CRISPR-Cas9-directed gene tagging using a single integrase-defective lentiviral vector carrying a transposase-based Cas9 off switch

Abstract

Locus-directed DNA cleavage induced by the CRISPR-Cas9 system triggers DNA repair mechanisms allowing gene repair or targeted insertion of foreign DNA. For gene insertion to be successful, availability of a homologous donor template needs to be timed with cleavage of the DNA by the Cas9 endonuclease guided by a target-specific single guide RNA (sgRNA). We present a novel approach for targeted gene insertion based on a single integrase-defective lentiviral vector (IDLV) carrying a Cas9 off switch. Gene insertion using this approach benefits from transposon-based stable Cas9 expression, which is switched off by excision-only transposase protein co-delivered in IDLV particles carrying a combined sgRNA/donor vector. This one-vector approach supports potent (up to >80%) knockin of a full-length EGFP gene sequence. This traceless cell engineering method benefits from high stable levels of Cas9, timed intracellular availability of the molecular tools, and a built-in feature to turn off Cas9 expression after DNA cleavage. The simple technique is based on transduction with a single IDLV, which holds the capacity to transfer larger donor templates, allowing robust gene knockin or tagging of genes in a single step.

Keywords: AAV; CRISPR-Cas9; DNA transposon; Donor template; Gene tagging; HDR; IDLV; MT: Delivery Strategies; lentivirus; piggyBac; protein delivery.

Graphical abstract

Figure 1

Protein transduction of hyPB^Exc+ efficiently excises genomically integrated transposons (A) Schematic of the GagPol-hyPB^Exc+ fusion vector and an overall outline of the reporter system used for evaluating hyPB^Exc+-mediated genomic excision. (B) LVNP-hyPB^Exc+ protein delivery noticeably increases genomic transposon excision in HeLa and HEK293 reporter cell lines. ∗∗∗∗p < 0.0001 (unpaired t test). (C) Comparison of PB transposition efficiency of hyPB protein variants after delivery by lentiviral protein transduction and transfection of pPBT/PGK-Puro. ∗∗∗p < 0.001 (unpaired t test). (D) Repeated LVNP dosing only marginally increases transposon excision rates. HeLa and HEK293 reporter cell lines were transduced with hyPBmut- or hyPB^Exc+-loaded LVNPs on 3 consecutive days. Excision rates were quantified by flow cytometry after single, double, and triple LVNP delivery, as indicated by black triangles. (E) Enrichment of transposon excision by negative FIAU selection. Experiments were performed in biological triplicates (individual wells); bars represent mean, with dots corresponding to individual replicates.

Figure 2

IDLV-mediated co-delivery of sgRNA and hyPB^Exc+ protein for highly efficient generation of Cas9-negative knockout cell lines (A) Schematic of the vectors used for creation of Cas9-PuroTK-expressing cell lines (pPBT/EFS-Cas9-PuroTK) and for co-delivering sgRNA and hyPB^Exc+ protein (pLV/guide-AFF1-mCherry). (B and C) AFF1 indel formation was quantified by ICE in HeLa PB/Cas9-PuroTK cell lines transduced with IDLV-hyPB^Exc+/sgRNA.AFF1 before FIAU (B) and after enrichment of PB/Cas9-PuroTK transposon excision by negative selection with FIAU (C). (D) HeLa cell lines carrying different copy numbers of the PB/Cas9-PuroTK transposon were treated with IDLV-hyPB^Exc+ co-delivering hyPB^Exc+ protein and a sgRNA targeting AFF1 (AFF1) or LVNPs delivering hyPBmut protein (mut). Enrichment was quantified by colony formation. ∗∗∗p < 0.0005, ∗∗∗∗p < 0.0005 (unpaired t test). (E) Quantification of transposon excision in single clones isolated from HeLa PB/Cas9-PuroTK cells transduced with the IDLV-hyPB^Exc+/sgRNA.AFF1; AFF1 genotype was assessed by ICE analysis and excision of transposon with ddPCR. Experiments were performed in biological triplicates (individual wells); bars represent mean, with dots corresponding to individual replicates.

Figure 3

Efficient, traceless, and Cas9-mediated HDR-based EGFP tagging of endogenous proteins (A) Schematic of the donor vector co-delivered with hyPB^Exc+ protein. (B) Vector schematics of KI donor vectors with homology arms and sgRNA or without arms (Ctrl) and sgRNA. (C–F) EGFP insertion into the LMNA locus. (G–J) EGFP insertion into the VIM locus. HeLaCas9#10 cells (C and G) and HeLaCas9#15 cells (D and H) were transduced with IDLV/donor or IDLV-hyPB^Exc+/donor corresponding to 40 ng P24. HDR-based EGFP tagging was quantified by flow cytometry. Cells were transduced with 40 ng P24 of IDLV/donor. (E and I) or AAV/donor at an MOI of 1 × 10⁵. (F and J) and immediately thereafter nucleofected with Cas9/sgRNA. Experiments were performed in biological triplicates (individual wells); bars represent mean, with dots corresponding to individual replicates. ∗∗p < 0.005, ∗∗∗p < 0.0005, ∗∗∗∗p < 0.0005 (unpaired t test).

Figure 4

Quantification of targeted EGFP KI in LMNA and VIM (A and B) Verification of transposon excision by western blot using Cas9 antibody after FIAU selection in HeLa PB/Cas9-PuroTK clone #10 (A) and clone #15 (B). (C and D) Representative images from confocal microscopy of cells expressing EGFP-tagged LMNA (C) and VIM (D) with antibody staining for the corresponding tagged protein. Scale bars, 5 μM. (E and F) Pearson correlation analysis of EGFP (tag) and Alexa Fluor 647 (antibody) intensity in each pixel, with the ROI focusing on the nuclear membrane for LMNA (E) and the cytoplasm for VIM (F). (G) Representative images of the samples used for ImageStream analysis. Data are shown for Hela/Cas9#10 expressing EGFP and two LMNA KI clones. (H) ImageStream gating strategy for separation of LMNA EGFP-tagged cells and cells with a diffuse EGFP expression pattern. Discrimination of cells is based on a nuclear confined mask (LMNA mask). (I) ImageStream gating strategy for separation of homozygous and heterozygous LMNA EGFP-tagged populations. (J) ImageStream-based quantification of correct LMNA EGFP-tagged cells and distribution of homozygous and heterozygous KI events. (K) Representative images of heterozygous cells (top row) and homozygous cells (bottom row) for HeLaCas9#10 and HeLaCas9#15 LMNA KI cells. (L) ImageStream-based quantification of bi-allelic and mono-allelic EGFP-tagged LMNA loci. Experiments were performed in biological triplicates (individual wells); bars represent mean, with dots corresponding to individual replicates.