Selectivity in the binding of psychotropic drugs to the variants of alpha-1 acid glycoprotein

An overview of albumin and alpha-1-acid glycoprotein main characteristics: highlighting the roles of amino acids in binding kinetics and molecular interactions

Although Albumin (ALB) and alpha-1-acid glycoprotein (AGP) have distinctive structural and functional characteristics, they both play a key role in binding a large variety of endogenous and exogenous ligands. An extensive binding to these plasma proteins could have a potential impact on drugs disposition (e.g. bioavailability, distribution and clearance), on their innocuity and their efficacy. This review summarizes the common knowledge about the structural and molecular characteristics of both ALB and AGP in humans, and about the most involved amino acids in their high-affinity binding pockets. However, the variability in residues found in binding pockets, for the same species, allows each plasma protein to interact differently with the ligands. The protein-ligand interaction influences differently the disposition of drugs that bind to either of these plasma proteins. The content of this review is useful for the design of new drug entities with high-binding characteristics, in qualitative and quantitative modelling (e.g. in vitro-in vivo extrapolations, 3D molecular docking, interspecies extrapolations), and for other interdisciplinary research.

Drug-binding energetics of human α-1-acid glycoprotein assessed by isothermal titration calorimetry and molecular docking simulations

This study utilizes sensitive, modern isothermal titration calorimetric methods to characterize the microscopic thermodynamic parameters that drive the binding of basic drugs to α-1-acid glycoprotein (AGP) and thereby rationalize the thermodynamic data in relation to docking models and crystallographic structures of the drug-AGP complexes. The binding of basic compounds from the tricyclic antidepressant series, together with miaserine, chlorpromazine, disopyramide and cimetidine, all displayed an exothermically driven binding interaction with AGP. The impact of protonation/deprotonation events, ionic strength, temperature and the individual selectivity of the A and F1*S AGP variants on drug-binding thermodynamics was characterized. A correlation plot of the thermodynamic parameters for all of the test compounds revealed that an enthalpy-entropy compensation is in effect. The exothermic binding energetics of the test compounds were driven by a combination of favorable (negative) enthalpic (∆Hº) and favorable (positive) entropic (∆Sº) contributions to the Gibbs free energy (∆Gº). Collectively, the data imply that the free energies that drive drug binding to AGP and its relationship to drug serum residency evolve from the complex interplay of enthalpic and entropic forces from interactions with explicit combinations of hydrophobic and polar side-chain sub-domains within the multi-lobed AGP ligand binding cavity.

Structure-activity relationships for the binding of polymyxins with human α-1-acid glycoprotein

Here, for the first time, we have characterized binding properties of the polymyxin class of antibiotics for human α-1-acid glycoprotein (AGP) using a combination of biophysical techniques. The binding affinity of colistin, polymyxin B, polymyxin B(3), colistin methansulfonate, and colistin nona-peptide was determined by isothermal titration calorimetry (ITC), surface plasma resonance (SPR) and fluorometric assay methods. All assay techniques indicated colistin, polymyxin B and polymyxin B(3) display a moderate binding affinity for AGP. ITC and SPR showed there was no detectable binding affinity for colistin methansulfonate and colistin nona-peptide, suggesting both the positive charges of the diaminobutyric acid (Dab) side chains and the N-terminal fatty acyl chain of the polymyxin molecule are required to drive binding to AGP. In addition, the ITC and fluorometric data suggested that endogenous lipidic substances bound to AGP provide part of the polymyxin binding surface. A molecular model of the polymyxin B(3)-AGP F1*S complex was presented that illustrates the pivotal role of the N-terminal fatty acyl chain and the D-Phe6-L-Leu7 hydrophobic motif of polymyxin B(3) for binding to the cleft-like ligand binding cavity of AGP F1*S variant. The model conforms with the entropy driven binding interaction characterized by ITC which suggests hydrophobic interactions coupled to desolvation events and conformational changes are the primary driving force for polymyxins binding to AGP. Collectively, the data are consistent with a role of this acute-phase reactant protein in the transport of polymyxins in plasma.

The pH dependency of the binding of drugs to plasma proteins in man

An analysis of pH-induced changes of drug binding may contribute to the understanding of the mechanisms involved and the clinical relevance. A literature search was performed, and acceptance criteria set up, to select reported data for quantitative evaluation. The relationship between percentage of unbound drug, fu, and pH was analyzed, and the relevance of physicochemical characteristics of the ligand drugs and the importance of hydrogen ion-induced changes in plasma proteins for the pH sensitivity of the binding were evaluated. With all basic and the majority of acidic drugs, fu depended linearly on pH. Basic drugs showed a consistent behavior with fu decreasing with increasing pH. Acidic compounds behaved differently: With some, fu increased, and with others fu decreased, with pH, and with a third group of acids fu was pH independent. Large differences in the pH sensitivity of the plasma protein binding among individual compounds were found. The fu in plasma for some bases and acids increased up to 136% and 95%, respectively, at pH values seen in severe acidemia or alkemia. These changes in fu could be clinically relevant with narrow-therapeutic-range drugs. Physicochemical properties and other characteristics of the ligands affect the pH sensitivity of the interaction with plasma proteins, but there was clear evidence indicating that pH-induced changes in the plasma proteins are also involved in the observed pH-dependent interaction with ligands. It is generally accepted that the unbound, free fraction in whole blood or plasma is an important determinant of the pharmacokinetics and pharmacodynamics of drugs. pH-dependent protein binding and consequent changes in the free fraction have been reported for many drugs. From a basic science point of view, the systematic study of pH-induced perturbations of the drug-protein interaction may provide insight into the mechanism and forces involved in the binding of drugs to plasma proteins. From a clinical viewpoint it may be of interest to know the extent of pH-induced changes in the unbound fraction of drugs under extreme acidemic or alkalemic conditions. Arterial blood pH values compatible with life reportedly range between 6.7 and 8.0. pH values as low as 6.3 have been measured in survivors of drowning accidents. To the best knowledge of the authors, a review and interpretation of pH-associated changes in the protein binding of drugs has not been attempted to date. The goals of this investigation were to (1) review published results of studies that determined the impact of pH changes on the protein binding of drugs in man, (2) select representative data using predetermined criteria, (3) determine relevant factors impacting the pH sensitivity of the drug-protein interaction, and (4) attempt to interpret the results and their clinical relevance.

Specific ligand binding on genetic variants of human α1-acid glycoprotein studied by circular dichroism spectroscopy

Human alpha1-acid glycoprotein displays genetic polymorphism. Different drug binding properties of the two main genetic products (F1-S and A variants) have been demonstrated. In search for specific circular dichroism (CD) probes, dicumarol and acridine orange were found to specifically bind to the F1-S and A variants, respectively. Dicumarol binding to the F1-S variant produced induced Cotton effects originating from the favored chiral conformation of the bound label. Acridine orange gave induced biphasic Cotton effects due to chiral intermolecular exciton interaction between label molecules bound to the A variant. Displacement of the CD probes by specific marker ligands was demonstrated. The induced CD spectrum of dicumarol was found to change sign in the presence of imipramine, as a manifestation of high-affinity ternary complex formation on the F1-S variant.

Drug interactions with patient-controlled analgesia

Patient-controlled analgesia (PCA) has become standard procedure in the clinical treatment of pain. Its widespread use in patients with all kinds of diseases opens a variety of possible interactions between analgesics used for PCA and other drugs that might be administered concomitantly to the patient. Many of these drug interactions are of little clinical importance. However, some drug interactions have been reported to result in serious clinical problems. Drug interactions can either predominantly affect the pharmacokinetics or pharmacodynamics of the drug. Most important pharmacokinetic drug interactions occur at the level of drug metabolism or protein binding. Acceleration of methadone metabolism caused by cytochrome P450 (CYP) 3A4 induction by antiretroviral drugs or rifampicin (rifampin) has caused methadone withdrawal symptoms. Lack of morphine formation from codeine as a result of CYP2D6 inhibition by quinidine results in an almost complete loss of the analgesic effects of codeine. Alterations of methadone protein binding caused by an inhibition of alpha1-acid glycoprotein synthesis by alkylating substances are another possibility for predominantly pharmacokinetically based drug interactions during PCA. Furthermore, inhibition of P-glycoprotein by anticancer drugs could result in altered transmembrane transport of morphine, methadone or fentanyl, although this has not been shown to be of clinical relevance. Synergistic effects of systemically administered opioids with spinally or topically delivered opioids or anaesthetics have been reported frequently. The same is true for the opioid-sparing effects of coadministered non-opioid analgesics. Antidepressants, anticonvulsants or alpha2-adrenoreceptor agonists have also been shown to exert additive analgesic effects when administered together with an opioid. Inconsistent findings, however, are reported regarding the treatment of patients with opioid-induced nausea and sedation, since coadministration of antiemetics either increased or decreased the respective adverse effects or revealed additional unwanted drug effects.

Steady-state pharmacokinetics of (R)- and (S)-methadone in methadone maintenance patients

AIMS

To investigate the steady-state pharmacokinetics of (R)- and (S)-methadone in a methadone maintenance population.

METHODS

Eighteen patients recruited from a public methadone maintenance program underwent an interdosing interval pharmacokinetic study. Plasma and urine samples were collected and analysed for methadone and its major metabolite (EDDP) using stereoselective h.p.l.c. Methadone plasma protein binding was examined using ultrafiltration, and plasma alpha1-acid glycoprotein concentrations were quantified by radial immunoassay.

RESULTS

(R)-methadone had a significantly (P < 0.05) greater unbound fraction (mean 173%) and total renal clearance (182%) compared with (S)-methadone, while maximum measured plasma concentrations (83%) and apparent partial clearance of methadone to EDDP (76%) were significantly (P < 0.001) lower. When protein binding was considered (R)-methadone plasma clearance of the unbound fraction (59%) and apparent partial intrinsic clearance to EDDP (44%) were significantly (P < 0.01) lower than for (S)-methadone, while AUCtau_¿u¿ss (167%) was significantly (P < 0. 001) greater. There were no significant (P > 0.2) differences between the methadone enantiomers for AUCtauss, steady-state plasma clearance, trough plasma concentrations and unbound renal clearance. Patients excreted significantly (P < 0.0001) more (R)-methadone and (S)-EDDP than the corresponding enantiomers. Considerable interindividual variability was observed for the pharmacokinetic parameters, with coefficients of variation of up to 70%.

CONCLUSIONS

Steady-state pharmacokinetics of unbound methadone are stereoselective, and there is large interindividual variability consistent with CYP3A4 mediated metabolism to the major metabolite EDDP; the variability did not obscure a significant dose-plasma concentration relationship. Stereoselective differences in the pharmacokinetics of methadone may have important implications for pharmacokinetic-pharmacodynamic modelling but is unlikely to be important for therapeutic drug monitoring of methadone, in the setting of opioid dependence.

The genetic variant A of human alpha 1-acid glycoprotein limits the blood to brain transfer of drugs it binds

The objective of this work was to check the effects of alpha-1 acid glycoprotein (AAG) and of its components, A and F1/S genetic variants, on the brain transfer of drugs they bind in plasma. The relevant extractions of six basic drugs, highly bound to AAG, were measured. We chose three drugs selectively bound to the A variant, disopyramide, imipramine and methadone, one drug mainly bound to the mixture F1/S, mifepristone, and two drugs which were simultaneously bound to the variant A and the mixture F1/S, propranolol and chlorpromazine. Their brain extraction were investigated in rats using the carotid injection technique and the capillary depletion method. Injected drugs were dissolved either in buffer, either in native AAG containing the three variants (A, F1 and S), either in variant A or in variant F1/S solutions. Brain extractions of disopyramide, imipramine and methadone were significantly reduced by native AAG and by variant A. Drug's plasma retention was related to their preferential and almost exclusive binding to A variant, both of them exhibiting the same decrease in brain transfer as compared to a buffered solution. At the opposite, there were no significative differences between the extraction either in buffer, either in AAG or in F1/S solutions, of drugs both bound to A variant and F1/S mixture (chlorpromazine and propranolol) or to the F1/S mixture (mifepristone). In serum, the retentional effect of the A variant on the extraction of disopyramide and imipramine was counteracted by the presence of albumin and lipoproteins, which simultaneously bind these two drugs at a high extent and act as permissive binders. We conclude that AAG binding decreases brain drug transfer when the A variant is mainly and almost exclusively involved in the binding. On the contrary, the entire fraction of the tested drugs when bound exclusively or partly to the mixture F1/S is available for transfer into the brain.

Single-step isolation method for six glycoforms of human alpha1-acid glycoprotein by hydroxylapatite chromatography and study of their binding capacities for disopyramide

A single-step isolation method for the glycoforms of human serum alpha1-acid glycoprotein (AAG) using a hydroxylapatite column under a gradient elution program was developed. The concentrations of N-acetylneuraminic acid and monosaccharides (fucose, N-acetylglucosamine, galactose and mannose) of six AAG glycoforms were determined by the pulsedamperometric detection method. For each AAG glycoform, significant sex-related differences in carbohydrate content have been observed only for AAG glycoforms two and six, and not for each AAG glycoform. The relationship between the extent of the branch in the glycan chain and the binding capacity to disopyramide were examined. Female AAG contained highly sialylated AAG glycoforms compared to male glycoforms. Conversely, male AAG was rich in the lower sialylated AAG glycoform. Furthermore, it was found that the drug binding capacity decreases with increasing branching of the glycan chain. This suggests that the binding sites of AAG are hindered by a relatively large carbohydrate moiety, such as tetraantennary structures.

Pharmacodynamic and pharmacokinetic considerations in geriatric psychopharmacology

Age-associated decreases in metabolism and elimination of drugs are sufficient to give one pause before prescribing drugs in the elderly particularly if multiple drugs are to be used. When one factors in concomitant drugs that may inhibit P450-mediated metabolism of other more toxic drugs, genetic deficiency of P450 enzymes, and medical illnesses such as liver and renal failure that will lead to further elevation of drug levels and delay in drug clearance, the likelihood of adverse events when multiple drugs are used in the elderly becomes truly dizzying. In consideration of the pharmacologic principles discussed, when prescribing drugs in the elderly, one might add to the often heard recommendation, "Start low and go slow," a third admonition, "Keep it as simple as possible!"

Abnormal microheterogeneity detected in one commercial alpha 1-acid glycoprotein preparation using chromatography on immobilized metal affinity adsorbent and on hydroxyapatite

The study of one commercial preparation of human alpha 1-acid glycoprotein (AAG) by isoelectric focusing and by different chromatographic methods, previously developed to purify and fractionate the genetic variants of AAG, revealed an abnormal heterogeneity for this preparation. In addition to the three main variants (F1, S and A) of AAG normally present, this preparation contained five other AAG variants (called here sigma, alpha, beta, delta and gamma), accounting for ca. 40% of the total. As it is very unlikely that the latter variants are rare AAG variants, the abnormal heterogeneity of this AAG preparation is most probably due to structural alterations occurring during the large scale isolation. The alpha and the sigma, beta, delta and gamma variants could correspond to altered forms of the A and the F1 and S variants, respectively, because of their similar retention behaviour on immobilized copper(II) ions and their similar drug binding properties. However, the elution of the variants from the immobilized metal affinity column suggested that sigma, alpha, beta, delta and gamma were desialylated. Chromatography on hydroxyapatite enabled the separation of the F1, S and A variants from the sigma, alpha, beta, delta and gamma variants. The inability of the latter variants to bind to hydroxyapatite suggests that the structural alterations might involve acidic amino acid residues. This proposal agreed with the isoelectric focusing study of variants sigma, alpha, beta, delta and gamma. Since the different separation methods used were able to resolve the variants of this AAG, this protocol could be used for characterization of commercial AAG proteins.

Drug-protein binding sites. New trends in analytical and experimental methodology

In the last few years, continuous progress in instrumental analytical methodology has been achieved with a substantial increase in the number of new, more specific and more flexible methods for ligand-protein assays. In general, the methods used for drug-protein binding studies can be divided into two main groups: separation methods (enabling the calculation of binding parameters, i.e. the number of binding sites and their respective affinity constants) and non-separation methods (describing predominantly qualitative parameters of the ligand-protein complex). This review will be focussed particularly on recent trends in the development of drug-protein binding methods including stereoselective and non-stereoselective aspects using chromatography, capillary electrophoresis and microdialysis as compared to the "conventional approach" using equilibrium dialysis, ultrafiltration or size exclusion chromatography. The advantages and limitations of various methods will be discussed including a focus on "optimal" experimental strategies taking into account in vitro, ex vivo and/or in vivo studies. Furthermore, the importance of some particular aspects concerning the drug binding to proteins (covalent binding of drugs and metabolites, stereoselective interactions and evaluation of binding data) will be outlined in more detail.

Drug binding in plasma. A summary of recent trends in the study of drug and hormone binding

The ligands are generally bound in plasma to a significant extent by several transport proteins (both high and low affinity), irrespective of their endogenous or exogenous origin. The protein binding of endogenous compounds (such as hormones) exhibits higher affinity and specificity than those of exogenous compounds (such as drugs). For plasma proteins that bind the same ligand(s), structural similarities or a common genetic origin may be found, although this is not a general rule. Alterations in ligand binding may be due to modifications of either the structure or the level of the binding protein. These modifications may result from genetic make up, physiology or pathology. In some situations, plasma binding may impair the distribution of drugs to tissues, with drug distribution then mainly restricted to the distribution compartment of the drug-binding protein. In other instances, the plasma drug-binding is permissive, and does not limit drug distribution to tissues. A given drug-transport protein system may have either a permissive or a restrictive effect on the drug distribution, depending on the tissue. The physiological significance of the high-affinity transport proteins is not completely understood. These proteins may increase the plasma concentration of poorly hydrosoluble ligands, ensure a more uniform tissue distribution and increase the life of the ligands. The life of the protein may also be increased by ligand binding. High-affinity transport proteins are also involved in some specific carrier mediated transfer mechanisms. It is possible to demonstrate structure-binding relationships or binding selectivity for the plasma transport proteins, but these are quite independent of relationships observed at the receptor level.

Evidence for differences in the binding of drugs to the two main genetic variants of human alpha 1-acid glycoprotein

1. Human alpha 1-acid glycoprotein (AAG), a plasma transport protein, has three main genetic variants. F1. S and A. Native commercial AAG (a mixture of almost equal proportions of these three variants) has been separated by chromatography into variants which correspond to the proteins of the two genes which code for AAG in humans: the A variant and a mixture of the F1 and S variants (60% F1 and 40% S). Their binding properties towards imipramine, warfarin and mifepristone were studied by equilibrium dialysis. 2. The F1S variant mixture strongly bound warfarin and mifepristone with an affinity of 1.89 and 2.06 x 10(6) l mol-1, respectively, but had a low affinity for imipramine. Conversely, the A variant strongly bound imipramine with an affinity of 0.98 x 10(6) l mol-1. The low degree of binding of warfarin and mifepristone to the A variant sample was explained by the presence of protein contaminants in this sample. These results indicate specific drug transport roles for each variant, with respect to its separate genetic origin. 3. Control binding experiments performed with (unfractionated) commercial AAG and with AAG isolated from individuals with either the F1/A or S/A phenotypes, agreed with these findings. The results for the binding of warfarin and mifepristone by the AAG samples were similar to those obtained with the F1S mixture: the mean high-affinity association constant of the AAG samples for each drug was of the same order as that of the F1S mixture: the decrease in the number of binding sites of the AAG samples, as compared with the F1S mixture, was explained by the smaller proportion of variants F1 and/or S in these samples. Conversely, results of the imipramine binding study with the AAG samples concurred with those for the binding of this basic drug by the A variant, with respect to the proportion of the A variant in these samples.

Orosomucoid system: 17 additional orosomucoid variants and proposal for a new nomenclature

There are two forms of orosomucoid (ORM) in the sera of most individuals. They are encoded by two separate but closely linked loci, ORM1 and ORM2. A number of variants have been identified in various populations. Duplication and nonexpression are also observed in some populations. Thus, the ORM system is very complicated and its nomenclature is very confusing. In order to propose a new nomenclature, ORM variants detected by several laboratories have been compared and characterized by isoelectric focusing (IEF) followed by immunoprinting. A total of 57 different alleles including 17 new ones were identified. The 27 alleles were assigned to the ORM1 locus, and the others to the ORM2 locus. The designations ORM*F1, ORM1*F2, ORM1*S and ORM2*M were adopted for the four common alleles instead of ORM1*1, ORM1*3, ORM1*2 and ORM2*1 (ORM2*A), respectively. The variants were designated alpha numerically according to their relative mobilities after IEF in a pH gradient of 4.5-5.4 with Triton X-100 and glycerol. For the duplicated genes a prefix is added to a combined name of two alleles, e.g. ORM1*dB9S. Silent alleles were named ORM1*Q0 and ORM2*Q0 conventionally. In addition, the effects of diseases to ORM band patterns after IEF are also discussed.

pH titration curves of the desialylated human alpha 1-acid glycoprotein variants by combined isoelectrofocusing-electrophoresis: utilization in the development of a fractionation method for the protein variants by chromatography on immobilized metal affinity adsorbent

Using a two-dimensional isoelectrofocusing (IEF)-electrophoresis technique, the pH titration curves of the three main desialylated variants (F1, S and A) of human alpha 1-acid glycoprotein (AAG) were studied to assist in the development of a fractionation method for the AAG variants. For this purpose, different AAG samples, each corresponding to one of the three main phenotypes of the protein (F1S/A, F1/A and S/A), were first purified by chromatographic separation of individual human plasma samples on immobilized Cibacron Blue F3G-A. The purified AAG samples were then disialylated and their heterogeneity was checked by analytical IEF. The pH-mobility curves of the desialylated AAG samples were displayed in polyacrylamide gel slabs, under a constant set of experimental conditions, by carrying out electrophoresis of the protein samples perpendicularly to two stationary pH gradients: a large gradient (pH 3.5-9.5) and a narrow gradient (pH 5-8). The curves showed that all the desialylated variants of AAG exhibited small charge differences and moved closely together between about pH 3.5-5.5 and pH 7.5-9.5. However, the variants were found to show microheterogeneity in their total charge between about pH 5.5 and 7.5 due to the titrated ionizable groups involved along this pH zone. This microheterogeneity was assumed to be accounted for by the existence of differences between the titratable histidyl residues of the AAG variants. Consequently, the interactions of the variants with immobilized transition metal ions were studied at pH 7, using affinity chromatography on an iminodiacetate Sepharose-Cu(II) gel. It was found that the A variant was strongly bound by immobilized Cu(II) ions, whereas the F1 and S variants interacted non-specifically with the immobilized ligand. This finding allowed the development of a rapid and effective fractionation method for desialylated AAG into its A and F1 or S variants, depending on the AAG phenotype, by chromatography on an immobilized affinity Cu(II) adsorbent. The quantitative relationships between immobilized Cu(II) ions and desialylated AAG (the apparent association constant and gel protein-binding capacity) were also determined using a non-chromatographic equilibrium binding technique.

High-performance anion-exchange chromatographic study of desialylated human alpha 1-acid glycoprotein variants. Development of a fractionation method for the protein slow variants

The three main desialylated variants (F1, S and A) of human alpha 1-acid glycoprotein (AAG), a serum acute-phase reactant, were analysed by high-performance anion-exchange chromatography in order to determine their optimum separation conditions. The analysis consisted of three steps, as follows: (1) A desialylated commercial AAG was separated into one "fast"- and one "slow"-migrating fraction by preparative isoelectrofocusing. The "fast" and "slow" fractions were shown to contain the F1 variant and a mixture of the S and A variants, respectively. (2) The pH titration curves of these two fractions were then measured by strong anion-exchange chromatography with several buffer systems of increasing pH. From the data obtained, it was not possible to select the optimum conditions to separate the "fast" variant F1 from the "slow" variants A and S. However, the S and A variants were shown to ionize very differently. (3) The specific fractionation of the S and A variants was therefore carried out by anion-exchange chromatography under operating conditions based on the data obtained from the study of their pH titration curves. This was performed both with the "slow"-migrating fraction obtained by preparative isoelectrofocusing of commercial AAG and with an AAG (containing only variants S and A) purified from an individual serum on immobilized Cibacron Blue F3G-A. Identification of the fractionated proteins was achieved by analytical isoelectrofocusing.

Binding of amitriptyline to alpha 1-acid glycoprotein and its variants

Binding studies have been performed between amitriptyline and i) native alpha 1-acid glycoprotein (AAG); ii) its desialylated form; iii) its two variants, S-AAG and F-AAG; and iv) a mixture of S-AAG and F-AAG. Scatchard analysis revealed the presence of two classes of binding sites on AAG. For native AAG, the first class (of high affinity) has an association constant (Ka1) of 1.5 x 10(6) L mol-1 and a number of binding sites per mole of protein (n1) of 0.25, while the second class (of low affinity) has a Ka2 of 3.2 x 10(4) L mol-1 and a n2 of 0.94. Similar data were found for desialylated AAG. S-AAG and F-AAG do not differ in their association constants measured with amitriptyline, but in their number of binding sites per mole of protein (n): S-AAG: n1 = 0.56, n2 = 0.52; F-AAG: n1 = 0.17, n2 = 0.71. These results confirm those of a previous study, in which a higher affinity of S-AAG towards various basic drugs in comparison with F-AAG has been found.