Proton Coupled Electron Transfer in Biomolecules

Computational description of PCET
We can get a complete picture of PCET by combining several different simulation approaches: 
Molecular mechanics methods focus on the overall protein environment and how it affects the donor and acceptor molecules. Quantum chemistry methods, on the other hand, precisely describe the actual electron and proton transfers. Together, these tools help us understand the complex interactions behind PCET.

How we do it
One particularly useful approach is constructing free energy surfaces (FES) of the PCET. Knowing FES makes it possible to infer the mechanism of the reaction (from the minimum energy pathway in the FES), the reaction energies as well as the reaction rates (the latter, from the barrier heights in the FES). To generate a FES, we need to define one or more so-called collective variables (CV) prior to running simulations. We use two CVs in the studies of PCET:

• one to describe the proton transfer (PT): the difference of distances of the proton being
transferred, from the proton donor and the proton acceptor; and

• another to describe the electron transfer (ET): the difference of summed Mulliken charges of the two amino acid residues participating in the PCET

Technical details
Much like many other biochemical reactions, PCET cannot be simulated with free molecular dynamics because that would be too slow: It would take extremely long simulation times to generate the required FES. Instead, we rely on the so-called extended sampling approaches, which make it possible to generate the FES much faster. Specifically, we use well-tempered metadynamics simulation, which involves a potential energy function that is modified in such a way that certain rarely occurring events are accelerated. In our molecular dynamics simulation running with this “biased” potential energy function, the PCET reactions take place frequently enough to generate reliable FES.

Modification of the potential energy function consists of adding a function of the CV for PT, as this turns out to accelerate not only the PT but also the ET reaction. After such a simulation is finished, we apply a reweighting technique to recover the unbiased probability distributions for both CVs, which directly provide the desired FES of the PCET, and can be interpreted in terms of mechanism and rate of the reaction.

Specific molecular complexes that we investigate
We study PCET in several different biomolecular complexes: for one thing, to simplify the process and accelerate the calculation, there are a few model systems such as a β-hairpin peptide inspired by the photosystem II and an α-helical protein. Then, we investigate some of the genuine protein complexes that support PCET, such as a ribonucleotide reductase (RNR) and a class II photolyase (MmCPDII).

Biomimetic peptides
What we have discovered for the PCET in the model peptides is that the exposure of the aromatic sidechains to the solvent influences the reaction mechanism significantly. The surrounding environment can reshape the free energy landscape and adjust the transfer barrier heights. For example, in the β-hairpin peptide, the PCET mechanism varies depending on whether the reaction occurs between a tyrosyl radical and a histidine, tryptophan, or tyrosine residue. We also observed that the relative orientation of the amino-acid residues involved plays a critical role in electron transfer efficiency. Additionally, we investigated an α-helical radical maquette, where depending on the environment (protein or water) the transfer mechanism and barrier height were altered.

Similar to what we observe in biomimetic peptides, the relative geometry of the residues
involved in PCET within RNR has a significant impact on the reaction barrier height and the free energy landscape. 

Our simulations show that the energy barrier is smaller when the residues adopt a π-stacked conformation compared to a flipped conformation. In the π-stacked conformation, the electron density is delocalized along the electron transfer axis, which facilitates the reaction. In contrast, the flipped orientation results in localized electron density, suggesting a concerted reaction mechanism, in which proton and electron transfer occurs simultaneously.

Preliminary studies of genuine proteins

The relative geometry of the residues involved in the PCET reaction significantly influences the reaction barrier height and the free energy landscape, with π-stacked
conformations leading to lower barriers compared to flipped orientations.

Preliminary simulations of MmCPDII suggest that the product state – formed after the
PCET between W388•− and Y345 – is more stable than the reactant state.

In MmCPDII the product state of the PCET reaction is more stable than the reactant state. The electron density seems more delocalized in the product state.

We are exploring different proton transfer mechanisms, including direct transfer to tryptophan or mediated transfer via nearby aspartate or histidine residues.
Additionally, we are comparing the PCET mechanism in MmCPDII with that in animal-like Cryptochrome (CraCRY) to better understand the variations in transfer pathways and reaction dynamics.

REFERENCES
Sibert, R. et al. Journal of the American Chemical Society. 129. 4393–4400. 2007
Berry, B. W. et al. Proceedings of the National Academy of Sciences. 109. 9739–9743. 2012
Tiwary, P. et al. The Journal of Physical Chemistry B. 119. 736–742. 2015
Gillet, N. et al. The Journal of Chemical Physics. 149. 072328. 2018
Maag, D. et al. The Journal of Chemical Physics. 158. 124107. 2023
Müller, P. et al. Chemical Science. 9. 1200–1212. 2018
Lacombat, F. et al. Journal of the American Chemical Society. 141. 13394–13409. 2019
Reinhardt, C. R. et al. Journal of the American Chemical Society. 142. 13768–13778. 2020
Zhong, J. et al. The Journal of Physical Chemistry Letters. 15. 1686–1693. 2024