Identification of diacylglycerol pyrophosphate as a novel metabolic product of phosphatidic acid during G-protein activation in plants.

We provide evidence that phosphatidic acid (PtdOH) formed during signaling in plants is metabolized by a novel pathway. In much of this study, 32Pi-labeled Chlamydomonas cells were used, and signaling was activated by adding the G-protein activator mastoparan. Within seconds of activation, large amounts of [32P]PtdOH were formed, with peak production at about 4 min, when the level was 5-25-fold higher than the control. As the level of [32P]PtdOH subsequently decreased, an unknown phospholipid (PLX) increased in radiolabeling; before activation it was barely detectable. The chromatographic properties of PLX resembled those of lyso-PtdOH and CMP.PtdOH but on close inspection were found to be different. PLX was shown to be diacylglycerol pyrophosphate (DGPP), the product of a newly discovered enzyme, phosphatidate kinase, whose in vitro activity was described recently (Wissing, J. B., and Behrbohm, H. (1993) Plant Physiol. 102, 1243-1249). The identity of DGPP was established by co-chromatrography with a standard and by degradation analysis as follows: [32P]DGPP was deacylated, and the product (glycerolpyrophosphate, GroPP) was hydrolyzed by mild acid treatment or pyrophosphatase to produce GroP and Pi as the only radioactive products. Since DGPP is the pyrophosphate derivative of PtdOH and is formed as the concentration of PtdOH decreases, we assumed that PtdOH was converted in vivo to DGPP. This was confirmed by showing that during a short labeling protocol while the specific radioactivity of DGPP was increasing, the specific radioactivity of the 32Pi derived from DGPP as above was higher than that of [32P]GroP. DGPP was also formed in suspension cultures of tomato and potato cells, and its synthesis was activated by mastoparan. Moreover, it was also found in intact tissues of a number of higher plants, for example, carnation flower petals, vetch roots, leaves of fig-leaved goosefoot, and common persicaria and microspores of rape seed. Our results suggest that DGPP is a common but minor plant lipid that increases in concentration when signaling is activated. Possible functions of DGPP in phospholpase C and D signaling cascades are discussed.

While the speed and magnitude of second messenger formation are important, cells must also down-regulate their effective concentrations to ensure that the response level is tightly coupled to the stimulation level. In general, PtdOH is thought to be converted by PtdOH phosphatase to DAG or by PLA 2 to lyso-PtdOH or to be converted to CMP⅐PtdOH for the resynthesis of certain phospholipids (Kent, 1995). Recently, a new mechanism of PtdOH attenuation was discovered (Van Blitterswijk and Hilkmann, 1993) in which PLD transfers the phosphatidyl moiety of phosphatidylcholine to endogenous DAG, resulting in the formation of the novel phospholipid bisphosphatidic acid (BisPtdOH). This uncommon lipid is therefore the condensation product of PLC and PLD signaling pathways and attenuates the formation of both DAG and PtdOH.
In plants, evidence for lipid-derived second messenger pathways is increasing, as reflected in the number of recent reviews (Drøbak, 1992(Drøbak, , 1993Crain, 1993, 1994). One of the best studied systems is the green alga Chlamydomonas that was first used to show that G-protein activators such as mastoparan (Law and Northrop, 1994;Ross and Higashijima, 1994) and EtOH (Hoek et al., 1992) rapidly activate the formation of PtdOH (Musgrave et al., 1992;Quarmby et al., 1992). The increase was originally explained as the combined activities of PLC and DAG kinase. Similar effects of mastoparan on PLC activation have since been presented for the higher plant cells soybean and carrot (Legendre et al., 1993;Drøbak and Watkins, 1994;Cho et al., 1995). However, Munnik et al. (1995) have made it clear for both Chlamydomonas and carnation that part of the PtdOH produced on G-protein activation arises from PLD activity. Thus in plants as in animals, G-protein activation triggers two signaling pathways that dramatically raise intracellular PtdOH levels, emphasizing the potential significance of this molecule. How the PtdOH levels are down-regulated in plants is unknown.
Recently, a new enzyme activity was extracted from many different higher plants that was able to phosphorylate PtdOH to diacylglycerol pyrophosphate (DGPP) and was called phosphatidate kinase Behrbohm, 1993a, 1993b;Wissing et al., 1994). As yet, all demonstrations of activity were in vitro. Here we show that when Chlamydomonas cells were treated with mastoparan, the initial rise in PtdOH formation was counteracted by its conversion to DGPP, illustrating that this novel lipid kinase is highly active in vivo when intracellular signaling is stimulated. DGPP was further shown to be present in several higher plants and their different tissues, and its synthesis was shown to be stimulated when cells were activated by mastoparan.

EXPERIMENTAL PROCEDURES
Materials-Silica 60 TLC plates and reagents for lipid extraction and analyses were from Merck. Mastoparan (Vespula lewisii), glycerophosphate (GroP), CMP⅐PtdOH, PLA 2 (Apis mellifera), and inorganic pyrophosphatase (Escherichia coli and Saccharomyces cerevisiae) were purchased from Sigma. PEI cellulose TLC polygram sheets were from Machery-Nagel (Dü ren, Germany). [ 32 P]Orthophosphate (carrier-free) was from Amersham International ('s-Hertogenbosch, The Netherlands), and [5-3 H]cytidine (27.8 Ci/mmol) was obtained from DuPont NEN. BisPtdOH (tetrapalmitoyl) was purchased from Serdary Research Laboratory Inc. (London) 32 P-labeled lyso-PtdOH was prepared as follows. Chlamydomonas cells were prelabeled with 32 P i for 30 min and treated for 5 min with 2 M mastoparan. Lipids were extracted, and nine-tenths of the material was separated by alkaline TLC. After autoradiography, [ 32 P]PtdOH was isolated from the TLC plate, extracted from the silica, dried, and resuspended in 100 mM Tris⅐HCl, pH 8.95, by sonication. The material was incubated at room temperature in the presence or absence of 50 units of PLA 2 in 100 l of 100 mM Tris⅐HCl, pH 8.95, 9.1 mM CaCl 2 . Lipids were extracted after 1 h and subsequently separated by TLC together with the remaining one-tenth of the starting material to compare the position of lyso-PtdOH with other Chlamydomonas phospholipids. The DGPP standard (dioleolylglycerol pyrophosphate) was provided by Dr. J. B. Wissing (Gesellschaft fü r Biotechnologische Forschung, Braunschweig, Germany).
Chlamydomonas Cell Culture, Metabolic Radiolabeling, and Stimulation-The unicellular, biflagellated green alga Chlamydomonas eugametos, strain 17.17.2, was cultivated on agar-containing M1 medium in Petri dishes and maintained at 20°C in a 12-h light/12-h dark regime with an average photon flux of 30 E m Ϫ2 s Ϫ1 , which was provided by Philips TL 65W/33 fluorescent tubes (Schuring et al., 1987).
Phospholipids were metabolically labeled by incubating cells (1-2 ϫ 10 7 cells per ml) with 100 Ci of carrier-free 32 PO 4 3Ϫ /ml in HMCK for the times indicated in the figure legends. When necessary, excess label was removed by washing the cells twice with HMCK. For the identification of CMP⅐PtdOH, cells were labeled for 5 h with 10 Ci of [ 3 H]cytidine/ sample (100 l). After extraction and TLC, the plate was sprayed with En 3 Hance (DuPont NEN) and exposed for 3 weeks to visualize 3 Hlabeled CMP⅐PtdOH by autoradiography.
Routinely, cells were treated with mastoparan in a total volume of 100 l for the time and concentrations indicated within the figure legends. Occasionally, for the preparation of labeled standards, larger volumes were used. Incubations were stopped by the addition of 3.75 volumes of CHCl 3 /MeOH/HCl (50:100:1 (v/v/v)), and lipids were extracted, separated, and quantified as described below.
Extraction and Analysis of Chlamydomonas Lipids-The extraction mixture was completed, and a two-phase system was induced by the addition of 3.75 volumes of CHCl 3 and 1 volume of 2 M HCl. Tubes were vortexed for 15 s and centrifuged for 2 min in a microcentrifuge. The organic lower phase was washed once with 3.75 volumes of CHCl 3 , MeOH, 1 M HCl (3:48:47 (v/v/v)) and dried by vacuum centrifugation. Lipids were dissolved in CHCl 3 and stored under nitrogen at Ϫ20°C.
Routinely, lipids were separated by TLC using an alkaline solvent system (CHCl 3 , MeOH, 25% NH 4 OH, H 2 O; 90:70:4:16 (v/v/v)) as described previously (Munnik et al., 1994a). For the preparation and purification of 32 P-labeled DGPP, two-dimensional TLC was used. After the alkaline solvent system in the first dimension, plates were dried and chromatographed in a second dimension using CHCl 3 /pyridine/ formic acid (35:30: Radiolabeled phospholipids were detected by autoradiography. Individual spots were scraped from the TLC plate for further processing or quantified by liquid scintillation counting. Alternatively, radioactivity was determined directly on the TLC plate using a PhosphorImager (Molecular Dynamics). Unlabeled phospholipid standards (ϳ10 g) were visualized by exposure to iodine vapor.
Controlled Breakdown of DGPP-32 P-Labeled DGPP was purified by two-dimensional TLC as described above. The lipid was scraped from the TLC plate and deacylated with monomethylamine at 53°C for 30 min as described in Munnik et al. (1994a). The resulting glyceropyrophosphate (GroPP) was dried by vacuum centrifugation, dissolved in H 2 O, and subsequently desalted on a cation-exchange column (Bio-Rad AG 50W ϫ8 200 -400, H ϩ form). Samples were concentrated by vacuum centrifugation and stored at Ϫ20°C in H 2 O.
Mild acid hydrolysis of the pyrophosphate bond of GroPP into P i and GroP was achieved by treating the sample with 1 M trichloroacetic acid at 100°C for 5 min. Trichloroacetic acid was removed by extracting the sample 3 times with water-saturated diethylether. Alternatively, GroPP was hydrolyzed using 4 M formic acid for 10 min at 100°C. Both treatments hydrolyzed [␥-32 P]ATP but left [ 32 P]GroP intact (data not shown); however, the formic acid method had the advantage that samples could be dried directly without leaving salt traces that interfered with subsequent analyses.
Enzymatic breakdown of GroPP was performed by treatment with inorganic pyrophosphatase. Two different sources were used, each applied according to the supplier's protocol (Sigma). The enzyme from S. cerevisiae (112 units, 0.8 mg, 85% buffer salts) was dissolved in 112 l of 10 mM Hepes at pH 7.2, while that from E. coli (1 mg, 100 units, 15% buffer salts) was dissolved in 100 l of 100 mM Tris⅐HCl, pH 8.9. Samples were incubated with 5 units of enzyme in 50 l of buffer for 2 h at room temperature. Protein was removed by EtOH precipitation. Samples were dried by vacuum centrifugation and stored in H 2 O at Ϫ20°C.
Samples were spotted on PEI cellulose anion-exchange TLC polygram sheets, and GroP, P i , and GroPP were separated using a mixture of 0.5 M ammonium formate and 0.2 M formic acid as a solvent. 32 Plabeled P i standard was from Amersham Corp. A [ 32 P]GroP standard was prepared by deacylation of metabolically labeled Chlamydomonas-PtdOH, which had been purified by two-dimensional TLC as described above. The identities of P i and GroP were confirmed by ionophoresis on Whatman No. 1 paper (56 cm) using 0.1 M sodium oxalate buffer, pH 1.5, at 2000 V⅐h as described by Seiffert and Agranoff (1965). Using this system, the migration rates relative to that of P i for GroP, 2,3-diphosphoglycerol, PP i , and GroPP were 1.44, 1.92, 2.11, and 2.0, respectively (results not shown).
Radioactive spots were detected by autoradiography and quantified by liquid scintillation counting of the corresponding regions of the chromatograph. Unlabeled phosphate standards (ϳ5 g) were visualized by spraying with a P i reagent that was specifically designed for cellulose-based chromatographs and was prepared as follows. 8 ml of stock solution A (4% ammonium molybdate (w/v) in 9% HClO 4 (w/v)) was mixed immediately before use with 24 ml of stock solution B (100 ml of H 2 O, 10 ml of HClO 4 , and 1 ml of concentrated HCl). The chro-matographs were then dried at room temperature and irradiated with a strong UV source to develop the blue colors.
Phospholipid Analysis of Other Plants-Carnation petal-discs (Dianthus caryophyllus L. cv. White Sim) were labeled with 32 P i for 1 h, and their lipids were extracted as described previously (Munnik et al., 1994b).
A potato (Solanum tuberosum L.) suspension culture, prepared from callus of an inbred line, was cultivated at 26°C in a Murashige-Skoog type liquid medium (Murashige and Skoog, 1962) supplemented with 30 g/liter sucrose, 5 mg/liter 1-naphthylacetic acid, and 0.1 mg/liter 6-benzyladenine. Cells were subcultured in intervals of 1 week and were used 4 days after transfer. Prior to labeling, cells were washed 3 times with medium without P i by decanting the supernatant after the culture had settled due to gravity after 2 min. Cells were labeled at room temperature in the same medium using 100 Ci of 32 P i /ml for the times indicated. Samples were treated with mastoparan by withdrawing aliqots of 90 l and adding them to 10 l of mastoparan or H 2 O (control) as specified within the figure. Lipids were extracted as described above for Chlamydomonas.
Leaves from common persicaria (Polygonum persicaria) and figleaved goosefoot (Chenopodium ficifolium) were harvested from mature plants growing in the university grounds. Leaf-discs were 32 P i -labeled for 3 h, and their lipids were extracted as described earlier for carnation petal discs (Munnik et al., 1994b).

RESULTS AND DISCUSSION
Mastoparan-induced Formation of a Novel Phospholipid PLX-When 32 P i -prelabeled Chlamydomonas cells were stimulated with the G-protein activator mastoparan, the subsequent analysis of their lipids revealed two clear responses (Fig.  1): a dramatic increase in the level of PtdOH and the appear-ance of an unknown 32 P-containing lipid, hereafter called phospholipid X (PLX). The PtdOH response is the result of Gprotein-activated PLD and PLC activities (Quarmby et al., 1992;Munnik et al., 1995). The product of PLC activity is actually DAG, but much of it is rapidly converted to PtdOH by DAG kinase. The origin and nature of PLX, however, were unknown and therefore studied further. Fig. 2A shows the kinetics of formation of PtdOH and PLX after mastoparan stimulation. The PtdOH response was very fast and transient, increased within seconds, peaked at about 4 min, and decreased quickly thereafter. The increase in 32 P-labeled PLX was slower but, significantly perhaps, correlated with the disappearance of [ 32 P]PtdOH. When different concentrations of mastoparan were used to stimulate cells for 7 min (Fig. 2B), the amount of PLX formed again seemed related to the amount of PtdOH formed. Based on these observations, we hypothesized that PLX was a metabolic product of PtdOH.

Mastoparan-induced Formation of PtdOH and PLX Is Timeand Dose-dependent-
PLX Is Not Lyso-PtdOH or CMP⅐PtdOH-The immediate biological response to treating Chlamydomonas cells with mastoparan is deflagellation. Since this is followed by the complete regeneration of new flagella within 90 min, lipid synthesis could be rapidly activated on adding mastoparan, and so we considered whether PLX was a key intermediate in the biosynthesis of lipids. Both lyso-PtdOH and CMP⅐PtdOH were good candidates since they are more polar derivatives of PtdOH. Accordingly, we isolated [ 32 P]PtdOH from 32 P-labeled cells that had been stimulated with mastoparan, treated it with PLA 2 to form lyso-PtdOH, and compared its chromatographic properties with those of PLX. As is obvious from Fig. 3, their R f values were different.
In order to identify CMP⅐PtdOH on chromatograms, Chlamydomonas cells were labeled with [ 3 H]cytidine, and the lipids were isolated and separated. While the only radioactive spot co-chromatographed with a CMP⅐PtdOH standard, it did not co-migrate with PLX (Fig. 4). These results established that PLX was not one of the well known metabolic products of PtdOH involved in lipid synthesis (Kent, 1995).
PLX Is Not BisPtdOH but Comigrates with the Novel Phospholipid DGPP-PtdOH is the biologically active product of PLD activity in Chlamydomonas cells (Munnik et al., 1995) and so must be considered a potential signal molecule. Since increases in levels of other signaling molecules are invariably attenuated by metabolism to less active derivatives, the same can be expected of PtdOH. For example, BisPtdOH is a newly discovered condensation product of PtdOH and DAG, whose formation in bradykinin-stimulated human fibroblasts attenuates both signals (Van Blitterswijk and Hilkmann, 1993). Since both PtdOH and DAG are formed in Chlamydomonas cells treated with mastoparan, we considered whether PLX is BisPt-dOH. However, we immediately found that the former is so much more polar than the latter (R f values in the alkaline TLC system were 0.39 and 0.92, respectively) that the two cannot be confused (results not shown). Another derivative of PtdOH has recently been reported by Wissing and Behrbohm (1993a). They discovered an in vitro enzyme activity in plants that utilized ATP and PtdOH to produce a new phospholipid Behrbohm, 1993a, 1993b). NMR analysis revealed it to be DGPP, and therefore they called the enzyme phosphatidate kinase. To test whether PLX could be DGPP, we compared their TLC properties using a standard generously provided by Dr. Wissing (GBF, Braunschweig, Germany). On a two-dimen-sional TLC combining the alkaline TLC system with an acidic second solvent, their chromatographic properties were found to be identical (Fig. 5). This result indicated that PLX could be DGPP, a phosphorylated form of PtdOH.
Identification of DGPP by Chemical and Enzymatic Degradation-To further establish the identity of PLX as DGPP, we devised the scheme shown in Fig. 6. It predicts that removal of the fatty acids will result in the formation of GroPP. However, because no standard was available, the molecule was further degraded. Since a pyrophosphate bond is much more acidsensitive than a monoester phosphate, mild acid hydrolysis was used to degrade GroPP to GroP and P i , for which standards were available. In addition, the sensitivity to inorganic pyrophosphatase was tested. 32 P-Labeled PLX was purified from mastoparan-stimulated Chlamydomonas cells by two-dimensional TLC and deacylated, and the water-soluble headgroup (i.e., PX or GroPP) was analyzed by anion-exchange TLC. Fig. 7 shows the deacylation product of PLX before and after mild acid hydrolysis. Before the hydrolysis, PX contained all of the radioactivity of PLX and migrated as a single spot whose R f was different from the phospholipid headgroups of PtdInsP 2 , PtdInsP, PtdOH, PtdIns, PtdEtn, and PtdGro (not shown). Hydrolysis in 4 M formic acid at 100°C for 10 min completely degraded PX, producing two products that contained all of the radioactivity and co-migrated exactly with standards of GroP and P i , as analyzed by anionexchange TLC (Fig. 7) and ionophoresis (see "Experimental Procedures"; not shown). In parallel analyses, GroP and P i were identified as the only degradation products when PX was hydrolyzed in 1 M trichloroacetic acid or treated with inorganic pyrophosphatase (Fig. 8A and B). Already after 4 min in 1 M trichloroacetic acid at 100°C, about 95% of the PX was degraded into P i and GroP (Fig. 8A). Incubations that were kept on ice also degraded PX and even storage in water at Ϫ20°C for 1-3 weeks resulted in 5-15% breakdown, emphasizing the instability of the molecule. In comparison, GroP was completely stable in 1 M trichloroacetic acid for at least 45 min at 100°C (not shown).
The inorganic pyrophosphatases from E. coli and yeast both degraded PX, although the latter was less effective (Fig. 8B). Together, these results indicate that PX is GroPP, and therefore we deduce that PLX is DGPP.
PtdOH Is Phosphorylated to DGPP in Vivo-The kinetics of formation of PtdOH and DGPP (Fig. 2) suggested that DGPP is formed by phosphorylation of PtdOH. To test this we used the strategy of Stephens and Downes (1990) and determined the relative specific radioactivities of the individual phosphate groups in DGPP after a short labeling period. Assuming that both phosphorylations use the ␥-phosphate of the same ATP pool, then while the specific radioactivity of ATP is still increasing, the phosphate with the highest specific activity is the one added last during synthesis. Accordingly, cells were labeled with 32 P i for 5 min and stimulated with mastoparan for 4 min, a period during which 32 P i is still being taken up into the cells and incorporated into the ATP pool. DGPP was isolated by two-dimensional TLC and hydrolyzed in the presence of excess nonradioactive GroP and P i to prevent a specific loss. The radioactivity in GroP and P i was then determined after separation on PEI cellulose. It revealed that 35.6% of the 32 Plabeled GroPP was in GroP and 64.4% in P i . Even in DGPP samples taken after longer periods of mastoparan stimulations (up to 25 min), the radioactivity in P i was always higher than in GroP, also see the data in Fig. 8. These results are consistent with DGPP being formed by phosphorylation of PtdOH.
Occurrence in Other Plants-To verify the presence of DGPP in other living plant tissues, we labeled leaves of Polygonum persicaria and Chenopodium ficifolium, microspores of Brassica napus, Vicia sativa roots, and flower petals of Dianthus caryophyllus with 32 P i and analyzed their lipids. All extracts contained a radioactive lipid that co-chromatographed with DGPP on both one-and two-dimensional TLC. Its headgroup comigrated with Chlamydomonas GroPP and was degraded into GroP and P i after mild acid hydrolysis (not shown). It is a minor lipid; for example after 1-3 h of incubation in 32 P i , the radioactivity in DGPP accounted for 2.5, 1.5, 1.0, 5.3, and 1.0% of the total present in the lipid fractions extracted from P. persicaria, C. ficifolium, B. napus, V. sativa, and D. caryophyllus, respectively. The molecular percentages are expected to be lower because DGPP contains two phosphates that are rapidly labeled via ATP and because no attempt was made to reach isotopic equilibrium. These results are in agreement with the data of Wissing and Behrbohm (1993b) who found in vitro PtdOH-kinase activity in the callus cell membranes of 16 different plants.
FIG. 7. Deacylated PLX before and after mild acid hydrolysis and TLC on PEI cellulose. 32 P-Prelabeled Chlamydomonas cells were treated with mastoparan for 30 min, and their lipids were extracted, separated by two-dimensional TLC, and visualized by autoradiography. 32 P-Labeled DGPP was isolated, and the fatty acids were removed by deacylation with monomethylamine. The water-soluble headgroup, presumed to be GroPP, was subsequently treated for 10 min with 4 M formic acid at 100°C or kept on ice as a control. Samples were chromatographed with a mixture of 0.5 M ammonium formate and 0.2 M formic acid on PEI cellulose together with nonradioactive P i and GroP. The positions of the standards were visualized by phosphate stain. Their position relative to the radioactive compounds detected by autoradiography is indicated by arrows.
suspension cultures were prelabeled with 32 P i and then treated with mastoparan. Lipids were then extracted and separated as illustrated in Fig. 9. In analogy with Chlamydomonas, the synthesis of PtdOH and DGPP was again stimulated by mastoparan. While this implies that these cells have receptors that are coupled by G-proteins to PLC, it also indicates that plants in general metabolize PtdOH to DGPP, in particular when signaling is stimulated. In a separate study, human platelets and U937 cells were prelabeled with 32 P i . The platelets were then stimulated with thrombin or mastoparan, and the U937 cells were stimulated with ATP. While these treatments increased the level of radiolabeled PtdOH, no radioactive DGPP spot was detected on TLC. 2 Thus the phosphorylation of PtdOH to form DGPP could be confined to plants.
Possible Functions-We have previously shown that G-protein activators such as mastoparan stimulate both PLC and PLD signaling in plant cells (Munnik et al., 1995). The subsequent rapid synthesis of PtdOH and its metabolism to DGPP indicate that both compounds could play an important role in signaling. Since PtdOH is the direct product of PLD activity, it must be considered a signal in its own right, just as in animal cells where it is reported to activate specific protein kinases (Bocckino et al., 1991;Epand et al., 1992;Nakanishi and Exton, 1992;Stasek et al. 1993;Khan et al. 1994;Limatola et al., 1994). In support, we have shown that PtdOH added to Chlamydomonas cells induces the same biological responses as the G-protein activators themselves (Munnik et al., 1995). Because cells must be able to respond to repeated stimuli, increased signal levels have to be constantly attentuated, and the conversion of PtdOH to DGPP may be an example of signal attenuation. This is represented in our model of signaling in a plant cell (Fig. 10), where the increase in PtdOH concentration is attenuated by PtdOH-kinase, which is known to be located at the plasma membrane (Wissing and Behrbohm, 1993b). While this hypothesis emphasizes the potential importance of PtdOH as a signal, it relegates DGPP to the realm of biologically inactive compounds, and that may not be the case. As a minor polar lipid that dramatically increases and then decreases in concentration when cells are activated, it has itself the potential to be a signal molecule. While there is no evidence for or against this (see question mark in Fig. 10), it remains possible that DGPP formation is not just a means of inactivating PtdOH.
FIG. 8. Trichloroacetic acid hydrolysis and pyrophosphatase digestion of deacylated PLX (GroPP). 32 P-labeled DGPP purfied by two-dimensional TLC from mastoparan-activated Chlamydomonas (15 min of labeling and 30 min of stimulation) was deacylated and treated with 1 M trichloroacetic acid at 100°C for the times indicated or kept on ice as a control (A) or incubated with inorganic pyrophosphatase from E. coli, or yeast, or with buffer only as described under "Experimental Procedures" (B). Labeled compounds were separated by PEI cellulose TLC and visualized by autoradiography. Radioactive spots were cut out of the polygram sheets and the radioactivity was quantified by liquid scintillation counting. Another possibility for DGPP stems from the fact that it is, like CMP⅐PtdOH, an activated phosphatidate molecule. In the presence of a phosphatidyl transferase and an appropriate substrate, e.g. inositol, it could be converted to other lipids such as phosphatidylinositol. As such it could help maintain the PtdIns cycle, whereby polyphosphoinositides hydrolyzed during PLC signaling are refurbished via DAG, PtdOH, DGPP, and PtdIns formation. In animal cells, the possibility of resynthesis of PtdIns from PtdOH in the plasma membrane remains controversial (e.g. Rara and Hokin (1990)) and it may be that PtdOH is transferred from the plasma membrane to the endoplasmic reticulum where it is converted via CMP⅐PtdOH into PtdIns. It is then retransferred to the plasma membrane via a PtdIns transfer protein, where it can be converted to PtdInsP 2 to complete the cycle. In plants, both DAG kinase and PtdOHkinase are localized in the plasma membrane (Lundberg and Sommarin, 1992;Wissing and Wagner, 1992;Wissing and Behrbohm, 1993b). The subsequent presence of DGPP in the plasma membrane implies that plants could retain a PtdIns cycle within the plasma membrane, without the need to shuttle lipids back and forth to the endoplasmic reticulum. The general importance of maintaining a PtdIns cycle has recently been emphasized by showing that the PtdIns transfer protein and CMP⅐PtdOH synthase are essential for insect and human cell lines to sustain signaling and normal development (Thomas et al., 1993;Cunningham et al., 1995;Martin, 1995;Wu et al., 1995). Perhaps because of this importance, plant cells have developed an alternative PtdIns cycle that operates locally during signaling. It could be justified by plants growing at lower temperatures than animals, for recycling lipids between the plasma membrane and endoplasmatic reticulum at low temperatures could limit cell signaling much more than recycling at the plasma membrane. A last possibility is that DGPP is a precursor for phosphatidylating proteins. PtdOH has recently been identified as the membrane anchor for ubiquitin of Autographa californica nuclear polyhedrosis virus infecting Sf9 cells (Guarino et al., 1995). We are now purifying relatively large quantities of DGPP to test its fate and biological activity in Chlamydomonas cells, in order to distinguish between these possibilities.