Effect of oxidation of low-density lipoprotein on drug binding affinity studied by high performance frontal analysis-capillary electrophoresis.

The effect of oxidation of low-density lipoprotein (LDL) on the enantioselective drug binding affinity was investigated using high performance frontal analysis--capillary electrophoresis (HPFA-CE). Verapamil and nilvadipine enantiomers were used as the chiral model drugs. LDL was oxidized with copper sulfate for 0, 0.5, 1, 2, and 12 h at 37 degrees C. The HPFA-CE method enabled microdetermination of unbound drug concentrations in native and oxidized LDL solutions. It was found that the bindings between LDL and the model drugs were not enantioselective at any oxidation stage. The total binding affinity (nK) between LDL and verapamil enantiomers was increased by 3.3-, 4.6-, 7.0-, and 19-fold after 0.5, 1, 2, and 12 h oxidation, respectively, whereas the nK value between nilvadipine and LDLwas increased by 1.3-, 1.4-, 1.4-, and 1.7-fold in the same reaction times, respectively. These results indicate that the LDL oxidation enhances the drug binding affinity, and the affinity of verapamil is increased more sensitively than that of nilvadipine. The nK value of each model drug increased steeply after the first 2 h oxidation, followed by the gradual increase after the next 10 h oxidation. It is considered that the net increase in the negative charges and/or the formation of hydroperoxides in the first 2 h oxidation enhances the drug-LDL binding more significantly than the formation of aldehydes or Schiff bases in the following 10 h oxidation.

Role of biantennary glycans and genetic variants of human alpha1-acid glycoprotein in enantioselective binding of basic drugs as studied by high performance frontal analysis/capillary electrophoresis.

PURPOSE: To establish a clear understanding of the role of biantennary branching glycans and genetic variants of alpha1-acid glycoprotein (AGP) in enantioselective bindings of basic drug. METHODS: Human native AGP was separated using concanavalin A affinity chromatography into two subfractions, the unretained fraction (UR-AGP, defect of biantennary glycan) and the retained fraction (R-AGP, possessing biantennary glycan(s)). Imminodiacetate-copper (II) affinity chromatography was used to separate human native AGP into A variant and a mixture of F1 and S variants (F1*S variants). The mixed solutions of the (R)- or (S)-isomer of the model drugs (15 microM disopyramide (DP) or 30 microM verapamil (VER)) and 40 microM of respective AGP species were subjected to high-performance frontal analysis/capillary electrophoresis (HPFA/CE) to determine the unbound drug concentrations. RESULTS: The unbound concentrations (Cu) of DP in UR-AGP solutions were lower than those in R-AGP solutions, whereas there was no significant difference in the enantiomeric ratios (Cu(R)/Cu(S)) of DP between UR- and R-AGP solutions. In case of genetic variant, the Cu(R)/Cu(S) values of DP in F1*S and A solutions were 1.07 and 2.37, respectively. On the other hand, the enantiomeric ratio of VER in F1*S and A variant solutions were 0.900 and 0.871, respectively. CONCLUSIONS: The biantennary glycan structures are related to binding affinity of DP to AGP, but not responsible for the enantioselectivity. Genetic variants give significant effect on the enantioselectivity in DP binding, but not in VER binding.

Enantioselective binding analysis of verapamil to plasma lipoproteins by capillary electrophoresis-frontal analysis.

Capillary electrophoresis coupled with frontal analysis was applied to the study of enantioselective binding of verapamil (VER) to plasma lipoproteins. The drug-lipoprotein mixed solution, which had been in the binding equilibrium, was hydrodynamically introduced into a non-coated fused-silica capillary. Since VER is positively charged in the neutral run buffer (pH 7.4), the unbound VER enantiomers migrated toward the cathodic end much faster than negatively charged lipoproteins and their bound forms. Once unbound VER migrated apart from lipoprotein, the bound VER was quickly released from the protein to maintain the binding equilibrium. Thus, VER migrated as a zone through the capillary and gave a trapezoidal peak with a plateau region on the electropherogram. The VER concentration in this plateau region was equal to the unbound VER concentration in the initial sample solution. It was found that the bindings of VER to high-density lipoprotein (HDL), low-density lipoprotein (LDL) and oxidized LDL were not site-specific and not enantioselective. Partition-like binding to lipid part of these lipoproteins seemed to be dominant. The total binding affinities of LDL to VER were about seven-times stronger than those of HDL, and the oxidation of LDL by copper ion enhanced the binding affinities significantly.

Comparison of motor reactivity of the human colon cold-stored in different preservative solutions to carbachol and noradrenaline in vitro.

We have compared the reactivity to carbachol and noradrenaline of circular smooth muscle isolated from the human colon which was cold-stored at 4 degrees C in three different preservative solutions: Krebs bicarbonate solution (KBS), phosphate buffer solution (PBS) or minimum essential medium (MEM). Concentration-dependent contractions in response to carbachol were reduced in terms of both their sensitivity (pEC50) and reactivity (Emax), depending on the period of cold storage. The reduction was more marked when the tissue was cold-stored in either MEM or KBS than in PBS. Similar reduction of the relaxation response to noradrenaline was also observed after cold storage. It is concluded that the cold storage of surgically resected human colon in PBS for two to three days best preserved smooth muscle functions for pharmacological examinations.

Binding study of semotiadil and levosemotiadil with alpha(1)-acid glycoprotein using high-performance frontal analysis.

High-performance frontal analysis (HPFA) was used to investigate the binding properties of human alpha(1)-acid glycoprotein (AGP) with semotiadil ((R)-isomer, Ca-channel blocker) and its antipode levosemotiadil ((S)-isomer, Ca- and Na-channel blockers). An on-line HPLC system consisting of a HPFA column, an extraction column, and an analytical HPLC column was used to determine the unbound concentrations of these enantiomers, and the experimental data were subsequently subjected to the Scatchard analyses to estimate their binding parameters. The binding affinity of the (R)-isomer (K = 3.17 x 10(7) M, n = 0.74) is approximately 1.2 times stronger than that of (S)-isomer (K = 2.59 x 10(7) M, n = 0.74). An enantioselective competitive binding study indicated that both enantiomers are bound at the same site on AGP molecules.

The role of branching glycan of human alpha1-acid glycoprotein in enantioselective binding to basic drugs as studied by capillary electrophoresis.

The role of the branching glycan structure of human alpha1-acid glycoprotein (AGP) in the interaction with basic drugs was investigated in terms of enantioselectivity in binding ability. AGP was separated by concanavalin A lectin affinity chromatography into two subfractions, the unretained AGP (UR-AGP) which has no biantennary glycan chain and the retained AGP (R-AGP) which possesses biantennary oligosaccharide chain(s). The unbound concentrations of propranolol (PRO) enantiomers and verapamil (VER) enantiomers in UR-AGP solution and R-AGP solution were determined by high-performance frontal analysis combined with capillary electrophoresis. It was found that (S)-PRO is bound to UR-AGP and R-AGP more strongly than (R)-PRO, whereas the reverse applies to VER enantiomers, and that such enantioselectivity is common to these proteins. This suggests that the branching type of glycan chains of AGP does not play significant role in the chiral recognition in binding these basic drugs.

High-performance frontal analysis for drug-protein binding study.

High-performance frontal analysis (HPFA) is a novel analytical method which enables simultaneous determination of total and unbound drug concentrations under drug-plasma protein binding condition. HPFA can be achieved using separation systems such as HPLC and CE. This paper deals with the principle and feature of HPFA method and its application to the stereoselective protein binding study. HPFA allows a simple analysis following direct sample injection, and does not suffer from undesirable drug adsorption on membrane nor leakage of bound drug through the membrane which are often encountered in conventional ultrafiltration and dialysis methods. HPFA can be easily incorporated into on-line HPLC system. By coupling HPFA with a chiral HPLC column, the unbound concentration of a racemic drug can be determined enantioselectively. The detection limit can be improved by coupling of HPFA with a preconcentration column. High-performance capillary electrophoresis/frontal analysis (HPCE/FA) enables to determine unbound concentrations enantioselectively with ultramicro injection volume, and is hence useful especially for the binding study of proteins which are scarce and difficult to obtain.

Binding analysis of nilvadipine to plasma lipoproteins by capillary electrophoresis-frontal analysis.

Capillary electrophoresis coupled with frontal analysis (HPCE/FA) was applied to the ultramicro analysis of enantioselective binding of nilvadipine (NV), a calcium channel blocker, to plasma lipoproteins. The drug lipoprotein mixed solution was hydrodynamically introduced into a non-coated fused silica capillary for capillary electrophoresis. Since NV has no electric charge in the run buffer (pH 7.4), the unbound NV moved towards the cathodic end by electroosmotic flow, which was faster than the electrophoretic migrations of negatively charged lipoproteins and the bound NV. Once unbound NV migrated apart from lipoprotein, and bound NV was quickly released from the protein to maintain the binding equilibrium. Thus, NV migrated as a zone with a plateau region. The concentration of NV in this plateau region appearing on the electrophorogram was the same as the unbound NV concentration in the initial sample solution. It was found that the binding of NV to high-density lipoprotein (HDL), low-density lipoprotein (LDL) and oxidized LDL was non-specific and not enantioselective. Partition-like binding to the lipid part of these lipoproteins seemed to occur dominantly. The total binding affinities of NV to LDL were about seven times stronger than those to HDL, and the oxidation of LDL enhanced the binding affinity significantly.

[Development of high-performance frontal analysis and application to drug-plasma protein binding study].

This review summarizes the principle and the features of high-performance frontal analysis (HPFA), a novel chromatographic method to determine the concentrations of unbound drugs under drug-protein binding equilibrium conditions, and its application to the study on the plasma protein binding. HPFA uses a "restricted-access" type HPLC column which retains a small molecule in the drugs in the micropores, while a large molecule in the plasma protein is size-excluded. After direct and continuous injection of a sample, the drug-protein binding equilibrium in the sample solution is regenerated in the column, and the constant concentration zone of the unbound drug appears from the equilibrium zone. The unbound drug is eluted as a trapezoidal peak with a plateau. The concentration in this plateau region is equal to that of the unbound drug in the sample solution, and the concentration of the unbound drug can be determined by subsequent on-line HPLC analysis. HPFA allows a simple analysis following direct injection of a sample, and does not cause undesirable drug adsorption on the membrane nor leakage of the bound drug from the membrane, which are often encountered in the conventional ultrafiltration or dialysis method. HPFA allows the simultaneous determination of the concentrations of total and unbound drugs in a single analysis. HPFA can be easily incorporated into the on-line HPLC system. By coupling HPFA with a chiral HPLC column, the concentration of an unbound racemic drug can be determined enantioselectively. The detection limit can be improved dramatically by coupling HPEA with a preconcentration column. Frontal analysis in capillary electrophoresis format (CE/FA) allows us an ultramicro binding assay. The concentration of the unbound racemic drug can be determined stereoselectively by coupling HPFA with a chiral CE technique.

Enantioselective protein binding of semotiadil and levosemotiadil determined by high-performance frontal analysis.

An on-line frontal analysis HPLC system was developed for the determination of the unbound concentrations of semotiadil, a new calcium antagonist with non-dihydropyridine structure, and its antipode (Levosemotiadil), and was applied to the enantioselective investigation of their plasma protein binding properties. This system consists of a high-performance frontal analysis (HPFA) column, an extraction column, and an analytical column, which are connected via two switching valves. After the direct injection of the sample solution into the HPFA column, the drug was eluted as a zonal peak with a plateau region. The unbound drug concentration was determined as the drug concentration in the plateau. As low as 1.04 nM of the unbound drug was determined with good reproducibility. Semotiadil (R-isomer) and levosemotiadil (S-isomer) are bound strongly and enantioselectively to human serum albumin (HSA) and human alpha 1-acid glycoprotein (AGP), and the enantioselectivity was reversed between these plasma proteins. While HSA binds S-isomer more strongly than the antipode, human AGP binds R-isomer more strongly. In human plasma, the unbound drug fraction was less than 1%, and the enantioselectivity was similar to that observed in AGP solution.

Enantioselective local disposition of semotiadil (R-enantiomer) and levosemotiadil (S-enantiomer) in perfused rat liver.

The enantioselective local disposition of semotiadil (R-enantiomer) and levosemotiadil (S-enantiomer) in rat liver was investigated in the single-pass perfusion system containing 1% bovine serum albumin (BSA). After an instantaneous injection of semotiadil, levosemotiadil, or Evans Blue (a marker of BSA), each outflow time profile from the liver was analyzed by a two-compartment dispersion model. The recovery ratio, FH (1.88 +/- 0.28%), of semotiadil was significantly smaller than that (8.99 +/- 1.40%) of levosemotiadil. The mean transit time, fH (0.146 +/- 0.014 min) of semotiadil was significantly smaller than that (0.191 +/- 0.012 min) of levosemotiadil. The biliary excretion kinetics of these enantiomers was also evaluated by moment analysis. The parent compound (semotiadil or levosemotiadil) was not detected in bile, but four metabolites generated from each parent enantiomer were found in the bile. A portion (16.5 +/- 1.2%) of the drug eliminated by the liver was recovered as R-metabolites in the bile within 1 hr after an injection of semotiadil, whereas 11.2 +/- 1.6% was recovered as S-metabolites in the bile within 1 hr after an injection of levosemotiadil. This excreted percentage into the bile was significantly different between R- and S-metabolites. The mean biliary excretion time MRTe (19.1 +/- 2.2 min) of total R-metabolites was significantly larger than that (14.8 +/- 1.1 min) of total S-metabolites. In conclusion, stereo-selectivity was suggested both at the hepatic elimination of the parent compound and at the biliary excretion of the metabolites.

Effect of sialic acid residues of human alpha 1-acid glycoprotein on stereoselectivity in basic drug-protein binding.

The function of sialic acid groups at the terminal of sugar chains of human alpha 1-acid glycoprotein (AGP) was investigated with respect to chiral discrimination between optical isomers of basic drugs, using high-performance capillary electrophoresis/frontal analysis (HPCE/FA), a novel analytical method developed for the determination of unbound drug concentration with ultramicrosample volume (100-200 nl). Native human AGP and desialylated AGP were used as test proteins, and propranolol (PRO) and verapamil (VER) were used as model drugs. The unbound concentration of (S)-VER was 1.31 times higher than that of (R)-VER in native AGP solution. This selectivity was not affected by desialylation. Further, enzymatic elimination of galactose residues, which neighbored sialic acid groups, did not change the binding of either isomer of VER. On the other hand, the unbound concentration of (R)-PRO was 1.27 times higher than that of (S)-PRO in native AGP solution. Desialylation caused the unbound concentration of (S)-PRO to rise to the same level of (R)-PRO, resulting in loss of enantioselectivity. Thus, it follows that sialic acid groups of AGP, as a whole, are not responsible for chiral recognition between enantiomers of VER but are involved in enantioselectivity toward the isomers of PRO.

Protein-binding high-performance frontal analysis of (R)- and (S)-warfarin on HSA with and without phenylbutazone.

Applicability of high-performance frontal analysis (HPFA) to the stereoselective study of drug-drug interaction upon plasma protein binding has been investigated. Racemic warfarin and phenylbutazone were used as model drugs. An on-line HPFA/HPLC system consisting of a HPFA column (diol-silica column), an extraction column, and a chiral separation column was developed, and human serum albumin solution containing racemic warfarin and/or phenylbutazone was injected directly to the HPFA column. When the injection volume was large enough, the binding equilibrium in the sample solution was reproduced in the column, and consequently a plateau region appeared on the chromatogram. This plateau region contains unbound drug(s). A given volume of eluent in the plateau part was transferred into the extraction column by column-switching. The concentrated drug(s) was then transferred to the chiral separation column to determine the unbound concentrations of the enantiomers and/or the competitor. The results agreed with those obtained by a conventional ultrafiltration-HPLC method. The influence of phenylbutazone upon the protein binding of warfarin is enantioselective. In warfarin and human serum albumin mixed solution, the unbound concentration of (R)-warfarin was 1.22 times higher than that of the S-isomer. By addition of phenylbutazone, the unbound concentration of (S)-warfarin increased more than that of (R)-warfarin, resulting in the reversed enantioselectivity, i.e., the unbound concentration of (S)-warfarin became 1.19 times larger than that of (R)-warfarin. The present method was also applicable to human plasma samples.

Separation mechanism of pullulan solution-filled capillary electrophoresis of sodium dodecyl sulfate-proteins.

The separation mechanism of capillary electrophoresis of sodium dodecyl sulfate (SDS)-proteins using pullulan with a molecular mass range of 50,000-100,000 as a separation matrix was investigated. The pullulan solution was filled into fused-silica capillaries whose inner walls were coated with linear polyacrylamide through chemically stable Si-C linkages. Baseline separations of SDS proteins were achieved at concentrations ranging from 3-10% w/v of pullulan. The entanglement threshold of pullulan solution was found to be around 0.5% w/v, indicating migration of SDS-proteins through an entangled pullulan network. Ferguson plots exhibited a linear relationship between log mobility and pullulan concentration. Linear relationships were also obtained for double logarithmic plots of the electrophoretic mobility and protein molecular mass. These results show that the separation is based on mass discrimination in accordance with the Ogston theory.

Effect of temperature and viscosity of sieving medium on electrophoretic behavior of sodium dodecyl sulfate-proteins on capillary electrophoresis in presence of pullulan.

Pullulan was used as a sieving medium in high-performance capillary gel electrophoresis to study the effect of temperature and viscosity on electrophoretic behavior of sodium dodecyl sulfate (SDS)-proteins with molecular masses in the range of 14,400-116,000 Da. The migration time decreased with increasing capillary temperature. Equations were derived relating mobility of SDS-proteins to capillary temperature and viscosity of pullulan solution. Linear relationships were obtained from experiments between logarithm of mobility and reciprocal temperature and between double logarithmic values of mobility and viscosity of pullulan solution. The experimental results agree well with the equations.

Prediction of the human pharmacokinetics of troglitazone, a new and extensively metabolized antidiabetic agent, after oral administration, with an animal scale-up approach.

We have attempted to predict the human pharmacokinetics of troglitazone after oral administration based on animal data. Troglitazone is a new antidiabetic agent that exhibits a high-metabolic clearance and is metabolized mainly in the liver to sulfate and glucuronide conjugates. The prediction of the area under the plasma concentration-time curve (AUCp.o.) and bio-availability (F) in humans after oral administration was initially attempted by use of allometric equations involving the oral plasma clearance of total (CLp.o.) or unbound drug (CLp.o.,fu), or the hepatic intrinsic clearance of unbound drug (CLuint) and animal body weight. The exponents in the allometric equations between the clearances and body weights were 0.63 to 0.82 with high correlation coefficients (r > .98), and there was no marked difference in predictability by the three methods. Next, the prediction of the range of plasma profiles after oral administration to humans was attempted by the following series of steps: (1) calculation of the exponent and coefficients in the allometric relationships between body weight and parameters, such as total body plasma clearance (CLi.v.) and various distribution volumes (Vss, V beta and Vc) based on animal data; (2) estimation of the absorption rate constant (ka) from allometric relationship to body weight, and estimation of F value from the predicted AUCp.o. (3) description of the plasma concentration-time profiles after oral administration by an equation involving the allometric exponents and coefficients, ka, F and body weight. The observed and simulated plasma profiles were similar and the predicted AUCp.o. values were 60 to 120% of those observed. These methodologies will be useful for predicting the human pharmacokinetics after oral dosing from animal data.

Theoretical study of high-performance frontal analysis: a chromatographic method for determination of drug-protein interaction.

High-performance frontal analysis (HPFA), a chromatographic method to determine unbound drug concentration in drug-protein binding equilibrium, has been considered on the basis of a theoretical plate model, where a rapid equilibrium of drug-protein binding in the mobile phase in the interstices of packing materials and a chromatographic partition equilibrium of the drug were taken into account simultaneously. When a certain excess volume of drug-protein mixed solution is injected directly into a HPFA column packed with a restricted-access type phase that excludes protein but retains drug in the micropores, the drug is eluted as a zonal peak with a plateau region. The elution profile can be well simulated by the mass balance equation derived according to a relatively simple plate theory concept, which confirms that the drug concentration in the plateau range agrees with the unbound drug concentration in the sample solution. The model was applied to the theoretical and systematic investigation of the dependence of the HPFA profile on several chromatographic conditions and the properties of the sample solution, such as injection volume of sample solution, drug and protein concentrations in sample solution, capacity factor of the drug, theoretical plate number, and binding parameters. The smaller capacity factor and the higher column efficiency lead to the larger plateau volume. The lower drug concentration, the higher protein concentration, and the stronger binding constant, which give the lower unbound drug fraction, lead to the larger plateau volume and allow frontal analysis with a smaller sample size.

Capillary electrophoresis/frontal analysis for microanalysis of enantioselective protein binding of a basic drug.

A new HPCE method was developed for the enantioselective determination of the unbound concentration of a basic drug under plasma protein binding equilibrium. The racemic basic drug verapamil (VER) and human serum albumin mixed solution was used as a model sample solution. The sample solution was introduced into a fused-silica capillary hydrodynamically or electrokinetically. During the electrophoresis following hydrodynamic injection, the unbound drug zone migrated apart from the sample zone and was separated into two zonal peaks with a plateau due to enantiomers by a chiral selector (trimethyl-beta-cyclodextrin) dissolved in the acidic running buffer solution (pH 2.5). By the electrokinetic introduction of the same sample solution from the anodic end, only the unbound drug entered the capillary and was separated into the enantiomers, which also gave the zonal peaks with plateau. The unbound concentration of each enantiomer was determined from the plateau peak height. The results obtained by the different methods for sample introduction agreed well with those determined by conventional ultrafiltration-chiral HPLC, which was employed as a reference method. The unbound concentration of (S)-VER was 1.7 times higher than that of the antipode. The sample size used in the present method was approximately 200 nL, which is about one-thousandth of that in the reference method. The electrokinetic introduction gave a better peak shape than the hydrodynamic introduction.

Chemical stability of polyacrylamide-coating on fused silica capillary.

The chemical stability of polyacrylamide coatings on a fused silica capillary for capillary electrophoresis was investigated with respect to temperature of capillary and pH as well as salt concentration of running buffer solution by evaluating their effects on electroosmotic mobility. The capillary covalently bonded to linear polyacrylamide through Si-C linkage was stable in buffer solutions of pH 2.3-8.0 at 30 degrees C for 30 days, whereas the capillary bonded through Si-O-Si-C linkage was damaged at pH above 4.6. The electroosmotic mobility observed in both capillaries increased with increasing temperature, though the degree of increase was smaller in the Si-C linked capillary than in the Si-O-Si-C linked capillary. The increase in the buffer concentration resulted in increased electroosmotic mobility in the Si-O-Si-C linked capillary, but no effect was observed for the Si-C linked capillary.

Michaelis-Menten analysis of immobilized enzyme by affinity capillary electrophoresis.

Michaelis constant of enzymatic reaction was evaluated by affinity capillary electrophoresis using beta-galactosidase as a model enzyme and o- and p-isomers of nitrophenyl-beta-galactoside as substrates. The enzyme was immobilized on the inner surface of a fused-silica capillary by the covalent bonding through a bridging group, and the substrates were introduced into the capillary. The reaction products migrated electrophoretically toward the detection side (anodic side), while the unreacted substrates moved toward the injection side (cathodic side) on a slow electroosmotic flow generated by the weak negative charge of the immobilized enzyme. The initial velocity of the enzymatic reaction was estimated from the peak height of the product, and the Michaelis constant was calculated according to Lineweaver-Burk equation. The results (Km, 2.34 mM for o-isomer and 1.09 mM for p-isomer) were reproducible (RSD < 11.8%, n = 5). Although the estimated Michaelis constants were larger than the reported values measured in homogeneous solution, the ratio of the Michaelis constants of o-/p-isomers was in good agreement with the literature value. The present method required as low as a few microgram amount of enzyme and nanogram amount of substrate which is far smaller than those required in a conventional affinity HPLC.