Genetically encoded voltage indicator

Even though the idea of optical measurement of neuronal activity was proposed in the late 1960s,^[7] the first successful GEVI that was convenient enough to put into actual use was not developed until technologies of genetic engineering had become mature in the late 1990s. The first GEVI, coined FlaSh,^[8] was constructed by fusing a modified green fluorescent protein with a voltage-sensitive K⁺ channel (Shaker). Unlike fluorescent proteins, the discovery of new GEVIs are seldom inspired by nature, for it is hard to find an organism which naturally has the ability to change its fluorescence based on voltage. Therefore, new GEVIs are mostly the products of genetic and protein engineering.

Two methods can be utilized to find novel GEVIs: rational design and directed evolution. The former method contributes to the most of new GEVI variants, but recent research using directed evolution have shown promising results in GEVI optimization.^[9][10]

Conceptually, a GEVI should sense the voltage difference across the cell membrane and report it by a change in fluorescence. Many different structures can be used for the voltage sensing function,^[11] but one essential feature is that it must be imbedded in the cell membrane. Usually, the voltage-sensing domain (VSD) of a GEVI spans across the membrane, and is connected to the fluorescent protein (FP). However, it is not necessary that sensing and reporting must happen in different structures - see, for example, the Archons.

By structure, GEVIs can be classified into four categories based on the current findings: (1) GEVIs contain a fluorescent protein FRET pair, e.g. VSFP1, (2) Single opsin GEVIs, e.g. Arch, (3) Opsin-FP FRET pair GEVIs, e.g. MacQ-mCitrine, (4) single FP with special types of voltage sensing domains, e.g. ASAP1. A majority of GEVIs are based on the Ciona intestinalis voltage sensitive phosphatase (Ci-VSP or Ci-VSD (domain)), which was discovered in 2005 from the genomic survey of the organism.^[12] Some GEVIs may have similar components, but in different positions. For example, ASAP1 and ArcLight both use a VSD and one FP, but the FP of ASAP1 is on the outside of the cell whereas that of ArcLight is on the inside, and the two FPs of VSFP-Butterfly are separated by the VSD, while the two FPs of Mermaid are relatively close to each other.

1. ^↑Names in italic denote GEVIs not named.

A GEVI can be evaluated by its many characteristics. These traits can be classified into two categories: performance and compatibility. The performance properties include brightness, photostability, sensitivity, kinetics (speed), linearity of response, etc., while the compatibility properties cover toxicity (phototoxicity), plasma membrane localization, adaptability of deep-tissue imaging, etc.^[45]

Different types of GEVIs are being developed in many biological or physiological research areas. Unlike earlier voltage detecting methods like electrode-based electrophysiological recordings or voltage sensitive dyes, GEVIs can be expressed stably, and can be targeted to particular cell types. GEVIs have subcellular spatial resolution^[46] and temporal resolution as low as 0.2 milliseconds, at least an order of magnitude faster than calcium imaging. This allows for spike detection fidelity comparable to electrode-based electrophysiology but without the invasiveness.^[33] Researchers have used them to probe neural communications of an intact brain (of Drosophila^[47] or mouse^[48]), electrical spiking of bacteria (E. coli^[21]), and human stem-cell derived cardiomyocyte.^[49][50]

Conversely, any form of voltage indication has inherent limitations.^[51] Imaging must be fast, or short voltage excursions will be missed. This means fewer photons per image exposure. Next, brightness per cell is inherently lower than calcium indicators, as about a 30-fold fewer voltage indicators can fit in the membrane compared to cytosolic calcium indicators.