Therapeutic Modulation of Complement Activity and DNaseI Expression using AAVmediated Gene Transfer

Abstract

Vectors based on adeno-associated virus (AAV) are emerging as systems of choice for clinical in vivo gene transfer for effective and persistent transgene expression. The promise for further success is heralded by discovery of AAV capsids with higher tissue tropisms and the rapid improvement in technology encourages development of novel approaches for disease treatment. The current project has sought to develop liver targeting strategies to modulate innate immunity through elevated serum protein levels. Vectors encoding DNaseI or complement regulatory proteins Factor I (FI) or a modified Factor H (FH) protein were packaged to permit efficient transduction of the murine liver. Studies with FI extend previous in vitro studies which found elevated FI levels normalises the complement Alternative Pathway in individuals with hyperactive “complotypes” at risk of disease. AAV delivery induced a 5-fold in vivo increase in murine serum FI levels leading to a 50% reduction in complement activity. Further studies to demonstrate bioactivity were hampered by availability of suitable models. Elevated FI did not prevent renal damage in models of ischemia reperfusion injury or FH deficiency but indicated that (i) AAV delivery of the human FH “minigene” is effective as gene replacement therapy and (ii) FI is not a limiting cause of pathology in FH-deficient mice. Elevation of serum DNaseI levels was pursued based on multiple reports indicating exogenous DNaseI is a promising approach to slow autoimmunity. AAV gene delivery induced high (>150-fold) and sustained (>6 months) activity of serum DNaseI in lupus-prone NZBWF1 mice resulting in decreased glomerular deposition of C3 and IgG. However, there were limited effects on autoantibodies, renal pathology and animal survival. Collectively, the current data demonstrate potential for continued development of AAV immunotherapy through optimised transgene delivery and further exploration in relevant models.