Many plant species demonstrate a systemic increase in phosphatidic acid (PA) levels after being wounded (Lee et al., 1997). To understand the role of PA in wound signal transduction, we investigated if PA can activate protein kinases in soybean (Glycine max L.). We found that a MAPK is activated in soybean seedlings in both wounded and neighboring unwounded leaves. The wound-activated soybean kinase is specifically recognized by an antibody against the alfalfa MAPK, SIMK. When PA production is inhibited with n-butanol, an inhibitor of phospholipase D, the wound-induced activation of the MAPK is suppressed, suggesting that an elevation in PA levels is essential for its activation. Supporting this is the observation that exogenous PA activates the MAPK in suspension-cultured soybean cells. Activation of the 49 kDa MAPK occurs almost exclusively by PA, as other lipids are unable to or can only weakly activate the kinase. PA-induced activation of the MAPK is not a direct effect on the kinase but is mediated by upstream kinases. Our results suggest that PA acts as a second messenger in wound-induced MAPK signaling in plants.
Here we show that in soybean plants, wounding activates a SIMK-like MAPK. The wound-activation of the MAPK is inhibited when PA production is suppressed, and exogenously applied PA specifically activates the MAPK in suspension-cultured soybean cells. This suggests that PA participates as a second messenger in wound signal transduction by activating a specific MAPK cascade.
Wounding activates a 49-kDa protein kinase in soybean leaves
To assess the role of PA in wound signal transduction, we studied the relationship between PA and MAPK activation. We used the soybean as our subject because the increase of PA due to wounding is more consistent and pronounced in the soybean compared with other plant species tested, including tomato, tobacco, pepper, sunflower and broad bean (Lee et al., 1997). At both wounded and unwounded neighboring leaves of soybean, PA level, expressed as the percentage of PA in control leaves from non-wounded plants, was elevated (Figure 1a), and a protein kinase of 49 kDa was activated (Figure 1b). The kinetics of the protein kinase activation and the PA level elevation were very similar (Figure 1). Both were apparent as early as 2 min after wounding, and were stronger and more persistent at the wounded leaves than at unwounded neighboring leaves. The 49 kDa wound-activated protein kinase (WAPK) could phosphorylate both histone and myelin basic protein (MBP), but preferred MBP as a kinase substrate (data not shown). The molecular mass and substrate preference of the soybean WAPK are similar to those of MAPKs.
At least two MAPKs, SIPK and WIPK, are activated in tobacco after wounding (Mizoguchi et al., 1996; Seo et al., 1995; Zhang and Klessig, 1998a). In order to determine which group the soybean WAPK belonged to, we used antibodies specific for synthetic peptides corresponding to the C-terminal amino acids of alfalfa MAPKs SIMK, MMK2, MMK3 and SAMK (Jonak et al., 1993; Jonak et al., 1995; Jonak et al., 1996). All of the antibodies except MMK3 antibody cross-reacted with proteins of the expected sizes from crude extracts of soybean leaves and suspension-cultured cells (Figure 2a). MMK3 antibody also cross-reacted with a protein of the expected size from soybean leaves, but not from suspension-cultured cells; MMK3-like protein kinase may be lacking in the cultured cells. Only the SIMK antibody immunoprecipitated a kinase that had been activated by wounding (Figure 2b).
PA activates the wound-activated 49 kDa protein kinase in soybean cells
To determine whether the WAPK activation is related to elevations in PA levels, we assessed the effect of n-butanol on soybean MAPK activation, using sec-butanol as a control. n-butanol has been widely used as an inhibitor of phospholipase D (PLD) in plant as well as animal systems (Chen et al., 1997; Jacob et al., 1999; Munnik et al., 1995; Nofer et al., 1997; Ritchie and Gilroy, 1998). PLD has been shown to be involved in systemic PA production in wounded castor bean leaves (Ryu and Wang, 1996). Two leaf-stage soybean seedlings were supplied with 0.1% of either n-butanol or sec-butanol through their cut stem for 30 min, and then one leaf was wounded. Five minutes after wounding, both wounded and unwounded neighboring leaves were harvested for measurement of either PA level or kinase activity. In the unwounded neighboring leaf, n-butanol completely inhibited both the PA increase (Figure 3a) and activation of MAPK (Figure 3b), while sec-butanol did not. However, in the wounded leaf, where PA levels are much higher than those seen systemically (Figure 1a, Lee et al., 1997; Ryu and Wang, 1996), n-butanol suppressed neither PA production nor MAPK activity (data not shown). At the wound site, non-specific hydrolases released from broken cells, rather than PLD, may be the major enzymes that produce PA and activate the WAPK.
As this experiment suggests that PA is involved in the wound-activation of the soybean MAPK, we examined whether exogenously applied PA can also activate the WAPK. When soybean leaves were infiltrated with either PA or water as a control according to previously described methods (Zhang and Klessig, 1998a), both induced activation of the WAPK (data not shown), perhaps because the infiltration process imposes mechanical or osmotic stresses. Intact soybean leaves may thus not be suitable for testing PA activation of the MAPK in this manner. Due to the limitations of the infiltration experiment, we decided to use suspension-cultured soybean cells to assess the effect of exogenous PA on soybean protein kinase. We found that 50 µm PA activates a 49-kDa protein kinase in these cells (Figure 4a). This protein kinase was similar to the 49 kDa WAPK, since it was immunoprecipitated by the SIMK but not the SAMK antibody (Figure 4b). Moreover, as with the seedling leaves, suspension-cultured soybean cells showed the 49 kDa WAPK activation in response to wounding, which was blocked by pretreatment with n-butanol (Figure 4c). In the suspension-cultured soybean cell system, PA activated the MAPK within 2–5 min after treatment (Figure 4a). The activity of the kinase peaked at 10 min and returned to basal levels 30 min after treatment. Control cells treated with water did not exhibit kinase activation at any time. The signals were not due to autophosphorylation because phosphorylation was not detected when MBP was omitted from the gel mixture (data not shown). The PA concentration-dependence of kinase activation showed a bell-shaped curve (Figure 5a); a PA concentration as low as 10 µm activated the MAPK, and the effect was maximal at 50 µm. At higher concentrations of PA (up to 200 µm), kinase activity was still higher than the controls but much lower than at 50 µm(Figure 5a). These observations indicate that the effect of PA on kinase activation is not due to a detergent effect.
PA selectively activates the soybean 49 kDa SIMK-like MAPK
To determine whether only PA can achieve activation of the soybean MAPK, soybean cells were treated for 5 min with 50 µm of other lipids, namely diacylglycerol and linoleic acid, and phospholipids such as lysophosphatidic acid, phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine, phosphatidylinositol, lysophosphatidylcholine, and lysophosphatidyl-ethanolamine. None of the lipids activated the protein kinase at levels comparable with that induced by PA, although slight activation was observed with lysophosphatidic acid and phosphatidylinositol (Figure 5b). The ineffectiveness of diacylglycerol, linoleic acid and lysophosphatidic acid shows that PA does not activate the protein kinase through its metabolites. However, it remains to be tested whether diacylglycerol pyrophosphate, a novel PA metabolite reported to form after PA signaling (Munnik et al., 2000; Van der Luit et al., 2000), activates the protein kinase. Dioctanoyl- and 1-palmitoyl, 2-linoleoyl-PA activated the protein kinase to similar extents as dipalmitoyl-PA did, suggesting that the length of the acyl chains in PA is not critical in its ability to activate the MAPK (data not shown).
PA-induced activation of the 49 kDa SIMK-like MAPK requires upstream phosphorylation
As MAPKs are activated by phosphorylation by upstream protein kinases, we tested whether staurosporin, a general protein kinase inhibitor, could alter PA-activation of the soybean MAPK. When suspension-cultured soybean cells were pretreated with 2 µm of staurosporin for 10 min, PA-induced activation of the 49 kDa protein kinase was completely suppressed (Figure 6a). We then reasoned that if PA activation of the protein kinase requires a phosphorylation step, activation might also be achieved by inhibiting dephosphorylation. Indeed, when soybean cells were treated for 30 min in the absence of PA with 1 µm calyculin A or okadaic acid (potent inhibitors of protein phosphatases 1 and 2 A), a soybean protein kinase was activated (Figure 6b). Like the PA-activated protein kinase (Figure 4b), this kinase could be immunoprecipitated by the SIMK antibody and was 49 kDa in size.
While PA has been shown to be involved in signal transduction pathways in plant cells (Jacob et al., 1999; Munnik et al., 2000; Ritchie and Gilroy, 1998), little is known about how PA acts as a signal mediator. Here, we show evidence that PA plays a critical role in wound signal transduction by activating a MAPK pathway.
Upon wounding, soybean leaves exhibit activation of a 49-kDa MAPK that crossreacts with an antibody that recognizes the alfalfa SIMK. Several lines of evidence support the involvement of PA production in the wound-induced activation of this MAPK. First, n-butanol, which inhibits the formation of PA by PLD, inhibits the wound-induced activation of the 49 kDa soybean protein kinase (Figures 3 and 4). Second, in correlation with PA levels in wounded plants (Figure 1a), the WAPK activity is higher in wounded than in unwounded neighboring leaves (Figure 1b). Third, treatment of suspension-cultured soybean cells with PA activates a 49-kDa SIMK-like MAPK that appears to be identical to the WAPK of soybean leaves, as both kinases are selectively recognized by anti-SIMK antibody (Figures 2 and 4), and their molecular weights and substrate preferences were identical. Fourth, elevation of systemic PA levels in wounded soybean parallels the activation of the 49 kDa WAPK (Figure 1). Also supporting our hypothesis, when PA is infiltrated into Arabidopsis leaves, it activates the 49 kDa AtMPK6 MAPK (our unpublished data).
Our data imply that elevations in PA levels are at least partly responsible for WAPK activation. This causal relationship can also be inferred from observations in other reports. In suspension-cultured alfalfa cells, hyper-osmotic stress activates SIMK (Munnik et al., 1999), and stimulates PLD to raise PA levels with kinetics that are similar or slightly faster than those for SIMK activation (Munnik et al., 2000). Thus, PA may activate SIMK not only in soybean and Arabidopsis, but also in alfalfa.
Lipid-activated protein kinases of plants have been reported previously. They belong to the Ca2+-dependent protein kinase (CDPK) family, and are activated by crude lipids and phospholipids such as lysophosphatidylcholine and lysophosphatidylinositol (Harper et al., 1993; Schaller et al., 1992). Recently a new type of CDPK that is specifically activated by PA has been reported (Farmer and Choi, 1999). However, the PA-activated 49 kDa protein kinase in soybean cells is clearly different from CDPKs: the activity of the PA-activated 49 kDa protein kinase is not dependent on Ca2+, as it remains active in the presence of 2 mm EGTA (Figure 4), and is specifically recognized by an antibody against an alfalfa MAPK. In fact, our data strongly suggest that the 49 kDa protein kinase is an ortholog of the alfalfa SIMK MAPK, because it is recognized by antibodies against SIMK, but not by antibodies against three other alfalfa MAPKs (Figure 4).
PA-activation of the MAPK is dependent upon upstream phosphorylation, as indicated by the effects of general protein kinase and phosphatase inhibitors (Figure 6). These results suggest that PA does not directly activate the soybean MAPK but does so by regulating other upstream protein kinases. In support of this idea, we could not observe PA-activation of the soybean MAPK when PA was added to the renatured protein present in a gel (our unpublished data). Furthermore, in animal cells, PA is known to activate protein kinases that are at the upstream end of MAPK pathways. For example, Raf, a MAPK kinase kinase, has a binding site for PA, and its binding is important for translocation of cytoplasmic Raf-1 to the plasma membrane, which is essential for Raf activation (Rizzo et al., 1999; Rizzo et al., 2000). In addition, some PKC isotypes, which function upstream of MAPK cascades in many signaling pathways in animal and yeast cells (Heinisch et al., 1999; Kolch et al., 1993; Seger and Krebs, 1995; Toda et al., 1996), are also activated by PA (Limatola et al., 1994; Yokozeki et al., 1998).
To the best of our knowledge, this is the first report that characterizes the mechanism by which PA acts as a signal mediator and as an upstream regulator of a MAPK in plant cells. Elevation in PA levels and MAPK activation in response to wounding have been reported separately for many different plant species (Bögre et al., 1997; Lee et al., 1997; Mizoguchi et al., 1996; Ryu and Wang, 1996; Seo et al., 1995; Zhang and Klessig, 1998a), but a causal relationship between these two events has not yet been established. In the light of the present and previous results, PA-mediated activation of wound-induced MAPK pathways may be a general mechanism in plants. Further studies on the role of PA and wound-induced MAPKs should help to understand wound signal transduction in plants better.
Soybean (Glycine max L., cv HwangKum) seeds were planted in vermiculite mixed with humus soil. The plants were grown in a greenhouse at 25 + 5°C with light/dark cycles of 16/8 h. Soybean plants were used at the two-leaf stage. Soybean cell suspensions were grown in Murashige and Skoog medium supplemented with 3% (w/v) sucrose, 10% (w/v) casein enzymatic hydrolysate, 3 mg l−1 2,4 D and 0.1 mg l−1 kinetin. Cells were subcultured every 7 d. Log phase cells were used 2–3 d after subculture. Lipids were purchased from Sigma, St. Louis, MO, USA, and dissolved in chloroform: methanol (1 : 1), except dipalmitoyl PA which was dissolved in chloroform: acetic acid (95 : 5, v/v) and diacylglycerol and linoleic acid which were dissolved in chloroform. Dissolved lipids were dried under nitrogen gas and then sonicated with water.
Plants were wounded with pliers (soybean) by pinching the leaf across the vein at the distal side. The wound area covered about 1/3 of each wounded leaf. For wounding of suspension-cultured soybean cells, the cell suspension was mixed with an equal volume of glass beads (425–600 microns; Sigma) and vortexed for 2 min, 3 times. The wounded cells were mixed with 9 times their volume of the same type of cells and incubated for 5 min before protein preparation.
The roots of soybean seedlings were removed with a sharp razor blade in 1 mm Mes buffer (pH 6.5) containing 0.1 mm KCl and 0.2 mm CaCl2. Seedlings were stabilized in the same buffer for at least 4 h. Before wounding, plants were transferred to vials containing 0.1% n- or sec-butanol in the buffer or the buffer alone, and incubated for 30 min. Five min after wounding, leaves were cut and frozen in liquid nitrogen. For lipid analysis, soybean seedlings were labeled with 555 kBq of 14C-acetate through their cut stem. After isotope uptake, the seedlings were incubated in the buffer described above for 4 h before wounding treatment.
Lipid extraction was performed as previously described (Lee et al., 1997). The lipids were separated on a TLC plate (MerckF254 60, Darmstadt, Germany) with an upper phase of ethylacetate: isooctane: acetic acid: water = 13 : 2 : 3 : 10. The separated lipid bands were identified by comparison with known lipid standards using iodine staining and autoradiograms. The PA amount was analyzed by either phosphate assay of the separated lipid bands or by a phospho-imager.
Preparation of protein extracts
Wounded and control leaves were harvested by rapid freezing in liquid nitrogen. Suspension cells (3 ml) were harvested by vacuum filtration and frozen in liquid nitrogen. Proteins were extracted by homogenizing the samples in extraction buffer (50 mm Hepes pH 7.5, 5 mm EGTA, 2 mm EDTA, 2 mm DTT, 10 mm NaF, 1 mm Na3VO4, 10 mm MgCl2, 50 mmβ-glycerophosphate, 1 µg ml-1 aprotinin, 1 µg ml-1 leupeptin, 1 µg ml-1 pepstatin, 100 µm phenylmethylsulfonyl fluoride). After centrifugation at 15 000 g for 20 min at 4°C, the supernatants (crude extracts) were saved for protein kinase assays. Protein concentrations were determined by the Bradford protein assay (Bio-Rad, Hercules, CA, USA) using BSA as a standard.
In-gel kinase assay
Twenty to 50 µg of total protein were electrophoresed on a 10% SDS-polyacrylamide gel embedded with 0.25 mg ml−1 myelin basic protein (MBP; Sigma, St. Louis, MO, USA) as a kinase substrate. After protein renaturation, the kinase reactions were carried out in the gel with γ-32P-ATP as described previously (Zhang and Klessig, 1997). Relative kinase activities were analyzed by a phospho-imager. Protein kinase sizes were estimated by using prestained molecular mass markers (Bio-Rad).
Polyclonal rabbit antibodies were produced against synthetic peptides encoding the C-terminal amino acids of four alfalfa MAPKs, namely, FNPEYQQ of SIMK (Jonak et al., 1993), VRFNPDPPNIN of MMK2 (Jonak et al., 1995), LNFCKEQILE of MMK3 and LNPEYA of SAMK (Jonak et al., 1996).
Crude extracts (50 µg and 70 µg of cell and leaf extracts, respectively) were separated on a 12% SDS-gel, and the proteins were transferred onto nitrocellulose membranes (Schleicher & Schuell, Postfach, Dassel, Germany). The membrane was blocked with TTBS buffer (25 mm Tris, pH 7.5, 150 mm NaCl, and 0.05% Tween 20) with 5% BSA, and probed with polyclonal antibodies directed against alfalfa MAPKs in TTBS for 2 h at room temperature. Alkaline phosphatase–conjugated goat antirabbit IgG (1 : 2000 dilution; Promega, Madison, WI, USA) was used as a secondary antibody, and the reaction was visualized by hydrolysis of the substrate tetrazolium 5-bromo-4-chloro-3-indolyl phosphate (Promega).
Crude extract (100 µg) was incubated with the previously mentioned antibodies for 2 h at 4°C in immunoprecipitation buffer (25 mm Tris–HCl, pH 7.5, 2 mm DTT, 150 mm NaCl, 1 mm EDTA, 1 mm EGTA, 1 mm Na3VO4, 10 mmβ-glycerophosphate, 2 µg ml-1 aprotinin, 2 µg ml-1 leupeptin, and 1% Triton X-100). Then protein A-agarose (RepliGen, Needham, MA, USA) was added, followed by incubation for another 1 h. The precipitates were washed three times with immunoprecipitation buffer and then re-suspended in the SDS sample buffer. Kinase activity of precipitated proteins was analyzed by the in-gel kinase assay as described previously.
We thank Drs S.-O. Eun and S.B. Ryu for critically reading the manuscript. This work is supported by a grant from the Korea Research Foundation (KRF-2000–015-DP0402) to Y.L., and by grants from the Austrian Science Foundation to H.H.