Pre-clinical studies indicate that efficient retrovirus-mediated gene transfer into hematopoietic stem cells and progenitor cells can be achieved by co-localizing retroviral particles and target cells on specific adhesion domains of fibronectin. In this pilot study, we used this technique to transfer the human multidrug resistance 1 gene into stem and progenitor cells of patients with germ cell tumors undergoing autologous transplantation. There was efficient gene transfer into stem and progenitor cells in the presence of recombinant fibronectin fragment CH-296. The infusion of these cells was associated with no harmful effects and led to prompt hematopoietic recovery. There was in vivo vector expression, but it may have been limited by the high rate of aberrant splicing of the multidrug resistance 1 gene in the vector. Gene marking has persisted more than a year at levels higher than previously reported in humans.

These results indicate the feasibility and safety of bone marrow gene therapy with a potentially therapeutic gene, the MDR gene.

The human multiple drug resistance (MDR) gene has been used as a selectable marker to increase the proportion of bone marrow cells that contain and express this gene by drug selection. By constructing retroviral vectors containing and expressing the MDR gene and a nonselectable gene such as the f-globin gene, enrichment for cells contining both of these genes can be achieved. A retroviral construct containing MDR cDNA in a Harvey virus-based vector has been used to transfect our ecotropic 3T3 retroviral packagin line GP+E86. Clones have been isolated by exposure of the retrovirally transfected cells (MDR producer cells) to colchicine (60 ng/ml), a selective agent that kills MDR-negative cells. Flow cytometry analysis (fluorescence-activated cell sorting) with an antibody to MDR demonstrates expression of human MDR protein on the surface of these colchicine-resistant producer clones. Untransfected GP+E86 cells are negative. Colchicineresistant clones were titered using clone supernatants and the highest titer clone (4 x 104 viral particles per ml) was cocultured with 10' donor mouse bone marrow cells for 24-48 hr.The donor cells were then injected into congenic irradiated mice, and the presence of the MDR gene was assayed by the polymerase chain reaction (PCR) analysis using MDR-specifIc primers. In one experiment eight of nine transduced mice were positive for MDR by PCR of peripheral blood 14 and 50 days posttransplantation; after 240 days three of nine transduced mice were positive. Bone marrow obtained from one of these positive animals was stained with the MDR monoclonal antibody and the granulocyte population was analyzed by FACS.

Insertion of a normally functioning human beta-globin gene into the hematopoietic stem cells (HSC) of patients with beta-thalassemia may be an effective approach to the therapy of this disorder. Safe, efficient gene transfer and long-term, high-level expression of the transferred human beta-globin gene in animal models are prerequisites for HSC somatic gene therapy. We have recently shown for the first time that, using a modified beta-globin retroviral vector in a mouse transplant model, long-term, high-level expression of a transferred human beta-globin gene is possible. The human beta-globin gene continues to be detected up to eight months post-transplantation of beta-globin-transduced hematopoietic cells into lethally irradiated mice. The transferred human beta-globin gene is detected in three of five mice surviving long-term (> 4 months) transplanted with bone marrow cells transduced with high-titer virus. The unrearranged 5.1 kb human beta-globin gene-containing provirus is seen by Southern blotting in two of these mice. More importantly, long-term expression of the transferred gene is seen in two mice at levels of 5% and 20% that of endogenous murine beta-globin. We document stem cell transduction by showing continued high-level expression of the human beta-globin gene in secondarily transplanted recipient mice. These results provide evidence of HSC transduction with a human beta-globin gene in animals and demonstrate that retroviral-mediated unrearranged human beta-globin gene transfer leads to a high level of human beta-globin gene expression in the long term for the first time. A gene therapy strategy may be a feasible therapeutic approach to the beta-thalassemias if consistent human beta-globin gene transfer and expression into HSC can be achieved.

We have performed a pilot study of MDR-1 gene transfer chain reaction (PCR) for vector-derived MDR-1 cDNA in patients receiving CD34-selected peripheral blood stem sequence in all cases. Analysis of peripheral blood and cell (PBSC) transplant for lymphoma. To ensure minimum bone marrow cells after transplant has, however, shown engraftment thresholds and facilitate CD34 purification, no evidence of in vivo gene transfer with a follow-up of 12, mobilisation of Ͼ2 × 10 6 CD34 cells/kg was a condition for 15 and 18 months. The effect of MDR-1 substrate drugs recruitment. Of 11 patients counselled for study entry, onlyhas not yet been tested as all patients remain in clinical five achieved this target in a single apheresis. In three conand radiological remission of their lymphoma. These senting patients, purified CD34 cells were exposed to results confirm the difficulty of achieving in vivo gene trans-A12M1 MDR-1 retroviral supernatant for 6 h, cryoprefer in human haemopoietic cells and indicate major logisserved then thawed and readministered following ablative tical constraints in PBSC mobilisation in patients with chemotherapy. No delay in engraftment was observed, relapsed and resistant disease in whom initial studies are although one patient received additional back-up cells.appropriate. Gene transfer was demonstrated by polymerase

Insertion of a normally functioning human β-globin gene into the hematopoietic stem cells (HSC) of patients with β-thalassemia may be an effective approach to the therapy of this disorder. Safe, efficient gene transfer and long-term, high-level expression of the transferred human β-globin gene in animal models are prerequisites for HSC somatic gene therapy. We have recently shown for the first time that, using a modified β-globin retroviral vector in a mouse transplant model, longterm, high-level expression of a transferred human β-globin gene is possible. The human β-globin gene continues to be detected up to eight months post-transplantation of β-globin-transduced hematopoietic cells into lethally irradiated mice. The transferred human β-globin gene is detected in three of five mice surviving long-term (>4 months) transplanted with bone marrow cells transduced with high-titer virus. The unrearranged 5.1 kb human β-globin gene-containing provirus is seen by Southern blotting in two of these mice. More importantly, long-term expression of the transferred gene is seen in two mice at levels of 5% and 20% that of endogenous murine β-globin. We document stem cell transduction by showing continued highlevel expression of the human β-globin gene in secondarily transplanted recipient mice. These results provide evidence of HSC transduction with a human β-globin gene in animals and demonstrate that retroviral-mediated unrearranged human β-globin gene transfer leads to a high level of human β-globin gene expression in the long term for the first time. A gene therapy strategy may be a feasible therapeutic approach to the β-thalassemias if consistent human β-globin gene transfer and expression into HSC can be achieved. P atients with β-thalassemia can be cured by allogeneic stem cell transplantation. 1,2 However, this treatment modality is associated with significant morbidity and mortality and is limited by the number of available HLA-compatible siblings. 3 Another potential approach to curing β-thalassemia is by the insertion of a normally functioning human β-globin gene into a 180 ANNALS NEW YORK ACADEMY OF SCIENCES FIGURE 1. A diagram of the p141 vector. The vector is flanked by the 5′ long terminal repeat (LTR) and 3′ deleted long terminal repeat (∆ LTR). The extended packaging signal is represented by ψ+. The β-globin gene is in reverse orientation with exons ( ), introns ( ), the deleted 3′ enhancer ( ) and the 372 bp deletion in intron 2 (∆ Rsa I). The LCR core sequences HS234 ( ) and the neomycin resistance gene (NEO) with the PGK promoter ( ) are as indicated.

One of the requirements for the use of retroviral vectors in human gene therapy is a packaging cell line which is incapable of producing replication-competent virus and which produces high titers of replication-deficient vector virus. Wild-type virus may be produced through recombinational events between the helper virus and a retroviral vector. We have constructed an ecotropic packaging cell line, GP + E-86, and an amphotropic packaging cell line, GP + envAm12, in which the viral gag and pol genes are on one plasmid and the viral env gene is on another plasmid. Both plasmids contain deletions of the packaging sequence and the 3' LTR. The fragmented helper virus genomes, when introduced into 3T3 cells, produce titers of retrovirus which are comparable to the titers produced from packaging cells containing the helper virus genome on a single plasmid. We have found no evidence for the generation of wild-type retrovirus using the GP + E-86 and GP + envAm12 packaging lines, either alone or in combination with the N2 retroviral vector. We also show that these packaging cell lines can be used to transfer the neoR gene of the N2 vector into mouse hematopoietic cells, followed by successful (48-52%), long-term (up to 200 days) transplantation into irradiated recipients. These results indicate that these packaging lines are safe and efficient for use in experiments designed for murine (using GP + E-86) and human (using GP + envAm12) gene therapy.

Lentiviral vectors derived from the HIV-1 genome offer great promise for gene therapy due to their ability to transduce non-dividing cells and sustain long-term expression of transgenes. The majority of current lentiviral vectors are pseudotyped with the vesicular stomatitis viral envelope (VSV-G). VSV-G equips lentiviral vectors with a broad host cell tropism and increased stability. Increased particle stability enables viral supernatants to be concentrated by high-speed centrifugation to enhance their infectivity. Despite its efficacy, VSV-G is cytotoxic - a feature that prohibits the development of stable cell lines that constitutively express this envelope. Therefore, non-toxic envelope proteins are being investigated. RD114 is an attractive alternative because it also provides increased particle stability and its receptor is widely expressed on hematopoietic stem cells (HSCs). In this study, the packaging efficiency of three envelope proteins, RD114, RDpro and VSV-G, were evaluated with two lentiviral vectors (TRIP GFP and HPV-402). RDpro is an RD114-HIV chimera designed to pseudotype lentiviral vectors more efficiently. In transient systems, VSV-G generated titers of 10(8) and 10(7) viral particles/mL for TRIP GFP and HPV-402. RDpro possessed titers of 10(7) and 10(6), while RD114 titers were one log lower for each vector. Despite having relatively lower titers, RD114 proteins are less toxic; this was demonstrated in the extension of transient transfection reactions from 48 to 96 h. VSV-G transfections are generally limited to 48 h. In regard to gene therapy applications, we show that RDpro supernatants efficiently transduce peripheral blood HSCs. The versatility of RD114 envelopes was again demonstrated by using a 'mixed' expression system; composed of stably expressed RD114 envelope proteins to pseudotype lentiviral vectors generated in trans (titer range 10(3)-10(5)). Our data show that RD114 envelope proteins are effective and versatile constructs that could prove to be essential components of therapeutic lentiviral gene transfer systems.