Stem Cell Reports Ar ticle High-Content Analysis of CRISPR-Cas9 Gene-Edited Human Embryonic Stem Cells Jared Carlson-Stevermer, 1,2 Madelyn Goedland, 1,2 Benjamin Steyer, 1 Arezoo Movaghar, 1,2 Meng Lou, 1 Lucille Kohlenberg, 1 Ryan Prestil, 1 and Krishanu Saha 1,2,3, * 1 Wisconsin Institute for Discovery, University of Wisconsin-Madison, Madison, WI 53715, USA 2 Department of Biomedical Engineering, University of Wisconsin-Madison, Madison, WI 53715, USA 3 Department of Medical History and Bioethics, University of Wisconsin-Madison, Madison, WI 53715, USA *Correspondence: ksaha@wisc.edu http://dx.doi.org/10.1016/j.stemcr.2015.11.014 This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). SUMMARY CRISPR-Cas9 gene editing of human cells and tissues holds much promise to advance medicine and biology, but standard editing methods require weeks to months of reagent preparation and selection where much or all of the initial edited samples are destroyed dur- ing analysis. ArrayEdit, a simple approach utilizing surface-modified multiwell plates containing one-pot transcribed single-guide RNAs, separates thousands of edited cell populations for automated, live, high-content imaging and analysis. The approach lowers the time and cost of gene editing and produces edited human embryonic stem cells at high efficiencies. Edited genes can be expressed in both plurip- otent stem cells and differentiated cells. This preclinical platform adds important capabilities to observe editing and selection in situ within complex structures generated by human cells, ultimately enabling optical and other molecular perturbations in the editing work- flow that could refine the specificity and versatility of gene editing. INTRODUCTION CRISPR-Cas9, an emerging genome surgery tool, exploits an engineered ribonucleoprotein complex consisting of two essential components: (1) a protein, Cas9; and (2) a sin- gle-guide RNA (sgRNA). Together, the Cas9-sgRNA complex cuts a specific target sequence in the genome. Human cells and tissues edited by CRISPR-Cas9 are important resources for drug target identification (Kasap et al., 2014; Shi et al., 2015; Smurnyy et al., 2014), regulatory science (Hsu et al., 2014), medicine (Doudna, 2015), and basic biology (Hsu et al., 2014; Sternberg and Doudna, 2015). However, human gene-editing experiments frequently require labo- rious cloning of expression plasmids for each sgRNA, and there are limited opportunities in these culture systems to watch and perturb genome surgery in action, as it is diffi- cult to isolate and image living mutant cells during and shortly after the DNA cleavage event. Overall, there is a need to expand the throughput and capabilities of current in vitro human culture systems where novel genome sur- gery approaches can be evaluated with human cells and tis- sues (Baltimore et al., 2015). Advanced capabilities with human pluripotent stem cells in particular could eventu- ally expand the suite of human preclinical model systems, ranging from patient-specific cell lines to complex human embryonic tissues established from stem cells. Current gene-editing techniques generate heterogeneous human cell populations that require significant subsequent characterization. It is crucial to analyze the genome of the edited cells by sequencing before continuing with other studies, and several protocols require destruction of mutant cell populations during sequencing analysis (Ding et al., 2013; Kasap et al., 2014; Mali et al., 2013; Miyaoka et al., 2014; Shi et al., 2015; Smurnyy et al., 2014; Yang et al., 2013). For example, targeted gene disrup- tion followed by selection and next-generation sequencing can identify drug targets, but a separate, subsequent gene- editing experiment is required to obtain living mutant cells for downstream analysis (Kasap et al., 2014; Sanjana et al., 2014; Shalem et al., 2014; Shi et al., 2015; Smurnyy et al., 2014; Wang et al., 2014), a process that is often infeasible for slowly dividing or primary cells. This slows epigenomic and functional characterization of properly edited cells, and it is currently unknown whether there are persistent epigenomic and functional problems within the edited cells (Bosley et al., 2015). Further sequence-level character- ization is also required at the single clone level, as there is frequent and variable disruption of, or insertion of donor DNA into, the non-targeted allele in edited cell lines (Mer- kle et al., 2015). Finally, efficiencies of isolating precisely edited cells remain a challenge with current methods, typi- cally with 20% or lower efficiencies to make near-precise deletions in the human genome (Byrne et al., 2015). Here, we describe a platform, termed ArrayEdit, that com- bines two capabilities: one-pot transcription, and the combi- nation of microcontact printed plates and high content analysis (HCA). First, we describe a method that can generate many sgRNAs in parallel, within hours, using chemically synthesized oligonucleotides ordered in a multi- well format. One-pot transcribed sgRNAs can be delivered without purification and can efficiently generate desired gene edits within human embryonic stem cells (hESCs) Stem Cell Reports j Vol. 6 j 109–120 j January 12, 2016 j ª2016 The Authors 109