Chemistry and Biochemistry
Jeffrey D. Evanseck
Mihaela Rita Mihailescu
Computational Chemistry, Parameterization, MD Simulation, Biophysics, G-quardruplexes, Decarboxylation, Hairpin, Peptide Nucleic Acids, Solvent, Androgen Receptor Allostery
Biophysical phenomena are modeled using a combination of quantum and classical methods to interpret and supplement three distinct and diverse problems in this dissertation. In the first project, decarboxylation reactions are ubiquitous across chemical and biological disciplines, yet the origin of non-catalytic solvent effects remains elusive. Specific solvent structure and energetics have not been well described for the monoanion of malonate, nor corrected from the gas-phase charge-assisted intramolecular hydrogen bond model known as “pseudochair”. In the aqueous phase, a low-lying energy conformer known as the “orthogonal conformation” is computed to be preferred by a three-water cluster of hydrogen bonding over the pseudochair intramolecular hydrogen bond accounting for 87% of the experimental activation enthalpy. The orthogonal conformation is further stabilized by water clusters that satisfy the ten theoretical donor/acceptor hydrogen bonding sites on malonate, reproducing the enthalpy of activation with errors under 0.6 kcal/mol, underscoring the relationship between solvent effects, conformation, and activation parameters for decarboxylation. In the second project, enzyme allostery can be induced by exogenous ligands, thereby impacting enzyme kinetics. Yet, a hypothesized allosteric effect of dichlorodiphenyldichloroethylene (p,p’-DDE) that inhibits androgen receptor’s (AR) activity, by inducing release of the endogenous ligand dihydrotestosterone (DHT) has not been well studied or defined. Through enhanced molecular dynamics, a series of residues were computed to transmit an allosteric response from the binding factor 3 (BF-3) site to the active site, destabilizing DHT through conformational changes of AR. The most probable potential path for the allosteric response is transmitted through a series of residues connecting the BF-3 site to the active site, inclusive of: Phe673, Val715, Leu722, Phe725, Ile737, Trp741, Met742, Leu744, Met745, Leu812, Phe813, Tyr834, and Ile899. Each amino acid changes rotameric state upon the stabilization of Phe673 and Tyr834 at the BF-3 site by p,p’-DDE. Three egress directions were identified, and the dissociation free energy of DHT when p,p’-DDE is bound and unbound at the BF-3 site were compared. For the three paths, the dissociation energy was lowered, relative to simulations without p,p’-DDE, suggesting that DHT is destabilized within the active pocket when p,p’-DDE is bound at the BF-3 site underscoring how allostery triggers dissociation. In the final project, to interpret the phenomena of aggregation of complementary gamma-modified peptide nucleic acids (gPNAs) as a preliminary step to understanding differential G-quadruplexes (GQ) and hairpin (HP) RNA binding, torsional and electrostatic parameters for the miniature polyethylene glycol (miniPEG) modified gPNA backbone were parameterized for MD simulation. The relative energetics for the backbone were parameterized to be within 0.2 kcal/mol. Our MD simulation indicates that the miniPEG reduces base pair hydrogen bonds selectively at A-T base pairs, ultimately promoting duplex dissociation to gPNA single strands. GQ and HP structures were independently simulated as a preliminary step to investigate gPNA binding. The parameterization of modified gPNA and structural derivation of GQ and HP structures are necessary first steps in identifying and exploiting important gPNA/RNA interactions.
Tamez, A. (2022). QUANTUM COMPUTATIONS AND MOLECULAR DYNAMICS SIMULATIONS: FROM THE FUNDAMENTALS OF ANTIMICROBIAL RESISTANCE TO NEUROLOGICAL DISEASES (Doctoral dissertation, Duquesne University). Retrieved from https://dsc.duq.edu/etd/2214