ALGEBRAIC STUDY OF RECEPTOR-LIGAND SYSTEMS: A DOSE-RESPONSE ANALYSIS

Abstract

The study of a receptor-ligand system generally relies on the analysis of its doseresponse (or concentration-effect) curve, which quantifies the relation between ligand concentration and the biological effect (or cellular response) induced when binding its specific cell surface receptor. Mathematical models of receptor-ligand systems have been developed to compute a dose-response curve under the assumption that the biological effect is proportional to the number of ligand-bound receptors. Given a dose-response curve, two quantities (or metrics) have been defined to characterize the properties of the ligand-receptor system under consideration: amplitude and potency (or halfmaximal effective concentration, and denoted by EC50). Both the amplitude and the EC50 are key quantities commonly used in pharmaco-dynamic modeling, yet a comprehensive mathematical investigation of the behavior of these two metrics is still outstanding; for a large (and important) family of receptors, called cytokine receptors, we still do not know how amplitude and EC50 depend on receptor copy numbers. Here we make use of algebraic approaches (Gröbner basis) to study these metrics for a large class of receptor-ligand models, with a focus on cytokine receptors. In particular, we introduce a method, making use of two motivating examples based on the interleukin- 7 (IL-7) receptor, to compute analytic expressions for the amplitude and the EC50. We then extend the method to a wider class of receptor-ligand systems, sequential receptor-ligand systems with extrinsic kinase, and provide some examples. The algebraic methods developed in this paper not only reduce computational costs and numerical errors, but allow us to explicitly identify key molecular parameters and rates which determine the behavior of the dose-response curve. Thus, the proposed methods provide a novel and useful approach to perform model validation, assay design and parameter exploration of receptor-ligand systems.