330. Using CRISPR/Cas9 to Edit Hematopoietic Stem and Progenitor Cells

Abstract

Engineered nucleases such as zinc finger nucleases (ZFNs) and CRISPR/Cas9 have the potential to improve the precision of gene therapies based on hematopoietic stem and progenitor cells (HSPC). ZFNs were the first of this class of reagents to be approved for the clinic, through the disruption of the CCR5 gene as an anti-HIV therapy. Since ZFNs only need to be expressed transiently to permanently disrupt a gene, ZFN mRNA electroporation has proved to be an effective method to deliver CCR5 ZFNs to HSCPs. Optimizing the delivery of CRISPR/Cas9 to HSPC requires, in addition, consideration of how to deliver multiple components - the Cas9 nuclease and one or more single guide RNAs (sgRNAs) - in a transient but coincidental window of expression. This is further complicated if a DNA homology template is also to be delivered, in order to direct site-specific gene editing or insertion. Some of the difficulties of working with HSPC include their sensitivity to DNA or viral vectors that can result in unacceptable levels of cytotoxicity, the possibility for only limited in vitro culturing, and the complexity of the studies needed to evaluate any impact of the treatment on HSPC biology. In addition, the relatively large size (4.1kb) of the standard S. pyogenes Cas9 (spCas9) makes it a challenge for certain delivery systems, although a smaller variant from S. aureus (saCas9, 3.2 kb) is also available. In this work, we evaluated several approaches for the delivery of both Cas9 and sgRNAs to HSPC, including AAV vectors, in vitro transcribed Cas9 mRNA, and synthesized sgRNAs. The use of the smaller saCas9 allowed us to evaluate AAV vectors for Cas9 delivery, as well as in combination with sgRNA, using AAV serotype 6 that we identified as having good tropism for human HSPC. Various combinations of the different platforms, and time between deliveries, were optimized to maximize nuclease activity without overt toxicity. Using in vitro synthesized Cas9 mRNA and sgRNAs, we found that a single co-electroporation step was preferable to reduce the toxicity of sequential electroporations, and that chemical modification of the sgRNA was necessary to stabilize it and allow its function when co-delivered in this way. For AAV vectors, although the smaller size of saCas9 allowed it to be co-packaged together with a sgRNA in a single AAV6 vector that worked in cell lines, this combination did not effectively function in HSC. However, the sgRNA component alone was effectively delivered by AAV6 vectors, and could be combined with a later delivery of saCas9 mRNA by electroporation. With this latter approach, we were able to disrupt the CCR5 locus at similar levels to those we achieve using the clinically optimized protocol for CCR5 ZFNs.