Electric field of DNA in solution: who is in charge?

In solution, DNA is a highly charged macromolecule which bears a unit of negative charge on each phosphate of its sugar-phosphate backbone. Although partially compensated by counterions adsorbed at or condensed near it, DNA still produces a substantial electric field in its vicinity, which is screened by buffer electrolyte at longer distances from the DNA. Such field has been explored so far predominantly within the scope of a primitive model of the electrolytic solution, not considering more complicated structural effects of the water solvent. In this paper we investigate the distribution of electric field around DNA using linear response nonlocal electrostatic theory, applied here for helix-specific charge distributions, and compare the predictions of such theory with specially performed fully atomistic large scale molecular dynamics simulations. The main finding of this study is that oscillations in the electrostatic potential distribution are present around DNA, caused by the overscreening effect of structured water. Surprisingly, electrolyte ions at physiological concentrations do not strongly disrupt these oscillations, and rather distribute according to these oscillating patterns, indicating that water structural effects dominate the short-range electrostatics. We also show that (i) structured water adsorbed in the grooves of DNA lead to a positive electrostatic potential core, (ii) the Debye length some 10 {\AA} away from the DNA is reduced, effectively renormalised by the helical pitch of the DNA, and (iii) Lorentzian contributions to the nonlocal dielectric function of water, effectively reducing the dielectric constant close to the DNA, enhances the overall electric field. The impressive agreement between the atomistic simulations and the developed theory substantiates the use of nonlocal electrostatics when considering solvent effects in molecular processes in biology.

Nucleosome-induced homology recognition in chromatin.

One of the least understood properties of chromatin is the ability of its similar regions to recognize each other through weak interactions. Theories based on electrostatic interactions between helical macromolecules suggest that the ability to recognize sequence homology is an innate property of the non-ideal helical structure of DNA. However, this theory does not account for the nucleosomal packing of DNA. Can homologous DNA sequences recognize each other while wrapped up in the nucleosomes? Can structural homology arise at the level of nucleosome arrays? Here, we present a theoretical model for the recognition potential well between chromatin fibres sliding against each other. This well is different from the one predicted for bare DNA; the minima in energy do not correspond to literal juxtaposition, but are shifted by approximately half the nucleosome repeat length. The presence of this potential well suggests that nucleosome positioning may induce mutual sequence recognition between chromatin fibres and facilitate the formation of chromatin nanodomains. This has implications for nucleosome arrays enclosed between CTCF-cohesin boundaries, which may form stiffer stem-like structures instead of flexible entropically favourable loops. We also consider switches between chromatin states, e.g. through acetylation/deacetylation of histones, and discuss nucleosome-induced recognition as a precursory stage of genetic recombination.