Premium
Low‐Energy‐Barrier Proton Transfer Induced by Electron Attachment to the Guanine⋅⋅⋅Cytosine Base Pair
Author(s) -
Szyperska Anna,
Rak Janusz,
Leszczynski Jerzy,
Li Xiang,
Ko Yeon Jae,
Wang Haopeng,
Bowen Kit H.
Publication year - 2010
Publication title -
chemphyschem
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.016
H-Index - 140
eISSN - 1439-7641
pISSN - 1439-4235
DOI - 10.1002/cphc.200900810
Subject(s) - guanine , chemistry , cytosine , ion , proton , crystallography , computational chemistry , nucleotide , dna , physics , biochemistry , organic chemistry , quantum mechanics , gene
The photoelectron spectrum of the anion of the guanine ⋅⋅⋅ cytosine base apair (GC) .− is recorded for the first time. The observed variation in the spectral peak‐height ratios with the source conditions suggests the presence of two or more anionic isomers. Two maxima of the broad bands in the photoelectron spectrum were measured at about 1.9 and about 2.6 eV. These values are very well reproduced by the vertical detachment energies of the B3LYP/6‐31++G(d,p) calculated low‐energy anionic structures, which are 1) the Watson–Crick base‐pair anion with proton transferred from N1 of guanine to N3 of cytosine, 2) its analogue in which the proton is transferred from N9 of guanine to N7 of guanine, and 3) the global minimum geometry, which is formed from the latter anion by rotation of guanine about the axis approximately defined by C2 of guanine and C4 of cytosine. Furthermore, a minor difference in the stabilities of the two lowest energy anions explains the experimentally observed source (temperature) dependence of the PES spectrum. A rational procedure, based on the chemistry involved in the formation of anionic dimers, which enables the low‐energy anions populated in the photoelectron spectrum to be identified is proposed. In contrast to the alternative combinatorial approach, which in the studied case would lead to carrying out quantum chemical calculations for 2000–2500 structures, the procedure described here reduces the computational problem to only 15 geometries.