Optimization of Cortisol-Selective Molecularly Imprinted Polymers Enabled by Molecular Dynamics Simulations

Today, we heavily rely on technology and increasingly utilize it to monitor our own health. The identification of sensitive, accurate biosensors that are capable of real-time cortisol analysis is one important potential feature for these technologies to aid us in the maintenance of our physical and mental wellbeing. Detection and quantification of cortisol, a well-known stress biomarker present in sweat, offers a noninvasive and potentially real-time method for monitoring anxiety. Molecularly imprinted polymers are attractive candidates for cortisol recognition elements in such devices as they can selectively rebind a targeted template molecule. However, mechanisms of imprinting and subsequent rebinding depend on the choice and composition of the prepolymerization mixture where the molecular interactions between the template, functional monomer, cross-linker, and solvent molecules are not fully understood. Here, we report the synthesis and evaluation of a molecularly imprinted polymer selective for cortisol detection. Molecular dynamics simulations were used to investigate the interactions between all components in the prepolymerization mixture of the as-synthesized molecularly imprinted polymer. Varying the component ratio of the prepolymerization mixture indicates that the number of cross-linker molecules relative to the template impacts the quality of imprinting. It was determined that a component ratio of 1:6:30 of cortisol, methacrylic acid, and ethylene glycol dimethacrylate, respectively, yields the optimal theoretical complexation of cortisol for the polymeric systems investigated. Experimental synthesis and rebinding results demonstrate an imprinting factor of up to 6.45. The trends in cortisol affinity predicted by molecular dynamics simulations of the prepolymerization mixture were also corroborated through experimental analysis of those modeled molecularly imprinted compositions, demonstrating the predictive capabilities of these simulations.