Targeted genome editing by lentiviral protein transduction of zinc-finger and TAL-effector nucleases

  1. Yujia Cai
  2. Rasmus O Bak
  3. Jacob Giehm Mikkelsen  Is a corresponding author
  1. Aarhus University, Denmark
5 figures, 1 table and 1 additional file

Figures

Figure 1 with 1 supplement
Targeted egfp gene disruption by lentiviral delivery of ZFN proteins.

(A) Schematic representation of the composition of Gag and GagPol polypeptides encoded by ZFN-encoding packaging constructs (top) and the production of ZFN-loaded LPs (bottom). Gag is composed of the N-terminal ZFN domain, an HIV-1 cleavage site (SQNY/PIVQ), the phospholipase C-δ1 pleckstrin homology (PH) domain, matrix (MA), capsid (CA), nucleocapsid (NC), and p6, whereas Pol consists of protease (PR), reverse transcriptase (RT), and integrase harboring the D64V mutation (IN-D64V). LPs harboring two types of ZFNs (indicated by blue and green dots inside the virion) are produced by co-transfecting 293T cells with pMD.2G (encoding VSV-G surface protein) and pZFNL-PH-gagpol-D64V and pZFNR-PH-gagpol-D64V encoding ZFNL and ZFNR, respectively. (B) and (C) Analysis of the contents of LPs by Western blot using HA- and p24-specific antibodies. HA-tagged ZFNs were incorporated in this LP batch, allowing detection of ZFNs and ZFN derivatives (left panel). ZFNs originating from non-intentional processing or cleavage at cryptic HIV-1 cleavage sites are indicated by arrows labeled with ZFN*. The same membrane was stripped and re-used for detection of p24 (right panels). It is indicated below each panel whether 0.2 μM of the protease inhibitor Saquinavir (SQV) was included during LP production. (D) egfp gene disruption by protein transduction of ZFNs in HEK293-eGFPmut reporter cells as measured by Surveyor nuclease-based detection of DNA mismatches. Cells were harvested for analysis 24 hr posttransduction. Arrowheads point to the specific cleavage products. Quantified locus modification rates (indel %) are indicated below relevant lanes. (E) egfp gene disruption by ZFN proteins at different time points after transduction. HEK293-eGFPmut reporter cells were transduced with 300 ng p24 LP-ZFNLR(gfp). Locus modification rates (indel %) are provided below the gel.

https://doi.org/10.7554/eLife.01911.003
Figure 1—figure supplement 1
Validation of the release of full-length ZFN protein within ZFN-loaded LPs.

Western blot analysis of the cleavage pattern of ZFN proteins was carried out on cellular protein, as control, and on protein prepared from ZFN-loaded virions. For the control, 293T cells were co-transfected with pcDNA3.1-HA-ZFNL(gfp) and pcDNA3.1-HA-ZFNL(gfp) expressing HA-tagged left and right ZFNs, respectively (left lane). Cell protein was purified for Western blot analysis 1 day after transfection. Right lanes show protein incorporated in LPs packaged with both HA-tagged left and right ZFNs. A short exposure of the blot is shown on the left, whereas a longer exposure of the blot is shown on the right. ZFNs originating from non-intentional processing are indicated by an asterisk (*).

https://doi.org/10.7554/eLife.01911.004
Targeted disruption of endogenous genes by protein transduction of ZFNs.

HEK293, normal human dermal fibroblasts (NHDFs) and primary human keratinocytes (HKs) were transduced with increasing amounts of LPs containing ZFNs targeting the CCR5 (left) and AAVS1 (right) loci. Cells were harvested for analysis 24 hr posttransduction. Arrowheads indicate specific cleavage products, whereas fragments marked with * were generated due to the presence of the CCR5 Δ32 allele in the analyzed cells. Quantified locus modification rates are indicated below relevant lanes.

https://doi.org/10.7554/eLife.01911.009
Figure 3 with 1 supplement
Targeted gene editing by ‘all-in-one’ IDLVs.

(A) Schematic representation of the production and intracellular action of IDLVs carrying two ZFNs (indicated by blue and green dots) and a vector with the donor sequence for HR-directed repair (indicated in red as an RNA homodimer). Upon endosomal escape and uncoating, the donor sequence is reverse-transcribed to double-stranded DNA that is imported to the nucleus, where it serves as a donor for repair either in the form of linear DNA or as 1-LTR or 2-LTR circles (only HR between linear DNA and the target is shown). Homologous sequences are highlighted in light yellow. The egfp gene harboring internal mutations (indicated by a red box) is repaired through ZFN-mediated cleavage and HR using the reverse-transcribed vector as a recombination donor. (BE) Correction of the egfp gene by lentiviral delivery of ZFN proteins. Flow cytometric analysis was performed 4 days after transduction or transfection. In (B), donor plasmid (pLV/egfp-donor-fw) was transfected 6 hr prior to ZFN protein transduction. Co-transfection of donor plasmid and ZFN-encoding plasmid DNA (pZFN) served as a positive control. In (C), the donor sequence was provided by IDLV/donor (MOI of 46) co-transduced with LP-ZFNLR(gfp), whereas in (D) correction was achieved by co-transduction with two IDLVs (IDLV-ZFNL(gfp)/donor and IDLV-ZFNR(gfp)/donor, respectively), both at an MOI of 9, loaded each with one of the two egfp-targeting ZFNs. In (E), ‘all-in-one’ IDLVs (IDLV-ZFNLR(gfp)/donor) induced potent gene correction. Gene editing was measured with virus loads ranging from an MOI of 2 (corresponding to 37 ng p24) to an MOI of 34 (corresponding to 600 ng p24). IDLVs without the VSV-G surface protein as well as reporter cells pretreated with of 1 μM Bafilomycin A1 (Baf A1) served as negative controls. (F) Targeted editing at endogenous CCR5 and AAVS1 loci. Schematic representation of PCR-based assay used for detection of gene editing at CCR5 and AAVS1 loci (left panel). Primers are indicated above the edited target sequence. LS, linker sequence. Gene editing at CCR5 and AAVS1 loci in HEK293 cells and NHDFs, as confirmed by PCR (right panel), was obtained with an MOI of 34. PCR fragments amplified from the AAVS1 locus and the CCR5 locus served as controls for CCR5- and AAVS1-directed LS insertion, respectively. The error bars represent ±SD from three independent replicates of the experiment.

https://doi.org/10.7554/eLife.01911.010
Figure 3—figure supplement 1
Determination of multiplicity of infection (MOI) of IDLVs carrying left and right ZFNs.

Quantitative PCR analysis was carried out on total DNA isolated from 105 HEK293-eGFPmut reporter cells treated in three independent experiments with a load of IDLV/donor and IDLV-ZFNLR(gfp)/donor corresponding to 30 ng p24. Genomic DNA was extracted 2 days after transduction. 48 ng genomic DNA was used for qPCR and each reaction was performed in triplicate. (A) Number of lentiviral DNA copies as measured by qPCR amplification of the WPRE element located in the lentiviral vector carrying the donor. (B) Determination of the number of copies of the albumin gene in the transduced cells. (C) Calculated multiplicity of infection (MOI) obtained in cells treated with a number of IDLVs corresponding to 30 ng p24. The MOI was calculated as follows: 2 × lentiviral DNA copy number/albumin gene copy number. (D) Determination of the number of infectious units (IU) per ng p24. The number of IU per ng p24 was calculated as follows: cell numbers × MOI/ng p24. The error bars represent ±SD from three independent replicates of the experiments using for each IDLV three separate vector preparations.

https://doi.org/10.7554/eLife.01911.011
Targeted egfp gene editing by lentiviral delivery of TALEN proteins.

(A) Schematic representation of the construction of the TALEN-GagPol polypeptide expression construct. GoldyTALEN was assembled as the Golden Gate assembly method into a shuttle plasmid pC-Goldy-TALEN-PH and was then cut out and cloned into pGFP-PH-gagpol-D64V to get the destination construct pTALEN-PH-gagpol-D64V expressing polypeptides composed of GoldyTALEN and GagPol connected by the HIV-1 protease cleavage site SQNY/PIVQ. NLS, SV40 nuclear localization signal, Δ152, 152 amino acids deletion from the wild-type TALE protein, RVDs, repeat variable di-residues, +63, 63 amino acids following the last repeat, PH, phospholipase C-δ1 pleckstrin homology domain. (B) egfp gene disruption by protein transduction of TALENs in HEK293-eGFPmut reporter cells as measured by Surveyor nuclease assay 24 hr posttransduction. (C) Sequences of egfp gene disruption by TALEN protein transduction in the HEK293-eGFPmut reporter cell line. Genomic DNA of HEK293-eGFPmut reporter cells transduced with 160 ng p24 LP-TALENLR(gfp) was used as PCR template for amplification and subsequent cloning of the part of the egfp gene encompassing the region recognized by the two TALENs. The wild-type sequence is shown at the top. The net change of length caused by the indels is indicated to the right of each sequence. Green dashes represent deleted nucleotides, whereas blue lower case letters illustrate inserted nucleotides. Three alleles out of 46 sequenced clones were found to be disrupted (disruption frequency: 6.5%). (D) Targeted egfp gene repair in HEK293-eGFPmut reporter cells. Cells were transfected with 1.8 μg donor plasmid (pLV/egfp-donor-fw) and transduced with 160 ng p24 of LP-TALENLR(gfp) 6 hr later. Cells treated only with LP-TALENLR(gfp) or donor served as negative controls. eGFP expression was analyzed by flow cytometry 15 days posttransduction. (E) Analysis of the contents of LP-HA-TALENR(gfp) by Western blot using HA-specific antibody. The left lane shows protein derived from 293T cells expressing right HA-TALEN(gfp) from transfected plasmid pcDNA3.1-Goldy-HA-TALENR(gfp). The expected size of the full-length TALEN is indicated, and truncated TALEN derivatives originating from non-intentional cleavage at cryptic, internal HIV-1 cleavage sites are indicated by an asterisk (*). The error bars represent ±SD from three independent replicates.

https://doi.org/10.7554/eLife.01911.012
On- and off-target activity by CCR5-targeting ZFNs delivered by LP protein transduction or plasmid transfection.

(A) Schematic representation of the PCR-based assay used for detection of simultaneous on-target and off-target cleavage at the CCR5 and CCR2 loci, respectively. Small blue and red boxes above each locus indicate the location of the two ZFN recognition sites in the CCR5 locus and the highly homologous sequences located within the CCR2 locus. Horizontal arrows indicate the location of sequences recognized by the primers used. (B) Detection of off-target cleavage by PCR-based detection of CCR2-CCR5 fusion fragments. HEK293 cells were seeded at a density of 1 × 105 cells/well on day 1 and were on day 2 either transduced with increasing amounts of LP-ZFNLR(CCR5) (50, 100, 200 ng p24, respectively; indicated by ‘LP’) or co-transfected with increasing amounts of the two ZFN-encoding plasmids (50 ng + 50 ng, 100 ng + 100 ng, 200 ng + 200 ng, respectively; indicated by ‘plasmid’). On day 3, cells were harvested for genomic DNA purification. (C) Disruption within the CCR5 and CCR2 loci after LP- and plasmid-directed ZFN delivery. HEK293 cells were treated with increasing amounts of LP-ZFNLR(CCR5) and ZFN-encoding plasmids as in (B). Sequence changes introduced by NHEJ-directed repair were identified by Surveyor nuclease-directed detection of mismatches in re-annealed PCR products. Analyses of CCR5- and CCR2-directed cleavage were performed on the same genomic DNA samples. Arrowheads indicate specific cleavage products. Quantified locus modification rates are indicated below each gel. (D) Distinct levels of CCR2 gene disruption using experimental conditions that support similar levels of CCR5 disruption after LP- and plasmid-directed ZFN delivery. HEK293 cells were treated by LP transduction (80 ng p24) or plasmid transfection (300 ng + 300 ng), essentially as described in (B), facilitating similar levels of gene disruption in the CCR5 locus. Surveyor nuclease assays were performed as in (C) except for the use of an alternative CCR5 primer set. Quantified locus modification rates are indicated below relevant lanes.

https://doi.org/10.7554/eLife.01911.013

Tables

Table 1

Summary of the targeting rates obtained by sequence-based identification of locus-targeted indels after LP-directed delivery of ZFN proteins

https://doi.org/10.7554/eLife.01911.005
Cell type
Target locusHEK293Normal human dermal fibroblastsPrimary human keratinocytes
egfp8/42 (19%)*N/AN/A
CCR57/42 (17%)8/46 (17%)11/45 (24%)
AAVS16/46 (13%)7/43 (16%)8/40 (20%)
  1. Provided ratios indicate the number of alleles with indels out of the total number of analyzed alleles after treatment with 600 ng p24 LP-ZFNLR(gfp), 200 ng p24 LP-ZFNLR(CCR5) or 200 ng p24 LP-ZFNLR(AAVS1), respectively.

  2. *

    indicates that data were obtained in HEK293-eGFPmut reporter cells; N/A, not available.

Table 1—source data 1

Sequences of egfp gene disruption by ZFN protein transduction in the HEK293-eGFPmut reporter cells.

Genomic DNA of HEK293-eGFPmut reporter cells transduced with 600 ng p24 LP-ZFNLR(gfp) was used as PCR template for amplification and subsequent cloning of the part of the egfp gene encompassing the region recognized by the two ZFNs. The wild-type sequence is shown at the top. The net change of length caused by indels is indicated to the right of each sequence. Green dashes represent deleted nucleotides, red lower case letters represent nucleotide substitutions, whereas blue lower case letters illustrate inserted nucleotides. If one particular sequence appeared in more than one clone, the exact number of clones with this particular sequence is provided in parenthesis.

https://doi.org/10.7554/eLife.01911.006
Table 1—source data 2

Sequences of CCR5 gene disruption by ZFN protein transduction of HEK293, NHDFs and HKs.

Genomic DNA of cells transduced with 200 ng p24 LP-ZFNLR(CCR5) was used as PCR template for amplification and subsequent cloning of a CCR5 amplicon encompassing the region recognized by the two ZFNs. The wild-type sequence is shown at the top. Types of indels are indicated as described in the legend to Table 1—source data 1.

https://doi.org/10.7554/eLife.01911.007
Table 1—source data 3

Sequences of AAVS1 gene disruption by ZFN protein transduction in HEK293, NHDFs and HKs.

Genomic DNA of cells transduced with 200 ng p24 LP-ZFNLR(AAVS1) was used as PCR template for amplification and subsequent cloning of an AAVS1 amplicon encompassing the region recognized by the two ZFNs. The wild-type sequence is shown at the top. Types of indels are indicated as described in the legend to Table 1—source data 1. If one particular sequence appeared in more than one clone, the exact number of clones with this sequence is provided in parenthesis.

https://doi.org/10.7554/eLife.01911.008

Additional files

Supplementary file 1

Oligonucleotides used in this study, ‘Targeted genome editing by lentiviral protein transduction of zinc-finger and TAL-effector nucleases’.

https://doi.org/10.7554/eLife.01911.014

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Yujia Cai
  2. Rasmus O Bak
  3. Jacob Giehm Mikkelsen
(2014)
Targeted genome editing by lentiviral protein transduction of zinc-finger and TAL-effector nucleases
eLife 3:e01911.
https://doi.org/10.7554/eLife.01911