In optogenetic experiments aiming at controlling neuronal activity light-sensitive molecules such as channelrhodopsins or proton pumps are activated through application of light from a physical light source, such as arc lamps, lasers or LEDs. An interesting extension of such experiments could be achieved by controlling light-sensing molecules with a biological light source. Luciferases are such light producing proteins; they are enzymes, used by a variety of bioluminescent organisms, which produce light by oxidizing a substrate molecule. Replacing the physical light source with a biological one, i.e. a light-emitting protein, should allow non-invasive activation of light-sensitive molecules. In addition, the advantage of genetically encoding both light production and light sensing introduces unique combinatorial possibilities.
Using bioluminescence in combination with optogenetic actuators hinges on the ability of luciferase to activate these actuators. In a recent paper published in PLoS ONE, Berglund and colleagues demonstrate that this is indeed possible. In this study the authors engineered fusion proteins of a luciferase from the marine copepod Gaussia princeps (GLuc) and channelrhodopsins, creating luminescent opsins, or “luminopsins”. Their study brings the proof-of-concept that light produced upon supplying the luciferase with its substrate, coelenterazine (CTZ), can activate the fused channelrhodopsin, thereby modulating neuronal activity.
Because luminopsins can be activated both optically by light and chemically by administration of CTZ, this combined approach essentially bestows chemical sensitivity on optogenetic actuators. In addition to permitting fast control of neuronal firing by light like native opsins do, luminopsins allow chronic and non-invasive control of entire neuronal population through chemical activation. A unique advantage of luminopsins is that activation of the opsin can be simultaneously documented as luminescence. Bioluminescence induced by CTZ application activates neurons; at the same time, it can be localized and quantified by optical imaging, allowing identification of the neurons that are being activated as well as the extent of their activation.
Genetically encoding the effector and the actuator allows unique possibilities. The use of stronger luciferase variants and more sensitive opsins might allow direct expression of each component by different promoters. This would permit reconstitution of luminopsins in neuronal subpopulations defined by the combinatorial expression of two genes. This might also allow the reconstitution of luminopsins across synapses, analogous to GRASP. Finally, combining opsins and bioluminescent proteins might allow gating opsin activation by neuronal activity itself. This could be done for example by using a calcium-sensitive version of the light-emitting component (such as aequorin), implementing a neuronal activation-controlled light switch.