Interactions of acyl carnitines with model membranes
2002; Elsevier BV; Volume: 43; Issue: 9 Linguagem: Inglês
10.1194/jlr.m200137-jlr200
ISSN1539-7262
AutoresJet Ho, R. Duclos, James A. Hamilton,
Tópico(s)Infectious Encephalopathies and Encephalitis
ResumoEsterification of fatty acids with the small polar molecule carnitine is a required step for the regulated flow of fatty acids into mitochondrial inner matrix. We have studied the interactions of acyl carnitines (ACs) with model membranes [egg yolk phosphatidylcholine (PC) vesicles] by 13C-nuclear magnetic resonance (NMR) spectroscopy. Using AC with 13C-enrichment of the carbonyl carbon of the acyl chain, we detected NMR signals from AC on the inside and outside leaflets of the bilayer of small unilamellar vesicles prepared by cosonication of PC and AC. However, when AC was added to the outside of pre-formed PC vesicles, only the signal for AC bound to the outer leaflet was observed, even after hours at equilibrium. The extremely slow transmembrane diffusion ("flip-flop") is consistent with the zwitterionic nature of the carnitine head group and the known requirement of transport proteins for movement of ACs through the mitochondrial membrane. The partitioning of ACs (8–18 carbons) between water and PC vesicles was studied by monitoring the [13C]carbonyl chemical shift of ACs as a function of pH and concentration of vesicles. Significant partitioning into the water phase was detected for ACs with chain lengths of 12 carbons or less. The effect of ACs on the integrity of the bilayer was examined in vesicles with up to 25 mol% myristoyl carnitine; no gross disruption of the bilayer was observed.We hypothesize that the effects of high levels of long-chain AC (as found in ischemia or in certain diseases) on cell membranes result from molecular effects on membrane functions rather than from gross disruption of the lipid bilayer. Esterification of fatty acids with the small polar molecule carnitine is a required step for the regulated flow of fatty acids into mitochondrial inner matrix. We have studied the interactions of acyl carnitines (ACs) with model membranes [egg yolk phosphatidylcholine (PC) vesicles] by 13C-nuclear magnetic resonance (NMR) spectroscopy. Using AC with 13C-enrichment of the carbonyl carbon of the acyl chain, we detected NMR signals from AC on the inside and outside leaflets of the bilayer of small unilamellar vesicles prepared by cosonication of PC and AC. However, when AC was added to the outside of pre-formed PC vesicles, only the signal for AC bound to the outer leaflet was observed, even after hours at equilibrium. The extremely slow transmembrane diffusion ("flip-flop") is consistent with the zwitterionic nature of the carnitine head group and the known requirement of transport proteins for movement of ACs through the mitochondrial membrane. The partitioning of ACs (8–18 carbons) between water and PC vesicles was studied by monitoring the [13C]carbonyl chemical shift of ACs as a function of pH and concentration of vesicles. Significant partitioning into the water phase was detected for ACs with chain lengths of 12 carbons or less. The effect of ACs on the integrity of the bilayer was examined in vesicles with up to 25 mol% myristoyl carnitine; no gross disruption of the bilayer was observed. We hypothesize that the effects of high levels of long-chain AC (as found in ischemia or in certain diseases) on cell membranes result from molecular effects on membrane functions rather than from gross disruption of the lipid bilayer. Acyl carnitines (ACs) participate in a crucial step in the shuttling of fatty acids and breakdown products of carbohydrates into the inner mitochondrial matrix for oxidative phosphorylation. Carnitine, a small molecule with a nontitratable positive charge at the quaternary amine position and a titratable carboxyl group, is covalently linked to fatty acid to form ACs (Fig. 1). Therefore, ACs are zwitterionic in aqueous solution at neutral pH. Two AC transferases (CPT I and CPT II) and a carnitine-AC translocase located within the inner mitochondrial membrane control the conversion between AC and acyl-CoA and regulate the flux of AC into the inner matrix (1Fritz I.B. Martelli E.A. Sites of action of carnitine and its derivatives on the cardiovascular system: interactions with membranes.Trends Pharmacol. Sci. 1993; 14: 355-360Google Scholar, 2Kerner J. Hoppel C. Fatty acid import into mitochondria.Biochim. Biophys. Acta. 2000; 1486: 1-17Google Scholar). This transport is regulated by both fatty acid and carbohydrate metabolism. ACs are also formed in the peroxisome as products of partial β-oxidation of long- and very-long-chain fatty acids (3Hettema E.H. Tabak H.F. Transport of fatty acids and metabolites across the peroxisomal membrane: evidence for a short-chain acyl-carnitine translocase in mitochondria specifically related to the metabolism of short-chain fatty acids.Biochim. Biophys. Acta. 2000; 1486: 18-27Google Scholar). The medium-chain ACs formed inside the peroxisome from partial degradation of very-long-chain saturated fatty acids and methyl-branched fatty acids are exported by a translocase. At the mitochondrion, these ACs bypass CPT I and go directly to a translocase (possibly a different translocase from that for long-chain unbranched fatty acids) in the inner membrane (4Roe D.S. Roe C.R. Brivet M. Sweetman L. Evidence for a short-chain carnitine-acylcarnitine translocase in mitochondria specifically related to the metabolism of branched-chain amino acids.Mol. Genet. Metab. 2000; 69: 69-75Google Scholar). Concentrations of carnitine and AC in cells and serum are influenced by a host of normal physiological and pathophysiological processes. Increased levels of AC in the blood and urine are associated with most inborn errors of fatty acid oxidation (5Schmidt-Sommerfeld E. Penn D. Duran M. Rinaldo P. Bennett M.J. Santer R. Stanley C.A. Detection and quantitation of acylcarnitines in plasma and blood spots from patients with inborn errors of fatty acid oxidation.Prog. Clin. Biol. Res. 1992; 375: 355-362Google Scholar). The concentration of short-chain ACs in cells increases and that of long-chain ACs decreases with exercise, although the ratio between total AC and acyl-CoA remains constant (6Friolet R. Hoppeler H. Krahenbuhl S. Relationship between the coenzyme A and the carnitine pools in human skeletal muscle at rest and after exhaustive exercise under normoxic and acutely hypoxic conditions.J. Clin. Invest. 1994; 94: 1490-1495Google Scholar). Under ischemic and hypoxic conditions, levels of long-chain ACs may increase 8- to 10-fold in myocytes and may be elevated up to 100-fold in sarcolemma (7Vogel W.M. Bush L.R. Cavallo G.C. Heathers G.P. Hirkaler G.M. Kozak M.Z. Higgins A.J. Inhibition of long-chain acylcarnitine accumulation during coronary artery occlusion does not alter infarct size in dogs.J. Cardiovasc. Pharmacol. 1994; 23: 826-832Google Scholar). Such levels of long-chain ACs elicit adverse pharmacological effects on cellular functions and make the heart more susceptible to arrhythmias and other dysfunction (8Dumonteil E. Barre H. Meissner G. Effects of palmitoyl carnitine and related metabolites on the avian Ca(2+)-ATPase and Ca2+ release channel.J. Physiol. 1994; 479: 29-39Google Scholar). These pathological effects may be caused by the binding of ACs to cell membranes and alteration of their functions. Limiting the formation of long-chain ACs by inhibiting CPT I might have protective effects against myocardial ischemia, although inhibiting CPT I did not reduce infarct size (7Vogel W.M. Bush L.R. Cavallo G.C. Heathers G.P. Hirkaler G.M. Kozak M.Z. Higgins A.J. Inhibition of long-chain acylcarnitine accumulation during coronary artery occlusion does not alter infarct size in dogs.J. Cardiovasc. Pharmacol. 1994; 23: 826-832Google Scholar, 9Madden M.C. Wolkowicz P.E. Pohost G.M. McMillin J.B. Pike M.M. Acylcarnitine accumulation does not correlate with reperfusion recovery in perfused rat hearts.Am. J. Physiol. 1995; 268: H2505-H2512Google Scholar). Malfunctions in the carnitine metabolic pathway can result in severe metabolic dysfunction, with poor clinical outcomes and early morbidity. Carnitine deficiency can cause a debilitating disease involving muscular and myocardial function by affecting fatty acid metabolism in cells (10Tien I. De Vivo D.C. Bierman F. Pulver P. De Meirleir L.J. Cvitanovic-Sojat L. Pagon R.A. Bertini E. Dionisi-Vici C. Servidei S. Impaired skin fibroblast carnitine uptake in primary systemic carnitine deficiency manifested by childhood carnitine-responsive cardiomyopathy.Pediatr. Res. 1990; 28: 247-255Google Scholar). Inborn errors of fatty acid oxidation occur with one in 10,000 to 15,000 live births (11Kelly D.P. Strauss A.W. Inherited cardiomyopathies.N. Engl. J. Med. 1994; 330: 913-919Google Scholar). The acyl-chain profile of ACs in fibroblasts of such patients is often sufficiently distinct to permit localization of the specific enzyme defect (12Schmidt-Sommerfeld E. Bobrowski P.J. Penn D. Rhead W.J. Wanders R.J. Bennett M.J. Analysis of carnitine esters by radio-high performance liquid chromatography in cultured skin fibroblasts from patients with mitochondrial fatty acid oxidation disorders.Pediatr. Res. 1998; 44: 210-214Google Scholar). Some of these infantile or juvenile onset myopathies, usually associated with mental retardation or regressions, are caused by defects in carnitine transport, carnitine-acylcarnitine translocase, and carnitine palmityl transferase (13Stanley C.A. Hale D.E. Berry G.T. Deleeuw S. Boxer J. Bonnefont J.P. Brief report: a deficiency of carnitine-acylcarnitine translocase in the inter mitochondrial membrane.N. Engl. J. Med. 1992; 327: 19-23Google Scholar). Deficiencies in serum AC sometimes are found in patients with chronic fatigue syndrome (14Kuratsune H. Yamaguti K. Takahashi M. Misaki H. Tagawa S. Kitani T.A. Acylcarnitine deficiency in chronic fatigue syndrome.Clin. Infect. Dis. 1994; 18: S62-S67Google Scholar). Despite the importance of ACs in human health and disease, physical studies of these common molecules have been sparse. The acyl chain alters the physical properties of carnitine and converts this water-soluble molecule into a molecule that, at a certain chain length, will partition favorably into the phospholipid bilayer of membranes. An understanding of how ACs interact with membranes is crucial for understanding the biophysical basis of the normal and pathological effects of these molecules. Previous studies have focused mainly on surfactant properties of ACs, such as their abilities to form micelles and to disrupt phospholipid bilayer structure. For example, surface-active properties of palmitoyl carnitine have been shown to lead to detergent effects at very high concentrations (e.g., 1 mol of AC per mol of phospholipid) (15Requero M.A. Goni F.M. Alonso A. The membrane-perturbing properties of palmitoyl-coenzyme A and palmitoylcarnitine.Biochemistry. 1995; 34: 10400-10405Google Scholar, 16Goni F.M. A Requero M. Alonso A. Palmitoylcarnitine, a surface-active metabolite.FEBS Lett. 1996; 390: 1-5Google Scholar). Limited information is available about the molecular interactions of ACs with membranes at physiological concentrations (17Haeyaert P. Verdonck A. Van Cauwelaert F.H. Influence of acylcarnitines of different chain length on pure and mixed phospholipid vesicles and on sarcoplasmic reticulum vesicles.Chem. Phys. Lipids. 1987; 45: 49-63Google Scholar, 18Echabe I. Requero M.A. Goni F.M. Arrondo J.L. Alonso A. An infrared investigation of palmitoyl-coenzyme A and palmitoylcarnitine interaction with perdeuterated phospholipid bilayers.Eur. J. Biochem. 1995; 231: 199-203Google Scholar). The thermodymanics and kinetics of transfer of pyrene-labeled AC between model and natural membranes were studied by fluorescence spectroscopy (19Wolkowicz P.E. Pownall H.J. Pauly D.F. McMillin-Wood J.B. Pyrenedodecanoylcarnitine and pyrenedodecanoyl coenzyme A: kinetics and thermodynamics of their intermembrane transfer.Biochemistry. 1984; 23: 6426-6432Google Scholar). A model for medium-chain AC, (1-pyrenebutyryl) carnitine showed interesting properties of transmembrane movement: as assessed by its extraction from phospholipid bilayer vesicles, it moved from the outer to the inner leaflet rapidly but appeared to be trapped in the inner leaflet (20Wolkowicz P.E. Pownall H.J. McMillin-Wood J.B. (1-Pyrenebutyryl) carnitine and 1-pyrenebutyryl coenzyme A: flouorescent probes for lipid metabolite studies in artificial and natural membranes.Biochemistry. 1982; 22: 2990-2998Google Scholar). Although definitive data about the transmembrane movement of natural ACs in model membranes are lacking, there is direct (21Classen J. Deuticke B. Haest C.W. Nonmediated flip-flop of phospholipid analogues in erythrocyte membrane as probed by palmitoylcarnitine: basic properties and influence of membrane modification.J. Membr. Biol. 1989; 111: 169-178Google Scholar) and indirect (22Arduini A. Mancinelli G. Radatti G.L. Dottori S. Molajoni F. Ramsay R.R. Role of carnitine and carnitine palmitoyltransferase as integral components of the pathway for membrane phospholipid fatty acid turnover in intact human erythrocytes.J. Biol. Chem. 1992; 267: 12673-12681Google Scholar) evidence that palmitoyl carnitine has a slow rate of transbilayer movement in erythrocyte membranes. Our study focuses on the binding of ACs with different chain lengths (8–18 carbons) to model membranes composed of phospholipid bilayers. Labeling with 13C permits the detection of low amounts of AC in model membranes and preserves the native structure of the lipid. We used 13C-NMR spectroscopy of carbonyl-labeled AC to address the questions of i) whether ACs bind to both leaflets of a small unilamellar vesicle; ii) whether ACs cross the lipid bilayer spontaneously in either direction; iii) how the acyl-chain length affects partitioning into the membrane; and iv) whether low levels of AC are disruptive to the bilayer structure. A series of six [octanoyl (8:0), decanoyl (10:0), lauroyl (12:0), myristoyl (14:0), palmitoyl (16:0), and oleoyl (18:1)] dl-O-acylcarnitine-(1′-13C) chloride analogs were prepared on scales of 100 to 300 mg from the corresponding carbonyl-labeled (99 atom% 13C) fatty acid (Cambridge Isotope Laboratories, Woburn, MA) and thionyl chloride followed by treatment of the intermediate fatty acid chloride with dl-carnitine chloride in trichloroacetic acid as solvent according to the reported method (23Ziegler H.J. Bruckner P. Binon F. O-acylation of dl-carnitine chloride.J. Org. Chem. 1967; 32: 3989-3991Google Scholar). The saturated fatty acyl derivatives were each recrystallized three times from isopropanol/acetone. The unsaturated oleoyl carnitine chloride was purified by a slight modification of the reported procedure (24Bremer J. Long-chain acylcarnitines.Biochem. Prep. 1968; 12: 69-73Google Scholar). The crude oleoyl carnitine chloride was washed with hexane followed by anhydrous diethyl ether and was then filtered through silica gel 60 (E. Merck), eluting with 9:1 chloroform-methanol (v/v). After these steps, three recrystallizations from methanol/anhydrous diethyl ether gave 143 mg (85% yield) of a white solid that melted/decomposed at 112°C to 114°C. The saturated AC chloride analogs melted/decomposed just above 150°C. The 13C-labeled AC chlorides were homogeneous by analytical TLC on silica gel 60 (E. Merck), eluting with chloroform-methanol-water (60:30:4, v/v/v). 1H-NMR showed the C3-H proton resonance of carnitine chloride starting material [δ 4.41 (DMSO-d6) or δ 4.68 (D2O)] shifted downfield to δ 5.44 (DMSO-d6) or 5.61 (D2O) for the 13C-labeled acylcarnitine chloride products, which differed from the corresponding unlabeled analogs only by the additional 2J -1H-13C = 7.3 Hz coupling observed for the [C2′-H2]methylene observed at δ 2.32 in DMSO-d6 and at δ 2.42 in D2O. The 13C-NMR spectra of the 13C-labeled AC chlorides showed strong C1′ carbonyl signals, and the acyl group C2′ carbons appeared as doublets (1J -13C-13C =57 Hz) at 34.4 ppm in D2O-H2O (1:3, v/v). Small unilamellar vesicles (SUV) composed of egg yolk phosphatidylcholine (PC) were used as a model system for cell membranes. Egg yolk PC in chloroform was purchased from Avanti Polar Lipid, Alabaster, AL (Lot #EPC-253, EPC-292). Its concentration was determined by evaporating a known amount of PC solution and measuring the dry weight on a Cahn C-31 microbalance. A measured volume of PC was transferred to a round-bottom Kimax centrifuge tube. Chloroform was removed by evaporation under a stream of nitrogen gas and lyophilization under vacuum (10 Torr) overnight. The dried PC sample, hydrated overnight in unbuffered 0.56% (75 mM) KCl solution for pH titration experiments or in 50 mM phosphate buffer solution (pH 7.4) for all other experiments to maintain a stable pH. D2O (20%, v/v), was added to the aqueous sample to provide a lock signal for the NMR experiments. To make AC available to both leaflets of the vesicles, vesicles were prepared either by adding AC to the chloroform solution of PC before lyophilizing or by hydrating PC in an aqueous solution of AC. The hydrated sample (2.0 ml total) was transferred to a thick-walled polycarbonate centrifuge tube for sonication with a Branson Sonifier cell disrupter model 350 equipped with a titanium tip (power level 3; 30% duty cycle) under a stream of nitrogen gas for 60 min to achieve a characteristic slightly translucent, nonturbid suspension. The preparation was then centrifuged on a tabletop centrifuge for 30 min to remove any metallic fragments that had dislodged from the sonication probe. The phospholipid concentration of selected samples was determined by Bartlett analysis (25Bartlett G.R. Phosphorous Assay in Column Chromatography.J. Biol. Chem. 1959; 234: 466-468Google Scholar), and was typically ∼90% of the concentration estimated from the initial weight measurement. Aqueous solutions of AC were prepared by dissolving weighed amounts of crystalline AC in water. Comparisons with samples prepared by cosonication of PC as described above were made by adding dissolved AC at a known concentration to sonicated vesicles. 13C-NMR studies were performed on i) a Bruker AMX-300 with the 10 mm QNP probe (13C = 75 MHz), with data acquired and processed by Bruker software UXNMR, or ii) a Bruker Avance system DMX-500 (13C = 125 MHz) with a 10-mm broadband probe, with data acquired and processed with UXNMR and XwinNMR. Continuous irradiation in the proton frequency provided decoupling and nuclear Overhauser enhancement. The spin lattice relaxation time (T1) was measured at 75 MHz for the carbonyl signal using the inversion-recovery method (26Hamilton J.A. Cistola D.P. Morrisett J.D. Sparrow J.T. Small D.M. Interactions of myristic acid with bovine serum albumin: a 13C NMR study.Proc. Natl. Acad. Sci. USA. 1984; 81: 3718-3722Google Scholar). The chemical-shift reference was tetramethylsilane in CDCl3 in a capillary tube. With such an external referencing configuration, the terminal methyl signal of phospholipids was seen at 13.9 ppm. 13C-NMR spectra were obtained for [13C]carbonyl-enriched ACs with different chain lengths (8–18 carbons) in water over a concentration range that was feasible with a time limit of 24 h for the most dilute samples (∼0.05 mg/ml). A single, narrow (<10 Hz) carboxyl peak was seen for all aqueous samples. The chemical shift and T1 of the carbonyl carbon of each AC were measured (Table 1). The chemical shift of the medium-chain ACs (octanoyl and decanoyl carnitine) was ∼175.5 ppm, and the T1 was ∼7 s at ∼1 mg/ml ( 14 carbons), which are micellar at the concentrations in our study (see below), was 173.5 ppm. Because the [13C]carbonyl is in the head group, we assumed that the length of the acyl chain would not affect its interaction with water in a given state (monomer or micelle) and that the limiting values obtained with shorter- and longer-chain ACs would be applicable to ACs with chain lengths of 12 (lauroyl) and 14 (myristoyl) carbons. At concentrations slightly above the CMC, monomers will be in fast exchange with the micelles and the carbonyl chemical shift is predicted to appear at values between 173.5 and 175.5 ppm. The concentration above which the carbonyl chemical shift first decreases below the value of 175.5 ppm signifies the CMC. Extrapolation of concentration-dependent chemical shift data for myristoyl carnitine (Fig. 1) yielded a value of 0.021 mg/ml, or 0.051 mM, a value somewhat lower than the previous value of 0.1 mM derived from light scattering measurements (29Yalkowsky S.H. Zografi G. Some micellar properties of long-chain acylcarnitines.J. Colloid Interface Sci. 1970; 34: 525-533Google Scholar). Lauroyl carnitine in aqueous solution showed similar concentration-dependent chemical shifts. Because the CMC is much higher than that for myristoyl carnitine, it was feasible to obtain data both below and above the CMC. Assuming that the monomer chemical shift is 175.5 and extrapolating from the chemical shift data (plotted as in Fig. 1), we estimated the CMC of lauroyl carnitine to be 4.3 mg/ml (1.15 mM), close to the published value of 1.2 mM (29Yalkowsky S.H. Zografi G. Some micellar properties of long-chain acylcarnitines.J. Colloid Interface Sci. 1970; 34: 525-533Google Scholar). The use of [13C]carbonyl-labeled AC permitted the detection of low amounts of AC in egg PC SUV (model membranes) by 13C-NMR spectroscopy. The interactions of ACs with vesicles were studied as a function of acyl chain length of AC, concentration of AC, and the solution pH. The chemical shift of a lipid carbonyl group is highly sensitive to its local environment and its proximity to the aqueous-lipid interface, which determines its hydration (27Yeagle P.L. Martin R.B. Hydrogen-bonding of the ester carbonyls in phosphatidylcholine bilayers.Biochem. Biophys. Res. Commun. 1976; 69: 775-780Google Scholar, 28Hamilton J.A. Small D.M. Solubilization and localization of triolein in phosphatidylcholine bilayers; a 13C NMR study.Proc. Natl. Acad. Sci. USA. 1981; 78: 6878-6882Google Scholar). The curvature difference between the outer and the inner leaflets in a small vesicle results in differences in the hydration of phospholipid carbonyls, allowing two [13C]carbonyl signals to be observed for the phospholipid (27Yeagle P.L. Martin R.B. Hydrogen-bonding of the ester carbonyls in phosphatidylcholine bilayers.Biochem. Biophys. Res. Commun. 1976; 69: 775-780Google Scholar, 28Hamilton J.A. Small D.M. Solubilization and localization of triolein in phosphatidylcholine bilayers; a 13C NMR study.Proc. Natl. Acad. Sci. USA. 1981; 78: 6878-6882Google Scholar). The natural-abundance 13C-NMR spectrum shows phospholipids localized in the outer and inner leaflets by peaks at 173.6 and 173.4 ppm (Fig. 2A). The intensity of the phospholipid signal at 173.6 ppm (outer leaflet) is ∼1.5 times that of the peak at 173.4 ppm (inner leaflet). The curvature effect can be exploited to locate other lipids in the two leaflets of vesicles (30Hamilton J.A. Bhamidipati S.P. Kodali D.R. Small D.M. The interfacial conformation and transbilayer movement of diacylglycerols in phospholipid bilayers.J. Biol. Chem. 1991; 266: 1177-1186Google Scholar, 31Cabral D.J. Hamilton J.A. Small D.M. The ionization behavior of bile acids in different aqueous environments.J. Lipid Res. 1986; 27: 334-343Google Scholar, 32Boylan J.G. Hamilton J.A. Interactions of acyl-coenzyme A with phosphatidylcholine bilayers and serum albumin.Biochemistry. 1992; 31: 557-567Google Scholar, 33Bhamidipati S.P. Hamilton J.A. Interactions of lyso 1-palmitoylphosphatidylcholine with phospholipids: a 13C and 31P NMR study.Biochemistry. 1995; 34: 5666-5677Google Scholar). The general features of the 13C-spectra (carbonyl region) of ACs in SUV are illustrated by the examples of decanoyl and lauroyl carnitine. For vesicles made by cosonication of lauroyl carnitine with PC, the intensity of the phospholipid peak at 173.4 ppm increased and a new signal was observed at 172.9 ppm (Fig. 2B). The two signals probably represent two distinct environments for the AC carbonyl group. In comparison, when lauroyl carnitine was added to preformed vesicles (Fig. 2C), the intensity of the signal at 173.4 ppm increased but the signal at 172.9 ppm was absent. The signals from the labeled lauroyl carnitine were visualized more clearly by computing a difference spectrum by subtracting the spectrum of SUV from that of SUV with added AC (Fig. 2, insets). The difference spectrum of the cosonicated preparation showed two distinct signals (Fig. 2B, inset), whereas that of the sample with AC added to preformed vesicles showed only one peak (Fig. 2C, inset). A signal at 173.4 ppm was observed in both preparations, whereas the signal at 172.9 ppm was observed only when lauroyl carnitine was added to preformed vesicles. The carbonyl spectra of decanoyl carnitine in SUV were also different for the two types of sample preparation (Fig. 3). Both preparations showed a signal at 174.0 to 174.1 ppm, whereas only decanoyl carnitine cosonicated with PC showed a signal at 172.9 ppm. The signal of ACo is shifted downfield (174.0 ppm; Fig. 3A) relative to that of lauroyl carnitine (173.4 ppm; Fig. 2) because of partitioning into the aqueous phase, as discussed below. Figure 3 (insets) illustrates difference spectra of the experiments for decanoyl carnitine. The cosonicated preparation showed two distinct signals at 174.0 ppm and 172.9 ppm (Fig. 3A), whereas the sample for which AC was added to preformed vesicles showed only one peak at 174.1 ppm (Fig. 3B). The line widths of the carbonyl peaks were significantly larger than the line width of decanoyl carnitine in aqueous solution without PC. The ratio of the intensities from the decanoyl carnitine signals at 174.0 and 172.9 ppm was measured in the difference spectrum of Fig. 3. The ratio (3:1) is about twice as high as the ratio of the intensities of the phospholipid signals at 173.6 ppm (outer leaflet) and 173.4 ppm (inner leaflet) and shows a preference of AC for the outer leaflet. The spectra of all ACs investigated showed two AC carbonyl signals for samples prepared by cosonication and one signal for samples prepared by adding AC to vesicles. These sites are assigned to AC in the outer (downfield peak) and inner (upfield peak) leaflet of the vesicle. The two signals are separated by 1.0 ppm (75 Hz), indicating that the exchange rate must be no greater than 34 s−1. The ability of AC to move spontaneously from the outer to the inner leaflet of the bilayers on a slower time scale was assessed by obtaining 13C-spectra as a function of time following the addition of AC to preformed vesicles. Even after 3 to 4 days of incubation of the sample at room temperature, no AC signal from the inner leaflet of the vesicles was observed. Spontaneous movement of AC from the outer to the inner leaflet would have resulted in the appearance of the AC carbonyl signal from the inner leaflet. The persistent absence of the signal from AC on the inner leaflet indicates that flip-flop is extremely slow. Because of the inductive effect from the titratable carnitine carboxyl group, the chemical shift of the nearby carbonyl carbon in the acyl chain (Fig. 1) is expected to be pH-sensitive. Figure 4shows pH-dependent spectra of decanoyl carnitine in cosonicated (A–C) and preformed (D–F) SUV in unbuffered 0.56% (75 mM) KCl solution. In both samples, the signals from AC in the outer leaflet of the vesicles shifted upfield by 1 ppm with the change from neutral to acidic pH. In the cosonicated sample, the signal from decanoyl carnitine on the inner leaflet of the vesicles (ACo) also shifted upfield, but to a lesser extent (Fig. 4A–C). This effect is a result of a leak of protons across the bilayer, whic
Referência(s)