Modeling Photoionization of Aqueous DNA and Its Components

Abstract

Radiation damage to DNA is usually considered in terms of UVA and UVB radiation. These ultraviolet rays, which are part of the solar spectrum, can indeed cause chemical lesions in DNA, triggered by photoexcitation particularly in the UVB range. Damage can, however, be also caused by higher energy radiation, which can ionize directly the DNA or its immediate surroundings, leading to indirect damage. Thanks to absorption in the atmosphere, the intensity of such ionizing radiation is negligible in the solar spectrum at the surface of Earth. Nevertheless, such an ionizing scenario can become dangerously plausible for astronauts or flight personnel, as well as for persons present at nuclear power plant accidents. On the beneficial side, ionizing radiation is employed as means for destroying the DNA of cancer cells during radiation therapy. Quantitative information about ionization of DNA and its components is important not only for DNA radiation damage, but also for understanding redox properties of DNA in redox sensing or labeling, as well as charge migration along the double helix in nanoelectronics applications. Until recently, the vast majority of experimental and computational data on DNA ionization was pertinent to its components in the gas phase, which is far from its native aqueous environment. The situation has, however, changed for the better due to the advent of photoelectron spectroscopy in liquid microjets and its most recent application to photoionization of aqueous nucleosides, nucleotides, and larger DNA fragments. Here, we present a consistent and efficient computational methodology, which allows to accurately evaluate ionization energies and model photoelectron spectra of aqueous DNA and its individual components. After careful benchmarking, the method based on density functional theory and its time-dependent variant with properly chosen hybrid functionals and polarizable continuum solvent model provides ionization energies with accuracy of 0.2-0.3 eV, allowing for faithful modeling and interpretation of DNA photoionization. The key finding is that the aqueous medium is remarkably efficient in screening the interactions within DNA such that, unlike in the gas phase, ionization of a base, nucleoside, or nucleotide depends only very weakly on the particular DNA context. An exception is the electronic interaction between neighboring bases which can lead to sequence-specific effects, such as a partial delocalization of the cationic hole upon ionization enabled by presence of adjacent bases of the same type.

Communication: UV photoionization of cytosine catalyzed by Ag+

“…[14][15][16][17]48 Usually, primary photoionization of guanine, the base with the lowest IP, 14 takes place upon direct interaction with ionizing radiation with photons near 8 eV. In this work, we show that Cyt can be also ionized, but at lower photon energies (less than 4 eV), when interacting with Ag + , through a CT state.…”

Efficient modeling of liquid phase photoemission spectra and reorganization energies: Difficult case of multiply charged anions

“…The first problem is connected to the evaluation of ionization energies (IE) with quantum chemical methods. The IE of isolated molecules can be calculated with a variety of methods, ranging from simple Koopmans‐based estimates, various Green function based schemes, density functional theory (DFT) approaches up to highly demanding multi‐reference methods, symmetry adapted cluster/configuration interaction approach or coupled cluster based approaches . Based on our experience, the DFT combined with its time‐dependent version (DFT/TDDFT) and suitably selected functional represents a pragmatic choice, an error is typically 0.1–0.3 eV for a medium sized molecule …”

Free Energies of Redox Half-Reactions from First-Principles Calculations

“…The system-size effects on the computed VIE were also discussed in the context of ionization of neutral species [8]. The linear dependence of VEGs, free energies, and reorganization energy on 1/R can be rationalized by considering dielectric media, splitting the solvent polarization response onto the slow (static) and fast (optical) components, and assuming a spherical charge distribution for the solute in the charged state [40][41][42].…”

“…aug−cc−pV T Z − VEG gas,DF T aug−cc−pV T Z (8) where VEG gas are ionization energies computed for the isolated (gas-phase) solute molecule at its equilibrium geometry, < ∆VEG b.s. > is the basis-set correction (see Sec.…”