Monomeric adenine decay dynamics influenced by the DNA environment

Abstract

We report on-the-fly surface-hopping dynamics simulations of single adenine embedded in solvated DNA oligomers, (dA)10 and (dA)10·(dT)10. Both model systems are found to decay from the S1 to the S0 state via distinct monomeric channels, on account of the strong hydrogen-bonding interactions between the Watson-Crick pair in the double-stranded oligomer. Surprisingly, the decay times (several picoseconds) for the current models are 10 times longer than those of adenine in the gas or aqueous phase, while matching one of the time constants observed experimentally. We discuss possible reasons for these longer decay times, including steric hindrance in the DNA strands, electronic effects of the environment, and the presence of other local excited-state minima. We present optimized geometries and relative energies for representative S0 and S1 minima as well as conical intersections related to the hopping events. We have also computed steady-state and time-dependent fluorescence spectra that may help understand the experimental observations.

Computational Modeling of Photoexcitation in DNA Single and Double Strands

“…3.2) may be suppressed when the base is paired with thymine (or uracil in RNA) through Watson-Crick hydrogen bonds. This prediction was confirmed in the QM/MM surface-hopping studies of a single adenine in (dA) 10 ·(dT) 10 by Lu et al [114,115]. Unlike the single-stranded (dA) 10 , the monomeric 6 S 1 channel in the double strand is completely locked, and the 2 E channel becomes dominant, since it does not require geometric deformations that perturb hydrogen bonds.…”

“…By contrast, electronic embedding considers the QM region as being immersed in a background of MM point charges (effective force-field charges), which leads to electronic QM polarization in response to the MM environment. Electronic QM/MM embedding was shown to be indispensable for correctly representing excited-state DNA systems, as it strikingly modulates the excited-state dynamics [76,114,115,144]. Some QM-only investigations on DNA excited states have employed implicit solvent models [145,146], e.g., the polarizable continuum model (PCM) [147][148][149][150][151][152][153][154][155][156][157][158].…”

“…Correspondingly, the ionization potentials of all four DNA nucleobases also increase in solvated QM/MM DNA models compared with the gas phase [217]. Thiel and coworkers [113][114][115] reported a small steady-state blue shift (0.09-0.17 eV) for a single adenine embedded in (dA) 10 and (dA) 10 ·(dT) 10 relative to the absorption energy of aqueous 9H-adenine. Although these QM/MM results reflect the solvent and environmental effects in a qualitatively correct manner, they should not be directly compared to the experimental spectra, because the calculations took into account only a single QM nucleobase (neglecting other bases as well as QM interbase interactions).…”

“…Lu et al [114,115] carried out QM/MM nonadiabatic dynamics simulations for a single adenine embedded in single-and double-stranded oligomers (dA) 10 and (dA) 10 · (dT) 10 , treating the QM adenine with the semiempirical OM2/MRCI approach. They found that the 6 S 1 and 2 E conical intersections (see above) remain the dominating decay channels, but the computed time constants for the monomeric excited-state decay increase dramatically, roughly by an order of magnitude compared with the gas or aqueous phase (~4.1-5.7 vs~0.4-0.6 ps at the same level of QM theory, see above).…”

Assessment of approximate computational methods for conical intersections and branching plane vectors in organic molecules

“…25,26 Taking advantage of its very low computational demands, numerous on-the-fly non-adiabatic molecular dynamics (MD) simulations have been performed with OM2/MRCI for electronically excited organic molecules. [27][28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43][44] Apart from a case study on adenine, 45 we are not aware of any systematic investigation on the ability of approximate quantum-chemical methods to properly describe the CIs and BPs of larger organic molecules (despite the growing number of such applications). This is the topic of the present article: we compare the results obtained from SF-TDDFT, SI-SA-REKS, and OM2/MRCI calculations with those from high-level ab initio MRCISD calculations covering a representative set of eight organic compounds with 12 CIs.…”