsive genes. Previous work has shown constitutive expression of FLO genes in mutants of Sen1, Nab3 and rnt1∆ cells leading to flocculation/aggregation of cells (Singhet al. 2015). Morphological analysis of flocculating cells (Sen1 mutants) suggests defects in the cell wall (Figure 1C). Therefore, we hypothesized that the CWI pathway might be required for regulation of yeast flocculation. The proteins encoded by FLO1, FLO5, FLO9, FLO10 and FLO11 genes are primarily responsible for flocculation and invasive growth. The proteins encoded by FLO genes are glycosylphosphatidylinositol-linked glycoproteins (Dranginiset al. 2007). There are evidences which suggest that the mechanism of flocculation/filamentous growth is functionally conserved. First; the expression of certain human genes such as DNA methyltransferase 1 (DNMT1) has been reported to induce yeast flocculation (Sugiyamaet al. 2015). Second; the expression of an activated extracellular signal-regulated kinase 1 (ERK1), one of the MAP kinases of mammalian systems in yeast has been shown to induce filamentous growth and cell wall remodeling (Atienzaet al. 2000). Third; most of the yeast cell wall flocculins and human cell membrane adhesins consists of amyloid-forming amino acid sequences (Rameauet al. 2016; Ramsooket al. 2010). Fourth, certain mutations in conserved histone proteins and deletion of a histone-modifying enzyme (swd3∆) also induce flocculation of yeast cells (Dietvorst and Brandt 2008). Mutations in histones and histone-modifying enzymes have been implicated in diseases including cancer (Shilatifard 2012). Fifth, during current studies we detected constitutive activation of a conserved MAP kinase (Slt2) dependent pathway known as Cell wall integrity (CWI) in flocculating cells (Figure 2B, Figure S4A and Figure 7A). It is quite possible that under certain stress conditions due to epigenetic or genetic changes, human cell membrane adhesins like yeast cell wall flocculins may induce morphological changes affecting cell-cell communication and pathological properties. Many cellular activities such as proliferation, differentiation, and cell death have been shown to be controlled by signaling pathways involving specific MAP kinases (Kim and Choi 2010). The defects in MAP kinases are implicated in many human diseases including Alzheimer’s disease, Parkinson’s disease, amyotrophic lateral sclerosis (ALS) and cancers (Kim and Choi 2010). Many mutations in human senataxin (yeast Sen1) have been correlated with neurological disorders, AOA2 and ALS4. However, the underlying mechanism for AOA2 and ALS4 as well as induction of flocculation due to mutations in yeast Sen1 is not clear. Our observations suggest that dysregulation in MAP kinase pathway might be responsible for AOA2 and ALS4 disorders associated with mutations in human senataxin.
4.2. The Rlm1 is required for the de-repression of FLO genes under stress conditions
To dissect the molecular mechanism of yeast flocculation, we performed a variety of experiments. First, we found that flocculating strains exhibit enlarged cell size, higher chitin content and rough cell surface morphology which is indicative of alterations in the cell wall. We then decided to examine the CWI pathway by testing the phosphorylation of Slt2, one of the key MAP kinases. Interestingly, we detected constitutive phosphorylation of Slt2 in flocculating strains, not only in Sen1 mutants but also in flocculating histone mutant cells, suggesting that the CWI signaling pathway indeed regulates the de-repression of FLO genes. It is known from several studies that phosphorylated form of Slt2 enters into the nucleus and which acts as a kinase to activate the Rlm1, a transcription factor which in turn induces the transcription of cell wall damage responsive genes by physically interacting with SWI/SNF chromatin remodeling complex and recruitment of general transcription factors. However, the significance of Rlm1 in de-repression of FLO genes has not been explored before. Excitingly, in addition to constitutive activation of Slt2, we also found constitutive recruitment of Pol-II and TBP, and reduced occupancy of Tup1 at the promoters of FLO genes in flocculating cells revealing the role of CWI signaling pathway in de-repression of FLO genes. Furthermore, we found reduced expression of FLO genes and suppression of flocculation phenotype upon deletion of SLT2 or RLM1 (Figure 4 and 5). To get more insight, we also measured the basal expression of FLO genes in deletion mutant cells of the CWI pathway. Remarkably we found a reduction in basal expression of FLO genes in all the deletion mutants of the CWI pathway except mlp1∆ cells. Since basal expression of FLO genes in mlp1∆ mutant did not decline, unlike other mutants, suggesting that expression of FLO genes is exclusively regulated through Rlm1 branch of CWI signaling pathway. These observations motivated us to hypothesize that Rlm1 is required for the transcription of FLO genes by binding to the promoters of FLO genes under stress conditions. We, therefore, searched for the Rlm1 binding site at the FLO genes by using a previously identified DNA sequences, TA(A/T)4TAG as Rlm1 binding sequences. We found that Rlm1 binding sites are indeed available at the upstream regions of FLO1, 5, 9 and 10 genes, overlapping with the Tup1 binding sites which were confirmed by ChIP assay (Figure 5A, B, C). Furthermore, to examine whether phosphorylation of Slt2 has any correlation or not with transcriptional induction of FLO genes during cell wall stress, we artificially induced cell wall damage condition by exposing wild-type cells to CFW and higher temperature. As per our assumption, in both the conditions, we observed an increase in phosphorylation of Slt2 and upregulation of FLO1, 5, and 9 leading to flocculation of cells. These results indicate the stress-dependent role of the CWI pathway in transcriptional de-repression of FLO genes via activation of Slt2 and Rlm1. Furthermore, to confirm the role of the CWI pathway in the regulation of FLO genes, first, we performed similar experiments by using additional flocculating strains other than Sen1 mutants. Through the screening of histone mutants, we found few mutants with the flocculating property. Out of these, we selected two mutants which showed better flocculation phenotype. We detected an increase in phosphorylation of Slt2 and induction of FLO genes in flocculating histone mutants as well, which further suggest that activation of Slt2p and Rlm1p-dependent CWI pathway is the general mechanism required for flocculation of yeast cells. Next, we also tested phosphorylation of Slt2 in CYC8 and TUP1 deleted cells which are known to possess strong flocculation property (Fleminget al. 2014). To our surprise, we did not detect any considerable increase in phosphorylation of Slt2 in flocculating CYC8 and TUP1 deleted cells. Moreover, upon deletion of SLT2 or RLM1 in tup1∆ cells also did not affect the expression of FLO genes. Future investigations would be required to understand the mechanism for Slt2-independent induction of FLO genes in tup1∆ and cyc8∆ cells.
We hypothesize two possibilities for Slt2-independent induction of FLO genes in the absence of CYC8 and TUP1; one, Rlm1-dependent and second, Rlm1-independent. Based on Rlm1-dependent mechanism we propose that in the absence of CYC8 or TUP1, the basal level of Rlm1p is sufficient to bind at the freely accessible Rlm1 binding sites at the promoter of FLO genes. Tup1p repressor in an unstressed condition probably occupies the Rlm1 binding site as binding sites for these two factors are overlapping. According to the second Rlm1-independent mechanism, possibly the nucleosome positioning is disrupted due to the opportunistic recruitment of SWI/SNF complex in the absence of Tup1p leading to the formation of PIC causing robust transcriptional de-repression of FLO genes. However, to support or refute these mechanisms further investigations are needed. It has been shown earlier that Cyc8-Tup1 repressor complex regulates the chromatin structure and gene expression of repressed genes by influencing Isw2-dependent nucleosome positioning (Rizzoet al. 2011; Zhang and Reese 2004) which supports our conclusions to some extent. Despite many studies, the precise mechanism for the regulation of FLO genes under cell wall stress conditions has not been explored before. Our results suggest that under cell wall stress conditions, activated Slt2 triggers the recruitment of active Rlm1 results in eviction of Tup1, recruitment of TBP and Pol-II, to induce the expression of FLO genes (Figure 9). Several studies have established the fact that nucleosome positioning influences the accessibility of the transcription factors binding sites at the promoters. The positioning of nucleosomes in yeast is primarily determined by two major chromatin remodeling complexes; ISW2 and SWI2/SNF2 (Tomaret al. 2009). Our ChIP assays suggest that Rlm1 is constitutively bound to the promoter of FLO genes in flocculating cells probably due to disruption of nucleosome positioning. The occupancy of Tup1 in the absence of Rlm1 perhaps maintains the repressed state of FLO genes due to nucleosome positioning. Further in vivo and in vitro investigations would be required to establish a precise relationship between Rlm1, Tup1 and chromatin remodeling.
Genetic (mutations in genes) and stress conditions (cell wall damaging agents and higher temperature) alters the biochemical composition of yeast cell wall results in activation of the CWI signaling pathway via phosphorylation of Slt2. Activated Slt2 enters into the nucleus and activates the Rlm1. Activated Rlm1 along with general transcription factors; TBP, Pol-II, and SWI/SNF is recruited to form Pre Initiation Complex (PIC) and dissociates the Tup1 repressor, converting repressed (OFF) state of FLO genes to transcriptional active (ON) state. To induce the expression of FLO genes, Sen1 probably cooperates with multiple proteins including Rlm1, factors of general transcription machinery leading to the eviction of Tup1 repressor.
5. Conclusion
The novel Rlm1-dependent mechanism for the transcription of FLO genes has certainly enhanced our knowledge about the molecular biology of yeast flocculation/biofilm formation. In summary, our studies are expected to serve two primary purposes, first; strengthening the understanding about yeast biofilm biology which may help us to tackle the antifungal resistance and therapies for biofilm-based fungal diseases and second; in biotechnological applications as flocculating yeast cells are useful for fermentation processes to develop antimicrobials and biofuel.
Author contribution
RST, SKS, VS conceived and designed the study. SKS and RK performed the experiments. Results were analyzed by RST, SKS, RK, and VS. RST, SKS, RK, and VS wrote the manuscript. SKS, RK, and VS prepared all the figures. All authors reviewed the results and approved the final version of the manuscript.
Funding
This research work was supported by funds from the Science and Engineering Research Board (SERB, Grant no: EMR/2015/001797), Govt. of India to RST.
Conflict of interest
Authors declare no conflict of interest.
Acknowledgments
We thank Michael Culbertson (University of Wisconsin), Jeffry Corden (Johns Hopkins University School of Medicine), Steve Buratowski (Harvard Medical School), Daniel Reines (Emory University), Anita H. Corbet (Emory University), Sherif Abou Elela (University of Sherbrooke) and David E. Levin (Boston University) for providing us the Sen1, Nrd1, Nab3, TRAMP, Rnt1 and CWI pathway mutant strains of yeast respectively. We acknowledge Joseph Reese (Pennsylvania State University) for the generous gift of TBP and Tup1 antibodies. Council of Scientific and Industrial Research (CSIR), India is acknowledged for fellowship support to SKS and RK. Members of the chromatin biology laboratory are recognized for their advice and helpful discussions throughout this work.
References
