Fluorescence imaging is an indispensable tool to understand how biological networks operate in the complex environment of the living cell. Small-molecule fluorescent dyes offer many advantages including their high modularity, favorable photophysical properties and wide spectral range. Polymethine dyes are among the most used fluorophores in single-molecule localization microscopy and in vivo imaging applications because of their high extinction coefficients and tunable emission wavelength ranging from the green to the shortwave infrared region. Their use for live-cell imaging, however, has been hampered by high background fluorescence and low membrane permeability.
In this thesis, a general strategy to transform polymethine compounds into fluorogenic derivatives is presented. These dyes are non-fluorescent in solution but become highly fluorescent when bound to their specific intracellular target. In order to introduce fluorogenicity, we incorporated a 5-exo-trig cyclization into the polymethine dye scaffold, in contrast to previously attempted 5-endo-trig ring closures. We demonstrate that a 5 exo¬-trig Cy5 probe conjugated to the self-labeling protein tag SNAPf-tag shows high fluorogenicity and a bright and specific fluorescence signal in live HeLa cells, whereas the corresponding 5-endo-trig probe showed low cell permeability, low fluorogenicity and high unspecific fluorescence signal. We further illustrate the versatility of this 5-exo-trig cyclization approach by creating a spontaneously blinking Cy5 dye as well as no-wash, turn-on polymethine dyes with emissions across the visible and near-infrared spectrum. These probes are not only compatible with self-labeling protein tags but also with small-molecule targeting ligands such as jasplakinolide for actin labeling and Hoechst 33258 dsDNA/nucleus labeling. Furthermore, these probes can be combined with rhodamine-based dyes in multi-color and fluorescence lifetime multiplexing imaging.
Since the reasons for ring opening and subsequent fluorescence turn-on were unclear, molecular mechanics simulations and protein mutagenesis were performed. These experiments indicate that ring opening is driven by specific interactions of the ring opening/closing group with specific amino acid residues at the protein/target surface. Furthermore, protein engineering was also carried out with the goal to improve the fluorescence turn-on upon SNAPf-tag binding, and to create a novel SNAPf-tag variant with improved crystallization propensity. Preliminary protein mutagenesis experiments resulted in a 2.7-fold improved fluorescence turn-on of our near-infrared Cy7 derivative upon binding to a novel SNAPf-tag variant, highlighting the immense potential of protein engineering as an avenue to improve fluorogenicity.
The fusion of a protein of interest to a fluorescent protein or self-labeling protein tag can perturb the folding, function, localization, and interaction of the protein of interest. Therefore, we also explored the possibility of a fluorescence turn-on upon binding to a short tetraserine peptide motif. Yeast display was carried out to generate a randomized peptide library containing a tetraserine motif, and fluorescence-activated cell-sorting will be carried out in the future to find a peptide capable of inducing a fluorescence turn-on in bis-boronic acid modified fluorogenic polymethine dyes.