A Comprehensive Live Cell Screening Approach for Developing Improved Microbial Rhodopsin-Based Voltage Biosensors

Fluorescence imaging of neuron activity would be greatly facilitated by the availability of brightly fluorescent genetically encoded voltage indicators with large and fast responses to membrane potential changes. The Cohen group has recently described a new class of genetically encoded voltage indicators based on the microbial rhodopsin protein archaerhodopsin-3 (Arch) with excellent properties. However, this class of voltage indicators suffers from poor quantum efficiency that limits the range of potential applications. To address this shortcoming, we developed a strategy for directed evolution for brighter Arch-based voltage indicators. Briefly, a random genetic library of Arch is fused to the N-terminus of the fluorescent protein mOrange2 and is expressed in E. coli. Hundreds of colonies on a Petri dish are ratiometrically imaged (i.e., the ratio of Arch to mOrange2 fluorescence) using a custom imaging system. The colonies with the highest ratio are picked and cultured in liquid media. The fluorescence of the overnight cultures is then recorded with a microplate reader. The brightest Arch variants are expressed in Hela cells and their voltage sensitivities are evaluated in a fluorescent microscope by electric field stimulation. The brightest functional variants are then used as the library template for the next round of directed evolution. Several iterative rounds of this screening procedure combined with further screening of site-directed mutagenesis libraries resulted in identification of our current best voltage-sensitive Arch (vArch) variant, vArch1.0. vArch1.0 is voltage sensitive, fast and several-fold brighter than wild-type Arch in mammalian cells. Unlike Arch, vArch1.0 doesn't generate photo-induced current that perturbs membrane potential of cells during imaging. vArch1.0 is capable of resolving electrically triggered action potentials in neurons in single trials with optical signal-to-noise ratio >30 and is a promising tool for optical interrogation of complex neural circuit.