Manganese Stress Tolerance Depends on Yap1 and Stress-Activated MAP Kinases

Understanding which intracellular signaling pathways are activated by manganese stress is crucial to decipher how metal overload compromise cellular integrity. Here, we unveil a role for oxidative and cell wall stress signaling in the response to manganese stress in yeast. We find that the oxidative stress transcription factor Yap1 protects cells against manganese toxicity. Conversely, extracellular manganese addition causes a rapid decay in Yap1 protein levels. In addition, manganese stress activates the MAPKs Hog1 and Slt2 (Mpk1) and leads to an up-regulation of the Slt2 downstream transcription factor target Rlm1. Importantly, Yap1 and Slt2 are both required to protect cells from oxidative stress in mutants impaired in manganese detoxification. Under such circumstances, Slt2 activation is enhanced upon Yap1 depletion suggesting an interplay between different stress signaling nodes to optimize cellular stress responses and manganese tolerance.


Introduction
Manganese is an essential, redox active metal of medical relevance. It serves as a metal co-factor for enzymes located in all cellular compartments and environmental exposure to manganese has been associated with the development of Parkinson-like neurological symptoms in human [1]. Budding yeast has been pivotal to identify genetic mutations affecting manganese metabolism and cellular availability. Manganese is sequestered by the vacuole or the Golgi, and Golgi manganese detoxification is mainly provided by the P-type Ca 2+ /Mn 2+ ATPase Pmr1 [2]. The Pmr1 protein can be replaced by its human homolog SPCA1 encoded by the ATP2C1 gene [3] and mutations affecting calcium and/or manganese transport activities of SPCA1 are linked to Hailey-Hailey disease (HHD, [4]). Noteworthy, yeast cells lacking Pmr1 are sensitive to peroxide (H 2 O 2 ; [5]) but at the same time, the absence of Pmr1 suppress oxygen-mediated phenotypes of superoxide dismutase 1 (Sod1) mutants [6].
To adapt to diverse extracellular stimuli and environmental changes, cells activate stress response pathways known to operate through sequential phosphorylation events that are termed protein kinase cascades. The stress-activated MAP kinases (SAPK) signaling cascades are evolutionary conserved prototypes of this kind of stress response signaling pathways. In yeast, the Hog1 SAPK plays a key role in reprogramming the gene expression pattern required for cell survival upon osmostress [7], while the Slt2 SAPK is the downstream kinase of the so-called cell wall integrity (CWI) pathway [8]. The human Hog1 paralog p38 plays a critical role in adaptive responses to environmental stress [9]. Slt2 is a functional homolog of human extracellular signal-regulated kinase 5 (ERK5), a MAPK that is activated in response to growth factors as well as physical and chemical stresses [10]. Together, Hog1 and Slt2 are needed to coordinate metabolic needs with cell cycle progression, therewith contributing to maintain genetic stability in yeast [11,12].
In an adaptive response to hydrogen peroxide (H 2 O 2 ) treatment-mediated oxidative stress, the transcription factor yeast activator (AP1-like) protein (Yap1) is activated and translocates into the nucleus to induce the expression of a protective transcriptional program [13,14]. During metal and heat-shock induced oxidative stress, Yap1 transcriptionally regulates the expression of GSH1, which encodes the first enzyme (γ-glutamylcysteine synthetase) involved in glutathione biosynthesis, leading to an increase in intracellular glutathione level [15,16].
A growing body of evidence provides a link between the loss of P-type Golgi Ca 2+ /Mn 2+ ATPases and genome instability in yeast and human. Loss of Pmr1 function leads to enhanced sensibility to a variety of DNA damaging agents, DNA damage formation and telomere shortening [5,17,18], while depletion of ATP2C1 in human cells was shown to boost the formation of reactive oxygen species (ROS) and to down-regulate the expression of DNA damage response (DDR) genes [19]. Finally, a genetic screen in Kluyveromyces lactis identified the Glutathione S-transferase θ-subunit (GTT1) as an oxidative stress suppressor in cells lacking Pmr1 [20], thereby providing a link between manganese and oxidative stress response.
These findings prompted us to further investigate stress response pathways related to oxidative damage and to determine if they contribute to manganese stress resistance in yeast. Here, we reveal that impaired manganese homeostasis leads to oxidative stress and that cellular tolerance to MnCl 2 requires the Yap1 transcription factor, although MnCl 2 addition stimulates Yap1 decay. Compromised manganese detoxification leads to a constitutive activation of the CWI effector MAPK Slt2, and Slt2 activation is further stimulated in the absence of Yap1. The activation of different stress signaling nodes pinpoints to a multifaceted impact of manganese on cellular metabolism that requires the concerted action of signaling kinases and transcription factors for manganese and oxidative stress tolerance.

Impaired Manganese Homeostasis Leads to Oxidative Stress
Several observations relate Golgi Ca 2+ /Mn 2+ transport to oxidative stress. The viability of yeast cells lacking a functional Pmr1 Golgi Ca 2+ /Mn 2+ ATPase is compromised upon treatment with H 2 O 2 [5] and oxidative stress and Notch1 activation were increased upon inactivation of the PMR1 homologue ATP2C1 in human cultured keratinocytes [19]. In yeast, several pathways participate in oxidative stress signaling, including the activation of the AP-1-like transcription factor Yap1 [21], and the SAPKs Hog1 and Slt2 [22,23] (see Figure 1A).
First, we assessed cell viability and ROS formation in pmr1∆ mutants using the fluorescent dyes propidium iodide (PI) and dihydroethidium (DHE), respectively ( Figure 1B). As compared to wild-type cells (Wt), we found that loss of Pmr1 led to a considerable increase in the fluorescence signal with either dyes indicative of loss of plasma membrane selective permeability or increased ROS formation in pmr1∆ mutants. The later finding is in concordance with previous observations showing that DHE fluorescence is increased in pmr1∆ mutants [24]. Next, we wondered if oxidative stress tolerance in cells lacking Pmr1 may depend on Yap1, a transcription factor needed to activate the transcription of antioxidant genes in response to H 2 O 2 [25,26]. In our strain background, no differences between Wt and pmr1∆ cells were observed in growth assays with H 2 O 2 supplemented media ( Figure 1C). However, mutant cells lacking Yap1 alone were very sensitive to H 2 O 2 and pmr1∆ yap1∆ double mutants became even more H 2 O 2 -sensitive than yap1∆ single mutants, supporting the previous notion that pmr1∆ mutants are prone to oxidative stress. We then analyzed whether H 2 O 2 induced Yap1 posttranslational modifications, observed as a shift in the electrophoretic mobility ( Figure 1D). The cells were treated for up to 1 h with 400 µM H 2 O 2 and protein samples were taken at various time points. In accordance with previous reports [27], Western blotting analysis revealed a rapid shift in Yap1 migration (7.5 min after treatment), suggesting an increase in Yap1 phosphorylation. While the kinetics of Yap1 protein modification was very similar in Wt and pmr1∆ cells, the initial levels of Yap1 protein appeared to be increased in cells lacking Pmr1 ( Figure 1D). In addition, we monitored intracellular Yap1-GFP protein localization to assess if the cytoplasmic-nuclear Yap1 transition in response to H 2 O 2 treatment is altered in pmr1∆ mutants ( Figure 1E). Fluorescence microscopy analysis showed that this was not the case as the Yap1-GFP protein localization was mainly cytoplasmic in untreated Wt and pmr1∆ mutant cells but became principally nuclear in both cell types shortly after H 2 O 2 treatment.

Manganese-Dependent Yap1 Decay Is Calcineurin B-Independent
Our finding that Yap1 is needed for oxidative stress resistance in the absence of Pmr1 led us to investigate whether or not Yap1 is required for growth on MnCl 2 supplemented medium. To do so, we performed a cell growth analysis of cells lacking Pmr1 and/or Yap1 in growth medium supplemented with 1, 2.5 or 5 mM MnCl 2 ( Figure 2A). As expected, the growth of pmr1∆ mutant cells was strongly inhibited by MnCl 2 addition. MnCl 2 addition also impaired the growth of yap1∆ mutants as compared to Wt control cells. We therefore wondered if MnCl 2 treatment would have an impact on Yap1 protein levels. Consequently, we performed Western blotting analyses of Wt and pmr1∆ cells expressing a Yap1-GFP protein upon treatment with MnCl 2 ( Figure 2B). MnCl 2 treatment caused a significant reduction of Yap1-GFP protein levels both in Wt and pmr1∆ cells after the addition of 10 mM MnCl 2 . MnCl 2 treatment did not alter the Yap1-GFP migration pattern suggesting that MnCl 2 is not associated with Yap1 posttranslational modifications. In order to assess if the reduction of Yap1 protein levels is associated with a change in its subcellular localization, we monitored Yap1-GFP distribution by fluorescence microscopy ( Figure 2C). However, in contrast to the nuclear Yap1-GFP accumulation observed in H 2 O 2 treated cells, Yap1-GFP did not accumulate in the nucleus upon MnCl 2 addition.
In order to distinguish whether Yap1 was down-regulated at the translational or post-translational level, we examined the effect of MnCl 2 on the stability of Yap1 in the presence of protein synthesis inhibitor cycloheximide (CHX, Figure 3A). Notably, the decrease of Yap1-GFP protein levels was very similar in the presence of MnCl 2 or CHX. No additive effect on Yap1 protein decay was observed in the presence of both MnCl 2 and CHX suggesting that MnCl 2 interferes with Yap1 protein synthesis. We next wondered if MnCl 2induced Yap1 decay is linked to the previously described CaCl 2 -dependent Yap1 decay due to activation of the calcium/calmodulin regulated serine/threonine protein phosphatase calcineurin [28]. Calcineurin is a heterodimeric enzyme comprising of a catalytic A and a regulatory B subunit, and is required for Crz1 transcription factor dependent gene expression, ion homeostasis and viability in yeast [29]. By taking advantage of calcineurin B (cnb1∆) mutant cells we confirmed that the calcineurin B subunit is required for MnCl 2 tolerance, in accordance with impaired activation of Crz1 target genes related to ion homeostasis ( Figure 3B). However, in our strain background the CHX-induced Yap1 decay was barely reduced in cnb1∆ mutant cells, and MnCl 2 -dependent Yap1 decay was similar in Wt cells as compared to cnb1∆ mutants ( Figure 3C). Further experimental settings will be needed to determine all the aspects of calcineurin-independent, MnCl 2 -driven Yap1-GFP decay. In order to distinguish whether Yap1 was down-regulated at the translational or post-translational level, we examined the effect of MnCl2 on the stability of Yap1 in the presence of protein synthesis inhibitor cycloheximide (CHX, Figure 3A). Notably, the decrease of Yap1-GFP protein levels was very similar in the presence of MnCl2 or CHX. No additive effect on Yap1 protein decay was observed in the presence of both MnCl2 and CHX suggesting that MnCl2 interferes with Yap1 protein synthesis. We next wondered if MnCl2-induced Yap1 decay is linked to the previously described CaCl2-dependent Yap1 decay due to activation of the calcium/calmodulin regulated serine/threonine protein phosphatase calcineurin [28]. Calcineurin is a heterodimeric en- (C) Fluorescence microscopy analysis of Yap1-GFP localization. The cells were analyzed before and after 15 min treatment with 1 or 10 mM MnCl 2 . Nup84-mCherry was used as a nuclear marker. Scale bar represents 5 µm. [29]. By taking advantage of calcineurin B (cnb1∆) mutant cells we confirmed th calcineurin B subunit is required for MnCl2 tolerance, in accordance with impaire vation of Crz1 target genes related to ion homeostasis ( Figure 3B). However, in our background the CHX-induced Yap1 decay was barely reduced in cnb1∆ mutant cell MnCl2-dependent Yap1 decay was similar in Wt cells as compared to cnb1∆ m ( Figure 3C). Further experimental settings will be needed to determine all the aspe calcineurin-independent, MnCl2-driven Yap1-GFP decay.

Pmr1 Depletion Leads to Constitutive Activation of the Slt2 MAP Kinase Pathway
MAPK cascades are among the major pathways by which extracellular stimu transduced into intracellular responses in eukaryotic cells. The yeast cell wall int (CWI) pathway operates through the sequential activation of protein kinases Mkk1/2 and the effector MAPK Slt2 (outlined in Figure 4A). Interestingly, CWI pa activation occurs in cells exposed to oxidative stress inducing agents, such as H2O In addition to Slt2, fungal resistance to a variety of stresses depends on the MAPK [22]. Since ROS are increased in cells lacking Pmr1, we asked whether Slt2 or Ho activated in pmr1∆ mutants. We therefore analyzed the levels of phosphorylated Sl Hog1 in Wt and pmr1∆ mutant cells ( Figure 4B). In contrast to Hog1, the levels of Sl phosphorylated Slt2 were enhanced in the absence of Pmr1 indicating a constituti tivation of the Slt2 MAPK pathway in this strain. Thus, we addressed whether ex lular MnCl2 addition would be sufficient to stimulate Slt2 activation ( Figure 4C). W pmr1∆ mutants were grown for 1 h in the presence of increasing amounts of MnCl to protein extraction. Subsequent immunoblotting revealed a MnCl2 conc tion-dependent increase in Slt2 and Hog1 phosphorylation. However, MnCl2 conc tion-dependent phosphorylation was more evident for Hog1 and we did not obs difference in Hog1 activation in pmr1∆ mutants as compared to Wt cells. To verif the constitutive Slt2 phosphorylation observed in pmr1∆ mutants depends upon i stream kinases Mkk1 and Mkk2, we compared MnCl2-and Congo red (CR)-induce phosphorylation in cells devoid of these kinases ( Figure 4D). Notably, the phosp

Pmr1 Depletion Leads to Constitutive Activation of the Slt2 MAP Kinase Pathway
MAPK cascades are among the major pathways by which extracellular stimuli are transduced into intracellular responses in eukaryotic cells. The yeast cell wall integrity (CWI) pathway operates through the sequential activation of protein kinases Bck1, Mkk1/2 and the effector MAPK Slt2 (outlined in Figure 4A). Interestingly, CWI pathway activation occurs in cells exposed to oxidative stress inducing agents, such as H 2 O 2 [23]. In addition to Slt2, fungal resistance to a variety of stresses depends on the MAPK Hog1 [22]. Since ROS are increased in cells lacking Pmr1, we asked whether Slt2 or Hog1 are activated in pmr1∆ mutants. We therefore analyzed the levels of phosphorylated Slt2 and Hog1 in Wt and pmr1∆ mutant cells ( Figure 4B). In contrast to Hog1, the levels of Slt2 and phosphorylated Slt2 were enhanced in the absence of Pmr1 indicating a constitutive activation of the Slt2 MAPK pathway in this strain. Thus, we addressed whether extracellular MnCl 2 addition would be sufficient to stimulate Slt2 activation ( Figure 4C). Wt and pmr1∆ mutants were grown for 1 h in the presence of increasing amounts of MnCl 2 prior to protein extraction. Subsequent immunoblotting revealed a MnCl 2 concentration-dependent increase in Slt2 and Hog1 phosphorylation. However, MnCl 2 concentration-dependent phosphorylation was more evident for Hog1 and we did not observe a difference in Hog1 activation in pmr1∆ mutants as compared to Wt cells. To verify that the constitutive Slt2 phosphorylation observed in pmr1∆ mutants depends upon its upstream kinases Mkk1 and Mkk2, we compared MnCl 2 -and Congo red (CR)-induced Slt2 phosphorylation in cells devoid of these kinases ( Figure 4D). Notably, the phosphorylated Slt2 signal was absent in mkk1∆ mkk2∆ (mkk1,2∆) double as well as in pmr1∆ mkk1∆ mkk2∆ triple mutants excluding the possibility that MnCl 2 could drive Slt2 phosphorylation by non-canonical kinases. Moreover, Hog1 phosphorylation was not affected by the concomitant absence of Mkk1 and Mkk2, as expected for Slt2-independent Hog1 phosphorylation in pmr1∆ mutants. These findings suggest that Slt2 phosphorylation is a hallmark of Pmr1 deficiency. lated Slt2 signal was absent in mkk1∆ mkk2∆ (mkk1,2∆) double as well as in pmr1∆ mkk1∆ mkk2∆ triple mutants excluding the possibility that MnCl2 could drive Slt2 phosphorylation by non-canonical kinases. Moreover, Hog1 phosphorylation was not affected by the concomitant absence of Mkk1 and Mkk2, as expected for Slt2-independent Hog1 phosphorylation in pmr1∆ mutants. These findings suggest that Slt2 phosphorylation is a hallmark of Pmr1 deficiency.

The Slt2 Target Rlm1 Is Up-Regulated in Cells Lacking Pmr1
SLT2 expression is known to be subjected to a feedback mechanism mediated by the CWI pathway specific transcription factor Rlm1 [30]. The observation that Slt2 protein levels are increased in pmr1∆ mutants ( Figure 4) suggests that Rlm1 activates SLT2 gene expression. To test for this possibility, we generated pmr1∆ rlm1∆ double mutants and analyzed the levels of phosphorylated and total Slt2 ( Figure 5A). The total Slt2 protein signal was reduced while phosphorylated Slt2 levels remained high in cells lacking Pmr1 and Rlm1. This finding is consistent with a model in which phosphorylated Slt2 activates Rlm1 to promote transcription of the SLT2 gene. We therefore took advantage of previously published microarray data [17], to figure out whether the expression of further Rlm1 target genes was increased in pmr1∆ mutants. This was indeed the case for a number of known Rlm1 target genes, including the SLT2 paralog KDX1 (MLP1), which was highly expressed in the absence of Pmr1 ( Figure 5B). SLT2 expression is known to be subjected to a feedback mechanism mediated by the CWI pathway specific transcription factor Rlm1 [30]. The observation that Slt2 protein levels are increased in pmr1∆ mutants ( Figure 4) suggests that Rlm1 activates SLT2 gene expression. To test for this possibility, we generated pmr1∆ rlm1∆ double mutants and analyzed the levels of phosphorylated and total Slt2 ( Figure 5A). The total Slt2 protein signal was reduced while phosphorylated Slt2 levels remained high in cells lacking Pmr1 and Rlm1. This finding is consistent with a model in which phosphorylated Slt2 activates Rlm1 to promote transcription of the SLT2 gene. We therefore took advantage of previously published microarray data [17], to figure out whether the expression of further Rlm1 target genes was increased in pmr1∆ mutants. This was indeed the case for a number of known Rlm1 target genes, including the SLT2 paralog KDX1 (MLP1), which was highly expressed in the absence of Pmr1 ( Figure 5B). We performed a Western blotting analysis and confirmed that Kdx1 protein levels are highly increased in pmr1∆ mutants ( Figure 5C). Kdx1 has been shown to interact with Rlm1 to drive the expression of stress responsive genes such as RCK1, which codes for a protein kinase involved in oxidative stress response [31]. It is noteworthy that Rck1 overexpression has been proposed to modulate Hog1 and Slt2 activation [32]. Next, we investigated if cells lacking Rlm1 and/or Kdx1 become more sensitive to H 2 O 2 or MnCl 2 ( Figure 5D). As compared to Wt and pmr1∆ mutants, colony formation efficiency was similar in the presence or absence of Rlm1 and/or Kdx1, suggesting that these factors are dispensable for manganese and oxidative stress resistance.

Slt2 Contributes to Oxidative Stress Resistance of pmr1∆ Mutants
Having excluded a role for Rlm1 and Kdx1 in the oxidative stress resistance of pmr1∆ mutants, we strove to assess the impact of Slt2 itself on oxidative stress resistance. Yeast cells have been previously shown to become highly sensitive to H 2 O 2 in the absence of Slt2 [23,33] and we therefore analyzed colony formation in the presence of H 2 O 2 ( Figure 6A). Notably, we did not observe a difference between the growth of Wt and slt2∆ mutants in H 2 O 2 supplemented medium, probably because different strain backgrounds were used in our and previous studies. Importantly, pmr1∆ slt2∆ double mutants were much more sensitive to H 2 O 2 than Wt, pmr1∆ or slt2∆ single mutant cells, providing evidence that Slt2 is required to withstand oxidative stress in the absence of Pmr1. The concomitant absence of Pmr1 and Yap1 resulted in further increased Slt2 phosphorylation compared to the pmr1∆ single mutant ( Figure 6B). These results suggest that manganese stress signalling is channeled into the CWI pathway in the absence of Yap1 and led us to assess whether oxidative stress is responsible for Slt2 phosphorylation. To this end, cells were treated with the antioxidant N-acetyl cystein [4] to see if Slt2 phosphorylation can be diminished ( Figure 6C). This was indeed the case, as N-acetyl cystein treatment strongly reduced Slt2 phosphorylation of pmr1∆ simple and pmr1∆ yap1∆ double mutants to similar levels. Since composite docking sites confer substrate recognition by both calcineurin and MAP kinases [34], we wondered if Slt2 regulates Yap1 protein levels upon MnCl 2 exposure. Therefore, cells lacking Pmr1 and/or Slt2 were grown for 1 h in MnCl 2 supplemented medium and Yap1-GFP signals analyzed by Western blotting (Supplementary Figure S1). However, MnCl 2 leads to a similar drop in Yap1 protein levels in the absence of Pmr1 and/or Slt2 and, thus, these proteins are not involved in Yap1 stability or synthesis. Taken together, our results define a relevant role for Slt2 and Yap1 in the response to manganese stress and open the possibility of an additional mechanism involved in the regulation of Yap1 transcription, translation or protein stability.

Yap1 Has a Hitherto Unknown Role in Manganese Stress Response
Manganese is a redox-active metal that can mimic superoxide dismutases by catalyzing the decomposition of O2•− to H2O2 and O2 in vitro [35]. The same effect can be achieved in vivo as impaired manganese detoxification in cells devoid of Pmr1 can bypass the lack of Sod1 [6]. If intracellular manganese stimulates H2O2 formation, intracellular manganese overload upon Pmr1 depletion would provide an explanation for why

Yap1 Has a Hitherto Unknown Role in Manganese Stress Response
Manganese is a redox-active metal that can mimic superoxide dismutases by catalyzing the decomposition of O 2 •− to H 2 O 2 and O 2 in vitro [35]. The same effect can be achieved in vivo as impaired manganese detoxification in cells devoid of Pmr1 can bypass the lack of Sod1 [6]. If intracellular manganese stimulates H 2 O 2 formation, intracellular manganese overload upon Pmr1 depletion would provide an explanation for why mutant cells suffer from increased ROS formation. Our findings corroborate the previously observed increase in ROS levels in ATP2C1 depleted human keratinocytes [19]. We find that Yap1 protein levels are increased in cells lacking Pmr1. However, there are substantial differences between oxidative stress activation of the mammalian stress transcription factor Nrf2 as compared to yeast Yap1, because a complex stress-sensing system mechanism is needed for Yap1 nuclear retention including the assistance of glutathione peroxidase 3 (Gpx3) [27,36,37]. Regarding this unique activation mechanism, the Yap1 system might stem from a different evolutionary origin than the Keap1-Nrf2 system (reviewed in [38]).
To our knowledge, the impact of manganese stress on Yap1 activation has not been thoroughly explored. Interestingly, we find that Yap1 is needed for MnCl 2 resistance, although MnCl 2 addition leads to a rapid and dramatic drop in Yap1 protein levels (see Figures 2 and 3). The nuclear accumulation of Yap1 upon oxidant challenge or due to impaired nuclear Yap1 export promotes its proteolytic degradation in a E3 Ubiquitin Ligase Not4-dependent manner [39]. MnCl 2 may stimulate cytoplasmic to nuclear Yap1 turnover, but there is no evidence of a MnCl 2 -dependent change in Yap1 localization suggesting another mechanistic cause for MnCl 2 -driven Yap1 decay. It is possible that manganese interferes with Yap1 translation as MnCl 2 was found to reduce total rRNA levels in a dose-dependent manner and to alter overall ribosome profiling [40], or one might even speculate that the 5 -UTR of YAP1 mRNA can form a riboswitch-like structure [41], whose interaction with manganese modulates YAP1 gene expression or mRNA stability. Yap1 has been shown to be regulated post-translationally in a calcineurin B-dependent manner [28], opening the possibility that MnCl 2 could alter the enzymatic activity of calcineurin to promote Yap1 decay [42]. Calcineurin B-independent Yap1 down-regulation upon H 2 O 2 [39] or MnCl 2 addition suggests that manganese could drive the modulation of other phosphatase or kinase activities that promote Yap1 decay. Along this line, we recently found that manganese stimulates the enzymatic activity of the TORC1 complex in vitro and in vivo [43]. In any case, although calcineurin does not mediate MnCl 2 -triggered Yap1 decay this phosphatase is needed to activate the stress response transcription factor Crz1. Calcineurin-driven Crz1 dephosphorylation is required for nuclear Crz1 translocation, and impaired Crz1 activation in cna1∆/cnb1∆ mutants renders these cells sensitive to manganese or arsenic stress [2,44,45]. Thus, it is likely that the balanced action of Yap1 and Crz1 activation is required for manganese stress tolerance.

SAPKs Activation Is a Read-Out for Manganese Stress
Deciphering the activation of different signaling pathways is important to define a manganese stress signature. The depletion of Pmr1 did not stimulate phosphorylation of the osmotic stress effector Hog1. In contrast, extracellular MnCl 2 did activate Hog1 phosphorylation in a dose-dependent manner. Notably, MnCl 2 addition results in a transient and similar induction of Hog1 phosphorylation in Wt and pmr1∆ mutants. Thus, Hog1 is required for an adaptive response to MnCl 2 extracellular addition.
In addition to Hog1, extracellular MnCl 2 supplementation promotes Slt2 phosphorylation in Wt and pmr1∆ mutants in a dose-dependent manner (see Figure 4). The total Slt2 levels and Slt2 phosphorylation were steadily increased in cells lacking Pmr1. The increase in Slt2 protein levels was dependent on the Slt2 downstream target Rlm1, a transcription factor that is part of a feedback loop to enhance SLT2 expression [30]. A hallmark for impaired cell wall integrity is the hypersensitivity to the cell wall damaging agents, a phenotype associated with the loss of Pmr1 function [46]. However, Slt2 activation could be a consequence of Golgi manganese depletion and consequential protein glycosylation and trafficking defects in the absence of Pmr1 [47].
Rlm1 has been shown to interact with the Slt2 paralog Kdx1 (Mlp1) to drive the expression of protein kinase Rck1 involved in oxidative stress response [31]. Indeed, we confirmed that Kdx1 protein levels were up-regulated in cells lacking Pmr1. However, genetic analysis failed to detect a growth defect if cells devoid of Pmr1 and/or Rml1/Kdx1 were grown in the presence of H 2 O 2 . Thus, it is unlikely that Rlm1/Kdx1 have a significant role in oxidative stress tolerance linked to manganese.

Activation of the CWI Stress Response in the Absence of Yap1
Corroborating previous observations that Slt2 protects cells from oxidative stress, a concomitant lack of Pmr1 and Slt2 renders cells hypersensitive to H 2 O 2 treatment. Slt2 has been shown to promote transcription activation and the elongation of stress-induced genes by catalytic and non-catalytic mechanisms [48][49][50].
It is likely that in the absence of Yap1, the oxidative stress response is channeled into the CWI pathway. The molecular bases of Slt2 activation in the absence of Yap1 remains to be explored, but Slt2 activation may result from impaired Slt2 dephosphorylation, as has occurred with genotoxic stress [51]. It will be interesting to determine if reduced phosphatase activity is linked to enhanced Slt2 phosphorylation in cells lacking Pmr1, and if Slt2 indeed drives the transcriptional activation of oxidative stress-induced genes.

Concluding Remarks and Perspectives
Each kind of stress requires an adequate response to optimize cell survival. How stress signaling networks manage to crosstalk with each other is not well understood, but mechanistic evidence have been provided on how oxidative stress inhibits pheromone signaling [14]. Here, we report the results on the stress response signature of cells supplemented with extracellular manganese and/or lacking the Mn 2+ /Ca 2+ Golgi transporter Pmr1. Thereby, we reveal the need for concomitant activation of various stress signaling pathways including Yap1 and SAPKs driven signaling as outlined in Figure 6D. A yet unanswered question is how stress signaling is channeled into the MAPK Slt2 in the absence of Yap1. In addition to the hitherto unknown role of Yap1 in manganese tolerance, we find that manganese induces a rapid reduction of Yap1 protein levels. However, the molecular bases of this manganese-driven Yap1 decay still remain to be explored in detail.

Yeast Strains, Plasmids, and Growth Conditions
Yeast strains and plasmids used in this study are listed in Tables 1 and 2, respectively. Yeast transformants were grown in liquid or solid, adenine supplemented YPD medium (YPAD). Yeast mutants were generated by according to standard protocols by direct gene knock-out or N-terminal tagging.

Drug Sensitivity Assays
Yeast cells were adjusted in concentration to an initial A 600 of 0.5, and were then serially diluted 1:10 and spotted onto plates supplemented with MnCl 2 (Sigma, St. Louis, MO, USA; 244589), or hydrogen peroxide (Alfa Aesar, Heysham, UK; L14000) at the indicated concentrations. The plates were incubated at 26 • C over the course of 2-3 days.

Yap1-GFP Localization
The cells were grown to exponential phase prior to fixation with 2.8% paraformaldehyde. Bright field images (DM-6000B, Leica, Wetzlar, Germany) were obtained at a 100× magnification, respectively. Fluorescence was detected using standard filters for mCherry (595 nm excitation/645-675 nm emission) and GFP (480 nm excitation/527 nm emission) and a digital charge-coupled device camera (DFC350, Leica) and pictures were processed with LAS AF (Leica).

Microarray Analysis
Microarray Analysis-Gene expression profiles were determined by using the "3 -expression microarray" technology by Affymetrix platform at the Genomics Unit of CABIMER (Seville, Spain) as described previously [5,17,18]. The microarray data can be derived from the GEO database using the identifier GSE29420.

ROS Detection Assay
For the detection of oxygen free radicals (ROS) and variations in the mitochondrial membrane potential or cell viability, yeast cells were grown on YPD and cultured as usual, and 2.5 µg/mL of dihydroethidium (DHE) or 5 µg/mL of propidium iodide (PI), respectively, were added to 1 mL of each sample and incubated at 24 • C for 5 and 30 min. Next, the samples were diluted 1:10 in PBS and analyzed on a FACScalibur flow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA). The data were analyzed with FlowJo software (Becton Dickinson).