The high-osmolarity glycerol (HOG) and cell wall integrity (CWI) signalling pathways interplay: a yeast dialogue between MAPK routes

Authors

  • Jose Manuel Rodríguez-Peña,

    1. Departamento de Microbiología II, Facultad de Farmacia, Universidad Complutense de Madrid (UCM), IRYCIS, 28040 Madrid, Spain
    Search for more papers by this author
  • Raúl García,

    1. Departamento de Microbiología II, Facultad de Farmacia, Universidad Complutense de Madrid (UCM), IRYCIS, 28040 Madrid, Spain
    Search for more papers by this author
  • César Nombela,

    1. Departamento de Microbiología II, Facultad de Farmacia, Universidad Complutense de Madrid (UCM), IRYCIS, 28040 Madrid, Spain
    Search for more papers by this author
  • Javier Arroyo

    Corresponding author
    1. Departamento de Microbiología II, Facultad de Farmacia, Universidad Complutense de Madrid (UCM), IRYCIS, 28040 Madrid, Spain
    • Departamento de Microbiología II, Facultad de Farmacia, Pza. Ramón y Cajal s/n, UCM, IRYCIS, 28040 Madrid, Spain.
    Search for more papers by this author

Abstract

Two mitogen-activated protein kinase (MAPK) pathways, viz. the high-osmolarity glycerol (HOG) and the cell wall integrity (CWI) pathways, regulate stress responses in the yeast Saccharomyces cerevisiae. Whereas the former is mainly involved in adaptation of yeast cells to hyperosmotic stress, the latter is activated under conditions leading to cell wall instability. Although MAPK signalling specificity can be conceived as requiring insulation of the different pathways, it is also becoming clear that the two pathways do not compete with each other but can be positively coordinated to regulate many stress responses. This review highlights our current knowledge about the collaboration between these two MAPK pathways to counteract different kinds of environmental stress. Copyright © 2010 John Wiley & Sons, Ltd.

Introduction

Saccharomyces cerevisiae lives in a constantly changing environment and therefore needs to cope with many stress situations. In response to these challenges, cells have developed mechanisms of counterbalance. The mitogen-activated protein kinase (MAPK) signalling pathways are particularly important for the adaptation of the cell to specific environments. The mechanisms mediated by these pathways are based on the detection of specific environmental signals and the development of appropriate cellular responses that allow the cells to survive in a particular niche. In the budding yeast, there are five MAPKs, Fus3, Kss1, Hog1, Slt2/Mpk1 and Smk1, which control mating, filamentation/invasion, high osmolarity, cell wall integrity and sporulation, respectively51.

Two MAPK pathways regulate the responses to stress: the HOG pathway and the CWI pathway (Figure 1). The HOG pathway is essential for orchestrating the adaptive response of yeast cells to hyperosmotic stress27, 28, in which two input branches activate the MAPK module by different mechanisms. The Sho1 branch requires Sho1 and the mucin-like proteins Hkr1 and Msb2 to sense osmostress61. Sho1-dependent signalling also requires the small G-protein Cdc42, and the PAK family members Ste20 and Cla4. The target of Ste20 is the MAPKKK Ste11, which activates the MAPKK Pbs2 under hyperosmotic conditions, resulting in Hog1 activation28. Another branch, leading to the activation of Pbs2, involves a ‘two-component’ phospho-relay signalling system with the participation of the transmembrane protein Sln1 and the response regulator proteins Ypd1 and Ssk150. In this branch, two redundant MAPKKKs (Ssk2/Ssk22) participate in the phosphorylation of Pbs2, which is responsible for the final activation of Hog1. Once Hog1 is activated, it coordinates the transcriptional program necessary for cellular adaptation to osmotic stress47, 49, 53. In addition to its main role in the regulation of hyperosmotic stress responses, the HOG pathway has also been shown to be activated in response to other stresses, including oxidative stress8, acid stress37, 45, methylglyoxal1, temperature downshift48, arsenite62, CsCl17, heat stress64 and zymolyase7.

Figure 1.

Schematic diagram of the yeast HOG and CWI MAPK pathways. Those stress conditions that activate both pathways and the known stress sensors required for this activation are indicated. The broken arrow indicates the sequential activation of HOG and CWI pathways required for zymolyase-adaptive response. See text for details.

The CWI pathway is mainly activated by cell wall stress38, 42. A pair of membrane proteins, Mid2 and Wsc1, act as the main sensors of this pathway ([34,52]; Rodicio and Heinisch, this issue). Under cell wall stress conditions, for instance caused by treatments with Congo red, Calcofluor white, heat shock or glucan synthase inhibitors (echinocandins), these sensors interact with the GEF Rom2, activating the small GTPase Rho1, which then activates Pkc138. The main role of activated Pkc1 is to trigger the specific MAPK module of this pathway. This is carried out by phosphorylation of the MAPKKK Bck1 that activates a pair of redundant MAPKKs [Mkk1 and Mkk2], which finally phosphorylate the MAPK Slt2. The phosphorylated form of this protein acts mainly on two transcription factors, the MADS-box transcription factors Rlm1 and SBF. The activation of the CWI pathway under these conditions leads to changes in the yeast transcriptional programme, mainly including those genes involved in cell wall remodelling that are necessary to protect cells from cell wall stress6. Genome-wide analyses have shown that Rlm1 plays a crucial role in the transcriptional activation of those genes induced by cell wall stress19, 20, 31. The CWI pathway is activated not only under those conditions that strictly stress the cell wall, but also under many other conditions, such as hypo-osmotic shock15, pheromone treatment12, actin depolarization35, alkaline pH58, oxidative stress4, low pH14, the unfolded-protein response [UPR]57, UV11 or treatments with caffeine, vanadate42 or rapamycin16.

MAPK pathways respond to different signals with specific physiological responses. Since they often share components, the exquisite specificity that they exert is a consequence of insulation mechanisms that avoid erroneous crosstalk. However, the evidence available indicates that a coordinated interaction/communication between these pathways is also required18. For instance, polarized growth in pseudohyphal development and mating requires, in addition to the control by the mating and filamentation signalling pathways, cell wall remodelling and, therefore, coordination between the CWI and other regulatory MAPK pathways9, 12, 66. Under these situations, cells are likely to be more sensitive to osmotic changes and have to monitor turgor pressure constantly. Since the cell wall plays a crucial role in resisting cellular turgor pressure, cell wall biogenesis and osmoregulation must be very precisely coordinated.

There is some evidence for a physiological role for the HOG pathway in cell wall biogenesis under non-stress conditions: (a) PBS2 and HOG1 have been involved in the transcriptional regulation of EXG1, encoding a yeast glucanase presumably involved in cell wall remodelling30, which affects the levels of β-1,6 glucan in the cell wall; (b) the overexpression of PBS2 also leads to significant alterations in the β-1,3 glucan network36; and (c) basal signalling through the HOG pathway in the absence of osmotic stress is required for the localization of the yeast glycosyltransferase Mnn1, a protein involved in the mannosylation of cell wall proteins at the Golgi apparatus54. A more recent detailed genome-wide transcriptional analysis of hog1 cells exposed to 0.5 M KCl47 revealed the existence of a small cluster of genes that require HOG1 for basal expression. Included in this cluster of genes are those encoding the cell wall proteins Cwp1, Sed1, Pir3 and Mlp1, all of them known to be induced by cell wall stress under the control of the CWI pathway. Apparently, the number of genes putatively regulated by this mechanism is very low. However, all these genes are included within the common signature of 18 genes that are induced in all the cell wall damage conditions studied so far6. These proteins are probably essential for cell wall biogenesis, supporting a functional role for the HOG pathway in cell wall maintenance under basal growth conditions.

HOG and CWI pathways cooperation under stress conditions

At first glance, the HOG and CWI pathways would appear to have opposite functions in the regulation of cellular stress responses: the HOG pathway is activated by hyperosmotic stress, whereas the CWI pathway is activated upon hypoosmotic shock15. However, several stressful conditions activate both HOG and CWI pathways, including heat shock, hyperosmotic stress, low pH, CsCl, oxidative stress and zymolyase-mediated cell wall damage (Figure 1), suggesting that, although pathway specificity is necessary, the HOG and CWI MAPK pathways can also be positively coordinated. Consistent with this notion, Cdc37, an Hsp90 co-chaperone, interacts in a complex with Hog1 and Slt2, controlling the functionality of the CWI and HOG cascades26. For the majority of these situations there is a lack of mechanistic insight and essentially little information about the connections between these MAPK pathways. However, there are also some examples, such as the cooperative crosstalk between HOG and CWI pathways that regulates cell wall damage caused by zymolyase in the β-1,3 glucan network, in which the elements of the two pathways and the molecular mechanisms involved has been characterized in more detail. Our aim in this review it is to present the state of the art of our current knowledge about the cooperation between HOG and CWI pathways in the regulation of yeast responses to different environmental stresses.

Heat shock activates the CWI pathway, and this activation depends on both sensors Wsc123 and Mid234. Plasma membrane stretches caused by an increase in the membrane fluidity and the accumulation of osmolytes in the cell, in particular trehalose, have been suggested as the mechanisms for CWI activation in response to heat stress32, 44. In fact, elements of this pathway are essential for cell survival under heat stress. It has also been reported that Hog1 is rapidly phosphorylated by heat stress and the Sho1 branch, but not the two-component branch of the HOG pathway, mediates this heat stress activation64. Although a hog1 mutant strain recovers more slowly from heat stress, Hog1 is not essential for survival under these conditions. What, then, is the role of Hog1 activation on adaptation to heat stress? Hog1 is not translocated to the nucleus under heat stress and, therefore, is not likely to alter gene expression. Therefore the physiological significance of Hog1 activation by heat stress remains obscure. Unfortunately, details of the possible connection between HOG and CWI pathways are lacking. It has been impossible to study whether Hog1 activation by heat stress is dependent on the CWI pathway, since mutants deleted in elements of the CWI pathway are hypersensitive to heat shock. However, the overexpression of the hyperactive allele MKK1-386 activates Hog1, but only when phosphatases Ptp2 and Ptp3 are absent, suggesting that the activation of the HOG pathway does not depend on MAPK Slt2, although the involvement of upstream elements of the CWI pathway cannot be ruled out.

A collaboration between the HOG and CWI pathways is also apparently required for the adaptation of yeast to hyperosmotic stress. This stress causes turgor pressure loss and a decrease in the volume of the cell. Under these circumstances, activation of the MAPK Hog1 leads to turgor pressure and volume recovery. However, the MAPK Slt2 is also transiently activated under these conditions through Mid222. One could speculate that the decrease in turgor pressure and the concomitant invagination of the plasma membrane is sensed by the CWI pathway. However, Slt2 activation under these conditions is dependent on the presence of Hog1 and it requires 45–60 min of stress treatment22, being delayed with respect to the rapid hyperphosphorylation of Hog1. This support the idea that activation of the MAPK Slt2 is not a consequence of the early lack of turgor pressure but a later event during the process of turgor pressure recovery. Glycerol accumulation after hyperosmotic shock depends both on an increase in glycerol-phosphate dehydrogenase (Gpd1) levels3 and on the downregulation of the glycerol channel protein (Fps1) function40. Intriguingly, the overexpression of GPD1 induces Slt2 activation even in the absence of stress65, whereas the deletion of FPS1 changes the activation of Slt2 by sorbitol from a transient to constitutive state22, suggesting that intracellular glycerol levels activate the CWI pathway. The biological significance of Slt2 activation by hyperosmotic stress could be related to the necessity of the cell to strengthen the cell wall. A transcriptional activation of SLT2 by high osmolarity (1 M sorbitol) that depends on Rlm1 and Hog1 has been reported25. The relevance of this transcriptional activation is doubtful, because genome-wide transcriptional analysis of the hyperosmotic stress response shows only a modest induction of several cell wall-related genes47, 53. In addition, yeast strains deleted in elements of the CWI pathway are not hypersensitive to hyperosmotic stress.

The response to caesium also involves both the HOG and CWI pathways. Caesium chloride induces the hyperphosphorylation of Hog1 and Slt2, with the Sln1 and Sho1 branches required for Hog1 hyperactivation13. Although a possible connection between both pathways under this stress has not been addressed, a small subcluster of genes related to cell wall biogenesis is also induced under these conditions17.

Regarding oxidative stress, it was initially described that mutants defective in elements of the Sln1 and Sho1 branches of this pathway, including hog1, displayed hydrogen peroxide and diamide hypersensitivity59. Accordingly, a slight Hog1 hyperphosphorylation has been observed, using a specific range of hydrogen peroxide concentrations, reaching a maximum at 5–7 mM24. Oxidative stress also causes the activation of the CWI pathway, with a different kinetics depending on the oxidizing agent studied. This has led to the proposal that different oxidizing agents should operate on different targets, inducing specific cellular responses that require the CWI pathway. The effect of linoleic acid hydroperoxide provided the first insight about Slt2 activation due to oxidative stress4. Additionally, studies on the effect of diamide and hydrogen peroxide showed that these two agents significantly activate Slt260, with Mtl1 as the sensor mainly responsible for signalling in case of diamide and Mid2 for hydrogen peroxide63. Slt2 participates in the global transcriptional response to oxidative stress caused by linoleic acid hydroperoxide4 and Rlm1-dependent transcriptional activation has also been observed60, 63, confirming the requirement of the CWI pathway in the response against oxidative stress.

Extracellular pH is a key environmental signal that influences yeast growth. Low pH conditions caused by inorganic acid activate the CWI pathway and consequently the transcriptional induction of genes such as SLT2 and PST1 in a Rlm1- and Mid2-dependent manner14. Although the activation of Hog1 under these conditions has not been detected, the participation of this pathway is suggested by the fact that low pH stress induces both the expression of a STRE–LacZ reporter in a Hog1-dependent manner56, and alterations in cell wall structure also dependent on Hog133. Interestingly, the transcriptional profile of yeast cells growing at low pH includes cell wall-related genes previously reported to be induced after osmotic shock33.

Low pH conditions caused by organic acids such as acetic acid induce the phosphorylation of both MAPKs Hog1 and Slt245, 46. The activation of Hog1 under these conditions depends on the Sln1 branch as reported for citric acid stress37. The phosphorylation of Slt2 in response to acetic acid stress depends moderately on the presence of Hog1 but completely on Wsc1. The possible connection between both pathways has not been studied in detail. However, the effects of a loss of Fps1 on the acetate-induced phosphorylation of these MAPKs are completely different. Although the presence of the aquaglyceroporin channel is essential for Hog1 activation, its absence leads to an increased in Slt2 phosphorylation46, suggesting that acetic acid stress may activates these MAPKs by different mechanisms.

The kinetics of activation of the two pathways under those stress conditions described above differs. In general, Hog1 shows a rapid and transient phosphorylation, whereas Slt2 activation is delayed and sustained. This suggests the participation of the CWI pathway to maintain cell wall stability under circumstances that activate the HOG pathway. Different sensors (Wsc1, Mid2 and/or Mtl1) are required for the increase in Slt2 phosphorylation (Figure 1), indicating that CWI activation is a consequence of sensing specific cell wall alterations.

Connections between HOG and CWI pathways under cell wall stress

Further support for the connections between the CWI and HOG pathways arise from recent studies on the regulation of cell wall stress responses provoked by zymolyase, an enzymatic cocktail with a predominant β-1,3 glucanase activity. The CWI pathway is essential for maintaining the integrity of the cell in a range of different conditions that interfere with the cell wall. Therefore, strains deleted in different elements of this pathway are hypersensitive to zymolyase. However, although mutants of the Sho1 branch of the HOG pathway are hypersensitive to zymolyase5, 7, HOG pathway mutants display an increased resistance to Calcofluor white21. These data suggest that, depending on the nature of cell wall insult, the HOG pathway could play different co-regulating roles within the CWI pathway. In this respect, cell wall perturbations elicited by Congo red, a compound that binds to chitin, activate a transcriptional response that almost completely depends on Slt2 but not on Hog119, 55, whereas the sequential activation of the HOG and CWI pathways is required for the cellular adaptation to zymolyase-triggered cell wall stress. Zymolyase activates both MAPKs, Hog1 and Slt2. In contrast to the other stresses described above, the molecular mechanisms involved in the connection between these pathways has been characterized in more detail. The increase in Slt2 phosphorylation under these conditions depends on the Sho1 branch of the HOG pathway and requires essential components of the CWI pathway such as Mkk1/Mkk2, Bck1 and Pkc1, except the sensors (Wsc1 and Mid2) and guanine nucleotide exchange factors of this pathway7. Elements of the Sho1 branch of the HOG pathway, including the mucin-like proteins Msb2 and/or Hkr1, recently described as putative sensors of this branch61, are also required for this signalling mechanism. The hyperactivation of Slt2, as a consequence of the previous activation of the Sho1 branch of the HOG pathway, activates Rlm1, which then regulates the bulk of the transcriptional response under these conditions20. The global transcriptional response to zymolyase also includes a group of genes that require Hog1, but not Slt2 or Rlm1, for its induction. The majority of these genes are also induced by hyperosmotic stress, suggesting that zymolyase also activates a minor osmotic-like transcriptional adaptation response that requires Msn2/420.

Additional interest regarding the interaction between the HOG and CWI MAPK signalling pathways has emerged from studies of the transcriptional responses induced by zymolyase in a slt2 mutant strain. In the absence of Slt2, the hyperosmotic-like transcriptional response is clearly increased as a consequence of Hog1 hyperactivation7, 20, suggesting that the CWI pathway has an inhibitory effect on the HOG pathway. The negative regulation of the HOG pathway by the CWI pathway is also supported by the fact that Hog1 MAPK is rapidly dephosphorylated in response to hypotonic shock in a Slt2-dependent manner15. Protein phosphatases play a crucial role as negative regulators of the MAPK pathways (for reviews, see Martin et al., 200541 and Molina et al., this issue). A mechanism of potential mutual regulation between the MAPK pathways is the activation of a phosphatase by one MAPK, which then acts on components of a second pathway. Ptp2 and Ptp3 phosphatases are known to act both on Slt2 and Hog141. Their activities prevent the lethal hyperactivation of Hog1 under heat stress29. The heat stress activation of the CWI pathway leads to the transcriptional induction of PTP2 in a Slt2-43 and Rlm1-dependent manner25. In the same way, the expression of PTP2 is also induced by zymolyase in a Slt2-dependent manner20. In the absence of Slt2, PTP2 is not induced by zymolyase, correlating with a higher degree of Hog1 phosphorylation under these circumstances and the corresponding induction of several osmotic genes7, 20. Our unpublished results indicate that Hog1 phosphorylation in a ptp2 mutant growing in the presence of zymolyase is higher than in wild-type strain and similar to that observed in cells lacking SLT2. Although direct mechanistic evidence is lacking, all these observations point out the involvement of Ptp2 in this regulatory circuit.

The Sln1-dependent response regulator Skn7 also links the CWI to the HOG pathway. The transcription factor Skn7 is a multifunctional protein that has been implicated in such diverse cellular processes as cell cycle, oxidative stress, heat shock response, cell wall metabolism or filamentous growth (see27, 38 and references therein). In response to hypoosmotic stress, the Sln1–Ypd1 osmosensing phosphorylating system is activated, leading to the inhibition of the HOG pathway. Whereas Sln1 activity, via Ypd1, inhibits Ssk1, it stimulates Skn7, resulting in the activation of at least one gene, OCH1, which encodes a mannosyltransferase involved in the maturation of cell wall glycoproteins39. Remarkably, Skn7 interacts genetically and physically not only with the HOG pathway but also with the CWI pathway2, 10. Particularly, the direct interaction between Skn7 and Rho1 associates Skn7 with the CWI pathway. However, how exactly Rho1 activates Skn7 is completely unknown. Therefore the potential link between the HOG and CWI pathways through Skn7 will require further mechanistic information.

Conclusions

As expected from a collaborative role, the HOG and CWI MAPK pathways seem to communicate with each other, at various levels. These connections could enhance the signalling capabilities of the HOG and CWI MAPK pathways to counter environmental stress. Although some of these interactions have been identified, the majority remain obscure. Therefore, there is still much work to do in mapping the complete yeast MAPK regulatory network at the molecular level.

Acknowledgements

Special thanks to members of our group for their work and discussions, especially to Clara Bermejo for her pioneering work in the study of the HOG–CWI connection. We are also indebted to Enrico Cabib, Humberto Martín and María Molina for critical reading of this manuscript. We apologize to those colleagues whose work we have been forced to omit through considerations of space. Work in our laboratory is supported by projects BIO2007-67821 from the MEC and S-SAL-0246/2006 from the Comunidad de Madrid. C.N. is head of the Merck Sharp & Dohme Special Chair in Genomics and Proteomics.

Ancillary