Engineering of a NIR fluorescent protein for live-cell nanoscopy
von Florian Habenstein
Datum der mündl. Prüfung:2020-09-03
Betreuer:Prof. Dr. Stefan Jakobs
Gutachter:Prof. Dr. Dirk Schwarzer
Gutachter:Prof. Dr. Stefan Hell
EnglischSince its first use as a genetically encoded fluorescent marker in 1994, GFP and its homologues have fundamentally revolutionized live-cell fluorescence imaging and became an essential tool for biomedical research. Many fluorescent proteins have been engineered from GFP-like proteins with excitation and emission covering the entirety of the visible spectrum. However, despite substantial efforts it was not possible to reach the near infrared (NIR) spectral region beyond 650nm with a GFP derived fluorescent protein. The NIR spectral region between 650 to 900nm is especially suited for live-cell deep-tissue fluorescence imaging due to reduced autofluorescence, light scattering and absorption of tissue in this spectral region. Additionally, photo toxicity of NIR light is reduced compared to UV/vis light. Recently, utilization of engineered bacterial phytochrome variants opened the NIR spectral region for fluorescence microscopy with genetically encoded fluorescent markers. Bacterial phytochromes gain their unique optical properties by incorporating the external chromophore and heme degradation product biliverdin which is ubiquitous in mammalian cells. A variety of bacterial phytochrome based fluorescent proteins has been successfully applied in various fluorescence microscopy techniques. However, to date all engineered bacterial phytochrome variants absorbing and emitting beyond 650nm suffer from a short fluorescence lifetime and a low fluorescence quantum yield, limiting their potential for fluorescence microscopy. In this work, an automated fluorescence lifetime screening microscope was built and applied to increase the fluorescence lifetime and quantum yield of the engineered fluorescent protein miRFP703 via directed evolution. The final protein variant (V410) had a fluorescence lifetime of 1.1 ns and a fluorescence quantum yield of 21%. With this it is the brightest NIR fluorescent protein described to date. V410 exhibited a good pH stability and a high extinction coefficient. In live-cell fluorescence microscopy, V410 performed well as fusion tag for various cellular structures. With STED microscopy, resolutions well beyond 80nm down to 40nm were measured on endogenously tagged vimentin filaments. In consecutive confocal and STED recordings, with 1000 and 100 consecutive frames, respectively, V410 demonstrated superior photo stability. Utilizing the fluorescence lifetime difference between V410 and the template miRFP703 of approximately 400 ps, two-color fluorescence lifetime confocal and STED imaging was performed entirely in the NIR spectral region with a single excitation beam at 660nm and a STED beam at 820 nm.
Keywords: protein-engineering; near-infrared; nanoscopy; STED; fluorescence lifetime; fluorescence quantum yield