•Phospholipase D (PLD) hydrolyzes phospholipids to produce phosphatidic acid (PA) and a head group, and is involved in the response to various environmental stresses, including salinity. Here, we determined the roles of PLDα and PA in the mediation of salt (NaCl)-stress signaling through the regulation of mitogen-activated protein kinase (MAPK or MPK) in Arabidopsis thaliana.
•NaCl-induced changes in the content and composition of PA were quantitatively profiled by electrospray ionization–tandem mass spectrometry (ESI-MS/MS). A specific PA species (a MAPK 16:0–18:2 PA), which was increased in abundance by exposure to NaCl, bound to a MPK6, according to filter binding and ELISA. The effect of PA on MPK6 activity was tested using in-gel analysis.
•16:0–18:2 PA stimulated the activity of MPK6 immunoprecipitated from Arabidopsis leaf extracts. Treatment with NaCl induced a transient activation of MPK6 in wild-type plant, but the activation was abolished in the pldα1 plant mutant. A plasma membrane Na+/H+ antiporter (SOS1) was identified as a downstream target of MPK6. MPK6 phosphorylated the C-terminal fragment of SOS1. The MPK6 phosphorylation of SOS1 was stimulated by treatment with NaCl, as well as directly by PA.
•These results suggest that PA plays a critical role in coupling MAPK cascades in response to salt stress.
Soil salinity has been regarded as a serious agricultural problem, affecting at least 20% of the irrigated land throughout the world (Yamaguchi & Blumwald, 2005; Munns & Tester, 2008). High salinity in soil inhibits the growth and development of plants through osmotic stress, ionic toxicity and ionic imbalance (Ward et al., 2003; Kim et al., 2007; Munns & Tester, 2008). To survive in high-salt soil, plants have evolved complex adaptive mechanisms to perceive and respond to salt stress (Rus et al., 2005). The maintenance of a high K+ : Na+ ratio in cytosol is an important strategy, because cytosolic enzymes in both glycophytes and halophytes show sensitivity to Na+ (or Cl−) (Yamaguchi & Blumwald, 2005). The control of Na+ influx, the extrusion of Na+ and the vacuolar compartmentation of Na+ are effective pathways for plants to maintain a high cytosolic K+ : Na+ ratio at the cellular level (Zhu, 2003; Yamaguchi & Blumwald, 2005; Munns & Tester, 2008). Although the mechanism of Na+ influx into plant cells still remains elusive, voltage-independent channel (VIC) or nonselective cation channels (NSCC) have been suggested as the routes for Na+ influx in root cells based on the results from electrophysiological analyses (Amtmann et al., 1997; Roberts & Tester, 1997; Apse & Blumwald, 2007). On the other hand, the extrusion of Na+ from cells is achieved by the plasma membrane Na+/H+ antiporter, SOS1 (Shi et al., 2000), which is driven by proton pumps on the plasma membrane. The transport of Na+ into the vacuole is mediated by the Na+/H+ antiporter, NHX1, in Arabidopsis (Apse et al., 1999). In the bryophyte Physcomitrella patents, the exclusion of Na+ from cells occurs via the Na+-ATPase, which is absent in vascular plants (Lunde et al., 2007). At the tissue level, OsHKT1;5 (previously OsHKT8 or SKC1), a Na+ selective transporter, which is expressed in xylem parenchyma cells, functions by removing Na+ from the xylem sap and reducing the transportation of Na+ to sheets in rice (Oryza sativa) (Ren et al., 2005).
The activities of transporters or channels are tightly regulated by complex signaling networks in plant cells. Calcium acts as a second messenger that mediates Na+ and K+ homeostasis in plant cells. Cytosolic calcium increase induced by salt stress is sensed by a calcineurin B-like protein, SOS3, which interacts physically with, and activates, a Ser/Thr protein kinase, SOS2, and the latter positively regulates the plasma membrane Na+/H+ antiporter, SOS1 (Zhu, 2003). More recently, SOS3-like calcium binding protein 8 (SCaBP8), also known as calcineurin B-like 10 (CBL10), was identified (Quan et al., 2007). Both SCaBP8 and SOS3 function in Arabidopsis salt tolerance by activating SOS2 (Liu et al., 2000; Quan et al., 2007). SCaBP8, not SOS3, is in turn phosphorylated by SOS2, and this phosphorylation is essential for the regulation of salt tolerance in Arabidopsis (Lin et al., 2009).
Other reported pathways of salt signaling include mitogen-activated protein kinase (MAPK or MPK) cascades (Nakagami et al., 2005). A MAPK cascade consists of a MAPK kinase kinase (MAPKkk)-MAPK kinase (MAPKK or MKK)-MAPK module that links receptors to downstream targets. MPK3, MPK4 and MPK6 are activated by NaCl stress (Ichimura et al., 2000). Overexpression of MKK2 leads to constitutive MPK4 and MPK6 activities and the increase of salt and freezing tolerance. Yeast two-hybrid assays reveal that MPK4 and MPK6 are MKK2 targets (Teige et al., 2004), whereas, in response to the plant hormone jasmonic acid and to ethylene, MPK6 is activated by MKK3 and MKK9, respectively (Takahashi et al., 2007; Yoo et al., 2008). These results suggest that different stresses or hormones activate MPK6 by different MAPKKs. In yeast, an activated MAPK (high osmolarity glycerol (Hog1)) phosphorylates an Na+/H+ antiporter (Proft & Struhl, 2004). This finding directly connects the MAPK pathway with salt tolerance in yeast. However, little is known about the upstream signaling events that activate MAPK cascades in plants and yeasts.
In animal cells, after the activation by stimuli, elevated amounts of phosphatidic acid (PA) from phospholipase D (PLD) hydrolysis recruit the MAPKKK Raf-1 to the membrane, where it is activated by RasGTP (which is activated by the guanine nucleotide exchange factor Sos) (Ghosh et al., 2003). PLD and PA are also involved in a diverse spectrum of developmental and environmental signal-transduction activities in plant cells (Testerink & Munnik, 2005; Wang et al., 2006; Hong et al., 2010; Zhang et al., 2010). PLDδ expression is induced by NaCl and dehydration (Katagiri et al., 2001). Knockout of PLDα3 increases sensitivity to salinity and water deficiency (Hong et al., 2008), while the double mutant, plda1/pldδ, was shown to be more sensitive to salt than either of the single mutants (Bargmann et al., 2009a). Phosphatidic acid has been found to bind to a Raf-1-like MAPKKK, constitutive triple response 1 (CTR1), which negatively regulates ethylene responses in Arabidopsis (Testerink et al., 2007). An elevation in the amount of PA is essential for MAPK activation induced by wounding in suspension-cultured soybean (Glycine max) cells (Lee et al., 2001). These results promoted us to investigate whether there is a relationship between lipid and MAPK signals in response to salt stress. Our results show that PLDα1-derived PA interacts with MPK6, which targets to SOS1.
Materials and Methods
Plant materials and growth condition
All mutants of Arabidopsis thaliana used in this study are in the Columbia ecotype background. The identification of pldα1 (a transfer DNA (T-DNA) insertional knockout mutant of PLDα1) was reported previously (Zhang et al., 2004). MPK6-RNA interference (RNAi) and MPK6-knockout (KO) seeds were kindly presented by Dr Brian Ellis from the University of British Columbia (Miles et al., 2005) and by Dr Justin Lee from the Leibniz Institute of Plant Biochemistry (Germany) (Bethke et al., 2009) respectively. Seeds were sown on half-strength Murashige and Skoog medium containing 3% sucrose or soil, and stored at 4°C for 2 d. Plants were grown in a growth room under a light intensity of 100 μmol m−2 s−1 and 12 h : 12 h (23°C : 18°C) day : night regimes.
Measurement of Na+ and K+ contents
Three-week-old seedlings grown in soil were transferred to a 2-l plastic pot containing aerated half-strength Hoagland solution. After 1 wk of culture at a light intensity of 100 μmol m−2 s−1 of and a 12 h : 12 h (23°C : 18°C) day : night regime, plants were treated with NaCl and the rosette leaves were harvested for Na+ and K+ assay, as described previously (Essah et al., 2003). Briefly, the samples were dried at 80°C for 2 d and extracted with 10 ml of 100 mM HNO3. The Na+ and K+ contents were determined using an atomic absorption spectrometer (TAS-986; Purkinje General Instrument Co., Beijing, China).
Electrolyte leakage assay
Salt stress-induced membrane injury was quantified by measuring ion leakage using a digital conductivity meter (DDS-11A; Shanghai Precision & Scientific Instrument Co., Shanghai, China). Rosette leaves were rinsed with deionized water, following 3 h of incubation in 10 ml of deionized water at 22°C with gentle shaking. The initial conductivity was then measured. Total conductivity was determined after boiling for 10 min. The conductivity was expressed as a percentage of the initial conductivity compared with the total conductivity. Leaves from three plants per genotype were assayed in each treatment, and three replicates of each treatment were carried out.
PLD activity analysis
PLDα activity was measured as described previously (Zhang et al., 2004) using protein extracted from leaves.
Four- to 5-wk-old plants grown in half-strength Hoagland culture solution were treated with NaCl for the indicated periods of time. Lipid extraction, electrospray ionization–tandem mass spectrometry (ESI-MS/MS) analysis and quantification were carried out as described previously (Devaiah et al., 2006).
Expression and purification of proteins in Escherichia coli
The complementary DNA (cDNA) sequence of the MPK6 gene was amplified from Arabidopsis leaf cDNA using the following primers: 5′-TATGGATCCATGGACGGTGGTTCAGGT-3′ (forward) and 5′-GACAAGCTTCTATTGCTGATATTCTGGATTGAA-3′ (reverse). The MPK6 cDNA was then cloned into the pET-28a vector (Novagen; Merk KGaA, Darmstacit, Germany), at BamHI and HindIII restriction sites, to produce a protein with six histidine residues fused at the N-terminus. The recombinant plasmid was transformed into E. coli BL21 (DE3) and the His–MPK6 recombinant protein thus produced was purified using Ni-nitrilotriacetic acid (Ni-NTA) agarose (Qiagen).
An Arabidopsis SOS1 cDNA clone was generously provided by Dr Yan Guo (National Institute of Biological Sciences, China). cDNA fragments of the SOS1 C-terminus (designated as SOS1CT) were amplified by PCR using SOS1 cDNA as the template. Seven cDNA fragments for SOS1CTs were named SOS1CT1 (444–1146 amino acids), SOS1CT2 (544–1146 amino acids), SOS1CT3 (644–1146 amino acids), SOS1CT4 (748–1146 amino acids), SOS1CT5 (848–1146 amino acids), SOS1CT6 (948–1146 amino acids) and SOS1CT7 (1048–1146 amino acids). The forward primers for these seven SOS1CT fragments were: SOS1CT1, 5′-CGGAATTCACCCAATTTGTTCTACGCCTTCTTC-3′; SOS1CT2, 5′-CGGAATTCTGGGAGATGCTT GATGAGGGCAGAA-3′; SOS1CT3, 5′-CGGAATTCTCCATTGTAATCAATGAAAGTGAAA-3′; SOS1CT4, 5′-CGG AATTCGCTCTTCCTCCTGCATTTTGTGAAC-3′; SOS1CT5, 5′-CGGAATTCATCGATGATTTCCTT TGGCAGGAAA-3′; SOS1CT6, 5′-CGGAATTCAGAGTCAGTTTCTCACAACAAGCAA-3′; and SOS1CT7, 5′-CGG AATTCTACAGAAGGAGTGTAAGTTTCGGTG-3′. The common reverse primer was 5′-ACGCGTCGACTCATAGATCGTTCCTGAAAACGATT-3′. The cDNA sequences of the Arabidopsis MKK2 gene were amplified from leaf cDNA using the following primers: 5′-GCGAATTCATGAAGAAAGGTGGATTC-3′ (forward) and 5′-CGGTCGACCTTACACGGAGAACG-3′ (reverse). The SOS1CT and MKK2 cDNA inserts were digested with EcoRI/SalI and ligated into pGEX-4T-1. All fusion constructs were transformed into E. coli BL21 (DE3) to produce the proteins with glutathione S-transferase (GST) fused at the N terminus. The GST recombinant proteins were purified with glutathione–Sepharose beads (GenScript, Piscataway, New Jersey , USA). All cDNAs and the junction of His- or GST-fusion proteins were confirmed by sequencing.
For protein expression, E. coli cultures grown at 37°C to an optical density of 0.5–1 at 600 nm were induced with 0.5 mM isopropyl thio-β-d-galactoside (IPTG) and grown for an additional 6 h at 22°C before harvesting.
The lipids used in this study were obtained from Avanti Polar Lipids (Alabaster, AL, USA). The binding of His–MPK6 to lipids immobilized on a nitrocellulose filter was performed as described previously (Zhang et al., 2004). Briefly, after the lipid-bound filter was treated with a protein solution, the filter was incubated with the antibody against His-tag (1:1000) (Sigma) and then incubated with a second antibody conjugated to alkaline phosphatase (1:2500) (Sigma). The MPK6 proteins bound to lipids on filters were visualized by staining for alkaline phosphatase activity.
The ELISA-based assay was carried out as described previously (Zhang et al., 2009). Phospholipids were coated on the wells of 96-well titer plates and incubated with His-tagged MPK6 or RbohD330 (RbohD100), followed by incubation with the antibody against His-tag and spectrometric measurements.
The GST fusion proteins expressed in E. coli BL21 (DE3) were incubated with glutathione–Sepharose beads rotating at 4°C for 6 h, then mixed with 40 μg of recombinant His-tagged MPK6 protein at 4°C for 2 h. The beads were precipitated by microcentrifugation, washed four times with wash buffer (20 mM Tris (pH 8.0), 200 mM NaCl, 1 mM EDTA, 0.5% Nonidet P-40, 2 μg ml−1 of aprotinin, 1 μg ml−1 of leupeptin, 1 mM phenylmethanesulfonyl fluoride) and resuspended in 5 × sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer. The resuspended protein sample was boiled at 100°C for 5 min, separated on a 10% SDS polyacrylamide gel, then immunoblotted using the antibody against His-tag.
MPK6 immunoblotting and the immunocomplex kinase activity assay
Proteins were extracted from leaves. To analyze MPK6 protein expression in wild-type (WT) and MPK6-null mutants, the extracts were separated by electrophoresis on a 12% SDS–polyacrylamide gel and blotted with antibodies against MPK6 (Sigma).
For the immunocomplex kinase activity assay, the protein extracts (100 μg) were incubated with 1.5 μg of the antibody against MPK6 in protein-extraction buffer on a rotator at 4°C overnight. Fifteen microlitres of a packed volume of protein A–sepharose beads (Sigma) was added and the incubation was continued for another 2 h. The beads were collected by centrifugation at 4000 g for 1 min, washed three times with 1 ml of wash buffer (25 mM Tris (pH 8.0), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM Na3VO4, 1 mM NaF, 10 mM β-glycerophosphate, 5 μg ml−1 of leupeptin, 5 μg ml−1 of aprotinin, 5 μg ml−1 of antipain and 0.1% Tween 20) and once with kinase buffer (25 mM Tris (pH 8.0), 1 mM EGTA, 5 mM MgCl2 and 1 mM dithiothreitol (DTT)). Kinase activity was assayed at 30°C for 30 min in a final volume of 30 μl containing 0.5 mg ml−1 of myelin basic protein (MBP) (or SOS1CT7), 10 μM ATP, 10 μCi of [γ32P]-ATP (> 4000 Ci mmol−1) and the beads with MPK6. The action was stopped by the addition of SDS-PAGE sample loading buffer. After electrophoresis on a 12% SDS–polyacrylamide gel, the phosphorylated substrates were visualized by autoradiography.
Abrogation of PLDα1 renders Arabidopsis seeds and seedlings sensitive to NaCl
To test whether the most abundant PLD, PLDα1, is involved in salt tolerance, we compared the seed germination and seedling growth between WT and pldα1 mutant at different concentrations of NaCl. Seeds were germinated on half-strength Murashige and Skoog medium, with or without NaCl. In the absence of NaCl, pldα1 and WT seeds had the same rate of germination. Both WT and pldα1 seeds showed a decrease in germination with increasing NaCl concentrations in the medium. However, 50–150 mM NaCl inhibited the germination of pldα1 seeds more than that of WT seeds (Fig. 1a). It is known that high salinity induces both osmotic stress and ionic toxicity to plant cells (Rains & Epstein, 1967). To determine the role of PLDα1 in response to osmotic and ionic toxicities of salt stress, we compared the effects of NaCl and osmotica mannitol on seed germination. As shown in Fig. 1(b), both 200 mM mannitol and 100 mM NaCl reduced the percentage of germination, but the magnitude of inhibition induced by mannitol was lower than that induced by NaCl. These data suggest that PLDα1 is involved in the response to both osmotic and ionic toxicities induced by salt stress.
When seedlings were transferred to half-strength Murashige and Skoog medium containing 150 mM NaCl, c. 80% of the WT seedlings and 28% of the pldα1 seedlings survived after 10 d. No significant difference in survival rates was found between WT and pldα1 seedlings when the NaCl concentration was 75 or 250 mM (Fig. 1c). Taken together, our results suggest that PLDα1 is involved in salt tolerance in seed germination and seedling growth in Arabidopsis.
pldα1 mutant accumulates more Na+ in leaves
To further dissect the physiological function of PLDα1 in salt tolerance, we analyzed Na+ and K+ accumulation in seedlings grown in culture medium with or without NaCl. As shown in Fig. 2, more Na+ was accumulated in the leaves of plda1 seedlings than in the leaves of WT seedlings when they were grown in culture medium containing100 mM NaCl for 6 d. Regarding K+ contents in the leaves, there was no difference between WT and pldα1 seedlings in the presence or absence of NaCl (Fig. 2a). Treatment with NaCl led to an increase in Na+ accumulation and a decrease in K+ accumulation in roots, and both WT and pldα1 seedlings showed a similar pattern in Na+ and K+ changes (Fig. 2b).
NaCl-induced cellular damage was quantitatively assessed by measurements of ion leakage, indicative of increased membrane permeability. Leaves from plda1 seedlings displayed more serious damage in plasma-membrane integrity after 6 d of exposure to NaCl compared with WT leaves (Fig. 2c).
NaCl induces PLDα activation and PA increase
To further confirm that PLDα1 is involved in the response to salt, we examined PLDα activity in Arabidopsis. In WT seedlings, PLDα was activated in response to treatment with 100 mM NaCl. The increase in PLDα activity peaked at 0.5 h after NaCl addition (Fig. 3a). In the plda1 mutant, PLDα activity was almost abolished, and the increase of the activity was negligible during NaCl treatment (Fig. 3a).
We next determined the change of the PLD hydrolysis product PA. The results shown in Fig. 3(b) indicate that treatment with NaCl leads to an increase in the amount of total PA produced in WT seedlings at 0.5 h of NaCl treatment. The total amount of PA produced decreased below that of the control (time 0) at 3 h of NaCl treatment. In the pldα1 mutant, there was a gradual decrease in PA content with prolonged NaCl treatment (Fig. 3b).
Analysis of PA molecular species revealed further distinguishable changes between WT and pldα1 seedlings after treatment with NaCl (Fig. 3c). The major molecular species of PA in WT were 34:2 (16:0–18:2), 34:3 (16:0–18:3) and 36:5 (18:2–18:3). After 0.5 h of NaCl treatment, PA species 34:2, 34:3, 34:6, 36:3 and 36:6 increased significantly compared with the control (time 0) (Fig. 3c). By contrast, most PA species decreased when the plda1 seedlings were treated with NaCl, except for 36:3 and 36:5. The results show that abrogation of PLDα1 leads to changes in the amount and type of molecular species of PA produced in response to treatment with NaCl.
PA interacts with MPK6 and stimulates its activity
To unravel the role of the NaCl-induced increase of PA, we investigated potential targets of PA involved in salt tolerance. The interaction of PA with MPK6 was studied because this MAPK was reported to be involved in the regulation of salt tolerance in Arabidopsis (Teige et al., 2004). MPK6 was expressed in E. coli as a His-tagged protein using its cDNA cloned from WT Arabidopsis (Fig. 4a). The purified protein was tested for lipid binding using a nitrocellulose filter-binding assay (Zhang et al., 2004). We used PA-binding respiratory burst oxidase homolog D (RbohD) fragment D330 (330 amino acids) as a positive control, and non-PA binding D100 (100 amino acids) as a negative control (Zhang et al., 2009). As shown in Fig. 4(b), MPK6 bound to natural PA, but not to other lipids, such as phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylglycerol (PG), phosphatidylserine (PS), or diacylglycerol (DAG). A series of PA species was then tested for the specificity of MPK6 binding. Both MPK6 and D330 showed significant binding to dioleoyl PA (di 18:1 PA), dilinoleoyl PA (di 18:2 PA), palmitoyl-oleoyl PA (16:0–18:1 PA), palmitoyl-linoleoyl PA (16:0–18:2 PA) and stearoyl-palmitoyl PA (18:0–18:2 PA). MPK6 bound weakly, while D330 did not bind at all, to dipalmitoyl PA (di 16:0 PA) and distearoyl PA (di 18:0 PA) (Fig. 4c). Weaker binding of MPK6 to di 16:0 PA or di 18:0 PA than to other PA species was not caused by more MPK6 being washed off the nitrocellulose filter, according to the comparison of the lipid association with the filter before and after protein–lipid incubation (Supporting Information Fig. S1).
PA–MPK6 binding was quantitatively characterized using ELISA analysis. 16:0–18:2 PA, a NaCl-induced and commercially available PA species, was coated onto microtiter plates and incubated with His-tagged MPK6. Compared with the negative control (RbohD100), MPK6 and RbohD330 (the positive control) showed significant binding to 16:0–18:2 PA. There was no significant difference for binding to 16:0–18:2 PC among the three proteins (Fig. 4d).
To address whether PA directly stimulates the MPK6 kinase, we tested the effect of exogenously added PA on MPK6 activity using MBP as a substrate. MPK6 was immunoprecipitated from the protein extracts of WT leaves using an MPK6-specific antibody. When 16:0–18:2 PA was incubated with immunoprecipitated MPK6 protein, it stimulated the kinase activity at 5 and 10 μM (Fig. 5a,b). 16:0–18:2 PC at the same concentrations as PA showed no stimulatory effect on the MPK6 kinase activity (Fig. 5a). To characterize the function of the PA–MPK6 interaction in the salt-stress response, we compared MPK6 kinase activity in WT and pldα1 seedlings. Endogenous kinase activity of MPK6 from pldα1 and WT leaf extracts was determined after NaCl treatment using immunocomplex kinase assays. Treatment with NaCl resulted in the rapid activation of endogenous MPK6 in WT seedlings, but not in pldα1 mutants (Fig. 5c, upper panel). The content of MPK6 protein was kept constant during the treatment with NaCl (Fig. 5c, bottom panel), suggesting that the activation of MPK6 is mainly induced by post-translational modification. Taken together, these data suggest that PLDα1-derived PA binds to MPK6 and leads to the kinase activation in response to NaCl stress.
MPK6-null seedlings show salt sensitivity that cannot be alleviated by exogenous PA
To establish if MPK6 plays a role in salt tolerance, we compared the salt tolerance between WT and MPK6-null plants. MPK6 protein expression was abolished in MPK6-RNAi and knockout plants, as detected by immunoblotting (Fig. 6a). Neither MPK6-RNAi nor knockout seedlings showed an obvious phenotype compared with WT seedlings under normal ambient conditions. However, they showed increased sensitivity to NaCl (150 mM) relative to WT seedlings, evident as yellower and smaller leaves and shorter roots (Fig. 6b,c). When seeds were germinated for 10 d on half-strength Murashige and Skoog medium containing 150 mM NaCl, WT, MPK6-RNAi and KO mutants had 36%, 12% and 7% germination, respectively. These results suggest that MPK6 is necessary for salt tolerance in Arabidopsis.
We next investigated whether the PA–MPK6 interaction was essential for salt tolerance in Arabidopsis. To address this question, exogenous PA (16:0–18:2 PA) was added to the medium together with NaCl. Although the sensitivity of MPK6–RNAi to NaCl was comparable with that of pldα1, the NaCl inhibition of growth on pldα1, but not on MPK6–RNAi, was restored by exogenously applied PA (Fig. 6d). The results suggest that the PA–MPK6 interaction is important in salt tolerance, and PA acts upstream of MPK6 in the salt-stress response.
MPK6 interacts with SOS1
We attempted to find the MPK6 target(s) in salt-stress signaling. Given that MPK6 and SOS1 are involved in salt tolerance in plants (Zhu, 2003; Fig. 6), we examined whether MPK6 interacts physically with SOS1 in vitro. It has been proposed that the SOS1 protein contains hydrophobic N-terminal transmembrane domains and hydrophilic C-terminal domains (Shi et al., 2000; Zhu, 2003). As it is unlikely that a MAPK would phosphorylate a transmembrane domain, we generated serial C-terminal fragments of SOS1 (SOS1CT) fused to GST and expressed the fused proteins in E. coli. (Fig. 7a). Then, His-tagged MPK6 was incubated with GST–SOS1CTs. The proteins pulled down by GST–SOS1CTs were blotted with the antibody against His-tag. The results shown in Fig. 7(b) (lower panel) indicate that a single band corresponding to the size of MPK6 was detected when GST–SOS1CT7 (amino acids 1048–1146) was incu-bated with His–MPK6, but GST alone did not pull down His–MPK6. As a positive control, GST–MKK2 could pull down MPK6 protein (Fig. 7b). This experiment suggests that MPK6 interacts directly with the soluble parts of SOS1.
We next established if MPK6 directly phosphorylates SOS1CT. MPK6 was immunoprecipitated from either nonstressed or salt-stressed WT plants with the antibody against MPK6. GST–SOS1CT7, which interacted with MPK6, was expressed in E. coli and purified as a substrate. As shown in Fig. 7(c), SOS1CT7 was efficiently phosphorylated in vitro by immunoprecipitated MPK6 from WT extracts. Furthermore, the phosphorylation was more efficient when MPK6 was immunoprecipitated from the plants treated with NaCl. Together with the earlier observations that NaCl induced the generation of PA (Fig. 3) and that MPK6 (with MBP as a substrate) was activated by NaCl in WT but not in pldα1 (Fig. 5), these results imply that NaCl-induced PA may activate MPK6, which phosphorylates SOS1CT. Therefore, we tested the effect of PA on MPK6 phosphorylation of SOS1CT. As shown in Fig. 7(c), PA increased the activity of MPK6 immunoprecipated from nonstressed Arabidopsis leaves, resulting in an increase in its phosphorylation of SOS1CT7.
The main site of Na+ toxicity for most glycophytes is the leaf, the main tissue in which photosynthesis takes place. Net Na+ accumulation in leaves is a result of deposition from the Na+ transpiration stream and Na+ exclusion from leaves (Apse & Blumwald, 2007; Munns & Tester, 2008). The transpiration rate and intensity depend mainly on the stomatal aperture in plant leaves. It is regarded that the most dramatic and readily measurable response to salt stress at the whole-plant level is a reduction in stomatal aperture to reduce water evaporation and Na+ (or Cl−) transport to shoots from the xylem (Munns & Tester, 2008). In previous work, we demonstrated that the loss of PLDα1 resulted in the insensitivity to ABA during stomatal closure (Zhang et al., 2004; Mishra et al., 2006). However, knockout of PLDα1 did not reduce NaCl-induced ABA accumulation (Fig. S2), but partially impaired NaCl-induced stomatal closure (Fig. S3). Therefore, an increase in the transportation of Na+ to the leaves of pldα1 could, in part, be caused by the unclosed stomata in the mutant (Fig. 2).
Arabidopsis is a glycophyte species that is sensitive to moderate amounts of NaCl. Therefore, it is believed that an active efflux of Na+ is required in all cells throughout the plant (Munns & Tester, 2008). The Na+/H+ antiporter, encoded by SOS1, is a transporter that has been shown to mediate sodium efflux in the plasma membrane (Shi et al., 2000). SOS1 is regulated by the SOS2/SOS3 pathway (Zhu, 2003), as well as by SCaBP8 and reactive oxygen species (ROS) (Quan et al., 2007; Chung et al., 2008). Our results show that SOS1 might be regulated by a MAPK. In particular, MPK6 interacts with the C-terminal domain of SOS1 (SOS1CT7). Furthermore, SOS1CT7 can be phosphorylated by MPK6, and this phosphorylation is activated by a rapid salt treatment (Fig. 7). The classical role of MAPKs is the phosphorylation of transcription factors that directly induce the expression of genes which alleviate this stress (Asai et al., 2002; Proft & Struhl, 2004; Nakagami et al., 2005). In yeast, Hog1 MAPK mediates the rapid activation of a Na+/H+ antiporter (Nha1), which precedes the transcriptional response (Proft & Struhl, 2004). Our current data suggest that the regulation of the Na+/H+ antiporter by MAPKs may also exist in plants.
Because SOS1 is a transmembrane Na+/H+ transporter (Shi et al., 2000), one can imagine that MPK6-induced phosphorylation of SOS1 in the early response to NaCl presumably occurs near the plasma membrane. The binding of PA drives the translocation of MPK6 to the plasma membrane, making the MPK6–SOS1 interaction possible. In vitro data suggest that PA directly increases MPK6 phosphorylation of SOS1CT7 (Fig. 7c). Furthermore, the PA species (16:0–18:2 PA) elevated by exposure to NaCl is identical to that used for MPK6 binding and activity (Figs 3, 4, 7), suggesting that such binding and activation might occur in vivo. By contrast, the wounding-induced MAPK activation is not affected in the pldα/δ double mutant, in which an increase in PA is abolished (Bargmann et al., 2009b). This may imply that the role of PA in MPK activation is different in the responses to salt and wounding, or that PA may activate a specific MAPK (e.g. MPK6 in this paper), rather than all MAPKs (20 MAPKs in Arabidopsis) in the stress response.
The PA–MPK6 binding is expected to be transient because, after NaCl treatment for 3 h, the amount of PA decreased to that detected at baseline (Fig. 3b). Simulta-neously with the decrease in PA, MPK6 activation also declined (Fig. 5a). The disassociation of MPK6 from the membranes permits it to move into nuclei and bind to transcription factors; however, spatial and temporal evidence for this move is missing. The loss of PLDα1 also led to a lower amount of PA as well as a reduced MPK6 activity (Figs 3b, 5c). These results suggest that PA is an essential regulator of MPK6.
A common model for the activation of MAPKs is through MAPKK-catalyzed dual phosphorylation of a -TXY- motif (Jonak et al., 2002). MPK6 is regulated by various MAPKKs and MAPKKKs in response to different stimuli. For example, MPK6 is upregulated by MKK2 (mitogen-activated protein kinase kinase 2) and MEKK1 (MAPK or ERK kinase kinase 1) in response to the presence of salt (Teige et al., 2004), and is upregulated by MKK5 and MEKK1 in response to pathogens (Chung et al., 2008). In animal and yeast cells, MAPK scaffolding molecules assemble specific MAPK pathway components in particular regions (Morrison & Davis, 2003). The scaffold proteins enhance downstream signaling by increasing the local concentration of the target kinases in the signaling cascade, or by providing tighter regulation of the signaling strength (Anderson, 2006). In animal cells, the kinase suppressor of Ras (KSR) acts as a scaffold protein for a MAPK cascade, Raf-MEK (MAPK or ERK kinase)-ERK (extracellular signal-regulated kinase). In this cascade, PA at least binds to KSR1, Raf-1 and guanine nucleotide exchange factor Sos (an activator of Ras, which activates Raf-1) (Ghosh et al., 1996; Zhao et al., 2007; Kraft et al., 2008). The generation of a local pool of PA by the stimuli is even proposed as a molecular scaffold in coupling of the ERK cascade (Kraft et al., 2008). In plants, PA was recently reported to bind to a Raf-1 like MAPKKK, CTR1 (Testerink et al., 2007). In addition to the PA–MPK6 interaction reported here, we found that MKK2 was bound by PA (data not shown). Therefore, it will be of interest to determine whether the salt-induced increase of PA acts a link to MAPK cascades. It should be noted that the PA stimulation of MPK6 activity does not rule out the activation by other factors, for example, MPKKs, ROS, OXI1 (oxidative signal-inducible 1), PDK1 and PTI1-2(Pto-interacting 1-2) (Asai et al., 2002; Anthony et al., 2004, 2006). PTI1-2 kinase is a partner of OXI1, which is regulated by PDK1, while PDK1 itself is a PA target. These results suggest that MPK6 may be regulated by a complicated network in plant cells.
All three mutants –pldα1, mpk6 and sos1– displayed sensitivity to NaCl (Shi et al., 2002; Figs 1, 6), but they were not fully identical in the physiological patterns of sensitivity. For example, a higher content of Na+ was found in the leaves of the plda1 mutant than in the leaves of the WT plant (Fig. 2a), while a higher Na+ content was found in both the leaves and roots of the sos1 mutant than in the WT plant (Shi et al., 2002). In addition, Na+ impaired the K+ permeability of sos1 root-cell membranes (Qi & Spalding, 2004). However, in pldα1 plants treated with NaCl, the K+ content was similar to that of the WT plant in shoots and roots (Fig. 2a,b). One explanation for these observations is that other PLD isoforms involved in the salt response, such as PLDα3 and PLDδ (Hong et al., 2008; Bargmann et al., 2009a,b), may also contribute to the regulation of SOS1. Moreover, PLD/PA is only one mediator of SOS1 pathways.
In conclusion, we report that salt stress induces a transient increase in the amount of PA. This PA binds to MPK6 and stimulates its kinase activity, which phosphorylates the SOS1 Na+/H+ antiporter. The PA functions as a salt signaling molecule through at least two routes: one via MPK6–SOS1 regulation to exclude Na+ from cells, and the other via its regulation of stomatal closure to reduce the transpiratory steam of Na+ to leaves, therefore reducing the accumulation of Na+ in leaves. Knockout of PLDα1 results in the generation of less PA and reduced MPK6 activity, leading to the accumulation of more Na+ in leaves and increased sensitivity to NaCl stress.
We thank Dr Xuemin Wang at University of Missouri St Louis for critical reading of the manuscript and for kindly providing the pldα1 mutant. We thank Dr Brian Ellis at University of British Columbia (Canada) for kindly providing MPK6-RNAi seeds, Dr Justin Lee at Leibniz Institute of Plant Biochemistry (Germany) for kindly providing MPK6-KO seeds, Dr Yan Guo at National Institute of Biological Sciences in China for kindly providing SOS1 cDNA, and Dr Ruth Welti at Kasas State University for lipid profiling. The Kansas Lipidomics Research Center was supported by NSF grants MCB0455318 and DBI 0521587, and NSF EPSCoR grant EPS-0236913 with matching support from the State of Kansas through Kansas Technology Enterprise Corporation and the Kansas State University. The work was supported by the grants from the Ministry of Science and Technology of China (2006CB100100, 2008AA10Z122), the National Science Foundation of China (30625027, 90817014), the Ministry of Education of China (B07030), and the Department of Education of Jiangsu Province (200910).