Title

Application of multi-dimensional NMR in the study of DNA structure and dynamics

Date of Award

1997

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Physics

Advisor(s)

Philip Borer

Keywords

hairpins, loop structure, base mismatch

Subject Categories

Biological and Chemical Physics

Abstract

This dissertation presents research in structure determination, analysis of NMR relaxation data, and data processing which are three important areas for the applications of multi-dimensional NMR techniques to DNA structure and dynamics.

The solution structure of an 18 residue DNA tetra-loop hairpin with a T-T mismatch in the middle of the stem was determined by 35 ps restrained molecular dynamics simulations followed by energy minimization calculations. 239 accurate distance constraints were derived by the MARDIGRAS program from 2D NOE experimental data. A total of 483 constraints including loose distances, H-bonds and chirality were obtained. The B-form stem structure is well determined. Two mismatched thymines are well stacked in the stem and protected from the solvent. The loop structure is less well-defined due to either an incomplete constraint set or the flexible nature of the loop. There is no evidence for hydrogen-bonding in the loop or mismatched bases.

Repeated measurements of the longitudinal relaxation time, T$\sb1,$ and the heteronuclear NOE were made prior to the work discussed here; the measurement were on a DNA duplex. The measurements were used to generate values of T$\sb1$ and NOE with normal distributions for each carbon corresponding to the measured errors. The order parameter, S$\sp2,$ and the effective internal correlation time, $\rm\tau\sb{e},$ in the Model-Free Approach, have been optimized from the distributions simulated by McMOLDYN. Between the two parameters, S$\sp2$ has a smaller relative error, estimated at 15% on average, which means that S$\sp2$ is a well-defined parameter. However, $\rm\tau\sb{e}$ is very poorly defined with the average relative error estimated at 85%; it is typically found in the range of 30-300 ps.

A computer program useful for automated two-dimensional baseplane NMR correction and subsequent $t\sb1$ and $t\sb2$ ridge suppression is described. The algorithm performs combined correction of smooth baseplane distortions and sharp ridges in 2D NMR spectra through five steps: (1) identification of resonance peaks and ridges, (2) extraction of an initial, putative global baseplane, (3) window filtering of the corresponding time domain, (4) construction of a 2D spectrum free of baseplane distortion, and (5) suppression of ridges. The optimal parameters for baseplane and ridge correction are automatically decided by the program, yielding a greatly improved spectrum, together with more accurate spectral information.

Access

Surface provides description only. Full text is available to ProQuest subscribers. Ask your Librarian for assistance.

http://libezproxy.syr.edu/login?url=http://proquest.umi.com/pqdweb?did=736867411&sid=2&Fmt=2&clientId=3739&RQT=309&VName=PQD