Improving Viral Vectors for Gene Targeting in Gene Therapy

Over 10,000 monogenic diseases in the world affect one out of every hundred live births (WHO). Gene targeting is a term that is used describe the manipulation of genetic material, either by adding a gene in a specific locus, creating a mutation at a specific locus, or correcting a gene at a specific locus. Here, unless otherwise noted, we will use the term to describe the correction of a gene with a homologous piece of donor genetic material whereby a mutant gene that causes monogenic disease is essentially replaced by a wild type copy through homologous recombination. Thus, gene targeting is inherently safer than classic gene therapy, where a gene is randomly introduced into the genome and can cause insertional mutagenesis. Although the rates of homologous recombination are low when simply delivering a donor substrate (1 in a million), creating a deoxyribonucleic acid (DNA) double-stranded break in or around the gene of interest using a nuclease, increases the rate of gene targeting 30,000-50,000 fold. The delivery of the nuclease and donor substrate to these cells is one of the major hurdles in achieving this type of therapy. However, for classic gene therapy there have already been many clinical trials using viral vehicles for gene delivery. One problem with using a virus for gene therapy is the low titer associated with some types of virus, in particular, lentivirus. In the first part of this dissertation, this problem is addressed by showing that the addition of caffeine during viral production can increase titer up to 8-fold. Besides lentivirus, other viruses, like Adeno-associated virus (AAV) have been used in clinical trials. There are nine AAV serotypes, but the most-well characterized is AAV2. Because there are situations where AAV is to be used in cells that cannot be transduced with AAV2, it is essential to know which serotype best infects the desired cell type. The second part of the dissertation describes a comprehensive survey of the ability of AAV1-9 and one engineered serotype to transduce primary and immortalized cells from human, mouse, hamster, and monkey origin. Overall, the results show that AAV1 and AAV6 transduce the most cell types at the highest efficiencies. Though gene targeting has been achieved using the homing endonuclease I-Sce in AAV2, targeting has never been achieved using two zinc-finger nucleases (ZFNs) in any AAV serotype. This is significant because the recognition site for I-Sce is not found in the human genome, while ZFNs are designed to specifically bind in or around a gene of interest. Based on the results from the AAV survey and the advantage of ZFNs, we created an AAV6 virus that carried the genetic information for both ZFNs and donor substrate for gene targeting in cells containing a GFP gene targeting system. We also created an AAV6 virus that carried the donor substrate alone. The third part of this dissertation reveals that dual infection at the optimal multiplicities of infection for both AAV viruses can achieve targeting efficiencies of ~3%, which is ~3-fold higher than by lipofection. Furthermore, we show that the addition of the proteasome inhibitor, MG132, increases the gene targeting level an additional 2-fold. This data suggests that AAV is a great choice for gene therapy by gene targeting. Chapters 3-5 within this body of work make significant contributions to the gene therapy field. The work and the contributions will be described in each section respectively as well as summarized in the last chapter.