Engineering photostable fluorescent proteins and biosensors using transcriptomic mining and massive-throughput single-cell screening

PROJECT SUMMARY/ABSTRACT

Fluorescent proteins are ubiquitous reagents in the biomedical sciences for reporting gene expression, protein and nucleic

acid localization, cell shape, and cellular activity. However, fluorescent proteins (FPs) become progressively dimmer —

they photobleach — with repeated or prolonged illumination. Photobleaching limits multiple types of biological experiments where photostability is essential, such as single-molecule biophysics and timelapse imaging of cellular activity during

development, learning, and aging. Photobleaching often cannot simply be addressed by increasing the excitation light, as

high illumination power can induce membrane blebbing, nuclear fragmentation, alterations in the cell cycle, changes to the

concentration of intracellular calcium, and, ultimately, cell death. While over two decades of FP engineering has led to a

toolbox of bright FPs, less attention has been devoted to improving photostability because of the greater difficulty and lower throughput endured when screening for photostable FPs. Moreover, few studies have attempted to improve photophysical

properties under two-photon illumination — a method of choice for deep-tissue imaging — because of technical challenges

associated with screening under this imaging modality. The overall objective of this research proposal is, therefore, to develop and apply a color palette of bright and photostable FPs for one- and two-photon imaging in mammalian cells. Our

proposal leverages two specialized and synergistic approaches to FP discovery and engineering: (1) SPOTlight, a new all-

optical screening approach developed in Dr. St-Pierre's lab that circumvents technical hurdles and enables rapid screening

of both brightness and photostability at the single-cell level under one- and two-photon illumination; and (2) transcriptomic and metagenomic mining for novel FPs from marine invertebrates, a technique pioneered by Dr. Shaner’s lab. SPOTlight relies on light patterning technology to selectively illuminate individual cells labeled with fluorophores that can be

photoactivated from a dim to a bright state. The cells are therefore tagged with a unique fluorescence signature that can then be distinguished and retrieved using Fluorescence Activated Cell Sorting (FACS). SPOTlight thus enables screening in

dense mixed cultures with single-cell resolution, thereby eclipsing the throughput of traditional well-based approaches. Mining for novel FPs in marine invertebrate transcriptomes and metagenomes will allow us to rapidly identify and characterize hundreds of novel FPs. From this pool of new FPs, we will select the most photostable for engineering with

the SPOTlight pipeline. We will also model their structures to guide site-directed mutagenesis. We propose to leverage

these new technologies and assays to develop FPs of different colors that are bright, monomeric, and sufficiently photostable for long-term imaging experiments. We also propose to apply these new FPs to increase the photostability of genetically encoded voltage indicators (GEVIs), which are fluorescent biosensors whose brightness reports changes in voltage. While

GEVIs are proposing tools for imaging neural electrical activity with exquisite temporal resolution, they require high

illumination power for detection and typically bleach in seconds or minutes. Overall, we anticipate that this project will

produce bright and photostable fluorophores and biosensors of broad utility for illuminating cellular dynamics and that our procedures will inspire further multi-parameter engineering of imaging probes for long-term imaging.