Utility of a mixed effects approach to defining target binding rate constants.
Neil Benson1, Nelleke Snelder2, Bart Ploeger2 , Carolyn Napier1, Harriet Sale1 and Piet van der Graaf1
1. Pfizer Global R&D 2.LAP&P Consultants
Objectives: The incorporation of target binding kinetics into PK/PD models has long been established (1). In recent years, both within our company and others, there has been a marked increase in interest in designing ‘slow-offset' ligands as a deliberate medicinal chemistry strategy to improve duration of action. In particular, targets where lead material does not meet conventional minimally acceptable pharmacokinetic criteria are of interest (2). However, despite the fact that methodologies and principles in this area have been developing for almost half a century (see 3, 4), defining transient target on-and off-rate kinetics remains a very resource intensive and reagent consuming process, with efficient parameter estimation often proving difficult in practice. This is a significant barrier for implementation as a routine methodology in drug discovery programs. In an attempt to address these issues, we have implemented receptor kinetic models in NONMEM. We have applied these models to analyse simulated and real datasets and to identify more optimal experiment design using simulation.
Methods: Test datasets were simulated using Gepasi (v3.30). Transient kinetic data were generated using competitive binding of radioactive tracer and test drug essentially as described in (4). Data were analysed and alternative designs simulated using NONMEM V.
Results: An initial comparison with conventional (naïve pool/sequential) data analysis methods demonstrated the superiority of the nonlinear mixed-effect modelling approach. Furthermore, the NONMEM model simulations, with representative experimental dispersion, suggested that significant reductions in experimental efforts could be made without compromising the quality of the transient kinetic parameter estimates.
Conclusions: In conclusion, we have developed a non-linear mixed effect modelling approach for the analysis of transient target binding kinetics and estimation of ligand dissociation and association rate constants. This methodology may aid in the further integration of target binding kinetics as a designable property in drug discovery programs. In principle this approach should also be generically applicable to other in vitro transient kinetic problems such as slow tight binding and suicide inhibition of enzymes.
. Shimada et al, Biol. Pharm. Bull., 19(3), 430-437, (1996).
. Copeland, RA et al, Nat Rev Drug Discov. 2006 Sep;5(9):730-9. Epub 2006 Aug 4.
. Paton, WDM, Proc. R. Soc. Lond. B., 154, 21-69 (1961)
. Motulsky and Mahan, Molecular Pharmacology, 25:1-9 (1983)