Main Text br Acknowledgments The

Main Text
Acknowledgments The authors would like to thank all delegates for their input into the workshops that led to the contents of this white paper (Table S2).
Introduction The type II clustered regularly interspaced short palindromic repeat (CRISPR) systems and the associated Cas9 nucleases have evolved in archaea and bacteria for sequence-specific recognition of DNA targets via a single-stranded RNA intermediate (Jinek et al., 2012). In an engineered version of the CRISPR system, the Streptococcus pyogenes Cas9 nuclease is directed by guide RNAs (gRNAs) targeting 20 bp sequences adjacent to a 5′-NRG-3′ sequence motif, and the resultant cleavage has been used to edit the genome in several species (Cho et al., 2013; Cong et al., 2013; Hwang et al., 2013; Jinek et al., 2013; Mali et al., 2013b). A mutated nuclease-inactive Cas9 (dCas9) regulates gene niclosamide by physically blocking transcription or through fusion to a transactivator (VP64, Ω subunit of RNA polymerase) or repressor domain (KRAB, SID) (Bikard et al., 2013; Cheng et al., 2013; Farzadfard et al., 2013; Gilbert et al., 2013; Kearns et al., 2014; Konermann et al., 2013; Maeder et al., 2013; Mali et al., 2013a; Perez-Pinera et al., 2013; Qi et al., 2013). While transgene overexpression has been used to achieve cellular reprogramming (Davis et al., 1987; Takahashi and Yamanaka, 2006), reprogramming via direct activation of an endogenous gene has only been recently demonstrated through the use of transcription activator-like effectors (TALEs) (Gao et al., 2013). However, difficulty in designing and codelivering multiple TALE expression constructs precludes simple screening and multiplexed gene activation that is straightforward with the dCas9-VP64 system. In this study, we used dCas9-based transactivators combined with an efficient lentivirus-based gene delivery system to induce cellular reprogramming.
Results and Discussion We tested the efficacy of a VP64dCas9-BFPVP64 fusion protein to activate expression of the endogenous Myod1 gene locus for a sufficient duration and magnitude to ultimately induce the reprogramming of mouse embryonic fibroblasts (MEFs) to skeletal myocytes (SkMs) (Figure 1A). Although we and others have previously used the single C-terminal fusion of VP64 to dCas9 to activate gene expression (Cheng et al., 2013; Farzadfard et al., 2013; Gilbert et al., 2013; Kabadi et al., 2014; Maeder et al., 2013; Perez-Pinera et al., 2013), preliminary studies indicated that this approach did not lead to levels of expression sufficient for cell reprogramming. Therefore, we tested whether two VP64 domains flanking dCas9 (VP64dCas9-BFPVP64) would yield higher Myod1 gene expression levels (Figures 1B and 1C and Figure S1A available online). dCas9 was also fused to blue fluorescent protein (BFP) to monitor expression (Figures 1B, 1C, S1A, and S1B). Lentiviral VP64dCas9-BFPVP64 was placed under the transcriptional control of a doxycycline-inducible promoter (Figures 1C and S1B). To avoid steric hindrance that may prevent transcriptional complex recruitment, we included flexible glycine-serine linkers adjacent to the VP64 domains (Figure S1A). We also added a third nuclear localization signal (NLS), which improved the nuclear localization of VP64dCas9-BFPVP64 by ∼10-fold (Figures 1C–1F). Initially, we transfected C3H10T1/2 cells with the VP64dCas9-BFPVP64 plasmid. However, transfection was inefficient as transgene expression was detectable only in a few cells, presumably due to the large size of the plasmid (13.5 kb). We then used a lentiviral gene delivery system, allowing stable transduction of more cells as evidenced by both BFP and immunofluorescence staining of the FLAG epitope (Figure 1F). Approximately 50% of the transduced cells were found to be BFP positive as compared with 5% by transfection (Figure 1G). To ensure efficient gRNA codelivery, we developed a lentivirus-based U6 promoter-driven gRNA delivery system (Figure 1C). In separate experiments, we observed an ∼6-fold upregulation of Myod1 mRNA levels when all components were delivered by transfection (Figure S1C) compared with ∼60-fold by lentiviral transduction (Figure 1H).