Riassunto analitico
Transplantation of autologous, genetically corrected epidermal stem cells (EpSC) is a potential treatment for a family of skin adhesion disorders called epidermolysis bullosa. Targeted transgene integration overcomes the issue of insertional mutagenesis associated with retroviral vectors, and may thus provide a safer gene transfer alternative. Recently, a new technology has been developed to increase the efficiency of targeted integration by introducing a double strand break (DSB) in a predetermined site in the genome. This system is based on the use of zinc-finger nucleases (ZFNs), which are artificial tailored nucleases composed of an array of zinc-finger DNA binding motifs fused to the catalytic domain of the endonuclease FokI. The introduced DSB can be repaired by homologous recombination (HR) using a donor DNA as a template. The donor DNA can either be the sister chromatid or an exogenous DNA molecule; thus if an expression cassette flanked by sequences homologous to the damaged site is provided to the cell, the repair machinery will insert the transgene specifically in the target site. In this work I developed a gene-targeting platform based on the use of ZFNs and integrase-defective lentiviral vectors (IDLVs) to insert a transgene by HR into the AAVS1 locus on chromosome 19. The AAVS1 locus has been proved to be a “safe harbour” for the integration of exogenous DNA sequences, because insertion of regulatory elements does not affect expression of the neighbouring genes and the organization of the chromatin grants high and persistent level of expression of the transgene. I evaluated the targeting efficiency in a keratinocyte cell line (HaCaT) by IDLV-mediated delivery of an AAVS1-specific ZFNs pair together with an HR construct driving the insertion of a GFP expression cassette into the site of cleavage. I achieved up to 25% of targeted insertion of single copies or concatamers of the GFP cassette into the AAVS1 locus, as analysed by PCR, Southern blotting and sequencing on individual HaCaT cell clones. The system was optimized by using an adenoviral vector to deliver the ZFNs combined with the IDLV carrying the donor construct. With this platform I was able to achieve up to 13% of site-specific gene addition in immortalized keratinocyte derived from a patient affected with a benign form of epidermolysis bullosa. I tested the toxicity and activity of the ZFNs in human primary keratinocytes derived from a healthy donor. I observed very low cytotoxicity and apoptosis induction, but robust activity of the nucleases resulting in up to 9% of disruption of the ZFNs-target site evaluated by Cel-1 assay and 454 Roche deep sequencing. Evidence of HR-mediated targeted integration was also obtained at a lower but significant frequency in human primary keratinocyte cultures, by transducing cells with the IDLV-donor and AdZFNs vectors. The strong discrepancy between the high cleavage activity of the ZFNs and the low efficiency of targeted gene addition suggests that HR constitutes the major limitation of ZFN-mediated targeted integration. To investigate if the system was able to achieve targeted gene addition in long-term repopulating keratinocyte stem cells, human skin equivalents derived from keratinocytes co-infected with IDLV donor and AdZFNs vectors were grafted onto immunodeficient (nu/nu) mice. GFP-positive spots were observed for at least 10 weeks in the grafted tissue, confirming that stable integration occurred in transplantable long-term repopulating cells. Molecular analysis confirmed the occurrence of targeted integration of the GFP expression cassette. Overall, these experiments provide proof of principle for the feasibility of ZFN-mediated transgene integration in a “safe harbour” genomic location in human keratinocyte stem cells, opening new possibilities for the treatment of genetic skin diseases.
|