TEXT

Top Abstract Text References

Retroviral vectors have become important tools for the study of biology and for the transfer of genes for therapeutic purposes. A crucial step in gene transfer by retroviral vectors is the binding of the virus envelope protein (Env) to a specific cellular receptor. Because cells vary in their expression of retroviral receptors, it is important to have available a variety of packaging cells that express different envelopes that in turn recognize different receptors. Viral interference experiments show that the recently described Mus dunni endogenous virus (MDEV) uses a novel receptor among murine retroviruses (10). Retroviral vectors pseudotyped by wild-type MDEV can infect different cell types from a variety of species, including mouse, rat, hamster, cat, dog, quail, and human (2). We have therefore constructed packaging cells that express the MDEV envelope to take advantage of its novel receptor usage and have evaluated the potential of vectors packaged by PD223 for use in hematopoietic gene therapy.

Generation of MDEV Env expression plasmids. The original molecular clones of MDEV contained a frameshift mutation in the envelope gene that precluded expression of complete Env proteins (15). Intact env genes were obtained by several methods and cloned into the expression plasmid pSX2 (Fig. 1), which was chosen because it was found to express the highest levels of 10A1 murine leukemia virus (MLV) Env among a series of constructions (7). pMDEV9ex contains an MDEV env from the original clone that was corrected by site-directed mutagenesis (15).

View larger version (24K): [in this window] [in a new window]   FIG. 1.   Envelope expression constructs. The constructs are drawn to scale, and the backbone plasmid sequences are not shown. Each plasmid contains the MoMLV LTR promoter (truncated at the 5' end to the Sau3AI site upstream of the enhancers), the MoMLV splice donor, the 10A1 MLV splice acceptor, an envelope coding sequence, and the simian virus 40 (SV40) early polyadenylation sequence. The pMDEV9ex plasmid was created by cloning a 2,165-bp BamHI-XmnI fragment containing the corrected MDEV env into the 3,948-bp fragment of pSX2 prepared by digestion with BsaBI and partial digestion with BamHI. The structure of pMEX represents 20 plasmids in which the MDEV env was amplified from various sources and cloned into the 3,835-bp DraIII-BsaBI fragment of pSX2.

Evaluation of MDEV Env expression plasmids. LGPS/LAPSN cells are NIH 3T3 cells that express Moloney murine leukemia virus (MoMLV) Gag and Pol (8) and contain the LAPSN retroviral vector, which expresses a human placental alkaline phosphatase (AP) gene (11). These cells secrete viral particles containing the LAPSN genome, but the particles are not infectious due to the lack of Env proteins. The MDEV Env expression plasmids were evaluated by transfecting them into LGPS/LAPSN cells and measuring the titer of infectious LAPSN vector particles 2 days after transfection. A preliminary experiment used to screen the 20 pMEX clones showed that all 5 clones containing the MDEV env amplified from plasmid DNA were functional, 5 out of 6 clones amplified from G355/LAPSN+MDEV cellular DNA were functional, but only 2 out of 9 clones amplified from dunni cell DNA were functional (data not shown). We have shown that elements related to MDEV exist in the M. dunni genome (2), and some of these nonfunctional pMEX plasmids may carry related but defective envelopes. We have not sequenced any of the nonfunctional plasmids.

Creation of the MDEV packaging cells and selection of clones. To generate the MDEV packaging cells, one of the active pMEX^dunni clones was stably introduced into LGPS cells (NIH 3T3 cells that express MoMLV Gag and Pol [9]) by cotransfection with the selectable plasmid pSV213-hyg (a gift from Paul Berg at Stanford University) at a ratio of 20:1. The plasmid pMEX^dunni was chosen because it likely contains the env sequence identical to that of the native M. dunni provirus. After transfection, the cells were selected in 0.4 mg of hygromycin per ml for 7 days, and 38 clones were isolated. The clones were designated "PD," for packaging cells based on the M. dunni endogenous virus envelope. Two screens confirmed that the PD clone 223 (PD223) packaged the LAPSN vector at the highest titer [>10⁵ AP⁺ focus-forming units (FFU)/ml]. Upon determining that PD223 packaged LAPSN at the highest titer, we cloned the PD223/LAPSN cells to obtain a high titer producer line which consistently produces the LAPSN vector at a titer of 4 × 10⁵ AP⁺ FFU/ml measured on D17 canine cells.

A vector packaged by PD223 cells uses the same receptor as MDEV. Viral interference analysis is used to classify retroviruses into groups according to receptor usage and relies on the observation that a cell producing a retroviral Env is resistant to infection by a retrovirus that recognizes the same receptor as the expressed Env. Interference assays were conducted to determine whether a vector packaged by the PD223 cells uses the same receptor as MDEV (Table 2). The LAPSN(PD223) and LAPSN(PA317) vectors contain the same MoMLV Gag and Pol proteins, so any differences observed are Env-specific and therefore at the level of viral entry. The entry of LAPSN(PA317) was not impeded by the presence of MDEV in dunni/N2 cells nor by the expression of retroviral proteins in the PD223 cells (which are based on NIH 3T3 cells), as expected, indicating that the amphotropic MLV envelope expressed by PA317 cells recognizes a different receptor than does MDEV. However, the entry of LAPSN(PD223) was dramatically impeded by the presence of MDEV in dunni/N2 cells (>24,000-fold interference), and the entry of LAPSN(MDEV) was impeded by the presence of retroviral constructs in PD223 cells (400-fold interference), indicating that the Env proteins contained on LAPSN(PD223) virions recognize the same receptor as MDEV.

PD223 cells package vectors without contaminating helper virus. A risk inherent to packaging cells is the generation of contaminating replication-competent retroviruses (RCR), or helper virus. One potential source of RCR in other packaging cells is homologous recombination among the engineered sequences in the packaging cells. This could not generate RCR in the PD223 cells due to lack of sequence similarity or overlap in the appropriate regions of the constructs. A homologous recombination event could join a long terminal repeat (LTR) from an LN-series vector (9) with the gag and pol sequences of pLGPS to generate the 5' half of an RCR, but a homologous recombination event between pLGPS and pMEX^dunni could not complete the set of retroviral genes due to a deletion of pol in pMEX^dunni. Additionally, there is no similarity between the 3' end of the MDEV env in pMEX^dunni and an LN-series vector that could supply a 3' LTR to a recombinant virus. However, incompletely characterized endogenous retroviral sequences could be involved in generation of RCR. Furthermore, rarer recombination events that involve no sequence homology could occur, making it necessary to directly test for RCR.

PD223 cells are useful for transduction of CHO cells. CHO cells provide an important tool for genetic analysis due to their functionally haploid nature (13) and are widely used in biotechnology for production of glycosylated therapeutic proteins. Unfortunately, CHO cells are relatively resistant to transduction by a variety of retroviral vectors, with a block at the level of viral entry (12). We have previously demonstrated that a vector pseudotyped by wild-type MDEV could efficiently transduce CHO cells, so we compared the LAPSN vector packaged by PD223 cells to LAPSN packaged by seven other packaging cells that express different envelopes for the ability to transduce CHO cells (Table 3). The titers of LAPSN packaged by the various cell lines were generally low when measured with CHO cells, as predicted. However, the titer on permissive cells (D17 or NIH 3T3) was high for each stock, demonstrating that there was functional virus present. The most efficient entry into CHO cells was achieved by LAPSN(PD223), which transduced CHO at least 25-fold more efficiently than vectors packaged by any of the seven other packaging cells.

Vectors packaged by PD223 cells transduce primary baboon and human CD34⁺ cells. Hematopoietic stem cells are an attractive target for gene therapy. Cells selected for the CD34 surface antigen contain hematopoietic stem cells and have been used here as a model for hematopoietic stem and progenitor cell transduction. Primary baboon and human CD34⁺ cells were transduced with LAPSN pseudotyped by PD223 and PG13 (8) packaging cells. PG13 is a packaging cell line that is preferred for transduction of baboon and human CD34⁺ cells (5). Additionally, the FLYRD packaging cell line was included in the human cell experiment because it shows promise for transduction of human cells (3); however, it does not infect baboon cells (data not shown). Flow cytometry using an anti-human placental AP antibody revealed two distinct populations of cells, allowing the determination of the percentage of transduced cells. In the baboon cell experiment, the LAPSN packaged by PD223 transduced the CD34⁺ cells with about twofold lower efficiency than did LAPSN packaged by PG13. In the human cell experiment, the LAPSN packaged by PD223 transduced the cells as efficiently as did LAPSN packaged by PG13 and more efficiently than did LAPSN packaged by FLYRD (Fig. 2). In both experiments, mock-transduced CD34⁺ cells showed 0.05% transduction. The rates of transduction in these experiments were lower than those we usually obtain, but are within the range typically observed for these assays. Because CD34⁺ cells are a heterogeneous population, it is possible that the vectors with different pseudotypes transduced different subpopulations of cells. The only true assay of a stem cell is the measurement of its ability to reconstitute hematopoiesis in a myeloablated animal, so additional experiments will be required to further assess the utility of PD223 cells for use in hematopoietic gene therapy.

View larger version (17K): [in this window] [in a new window]   FIG. 2.   The LAPSN vector packaged by PD223 cells transduces baboon and human CD34+ cells relatively efficiently. Fresh normal bone marrow cells were selected for the presence of the CD34+ antigen by using an immunomagnetic column procedure (7). On day 1, the cells were prestimulated with flt-3 ligand, interleukin 6, stem cell factor (100 ng/ml each), and megakaryocyte growth and development factor (50 ng/ml) during culture in RPMI medium plus 10% fetal bovine serum (FBS). On day 2, cells were plated at 50,000 cells per well in a 24-well Pronectin F culture plate (for the experiment with human cells) (Protein Polymer Technologies) or in the presence of fibronectin fragment CH-296 (for the experiment with baboon cells) (Retronectin; Takara Shuzo Co., Ltd., Shiga, Japan) (4), in 500 µl of medium and exposed to 500 µl of LAPSN(PD223), LAPSN(PG13), or LAPSN(FLYRD) in the presence of Polybrene (4 µg/ml). Fresh virus was added on day 3, and the percentage of AP-expressing cells was evaluated by flow immunocytometry on day 5. For analysis, the cells were harvested, resuspended in hybridoma 2.4G2-conditioned medium to block FcII receptors (14), and labeled using anti-human placental alkaline phosphatase (clone 8B6; Dako, Carpinteria, Calif.) which had been conjugated with biotin. After washing, the cells were stained using streptavidin phycoerythrin (Dako) and further washed. Prior to analysis, cells were resuspended in phosphate-buffered saline containing 2% FBS and propidium iodide (1 µg/ml; Molecular Probes, Eugene, Oreg.). Analysis was performed on a FACSCalibur (Becton Dickinson, San Jose, Calif.) using CellQuest II software. Dead cells and debris were excluded by gating on forward and high angle light scatter and by the absence of propidium iodide staining. Each value represents the average of three separate transductions, with the standard deviations indicated. Human cells were obtained by using a protocol approved by the Institutional Review Board, and baboon cells were obtained by using a protocol approved by the American Association for the Accreditation of Laboratory Animal Care.