Enhanced Tetrahedral Ordering of Water ... - ACS Publications

Feb 19, 2010 - The Journal of Physical Chemistry B 2017 121 (29), 7016-7026 ... Molecular Dynamics Study of the Role of the Spine of Hydration in DNA ...
0 downloads 0 Views 2MB Size
Download PDF
J. Phys. Chem. B 2010, 114, 3633–3638

3633

Enhanced Tetrahedral Ordering of Water Molecules in Minor Grooves of DNA: Relative Role of DNA Rigidity, Nanoconfinement, and Surface Specific Interactions Biman Jana,† Subrata Pal,‡ and Biman Bagchi*,†,§ Solid State and Structural Chemistry Unit, Indian Institute of Science, Bangalore - 560012, India., Department of Chemistry, Indian Institute of Technology, Gandhinagar, Ahmedabad - 382424, India, JNCASR, Jakkur, Bangalore 560064, India ReceiVed: August 4, 2009; ReVised Manuscript ReceiVed: NoVember 13, 2009

Confinement and surface specific interactions can induce structures otherwise unstable at that temperature and pressure. Here we study the groove specific water dynamics in the nucleic acid sequences, poly-AT and poly-GC, in long B-DNA duplex chains by large scale atomistic molecular dynamics simulations, accompanied by thermodynamic analysis. While water dynamics in the major groove remains insensitive to the sequence differences, exactly the opposite is true for the minor groove water. Much slower water dynamics observed in the minor grooves (especially in the AT minor) can be attributed to an enhanced tetrahedral ordering (〈th〉) of water. The largest value of 〈th〉 in the AT minor groove is related to the spine of hydration found in X-ray structure. The calculated configurational entropy (SC) of the water molecules is found to be correlated with the self-diffusion coefficient of water in different region via Adam-Gibbs relation D ) A exp(-B/TSC), and also with 〈th〉. I. Introduction Water molecules in many of the natural and biological systems (aqueous DNA and proteins, micelles and reverse micelles, etc.) face two distinct, external influences: (a) confinement and (b) surface specific interactions.1 The modification of water structure and dynamics can thus be a combined effect of both. Recent studies of Fayer and co-workers on the water relaxation inside the reverse micelles (RM) suggest that confinement has a larger effect than surface specific interactions.2-4 Computer simulations5,6 and other experimental studies7,8 have also addressed the role of surface interactions in the observed slow dynamics in these systems. In particular, the variation of relaxation time with the size of the micelle is well-documented. The study of the effect of confinement on the structure and dynamics of liquids has a long history. Several groups have reported that a confined liquid behaves dynamically as a bulk liquid but at a lower temperature, indicating that intermolecular correlations increase due to confinement. Tanaka and co-workers have reported that the nature and also the liquid-liquid transition temperature of triphenyl phosphite depend strongly on confinement.9 Increase of intermolecular correlations due to confinement might originate from the choice of low energy local structure. In a different context, they have observed the formation of low energy locally favorable structures in colloidal fluids, gels, and glasses, which explains the characteristic slow dynamics observed in these systems.10-12 In this context, B-DNA duplex provides us a unique opportunity to study these two effects. Here we can compare the results of confinement by studying dynamics at both major and minor grooves, which offer confined domains of varying sizes. We can additionally change interaction with water by choosing different sequences, like poly-GC and poly-AT. Therefore, we * Corresponding author. E-mail: [email protected]. † Solid State and Structural Chemistry Unit, Indian Institute of Science. ‡ Department of Chemistry, Indian Institute of Technology. § JNCASR.

can vary and monitor the effects of interactions and confinement on the dynamics. Water dynamics in DNA grooves is otherwise also a problem of great interest because many drugs and small ligands bind to AT rich minor grooves, for reasons that are beginning to be understood.13,14 Several experimental techniques, such as NMR, time domain fluorescence Stokes shift have earlier provided valuable estimates of the residence times of water molecules in the grooves of DNA. NMR can distinguish between different grooves, but its time resolution is rather limited and can only detect an average behavior while dynamics at a heterogeneous surface is often nonexponential. While solvation dynamics has a much better temporal resolution, its spatial resolution is rather limited. Berg and co-workers have found an interesting slow component in the solvation dynamics of an excited Coumarin102 dye probe embedded in a 17-mer duplex.15,16 This slow component can be fitted to a logarithmic time dependence over three decades of time (40 ps to 40 ns), indicating a complex dynamics.15-17 A recent study devoted to the solvation dynamics of DNA hydration have reported somewhat different results compared to the results obtained by Berg and co-workers.18 However, the difference could possibly be explained by considering the different size of the probe used in the two studies. In the study of Berg et al. a small dye (Coumarin102) was used15,16 while Furse and Corcelli used a larger Hoechst dye.18 It is observed that for a dye larger in size, two layers of hydration water might not be sensitive to the slowness of water molecules in the minor grooves.18 Our study shows that water dynamics in the major groove remain almost completely insensitive to the sequence specificity. Exactly the opposite is found to be true for the minor groove. Water dynamics is much slower in the minor groove than that in the major groove for both AT and GC sequences. We find that the dynamics of water molecules is slowest in the AT minor grooves, which can be explained in terms of an enhanced tetrahedrality of water molecules in the AT minor grooves due to largest confinement.

10.1021/jp907513w  2010 American Chemical Society Published on Web 02/19/2010

3634

J. Phys. Chem. B, Vol. 114, No. 10, 2010

Jana et al.

In supercooled liquid studies, one finds that the diffusion constant is correlated with configurational entropy, by the following Adam-Gibbs relation,

(

D ) A exp -

B TSC

)

(1)

where T is the absolute temperature, SC is the configurational entropy, and A and B are constants. We also calculate the entropy of the water molecules by using the 2PT method19,20 and find that the slowest water dynamics in the AT minor groove can be correlated with the lowest configurational entropy among all the five systems (water in the GC and AT major and minor grooves and in the bulk). In fact, both the orientational relaxation time and the self-diffusion coefficient of water are found to be surprisingly well-correlated with the configurational entropy and follow the Adam-Gibbs relation (eq 1) closely. Since both the major and minor grooves are separately connected among themselves, they act as channels. We have, on the average, 440 water molecules in the major grooves, 220 water molecules in the GC minor grooves, and 140 water molecules in the AT minor groove. These are sufficiently large numbers to define average properties and time correlation functions.

Figure 1. Groove structure of poly-GC (a) and poly-AT (b) sequences, exhibiting major (M) and minor (m) grooves. The snapshots are taken from the present simulations. Note the narrower and deeper minor groove (m) of poly-AT sequence than that of the poly-GC sequence.

II. Systems and Simulation Details The 38 base pair long sequences used in the present simulations are d(GGAAAAAAAAAAAAA AAAAAAAAAAAAA AAAAAAAAGG) and d(GGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGG). These sequences are termed as poly-AT and poly-GC, respectively. The sequences have been generated using the “nucgen” module of the AMBER software.21 All MD simulations reported in this article used the AMBER7 software package with the all-atom AMBER95 force field (FF).22,23 The electrostatic interactions were calculated with the particle mesh Ewald (PME) method.24 Using the LEAP module of the AMBER software, each sequence was immersed in the TIP3P water box. The box dimensions were chosen to ensure a 10 Å thick solvation shell around the DNA structure. In addition, some water molecules were replaced by Na+ counterions to neutralize the negative charge on the phosphate groups of the backbone of the DNA. These procedures resulted in solvated structures, containing around 25 000 atoms, which include ∼2400 DNA atoms, 74 counterions, and ∼7550 water molecules for both sequences. The solvated structures were then subjected to 1000 steps of steepest descent minimization of the potential energy, followed by 2000 steps of conjugate gradient minimization. During this minimization the DNA molecule was kept fixed in its starting conformations using harmonic constraints. This allowed the water molecules to reorganize to eliminate bad contacts with the DNA molecule. The minimized structure was then subjected to 500 ps of molecular dynamics (MD), using a 2 fs time step for integration. Subsequently, MD was performed under constant pressureconstant temperature conditions (NPT) with temperature regulation achieved using the Berendsen weak coupling method (0.5 ps time constant for heat bath coupling and 0.2 ps pressure relaxation time). This was followed by another 5000 steps of conjugate gradient. We then carried out 100 ps of unconstrained NPT MD to equilibrate the system. Finally, for analysis of structures and properties, we carried out 15 ns of NVT MD simulation for both sequences.

III. Results and Discussions In Figure 1 we have shown the snapshot of poly-GC and poly-AT DNA to show the groove structure in these two sequences. Clearly the major groove is wider for both sequences (width ∼ 10-12 Å). A large difference exists between the minor groove structures of the two sequences. The minor groove of the AT sequence is narrower (width ∼ 3.5 Å) and deeper than that of the poly-GC sequence (width ∼ 6 Å). This implies that water molecules present in the minor groove of the AT sequence will be more confined compared to other types of grooves. We now discuss the results, with analysis, on water structure and dynamics in the grooves. A. Tetrahedral Ordering in Groove Water. We have calculated the O-O-O angle distribution of the water molecules in the major and minor grooves of both sequences. This angle distribution gives information about the ordering among water molecules. The distribution has two characteristic peaks. The peak at ∼100° in water directly probes the amount of tetrahedrality present in the system.25 However, the peak at ∼60° probes the amount of interstitial water present in the system. In Figure 2a, we plot the O-O-O angle distribution of water molecules present in the minor grooves of the poly-AT sequence along with the bulk water distribution. Note the enhancement of the peak at ∼100° (the peak is shifted to ∼120°) for the AT minor groove water compared to bulk water distribution. The shift is due to the formation of an ice-like structure, which is a trademark of the spine of hydration. This provides clear evidence of the presence of strong tetrahedral arrangement of water molecules in the AT minor grooves. However, we find a reduction of peak value for the interstitial water molecules (∼60°) in the AT minor groove as compared to that for bulk water. This enhancement of the tetrahedral ordering in the minor grooves of the poly-AT sequence can be understood as a consequence of strong confinement, combined with surface effects. This result is also in agreement with the observation of the spine of hydration along the AT minor grooves.

Water Molecules in Minor Grooves of DNA

J. Phys. Chem. B, Vol. 114, No. 10, 2010 3635

〈(

)〉

12 9 cos(θOOO) + 4 3 9 π 12 cos(θOOO) + P(θOOO) dθOOO ) 10 4 3

〈th〉 ) 1 -

∫

(

)

(2)

Figure 2. O-O-O angle distribution among the water molecules in the AT minor groove (a), GC minor groove (b), and major grooves (c) region of both sequences. Note the enhanced peak at ∼100° for the AT minor groove water, indicating larger tetrahedral order among them.

In Figure 2b, we show the O-O-O distribution of water molecules present in the minor grooves of the poly-GC sequence along with the bulk water distribution. We find no such enhancement of the peak ∼100° as compared to the AT minor groove, indicating that the water molecules in the GC minor groove cannot form a strong tetrahedral network. These results can be understood as follows. The presence of a strong intrabase interaction in the GC strand (3 hydrogen bonds between G and C) makes the groove structure quite rigid, and hence the interaction of water with the DNA atoms is frustrated. The DNA atoms do not undergo deformations in position to accommodate the preferred orientation of water to form a stable tetrahedral structure. The opposite is true for the AT minor groove (2 hydrogen bonds between A and T) where the intrabase interaction is less compared to that for the GC strand. Here the groove structure can be deformed in such a way that water molecules can attain their preferred orientation to form stable tetrahedral structure. In Figure 2c, we present the O-O-O distribution of water molecules present in the major grooves of both sequences. We find hardly any difference between the two distributions. This signifies that the water molecules are structurally quite similar in the major grooves of both sequences. Next, we have quantified the average tetrahedral order parameter, 〈th〉, for water molecules in the different region using the relation,26

Here θOOO is the O-O-O angle and P(θOOO) is the distribution. We find the values of 〈th〉 are 0.41, 0.47, 0.48, 0.52, and 0.57 for bulk water, major groove water of poly-GC, major groove water of poly-AT, minor groove water of poly-GC, and minor groove water for poly-AT, respectively. Thus tetrahedral ordering in the groove of DNA increases with increasing confinement. Note that for low density liquid (LDL) at T ) 230 K, the value of the tetrahedral order parameter 〈th〉 is found to be 0.8. This difference in the structure of water molecules present in the different regions of the poly-AT and poly-GC sequences provides microscopic explanation of the observed thermodynamics and dynamics of water molecules in these regions, as will be discussed later. B. Entropy of Groove Water. We have calculated the entropy of the water molecule in the different grooves of the two DNA sequences using the recently developed 2PT method.19,20 The main input for calculation of entropy using this method is the velocity autocorrelation functions (CV(t)) of the groove water molecules. Details of the method can be found elsewhere.19,20 It is believed that the 2PT method provides a reliable estimate of the entropy of water. The computed entropy values along with the diffusion coefficients are listed in Table 1. The calculated entropy values for major groove water of both sequences are found to be very close to each other, which are in good agreement with the similar water structure and dynamics (VAC and C2(t)) observed in the major grooves of both sequences (will be discussed later). However, entropy values for minor groove water for the two different sequences are found to be rather different from each other. The lower value (6.28 kcal/mol) of the AT sequence compared to the GC sequence (6.51 kcal/mol) suggests that the water molecules are significantly more constrained in the minor groove of the former. This result is also in agreement with the observed very different water structure and dynamics (VAC and C2(t)) observed in the two minor grooves (will be discussed later). We now examine the dependence of translational diffusivity of water molecules in the different regions of both sequences on the configurational entropy of the respective regions. Figure 3 displays the dependence of the logarithm of translational diffusivity versus 1/TSC, which shows a linear dependence in agreement with the Adam-Gibbs relation (eq 1), with a negative slope of 1.58 kcal/mol. Further, if one assumes Stokes-Einstein relation, where the translational diffusion constant and viscosity are inversely related, then we have an interesting correlation between microviscosity and configurational entropy. Note that such a correlation between the diffusion coefficient and configurational entropy is found to be present in a supercooled Lennard-Jones liquid with a negative slope of 0.47 kcal/mol.27 The larger value of the slope observed here originates from stronger intermolecular interactions in water and partly due to confinement. In the same plot, we have also shown the correlation between 〈th〉 and the configuational entropy of water in different regions. It is clear from the figure that 〈th〉 increases with decreasing configurational entropy. C. Velocity Time Correlation Function and Translational Dynamics. The translational mobility of the water molecules in the grooves of DNA can be understood by studying their respective velocity autocorrelation functions. The normalized

3636

J. Phys. Chem. B, Vol. 114, No. 10, 2010

Jana et al.

TABLE 1: Computed Entropy Values and Diffusion Coefficients of Water Molecules in the Major and Minor Grooves of Poly-GC and Poly-AT Sequences sequence

groove

poly-GC

major minor major minor

poly-AT

entropy (TS) (kcal/mol)

diffusion constant (10-5 cm2/s)

6.73 6.51 6.70 6.28 7.27

2.08 1.37 1.92 0.47 4.89

bulk water

Figure 3. Correlations between diffusion coefficient, configurational entropy, and tetrahedral order parameter (〈th〉). Note that left side of the Y-axis represents the logarithm of diffusivity and the right side of the Y-axis represents 〈th〉. The straight line fitting of the data validates the Adam-Gibbs relation. The dashed line shows the correlation between 〈th〉 and configurational entropy.

velocity autocorrelation function of a tagged water molecule is defined as

CV(t) )

b(t) · b 〈V V(0)〉 b 〈V(0) · b V(0)〉

(3)

where b V(t)and b V(0)are the velocity of the tagged water molecule at time t and zero, respectively. We have located the water molecules present at the major and the minor grooves of the DNA using the standard procedure where the distance (3.5 Å) between the oxygen atom of water molecules and the nearest DNA atoms of the respective groove was used as criteria. In Figure 4a, we show the calculated VAC of the water molecules present in the major groove of the poly-AT and polyGC sequences. Translational motion of the major groove water is found to be more constrained as compared to the bulk, showing more backscattering in the VAC than the bulk, as expected. However, we find no significant difference between the decay pattern of the VAC of the water molecules between the major grooves of both sequences and the two curves are almost overlapping. We have calculated the diffusion coefficient of water molecules in the major grooves of both sequences and for the bulk water molecules by integrating the VAC (D ) 1 /3∫∞0 CV(t) dt). We find that the diffusion coefficients in the major groove region of poly-GC and poly-AT sequences are ∼2.08 × 10-5 and ∼1.92 × 10-5 cm2/s, respectively. The diffusion coefficient for the bulk water is ∼4.89 × 10-5 cm2/s. The calculated similar values of the diffusion coefficient of water molecules in the two major grooves are in agreement with the observed similarity between the structures of water molecules in the two major grooves. However, a markedly different scenario emerges when the VAC of water molecules located in the minor groove of the two sequences, as shown in Figure 4b, is considered. We find

Figure 4. Velocity autocorrelation function of water molecules in the major groove (a) and the minor groove (b) region of both sequences. Note the highest backscattering in the case of minor groove water of poly-AT sequence. Note also the sequence insensitivity for the major groove water.

that the water molecules in the minor groove of poly-AT sequence are significantly more constrained translationally than those in the minor groove of the poly-GC sequence which are in turn much more constrained than the bulk. This feature is manifested in the large back scattering observed in the VAC in the minor groove of the poly-AT sequence as compared to the same in the poly-GC sequence. The computed diffusion coefficients of water molecules in the minor groove of poly-GC and poly-AT sequences are found to be ∼1.37 × 10-5 and ∼0.47 × 10-5 cm2/s. The AT minor groove is deeper and narrower than the GC minor groove, and the water molecules in the AT minor groove are more confined.28 This confinement plays a crucial role in several biological processes (drug-DNA intercalation, protein-DNA interaction, etc.). The difference in the diffusion coefficients of water molecules between the two minor grooves is also in agreement with the observed structural differences of water molecules between them. While the water molecules in the grooves can perform three-dimensional diffusion in the groove region at short times, at long times water molecules move in a channel-like path constituted by the connected adjacent grooves in the long chain. So, the confinement occurs in narrow channels of major/minor grooves. D. Orientational Relaxation. We probe the orientational dynamics of water molecules in the major and minor grooves of the two sequences by the rotational anisotropy which is directly proportional to the second rank orientational correlation function (C2(t)) of the O-H vector and is defined as

∑ P (e (t) · e (0))〉 〈 C (t) ) 〈 ∑ P (e (0) · e (0))〉 2

i

i

2

i

i

i

2

(4)

i

where ei(t) the O-H bond unit vector at time t. In Figure 5a, we show the decay of C2(t) of water molecules in the major grooves of the poly-AT and poly-GC sequences. We find a slower decay for the major grooves of both sequences as

Water Molecules in Minor Grooves of DNA

J. Phys. Chem. B, Vol. 114, No. 10, 2010 3637

Figure 5. Decay of the rotational anisotropy of the water molecules in the major groove (a) and the minor groove (b) region of both sequences. Note the slowest decay for the AT-minor groove water.

compared to the bulk, as expected. We have also observed that the decay of C2(t) in the major grooves of both sequences are quite similar. This implies that the orientational dynamics in the major groove region of the two sequences are very similar to each other, as in the case of translational dynamics. We have calculated the average orientational relaxation time by fitting the observed C2(t). The average orientational relaxation times for major grooves of the poly-GC and poly-AT sequences are found to be ∼4.2 and ∼5.2 ps. The relaxation time for bulk water is ∼0.7 ps. Similar to the translational dynamics, we observed a significant difference in the orientational dynamics between the minor grooves of the two sequences as shown in Figure 5b. We find the decay of C2(t) for the poly-AT sequences is slower than that of the poly-GC sequence, which is in turn slower than that of bulk water. This implies that the water molecules in the minor groove of poly-AT sequence are orientationally more constrained than poly-GC sequence, similar to what has been observed for the translational dynamics. The average orientational relaxation times in the minor grooves of the poly-GC and poly-AT sequences are found to be ∼8.1 and ∼46.7 ps. This again demonstrates that the orientational dynamics in the minor grooves of the AT sequence is quite slow. The slow component (102 ps) arises largely due to the confinement of the water molecules at the AT minor groove. E. Anisotropy in the Orientational Relaxation. Next, we discuss the anisotropy of the orientational relaxation of water molecules in different regions of poly-GC and poly-AT DNA. To this end, we have calculated the orientational correlation function of dipole, O-H and HH vectors, and the correlation function is defined as

∑ e (t) · e (0)〉 〈 C(t) ) 〈 ∑ e (0) · e (0)〉 i

i

i

i

i

i

(5)

Figure 6. Decay of CHH(t), CO-H(t), and Cµ(t) of the water molecules in the major groove (a), in the GC minor groove (b), and in the AT minor groove (c). Note the faster decay of CHH(t) and the slower decay of Cµ(t). Note also the highest separation in the time scale of the decay for the AT minor groove water showing the largest anisotropy in the relaxation behavior.

Here ei(t) indicates the unit vectors along dipole or O-H or HH vectors and the corresponding correlation functions are denoted as Cµ(t) or CO-H(t) or CHH(t). For bulk water, the decay of these correlations is found to be very close to one another, indicating no significant anisotropy in the orientational relaxation. However, the relaxation becomes anisotropic in the major grooves of both the DNA. The HH vector rotation is found to be fastest among the three and dipole vector rotation is the slowest (as shown in Figure 6a). Relaxation of the O-H vectors shows intermediate behavior. We have calculated the correlation functions for the minor groove water molecules of both the DNA and found that the anisotropy becomes more prominent, especially in the AT minor groove (Figure 6b,c). The relative rate of the rotation is found to be similar to what was observed in the major grooves, HH > O-H > dipole. Thus, confinement induces anisotropy in the orientational relaxation of the water molecules severely and the rotation of the HH vectors becomes faster and the dipole vector rotation becomes slower. The highest anisotropy observed in the AT minor groove water can be attributed to the highest confinement. However, the qualitative behavior of these three correlation functions in the major and minor grooves of both the DNA (as discussed in section D) remains similar. F. Tetrahedral Ordering, Surface Specific Interaction and Nanoconfinement. We know that the total energy of the WatsonCrick hydrogen bonds of the GC pair (three hydrogen bonds) is 16.8

3638

J. Phys. Chem. B, Vol. 114, No. 10, 2010

kcal/mol and that of the AT pair (two hydrogen bonds) is 7.0 kcal/ mol. Thus, the average single hydrogen bonding energy of the GC pair is 5.6 kcal/mol and the same for the AT pair is 3.5 kcal/mol.29 This indicates that the DNA atoms in the AT grooves are more amendable to move (bend or distort) to ensure the interaction with water molecules, and thus the interaction energy between groove water molecules and groove atoms may become more for AT grooves than that for GC grooves. Similar dynamical behavior of water molecules in the major grooves of both poly-AT and poly-GC sequences demonstrates that such interaction effects are not important for the major grooves. The major grooves of both sequences have similar widths of ∼10 -12 Å, and we also find similar values of tetrahedral order parameters in the major grooves. This result clearly implies the larger effect of confinement, which is directly correlated with the tetrahedral ordering the grooves, of water in the grooves rather than surface specific interactions on the water dynamics. We find slower dynamics of water molecules in the AT minor grooves compared to the dynamics in the GC minor groove. This result is in accordance with the facts that (1) water molecules have larger interaction energies in the AT minor groove rather than in GC and (2) water molecules in the AT minor groove (width ∼3.5 Å, higher value of tetrahedral parameter) experience larger confinement than in the GC minor grooves (width ∼6 Å, lower values of tetrahedral order parameter). However, comparison of the dynamics in the major grooves reveals that confinement has the larger effect and surface specific interactions have very little effect on the dynamics. Hence, the slowest dynamics of water molecules in the AT minor groove is primarily due to the larger confinement, which enhances the tetrahedral ordering in the AT minor groove rather than the stronger surface specific interactions. IV. Conclusions We are not aware of any previous study that explored the combined effects of nanoconfinement, surface specific interactions, and DNA rigidity on dynamics of water molecules in grooves of DNA. DNA offers a unique scope for such a study because of the existence of the two grooves of different volumes that can be formed with two different sequences. We find that the confinement increases the order (tetrahedral) and hence enhances the correlation between the water molecules. The slowest dynamics observed in the AT-minor groove region can be attributed to the greater order observed due to larger confinement. The results of the present study are in agreement with the observed lowering of static dielectric constant of water when confined to the nanocavity.30 In the above study there was no electrostatic interaction between water and the cavity surface. So, the reduction of the dielectric constant observed is probably due to the increase in ordering among the water molecules inside the nanocavity as a consequence of confinement.30 Reduction of the dielectric constant has also been observed for Stockmayer fluids confined to a spherical cavity.31 The correlations among configuration entropy, diffusion coefficient, 〈th〉, and orientational relaxation time are important results of this study. The agreement with the Adam-Gibbs relation is indeed impressive. This further suggests the role of confinement, which effectively decreases the entropy of water molecules. That is, in the entropy-enthalpy balance present in a dynamic equilibrium, nanoconfinement favors enthalpy over entropy. It is clear that confinement removes important volume and number fluctuations, allowing only entropy fluctuations. According to Einstein’s theory of fluctuation, the probability of a thermodynamic fluctuation is given by32

Jana et al.

P(∆S) ∝ exp((-∆T∆S)/kB) ∝ exp((-2(∆S)2 /kBCP)

(6)

Because CP of ordered water is less than that for the bulk, the above equation shows a lower probability of entropy fluctuation in the minor grooves. We propose that the flexibility of the groove structure is partially responsible for the observed tetrahedrality of the water molecules inside the groove. GC groove atoms are more rigidly bound to the DNA frame and not easily deformable. The larger deformability of the AT groove atoms helps in the formation tetrahedral order in AT minor groove water. Acknowledgment. This work was supported in parts by grants from DST and CSIR, India. B.J. thanks CSIR for providing SRF, and B.B. thanks DST for a J.C. Bose Fellowship. References and Notes (1) Bagchi, B. Chem. ReV. 2005, 105, 3197. (2) Park, S.; Fayer, M. D. Proc. Natl. Acad. Sci., U.S.A. 2007, 104, 16731. (3) Moilanen, D. E.; Levinger, N.; Spry, D. B.; Fayer, M. D. J. Am. Chem. Soc. 2007, 129, 14311. (4) Tan, H.-S.; Piletic, I. R.; Riter, R. E.; Levinger, N. E.; Fayer, M. D. Phys. ReV. Lett. 2005, 94, 057405. (5) Faeder, J.; Ladanyi, B. M. J. Phys. Chem. B 2005, 109, 6732. (6) Senapati, S.; Berkowitz, M. L. J. Phys. Chem. A 2004, 108, 9768; J. Phys. Chem. B 2003, 107, 12906. (7) Bhattacharyya, K. Acc. Chem. Res. 2003, 36, 95; Chem. Commun., 2008, 2848. (8) Pant, D.; Levinger, N. E. Langmuir 2000, 16, 10123. (9) Kurita, R.; Tanaka, H. Phys. ReV. Lett. 2007, 98, 235701. (10) Royall, C. P.; Williams, S. R.; Ohtsuka, T.; Tanaka, H. Nat. Mater. 2008, 7, 556. (11) Shintani, H.; Tanaka, H. Nat. Phys. 2006, 2, 200. (12) Ohtsuka, T.; Royall, C. P.; Tanaka, H. Eur. Phys. Lett. 2008, 84, 46002. (13) Mukherjee, A.; Lavery, R.; Bagchi, B.; Hynes, J. T. J. Am. Chem. Soc. 2008, 130, 9747. (14) Nguyen, B.; Neidle, S.; Wilson, W. D. Acc. Chem. Res. 2008, 42, 11. Flatters, D.; Lavery, R. Biophys. J. 1998, 75, 372. (15) Andreatta, D.; Pe´rez Lustres, J. L.; Kovalenko, S. A.; Ernsting, N. P.; Murphy, C. J.; Coleman, R. S.; Berg, M. A. J. Am. Chem. Soc. 2005, 127, 7270. (16) Berg, M. A.; Coleman, R. S.; Murphy, C. J. Phys. Chem. Chem. Phys. 2008, 10, 1229. (17) Pal, S.; Maiti, P. K.; Bagchi, B.; Hynes, J. T. J. Phys. Chem. B 2006, 110, 26396. (18) Furse, K. E.; Corcelli, S. A. J. Am. Chem. Soc. 2008, 130, 13103. (19) Lin, S. T.; Blanco, M.; Goddard, W. A. J. Chem. Phys. 2003, 119, 11792. (20) Jana, B.; Pal, S.; Maiti, P. K.; Lin, S. T.; Hynes, J. T.; Bagchi, B. J. Phys. Chem. B 2006, 110, 19611. (21) Case, D. A.; Pearlman, D. A.; Caldwell, J. W.; Cheatham, T. E.; Wang, J.; Ross, W. S.; Simmerling, C.; Darden, T.; Merz, K. M.; Stanton, R. V. AMBER, 7th ed.; University of California: San Francisco, 1999. (22) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman, P. A. J. Am. Chem. Soc. 1995, 117, 5179. (23) Feig, M.; Pettitt, B. M. Biophys. J. 1998, 75, 134; J. Mol. Biol. 1999, 286, 1075. (24) Essmann, U.; Perera, L.; Berkowitz, M. L.; Darden, T.; Lee, H.; Pedersen, L. G. J. Chem. Phys. 1995, 103, 8577. (25) Giguere, P. A. J. Raman Spectrosc. 1984, 15, 354. (26) Chau, P.-L.; Hardwick, A. J. Mol. Phys. 1998, 93, 511. (27) Nave, E. L.; Sastry, S.; Sciortino, F. Phys. ReV. E. 2006, 74, 050501. (28) Alexeev, D. G.; Lipanov, A. A.; Skuratovskii, I. Y. Nature 1987, 325, 821. (29) Yakishevich, L. V. Nonlinear Physics of DNA; Wiley-VCH: Weinheim, 2004; p 6. (30) Senapati, S.; Chandra, A. J. Phys. Chem. B 2001, 105, 5106. (31) Senapati, S.; Chandra, A. J. Chem. Phys. 1999, 111, 1223. (32) Hansen, J.-P.; McDonald, I. R. Theory of Simple Liquids, 2nd ed.; Academic Press: New York, 1968.

JP907513W