Abstract: Bioluminescent enzymes (luciferases) are among the most popular reporters for illuminating biological processes in vivo. Luciferases emit light by catalyzing the oxidation of small molecule luciferins. Since no excitation light is needed, there is virtually no background signal. Thus, bioluminescence imaging (BLI) is extremely sensitive and well-suited for applications in tissues and whole organisms.

While powerful, bioluminescence has been historically limited in scope. Traditional luciferase-luciferin pairs have been slow to transition to imaging more than one target at a time, owing to a lack of distinguishable probes and detection methods. Naturally existing luciferases process similar substrates and exhibit broad emission spectra.  Thus, they cannot be readily used in tandem. Even fewer probes are both distinguishable and produce significant amounts of tissue penetrant light (>650 nm). Current protocols for visualizing multiple targets are also time intensive, requiring multiple hours (if not days). Consequently, bioluminescence is rarely used for multiplexed imaging of dynamic events. Existing bioluminescent tools are also too dim to register on standard microscopes, precluding further studies of biomolecule localization and other subcellular features. Such issues have stymied efforts to visualize complex networks – like the interplay between immune function and metabolism – for decades.

To address these limitations, I developed novel small molecules and enzymes expand to the bioluminescent toolkit for multicomponent imaging in deep tissues. In one area, I synthesized pi-extended luciferins featuring intervening aryl moieties. The flexible frameworks provided long wavelengths of emission only when paired with luciferases capable of enforcing planarity. These analogs were initially poor substrates for the native luciferase. I subsequently evolved complementary luciferases for the analogs by screening a combination of rational and in silico designed libraries. The evolved luciferases displayed improved brightness, red-shifted emission, and substrate specificity. A similar engineering strategy was used to create a second class of NIR-emitting probes based on fluorescent coumarin scaffolds. Lastly, I demonstrated the engineered luciferase-luciferin probes can be used in tandem with other bioluminescent tools for multicomponent applications.

In addition to crafting designer probes, I also developed new imaging platforms to broaden the scope of bioluminescence imaging. This work involved a combination of sequential imaging with linear unmixing, enabling rapid readouts on combinations of luciferase reporters. Multiple bioluminescent signals could be registered on the minutes-to-hours time scale, a vast improvement over conventional protocols. I also worked to develop a new microscopy method that merges bioluminescence with phasor analysis, an optical method commonly used to differentiate spectrally similar fluorophores. Bioluminescent phasor achieved high resolution, multicomponent imaging with probes that cannot be separated by wavelength alone. This technology also leverages widely available optical components and user-friendly probes, making it easily accessible to non-specialists. These new detection platforms and probes will broden the scope of bioluminescence imaging and further transform the way we see and interrogate biological processes.