Huu-yi Chang

The potential for platelet-activating factor synthesis in brain: Properties of cholinephosphotransferase and 1-alkyl-sn-glycero-3-phosphate acetyltransferase in microsomal fractions of immature rabbit cerebral cortex

The synthesis of platelet-activating factor (PAF) was studied in microsomal fractions of cerebral cortices of 15-day-old rabbits. These included: a total microsomal fraction P3, rough and smooth microsomes, R and S, and microsomal fraction P derived from isolated nerve cell bodies. Cholinephosphotransferase (CPT) generating PAF from alkylacetylglycerol had the highest specific activities in fractions R and P (24 and 6 times the homogenate values, based on membrane phospholipid content). This CPT activity differed from that which synthesized phosphatidylcholine as the latter was sensitive to dithiothreitol inhibition and was more readily inhibited by Triton X-100. As the CPT activity for PAF synthesis relies on the production of alkylacetylglycerol we studied the acetyltransferase which forms 1-alkyl-2-acetyl-sn-glycero-3-phosphate (AAGP). This enzyme had the highest specific activity in fraction R, followed by fractions P3 and P. There was evidence that the acetyltransferase was more active in a phosphorylated form. NaF maximized the recovery of AAGP products in the assays. The pH optimum for acetylation was in a range of 8.0-9.0. Lyso PAF did not inhibit the formation of AAGP and the rates of formation of PAF by acetylation were less than 5% of values for AAGP synthesis. During AAGP formation there was no evidence for subsequent alkylacetylglycerol formation in the absence of NaF, but a small formation of radioactive PAF could be demonstrated from AAGP under the CPT assay conditions.(ABSTRACT TRUNCATED AT 250 WORDS)

Alkylglycerophosphate acetyltransferase and lyso platelet activating factor acetyltransferase, two key enzymes in the synthesis of platelet activating factor, are found in neuronal nuclei isolated from cerebral cortex

Neuronal nuclear fractions (N1) isolated from cerebral cortices of 15-day-old rabbits were enriched in two acetyltransferases involved in biosynthetic pathways leading to platelet activating factor (PAF). Alkylglycerophosphate (AGP) acetyltransferase of the de novo biosynthetic path had specific activities in fraction N1 which were 3-times those of the microsomal fraction (P3D) from cerebral cortex. Lyso PAF acetyltransferase of the remodelling path had specific activities in N1 which were 16-times those of P3D and 51-times those of the homogenate. The maximum specific activity observed for the N1 AGP acetyltransferase was 1.4-times the corresponding N1 lyso PAF acetyltransferase value. The pH optimum for the N1 AGP acetyltransferase was within the alkaline range (pH 8-9), while the N1 lyso PAF acetyltransferase showed a much broader pH optimal range which extended over the neutral and physiological pH values. Both acetyltransferases were inhibited by MgATP (0.125-1 mM) or oleoyl CoA (2-10 microM). However, the N1 AGP acetyltransferase could be distinguished from the N1 lyso PAF acetyltransferase by a greater sensitivity to MgATP inhibition. When NaF was not present in the assays, less of the product of N1 AGP acetyltransferase was recovered, likely indicating a hydrolysis of the acetylated AGP. When the AGP and lyso PAF substrates were combined in acetyltransferase assays, the two N1 acetylations appeared to proceed independently. The enrichment of the acetyltransferases, and particularly the lyso PAF acetyltransferase, within the neuronal nuclear fraction is of particular interest with respect to the intracellular effects of PAF which are considered to be involved in nuclear signalling mechanisms.

Neuronal nuclear acetyltransferases involved in the synthesis of platelet-activating factor are located in the nuclear envelope and show differential losses in activity

Neuronal nuclear fraction N1 was isolated from cerebral cortices of 15-day-old rabbits, and nuclear subfractions prepared, in order to study the location of nuclear lyso platelet-activating factor (lyso-PAF) acetyltransferase and alkylglycerophosphate (AGP) acetyltransferase, and factors that affect the loss of these two nuclear activities. Subfractionation of prelabelled N1 indicated that the nuclear envelope had the highest percentage of the radioactive acetylated products alkylacetylglycerophosphate (AAGP) and PAF, and the distribution of these phospholipids reflected phospholipid distributions in the nuclear subfractions. The majority (95%) of radioactive AAGP and PAF was also recovered in Triton X-100 extracts of prelabelled nuclei, suggesting that these acetylated lipids are located in nuclear membranes rather than in the nuclear matrix/chromatin. Of the nuclear subfractions, the envelope had the highest AGP and lyso-PAF acetyltransferase specific activities which were close to corresponding values seen in the parent N1 fraction. Thus the nuclear AGP and lyso-PAF acetyltransferases were principally localized to the nuclear membranes. Differentials in activity loss were seen for the two acetyltransferase activities. In the nuclear envelope fractions, the lyso-PAF acetyltransferase was the more susceptible to oxidation reactions which could be reversed or blocked by the use of reducing agents. In preincubations, N1 showed greater losses in lyso-PAF acetyltransferase activity than in AGP acetyltransferase activity, losses which were not attributable to oxidation. Addition of cytosolic fraction S3 to preincubations promoted losses for each acetyltransferase in N1, and gave evidence for cytosolic and endogenous nuclear contributions to the activity loss. Addition of okadaic acid to the preincubations did not prevent the decline of either acetyltransferase in intact nuclei, but did diminish the loss of nuclear lyso-PAF acetyltransferase activity promoted by S3 addition, and also blocked the loss of this acetyltransferase seen in preincubations of isolated nuclear envelopes. This suggests that nuclear lyso-PAF acetyltransferase is susceptible to okadaic acid-sensitive nuclear and cytosolic protein phosphatase activities, while AGP acetyltransferase may lose activity by the action of other phosphatases or by other mechanisms within the nucleus.

Phosphatidylinositol synthetase activities in neuronal nuclei and microsomal fractions isolated from immature rabbit cerebral cortex.

The synthesis of phosphatidylinositol was studied using a nuclear fraction N1, a microsomal fraction P3, rough (R) and smooth (S) microsomal fractions and a microsomal fraction P derived from isolated nerve cell bodies. Each fraction was prepared using cerebral cortices of 15-day-old rabbits. In assays using CDP-diacylglycerol (prepared from egg phosphatidylcholine) and myo[3H]inositol at pH 7.4, fraction N1 had the highest maximal specific rates of phosphatidylinositol synthetase (EC 2.7.8.11) (expressed per mumol phospholipid in the fraction). However the three microsomal fractions achieved maximal specific activities at liponucleotide concentrations close to 50 microM, while fraction N1 required 200 microM concentrations. In certain cases (25-120 microM CDP-diacylglycerol, and at higher pH values) fraction R had specific activities which equalled or surpassed those of N1. However, with respect to inositol, fraction N1 had a distinctly lower Km than was shown for fractions R or P3. Each of the microsomal fractions and N1 required Mg2+ for the reaction, but for N1, maximal rates could be sustained at 0.1 mM, while for the microsomal fractions the optimal Mg2+ concentration was 1 mM. For each fraction Mn2+ could not replace Mg2+ in the reaction and Mn2+ was inhibitory. The optimal pH for the reaction was between 8.0 and 9.0. Phosphatidylinositol synthetase could also be shown using fraction N1 enriched in endogenous CDP-diacylglycerol. The relatively high specific activities of fraction N1, and the differences found between N1 and the microsomal fractions, for optimal CDP-diacylglycerol and Mg2+ concentrations and for Km values for inositol, support the existence of a neuronal nuclear phosphatidylinositol synthetase.

A metabolic path for the degradation of lysophosphatidic acid, an inhibitor of lysophosphatidylcholine lysophospholipase, in neuronal nuclei of cerebral cortex.

Neuronal nuclei isolated from rabbit cerebral cortex were found to be enriched in an NEM-insensitive lysophosphatidic acid (lysoPA) phosphohydrolase activity. LysoPA is an inhibitor of the nuclear lysophosphatidylcholine (lysoPC) lysophospholipase, and by preserving lysoPC levels, lysoPA boosted the nuclear production of the acyl analogue of platelet-activating factor by promoting the acetylation of lysoPC (Baker and Chang, Mol. Cell Biochem., 1999, in press). The nuclear phosphohydrolase converts lysoPA to 1-monoacylglycerol, and thus eliminates this lysoPA inhibition of lysoPC lysophospholipase. The nuclear lysoPA phosphohydrolase specific activity was more than three times that observed for the nuclear lysoPA lysophospholipase (Baker and Chang, Biochim. Biophys. Acta 1438 (1999) 253-263) and represents a more active route for nuclear lysoPA removal. The neuronal nuclear lysoPA phosphohydrolase was inhibited at acidic pH, and also inhibited by calcium ions. The 1-monoacylglycerol product of the phosphohydrolase is rapidly degraded by neuronal monoacylglycerol lipase, an enzyme some sevenfold more active than the phosphohydrolase and sensitive to inhibition by arachidonoyl trifluoromethyl ketone (AACOCF(3)). Both acidic pH and free fatty acid inhibited the lipase. In the absence of AACOCF(3), production of fatty acid from lysoPA substrate could be largely attributed to the sequential actions of the nuclear phosphohydrolase and lipase. This facilitates fatty acid recycling back into phospholipid by lysophospholipid acylation when ATP levels are restored following periods of brain ischemia. At relatively low concentrations, sphingosine-1-phosphate, and alkylglycerophosphate were the most effective phosphohydrolase inhibitors while phosphatidic acid, alkylacetylglycerophosphate and ceramide were without effect. LysoPA is an interesting regulatory molecule that can potentially preserve lysophosphatidylcholine within the nuclear membrane for use in acetylation reactions. Thus conditions relevant to brain ischemia such as falling pH, falling ATP concentrations, rising fatty acid and intracellular calcium levels may, by slowing this metabolic path for lysoPA loss, promote the production of acyl PAF and contribute to the increased levels of the acetylated lipids noted in ischemia.

The rapid incorporation of radioactive fatty acid into triacylglycerols during the in vitro acylation of native lipids of neuronal nuclei

Using neuronal nuclei (N1) and microsomes (P3) isolated from cerebral cortices of 15-day-old rabbits, the incorporation of [14C]oleate was followed in vitro, making use of fatty acid activation factors and endogenous membrane acyl acceptors. Of the lipids of N1, it was triacylglycerol which showed the highest rates of labelling and which represented 71-80% of the total incorporated radioactivity in this fraction. Specific rates of N1 triacylglycerol formation were 63-166 times those of P3 triacylglycerols (based upon membrane phospholipid content). In P3, phospholipids made up 85% of the total microsomal lipid labelling. The incorporation of oleate was dependent upon ATP and coenzyme A, and acyl-CoA synthetase activities were demonstrated in N1 and P3 (specific activity ratio, N1:P3 = 4.5). Using exogenous [14C]oleoyl-CoA, high rates of N1 triacylglycerol labelling were still seen relative to P3, but rates of diacylglycerol and phospholipid labelling were substantially elevated in both fractions in contrast to rates found using [14C]oleate. By increasing levels of endogenous diacylglycerol using preincubations with phospholipase C, a 3-fold increase was seen in specific rates of triacylglycerol formation in both fractions in subsequent assays with [14C]oleate. A 4.5-fold increase in N1 diacylglycerol concentrations was found when N1 was incubated for 10 min in the absence of fatty acid, ATP and coenzyme A. It is concluded that neuronal nuclei have a very active diacylglycerol acyltransferase as well as the ability to generate diacylglycerol substrates.

A comparison of lysophosphatidylcholine acyltransferase activities in neuronal nuclei and microsomes isolated from immature rabbit cerebral cortex.

Using neuronal nuclei (N1) and microsomes (P3) isolated from cerebral cortices of 15-day-old rabbits, the activity of lysophosphatidylcholine acyltransferase (acyl CoA: 1-acyl-sn-glycerol-3-phosphorylcholine acyltransferase) was studied using palmitoyl-, oleoyl- and arachidonoyl-CoA and a pool of lysophosphatidylcholine labelled with [3H]palmitate, [3H]stearate or [3H]oleate. Generally, in the acylation of the three radioactive lysophosphatidylcholines with arachidonoyl-CoA, the N1-specific acylation activities were two to seven times those of P3. For oleoyl-coA smaller N1 : P3 specific activity ratios were found, differing significantly from unity for only the 1-palmitoyl and 1-stearoyl lysophosphatidylcholines. The N1 : P3 ratios for the two unsaturated acyl-CoA thioesters were usually found to increase as the lysophosphatidylcholine concentration was lowered from 100 to 25 microM. Thus, nuclear acylation rates, particularly with arachidonoyl-CoA, were less affected by lowering the acceptor concentration than were microsomal activities. At the high lysophosphatidylcholine concentration (100 microM), arachidonoyl-CoA was a superior substrate to oleoyl-CoA in the nuclear acylations of the 1-palmitoyl or 1-stearoyl acceptors. Such a preference was never seen for the microsomal fraction. Using oleoyl- and arachidonoyl-CoA, the nuclear enzymes also showed a greater preference for the 1-palmitoyl homologue over the 1-oleoyl homologue than did the microsomal enzymes. These results support the existence of neuronal nuclear lysophosphatidylcholine acyltransferases with different substrate preferences than shown by the microsomal fraction.

Evidence for two distinct lysophospholipase activities that degrade lysophosphatidylcholine and lysophosphatidic acid in neuronal nuclei of cerebral cortex.

Neuronal nuclei were isolated from immature rabbit cerebral cortex and nuclear lysophospholipase activities studied using two different 1-acyl lysophospholipids: lysophosphatidylcholine (lysoPC) and lysophosphatidic acid (lysoPA). Our interest in these two lysolipids arose from the observation that lysoPA could promote the acetylation of lysoPC by substantially inhibiting a very active nuclear lysoPC lysophospholipase activity, in a competitive manner (R.R. Baker, H. -y. Chang, Mol. Cell. Biochem. (1999) in press). As there was also evidence for nuclear lysoPA deacylation, it was of interest to see whether one activity could possibly utilize both lysolipid substrates. We now have evidence for two separate lysophospholipase activities in neuronal nuclei. The lysoPC lysophospholipase activity was the more active, more highly enriched in the neuronal nuclei, and showed optimal activity at pH 8.4-9, while the lysoPA lysophospholipase activity was maintained over a much broader pH range. The lysoPC activity was substantially inhibited by free fatty acid, and showed considerable stimulation by serum albumin, while the activity utilizing lysoPA was much less affected by these agents. When lysoPC was added to incubations containing radioactive lysoPA, there was no significant inhibition found in rates of release of radioactive fatty acid, indicating that the lysoPA lysophospholipase activity did not utilize the lysoPC substrate. In incubations with lysoPC, MgATP and CoA brought about a sizable formation of phosphatidylcholine whose radioactivity was equally distributed between the sn-1 and sn-2 positions suggesting labelling both directly from the lysoPC substrate and from fatty acid produced by the lysophospholipase activity. By comparison, with the radioactive lysoPA substrate, MgATP and CoA promoted relatively lower levels of phosphatidic acid formation whose principal labelling came directly from the radioactive lysoPA. Largely because of the high activity of the nuclear lysoPC lysophospholipase, there is considerable potential in the neuronal nucleus to limit the use of lysoPC in other reactions, such as the formation of acylPAF (1-acyl analogue of platelet activating factor). It is of interest that conditions associated with brain ischaemia such as increased free fatty acid levels, falling pH and declines in MgATP may allow a preservation of neuronal nuclear lysoPC levels for acetylation. The existence of a separate lysophospholipase activity for lysoPA allows an independent control of lysoPA which can serve as an important regulator of the nuclear lysoPC lysophospholipase.

MgATP inhibits the synthesis of 1-alkyl-2-acetyl-sn-glycero-3-phosphate by microsomal acetyltransferase of immature rabbit cerebral cortex

The activity of 1-alkyl-sn-glycero-3-phosphate (AGP) acetyltransferase was studied using microsomal fractions isolated from cerebral cortices of 15-day-old rabbits. Fraction P3A was isolated using buffered 0.32 M sucrose containing mercaptoethanol, EDTA and NaF. This fraction had specific AGP acetyltransferase activities which were 4.9-times those of microsomal fraction P3B isolated in 0.32 M sucrose alone. This P3B activity was increased 2.4-times after a preincubation in the presence of ATP, MgCl2 and a high-speed supernatant fraction from cerebral cortex. Further, the activities of both P3A and P3B were almost completely eliminated by preincubation in the presence of alkaline phosphatase. Thus an activation of the AGP acetyltransferase by phosphorylation was indicated. While there was little inhibition of the P3A AGP acetyltransferase in the presence of added ATP, the magnesium salt form of ATP (1 mM) was severely inhibitory, bringing about 86% inhibition for P3A and 91% for P3B. The inhibitory effects of MgADP and MgAMP were smaller, and MgATP was a much more effective inhibitor than MgCTP, MgGTP and MgUTP which brought about 20-38% inhibitions of P3A activity at 1 mM concentrations. The effect of MgATP may be of particular relevance to the synthesis of platelet activating factor (PAF) following a period of ischemia in brain. Falling MgATP levels during energy failure could relieve the inhibition of AGP acetyltransferase seen in healthy cells and allow the formation of 1-alkyl-2-acetyl-sn-glycero-3-phosphate, which is the first committed intermediate in the de novo pathway of PAF synthesis.

The hydrolysis of natural phosphatidylethanolamines by phospholipase A2 from rat serum: a degree of selectivity is shown for docosahexaenoate release.

The selectivity of phospholipase A2 from serum was evaluated using radioassays and mass analyses of fatty acids liberated from phosphatidylcholine and phosphatidylethanolamine. These natural phospholipid substrates were labelled at the sn-2 position with radioactive oleate, linoleate and arachidonate. The rates of release of fatty acids were compared with their abundance at the sn-2 position of these phospholipid substrates. While there was little or no selectivity in the liberation of these fatty acids from phosphatidylcholine, there was some evidence for a preferential release of arachidonate with respect to linoleate from phosphatidylethanolamine. Mass analyses of free fatty acid products revealed that docosahexaenoate was consistently liberated at levels that exceeded its abundance at the sn-2 position of phosphatidylethanolamine. Three different, natural phosphatidylethanolamines with varying levels of docosahexaenoate showed a 1.2-1.8-fold enrichment of this polyunsaturate in the free fatty acid products compared with its abundance at the sn-2 position. This preference could also be shown when phosphatidylethanolamine was mixed with synthetic phosphatidylcholine as co-sonicated substrates. This preferential release of docosahexaenoate by serum phospholipase A2 is of considerable significance in the nervous system which is enriched in this polyunsaturate. The potential competition between liberated docosahexaenoate and arachidonate may be of fundamental importance in the response of brain to hemorrhage.