Gene therapy has vast potential for treating and potentially curing a wide variety of disorders. However, gene delivery technologies require significant improvements in safety, efficiency, and expression stability before the majority of these diseases can be treated. Vectors based on adeno-associated virus (AAV) have proven themselves to be highly promising, both in the laboratory and the clinic, but they still suffer from several shortcomings. In particular, the majority of the human population has been exposed to AAV serotype 2, as well as other serotypes, and as a result the immune system is primed to neutralize AAV. Antibody neutralization of AAV vectors is an established problem, and cellular immune responses may also be a challenge. We will attempt to solve the former problem and will further investigate basic mechanisms involved in the latter. For the former, we have developed novel directed evolution technology to generate new mutants of AAV with new properties. Specifically, large libraries of virus with random point mutations in the capsid gene encoding the viral coat protein are generated, and variants with novel properties are selected using high throughput screens. We have utilized this approach to generate variants with altered receptor binding properties, as well as variants that escape neutralization by antibodies that greatly inhibit AAV gene delivery by the wild type capsid or coat proteins. We will study the potential of human antibody evading variants to mediate high efficiency gene delivery of the therapeutic gene erythropoietin to the muscle and liver of animals carrying anti-AAV antibodies. In addition, while AAV neutralization by antibodies is an established problem, much less is known about AAV interactions with other components of the immune system. Therefore, the basic mechanisms of immune neutralization of this virus by complement [and T cells] will be investigated to both in vitro and in vivo. In summary, viruses have naturally evolved for their own ends, which do not always meet the needs of a human therapeutic. The novel approaches developed in this work to re-evolve viruses into enhanced human therapeutics will therefore have broad and general impact on the molecular engineering of enhanced viral gene delivery vehicles, including alternate AAV serotypes as well as other vectors. Furthermore, it will yield insights into the responses of other immune system components to AAV.