Masters Theses

Date of Award


Degree Type


Degree Name

Master of Science



Major Professor

Solon Georghiou

Committee Members

Henry Simpson, Chia C. Shih, Engin Serpersu


The photophysical properties of deoxyguanosine, dG, have been studied in aqueous solution as well as in organic solvents of varying polarity. The fluorescence quantum yield, q, is found to increase as the solvent polarity is decreased: in diethyl ether, the solvent of lowest polarity used in the present study, q has been found to increase by a factor of 35 relative to the value of q in aqueous buffer. This suggests that hydrophobic interactions reduce considerably the rate constant of the radiationless process of internal conversion. On the other hand, upon increasing the viscosity, η (through addition of sucrose), q has been found to increase by a factor of 1.7 for η = 14 cP and by a factor of 7.4 for η = 149 cP. These results suggest that the increase in the rigidity of dG and of the hydrophobic nature of its environment, which occur when dG becomes part of DNA, are not responsible for the very small value of q in DNA * 0.8 x 10-5 as compared to that of 0.8 x 10-4 for free dG m water. An attractive alternative is provided by differences in the hydration network between free dG and dG in DNA. Support for this proposal was obtained from an experiment with an 8% water/butanol mixture which suggested that the hydration network affects profoundly the photophysical properties of free dG. Measurements in this water/butanol mixture were pursued further Surprisingly, it was found that the results depended on the sequence of the steps used in preparing the solution. This is consistent with the reported very large, about - 20 kcal/mol, solvation free energy of guanine We measured the fluorescence of dG at room temperature after the solution was incubated for 10 minutes at a number of temperatures ranging from 23 to 40 °C. An Arrhenius plot yielded ΔH ≈ - 6 kcal/mol and ΔS ≈ 48 cal/mol K. A potential implication of these findings is that intermolecular hydrogen bonds in DNA, once broken by thermal fluctuations, would tend to remain broken; consequently, the helix would appear to contain a number of GC open or "unzipped" base pairs. Because of the known coupling between base pair opening and bending, it would take relatively little energy for the DNA to bend at GC open sites. We have also employed the theoretical analysis of Jarque and Buckingham, for two ions of the same charge embedded in a polarizable medium, to calculate the contribution which is made by the polarization effects to the electrostatic interaction between the negatively-charged phosphate groups in DNA. It is found that many-body, non-additive polarization interactions greatly diminish the effect of the repulsive Coulomb interactions between phosphates both across the DNA minor groove (but not across the major groove) and along the same DNA strand. This reduction is found to be dramatic for groove widths in the range of about 3.5-5 Å for a nearest neighbor distance β of about 3.5 Å. This value of β appears to be realistic for DNA. Thus, variations in the minor groove width caused by sequence effects or by thermal fluctuations induce local anisotropic interactions with concomitant local DNA bending toward the groove with the smaller width. Other values of &beta ranging from 3.2 to 5 Å have also been used. From the derivative of the effective potential with respect to the groove width, the overall force F was calculated and its magnitude and direction were found to be strongly dependent on the value of β for small groove widths F is negative, i e attractive for β ≤ 4 Å, and positive for larger β values. Thus, the conformation of DNA exerts a profound effect on the ability of DNA to control its dynamics. Based on the fact that the A-tracts have a narrow groove width of about 3.5 Å, the present findings suggest that they would tend to bend toward that groove. This prediction is in agreement with the results of electrophoresis studies. Because of the known coupling between DNA bending and opening of its base pairs, the present findings suggest that base pair opening is a process that does not appear to be very infrequent and is driven by both sequence effects (of static origin) and by thermal fluctuations (of dynamic origin). Such a process may contribute to the creation of the intercalation cavity during drug-DNA interaction. This analysis also offers an explanation for the wrapping of DNA around the histone octamer in nucleosomes: because the A-tracts in DNA are known to bind with their minor grooves facing in toward the histone octamer, the DNA in those regions would bend toward those tracts. Because of the known gradual decrease of the width of the grooves of these tracts from 5' to 3' and the large increase of the strength of the polarization interaction that occurs with this groove width decrease, the DNA wrapping would be done with considerable force. The present analysis brings into question the validity of the continuum dielectric screening model currently being used in the literature to calculate electrostatic interactions.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."