(M1530-01-01) Exploring AAV Capsid Aggregation: Molecular Simulations of Capsid Interactions in Varied Ionic and Surfactant Environments

Purpose: Adeno-associated virus (AAV) is a non-enveloped virus with approximately 26 nm diameter, which exhibits great potential as a gene delivery tool. However, its manufacturing faces challenges, including lower yields and moderate purity, primarily due to degradation at various production stages. This degradation occurs potentially via aggregation at low ionic strength and insufficient surfactant concentration. These mechanisms of degradation remain poorly understood, and there is a lack of a systematic approach to formulation development. To address this, a coarse-grained molecular dynamics (CG-MD) simulations approach was utilized to understand the molecular level of capsid-capsid interaction under different types and concentrations of salts and surfactants. The model was validated with experimental data.

Methods: CG-MD simulations were performed using the GROMACS package with MARTINI force fields. AAV8 (PDB: 6v12) with 2.2kb single-strand DNA was selected as the model viral capsid. Energy minimization used the steepest descent algorithm, followed by a 10 ns equilibration step. Production runs were performed using the Nosé−Hoover thermostat and the Parrinello−Rahman barostat at 1 bar pressure for each 20-fs time step in an isothermal-isobaric ensemble. Trajectory generation was conducted through numerical calculations on the High-Performance Center Supercomputer Cluster. This sophisticated computing resource, which is equipped with Nvidia Tesla v100 units provided by Nvidia Corp, is hosted by ITS and the University of Connecticut in the United States.

Results: We investigated the effect of different types and concentrations of salts on AAV vector aggregation by using different analysis tools. To evaluate the capsid interaction and aggregation, the variation of the COM distance between two AAV capsids in systems containing different types and concentration of salts and surfactant was first measured. The Initial center of mass of two capsids in zero-time step was set 30 nm. Overall, the COM distance decreased as the simulation progressed, indicating an increased attraction between the two capsids at a 0.05 M NaCl concentration. However, for higher concentrations of NACL salt and MgCl₂ salts , the COM distance increases. These results are confirmed by total interaction energy  analysis and RMSD analysis between two capsids. Our results showed that NaCl and MgCl₂ salts at various concentrations could prevent aggregation to some extent. AAV vector aggregation was influenced by the ionic strength of the solution, and NaCl salts were effective at preventing aggregation when present at a concentration above 0.05 M. Additionally, surfactants such as poloxamer 188 at concentrations ranging from 0.001% to 0.01% (w/v) were tested, and our findings suggest that the type and concentration of surfactants are critical factors that affect AAV vector aggregation.

Conclusion: Our developed computational model, with a microsecond timescale, evaluated the behavior of different salts and surfactants with varying concentrations to understand the factors preventing AAV8 aggregation. It was discovered that a concentration of 0.05 M of NaCl leads to the aggregation of AAV capsids, and concentrations above 0.05 M can inhibit this aggregation. Additionally, it was observed that MgCl₂ increases the distance between the centers of mass, which reduces the likelihood of aggregation. This analysis provided insights that would be difficult to obtain from experimental studies. The CG-MD simulations proved powerful in simulating macromolecular complexes and guiding the interpretation, and direction of experiments. Our simulations provided important insights into AAV vector formulation development, which could lead to the development of more efficient and stable gene delivery systems.


Fig. 1. Measurement of center of mass (COM) distance between two full capsids in systems with different NACL concentration



Fig.2. Total interaction energy (the van der Waals energy) between two full capsids in systems with different NACL concentration



Fig.5. snapshot showing aggregation of two AAV capsids in system with 0.05 M of NaCl. Time= 2000 ns