The perioperative use of β-adrenoceptor-blocking agents or β-blockers during anaesthesia has been suggested to reduce and prevent the risk of adverse events such as myocardial ischaemia, perioperative arrhythmia and cardiac morbidity.1–3 Based on the affinity to two distinct receptors, clinically relevant β-blockers are divided into nonselective blockers, which affect both β1-adrenoceptors and β2-adrenoceptors, and selective (cardioselective) β1-blockers, which mainly block β1-adrenoceptors. Cardioselective β1-blockers are preferable to nonselective β-blockers for patients with bronchial asthma and peripheral vascular diseases because bronchioles and peripheral blood vessels contain β2-adrenoceptors, which increasingly mediate bronchial and arterial dilatation.4 β1-Selectivity offers an advantage over nonselectivity in perioperative systemic management during anaesthesia.
The adrenoceptor blockade and the selectivity of β-blockers have been exclusively explained by the specific bindings of drugs to proteinaceous receptors as the primary mode of action. However, certain β-blockers have an additional mechanism, which produces their pharmacological characteristics. Recent studies suggest that it may be premature to exclude the interaction with membranous lipids from the mechanistic contribution to receptor-interacting drugs.5 Membrane lipid bilayers provide β1-adrenoceptors and β2-adrenoceptors with their surrounding environments. β-Blockers interact with lipid membranes to induce fluidity changes,6–8 and the subsequent alteration of receptor protein conformation could influence β-adrenoceptor activity.9
Most β-blockers are structurally characterized as isoprenaline derivatives containing an alkanolamine side-chain terminating with a secondary amino group.10 They are present in cationic and neutral form at physiological pH. The hydrophobicity as well as the ionization state is important for the membrane interaction of drugs. Although the pKa values of different β-blockers are within a relatively narrow range, their hydrophobicities estimated by octanol–water partition coefficients (Log P) significantly vary.10,11 In particular, a difference in hydrophobicity seems to exist between selective and nonselective β1-blockers.11 Membrane lipid bilayers are also able to discriminate even slightly different structures of drugs and cause membrane fluidity changes depending on drug structures.12 Therefore, the comparative potencies to interact with lipid membranes and modify their fluidity may offer a novel insight into the differentiation of β-blockers’ selectivity and nonselectivity apart from the receptor-binding characteristics.
We aimed to address whether selective and nonselective β-blockers are differentiated in the light of the interaction with lipid membranes. β-Blockers with the common alkanolamine side-chain, a 2-hydroxy-3-isopropylaminopropoxyl substituent, were chosen for this comparative study. Selective β1-blockers, including atenolol, metoprolol and esmolol, and nonselective β-blockers, including alprenolol, oxprenolol and propranolol (see Fig. 1 for their structures), were reacted with liposomal model membranes and the fluidity changes that they induced were determined and their relative hydrophobicities were evaluated.
As relative drug hydrophobicity associated with interactivity with phosphatidylcholine bilayers is evaluated by reversed-phase liquid chromatographic retention on an octadecyl solid phase, β-blockers were subjected to chromatography as reported previously.13 The Shimadzu chromatographic system (Kyoto, Japan) consisted of an LC-10ADVP liquid chromatograph connected to an SIL-10ADVP autosampler, a DGU-4A degasser, a CTO-6A column oven and an SPD-M10AVP diode-array detector. Atenolol, metoprolol, esmolol, alprenolol, oxprenolol and propranolol (1 mmol l−1 for each) were dissolved in the mobile phase solution, 35% (v/v) acetonitrile in Dulbecco's phosphate buffered saline, pH 7.4, (Dainippon Sumitomo Pharma, Osaka, Japan). Esmolol was supplied by Maruishi Pharmaceutical (Osaka, Japan) and the other β-blockers were purchased from Wako Pure Chemicals (Osaka, Japan). An aliquot (5 μl) of the resulting solutions was repeatedly (n = 5) injected onto a Shim-pack CLC-ODS column (150 mm × 6.0 mm internal diameter particle size of 5 μm; Shimadzu). The mobile phase was delivered at a flow rate of 1.0 ml min−1 and at a column temperature of 37°C. Column eluates were detected at the absorption wavelength of 228 nm. The capacity factors of β-blockers were calculated by the defined formula (tR−t0)/t0, in which tR was the obtained retention time of each blocker and t0 (the retention time of a nonretained compound) was determined by a previous described method.14 The hydrophobicities of compounds are expressed as the partition coefficients in the biphasic solvent system octanol–water, Log P. As the retention of compounds on reversed-phase liquid chromatography is governed by the hydrophobic interaction, a linear relationship exists between the capacity factor and Log P, whereby the hydrophobic properties of β-blockers can be comparatively evaluated using their capacity factors.
Unilamellar liposomes, liposomal membranes with a lipid bilayer structure, were prepared by a previously described method with some modifications, as follows.15 Different lipids (total lipids of 10 mmol l−1) and a fluorescent probe: 1,6-diphenyl-1,3,5-hexatriene (DPH), n-(9-anthroyloxy)stearic acid (n = 2, 6, 9 and 12; n-AS) or 16-(9-anthroyloxy)palmitic acid (16-AP) were dissolved in ethanol with a molar ratio of total lipids to each probe of 200: 1. Fluorescence probes were purchased from Molecular Probes (Eugene, Oregon, USA). An aliquot (250 μl) of the ethanol solutions was injected four times into 199 ml of Dulbecco's phosphate buffered saline, pH 7.4, under stirring above the phase transition temperatures of phospholipids. The composition of membrane lipids was 100 mol% 1,2-dipalmitoylphosphatidylcholine (DPPC) for DPPC liposomal membranes and a mixture of 1-palmitoyl-2-oleoylphosphatidylcholine, 1-palmitoyl-2-oleoylphosphatidylethanolamine, cerebroside, sphingomyelin and cholesterol (1: 1: 1: 1: 2, molar ratio) for biomimetic membranes.16
The interactivities of β-blockers with lipid membranes were compared on the basis of their induced changes in membrane fluidity, which were determined by the fluorescence polarization method.17 Atenolol, metoprolol, esmolol, alprenolol, oxprenolol and propranolol (a final concentration of 0.2 and 1 mmol l−1 for each) were reacted with the membrane preparations by adding 0.1 ml of their aqueous solutions to 2.9 ml of liposome suspensions. After the reaction at 37°C for 10 min, fluorescence polarization was measured by an RF-540 spectrofluorometer (Shimadzu) equipped with a polarizer and a cell holder thermocontrolled at 37°C under the analytical conditions reported previously.18 Excitation and emission wavelengths were 360 and 430 nm for DPH polarization, and 367 and 443 nm for n-AS(P) polarization. The potencies to modify membrane fluidity were determined by DPH and n-AS(P) polarization differences from controls (the membrane treated with vehicle alone). Fluorescent probes penetrate into membrane lipid bilayers to align with phospholipid acyl chains. As they are subject to the rotational restriction imparted by membrane rigidity (lipid order), more fluid (disordered) membranes facilitate probe rotation emitting the absorbed light in all directions, resulting in a lower polarization ratio. Compared with controls, a decrease in fluorescence polarization is indicative of an increase in membrane fluidity (membrane disordering). As n-(9-anthroyloxy) fatty acids (n = 2–16) selectively locate at a graded series of levels in lipid membranes, fluorescence polarization reflects the fluidity gradient extending from the surface to the centre of membranes with an increase in n. The n-AS(P) polarization values decrease with increasing n because the deeper regions of lipid bilayers are more fluid than the surface regions. Therefore, the degree of polarization changes (%) relative to control polarization values were used for comparing the membrane interactivity between DPPC liposomes and biomimetic membranes and between different membrane regions.
Membrane samples were freshly prepared on each experimental day and the comparative analyses of selective and nonselective β-blockers were carried out on the same day. The analytical precisions of membrane interactivity determinations were evaluated using the polarization values obtained from replicated samples (n = 7 for each analysis). Intraassay coefficients of variation ranged from 0.2 to 3.3% in all analyses. From such high analytical reproducibility, the differences in polarization changes conclusively indicate differences in inherent membrane interactivities of β-blockers, and are not derived from experimental artifact.
All results are expressed as mean ± SE (n = 7 for membrane experiments and n = 5 for chromatographic experiments). Data were analysed using the computer statistics package StatView 5.0 (SAS Institute, Cary, North Carolina, USA). Statistical tests included a one-way analysis of variance (ANOVA) followed by a post-hoc Fisher's protected least significant difference (PLSD) test. Statistical significance was accepted at the 95% confidence level (P < 0.05).
In reversed-phase liquid chromatography, the capacity factor was 0.05 ± 0.003 for atenolol, 1.31 ± 0.002 for metoprolol, 1.65 ± 0.024 for esmolol, 2.33 ± 0.030 for oxprenolol, 5.11 ± 0.057 for alprenolol and 4.78 ± 0.041 for propranolol. Selective β1-blockers showed smaller values than nonselective β-blockers.
Oxprenolol, alprenolol and propranolol fluidized DPPC liposomal membranes in an increasing order of intensity, as shown by the DPH polarization decrease (Table 1). The potency indicated by relative polarization change (%) was 1.5 ± 0.5 for oxprenolol, 10.4 ± 0.3 for alprenolol and 20.9 ± 0.1 for propranolol. Propranolol and alprenolol showed membrane-fluidizing effects at 20–50 μmol l−1. However, atenolol, metoprolol and esmolol did not influence membrane fluidity even at 1 mmol l−1.
The sites of action in membrane lipid bilayers were characterized using selective and nonselective β-blockers: atenolol (1-p-carbamoylmethylphenoxy-3-isopropylamino-2-propanol) and alprenolol (1-o-allylphenoxy-3-isopropylamino-2-propanol), both of which have a similar basic structure suitable for comparison. When reacted with n-AS(P)-labelled DPPC liposomal membranes, alprenolol decreased n-AS(P) polarization with increasing n, whereas atenolol showed no changes in any polarization (Fig. 2).
β-Blockers interacted with biomimetic membranes to increase their fluidity with the potency being oxprenolol < alprenolol < propranolol (Table 2). However, membrane fluidization was not induced by atenolol and nor by metoprolol or esmolol. The comparative results on biomimetic membranes (the intensity of membrane interaction and the rank order of interactivity) agreed with those on DPPC liposomal membranes, though the relative polarization changes (%) were not as great as in DPPC liposomal membranes.
The comparative interactions with DPPC liposomal membranes have indicated that β1-blockers are differentiated into membrane-fluidizing and nonfluidizing drugs. The membrane interactivity is confined to nonselective β-blockers consisting of oxprenolol, alprenolol and propranolol. Membrane-interactive β-blockers rearrange the intermolecular hydrogen-bonded network among phospholipid molecules and also change the orientation of the P–N dipole of phospholipid molecules, resulting in an increase in membrane fluidity. In the depth-dependent interactions with DPPC liposomal membranes by structurally similar β-blockers, nonselective alprenolol acts more intensively on the deeper regions of DPPC bilayers, whereas β1-selective atenolol is not active on any membrane regions. The preferential interaction with the hydrophobic core is consistent with the previous report.19 As DPH is also localized in the acyl chain region of phospholipid bilayers,20 this probe is suitable for indicating the fluidity change at the β-blocker's site of action.
Nonselective β-blockers with membrane interactivity show a higher hydrophobicity (larger capacity factors), whereas selective β1-blockers without membrane interactivity show a lower hydrophobicity (smaller capacity factors). However, the rank order of hydrophobicity of nonselective β-blockers (alprenolol > propranolol > oxprenolol) does not necessarily agree with that of membrane interactivity (propranolol > alprenolol > oxprenolol).
Although hydrophobicity facilitates the interaction of drugs with lipid bilayers, certain chemical structures seem to be required for β-blockers to modify membrane fluidity. Selective β1-blockers (atenolol, metoprolol and esmolol) have the 2-hydroxy-3-isopropylaminopropoxyphenyl structure with side-chains at the para-position in common. Although nonselective β-blockers are also 2-hydroxy-3-isopropylaminopropoxyl derivatives, alprenolol and oxprenolol have a phenyl nucleus with side-chains at the ortho-position, and propranolol has an α-naphthyl nucleus instead of a phenyl one. As shown by DPH and n-AS(P) polarization changes, β-blockers are assumed to penetrate into phospholipid bilayers and align between the acyl chains. When two side-chains are in the ortho-positions, they occupy more space, which perturbs the alignment of phospholipid acyl chains and results in an increase in membrane fluidity. When two side-chains are in the para-positions, the molecules show an almost linear configuration. Therefore, selective β1-blockers align approximately parallel to phospholipid acyl chains, inducing no significant changes in membrane fluidity. Selective β1-blockers also have fewer hydrophobicity requisites for membrane interaction. Nonselective and selective β1-blockers are thought to interact differently with lipid membranes because of differences in their respective conformations. Compared with the phenyl nucleus of alprenolol, the α-naphthyl moiety of propranolol is more bulky, enough to cause greater perturbation of membranes. Propranolol could produce greater fluidization than alprenolol, despite having lower hydrophobicity than alprenolol.
All selective β1-blockers are phenyl derivatives with an alkanolamine side-chain and different groups in the para-positions.11 In contrast, nonselective β-blockers consist of either phenyl derivatives with an alkanolamine side-chain and different groups in the ortho-positions or alkanolamine derivatives with a bulky ring moiety such as naphthalene, indole, quinoline and carbazole.8,10 The membrane interactivity whereby selective and nonselective β1-blockers are differentially characterized may be widely applicable to β-blockers.
β-Adrenoceptor-acting drugs modify membrane fluidity, and neurotransmitters indirectly act on their receptors by changing the membrane properties in addition to directly interacting with their receptors.21 The membrane affinity also controls the effects of β-adrenergic drugs.22 It has become clear that biological membranes are not a simple bilayer structure but include microdomains known as lipid rafts, which are biophysically different from the bulk membranes.23 Recent studies have suggested that highly ordered membrane domains encompass β-adrenoceptors to determine their functions.24,25 Lipid rafts are formed by the tight packing of cholesterol and sphingolipids to play a role as the potential platform for functional proteins.
As membrane fluidity modifies the functions of cardiac β-adrenoceptors,9,26,27 the differential interactivities with lipid membranes may represent one of the possible mechanisms whereby selective and nonselective β1-blockers influence different β-adrenoceptor activities. Although nonselective β1-blockers fluidized both kinds of membrane preparations, the degree of fluidization (polarization decrease) was smaller in biomimetic membranes than in DPPC liposomal membranes. In this study, the biomimetic membranes were prepared with sphingomyelin, cholesterol and phospholipids to mimic membrane lipid rafts.16 Such a lipid composition would make them more ordered, which may explain the relative resistance to drug-induced fluidization. Lipid rafts also form caveolae by polymerizing with caveolins, the palmitoylated integral membrane proteins that bind to cholesterol.23 Adrenoceptors show a different distribution of subtype receptors on cardiac myocytes, in which β2-adrenoceptors are predominantly found with caveolae. β2-Adrenoceptors, but not β1-adrenoceptors, require carveolar localization for physiological signals.24 The membrane lipid domains influence β2-adrenoceptor signal transduction and nonselective β1-blockers exclusively fluidize biomimetic membranes. Membrane fluidization results in the decreased function of β2-adrenoceptors.28 Nonselective β1-blockers may interact with lipid membranes and induce their fluidization, which decreasingly changes the β2-adrenoceptor activity cooperatively with the specific interaction with β2-adrenoceptors to block them.
The membrane fluidization associated with β-adrenoceptor blockade has been inconclusive at clinically relevant concentrations. Hydrophobic drugs such as nonselective β-blockers penetrate into lipid bilayers to be concentrated in the deeper regions. Nonselective β-blockers – propranolol, alprenolol and carvedilol – are able to fluidize phospholipid membranes at micromolar concentrations as revealed by this and previous studies.8 Membrane-active agents are more effective on biological membranes than on liposomal model membranes.29,30 The membrane-fluidizing effects could partly contribute to the nonselective blockade of β-adrenoceptors.
In conclusion, selective and nonselective β1-blockers are differentially characterized by high or no membrane interactivities. Such differentiation may reflect not only the structural difference but also the β-adrenoceptor-blocking difference.
The present study was supported by Grant-in-Aids for Scientific Research (C) 15592145 (to H.T.) and 20592381 (to H.T.) from the Japan Society for the Promotion of Science.
The authors have no conflicts of interest to disclose.
1 Jakobsen CJ, Blom L. Effect of preoperative metoprolol on cardiovascular and catecholamine response and bleeding during hysterectomy. Eur J Anaesthesiol 1992; 9:209–215.
2 Auerbach AD, Goldman L. β-Blockers and reduction of cardiac events in noncardiac surgery. JAMA 2002; 287:1435–1444.
3 Lombaard SA, Robbertze R. Perioperative use of β-blockers in the elderly patient. Anesthesiol Clin 2009; 27:581–597.
4 Aronson JK. Changing beta-blockers in heart failure: when is a class not a class? Br J Gen Pract 2008; 58:387–389.
5 Lynch C III. Meyer and Overton revised. Anesth Analg 2008; 107:864–867.
6 Shi B, Tien HT. Action of calcium channel and β-adrenergic blocking agents in bilayer lipid membranes. Biochim Biophys Acta 1986; 24:125–134.
7 Varga E, Szöllösi J, Antal K, et al
. Nonspecific membrane effects of CH-103: hydrophobicity
, surface activity and membrane fluidity studies in comparison with propranolol and practolol. Pharmazie 1999; 54:380–384.
8 Butler S, Wang R, Wunder SL, et al
. Perturbing effects of carvedilol on a model membrane system: role of lipophilicity and chemical structure. Biophys Chem 2006; 119:307–315.
9 Ma Z, Lee SS, Meddings JB. Effects of altered cardiac membrane fluidity on β-adrenergic receptor signalling in rats with cirrhotic cardiomyopathy. J Hepatol 1997; 26:904–912.
10 Carda-Broch S, Berthod A. pH dependence of the hydrophobicity
of β-blocker amine compounds measured by counter-current chromatography. J Chromatogr A 2003; 995:55–66.
11 Detroyer A, Heyden YV, Carda-Broch S, et al
. Quantitative structure-retention and retention-activity relationships of β-blocking agents by micellar liquid chromatography. J Chromatogr A 2001; 912:211–221.
12 Mizogami M, Tsuchiya H, Ueno T, et al
. Stereospecific interaction of bupivacaine enantiomers with lipid membranes. Reg Anesth Pain Med 2008; 33:304–311.
13 Tsuchiya H, Mizogami M, Takakura K. Reversed-phase liquid chromatographic retention and membrane activity relationships of local anesthetics. J Chromatogr A 2005; 1073:303–308.
14 Kraak JC, van Rooij HH, Thus JLG. Reversed-phase ion pair systems for the prediction of n-octanol-water partition coefficients of basic compounds by high-performance liquid chromatography. J Chromatogr A 1986; 352:455–463.
15 Tsuchiya H, Mizogami M. Membrane interactivity
of charged local anesthetics derivative and stereoselectivity in membrane interaction of local anesthetic enantiomers. Local Reg Anesth 2008; 1:1–9.
16 Schroeder R, London E, Brown D. Interactions between saturated acyl chains confer detergent resistance on lipids and glycosylphosphatidylinositol (GPI)-anchored proteins: GPI-anchored proteins in liposomes and cells show similar behavior. Proc Natl Acad Sci U S A 1994; 91:12130–12134.
17 Tsuchiya H. Structure-specific membrane-fluidizing effect of propranolol. Clin Exp Pharmacol Physiol 2001; 28:292–299.
18 Tsuchiya H, Nagayama M. Garlic allyl derivatives interact with membrane lipids to modify the membrane fluidity. J Biomed Sci 2008; 15:653–660.
19 Ondrias K, Stasko A, Jancinova V, Balgavy P. Comparison of the effect of eleven β-adrenoceptor blocking drugs in perturbing lipid membrane: an ESR spectroscopy study. Mol Pharmacol 1987; 31:97–102.
20 Repáková J, Holopainen JM, Morrow MR, et al
. Influence of DPH on the structure and dynamics of a DPPC bilayer. Biophys J 2005; 88:3398–3410.
21 Cantor RS. Receptor desensitization by neurotransmitters in membranes: are neurotransmitters the endogenous anesthetics? Biochemistry 2003; 42:11891–11897.
22 Austin RP, Barton P, Bonnert RV, et al
. QSAR and the rational design of long-acting dual D2-receptor/β2
-adrenoceptor agonists. J Med Chem 2003; 46:3210–3220.
23 Simons K, Toomre D. Lipid rafts and signal transduction. Nat Rev Mol Cell Biol 2000; 1:31–39.
24 Xiang Y, Rybin VO, Steinberg SF, Kobilka B. Caveolar localization dictates physiologic signaling of β2
-adrenoceptors in neonatal cardiac myocytes. J Biol Chem 2002; 277:34280–34286.
25 Ianoul A, Grant DD, Rouleau Y, et al
. Imaging nanometer domains of β-adrenergic receptor complexes on the surface of cardiac myocytes. Nat Chem Biol 2005; 1:196–202.
26 Hirata F, Axelrod J. Phospholipid methylation and biological signal transmission. Science 1980; 209:1082–1090.
27 Gazawi H, Ljubuncic P, Cogan U, et al
. The effects of bile acids on β-adrenoceptors, fluidity, and the extent of lipid peroxidation in rat cardiac membranes. Biochem Pharmacol 2000; 59:1623–1628.
28 Lombardi D, Cuenoud B, Krämer SD. Lipid membrane interactions of indacaterol and salmetelol: do they influence their pharmacological properties? Eur J Pharm Sci 2009; 38:533–547.
29 Muller CP, Krueger GRF. Modulation of membrane proteins by vertical phase separation and membrane lipid fluidity. Basis for a new approach to tumor immunotherapy. Anticancer Res 1986; 6:1181–1194.
30 Yun I, Cho E-S, Jang H-O, et al
. Amphiphilic effects of local anesthetics on rotational mobility in neuronal and model membranes. Biochim Biophys Acta 2002; 1564:123–132.