Urea Mimics Nucleobases by Preserving the Helical Integrity of B- DNA Duplexes via Hydrogen Bonding and Stacking Interactions

Transcription

1 Urea Mimics Nucleobases by Preserving the Helical Integrity of B- DNA Duplexes via Hydrogen Bonding and Stacking Interactions Thesis submitted in partial fulfillment of the requirements for the degree of Master of Science by Research in Computational Natural Science By Indrajit Patil INTERNATIONAL INSTITUTE OF INFORMATION TECHNOLOGY (Deemed to be University) Hyderabad , India July, 2016

2 Copyright Indrajit Patil, 2016 All Rights Reserved 2

3 3

4 4

5 5

6 6

7 ACKNOWLEDGEMENT me during my stay here at IIIT. It is my great pleasure to acknowledge the many people who have helped First and foremost, I would like to thank my guide, Dr. U Deva Priyakumar. Sir has been an inspiration for me throughout my stay here and a wonderful guide during my thesis. He patiently addressed all my queries and provided constant support when I needed it the most. I have learnt so many things from him and would try and emulate his very qualities in my life ahead. I would also like to thank all the CCNSB faculty for their constant support and patience during our course work. I would also like to thank all the faculty here at IIIT Hyderabad. In hindsight, the quality of education imparted here is unparalleled and of the highest standard. I would always cherish and respect all the faculty for shaping us into what we are today. They would have a very big role in whatever we achieve in our respective careers ahead. I would like to thank all my lab seniors for helping me and guiding me during the course of my thesis. Suresh sir, Siladitya sir, Pratyusha mam, Koushik sir, Gayathri, Shwetha mam, Sandhya mam, Ramki sir, Ragini, Antarip sir who have supported and guided me at every point. I would also like to thank all my batchmates for making the stay a very memorable one. There have been a lot of fun memories which will last a lifetime. Also, the healthy competition and quality of assignments have helped shape our characters and 7

8 personalities. I would also like to thank Kuldeep and Supriya for motivating me and supporting me during the course of my thesis. I would also like to thank Mr. Girish Byrappa and Mr. Balsantosh the CCNSB administrative assistants for providing assistance. I would also like to thank Mr. Umesh for maintaining an academic atmosphere in the lab by keeping it clean at all times. I would also like to thank my parents for making me whatever I am today. My parents taught me and sacrificed so many things to see me succeed in life. I hope I am able to live up to their expectations and take care of them. 8

9 ABSTRACT One of the DNA lesions formed due to free radical damage of thymine is urea, occurrence of which in DNA blocks DNA polymerases, and has been shown to be lethal. Recently, it has been shown that urea is capable of forming hydrogen bonding and stacking interactions with nucleobases, which are responsible for the unfolding of RNA in aqueous urea. Base pairing and stacking are inherent properties of nucleobases; since urea is able to form both, this study attempts to investigate if urea can mimic nucleobases in the context of the nucleic acid structures by examining the effect of introducing urea lesions complementary to the four different nucleobases on the overall helical integrity of B-DNA duplexes and their thermodynamic stabilities using molecular dynamics (MD) simulations. The MD simulations resulted in stable duplexes without significant changes in the global B-DNA conformation. In agreement with experimental results, the urea lesions occupy intrahelical positions by forming hydrogen bonds with nitrogenous nucleobases. Further, these urea lesions form hydrogen bonding and stacking interactions with other nucleobases of the same and partner strands, which are sometimes similar to the nucleobases in typical B-DNA duplexes. Direct hydrogen bond interactions are observed for the urea-purine pairs within DNA duplexes, whereas two different modes of interactions, namely direct hydrogen bonds and water-mediated hydrogen bonds, are observed for the urea-pyrimidine pairs. The latter explains the complexities involved in interpreting the experimental NMR data reported previously. Binding free energy calculations were further performed to understand the thermodynamic stability of the urea incorporated DNA duplexes with respect to pure duplexes. This study suggests that urea potentially mimics nucleobases by pairing opposite to all the four nucleobases and maintains the overall structure of the B-DNA duplexes. 9

10 10

11 TABLE OF CONTENTS CHAPTER 1: Introduction...17 CHAPTER 2: Computational Methods Quantum Mechanical Calculations Molecular Dynamics Simulations Free Energy Calculations...24 CHAPTER 3: Results and Discussions QM calculations on urea-nucleobase pairs Stable structures obtained from initial simulations Discussions based on long MD simulations...38 CHAPTER 4: Conclusions...57 References

12 LIST OF FIGURES Figure 2.1.1: Model systems of urea-nucleobase pairs and DNA duplexes. (a-g) Representation of the possible hydrogen bonding between urea and all the four nucleobases (A, G, T and C), and (h) the DNA sequence used to model the pure and mismatched B-DNA duplexes Figure 3.1.1: Various conformations depicting hydrogen bonding between urea and the nucleobases. The interaction energies for each of the conformations are also given Figure 3.1.2: Average RMSD values for all the urea incorporated DNA duplexes averaged over the simulation period...27 Figure : Conformations changing to other conformations within a few ns of MD simulations Figure 3.1.4: Times series of root mean square (RMS) deviations for all the duplexes in reference to their respective initial structure Figure 3.1.5: Probability distribution of the two hydrogen bonds between urea and the complementary bases present in all the DNA duplexes Figure 3.1.6: Probability distribution of all base pairs in the double helix except urea-base pair...32 Figure 3.1.7: Probability distributions of backbone alpha angles corresponding to urea and its complementary nucleotides in all the stable urea incorporated DNA duplexes

13 Figure 3.1.8: Probability distributions of backbone beta angles corresponding to urea and its complementary nucleotides in all the stable urea incorporated DNA duplexes Figure 3.1.9: Probability distributions of backbone epsilon angles corresponding to urea and its complementary nucleotides in all the stable urea incorporated DNA duplexes Figure : Probability distributions of backbone gamma angles corresponding to urea and its complementary nucleotides in all the stable urea incorporated DNA duplexes Figure : Probability distributions of backbone zeta angles corresponding to urea and its complementary nucleotides in all the stable urea incorporated DNA duplexes Figure 3.2.1: Time series of root mean square deviations (in Å) of the urea incorporated B-DNA duplexes Figure 3.2.2: Probability distributions of distances (in Å) corresponding to the hydrogen bonds between urea and all the four nucleobases that are present in urea incorporated B-DNA duplexes Figure 3.2.3: Probability distributions of hydrogen bond distances (in Å) of all the base pairs in the urea incorporated B-DNA duplexes during the MD simulations Figure 3.2.4: : Interaction energies of the urea-nucleobase mismatched pairs and their neighbour AT and GC base pairs of all the urea incorporated B-DNA duplexes (blue). The interaction energies corresponding to base pairs in pure DNA are also included for comparison (red) Figure 3.2.5: : Probability distribution of the distances (in Å) between connecting nitrogen atoms of nucleobase or urea with furanose sugar for (a) urea-nucleobase mismatched pairs and 13

14 (b) base pairs next to the mismatched base pair (CG:red and AT:black) present in all the urea incorporated B-DNA duplexes Figure 3.2.6: Two modes of interactions between urea and pyrimidine bases. Depiction of direct hydrogen bonds and water mediated hydrogen bonds of urea with its complementary bases, thymine (a:t2c) and cytosine (b:c2b), during the course of the molecular dynamics simulations Figure 3.2.7: Probability distributions of pseudorotation angles of furanose sugars of urea nucleotides (Ur) and their respective complementary nucleotides (cu) in all the urea incorporated B-DNA duplexes. The probabilities corresponding to Ur and cu are shown in blue and green respectively, while the probability corresponding to pure DNA duplexes is shown in purple Figure 3.2.8: Probability distributions of pseudorotation angles of furanose sugars for the (a) full DNA duplex, (b) full DNA duplex without urea-nucleobase pair, and (c, d) neighbour base pairs in all the urea incorporated DNA duplexes. The pseudorotation angles corresponding to the furanose sugars of typical B-DNA duplexes are shown in black while the urea incorporated DNA duplexes are shown in red Figure 3.2.9: Probability distributions of sugar-base glycosidic angle (in degree) for urea nucleotides (a) and respective complementary nucleobases (b) of all the urea incorporated B- DNA duplexes Figure : Probability distributions of backbone dihedral angles (in degree) corresponding to the urea nucleotides, its complementary nucleotide and their neighbour nucleotides in all the urea incorporated DNA duplexes

15 Figure : Probability distributions of phosphodiester bond angle (in degree) corresponding to the urea nucleotides, its complementary base and their neighbour nucleotides in all the urea incorporated DNA duplexes

16 LIST OF TABLES Table : Stacking interaction energies of urea and its complementary nucleobases with the nucleobases of the same and their complementary strands. All the energies are in kcal mol Table 3.2.2: Number of water molecules present around the mismatched base pair (urea and complementary base) for all the urea incorporated B-DNA duplexes considered in the present study Table 3.2.3: Absolute (ΔG) and relative (ΔΔG) binding free energy values calculated for the duplex formation of B-DNA duplexes considered in the present study. The relative binding free energy values are calculated with respect to their pure DNA duplex. All the energy values are in kcal mol

17 Chapter 1 INTRODUCTION Many exogenous and endogenous agents are capable of damaging cellular DNA, termed as DNA damage, thereby modifying the properties of DNA or triggering mutations. Such modifications have been shown to be responsible for mutagenesis, carcinogenesis and cell lethality [1]. DNA damage includes base and furanose sugar modifications, abasic site formation, single and double strand breaks, and cross linkages [1-3], and these factors play a central role in the etiology of cancer, neurological disorders and aging related diseases [4-8]. One of the most common DNA damages is the formation of thymine glycol (Tg) from thymine upon oxidative addition. Thymine glycol undergoes further ring fragmentation to form N-(2-deoxy-β-D-erythropentofuranosyl) formamide and N-(2-deoxy-β-D-erythro-pentofuranosyl) urea [9-13]. These fragmented products are considered as intermediate structures between nucleobases and abasic sites. These urea lesions were reported as being non-instructive [14, 15] and are known to be mutagenic, similar to abasic sites [16]. Previous studies have shown that urea lesions are capable of forming hydrogen bonds with nucleobases after losing part of their original coding information [9, 10]. Urea lesions are poorly bypassed by DNA polymerase and they block DNA polymerase in in vitro [15, 17]. These urea lesions are recognized by enzymes like endonuclease III, which are known to remove the lesions in Escherichia coli [18]. If these modifications are not repaired, the polymerase can potentially incorporate any base opposite to it, which results in a mutation. It has been observed that the presence of urea and Tg affects the cleavage rate of RNA-DNA hybrids by ribonuclease H, and it preferentially redirects the cleavage site lying next to the mismatch base pair [19]. Previous studies have also shown that the damaged site increases 17

18 the flexibility of duplex DNA that increases protein recognition of DNA during DNA repair mechanisms [20]. NMR studies have been performed on the B-DNA duplexes containing urea or formamide lesion as a bulge or complementary to the nucleobases [21-29]. NMR studies of DNA with urea lesion opposite to the thymine base have indicated that the urea nucleotide adopts cis and trans conformations around the uridic bond, and these two isomers were present in the ratio of 2:3 in solution [23]. It was also observed that urea-thymine base pair has two hydrogen bonds similar to regular AT base pair [23]. Furthermore, the mismatched base pair occupied an intrahelical position and was held by hydrogen bonds with other nucleobases. NMR studies have also reported the structures of B-DNA helix with urea lesion as a bulge opposite to the abasic site, and explained their role in frameshift mutagenesis [26]. This study had also suggested that a kink was produced in the double helix due to the formation of hydrogen bonds with the nucleobases in the same strand [26]. During replication, if the polymerase does not include a nucleobase opposite to the urea lesion, then it creates frameshift mutagenesis. Such a structure was reported in previous studies [27]. But these studies were unable to characterize the individual structures of DNA duplexes with urea lesion in cis and trans isomers due to resonance overlap. Unlike formamide, the urea lesion occupies intrahelical and extrahelical positions because of the capability of urea lesion to form unusual hydrogen bonds. Similarly, apurinic and apyrimidinic abasic sites in DNA duplexes occupy both intrahelical and extrahelical conformations as observed in several experimental studies [30-35]. Recently we have investigated the nonbonded interactions that are responsible for the unfolding of RNA in aqueous urea [36]. It has been shown that urea is capable of forming hydrogen bonding interactions and more interestingly stacking interactions with nucleobases, 18

19 which is responsible for stabilizing the bases in their extrahelical conformation and hence favoring the unfolded states, which is otherwise not possible in presence of only water [36-38]. Hydrogen bonding and stacking interactions between nucleobases are two important phenomena that are responsible for the stability of the duplex structures of DNA. Since urea is capable of forming both, the natural question is how well urea can mimic the nucleobases within the nucleic acid structures. Given that urea is a thymine damaged product and that it is present in DNA structures, this question assumes higher significance in terms of understanding the structures, dynamics, and thermodynamic stabilities of nucleic acid duplexes containing urea lesions. Although NMR experiments provide valuable information about the structures of urea and formamide incorporated B-DNA duplexes, the complete characterization of DNA duplexes with urea lesions has been difficult because of the resonance overlapping of the signals [23, 26]. Molecular dynamics (MD) simulations are, in general, very useful in studying the structure, dynamics and thermodynamic stability of damaged and chemically modified nucleic acid duplexes [39-42]. In the present study, MD simulations have been performed on the B-DNA duplexes with urea lesion opposite to four nucleobases adenine (A), thymine (T), guanine (G), and cytosine (C) in Watson-Crick (WC), Hoogsteen, and sugar edged interactions. Analyses of the MD trajectories suggest that the urea lesions prefer to form WC-like hydrogen bond interactions with the nucleobases, especially purines and can potentially mimic the nucleobases in B-DNA duplexes. 19

20 Chapter 2 COMPUTATIONAL METHODS 2.1 Quantum Mechanical Calculations To investigate the paring nature of urea with nitrogeneous bases, all the canonical and noncanonical orientations like Watson-Crick and Hoogsteen etc were considered between urea and nitrogeneous bases. The initial geometries corresponding to pairs of urea and nucleobases A, G, C, T were generated using GaussView 05 [43] software. All these geometries were optimized using aug-cc-pvdz basis set at MP2 level of theory with resolution of identity (RI) approximation (RI-MP2) in TURBOMOLE 6.2 program [44] and the interaction energies were corrected according to Basis set super position errors (BSSE) method using RI-MP2 level of theory in GAMESS [45] software. The interaction energies were calculated as the difference between energies of urea-nucleobase pair and sum of urea and nucleobase energies. All the ureanucleobase binary complexes from this study were considered for generating the initial configurations of the urea incorporated DNA duplexes. Based on the structures of typical DNA duplex structures, it is known that the distance between nitrogen atoms (N9 for purines and N1 for pyrimidines) connected to the furanose sugars of the nucleobases from the two complementary strands is approximately 9 Å. Based on this, a cut-off distance of 9 ± 2 Å between either of the N atom of urea and N9 of purines or N1 of pyrimidines was used to select urea-nucleobase pairs that are suitable to fit within the B-DNA duplex (Figure 2.1.1a-g). Initial models for urea incorporated DNA duplexes: This selection criterion resulted in thirteen different structures of urea-nucleobase pairs with WC, Hoogsteen and sugar edge-like 20

21 interactions. The 10-mer DNA sequence from the available NMR experimental study on ureathymine base pair [23] was considered as the template sequence for modeling all the pure and urea incorporated B-DNA duplexes (Figure 2.1.1h). The thymine-urea pair present in this structure was replaced by each of the urea-nucleobase pairs resulting in thirteen duplex structures containing the urea lesion (Figure 2.1.1a-g). Corresponding DNA duplexes with the typical GC, CG, AT and TA base pairs were also considered for comparison. All the urea modifications were done by using the SYBYL program (Tripos Inc). 2.2 Molecular dynamics simulations: CHARMM bimolecular simulation program [46] was used to prepare the initial systems of urea incorporated B-DNA duplexes, to setup the MD simulations, to equilibrate the systems and to perform analyses of the MD trajectories. All the production simulations were performed by employing the NAMD biomolecular simulation program [47]. CHARMM36 all-atom parameters [42, 48-50] were used for B-DNA duplex and urea. The topology and parameters corresponding to urea nucleotide were derived from the CHARMM Generalized Force Field (CGenFF) by combining urea, nucleotide and amide parameters [51]. 500-step minimization was performed on each duplex using the steepest descent (SD) method by applying NOE constraints on the hydrogen bonds present between urea and its complementary base in all the B-DNA duplexes. Another 500-step SD minimization was performed on these systems with harmonic restraints on the heavy atoms. Then these systems were overlaid into an orthorhombic waterbox whose dimensions were selected by extending 10 Å beyond the duplex dimensions. Modified TIP3P water model [52] was used for water. 21

22 Figure 2.1.1: Model systems of urea-nucleobase pairs and DNA duplexes. (a-g) Representation of the possible hydrogen bonding between urea and all the four nucleobases (A, G, T and C), and (h) the DNA sequence used to model the pure and mismatched B-DNA duplexes. This sequence is taken from an earlier NMR spectroscopic study on 10-mer DNA duplex with urea-thymine mismatch pair [21]. The terms used for representing the urea incorporated DNA duplexes and throughout the thesis are also mentioned. 22

23 The negative charge of the phosphate groups of DNA duplexes were neutralized by adding sodium ions, and then the systems were minimized by performing 500 steps each of SD and adopted basis Newton Raphson (ABNR) minimizations while harmonically restraining the DNA heavy atoms. A 100 ps MD simulation was performed on each of these systems in NVT ensemble under the same conditions used above. The final coordinates obtained after the equilibration were then used for running production simulations in NPT ensemble. SHAKE algorithm [53] was used to constrain the covalent bonds involving hydrogen atoms and CRYSTAL module [54] in CHARMM was used to represent periodic boundary conditions during equilibration. Particle mesh Ewald (PME) summation method [55, 56] was used to treat the long range electrostatic interactions. Lennard-Jones (LJ) potential was truncated at 14 Å by using a force switch function [57]. An integration time step of 2 fs was used and the coordinates were saved every 5 ps for further analysis. All the urea incorporated DNA duplexes were simulated for 25 ns without any restraints for further equilibration, which were considered as the second stage of equilibration. Further, 100 ns MD simulations were performed on the duplexes in which the urea lesions occupy intrahelical positions. The stable conformations obtained at the end of the 25 ns MD simulations were used to run these simulations (see later). The temperature and pressure were maintained using the Nosé-Hoover thermostat [58] and Langevin piston [59] respectively during the simulations. Four additional simulations of duration 100 ns each were performed on pure DNA duplexes corresponding to AT, GC, TA and CG base pairs for comparison (Figure 2.1.1h). Visual Molecular Dynamics (VMD) program [60] was used for viewing the structures and rendering the images shown in the manuscript. 23

24 2.3 Free energy calculations: The molecular mechanics-generalised Born with surface area (MM-GBSA) method was used to calculate the binding free energies corresponding to formation of duplex from individual strands which uses molecular mechanical energy, solvation energy and entropy of the system [36, 37]. where represents the free energy of the system, represents the molecular mechanical energy which is the sum of all the energy terms in the force field equation, is the solvation free energy which is the sum of electrostatic ( ) and non-polar ( ) solvation free energies, is the absolute temperature and is the molecular mechanical entropy of the system. The electrostatic and non-polar solvation free energies were calculated by using the Generalised-Born molecular volume (GBMV) method and, where = kcal mol -1 Å -2 respectively. The represents the solvent accessible surface area calculated by using a small probe with radius 1.4 Å. The binding free energy can be calculated as the difference in the free energies of the duplex, and the sum of the free energies of the individual components. where represent the free energies of the duplex and individual strands present in the duplex calculated using the above equation. 24

25 Chapter 3 RESULTS AND DISCUSSION 3.1 QM calculations on urea-nucleobase pairs Figure 3.1.1: Various conformations depicting hydrogen bonding between urea and the nucleobases. The interaction energies for each of the conformations in kcal/mol are also given. 25

26 We have performed QM calculations to investigate all possible hydrogen bonded interactions of urea with all the DNA nucleobases and have observed that such hydrogen bonded interactions give rise to stabilization energies in the range of 6 to 21 kcal/mol (Figure 3.1.1). The interaction energies of nucleobases and urea calculated from the above study are comparable to the interaction energies of the usual base pairs. We can thus conclude that urea forms strong hydrogen bonded interactions with all the nucleobases in many of the conformations. To study how urea would interact with nucleobases inside the DNA double helix we introduce a mismatched base pair with urea replacing a nucleobase in one of the strands. Using the selection criteria based on the cut-off distance of 9 ± 2 Å between either of the N atom of urea and N9 of purines or N1 of pyrimidines described above, we modeled the mismatched base pair according to the conformations studied in the QM calculations and performed molecular dynamics simulations to study the stability of the double helix which incorporates the mismatched base pair in various conformations. 3.2 Stable structures obtained from initial simulations Structural Deviations and Dynamics: The propagation of urea incorporated DNA duplexes are assessed by calculating root mean square (rms) deviations. The rms deviations corresponding to all the 13 urea incorporated systems were calculated with respect to their starting conformations along the simulation time (Figure 3.1.2). The rms deviations corresponding to bases and backbone of the duplexes were also calculated. The variations in rms deviations of full duplex, bases and backbone indicate that the structural deviations are small compared to their respective starting structure, indicating close 26

27 behaviour with the initial conformation. Visual inspection of the trajectories showed that the DNA duplex with C2c conformation lost its base pairing interactions and became unstable. This can be seen by the higher rms deviation values corresponding to the duplex with conformation C2c than the other duplexes (Figure 3.1.2). Figure 3.1.2: Average RMSD values for all the urea incorporated DNA duplexes averaged over the simulation period. Thus, this conformation, C2c, is not considered for further analysis. Similar to previous NMR studies, all the urea nucleotides occupied intrahelical position in the duplex. Interestingly, some of the urea-nucleobase pair conformations simply transformed into other conformations within a 27

28 few ns of the simulation time, indicating the stability of the duplexes with certain base pair conformations. The duplexes with conformations A3b, A3c, A6b and A6c converted to other conformations A3a, A5c, A2b and A2c respectively (Figure 3.1.3). Figure 3.1.3: Conformations changing to other conformations within a few ns of MD simulations. So the further analysis is restricted to the duplexes with stable conformations only. Hence the total number of final structures considered for further analysis were reduced from 13 to 8. 28

29 To understand the similarity of the urea incorporated DNA duplexes with typical B-DNA structure, rms deviations with respect to the typical B-DNA duplex conformation were also calculated for all the duplexes. The time series of rmsd plots indicate that the urea incorporated duplexes do not deviate much from the typical B-DNA conformation, indicating the preservation of B-form characteristics of the urea modified duplexes (Figure 3.1.4). The small deviations observed in both backbone and base of all these stable duplexes further support the B-form nature of the duplexes. Figure 3.1.4: Times series of root mean square (RMS) deviations for all the duplexes in reference to their respective initial structure. 29

30 Figure 3.1.5: Probability distribution of the two hydrogen bonds between urea and the complementary bases present in all the DNA duplexes. 30

31 Hydrogen Bonds Between Urea and Nucleobases: The initial structures have two hydrogen bonds between urea and nucleobases (Figure 3.1.1). The time series plots corresponding to these two hydrogen bonds were calculated for all the ureanucleobase pairs present in the urea incorporated duplexes studied here (Figure 3.1.5). As shown in Figure 3.1.5, the two hydrogen bonds between urea and cytosine paired in C2c conformation are not stable as seen from the high bond distances (>4.5 Å), indicating instability of the duplex in this conformation. In support with the rmsd values, the variations observed for the distances corresponding to the two hydrogen bonds present in A3b, A3c, A6b, and A6c structures indicate their transformation into other conformations. The remaining structures have at least one stable hydrogen bond throughout the simulation, indicating the stability of the urea-nucleobase base pairs. It is also noticed that frequent formations and breakages of the hydrogen bonds are observed in urea-pyrimidine base pairs while such kind of transformations are not observed for urea-purine base pairs. It is seen throughout the simulation time that the hydrogen bonds between the remaining base pairs in the double helix are not affected and are like the purine-pyrimidine base pairing in the usual DNA double helix as shown by the probability distributions in the Figure Conformation of the Urea Nucleotide Sugar and Backbone: The effect of the urea nucleotides on the conformational properties of the DNA duplexes were examined in terms of pseudorotation angles and backbone dihedral angles such as α, β, χ, ε, γ, δ. 31

32 Figure 3.1.6: Probability distribution of all base pairs in the double helix except urea-base pair. The deoxyribose sugars generally show high populations in C2'-endo and small populations in O4'-endo regions. But the incorporated urea strongly influences this ratio and increases populations in C2'-endo region. Considerable changes are not seen in the conformation of the furanose sugars of bases complementary to the urea nucleotides. 32

33 Figure 3.1.7: Probability distributions of backbone alpha angles corresponding to urea and its complementary nucleotides in all the stable urea incorporated DNA duplexes. Here the Nc - Base complementary to urea, N1b - Base above Nc, N2b - Base below Nc, N1a - Base above Urea, and N2a - Base below urea The urea incorporation does not induce considerable changes in the backbone conformation revealed by the probability distributions of the characteristic backbone dihedral angles corresponding to urea nucleotide and its neighbour nucleotides (Figures ). 33

34 Figure 3.1.8: Probability distributions of backbone beta angles corresponding to urea and its complementary nucleotides in all the stable urea incorporated DNA duplexes. Here the Nc - Base complementary to urea, N1b - Base above Nc, N2b - Base below Nc, N1a - Base above Urea, and N2a - Base below urea 34

35 Figure 3.1.9: Probability distributions of backbone epsilon angles corresponding to urea and its complementary nucleotides in all the stable urea incorporated DNA duplexes. Here the Nc - Base complementary to urea, N1b - Base above Nc, N2b - Base below Nc, N1a - Base above Urea, and N2a - Base below urea 35

36 Figure : Probability distributions of backbone gamma angles corresponding to urea and its complementary nucleotides in all the stable urea incorporated DNA duplexes. Here the Nc - Base complementary to urea, N1b - Base above Nc, N2b - Base below Nc, N1a - Base above Urea, and N2a - Base below urea 36

37 Figure : Probability distributions of backbone zeta angles corresponding to urea and its complementary nucleotides in all the stable urea incorporated DNA duplexes. Here the Nc - Base complementary to urea, N1b - Base above Nc, N2b - Base below Nc, N1a - Base above Urea, and N2a - Base below urea 37

38 3.3 Discussions based on long MD simulations The structural deviations of urea incorporated duplexes were assessed by calculating root mean square deviations with respect to their initial conformation. Rmsd values corresponding to all the eight urea incorporated systems were calculated with respect to their starting conformations. The deviations were considerably small for all the duplexes within 25 ns simulation time. Interestingly, the non-wc urea-nucleobase pairs either transformed into WC conformations or simply move out of the duplex within few nanoseconds of the simulation time. Elimination of the structures based on the stability and helical position of the urea nucleotides decreased the number of structures under study from 8 to 5. These five structures are A2b, G1a, G1c, C2b and T2c, on which a new set of simulations for 100 ns each were performed. The current section describes the results of these five urea incorporated duplexes obtained from the 100 ns MD simulations. Structure and dynamics of urea incorporated B-DNA duplexes: The structural closeness of the damaged DNA duplexes was examined by calculating the rmsd with respect to their initial conformation that resembles a typical B-form DNA. As shown in Figure 3.2.1, the rmsd values indicate that the structural deviations are small compared to their respective starting structure, indicating their similarity with an ideal B-form DNA. Visual inspection of MD snapshots indicates that the urea lesions are placed well inside the duplex for all the urea incorporated B-DNA during the course of these simulations. Furthermore, the rmsd values were calculated with respect to typical B-DNA conformation, and the small values observed indicate that the presence of urea does not affect 38

39 their overall structure significantly. This suggests that the B-DNA duplexes preserve their global characteristics after the incorporation of urea lesions opposite to all the four nucleobases. High deviations were observed for short time within these simulations, especially for T2c and C2b, which are attributed to local structural transitions that are discussed later. The simulations study resulted in stable duplexes that are similar to the structures observed in earlier NMR spectroscopy studies on DNA duplex with urea complementary to thymine base [23]. Figure 3.2.1: Time series of root mean square deviations (in Å) of the urea incorporated B-DNA duplexes. These values are calculated with respect to their initial conformation used to run the simulations. 39

40 Hydrogen bonding interactions between urea and nucleobases: Previous studies have shown that urea has strong capability to form hydrogen bonds with the nucleobases [40-42]. There are two potential hydrogen bonding sites on the urea molecule to pair with the nitrogenous bases of the complementary strand. These are oxygen atom on the carbonyl group of urea and nitrogen atom of the amino group which is not attached to the deoxyribose sugar of the double helix. In the beginning, all the urea-nucleobase pairs have two hydrogen bonds between urea and nucleobases (Figure 2.1.1a-g). Stability of these hydrogen bonds was monitored by calculating the hydrogen bond distances corresponding to the donor/acceptor atoms involved in hydrogen bonding between urea and nucleobases during the simulations. Figure shows the probability distributions of such hydrogen bond distances for all the ureanucleobases pairs present in the urea incorporated duplexes. These distributions show interesting trends that differentiate the urea-purine and urea-pyrimidine pairs. In the duplexes containing urea-purine pairs, a strong peak at 3.5 Å suggests that at least one stable hydrogen bond is observed during the simulations in spite of deviations in the second hydrogen bond (Figure 3.2.2). The probability distributions of G1c indicate the GUA-urea pair to be alternating between two-hydrogen bonded state and a bifurcated hydrogen bonded state. In case of urea-pyrimidine pairs, the peak at 3.5 Å corresponds to the direct hydrogen bond interactions between urea and nucleobase, whereas the peak at distance 5.5 Å corresponds to water mediated hydrogen bond formation between urea and nucleobase. This suggests that, while urea interacts with the purine bases through direct hydrogen bonding interactions, urea-pyrimidine pairs alternate between direct hydrogen bonding and water mediated interactions within the duplex. The nature of such water mediated interaction between the pyrimidine nucleobase and urea are discussed later. The base pairs other than the mismatched urea and its neighbors are reasonably stable with WC base 40

41 pairing (Figure 3.2.3) indicating that inclusion of urea lesion in the B-DNA duplexes does not disturb their structural integrity. Figure 3.2.2: Probability distributions of distances (in Å) corresponding to the hydrogen bonds between urea and all the four nucleobases that are present in urea incorporated B-DNA duplexes. The hydrogen bonding between urea and nucleobase are shown in a separate panel on the right hand side. The color scheme used to represent probability distribution lines is same as the color scheme of hydrogen bonds between the urea-nucleobase pairs. 41

42 Figure 3.2.3: Probability distributions of hydrogen bond distances (in Å) of all the base pairs in the urea incorporated B-DNA duplexes during the MD simulations. Energetics of the hydrogen bonding and stacking interactions between urea and nucleobases: Base pairing and stacking interactions among the nucleobases are important for the stability of nucleic acid duplexes. The strength of these interactions in the presence of the urea lesion were calculated and compared to the regular DNA duplexes. The mean interaction energies of urea 42

43 with its complementary WC base are shown in Figure The interaction energy values in the range from ~ -5 to -15 kcal/mol suggest that urea strongly interacts with all the nucleobases though most of the energies are less favorable compared to those of the regular A-T and G-C base pairs (~ -14 and -24 kcal/mol for the canonical AT and GC base pairs) of a typical B-DNA duplex. The effect of the presence of urea on the strength of neighboring base pairs was also evaluated (Figure 3.2.4), which indicate no significant change in the interaction energies due to the inclusion of urea lesion. Though the interaction energies between the nucleobases and urea are comparable or less favorable compared to canonical AT base pair, they are significant and seem strong enough to stabilize these duplexes. Stacking interactions of urea moiety with the nitrogenous bases present above and below in the same strand and complementary strand were also calculated, and are shown in Table The energy values indicate that the urea moiety present in DNA duplexes forms significant stacking interactions with the neighboring bases. However, similar to the hydrogen bonding interactions, the magnitudes of stacking interaction energies of urea moiety are less favorable than that observed in the control simulations. However, these interactions are sufficient to keep the urea lesions inside the double helix in a manner characteristic of regular bases. Such strong stacking interactions between urea and nucleobases were also observed in previous quantum chemical and MD simulation studies in the context of RNA unfolding in presence of aqueous urea [36-38]. The incorporation of urea lesion leads to small deviations in the stacking interaction energies of nucleobases complementary to urea (Table 3.2.1). Therefore, the energy values from hydrogen bond interactions and stacking interactions though less favorable than in canonical DNA duplexes suggest that urea can potentially replace nucleobases in DNA duplexes without distorting their overall structures. The 43

44 effect of incorporating a urea lesion on the overall stability of the duplex structures is discussed later. Figure 3.2.4: Interaction energies of the urea-nucleobase mismatched pairs and their neighbour AT and GC base pairs of all the urea incorporated B-DNA duplexes (blue). The interaction energies corresponding to base pairs in pure DNA are also included for comparison (red). All the interaction energies are in kcal mol -1. Smaller size of urea compared to the purine base leads to interesting hydrogen bond dynamics: All canonical base pairs in nucleic acids are formed between a purine and a pyrimidine nucleobase. 44

45 Table 3.2.1: Stacking interaction energies of urea and its complementary nucleobases with the nucleobases of the same and their complementary strands. All the energies are in kcal mol -1. Duplex Urea Complementary B-DNA ± ± 0.0 A2b -5.6 ± ± 0.0 B-DNA ± ± 0.0 G1a -5.6 ± ± 0.1 G1c -7.3 ± ± 0.0 B-DNA ± ± 0.0 T2c -6.4 ± ± 0.1 B-DNA ± ± 0.1 C2b -5.2 ± ± 0.2 In the duplexes considered here, one of the nucleobases is replaced by the urea moiety. While the size of the urea moiety and the positions of the hydrogen bonding groups are comparable to those of the pyrimidine base, there exists a huge mismatch in terms of the size when it comes to replacing a purine base by urea. As discussed above, at least one conventional hydrogen bond is formed between the pairs involving urea and GUA/ADE. In case of urea-cyt/thy pairs, the sugar moieties have to move towards each other to form hydrogen bonds. The distance between the N9 of purine and N1 of pyrimidine corresponding to a typical base pair is found to be approximately 9 Å. The distance between the N atom of urea covalently bonded to the sugar moiety and the paired base (N9 for purine and N1 for pyrimidine) were calculated and the distributions are given in Figure 3.2.5a. 45

46 Figure 3.2.5: Probability distribution of the distances (in Å) between connecting nitrogen atoms of nucleobase or urea with furanose sugar for (a) urea-nucleobase mismatched pairs and (b) base pairs next to the mismatched base pair (CG:red and AT:black) present in all the urea incorporated B-DNA duplexes. The distributions involving urea-gua/ade pairs sample the distance close to 9 Å with minor deviation for G1a. However, for both the urea-thy and urea-cyt pairs, two distinct states centered at approximately 7.5 and 9.5 Å are sampled with no significant probability in the intermediate distances. Careful analysis of the trajectories indicates that they exist in dynamic 46

47 equilibrium between two states corresponding to these distances. Time series plot of the distances indicate that the systems alternate between these two states in the ns timescale. Analysis of the structures reveal that the state corresponding to a distance close to 7.5 is when there is direct hydrogen bond between the nucleobase and urea, whereas the other state corresponds to the state where the two moieties interact via bridging water molecules (Figure 3.2.6). Figure 3.2.6: Two modes of interactions between urea and pyrimidine bases. Depiction of direct hydrogen bonds and water mediated hydrogen bonds of urea with its complementary bases, thymine (a:t2c) and cytosine (b:c2b), during the course of the molecular dynamics simulations. Such dynamics will lead to detection of the imino protons of this pair difficult/impossible. Notably, previous NMR studies had difficulties in studying nucleic acid structures with urea- 47

48 nucleobase pairs possibly because of this effect. This phenomenon also is expected to alter the solvation dynamics of the nucleic acid significantly [26, 28]. To understand this further, the number of water molecules present around the urea:nucleobase pair were calculated. Table shows the average numbers of water molecules around the urea-nucleobase pairs, and these values indicate that more number of water molecules are present around the urea-pyrimidine base pairs compared to urea-purine base pairs. This is because of the difference in mode of interaction of urea lesion with purine and pyrimidine bases. As discussed above, urea lesion forms direct hydrogen bonds with purine bases while for urea-pyrimidine base pairs, water mediated hydrogen bonds are also observed in addition to direct hydrogen bonds and an equilibrium between these two states (Figure 3.2.6). No such water mediated hydrogen bond formations are found when the urea lesion forms hydrogen bonds with purine nucleobases. This further indicates that water molecules play a significant role in stabilizing the urea-pyrimidine base pairs in DNA double helix. Table 3.2.2: Number of water molecules present around the mismatched base pair (urea and complementary base) for all the urea incorporated B-DNA duplexes considered in the present study. Duplex Hydration Number A2b ± 0.04 G1a ± 0.04 G1c ± 0.02 T2c ± 0.04 C2b ±

49 Figure 3.2.7: Probability distributions of pseudorotation angles of furanose sugars of urea nucleotides (Ur) and their respective complementary nucleotides (cu) in all the urea incorporated B-DNA duplexes. The probabilities corresponding to Ur and cu are shown in blue and green respectively, while the probability corresponding to pure DNA duplexes is shown in purple. 49

50 Local and global conformational properties of B-DNA duplexes after the incorporation of the urea nucleotide: Conformational properties of the DNA duplexes and changes in their conformation after the incorporation of urea lesion were examined by calculating the pseudorotation angles and characteristic backbone dihedral angles such as α (O3 -P-O5 -C5 ), β (P-O5 -C5 -C4 ), γ (O5 - C5 -C4 -C3 ), ε (C4 -C3 -O3 -P), and δ (C3 -O3 -P-O5 ). The probability distributions of pseudorotation angles indicate that the conformation of the furanose sugars of urea nucleotide depends on its complementary base (Figure 3.2.7). The deoxyribose sugars generally sample populations majorly in C2'-endo and minor populations in O4'-endo regions [61, 62]. The furanose sugars corresponding to urea lesions opposite to guanine and cytosine predominantly sample in C2'-endo regions whereas furanose sugars corresponding to urea lesion complementary to adenine and thymine nucleobases sample conformations in C2'- endo and C3'-endo regions. The changes in pseudorotation profiles of furanose sugars of urea lesion are restricted to local changes and do not alter the global conformation of B-DNA duplexes (Figure 3.2.8a and b). Similar conformations have been observed in NMR studies for the furanose sugars corresponding to the modified sites in B-DNA duplexes with urea or formamide lesions [26, 28]. Sampling of the population in O4 -endo region increases for the furanose sugars of nucleobases that are opposite to urea lesion. To understand the conformation of the urea and complementary nucleotides, glycosidic dihedral angles corresponding to sugar-base or sugar-urea bond were calculated. 50

51 Figure 3.2.8: Probability distributions of pseudorotation angles of furanose sugars for the (a) full DNA duplex, (b) full DNA duplex without urea-nucleobase pair, and (c, d) neighbour base pairs in all the urea incorporated DNA duplexes. The pseudorotation angles corresponding to the furanose sugars of typical B-DNA duplexes are shown in black while the urea incorporated DNA duplexes are shown in red. Probability distributions shown in Figure 3.2.9a indicate that all the urea lesions exhibit anti conformation except for guanine in G1a conformation, in which urea exhibits both syn and anti-conformations. As shown in Figure 3.2.9b, the probability distributions indicate that the 51

52 complementary purine bases sample conformations that are sampled by the bases in the regular B-DNA duplex. However, the pyrimidine bases sample regions corresponding to regions sampled by bases in regular B-form and A-form duplexes. This might be associated with the change in pseudorotation profile of the furanose sugar that requires rearrangement in the orientation of sugar-base glycosidic dihedral angle. The urea incorporation does not induce considerable changes in the backbone conformation as revealed by the probability distributions of the characteristic backbone dihedral angles corresponding to the urea nucleotide and its neighboring nucleotides (Figure ). Figure 3.2.9: Probability distributions of sugar-base glycosidic angle (in degree) for urea nucleotides (a) and respective complementary nucleobases (b) of all the urea incorporated B- DNA duplexes. 52

53 This suggests that incorporation of urea lesion does not alter the backbone conformation of the attached DNA strand significantly. The incorporation of the urea lesion preserves the B- form characteristics of the DNA duplexes, which is in agreement with previous studies [23]. Figure : Probability distributions of backbone dihedral angles (in degree) corresponding to the urea nucleotides, its complementary nucleotide and their neighbour nucleotides in all the urea incorporated DNA duplexes. 53

54 Figure : Probability distributions of phosphodiester bond angle (in degree) corresponding to the urea nucleotides, its complementary base and their neighbour nucleotides in all the urea incorporated DNA duplexes. Thermodynamic stability of urea incorporated duplexes: To understand the effect of incorporation of urea lesion on the thermodynamic stability of DNA duplexes, binding free energies corresponding to duplex formation were calculated using MM-GBSA method (Table 3.2.3). 54

55 Table 3.2.3: Absolute ( G) and relative ( G) binding free energy values calculated for the duplex formation of B-DNA duplexes considered in the present study. The relative binding free energy values are calculated with respect to their pure DNA duplex. All the energy values are in kcal mol -1. Duplex G G (a) Adenine Pure DNA A2b (b) Guanine Pure DNA G1a G1c (c) Thymine Pure DNA T2c (d) Cytosine Pure DNA C2b The relative binding free energy values are shown with respect to their corresponding pure DNA duplexes, and these values indicate that the urea incorporation marginally decreases the thermodynamic stabilities of B-DNA duplexes. It is also seen that the extent of destabilization depends on the size of the nucleobase base complementary to urea lesion. 55

56 Reduction of duplex stability is high when the urea lesion replaces purine bases compared to pyrimidine bases. This is because the replacement of purine bases results in urea-pyrimidine base pairs which are stabilized by two modes of interactions (direct and water mediated hydrogen bonds) during the base pair formation. The stabilization due to bridging water molecules are not accounted in the free energy calculation due to the methodological limitations. However, if urea replaces pyrimidine bases, it results in a urea-purine base pair that is stabilized by direct hydrogen bond interactions between urea and purine bases. This kind of mode of interaction might be the factor responsible for the lower impact on the thermodynamic stabilities of DNA duplexes when replacing pyrimidine bases compared to purine bases. This indicates that urea can potentially replace the pyrimidine bases in purine-pyrimidine base pairs and mimic the pyrimidine without significantly altering the duplex properties. Future studies can possibly include DNA duplexes with homo AT and GC base pairs in which all the pyrimidine bases are replaced by urea. This can provide further support for the pyrimidine-mimicking nature of urea. 56