Introduction

The yeast Saccharomyces cerevisiae responds to cellular stress through dramatic and global changes in gene expression, translation, and protein/enzyme activity. The environmental stress response (ESR) lies at the center of these changes and represents a common gene expression program that is triggered by numerous distinct external stresses (Gasch 2003, 2007; Gasch et al. 2000). These gene expression changes include ~ 900 genes that have similar responses to osmolytes, oxidants, thermal shock, nitrogen depletion, amino acid starvation, reducing conditions (DTT), and lead to adaption and/or progression into stationary phase (Gasch 2003, 2007; Gasch et al. 2000). In general, the ESR genes are categorized as repressed genes (rESR), such as ribosome biogenesis (RiBi) genes that are associated with proliferation, and induced genes (iESR), such as protein chaperones that provide adaptive responses to intracellular stress. Each of these stresses also generates a specific transcriptional response in addition to the ESR (Brewster and Gustin 2014; Morano et al. 2012). In response to nitrogen and carbon deprivation, cells can also enter into the developmental pathways of sporulation and pseudohyphal formation instead of progression into stationary phase (Gancedo 2001; Neiman 2011). Unexpectedly and as described below, the cellular responses to these stresses require inositol pyrophosphates.

Inositol pyrophosphates (PP-InsPs) are composed of a phosphorylated myo-inositol ring that has one or more β-phosphates (pyrophosphates; see Fig. 1). We refer readers to several excellent reviews to learn more about the enzymes that metabolize inositol polyphosphates, the cellular processes affected by them, and their multiple mechanisms of action (Eskes et al. 2017; Mutlu and Kumar 2019; Shears 2015, 2017; Thota and Bhandari 2015; Wickner et al. 2018; Wilson et al. 2013). Here, we focus on the functional roles for PP-InsPs in the cellular responses of S. cerevisiae to stress with a brief review of the key metabolic enzymes.

Fig. 1
figure 1

Schematic depicting the metabolism of inositol pyrophosphates. The soluble inositol phosphate pathway begins with cleavage of phosphatidylinositol (4,5) bisphosphate by phospholipase C (not shown) to generate inositol (1,4,5) trisphosphate (IP3). Arg82 (inositol phosphate multikinase) phosphorylates the 3 and 6 positions to generate inositol (1,3,4,5,6) pentakisphosphate (IP5). Ipk1 (inositol phosphate kinase) phosphorylates the 2 position to generate inositol (1,2,3,4,5,6) hexakisphosphate (IP6). The preferred pathway for synthesis of inositol pyrophosphates with heavy arrows (top): IP6 is pyrophosphorylated by Kcs1 (IP6K) on the 5 position to generate 5PP-InsP5 (aka, 5-IP7), and Vip1 (PP-IP5K) pyrophosphorylates 5-IP7 to generate 1,5PP-InsP4 (IP8). Siw14 specifically dephosphorylates the β-phosphate on the 5 position, and Vip1 and Ddp1 dephosphorylate the β-phosphate on the 1 position. Below is the less preferred pathway, in which Vip1 acts first on the 1 position of IP6 to generate 1PP-InsP5 (1-IP7) and Kcs1 pyrophosphorylates the 5 position on 1PP-InsP5 to generate 1,5PP-InsP4 (IP8)

Full size image

The soluble inositol polyphosphate pathway begins with the cleavage of membrane-associated phosphoinositide, phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2) by phospholipase C (PLC1) to form inositol (1,4,5) trisphosphate (abbreviated as InsP3 and further shortened to IP3). IP3 is phosphorylated by inositol multikinase (ARG82) and inositol phosphate kinase (IPK1) to produce inositol (1,2,3,4,5,6) hexakisphosphate (IP6). The kinase Kcs1 (aka, IP6K) phosphorylates the C5 α-phosphate of IP6 to generate 5PP-InsP5 (shortened to 5-IP7) (Saiardi et al. 1999), and the phosphatase Siw14 reverses this reaction by hydrolyzing the C5 β-phosphate (Steidle et al. 2016). Vip1 has dual kinase and phosphatase domains and, therefore, both adds and removes the C1 β-phosphate on 5-IP7 to generate 1,5PP-InsP4 (IP8) or it produces 1PP-InsP5 (1-IP7) from InsP6 (Lee et al. 2007; Mulugu et al. 2007). The DDP1 gene encodes a nonspecific polyphosphate phosphatase with many substrates, including 1-IP7, IP8, polyphosphate, and diadenosine polyphosphates (Kilari et al. 2013; Safrany et al. 1999; Yang et al. 1999). As such, the vip1∆ mutant cannot synthesize 1-IP7 or IP8 and has increased levels of its substrate 5-IP7 (Onnebo and Saiardi 2009; Steidle et al. 2016). The kcs1∆ mutant produces neither 5-IP7 nor IP8 (Dubois et al. 2002). The kcs1∆ vip1∆ double mutant has undetectable levels of all inositol pyrophosphates (Mulugu et al. 2007). The siw14∆ mutant has increased levels of both 5-IP7 and IP8 (Steidle et al. 2016). Many studies use the mutants described above to vary the levels of PP-InsPs to examine isoform-specific phenotypes and to characterize mechanisms of action.

Inositol polyphosphates currently have three known mechanisms of action. They serve as structural elements enabling proteins to fold into biologically active molecules (Dick et al. 2018; Macbeth et al. 2005; Millard et al. 2013; Ouyang et al. 2016; Watson et al. 2012, 2016). They act as regulatory molecules to affect signaling pathways, most commonly by competing with membrane-associated phosphatidylinositides for protein binding (Gokhale et al. 2013; Luo et al. 2003; Nagata et al. 2010; Wild et al. 2016). Finally, inositol pyrophosphates can donate the β-phosphate group to phosphoserine in proteins—without requiring a kinase—to alter protein activity (Azevedo et al. 2009; Bhandari et al. 2007; Chanduri et al. 2016; Pulloor et al. 2014; Ravi et al. 2019; Szijgyarto et al. 2011; Thota et al. 2015).

Inositol pyrophosphates are required for cellular responses to stress

Several studies have shown that inositol pyrophosphates are required for cell survival in response to stress. In one of the earliest studies, the kcs1∆ mutant was discovered to be sensitive to osmotic and cell wall (caffeine) stresses (Dubois et al. 2002). The kcs1Δ mutant is also sensitive to heat stress (Auesukaree et al. 2009; Gibney et al. 2013; Horigome et al. 2009; Jarolim et al. 2013; Ruiz-Roig et al. 2010; Saiardi et al. 2000; Sinha et al. 2008). Perhaps unexpectedly, both the kcs1∆ and vip1∆ mutants are resistant to hydrogen peroxide (Onnebo and Saiardi 2009). Although cell survival phenotypes for the kcs1∆ vip1∆ double mutant have not been reported, this strain is unable to activate the ESR in response to oxidative and osmotic stresses as measured by global gene expression assays (Worley et al. 2013).

We recently reported that the siw14Δ mutant was resistant to osmotic, heat, and oxidative stresses and nutrient deprivation (Steidle et al. 2020), which was also found during high-throughput analyses (Brown et al. 2006; Garay et al. 2014; Jarolim et al. 2013). Unlike the kcs1∆ vip1∆ double mutant, the siw14∆ mutant can generate the ESR; in fact, the siw14∆ mutant has an amplified transcriptional response to stress compared to the wild-type (Steidle et al. 2020). These results in the siw14∆, kcs1∆, and vip1∆ mutants indicate that formation of 5-IP7 (or possibly IP8) is important for the adaptive stress responses, although the different stress responses may utilize the PP-InsPs uniquely.

Cells also require inositol pyrophosphates to undergo the developmental changes of sporulation and pseudohyphal development in response to carbon and nitrogen starvation. The kcs1Δ mutant is defective for both sporulation (Enyenihi and Saunders 2003; Kloimwieder and Winston 2011) and pseudohyphal formation (Norman et al. 2018), and the vip1Δ strain is unable to form pseudohyphae (Pöhlmann et al. 2014). Increased levels of 5-IP7 due to overexpression of KCS1 or deletion of siw14Δ result in hyperfilamentation (Norman et al. 2018). Thus, the phenotypes of the siw14Δ mutant are largely the reverse of those in the kcs1Δ mutant, suggesting that synthesis of 5-IP7 is important for the adaptive stress responses as well as the developmental and morphological changes in response to nutrient deprivation.

Mechanisms to change inositol pyrophosphate levels in response to stress

Given the requirement for inositol pyrophosphates in the cellular responses to stress, the question was raised whether the levels of PP-InsPs are altered upon stress. Two studies showed opposite results when cells were challenged with hydrogen peroxide. In one, hydrogen peroxide treatment led to a decrease in IP7 levels to ~ 20% of the wild-type in vivo, and activity of recombinant Kcs1 was inhibited by oxidation in vitro, consistent with the lower levels of IP7 (Onnebo and Saiardi 2009). In the other, hydrogen peroxide treatment led to a twofold increase in IP7 levels and a fivefold increase in IP8 levels in vivo, and activity of recombinant Siw14 was reversibly inhibited by oxidation in vitro, consistent with the higher levels of IP7 (Steidle et al. 2020). The differences in the findings from the two studies have not been resolved; however, the differences could be due to strain backgrounds or experimental conditions. Both reports suggested that direct oxidation of the enzymes could occur within cells to alter inositol pyrophosphate levels.

Phosphorylation of enzymes is another possible mechanism to alter inositol pyrophosphates levels. IP7 levels increase under phosphate- and nitrogen-limiting conditions (Lee et al. 2007; Norman et al. 2018). Pseudohyphal growth depends upon several signaling pathways and their kinases, including the mitogen-activated protein kinase (MAPK) Kss1, the AMP-activated kinase Snf1, and the Ras/Protein Kinase A (PKA) pathway kinase Tpk1/2 (Celenza and Carlson 1984; Cook et al. 1996; Cullen and Sprague 2012; Erdman and Snyder 2001; Pan and Heitman 1999; Robertson and Fink 1998; Vyas et al. 2003). Both Kcs1 and Vip1 (but not Siw14) are phosphorylated under pseudohyphal-inducing conditions by the Kss1 and Snf1 kinases (Shively et al. 2015). Importantly, the ratio of 5-IP7 to 1-IP7 increased under pseudohyphal-inducing conditions (Norman et al. 2018), suggesting that phosphorylation of Kcs1 increases its catalytic activity and/or that phosphorylation of Vip1 either decreases its kinase activity or increases its phosphatase activity. These results suggest that phosphorylation could be a cellular mechanism to regulate inositol pyrophosphate levels; however, a direct demonstration of altered enzymatic activity is lacking. Further research will be necessary to determine if there are additional mechanisms to modulate inositol pyrophosphate levels.

Inositol pyrophosphates affect the ESR through both Rpd3L and Msn2/4

Two studies sought to determine the mechanism(s) by which inositol pyrophosphates influence the stress response, given the phenotypes associated with these molecules (Steidle et al. 2020; Worley et al. 2013). The general stress response in yeast depends on two distinct regulators, Msn2/4 and Rpd3L. Msn2 and Msn4 (abbreviated Msn2/4) are two partially redundant transcription factors that bind to stress response elements to promote the expression of their target genes (Görner et al. 1998; Sadeh et al. 2011). Rpd3L is a histone deacetylase (HDAC) that activates stress response genes and represses RiBi genes in response to stress (Alejandro-Osorio et al. 2009; Ruiz-Roig et al. 2010).

As noted above, one mechanism of action for inositol phosphates is structural. Mammalian Class I HDACs require IP4 to assemble into fully active histone deacetylase complexes (Millard et al. 2013; Watson et al. 2012, 2016). However, the authors note that it is possible that InsP4 regulates the enzymatic activity of the Class I HDACs because the binding pocket is accessible to soluble InsP4 and the apparent Kd for binding inositol phosphates is consistent with catalytic activity being affected by small changes in intracellular concentrations (Millard et al. 2013). The yeast Rpd3 protein has a putative inositol phosphate binding pocket based on sequence conservation with the mammalian enzymes (Worley et al. 2013). An important difference between yeast and mammalian HDACs is that full activity of the yeast Rpd3L complex requires inositol pyrophosphates (Worley et al. 2013). Substitution mutations in the residues proposed to bind inositol phosphate in Rpd3 (rpd3ibs) led to a pattern of gene expression similar to that of strains that either lack the HDAC (rpd3Δ) or are unable to synthesize inositol pyrophosphates (kcs1Δ vip1Δ) (Worley et al. 2013). In support of this, the Rpd3L target genes were not deacetylated upon osmotic stress in the kcs1∆ vip1∆ mutant. These findings demonstrated that inositol pyrophosphates are required for Rpd3L deacetylase activity (Worley et al. 2013). However, current results from the experiments in yeast cannot determine whether the PP-InsPs are exchangeable and regulate the HDAC activity or serve solely a structural role (Steidle et al. 2020; Worley et al. 2013).

Under normal growth conditions, Msn2 is phosphorylated and sequestered in the cytoplasm with low frequency shuttling into the nucleus; however, upon stress Msn2 is dephosphorylated and shuttles into the nucleus at higher frequency (Cai et al. 2008; Gonze et al. 2008; Jacquet et al. 2003; Petrenko et al. 2013). The siw14∆ mutant shows increased nuclear localization of Msn2-GFP and increased expression of Msn2 target genes under non-stress conditions (Steidle et al. 2020). These results indicated that high levels of PP-InsPs likely enhance activation of the iESR genes through Msn2/4. Further experiments are necessary to determine how inositol pyrophosphates influence Msn2/4 activity, considering that Msn2/4 are regulated by multiple pathways, including TORC1, PKA, Snf1, and Psr1/2-Whi2 (Beck and Hall 1999; Boy-Marcotte et al. 1998; Görner et al. 1998; Kaida et al. 2002; Lee et al. 2013; Smith et al. 1998).

Do inositol pyrophosphates affect signaling through TORC1?

The Target of Rapamycin (TOR) kinase is a key regulator that integrates signals of nutrient abundance and cellular stress to regulate proliferation and stress responses (reviewed in Deprez et al. 2018; Zaman et al. 2008). The TORC1 complex regulates proliferation by affecting ribosome synthesis, gene expression, and translation initiation (Fig. 2). During favorable conditions, TORC1 phosphorylates the S6 kinase Sch9 to activate it (Urban et al. 2007); in turn, Sch9 phosphorylates Dot6 and Tod6 to prevent repression of Ribi genes (Huber et al. 2011). Under stress conditions, TORC1 inhibits Sch9 activity (Urban et al. 2007), thus allowing Dot6/Tod6 to repress RiBi genes. Under favorable conditions, TORC1 also phosphorylates Tap42 to inhibit the activity of the associated Pph21/22 phosphatases (Di Como and Arndt 1996; Wang and Jiang 2003). Upon TORC1 inhibition, the Tap42 complex dissociates allowing for activation of transcription factors, such as Gln3/Gat1 and Msn2/4, and expression of the target genes (Beck and Hall 1999).

Fig. 2
figure 2

Simplified depiction of the TORC1 regulated genes. (Left) Under favorable conditions, TORC1 is active to phosphorylate Tap42. Tap42 interacts with the PP2A phosphatases to inhibit their activity, leading to lower activity of the downstream transcription factors. Active TORC1 phosphorylates the Sch9 kinase to active it; the repressors Dot6/Tod6 are inactivated by Sch9 phosphorylation. (Right) Under unfavorable conditions, TORC1 is inactive, Tap42 releases the PP2A phosphatases that dephosphorylate the downstream targets to activate the transcription factors, and inactive Sch9 does not inhibit the repressors Dot6/Tod6, leading to RiBi gene repression. High levels of inositol pyrophosphates in the siw14Δ led to increased expression of target genes in the Tap42 branch and no difference in the Sch9 branch (Steidle et al. 2020), consistent with partial activation of the Tap42 branch. Lack of inositol pyrophosphates in the kcs1Δ vip1Δ double mutant led to no change in the phosphorylation of Dot6/Tod6 (Worley et al. 2013) under favorable growth conditions

Full size image

In our analysis of microarray data obtained for an siw14Δ strain that overproduces 5-IP7 and IP8 (Steidle et al. 2020), we noted that approximately 75% of genes misregulated in the absence of stress in the siw14Δ mutant are regulated by transcription factors that are downstream of TORC1. In that report, we focused on the role of Msn2/4 because of the connection with stress-resistance phenotypes and stress-induced transcriptional changes (Steidle et al. 2020). However, target genes for the transcription factors Gln3/Gat1, Rap1, Yap1, Gcn4, Gis1, and Sfp1 also showed increased expression that was not seen in wild-type cells treated with osmotic or oxidative stresses (Fig. 3, Table 1). Increased nuclear localization of Gln3 in the siw14Δ mutant was previously reported (Hirasaki et al. 2010), and our data showing altered expression of target genes is consistent with this. Gln3 is inhibited by TORC1 (Beck and Hall 1999) through the Npr1 kinase (Tate et al. 2006), and the siw14Δ and npr1Δ mutants have opposing effects on Gln3 localization that correspond to increased or decreased activation of Gln3 target genes, respectively (Hirasaki et al. 2007). Multiple studies have found that the siw14Δ mutant is sensitive to caffeine (Alic et al. 2003; Brown et al. 2006; Care et al. 2004; Hirasaki et al. 2007; Romá-Mateo et al. 2011; Sakumoto et al. 2002). Caffeine imposes a cell wall stress through the inhibition of the cell wall integrity pathway as it intersects with TORC1 (Yan et al. 2012). Therefore, the high levels of PP-InsPs present in the siw14Δ strain could partially inhibit TORC1, leading to increased nuclear localization of downstream transcription factors (e.g., Msn2, Gln3) and increased transcription of some of their target genes.

Fig. 3
figure 3

© The authors E.A. Steidle, V.A. Morrissette, K. Fujimaki, L. Chong, A.C. Resnick, A.P. Capaldi, and R.J. Rolfes

Heat map showing differential gene expression. Heat maps comparing differentially expressed genes from DNA microarray analysis. Differentially expressed genes with lower expression (blue) or higher expression (yellow) on a Log2 scale. There are 444 genes misregulated in the siw14∆ mutant (top bar), aligned with the corresponding genes in the wild-type strain stressed with either 1 mM H2O2 (middle) or 1.3 M KCl (bottom). Genes with increased expression are regulated by Msn2/4 (left portion in the siw14Δ/WT comparison) or by multiple transcription factors—Gln3/Gat1, Rap1, Yap1, Gcn4, Gis1, and Sfp1—that are regulated by TORC1 (under the label “TFs, Table 1”). There are other genes regulated by TORC1 but the factors are not annotated. Finally, some differentially expressed genes are affected by PP-InsPs and are not regulated by TORC1. The gene expression data for this figure was originally published in the Journal of Biological Chemistry. Steidle et al. (2020).

Full size image
Table 1 Transcription factors that regulate the genes showing altered expression in siw14Δ
Full size table

Capaldi and colleagues examined the consequences of the absence of inositol pyrophosphates on the TOR pathway (Worley et al. 2013). They focused on the phosphorylation of the repressors Dot6 and Tod6, as well as expression of RiBi genes in the kcs1Δ vip1Δ double mutant (see Fig. 2). They found that phosphorylation of Dot6/Tod6 in the kcs1Δ vip1Δ double mutant was the same as in the wild-type strain (Worley et al. 2013), indicating that signaling through the Sch9 branch was unaffected by the loss of inositol pyrophosphates. This finding is not unexpected if the signal for stress is increased PP-InsPs, and this interpretation is consistent with the wild-type expression of RiBi genes in the siw14Δ mutant (Steidle et al. 2020). Together, the results from both reports (Steidle et al. 2020; Worley et al. 2013) suggest that inositol pyrophosphates do not affect the Sch9 branch, but instead may affect the Tap42 branch of TORC1 signaling (Fig. 2). The mechanism(s) by which inositol pyrophosphates affect the TOR pathway is not yet known but could be through direct binding of PP-InsPs or pyrophosphorylation. Further research is necessary to determine the mechanism(s) by which inositol pyrophosphates affect TORC1 signaling.

Beyond Saccharomyces

Inositol pyrophosphates regulate and affect cellular responses to stress in not only Saccharomyces but they also play similarly important roles in other eukaryotes. In fungi, the focus has been on developmental changes. Strains of the fungal pathogens Aspergillus nidulans and Ustilago maydis with mutations in the VIP1 ortholog (ASP1) displayed abnormal cell morphologies, defects in polarized cell growth, and defects in filamentation (Pöhlmann et al. 2014). In Schizosaccharomyces pombe, mutations in asp1 (vip1) blocked the dimorphic switch (Pöhlmann and Fleig 2010). In the pathogenic yeast Cryptococcus neoformans, kcs1Δ mutants were avirulent and displayed lower melanization, decreased laccase activity, and increased capsule formation (Lev et al. 2015; Li et al. 2016a; Liao et al. 2018). These results indicate that disruptions in the synthesis of the different types of inositol pyrophosphates has dramatic effects on morphology and virulence. Determining the mechanisms of action will lead to increased understanding of these important cellular processes and provide potentially new therapeutic targets (Kang 2018; Li et al. 2016b; Saiardi et al. 2018).

Inositol pyrophosphates also affect the stress responses in multicellular organisms. The human antiviral immune response requires the production of 1-InsP7 to activate the transcription factor IRF3 and increase expression of Type 1 interferon (Pulloor et al. 2014). Murine skeletal muscle cells express the KCS1 orthologue IP6K3 and gene expression increases in response to nutrient deprivation (fasting) (Moritoh et al. 2016). Murine IP6K3 is also implicated in controlling glucose and insulin levels (Moritoh et al. 2016). Mammalian cells respond to osmotic and thermal stresses by increasing the synthesis of InsP8 (Choi et al. 2005; Pesesse et al. 2004). In plants, inositol pyrophosphates play an important role in polarized growth (Heilmann 2016), jasmonate perception, and responses to herbivorous insects (Laha et al. 2015). Thus, the role(s) for inositol pyrophosphates in stress responses are found broadly across eukaryotes, although there is not a single common response.