Dominant selection of hematopoietic progenitor cells with retroviral MDR1 co-expression vectors

Chemoprotection of murine hematopoietic cells by combined gene transfer of cytidine deaminase (CDD) and multidrug resistance 1 gene (MDR1)

Background

Hematologic toxicity represents a major side effect of cytotoxic chemotherapy frequently preventing adequately dosed chemotherapy application and impeding therapeutic success. Transgenic (over)expression of chemotherapy resistance (CTX-R) genes in hematopoietic stem- and progenitor cells represents a potential strategy to overcome this problem. To apply this concept in the context of acute myeloid leukemia and myelodysplasia, we have investigated the overexpression of the multidrug resistance 1 (MDR1) and the cytidine deaminase (CDD) gene conferring resistance to anthracyclines and cytarabine (Ara-C), the two most important drugs in the treatment of these diseases.

Methods

State-of-the-art, third generation, self-inactivating (SIN) lentiviral vectors were utilized to overexpress a human CDD-cDNA and a codon-optimized human MDR1-cDNA corrected for cryptic splice sites from a spleen focus forming virus derived internal promoter. Studies were performed in myeloid 32D cells as well as primary lineage marker negative (lin⁻) murine bone marrow cells and flow cytometric analysis of suspension cultures and clonogenic analysis of vector transduced cells following cytotoxic drug challenge were utilized as read outs.

Results

Efficient chemoprotection of CDD and MDR1 transduced hematopoietic 32D as well as primary lin⁻ cells was proven in the context of Ara-C and anthracycline application. Both, CTX-R transduced 32D as well as primary hematopoietic cells displayed marked resistance at concentrations 5–20 times the LD₅₀ of non-transduced control cells. Moreover, simultaneous CDD/MDR1 gene transfer resulted in similar protection levels even when combined Ara-C anthracycline treatment was applied. Furthermore, significant enrichment of transduced cells was observed upon cytotoxic drug administration.

Conclusions

Our data demonstrate efficient chemoprotection as well as enrichment of transduced cells in hematopoietic cell lines as well as primary murine hematopoietic progenitor cells following Ara-C and/or anthracycline application, arguing for the efficacy as well as feasibility of our approach and warranting further evaluation of this concept.

Electronic supplementary material

The online version of this article (doi:10.1186/s13046-015-0260-4) contains supplementary material, which is available to authorized users.

Knockdown of HPRT for selection of genetically modified human hematopoietic progenitor cells

The inability to obtain sufficient numbers of transduced cells remains a limitation in gene therapy. One strategy to address this limitation is in vivo pharmacologic selection of transduced cells. We have previously shown that knockdown of HPRT using lentiviral delivered shRNA facilitates efficient selection of transduced murine hematopoietic progenitor cells (HPC) using 6-thioguanine (6TG). Herein, we now extend these studies to human HPC. We tested multiple shRNA constructs in human derived cell lines and identified the optimal shRNA sequence for knockdown of HPRT and 6TG resistance. We then tested this vector in human umbilical cord blood derived HPC in vitro and in NOD/SCID recipients. Knockdown of HPRT effectively provided resistance to 6TG in vitro. 6TG treatment of mice resulted in increased percentages of transduced human CD45(+) cells in the peripheral blood and in the spleen in particular, in both myeloid and lymphoid compartments. 6TG treatment of secondary recipients resulted in higher percentages of transduced human cells in the bone marrow, confirming selection from the progeny of long-term repopulating HPCs. However, the extent of selection of cells in the bone marrow at the doses of 6TG tested and the toxicity of higher doses, suggest that this strategy may be limited to selection of more committed progenitor cells. Together, these data suggest that human HPC can be programmed to be resistant to purine analogs, but that HPRT knockdown/6TG-based selection may not be robust enough for in vivo selection.

Cotransduction with MGMT and Ubiquitous or Erythroid-Specific GFP Lentiviruses Allows Enrichment of Dual-Positive Hematopoietic Progenitor Cells In Vivo

The P140K point mutant of MGMT allows robust hematopoietic stem cell (HSC) enrichment in vivo. Thus, dual-gene vectors that couple MGMT and therapeutic gene expression have allowed enrichment of gene-corrected HSCs in animal models. However, expression levels from dual-gene vectors are often reduced for one or both genes. Further, it may be desirable to express selection and therapeutic genes at distinct stages of cell differentiation. In this regard, we evaluated whether hematopoietic cells could be efficiently cotransduced using low MOIs of two separate single-gene lentiviruses, including MGMT for dual-positive cell enrichment. Cotransduction efficiencies were evaluated using a range of MGMT : GFP virus ratios, MOIs, and selection stringencies in vitro. Cotransduction was optimal when equal proportions of each virus were used, but low MGMT : GFP virus ratios resulted in the highest proportion of dual-positive cells after selection. This strategy was then evaluated in murine models for in vivo selection of HSCs cotransduced with a ubiquitous MGMT expression vector and an erythroid-specific GFP vector. Although the MGMT and GFP expression percentages were variable among engrafted recipients, drug selection enriched MGMT-positive leukocyte and GFP-positive erythroid cell populations. These data demonstrate cotransduction as a mean to rapidly enrich and evaluate therapeutic lentivectors in vivo.

In vivo proliferation of postmitotic cochlear supporting cells by acute ablation of the retinoblastoma protein in neonatal mice

Cochlear hair cells (HCs) are mechanosensory receptors that transduce sound into electrical signals. HC damage in nonmammalian vertebrates induces surrounding supporting cells (SCs) to divide, transdifferentiate and replace lost HCs; however, such spontaneous HC regeneration does not occur in the mammalian cochlea. Here, we acutely ablate the retinoblastoma protein (Rb), a crucial cell cycle regulator, in two subtypes of postmitotic SCs (pillar and Deiters' cells) using an inducible Cre line, Prox1-CreER(T2). Inactivation of Rb in these SCs results in cell cycle reentry of both pillar and Deiters' cells, and completion of cell division with an increase in cell number of pillar cells. Interestingly, nuclei of Rb(-/-) mitotic pillar and Deiters' cells migrate toward the HC layer and divide near the epithelial surface in a manner similar to the SCs in the regenerating avian auditory epithelium. In contrast to postmitotic Rb(-/-) HCs which abort cell division, postmitotic Rb(-/-) pillar cells can proliferate, maintain their SC fate and survive for more than a week. However, no newly formed HCs are detected and SC death followed by HC loss occurs. Our studies accomplish a crucial step toward functional HC regeneration in the mammalian cochlea in vivo, demonstrating the critical role of Rb in maintaining quiescence of postmitotic pillar and Deiters' cells and highlighting the heterogeneity between these two cell types. Therefore, the combination of transient Rb inactivation and further manipulation of transcription factors (i.e., Atoh1 activation) in SCs may represent an effective therapeutic avenue for HC regeneration in the mammalian cochlea.

The cyclin-dependent kinase inhibitor roscovitine and the nucleoside analog sangivamycin induce apoptosis in caspase-3 deficient breast cancer cells independent of caspase mediated P-glycoprotein cleavage: implications for therapy of drug resistant breast cancers

Resistance to multiple chemotherapeutic agents is a common clinical problem which can arise during cancer treatment. Drug resistance often involves overexpression of the multidrug resistance MDR1 gene, encoding P-glycoprotein (P-gp), a 170-kDa glycoprotein belonging to the ATP-binding cassette superfamily of membrane transporters. We have recently demonstrated apoptosis-induced, caspase-3-dependent P-gp cleavage in human T-lymphoblastoid CEM-R VBL100 cells. However, P-gp contain many aspartate residues which could be targeted by caspases other than caspase-3. To test whether other caspases could cleave P-gp in vivo, we investigated the fate of P-gp during roscovitine- and sangivamycin- induced apoptosis in MCF7 human breast cancer cells, as they lack functional caspase-3. MCF7 cells were stably transfected with human cDNA encoding P-gp. P-gp was cleaved in vitro by purified recombinant caspase-3, -6 and -7. However, P-gp cleavage was not detected in vivo in MCF7 cells induced to undergoing apoptosis by either roscovitine or sangivamycin, despite activation of both caspase-6 and -7. Interestingly, P-gp overexpressing MCF7 cells were more sensitive to either roscovitine or sangivamycin than wild-type cells, suggesting a novel potential therapeutic strategy against P-gp overexpressing cells. Taken together, our results support the concept that caspase-3 is the only caspase responsible for in vivo cleavage of P-gp and also highlight small molecules which could be effective in treating P-gp overexpressing cancers.

Chemoprotection by transfer of resistance genes

Dose-limiting toxicity of chemotherapeutic agents, i.e., myelosuppression, can limit their effectiveness. The transfer and expression of drug-resistance genes might decrease the risks associated with acute hematopoietic toxicity. Protection of hematopoietic stem/progenitor cells by transfer of drug-resistance genes provides the possibility of intensification or escalation of antitumor drug doses and consequently an improved therapeutic index. This chapter reviews drug-resistance gene transfer strategies for either myeloprotection or therapeutic gene selection. Selecting candidate drug-resistance gene(s), gene transfer methodology, evaluating the safety and the efficiency of the treatment strategy, relevant in vivo models, and oncoretroviral transduction of human hematopoietic stem/progenitor cells under clinically applicable conditions are described.

An approach to achieve long-term expression in skin gene therapy

For gene therapy purposes, the skin is an attractive organ to target for systemic delivery of therapeutic proteins to treat systemic diseases, skin diseases, or skin cancer. To achieve long-term stable expression of a therapeutic gene in keratinocytes (KC), we have developed an approach using a bicistronic retroviral vector expressing the desired therapeutic gene linked to a selectable marker (multidrug resistant gene, MDR) that is then introduced into KC and fibroblasts (FB) to create genetically modified human skin equivalent (HSE). After grafting the HSE onto immunocompromised mice, topical colchicine treatment is used to select and enrich for genetically modified keratinocyte stem cells (KSC) that express MDR and are resistant to colchicine's antimitotic effects. Both the apparatus for topical colchicine delivery and the colchicine doses have been optimized for application to human skin. This approach can be validated by systemic delivery of therapeutic factors such as erythropoietin and the antihypertensive atrial natriuretic peptide.

Stable differentiation and clonality of murine long-term hematopoiesis after extended reduced-intensity selection for MGMT P140K transgene expression

Efficient in vivo selection increases survival of gene-corrected hematopoietic stem cells (HSCs) and protects hematopoiesis, even if initial gene transfer efficiency is low. Moreover, selection of a limited number of transduced HSCs lowers the number of cell clones at risk of gene activation by insertional mutagenesis. However, a limited clonal repertoire greatly increases the proliferation stress of each individual clone. Therefore, understanding the impact of in vivo selection on proliferation and lineage differentiation of stem-cell clones is essential for its clinical use. We established minimal cell and drug dosage requirements for selection of P140K mutant O6-methylguanine-DNA-methyltransferase (MGMT P140K)-expressing HSCs and monitored their differentiation potential and clonality under long-term selective stress. Up to 17 administrations of O6-benzylguanine (O6-BG) and 1,3-bis(2-chloroethyl)-1-nitroso-urea (BCNU) did not impair long-term differentiation and proliferation of MGMT P140K-expressing stem-cell clones in mice that underwent serial transplantation and did not lead to clonal exhaustion. Interestingly, not all gene-modified hematopoietic repopulating cell clones were efficiently selectable. Our studies demonstrate that the normal function of murine hematopoietic stem and progenitor cells is not compromised by reduced-intensity long-term in vivo selection, thus underscoring the potential value of MGMT P140K selection for clinical gene therapy.

Gamma-glutamylcysteine synthetase-based selection strategy for gene therapy of chronic granulomatous disease and graft-vs.-host disease

Efficient ex vivo/in vivo selection of genetically modified hematopoietic stem/progenitor cells (HPCs) and T lymphocytes could greatly improve several gene therapy strategies. We have previously reported that primary murine HPCs, transduced with a bicistronic retroviral vector, co-expressing the catalytic subunit of gamma-glutamylcysteine synthetase (gamma-GCSh) and eGFP, could be selected by l-buthionine-S,R-sulfoximine (BSO). Upon ex vivo transduction with a low, defined gene dosage and BSO selection, HPCs were able to repopulate the bone marrow of syngeneic myeloablated hosts, showing multi-lineage expression [Hum Gene Ther, 16 (2005), 711]. We now provide 'proof-of-principle' that the same strategy can be applied to the gene therapy of graft-vs.-host disease (GVHD) subsequent to allogeneic bone marrow transplantation (ABMT), and of chromosome X-associated chronic granulomatous disease (CGD). Transfer of the herpes simplex virus-thymidine kinase (HSV-Tk) 'suicide' gene into donor T lymphocytes is a potential method to control GVHD after ABMT. However, an efficient selection system is required to eliminate non-HSV-Tk-expressing T lymphocytes before administration to the patient. We now report that, upon transduction with a retroviral vector, co-expressing gamma-GCSh and eGFP, and subsequent selection by BSO, over 95% human T lymphocytes were found to express eGFP; moreover, upon transduction with a novel retroviral vector co-expressing gamma-GCSh and HSV-Tk, and subsequent BSO treatment, over 95% of T lymphocytes could be eliminated by ganciclovir. The efficacy of the gamma-GCSh-BSO selection strategy was then tested on an in vitro model of CGD. Upon transduction of gp91 (phox)-deficient PLBKO cells with a novel bicistronic retroviral vector co-expressing human gp91 (phox) and gamma-GCSh, exposure to BSO for 48 h eliminated most non-transduced cells, resulting in selection of gp91 (phox)-expressing cells, and reconstitution of NADPH oxidase activity.

Reliable Generation of Stable High Titer Producer Cell Lines for Gene Therapy

OBJECTIVE

Retroviral vectors represent one of the most robust technologies for in vivo expression of heterologous gene sequences and are still the most commonly used vectors in clinical gene therapy trials. The production of high titer retroviral preparations, however, can be a problematic procedure for certain constructs.

METHODS

GALV- or RD114-pseudotyped retroviral particles carrying selectable fluorescence markers or drug resistance genes, such as the green fluorescent protein (GFP) or the O(6)-methylguanine-DNA-methyltransferase (MGMT) mutants, were used to stably transduce Phoenix-(FNX-)eco cells. Thereafter, a polyclonal population of producer cells was generated by enriching cells with high marker gene expression. In addition, single producer clones were selected by limiting dilution.

RESULTS

Retroviral titers were increased 1-2 logs by enriching for a polyclonal population of producer cells, and selection of single producer clones allowed another 1- to 2-log increase in titers. Using this method, reproducibly high titer viral preparations allowing efficient transduction of hematopoietic stem cells were generated.

CONCLUSION

A reliable and time-effective method to generate stable high titer producer cells based on the FNX-cell line for problematic retroviral vector constructs is described.

Gene transfer of cytidine deaminase protects myelopoiesis from cytidine analogs in an in vivo murine transplant model

Hematopoietic stem cell gene transfer of the drug-resistance gene cytidine deaminase (CDD) protecting cells from the cytotoxic cytidine analogs cytarabine and gemcitabine was investigated in a murine transplant model. Following transplantation of CDD-transduced cells and cytarabine application (500 mg/kg; days 1-4; intraperitoneally) significant myeloprotection was demonstrated with nadir counts of peripheral blood granulocytes and thrombocytes of 2.9 ± 0.6/nL versus 0.7 ± 0.1/nL (P < .001) and 509 ± 147/nL versus 80 ± 9/nL (P = .008), respectively (CDD versus control). Protection also was observed from otherwise lethal gemcitabine treatment (250 mg/kg; days 1-3). Stable levels of gene-marked cells in primary and secondary recipients were demonstrated for up to 9 months, and whereas CDD overexpression clearly reduced B- and T-lymphocyte numbers, no major toxicity was observed in the myeloid compartment. Despite the profound myeloprotective properties, however, CDD overexpression did not allow for pharmacologic enrichment of transduced hematopoiesis in our model. Thus, in summary, our data establish CDD as a drug-resistance gene highly suitable for myeloprotective purposes, which, given the lack of selection observed in our hands, might best be used in combination with selectable drugresistance genes such as MGMT (P140K) or MDR1.

Resistance to cytarabine and gemcitabine and in vitro selection of transduced cells after retroviral expression of cytidine deaminase in human hematopoietic progenitor cells

Overexpression of the detoxifying enzyme cytidine deaminase (CDD) renders normal and leukemic hematopoietic cells resistant to cytarabine (1-beta-D-arabinofuranosylcytosine), and studies on murine cells have suggested transgenic CDD overexpression as a way to reduce the substantial myelotoxicity induced by the deoxycytidine analogs cytarabine and gemcitabine (2',2'-difluorodeoxycytidine). We now have investigated CDD (over-)expression in the human hematopoietic system. Retroviral gene transfer significantly increased the resistance of CDD-transduced cord blood and peripheral blood-derived progenitor cells for doses ranging from 20-100 nM cytarabine and 8-10 nM gemcitabine. Protection was observed for progenitors of erythroid as well as myeloid differentiation, though the degree of protection varied for individual drugs. In addition, significant selection of CDD-transduced cells was obtained after a 4-day culture in 30-100 nM cytarabine. Thus, our data demonstrate that overexpression of CDD cDNA results in significant protection of human progenitors from cytarabine- as well as gemcitabine-induced toxicity, and allows in vitro selection of transduced cells. This strongly argues for a potential therapeutic role of CDD gene transfer in conjunction with dose-intensive cytarabine- or gemcitabine-containing chemotherapy regimen.

γ-Glutamylcysteine Synthetase and L-Buthionine-(S,R)-Sulfoximine: A New Selection Strategy for Gene-Transduced Neural and Hematopoietic Stem/Progenitor Cells

In most experimental gene therapy protocols involving stem/progenitor cells, only a small fraction of cells, often therapeutically inadequate, can be transduced and made to express the therapeutic gene. A promising strategy for overcoming this problem is the use of a dominant selection marker, such as a drug resistance gene. In this paper, we explore the potential of the heavy subunit of gamma-glutamylcysteine synthetase (gamma-GCSh) to act as a selection marker. We found that 3T3 fibroblasts transduced with the bicistronic retroviral vector SF91/GCSh-eGFP, encoding gamma-GCSh and the enhanced green fluorescent protein (eGFP), were highly resistant to L-buthionine-(S,R)-sulfoximine (BSO), a gamma-GCS inhibitor with a low clinical toxicity profile. The level of resistance was not proportional to the increase in intracellular glutathione. In fact, cells overexpressing both heavy and light gamma-GCS subunits had higher intracellular GSH levels, and a lower level of resistance to the cytotoxic activity of BSO, compared with cells overexpressing gamma-GCSh alone. 3T3 fibroblasts overexpressing gamma-GCSh could be selected from cultures containing both naive and gene-modified cells by application of exogenous BSO selection pressure for 4 days. Also, primary neural stem/progenitor cells derived from the lateral ventricles of mouse neonatal brains and primary hematopoietic stem/progenitor cells (HSCs/HPCs) from mouse bone marrow, transduced with the gamma-GCSh-eGFP vector, could be selected by BSO treatment in vitro. On ex vivo BSO selection and reimplantation into a syngeneic myeloablated host, donor HSCs/HPCs repopulated the marrow and continued to express the transgene(s). These results provide proof of principle that somatic stem/progenitor cells, transduced simultaneously with a potentially curative gene and gamma-GCSh, can be selected by treatment with BSO before in vivo transplantation.

In vivo selection of human hematopoietic cells in a xenograft model using combined pharmacologic and genetic manipulations

Strategies that increase the ability of human hematopoietic stem and progenitor cells to repair alkylator-induced DNA damage may prevent the severe hematopoietic toxicity in patients with cancer undergoing high-dose alkylator therapy. In the context of genetic diseases, this approach may allow for selection of small numbers of cells that would not otherwise have a favorable growth advantage. No studies have tested this approach in vivo using human hematopoietic stem and progenitor cells. Human CD34(+) cells were transduced with a bicistronic oncoretrovirus vector that coexpresses a mutant form of O(6)-methylguanine DNA methyltransferase (MGMT(P140K)) and the enhanced green fluorescent protein (EGFP) and transplanted into nonobese diabetic/severe combined immunodeficient (NOD/SCID) mice. Mice were either not treated or treated with O(6)-benzylguanine (6BG) and 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU). At 8-weeks postinjection, a 2- to 8-fold increase in the percentage of human CD45(+)EGFP(+) cells in 6BG/BCNU-treated versus nontreated mice was observed in the bone marrow and was associated with increased MGMT(P140K)-repair activity. Functionally, 6BG/BCNU-treated mice demonstrated multilineage differentiation in vivo, although some skewing in the maturation of myeloid and B cells was observed in mice transplanted with granulocyte-colony stimulating factor (G-CSF)-mobilized peripheral blood compared to umbilical cord blood. Expansion of human cells in 6BG/BCNU-treated mice was observed in the majority of mice previously transplanted with transduced umbilical cord blood cells. In addition, a significant increase in the number of EGFP(+) progenitor colonies in treated versus nontreated mice were observed in highly engrafted mice indicating that selection and maintenance of human progenitor cells can be accomplished by expression of MGMT(P140K) and treatment with 6BG/BCNU.

Efficient expansion and gene transduction of mouse neural stem/progenitor cells on recombinant fibronectin

Neural stem/progenitor cells (NSCs) are commonly grown as floating neurospheres in medium containing basic fibroblast growth factor and epidermal growth factor. Under these conditions, about 1% of the cells retain multipotentiality. We developed a protocol based on culture of NSCs in adherence on recombinant fibronectin (rFN) to transduce up to 90% NSCs at a multiplicity of infection of 2 with no need for viral concentration or production of serum-free retroviral supernatants. NSCs grew faster on rFN than as neurospheres on tissue culture plastic and did not lose their stem cell nature or multipotentiality. Furthermore, retroviral-mediated transgene expression was sustained with time in culture and upon differentiation into neurons and astrocytes. These experimental conditions may be utilized to study the function of various genes in NSCs, and to manipulate NSCs for gene and cell therapy of several neurological diseases.

Efficient protection from methotrexate toxicity and selection of transduced human hematopoietic cells following gene transfer of dihydrofolate reductase mutants

OBJECTIVE

While retrovirally mediated gene transfer of dihydrofolate reductase mutants (mutDHFR) has convincingly been demonstrated to confer methotrexate (MTX) resistance to murine hematopoietic cells, clinical application of this technology will require high efficacy in human cells. Therefore, we investigated retroviral constructs expressing various point mutants of human DHFR for their ability to confer MTX resistance to human clonogenic progenitor cells (CFU-C) and to allow for in vitro selection of transduced CFU-C.

METHODS

Primary human hematopoietic cells were retrovirally transduced using MMLV- and SFFV/MESV-based vectors expressing DHFR(Ser31), DHFR(Phe22/Ser31), or DHFR(Tyr22/Gly31). MTX resistance of unselected and in vitro-selected CFU-C was determined using MTX-supplemented methylcellulose cultures and gene transfer efficiency was assesed by single-colony PCR analysis.

RESULTS

While less than 1% mock-transduced CFU-C survived the presence of > or =5 x 10(-8) M MTX, MMLV- and SFFV/MESV-based vectors expressing DHFR(Ser31) significantly protected CFU-C from MTX at doses ranging from 2.5 to 30 x 10(-8) M. Vectors expressing DHFR(Phe22/Ser31) or DHFR(Tyr22/Gly31) were even more protective and MTX-resistant CFU-C were observed up to 1 x 10(-5) M MTX. Three-day suspension cultures in the presence of 10-20 x 10(-8) M MTX resulted in significant selection of mutDHFR-transduced CFU-C. The percentage of CFU-C resistant to 10 x 10(-8) M MTX increased fourfold to 20-fold and provirus-containing CFU-C increased from 27% to 79-100%.

CONCLUSION

Gene transfer of DHFR using suitable retroviral backbones and DHFR mutants significantly increases MTX resistance of human CFU-C and allows efficient in vitro selection of transduced cells using a short-term selection procedure.

Impact of splice-site mutations of the human MDR1 cDNA on its stability and expression following retroviral gene transfer

The multidrug resistance 1 (MDR1) gene transfer to hematopoietic cells for protection against cytotoxic drugs has received considerable attention in gene therapy. However, ectopic expression of MDR1 from retroviral vectors has been hampered by its genetic instability resulting from cryptic splice sites within the cDNA. We have evaluated the efficiency of retroviral MDR1 vectors with introduced mutations of the MDR1 cryptic splice donor (cSD) located at nucleotide +339 and of the cryptic splice acceptor (cSA) at nucleotide +2319 of the cDNA. Sequence alterations of the cSD reduced the expression of MDR1 P-glycoprotein (P-gp), even when generated as silent mutations. A silent mutation of the cSA reduced the splicing activity shifting the splice acceptor site one base downstream; however, it significantly improved the expression of P-gp. The incidence of wild-type MDR1 pregenome splicing was markedly reduced when vectors were produced in human 293 packaging cells as opposed to murine PG13 and GP+envAm12. We conclude that complete splice correction of MDR1 in retroviral vectors may only be achieved with extensive alterations of the cDNA or neighboring vector sequences and that the splicing is significantly influenced by the choice of the packaging cells.

Long-term high-level reconstitution of NADPH oxidase activity in murine X-linked chronic granulomatous disease using a bicistronic vector expressing gp91phox and a Delta LNGFR cell surface marker

A murine model of X-linked chronic granulomatous disease (X-CGD), an inherited immune deficiency with absent phagocyte NADPH oxidase activity caused by defects in the gp91(phox) gene, was used to evaluate a bicistronic retroviral vector in which expression of human gp91(phox) and a linked gene for Delta LNGFR, a truncated form of human low-affinity nerve growth factor receptor, are under the control of a spleen focus-forming virus long-terminal repeat (LTR). Four independent cohorts of 11-Gy irradiated X-CGD mice (total, 22 mice) were transplanted with or without preselection of transduced X-CGD bone marrow (BM). Transplanted mice had high-level correction of neutrophil gp91(phox) expression and reconstitution of NADPH oxidase activity. Expression lasted for at least 14 months in primary transplants, and persisted in secondary and tertiary transplants. Both gp91(phox) and Delta LNGFR were detected on circulating granulocytes, lymphocytes, lymphoid, and (for Delta LNGFR) red blood cells. Mice receiving transduced bone marrow [BM] preselected ex vivo for Delta LNGFR expression had high-level (= 80%) reconstitution with transduced cells, with an improved fraction of oxidase-corrected neutrophils posttransplant. Analysis of secondary and tertiary CFU-S showed that silencing of individual provirus integrants can occur even after preselection for Delta LNGFR prior to transplantation, and that persistent provirus expression was associated with multiple integration sites in most cases. No obvious adverse consequences of transgenic protein expression were observed.

Role of the Multidrug Resistance Protein 1 in protection from heavy metal oxyanions: investigations in vitro and in MRP1-deficient mice

The Multidrug Resistance Protein 1 (MRP1) is a membrane pump that mediates the efflux of a wide variety of xenobiotics, including arsenical and antimonial compounds, as demonstrated by the study of MRP1-transfected cell lines. We have previously shown that mrp1(-/-) cells are hypersensitive to sodium arsenite, sodium arsenate, and antimony potassium tartrate. We now report that the retroviral vector-mediated overexpression of MRP1 and of the two subunits of gamma-GCS (heavy and light) resulted in higher intracellular glutathione levels and in a greater level of resistance to sodium arsenite and antimony potassium tartrate, compared to the overexpression of MRP1 and gamma-GCS heavy alone. These observations further demonstrate that glutathione is an important component of MRP1-mediated cellular resistance to arsenite and antimony. However, the constitutive expression of MRP1 did not protect mice from the lethality of sodium arsenite and antimony potassium tartrate nor reduced the tissue accumulation of arsenic in mice injected i.p. with sodium arsenite. It is conceivable that, in vivo, other pump(s) effectively vicariate for MRP1-mediated transport of heavy metal oxyanions.

Retroviral transfer and expression of human MDR-1 in a murine haemopoietic stem cell line does not alter factor dependence, growth or differentiation characteristics

In view of the recent report of a myeloproliferative syndrome in mice that had received an MDR-1-transduced haemopoietic graft, we have investigated the potential effects of MDR-1 expression on primitive haemopoietic cell growth and differentiation. Retroviral gene transfer was used to achieve exogenous expression of either MDR-1 or truncated nerve growth factor receptor (tNGFR) in the multipotent murine haemopoietic progenitor cell line, FDCP-mix. Following gene transfer, clonal lines were derived and FACS analysis confirmed appropriate expression of each transgene. MDR-1 (but not tNGFR) expression was associated with verapamil-sensitive rhodamine efflux and resistance to killing by etoposide. When growth factor responsiveness, proliferative capacity and differentiation capacity were examined, MDR-1 expressing FDCP-mix cells exhibited a normal phenotype and mimicked the response of tNGFR-expressing or untransduced FDCP-mix cells. Thus, in the model system we have used, MDR-1 does not perturb haemopoietic cell growth and development and our data do not support a myeloproliferative role for MDR-1.

Novel bicistronic retroviral vector expressing gamma-glutamylcysteine synthetase and the multidrug resistance protein 1 (MRP1) protects cells from MRP1-effluxed drugs and alkylating agents

We have constructed two retroviral vectors, one expressing multidrug resistance protein 1 (MRP1) alone (SF91MRP) and the other expressing MRP1 and gamma-glutamylcysteine synthetase (gamma-GCS), the rate-limiting enzyme of glutathione biosynthesis (SF91GCS-MRP). We have utilized the hybrid FMEV (Friend mink cell focus-forming/murine embryonic stem cell virus) backbone, previously shown to be efficient in early hematopoietic cells, even when coexpressing two distinct genes. In SF91GCS-MRP, the cDNAs were combined via an internal ribosomal entry site (IRES) sequence from poliovirus, resulting in a bicistronic mRNA produced via the long terminal repeat (LTR). Producer Fly-eco clones were established by trans-infection with vesicular stomatitis virus glycoprotein (VSV-G)-pseudotyped retroviral supernatants. Drug-resistant producer clones were subsequently selected with antimony potassium tartrate, a nonmutagenic MRP1 substrate. By RNA slot-blot and transduction of 3T3 fibroblasts, titers of both SF91MRP and SF91GCS-MRP were found to be greater than 10(6) viral particles/ml. The correct viral integration in the genome was established by Southern blotting. By flow cytometry, both MRP1 and bicistronic clones showed an increase in expression of the MRP1 protein. The bicistronic producer clones, as well as 3T3 cells transduced with SF91GCS-MRP, presented an increase in intracellular glutathione levels, compared with the parental counterparts. Producer cells, 3T3 fibroblasts transduced with either SF91MRP or SF91GCS-MRP, and primary murine myeloid progenitor cells transduced with SF91GCS-MRP were resistant to MRP1-effluxed drugs. However, only bicistronic producers, 3T3 fibroblasts transduced with SF91GCS-MRP, and primary murine myeloid progenitor cells transduced with SF91GCS-MRP were also resistant to alkylating agents. We conclude that the retrovirus SF91GCS-MRP has features that make it a suitable vector to induce bone marrow resistance to multiple classes of chemotherapeutic agents. The strategy of coexpressing gamma-GCS and MRP1 may help to design an effective in vivo selection for various clinical protocols of gene therapy.

MDR1 bicistronic vectors: analysis of selection stringency, amplified gene expression, and vector stability in cell lines

The human multidrug resistance-1 gene (MDR1) is a dominant selectable and amplifiable marker in mammalian tissue culture cells. MDR1 is also being investigated as a gene therapy tool, both to protect normal cells against chemotherapy-related toxicity and to serve as an in vivo selectable marker for the overexpression of non-selectable therapeutic genes. The success of these strategies will depend on whether MDR1 expression can be sustained at levels high enough to confer a survival advantage on target cells. However, the MDR1 selection system is quite stringent, requiring high gene expression for transduced cells to survive in the presence of drug. The current report is a detailed molecular analysis of MDR1 selection stringency compared with the common neo selectable marker. A bicistronic vector encoding MDR1 and neo genes linked through an internal ribosome entry site was transferred into NIH 3T3 mouse fibroblasts and K562 human leukemia cells; cells were then exposed to colchicine (to select for MDR1 expression) or to G418 (to select for neo expression). Surviving populations and individual clones of cells were analyzed for expression levels of MDR1 and neo gene products; resistance to colchicine, paclitaxel, and G418; level and integrity of bicistronic mRNA; and structural integrity, integration number, and copy number of vector DNA. These studies provide direct evidence that colchicine selection is more stringent than G418 selection; that increased selection pressure with colchicine leads to increased gene expression; that increased gene expression can be accommodated primarily by gene amplification, even within an individual transduced clone and starting from a single-copy proviral integration event; and that the clonal diversity of a transduced population of cells is influenced significantly by the stringency of selection. Taken together, these results have important implications for the potential utility of MDR1 as a selectable marker and as a gene therapy tool in hematopoietic cells.

Protection of hematopoietic cells from O(6)-alkylation damage by O(6)-methylguanine DNA methyltransferase gene transfer: studies with different O(6)-alkylating agents and retroviral backbones

Overexpression of O(6)-methylguanine DNA methyltransferase (MGMT) can protect hematopoietic cells from O(6)-alkylation damage. To identify possible clinical applications of this technology we compared the effect of MGMT gene transfer on the hematotoxicity induced by different O(6)-alkylating agents in clinical use: the chloroethylnitrosoureas ACNU, BCNU, CCNU and the tetrazine derivative temozolomide. In addition, various retroviral vectors expressing the MGMT-cDNA were investigated to identify optimal viral backbones for hematoprotection by MGMT expression. Protection from ACNU, BCNU, CCNU or temozolomide toxicity was evaluated utilizing a Moloney murine leukemia virus-based retroviral vector (N2/Zip-PGK-MGMT) to transduce primary murine bone marrow cells. Increased resistance in murine colony-forming units (CFU) was demonstrated for all four drugs. In comparison to mock-transduced controls, after transduction with N2/Zip-PGK-MGMT the IC50 for CFU increased on average 4.7-fold for ACNU, 2.5-fold for BCNU, 6.3-fold for CCNU and 1.5-fold for temozolomide. To study the effect of the retroviral backbone on hematoprotection various vectors expressing the human MGMT-cDNA from a murine embryonic sarcoma virus LTR (MSCV-MGMT) or a hybrid spleen focus-forming/murine embryonic sarcoma virus LTR (SF1-MGMT) were compared with the N2/Zip-PGK-MGMT vector. While all vectors increased resistance of transduced human CFU to ACNU, the SF1-MGMT construct was most efficient especially at high ACNU concentrations (8-12 microg/ml). Similar results were obtained for protection of murine high-proliferative-potential colony-forming cells. These data may help to optimize treatment design and retroviral constructs in future clinical studies aiming at hematoprotection by MGMT gene transfer.

Bone marrow stem cell gene therapy of arylsulfatase A-deficient mice, using an arylsulfatase A mutant that is hypersecreted from retrovirally transduced donor-type cells

Arylsulfatase A (ASA)-deficient mice represent an animal model for the fatal lysosomal storage disease metachromatic leukodystrophy, which is characterized by widespread intralysosomal deposition of sulfatide. Bone marrow stem cell gene therapy in mice, using a retroviral vector mediating expression of wild-type human ASA, has the potential to ameliorate the visceral pathology, but improves the prevailing brain disease and neurologic symptoms only marginally. One factor that influences the efficacy of bone marrow transplantation therapy in lysosomal storage diseases is the secretion level of the therapeutic enzyme from donor-type cells. Here we test the potential of a hypersecreted glycosylation variant of ASA. Although this mutant lacks mannose 6-phosphate residues it is taken up by cells by a mannose 6-phosphate receptor-independent pathway and causes partial metabolic correction of ASA-deficient mouse cells. Retrovirally mediated transfer of the mutant cDNA into ASA-deficient mice results in the sustained expression of the transgene. Serum levels argue for an increased secretion of the glycosylation mutant also in vivo. Tissue levels were reduced to 2% in liver and up to 40% in kidney compared with animals treated with the wild-type enzyme, indicating reduced endocytosis. Thus, the limited uptake of the variant enzyme outweighs the putative advantageous effect of improved supply. Although the mutant enzyme is able to correct the metabolic defect partially, histological examinations did not reveal any reduction of sulfatide storage in treated animals. Surprisingly, analysis of neurologic symptoms indicated a significant improvement of the gait pattern.

Membrane-anchored peptide inhibits human immunodeficiency virus entry

Peptides derived from the heptad repeats of human immunodeficiency virus (HIV) gp41 envelope glycoprotein, such as T20, can efficiently inhibit HIV type 1 (HIV-1) entry. In this study, replication of HIV-1 was inhibited more than 100-fold in a T-helper cell line transduced with a retrovirus vector expressing membrane-anchored T20 on the cell surface. Inhibition was independent of coreceptor usage.

Lack of superinfection interference in retroviral vector producer cells

Vesicular stomatitis virus G protein (VSV-G)-pseudotyped retroviral vectors have become more feasible for clinical gene transfer protocols since stable tetracycline (tet)-regulated packaging cell lines have become available. Here, we analyzed superinfection interference in VSV-G-pseudotyped and classic amphotropic packaging cell lines. No superinfection interference was observed in VSV-G-pseudotyped packaging cell lines. Thus, integrated retroviral vector genomes accumulated during culture. Similar results were obtained with the amphotropic packaging cells, but to a lesser degree. In addition, VSV-G packaging cells were susceptible to infection with vector particles devoid of envelope proteins, which are produced by these cells in high titers when VSV-G expression is suppressed by tetracycline. For both packaging systems, superinfection could be blocked by azidothymidine (AZT). With regard to safety, this study suggests that in clinical protocols amphotropic producer clones should be tested for superinfection interference and VSV-G packaging cells should always be cultured in the presence of AZT.

Improved post-transcriptional processing of an MDR1 retrovirus elevates expression of multidrug resistance in primary human hematopoietic cells

We describe the functional analysis of a novel retroviral vector, SF91m3, which was designed for improved expression of the in vivo selectable marker, multidrug resistance 1 gene (MDR1), in hematopoietic cells. SF91m3 combines several promising features. The vector backbone lacks viral coding sequences and AUG-start codons 5' of the MDR1 cDNA. A point mutation of a cryptic splice acceptor of the MDR1 cDNA increases the probability of transferring an intact provirus. The titer of a PG13 packaging cell clone containing a single proviral integration is high (>2 x 10(6) particles/ml from frozen stocks of serum-free vector harvests). Human hematopoietic cells transduced with SF91m3 reliably express MDR1 before and after passage through NOD/SCID mice, as shown by quantitative PCR and efflux assays with rhodamine 123 or Hoechst 33342. Finally, SF91m3 mediates resistance to escalated doses of cytotoxic agents, as shown by survival and differentiation of transduced colony-forming cells in the presence of colchicine at 48 ng/ml (>10 x IC(50)). Thus, SF91m3 may represent an interesting candidate for future trials addressing the safety and utility of MDR1 gene transfer; moreover, this study demonstrates that sequence alterations improving post-transcriptional processing of retroviral vectors have a substantial impact for gene expression in hematopoietic cells.

Protection and in vivo selection of hematopoietic stem cells using temozolomide, O6-benzylguanine, and an alkyltransferase-expressing retroviral vector

Transfer of drug resistance genes to hematopoietic stem cells offers the potential to protect cancer patients from drug-induced myelosuppression and to increase the number of gene-modified cells by in vivo selection. In this study, a retroviral vector expressing both a P140K variant of human O6-methylguanine-DNA methyltransferase (MGMT) and an EGFP reporter gene was evaluated for stem cell protection in a murine transplant model. Mice transplanted with vector-transduced cells showed significant resistance to the myelosuppressive effects of temozolomide (TMZ), an orally administered DNA-methylating drug, and O6-benzylguanine (BG), a drug that depletes cells of wild-type MGMT activity. Following drug treatment, increases in EGFP(+) peripheral blood cells were seen in all peripheral blood lineages, and secondary transplant experiments proved that selection had occurred at the stem cell level. In a second set of experiments in which transduced cells were diluted with unmarked cells, efficient stem cell selection was noted together with progressive marrow protection with repeated treatment courses. Altogether, these results show that P140K MGMT gene transfer can protect stem cells against the toxic effects of TMZ and BG and that this vector/drug system may be useful for clinical myeloprotection and for in vivo selection of transduced stem cells.

Functional reconstitution of class II MHC-restricted T cell immunity mediated by retroviral transfer of the alpha beta TCR complex

Transfer of the alphabeta TCR genes into T lymphocytes will provide a means to enhance Ag-specific immunity by increasing the frequency of tumor- or pathogen-specific T lymphocytes. We generated an efficient alphabeta TCR gene transfer system using two independent monocistronic retrovirus vectors harboring either of the class II MHC-restricted alpha or beta TCR genes specific for chicken OVA. The system enabled us to express the clonotypic TCR in 44% of the CD4+ T cells. The transduced cells showed a remarkable response to OVA323-339 peptide in the in vitro culture system, and the response to the Ag was comparable with those of the T lymphocytes derived from transgenic mice harboring OVA-specific TCR. Adoptive transfer of the TCR-transduced cells in mice induced the Ag-specific delayed-type hypersensitivity in response to OVA323-339 challenge. These results indicate that alphabeta TCR gene transfer into peripheral T lymphocytes can reconstitute Ag-specific immunity. We here propose that this method provides a basis for a new approach to manipulation of immune reactions and immunotherapy.

Quantitative assessment of retroviral transfer of the human multidrug resistance 1 gene to human mobilized peripheral blood progenitor cells engrafted in nonobese diabetic/severe combined immunodeficient mice

Mobilized peripheral blood progenitor cells (PBPC) are a potential target for the retrovirus-mediated transfer of cytostatic drug-resistance genes. We analyzed nonobese diabetic/severe combined immunodeficient (NOD/SCID) mouse-repopulating CD34+ PBPC from patients with cancer after retroviral transduction in various cytokine combinations with the hybrid vector SF-MDR, which is based on the Friend mink cell focus-forming/murine embryonic stem-cell virus and carries the human multidrug resistance 1 (MDR1) gene. Five to 13 weeks after transplantation of CD34+ PBPC into NOD/SCID mice (n = 84), a cell dose-dependent multilineage engraftment of human leukocytes up to an average of 33% was observed. The SF-MDR provirus was detected in the bone marrow (BM) and in its granulocyte fractions in 96% and 72%, respectively, of chimeric NOD/SCID mice. SF-MDR provirus integration assessed by quantitative real-time polymerase chain reaction (PCR) was optimal in the presence of Flt-3 ligand/thrombopoietin/stem-cell factor, resulting in a 6-fold (24% ± 5% [mean ± SE]) higher average proportion of gene-marked human cells in NOD/SCID mice than that achieved with IL-3 alone (P &lt; .01). A population of clearly rhodamine-123dull human myeloid progeny cells could be isolated from BM samples from chimeric NOD/SCID mice. On the basis of PCR and rhodamine-123 efflux data, up to 18% ± 4% of transduced cells were calculated to express the transgene. Our data suggest that the NOD/SCID model provides a valid assay for estimating the gene-transfer efficiency to repopulating human PBPC that may be achievable in clinical autologous transplantation. P-glycoprotein expression sufficient to prevent marrow aplasia in vivo may be obtained with this SF-MDR vector and an optimized transduction protocol.

Establishment of an optimised gene transfer protocol for human primary T lymphocytes according to clinical requirements

Current gene therapeutic protocols directed towards the treatment of inherited disorders (eg ADA-SCID) and viral infections (eg AIDS), as well as adoptive immunotherapy approaches are based on the use of genetically modified lymphocytes. Since only insufficient transduction of T cells is obtained using existing techniques, the development of more efficient gene transfer protocols into these cells is of great importance. We present here a protocol for the highly efficient transduction of human primary T cells at high densities (1 x 106/ml) by retroviral infection. Using retroviral vectors encoding a truncated human low-affinity nerve growth factor receptor (DeltaLNGFR) as a gene transfer marker, we obtained transduction frequencies of more than 70% of CD3+ cells after two cycles of infection. Our protocol is based on the use of FBS-free media for both the production of retrovirus-containing supernatant and the cultivation of the primary T cells. Since the protocol presented here works just as efficiently under large-scale conditions, it may be easily adapted to clinical needs and 'good manufacturing practice' (GMP) standards.

Retroviral transfer of human MDR1 gene to hematopoietic cells: effects of drug selection and of transcript splicing on expression of encoded P-glycoprotein

Protection of hematopoietic cells of patients undergoing anticancer chemotherapy by MDR1 gene transfer is currently being studied in clinical trials. From animal studies, it has been suggested that aberrant splicing due to cryptic donor and acceptor sites in the MDR1 cDNA could be a major reason for failure to obtain high-level expression of P-glycoprotein in bone marrow. We investigated effects of drug selection on protein expression levels and on splicing of MDR1 transcripts in murine bone marrow cells (BMCs) in vitro. To this end, retroviruses were generated through an identical plasmid, pHaMDR1/A, introduced into different packaging cells. GP + E86- but not PA317-derived producer cells were found to express truncated in addition to full-length message. In BMCs transduced with GP + E86-derived viruses, both messages were increased after treatment with colchicine or daunomycin. Similar results were obtained with NIH 3T3 fibroblasts. However, transduced and drug-selected BMCs displayed the spliced transcript even if the respective PA317-derived producer cells contained no truncated RNA as detected in transduced NIH 3T3 fibroblasts. Short-term drug selection in BMCs transduced with either ecotropic or amphotropic retroviruses resulted in a striking increase in P-glycoprotein expression. Thus, aberrant splicing failed to abrogate P-glycoprotein expression in BMCs. We also studied a vector in which MDR1 was coexpressed with glucocerebrosidase, using an internal ribosomal entry site. Although chemoprotection was less efficient than with pHaMDR1/A, augmentation of protein expression was observed at low selecting drug concentrations. Our study shows that drug selection can partially compensate for inefficient transduction of hematopoietic cells, and may help to develop strategies by which unstable expression of transduced genes can be overcome.

A novel dual function retrovirus expressing multidrug resistance 1 and O6-alkylguanine-DNA-alkyltransferase for engineering resistance of haemopoietic progenitor cells to multiple chemotherapeutic agents

Following transduction with a retrovirus (SF1MIH) expressing both the multidrug resistance 1 (MDR1) and O6-alkylguanine-DNA-alkyltransferase (ATase) proteins, human erythroleukaemic progenitor (K562) cells were isolated which were resistant to killing by the MDR1 substrate, colchicine. In colony-forming survival assays, K562-SF1MIH cells exhibited resistance to colchicine and doxorubicin, as well as to the O6-alkylating agents N-Methyl-N-nitrosourea (MNU) and temozolomide. Furthermore, the resistance to doxorubicin was abolished by preincubation with the MDR1 inhibitor verapamil while resistance to MNU was ablated by the specific ATase inactivator, O6-benzylguanine (O6-beG) confirming that resistance to doxorubicin and MNU was conferred by MDR1 and ATase, respectively. When K562-SF1MIH were exposed to combinations of colchicine and MNU or doxorubicin and temozolomide, simultaneous resistance to these agents was observed. Thus, transduction of K562 with SF1MIH conferred dual resistance to these cells. These data offer the prospect of designing vectors that will confer resistance to entire regimens of chemotherapy rather than just to individual components of such drug cocktails, thereby substantially increasing the efficacy of therapy. Furthermore, the use of such dual expression constructs is likely to be highly informative for the design of effective in vivo selection protocols, an issue likely to make a major impact in a clinical context in gene therapy in the near future.

Efficient gene transfer by hybrid retroviral vectors to murine spermatogenic cells

Using murine spermatogenic cell lines GC-1 spg and GC-2 spd(ts) as target cells, an attempt was made to design a retroviral vector that would transduce genes efficiently. Promoter activities of various retroviral long terminal repeats (LTRs) were examined by using chloramphenicol acetyltransferase (CAT) as a reporter. The U3 region of spleen focus-forming virus (SFFVp) showed higher enhancer activity than that of Moloney murine leukemia virus (Mo-MuLV) in both cell lines. The U3 region of myeloproliferative sarcoma virus (MPSV) showed higher activity only in GC-1 spg cells. Expression was suppressed by the repressor element of the primer-binding site (PBS) of the Moloney-related virus. The efficiency of transduction of the multidrug-resistance gene (mdr-1) by an Mo-MuLV-based vector was compared with hybrid vectors consisting of the murine embryonic stem cell virus (MESV) PBS and the LTR of either SFFVp or MPSV. Rhodamine efflux assays and colchicine-resistant colony-forming assays demonstrated higher gene expression by the hybrid vectors. Amphotropic and ecotropic receptors were found to be expressed and functional in both cell lines. Thus, these hybrid vectors represent a powerful tool by which to transfer genes into spermatogenic cells.

Bicistronic retroviral vectors for combining myeloprotection with cell-surface marking

We have developed a retroviral vector coexpressing the multidrug-resistance 1 (MDR1) cDNA for inducing cancer drug resistance and the truncated version of the low-affinity nerve growth factor receptor (DeltaLNGFR) for cell-surface marking of transduced cells. The vector is based on the FMEV backbone which mediates high levels of gene expression in hematopoietic cells. To achieve optimal expression levels of both cDNAs, untranslated regions from MDR1 and DeltaLNGFR were removed and three different connections were tested: retroviral splice signals, an internal ribosomal entry site (IRES) from encephalomyocarditis virus, and an internal promoter from the chicken beta-actin gene. As determined by two-color flow cytometry, the best correlation of the expression of both cDNAs was obtained using the vector SF1mSdelta which utilized retroviral splice signals for co-expression. Simultaneous expression of both cDNAs at the single cell level was also shown by confocal laser microscopy. Lymphoid and hematopoietic progenitor cells, including primary human CD34+ cells, transduced with SF1mSdelta acquired dominant multidrug resistance. Transduced primary CD34+ cells could be enriched in vitro based on expression of DeltaLNGFR, avoiding exposure to cytostatic agents. Thus, monitoring the selection of chemotherapy-resistant cells and analyzing their biological properties may be alleviated, both in vitro and in vivo.

Soluble bone marrow stroma factors improve the efficiency of retroviral transfer of the human multidrug resistance 1 gene to human mobilized peripheral blood progenitor cells

Hematopoietic stem cells (HSCs) are a potential target for the retrovirus-mediated transfer of chemotherapeutic drug resistance genes. For integration of the proviral DNA in the HSC genome cell division is required. In the bone marrow (BM) hematopoiesis occurs in the vicinity of stroma cells. Soluble stroma components were shown to play a permissive role for the proliferation of lineage-committed and primitive hematopoietic progenitors in conjunction with cytokines. We investigated the effect of stroma-conditioned medium (SCM) of the FBMD1 cell line on the gene transfer rate of the human multidrug resistance 1 (MDR1) gene contained in the retroviral SF-MDR vector into human mobilized peripheral blood progenitor cells (PBPCs) from tumor patients (n = 14) during transwell transduction in the presence of the recombinant fibronectin fragment CH-296. Addition of SCM during transduction increased the gene transfer efficiency into myeloid lineage-committed colony-forming cells by an average of 1.5-fold (p = 0.02) as detected by an SF-MDR provirus-specific polymerase chain reaction (PCR). These data were paralleled by significantly (p = 0.04 to p = 0.007) higher proportions of MDR1-expressing myelo-monocytic progeny after transduction in SCM plus interleukin 3 (IL-3), IL-3/Flt3 ligand (FL), IL-3/IL-6/FL, or IL-3/IL-6/stem cell factor (SCF) when compared with transductions without SCM as measured by rhodamine-123 exclusion. A similar trend was observed for SCM employed in combination with IL-3/IL-6/SCF/FL or FL/thrombopoietin (TPO)/SCF during transduction. The latter combination plus SCM yielded the highest proportion, 19.16 +/- 3.10% Rh-123dull cells. The beneficial effect of SCM on transduction efficiency was confirmed in additional four patients' samples, using a serum-free viral supernatant transduction protocol. As soluble BM stroma factors are able to increase the efficiency of retrovirus-mediated gene transfer into committed progenitor cells, beyond that achieved with fibronectin fragment CH-296, their effect on gene transfer into primitive repopulating hematopoietic cells may also prove beneficial.

Internal ribosome entry sites (IRESs): reality and use

IRESs are known to recruit ribosomes directly, without a previous scanning of untranslated region of mRNA by the ribosomes. IRESs have been found in a number of viral and cellular mRNAs. Experimentally, IRESs are commonly used to direct the expression of the second cistrons of bicistronic mRNAs. The mechanism of action of IRESs is not fully understood and a certain number of laboratories were not successful in using them in a reliable manner. Three observations done in our laboratory suggested that IRESs might not work as functionally as it was generally believed. Stem loops added before IRESs inhibited mRNA translation. When added into bicistronic mRNAs, IRESs initiated translation of the second cistrons efficiently only when the intercistronic region contained about 80 nucleotides, and they did not work any more effectively with intercistronic regions containing at least 300-400 nucleotides. Conversely, IRESs inserted at any position into the coding region of a cistron interrupted its translation and initiated translation of the following cistron. The first two data are hardly compatible with the idea that IRESs are able to recruit ribosomes without using the classical scanning mechanism. IRESs are highly structured and cannot be scanned by the 40S ribosomal subunit. We suggest that IRESs are short-circuited and are essentially potent stimulators favoring translation in particular physiological situations.

Design of 5' untranslated sequences in retroviral vectors developed for medical use

Utilizing genetic modules of simple retroviruses, we have developed a novel generation of gene transfer vectors with improved therapeutic potential. In the 5' untranslated "leader" sequences, all AUG codons which may aberrantly initiate translation and all viral coding sequences were removed. Thus, the probability of expressing unwanted peptides and the potential for homologous recombination with retroviral genes were largely reduced, and the cloning capacity was increased. The transgene was inserted to replace the viral gag sequences, and a new minimal splice acceptor was introduced, resulting in increased expression with all genes tested (those coding for human multidrug resistance 1 and enhanced green fluorescent protein, as well as the lacZ gene). These vectors may represent attractive tools for human gene therapy, because they increase the efficiency of transgene expression and may also increase safety in medical applications.

Retrovirus-mediated gene transfer of the human multidrug resistance-associated protein into hematopoietic cells protects mice from chemotherapy-induced leukopenia

Utilization of chemotherapy for the treatment of tumors is mainly limited by its hematological toxicity. Because of the low-level expression of drug resistance genes, transduction of hematopoietic progenitors with multidrug resistance 1 (MDR1) or multidrug resistance-associated protein (MRP) genes should provide protection from chemotherapeutic agent toxicity. Successful transfer of drug resistance genes into hematopoietic cells may allow the administration of higher doses of chemotherapy and, thus, increase regression of chemosensitive tumors. The interest in the use of MRP as an alternative to MDR1 for bone marrow protection lies in its different modulation. This would allow, in the same patient, the use of MDR1 reversal agents to decrease MDR1 tumor resistance without reversing bone marrow (BM) protection of the MRP-transduced hematopoietic cells, since MRP expression is not reversed by these agents. We have constructed MRP-containing retroviral vectors using the phosphoglycerate kinase promoter and generated ecotropic producer cells. Lethally irradiated mice were engrafted with BM cells transduced by coculture with MRP producer cells. Evidence of long-term (9 months) gene transfer was provided by PCR of peripheral blood from MRP-transduced mice. Southern blot analysis confirmed the integrity of the provirus in the MRP-transduced mice. Long-term MRP expression (>5 months) was detected by RT-PCR and fluorescence-activated cell sorting of blood from living mice. High-level expression of MRP in murine hematopoietic cells reduces doxorubicin-induced leukopenia and mortality. Furthermore, we show in vivo selection of MRP-transduced cells following doxorubicin administration, with better and more significant chemoprotection after the second chemotherapy cycle. These data indicate that MRP retroviral gene transfer may be useful for chemoprotection and selection.

Tetracycline-inducible expression systems with reduced basal activity in mammalian cells

We describe a modification of the tetracycline-inducible eukaryotic gene expression system with decreased basal levels of expression in HeLa cells. It employs the tetracycline-inducible transactivator and a tetracycline-regulated repressor fusion acting on the same promoter. To avoid heterodimerization or competition for the same DNA site, each was provided with different DNA recognition and/or protein dimerization specificities. We achieved active silencing in the uninduced state resulting in approximately 6-fold reduced levels of basal transcription and several hundred-fold activation of gene expression upon addition of tetracycline.

Poor expression of MDR1 transgene in HeLa cells by bicistronic Moloney murine leukemia virus-based vector

Cotransfer of a therapeutic gene together with the human MDR1 gene provides an opportunity to increase the number of transduced marrow cells, expressing the therapeutic gene, by in vivo selection for MDR1. We have used an Lg-MDR1-IRES-neo (LgMIN) retroviral vector, containing MDR1 and neo genes, separated by the EMCV IRES. Human HeLa or canine CTAC cells, transduced with GALV env pseudotyped LgMIN at an MOI of less than 0.01 to ensure 1 proviral copy/genome, were selected with either G418 for neo expression or colchicine for MDR1 expression. The titer determined on HeLa cells with G418 selection was eight-fold higher than that with colchicine selection. In contrast, the same viral supernatant exhibited only a 1.4-fold difference between neo- and MDR1-based viral titer values for CTAC cells. The transduced HeLa cells, with one intact proviral copy per genome, exhibited a 55-fold higher resistance to G418 but only a 4-fold higher resistance to colchicine and a 2-fold higher resistance to Taxol compared with nontransduced cells. About 23% of the transduced cell population did not express vector-derived P-glycoprotein (P-gp) as detected by anti-human P-gp MAb MRK-16. This could explain the difference in viral titers obtained on CTAC cells but not that obtained on HeLa cells. The vector-mediated increase in expression of P-gp was about 20-fold higher in CTAC cells as compared with HeLa cells. These results indicated suppression of expression of vector-derived MDR1 in HeLa cells, in contrast with CTAC cells. To investigate further the possible reasons for this difference, genomic DNA was isolated from the G418-resistant individual colonies of infected cells and analyzed by PCR for full-length proviral MDR1. For transduced CTAC and HeLa cells, selected at a G418 concentration of 1 mg/ml, PCR detected aberrant forms of MDR1 in 17 to 25% of colonies tested. The aberrant forms consisted of MDR1 genes with 2- and 0.7-kb deletions. DNA sequencing across the 2-kb and the 0.7-kb deletion junction suggests cryptic splicing in the producer cell line as the origin of these deletions. The 2-kb deletion corresponds to MDR1 mRNA cryptic splicing via donor (codon 113) and acceptor (codon 773). The 0.7-kb deletion corresponds to splicing via the same donor and a different acceptor (codon 344). When transduced HeLa cells were selected at a higher concentration of G418 (3 mg/ml), the aberrant forms were detected at an increased frequency of about 50% of colonies tested. These results indicate that vector-derived MDR1 is a poor selective marker in HeLa cells but not in CTAC cells and that deletions, which inactivated the MDR1 gene in a bicistronic Mo-MuLV vector, may provide an advantage for expression of the second transgene in HeLa cells.