Investigation of Thermal Stability and Functional Dynamics of Nucleic Acid Nanostructures using Molecular Dynamics Simulations

Abstract

DNA has emerged not only as the carrier of genetic information but also as one of the most studied and most versatile nanoscale materials in modern science. Since Ned Seeman first proposed the concept of DNA nanotechnology 40 years ago, DNA has been used as a programmable scaffold for engineering various multifunctional structures (e.g., DNA origami nanotubes, flasks, and polyhedra cages of different sizes and shapes) in applications such as drug delivery, biosensing, nanomedicine, and molecular electronics. Along with DNA, its synthetic charge-neutral analogue, ”PNA” offers exceptional thermal stability and sequence-specific binding affinity to DNA/RNA, making it a promising nanomaterial for therapeutic applications. However, real-life applications of DNA/PNA nanostructures are hindered by limited molecular understanding of their thermal stability, enzymatic-degradation resistance, drugloading mechanisms, and sequence-dependent interactions; questions that remain challenging to probe experimentally. In this thesis, we use atomistic molecular dynamics (MD) simulations, enhanced sampling techniques, and free energy calculations to understand the structural, mechanical, and thermodynamic properties that govern DNA and PNA-based nanostructures. We first demonstrate that terminal hydrogen bond “fraying-peeling” dynamics and backbone charge dictate fundamentally different melting pathways in nanoscale triplexes: sequential “two-step” unzipping in DNA triplex vs. cooperative “onestep” dissociation in PNA triplex, with PNA triplexes showing very high thermal stability and sharp melting profiles. We further investigate crossover-rich PX/JX DNA motifs to understand their resistance to nuclease degradation. Using umbrella sampling PMF calculations and crossover-induced mechanical rigidity, we demonstrate that crossover geometry significantly restricts DNase I binding, establishing PX/JX as highly stable candidates for enzyme-resistant nanoscale drug-delivery platforms. We also developed a tetrahedral DNA nanostructure (TDN) and uncovered why its unique topology and multimodal binding regions make it exceptionally suitable for drug delivery. Integrating MD simulations with experimental validation, we show that TDNs can efficiently encapsulate and transport dopamine neurotransmitters across brain cell membranes, with superior blood-brain-barrier (BBB) permeability and 90% loading efficiency, highlighting the translational potential of DNA nanostructures for neurological disorders (e.g., Parkinson’s disease) Finally, with an understanding of the thermal stability, enzymatic resistance, and drug loading mechanism of DNA/PNA nanostructures, we investigate how tuning the sequence-dependent AT/GC components modulates DNA compaction by short arginine-rich protamine peptide in sperm cells. Our MD simulation results reveal distinct protamine binding modes: GC-rich DNA favours major groove binding, while AT-rich DNA prefers minor groove binding. These results demonstrate that protaminemediated DNA condensation is highly sequence-dependent, which can influence DNA mechanical properties, compaction behaviour, and chromatin accessibility. The works presented in this thesis thus bridge molecular-scale interaction insights with the real-world translational potential for the next generation of nucleic acid nanotechnology.