Date of Award

Spring 2007

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Chemistry - (Ph.D.)

Department

Chemistry and Environmental Science

First Advisor

Carol A. Venanzi

Second Advisor

Joseph W. Bozzelli

Third Advisor

Tamara M. Gund

Fourth Advisor

Edgardo Tabion Farinas

Fifth Advisor

Christopher C. Van Dyke

Abstract

Ligand-based drug design (LBDD) techniques are applied when the structure of the receptor is unknown but when a series of compounds or ligands have been identified that show the biological activity of the interest. Generally, availability of a series of compounds with high activity, with no activity, and also with a range of intermediate activities for the desired biological target is required. It is common that structures of membrane-bound proteins (for example, monoamine transporter proteins and opioid receptor proteins) are unknown as these proteins are notoriously difficult to crystallize.

In Part I of this study, analogs of the flexible dopamine reuptake inhibitor, GBR 12909, may have potential usefulness in the treatment of cocaine abuse. As a first step in the 3D-QSAR modeling of the dopamine transporter (DAT)/serotonin transporter (SERT) selectivity of these compounds, conformational analysis of a piperazine and related piperidine analog of GBR12909 is performed. These analogs have eight rotatable bonds and are somewhat easier to deal with computationally than the parent compound. Ensembles of conformers consisting of local minima on the potential energy surface of the molecule were generated in the vacuum phase and implicit solvent (also known as continuum solvent) by random search conformational analysis using the molecular mechanics methods and the Tripos and MMFF94 force fields. These conformer populations were classified by relative energy, molecular shape, and their behavior in 2D torsional angle space in order to evaluate their sensitivity to the choice of charges and force field. Some differences were noted in the conformer populations due to differences in the treatment of the tertiary amine nitrogen and ether oxygen atom types by the force fields.

In Part II of this study, 3D-QSAR studies of salvinorin A analogs as kappa opioid (K) receptor agonists were performed. Salvinorin A is a naturally-occurring diterpene from the plant Salvia divinorum which activates the kappa opioid receptor (KOR) selectively and potently. It is the only known natural non-nitrogenous agent active at the human KOR. Salvinorin A may represent a novel lead compound with possible potential in the treatment of addiction and pain. The primary aim of the current study was to develop Comparative Molecular Field Analysis (CoMFA) models to clarify the correlation between the molecular features of the 2-position analogs of salvinorin A and their KOR binding affinity. The final, stable CoMFA model has predictivity given by q2 of 0.62 and fit given by r2 of 0.86. The steric and electrostatic contributions were 47% and 53%, respectively. The CoMFA contour map indicated that the presence of a negative environment and steric region near the 2-position would lead to improved binding affinity at the KOR. Novel salvinorin A analogs with improved binding affinity were predicted based on the stable and predictive CoMFA model. Novel analogs were synthesized by Dr. Thomas Prisinzano of the University of Iowa and preliminary biological results are available from the Rothman laboratory at the National Institute on Drug Abuse. These novel analogs appear to be KOR selective.

Included in

Chemistry Commons

Share

COinS