Reconstitution of purified, active and malonyl-CoA-sensitive rat liver carnitine palmitoyltransferase I: relationship between membrane environment and malonyl-CoA sensitivity

Abstract

Carnitine palmitoyltransferase I (CPT I) catalyses the initial step of fatty acid import into the mitochondrial matrix, the site of β-oxidation, and its inhibition by malonyl-CoA is a primary control point for this process. The enzyme exists in at least two isoforms, denoted L-CPT I (liver type) and M-CPT I (skeletal-muscle type), which differ in their kinetic characteristics and tissue distributions. A property apparently unique to L-CPT I is that its sensitivity to malonyl-CoA decreases in vivo with fasting or experimentally induced diabetes. The mechanism of this important regulatory effect is unknown and has aroused much interest. CPT I is an integral outer-membrane protein and displays little activity after removal from the membrane by detergents, precluding direct purification of active protein by conventional means. Here we describe the expression of a 6×His-tagged rat L-CPT I in Pichia pastoris and purification of the detergent-solubilized enzyme in milligram quantities. Reconstitution of the purified product into a liposomal environment yielded a 200-400-fold increase in enzymic activity and restored malonyl-CoA sensitivity. This is the first time that a CPT I protein has been available for study in a form that is both pure and active. Comparison of the kinetic properties of the reconstituted material with those of L-CPT I as it exists in mitochondria prepared from yeast over-expressing the enzyme and in livers from fed or fasted rats permitted novel insight into several aspects of the enzyme's behaviour. The malonyl-CoA response of the liposomal enzyme was found to be greater when the reconstitution procedure was carried out at 22 °C compared with 4 °C (IC50 ≈ 11 μM versus 30 μM, respectively). When the sensitivities of L-CPT I in each of the different environments were compared, they were found to decrease in the following order: fed liver > fasted liver≈ liposomes prepared at 22 °C≈ P. pastoris mitochondria > liposomes prepared at 4 °C. In addition, pre-treatment of L-CPT I liposomes with the membrane-fluidizing reagent benzyl alcohol caused densensitization to the inhibitor. In contrast with the variable response to malonyl-CoA, the liposomal L-CPT I displayed a pH profile and kinetics with regard to the carnitine and acyl-CoA substrates similar to those of the enzyme in fed or fasted liver mitochondria. However, despite a normal sensitivity to malonyl-CoA, L-CPT I in P. pastoris mitochondria displayed aberrant behaviour with regard to each of these other parameters. The kinetic data establish several novel points. First, even after stringent purification procedures in the presence of detergent, recombinant L-CPT I could be reconstituted in active, malonyl-CoA sensitive form. Second, the kinetics of the reconstituted, 6×His-tagged L-CPT I with regard to substrate and pH responses were similar to what is observed with rat liver mitochondria (whereas in P. pastoris mitochondria the enzyme behaved anomalously), confirming that the purified preparation is a suitable model for studying the functional properties of the enzyme. Third, wide variation in the response to the inhibitor, malonyl-CoA, was observed depending only on the enzyme's membrane environment and independent of interaction with other proteins. In particular, the fluidity of the membrane had a direct influence on this parameter. These observations may help to explain the mechanism of the physiological changes in the properties of L-CPT I that occur in vivo and are consistent with the current topographical model of the enzyme.