Multiple Genes Surrounding Bcl-x L,a Common Retroviral Insertion Site,Can Influence Hematopoiesis Individually or in Concert

Abstract

Retroviral insertional mutagenesis (RIM) is both a relevant risk in gene therapy and a powerful tool for identifying genes that enhance the competitiveness of repopulating hematopoietic stem and progenitor cells (HSPCs). However, focusing only on the gene closest to the retroviral vector insertion site (RVIS) may underestimate the effects of RIM, as dysregulation of distal and/or multiple genes by a single insertion event was reported in several studies. As a proof of concept, we examined the common insertion site (CIS) Bcl-x_L , which revealed seven genes located within ±150 kb from the RVIS for our study. We confirmed that Bcl-x_L enhanced the competitiveness of HSPCs, whereas the Bcl-x_L neighbor Id1 hindered HSPC long-term repopulation. This negative influence of Id1 could be counteracted by co-expressing Bcl-x_L . Interestingly, >90% of early reconstituted myeloid cells were found to originate from transduced HSPCs upon simultaneous overexpression of Bcl-x_L and Id1, which implies that Bcl-x_L and Id1 can collaborate to rapidly replenish the myeloid compartment under stress conditions. To directly compare the competitiveness of HSPCs conveyed by multiple transgenes, we developed a multiple competitor competitive repopulation (MCCR) assay to simultaneously screen effects on HSPC repopulating capacity in a single mouse. The MCCR assay revealed that multiple genes within a CIS can have positive or negative impact on hematopoiesis. Furthermore, these data highlight the importance of studying multiple genes located within the proximity of an insertion site to understand complex biological effects, especially as the number of gene therapy patients increases.

Introduction

Retroviruses and their derived vectors are well known to deregulate neighboring genes in the genome where the provirus is inserted, a phenomenon known as retroviral insertional mutagenesis (RIM). RIM has been shown to cause adverse effects in gene therapy settings^1

–4; it is also a robust tool to screen for genetic components involved in oncogenesis and stem cell biology.^5

–8 It was demonstrated that sequences of common insertion sites (CIS) recovered from RIM-induced dominant hematopoietic clones overlapped with known stem cell regulators⁷ and nonmalignant clonal dominance associated with vector integration site was observed in lentiviral gene therapy.^8,9

Moreover, upregulation of genes located >100 kb from the retroviral vector insertion site (RVIS)^2,10
–12 and dysregulation of multiple neighboring genes by single retroviral vector integrations were reported.^2,12,13 Nonetheless, the dysregulation of genes distal from the RVIS and the cooperative effects of multiple dysregulated genes within the proximity of an RVIS are often overlooked in published descriptions of RVIS. We hypothesized that more than one RVIS neighbor could collaboratively or individually influence hematopoiesis when upregulated by a single integration event.

As a proof-of-principle study, we explored the Bcl-x_L locus—a CIS annotated in the Insertional Dominance Database^7,14 and the Retroviral Tagged Cancer Gene Database.¹⁵ The importance of this locus in stem cell biology is highlighted by studies from the International Stem Cell Initiative, which demonstrated that a minimal amplicon duplication at chromosome 20q11.21 containing the ID1 and BCL-X_L genes confers competitive advantages to human pluripotent cells independent of culture conditions or cell origin from 38 laboratories worldwide.^16,17

Furthermore, both Bcl-x_L and Id1 are also known to regulate the stemness and differentiation of murine hematopoietic stem and progenitor cell (HSPC). BCL-X_L, encoded by the gene BCL2-like 1 (BCL2L1), is an antiapoptotic member of the Bcl2 family¹⁸ that has a highly homologous amino acid sequence and protein structure similarity with the better studied BCL2 protein.¹⁹ Both Bcl2 and Bcl-x_L were shown to enhance hematopoietic cell survival,²⁰ and their overexpression led to increased murine HSPC numbers and repopulating potential.^21

–25 ID1 (Inhibitor of DNA binding 1) is a helix-loop-helix (HLH) protein that forms heterodimers with the basic HLH transcription factors.^26,27 ID1 is expressed in hematopoietic stem cells (HSC),^28,29 and Id1 knock-out mice showed reduced HSPC numbers with impaired engraftment capacity.^28,29 Id1 also affected cell fate decisions in vivo when overexpressed in HSPCs, notably by impaired B cell production and promoted myeloid differentiation.^30

–33

In addition to Bcl-x_L and Id1, six other protein-coding genes (H13, Mcts2, Cox4i2, Tpx2, Mylk2, and Fkhl18) are located within ±150 kb of the Bcl-x_L -RVIS. To understand whether numerous genes can influence the competitiveness of HSPCs using conventional transplantation settings, individual genes of interest (GOI) have to be studied in different subsets of recipient mice, which requires high numbers of mice. As our interest was to measure the competitiveness of HSPCs conveyed by different transgenes, we developed a multiple competitor competitive repopulation (MCCR) assay to simultaneously screen multiple transgenes for effects on HSPC repopulating capacity in a single mouse. The MCCR assay allowed us to reduce the number of mice required to measure transgene effects on HSPC engraftment in vivo.

Surprisingly, we found that the overexpression of Id1 led to the reduced competitive output of Id1-HSPC after transplantation. However, coexpressing Bcl-x_L restored the competitive output of Id1-expressing HSPCs. Coexpressing Bcl-x_L and Id1 also contributed to rapid generation of myeloid cells shortly after transplantation. Our work presents an approach to understand complex insertional mutagenic effects in one genomic locus as we found that multiple genes surrounding a single common RVIS when upregulated can influence hematopoiesis individually or even collaboratively.

Materials and Methods

Lentiviral vectors production

The complementary DNAs (cDNAs) for H13, Id1, Cox4i2, Bcl-x_L , Tpx2, Mylk2, and Fkhl18 were each cloned into individual third-generation self-inactivating lentiviral vectors (LVs), in between an internal spleen focus-forming virus (SFFV) promoter and the internal ribosomal entry site, which preceded the enhanced green fluorescence protein (eGFP).³⁴ For eGFP vector control, a non-coding spacer sequence (Spcr; 243 bp) was inserted for tagging purposes. The vector coexpressing Bcl-x_L and Id1 (BI) was cloned by fusing the cDNA of Bcl-x_L and Id1 with porcine teschovirus-1 2A self-cleaving peptides (P2A; Fig. 1A). Lentiviral supernatants were produced, concentrated, and titrated as previously described.^35,36

Figure 1.

Id1-overexpressing HSPCs impaired B lineage. (A) Depiction of SIN-LV design of control vector (Spcr, spacer sequence between the promoter and IRES element), Id1-, Bcl-x_L-, and BI-cDNA containing vectors. All vectors express eGFP as reporter. (B) Transplantation experiment design. LSK cells were sorted and transduced with control-, Id1-, Bcl-x_L-, and BI-LVs before transplanting into lethally irradiated recipients at a cell dose of 2,000 LSK (input equivalent) and 2 × 10⁵ BM short-term supporting cells. (C) Transduction rate of LSK cells measured before transplantation. Open circles indicate data from experiment 1 and closed circles represent data from experiment 2. (D) Simplified gating strategy to analyze the lineage composition within the eGFP⁺ and eGFP⁻ PB fractions. (E) Percentages of myeloid (Gr-1⁺ and Mac-1⁺) and (F) B cells (B220⁺) within the PB of recipients 12 weeks after transplantation. Primary transplantation experiments were performed twice (eGFP: n = 7; Id1: n = 9; Bcl-x_L : n = 8; and BI: n = 10 mice). Open circles indicate data from experiment 1 and closed circles represent data from experiment 2. Symbols in gray represent the eGFP⁺ fraction and in black the eGFP⁻ fraction, medians are shown with solid lines for each data set (C, E, F). Mann–Whitney test results are shown with bars annotated with asterisks. Horizontal bars in black compare the groups in the eGFP⁻ fraction and horizontal bars in gray compare groups in the eGFP⁺ fraction. Vertical bars below the graphs compare the differences between eGFP⁺ and eGFP⁻ fractions within each group (i.e., Bcl-x_L+ vs. Bcl-x_L- ) (E, F). Not significant or p > 0.05 are not annotated; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; and p values adjusted with Benjamini–Hochberg methods. BM, bone marrow; BI, coexpression of both Bcl-x_L and Id1; cDNA, complementary DNA; eGFP, enhanced green fluorescence protein; FSC, forward scatter; HSPC, hematopoietic stem and progenitor cells; IRES, internal ribosomal entry site; LSK, Lineage⁻, Sca-1⁺, and c-Kit⁺; LTR, long terminal repeat; LV, lentiviral vector; MOI, multiplicity of infection; PB, peripheral blood; SFFV, spleen focus-forming virus; SIN, self-inactivating.

Cell sorting and flow cytometric analysis

LSK (Lineage⁻, Sca-1⁺, and c-Kit⁺) cells were sorted from bone marrow (BM) cells of C57BL/6J mice (Janvier Labs, Le Genest-Saint-Isle, France). Briefly, BM cells were stained with biotin anti-mouse lineage panel (Cat.133307; Biolegend, London, UK) and subsequently stained with FITC-Streptavidin (Cat.405201), PE-Sca-1 (Cat.108108) and APC-c-Kit (Cat.105812). To isolate B220⁺, CD3e⁺, and myeloid (Gr-1⁺, CD11b⁺) cells from recipient BM, PE-CD3e (Cat.100308), PE-Cy7-B220 (Cat.103222), APC-Gr-1 (Cat.108412), and APC-CD11b (Cat.101212) antibodies were used. Cells were stained with 1 μg/mL 4’,6-diamidino-2-phenylindole (Cat.D8417; Sigma-Aldrich) before sorting. To analyze leukocyte composition from peripheral blood (PB), PE-B220 (Cat.103208), PE-CD3e, APC-CD11b, and APC-Gr-1 antibodies and 1 μg/mL propidium iodide (Cat.81845; Sigma-Aldrich) were used. All BM and PB were first blocked with CD16/32 (Cat.14-0161-86; eBioscience, Frankfurt, Germany) antibody after NH₄Cl-mediated erythrocyte lysis. Samples were sorted with FACSAria II or analyzed on a FACSCalibur flow cytometer (Becton Dickinson, Heidelberg, Germany). All antibodies were purchased from Biolegend unless otherwise specified and were titrated before use.

Transduction of LSK HSPCs

After isolation and until the point of transplantation, LSK cells were cultured in HSPC expansion medium (StemSpan medium; Stem Cell Technologies, Vancouver, Canada) supplemented with STIF cocktail (50 ng/mL murine stem cell factor, 20 ng/mL murine thrombopoietin, 20 ng/mL murine insulin-like growth factor 2, and 10 ng/mL human fibroblast growth factor 1; all from Peprotech, Hamburg, Germany), 2% penicillin/streptomycin, and 1% glutamine (PAA, Pasching, Austria). Cells were prestimulated overnight before transduction with LV preloaded on RetroNectin (TaKaRa, Otsu, Japan)-coated culture plates.³⁷ Transduction efficiencies were estimated by measuring the percentage of eGFP⁺ cells 48 h after transduction of primary cells and before transplantation.

Transplantation and MCCR assay

C57BL/6J recipients were lethally irradiated (10 Gy) 1 day before transplantation. Each primary recipient received 2,000 LSK cells (input equivalent) after LV transduction along with 2 × 10⁵ BM cells. For secondary transplantation, BM cells from 2 to 4 primary recipients with a comparable level of eGFP marking were pooled, and 2 × 10⁶ BM cells were transplanted into each secondary recipient. For the MCCR assay, LSK cells were transduced with individual GOI-LV and individual GOI-HSPCs were pooled for transplantation. Each recipient received 8,000 or 16,000 (input equivalent; 2,000 per GOI-transduced group) pooled cells along with 2 × 10⁵ freshly isolated BM cells. In all transplantation settings, LV-tagged (transduced) and non-LV-tagged (nontransduced) cells for a given GOI-transduction procedure were transplanted as bulk without prior sorting for transduced cells. Drinking water containing 100 mg/mL ciprofloxacin (Bayer Vital GmbH, Leverkusen, Germany) was given to all recipients for 2 weeks following irradiation. PB was collected every 6 weeks after transplantation before mice were sacrificed during week 18. All animal experiments were approved by the local government and accomplished according to the rules of the central animal facility.

Vector copy number determination and population composition in the MCCR assay

Vector copy number (VCN) was determined by quantitative PCR (qPCR) as previously described.^36,38 In MCCR assays, transgene-specific primers (Supplementary Table S1) were used to differentiate individual GOI-transduced populations and normalized against the number of total LV integrants by measuring the DNA sequence of woodchuck hepatitis virus posttranscriptional element within the viral construct. qPCR was performed in technical triplicates using Quantitect SYBR Green Kit (Qiagen, Hilden, Germany) on a StepOnePlus System (Applied Biosystems, Foster City, CA).

INSPIIRED-PCR and Illumina sequencing

INSPIIRED-PCR data were processed as described in Sherman et al. ³⁹ Genomic DNA samples were resuspended in 120 μL nuclease-free water and subjected to shearing with Covaris 220 (Covaris Ltd., Brighton, UK) in AFA fiber microtubes (Covaris Ltd.) for 60 s (peak power of 50 W, 5% duty factor, 200 cycles/burst, and 4°C water temperature). Samples were purified with AMPure XP beads (Beckman Coulter GmbH, Krefeld, Germany) using a 0.7-fold bead to sample ratio. After using the NEBNext Ultra End Repair/dA-Tailing Module (New England Biolabs; NEB, Frankfurt am Main, Germany), linkers (linker blunt + sample-specific linker) were ligated with the NEBNext Ultra Ligation Module (NEB). Ligated samples were purified with AMPure XP beads (0.7-fold bead to sample ratio) and used in exponential PCR1 (20 cycles) using Clontech Advantage PCR kit (Takara Bio Europe SAS, Saint-Germain-en-Laye, France) with the addition of linker-specific primers (300 nM), long terminal repeats (LTR)-specific primer 1 (300 nM), blocking oligo (1 μM), dNTPs (200 nM), and linker-ligated DNA (0.6 v/v ratio). Products from PCR1 were used as a template for PCR2 (15 cycles) with a 0.08 v/v ratio to the complete sample volume. Specific index primers and sample-specific linker primers were used in all reactions (sequences available upon request). Products from PCR2 were subjected to quality control using 2% TAE-buffered agarose gel visualization and/or by Qubit Fluorometer (ThermoFisher, Foster City, CA) before being mixed in equal volumes to generate final libraries (six individual libraries were generated to cover all samples). The library was further purified with two rounds of AMPure XP beads (0.7-fold and 0.6-fold ratio of beads to sample volume) and transferred to the Hannover Medical School Research Core Unit Genomics (RCUG) for quality control using Bioanalyzer 2100 (Agilent Technologies Germany GmbH & Co. KG, Waldbronn, Germany) and analyzed on an Illumina sequencer (Illumina Inc., San Diego, CA).

LV insertion site analysis

Data from Illumina sequencing were processed as described by Berry et al.⁴⁰ Scripts required for INSPIIRED pipeline were acquired from GitHub. Sequence data in FASTQ format were demultiplexed based on the index primers used in PCR2. A quality score of 30, corresponding to an error probability of 0.001, was used as a quality threshold. A total of 5 bp below the quality threshold per amplicon was allowed and a quality sliding window of 10 bp was chosen. Processing parameters, such as: 1. Start of the alignment no later than five bp after the LTR; 2. a maximum fragment length of 2500 bp; and 3. a minimal overlap with reference genome of 20 bp with 95% homology, were set in the INSPRIIRED bioinformatics pipeline. Individual files were aligned and annotated to the mouse genome build mm9. Protein-coding genes located within ±150 kb from the integration site were compared with oncogene and stem cell-related gene (OSCG) databases. Oncogene databases used as references include the Network of Cancer Genes 6.0⁴¹ and “allOnco” cancer gene list 40 compiled by the Bushman group. For stem cell-related genes, we referred to the HSC-Explorer database⁴² and matched against genes that (1) increased or maintained self-renewal, (2) increased HSC quantity, (3) decreased HSC differentiation, (4) had increased expression in HSC compared to later progeny, and (5) were expressed in other stem cell types.

Statistics and plots

Graphs and statistical analyses were done with Prism 5 (GraphPad Software, San Diego, CA) using the Mann–Whitney test to compare medians unless otherwise specified, and Benjamini–Hochberg correction was applied for multiple comparisons.

Results

Id1 impaired B-lymphopoiesis with a bias toward myelopoiesis

Both Bcl-x_L and Id1 are known to influence HSPC fate. To mimic the upregulation of single or multiple genes in the context of RIM, we overexpressed Id1 and Bcl-x_L individually or simultaneously (BI) in HSPCs with LVs (Fig. 1A). A total of 2,000 LSK cells were cultured in serum-free HSPC expansion medium (StemSpan with STIF cytokine cocktail)⁴³ and transduced with Id1, Bcl-x_L , BI, or control LV at an multiplicity of infection of 10. Transduced cell populations were transplanted into lethally irradiated mice without prior sorting so that transplanted cell populations also contained nontransduced cells that were subjected to identical treatment as the transduced cells, with the exception that non-transduced cells were not influenced by the transgene and vector integration (Fig. 1B).

All vectors express eGFP as a reporter, and pretransplant transduction efficiencies were 10–30% (Fig. 1C). PB samples were analyzed every 6 weeks over the 18-week experiment. Overall, the numbers of white blood cells (WBCs), red blood cells, and platelets in the donors were similar and remained within the normal blood count range (Supplementary Fig. S1A). No bias in lineage distribution was found in most transplanted mice across different transplant groups (Supplementary Fig. S1B). However, comparison of transgene-positive (eGFP⁺) and transgene-negative (eGFP⁻) PB fractions (Fig. 1D) showed that Id1-expressing HSPCs were biased toward myeloid lineages (Fig. 1E) and had reduced B-lineage contribution (Fig. 1F) regardless of Bcl-x_L coexpression.

Bcl-x_L enhanced progeny output of HSPCs by producing more mature blood cells

To compare the competitiveness of the different transgene groups (ctrl, Id1, Bcl-x_L , and BI), the number of eGFP⁺ or eGFP⁻ WBCs in the PB of the recipients was normalized to the number of transplanted eGFP⁺ or eGFP⁻ LSK cells (Fig. 1C), and the amount of progeny generated by a single LSK cell in each recipient was compared among transgene-positive fractions (Fig. 2) and between transgene-positive and transgene-negative fractions in each recipient (Supplementary Fig. S2A–D). The Id1-LSK cells produced similar amounts of mature progeny as the eGFP control group, but fewer than the Bcl-x_L or BI groups (Fig. 2A–D), even though the transduction rate of Id1-LSK was twofold higher (comparable to Bcl-x_L and BI) than the vector control LSK. The output capacity of Id1 cells was gradually lost over time (Fig. 2A–C), while the nontransduced progeny output within the Id1 animals remained stable (Supplementary Fig. S2A–C).

Figure 2.

Progeny output from a single transplanted LSK cell. The number of WBCs in 1 μL of PB was divided by the number of transplanted LSK cells based on eGFP positivity. The resulting progeny output from a single transplanted GOI-LSK cell in 1 μL of PB obtained at (A) week 6, (B) week 12, and (C) week 18 after transplantation is shown. (D) Number of daughter LSK cells generated from transplanted GOI-positive LSK cells 18 weeks after transplantation. The numbers were obtained by dividing the number of LSK cells (from both hip bones, both femurs, and one tibia) with the number of transplanted LSK cells based on eGFP positivity. Primary transplantation experiments were performed twice (eGFP control: n = 7; Id1: n = 9; Bcl-x_L : n = 8; and BI: n = 10 mice). Open circles indicate data from experiment 1 and closed circles represent data from experiment 2. Mann–Whitney test was performed and medians are shown with solid lines for each data set. Benjamini–Hochberg adjusted p values are shown for all groups. GOI, genes of interest; WBCs, white blood cells.

After 18 weeks, only 3 (out of 9) Id1 primary recipient animals had transgene-positive cells that were above 3% in the PB (28%, 17%, and 32%; Supplementary Fig. S3A). A similar trend was also observed in the eGFP vector control group, where the vector-transduced LSK cells failed to generate progeny in several animals throughout the experiment (Fig. 2B, C; Supplementary Figs. S2 and S3). In contrast, all Bcl-x_L -overexpressing HSPCs exhibited enhanced progeny output compared to the nontransduced cells transplanted in the same animal (Supplementary Fig. S2A–C), regardless of Id1 coexpression, and outperformed both the eGFP control and Id1 alone (Fig. 2A–C). This competitive advantage can also be seen in the LSK-gated BM fractions (Fig. 2D; Supplementary Fig. S2D). Id1 expression in the BM of secondary transplanted animals was 7- to 12-fold higher in the eGFP⁺ fraction in the Id1-group, and 16- to 37-fold higher in the BI-group. Bcl-x_L expression was about 80-fold higher in BM from the Bcl-x_L group and 18- to 54-fold higher in BM from the BI group (Supplementary Fig. S3B).

Loss of Id1-HSPCs in recipient hematopoietic systems

As the transgene-positive cells were measured by surrogate eGFP expression, the decrease of eGFP⁺ cells in the Id1 and vector control groups can either be caused by the reduction of the transduced cell population or vector silencing. To analyze the presence of gene-modified cells, genomic DNA from recipient BM was subjected to qPCR to determine the VCN, and INSPIIRED-PCR followed by Illumina sequencing for vector insertion site (VIS) analyses. Both the VCN and the number of VIS recovered from the BM of Id1 primary and secondary recipients were significantly lower compared to Bcl-x_L or BI recipients (Fig. 3A–C). The VCN and VIS analyses suggested that the Id1-HSPC population diminished, but not that the expression cassette was silenced. Plotting the number of VIS recovered from BM of primary recipients from Fig. 3B against the transduction rate of transplanted LSK shown in Fig. 1C revealed that the number of VIS recovered in the Id1 group was low (Fig. 3D), even though the initial transduction rate was comparable to the Bcl-x_L and BI groups.

Figure 3.

VCN and retroviral insertion site analyses. (A) VCNs were determined by qPCR in BM of primary recipients. (B) Number of actual recovered VIS by INSPIIRED-PCR and Illumina sequencing from the BM of primary transplanted mice and (C) secondary transplanted mice. (D) The numbers of VIS recovered from primary BM depicted in (B) were plotted against the transduction rate of LSK cells at the time of transplantation as depicted in Fig. 1C. Percentage of VIS containing at least one OSCG within ±150 kb is plotted for (E) primary recipients and (F) secondary recipients. Fisher's exact test was performed among VIS retrieved from different groups, and the p values are shown in tables under figures (E, F). Medians are depicted with solid lines for each data set. Open circles in (A, B) represent data from experiment 1 and closed circles represent data from experiment 2. Not significant or p > 0.05 are not annotated; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001. OSCG, oncogene or stem cell-related gene; qPCR, quantitative PCR; VCN, vector copy number; VIS, vector insertion site.

Furthermore, agarose gel electrophoresis of integration site amplicons from LM-PCR suggested that many of the Id1-transduced cells were lost before the first sampling time (6 weeks after transplantation). Only low and decreasing numbers of individual amplicons were visible as discrete bands in PB samples over time in the Id1 animals (Supplementary Fig. S4, mouse #86, #88, and #90). In contrast, this was not the case for the vector control, Bcl-x_L , or BI groups.

Clonal restriction in secondary recipients

From all Id1-HSPC-transplanted animals, only three mice (#40, #41, #89) had an eGFP-positive cell population above 3% in the PB (Supplementary Fig. S3A). Analysis of the integration sites revealed oligoclonal distribution of LV-tagged clones in the BM of these animals (Supplementary Fig. S5A). When the BM of the three Id1 mice was transplanted into secondary recipients, 2–3 clones were overrepresented in all mice (Supplementary Fig. S5A). Integrations inside Tle4 (intron 8) and Celf2 (intron 5) accounted for 57.8 ± 6.4% and 34.0 ± 13.0% of the sequence pool in the BM of secondary animals. Transcripts of both genes were found to be altered in the eGFP⁺ BM fraction of the secondary animals (Supplementary Fig. S5B, C). Interestingly, even though the Id1 group had a low overall number of VIS, the percentage of VIS near OSCG is the highest in the Id1 group, and all except one Id1 LV insertion site (±150 kb) in secondary recipients overlapped with OSCG databases.^15,41,42 This degree of OSCG enrichment was only observed for the Id1 group (Fig. 3F).

Selection of individual clones in secondary recipients was also observed in the other treatment groups (Supplementary Fig. S6). For instance, in the eGFP control group, clones with integrations near Stim2 and Ddx60 were predominantly found across the secondary animals (Supplementary Fig. S6A). When BM of mice #48, #93, and #94 of the Bcl-x_L group was serially transplanted, two clones with either integrations, 13 kb downstream of Bcl2 or within Stag1, were found in all secondary recipients (Supplementary Fig. S6B). In another Bcl-x_L transplantation group, the Crisp1 clone was most prominent in most mice (Supplementary Fig. S6C). Similarly, in the BI-group, we performed three separate serial transplantation experiments and observed 3–5 prevalent clones in the secondary recipients (Supplementary Fig. S6D–F).

Coexpressing Bcl-x_L prevented Id1-mediated HSPC loss, did not release Id1 suppression of B cell development, and promoted short-term myelopoiesis

While the Id1-HSPCs were lost in recipients, coexpression with Bcl-x_L (BI) prevented this loss, increased the progeny output (Fig. 2; Supplementary Fig. S2), and supported long-term reconstitution in secondary recipients (Supplementary Fig. S3A). The expression of Bcl-x_L did not overcome the well-studied^30,33 Id1-mediated B-lineage suppression as depicted in Fig. 1F. The contribution of BI-HSPCs to the B lineage was lower compared with the nontransduced HSPCs (nt-HSPCs) within the BI group as well as both transgene-positive and transgene-negative fractions of the vector control group (Fig. 1F).

Moreover, coexpression of BI contributed toward a higher myeloid output (Fig. 1E), especially during short-term repopulation measured during week 6 after transplantation (Fig. 4). When gated on the Gr-1- and Mac-1-positive fraction on PB (Fig. 4A), over 90% of week 6 myeloid cells in primary recipients (Fig. 4B, C) and 77–98% in secondary recipients (Fig. 4C) were observed to be BI-transgene positive. Despite the high output from the BI-HSPCs, the overall WBC count (Supplementary Fig. S1A) and the myeloid count (granulocytes and monocytes; not shown) were normal, indicating the absence of abnormal myeloid expansion in these animals. Furthermore, the proportion of BI-myeloid and nontransduced myeloid cells equilibrated over time (Fig. 4C), suggesting that the high BI-myeloid content shortly after transplantation might be the result of rapid hematopoietic regeneration under transplantation stress.

Figure 4.

BI-HSPCs dominate the myeloid compartment during short-term engraftment. (A) Analyses were done by gating for the myeloid fraction in PB and quantifying the percentage of eGFP⁺ cells. (B) The percentage of eGFP⁺ myeloid cells in week 6 PB of primary recipient mice of each group. (C) The percentage of eGFP⁺ cells in the myeloid gated fraction of PB collected from the BI-group at different time points in both primary (1°) and secondary (2°) recipients. The different gray scales in secondary recipients indicate different transplanted groups. Open circles represent data from experiment 1 and closed circles represent data from experiment 2. Medians are depicted with solid lines for each data set. Data were subjected to the Mann–Whitney test, p values were adjusted based on Benjamini–Hochberg method and depicted as *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; not significant or p > 0.05 are not annotated.

Multiplexed in vivo competition assay reproduced observations from conventional transplantation experiments with fewer animals

To investigate whether dysregulation of other RVIS neighbors near Bcl-x_L might influence hematopoiesis, we cloned cDNAs of the other 5 genes located within ±150 kb from the RVIS (Supplementary Fig. S7A) into our LV backbone (Supplementary Fig. S7B). However, to verify all candidates in vivo, a significant number of recipient mice would be required to achieve statistical relevance. To comply with the “reduction” and “refinement” of the 3R principle, we designed a multiplexed transplantation assay that allowed us to perform MCCR. To this end, LSK HSPCs were transduced with individual GOI-LV before being pooled 2 days later for transplantation without prior sorting so that transplanted cell populations also contained nontransduced cells that were subjected to identical treatment as the transduced cells, with the exception that nontransduced cells were not influenced by the transgene and vector integration. The GOI-LV transduction rate was measured individually and as bulk using eGFP as a reporter. Since each transduced population was distinctly tagged by the transgene, the progress and outcome of the competition could be traced by LV-GOI-specific-primer sets (Supplementary Table S1) using qPCR on the genomic DNA.

Assuming that each cell had a single integration event, the copy numbers of individual GOI were then normalized by the number of transplanted GOI-tagged LSK cells to obtain the output capacity of individual GOI-LSK. In our earlier integration analysis, we noticed that clonal selection occurred in most animals transplanted with transduced HSPCs, including the eGFP vector control (Supplementary Figs. S5A and S6). Therefore, the eGFP⁻ population in the recipients was also normalized with the number of nt-HSPCs transplanted to serve as the baseline output capacity of nonmodified HSPCs. This calculation ignored the residual HSPCs in the recipients, which may lead to overestimation of the nt-HSPC output capacity, but should still serve as a more neutral reference than the eGFP⁺ population, which can be influenced by vector modification in terms of actual HSPC output. In the MCCR assay, the output capacity of nontransduced cells was used as a reference along with the vector control group.

To test if the MCCR assay yields similar results as conventional transplantation assays, an initial MCCR assay was performed with the same four populations analyzed above (Fig. 1A, B) by multiplexing the Bcl-x_L , Id1, BI, and eGFP vector-transduced HSPCs (Fig. 5A). The results of the MCCR assay 18 weeks after transplantation were similar to our observations from the conventional transplantation study (Figs. 1 and 2). After 18 weeks, the Id1-LSK cells contributed the least in total BM cells, while the Bcl-x_L -containing LSK (regardless of Id1 coexpression) had higher progeny output capacity compared to other LSK cells (Fig. 5B). Furthermore, B220⁺, CD3e⁺, and myeloid (Gr-1⁺ Mac-1⁺) cell populations in the recipient BM were sorted and the proportion of GOI-tagged cells was determined (Fig. 5C–E). The MCCR assay showed that Bcl-x_L -LSK had the highest output of B220-positive cells, while Id1- and BI-LSK had a lower contribution to the B220⁺ fraction (Fig. 5C) as observed earlier in the traditional transplantation experiment. This showed that the MCCR assay is capable of gene function studies in different lineages as well. In the myeloid fraction, Bcl-x_L - and BI-tagged cells showed higher output, whereas Id1-tagged cells had reduced output capacity (Fig. 5D), possibly due to overall reduction of Id1-tagged HSPCs that were depleted earlier. Results from the MCCR assay also suggested that Id1 expression reduced the output of CD3e⁺ cells as this population was repressed in Id1- and BI-tagged fractions (Fig. 5E).

Figure 5.

MCCR assay competing vector control, Bcl-x_L -, Id1-, and BI-transduced HSPC. (A) Experimental scheme describing the MCCR assay. LSK cells were sorted and transduced with individual vectors. Transduced populations were mixed before transplantation. Each mouse received 2,000 input equivalent to LSK per GOI and 2 × 10⁵ freshly isolated BM cells. Progeny output capacity of each GOI-HSPC from (B) whole BM, (C) B220⁺, (D) Gr-1⁺Mac-1⁺, and (E) CD3e⁺ sorted BM cells was compared to the eGFP control group and the output capacity of nt-HSPC. Gray dotted lines and black solid lines represent the minimum and maximum output capacity of eGFP control vector-transduced HSPCs and nt-HSPC, respectively. The regions in between are marked with two shades of gray, light gray marks the region covered by the vector control group and dark gray marks the region of nontransduced output. Two independent transplantation experiments were performed (n = 10 mice). Open circles indicate data from experiment 1 and closed circles represent data from experiment 2. Data were subjected to the Mann–Whitney test, p values were adjusted by Benjamini–Hochberg method. MCCR, multiple competitor competitive repopulation; ns, not significant; nt, nontransduced.

Multiple genes surrounding the CIS in the Bcl-x_L locus can influence hematopoiesis

As the MCCR assay recapitulated our earlier observations, we next investigated all eight GOIs in a larger-scale MCCR assay. Each GOI was transduced into individual HSPC cultures before pooling them for transplantation (Supplementary Fig. S7C). All recipients (n = 10) from this assay exhibited comparable percentages of eGFP⁺ cells in the PB (mean ± standard deviation of 28.7 ± 4.1%) and showed no sign of abnormality in blood counts throughout the 18-week competition period (Supplementary Fig. S7D, E). Transgene expression levels determined in the BM cells of experimental animals demonstrated 19,000- to 60,000-fold higher expression in the eGFP⁺ compared to the eGFP⁻ fraction (Supplementary Fig. S7F).

The MCCR assay was again shown to be consistent and reproducible as the reduced progeny output capacity of Id1-LSK cells and increased capacity of Bcl-x_L -LSK cells were also observed in this more complex setting (Fig. 6). The reduced competitive output of Id1-LSK was seen in several PB samples acquired at different time points, as well as in spleen and BM. Based on the output capacity of each GOI in PB (Fig. 6) and referenced to the vector control-LSK and nt-LSK output capacities, we categorized the applied transgenes into three different phenotypic groups: (1) the neutral group, which did not significantly influence hematopoiesis, including the vector control, H13, and Cox4i2; (2) the reduction group represented by Id1, Mylk2, and Fkhl18, which negatively influenced hematopoiesis; and (3) the enhancement group with Bcl-x_L and Tpx2 that increased the output capacity of HSPC.

Figure 6.

The progeny output capacity of HSPC was influenced by Bcl-x_L -surrounding genes. Output capacity of individual GOI-HSPCs and nt-HSPCs in PB from (A) week 6, (B) week 12, and (C) week 18 after transplantation. (D) Output capacity was also quantified in splenocytes (Spl) and (E) BM 18 weeks after transplantation. Output capacity of individual GOI-HSPC is color coded. Solid lines represent the maximum and minimum output capacity from nt-HSPCs (nt) and dotted lines show the maximum and minimum values (in yellow) of output capacity from vector control HSPCs (eGFP). Data were subjected to the Mann–Whitney test, p values were adjusted by Benjamini–Hochberg method. Color images are available online.

Discussion

As previously reported by others and supported by our own observations (Supplementary Fig. S8), a single retroviral integration can dysregulate multiple surrounding genes.^2,10

–13 However, combinatory effects of two or more genes surrounding the RVIS are insufficiently covered in the literature. In this study, we demonstrated that more than one gene in the 150 kb vicinity of an RVIS Bcl-x_L could influence hematopoiesis, and that cooperative interactions between neighboring genes may override the activity of single genes.

Under stress conditions (transplantation, myelodepletion, etc.), HSC rapidly enter into cell cycle to replenish the hematopoietic system and return to a quiescent state when homeostasis is reestablished. E47, the HLH transcription factor encoded by the E2A gene, is required to prevent HSC exhaustion during repopulating stress^44,45 and is essential for maintaining the integrity of HSC.^45,46 The ID1 protein, however, is a dominant negative inhibitor that prevents E47 from forming homodimers and, thus, prevents the transcription of downstream genes, including p21(Cdkn1a) and p27(Cdkn1b), which are crucial for HSC quiescence.^44

–49 It is reasonable to speculate that our Id1-HSPC were unable to establish quiescence under engraftment stress, which led to the observed loss of Id1-transduced cells. This is in line with a recent study by Singh et al., which demonstrated the increased expression of p21 and p27 in Id1 knockout HSC compared to wild-type HSC.⁵⁰ These Id1 ^−/− HSC also demonstrated enhanced reconstitution ability as they were able to reconstitute murine hematopoiesis up to seven serial transplantations, while the wild-type BM cells were depleted at the 4th serial transplantation.⁵⁰

Similar to E2A-knockout,^44
–46 Ku70-deficient HSC demonstrated loss of quiescence through downregulation of p27 and p57 expression, which leads to defects in self-renewal and competitive repopulation ability. Overexpression of Bcl2 restored quiescence and rescued HSC repopulation capacity in this setting.⁵¹ Like its family member Bcl2, Bcl-x_L can delay cell cycle entry or restore quiescence through p27 ^52,53 and p57(Cdkn1c)⁵¹ upregulation. Therefore, we speculate that coexpression of Bcl-x_L overrides the Id1-induced loss of HSPC by restoring quiescence through cell cycle regulatory pathways in BI-HSPC as Bcl2 does in Ku70 ^−/− HSC.⁵¹ It would be possible to further dissect the interaction among these proteins in their respective knockout or transgenic animals.^24,45,54

Our cell cycle analyses of the lineage negative BM populations isolated from primary recipients showed an increased G0/G1 population for Bcl-x_L - and BI-transduced cells (eGFP⁺) compared to their nontransduced (eGFP⁻) counterpart, which indicates that lineage-negative progenitor cells are more quiescent when Bcl-x_L is expressed. In contrast, Id1 ⁺ lineage-negative progenitor cell profiles were similar between transduced and nontransduced populations (Supplementary Fig. S9A, B). Furthermore, reverse transcription quantitative PCR (RT-qPCR) experiments to determine expression of cell cycle regulators (p21, p27, and p57) and indicator (Ki67) from eGFP⁺ and eGFP⁻ LSK cells sorted from the BM of primary recipients indicated lower p21 expression in Id1 LSK, with no significant changes in p27, p57, and Ki67 expression. In contrast, most Bcl-x_L and BI-LSK showed elevated p57 accompanied with decreased Ki67 expression, a signature of the quiescent state (Supplementary Fig. S9C-F; qPCR protocols described in Supplementary Data, primers are listed in Supplementary Table S2). As these Id1-LSK cells were harvested from animals with clones selected to avoid depletion, they might have been selected for features that enabled them to restore quiescence after stress hematopoiesis, such as replenishing the recipient's BM after transplantation. Therefore, 18 weeks after transplantation, these Id1-LSK cells showed no difference in ki67 expression level when compared to the eGFP⁻ LSK. Interestingly, the sample obtained from mouse #41 (Supplementary Fig. S9C, in red) expressed higher levels of cyclin inhibitors and lower Ki67 compared to samples from the other two Id1 animals. The Tle4 clone that later developed into the most prominent clone in all secondary animals was detected in this mouse (Supplementary Fig. S5A). The expression level of E2A, an upstream regulator of all Cdkn1, remained relatively stable in all transgene groups (Supplementary Fig. S9C–F).

Overexpression of Id1 is known to impair B cell development,³⁰ likely through binding with E2A proteins that are crucial for both HSC integrity and B cell development.^45,46,54,55 Coexpression of Bcl-x_L with Id1 restored the engraftment capability, but not the B-lineage impairment in BI-HSPC. This suggests that Bcl-x_L restored the engraftment capability of these cells through distinct pathways, which are independent of the Id1 and E47 interactions. Concurrently, Id1 is also a crucial gene for myelopoiesis and a downstream effector of the IL-3 signaling pathway. We observed myeloid bias in all Id1-HSPC-transplanted mice, regardless of Bcl-x_L coexpression (Figs. 1E and 5D), with the exception in Fig. 5D, in which the loss of Id1-derived myeloid cells may be due to overall loss of Id1 cells (Fig. 5B). Even here, the few mice that retained Id1 cells in BM (Fig. 5B) contributed mainly in the myeloid compartment (Fig. 5D) compared to the lymphoid compartment (Fig. 5C, E).

Unexpectedly, we observed that nearly all myeloid cells detected during early hematopoietic reconstitution originated from BI-HSPCs in BI-primary and BI-secondary recipients (Fig. 4B, C) without disruption of the myeloid compartment or induction of adverse effects (Supplementary Fig. S1). This suggests that Bcl-x_L and Id1 collaborate to replenish the myeloid compartment under stress conditions rapidly. This observation may have clinical relevance, for example, to foster more rapid myeloid recovery after transplantation in case of invasive fungal or bacterial infections by transiently elevating the expression of Id1 and Bcl-x_L with small molecules. BI coexpression also increased the output capacity in both mature and primitive compartments (Fig. 2).

Self-inactivating lentiviral vectors (SIN-LV) have not been observed to cause adverse effects in gene therapy patients to date, and the SIN configuration is considered to be safer in comparison to earlier LTR-driven vectors. However, disruption of neighboring genes by SIN-LV has been documented in gene therapy laboratory models and patients.^{8,9,11,56
–58} In our LV integration analysis, limited numbers of clones (<5) were found in primary transplanted animals of all transgene groups (Supplementary Figs. S5A and S6). Further clonal restriction was observed in all secondary animals as 2–3 clones prevailed in all mice within the same secondary transplant group (Supplementary Figs. S5A and S6). However, only the Id1-vector demonstrated a preference toward integrations near (±150 kb) OSCG in the secondary animals. Among all eight recovered VIS, only the insertion site near Oxct1 has no OSCG within the ±150 kb window. Yet, AW549877, a gene that matched the stem cells database,⁴² was found to be 2.5 kb outside our ±150 kb window. Of the two most prominent clones in all secondary animals, Tle4 is mainly described as a tumor suppressor.⁵⁹ Loss of Tle4 function confers proliferative advantage to AML1-ETO leukemic cells⁵⁹ and was described as one of the recurring somatic mutations in AML.⁶⁰ Celf2 is documented as a stem cell-related gene that is expressed higher in HSC compared to later progeny⁶¹ and is also listed in the Bushman oncogene list. Coincidentally, integration near Tle4 was found in dominant clones in two other lentiviral-based studies.^37,62

As we used a strong SFFV promoter/enhancer element in all constructs, clonal imbalance after serial transplantation can be expected. RT-qPCR on eGFP⁺ and eGFP⁻ sorted BM of Id1-secondary recipients revealed altered expression of Tle4 and Celf2, the two most prominent VIS in the Id1-group (Supplementary Fig. S5B, C). The dysregulation of the two genes might be associated with the prevalence of Id1-HSPC, but these observations require follow-up studies before definitive conclusions can be drawn. In the Bcl-x_L group, a clone with an integration near the proto-oncogene Bcl2 became readily apparent.

Considering the costs and welfare of experimental animals, in vitro assays⁶³ are sometimes preferred over transplantation experiments to study gene effects in HSPCs. While in vitro assays have some beneficial characteristics, misinterpretation of data due to artificial in vitro systems is a common limitation. For example, Id1-HSPC exhibited reduced progeny output in our experiments (Figs. 2, 5, and 6), but outcompeted the nt-HSPCs in our in vitro long-term cultures (Supplementary Fig. S10). This is not surprising as others have also demonstrated that Id1 immortalizes hematopoietic progenitors in vitro.³² This is also the reason that in vivo transplantation assays are still the gold standard for HSPC studies.

To overcome these challenges, and to comply with the “reduction” and “refinement” of the 3R principle, we developed the MCCR assay inspired by the competitive repopulation assay described by Harrison et al.⁶⁴ to screen multiple GOIs simultaneously in vivo using far fewer animals than would be required in conventional transplantation settings. For example, in our larger MCCR assay shown in Supplementary Fig. S7 and Fig. 6 and 10, recipients were transplanted with 8 transduced populations. In a conventional transplantation assay, at least 24 mice (3 mice per transduced population) would have been used for such a screening experiment. Thus, the MCCR increases animal welfare by reducing the number of experimental mice needed, which results in reduced animal costs per experiment as well as the required hands on time for animal experiments.

In our study, the MCCR assay not only recapitulated our conventional transplantation experiments by showing the effects of Id1, Bcl-x_L , and BI on HSPC (Fig. 5), but also allowed us to classify our GOIs in the Bcl-x_L locus into three categories (Fig. 6): (1) the neutral group (vector control, H13, and Cox4i2); (2) the reduction group (Id1, Mylk2, and Fkhl18); and (3) the enhancement group (Bcl-x_L and Tpx2). The assay can also be employed to study any desired lineage and screen for multiple genetic elements (cDNAs of different genes, cDNA of different splice variants of a gene, microRNA, long non-coding RNA, circular RNA, etc.) that influence lineage development in the native in vivo microenvironment (Fig. 5C–E). For example, we detected Id1-induced B cell suppression in BI-HSPCs when the B220⁺ fraction from BM was sorted before performing GOI-specific qPCR (Fig. 5C). Furthermore, in combination with integrating DNA barcode vectors,^62,65 the MCCR has the potential to be used to study the impact of pretransplantation manipulation (e.g., different culture conditions or handling) on hematopoietic cells, thus not only reducing the required numbers of recipient animals, but also the likelihood of in vitro artifacts.

Our MCCR assay identified Bcl-x_L and Tpx2 as genes that can confer improved competitive output to HSPCs. While Bcl-x_L was previously shown to increase HSPC fitness,^21,22 Tpx2 has not been described in this context. Tpx2 is vital for mitosis and spindle assembly,⁶⁶ is highly expressed in a wide variety of tumors,⁶⁷ and is a collaborating partner that stabilizes the proto-oncogene Aurora A kinase.⁶⁸ Further experiments should be conducted to understand the impact of Tpx2 on HSPCs.

Like with any screening method, there are limitations to the MCCR. The most obvious limitation is that the VCN of individual GOI-LV in each transduced cell should be equal, and ideally, a single copy in each cell is desired, especially when barcoded vectors are used. In our study, we tried to control the VCN by limiting the transduction rate. Based on several earlier publications,^69
–71 at the highest transduction rate achieved in our experiments (data not shown; 40% in both H13 and Cox4i2), >70% of our transduced cells are expected to have single integrations. However, this should be taken only as a reference, since selection might favor or be biased against certain GOIs depending on the VCN (expression level) at the end of the competition. Although this can be further addressed, for example, by measuring the VCN of individual clones grown in semisolid culture, confirmatory experiments are recommended as for any other screening method.

Furthermore, selection based on vector integration site can occur, including the cells modified with the control vector. Retrospectively, this could have been overcome by using a three population congenic model utilizing the CD45.1, CD45.2, and CD45.1/CD45.2 hybrid, to distinguish GOI-transduced, nontransduced, and recipient-generated progenies. In the MCCR assays presented in this study, we have only tested the overexpression aspect of RIM dysregulation and have not tested other possible genotoxic outcomes, such as vector-induced truncation of genes, altered splice forms, gene fusion or gene inactivation. We have also tested only one combinatory event with the Bcl-x_L and Id1 coexpression. Combinatory events should be tested in the future to understand the impact of the genes surrounding a retroviral CIS as well as potential interactions among gene products that may influence cell behavior. In addition, systematic evaluation of individual transgene expression levels in the MCCR-transplanted animals could offer more insight, such as the selection of transduced populations based on transgene expression levels. With appropriate MCCR design, these questions can be answered with fewer experimental animals.

In summary, the pilot MCCR assay developed in this study can aid future studies to adhere to the 3R principles by reducing the number of experimental animals when screening activities of multiple genes in vivo. Furthermore, we have demonstrated that more than one gene within an RVIS can influence hematopoiesis individually or cooperatively when overexpressed, and that the cooperative interactions between neighboring genes may override the activity of single genes. Therefore, to better understand RIM in clinical cases and to avoid neglecting potential candidates when searching for novel OSCG, it is important to include genes and elements surrounding the VIS of interest as study subjects.

Footnotes

Acknowledgments

The authors would like to thank Tobias Maetzig for his critical input to this study, Matthias Ballmaier (Sorter Lab, Hannover Medical School) for cell sorting procedures, Jörg Frühauf (Institute of Radiology, Hannover Medical School) for assistance with the mouse irradiation procedure, Rena-Mareike Gastmeier (Institute of Experimental Hematology, Hannover Medical School) and staffs at the animal facility of Hannover Medical School for mouse visiting and caring, and the team at Hannover Medical School RCUG for the processing of samples for Illumina sequencing. We are also grateful to Peter Andrews (Department of Biomedical Science, The University of Sheffield) for critical input on stem cell biology and constructive suggestions related to this work.

Authors' Contributions

T.-C.H., O.K., B.F., A.S., and C.B. designed the study. T.-C.H., M.S., M.R., G.P., V.D., and O.K. performed the experiments and collected data. T.-C.H., M.R., M.M., and M.B. analyzed and interpreted data. T.-C.H., M.M., M.R., O.K., A.S., and C.B. wrote the article.

Author Disclosure

The authors declare no conflicts of interest.

Funding Information

This project has received funding from the European Union's Horizon 2020 research and innovation program under grant agreement No. 666908, the Deutsche Forschungsgemeinschaft [DFG; through SPP1230, SFB738, Excellence Cluster REBIRTH (EXC62/2), FE568/11-1, RO 5102/1-1], the German Federal Ministry of Education and Research (BMBF; through PIDNET and IFB-Tx), and the Federal State of Lower Saxony (through research project R2N).

Supplementary Material

Supplementary Data

Supplementary Table S1

Supplementary Table S2

Supplementary Figure S1

Supplementary Figure S2

Supplementary Figure S3

Supplementary Figure S4

Supplementary Figure S5

Supplementary Figure S6

Supplementary Figure S7

Supplementary Figure S8

Supplementary Figure S9

Supplementary Figure S10

References

Hacein-Bey-Abina

, Von Kalle

, Schmidt

, et al. LMO2-associated clonal T cell proliferation in two patients after gene therapy for SCID-X1. Science, 2003; 302:415–419.

Howe

, Mansour

, Schwarzwaelder

, et al. Insertional mutagenesis combined with acquired somatic mutations causes leukemogenesis following gene therapy of SCID-X1 patients. J Clin Invest, 2008; 118:3143–3150.

Stein

, Ott

, Schultze-Strasser

, et al. Genomic instability and myelodysplasia with monosomy 7 consequent to EVI1 activation after gene therapy for chronic granulomatous disease. Nat Med, 2010; 16:198–204.

Braun

, Boztug

, Paruzynski

, et al. Gene therapy for Wiskott-Aldrich syndrome—long-term efficacy and genotoxicity. Sci Transl Med, 2014; 6:227ra33–227ra33.

Selten

, Cuypers

, Zijlstra

, et al. Involvement of c-myc in MuLV-induced T cell lymphomas in mice: frequency and mechanisms of activation. EMBO J, 1984; 3:3215–3222.

. Cooperating cancer-gene identification through oncogenic-retrovirus-induced insertional mutagenesis. Blood, 2005; 106:2498–2505.

Kustikova

, Geiger

, Li

, et al. Retroviral vector insertion sites associated with dominant hematopoietic clones mark “stemness” pathways. Blood, 2007; 109:1897–1907.

Cavazzana-Calvo

, Payen

, Negre

, et al. Transfusion independence and HMGA2 activation after gene therapy of human β-thalassaemia. Nature, 2010; 467:318–322.

Fraietta

, Nobles

, Sammons

, et al. Disruption of TET2 promotes the therapeutic efficacy of CD19-targeted T cells. Nature, 2018; 558:307–312.

10.

Kustikova

, Fehse

, Modlich

, et al. Clonal dominance of hematopoietic stem cells triggered by retroviral gene marking. Science, 2005; 308:1171–1174.

11.

Singhal

, Deng

, Chenchik

, et al. Long-distance effects of insertional mutagenesis. PLoS One, 2011; 6:e15832.

12.

Sauvageau

, Miller

, Lemieux

, et al. Quantitative expression profiling guided by common retroviral insertion sites reveals novel and cell type specific cancer genes in leukemia. Blood, 2008; 111:790–799.

13.

Tsichlis

, Lee

, Bear

, et al. Activation of multiple genes by provirus integration in the Mlvi-4 locus in T-cell lymphomas induced by Moloney murine leukemia virus. J Virol, 1990; 64:2236–2244.

14.

von Keudell

, Cornils

, Badbaran

, et al. Clonal dominance after reconstitution of the haematopoietic system with bone marrow cells retroviral. In: Bilko

, Fehse

, Ostertag

, et al., eds. Stem Cells and their Potential for Clinical Application. Dordrecht: Springer Netherlands, 2008, 1–15.

15.

Akagi

, Suzuki

, Stephens

, et al. RTCGD: retroviral tagged cancer gene database. Nucleic Acids Res, 2004; 32:D523–D527.

16.

Närvä

, Autio

, Rahkonen

, et al. High-resolution DNA analysis of human embryonic stem cell lines reveals culture-induced copy number changes and loss of heterozygosity. Nat Biotechnol, 2010; 28:371–377.

17.

Amps

, Andrews

, Anyfantis

, et al. Screening ethnically diverse human embryonic stem cells identifies a chromosome 20 minimal amplicon conferring growth advantage. Nat Biotechnol, 2011; 29:1132–1144.

18.

Boise

, González-García

, Postema

, et al. bcl-x, a bcl-2-related gene that functions as a dominant regulator of apoptotic cell death. Cell, 1993; 74:597–608.

19.

Petros

, Olejniczak

, Fesik

. Structural biology of the Bcl-2 family of proteins. Biochim Biophys Acta Mol Cell Res, 2004; 1644:83–94.

20.

Chao

, Linette

, Boise

, et al. Bcl-XL and Bcl-2 repress a common pathway of cell death. J Exp Med, 1995; 182:821–828.

21.

Domen

, Cheshier

, Weissman

. The role of apoptosis in the regulation of hematopoietic stem cells: overexpression of BCL-2 increases both their number and repopulation potential. J Exp Med, 2000; 191:253–264.

22.

Mata

, Chiffoleau

, Adler

, et al. Bcl-XL expression in stem cells facilitates engraftment and reduces the need for host conditioning during bone marrow transplantation. Am J Transplant, 2004; 4:58–64.

23.

Domen

, Weissman

. Hematopoietic stem cells need two signals to prevent apoptosis; BCL-2 can provide one of these, Kitl/c-Kit signaling the other. J Exp Med, 2000; 192:1707–1718.

24.

Domen

, Gandy

, Weissman

. Systemic overexpression of BCL-2 in the hematopoietic system protects transgenic mice from the consequences of lethal irradiation. Blood, 1998; 91:2272–2282.

25.

Suzuki

, Yamazaki

, Ema

, et al. Homeostasis of hematopoietic stem cells regulated by the myeloproliferative disease associated-gene product Lnk/Sh2b3 via Bcl-xL. Exp Hematol 2012;40:166.e3–174.e3.

26.

Benezra

, Davis

, Lockshon

, et al. The protein Id: a negative regulator of helix-loop-helix DNA binding proteins. Cell, 1990; 61:49–59.

27.

Sun

, Copeland

, Jenkins

, et al. Id proteins Id1 and Id2 selectively inhibit DNA binding by one class of helix-loop-helix proteins. Mol Cell Biol, 1991; 11:5603–5611.

28.

Perry

, Zhao

, Nie

, et al. Id1, but not Id3, directs long-term repopulating hematopoietic stem-cell maintenance. Blood, 2007; 110:2351–2360.

29.

Jankovic

, Ciarrocchi

, Boccuni

, et al. Id1 restrains myeloid commitment, maintaining the self-renewal capacity of hematopoietic stem cells. Proc Natl Acad Sci U S A, 2007; 104:1260–1265.

30.

Sun

. Constitutive expression of the Id1 gene impairs mouse B cell development. Cell, 1994; 79:893–900.

31.

Leeanansaksiri

, Wang

, Gooya

, et al. IL-3 induces inhibitor of DNA-binding protein-1 in hemopoietic progenitor cells and promotes myeloid cell development. J Immunol, 2005; 174:7014–7021.

32.

Suh

, Leeanansaksiri

, Ji

, et al. Id1 immortalizes hematopoietic progenitors in vitro and promotes a myeloproliferative disease in vivo. Oncogene, 2008; 27:5612–5623.

33.

Cochrane

, Zhao

, Welner

, et al. Balance between Id and E proteins regulates myeloid-versus-lymphoid lineage decisions. Blood, 2009; 113:1016–1026.

34.

Dull

, Zufferey

, Kelly

, et al. A third-generation lentivirus vector with a conditional packaging system. J Virol, 1998; 72:8463–8471.

35.

Schambach

, Galla

, Modlich

, et al. Lentiviral vectors pseudotyped with murine ecotropic envelope: increased biosafety and convenience in preclinical research. Exp Hematol, 2006; 34:588–592.

36.

Maetzig

, Kuehle

, Schwarzer

, et al. All-in-One inducible lentiviral vector systems based on drug controlled FLP recombinase. Biomaterials, 2014; 35:4345–4356.

37.

Kustikova

, Stahlhut

, Ha

T-C

, et al. Dose response and clonal variability of lentiviral tetracycline-regulated vectors in murine hematopoietic cells. Exp Hematol 2014;42:505.e7–515.e7.

38.

Heinz

, Schambach

, Galla

, et al. Retroviral and transposon-based tet-regulated all-in-one vectors with reduced background expression and improved dynamic range. Hum Gene Ther, 2011; 22:166–176.

39.

Sherman

, Nobles

, Berry

, et al. INSPIIRED: a pipeline for quantitative analysis of sites of new DNA integration in cellular genomes. Mol Ther Methods Clin Dev, 2017; 4:39–49.

40.

Berry

, Nobles

, Six

, et al. INSPIIRED: quantification and visualization tools for analyzing integration site distributions. Mol Ther Methods Clin Dev, 2017; 4:17–26.

41.

, Dall'Olio

, Mourikis

, et al. NCG 5.0: updates of a manually curated repository of cancer genes and associated properties from cancer mutational screenings. Nucleic Acids Res, 2016; 44:D992–D999.

42.

Montrone

, Kokkaliaris

, Loeffler

, et al. HSC-explorer: a curated database for hematopoietic stem cells. PLoS One, 2013; 8:e70348.

43.

Zhang

, Lodish

. Murine hematopoietic stem cells change their surface phenotype during ex vivo expansion. Blood, 2005; 105:4314–4320.

44.

Yang

, Kardava

, St Leger

, et al. E47 controls the developmental integrity and cell cycle quiescence of multipotential hematopoietic progenitors. J Immunol, 2008; 181:5885–5894.

45.

Santos

, Ding

, Borghesi

. Cell-intrinsic in vivo requirement for the E47-p21 pathway in long-term hematopoietic stem cells. J Immunol, 2014; 192:160–168.

46.

Semerad

, Mercer

, Inlay

, et al. E2A proteins maintain the hematopoietic stem cell pool and promote the maturation of myelolymphoid and myeloerythroid progenitors. Proc Natl Acad Sci USA, 2009; 106:1930–1935.

47.

Prabhu

, Ignatova a, Park

, et al. Regulation of the expression of cyclin-dependent kinase inhibitor p21 by E2A and Id proteins. Mol Cell Biol, 1997; 17:5888–5896.

48.

Yang

, Esplin

, Borghesi

. E47 regulates hematopoietic stem cell proliferation and energetics but not myeloid lineage restriction. Blood, 2011; 117:3529–3538.

49.

Cheng

. Hematopoietic stem cell quiescence maintained by p21cip1/waf1. Science (80-), 2000; 287:1804–1808.

50.

Singh

, Singh

, Gadomski

, et al. Id1 ablation protects hematopoietic stem cells from stress-induced exhaustion and aging. Cell Stem Cell 2018;23:252.e8–265.e8.

51.

Qing

, Wang

, Bunting

, et al. Bcl2 overexpression rescues the hematopoietic stem cell defects in Ku70-deficient mice by restoration of quiescence. Blood, 2014; 123:1002–1011.

52.

Greider

, Chattopadhyay

, Parkhurst

, et al. BCL-x(L) and BCL2 delay Myc-induced cell cycle entry through elevation of p27 and inhibition of G1 cyclin-dependent kinases. Oncogene, 2002; 21:7765–7775.

53.

Janumyan

, Cui

, Yan

, et al. G0 function of BCL2 and BCL-xL requires BAX, BAK, and p27 phosphorylation by Mirk, revealing a novel role of BAX and BAK in quiescence regulation. J Biol Chem, 2008; 283:34108–34120.

54.

Bain

, Maandag

, Izon

, et al. E2A proteins are required for proper B cell development and initiation of immunoglobulin gene rearrangements. Cell, 1994; 79:885–892.

55.

Zhuang

, Soriano

, Weintraub

. The helix-loop-helix gene E2A is required for B cell formation. Cell, 1994; 79:875–884.

56.

Cesana

, Ranzani

, Volpin

, et al. Uncovering and dissecting the genotoxicity of self-inactivating lentiviral vectors in vivo. Mol Ther, 2014; 22:774–785.

57.

Bokhoven

, Stephen

, Knight

, et al. Insertional gene activation by lentiviral and gammaretroviral vectors. J Virol, 2009; 83:283–294.

58.

Ranzani

, Cesana

, Bartholomae

, et al. Lentiviral vector-based insertional mutagenesis identifies genes associated with liver cancer. Nat Methods, 2013; 10:155–161.

59.

Shin

, Brynczka

, Dayyani

, et al. TLE4 regulation of wnt-mediated inflammation underlies its role as a tumor suppressor in myeloid leukemia. Leuk Res, 2016; 48:46–56.

60.

Greif

, Eck

, Konstandin

, et al. Identification of recurring tumor-specific somatic mutations in acute myeloid leukemia by transcriptome sequencing. Leukemia, 2011; 25:821–827.

61.

Chambers

, Boles

, Lin

K-YK

, et al. Hematopoietic fingerprints: an expression database of stem cells and their progeny. Cell Stem Cell, 2007; 1:578–591.

62.

Cornils

, Thielecke

, Hüser

, et al. Multiplexing clonality: combining RGB marking and genetic barcoding. Nucleic Acids Res, 2014; 42:e56.

63.

van Os

, Kamminga

, de Haan

. Stem cell assays: something old, something new, something borrowed. Stem Cells, 2004; 22:1181–1190.

64.

Harrison

. Competitive repopulation: a new assay for long-term stem cell functional capacity. Blood, 1980; 55:77–81.

65.

, Neff

, Quake

, et al. Tracking single hematopoietic stem cells in vivo using high-throughput sequencing in conjunction with viral genetic barcoding. Nat Biotechnol, 2011; 29:1–7.

66.

Boleti

, Karsenti

, Vernos

. Xklp2, a novel xenopus centrosomal kinesin-like protein required for centrosome separation during mitosis. Cell, 1996; 84:49–59.

67.

Neumayer

, Belzil

, Gruss

, et al. TPX2: of spindle assembly, DNA damage response, and cancer. Cell Mol Life Sci, 2014; 71:3027–3047.

68.

Giubettini

, Asteriti

, Scrofani

, et al. Control of Aurora-A stability through interaction with TPX2. J Cell Sci, 2011; 124:113–122.

69.

Kustikova

, Wahlers

, Ku

, et al. Dose finding with retroviral vectors: correlation of retroviral vector copy numbers in single cells with gene transfer efficiency in a cell population. Blood, 2003; 102:3934–3937.

70.

Fehse

, Kustikova

, Bubenheim

, et al. Pois(s)on—It's a Question of Dose…. Gene Ther, 2004; 11:879–881.

71.

Charrier

, Ferrand

, Zerbato

, et al. Quantification of lentiviral vector copy numbers in individual hematopoietic colony-forming cells shows vector dose-dependent effects on the frequency and level of transduction. Gene Ther, 2011; 18:479–487.

Multiple Genes Surrounding Bcl-x L,a Common Retroviral Insertion Site,Can Influence Hematopoiesis Individually or in Concert

Abstract

Introduction

Materials and Methods

Lentiviral vectors production

Cell sorting and flow cytometric analysis

Transduction of LSK HSPCs

Transplantation and MCCR assay

Vector copy number determination and population composition in the MCCR assay

INSPIIRED-PCR and Illumina sequencing

LV insertion site analysis

Statistics and plots

Results

Id1 impaired B-lymphopoiesis with a bias toward myelopoiesis

Bcl-xL enhanced progeny output of HSPCs by producing more mature blood cells

Loss of Id1-HSPCs in recipient hematopoietic systems

Clonal restriction in secondary recipients

Coexpressing Bcl-xL prevented Id1-mediated HSPC loss, did not release Id1 suppression of B cell development, and promoted short-term myelopoiesis

Multiplexed in vivo competition assay reproduced observations from conventional transplantation experiments with fewer animals

Multiple genes surrounding the CIS in the Bcl-xL locus can influence hematopoiesis

Discussion

Footnotes

Acknowledgments

Authors' Contributions

Author Disclosure

Funding Information

Supplementary Material

References

Bcl-x_L enhanced progeny output of HSPCs by producing more mature blood cells

Coexpressing Bcl-x_L prevented Id1-mediated HSPC loss, did not release Id1 suppression of B cell development, and promoted short-term myelopoiesis

Multiple genes surrounding the CIS in the Bcl-x_L locus can influence hematopoiesis