Recombinant adeno-associated viruses (rAAV) represent the basis of modern gene therapy. They can be modified extensively at the genomic and capsid level enabling personalized therapies. The high potential of this technology and the great successes that have been achieved in this field require methods that allow for the development of modified rAAV variants, large production scales of viral vectors and finally the generation of high purity products for medical therapy. This thesis dealt with the expansion of a plasmid system for recombinant adeno-associated virus serotype 2 (rAAV2) production, the generation of rAAV producer cell lines and the development of a novel rAAV2 purification method based on affinity chromatography.
For a deeper understanding of the existing plasmid-based expression system, a stability assay was developed which aims at the biologically relevant release of rAAV DNA under increasing temperature exposure. Using this technique, it was discovered that, the thermal stability of capsid-modified rAAV2 particles bearing glycine-serine amino acid linker insertions at position 587, decreases with increasing linker length. In addition, a rAAV2 variant was generated which presents 60 active β-lactamase enzymes on its capsid. This variant had a similar thermal stability as the wild type variant and proved the enormous modifiability of the rAAV2 capsid.
Within the scope of cell line generation experiments an attempt was made to insert fluorescent reporter genes into suitable genomic target sites for the later insertion of the adenoviral E1a/b genes into the CHO K1 (chinese hamster ovary K1) genome using a CRISPR/Cas9 mediated rational approach. These experiments did not lead to the desired success, but it could be demonstrated that it is possible to produce rAAV2 particles using CHO K1 cells and that the production can be improved by the addition of the E1a/b genes. Later, monoclonal HEK293 cell lines could be generated, providing the helper genes required for rAAV production through random genomic integration, thus reducing the cost and effort of rAAV production.
In order to improve rAAV2 production, an attempt was made to generate a novel affinity chromatography based on a single-chain Fv (scFv) antibody designed using the murine A20 antibody structure. The construct was expressed in the periplasm of E. coli. Silica material was used as base material for the chromatography and the protein was immobilized via a fused silica tag. The principal functionality of the construct was verified. However, it turned out that silica is not suitable for rAAV purification due to its non-specific binding properties. A second construct, in which the scFv was fused to a human IgG-Fc part was produced in HEK293 suspension cells and has since been used for AAV capsid ELISAs. In the search for another suitable affinity ligand that can be expressed bacterially, the natural AAV receptor (AAVR) came into focus. A fusion protein was designed, which consists of the PKD2 domain (polycystic kidney disease 2) from the AAVR in combination with a cellulose binding domain. The suitability of this construct in the context of rAAV2 affinity chromatography was proven and the production could be significantly improved by the genetic fusion with a maltose binding protein. In further studies, a fast, cost-effective and very efficient purification process for rAAV2 particles based on PKD2 was developed.
Ultimately, this thesis led to an improvement of rAAV production and modification, resulting in a contribution to rAAV-based gene therapy.
