Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Methotrexate selection of long-term culture initiating cells following transduction of CD34+ cells with a retrovirus containing a mutated human dihydrofolate reductase gene


A limitation of successful stem cell gene transfer to hematopoietic stem cells is low transduction efficiency. To overcome this hurdle and develop a gene transfer strategy that might be clinically feasible, retroviral vectors containing a drug resistance gene were utilized to transduce human CD34+-enriched cells and select gene-modified cells by drug administration. We constructed a high-titer retroviral vector containing a fusion gene (F/S-EGFP) consisting of a mutated dihydrofolate reductase (DHFR) (Leu22→Phe22, Phe31→Ser31; F/S) gene and enhanced green fluorescent protein (EGFP) cDNA. To test whether the fusion gene could function as a selectable marker, transduced CD34+ cells were assayed in long-term stromal co-cultures with and without addition of methotrexate (MTX). Without MTX exposure, the vector-transduced CD34+ cells generated 22–50% EGFP+ cobblestone area forming cells (CAFC) at week 5. By contrast, the vector-transduced cells cultured with MTX produced 96–100% EGFP+ CAFC in four separate experiments. These are the first investigations to demonstrate selection for transduced long-term culture initiating cells using MTX. The DHFR/MTX system holds promise for improving selection of gene-transduced hematopoietic progenitor cells in vivo.


Hematopoietic stem cell gene transfer as a method to treat genetic diseases and patients with cancer has been receiving increasing attention. However, long-term persistence of only a small number of transduced hematopoietic progenitor cells was demonstrated in gene marking studies using the neomycin phosphotransferase gene (Neo)1,2 and in'human clinical trials studies using the multidrug resistance gene (mdr1).3,4 To improve retroviral stem cell transduction, many different strategies aimed at increasing viral attachment and stem cell proliferation without differentiation have been explored.5 Improved transduction of CD34+ progenitors was recently demonstrated (50–80%) with the use of fibronectin fragments (retronectin) or stromal cell layer–coated plates, which enhance co-localization of virus.6 More recently, improved transduction of human CD34+ cells was reported in a clinical trial that employed retronectin to modify cells with the mdr1 gene.7 Application of the cytokines TPO, FLT3 and c-kit ligand improves transduction of primitive hematopoietic progenitor cells by promoting cell cycle and stem cell self-renewal.8

The use of drug resistance genes as selectable markers may provide a means for allowing persistence and even enrichment of transduced stem cells in vivo. Our group and others have developed drug resistance genes with improved ability to protect host cells against chemotherapeutic toxicity. Various genes that confer drug resistance to hematopoietic cells, e.g., mutant human dihydrofolate reductase (DHFR) genes,9,10,11,12 human DNA alkyltransferase genes,13,14 and the mdr1 gene (p-glycoprotein),3,4,7,15 have been generated. Our group and others have previously tested various models to show hematopoietic cell selection by methotrexate (MTX). Flasshove et al11 have demonstrated selection for a pre-CFC population of human CD34+-enriched cells transduced with mutant DHFR in liquid culture system. Allay et al12 used in vivo competitive repopulation assay in a mouse model to show that mouse bone marrow cells transduced with mutant DHFR were selectively enriched in vivo by simultaneous administration of trimetrexate (TMTX) treatment of thymidine transport inhibitor. However, to our knowledge, there have been no reports demonstrating MTX selection of primitive human hematopoietic cells.

In this study, we constructed a retrovirus containing a human mutated DHFR (Phe22/Ser31; F/S) cDNA16 fused to an enhanced green fluorescent protein (EGFP) cDNA. The “humanized” EGFP gene from the jellyfish Aequorea victoria has been widely used as a tool to trace retrovirally transduced hematopoietic progenitors. It allows for rapid identification of transduced cells by fluorescence microscopy and flow cytometry.17,18 Transduced human hematopoietic progenitors, including week 5 cobblestone area forming cell (CAFC) and LTC-IC, were directly visualized in long-term culture (LTC) by fluorescence microscopy. Week 5 CAFC and LTC-IC are in vitro surrogates for stem cells.19,20 The CAFC assay and the LTC-IC assay allow for frequency analysis of primitive cells capable of long-term repopulation in vitro.21 Based on previous studies, this assay may be used to assess the effect of manipulation of human hematopoietic progenitor cell populations by gene transfer methods.22 Recent studies showed that retroviral gene transduction can be assessed in both CAFC and in the NOD/SCID mice transplantation model using purified CD34+ and CD34+ CD38 umbilical cord blood (UCB) cells.18

The main goal of this study was to test the efficacy of MTX selection for DHFR-transduced human CAFC.23We used the mutated DHFR (F/S) and EGFP fusion (F/S-EGFP) retroviral constructs in a long term stromal co-culture system, thus ensuring that both genes were equally expressed. MTX exposure was found to select for EGFP+ CAFC and secondary CFU-GM (LTC-IC) measured at 5'weeks. These investigations have implications for the use of gene transfer of DHFR to limit MTX-induced myelotoxicity in cancer patients.

Materials and methods

Human peripheral blood progenitor cells (PBPCs) and CD34+ cell selection

All leukapheresis products used for PBPC isolation were specimens from deceased cancer patients. Based on the IRB-approved consent and the family member's agreement, we used the specimens that were designated to be discarded. Following PBPC mobilization with chemotherapies and G-CSF, leukapheresis specimens were harvested, cryopreserved, and stored in the Stem Cell Laboratory at Memorial Sloan-Kettering Cancer Center. Mononuclear cells were isolated from thawed specimens by Ficoll-Paque (Pharmacia, Piscataway, NJ) density gradient centrifugation. CD34+ cell selection was performed using the StemSep system per manufacturer's instructions (StemCell Technologies, Vancouver, Canada). CD34+-selected cells were incubated with R-phycoerythrin (PE)–conjugated antibodies against human CD34 (C34-PE; Becton Dickinson, San Jose, CA) for 30 minutes on ice in PBS supplemented with 2% fetal bovine serum (FBS). After incubation, the cells were washed twice, resuspended in PBS, and the percentage of CD34+ cells was determined using FACS (FACScalibar; Becton Dickinson).

Generation of a fusion protein F/S-EGFP

The coding sequence of human mutated DHFR (F/S), fused to a sequence encoding a consensus thrombin cleavage site, was generated in the bacterial expression vector pkT7 (Clontech Laboratories, Palo Alto, CA) to yield a plasmid pkT7 F/S. A 720-bp fragment of EGFP was generated by polymerase chain reaction (PCR) amplification of pEGFP-N1 (Clontech Laboratories) using the following primers: EGFPBHI: 5′-GTTCCGCGTGGATCCCCGGGAATTGTGAGCAAGGGCGAGGAGCTG; EGFPSTOPIII: 5′-ATCATCATCAAGCTTTTACTTGTACAGCTCGTCCATGCCGAG. The forward primer EGFPBHI contained BamHI and SmaI restriction sites and a reverse primer EGFPSTOPIII contained XhoI and HindIII restriction sites.

The PCR-amplified fragment and the pKT7-F/S vector were digested with BamHI/HindIII. The purified EGFP PCR product was ligated into the pKT7-F/S vector backbone to yield pKT7-F/S-EGFP. This sequence was confirmed by restriction digestion and DNA sequencing. Escherichia coli BL21 cells were transformed with pKT7-F/S-EGFP, grown to a density of OD 0.6–0.8, and induced with 1 mM isopropyl β-D-thiogalactopyranoside (IPTG) for 5 hours. The transformed cells were harvested and resuspended in buffer (10'mM Tris, 0.1 mM EDTA), sonicated (5 seconds every 15'seconds ×4 on ice), followed by incubation at 95°C for 5'minutes. The crude lysate was obtained after centrifugation at 15,000 rpm for 20 minutes. Next, the supernatant and the whole lysate were electrophoresed, side by side, on a sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) minigel (12%) to confirm the correct molecular weight of the fusion protein product. Noninduced lysates served as controls. The detailed enzyme kinetics of F/S-EGFP fusion protein and other F/S fusions constructed in our laboratory is the subject of another manuscript in preparation. Quantitation of GFP fluorescence of F/S-EGFP–transfected cells is discussed in the Generation of Amphotropic and Ecotropic Producer Cell Lines section.

Construction of retroviral vectors SFG-F/S-EGFP and SFG-EGFP-IRES-Neo

The MMLV-based SFG retroviral vector, SFG-F/S-EGFP, contained the fusion gene F/S-EGFP cDNA (vide supra) (Fig 1). The vector construct SFG-EGFP-IRES-Neo was used as a control. The Neo gene was included in the control vector to facilitate selection of producer cell lines. The SFG-F/S-EGFP (Fig 1A) was constructed from pkT7-F/S-EGFP and cloned into an SFG-based retroviral vector backbone.16 The SFG-EGFP-IRES-Neo (Fig 1B) was generated by subcloning the EGFP fragment obtained from the pEGFP-N1 vector (Clontech Laboratories) into the NcoI/NotI site of the SFG backbone. The resultant plasmid, SFG-EGFP, was then digested with NotI/XhoI restriction sites. An internal ribosomal entry site (IRES)-Neo fragment, excised from pBlue IRES-Neo plasmid (subcloned in our laboratory from SFG-F/S-IRES-Neo vector), was ligated with SFG-EGFP to yield the SFG-EGFP-IRES-Neo vector.

Figure 1

The structure of the F/S-EGFP and EGFP-IRES-Neo retroviral vectors. The chimeric 3′ LTR contains the myeloproliferative sarcoma virus (MPSV) promoter/enhancer. Splice donor (SD) and splice acceptor (SA) sites are located between the 5′LTR and the NcoI site. The unique restriction sites used for cloning are shown. A: Contains a fusion gene of DHFR (F/S) and EGFP with a linker that contains thrombin cleavage site, BamHI, and SmaI restriction sites. B: The SFG-EGFP-IRES-Neo construct.

Generation of amphotropic and ecotropic producer cell lines

GP-E86 SFG-F/S-EGFP, an ecotropic producer cell line, and GP-AM12 SFG-F/S-EGFP, an amphotropic packaging line,24,25 were generated by co-transfection of SFG-F/S-EGFP and pSV2Neo at a molar ratio of 10:126 using DOTAP reagent (Boehringer-Mannheim, Indianapolis, IN). GP-E86 SFG-EGFP-IRES-Neo and GP-AM12 SFG-EGFP-IRES-Neo were also generated by transfection using DOTAP reagent. Transfected cells were selected with G418 (750 μg/mL) for 14 days. The selection of the best producer cell line was carried out by two-step procedure. First, each producer cell clone was analyzed by flow cytometry according to the intensity of GFP. Both SFG-F/S-EGFP– and SFG-EGFP-IRES-Neo–transfected producer cell clones showed 103–104 fluorescent intensity compared to that of control cells, which were in the range of 100–102. The degree of GFP intensity in fusion construct was comparable to that of nonfusion construct. The percentage of GFP+ cells detected by FACS analysis and fluorescent microscopic analysis was identical. The second step was to submit preselected producer clones for the viral titer estimated by infecting NIH3T3 cells and then selecting with MTX (150 nM) or G418 (750 μg/mL) for 14 days as described previously.27 We chose the highest titer clone, GP-AM12 SFG-F/S-EGFP, for the following procedure to further increase the titer. We infected the amphotropic producer cell line (GP-AM12 SFG-F/S-EGFP) with the supernatant from the ecotropic producer cell line (GP-E86 SFG-F/S-EGFP) 10 times over 10 days in the presence of fresh 8 μg/mL polybrene. The final viral titer of the AM12 SFG-F/S-EGFP was 1–2×107 colony-forming units/mL (cfu/mL) showing approximately 1 log improvement of the titer. The titer of GP-AM12 SFG-EGFP-IRES-Neo was 5×105 cfu/mL. These producer cell lines were free of replication-competent retrovirus as demonstrated by failure of MTX or G418 resistance to be transferred from virally transduced NIH3T3 cells to nontransduced NIH3T3 cells as described previously.17 The selected producer cell lines were confirmed by FACS and fluorescent microscopy for GFP expression.

Retroviral transduction of NIH3T3 cells

NIH3T3 cells were exposed to virus-containing supernatant for 6 hours in the presence of 8 μg/mL polybrene at an MOI of 5. The following day, the medium was changed to either G418 (750 μg/mL) or TMTX (20 nM) for 10 days, for the vectors containing the Neo gene or F/S gene, respectively. This concentration of TMTX is 100% lethal to nontransduced NIH3T3 cells, but is not inhibitory to F/S-DHFR infected cells.28

PCR analysis of genomic DNA from NIH3T3 cells

Genomic DNA from NIH3T3 cells was isolated as described previously.27 Two hundred nanograms of genomic DNA was used for each PCR reaction. EGFP cDNA was amplified using the following primers: forward EGFP1 (5′-GCCACAAGTTCAGCGTGTCC-3′) and reverse EGFP2 (5′-AGCTCGATGCGGTTCACCAG-3′) primers.29 The resultant PCR product was approximately 300 bp.

Protection of NIH3T3 cells infected with retroviral vectors containing a fusion gene of human mutated DHFR (F/S) and EGFP from MTX toxicity

MTX was prepared from a stock solution (10−2 M) stored at −70°C. MTX cytotoxicity was measured by a colorimetric assay using XTT (sodium 3′-{1-[(phenylamino)-carbonyl]-3,4-tetrazoliu}-bis(4-methyoxy-6-nitro) benzene-sulfonic acid hydrate) as described previously.30 NIH3T3 cells that were infected with viral supernatant and selected as described (vide supra) were plated in 96-well plates (400 cells per well). The following day, infected cells were exposed to varying doses of MTX, using media containing dialyzed FBS. After 5 days of exposure to MTX, 50 μL of 1 mg/mL XTT and 0.025 mM phenazine methosulfate (PMS) were added. The absorbance of the supernatant was measured at 450 and 630 nm with wells without drug (with cells) as 100% and wells without cells as 0%. To assess the GFP sensitivity in transduced cells by both fusion vector and control vector, we analyzed an aliquot of the same cells used for the cytotoxicity assay by FACS analysis and fluorescence microscopy.

Western blotting

Protein lysates were prepared from cells transduced by SFG-F/S-EGFP, SFG-EGFP-IRES-Neo, SFG-F/S-IRES-Neo,16 mock-transduced cells, and the DHFR−/− DG44 CHO cell line (used as a negative control). Cells were lysed in a buffer containing proteinase inhibitors as previously described.31 After a 20-minute incubation on ice and centrifugation at 15,000×g for 20 minutes, the supernatant was recovered. Protein concentration was determined by the Bio-Rad (Hercules, CA) reagent per manufacturer's instructions using BSA as a standard. Lysate, containing 75 μg of protein, was used for electrophoresis on 12.5% SDS-PAGE. After transfer to a nitrocellulose membrane, the membrane was incubated for 2'hours with antihuman DHFR antibody, a rabbit polyclonal antiserum raised against recombinantly expressed human DHFR protein. The membrane was then incubated in a goat antirabbit secondary antibody (Santa Cruz Biotech, Santa Cruz, CA). The protein bands were visualized on X-ray film using the enhanced chemiluminescence reagent from Amersham (Arlington Heights, IL). After autoradiography, the membrane was stripped and incubated with an anti-GFP antibody (Clontech Laboratories), followed by antimouse secondary antibody (Santa Cruz Biotech). The bands were visualized as described above.

Retroviral transduction of PB CD34+ cells

Supernatants containing recombinant retrovirus were generated from the producer cells described above. The producer cell line was expanded and aliquots were frozen in multiple tubes until use. The titer of the producers was measured immediately after the first round of cell culture passage and then again after the cells were expanded in culture. Viral titer was again measured after expanding and splitting the cells three times. The titer was found to be consistent throughout this procedure. Because our transduction procedure involved multiple rounds of infection in fresh supernatant, we also titered the supernatant used in the first and last round of infections. Infections were performed on Retronectin® (Takara Shuzo, Otsu, Japan) coated dishes, as per manufacturer's instructions. Following overnight prestimulation in media containing human recombinant interleukin-6 (IL-6; 20 ng/mL; R&D Systems, Minneapolis, MN), thrombopoietin (TPO; 100 ng/mL; R&D Systems), stem cell factor (SCF; 20 ng/mL; Immunex, Seattle, WA), Flt-3 ligand (FLT3; 100 ng/mL; Imclone, New York, NY), and human granulocyte–macrophage colony-stimulating factor (GM-CSF; 100 ng/mL; Immunex), the CD34+-enriched population was transduced with either AM12 SFG-F/S-EGFP, AM12 SFG-EGFP-IRES-Neo, or mock AM12 viral supernatants. Retroviral infection was performed twice a day for 4 days at an MOI of 20 for SFG-F/S-EGFP and at an MOI of 1 for SFG-EGFP-IRES-Neo. Each infection was performed by resuspending the cells in fresh virus supplemented with cytokines. On day 5, the cells were harvested and FACS analysis, human granulocyte–macrophage CFU (CFU-GM), erythroid burst-forming units (BFU-E) assays, and LTC-IC assays performed.

Flow cytometry

Cell samples were analyzed using a FACScalibar (Becton Dickinson). Immunophenotyping of EGFP-transduced cells was performed by staining with PE-conjugated CD34 (vide supra).

In vitro colony assay

Transduced progenitors were assayed for CFU-GM and BFU-E colony formation in methylcellulose as previously described.11 The cells were plated with increasing doses of MTX mixed in methylcellulose culture medium with BSA (StemCell Technologies) and additional human recombinant erythropoietin (EPO; 3 U/mL) (Amgen, Thousand Oaks, CA). Cells were plated in triplicate. The number of colonies was determined after 14–16 days of culture in a humidified atmosphere of 5% CO2 at 37°C. EGFP containing colonies were scored using fluorescence microscopy.

LTC and CAFC assay

The MS-5 murine marrow stromal cell line (Kirin Pharmaceuticals, Tokyo, Japan) was used to support hematopoietic progenitors during the LTC.34 MS-5 cells were seeded in T25 flasks and grown to confluence prior to the initiation of LTC. An equal number of mock-transduced or vector-transduced CD34+ enriched cells were seeded onto the stroma (0.8–1.125×105 cells per flask for each experiment). Cells were maintained in modified LTC media containing thymidine phosphorylase–treated FBS.32 Thymidine phosphorylase–treated FBS was used in LTC media to decrease background colonies as the cytotoxicity of antifolates was attenuated by the thymidine salvage. FBS contains a high level of this metabolite.9 Mock-transduced, control vector–transduced, and fusion protein vector–transduced cells were assayed in LTCs with MTX (5×10−8 M) and without MTX. Each culture was run in triplicate. Each week, the LTC was demi-depopulated and fresh LTC medium with or without MTX was added (5×10−8 M). Demi-depopulation involves removal of half the total culture media and suspension cells and then replenishment with the same volume of fresh media, such that the total culture volume and concentration of MTX remain constant throughout the 5-week period. This MTX concentration was not cytotoxic to MS-5 stromal layers, but was 100% lethal to nontransduced peripheral blood CD34+-enriched cells by CFU-GM assay (unpublished data). Cell counts and FACS analysis were performed each week. All LTCs were carried out for 5 weeks. CFU-GM assays were performed at weeks 0 and 5. Phase dark hematopoietic clones of at least eight cells (i.e., CAFCs) beneath the stromal layer was determined at week 5 using an inverted microscope. Green fluorescent cobblestone areas were screened in the same way, but with an UV light excitation source (Eclipse TE200; Nikon, Melville, NY). LTC-IC values were calculated from the colony numbers detected in the secondary CFU-GM assay at week 5.

EGFP cDNA detection in CFU-GM colonies by PCR

After 14 days of culture in methylcellulose, CFU-GM colonies were analyzed by fluorescence microscopy as well as by PCR. Individual large colonies were harvested by a pipette and transferred into 0.5-mL Eppendorf tubes, suspended in PBS, and washed twice. The pellet was resuspended in Higuchi's buffer33 with proteinase K, incubated at 55°C for 3 hours, followed by inactivation of proteinase K (95°C for 10 minutes). For PCR, β-actin primers were used to test the quality of genomic DNA, and EGFP primers (forward: EGFP1 and reverse: EGFP2, sequence described as above) were used for transgene detection. The primer sequences for β-actin used are as follows: β-actin upper: 5′-GTGGGGCGCCCCAGGCACCA-3′; β-actin lower: 5′-CTCCTTAATGTCACGCACGATTTC-3′. PCR conditions were 95°C for 30 seconds, 60°C for 1 minute, and 72°C for 1 minute (EGFP) or 2 minutes (β-actin) for 40 cycles. Five microliters of genomic DNA was used in the total volume of 50 μl for the reaction.


Retroviral vector expression in NIH3T3 cells

Retroviral vectors shown in Figure 1 were first tested by transduction of NIH3T3 cells. NIH3T3 cells were used as target cells for testing vector expression because this cell line is very sensitive to MTX, which enabled accurate testing of MTX resistance. Also, the antibody to human DHFR preferentially recognizes the human and not the mouse enzyme. NIH3T3 cells were transduced with amphotropic retrovirus encoding the human F/S-EGFP fusion gene, or a retrovirus encoding EGFP and Neo as a control. After appropriate drug selection using G418 or MTX (see Materials and Methods), both control vector and SFG-F/S-EGFP vector–transduced cells were nearly 100% positive for EGFP as determined by both either fluorescence microscopy and FACS analysis. The intensity of EGFP fluorescence in SFG-F/S-EGFP–transduced cells was also comparable to fluorescence of SFG-EGFP-IRES-Neo–transduced cells.

Proviral DNA was detected in transduced NIH3T3 cells by PCR amplification of a 310-bp EGFP fragment (Fig 2). As expected, no EGFP DNA was detected in mock-transduced NIH3T3 cells. In order to demonstrate protein expression, cell lysates from vector-transduced and mock-transduced clones were analyzed by Western blot using antibodies to human DHFR and EGFP (Fig 3). Pure DHFR (50 ng) protein and lysate from NIH3T3 cells transduced with SFG-F/S-IRES-Neo were loaded as positive controls. No endogenous human DHFR or EGFP protein was detected in mock-transduced NIH3T3 cells. A 48-kDa band, corresponding to the size of the F/S-EGFP protein, was detected by anti-DHFR in the cell line transduced with SFG-F/S-EGFP, but not in the cell line transduced with SFG-EGFP-IRES-Neo. The 48-kDa band was also detected in SFG-F/S-GFP–transduced cells by the antibody to EGFP. A 27-kDa band, corresponding to the size of the EGFP protein, was detected by the anti-EGFP antibody in cells transduced with SFG-EGFP-IRES-Neo.

Figure 2

PCR analysis of NIH3T3 cells transduced with retroviral vectors. Lane 1: Control (mock-transduced) NIH3T3 cells. Lane 2: NIH3T3 cells transduced with the F/S-EGFP construct. Lane 3: NIH3T3 cells transduced with the EGFP-IRES-Neo construct. A φx molecular marker is shown next to lane 3. Primers designed to amplify the EGFP sequence were used to amplify the integrated retroviral construct from genomic DNA (see Materials and Methods). An EGFP band is detected in both vector-transduced cells at the 310-bp size φx marker.

Figure 3

Western blot analysis of cellular extracts from NIH3T3 mouse fibroblasts. A: Represents Western blots stained with human DHFR antibody. Lane 1 is pure DHFR protein (21 kDa); lane 2 is DG44 cell line, a negative control from a DHFR−/− containing cell line (DG44 CHO cells); lane 3 is mock-infected NIH3T3 cells; lane 4 is from control vector–infected cells (EGFP-IRES-Neo–transduced NIH3T3 cells); lane 5 is a positive control cell line (F/S-IRES-Neo–transduced NIH3T3 cells) (see Materials and Methods); and lane 6 is from cells infected with the fusion gene (F/S-EGFP–transduced NIH3T3 cells), 48 kDa. B: Shows the same Western blot with DHFR antibody staining, which was destained and restained with EGFP antibody. The molecular mass of EGFP protein is approximately 27 kDa. There are bands at the level of 27 kDa corresponding to the size of EGFP in lane 4 and the minor bands in lanes 5 and 6. These bands are likely as an overflow product of EGFP protein from band 4.

MTX cytotoxicity assays in NIH3T3 cells

To test whether increased expression of F/S-EGFP fusion protein led to MTX resistance, SFG-F/S-EGFP and mock-transduced NIH3T3 cells were exposed to increasing doses of MTX. SFG-F/S-EGFP–transduced cells were selected by treatment with 20 nM TMTX for 10 days. The IC50 for MTX cytotoxicity toward SFG-F/S-EGFP–transduced cells was determined by the XTT assay to be 5.2±0.6 μM (results from three separate experiments were 4.5, 6, and 5'μM), a 250-fold increase over mock-transduced control cells (Fig 4).

Figure 4

MTX cytotoxicity in NIH3T3 cells after MTX transduction. Cells were exposed to MTX for 5 days (see Materials and Methods). The SFG-F/S-EGFP–transduced cells were 250-fold more resistant to MTX compared to untransduced NIH3T3 cells. The graph was plotted using the mean and the standard deviation from three separate experiments. NIH3T3 cells (++), NIH3T3 cells transduced with SFG-F/E-EGFP (--).

Transduction efficiency in PBSC CD34+-enriched cells with the F/S-EGFP vector

CD34+-enriched cells were transduced with either mock, SFG-EGFP-IRES-Neo, or SFG-F/S-EGFP viral supernatant for 4 days in fibronectin-coated dishes with an MOI of 1 for SFG-EGFP-IRES-Neo and an MOI of 20 for SFG-F/S-EGFP construct. Fresh media and cytokines were added twice a day during the period of transduction (see Materials and Methods). The experiment was repeated four times using CD34+-enriched progenitor cells isolated from mobilized peripheral blood (mPB). The mPB samples were from different patients in each experiment. The percentages of CD34+ cells in the progenitor-enriched samples, before and after retroviral infection, are summarized in Table 1. Transduction with SFG-F/S-EGFP, as determined by FACS analysis for GFP expression, is also shown in Table 1. The dual positivity (CD34+ and EGFP+) was decreased compared to the EGFP+ percentage in experiments 1, 3, and 4. This reflected the decrease in the percentage of CD34+, which occurred during the transduction procedure and ex vivo culture for 5 days. The lower percentage of cells transduced with SFG-EGFP-IRES-Neo compared with SFG-F/S/EGFP was likely due to the difference in the MOI (20 vs 1). The mock-transduced group was analyzed in the same manner as the vector-transduced group for expression of CD34 and EGFP.

Table 1 Percentage of CD34+ cells and GFP+ cells pre- and posttransduction procedure

Progenitor cell proliferation during LTC

Following retroviral transduction or mock infection, CD34+-enriched cells were plated in LTC with MS-5 stromal cells. LTCs were propagated for 5 weeks with or without MTX. Each week, suspension cells were demi-depopulated and total cells per flask were enumerated. Figure 5 shows the total number of cells at weekly time points in all four experiments. In experiments 1, 3, and 4, proliferation of SFG-F/S-EFGP cells was similar to control cultures. In the absence of MTX, proliferation was identical in all experiments. Control vector–transduced cells did not survive more than 2 weeks in the presence of MTX in each experiment. In contrast, cells transduced with SFG-F/S-EFGP, which were cultured with MTX, survived throughout the entire period of culture in all experiments. The number of cells surviving MTX treatment in the SFG-F/S-EFGP culture was less in experiment 2 than in other experiments. This was most likely due to the lower efficiency of transduction with the fusion vector in experiment 2 [%GFP+input was 16% (experiment 2) vs 40% (experiment 4); week 5 GFP+CAFC was 28% (experiment 2) vs 81% (experiment 4)].

Figure 5

Total cell counts per flask during the LTC of transduced CD34+ cells on MS-5 stroma ±MTX (5×10−8 M weekly) (suspension cell fraction). Cells in suspension were counted weekly (Materials and Methods). (A), (B), (C), and (D) represent experiments 1, 2, 3, and 4, respectively. B: Showed the lowest GFP+ percent input (16%) and (D) showed the highest input (40%) at the time of starting the LTC. A, C: Contain the middle range input. The graph showed the mean volume and standard deviation from three separate flasks per each transduced cell group. A: Includes mock without drug (--), mock with MTX (-•-), fusion vector without drug (--), and fusion vector with MTX (--). (B), (C), and (D) include control vector without drug (--), control vector with MTX (--), fusion vector without drug (--), and fusion vector with MTX (--).

FACS analysis of suspension cell population

Suspension cells from SFG-F/S-EGFP LTCs were analyzed for EGFP expression by FACS analysis (Fig 6). Cells analyzed for GFP expression were harvested at the time of weekly demi-depopulation. In each experiment, the percentage of GFP+ cells in the MTX-treated group increased to 96–98% by week 5. In contrast, in LTCs without MTX, GFP+ cells were either maintained or decreased with time. These results demonstrate selection for SFG-F/S-EGFP+ cells in the presence of MTX. In experiment 2, 3 weeks of MTX were required for enrichment of the suspension cells to >90% GFP positivity. This time lag probably reflects the lower percentage of GFP+ cells at the initiation of the LTC compared with the other groups (experiment 2 had only 16% GFP+ cells versus experiments 1,3, and 4 with 80%, 56%, and 51% GFP+ cells, respectively). The decline in GFP+ cells in cultures without MTX in experiments 1 and 3 was most likely due to overgrowth of nontransduced cells. Retroviral vectors tend to target more mature subsets of progenitor cells. The percentage of vector-transduced cells therefore declines at later time points as more primitive (nontransduced) cells take over the culture.

Figure 6

GFP+ suspension cells in co-culture of F/S-EGFP–transduced CD34+ cells on MS-5 stroma cells with () or minus (×) 5×10−8 M MTX. GFP+ percentage in the total suspension cell fraction was analyzed by FACS analysis at weeks 0, 1, 3, and 5 during the LTC. The suspension cell fractions from triplicate culture flasks were pooled and the percentage of GFP+ cells was demonstrated in the table. Each experiment contains the results from both MTX with () and without (×). The Y axis/week 0 shows the %GFP+input and Y axis/week 5 shows the final %GFP with and without MTX selection. All experiments showed high %GFP (98%, 97%, 96%, and 97%) as a consequence of MTX selection, but the nonselection group did not show a significant increment of %GFP (61%, 20%, 17%, and 53%).

Transduction efficiency and MTX selection of CAFC subsets

The ability of transduced cells to form cobblestone areas was evaluated at week 5 of LTC (Fig 7). The absolute numbers of CAFC derived from mock, control, and fusion protein vector–transduced cells without MTX treatment were comparable within each experiment. The percentage of GFP+ CAFC in the SFG-EGFP-IRES-Neo cultures was lower than that of the fusion vector–transduced cultures. This finding reflected the lower initial transduction efficiency of the control vector (5–12%) compared with the fusion vector (16–78%). The absolute number of CAFC was variable between experiments, ranging from 40 to 400 per flask. This may be partially due to the differences of CD34+ input at the initiation of LTC (0.8–1.1×105 CD34+ cells). The most striking and consistent observation was that regardless of differences in transduction efficiency at week 0, CAFCs in the fusion vector–transduced LTCs exposed to MTX were approximately 100% GFP+ in all experiments (Table 2). The percentage of GFP+ CAFC at week 5 without MTX in experiments 1, 2, 3, and 4 was 53%, 28%, 24%, and 81%, whereas in the presence of MTX, the percentage was 100%, 96%, 100%, and 100%, respectively.

Figure 7

Week 5 CAFC in MS-5 stromal co-culture ±5×10−8 M MTX. (A), (B), (C), and (D) represent the data of total CAFC and GFP+ CAFC counts from experiments 1, 2, 3, and 4 containing the mock-transduced, the control (SFG-EGFP-IRES-Neo)–transduced, and the F/SEGFP (SFG-F/S-EGFP)–transduced groups. The each group was tested with and without MTX. CAFC did not survive at all in the mock and the control groups. Therefore, MTX-treated data for these two groups were omitted. Only SFG-F/S-EGFP–transduced group showed CAFC with MTX. MTX treated or not treated is shown as (+) or (−) at the bottom of the each group. CAFC count performed without fluorescence is shown as GFP(−) and with fluorescence is shown as GFP(+). The control vector was not used in experiment 1. The values are shown as mean±SD from three different plates.

Table 2 GFP+ CAFC (%) in week 5 LTCs

MTX resistance conferred by the SFG-F/S-EGFP vector measured by CFU-GM and BFU-E assays

Aliquots of cells were assayed for CFU-GM and BFU-E at weeks 0 and 5 of LTC. Clonogenic assays were performed with and without MTX. The mock or SFG-EGFP-IRES-Neo vector–transduced groups did not show any surviving colonies at the concentration of 5×10−8 M MTX at week 0 or 5 in all experiments (Table 3). In contrast, week 0 CFU-GM and BFU-E that were transduced with SFG-F/S-EGFP demonstrated resistance to MTX, with 6–80% survival at 5×10−8 M MTX and 4–35% survival at 10−6 M MTX. CFU-GM colonies from week 5 of LTCs, either from adherent stromal layers or suspension, showed resistance to 5×10−8 and 10−6 M MTX. The percentage of the GFP+ colonies generated by SFG-F/S-EGFP cells in the absence of MTX at week 0 ranged from 335 to 82%. The percentage of week 0 GFP+ SFG-F/S-EGFP colonies increased to 88–100% in cultures with 5×10−8 M MTX, and 100% at 10−6 M MTX. Similarly, CFU-GM colonies generated by week 6 suspension cells treated with 10−6 M MTX were 100% GFP+.

Table 3 MTX resistance in virally transduced CFU-GM*

EGFP detection in CFU-GM colonies by PCR

PCR analysis was performed to detect proviral sequences in weeks 0 and 5 CFU-GM (Fig 8). DNA from seven colonies at week 0, generated by SFG-F/S-EGFP and mock-transduced cells and grown in the absence of MTX, was analyzed for EGFP cDNA. While all mock-transduced colonies were negative (data lanes 8 and 18 and additional data not shown), six of seven SFG-F/S-EGFP colonies were positive (lane 5) for EGFP DNA (86% transduction efficiency). Nine colonies grown in the presence of 5×10−8 M MTX (at this concentration, none of the mock-transduced cells survived) were positive for EGFP (lanes 9–17). This result was consistent with the fluorescence microscopy results from experiment 4 (100% of colonies were green fluorescent). Eight colonies from CFU-GM assay without MTX, generated by week 5 LTC cells without MTX, were also analyzed for vector DNA (Fig 9). Four of eight colonies were positive (lanes 3, 5, 6, 7). This result was consistent with the relatively lower transduction efficiency in this experiment (56%) as described above.

Figure 8

Detection of EGFP sequences from CFU-GM assays at week 0 of LTC. PCR amplification was performed on individual colonies after extraction of genomic DNA using primers coding for the EGFP sequence and β-actin coding sequence (see Materials and Methods). The top lane shows EGFP amplification with detection of the expected size product from CFU-GM colonies at week 0 (experiment 4). Lanes 1–7: Fusion protein vector–transduced group with no MTX treatment shows 5/7 (70%) GFP+ colonies. Lane 8: Mock-transduced group without drug treatment. Lanes 9–17: Fusion protein vector–transduced group with MTX treatment at 5×10−8 M shows 9/9 (100%) GFP+ colonies. Lane 18: Mock-transduced group without drug treatment. The bottom lane represents β-actin amplification to control for loading.

Figure 9

PCR detection of EGFP from CFU-GM colonies at week 5 of LTC. CFU-GM colonies from week 5 LTC without MTX was examined for EGFP detection. PCR amplification using EGFP primers was performed as described in Figure 8). Lanes 3, 5, 6, and 7 show the expected EGFP-amplified products. The bottom lane shows the expected β-actin–amplified products.


EGFP is a useful marker that can be used to optimize transduction conditions and evaluate the effect of hematopoietic growth factors on stem cell activation and expansion.18,29 We have generated a retroviral vector containing a fusion cDNA of EGFP and mutated DHFR. This vector was employed to confer drug resistance and investigate the selection process at the stem cell level in LTC using CAFC analysis. The expression of EGFP from both our control vector and fusion vector was equivalent by FACS analysis and fluorescence microscopy. Fusion vector–transduced NIH3T3 cells conferred MTX resistance comparable to that of our previously published nonfusion vector, SFG-F/S-IRES-Neo, transduced cells.16 As we hypothesized earlier,16 we now demonstrate that human primitive hematopoietic cells expressing a mutated DHFR (F/S) gene can be selected at the level of CAFC in vitro LTC. Moreover, there was also expression of the second gene (EGFP) with nearly 100% concordance.

The CD34+-enriched cells were obtained from cancer patients who had been exposed to variable degrees of chemotherapy prior to stem cell mobilization. The impact of this on the number and quality of stem cells recovered and subsequently transduced is an important variable. The variations in transduction and proliferative capacity illustrated in these investigations provide us with a realistic model of the clinical situation where this form of gene therapy may be useful.

In these investigations, CD34+-enriched cells from four different patient samples were transduced. Each patient had been subjected to heavy, albeit diverse, chemotherapy courses and different mobilization regimens. Despite our application of the same transduction method, we observed a wide range of transduction efficiencies in the four sets of experiments. The lowest transduction efficiency was observed in experiment 2. This low transduction efficiency may have been due to resistance of the stem cells to entry into the cell cycle because the FACS analysis showed that 94% of cells was still CD34+ after the 5-day transduction procedure. We cannot rule out possible toxic effects of the virus preparation during the transduction period, but variability within patient samples is a more likely explanation for the different transduction efficiencies. MTX treatment resulted in nearly 100% selection of mature progenitors (CFU-GM, BFU-E) in the suspension cell population throughout the LTC. This finding was supported by the increased incidence of EGFP+ cells by the second or third week. The reason for the delay in reaching the maximum of GFP positivity in experiment 2 is probably due to lower number of transduced cells (only 16% GFP+ cells at the initiation of the LTC).

CAFC assay for these four separate experiments also consistently demonstrated a result similar to the suspension cell data. The absolute number of CAFC was variable in each experiment, which could be due to experimental variation but most likely patient variability. Experiments 1 and 4 with fusion vector–transduced groups showed the maximum relative number of CAFC with MTX exposure compared to the mock- or control vector–transduced group without MTX. This finding is probably related to the higher number of total progenitor cells seeded in the flask for the fusion vector group as well as high transduction efficiency seen with this vector, which resulted in a high number of surviving CAFC following MTX treatment. An interesting observation through all sets of experiments was that some cobblestone areas derived from the fusion protein–transduced cells with no MTX treatment frequently appeared chimeric with both GFP+ and GFP phase dark cells in a cobblestone area instead of totally GFP+ cobblestone areas observed in the MTX treatment group (data not shown). The cobblestone areas are considered to be clonal so that the GFP negativity may be caused by a vector-silencing phenomenon.

All experimental sets demonstrated a decreased total cell count after 3 or 4 weeks of culture except for set 1, which showed a continuous increase in the mock-transduced total cell count after week 5 of LTC. This difference could be due to patient variability and the fact that mock cells were not exposed to the retronectin-coated plates during the transduction period. Possibly, multiple retronectin adhesion steps deplete cells needed to expand cell production long term. This is an observation from a small number of experiments, but it may need to be explored in the future in order to optimize transduction and maintenance of long-term repopulating cells.

Although GFP positivity by fluorescence microscopy was sufficiently sensitive to allow detection of transduced CFU-GM, fluorescence microscopy may underestimate the number of transduced colonies due to silencing or low levels of expression. The correlation between MTX resistance and GFP positivity was demonstrated by the percentage of GFP+ cells in CFU-GM assay (nearly 100% GFP positivity, except for experiment set 1 which showed only 62% positivity) at MTX concentrations of 5×10−8 M where only transduced colonies survived. We noted that the brightness of green fluorescence was enhanced by MTX treatment, especially in the methylcellulose colony assays, which might have increased the efficiency of detection of transduced colonies by fluorescence microscopy.

In summary, we have shown that genetically modified human primitive hematopoietic progenitor cells, demonstrated as week 5 CAFC and week 5 CFU-C, may be selected in LTC by drug treatment. The mutated DHFR and the MTX selection system may overcome low transduction efficiencies of human long-term repopulating stem cells, which has been one of the major hurdles in gene therapy for genetic disorders of the lymphohematopoietic system. In addition, such drug-resistant stem cells may afford significant myeloprotection in cancer patients treated with high-dose MTX treatment.


  1. 1

    Brenner MK, Rill DR, Holladay MS et al. Gene marking to determine whether autologous marrow infusion restores long-term haemopoiesis in cancer patients Lancet 1993 342: 1134–1137

    CAS  Article  Google Scholar 

  2. 2

    Dunbar CE, Cottler-Fox M, O'Shaughnessy JA et al. Retrovirally marked CD34-enriched peripheral blood and bone marrow cells contribute to long-term engraftment after autologous transplantation Blood 1995 85: 3048–3057

    CAS  PubMed  PubMed Central  Google Scholar 

  3. 3

    Hesdorffer C, Ayello J, Ward M et al. Phase I trial of retroviral-mediated transfer of the human MDR1 gene as marrow chemoprotection in patients undergoing high-dose chemotherapy and autologous stem-cell transplantation J Clin Oncol 1998 16: 165–172

    CAS  Article  Google Scholar 

  4. 4

    Rahman Z, Kavanagh J, Champlin R et al. Chemotherapy immediately following autologous stem-cell transplantation in patients with advanced breast cancer Clin Cancer Res 1998 4: 2717–2721

    CAS  PubMed  Google Scholar 

  5. 5

    Moore MA, MacKenzie KL . Optimizing conditions for gene transfer into human hematopoietic cells Prog Exp Tumor Res 1999 36: 20–49

    CAS  Article  Google Scholar 

  6. 6

    Hanenberg H, Xiao XL, Dilloo D et al. Colocalization of retrovirus and target cells on specific fibronectin fragments increases genetic transduction of mammalian cells Nat Med 1996 2: 876–882

    CAS  Article  Google Scholar 

  7. 7

    Abonour R, Williams DA, Einhorn L et al. Efficient retrovirus-mediated transfer of the multidrug resistance 1 gene into autologous human long-term repopulating hematopoietic stem cells Nat Med 2000 6: 652–658

    CAS  Article  Google Scholar 

  8. 8

    Dao MA, Hannum CH, Kohn DB et al. FLT3 ligand preserves the ability of human CD34+ progenitors to sustain long-term hematopoiesis in immune-deficient mice after ex vivo retroviral-mediated transduction Blood 1997 89: 446–456

    CAS  Google Scholar 

  9. 9

    Zhao SC, Li MX, Banerjee D et al. Long-term protection of recipient mice from lethal doses of methotrexate by marrow infected with a double-copy vector retrovirus containing a'mutant dihydrofolate reductase Cancer Gene Ther 1994 1: 27–33

    CAS  PubMed  Google Scholar 

  10. 10

    May C, Gunther R, McIvor RS . Protection of mice from lethal doses of methotrexate by transplantation with transgenic marrow expressing drug-resistant dihydrofolate reductase activity Blood 1995 86: 2439–2448

    CAS  PubMed  Google Scholar 

  11. 11

    Flasshove M, Banerjee D, Mineishi S et al. Ex vivo expansion and selection of human CD34+ peripheral blood progenitor cells after introduction of a mutated dihydrofolate reductase cDNA via retroviral gene transfer Blood 1995 85: 566–574

    CAS  PubMed  Google Scholar 

  12. 12

    Allay JA, Persons DA, Galipeau J et al. In vivo selection of retrovirally transduced hematopoietic stem cells Nat Med 1998 4: 1136–1143

    CAS  Article  Google Scholar 

  13. 13

    Allay JA, Davis BM, Gerson SL . Human alkyltransferase–transduced murine myeloid progenitors are enriched in vivo by BCNU treatment of transplanted mice Exp Hematol 1997 25: 1069–1076

    CAS  PubMed  Google Scholar 

  14. 14

    Davis BM, Reese JS, Koc ON et al. Selection for G156A O6-methylguanine DNA methyltransferase gene-transduced hematopoietic progenitors and protection from lethality in mice treated with O6-benzylguanine and 1,3-bis(2-chloroethyl)-1-nitrosourea Cancer Res 1997 57: 5093–5099

    CAS  Google Scholar 

  15. 15

    Podda S, Ward M, Himelstein A et al. Transfer and expression of the human multiple drug resistance gene into live mice Proc Natl Acad Sci USA 1992 89: 9676–9680

    CAS  Article  Google Scholar 

  16. 16

    Ercikan-Abali EA, Mineishi S, Tong Y et al. Active site-directed double mutants of dihydrofolate reductase Cancer Res 1996 56: 4142–4145

    CAS  PubMed  Google Scholar 

  17. 17

    Persons DA, Allay JA, Allay ER et al. Retroviral-mediated transfer of the green fluorescent protein gene into murine hematopoietic cells facilitates scoring and selection of transduced progenitors in vitro and identification of genetically modified cells in vivo Blood 1997 90: 1777–1786

    CAS  PubMed  Google Scholar 

  18. 18

    van Hennik PB, Verstegen MM, Bierhuizen MF et al. Highly efficient transduction of the green fluorescent protein gene in human umbilical cord blood stem cells capable of cobblestone formation in long-term cultures and multilineage engraftment of immunodeficient mice Blood 1998 92: 4013–4022

    Google Scholar 

  19. 19

    Ploemacher RE, van der Sluijs JP, Voerman JS et al. An in vitro limiting-dilution assay of long-term repopulating hematopoietic stem cells in the mouse Blood 1989 74: 2755–2763

    CAS  PubMed  PubMed Central  Google Scholar 

  20. 20

    Sutherland HJ, Lansdorp PM, Henkelman DH et al. Functional characterization of individual human hematopoietic stem cells cultured at limiting dilution on supportive marrow stromal layers Proc Natl Acad Sci USA 1990 87: 3584–3588

    CAS  Article  Google Scholar 

  21. 21

    Breems DA, Blokland EA, Neben S et al. Frequency analysis of human primitive haematopoietic stem cell subsets using a cobblestone area forming cell assay Leukemia 1994 8: 1095–1104

    CAS  PubMed  PubMed Central  Google Scholar 

  22. 22

    Breems DA, Van Driel EM, Hawley RG et al. Stroma-conditioned medium and sufficient prestimulation improve fibronectin fragment-mediated retroviral gene transfer into human primitive mobilized peripheral blood stem cells through effects on their recovery and transduction efficiency Leukemia 1998 12: 951–959

    CAS  Article  Google Scholar 

  23. 23

    Banerjee D, Zhao SC, Li MX et al. Gene therapy utilizing drug resistance genes: a review Stem Cells 1994 12: 378–385

    CAS  Article  Google Scholar 

  24. 24

    Markowitz D, Goff S, Bank A . A safe packaging line for gene transfer: separating viral genes on two different plasmids J Virol 1988 62: 1120–1124

    CAS  PubMed  PubMed Central  Google Scholar 

  25. 25

    Markowitz D, Goff S, Bank A . Construction and use of a safe and efficient amphotropic packaging cell line Virology 1988 167: 400–406

    CAS  Article  Google Scholar 

  26. 26

    Riviere I, Sadelain M . Methods for the construction of retroviral vectors and the generation of high titer producers In: Robbins PD, Ed Gene Therapy Protocols 1st ed Totowa, NJ: Human Press 1997 59

    Google Scholar 

  27. 27

    Takebe N, Nakahara S, Zhao SC et al. Comparison of methotrexate resistance conferred by a mutated dihydrofolate reductase (DHFR) cDNA in two different retroviral vectors Cancer Gene Ther 2000 7: 910–919

    CAS  Article  Google Scholar 

  28. 28

    Mineishi S, Nakahara S, Takebe N et al. Co-expression of the herpes simplex virus thymidine kinase gene potentiates methotrexate resistance conferred by transfer of a mutated dihydrofolate reductase gene Gene Ther 1997 4: 570–576

    CAS  Article  Google Scholar 

  29. 29

    Limon A, Briones J, Puig T et al. High-titer retroviral vectors containing the enhanced green fluorescent protein gene for efficient expression in hematopoietic cells Blood 1997 90: 3316–3321

    CAS  PubMed  Google Scholar 

  30. 30

    Roehm NW, Rodgers GH, Hatfield SM et al. An improved colorimetric assay for cell proliferation and viability utilizing the tetrazolium salt XTT J Immunol Methods 1991 142: 257–265

    CAS  Article  Google Scholar 

  31. 31

    Banerjee D, Schnieders B, Fu JZ et al. Role of E2F-1 in chemosensitivity Cancer Res 1998 58: 4292–4296

    CAS  Google Scholar 

  32. 32

    Gartner S, Kaplan HS . Long-term culture of human bone marrow cells Proc Natl Acad Sci USA 1980 77: 4756–4759

    CAS  Article  Google Scholar 

  33. 33

    Xu LC, Kluerfel-Stahl S, Blanco M, Schiffmann R, Dunbar C, Karlsson S . Growth factors and stromal support generate very efficient retroviral transduction of peripheral blood CD34+ cells from Gaucher patients Blood 1995 86: 141–146

    CAS  PubMed  Google Scholar 

  34. 34

    Itoh K, Tezuka H, Sakoda H et al. Reproducible establishment of hematopoietic supportive stromal cell lines from murine bone marrow Exp Hematol 1989 17: 145–153

    CAS  PubMed  PubMed Central  Google Scholar 

Download references


We gratefully acknowledge the assistance of Diane Domingo for FACS analysis, Jessica Boklan for providing optimal conditions for colony PCR, and Emine Ercikan-Abali for providing the Pkt7 plasmid. Supported by CA59350, CA08748, and CA09512 (NT).

Author information



Corresponding author

Correspondence to Malcolm AS Moore.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Takebe, N., Xu, LC., MacKenzie, K. et al. Methotrexate selection of long-term culture initiating cells following transduction of CD34+ cells with a retrovirus containing a mutated human dihydrofolate reductase gene. Cancer Gene Ther 9, 308–320 (2002).

Download citation


  • stem cell selection
  • methotrexate
  • double mutant dihydrofolate reductase
  • enhanced green fluorescent protein
  • fusion protein
  • gene therapy

Further reading


Quick links