The clustered regularly interspaced short palindromic repeats (CRISPR) and the CRISPR-associated proteins 9 (Cas9) systems are powerful tools for gene editing. Ribonucleoprotein (RNP) complex composed of Cas9 protein and sgRNA binds to specific genomic loci and generate DNA double strand breaks. While plasmids expressing Cas9 protein and sgRNA are routinely transfected into various cell lines to perform gene editing (Cong et al., 2013; Mali et al., 2013; Ran et al., 2013), direct delivery of Cas9-sgRNA RNP has shown higher efficiency and lower off-target effects (Kim et al., 2014), especially in human primary cells such as T cells (Hendel et al., 2015; Schumann et al., 2015). SgRNA can be generated by either in vitro transcription (IVT) or chemical synthesis. IVT is widely used to generate sgRNAs, since it can be easily performed in most labs. Gene editing of human primary cells, such as human CD34+ hematopoietic stem and progenitor cells (HSPCs), T cells and chimeric antigen receptor T (CAR-T) cells, is important for studying gene functions in these cell types, and holds great promise to further improve cell therapy (Schumann et al., 2015; Dever et al., 2016; Liu et al., 2016; Ren et al., 2017; Zhang et al., 2017). Our previous study showed that multiplex gene editing using CRISPR-Cas9 RNP hampered the proliferation of CAR-T cells (Liu et al., 2016). It was also reported that Cas9/hCD45sg1 RNP-treated human HSPCs had lower cell number compared to Cas9 proteintreated control (Gundry et al., 2016). In this study, we investigated the mechanism of cell loss after electroporation of Cas9-sgRNA RNP and find an easy method to resolve it. As our initial attempt of gene editing in human CD34+ HSPCs, we delivered Cas9-sgRNA RNPs into HSPCs by electroporation, using sgRNA in vitro transcribed by T7 polymerase. We observed significantly decreased cell number and reduced CD34 expression in survived cells 48 h after electroporation in experiments using five different sgRNA (Fig. S1A and S1B). Accordingly, the colony forming ability of HSPCs after RNP electroporation was markedly compromised (Fig. S1C). To identify the factor that contributed to this effect, we delivered IVT sgRNA (sgRNA-IVT), RNP complex consisting of Cas9 protein and IVT sgRNA (RNP-IVT), or Cas9 protein alone individually into primary

HSPCs by electroporation. We observed lower cell viability, reduced CD34 expression and decreased colony forming ability in sgRNA-IVT and RNP-IVT groups, while the cells in Cas9 protein group and mock electroporation group survived well (Fig. 1 A–C), indicating that sgRNA-IVT reduced HSPC stemness in addition to causing cell death. To validate whether sgRNA could also cause the death of T cells, we electroporated different amounts of IVT sgRNA into human primary CD3+ T cells. Reduced cell viability and increased immune stimulation were observed when higher amounts of sgRNAs were used (Fig. S2A). All sgRNAs we tested had a negative effect on the survival of HSPCs and T cells, suggesting that some common feature of these IVT sgRNA was responsible for this effect. Based on existing reports, we hypothesized that the 5’triphosphate of IVT sgRNA induced type I IFN production, leading to cell death in HSPCs and T cells. To test this hypothesis, we electroporated sgRNA-IVT targeting different genomic loci into CD34+ HSPCs and CD3+ T cells respectively, and measured the concentration of type I IFN in the culture medium using enzyme-linked immunosorbent assay (ELISA). Indeed, we detected significant release of IFN I in all samples electroporated with RNP-IVT (Figs. 1 D and S2B).