Immobilization of alpha1-acid glycoprotein for chromatographic studies of drug-protein binding II. correction for errors in association constant measurements

Abstract

A new method for the immobilization of alpha(1)-acid glycoprotein (AGP) in high-performance liquid chromatography (HPLC) columns was recently described for applications such as drug binding studies. Part of this earlier work used self-competition zonal elution studies to measure association equilibrium constants between immobilized AGP and R- or S-propranolol. It was later found that analysis of these data by a common equation derived for linear elution conditions gave erroneous values for experiments actually conducted under nonlinear conditions. This report discusses the nature of this error and uses frontal analysis to estimate the true binding strength between R- and S-propranolol and HPLC columns containing immobilized AGP.

Figure 1

Frontal analysis results for R-propranolol at pH 7.4 and 37.0°C on a hydrazide-linked AGP column. The measured values for m_Lapp had precisions of ±0.4 to 13% over the range of propranolol concentrations that were examined. The best-fit line shown in this graph is for the equation m_Lapp = (m_L1K_a1 [Propranolol])/(1 + K_a1[Propranolol]) + (m_L2K_a2) [Propranolol], where K_a1 and K_a2 are the association equilibrium constants for a set of selective and non-selective sites for propranolol on AGP, with m_L1 and m_L2 being the moles of these binding sites that are present in the AGP column. Using the total estimated amount of AGP in the column (m_L,AGP), the following relationships were used with the best-fit results to obtain the values for n₁ and n₂ K_a2 that are given in the text: m_L1/m_L,AGP = n₁ and (m_L2 K_a2)/m_L,AGP = n₂ K_a2. The dashed line in the inset shows the response predicted for the high affinity selective site in the two-site model.