3.3 Genome Modification with Zinc Finger Nucleases 39two sites to introduce two DSBs that may result in a large deletion or chromo-
somal translocation [51].
HR-based methods can be used to introduce transgene sequences or small
mutations at target sites by providing a homologous donor DNA template.
This template should include 700 bp homology arms if it is typical double-
stranded circular DNA [27]. For linear DNA, only 50 bp of homology is required
[19]. Single-stranded DNA oligonucleotides have also been used to achieve
point mutagenesis, deletions, or insertions [52, 53]. As compared with HR
methods that involve simply introducing the homologous sequence with the
desired mutation, introducing a targeted double-stranded DNA break enhances
the efficiency of genome editing by many orders of magnitude [18, 54–57].
HR-based methods do not work in every cell type, however, as they require the
presence of the HR machinery, which is only available during the S and G2
phases of the cell cycle just prior to mitosis. Strategies employing HR can be
achieved at desired rates in early stem cells. However, HR-directed genomic
modification cannot be achieved at appreciable rates in many differentiated
cell types because these cells are not dividing and do not have the HR machin-
ery available. This presents a major roadblock for the design of a gene therapy-
type strategy based on ZFN-induced HR. There are also species-specific
differences in the frequency of HR to consider: for example, mouse embryonic
stem (ES) cells are more prone to HR and thus easier to modify than human ES
cells [58, 59].
HR is an elegant and seamless method to create perfectly tailored DNA
sequences in the genome, but many somatic cells rely on the NHEJ repair path-
way instead. NHEJ-based gene disruption is much easier to achieve than HR,
although the resulting mutations are not predictable. Thus far the only clinical
trial to date using a genome modification system based on nucleases uses a ZFN
pair to disrupt the CCR5 locus [13, 14]. Since the CCR5 cell surface receptor is
required for most HIV infection, disruption of this locus was used to create CD4
T cells that are unable to be infected by the HIV [13, 14]. The phase I clinical trial
indicated that patients infused with T cells that were modified via ZFN technol-
ogy to lack functional CCR5 receptors exhibited a slower rate of decline in the
modified CD4 T cells relative to unmodified T cells [13, 14]. Among the 12 clini-
cal trial participants, one serious adverse event was reported of a patient suffer-
ing fever, chills, and joint pain the day following infusion [14]. Nevertheless, the
authors concluded that the autologous CD4 T-cell infusions were safe [14]. They
also found that the blood level of HIV DNA decreased in most patients and one
out of four patients tested had no detectable traces of HIV RNA in the tested
samples, suggesting efficacy [14].
In addition to gene disruption via NHEJ, pairs of ZFNs may be designed to
induce chromosomal translocations [60] by causing two simultaneous DSBs at
desired locations. This technique could be used to study translocations that are
important for cancer formation. The same two-DSB strategy will also produce
deletions of up to 15 Mb [51]. These deletions could be used to remove an exon,
an entire genomic locus, or even a number of genes from the genome. Therefore,
multiple options exist both for achieving the desired genome modification and
the methods by which to achieve those modifications using ZFNs.