Effect of albumin conformation on binding of phenylbutazone and oxyphenbutazone to human serum albumin

A critical view on the analysis of fluorescence quenching data for determining ligand-protein binding affinity

The past decade has seen an increase in the number of research papers on ligand binding to proteins based on fluorescence spectroscopy. In most cases, determination of the binding affinity is made by analyzing the quenching of protein fluorescence induced by the ligand. However, many such articles, even those published in reputed journals, suffer from several mistakes with regard to analysis of fluorescence quenching data. Using the binding of phenylbutazone to human serum albumin as a model, we consider some of these mistakes and show how they affect the values of the association constant. In particular, the failure to correct for the inner filter effect and the use of unsuitable equations are discussed. Ligand binding data presented in these articles should be treated with caution, especially in the absence of data from complementary techniques.

Alteration of human serum albumin binding properties induced by modifications: A review

Albumin, a major transporting protein in the blood, is the main target of modification that affects the binding of drugs to Sudlow's site I and II. These modification of serum protein moderates its physiological function, and works as a biomarker of some diseases. The main goal of the paper was to explain the possible alteration of human serum albumin binding properties induced by modifications such as glycation, oxidation and ageing, their origin, methods of evaluation and positive and negative meaning described by significant researchers.

A fluorescent reporter detects details of aromatic ligand interference in drug-binding sites of human serum albumin

Human serum albumin (HSA) transports many ligands including small aromatic molecules: metabolites, drugs etc. Phenylbutazone is an anti-inflammatory drug, which binds to the drug-binding site I of HSA. Its interaction with this site has been studied using a fluorescent dye, CAPIDAN, whose fluorescence in serum originates from HSA and is sensitive to the changes in HSA site I in some diseases. Its fluorescence in HSA solutions is strongly suppressed by phenylbutazone. This phenomenon seems to be a basic sign of a simple drug-dye competition. However, a more detailed study of the time-resolved fluorescence decay of CAPIDAN has shown that phenylbutazone lowers fluorescence without changing the total amount of bound dye. In brief, the HSA-bound dye forms three populations due to three types of environment at the binding sites. The first two populations probably have a rather strong Coulomb interaction with the positive charge of residues Arginine 218 or Arginine 222 in site I and are responsible for approximately 90% of the total fluorescence. Phenylbutazone blocks this interaction and therefore lowers this fluorescence. At the same time the binding of the third population increases considerably in the presence of phenylbutazone, and, as a result, the actual number of bound dye molecules remains almost unchanged despite the ligand competition. So, time resolved fluorescence of the reporter allows to observe details of interactions and interference of aromatic ligands in drug binding site I of HSA both in isolated HSA and in serum.

A spectroscopic study of phenylbutazone and aspirin bound to serum albumin in rheumatoid diseases

Interaction of phenylbutazone (PBZ) and aspirin (ASA), two drugs recommended in rheumatoid diseases (RDs), when binding to human (HSA) and bovine (BSA) serum albumins, has been studied by quenching of fluorescence and proton nuclear magnetic resonance ((1)HNMR) techniques. On the basis of spectrofluorescence measurements high affinity binding sites of PBZ and ASA on albumin as well as their interaction within the binding sites were described. A low affinity binding site has been studied by proton nuclear magnetic resonance spectroscopy. Using fluorescence spectroscopy the location of binding site in serum albumin (SA) for PBZ and ASA was found. Association constants K(a) were determined for binary (i.e. PBZ-SA and ASA-SA) and ternary complexes (i.e. PBZ-[ASA]-SA and ASA-[PBZ]-SA). PBZ and ASA change the affinity of each other to the binding site in serum albumin (SA). The presence of ASA causes the increase of association constants K(aI) of PBZ-SA complex. Similarly, PBZ influences K(aI) of ASA-SA complex. This phenomenon shows that the strength of binding and the stability of the complexes increase in the presence of the second drug. The decrease of K(aII) values suggests that the competition between PBZ and ASA in binding to serum albumin in the second class of binding sites occurs. The analysis of (1)HNMR spectral parameters i.e. changes of chemical shifts and relaxation times of the drug indicate that the presence of ASA weakens the interaction of PBZ with albumin. Similarly PBZ weakens the interaction of ASA with albumin. This conclusion points to the necessity of using a monitoring therapy owning to the possible increase of uncontrolled toxic effects.

Structural and ligand-binding properties of serum albumin species interacting with a biomembrane interface

In the process of drug development, preclinical testing using experimental animals is an important aspect, for verification of the efficacy and safety of a drug. Serum albumin is a major binding protein for endogenous and exogenous ligands and regulates their distribution in various tissues. In this study, the structural and drug-binding properties of albumins on a biomembrane surface were investigated using reverse micelles as a model membrane. In reverse micelles, the secondary structures of all albumins were found, to varying degrees, to be intermediate between the native and denatured states. The tertiary structures of human and bovine albumin were similar to those of the native and intermediate states, respectively, whereas those of the dog, rabbit, and rat were in a denatured state. Thus, bovine albumin is an appropriate model for studying structural changes in human albumin in a membrane-water phase. Binding studies also showed the presence of species difference in the change in binding capacity of albumins during their interaction with reverse micelles. Among the albumins, rat albumin appears to be a good model for the protein-mediated drug uptake of human albumin in a biomembrane environment. These findings are significant in terms of the appropriate extrapolation of pharmacokinetics and pharmacodynamics data in various animals to humans.

Predicting plasma protein binding of drugs: a new approach

In spite of the large amount of plasma protein binding data for drugs, it is not obvious and there is no clear consensus among different disciplines how to deal with this parameter in multidimensional lead optimization strategies. In this work, we have made a comprehensive study on the importance of plasma protein binding and the influencing factors in order to get new insights for this molecular property. Our analysis of the distribution of percentage plasma protein binding among therapeutic drugs showed that no general rules for protein binding can be derived, except for the class of chemotherapeutics, where a clear trend towards lower binding could be observed. For the majority of indication areas, however, empirical rules are missing. We present here an extensive list of multiply determined primary association constants for binding to human serum albumin (HSA) for 138 compounds from the literature. Correlating these binding constants with the percentage fraction of protein bound showed that the percentage data above 90%, corresponding to a binding constant below 6 microM, are of insufficient accuracy. Furthermore, it could be demonstrated that the lipophilicity of drugs, traditionally felt to dominate binding to HSA, is not the only relevant descriptor. Here, we report a generic model for the prediction of drug association constants to HSA, which uses a pharmacophoric similarity concept and partial least square analysis (PLS) to construct a quantitative structure-activity relationship. It is able to single out the submicromolar to nanomolar binders, i.e. to differentiate between 99.0 and 99.99% plasma protein binding. Depending on the system, this can be important in medicinal chemistry programs and may together with other computed physicochemical and ADME properties assist in the prioritization of synthetic strategies.

Interaction of ochratoxin A with human serum albumin. Binding sites localized by competitive interactions with the native protein and its recombinant fragments

Competitive interactions of ochratoxin A (OTA) and several other acidic compounds were utilized to gain insight into the localization of binding sites and the nature of binding interactions between anionic species and human serum albumin (HSA). Depolarization of OTA fluorescence in the presence of a competing anion was used to quantify ligand-protein interactions. The results obtained were rationalized in terms of OTA displacement from its major binding site. Based on their ability to displace OTA, two distinct groups of the anionic ligands were revealed. The first group contained structurally diverse compounds that shared a common binding site in subdomain IIA (Sudlow Site I). The second group consisted of three non-steroidal anti-inflammatory drugs, which showed much lower affinity to Site I than the OTA dianion. The major site for these drugs was located in domain III. Fluorescence spectroscopy measurements of OTA, warfarin (WAR) and naproxen (NAP) complexes with recombinant proteins corresponding to the domains of HSA (D1-D3) revealed binding to all domains but with different affinities. The binding constants for OTA and WAR decreased in the series D2z.Gt;D3>D1. In contrast, NAP showed the most favorable interaction with D3 and comparable affinities to the two remaining domains. The OTA binding constant for D2, 7.9 x 10(5) M(-1), was smaller than the largest constant for HSA by a factor of approximately 7. The binding constant for OTA with D3, 1.1 x 10(5) M(-1), was very close to that of the secondary binding site for HSA.

Articular diffusion of meloxicam after a single oral dose: relationship to cyclo-oxygenase inhibition in synovial cells

OBJECTIVE

To investigate the distribution of meloxicam in the human knee joint and to compare it with the inhibition of cyclo-oxygenase (COX) activity in synovial cells.

DESIGN

Prospective pharmacokinetic study and in vitro laboratory investigation.

PATIENTS AND PARTICIPANTS

42 male and female patients aged 26 to 85 years hospitalised for rheumatic disease and requiring a diagnostic and/or therapeutic knee puncture.

METHODS

After a single oral dose of meloxicam 15mg, synovial fluid and blood samples were collected once per patient at various intervals after administration. Meloxicam concentrations were determined by a validated high performance liquid chromatography assay, protein binding by equilibrium dialysis, and pharmacokinetic parameters were calculated by noncompartmental analysis from the mean drug concentration-time profiles. The inhibitory effect of meloxicam on COX activity was investigated separately in unstimulated or interleukin-1beta-stimulated human synovial cells from osteoarthritic patients.

RESULTS

Meloxicam was found in synovial fluid at the earliest sampling time (1 hour). Peak concentrations were reached approximately 6 hours postdose in both plasma (842 microg/L) and synovial fluid (320 microg/L). A plateau was observed after the distribution phase (6 hours), corresponding to a constant ratio of drug concentration between synovial fluid and plasma of about 0.47. This ratio was higher in patients with acute inflammation (0.58) than in those with no inflammation (0.38). Meloxicam was extensively bound to protein, mainly to serum albumin. The area under the drug concentration-time curve (AUC) in plasma was more than 2.5 times that in synovial fluid. The AUC for free meloxicam was similar in plasma and synovial fluid. The 50% inhibitory concentrations (IC50) for basal and stimulated COX activity in human synovial cells were 33.7 nmol/L (11.8 microg/L) and 2.0 nmol/L (0.70 microg/L), respectively. The free concentration of meloxicam in synovial fluid was higher than the IC50 for stimulated COX activity from 6 to 36 hours postdose.

CONCLUSION

On the basis of free synovial concentrations and the IC50 for stimulated COX activity, meloxicam is expected to have a long duration of action. Inhibition of COX activity is expected to be more marked in inflamed synovium compared with non-inflamed synovium.

Using difference spectrophotometry to study the influence of different ions and buffer systems on drug protein binding

Difference spectrophotometry was used to investigate the effect of different ions and buffer systems on the binding of the anti-inflammatory drug tenoxicam to human serum albumin (HSA). Chloride anions, as well as sodium cations, were found to decrease the binding affinity. The effect of chloride ions was greater on the primary binding constant K1, while sodium ions had a greater effect on the secondary binding constant K2. The number of binding sites n1 and n2 were not affected except at 0.12% HSA, for which the presence of sodium ions halved n2. Potassium ions significantly increased K1. The presence of potassium instead of sodium ions increased binding affinity at lower HSA concentrations. The number of binding sites n1 and n2 were fewer in presence of potassium than in the presence of sodium ions except at 0.12% HSA. The divalent calcium and magnesium cations increased the binding affinity of HSA to tenoxicam, with a greater effect on K1. The effect of magnesium ions on K1 occurred when the MgCl2 concentration was increased to 3 and 9 mM, with the former seeming to be a critical concentration. The number of primary binding sites n1 was not affected by calcium ions, but was halved by 1 mM MgCl2. Both calcium and magnesium cations decreased n2, which was halved when the concentration of either cation was increased to 9 mM. The effect of buffer systems on tenoxicam binding to HSA was dependent on HSA concentration. The value of K1 was higher in Sorensen's phosphate buffer than in Tris [tris(hydroxymethyl) aminomethane HCl] buffer when the HSA concentration was 0.04% and 0.16%, while the reverse was observed at 0.08% and 0.12% HSA. The other binding parameters (K2, n1, and n2) were higher in Sorensen's phosphate than in Tris buffer. However; at certain HSA concentrations, the values of such parameters were comparable in both buffer systems.

Species differences of serum albumins: III. Analysis of structural characteristics and ligand binding properties during N-B transitions

PURPOSE

The aim of this study was to investigate the characteristics of the structural transitions and changes in ligand binding properties of different albumins during the pH-dependent structural transition, often referred to as the N-B transition.

METHODS

Structural transitions were evaluated by means of spectrometry, differential scanning calorimetry and chemical modification. In addition, ligand binding properties were investigated using typical site-specific bound drugs (warfarin, phenylbutazone, ibuprofen and diazepam).

RESULTS

Conformational changes, including N-B transition, clearly occurred in albumins from all species used in this study. The conformational stabilities of all the albumins were clearly lost in the weakly alkaline pH range. This was probably the result of the destruction of salt bridges between domain I and domain III in the albumin molecule. In addition, the profiles of the ANS-induced fluorescence were different and could be classified into two patterns, suggesting that hydrophobic pockets in the albumin molecules were different for the different species. The data suggest that the amino acid residues responsible for the transitions were some of the His residues located in domain I. Further, the ligand binding properties of the albumins were slightly different but statistically significant.

CONCLUSIONS

The overall mechanisms of the N-B transition may be similar for all the albumins, but its impact is considerably different among the species in terms of both structural characteristics and ligand binding properties. Furthermore, the transitions appear to be multi-step transitions.

Determination of binding conformations of drugs to human serum albumin by transferred nuclear overhauser effect measurements and conformational analyses using high-temperature molecular dynamics calculations

The binding conformations of oxyphenbutazone (OXY), Nepsilon-dansyl-L-lysine (DNS-LYS), and furosemide (FU) to human serum albumin (HSA) have been investigated by molecular dynamics (MD) calculations and transferred nuclear Overhauser effect (TRNOE) measurements. We have combined distance information obtained from the Conformational Analyzer with Molecular Dynamics And Sampling (CAMDAS) calculation and experimental NOE spectroscopy measurements to determine a "binding conformation" for each drug which binds to site I of HSA. For OXY, only one conformer (conf9) among the conformer set generated by MD calculation satisfied the distance restraint conditions obtained from TRNOE measurements. For DNS-LYS and FU, 17 and 5 conformers satisfied distance restraint conditions, respectively. The structure of conf9 of OXY was taken as a "template" to choose binding conformers for DNS-LYS and FU. By fitting the "template" to the 17 conformers of DNS-LYS and 5 conformers of FU, we could efficiently obtain one binding conformer for DNS-LYS (conf144) and FU (conf26). It is suggested from the feature of the binding conformation that the three-dimensional location of a hydrophobic aromatic ring, alkyl chain, and electronegative functional group is important for binding to site I of HSA. This method, which combines MD calculations and NOE information, is thought to be effective for determining the binding conformation of drugs to HSA.

Stereospecific and competitive binding of drugs to human serum albumin: a difference circular dichroism approach

Simultaneous binding of two drugs to human serum albumin (HSA) was investigated by difference circular dichroism (delta CD) spectroscopy. Phenylbutazone and diazepam were chosen as specific markers for binding areas I (cumarines) and II (indoles), respectively, and their stereospecific interactions with protein were selectively characterized. Displacers were drugs known to specifically bind to areas I (salicylate) and II (racemic ibuprofen). The results indicate two different interaction mechanisms: a direct competition one (diazepam-ibuprofen and phenylbutazone-salicylate) and an indirect competition one (diazepam-salicylate and phenylbutazone-ibuprofen). The two major binding areas on HSA are distinct, but not independent, entities. Finally, the dissociation constants of marker ligands and competitors complexed to HSA were determined by quantitative analysis of CD data.

Conformational effects in the interaction of phenylbutazone with albumin studied by circular dichroism

The binding of phenylbutazone (PB) to human serum albumin (HSA) at different pH and in the presence of different NaSCN and urea concentrations that alter the conformation of the protein was examined qualitatively on the basis of extrinsic elliptical strength at 288 nm by means of circular dichroism (CD). The values of the binding index expressed as a ratio of [theta]max/[theta]pH7.4(288) at each extrinsic rotational strength in the presence of various concentrations of NaSCN, urea and hydrogen ion were directly proportional to the alpha-helix content based on the peptide backbone alteration of HSA by NaSCN, urea and hydrogen ion except for the pH range of 5.0 to 10.0. The values in the pH range of 7.4 to 10.0 depended on the concentration of hydrogen ion and not on the alpha-helix content, showing a significant effect of the hydrogen ion on the tertiary conformation with respect to the binding sites of the amino acid chain rather than the peptide backbone of HSA. The increases in the binding index observed in the pH range of 7.4 to 10.0 were not observed at all in the case of NaSCN and urea at the concentrations studied. It was demonstrated that the binding of PB to HSA increased with the change in the tertiary conformation caused by hydrogen ions but decreased with that in the secondary conformation caused by a concentration change of NaSCN and urea. Thus, the binding was closely associated with skeletal conformational alterations as well as changes in the binding sites of the amino acid chains of the protein.

Ligand binding at membrane mimetic interfaces. Human serum albumin in reverse micelles

The behaviour of human serum albumin in the presence of three chemically distinct ligands: oxyphenylbutazone, dansylsarcosine and hemin, has been compared in buffer and in reverse micelles of isooctane, water, and either sodium bis(2-ethylhexyl)sulfosuccinate or hexadecyl trimethylammonium bromide, systems selected to mimic the membrane-water interface. Upon micellar incorporation, the dansylsarcosine-albumin complex dissociated, as evidenced by fluorescence emission spectroscopy (red shift from 485 nm to 570 nm) and by fluorescence polarization measurements. In contrast, the hemin-albumin complex remained stable in reverse micelles, as judged from the Soret absorption band at 408 nm and the molar absorption coefficient of 8.4 x 10(4) M-1 cm-1. The oxyphenylbutazone to albumin binding curves reveal that while the association constant remained unchanged (Ka approximately 1.0 x 10(5) M-1), only a fraction of the albumin molecules present reacted with the ligand. The results were unaffected by the nature and the concentration of the surfactant. These findings can be interpreted in the light of conformational changes induced in human serum albumin by the large micellar inner surface area. The blue shift of the fluorescence emission maximum from 344 nm in buffer to 327 nm in sodium bis(2-ethylhexyl)sulfosuccinate micelles and the lesser reactivity/accessibility of the fluorophore to oxidation by N-bromosuccinimide, indicate perturbations of the sole tryptophan-214 microenvironment. However, the distance between the indole residue and tyrosine-411 does not seem substantially modified by the 15% decrease affecting the alpha helices of the albumin molecule. It is proposed that the results reported herein reflect the interactions of albumin with a membrane-like interface which generates two protein subpopulations differing in their membrane-surface and ligand affinities. Overall and local conformational changes, originating from this surface-induced effect, may thus constitute a ligand-release facilitating mechanism acting at cellular membrane levels.

Separation procedures used to reveal and follow drug-protein binding

The review gives a critical evaluation of the different separation procedures used to study drug-protein interactions and describes their various fields of application. For pharmacological studies, the most widely used methods are dialysis and ultrafiltration, because they allow measurements with solutions of high protein concentrations, such as those found in therapeutic conditions. Both techniques use membrane devices, which may induce additional binding effects. Another drawback of these techniques is the need for radiolabelled compounds. Chromatographic methods, which now take advantage of the technology of high-performance liquid chromatography, are generally faster and do not use drug labelling because of the higher sensitivities of the detectors. Two different approaches are possible: either all the interacting species (protein and drug) are dissolved in the mobile phase, or one of them (protein or drug) is immobilized on the support. Several chromatographic methods are available for studies in solution that differ according to the sample injection mode (frontal or zonal elution) and the nature of the mobile phase used. They include quantitation of the drug-protein complex by zonal elution, the Hummel and Dreyer method, frontal elution, the vacancy peak method, and retention analysis by zonal elution. Frontal elution is the most rigorous method since all the species at equilibrium are present in the mobile phase with known and constant concentrations. The most promising one is the Hummel and Dreyer method, because of the very small amount of protein injected in the mobile phase containing the drug. Drug-protein interactions may be studied by affinity chromatography by immobilizing one of the interacting species on the support. Comparison of the constants obtained with methods when both the drug and the protein are in solution is questionable, since the immobilized species in affinity separations differ in their physical properties from those in solution. The main advantage with studies on immobilized proteins is the easy comparison of the binding properties of various drugs, especially when they are enantiomeric. The results of the binding constants measured by different separation methods are given for the albumin-phenylbutazone and albumin-warfarin systems. Good agreement is generally obtained, which proves the validity of using chromatography as a tool to study drug-protein interactions.

Structural requirements for binding of nonsteroidal anti-inflammatory drugs to human serum albumin

The binding of representative chemical classes of nonsteroidal anti-inflammatory drugs (NSAIDs) to human serum albumin (HSA) was investigated by equilibrium dialysis. Warfarin enantiomers were used as specific markers in displacement studies. Data were analyzed by a computerized nonlinear least squares approach designed for binding of small ligands to macromolecules at equilibrium. The binding data indicated comparable affinities to the primary site by the warfarin enantiomers, phenylbutazone, and meclofenamate sodium. Naproxen, sulindac, and zomepirac showed lower affinity by one order of magnitude. The displacement data revealed stereoselectivity. The R(+) isomer was displaced to a significantly greater extent than the S(-) isomer by meclofenamate sodium, while the reverse was observed for phenylbutazone. Naproxen displaced both isomers to the same extent. No significant displacement of either isomer was seen with sulindac or zomepirac. Examination of the chemical structures of the high affinity compounds indicated the common feature of a hydrophobic area bearing a widely delocalized negative charge. Hydrophobic binding of these compounds to HSA at the warfarin site is possibly stabilized by the attraction of the delocalized negative charge to the basic lysine and arginine residues adjoining the lone tryptophan.

Evidence for the fatty acid-induced heterogeneity of the N and B conformations of human serum albumin

The influence of oleic acid on the interaction between albumin and warfarin, oxyphenbutazone or diazepam has been studied by circular dichroism and equilibrium dialysis. The pH dependences of the molar ellipticity of the drug-albumin complexes and of the free fraction of drug are completely changed by the presence of oleic acid. This phenomenon is attributed to an oleic acid-induced conformational change in both the neutral (N) and the basic (B) conformation of albumin, a change to which the warfarin-oxyphenbutazone binding area and the diazepam binding site is sensitive. The oleic acid-induced conformational states of albumin, the so-called N* and B* conformations, show binding properties that are different from the binding properties of the N and B conformations.