Electron Paramagnetic Resonance Parameters of Semiquinone Bioradicals

Electron Paramagnetic Resonance (EPR) spectroscopy is a standard method of detecting and characterising paramagneticspecies. In recent years, theoretical analyses of EPR parameters have become an important complement to experimental EPR, as improvements in method and increases in computational power have expanded the ability of theory to predict, analyse and confirm xperimental results.

Semiquinone radical anions (C6 R4 O2 •-) are an important class of bioradicals which serve as electron transport agents in photosynthesis and respiration in all living things. They have been extensively characteried by EPR studies, and theoretical analyses of their EPR parameters go back decades. While initially concerned with deriving electronic spin density from experimental data, theoretical studies are now able to calculate the electronic orbitals ab initio and predict experimental data from them, with reasonable accuracy when dealing with main-group radicals. Most calculations on semiquinone radical anions have been limited to considering only the immediate environment of the molecule (the first solvation sphere, if in solution), and the effects of thermal motion have been ignored or estimated in a rough manner. These are approximations frequently made in computational chemistry, due to the computational costs involved in doing otherwise. Recently, however, supercomputing power has been used to simulate semiquinone behaviour more realistically: both by including a larger and more realistic solvation environment, and by explicitly calculating the motion of the system over time (a technique termed “Molecular Dynamics”, or MD) at a quantum mechanical level of theory (Density Functional Theory, or DFT).

The system thus treated was benzosemiquinone radical anion (C6 R4 O2 •-, or BQ•-), the simplest, most prototypical of the semiquinone bioradicals, included with 60 water molecules in a periodically- repeating box (see Figure 1). The interaction of BQ with solvent was closely studied, as both hydrogen bonding and the presence of a dielectric medium strongly affect the EPR parameters. Significantly, hydrogen bonding to BQ•- is more extensive than was believed from smaller calculations. The MD simulations confirmed the theoretical prediction, and tentative experimental indication, of “T-stacking” hydrogen bonding to the C6 ring, as these were found to occur frequently. The EPR parameters, both electronic g-tensors and hyperfine coupling data (which parameterize the interaction between electronic and nuclear spins), were calculated using DFT-based methods. By varying which water molecules were included in the calculations (e.g. none, or only those H-bonded to BQ, or only those within 4 Å, and so forth) and comparing the results, it was also possible to elucidate the effects of different effects on the EPR parameters. For instance, the effects of T-stacked hydrogen bonding alone, or of hydrogen bonding to oxygen, or both together, could be identified. The effects of the bulk solvent beyond the first solvation sphere were found to be significant, both for g-tensors and for hyperfine data.

Simulating the behaviour of BQ•- over time allowed, for the first time, an assessment of how thermal motion affects the EPR data. One notable result was that several parameters display marked short-range (~30 fs) oscillations (seen in the plot of the x-component of the g-tensor in Figure 2). This arises because of the sensitivity of numerous parameters to the C-O bondlength; these periodic oscillations reflect the C-O bond-stretching vibrational motion. Thermal motion also has a small overall effect on the timeaveraged EPR data, which can be seen by comparison with the results of static calculations. Together with the information on the solvation sphere, this should allow more informed use of static calculations using DFT methods to predict and analyse semiquinone EPR in the future. The next step is to extend this methodology to biologically important semiquinones, such as ubi- or plastosemiquinone. These are considerably larger than benzosemiquinone, and have long hydrocarbon side chains which make them lipophilic enough to traverse lipid membranes. Work has begun on simulating ubisemiquinone, a ubiquitous bioradical important in both respiration and photosystem II. Ubisemiquinone radical anion, or UQ•-, has an additional level of complexity in its behaviour due to the presence of methoxy side groups: these also attract hydrogen bonds, and their orientation affects the EPR parameters significantly.

These calculations were performed on the SR-8000 machine of the Leibniz-Rechenzentrum. The MD was performed with the CPMD code, a fast and wellparallelized code developed by J. Hutter et al. at IBM Zürich and the Max-Planck- Institute at Stuttgart. Property calculations were carried out using the Turbomole code of R. Ahlrichs et al. at the University of Karlsruhe, and the MAGReSpect code of V. Malkin et al. at the University of Würzburg and the Slovak Academy of Science in Bratislava.

• James Asher

Anorganische Chemie, Julius-Maximilians- Universität Würzburg

  top