Integrated design, execution, and analysis of arrayed and pooled CRISPR genome-editing experiments

Abstract

CRISPR (clustered regularly interspaced short palindromic repeats) genome-editing experiments offer enormous potential for the evaluation of genomic loci using arrayed single guide RNAs (sgRNAs) or pooled sgRNA libraries. Numerous computational tools are available to help design sgRNAs with optimal on-target efficiency and minimal off-target potential. In addition, computational tools have been developed to analyze deep-sequencing data resulting from genome-editing experiments. However, these tools are typically developed in isolation and oftentimes are not readily translatable into laboratory-based experiments. Here, we present a protocol that describes in detail both the computational and benchtop implementation of an arrayed and/or pooled CRISPR genome-editing experiment. This protocol provides instructions for sgRNA design with CRISPOR (computational tool for the design, evaluation, and cloning of sgRNA sequences), experimental implementation, and analysis of the resulting high-throughput sequencing data with CRISPResso (computational tool for analysis of genome-editing outcomes from deep-sequencing data). This protocol allows for design and execution of arrayed and pooled CRISPR experiments in 4-5 weeks by non-experts, as well as computational data analysis that can be performed in 1-2 d by both computational and noncomputational biologists alike using web-based and/or command-line versions.

Figure 1

Schematic of an arrayed genome-editing experiment. Arrayed genome-editing experiments are performed by designing one sgRNA using CRISPOR. This schematic demonstrates the design of an sgRNA to mutagenize a GATA motif. After designing the optimal sgRNA, it is cloned into pLentiGuide-puro, lentivirus is produced, cells are transduced, and successful transductants are selected (successful transduction is indicated by red curved lines). After conclusion of the experiment, cells are pelleted, and genomic DNA is extracted. Locus-specific PCR primers are used to amplify regions flanking the double-strand break site. Deep sequencing of the amplicon generated by locus-specific PCR is subsequently performed. Quantification of editing frequency and indel distribution is determined by CRISPResso. The GATA motif is underlined, the triangle indicates the double-strand break position, and the PAM sequence is shown in blue.

Figure 2

Schematic of a pooled genome-editing experiment. Pooled genome-editing experiments are performed by designing multiple sgRNA using CRISPOR. After designing the sgRNAs, they are batch-cloned into pLentiGuide-puro, lentivirus is produced, cells are transduced at low multiplicity, and successful transductants are selected (successful transduction is indicated by curved lines). Phenotypic selection is performed (e.g., FACS, drug/toxin resistance, drug/toxin sensitivity, cell lethality/gene essentiality, cellular fitness/proliferation). After conclusion of the experiment, cells are pelleted, and genomic DNA is extracted. PCR primers specific (primer sequence is underlined) to the pLentiGuide-Puro construct are used to amplify regions flanking the cloned sgRNA sequence. Deep sequencing of the amplicon generated by construct-specific PCR is subsequently performed. sgRNAs present within the sample are enumerated by CRISPRessoCount.

Figure 3

Design, experimental execution, and data analysis workflows for arrayed and pooled genome-editing experiments. sgRNA design steps by CRISPOR are shown in gray, experimental execution steps are shown in blue, and data analysis steps by CRISPResso are shown in red/pink.

Figure 4

Locus-specific deep-sequencing analysis of coding and noncoding targeting by CRISPResso. (a) Frequency distribution of alleles with indels (shown in blue) and without indels (shown in pink) for an sgRNA targeting BCL11A exon 2. (b) All reads with sequence modifications (insertions, deletions, and substitutions) are mapped to a position within the BCL11A exon 2 reference amplicon. The vertical dashed line indicates the position of predicted Cas9 cleavage. The position of the sgRNA is shown in gray. (c) Distribution of indel sizes when targeting BCL11A exon 2. Percentage of unmodified sequences is shown in red and percentages of modified sequences are shown in blue. (d) Frameshift analysis of BCL11A exon 2 coding sequence targeted reads. Frameshift mutations are shown in red and in-frame mutations are shown in tan. (e) sgRNA enrichment based on analysis of fetal hemoglobin (HbF) levels when performing saturating mutagenesis of BCL11A exon 2 and analysis of the functional core of the BCL11A enhancer using NGG- and NGA-restricted sgRNAs from two previously published studies11,12. Nontargeting sgRNAs are pseudomapped with 5-bp spacing.