Search

HIV entry into target cells requires binding of the envelope protein GP120 to the CD4 co-receptors CCR5 or CXCR4. Loss of these HIV co-receptors is therefore attractive as a potential therapy for HIV patients, yet no current methods permit the efficient therapeutic disruption of a chosen gene in the human genome. To this end, we have developed designed zinc-finger protein nucleases (ZFNs) to target predetermined sequences within the CCR5 gene. Once these ZFNs create a double strand break (DSB) in the targeted DNA, natural DNA repair pathways, including Non-Homologous End Joining (NHEJ), subsequently repair the DSB. Importantly, however, NHEJ is error prone and thus can result in permanent disruption of the targeted gene.

Zinc finger nucleases have been shown to be effective tools for high-frequency genome editing with potential therapeutic applications (Urnov et al., Nature 435:646-651). Such nucleases are created by fusing the cleavage domain of the type IIs restriction enzyme Fok I to engineered zinc finger proteins. One limitation of this architecture is that the natural Fok I cleavage domain must dimerize in order to cleave DNA; therefore two different zinc finger protein-Fok I fusions are required to introduce a double-stranded break at an asymmetric target site. As a consequence, use of this technology for genome editing may expose the cell to three different nuclease species |[ndash]| the desired heterodimer as well as two homodimers which are irrelevant to the desired process but which may still contribute to off-target cleavage events.

We have shown previously that human genome editing can be performed at high efficiency using designed zinc finger nucleases (ZFNs). ZFNs can be engineered to generate a double-strand break at a specific chromosomal location both in transformed and primary human cells. These breaks are repaired by the cell's own DNA repair machinery, and when a homologous extrachromosomal donor molecule is provided, a homology-directed repair process leads to highly efficient transfer of single-base-pair changes into the chromosome (Urnov et al. Nature 435: 646). ZFNs can be engineered to target virtually any chromosomal location (Pabo et al., Ann. Rev. Biochem. 70: 313), and function robustly in human hematopoietic stem cells, hence such localized gene modification may potentially be useful in the correction and treatment of certain monogenic diseases. Here we show that ZFNs can also be used for the preciseinsertion of a novel sequence into a pre-determined location in the human genome. Remarkably, this novel sequence can constitute a short patch sequence or several kilobases of one or more transgenes. We have observed ZFN-driven targeted integration into endogenous chromosomal loci in human cells of entire open reading frames and promoter-transcription units at a frequency of 5% in the absence of any selection. We also show the use of this process to disrupt with high efficiency an endogenous chromosomal locus with a selectable marker ORF. Finally, we describe our work on using ZFN-driven integration to insert a therapeutically-relevant transgene into a |[ldquo]|safe-harbor|[rdquo]| locus in the human genome, potentially avoiding the problem of insertional mutagenesis. Hence, the homology-directed repair process invoked by the ZFNs can be used to carry out high-efficiency targeted integration to potentially improve the safety and efficacy of gene therapy.

HIV entry into target cells requires binding of the envelope protein GP120 to the CD4 co-receptors CCR5 or CXCR4. Loss of these HIV co-receptors is therefore attractive as a potential therapy for HIV patients, yet no current methods permit the efficient therapeutic disruption of a chosen gene in the human genome.

Wiskott-Aldrich Syndrome (WAS) is an X-linked, recessive immune disorder caused by hypomorphic mutations in the gene encoding Wiskott-Aldrich Syndrome Protein (WASP), a major cytoskeletal regulator expressed in hematopoietic cells. WAS is therefore a good candidate for ex vivo gene correction therapy as the restoration of a single wild-type allele in a sub-population of patient cells is expected to provide therapeutic benefit (Charrier S et al., Gene Therapy Dec 2004).While gene replacement strategies seek to provide a new artificial copy of the defective gene, the gene correction approach aims to |[ldquo]|repair|[rdquo]| the clinically relevant mutations in the endogenous copy of the WASP gene, and thus restore normal gene function. This strategy eliminates the need for integrated transgenes, maintains the normal regulation of the corrected gene, and yet results in a permanent effect on cell phenotype.

Top of pageAbstract We have developed an efficient strategy for introducing precise sequence changes into the human genome, and have initiated studies to apply our methods to the treatment of human disease. Our approach relies on our capacity to design Zinc Finger Protein Nucleases (ZFNs) to bind and cleave virtually any chosen location in the genome. In the first step, a double-stranded break is introduced within the targeted locus using a pair of designed ZFNs that function as a heterodimer. Next, a homologous donor DNA molecule is provided as a recombination substrate to direct precise alterations in the sequence flanking the break site. Recently it was demonstrated that this method can be used to modify an endogenous gene sequence with very high efficiency (>15% of all target sequences)1. We have now begun developing this approach for the treatment of monogenic diseases. In the present application, ZFNs have been developed to target the |[beta]|-globin gene sequence at the precise location of the mutation causing sickle cell anemia. Development proceeded in several stages. First, a panel of candidate ZFNs was designed and tested for the ability to bind DNA in vitro. The ZFN pair with the highest affinity was then shown to efficiently cleave a DNA fragment corresponding to the |[beta]|-globin locus at the chosen target site. Importantly, the lead ZFN dimer also discriminated well against the corresponding sequence from the highly sequence-related globin paralogs. Next, the binding properties of the lead pair were further improved via site directed mutagenesis and the expansion of the DNA-binding domain of one of the ZFNs from four to six fingers, increasing the DNA recognition sequence from 12 to 18 bp. Finally, the resultant lead pair was tested in cells to determine its ability to cleave the globin target sequence in a chromosomal context. In this assay, the target sequence is embedded within a defective GFP reporter gene that is integrated into the genome of 293 cells. Cleavage of the target and subsequent repair of the GFP gene can be monitored by recovery of green fluorescence, and we observed efficient GFP conversion in this system. These experiments suggest that the current lead pair may be suitable for development of a therapeutic gene correction approach to sickle cell anemia, and we have initiated studies to test the ability of these ZFPs to mediate gene correction at the endogenous |[beta]|-globin locus.

Human monogenic disorders, such as X-linked SCID, cystic fibrosis, sickle cell anemia, and hemophilia, are caused by hypomorphic or null mutations in a single gene, and restoration of even a single wild-type allele in a sub-population of cells can be curative. In practice, however, resistance to homologous recombination (HR) is a universal feature of mammalian genomes. An important exception to this rule occurs during repair of double-strand breaks (DSBs), which are healed by "gene conversion" – a unidirectional transfer of genetic information missing at the break from a homologous sequence such as a sister chromatid. Extensive analysis with reporter genes engineered to contain recognition sites for rare endonucleases validated the notion that a DSB potentiates homologous recombination by several orders of magnitude. The concept of targeted DSB-driven genome editing can be expanded from reporter genes to any endogenous locus via the use of zinc finger protein-based chimeric endonucleases (ZFNs): enzymes whose DNA cleavage specificity is determined by a designed zinc finger protein, engineered to bind specifically to a sequence at the mutated locus. We present data on the use of such ZFNs to introduce defined nucleotide changes at several endogenous loci in the human genome, including the IL2Rg and b-globin loci. Use of such "gene correction" in the context of the clinic will require high-frequency targeting. We present data on optimization of the targeting protocol that reproducibly yields > 10' gene-targeted cells in a given population in the absence of any selective pressure. This approach to gene correction could be used to treat monogenic disorders such as X-linked SCID and Sickle Cell Anemia.

Zinc finger protein transcription factors (ZFP TFs) can be engineered to regulate expression of any endogenous target gene, and thus represent a novel approach to the treatment of disease. In the area of diabetic neuropathy, for example, where no current treatment options are available, ZFP TF technology offers a unique solution to an unmet medical need.