The origin of efficient triplet state population in sulfur-substituted nucleobases

Abstract

Elucidating the photophysical mechanisms in sulfur-substituted nucleobases (thiobases) is essential for designing prospective drugs for photo- and chemotherapeutic applications. Although it has long been established that the phototherapeutic activity of thiobases is intimately linked to efficient intersystem crossing into reactive triplet states, the molecular factors underlying this efficiency are poorly understood. Herein we combine femtosecond transient absorption experiments with quantum chemistry and nonadiabatic dynamics simulations to investigate 2-thiocytosine as a necessary step to unravel the electronic and structural elements that lead to ultrafast and near-unity triplet-state population in thiobases in general. We show that different parts of the potential energy surfaces are stabilized to different extents via thionation, quenching the intrinsic photostability of canonical DNA and RNA nucleobases. These findings satisfactorily explain why thiobases exhibit the fastest intersystem crossing lifetimes measured to date among bio-organic molecules and have near-unity triplet yields, whereas the triplet yields of canonical nucleobases are nearly zero.

Figure 1. Transient absorption spectra.

(a) Early time delays (−360 to 320 fs) and (b) later time delays (320 to 3,760 fs). The spectra were recorded in phosphate buffer solution (pH 7.4) using an excitation wavelength of 308 nm. Arrows indicate the movement of the absorption bands with time. The grey boxes block the overtone band of the excitation light. The inset shows the molecular structure of 2-thiocytosine.

Figure 2. Target and global analysis of the TAS.

(a) Decay associated spectra obtained by globally fitting the transient absorption data to a two-component sequential kinetic model (A and B) with a constant offset (C). (b) Kinetic traces and fitted functions from the global fit analysis, taken at representative probe wavelengths, as indicated with coloured dashed lines in (a).

Figure 3. Simulated absorption spectra of the excited states.

(a) Calculated absorption spectra from the given excited states at their respective minimum geometries (MS-CASPT2, convoluted using Gaussians with 0.7 eV full-width at half-maximum). (b) Linear combinations of the calculated spectra with the given proportions. (c) The experimental transient absorption spectra at the indicated time delays.

Figure 4. Schematic representation of possible relaxation pathways.

All energies were computed at the MS-CASPT2 level of theory and are given in eV. Additional information can be found in Supplementary Note 2.

Figure 5. Excited-state populations from surface hopping simulations.

Simulations were performed with the SHARC method coupled to the MRCIS electronic structure method. Thin lines represent the simulated electronic populations, thick lines show the fitted functions.

Figure 6. General relaxation mechanism of 2-thiocytosine.

The scheme is based on the presented experimental and computational data. The τ₁ and τ₂ are the experimental time constants.