TEXT

Top Abstract Text References

Gene transfer into hematopoietic stem cells (HSCs) of humans and large animals has been inefficient using oncoretrovirus vectors (18, 19, 25). This is to a large extent due to the quiescent nature of HSCs since Moloney murine leukemia virus (MMLV)-based vectors require dividing target cells for successful transduction (28). This has led to a recent interest in lentivirus vectors as potential gene delivery vehicles to human HSCs (6, 22, 33), since these have the ability to transduce quiescent cells (1, 5, 15, 23, 24, 31). To study gene expression characteristics of lentivirus vectors in hematopoietic cells, we transduced embryonic stem (ES) cells with lentivirus vectors and developed lentivirus-transduced ES cell clones that subsequently differentiated into hematopoietic cells. In this fashion, we could examine whether persistence of gene expression can be maintained during development of lentivirus-transduced ES cells from the undifferentiated state to hematopoietic colonies in vitro. Similarly, we were able to investigate whether the level of expression is copy number related or whether it is subjected to position effects (integration site dependent).

MMLV vectors can transduce ES cells efficiently, but expression from the viral long terminal repeat (LTR) is not active due to transcriptional silencing immediately following infection which is attributable to trans-acting factors (12, 21, 30). A recombinant viral LTR in a virus called the murine ES cell virus (MESV) allows effective expression in undifferentiated ES cells (12). This virus has a high-affinity binding site for the Sp1 transcription factor (13, 27), elimination of a binding site in the LTR for a transcriptional repressor (4, 12, 13, 34), and elimination of the negative regulatory element in the primer binding site (2, 12, 21, 35). These changes in and around the LTR do not allow expression in transduced ES cells that are differentiated in vitro, including cells that are differentiated into hematopoietic cells. During the differentiation process, expression is turned off by a cis-acting mechanism (20). We decided to examine whether a human immunodeficiency virus type 1 (HIV-1)-based vector that lacks the tat gene and uses an internal phosphoglycerate kinase (PGK) promoter to drive the enhanced green fluorescent protein (GFP) gene would be able to express the gene throughout differentiation from undifferentiated ES cells through embryonic bodies (EBs) and to hematopoietic colonies grown in methylcellulose (16, 17, 36). Our hypothesis was that the internal promoter would be active and this experimental model system could be used to study lentivirus gene expression in hematopoietic cells in vitro following transduction of ES cells. The basis for our hypothesis was that early work using MMLV vectors with internal promoters in embryonic carcinoma cells or after infection of mouse embryos showed expression from internal promoters (29, 34). Also, there is no need for active transcriptional repression of the HIV-1 LTR by the cells since without the Tat protein, the HIV-1 LTR is transcriptionally inactive, although this does not exclude an additional inactivation mechanism by the cell (8, 9).

Lentivirus vector gene transfer efficiency of ES cells. We first examined whether it was possible to transduce ES cells by using lentivirus vectors and, if so, how the transduction efficiency compared that of to standard oncoretrovirus vectors. The feeder-independent CCE ES cells were transduced for 48 h by the lentivirus vector HIV-PGK-GFP packaged in 293T cells with vesicular stomatitis virus G protein (VSV-G) and the oncoretrovirus vector MGirL22Y packaged in the ecotropic packaging cell line GP+E86. The HIV-PGK-GFP vector has the same backbone structure as HR'CMV-GFP (38), where the PGK promoter replaces the cytomegalovirus promoter. The structures of these vectors are shown in Fig. 1. The lentivirus vector HIV-PGK-GFP was produced by transient transfection into 293T cells as previously described (3, 7, 24). A total of 40 µg of plasmid DNA was used for transfection of one 10-cm-diameter dish: 10 µg of the envelope plasmid pMD.G encoding VSV-G, 10 µg of packaging plasmid pCMVR 8.91 (37) expressing Gag, Pol, Tat, and Rev, and 20 µg of vector plasmid HIV-PGK-GFP, which contained the enhanced GFP gene under the control of the murine PGK promoter (Fig. 1). The conditioned medium was collected after 96 h, filtered through 0.45-µm-pore-size cellulose acetate filters (Nalge Company, Rochester, N.Y.), and ultracentrifuged with a Beckman T28W rotor at 50,000 × g for 1.5 h. The oncoretrovirus vector MGirL22Y (Fig. 1) was harvested from the conditioned medium of MGirL22Y producer cells (kindly provided by D. A. Persons and A. W. Nienhuis, St. Jude Children's Hospital, Memphis, Tenn.) (26). The MGirL22Y vector contains the MESV LTR (12) to drive the GFP gene. One million MGirL22Y cells were seeded on a 10-cm-diameter dish; the conditioned medium was collected after 96 h, filtered through 0.45-µm-pore-size cellulose acetate filters, and concentrated in a Centricon (Amicon, Beverly, Mass.). The titer of the vectors was determined on NIH 3T3 cells. NIH 3T3 cells (10⁵) were plated in 12-well plates and transduced in the presence of protamine sulfate (Sigma Chemical Co., St. Louis, Mo.). GFP expression was evaluated by fluorescence-activated cell sorting (FACS) analysis in a FACSCalibur (Becton Dickinson Immunocytometry Systems, San Jose, Calif.) after 48 h, and virus titers were estimated. The lentivirus and oncoretrovirus vector titers on NIH 3T3 cells were 1.18 × 10⁷ and 1.70 × 10⁷ TU/ml, respectively. CCE ES cells were maintained in Dulbecco's modified Eagle's medium with 15% fetal calf serum (FCS; GIBCO Life Technologies, Gaithersburg, Md.), 1 mM glutamine (GIBCO), 1 mM sodium pyruvate (GIBCO), 150 mM -monothioglycerol (Sigma), and 1,000 U of leukemic inhibitory factor (LIF; GIBCO) on gelatinized plates. CCE cells (10⁴) were transduced with the viral vectors in the gelatinized T25 bottles in the presence of protamine sulfate. The efficiency of viral transfer in the bulk population was estimated by determining the ratio of GFP-expressing cells with a FACSCalibur. As shown in Table 1, the efficiency of lentivirus transduction was 99.8% ± 0.1% at a multiplicity of infection (MOI) of 59, and the efficiency increased in proportion to the MOI. A similar correlation was seen between MOI and transduction efficiency for oncoretrovirus transduction, but the latter was lower than that of lentivirus vectors at the same MOI (5.9 to 59). The efficiency of oncoretrovirus transduction increased to 56.2% ± 2.8% at an MOI of 590 (Table 1). After transduction, part of the ES cell bulk population was washed and then replated in 96-well plates at 0.5 to 2 cells per well by the limiting dilution method to create clones from single, transduced ES cells. Undifferentiated ES colonies were formed after 7 days of culture in 96-well plates in the presence of LIF. GFP-expressing colonies were scored by fluorescence microscopy. Almost all of the ES colonies were transduced by the lentivirus vector at an MOI of 59 (94.8% ± 1.5% positive for GFP expression [Table 1]). This number was almost identical to the number of GFP-positive ES cells from the bulk population as estimated by FACS (99.8% ± 0.1%, MOI = 59). There was a good correlation between the proportion of GFP-positive ES cells from the bulk population and the number of positive ES cell colonies that scored visually positive in the microscope at all MOIs used (5.9 to 59 [Table 1]). These results suggest that the lentivirus and oncoretrovirus transgenes are integrated stably into the genomes of their ES target cells (see below). These results demonstrate that transduction of ES cells with VSV-G-pseudotyped lentivirus vectors is very efficient. The transduction efficiency was close to 100% at an MOI of 59. Ecotropic oncoretrovirus vectors transduce ES cells less efficiently, requiring very high MOIs to transduce approximately 50% of the cells.

View larger version (18K): [in this window] [in a new window]   FIG. 1.   Vectors. The lentivirus vector (HIV-PGK-GFP) uses the internal PGK promoter to drive expression of GFP gene. The Rev-responsive element (RRE) is indicated. The oncoretrovirus vectors (MGirL22Y [28]) and MSV-PGK-GFP) contain the same vector backbone using the MESV LTR, the same primer binding site, and the same length of the gag region (the figure is not drawn to scale) (14, 28). MGirL22Y contains the GFP gene followed by an internal ribosomal entry site (IRES) from the encephalomyocarditis virus linked to a mutant dihydrofolate reductase gene (L22Y). MSV-PGK-GFP contains the MESV LTR and the internal PGK promoter followed by the GFP gene.

ES cell transduction with VSV-G-pseudotyped oncoretrovirus vectors. To determine whether the difference in lentivirus and oncoretrovirus transduction efficiency of ES cells was due to the differences in the envelope protein used, we pseudotyped two oncoretrovirus vectors, MGirL22Y and MSV-PGK-GFP (Fig. 1), with VSV-G. MSV-PGK-GFP (provided by Keith Humphries, Vancouver, British Columbia, Canada) was derived from the MSCV vector that contains the internal PGK promoter driving the neomycin resistance (neo) gene (14) where the GFP gene replaces the neo gene. It has the same vector backbone as MGirL22Y and contains the MESV LTR. Amphotropic vector MSV-PGK-GFP was produced by transient transfection into Phoenix amphotropic cells (provided by G. Nolan, Stanford University, Palo Alto, Calif.). Ecotropic MSV-PGK-GFP vector producer cell lines were established by transduction of GP+E86 cells with the amphotropic MSV-PGK-GFP vector supernatant. VSV-G-pseudotyped MGirL22Y and MSV-PGK-GFP producer cell lines were established by transduction of 293GPG cells with each amphotropic vector. Table 1 shows that VSV-G-pseudotyped oncoretrovirus vectors transduce ES cells with higher efficiency than the same vectors packaged in ecotropic packaging cells. To transduce practically 100% of the cells, an MOI of 59 was needed for the HIV-1-based lentivirus vector. VSV-G pseudotyping of oncoretrovirus vectors improved their transduction efficiency of ES cells substantially. These results show that VSV-G pseudotyping of both lentivirus and oncoretrovirus vectors leads to very efficient transduction of ES cells, although a very high MOI (590) is needed to reach the oncoretrovirus maximum transduction level of approximately 100% of the cells.

Integration of viral vectors in ES cells. As shown above, the lentivirus-transduced clones seemed to be stably transduced, since they were expressing the transgene for more than 10 days after transduction. To determine whether the lentivirus genome had integrated into the chromosomal DNA of the target cell, genomic DNA from the transduced ES cell clones was subjected to Southern blot analysis. Genomic DNAs of the lentivirus- and oncoretrovirus-transduced clones were digested with BamHI and EcoRI, respectively, and then hybridized with a ³²P-labeled GFP probe (Fig. 2). Lentivirus-transduced clones that expressed the GFP transgene 11 days after transduction, as detected visually in a fluorescence microscope, proved to have one or more integrated copies of the HIV-PGK-GFP provirus when analyzed by Southern blot analysis. As shown in Fig. 2, the lentivirus-transduced clones contained approximately 10 integrated copies in the ES genome. The lentivirus-transduced clones expressing the transgene at day 11 following transduction were all found to harbor an integrated provirus. Since the ES cells were rapidly dividing during transduction and after plating into the microtiter plates, the data suggest that integration takes place relatively rapidly following transduction.

View larger version (43K): [in this window] [in a new window]   FIG. 2.   Expression level of GFP and number of integrated proviral copies in transduced ES clones. (A) Three lentivirus (HIV-PGK-GFP vector)- and three oncoretrovirus (MGirL22Y vector)-transduced ES clones (PG and MS clones, respectively) which showed high GFP expression levels. (B) Southern blot analysis of the clones. Genomic DNAs of the PG and MG clones were digested with BamHI and EcoRI, respectively, and hybridized with GFP cDNA. (C) Relationship between the integrated copy number and expression level of GFP.

Transgene expression in ES cells during development in vitro. The transduced ES cell clones were used to monitor vector expression during hematopoietic development in culture to examine whether gene expression persists as differentiation to hematopoietic lineages proceeds. Expression of the vector transgene was monitored in undifferentiated ES cells, in ES cell-derived day 6 EBs, and in hematopoietic colonies. Two days before the initiation of differentiation, ES cells, maintained in the complete medium mentioned above, were transferred to Iscove's modified Dulbecco's medium (IMDM) containing 15% FCS, 1 mM glutamine, 150 mM -monothioglycerol, and 1,000 U of LIF. After 2 days, ES cells were plated at 900 cells/ml into differentiation medium containing IMDM supplemented with 15% FCS, 2 mM glutamine, 0.5 mM ascorbic acid (Sigma), and 450 mM -monothioglycerol. The differentiation cultures were maintained in petri dishes for 6 days. Day 6 EBs were trypsinized into single-cell suspensions and plated at 5 × 10⁴ cells per 3.5-cm-diameter plate in IMDM containing 1.0% methylcellulose, 10% plasma-derived serum (Animal Technologies, Tyler, Tex.), 2 mM glutamine, transferrin (300 µg/ml; Boehringer Mannheim, Mannheim, Germany), interleukin-3 (20 ng/ml; Immunex, Seattle, Wash.), c-Kit ligand (50 ng/ml; Amgen, Thousand Oaks, Calif.), erythropoietin (2 U/ml; Janssen-Cilag, Sollentuna, Sweden), and 5% protein-free hybridoma medium (GIBCO). The cells were cultured for 10 days.

View larger version (90K): [in this window] [in a new window]   FIG. 3.   ES-derived differentiated cells. (A and B) Day 6 EBs derived from the lentivirus-transduced clone PG20 (magnification, ×100). Panel B is visualized by a fluorescence microscope. (C) Day 10 hematopoietic colony derived from ES cells (×150).

View larger version (44K): [in this window] [in a new window]   FIG. 4.   GFP expression in lentivirus- and oncoretrovirus-transduced clones during hematopoietic differentiation. FACS analysis shows the number of GFP-positive cells in undifferentiated ES cells, in day 6 EBs, and in hematopoietic colonies from methylcellulose cultures. Cells from day 6 EBs and hematopoietic colonies were analyzed by gating away dead cells stained with propidium iodide. PG, MG, and MPG indicate clones transduced with HIV-PGK-GFP, MGirL22Y, and MSV-PGK-GFP, respectively.

Transcriptional control of the GFP transgene in ES cells. To determine whether the lentivirus transgene in the ES clones is expressed from the internal PGK promoter, Northern blot analysis was performed. Total RNA was prepared from four ES clones, from 293T cells transiently transfected with the lentivirus vector construct, and from the oncoretrovirus (MGirL22Y) producing cells. As shown in Fig. 5A, the MGirL22Y-transduced clone contained messages of 2.5 and 3.2 kb, derived from the oncoretrovirus LTR. In contrast, a 1.4-kb mRNA was detected in all of the lentivirus-transduced clones, suggesting that the internal PGK promoter controls the transgene. No genomic lentivirus mRNA was detected in the ES clones, indicating that the lentivirus LTR is silent in ES cells. The lentivirus GFP is expressed during hematopoietic differentiation in PG clones 20 and 14 (Fig. 4), due to the activity of the lentivirus internal PGK promoter throughout differentiation as shown in Fig. 5A. There is no major reduction in RNA levels in PG clone 20 following differentiation, but the reduction is substantial in PG clone 14. Similarly, the internal PGK promoter is the main generator of GFP containing RNA in the MSV-PGK-GFP-transduced ES clones. Northern blot analysis shows that the main RNA species in the undifferentiated clones was generated by the internal promoter (Fig. 5B). Following differentiation to day 6 EBs, levels of the LTR-generated RNA species and the species derived from the internal promoter were both reduced. Therefore, the internal PGK promoter seems to function well within the context of an oncoretrovirus genome in undifferentiated ES cells, but the expression level is dramatically reduced in their differentiated progeny cells.

View larger version (55K): [in this window] [in a new window]   FIG. 5.   Northern blot analysis of GFP expression in ES clones transduced with lentivirus and oncoretrovirus vectors. Schematic drawings of the lentivirus vector, HIV-PGK-GFP (A), and the retrovirus vector, MSV-PGK-GFP (C), depict the species of RNA generated by the internal promoter (solid line, shorter transcript) and the viral LTR (broken line, longer transcript). The splice donor and acceptor sites (SD and SA) are indicated. (B and D) RNA was extracted from virus-transduced ES cell clones in undifferentiated (ES) and differentiated (EB Day 6) cells. Twenty micrograms of RNA was transferred onto a nylon membrane and hybridized with a radioactively labeled DNA fragment specific for GFP and glyceraldehyde 3-phosphate dehydrogenase (G3PDH). RNA from 293T cells producing HIV-PGK-GFP, the MGirL22Y producer cells, and ecotropic Phoenix cells transfected with the MSV-PGK-GFP vector are shown for comparison. The MG lane in panel B shows two bands of 3.2 kb (unspliced) and 2.5 kb (spliced). In the lentivirus vector clones, a 1.4-kb mRNA species is generated by the PGK promoter. PG, MG, and MPG indicate clones transduced with HIV-PGK-GFP, MGirL22Y, and MSV-PGK-GFP, respectively. CCE, untransduced ES cells.