Bridging Evolution and Design: Mapping the Diversity of LOV Photosensors

Light-sensitive proteins allow organisms to perceive and respond to their environment, and have diversified over billions of years. Among these, Light-Oxygen-Voltage (LOV) domains are widely distributed photosensors that control diverse physiological processes. Despite their broad biological roles and increasing use in optogenetics, the functional diversity of natural LOV domains and the evolutionary constraints shaping their dynamics remain poorly resolved. A key unresolved problem is how evolution modulates the timescales and efficiencies of LOV photocycles and how this kinetic flexibility relates to biological function. Here we systematically map the photodynamics of 21 natural LOV domains - including 18 previously uncharacterized variants - and one de novo photosensor generated by artificial intelligence-guided protein design. We uncover an exceptional kinetic diversity spanning picoseconds to days and identify distinct functional classes within the LOV family. These patterns holistically reveal that billion years of evolutionary adaptation led to branching photocycle kinetics, matching physiological requirements. Moreover, by extending the natural catalog of LOV photosensors with a de novo designed LOV variant, we demonstrate how computational protein design can access new biophysical niches. This work expands the optogenetic toolkit and offers a framework to dissect and harness the evolutionary design principles of light-responsive proteins.