TARGETED DNA INTEGRATION WITH LENTIVIRAL VECTORS AND USES THEREOF

Abstract

The present invention relates to a composition of lentiviral particles comprising a mutated integrase and an RNA-guided nuclease, for targeted gene insertion.

Claims

1. A composition comprising: (i) a nucleic acid encoding a lentiviral vector with impaired integration activity, (ii) a nucleic acid encoding a polypeptide or protein comprising an RNA-guided nuclease or nickase, (iii) a nucleic acid encoding a guide RNA (gRNA) fused to an aptamer, (iv) a nucleic acid encoding a transgene of interest, and (v) a nucleic acid encoding a fusion protein comprising: a first protein comprising or consisting of a GAG polyprotein, a second protein comprising or consisting of an aptamer-binding protein.

2. The composition according to claim 2, wherein: said aptamer-binding protein is a MS2 bacteriophage coat protein (MCP), and said aptamer is a MS2 RNA tetraloop binding sequence.

3. The composition according to claim 1, wherein said RNA-guided nuclease or nickase is not fused to an integrase.

4. The composition according to claim 1, wherein said RNA-guided nuclease or nickase is a Cas protein.

5. The composition according to claim 1, wherein said RNA-guided nuclease or nickase is bound to a mutated hyperactive PiggyBac transposase.

6. (canceled)

7. The composition according to claim 1, wherein the integrase of (i) and/or of (v) comprises at least one amino acid mutation selected from the group consisting of D10K, E11K, E13K, D64A, D64E, G94D, G94E, G94R, G94K, D116A, D116E, N117D, N117E, N117R, N117K, S119A, S119P, S119T, S119G, S119D, S119E, S119R, S119K, N120D, N120E, N120R, N120K, T122K, T122I, T122V, T122A, T122R, A124D, A124E, A124R, A124K, A128T, E152A, E152D, D164N, Q168L, Q168A, E170G, F185K, K186E, R231G, R231K, R231D, R231E, R231S, K264R, K266R, and K273R of the HIV-1 integrase of SEQ ID NO: 1 or at a corresponding position in another integrase.

8. The composition according to claim 1, wherein the integration activity of the lentiviral vector of (i) is decreased or abolished by at least one mutation in the integrase sequence selected from the group consisting of a D116N substitution, numbering based on the HIV-1 integrase of SEQ ID NO: 1; a D164N substitution, numbering based on the HIV-1 integrase of SEQ ID NO: 1; an insertion of a premature stop codon; and a deletion of the integrase gene (IN).

9. (canceled)

10. (canceled)

11. A population of lentiviral particles comprising lentiviral particles comprising or consisting of: a. lentiviral proteins and/or genes or parts thereof, wherein the lentiviral integrase has impaired integration activity, b. an RNA-guided nuclease or nickase, or a nucleic acid sequence coding therefor, c. a guide RNA fused to an aptamer, and d. a fusion protein comprising (1) a first protein comprising or consisting of a GAG polyprotein, and (2) a second protein comprising or consisting of an aptamer-binding protein, or a nucleic acid sequence coding therefor.

12. (canceled)

13. (canceled)

14. (canceled)

15. A method for treating a genetic disease in a subject in need thereof, comprising administering to said subject a therapeutically effective amount of a population of lentiviral particles according to claim 11, wherein the population of lentiviral particles comprises a transgene of interest or wherein a transgene of interest is to be delivered to the subject before, concomitantly with or after the population of lentiviral particles.

16. The composition according to claim 1, wherein the nucleic acid encoding a lentiviral vector with impaired integration activity comprises an aptamer-binding protein.

17. The composition according to claim 3, wherein said MCP shares at least 75% identity with SEQ ID NO: 61.

18. The composition according to claim 3, wherein said MS2 RNA tetraloop binding sequence shares at least 75% identity with SEQ ID NO: 63.

19. The composition according to claim 1, wherein said nucleic acid encoding a fusion protein further comprises a third protein comprising or consisting of a POL polyprotein.

20. The composition according to claim 1, wherein said nucleic acid encoding a fusion protein further comprises a third protein comprising or consisting of a POL polyprotein, and wherein the integrase of (i) and/or the integrase comprised in the POL polyprotein of (v) comprises at least one amino acid mutation at a position selected from the group consisting of 10, 11, 13, 64, 94, 116, 117, 119, 120, 122, 124, 128, 152, 164, 168, 170, 185, 186, 231, 264, 266 of the HIV-1 integrase of SEQ ID NO: 1 or at a corresponding position in another integrase.

21. The population of lentiviral particles according to claim 11, wherein said RNA-guided nuclease or nickase is a Cas9 protein, or a nucleic acid sequence coding thereof.

22. The population of lentiviral particles according to claim 11, wherein said guide RNA is fused to at least one MS2 RNA tetraloop-binding sequence.

23. The population of lentiviral particles according to claim 11, wherein said fusion protein further comprises (3) a third protein comprising or consisting of a POL polyprotein, or a nucleic acid sequence coding thereof.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0473] FIGS. 1A-D relate to several packaging strategies for achieving programmable insertion. FIG. 1A shows that simultaneous loading with both Nuclease and payload DNA (co-packaged) are more efficient in targeted integration than separate addition of the components (editor+DNA). FIG. 1B shows the targeted integration of LV payload with the aptamer binding protein (MCP) fused to either the polyprotein GAG or the polyprotein GAG-POL. FIG. 1C shows the targeted integration of LV payload with the different packaging strategies: co-packaging with either NLS (NLS-Cas9), integrase fused with viral protein R (VPRIN-Cas9), LEDGF (p75-LEDGF-Cas9) and Gag-Pol (GAGPOL-Cas9). FIG. 1D shows how integrase mutants lentiviral proteins determine efficiencies of on-target integration levels upon simultaneous SpCas9 nuclease RNP delivery and LV vector infection.

[0474] FIG. 2 is a graph showing targeted DNA integration with different RNA guided nucleases: SpCas9, SaCas9, Cas12f and TnpB can be efficiently used for targeted delivery.

[0475] FIG. 3 shows the percentage of NHEJ activated cells following the delivery of either Cas9 nuclease or Cas9 fused to a Programmable Transposase (FiCAT), in a lentiviral particle using the MS2 system, in either DNA, RNA or protein form. Loading is measured by NHEJ performed in a reporter cell line.

[0476] FIGS. 4A-C relate to the integrated junction sequencing. FIG. 4A shows a comparison of the titer of targeted integration (TU/mL) obtained with a wild type lentiviral vector (WT-LV) and the co-delivery of a mutated integrase fused to MCP, a Cas9 protein and a MS2-containing gRNA (PILV), in HEK293T cells, with a comparison between on-target and off-target integration. FIG. 4B shows insertion events detected on each chromosome of HEK293T cells; the larger band on chromosome 19 corresponds to on-target integration. FIG. 4C exemplifies the integrated junction sequencing.

[0477] FIGS. 5A and 5B show loading strategies for Cas9 incorporation into lentiviral particles. FIG. 5A shows a viral production overview and corresponding genetic components. FIG. 5B illustrates seven (1 to 7) loading strategies to Cas9 into the VLPs. Line 1) Cas9 overexpression; line 2) fusion to VPR (a viral capsid protein); line 3) fusion to the viral Integrase protein either wild-type or line 4) mutated; line 5) fusion of Cas9 to both VPR and mutated viral Integrase; line 6) fusion to GAP-POL viral genes or line 7) fusion to human endogenous p75 protein.

[0478] FIGS. 6A and 6B shows loading of Cas9 in VLP and activity. FIG. 6A results of a Western blot and shows the relative Cas9 loading efficiency in the seven loading strategies described above from 1 to 7; condition 0 represents the non-integrative control without Cas9 presence. FIG. 6B shows editing activity measured by NHEJ reads in next generation illumina targeted sequencing of cells infected with the viral particles depending on the seven encapsulation strategies described above, with a regular RNA (negative) and with a gRNA harboring a tetravalent MS2 domain for MCP specific binding.

[0479] FIGS. 7A and 7B illustrate the MS2-MCP system for loading of Cas9 in VLP and activity. FIG. 7A is a scheme of the MS2-MCP system as an aptamer in the gRNA of Cas9 for efficient loading into the VLP. FIG. 7B results of the western blot showing the relative Cas9 loading efficiency in the seven loading strategies described above from 1 to 7; condition 0 represents the non-integrative control without Cas9 presence. In all cases, the gRNA expressed in the produced cells was the gRNA-MS2 variant.

[0480] FIGS. 8A and 8B show PILV (Programmable Insertion Lentiviral Vector) performance in programmable insertion reporter cell line. FIG. 8A is a schematic representation of the split GFP reporter system for programmable integration detection. The reporter payload was cloned in a Lentiviral vector with LTR instead of ITRs. FIG. 8B illustrates programmable insertion (GFP) and overall insertion (RFP) capacity of the different packaging systems, tested in the Hershey reported system. Efficiency was measured and compared by flow cytometry analysis.

[0481] FIG. 9 illustrates PILV performance in programmable insertion in endogenous targets in cell line models. After infection of the PILV packaging strategies and controls in HEK293T cells both the targeted integration measured by junction qPCR and the targeted NHEJ activity measured by Illumina sequencing were plotted to observe correlations. The packaging strategies were as follows: Clone, clonal cell line with a known targeted insertion event as positive control; NI, Non integrative negative control; PILV, packaging strategy with MCP-GAGPOL, MS2-gRNA and Cas9 overexpression; MCP-GAG, MCP was fused to GAG protein and gRNA-MS2 used to package Cas9; Cas9_Insynthetic, strategy 3) fusion to the viral Integrase protein wild-type; Cas9_IND64Vsynth, strategy 4) fusion to the viral Integrase protein mutated; Cas9_Ms2_mRNA, fusion of MCP to GAG and expression of Cas9 with MS2 tag to encapsulate the mRNA instead of the RNP; VPRIN_Cas9-186+11, strategy 5) fusion of Cas9 to both VPR and mutated viral Integrase; GAGPOL-NLS-Cas9, strategy 6) fusion to GAP-POL viral genes.

[0482] FIGS. 10A, 10B and 10C show PILV performance for improved HDR donors. FIG. 10A is a schematic representation of the HDR designed donors for PILVs and mechanisms of programmable integration using HDR. FIG. 10B shows the GFP signal of non-integrating and PILV vector for all three donors in HEK293T cells. FIG. 10C shows GFP signal at 7 days post transduction of HDR payloads using non-integrative lentiviral particles (NI) or PILVs.

[0483] FIGS. 11A and 11B show PILV performance for improved HDR donors for T-cell engineering. FIG. 11A is a schematic representation of the HDR designed donors for PILVs for the TRAC locus. FIG. 11B represents GFP signal of non-integrating GFP control and PILV vectors for TRAC programmable insertion in primary T-cells in two alternative PBMC donors.

EXAMPLES

[0484] The present invention is further illustrated by the following example.

Example 1: Targeted DNA Integration with Lentiviral Vectors

Materials and Methods

Plasmid Construction

The Following Plasmids were Obtained from Addgene: [0485] hCas9 (plasmid #41815), a pcDNA3.3-TOPO plasmid expressing human codon-optimized Cas9 nuclease under the control of a CMV promoter; [0486] psPAX2 (plasmid #12260), an empty second-generation lentiviral packaging plasmid comprising the gag, pol, tat and rev genes; [0487] pCMV-VSV-G (plasmid #8454), a plasmid expressing the G glycoprotein of the vesicular stomatitis virus under the control of a CMV promoter; [0488] pMDLg/pRRE (plasmid #12251), a third-generation lentiviral packaging plasmid expressing HIV-1 Gag (coding for the virion main structural proteins), HIV-1 Pol (coding for HIV-1 protease, HIV-1 reverse transcriptase and HIV-1 integrase) and HIV-1 Rev response element (RRE, binding site for the Rev protein which facilitates export of the RNA from the nucleus); [0489] pRSV-Rev (plasmid #12253), a third-generation lentiviral packaging plasmid expressing the Rev protein.

[0490] Mutations in the integrase were obtained with QuikChange Lightning Multi Site-Directed Mutagenesis Kit (Agilent #210513). The sequences of integrase mutants are disclosed in SEQ ID NO: 2 to SEQ ID NO: 9.

[0491] All the remaining vectors were built by Golden Gate assembly using Esp3I and T4 ligase. Plasmid pSICO (Addgene reference Plasmid #41815; a generic vector comprising a transgene of interest) of SEQ ID NO: 74, pRRL dual of SEQ ID NO: 75 (a generic vector comprising a transgene of interest), psMCP-GAG of SEQ ID NO: 76, and psMCP-GAGPOL of SEQ ID NO: 77 were used.

Cell Culture

[0492] HEK293T cells (ATCC CRL-3216) were grown in Dulbecco's modified Eagle medium (DMEM), supplemented with high glucose (Gibco, Thermo Fisher), 10% fetal bovine serum, 2 mM glutamine and 100 U penicillin, 0.1 mg/mL streptomycin.

[0493] K-562 cells (ATCC CRL-3343) and Jurkat-T cells (Clone E6-1, ATCC IB-152) were grown in RPMI 1640 medium (Gibco) supplemented with 10% FBS, 1% penicillin-streptomycin (Gibco) and 1% GlutaMAX (100) (Gibco).

[0494] All cell lines were grown at 37 C. under 95% air and 5% CO.sub.2.

Lentiviral Production

[0495] Lentiviral vectors were produced following protocol available at Addgene website at https://www.addgene.org/protocols/lentivirus-production/, and reproduced below:

INTRODUCTION

[0496] This protocol can be used to produce lentivirus from a lentiviral vector transfected into Lenti-X 293T cells using a polyethylenimine (PEI) transfection protocol. This procedure can be modified for alternative packaging cell lines or transfection reagents. Once produced, lentivirus can be used for a variety of downstream applications such as stable-cell line generation.

Workflow Timeline

[0497] Day 0: Seed 293T packaging cells [0498] Day 1 (pm): Transfect packaging cells [0499] Day 2 (am): 18 hours post transfection. Remove media, replace with fresh media [0500] Day 3-4 (am): Harvest virus

Reagent Preparation

[0501] 1. DMEM Complete: 10% v/v FBS and 4 mM L-alanyl-L-glutamine [0502] To a 500 mL bottle of DMEM high glucose, add 55 mL of heat inactivated FBS and 11 mL of 200 mM L-alanyl-L-glutamine. Store at 4 C. [0503] 2. 25 mM chloroquine diphosphate Dissolve 0.129 g of chloroquine diphosphate salt in 10 mL of sterile water. [0504] Filter sterilize through a 0.22 m filter. [0505] Aliquot 50-100 L and store at 20 C. [0506] Aliquots can be thawed and stored at 4 C. prior to use. Thawed aliquots should be discarded after 1-2 months. [0507] 3. 1 mg/mL PEI, linear MW 25,000 Da [0508] Dissolve 100 mg of powder in 100 mL of deionized water. [0509] While stirring, slowly add hydrochloric acid until the solution clears. [0510] Check the pH of the solution. [0511] Use hydrochloric acid or sodium hydroxide to adjust the pH to 7.0. Typically the solution will be basic and will need adjustment with hydrochloric acid first. [0512] Allow the solution to mix for 10 min and then recheck the pH to ensure that it has not drifted. [0513] Filter the solution through a 0.22 m membrane. [0514] Aliquot 500-1000 L into sterile tubes. [0515] Store the tubes at 80 C. [0516] After thawing, the solution can be stored at 4 C. for up to 2 months. After 2 months, discard the tube and thaw a new working stock. [0517] The optimal mass DNA:mass PEI ratio will need to be empirically determined for each new batch of 1 mg/mL PEI and for each cell line.

Procedure

[0518] Seed 293T packaging cells at 3.810.sup.6 cells per plate in DMEM complete in 10 cm tissue culture plates. [0519] Incubate the cells at 37 C., 5% CO2 for 20 hours. [0520] Gently aspirate media, add 10 mL fresh DMEM complete containing 25 M chloroquine diphosphate and incubate 5 hours. [0521] For 10 mL of DMEM complete, add 10 L of 25 mM chloroquine diphosphate. [0522] Prepare a mixture of the 3 transfection plasmids: [0523] psPAX2: 1.3 pmol [0524] pMD2.G: 0.72 pmol [0525] Transfer plasmid. 1.64 pmol [0526] OptiPro SFM to total volume of 500 L [0527] Dilute the above 500 L mixture into 500 L PEI-OptiPro SFM with enough PEI such that the ratio of g DNA:g PEI is 1:3 (1000 L total per 10 cm dish). [0528] Using transfer plasmid pHAGE TRE dCas9-KRAB (total ug of plasmid DNA 27.8 g), this would be 83.4 L of 1 mg/mL PEI in 416.6 L of OptiPro SFM per 10 cm dish. [0529] Gently add the diluted PEI to the diluted DNA. Add the diluted PEI dropwise while gently flicking the diluted DNA tube. Incubate the mixture 15-20 min at room temperature. [0530] Carefully transfer the transfection mix to the Lenti-X 293T packaging cells. Add the transfection mix dropwise being careful not to dislodge the cells. [0531] Incubate the cells for 18 hours, or until the following morning. [0532] The following morning, carefully aspirate the media. Replace the media with 15 mL of DMEM complete. [0533] Incubate the cells. [0534] Virus can be harvested at 48, 72, and 96 hours post transfection in individual harvests or a combined harvest where all the individual harvests are pooled. If pooling harvests, transfer the harvested media to a polypropylene storage tube and store at 4 C. between harvest. [0535] Centrifuge the viral supernatant at 500 g for 5 minutes to pellet any packaging cells that were collected during harvesting. [0536] Filter supernatant through a 0.45 m PES filter. [0537] The viral supernatant can be stored at 4 C. for several hours but should be aliquoted and snap frozen in liquid nitrogen and stored at 80 C. as soon as possible to avoid loss of titer.

[0538] Cells were produced in 10-cm dishes seeded with 4.910.sup.6 HEK293T cells a day prior to transfection, using 0.72 pmol of pCMV-VSV-G envelope plasmid, 1.64 pmol of pSICO or pRRL dual payloads (transfer plasmid) and 1.30 pmol psPAX2 plasmid, either in wild-type form or with pol mutants.

[0539] For nuclease packaging, 0.65 pmol of psPAX2 and 0.65 pmol of psMCP-GAG or psMCP-GAGPOL plasmids were used together with 0.65 pmol of the nuclease plasmid.

[0540] For transfection, plasmids were mixed in 500 L of Optimem and 100 mg of polyethylenimine (PEI). Two days after plasmid transfection, the supernatant was harvested and filtered and centrifuged over night at 4000 g and 4 C. Supernatant was discarded and lentiviral particles were resuspended to achieve 100 vector concentration.

Determination of Nuclease Activity

[0541] For VPR/IN/LEDGF nuclease fusion activity, traffic light reporter (TLR) (Certo et al., 2011. Nat Methods. 8(8):671-676) HEK293T cell line, which contains an out-of-frame mCherry ORF that gets reconstituted by NHEJ activity upon Cas9 targeting, was used. Fluorescence was measured by flow cytometry using a BD LSR Fortessa instrument (Yellow green 561 nm laser with 610/20 filter).

[0542] SaCas9, AsCas12f and TnpB editing and MS2 scaffold in gRNA impact HEK293T were assessed by transfecting cells with gRNA in p12 cells with 240,000 cells/well. Three days after nuclease addition, cells were pelleted and DNA extracted with KAPPA quick extract. PCR amplification of the respective target site was performed, and sequenced on Illumina sequencing by synthesis MiSeq platform with MiSeq Reagent Kit v2 (300 cycles, 2150 configuration).

Determination of On-Target Integration

[0543] A reporter cell line (Hershey) was built to asses targeted integration, in which reconstitution of a fluorescent protein ORF upon on-target integration led to a measurable fluorescent signal (GFP).

[0544] A viral payload plasmid containing of Emerald GFP (emGFP) and of intron was packaged in a lentiviral vector with corresponding packaging system and/or integrase mutant. A promoter-less C-terminal (C-t) half of emGFP preceded by a splicing acceptor was randomly inserted in the genome of HEK293T cells to build a reporter cell line. A target site was added upstream of the C-t emGFP.

[0545] To perform the experiment, 200,000 HEK293T Hershey reporter cells were seeded in 12 well/plates and infected with lentiviral particles the next day, at the same time of transfection, that led to emGFP reconstitution upon addition of a payload plasmid.

[0546] Lentiviruses were produced using standard protocols. Four days after infection, emGFP fluorescent signal was measured by FACS (BD LSR Fortessa instrument; blue 488 nm laser with 530/30 filter, Yellow Green 561 nm laser with 610/20 filter).

Integrated Junction Sequencing

[0547] INSERTseq method combines targeted amplification of integrated DNA, UMI-based correction of PCR bias and Oxford Nanopore long-read sequencing for robust analysis of DNA integration in a genome. INSERT-seq is capable of detecting events occurring at a frequency of up to 0.1%. INSERT-seq presents a complete handling of all insertions independently of repeat size.

Library Preparation and Sequencing

[0548] DNA was extracted using Nanobind kit (Circulomics, catalog no. NB-900-001-01), sheared to 2 kb fragments using g-TUBE (Covaris, catalog no. 520079). WGP primer mix from Nanopore PCR Barcoding Kit (SQK-PBK004) was additionally added to the second PCR. Sequencing was performed in Flongle R9.4.1 flow cells obtaining a total output of 300,000 reads. For the calculation of the limit of detection (LOD), a monoclonal sample from HEK293T cells with one true lentiviral insertion was diluted with WT genomic material at the proportions 1/100, 1/1,000 and 1/10,000. Dilutions were sequenced a Flongle R9.4.1 flowcell obtaining a sequencing output of 500,000, 500,000 and 150,000 reads respectively. The analysis was performed following the INSERTseq analysis.

Integration Site Analysis

[0549] Nanopore raw reads were base-called using Guppy 4.0.11 (made available by ONT via their webpage https://community.nanoporetech.com).

[0550] Read quality was assessed with NanoStats NanoQC and NanoPlot from NanoPack (De Coster et al., 2018. Bioinformatics. 34(15):2666-2669). Reads were filtered by quality (>10) and length (>200) with NanoFilt from NanoPack. Reads were clustered by UMI using a combination and adaptation of two previously published pipelines (pipeline-umi-amplicon distributed by ONT https://github.com/nanoporetech/pipeline-umi-amplicon). The clustering was performed by extracting the UMI sequences with Python scripting, sequences were clustered with vsearch (Rognes et al., 2016. PeerJ. 4:e2584) and the consensus sequence of the clusters was obtained by performing two rounds of polishing with racon (Vaser et al., 2017. Genome Res. 27(5):737-746) and two rounds of medaka (https://github.com/nanoporetech/medaka).

[0551] For the analysis of insertions, reads were filtered to force the presence of used adapters and trimmed to remove the adapter and insert sequence from the read with cutadapt (Martin, 2011. EMBnet j. 17(1):10-12). Reads were mapped against the reference genome with minimap2 (Li, 2018. Bioinformatics. 34(18):3094-3100) map-ont default parameters and filtered with Python scripting, selecting uniquely mapping reads with a map quality higher than 30. A first output is returned with bedtools (Quinlan & Hall, 2010. Bioinformatics. 26(6):841-842) in bed format containing all mapped reads. Afterwards, a peak calling step is performed with Python scripting where peaks are filtered by shape. A peak is considered to pass the shape filter when the Residuals Sum of Squares (RSS) of fitting the peak coverage to a beta distribution is lower than 1.

Implementation of the Optimized INSERTseq Protocol

[0552] Genomic DNA was extracted, fragmented, end-repaired and A-tailed followed by ligation of an adaptor that contains an UMI for read clustering and a barcode for sample demultiplexing in the computational pipeline. Reads were clustered by UMI, integration sequence and adapters were filtered and trimmed, reads were mapped against the reference genome and significant peaks were reported and annotated.

Results

Nuclease Packaging Strategies

[0553] Given that lentiviral DNA integrates in double-strand breaks (DSB), it was reasoned that packaging functional Cas9-gRNA into a lentiviral particle should generate targeted integration.

[0554] Six different packaging system configurations were tested to achieve targeted integration via packaging of SpCas9 protein in lentiviruses: [0555] (i) co-expression of SpCas9 upon vector production, [0556] (ii) co-expression of SpCas9 fusion with virion-targeted protein Vpr p6-GAG interacting domain, as previously described (Indikova & Indik, 2020. Nucleic Acids Res. 48(14):8178-8187), [0557] (iii) SpCas9 fusion with the integrase protein, [0558] (iv) fusion to VPR integrase as previously described (Montagna et al., 2018. Mol Ther Nucleic Acids. 12:453-462), [0559] (v) fusion to the integrase-binding domain of the LEDGF-p75 chromatin docking factor as previously described (Hare et al., 2009. PLoS Pathog. 5(1):e1000259), and [0560] (vi) direct fusion to GAGPOL polyprotein in the packaging plasmid.

[0561] SpCas9 protein was detected in all of the packaging systems, but no non-homologous end joining (NHEJ) editing was detected upon lentiviral administration to cultured cells. One hypothesis was that lack of editing could be due to missing SpCas9 gRNA, as U6-driven expression localizes the gRNA in the nucleus, and cytoplasmic gRNA expression has been shown to rescue packaging of gRNA in lentivirus-derived vesicles.

[0562] An MCP-GAGPOL packaging fusion and MS2 stem-loop containing SpCas9 gRNA were used to recruit the gRNA to lentiviral particles. Using this configuration, the editing activity upon transduction with lentiviral particles was rescued (FIG. 1A).

[0563] The efficiency of targeted integration was compared between lentiviral particles comprising the aptamer binding protein MS2 coat protein (MCP) fused either the lentiviral polyprotein GAG-POL or the polyprotein GAG alone (FIG. 1B). Both conditions led to a similar targeted integration rate.

[0564] Targeted integration was observed upon simultaneous lentiviral infection and Cas9 RNP delivery in Jurkat-T cells. Four different packaging strategies were used to generate a programmable lentivirus (PILV): Cas9 was co-packaged with either NLS (NLS-Cas9), with the integrase fused with viral protein R (VPRIN-Cas9), with LEDGF (p75-LEDGF-Cas9) or with Gag-Pol (GAGPOL-Cas9). NLS-Cas9 and VPRIN-Cas9 produced the highest insertion rate (FIG. 1C).

[0565] Several integrase (IN) mutants were tested, and all showed better targeted integration performances than the wild-type integrase (FIG. 1D).

[0566] Finally, the proportion of NHEJ-activated cells was compared when the DNA payload was delivered using an editing tool that was either a Cas9 nuclease, or a fusion protein comprising a Cas9 and a programmable transposase (PT condition), in a lentiviral particle using the MS2 system. In addition, the editing tool was provided in either DNA, RNA or protein (RNP) form.

Compatibility with Other Nucleases

[0567] Given the vast repertoire of programmable nucleases that have been described since SpCas9, the programmable lentivirus packaging system (PILV) was tested with different programmable nucleases (AsCas12 and TnpB). Targeted integration was detected in all of the systems (FIG. 2), showing that the system is compatible with any nuclease enzyme.

Co-Delivery with a Programmable Transposase

[0568] Using the MCP-MS2 system, either Cas9 or a fusion protein comprising Cas9 and a hyperactive PiggyBac transposase (FiCAT, with R372A/K375A/D450N substitutions) were loaded in lentiviral particles. Editing activity was measured in traffic-light reporter HEK293T cells transduced with lentiviral particles loaded with the gene editors either in mRNA or protein (RNP) form (FIG. 3).

[0569] The results show gene editing activity for both conditions, although better results were obtained when the gene editor was delivered in protein form.

Measurement of On-Target Integration

[0570] On-target integration was measured using the INSERTseq approach as illustrated in FIG. 4B and FIG. 4C.

[0571] As shown on FIG. 4A, the overall integration and the on-target integration activity were compared between a wild-type lentiviral particle (WT-LV) and the programmable lentivirus (PILV). PILV enabled precise on-target integration compared to WT-LV, making it an optimal system for therapeutic applications.

Example 2: Strategies for Cas9 Incorporation into Lentiviral Particles

[0572] To incorporate Cas9 nuclease into the viral like particles (VLPs), several nuclease expression approaches in the lentiviral producing cell can be adopted. In addition to overexpression on the cytoplasm of producing cells and passive loading (FIG. 5A) 6 other strategies for Cas9 loading were explored. These strategies (FIG. 5B) include: 1) Cas9 overexpression; 2) fusion to VPR (a viral capsid protein); 3) fusion to the viral Integrase protein either wild-type or 4) mutated; 5) fusion of Cas9 to both VPR and mutated viral Integrase; 6) fusion to GAP-POL viral genes or 7) fusion to human endogenous p75 protein.

Materials and Methods

Cell Culture & Plasmid Cloning

[0573] HEK293T cells (ATCC CRL-3216) were grown in Dulbecco's modified Eagle medium (DMEM), supplemented with high glucose (Gibco, Thermo Fisher), 10% fetal bovine serum, 2 mM glutamine and 100 U penicillin, 0.1 mg/mL streptomycin.

[0574] K-562 cells (ATCC CRL-3343) and Jurkat-T cells (Clone E6-1, ATCC IB-152) were grown in RPMI 1640 medium (Gibco) supplemented with 10% FBS, 1% penicillin-streptomycin (Gibco) and 1% GlutaMAX (100) (Gibco).

[0575] All cell lines were grown at 37 C. under 95% air and 5% CO.sub.2.

[0576] Additional vectors not included in example 1 were built by Golden Gate assembly using Esp3I and T4 ligase. Plasmid pRRL_Cas9_Insynth containing the fusion of the codon optimized IN and Cas9 of SEQ ID NO: 79, pRRL_VPR-IN186opt-Cas9 containing the fusion of VPR, mutated IN and Cas9 of SEQ ID NO: 80, were used.

Lentiviral Production

[0577] Lentiviral vectors were produced following protocol available at Addgene website at https://www.addgene.org/protocols/lentivirus-production/, and reproduced below:

Introduction

[0578] This protocol can be used to produce lentivirus from a lentiviral vector transfected into Lenti-X 293T cells using a polyethyleneimine (PEI) transfection protocol. This procedure can be modified for alternative packaging cell lines or transfection reagents. Once produced, lentivirus can be used for a variety of downstream applications such as stable-cell line generation.

Workflow Timeline

[0579] Day 0: Seed 293T packaging cells [0580] Day 1 (pm): Transfect packaging cells [0581] Day 2 (am): 18 hours post transfection. Remove media, replace with fresh media [0582] Day 3-4 (am): Harvest virus

Reagent Preparation

[0583] 1. DMEM Complete: 10% v/v FBS and 4 mM L-alanyl-L-glutamine [0584] To a 500 mL bottle of DMEM high glucose, add 55 mL of heat inactivated FBS and 11 mL of 200 mM L-alanyl-L-glutamine. Store at 4 C. [0585] 2. 25 mM chloroquine diphosphate [0586] Dissolve 0.129 g of chloroquine diphosphate salt in 10 mL of sterile water. [0587] Filter sterilize through a 0.22 m filter. [0588] Aliquot 50-100 L and store at 20 C. [0589] Aliquots can be thawed and stored at 4 C. prior to use. Thawed aliquots should be discarded after 1-2 months. [0590] 3. 1 mg/mL PEI, linear MW 25,000 Da [0591] Dissolve 100 mg of powder in 100 mL of deionized water. [0592] While stirring, slowly add hydrochloric acid until the solution clears. [0593] Check the pH of the solution. [0594] Use hydrochloric acid or sodium hydroxide to adjust the pH to 7.0. Typically the solution will be basic and will need adjustment with hydrochloric acid first. [0595] Allow the solution to mix for 10 min and then recheck the pH to ensure that it has not drifted. [0596] Filter the solution through a 0.22 m membrane. [0597] Aliquot 500-1000 L into sterile tubes. [0598] Store the tubes at 80 C. [0599] After thawing, the solution can be stored at 4 C. for up to 2 months. After 2 months, discard the tube and thaw a new working stock. [0600] The optimal mass DNA:mass PEI ratio will need to be empirically determined for each new batch of 1 mg/mL PEI and for each cell line.

Procedure

[0601] Seed 293T packaging cells at 3.810.sup.6 cells per plate in DMEM complete in 10 cm tissue culture plates. [0602] Incubate the cells at 37 C., 5% CO2 for 20 hours. [0603] Gently aspirate media, add 10 mL fresh DMEM complete containing 25 M chloroquine diphosphate and incubate 5 hours. [0604] For 10 mL of DMEM complete, add 10 L of 25 mM chloroquine diphosphate. [0605] Prepare a mixture of the 3 transfection plasmids: [0606] psPAX2: 1.3 pmol [0607] pMD2.G: 0.72 pmol [0608] Transfer plasmid: 1.64 pmol [0609] OptiPro SFM to total volume of 500 L [0610] Dilute the above 500 L mixture into 500 L PEI-OptiPro SFM with enough PEI such that the ratio of g DNA:g PEI is 1:3 (1000 L total per 10 cm dish). [0611] Using transfer plasmid pHAGE TRE dCas9-KRAB (total ug of plasmid DNA 27.8 g), this would be 83.4 L of 1 mg/mL PEI in 416.6 L of OptiPro SFM per 10 cm dish. [0612] Gently add the diluted PEI to the diluted DNA. Add the diluted PEI dropwise while gently flicking the diluted DNA tube. Incubate the mixture 15-20 min at room temperature. [0613] Carefully transfer the transfection mix to the Lenti-X 293T packaging cells. Add the transfection mix dropwise being careful not to dislodge the cells. [0614] Incubate the cells for 18 hours, or until the following morning. [0615] The following morning, carefully aspirate the media. Replace the media with 15 mL of DMEM complete. [0616] Incubate the cells. [0617] Virus can be harvested at 48, 72, and 96 hours post transfection in individual harvests or a combined harvest where all the individual harvests are pooled. If pooling harvests, transfer the harvested media to a polypropylene storage tube and store at 4 C. between harvest. [0618] Centrifuge the viral supernatant at 500 g for 5 minutes to pellet any packaging cells that were collected during harvesting. [0619] Filter supernatant through a 0.45 m PES filter. [0620] The viral supernatant can be stored at 4 C. for several hours but should be aliquoted and snap frozen in liquid nitrogen and stored at 80 C. as soon as possible to avoid loss of titer.

[0621] Cells were produced in 10-cm dishes seeded with 4.910.sup.6 HEK293T cells a day prior to transfection, using 0.72 pmol of pCMV-VSV-G envelope plasmid, 1.64 pmol of pSICO or pRRL dual payloads (transfer plasmid) and 1.30 pmol psPAX2 plasmid, either in wild-type form or with pol mutants.

[0622] For nuclease packaging, 0.65 pmol of psPAX2 and 0.65 pmol of psMCP-GAG or psMCP-GAGPOL plasmids were used together with 0.65 pmol of the nuclease plasmid.

[0623] For transfection, plasmids were mixed in 500 L of Optimem and 100 mg of polyethylenimine (PEI). Two days after plasmid transfection, the supernatant was harvested and filtered and centrifuged over night at 4000 g and 4 C. Supernatant was discarded and lentiviral particles were resuspended to achieve 100 vector concentration.

Determination of Nuclease Activity

[0624] Produced lentiviral particles with the different strategies for loading Cas9 and gRNA were use to infect HEK293T or other cell line models. Three days after nuclease addition in the viral particles, cells were pelleted and DNA extracted with KAPPA quick extract. PCR amplification of the respective target site was performed, and sequenced on Illumina sequencing by synthesis MiSeq platform with MiSeq Reagent Kit v2 (300 cycles, 2150 configuration).

Determination of Nuclease Activity

[0625] Lentiviral particles were lysed in RIPA buffer for protein. Proteins were analyzed by western blot following standard methods, in brief: Proteins were loaded and run in SDS-PAGE gels (Invitrogen) and then were transfer to a nitrocellulose blot for protein detection using anti-Cas9 and anti-p24 antibodies (AbCam) and HRP-conjugated secondary antibodies. Proteins were detected and quantified with chemiluminescence development, using luminol as a substrate.

Results

[0626] Although presence of Cas9 can be achieved with most of the 7 tested strategies (FIG. 6A), no nuclease activity could be detected using these approaches (FIG. 6B).

[0627] We next tested several alternative loading strategies with the MS2-MCP RNA-protein system to take advantage of the gRNA interaction with Cas9. In one of these strategies, we co-expressed a tetravalent MS2 in the gRNA with a modified GAG-POL viral gene fused with MCP protein for active loading of the Cas9 into the viral particle, as well as loading of the gRNA for adequate targeting and cleavage of the target DNA in the infected cell (FIG. 6A).

[0628] This approach showed an increased presence of Cas9 when the MCP-MS2 system was used (FIG. 7B) but also a recovery of the nuclease activity in the infected cell, but surprisingly a complete recovery of the nuclease activity in the infected cell (FIG. 6B) when the MCP-MS2 system was used.

Example 3: Validation of Best Performing Cas9 Loading PILV Prototype

Materials and Methods

[0629] Produced lentiviral particles with different loading strategies, produced following the same methods as in example 1 and 2, with the different strategies for loading Cas9 and gRNA were used to infect HEK293T reporter cells or K562 cells. Infected cells were analyzed as follows.

Determination of Nuclease Activity

[0630] Three days after nuclease addition in the viral particles, cells were pelleted and DNA extracted with KAPPA quick extract. PCR amplification of the respective target site was performed, and sequenced on Illumina sequencing by synthesis MiSeq platform with MiSeq Reagent Kit v2 (300 cycles, 2150 configuration).

Determination of On-Target Integration

[0631] A reporter cell line (Hershey) was built to asses targeted integration, in which reconstitution of a fluorescent protein ORF upon on-target integration led to a measurable fluorescent signal (GFP).

[0632] A viral payload plasmid containing of Emerald GFP (emGFP) and of intron was packaged in a lentiviral vector with corresponding packaging system and/or integrase mutant. A promoter-less C-terminal (C-t) half of emGFP preceded by a splicing acceptor was randomly inserted in the genome of HEK293T cells to build a reporter cell line. A target site was added upstream of the C-t emGFP.

[0633] To perform the experiment, 200,000 HEK293T Hershey reporter cells were seeded in 12 well/plates and infected with lentiviral particles the next day, at the same time of transfection, that led to emGFP reconstitution upon addition of a payload plasmid.

[0634] Lentiviruses were produced using standard protocols. Four days after infection, emGFP fluorescent signal was measured by FACS (BD LSR Fortessa instrument; blue 488 nm laser with 530/30 filter, Yellow Green 561 nm laser with 610/20 filter).

[0635] For K562 cells, genomic DNA was extracted 5 days upon infection, and junction PCR of the target site and viral payload was amplified by semiquantitive standard methods (SYBRgreen qPCR, Applied Bio Systems) an compared to total viral copies and endogenous gene copy numbers.

Results

[0636] Once the loading system of both Cas9 and gRNA was established for efficient nuclease activity into the infected donors, several Integrase variations were tested in the context of Integrase defective viral particles to achieve programmable integration.

[0637] To assess programmable integration in a defined target site by the gRNA were used a reporter HEK293T cell line based on a Split GFP reporter (FIG. 8A). The efficiency of each packaging system was assessed and the best performing system was shown to be the MCP-GAG fusion with the gRNA-MS2 as well as the mRNA encapsulation with the MCP-MS2 system instead of the mRNA. PILV with MCP-GAGPOL showed very promising results (FIG. 8B). PILV with MCP-GAGPOL was improved in subsequent repeats by improving production rates. Direct fusion with Integrase or POL showed the worst results.

[0638] In addition to the cellular reporter system for programmable integration, were tested and validated the programmable insertion efficiency of PILV in endogenous target sites of alternative cell line model K562. Programmable integration efficiency after episomal decay was assessed by junctions targeted qPCR and relative copy numbers were plotted and compared to targeted NHEJ activity (FIG. 9). Strategy 1 consisting in MCP-GAGPOL fusion and MS2-MCP interaction with gRNA and MCP-GAG fusion were shown to be the most promising ones followed by mRNA encapsulation in the Viral like particle instead of the RNP.

Example 4: Programmable Integration of PILV Genomes by Homology Dependent Recombination (HDR)

Materials and Methods

Cell Culture & Plasmid Cloning

[0639] HEK293T cells (ATCC CRL-3216) were grown in Dulbecco's modified Eagle medium (DMEM), supplemented with high glucose (Gibco, Thermo Fisher), 10% fetal bovine serum, 2 mM glutamine and 100 U penicillin, 0.1 mg/mL streptomycin.

[0640] K-562 cells (ATCC CRL-3343) and T cells (isolated from human donors, and activated) were grown in RPMI 1640 medium (Gibco) supplemented with 10% FBS, 1% penicillin-streptomycin (Gibco) and 1% GlutaMAX (100) (Gibco).

[0641] T-cells were isolated form Buffy coats of up to 4 human donors; PBMCs were purified and T-cells were activated using standard methods for active proliferation before infection.

[0642] All cell were grown at 37 C. under 95% air and 5% CO.sub.2.

[0643] Additional vectors not included in example 1 and 2 were built by Golden Gate assembly using Esp3I and T4 ligase. Plasmid pLV RAB11A-HA-GFP-HA a lentiviral payload donor with GFP tag for targeted integration at the RAB11A site of SEQ ID NO: 81; pLV RAB11A-HA-GFPinv-HA a lentiviral payload donor with GFP tag in inverted orientation for targeted integration at the RAB11A site of SEQ ID NO: 82; pLV FBL-HA-mGFP-HA a lentiviral payload donor with GFP tag for targeted integration at the FBL site of SEQ ID NO: 83; pLV GFP-TRAC-HA a lentiviral payload donor with GFP tag for targeted integration at the TRAC site of SEQ ID NO: 84; pLV eGFP-TRAC-HA a lentiviral payload donor with eGFP tag for targeted integration at the TRAC site of SEQ ID NO: 85, were used.

Determination of On-Target Integration

[0644] Primary T-cells, or HEK293T, K562 cell line models were infected with the viral particles generated using the same methods as example 1 and 2, with the HDR genetrap reporters on endogenous genes, in which a GFP tag was added to the protein, for site specific insertion detection.

[0645] To perform the experiment, 200,000 HEK293T or K562, or 500,000 activated T-cells were infected with lentiviral particles. GFP fluorescent signal was measured by FACS (BD LSR Fortessa instrument; blue 488 nm laser with 530/30 filter, Yellow Green 561 nm laser with 610/20 filter) in different time points to track signal.

Results

[0646] The cargo donors were improved by generating viral genomes for programmable insertion with homology arms to the target site of interest to boost efficiency on actively dividing cells. Three alternative PILV cargoes were designed and generated (FIG. 10A). In HEK293T cells (FIG. 10B), the RAB11A reporter construct gave a high level of nonspecific signal, but the inverted version of this promoter drastically reduced noise signal and increased fluorescence for the PILV condition, consistent with the programmable insertion of the payload. FIG. 10C shows a modest increase in the PILV signal on both cell lines consistent with the programmable insertion of the HDR payload.

[0647] In addition, FBL reporters showed similar results with up to 10% of specific GFP signals for PILVs in HEK293T cells (FIG. 10B). Similar results were observed for the HDR payloads in cell models K562 and JurcaTs, with episomal leaking expression present and increase of signal for PILV in both payloads (FIG. 10C).

[0648] To validate the therapeutic potential of the programmable integration of lentiviral vectors in the context of cell engineering, a TRAC specific gene trap was designed and built for generating T-cells with therapeutic cargos inserted in the TRAC locus (FIG. 11A). The programmable insertion of the HDR based TRAC specific PILV payload was tested in primary T-cells of two different PBMC donors. Transduction efficiencies of above 50% were obtained with the programmable insertion vectors compared to none for the non-integrative control (FIG. 11B).

[0649] These results validate the transfer potential of PILVs as a safe and controllable manner of doing T-cell and other primary cell engineering for therapeutic purposes.

CONCLUSION

[0650] Targeted DNA integration of gene sized fragments in mammalian genomes can enable precise addition of therapeutic messages and new functionalities. Lentiviral vectors are therapeutically used for uncontrolled insertion of transgenes both for ex vivo applications for advance cell therapy and in vivo for gene therapy. By mutating the integrase protein and incorporating SpCas9 nuclease in lentiviral particles we observe that lentiviral vectors can be reprogramed to precisely integrate transgenic DNA.

[0651] Precise DNA integration was consistent in immortalized cell models and primary cells. PILV could also deliver a transgene in the ROSA26 locus in a mouse liver. We also detected PILV cis integration upon simultaneous delivery of two transgenes, providing a robust method for multiple KO-KI in ex vivo models. By changing two packaging vectors upon lentiviral production, lentiviral vectors can be made programmable enabling a targeted DNA addition with applications in ex vivo cell manufacturing, in vivo gene replacement, and mammalian genome engineering.