On the Mechanism of the Mutagenic Action of 5-Bromouracil: A DFT Study of Uracil and 5-Bromouracil in a Water Cluster Victor I. Danilov, † Tanja van Mourik,* ,‡ Noriyuki Kurita, § Hajime Wakabayashi, § Takayuki Tsukamoto, § and Dmytro M. Hovorun † Department of Molecular and Quantum Biophysics, Institute of Molecular Biology and Genetics, National Academy of Sciences of Ukraine, 150 Zabolotny Street, KyiV-143, 03143, Ukraine, School of Chemistry, UniVersity of St Andrews, North Haugh, St. Andrews, Fife, KY16 9ST, Scotland, U.K., and Department of Knowledge-based Information Engineering, Toyohashi UniVersity of Technology, Tempaku-cho, Toyohashi 441-8580, Japan ReceiVed: December 13, 2008 Density functional theory calculations on the canonical (keto) and rare (enol) tautomeric forms of uracil and 5-bromouracil in a cluster consisting of 50 water molecules are presented. The keto form of uracil is favored over the enol tautomer in both the gas phase and solution. However, the presence of the water cluster reverses the tautomeric preference of 5-bromouracil, rendering the rare tautomeric form to be preferred over the canonical form in aqueous solution. This effect is, to a large extent, due to the more favorable water-water interactions in the cluster around 5-bromouracil and can therefore only be obtained by including explicit water-water interactions in the calculations. The mechanism of the mutagenic activity of 5-bromouridine has been under some debate over the last few decades. The canonical keto form of 5-bromouracil (5BrU) is complementary to adenine. However, this tautomeric form cannot bind with guanine. Mechanisms that have been brought forward to explain the pairing of 5BrU with guanine include keto f enol tautomerization of 5BrU, ionization (deprotonation) of 5BrU, and the involvement of “wobble” hydrogen bonds (see, for example, ref 1). Additionally, a mutagenic mechanism involving triplet exited states of U and 5BrU has been suggested. 2 A mechanism involving base pairing between the ionic form of BrU and guanine is not very likely. The reason for this is that, at interbase distances normally occurring in DNA duplex structures, such base pairs prefer nonplanar structures, exhibiting very large propeller twist angles due to the unfavorable arrangement of the oxygen atoms. The nonplanarity of these structures hinders incorporation into DNA. 3 The possible formation of wobble hydrogen bonds should not depend on the presence or absence of bromine and can therefore also be discarded as a likely cause for the mutagenicity of 5-bromo- uridine. The involvement of excited states is very unlikely as the mutation reaction is a dark reaction, that is, no light or other type of radiation is required to induce the mutations. The only remaining mutagenic mechanism, the tautomeric model of the mutagenic activity of 5BrU, has also been discarded by several authors. 1,2 The rejection of the tautomeric model was based on quantum chemical calculations that showed that solvation in water does not change the known gas-phase preference for the canonical form of 5BrU. As a result, the tautomeric model is currently not considered as a viable candidate to explain the mutagenic activity of 5BrU. However, these calculations employed continuum solvation models to describe the bulk water, which lack explicit interactions between individual water molecules and may therefore not be accurate enough to reliably obtain the tautomeric preferences of solvated 5BrU. In addition, it should be noted that Katritzky and Waring 4 showed that a bromine in the five position considerably increases the propor- tion of the enol form present in 1-methyluracil. In this Letter, we show that consideration of explicit water molecules dramati- cally alters the tautomeric preference of 5BrU. Our results therefore reinstate the tautomeric mechanism as a possible explanation of the mutagenic properties of 5BrU. To reliably predict the tautomeric preferences of U and 5BrU, one needs to treat the water-water and water-base interactions at the molecular level by quantum chemistry methods. In addition to model bulk water, a sufficiently large cluster of water molecules needs to be employed. Previous calculations on uracil and thymine complexes comprising 11 water molecules 5 and cytosine and adenine complexes comprising 14 and 16 water molecules, respectively, 6 showed that structures with a clustering of the water molecules were preferred over structures with the water molecules distributed around the central base. This is due to the increased attractive interaction between the water molecules in the clustered complexes. The results indicate that the limited-size water clusters used in these calculations do not accurately represent bulk water; in the clustered structures, the water only interacts with one face of the central base, whereas in the distributed structures, the attractive water-water interac- tion is underestimated as compared to the clustered structures and, consequently, also to bulk water. It was recently shown that a RNA base pair in a sufficiently large nanodroplet (water cluster) mimics the behavior of a base pair in bulk water. 7 For the RNA base pairs, the stability observed in bulk water was * To whom correspondence should be addressed. E-mail: tanja.vanmourik@st-andrews.ac.uk. † National Academy of Sciences of Ukraine. ‡ University of St Andrews. § Toyohashi University of Technology. 2233 10.1021/jp811007j CCC: $40.75  2009 American Chemical Society Published on Web 02/13/2009 2009, 113, 2233–2235