The catabolism of phosphatidylinisitol by an EDTA-insensitive phospholipase A1 and calcium-dependent phosphatidylinositol phosphodiesterase in rat brain

1. A rat brain supernatant and microsomal fraction contained a phospholipase A1 enzyme which hydrolysed phosphatidylinositol at pH 8 in the absence of calcium. Triolein and phosphatidylcholine were not attacked under the same incubation conditions. 2. No evidence could be obtained for a phospholipase A2 in the microsomal preparation, and in the presence of Ca2+ the release of fatty acid observed was due to phosphatidylinositol phosphodiesterase followed by diacylglycerol lipase action. 3. Brain phosphatidylinositol phosphodiesterase showed extensive activity in the alkaline range (7-8.5) as well as at pH 5-5.5. The activity at higher pH values required higher calcium concentrations and disappeared on purification of the soluble enzyme by ammonium sulphate fractionation. 4. In general the ratio between inositol 1,2-(cyclic)phosphate and inositol 1-phosphate produced by phosphodiesterase action decreased with increasing pH.     

Protein-catalyzed exchange of phosphatidylinositol between rat brain microsomes and myelin.

Exchange of phosphatidylinositol and phosphatidylcholine between microsomal and myelin membranes has been demonstrated. This exchange is reversible and catalyzed by soluble proteins from the brain homogenate precipitated at pH 5.1. The extent of exchange of phosphatidylinositol from microsomal membrane to myelin is dependent upon pH and temperature, with an optimum around pH 7 and at 50 degrees C. Maximum exchange was observed at approximately equal amounts of microsomal, myelin, and supernatant proteins. The extent of the catalyzed exchange increases 4- to 8-fold upon using sonicated or heat-treated myelin as an acceptor membrane. Heating of microsomal membranes results in no change. The extent of catalyzed exchange of phosphatidylcholine is less than that of the phosphatidylinositol. The exchange of other phospholipids and glycolipids between microsomal and myelin membranes cannot be demonstrated. The catalytic activity of the pH 5.1 supernatant proteins in rat brain for the exchange of phosphatidylinositol increases with age after birth and reaches a maximum around 21 days of age analogous to the process of myelination. The pH 5.1 supernatant proteins from quaking and jimpy mutant mice has normal catalytic activity.     

The diphosphoinositide kinase of rat brain

1. The supernatant fraction of adult rat brain contains a diphosphoinositide kinase. 2. Formation of triphosphoinositide by the enzyme in the presence of ATP and Mg(2+) ions was shown with labelled ATP or labelled diphosphoinositide. 3. The kinase was also activated by Ca(2+), Mn(2+) and Co(2+) ions, but to a smaller extent than by Mg(2+) ions. 4. In the presence of optimum Mg(2+) ion concentration the enzyme was inhibited by Ca(2+) ions. 5. Activity did not depend on thiol groups and the pH optimum was 7.3. 6. The dialysed supernatant fraction had no diglyceride kinase activity and negligible phosphatidylinositol kinase activity. 7. Triphosphoinositide phosphomonoesterase was present but showed little activity under the conditions used to assay the kinase. 8. Diphosphoinositide kinase was purified by ammonium sulphate fractionation, ethanol treatment and chromatography on Sephadex G-200. 9. This purification removed much of the triphosphoinositide phosphomonoesterase.     

Characterization of phospholipase C-mediated phosphatidylinositol degradation in rat heart ventricle

Phosphoinositide-specific phospholipase C (PI-PLC) activity was investigated in the rat heart ventricle. Incubation of ventricle homogenate or 100,000g supernatant fraction with [3H]myoinositol or [3H]arachidonate-labeled phosphatidylinositol in the presence of Ca2+ resulted in a decrease in phosphatidylinositol with a concomitant increase in water-soluble [3H]inositol phosphate or [3H]diglyceride, respectively. Total overt homogenate PI-PLC activity could be accounted for in the supernatant fraction. Neutral, zwitterionic, cationic, or anionic detergents did not unmask membrane-associated activity. While cytosolic phospholipase C was active against a pure phosphatidylinositol substrate in the presence of Ca2+, no hydrolytic activity was detected when phosphatidylinositol was presented as a component (4-5%) of a mixture of phospholipids. However, addition of deoxycholate to the incubation mixture (pH 6.5, Ca2+ 10(-3) M) containing mixed phospholipids resulted in the exclusive hydrolysis of inositol phospholipids. Ventricular supernatant phospholipase C-mediated phosphatidylinositol degradation has a sharp pH optimum at 5.5 and a specific requirement for Ca2+. Activity is maximal at 1 to 2 X 10(-3) M Ca2+, with inhibition occurring at higher levels. Under optimized conditions phosphatidylinositol is hydrolyzed at a rate of 20-25 nmol/min/mg protein. Multivalent cations inhibit Ca2+-dependent PI-PLC activity while monovalent cations and anions have no effect. There is no apparent selectivity for specific fatty acid moieties on phosphatidylinositol. Soluble PI-PLC is inhibited by sulfhydryl reagents, neomycin, mepacrine, trifluoperazine, and propranolol. Chlorpromazine, dibucaine, and tetracaine exert a biphasic influence, stimulating at lower and inhibiting at higher concentrations.     

THE EXCHANGE OF PHOSPHATIDYLCHOLINE AND OF PHOSPHATIDYLINOSITOL IN SHEEP BRAIN

—The exchange of phospholipids between liposomes and brain mitochondria has been studied in the presence of pH 5·1 supernatant fluids derived from rat, guinea pig, sheep and ox brains. The exchange phenomenon was similar to that observed in liver and heart, but phosphatidylinositol and not phosphatidylcholine was the most rapidly exchanging phospholipid. The phosphatidylcholine exchange activity was purified 186-fold from sheep brain and the protein fraction contained two major and several minor protein species. The phosphatidylcholine and phosphatidylinositol exchange activities have been shown to have very similar molecular weights and isoelectric points. However, their behaviour in response to changes in liposomal surface charge suggested that separate proteins might be involved in stimulating the exchange of the two phospholipid classes.     

SUBCELLULAR DISTRIBUTION OF 5-HYDROXYTRYPTAMINE IN RAT BRAIN DURING DEVELOPMENT: EFFECT OF DRUGS AND FASTING

Abstract— Distribution of brain 5-HT content between the high-speed supernatant and particulate fractions under normal and experimental conditions was studied in postnatal and adult rats. In adult and 35-day-old rats the 5-HT content of the supernatant fraction was about 25% of that of the total homogenate and significantly higher than that in 1, 7 and 21-day-old rats. In 1-day-old rats fasting caused an increase of 100% in the turnover, 50% in the content and no alteration in the subcellular distribution of brain 5-HT, which suggests that under normal conditions 5-HT stores may be filled near to capacity. After 5-hydroxytryptophan administration, the 5-HT content of the adult rat brain increased 3-fold and that of the supernatant fraction to 35% of 5-HT content of the total homogenate. In postnatal rats, the brain 5-HT content rose to an adult level and the supernatant 5-HT percentage to a markedly higher than adult level, probably because of the known higher than adult 5-hydroxytryptophan decarboxylase activity of brain capillaries. Administration of tranylcypromine to adult rats caused a 2.6-fold increase of brain 5-HT content and a slight increase of the supernatant 5-HT percentage. At various times after the administration of the MAO inhibitors (tranylcypromine or pargyline) and fasting to the 1-day-old rats, brain 5-HT content increased 4, 5 and 7-fold, respectively, and the supernatant 5-HT rose consistently and, as in the adult, to about 30% of the 5-HT content of the total homogenate. After pargyline following reserpine pretreatment, the 5-HT content of the adult and 1-day-old rat brain increased to 2–3 times the control level and that of the supernatant fraction to about 40% of the 5-HT content of the total homogenate. The adult values for 5-HT in the particulate fraction of the 1-day-old rats after the drug treatments are in sharp contradiction to the low endogenous 5-HT content and known lack of nerve terminals and synaptic vesicles in their brains, and suggest that after MAO inhibition brain 5-HT neurons may bind the amine by some other mechanism than the Mg²⁺-ATP-dependent, reserpine-sensitive granular storage.     

The protein-mediated net transfer of phosphatidylinositol in model systems

The phospholipid monolayer technique has been used to study the transfer activity of the phospholipid exchange protein from beef brain. In measuring the transfer between a monolayer consisting of equimolar amounts of phosphatidylcholine and phosphatidylinositol and liposomes consisting of 98 mol% phosphatidylcholine and 2 mol% phosphatidylinositol, the beef brain protein demonstrates an 8-fold higher transfer activity for phosphatidylinositol than for phosphatidylcholine. Under similar conditions the phosphatidylcholine exchange protein from beef liver showed a great preference for phosphatidylcholine. Phosphatidylcholine liposomes devoid of phosphatidylinositol still functioned as receptors of phosphatidylinositol when the beef brain exchange protein was present. This indicates that this protein can catalyse a net transfer of phosphatidylinopsitol. Binding of both phosphatidylinositol and phosphatidylcholine to the beef brain protein was shown.     

Incorporation of injected [32P]phosphate into the phosphoinositides of subcellular fractions from young rat brain

1. Six subcellular fractions were prepared from the brain of 3-week-old rats after the injection of [(32)P]orthophosphate into the subarachnoidal space. The fractions were characterized by chemical and enzymic determinations and by electron microscopy. 2. The highest concentration of phosphoinositides was found in the microsomal fraction. 3. After an exchange period of 4hr. in vivo, the specific radioactivity of phosphatidylinositol was high in the nuclear, mitochondrial and supernatant fractions, and that of diphosphoinositide was high in the nuclear, microsomal and supernatant fractions. Triphosphoinositide specific activity was highest in the myelin fraction. 4. Specific activities (counts/100sec./mug. of P) were in the following order for all fractions except myelin: diphosphoinositide>phosphatidylinositol>triphosphoinositide. For myelin the order was: diphosphoinositide>triphosphoinositide>phosphatidylinositol. 5. Phosphatidylinositol extracted from a tissue fraction by acidified chloroform-methanol had roughly twice the specific activity of that extracted by neutral solvents. The same applied to diphosphoinositide.     

A Low-Km 5'-Nucleotidase from Rat Brain Cytosolic Fraction: Purification, Kinetic Properties, and Description of Regulation by a Novel Factor that Increases Sensitivity to Inhibition by ATP and ADP

Abstract: A readily soluble 5'-nucleotidase was purified 1,800-fold from rat brain 105,000- g supernatant. The enzyme showed similarity to the 5'-nucleotidase ectoenzyme of plasma membranes. It exhibited a low K _m for AMP, which was preferred over IMP as substrate. It was inhibited by free ATP and ADP and by α,β-methylene ADP. The enzyme appeared to be a glycoprotein on the basis of its interaction with concanavalin A. It contained a phosphatidylinositol moiety because treatment with phosphatidylinositol-specific phospholipase C increased its hydrophilicity. A single subunit of M_r = 54,300 ± 800 was observed, which is appreciably smaller than published values for the 5'-nucleotidase ectoenzyme or for other low- K _m"soluble" 5'-nucleotidases. The soluble 5'-nucleotidase showed an elution profile on AMP-Sepharose affinity chromatography or on Mono Q ion-exchange chromatography different from that of the brain ectoenzyme. Forty-two percent of the soluble 5'-nucleotidase in brain 105,000- g supernatant did not bind to a Mono Q ion-exchange column because of its interaction with a soluble factor. This factor could be removed by chromatography on concanavalin A-Sepharose. The factor had the novel property of increasing the sensitivity of the purified soluble 5'-nucleotidase toward the inhibitor ATP by 20-fold. This factor was also able to increase the inhibition of brain 5'-nucleotidase ectoenzyme by ATP.     

The effect of various PGI2 analogues on free and phospholipid-bound fatty acids in rat brain, following complete ischaemia, normobaric and hypobaric hypoxia.

The effect of various prostacyclin analogues on free and phospholipid-bound fatty acids in rat brain following complete ischaemia, normobaric and hypobaric hypoxia were investigated. Both ischaemic and hypoxic conditions cause a significant rise in both unsaturated (arachidonic and docosahexaenoic) and saturated (palmitic and stearic) fatty acids. Under same conditions, a significant fall in phosphatidylinositol level in brain was observed. Prostacyclin and its analogues significantly reduce the rise of free fatty acids, as well as the decrease of phosphatidylinositol.     

Phospholipid transfer activities in toad oocytes and developing embryos

The role of lipid transfer proteins during plasma membrane biogenesis was explored. Developing amphibia embryos were used because during their growth an active plasma membrane biosynthesis occurs together with negligible mitochondrial and endoplasmic reticulum proliferation. Sonicated vesicles, containing 14C-labeled phospholipids and 3H-labeled triolein, as donor particles and cross-linked erythrocyte ghosts as acceptor particles were used to measure phospholipid transfer activities in unfertilized oocytes and in developing embryos of the toad Bufo arenarum. Phosphatidylcholine transfer activity in pH 5.1 supernatant of unfertilized oocytes was 8-fold higher than the activity found in female toad liver supernatant, but dropped steadily after fertilization. After 20 hr of development, at the stage of late blastula, the phosphatidylcholine transfer activity had dropped 4-fold. Unfertilized oocyte supernatant exhibited phosphatidylinositol and phosphatidylethanolamine transfer activity also, but at the late blastula stage the former had dropped 18-fold and the latter was no longer detectable under our assay conditions. Our results show that fertilization does not trigger a phospholipid transport process catalyzed by lipid transfer proteins. Moreover, they imply that 75% of the phosphatidylcholine transfer activity and more than 95% of the phosphatidylinositol and phosphatidylethanolamine transfer activities present in pH 5.1 supernatants of unfertilized oocytes may not be essential for toad embryo development. Our findings do not rule out, however, that a phosphatidylcholine-specific lipid transfer protein could be required for embryo early growth.     

A Possible Pathway of Phosphoinositide Metabolism Through EDTA-Insensitive Phospholipase A1 Followed by Lysophosphoinositide-Specific Phospholipase C in Rat Brain

Abstract— Incubation of [2-³H]glycerol-labeled phosphatidylinositol with a crude cytosol fraction of rat brain in the presence of EDTA yielded [³H]lysophosphatidylinositol predominantly without accumulation of labeled monoacylglycerol and diacylglycerol. The pH optimum of this Phospholipase A activity was 8.0. The activity for phosphatidylinositol was twofold higher than for phosphatidylethanolamine, whereas phosphatidylcholine, phosphatidylserine, and phosphatidic acid were not hydrolyzed significantly under the conditions used. The phospholipase A activity for phosphatidylethanolamine was resolved in part from that for phosphatidylinositol by ammonium sulfate fractionation of the cytosol, indicating the existence of at least two forms of EDTA-insensitive phospholipase A. The positional specificity of the phosphatidylinositol-hydrolyzing activity was found to be that of a phospholipase A₁, as radioactive lysophosphatidylinositol was produced from 1 -stearoyl-2-[1-¹⁴C]arachidonyl- sn -glycero-3-phosphoinositol without release of free arachidonate. A phospholipase C activity specific for lysophosphoinositides was found in a membrane fraction from rat brain, which was similar to that characterized in porcine platelets. The phospholipase C was demonstrated to hydrolyze the 2-acyl isomer as well as the 1-acyl isomer of lysophosphatidylinositol. Taken together, our results suggest a possible pathway through which phosphatidylinositol is selectively degraded to the 2-acyl isomer of lysophosphatidylinositol in a Ca²⁺-independeht manner, and subsequently converted to 2-monoacylglycerol in rat brain.     

Lysophosphoinositide-specific phospholipase C in rat brain synaptic plasma membranes

A membrane preparation from rat brain catalyzed the hydrolysis of [2-³H]glycerol-labeled lysophosphatidylinositol (lysoPI) to yield monoacylglycerol (MG) and inositolphosphates. This phospholipase C activity had an optimal pH of 8.2. The membrane preparation did not require the addition of Ca²⁺ for its maximum activity, but the activity was inhibited by addition of 0.1 mM EDTA to the assay mixture and was restored by simultaneous addition of 0.2 mM Ca²⁺. The activity was found to be localized in synaptic plasma membranes prepared by Ficoll and Percoll density gradients. The phospholipase C was highly specific for lysoPI; diacylglycerol formation from phosphatidylinositol, and MG formation from lysophosphatidylcholine, lysophosphatidylethanolamine, and lysophosphatidylserine were below 5% of that observed with lysoPI under the conditions used. We concluded that there is a pathway for phosphatidylinositol metabolism in brain synaptic membranes which is different from the well-characterized phosphoinositide-specific phospholipase C pathway.Abbreviations PI phosphatidylinositol - lysoPI lysophosphatidylinositol - lysoPI-PLC lysophosphoinositide-specific phospholipase C - PI-PLC phosphoinositide-specific phospholipase C - MG monoacylglycerol - PLC phospholipase C To whom to address reprint requests.     

Purification of phosphoinositide-specific phospholipase C from a particulate fraction of bovine brain

The coupling of various agonist receptors to the hydrolysis of phosphoinositides has generated much interest in the nature of the phospholipase C that is activated. Here we report the purification of a bovine brain phospholipase C derived from the particulate fraction. A 1000-fold purification was achieved by a combination of heparin-Sepharose, DEAE-cellulose and gel-permeation chromatography. The purified enzyme appears to be monomeric and under denaturing conditions shows a single staining major polypeptide of molecular mass 154 kDa in SDS gels. The enzyme is specific for phosphoinositides although it shows a marked preference for the polyphosphoinositides. With phosphatidylinositol 4,5-bisphosphate as substrate the enzyme expresses a specific activity of greater than 100 mumol min-1 mg-1. The phospholipase C is activated by Ca2+ (0.1-10 microM). The behaviour of this particulate enzyme is discussed in the context of a agonist-induced phosphatidylinositol hydrolysis.     

Inositol phospholipid levels of rat forebrain obtained by freeze-blowing method

Four different techniques of handling rat brain prior to lipid extraction and assay were tested to investigate the levels of inositol phospholipids in the brain. In these four techniques, the rat forebrains were either (1) freeze-blown followed by being preserved in liquid N2, (2) subjected to microwave irradiation prior to decapitation, (3) removed and frozen in liquid N2, or (4) removed at room temperature and subjected to lipid extraction as rapidly as possible. There was little change in phosphatidylinositol levels under any of these conditions; however, higher levels of phosphatidylinositol 4-phosphate were observed in freeze-blown and microwave-irradiated samples compared to the other samples. Even more striking differences were seen in phosphatidylinositol 4,5-bisphosphate fractions. The highest level of this lipid, 763 +/- 39 nmol/g tissue, which was obtained from freeze-blown samples, was more than 2-fold higher than that of the lowest values which were obtained by extraction without prior inactivation. These results indicate that the values of phosphatidylinositol 4,5-bisphosphate in brain in situ are higher than those generally reported, and that the freeze-blowing method has an advantage for further investigation of inositol phospholipid metabolism in brain due to the rapid breakdown of these compounds.     

Phospholipase C activity in palate mesenchyme cells: calcium and pH requirements, substrate specificity, and subcellular localization

Primary cultures of mouse embryo palate mesenchyme cells were incubated with [3H]arachidonic acid and [14C]stearic acid in order to radiolabel their lipids. The cells were then washed, collected by centrifugation, and homogenized. Incubation of the homogenates under various conditions revealed that deoxycholate inhibited phospholipase A activity and stimulated a phospholipase C activity in these cells which preferentially degraded phosphatidylinositol (PI) compared to phosphatidylcholine (PC), -ethanolamine (PE), and -serine (PS). Expression of this phospholipase C (E.C. 3.1.4.10) activity was dependent on Ca2+ and had a pH optimum of no more than 7.0-7.5. Centrifugation of the homogenates at 105,000g for 30 min produced a membranous fraction that contained phospholipase C activity with characteristics similar to those of the enzyme found in the supernatant. Such a dual distribution of this enzyme may reflect that mouse embryo palate mesenchyme cells are neural crest in origin.     

Solubilization of the Membrane-Bound Sialidase from Pig Brain by Treatment with Bacterial Phosphatidylinositol Phospholipase C

The total pellet from pig forebrain (from which the cytosolic sialidase was completely washed out) was treated with phosphatidylinositol phospholipase C (PIPLC) and centrifuged at high speed. The supernatant contained sialidase and 5'-nucleotidase activities. The greatest liberation of sialidase was obtained after incubation for 20 min with PIPLC at 37 degrees C using pH 6.0 and a ratio between PIPLC (as units) and protein of 1.6. Under these conditions, the release of sialidase, 5'-nucleotidase, and protein was 22, 50, and 18.5%, respectively. On treatment with PIPLC, a purified preparation of pig brain neuronal (synaptosomal) membranes released 28% of its sialidase whereas a purified preparation of pig brain lysosomes did not liberate any sialidase activity. The pH optimum of sialidase present in the supernatant obtained after PIPLC treatment of the total pellet was 4.2, the same as that of the enzyme embedded in the membrane. When this supernatant was subjected to ammonium sulfate fractionation, 88% of its sialidase, having a pH optimum of 4.2, was recovered in the fraction precipitated between 20 and 45% of salt saturation and subsequently dialyzed. Ammonium sulfate treatment caused the appearance of a second sialidase activity, having a pH optimum of 6.6 and behaving on fractionation similarly to the pH 4.2 sialidase. The Km and Vmax values of pH 4.2 and pH 6.6 sialidase were similar (1.48 x 10(-4) and 0.98 x 10(-4) M for Km and 1.6 and 1.4 mU/mg of protein for Vmax, respectively), whereas the stability on standing at 4 degrees C or exposure to freezing and thawing cycles was greater for pH 4.2 sialidase.(ABSTRACT TRUNCATED AT 250 WORDS)     

The metabolism of phosphatidylinositol in the thyroid gland of the pig

1. The metabolism of phosphatidylinositol in pig thyroid has been investigated as a basis for understanding the specific stimulation of the synthesis of this phospholipid in the gland by thyrotropin. 2. The gland contained an active Ca(2+)-dependent phosphatidylinositol-splitting enzyme with an optimum pH of 5.3-5.5. 3. The major water-soluble product (65%) formed by this catabolic enzyme was not phosphorylinositol but a related compound, which may be a cyclic phosphorylinositol. Both this and phosphorylinositol (35%) were released simultaneously from the phosphatidylinositol substrate. 4. The phosphatidylinositol-splitting enzyme was found almost exclusively in the supernatant fraction obtained by homogenization of the gland. It was not present in the acid-phosphatase-containing particulate fraction. 5. The incorporation of [2-(3)H(1)]inositol into phosphatidylinositol in the presence of either CDP-diglyceride or CTP+ATP was most active in the microsomal fraction. 6. When thyroidal microsomes were labelled with [(3)H]inositol and (32)P, and then incubated with unlabelled inositol, there was a dramatic loss of (3)H labelling from the phosphatidylinositol, which was not accompanied by an equivalent loss of (32)P from the phosphate moiety. This turnover of the inositol moiety required nucleotide coenzymes. It is postulated that the phosphatidylinositol is split into inositol and a phosphorus-containing lipid precursor of the phospholipid that remains on the microsomal membrane and is recycled. 7. Isolated thyroidal mitochondria synthesized phosphatidylinositol from [2-(3)H(1)]inositol only because of their contaminating microsomal component. 8. Some evidence has been obtained of a rapid transfer of phosphatidylinositol molecules from thyroidal microsomes to mitochondria when these were incubated together in the presence of a supernatant fraction. 9. Both phosphatidylinositol breakdown by the supernatant fraction of the gland and synthesis by the microsomes were totally inhibited by 1mm-chlorpromazine. This drug is known to suppress thyrotrophin-induced stimulation of activity in thyroid slices.     

Polyamines stimulate the phosphorylation of phosphatidylinositol in membranes from A431 cells

The phosphorylation of phosphatidylinositol in plasma membranes from A431 cells was investigated using [gamma-32P]ATP as the substrate. Phosphatidylinositol 4-phosphate was found to be the major product after an incubation time of 5-10 min. Little, if any, phosphatidylinositol 4,5-bisphosphate was found under these conditions. Epidermal growth factor (EGF) had no effect on the formation of phosphatidylinositol 4-phosphate or phosphatidylinositol 4,5-bisphosphate. On the other hand, the polyamines spermidine and spermine stimulated the phosphatidylinositol kinase activity about eightfold yielding almost exclusively phosphatidylinositol 4-phosphate as the reaction product. Half-maximum stimulation by spermidine occurred under near physiological conditions (1.5 mM). Furthermore various proteins and amino acid polymers containing clustered basic amino acid residues (e.g. histones and polylysine) stimulated the formation of phosphatidylinositol 4-phosphate to a similar extent. Half-maximal concentrations for the activation were considerably lower ranging from 1.5 microM to 80 microM. The ATP specificity of the phosphatidylinositol kinase(s) was investigated with a small set of selected ATP derivatives. In the presence of spermidine the specificity changed significantly indicating that (a) spermidine acts on a kinase and not on a phosphatase, (b) this activity is distinct from the EGF-receptor protein kinase activity. The results do not suggest an involvement of the EGF receptor in the growth-factor-dependent formation of phosphatidylinositol phosphates. It is proposed that the phosphorylation of phosphatidylinositol by polyamines might be a mechanism to replenish the pool of inositolphospholipids.