An efficient computational method has been identified that uses B3LYP density functional theory, IEF-PCM solvation modeling with a modified UFF cavity, and Boltzmann weighting of tautomers to predict the site-specific and global pKa of DNA nucleobases and their oxidation products. The method has been used to evaluate the acidity of guanidinohydantoin (Gh) and spiroiminodihydantoin (Sp), two highly mutagenic guanine oxidation products. The trend observed for the pKa values of Gh (9.64 and 8.15) is consistent with the experimentally observed values for guanidine cation (13.7) and hydantoin (9.16). The pKa1(calc) value for deprotonation of Sp cation (Sp+ --> Sp) is very close to the experimentally observed pKa1 for 8-oxoG and is consistent with the similarity in their structures. The data suggest that the imide (N7) proton in Sp is considerably more acidic than that in Gh, possibly due to the presence of the through-space electronic effects of the carbonyl group located at C6. This difference in the acidity of Gh and Sp may be an indication of their potential toxicity and mutagenicity in vivo and remains a fertile area for experimental study.

One contribution of 8 to a Theme Issue 'Density functional theory across chemistry, physics and biology' .

No abstract

Peptide Nucleic Acids (PNAs) can efficiently target DNA or RNA acting as chemical tools for gene regulation. Their backbone modification and functionalization is often used to increase the affinity for a particular sequence improving selectivity. The understanding of the trading forces that lead the single strand PNA to bind the DNA or RNA sequence is preparatory for any further rational design, but a clear and unique description of this process is still not complete. In this paper we report further insights into this subject, by a computational investigation aiming at the characterization of the conformations of a single strand PNA and how these can be correlated to its capability in binding DNA/RNA. Employing Metadynamics we were able to better define conformational pre-organizations of the single strand PNA and γ-modified PNA otherwise unrevealed through classical molecular dynamics. Our simulations driven on backbone modified PNAs lead to the conclusion that this γ-functionalization affects the single strand preorganization and targeting properties to the DNA/RNA, in agreement with circular dichroism (CD) spectra obtained for this class of compounds. MD simulations on PNA:RNA dissociation and association mechanisms allowed to reveal the critical role of central bases and preorganization in the binding process.