Computational studies of the mechanism of fluorescence enhancement and spectral tuning in arch neuronal optogenetic reporters

Several biological systems have evolved structures that can interact with light and exploit the photons energy for their biological functions. In fact, thanks to specialized organic molecules spread across all life kingdoms, life forms use light to trigger and regulate an incredible variety of different activities, including vision, photosynthesis, circadian rhythm, bioluminescence and in general a wide spectrum of biochemical processes. The common thread to all these activities is that they are invariably initiated by a photochemical reaction. Among these specialized structures, rhodopsins, membrane proteins harbouring a covalently bound retinal chromophore, play a preeminent role. In fact, while the first member of the rhodopsin family identified in the 1930s was the visual pigment of the animal retina, it was then found that this class of light-triggered proteins populates the whole spectrum of living organisms, where it promotes an astounding variety of biochemical processes, ranging from vision to cellular metabolism and sensorial function. More recently, rhodopsin became especially important in Optogenetics, a collection of neurobiology techniques where photo-sensitive proteins are used to control and visualize neural activity. The imaging of networks of interacting neurons is a particularly challenging task, since common fluorescence microscopy techniques require bright fluorescent probes localized in the neuron membrane. Ideally, such probes would also display a red-shifted maximum absorption wavelength, allowing to use excitation lasers with higher tissue penetration and reduced phototoxicity. Since rhodopsins are light-responsive membrane proteins, they are ideal candidates to report on action potentials. Archaerhodopsin-3 (Arch3) a microbial rhodopsin from Halorubrum Sodomense, was the first proposed rhodopsin-based fluorescent reporter. However, Arch3 has a fluorescence quantum yield (FQY) of ca. ~0.0001, which impairs the possibility to efficiently visualize the activity of neuronal populations. Throughout the years, several experiments of random and site-directed mutagenesis on Arch3 culminated in the discovery of variants with increased FQY, such as the Archers, the Archons, Arch5 and Arch7. These variants feature a brighter fluorescence enabling applications in imaging of acute brain slices, but also in living mammals and invertebrates. However, the FQY value are still in the range 0.001-0.01 and therefore not yet as bright as desirable. Therefore, it would be highly beneficial to develop computational tools for the high-throughput rational-design of rhodopsins with desired photochemical and/or photophysical properties. In my thesis, we provide the necessary theoretical foundations to envision the development of such tools. By constructing multiconfigurational quantum chemistry (MCQC) model of Arch3 and six of is variants, we establish a theory connecting the amino acid sequence to the increase of fluorescent brightness. We show that the observed experimental FQY trend correlates with the decrease in energy difference between the planar fluorescent emitting state and a newly characterized exotic diradical intermediate intercepted by a nearby photoisomerization channel. Investigating the molecular-level factors modulating this critical quantity, we show that the electronic structure of the retinal chromophore at the two minima is substantially different and that this is reflected by their different charge distributions. This is important because it indicates that a variation in the protein electrostatic potential that simultaneously stabilize the fluorescent state and destabilize the region of the decay channel would dramatically increase the FQY, suggesting an ideal target for microbial rhodopsins fluorescence engineering. Since all the reported Arch3 variants with increased FQY display also red-shifted absorption maximum, another appealing feature in Optogenetics, we used Arch3 as template to develop a new computational tool which allowed us to identify the critical opsin electrostatic variations that contribute to the spectral tuning. By designing an optimization procedure based on a variational protocol, we show that the “effective delocalization” of the counterion is the critical determinant of Arch3 absorption maximum. This theoretical framework agrees with the experimental findings, which heuristically identified in the counterion complex the most sensible replacements necessary to shift to the red the absorption of Arch3. Finally, using the same tool, we show that, in this family of rhodopsins, there is indeed a first-order relationship between maximum absorption wavelength and FQY