David Knapp - Genetic engineering in primary human hematopoietic stem cells

Precise editing of human hematopoietic stem and progenitor cells (HSPC) is of interest both for the treatment of genetic disorders and for disease modeling. The advent of CRISPR/Cas has made such editing possible, however, current strategies have remained limited to 10-20% efficiency. Uses requiring purity thus need to include selection markers. This also limits donor selection to adeno-associated virus (AAV), as single stranded oligonucleotide donors (ssODNs) are not long enough to incorporate them. To address these issues, we optimized a homology-directed repair (HDR)-based CRISPR/Cas approach to achieve near-unit precise editing efficiency directly in primary human HSPC.

Through a combination of optimal growth conditions, delivery timing, RNP selection, donor design, and the addition of small molecule DNA repair modulators, we can now consistently achieve >90% precision edits in human HSPC with minimal toxicity. This functions both for AAV, as well as ssODN donors, thus facilitating the rapid prototyping of donors. Importantly we have shown that this editing is consistent across the hematopoietic hierarchy, from the stem to the committed progenitor compartments. We also demonstrate that the editing process does not impact the proliferation and differentiation potential of the HSPCs. Applying our system in a disease modelling context, we introduced the leukemia-associated mutation SRSF2 P95H. This mimicked the disease phenotypes observed in patients including an erythroid differentiation deficit, and a competitive advantage over wild-type cells. Moreover, it gave important insight into the cellular mechanism by which the mutation acts, reconciling mouse and over-expression models which had failed to fully recapitulate competitive benefits observed in patients.

We developed a high efficiency system for the introduction of precise edits into primary human HSPC and have shown that it can give important insights in a disease modelling context. We anticipate that our system will enable disease modelling for many hematopoietic disorders and improve therapeutic editing strategies.

co-authors: Cloarec-Ung FM, Beaulieu J, Suthananthan A, Lehnertz B, Sauvageau G, Sheppard H