A generalizable method for controlling protein activities with light

If we could control the activity of any intracellular protein of interest with light, it would revolutionize how we study biology and engineer gene- or cell-based therapies. For instance, optical protein control would allow testing of protein functions in cells with high spatiotemporal precision, so we can observe how protein activation at specific times and places affects downstream biochemical reactions and cellular behavior, without secondary effects from chronic or ubiquitous activation. Optical protein control would also allow four-dimensional regulation of cellular proliferation, survival, or differentiation in tissues or animals for basic research or therapy. Thus controlling a wide variety of proteins with light is one of the major long-term goals of optogenetics.

Meeting this challenge will require the development of easily generalizable methods for optical control of protein activity. Considerable efforts have been made to adapt natural light-responsive signaling proteins to regulate specific proteins of interest. In recent years, phytochrome-PIF, cryptochrome-CIB, and FKF-gigantea light-dependent interactions from plants have been adapted to control heterodimerization of proteins in mammalian cells. The LOV domain from phototropin has also been used to create single-chain photoactivatable proteins via light-modulated allosteric mechanisms. However, these methods have certain disadvantages. Light-controlled heterodimerization can not effectively control activities of many protein types, while making light-controllable proteins using LOV domains requires extensive construction and screening. The chromophore for phytochromes is a plant compound that needs to be supplied, while the flavin chromophores used by the other systems can produce toxic reactive oxygen species upon illumination. Due to these limitations, these previous methods have been successfully used to control only a few proteins.

In a recent paper in Science, Lin et al. make the unexpected finding that a tetrameric variant of Dronpa, a photoswitchable fluorescent protein of the GFP superfamily, undergoes dissociation upon exposure to cyan light. They then created light-inducible proteins (FLIPs) by fusing Dronpa domains to the N- and C-termini of proteins of interest. They found that these proteins were inactive at baseline, presumably due to steric hindrance to the active site by the Dronpa tetramer, and could be activated by cyan light. The design was generalized to two different protein domains, a Dbl-family guanine nucleotide exchange factor, and a protease. Uniquely, the process of protein uncaging could be tracked as it occurs using the off-switching of Dronpa fluorescence. This should help experimenters find the right light dosage to use when testing constructs for effectiveness and when actually using validated constructs in different expeirmental systems. Perhaps the most interesting aspect of this project is that it forces us to rethink the very nature of fluorescent proteins — rather than considering them strictly as imaging modules, as we have become accustomed to, it may be useful to also think of them similarly to other natural chromophore-containing domains, as being capable of translating the energy of photons to control or drive cellular functions.

Model of a light-inducible intersectin protein in the basal state made by fusing Dronpa fluorescent protein domains (green) to each end of the Dbl-homology (DH) domain of intersectin (blue). The tetramerization of Dronpa domains effectively cages the catalytic activity of the DH domain. Illumination leads to dissociation of the Dronpa tetramer and allows the DH domain to interact with its downstream target, the Rho-family GTPase Cdc42. From Zhou et al. 2012.

This entry was posted in Uncategorized and tagged . Bookmark the permalink.
Add Comment Register

Leave a Reply