Computationally Assisted Design of High Signal-to-Noise Photoinduced Electron Transfer-Based Voltage-Sensitive Dyes

High signal-to-noise optical voltage indicators will enable simultaneous interrogation of membrane potential in large ensembles of neurons. However, design principles for voltage sensors with high sensitivity and brightness remain elusive, limiting the applicability of voltage imaging. In this paper, we use molecular dynamics (MD) simulations and density functional theory (DFT) calculations to guide the design of a bright and sensitive green-fluorescent voltage-sensitive fluorophore, or VoltageFluor (VF dye), that uses photoinduced electron transfer (PeT) as a voltage-sensing mechanism. MD simulations predict an 11% increase in sensitivity due to membrane orientation, while DFT calculations predict an increase in fluorescence quantum yield, but a decrease in sensitivity due to a decrease in rate of PeT. We confirm these predictions by synthesizing a new VF dye and demonstrating that it displays the expected improvements by doubling the brightness and retaining similar sensitivity to prior VF dyes. Combining theoretical predictions and experimental validation has resulted in the synthesis of the highest signal-to-noise green VF dye to date. We use this new voltage indicator to monitor the electrophysiological maturation of human embryonic stem cell-derived medium spiny neurons.