|Home | About | Journals | Submit | Contact Us | Français|
Yeast has numerous mechanisms to survive stress. Deletion of myosin type II (myo1Δ) in Saccharomyces cerevisiae results in a cell that has defective cytokinesis. To survive this genetically induced stress, this budding yeast up regulates the PKC1 cell wall integrity pathway (CWIP). More recently, our work indicated that TOR, another stress signaling pathway, was down regulated in myo1Δ strains. Since negative signaling by TOR is known to regulate PKC1, our objectives in this study were to understand the cross-talk between the TOR and PKC1 signaling pathways and to determine if they share upstream regulators for mounting the stress response in myo1Δ strains.
Here we proved that TORC1 signaling was down regulated in the myo1Δ strain. While a tor1Δ mutant strain had increased viability relative to myo1Δ, a combined myo1Δtor1Δ mutant strain showed significantly reduced cell viability. Synthetic rescue of the tor2-21ts lethal phenotype was observed in the myo1Δ strain in contrast to the chs2Δ strain, a chitin synthase II null mutant that also activates the PKC1 CWIP and exhibits cytokinesis defects very similar to myo1Δ, where the rescue effect was not observed. We observed two pools of Slt2p, the final Mitogen Activated Protein Kinase (MAPK) of the PKC1 CWIP; one pool that is up regulated by heat shock and one that is up regulated by the myo1Δ stress. The cell wall stress sensor WSC1 that activates PKC1 CWIP under other stress conditions was shown to act as a negative regulator of TORC1 in the myo1Δ mutant. Finally, the repression of TORC1 was inversely correlated with the activation of PKC1 in the myo1Δ strain.
Regulated expression of TOR1 was important in the activation of the PKC1 CWIP in a myo1Δ strain and hence its survival. We found evidence that the PKC1 and TORC1 pathways share a common upstream regulator associated with the cell wall stress sensor WSC1. Surprisingly, essential TORC2 functions were not required in the myo1Δ strain. By understanding how yeast mounts a concerted stress response, one can further design pharmacological cocktails to undermine their ability to adapt and to survive.
The calcium-dependent protein kinase (Pkc1p) and target-of-rapamycin (TOR) signaling pathways are conserved in yeast and other fungi and are important for stress response and fungal survival. In addition to regulating growth and metabolic activity in normal cells, these pathways also regulate the cellular response to transient cell wall stress during the normal yeast life cycle, and during exposure to heat shock, cell wall damage, or other stressors that can compromise cellular integrity [1-3]. Our studies with myosin type II-deficient (myo1Δ) strains of the budding yeast Saccharomyces cerevisiae, which we have characterized previously as stress mutants, showed that the Pkc1p pathway is activated and essential for myo1Δ strain survival [4-6]. It has been our contention that this activation is due to cell wall stress caused by morphological abnormalities in the lateral cell wall and bud neck architecture [7,8]. In response to cell wall damage, heat shock, and other types of environmental stress, Rho1p activates the PKC1 cell wall integrity pathway (CWIP), which in turn activates Slt2p (Mpk1p), the Serine/Threonine (Ser/Thr) MAPK at the end of this cascade [1-3]. This leads to transcriptional up regulation of cell wall-related genes by the Rlm1p transcription factor [9-12]. In addition to regulating the genetic program for cell wall integrity through the transcription factor Rlm1p [9,13,14], Slt2p may also modulate PKC1 activity indirectly by a previously proposed feedback mechanism that phosphorylates and down regulates the Rho1p GDP-GTP Exchange Factor (GEF) Rom2p . Rho1p also functions as the regulatory subunit of Fks1p, a β-1,3-glucan synthase for lateral cell wall fortification .
In prior studies, we have shown that similar to wild-type (wt) cells under stress conditions, the myo1Δ mutant (a genetically induced stress caused by the deletion of myosin II heavy chain that inhibits normal cytokinetic ring assembly) also activates the PKC1 CWIP, but uses a different repertoire of genes [4,5]. Further characterization of the genes of the myo1Δ mutant at the post-transcriptional level showed that only a subset of cell wall integrity genes was activated. Thus, the myo1Δ mutant may serve as a simplified model for studying the cell wall stress response. Furthermore, we found that translation and ribosome biogenesis were down regulated in the myo1Δ strain . This observation led us to investigate the role of TOR in the myo1Δ strain survival and how it may complement the reduced CWIP response.
Yeast TOR consists of two proteins - Tor1p and Tor2p - which are contained in two protein complexes TORC1 and TORC2 [18,19]. The TORC1 complex that is sensitive to rapamycin treatment contains proteins Tor1p or Tor2p, Kog1p, Tco89p and Lst8p [18,20-22]. TORC2 that is resistant to rapamycin treatment contains Tor2p, Avo1p, Avo2p, Avo3p, Bit61p, and Lst8p [18,20]. Recent subcellular localization studies showed that Tor1p was concentrated near to the vacuolar membrane while Tor2p was predominantly in punctuate structures near to the cytoplasmic surface of the plasma membrane . Their differences in composition, sensitivity to rapamycin, and cellular localization support the idea that they function as two separate complexes [18,20,23]. TOR is important for nutrient sensing and is believed to play an important role in life span extension [24-27]. While TOR is conserved structurally and functionally from yeast to human, their roles are not biologically identical and warrant careful characterization of TOR from both species.
Rho1p is regulated by two mechanisms, a TOR-independent mechanism that is activated by cell wall stress (discussed above) and a separate TORC2-dependent mechanism that regulates actin cytoskeleton reorganization through the Rho1p-dependent activation of PKC1 . In this latter pathway, Rom2p activity is indirectly modulated by the essential phosphatidylinositol kinase TORC2 via a GTPase switch consisting of Rho1p, Rho2p, Rom2p, and Sac7p (a GTPase activating protein of Rho1p) [29,30]. In short, the TORC2-dependent association of Rho1p (and Rho2p) with the Rom2p phosphatidylinositol-binding domain promotes Rom2p activation and downstream events [30,31]. In this manner Rom2p functions as the relay by which TORC2 regulates polarization of the actin cytoskeleton via Pkc1p. Therefore, functional interconnections between Pkc1p and TORC2 have been proposed through a mechanism integrated by the Rho1p-Rom2p complex [30,31] (Figure (Figure11).
The stress sensor proteins Wsc1p, Wsc2p, Wsc3p, Mid2p and Mtl1p are involved in the activation of cell integrity signaling [2,32-37]. These cell surface sensors span the plasma membrane and are attached to the extracellular cell wall. The Mid2p homologue Mtl1p, that shares 50% sequence identity with Mid2p, appears to have a minor role in PKC1 signaling . These sensors react differently under specific stress conditions . It has been reported that cells lacking WSC1 are hypersensitive to drugs interfering with the cell wall and plasma membrane like Calcofluor white, Congo red, Caspofungin, Chlorpromazine and tea tree oil [1,38-40]. Additionally, Wsc1p responds to hypo-osmotic and alkaline pH conditions [39,41]. A mid2Δ mutant is hypersensitive to pheromone treatment, is hyperresistant to Calcofluor white, tea tree oil and Congo Red, and it senses acidic conditions and vanadate [32,37,39,40,42-44]. Wsc1p and Mid2p are also involved in the response to heat shock [2,33,35,45]. WSC2 and WSC3 act as suppressors of mutants defective in glycerol synthesis , while Mtl1p is associated with response to oxidative stress and glucose starvation [46,47]. The Wsc family of proteins and Mid2p have been shown to interact with specific signaling proteins that transmit stress signals from the fungal cell wall sensors to the Pkc1p and TOR signaling pathways. For example, Rom2p, the GEF that regulates Pkc1p, physically interacts with Wsc1p, Wsc2p, and Mid2p to activate the PKC1 CWIP in the response to cell wall stress [36,48,49]. To define the nature of these signaling interactions in myo1Δ strains, we demonstrate here that TORC1 and Pkc1p activities were inversely correlated, which suggests cross-talk between the two pathways. Furthermore, we found that TORC1 was down regulated in myo1Δ strains by a mechanism that required expression of Wsc1p but not the other cell wall stress sensors. Surprisingly, Tor2p functions were not essential for survival in myo1Δ cells.
To test the hypothesis that the TORC1 pathway was down regulated in the myo1Δ strain, we measured the levels of phosphorylated or dephosphorylated Npr1p to assess the TORC1 status . TORC1 signaling regulates phosphorylation of the Ser/Thr protein kinase Npr1p at 22 potential phosphorylation sites [51,52]. At steady state, Npr1p is maintained inactive by phosphorylation  (Figure (Figure2A,2A, left diagram). Inhibition of TORC1 activity by nutrient starvation or application of the antiproliferative drug rapamycin results in dephosphorylation and subsequent activation of Npr1p by the protein phosphatase Sit4p [50,51,53] (Figure (Figure2A,2A, right diagram). To detect Npr1p in the myo1Δ strains, an expression plasmid containing a functional N-terminal hemagglutinin (HA)-tagged NPR1 gene (HA-NPR1) was transformed into wt and mutant strains .
The phosphorylation state of Npr1p can be deduced from its relative electrophoretic mobility on SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and detected by Western blot analysis using an anti-HA antibody. In wt whole cell extracts, Npr1p was detected as a slower-migrating band, which corresponds to the hyperphosphorylated form, Npr1pP (Figure (Figure2B,2B, lane 1). Treatment of wt whole cell extracts with exogenous Calf Intestinal Alkaline Phosphatase (PPase) converted the slower-migrating band to a faster-migrating band which corresponds to the in vitro dephosphorylated form, Npr1p (Figure (Figure2B,2B, lane 2). Treatment of wt cell cultures with rapamycin produced a faster-migrating Npr1p band that co-migrated with the PPase treated band, consistent with the inhibition of TORC1 activity by rapamycin (Figure (Figure2B,2B, lane 3). This experiment established that the activity of TORC1 could be assessed indirectly by observing the relative electrophoretic mobility of Npr1p by SDS-PAGE [50,53].
In myo1Δ whole cell extracts, no change in the electrophoretic mobility of Npr1p was observed as judged by the co-migration of Npr1p bands in extracts from rapamycin treated and untreated cells (Figure (Figure2C,2C, lanes 1 and 3). This result suggested that Npr1p is dephosphorylated at steady state in the myo1Δ strain. A chs2Δ strain (a chitin synthase II null mutant defective in contractile ring function) and an fks1Δ strain (a β-1,3-glucan synthase null mutant deficient in cell wall synthesis and maintenance) were incorporated as controls representing strains under cell wall stress [55,56]. In contrast, the electrophoretic mobility of Npr1p in chs2Δ (Figure (Figure2D)2D) and fks1Δ (Figure (Figure2E)2E) control strains behaved similar to the wt (Figure (Figure2B).2B). Phosphatase treatment did not change the electrophoretic mobility of Npr1p in the myo1Δ extracts supporting our hypothesis that TORC1 is down regulated (Figure (Figure2C,2C, lanes 1 and 2). Introduction of a tor1Δ mutation into the myo1Δ strain did not produce any change in the electrophoretic migration of Npr1p compared to the myo1Δ single mutant (Figure (Figure3A).3A). This was also consistent with the notion that TORC1 is inhibited in the myo1Δ strain. These observations suggest that the myo1Δ strain is distinct from the other cell wall stress models because despite the similarity between these strains which have the PKC1 CWIP activated, the TORC1 activity was inhibited in the myo1Δ strain but not in the cell wall mutants tested (chs2Δ and fks1Δ) (see Discussion for details).
The SIT4 gene encodes a protein phosphatase that is responsible for dephosphorylation of Npr1p in vivo during nutrient starvation (Figure (Figure2A)2A) . TORC1 when active, phosphorylates Tap42p, which then binds and keeps Sit4p inactive . Thus, Sit4p activity is negatively regulated by TORC1 . In previous studies, Npr1p was shown to maintain the hyperphosphorylated state in a sit4Δ mutant treated with rapamycin indicating that its dephosphorylation was directly dependent on Sit4p activity [50,53]. To establish that dephosphorylation of Npr1p employs the same mechanism in myo1Δ, we conducted a Western blot analysis of Npr1p in a myo1Δsit4Δ strain (Figure (Figure3B).3B). Absence of Sit4p activity in the myo1Δ strain resulted in the accumulation of the slow-migrating hyperphosphorylated Npr1p. This result supports that the myo1Δ dephosphorylation of Npr1p is via Sit4p.
TOR signaling activity was previously reported to negatively regulate the PKC1 CWIP  because rapamycin treatment resulted in up regulation of PKC1 activity. Our new observations show that TORC1 was repressed in myo1Δ cells while we had previously shown that PKC1 activity was up regulated in these strains [4,6]. To determine if the repression of TORC1 activity in myo1Δ cells is responsible for up regulation of the PKC1 pathway, we treated myo1Δ cells with rapamycin and monitored Slt2p/Mpk1p hyperphosphorylation (referred to as P-Slt2p from here on) by Western blot analysis (see Methods). As previously reported, down regulation of TORC1 activity in wt cells treated with rapamycin resulted in up regulation of PKC1, reflected by an accumulation of P-Slt2p (Figure (Figure4A).4A). In untreated myo1Δ cells, there was an accumulation of P-Slt2p as described previously (Figure (Figure4A).4A). There was no significant increase in P-Slt2p levels following rapamycin treatment in these cells (Figure (Figure4B,4B, lanes 1 and 2). This observation supports the aforementioned result  that down regulation of TORC1 activity is correlated to the up regulation of PKC1.
To analyze this putative cross-talk further, we assayed the relative levels of P-Slt2p in a myo1Δtor1Δ strain. In the absence of Tor1p, a myo1Δtor1Δ strain maintained significant steady state levels of P-Slt2p at approximately 50% of the myo1Δ- levels (Figure (Figure4B,4B, lane 5), while treatment with rapamycin did not generate a significant change in these levels (Figure (Figure4B,4B, lane 6). A tor1Δ single mutant activated PKC1 at low levels (Figure (Figure4B,4B, lanes 3 and 4).
To determine if these strains presented a growth defect we tested wt, myo1Δ, tor1Δ and myo1Δtor1Δ strains for cell viability using a serial dilution assay (Figure (Figure4C).4C). The myo1Δ strain exhibited a viability range similar to the wt strain. Surprisingly, the tor1Δ strain showed increased viability through the 102 cells/ml range, consistent with previous studies that showed that Tor1p functions were not essential for cell viability. Despite having P-Slt2p present (Figure (Figure4B,4B, lane 5) the myo1Δtor1Δ strain presented a reduction in cell viability of approximately four orders of magnitude (Figure (Figure4C,4C, bottom row). Therefore, down regulation of TORC1 appears to be favorable to maintain viability in the myo1Δ strain while a complete absence of Tor1p in this strain is detrimental. These results imply that Tor1p may have a predominant role in the TORC1 functions with less activity attributed to the Tor2p in this complex. However, the residual activity in the TORC1 complex was essential for myo1Δtor1Δ strain survival because five days treatment with the IC50 of rapamycin (44nM) resulted in a 10-fold further reduction of growth (data not shown).
TORC1 is not essential in wt  or myo1Δ strains. In contrast, TORC2 carries out essential functions in yeast cells that are not shared with TORC1 [21,22,61]. Growth at the permissive (26°C) and restrictive (37°C) temperatures was assayed for strains wt, myo1Δ, chs2Δ (each bearing a genomic copy of wild type TOR2), SH121 (a control tor2Δ strain that contains a plasmid-borne copy of a temperature_sensitive tor2-21 allele, ptor2ts) and combination strains bearing also the ptor2ts plasmid. Since TORC2 is responsible for essential cellular functions, the expected outcome was that repression of the tor2-21ts allele at 37°C would be lethal in any strain. Parental strains and mutant strains bearing ptor2ts were therefore assessed for growth on agar at 26°C and 37°C. As expected, the wt, myo1Δ, and chs2Δ strains did not exhibit temperature sensitive growth (Figure (Figure5A,5A, top row). The tor2Δ ptor2ts (control), wt ptor2ts, and chs2Δ ptor2ts strains were viable at 26°C and temperature sensitive for growth at 37°C indicating dominance of the tor2-21ts mutation (Figure (Figure5A,5A, upper left and upper right respectively). In contrast, the myo1Δ ptor2ts strain presented viable growth at both temperatures (Figure (Figure5A,5A, top row). The suppression of tor2-21ts lethality by myo1Δ in this strain (YJR13) was also confirmed in the SH121 strain background (Figure (Figure5A,5A, bottom left and bottom right, respectively) to exclude genetic background effects. Growth was also assessed on Leucine dropout medium plates to ascertain that strains transformed with the ptor2ts allele were expressing the plasmid, as observed by the normal growth at 26°C [Additional file 1, left plates], and again that the tor2-21ts defect was suppressed in the absence of the MYO1 gene at 37°C independently of the strain background [Additional file 1, right plates]. The positive genetic interaction between myo1Δ and tor2-21ts mutations was reverted by complementation with a plasmid-borne copy of the wild type MYO1 gene in myo1Δtor2Δptor2tspMYO1 (SH121) (Figure (Figure5B)5B) and also myo1Δptor2tspMYO1 (YJR13) strains (data not shown).
TORC2 has been shown to have a strong regulatory effect on PKC1 activity in cell wall mutants [29,30]. To explain the observed synthetic rescue of tor2-21ts lethality by myo1Δ, we conducted a Western blot analysis of P-Slt2p levels in the wt, myo1Δ (Figure (Figure5C,5C, box1), tor2Δ ptor2ts and myo1Δ ptor2ts strains (Figure (Figure5C,5C, box 2) at the permissive temperature. Wt and tor2Δ ptor2ts strains showed similarly low P-Slt2p levels consistent with growth under non-stress conditions (Figure (Figure5C,5C, box 1 and box 2 respectively). The myo1Δ (Figure (Figure5C,5C, box1) and myo1Δ ptor2ts (Figure (Figure5C,5C, box2) strains both presented a higher level of P-Slt2p relative to wt strains at 26°C, which was also consistent with previous observations that PKC1 is activated in these strains (Figure (Figure4A,4A, lane 3).
To assess if rescue of tor2-21ts lethality at the restrictive temperature by myo1Δ was accompanied by a change in PKC1 activity levels, P-Slt2p was analyzed in whole cell extracts from cultures taken at 37°C (Figure (Figure5C5C boxes 3 and 4). The temperature shift to 37°C produced an increase in P-Slt2p levels in all four strains attributable to the heat shock effect that is known to activate the PKC1 pathway [1,62]. This suggests that there may exist two pools of Slt2p, one that is activated by the myo1Δ mutation and one that is activated (or phosphorylated) by the heat stress. Densitometric quantification and normalization of autoradiographs from duplicate experiments established that P-Slt2p levels in the myo1Δ ptor2ts strain at 37°C were 5-fold higher than in the tor2Δ ptor2ts strain yet were very similar to the myo1Δ single mutant strain at 37°C (Figure (Figure5C,5C, bottom panel). However, when we compared P-Slt2p levels between the tor2Δ ptor2ts and myo1Δtor2Δptor2ts strain, there was no significant difference between them supporting that the rescue effect was not due to Tor2p-dependent P-Slt2p up regulation (Figure (Figure5C,5C, box 5) or differences in P-Slt2p levels.
Inhibition of TOR functions activates multiple cell wall stress sensor proteins located in the plasma membrane that interact with signaling intermediates through their C-terminus in the cytoplasm . Cell wall stress sensor proteins that belong to the Wsc family and also include the Mid2 proteins, signal positively to activate the PKC1 CWIP [2,32-37]. Under cell wall stress conditions, Wsc1p, Wsc2p and Mid2p are reported to be the principal sensors responsible for activating this pathway [36,48,49]. Consistent with the notion that cell wall stress sensors may mediate the stress response in myo1Δ strains, we have presented evidence that the PKC1 CWIP is activated and essential [4-6] and that TORC1 activity is down regulated in myo1Δ strains (this study). To test if cell wall stress sensor proteins could, by a cross-talk mechanism, be involved in down regulating the TOR pathway, the myo1Δ mutation was inserted in wsc1Δ, wsc2Δ, wsc3Δ and mid2Δ mutant strains by standard genetic techniques (see Methods). The single mutants (wsc1Δ, wsc2Δ, wsc3Δ, mid2Δ) and their corresponding double mutant strains (myo1Δwsc1Δ, myo1Δwsc2Δ, myo1Δwsc3Δ, myo1Δmid2Δ) were tested for TORC1 activity and cell viability (only the results for Wsc1p are shown, Figure Figure66).
Relative TORC1 activity levels for the wt and myo1Δ strains were previously shown (Figures2B and 2C respectively), while the results for myo1Δwsc2Δ, myo1Δwsc3Δ and myo1Δmid2Δ strains were also consistent with down regulated TORC1 activity in these strains (data not shown). In contrast, the myo1Δwsc1Δ double mutant strain exhibited a result that was consistent with a fully active TORC1 (Figure (Figure6A,6A, lane 1) as judged by the relative decrease in electrophoretic mobility normally exhibited by Npr1pp, and the restored sensitivity of Npr1pp electrophoretic mobility to rapamycin treatment (Figure (Figure6A,6A, lane 2). Also, like the myo1Δ and wsc1Δ strains (Figure (Figure6A,6A, lanes 3 and 4), preliminary results show that wsc2Δ, wsc3Δ, and mid2Δ single mutant strains (data not shown) exhibited a rapamycin-insensitive Npr1p electrophoretic mobility that was consistent with down regulation of TORC1. These results indicate that absence of these cell wall stress sensors represents a cell stress and supports the idea that they also play a role during normal cell growth. We therefore conclude from these results that Wsc1p may be associated with the regulation of TORC1 in both the wt and myo1Δ strains.
Because TORC1 and PKC1 activities maintain an inverse relationship , we predicted that a re-activation of the TORC1 observed in the myo1Δwsc1Δ strain would exert an inhibitory effect on the PKC1 pathway. Consistent with this hypothesis, the myo1Δwsc1Δ strain failed to activate the PKC1 pathway as evidenced by undetectable levels of P-Slt2p relative to a myo1Δ single mutant where PKC1 was activated (Figure (Figure6B).6B). Cell viability analysis revealed a reduction in growth of approximately one order of magnitude in the myo1Δwsc1Δ double mutant strain relative to the myo1Δ strain and two orders of magnitude relative to wt and wsc1Δ strains that grew comparably well (Figure (Figure66C).
We then tested eIF2α phosphorylation levels as an additional readout of TORC1 status . Previous studies showed that TORC1 down regulation in myo1Δ was evidenced by a large (2 fold) accumulation of phosphorylated eIF2α (eIF2α-P) produced by a TOR-dependent activation of Gcn2p protein kinase (Figure (Figure7,7, top panel) . The myo1Δwsc1Δ mutant combination that restored wild type electrophoretic mobility to Npr1p (Figure (Figure6A)6A) also restored eIF2α-P to its active unphosphorylated state, eIF2α (Figure (Figure7,7, top panel, lane 4), thereby confirming that TORC1 was being reactivated in this mutant. Likewise, the accumulation of eIF2α-P in the wsc1Δ mutant confirmed that TORC1 activity was down regulated in this mutant (Figure (Figure7,7, top panel, lane 3). Control extracts from the wt and myo1Δ strains were consistent with previously reported results . Total eIF2α confirmed that the proposed changes in eIF2α-P were due to phosphorylation rather than a change in steady state levels of eIF2α (Figure (Figure7,7, middle panel).
Yeast cells must respond rapidly and effectively to alterations in the environment in order to survive stressful conditions. These processes require the involvement of signal transduction pathways such as TOR and PKC1. The PKC1 dependent CWIP is the first line of response to cell wall damage in the yeast Saccharomyces cerevisiae[13,14]. Transduction of the signal begins with the cell wall stress sensor proteins Wsc1p and Mid2p at the plasma membrane and proceeds through Rom2p and Rho1p to the PKC1 CWIP that ends with activation of the MAP kinase, Slt2p [13,48] (Figure (Figure1).1). Downstream, the transcription factor Rlm1p activates nuclear genes involved in cell wall synthesis and remodeling to produce a cell wall stress response that increases the survival potential of the yeast cell [9,10]. We have shown in prior studies that the PKC1 pathway is continuously activated in myo1Δ strains [4,6]. This response is further characterized here in the myo1Δ strains. In addition to the up regulation of the CWIP we found that TORC1 was down regulated to enhance cell survival and we provide evidence of cross-talk between the two signaling pathways.
Npr1p is a protein kinase that regulates the amino acid permease Gap1p to transport secondary nitrogen sources into the cell for the restoration of amino acid precursor levels and protein synthesis [51,64,65]. When TORC1 is down regulated by nutrient starvation or rapamycin treatment, Npr1p becomes dephosphorylated by the protein phosphatase Sit4p, thereby activating its biochemical function [50,51,53] Furthermore, inactivation of TORC1 results in downregulation of ribosome and protein synthesis . When we assayed the relative status of TORC1 activity in a myo1Δ strain, we observed that Npr1p was maintained in the dephosphorylated state and demonstrated that the Npr1p phosphorylation state was directly dependent on TORC1 and Sit4p activities. Therefore, we established that the TORC1 complex is found in a predominantly inactive state in the myo1Δ strain. The implications of such a metabolic state led us to believe that the survival of this strain is directly linked to this observation. However, the complete absence of Tor1p by genetic deletion (tor1Δ) was detrimental for survival of the myo1Δstrain, supporting that a precise level of TORC1 activity must be maintained for its survival. Furthermore, complete inhibition of TORC1 activity by rapamycin treatment of a myo1Δtor1Δ strain was lethal for growth, further supporting the idea that minimal levels of TORC1 activity are essential. Conversely, the tor1Δ single mutant was shown to acquire increased fitness, which was consistent with the proposed role of mTOR and TOR in regulating longevity and replicative life span extension respectively [24-27].
It was previously known that the TORC1 pathway plays a role in the response to cell wall stress by a negative regulation of the PKC1 CWIP under nutrient rich conditions  (Figure (Figure1).1). Under myo1Δ conditions, we observed a clear inverse biochemical correlation between TORC1 and Pkc1p activities. This raised the question, is there a common upstream regulator of the two pathways? Our results are strongly suggestive that Wsc1p acts as a common upstream regulator of TORC1 and PKC1 by exerting a positive role in the activation of the PKC1 CWIP and a negative role in the down regulation of TORC1 by an unknown mechanism. We do not propose that TORC1 and Wsc1p interact directly. Most likely, the mechanism involves an undetermined protein interactor of Wsc1p (labeled “X” in Figure Figure8).8). Recently, we have identified several novel Wsc1p interacting proteins (unpublished results) and suspect that some may function as a signaling intermediate between Wsc1p and TORC1. However, pending future studies, it is not known whether or not these Wsc1p interactors inhibit TORC1.
Our findings showed that chs2Δ and fks1Δ mutant strains strongly activated the PKC1 CWIP yet maintained normal TORC1 activity levels. We are therefore confronted with variable signaling outputs exiting from the cell wall stress sensors. In particular, disruption of cell wall integrity by these mutants leads to activation of the PKC1 CWIP; however, only the disruption of cytokinesis in a myo1Δ strain leads to both activation of the PKC1 CWIP and down regulation of TORC1 (Figure (Figure8).8). This finding is consistent with the transcriptional profiles we have determined previously, where myo1Δ only regulated half of the CWIP fingerprint genes, while fks1Δ and chs2Δ profiles were more like other cell wall damage profiles [4,5]. This reinforces the idea that myo1Δ activates a cell signaling program that is distinct from other cell wall mutants. We propose that in addition to its filament assembly function, the tail domain may serve as a scaffold (or guide) for the assembly of interacting protein complexes at the cytokinetic ring that are important for myosin function . Therefore, disruption of these putative protein assemblies by a genetic deletion of the MYO1 gene may activate the cell wall stress sensors Wsc1p and Mid2p in a different manner than in the chs2Δ and fks1Δ mutants. The non-muscle myosin heavy chain (Myo1p) of budding yeast has been shown to have independent functions associated with the head and tail domains of the protein . The tail domain contains a Minimum Localization Domain (MLD) that is sufficient to target the myosin heavy chain to the bud neck independently of the actin-binding site that is encoded within the head domain . Therefore, despite the common activation of the PKC1 CWIP among the myo1Δ, chs2Δ and fks1Δ mutant strains, we hypothesize that the inhibition of TORC1 by Wsc1p is unique to the myo1Δ mutant and may be triggered by the disruption of specific protein-protein interactions in the putative Myo1p scaffold at the cytokinetic ring.
The final question that arises from these results is, how does the myo1Δ mutant rescue tor2-21ts lethality? Strains that carry the temperature-sensitive gene of TOR2 (tor2ts) arrest growth at the restrictive temperature (37°C). This lethality is thought to be caused by the lack of TORC2 activity, decreased RHO1 activation , the lack of actin organization and cell lysis probably due to cell wall defects . The lethality has been shown to be rescued in several different ways. One way is by growth on nonfermentative carbon sources (i.e. raffinose) but not by nonfermentable carbon sources (i.e. glycerol or ethanol) . A second way in which tor2ts lethality can be circumvented is by treatment with agents that cause cell integrity stress (i.e. 0.005% SDS) . A third way in which the tor2ts lethality can be rescued is by the osmotic stabilizer, sorbitol, again suggesting that the cell wall is somehow compromised. Finally, there are several genes that have been shown to suppress the lethality of tor2ts lethality. One example is yeast PAS kinase overexpression (a gene involved in glucose partitioning in the cell) which is thought to suppress the tor2ts lethality by RHO1-dependent activation of PKC1 and actin rearrangement, activation of FKS1 and cell wall synthesis, or both . These observations have lead to the idea that cell growth and survival is a product of signals derived from cell integrity and nutrient availability . In this work we provide evidence for rescue of tor2ts lethality by the deletion of the MYO1 gene. We propose that myo1Δ rescues the tor2ts lethality by invoking both strategies described above, namely, by activating a starvation type response (TOR) and the cell wall integrity pathway (PKC1 CWIP), most likely through the reorganization of the actin cytoskeleton. However, unlike the results of Cardon et al.  where the essential Rho1p GEF was Rom2p, the roles of Rom1p and Rom2p appear to be redundant for the proposed myo1Δ rescue mechanism (data not shown).
We have shown that cross-talk between the PKC1 and TOR signaling cascades occur under the myo1Δ stress condition. TORC1 activity was found to be inversely correlated with activation of the PKC1 pathway while both Tor1p and Pkc1p act as positive regulators of viability in the myo1Δ strain. Synthetic rescue of tor2-21ts lethality by myo1Δ points to the PKC1-dependent reorganization of the actin cytoskeleton as the possible rescue mechanism. The data presented supports that in addition to its known role in signaling to the PKC1 CWIP, Wsc1p may also function as an upstream regulator of TORC1.
Saccharomyces cerevisiae strains used in this study are listed in Table Table1.1. Dr. Brian C. Rymond kindly provided the TOR1, SIT4 and WSC1 null mutation strains. The myo1Δtor1Δ, myo1Δsit4Δ, myo1Δwsc1Δ and myo1Δtor2Δptor2ts double mutants were constructed by a disruption of the MYO1 gene with a HIS5 module by homologous recombination using a PCR based method. Strains tor2ΔpTOR2 and tor2Δptor2ts were kindly provided by Dr. Michael N. Hall. The composition of complete synthetic media (CSM) for wild type cells was complemented with 2% Glucose and 1X Nitrogen base without amino acids. The composition of rich medium (YPD)/G for tor1Δ, sit4Δ, and wsc1Δ strains was complemented with 2% Glucose and 200ug/mLG418 (Geneticin). Histidine dropout media (CSM HIS-) was used for double mutant lacking the MYO1 gene; while Leucine dropout media (CSM LEU-) was used for the maintenance of plasmids, each one complemented with 2% Glucose and the appropriate nitrogen base without amino acids. Cultures were grown overnight at 26°C to mid-logarithmic phase with an optical density between 0.5–0.8AU (OD600) with continuous shaking at 225rpm. A 1mg/mL stock solution of rapamycin (SIGMA) was dissolved in the drug vehicle 100% DMSO. To inhibit TORC1, cultures were treated with rapamycin for 1h with drug vehicle alone (DMSO) 100% or with a half maximal (50%) growth inhibitory concentration of rapamycin at a final concentration of 60nM (54.85ng/mL) for wt cells and 44nM (36.56ng/mL) for mutant cells, prior to harvesting. Cultures bearing plasmid ptor2ts were grown to mid-log phase in CSM LEU- medium, and were then diluted to an OD600 of 0.5 in pre-warmed media. Cultures were shaken at 225rpm at 37°C for 1h.
Plasmid pHA-NPR1 (pEJ23) consists of YEplac181 (LEU2) expressing a functional N-terminally HA-tagged NPR1 under its own promoter , kindly provided by Dr. Estela Jacinto. Plasmid YCplac111::tor2-21ts (LEU2) containing a temperature sensitive tor2-21ts allele  was kindly provided by Dr. Michael N. Hall. Escherichia coli strain DH5α was used for the propagation and isolation of plasmids. Yeast transformations were performed by the Lithium acetate procedure. Yeast plasmid DNA was isolated by an adaptation from the QIAGEN QIAprep Spin miniprep kit.
Whole yeast cell protein extracts were prepared by harvesting and lysing cell cultures by vortexing with glass beads for 20s with 3min intervals on ice (repeated 3 times). Lysis buffer contained 50mM Tris–HCl pH 7.5, 10% Glycerol, 1% TritonX-100, 0.1%SDS, 150mN NaCl, and 5mM EDTA, supplemented with 5X Protease Inhibitor Cocktail (50X stock; Roche) and 10mM PMSF. Cell lysates were centrifuged at 13,000rpm for 10min at 4°C; the supernatant was removed and quantified using the DC Protein Assay method (Bio-Rad, Hercules, CA).
Whole protein extracts were denatured at 95°C for 5min, separated on 10% SDS-polyacrylamide gels and transferred to nitrocellulose membrane at 0.37 Amps for 1h at 4°C in a Mini Trans Blot Cell (Bio-Rad, Hercules, CA). Npr1p was consistently expressed more abundantly in wt cells than in any of the mutant cells. Therefore, the loading volumes were adjusted accordingly. The reason for these differences in Npr1p levels between strains is not known, but it has been speculated that it could be due to differences in the stability of the protein . For analysis of HA-NPR1, membranes were probed with anti-HA rat monoclonal antibody (3F10, Roche, 1:1000) in blocking solution containing 0.5% Western Blocking Reagent (Roche) diluted in 1X TBS (Tris Buffered Saline, Sigma Aldrich) at 4°C overnight and washed in 1X TBS/0.1% Tween-20 (TBS/T) (Sigma Aldrich). Membranes were counter-probed with a Horseradish Peroxidase (HRP) conjugated secondary Goat anti-rat IgG antibody (Pierce, 1:5000). For phosphorylated Slt2p (P-Slt2p), membranes were incubated with anti-phospho-p44/42 MAPK rabbit monoclonal antibody (Cell Signaling, 1:1000) in 5% BSA (Bovine Serum Albumin, Sigma Aldrich) plus TBS/T buffer at 4°C overnight. HRP-conjugated secondary antibody was Goat anti-rabbit IgG antibody (Pierce, 1:10000) diluted in blocking solution. For analysis of phosphorylated eukaryotic Initiation Factor α (eIF2α-P), the membrane was incubated with anti-phospho-eIF2α polyclonal antibody (Invitrogen, 1:1000) in blocking solution at 4°C overnight. Membranes were stripped and reprobed with a rabbit polyclonal antibody that recognizes both the phosphorylated and unphosphorylated forms of eIF2α (eIF2α)(kindly provided by Dr. Thomas E. Dever). HRP-conjugated secondary antibody was Goat anti-rabbit IgG antibody (Pierce, 1:10000) diluted in blocking solution. Membranes were also probed with a mouse monoclonal antibody against Phosphoglycerate kinase (Pgk1p) (Molecular Probes, Invitrogen, 1:500) as a loading control.
Proteins were detected using a chemiluminescent substrate (SuperSignal West Pico, Thermo Scientific), and membranes were exposed to X-ray film, which were then scanned with a Molecular Imager FX Pro Plus (Bio-Rad, Hercules, CA). Digital image intensity was quantified using Quantity One 4.5.2 software (BioRad). Protein bands were quantified according to the ratio of the intensity of the test protein relative to the intensity of its Pgk1p loading control. The obtained values were averaged from duplicate experiments. Quantitative units were expressed as CNT*mm2 or Contour Quantity. This is described as the sum of the intensities of all the pixels within the band boundary multiplied by the area of each pixel (Quantity One, Bio-Rad). Error bars represent the Standard Error of the mean (STDError mean), calculated as the standard deviation (STDEV)/Square root (SQRT) of the count.
To generate dephosphorylated proteins, 50μg of whole yeast cell protein extract were incubated with 50U (1U/μg) of Calf Intestinal Alkaline Phosphatase (CIP or PPase, New England Biolabs) in the presence of 1X CIP buffer (10X NEB 3, New England Biolabs) and 5X Protease Inhibitors cocktail, EDTA Free (50X stock, Roche) for 30min at 37°C. Samples were denatured at 95°C for 5min and subjected to SDS-PAGE and Western blot analysis.
Wt, myo1Δ, tor1Δ, myo1Δtor1Δ, wsc1Δ and myo1Δwsc1Δ strains were grown to OD600 between 0.5–0.8AU at 26°C with continuous shaking at 226rpm. 5uL of serial dilutions ranging from 1x107–1x102 cells/mL were spotted onto CSM or selection media agar plates containing 2% Glucose and 1X Nitrogen base. Plates were incubated at 26°C to observe growth after three days of incubation. Strains expressing the temperature sensitive tor2-21ts mutation were streaked on CSM or selection media agar plates, and were incubated at 26°C and 37°C for 2.5days.
The authors declare that they have no competing interests.
GP-M performed genetic knockout experiments, Western blotting, growth assays, data analysis and interpretation, and writing sections of the manuscript. ES-C participated in genetic knockout and Western blot experiments of WSC strains and growth assays. PA contributed to the data analysis and interpretation and to the writing and revision of sections of the manuscript. JRR-M as principal investigator, conceived the study, designed experiments, carried out data analysis and interpretation, wrote and revised the manuscript. All authors read and approved the final manuscript.
Pagán-Mercado, Santiago-Cartagena, Akamine, and Rodríguez-Medina. Assay for viability of yeast strains by growth at 26°C and 37°C in Leucine-deficient dropout agar medium. Strains wt (YJR24), myo1Δ, chs2Δ, wt’ (JK9-3da), tor2Δ ptor2ts, wt ptor2ts, myo1Δ ptorts, chs2Δ ptor2ts, tor2Δ pTOR2, myo1Δtor2Δptor2ts were tested for presence of the ptor2ts plasmid containing the LEU2 marker.
The authors thank Drs. Brian C. Rymond, Estela Jacinto, Thomas E. Dever, and Michael N. Hall for their kind contribution of essential reagents and yeast strains. We also thank Sahily González-Crespo and Lilliam Villanueva-Alicea for their excellent technical support. This work was supported by a SCORE Award number (5-SC1AI081658-04) from the National Institute of Allergy and Infectious Diseases (NIAID) and National Institute of General Medical Sciences (NIGMS). Partial support for this project was provided through Awards by RCMI (G12RR-03051-26) & (8G12-MD007600) and MBRS-RISE (R25GM061838).