Uncovering the Principles of Light-Switchable Protein Design through Genome-Wide Interrogation of Light Oxygen Voltage Domain Insertion Sites in Yeast

Optogenetic tools have transformed modern biology by massively increasing the level of spatiotemporal precision and control over cellular systems. Indeed, with their recent application to metabolic engineering, bacterial growth patterning,  stem cell differentiation, and even the development of photo-programmable biomaterials, optogenetic tools are poised to revolutionize biotechnology as well. However, the development and optimization of these bespoke photo-switchable proteins has proven to be inefficient­—more of an art than a science. To overcome this obstacle, this project seeks to utilize a high-throughput CRISPR-based genome engineering technology, called MAGESTIC after its ability to generate Multiplexed Accurate Genome Editing with Short, Trackable, Integrated Cellular barcodes. This will enable us to construct, and then rank in efficacy, a library of >10⁴ photo-switchable proteins in S. cerevisiae (yeast). We then will leverage this library as training data to construct a neural network to predict future protein insertion sites. By combining these approaches, we aim to develop a high-throughput pipeline for constructing synthetic, light-responsive proteins as well as a design framework for engineering photo-responsive proteins from any organism. Thus, we aim to unlock the full potential of photo-programmable proteins for applications ranging from synthetic biology to materials science.

The opto-tools and design algorithms proposed here have potential applications to the development of synthetic programmable organisms for fuel production, chemical and biological contaminant sensing, on-demand development of medicines, and bio-derived self-healing materials.