We have developed a procedure for the reconstitution of Escherichia coli diacylglycerol kinase (DGK) into phospholipid bilayers containing diacylglycerol substrate. When DGK is reconstituted into a series of phosphatidylcholines containing monounsaturated fatty acyl chains, activity against dihexanoylglycerol (DHG) as a substrate was found to be markedly dependent on the fatty acyl chain length with the highest activity in dioleoylphosphatidylcholine [di(C18:1)PC] and a lower activity in bilayers with shorter or longer fatty acyl chains. Low activities in the short chain phospholipid dimyristoleoylphosphatidylcholine [di(C14:1)PC] followed from an increase in the K(m) value for DHG and ATP, with no effect on v(max). In contrast, in the long chain lipid dierucoylphosphatidylcholine [di(C24:1)PC], the low activity followed from a decrease in v(max) with no effect on K(m). In mixtures of two phosphatidylcholines with different chain lengths, the activity corresponded to that expected for the average chain length of the mixture. Cholesterol increased the activity in di(C14:1)PC but slightly decreased it in di(C18:1)PC or di(C24:1)PC, effects that could follow from changes in bilayer thickness caused by cholesterol.

Diacylglycerol kinase (DGK) of Escherichia coli has been reconstituted into a variety of phospholipid bilayers and its activity determined as a function of lipid headgroup structure and phase preference. The anionic phospholipids dioleoylphosphatidic acid, dioleoylphosphatidylserine, and cardiolipin were all found to support activities lower than that supported by dioleoylphosphatidylcholine. In mixtures of dioleoylphosphatidylcholine and 20 mol % anionic phospholipids, the presence of anionic phospholipids all resulted in lower activities than in dioleoylphosphatidylcholine, except for dioleoylphosphatidylglycerol whose presence had little effect on activity. In some cases, the low activity in the presence of anionic phospholipid followed from a decrease in v(max); in some cases, it followed from an increase in the K(m) for diacylglycerol, and in the case of dioleoylphosphatidic acid, it followed from both. Activities in mixtures containing 80 mol % dioleoylphosphatidylethanolamine were lower than in dioleoylphosphatidylcholine at temperatures where both lipids adopted a bilayer phase; at higher temperatures where dioleoylphosphatidylethanolamine preferred a hexagonal H(II) phase, the differences in activity were greater. These experiments suggest that the presence of lipids preferring a hexagonal H(II) phase leads to low activities. Activities of DGK are low in a gel phase lipid.

Accumulation of Ca2+ by the Ca2+-ATPase of skeletal-muscle sarcoplasmic reticulum has been measured in reconstituted, sealed vesicles as a function of lipid composition. Measurements were performed in the presence of carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP) to eliminate any effects of H+ transport; in the presence of FCCP, addition of valinomycin had no effect on the level or rate of accumulation of Ca2+ showing that, in the presence of FCCP, no electrical potential built up across the membrane. Levels of accumulation were low when the phospholipid was dioleoylphosphatidylcholine (DOPC), even though DOPC supports high ATPase activity. Inclusion of 10 mol% anionic phospholipid [dioleoylphosphatidic acid (DOPA) or dioleoylphosphatidylserine (DOPS)] led to higher levels of accumulation of Ca2+, 10 mol% being the optimum concentration. Cardiolipin or phosphatidylinositol 4-phosphate were more effective than DOPA or DOPS in increasing accumulation of Ca2+. Effects of anionic phospholipids were seen in the presence of an ATP-regenerating system to remove ADP, and in the presence of phosphate within the reconstituted vesicles to precipitate calcium phosphate. Rates of passive leak of Ca2+ from the reconstituted vesicles were slow. The Ca2+-accumulation process was simulated assuming either simple passive leak of Ca2+ from the vesicles or assuming slippage on the ATPase, a process in which the phosphorylated intermediate of the ATPase releases bound Ca2+ on the cytoplasmic rather than the lumenal side of the membrane. The experimental data fitted to a slippage model, with anionic phospholipids decreasing the rate of slippage.

The major phospholipid found in mammalian membranes is the bilayer-favouring, zwitterionic, phosphatidylcholine. However, membranes also contain phosphatidylethanolamine, which favours a non-bilayer structure, and a variety of anionic phospholipids including phosphatidylserine and the phosphatidylinositols. The roles of these lipids can be studied in simplified, model systems in which the Ca2+-ATPase purified from skeletal muscle sarcoplasmic reticulum is reconstituted into phospholipid bilayers of defined composition [l]. If the reconstitution is performed to give membrane fragments, the scalar process of ATP hydrolysis by the ATPase can be studied uncomplicated by vectorial processes such as accumulation of Caz+. These studies show that ATPase activity is highest in bilayers of dioleoylphosphatidylcholine (DOPC) and that dioleoy lphosphatidic acid (DOPA) or dioleoy lphosphatidykerine (DOPS) inhibit ATPase activity at high concentrations, whilst having no effect at low concentrations [2]. Fluorescence quenching studies show that DOPS and DOPA bind to the ATPase with affities equal to that for DOPC, and suggest that the anionic lipids only bind to the ATPase at the bulk lipid sites around the ATPase. Time (secs) Figure 1. The level of accumulation of Ca2+ recorded for the ATPase reconstituted into vesicles of DOPC (bottom trace), DOPC + I 0 nwl % DOPA (middle trace) and DOPC + 10 rnol % cardiolipin (top trace). The solid lines show the experimental data. and the broken lines are simulations with parameters as described in the text.1 El-El--ElCa,ATP I I indd. * E2-E2p, -E2P -E2Pca, t I The ATPase can also be reconstituted into sealed vesicles, allowing measurements of ATPdriven accumulation of Ca2 + .Surprisingly, uptake of Ca2+ into vesicles of DOPC is low, but is increased significantly in the presence of 10 mol % DOPA or DOPS; levels of anionic lipid higher than 10 mol % have no additional effect (Figure 1). Sizes of the reconstituted vesicles are not affected significantly by the presence of anionic lipid. Transport of Caz+ into sealed vesicles is accompanied by a build-up of a membrane potential, inside positive, and transport of H+ out of the vesicles. Effects of anionic lipid are seen in the presence of valinomycin and K+ (to collapse membrane potentials) and in the presence of FCCP (to equilibrate H+ across the membrane). The presence of phosphate trapped within the vesicles leads to increased levels of accumulation of Caz+; effects of phosphate are more marked in the presence than in the absence of anionic lipid. These experiments suggest both that the effect of anionic lipid cannot be to bind the accumulated Caz+ (leading to reduction in the free Caz+ concentration in the vesicle lumen) and that there is a significant rate of Caz+ leak from the vesicles.Simulations have been performed according to the model shown in Scheme 1. The pattern of accumulation of Caz+ cannot be reproduced if it is assumed that Caz+ leak follows a simple, passive pathway -in a simple leak model, Caz+ accumulation increases almos...