Phosphatidylinositol 3-kinase: inhibition of intrinsic protein-serine kinase activity by phosphoinositides, and of lipid kinase activity by Mn2+.

Phosphatidylinositol (PI) 3-kinase is composed of 110 kDa catalytic and 85 kDa regulatory subunits. The 110 kDa subunit has two intrinsic kinase activities, i.e., Mn(2+)-dependent protein-serine kinase and Mg(2+)-dependent lipid kinase activities. These intrinsic kinases have been reported to be interdependent: protein-serine kinase phosphorylates the 85 kDa subunit of PI 3-kinase, which upon phosphorylation inhibits the lipid kinase activity of PI 3-kinase. We report here that phosphoinositides can selectively inhibit the protein-serine kinase activity of PI 3-kinase without affecting lipid kinase activity. This inhibition depends on the phosphorylation status of the phosphoinositides, i.e., PI 4,5-bisphosphate > PI 4-phosphate >> PI. Mn2+ (2 mM) protected protein kinase activity from phosphoinositides-mediated inhibition if added prior to interaction of PI 3-kinase with phosphoinositides. On the other hand, Mn2+ (2 mM) inhibited lipid kinase activity independent of its effect on the protein kinase activity of PI 3-kinase. The present study suggests that the protein-serine kinase and the lipid kinase activities of PI 3-kinase can be selectively inhibited by phosphoinositides and Mn2+ respectively.

Differential effects of spermine on phosphatidylinositol 3-kinase and phosphatidylinositol phosphate 5-kinase.

The metabolism of phosphoinositides plays an important role in the signal transduction pathways. We report here that naturally occurring polyamines affect the activities of phosphatidylinositol (PI) 3-kinase and PI 4-phosphate (PIP) 5-kinase differently. While polyamines inhibited the PI 3-kinase activity, they stimulated the activity of PIP 5-kinase in the order of spermine > spermidine > putrescine. Spermine inhibited the PI 3-kinase activity in a concentration-dependent manner with an IC50 of 100 microM. On the other hand, spermine (5 mM) stimulated the activity of PIP 5-kinase 2-3 fold. Kinetic studies of spermine-mediated inhibition of PI 3-kinase revealed that it was noncompetitive with respect to ATP. The effect of Mg2+ and PIP2 concentration on kinase activity was sigmoidal, with spermine inhibiting PI 3-kinase activity at all PIP2 concentrations. While 1 mM calcium stimulated PI 3-kinase activity at submaximal concentrations of Mg2+ (1.25 mM), inhibition was observed at optimal concentration of Mg2+ (2 mM). We propose that spermine may modulate the cellular signal by virtue of its differential effects on phosphoinositide kinases.

Interaction of protein kinase C and phosphoinositides: regulation by polyamines.

Phosphatidylinositol 4,5-bisphosphate (PIP2) activates protein kinase C (PKC) in the presence of phosphatidylserine and calcium. Recently it has been demonstrated that direct interaction of PKC with PIP2 in the absence of divalent cation inactivates this kinase. In the present study, the interaction of natural aliphatic polyamines with phosphoinositides was investigated for its possible relevance to PKC-mediated protein phosphorylation. PKC/phosphoinositide interaction was studied by monitoring the changes in (a) intrinsic fluorescence of the enzyme, and (b) PKC activity (protamine sulphate or histone III-S as substrate). All the phosphoinositides: PIP2, phosphatidylinositol 4-phosphate (PIP) and phosphatidylinositol (PI) inactivated PKC with an IC50 of 0.4 microM for PIP2, 5 microM for PIP and 10 microM for PI. Hydrogenated PIP2 behaved similarly to that of natural PIP2. Time-dependent studies showed very rapid inactivation of PKC by PIP2. The polyamines spermine and spermidine at physiological concentrations protected PKC from phosphoinositides-mediated inactivation when added prior to PKC interaction with phosphoinositides. Putrescine was least effective. Addition of spermine or spermidine to PKC/phosphoinositides incubation mixture did not reverse PKC activity indicating that the inactivation of PKC by phosphoinositides is irreversible. Fluorescence quenching experiments showed that phosphoinositides inactivate PKC by inducing conformational changes of the enzyme that are prevented by spermine. We propose that polyamines protect PKC and possibly other protein kinase from phosphoinositides-mediated inactivation, and that inactivation of protein kinases by phosphoinositides may not have physiological relevance.

Amyloid beta-protein stimulates casein kinase I and casein kinase II activities.

Amyloid beta-protein (A beta) is the major protein of cerebrovascular and plaque amyloid in Alzheimer's disease (AD). Extensive evidence has demonstrated abnormal protein phosphorylation in this disease. We investigated the effect of synthetic A beta with the amino-acid sequence corresponding to cerebrovascular A beta and plaque A beta on the activities of casein kinase I (CK I) and casein kinase II (CK II). These enzymes were purified from bovine brain and casein was used as a substrate. A beta was found to stimulate markedly CK I- and CK II-mediated phosphorylation of casein in a concentration-dependent manner. The effect of plaque A beta was considerably higher than that of cerebrovascular A beta. Heparin, which is known to be a specific inhibitor of CK II, completely inhibited A beta-stimulated CK II activity. A beta itself was not a substrate for casein kinases. These findings were confirmed using other substrates for CK I and CK II. The experiments with synthetic CK II-substrate peptide (Leu-Glu-Leu-Ser-Asp-Asp-Asp-Asp-Glu) and the phosphorylation of erythrocyte membrane proteins by intrinsic membrane-bound CK I in erythrocytes showed marked stimulation in activities of casein kinases in the presence of A beta 1-40 or blocked A beta. We propose that A beta, by stimulating casein kinases, may contribute to abnormal protein phosphorylation in AD, in particular to increased phosphorylation of microtubule-associated proteins, leading to the neurofibrillary tangles formation and neurodegeneration in this disease. Interaction of A beta with protein kinases, thus, may characterize the beginning of the disease.

Activation of protein kinase C by phosphatidylinositol 3,4,5-trisphosphate.

Phosphatidylinositol 3-kinase (PI 3-kinase) was partially purified from rat liver cytosol and used to synthesize phosphatidylinositol 3,4,5-trisphosphate (PIP3), using phosphatidylinositol 4,5-bisphosphate (PIP2) as a substrate. Purified PIP3 (free of chromatographic oxalate) activated protein kinase C (PKC) in the presence of phosphatidylserine and calcium (PKC -cofactors) in a concentration-dependent manner. In the absence of these cofactors, effect of PIP3 was not observed. Comparison of the effects of PIP3 and PIP2 on PKC activity indicates that PIP3 is a more potent PKC-activator than PIP2. The affinity of PKC to PIP3 was 4 fold higher than that to PIP2 (KPIP3 = 0.022 and KPIP2 = 0.087 mol %), while its maximal velocity (Vmax) was similar to that of PIP2-stimulated PKC activity (0.4 - 0.5 mumol/mg/min). These results suggest a physiological role for PIP3 in signal transduction, and support the previous finding (Chauhan et al. (1991) Arch. Biochem. Biophys. 287,283) that PKC-activation by phosphoinositides increases with the state of phosphorylation of these lipids. We propose that PIP3 by activating PKC may initiate a cascade of events from PIP3-->PKC- activation-->effects on other protein kinases such as MAP-kinase-->gene expression.

Magnesium protects phosphatidylinositol-4,5-bisphosphate-mediated inactivation of casein kinase I in erythrocyte membrane.

Recent reports suggest that membrane-bound casein kinase I (MBCK I) activity in erythrocytes is inactivated by exogenously added phosphatidylinositol 4,5-bisphosphate (PIP2) (Bazenet et al. (1990) J. Biol. Chem. 265, 7369-7376; Brockman and Anderson (1991) J. Biol. Chem. 266, 2508-2512). Here we report that PIP2-mediated inhibition of MBCK I in erythrocytes is only observed if exogenous PIP2 and the kinase are allowed to interact in the absence of Mg2+. Prior incubation of PIP2 with 1 mM Mg2+ prevents the inactivation of MBCK I by PIP2. Other divalent cations (Ni2+, Co2+, Mn2+, Cd2+, Ca2+) and trivalent metal ions (La3+, Cr3+, Al3+) did not protect MBCK I from PIP2-mediated inactivation, indicating that the protective effect is specific for Mg2+ only. We propose a role of Mg2+ in the interaction of CK I with phosphoinositides, and that PIP2-mediated inhibition of protein kinase(s) may be a non-physiological phenomenon.

Barbiturates inhibit intracellular Ca2+ rise induced by thrombin in rat platelets.

The effects of a number of barbiturates (anesthetic as well as anticonvulsant) on thrombin-induced calcium mobilization were tested in rat platelets using the fluorescent Ca2+ probe Fura-2. All drugs, except barbituric acid and Na-barbital, inhibited the thrombin-induced intracellular Ca2+ rise. Both the uptake of extracellular Ca2+ and the release of calcium from intracellular organelles were affected but influx was inhibited more strongly and at lower concentrations of the drugs (e.g. IC50 of thiopental was 0.83 mM for influx and 1.2 mM for intracellular release). Inhibitory potencies of the various barbiturates were markedly different. Thiopental was the most and barbital the least potent inhibitor. The order of inhibitory potency of the drugs appeared generally to follow their lipid solubility and the order of their hypnotic efficiency, with hexobarbital as the most conspicuous exception. Therefore, barbiturate treatment of cells perturbs agonist-induced calcium mobilization. This effect may be partially linked to their previously reported inhibitory action on two kinases, protein kinase C and phosphatidylinositol 4-phosphate kinase [1, 2].

Interaction of protein kinase C with phosphoinositides.

Calcium/phosphatidylserine-dependent protein kinase C (PKC) is activated by phosphatidylinositol 4,5-bisphosphate (PIP2), as well as by diacylglycerol (DG) and phorbol esters. Here we report that PIP2, like DG, increases the affinity of PKC for Ca2+, and causes Ca(2+)-dependent translocation of the enzyme from the soluble to a particulate fraction (liposomes). Phosphatidylinositol 4-phosphate (PIP) also displaces phorbol ester from PKC and causes Ca(2+)-dependent translocation of the enzyme to liposomes, but is much less efficient than PIP2, and a much weaker activator, with a histone phosphorylation v(PIP)/v(PIP2) of approximately 0.15. Scatchard analysis indicates competitive inhibition between PIP and phorbol ester with Ki(PIP) = 0.26 mol% as compared with Ki(PIP2) = 0.043 mol%. No effect of phosphatidylinositol (PI) on phorbol ester binding to PKC, translocation of PKC, or activation of PKC was observed. These results suggest that both PIP and PIP2 can complex with PKC, but full activation of the enzyme takes place only when PIP is converted to PIP2. We suggest that an inositide interconversion shuttle has a role in the regulation of protein phosphorylation.

Activation of protein kinase C by phosphatidylinositol 4,5-bisphosphate: possible involvement in Na+/H+ antiport down-regulation and cell proliferation.

Phosphatidylinositol 4,5-bisphosphate (PIP2) as well as diacylglycerol (DG) activate protein kinase C (PKC) in the presence of calcium and phosphatidylserine. The pH at half-activation (pK) is 6.2 for DG.PKC and 7.7 for PIP2.PKC. Since the second monophosphate proton in position 5 of the PIP2 inositol (i.e., the last ionizable proton) has a pK of 7.7 (Van Paridon et al., (1986) Biochim. Biophys. Acta. 877, 216), the active effector is a fully deprotonated PIP2. Activation of PKC by PIP2 thus may follow intracellular alkalinization and be tied to the down-regulation of the Na+/H+ antiport mechanism. Since alkalinization is obligatory for cell proliferation, PIP2(5-).Ca.PKC may also be the gate that opens the pathways toward this and connected cellular reactions. A PIP2 analog in which inositol carbons 2-4 and the 4-phosphate have been removed, 1-phosphatidyl-rac-glycerol-3-phosphate (PGP), is completely inactive as PKC effector; this suggests that both 4-and 5-phosphate are engaged in the PIP2(5-).Ca.PKC complex. A model of the activated kinase takes this into account.

Action of amyloid beta-protein on protein kinase C activity.

Amyloid beta-protein (A beta), the major protein of cerebrovascular and plaque amyloid in Alzheimer disease, is considered a primary factor in the pathology of this disease. The effect of synthetic A beta (1-40) on the activity of protein kinase C (PKC) was studied with histones for a substrate in a mixed micellar assay, and with calmodulin-depleted soluble brain proteins in a liposomal system. We report here that A beta affects PKC activity in a biphasic manner. An initial stimulation of PKC was noted at low concentrations of A beta (less than 2.5 microM); while PKC-inhibition was observed in a concentration-dependent manner at higher concentrations of A beta. The in vitro phosphorylation of 20, 47, and 87 kDa brain proteins (known PKC substrates) was significantly reduced by 60 microM A beta. The role of 20 kDa in memory storage, of 87 kDa in neurotransmission and neurosecretory processes, and of 47 kDa in long-term potentiation or memory is well recognized, and A beta is known to have both neurotrophic and neurotoxic effects. Since PKC plays an important role in neuronal function, it is suggested that dual modulation of PKC by A beta may be linked to its neurotrophic and neurotoxic effects. We propose that at low concentrations A beta, by stimulating PKC, may contribute to neurites generation; and at higher concentrations A beta, by inhibiting PKC activity, might lead first to memory impairment, and then to neuronal loss.

Lipid activators of protein kinase C.

Among the many reported lipid activators of protein kinase C only those of high affinity can be considered true physiological effectors, at present the tumor promoters, e.g., phorbol esters; 1,2-diacyl-sn-glycerols; and phosphatidylinositol 4,5-bisphosphate. Many other compounds (including arachidonic acid) are activators at high, unphysiological concentrations only, and they seem to be sterically unsuited for bonding to the enzyme. Such pseudo-activators possibly act by scrambling the structure of the regulatory moiety of the kinase.

Mechanism of anesthesia: Anesthetics may restructure the hydrogen belts of membranes.

It is suggested that general anesthetics invade the "hydrogen belts" of neuronal plasma membranes, i.e. the regions of the bilayer containing hydrogen bond acceptors (CO of phospholipids) and donors (OH of cholesterol, sphingosin, ?-hydroxy fatty acids, proteins), and that they restructure the H-bond patterns between membrane lipids and proteins. The evidence is 2-fold. (1) The postulated existence of protein lipid hydrogen bonding has previously been demonstrated for glucose-6-phosphatase and for protein kinase C. (2) The prediction that changes in the H-bonding part of an anesthetic may influence its potency, but changes in the lipophilic part should not, can be verified. For a structurally widely varied group of monohydroxy compounds the range of anesthetic potency (inverse of intramembrane ED(50)(M) for loss of tadpole righting reflex) was smaller than 4-fold, while for n-hexane derivatives with different H-bonding headgroups the range of ED(50)(M) was about 100-fold. Although the results permit contributions by other factors they suggest that a restructuring of hydrogen belt H-bonding patterns is a general step in anesthesia.

Effect of barbiturates on polyphosphoinositide biosynthesis and protein kinase C activity in synaptosomes.

Previously, it has been found that phenobarbital inhibited protein kinase C (PKC) and the enzymes of the metabolism of polyphosphoinositide, especially phosphatidylinositol 4-phosphate (PIP) kinase (PIP-kinase). As a continuation of these studies, a number of barbiturates (barbituric acid, barbital, butabarbital, pentobarbital, amobarbital, phenobarbital, secobarbital and hexobarbital) were tested for inhibition of these enzymes and also of phosphatidylinositol (PI) kinase (PI-kinase), in a synaptosomal preparation at pH 7.8 from the brain of rat. All compounds, except barbituric acid (and Na-barbital for PI-kinase) inhibited the three kinases. However, PKC was approximately 3-5 fold more sensitive to inhibition by the drugs (measured by Ki values) than PIP-kinase, which was 2- to 4-fold more sensitive than PI-kinase. The inhibitory potency of the drugs increased with their lipophilicity, although to a lesser degree than expected from the differences in partition coefficients; the largest deviation from a positive correlation (i.e. hexobarbital) may be the result of the blockade of an imide (-NH) group at one position of the barbituric ring. Concentrations of drugs (after correction for the greater than normal ionization (pH 7.8) of the drugs in the assays) necessary for half-maximal inhibition were well within, or smaller than, those reported necessary for in vitro blocking of nerves. The possibility, therefore, exists that the physiological effects of the barbiturates are, in part, the result of an inhibition of protein kinase C and PIP-kinase.

Phosphatidylinositol 4,5-bisphosphate competitively inhibits phorbol ester binding to protein kinase C.

Calcium phospholipid dependent protein kinase C (PKC) is activated by diacylglycerol (DG) and by phorbol esters and is recognized to be the phorbol ester receptor of cells; DG displaces phorbol ester competitively from PKC. A phospholipid, phosphatidylinositol 4,5-bisphosphate (PIP2), can also activate PKC in the presence of phosphatidylserine (PS) and Ca2+ with a KPIP2 of 0.04 mol %. Preliminary experiments have suggested a common binding site for PIP2 and DG on PKC. Here, we investigate the effect of PIP2 on phorbol ester binding to PKC in a mixed micellar assay. In the presence of 20 mol % PS, PIP2 inhibited specific binding of [3H]phorbol 12,13-dibutyrate (PDBu) in a dose-dependent fashion up to 85% at 1 mol %. Inhibition of binding was more pronounced with PIP2 than with DG. Scatchard analysis indicated that the decrease in binding of PDBu in the presence of PIP2 is the result of an altered affinity for the phorbol ester rather than of a change in maximal binding. The plot of apparent dissociation constants (Kd') against PIP2 concentration was linear over a range of 0.01-1 mol % with a Ki of 0.043 mol % and confirmed the competitive nature of inhibition between PDBu and PIP2. Competition between PIP2 and phorbol ester could be demonstrated in a liposomal assay system also. These results indicate that PIP2, DG, and phorbol ester all compete for the same activator-receiving region on the regulatory moiety of protein kinase C, and they lend support to the suggestion that PIP2 is a primary activator of the enzyme.

Phosphatidylinositol-4,5-bisphosphate may antecede diacylglycerol as activator of protein kinase C.

Phosphatidylserine/calcium-dependent protein kinase C (PKC) from rat brain is activated fifty times more efficiently by phosphatidylinositol-4,5-bisphosphate (PIP2) (Kapp = 0.04 mole% in Triton-lipid micelles) than by diacylglycerol (DG) (Kapp = 2 mole%). Both effector lipids appear to bind to the same site but PIP2 may confer a narrower substrate specificity on the kinase. DG, which together with inositol trisphosphate (IP3) is generated by hydrolysis from PIP2 after cell stimulation, has been considered the natural activator of the kinase but it is likely to be anteceded in this function by PIP2; DG may perhaps retain the function of a back-up activator. The lack of PKC-activation by phosphatidylinositol (PI) or phosphatidylinositol-4-phosphate (PIP) opens the possibility that the Inositide Shuttle, PI reversible PIP reversible PIP2, has a role in controlling the activity of the kinase.

Calcium diphosphatidate membrane traversal is inhibited by common phospholipids and cholesterol but not by plasmalogen.

Phosphatidate-mediated Ca2+ membrane traversal is inhibited by phospholipids (PL) such a phosphatidylcholine (PC), phosphatidylinositol (PI), phosphatidylserine (PS), sphingomyelin and lysoPC, but not by PC-plasmalogen. Kinetics of Ca2+ traversal through a 'passive' bilayer consisting of OH-blocked cholesterol show competition between PC and phosphatidic acid (PA); it appears likely that a Ca(PA.PC) complex is formed which is not a transmembrane ionophore but will reduce the amount of phosphatidic acid available for the formation of the ionophore, Ca(PA)2. PS and PI may inhibit Ca2+-traversal in the same manner by forming Ca(PA.PL) complexes. We suggest that PC-plasmalogen, with one of the Ca2+-chelating ester CO groups missing, cannot engage in calcium cages, i.e., Ca(PA.PL) complexes, and thus does not interfere with Ca(PA)2 formation. Double-reciprocal plotting of Ca2+ traversal rates in cholesterol-containing liposomes vs. calcium concentration suggests that cholesterol inhibits Ca2+ traversal by competing with Ca2+ for PA. The inhibition does not seem to be caused by a restructuring or dehydration of the membrane 'hydrogen belts' affected by cholesterol; most probably, it is due to hydrogen bonding of the cholesterol-OH group to a CO group of PA; this reduces the amount of PA available for the calcium ferry. The inhibition by sphingomyelin and lysoPC may also be explained by their OH group interacting with PA via hydrogen bonding. The pH dependence of Ca2+ traversal suggests that H[Ca(PA)2]- can serve as Ca2+ cross-membrane ferry but that at physiological pH, [Ca(PA)2]2- is the predominant ionophore. In conclusion, the results indicate that Ca2+ traversal is strongly dependent on the structure of the hydrogen belts, i.e., the membrane strata occupied by hydrogen bond acceptors (CO of phospholipids) and donors (OH of cholesterol, sphingosine), and that lipid hydrogen belt structures may regulate storage and passage of Ca2+.

Preparation of passive bilayer liposomes.

In studies of in-membrane molecular interactions, need may arise for a matrix that cannot itself interact, except hydrophobically, with the reactants. Such a bilayer matrix should, ideally, consist of only a hydrophobic zone without ionic outer layers and without hydrogen belts (the membrane strata containing CO and OH groups). However, because of the necessity of anchoring the bilayer to its aqueous surroundings, there must be polar substituents. Hydrophilic ether groups in the form of polyoxyethylenes can provide nearly sufficient anchoring and yet not confer unwanted reactivity to the membrane since they are only very weak H-bond acceptors. The stability of the bilayer is ensured by the presence of a few percent of an amphiphile (which may be the substrate to be studied, e.g. a phospholipid) or by a free polyethylene hydroxy group far remote from the original hydrogen belt region. Our most impermeable liposomes consisted of O-methylcholesterol/O-methoxyethoxyethoxyethylcholesterol; the most readily prepared liposomes were made from O-methylcholesterol and hydroxy(ethoxy)4dodecane (Brij 30) or Triton.

Effect of phenobarbital on metabolism of polyphosphoinositides of rat brain synaptosomes.

Synthesis and degradation of polyphosphoinositides in a rat brain synaptosome preparation were depressed by phenobarbital. Phosphatidylinositol-4-phosphate kinase (PIP-kinase), the enzyme which synthesizes phosphatidylinositol-4,5-bisphosphate (PIP2) was most strongly affected (50% inhibition at 3 mM phenobarbital); phosphatidylinositol (PI-kinase) followed (50% at 15 mM). The phosphoesterases were less sensitive: PIP-monoesterase (50% at 39 mM), PIP2-monoesterase (at 47 mM), and, least inhibited, PIP-diesterase (50% at 65 mM) and PIP2-diesterase (at 68 mM). Phenobarbital by inhibiting PIP-kinase may reduce the membrane concentration of PIP2 and thus dampen the stimulus-response which leads to the hydrolysis of PIP2 and the formation of the second messenger, inositol-1,4,5-trisphosphate (IP3), involved in mobilization of intracellular Ca2+.

Phenobarbital competes with diacylglycerol for protein kinase C.

Phenobarbital inhibits protein kinase C of rat brain by competitively displacing the effector of the enzyme, diacylglycerol. The drug appears to occupy the triple hydrogen bonding site which bonds diacylglycerol - and also phorbol esters - to the enzyme. It remains to be seen if the effect is responsible for the pharmaceutical activity of the drug; even so, it provides an example of a restructuring of lipid-protein hydrogen bonding, in the hydrogen belt of the membrane, in a manner postulated as a mechanism of anesthesia.