Although the rhodopsins contain the same chromophore, a protonated retinal Schiff base, the absorption maximum of the protein-bound chromphore varies over an extreme range from 345 nm (ultraviolett) to 610 nm (red). This provides organisms with the ability to detect color, on the one hand, and the possibility to adapt the pigments to be most sensitive to the predominating light conditions of the habitat and respective function on the other hand. The wavelength regulation or color tuning problem means therefore at least at a molecular level to identify the mechanisms that determine the wavelength pigments absorb at. The unique protein-chromophore interaction in each pigment play a crucial role in this regard. Furthermore, the photoisomerization in the protein differs significantly from that in solution, concerning its velocity, the reaction coordinate and quantum yield.

In this project, we investigate the specific interactions of the proteins with the chromophore in its electronically excited state to understand the mechanisms that determine the absorption maximum. The influence of steric and electrostatic interactions of amino acids on the optical properties of the chromophore as well as the spectral changes due to rearrangement of the protein during the photocycle are investigated. This involves the study of different rhodopsins, cone pigments, archaeal rhodopsins, and photo-intermediates thereof, but includes also methodologic work that aims at improving the predictive power of the calculated shifts in the absorption maximum.

Methodological developments include the implementation of semi-empirical excited-state methods based on TD-DFTB and OM2/MRCI, charge and polarization models for proteins, and QM/MM coupling schemes.

A broad range of QM methods, including state-of-the-art ab initio multi-reference approaches, has been assessed concerning the response of the retinal geometry and excitation energy to the steric and electrostatic influence of the environment. Rigorous deficiencies of TDDFT regarding the description of intra-molecular charge transfer have been documented as well as the systematic errors of various ab initio methods. Based on this experience, we developed a reliable computational hybrid scheme, combining the accurate DFTB description of retinal´s geometry in the electronic ground state with the balanced incorporation of electron correlation in the SORCI and OM2/MRCI methods.

Further, the influence of protein polarization, dispersion, and charge-transfer on the absorption maximum are studied using extended QM/MM, QM/QM/MM, and polarizable protein force fields.

Color Tuning in Archaeal Rhodopsins. In the first subproject the mechanism of color tuning in the rhodopsin family of proteins has been studied by comparing the optical properties of the light-driven proton pump bacteriorhodopsin (bR) and the light detector sensory rhodopsin II (sRII). Despite a high structural similarity, the maximal absorption is blue-shifted from 568 nm in bR to 497 nm in sRII. The molecular mechanism of this shift is still a matter of debate, and its clarification sheds light onto the general mechanisms of color tuning in retinal proteins. The calculations employ a combined quantum mechanical/molecular mechanical (QM/MM) technique, using  a DFT-based method for ground state properties and the semiempirical OM2/MRCI method and ab initio SORCI method for excited state calculations. The high efficiency of the methodology has allowed us to study a wide variety of aspects including dynamical effects. The absorption shift as well as various mutation experiments and vibrational properties have been successfully reproduced. Our results indicate that several sources contribute to the spectral shift between bR and sRII. The main factors are the counterion region at the extracellular side of retinal and the amino acid composition of the binding pocket. Our analysis allows a distinction and identification of the different effects in detail and leads to a clear picture of the mechanism of color tuning, which is in good agreement with available experimental data.

Color Tuning in Cone Visual Pigments. In the second subproject the mechanism of color tuning in cone pigments is studied using the same quantum mechanical/molecular mechanical (QM/MM) methods as described in the previous paragraph (DFTB/CHARMM, OM2/MRCI, SORCI). Since crystal structures are not available for the cone pigments, we use the strong similarity of the cone pigments with rhodopsin and start from the crystal structure of bovine rhodopsin to investigate the effect of single and multiple mutations on the absorption maximum. The maximal absorption is blue shifted from 498 nm in Rh to 420 nm in the blue cone pigment. Green and red cone pigment are red-shifted to 534 and 564 nm. The molecular mechanism of this shift is still a matter of debate and its clarification sheds light onto the general mechanisms of color tuning in retinal proteins. Our  methodology allows us to investigate the role of different factors on the excitation energy. In particular, we focus on the color tuning mechanisms between rhodopsin and the blue cone pigment. Single amino acids  are mutated to investigate the influence on the absorption shift. A mutation of only 10 residues in bovine rhodopsin results in nearly the full blueshift in good agreement with experimental results.