Efficient Installation of Heterozygous Mutations in Human Pluripotent Stem Cells Using Prime Editing

Abstract

The utility of human pluripotent stem cells (hPSCs) is greatly enhanced by the ability to introduce precise, site-specific genetic modifications with minimal off-target effects. Although Cas9 endonuclease is an exceptionally efficient gene-editing tool, its propensity for generating biallelic modifications often limits its capacity for introducing heterozygous variants. Here, we use prime editing (PE) to install heterozygous edits in over 10 distinct genetic loci, achieving knock-in efficiencies of up to 40% without the need for subsequent purification or drug selection steps. Moreover, PE enables the precise introduction of heterozygous edits in paralogous genes that are otherwise extremely challenging to achieve using endonuclease-based editing approaches. We also show that PE can be successfully combined with reprogramming to derive heterozygous induced pluripotent stem cell clones directly from human fibroblasts and peripheral blood mononuclear cells. Our findings highlight the utility of PE for generating hPSCs with complex edits and represent a powerful platform for modeling disease-associated dominant mutations and gene-dosage effects in an entirely isogenic context.

Introduction

Human induced pluripotent stem cells (iPSCs) hold incredible promise for many exciting applications, including autologous cell-based therapies, disease modeling, and drug discovery. However, harnessing the full potential of iPSCs depends largely on an ability to genetically modify cells in a manner that is both highly precise and efficient. In the context of disease modeling, this may entail correcting mutations in patient-derived iPSCs or introducing specific genetic variants into an otherwise healthy iPSC line. Indeed, the introduction of specific genetic variants into a well-characterized human pluripotent stem cell (hPSC) clone enables its functional consequence to be studied in a completely isogenic context, thereby eliminating the substantial variability that can arise from genetic background differences. As an example, we recently described the introduction of a series of homozygous mutations in the NPHS2 locus in wildtype iPSCs to reveal a previously unappreciated reduction in variant protein (PODOCIN), variant-specific subcellular localization, as well as specific effects on association with another important glomerular protein, NEPHRIN.¹

While technologies such as CRISPR-Cas9 have revolutionized genome engineering workflows and offer the ability to edit the genome with high precision and efficacy,^2–4 achieving heterozygous modifications—where only one allele carries the desired edit—remains challenging. This can be attributed to the remarkable efficiency of CRISPR-Cas9 in generating DNA double-stranded breaks, which typically result in cells containing the intended (knock-in) or unintended (indel) edit within both alleles of a given target locus.

Here, we explore the use of prime editing (PE) to reliably and efficiently install heterozygous edits in iPSCs without impacting the second allele. The PE system is comprised of a Cas9-nickase fused to a reverse transcriptase (nCas9-RT) complexed to a programmable PE guide RNA (pegRNA).⁵ The pegRNA functions both to direct nCas9-RT and, once reverse transcribed, as a repair template at the target site. Compared with genome editing methods that rely on DNA double-stranded breaks, PE provides a significantly higher ratio of successful edits to unintended insertions or deletions (indels) and is also thought to be less dependent on endogenous repair mechanisms. PE has also been shown to exhibit minimal off-target effects,⁶ further strengthening its appeal in genome engineering applications.

While successful application of PE in hPSCs has been previously described,^7–10 we extend this work further to show that PE is an effective strategy for generating hPSCs with “clean” heterozygous edits in a broad range of genetic loci, including within genes that harbor identical sequences within closely related paralogues. We also demonstrate that PE can be combined with episomal reprogramming to derive iPSC clones with heterozygous edits directly from skin fibroblasts or patient blood samples. Collectively, these findings highlight the utility of PE as an effective strategy for generating iPSC clones with complex genetic modifications.

Methods

PE experimental design

For all PE experiments, protospacers as close as possible to the desired edit were prioritized. Each epegRNA construct encoded a 12 or 13 nt primer binding site (PBS) sequence, based on recommendations by David Lui’s group, who suggest starting with PBS length of 11–13 nt.⁵ The RT templates ranged from 18 to 22 nt in length and, where possible, extended at least 7 nt beyond the desired edit as recommended previously.¹¹ The only exception was the epegRNA used to introduce the SFTPA c.532G>A single nucleotide variant (SNV), which due to the location of a suitable protospacer adjacent motif (PAM), encoded a 30 nt RT template with only a 2 nt extension beyond the desired edit. In experiments performed with a secondary nicking sgRNA, protospacers ∼50 bp downstream and on the opposite strand of the pegRNA-guided nick were selected.

Vector construction

PegRNA plasmids were generated by ligating two sets of annealed oligo pairs encoding the pegRNA scaffold, evopreQ₁ motif, and locus/variant-specific Cas9-target (spacer), PBS, and RT sequences into the BsaI-digested pU6-peg-GG-acceptor plasmid (Addgene #132777; gift from D. Liu). Annealed oligo pairs were treated with T4 DNA kinase, followed by heat inactivation, prior to ligation. Nicking sgRNA plasmids were generated by ligating annealed oligo pairs into the BbsI-digested pSMART-sgRNA (Addgene #80427) plasmid. Transfection-grade plasmid DNA was prepared using the QIAGEN Plasmid Midi kit. See Supplementary Table S1 for the full list of oligonucleotides used to generate epegRNA and sgRNA constructs.

In vitro transcription

Capped and polyadenylated in vitro transcribed mRNA encoding PEmax and PE6b were generated from PmeI-digested pCMV-PEmax (Addgene #174820; gift from D. Liu) and pCMV-PE6b (Addgene #207852; gift from D. Liu) plasmid DNA, respectively. A gBlock (Integrated DNA Technologies) encoding a T7 promoter and MLH1dn coding sequence was used for in vitro transcription of mRNA encoding MLH1dn. In vitro transcription was performed using the mMESSAGE mMACHINE T7 ULTRA transcription kit (Thermo Fisher) according to the manufacturer’s recommendations. LiCl was used to precipitate mRNA prior to resuspension in nuclease-free H₂O.

Amplicon next-generation sequencing

Sequencing was performed using Illumina MiSeq technology, following the 16S metagenomic Sequencing Library Preparation protocol for amplicon library preparation and sequencing. Briefly, genomic DNA was extracted from bulk transfected cultures using the DNeasy Blood and Tissue Kit (QIAGEN) or from isolated iPSC colonies using Quick Extraction Buffer, and ∼200 bp amplicons were generated by polymerase chain reaction (PCR) using adapter primers that flank the target site (see Supplementary Table S3 for the full list of gene-specific primers). Exonuclease I (NEB) was then used to catalyze the removal of excess oligonucleotides, and amplicons were indexed with dual-barcode primers (Nextera XT DNA Indexes, Illumina). Indexed amplicons underwent fragment analysis and quality control using capillary electrophoresis (QIAxcel connect system, Qiagen) before samples were pooled in equimolar concentrations and cleaned up using Ampure XP magnetic beads (Beckman Coulter) at a 0.8 ratio. The pooled library was denatured and loaded onto the MiSeq sequencing platform with a final concentration of 6 pM + 15% PhiX spike-in, and 300 bp paired-end sequencing was performed (MiSeq Reagent Nano Kit v2 300 cycles, Illumina). All PCR amplifications were carried out using Q5 High-Fidelity DNA polymerase (NEB).

Sequencing reads were aligned to the GRCh38 reference genome using the “bwa mem” tool. Reads were reformatted with “samtools” to binary alignment map (BAM) files. A custom R script using the “CrispRVariants” package¹² was used to evaluate and plot gene editing outcomes. Briefly, using the “readsToTarget ()” function, reads aligned to the target site were extracted from the BAM files, producing a “CrisprSet” object. Sequencing variants near the target location were quantified and then visualized using the “plotVariants()” function, generating heat maps of variant frequencies across samples. Coverage at target SNV sites averaged at ∼8,000 reads and ranged from 5,000 to 11,000 reads. All next-generation sequencing (NGS) data related to bulk transfected cultures, isolated colonies, and subcloning analyses are available through Mendeley Data (DOI: 10.17632/yn7kpg48ym.2) in agreement with FAIR principles. https://data.mendeley.com/datasets/yn7kpg48ym/2.

PCR and Sanger sequencing

Allele-specific PCR was performed using GoTaq Green PCR Mastermix (Promega) with primers depicted in Supplementary Table S4. Prior to sequencing, PCR amplicons were treated with rAPid Alkaline Phosphatase (Roche Life Science) and Endonuclease I (NEB) for 30 min at 37°C, followed by heat inactivation at 80°C for 15 min. Sequencing reactions were performed using BigDye Terminator v3.1 (Thermo Fisher Scientific). Reactions were purified and sequenced by the Australian Genome Research Facility.

Cell culture, cryopreservation, and quality assurance

The iPSC lines used in this study are listed in Supplementary Table S4. Generation and characterization of PB010.5, PB005.1, and PB001.1 iPSC lines have been described previously.¹³ iPSCs were maintained and expanded at 37°C, 5% CO₂, and 20% O₂ in Essential 8 (Thermo Fisher Scientific) or mTesR+ media (Stem Cell Technologies) on hESC-qualified Matrigel (Corning)-coated plates with daily media changes. No antibiotics were used in the culture medium. iPSCs were passaged using 0.5 mM EDTA in 1X PBS every 3–4 days when cultures reached ∼80% confluency and/or when individual colonies were 1–2 mm in size, as previously described.¹⁴ For colony picking, single iPSC colonies (∼1–2 mm in size) were detached with a sterile 200 µL filter tip and transferred to one well of a 48-well Matrigel-coated plate. All SNV knock-in experiments described in this study were requested by users of the iPSC Gene Editing Core Facility (Murdoch Children’s Research Institute, Australia), and iPSC clones confirmed to harbor heterozygous knock-in of the requested variant were subjected to standard quality assurance procedures (molecular karyotyping, assessment of pluripotency markers, mycoplasma testing) after expansion and cryopreservation but prior to handover. Molecular karyotyping was performed by Victorian Clinical Genetics Services (Parkville, Australia), using Illumina Infinium CoreExome-24 v3.0 SNP arrays. Molecular karyotype reports for iPSC clones generated throughout this study are available through Mendeley Data (DOI: 10.17632/yn7kpg48ym.2) https://data.mendeley.com/datasets/yn7kpg48ym/2. SNPduo analyses were performed to confirm gene-edited clones were derived from the corresponding parental iPSC line. For assessment of pluripotency markers, iPSCs were harvested with TrypLE (Thermo Fisher Scientific) and incubated with antibodies to cell surface markers TRA-1-60 BV421 (BD Horizon), TRA-1-81 AF647 (Biolegend), and SSEA4 PeCy7 (Biolegend) as per manufacturer’s specifications. For intracellular OCT4 staining, cells were treated with Foxp3 fixation/permeabilization buffer (Thermo Fisher Scientific) and stained with OCT3/4 PE antibody (Thermo Fisher Scientific). Cells were analyzed by flow cytometry using an LSRFortessa x-20 flow cytometer (BD Biosciences). Unstained iPSCs were used as a gating control. Mycoplasma testing was performed by Cerberus Sciences (Scoresby, Australia). For cryopreservation, iPSCs were dissociated 2 days post-passaging (at ∼50% confluency) using 0.5 mM EDTA in 1x PBS and resuspended in Essential 8 medium. An equal volume of freezing medium (20% dimethyl sulfoxide in Essential 8 medium) was added dropwise, and the cell suspension was mixed by gentle agitation prior to aliquoting (1–1.5 mL) into cryotubes for controlled-rate freezing. Thawed cells were transferred to 15 mL Falcon tubes, and 5 mL Essential medium was added dropwise to the cell suspension prior to centrifugation (300 g for 3 min). The cell pellet was resuspended in 2 mL Essential 8 medium and plated at varying densities across two wells of a Matrigel-coated 24-well plate.

Subcloning

Subcloning was performed by dissociating iPSC cultures 1–2 days after EDTA passaging with TrypLE. Cells were plated at low density (100–1,000 cells per well) in 6-well Matrigel-coated plates in Essential 8 medium supplemented with CloneR2 (Stem Cell Technologies). The next day, colonies consisting of 2–3 cells were marked and allowed to expand for ∼1 week prior to picking, expansion, and subsequent NGS analysis.

RNP-mediated HDR

A homology-directed repair (HDR)-Donor oligo (Integrated DNA Technologies) encoding the TTN c.72663delA mutation, flanked by 67 nt and 44 nt homology arms, was used in conjunction with RNP for Cas9-based editing of the TTN locus. A TTN c.72657T>A synonymous change was also included in the donor oligo to act as a Cas9-blocking mutation and prevent re-cutting following successful HDR. The synonymous change is unlikely to affect functional assessment of the intended TTN c.72663delA truncating mutation, which leads to a frameshift and premature termination of the TTN transcript. RNP complexes (Integrated DNA Technologies) were formed by combining an equal volume of sgRNA (resuspended to a concentration of 44 µM) and Cas9 enzyme (diluted to 62 µM) and incubated at room temperature for 15 min. Transfections were performed as described below using 5 µL of RNP and 0.3 nm of the HDR oligo per transfection.

Transfection

iPSC transfections were performed using the Neon Transfection System (Thermo Fisher Scientific) as previously described.¹⁵ Briefly, cells were dissociated with TrypLE and resuspended in Buffer R at a final concentration of 1 × 10⁷ cells/mL. Transfections were performed in 100 µL tips using the following conditions: 1,100 V, 30 ms, one pulse. Following transfection, cells were transferred to Matrigel-coated plates containing mTesR+ medium supplemented with CloneR2, which was omitted in subsequent media changes. Simultaneous reprogramming and gene editing of human fibroblasts (GM21808, Coriell) were performed as previously described.¹⁶ Briefly, fibroblasts were dissociated with TrypLE 2 days after passaging and resuspended in Buffer R at a final concentration of 1 × 10⁷ cells/mL. Electroporation was performed in 100 µL tips using the Neon Transfection device using the following conditions: 1,400 V, 20 ms, two pulses. Cells were subsequently seeded on Matrigel-coated plates and maintained in fibroblast medium (Dulbecco's Modified Eagle Medium supplemented with 10% fetal bovine serum) until 3 days post-transfection, then switched to E7 medium (E8 medium without transforming growth factor β) supplemented with 100 µM sodium butyrate and changed every other day. Sodium butyrate was removed from the medium after the appearance of the first iPSC colonies at around day 10.

Results

Truncating mutations in TTN are one of the leading causes of dilated cardiomyopathy.¹⁷ We initially assessed the ability of PE to introduce a heterozygous TTN c.72663delA mutation in the wildtype iPSC line, PB010.5 (Supplementary Fig. S1A). A pegRNA comprising a TTN-specific spacer, 12 bp PBS, and 22 bp reverse transcription template (RTT) encoding the truncating TTN c.72663delA mutation and a nearby synonymous change (TTN c.72657T>A) (Supplementary Fig. S1B) was incorporated into a plasmid-based vector for expression from a U6 promoter. The evopreQ₁ motif was also included downstream of the RTT to enhance pegRNA stability, as previously described.¹⁸ The resulting epegRNA plasmid was introduced into iPSCs via electroporation using our previously optimized workflow,¹⁵ along with mRNA encoding either the PEmax¹⁹ or the more recently described PE6b²⁰ prime editors. We also evaluated the effect of a secondary downstream “nicking” sgRNA specific to the complementary strand, as described in the PE3 (and PE5) system,¹⁹ and/or expression of the engineered dominant-negative MLH1dn mutant, as described in the PE4 and PE5 systems.^11,19 Gene editing outcomes were evaluated in the transfected bulk cell population 2–3 days after electroporation (Fig. 1A) and in isolated colonies (Fig. 1B) by amplicon NGS. To attain independent iPSC clones harboring heterozygous knock-in of the TTN c.72663delA mutation, we selected colonies/wells with a TTN c.72663delA allele frequency >10% but <50% for subcloning. At least 10 subclones were then isolated and genotyped by NGS to attain purely heterozygous subclones harboring a 1:1 ratio of TTN c.72663delA to wildtype reads. An overview of the workflow used throughout this study is depicted in Supplementary Figure S2.

FIG. 1.

Evaluation of prime editing strategies in iPSCs. (A) Evaluation of knock-in efficiency, measured as TTN c.72663delA allele frequency, in bulk cultures following introduction of PE factors in healthy iPSCs. N = 3. Error bars indicate standard deviation (SD), N = number of biological replicates. (B) Genotypic analysis of individual colonies isolated from PE6b + MLH1dn transfection experiments performed with and without the nicking sgRNA. (C) CrispRVariants plot of three representative iPSC clones attained from the “PE6b + MLH1dn + nick” condition harboring either homozygous knock-in (HOM KI), clean heterozygous knock-in (HET KI), or heterozygous knock-in with a secondary indel mutation (KI/HET). Allele frequency is indicated by the heatmap. The TTN c.72663delA mutation (delA, red box) and accompanying TTN c.72657T>A synonymous change (SC, blue box) are shown. The PAMs are shown on the reference (REF) sequence, whereby PAM1 corresponds to the spacer encoded by the epegRNA and PAM2 corresponds to the spacer encoded by the complementary nicking sgRNA. iPSC, induced pluripotent stem cell; PE, prime editing.

Compared with PEmax, PE6b resulted in a consistent, albeit modest, improvement in editing efficiency, as detected by NGS analysis of bulk transfected cultures (Fig. 1A). While the inclusion of either MLH1dn mRNA or the nicking sgRNA alone enhanced editing efficiency, the most dramatic increase occurred when both were used together, resulting in ∼40% of total alleles harboring the TTN c.72663delA mutation (Fig. 1A; PE6b + MLH1dn + nick condition). We next performed NGS analysis on isolated colonies, with a focus on the PE6b + MLH1dn ± nick conditions (Fig. 1B). In the PE6b + MLH1dn condition, of 42 colonies screened we identified 2 (5%) heterozygous for the TTN c.72663delA mutation, which was confirmed following subcloning analysis (Supplementary Fig. S2B). Notably, of the 36 colonies screened from the PE6b + MLH1dn + nick condition, the TTN c.72663delA variant was detected in 25 (69%) colonies, of which 10 were likely heterozygous (≤50% TTN c.72663delA alleles), 13 homozygous (>50% TTN c.72663delA alleles), and 2 harbored heterozygous knock-in but also a deletion spanning the intervening sequence between the PE and nicking protospacers (Fig. 1C).

We also conducted a direct comparison with editing outcomes using a typical homology-directed repair (HDR) strategy. An RNP complex consisting of Cas9 nuclease and the same TTN-specific sgRNA sequence encoded by the epegRNA, along with a single-stranded oligonucleotide (ssODN) encoding the TTN c.72663delA mutation and TTN c.72657T>A synonymous change (Supplementary Fig. S1C), was introduced into MCRIi010-A iPSCs, with gene editing outcomes assessed as described above. NGS analysis of the bulk transfected population revealed knock-in efficiencies comparable with those attained for the best-performing PE condition, with ∼42% of total alleles carrying both the TTN c.72663delA mutation and synonymous change (Supplementary Fig. S3A). We also detected a high frequency of “partial knock-in” (∼11% of total alleles), where incorporation of the synonymous change occurred in the absence of the TTN c.72663delA mutation. Such partial knock-in events were far less common across all tested PE conditions, with the highest incidence (∼1% of total alleles) observed in the PE6b + MLH1dn + nick condition (Supplementary Fig. S3B, E). As expected, we also observed a much higher incidence of indel mutations (∼16% of total alleles) in the RNP condition (Supplementary Fig. S3D) compared with all PE conditions. Indels were barely detectable (<0.1% of total reads) in the absence of the nicking sgRNA and remained low (∼1% of total reads) in PE conditions that included the nicking sgRNA (Supplementary Fig. S3E). NGS analysis of individual colonies isolated from the RNP condition revealed a broad spectrum of genotypes, with complete homozygous knock-in of the TTN c.72663delA mutation and accompanying synonymous change being the most prevalent outcome (∼40% of total colonies screened). Reflective of the bulk analysis, indel mutations were also common, detected in more than a third (∼36%) of iPSC colonies screened. Notably, heterozygous knock-in of the TTN c.72663delA mutation (in the absence of any secondary indel mutations) was only detected in two colonies, although both were homozygous for the TTN c.72657T>A synonymous change, indicating knock-in events on both alleles (one being partial).

We further assessed the capacity of PE to introduce heterozygous SNVs in an additional six genetic loci (FND3CB, SFTPA2, MYBPC3, RTEL1, INF2, RAG1) across four additional iPSC lines (Table 1). We again performed a head-to-head comparison with PE6b and PEmax prime editors for installation of the TTN c.72663delA mutation in the female iPSC line (MCRIi005-A), as well as for two additional SNVs (FNDC3B c.2206G>A and SFTPA2 c.160C>T) in the commercially available ChiPSC18 line (Cellartis/Takara Bio). In agreement with our previous findings targeting the TTN locus in MCRIi010-A iPSCs, PE6b consistently resulted in a modest increase in editing efficiency compared with PEmax. We also evaluated the effect of a secondary nick in six separate editing experiments (MYBPC3 c.1928-2A>G, RTEL1 c.3791G>A, RTEL1 c.2920C>T, RAG1 c.108C>T, INF2 c.599T>G, and INF2 c.643A>C) and generally observed pronounced increases in editing efficiency in experiments that included a nicking sgRNA. Indeed, for two SNVs (MYBPC3 c.1928-2A>G and RTEL1 c.2920C>T), the addition of a nicking sgRNA was essential, with knock-in reads undetectable in bulk transfected cultures where the nicking sgRNA was omitted (Table 1). While we observed a broad variation in editing efficiency (ranging from 0.3 to 30.5% in bulk cultures), we successfully derived at least one, but typically multiple, independent iPSC clones harboring heterozygous incorporation of each of the SNVs tested (Fig. 2). Although several iPSC clones carrying indel mutations were also identified, these were predominantly observed in PE conditions that included a secondary nicking sgRNA and were typically associated with knock-in of the gene-specific SNV (Fig. 2).

FIG. 2.

Genotypic analysis of iPSC colonies from prime-editing experiments (marked by + in Table 1). Plasmids encoding gene-specific epegRNA and complementary nicking sgRNA (where indicated; + nick) were used in all transfection experiments along with mRNA encoding either PEmax or PE6b. MLH1dn mRNA was included in all transfections. Individual colonies were isolated and analyzed by amplicon NGS. Colonies harboring >50% of the corresponding SNV knock-in allele were classified as homozygous (HOM KI), whereas colonies harboring 10–50% of the SNV allele were classified as heterozygous (HET KI). Colonies classified as heterozygous knock-in were also confirmed following subclone analysis. iPSC, induced pluripotent stem cell; NGS, next-generation sequencing; SNV, single nucleotide variant.

Table 1.

Summary of PE Experiments Performed Across Four Different Healthy iPSC Lines, Targeting Seven Genetic Loci

SNV	Cell name	PEmax/PE6b	Nick	KI allele frequency (% total reads)
TTN c.72663delA	PB005 (hPSCreg:MCRIi005-A)	PEmax PE6b	No No	1.97 2.13
FNDC3B c.2206G>A	ChiPSC18 (Cellartis)	PEmax PE6b	No No	1.69 2.09^a
SFTPA2 c.160C>T	ChiPSC18 (Cellartis)	PEmax PE6b	No No	4.45 4.84^a
MYBPC3 c.1928-2A>G	PB010 (hPSCreg:MCRIi010-A)	PE6b PE6b	No Yes	0 0.30^a
RTEL1 c.3791G>A^b	ChiPSC18 (Cellartis)	PE6b PE6b	No Yes	3.37 ± 0.29^a 18.28 ± 10.67^a
RTEL1 c.2920C>T	ChiPSC18 (Cellartis)	PE6b PE6b	No Yes	0 8.35^a
RAG1 c.108C>T	PB001 (hPSCreg:MCRIi001-A)	PE6b PE6b	No Yes	1.57 13.07^a
INF2 c.599T>G	PB001 (hPSCreg:MCRIi001-A)	PE6b PE6b	No Yes	0.54 9.83
INF2 c.643A>C^b	PB001 (hPSCreg:MCRIi001-A)	PE6b PE6b	No Yes	12.74 ± 4.61^a 16.77 ± 7.71

Plasmids encoding gene-specific epegRNA and complementary nicking sgRNA (where indicated) were used in all transfection experiments along with mRNA encoding either PEmax or PE6b. MLH1dn mRNA was included in all transfections. Editing efficiency in bulk transfected cultures was measured by amplicon NGS.

Indicates experiments where individual colonies were isolated and genotyped by NGS, shown in Figure 2.

Indicates experiments performed in triplicate, where mean ± SD is shown. All other experiments were performed once.

PE, prime editing; iPSC, induced pluripotent stem cell; SNV, single nucleotide variant; NGS, next-generation sequencing; SD, standard deviation.

We also sought to evaluate the capacity of PE to generate more complex edits, specifically in loci containing identical sequences within closely related paralogues. The MYH6 and MYH7 genes, which encode the alpha and beta isoforms of cardiac myosin heavy chain, respectively, share a common evolutionary origin and have both been linked to various human cardiac pathologies.²¹ We attempted to introduce a heterozygous SNV in the MYH7 locus (c.2155C>T) without affecting corresponding sequences within the MYH6 locus (c.2161C>T). Located in tandem, the MYH6 and MYH7 genes likely emerged from a previous gene duplication event and share >91% sequence identity,²² rendering them less amenable to editing strategies that rely on Cas9 endonuclease. Given all potential sgRNAs within an 80 bp window of the MYH7 c.2155C>T SNV also recognize corresponding sequences in MYH6, Cas9 nuclease would likely induce DSBs within both genes, potentially leading to deletion of the intervening (∼30 kb) sequence. We designed two epegRNAs, which encode the C>T SNV as well as spacers predicted to bind MYH7 and MYH6 with equal efficiency (Fig. 3A). To reduce indel formation and minimize multiple editing events, we chose to omit a secondary nicking sgRNA. Plasmids encoding either epegRNA-1 or epegRNA-2 along with mRNA encoding PE6b and MLH1dn were introduced into MCRIi010-A iPSCs, and editing efficiency was measured in bulk transfected cultures by amplicon NGS, with MYH6 and MYH7 alleles distinguishable by two single nucleotide differences located upstream and downstream of the target site (Fig. 3A). Despite epegRNA-1 yielding a higher overall editing efficiency, we selected the epegRNA-2 condition for iPSC colony isolation due to a higher ratio of MYH7 to MYH6 KI events (Fig. 3B). Of the 44 colonies screened, 5 (11.4%) exhibited SNV knock-in exclusively within MYH7, 2 (4.5%) exclusively within MYH6, and 6 (13.6%) within both MYH6 and MYH7 (Fig. 3B). All remaining colonies were unedited (≥95% wildtype alleles), and no indel mutations were detected in any of the colonies screened. Next, three colonies, harboring 23–30% MYH7 c.2155C>T but <1% MYH6 c.2161C>T alleles, were selected for subcloning and subsequent NGS analysis. This enabled successful isolation of multiple iPSC clones from each parent colony, with clean heterozygous knock-in of the MYH7 c.2155 SNV (as determined by an equal ratio of MYH7 c.2155C>T to wildtype alleles). Importantly, no subclone showed >50% MYH7 c.2155C>T alleles, indicating homozygous knock-in had not occurred in any of the parent colonies selected for subcloning. SNV knock-in at the MYH7 locus, but not at MYH6, was further confirmed using gene-specific primers to separately amplify MYH6 and MYH7, followed by Sanger sequencing of the resulting amplicons (Fig. 3D).

FIG. 3.

Harnessing PE to install heterozygous edits in paralogous genes. (A) Schematic of the tandemly arranged MYH7 and MYH6 genes. The spacers encoded by the two epegRNA constructs, the nucleotide intended for editing (highlighted in red), and the two nucleotides used to distinguish MYH6 and MYH7 (blue type) are shown. (B) Percentage of MYH7 c.2155C>T (black bars) and MYH6 c.2161C>T alleles (white bars) in bulk transfected iPSC cultures. (C) Proportion of iPSC colonies harboring knock-in of MYH7 c.2155C>T, MYH6 c.2161C>T, or both MYH7 c.2155C>T and MYH6 c.2161C>T SNVs. (D) Sanger sequencing analysis of a representative iPSC clone harboring heterozygous knock-in of the MYH7 c.2155C>T SNV and an unedited MYH6 locus. The MYH7 c.2155C>T SNV (highlighted in red) and nucleotides used to distinguish MYH6 and MYH7 (highlighted in blue) are shown. PE, prime editing; iPSC, induced pluripotent stem cell; SNV, single nucleotide variant.

We also utilized PE to introduce a heterozygous SNV (SFTPA c.532G>A) in either the SFTPA1 or SFTPA2 genes. SFTPA1 and SFTPA2 exhibit >98% sequence homology and have been implicated in adult-onset pulmonary fibrosis and lung cancer.^16,23 Although in reverse orientation to one another, SFTPA1 and SFTPA2 are also arranged in tandem (Supplementary Fig. S4A), making them incompatible with endonuclease-based editing strategies. A nicking sgRNA was included in the PE cocktail along with mRNA encoding PE6b and MLH1dn, since editing efficiency was extremely low (<0.5%) in its absence. A single nucleotide difference located 74 bp downstream of the SFTPA c.532G>A target was used to distinguish SFTPA1 and SFTPA2 alleles (Supplementary Fig. S4). Of 72 colonies screened, 4 (5.6%) exhibited knock-in of the c.532G>A SNV within SFTPA1 and 3 (4.2%) within SFTPA2. However, two of the SFTPA2 knock-in clones also harbored a 9 bp deletion—located within the SFTPA1 gene in one clone and adjacent to the SFTPA2 c.532G>A SNV in the other—which spanned sequences between the epegRNA and nicking PAM sites. Large deletions exceeding the full length of the NGS amplicon (>200 bp) were also detected in four (4.2%) clones lacking SNV knock-in: two within SFTPA1 and two in SFTPA2. Nonetheless, we subsequently isolated subclones with clean heterozygous incorporation of the c.532G>A mutation exclusively within SFTPA1 or SFTPA2, which was confirmed using gene-specific primers followed by Sanger sequencing of the resulting amplicons (Supplementary Fig. S4C, D).

Lastly, we evaluated the feasibility of utilizing PE in our previously described one-step genome editing and reprogramming workflow, which enables the derivation of genetically engineered iPSC clones directly from human fibroblasts.^16,24 We initially aimed to derive iPSCs from healthy foreskin fibroblasts carrying a heterozygous SNV in ALPK3 (Fig. 4A), a gene that has previously been linked to familial cardiomyopathy.²³ An epegRNA encoding the ALPK3 c.1792C>T SNV and an adjacent synonymous change was introduced into primary foreskin fibroblasts (GM21808; Coriell) along with mRNA encoding PEmax in addition to the episomal vectors required for reprogramming. We also performed a direct comparison with our previously established protocol that uses a gene-editing cocktail consisting of a plasmid encoding the gene-specific sgRNA, ssODN repair template, and mRNA encoding Cas9-Gem—a Cas9 variant that is associated with lower rates of nonhomologous end joining.¹⁵ In our one-step workflow, iPSC colonies typically emerge from a single fibroblast cell and are therefore clonal in nature, eliminating the need for successive rounds of subcloning. As such, identification of successfully edited iPSC clones can be achieved using basic techniques such as allele-specific PCR and/or Sanger sequencing analysis (Fig. 4B, C). In the absence of MLH1dn, we successfully identified one (3.1%) out of 32 iPSC clones harboring heterozygous incorporation of the ALPK3 c.1792C>T SNV (Fig. 4D). The inclusion of MLH1dn mRNA led to over a fourfold increase in editing efficiency, with 6 (14%) out of 43 iPSC clones exhibiting heterozygous incorporation of the ALPK3 c.1792C>T SNV. With respect to the Cas9-Gem mRNA/ssODN condition, we identified 2 (4.7%) out of 43 iPSC clones harboring homozygous incorporation of the SNV and 2 (4.7%) clones with heterozygous SNV knock-in. However, both heterozygous clones also harbored indel mutations (1 bp or 2 bp deletion) in the second ALPK3 allele (Fig. 4C, D).

FIG. 4.

One-step prime editing and reprogramming of primary fibroblasts. (A) ALPK3 locus showing location of the ALPK3 c.1792C>T SNV (highlighted in red) and accompanying synonymous change (highlighted in blue). (B) Overview of the one-step PE/reprogramming workflow used to generate gene-edited iPSCs directly from primary fibroblasts. iPSC clones harboring knock-in of ALPK3 c.1792C>T were identified by allele-specific PCR and confirmed by Sanger sequencing (C). (D) Proportion of clones harboring heterozygous knock-in of the ALPK3 c.1792C>T SNV isolated from PE experiments performed with and without MLH1dn. Only clones with biallelic modifications were identified from the Cas9-Gem mRNA/ssODN condition. SNV, single nucleotide variant; PE, prime editing; iPSC, induced pluripotent stem cell; ssODN, single-stranded oligonucleotide.

We also tested whether PE can be used in conjunction with reprogramming to derive gene-edited iPSCs directly from patient peripheral blood mononuclear cells (PBMCs) from an infant with a severe form of autosomal recessive polycystic kidney disease (ARPKD) who harbors compound heterozygous mutations in the PKHD1 gene (Supplementary Fig. S5). We successfully used PE to correct one mutation, PKHD1 c.107C>T, which is the most common mutation associated with ARPKD.²⁵

Discussion

The development of advanced genome engineering systems has profoundly enhanced the potential of hPSCs in developmental biology, disease modeling, and regenerative medicine applications. However, achieving reliable knock-in of SNVs in a heterozygous context can be challenging. Here, we comprehensively demonstrate that PE is a robust gene-editing tool for introducing heterozygous edits in a broad range of genetic loci. We successfully generated iPSC lines carrying 15 unique SNVs across 8 different genetic backgrounds. Moreover, in nearly all instances, we successfully obtained multiple independent clones with heterozygous incorporation of the gene-specific SNV. This was accomplished without the need to screen large numbers of colonies, with no more than 72 individual colonies genotyped per experiment using a workflow that is free of subsequent cell purification or drug selection steps. We also demonstrate that PE can effectively introduce heterozygous edits in paralogous genes exhibiting very high sequence homology. Paralogues typically arise from duplication events and are generally not suitable for endonuclease-based editing strategies, as dual sgRNA recognition frequently leads to large deletions, inversions, or other chromosomal rearrangements. Indeed, dual sgRNA/Cas9 systems are commonly employed to intentionally induce such outcomes.^26–28

With respect to epegRNA design, we found a PBS length of 12–13 nt and an RT template length of 18–22 nt consistently produced workable editing efficiencies for the vast majority of the epegRNAs tested throughout this study. However, it remains possible that further optimization through systematic evaluation of different PBS and RT template lengths, as previously described,^5,11 could yield even higher editing efficiencies. Consistent with previous reports, we also found the inclusion of a secondary nicking sgRNA resulted in higher editing efficiency, but also led to an increased indel byproducts.¹¹ Nonetheless, compared with Cas9 endonuclease, PE was associated with a lower overall indel frequency. When direct comparisons between Cas9 nuclease and PE were performed, the overall editing efficiency was similar in both iPSCs and fibroblasts, where editing was combined with reprogramming. However, PE was by far the superior method for generating clones with heterozygous edits in both instances. While heterozygous knock-in was the main focus of this study, PE also induced homozygous edits, particularly when a secondary nicking sgRNA was included. While starting with a nicking sgRNA ∼50 bp downstream of the epegRNA-guided nick may typically be sufficient, evaluating the effect of additional nicking sgRNAs positioned both upstream and downstream of the edit may yield higher editing efficiency and/or reduced indel frequencies. Going forward, we recommend conducting PE experiments both with and without nicking sgRNAs and evaluating outcomes in bulk transfected cells. If sufficient editing efficiency is achieved without the nicking sgRNA, we suggest proceeding with colony isolation and screening.

The inclusion of mRNA encoding the dominant negative mismatch repair protein, MLH1dn, also consistently improved editing efficiencies. This was particularly evident in the instances where PE was combined with reprogramming to generate gene-edited iPSCs directly from primary fibroblasts or PBMCs. Importantly, transient MLH1dn expression was not associated with any genomic aberrations, with no aneuploidies or copy number variations detected in any of the clones generated throughout this study, as determined by SNP array molecular karyotyping analyses.

Taken together, our results demonstrate that PE is a highly effective approach for generating iPSCs with complex genetic modifications, facilitating the study of heterozygous variants within a consistent genetic background. This provides a powerful platform for investigating disease-associated dominant mutations in a fully isogenic context, with the potential to uncover functional outcomes that have yet to be explored.

Authors’ Contributions

S.E.H. designed experiments and wrote the article. A.S. performed NGS experiments with assistance from J.Y.K. and J.C. J.C. and A.S. constructed, prepped, and sequence-verified all epegRNA and sgRNA constructs. A.S. performed NGS analysis. A.S., J.Y.K., K.V., A.G., K.S., and C.G. all provided assistance with cell maintenance, transfection, cryopreservation, colony picking, and quality assurance. F.J.R., M.S., and M.R. provided bioinformatics support, including the setup of the custom CrispRVariants pipeline.

Footnotes

Acknowledgments

The authors thank users of the MCRI iPSC Gene Editing Facility for allowing them to share information around the generation of their requested iPSC lines.

Funding Information

The iPSC Gene Editing Facility was established using a generous donation from the Stafford Fox Medical Research Foundation and is currently supported by Phenomics Australia (PA) and the Novo Nordisk Foundation reNEW Center for Stem Cell Medicine (NNF21CC0073729). PA is supported by the Australian Government through the National Collaborative Research Infrastructure Strategy program.

Supplemental Material

References

Dorison

, Ghobrial

, Graham

, et al. Kidney organoids generated using an allelic series of NPHS2 point variants reveal distinct intracellular podocin mistrafficking. J Am Soc Nephrol, 2023; 34(1):88–109; doi: 10.1681/asn.2022060707

Cong

, Ran

, Cox

, et al. Multiplex genome engineering using CRISPR/Cas systems. Science, 2013; 339(6121):819–823; doi: 10.1126/science.1231143

Mali

, Yang

, Esvelt

, et al. RNA-guided human genome engineering via Cas9. Science, 2013; 339(6121):823–826; doi: 10.1126/science.1232033

Jinek

, East

, Cheng

, et al. RNA-programmed genome editing in human cells. Elife, 2013; 2:e00471; doi: 10.7554/eLife.00471

Anzalone

, Randolph

, Davis

, et al. Search-and-replace genome editing without double-strand breaks or donor DNA. Nature, 2019; 576(7785):149–157; doi: 10.1038/s41586-019-1711-4

Kim

, Moon

, Ko

, et al. Unbiased investigation of specificities of prime editing systems in human cells. Nucleic Acids Res, 2020; 48(18):10576–10589; doi: 10.1093/nar/gkaa764

, Busquets

, Verma

, et al. Highly efficient generation of isogenic pluripotent stem cell models using prime editing. Elife, 2022; 11:e79208; doi: 10.7554/eLife.79208

, Zhong

, Wu

, et al. Transient inhibition of p53 enhances prime editing and cytosine base-editing efficiencies in human pluripotent stem cells. Nat Commun, 2022; 13(1):6354; doi: 10.1038/s41467-022-34045-7

Chemello

, Chai

, Li

, et al. Precise correction of Duchenne muscular dystrophy exon deletion mutations by base and prime editing. Sci Adv, 2021; 7(18):eabg4910; doi: 10.1126/sciadv.abg4910

10.

Sürün

, Schneider

, Mircetic

, et al. Efficient generation and correction of mutations in human iPS cells utilizing mRNAs of CRISPR base editors and prime editors. Genes (Basel), 2020; 11(5):511; doi: 10.3390/genes11050511

11.

Doman

, Sousa

, Randolph

, et al. Designing and executing prime editing experiments in mammalian cells. Nat Protoc, 2022; 17(11):2431–2468; doi: 10.1038/s41596-022-00724-4

12.

Lindsay

, Burger

, Biyong

, et al. CrispRVariants charts the mutation spectrum of genome engineering experiments. Nat Biotechnol, 2016; 34(7):701–702; doi: 10.1038/nbt.3628

13.

Vlahos

, Sourris

, Mayberry

, et al. Generation of iPSC lines from peripheral blood mononuclear cells from 5 healthy adults. Stem Cell Res, 2019; 34:101380; doi: 10.1016/j.scr.2018.101380

14.

Chen

, Gulbranson

, Hou

, et al. Chemically defined conditions for human iPSC derivation and culture. Nat Methods, 2011; 8(5):424–429; doi: 10.1038/nmeth.1593

15.

Howden

, McColl

, Glaser

, et al. A Cas9 Variant For Efficient Generation OF Indel-Free knockin or gene-corrected human pluripotent stem cells. Stem Cell Reports, 2016; 7(3):508–517; doi: 10.1016/j.stemcr.2016.07.001

16.

Howden

, Thomson

, Little

. Simultaneous reprogramming and gene editing of human fibroblasts. Nat Protoc, 2018; 13(5):875–898; doi: 10.1038/nprot.2018.007

17.

Herman

, Lam

, Taylor

, et al. Truncations of titin causing dilated cardiomyopathy. N Engl J Med, 2012; 366(7):619–628; doi: 10.1056/NEJMoa1110186

18.

Nelson

, Randolph

, Shen

, et al. Engineered pegRNAs improve prime editing efficiency. Nat Biotechnol, 2022; 40(3):402–410; doi: 10.1038/s41587-021-01039-7

19.

Bonnycastle

, Swift

, Mansell

, et al. Generation of human isogenic induced pluripotent stem cell lines with CRISPR prime editing. Crispr J, 2024; 7(1):53–67; doi: 10.1089/crispr.2023.0066

20.

Doman

, Pandey

, Neugebauer

, et al. Phage-assisted evolution and protein engineering yield compact, efficient prime editors. Cell, 2023; 186(18):3983–4002.e26; doi: 10.1016/j.cell.2023.07.039

21.

Marian

, Braunwald

. Hypertrophic cardiomyopathy: Genetics, pathogenesis, clinical manifestations, diagnosis, and therapy. Circ Res, 2017; 121(7):749–770; doi: 10.1161/circresaha.117.311059

22.

McGuigan

, Phillips

, Postlethwait

. Evolution of sarcomeric myosin heavy chain genes: Evidence from fish. Mol Biol Evol, 2004; 21(6):1042–1056; doi: 10.1093/molbev/msh103

23.

Legendre

, Butt

, Borie

, et al. Functional assessment and phenotypic heterogeneity of SFTPA1 and SFTPA2 mutations in interstitial lung diseases and lung cancer. Eur Respir J, 2020; 56(6):2002806; doi: 10.1183/13993003.02806-2020

24.

Howden

, Maufort

, Duffin

, et al. Simultaneous reprogramming and gene correction of patient fibroblasts. Stem Cell Reports, 2015; 5(6):1109–1118; doi: 10.1016/j.stemcr.2015.10.009

25.

Goggolidou

, Richards

. The genetics of Autosomal Recessive Polycystic Kidney Disease (ARPKD). Biochim Biophys Acta Mol Basis Dis, 2022; 1868(4):166348; doi: 10.1016/j.bbadis.2022.166348

26.

Blasco

, Karaca

, Ambrogio

, et al. Simple and rapid in vivo generation of chromosomal rearrangements using CRISPR/Cas9 technology. Cell Rep, 2014; 9(4):1219–1227; doi: 10.1016/j.celrep.2014.10.051

27.

Chen

, Xu

, Zhu

, et al. Dual sgRNA-directed gene knockout using CRISPR/Cas9 technology in Caenorhabditis elegans. Sci Rep, 2014; 4:7581; doi: 10.1038/srep07581

28.

Choi

, Meyerson

. Targeted genomic rearrangements using CRISPR/Cas technology. Nat Commun, 2014; 5:3728; doi: 10.1038/ncomms4728

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.02 MB

2.50 MB

0.00 MB