Activation of phospholipase D and the possible mechanism of activation in wound-induced lipid hydrolysis in castor bean leaves

BiochimicaL et Biouhvsica Acta

ELSEVIER

Biochimica

et Biophysics

Acta 1303 (1996) 243-250

Activation of phospholipase D and the possible mechanism of activation in wound-induced lipid hydrolysis in castor bean leaves Stephen B. Ryu, Xuemin Wang Department of Biochemistq Received

*

Kansas State University, Manhattan, KS 66506, USA

12 December

1995; accepted 7 June 1996

Abstract Hydrolysis of membrane lipids has been suggested to provide messengers mediating defense gene expression in the wound signaling process. It is, however, unknown which lipolytic enzyme is involved in the signaling pathway. This study investigated the temporal and spatial activation of phospholipase D (PLD; EC 3.1.4.4) and the possible activation mechanism in response to wounding in castor bean (Ricinus communis L.) leaves. Wounding triggered a rapid activation of PLD-mediated phospholipid hydrolysis, as indicated by the in vivo increase in phosphatidic acid and free choline, at not only the site of wounding but also the undamaged area of wounded leaves. RNA blotting analysis indicated that PLD gene expression was not involved in the early phase of wounding-activation of PLD. Measurements of PLD by activity assay and immunoblotting suggest that the wounding-activation of PLD at unwounded cells results from intracellular translocation of PLD from cytosol to membranes. A similar translocation pattern of PLD was also obtained as a function of increased free calcium at physiological concentrations in a homogenization buffer. Based on the above results, it is proposed that wounding induces activation of PLD leading to phospholipid hydrolysis, and that the activation results from translocation of PLD to membranes, which is mediated by an increase in cytoplasmic calcium upon wounding. Keywords: Phospholipase

D; Phosphatidylcholine

hydrolysis:

Wounding;

1. Introduction Hydrolysis of membrane lipids has been suggested to couple the signal perception and the synthesis of lipid-derived messengers such as jasmonate in the octadecanoid signaling pathway [l-7]. The enzymatic mechanism for the lipolytic process is, however, unresolved. Several classes of membrane lipid-hydrolyzing enzymes have been reported in plants, which include phospholipase A (PLA), phospholipase C (PLC), phospholipase D (PLD), galactolipase and non-specific acyl hydrolase [8,9]. Among those, PLD and non-specific acyl hydrolases appear to be more active in plant tissues when assayed in vitro. The former hydrolyzes phospholipids at the terminal phosphodiester bond. The latter removes acyl groups from several classes

Abbreviations: DAG, diacylglycerol; g,, , average gravity; PA, phosphatidic acid; PC, phosphatidylcholine: PG. phosphatidylglycerol; PLA, phospholipase A; PLC, phospholipase C; PLD, phospholipase D * Corresponding author. Fax: + 1 913 532 7278; e-mail: [email protected] 00052760/96/$15.00 Published PII SOOOS-2760(96)00096-3

by Elsevier Science B.V.

Signal transduction;

(Ricinus communis)

of lipids including galactolipids, phospholipids, sulfolipids and mono- and diacylglycerol 181. PLD and acyl hydrolase have been suggested to act in sequence to release free fatty acids [lo- 121. The activity of PLD forms phosphatidic acid (PA), which is dephosphorylated by PA phosphatase, and the resulting diacylglycerol (DAG) is deacylated by acyl hydrolase. The presence of the phospholipid degradation pathway in plants was originally proposed by Kates [ 131 about four decades ago. Later studies have linked the concerted lipolytic reactions releasing polyunsaturated fatty acids to membrane deterioration in senescence, ageing and stress injuries. It has been proposed that the liberated polyunsaturated fatty acids serve as substrates for lipoxygenase that produces activated oxygen and lipid peroxides leading to membrane damage [ 10,l 1,141.In view of the recent findings on the role of lipid metabolites in the signaling pathways, the PLD-initiated lipolytic process may also be involved in generating lipid messengers mediating plant defense responses. A regulatory role of PLD and its product PA has been proposed in transmembrane signaling, respiratory burst, membrane trafficking and protein secretion [151.

244

S.B. Ryu, X. Wang / Biochimica et Biophysics Acta 1303 (1996) 243-250

PLD-catalyzed hydrolysis is suggested to be an integral part of the receptor-linked hydrolysis of phospholipids involving a network of PLA, and PLC in stimulated animal cells [16]. The PLD-initiated lipolytic process suggested for plant membrane deterioration has an obvious analogy to the biochemical process in which PLD working together with other lipolytic enzymes mediates cellular responses in animal systems. However, the signaling role of PLD in plants has been unexplored. In a model depicting the production of free polyunsaturated fatty acids in the octadecanoid signaling pathway, it has been postulated that a lipolytic process is activated through the interaction of defense signals such as systemin and oligouronides with membrane receptors during wounding responses [4,5]. The activation of lipid hydrolysis was proposed based on indirect evidence such as wounding-induced systemic effect on protoplast fragility and the parallel time courses between loss of membrane integrity and systemic signal production [ 17,181. There has been, however, no direct evidence for the wounding-activation of a specific lipid-hydrolyzing enzyme. This study was undertaken to determine the spatial and temporal activation of PLD and the mechanism of PLD activation in response to wounding in castor bean leaves.

activation of lipolytic activities during extraction. Lipids were extracted according to the procedure of Bligh and Dyer [20] with following modifications. After 1 ml chloroform and 0.8 ml water were added, the extract was incubated at room temperature for 1 h with agitation. The leaf disks were re-extracted two more times with 2 ml chloroform/methanol (2:l; v/v> for 30 min. The combined extracts were washed with 1 ml of 1 M KCl. For experiments in which free choline and phosphocholine were assayed, 1 ml water instead of 1 M KC1 was added to separate chloroform and aqueous phase. The chloroform phase was then washed with 2 ml of 0.5 M KCl. Butylated hydroxytoluene was added to the solvent (O.Ol%, w/v) to reduce lipid oxidation. Phospholipids were separated by TLC (Silica Gel 60; Merck) with the solvent of chloroform/methanol/acetic acid/water (85:15:12.5:3.5, v/v) [21]. Individual lipids were made visible with iodine vapor, identified by co-chromatography with authentic standards and verified by the use of specific spray reagents [22]. Various phospholipids were separated under this system, and the relative mobilities (R,) were: PA 0.52, phosphatidylethanolamine 0.39, phosphatidylglycerol (PG) 0.33, phosphatidylcholine (PC) 0.21, phosphatidylserine 0.17 and phosphatidylinositol 0.1. Phospholipid spots were scraped into test tubes and quantitated by calorimetric determination of phosphorus content [23].

2. Materials and methods

2.3. Assays for choline and phosphocholine

2.1. Plant materials and sample treatment Coatless castor bean (Ricinus communis L. var. Hale) seeds were germinated in the dark in moist vermiculite for 3 days. The seedlings were individually transplanted into plastic pots containing a mixture of vermiculite and perlite (1: 1, v/v) subirrigated with Hoagland nutrient solution [ 191. Plants were grown under cool white fluorescent lights at 23 f 3°C with a 14-h photoperiod. Wounding of a plant was initiated by excising with a cork borer two leaf discs (1.5 cm in diameter), which represented about 2% of the total leaf area, from a fully expanded leaf of an approximately 2-month-old plant. Meanwhile, these discs were used as prior-to-wounding control. Two leaf discs per interval were subsequently sampled from the leaf at the intervals of 5, 30 and 60 min after initial wounding. The sites of sampling were approximately 10 cm apart from the initial wound sites and among subsequential sampling sites, which were referred to as distal site throughout the text. To examine the wounding effect on PLD at the site of wounding, a physiologically similar leaf as in distal site was wounded multiple times with pliers. The wounded areas of the leaf were sampled with a cork borer at the time intervals of 0.5, 1, 2 and 5 min after initial wounding. 2.2. Lipid extraction and phospholipid

analysis

The leaf discs were immersed in 2 ml hot isopropanol (75°C) for 15 min immediately after sampling to avoid

The aqueous phase obtained from lipid extraction was dried in a Speed Vat. Choline and phosphocholine were measured according to an enzymic phosphorylation procedure [24]. Briefly, free choline in aqueous leaf extracts was converted to [32P]phosphocholine in the presence of [y32P]ATP and choline kinase (Sigma). Unincorporated phosphoi3* PIATP was removed from 32P-containing choline by Dowex l-X8 anion exchange column chromatography. Newly synthesized phosphocholine was quantitated by scintillation counting of [ 32Plphosphocholine. For assaying phosphocholine, equal volumes of the aqueous leaf extracts were first treated with alkaline phosphatase to convert phosphocholine into free choline, which was then measured as described above. The amount of phosphocholine in each sample is calculated as the difference between the levels of choline measured before and after alkaline phosphatase treatment. 2.4. Tissue fractionation

and PLD activity assay

For determination of PLD activity, the leaf discs were sampled in the same manner as those for lipid analysis and ground with a mortar and pestle at 4°C in an extraction buffer containing 50 mM Tris-HCl (pH 8.0), 10 mM KCl, 1 mM EDTA, 0.5 mM PMSF and 2 mM DTT. The homogenate was centrifuged at 6000 X g for 10 min. The supematant was centrifuged at 100000 X g, for 30 min.

S.B. Ryu, X. Wang/ Biochimica et Biophysics Acta 1303 (1996) 243-250

The resultant supematant (soluble fraction) and pellet (membrane fraction) were suspended in the same homogenization buffer. The microsomal fractions were centrifuged again to remove contaminating soluble proteins. The washed pellets were resuspended in the same buffer followed by sonication. Protein content was determined with a Coomassie blue dye-binding assay according to manufacturer’s instructions (Bio-Rad). The effect of Ca*+ on the intracellular distribution of PLD was determined as described previously [25] with some modifications. Leaf tissues were homogenized with a mortar and pestle on ice in the extraction buffer described above. Different concentrations of Ca*+, 0, 10, 50 and 200 PM, were added to the aliquots of homogenate and free Ca*+ concentration in the homogenates was measured with a Ca*+ -sensitive electrode. Sub-micromolar Ca*+ levels were estimated by interpolation between adjacent points and a standard Ca*+ curve was made based on the electrode measurements of known concentrations of CaCl,. After 30-min incubation on ice, the homogenates were then fractionated into soluble and membrane fractions as described above. PLD activity was determined by measuring the formation of both PA and phosphatidylethanol. The transphosphatidylation reaction of PLD serves as a better indicator of PLD activity than formation of PA since phosphatidylethanol is not readily metabolized further. A standard enzyme assay mixture contained 100 mM MES/NaOH (pH 6.5), 25 mM CaCl,, 0.5 mM SDS, 0.4 pmol of PC (egg yolk; Sigma), 1% ethanol and 20 ~1 enzyme solution in a total volume of 200 ,ul. The assay conditions, reaction products separation and quantitation were detailed elsewhere [19,26]. Briefly, the reaction products were spotted onto a TLC plate (silica gel G) which was developed with chloroform/methanol/NH,OH (65:35:5). Lipids on plates were visualized by exposure to iodine vapor, and their R, values were compared with known lipid standards. The R, values of PA, PC and phosphatidylethanol were 0.03, 0.28 and 0.71, respectively. Phosphatidylethanol was scraped and quantitated by measuring phosphorus content as described 1231. 2.5. Electrophoresis, analysis

immunoblotting

and RNA

with a video densitometer (Bio-Rad). Castor bean RNA was isolated using a cetyltrimethylammonium bromide extraction method as described previously [28]. Total RNA was subjected to denaturing formaldehyde/agarose gel electrophoresis and transferred onto a nylon membrane. After transfer, the RNA was fixed on the filters by crosslinking by UV illumination. Northern blots were prehybridized in a solution of 6 X SSC, 0.5% SDS, 5 X Denhardt’s solution and 100 pg/ml salmon sperm DNA at 65°C. The probe was the 2834-bp EcoRIKpnI fragment of a PLD cDNA [29], which was random primer labelled with [ w3* P]dCTP. Hybridization was performed in the same solution at 65°C overnight. The blots were washed with 2 X SSC and 0.1% SDS followed by 0.1 X SSC and 0.5% SDS at 65°C and exposed to X-ray film.

3. Results 3.1. Increases

in PA and choline at wound and distal sites

The extent of increased PA and choline was smaller at the distal site than in the site of wounding (Fig. 1). The largest increases in PA and free choline were 0.14 and 0.16 pmol/g dry wt., respectively, at the distal site (Fig. 1A and C). Choline level was highest 30 min after wounding whereas the increase of PA peaked 5 min after initial wounding. The temporal difference in the increases between PA and choline indicate that the initial rise of PA resulted from the PLD hydrolysis of other phospholipids in addition to PC. Measurements of phospholipids revealed that the cellular levels of PC and PG decreased (Fig. 2A and C), whereas the amounts of PE and PI remained unchanged (data not shown) in the first 5 min after wound-

0.7

B. wound site

0.6

0.2 0.6

blotting

Proteins from leaf disc extracts were mixed with SDS gel loading buffer and heated at 95°C for 3 min, and then resolved on 8% slab SDS-polyacrylamide minigels (3.5% stacking, 0.75 mm thick) with a constant voltage of 120 V. Subsequent procedures for transferring the proteins onto polyvinylidine difluoride membranes and immunoblotting with anti-PLD antibodies were the same as previously reported [27]. Polyclonal PLD antibodies were raised in rabbits against purified PLD from 2-day germinating endosperm [27]. PLD protein amounts on immunoblots were estimated by densitometric scanning of the band intensity

245

2

0.20

10 20304050

Time (min)

60 70

0123456

Time (min)

Fig. 1. Wound-induced increase in PA and choline at wound and distal sites in wounded castor bean leaves. Distal sites (left panel) refer to the leaf samples collected 10 cm distance away from the initial wound sites that were inflicted by excision of two leaf discs (1.5 cm in diameter). Wound sites (right panel) refer to the leaf discs collected from the areas directly wounded with pliers. Note the difference in sampling time intervals between the distal and wound sites. Values are means + S.E. of three experiments

S.B. Ryu, X. Wang / Biochimica et Biophysics Acta 1303 (1996) 243-250

246

8.5 8.0 7.5

z a Kc.

m

7.0

7.0

4.0

1.0

D.

C. 3.5

3.5

g

3.0

3.0

g n

2s

2.5 i-2.0’ 0

n ” n I IO 20 30 40 Time @in)

( 50

Z.Ot---J n ” 60 70 o 1

2 Tiie

3 4 (min)

5

crease in free choline content was larger than that of PA. At the first minute after wounding the increase of choline was 0.40 ,umol/g dry wt., whereas that of PA was 0.29 pmol/g dry wt. The lower increase in PA than choline may result from the metabolic instability of PA as shown in distal wound response. To determine whether PC-PLC was activated by wounding, the content of phosphocholine was analyzed in parallel with choline assay. No change in phosphocholine content was detected at the site of wounding and distal site, ruling out the possibility of wounding-induced activity of PC-PLC at the time period (data not shown).

6

Fig. 2. Wound-induced changes in PC and PG at distal (left panel) and wound (right panel) sites in wounded castor bean leaves. PC and PG were separated by TLC and quantitated by calorimetric determination of phosphorus content. Values are means + S.E. of three experiments.

ing. The PC decrease was transient; the prior-to-wounding level of PC was almost recovered 60 min after wounding. The time course of increased choline matched that of the decrease in PC (Fig. 2A), but PA was not further increased after the first 5 min. This discrepancy between two PLD products may reflect the metabolic instability of PA because it can be further hydrolyzed by PA phosphatase and acyl hydrolases [ 101. Experiments were also conducted to examine whether the above wound-induced distal changes were dependent on the distance from the wounding site and the intensity of wounding. The distance effect was examined by excising two leaf discs followed by measuring the changes in PA and choline content at different locations on the same leaf. No significant difference was observed at the time examined (data not shown), suggesting that the initial signaling process that triggered PLD activation might be completed in a wound leaf in 5 min. The dependence of wounding intensity was tested by comparing the PA and choline changes between the gentle excision of two leaf discs (less than 5% of total leaf area) and injury inflicted by pliers in more than 30% of total leaf area. The increase in PA and choline in distal sites was similar in both wounding methods. At the site of wounding, rapid increases in PA and choline were evident within 30 s after initial wounding, indicating a rapid rise of PLD-mediated hydrolysis (Fig. 1B and D). Significant losses of PC and PG occurred at the first 2 min after wounding (Fig. 2B and D). The PC and PG content 5 min after wounding were about 12 and 20%, respectively, lower than that prior-to-wounding. In the 5-min post-wounding period examined, PA content was highest at the first minute after wounding (0.58 ,umol/g dry wt), which was 2-fold higher than that of the prior-towounding PA concentration (0.29 pmol/g dry wt). While the overall pattern for PA change in this period was similar to that of free choline change, the wounding-induced in-

3.2. Wound-induced ciated PLD

transient increase in membrane-asso-

To elucidate the mechanism(s) for the wounding activation of PLD in the distal, intact cells, the distribution of PLD in soluble and membrane fractions was measured at various time intervals after wounding at the distal areas of wounded leaves. Approximately 20% increase in membrane-associated specific PLD activity was observed 5 min after wounding (Fig. 3). Membrane-associated PLD activity increased to a plateau (30%) 30 min after wounding and declined afterwards. Changes in soluble PLD showed an inverse pattern compared to membrane-associated PLD. The lowest activity (20% decrease) occurred at 30 min after wounding. The soluble PLD almost returned to the prior-to-wounding level 60 min after wounding. The amounts of PLD protein in the membrane and soluble fractions were analyzed by immunoblotting using anti-PLD antibodies (Fig. 4). At 30 min after wounding, membrane associated PLD increased by 20%, whereas soluble PLD decreased by 20%. The pattern of the wounding-induced change in PLD protein concentrations was consistent with that of the PLD activity change in the two subcellular fractions (Fig. 3), suggesting that the change of PLD activity resulted from a difference in protein content

16 14 12

.‘~‘~‘~m~‘~‘A

1”

0

10

20

30

Time

40

50

6n

70

(min)

Fig. 3. Wound-induced change in membrane-associated and soluble PLD activity at the distal site of wounded castor bean leaves, Membrane PLD was from the pellet of 100000X g,, centrifugation of the 6000x g supernatant, and soluble PLD was from the supematant after 100000 X g,, centrifugation. Values are means f S.E. of three experiments.

S.B. Ryu, X. Wang/Biochimica

0

5

30

et Biophysics Acta 1303 (1996) 243-250

247

60 Min

Mic PL&

0

’ 10

’ 20

I

. I

. I

30

40

50

. 60

Time (min) Fig. 4. Immunoblotting analysis of membrane-associated and soluble PLD at the distal site of wounded castor bean leave. (Left panel) Immunoblot of membrane (Mic; 10 pg/lane) and soluble (Sol; 30 pg/lane) PLD. Proteins were resolved by 8% SDS-PAGE and PLD on the blot was made visible using alkaline phosphatase conjugated to goat antibodies against rabbit immunoglobulin. (Right panel) Densitometric scanning of the immunoblot showing the changes in membrane-associated and soluble PLD protein amount after wounding. The band intensity of microsomal and soluble PLD was calculated based on the same amount of protein loaded to the gel.

in the membrane and soluble fractions. At the site of wounding, there was no significant change in membraneassociated PLD during the first 5 min after wounding (data not shown). To examine how PLD was associated with membranes, membrane pellets were suspended with 0.2 M KC1 at 4°C for 30 min followed by centrifugation at 100000 X g,, to remove some peripheral proteins from membranes. PLD in the salt-washed membranes was assayed for activity and immunoblotted (Fig. 5). The salt washing removed about 55% PLD from the prior-to-wounding membranes, but only about 35% from the membranes of 30 and 60 min after wounding. The immunoblotting analysis also showed the same result; a higher amount of PLD remained associated with membranes in the 30 and 60 min post-wounding than the prior-to-wounding control after salt washing (Fig. 5). These results indicated that almost one-half of membrane-bound PLD in castor bean leaves was loosely associated with membranes, and that most the membrane-associated PLD that increased in response to wounding was not removed by 0.2 M KC1 washing of membranes. There appeared to be a discrepancy in 60 min wound response between Figs. 4 and 5 on the increased PLD

association with membranes, Fig. 4 shows the change in PLD protein levels based on the equal amounts of total protein (PLD/mg total membrane-associated or soluble protein) from the various treatments, whereas Fig. 5 reflects the total PLD activity per leaf disc. Thus, the difference in the time courses could be due to an increase in membrane-associated protein content in the leaves 60 min after wounding. Meanwhile, leaf-to-leaf variations might also cause part of this discrepancy. 3.3. Association

of PLD with membranes increased free calcium

of

The effect of Ca2+ on the partitioning of PLD between membrane and soluble fractions was tested in order to gain a clue as to the factor that promotes PLD association with membranes. The amount of PLD associated with membranes was influenced by the inclusion of Ca*+ in the homogenate at physiological concentrations (Fig. 6). Ca*+ from 0.2 to 5 PM increased membrane-associated PLD with a concomitant decrease in soluble PLD. Free Ca*+ at 1 PM caused an increase in membrane-associated activity by 70 nmol/min per g fresh wt. and a decrease in soluble

Diiiated

5

as a function

Associated

0 10 30 60 0 10 30 60 Min

4

3 2 1

0

afterwash

with

20 Time

40

KC1 60

(min)

Fig. 5. Removal of membrane-associated PLD after washing with 0.2 M KCI. (Left panel) Total membrane-associated PLD activity before and after KC1 wash. (Right panel) Immunoblot of PLD dissociated from and associated with membranes after KC1 wash. An equal volume of the 100000 x g,,-supematant (Dissociated) and pellet (Associated) after salt wash was loaded to each lane, and proteins were resolved by 8% SDS-PAGE. Results were representative of two separate experiments.

S.B. Ryu, X. Wang/Biochimica

248

et Biophysics Acta 1303 (1996) 243-250

3.4. PLD gene expression

Free calcium (PM)

in wounding

To test if PLD gene expression plays a role in the early phase of wounding-induced changes in PLD, the total RNA isolated from the wounded leaves was probed with a 2.8-kb PLD cDNA cloned recently from castor bean [29]. PLD mRNA level remained unchanged at the distal site throughout the first 2 h after wounding, indicating that the rapid increase in PLD activity does not involve activation of PLD gene expression (Fig. 7). At the site of wounding, a gradual decline of PLD mRNA was observed starting 5 min after wounding.

4. Discussion Free

calcium

(KM)

Fig. 6. Distribution of PLD between soluble and membrane fractions as a function of free Ca2+ concentrations in homogenization medium. Various concentrations of CaCI, were added to the homogenate extracted in an EDTA-containing buffer, and free Ca2+ in the extracts was measured with a Ca’+-sensitive electrode. The effect of Ca” on PLD distribution between soluble and membrane fractions is expressed according to specific (upper panels) and total (lower panel) PLD activities.

activity by 60 nmol/min per g fresh wt. Immunoblotting analysis of PLD showed that the difference in PLD activity was due to the change in PLD protein content (data not shown). At 1 PM free Ca*+, the specific activity of membrane-associated PLD increased from 67 to 79 nmol/min per mg protein and that of soluble PLD decreased from 7.7 to 4.3 nmol/min per mg protein. The inverse changes in the specific activity between the membrane and soluble PLD indicated that the Ca*+-promoted increase in membrane-association of PLD was specific, rather than a general increase in total membrane-associated proteins by Ca2+. The effect of Mg*+ on the PLD association with membranes was also examined, and no inverse changes between membrane-associated and soluble PLD were observed. This result indicated that the Ca2+-promoted intracellular translocation of PLD did not result from a general cation effect.

0

Wound site

Distal site

5 30

5 30 120 Min

120

Fig. 7. Northern blot of PLD transcript at the wound and distal sites of wounded castor bean leaves, 5, 30 and 120 min after initial wounding. Total RNA (20 pg/lane) was electrophoresed on an 1% agarose gel, transferred to a nylon membrane and probed with a 2.8-kb PLD cDNA. Results were representative of three separate experiments.

The present study shows a rapid increase in PA and choline in castor bean leaves after wounding, suggesting a rapid activation of PLD-mediated hydrolysis of membrane lipids. Furthermore, this study reveals that the woundingstimulation of PLD is not limited to the direct sites of wounding and also occurs in the distal, intact cells. The PLD-mediated hydrolysis at the site of wounding and distal site may involve two distinct categories of biochemical events. The lipid hydrolysis at the site of wounding is most likely to result from the release/deregulation of PLD since wounding disrupts cell membranes, leading to the release of PLD from its original stores. A recent immunological study has shown that PLD is localized in vacuoles, cytoplasm, ER and plasma membranes in castor bean leaves [30]. This process, which is analogous to the rapid lipolysis observed during tissue homogenization, should be less regulated and unique to the process involving physical damage of membranes. The category of events at the distal, unwounded cells, however, should be distinct from that at the site of wounding because there is no physical decompartmentalization of PLD. The transient increases in PA and choline suggest that the wounding-induced lipid turnover in intact cells is highly regulated. Measurements of PLD activity and protein in subcellular fractions suggest that one mechanism that activates PLD in response to wounding may involve intracellular translocation of PLD from cytosol to membranes. In the intact, distal cells, wounding stimulated a transient increase in membrane-associated PLD, accompanied by the concomitant decrease in soluble PLD. The time course for the wound-stimulated increase in membranous PLD correlated with that of decreased phospholipids. The increase in membrane-associated PLD has been suggested to result in lipid hydrolysis in other studies. For example, irradiation of cauliflower florets caused an immediate rise in membrane-associated PLD activity, which was attributed to accelerated membrane deterioration [12]. The increase in membrane-associated PLD was also correlated with the rate of phospholipid degradation and leaf senescence in castor bean leaves [26]. Therefore, it is suggested that the increased PLD hydrolysis stimulated by wounding results

S.B. Ryu, X. Wang/Biochimica

et Biophysics Acta 1303 (1996) 243-250

from increased PLD association with membranes. Membrane association may increase access of PLD to its substrates and thereby stimulate lipid hydrolysis. It is conceivable that the intracellular translocation constitutes an early step in rapid PLD activation in response to wounding. The reversible association with membranes has been a mechanism of activation proposed for various phospholipid-related enzymes in animals, including cytidylyltransferase, intracellular PLA 2, PA phosphohydrolase and protein kinase C. Recent intracellular localization study showed that more than half of the cellular PLD in fully expanded castor bean leaves is present in the cytoplasmic phase and plasma membranes [30]. Thus, endoplasmic reticulum and plasma membrane may be the likely sites for the increased membrane association of PLD [30]. On the other hand, a substantial amount of PLD was already associated with membranes in unwounded leaves. More than 50% of the pre-existing membrane associated PLD was removed from membranes after salt washing, whereas most of the PLD that increased in response to wounding was not removed by the washing. The structural and functional relationships remain to be established between the pre-existing membrane-associated PLD and the stimulus-responsive, freshly translocated PLD. The observation that association of PLD to membranes in vitro was promoted by physiological concentrations of Ca*+ suggests that the intracellular translocation of PLD may be mediated by Ca *+ . The transduction of many hormonal and environmental signals involves Ca*+ and mechanical perturbations immediately increase cytosolic Ca2+ [31]. Thus, the increased PLD association with membranes in response to wounding may be triggered by an influx of Ca2+ to cytoplasm. Ca2’ has been shown to modulate plant PLD activity in vitro and in vivo 110,321. Recent cloning of plant PLD has revealed a consensus sequence for a calcium-binding domain present in the enzyme from castor bean, rice and maize [29,33]. A study using PLD from Streptumyces indicated that PLD binds to membrane vesicles in a Ca’+-dependent manner. It was suggested that the binding of Ca2+ to PLD changes its conformation, promoting the association of PLD with membranes [34]. However, the mechanism for Ca*+ action on plant PLD remains to be elucidated. An important question that arises from this study is the physiological significance of the wound activation of PLD. PA is a potent regulator of various signaling processes in animal cells [ 15,161. While a direct role of PA in plant signal transduction remains to be studied, several routes for PA metabolism, which are known to occur in plants, may be involved in the wounding response. For example, PA is a central intermediate in glycerolipid biosynthesis, so it is possible that the PA generated by PLD may be used for the synthesis and remodeling of membrane lipid composition in wounding response. It has been observed that the wound-induced decrease in phospholipid content is transient in the distal cells (Fig. 2A and Cl, indicating the

249

presence of wound-stimulated synthesis of phospholipids primed by degradation, On the other hand, PA can be dephosphorylated to DAG or deacylated. It has been proposed that DAG derived form PA is deacylated by nonspecific acyl hydrolase to release free fatty acids, and that the free polyunsaturated fatty acids are peroxidized by lipoxygenase [ 10,ll I. The liberated polyunsaturated fatty acids may serve as substrates for the synthesis of oxylipins signaling plant wounding response. The present data also indicated that leaves of different physiological stages show varied degree of lipid turnover in response to wound treatment. In this study, before- and after-wounding samples were collected from the same leaf to provide the same physiological background. The standard errors in Figs. l-3 reflected leaf-to-leaf variations. Such variations were not surprising in light of recent findings that the intracellular distribution of PLD and its levels of activity and protein change during leaf development 119,301. Some discrepancies in the time courses between the PLD activity (Fig. 3) and protein level (Fig. 4 and Fig. 5) could result from the sampling heterogeneity and/or the method of detection. For example, the presence of PLD inhibitor or activator in the extract would lead to an under- or over-estimation of PLD activity, but have little effect on the immunoblotting measurement. In summary, results from this study suggest that a rapid activation of PLD is involved in the wound-induced lipid hydrolysis. The events for the activation of PLD in wounded castor bean leaves has been proposed to occur via following sequence. Wounding stimulates an increase in cytoplasmic Ca*+ , this increase promotes translocation of cytosolic PLD to bind microsomal membranes, and such translocation increases PLD-mediated hydrolysis of phospholipids. The controlled induction of PLD-mediated hydrolysis may point to its significance in generating lipid messengers in plant stress response.

Acknowledgements The authors are grateful to Dr. L. Davis for his valuable suggestions throughout this study and to Drs. L. Davis, M. Kanost and Mr. J. Dyer for their critical reading of the manuscript. This research was supported by National Science Foundation Grant IBN-9511623, and this is contribution 95-207-J of the Kansas Agricultural Experiment Station.

References [ll Bell, E., Creelman, R.A. and Mullet, J.E. (1995) Proc. Natl. Acad. Sci. USA 92, 8675-8679. S., Brodschelm, W., Holder, S., Kammerer, L., Kutchan. T.M., Mueller, M.J., Xia, Z.-Q. and Zenk, M.H. (1995) Proc. Natl. Acad. Sci. USA 92, 4099-4105. [31Constabel, C.P., Bergey, D.R. and Ryan, CA. (1995) Proc. Nat]. Acad. Sci. USA 92, 407-411.

I21Blechert,

250

S.B. Ryu, X. Wang / Biochimica et Biophysics Acta 1303 (1996) 243-250

[4] Doares, S.H., Syrovets, T., Weiler, E.W. and Ryan, CA. (1995) Proc. Natl. Acad. Sci. USA 92, 4095-4098. [5] Farmer, E.E. and Ryan, C.A. (1992) Plant Cell 4, 129-134. [6] Mueller, M.J., Brodschelm, W., Spannagl, E. and Zenk, M.H. (1993) Proc. Natl. Acad. Sci. USA 90, 7490-7494. [7] Pena-Cortes, H., Fisahn, J. and Willmitzer, L. (1995) Proc. Natl. Acad. Sci. USA 92, 4106-4113. [s] Huang, A.H.C. (1993) in Lipid Metabolism in Plants (Moore, T.S., ed.), pp. 473-503, CRC Press, Boca Raton, FL. [9] Wang, X. (1993) in Lipid Metabolism in Plants (Moore, T.S., ed.), pp. 505-525, CRC Press, Boca Raton, FL. [lo] Paliyath, G. and Droillard, M.J. (1992) Plant Physiol. Biochem. 30, 789-812. [Ill Thompson, J.E., Brown, J.H., Paliyath, G., Todd, J.F. and Yao, K. (1991) in Active Oxygen/Oxidative Stress and Plant Metabolism (Pell, E.J. and Steffen, K.L., eds.), pp. 57-66, The American Society of Plant Physiologists, Rockville, MD. [12] Voisine, R., Vezina, L.-P. and Willemot, C. (1993) Plant Physiol. 102, 213-218. [13] Kates, M. (1954) Can. J. Biochem. Physiol. 32, 571-583. [14] Salama, A.M. and Pearce, R.S. (1993) J. Exp. Botany 44,1253-1265. [15] Exton, J.H. (1994) Biochim. Biophys. Acta 1212, 26-42. [16] Liscovitch, M. (1992) Trends B&hem. Sci. 17, 393-399. [17] Walker-Simmons, M., Hollaender-Czytko, H., Anderson, J.K. and Ryan, CA. (1984) Proc. Natl. Acad. Sci. USA 81, 3737-3741. 1181 Nelson, C.E., Walker-Simmons, M., Makus, D., Zuroske, G., Graham, J. and Ryan, C.A. (1983) in Plant Resistance to Insects, American Chemical Society Symposium Series 208 (Hedin, P.A., ed.), pp. 103-122, American Chemical Society, Washington, DC. [19] Dyer, J.H., Ryu, S.B. and Wang, X. (1994) Plant Physiol. 105, 715-724.

[20] Bligh, E.G. and Dyer, W.J. (1959) Can. J. Biochem. Physiol. 37, 911-917. [21] Whitaker, B.D. (1991) J. Am. Sot. Hort. Sci. 116, 528-533. [22] Vioque, E. (1984) in CRC Handbook of Chromatography, Lipids (Mangold, H.K., ed.), Vol. II, pp. 309-317, CRC Press, Boca Raton, FL. [23] Rouser, G., Fleisher, S. and Yamamoto, A. (1970) Lipids 5.494-496. [24] Bonser, R.W. and Thompson, N.T. (1992) in Signal Transduction, A Practical Approach (Milligan, G., ed.), pp. 123-151, IRL Press, New York, NY. [25] Channon, J.Y. and Leslie, C.C. (1990) J. Biol. Chem. 265, 54095413. [26] Ryu, S.B. and Wang, X. (1995) Plant Physiol. 108, 713-719. [271 Wang, X., Dyer, J.H. and Zheng, L. (1993) Arch. Biochem. Biophys. 306,486-494. [28] Wang, X. and Hildebrand, D.F. (1988) Physiol. Plant. 72, 127-132. 1291 Wang, X., Xu, L. and Zheng, L. (1994) J. Biol. Chem. 269, 20312-20317. [30] Xu, L., Paulsen, A.Q., Ryu S.B. and Wang, X. (1996) Plant Physiol. lll, lOl-107. [31] Knight, M.R., Campbell, A.K., Smith, S.M. and Trewavas, A.J. (1991) Nature 352, 524-526. 1321 Cheour, F., And, J., Makhlouf, J. and Willemot, C. (1992) Plant Physiol 100, 1656-1660. [33] Ueki, J., Morioka, S., Komari, T. and Kumashiro, T. (1995) Plant Cell Physiol. 36, 903-914. [34] Yamamoto, J., Konto, A., Handa, T. and Miyajima, K. (1995) Biochim. Biophys. Acta 1233, 21-26.