Figure 4.

Schematic of sequential protein engineering to reprogram meganuclease specificity. Meganucleases are first retargeted to ‘round 1′ targets, DNA sequences containing only one cluster of consecutive nucleotide substitutions from the original target (WT). Following rounds of in vitro selection, selected endonucleases are further mutated and screened to collect endonucleases that cleave next rounds of target sites, which contain additional base-pair mismatches. This iterative approach results in engineering of variant enzymes that target ‘half’ altered sites. Half domains of such proteins are shuffled to create meganucleases that recognize ‘fully’ altered targets.