Robust and inducible genome editing via an all-in-one prime editor in human pluripotent stem cells

Abstract

Prime editing (PE) allows for precise genome editing in human pluripotent stem cells (hPSCs), such as introducing single nucleotide modifications, small insertions or deletions at a specific genomic locus. Here, we systematically compare a panel of prime editing conditions in hPSCs and generate a potent prime editor, "PE-Plus", through co-inhibition of mismatch repair and p53-mediated cellular stress responses. We further establish an inducible prime editing platform in hPSCs by incorporating the PE-Plus into a safe-harbor locus and demonstrated temporal control of precise editing in both hPSCs and differentiated cells. By evaluating disease-associated mutations, we show that this platform allows efficient creation of both monoallelic and biallelic disease-relevant mutations in hPSCs. In addition, this platform enables the efficient introduction of single or multiple edits in one step, demonstrating potential for multiplex editing. Our method presents an efficient and controllable multiplex prime editing tool in hPSCs and their differentiated progeny.

Fig. 1. Comparison of different prime editing tools in hPSCs using an “H2B-turn-on reporter” cell line.

a Schematic of the H2B-turn-on reporter for evaluating prime editing efficiency. b Sequence of the region in the H2B-turn-on reporter cells (right panel) containing a “C” deletion and the sequence after restoration of the “C” by prime editing (left panel). c Detection of tdTomato under a fluorescence microscope before (Neg) and 48 h after prime editing with indicated conditions in the reporter cells. Bright-filed images are provided in the lower panel. Scale bar: 10 µm. One of four independent experiments (n = 4) is shown. d, e Representative FACS plots (d) and bar graph (e) showing the percentage of tdTomato-positive cells 48 h after electroporation. n = 4 independent electroporation reactions for Neg, PE2, PE4, PE2 + mP53DD, PE4 + mP53DD, PE2max, PE2max + mP53DD; n = 7 for PE4max, PE4max + mP53DD, PE4max + mP53DD + tev, PE4max + mP53DD + tmp; n = 3 for PE4max + tev and PE4max + tmp. f Miseq analysis of the desired “C” insertion and the byproduct frequencies 48 h after electroporation with different prime editing conditions (n = 3 independent electroporation reactions). g Prime editing outcome purity calculated by edit/byproduct ratio (n = 3 independent electroporation reactions). h Fold change in the percentage of tdTomato-positive cells under the indicated prime editing conditions relative to PE4max. n = 7 independent electroporation reactions for PE4max, PE4max + mP53DD, PE4max + mP53DD + tev, PE4max + mP53DD + tmp; n = 3 for PE4max+tev and PE4max + tmp. i Fold change in desired “C” insertion frequencies under the indicated editing conditions normalized to that of PE4max (n = 3 independent electroporation reactions). j–n Miseq analysis of desired and undesired edits of a 2nt deletion (j), a 30nt deletion (k), a 34nt “Loxp” insertion at the HEK3 locus (l), a 10nt deletion (m), and a 40nt deletion (n) at the SOX2 locus with indicated prime editing conditions (n = 3 independent electroporation reactions). Data in e–n are presented as mean ± S.D. p-values were calculated by one-way ANOVA with Tukey’s multiple comparison test (e–n). Source data are provided as a Source Data file for (e–n).

Fig. 2. Generation of all-in-one prime editor co-expressing PEmax, hMLH1dn, and hP53DD.

a Construction of all-in-one PEmax plasmid incorporating hMLH1dn and hP53DD simultaneously. The components are linked with PEmax via direct fusion or linkages, including P2A or IRES, as indicated. The PE-Plus plasmid consists of PEmax, hP53DD, and hMLH1dn linked with P2A and IRES in between. b, c Representative FACS plots (b) and bar graph (c) showing the proportions of tdTomato-positive cells at 48 h post-electroporation with the indicated prime editors together with the pegRNA in H2B-turn-on reporter cells. n = 6 independent electroporation reactions. Bars represent the mean ± S.D. d Schematic overview of experimental design to evaluate genome-wide off-target effects induced by PE-Plus and PEmax. The edited cells were isolated by FACS sorting of tdTomato-positive cells with frame restoration in the “H2B-turn-on reporter”. SNVs and indels induced by these two prime editors were identified by comparing them to the unedited parental cells. e Number of SNVs, insertion, and deletions identified in the PEmax and PE-Plus edited cells. f Number of different types of SNVs (left panel) and their relative proportion (right panel) in the PEmax and PE-Plus edited cells. Source data are provided as a Source Data file for (c, e, and f).

Fig. 3. Inducible prime editing in hPSCs with the iPE-Plus platform.

a Schematic workflow of iPE-Plus platform generation in hPSCs and induction of intended edits in the genome. b Schematic of doxycycline-inducible correction of a frameshift mutation in H2B with iPE-Plus platform in H2B-turn-on reporter. c Fluorescence images showing tdTomato activation using the iPE-Plus platform at indicated time points after doxycycline treatment. The iPE-Plus lines were transduced with either pegRNA (upper panel) or epegRNA (lower panel) lentivirus. Scale bar: 5 µm. Representative images from three independent experiments are shown. d Representative histograms at the indicated time points showing tdTomato-positive cell populations after doxycycline treatment in the presence of pegRNA (upper) or epegRNA (lower). Untreated (blue), doxycycline-treated (red). e Summary plot showing the average tdTomato percentage from three single-cell clones transduced with pegRNA and epegRNA lentivirus at indicated days of doxycycline treatment. Data represent the mean ± S.D. from 3 independent experiments. f Schematic of inducible prime editing to correct frameshift mutations in H2B during neuroectoderm induction and maintenance. Doxycycline was added for 7 days at indicated stages. g Representative images from three independent experiments of immunofluorescence staining of PAX6 and co-expression with the tdTomato reporter gene in NPC cells after 7 days of doxycycline treatment during neuroectoderm induction or NPC maintenance. Scale bar: 5 µm. h, i Representative FACS plots (h) showing tdTomato-positive cells at day 7 or day 14 upon 7 days of doxycycline treatment. Corresponding cells without doxycycline treatment served as controls. The bar graph (i) depicts the mean percentage of edited cells ± S.D from 3 independent experiments. Source data are provided as a Source Data file for (e and i).

Fig. 4. Inducible installation of disease-related mutations in hPSCs.

a Schematic of quantifying mutation rates by ddPCR. b, c N370S mutation rate in GBA gene (b) and L858R mutation rate in EGFR gene (c) generated by iPE, iPEmax, or iPE-Plus platforms with different days of doxycycline induction, as determined by ddPCR. Data represents the mean from two single-cell clones for each type of inducible line. d Schematic of evaluating prime editing outcomes by Mi-seq. e, f Mi-seq analysis of intended editing and by-products of GBA N370S (e) and EGFR L858R (f) mutation induction using iPE, iPEmax, or iPE-Plus platforms before or after 7 days of doxycycline induction. Data represent the mean from two clones for each type of cell line. g Evaluation of prime editing outcome purity by iPE, iPEmax, and iPE-Plus. Data are represented as the mean from two single-cell clones for each inducible line. h, i Miseq analysis of intended and unintended edits of the LRRK2 G2019S mutation induction (h) and a “Loxp” insertion at the HEK3 locus (i) using iPEmax or iPE-Plus. Bars represent the mean from two single-cell clones for each type of inducible line. j Prime editing purity calculated from h and i. Data are represented as the mean from two single-cell clones. Source data are provided as a Source Data file for (b, c, and e–j).

Fig. 5. Generation of single-cell clones carrying disease-related mutations using iPE, iPEmax and iPE-Plus platforms.

a–c Genotyping of single-cell clones from iPE, iPEmax, and iPE-Plus platforms with 7 days of induction of N370S mutation in GBA. Sequences of unedited, heterozygous, and homozygous mutated clones were determined by Sanger sequencing. The targeted locus is indicated by the red arrows (a). Genotyping for each clone is demonstrated with icons (b) and the proportions of different genotypes are summarized with a bar graph (c). d–f Genotyping of single-cell clones from the three platforms with 7 days of induction of EGFR L858R mutation via Sanger sequencing (d). Genotypes for single-cell clones were analyzed (e), and the percentages for each genotype were calculated (f) accordingly. Source data are provided as a Source Data file for (c, f).

Fig. 6. One-step installation of multiplex mutations in hPSCs by the iPE-Plus platform.

a Construction of dual epegRNAs (upper) and dual nicking sgRNAs (lower) driven by tandem U6 promoters for the induction of L858R and T790M mutations. b The induction rate of two EGFR mutations by iPE-Plus platform at indicated time points, as determined by ddPCR. Data represents the mean from two iPE-Plus clones. c Miseq evaluation of the intended L858R and T790M mutations in EGFR as well as byproducts after 7 days of induction by the iPE-Plus platform. Bars represent mean from two iPE-Plus clones (d, e) Genotyping of single cell clone after 7-days of dual EGFR mutation induction using the iPE-Plus platform (d). Single clones with precise single mutations or double mutations were summarized with pie charts. Source data are provided as a Source Data file for (b, c).