Fast Method for Excited-State Dynamics in Complex Systems and Its Application to the Photoactivation of a Blue Light Using Flavin Photoreceptor

Abstract

The excited-state dynamics of molecules embedded in complex (bio)matrices is still a challenging goal for quantum chemical models. Hybrid QM/MM models have proven to be an effective strategy, but an optimal combination of accuracy and computational cost still has to be found. Here, we present a method which combines the accuracy of a polarizable embedding QM/MM approach with the computational efficiency of an excited-state self-consistent field method. The newly implemented method is applied to the photoactivation of the blue-light-using flavin (BLUF) domain of the AppA protein. We show that the proton-coupled electron transfer (PCET) process suggested for other BLUF proteins is still valid also for AppA.

Integrated Computational Study of the Light-Activated Structure of the AppA BLUF Domain and Its Spectral Signatures

“…Our simulations confirmed that the dark-adapted state of AppA is compatible with the structure by Jung et al, containing Met106 in the binding site instead of the Trp104 . Very recently, starting from this dark-adapted state, we have simulated the photoactivation mechanism using polarizable QM/MM dynamics simulations in the excited and ground states . These simulations showed that the PCET mechanism is indeed possible for the AppA protein, suggesting a conserved mechanism among different BLUF domains.…”

“…In a recent study, we have shown that a PCET mechanism of photoactivation is possible for AppA . As shown in Figure , this mechanism results in an important change in the glutamine placed very close to the flavin, which transforms in its ZZ imidic acid tautomer.…”

QM/AMOEBA description of properties and dynamics of embedded molecules

“…Tamm–Dancoff approximation (TDA) DFT/AMOEBA dynamics have been used to study this mechanism 142 : the active site (i.e., FMN and Tyr and Gln side‐chains) was treated at the QM level, whereas the surrounding protein and water molecules were described with the AMOEBA force field. This procedure was useful to predict the LE CT transition, but not the proton transfers due to the intrinsic limitations of (TD or TDA)DFT/AMOEBA methodology.…”

“…This procedure was useful to predict the LE CT transition, but not the proton transfers due to the intrinsic limitations of (TD or TDA)DFT/AMOEBA methodology. Instead, a SCF strategy 143–145 was coupled to the QM/AMOEBA machinery, 146 to perform SCF/AMOEBA excited‐state dynamics 142 . The SCF/AMOEBA approach showed to be less computationally demanding than (TD)DFT/AMOEBA.…”

“…Instead, a ΔSCF strategy [143][144][145] was coupled to the QM/AMOEBA machinery, 146 to perform ΔSCF/AMOEBA excited-state dynamics. 142 The ΔSCF/AMOEBA approach showed to be less computationally demanding than (TD)DFT/AMOEBA. Moreover, thanks to its self-consistent formulation, ΔSCF/AMOEBA includes intrinsically the state-specific effect of the environment, which is mandatory to describe excited states with a large CT character.…”