Past Event: Oden Institute Seminar

Molecular Simulation Studies of Nucleic Acid Systems for Biological Applications

Arthi Jayaraman, University of Colorado, Boulder

Abstract

Molecular simulations serve as valuable tools that connect molecular-level interactions to macroscopic behavior in materials and biological systems. In this talk I will present two specific systems where my research group uses molecular simulations to guide design of materials for biological applications. The first example involves use of simulations to guide design of vectors for gene therapy, which involves the delivery or transfection of therapeutic DNA into the genome of target cells. Viruses are effective at transfection but elicit dangerous immunogenic responses. Non-viral gene delivery agents while not as effective as viral vectors have a significant advantage of being non-immunogenic. Polycations have emerged as promising non-viral delivery agents due to their propensity to bind the polyanionic DNA backbone, neutralizing the negative charge of DNA backbone, and facilitating gene delivery. Recent experiments by Emrick and coworkers has shown that the architecture of poly-lysines affects, in a non-trivial manner , the efficiency of these polycations as transfection agents . Our work using molecular dynamics simulations connects poly-L-lysine architecture to the strength of binding to DNA and explains the experimentally observed trends in transfection. The second example involves using molecular simulations to explain how repair proteins recognize drug-induced DNA damage. Platinum-based anti-cancer drugs are designed to covalently bind to DNA to form adducts in order to inhibit continuous cell growth. Differential recognition of various drug-DNA adducts by repair proteins, like HMGB1a, has been linked to differential tumor resistance. We use molecular simulations to show the molecular reasons behind why HMGB1a differentially binds to adducts of cisplatin and oxaliplatin, two anti-cancer drugs used for treatment of cancer. We also explain why the binding affinity studies of HMGB1 with these two drug-DNA adducts is dependent on sequence context i.e. bases flanking the drug-DNA adduct. Our work has shown that the structure of drug-DNA bound to HMGB1a protein cannot explain differential recognition. Instead the differences in conformational dynamics of the Cp-DNA and Ox-DNA in various sequence contexts before the protein binds and the free energy of deforming (e.g. bending and minor groove opening) the drug-DNA during protein binding explain the differential protein binding affinity for the two drugs in the various sequence contexts.