Mechanistic Insights into Dynamics and Thermodynamics of Biomolecular Processes: Protein Unfolding and Aggregation, DNA Nanomechanics, and Drug Delivery

Abstract

Biophysics has seen unprecedented progress, applying concepts from physics to study intriguing biological phenomena. Further advances in this field require fundamental understanding of various processes at the nanoscale and development of appropriate methods and models for different applications. Molecular simulation is playing an ever-increasing role for these purposes. In this thesis, I have examined the structure, dynamics and thermodynamics of various biomolecules of interest using molecular simulation and theoretical modeling. The thesis is organized as follows. In the 1st chapter, I briefly introduce various bioactive molecules and relevant biological phenomena. The 2nd chapter consists of detailed descriptions of simulation methodologies and theoretical frameworks. These include classical molecular dynamics (MD) simulations, advanced sampling techniques for free energy calculations, and various entropy calculation methods. In chapter 3, we propose a carbon nanotube (CNT)-based drug-delivery method. One of the major challenges of nanomedicine and gene therapy is the effective delivery of drugs and genes across cell membranes. Generally, bioactive molecules used as drugs or drug-delivery vehicles cannot passively pass through the cell membrane due to the high penalty associated with membrane rupture. We show via MD simulations that molecules of various shapes, sizes and chemistries can spontaneously enter a membrane-spanning CNT nanopore. We study the thermodynamics of entry of several molecules of interest, such as dendrimers, asiRNA, ssDNA and ubiquitin protein. We show that another free CNT can spontaneously enter the CNT nanopore and eject the encapsulated molecule out of the nanopore. In this way, a macromolecule can be translocated across the cell membrane. We also verify the thermodynamic feasibility of the proposed method. This method should work for other molecules as well, and hence could be potentially useful for drug-delivery applications. The fourth chapter deals with the understanding of complex force-dependent protein unfolding kinetics. For some proteins, e.g., ubiquitin, the unfolding rate at very low forces doesn't vary much up to a critical force, after which the rate increases exponentially by increasing the force further. This crossover in the unfolding rate can be due to one of the following two scenarios. First, there are two unfolding pathways for the protein and pathway-switch occurs after the critical force. Second, the unfolding pathway can change continuously due to force-dependent modifications in the free-energy landscape. By performing nonequilibrium MD simulations of ubiquitin at forces ranging 20–800 pN, we find a crossover in the unfolding rate and show that the crossover is due to the second scenario. We rationalize the results by using multidimensional transition state theory. The findings from this chapter will have implications in understanding the folding/unfolding kinetics of protein which is one of the outstanding problems of the current century. In the 5th chapter, we decipher the molecular mechanism and thermodynamic driving force for lower critical solution temperature (LCST) phase behavior of the aqueous solution of proteins induced by multivalent ions, observed recently in experiments. LCST phase behavior manifests itself as phase separation of the protein–salt solution upon heating. This has been attributed to entropy effects. We use MD simulation along with the two-phase thermodynamic method for entropy calculation. Our simulations reveal two key steps that help in explaining the LCST phase behavior. First, the cations binding to the protein: this requires the release of tightly bound water molecules from the solvation shells of cations and partial desolvation of the protein surface residues, which are indeed entropy driven. Second, the protein-bound cations attract other proteins present in the solution, whose binding is again entropy driven, resulting in LCST behavior. By performing series of simulations of protein in chloride solutions of various cations (Na+, Ca2+, Mg2+ and Y3+), at temperatures ranging 283–323 K, we suggest that multivalent cation binding to any negatively charged surface can be entropy driven. These findings have direct implications for tuning the phase behavior of soft matter systems, such as reentrant condensation and protein crystallization. In a broader context, molecular-level understanding of interactions of heavy metals—usually not found in healthy cells—with different biomolecules can provide insights for carcinogenicity and neurotoxicity induced by exposure to such environmental contaminants. In chapter 6, we provide a molecular-level understanding of how intercalation of a drug affects DNA mechanics. Most of the anticancer drugs are known to intercalate in-between two consecutive base-pairs of a double-stranded DNA (dsDNA). These DNA-intercalators are believed to hinder DNA replication and transcription, eventually leading to cell death—thus acting as anticancer drugs. We probe, using MD simulations, change in the mechanical properties of the intercalated drug–DNA complexes for two intercalators, daunomycin and ethidium. We find that, upon drug intercalation, the persistence length and bending modulus of dsDNA don’t change significantly, whereas its stretch modulus increases by as much as 65%. Steered MD simulations also reveal that it requires higher forces to stretch the drug-intercalated dsDNA complexes than the bare dsDNA. Adopting various pulling protocols to study force-induced DNA melting, we find that dissociation of the dsDNA complex becomes difficult in the presence of intercalators. The results obtained here provide a plausible mechanism of action of the anticancer drugs—i.e., via modifying the mechanical properties of DNA. Finally, in chapter 7, I summarize all the results with concluding remarks and future outlooks