Cells
HCT116 colorectal human cells (American Type Culture Collection—ATCC, Manassas, VA) which are near diploid with 45 median chromosome count were cultured at 37 °C and 5% CO2 in complete DMEM medium (DMEM with 4.5 g/L glucose (Gibco, Waltham, MA) supplemented with 10% (v/v) fetal calf serum (Eurobio, Les Ullis, France), 1% (vol./vol.) GlutaMAX® (Gibco) and 1% (v/v) penicillin/streptomycin (Gibco)). A subculture of HEK293T human embryonic kidney cells was derived at Genethon and used to produce LV for research or for clinical use as previously reported [4]. Such cells were derived from a working cell bank which was authenticated and tested negative for mycoplasma.
Lentiviral vector production and titration
A VSV-G pseudotyped self-inactivating lentiviral vector coding for the truncated human nerve growth factor receptor (dNGFR/CD271) under the control of a human phosphoglycerate kinase (PGK) promoter was produced by transient transfection of 293T cells using calcium phosphate and 4 plasmids (pRRL-hPGK-dNGFR-WPRE, HIV-1 gagpol, HIV-Rev, and VSV-G plasmids). The harvested particles were concentrated about 500 fold by ultracentrifugation (50,000 × g, 2 h, 12 °C) suspended in phosphate buffered saline and cryopreserved at −80 °C. The infectious titer of the vector, determined on HCT116 cells using qPCR as previously reported [9], was 2.1E9 infectious genome (IG)/mL.
Magnetic cell sorting and flow cytometry
Cells expressing the dNGFR transgene were positively selected by CD271 magnetic bead cell sorting (Miltenyi Biotec, Bergisch Gladbach Germany) using manufacturer’s instructions. Transgene expression was measured by flow cytometry using the LSR II cytometer (BD Biosciences Franklin Lakes, NJ), anti-CD271 antibodies (CD271 NGFR APC # 130-110-080, Miltenyi Biotec) and 7AAD (Sigma-Aldrich, St. Louis, MO) to exclude dead cells.
DNA extraction and PCR methods
DNA was extracted using the Wizard Genomic DNA purification kit (Promega, Madison, WI USA) according to the manufacturer recommendations. Vector copy number (VCN) was measured by quantitative real time PCR (qPCR) or by digital droplet PCR (ddPCR). The names of the primers and probes are indicated in italics and the sequences are listed in Supplementary Table S1. Amplification of the human ALB gene with Alb.fw, Alb.rv and Alb.pr was used to determine the number of diploid genome. The amplification of the vector HIV-1 PSI sequence with PSI.fw, PSI.rv, and PSI.pr or LTR- sequence with PRO.fw, PRO.rv, PRO.pr were respectively used to determine the vector copies. For qPCR, the reaction was performed using 100 ng of gDNA per reaction using the Light Cycler 480 device (Roche Life Science Penzberg, Germany) according to the manufacturer recommendations (ABsolute qPCR ROX Mix, Thermo Scientific, Waltham, MA USA). A qPCR standard curve was used to convert Ct values to copy numbers by amplifying serial dilutions of a plasmid containing equimolar ratios of the PSI and ALB targeted sequences. For ddPCR, the reaction was performed on the Biorad system (Biorad QX200 autoDG and PCR systems Hercules, CA USA) according to the manufacturer recommendations using 3 units of the HaeIII restriction enzyme (New England Biolabs (NEB), Ipswich MA USA) in the mix (Biorad ddPCR Supermix for Probes (No dUTP) and 40 ng of gDNA per reaction unless indicated otherwise.
Vector IS analysis
Vector IS were identified using a Linker-Mediated PCR (LM-PCR) technique with paired-end Illumina sequencing to amplify the junction between the integrated provirus 3′LTR and the genomic DNA at IS. The sequences of primers and oligonucleotides used are found in Supplementary Table S1. The gDNA (500 ng in 50 µl) was sonicated for 5 cycles of 15″ON/30″OFF (Diagenode, Bioruptor Pico, Liege, Belgium) to obtain a mean fragment size of 600 bp. Fragmented DNA was then end-repaired and a protruding 3′ A was added following the manufacturer recommendations (NEB). The linker was assembled by mixing oligos Linker+ and Linker− to a final concentration of 20 µM in 10 mM Tris, pH 7.5–8.0, 50 mM NaCl and 1 mM EDTA. The mix was heated at 95 °C for 5 min and slowly cooled down to room temperature at a rate of 1.5 °C/min and stored at −20 °C. The double stranded linker was ligated to the repaired fragmented DNA with a molar ratio of 5:1 (NEB) and the ligation product purified using spin columns using a 5:1 buffer dilution to eliminate fragments smaller than 200 bp (Macherey Nagel, Dueren Germany). A first PCR was performed for 30 cycles to amplify vector/genome junctions with primers VISA1.vector and VISA1.linker (TM = 64 °C) using 4 µl of DNA in 25 µl reaction volume (NEB). A Blocking oligo (10 µM) containing bridged nucleic acid (BNA) bases was used to suppress the generation of unwanted sequences elongated from the 5′LTR end of the provirus, thereby increasing the number of reads containing the vector/junction from the 3′LTR as described [15]. After purification, 2 µl of DNA was amplified by 20 cycles of nested PCR in order to add Illumina NEXTERA adaptors and 4 bp barcodes using primers VISA2.TAGx.vector, VISA2.linker, and 10 µM of the blocking oligo. After PCR, amplicons ranging from 400 to 1 kb were gel purified and sequenced using either Sanger sequencing (Genewiz, Leipzig Germany) after cloning of PCR products in Top10 E.coli (Invitrogen, Whaltham, MA) or MiSeq Illumina paired-end sequencing (301 bp, 2E6 reads, IGATech, Italy). For the PCR verification of the vector/genome junction, we used specific primers recognizing gDNA sequences around the identified IS and the VISA1.vector primer with the OneTaq DNA polymerase (NEB). Amplicon were size-selected on gel and sequenced by Sanger sequencing using the VISA2.vector primer (Genewiz, Germany).
Data processing and statistics
To identify IS from sequencing data, raw reads from Sanger sequencing or from Illumina sequencing were first demultiplexed (cutadapt v2.2) based on the sample TAG added during the nested PCR and both the LTR and the linker sequences were trimmed (cutadapt v2.2). Resulting sequences longer than 20 bp and not containing the proviral or plasmid sequences were aligned on the reference genome (hg19, bowtie2 v2.3.4.2) and insertion point inferred from the alignment coordinates. Reads mapping to the same IS were aggregated and the number of different sonication fragments was estimated using the sonic-Length method [16]. Each IS was then annotated using the gencode database release 19. The availability of the code used for this study is subject to restrictions.
Data processing, statistics and figures were produced using the R software 3.6.0. Information about sample size and statistical tests are found in the figure legends.
