Genetic and Biochemical Analysis of the SLN1 Pathway in Saccharomyces cerevisiae

Abstract

The histidine kinase-based signal transduction pathway was first uncovered in bacteria and is a prominent form of regulation in prokaryotes. However, this type of signal transduction is not unique to prokaryotes; over the last decade two-component signal transduction pathways have been identified and characterized in diverse eukaryotes, from unicellular yeasts to multicellular land plants. A number of small but important differences have been noted in the architecture and function of eukaryotic pathways. Because of the powerful genetic approaches and facile molecular analysis associated with the yeast system, the SLN1 osmotic response pathway in Saccharomyces cerevisiae is particularly useful as a eukaryotic pathway model. This chapter provides an overview of genetic and biochemical methods that have been important in elucidating the stimulus-response events that underlie this pathway and in understanding the details of a eukaryotic His-Asp phosphorelay.

Introduction

The SLN1 two-component signaling pathway of Saccharomyces cerevisiae is a branched multistep phosphorelay containing a single hybrid histidine kinase, a histidine-containing phosphotransfer protein, and two downstream response regulators (Fig. 16.1A). Like the well-studied bacterial two-component pathways, the activity of the SLN1 pathway responds to environmental conditions. Phosphorelay from the Sln1 histidine kinase to the cytoplasmic Ssk1 response regulator is important in dampening the activity of the HOG1 MAP kinase pathway under normal osmotic conditions (Maeda et al., 1994, Posas & Saito, 1997, Reiser et al., 2003) while phosphorelay from the Sln1 histidine kinase to the nuclear Skn7 response regulator appears to play an important role in the cellular response to cell wall perturbations (Shankarnarayan et al., 2008).

Like most eukaryotic histidine kinases, Sln1 is a hybrid protein containing both kinase and receiver domains (Ota and Varshavsky, 1993) (Fig. 16.1B). In this configuration the first phosphotransfer step is between H576, the phoshorylatable histidine within the kinase domain of Sln1 to D1144, the phosphoaccepting aspartate within the receiver domain of Sln1. The second phosphotransfer step is between Sln1 D1144 and histidine H64 in the phosphotransfer protein, Ypd1. The final phosphotransfer steps in the pathway occur between H64 of Ypd1 and the phosphoaccepting aspartate D554 in the receiver domain of the cytoplasmic response regulator, Ssk1 and between H64 of Ypd1 and aspartate D427 in the receiver domain of the nuclear response regulator, Skn7 (Fassler et al., 1997, Li et al., 1998, Posas et al., 1996) (Fig. 16.1A).

The kinase activity of Sln1 is regulated in response to the environment. Under normal growth conditions, a modest level of kinase activity appears to be crucial for viability (Fig. 16.1A). Deletion of SLN1 or mutation in any of the phosphorylatable residues in Sln1, Ypd1, or Ssk1 leads to inviability that is suppressible by inactivating mutations in components of the HOG1 MAP kinase pathway and by overexpression of HOG1 pathway phosphatases (Maeda et al., 1994, Ota & Varshavsky, 1992, Wurgler-Murphy et al., 1997). This type of evidence led to our current understanding of the SLN1–YPD1–SSK1 pathway as important in negatively regulating the activity of the HOG1 pathway in the absence of stress.

SLN1 pathway activity is diminished in the presence of elevated osmotic conditions (Ostrander and Gorman, 1999) leading to decreased turgor (Reiser et al., 2003) (Fig. 16.1A and C). These conditions lead to the accumulation of SLN1 pathway components in the dephosphorylated form. Dephosphorylated Ssk1 interacts with and activates the kinase activity of the Ssk2 and Ssk22 MAPKKKs (Maeda et al., 1995, Posas & Saito, 1998) of the HOG1 pathway thus setting in motion the signaling cascade that ultimately results in changes in expression of osmotic response genes including those involved in biosynthesis of the compatible osmolyte, glycerol (Albertyn et al., 1994a, Albertyn et al., 1994b).

SLN1 pathway activity can also be increased (Fig. 16.1A and C). This was first shown by the isolation of mutations exhibiting elevated expression of SKN7 target genes. These have been dubbed sln1* activating mutants since they increase the activity of the Sln1 phosphorelay (Ault et al., 2002, Fassler et al., 1997, Tao et al., 2002). Both the Skn7 and the Ssk1 response regulators appear to be more highly phosphorylated in these mutants. The elevated phosphorylation of Skn7 is responsible for the increase in expression of Skn7-dependent target genes (Li et al., 1998, Li et al., 2002) and the elevated phosphorylation of Ssk1 causes defects in the cellular response to hyperosmotic conditions due to changes in the kinetics of Hog1 phosphorylation (Fassler et al., 1997). In principle, sln1 activating mutations leading to increased phosphorylation of the Ssk1 and Skn7 response regulators could be attributable to a variety of mechanisms, including, for example, increased rate of Sln1 autophosphorylation or phosphotransfer activity, or diminished Sln1-directed phosphatase activity. In practice, sln1 activating mutations map to both the Sln1 receiver domain and to the coiled-coil domain located between the second transmembrane domain and the kinase domain (Ault et al., 2002, Fassler et al., 1997, Tao et al., 2002). The sln1* P1148S activating allele is a Pro to Ser mutation of a helix capping proline that is conserved in the receiver domain of most response regulators and is equivalent to P61 of Escherichia coli CheY (Ault et al., 2002, Fassler et al., 1997). In vitro phosphotransfer analysis of the P1148S mutant revealed a shift in the phosphorelay equilibrium from Sln1 to Ypd1 but no change in the rate of hydrolysis of the aspartyl phosphate on the Sln1 receiver domain. Consistent with the increase in expression of SLN1–SKN7-dependent genes observed in the sln1* P1148S mutant, in vitro phosphorelay assays revealed a twofold increase in accumulation of phosphorylated Skn7 with the mutant Sln1* receiver domain as the phosphodonor compared to the wild-type receiver (Fig. 16.5B) (Ault et al., 2002).

The sln1* T550I activating mutation located just upstream of H576 led to the identification of a functional coiled-coil (CC) domain in the linker region of Sln1 between the extracellular domain and the kinase domain (Fig. 16.1B). This region was recognizable as a coiled-coil using the LearnCoil algorithm (Singh et al., 1998) that was developed to detect weak coiled-coils thought to be a general feature of histidine kinases occurring just upstream of the H box. The impact of the sln1 T550I mutation and of the coiled-coil domain was studied using a functional kinase domain construct consisting of aa 537–950. Derivatives of that construct that lack aa 537–570 showed no evidence of autophosphorylation at 40 min, a period of time sufficient for autophosphorylation of a coiled-coil containing construct to plateau (Tao et al., 2002). At 3 h, the mutant was phosphorylated to about 20% the extent of wild-type kinase (Ault, 2001). These observations are consistent with the inviable phenotype of the sln1ΔCC mutant (Tao et al., 2002). The activation phenotype of the sln1 T550I mutant correlates with the increased hydrophobicity afforded by the isoleucine residue at position 550. Additional mutations in the region that increased hydrophobicity at the “a” and “d” positions of the helix were likewise activating while mutations at other positions had no effect (Tao et al., 2002). The absence of an effect of CC domain activating mutations on Sln1 homodimerization led to the hypothesis that the rotational flexibility of the coiled-coil in Sln1 is important for normal Sln1 activity (Tao et al., 2002) (Fig. 16.1B). Consistent with this hypothesis, the coiled-coil domain of the strongly dimerizing mammalian C/EBP transcription factor rescued viability of the inviable sln1ΔCC mutant (Tao et al., 2002), but like sln1* mutants, this chimeric SLN1 allele caused a salt-sensitive phenotype reflecting a defect in the ability of the kinase to be downregulated in response to osmotic stress (Tao et al., 2002).

Although the isolation of activating mutations in SLN1 suggest that Sln1 kinase activity can be stimulated, it has been difficult to define the environmental conditions that trigger an increase in Sln1 kinase activity. The study of fps1 mutants led to the hypothesis that Sln1 kinase activity and/or signaling is stimulated by osmotic imbalance (Tao et al., 1999). The FPS1 gene encodes the major glycerol channel responsible for glycerol efflux (Tao et al., 1999) and fps1 mutants accumulate intracellular glycerol even in the absence of osmotic stress (Luyten et al., 1995). Interestingly, although SLN1 pathway activity is elevated in the fps1 mutant, transient osmotic imbalance caused, for example, by shifting an FPS1⁺ strain from high to low osmotic environments does not activate the SLN1 pathway (Li, 2001, Shankarnarayan, 2007). Efforts to clarify the difference in physiology between transient and prolonged hypoosmotic stress led to the identification of a cell wall protein, encoded by the CCW12 gene. Strains lacking the CCW12 gene exhibit constitutive activation of the Sln1 kinase as do strains in which the Ccw12 protein is enzymatically removed from the wall (Shankarnarayan et al., 2008). Interestingly, the fps1 mutant is deficient in wall-associated Ccw12 protein suggesting that the activity of the Sln1 kinase is regulated by aspects of the environment that trigger specific wall perturbations (Shankarnarayan et al., 2008).

Increased Sln1 kinase activity causes changes in gene expression via changes in the activity of the Skn7 response regulator. Skn7 is a transcription factor that binds to Skn7 response elements regardless of its phosphorylation state, but stimulates expression of SLN1-dependent genes only in response to phosphorylation of D427 in the Skn7 receiver domain. The nonphosphorylatable skn7D427N mutant fails to respond to elevated Sln1 signaling and the constitutive skn7D427E mutant increases expression of SLN1-dependent genes independent of SLN1 (Li et al., 1998, Li et al., 2002).

The branched architecture of the pathway (Fig. 16.1A) in which a single histidine kinase signals to the Skn7 as well as the Ssk1 response regulator was suggested by genetic analysis and confirmed by biochemical analysis (Ketela et al., 1998, Li et al., 1998). Sln1 and Ypd1-dependent phosphorylation of the Skn7 response regulator was confirmed by reconstitution of the phosphorelay pathway using purified recombinant proteins consisting of single domains of each protein (Ault et al., 2002, Li et al., 1998). The radiolabeled phosphoryl group on the Sln1 kinase domain could be distributed in turn to the Sln1 receiver domain, the Ypd1 phosphorelay protein and to the receiver domain of Skn7. Furthermore, the distribution of the phosphoryl groups to each of these domains was dependent on the presence of domains participating in earlier steps in the pathway and on the phosphoaccepting residues in each domain (Li et al., 1998). In vitro reconstitution of the Sln1 phosphorelay pathway allowed for biochemical characterization of sln1 and ypd1 mutants (Ault et al., 2002, Janiak-Spens & West, 2000, Janiak-Spens et al., 2000, Janiak-Spens et al., 2005, Tao et al., 2002) and for measurements of the phosphorylated lifetime of receiver domains (Janiak-Spens et al., 1999, Janiak-Spens et al., 2000). It also paved the way for structural analysis of pathway components (Xu & West, 1999, Xu et al., 2003, Zhao et al., 2008). In this chapter we describe genetic and biochemical methods for analysis of SLN1 pathway activity in response to relevant mutational and environmental factors.