Virus-like particles (VLPs) are biological nanoparticles formed from the self-assembly of viral subunits into a shell or capsid, which lacks the infectious DNA or RNA genome of the virus. These VLPs are empty, non-replicative, non-infectious virus shells and find commercial use as vaccines (e.g. Merck's Gardasil vaccine for cervical cancer). The empty ‘shells' can also be packaged with therapeutic substances or DNA opening options for new products in the areas of drug delivery and gene therapy. However, the large-scale manufacture of VLPs is still challenging and current processes result in products having low architectural stability and low molecular homogeneity. We are presented with a processing challenge to create a VLP that will form homogenous and stable quaternary structures at large-scale, without losing therapeutic biological function. Our approach to this processing challenge is to create a VLP that either has better quaternary stability or less proteolysis, or ideally both. Mutational studies of virus capsids have provided important insights into the protein's structure-function relationships, particularly the importance of individual residues in capsid assembly, disassembly, and stability. The mutation selection procedure is a key step for efficient mutational studies. We suggest that proteolysis of proteins may account for some heterogeneity of VLPs. Assembly of truncated proteins would result in quaternary structures with varied size and shape. Through mutagenesis we can eliminate protease cleavage points on the molecule, without losing the innate function of the protein. The assembly and the overall stability of the viral capsid can also be enhanced with mutagenesis. By introducing structurally significant interactions within the protein we can modify the structure to stabilize the VLP, whilst maintaining surrounding interactions. Using bioinformatics techniques we can isolate specific points of the protein to target based on predictions of the effects of the mutation. This allows us to change the molecule to increase the processability of the VLP, without destroying the self-assembly characteristics. The homogeneity of VLP products produced at large-scale should increase as a result of the changes, as well as providing a more stable VLP. Here we show an application of our approach using the Murine Polyomavirus capsid. Through an extensive computational analysis, we were able to propose a list of mutations to improve VLP stability and to reduce proteolysis, based on evolutionary comparisons, protease cleavage predictions, and disulfide bonding. Preliminary results for the experimental validation of the mutations will be presented, including expression, solubility, and assembly data. By taking an in-silico approach to this process-inspired biomolecular engineering problem, we attempt to predict the expected outcomes of in-vitro mutational analysis, and in turn increase the efficiency of both the investigational and biomanufacturing processes.

Extended Abstract Status: Not Uploaded