An engineered baculoviral protein and DNA co-delivery system for CRISPR-based mammalian genome editing

Abstract

CRISPR-based DNA editing technologies enable rapid and accessible genome engineering of eukaryotic cells. However, the delivery of genetically encoded CRISPR components remains challenging and sustained Cas9 expression correlates with higher off-target activities, which can be reduced via Cas9-protein delivery. Here we demonstrate that baculovirus, alongside its DNA cargo, can be used to package and deliver proteins to human cells. Using protein-loaded baculovirus (pBV), we demonstrate delivery of Cas9 or base editors proteins, leading to efficient genome and base editing in human cells. By implementing a reversible, chemically inducible heterodimerization system, we show that protein cargoes can selectively and more efficiently be loaded into pBVs (spBVs). Using spBVs we achieved high levels of multiplexed genome editing in a panel of human cell lines. Importantly, spBVs maintain high editing efficiencies in absence of detectable off-targets events. Finally, by exploiting Cas9 protein and template DNA co-delivery, we demonstrate up to 5% site-specific targeted integration of a 1.8 kb heterologous DNA payload using a single spBV in a panel of human cell lines. In summary, we demonstrate that spBVs represent a versatile, efficient and potentially safer alternative for CRISPR applications requiring co-delivery of DNA and protein cargoes.

Graphical Abstract

Figure 1.

Large pocket formed by the membrane envelope of baculovirus budded virions entraps soluble recombinant proteins overexpressed in insect cells. (A) Negative stain imaging of a baculovirus budded virion by electron microscopy (P = Pocket, N = Nucleocapsid, E = Envelope). (B) Schematic representation of the experimental setup to validate stochastic protein entrapment in budded virions (BVs) in Spodoptera frugiperda Sf21 cells and subsequent delivery in HEK293T human cells. (C, D) mCherry fluorescence detection in HEK293T 24 h post-transduction with the indicated protein-loaded BV (pBVs) carrying either mCherry or mCherry::NLS protein cargoes. (C) Representative flow-cytometry histograms of 10× concentrated pBV and (D) percentages of mCherry + cells upon serial-dilutions of concentrated viral stocks. Data are mean + s.d. of n = 3 independent biological replicates analysed by flow-cytometry. Multiplicity of infections of undiluted pBV are: mCherry-pBV ≈ 1, mCherry-pBV VSVG ≈ 50 and mCherry::NLS-pBV VSVG ≈ 20. (E, F) Time-course of DNA (eGFP) (E) and protein (mCherry) (F) delivery in HEK293T transduced with mCherry-pBV VSVG at multiplicity of transduction (MOT) of 50 with or without spinoculation. Data are mean + s.d. of n = 3 independent biological replicates analysed by flow-cytometry. (G, H) Live confocal microscopy of HEK293T at 24 or 48 h post-transduction with mCherry-pBV VSVG or mCherry::NLS-pBV VSVG at MOT = 10. Hoechst dye counterstains nuclei. Scale bar is 50 μm.

Figure 2.

Overexpressed Cas9 is abundant in pBV and enables rapid and highly efficient genome editing in HEK293T. (A) Schematic representation Cas9-pBV production in Sf21 insect cells. (B) Histogram of protein intensities normalised to VSVG (top 30 most abundant proteins identified in concentrated Cas9-pBV by UPLC–MS/MS) and pie-chart of discovered proteins abundancies. Colours indicate recombinant (purple), baculovirus (turquoise) and host (grey) encoded proteins. Mean ± s.d. of n = 3 independent biological replicates. (C, D) Comparison of mCherry knock-out efficiencies in HEK293T stably expressing mCherry and mCherry targeting sgRNAs following transfection or transduction with the indicated (C) Cas9 encoding plasmid or Cas9-pBV. (D) mCherry- cells % at 7 days post-transfection (500 ng plasmid) or Cas9-pBV transduction (MOT = 50). Data are mean + s.d. of n = 3 independent biological replicates analysed by flow-cytometry. *** P < 0.001, ns = not significant, Student's t-test. (E-G) Indel formation dynamics in HEK293T stably expressing mCherry following transduction or transfection with plasmids or Cas9-pBV indicated in (E). (F) Indels formation at the indicated time points following transfection or transduction with plasmids or Cas9-pBV and (G) representative Sanger sequencing chromatograms at 7 days. Data are mean ± s.d. of Sanger sequencing decomposition analysis (ICE) of n = 3 independent biological replicates. *** P < 0.001, Student's t-test. (H–M) mCherry knock-out via base editors delivered through pBV in HEK293T stably expressing mCherry. (H) Schematic representation of BE3 pBV with or without sgRNA and (I) overview of the targeted mCherry region for the introduction of a stop codon using BE3. In the editing window target and non-target G-nucleotides are highlighted in red and purple, respectively. (J-K) Flow-cytometry analysis of mCherry loss at 7 days post-transduction with 1, 2 or 3 consecutive administrations of PE3-pBV + sgRNA each at MOT ≈ 50. (J) Representative flow-cytometry histograms and (K) percentages of mCherry- cells. Mean ± s.d. of n = 3 independent biological replicates. (L-M) G-to-A conversion efficiencies in the editing window (L) and representative individual base editing rates (M). Data are mean ± s.d. of Sanger sequencing data analysed with EditR. n = 3 independent biological replicates.

Figure 3.

Selective incorporation of Cas9 in pBV (spBV) leads to higher editing efficiencies in a panel of human cell lines while maintaining undetectable off-target editing. (A-C) Development of an abscisic acid (ABA) inducible selective protein packaging strategy. (A) Pyl1 and ABI1 domains are fused to the N-terminal and C-terminal domain of Cas9 and VSVG, respectively. During viral packaging, addition of ABA promotes selective packaging of Cas9 into budded virions. (B) Workflow of ABA-induced Cas9-spBV production in Sf21 insect cells. (C) Western blot of Cas9 and gp64 in concentrated spBV cultured in presence or absence of ABA at the indicated concentrations [μM]. (D, E) DNA delivery and mCherry KO efficiencies in HEK293T stably expressing mCherry and sgRNA 3 following transduction with serial dilution of Cas9 spBV amplified in presence or absence of ABA at the indicated concentrations. (D) % of eGFP + cells assessed 24 h post-transduction and (E) mCherry KO efficiency assessed at 10 days post-transduction. MOT = multiplicity of transduction. Histogram of flow-cytometry data. mean ± s.d. of n = 3 independent biological replicates. (F) Schematic representation of multiplexed genome editing approach via spBV in HEK293T using a single Cas9-spBV carrying two sgRNA cassettes (HEKs1 and HEKs3). (G-H) Indel frequencies at HEKs1 and HEKs3 loci in HEK293T transduced with dual sgRNA Cas9-spBV at MOT 200 amplified in presence or absence of 100 μM ABA. (G) Representative Sanger sequencing chromatograms and (H) histogram of Sanger sequencing deconvolution data (ICE), n = 3 independent biological replicates. (I–K) Multiplexed spBV editing in HeLa, SH-SY5Y, RPE1-hTERT, A549 and HEK29T cells. (I) Live confocal microscopy at 24 h post-transduction (scalebar is 50 μm); (J) eGFP expression levels at 24 h post-transduction, mean ± s.d. of n = 3 independent biological replicates. (K) Indel frequencies at HEKs1 and HEKs3 loci in the indicated cell lines transduced with dual sgRNA Cas9-spBV amplified in presence of 100 μM ABA at MOT 10, 50 and 100. Data are mean + s.d. of Sanger sequencing deconvolution data (ICE), n = 3 independent biological replicates. (L–O) On-target and off-target indels frequencies in HEK293T following co-transfection with CMV Cas9 + individual sgRNA plasmids (500 ng) or all-in-one Cas9-spBV equipped with either EMX1 or VEGFA sgRNAs (MOT 10). (L) DNA delivery efficiency and (M) DRAQ7- cells (viability) at 48 h post-transfection or 24 h post-transduction. Histograms of flow-cytometry data, mean ± s.d. of n = 3 independent biological replicates. ** P < 0.01, Student's t-test. (N, O) Indel frequencies at on- and off- targets sites at 72 h post-transfection or transduction. (N) Absolute indel frequencies and (O) Normalised indel frequencies relative to transduction efficiencies in (L). Histograms of Sanger sequencing deconvolution data (ICE), mean + s.d. of n = 3 independent biological replicates. *** P < 0.001, Student's t-test.

Figure 4.

Knock-in using all-in-one pBV or spBV in diverse cell types. (A–E) pBV mediated ACTB C-terminal exon tagging via HDR or HITI-2c strategies. (A) Schematic representation of human ACTB locus. Red-triangles represent position and orientation of the selected ACTB sgRNA target site. (B) Constructs used to generate all-in-one HDR or HITI-2c pBV differ only for the DNA donor. The HDR donor is equipped with homology arms, while the HITI-2c donor is flanked by sgRNA targeted sequences. The knock-in cassette encodes a synthetic C-terminal exon fused to T2A::mCherry::P2A::Puro-R. Genotyping oligonucleotides are indicated by black arrows in (A, B and F). (C-D) mCherry detection in HEK293T 5 days post-transduction with the indicated all-in-one HDR or HITI-2c pBV at MOT 100. (C) Representative flow-cytometry dot plots and (D) histogram of flow-cytometry data. Mean ± s.d. of n = 3 independent biological replicates. n.s.=not significant, Student's t-test. (E) PCR genotypes of puromycin selected HEK239T transduced with all-in-one HDR or HITI-2c with 5′ junction or wild-type allele specific primer pairs. (F-M) spBV mediated ACTB C-terminal exon tagging via HITI-2c in a panel of human cell lines. (F) Construct used to generate all-in-one HITI-2c pBV. The knock-in cassette encodes a synthetic C-terminal exon fused to mCherry::T2A::Puro-R. (G) Knock-in efficiencies in HeLa, A549, RPE-1 hTERT, SH-SY5Y and HEK293T at 48 h post-transduction with HITI-2c-spBV at the indicated MOTs, amplified in presence or absence of 100 μM ABA. Data are mean + s.d. of flow-cytometry data. n = 3 independent biological replicates. (H) Representative live confocal images of unselected HITI-2c spBV ABA transduced cells at MOT 100 at 48 h post-transduction. Scalebar is 25 μm. (I) PCR genotype with wild-type, 5′ junction and 3′ junction specific primer pairs following puromycin selection for 4 days and amplification in absence of selective pressure for 15 days (HeLa, A549, SH-SY5Y and HEK293T), or at 4 days post-transduction in absence of selective pressure (RPE-1 hTERT). (J–M) Live confocal imaging and PCR genotype of puromycin selected Jurkat (J-K) and H4 (L-M) cell lines following transduction with HITI-2c spBV + ABA at MOT 10 and 100, respectively. Scalebar is 100 μm.