1.1 Hog1 pathway is constitutively active in flocculating tup1∆/∆ and cyc8∆/∆ cells.
Flocculation is one of the prominent characteristic features of yeast cells required for survival in harsh environmental conditions. Floc formation is influenced by the availability of nutrients, cations, dissolved oxygen, pH, agitation, and temperature [28, 29, 60]. The tup1-Cyc8 corepressor complex is known to suppress yeast flocculation [2]. Yeast cells possess a variety of kinases that are essential to generating appropriate responses required to survive under environmental stress conditions. During environmental stress conditions, yeast cells activate specific pathways. Many of these pathways show crosstalk with each other and share common signaling mechanisms. Out of three major MAPK pathways (Fus3/Kss1, HOG, and CWI pathway), the CWI and HOG pathways are majorly involved in environmental stress responses [61]. These MAPK pathways are involved in the regulation of stress-specific genes through the activation of transcription repressors or activator proteins. The MAPK pathways and flocculation phenotype are required for yeast to survive in adverse conditions, however, whether both are interlinked or not, has not been explored till now. To identify the involvement of MAPK pathway in yeast flocculation, we first tested the Hog1p and Slt2p phosphorylation in constitutively flocculating TUP1/CYC8 deleted cells. Interestingly we observed hyper-phosphorylation of Hog1 in flocculating TUP1 and CYC8 deleted cells, which indicates involvement of the HOG MAPK pathway in flocs formation (Fig. 1B). However, we did not find any significant difference in the Slt2 phosphorylation in flocculating strains compared to wild-type cells (Fig. 1C). The constitutive phosphorylation of Hog1 in flocculating cells, even in untreated condition, suggested that the HOG MAPK pathway might have some role in yeast cells’ flocculation.
1.2 Hog1 downstream genes are upregulated in flocculating tup1∆/∆ and cyc8∆/∆ cells.
During environmental stress conditions such as osmotic stress, heat stress, arsenite stress, cold stress, acetic acid stress, and hypoxic stress, the Hog1 protein become active through phosphorylation [49–51, 54]. After activation Hog1 moves towards the nucleus where it works as a transcription factor to activate the expression of genes that are involved in stress tolerance. For further confirmation that in flocculating cells HOG MAPK pathway is constitutively active, we checked the basal level expression of Hog1-regulated genes such as GPD1, CTT1, and GRE2 in TUP1 and CYC8 deleted strains by qRT-PCR. GPD1 encodes a NAD-dependent glycerol-3-phosphate dehydrogenase which is a key enzyme for glycerol synthesis required for growth under osmotic stress and also regulates the heat shock response [62]. CTT1 encodes a cytosolic catalase T, involved in protection from oxidative damage. The expression of CTT1 is also induced under osmotic and carbon starvation stress conditions [63, 64]. GRE2 encodes 3-methyl butanal reductase and NADPH-dependent methylglyoxal reductase. The expression of GRE2 is induced during osmotic, heat shock, heavy metals, and oxidative stress conditions[65, 66]. As expected, we found upregulation of Hog1-regulated genes in flocculating cells in unstressed conditions (Fig. 1D) which further confirms the constitutive activation of the HOG MAPK pathway in flocculating cells.
2.1 The expression of FLO genes is induced by Hog1 Phosphorylation in wild-type cells.
The above results suggest that the Hog1 MAPK pathway is constitutively active in flocculating strains. To further investigate the involvement of this pathway in yeast flocculation, we artificially induced the Hog1 phosphorylation and checked the expression of FLO genes. The wild-type cells were treated with sodium chloride (NaCl), a well-known osmotic stress inducer that induces Hog1 phosphorylation [67]. First, we checked the Hog1 phosphorylation in wild-type cells BY4743, treated with different concentrations of NaCl (0.5, 0.8, and 1.0M) for 1 hour. We observed a gradual increase in Hog1 phosphorylation with increasing concentrations of NaCl (Fig. 2A). Further, we tested the expression of FLO genes (FLO1, 5, and 9) in wild-type cells treated with 0.5, 0.8, and 1.0 M NaCl for 1 hour. The FLO1, FLO5, and FLO9 are a member of the subtelomeric genes family which encodes specific lectin proteins that are responsible for yeast flocculation [68]. The qRT-PCR analysis indicates a significant induction in the expression of FLO genes as shown in Fig. 2B, which suggests that the HOG1 MAPK pathway positively regulates the yeast FLO genes. For further validation that Hog1 phosphorylation induces FLO genes expression, we used other well-known inducers of Hog1 phosphorylation such as potassium chloride and arsenic (III) oxide [69] (Ref.). After the treatment, we checked the FLO1 expression and found the upregulation in the FLO1 expression as expected (Fig. 2C). Based on this data if Hog1 phosphorylation is important for FLO genes expression then deletion of HOG1 in wild-type cells, should reduce the FLO genes expression. For this, we deleted the HOG1 gene in WT cells and perform the qPCR to check the FLO expression. Interestingly we found a significant reduction in FLO1 expression in HOG1-deleted cells compared the wild-type cells (Fig. 2D). We also tested the flocculation phenotypic analysis of cells upon Hog1 phosphorylation. For this, we performed the plate assay upon 0.8 M NaCl treatment of wild-type cells but we did not find any significant induction in the flocculation phenotype (Fig S1A). The possible reason for the insignificant flocculation of wild-type cells upon NaCl treatment could be due to the suppression of FLO genes by the Tup1-Cyc8 repressor complex. This observation suggests that 2–5 fold upregulation of the FLO genes upon NaCl exposure is not sufficient to induce a strong visible flocculation phenotype. Whereas in the TUP1/CYC8 deleted cells, in absence of a repressor complex, FLO genes are highly expressed (150 to 200 fold), compared to wild-type cells and show hyper-flocculation (Fig. S1B).
2.2 Hog1p occupancy at FLO1 promoter increased while Tup1 occupancy decreased upon Hog1 phosphorylation in wild-type cells
Gene expression is a well-regulated process in which target-specific transcription regulators either activators or repressors bind to the promoter and regulates the chromatin remodelers’ binding, to modulate the gene expression. Based on the physiological conditions and growth phase one or more transcription regulators bind to the FLO promoter either permanently or temporarily to regulate the FLO genes expression [2]. As discussed in the introduction part, the Tup1-Cyc8 complex binds to the promoter of the FLO genes and represses their expression. The FLO1, FLO5, and FLO9 are major FLO genes responsible for yeast flocculation, and FLO8 works as a transcription activator for FLO1, and FLO11 [68, 70]. In most of the laboratory strains (W303-1A and S288C) the FLO8 gene is mutated by nonsense mutation to create transcription silencing of flocculation genes [71]. As in the S288C background, FLO8 is absent but based on our data shown in Fig. 2B, Hog1 phosphorylation induces the FLO genes expression. To narrow down the Hog1 role in FLO genes regulation, we checked the Hog1 occupancy at the FLO1 promoter upon NaCl treatment through chromatin immunoprecipitation (ChIP) assay. For the ChIP experiment, we treated the wild-type cells with 0.8 M NaCl for 1 hour, and ChIP DNA was amplified using the primer sets of FLO1 promoter (ranging + 14 to -817) by qRT-PCR (Fig. E). Based on the qRT-PCR analysis, we found significant enrichment in Hog1 occupancy at the FLO promoter upon NaCl treatment (Fig. 2F). It is well established that the Tup1-Cyc8 corepressor complex is a potent repressor of FLO genes expression and our above data also suggest that upon NaCl exposure FLO genes expression increased. Next in this line with this, we decided to check the Tup1 occupancy at the FLO1 promoter upon NaCl treatment. As shown in Fig. 2G we found a significant reduction in Tup1 occupancy at the promoter upon NaCl treatment (Fig. 2G). These results suggest that under the stress conditions phosphorylated Hog1p bind to the FLO1 promoter and work as a transcription activator for FLO genes expression and its binding decrease/inhibit the binding of Tup1p at the FLO promoter which is important for the induction of FLO genes.
3. Flocculating cells show hypersensitivity to GPI-anchored proteins targeting molecule, cantharidin.
It is a well-established fact that flocculins are GPI-anchored proteins, involved in flocculation through cell-to-cell interaction (Fig. 3A) [30]. GPI-anchored proteins have different functions in different organisms. In higher eukaryotes (mammals), GPI-anchored proteins are involved in fertilization, embryonic development, neurogenesis, immune response, etc. [72, 73].In yeast cells, GPI-anchored proteins are involved in the regulation of cell wall composition and architecture maintenance, adhesion, invasion, and yeast flocculation. To get more insight, we treated the flocculating cells (tup1∆/∆ and cyc8∆/∆) with a chemical compound, cantharidin that is known to affect the GPI-anchored protein sorting. Cantharidin is a terpenoid that is produced by blister beetles and targets the remodeling processes of GPI-proteins by affecting the Cdc1 activity in ER and impair their sorting which causes aggregation of GPI-anchored proteins [74]. In yeast cells, the cantharidin-resistant gene CRG1 and ABC transporter Pdr5 are required for cantharidin tolerance [75]. Therefore, we first checked the growth of flocculating cells upon cantharidin exposure. For this, we performed the spot assay analysis of wild-type cells and flocculating mutants in the presence of different concentrations of cantharidin. Interestingly we found that flocculating cells (tup1∆/∆ and cyc8∆/∆) were hyper-sensitive in the presence of cantharidin as shown in Fig. 3B. For the further validation of sensitivity toward cantharidin, we performed the growth curve assay and found cells to be hypersensitivity (Fig. 3C, D). These data suggest that the sensitivity of flocculating cells might be due to the altered sorting of GPI-proteins. To check the effect of cantharidin on flocculation phenotype, we performed the tube assay. We found a significant reduction in the flocculation phenotype of flocculating (tup1∆/∆ and cyc8∆/∆) cells upon cantharidin treatment (Fig. 3E, F). These data suggest that cantharidin suppresses yeast flocculation by affecting sorting of GPI-anchored flocculin proteins.
4. Cantharidin suppresses Hog1 phosphorylation and FLO gene expression
As cantharidin suppresses yeast flocculation, we further investigated to find the mechanism. Our earlier data clearly suggest (Fig. 2B, D) the involvement of Hog1 phosphorylation in yeast flocculation. Further, we wanted to check whether cantharidin affects Hog1 phosphorylation. To this end, we treated the wild-type and flocculating cells (tup1∆/∆ and cyc8∆/∆) with different concentrations of cantharidin in the liquid medium. To determine the sublethal concentration of cantharidin for the liquid medium, the colony-forming unit assay was performed. Cells were treated with different concentrations of cantharidin for 2 hours and plated. As shown in Fig. S2, cells treated with 15 µM and 30 µM of cantharidin for 2 hours, showed around more than 90% viability. Based on the CFU assay, we decided to use 15µM and 30µM concentrations of cantharidin for testing the effect on Hog1 phosphorylation. To our surprise, we found a drastic reduction in the Hog1 phosphorylation upon cantharidin treatment in flocculating (tup1∆/∆ and cyc8∆/∆) cells as compared to untreated cells (Fig. 4A, B). As we showed previously (Fig. 2D) phosphorylated Hog1 binds to the promoters of FLO genes causing an increase in expression. In this direction, as cantharidin suppresses the Hog1 phosphorylation, we wanted to test whether or not cantharidin suppresses the expression of FLO genes. The expression of FLO genes was measured by qRT-PCR and semi-qPCR in cantharidin-treated cells. Interestingly we found a significant reduction in FLO1, 5, and 9 expressions as expected (Fig. 4C-E). These results confirm that the HOG1 MAPK pathway positively regulates the FLO gene expression and the phosphorylated form of Hog1 is required for the proper expression of FLO genes. As cantharidin suppresses the Hog1 phosphorylation, the reduction in the flocculation phenotype of cells (tup1∆/∆ and cyc8∆/∆) was observed (Fig. 3E and F) upon treatment.
As cantharidin suppresses the Hog1 phosphorylation, we wanted to check whether cantharidin effects are only limited at the protein level or is it affecting the HOG1 transcript levels also. To explore this possibility, we checked the HOG1 mRNA expression with and without cantharidin exposure. We found that in untreated conditions, HOG1 mRNA expression in flocculating cells (tup1∆/∆ and cyc8∆/∆) is significantly lower than in wild-type cells. Interestingly, upon cantharidin treatment, the HOG1 mRNA level increases significantly in wild-type and flocculating mutant cells although the phosphorylation of Hog1 protein was reduced upon treatment (Fig. 4F). This suggests that cantharidin does not suppress the HOG1 expression, it prevents the phosphorylation of the Hog1 protein. Further, to check the other possibility which is Hog1 phosphatases expression, we hypothesized that the up-regulation of Hog1 phosphatases upon cantharidin treatment may contribute to the reduction in phosphorylation of Hog1. For this, we tested the expression of Hog1 phosphatases; PTC1, and PTP2. PTP1 encodes protein serine/ threonine phosphatase (type 2C protein phosphatase) whereas PTP2 are phospho-tyrosine-specific phosphatases. A qRT-PCR was performed which revealed that cantharidin exposure induces the expression of Hog1 phosphatases (Fig. 4F) resulting in a reduction in the Hog1 phosphorylation levels followed by decreased FLO expression (Fig. 4A-E). Altogether these results suggest that cantharidin targets the Hog1 phosphorylation by upregulation of phosphatases which ultimately reduces the FLO genes expression and flocculation.
5. Cantharidin abrogates the expression of GPI-anchored proteins and alters cell surface morphology of flocculating cells
Flocculins are glycosylphosphatidyl-inositol (GPI) anchored proteins and are required for yeast flocculation [30]. If the structure and expression of flocculins are altered, it will affect the flocculation phenotype. Our results suggest that cantharidin suppresses the Hog1 phosphorylation and expression of FLO genes resulting in the down-regulation of GPI-anchored protein (flocculins). To further re-confirm this analysis, we checked the protein level of another GPI-anchored protein, Gas1. Gas1 is a GPI-anchored cell wall protein that is required for cell wall biogenesis, assembly, and maintenance [76]. It is mainly localized to the cell surface but is also present at the nuclear periphery[77]. To test this, first, we transformed the wild-type and flocculating cells with a pRS415-Gas1-GFP plasmid. Upon cantharidin treatment (15 µM and 30 µM) for 2 hours, western blotting was performed to check the expression of Gas1p. As expected, we observed less expression of Gas1p in flocculating strains compared to wild-type cells upon cantharidin treatment (Fig. S3).
This observation suggests that cantharidin affects the cell surface GPI- anchored proteins which can affect cell surface morphology. To confirm this, we checked the cell morphology of wild-type and flocculating TUP1 deleted cells by scanning electron microscopy. In the untreated conditions, the wild-type cells exhibited an oval shape and smooth cell surface whereas many of the flocculating cells showed a rough surface, and two adjacent cells were attached to each other through a thread-like structure (Fig. 5). To better understand the effect of cantharidin on cell morphology, next we treated the cells with cantharidin for 3 hours and check the morphology by scanning electron microscopy. Interestingly, we found that the roughness of flocculating cells is decreased compared to untreated cells. We also observed that most of the cantharidin-treated cells exhibited regional invaginations in the cell wall (Fig. 5). This data suggests that cantharidin alters the cell surface morphology and structure and decreases the cell-to-cell interaction causing a decrease in yeast flocculation phenotype.
6. Deletion of HOG1 in flocculating yeast cells shows a significant reduction in yeast flocculation
As our above data suggests that Hog1 phosphorylation is essential for yeast flocculation, we hypothesized that the deletion of Hog1 should reduce the flocculation in flocculating cells. In this line first, we delete the HOG1 gene in flocculating TUP1∆/∆ and CYC8∆/∆ cells but even after multiple attempts we did not succeed. The failure to delete the HOG1 gene could be due to the diploid background (BY4743) as it is difficult to delete two copies of a gene together in the same cells. To resolve this problem next, we used a haploid background (BY4741) flocculating yeast strain, TUP1∆, and CYC8∆. Upon HOG1 deletion in TUP1∆, and CYC8∆, we found a reduction in FLO genes expression (Fig. 6A). and a significant difference in flocculation phenotype in plate assay (Fig. 6B). As it is mentioned before that Hog1 phosphorylation induces the flocculation and flocculating mutants shows constitutive phosphorylation (Fig. 4A, B). We thought if we treat the flocculating cells with an inducer (NaCl) of Hog1 phosphorylation then, Hog1 will be more phosphorylated and flocculation will be increased further. To achieve this, we treated the flocculating cells with 0.5M sodium chloride. On contrary to our proposed hypothesis, we found that upon NaCl treatment the flocs completely disappeared (Fig. 6C, D). Since NaCl affects cellular health in multiple ways, so possibly sodium chloride caused the cessation of cell growth, triggered changes in cell morphology, and reduced surface hydrophobicity. In flocculating strains NCYC 1195 (an ale-brewing strain) reduced hydrophobicity has been correlated with the reduction in flocculation phenotype. Since TUP1∆and CYC8∆ are also flocculating like NCYC 1195, the reduction in flocculation might be due to morphological changes in these flocculating cells [59].
7. HOG1 Deletion in flocculating H3R63A cells shows a significant reduction in yeast flocculation.
As discussed above that in flocculating TUP1∆ and CYC8∆ cells, the repressor complex of FLO genes is absent so it is difficult to find a significant difference in flocculation phenotype upon HOG1 deletion in these mutants. To confirm the role of Hog1 phosphorylation in yeast flocculation we used another flocculating mutant in which repressor complex Tup1-Cyc8 is present. For this purpose, we used histone H3R63A mutant which shows flocculation. Histone H3R63A is a mutant of histone H3 in which R (Arginine) is replaced by A (Alanine). However, the flocculation phenotype intensity of this mutant is less than TUP1∆/∆ cells but it was significant, compared to the wild-type cells (Fig. 7A). In this direction, before deleting the HOG1 in this mutant we first check the Hog1 phosphorylation to confirm whether, in this flocculating mutant, Hog1 is constitutively phosphorylated or not. Interestingly we found the constitutive phosphorylation in this mutant also (Fig. 7B), similar to other flocculating (TUP1∆/∆ and CYC8∆/∆) cells (Fig. 1B). Next, we deleted the HOG1 gene in this mutant strain. Upon HOG1 deletion, we performed qRT-PCR to check the FLO gene expression and qRT-PCR data analysis indicates a significant reduction in FLO1 and FLO5 gene expression (Fig. 7C). We also checked the flocculation at the phenotypic level. For this, we performed the plate assay to check the difference in flocculation and we found a significant reduction in flocculation in HOG1 deleted cells (Fig. 7D), which is consistent with previous observations. Next, we checked the Hog1 and Tup1 occupancy at the FLO1 promoter in histone H3 wild-type and H3R63A mutant. For this, ChIP DNA was amplified using the primer sets of FLO1 promoter (ranging from + 14 to -817) by qRT-PCR. Based on the qRT-PCR analysis, we found significant enrichment in Hog1 occupancy at the FLO promoter in the H3R63A mutant compared to wild-type cells (Fig. 7E). We also found a significant reduction in Tup1 occupancy at the FLO promoter in this mutant (Fig. 7F). This data clearly suggests that Hog1 is essential for the FLO gene expression, and deletion of HOG1 or suppression of Hog1 phosphorylation significantly affects yeast flocculation. Phosphorylated Hog1 works as a transcription activator for FLO genes and its binding at the FLO promoter is important for the induction of FLO genes.