Dear Editor,

Genome editing of human embryonic stem cells (hESCs) is critical for basic biological research and regenerative medicine. However, until a few years ago, gene targeting technologies to disrupt, repair or overexpress genes in hESCs had been very inefficient and thus could not be routinely used. Recent technical breakthrough includes bacterial artificial chromosome based high efficiency gene targeting (Song et al., 2010), as well as the successful engineering of two systems of site-specific nucleases, zinc-finger nucleases (ZFNs) and transcription activator-like effector nucleases (TALENs) (Hockemeyer et al., 2009; Hockemeyer et al., 2011). The two engineered nucleases are composed of programmable and sequence-specific DNA-binding modules, which bring the nucleases to specific genomic site to introduce a DNA double-strand break. However, these two technologies have several limitations, including the time-consuming and labor-intensive experimental design, and the risk of off-targeting mutations (Gaj et al., 2013). More recently, a new genome-editing technology, denoted the clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated (Cas) system, has been developed for efficient gene targeting in cells of various species, including zebrafish (Hwang et al., 2013), mouse (Wang et al., 2013), monkey (Niu et al., 2014), and human (Cong et al., 2013; Mali et al., 2013b). In this system, Cas9 nuclease is targeted to a specific genomic site by complexing with a guide RNA, which hybridizes a 20-nucleotide DNA sequence immediately preceding an NGG motif, introducing a double-strand break three nucleotides upstream of the NGG motif (**ek et al., 2012). Compared to ZFNs and TALENs, CRISPR/Cas9 system offers simple experimental design and very high targeting efficiency (Ding et al., Full size image