Activation of phospholipase D by PKC and GTPgammaS in human neuroblastoma cells overexpressing MARCKS.

Regulation of phospholipase D (PLD) activity participating in signal transduction involves complex interactions with small G-proteins (ARF, Rho) and protein kinase C isoforms (PKCalpha). In SK-N-MC human neuroblastoma cells, phorbol ester (TPA) activation of PLD was enhanced by overexpressing myristoylated alanine-rich C kinase substrate (MARCKS). To study MARCKS interactions with PLD, we investigated PLD isoform expression and activation by TPA and GTPgammaS in intact and digitonin-permeabilized clones transfected with MARCKS (M22). PLD2 was in both cytosol and membrane fractions while PLD1 was primarily membrane-associated in both vector control and M22 cells; location or quantities were unaltered by TPA treatment. TPA-stimulated PLD activity was higher in both intact and digitonin-permeabilized M22 cells than in vector controls. In contrast, GTPgammaS-stimulated PLD activity was independent of MARCKS expression but was additive with MARCKS-PKC-dependent activation in permeabilized cells. Combinations of PKC inhibition and down-regulation in intact and permeabilized (with GTPgammaS present) cells indicated that a PKC-mediated phosphorylation event was necessary in intact cells without access to GTPgammaS, stimulation of PLD mediated by GTPgammaS was independent of PKC, and PLD activation by PKC in permeabilized cells was kinase-independent. Western blot analysis showed that MARCKS, PKCalpha, PLD1 and PLD2 were present in a detergent-insoluble fraction (DIF); GTPgammaS increased recovery of PLD2 in DIF. Disruption of cholesterol-rich DIFs with digitonin, cyclodextrin or filipin potentiated activation of PLD by TPA. Our studies suggest that activation of PLD by PKC requires MARCKS and can involve both phosphorylation-independent and -dependent processes. As PLD activation by GTPgammaS is PKC-MARCKS-independent, MARCKS may provide a fine tuning component in conjunction with G-protein-mediated mechanisms for regulation of PLD.

Overexpression of myristoylated alanine-rich C-kinase substrate enhances activation of phospholipase D by protein kinase C in SK-N-MC human neuroblastoma cells.

Signal transduction can involve the activation of protein kinase C (PKC) and the subsequent phosphorylation of protein substrates, including myristoylated alanine-rich C kinase substrate (MARCKS). Previously we showed that stimulation of phosphatidylcholine (PtdCho) synthesis by PMA in SK-N-MC human neuroblastoma cells required overexpression of MARCKS, whereas PKCalpha alone was insufficient. We have now investigated the role of MARCKS in PMA-stimulated PtdCho hydrolysis by phospholipase D (PLD). Overexpression of MARCKS enhanced PLD activity 1.3-2.5-fold compared with vector controls in unstimulated cells, and 3-4-fold in cells stimulated with 100 nM PMA. PMA-stimulated PLD activity was blocked by the PKC inhibitor bisindolylmaleimide. Activation of PLD by PMA was linear with time to 60 min, whereas stimulation of PtdCho synthesis by PMA in clones overexpressing MARCKS was observed after a 15 min time lag, suggesting that the hydrolysis of PtdCho by PLD preceded synthesis. The formation of phosphatidylbutanol by PLD was greatest when PtdCho was the predominantly labelled phospholipid, indicating that PtdCho was the preferred, but not the only, phospholipid substrate for PLD. Cells overexpressing MARCKS had 2-fold higher levels of PKCalpha than in vector control cells analysed by Western blot analysis; levels of PKCbeta and PLD were similar in all clones. The loss of both MARCKS and PKCalpha expression at higher subcultures of the clones was paralleled by the loss of stimulation of PLD activity and PtdCho synthesis by PMA. Our results show that MARCKS is an essential link in the PKC-mediated activation of PtdCho-specific PLD in these cells and that the stimulation of PtdCho synthesis by PMA is a secondary response.

Protein kinase C isoforms and growth, differentiation and phosphatidylcholine turnover in human neuroblastoma cells.

Neuroblastoma and glioma cells differentially express isoforms of protein kinase C (PKC) and myristoylated PKC substrates (e.g. MARCKS). Correlation with metabolism of membrane phospholipids suggests that PKC-alpha and MARCKS may be required to mediate phosphatidylcholine turnover stimulated by phorbol ester (beta-TPA). To evaluate relationships to neural cell differentiation, SK-N-SH human neuroblastoma cells were treated with 20 nM beta-TPA. In beta-TPA-treated cells, growth arrest and differentiation occurred (neurite extension; 40-60% decrease in cell number, total protein and RNA). By day 4, mRNA for PKC-alpha and MARCKS increased and, after an initial decrease, PKC-alpha protein also increased. At day 4, phosphatidylcholine synthesis was 3-5 fold greater than in control cells. In contrast, C6 glioma cells treated with beta-TPA showed no growth arrest, decreased PKC-alpha protein (< 20%) and lower phosphatidylcholine synthesis. Thus, induced differentiation of human neuroblastoma cells involved increased expression of PKC-alpha and MARCKS and synthesis of phosphatidylcholine, consistent with involvement of PKC-alpha and MARCKS in regulation of phosphatidylcholine turnover during neurite growth.

Lipopolysaccharide stimulates differential expression of myristoylated protein kinase C substrates in murine microglia.

Microglia rapidly respond to lipoplysaccharide (LPS) by transformation from resting to active states and secretion of several neuro- and immuno-regulators including tumour necrosis factor alpha (TNF-alpha), interleukin 1 beta (IL-1 beta), and interleukin 6 (IL-6). With longer LPS treatment, microglia are converted to reactive or phagocytic states with characteristics similar to macrophages in inflammation and injury processes. We have investigated LPS-mediated changes in two myristoylated substrates of protein kinase C (PKC): MARCKS (myristoylated alaninerich C kinase substrate) and MRP (MARCKS-related protein). Within 6 hours of addition, LPS induced a twofold increase in [3H]myristoylated and immunoreactive MARCKS protein and a sevenfold increase in MRP. The differential effect of LPS on expression of MRP vs. MARCKS was even more dramatic at the level of transcription: S1 nuclease protection assays revealed a 40-fold increase in MRP mRNA levels (maximum at 4-6 hours), whereas a threefold increase was observed for MARCKS. TNF alpha and colony-stimulating factor 1 (CSF-1), two cytokines which are induced by LPS, did not reproduce the observed effect of LPS on MARCKS and MRP gene transcription. CSF-1 also induced differential transcription of MRP, but of lower magnitude (threefold) and more sustained than by LPS. Accordingly, these two substrates for PKC are differentially up-regulated by LPS, apparently independent of TNF alpha or CSF-1.

Overexpression of MARCKS, but not protein kinase C-alpha, increases phorbol ester-stimulated synthesis of phosphatidylcholine in human SK-N-MC neuroblastoma cells.

To investigate the regulation of phorbol ester-stimulated synthesis of phosphatidylcholine (PtdCho), myristoylated alanine-rich protein kinase C substrate (MARCKS) and the alpha-isoform of protein kinase C (PKC-alpha) were overexpressed in a human neuroblastoma (SK-N-MC) cell line that does not increase PtdCho synthesis in response to 4beta-12-O-tetradecanoylphorbol 13-acetate (TPA). In five clones with a less than fivefold increase in MARCKS protein level, the synthesis of PtdCho from [methyl-3H] choline was stimulated 1.88-2.34-fold in the presence of 100-200 nM TPA. In clones overexpressing PKC-alpha (30-40-fold increased level of protein) or in mock-transfected vector controls, TPA had much less of a stimulatory effect (1.04-1.43 fold) on PtdCho synthesis. TPA caused translocation of PKC-alpha and increased phosphorylation of MARCKS, indicating that both overexpressed proteins responded to stimulation. Thus, in SK-N-MC cells, MARCKS is required for TPA-stimulated synthesis of PtdCho and PKC-alpha alone is insufficient for supporting enhanced synthesis.

Phospholipid and cation activation of chimaeric choline/ethanolamine phosphotransferases.

The Saccharomyces cerevisiae CPT1 and EPT1 genes encode for a cholinephosphotransferase (CPT) and choline/ethanolaminephosphotransferase, respectively. Both Cpt1p and Ept1p activities display an absolute requirement for cations and phospholipids. A mixed-micelle assay was employed to determine cation and lipid activators of parental and chimaeric Cpt1p/Ept1p enzymes to gain insight into their mechanism(s) of activation. Mg2+, Mn2+ and Co2+ were the only cations capable of activating Cpt1p and Ept1p in vitro. Kinetic data revealed that only Mg2+ is present in appropriate amounts to activate CPT activity in vivo. Kinetic data revealed that only Mg2+ is present in appropriate amounts to activate CPT activity in vivo. The two enzymes displayed distinct activation profiles on the basis of their relative affinities for Mg2+, and Mn2+ and Co2+. This allowed the use of chimaeric enzymes to determine the mechanism of cation activation. Cations do not activate Cpt1p or Ept1p by complexing with the substrate, CDP-choline, but instead bind to disparate regions within the enzymes themselves. Cpt1p and Ept1p also displayed distinct phospholipid activation profiles. Phospholipid activation required a phosphate and/or glycero-phosphoester linkage, with the phospho-head group moiety positioned at the surface of the micelle. Assays with parental and chimaeric Cpt1p/Ept1p constructs revealed that the phospholipid binding/activation domains are not located within linear segments of the protein, but instead are contained within distinct and separate regions of the proteins that require an intact tertiary structure for formation. Phosphatidylcholine (and its structural analogue sphingomyelin) were the best lipid activators of Cpt1p, the main biologically relevant CPT activity in S. cerevisiae. Hence CPT displays product activation. Because phosphatidylcholine is an efficient activator of CPT activity (and hence Cpt1p is not subject to feedback inhibition by its product), Cpt1p is incapable of functioning as a direct monitor of membrane phosphatidylcholine composition.

Phorbol ester stimulation of phosphatidylcholine synthesis in four cultured neural cell lines: correlations with expression of protein kinase C isoforms.

Phosphatidylcholine (PtdCho) can provide lipid second messengers involved in signal transduction pathways. As a measure of phospholipid turnover in response to extracellular stimulation, we investigated differential enhancement of [3H]choline incorporation into PtdCho by phorbol esters. In C6 rat glioma and SK-N-SH human neuroblastoma cells, [3H]PtdCho synthesis was 2-4 fold stimulated by beta-12-O-tetradecanoylphorbol-13-acetate (beta-TPA) when [3H]choline was incubated simultaneously with, or 15 min prior to, beta-TPA treatment. By contrast, in N1E-115 mouse and SK-N-MC human neuroblastoma cells, phorbol esters had no appreciable effect on [3H]choline incorporation; however, in all cells, 200 microM oleic acid enhanced PtdCho synthesis, indicating a stimulable process. Alterations by thymeleatoxin (TMT), an activator of conventional PKC isoforms (alpha, beta and gamma), were similar to beta-TPA. We investigated whether expression of specific PKC isoforms might correlate with these effects of phorbol esters on PtdCho synthesis. All cell lines bound phorbol esters, had PKC activity that was translocated by phorbol esters and differentially expressed isoforms of PKC. Northern and western blot analyses, using specific cDNA and antibodies for PKC-alpha, -beta, -gamma, -delta, -epsilon, and -zeta, revealed that expression of alpha-isoform predominated in C6 and SK-N-SH cells. In contrast, TPA-responsive beta-isoform predominated in SK-N-MC cells. gamma-PKC was not detected in any cells and only in C6 cells was PKC-delta present and translocated by beta-TPA treatment. PKC-epsilon was not detected in SK-N-MC cell lines but translocated with TPA treatment in the other three cell lines. PKC-zeta was present in all cells but was unaltered by TPA treatment. Accordingly, stimulation of PtdCho turnover by phorbol esters correlated only with expression of PKC-alpha; presence of PKC-beta alone was insufficient for a TPA response.

Studies employing Saccharomyces cerevisiae cpt1 and ept1 null mutants implicate the CPT1 gene in coordinate regulation of phospholipid biosynthesis.

The Saccharomyces cerevisiae CPT1 and EPT1 genes are structural genes encoding sn-1,2-diacylglycerol choline phosphotransferase and sn-1,2-diacylglycerol choline/ethanolamine phosphotransferase, respectively. Incorporation of 32Pi into phosphatidylcholine, phosphatidylethanolamine, and phosphatidylserine in wild type and ept1 strains was decreased in the presence of exogenous inositol. In contrast, inositol did not affect 32Pi incorporation into phospholipid in cpt1 or cpt1ept1 strains. In membranes isolated from wild type and ept1 strains grown in the presence of inositol or inositol/choline, the CPT1-derived cholinephosphotransferase activities were reduced 40-50 and 65%, respectively. Inositol-dependent reductions in CPT1 derived choline-phosphotransferase activity correlated with transcript levels in both wild type and ept- backgrounds. The ethanolaminephosphotransferase activity of the EPT1 gene product in wild type cells was reduced 40% by exogenous inositol alone and 50% by inositol/choline. In the cpt1 strain, however, the ethanolaminephosphotransferase activity was unaffected by exogenous inositol or inositol/choline. The inositol-dependent reduction of ethanolaminephosphotransferase activity observed in wild type cells correlated with reduced levels of EPT1 transcripts; in the cpt1 strain, EPT1 transcript levels were not affected by inositol. These results indicate that 1) a functional CPT1 gene or gene product is required for inositol-dependent regulation of phospholipid synthesis; 2) the enzyme activities of both the CPT1 and EPT1 gene products are repressed by inositol and inositol/choline, and require an intact CPT1 gene; 3) inositol mediates its regulatory effects on phospholipid synthesis via a transcriptional mechanism.

Chimeric enzymes. Structure-function analysis of segments of sn-1,2-diacylglycerol choline- and ethanolaminephosphotransferases.

The Saccharomyces cerevisiae CPT1 and EPT1 genes represent structural genes that encode distinct choline- and choline/ethanolaminephosphotransferases, respectively. To explore the function of linear segments of these enzymes, a series of 14 EPT1-CPT1 chimeric gene constructs and the parental wild-type genes were expressed in a cpt1 ept1 double null mutant background completely devoid of phosphoamino alcohol transferase activity. Eleven of the chimeric genes expressed functional enzymes. The CDP-amino alcohol and sn-1,2-diacylglycerol (DAG) substrate specificities and essential phospholipid cofactor requirements of the parental and chimeric enzymes were investigated using a mixed micellar assay system. Chimeric enzymes exhibited a pattern of CDP-amino alcohol affinities that defined a structural domain sufficient to confer CDP-amino alcohol specificity. When wild-type enzymes were investigated using a chemically defined series of DAGs, each possessed a distinct characteristic pattern of utilization. Chimeric enzymes exhibited DAG acyl chain specificity profiles that either conformed to parental wild-type patterns or represented novel substrate specificities. Correlation of these outcomes with their underlying structural modifications permitted the assignment of an internal, linear region of 218 amino acids sufficient to confer DAG acyl chain specificity; this region contained three predicted transmembrane segments. Neither wild-type enzyme showed significant acyl chain selectivity with respect to phospholipid activation when a homologous series of chemically defined phosphatidylcholines were employed, suggesting that enzyme recognition of the fatty acyl moieties of the DAG substrate and phospholipid activator is fundamentally different. Analysis of chimeric enzymes dependence on phospholipid activators suggested the involvement of discontinuous protein segments participating in the interaction with phospholipid cofactors.

Lysophosphatidylcholine as an intermediate in phosphatidylcholine metabolism and glycerophosphocholine synthesis in cultured cells: an evaluation of the roles of 1-acyl- and 2-acyl-lysophosphatidylcholine.

Previous studies in our laboratory have shown that the principal pathway of phosphatidylcholine (PtdCho) degradation in cultured mouse N1E-115 neuroblastoma, C6 rat glioma, primary rat brain glia and human fibroblasts is PtdCho----lysophosphatidylcholine (lysoPtdCho)----glycerophosphocholine (GroPCho)----glycerophosphate plus choline (Morash, S.C. et al. (1988) Biochim. Biophys. Acta 961, 194-202). GroPCho is the first quantitatively major degradation product in this pathway, and could be formed by phospholipases A1 or A2, followed by lysophospholipase, or by a co-ordinated attack releasing both fatty acids by phospholipase B. The quality and quantities of lysoPtdCho present in cells reflect the nature of the initial hydrolysis step (A1 or A2), specificities of the lysophospholipases, and activities of acyltransferases that form PtdCho from lysoPtdCho. The present study was undertaken to elucidate the relative importance of these pathways by examining the fate of exogenous 1-acyl and 2-acyl-lysoPtdCho incubated with N1E-115 and C6 cells in culture. By fatty acid composition, endogenous lysoPtdCho was found to be mainly 1-acyl in both cell types based on a predominance of saturated acyl species; this suggested either preferential further deacylation or reacylation of 2-acyl-lysoPtdCho, or that 2-acyl-lysoPtdCho was not formed. Exogenous 1- and 2-acyl-lysoPtdCho specifically radiolabelled with choline and/or fatty acid were incubated either singly or as equimolar mixtures with cells. Cell association was rapid and not reversible by washing and both species were taken up at similar rates. The 2-acyl species was acylated to PtdCho faster than the 1-acyl species in both cell lines. Acylation of both lyso species was higher in C6 compared to N1E-115 cells. Hydrolysis of lysoPtdCho to GroPCho was higher in N1E-115 cells and with 1-acyl-lysoPtdCho. Transacylation between two molecules of lysoPtdCho was a minor pathway. These results document the variety and relative importance of reactions of lysoPtdCho metabolism; under similar conditions, 1- and 2-acyl-lysoPtdCho are handled differently. Both species turn over actively, but only the 1-acyl species accumulates while 2-acyl-lysoPtdCho is likely to be reacylated to form PtdCho.

Phosphatidylcholine biosynthesis in cultured glioma cells: evidence for channeling of intermediates.

The major pathway of choline (Cho) incorporation into phosphatidylcholine (PtdCho) in mammalian cells is sequential conversion of Cho to phosphocholine (PCho), cytidinediphosphate choline (CDP-Cho) and PtdCho. In intact cells, this sequence is usually demonstrated using radiolabeled Cho since PCho and CDP-Cho do not enter the cell intact. We have studied the incorporation of radiolabeled Cho, PCho and CDP-Cho into rat glioma (C6) cells following electropermeabilization. C6 cells were permeable as judged by [U-14C]sucrose and Erythrosin B uptake and more rapid incorporation of [1,2,3-3H]glycerol into cell lipids, and viable as assessed by uptake and incorporation of [methyl-3H]Cho, [1-14C]oleate and [1,2,3-3H]glycerol into complex lipids. Despite rapid incorporation of [methyl-3H]Cho into PtdCho in permeabilized cells, there was no incorporation of [methyl-14C]PCho or CDP-[methyl-14C]Cho into PtdCho. PCho (300 microM) and CDP-Cho (300 microM) failed to significantly reduce incorporation of 28 microM [methyl-3H]Cho into PtdCho. Radioactivity in PtdCho of cells prelabeled with [methyl-3H]Cho prior to permeabilization could be chased with 4 mM Cho but not with 4 mM PCho or 4 mM CDP-Cho. The water-soluble products of Cho metabolism--PCho, CDP-Cho and glycerophosphocholine--were retained at 37 degrees C in permeabilized cells compared with controls while there was uniform leakage from permeabilized cells at 4 degrees C. Hemicholinium-3, an inhibitor of high-affinity Cho transport, decreased [methyl-3H]Cho incorporation into PtdCho in permeabilized cells, as in controls, suggesting that even in permeabilized cells, Cho incorporation into PtdCho is linked to the transport system. We propose that individual steps of the cytidine pathway of PtdCho biosynthesis are functionally linked and that reaction intermediates are not freely diffusible within the cell but are channeled to PtdCho biosynthesis.

Phosphatidylcholine metabolism in cultured cells: catabolism via glycerophosphocholine.

The catabolism of phosphatidylcholine (PtdCho) has been studied in cultured murine neuroblastoma (N1E-115), C6 glioma, rat brain primary glia, and human fibroblast cells. Cells were pulse labelled for 96 h with [methyl-3H]choline followed by a chase for up to 24 h in medium containing 4 mM choline. Measurement of the radioactivity and mass of choline-containing compounds in these cells indicated that the major degradative pathway is PtdCho----lysophosphatidylcholine (lysoPtdCho)----glycerophosphocholine (GroPCho)----choline. At all times during the chase, PtdCho, sphingomyelin and lysoPtdCho comprised 72-92% of the cell-associated radioactivity; the remaining 10-30% was water-soluble and was chiefly GroPCho (30-80%) in all cell lines. In fibroblasts, however, phosphocholine (PCho) was also a major labelled water-soluble component (33-54%). The specific activity of GroPCho closely parallelled that of PtdCho in fibroblasts, but decreased faster than PtdCho in C6 and N1E-115 cells. We postulate that this may be due to distinct pools of PtdCho in the cell with differing rates of turnover. The changes in specific activity of PCho suggest that the major portion is formed by synthesis rather than as a degradative product. However, the inability to reduce the specific activity of this fraction to that of the intracellular choline suggests that a portion may be derived from either PtdCho or GroPCho.