379. New Zinc Finger Protein Nuclease Architectures for More Efficient Gene Modification Therapies

Abstract

Top of pageAbstract Recent studies have shown that zinc finger protein nucleases (ZFNs) may be used to introduce precise, therapeutically beneficial changes into the genome. This approach offers considerable promise for the correction of monogenic diseases, and should also allow the ablation of genes whose expression is harmful in particular disease settings. Realizing the full potential of ZFN-mediated gene modification therapy, however, will require the development of methods for routinely engineering ZFNs that cleave with high efficiency at virtually any DNA sequence, since a separation distance of as few as forty bp between the sites of nuclease cleavage and desired sequence change can lead to a significant reduction in gene modification efficiencies. In order to address these issues we have engineered and characterized several improvements to the architecture of our designed ZFNs. In initial studies, we systematically varied the length and composition of the linker connecting the zinc finger and nuclease domains, and determined the impact of these changes on the cleavage efficiency and optimal target arrangement of the resultant ZFN (which functions as a dimer). It was observed both in vitro and in vivo that ZFNs could be efficiently designed to cleave targets separated by 4-8 bps, with maximal activity obtained when monomer sites were separated by 4, 5 or 6 bp. Next, structure-based design was used to engineer asymmetric contacts into the dimer interface in order to create ZFNs that could function only as heterodimers. This was of interest from the standpoint of enhancing both the specificity and efficiency of ZFN performance, since prohibiting formation of the therapeutically irrelevant homodimer would eliminate the possibility of off-target effects. Biochemical and cellular studies revealed that the resultant ZFNs functioned as desired. Finally, we examined whether re-ordering of the zinc finger and nuclease domains within the ZFN polypeptide would enable functional assembly of novel dimer arrangements on the target DNA (|[ldquo]|head to head' and |[ldquo]|head to tail|[rdquo]|, in addition to the natural |[ldquo]|tail to tail|[rdquo]| arrangement). This was of interest because the availability of an increased number of productive arrangements would directly enhance our capacity to design ZFNs to cleave at any chosen base. We have demonstrated that the |[ldquo]|re-ordered|[rdquo]| ZFNs may be combined with each other and with the parental architecture to yield dimers that cut preferentially in all three configurations (|[ldquo]|head to head|[rdquo]|, |[ldquo]|tail to tail|[rdquo]| and |[ldquo]|head to tail|[rdquo]|). Together, these approaches have significantly improved the cleavage efficiency and targeting specificity of the ZFN platform, and have enhanced our capacity to design ZFNs for use in therapeutic gene modification.