CHAPTER 1
Photoinduced Processes in DNA: Basic Theoretical and Experimental Features
Roberto Improta*a and Thierry Douki*b
a CNR â Consiglio Nazionale delle Ricerche, Istituto di Biostrutture e Bioimmagini (IBB-CNR), via Mezzocannone 16, Italy,
This chapter is intended to provide a concise overview of some basic and general concepts of photophysics/photochemistry useful for the understanding of the following chapters. After a general and easy introduction to the theoretical basis of the interaction between light and molecules, we shall briefly describe the principle of the most used steady-state and time-resolved experimental spectroscopy techniques. Then we shall introduce the most used analytical techniques to quantify and characterize the photodamage and describe the main biological phenomena associated with DNA damage and repair.
1.1 Introduction
This book describes the main processes triggered by the interaction of ultraviolet (UV) light with nucleic acids, either directly or indirectly through the mediation of other compounds, and discusses some of the resulting biological consequences. It is not possible, obviously, to treat, even rapidly, all the scientific and technical aspects involved in the study of such a large variety of phenomena. However, it can be helpful for readers to have a concise review of some basic features of the techniques, experimental or computational, most commonly used to investigate these processes, and the general concepts/models employed for their interpretation, focussing essentially on those most thoroughly used in the following chapters. This is the goal of the present introductory chapter. Some parts will be obvious to chemists and some to biologists. Hopefully, all readers will share a common language and will be able to gather useful information from the whole volume.
1.2 A Theoretical Basis
According to Quantum Mechanics (QM), the behaviour of atoms and molecules is governed by the Hamiltonian operator, which contains all the energy terms associated with electrons and nuclei. 1,2 The solution of the associated Schrödinger equation, which allows prediction of the time evolution of the system, is made simpler by resorting to the BornâOppenheimer (BO) approximation, which provides the separation of the motion of electrons and nuclei, the latter being much heavier and, therefore, much slower. In this framework, we can identify different electronic states for a molecule (i.e. the solution of the electronic Hamiltonian) differing on the distribution of the electrons in the molecular orbitals (MOs). It is then possible to compute for each molecule a ground electronic state with its equilibrium geometry, corresponding to the structure in which the total energy (electron + nuclei) of the molecule reaches its minimum. Please remember that the total energy is negative, so a decrease in energy corresponds to a stabilization of the system. When the ground electronic state (sometimes known as âGSâ) is a singlet (i.e. all the electrons are spin paired in their MOs), it is very often described as âS0â.
As schematically depicted in Figure 1.1, when a molecule absorbs a photon of appropriate energy, i.e. corresponding to the energy difference between the âgroundâ and an âexcitedâ electronic state, an electron is promoted to some vibrational level in the excited singlet manifold (S1, S2, S3 in order of increasing energy). 3 However, not all the transitions are âallowedâ, i.e. for symmetry reasons light absorption can populate some excited states (described as âbright excited statesâ) and not others (âdark excited statesâ). For example, in the molecule depicted above (thymine), the lowest energy bright excited state corresponds to S2 and can be described as due to a ÏÏ* (indicating that the transfer is from a Ï to a Ï* symmetry MO) transition. In thymine it involves the transfer of an electron from the Highest Occupied Molecular Orbital (HOMO, a Ï MO) to the lowest unoccupied MO (LUMO, a Ï* MO). A similar transition is usually described as HOMO â LUMO. S1, i.e. the lowest energy excited state in the ground state minimum, is a dark nÏ*, involving a lone pair of the carbonyl oxygen, which is the second highest occupied molecular orbital (HOMO-1) and the same Ï* LUMO. This transition is usually labelled as HOMO-1 â LUMO. In addition to excited states having the same multiplicity as the ground electronic state (a singlet in our example), we can have excited electronic states with different multiplicity, as the triplets (T1, T2, T3 in order of increasing energy), where the two electrons in the âhalf-filledâ MOs have the same spin. These excited states, which can be described by following a similar framework used below for the singlets, i.e. based on the occupied MOs, cannot be reached directly by light irradiation but, as discussed below, can also play a role in photoactivated dynamics. We highlight that we have just described the different adiabatic states (S1, S2, T1, etc.) according to their âdiabaticâ nature, related to the occupancy of the different electronic states. The âadiabaticâ description simply denotes the relative energy of the different âadiabaticâ electronic state for a given nuclear structure. As a consequence, the mapping between adiabatic and diabatic pictures can thus depend on the considered geometry. In other words, as we'll see below, we can find nuclear arrangements where, in the example we have used, S1, i.e. the lowest energy adiabatic state, corresponds to a ÏÏ* transition, i.e. to a different diabatic state than in the ground state minimum.
Figure 1.1 Schematic description of some of the changes in the electronic states population induced by light irradiation of a molecule (using thymine as an example).
After light absorption, our molecule is now in an S2 electronic state. The process of light absorption is extremely rapid (â€1 fs) and the nuclei can be considered frozen on that timescale, i.e. they still have the âoptimalâ arrangement for S0. Within the BO approximation, as we discussed above for S0, we can associate each electronic state with a potential energy surface (PES), which tells us how the energy of this state changes due to the motion of the nuclei. The molecule thus starts evolving on the S2 PES, i.e. the nuclei start readjusting to find the new minimum energy structure and the total energy of S2 starts decreasing, and can reach its minimum. In its minimum it can be the case that the ÏÏ* diabatic state is the lowest energy excited state (S1), whereas the nÏ* state now corresponds to S2.
One possibility is then that the molecule emits light, and so we have a radiative decay to S0. This process is known as fluorescence when emission takes place from a single state, but eventually it can also happen from a triplet state and is called phosphorescence (Figure 1.2, straight arrows). The ratio between the photons absorbed and emitted is defined as the fluorescence quantum yield (often abbreviated as QY and sometimes as Ï). The concept of QY is very important in the study of radiation-induced processes and is not limited to fluorescence, but it can be extended to any possible photoactivated event, as the ratio between the number of events occurring in the system and the number of photons absorbed. For example, in the following chapters we shall often refer to the QY of the different photochemical processes.
Figure 1.2 Jablonski diagram of the main photophysical processes in a molecule photoexcited on S2.
As depicted in Figure 1.2 (curved arrows) a bright excited state can also decay non-radiatively by internal conversion to dark excited states (both singlet and triplets) or, directly, to S0. This process is usually described as âinternal conversionâ (IC) or as âintersystem crossingâ (ISC), depending on whether it involves states with the same or different spin multiplicity.
The non-radiative decay is particularly effective in the proximity of the surface crossings, i.e. when two different PESs are degenerate (Figure 1.3). In these regions, the BO approximation is not valid, and the so-called non-adiabatic coupling effects and, for singlet/triplet crossings, spin-orbit couplings need to be taken into account to correctly describe the evolution of the molecular system. For only two coordinates, the intersection between the PESs of two states of different symmetry is a single point and the PESs can adopt a typical âconicalâ shape (with the degeneracy point being the vertex), giving account of the term âconical intersectionâ (CI), which is usually used to define the lowest energy structure of a crossing seam between two singlet states. In a similar structure for crossing between states of different multiplicity, the more general label of minimum energy crossing point (MECP) can be used. The presence of an easily accessible crossing region between S0 and the PES of a bright excited state leads to a very fast and effective non-radiative ground state recovery, mirrored by an ultrashort excited state lifetime.
Figure 1.3 Schematic picture of the potential energy surfaces of the ground (S0) and an excited electronic state (S1) and of the main photophysical and photochemical processes involved (see text for details).
Strongly fluorescent molecules have a QY close to 1 and a radiative lifetime of several nanoseconds. The fluorescence QY of DNA and of its component is instead extremely low (10â3âŒ10â4) and the lifetime of the bright excited states is of a few ps (at maximum). Actually, for nucleobases an almost barrierless path on the PES of the bright excited state leads to a crossing region with S0 (see Chapter 2). Moreover, for oligonucleotides, the bright excited state can decay to almost-dark excited states (for example transitions with charge transfer character) which then return to S0, non-radiatively, by charge recombination.
Figure 1.3 shows another possible outcome of the photoactivated dynamics, that the system reaches another stable arrangement of the nuclei, i.e. producing a ânewâ molecule stable at room temperature. Examples of these âphotochemical reactionsâ are the photodimerizations discussed in Chapters 2 and 4 or the ionizations described in Chapter 3.
From the computational point of view, a full understanding of the photoactivated dynamics requires the complete characterization of the PESs of the electronic states, of their crossings, stationary points, and the characterization of the spectral signatures of such stationary points (e.g. absorption and emission spectra). This âstaticâ picture can already give useful insights on the most important excited state processes, and it is nowadays quite routinely applied to many systems, including oligonucleotides. The simulation of the excited state dynamics requires also the determination of the non-adiabatic couplings between the different electronic states and the application of dynamical methods, either semiclassical (i.e. the motion of the nuclei is described according the classical law of mechanics) or quantum dynamical (i.e. describing also the motion of the nuclei at a QM level). Thanks to the development of more accurate computational methods, efficient software and the continuously increasing computing power, the possibility of studying, at the computation level, the excited state dynamics of molecular systems has known impressive advances in the last years, as recorded in many chapters of this book. 4
1.3 Notion of Spectrum in Photobiology
All photochemical reactions...