Efficient, footprint-free human iPSC genome editing by consolidation of Cas9/CRISPR and piggyBac technologies

Abstract

Genome editing of human induced pluripotent stem cells (hiPSCs) offers unprecedented opportunities for in vitro disease modeling and personalized cell replacement therapy. The introduction of Cas9-directed genome editing has expanded adoption of this approach. However, marker-free genome editing using standard protocols remains inefficient, yielding desired targeted alleles at a rate of ∼1-5%. We developed a protocol based on a doxycycline-inducible Cas9 transgene carried on a piggyBac transposon to enable robust and highly efficient Cas9-directed genome editing, so that a parental line can be expeditiously engineered to harbor many separate mutations. Treatment with doxycycline and transfection with guide RNA (gRNA), donor DNA and piggyBac transposase resulted in efficient, targeted genome editing and concurrent scarless transgene excision. Using this approach, in 7 weeks it is possible to efficiently obtain genome-edited clones with minimal off-target mutagenesis and with indel mutation frequencies of 40-50% and homology-directed repair (HDR) frequencies of 10-20%.

Figure 1. Enhanced genome editing with Dox-inducible Cas9

a. Overview. A piggy-Bac transposon encapsulating the reverse tet activator, a tet-activator responsive promoter driving humanized Cas9, and a puromycin resistance cassette were integrated into the genome of wild-type human iPSCs. Treatment with Dox and co-transfection with gRNA and donor DNA oligonucleotide efficiently yields mutant iPSC clones. The transposon is efficiently removed by transfection with an excision-only piggyBac transposase, either as a separate step or concurrently with the gRNA and donor oligo. b. PCR genotyping of PGP1 cells confirming hCas9-PB stable integration.

Figure 2. Efficiency of genome editing with Dox-inducible Cas9

a. Surveyor mutation assay to detect genome modification. The indicated hiPSC lines were treated or not treated with Dox. Modification efficiency in genomic DNA at the TAZ locus was assessed using Surveyor nuclease followed by native gel separation of reaction products. Arrowheads indicate nuclease cleavage products. b. Deep sequencing analysis of the frequency of HDR or NHEJ genome modification at the TAZ locus. A PCR amplicon encompassing the TAZ gRNA target site was sequenced using a MiSeq Illumina sequencer at a minimum depth of 100,000 reads per amplicon. Amount of gRNA expression construct is shown in μg. c. After Dox-induced genome editing at the TAZ locus on the X chromosome of a male iPSC line, individual clones were picked and genotyped by Sanger sequencing. The pie chart displays the frequency of TAZ modification by HDR or NHEJ. d. Representative Sanger sequencing chromatograms, showing a clone that underwent HDR-mediated genomic modification (red arrow indicating one base HDR-programmed deletion) compared to a control.

Figure 3. Excision of Cas9-bearing transposon using piggyBac transposase

a. PGP1-hCas9-PB-TAZc.821delG cells were transfected with piggyBac expression vector. Puromycin resistant clones, the clones that failed to undergo transposon excision, were visualized by crystal violet staining. b. PCR genotyping of individual clones with or without transfection of piggyBac expression vector. Representative examples of genotyping results of positive and negative clones are shown. Pie chart summarizes the genotyping results of 34 clones.

Fig. 4. Analysis of Cas9 off-target activity

a. Off-target activity at 31 computationally predicted potential off-target sites for TAZ gRNA. Cas9-PB iPSCs were treated with DOX + TAZ gRNA + donor. Off-target activity at 31 sites was analyzed by PCR amplification of the candidate sites from pooled genomic DNA followed by deep sequencing (greater than 100,000 sequences per site). The candidate sites had 3 mismatches from the gRNA, with the exception of site 28, which had a single nucleotide polymorphism that created a 2 bp mismatch as we previously reported. “Cas9 no PB” and “Cas9 with PB” refer to omission or inclusion of piggyback transposase expression plasmid in the transfection of gRNA and DNA donor. “TAZ HDR” and “TAZ NHEJ” indicate the frequency of on target homology directed repair and non-homologous end joining at the TAZ locus. b. Whole genome sequencing on 6 independently isolated clones derived from PGP1-Cas9-PB (listed under WGS sample) after targeting at 3 loci (TAZ, DNAJC19, and JUP). HDR or NHEJ indicates the type of mutation found at the target site. In each whole genome sequence we identified 10–15 indels. These were analyzed for homology to the gRNA, presence of proto-spacer adjacent motf (PAM) sequences (blue) and recurrence in multiple clones or genomic locations. Based on this analysis, most of the indels were unlikely to be related to the gRNA sequences and may have arisen spontaneously during clonal expansion from a single cell. Those that were related to the gRNA sequence are listed as “off-target indels” and named with a number if it was among the 31 predicted potential off target sites (site 28, panel a) or a letter if it was not among these top predicted off target sites. Red letters indicate differences from the gRNA, the sequence of which is shown next to the locus name. The three separate TAZ clones sequenced all shared indels at the same two sites, Site 28 and Site A. Interestingly site A differs from the gRNA target by only one nucleotide, but was not computationally predicted because of its atypical PAM (CAG, pink underline). The two DNAJC19 clones had different indels at the same site, Site B. An identical genomic sequence on a different chromosome (Site C) also contained in indel in one of these clones. However, Sites B and C had neither a functional PAM (orange underline) nor close homology to the gRNA withing the Cas9 seed sequence. ^ indicates a site that is listed for completeness but may not have arisen from Cas9 activity (single occurrence; lack of PAM; multiple gRNA mismatches). CNV analysis of the whole genome sequencing data found no significant copy number variation in these clones.

Fig. 5. Design of genome editing reagents

A. gRNA recognition site. Red arrow indicates the gRNA cleavage site. Underline indicates the 12 nucleotide “seed” sequence that is generally considered most important for Cas9 targeting. The PAM (protospacer adjacent motif) sequence is required for Cas9 targeting. B. A gRNA is designed to introduce a double strand break (DSB, red lines) close to the desired modification site (blue). The HDR donor oligo is design to have homology arms that are at least 45 nt long on either side of the desired sequence modification (purple). Two additional sequence modifications are desirable if they can be achieved without impacting the ultimate experimental goal. First, introduce “silent” sequence variants that eliminate the gRNA recognition sequence after HDR (purple). Second, introduce or remove a restriction enzyme site so that HDR genomes can be identified by restriction digestion of PCR products (green). Genotyping PCR primers are designed flanking the targeted region. These genotyping primers should not overlap the homology arms.

Fig. 6. Surveyor nuclease assay

Examples of good, moderate, and poor gRNA-directed Cas9 cleavage followed by NHEJ, as determined by the surveyor assay. Arrows indicate surveyor nuclease cleavage products.