The recovery of the parameters of the kinetic properties of carnitine palmitoyltransferase (CPT) I in liver mitochondria of starved rats was studied after re-feeding animals for various periods of time. There were no significant changes either in the activity of the enzyme at high palmitoyl-CoA concentrations or in the affinity of the enzyme for palmitoyl-CoA, or in the sensitivity of CPT I to malonyl-CoA inhibition after 3 h or 6 h re-feeding. After 24 h re-feeding, both the affinity of the enzyme for palmitoyl-CoA and the activity of the enzyme were still not significantly different from those for the enzyme in mitochondria from 24 h-starved animals. By contrast, the sensitivity of CPT I to malonyl-CoA inhibition was largely, but not fully, restored to that observed in mitochondria from fed rats.

We have tested the possibility that alterations in the fluidity of the outer membrane of rat liver mitochondria could result in changes in the sensitivity of overt carnitine palmitoyltransferase (CPT I) to malonyl-CoA [Zammit (1986) Biochem. Soc. Trans. 14,[676][677][678][679]. The sensitivity of CPT I to malonyl-CoA inhibition was measured by using highly purified mitochondrial outer membranes prepared from fed or 48 h-starved rats in the presence and absence of agents that increase membrane fluidity by perturbing membrane lipid order [benzyl alcohol, isoamyl alcohol (3-methylbutan-1-ol) and 2-(2-methoxyethoxy)ethyl-8-(cis-2-n-octylpropyl)octanoate (A2C)]. All these agents resulted in marked decreases in the ability of malonyl-CoA to inhibit CPT I. This effect was accompanied by a modest increase in the absolute activity of CPT I in the absence ofmalonyl-CoA when the short-chain alcohols were used, but not when A2C was used, suggesting that the effect of increased membrane fluidity to decrease the malonyl-CoA sensitivity of CPT I may occur independently from other actions that may affect more directly the active site of the enzyme. In confirmation of the potential importance of fluidity changes, we showed that a marked increase in sensitivity of CPT I to malonyl-CoA could be produced when assays were performed at lower temperatures than those normally employed. These observations are discussed in the context of the slowness of the changes in CPT I sensitivity to malonyl-CoA inhibition that are induced by physiological perturbations. INTRODUCTIONThe sensitivity of mitochondrial outer-membrane carnitine palmitoyltransferase (CPT I) to malonyl-CoA inhibition changes with physiological state in rat liver (see [1] for review). Under conditions of increased fatty acid oxidation, e.g. starvation, diabetes, the enzyme becomes less sensitive to malonyl-CoA inhibition. This adaptation amplifies the effect of changes in hepatic malonyl-CoA concentration, as decreases in malonylCoA content of the liver are accompanied by decreased sensitivity [2]. Moreover, the two parameters appear to be positively related

Carnitine palmitoyltransferase I (CPT I) is a key enzyme in the regulation of beta-oxidation. The topology of this enzyme has been difficult to elucidate by biochemical methods. We studied the topology of a fusion protein of muscle-type CPT I (M-CPT I) and green fluorescent protein (GFP) by microscopical means. To validate the use of the fusion protein, designated CPT I-GFP, we checked whether the main characteristics of native CPT I were retained. CPT I-GFP was expressed in HeLa cells after stable transfection. Confocal laser scanning microscopy in living cells revealed an extranuclear punctate distribution of CPT I-GFP, which coincided with a mitochondrial fluorescent marker. Immunogold electron microscopy localized CPT I-GFP almost exclusively to the mitochondrial periphery and showed that the C-terminus of CPT I must be on the cytosolic face of the mitochondrial outer membrane. Western analysis showed a protein that was 6 kDa smaller than predicted, which is consistent with previous results for the native M-CPT I. Mitochondria from CPT I-GFP-expressing cells showed an increased CPT activity that was inhibited by malonyl-CoA and was lost on solubilization with Triton X-100. We conclude that CPT I-GFP adopts the same topology as native CPT I and that its C-terminus is located on the cytosolic face of the mitochondrial outer membrane. The evidence supports a recently proposed model for the domain structure of CPT I based on biochemical evidence.

The release of carnitine palmitoyltransferase (CPT) activity from rat liver mitochondria by increasing concentrations of digitonin was studied for mitochondrial preparations from fed, 48 h-starved and diabetic animals. A bimodal release of activity was observed only for mitochondria isolated from starved and, to a lesser degree, from diabetic rats, and it appeared to result primarily from the enhanced release of approx. 40% and 60%, respectively, of the total CPT activity. This change in the pattern of release was specific to CPT among the marker enzymes studied. For all three types of mitochondria there was no substantial release of CPT concurrently with that of the marker enzyme for the soluble intermembrane space, adenylate kinase. These results illustrate that the bimodal pattern of release of CPT reported previously for mitochondria from starved rats [Bergstrom & Reitz (1980) Arch. Biochem. Biophys. 204, 71-79] is not an immutable consequence of the localization of CPT activity on either side of the mitochondrial inner membrane. Sequential loss of CPT I (i.e. the overt form) from the mitochondrial inner membrane did not affect the concentration of malonyl-CoA required to effect fractional inhibition of the CPT I that remained associated with the mitochondria. The results are discussed in relation to the possibility that altered enzyme-membrane interactions may account for some of the altered regulatory properties of CPT I in liver mitochondria of animals in different physiological states.

1. The synthesis and secretion of glycerolipid by monolayer cultures of rat hepatocytes was measured by determining the incorporations of [3H]glycerol, [3H]oleate and [14C]choline and by the absolute concentration of triacylglycerol. 2. The presence of albumin in the medium stimulated the accumulation of lysophosphatidylcholine in the medium by 11-13-fold. 3. Cycloheximide did not significantly alter the accumulation of lysophosphatidylcholine. 4. This process was particularly sensitive to inhibition by chlorpromazine and verapamil, compared with the secretion of triacylglycerol and phosphatidylcholine. By contrast, it was relatively less sensitive to EGTA. 5. It is suggested that intracellular Ca2+ may be important in the production of lysophosphatidylcholine, which then accumulates in the medium by binding to albumin. In vivo this lysophosphatidycholine may be a means of delivering choline and polyunsaturated fatty acids to other organs.