Coarse-grained molecular dynamics simulations of thermoresponsive polymers

Date
2017
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Thermoresponsive polymers are present in many biomedical applications such as drug delivery and tissue engineering for their ability to change phase or morphology with changes in temperature. This thesis presents a coarse-grained (CG) molecular dynamics approach to understanding thermoresponsive behavior in three systems: elastin-like polypeptide (ELP), collagen-like polypeptide (CLP) and oligonucleic acid (ONA). CG molecular dynamics simulations allow for improved computational efficiency and facilitate study of macromolecular systems at length and time scales relevant to thermoresponsive behavior such as melting, aggregation, and assembly. The first system presented in this thesis is the elastin-like polypeptide (ELP)-collagen-like polypeptide (CLP) conjugate in an aqueous solution. ELP-CLP conjugates self-assemble above the lower critical solution temperature (LCST) of ELP into vesicles with drug delivery applications. An implicit solvent CG model of ELP-CLP conjugates was developed where ELP strands are considered as bead-spring polymers and CLP triple helices are considered as rigid rods. Through molecular dynamics simulations over a range of ELP-ELP interaction strengths, it was determined that ELP-CLP conjugates with multiple ELP strands conjugated to the CLP rigid body undergo the onset of aggregation at a lower ELP-ELP interaction strength than a corresponding system of ELP strands that are not conjugated to CLP. In addition, simulations of a system with only one ELP strand conjugated to a CLP rigid body did not show this decrease in the onset of interaction, thus supporting the hypothesis that ELP crowding due to conjugation lowers the LCST of ELP-CLP conjugates in experiments with respect to the LCST of free ELP strands. ☐ The second system presented in this thesis is the CLP triple helix in aqueous solution. This study primarily focuses on developing a phenomenological CG model of the CLP triple helix that mimics melting of CLP triple helix. By developing and implementing this CG model in the ELP-CLP conjugate system, analysis of ELP-CLP conjugates where CLP is in the melted state is possible. Using implicit solvent molecular dynamics simulations, the phenomenological CG model of CLP was able to qualitatively capture trends in melting temperature of the CLP triple helix with respect to CLP strand length and with substitution with charged amino acids. Further, this new model for CLP was substituted for the rigid body in the ELP-CLP conjugate system and simulations of this updated model demonstrated qualitative agreement with the original model with the added property of being able to undergo a melting transition. ☐ The last system that was studied in this thesis is the oligonucleic acid (ONA)-star polymer conjugate system. A CG model for ONA-star polymer conjugates was developed where the ONA portion of the model was taken from a previous two site per nucleotide model that captures specific directional hydrogen bonding (H-bonding) interactions. The star polymer portion was modeled as a bead-spring polymer. Utilizing implicit solvent CG molecular dynamics simulations, both melting transition behavior and assembly transition behavior was studied with respect to changing ONA backbone flexibility, backbone charge, and number of star polymer arms. Increasing melting temperature was observed for ONA-star polymer conjugate systems with decreasing backbone flexibility and backbone charge. Assembly transition temperature increases accordingly, and, in addition, increases with increasing number of star polymer arms per ONA-star polymer conjugates. ☐ This thesis demonstrates the usefulness and importance of CG molecular dynamics simulations in screening large materials design parameter space, specifically for thermoresponsive polymers. These simulations provide a better understanding of the thermodynamics and driving forces for thermoresponsive behavior, and thus better inform scientists on how chemical and physical features of the polymer influences LCST, melting, and assembly behavior. In addition, using a CG framework, instead of experimental or atomistic framework, reduces the time and costs associated with testing a potentially vast design space of a thermoresponsive polymer. Future improvements in the models listed in this thesis may allow for better qualitative and quantitative characterization of thermoresponsive and self-assembly behavior in polymers.
Description
Keywords
Applied sciences, Biomimetic polymers, Coarse-grained modeling, Molecular dynamics
Citation