Cell and Gene Therapy Strategies for Treatment of Bone Fragility Disorders

Abstract

Current therapies for bone fragility disorders such as Osteogenesis Imperfecta (OI) can reduce fracture risk by improving bone quantity but not bone quality. Cell and/or gene therapy strategies hold promise for addressing the fundamental deficiencies in genetic bone disease but involve a number of technical hurdles that need to be overcome. This thesis takes a stepwise approach to address some of these challenges. Allogenic bone marrow transplantation (BMT) from healthy donors has been suggested for several decades to be able to repopulate the bone compartment with genetically healthy cells. However prior attempts have often featured poor osteoblastic engraftment. We describe the application of cell therapy to the mild-moderate severity Col1a2G610C OI mouse model. The effects of sub-lethal irradiation followed by transplantation of BMT from wild type (WT) mice into OI mice were analysed via DEXA, microCT, and mechanical testing. No differences were observed between the OI transplanted with WT cell group and the naïve WT and OI control groups in any measure. OI cells transplanted into OI mice were also included an additional control group to test for the paracrine effects of BMT, but again no significant differences were found compared to naïve OI controls. Lineage tracking using mice irradiated then transplanted with fluorescently labelled bone marrow cells revealed that most engrafted donor cells expressed the osteoclast marker tartrate-resistant acid phosphatase (TRAP). These results together indicate the inefficacy of irradiation and BMT on the osteopoietic compartment and suggest that alternative novel methods would be needed to increase engraftment for OI. In order to facilitate gene therapy approaches for OI, we aimed to engineer a system allowing the specific targeting of bone cells (osteoblasts and osteocytes) throughout the skeleton. Adeno associated viruses (AAVs) emerged as a prime vector candidate due to their small size, non-immunogenicity, and tissue specificity. A panel of 18 AAV variants expressing Cre recombinase and GFP under CAG ubiquitous promoter were first trialled via local delivery in a murine fracture model and in vitro on a human osteoblastic cell line. High performing variants, AAV8 and AAV-DJ were then used in systemic delivery experiments where vectors driving Cre expression via bone-cell specific promoters were designed and generated. The AAV8-Sp7-Cre vector was demonstrated to specifically and efficiently transduce osteoblasts and osteocytes throughout the skeleton. In a final series of experiments, delivery methods for the Cre constructs were compared and CRISPR/Cas9 gene editing constructs were designed and generated based on the Cre constructs design. Intraperitoneal (IP) delivery of the AAV8-CAG-Cre construct showed a similar transduction profile to intravenous (IV) delivery throughout the organs and bones. The one exception was skeletal muscle where IP delivery was able to transduce some skeletal muscle surrounding the tibia. In utero delivery was also trialled via IV delivery to pregnant female mice at ED17. This did not result in transduction of the pups, and IP injection of the pregnant female mice or direct injection of the pups in utero should be trialled. Finally, two AAV8 CRISPR/Cas9 constructs (a self-assembling intein system) able to drive gene editing of the Ai9 locus were produced. Preliminary studies showed a lack of gene editing in target tissues, and hence further studies were conducted to troubleshoot the constructs. HEK293 cells transduced with the virus in vitro showed staining of the N-terminal Cas9 intein, however the C-terminal Cas9 intein has yet to be validated. Further studies will be taken to resolve issues with these constructs to produce vectors able to mediate global skeletal gene editing in the Ai9 mouse. In summary, the published papers and subsequent experiments detailed in this thesis represent a stepwise approach for developing a gene therapy solution to genetic bone diseases. The creation of a bone specific Cre expressing AAV vector is expected to have remarkable utility as a tool for generating timed bone specific knockouts in floxed mouse lines. Its specificity and efficiency are particularly notable features. It is anticipated that rectification of one or more of the components of the split CRISPR/Cas9 approach will ultimately enable high-efficiency gene editing in bone, which will be a major advance for the field.