Protein Response Effects on Cofactor Excitation Energies from First Principles: Augmenting Subsystem Time-Dependent Density-Functional Theory with Many-Body Expansion Techniques

Abstract

We investigate the possibility of describing protein response effects on a chromophore excitation by means of subsystem time-dependent density-functional theory (sTDDFT) in combination with a many-body expansion (MBE) approach. While sTDDFT is in principle intrinsically able to include such contributions, addressing cofactor excitations in protein models or entire proteins with full environment-response treatments is currently out of reach. Taking different model structures of the green fluorescent protein (GFP) and bovine rhodopsin as examples, we demonstrate that an embedded-MBE approach based on sTDDFT in its simplest version leads to a good agreement of the predicted protein response effect already at second order. To reproduce reference response effects from nonsubsystem TDDFT calculations quantitatively (error ≤ 5%), however, a third-or even fourth-order MBE may be required. For the latter case, we explore a selective inclusion of fourth-order terms that drastically reduces the computational burden. In addition, we demonstrate how this sTDDFT-MBE treatment can be utilized as an analysis tool to identify residues with dominant response contributions. This, in turn, can be employed to arrive at smaller structural models for light-absorbing proteins, which still feature all of the main characteristics in terms of photoresponse properties.

The Seamless Connection of Local and Collective Excited States in Subsystem Time-Dependent Density Functional Theory

“…This methodology has led to an accurate description of van der Waals interactions between weakly interacting subsystems , and the determination of C 6 coefficients for molecules adsorbed on surfaces . In addition, a many-body expansion based on FDEu-TDDFT has been used to dissect protein response effects on cofactor excitation energies …”

“…132 In addition, a many-body expansion based on FDEu-TDDFT has been used to dissect protein response effects on cofactor excitation energies. 133 As mentioned earlier, sTDDFT in combination with NAKE functionals offers an efficient and accurate way for the determination and analysis of excitation energies and response properties associated with each subsystem (FDEu-TDDFT) or delocalized over the entire system (FDEc-TDDFT). Nevertheless, sTDDFT in combination with approximated local and semilocal nonadditive functionals is incapable of describing CT excitations, even in principle.…”

Rhodopsins: An Excitingly Versatile Protein Species for Research, Development and Creative Engineering

“…This further inspired detailed analysis with modified and/or labeled retinals ( 13 C, 2 H, F) and protein residues ( 13 C, 15 N, F, azido, spin labels) using fluorescence, vibrational, EPR and NMR spectroscopy ( Table 5 ). This profited from as well as steered development of sophisticated theoretical and in-silico procedures, like DFT, QM/MM and molecular dynamics ( Altun et al, 2008 ; Collette et al, 2018 ; Del Carmen Marín et al, 2019b ; Dokukina et al, 2019 ; Pieri et al, 2019 ; Shao et al, 2020 ; Nikolaev et al, 2021 ; Scholz and Neugebauer, 2021 ; Shen et al, 2021 ; Pedraza-González et al, 2022 ). All these elements have already profoundly deepened our insight into the structure and mechanism of bovine rod rhodopsin and bacteriorhodopsin, the frontrunners of type-2 and -1, respectively.…”

Massively Parallel Fragment-Based Quantum Chemistry for Large Molecular Systems: The Serestipy Software

“…In particular, we will employ subsystem time-dependent densityfunctional theory [32][33][34][35] (subsystem TDDFT) with parallel FDE calculations as a reference [36,37] to efficiently obtain excited states. Besides the presented example, this massively-parallel framework could be employed to efficiently perform tasks such as embarrassingly-parallel potential-energy surface constructions [38][39][40][41][42][43], parallel mode-and intensity-tracking or general semi-numerical frequency calculations [44][45][46][47], or many-body expansion calculations [48][49][50][51][52].…”