Simple thermodynamic model of unassisted proton shuttle uncoupling and prediction of activity from calculated speciation, lipophilicity, and molecular geometry

A mechanistic model of uncoupling of oxidative phosphorylation by lipophilic weak acids (i.e. proton shuttles) was developed for the purposes of predicting the relative activity of xenobiotics of widely varying structure and of guiding the design of optimized derivatives. The model is based on thermodynamic premises not formulated elsewhere that allow for the calculation of steady-state conditions and of rate of energy dissipation on the basis of acid-dissociation and permeability behavior, the later estimated from partitioning behavior and geometric considerations. Moreover, permeability of either the neutral or of the ionized species is proposed to be effectively enhanced under conditions of asymmetrical molecular distribution. Finally, special considerations were developed to accommodate multi-protic compounds. The comparison of predicted to measured activity for a diverse testset of 48 compounds of natural origin spanning a wide range of activity yielded a Spearman’s rho of 0.90. The model was used to tentatively identify several novel proton shuttles, as well as to elucidate core structures particularly conducive to proton shuttle activity from which optimized derivatives can be designed. Principles of design were formulated and examples of derivatives projected to be active at concentrations on the order of 10−7M are proposed. Among these are di-protic compounds predicted to shuttle two protons per cycle iteration and proposed to maximally exploit the proton shuttle mechanism. This work promotes the design of highly active, yet easily-metabolized uncouplers for therapeutic applications, namely the indirect activation of AMP-kinase, as well as for various industrial applications where low persistence is desirable. [Display omitted] ► A physiological model of uncoupling by proton shuttles was developed. ► It allows activity to be predicted from easily-estimated physicochemical properties. ► It effectively discriminates actives from related compounds with little to no activity. ► The model was used to screen a library of naturally-occurring compounds in-silico. ► The model was used to formulate principles of design for optimized activity.