OUP user menu

The wheat MAP kinase phosphatase 1 confers higher lithium tolerance in yeast

Ikram Zaidi, Asier González, Majdi Touzri, María C. Alvarez, José Ramos, Khaled Masmoudi, Joaquín Ariño, Moez Hanin
DOI: http://dx.doi.org/10.1111/j.1567-1364.2012.00827.x 774-784 First published online: 1 November 2012


The durum wheat TMKP1 gene encodes a MAP kinase phosphatase. When overexpressed in Saccharomyces cerevisiae, TMKP1 leads to salt stress tolerance (especially LiCl), which is dependent on the phosphatase activity of the protein. The TMKP1-associated Li+ resistance is restricted to a galactose-containing medium. Interestingly, this salt tolerance is abolished in the absence of one member of the yeast type 2C Ser/Thr protein phosphatase family (Ptc1) but not when other members such as Ptc2 or Ptc3 are lacking. Increased Li+ tolerance is not mediated by regulation of the P-type ATPase Ena1, a major determinant for salt tolerance. In contrast, the effect of TMKP1 depends on Hal3 (a negative regulator of Ppz phosphatases) and on the presence of the high-affinity potassium transporters Trk1/Trk2. Tolerance to Li+ is also abolished in cells lacking the aldose reductase Gre3, previously shown to be involved in the resistance to this cation. This study provides evidence that the wheat TMKP1 phosphatase is contributing to reduce the exacerbated lithium toxicity in galactose-grown cells, in a way that depends on the presence of the potassium Trk transporters.

  • MAP kinase phosphatase
  • salt stress tolerance
  • wheat
  • yeast
  • MAPK pathways


Reversible protein phosphorylation is the most common mechanism for regulation of a myriad of cellular processes that enables eukaryotic organisms to grow and to reproduce successfully. Mitogen-activated protein kinases (MAPKs) constitute a conserved family of proteins which play a central role in such complex signal transduction networks and allow eukaryotic cells to respond properly to a broad set of environmental signals. MAPKs are activated by dual phosphorylation on serine/threonine and tyrosine residues. This dual phosphorylation needs to be tightly regulated (Marshall, 1995; Camps et al., 2000). Phosphatases are important negative regulators of MAPK signaling (Keyse, 2008) as they control the magnitude and duration of MAPK activities. They can be grouped into three major groups: tyrosine phosphatases (PTPS), serine–threonine phosphatases (PSTPs), and dual-specificity (Ser/Thr and Tyr) phosphatases (DSPs) or MAPK phosphatases (MKPs). Given that full inactivation of MAPKs requires dual dephosphorylation, MKPs are an important group of phosphatases dedicated to the regulation of MAPK signaling (Camps et al., 2000; Keyse, 2008).

Whereas relevant progress has been made toward understanding the regulatory function of several MKPs in mammals (Boutros et al., 2008), in plants our knowledge is largely restricted to a small subset of MKPs, and only a few reports have been published for a decade (for review, see Bartels et al., 2009). The Arabidopsis MKP1 (AtMKP1) with its rice and tobacco counterparts are the most characterized MKP in plants. It has been shown that AtMKP1 is required for the control of various stress responses including salinity, UV, and pathogen response (Ulm et al., 2001, 2002; Bartels et al., 2009). The rice osMKP1 is implicated in the wound response and the regulation of overall plant growth and development (Katou et al., 2007). In the case of tobacco, NtMKP1 controls wound-induced activation of SIPK and WIPK (Yamakawa et al., 2004).

We have recently identified in durum wheat the first MKP (TMKP1) highly similar to OsMKP1 (Zaidi et al., 2010). The TMKP1 protein (752 aa) has four characteristic domains conserved among plant MKPs. A catalytic domain (residues 172–287) including the signature motif for DSPs (VHCcqGvsRSTSLVIAYLM), with the conserved cysteine (Cys 214), occupies the N-terminal region. The catalytic domain is followed by another one similar to the actin-binding protein gelsolin (aa 291–379). A central region (aa 398–449) shows high homology with the previously characterized calmodulin-binding domains (CaMBD) of NtMKP1 and OsMKP1 (Yamakawa et al., 2004; Katou et al., 2005). Finally, a serine tract region (especially between residues 489–550) is present at the C-terminal part of TMKP1 (Fig. 1a). TMKP1 was shown to be differentially regulated at the transcriptional and translational levels in two durum varieties with different tolerance to salt and drought stress. We have in addition observed that the protein accumulates in nuclear/nucleolar compartments and controls the subcellular localization of its MAPK partners TMPK3 and TMPK6. These features make TMKP1 particularly attractive and may help to gain insight into the modulation of stress responses in wheat (Zaidi et al., 2010).

Figure 1

Overexpression of TMKP1 confers tolerance to LiCl. (a) Schematic depiction of the structure of TMKP1 and the N-terminal and C-terminal extension lacking versions showing the positions of the conserved domains (DSP, catalytic dual specificity domain; Gel, gelsolin domain; CaM, calmodulin-binding domain; Ser, Ser-rich region). (b) Effect of TMKP1 overexpression in wild-type cells (BY4741). The strain was transformed with the indicated pYES2-derived plasmids. Cells were grown overnight in synthetic medium lacking uracil and containing 2% galactose and then were spotted on YPGal plates with different concentrations of LiCl as indicated. Growth was monitored after 4 days.

Heterologous expression of plant genes in the well-defined model organism Saccharomyces cerevisiae has been proven to be an efficient approach to initiate functional analyses of the explored gene/protein (Trueman, 1995; Byrne et al., 2005). To investigate the function of TMKP1 in salt stress response, we decided in a first step to explore the effect of its overexpression in S. cerevisiae.

In S. cerevisiae, tyrosine, serine/threonine, and DSP phosphatases co-ordinately dephosphorylate and thereby inactivate MAPKs. At present, these phosphatases have been demonstrated to act on Fus3, Hog1, and Slt2 kinases controlling pheromones, high osmolarity, and cell wall integrity signaling pathways, respectively. The complexity of MAPK-mediated signaling pathways is illustrated by the fact that each MAPK is controlled by several phosphatases which in turn can inactivate more than one MAPK. Only two DSPs, Msg5, and Sdp1 have been identified in yeast. Msg5, which promotes adaptation to pheromone response by dephosphorylating Fus3 (Doi et al., 1994; Zhan et al., 1997), can in addition regulate the cell wall integrity pathway by controlling the phosphorylation status of Slt2 (Martin et al., 2000). The second dual phosphatase Sdp1 has been also shown to dephosphorylate Slt2, upon a heat shock (Collister et al., 2002; Hahn & Thiele, 2002).

Among the four basic families of serine/threonine phosphatases [type 1 (PP1), type 2A (PP2A), type 2B (PP2B), and type 2C (PP2C)], only three members of the last family (Ptc1, Ptc2, and Ptc3) have been identified as direct MAPK-negative regulators linked to the HOG pathway (Martin et al., 2005; Arino et al., 2011). Interestingly, Ptc3 and Ptc1 affect salt tolerance in yeast, although through different mechanisms: overexpression of PTC3 confers lithium tolerance in wild-type cells in a Hog1-dependent manner, whereas cells lacking PTC1 display a pronounced sensitivity to LiCl but not NaCl (Ruiz et al., 2006). This lithium sensitivity is comparable to that produced by lack of well-known regulators of saline tolerance, such as the phosphatase calcineurin or Hal3, an inhibitory subunit of Ppz1/2 phosphatases (Arino, 2002), although this phenotype is not attributable to the role of Ptc1 in Hog1 regulation. It has been proposed that Ptc1 may control the function of Ena1 Na+-ATPase, which pumps out of the cell toxic sodium and lithium cations (Haro et al., 1991), probably through the regulation of the Hal3/Ppz system (Ruiz et al., 2006).

Lithium cations are exceedingly toxic when cells are grown in the presence of galactose as carbon source (Masuda et al., 2000, 2001). In this work, we report that TMKP1 overexpression in yeast confers enhanced tolerance to LiCl but exclusively in galactose-containing growth media. The observed Li+ resistance is linked to Ptc1 (but neither to Ptc2 nor Ptc3), Hal3, and Trk1/2, the two main transporters involved in potassium uptake in S. cerevisiae. The observation that increased tolerance also depends on the presence of GRE3 suggests a link between this aldose reductase, involved in galactose metabolism and methylglyoxal detoxification, and TMKP1 function.

Materials and methods

Strains and growth conditions

Escherichia coli TOP 10 cells were used as plasmid DNA host and were grown at 37 °C in Luria–Bertani supplemented with 100 μg mL−1 of appropriate antibiotic for plasmid selection. Saccharomyces cerevisiae cells were grown at 28 °C in YP (10 g L−1 yeast extract, 20 g L−1 peptone) plus 20 g L−1 of the indicated carbon source or, when required, in synthetic medium containing 2% glucose or galactose and lacking the appropriate selection requirements. The relevant genotypes of the strains used in this work are listed in Table 1. Single kanMX mutants in the BY4741 background were generated in the context of the Saccharomyces deletion project (Winzeler et al., 1999). The BY4741 derivative strain BYT5 is a kind gift from Hana Sychrová (Institute of Physiology Academy of Sciences of the Czech Republic) and it was constructed by homologous recombination using the Cre–loxP system (Guldener et al., 1996). Bacterial and yeast cells were transformed using standard techniques. Restriction reactions, DNA ligations, and other standard recombinant DNA techniques were carried out as described elsewhere (Sambrook & Russell, 2001).

View this table:
Table 1

Saccharomyces cerevisiae strains used in this work

NameRelevant genotypeSource or reference
BY4741Mat a his3∆1 leu2∆ met15∆ ura3∆ Winzeler et al. (1999)
MAR143BY4741 ptc1::nat1 Ruiz et al. (2006)
BYT12BY4741 trk1::loxP trk2::loxP Petrezselyova et al. (2010)
BYT5BY4741 ena1-5::loxP H. Sychrová
JA100MATa ura3-52 leu2-3,112 his4 trp1-1 can-1 de Nadal et al. (1998)
  • Single kanMX mutants are not listed here.

Plasmid construction and site-directed mutagenesis

The construction of plasmids pY-TMKP1, pY-TMKP1_∆ 5′, and pY-TMKP1_∆ 3′ was performed as follows. One PCR product of 2.2 kb and two others of approximately 1.8 kb corresponding to full length, 5′ deleted (405 bp) and 3′ deleted (393 bp) TMKP1 ORF respectively, were amplified with additional EcoRI and XbaI restriction sites at their ends. These products were then digested with EcoRI and XbaI and subcloned into the same sites of pYES2 expression vector (URA3 marker). To obtain pWS–TMKP1 and pSK–TMKP1 constructs, an EcoRI/SalI 2.2-kb PCR fragment derived from pY-TMKP1 was inserted into the same restriction sites of pWS93 plasmid (URA3 marker, Song & Carlson, 1998) or pSK93 (the same as pWS93 but URA3 is replaced by TRP1 marker). Two mutated forms of TMKP1, TMKP1P303T (proline replaced by threonine), and TMKP1C214G (cysteine 214 replaced by glycine) were generated by PCR as described previously (Atanassov et al., 2009). Both mutated versions were cloned as PCR products in pYES2 (in EcoRI site), resulting in pY-TMKP1P303T and pY-TMKP1C214G constructs.

β-galactosidase assays

Analysis of the ENA1 promoter activity was carried out by co-transforming the appropriate yeast strain (JA100) with plasmid pKC201, which carries the ENA1 promoter fused to the LacZ reporter gene (Alepuz et al., 1997), and with pSK-TMKP1 or pSK93 constructs. Positive clones were selected in medium lacking uracil and tryptophan. The resulting cultures were then diluted in YPD and YPGAL medium (5 mL) and grown for 4 h to an A 660 0.8–1. Next, aliquots of 1 mL cultures were centrifuged and resuspended in the same volume of the appropriate YPD/YPGAL containing or not containing 200 mM LiCl. Growth was resumed for 60 min, and then cells were collected and processed for β-galactosidase assay as described previously (Reynolds et al., 1997).

Growth tests

Sensitivity of yeast cells to lithium or sodium chloride was evaluated by growth on plates (drop test) as previously described (Posas et al., 1995). For growth tests in liquid medium under limiting concentrations of KCl, cells were grown overnight until saturation, and then collected and diluted to an A 660 of 0.005 in Translucent K-free medium (which carries a negligible concentration of potassium, approximately 15 μM, and 68 mM ammonium) containing 2% glucose or 2% galactose and various concentrations of KCl (Navarrete et al., 2010). Cells were finally incubated for further growth during 17 h. Growth was analyzed by measuring optical density at 660 nm.

Protein extraction and immunoblot analysis

The indicated yeast strains were grown to an A 660 of 0.4 in YPD or YPGAL media. Cells were then incubated in YPD containing 200 mM LiCl or YPGAL containing 20 mM LiCl for 1 h, harvested by centrifugation, and frozen at −80 °C. For protein extraction, cells were resuspended in homogenization buffer [50 mM Tris–HCl, pH 7.5, 150 mM NaCl, 1 mM DTT, 10% glycerol, 0.5 mM PMSF, 0.1% (v/v) Triton X-100 and complete inhibitor mixture (Roche Applied Science)], in the presence of acid-washed glass beads (Sigma). Cells were then broken at 4 °C by vigorous shaking in a Fast Prep cell breaker (Bio 101 Inc.) at an intensity of 5.5 for 30 s (four repeats spaced each time by 1 min to avoid overheating). The lysate was collected after centrifugation for 30 s at 500 g at 4 °C and subjected to further centrifugation (10 min at 500 g at 4 °C). The supernatant was mixed with Laemmli sample buffer (4×) and electrophoresed, and immunoblots were performed using anti-HA antibody (Roche). The horseradish peroxidase–conjugated anti-mouse IgG (GE, 1/20.000 dilution) was added as a secondary antibody. Signal detection was performed as described in the ECL western detection kit (Amersham Biosciences).

Phosphatase activity

The phosphatase assay was performed on purified GST::TMKP1, GST::TMKP1C214G using the 3-O-methylfluorescein phosphate (OMFP; Sigma) as a substrate. Phosphatase activity was measured as described previously (Zaidi et al., 2010).

Cation content measurements

Intracellular Li+ and K+ contents were measured as described in Casado et al. (2010) and Marquina et al. 2012), respectively.


TMKP1 confers increased tolerance to LiCl in yeast cells

The TMKP1 cDNA was cloned in the multicopy expression vector pYES2 (pY-TMKP1) under the galactose-inducible promoter and transferred to the wild-type S. cerevisiae strain BY4741. As can be observed (Fig. 1b), the overexpression of TMKP1 in cells growing on a galactose-containing medium results in an enhanced tolerance to salt stress, particularly to LiCl (note that the concentration of LiCl in our assays does not exceed 40 mM because, as it has been previously described (Masuda et al., 2001), lithium is highly toxic to yeast when grown in galactose medium). A very slight increased tolerance to NaCl upon TMKP1 overexpression is also detected (data not shown). Additionally, versions of TMKP1 lacking N-terminal or C-terminal extensions (removing the first 135 a.a or the last 131 a.a, respectively), which contain the four conserved domains of TMKP1, are also able to increase the LiCl tolerance of the BY4741 strain (Fig. 1b).

The phosphatase activity of TMKP1 is essential to improve lithium tolerance

To investigate whether the tolerance to lithium chloride requires the phosphatase activity of TMKP1, we engineered a mutant where the conserved Cystein 214 in the catalytic domain is replaced by a Glycine (TMKP1C214G). First, a phosphatase assay was performed in vitro using both wild-type (GST::TMKP1) and mutated (GST::TMKP1C214G) recombinant forms produced in E. coli. In this assay, the phosphatase activity is determined by the hydrolysis of the OMFP releasing the 3-O-methylfluorescein adducts measurable by its absorbance at 477 nm. As observed in Fig. 2a, in the presence of GST::TMKP1, a time-dependent increase in the reaction product was detected, reflecting a constant phosphatase activity of TMKP1. However, when GST::TMKP1C214G was assayed, no phosphatase activity was detected, similar to what is observed for GST alone, suggesting that the C214G substitution inactivates the enzyme. Furthermore, the ability of the overexpressed TMKP1 to confer lithium tolerance is eliminated when the protein contains a C214G mutation (Fig. 2b), whereas the phenotype of the yeast cells transformed with a version carrying a mutation outside the catalytic domain of TMKP1 (TMKP1P303T, used here as an additional control) is indistinguishable from the strain overexpressing TMKP1. All these results serve as a proof of concept and confirm that TMKP1 confers the lithium tolerance phenotype via its phosphatase activity.

Figure 2

The TMKP1 dead phosphatase mutant is unable to enhance lithium tolerance. (a) In vitro phosphatase activity of the recombinant TMKP1 proteins produced in E. coli. Time courses of TMKP1 phosphatase activity using OMFP as a substrate. Activities were registered on wild-type GST::TMKP1, mutated GST::TMKP1C214G, and GST proteins. Experiments were repeated three times, and the data from one representative assay are shown here. (b) The wild-type BY4741 strain was transformed with the pYES2 vector (pø), pY-TMKP1, pY-TMKP1C214G (containing the dead phosphatase mutant), and pY-TMKP1P303T (containing a mutation outside the DSP catalytic domain). Cells were grown as in Fig. 1b and spotted on YPGal plates containing the indicated amounts of LiCl. Growth was monitored after 4 days.

The TMKP1-mediated lithium tolerance depends on galactose as the sole carbon source

The use of a pYES2 vector, which is based on the GAL1 regulatable promoter, for the expression of TMKP1 forced us to use YPGAL medium in our growth assays in the presence of the toxic Li+ cations. To overcome this constraint, we cloned the TMKP1 cDNA in the pWS93 plasmid under the control of the strong constitutive ADH1 promoter and in frame with three HA epitopes. The growth of cells harboring pWS-TMKP1 was then tested on YP media containing different carbon sources (galactose, glucose, fructose, raffinose, and sucrose) in the presence of LiCl. As it can observed in Fig. 3a, the enhanced lithium tolerance in comparison with the strain transformed with the empty plasmid is only observed in the presence of galactose as carbon source. It must be noted that Li+ is highly toxic when galactose is used as the only carbon source. Therefore, our results suggest that the beneficial effect of TMKP1 expression could be related to the mechanisms leading to abnormal toxicity of Li+ in galactose-grown cells. The failure of cells to tolerate LiCl when growing on carbon sources other than galactose cannot be attributed to deficient protein expression. As shown in Fig. 3b, in cells grown on glucose, TMKP1 accumulates even at higher amounts than when grown on galactose. This result confirms again that TMKP1-mediated lithium tolerance is related to galactose.

Figure 3

The TMKP1-mediated lithium tolerance occurs in medium containing galactose but not other carbon sources. (a) The wild-type strain BY4741 was transformed with the plasmid pWS93 (pWSØ) or pWS93-TMKP1 (pWS-TMKP1). Cells were grown overnight in synthetic medium lacking uracil and with 2% glucose as carbon source, and then cultures were deposited on YP plates containing galactose, glucose, fructose, sucrose, or raffinose as a carbon source, and the indicated concentrations of LiCl. Growth was monitored after 4 days. (b) Immunoblot analysis of TMKP1 levels. Cultures of BY4741 strains harboring pWS93 or pWS-TMKP1 constructs were grown to an early exponential phase in YPD or in YPGal media and treated with LiCl (200 mM for cells growing in YPD or 20 mM for cells cultivated in YPGal) for 1 h. Thirty micrograms of protein were electrophoresed, transferred, and incubated with anti-HA monoclonal antibodies.

Deletion of PTC1 blocks the effects of overexpression of TMKP1

As mentioned in the Introduction, the type 2C protein phosphatases Ptc1 and Ptc3 affect lithium tolerance in S. cerevisiae. To test whether the LiCl hypertolerance conferred by overexpression of TMKP1 could be related to these phosphatases, we expressed TMKP1 in cells lacking Ptc1, Ptc2, or Ptc3. As observed in Fig. 4, the overexpression of TMKP1 was completely unable to increase tolerance to Li+ in the absence of Ptc1, whereas in strains lacking PTC2 or PTC3 we could detect an improvement in tolerance similar to that of wild-type strain. These results suggest that the effect of TMKP1 depends on Ptc1 function.

Figure 4

Effect of TMKP1 overexpression on strains lacking Ptc1–3. (a) The BY4741 strain and its ptc1, ptc2, and ptc3 derivatives were transformed with the indicated plasmids. Cells were grown until saturation in synthetic medium lacking uracil and with 2% galactose as carbon source. Then, cultures were spotted on the indicated plates, and growth was monitored after 4 days.

Increased lithium tolerance owing to overexpression of TMKP1 is not caused by deregulation of the ENA1 gene

To gain insight into the possible mechanism through which TMKP1 enhances lithium tolerance in yeast and taking into account that Ptc1 affects lithium tolerance modulating the Ena1 Na+-ATPase function (which pumps out of the cell the toxic Li+ and Na+ cations), we addressed the question whether Ena1 is involved in the TMKP1-related phenotype. For that purpose, we overexpressed TMKP1 in cells lacking the ENA1-5 cluster. As shown in Fig. 5a, we could detect an improvement in lithium tolerance in this hypersensitive strain. This finding suggests that TMKP1 increases lithium tolerance regardless the function of Ena1. Because ENA1 is essentially regulated at the transcriptional level (for a review, see Ruiz & Arino, 2007), we examined the effect of TMKP1 overexpression on the expression levels of ENA1. To this end, we monitored the ENA1 promoter activity through measuring the β-galactosidase activity in yeast cells harboring an ENA1LacZ fusion reporter. The basal β-galactosidase activities registered in the absence or in the presence of TMKP1 were similar in cells grown in YPD or YPGAL media (Fig. 5b). As the expression of ENA1 is induced by LiCl, we also investigated whether TMKP1 affects this induction. After an exposure of 1 h to 200 mM LiCl, the ENA1 promoter was activated in the presence of TMKP1 almost to the same extent as the strain transformed with the empty plasmid. Altogether, these results indicate that the observed phenotype could not be attributed to increased ENA1 expression, suggesting that TMKP1 confers lithium tolerance through an Ena1-independent mechanism. Consistent with this hypothesis is that intracellular Li+ contents are not altered in yeast overexpressing TMKP1 when compared to wild-type cells (Supporting Information, Fig. S1).

Figure 5

The effect of overexpression of TMKP1 is not mediated by regulation of the ENA1 gene. (a) BY4741 strain and its ena1–5 derivative were transformed with pY-TMKP1 or with pø (empty pYES2 vector) and grown overnight in synthetic medium lacking uracil and with 2% galactose as carbon source. Then, cultures were spotted on YPGal plates containing increasing concentrations of LiCl and incubated for 4 days. (b) Wild-type strain JA100 was co-transformed with plasmid pKC201 (containing the ENA1 promoter fused to LacZ) (Alepuz et al., 1997) and with pSK-TMKP1 or pSK93 constructs. Cells were selected and grown as indicated in . Then, cultures were treated or not with 0.2 M LiCl for 1 h, after which they were processed for β-galactosidase measurements as described in the section. Results are mean from three independent assays.

Involvement of Trk1/2 and Hal3 in the role of TMKP1 in lithium tolerance

In an attempt to further understand how TMKP1 improves lithium tolerance in yeast, we overexpressed TMKP1 in strains lacking different elements related to lithium sensitivity. Our results show that TMKP1 can still enhance LiCl tolerance in the hypersensitive cnb1, hog1, tor1 (Fig. 6a–c), and sit4 (data not shown) genetic backgrounds. In contrast, in the case of strains lacking hal3 or trk1,2 (devoid of the high-affinity potassium transport system), an increase in tolerance to Li+ was not observed (Fig. 6a and b). These results suggest that the effect of TMKP1 depends on the activities of the K+ transporters Trk1,2 and of Hal3 (the inhibitory subunit of Ppz1,2 phosphatases). Transport of potassium through the Trk1,2 transporters is an important factor for normal salt tolerance in yeast (Gomez et al., 1994; Yenush et al., 2002, 2005). It is therefore plausible that TMKP1 may act to enhance the activity of Trk1,2 transporters especially when cells are exposed to toxic concentrations of lithium. If such an effect exists, we would expect a better growth of TMKP1-overexpressing yeast cells in K+ -limiting conditions. To check this possibility, we compared the relative growth of yeast cells expressing or not expressing TMKP1 under low potassium concentrations (0.25–2 mM). Under such conditions, a slight growth improvement of cells overexpressing TMKP1 was observed, especially in the presence of 1.5–2 mM of KCl (Fig. 7). Such data are in agreement with a positive effect of TMKP1 on Trk1,2 transporters. However, the intracellular K contents registered in TMKP1-overexpressing strain exposed for 1 h to LiCl in the presence and in the absence of KCl are comparable to the control strain submitted to the same treatments (Fig. S1). Therefore, although the Li+ tolerance associated to TMKP1 requires Trk1,2 function, the positive effect of TMKP1 on these transporters is not likely caused by changes in the ability of the transporters to discriminate Li+ from K+.

Figure 6

Effect of overexpression of TMKP1 on the lithium tolerance of several mutants involved in saline homeostasis. Cells of the BY4741 strain and the indicated derivatives cnb1, trk1,2 (a); hog1, hal3 (b) and tor1 (c) were transformed with pY-TMKP1 or pØ (empty pYES2 vector) and were grown overnight in synthetic medium lacking uracil and with 2% galactose as carbon source. Cultures were spotted on YPGal plates containing different concentrations of LiCl as described earlier. Growth was recorded after 4 days.

Figure 7

TMKP1 improves yeast growth under low KCl concentrations. Cells of wild-type strain BY4741 transformed with pYES2 vector or pY-TMKP1 were grown until saturation in synthetic medium lacking uracil and with 2% glucose and inoculated on K+-free Translucent medium containing the indicated amounts of KCl and in the presence of glucose or galactose as a carbon source at A660 of 0.005. Growth was resumed for 16 h. Data are expressed as percentage of growth respective to equivalent cultures grown on 50 mM KCl and are mean from three independent experiments. Asterisks denote P< 0.05.

It was previously reported that the particular high Li+ toxicity in the presence of galactose can be attributed to its capacity to induce a ‘galactosemia-like phenotype’ by reducing the rate of fermentation and promoting the accumulation of toxic byproducts. Thus, a beneficial effect of overexpressing GRE3, encoding an aldose reductase able to convert galactose to galactitol, was reported (Masuda et al., 2008). We show here that the overexpression of TMKP1 is unable to increase Li+ tolerance in cells lacking GRE3 (Fig. 8). In contrast, the effect can be still observed in cells lacking GLO1 or GLO2 (see ). Similarly, the beneficial effects of overexpression of GRE3 and TMKP1 are not additive (data not shown). This suggests that TMKP1 could be impinging on the Li+ toxicity associated with growth on galactose.

Figure 8

The TMKP1-mediated lithium tolerance in galactose-containing medium requires GRE3. The BY4741 strain and its gre3, glo1, or glo2 derivatives were transformed with pYES-TMKP1 or with empty pYES2 vector and grown overnight in synthetic medium lacking uracil and with 2% galactose as a carbon source. Then, cultures were spotted on YPGal plates with or without 25 mM of LiCl and incubated for 4 days.


In all eukaryotes, the MAPK-signaling pathways are the most relevant modules orchestrating cellular responses to external stimuli. The magnitude as well as the duration of MAPK activation by phosphorylation are both essential in determining the biological outcome (Qi & Elion, 2005; Kondoh & Nishida, 2007). This highlights the importance of negative regulatory mechanisms which engage protein phosphatases able to control and dephosphorylate specific MAPKs. In this context, the role of DSP, known also as MKPs, in regulating cellular stress responses via MAPK dephosphorylation is getting increasing interest. This work is an attempt to investigate the role of wheat MKP (TMKP1) in the cellular response to salt stress. For this purpose, we decided first to overexpress TMKP1 in the well-established eukaryotic model S. cerevisiae. The yeast cells overexpressing TMKP1 were shown to be more tolerant to salt stress and especially to LiCl. This tolerance seems to be rather specific as these cells are not more resistant to other toxic cations such as hygromycin, tetramethylammonium (TMA), and spermine (data not shown). Increased tolerance to alkaline cations, such as Li+ and Na+, can be gained by increasing expression of the Ena1 Na+-ATPase, which pumps Na+ and Li+ cations out of the cell (Ruiz & Arino, 2007). However, the TMKP1-mediated lithium tolerance cannot be linked to Ena1 Na+-ATPase because TMKP1 still can increase the tolerance to LiCl in the ena1-5 background, and furthermore, the lithium-induced expression of ENA1 is not significantly affected by TMKP1. Consistent with this is that intracellular Li+ contents are not altered in yeast overexpressing TMKP1 when compared to wild-type cells.

Intriguingly, the TMKP1-mediated lithium tolerance is observed only when galactose was used as a sole carbon source. The presence of galactose is not crucial for inducing a high expression of TMKP1 but for another unknown mechanism leading to the enhanced lithium tolerance. As in humans, galactose is metabolized in yeast by the Leloir pathway where the glucose 1-phosphate is converted by the phosphoglucomutase (Pgm2) to glucose 6-phosphate, which is further metabolized through glycolysis (Boles et al., 1994). It has been previously reported that lithium inhibits the rate of fermentation by blocking the Pgm2 activity (Masuda et al., 2001). As a result, lithium induces in the presence of galactose a cellular toxicity known as a ‘galactosemia-like’ phenotype that can be ameliorated by overexpression of GRE3. likely by transforming galactose 1-phosphate into galactitol. We observe here that the effects on Li+ tolerance of overexpressing both GRE3 and TMKP1 are not additive and that the effect of overexpression of the wheat phosphatase is lost in gre3 cells. Gre3 is also involved in the degradation of the toxic intermediate methylglyoxal (Aguilera & Prieto, 2001), by transforming this compound into l-lactaldehyde. However, our data suggest that this function is not at the basis of the Li+-tolerant phenotype of TMPK1-expressing cells, because the effect of overexpressing this phosphatase is still observable in cells lacking GLO1 or GLO2, encoding enzymes able to detoxify methylglyoxal to d-lactic acid by an independent pathway (Inoue & Kimura, 1995). Therefore, TMKP1 may contribute to reduce the exacerbated lithium toxicity in cells grown in galactose-containing medium.

To investigate further the involvement of TMKP1 in lithium tolerance, we undertook growth assays of mutant series overexpressing this phosphatase. Our data reveal that this tolerance does not occur in ptc1, hal3. or trk1,2 mutants which are all known as hypersensitive to LiCl (see Arino et al., 2010 and references therein). It is worth noting that only the lack of Ptc1 (but not the other six members of the PP2C family) results in a rather strong phenotype of lithium sensitivity, comparable to that produced by lack of well-known regulators of saline tolerance, such as Hal3 (Ruiz et al., 2006 and data not shown). In this regard, it must be noted that deletion of Hal3 does not aggravate the Li+-sensitive ptc1 phenotype (and vice versa), thus configuring a signaling pathway that affects the Trk transporters in a still obscure way (Ruiz et al., 2006). The fact that overexpression of TMKP1 fails to confer Li+ tolerance in hal3 or ptc1 mutants suggests that, somehow, the plant phosphatase might be affecting this pathway. The observation that the effect of TMKP1 is also dependent on the Trk transporters, a target of the above-mentioned pathway, is coherent with this hypothesis. Our finding suggests that TMKP1 leads to enhanced lithium tolerance by ultimately promoting the activity of Trk1,2 transporters. Trk1,2 can transport Li+ in addition to K+ (Haro & Rodriguez-Navarro, 2003). As we do not observe a decrease in intracellular lithium upon overexpression of TMKP1, a possibility would be that transport of potassium could be (even moderately) increased in the presence of stressing amounts of Li+ in the medium. This possibility would agree with our observation that overexpression of TMKP1 slightly improves gvates the enzyme. Furthermore, the ability of the overexpressed TMKP1 to confer lithium tolerance is eliminated when the protein contains a C214G mutation (Fig. 2b), whereas the phenotype of the yeast cells transformed with a version carrying a mutation outside the catalytic domain of TMKP1 (TMKP1P303T, used here as an additional control) is indistinguishable from the strain overexpressing TMKP1. All these results serve as a proof of concept and confirm that TMKP1 confers the lithium tolerance phenotype via its phosphatase activity.

Supporting Information

Additional Supporting Information may be found in the online version of this article:

Fig. S1. Measurements of intracellular K+ and Li+ contents.

Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.


We thank H. Mahjoubi and M. Robledo for technical help. We acknowledge Francisca Rández-Gil and José A. Prieto for reagents and advice. This work was supported by grants from the Ministry of Higher Education and Scientific Research and the Tunisian-Spanish Grant A/025376/09, and grants BFU2011-30197-C3-01 (MICINN) and A/016028/08 (AECID, Ministry of Foreign Affairs and Cooperation) to J.A.


  • Editor: Ian Dawes


View Abstract