Genome Editing Via NHEJ-Mediated Targeted Integration Retains Engraftment Potential of Edited Hematopoietic Stem Cells

The ability to edit the genome of long-term-engrafting hematopoietic stem cells (LT-HSCs) would provide a curative therapy for a wide range of hematological disorders. While mixed LT/ST-HSC (ST, short-term) pools edited via non-homologous end joining (NHEJ) yield engraftable transgenic cells, pools made transgenic via homology-dependent repair (HDR) do not. This observation is likely explained by the absence of DNA replication required for HDR in LT-HSCs. Quiescence renders LT-HSCs refractory to transgene insertion via HDR, with the transgene-positive population consisting overwhelmingly of actively dividing, short-term-engrafting HSCs. In contrast to HDR, NHEJ is a DNA repair mechanism active in non-dividing cells. We hypothesized that CRISPR/Cas9-mediated transgenesis of HSC pools via NHEJ-mediated targeted integration (NHEJ-TI) would allow modification of LT-HSCs and produce edited, long-term engrafting progeny.

First, we investigated whether NHEJ-TI is a feasible method of genome editing in HSCs. Utilizing CRISPR/Cas9 to induce a DSB in the AAVS1 (PPP1R12C) locus and a self-complementary rAAV6 (scAAV6) to deliver a GFP expression cassette dependent on the endogenous promoter, we optimized RNP nucleofection, scAAV6 infection, and HSC culture conditions for efficient NHEJ-TI in HSCs. Flow cytometry for eGFP and in-out PCR was used to confirm transgene integration at frequencies ranging from 4-6%.

Next, we investigated whether the NHEJ-TI edited pool of mixed LT/ST-HSCs contained engraftable, edited cells by injection into irradiated NOD/SCID/IL2Rg^-/- (NSG) mice. We compared the engraftment potential of transgenic HSCs cultured for either 2 days (2D) or 2 hours (2H) prior to editing. While the overall editing efficiency was ~20% higher in the 2D condition, the near absence of cell prestimulation in the 2H condition resulted in increased levels of overall engraftment, doubling the total engraftment level of edited hematopoietic cells at 10 weeks post-injection. Importantly, the percentage of edited cells among the engrafted human hematopoietic cell population remained essentially unchanged relative to the starting efficiency (2D: 4.0%, 2H: 3.6%; 10 weeks post-injection). Analysis of bone marrow-resident CD34-positive cells at 16 weeks post-engraftment will reveal the fraction of edited LT-HSCs. Our results show that NHEJ-TI is a feasible method of inserting transgenes into HSCs whilst retaining their engraftment potential. To our knowledge, this is the first study to achieve successful NHEJ-mediated targeted integration in HSCs while retaining engraftment potential and (perhaps) transgenesis as well.