Theoretical Insights on the Inefficiency of RNA ... - ACS Publications

Dec 5, 2017 - provided in the Supporting Information (Figure S6−S8 and. Table S4−S6). ... in all of its corresponding cis-(5R,6S) peroxyl species ...
0 downloads 0 Views 561KB Size
Subscriber access provided by READING UNIV

Article

Theoretical Insights on the Inefficiency of RNA Oxidative Damage Under Aerobic Conditions Shuang Zhao, Leif A Eriksson, and Ru-Bo Zhang J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.7b10711 • Publication Date (Web): 05 Dec 2017 Downloaded from http://pubs.acs.org on December 5, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry A is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Theoretical Insights on the Inefficiency of RNA Oxidative Damage under Aerobic Conditions Shuang Zhao a, Leif. A. Erikssonb,* and Ru-bo Zhang a,* aSchool

of Chemistry and Chemical Engineering, Beijing Institute of Technology, South Street No. 5, Zhongguancun, Haidian

District Beijing 100081, China

bDepartment

of Chemistry and Molecular Biology, University of Gothenburg, Box 462, 405 30 Göteborg, Sweden

* Corresponding author. E-mail: [email protected]; [email protected]

ABSTRACT Oxidative damage to RNA has been linked to change or loss of RNA function and development of many human age-related diseases. However, knowledge on the nature of RNA oxidative damage is relatively limited. In this study, oxidative damage to RNA is investigated under anaerobic and aerobic conditions by exploring the properties and reactions of 5-hydroxyl-2’-uridin-6-yl and its peroxyl diastereoisomers in the RNA strand, respectively. Selective addition of OH to the nucleic base from the 5’-end is studied at the molecular level for the first time, explaining the large number of the 5S-isomer available for further reactions. Our results provide clear evidence that the efficiency of C2’-H2’ bond activation in the peroxyl isomers is lower than in the carbon radical species. An exception is observed for the isomer cis-(5S,6R)-A1, whose internucleotidyl H2’-abstraction barrier is far smaller than that in the corresponding C6-yl radical. However, analysis of the equilibrium species distribution reveals that the amount of cis-(5S,6R)-A1 is very small among the peroxyl diastereoisomers, and hence the resulting products from direct strand scission should be a less important component in RNA oxidative damage. The species with maximum distribution is the cis-(5S,6R)-B1 isomer, which is derived from cis-(5S,6R)-A1, and has a moderate intranucleotidyl H2’-abstraction barrier. More importantly, the reaction is mildly exothermic. These results show that the main fraction of the intranucleotidyl H2’-abstraction intermediates can be formed from the cis-(5S,6R)-B1 isomer. The absolute reduction potentials, the hydrogen atom binding energies and the key structural parameters of the C6-peroxyl species are used to understand the diverse reactivity of the cis-(5S,6R) diastereoisomers towards the C2’-H2’ bonds activation. The present study shows that in addition to the selectivity of the OH radical addition, there is a strong correlation between the conformation of the modified uracil base and its reactivity in RNA oxidative damage.

1

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 25

1. Introduction The hydroxyl radical, being one of the reactive oxygen species (ROS), readily attacks biomolecules with resulting loss of or change in their biological functions.1 Compared to DNA oxidative damage, little attention has been given to RNA oxidation although it accounts for 80~90% of the total cellular nucleic acids in the cells.2 Based on abundance, it is thus not unlikely that RNA may be the major target of nucleic acid-damaging agents.3 There is significant evidence linking RNA oxidation to various age-related neurodegenerative diseases and other ROS-induced physiological damage.4-6 The RNA molecules should be more easily oxidized than DNA since the 2’-OH group of the ribose moiety greatly contributes to increased acidity of the remaining H2’ atom. Recent experiments7-10 have demonstrated that H2’ is the most facile position to be abstracted by uracil C6-yl in an internucleotidyl process. Our previous theoretical studies11 showed that the efficiency of C2’-H2’ bond activation in the C6-yl radical is quite high under anaerobic conditions, which in turn leads to C3’-O3’(P) bond breakage and direct strand scission. RNA oxidation products were in the past mainly monitored through formation of 8-hydroxyguanosine (8-OH-Gua), an approach that lately has been challenged by Fields et al.12,13 Recent in vitro experiments revealed that RNA strand scission is significantly more efficient under anaerobic conditions than under aerobic conditions.14 The results are in sharp contrast to the generally accepted oxygen enhancement effects,15 namely that the peroxyl radical should be more reactive to nucleic acids than the corresponding carbon radical. The conflicting opinions are indicative of the limited knowledge about RNA oxidative damage under aerobic conditions, and hence the discrepancy could be resolved through answering some essential questions. For instance, is it correct that the 5-hydroxyl-2’-uridin-6-yl radical is more reactive than the corresponding peroxyl 2

ACS Paragon Plus Environment

Page 3 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

species?

Can

the

H2’

atom

of

the

ribose

be

effectively

abstracted

by

the

5-hydroxyl-2’-uridin-6-peroxyl species in an intra- or internuleotidyl manner? What factors influence reactivity of the peroxyl species and the product distribution in RNA strand breakage? We herein attempt to address these questions through careful selection of the appropriate RNA stacking models and usage of reliable DFT calculations with a particular focus to unveil the origin of the reactivity of the peroxyl species in RNA. In addition to a remarkable consistency with previous experimental results,14 our theoretical results furthermore show at a molecular level that in the modified RNA, there is a strong correlation between conformation and reactivity of the peroxyl diastereoisomers, which influences the efficiency and the distribution of strand scission products in RNA oxidative damage.

2. Computational Model and Methods The standard A-type (rU)3 stacking structure was constructed from ideal geometries as given in NDB.16 U and r stand for uracil base and ribose, respectively. The phosphate groups were neutralized through protonation to guarantee correct electronic structures. The method is common practice in computational chemistry and has been validated in recent studies.17,18 Previous investigations indicate that OH attack on the C5-position of uridine, which results in formation of the C6-yl radical, accounts for ~70% of the reactions between hydroxyl and poly(U).19,20 In the present studies, the C6-yl radical was obtained through addition of OH radical to the C5-position of the middle uracil base of the stacked trimer. The RNA C6-yl species is known to be reactive towards C2’-H2’ bond activation under anaerobic conditions.7-10 Under aerobic conditions, on the other hand, extrinsic O2 molecules will preferentially

react

with

the

C6-yl

radical

site

to

3

ACS Paragon Plus Environment

produce

the

corresponding

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 25

5-hydroxyl-2’-uridin-6-peroxyl reactive species. The primary reaction processes and the important atomic labels of the C6-peroxyl model are shown in Scheme 1.

Scheme 1. Schematic view of the formation process of the 5-hydroxyl-2’-uridin-6- peroxyl radical.

All calculations were carried out using Gaussian 09.21 The structures were optimized at the M06-2X/6-31G(d,p) level22,23 and assessed in order to avoid formation of unrealistic hydrogen bonds. The M06-2X functional is designed in part to yield more accurate noncovalent interactions that contain significant dispersion contributions, as well as reliable thermochemical data.23,24 The double-ζ Pople basis set 6-31G(d,p) was used throughout, to be consistent with previous studies of the reactivity of the DNA peroxyl radical.25 The conductor-like polarizable continuum model (CPCM)26,27 was employed with ε=78.6 to simulate an aqueous environment. Frequency calculations were carried out at the same level of theory to confirm that the optimized geometries corresponded to stationary points and to extract zero-point vibration energies, thermal enthalpies and free energies. The transition state (TS) structures are characterized by the existence of a unique imaginary frequency, which corresponds to formation or breakage of the targeted bond. Intrinsic reaction coordinate (IRC) 4

ACS Paragon Plus Environment

Page 5 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

calculations 28 were performed for the TS structures as applicable. Single-point energies were computed on the optimized structures, at the M06-2X/6-311++G(d,p) level. To better characterize the kinetics of the H2’-abstraction reaction by the carbon-centered radical and its corresponding peroxyl species, thermal tunneling-corrected canonical transition state theory was implemented to calculate the rate constants of the major paths from the following equation29-31:

K ( T ) = k( T )

σK BT QTS ( T ) h

E≠ exp( − ) QR ( T ) RT

(1)

where σ, KB, h and T are the reaction symmetry number, Boltzmann’s constant, Planck’s constant and temperature (Kelvin), respectively. k(T) is the thermal tunneling coefficient, which was obtained through calculation of the one-dimensional tunneling probability. The Eckart potential32 can generally provide an accurate representation of the barrier, and was used herein to calculate the tunneling probability. E≠ is the classical barrier height without ZPE correction. QTS(T) and QR(T) represent the partition function calculated, respectively, for the transition state (TS) and the reactant (R). For the practical use in chemical kinetic modeling, three kinetic parameters A, n and Ea for each reaction of the major thermal decomposition paths were obtained by fitting the calculated rate constants over the temperature range of 273 ~ 313 K to a modified Arrhenius expression 29-31:

K (T ) = Ak (T )T n exp(−

Ea ) RT

(2)

3. Results and discussion 3.1 Reactivity of the 5-hydroxyl-2’-uridin-6-yl species in single-strand RNA under anaerobic conditions 5

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 25

Hydroxyl radical attack to the C5 position of the middle uracil base of the (rU)3 model leads to the formation of 5S-(rU)3 or 5R-(rU)3 chiral diastereoisomers, which corresponds to OH radical addition to the middle uracil base from the RNA 5’- or 3’-end, respectively. The stereoselectivity of OH radical addition to the RNA strand as opposed to a single nucleic base or base pair33-35 is very important for the distribution of the radical adducts, but has been addressed to a very small degree in the past. The structures of the RNA strand models were optimized and are displayed in supporting material Fig. S1. In most of the radical-molecule reactions, OH addition channels consist of an initial formation of a reaction complex (RC) in a reversible and barrierless process, followed by irreversible formation of the products; Eqs. (3) and (4): Step 1 :

(rU)3 + OH ↔ (rU)3···OH

Step 2 :

(rU)3···OH → HO-(rU)3 (5S or 5R)

(3) (4)

The mechanism for the radical-molecule reactions was first proposed by Singleton and Cvetanovic

36

and has been used successfully in quantum chemical TST calculations.29-31 If k1 and k−1 are the forward and reverse rate constants for the first step, respectively, and k2 is the rate constant for the second step, a steady-state analysis leads to a rate constant of the overall reaction which can be approximated as (Eq. (5)):

k=

k1 k 2 = K eq k 2 k −1

(5)

Herein, Keq is the equilibrium constant between the isolated reactants and the reaction complex. According to statistical thermodynamic principles, Keq can be written as (Eq. (6)):

K eq =

Q( rU )3 ⋅⋅⋅OH Q( rU )3 QOH

exp (

E R − E( rU )3 ...OH RT

6

ACS Paragon Plus Environment

)

(6)

Page 7 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

where Q(rU)3, QOH, and Q(rU)3...OH are the partition functions of reactants (rU)3, OH, and the reaction complex, respectively; ER and E(rU)3...OH are the respective total energies of the reactants and the reaction complex, R is the ideal gas constant, and T is the absolute temperature.

Table 1 Equilibrium constant Keq of step 1, rate constant k2, activation barriers (in kcal/mol) including zero-point correction energy (ΔE‡) and enthalpy (ΔH‡) of step 2 at the M06-2X/6-311++G(d,p)//M06-2X/6-31G(d,p) level at T = 298 K.

aQ

OH

step 1a

Q(rU)3...OH

EOH

E(rU)3...OH

Keq

·OH+(rU)3→5R-(rU)3···OH ·OH+(rU)3→5S-(rU)3···OH

10-294.1 10-292.4

-75.731892

-3869.937897 -3869.938489

6.4 × 10-10 3.2 × 10-8

step 2

ΔE‡

ΔH‡

νTS(cm-1)

k2

5R-(rU)3···OH→5R-(rU)3 5S-(rU)3···OH→5S-(rU)3

3.9 2.8

4.2 3.1

-436 -361

5.50 × 109 3.81 × 1010

and Q(rU)3 are1011.6 and 10-287.3, respectively.

The Keq values from step 1 are presented in Table 1. From the Keq values, formation of the 5S-(rU)3···OH reaction complex is favored more than 50-fold over that of 5R-(rU)3···OH at 298 K. It should be noted that both reaction complexes were derived from IRC calculations (Fig. S2) of the (rU)3···OH → HO-(rU)3 reactions, and further optimized at the M06-2X/6-31G(d,p) level. The rate constants k2 of the addition reactions were thus calculated using classical TST theory. As shown in Table 1, the rate constant k2 of the 5S-(rU)3···OH → 5S-(rU)3 reaction is 7-fold greater than that of the 5R-(rU)3···OH → 5R-(rU)3 reaction. Hence, since the overall reaction rate constant k is strongly dependent on both Keq and k2 (Eq. 5), formation of 5S-(rU)3 should be strongly favored. The product ratio of the 5S-(rU)3 to 5R-(rU)3 can approximately be estimated through the ratio of their overall rate constants, k5S/k5R. As shown in Fig. 1, the ratio ranges from 404 to 321 during the temperature range 273 - 313 K. To the best of our knowledge, this is the first application of the theoretical

7

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 25

formulas to explain why the yield of the 5S-(rU)3 OH adduct to the uracil base from the 5’- end of RNA is much more abundant than that of 5R-(rU)3 formed by attack from the 3’-end.8

Fig. 1. Ratio of formation rate constants (k5S/k5R) of the 5S-(rU)3 and 5R-(rU)3 diastereoisomers along the temperature range 270 – 320 K.

Based on the optimized structures of the C6-yl radical species, the formed radical could abstract H2’ in an intranucleotidyl or internucleotidyl manner. The procedure is strongly relevant to direct strand scission of RNA, as indicated in our previous studies and recent experiments.9,11,14 The activation barriers for C2’-H2’ abstraction by the 5R-(rU)3 and 5S-(rU)3 C6-yl radicals are shown in Table 2. Note that irrespective of 5R- or 5S- (rU)3, the internucleotidyl C2’-H2’ bond activation barrier is markedly lower than the intranucleotidyl one, which means that H2’ from a ribose of a neigbouring nucleotide is more facile to be abstracted by the C6-yl radical. This is completely consistent with experimental observations.8,9,14 For 5S-(rU)3, the C2’-H2’ bond activation enthalpy is estimated to be 18.3 kcal/mol, while the entropy effects are unfavorable for the H2’-abstraction reaction and lead to a small increase in the barrier to 20.4 kcal/mol.

8

ACS Paragon Plus Environment

Page 9 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Table 2 Activation energies (in kcal/mol) of C2’-H2’ bond abstraction by the C6-yl radicals including zero-point energy correction (ΔE‡), enthalpy (ΔH‡) and Gibbs energy (ΔG‡), at the M06-2X/6-311++G(d,p)//M06-2X /6-31G(d,p) level. isomer 5R-(rU)3 5S- (rU)3

manner

ΔE‡

ΔH‡

ΔG‡

νTS (cm-1)

interintrainterintra-

21.3 30.9 19.0 35.0

20.3 30.2 18.3 34.6

24.3 32.6 20.4 35.3

-1655 -1904 -1661 -1865

3.2 C2’-H2’ bond activation by the 5-hydroxyl-2’-uridin-6-peroxyl species in single-strand RNA under aerobic conditions Under aerobic conditions, the C6-peroxyl species can be formed through diffusion controlled addition of one O2 molecule at the C6 position of the 5S/R-(rU)3 hydoxyl radical adduct. Based on our estimated product ratio of 5S-(rU)3 to 5R-(rU)3, there should be a preferential abundance for the C6-peroxyl species with 5S chirality. In addition, previous experiments have shown that the added -OH and -OO groups were added from the same side of the base.37 Hence, the C6-peroxyl species with 5S chirality were also optimized at the M06-2X/6-31G(d,p) level and their structures are shown in Fig. S3. From both previous38 and the present studies (cf. Fig. S4), the peroxyl group is found to exist in two rotational isomers around the C6-Oa bond. Thus, the cis-(5S,6R)-A1 and A2 conformers were optimized with the ∠H6-C6-Oa-Ob dihedral angle close to 180° and 0°, respectively. Compared to the structure of the wild type (rU)3 trimer, one of the distinct structural features observed in the A1 and A2 species is the loss of the π-π stacking conformation, which is brought about mainly by insertion of the OO group into the space between two adjacent nucleobases. 18 One would think that ‘oxygen enhancement effects’ should make the peroxyl species more reactive towards C-H bond activation than the corresponding carbon radical, which is absolutely true for DNA systems.39,40 Our previous studies showed that H1’ of deoxyribose was more facile to be 9

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 25

abstracted by a peroxyl species than by the corresponding carbon radical in single-stranded DNA.41 However, recent experimental results showed that the reverse is true for the RNA oxidative damage.14 This distinct difference between DNA and RNA prompted us to study the reactivity of the carbon radical and its corresponding peroxyl species towards C2’-H2’ bond activation in cis-(5S,6R)-A1 and A2. For the most accessible C2’-H2’ bonds, the closest H2’ atoms to the distal Ob of the C6-peroxyl in cis-(5S,6R)-A1 and A2 arise from the 5’-adjacent sugar, at distances 2.727 and 2.663 Å, respectively. The H2’ in the sugar moiety of the same nucleoside can not be abstracted by the C6-peroxyl species due to large steric hindrance. As shown in Table 3, the C2’-H2’ bond activation enthalpy and free energy in the cis-(5S,6R)-A1 are 15.3 and 16.6 kcal/mol, respectively, calculated at the M06-2X/6-311++G(d,p)//M06-2X/6-31G(d,p) level. The corresponding values for cis-(5S,6R)-A2 are 18.7 and 22.2 kcal/mol, respectively. The two reactions are hence endothermic. Compared to the barriers for the corresponding carbon radical of 5S-(rU)3 as shown in Table 2, the present activation energies of both cis-(5S,6R)-A1 and A2 peroxyl species are clearly lower or much lower than those of the carbon radical in 5S-(rU)3. The reaction rate constants of C2’-H2’ bond activation by the C6-yl in 5S-(rU)3 and its C6-peroxyl in cis-(5S,6R)-A1 and A2 were calculated in a biologically significant temperature range, as shown in Fig. 3. Tunnelling effects on the hydrogen atom transfer was also considered in Fig. 3(b). According to Fig. 3, the tunnelling effect is essential to increase of the reaction rate constant of each C2’-H2’ bond activation. Moreover, the difference in rate constants between the C6-yl in 5S-(rU)3 and the C6-peroxyl in cis-(5S,6R)-A2 decrease upon consideration of the tunnelling effects, as shown in Fig. 3. Interestingly, the rate constant for the C6-peroxyl in cis-(5S,6R)-A1 is much larger than for C6-yl in 5S-(rU)3. The results are consistent with the generally accepted oxygen enhancement effects but contradict the recent experimental results.8,9,14 10

ACS Paragon Plus Environment

Page 11 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Fig. 3. Rate constants for C2’-H2’ bond activation reactions of the cis-(5S,6R) peroxyl species and the carbon radical 5S-(rU)3 along the temperature change from 273K to 313K without (a) or with (b) tunneling effect. The color label is black: 5S-(rU)3; red: cis-(5S,6R)-A1; blue: cis-(5S,6R)-A2, pink: cis-(5S,6R)-B1 and green: cis-(5S,6R)-B2, respectively.

In the previous experiments,42 the precursors were designed with a tert-butyl ketone group attached to the C6 position of thymidine or uridine in the nucleic acid strand. Upon radiation, the formed radical hotspot on C6 occurred naturally in the strand. The authors showed the proposed equilibrium between the trans- and cis- conformers in DNA, obtained through rotation of the modified base around the N-glycosidic bond. For wild type DNA or RNA, the cis- and trans- purine nucleosides are in dynamic equilibrium, whereas the intact pyrimidine nucleoside exist only in the trans- conformer. In the present studies, the rotational potential energy profiles were determined through rotation of the modified uracil bases around the N-glycosidic bonds in cis-(5S,6R)-A1 and A2, see Fig. S5. The data shows that new rotational isomers can exist, and these were optimized giving the two stable cis-(5S,6R)-B1 and B2 diastereoisomers shown in Fig. S3, respectively. For the cis-(5S,6R)-B1 and B2 isomers, the added –OO and –OH groups are directed towards the 3’-adjacent uridine. Only the intranucleotidyl H2’ atoms are this time available to the distal Ob, at distances

11

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 25

2.608 and 2.520 Å, respectively. As seen in Table 3, the activation enthalpies of the intranucleotidyl H2’-abstraction reactions are 21.9 kcal/mol for cis-(5S,6R)-B1 and 26.4 kcal/mol for cis-(5S,6R)-B2, respectively. Note that the reaction enthalpy for cis-(5S,6R)-B1 is slightly negative, which shows that the reaction is thermally favored. The entropy effects lead to small increase in reaction barriers and energies.

Table 3 Activation energies and reaction energies including zero-point correction energy (ΔE‡ and ΔE), enthalpy (ΔH‡ and ΔH) and Gibbs energy (ΔG‡ and ΔG) of the C2’-H2’ bond activation reactions in the four cis-(5S,6R) isomers in single-strand RNA, obtained at the M06-2X/6-311++G(d,p)//M06-2X/6-31G(d,p) level (in kcal/mol).

a

isomers

ΔE‡

ΔE

ΔH‡

ΔH

ΔG‡

ΔG

νTS(cm-1)

cis-(5S,6R)-A1a cis-(5S,6R)-A2a cis-(5S,6R)-B1b cis-(5S,6R)-B2b

15.7 19.6 22.5 27.0

8.6 7.6 -1.2 1.9

15.3 18.7 21.9 26.4

8.6 7.3 -1.5 1.8

16.6 22.2 24.6 28.7

8.9 9.2 0.2 2.7

-1755 -1999 -1915 -2037

and b represent the H2’ abstraction in an inter- and intranucleotidyl manner, respectively.

The mutual relationships between the peroxyl diastereoisomers are displayed in Fig. 4. The Gibbs

free

energies

at

298

K

of

cis-(5S,6R)-A1,

A2,

B1

and

B2

at

the

M06-2X/6-311++G(d,p)//M06-2X/ 6-31G(d,p) level of theory were used to calculate the equilibrium constants of the rotational isomerization reactions according to equation (7). The calculated equilibrium constants K1 - K4 are also shown in Fig. 4.

G product − Greac tan t ∆G 0 ln K = − =− RT RT

(7)

Based on the calculated equilibrium constants K1 - K4, the species distribution of each rotational isomer can be easily calculated, which are also presented in Fig. 4. The species with maximum distribution is observed to be cis-(5S,6R)-B1 and its population distribution is estimated to constitute 76.8 % of all four cis-(5S,6R) isomers. The cis-(5S,6R)-A1 and A2 diastereoisomers account for 12

ACS Paragon Plus Environment

Page 13 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

only 2.2 % and 0.7 % of the structures, respectively. Hence, cis-(5S,6R)-B1 is by far the most dominant among the four peroxyl diastereoisomers.

Fig. 4. Mutual relationship among the four isomers cis-(5S,6R)-A1, A2, B1 and B2. Activation Gibbs energies ΔG‡ and reaction Gibbs energies ΔG for C2’-H2’ bond activation by the 5-hydroxyl-2’-uridin-6-peroxyl radical are obtained at the M06-2X/6-311++G(d,p)//M06-2X/ 6-31G(d,p) level. K1 - K4 are the equilibrium constants between the four isomers. The direction of the arrow closest to K represents the favoured rotational direction. All energies in kcal/mol.

The small species distribution for cis-(5S,6R)-A1 could be one of the reasons why only a small number of the internucleotidyl H2’-abstraction products could be observed experimentally,8,9,14 even though its reactivity is much higher than the other peroxyl diastereoisomers as well as the carbon radical species. Thus, the internuleotidyl H2’-abstraction product yield from cis-(5S,6R)-A1 is expected to be rather small. Cis-(5S,6R)-B1 is by far the most dominant among the studied diastereoisomers and the H2’ abstraction reaction for cis-(5S,6R)-B1 has a moderate barrier. More importantly, the reaction is slightly exothermic. Hence, the largest amount of products observed

13

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 25

experimentally should be from the intranucleotidyl H2’ abstraction in cis-(5S,6R)-B1.14 In addition, with respect to the difference in reaction barriers and rates as shown in Tables 2 and 3 and the reaction rates in Fig. 3, the C6-yl radical in 5S-(rU)3 is more reactive towards C2’-H2’ bond activation than the peroxyl radical in cis-(5S,6R)-B1. Thus, our present conclusions are completely consistent with the recent experimental findings.8,9,14 Our present study highlights the fact that at the molecular level the efficiency of RNA damage induced by the peroxyl radical is strongly dependent on the locally modified RNA conformation. The peroxyl radical isomer with the highest reaction efficiency has a very small species distribution and should provide only a minor contribution to the overall RNA damage and subsequent product distribution. The peroxyl isomer with the maximum species distribution has a higher reaction barrier and a lower reaction rate, resulting in low-efficiency oxidative damage to RNA. The C6-yl species has an apparently lower barrier and faster reaction rate constant than the peroxyl species with maximum species distribution, showing that the carbon radical formed under anaerobic conditions should be a more competent candidate to RNA damage. Indirect strand scission of RNA is shown to be less prominent based on our present calculations. Indirect strand scission of RNA is initiated from H1’ or H3’ atom abstraction, rather than H2’, by the peroxyl species. It is then followed by H2’ atom transfer to the C1’ or C3’ position to form the C2’ radical, from which the RNA strand scission can then occur.14 Based on our present results (Table S1), the activation barriers of the C3’-H3’ bond in cis-(5R,6S)-A2 and of the C1’-H1’ bond in cis-(5R,6S)-B2 are higher than for H2’ abstraction, hence, the indirect strand scission of RNA appears less likely. Our theoretical results clarify the ambiguous statement issued earlier that RNA damage could be formed from the indirect strand scission 14.

14

ACS Paragon Plus Environment

Page 15 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

In addition, the reactivities of the peroxyl isomers with cis-(5R,6S) chirality were also studied. The relevant data are provided in the supporting material (Fig. S6-S8 and Table S4-S6). It is worth to note that the reaction rate constant of C2’-H2’ bond activation is much smaller in 5R-(rU)3 than in all of its corresponding cis-(5R,6S) peroxyl species (Fig. S4), which is inconsistent with experimental findings.14 However, according to our above studies on the selective addition of OH to nucleic bases in the RNA strand, 5S-(rU)3 and its corresponding cis-(5S,6R) peroxyl species dominate the species distribution, which thus determines the efficiency of the RNA oxidation damage.

3.3 Reasons for the diverse reactivity of the peroxyl diastereoisomers towards C2’-H2’ bond activation It is clear that the internucleotidyl H2’-abstraction barrier in cis-(5S,6R)-A1 is markedly lower than those in the other peroxyl diastereoisomers. This interesting fact prompted us to further explore the nature of the H2’ atom transfer reaction. Previous data indicate that there is a good correlation between hydrogen-atom abstraction reaction efficiencies for the aryl radicals and their vertical electron affinities.43 However, this correlation is not as clear in neither our recent studies of DNA peroxyl species 41 nor in this work. The harmonic stretch vibration frequencies of the C2’-H2’ bond (νC2’-H2’), the absolute reduction potentials (Ered) and the hydrogen atom binding energies (BE) for the four peroxyl diastereoisomers were calculated, and the results are presented in Table 4. The νC2’-H2’ values of cis-(5S,6R)-A1 and A2 are almost identical, which indicates that their C2’-H2’ bond strengths are highly similar. The C2’-H2’ bond strengths in cis-(5S,6R)-B1 and B2 isomers are also almost equal. The absolute reduction potential of cis-(5S,6R)-A1 and B1 are 4.75 and 4.74 eV, respectively, which are higher by

15

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 25

0.1 - 0.3 eV than those of cis-(5S,6R)-A2 and B2 (Table 4). Therefore, neither cis-(5S,6R)-A1 nor B1 has a higher tendency to capture an electron than the cis-(5S,6R)-A2 and B2 isomers. In addition, the hydrogen atom binding energies in cis-(5S,6R)-A1 and B1 were found to be almost identical. The unpaired electron spin densities of the cis-(5S,6R) isomers are mainly localized on Ob, as shown in Table 4. Taken together, we conclude that the peroxyl group in cis-(5S,6R)-A1 and B1 should have the similar reaction efficiency of C2’-H2’ bond activation among the peroxyl diastereoisomers.

Table 4 Single-electron spin density (SD, in e) of Ob, electron affinities (EAs, in eV), hydrogen atom binding energies (BE, in kcal/mol), absolute reduction potentials (Ered, in eV) and the stretching vibrational frequency of the C2’-H2’ bond (νC2’-H2’, in cm-1) in the four isomers cis-(5S,6R)-A1, A2, B1 and B2 in RNA, obtained at the M06-2X/6-311++G(d,p)//M06-2X /6-31G(d,p) level. isomer cis-(5S,6R)-A1 cis-(5S,6R)-A2 cis-(5S,6R)-B1 cis-(5S,6R)-B2

SD 0.69 0.71 0.69 0.72

EA 4.84 (3.97)a 4.57 (3.58) 4.84 (4.06) 4.66 (3.94)

BE

Ered

νC2’-H2’

-88.3 -88.5 -88.8 -87.9

4.75 4.51 4.74 4.63

3183b/3106c 3190/3104 3125/3138 3126/3147

a

Values in parentheses are vertical electron affinities. b,c Stretching vibration frequency of C2’-H2’ in the 5’-adjacent nucleoside (b) and in the middle nucleoside (c), respectively.

The important structural parameters of the reactants and the C2’-H2’ bond activation transition states listed in Table 5. For the reactants, the Oa-Ob bond lengths are almost the same in all four diastereoisomers. The same observation is also made for the C2’-H2’ bond lengths. Based on the ∠OaObH2’ angle in the transition structure, H2’ is clearly seen to be transferred to an Oa-Ob orbital of π/π∗ symmetry. It is well known that the efficiency of C-H bond activation is optimal if the σC-H donor orbital can obtain maximum overlap with the πO-O acceptor orbital. For the present systems, the complex needs to rearrange a vertical attack of the cleaving C2’-H2’ bond toward the Oa-Ob 16

ACS Paragon Plus Environment

Page 17 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

radical fragment such that the overlap between the involved orbitals are maximized. For the isomer cis-(5S,6R)-A1, the ∠C2’H2’ObOa and ∠OaObH2’ values are 82.3° and 101.6°, both of which are much closer to 90° than those observed for the other reactants, cis-(5S,6R)-A2, B1 and B2. This indicates that the cleaving C2’-H2’ bond in cis-(5S,6R)-A1 is more close to vertical attack by the Oa-Ob radical core, i.e., that the largest orbital interaction should be found in cis-(5S,6R)-A1. Hence, the C2’-H2’ bond activation barrier in cis-(5S,6R)-A1 must be smaller than those in cis-(5S,6R)-A2, B1 and B2.

Table 5 Key geometric parameters of the reactants and the C2’-H2’ bond activation transition states (TSH2’). Bond lengths in Å. isomers cis-(5S,6R)-A1a cis-(5S,6R)-A2a cis-(5S,6R)-B1b cis-(5S,6R)-B2b a,b

Reactant

TSH2’

dOa-Ob

dC2’-H2’

∠C2’H2’ObOa

∠OaOb H2’

dOa-Ob

dC2’-H2’

∠OaOb H2’

1.303 1.304 1.302 1.302

1.092 1.092 1.094 1.094

101.6° -173.8° 2.4° 105.3°

82.3° 56.8° 97.6° 104.3°

1.384 1.384 1.383 1.387

1.276 1.319 1.312 1.326

102.2° 106.0° 102.1° 108.7°

H2’ abstraction in an internucleotidyl (a) and intranucleotidyl (b) manner, respectively.

4. Conclusions In this study, the DFT method M06-2X was used to gain insight into the lower efficiency of RNA oxidative damage under aerobic than anaerobic conditions, through comparison of the reactivity of the 5-hydroxyl-2’-uridin-6-yl radical and its corresponding peroxyl species in single-strand RNA. C2’-H2’ bond activation by the peroxyl group of the cis-(5S,6R)-A1, A2, B1 and B2 diastereoisomers is associated with RNA damage in a direct strand scission manner. Harmonic vibrational frequencies of the C2’-H2’ bonds, absolute reduction potentials, binding energies of the

17

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 25

hydrogen atom, and structural parameters of the reactive hotspots were calculated to unveil the reactivity difference among the C6-peroxyl species. Based on our theoretical studies, 5S-r(U)3 formation is favored over 5R-r(U)3, which is seen from the highly selective addition of the OH radical to the RNA strand from the 5’-end. The activation barriers of the C2’-H2’ bond by the peroxyl species derived from 5S-(rU)3 range from 16.6 to 28.7 kcal/mol. The corresponding reaction barrier by the uracil C6-yl radical in 5S-r(U)3 is ca. 20.4 kcal/mol. C2’-H2’ bond activation barriers by peroxyl species are normally higher than those of their corresponding carbon-centered radicals. However, an exception is found for the cis-(5S,6R)-A1 isomer, in that its C2’-H2’ bond activation barrier is much lower than those in the C6-yl species. The direct strand-break products from cis-(5S,6R)-A1 will however not be observed at any great amount since it corresponds to only about 2.2% of the species distribution among the cis-(5S,6R)-A1, A2, B1 and B2 isomers. Instead, the species distribution of the cis-(5S,6R)-B1 isomer is the absolutely dominant one. The cis-(5S,6R)-B1 isomer has a moderate C2’-H2’ bond activation barrier, which however is higher than that of 5S-r(U)3. More importantly, the reaction is slightly exothermic. Hence, intranucleotidyl H2’ abstraction by way of cis-(5S,6R)-B1 should be the preferred one, leading eventually to direct strand scission of RNA. The present theoretical results unveil the strong correlation between the conformation of the locally modified RNA and its oxidative damage efficiency, which governs the distribution of the final products. The present results also explain why the C2’-H2’ bond activation reaction by the uracil C6-yl species is more efficient under anaerobic conditions than under aerobic conditions. The reasons for the smaller activation barrier of the C2’-H2’ bond by the peroxyl group in the cis-(5S,6R)-A1 isomer than those in cis-(5S,6R)-A2, B1 and B2 were also investigated. We note that this isomer exhibits the most favorable orbital interaction between the H2’-donor and its acceptor in 18

ACS Paragon Plus Environment

Page 19 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

the reactant and transition state compared to the other isomers, thereby contributing largely to the decrease in C-H bond activation barrier. Moreover, the cis-(5S,6R)-A1 isomer has the largest absolute reduction potential, which could be another reason why the C2’-H2’ bond activation barrier in cis-(5S,6R)-A1 is the lowest.

Acknowledgements We express our deep thanks to Dr. S.W. Zhang for many helpful discussions. LAE gratefully acknowledges financial support from the Swedish research council (VR) and the Faculty of science at the University of Gothenburg.

Supplementary material See supplementary material for detailed information on rotational energy surfaces, reactivity of the cis-(5R,6S) isomers, and energies obtained at the M06-2X level together with 6-31G(d,p), 6-31+G(d,p) and 6-311++G(d,p) basis sets.

References (1) Thannickal,V.J. ; Fanburg,B. L. Reactive oxygen species in cell signaling. Am. J. Physiol. Lung Cell. Mol. Physiol. 2000, 279, L1005-1028. (2) Evans, M. D.; Dizdaroglu, M.; Cooke,M.S. Oxidative DNA damage and disease: induction, repair and significance. Mutat. Res. 2004, 567, 1-61. (3) Hofer, T. ; Badouard, C.;

Bajak, E.; Ravanat, J.L.; Mattsson, A.; Cotgreave, I. A.

Hydrogen peroxide causes greater oxidation in cellular RNA than in DNA. Biol. Chem. 2005,386, 333-337. 19

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 25

(4) Stadtman, E.R. and Berlett, B.S. Reactive oxygen-mediated protein oxidation in aging and disease. Chem. Res. Toxicol. 1997, 10, 485-494. (5) Zhang, J.; Perry, G.; Smith, M.A.; Robertson, D.; Olson, S. J.; Graham, D.G.; Montine, T. J. Parkinson’s disease is associated with oxidative damage to cytoplasmic DNA and RNA in substantia nigra neurons, Am. J. Pathol. 1999,154,1423-1429. (6) Lovell, M.A.; Markesbery, W. R. Oxidative damage in mild cognitive impairment and early Alzheimer’s disease. J. Neurosci. Res. 2007, 85,3036-3040. (7) Li, M. J.; Liu, L.; Wei, K.; Fu, Y.; Guo, Q. X. Significant effects of phosphorylation on relative stabilities of DNA and RNA sugar radicals: remarkably high susceptibility of H-2’ abstraction in RNA. J. Phys. Chem. B 2006,110, 13582-13589. (8) Jacobs, A. C.; Resendiz, M.J.E.; Greenberg,M. M. Direct strand scission from a nucleobase radical in RNA.J. Am. Chem. Soc. 2010,132, 3668-3669. (9) Paul, R.; Greenberg,M. M. Rapid RNA strand scission following C2’-hydrogen atom abstraction.J. Am. Chem. Soc. 2015,137,596-599. (10) Paul, R.; Greenberg, M. M. Mechanistic studies on RNA strand scission from a C2’-radical. J. Org. Chem. 2016,81,9199-9205. (11) Zhang, R. B.;Eriksson, L. A. Distinct hydroxy-radical-induced damage of 3’-uridine monophosphate in RNA: a theoretical study. Chem. Eur. J. 2009, 15, 2394-2402. (12) McKinlay, A.; Gerard, W.; Fields, S. Global analysis of RNA oxidation in Saccharomyces cerevisiae. Biotechniques. 2012,52,109-111. (13) Goto, M.; Ueda, K.; Hashimoto, T.; Fujiwara, S.; Matsuyama, K.; Kometani, T.; Kanazawa, K. A formation mechanism for 8-hydroxy-2’-deoxyguanosine mediated by peroxidized 2’-deoxythymidine. Free Radical Biology & Medicine 2008, 45, 1318-1325. 20

ACS Paragon Plus Environment

Page 21 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

(14) Jacobs, A. C.; Resendiz, M.J.E.; Greenberg, M. M. Product and mechanistic analysis of the reactivity of a C6-pyrimidine radical in RNA. J. Am. Chem. Soc. 2011, 133, 5152-5159. (15) C. von Sonntag, The Chemical Basis of Radiation Biology, Taylor & Francis, London, 1987. (16) Narayanan, B. C.; Westbrook, J.; Ghosh, S.; Petrov, A. I.; Sweeney, B.; Zirbel, C. L.; Leont, N. B.; Berman, H. M. The Nucleic Acid Database: new features and capabilities. Nucleic Acids Res. 2013, 42, D114-122. (17) Labet, V.; Morell, C.; Grand, A.; Cadet, J.; Cimino, P.; Barone, V. Formation of cross-linked adducts between guanine and thymine mediated by hydroxyl radical and one-electron oxidation: a theoretical study. Org. Biomol. Chem. 2008, 6, 3300-3305. (18) Dupont, C.; Patel, C.; Ravanat, J. L.; Dumont, E. Addressing the competitive formation of tandem DNA lesions by a nucleobase peroxyl radical: a DFT-D screening. Org. Biomol. Chem. 2013,11, 3038-3045. (19) C. von Sonntag, Free-Radical-Induced DNA Damage and Its Repair, Springer-Verlag, Berlin, 2006. (20) Newman, C. A.; Resendiz, M. J. E.; Sczepanski, J. T.; Greenberg, M. M. Photochemical generation and reactivity of the 5,6-dihydrouridin-6-yl radical. J. Org. Chem. 2009, 74, 7007-7012. (21) Frisch, M. J.; Trucks, G.W. ; Schlegel, H.B.; Scuseria, G.E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A. et al , Gaussian 09, Revision A.02, Gaussian, Inc., Wallingford CT, 2009. (22) Zhao, Y.; Truhlar, D. G. Density functionals with broad applicability in chemistry. Acc. Chem. Res. 2008, 41, 157-167. 21

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 25

(23) Zhao, Y.; Truhlar, D. G. The M06 suite of density functionals for main group thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and transition elements: two new functionals and systematic testing of four M06-class functionals and 12 other functionals. Theor. Chem. Acc. 2008,120, 215-241. (24) Kellie,J. L.; Navarro-Whyte, L.; Carvey, M. T.; Wetmore, S. D. Combined effects of π−π stacking and hydrogen bonding on the (N1) acidity of uracil and hydrolysis of 2’-deoxyuridine. J. Phys. Chem. B 2012,116, 2622-2632. (25) Olivella, S.; Sole, A.; Bofill, J. M. Theoretical mechanistic study of the oxidative degradation of benzene in the troposphere: reaction of benzene-HO radical adduct with O2. J. Chem. Theory Comput. 2009,5, 1607-1623. (26) Cances, E.; Mennucci, B.; Tomasi, J. A new integral equation formalism for the polarizable continuum model: Theoretical background and applications to isotropic and anistropic dielectrics. J. Chem. Phys. 1997, 107, 3032-3041. (27) Cossi, M.; Barone, V.; Mennucci, B.; Tomasi, J. Ab initio study of ionic solutions by a polarizable continuum dielectric model. Chem. Phys. Lett. 1998, 286, 253-260. (28) Hratchian, H.P.; Schlegel, H. B. Theory and Applications of Computational Chemistry: The First 40 Years (Elsevier, Amsterdam, 2005) 195-249. (29) Alvarez-Idaboy, J. R.; Mora-Diez, N.; Vivier-Bunge, A. A quantum chemical and classical transition state theory explanation of negative activation energies in OH addition to substituted ethenes. J. Am. Chem. Soc. 2000, 122, 3715-3720. (30) Vega-Rodriguez, A.; Alvarez-Idaboy, J. R. Quantum chemistry and TST study of the mechanisms and branching ratios for the reactions of OH with unsaturated aldehydes. Phys. Chem. Chem. Phys. 2009, 11, 7649-7658. 22

ACS Paragon Plus Environment

Page 23 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

(31) Sun, C. H.; Lv, L. Q.; Zhang, S. W. Theoretical investigation on the reaction mechanism and kinetics of benzyl alcohol with OH radical. Theor. Chem. Acc .2016, 135, 1-9. (32) Miller, W.H. Tunneling corrections to unimolecular rate constants, with application to formaldehyde. J. Am. Chem. Soc. 1979, 101, 6810-6814. (33) Bera, P. P.; Schaefer, H. F. (G-H)*-C and G-(C-H)* radicals derived from the guanine-cytosine base pair cause DNA subunit lesions. PNAS 2005, 102, 6698-6703. (34) Lind, M.C.; Bera, P. P.; Richardson, N. A.; Wheeler, S. E.; Schaefer, H. F. The deprotonated guanine-cytosine base pair. PNAS 2006, 103, 7554-7559. (35) Zhang, R. B.; Eriksson, L. A. Effects of OH radical addition on proton transfer in the guanine-cytosine base pair. J. Phys. Chem. B 2007,111, 6571-6576. (36) Singleton, D. L.; Cvetanovic, R. J. Temperature dependence of the reaction of oxygen atoms with olefins. J. Am. Chem. Soc.1976, 98, 6812-6819. (37) Wagner, J. R.; Vanlier, J. E.; Berger, M.; Cadet, J. Thymidine hydroperoxides: structural assignment, conformational features, and thermal decomposition in water. J. Am. Chem. Soc. 1994, 116, 2235-2242. (38) Schyman, P.; Eriksson, L.A.; Laaksonen, A. Hydrogen abstraction from deoxyribose by a neighboring 3’-uracil peroxyl radical. J. Phys. Chem. B 2009, 113, 6574-6578. (39) Samson-Thibault, F.; Madugundu, G. S.; Gao, S.; Cadet, J. Profiling cytosine oxidation in DNA by LC-MS/MS. Chem. Res. Toxicol. 2012, 25, 1902-1911. (40) Madugundu, G. S.; Cadet, J.; Wagner, J. R. Hydroxyl-radical-induced oxidation of 5-methylcytosine in isolated and cellular DNA, Nucleic Acids Res. 2014, 42, 7450-7460. (41) Zhao,S. ; Zhang, R. B.; Li, Z. S. A new understanding towards reactivity of DNA peroxy radical. Phys. Chem. Chem. Phys. 2016, 18, 23763-23768. 23

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 25

(42) Hong, I. S.; Carter, K. N.; Greenberg, M. M. Evidence for glycosidic bond rotation in a nucleobase peroxyl radical and its effect on tandem lesion formation. J. Org. Chem. 2004, 69, 6974-6978. (43) Jing, L.H.; Nash, J. J.; Kenttamaa, H. I. Correlation of hydrogen-atom abstraction reaction efficiencies for aryl radicals with their vertical electron affinities and the vertical ionization energies of the hydrogen-atom donors. J. Am. Chem. Soc. 2008, 130, 17697-17709.

24

ACS Paragon Plus Environment

Page 25 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

TOC GRAPHIC

25

ACS Paragon Plus Environment