Preclinical demonstration of lentiviral vector-mediated correction of immunological and metabolic abnormalities in models of adenosine deaminase deficiency

Abstract

Gene transfer into autologous hematopoietic stem cells by γ-retroviral vectors (gRV) is an effective treatment for adenosine deaminase (ADA)-deficient severe combined immunodeficiency (SCID). However, current gRV have significant potential for insertional mutagenesis as reported in clinical trials for other primary immunodeficiencies. To improve the efficacy and safety of ADA-SCID gene therapy (GT), we generated a self-inactivating lentiviral vector (LV) with a codon-optimized human cADA gene under the control of the short form elongation factor-1α promoter (LV EFS ADA). In ADA(-/-) mice, LV EFS ADA displayed high-efficiency gene transfer and sufficient ADA expression to rescue ADA(-/-) mice from their lethal phenotype with good thymic and peripheral T- and B-cell reconstitution. Human ADA-deficient CD34(+) cells transduced with 1-5 × 10(7) TU/ml had 1-3 vector copies/cell and expressed 1-2x of normal endogenous levels of ADA, as assayed in vitro and by transplantation into immune-deficient mice. Importantly, in vitro immortalization assays demonstrated that LV EFS ADA had significantly less transformation potential compared to gRV vectors, and vector integration-site analysis by nrLAM-PCR of transduced human cells grown in immune-deficient mice showed no evidence of clonal skewing. These data demonstrated that the LV EFS ADA vector can effectively transfer the human ADA cDNA and promote immune and metabolic recovery, while reducing the potential for vector-mediated insertional mutagenesis.

Figures

Figure 1

Viral constructs and transgene expression in murine lineage depleted bone marrow and human cord blood CD34+ cells. (a,b) Schematic representation of viral vectors. (a) Retroviral vectors (gRV): the γ-retroviral vector, MND-MFG-ADA (gRV MND ADA) contains the MND retroviral LTRs flanking the wild-type human adenosine deaminase cDNA (hADA) with the Moloney murine leukemia virus packaging region (Ψ) and env splice acceptor fragment (env SA). The gRVSFada/W vector contains hADA driven by SFFV LTR. (b) Lentiviral vectors (LV): All lentiviral vectors contain the enhancer-deleted “SIN” LTR (indicated by the X in the U3 region), the primer binding site (Θ), the HIV-1 packaging signal (Ψ), the central polypurine tract (cPPT), the rev-responsive element (RRE). LV MND ADA contains the MND LTR U3 region enhancer/promoter (MND) driving expression of the hADA cDNA. LV EFS ADA contains the human elongation factor-α gene “short” promoter (EFS) driving expression of the codon-optimized human ADA cDNA (co-hADA) and a Woodchuck Hepatitis Virus posttranscriptional regulatory element (wPRE). The LV EFS GFP vector contains the EFS promoter and green fluorescent protein (GFP). (c,d) Murine ADA−/− bone marrow lineage negative (Lin-) progenitors were transduced for 24 hours with lentiviral or retroviral vectors at a multiplicity of infection (MOI) of 20 after 24 or 72 hours preactivation, respectively (Mean ± SD). ADA expression was analyzed 72 hours after transduction by (c) enzymatic activity assay and (d) Western Blot with whole cell lysates. Mean and standard deviation of ADA activities were calculated from experiments performed with cells obtained from three different ADA−/− donors. (e–g) Human cord blood CD34+ cells were transduced with the vectors at the indicated vector concentration, grown for 2 weeks in myeloid differentiation culture, and assayed for (e) vector copy number (VCN) by quantitative polymerase chain reaction (qPCR) and for (f) ADA enzyme activity by colorimetric assay. (g) The ADA enzyme activity present per VC was calculated. Horizontal bars indicate Mean ± SEM.

Figure 2

Survival rate, peripheral blood (PB) analysis, and immunophenotype of ADA−/− recipients. (a) Survival: Kaplan–Meier curves of ADA−/− recipients were transplanted with transduced ADA−/− BM Lin- cells (LV EFS ADA, n = 6) and gRVSFada/W (SFada/W, n = 5), respectively, at a multiplicity of infection (MOI) of 20. Control mice were injected with untransduced BM Lin- cells from ADA+/+ donors (WT Lin-, n = 5). All LV EFS ADA and WT mice were alive at 13 weeks compared to gRVSFada/W group, where two mice died at 7 weeks and one died at 12 weeks with the remaining two alive at 13 weeks (P = 0.02). All surviving mice were euthanized for analysis unless otherwise indicated. (b) Percentage of donor cells in total PB mononuclear cells (PBMCs) by quantitative PCR (qPCR). (c) Vector copy number (VCN) in PBMCs of transplanted ADA−/− mice. Percentage of DNA with Y chromosome were evaluated in sex-mismatched transplants indicated in 1a (Mean ± SD). (d) FACS analysis of circulating mature T and B cells in peripheral blood of ADA−/− transplants. Untreated ADA−/− mice (untreated, 18 days old, n = 2) and 4–5 months old PEG-ADA treated ADA−/− mice (ERT, n = 2) were analyzed as controls. Data are displayed as percentage of CD3+, CD4+, CD8+, and B220+ cells in PBMCs. Horizontal bars indicate the average values. (e) Total mononucleated cell counts in thymi and spleens (*P < 0.001; **P < 0.05). Results are given as Mean ± SD. (f) FACS analysis of thymocytes of ADA−/− recipients and control mice. Data are presented as percentage of total CD3+ and CD4−CD8− cells in mononucleated cells. Horizontal bars indicate the average values. (g) FACS analysis of splenocytes in ADA−/− recipients and control mice. Data are presented as percentage of CD3+ and B220+ cells in total mononucleated cells. Horizontal bars indicate the average values.

Figure 3

Engraftment of transduced cells, immune reconstitution, and systemic detoxification in ADA−/− recipients. (a) Vector copy number analysis of thymus, spleen, and bone marrow in ADA−/− recipients. (Mean ± SD). (b) Percentage of donor cells in thymus, spleen, and bone marrow by quantitative PCR (qPCR). Percentage of DNA with Y chromosome were evaluated in sex-mismatched transplants indicated in 1a (Mean ± SD). (c) ADA activity and (d) SAHH activity in red blood cells, BM cells, thymocytes, and lung tissue of ADA−/− transplants and control mice were measured by enzymatic activity assay as indicated (Mean ± SD). (e) Histopathologic analysis of lung sections from ADA−/− transplanted with ADA−/− Lin- gRVSFada/W or ADA−/− Lin- LV EFS ADA or ADA+/+ WT Lin- cells compared to lung sections from 18-day-old untreated ADA−/− and ADA−/− mice under ERT (ADA−/− ERT) mice. All sections have been stained with hematoxylin and eosin.

Figure 4

The role of IL-3 in EFS-ADA lentiviral transduction of human cord blood CD34+ cells and engraftment in NSG mice. Human cord blood CD34+ cells were transduced with LV EFS-ADA (3 × 107 TU/ml) in medium with recombinant human cytokines SCF/ckit ligand, flt-3 ligand, and thrombopoietin (TPO), with or without interleukin-3 (IL-3). (a-d) In vitro: (a) transduced cells cultured for 14 days in vitro under myeloid differentiation conditions and analyzed for VCN (14d VCN) (N.S, not significant). (b,c) Transduced CD34+ cells were grown in colony-forming unit (CFU) assay in methylcellulose and assayed after 2 weeks. Colony (b) enumeration and (c) types formed by CD34+ cells in CFU assay. (d,e) CFU were harvested and DNA analyzed by qPCR for VCN (d) Transduction efficiency was measured by the percentage of colonies positive for vector sequence by PCR for the human ADA cDNA (%PCR (+) CFU). (e) VCN was quantified in DNA extracted from individual CFU by qPCR (*P value = 0.001). (f-h) In Vivo: other portions of the transduced CD34+ cells were transplanted into NSG mice and analyzed after 4 months for engraftment of human cells based on FACS analysis of huCD45 expression and for VCN by qPCR. Engraftment of human cells in (f) bone marrow, thymus (when present) and spleen by FACS of tissue cell suspensions immunostained with anti-human CD45 (%hCD45+). (g) EFS-ADA VCN in bone marrow, thymus (when enough cells were available for analysis; total n = 3),and spleen. (h) Immunophenotypic analysis of human CD45+ cells in NSG bone marrow (CD34+ and CD33+), thymus (CD4−/CD8− double-negative (DN), CD4+/CD8+ double-positive (DP), CD4+ single-positive (SP-4) and CD8+ single-positive (SP-8)) and spleen (CD19+ and CD3+).

Figure 5

EFS-ADA transduction of normal and ADA-deficient human cord blood and bone marrow CD34+ cells analyzed in vitro and in vivo. ADA-deficient severe combined immunodeficiency (SCID) bone marrow CD34+ cells from two donors in three separate experiments were isolated and transduced with the EFS-ADA vector at 3 × 107 TU/ml or mock-transduced, cultured in short-term myeloid culture for 2 weeks and then harvested and analyzed. (a–c). In vitro: CFU progenitor assays. (a) Enumeration of lineage committed progenitors (b) the frequency of colonies of different lineages. (c) Transduction efficiency determined by the presence of vector sequence in DNA from isolated colonies. (d) In vitro ADA activity (U) measured in mock-transduced and in EFS-ADA transduced bone marrow CD34+ cells and the VCN and expressed ADA activity (U)/VC measured in the EFS-ADA transduced cultures. (e–i) In Vivo: NSG humanized mice. (e) Engraftment of human (%hCD45+) cells in the bone marrow, thymus, and spleen of NSG mice 4 months after transplantation with mock-transduced or EFS-ADA-transduced human ADA-deficient SCID bone marrow CD34+ cells. (f) Human CD45+ leukocyte populations by immunophenotype in bone marrow, thymus, and spleens from NSG mice (g) ADA enzyme activity (U) (h) EFS-ADA VCN and (I) ADA activity (U)/VC) in cells isolated from the bone marrow (huCD45-selected), thymus (total thymocytes), and spleen (huCD45-selected).

Figure 6

Assessments of genotoxicity of EFS-ADA lentiviral vector. In vitro immortalization (IVIM) Assay. Murine Lin- cells transduced by the indicated vectors were expanded as mass cultures for 2 weeks. An aliquot was taken for qPCR for VCN measurement. On day 15, cells were plated into 96-well plates at a density of 100 cells/well or 1,000 cells/well in 100 µl medium. Two weeks later, the wells showing abundant cell growth were counted as positive, and the frequency of replating cells was calculated based on Poisson statistics using L-Calc Software (Stemcell Technologies, Vancouver, Canada). Horizontal bars indicate mean values. (a) Replating frequency corrected for VCN group by investigators at GOSH, UK. (b) Replating frequency corrected for VCN group by investigators at University of California, Los Angeles, USA (c,d) Vector integration site analysis in human ADA-deficient bone marrow in vitro and in vivo. (c) The percentages of unique integration sites in human cells (isolated from primary NSG mouse recipient bone marrow) near cancer-related genes were determined in vitro (n = 9,822 unique sites) or in vivo (n = 3,141 unique sites). Integration sites in genes or within 300 kb of gene TSS were considered “near” and cancer-related genes were defined as in Higgins et al. .(d) The EFS-ADA vector integration sites were mapped relative to transcriptional start sites (TSS) in vitro (n = 1,610 unique sites) and in vivo (n = 517), and compared to a published data set for MLV1 (n = 828). Grey line represents the theoretical random distribution (n = 12,837).