Functional estimation of endothelin-1 receptor antagonism by bosentan, macitentan and ambrisentan in human pulmonary and radial arteries in vitro

Abstract

Background

Endothelin receptor antagonists are approved for pulmonary arterial hypertension. Development of selective ET_A-receptor antagonists over mixed or dual receptor antagonists has depended on a range of receptor binding assays, second messenger assays and functional blood vessel assays. This study compared the 3 clinically-approved endothelin receptor antagonists in assays of human isolated pulmonary and radial arteries in vitro.

Methods

Human isolated pulmonary (i.d. 5.5 mm) and human radial (i.d. 3.23 mm) artery ring segments were mounted in organ baths for isometric force measurement. Single concentration-contraction curves to endothelin-1 were constructed in the absence or presence of bosentan (1–10 µM), macitentan (0.03–0.3 µM) or ambrisentan (0.1–1 µM).

Results

All 3 endothelin antagonists caused competitive rightward shifts in the endothelin-1 concentration-response curves in both arteries. The Clark plot and analysis gave the following pK_B values: bosentan, pulmonary artery 6.28±0.13 and radial artery 6.04±0.10; macitentan, pulmonary artery 8.02±0.13 and radial artery 7.49±0.08; and ambrisentan, pulmonary artery 7.38±0.13 and radial artery 6.96±0.10.

Conclusions

Noting the maximum plasma levels attained from recommended oral doses of each antagonist in volunteers, the pK_B findings here show that there would be significant antagonism of endothelin-1 contraction in the pulmonary and radial arteries at therapeutic plasma levels. This functional assay confirms in human tissue that much higher plasma concentrations of endothelin-1 receptor antagonists are required to be effective than those predicted from binding or other biochemical assays.

Introduction

There is growing interest in the question of whether selectivity of newer endothelin-1 receptor antagonists for the ET_A receptor based on binding studies bears any clinical importance compared with “dual” or “mixed” non-selective ET_A and ET_B receptor antagonists. Indeed, secondary properties such as limiting side effects, longer half-times for once a day dosing and fewer serious drug-drug interactions may prevail as being more important characteristics than receptor selectivity based on binding (Opitz et al., 2008). Historically, much of the quantitative analysis of a vast range of endothelin-1 antagonists from four chemical classes has relied on the determination of K_i values, the dissociation constant of unlabelled ligands in competition binding experiments in human cloned ET_A and ET_B endothelin receptor assays. In “clinically relevant” human cardiovascular tissues such as samples from left ventricle, coronary artery and homogenates of saphenous vein, the binding assay readout for K_i values of a range of endothelin-1 antagonists were compared to the “functional” K_B values, i.e. the antagonist concentration that shifted the agonist endothelin-1 EC₅₀ two-fold to the right (Maguire et al., 2012). These authors reported a significant discrepancy between the K_i and K_B values that was not correlated with the structural class of the endothelin antagonists.

In this work we have determined the functional K_B values of bosentan, macitentan and ambrisentan in isolated ring segments of human pulmonary and radial arteries (discarded from surgical procedures) in organ baths and compared these K_B values to the literature-reported plasma levels of these drugs used in the treatment of pulmonary hypertension. One advantage of these two human large conduit vessels is that the ET_A receptor appears to be the primary functional endothelin-1 receptor subclass that mediates the contraction (Conant et al., 2002, Hay et al., 1993, Liu et al., 1996, Maguire and Davenport, 1995).