Introduction

Mouse models and clinical observations have revealed that insertional mutagenicity represents a dose-limiting long-term toxicity of integrating transgene vectors^1,2,3,4. It is likely that the architecture of conventional gammaretroviral vectors with potent enhancer–promoter elements in the long terminal repeats (LTRs) increases the risk of insertional proto-oncogene activation⁵. Self-inactivating (SIN) vectors do not contain the U3 enhancer–promoter in the LTRs⁶ and are therefore likely to be less potent in activating cellular genes at the integration site.

When using improved transduction conditions, hematopoietic clones may accumulate several independent transgene insertions, which elevates the risk of insertional transformation^3,7,8. Synergistic mutations might also occur if clinical conditions require the application of genotoxic drugs. This concern is of special relevance for the proposed use of cancer drug resistance genes to provide a selective advantage to gene-modified cells in vivo, as exemplified with O⁶-methylguanine–DNA methyltransferase (MGMT)^9,10,11.

MGMT is a nuclear protein that reverses toxic and mutagenic lesions produced at the O⁶-position of guanines by DNA-alkylating agents. Among these agents are potent anti-cancer drugs such as the chloroethylating agent 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU). To overcome the dose-limiting bone marrow toxicity of these drugs, gene transfer of MGMT into healthy bone marrow cells has been proposed and recently implemented in a clinical trial^9,10,11. Evaluation in preclinical animal models, including large animals, has shown that MGMT gene transfer and subsequent chemoselection represents the most powerful tool available to date to increase the contribution of gene-modified cells to hematopoiesis, eventually resulting in a complete chimerism^12,13,14,15. MGMT gene transfer into bone marrow cells is thus also of interest for the treatment of inherited disorders of hematopoiesis^16,17. Because of the unique efficiency of this approach, preparative conditioning of recipients prior to infusion of gene-modified cells may even be unnecessary^13,18.

To overcome resistance of tumor cells, which can express higher levels of endogenous MGMT, mutant forms of MGMT have been introduced that are resistant to pharmacologic inhibitors of wild-type MGMT, such as O⁶-benzylguanine (O⁶BG). Currently the most potent of these mutants is MGMT-P140K^12,19,20. However, relatively high levels of MGMT-P140K expression might be required for cellular detoxification, because MGMT becomes inactivated and degraded after removing alkyl residues from DNA in a stoichiometric fashion. Moreover, the repair rate of the P140K mutant is less than that of the wild-type protein²⁰. Encouragingly, not only gammaretroviral LTR-driven vectors²¹ but also lentiviral SIN vectors expressing MGMT-P140K have been shown to protect hematopoietic cells against cytotoxic effects of alkylating agents^22,23. However, systematic comparisons of expression cassettes or side-by-side comparisons of gammaretroviral and lentiviral vectors have not yet been carried out.

Toward defining a suitable construct for future clinical studies, we developed and compared 11 monocistronic retroviral vectors driving identical MGMT-P140K transgenes. Five vectors were based on murine leukemia virus (MLV) gammaretroviral technology, the remaining 6 on HIV-1 lentiviral constructs²⁴. To allow for an unbiased comparison, we generated a similar design of expression cassettes in some MLV and HIV vectors, using the strong enhancer promoter of spleen focus-forming virus (SF)²⁵ positioned either in the LTRs or in an internal position of the SIN vectors. In the SIN context, we also tested the impact of an additional intron and two cellular promoters (elongation factor 1 (EF1) and phosphoglycerate kinase (PGK)) that have been shown to direct rather potent transgene expression in hematopoietic cells^26,27,28,29. We obtained several significant findings: Gammaretroviral vectors with intact LTRs containing enhancer–promoter sequences showed both higher titers and higher expression levels than the lentiviral counterpart, likely a result of suboptimal RNA processing of the lentiviral leader region. In the SIN context, gammaretroviral and lentiviral vectors with comparable internal cassettes had similar expression properties. Gammaretroviral SIN vectors pseudotyped with RD114/TR³⁰ had a higher gene transfer efficiency on proliferating human CD34⁺ cells than lentiviral counterparts. Comparative evaluation of different retroviral vector families will thus pave the way for the design of novel hybrid vectors that fulfill the desired features of both efficiency and increased biosafety.

Results and discussion

Vector design

To compare gammaretroviral (MLV-based) and lentiviral (HIV-1-based) vector technology, we cloned vectors with largely identical internal transgene cassettes consisting of MGMT-P140K followed by the posttranscriptional regulatory element (PRE) of woodchuck hepatitis virus³¹. The lentiviral vectors contained the original truncated PRE, as present in most lentiviral vectors³¹. In the gammaretroviral SIN vectors, we introduced a safety-modified PRE that maintains the posttranscriptional improvement of RNA processing³².

Both in lentiviral and in gammaretroviral vectors, we placed the SF enhancer–promoter into the U3 region of the LTR. In the case of the gammaretroviral vector SF91 the 5'UTR was devoid of aberrant AUG translational initiation codons and the packaging signal was flanked by splice sites³³. MGMT was cloned 3' of the splice acceptor, and the PRE was placed upstream of the 3'LTR. The construct MP71 had a similar design but contained the myeloproliferative sarcoma virus U3 region instead of SF. The LTR-driven lentiviral vector RRL.SFLTR contained SF in the LTR, and the internal promoter of the lentiviral backbone was deleted. The cDNA was inserted behind the splice acceptor of the env-derived fragment cotaining the Rev-responsive element (RRE). This results in an intron formed by the major HIV splice donor and the env-derived splice acceptor located upstream of the cDNA. Thus, both LTR vectors were splice competent (Fig. 1A), although using different splice signals.

Figure 1.

Schematic design of the five gammaretroviral and six lentiviral vectors used in this study. The modular composition of the vectors is shown with the respective LTRs (U3, R, U5), splice sites in the 5'UTR, internal promoter (if applicable), MGMT-P140K transgene, and posttranscriptional regulatory element (PRE) of the woodchuck hepatitis virus. The LTR-driven vectors are shown in (A) and the SIN (self-inactivating) vectors in (B). A SIN configuration is indicated by the deletion of the U3 () of the 3'LTR and the presence of an internal promoter. All vectors are shown as proviruses.

Full figure and legend (131K)

We also constructed SIN vectors with SF placed internally in both the gammaretroviral (Sin.SF) and the lentiviral backbone (RRL.PPT.SF). The lentiviral SIN vectors contained the central polypurine tract (PPT), which is expected to increase the efficiency of reverse transcription and potentially also nuclear translocation³⁴. The gammaretroviral SIN vectors (Sin series) were derived from Sin110.SF, restoring the splice donor upstream of the packaging signal to improve viral titers³⁵. These SIN vectors contained no intron between the promoter and the cDNA. We also tested whether the gammaretroviral intron could be introduced into the lentiviral backbone. To do this, we inserted a fragment containing the gammaretroviral intron derived from SINSF91³⁵ downstream of SF and upstream of the cDNA. The resulting construct was named RRL.PPT.SF91 (Fig. 1B).

In both types of SIN vectors, we also tested internal cellular promoters. In addition to SF, which was previously shown to represent a very strong constitutive promoter for transgene expression in hematopoietic cells²⁵, we inserted the human EF1 promoter, its intron-deleted version (EF1 short or EFS), or the human PGK promoter^{27,28,29,36,37}. EF1 contains an intron in the 5'UTR, resulting in a splice-competent SIN vector. The resulting lentiviral vectors were named RRL.PPT.EF1, RRL.PPT.EFS, and RRL.PPT.PGK, respectively, the corresponding retroviral vectors Sin.EFS and Sin.PGK. To shorten vector names, the uniform presence of the MGMT-PRE cassette is not included in the abbreviations used in the present report.

SIN vectors show correct RNA processing in target cells

All vectors were produced by transient cotransfection of the vector plasmids and packaging constructs into 293T-cell-based packaging systems. In our initial analyses we compared five SIN vectors (Sin.SF, RRL.PPT.SF, RRL.PPT.SF91, RRL.PPT.EF1, RRL.PPT.PGK) and three LTR vectors (SF91, MP71, RRL.SFLTR), following pseudotyping with the ecotropic envelope^38,39. We determined titers following limiting dilution transduction of murine SC-1 fibroblasts (data not shown), using a sensitive FACS assay described below. As shown in Fig. 2A, all SIN vectors except RRL.PPT.EF1 generated 10⁶ infectious particles per milliliter of unconcentrated tissue culture supernatant.

Figure 2.

Vector titers and correct processing of RNA and protein. (A) Comparative titer analysis of gammaretroviral and lentiviral vectors, as determined on murine SC-1 fibroblasts. (B) Vector RNA processing in 32D myeloid cells (Northern blot of total RNA). Transcript-specific RNA was detected with a radiolabeled PRE probe. Lane 1 represents mock-transduced cells, lanes 2 to 9 represent polyclonal populations transduced at equivalent m.o.i. with the different vectors, same order as shown in (A). U, unspliced RNA; S, short RNA, either spliced or internal promoter driven; G, GAPDH loading control. (C) Western blot analysis of 32D cells transduced with gammaretroviral (lanes 2–4) or lentiviral vectors (lanes 5–9, same order as in (A)). The Erk2 protein served as the internal control. (D) Immunofluorescence analysis of SC-1 cells mock-treated and transduced with the gammaretroviral vector Sin.SF. The lentiviral vector RRL.PPT.SF gave data similar to those of Sin.SF (not shown). MGMT protein was detected using a monoclonal MGMT antibody and a secondary FITC-conjugated antibody. Nuclei were stained with DAPI (blue).

Full figure and legend (136K)

Using a m.o.i. of 2, we transduced 32D cells, a murine interleukin-3 (IL-3)-dependent myeloid cell line. We prepared RNA from unselected cells to analyze transcript processing of the vectors following target cell transduction by Northern blot. Fig. 2B shows that the gammaretroviral LTR vectors (lanes 2 and 3) generated mainly spliced transcripts; the larger unspliced transcripts were clearly underrepresented. The splicing efficiency in 32D cells was thus stronger than in our previous studies in fibroblasts³⁵. The LTR-driven lentiviral vector RRL.SFLTR (lane 5) almost exclusively expressed spliced RNA; the genomic band was barely detectable. As the accessory lentiviral protein Rev is required for nuclear export of the lentiviral vector's unspliced RNA²⁴ but not expressed in 32D cells, this result was expected.

As indicated by phosphoimager analyses, the spliced RNA species was weaker in the case of the lentiviral LTR vector than in the case of its gammaretroviral counterpart, despite the presence of largely identical U3 regions (Fig. 2B, compare lanes 5 and 3). This indicates that the lentiviral intron is less efficient than the gammaretroviral intron. The weak unspliced band of the lentiviral LTR vector suggests nuclear retention and/or degradation. The precise mechanism is under investigation. According to the Northern blot, lentiviral vectors containing the gammaretroviral intron (RRL.PPT.SF91) or the cellular intron (RRL.PPT.EF1) showed correct splicing in the target cells (Fig. 2B, lanes 6 and 9).

Using a monoclonal antibody directed against human MGMT, we performed Western blot analyses of cell lysates harvested from the transduced but unselected 32D cell mass cultures (Fig. 2C). A single band of the expected molecular weight was detected in all cases. MGMT protein levels detected by Western blot largely correlated with the RNA levels shown in Fig. 2B. Fluorescence microscopy of transduced cells showed that ectopically expressed MGMT-P140K is correctly localized in the nucleus (Fig. 2D).

Together, these data demonstrate that all vectors mediate correct cellular processing of MGMT-P140K RNA and protein.

FACS analysis allows sensitive detection of MGMT expression in transduced cells and reveals significant differences in MGMT levels depending on vector design

To monitor MGMT expression at the single-cell level, we adapted the FACS assay described by Gerson et al.⁴⁰. Transduced and unselected cells were fixed, permeabilized, and stained with the monoclonal antibody described above. As shown in Fig. 3A for 32D cells, this FACS assay allowed a clear separation of cells overexpressing MGMT from untransduced controls. Interestingly, the higher the MGMT was expressed in the transduced population, the more the fluorescence intensity increased in the negative population. It remains to be tested whether this reflects diffusion of the protein before or after fixation and permeabilization. To separate the two populations, we adjusted the gate settings of the flow cytometer (Fig. 3A).

Figure 3.

MGMT protein expression in transduced 32D cells. (A) FACS analysis of transduced 32D cells in a representative experiment. Intracellular staining was performed using a monoclonal MGMT antibody and a PE-conjugated secondary antibody. A marker was set to separate transduced and untransduced cells. (B) MGMT activity (fmol/g DNA) in extracts of the cells indicated. Raji cell extract served as a control. No MGMT activity was detected in mock-transduced cells (data not shown).

Full figure and legend (173K)

The FACS assay thus allowed a more accurate quantification of MGMT protein levels on a single-cell basis. The mean fluorescence intensity (MFI; Fig. 3A) correlated with the Northern (Fig. 2B) and the Western blots (Fig. 2C). To exclude effects of varying m.o.i., we included only those mass cultures in which the MGMT⁺ cell population did not exceed 30%, which likely reflects conditions of a single transgene copy in the majority of the transduced population⁷. The mass culture transduced with RRL.PPT.PGK had a comparatively low level of MGMT-P140K expression, despite efficient transduction (42% positive cells, Fig. 3A).

According to the MFI, the highest levels of MGMT protein expression were mediated by the vectors containing the SF enhancer–promoter, irrespective of vector architecture (LTR or SIN). In the lentiviral context, the LTR-driven backbone achieved only 25% of the SIN level, which can be explained by the inefficient splicing of the lentiviral leader region in transduced cells (see Fig. 2B). Gammaretroviral and lentiviral SIN vectors with the same internal promoter (SF) showed no significant differences with respect to MGMT-P140K protein expression. This analysis also suggested that EFS and especially PGK were much less powerful internal promoters than SF (Fig. 3A).

MGMT repair activity reveals substantial differences depending on the vectors' cis elements and architecture

Antibody-based assays for MGMT do not necessarily reflect the presence of active protein, as inactivated MGMT is still immunoreactive for more than 18 h⁴¹. To investigate the DNA repair activity following target cell transduction, we harvested pellets of transduced but unselected 32D cells to perform a biochemical assay measuring the functional activity of MGMT, normalized for the amount of DNA present and for transduction frequencies (Fig. 3B). This repair assay confirmed the antibody-based results but revealed even greater differences depending on the design of the lentiviral expression cassettes. Repair activity mediated by PGK was thus at least eightfold less than that mediated by SF.

This assay also allowed us to determine the kinetics of methyl transfer by the transgenic protein. Consistent with previous studies²⁰, the mutant MGMT-P140K present in all vectors had a slower rate of methyl transfer from substrate DNA compared to wild-type protein (supplementary material). Whether this might result in insufficient repair in a rapidly dividing cell population remains to be determined.

Compared to lentiviral SIN vectors, gammaretroviral SIN vectors pseudotyped with RD114/TR envelope show an equal expression but higher titers in human hematopoietic cells

To determine whether murine cells were predictive of vector performance in human cells, we extended our analyses to HT1080 fibrosarcoma cells, K562 leukemia cells, and primary human CD34⁺ cells obtained from GCSF-mobilized peripheral blood. Following reports that the RD114 envelope protein harboring an amphotropic cytoplasmic tail (RD114/TR) provides efficient lentiviral transduction of human hematopoietic cells^30,42, all constructs were pseudotyped with RD114/TR. For these studies we chose seven SIN vectors, three gammaretroviral (Sin.PGK, Sin.EFS, and Sin.SF) and four lentiviral (RRL.PPT.PGK, RRL.PPT.EFS, RRL.PPT.EF1, and RRL.PPT.SF). Using biochemical assays for MGMT detection (as in Fig. 3B), initial analyses of the lentiviral series in K562 cells revealed that SF was the strongest promoter, followed by EF1, EFS, and PGK (Fig. 4A). By flow cytometry of MGMT, we obtained similar data in K562 cells, HT1080 cells, and primary CD34⁺ cells (Figs. 4B and 4C). Interestingly, the difference between EF1 and SF was somewhat smaller than in the murine background. As previously observed with murine cells, gammaretroviral and lentiviral SIN vectors expressed equal levels of MGMT-P140K if containing identical expression cassettes (Fig. 4C). The kinetics of MGMT-P140K-mediated dealkylation also mirrored the murine data (Fig. 4D, compare with supplemental data).

Figure 4.

MGMT expression and titers of RD114/TR pseudotypes in human cells. (A) MGMT activity (fmol/g DNA) in extracts of transduced human K562 cells reveals differential promoter strength. The cells were transduced with the given lentiviral vectors in duplicate. MGMT activity was normalized for transduction efficiency. (B) Representative FACS dot plots are shown of transduced human CD34⁺ cells 5 days after transduction with the lentiviral vectors as indicated. The data were reproduced in three independent experiments (not shown). (C) K562, HT1080, and human CD34⁺ cells were transduced with low m.o.i. of SIN vectors (gammaretroviral Sin, lentiviral RRL.PPT). Means and standard deviations of at least three independent experiments. (D) Kinetics of methyl transfer to wild-type MGMT (present in Raji cells, ) and the P140K mutant present in transduced K562 cells ().

Full figure and legend (151K)

While the lentiviral EF1 vector (RRL.PPT.EF1) mediated promising expression levels of MGMT in CD34⁺ cells (Fig. 4C), it reproducibly gave the lowest titers (Fig. 5). Limiting dilution of RD114/TR pseudotypes on HT1080 cells revealed that gammaretroviral SIN vectors generally had higher titers than their lentiviral counterparts (2–5 times, Fig. 5A). When transducing proliferating hematopoietic cells (K562 or CD34⁺ primary cells), these differences became even more pronounced (10–20 times lower transduction by lentiviral vectors after using the same number of HT1080 transducing units, Figs. 5B and 5C). Being aware of these differences, the above expression data were obtained with comparable transduction efficiency (7–30%).

Figure 5.

Titers of RD114/TR pseudotypes in human cells. (A) End-point titration of retroviral and lentiviral supernatants pseudotyped with the RD114/TR envelope on HT1080 cells and relative transfer efficiency in (B) K562 and (C) primary CD34⁺ cells. Equal titers according to HT1080 results were used in (B) and (C), and the transduction efficiency achieved with RRL.PPT.EF1 was set to 1. Means and standard deviations of three independent experiments.

Full figure and legend (96K)

High levels of MGMT-P140K are required for optimal chemoresistance in primary hematopoietic cells

Finally we investigated whether the vectors expressing relatively low levels of MGMT would still be sufficient for chemoprotection. In all cultures of cell lines tested (K562 and 32D), gammaretroviral or lentiviral SIN vectors expressing MGMT-P140K from different promoters (PGK, EF1, or SF) mediated a survival advantage upon treatment with O⁶BG and BCNU (data not shown). To demonstrate chemoresistance in a more primitive hematopoietic compartment, we performed progenitor assays using colony formation in methylcellulose, following transduction of primary murine bone marrow cells under conditions of a low m.o.i. All vectors tested mediated sufficient levels of MGMT expression to allow detection by flow cytometry (Fig. 6A shows examples for RRL.PPT.PGK and RRL.PPT.SF91). Equivalent levels of protection were achieved with the strong gammaretroviral (Sin.SF) and lentiviral (RRL.PPT.SF91) SIN vectors (Fig. 6B). All surviving colonies contained the MGMT transgene, as determined by PCR (data not shown). However, the PGK promoter (RRL.PPT.PGK) did not provide a significant growth advantage (Fig. 6C). The PGK cassette might be strong enough for chemoprotection only when introduced at more than one copy per cell^23,43.

Figure 6.

MGMT-P140K expressed from the internal SF promoter mediates polyclonal protection of primary bone marrow cells. (A) FACS analysis of murine lineage-depleted bone marrow cells 5 days after transduction with vector RRL.PPT.SF91 or RRL.PPT.PGK. (B) Protection of clonogenic bone marrow cells against alkylation damage subsequent to gene transfer with gammaretroviral and lentiviral vectors expressing high levels of MGMT-P140K under control of the SF promoter. Following transduction, 5 10⁴ cells were treated with 20 M BG for 1 h and plated in methylcellulose containing up to 20 M BCNU. Colonies were enumerated after 7 days. Error bars indicate standard deviations of triplicates. Equal potency of gammaretroviral and lentiviral SIN vectors with internal SF promoters was reproduced in an independent experiment that had a higher transduction efficiency (50% surviving CFU at 10 M BCNU for both Sin.SF and RRL.PPT.SF91). (C) When exposed to stringent drug treatment, chemoprotected colony-forming cells were obtained after transduction of bone marrow cells with RRL.PPT.SF91 (high MGMT expression), but not with RRL.PPT.PGK (low MGMT expression). Following transduction, 5 10⁵ cells were treated with 20 M BG for 1 h and plated in methylcellulose containing 40 M BCNU. Colonies were enumerated after 7 days. Error bars indicate standard deviations of six replicates from two independent experiments. RRL.PPT.PGK also did not provide any significant selective advantage when the dose of BCNU was lowered to 10 and 5 M (data not shown).

Full figure and legend (104K)

Additional studies in vivo in mice, including studies using primary human cells xenografted and drug-exposed in immunodeficient animals^13,44, would be required to reveal whether the EF1 promoter will be sufficient for chemoprotection. The choice of the optimal internal promoter will thus be a balance between the danger of insufficient DNA repair (more likely to occur with a weaker promoter) and the risk of insertional mutagenesis (likely to be more relevant with a stronger promoter or a weaker construct introduced at a high copy number).

While it is tempting to speculate that the increased risk of gammaretroviral vectors to insert in promoter-proximal positions may result in a comparatively reduced safety profile compared with lentiviral vectors⁴⁵, experimental evidence is still needed to address this hypothesis adequately. In our studies investigating insertional leukemias and potentially preleukemic clonal expansions triggered by insertional mutagenesis, hits in the promoter-proximal window were not strongly overrepresented compared to unselected retrovirally transduced cells^3,4. Moreover, the proposed relative reduction of "dangerous" hits when using lentiviral vectors might become meaningless if higher copy numbers and engraftment rates are achieved with this technology⁸. It also needs to be established if gammaretroviral SIN vectors show a different insertion pattern compared to that of LTR-driven counterparts, because the protein–protein interactions that are expected to guide retroviral integration⁴⁶ might partially depend on transcription factors binding to the LTR. Using dose-escalation models³, our vector series could be useful experimentally to address the impact of vector design on insertion pattern and insertional gene activation.

In summary, advantageous features of gammaretroviral vectors are mostly related to the complete absence of any viral gene remnants in the transfer vector, efficient pseudotyping with potent envelopes such as RD114/TR resulting in high transduction rates in proliferating hematopoietic cells, and the lack of mobilization by human-infectious viruses. Compared to HIV-derived vectors, there are also minor concerns related to potential seroconversion following exposure to particle remnants in vivo. Lentiviral vectors are clearly superior for the transduction of cells that reside in the G1 phase of the cell cycle and show a different insertional pattern with a potentially reduced danger of transformation. The present study reveals that both vectors are equally potent in terms of expression properties when containing similar internal expression cassettes. Combining advantageous features in a hybrid technology represents an important aim for future studies.

Material and methods

Plasmids, cell lines, transfections, and transductions

Details of the vector plasmids and the methods used for cell culture and transduction of cell lines are available in the supportive online material.

Primary cells

Bone marrow was harvested from C57BL/6 mice obtained from Charles River Laboratories or The Jackson Laboratory and single-cell suspensions were prepared by gentle pipetting. Low-density mononuclear cells were isolated from bone marrow preparations by centrifugation at 1700 rpm for 20 min on Histopaque-1083 (Sigma Diagnostics, St. Louis, MO, USA). The cells were subsequently cultured for 2 days in Iscove's modified Dulbecco's medium (Mediatech, Inc., Herndon, VA, USA) supplemented with 10% FBS (HyClone, Logan, UT, USA), 100 ng/ml recombinant rat stem cell factor (rrSCF) (Amgen, Thousands Oaks, CA, USA), 100 ng/ml IL-11, 100 ng/ml Flt-3 ligand, 20 ng/ml murine IL-3 (all from Peprotech, Rocky Hill, NJ, USA). Transduction of cultured cells was performed on non-tissue culture-treated six-well plates (Becton–Dickinson Labware, Franklin Hills, NJ, USA) coated with fibronectin fragment CH-296 (8 g/cm²; Takara, Shiga, Otsu, Japan). Four days after the first transduction cells were removed from the wells using Cell Dissociation Buffer (Invitrogen, Grand Island, NY, USA) and transferred to new plates containing fresh medium, and transduction efficiency was analyzed by MGMT expression using flow cytometry as described previously.

A standard progenitor assay was performed to assess the protection against alkylating damage induced by the combination of O⁶BG and BCNU. Briefly, transduced bone marrow cells were treated for 1 h with 10–20 M O⁶BG and washed with PBS. Subsequently, 5 10⁴ to 5 10⁵ cells were plated in 1 ml of medium containing 1% methylcellulose (Stem Cell Technologies, Inc., Vancouver, BC, Canada), 30% fetal calf serum (HyClone), 1% BSA, 10^-4 mol/L 2-mercaptoethanol, 4 U/ml erythropoietin (Amgen), 100 ng/ml rrSCF, 100 ng/ml murine IL-3 (Peprotech), and up to 40 M BCNU (Sigma). Cultures were incubated at 37°C and 5% CO₂, and colonies were scored after 7 days.

Human CD34⁺ cells purified from G-CSF-mobilized peripheral blood of healthy voluntary donors were kindly provided by Cytonet GmbH (Hannover, Germany). Cultivation and transduction were performed as described above except for the cytokine composition. CD34⁺ cells were cultured in 100 ng/ml hFlt-3 ligand, 100 ng/ml hSCF, 20 ng/ml hTPO, and 20 ng/ml hIL-6 (all Peprotech). For transduction an m.o.i. of 5 was used on 2 consecutive days.

MGMT immunofluorescence staining and microscopy

Cells grown on coverslips were fixed with 4% paraformaldehyde for 20 min. Cells were then incubated with 50 mM ammonium chloride for 10 min and permeabilized with 0.1% Triton X-100 for 5 min. MGMT protein was detected using a 1:100 dilution of a murine anti-MGMT monoclonal antibody (clone MT3.1; Chemicon), followed by a 1:100 dilution of a goat anti-mouse FITC antibody (Becton–Dickinson). Incubation with the antibodies was performed for 1 h in a 37°C moist chamber. Fluorescent cells were analyzed on a Zeiss Axioscop fluorescence microscope.

Flow cytometry

For intracellular staining of MGMT, the Cytofix/Cytoperm Kit (Becton–Dickinson) was used according to the manufacturer's instructions. In brief, at least 5 10⁵ cells were harvested and washed in PBS. Cytofix/Cytoperm fixative (4% paraformaldehyde, 250 l) was added for 20 min at 20°C. Washing with 1 ml Perm/Wash buffer was followed by incubation for 30 min at 4°C with 0.25 g of a murine anti-MGMT monoclonal antibody (Chemicon). After two washing steps with Perm/Wash buffer, 1 g of a goat anti-mouse PE-conjugated secondary antibody (Becton–Dickinson) was added for 30 min at 4°C. After two additional washing steps, the samples were analyzed in a FACSCalibur using CellQuest software (Becton–Dickinson). A gate was set on a homogeneous cell population, as determined by scatter characteristics, and 20,000 events were monitored. A marker was set to calculate the percentage and mean fluorescence intensity of positive cells.

Western blot

Cell lysates normalized for protein levels were fractionated on SDS–polyacrylamide gels containing 17.5% polyacrylamide (200:1 ratio of acrylamide to N,N-methylenbisacrylamide). Following electrophoresis, the proteins were transferred to a nitrocellulose membrane (0.45 m; Schleicher and Schuell, Whatman, UK) for 90 min at 20°C. The membrane was blocked with 10% dry milk in PBS for 1 h and stained with a MGMT antibody (MT3.1, diluted 1:2000; Chemicon) overnight, followed by an additional 1-h blocking step and incubation with a peroxidase-conjugated anti-rabbit secondary antibody (Santa Cruz Biotechnology) for 2 h. Detection was by enhanced chemiluminescence (Amersham) according to the manufacturer's protocol.

Northern blot

Total RNA preparation was performed using the RNAzol extraction method (Wako Chemicals, Steinbach, Germany). Ten micrograms of each RNA was separated at 0.6 V/cm² for 4 h in denaturing formaldehyde gels. Subsequently, RNA was transferred to Biodyne-B membranes (0.45 m; Pall Corp., Pensacola, FL, USA) by capillary transfer and heat fixed for 2 h at 80°C. Hybridization was performed using standard procedures as previously described. Specific probes (100 ng) corresponding to the PRE fragment, present in the respective retroviral and lentiviral vectors, were radiolabeled using the DecaLabel DNA labeling kit (MBI Fermentas) to an activity of at least 10⁶ cpm/ml hybridization solution and separated from unincorporated nucleotides on spin columns (Mobitec). Filters were washed, sealed, and exposed to X-ray film (Kodak X-Omat-AR) and quantified by Phosphoimager (Fuji) analysis.

MGMT activity and selection

Transduced 32D cells were harvested 6 days after transduction. MGMT activity was determined in cell-free sonicates as previously described⁴⁷ by quantification of the transfer of [³H]methyl groups from N-[³H]methyl-N-nitrosourea-methylated calf thymus DNA substrate to MGMT protein. Cell pellets of the human lymphoblastoid cell line Raji were used as an assay control. Activity, determined under protein-limiting conditions, was expressed as femtomoles per microgram of total DNA in the extract. The kinetics of methyl transfer was determined by incubation at 37°C for the times indicated under excess substrate conditions. Cells transduced with RRL.SFLTR and Raji cells were used as sources of MGMT-P140K and human wild-type protein, respectively.

For selection of MGMT-transduced cells 1 10⁶ 32D cells were seeded in six-well plates. O⁶BG (25 M; Sigma–Aldrich) was added and, after incubation for 2 h at 37°C, 25 M BCNU (Bristol–Myers Squibb) was added. Five days later, cells were analyzed by flow cytometry as described above.

References

1. Li, Z., et al. (2002). Murine leukemia induced by retroviral gene marking. Science. 296: 497. | Article | PubMed | ISI | ChemPort |
2. Hacein-Bey-Abina, S., et al. (2003). LMO2-associated clonal T cell proliferation in two patients after gene therapy for SCID-X1. Science. 302: 415–419. | Article | PubMed | ISI | ChemPort |
3. Modlich, U., et al. (2005). Leukemias following retroviral transfer of multidrug resistance 1 are driven by combinatorial insertional mutagenesis. Blood. 105: 4235–4246. | Article | PubMed | ISI | ChemPort |
4. Kustikova, O. S., et al. (2005). Clonal dominance of hematopoietic stem cells triggered by retroviral gene marking. Science. 308: 1171–1174. | Article | PubMed | ISI | ChemPort |
5. Baum, C., et al. (2003). Side effects of retroviral gene transfer into hematopoietic stem cells. Blood. 101: 2099–2114. | Article | PubMed | ISI | ChemPort |
6. Yu, S. F., et al. (1986). Self-inactivating retroviral vectors designed for transfer of whole genes into mammalian cells. Proc. Natl. Acad. Sci. USA. 83: 3194–3198. | Article | PubMed | ChemPort |
7. Wahlers, A., et al. (2001). Influence of multiplicity of infection and protein stability on retroviral vector-mediated gene expression in hematopoietic cells. Gene Ther. 8: 477–486. | Article | PubMed | ChemPort |
8. Woods, N. B., et al. (2003). Lentiviral vector transduction of NOD/SCID repopulating cells results in multiple vector integrations per transduced cell: risk of insertional mutagenesis. Blood. 101: 1284–1289. | Article | PubMed | ISI | ChemPort |
9. Gerson, S. L. (2004). MGMT: its role in cancer aetiology and cancer therapeutics. Nat. Rev. Cancer. 4: 296–307. | Article | PubMed | ISI | ChemPort |
10. Milsom, M. D. and Fairbairn, L. J. (2004). Protection and selection for gene therapy in the hematopoietic system. J. Gene Med. 6: 133–146. | Article | PubMed | ISI | ChemPort |
11. Moritz, T. and Williams, D. A. (2003). Methods for gene transfer: genetic manipulation of hematopoietic stem cells. In: Blume K. G., Forman S. J., and Appelbaum F. R. (Eds), 3rd ed., Thomas' Hematopoietic Cell Transplantation. Blackwell Sci. Malden, MA. 107–117.
12. Ragg, S., et al. (2000). Direct reversal of DNA damage by mutant methyltransferase protein protects mice against dose-intensified chemotherapy and leads to in vivo selection of hematopoietic stem cells. Cancer Res. 60: 5187–5195. | PubMed | ISI | ChemPort |
13. Zielske, S. P., Reese, J. S., Lingas, K. T., Donze, J. R. and Gerson, S. L. (2003). In vivo selection of MGMT(P140K) lentivirus-transduced human NOD/SCID repopulating cells without pretransplant irradiation conditioning. J. Clin. Invest. 112: 1561–1570. | Article | PubMed | ISI | ChemPort |
14. Neff, T., et al. (2003). Methylguanine methyltransferase-mediated in vivo selection and chemoprotection of allogeneic stem cells in a large-animal model. J. Clin. Invest. 112: 1581–1588. | Article | PubMed | ISI | ChemPort |
15. Neff, T., et al. (2005). Polyclonal chemoprotection against temozolomide in a large animal model of drug resistance gene therapy. Blood. 105: 997–1002. | Article | PubMed | ISI | ChemPort |
16. Richard, E., et al. (2004). Hematopoietic stem cell gene therapy of murine protoporphyria by methylguanine–DNA–methyltransferase-mediated in vivo drug selection. Gene Ther. 11: 1638–1647. | Article | PubMed | ChemPort |
17. Persons, D. A., et al. (2003). Successful treatment of murine beta-thalassemia using in vivo selection of genetically modified, drug-resistant hematopoietic stem cells. Blood. 102: 506–513. | Article | PubMed | ISI | ChemPort |
18. Bowman, J. E., Reese, J. S., Lingas, K. T. and Gerson, S. L. (2003). Myeloablation is not required to select and maintain expression of the drug-resistance gene, mutant MGMT, in primary and secondary recipients. Mol. Ther. 8: 42–50. | Article | PubMed | ISI | ChemPort |
19. Crone, T. M., Goodtzova, K., Edara, S. and Pegg, A. E. (1994). Mutations in human O⁶-alkylguanine–DNA alkyltransferase imparting resistance to O⁶-benzylguanine. Cancer Res. 54: 6221–6227. | PubMed | ISI | ChemPort |
20. Davis, B. M., et al. (1999). Characterization of the P140K, PVP(138–140)MLK, and G156A O⁶-methylguanine–DNA methyltransferase mutants: implications for drug resistance gene therapy. Hum. Gene Ther. 10: 2769–2778. | Article | PubMed | ISI | ChemPort |
21. Jansen, M., et al. (2001). Protection of hematopoietic cells from O(6)-alkylation damage by O(6)-methylguanine DNA methyltransferase gene transfer: studies with different O(6)-alkylating agents and retroviral backbones. Eur. J. Haematol. 67: 2–13. | Article | PubMed | ISI | ChemPort |
22. Zielske, S. P. and Gerson, S. L. (2002). Lentiviral transduction of P140K MGMT into human CD34(+) hematopoietic progenitors at low multiplicity of infection confers significant resistance to BG/BCNU and allows selection in vitro. Mol. Ther. 5: 381–387. | Article | PubMed | ISI | ChemPort |
23. Chinnasamy, D., et al. (2004). Lentivirus-mediated expression of mutant MGMTP140K protects human CD34⁺ cells against the combined toxicity of O⁶-benzylguanine and 1,3-bis(2-chloroethyl)-nitrosourea or temozolomide. Hum. Gene Ther. 15: 758–769. | Article | PubMed | ChemPort |
24. Dull, T., et al. (1998). A third-generation lentivirus vector with a conditional packaging system. J. Virol. 72: 8463–8471. | PubMed | ISI | ChemPort |
25. Baum, C., Hegewisch-Becker, S., Eckert, H. G., Stocking, C. and Ostertag, W. (1995). Novel retroviral vectors for efficient expression of the multidrug-resistance (mdr-1) gene in early hemopoietic cells. J. Virol. 69: 7541–7547. | PubMed | ISI | ChemPort |
26. Lim, B., Williams, D. A. and Orkin, S. H. (1987). Retroviral-mediated gene transfer of human adenosine deaminase: expression of functional enzyme in murine hematopoietic stem cells in vivo. Mol. Cell. Biol. 7: 3459–3465. | PubMed | ChemPort |
27. Salmon, P., et al. (2000). High-level transgene expression in human hematopoietic progenitors and differentiated blood lineages after transduction with improved lentiviral vectors. Blood. 96: 3392–3398. | PubMed | ISI | ChemPort |
28. Woods, N. B., et al. (2000). Lentiviral gene transfer into primary and secondary NOD/SCID repopulating cells. Blood. 96: 3725–3733. | PubMed | ISI | ChemPort |
29. Woods, N. B., Ooka, A. and Karlsson, S. (2002). Development of gene therapy for hematopoietic stem cells using lentiviral vectors. Leukemia. 16: 563–569. | Article | PubMed | ISI | ChemPort |
30. Sandrin, V., et al. (2002). Lentiviral vectors pseudotyped with a modified RD114 envelope glycoprotein show increased stability in sera and augmented transduction of primary lymphocytes and CD34⁺ cells derived from human and nonhuman primates. Blood. 100: 823–832. | Article | PubMed | ISI | ChemPort |
31. Zufferey, R., Donello, J. E., Trono, D. and Hope, Z. J. (1999). Woodchuck hepatitis virus posttranscriptional regulatory element enhances expression of transgenes delivered by retroviral vectors. J. Virol. 73: 2886–2892. | PubMed | ISI | ChemPort |
32. Egelhofer, M., et al. (2004). Inhibition of HIV-1 entry in cells expressing Gp41-derived peptides. J. Virol. 78: 568–575. | Article | PubMed | ISI | ChemPort |
33. Schambach, A., et al. (2000). Context dependence of different modules for posttranscriptional enhancement of gene expression from retroviral vectors. Mol. Ther. 2: 435–445. | Article | PubMed | ISI | ChemPort |
34. Follenzi, A., Ailles, L. E., Bakovic, S., Geuna, M. and Naldini, L. (2000). Gene transfer by lentiviral vectors is limited by nuclear translocation and rescued by HIV-1 pol sequences. Nat. Genet. 25: 217–222. | Article | PubMed | ISI | ChemPort |
35. Kraunus, J., et al. (2004). Self-inactivating retroviral vectors with improved RNA processing. Gene Ther. 11: 1568–1578. | Article | PubMed | ChemPort |
36. Mikkola, H., et al. (2000). Lentivirus gene transfer in murine hematopoietic progenitor cells is compromised by a delay in proviral integration and results in transduction mosaicism and heterogeneous gene expression in progeny cells. J. Virol. 74: 11911–11918. | Article | PubMed | ChemPort |
37. Woods, N. B., et al. (2001). Lentiviral-mediated gene transfer into haematopoietic stem cells. J. Intern. Med. 249: 339–343. | Article | PubMed | ChemPort |
38. Li, Z., et al. (2003). Predictable and efficient retroviral gene transfer into murine bone marrow repopulating cells using a defined vector dose. Exp. Hematol. 31: 1206–1214. | Article | PubMed | ISI | ChemPort |
39. Hanawa, H., et al. (2002). Comparison of various envelope proteins for their ability to pseudotype lentiviral vectors and transduce primitive hematopoietic cells from human blood. Mol. Ther. 5: 242–251. | Article | PubMed | ISI | ChemPort |
40. Zielske, S. P. and Gerson, S. L. (2004). Limited lentiviral transgene expression with increasing copy number in an MGMT selection model: lack of copy number selection by drug treatment. Mol. Ther. 9: 923–931. | Article | PubMed | ISI | ChemPort |
41. Liu, L., et al. (2001). Differential degradation rates of inactivated alkyltransferase in blood mononuclear cells and tumors of patients after treatment with O(6)-benzylguanine. Clin. Cancer Res. 7: 2318–2324. | PubMed | ISI | ChemPort |
42. Relander, T., et al. (2005). Gene transfer to repopulating human CD34⁺ cells using amphotropic-, GALV-, or RD114-pseudotyped HIV-1-based vectors from stable producer cells. Mol. Ther. 11: 452–459. | Article | PubMed | ChemPort |
43. Richard, E., et al. (2003). A bicistronic SIN-lentiviral vector containing G156A MGMT allows selection and metabolic correction of hematopoietic protoporphyric cell lines. J. Gene Med. 5: 737–747. | Article | PubMed | ISI | ChemPort |
44. Pollok, K. E., et al. (2003). In vivo selection of human hematopoietic cells in a xenograft model using combined pharmacologic and genetic manipulations. Hum. Gene Ther. 14: 1703–1714. | Article | PubMed | ISI | ChemPort |
45. Wu, X., Li, Y., Crise, B. and Burgess, S. M. (2003). Transcription start regions in the human genome are favored targets for MLV integration. Science. 300: 1749–1751. | Article | PubMed | ISI | ChemPort |
46. Bushman, F. D. (2003). Targeting survival: integration site selection by retroviruses and LTR-retrotransposons. Cell. 115: 135–138. | Article | PubMed | ISI | ChemPort |
47. Watson, A. J. and Margison, G. P. (2000). O⁶-Alkylguanine–DNA alkyltransferase assay. Methods Mol. Biol. 152: 49–61. | PubMed | ChemPort |

Appendices

Appendix A

Supplementary data

Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ymthe.2005.08.012.

Acknowledgements

This work was supported by grants from the European Union (CONSERT, LSHB-CT-2004-005242) and the Deutsche Forschungsgemeinschaft (DFG BA 1837/4). We are grateful to Cornelia Rudolph for assistance with fluorescence microscopy. We thank the CCHMC Translational Trials Development and Support Laboratory for CD34⁺ cells, Maimona Id for technical assistance, and Michael Milsom (Paterson Institute for Cancer Research, Manchester, UK) for fruitful discussions. This work is dedicated to Leslie Fairbairn.