Repair via non-homologous end joining (NHEJ) often results in small insertions or deletions and can be used for the creation of specific gene knockouts. By using highly specific and programmable nucleases ( 4), such as zinc-finger nucleases (ZFNs) ( 5–9), homing endonucleases ( 10–12), chemical nucleases ( 13, 14) or TALE nucleases ( 15–17), specific DNA double-strand breaks (DSBs) can be introduced next to or within defective genes to trigger the cellular DSB repair ( 18) required for DSB-mediated gene surgery. Specific targeting and modification of defined DNA sequences within the human genome is a major challenge for gene therapy ( 1–3). These results present the ZF-PvuII platform as a valid alternative to conventional ZFNs. In contrast to the ‘analogous’ ZF-FokI nucleases, neither excess of enzyme over substrate nor prolonged incubation times induced unaddressed cleavage in vitro. The optimized ZF-PvuII fusion constructs cleave DNA at addressed sites with a >1000-fold preference over unaddressed PvuII sites in vitro as well as in cellula. To achieve this, we introduced amino acid substitutions into PvuII that alter K m and k cat and increase fidelity. While PvuII adds an extra element of specificity when combined with ZFs, ZF-PvuII constructs must be designed such that only PvuII sites with adjacent ZF-binding sites are cleaved. Here, we have fused a three ZF array to the restriction endonuclease PvuII to generate an alternative ZFN. TALE sequences, have been considered as alternatives to ZFs, no efforts have been undertaken so far to replace the catalytic domain of FokI as the cleavage module in ZFNs. In this configuration, the ZFs constitute the binding module and the FokI domain the cleavage module. Zinc-finger nucleases (ZFNs) typically consist of three to four zinc fingers (ZFs) and the non-specific DNA-cleavage domain of the restriction endonuclease FokI.

Palindromic sequence zinc finger