Studies of drug binding to plasma proteins using a variant of equilibrium dialysis
Introduction
The two predominant techniques for studying drug protein binding in plasma are equilibrium dialysis (ED) and ultrafiltration (UF). UF today, which is probably the most common method for the determination of unbound plasma concentrations (Cu), is a rapid procedure (15–45 min being a typical range) and is very simple to use, since commercial devices are available. Recently, 96-well plates for UF were also introduced (Millipore Corp., Danvers, MA, USA), making this technique even more attractive. A major drawback is that adsorption to the UF device and the filter can be a source of significant error in this type of experiment. Non-specific adsorption must be checked for all compounds where UF is used for generating unbound fractions or concentrations. Another drawback is protein leakage across the filter, which can cause erroneously high free concentrations for highly bound compounds.
The other common technique for determining the protein binding of a drug is equilibrium dialysis. The ED device contains two chambers, divided by a semipermeable membrane that only allows the passage of molecules with a molecular weight less than the molecular weight cut-off (MWCO) of the membrane. In a typical ED experiment, plasma (the ‘retentate’)—containing the drug—is on one side of the membrane and a buffer (the ‘dialysate’) is placed in the other dialysis chamber. While ED is not as simple to use as UF, it has the advantage that non-specific adsorption can be compensated for if the concentration at equilibrium on each side is measured, which means that the adsorption will not affect the concentration ratio at equilibrium, only the mass balance. One drawback is the relatively long times (>20 h) often needed to reach equilibrium, which could give rise to degradation and changes in the pH of the plasma in the course of the dialysis. Another drawback is that the unbound fraction at equilibrium differs from the initial unbound fraction if the binding is concentration-dependent, thus the unbound fraction should always be related to the total concentration measured on the plasma side at equilibrium. This also means that the actual free concentration of an in vivo sample cannot be determined using ED if the binding is concentration dependent. A frequently discussed problem of ED experiments is the volume shift [1], the flow of fluid from the buffer side to the plasma side. The volume shift is caused by an influx of the dialysate to the retentate, due to the osmotic pressure of the plasma proteins. The effects of volume shifts are often complicated by the fact that plasma diluted with buffer might change its binding properties in an unpredictable way, i.e. the dilution is not ideal and uniform. Factors, such as changes in ionic strength or pH, could result in significant changes in the binding properties of the proteins involved. Another commonly identified problem with the ED method is the uneven distribution of low molecular weight ions, the Gibbs–Donnan effect, due to the fact that charged proteins cannot pass through the membrane, which results in a flow of small ions across the membrane to achieve electroneutrality. The problem is reduced by using an isotone phosphate buffer to which electrolytes (i.e. NaCl) are added to diminish the difference in ionic strength on either side of the membrane [2]. ED has also the reputation as being a rather laborious and time consuming technique although some recent papers show that high throughput approaches using the 96-well format are possible [3], [4].
From the above, it can be understood that both techniques have their problems and that all unbound fraction/unbound concentration results must be looked at critically. The search for a general methodology has resulted in a plethora of publications describing alternative methods, including ultracentrifugation, frontal analysis chromatography, affinity chromatography, negligible extraction methods and erythrocyte/plasma distribution [5], [6], [7].
The objective of this study was to investigate the usefulness of a recently published variant of the ED method where the drugs valproate and monoacetyl dapsone were studied, both compounds showing high plasma protein binding [8]. This method will be referred to as comparative equilibrium dialysis (CED). In CED, plasma (usually spiked) from two different sources (species, individuals, etc.) is placed on either side of the dialysis membrane instead of, as in traditional ED, plasma on one side and buffer on the other. At equilibrium, the Ctot measured on either side of the membrane reflects the ratio between the respective unbound fractions (see Fig. 1). This relative binding for two species would be very valuable when, for example, scaling the pharmacokinetics from animals to humans, since information about the ‘equivalent total concentrations’ for a given unbound drug concentration is ideally obtained in this way. Also, in the case of drugs with extremely high protein binding, the low unbound concentrations can be difficult to measure with acceptable precision. In a CED experiment, this problem is circumvented, as high total concentrations are determined instead. CED would therefore also be an excellent tool for verifying whether two determined unbound fractions are correct, e.g. whether the species with a determined protein binding of 99.4% really has a three times higher unbound fraction than the species with a protein binding of 99.8%. Another interesting possibility opens up for compounds with very large species variations in protein binding: if the absolute unbound fraction in the species with the lowest protein binding can be determined with acceptable precision, it would be possible to obtain a good estimate of the absolute unbound fraction for all species via a series of CED experiments.
The relationship between the unbound and total drug concentrations in plasma in vivo is described by:where Cu is determined by dose rate and unbound clearance and fu is a function of drug–protein affinity and the concentration of binding sites. The total plasma concentration then becomes a function of Cu and fu.
In a CED experiment, using plasma from two different species, denoted I and II, at equilibrium the unbound concentrations (Cu) are the same on both sides:From Eq. (1), it can be seen that at a given unbound concentration the total concentration in each chamber will be governed by the unbound fraction of the drug in the chamber. Thus,or (rearranged)Eq. (4) shows that the ratio between the determined total concentrations of two different types of plasma is the same as the inverse ratio between the unbound fractions of the same plasma types. Thus, determination of the total concentration in each chamber can be used to study the relationship between the unbound fractions of two species, for example, and if the absolute unbound fraction is known for one species, the other unbound fraction can be calculated.
Three model compounds with differing protein-binding properties were used to study this promising variant of ED, some properties being shown in Table 1. The first compound, referred to as Compound 1, shows very high plasma protein binding in humans (predominantly to albumin, unbound fraction about 0.05%). The unbound fraction in the rabbit is considerably higher, 1.5%, offering good possibilities for comparison with human plasma in CED experiments. The second compound, NAD-299, was chosen since it does not bind to plasma proteins to the same extent [9]. It is a basic drug, binding mainly to α1-acid glycoprotein (α1-AGP). The between-species variation of the unbound fraction is also large (Table 1), which should facilitate the interpretation and validation of the CED method. Tolterodine, which also shows large species differences in the degree of plasma protein binding [10], was chosen as a third model compound to further investigate the CED method.
Section snippets
Materials
All water was purified in a Milli-Q filtration system (Millipore Corp., Bedford, MA, USA). For the phosphate buffer, 34.5 g NaH2PO4·H2O (Merck) and 22.2 g Na2HPO4·2H2O (Merck) were each dissolved in 250 mL of water, giving concentrations of 1.00 and 0.50 M, respectively. Fifteen millilitres of the solution containing Na2HPO4 was diluted to 0.015 M and the pH was adjusted to 7.4 with the solution containing NaH2PO4. For soaking of the dialysis membranes, an isotone buffer [3] was prepared: 4.00 g Na2
Compound 1 experiments
As an initial experiment, the time required to reach equilibrium between two dialysis chambers was investigated, i.e. when the values of Ctot on both sides of the dialysis membrane were constant if plasma from different species was used on each side or, if the same plasma was used on both sides, when Ctot was the same on both sides. In Fig. 2(a) the time course of Ctot is shown for a CED experiment with human plasma in both chambers, where one side was spiked with 100 μM of Compound 1 and the
Conclusions
When exploring the CED method for the determination of relative unbound concentrations, several difficulties were encountered. The time to reach equilibrium was strongly dependent on the free faction. For Compound 1, unusually strongly bound in human plasma, equilibrium could no be reached within a reasonable time (<24 h). For the other two model compounds a more rapid equilibration (<16 h) was seen. CED experiments were done with plasma from three different species (human, dog and rat) and
Acknowledgment
Gerd Ackehed is gratefully acknowledged for providing data and bioanalytical method for one of our model compounds.
References (11)
- et al.
J. Pharm. Sci.
(2003) - et al.
J. Pharm. Sci.
(2001) - et al.
J. Chromatogr. B
(1996) - et al.
Trends Anal. Chem.
(2003) - et al.
J. Pharm. Sci.
(2000)
Cited by (24)
Stereoselective protein binding studies of 2-(2-hydroxypropanamido) benzoic acid enantiomers in rat, beagle dog and human plasma
2017, Journal of Chromatography B: Analytical Technologies in the Biomedical and Life SciencesCitation Excerpt :To the best of our knowledge, ED method is one of the most commonly used techniques to study plasma protein binding in drug discovery and drug development because of it simplicity and general applicability to many different systems in vitro and ex vivo. ED method has the advantage to compensate for the non-specific adsorption if the concentration at equilibrium on each side is measured, which means that the concentration ratio will not be affected by the adsorption at equilibrium, only the mass balance [14]. (R)-/(S)-2-(2-hydroxypropanamido) benzoic acid ((R)-/(S)-HPABA, Fig. 1A, B) was initially obtained from the fermentation broth of a marine fungus Penicillium chrysogenum. (
Human serum albumin-mimetic chromatography based hexadecyltrimethylammonium bromide as a novel direct probe for protein binding of acidic drugs
2015, Journal of Pharmaceutical and Biomedical AnalysisCitation Excerpt :BMC can neither describe active and metabolic processes nor determine enantioselective differences between the enantiomers of a chiral drug [19]. Despite the mentioned limitation of present method, it has some important advantages over other in vitro techniques used to protein binding assay (e.g. equilibrium dialysis [22], ultrafiltration [22], biosensors and 96-well fluorescence plate readers [23,24]): the preparation of the chromatographic system is rapid, simple and economical, the reproducibility intra- and inter-day of the retention data is very high that permits the protein binding estimation with high accuracy [25]. Also, this approach can provide key information about the potential protein binding properties of new acidic drugs during the drug discovery processes and so, can be very useful in medicinal chemistry and pharmaceutical research.
A pulsed tangential-flow ultrafiltration technique for studying protein-drug binding
2013, Journal of Pharmaceutical SciencesCitation Excerpt :Membrane-based techniques utilize the differential barrier property of membranes toward free and protein-bound drug molecules, that is, one permeates, whereas the other is retained. Equilibrium dialysis (ED) that is commonly used for determining the fraction of drug bound to protein relies on diffusion of free drug molecules through a membrane from a donor compartment to a receiving compartment, till equilibrium is attained.13,15 One of the major drawbacks of ED is that it is slow as the transport of drug takes place by diffusion.6,7
Biopartitioning micellar chromatography with sodium dodecyl sulfate as a pseudo α<inf>1</inf>-acid glycoprotein to the prediction of protein-drug binding
2013, Journal of Chromatography B: Analytical Technologies in the Biomedical and Life SciencesCitation Excerpt :Recently, biopartitioning micellar chromatography (BMC) has been proposed as a high-throughput screening platform and attracted considerable attention as an in vitro model to predict the pharmacological and pharmacokinetic properties of drugs [17–20]. The BMC methodology has important advantages over other in vitro techniques used to protein binding assay (e.g. equilibrium dialysis [21], ultrafiltration [21], biosensors and 96-well fluorescence plate readers [22,23]) including accuracy, low cost, experimental simplicity, speed and reproducibility. In this chromatographic system, polioxyethylene 23 lauryl ether (Brij-35) micellar mobile phases and C18 reversed stationary phase in adequate experimental conditions are used.