Partition of Dopamine Antagonists into Synthetic Lipid Bilayers: the Effect of Membrane Structure and Composition

Abstract— Partition coefficients, K_p, of four dopamine antagonists (pimozide, fluspirilene, haloperidol and domperidone) between the aqueous phase and lipid bilayer vesicles were determined as a function of lipid chain length, unsaturation and temperature encompassing the range of the lipid phase transition. Model membranes of egg phosphatidylcholine (PC), dimyristoyl (DMPC)-, dipalmitoyl (DPPC)-, distearoyl (DSPC)- and dioleoyl (DOPC)-phosphatidylcholines were studied. K_p values of the drugs are different in the various membranes under study and depend on temperature, aliphatic carbon chain-length and on the presence of unsaturation in the aliphatic lipid chain. First-order transition of membrane lipids from the gel to the liquid crystalline state is accompanied by a sharp increase of the partition coefficient of pimozide and fluspirilene in DMPC, DPPC and DSPC bilayers. For domperidone, K_p values are maximal within the midpoint of phase transition of DMPC and DPPC, while for DSPC K_p values increase progressively with increasing temperature. Haloperidol K_p values display a maximum at the mid-point of phase transition of DMPC, while a progressive increase of K_p is observed in DPPC and DSPC. The four drugs are easily accommodated in bilayers of short aliphatic chain lipids (DMPC), the partition coefficients being 17137 for pimozide, 18 700 for fluspirilene, 686 for domperidone and 722 for haloperidol, at temperatures 10°C below the mid-point of the lipid phase transition. Except for haloperidol, the partition of the drugs in DOPC (18:1) is higher than that in DSPC (18:0) bilayers at a temperature above the phase transition temperature of both lipids. From our experiments we can conclude that artificial membranes are useful models to understand the physicochemical mechanisms involved in the interaction of dopamine antagonists with biological membranes.     

Partition of Ca2+ antagonists in brain plasma membranes

The partition coefficients (Kp) of three prototype Ca2+ antagonists, nitrendipine, (-)-desmethoxyverapamil and flunarizine were determined in native synaptic plasma membranes (SPM) isolated from sheep brain cortex and in liposomes prepared with the total lipids extracted from the membranes. We found that at 25 degrees and at 5 x 10(-6) M drug concentration the Kp values of the drugs for native SPM are higher than those obtained for liposomes, and are of the order of 334 +/- 53, 257 +/- 36 and 23 X 10(3) for nitrendipine, (-)desmethoxyverapamil and flunarizine, respectively, whereas the Kp values in liposomes are 190 +/- 41, 118 +/- 10 and 6 x 10(3) for the same drugs. The results suggest that the presence of membrane proteins favors the incorporation of the drugs in the membranes. Furthermore, the Kp values of the three Ca2+ antagonists studied increase with temperature in native membranes, but not in liposomes. It is concluded that the physical partitioning in membranes of drugs which act on Ca2+ channels may play some role in the mechanism of interaction of these drugs with the Ca2+ channel proteins.     

Antioxidant effect of calcium antagonists on microsomal membranes isolated from different brain areas.

The antioxidant effect of Ca2+ antagonists on sheep brain microsomal membranes and on liposomes prepared with total lipids extracted from the membranes was studied. Microsomal membranes were isolated from three brain areas: frontal cortex, hippocampus and caudate nucleus. Lipid peroxidation was induced by ascorbic acid and measured by malondialdehyde formation. Seven Ca2+ antagonists representative of the major chemical classes (dihydropyridines, benzothiazepines, phenylalkylamines, alkylamines, diphenylpiperazines) were tested for their antioxidant activity over a wide range of concentrations (0-500 microM). The order of antioxidant activity on frontal cortex membrane phospholipids, expressed as 50% inhibition of peroxidation (antioxidant IC50), was: nifedipine (IC50 = 4 microM) greater than flunarizine (IC50 = 48 microM) greater than bepridil (IC50 = 50 microM) greater than verapamil (IC50 = 74 microM). The dihydropyridines, nitrendipine and nimodipine, and the benzothiazepine, diltiazem, did not affect peroxidation even at a concentration of 500 microM. Membrane phospholipids are the substrate for free radical-induced damage since the extent of peroxidation in brain microsomal membranes was equal to that produced in liposomes prepared from membrane lipids. Although the lipophilicity of certain Ca2+ channel antagonists can enhance their antioxidant activity, our data suggest that Ca2+ antagonists inhibit peroxidation of the membrane lipid bilayer by a free radical scavenger effect that may be related to their chemical structure.     

Binding of [3H]haloperidol to dopamine D2 receptors in the rat striatum.

The present study was designed to examine the properties of [3H]haloperidol binding to dopamine D2-receptors in rat striatum membranes, displacement potencies of various chemicals and differences between the affinities of various chemicals and two new 5-hydroxytryptamine (5-HT2) receptor antagonists, MCI-9042, (+/-)-2-(dimethylamino)-1-[[o-(m-methoxyphenetyl)phenoxy]methyl]et hyl hydrogen succinate hydrochloride and one of its metabolites. The plots of specific binding for the striatum membranes obtained from the Scatchard analysis using [3H]haloperidol were monophasic when non-specific binding was determined with 10 microM chlorpromazine, and the Kd and Bmax values were 7.42 +/- 1.03 nM and 1.58 +/- 0.20 pmol (mg protein)-1 (n = 10), respectively. The displacement potencies of D2 receptor, 5-HT2 receptor, histamine H1-receptor, and adrenoceptor antagonists were characterized by [3H]haloperidol binding to D2 receptors. The pKi values of a new antiplatelet agent, MCI-9042, and its metabolite were 5.02 and 5.53, respectively, and these values were lower than those of the D2-receptor antagonists, fluphenazine, spiperone, haloperidol, prochlorperazine, chlorpromazine, thioridazine, and sulpiride. They were also lower than the pKi values of the 5-HT2-receptor antagonists, pirenperone, ketanserin, methysergide, and mianserin. We conclude that the binding site of [3H]haloperidol in the rat striatum is the D2 receptor, that MCI-9042 and its metabolite have lower affinities for D2 receptors than for 5-HT2 receptors, and that this radioreceptor assay is useful for assessing the affinities of various agents.     

Cholesterol alters the binding of Ca2+ channel blockers to the membrane lipid bilayer.

X-ray diffraction and equilibrium binding techniques were used to study the effect of cholesterol on membrane binding of the charged 1,4-dihydropyridine (DHP) Ca2+ channel antagonist amlodipine and uncharged isradipine, nimodipine, and nitrendipine. Increases in membrane cholesterol content resulted in a marked decrease in DHP binding to cardiac phospholipid membranes, as expressed by the equilibrium partition coefficient (Kp[mem]). Between a 0:1 and 0.3:1 cholesterol to phospholipid mole ratio, the Kp(mem) values for isradipine, nimodipine, and nitrendipine decreased by greater than 50%, whereas that for amlodipine decreased by only 10%. Electron density profiles calculated from the X-ray diffraction data showed that the time-averaged locations for the DHPs and cholesterol in the membrane overlap, leading to the conclusion that the addition of cholesterol alters the lipid bilayer hydrocarbon core structure in a manner that makes drug partitioning into the membrane less energetically favorable. These data support the idea that drug interactions with the anisotropic membrane environment are complex and may be greatly influenced by cholesterol composition. This effect of cholesterol was also observed for phenylalkylamine (verapamil) and benzothiazepine (diltiazem) Ca2+ channel blockers. The DHP amlodipine had the highest membrane partition coefficient (Kp[mem] greater than 10(4) and the slowest rate of dissociation and was affected least by membrane cholesterol content. The combination of electrostatic and hydrophobic bonding between amlodipine and membrane phospholipid may explain the high affinity of this drug for the membrane bilayer with normal and elevated cholesterol. The results of this study show that cholesterol content differentially affects the membrane-binding properties of the charged DHP amlodipine, compared with other Ca2+ channel blockers. These data help explain the biological distribution of these drugs and the distinct pharmacokinetics of amlodipine versus other Ca2+ channel blockers.     

Evaluation of estimations in vitro of tissue/blood distribution coefficients for organothiophosphate insecticides

Physiologically based pharmacokinetic (PBPK) modeling of foreign chemicals is dependent on the accurate determination of their tissue/blood distribution coefficients, Kp (partition coefficients). The present study was undertaken to evaluate the validity of the in vitro estimation of the Kp values of the organothiophosphate insecticides parathion and methyl parathion by equilibrium dialysis. Data derived from previously published studies that utilized single-pass perfusions of mouse livers in situ with parathion or methyl parathion were analyzed to determine liver/perfusate Kp values from the equation Kp = (t 1/2ss) (Q)/(0.693) (VH), where Kp is the liver/perfusate distribution ratio, t 1/2ss is the half-life for approach to steady state of the chemical, VH is the liver volume, and Q is the perfusate flow rate. Kp values for methyl parathion were calculated to be 16.4 +/- 7.5 and 9.5 +/- 2.7 (mean +/- SD) for perfused livers and equilibrium dialysis, respectively, while estimates of Kp for parathion were 15.6 +/- 6.3 and 19.5 +/- 5.5 for perfused livers and equilibrium dialysis, respectively. These results indicate that equilibrium dialysis can be utilized to give an accurate estimate of tissue partitioning of parathion and methyl parathion from perfusate into perfused mouse livers.     

A comparative in vitro study of transdermal absorption of a series of calcium channel antagonists.

The in vitro transdermal absorption of five calcium channel antagonists (nifedipine, nitrendipine, nicardipine, felodipine, and nimodipine) was studied using the skin of hairless rats as a membrane. The aim of this study was to determine the penetration parameters [permeability constant (Kp), lag time (T1, and flux (J)] as measures of the intrinsic transdermal permeabilities of these drugs, in order to predict the potential capacity of these drugs to be formulated in a therapeutical transdermal system (TTS). Reliable prediction of Kp values, using K'w (extrapolated capacity factor in 100% water) and P (n-octanol-water partition coefficient) values as independent variables in the parabolic, bilinear, and hyperbolic functions, were assayed. Nicardipine showed a higher mean transdermal penetration (Kp; 4.9 x 10(-3) cm.h-1) value than the other dihydropyridines. Nifedipine showed the shortest mean T1 value (5.1 h). The permeability rate constants of the calcium channel antagonists assayed can be predicted from their n-octanol-water partition coefficients, using the parabolic function (r = 0.984, p less than 0.01). Nicardipine would be the most suitable candidate for formulation in a TTS design.     

Parallel artificial membrane permeability assay: a new membrane for the fast prediction of passive human skin permeability

This work was devoted to the search for new artificial membranes allowing a rapid evaluation of passive human skin permeation of compounds with a parallel artificial membrane permeability assay (PAMPA). Effective permeability coefficients (Pe) determined for a set of compounds using the PAMPA technique with isopropyl myristate (IPM) and silicone oil, alone or in mixture, were compared to the corresponding human skin permeability coefficient values (Kp). A good correlation between Pe and Kp was found for compounds tested through a membrane consisting of 70% silicone and 30% IPM. Moreover, positive correlation between the membrane retention of compounds and stratum corneum/water partition coefficients (PSC) was established. These results showed that this new artificial membrane, defined as PAMPA-skin, is able to mimic the main barrier properties of human stratum corneum and can be used for the fast prediction of passive human skin permeability coefficients.     

The dependence of the lipid bilayer membrane: buffer partition coefficient of pentobarbitone on pH and lipid composition.

1 The membrane/buffer partition coefficient of [14C]-pentobarbitone has been determined as a function of the lipid composition of bilayer membranes. 2 A new technique based on ultrafiltration gave comparable results to conventional techniques but required less time for equilbration. 3 The membrane/buffer coefficient was independent of pentobarbitone concentration in the range studies. 4 The apparent partition coefficient varied with pH and was a linear function of the degree of dissociation of pentobarbition. 5 Both the charged and uncharged forms of pentobarbitone partitioned into the membrane, the latter to a much greater extent than the former. 6 At low pH the highest partition coefficient observed was in egg phosphatidylcholine bilayer membranes. 7 Incorporation of cholesterol or phosphatidic acid into phosphatidylcholine membranes greatly reduced the partition coefficient. 8 High pressures do not greatly change these partition coefficients.     

Trapping of labelled ligands in intact cells: a pitfall in binding studies

Binding on/in whole cells seems to be a more appropriate approach for studying receptor sites in physiological conditions. However, certain difficulties encountered throughout the characterization of [3H]spiperone binding in human lymphocytes led us to reconsider this problem. The IC50 values of [3H]spiperone binding to human lymphocytes did not correlate with those found in rat striatum; domperidone was inactive in lymphocytes whereas it is one of the most potent dopamine antagonists in rat striatal preparations in vitro. In contrast, chloroquine, a lysosomotropic drug, displaced [3H]spiperone at low concentration in intact lymphocytes but did not in the striatum. [3H]Spiperone binding was not displaceable in the membrane preparation of lymphocytes. Similar results were obtained with other intact cells, fibroblasts, hepatocytes and neuroblastoma cells using [3H]spiperone and other ligands, such as [3H]haloperidol, [3H]pyrilamine and [3H]ketanserin. Here again, displaceable binding was only present in intact cells but not in membrane fractions. Such a 'displaceable' binding was not related to receptor sites but may be regarded as non-specific binding which should correspond to a trapping phenomenon presumably in the lysosomes. Binding studies on intact cells need more caution than when performed on membrane preparations; indeed, permeation or trapping of ligands in the nanomolar range represents a serious drawback which, sometimes, can give the illusion of specific binding.     

Transport of anthracyclines and mitoxantrone across membranes by a flip-flop mechanism

The objectives of the present work are to characterize the transport of mitoxantrone and three anthracyclines in terms of binding to the membrane surface, flip-flop across the lipid core of the membrane, and release into the medium. Mitoxantrone and anthracyclines are positively charged amphipathic molecules, and as such are located at the surface of membranes among the headgroups of the phospholipids. Therefore, their transport across membranes occurs by a flip-flop mechanism, rather than by diffusion down a continuous concentration gradient located in the lipid core of the membrane. Flip-flop rates have been estimated with liposomes labeled at their surface with 7-nitrobenzo-2-oxa-1,3-diazol-4-yl (NBD) moiety attached to the headgroup of phosphatidylethanolamine. Flip-flop of mitoxantrone, doxorubicin, daunorubicin, and idarubicin occurred with half-lives of 6, 0.7, 0.15, and 0.1min, respectively. Partition of the drugs into the membrane occurred with lipid phase/aqueous medium coefficients of 230,000, 8600, 23,000, and 40,000 for mitoxantrone, doxorubicin, daunorubicin, and idarubicin, respectively, which are much higher than their corresponding octanol/aqueous medium values. There was no direct correlation between the lipophilicity of the drugs and their lipid phase/aqueous medium partition coefficient or their flip-flop rate. Mitoxantrone exhibited the highest affinity toward liposome membranes, but the slowest flip-flop across the lipid core of the membranes. Simulation of drug uptake into liposomes revealed that transmembrane movement of the mitoxantrone and anthracyclines is determined by their flip-flop rate and affinity toward membranes.     

Transport across 1,9-decadiene precisely mimics the chemical selectivity of the barrier domain in egg lecithin bilayers

The barrier domain solubility-diffusion theory of lipid bilayer permeability relates the permeability coefficient (P(m)) to the solute's partition coefficient (PC(barrier/w)) and diffusion coefficient (D(barrier)) in the ordered chain region of the bilayer that serves as the barrier region for polar permeants. To select the best solvent to mimic the barrier domain, permeability coefficients across a layer of 1,9-decadiene were compared with permeability coefficients from bilayer transport. Rate constants for transport, k, of alpha-methyl substituted analogues of p-toluic and p-methylhippuric acid were measured across a layer of 1,9-decadiene embedded in a PTFE filter membrane placed between two aqueous solutions in side-by-side diffusion cells. Permeability coefficients (P(1,9-decadiene)) were normalized to that obtained for p-toluic acid, which was included in donor solutions. The correlation of log(P(bilayer)) versus log(P(1,9-decadiene)) was linear with a slope of 0.99 +/- 0.02 SD, indicating that 1,9-decadiene precisely mimics the egg lecithin bilayer barrier domain in its chemical selectivity. Using the decadiene membrane transport method to indirectly estimate partition coefficients for similarly sized permeants extended the range of measurable values beyond those readily attainable by the traditional shake-flask method, allowing measurement of 1,9-decadiene/water PCs as low as 3 x 10(-7).     

Mechanism of action of dopamine on the guinea-pig gastro-oesophageal junction in vitro.

1 The effect of dopamine on longitudinal muscle strips of the guinea-pig isolated gastro-oesophageal junction was compared with the response obtained to phenylephrine, isoprenaline and clonidine. Phenylephrine (5 x 10(-7) to 5 x 10(-5) M) produced a dose-related contraction, whilst dopamine (10(-6) to 10(-4) M) and isoprenaline (5 x 10(-7) to 2 x 10(-5) M) produced dose-related relaxations. Clonidine was ineffective in doses up to 10(-5) M. 5-Hydroxytryptamine (5-HT) produced a contraction. 2 Phenylephrine was antagonized by alpha 1-adrenoceptor antagonists but unaffected by beta-adrenoceptor antagonists, whilst the opposite was the case for isoprenaline. A mixture of alpha- and beta-adrenoceptor antagonists was required to inhibit completely dopamine-induced relaxations. 5-HT (3 x 10(-7) M) was specifically antagonized by methysergide (3 x 10(-6) M). 3 pA2 values for a range of alpha-adrenoceptor and dopamine receptor antagonists were determined against dopamine and phenylephrine. The relative order of potency of the antagonists was the same for both antagonists and was prazosin greater than spiroperidol greater than phentolamine greater than domperidone greater than haloperidol, with pimozide and metoclopramide being inactive. 4 Tyramine caused dose-related relaxations of the gastro-oesophageal strips which were susceptible to the same range of antagonists as dopamine. 5 Cocaine (6 x 10(-6) M) and desmethylimipramine (3 x 10(-7) M) reduced the relaxations induced by dopamine and tyramine but there were quantitative differences in the antagonism. 6 Tissue from reserpine pretreated guinea-pigs was insensitive to tyramine but the response to dopamine was only partly reduced. 7 Histological examination of the strips revealed the presence of smooth muscle but only a sparse adrenergic innervation. 8 The results suggest that dopamine acts partly indirectly and partly directly on postjunctional alpha- and beta-adrenoceptors. There is no evidence for an action on specific dopamine receptors.     

Influence of the level of ceramides on the permeability of stratum corneum lipid liposomes caused by a C12-betaine/sodium dodecyl sulfate mixture.

The sublytic interactions of a mixture of N-dodecyl-N, N-dimethylbetaine dodecyl betaine (C12-Bet)/sodium dodecyl sulfate (SDS) (mole fraction of the zwitterionic surfactant=0.6) with stratum corneum (SC) lipid liposomes varying the proportion of ceramides type III (Cer) were investigated. The surfactant/lipid molar ratios (Re) and the bilayer/aqueous phase partition coefficients (K) were determined by monitoring the changes in the fluorescence intensity of liposomes due to the 5(6) carboxyfluorescein (CF) released from the interior of vesicles. The fact that the free surfactant mixture concentration was always lower than its critical micelle concentration indicates that permeability changes were ruled by the action of surfactant monomers in all cases. Higher and lower Cer proportions than that of the SC lipids led to a fall and to a rise in the activity of the surfactant mixture on these bilayer structures. However, the surfactant partitioning into liposomes (or affinity with these bilayer structures) increased as the proportion of Cer increased up to the highest value was achieved for a Cer proportion similar to that in the SC lipids (about 40-45%). Thus, at low Cer proportions the ability of the surfactant mixture to alter the permeability of these bilayer structures was higher than that for liposomes approximating the SC lipid composition despite their reduced partitioning into liposomes. These findings are in agreement with the recently reported dependencies of the level of ceramides in skin lipids and function barrier abnormalities and could explain in part these dependencies.     

Interaction of clonixin with EPC liposomes used as membrane models

In this work, an overall analysis of clonixin interaction with liposomes was achieved using different techniques, which allowed the evaluation of the change in different membrane's characteristics as well as the possible location of the drug in the membrane. Clonixin acidity constants were obtained and the values are 5.5 +/- 0.08 and 2.2 +/- 0.04. Clonixin partition coefficient (K(p)) between liposomes and water was also determined using derivative spectrophotometry, fluorescence quenching, and zeta-potential (zeta-potential). These three techniques yielded similar results. zeta-potential measurements were performed and an increase of the membrane negative charge with an increase of drug concentration was observed. Drug location within the bilayer was performed by fluorescence quenching using a set of n-(9-anthroyloxy) fatty acid probes (n = 2, 6, 9, and 12). The fluorescence intensity of all probes was quenched by the drug. This effect is more noticeable for the outer located probe, indicating that the drug is positioning in the external part of the membrane. These same probes were used for steady-state anisotropy measurements to determine the perturbation in membrane structure induced by clonixin. Clonixin increased membrane fluidity in a concentration dependent manner, with the highest perturbation occurring nearby the 2-AS probe, closely located to the bilayer surface.     

Competitive inhibition of [3H]spiperone binding to D-2 dopamine receptors in striatal homogenates by organic calcium channel antagonists and polyvalent cations

The specific binding of [3H]spiperone (30 pM) to D-2 dopamine receptors in homogenates of the rat corpus striatum, as defined by the D-2 antagonist haloperidol (100 nM), was displaced by organic calcium channel antagonists and by polyvalent cations. Both classes of agents were able to totally displace [3H]spiperone binding by an apparently competitive mechanism in that the dissociation constant was increased while the density of binding sites was unchanged. The rank order of inhibition potency for the cations was Zn2+ greater than Cd2+ greater than La3+ greater than Cu2+ greater than Ni2+ greater than Co2+ greater than Mn2+ greater than Mg2+, Ca2+ greater than Ba2+. Of the organic calcium channel antagonists, D600 and verapamil were the most potent displacers of [3H]spiperone binding (IC50 values both 2.0 microM), while diltiazem possessed an IC50 of 33 microM. Nicardipine (IC50 6.0 microM) was the only 1,4-dihydropyridine to inhibit [3H]spiperone binding. The results suggest that sites labelled by [3H]spiperone also bind organic calcium channel antagonists and polyvalent cations.     

Free Fatty Acids Cause pH-Dependent Changes in Drug-Lipid Membrane Interactions Around Physiological pH

Purpose. The influence of oleic acid (OA) and phosphatidylethanolamine (PhE) as membrane constituents on the partition behavior of (RS)-[³H]propranolol between unilamellar liposomes and buffer was studied as a function of pH. Methods. Partition studies were performed by means of equilibrium dialysis at 37°C over a broad pH range at a molar propranolol to lipid ratio in the membrane of 10^–6. Results. As compared to the standard phosphatidylcholine (PhC)-liposome/buffer partition system PhE and OA have an enhancing effect on the apparent partition coefficient (D) of (RS)-[³H]propranolol between pH 6 and 11. Data analysis with Henderson-Hasselbalch equations revealed that the neutral propranolol has a higher affinity to membranes containing net neutral charged PhE than to pure PhC-liposomes. Net negatively charged PhE seems to have no significant influence on the partitioning. Deprotonated OA caused an increase in the true partition coefficient (P) of the protonated propranolol. Neutral OA showed no influence on the partitioning. From the fit D vs pH curves and from zeta potential measurements of the liposomes the intrinsic pK_a values of the membranous lipids were calculated as 7.5 to 7.8 for OA and 9.7 to 9.8 for PhE. Conclusions. Since the pK_a of membranous OA is close to the physiological pH and D depends on the ionisation state of OA, small pH changes around the physiological pH may cause large differences in drug-lipid membrane interactions.     

Equilibrium and kinetic studies of the interactions of salmeterol with membrane bilayers.

The interaction of salmeterol with model membranes has been studied with regard to equilibrium and kinetic behavior, including determination of the membrane-based partition coefficient, the rate of dissociation of salmeterol from membranes, and the rate of association. These data were obtained in various membrane preparations and under various conditions (e.g., temperature, cholesterol content). The compound is very lipophilic, compared with other beta 2 agonists such as salbutamol, and has a rapid association rate and a moderate dissociation rate. The equilibrium data support the assertion that the salmeterol action measured in perfused tissue involves an exo-site for nonspecific binding that may be identified with or related to the lipid bilayer. The kinetic data in unilamellar and multilamellar liposomes of synthetic lipids further suggest that the approach to the exo-site and the active site may involve components in the native system other than the lipid bilayer in which the beta 2 receptor is located. These additional components may explain the slow onset and the extraordinarily long duration of action.     

Interactions of chlorpromazine and imipramine with artificial membranes investigated by equilibrium dialysis, dual-wavelength photometry, and fluorimetry.

Binding of chlorpromazine and imipramine to liposomes of various synthetic lipids was investigated by equilibrium dialysis, dual-wavelength photometry, and fluorimetry. As proved by equilibrium dialysis, liposomes of all investigated lipids bound chlorpromazine to a greater extent than imipramine. At temperatures above the lipid phase transition the membranes bound more of both drugs than they did below. These results were confirmed by dual-wavelength photometry at low drug concentrations. Chlorpromazine and imipramine fluidized lipid bilayers below their transition temperatures. Less imipramine bound to positively charged liposomes as compared with membranes of zwitterionic phosphocholine. This emphasizes the importance of ionic interactions for the binding of imipramine to lipid membranes. Chlorpromazine binding, however, was little affected by ionic interactions. Both chlorpromazine and imipramine had the same effect on the fluorescence of 1-anilino-8-naphthalene-sulfonate (a probe for the polar part of the membrane), whereas the fluorescence of perylene (a probe for the inner hydrocarbon region of bilayers) was quenched by chlorpromazine but not by imipramine. From the results obtained with the three complementary methods it is concluded that chlorpromazine binds to the surface of membranes and also penetrates into the inner hydrocarbon phase of the bilayer, whereas imipramine only binds near the surface of the liposomes.     

Spiperone: a ligand of choice for neuroleptic receptors. 1. Kinetics and characteristics of in vitro binding.

A binding assay for neuroleptic receptors has been developed with spiperone as the labelled ligand. As compared to haloperidol, spiperone showed a 2-times higher ratio of specific versus aspecific binding, a 10-fold greater association constant and a slower dissociation of the receptor ligand complex. The receptor sites labelled by spiperone appeared to be for a great deal similar to those of haloperidol, however certain differences were apparent; the number of receptor sites per gram of tissue was found to be higher for the former; spiperone showed a biphasic receptor ligand dissociation curve which was not observed for haloperidol; also a slight difference in physical stability between spiperone and haloperidol binding sites was noted. Inhibition studies using antagonists and agonists in comparison with the pharmacological profile of the compounds showed that the receptor sites labelled by both ligands are mainly of dopaminergic nature, but also a serotonergic and to a minor extent a noradrenergic component should be involved. Within the striatum haloperidol binding sites seemed to be relatively more related to dopaminergic sites whilst the spiperone binding sites appeared to comprise a higher serotonergic component.It is concluded that spiperone is a more suitable ligand than haloperidol for studying the neuroleptic receptors. The use of different labelled ligands provided evidence for the heterogeneity of the neuroleptic binding sites in the striatum.