Design of dyes for bioimaging studies
Techniques for imaging live biological cells are becoming increasingly sophisticated and would benefit tremendously from a broader array of dyes, such that one
could monitor multiple targets within the same cell. Recent advances in combinatorial chemistry allow one to first design a dye with desirable photophysical properties
and then identify a small amino acid sequence that can serve as the binding region for the dye. We continue designing such dyes with desirable fluorescence properties.
Last year, we published our results in JACS showing that semiempirical quantum chemical modeling of the torsional potential can account for the observed fluorescence properties
of fluorine substituted thiazole orange dyes: the computed barrier heights for twisting between a planar emissive structure and a dark non-planar structure account for the relative
quantum yields in solution.
This year, we stayed focused on the effects of electronic substituents on the fluorogenic properties of thiazole orange. We developed a general computational
scheme to identify dye candidates that simultaneously meet multiple design criteria. Some of the design criteria are fairly subtle, including especially the excited-state barrier
that determines whether the molecule will remain planar, and emissive, in the excited state or twist into a dark global minimum.
We recently found that the electronic substituent
effects are additive, even for these rather subtle features of the excited electronic surface.
Additivity made a computational search over millions of candidate dyes feasible.
Furthermore, we've also shown that additivity can apply even when the substituent effects are complex by other measures. This indicates that our work may have broader implications
for physical organic chemistry. The manuscript featuring this work is published in JPC and can be found
here .
In addition to the substituent effects, we also considered use of steric groups to enhance fluorogenicity of the dye based on thiazole orange. In order to simultaneously track
multiple biomolecules in vivo, we need fluorogens that span a range of wavelengths. One possible strategy is to increase the length of the bridge between the heterocycles.
As expected, increasing the length of the bridge shifts the absorption wavelength. However, this shift is accompanied by a significant decrease in fluorogenicity because the dye
fluoresces even in solution. Our calculations suggest that this is due to a large increase in the excited-state barrier between the planar and twisted geometries. The strategy, here,
is substitution of steric groups at relevant positions on heterocyclic rings of thiazole orange derivatives to modify the excited-state surface such that the excited-state barrier
doesn't drastically increase upon lengthening the bridge, and thereby enhance the fluorogenicity. We found that steric effects are similar in ground and excited state, and our
computations suggest that the steric effects destabilize the planar structure relative to that of the twisted structure, which in turn lowers the excited-state barrier. The drop
in the barrier height flattens the excited-state torsional potential such that the molecule preferentially takes the non-emissive pathway, in solution, but when bound to a binding element,
the environmental constraints will overcome the steric effects and keep the dyes fluorogenic. A manuscript featuring this work is available and can be seen upon request.