Retroviral replication relies on atomic-scale phenomena such as the quantum chemistry of bond formation and large scale processes such as protein self-assembly. These processes are fundamentally multiscale in nature, since they span time and length scales from the molecular to the mesoscopic. For instance, during viral particle maturation, proteolytic cleavage of the group specific antigen polyprotein (Gag) releases capsid (CA) proteins, which are subsequently reassembled into a conical capsid. Our long-term goal is to develop and apply novel multiscale computational approaches to study retroviral replication. To achieve this goal, there are two main foci. The first focus is to develop multiscale simulation methods for which there are two separate strategies. (1) Strategies that link atomic-level structures and simulations to lower resolution models using rigorous statistical mechanical approaches or so-called ?bottom-up? coarse-graining methods. (2) Strategies that use experimental data to directly develop coarse models, which are subsequently refined at higher resolutions. These are ?top-down? approaches. The second focus is to apply these methods to uncover the mechanisms by which key proteins might influence retroviral replication. This proposal will concentrate on three biomolecular systems to probe the viral replication process: the innate immune sensors, TRIM5?, that recognize and destroy viral capsids, the CA protein that encases the viral DNA, and CA hexamers that form pores in the capsid. Investigations into the biophysical mechanisms involved will be guided by three specific aims (1) to determine how TRIM5? nucleate and grow protein lattices that encage capsids, (2) to identify the origins for capsid polymorphism, and (3) to investigate chemical features of capsid pores that contribute to capsid stability or viral function. Top-down coarse-grained models of TRIM5? and CA proteins will be developed and made more physically accurate using available experimental, structural, and biochemical data. Bottom-up coarse-grained models for CA will be constructed using atomistic simulations of viral capsids. Microsecond timescale atomistic simulations of CA hexamers capsid pores will also be performed to lay the groundwork for reactive molecular dynamics simulations approaches that allow for proton transport and chemical bond formation. A key approach, at all stages of our studies, is to develop multiscale computational methods that couple each level of description to the next. Our computational predictions on retroviral mechanism will be validated and/or refined in collaboration with three leading experimentalists. Collectively, insights from these studies will broadly impact the fields of molecular simulation, virology, and biophysics. Findings from these studies have the potential to aid in the development of new therapeutic strategies to combat retroviral infection.
This project concerns the development and application of multiscale computer simulation methods to study retroviral replication in collaboration with experimental structural biologists and biochemists. The specific systems to be studied are implicated in the anti-retroviral innate immune response, infection of HIV, and assembly and disassembly of viral capsids. The findings from this work will advance fundamental knowledge in molecular simulation, virology and biophysics and have the potential to aid in the development of new antiviral therapies.
