Synergistic roles of the phospholipase B homolog Plb1 and the cAMP-dependent protein kinase Pka1 in the hypertonic stress response of Schizosaccharomyces pombe

The phospholipase B homolog Plb1 and the cAMP-dependent protein kinase (PKA) pathway are required by fission yeast, also known as to Schizosaccharomyces pombe, to grow under KCl-stress conditions. Here, we report the relative contributions of Plb1 and the cAMP/PKA pathway during the hypertonic stress response. We show that the plb1∆, cyr1∆, and pka1∆ single mutants are sensitive to high concentrations of KCl but insensitive to sorbitol-induced osmotic stress. In contrast, the plb1∆ cyr1∆ and plb1∆ pka1∆ double mutants are hypersensitive to KCl and sorbitol. The cyr1∆ pka1∆ double mutants showed the same phenotype of each single mutant. Growth inhibition due to hypertonic stress in the plb1∆, plb1∆ cyr1∆, and plb1∆ pka1∆ strains was partially rescued by cgs1 deletion—cgs1∆ has constitutively active Pka1—or by the deletion of transcription factor Rst2, which is negatively regulated by Pka1. Pka1-GFP localized in the nucleus and cytoplasm in plb1∆, whereas it is localized only in the cytoplasm in cyr1∆, indicating that Plb1 does not regulate Pka1 localization. Glucose limitation downregulates the PKA pathway, and it was accordingly observed that glucose limitation in plb1∆ further increased the strain’s sensitivity to KCl. Growth inhibition by KCl in plb1∆ under glucose-limited conditions was significantly rescued by cgs1∆ and slightly rescued by rst2∆. These findings indicate that, in fission yeast, Plb1 and the glucose-sensing cAMP/PKA pathway play a synergistic role in responding to hypertonic stress.


Introduction
Signal transduction pathways play an important role in adapting to various environmental conditions. In the fission yeast, Schizosaccharomyces pombe, the stress-activated protein kinase (SAPK), the target of rapamycin (TOR), and the cAMP-dependent protein kinase (PKA) pathways respond to various environmental conditions. The SAPK pathway is activated in response to osmotic, oxidative, or heat stress (Shieh et al. 1997(Shieh et al. , 1998Shiozaki and Russell 1996;Wilkinson et al. 1996). The TOR pathway is regulated in response to nutrient starvation, such as nitrogen or glucose limitation (Halova et al. 2013;Saitoh et al. 2015). The cAMP/PKA pathway mainly responds to glucose concentration (Hoffman 2005;Inamura et al. 2021;Welton and Hoffman 2000).
Phospholipase B (Plb1) hydrolyses the acyl chains sn1 and sn2, and has fatty acid acyltransferase activity that converts lysophospholipids to phospholipids (Ansell and Hawthorne 1964). In S. pombe, plb1∆ shows a high frequency of mated cells and a KCl-sensitive phenotype that is suppressed by Git3 or Gpa2 overexpression (Yang et al. 2003). Based on these results, Yang et al. suggested that Plb1 is a mediator of an osmotic stress response that is linked to the cAMP/ PKA pathway. However, the relationship between Plb1 and the cAMP/PKA pathway is still unclear due to insufficient analyses.
In this study, we characterize the genetic relationship between Plb1 and the cAMP/PKA pathway in the context of the hypertonic stress response. We also show that Pka1 activation rescued the hypertonic-stress-induced growth defect of plb1∆. Glucose limitation enhanced the KCl-sensitive phenotype of plb1∆, and a gain-of-function of Pka1 suppressed this effect. Finally, we show that Plb1 is not involved in regulating Pka1 localization or phosphorylation. Together, our findings demonstrate that Plb1 and the cAMP/PKA pathway play a synergistic role in responding to hypertonic stress in S. pombe.

Fluorescence microscopy of GFP fusion protein
S. pombe cells were grown in YES liquid medium to the mid-log phase at 30 °C. GFP-tagged Plb1 and Pka1 were visualized in living cells, and images were taken by a BX51 microscope (Olympus) equipped with a DP74 digital camera (Olympus).

Preparation of cell lysates and detection of 13Myc fusion protein using immunoblotting
S. pombe cell lysates were prepared as previously described (Matsuo et al. 2004). Lysate proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), after which western blot analysis was performed using a Western Lightning enhanced chemiluminescence (ECL) Pro detection system (PerkinElmer) according to the supplier's instructions. Mouse monoclonal anti-Myc (diluted 1:1000) and rabbit polyclonal anti-PSTAIRE (Cdc2; diluted 1:1000) antibodies were purchased from Santa Cruz Biotechnology. Horseradish-peroxidase-conjugated antimouse IgG (Santa Cruz Biotechnology) and anti-rabbit IgG antibody (Promega) were used as secondary antibodies.

Pka1 overexpression rescues the KCl-sensitive phenotype of the plb1∆ strain
It has been proposed that the cAMP/PKA pathway is a downstream target of Plb1 during the osmotic stress response induced by a high concentration of KCl (Yang et al. 2003). To clarify how Plb1 might affect the cAMP/PKA pathway, we first analyzed cell growth in the presence of a high concentration of KCl in a plb1∆ strain with Pka1 overexpression and in a pka1∆ strain with Plb1 overexpression. Pka1 overexpression rescued the growth defect of plb1∆ in the presence of 1.2 M KCl (Fig. 1a), whereas Plb1 overexpression did not recue the growth defect of pka1∆ (Fig. 1b). These results are compatible with the idea that the cAMP/PKA pathway is downstream of Plb1 in the context of responding to KCl stress.

S. pombe mutants lacking Plb1 and PKA are hypersensitive to hypertonic stress
We next analyzed the growth of single and double mutants under high-osmolarity stress. If Plb1 linearly affects the  (2017) This study cAMP/PKA pathway, we expected that the plb1∆ cyr1∆ and plb1∆ pka1∆ double mutants would show the same KCl-sensitivity as the respective single mutants. Likewise, if cAMP only regulates PKA, we predicted that the cyr1∆ pka1∆ strain would exhibit the same sensitivity to hypertonic stress as the cyr1∆ and pka1∆ single mutants. The plb1∆ cyr1∆ and plb1∆ pka1∆ double mutants significantly exhibited the KCl-sensitive phenotype on a low concentration KCl (0.5 M), whereas the plb1∆, cyr1∆, and pka1∆ single mutants showed the KCl-sensitive phenotype on a high concentration of KCl (1.2 M). However, while the cyr1∆ pka1∆ strain grew normally on 0.5 M KCl, it showed the KCl-sensitive phenotype on 1.2 M KCl (Fig. 2a). The double mutants plb1∆ cyr1∆ and plb1∆ pka1∆ displayed the sorbitol-sensitive phenotype on 2 M sorbitol, while plb1∆, cyr1∆, pka1∆, and cyr1∆ pka1∆ mutants grew normally on the same plate (Fig. 2a). These results demonstrate that Plb1 and the cAMP/PKA pathway respond cooperatively to hypertonic stress. Loss of functional Pka1 inhibits cell growth on media containing CaCl 2 (Matsuo and Kawamukai 2017). We next investigated whether the single and double mutants are sensitive to CaCl 2 and NaCl. While cyr1∆ and pka1∆ showed growth defects with 0.3 M CaCl 2 , wild type and plb1∆ cells grew normally (Fig. S2). All the single and double mutants showed normal growth with 0.1 M CaCl 2 or 0.2 M NaCl (Fig. S2), indicating that Plb1 is not involved in responding to CaCl 2 .
Git3 or Gpa2 overexpression rescues the growth defect of plb1∆ under KCl-stress conditions (Yang et al. 2003). To better understand the relationship between Plb1 and the upstream component of the cAMP/PKA pathway, we analyzed the effect of the KCl stress on the plb1∆ git3∆ strain. This double mutant significantly exhibited the KCl-sensitive phenotype with a low concentration of KCl (0.5 M), but the plb1∆ and git3∆ single mutants grew normally on a plate containing 0.5 M KCl, only showing the KCl-sensitive phenotype with 1.2 M KCl (Fig. 2b). This indicates that Plb1 functions independently of the upstream cAMP/ PKA pathway in the context of KCl stress.

Gain of functional Pka1 partially recues growth defects of plb1∆, cyr1∆, and cyr1∆ plb1∆ under hypertonic stress conditions
It has been shown that cgs1∆ rescued growth defects associated with the CaCl 2 -and TBZ-sensitive phenotypes of the cyr1∆ strain (Matsuo and Kawamukai 2017;Tanabe et al. 2019). These results suggest that gain of functional Pka1 would rescue the stress-sensitive phenotype of cyr1∆, and also of plb1∆ if Plb1 is indeed upstream of Pka1 in the context of hypertonic stress response. We investigated whether deleting cgs1 suppresses the growth defects of plb1∆, cyr1∆, and plb1∆ cyr1∆ upon KCl stress. If Plb1 is required for modulating Pka1 activity upon KCl stress, deleting cgs1 would completely rescue the hypertonic-stress-sensitive phenotype of plb1∆. As could be expected, we found that cgs1∆ significantly recued the KCl-sensitive phenotype of cyr1∆ with both 1.2 and 1.5 M KCl (Fig. 3a). Additionally, cgs1∆ rescued the growth of plb1∆ with 1.2 M KCl but not with 1.5 M KCl (Fig. 3b). Intriguingly, cgs1∆ partially rescued the growth defect of plb1∆ cyr1∆ upon KCl stress and fully rescued the sorbitol-sensitive phenotype (Fig. 3b). These results indicate that while Pka1 activity is required for hypertonic stress response, Plb1 is not dispensable when responding to more serve hypertonic stress. Loss of functional Rst2 rescues growth defects of plb1∆, pka1∆, and plb1∆ pka1∆ under hypertonic stress conditions The transcription factor Rst2 is negatively regulated by Pka1 (Inamura et al. 2021;Takenaka et al. 2018). Loss of functional Rst2 rescues the phenotypes such as high expression of mug14 and ste11 mRNAs of the pka1∆ strain (Higuchi et al. 2002;Inamura et al. 2021). To investigate the role of Rst2 in hypertonic stress response, we next tested if rst2∆ rescues the growth defects of plb1∆, pka1∆, and plb1∆ pka1∆ under the hypertonic stress conditions. Our results showed that rst2∆ rescued the KCl-sensitive phenotype of plb1∆ with 1.0 M KCl but did not with 1.5 M KCl (Fig. 4a). Meanwhile, rst2∆ completely rescued the growth defect of pka1∆ on 1.5 M KCl (Fig. 4b). Finally, rst2∆ partially rescued the growth defects of plb1∆ pka1∆ on KCl-containing plates and fully rescued the growth defects on sorbitol-containing plates (Fig. 4c). These results indicate that Rst2 is the downstream target of the cAMP/PKA pathway in the context of responding to KCl stress, and that Plb1 is not dispensable when responding to higher levels of KCl.

Plb1 does not regulate Pka1 localization
We next analyzed the localization of Plb1 in the presence of KCl stress. We first made the plb1-GFP strain by tagging GFP at the C-terminus of Plb1, but we did not observe GFP fluorescence (data not shown). To overcome this problem, we next constructed a GFP-Plb1-expressing plasmid, in which Plb1 tagged with GFP at its N-terminus is expressed under the nmt41 promoter. GFP-Plb1 rescued the growth of the plb1∆ strain upon KCl stress, indicating that GFP-Plb1 is functional (Fig. S3). Plb1 localized at the endoplasmic reticulum (ER) and highly accumulated as the dots under normal conditions. Upon KCl stress, Plb1 localization to the ER did not change, but its localization to the dots decreased (Fig. 5a). Because the KCl-sensitive phenotype of plb1∆ is dependent on Pka1 activity, we next analyzed whether Pka1 localization is regulated by Plb1. Pka1-GFP mainly a b  localizes in the nucleus and diffusely in the cytoplasm in wild-type cells (Matsuo et al. 2008). In the presence of inactive endogenous Pka1 in the cyr1∆ strain, Pka1-GFP only localized to the cytoplasm. In the context of constitutive Pka1 activation in cgs1∆ under normal conditions, Pka1-GFP localized to the nucleus and diffusely to the cytoplasm as observed in the wild-type strain (Matsuo et al. 2008 and Fig. 5b). In plb1∆, Pka1-GFP exhibited the same localization as in wild type and cgs1∆ cells. Under KCl stress, the localization of Pka1-GFP changed from the nucleus to the cytoplasm in wild-type cells, and still localized only in the cytoplasm in cyr1∆. In cgs1∆, the nuclear localization of Pka1-GFP decreased, and we observed accumulation as dots distributed through the cytoplasm. Under KCl-stress conditions in plb1∆, Pka1-GFP changed its localization to the cytoplasm, as observed in wild-type   (Fig. 5b, S4). These results indicate that Plb1 does not regulate Pka1 localization, suggesting that Plb1 likely is not involved in the regulation of Pka1 activity. We next tested whether deleting plb1 would affect Pka1 phosphorylation. Pka1 is phosphorylated at threonine 356 in cyr1∆, whereas this phosphorylation site is not detected in wild type or cgs1∆ strains (Gupta et al. 2011a;McInnis et al. 2010). Under normal conditions, Pka1-13Myc was detected in a non-phosphorylated form in wild type and cgs1∆ cells, while it was phosphorylated in the cyr1∆ strain (Fig. 5c). However, Pka1-13Myc was not phosphorylated under normal conditions in plb1∆ (Fig. 5c)

Glucose limitation induces growth defects in wild type and plb1∆ upon the KCl stress
Because the cAMP/PKA pathway is tightly regulated by extracellular glucose concentration (Byrne and Hoffman 1993;Inamura et al. 2021;Tanabe et al. 2020;Welton and Hoffman 2000), we next analyzed whether glucose limitation induces the KCl-sensitive phenotype. Wild type, plb1∆, and pka1∆ grew, but plb1∆ pka1∆ did not grow, on glucose-rich (3% glucose) medium with 0.3 M KCl (Fig. 6a). Additionally, plb1∆, pka1∆, and plb1∆ pka1∆ exhibited a KCl-hypersensitive phenotype on low-glucose (0.1% glucose) medium than on normal glucose (3% glucose) medium (Fig. 6a). Next, we analyzed the localization of GFP-Plb1 under the glucose-limited conditions. Localization of GFP-Plb1 showed the same pattern, localization at the ER and in the dots, in the glucose-rich (3% glucose) and glucose-limited (0.1% glucose) media, but the dots structure was decreased in the glucose-limited media (Fig. 6b), indicating that glucose levels affects Plb1 localization.

Gain of functional Pka1 rescues the KCl-sensitive phenotype of plb1∆ in glucose-limited media
We next tested whether gain of functional Pka1 affects growth during KCl stress under glucose-limited conditions. To do this, we compared the effect of cgs1 deletion in plb1∆ and plb1∆ cyr1∆ under glucose-limited conditions. While plb1∆ and plb1∆ cyr1∆ significantly exhibited the KCl-sensitive phenotype with 0.3 M KCl, plb1∆ cgs1∆ and plb1∆ cyr1∆ cgs1∆ grew on 0.5 M KCl under glucose-limited conditions (Fig. 7a). These results indicate that increased Pka1 activity overcomes sensitivity to KCl stress in plb1∆ under glucose-limited conditions.
We finally analyzed whether rst2∆ rescues the KCl-sensitive phenotype in the context of glucose limitation. The plb1∆ and plb1∆ pka1∆ strains exhibited the KCl-sensitive phenotype with 0.3 M KCl under glucose-limited conditions. Under the same conditions, the plb1∆ rst2∆ and plb1∆ pka1∆ rst2∆ strains grew, and the plb1∆ pka1∆ rst2∆ strain also grew with 0.5 M KCl under glucose-limited conditions (Fig. 7b). Therefore, loss of functional Rst2 rescued growth during KCl stress under glucose-limited conditions.

Discussion
Cells encounter various environmental stresses and use signal transduction systems to adapt to such stresses. In this study, we focused on the relationship between KCl stress and the cAMP/PKA pathway in fission yeast. Our findings show that Plb1 (a phospholipase B homolog) and the cAMP/PKA pathway play a cooperative role in the KCl-stress response. This idea contradicts the previously proposed idea that Plb1 acts upstream of the cAMP/PKA pathway (Yang et al. 2003), which was based on the observation that overexpression of Gpa2 or Git3 rescued the KCl-sensitive phenotype of plb1∆. We also observed that Pka1 overexpression rescued the KCl-sensitive phenotype of plb1∆ (Fig. 1). However, our observations that plb1∆ enhanced the KCl-sensitive phenotype of both pka1∆ (Fig. 2a) and of git3∆ (Fig. 2b) indicate that Plb1 and the cAMP/PKA pathway act separately to counteract KCl stress. Git3 is the G-protein-coupled receptor that senses extracellular glucose and transfers the signal through Gpa2 to the cAMP/PKA pathway (Fig. S1), which is downregulated by glucose limitation and upregulated by higher glucose concentration (Hoffman 2005;Inamura et al. 2021;Matsuo et al. 2008;Tanabe et al. 2020;Welton and Hoffman 2000). It is therefore unlikely that Plb1 interacts with the glucose-sensing cAMP/PKA pathway due to our observation that deletion of git3 enhanced the KCl-sensitive phenotype of plb1∆. Growth inhibition by KCl in plb1∆ was also enhanced by glucose limitation (0.1% glucose) when compared to that by normal conditions (3% glucose) (Fig. 6a, 7) supporting the idea that the cAMP/PKA pathway works independently from Plb1. Cgs1 negatively regulates Pka1 by forming a heterotetramer with Pka1, preventing it from being active. Therefore, deletion of cgs1 results in constitutive Pka1 activity (Gupta et al. 2011a). Our genetic observation regarding cgs1 further strengthen the idea that the glucose-cAMP-PKA pathway contributes to the KCl-stress response: (1) Deletion of cgs1 rescued the 1.5 M KCl-sensitive phenotype of cyr1∆ to wild-type levels (Fig. 3a).
The transcription factor Rst2 is negatively regulated by Pka1 (Higuchi et al. 2002;Inamura et al. 2021). Our observation of rst2∆ combined with various other deletion mutants indicate that Rst2 is important for KCl tolerance. Deleting rst2 partially rescued the growth defects of plb1∆ and plb1∆ pka1∆ with 1.0 M KCl, whereas it fully rescued pka1∆ with 1.5 M KCl (Fig. 4). These observations are conserved under glucose limitation. We also observed that plb1∆ cyr1∆ cgs1∆ and plb1∆ pka1∆ rst2∆ exhibited the same growth rate under KCl stress as cgs1∆ and rst2∆, respectively (Fig. 7). These results also support the idea that the glucose-cAMP-PKA pathway plays a major role in the KCl-stress response.
Localization of Pka1 changed from the nucleus to the cytoplasm with 1.2 M KCl in both wild type and plb1∆ (Fig. 5b). Loss of functional Plb1 did not affect the phosphorylation profile of Pka1 (Fig. 5c), consistent with the idea that Plb1 does not directly regulate the cAMP/PKA pathway. Pka1 is hyperphosphorylated in the cytoplasm when it is constitutively inactivated via deletion of cyr1 (Fig. 5b, c, and Gupta et al. 2011a, b, Matsuo et al. 2008, McInnis et al. 2010). In the Pka1-activated cgs1∆ strain, we could not detect phosphorylated Pka1 by western blotting, but we observed Pka1 localized to the cytoplasm and nucleus upon KCl stress (Fig. 5b, c), as observed previously (Matsuo et al. 2008;McInnis et al. 2010). Cytoplasmic localization of Pka1 prevents it from negatively regulating Rst2, and Rst2 therefore remains in an active form. Our results consistently indicate that the glucosesensing pathway mediated by cAMP/PKA regulates Rst2 to responds to KCl stress.
We also would like to highlight our observation that glucose limitation enhanced the KCl-sensitive phenotype of pka1∆ and plb1∆ pka1∆ (Fig. 6a), suggesting that cAMP-PKA is not alone in responding to glucose limitation. Responding to glucose limitation is likely additionally mediated by other glucose-responding pathways such as the SAPK and TORC2 pathways (Cohen et al. 2014;Fraile et al. 2020;Ikai et al. 2011;Saitoh et al. 2015;Sanchez-Mir et al. 2018;Zuin et al. 2010). While we have not conducted any experiments investigating these pathways in the context of glucose limitation, many other works on the SAPK pathway clearly indicate it is important for the KCl-stress response (Matsuo and Kawamukai 2017;Matsuo et al. 2008;Shieh et al. 1997Shieh et al. , 1998Shiozaki and Russell 1996;Stiefel et al. 2004;Yang et al. 2003). Deleting sty1 causes sensitivity to 0.6 M KCl (Nunez et al. 2009), while cells lacking pka1 tolerates 0.7 M KCl (our unpublished data). The SAPK pathway is responsible for responding to lower concentrations of KCl, while cAMP/PKA is responsible for responding to higher concentrations (Matsuo and Kawamukai 2017;Matsuo et al. 2008). While loss of a functional SAPK pathway results in oxidative-, osmotic-, and heat-stress-sensitive phenotypes and decreases mating efficiency (Shieh et al. 1998;Shiozaki and Russell 1996), plb1∆ and pka1∆ do not show these phenotypes (Figs. 2a, 3b, 4c, S2 and Yang et al. 2003) and display a high frequency of mating and sporulation (Gupta et al. 2011a, b;Maeda et al. 1994;Yang et al. 2003). The SAPK pathway, cAMP/PKA pathway, and Plb1 are commonly involved in responding to only KCl stress. It is possible that the SAPK pathway is modulated by Plb1 since it has been shown that mRNA expression of gpd1 decreases in plb1∆ under KCl stress (Yang et al. 2003) and that it is upregulated by KCl through the SAPK pathway (Shieh et al. 1997;Wilkinson et al. 1996).
Our findings show that the glucose-activated cAMP/ PKA pathway regulates Rst2 to produce a hypertonic stress response. In S. pombe, this signal transduction pathway is particularly important when responding to higher concentrations of KCl. Our study demonstrates that Plb1 and the cAMP/PKA pathway cooperatively but independently function to respond KCl-induced hypertonic stress.