The effect of chloride on the binding of warfarin to albumin as a function of pH

NMR metabolomics profiling of blood plasma mimics shows that medium- and long-chain fatty acids differently release metabolites from human serum albumin

Metabolite profiling by NMR of body fluids is increasingly used to successfully differentiate patients from healthy individuals. Metabolites and their concentrations are direct reporters of body biochemistry. However, in blood plasma the NMR-detected free-metabolite concentrations are also strongly affected by interactions with the abundant plasma proteins, which have as of yet not been considered much in metabolic profiling. We previously reported that many of the common NMR-detected metabolites in blood plasma bind to human serum albumin (HSA) and many are released by fatty acids present in fatted HSA. HSA is the most abundant plasma protein and main transporter of endogenous and exogenous metabolites. Here, we show by NMR how the two most common fatty acids (FAs) in blood plasma - the long-chain FA, stearate (C18:0) and medium-chain FA, myristate (C14:0) - affect metabolite-HSA interaction. Of the set of 18 common NMR-detected metabolites, many are released by stearate and/or myristate, lactate appearing the most strongly affected. Myristate, but not stearate, reduces HSA-binding of phenylalanine and pyruvate. Citrate signals were NMR invisible in the presence of HSA. Only at high myristate-HSA mole ratios 11:1, is citrate sufficiently released to be detected. Finally, we find that limited dilution of blood-plasma mimics releases HSA-bound metabolites, a finding confirmed in real blood plasma samples. Based on these findings, we provide recommendations for NMR experiments for quantitative metabolite profiling.

Interaction of Non-Steroidal Anti-Inflammatory Drugs, Etoricoxib and Parecoxib Sodium, with Human Serum Albumin Studied by Fluorescence Spectroscopy

The mechanism of interaction of the non-steroidal anti-inflammatory drugs, etoricoxib and parecoxib sodium, with human serum albumin (HSA) was studied using fluorescence spectroscopy. There was only one class of binding site with association constants of the order of 10(4). Thermodynamic parameters suggest that van der Waals and hydrogen bonding interactions in the case of etoricoxib, and electrostatic and hydrogen bonding interactions in the case of parecoxib sodium, are predominantly involved in the binding. Studies in the presence of the hydrophobic probe, 1-anilinonaphthalene-8-sulfonate (ANS), showed that hydrophobic interactions are not involved in the binding of these drugs to HSA. Displacement studies using the site-specific probe, dansylsarcosine piperidinium salt (DSS), showed that the drugs are bound at site II on the HSA molecule. However, etoricoxib and parecoxib sodium are bound at different regions within site II. Increase of pH and the presence of salt caused significant changes in the association constants and the concentration of free pharmacologically active drug. Stern-Volmer analysis of the binding data indicated that the tryptophan residues of albumin are not fully accessible to anionic parecoxib sodium and a predominantly static quenching mechanism is operative in each case.

Mechanism of interaction of the non-steroidal antiinflammatory drugs meloxicam and nimesulide with serum albumin

The mechanism of interaction of the non-steroidal antiinflammatory drugs meloxicam and nimesulide with human and bovine serum albumin has been studied using fluorescence spectroscopy. There was only one high affinity site on serum albumin for both the drugs with association constants of the order of 10(5). Negative enthalpy (DeltaH(0)) and positive entropy (DeltaS(0)) values in the case of both meloxicam and nimesulide showed that both hydrogen bonding and hydrophobic interactions play a role in the binding of these drugs. Binding studies in the presence of the hydrophobic probe 1-anilinonaphthalene-8-sulfonate (ANS) showed that the binding of meloxicam and nimesulide to serum albumin involves predominantly hydrophobic interactions. Stern-Volmer analysis of the quenching data showed that quenching is highly efficient and that the tryptophan residues in hydrophobic regions of the proteins are fully exposed to the drugs. Thus these drugs are bound to albumin by hydrophobic interactions as well as hydrogen bonding at a site, which is close to the tryptophan residues. An increase of the pH and ionic strength caused an increase in the concentration of free drug, although the effect was not very significant.

Mechanism of binding of warfarin enantiomers to recombinant domains of human albumin

Domain fragments of human serum albumin corresponding to domains 1 and 2 (D12) and domains 2 and 3 (D23) were expressed in yeast. The kinetics of warfarin binding to these fragments were investigated using stopped-flow fluorescence spectroscopy. Binding can be characterized by a two-step process, a rapid diffusion-controlled step and a slower rate-limiting step in which a stable drug-protein complex is formed. The equilibrium constant for step 1 is greater for both D12 and D23 than for albumin, probably as a result of reduced steric hindrance offered by the domain fragments. Binding step 2, thought to be the result of a conformational change as warfarin is accommodated by the protein, is faster for D12 and D23. Albumin and the domain fragments show an increased preference for the R enantiomer, but the preference is particularly enhanced for domain fragment D12. These preferences can largely be explained by the domains having different rates for step 2 of the binding process.

Practical aspects of the ligand-binding and enzymatic properties of human serum albumin

Recent work with approaches like recombinant mutants and X-ray crystallography has given much new information about the ligand-binding properties of human serum albumin (HSA). The information increases the understanding of this unique transport and depot protein and could give a structural basis for the possible construction of therapeutic agents with altered HSA-binding properties. A tabulation of high-affinity binding sites for both endogenous and exogenous compounds has been made; it could be useful for the above-mentioned purpose, but it could also be of value when trying to predict potential drug interactions at the protein-binding level. Drug displacement is not always a complication to therapy; it can be used to increase the biological effect of a drug. However, due to rebinding at other sites, the increase in the free concentration of a displaced ligand can be less than expected. Drugs and drug metabolites can also interact covalently with HSA; thiol-containing drugs often bind to the single free cysteine residue of HSA, and glucuronidated drugs react irreversibly with other residues of the protein. Reversible binding of ligands is often stereospecific, and therefore immobilized HSA can be used to separate drug isomers. Albumin-containing dialysates are useful for extracorporeal removal of endogenous toxins and in the treatment of drug overdoses. HSA has different types of hydrolytic activities, which also can be stereospecific. The esterase-like property seems especially useful in converting prodrugs to active drugs in plasma.

High prostaglandin-E1 binding to serum protein in allergic subjects

Protein binding to both salicylate and 3H-labelled prostaglandin-E1 ([3H]PGE1) was examined in the sera of 22 allergic and 16 normal individuals. Protein binding to salicylate (P less than 0.001) and to [3H]PGE1 (P less than 0.01) was significantly greater in the allergic than in the normal group. The nature of the binding sites of salicylate and PGE1 was investigated with two fluorescent probes, dansylamide and dansylsarcosine as specific marker ligands for established Sites I and II, found to be specific for anionic drugs. The serum protein of 11 allergic subjects showed a higher binding at Site I (P less than 0.05) and a lower binding at Site II (P less than 0.05) than that of seven normal subjects. Salicylate and [3H]PGE1 bound competitively at the two sites. It was concluded, when comparing allergic to normal subjects, that the high protein binding of allergic individuals to salicylate and PGE1 could be attributed to qualitative and/or quantitative differences in the lipophilic substances which are tightly bound to the albumin of normal sera, causing a reduction in binding ability at Site I.

Location and characterization of the suramin binding sites of human serum albumin

The objective of the present study was to investigate the location of the high-affinity suramin binding sites on the human serum albumin molecule. For this purpose, circular dichroism and equilibrium dialysis experiments were performed on the interaction between suramin and a large peptic and a large tryptic fragment of albumin, the former comprising domains one and two of the albumin structure and the latter domains two and three. The equilibrium dialysis experiments revealed that albumin and the fragments have a comparable total affinity for suramin. Furthermore, all three proteins display a similar pH dependence of the unbound fraction of suramin. The circular dichroism experiments revealed that only the suramin-albumin and the suramin-peptic fragment complexes can undergo the pH dependent neutral-to-base or N-B conformational change, whereas the suramin-tryptic fragment complex lacks this ability. It is likely that the main parts of the high-affinity binding sites for suramin are located in domain two of the albumin molecule. The nature of these binding sites is discussed. The deprotonation of histidine and other positively charged residues taking part in salt bridges between suramin and albumin is, in all probability, the main cause of the decrease in affinity of suramin for albumin as the pH is raised from 6 to 9.

Location and characterization of the warfarin binding site of human serum albumin. A comparative study of two large fragments

The warfarin binding behaviour of a large tryptic fragment (residues 198-585 which comprise domains two and three) and of a large peptic fragment (residues 1-387 which comprise domains one and two) of human serum albumin has been studied by circular dichroism and equilibrium dialysis in order to locate and characterize the primary warfarin binding site. The induced ellipticity of the warfarin-peptic fragment complex turned out to be pH dependent. This pH dependence occurs in the region of the neutral-to-base transition of the albumin molecule. The induced ellipticity of the warfarin-tryptic fragment complex is pH independent. Difference CD-spectra showed that the peptic fragment and albumin have similar warfarin binding properties whereas the tryptic fragment has deviant warfarin binding properties. The equilibrium dialysis experiments showed that the affinity of warfarin to the peptic fragment and to albumin is practically the same whereas the affinity of warfarin to the tryptic fragment is a factor 2-8 lower than the affinity of warfarin to albumin. Our results indicate that the main part of the primary warfarin binding site is located in domain two of the albumin structure and that domain one plays an important role in the N-B transition of albumin.

Drug-binding and other physicochemical properties of a large tryptic and a large peptic fragment of human serum albumin

The diazepam-binding behaviour of a large tryptic and a large peptic fragment of human serum albumin has been studied by circular dichroism and equilibrium dialysis in order to locate the primary diazepam-binding site on the albumin molecule. The analytical set-up of the FPLC was used to find the optimum experimental conditions for isolating the fragments. Conventional columns with a 100-fold higher loading capacity than the analytical FPLC columns were used to isolate large amounts of the fragments. The isolation procedure for the tryptic fragment (45 kDa, domains two and three of the albumin structure) is described in this paper. The description of the isolation procedure for the peptic fragment (46 kDa, domains one and two of the albumin structure) is published elsewhere (Bos, O.J.M., Fischer, M.J.E., Wilting, J. and Janssen, L.H.M. (1988) J. Chromatogr. 424, 13-21). The induced ellipticity of the diazepam-fragment complexes as well as the affinity of diazepam to the fragments turned out to be pH dependent. This pH dependence occurs in the region of the neutral to base transition of the albumin molecule. Difference CD-spectra of the proteins showed that the tryptic fragment and albumin have similar diazepam-binding properties, whereas the peptic fragment has different diazepam-binding properties. This result is in line with our equilibrium dialysis experiments which showed that the affinity of diazepam to the tryptic fragment and to albumin is of the same order of magnitude, whereas the affinity of diazepam to the peptic fragment is several orders of magnitude lower. On the basis of these results, it can be concluded that the tryptic fragment contains the primary diazepam-binding site and the peptic fragment one or more secondary diazepam-binding sites. This means that at least the main part of the primary diazepam-binding site is located in domain three of the albumin structure.

The fatty-acid-induced conformational states of human serum albumin investigated by means of multiple co-binding of protons and oleic acid

The binding of oleic acid to human serum albumin causes progressive changes in (a) the pK of some amino acid residues, as detected by pH-stat titration and (b) the induced molar ellipticities of albumin-bound drugs (diazepam and oxyphenbutazone), as measured by c.d. It is concluded that albumin undergoes several conformational transitions as the amount of oleic acid bound increases from 0 to about 9 molecules/molecule of protein. At least three different conformations of the protein seem to be involved. These conformations can be linked with the three classes of oleic acid-binding sites on albumin.

Procedure for the isolation of a large peptic fragment of human serum albumin

For a thorough investigation of the drug-binding behaviour and other physicochemical properties of human serum albumin, one needs large amounts of specific fragments of albumin. Such fragments were obtained by careful proteolysis of the native protein with pepsin at pH 3.70. The fast protein liquid chromatographic technique was used to find the optimum experimental conditions for the separation of the fragments. By means of anion-exchange chromatography, chromatofocusing and gel permeation, it was possible to obtain a large fragment with a relative molecular mass of 46,000. The fragment could be assigned to segment 1-387 and therefore consists of domains 1 and 2 of the albumin structure. A 1-g amount of albumin produced 50 mg of a fragment that was 98% homogeneous.

Application of fluorine-19 nuclear magnetic resonance to the determination of plasma-protein binding of 5′-deoxy-5-fluorouridine, a new antineoplastic fluoropyrimidine

Two distinct fluorine-19 nuclear magnetic resonance ((19)F NMR) signals have been observed in human serum for free and plasma-protein bound 5'-deoxy-5-fluorouridine (5'dFUrd). The binding of this drug was studied directly in serum using (19)F NMR. To evaluate the validity of this method, a parallel study was conducted with equilibrium dialysis as the reference method. Two assay methods were applied after equilibrium dialysis, UV spectrophotometry and (19)F NMR spectrometry, the UV assay being used to validate the (19)F NMR assay. A study of the binding of 5'dFUrd to human serum albumin was also reported. The reliability of (19)F NMR as a technique to measure directly the binding of the drug and as an assay after equilibrium dialysis was demonstrated. The percentage of 5'dFUrd bound to plasma proteins is low and concentration-dependent in the 0.04-3.5 mmol l(-1) range.

A study on the allosteric interaction between the major binding sites of human serum albumin using microcalorimetry

The binding of warfarin and oxyphenbutazone to albumin has been studied at pH 6.8 and pH 9.2 by measuring the heat of binding of these ligands to their high-affinity binding sites on albumin (delta Ho'1). The -delta Ho'1 values for the binding of warfarin at pH 6.8 and 9.2 and oxyphenbutazone at pH 6.8 and 9.2 were found to be 16.9(+/- 0.6), 28.8(+/- 0.6), 10.5(+/- 0.4) and 17.4(+/- 0.6) kJmol-1, respectively. The Gibbs energies (delta Go'1) corresponding to these delta Ho'1 values cover a much smaller range. The pH dependences of delta Go'1 and delta Ho'1 are explained in terms of pK shifts in the albumin upon binding warfarin or oxyphenbutazone. Diazepam, which binds to a site on albumin which is different from the warfarin-oxyphenbutazone binding site, increases - delta Ho'1 for the binding of warfarin and oxyphenbutazone to albumin at pH 6.8, but it does not influence the -delta Ho'1 at pH 9.2. This phenomenon may be attributed to an allosteric interaction between the diazepam binding site and the warfarin binding site. This allosteric interaction must have its origin in a phenomenon other than the N-B transition.

Inhibition of salicylate binding to normal plasma by extracts of uremic fluids

We previously reported that an extract of uremic plasma reduces binding of phenytoin and tryptophan by normal plasma and plasma albumin. This effect appears to reproduce the impaired binding of many drugs and several endogenous metabolites by uremic plasma. In the present study we further characterized the properties of extracts from uremic sera and body fluids using binding of salicylate as a model. Salicylate was chosen because it binds to both of the main albumin binding loci for aromatic, acidic drugs. Using a computer-assisted, least-squares, curve-fitting program, LIGAND, we found that the most satisfactory model for salicylate binding to 1:10 diluted normal plasma was a binding number (n) of 2 mol of salicylate per mole of albumin with an association constant (k) of 2.85 X 10(4) L/mol, an additional binding of 0.5 mol to other sites on albumin or to other proteins, and nonspecific binding of 21%. Addition of uremic pleural fluid extract to diluted normal plasma produced a monotonic decline in k to 0.17 X 10(4) L/mol with no change in n except possibly at the highest dose of uremic inhibitor. This pattern of competitive inhibition indicates presence of unknown ligands in the uremic extract that compete at both binding loci. More efficient extraction methods might also yield additional ligand(s) that inhibit through a noncompetitive mechanism.

Pethidine binding in plasma: effects of methodological variables

Several methodological variables potentially influencing the plasma protein binding of [14C]-pethidine in vitro were investigated using equilibrium dialysis and rigorous pH control. Ionic strength of buffer, pethidine concentration and 21 days of plasma samples frozen to -8 degrees C did not affect the outcome of the binding experiments. Unbound fraction decreased with increasing temperatures between 25 degrees C and 37 degrees C. Unbound fraction decreased with increased pH between pH 7.0 and 8.0; the mean unbound fraction at an equilibrium pH of 7.4 and at 37 degrees C was 0.58 (s.d. 0.03, n = 58). It is likely that previous reports of pethidine unbound fraction as being between 0.2 and 0.4 represent artefacts caused by inadequate pH control during dialysis.

Allosteric properties of the oxyphenbutazone--human serum albumin complex

The conformational change of albumin which occurs around physiological pH, the so-called N-B transition, has been studied by measuring the induced circular dichroic signal of the oxyphenbutazone-albumin complex. This N-B transition has been characterized by a set of parameters according to the two-state model of Monod, Wyman and Changeux. The influence of calcium ions on the N-B transition has been interpreted in terms of a change in some of the parameters describing the two-state model, viz. a decrease of the apparent pK value of the histidines and of the apparent allosteric constant of the oxyphenbutazone-albumin complex. This apparent pK change increases with increasing Ca2+ concentration, whereas the apparent allosteric constant approaches a final value at 5 mM Ca2+. From acid-base titration curves of albumin in the presence and in the absence of Ca2+ it could be concluded that in the presence of Ca2+ less histidines are titratable than in the absence of Ca2+. Assuming that these histidines are not involved in the N-B transition it follows that at least four to five histidines are involved in the N-B transition.

Consequences of the N-B transition of albumin for the binding of warfarin in human serum

The protein binding of warfarin in serum has been studied by means of circular dichroism and equilibrium dialysis. Evidence was found that the N-B transition of albumin, occurring around physiological pH, takes place not only in solutions of pure albumin but also in serum. The protein binding of warfarin in serum is pH-dependent and increases with pH especially around physiological pH. This pH-dependent serum binding of warfarin can be reasonably explained by the N-B transition of albumin. The effect of Ca2+ and Mg2+ on the protein binding of warfarin in serum is negligible at pH 7.4, whereas at this pH Cl- increases the free-warfarin concentration by a competitive displacement.

The effects of N-B transition of human serum albumin on the specific drug-binding sites

In the pH range 6-9, human serum albumin undergoes a conformational change termed the neutral-base (N-B) transition. Recently, it has been shown that the N-B transition causes enhanced binding at the warfarin-binding site (site I). The present study used fluorescence and equilibrium dialysis to investigate the effects of the N-B transition, chloride, calcium and fatty acids on the specific binding sites I and II on human serum albumin. The effect of the N-B transition of human serum albumin provides a further distinction between site I and II binding characteristics. The N-B transition of albumin caused a change in conformation at site I which resulted in increased binding of drugs and fluorescent probes at this site, whereas there was no effect on acidic drug binding at site II. These effects on site I and II are qualitatively similar to those induced by fatty acids (increased drug binding at site I and no change at site II). However, the effects of increasing pH and fatty acids were additive, showing that they were caused by two different conformational changes. The effect of Cl- on site I binding was pH-dependent and was abolished by the presence of fatty acid. Ca2+ reduced the fluorescence of site I probes but had no effect on a site II fluorescent probe. Effects of pH were also investigated with drugs not binding to site I or II. Increasing pH caused a decrease in binding to indomethacin, increases in binding of L-tryptophan, tolmetin and quinidine and no change in the binding of salicylic acid, diflunisal and phenytoin.

A comparative study of some physico-chemical properties of human serum albumin samples from different sources--II. The characteristics of the N-B transition and the binding behaviour with regard to warfarin and diazepam

A comparative study of the N-B transition and the drug binding properties of human serum albumin samples from various sources has been carried out with the help of circular dichroism and equilibrium dialysis. It was found that when warfarin was used as a marker the midpoint pH and the cooperative nature of the N-B transition that occurs in the albumin around physiological pH varied with the albumin sample. The midpoint pH was found to be related to the cooperative nature of the albumin samples. A similar relationship has been found for allosteric proteins. However, when diazepam was used as a marker molecule for the N-B transition, variations in the midpoint pH and cooperative nature of the N-B transition disappeared. This is attributed to the strong allosteric effect of diazepam on binding. The affinity of warfarin for albumin depends strongly on the sample, but this is not the case with diazepam. The cooperative binding properties found for the albumin samples are compared with those found for the albumin in serum. After a discussion it is concluded that the cooperative binding properties of a particular albumin sample should be taken into account when that sample is used in binding studies.

A comparative study of some physico-chemical properties of human serum albumin samples from different sources--I. Some physico-chemical properties of isoionic human serum albumin solutions

Human serum albumin samples from different sources were investigated. The fatty acid content of the albumin before and after deionization on a mixed bed ion-exchange column varied from sample to sample. When an albumin sample from one source was deionized under standard conditions the amount of fatty acid bound by the albumin was reduced to a reproducible amount. In samples from different sources, however, the amount bound varied considerably. Also the isoionic pH of the albumin varied from sample to sample. This variation could be attributed to the difference in the fatty acid content and the different number of titratable histidines and acid amino-acid residues in the albumin from different sources. It can be concluded from the specific conductance of these isoionic solutions that ions other than H-, OH and protein are effectively removed by a mixed bed ion-exchange column. The specific conductance of the albumin samples is directly related to the isoionic pH. Therefore, the isoionic pH and the specific conductance of the albumins reflect the heterogeneity of the albumin samples with respect to their primary structure and fatty acid content.