Extensive Binding of the Bioflavonoid Quercetin to Human Plasma Proteins

Although the bioflavonoids, a large group of polyphenolic natural products, exert chemopreventive effects in cardiovascular disease and cancer, there is little information about the disposition of these dietary components in man. The objective of this study was to investigate the plasma-protein binding of the most abundant bioflavonoid, quercetin, using ¹⁴C-labelled quercetin. An ultracentrifugation assay (170 000 g for 16 h at 20°C) was shown to sediment plasma proteins. Binding of quercetin to normal plasma was extensive (99.1 ± 0.5%, mean ± s.d., n = 5). The unbound fraction varied as much as 6-fold, 0.3–1.8%, between subjects. This high binding was independent of quercetin concentration over the range 1.5–15 μM (0.5–5 μg mL^?1). Human serum albumin was the primary protein responsible for the binding of quercetin in plasma (99.4 ± 0.1%). Binding by α₁-acid glycoprotein (39.2 ± 0.5%) and very-low-density lipoproteins (< 0.5% of total quercetin) did not make substantial contributions to overall plasma binding. The equilibrium association constant for the binding of quercetin to serum albumin was 267 ± 33 times 10³ M^?1 (n=15). Thermodynamic data for the binding of quercetin to serum albumin indicated spontaneous, endothermic association. Displacement studies suggested that in man the ‘IIA’ subdomain binding site of human serum albumin was the primary binding site for quercetin. Association of quercetin with erythrocytes was significantly (P < 0.001) reduced by plasma protein binding. These data indicate poor cellular availability of quercetin because of its extensive binding to plasma proteins.     

The Binding of Thiopental to Serum Proteins Determined by Ultrafiltration and Equilibrium Dialysis

Abstract: An ultra filtration method is described by which it is possible to estimate the protein bound fraction of a drug at its original concentration in a serum sample. The determination is carried out at 37° and pH 7.4. For thiopental no difference was found between the values of protein binding whether they were determined by the ultrafiltration method or by a less time consuming equilibrium dialysis against an equal volume of phosphate buffered isotonic sodium chloride solution. The equilibrium dialysis was used to measure the concentration of bound and unbound thiopental molecules; and the binding parameters in a two class binding model were determined. No evidence was found for binding to other proteins than albumin. About one binding site belonging to the first class was found per 1000 albumin molecules whereas an average of about 5 secondary sites were found for each albumin molecule. The association constants for the primary class of binding sites were 3.4×10⁶M^?1 for albumin and 13.3×10⁶M^?1 for serum while the values for the secondary class were 2.2 and 1.2×10³M^?1 for albumin and serum respectively. The estimates of association constants and number of binding sites were based on experiments with total thiopental concentrations ranging from 0.4 to 80 μg/ml. In a phosphate buffered albumin solution with 2 g albumin per 100 ml, a pH increase from 5 to 8 caused an increase in protein binding from 36 to 76%. A Scatchard plot using data from experiments with increasing albumin concentrations resulted in a “plot” with positive slope. The use of a curve like this is discussed and it is concluded that the binding parameters for thiopental are influenced by the albumin concentration.     

Plasma protein binding of fentanyl

Whole plasma of 15 normal volunteers bound 79·16% ± 0·16 s.e. (n = 45) of fentanyl, at a concentration of 0·6 ng ml^?1. Measurement was by equilibrium dialysis at 37°C, pH 7·4. The largest contribution to binding appears to be due to serum albumin, since 45·52% ± 0·40 s.e. (n = 3) of fentanyl was bound to a solution of purified albumin at a concentration of 46 g litre^?1 in buffer. Physiological concentrations of isolated very low density, low density and high density lipoprotein fractions in buffer bound 18·39% ± 0·65 s.e.; 39·14% ± 0·42 and 21·18% ± 0·51 (n = 15) respectively. A significant correlation was found between percent binding and serum albumin concentration (r = 0·745, P = 0·0022) and oestrogen and progestagen therapy (r = 0·766, P = 0·0014). There was no significant correlation with fasting serum cholesterol, triglyceride, age, sex or the concentration of total protein minus albumin. Binding to fibrinogen and α₁-acid glycoprotein did not occur. Binding increased with increasing pH, temperature and ionic strength of the buffer. The results were compatible with hydrophobic bond formation between fentanyl and proteins. Fentanyl concentration did not affect the percent bound to whole plasma or the protein fractions over a range of 0·6 ng-10 mg ml^?1. Dilution of plasma with buffer gave a linear relation of percent bound or free to log plasma dilution. The binding of fentanyl to pooled plasma of the normal subjects was not affected by a wide variety of anionic, cationic and uncharged drugs when these were tested at a concentration of 20 μg ml^?1. At higher concentrations, aspirin, phenylbutazone and quinidine caused inhibition of fentanyl binding. A linear relation was found between percent bound and concentration of the inhibitory ligand. For aspirin, r = 0·873, P <0·01; for phenylbutazone, r = 0·81, P <0·05; and for quinidine, r = 0·982, P <0·01. Aspirin and phenylbutazone inhibited binding of fentanyl to albumin, while quinidine caused inhibition of binding to lipoproteins of all three densities, but not to albumin. Changes in the concentrations of the common ions of plasma (except H⁺), of free fatty acids and of creatinine did not affect fentanyl binding to whole plasma. 8 M urea reduced binding by 25% of the normal value.     

Protein binding of sotalol enantiomers in young and elderly human and rat serum using ultrafiltration

The protein binding of sotalol (STL) enantiomers was evaluated using an ultrafiltration technique with serum from young (32±2 years, n=5) and elderly (73±6 years, n=5) male and female humans, and young (8 weeks, n=4) and elderly (60 weeks, n=3) male Sprague—Dawley rats. Serum samples were collected and immediately frozen at ?20°C. Within 1 week, the serum samples were thawed at room temperature, and adjusted to pH 7.4 using 0.05 M phosphate buffer, pH 5.0. Aliquots were spiked with 250 ng mL^?1 and 500 ng mL^?1 of each STL enantiomer, placed in ultrafiltration sets (Microsep, 30K molecular weight cut-off), capped, equilibrated to 37°C, and centrifuged at 1850g for 1.5h at 37°C. Aliquots of ultrafiltrate and unspun serum were analysed for STL enantiomer concentration using a stereospecific HPLC assay. In all groups, bound fraction was less than 7% for both STL enantiomers. There were no significant differences in bound fraction between groups, or between enantiomers. Adsorption of STL enantiomers to the ultrafiltration device and membrane, evaporative loss of serum samples during centrifugation, and protein concentration in each ultrafiltrate sample were all negligible. It is concluded that the binding of STL in human and rat serum at therapeutic concentrations and physiological temperature and pH is negligible and non-stereoselective.     

Pharmacokinetics of reboxetine in healthy volunteers. Single oral doses,linearity and plasma protein binding

The pharmacokinetics of reboxetine, a new antidepressant agent, were found to be close to linear in a crossover study comparing administration of single 2, 3, 4 and 5 mg capsule doses in 15 healthy male volunteers, and in the same study the capsules were bioequivalent to the proposed therapeutic tablet formulation (4mg). Kinetic analysis was based on HPLC assay of reboxetine in plasma and urine collected up to 72 h after each administration. Plasma levels indicated a rapid absorption (t_max?2h) and an elimination half-life of about 13 h. Clearance and volume of distribution were modest (ratios to bioavailability: CL/F?29 mL min^?1; V_z/F?32L); urinary excretion was ～9% of dose, corresponding to a renal clearance of only 3 mL min^?1 (a value consistent with the rate of glomerular filtration of unbound drug). In vitro, binding to plasma proteins, estimated from radioactivity levels following dialysis of ¹⁴C-labelled reboxetine, appeared to be dominated by α₁-acid glycoprotein without marked saturation up to plasma concentrations of over 500 ng mL^?1 (2.8–3.1% unbound with human plasma from three additional volunteers; 1.8–2.0% for 2gL^?1 orosomucoid α₁-acid glycoprotein, and 46.4–47.4% for 40 gL^?1 albumin), whilst the mean C_max in the current study was much lower (164 ng mL^?1 after a 5 mg dose).     

Protein Binding of Doxepin and Desmethyldoxepin

Abstract: The protein binding of doxepin (DOX) and desmethyldoxepin (DDOX) were studied in serum and plasma samples from healthy volunteers and psychiatric patients. Binding was measured by equilibrium dialysis (16 hrs at 37°) and drug concentrations by radioimmunoassay. In addition, albumin and α₁-acid glycoprotein concentrations of the samples were measured by radial immunodiffusion. The mean ± SEM percentages of unbound DOX were: 20.4 ± 1.2 and 15.9 ± 1.2 in healthy subjects (n = 16) and patients (n= 15) respectively. and those of DDOX: 21.4 ± 0.9 and 19.0 ± 1.4 for healthy subjects and patients, respectively. There was a significant negative correlation between serum α₁-acid glycoprotein concentration and free fraction of DOX in both groups. In healthy subjects a significant negative correlation was also found between albumin concentrations and free fraction of both DOX and DDOX. Binding experiments with isolated protein fractions revealed that all of the total binding in plasma could be explained by binding to albumin and α₁-acid glycoprotein. The observed 2–4-fold interindividual variability in the free fractions of these drugs is probably less important than the much larger variability in the total serum concentrations.     

Pharmacokinetics of ibuprofen in man—III: Plasma protein binding

Plasma protein binding of ibuprofen was measured by equilibrium dialysis on 406 plasma samples collected from 15 normal volunteers following doses of 400, 800, and 1200 mg of ibuprofen as tablets (N=102, 100, 104, respectively) and 420 mg as an aqueous solution (N=100). Individual subject bound concentration at dialysis equilibrium (C_bd) vs. free concentration at dialysis equilibrium (C_fd) were well fitted via computer to the Scatchard equation with one class of binding sites. The binding capacity averaged 1231 μM (range 848–1658 μM), and the association constant averaged 1.76 × 10⁵M^?1 (range 1.15 × 10⁵ to 2.73 × 10⁵M^?1). Distributional analysis was performed on the free fraction (f_d) and bound/free ratios (C_bd/C_fd=1/f_d?1) at dialysis equilibrium for each treatment. Using pooled data of all four treatments, distributional analysis was also performed on the free fractions (f) and bound/free ratios (C_b/C_f=1/f?1) corresponding to the plasma drug concentrations in blood as it was withdrawn from the subjects. The bound/free ratios were normally distributed, whereas the distributions of the free fractions were skewed towards higher values.     

In-vivo Distribution and Erythrocyte Binding Characteristics of Cyclosporin in Renal Transplant Patients

The pharmacokinetic parameters of cyclosporin, a potent immunosuppressive agent, show large intra-and inter-individual variability, possibly because of the different analytical methods used. A recently developed cyclosporin-specific radioimmunoassay has been used to study the in-vivo distribution and binding characteristics of cyclosporin in whole blood, plasma and erythrocytes of fifteen renal transplant patients. The profiles of cyclosporin concentration-time curves after an oral dose of cyclosporin had either one peak (ten patients, group A) or two (five patients, group B). Essentially no difference was observed between the two groups in the relationship between equilibrium cyclosporin concentrations in erythrocyte and plasma as a function of whole-blood concentration. The equilibrium in-vivo cyclosporin concentrations in erythrocytes and plasma were, however, markedly lower than those previously observed under in-vitro conditions. The ratio of cyclosporin concentration in erythrocytes (C_E) to that in plasma (C_P) changed with time, in inverse proportion to the change in cyclosporin concentration in blood, over the range 0.63-2.80 in individual patients with an average of 1.36 ± 007 (mean ± s.e.m.) for group A and 1.42 ± 0.23 for group B. The apparent cyclosporin binding affinity (K_d) to erythrocytes under in-vivo conditions averaged 452.2 ± 47.6 nm (543.5 ± 57.2 ng mL^?1) for group A and 419.4 ± 41.2 nm (504.1 ± 49.5 ng mL^?1) for group B, whereas apparent cyclosporin binding capacity (B_max) of the blood cell averaged 0.83 ± 0.07 nmol mL^?1 for group A and 0.78 ± 0.07 nmol mL^?1 for group B. Significantly reduced average K_d (262.7 ± 40.2 nm or 315.8 ± 48.9 ng mL^?1, P < 001) and B_max (0.56 ± 008 nmol mL^?1, P < 005) values were observed during the period after T_max (4–12 h after the drug ingestion) in group A patients. Apparent K_d and B_max, determined by a nonlinear regression technique, were 131.6 ± 29.4 and 1088.0 ± 114.7 nm (158.2 ± 35.4 and 1307.8 ± 137.9 ng mL^?1) and 0.178 ± 0.024 and 0.814 ± 0.078 nmol mL^?1, respectively, during the 4–12 h period in group A patients. These findings reveal distinct differences in in-vivo distribution of cyclosporin and the binding characteristics of the compound to erythrocytes from those previously observed under in-vitro conditions. The significantly lower K_d of cyclosporin binding to erythrocytes during the elimination phase suggests a potential effect of cyclosporin-containing erythrocytes or of cyclosporin contained in erythrocytes during cyclosporin treatment.     

Drug-Biomacromolecule Interaction XII: Comparative binding study of sulfaethidole to bovine serum albumin by equilibrium dialysis,fluorescence probe technique,uv difference spectrophotometry and circular dichroism

Binding of sulfaethidole to bovine serum albumin was studied by equilibrium dialysis, fluorescence probe technique, uv difference spectrophotometry and circular dichroism. Equilibrium dialysis method enabled us to estimate the total number of drug binding sites of albumin molecule. For sulfaethidole, albumin had 6 primary and 40 secondary binding sites. The primary and secondary binding constants were 0.9×10⁵ M⁻¹ and 0.2×10⁶ M⁻¹, respectively. 1-Anilino-8-naphthalenesulfonate (ANS) and 2-(4′-hydroxylbenzeneazo)-benzoic acid (HBAB) were used as the fluorescence probe and the uv spectrophotometric probe, respectively. In fluorescence probe technique, results indicated that the number of higher affinity drug binding site of albumin was 1 and the number of lower affinity drug binding sites of albumin was 3, and the primary and secondary drug binding constants for bovine serum albumin were 2.15×10⁵ M⁻¹ and 1.04×10⁵ M⁻¹, respectively. In uv difference spectrophotometry, binding sites were 3 and binding constant was 1.88×10⁵ M⁻¹. The above results suggest that several different methods should be used in ompensation for insufficient information about drug binding to albumin molecule given by only one method.     

Variable binding of propranolol in human serum.

Human serum proteins were fractionated by ultracentrifugation and gel filtration. Binding of propranolol was determined by equilibrium dialysis. Propranolol was distributed to lipoproteins independent of drug concentration. Two groups of propranolol binding sites were found to be present in the protein preparation containing albumin, α₁-acid glycoprotein, transferrin and prealbumin. The first binding site with a dissociation constant of 7.5 × 10^?7 was present in number equivalent to concentration of α₁-acid glycoprotein. The propranolol binding to serum samples from 21 healthy males expressed as binding ratio B/F and per cent binding ranged from 7.5 to 19.2 and 88.2 to 95.0 respectively. The binding ratio was correlated to concentration of α₁-acid glycoprotein (r = 0.85, P < 0.001), but not to concentrations of albumin and lipoproteins. The results indicate that α₁-acid glycoprotein is the main propranolol binding protein in human serum.     

Factors influencing the protein binding of azosemide using an equilibrium dialysis technique

Various factors most likely to influence the plasma protein binding of azosemide to 4% human serum albumin (HSA) were evaluated using equilibrium dialysis at the initial azosemide concentration of 10 μg mL^?1. It took approximately 8h of incubation to reach an equilibrium between 4% HSA and isotonic phosphate buffer of pH 7.4 containing 3% dextran (the ‘buffer’) using a Spectra/Por 2 membrane (molecular weight cut-off 12000–14000) in a water bath shaker kept at 37°C and a rate of 50 oscillations min^?1. Azosemide was fairly stable both in 4% HSA and in the ‘buffer’ for up to 24h. The binding of azosemide to 4% HSA was constant (95.5 ± 0.142%) at azosemide concentrations ranging from 5 to 100 μg mL^?1. However, the extent of binding was dependent on HSA concentration: the values were 88.4, 91.0, 92.2, 94.2, 94.9, 94.9, and 94.9% at albumin concentrations of 0.5, 1, 2, 3, 4, 5, and 6% respectively. The binding was also dependent on incubation temperature; the binding values were 97.0, 94.9, and 94.9% when incubated at 6, 28, and 37°C, respectively. The binding of azosemide was also influenced by buffers containing various chloride ion concentrations and buffer pHs. The binding values were 95.3, 94.9, and 93.6% for the chloride ion concentrations of 0, 0.249, and 0.546%, respectively, and the unbound values were 6.8, 5.1, 3.8, 3.4, and 3.3% for buffer pHs of 5.8, 6.4, 7.0, 7.4, and 8.0, respectively. The binding of azosemide was independent of the quantity of heparin (up to 40 UmL^?1), AAG (up to 0.16%), sodium azide (NaN₃, up to 5%), its metabolite, Ml (up to 10 μg mL^?1), and anticoagulants (EDTA and citrate).     

Protein binding of methohexital. Study of parameters and modulating factors using the equilibrium dialysis technique

This paper describes the parameters that characterize methohexital—albumin binding and the influence of physiological or analytical factors on this binding. Two useful and reproducible methods for measuring the free concentration—equilibrium dialysis (ED) and ultrafiltration (UF)—are described and their performances are compared. Methohexital binds exclusively to albumin according to a two-class binding model. The first is a saturable class site of high affinity constant (K_A = 11 200 M⁻¹) and a number of sites per albumin molecule of 1. The second is a non-saturable site of poorer affinity (nK_A = 810 M⁻¹). The bound fraction of methohexital in the therapeutic range and at physiological albumin concentration is 86.7 ± 0.9% in isolated albumin solution. In serum, it ranges from 80 to 84.5%, according to subjects (n = 6). Binding is inhibited by the presence of endogeneous compounds of serum (for a given albumin concentration the bound fraction decreases from 90.3% in isolated albumin solution to 82.6% in serum), probably by free fatty acids. An increase in the bound fraction is observed when the pH is increased from 7 to 9. This phenomenon may be explained by a higher affinity of the drug towards the basic (B-form) conformation of the albumin molecule, in analogy with the close barbiturate thiopental. A decrease in the bound fraction against temperature is shown, as though binding forces diminished with increase in temperature. Indeed, the binding modification is less pronounced in the presence of serum endogenous compounds. As expected, there is no evidence of any effect of heparin anticoagulant on the bound fraction. Methohexital binding is strongly modified by the albumin concentration; the bound fractions change from 67 to 91% in the albumin range 150–900 μM.     

Determination of Protein Binding of a Highly Lipophilic Drug,Isocarbacyclin Methyl Ester (TEI-9090), Using a Polydimethylsiloxane-coated Glass Beads Assay

Abstract— An adsorption technique with polydimethylsiloxane-coated glass beads (PDMS-GB) was developed to determine the protein binding of a highly lipophilic and hydrophobic drug. The present assay method is based on the quantitative adsorption of unbound drug to the PDMS-GB. This method of batch separation in a glass assay tube has an advantage of simplicity and rapidity. To evaluate the reliability of PDMS-GB assay, we compared the protein binding of diazepam in serum in-vitro measured by ultrafiltration and PDMS-GB assay. There was no significant difference between the extent of binding measured by each method. Using PDMS-GB assay, we determined the protein binding of the prostaglandin I₂ (PGI₂) analogue isocarbacyclin methyl ester (TEI-9090), whose binding cannot be measured by commonly employed techniques (equilibrium dialysis, ultrafiltration, gel filtration or ultracentrifugation) because of a high degree of adsorption to membranes, resins or tubes. The percentage of TEI-9090 bound in human serum, 4% human serum albumin (HSA, fatty acid-free) and dog serum were ～98, ～87 and ～95%, respectively, and these values were independent of TEI-9090 concentration up to 10 ng mL^?1. The binding of isocarbacyclin (TEI-7165) to serum protein in man, dogs, rabbits and rats, determined by ultrafiltration, was also high (>90%). While the displacement of TEI-9090 and TEI-7165 binding to HSA by aspirin, salicylic acid and indomethacin was not observed, clofibric acid and free fatty acids significantly inhibited the protein binding of both compounds. These results indicate that the binding site of TEI-9090 and TEI-7165 on HSA could be identical with the possible binding site of PGI₂.     

Renal Excretion and Accumulation Kinetics of 2-Methylbenzoylglycine in the Isolated Perfused Rat Kidney

The effect of protein binding on kidney function has been studied by investigating the renal accumulation and secretion of the hippurate analogue 2-methylbenzoylglycine in the isolated perfused rat kidney in the absence and presence of bovine serum albumin (BSA). Experiments were performed with either 2.5% pluronic or a combination of 2.2% pluronic and 2% BSA as oncotic agents; a wide concentration range (1–190 μg mL^?1) of 2-methylbenzoylglycine was studied. Tubular secretion appeared to be a function of the amount of unbound drug in the perfusate and was best described by a model consisting of a high and low affinity Michaelis-Menten term. Parameters obtained after the analysis of renal excretion data were maximum transport velocity for the high affinity site (T_M,H) = 3.0 ± 2.8 μg min^?1, Michaelis-Menten constant for tubular transport for the high affinity site (K_T,H) = 0.5 ± 0.8 μg mL^?1, maximum transport velocity for the low affinity site (T_M,L) = 250 ± 36 μg min^?1, and Michaelis-Menten constant for tubular transport for the low affinity site (K_T,L) = 62 ± 17 μg mL^?1. The compound accumulated extensively in kidney tissue, ratios up to 175 times the perfusate concentration were reached. Accumulation data were best analysed by a two-site model similar to the model used to describe renal excretion. Calculated parameters were theoretical maximum capacity of the high affinity site (R_M,H) = 26 ± 23 μg g^?1, affinity constant for renal accumulation at the high affinity site (K_A,H) = 0.2 ± 0.4 μg mL^?1, theoretical maximum capacity of the low affinity site (R_M,L)= 1640 ± 1100 μg g^?1 and affinity constant for renal accumulation at the low affinity site (K_A,L) = 60 ± 58 μg mL^?1. The very high accumulation in kidney tissue could be explained by active tubular uptake, mediated by the secretory mechanisms involved, and dependent on the amount of free drug in the perfusate. This study shows that anionic drugs, subject to active secretion, may reach high concentrations in tubular cells even at low plasma concentrations.     

Influence of Fatty Acids on the Binding of Warfarin and Phenprocoumon to Human Serum Albumin with Relation to Anticoagulant Therapy

Warfarin and phenprocoumon binding to human serum albumin was studied by equilibrium dialysis. The first stoichiometric binding constant was 1.89 × 10⁵ M^?1 for warfarin and 2.40 × 10⁵ M^?1 for phenprocoumon. The affinity of warfarin was markedly increased on addition of up to 3 mol mol^?1 albumin of palmitic, stearic, oleic or linoleic acids with energetic couplings for co-binding of one molecule of each of the fatty acids and one molecule of warfarin of 0.9, 1.1, 0.7 and 0.6 kJ mol^?1, respectively. The affinity of phenprocoumon was only increased slightly on addition of palmitate with an energetic coupling of 0.3 kJ mol^?1. Six consecutive serum samples were obtained from each of 14 patients undergoing surgery. The serum affinity of the drugs varied considerably corresponding to free drug concentrations between 0.7 and 2.7% for warfarin and between 0.8 and 4.9% for phenprocoumon. The affinity of warfarin but not of phenprocoumon was correlated to the increasing plasma fatty acid concentration. Anticoagulant therapy with phenprocoumon may thus be less sensitive than warfarin to changes in the fatty acid concentration of plasma.     

Plasma protein binding and interaction studies with diflunisal,a new salicylate analgesic

The binding of diflunisal to human serum albumin and normal human plasma has been studied by equilibrium dialysis at 37°, pH 4. The plasma protein binding data were analysed according to a Scatchard model with two independent classes of binding sites. The number of binding sites and the corresponding association constants have been estimated by nonlinear least-squares regression analysis: N₁ = 2.1, k₁ = 5.28 × 10⁵M^?1, N₂ = 7.7 and K₂ = 0.17 × 10⁵M^?1. At a difluni concentration of 50 μg/ml on average 99.83 per cent of the drug was bound to plasma proteins. The in vitro plasma protein binding of diflunisal was impaired by salicylic acid and phenprocoumon, while diflunisal itself was displaced from its primary binding sites in plasma by salicylic acid and bilirubin. Tolbutamide had no effect on the binding of diflunisal to plasma proteins.     

Plasma Protein Binding of the Experimental Antitumour Agent Acridine-4-carboxamide in Man,Dog, Rat and Rabbit

Abstract— The plasma binding of N-[2-(dimethylamino)ethyl]acridine-4-carboxamide (AC) was investigated in-vitro by equilibrium dialysis for 3 h at 37°C against isotonic phosphate buffer (pH 7·35) using [³H]AC. There were significant species differences with the smallest % free fraction (mean ± s.d.) occurring in human plasma (3·4 ± 0·2), followed by dog (8·1 ±0·4), mouse (14·8 ± 0·8), rat (16·3 ± 0·9) and rabbit (20·2 ± 0·7). In plasma from healthy individuals (n = 5), the % free fraction ranged from 2·7 to 3·8. In physiological solutions of human proteins, the greatest binding was observed for α-acid glycoprotein (AAG) (0·75 g L^?1) with a mean free fraction of 24·1 ± 2·2%, followed by albumin (40 g L^?1) with 31·6 ± 0·7 and 39·8 ± 2·5% for fatty-acid-free and globulin-free, respectively. There was also some binding to globulins (5 g L^?1) with a mean % free fraction of 70·3 ± 1·6 and 84·8 ± 2·2 for Conn's fraction I and IV, respectively. Binding data from the displacement of [³H]AC by increasing concentrations of AC in human AAG (0·75 g L^?1) or albumin solution (40 g L^?1) indicated that AAG had 10-fold greater binding affinity for AC (K_a, 7·8 × 10⁴ m^?1) compared with albumin (K_a, 6·8 × 10³ m^?1). In human plasma enriched with AAG there was a significant negative linear correlation (r = 0·932; P < 0·001) between % AC free fraction and increasing AAG concentration over the range 0·6–4·5 g L^?1. Small but significant (P < 0·05) increases in AC free fraction occurred in the presence of various metabolites (50 and 100 μm) but, of those tested, only N-monomethyl-acridine carboxamide increased the free fraction to the same extent as parent AC.     

Disposition and bioavailability of the β1-adrenoceptor antagonist talinolol in man

In an open randomized crossover study, the pharmacokinetics and bioavailability of the selective β₁-adrenoceptor antagonist talinolol (Cordanum®—Arzneimittelwerk Dresden GmbH, Germany) were investigated in twelve healthy volunteers (five female, seven male; three poor and nine extensive metabolizers of the debrisoquine hydroxylation phenotype) after intravenous infusion (30 mg) and oral administration (50 mg), respectively. Concentrations of talinolol and its metabolites were measured in serum and urine by HPLC or GC-MS. At the end of infusion a peak serum concentration (C_max) of 631 ± 95 ng mL^?1 (mean ± SD) was observed. The area under the serum concentration-time curve from zero to infinity (AUC_0-∞) was 1433 ± 153 ng h mL^?1. The following parameters were estimated: terminal elimination half life (t_1/2), 10.6 ± 3.3 h; mean residence time, 11.6 ± 3.1 h; volume of distribution, 3.3 ± 0.5 L kg^?1; and total body clearance, 4.9 ± 0.6 mL min^?1 kg^?1. Within 36 h 52.8 ± 10.6% of the administered dose was recovered as unchanged talinolol and 0.33 ± 0.18% as hydroxylated talinolol metabolites in urine. After oral administration a C_max of 168 ± 67 ng mL^?1 was reached after 3.2 ± 0.8h. The AUC_0-∞ was 1321 ± 382 ng h mL^?1. The t_1/2 was 11.9 ± 2.4 h. 28.1 ± 6.8% of the dose or 55.0 ± 11.0% of the bioavailable talinolol was eliminated as unchanged talinolol and 0.26 ± 0.17% of the dose as hydroxylated metabolites by kidney. The absolute bioavailability of talinolol was 55 ± 15% (95% confidence interval, 36–69%). Talinolol does not undergo a relevant first-pass metabolism, and its reduced bioavailability results from incomplete absorption. Talinolol disposition is not found to be altered in poor metabolizers of debrisoquine type.     

Binding of gefitinib, an inhibitor of epidermal growth factor receptor-tyrosine kinase, to plasma proteins and blood cells: in vitro and in cancer patients

Summary Gefitinib exhibits wide inter-subject pharmacokinetic variability which may contribute to differences in treatment outcome. Unbound drug concentrations are believed to be more relevant to pharmacological and toxicological responses than total drug. Thus it is desirable to determine gefitinib binding in plasma and factors affecting this process. An equilibrium dialysis method using 96-well microdialysis plates was optimized and validated for determining the fraction unbound (fu) gefitinib in human plasma. Gefitinib binding in plasma from four different species and isolated protein solutions as well as drug partitioning in human blood cells were investigated. Unbound gefitinib plasma concentrations were measured in 21 cancer patients receiving daily oral gefitinib 250 mg or 500 mg. It was found that gefitinib was extensively bound in human rat mouse and dog plasma with mean fu values of 3.4%, 3.8%, 5.1% and 6.0% respectively. In isolated protein solutions approximately 90% and 78% of gefitinib was bound to human serum albumin (HSA) (40 mg/dL) and alpha1-acid glycoprotein (AAG) (1.4 mg/dL) with binding constants of 1.85 × 10⁴ M⁻¹ and 1.13 × 10⁵ M⁻¹ respectively. In whole blood 2.8% of gefitinib existed as the free drug while 79.4% and 17.8% was bound to plasma proteins and blood cells respectively. In plasma from cancer patients fu at pre-treatment varied 2.4-fold (mean 3.4 ± 0.6%; range 2.2–5.4%) and fu was constant over the 28-days of treatment (P > 0.05). Pre-treatment AAG concentration was negatively correlated with pre-treatment fu (R² = 0.28, P = 0.01). In conclusion gefitinib is highly protein bound (∼ 97%) in human plasma. Variable AAG concentrations observed in cancer patients may affect gefitinib fu with implications for inter-subject variation in drug toxicity and response.     

Rat Protein Binding and Cerebral Phospholipid Affinity of the H3-receptor Antagonist Thioperamide

The binding of thioperamide, a known H₃-receptor antagonist, to rat plasma proteins and its affinity for rat cerebral phospholipids are investigated. Thioperamide is strongly bound to plasma proteins (95?80% at plasma concentrations of 3.5?400 μg mL^?1), and its binding can be resolved into two components: a high-affinity, saturable component and a non-specific component. The drug has a high affinity for cerebral phospholipids, with a partition coefficient of approximately 100 (log K = 2.06 ± 0.14), which should promote brain penetration and accumulation. Protein binding and cerebral phospholipid affinity can suggest the explanation of some differences reported in the literature on thioperamide distribution data: at low plasma concentrations of the drug, its protein binding (95% at 3.5 μg mL^?1) can prevent brain accumulation, while at higher concentrations the free plasma fraction suddenly increases (> 10% at 18 μg mL^?1) and it allows passive distribution to lipophilic tissues such as brain tissue.