Evaluating Candidate Lead Compounds by Rapid Analysis of Drug Interactions with Human Serum Albumin

Human serum albumin (HSA) plays a fundamental role in the transport of drugs, metabolites, and endogenous ligands. Binding to HSA controls the free, active concentration of a drug; provides a reservoir for a long duration of action; and ultimately affects drug absorption, distribution, metabolism, and excretion (ADME), making HSA binding a key parameter in evaluating candidate lead compounds during drug design.1 A broad range of approaches has been used to study drug–protein binding, including equilibrium dialysis, filtration, centrifugation, and chromatography. However, the resolution, reproducibility, and throughput of these methods are limiting.

Surface plasmon resonance (SPR) technology offers several advantages over traditional methods for analyzing drug interactions with proteins. Neither radiochemical nor fluorescent labels are required to provide real-time data on the affinity, specificity, and kinetics of drug–protein interactions. This means that all compounds can be analyzed, and binding measurements are not limited only to those proteins that can be derivatized with label moieties. SPR is also a highly sensitive technique that requires much less sample than traditional techniques, while enabling the detection of picomolar levels of proteins and molecules as small as a few hundred daltons, thus providing detailed information about drug–protein interactions. Since SPR provides real-time measurements, it delivers kinetic data on both complex formation and dissociation, unlike equilibrium-based methods. Finally, SPR can be highly automated for high throughput.

While SPR has been used to provide kinetic data on HSA–drug interactions,2,3 until recently, most SPR experiments for the determination of kinetic rate constants could only be run sequentially, limiting the throughput in the drug development process. An alternative approach utilizes the ProteOn™ XPR36 protein interaction array system and a One-shot Kinetics™ approach4 (Bio-Rad Laboratories, Hercules, CA). The system is a multiplexed SPR device that integrates a 6 × 6 interaction array for the analysis of up to six ligands with panels of up to six analytes, producing 36 data points in a single injection (Figure 1). Multiplexing improves and expands the capabilities of traditional SPR technology and work flow by enabling multiple quantitative protein binding experiments in parallel. Because multiple conditions can be tested in parallel, robust kinetic analysis of an analyte concentration series can be handled in one experiment. This one-shot parallel approach generates a complete kinetic profile of a biomolecular interaction without the need for regeneration, in one experiment, using a single sensor chip.

Figure 1 - ProteOn XPR36 system 6 × 6 ligand–analyte interaction array. a) HSA is immobilized in all six vertical channels. b) Six different concentrations of a drug are injected into the six horizontal channels. c) Detail of a single HSA–drug interaction spot (green).

This article describes a study designed to demonstrate the utility of the ProteOn XPR36 system for the rapid and accurate determination of binding parameters of small-molecule drugs to HSA. The binding of the following drugs, which have different molecular masses and affinities for HSA, was studied: corticosterone, a steroid hormone; naproxen and phenylbutazone, nonsteroidal antiinflammatory drugs (NSAIDs); quinine, used to prevent and treat malaria; and warfarin, an anticoagulant with well-characterized binding properties to HSA. The binding kinetics of dansylamide, a fluorescent probe used in traditional HSA binding methods, were also measured.

Real-time, label-free kinetic analysis of HSA–drug interactions

For SPR analysis, the ligand—in this case HSA—is covalently bound to a biosensor surface, and the binding partner—in this case a small-molecule drug—is injected across the surface of the sensor chip. As a binding event occurs, SPR detectors monitor the increase in refractive index near the sensor surface, leading to a shift in the SPR angle. When the injection solution is switched from analyte (drug) to buffer, the protein complex on the sensor surface dissociates, and the SPR angle shifts back. The shift in the SPR angle is measured in response units (RU) and recorded as a function of time in the form of a sensorgram (Figure 2). These changes in refractive index are proportional to mass changes at the surface of the sensor chip. The sensorgram thus displays the time course of binding of the drug to HSA. Optimized association, dissociation, and equilibrium constants are calculated by fitting response data for a range of analyte concentrations under identical reaction conditions to a computational model. In this case, drugs bind and dissociate from HSA very rapidly, as shown by the “square wave” shape of the binding curves.

Figure 2 - Representative sensorgram from SPR analysis of drug binding to HSA. Each of the six curves represents one of the six horizontal channels on the ProteOn XPR36 system, each of which contains a different concentration of the drug corticosterone. This sensorgram has a “square wave" shape due to the rapid binding and dissociation kinetics.

Drug affinities for HSA

The kinetics of six drugs can be measured using a single ProteOn GLM sensor chip to which HSA is immobilized in six vertical channels (Figure 1). Different concentrations of one drug can then be injected into the six horizontal channels simultaneously, thus obtaining a complete kinetic profile of the interaction of this drug with HSA, in a single injection. Six different concentrations of the second drug are then injected in the horizontal channels. No regeneration is required between drug injections, since the dissociation is very rapid, and the signal drops to zero baseline. Figure 3a shows the sensorgrams for each of six drugs tested, each curve (sensorgram) representing a different drug concentration. For the data shown in Figure 3, some drugs were tested at more than six concentrations for comparison to data reported in the literature. The signal at equilibrium was then plotted versus the drug concentration in order to determine the equilibrium constants (Figure 3b). From this plot, the equilibrium constant, KD, is fitted using the following equation:

where Req is the signal at equilibrium for each concentration, Rmax is the maximal signal when all the albumin binding sites are populated, and A is the analyte concentration. For the high-affinity compounds, a different model that includes two binding sites was used.3

Figure 3 - Drug-binding affinities to human serum albumin. a) Sensorgrams from binding kinetics studies of six drugs, at various concentrations, to HSA. b) Plot of signal at equilibrium versus the drug concentrations, to determine affinities.

The KD values calculated using the equations for the fitting are listed in Table 1, which also includes the literature values for these drugs determined by conventional SPR methods. The affinities vary as much as 1000-fold from a weakly binding drug such as quinine to a very tightly bound drug such as naproxen. Equilibrium binding data have traditionally been stated as percent bound values, which were originally calculated from equilibrium dialysis experiments. The KD values in Table 1 were used to calculate percent bound values as described in Ref. 2, and are also listed. It has been shown previously that affinity values measured by SPR correlate well with those determined using traditional techniques.3 While the values determined using the ProteOn XPR36 system agree quite well with those from the literature, the 6 × 6 interaction array and One-shot Kinetics approach of the system produce the same amount of data in about one-tenth the analysis time required by the SPR systems used to produce the values found in the literature.

Conclusion

It is advantageous to the drug discovery process to gather as much information about a candidate chemical compound as early in the process as possible, driving the need for high-throughput methods for analyzing HSA–drug interactions. While SPR is a high-resolution method for studying drug binding to HSA, until recently, its throughput was limited by the sequential approach used to perform binding analyses. These systems utilized a single injection needle, which meant that each concentration of each drug had to be injected sequentially. Since several concentrations of each drug are required to measure binding affinities, and at least 10–20 compounds may be tested in each study, this sequential injection approach greatly limited the throughput of SPR methods.

The ProteOn XPR36 system features six injection needles and the ability to perform injections in both the horizontal and vertical orientations of the sensor chip. This means that HSA and other plasma proteins can be bound to the chip in six vertical channels, and the sensor chip then rotated, so that different concentrations of a drug candidate compound can be injected across each of the plasma protein channels simultaneously, allowing robust kinetic analysis in one injection (One-shot Kinetics method). Through regeneration, multiple drug candidates can be screened using the same chip. The end result is that the ProteOn XPR36 system can significantly reduce the experimental time from days to hours, while generating the same amount of data as other SPR systems. With this significant increase in throughput, SPR is now a more valuable tool that can better meet the needs of drug discovery for rapid screening and accurate characterization of the affinity of chemical compounds for plasma proteins.

References

  1. Bertucci, C.; Domenici, E. Reversible and covalent binding of drugs to human serum albumin: methodological approaches and physiological relevance. Curr. Med. Chem. 2002, 9, 1463–81.
  2. Rich, R.L.; Day, Y.S.N.; Morton, T.A.; Myszka, D.G. High-resolution and high-throughput protocols for measuring drug/human serum albumin interactions using Biacore. Anal. Biochem.2001, 296, 197–207.
  3. Day, Y.S.N.; Myszka, D.G. Characterizing a drug’s primary binding site on albumin. J. Pharm. Sci.2003, 92, 333–43.
  4. Bravman, T.; Bronner, V.; Lavie, K.; Notcovich, A.; Papalia, G.A.; Myszka, D.G. Exploring “One-shot” Kinetics and small molecule analysis using the ProteOn XPR36 array biosensor. Anal. Biochem.2006, 358, 281–8.

Ms. Bronner and Mr. Nahshol are ProteOn Applications Scientists, and Dr. Bravman is a ProteOn Applications Manager, Bio-Rad Laboratories, Life Science Group, 2000 Alfred Nobel Dr., Hercules, CA 94547, U.S.A.; tel.: 800-424-6723; fax: 800-849-2289; e-mail: [email protected].