19th-banner-rev.gif

Book of Abstracts: Albany 2009

category image Albany 2009
Conversation 16
June 16-20 2009
© Adenine Press (2008)

Molecular Dynamics Studies of Nucleosome Positioning

There are approximately 30 high resolution crystallographic structures of the nucleosome available in the protein data bank. All have essentially the same sequence of DNA, nuc147. Some have specific structural modifications. We previously compared 24 of these structures to determine the necessary and sufficient distribution of DNA helical parameters (tilt, roll, twist, shift, slide, rise) required to recreate all atom models that are within 3Å RMSD of the initial x-ray structure. We found that the distribution of roll, slide and twist is highly conserved in all structures but that rise, tilt, and shift vary. Here we use a combination of all atom molecular mechanics and elastic rod modeling techniques to investigate sequence dependencies in the structure, dynamics and energetics of the nucleosome. For this purpose we have constructed several thousand all atom models of the nucleosome with different sequences of DNA and systematically varied the folding of free DNA into a nucleosome for the sequence nuc147. The sequences studied include the 1489nt sequence of the mouse mammary tumor virus promoter (MMTV, genbank id V01175) which positions six nucleosomes and nearly 100 sequences of DNA that span the range of binding free energies. We find that threading different sequences onto the histone core is a suitable starting point for all-atom simulations for both solvent free and fully solvated systems and that molecular mechanics energies are well correlated with an elastic rod model that includes a screened electrostatics term for long range interactions (Debye-Huckel approximation). Without a long range term the elastic rod model fundamentally differs from molecular mechanics models. Yet, assuming the conformation of nucleosomal DNA is sequence invariant, this electrostatic term is constant. Molecular dynamics simulations are being employed to assess this assumption.

This work supported by a grant from the NIH (R01GM76356).

Thomas C. Bishop

Center for Computational Science
Tulane University
New Orleans, LA 70118

Tel: 504-862-3370
Fax: 504-862-8392
email Tom Bishop