Quantitative studies of allosteric effects by biointeraction chromatography: analysis of protein binding for low-solubility drugs

Abstract

A new chromatographic method was developed for characterizing allosteric interactions between an immobilized binding agent and low-solubility compounds. This approach was illustrated by using it to characterize the interactions between tamoxifen and warfarin during their binding to the protein human serum albumin (HSA), with beta-cyclodextrin being employed as a solubilizing agent for these drugs. It was confirmed in this work through several experiments that warfarin had a single binding site on HSA with an association equilibrium constant of (2-5) x 10(5) M(-1) (average, 3.9 x 10(5) M(-1)) at 37 degrees C, in agreement with previous reports. It was also found that tamoxifen had a single major binding site on HSA, with an association equilibrium constant of (3-4) x 10(7) M(-1) (average, 3.5 x 10(7) M(-1)) at 37 degrees C. When warfarin was used as a mobile-phase additive in competition studies with tamoxifen, this had a positive allosteric effect on tamoxifen/HSA binding, giving a coupling constant of 2.3 (+/-0.3). Competitive studies using tamoxifen as a mobile-phase additive indicated that tamoxifen had a negative allosteric effect on warfarin/HSA binding, providing a coupling constant of 0.79 (+/-0.03). A unique feature of the technique described in this report was its ability to independently examine both directions of the warfarin/tamoxifen allosteric interaction. This approach is not limited to warfarin, tamoxifen, and HSA but can also be used to study other solutes and binding agents.

Figure 1

Structures of warfarin and tamoxifen.

Figure 2

General model for (a) direct competition or (b) allosteric competition for the binding of analyte A and mobile additive I with immobilized binding agent L in the presence of solubilizing agent S. The other symbols are defined in the text.

Figure 3

Chromatograms for (a) self-competition zonal elution studies for warfarin, (b) zonal elution studies examining the competition of injected tamoxifen in the presence of various mobile phase concentrations of warfarin, and (c) zonal elution studies examining the competition of injected warfarin in the presence of various mobile phase concentrations of tamoxifen. The dashed lines show the mean retention times measured for the injected analytes in the presence of no competing agent (Note: in (c) the dashed line is shown for the second warfarin peak, which is due to the S-enantiomer of warfarin). The total concentration of warfarin (C_Warfarin) or tamoxifen (C_Tamoxifen) that was used as a competing agent in the mobile phase is indicated to the right of each chromatogram. Other conditions are given in the text.

Figure 4

Plots of (a) 1/k versus the total mobile phase concentration of warfarin (C_Warfarin) and (b) k₀/(k - k₀) versus 1/C_Warfarin for self-competition zonal elution studies with warfarin in the presence of 2.2 mM β-cyclodextrin as a solubilizing agent. The equations for the best-fit lines were: (a) y = 0.030 (± 0.001) × + 0.120 (± 0.003), with a correlation coefficient of 0.997 (n = 5); and (b) y = -4.42 (± 0.24) × - 0.86 (± 0.14), with a correlation coefficient of 0.997 (n = 4). The error bars in (a) and (b) represent a range of ± 2 S.D.

Figure 5

Plots of (a) 1/(k - x) versus the total mobile phase concentration of tamoxifen (C_Tamoxifen) and (b) (k₀ - x)/(k - k₀) versus 1/C_Tamoxifen for self-competition zonal elution studies with tamoxifen in the presence of 2.2 mM β-cyclodextrin. The term x here represents the non-specific binding of tamoxifen to the support, as measured on the control column. The equations for the best-fit lines in this figure were (a) y = 0.079 (± 0.009) × + 0.043 (± 0.015), with a correlation coefficient of 0.98 (n = 5); (b) y = -0.24 (± 0.06) × - 1.08 (± 0.05), with a correlation coefficient of 0.94 (n = 4). The error bars in (a) and (b) represent a range of ± 2 S.D.

Figure 6

Plots of (a) 1/(k - x) versus the total mobile phase concentration of warfarin (C_Warfarin) and (b) (k₀ - x)/(k - k₀) versus 1/(C_Warfarin) for tamoxifen binding to immobilized HSA in the presence of 2.2 mM β-cyclodextrin and various concentrations of warfarin in the mobile phase. In these plots, k and k₀ are the retention factors for tamoxifen in the presence or absence of warfarin in the mobile phase, respectively. The term x is the retention factor due to non-specific binding. The equation for the best-fit line in (b) is y = 7.6 (± 0.3) × + 0.79 (± 0.17), with a correlation coefficient of 0.999 (n = 4). The error bars in (a) and (b) represent a range of ± 2 S.D.

Figure 7

Plots of (a) 1/k versus the total mobile phase concentration of tamoxifen (C_Tamoxifen) and (b) k₀/(k - k₀) versus 1/(C_Tamoxifen) for injections of warfarin in the presence of 2.2 mM β-cyclodextrin and various concentrations of tamoxifen in the mobile phase. In these plots, k and k₀ are the retention factors for warfarin in the presence or absence of tamoxifen in the mobile phase, respectively. The equations for the best-fit lines are as follows:(a) y = 0.0064 (± 0.0017) × + 0.093 (± 0.003), with a correlation coefficient of 0.91 (n = 5); (b) y = -3.3 (± 1.1) × - 4.8 (± 0.8), with a correlation coefficient of 0.91 (n = 4). The error bars in (a) and (b) represent a range of ± 2 S.D.

Figure 8

General model for the interactions between warfarin and tamoxifen on HSA. The association equilibrium constants (K) given here are the averages of the various values measured in this study. The coupling constants (β) were determined from the competition experiments described in the text between warfarin and tamoxifen. The overall range of association constants found in this report for warfarin with HSA was 2-5 × 10⁵ M^-1 (n = 3) at 37°C. The range of association equilibrium constants determined for tamoxifen for HSA was 3-4 × 10⁷ M^-1 (n = 4) at 37°C (Note: one additional value of 1.1 × 10⁸ M^-1 was determined, but this was not included in the given average since it was a suspected outlier; when this was included, the average association equilibrium constant for tamoxifen with HSA was 5.0 (± 3.9) × 10⁷ M^-1).