Rapid EC50 determination of hydrophobic toxicants in continuous droplet biomicrofluidics

Abstract

Droplet-based bioassays are attractive for increasing throughput while minimizing reagent consumption. However, the choice of continuous oil phase within the bioassay can substantially alter assay outcomes when hydrophobic reagents and analytes dynamically partition between the oil and aqueous phases. Fluorinated continuous phases have been recognized as a better alternative to hydrocarbons to prevent leakage from droplets but variations among them has not been quantitatively addressed and analysis is complicated by the presence of surfactants. Herein, surface modification strategies to avoid surfactants are demonstrated with fluorocarbon oils for the first time. Perfluorocarbons, hydrofluoroether and hexadecane are quantitatively compared based on their aqueous distribution coefficient with a model industrial toxicant, pentachlorophenol (PCP). PCP partition in the oil phase skews cytotoxicity evaluation, wherein a bacterial inhibition assay displays a wide range of EC50 values for the same toxicant. Prevention of PCP extraction in a real-time viability assay reveals a 40-fold difference in LD50 toxicity vs. hexadecane control. Theoretical and empirical approaches to limit bioassay partitioning are established for expansion outside cell viability analyses employed herein. Structure activity relationships of surface modification and oil-water partition are related to Hamaker constants and Hansen solubility parameters towards translation on disposable plastics beside poly(methyl acrylate). The results provide a theoretical basis towards design and selection of continuous phase oils and microfluidic substrates to achieve a range of aqueous/oil partitioning for droplet microfluidic bioassays.