Poster ASGCT2020 MDG.pdf

The liver is an organ of particular interest for gene therapy, both for correcting inherited metabolic disorders, as well as a biofactory for secretion of therapeutic proteins. The CRISPR/Cas9 genome editing system has greatly improved specificity and precision of gene targeting and can be efficiently delivered to the liver with adeno associated viral vectors (AAV). The goal of this project is to identify novel safe harbor sites in the liver that could drive high expression of therapeutic proteins. We performed ChIP-Seq for histone H3K27 acetylation and RNA Pol II binding and ATAC-Seq, and identified Apolipoprotein a1 (Apoa1) as one of the most highly expressed and accessible loci in mouse and human liver. To target the Apoa1 locus, we injected adult or postnatal 4 (P4) C57BL/6J mice with a dual AAV8 system, consisting of the CRISPR/Staphylococcus aureus Cas9 machinery (AAV-CRISPR) and a donor cassette (AAV-Donor), which includes the final coding exon of Apoa1 (Exon 4) fused to a promoterless marker gene (mKate2) by a 2A self-cleaving peptide coding sequence, flanked by homology arms to the Apoa1 locus. By deep sequencing, we observed efficient on-target nuclease activity at the Apoa1 locus (indels frequency: 51.5 ± 3.2%) and no off-target activity in any of the predicted sites. We detected two major integration events at the Apoa1 locus: the expected homology directed repair (HDR)-mediated integration of the donor cassette and the non-homologous end joining (NHEJ)-insertion of the whole AAV genome. By droplet digital PCR, we observed 11.8 ± 2.5 and 1.1 ± 0.6 % of NHEJ- and HDR-insertion in adult injected-mice, which resulted in 5.7 ± 2.9 % of mKate2-positive hepatocytes. In mice injected at P4, the NHEJ- and HDR-insertion frequency was respectively 21.5 ± 8 and 9.5 ± 3.3 % and the percentage of mKate2-positive hepatocytes increased up to 16.4 ± 3.6 %. The genome editing of the Apoa1 locus did not adversely affect the endogenous level of apoA1 as well as neighboring genes, and no liver toxicity or histopathological abnormalities were observed. To test whether the Apoa1 locus could support the production of a secreted therapeutic protein, we designed an AAV-Donor carrying the coding sequence of human Factor IX (FIX) fused to the 2A sequence. We observed the efficient and sustained expression of FIX from the Apoa1 locus over a period of six months post-injection (0.23 ± 0.06 µg/ml at 24 weeks). Finally, we sought to determine whether a targeted gene therapy at the Apoa1 locus could correct an inherited metabolic liver disorder. We injected a mouse model of Hereditary tyrosinemia type I, fumarylacetoacetate hydrolase deficient (Fah-/-) mice, with AAV-CRISPR plus an AAV-Donor carrying human FAH. The targeted expression of FAH by the Apoa1 locus resulted in rescue of lethality, repopulation of the liver by FAH positive cells, recovery of normal histology and liver function. In addition, the targeted expression of human apolipoprotein E (APOE) by the Apoa1 locus resulted in APOE restoration and amelioration of plasma cholesterol levels in a mouse model of hypercholesterolemia (Apoe-/- mice). Overall, these results show Apoa1 as novel safe harbor site for targeted gene therapy in the liver.