A high throughput platform for measuring and predicting vitrification behavior in multicomponent aqueous solutions

Abstract

Cryopreservation depends critically on suppression of ice formation by cryoprotective agents (CPAs), but limited data is available on the CPA concentration required for vitrification (Cv). Here, we introduce a high-throughput 384-well platform that integrates automated liquid handling, randomized plate layouts, and a binary-search strategy to rapidly determine Cv across hundreds of formulations. Relative to conventional methods, this approach increases throughput by [~]50-fold, compressing a year of measurements into one week, while markedly reducing manual labor. Across [~]200 CPA compositions, we demonstrate that environmental boundary conditions strongly influence vitrification behavior: plates sealed with silicone mats exhibited lower Cv than open plates, indicating that sealed configurations promote vitrification. Further, the data reveal a decrease in Cv with increasing CPA molecular weight, consistent with enhanced ice suppression by larger molecules. We also present a simple mixture model that accurately predicts Cv for a broad range of CPA formulations, including mixtures containing up to seven CPAs (R{superscript 2} [≥] 0.93), and use this model to evaluate published CPA toxicity data to identify formulations that operate near their vitrification threshold while maintaining relatively low toxicity. Together, these results establish a framework for rapid Cv determination, predictive modeling of vitrification behavior, and rational design of CPA formulations.