Sequence-dependent Elasticity Research Articles

Translational nucleosome positioning: A computational study.

About three-quarters of eukaryotic DNA is wrapped into nucleosomes; DNA spools with a protein core. The affinity of a given DNA stretch to be incorporated into a nucleosome is known to depend on the base-pair sequence-dependent geometry and elasticity of the DNA double helix. This causes the rotational and translational positioning of nucleosomes. In this study we ask the question whether the latter can be predicted by a simple coarse-grained DNA model with sequence-dependent elasticity, the rigid base-pair model. Whereas this model is known to be rather robust in predicting rotational nucleosome positioning, we show that the translational positioning is a rather subtle effect that is dominated by the guanine-cytosine content dependence of entropy rather than energy. A correct qualitative prediction within the rigid base-pair framework can only be achieved by assuming that DNA elasticity effectively changes on complexation into the nucleosome complex. With that extra assumption we arrive at a model which gives an excellent quantitative agreement to experimental in vitro nucleosome maps, under the additional assumption that nucleosomes equilibrate their positions only locally.

Effects of Protein-Induced Local Bending and Sequence Dependence on the Configurations of Supercoiled DNA Minicircles.

The statistical mechanics of a short circularized DNA molecule, a DNA minicircle, with a prescribed linking number depends heavily on the mechanical and geometrical properties of the DNA, which are known to be functions of the sequence. A description of a general numerical scheme used for the performed advanced simulation is presented with examples that reveal the effects of sequence dependence and local deformations, caused by protein binding, on the configurations in a canonical ensemble of a supercoiled minicircle. Using a realistic course grain model in which the sequence-dependent elasticity, the intramolecular electrostatic interactions, and the impenetrability of the DNA molecule are taken into account, the bifurcation of equilibria of supercoiled minicircle DNA with the induced bending as a parameter is shown. The unique Monte Carlo scheme for constrained DNA molecules was utilized in order to calculate site-to-site distance probabilities for each pair of sites in the molecule. The simulated examples show not only that the sequence alone can play a significant role in bringing two remote sites to a contact but also the possible existence of several competing regions of contact. The results suggest that the local deformation caused by protein binding can yield a global configurational change, dominated by slithering motion, which brings two (originally) remote sites to close proximity, and that the nature of such effect is related to the sequence architecture.

Understanding How Proteins Shape DNA Using Energy Minimization

Several proteins are known to produce significant changes in DNA shape by locally deforming the double helix or by mediating the formation a loop. For example, the binding of the bacterial proteins HU or Hbb produces large bends in DNA and the tetrameric Lac repressor protein induces the formation of a loop between two distant DNA operators. In order to understand how such proteins sculpt DNA, we have developed a novel optimization method for DNA at the base-pair level. Our method accounts for the sequence-dependent elasticity of DNA and can be applied to a DNA fragment in which the first and last base pairs are spatially constrained. Moreover, our approach makes it possible to constrain intervening parts of the DNA in order to model the presence of bound proteins. We can therefore compute the energy landscape for a wide variety of protein-DNA systems. For example, we studied the effect of the presence of multiple HU or Hbb proteins on DNA minicircles and identified the most likely configurations for given DNA chain lengths and numbers of bound proteins. We also investigated how the presence of HU can enhance or diminish the apparent flexibility of DNA. In addition, we performed a detailed analysis of the energy landscape for loops induced by the Lac repressor. We focused on the binding of the Lac repressor protein on DNA minicircles and studied the competition between large and small loops on supercoiled molecules. We also found that changes in the repressor structure can lead to different loop topologies. Our method makes it possible to conduct thorough analyses of the geometry and mechanics of protein-DNA systems by computing the associated energy landscape and paves the way for the study of other biologically relevant system such as the SV40 minichromosome.

Rigid-body molecular dynamics of DNA inside a nucleosome

The majority of eukaryotic DNA, about three quarter, is wrapped around histone proteins forming so-called nucleosomes. To study nucleosomal DNA we introduce a coarse-grained molecular dynamics model based on sequence-dependent harmonic rigid base pair step parameters of DNA and nucleosomal binding sites. Mixed parametrization based on all-atom molecular dynamics and crystallographic data of protein-DNA structures is used for the base pair step parameters. The binding site parameters are adjusted by experimental B-factor values of the nucleosome crystal structure. The model is then used to determine the energy cost for placing a twist defect into the nucleosomal DNA which allows us to use Kramers theory to calculate nucleosome sliding caused by such defects. It is shown that the twist defect scenario together with the sequence-dependent elasticity of DNA can explain the slow time scales observed for nucleosome mobility along DNA. With this method we also show how the twist defect mechanism leads to a higher mobility of DNA in the presence of sin mutations near the dyad axis. Finally, by performing simulations on 5s rDNA, 601, and telomeric base pair sequences, it is demonstrated that the current model is a powerful tool to predict nucleosome positioning.

Sequence-Dependent Mechanics of DNA

Double-stranded DNA is a semiflexible polymer that can naturally bend on length scales comparable to the size of large DNA-protein complexes like nucleosomes or protein-mediated DNA loops. The sequence of the substrate DNA does not only provide biochemical binding sites for the proteins, but also affects the local mechanical properties of the DNA. Notably, sequence can affect the intrinsic curvature of the DNA, as well as its bendability, or elasticity. While intrinsic bends in DNA and their role in protein-DNA complex formation are well studied, sequence-dependent elasticity still remains only vaguely explored. In order to separate sequence effects on elasticity from those on intrinsic curvature, we have designed sequences of DNA which have nearly identical curvatures but varied AT content and directly measured their mechanical elasticity using constant force axial optical tweezers. We found the persistence length to be highly dependent on the AT content of the DNA, differing almost thirty percent between sequences with nearly identical curvature but different sequence composition. This is a departure from conventional dinucleotide and trinucleotide models, which predict a much smaller difference between the two sequences, but consistent with estimates obtained from the crystallographic structures of protein-DNA complexes.