Fig. 1

Gene editing technologies used in cell therapies. Depicted are the three basic structures and main characteristics of each editing platform used clinically in cell therapies showing how the editing agent interacts with the DNA in order to initiate the double-strand break. a Zinc-finger nucleases (ZFNs) consist of Zinc-finger proteins bound directly to an endonuclease such as FokI. The zinc finger proteins work as DNA-binding domains recognising trinucleotide DNA sequences, with proteins linked in series to enable recognition of longer DNA sequences, thereby generating sequence recognition specificity. The fused FokI functions as a dimer so ZFNs are engineered in pairs to recognise nucleotide sequences in close proximity ensuring DSBs are only produced when two ZFNs simultaneously bind to opposite strands of the DNA. b Transcription activator-like effector nucleases (TALENs) consist of bacterial TALE proteins fused to endonucleases such as FokI. As with ZFNs this requires paired binding to initiate the DNA break. Here the DNA targeting specificity comes from the modular TALE arrays which are linked together to recognize flanking DNA sequences, but each TALE recognises only a single nucleotide. c The CRISPR/Cas9 platform does not rely on protein-DNA binding as with ZFNs and TALENs but gets its DNA targeting specificity from Watson–Crick RNA–DNA base pairing of the guide RNA (gRNA) with the recognition site. Initially the Cas9 binds to a protospacer adjacent motif (PAM) this is a 2–6 base pair DNA sequence which is specific for each Cas protein. Without the correct PAM sequence the Cas will not bind or cut the DNA. Following correct PAM identification, the Cas melts the remaining target DNA to test sequence complementarity to the gRNA. PAM binding allows the Cas protein to rapidly screen potential targets and avoid melting lots of non-target sequences whilst searching for fully complementary sequences