Regulatory aspects of mitochondrial phospholipase A2 from rat liver: effects of proteins, phospholipids and calcium ions

Mitochondrial iPLA2 Activity Modulates the Release of Cytochrome c from Mitochondria and Influences the Permeability Transition

The mitochondrial Ca(2+)-independent phospholipase A(2) is activated during energy-dependent Ca(2+) accumulation under conditions where there is a sustained depression of the membrane potential. This activation is not dependent on induction of the mitochondrial permeability transition. Bromoenol lactone, which inhibits the phospholipase, is effective as an inhibitor of the transition, and this action can be overcome by low levels of exogenous free fatty acids. Apparently, activation of the Ca(2+)-independent phospholipase is a factor in the mechanisms by which depolarization and Ca(2+) accumulation promote opening of the permeability transition pore. Sustained activity of the Ca(2+)-independent phospholipase A(2) promotes rupture of the outer mitochondrial membrane and spontaneous release of cytochrome c on a time scale similar to that of apoptosis occurring in cells. However, more swelling of the matrix space must occur to provoke release of a given cytochrome c fraction when the enzyme is active, compared with when it is inhibited. Through its effects on the permeability transition and release of intermembrane space proteins, the mitochondrial Ca(2+)-independent phospholipase A(2) may be an important factor governing cell death caused by necrosis or apoptosis.

Relation between accumulation of phospholipase A2 reaction products and Ca2+ release in isolated liver mitochondria

A Ca(2+)-dependent stimulation of mitochondrial phospholipase A2 is often assumed to play a role in mitochondrial Ca2+ release. We sought to clarify this relation by measuring Ca2+ transport and determining phospholipase A2 reaction products from the same sample of isolated, incubated rat liver mitochondria. When mitochondria had accumulated and spontaneously released again Ca2+, most probably by membrane permeability transition, there was no increase of phospholipase A2 reaction products. However, when the incubation was continued after Ca2+ release, significant increases of the content of lysophosphatidylcholine and unesterified fatty acids could be seen. Quinacrine, an inhibitor of phospholipase A2 activity, prevented Ca2+ release and p-hydroxymercuribenzoic acid, an inhibitor of lysophospholipid reesterification, induced a fast release of Ca2+ from isolated mitochondria. Such effects are usually taken as indirect evidence for a participation of phospholipase A2 in mitochondrial Ca2+ release, but analysis of the mitochondrial lipids revealed that no significant changes of the mass of phospholipase A2 reaction products had occurred. These experiments suggest that the accumulation of phospholipase A2 reaction products in mitochondria is the consequence rather than the cause of the membrane permeability transition. Exogenous phospholipase A2 products, lysophosphatidylcholine and arachidonic acid, induced mitochondrial Ca2+ release after a time lag, which decreased with aging of the mitochondrial preparation. The amount of lysophosphatidylcholine taken up by the mitochondria from the incubation medium during these experiments was measured and compared to the amount of lysophosphatidylcholine produced endogenously by mitochondrial phospholipase A2. From these data it appears likely that the amount of lysophosphatidylcholine generated in the mitochondria after the permeability transition is sufficient to sustain the permeable state. An accumulation of mitochondrially generated phospholipase A2 reaction products after the permeability transition could thus be a decisive factor for the limited reversibility of the membrane permeability transition.

Lipid composition of glucose-stimulated pancreatic islets and insulin-secreting tumor cells

The effect of glucose stimulation (25 mM for 5 min) on the phospholipid and neutral lipid composition of isolated pancreatic islets was studied to find out whether there is a change in the mass of potential lipid mediators or modulators of insulin secretion. For comparison, the lipid compositions of homogenates and subcellular fractions from RINm5F insulin-secreting tumor cells and of glucose-stimulated streptozotocin/nicotinamide-induced islet cell tumors were analyzed. After separation of the lipid extract into a neutral and an acidic fraction by anion-exchange chromatography, lipids were separated by high-performance thin-layer chromatography and quantitated by in situ densitometry of the cupric sulfate-charred bands. In glucose-stimulated islets, the molar percentages of phosphatidic acid (PA) and of phosphatidylinositol were significantly increased (3.1 vs. 4.7 mol% and 8.6 vs. 11.8 mol%), while those of all other phospholipids and neutral lipids, including 1,2-diacylglycerol, were not significantly changed. In stimulated islet cell tumors, an increase of PA was visible in the microsomal fraction, and there was an increase of lysophosphatidylcholine in the mitochondrial fraction. However, in both tumoral tissues, particularly in RINm5F cells, the lipid distribution pattern showed abnormalities which can be regarded as a loss of differentiation and which limit the usefulness of these tissues for the study of the physiological regulation of lipid metabolism during glucose stimulation. In conclusion, the data are in accordance with a role of PA early in stimulus-secretion coupling. The well-known stimulation of phospholipid synthesis in pancreatic islets during glucose-induced insulin secretion does not result in an increase in the total phospholipid mass.

Demonstration of similar calcium dependencies by mammalian high and low molecular mass phospholipase A2

The in vitro Ca2+ dependencies of arachidonyl (AA)-selective high molecular mass phospholipase A2 (HMM, 85 kDa-PLA2) and human low molecular mass (LMM-Type II, 14 kDa)-PLA2 were compared. When the LMM-PLA2 and HMM-PLA2 enzymes were examined for hydrolysis against [3H]AA Escherichia coli in an ethyleneglycol-bis(beta-aminoethyl ether) N,N,N',N'-tetraacetic acid (EGTA)-free buffer system, neither enzyme demonstrated activity below 10 microM free Ca2+. Beyond 11 microM Ca2+ both enzyme activities increased steadily exhibiting 50% of maximal activity at 0.1 and 1.0 mM, respectively. Using EGTA-regulated free Ca2+ buffers, both enzymes responded in a biphasic manner, achieving 50% of the maximum response by 0.5 microM Ca2+, stabilizing up to 0.1 mM, then further increasing with exposure to millimolar Ca2+ concentrations. Replacement of [3H]AA-labeled phosphatidylethanolamine vesicles for [3H]AA E. coli or using Tris-HCl buffer instead of HEPES buffer did not alter these findings significantly. The presence of EGTA had a pronounced concentration-dependent effect on the activity of both the HMM- and LMM-PLA2 enzymes but only in the range of 0 to 100 microM free Ca2+. EGTA (EC50 approximately 200 microM) reduced the concentration of Ca2+ required by PLA2 to achieve 50% of maximal acylhydrolysis. In contrast, the Type I bovine pancreatic PLA2 required millimolar Ca2+ concentrations to elicit 50% of the maximal response in both EGTA-free or EGTA-containing systems, which is concordant with its extracellular role as a digestive enzyme. These data suggest that the LMM-Type II PLA2 and HMM-PLA2 are both activated at submicromolar, intracellularly relevant, Ca2+ concentrations and therefore have the ability to contribute to cellular lipid metabolism.

Cardiolipins and biomembrane function

Evidence is discussed for roles of cardiolipins in oxidative phosphorylation mechanisms that regulate State 4 respiration by returning ejected protons across and over bacterial and mitochondrial membrane phospholipids, and that regulate State 3 respiration through the relative contributions of proteins that transport protons, electrons and/or metabolites. The barrier properties of phospholipid bilayers support and regulate the slow proton leak that is the basis for State 4 respiration. Proton permeability is in the range 10(-3)-10(-4) cm s-1 in mitochondria and in protein-free membranes formed from extracted mitochondrial phospholipids or from stable synthetic phosphatidylcholines or phosphatidylethanolamines. The roles of cardiolipins in proton conductance in model phospholipid membrane systems need to be assessed in view of new findings by Hübner et al. [313]: saturated cardiolipins form bilayers whilst natural highly unsaturated cardiolipins form nonlamellar phases. Mitochondrial cardiolipins apparently participate in bilayers formed by phosphatidylcholines and phosphatidylethanolamines. It is not yet clear if cardiolipins themselves conduct protons back across the membrane according to their degree of fatty acyl saturation, and/or modulate proton conductance by phosphatidylcholines and phosphatidylethanolamines. Mitochondrial cardiolipins, especially those with high 18:2 acyl contents, strongly bind many carrier and enzyme proteins that are involved in oxidative phosphorylation, some of which contribute to regulation of State 3 respiration. The role of cardiolipins in biomembrane protein function has been examined by measuring retained phospholipids and phospholipid binding in purified proteins, and by reconstituting delipidated proteins. The reconstitution criterion for the significance of cardiolipin-protein interactions has been catalytical activity; proton-pumping and multiprotein interactions have yet to be correlated. Some proteins, e.g., cytochrome c oxidase are catalytically active when dimyristoylphosphatidylcholine replaces retained cardiolipins. Cardiolipin-protein interactions orient membrane proteins, matrix proteins, and on the outerface receptors, enzymes, and some leader peptides for import; activate enzymes or keep them inactive unless the inner membrane is disrupted; and modulate formation of nonbilayer HII-phases. The capacity of the proton-exchanging uncoupling protein to accelerate thermogenic respiration in brown adipose tissue mitochondria of cold-adapted animals is not apparently affected by the increased cardiolipin unsaturation; this protein seems to take over the protonophoric role of cardiolipins in other mitochondria. Many in vivo influences that affect proton leakage and carrier rates selectively alter cardiolipins in amount per mitochondrial phospholipids, in fatty acyl composition and perhaps in sidedness; other mitochondrial membrane phospholipids respond less or not at all.(ABSTRACT TRUNCATED AT 400 WORDS)

Evidence that the rate of phosphatidylcholine catabolism is regulated in cultured rat hepatocytes

The regulation of phosphatidylcholine (PC) catabolism has been studied in choline-deficient rat hepatocytes. Supplementation of choline-deficient hepatocytes, prelabeled with [3H]choline, with 100 microM choline increased the rate of PC catabolism by approx. 2-fold. The major product of PC degradation was glycerophosphocholine in both choline-deficient and choline-supplemented cells. Choline supplementation decreased the radioactivity recovered in lysoPC by 50%. This effect was accompanied by a 2-fold increase of labeled glycerophosphocholine. Comparable results were obtained when PC of the cells was prelabeled with [3H]methionine or [3H]glycerol. The activity of phospholipase A in cytosol, mitochondria and microsomes isolated from choline-deficient rat liver was similar to the activity in control liver, when determined with [3H]PC vesicles as the substrate. Measurement of the activity of phospholipase A with endogenously [3H]choline-labeled PC showed that the formation of lysoPC in mitochondria isolated form choline-supplemented cells was 40% lower than in choline-deficient cells. Alternatively, the formation of [3H]glycerophosphocholine and [3H]choline in microsomes from choline-supplemented cells was significantly higher (1.4-fold) than in microsomes from choline-deficient cells. These results suggest that the rate of PC catabolism is regulated in rat hepatocytes and that the concentration of PC might be an important regulatory factor.

Synergistic effect of magnesium and calcium ions in the activation of phospholipase A2 of liver macrophages

In cell-free extracts of rat liver macrophages (Kupffer cells) phospholipase A2 was found to be strongly activated at free Ca2+ concentrations from 100 nM to 1 microM in the presence of 4 mM free Mg2+. This is within the range of intracellular free Ca2+ reported for basal and various stimulated conditions, respectively. Ca2+ alone increased phospholipase A2 activity at high Ca2+ concentrations (1 mM) whereas Mg2+ alone had only little stimulatory effect. Calmodulin does not seem to participate in the regulation of phospholipase A2 although it relieved the inhibition of phospholipase A2 activity by calmodulin antagonists.

Lysocardiolipin formation and reacylation in isolated rat liver mitochondria

Liver mitochondrial cardiolipin (CL) is distinguished from other phospholipids by the presence of linoleoyl in almost all molecular species, and the biosynthesis of these species is not yet understood. The present study was carried out in order to test the hypothesis that the linoleoyl proportion of CL may be specifically enriched by a deacylation-reacylation cycle. Incorporation of [14C]glycerol 3-phosphate into the metabolites of the CL pathway was accompanied by formation of 14C-labelled monolyso- and dilyso-CL. Labelling of dilyso-CL was increased or decreased by stimulation or inhibition respectively of mitochondrial phospholipase A2. These data suggest a rapid deacylation of newly formed [14C]CL by phospholipase A2, whereas endogenous mitochondrial CL was very resistant to hydrolytic degradation. Unlike dilyso-CL, monolyso-CL could be reacylated by [14C]linoleoyl residues. [14C]Linoleoyl incorporation into CL was also observed when exogenous CL was added instead of monolyso-CL, thus indicating the concerted action of de- and re-acylation. Although 1-palmitoyl-2-[14C]linoleoyl-phosphatidylcholine was a suitable acyl donor under experimental conditions, the reaction was not a transacylation but required splitting of [14C]linoleic acid from phosphatidylcholine and formation of [14C]linoleoyl-CoA as an intermediate. The [14C]linoleoyl was mainly bound to the sn-2(2") position of CL, and a small portion (about 20%) to the sn-1(1") position. It is concluded that a cycle, comprising CL deacylation and monolyso-CL reacylation by linoleoyl-CoA, provides a potential mechanism for the remodelling of molecular species of newly formed CL.

Phospholipid fatty acid composition of various mouse tissues after feeding alpha-linolenate (18:3n-3) or eicosatrienoate (20:3n-3)

The selective incorporation of dietary alpha-linolenate (18:3n-3) and its elongation product, eicosatrienoate (20:3n-3), into various phospholipids (PL) of mouse liver, spleen, kidney, and heart, was examined in a two-week feeding trial by assessing mol % changes in associated fatty acids. Mice were fed fat-free AIN 76A diets modified with either 2 wt% safflower oil (control); 1% safflower and 1% linolenate; or 1% safflower and 1% eicosatrienoate. After linolenate or eicosatrienoate feeding, 20:4n-6 was reduced by 36-50% in liver phosphatidylcholine (PC) and in liver and spleen phosphatidylethanolamine (PE). Linolenate was minimally incorporated into PL, but was desaturated and elongated to 20:5n-3, 22:5n-3, and 22:6n-3, with notable differences in the quantity of these n-3 derivatives associated with different tissues and PL. Eicosatrienoate was uniquely incorporated into the cardiolipin (CL) pool of all organs. There was also considerable retroconversion of 20:3n-3 to 18:3n-3 (PC,PE). Dietary eicosatrienoate may therefore affect metabolism in diverse ways--20:3n-3, which is retroconverted to 18:3n-3, may provide substrate for 20:5n-3 and 22:6n-3 syntheses, whereas intact 20:3n-3 may be incorporated into the CL pool. Acyl modifications of CL are known to affect the activity of key innermitochondrial enzymes, such as cytochrome c oxidase.

Intramembraneous hydrogenation of mitochondrial lipids reduces the substrate availability, but not the enzyme activity of endogenous phospholipase A. The role of polyunsaturated phospholipid species

(1) Isolated rat liver mitochondria were subjected to catalytic hydrogenation using a water-soluble Pd complex and molecular H2. This treatment resulted in a reduction of double bonds on phospholipid acyl chains as judged by gas chromatography of fatty acid methyl esters and HPLC of dinitrobenzoyldiacylglycerols. (2) After hydrogenation, mitochondria lost their ability to hydrolyze endogenous phospholipids in alkaline, Ca2+ containing medium, while phospholipase A2 retained full activity against exogenous substrates, regardless of whether those substrates were hydrogenated or not. (3) Inhibition by hydrogenation of endogenous phospholipid hydrolysis correlated with the loss of polyunsaturated fatty acyls, rather than with changes of the bulk membrane fluidity as measured by ESR and fluorescence studies. (4) These data suggest that the unsaturation of mitochondrial membrane lipids might be important for regulation of phospholipid breakdown by endogenous phospholipases. In particular, polyunsaturated molecular species seem to be involved in making phospholipids accessible to phospholipase A-mediated hydrolysis.

Regulation of transmembrane ion transport by reaction products of phospholipase A2. I. Effects of lysophospholipids on mitochondrial Ca2+ transport

Lysophospholipids inhibited mitochondrial Ca2+ uptake, induced a net Ca2+ efflux, and thereby increased the extramitochondrial Ca2+ concentration. The inhibitory potency decreased in the order lysophosphatidylcholine (LPC) = lysophosphatidylglycerol (LPG) greater than lysophosphatidylinositol (LPI) greater than lysophosphatidylserine (LPS) much greater than lysophosphatidylethanolamine (LPE). This relative order is in inverse relation to the ability of the various phospholipid head-groups to build up intermolecular hydrogen bonds with neighbouring membrane lipids. This indicates that changes in Ca2+ transport induced by lysophospholipids are mediated by the interaction of the lysophospholipids with the mitochondrial membrane bilayer structure. The mitochondrial membrane potential, which is the main driving force for mitochondrial Ca2+ uptake, was affected in the same order by the various lysophospholipids. This reduction of the mitochondrial membrane potential may be the underlying cause for the inhibition of the mitochondrial Ca2+ uniport and the resulting release of Ca2+ from the mitochondria.

Synthetic peptide from lipocortin I has no phospholipase A2 inhibitory activity

Two anti-inflammatory peptides corresponding to a high amino acid similarity region between lipocortins were synthesized and tested on their ability to inhibit porcine pancreatic phospholipase A2. Kinetic assays using monomeric and aggregated phospholipids did not reveal any phospholipase A2 inhibitory activity. The peptides did not inhibit phospholipase A2 activity on monolayers of negatively charged substrate and did not prevent phospholipase A2 action on mixed micelles of 1-stearoyl-2-arachidonoyl-sn-glycero-3-phosphocholine and sodiumdeoxycholate. Ultraviolet difference spectroscopy did not show binding of the peptides to phospholipase A2. Therefore we conclude that these anti-inflammatory peptides do not inhibit pancreatic phospholipase A2 in vitro, in contrast to the results recently published [(1988) Nature 335, 726-730].

A novel phospholipase A2 associated with nuclear matrix: stimulation of the activity and modulation of the Ca2+ dependency by polyphosphoinositides

Neutral phospholipase A2 activity, which hydrolyzed phosphatidylcholine and phosphatidylethanolamine with the same efficiency, was identified in the nuclear matrix prepared from purified nuclei of rat ascites hepatoma cells (AH 7974). The enzyme activity was optimal at pH 7.0 and required Ca2+ absolutely. Concentrations of Ca2+ for a maximal and a half-maximal activation were 1.10(-2) and 1.10(-3) M, respectively, and little activity was detected at Ca2+ concentrations lower than 1.10(-5) M. Addition of acidic phospholipids markedly stimulated the enzyme activity, and further, lowered the minimum Ca2+ concentration required for activation. In particular, the polyphosphoionositides phosphatidylinositol 4-monophosphate and 4,5-diphosphate were most effective. These two polyphosphoinositides lowered the Ca2+ concentration required for half-maximal activation to 10(-5) M and dramatically stimulated the activity at that Ca2+ concentration (greater than 30-fold). The neutral phospholipase A2 activity such as characterized in the present study was very low in the other subcellular fractions including mitochondria, microsome, plasma membrane and cytosol.

Phospholipase A2--regulation and inhibition

Phospholipase A2 (PLA2) has been implicated in the pathogenesis of different diseases. Thus, the pharmacological intervention of PLA2 activity by specific inhibitors is of great therapeutical value in ameliorating pathological conditions. Despite a great number of published data regarding PLA2 inhibitors none has reached clinical application. Since enzyme activity can be greatly influenced by the experimental conditions of the test system used, a potent in vitro enzyme inhibitor does not indicate therapeutic effectiveness per se. In order to enhance the predictable value of an in vitro screening system for PLA2 inhibitors, a battery of test systems each measuring certain parameters should be applied. Considering the complex mechanism(s) of PLA2 it is extremely important to elucidate the exact inhibition mechanism of those compounds, which have passed these first filters. True inhibitors of PLA2 should then be evaluated in suitable ex vivo, in vivo models.