Effects of Extending the Computational Model on DNA–Protein T-shaped Interactions: The Case of Adenine–Histidine Dimers
journal contributionposted on 17.11.2011, 00:00 by Lesley R. Rutledge, Lex Navarro-Whyte, Terri L. Peterson, Stacey D. Wetmore
The MP2/6-31G*(0.25) π–π or π+–π T-shaped (edge-to-face) interactions between neutral or protonated histidine and adenine were considered using computational models of varying size to determine the effects of the protein and DNA backbones on the preferred dimer structure and binding strength. The overall consequences of the backbones are reasonably subtle for the neutral adenine–histidine T-shaped dimers. Furthermore, the minor changes in the binding strengths of these dimers upon model extension arise from additional (attractive) backbone−π (bb−π) contacts and changes in the preferred π–π orientations, which is verified by the quantum theory of atoms in molecules (QTAIM). Since the binding strength of the extended dimer equals the sum of the individual backbone−π and π–π contributions, the π–π component is not appreciably affected by polarization of the ring upon inclusion of the biological backbone. In contrast, the larger effect of the backbone on the protonated histidine dimers cannot simply be predicted as the sum of changes in the π–π and bb−π components regardless of the dimer type or model. This suggests, and QTAIM qualitatively supports, that the magnitude of the π+–π contribution changes, which is likely due to alterations in the electrostatic landscape of the monomer rings upon inclusion of the biological backbone that largely affect T-shaped dimers. These findings differ from those previously reported for (neutral) π–π stacked and (metallic) cation−π interactions, which highlights the distinct properties of each (π–π, π+–π, and cation−π) classification of noncovalent interaction. Furthermore, these results emphasize the importance of considering backbone−π interactions when analyzing contacts that appear in experimental crystal structures and cautions the use of truncated models when evaluating the magnitude of the π+–π contribution present in large biological complexes.