Journal of Neicrochemistry Raven Press, Ltd., New York 8 1992 International Society for Neurochemistry
Phorbol Ester-Mediated Stimulation of Phospholipase D Activity in Sciatic Nerve from Normal and Diabetic Rats Bart J. L. Eggen and Joseph Eichberg Department of Biochemical and Biophysical Sciences, University c?f Houston, Houston, Texas, U.S.A.
Abstract: Evidence for the presence of phospholipase D activity in sciatic nerve was obtained by incubation of 32P-prelabeled nerve segments in the presence of ethanol and mea(PEth) formation exsurement of [32P]phosphatidylethanol pressed as a fraction of total phospholipid radioactivity. PEth synthesis was enhanced with increasing concentrations of ethanol (100 mM-2 M ) . 4-@-Phorboldibutyrate (100 nM-I p M ) stimulated PEth formation up to twofold in a time- and dosedependent manner. The stirnulatory effect evoked by 100 nMphorbol ester was completely abolished by Ro 3 1-8220 (compound 3), a selective protein kinase C inhibitor. Efforts to identify the phospholipid precursor of PEth were unsuccessful, suggesting this product arises from a small discrete precursor pool. On subcellular fractionation of nerve, the ratio of basal and 4-P-phorbol dibutyrate-stimulated phospholipase D activity recovered in a myelin-enrichcd fraction, compared with a nonmyelin fraction, was 0.5 when results are expressed as a percentage
of total phospholipid radioactivity. This ratio rises to 1.2 if the results are calculated assuming only phosphatidylcholine and phosphatidylethanolamine are potential precursors. The results suggest that myelin is a major locus of phospholipase D activity. Nerve from streptozotocin-induced diabetic and control animals displayed the same basal phospholipase D activity, but the enzyme in diabetic nerve was stimulated to a greater extent by a suboptimal concentration of 4-P-phorbol dibutyrate. These results support the conclusion that protein kinase C modulates phospholipase D activity in nerve and suggest that in diabetic nerve the cnzyme activation mechanism may possess increased sensitivity. Key Words: Phospholipase D-Myelin-Diabetic neuropathy-Peripheral nerve-Phosphatidylethanol. Eggen B. J. L. and Eichberg J. Phorbol estermediated stimulation of phospholipase D activity in sciatic nerve from normal and diabetic rats. J. Neurochern. 59, 1467-1473 (1992).
The involvement of phosphoinositidesin cell signal transduction mechanisms is well established (Rana and Hokin, 1990). Many hormones, growth factors, and other agents elicit receptor-mediated stimu!2tion of a phospholipase C that degrades phosphatidylinosito1 4,5-bisphosphateto give two biologically active hydrolysis products, inositol 1,4,5-trisphosphate,which mobilizes Ca2+from sequestered intracellular stores, and 1,2-diacylglycerol (DAG), which activates protein kinase C and promotes its translocation to the plasma membrane. More recently, the phospholipase C-catalyzed degradation of other glycerophospholipids, especially phosphatidylcholine, has been recognized to be stimulated on activation of a variety of receptors (Billah and Anthes, 1990; Exton, 1990). This mechanism is of interest in that it provides a means for the generation of DAG without the concomitant production of inositol trisphosphate.
In recent years, evidence has accumulated that phosphatidylcholine can also undergo agonist-induced breakdown via the action of phospholipase D (PLD) in if wide variety of tissues and cells to yie!d phosphatidic acid (PA) and choline (Billah and Anthes, 1990; Exton, 1990). PA can also serve as a precursor of DAG via the action of PA phosphatase. However, PA itself also has a variety of biological effects, including the inhibition of stimulated cyclic AMP formation elicited by certain agonists (Murayama and Ui, 1987),promotion of arachidonic acid release (Kroll et al., 1989), and an ability to act as a mitogenic factor for some cell types (Moolenaar et al., 1986; Wu et al., 1988; Knauss et al., 1990). The addition of either PLD or PA to pancreatic islets has recently been shown to enhance insulin secretion (Dunlop and Larkins, 1990;Metz and Dunlop, 1990). The presence of PLD in the nervous system has
Received June 28, 1991; final revised manuscript receivcd March I I , 1992; accepted April 7, 1992. Address correspondence and reprint rcquests to Dr. J. Eichberg at Department of Biochemical and Biophysical Sciences, University of Houston, Houston, T X 77204-5934, U.S.A.
Abbreviations used: DAG, I ,2diacylglycerol; PA, phosphatidic acid; PDB, 4-0-phorbol 12,13-dibutyratc; PEth, phosphatidylethanol; PLD, phospholipase D.
1467
B. J. L. EGGEN AND J. EICHBERG
I468
been documented for brain extracts and synaptosoma1 preparations and in both transformed and pnmary cultured cclls of nervous system origin (Taki and Kanfer, 1979; Hattori and Kanfer, 1985; Qian and Drewes, 1989; Gustavsson and Hansson, 1990; Martinson et al., 1990; Sandmann and Wurtman, 1991). A report of PLD activity associated with the “microsomal myclin” fraction of sciatic nerve (Chattopadhyay et al., 199 1) appeared after this current work was completed. Previous findings in our laboratory and others have documented disturbances in phosphoinositide metabolism in peripheral nerve from experimentally diabetic rats. Evidence was obtained that the monoesterified phosphate moieties of myelin polyphosphoinositides undergo increased turnover in diabetic nerve (Bell et al., 1982; Lowery et al., 1989). In contrast, the incorporation of rny~-[~H]inositol into phosphoinositides is decreased (Hothersall and McLean, 1979; Bell and Eichberg, 1985) and a pool of myo-inositol used for phosphatidylinositol synthesis is depleted in diabetic nerve (Zhu and Eichberg, 1990a). Moreover, the content of total DAG is significantly reduced in diabetic nerve and the level of arachidonyl-containing molecular species of this neutral lipid explains most of this decrease (Zhu and Eichberg, 1990b). Quite recently, muscarinic cholinergic receptors that mediate enhanced phosphoinositide breakdown by phospholipase C-catalyzed hydrolysis have been characterized in sciatic nerve (Day et al., 199 1). The present study was undertaken to determine whether PLD activity is demonstrable in intact peripheral nerve and whether the enzyme can be stimulated either by a muscannic agonist or an activator of protein kinase C. Inasmuch as DAG can be generated by PLD action via PA, another goal was t o evaluate whether the activity of the enzyme differs in nerve from normal compared with diabetic animals. To measure phospholipase D activity, the well-known transphosphatidylating activity of the enzyme to form phosphatidylethanol (PEth) in the presence of ethanol and phospholipid substrate (Dawson, 1967; Yang et al., 1967; Kobayashi and Kanfer, 1987) was used.
+
MATERIALS AND METHODS Treatment of animals and preparation of labeled nerve segments Male Sprague-Dawley rats ( 150- 175 g; Harlan) were made diabetic by intraperitoneal injection of streptozotocin (60 mg/kg of body weight; Sigma Chemical Co.). These animals and agc-matched controls were allowed food and water ad libitum. Rats were killed 8-10 weeks later and blood was immediately collected for determination of serum glucose by using a glucose diagnostic kit (Sigma no. 510). Diabetic animals exhibited severe weight loss (3050%)and pronounced hyperglycemia (>600 mg/dl). Sciatic nerves were dissected, desheathed, and cut into 1S-3-mm segments. The segments (eight to twelve per incubation) were equilibrated by incubation for 30 min at J. Neurochem., Vol. 59, No. 4, 1992
37°C in 300 pl of Krebs-Ringer bicarbonatc medium containing 5 mMglucosc. Subsequently, 300 pl of this medium containing 200 pCi of 32Piwas added and the segments were labelcd for 3 h. All animal use procedures were in strict accordance with the NIH Guide for the Care and Use of LaboratoryAnimals and were approved by the institutional animal care committee.
Incubation of 32P-labelednerve segments with ethanol The radioactive Krebs-Ringer medium was replaced by 600 pl of fresh radioactivity-free medium containing cthanol. When 4-/3-phorbol dibutyrate (PDB; LC Services) was present, it was dissolved in dimethyl sulfoxide and added in 300 pl of medium followed after 5 min by ethanol and other additions as desired in an additional 300 pl of medium. The final concentration of dimethyl sulfoxide was 0.1% in PDBcontaining and control media. Incubations were performed for 5-60 min and stopped by placing the samples on ice and adding several milliliters of ice-cold 0.9% NaCI. The segments were washed twice with saline and the nerve segments were then divided into groups, each containing three or four pieces, for analysis.
Lipid extraction and separation To each group of segments was added 1 ml chloroform/ methanol ( 1 :I, vol/vol). Lipids were extracted essentially according to Bligh and Dyer (1959) with substitution of 0.9% NaCl for water. The resulting lower phase was washed with 0.9% NaCl/methanol (1 :1, vol/vol), made to one phase by addition of methanol and dried under a stream of nitrogen. The residue was dissolved in 200 p1 chloroform/methanol/water (75:25:2, by vol) and aliquots removed to determine total phospholipid radioactivity. Individual phospholipid classes were separated by TLC on silica gel 60A (Whatman) in chloroform/methanol/NH.,OH (65:25:5 by vol; solvent system A). The plate was autoradiographed and the area of silica gel containing PEth was scraped off and countcd. The quantity of PEth formed was expressed as a percentage of total phospholipid radioactivity.
Preparation of standard PEth Authentic PEth was prepared by a procedure modified from that of Kobayashi and Kanfer (1 987). [32P]Phosphatidylcholinc was purified from a 32P-labelednerve lipid extract by TLC as described above and eluted from silica gel using chlorofonn/methanol/water (1 :1:0.2, by vol). The radioactive phospholipid was hydrolyzed overnight at 30°C in a reaction mixture composed of 0.1 M sodium acetate buffer, pH 5.6, 38 mM CaCI,, 0.5 mM sodium dodecyl sulfate, 2 M ethanol, and 0.25-1.0 mg of cabbage PLD (Boehringer-Mannheim) in a total volume of 1 ml. The resulting radioactive products, PEth, PA, and undigested phosphatidylcholine, were extractcd and separated by T I X as previously described. Alternatively, PEth was isolated by two-dimensional TLC, using solvent system A in the first dimension and chloroform/methanol/acetone/ghcial acetic acid/water (75: 15:30:15:7.5. by vol; solvent system B) as the solvent system in the second dimension. Radioactive lipids were located by autoradiography and eluted with chloroform/mcthanol/watcr ( 1 :1:0.2, by vol).
Enzymatic degradation of [32P]PEth Purified PEth was hydrolyzed overnight with Bacillus cereus phospholipase C (Sigma Chemical Co.) by the procedure of Mavis et al. ( 1 972). The phosphoethanol formed was separated by TLC on cellulose (MN 300; Analtech) in
PHOSPHOLIPASE D IN SCIATIC NERVE H,O/NH,OH (28%)/triehloroaceticacid (52335: 10:3.5, by vol; solvent system C) (Tettenborn and Muller, 1987).
Preparation and extraction of myelin Peripheral myelin was purified according to a modified procedure of Toews et al. (1987). All steps were performed at 4°C. Nerves were homogenized using a motor-driven glass-glass homogenizer in 2 ml of 0.32 M sucrose/25 mM Tris buffer, pH 8.5. The homogenatc derived from one nerve was placed on a 1.8-ml layer of 0.85 M sucrose/25 mM Tris buffer, pH 8.5, and centrifuged for 45 rnin at 82,000 g in a Beckman SW 60 Ti rotor. The myelinenriched layer floating on 0.85 Msucrose and the “nonmyelin” pellet at the bottom of the tube were recovered, diluted 20 times with 25 mM Tris buffer, pH 8.5, and resedimented for 15 rnin at 82.000 g. The resulting pellets were resuspended in Tris buffer and lipids were extracted as previously described except that 0.1 M HC1 was substituted for 0.9% NaCI. The washed lower phases were neutralized with 28% NH,OH before being dried. Lipids were redissolved in chloroform/methanol/H,O (75:25:2, by vol) and separated by TLC on silica gel in either solvent system A or in chlore form/methano1/40% methylamine/H,O (60:36:5:5, by vol).
Statistics Statistical analysis of data was performed by one-way analysis of variance followed by a Tukey-Kramer test. The number (n) ofeither incubations or rats used for these calculations is indicated in the legend to each figure.
RESULTS Incubation of 32P-prelabelednerve segments for 15 rnin in the presence of 100 mMethanol resulted in the formation of a radioactive phospholipid, which on TLC migrated above phosphatidylethanolamine in solvent system A (Fig. 1, lanes 1 and 2). When radioactive nerve lipids were separated by two-dimensional TLC, the unknown spot comigrated with authentic PEth prepared from phosphatidylcholine by cabbage PLD-catalyzed hydrolysis (data not shown). On phospholipase C-catalyzed degradation of purified putative PEth, a radioactive product was obtained - PEth
- PE - PC - Other Phospholipids - Origin 1
2
3
4
5
6
FIG. 1. Autoradiogram showing formation of [=P]PEth in sciatic nerve. Prelabeled nerve segments were incubated for 15 rnin in Krebs-Ringer bicarbonate medium containing glucose, loo mM ethand and no PDB (lanes 1 and-2).1 0 0 nM PDB (lanes 3 and 4). or 1 p M PDB (lanes 5 and 6). Lipids were extracted, separated by TLC in Solvent system A and autoradiographed as described in Materials and Methods. PC, phosphatidylcholine; PE, phosphatidylethanolamine.
1469
that migrated to the same position on cellulose TLC as phosphoethanol prepared from standard PEth (Fig. 2). The extent of basal PEth formation in nerve was less than 0.5% of radioactivity in total phospholipids extracted with neutral solvent, i.e., excluding polyphosphoinositides. PEth synthesis was stimulated in a dose-dependent fashion by PDB (Fig. 1, lanes 3-6, and Fig. 3 ) and was enhanced up to twofold at 1 p M PDB, the highest concentration used. PEth formation was not increased when 1 1 pM 4-a-phorbol, which does not activate protein kinase C, replaced PDB. In some experiments, PDB was added to amplify the signal. In the presence of this phorbol ester, PEth synthesis increased steadily as the concentration of ethanol in the incubation medium was varied from 100 mM to 2 M (Fig. 4). In most experiments, 100 mM (0.5%) ethanol was used to minimize possible perturbation of membrane structure. The time course of enhanced PEth formation in the presence of PDB showed a lag in that increased synthesis was not evident after 5 rnin (Fig. 5). In the experiment shown, PDB failed to elevate PEth levels significantly after 15 min, but in other experiments the phorbol ester did elicit a measurable stimulatory response after this time (cf. Fig. 6). PEth continued to accumulate slowly up to 60 rnin (Fig. 5). The stimulation of PEth generation by phorbol ester suggested the involvement of protein kinase C in this phenomenon. Consequently, the effect of Ro 3 18220 [described as compound 3 in Davis et al. (1989)], a novel, potent, and rather specific protein kinase C inhibitor, on PEth synthesis was examined (Fig. 6). In the presence of 100 nM PDB, 100 nMRo 3 1-8220 partially prevented enhanced PEth formation and l p M Ro 3 1-8220 was completely effective. When 1 pM PDB was present, 1 p M Ro 31-8220 tended to reduce the more pronounced stimulation of PEth synthesis, although the decrease failed to attain statistical significance. The inhibitor alone had no effect on basal PEth synthesis. Efforts to demonstrate increased PEth formation by the muscarinic cholinergic agonist, carbamylcholine, yielded inconsistent results. Up to a 60% rise in PEth generation, which was maximal after a 5-min incubation with 100 mM ethanol, was Seen in some experiments, but no stimulation was observed in others. The reason for this variability is not clear. Several attempts were made to identify the principal phospholipid precursor of PEth by isolation of the radioactive product and determination of its molecular species composition (Zhu and Eichberg, 1990h). This approach was adopted because the molecular species profile of phosphatidylcholine is readily distinguishable from that of phosphatidylethanolamine and all other glycerophospholipids in nerve (Zhu and Eichberg, unpublished results). However, starting with four labeled nerves incubated in the presence of 1 M ethanol and 1 pM PDB, a sufficient amount of J . Neurochem., Vol. 59, No. 4, 1992
B. J. L. EGGEN AND J. EICHBERG
1470
FIG. 2. Autoradiogram showing the liberation of [32P]phosphoethanol by enzymatic degradation of [32P]PEthformed in nerve. Putative r3'P]PEth was purified by two-dimensional TLC, hydrolyzed with 6. cereus phospholipase C, and the water-soluble products separated by TLC in solvent system C as described in Materials and Methods. Radioactive spots were visualized by autoradiography. Lane 1, inorganic phosphate; lane 2, hydrolysis product from nerve [=P]PEth; lane 3, authentic [32P]phosphoethand.
0.10
050
0.20
0.75
I S
1.m
200
Ethanol Concentration(M)
FIG. 4. Effect of varying ethand concentration on PEth formation. Preiabeled nerve segments were incubated for 15 min in the presence of 100 nM PDB and ethanol concentrations as shown. PEth synthesis is expressed as a mean percentage of total phospholipid radioactivity soluble in chloroform/methanol 2 SEM. Each value was derived from at least n = 6 nerve incubations from a minimum of two separate experiments for each condition.
purified PEth could not be isolated to permit molecuIar species determination. The ratio of radioactivity in phosphsiidylchohe to PEth was -4G: 1 under these incubation conditions and four sciatic nerves contain > 3 pmol of phosphatidylcholine. A complete molecular species analysis can be performed on I0 nmol of DAG enzymatically derived from phosphatidylcholine.,Thus, the failure to recover an appreciable mass of PEth suggests that the labeled product must be formed from a small, discrete pool of phosphatidylcholine. A similar argument can be advanced in regard to a precursor role for any of the other major nerve phospholipids. The subcellular distribution of PLD activity in nerve was examined by measurement of PEth fonnation in myelin-enriched and nonmyelin fractions prepared after prelabeled nerve segments were incubated for 15 min with 100 mM ethanol either in the presence or absence of 1 p M PDB. In these experiments,
both neutral and acidic solvents were used to extract all lipids, including polyphosphoinositides, from the fractions. The pattern of phospholipid labeling differed markedly in the two fractions. In myelin, polyphosphoinositides comprised 65-70% phosphatidylcholine 12- 1496, and phosphatidylethanolamine 2-3% of incorporated radioactivity, whereas in nonmyelin the corresponding values were 29-37%, 3042%, and 3-5%, respectively. When expressed as a percentage of total phospholipid radioactivity, myelin exhibited 50%of the PEth-forming activity of nonmyelin and formation of this product was significantly stimulated in the presence of PDB in nonmyelin, but not in myelin (Fig. 7, left panel). However, when PEth synthesis was calculated as a percentage of radioactivity in phosphatidylcholine + phosphatidylethanolamine, the most likely phospholipid precursors, myelin and nonmyelin, displayed essentially equal PLD activities (Fig. 7, right panel).
-
;8{+:I
9
NoPDB
.-n
.a
k
02-
0
L 0
0.1
0.3
1.0
4-6-PDB 0
FIG. 3. Effect of varying concentrations of PDB on PEth synthesis. Prelabeled nerve segments were incubated for 15 min in 100 mM ethanol and either no PDB (basal) or PDB concentrationsas shown. Results are expressed as a percentage of basal PEth formation f SEM and are derived from a total of at least n = 6 nerve incubations from a minimum of two separate experiments for each condition. Unless otherwise indicated, statistical analysis for data in this and subsequent figures was performed as described in Materials and Methods. "p 0.05; **p < 0.01; '"*p < 0.005.
J. Neurochem.. Vol. 59, No. 4, 1992
0.0
.
,
.
,
.
,
'
,
.
l
'
i
PHOSPHOI,IPASE D IN SCIATIC NERVE
1471
Ro 3 1-8220 concentration:
0 None 0 loonM
.a .-
lo
B
O8
0.
.
I
2
5
0.6
8
5
0.4
bR
02
No PDB
00
No PDB
0.1 ph4 PDB
IWPDB
FIG. 6. Inhibition of [32P]PEthformation by Ro 31-8220 (compound 3). Prelabeled nerves were incubated for 15 min with 100 nM ethanol, PDB concentrations as indicated, either in the absence or presence of Ro 31-8220. PEth synthesis is expressed as a percentageof total phospholipid radioactivity soluble in neutral solvent. Results were derived from at least n = 8 nerve incubations from a minimum of two separate experiments for each condition. ' p < 0.05; '*p c 0.005.
Indications that phosphoinositide metabolism is altered in nerve from diabetic animals prompted us to compare PLD activity in normal and diabctic nerve incubated in vitro. The formation of PEth in nerve from diabctic and age-matched normal rats incubated under basal conditions was very similar (Fig. 8). However, in the presence of 100 nMPDB, diabetic nerve exhibited a 74% stimulation of PLD activity, whereas PEth formation in normal tissue rose only 35%. At 1 pM PDB, PEth synthesis was further enhanced to 72% over basal in normal nerve, but increased just
'1 -
EQ
0.4
0.0
Myelin
Non Myelin
Myelin
Non Myelin
FIG. 7. Subcellular distribution of PLD activity in nerve. Nerve segments were prelabeledwith "P and then incubated for 15 min with 100 mM ethanol either in the absence (open bars) or presence (shaded bars) of 1 pM PDB as described in Materials and Methods. Left PEth synthesis is expressed as a percentage of total phospholipid radioactivity. Right: PEth synthesis is expressed as a percentage of the sum of radioactivity in phosphatidylcholine (PC) and phosphatidylethanolamine (PE). Results shown are the mean I SEM from two experiments. In one experiment, duplicate incubationsof nerves from two normal rats were performed. In the second experiment. one normal and one diabetic rat were used. Because all values were similar, they were combined to give a total of n = 4 incubations. Differences between PEth in the absence and presenceof PDB were statistically evaluated by means of Student's t test for myelin and nonmyelin fractions. *p < 0.05, different from no PDB.
0.I pM 4-R-PDB
1 ph4 4-R-PDB
FIG. 8. PDB stimulation of [=P]PEth formation in control and diabetic nerve. Prelabeled nerve segments from diabetic and agematched control rats were incubated for 15 min with 100 nM ethanol either in the absence or presence of PDB. Results are expressed as a percentageof basal PLD activity in normal nerve. The values for each condition were derived from three separate experiments and a total of seven to nine normal rats and six to eight diabetic rats. *p < 0.01 ; * p < 0.001.
slightly to 83% in diabetic nerve, so that the significant difference in response observed at the lower phorbol ester concentration was no longer seen. DISCUSSION
Investigations of PLD in animal tissues have usually relied on quantifying the release of radioactive choline or PA from labeled phospholipids or, alternatively, the formation of PEth from lipid substrate in the presence of ethanol. In our study, initial efforts to demonstrate the liberation of [3H]cholineon incubation of nerve segments prelabeled with this phospholipid precursor were not successful, possibly because of the high basal level of radioactive choline in the tissue. Thus, we turned to determination of PEth liberation from 32P-labelednerve phospholipids, an a p proach that is well suited to detect the relativcly small amounts of radioactive PEth formed. Our rcsults provide strong evidence that PLD is active in intact sciatic nerve scgments and that the enzyme can be stimulated by PDB. Efforts were also made to measure changes in [32P]PAformation in the presence compared with the absence of ethanol. Approximately 10% of isotope incorporated into prelabeled nerve phospholipids was in PA and, consequently, the magnitude of the transphosphatidylation response (ranging from 0.2 to 1.O% as a proportion of total phospholipid label) is such that any alterations in radioactive PA would have fallen within experimental error. Thus, our measurements suggest that the extent of PA formation by PLD action in nerve is very small. Further, it cannot be absolutely ruled out that in addition to PLD, other uncharacterized reactions could contribute to the formation of [32P]PEth. Several other studies have found that phorbol-12myristate- 13-acctate enhances PLD activity in nervous tissue preparations (Gustavsson and Alling, 1987; Gustavsson and Hansson, 1990; Martinson et J. Neurochem., Vol. 59, No. 4, 1992
1472
B. J. L. EGGEN AND J. EKHBERG
al., 1989)and the stimulation of the enzyme by active phorbol diesters is evidently a characteristic response in many tissues and cells (Billah and Anthes, 1990). The complete blockade of PDB-mediated stimulation by Ro 3 1-8220 at the lower concentration of phorbol ester used suggeststhat protein kinase C plays a role in enhancing PLD activity. This inhibitor has a structure based on those of staurosporine and K252a and has been reported to exhibit affinities for protein kinase C that are 1,700-fold and 150-fold greater than for calmodulin-dependent protein kinase and cyclic AMPdependent kinase, respectively (Davis et al., 1989). The inability of this substance to strongly inhibit at the higher PDB concentration is most likely due to an insufficient quantity being present. However, it remains possible that PDB in sufficient quantity interacts directly with PLD as well as protein kinase C to activate the enzyme. There is evidence that PLD a d vation can occur by both protein kinase C-dependent and -independent mechanisms (Billah and Anthes, 1990; Martinson et al., 1990). Although we were not able to identify the phospholipid source of PEth, if it is assumed that phosphatidylcholine is the primary precursor, then only a small amount of this lipid in nerve would be accessible to PLD. Subcellular fractionation of nerve after incubation with ethanol and PDB indicates that the location of this responsive pool as well as PLD is at least partly in myelin or closely associated structures. The tentative identification of PLD as a peripheral myelin constituent adds to the growing list of enzymes present in the myelin sheath (Ledeen, 1984) and strengthens accumulating evidence that myelin possesses much of the signal transduction machinery found in the plasma membrane, including muscarinic receptors, G proteins, phospholipase C , and adenylate cyclase (Larocca et al., 1987a,h; Golly et al., 1990). Basal PLD activity, as assessed by PEth formation, was the same in nerves from diabetic as from normal animals, but the enzyme was stimulated to a greater extent in diabetic nerve in the presence of a level of PDB that was suboptimal for normal tissue. In seeking to explain the significance of this difference, it is well known that Na+,K+-ATPase activity in diabetic nerve is diminished and can be corrected by protein kinase C activators, including phorbol esters and dioctanoylglycerol (Lattimer et al., 1989; Bianchi et al., 1991). The reduced level of endogenous DAG and/or arachidonyl-containing species of this neutral lipid in diabetic nerve (Zhu and Eichberg, 19906) may contribute to loss of Na+,K+-ATPaseactivity by interfering with a critical protein kinase C-dependent activation step. In this context, the greater response of PLD to PDB stimulation in the diabetic statc could reflect a heightened sensitivity of protein kinase C in the tissue to natural activators as a result of an adaptation to compensate for the decreased level of DAG. Consequently, PLD would be more readily stimulated in diabetic nerve, leading in turn to a greater J . Neurochem., Vol. 59, No. 4, I992
production of PA, a precursor of DAG. The possibility that PLD may play a role in the normal maintenance of Na+,K+-ATPaseactivity can by no means be excluded and it is attractive to speculate that in nerve of diabetic animals in vivo, the production of DAG by the PLD-mediated pathway may assume increased importance. Acknowledgment: This work was supported by NIH grants DK30577 and RR07147. We thank Dr. P. D. Davis of Roche Products, Ltd., for a generous gift of Ro 3 1-8220.
REFERENCES Bell M. E. and Eichberg J. (1985) Decreased incorporation of ['HIinositol and [3H]glycerol into glycerolipids of sciatic nerve from the streptozotocin diabetic rat. J. Neurochem. 45,465469. Bell M. E., Peterson R. G., and Eichberg J. (1982) Metabolism of polyphosphoinositides and other phospholipids in peripheral nerve from rats with chronic s~reptozotocin-induceddiabetes: increased turnover of phosphatidylinositol 4,5-bisphosphate. J. Neurochem. 39, 192-200. Besterman J. M., Duronio V., and Cuatrecasas P. (1986) Rapid formation of diacylglycerol from phosphatidylcholine:a pathway for generation of a second mcssenger. Proc. Natl. Acad. Sci. USA 83,6785-6789. Bianchi R., BedMattera L. N., Eichberg J., Triban C., and Fiori M. G. (1991) Studies in vitro and in vivo confirm ganglioside efficacy on Nat, K+-ATPaseactivity in experimental diabetic neuropathy, in Advances in i'hysiological Sciences (Manchanda s. K., Selvamurthy V., and Kumar N., ds), Macmillian India Ltd., New Delhi. Billah M. M. and Anthes J. C . (1990) The regulation and cellular functions of phosphatidylcholine hydrolysis. Biochem. J. 269, 28 1-29 I . Bligh E. A. and Dyer W. J. (1959) A rapid method of total lipid extraction and purification. Cm. J. Biochem. Physiol. 37, 911-917. Chattopadhyay J., Natarajan V., and Schmid H. H. 0. (1991) Membrane-associated phospholipase D activity in rat sciatic nerve. J. Neurochem. 57, 1429-1436. Davis P. D., Hill C. H ., Keech E., Lawton G., Nixon J. S., Sedgwick A. D., Wadsworth J., Westmacott D., and Wilkinson S. E. (1989) Potent selective inhibitors of protein kinase C. FEBS Lett. 259, 61-63. Dawson R. M. C. (1967) The formation of phosphatidylglycerol and other phospholipids by the transferase activity ofphospholipase D. Biochem J. 102, 205-210. Day N., Berti-Mattera L. N., and Eichberg J. (1991) Muscarinic cholincrgic receptor-mediated phosphoinositide metabolism in peripheral nerve. J. Neurochern. 56, 1905-1913. Dunlop M. E. and Larkins R. G. (1 990) Effects ofphosphatidic acid on islet cell phosphoinositide hydrolysis, Caz+and adenylate cyclax. Diubetes 38, 1187-1 192. Exton J. (19YO) Signalingthrough phosphatidylcholine breakdown. J. Biol. Chem. 265, 1-4. Golly F., Larocca J. N., and Ledeen R. W.(1990) Phosphoinositide breakdown in isolated myelin is stimulatcd by GTP analogs and calcium. J. Neurosci. Res. 27, 342-348. Gustavsson L. and Alling C. (1 987) Formation of phosphatidylethanol in rat brain by phospholipase D. Biochem. Biophys. Res. Commun. 142,958-963. Gustavsson L. and Hansson E. (1990) Stimulation of phospholipase D activity by phorbol esters in cultured astrocytes. J. Neurochem. 54,737-742. Hattori H. and Kanfer J. (1 985) Synaptosomal phospholipase D: a potential role in providing choline for acetylcholine synthesis. J. Neurochem. 45, 1578- 1584.
PHOSPHOLIPASE D IN SCIATIC NERVE Hothersall J. A. and McLean P. (1979) Effect ofexpenmental diabetes and insulin on phosphatidylinositol synthesis in rat sciatic nerve. Biochem. Biophys. Res. Commun. 88, 477-484. Knauss T. C., Jaffer F. E., and Abboud H. E. (1990) Phosphatidic acid mediates DNA synthesis, phospholipase C and plateletderived growth factor mRNAs in cultured mesangial cells. J. Bid. Chem. 265, 14457-14463. Kobayashi M. and Kanfer J. N. (I 987) Phosphatidylethanol formation via transphosphatidylation by rat brain synaptosomal phospholipase D.J. Neurochem. 48, 1597-1 603. KrollM. H.,ZovoicoG.B.,andSchaferA. L(1989)Secondmessenger function of phosphdtidic acid in platelet activation. J. Cell Phvsiol. 139, 558-564. Larocca J. N., Ledeen R. W., Dvorkin B., and Makman M. H. ( 1 9 8 7 ~Muscarinic ) receptor binding and muscarinic receptormediated inhibition of adenylate cyclase in rat brain myelin. J. Netirosci. 7, 3869-3876. Larocca J. N., Cervone A., and Ledeen R. W. (19876) Stimulation of phosphoinositide hydrolysis in myelin by muscannic agonists and potassium. Brain Res. 436,357-362. Lattimer S. A., Sima A. A. F., and Greene D. A. (1989) In vitro correction of Na+,K+-ATPasein diabetic nerve by protein kinase C agonists. Am. J. Physiol. 256, E264-E269. Ledeen R. W. (1984) Lipid-metabolizing enzymes of myelin and their relation to the axon, J. Lipid Res. 25, 1548-1554. Loffelholz K. (1989) Receptor regulation of choline phospholipid hydrolysis. Biochem. Pharmacol. 38, 1543-1 549. Lowery J. M., Bed-Mattera L. N., Day S.-F., Peterson R. G., and Eichberg J. (1989) Relationship of ATP turnover, polyphosphoinositide metabolism, and protein phosphorylation in sciatic nerve and derived peripheral myelin subfractions from normal and streptozotocin diabetic rats. J. Nezirochem. 52, 921-932. Martinson E. A., Goldstein D., and Brown J. H. (1 989) Muscarinic receptor activation of phosphatidylcholine hydrolysis. J. Biol. Chem. 264, 14748-14754. Martinson E. A,, Tnlivas l., and Brown J. H. (1990) Rapid protein kinase C-dependent activation of phospholipase D leads to delayed 1,2diglyceride accumulation. J. Biol. Chem. 265, 22282-22287. Mavis R. D., Bell R. M., and Vagelos P. R. (1 972) Effect of phos-
1473
pholipase C hydrolysis of membrane phospholipids on membranous enzymes. J. Biol. Chem. 247,2835-2841. Metz S. A. and Dunlop M. (1990) Stimulation of insulin release by phospholipase D a potential role for endogenousphosphatidic acid in pancreatic islet function. Biochem. J. 270,427-435. Moolenaar W. H., Kruijer W., Tilly B. C., Verlaan I., and de Laat S. W. (1986) Growth factor-like action of phosphatidic acid. Nature 323, 1 7 1 - 1 73. Murayarna T. and Ui M. (1987) Phosphatidic acid may stimulate membrane receptors mediating adenylate cyclase inhibition and phospholipid breakdown in 3T3 fibroblasts. J. Biol. Chem. 262,5522-5529. Qian Z . and Drewes L. R. (1989) Muscarinic acetylcholinereceptor regulates phosphatidylcholine phospholipase D in canine brain. J. Biol. Chem. 264,21720-21724. Rana R. S. and Hokin L. E. (1990) Role of phosphoinositides in transmembrane signaling. Physiol. Rev. 70, 1 15-163. Sandrnann J. and Wurtman R. J. (1991) Stimulation of phospholipase D activity in human neuroblastoma (LA-N-2) cells by activation of muscarinic acetylcholine receptors or by phorbol esters: relationship to phosphoinositide turnover. J. Neurochem. 56, 1312-1319. Taki T. and Kanfer J. N. (1979) Partial purification and properties of a rat brain phospholipaseD. J. Biol. Chem. 254,976 1-9765. Tettenborn C. S. and Mueller G. C. (1987) Phorbol esters activate the pathway for phosphatidylcthanol synthesis in differentiating HL-60 cells. Bicxhim. Biophys. Acta 931,242-250. Toews A. D., Fischer H. R., Goodrum J. F., Windes S., and Morel! P. (1987) Metabolism of phosphate and sulfate groups modifying the Po protein of peripheral nervous system myelin. J. Neurochem. 48,883-887. Wu C.-L., Tsai M.-H., and Stacey D. W. (1988) Cellular ras activity and phospholipid metabolism. Cell 52,68-71. Yang S. F., Freer S., and Benson A. A. (1967) Transphosphatidylation by phospholipase D. J. Biol. Chem. 242,477-484. Zhu X . and Eichberg J. ( 1 9 9 0 ~A ) pool of myeinositol needed for phosphatidylinositol synthesis is depleted in diabetic nerve. Proc. Natl. Acad. Sci. USA 87, 98 18-9822. Zhu X. and Eichberg J. (19904 1,2-Diacylglycerolcontent and its arachidonyl-containing molecular species are reduced in sciatic nerve from streptozotocin-induceddiabetic rats. J. Neurochem. 55, 1087-1090.
J. Neurochem.. Vol. 59, No. 4, 199.2
Phorbol Ester-Mediated Stimulation of Phospholipase D Activity in Sciatic Nerve from Normal and Diabetic Rats Bart J. L. Eggen and Joseph Eichberg Department of Biochemical and Biophysical Sciences, University c?f Houston, Houston, Texas, U.S.A.
Abstract: Evidence for the presence of phospholipase D activity in sciatic nerve was obtained by incubation of 32P-prelabeled nerve segments in the presence of ethanol and mea(PEth) formation exsurement of [32P]phosphatidylethanol pressed as a fraction of total phospholipid radioactivity. PEth synthesis was enhanced with increasing concentrations of ethanol (100 mM-2 M ) . 4-@-Phorboldibutyrate (100 nM-I p M ) stimulated PEth formation up to twofold in a time- and dosedependent manner. The stirnulatory effect evoked by 100 nMphorbol ester was completely abolished by Ro 3 1-8220 (compound 3), a selective protein kinase C inhibitor. Efforts to identify the phospholipid precursor of PEth were unsuccessful, suggesting this product arises from a small discrete precursor pool. On subcellular fractionation of nerve, the ratio of basal and 4-P-phorbol dibutyrate-stimulated phospholipase D activity recovered in a myelin-enrichcd fraction, compared with a nonmyelin fraction, was 0.5 when results are expressed as a percentage
of total phospholipid radioactivity. This ratio rises to 1.2 if the results are calculated assuming only phosphatidylcholine and phosphatidylethanolamine are potential precursors. The results suggest that myelin is a major locus of phospholipase D activity. Nerve from streptozotocin-induced diabetic and control animals displayed the same basal phospholipase D activity, but the enzyme in diabetic nerve was stimulated to a greater extent by a suboptimal concentration of 4-P-phorbol dibutyrate. These results support the conclusion that protein kinase C modulates phospholipase D activity in nerve and suggest that in diabetic nerve the cnzyme activation mechanism may possess increased sensitivity. Key Words: Phospholipase D-Myelin-Diabetic neuropathy-Peripheral nerve-Phosphatidylethanol. Eggen B. J. L. and Eichberg J. Phorbol estermediated stimulation of phospholipase D activity in sciatic nerve from normal and diabetic rats. J. Neurochern. 59, 1467-1473 (1992).
The involvement of phosphoinositidesin cell signal transduction mechanisms is well established (Rana and Hokin, 1990). Many hormones, growth factors, and other agents elicit receptor-mediated stimu!2tion of a phospholipase C that degrades phosphatidylinosito1 4,5-bisphosphateto give two biologically active hydrolysis products, inositol 1,4,5-trisphosphate,which mobilizes Ca2+from sequestered intracellular stores, and 1,2-diacylglycerol (DAG), which activates protein kinase C and promotes its translocation to the plasma membrane. More recently, the phospholipase C-catalyzed degradation of other glycerophospholipids, especially phosphatidylcholine, has been recognized to be stimulated on activation of a variety of receptors (Billah and Anthes, 1990; Exton, 1990). This mechanism is of interest in that it provides a means for the generation of DAG without the concomitant production of inositol trisphosphate.
In recent years, evidence has accumulated that phosphatidylcholine can also undergo agonist-induced breakdown via the action of phospholipase D (PLD) in if wide variety of tissues and cells to yie!d phosphatidic acid (PA) and choline (Billah and Anthes, 1990; Exton, 1990). PA can also serve as a precursor of DAG via the action of PA phosphatase. However, PA itself also has a variety of biological effects, including the inhibition of stimulated cyclic AMP formation elicited by certain agonists (Murayama and Ui, 1987),promotion of arachidonic acid release (Kroll et al., 1989), and an ability to act as a mitogenic factor for some cell types (Moolenaar et al., 1986; Wu et al., 1988; Knauss et al., 1990). The addition of either PLD or PA to pancreatic islets has recently been shown to enhance insulin secretion (Dunlop and Larkins, 1990;Metz and Dunlop, 1990). The presence of PLD in the nervous system has
Received June 28, 1991; final revised manuscript receivcd March I I , 1992; accepted April 7, 1992. Address correspondence and reprint rcquests to Dr. J. Eichberg at Department of Biochemical and Biophysical Sciences, University of Houston, Houston, T X 77204-5934, U.S.A.
Abbreviations used: DAG, I ,2diacylglycerol; PA, phosphatidic acid; PDB, 4-0-phorbol 12,13-dibutyratc; PEth, phosphatidylethanol; PLD, phospholipase D.
1467
B. J. L. EGGEN AND J. EICHBERG
I468
been documented for brain extracts and synaptosoma1 preparations and in both transformed and pnmary cultured cclls of nervous system origin (Taki and Kanfer, 1979; Hattori and Kanfer, 1985; Qian and Drewes, 1989; Gustavsson and Hansson, 1990; Martinson et al., 1990; Sandmann and Wurtman, 1991). A report of PLD activity associated with the “microsomal myclin” fraction of sciatic nerve (Chattopadhyay et al., 199 1) appeared after this current work was completed. Previous findings in our laboratory and others have documented disturbances in phosphoinositide metabolism in peripheral nerve from experimentally diabetic rats. Evidence was obtained that the monoesterified phosphate moieties of myelin polyphosphoinositides undergo increased turnover in diabetic nerve (Bell et al., 1982; Lowery et al., 1989). In contrast, the incorporation of rny~-[~H]inositol into phosphoinositides is decreased (Hothersall and McLean, 1979; Bell and Eichberg, 1985) and a pool of myo-inositol used for phosphatidylinositol synthesis is depleted in diabetic nerve (Zhu and Eichberg, 1990a). Moreover, the content of total DAG is significantly reduced in diabetic nerve and the level of arachidonyl-containing molecular species of this neutral lipid explains most of this decrease (Zhu and Eichberg, 1990b). Quite recently, muscarinic cholinergic receptors that mediate enhanced phosphoinositide breakdown by phospholipase C-catalyzed hydrolysis have been characterized in sciatic nerve (Day et al., 199 1). The present study was undertaken to determine whether PLD activity is demonstrable in intact peripheral nerve and whether the enzyme can be stimulated either by a muscannic agonist or an activator of protein kinase C. Inasmuch as DAG can be generated by PLD action via PA, another goal was t o evaluate whether the activity of the enzyme differs in nerve from normal compared with diabetic animals. To measure phospholipase D activity, the well-known transphosphatidylating activity of the enzyme to form phosphatidylethanol (PEth) in the presence of ethanol and phospholipid substrate (Dawson, 1967; Yang et al., 1967; Kobayashi and Kanfer, 1987) was used.
+
MATERIALS AND METHODS Treatment of animals and preparation of labeled nerve segments Male Sprague-Dawley rats ( 150- 175 g; Harlan) were made diabetic by intraperitoneal injection of streptozotocin (60 mg/kg of body weight; Sigma Chemical Co.). These animals and agc-matched controls were allowed food and water ad libitum. Rats were killed 8-10 weeks later and blood was immediately collected for determination of serum glucose by using a glucose diagnostic kit (Sigma no. 510). Diabetic animals exhibited severe weight loss (3050%)and pronounced hyperglycemia (>600 mg/dl). Sciatic nerves were dissected, desheathed, and cut into 1S-3-mm segments. The segments (eight to twelve per incubation) were equilibrated by incubation for 30 min at J. Neurochem., Vol. 59, No. 4, 1992
37°C in 300 pl of Krebs-Ringer bicarbonatc medium containing 5 mMglucosc. Subsequently, 300 pl of this medium containing 200 pCi of 32Piwas added and the segments were labelcd for 3 h. All animal use procedures were in strict accordance with the NIH Guide for the Care and Use of LaboratoryAnimals and were approved by the institutional animal care committee.
Incubation of 32P-labelednerve segments with ethanol The radioactive Krebs-Ringer medium was replaced by 600 pl of fresh radioactivity-free medium containing cthanol. When 4-/3-phorbol dibutyrate (PDB; LC Services) was present, it was dissolved in dimethyl sulfoxide and added in 300 pl of medium followed after 5 min by ethanol and other additions as desired in an additional 300 pl of medium. The final concentration of dimethyl sulfoxide was 0.1% in PDBcontaining and control media. Incubations were performed for 5-60 min and stopped by placing the samples on ice and adding several milliliters of ice-cold 0.9% NaCI. The segments were washed twice with saline and the nerve segments were then divided into groups, each containing three or four pieces, for analysis.
Lipid extraction and separation To each group of segments was added 1 ml chloroform/ methanol ( 1 :I, vol/vol). Lipids were extracted essentially according to Bligh and Dyer (1959) with substitution of 0.9% NaCl for water. The resulting lower phase was washed with 0.9% NaCl/methanol (1 :1, vol/vol), made to one phase by addition of methanol and dried under a stream of nitrogen. The residue was dissolved in 200 p1 chloroform/methanol/water (75:25:2, by vol) and aliquots removed to determine total phospholipid radioactivity. Individual phospholipid classes were separated by TLC on silica gel 60A (Whatman) in chloroform/methanol/NH.,OH (65:25:5 by vol; solvent system A). The plate was autoradiographed and the area of silica gel containing PEth was scraped off and countcd. The quantity of PEth formed was expressed as a percentage of total phospholipid radioactivity.
Preparation of standard PEth Authentic PEth was prepared by a procedure modified from that of Kobayashi and Kanfer (1 987). [32P]Phosphatidylcholinc was purified from a 32P-labelednerve lipid extract by TLC as described above and eluted from silica gel using chlorofonn/methanol/water (1 :1:0.2, by vol). The radioactive phospholipid was hydrolyzed overnight at 30°C in a reaction mixture composed of 0.1 M sodium acetate buffer, pH 5.6, 38 mM CaCI,, 0.5 mM sodium dodecyl sulfate, 2 M ethanol, and 0.25-1.0 mg of cabbage PLD (Boehringer-Mannheim) in a total volume of 1 ml. The resulting radioactive products, PEth, PA, and undigested phosphatidylcholine, were extractcd and separated by T I X as previously described. Alternatively, PEth was isolated by two-dimensional TLC, using solvent system A in the first dimension and chloroform/methanol/acetone/ghcial acetic acid/water (75: 15:30:15:7.5. by vol; solvent system B) as the solvent system in the second dimension. Radioactive lipids were located by autoradiography and eluted with chloroform/mcthanol/watcr ( 1 :1:0.2, by vol).
Enzymatic degradation of [32P]PEth Purified PEth was hydrolyzed overnight with Bacillus cereus phospholipase C (Sigma Chemical Co.) by the procedure of Mavis et al. ( 1 972). The phosphoethanol formed was separated by TLC on cellulose (MN 300; Analtech) in
PHOSPHOLIPASE D IN SCIATIC NERVE H,O/NH,OH (28%)/triehloroaceticacid (52335: 10:3.5, by vol; solvent system C) (Tettenborn and Muller, 1987).
Preparation and extraction of myelin Peripheral myelin was purified according to a modified procedure of Toews et al. (1987). All steps were performed at 4°C. Nerves were homogenized using a motor-driven glass-glass homogenizer in 2 ml of 0.32 M sucrose/25 mM Tris buffer, pH 8.5. The homogenatc derived from one nerve was placed on a 1.8-ml layer of 0.85 M sucrose/25 mM Tris buffer, pH 8.5, and centrifuged for 45 rnin at 82,000 g in a Beckman SW 60 Ti rotor. The myelinenriched layer floating on 0.85 Msucrose and the “nonmyelin” pellet at the bottom of the tube were recovered, diluted 20 times with 25 mM Tris buffer, pH 8.5, and resedimented for 15 rnin at 82.000 g. The resulting pellets were resuspended in Tris buffer and lipids were extracted as previously described except that 0.1 M HC1 was substituted for 0.9% NaCI. The washed lower phases were neutralized with 28% NH,OH before being dried. Lipids were redissolved in chloroform/methanol/H,O (75:25:2, by vol) and separated by TLC on silica gel in either solvent system A or in chlore form/methano1/40% methylamine/H,O (60:36:5:5, by vol).
Statistics Statistical analysis of data was performed by one-way analysis of variance followed by a Tukey-Kramer test. The number (n) ofeither incubations or rats used for these calculations is indicated in the legend to each figure.
RESULTS Incubation of 32P-prelabelednerve segments for 15 rnin in the presence of 100 mMethanol resulted in the formation of a radioactive phospholipid, which on TLC migrated above phosphatidylethanolamine in solvent system A (Fig. 1, lanes 1 and 2). When radioactive nerve lipids were separated by two-dimensional TLC, the unknown spot comigrated with authentic PEth prepared from phosphatidylcholine by cabbage PLD-catalyzed hydrolysis (data not shown). On phospholipase C-catalyzed degradation of purified putative PEth, a radioactive product was obtained - PEth
- PE - PC - Other Phospholipids - Origin 1
2
3
4
5
6
FIG. 1. Autoradiogram showing formation of [=P]PEth in sciatic nerve. Prelabeled nerve segments were incubated for 15 rnin in Krebs-Ringer bicarbonate medium containing glucose, loo mM ethand and no PDB (lanes 1 and-2).1 0 0 nM PDB (lanes 3 and 4). or 1 p M PDB (lanes 5 and 6). Lipids were extracted, separated by TLC in Solvent system A and autoradiographed as described in Materials and Methods. PC, phosphatidylcholine; PE, phosphatidylethanolamine.
1469
that migrated to the same position on cellulose TLC as phosphoethanol prepared from standard PEth (Fig. 2). The extent of basal PEth formation in nerve was less than 0.5% of radioactivity in total phospholipids extracted with neutral solvent, i.e., excluding polyphosphoinositides. PEth synthesis was stimulated in a dose-dependent fashion by PDB (Fig. 1, lanes 3-6, and Fig. 3 ) and was enhanced up to twofold at 1 p M PDB, the highest concentration used. PEth formation was not increased when 1 1 pM 4-a-phorbol, which does not activate protein kinase C, replaced PDB. In some experiments, PDB was added to amplify the signal. In the presence of this phorbol ester, PEth synthesis increased steadily as the concentration of ethanol in the incubation medium was varied from 100 mM to 2 M (Fig. 4). In most experiments, 100 mM (0.5%) ethanol was used to minimize possible perturbation of membrane structure. The time course of enhanced PEth formation in the presence of PDB showed a lag in that increased synthesis was not evident after 5 rnin (Fig. 5). In the experiment shown, PDB failed to elevate PEth levels significantly after 15 min, but in other experiments the phorbol ester did elicit a measurable stimulatory response after this time (cf. Fig. 6). PEth continued to accumulate slowly up to 60 rnin (Fig. 5). The stimulation of PEth generation by phorbol ester suggested the involvement of protein kinase C in this phenomenon. Consequently, the effect of Ro 3 18220 [described as compound 3 in Davis et al. (1989)], a novel, potent, and rather specific protein kinase C inhibitor, on PEth synthesis was examined (Fig. 6). In the presence of 100 nM PDB, 100 nMRo 3 1-8220 partially prevented enhanced PEth formation and l p M Ro 3 1-8220 was completely effective. When 1 pM PDB was present, 1 p M Ro 31-8220 tended to reduce the more pronounced stimulation of PEth synthesis, although the decrease failed to attain statistical significance. The inhibitor alone had no effect on basal PEth synthesis. Efforts to demonstrate increased PEth formation by the muscarinic cholinergic agonist, carbamylcholine, yielded inconsistent results. Up to a 60% rise in PEth generation, which was maximal after a 5-min incubation with 100 mM ethanol, was Seen in some experiments, but no stimulation was observed in others. The reason for this variability is not clear. Several attempts were made to identify the principal phospholipid precursor of PEth by isolation of the radioactive product and determination of its molecular species composition (Zhu and Eichberg, 1990h). This approach was adopted because the molecular species profile of phosphatidylcholine is readily distinguishable from that of phosphatidylethanolamine and all other glycerophospholipids in nerve (Zhu and Eichberg, unpublished results). However, starting with four labeled nerves incubated in the presence of 1 M ethanol and 1 pM PDB, a sufficient amount of J . Neurochem., Vol. 59, No. 4, 1992
B. J. L. EGGEN AND J. EICHBERG
1470
FIG. 2. Autoradiogram showing the liberation of [32P]phosphoethanol by enzymatic degradation of [32P]PEthformed in nerve. Putative r3'P]PEth was purified by two-dimensional TLC, hydrolyzed with 6. cereus phospholipase C, and the water-soluble products separated by TLC in solvent system C as described in Materials and Methods. Radioactive spots were visualized by autoradiography. Lane 1, inorganic phosphate; lane 2, hydrolysis product from nerve [=P]PEth; lane 3, authentic [32P]phosphoethand.
0.10
050
0.20
0.75
I S
1.m
200
Ethanol Concentration(M)
FIG. 4. Effect of varying ethand concentration on PEth formation. Preiabeled nerve segments were incubated for 15 min in the presence of 100 nM PDB and ethanol concentrations as shown. PEth synthesis is expressed as a mean percentage of total phospholipid radioactivity soluble in chloroform/methanol 2 SEM. Each value was derived from at least n = 6 nerve incubations from a minimum of two separate experiments for each condition.
purified PEth could not be isolated to permit molecuIar species determination. The ratio of radioactivity in phosphsiidylchohe to PEth was -4G: 1 under these incubation conditions and four sciatic nerves contain > 3 pmol of phosphatidylcholine. A complete molecular species analysis can be performed on I0 nmol of DAG enzymatically derived from phosphatidylcholine.,Thus, the failure to recover an appreciable mass of PEth suggests that the labeled product must be formed from a small, discrete pool of phosphatidylcholine. A similar argument can be advanced in regard to a precursor role for any of the other major nerve phospholipids. The subcellular distribution of PLD activity in nerve was examined by measurement of PEth fonnation in myelin-enriched and nonmyelin fractions prepared after prelabeled nerve segments were incubated for 15 min with 100 mM ethanol either in the presence or absence of 1 p M PDB. In these experiments,
both neutral and acidic solvents were used to extract all lipids, including polyphosphoinositides, from the fractions. The pattern of phospholipid labeling differed markedly in the two fractions. In myelin, polyphosphoinositides comprised 65-70% phosphatidylcholine 12- 1496, and phosphatidylethanolamine 2-3% of incorporated radioactivity, whereas in nonmyelin the corresponding values were 29-37%, 3042%, and 3-5%, respectively. When expressed as a percentage of total phospholipid radioactivity, myelin exhibited 50%of the PEth-forming activity of nonmyelin and formation of this product was significantly stimulated in the presence of PDB in nonmyelin, but not in myelin (Fig. 7, left panel). However, when PEth synthesis was calculated as a percentage of radioactivity in phosphatidylcholine + phosphatidylethanolamine, the most likely phospholipid precursors, myelin and nonmyelin, displayed essentially equal PLD activities (Fig. 7, right panel).
-
;8{+:I
9
NoPDB
.-n
.a
k
02-
0
L 0
0.1
0.3
1.0
4-6-PDB 0
FIG. 3. Effect of varying concentrations of PDB on PEth synthesis. Prelabeled nerve segments were incubated for 15 min in 100 mM ethanol and either no PDB (basal) or PDB concentrationsas shown. Results are expressed as a percentage of basal PEth formation f SEM and are derived from a total of at least n = 6 nerve incubations from a minimum of two separate experiments for each condition. Unless otherwise indicated, statistical analysis for data in this and subsequent figures was performed as described in Materials and Methods. "p 0.05; **p < 0.01; '"*p < 0.005.
J. Neurochem.. Vol. 59, No. 4, 1992
0.0
.
,
.
,
.
,
'
,
.
l
'
i
PHOSPHOI,IPASE D IN SCIATIC NERVE
1471
Ro 3 1-8220 concentration:
0 None 0 loonM
.a .-
lo
B
O8
0.
.
I
2
5
0.6
8
5
0.4
bR
02
No PDB
00
No PDB
0.1 ph4 PDB
IWPDB
FIG. 6. Inhibition of [32P]PEthformation by Ro 31-8220 (compound 3). Prelabeled nerves were incubated for 15 min with 100 nM ethanol, PDB concentrations as indicated, either in the absence or presence of Ro 31-8220. PEth synthesis is expressed as a percentageof total phospholipid radioactivity soluble in neutral solvent. Results were derived from at least n = 8 nerve incubations from a minimum of two separate experiments for each condition. ' p < 0.05; '*p c 0.005.
Indications that phosphoinositide metabolism is altered in nerve from diabetic animals prompted us to compare PLD activity in normal and diabctic nerve incubated in vitro. The formation of PEth in nerve from diabctic and age-matched normal rats incubated under basal conditions was very similar (Fig. 8). However, in the presence of 100 nMPDB, diabetic nerve exhibited a 74% stimulation of PLD activity, whereas PEth formation in normal tissue rose only 35%. At 1 pM PDB, PEth synthesis was further enhanced to 72% over basal in normal nerve, but increased just
'1 -
EQ
0.4
0.0
Myelin
Non Myelin
Myelin
Non Myelin
FIG. 7. Subcellular distribution of PLD activity in nerve. Nerve segments were prelabeledwith "P and then incubated for 15 min with 100 mM ethanol either in the absence (open bars) or presence (shaded bars) of 1 pM PDB as described in Materials and Methods. Left PEth synthesis is expressed as a percentage of total phospholipid radioactivity. Right: PEth synthesis is expressed as a percentage of the sum of radioactivity in phosphatidylcholine (PC) and phosphatidylethanolamine (PE). Results shown are the mean I SEM from two experiments. In one experiment, duplicate incubationsof nerves from two normal rats were performed. In the second experiment. one normal and one diabetic rat were used. Because all values were similar, they were combined to give a total of n = 4 incubations. Differences between PEth in the absence and presenceof PDB were statistically evaluated by means of Student's t test for myelin and nonmyelin fractions. *p < 0.05, different from no PDB.
0.I pM 4-R-PDB
1 ph4 4-R-PDB
FIG. 8. PDB stimulation of [=P]PEth formation in control and diabetic nerve. Prelabeled nerve segments from diabetic and agematched control rats were incubated for 15 min with 100 nM ethanol either in the absence or presence of PDB. Results are expressed as a percentageof basal PLD activity in normal nerve. The values for each condition were derived from three separate experiments and a total of seven to nine normal rats and six to eight diabetic rats. *p < 0.01 ; * p < 0.001.
slightly to 83% in diabetic nerve, so that the significant difference in response observed at the lower phorbol ester concentration was no longer seen. DISCUSSION
Investigations of PLD in animal tissues have usually relied on quantifying the release of radioactive choline or PA from labeled phospholipids or, alternatively, the formation of PEth from lipid substrate in the presence of ethanol. In our study, initial efforts to demonstrate the liberation of [3H]cholineon incubation of nerve segments prelabeled with this phospholipid precursor were not successful, possibly because of the high basal level of radioactive choline in the tissue. Thus, we turned to determination of PEth liberation from 32P-labelednerve phospholipids, an a p proach that is well suited to detect the relativcly small amounts of radioactive PEth formed. Our rcsults provide strong evidence that PLD is active in intact sciatic nerve scgments and that the enzyme can be stimulated by PDB. Efforts were also made to measure changes in [32P]PAformation in the presence compared with the absence of ethanol. Approximately 10% of isotope incorporated into prelabeled nerve phospholipids was in PA and, consequently, the magnitude of the transphosphatidylation response (ranging from 0.2 to 1.O% as a proportion of total phospholipid label) is such that any alterations in radioactive PA would have fallen within experimental error. Thus, our measurements suggest that the extent of PA formation by PLD action in nerve is very small. Further, it cannot be absolutely ruled out that in addition to PLD, other uncharacterized reactions could contribute to the formation of [32P]PEth. Several other studies have found that phorbol-12myristate- 13-acctate enhances PLD activity in nervous tissue preparations (Gustavsson and Alling, 1987; Gustavsson and Hansson, 1990; Martinson et J. Neurochem., Vol. 59, No. 4, 1992
1472
B. J. L. EGGEN AND J. EKHBERG
al., 1989)and the stimulation of the enzyme by active phorbol diesters is evidently a characteristic response in many tissues and cells (Billah and Anthes, 1990). The complete blockade of PDB-mediated stimulation by Ro 3 1-8220 at the lower concentration of phorbol ester used suggeststhat protein kinase C plays a role in enhancing PLD activity. This inhibitor has a structure based on those of staurosporine and K252a and has been reported to exhibit affinities for protein kinase C that are 1,700-fold and 150-fold greater than for calmodulin-dependent protein kinase and cyclic AMPdependent kinase, respectively (Davis et al., 1989). The inability of this substance to strongly inhibit at the higher PDB concentration is most likely due to an insufficient quantity being present. However, it remains possible that PDB in sufficient quantity interacts directly with PLD as well as protein kinase C to activate the enzyme. There is evidence that PLD a d vation can occur by both protein kinase C-dependent and -independent mechanisms (Billah and Anthes, 1990; Martinson et al., 1990). Although we were not able to identify the phospholipid source of PEth, if it is assumed that phosphatidylcholine is the primary precursor, then only a small amount of this lipid in nerve would be accessible to PLD. Subcellular fractionation of nerve after incubation with ethanol and PDB indicates that the location of this responsive pool as well as PLD is at least partly in myelin or closely associated structures. The tentative identification of PLD as a peripheral myelin constituent adds to the growing list of enzymes present in the myelin sheath (Ledeen, 1984) and strengthens accumulating evidence that myelin possesses much of the signal transduction machinery found in the plasma membrane, including muscarinic receptors, G proteins, phospholipase C , and adenylate cyclase (Larocca et al., 1987a,h; Golly et al., 1990). Basal PLD activity, as assessed by PEth formation, was the same in nerves from diabetic as from normal animals, but the enzyme was stimulated to a greater extent in diabetic nerve in the presence of a level of PDB that was suboptimal for normal tissue. In seeking to explain the significance of this difference, it is well known that Na+,K+-ATPase activity in diabetic nerve is diminished and can be corrected by protein kinase C activators, including phorbol esters and dioctanoylglycerol (Lattimer et al., 1989; Bianchi et al., 1991). The reduced level of endogenous DAG and/or arachidonyl-containing species of this neutral lipid in diabetic nerve (Zhu and Eichberg, 19906) may contribute to loss of Na+,K+-ATPaseactivity by interfering with a critical protein kinase C-dependent activation step. In this context, the greater response of PLD to PDB stimulation in the diabetic statc could reflect a heightened sensitivity of protein kinase C in the tissue to natural activators as a result of an adaptation to compensate for the decreased level of DAG. Consequently, PLD would be more readily stimulated in diabetic nerve, leading in turn to a greater J . Neurochem., Vol. 59, No. 4, I992
production of PA, a precursor of DAG. The possibility that PLD may play a role in the normal maintenance of Na+,K+-ATPaseactivity can by no means be excluded and it is attractive to speculate that in nerve of diabetic animals in vivo, the production of DAG by the PLD-mediated pathway may assume increased importance. Acknowledgment: This work was supported by NIH grants DK30577 and RR07147. We thank Dr. P. D. Davis of Roche Products, Ltd., for a generous gift of Ro 3 1-8220.
REFERENCES Bell M. E. and Eichberg J. (1985) Decreased incorporation of ['HIinositol and [3H]glycerol into glycerolipids of sciatic nerve from the streptozotocin diabetic rat. J. Neurochem. 45,465469. Bell M. E., Peterson R. G., and Eichberg J. (1982) Metabolism of polyphosphoinositides and other phospholipids in peripheral nerve from rats with chronic s~reptozotocin-induceddiabetes: increased turnover of phosphatidylinositol 4,5-bisphosphate. J. Neurochem. 39, 192-200. Besterman J. M., Duronio V., and Cuatrecasas P. (1986) Rapid formation of diacylglycerol from phosphatidylcholine:a pathway for generation of a second mcssenger. Proc. Natl. Acad. Sci. USA 83,6785-6789. Bianchi R., BedMattera L. N., Eichberg J., Triban C., and Fiori M. G. (1991) Studies in vitro and in vivo confirm ganglioside efficacy on Nat, K+-ATPaseactivity in experimental diabetic neuropathy, in Advances in i'hysiological Sciences (Manchanda s. K., Selvamurthy V., and Kumar N., ds), Macmillian India Ltd., New Delhi. Billah M. M. and Anthes J. C . (1990) The regulation and cellular functions of phosphatidylcholine hydrolysis. Biochem. J. 269, 28 1-29 I . Bligh E. A. and Dyer W. J. (1959) A rapid method of total lipid extraction and purification. Cm. J. Biochem. Physiol. 37, 911-917. Chattopadhyay J., Natarajan V., and Schmid H. H. 0. (1991) Membrane-associated phospholipase D activity in rat sciatic nerve. J. Neurochem. 57, 1429-1436. Davis P. D., Hill C. H ., Keech E., Lawton G., Nixon J. S., Sedgwick A. D., Wadsworth J., Westmacott D., and Wilkinson S. E. (1989) Potent selective inhibitors of protein kinase C. FEBS Lett. 259, 61-63. Dawson R. M. C. (1967) The formation of phosphatidylglycerol and other phospholipids by the transferase activity ofphospholipase D. Biochem J. 102, 205-210. Day N., Berti-Mattera L. N., and Eichberg J. (1991) Muscarinic cholincrgic receptor-mediated phosphoinositide metabolism in peripheral nerve. J. Neurochern. 56, 1905-1913. Dunlop M. E. and Larkins R. G. (1 990) Effects ofphosphatidic acid on islet cell phosphoinositide hydrolysis, Caz+and adenylate cyclax. Diubetes 38, 1187-1 192. Exton J. (19YO) Signalingthrough phosphatidylcholine breakdown. J. Biol. Chem. 265, 1-4. Golly F., Larocca J. N., and Ledeen R. W.(1990) Phosphoinositide breakdown in isolated myelin is stimulatcd by GTP analogs and calcium. J. Neurosci. Res. 27, 342-348. Gustavsson L. and Alling C. (1 987) Formation of phosphatidylethanol in rat brain by phospholipase D. Biochem. Biophys. Res. Commun. 142,958-963. Gustavsson L. and Hansson E. (1990) Stimulation of phospholipase D activity by phorbol esters in cultured astrocytes. J. Neurochem. 54,737-742. Hattori H. and Kanfer J. (1 985) Synaptosomal phospholipase D: a potential role in providing choline for acetylcholine synthesis. J. Neurochem. 45, 1578- 1584.
PHOSPHOLIPASE D IN SCIATIC NERVE Hothersall J. A. and McLean P. (1979) Effect ofexpenmental diabetes and insulin on phosphatidylinositol synthesis in rat sciatic nerve. Biochem. Biophys. Res. Commun. 88, 477-484. Knauss T. C., Jaffer F. E., and Abboud H. E. (1990) Phosphatidic acid mediates DNA synthesis, phospholipase C and plateletderived growth factor mRNAs in cultured mesangial cells. J. Bid. Chem. 265, 14457-14463. Kobayashi M. and Kanfer J. N. (I 987) Phosphatidylethanol formation via transphosphatidylation by rat brain synaptosomal phospholipase D.J. Neurochem. 48, 1597-1 603. KrollM. H.,ZovoicoG.B.,andSchaferA. L(1989)Secondmessenger function of phosphdtidic acid in platelet activation. J. Cell Phvsiol. 139, 558-564. Larocca J. N., Ledeen R. W., Dvorkin B., and Makman M. H. ( 1 9 8 7 ~Muscarinic ) receptor binding and muscarinic receptormediated inhibition of adenylate cyclase in rat brain myelin. J. Netirosci. 7, 3869-3876. Larocca J. N., Cervone A., and Ledeen R. W. (19876) Stimulation of phosphoinositide hydrolysis in myelin by muscannic agonists and potassium. Brain Res. 436,357-362. Lattimer S. A., Sima A. A. F., and Greene D. A. (1989) In vitro correction of Na+,K+-ATPasein diabetic nerve by protein kinase C agonists. Am. J. Physiol. 256, E264-E269. Ledeen R. W. (1984) Lipid-metabolizing enzymes of myelin and their relation to the axon, J. Lipid Res. 25, 1548-1554. Loffelholz K. (1989) Receptor regulation of choline phospholipid hydrolysis. Biochem. Pharmacol. 38, 1543-1 549. Lowery J. M., Bed-Mattera L. N., Day S.-F., Peterson R. G., and Eichberg J. (1989) Relationship of ATP turnover, polyphosphoinositide metabolism, and protein phosphorylation in sciatic nerve and derived peripheral myelin subfractions from normal and streptozotocin diabetic rats. J. Nezirochem. 52, 921-932. Martinson E. A., Goldstein D., and Brown J. H. (1 989) Muscarinic receptor activation of phosphatidylcholine hydrolysis. J. Biol. Chem. 264, 14748-14754. Martinson E. A,, Tnlivas l., and Brown J. H. (1990) Rapid protein kinase C-dependent activation of phospholipase D leads to delayed 1,2diglyceride accumulation. J. Biol. Chem. 265, 22282-22287. Mavis R. D., Bell R. M., and Vagelos P. R. (1 972) Effect of phos-
1473
pholipase C hydrolysis of membrane phospholipids on membranous enzymes. J. Biol. Chem. 247,2835-2841. Metz S. A. and Dunlop M. (1990) Stimulation of insulin release by phospholipase D a potential role for endogenousphosphatidic acid in pancreatic islet function. Biochem. J. 270,427-435. Moolenaar W. H., Kruijer W., Tilly B. C., Verlaan I., and de Laat S. W. (1986) Growth factor-like action of phosphatidic acid. Nature 323, 1 7 1 - 1 73. Murayarna T. and Ui M. (1987) Phosphatidic acid may stimulate membrane receptors mediating adenylate cyclase inhibition and phospholipid breakdown in 3T3 fibroblasts. J. Biol. Chem. 262,5522-5529. Qian Z . and Drewes L. R. (1989) Muscarinic acetylcholinereceptor regulates phosphatidylcholine phospholipase D in canine brain. J. Biol. Chem. 264,21720-21724. Rana R. S. and Hokin L. E. (1990) Role of phosphoinositides in transmembrane signaling. Physiol. Rev. 70, 1 15-163. Sandrnann J. and Wurtman R. J. (1991) Stimulation of phospholipase D activity in human neuroblastoma (LA-N-2) cells by activation of muscarinic acetylcholine receptors or by phorbol esters: relationship to phosphoinositide turnover. J. Neurochem. 56, 1312-1319. Taki T. and Kanfer J. N. (1979) Partial purification and properties of a rat brain phospholipaseD. J. Biol. Chem. 254,976 1-9765. Tettenborn C. S. and Mueller G. C. (1987) Phorbol esters activate the pathway for phosphatidylcthanol synthesis in differentiating HL-60 cells. Bicxhim. Biophys. Acta 931,242-250. Toews A. D., Fischer H. R., Goodrum J. F., Windes S., and Morel! P. (1987) Metabolism of phosphate and sulfate groups modifying the Po protein of peripheral nervous system myelin. J. Neurochem. 48,883-887. Wu C.-L., Tsai M.-H., and Stacey D. W. (1988) Cellular ras activity and phospholipid metabolism. Cell 52,68-71. Yang S. F., Freer S., and Benson A. A. (1967) Transphosphatidylation by phospholipase D. J. Biol. Chem. 242,477-484. Zhu X . and Eichberg J. ( 1 9 9 0 ~A ) pool of myeinositol needed for phosphatidylinositol synthesis is depleted in diabetic nerve. Proc. Natl. Acad. Sci. USA 87, 98 18-9822. Zhu X. and Eichberg J. (19904 1,2-Diacylglycerolcontent and its arachidonyl-containing molecular species are reduced in sciatic nerve from streptozotocin-induceddiabetic rats. J. Neurochem. 55, 1087-1090.
J. Neurochem.. Vol. 59, No. 4, 199.2
