Direct Observation of a Hydrogen Atom Adduct to C-5 in Uracil. A Neutralization-Reionization Mass Spectrometric and ab Initio Study

The DNA Radical Code. Resolution of Identity in Dissociations of Trinucleotide Codon Cation Radicals in the Gas Phase

Sixty DNA trinucleotide cation radicals covering a large part of the genetic code alphabet were generated by electron transfer in the gas phase, and their chemistry was studied by collision-induced dissociation tandem mass spectrometry and theoretical calculations. The major dissociations involved loss of nucleobase molecules and radicals, backbone cleavage, and cross-ring fragmentations that depended on the nature and position of the nucleobases. Mass identity in dissociations of symmetrical trinucleotide cation radicals of the (XXX+2H)^+• and (XYX+2H)^+• type was resolved by specific ¹⁵N labeling. The specific features of trinucleotide cation radical dissociations involved the dominant formation of d₂⁺ ions, hydrogen atom migrations accompanying the formation of (w₂+H)^+•, (w₂+2H)⁺, and (d₂+2H)⁺ sequence ions, and cross-ring cleavages in the 3'- and 5'-deoxyribose moieties that depended on the nucleobase type and its position in the ion. Born-Oppenheimer molecular dynamics (BOMD) and density functional theory calculations were used to obtain structures and energies of several cation-radical protomers and conformers for (AAA+2H)^+•, (CCC+2H)^+•, (GGG+2H)^+•, (ACA+2H)^+•, and (CAA+2H)^+• that were representative of the different types of backbone dissociations. The ion electronic structure, protonation and radical sites, and hydrogen bonding were used to propose reaction mechanisms for the dissociations.

Charge-Tagged Nucleosides in the Gas Phase: UV-Vis Action Spectroscopy and Structures of Cytidine Cations, Dications, and Cation Radicals

Cytidine ribonucleosides were furnished at O5' with fixed-charge 6-trimethylammoniumhexan-1-aminecarbonyl tags and studied by UV-vis photodissociation action spectroscopy in the gas phase to probe isolated nucleobase chromophores in their neutral, protonated, and hydrogen-adduct radical forms. The action spectrum of the doubly charged cytidine conjugate showed bands at 310 and 270 nm that were assigned to the N3- and O2-protonated cytosine tautomers formed by electrospray, respectively. In contrast, cytidine conjugates coordinated to dibenzo-18-crown-6-ether (DBCE) in a noncovalent complex were found to strongly favor protonation at N3, forming a single-ion tautomer. This allowed us to form cytidine N3-H radicals by electron transfer dissociation of the complex and study their action spectra. Cytidine radicals showed only very weak absorption in the visible region of the spectrum for dipole-disallowed transitions to the low (A and B) excited states. The main bands were observed at 360, 300, and 250 nm that were assigned with the help of theoretical vibronic spectra obtained by time-dependent density functional theory calculations of multiple (>300) radical vibrational configurations. Collision-induced dissociations of cytidine radicals proceeded by major cleavage of the N1-C1' glycosidic bond leading to loss of cytosine and competitive loss of N3-hydrogen atom. These dissociations were characterized by calculations of transition-state structures and energies using combined Born-Oppenheimer molecular dynamics and DFT calculations. Overall, cytidine radicals were found to be kinetically and thermodynamically more stable than previously reported analogous adenosine and guanosine radicals.

On the hydrogen loss from protonated nucleobases after electronic excitation or collisional electron capture

In this work, we have subjected protonated nucleobases MH(+) (M = guanine, adenine, thymine, uracil and cytosine) to a range of experiments that involve high-energy (50 keV) collision induced dissociation and electron capture induced dissociation. In the latter case, both neutralisation reionisation and charge reversal were done. For the collision induced dissociation experiments, the ions interacted with O(2). In neutral reionisation, caesium atoms were used as the target gas and the protonated nucleobases captured electrons to give neutrals. These were reionised to cations a microsecond later in collisions with O(2). In choosing Cs as the target gas, we have assured that the first electron transfer process is favourable (by about 0.1-0.8 eV depending on the base). In the case of protonated adenine, charge reversal experiments (two Cs collisions) were also carried out, with the results corroborating those from the neutralisation reionisation experiments. We find that while collisional excitation of protonated nucleobases in O(2) may lead to hydrogen loss with limited probabilities, this channel becomes dominant for electron capture events. Indeed, when sampling reionised neutrals on a microsecond timescale, we see that the ratio between MH(+) and M(+) is 0.2-0.4 when one electron is captured from Cs. There are differences in these ratios between the bases but no obvious correlation with recombination energies was found.

Modeling deoxyribonucleic acid and ribonucleic acid damage in the gas phase

This short review outlines the tandem mass spectrometric methods for the generation and analysis of transient nucleobase radicals relevant to deoxyribonucleic acid and ribonucleic acid damage. Radical hydrogen atom adducts to uracil, adenine, cytosine and N-methylcytosine were generated by femtosecond electron transfer to the corresponding gas-phase cations in fast beams at 8 keV kinetic energy. Radical unimolecular dissociations were monitored by product analysis following collisional ionization to cations or anions using neutralization-reionization mass spectrometry. The radical energetics and dissociation kinetics were further analyzed by mapping the potential energy surfaces by high-level ab initio calculations in combination with Rice-Remsberger-Kassel-Marcus calculations of unimolecular rate constants. This first- principles-based approach allows one to model radical dissociations occurring from doublet ground electronic states of radical intermediates, assign reaction mechanisms and derive quantitative branching ratios for dissociation channels that are in agreement with experiments. Theoretical analysis also provides distinction between radical dissociations occurring on the ground and excited electronic state potential energy surfaces. Specific characterization of excited state dissociations of nucleobase and other polyatomic radicals remains a challenging topic for both experimentalists and computational chemists.

Use of DFT-based reactivity descriptors for rationalizing radical addition reactions: applicability and difficulties

The use of density functional theory-based reactivity descriptors, including global (hardness, electronegativity) and local (Fukui function, local softness) indices in rationalizing the reactivity and regioselectivity of radical addition reactions has been critically analyzed. We demonstrate that there is a severe inherent deficiency in the current way of defining the Fukui functions and local softness for radical attack parameters (f(0) and s(0)), and propose a (preliminary) alternative.

The effect of radical trap moieties on electron capture dissociation spectra of substance P

To further test the hypothesis that electron capture dissociation (ECD) involves long-lived radical intermediates and radical migration occurs within these intermediates before fragmentation, radical trap moieties were attached to peptides with the assumption that they would reduce fragmentation by decreasing the mobility of the radical. Coumarin labels were chosen for the radical traps, and unlabeled, singly-labeled, and doubly-labeled Substance P were analyzed by ECD. The results demonstrated a correlation between the number and position of tags on the peptide and the intensity of side-chain cleavages observed, as well as an inverse correlation between the number of tags on the peptide and the intensity of backbone cleavages. Addition of radical traps to the peptide inhibits backbone cleavages, suggesting that either radical mobility is required for these cleavages, or new noncovalent interactions prevent separation of backbone cleavage fragments. The enhancement of side-chain cleavages and the observation of new side-chain cleavages associated with aromatic groups suggest that the gas-phase conformation of this peptide is substantially distorted from untagged Substance P and involves previously unobserved interactions between the coumarin tags and the phenylalanine residues. Furthermore, the use of a double resonance (DR)-ECD experiment showed that these side-chain losses are all products of long-lived radical intermediate species, which suggests that steric hindrance prevents the coumarin-localized radical from interacting with the backbone while simultaneously increasing the radical rearrangements with the side chains.

Simple b ions have cyclic oxazolone structures. A neutralization-reionization mass spectrometric and computational study of oxazolone radicals

The 2-methyloxazol-5-on-2-yl radical (3) and its deuterium labeled analogs were generated in the gas-phase by femtosecond electron-transfer and studied by neutralization-reionization mass spectrometry and quantum chemical calculations. Radical 3 undergoes fast dissociation by ring opening and elimination of CO and CH(3)CO. Loss of hydrogen is less abundant and involves hydrogen atoms from both the ring and side-chain positions. The experimental results are corroborated by the analysis of the potential energy surface of the ground electronic state in 3 using density functional, perturbational, and coupled-cluster theories up to CCSD(T) and extrapolated to the 6-311 ++ G(3df,2p) basis set. RRKM calculations of radical dissociations gave branching ratios for loss of CO and H that were k(CO)/k(H) > 10 over an 80-300 kJ mol(-1) range of internal energies. The driving force for the dissociations of 3 is provided by large Franck-Condon effects on vertical neutralization and possibly from involvement of excited electronic states. Calculations also provided the adiabatic ionization energy of 3, IE(adiab) = 5.48 eV and vertical recombination energy of cation 3(+), RE(vert) = 4.70 eV. The present results strongly indicate that oxazolone structures can explain fragmentations of b-type peptide ions upon electron capture, contrary to previous speculations.

Modeling deoxyribose radicals by neutralization-reionization mass spectrometry. Part 2. Preparation, dissociations, and energetics of 3-hydroxyoxolan-3-yl radical and cation

The title radical (1) is generated in the gas-phase by collisional neutralization of carbonyl-protonated oxolan-3-one. A 1.5% fraction of 1 does not dissociate and is detected following reionization as survivor ions. The major dissociation of 1 (approximately 56%) occurs as loss of the hydroxyl H atom forming oxolan-3-one (2). The competing ring cleavages by O[bond]C-2 and C-4[bond]C-5 bond dissociations combined account for approximately 42% of dissociation and result in the formation of formaldehyde and 2-hydroxyallyl radical. Additional ring-cleavage dissociations of 1 resulting in the formation of C(2)H(3)O and C(2)H(4)O cannot be explained as occurring competitively on the doublet ground (X) electronic state of 1, but are energetically accessible from the A and higher electronic states accessed by vertical electron transfer. Exothermic protonation of 2 also produces 3-oxo-(1H)-oxolanium cation (3(+)) which upon collisional neutralization gives hypervalent 3-oxo-(1H)-oxolanium radical (3). The latter dissociates spontaneously by ring opening and expulsion of hydroxy radical. Experiment and calculations suggest that carbohydrate radicals incorporating the 3-hydroxyoxolan-3-yl motif will prefer ring-cleavage dissociations at low internal energies or upon photoexcitation by absorbing light at approximately 590 and approximately 400 nm.