If you build it, they will come: engineering the next generation of optical tools to image neural activity deep within the living brain

Abstract

To see the activity of large, integrated neural circuits functioning in real-time inside of a living brain, neuroscientists will need multiple genetically-encoded fluorescent activity sensors that can be individually targeted to specific cell types, are fast enough to resolve multiple action potentials, can be distinguished from one another and imaged deep within the brain. The goal of this work is to better understand and improve upon the most recent generations of genetically-encoded Ca 2+ and voltage sensors, and to expand biosensor utility in two-photon excitation, which will be necessary to image neural activity deep within the brain. Genetically-encoded Ca 2+ sensors measure the intracellular Ca 2+ release that occurs downstream of an action potential. The GCaMP6 series are the best Ca 2+ sensors available, however little is known about how they work. Measurements of four different states in GCaMP6m reveal that its large Ca 2+-dependent change in 470 nm excited fluorescence is due to a redistribution of the chromophore protonation state, from a neutral form excited at ~400 nm to an anionic form excited at ~470 nm, via a change in pK a. Making pK a-altering mutations in GCaMP6m changes the Ca 2+-dependent fluorescence response. This highlights the importance of Delta pK a and identifies key amino acid positions that will be important for improving GCaMP6m and GCaMP-like biosensors. A direct readout of an action potential would be ideal for capturing complex signal transduction in the brain. This will require a bright, fast voltage sensor. ElectricPk is the first genetically-encoded voltage sensor with a fluorescence response fast enough to resolve multiple action potentials in mammalian neurons. This design indicates it is possible to couple a fluorescence change with a very fast (~1 ms) voltage-dependent movement in the Ciona intestinalis voltage-sensitive phosphatase protein. Whether imaging a downstream Ca 2+ signal or a direct change in membrane potential, to image neuronal activity in deep brain tissue biosensors will need to be brightly fluorescent in two-photon excitation. The two-photon directed evolution of green fluorescent proteins presented here is a proof-of-principle design that shows a high-throughput screen focused on improving the two-photon properties of a fluorescent protein is possible.