The B- to A-DNA transition and the reorganization of solvent at the DNA surface

Abstract

DNA geometry depends on relative humidity. Using the CHARMM22 force field to push B-DNA to A-DNA, a molecular dynamics simulation of a mixed-sequence 24-basepair DNA double-stranded oligomer, starting from B-DNA, was carried out to explore both the mechanism of the transition and the evolution of hydration patterns on the surface of DNA. Over the 11-ns trajectory, the transition recapitulates the slide-first, roll-later mechanism, is opposed by DNA electrostatics, and is favored by an increasing amount of condensed sodium ions. Hydration was characterized by counting the hydrogen bonds between water and DNA, and by the number of water bridges linking two DNA atoms. The number of hydrogen bonds between water and DNA remains constant during the transition, but there is a 40% increase in the number of water bridges, in agreement with the principle of economy of hydration. Water bridges emerge as delicate sensors of both structure and dynamics of DNA. Both local flexibility and the frustration of the water network on the surface of DNA probably account for the low populations and short residence times of the bridges, and for the lubricant role of water in ligand-DNA interactions.

FIGURE 1

Root mean-square deviation of the oligomer structure to canonical A- (open symbols) and B-DNA (solid symbols), as a function of time. One-nanosecond averages, mean ± SD are shown as a function of time for the three regions defined in the text. (A) Region 1; (B) Region 2; and (C) Region 3. (Circles, native simulation; squares, control simulation with uncharged sodium ions.)

FIGURE 2

Internal energy decomposition for the B- to A-DNA transition. One-nanosecond averages, mean ± SD are shown as a function of time. (A) Total internal energy (kcal/mol); (B) electrostatic energy (kcal/mol); and (C) sodium-DNA energy (kcal/mol). (Solid symbols, native simulation; open symbols, control simulation with uncharged sodium ions.) (D) Number of sodium ions within a 6 Å radius of the surface of DNA ± twice the mean SD over 2-ns intervals. (E) Number of sodium ions within a 6 Å radius of the surface of DNA over 2-ns intervals, for the minor groove (solid circles), major groove (open circles), and phosphate functional groups (solid diamonds).

FIGURE 3

Geometrical characterization of the B- to A-DNA transition. One-nanosecond averages, mean ± SD are shown as a function of time. (A) Roll (continuous line) and slide (dashed line) versus time. (Solid symbols, native simulation; open symbols, control simulation.) (B) Correlation between δ and χ; the corresponding time interval is labeled for 1, 2, 3, 4, and 11 ns (solid symbols and continuous line, native simulation; open symbols and dashed line, control simulation). (C) Population percentage for N (solid symbols) and E + S (open symbols) sugar puckers, as a function of time, for 1-ns intervals (circles, native simulation; squares, control simulation). (D) Consecutive P–P distance histograms for the native simulation, as a function of time, for 1-ns intervals (the first and last nanoseconds are shown in thicker lines). (E) Population percentage for consecutive P–P distances within 6.4 Å (solid symbols, native simulation; open symbols, control simulation). (F) The values ɛ- and ζ-conformations as a function of time, for 1-ns intervals (BI in circles, BII in squares, and A in triangles; solid symbols, native simulation; open symbols, control simulation).

FIGURE 4

Solvent-accessible surface area along the B- to A-DNA transition. One-nanosecond averages, mean ± SD are shown as a function of time. (A) Surface area for the complete oligomer; (B) minor groove surface area; (C) major groove surface area; and (D) sugar surface area (purines in circles and pyrimidines in squares). Solid symbols, native simulation; open symbols, control simulation.

FIGURE 5

H-bonds and water bridges during the B- to A-DNA transition. (A) Corresponds to the fixed A-DNA simulation. Normalized with respect to the fixed B-DNA simulation (B). Average H-bonds and water bridges found in 1-ns intervals. (H-bonds in open circles and water bridges in solid circles.)

FIGURE 6

Water-bridge populations that change during the B- to A-DNA transition. Normalized to the total number of possible sites forming each particular type of water bridge. (A) Backbone bridges (O1P–O1P in solid circles, O1P–O2P in open circles, and O1P–O5′ in solid squares). (B) Minor groove bridges (N3–O4′ in solid circles, O2–O4′ in open circles, and N3–O2 in solid squares).

FIGURE 7

Water bridges formed by O1P. Snapshot containing the three bridges in Fig. 6 A; all atoms in standard CPK coloring except for O5′ in pink and O2P in purple. Water molecules involved in the bridges colored green. One water molecule is making a three-point contact involving O1P from nucleotide i, O1P from nucleotide i–1, and O5′ from nucleotide i–1. The other water-molecule bridges O1P from nucleotide i to O2P from nucleotide i–1.

FIGURE 8

Atomic fluctuations as a function of time, and their relation to water-bridge lifetime. (A) Average atomic fluctuations of all polar atoms, in 1-ns intervals mean ± SE, as a function of time. (B) Correlation between O1P fluctuations and the average lifetime of the three water bridges shown in Fig. 6 A. (O1P–O1P in solid circles and solid line; O1P–O2P in open circles and dashed line; and O1P–O5′ in solid squares and dashed line.)

FIGURE 9

Competition between equivalent sites at the minor groove. All atoms in standard CPK colors, except water, which is shown in green. N3, O2, and O4′ atoms involved in the bridges are shown in space-filling mode. (A) N3–O4′ bridge. (B) N3–O2 bridge, and water-water bridge between a water molecule H-bonded to O4′ and the one making the N3–O2 bridge.