Author ORCID Identifier

Defense Date


Document Type


Degree Name

Doctor of Philosophy


Pharmaceutical Sciences

First Advisor

Martin K. Safo

Second Advisor

Glen E. Kellogg


Naturally developed proteins are capable of carrying out a wide variety of molecular functions due to their highly precise three-dimensional structures, which are determined by their genetically encoded sequences of amino acids. A thorough knowledge of protein structures and interactions at the atomic level will enable researchers to get a deep foundational understanding of the molecular interactions and enzymatic processes required for cells, resulting in more effective therapeutic interventions. This dissertation intends to use structural knowledge from solved protein structures for two distinct objectives.

In the first project, we conducted a bioinformatics structural analysis of experimental protein structures using our novel paradigm "3D Interaction Homology". The three-dimensional structure of biological macromolecules, particularly proteins, provides us with a better understanding of protein interactions and functions, enabling us to establish hypotheses about how to modulate, regulate, or modify their functions. Therefore, the Kellogg lab proposed a new paradigm for the building and refinement of protein structure models using 3D hydropathic maps. To accomplish this goal, we have been characterizing the hydropathic interaction residue environments by compiling a database of residue type- and backbone angle-dependent 3D maps. In this work, 3D hydropathic interaction maps feature of the HINT (Hydropathic INTeraction) program enabled calculation and visualization of the 3D hydropathic environments of the three aromatic amino acid residues. We have shown that these 3D maps are information rich descriptors of preferred conformations, interaction types and energetics, and solvent accessibility. We calculated and analyzed sidechain-to-environment 3D maps for over 70,000 phenylalanine, tyrosine, and tryptophan residues. Moreover, significant and occurrence of some special non-covalent interactions (π-π and π-cation) were calculated and analyzed. This recognition of even these subtle interactions in the 3D hydropathic environment maps is key support for our interaction homology paradigm of protein structure elucidation and possibly prediction.

In the second project, we aimed to investigate the physical interaction between a vitamin B6-salvage enzyme, pyridoxine-5' phosphate oxidase (PNPO), and a vitamin B6-dependent enzyme, dopa decarboxylase (DDC), employing different approaches, including molecular modeling, biophysical, enzyme kinetics, and site-directed mutagenesis studies. PLP, the active vitamer of B6, serves as a cofactor for approximately 180 B6-dependent (PLP-dependent) enzymes and play crucial roles on many of cellular functions, e.g., heme, amino acid, neurotransmitter, DNA/RNA biosynthesis. Vitamin B6 deficiency is suspected to contribute to several pathologies, e.g., seizures, autism, schizophrenia, epilepsy, and Alzheimer’s disease. High levels of vitamin B6 are also linked to neurotoxic effects due in part to potential toxicity by free PLP in the cell. Therefore, the cellular content of free PLP is kept very low. Understanding the role of this vitamin in these pathologies requires knowledge on its metabolism and regulation, and subsequent transfer to dozens of apo-B6 enzymes. We hypothesize that the transfer of PLP from the donor PNPO salvage enzyme to the acceptor apo-B6 enzyme DDC requires that both enzymes form a complex to offer an efficient and protected means of delivery of the highly reactive PLP. Knowledge of the 3D protein structures of PNPO and DDC (in both active state or holo-form and inactive state or apo-form) enabled us to undertake protein-protein docking and molecular dynamics simulations studies to predict the most likely near-native structure of the complex. The physical binding between PNPO and DDC were experimentally characterized using fluorescence polarization (FP), surface plasmon resonance (SPR), and isothermal calorimetry (ITC) techniques. The dissociation constants (KD) was observed to be in low micromolar range. Expectedly, interactions between PNPO and apoDDC was found to be about 3-fold stronger than interactions between PNPO and holoDDC, with KD values of 0.92 ± 0.07 μM and 2.59 ± 0.11 μM, respectively. PLP transfer studies were carried out to demonstrate that PLP is capable of transferring from PNPO and activating the apoDDC. Site mutation investigations of critical residues identified by computational/modeling studies to be important in protein-protein interaction were carried out but showed negligible effect on the complex formation.


© Mohammed H. AL Mughram

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission