Methotrexate (MTX) dose-escalation studies were conducted in C57BL/6 mice to determine the chemoprotective effect of transplantation using bone marrow transduced with lentivirus vectors expressing a drug-resistant variant of murine dihydrofolate reductase (DHFR). Methotrexate-resistant dihydrofolate reductase [tyrosine-22 (Tyr22)DHFR] and enhanced green fluorescent protein (GFP) coding sequences were inserted into self-inactivating lentiviral vectors as part of a genetic fusion or within the context of a bicistronic expression cassette. MTX-treated animals that received Tyr22DHFR-transduced marrow recovered to normal hematocrit levels by 3 weeks post-transplant and exhibited significant GFP marking in myeloid and lymphoid lineage-derived peripheral blood mononuclear cells (PBMCs). In contrast, MTX-treated animals transplanted with control GFP-transduced marrow exhibited extremely reduced hematocrits with severe marrow hypoplasia and did not survive MTX dose escalation. To minimize cell manipulation, we treated unfractionated marrow in an overnight exposure. Transduction at a multiplicity of infection of 10 resulted in up to 11% vector-modified PBMCs in primary recipients and successful repopulation of secondary recipients with vector-marked cells. Experimental cohorts exhibited sustained proviral expression with stable GFP fluorescence intensity. These results demonstrate the effectiveness of lentivirus vectors for chemoprotection in a well developed animal model, with the potential for further preclinical development toward human application.
Methotrexate (MTX; 4-amino-10-methylfolic acid) is an antiproliferative chemotherapeutic that interrupts folate metabolism by inhibition of dihydrofolate reductase (DHFR; EC 188.8.131.52). DHFR catalyzes the reduction of dihydrofolate to tetrahydrofolate, a required precursor for cofactors involved in macromolecule biosynthesis, one-carbon transfer reactions, and other cellular metabolic pathways (Blakley and Benkovic, 1984). MTX has been successfully used to treat a number of malignancies, such as acute lymphoblastic leukemia, non-Hodgkin's lymphoma, and osteosarcoma (Bertino, 1993). However, the therapeutic dose that can be administered is limited by toxicity to rapidly dividing cells of the bone marrow and gastrointestinal tissues (Sirotnak and Moccio, 1980; Rivera et al., 1985).
One potential way to protect against MTX toxicity is by the expression of drug-resistant DHFR in normal drug-sensitive cells and tissues. MTXr-DHFR variants have been shown to mediate protection from antifolate toxicity when expressed in DHFR transgenic hematopoietic cells after transplantation (May et al., 1995, 1996; Morris et al., 1996; James et al., 1997, 2000). Improved antifolate chemotherapy has also been shown in mice transplanted with DHFR transgenic marrow or with γ-retrovirus-transduced marrow cells expressing drug-resistant DHFR (Mayer-Kuckuk et al., 2002; Sweeney et al., 2003). Selective expansion of tyrosine 22 (Tyr22) DHFR-transduced hematopoietic cells reported in murine and nonhuman primate models (Allay et al., 1997; Persons et al., 2004) also supports the potential utility of DHFR gene transfer for protection of normal cells during administration of chemotherapy using antifolates.
Gamma-retrovirus vectors transducing MTXr-DHFR cDNAs, sometimes in combination with other drug-resistance genes, have been evaluated extensively for gene transfer and drug resistance in mammalian cells and in mouse models (Williams et al., 1987; Corey et al., 1990; Li et al., 1994; Budak-Alpdogan et al., 2004; Capiaux et al., 2004). However, clinical application of MTXr-DHFR expression will require the most effective means available to achieve gene transfer into hematopoietic cell targets. Lentivirus vectors have been shown to be highly effective in mediating gene transfer into HSCs, particularly within the context of ex vivo gene transfer into HSCs of dogs and nonhuman primates (Hanawa et al., 2004; Horn et al., 2004). Here, we report the generation of lentivirus vectors expressing murine MTXr-DHFR (Tyr22), and we demonstrate effective protection of mice from lethal doses of MTX by transplantation with congenic marrow after a single overnight vector transduction. These results suggest the development of a clinically applicable ex vivo DHFR gene transfer procedure for reduced toxicity associated with antifolate chemotherapy.
Materials and Methods
Lentivirus Vector Construction. Recombinant plasmids were generated using standard molecular cloning techniques. Lentivirus vector plasmids pCSIIEG and pCSII have been described previously (Agarwal et al., 2006). To construct pEFDIG, a Tyr22DHFR coding sequence was obtained as a XhoI-ClaI (blunt) fragment from pLasBD (Zhao et al., 1997) and cloned between XhoI and BamHI (blunt) of pCSII-CMV-12G (N. Somia, unpublished data) to form pCCDG. A DHFR-internal ribosomal entry site (IRES)-GFP fragment was then generated by polymerase chain reaction (PCR) from pCCDG template plasmid using sense (5′-GCGAATTCTCGAGGGTCCTCTAGAGCAAG-3′) and anti-sense (5′-CGCTGCAGCCTCGATGTTAACTCTAGAGTCG-3′) primers to introduce EcoRI and PstI restriction sites at the 5′ and 3′ fragment ends, respectively, and cloned into pCSII. For construction of DL2G, a DHFR-GFP fusion sequence was generated by two-step PCR overlap extension. First, an antisense oligonucleotide (5′-ACCTCCTCCtcaTCCACCACCCCTGTCTTTCTTCTCGTAGACTTCAAACTT-3′) was designed to replace the TGA stop codon (underlined) of the Tyr22DHFR cDNA with glycine while providing additional sequences encoding eight amino acids of a [Gly4Ser]2 linker (glycine in italics, serine in lowercase), included to impart three-dimensional flexibility between the Tyr22DHFR and enhanced (e)GFP domains of the fusion protein (Lewis et al., 2003). The Tyr22DHFR cDNA was amplified using this antisense primer along with a sense oligonucleotide (5′-GCGAATTCTCGAGGGTCCTCTAGAGCAAG-3′) to introduce an EcoRI site (underlined) at the 5′ end. A sense oligonucleotide (5′-GGTGGAtcaGGAGGAGGTGGTtctGCGGTGAGCAAGGGCGAGCTGTTCACCGGG-3′) was designed to provide sequences overlapping the DHFR antisense primer encoding two glycine (italics) and two serine (lowercase) residues of the first glycine-serine repeat, add a second Gly4Ser repeat, and convert the eGFP start codon to an alanine codon (underlined). This sense oligonucleotide was used along with an antisense oligonucleotide (5′-CGCTGCAGCCTCGATGTTAACTCTAGAGTCG-3′) to amplify the eGFP coding sequence and to introduce a PstI site (underlined) at the 3′ end. The amplified Tyr22DHFR-Gly4Ser-Gly3 and Gly2Ser-Gly4Ser-eGFP products were fused by overlap extension [94°C for 5 min, slow cool for 30 min in air, add dNTPs and Taq/Tgo DNA polymerase mixture (Expand High Fidelity Enzyme mix; Roche Applied Science, Penzberg, Germany), and then heat to 72°C for 5 min], and the final product was amplified by PCR using the far 5′ sense (DHFR) and far 3′ antisense (eGFP) oligonucleotides described above. The final PCR product was ligated into pCSII between EcoRI and PstI. Correct vector construction was verified by restriction endonuclease mapping and DNA sequencing. All restriction enzymes were from New England Biolabs (Beverly, MA). pCCDG was used as template for both DHFR and GFP amplifications, with PCR reactions carried out under the following conditions: 94°C for 10 min, 1 cycle; 94°C for 30 s, 55°C for 30 s, 72°C for 3 min, 30 cycles; and 72°C for 10 min, 1 cycle. All PCR reactions and overlap extensions were carried out using an Applied Biosystems (Foster City, CA) Thermal Cycler model 2720.
Preparation of High-Titer Vector Stocks. Lentiviral vectors were packaged using a three-plasmid transient transfection procedure as described previously (Zufferey et al., 1997). Lentivirus vector plasmids (pCSIIEG, pEFDIG, or pDL2G) were cotransfected with pΔNRF to provide the gag, pol, and rev proteins and pMD.G for pseudotyping with vesicular stomatitis virus G protein envelope protein. In brief, 6 × 106 human embryonic kidney 293T cells were seeded onto poly-l-lysine-coated 15-cm2 plates 24 h before transfection in Dulbecco's modified Eagle's medium (DMEM) supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin (P/S) (all from Invitrogen, Carlsbad, CA), and 10% heat-inactivated fetal bovine serum (FBS; Atlas Biologicals, Fort Collins, CO) in a humidified atmosphere containing 3% CO2 in air at 37°C. For the production of vector particles, 8 μg of pMD.G, 30 μgofpΔNRF packaging plasmid, and 25 μg of lentiviral vector plasmid were cotransfected into 293T cells using the DNA-calcium phosphate coprecipitation technique (Tiscornia et al., 2006). After overnight transfection, cells were washed with PBS, provided with fresh DMEM plus 10% FBS, 1% P/S, and 10 mM sodium butyrate (Sigma-Aldrich, St. Louis, MO), and incubated in a humidified atmosphere containing 10% CO2 in air at 37°C. Vector supernatants were collected 8 h later and replaced with DMEM plus 10% FBS and 1% P/S for two additional 12-h incubations. Supernatants from all three harvests were pooled, concentrated 100-fold by centrifugation at 23,000g, and resuspended in unsupplemented Iscove's modified Dulbecco's medium (Invitrogen). Concentrated supernatants were diluted and titrated on NIH3T3 TK– (lacking thymidine kinase) murine fibroblasts. Cells and virus were incubated for 48 h in the presence of 8 μg/ml polybrene (hexadimethrine bromide; Sigma-Aldrich) followed by fluorescence-activated cytometric analysis to determine the percentage of GFP+ cells using a FACSCalibur instrument (BD Biosciences, San Jose, CA). Titer was also assessed by drug-resistant colony formation after subculturing transduced 3T3 TK– cells into selective medium containing 0.15 μM MTX (Vinh and McIvor, 1993). After 2 weeks, colonies were stained with crystal violet, and the MTXr-DHFR titer was calculated as the number of colony-forming units per milliliter. Vector concentrations were increased approximately 100-fold with 70% recovery of transducing units, and final titers ranged between 107 and 108 transducing units per milliliter as assessed both by drug resistance (MTXr-DHFR) and flow cytometry (GFP).
Fluorescence Microscopy. 3T3 TK– fibroblasts were transduced with EFDIG or DL2G lentiviral vector and subcultured in 100 μM MTX. After 2 weeks, MTX-resistant clones were isolated and expanded in culture. For fluorescence microscopy, MTX-resistant clones and uninfected cells were subcultured onto coverslips coated with 0.002% poly-l-lysine (Sigma-Aldrich). Confluent monolayers were fixed with 4% paraformaldehyde (Sigma-Aldrich), washed with PBS, and mounted onto glass slides with VectaMount (Vector Laboratories, Burlingame, CA). Fluorescence was visualized on an Olympus BX60 upright microscope (Olympus, Tokyo, Japan).
Western Blot Analysis. MTXr-clones (described in the preceding section) and uninfected cells were harvested by trypsinization, washed with phosphate-buffered saline (PBS; Cambrex Bio Science, Walkersville, MD), and lysed with 2% sodium dodecyl sulfate (SDS; Sigma-Aldrich). Cleared lysates were prepared by centrifugation at 15,000g for 20 min. Protein extracts (25 μg, quantified by Bradford assay; Pierce Chemical, Rockford, IL) were boiled for 5 min in the presence of sample loading buffer and electrophoresed through 12% polyacrylamide-SDS. Proteins were transferred onto nitrocellulose membranes using the iBLOT Dry Blotting System (Invitrogen) and then incubated sequentially with either a polyclonal rabbit antibody raised against human DHFR (a kind gift from Dr. B. Dolnick, Roswell Park Memorial Institute, Buffalo, NY; Morris and McIvor, 1994) or a monoclonal rabbit antibody against eGFP (Clontech, Mountain View, CA) followed by a secondary anti-rabbit IgG antibody conjugated with horseradish peroxidase (Invitrogen).
DHFR Enzyme Assay. 3T3 TK– cells were transduced with CSIIEG, EFDIG, or DL2G lentivirus vector at a multiplicity of infection of 1, harvested by trypsinization, washed with PBS, and then resuspended in lysis buffer [50 mM Tris buffer, pH 7.5, 150 mM KCl, and 25 mM MgCl2 (all from Sigma-Aldrich), 10 mM β-mercaptoethanol (Bio-Rad, Hercules, CA)]. The cells were then sonicated with a Branson 250/450 Sonifier at a setting of 5 for 20 s using a microtip and cleared by centrifugation (16,000g for 30 min at 4°C). Extracts were diluted into reaction buffer (100 mM Tris buffer, pH 7.5, 150 mM KCl, and 10 mM β-mercaptoethanol). When including MTX in the reaction, samples were incubated with drug at room temperature for 10 min before enzyme assay. Reactions were started by adding NADPH (β-nicotinamide adenine dinucleotide phosphate reduced tetrasodium salt) and dihydrofolic acid (both from Sigma-Aldrich) to final concentrations of 120 and 30 μM, respectively, following the change in absorbance at 340 nm on a DU40 spectrophotometer (Beckman Coulter, Fullerton, CA). One unit was defined as the amount of enzyme required to reduce 1 nmol of dihydrofolic acid per minute (ϵ = 12,300 M–1 cm–1) (Carr et al., 1983).
Mouse Bone Marrow Transplantation, ex Vivo Transduction, and Methotrexate Administration. All procedures were reviewed and approved by the University of Minnesota Institutional Animal Care and Use Committee. Six-week-old C57BL/6 CD45.2 and CD45.1 female mice were obtained from the National Cancer Institute (Frederick, MD) and provided with food and water ad libitum. Donor CD45.2 mice were administered 150 mg/kg 5-fluorouracil (Sigma-Aldrich) 2 days before marrow harvest. Bone marrow was flushed from the hind limbs of donor mice and processed to a single-cell suspension in DMEM plus 10 U/ml heparin and 10% FBS. Cell suspensions were quantitated and assayed for viability by trypan blue exclusion. Following marrow harvest and processing, cells were resuspended in complete Stempro medium supplemented with 2 mM l-glutamine, 100 IU/ml penicillin, 100 μg/ml streptomycin (Invitrogen), cytokines (20 ng/ml human IL-3, 50 ng/ml human IL-6, 100 ng/ml murine SCF; all from R&D Systems, Minneapolis, MN) and 8 μg/ml polybrene (Sigma-Aldrich). All transductions were carried out as 12-h vector exposures following the addition of 100-fold concentrated vector to 2.25 × 108 cells plated in three 15-cm2 tissue culture plates. Concentrated vector was added to a final volume of 15 ml/plate at a multiplicity of infection of 10 (MTXr CFU/marrow cell). Transduced cells were washed, enumerated, and resuspended at a final concentration of 1 to 2 × 107/ml DMEM, and then equal numbers (2 or 3 × 106) were injected through the lateral tail vein into 12- to 16-week-old sublethally irradiated (700 cGy) congenic recipients (CD45.1). Mice that received MTX (+ amethopterin; Bedford Laboratories, Bedford, OH) were injected intraperitoneally with 0.25 mg/kg MTX on days 1 to 4, 0.5 mg/kg MTX on days 5 to 8, 1 mg/kg MTX on days 9 to 25, and 1 to 2 mg/kg on days 26 to 35 or until recovery of hematocrit to the normal range. Control mice were injected with equivalent volumes of PBS. For secondary transplant, marrow harvested from primary recipients was injected into lethally irradiated (800–850 cGy) CD45.1 secondary recipients, which were allowed to recover for 4 months before analysis of engraftment and transduction.
Hematology and Flow Cytometry. Blood was obtained from animals at regular intervals for determination of hematocrit and for flow cytometric analysis. Whole blood samples were resuspended in hemolysis buffer (0.15 M NH4Cl, 1 M NaHCO3, and 0.1 M Na2EDTA, pH 7.2), and leukocytes were pelleted by low-speed centrifugation, washed with PBS, and stained with fluorochrome-conjugated monoclonal antibodies for determination of immunophenotype. GFP expression was analyzed as described previously (Kurre et al., 2004). Engraftment and leukocyte immunophenotyping were conducted using anti-mouse monoclonal antibodies: allophycocyanin-conjugated CD45.2, phycoerythrin-conjugated CD3e (T lymphocytes), B220 (B lymphocytes), CD11b and Gr-1 (myeloid lineages) (all purchased from eBiosciences, San Diego, CA). Data were collected on a FACS-Calibur instrument (BD Biosciences) and analyzed using CellQuest Pro (BD Biosciences) and FlowJo (Tree Star, Inc., Ashland, OR) software, respectively.
Histopathologic Analysis. Animals were euthanized, and tissue samples were harvested, including sternum and ileum. Tissues were fixed in 10% formalin (further decalcifying bones in 1% formic acid; both from Sigma-Aldrich), embedded in paraffin, sectioned, mounted, stained (hematoxylin and eosin), and analyzed microscopically without prior knowledge of sample identity.
Real-Time Quantitative PCR. DNA was extracted from samples using the Puregene DNA Purification System (Gentra Systems, Inc., Minneapolis, MN) and quantitated spectrophotometrically (A260). Reactions were run in duplicate under the following conditions: 60 ng of DNA, 200 nM forward (5′-CACATGAAGCAGCACGACTT-3′) and reverse (5′-GGTCTTGTAGTTGCCGTCGT-3′) GFP-specific primers, 100 nM GFP-specific probe (5′/56-5-carboxyfluorescein/AGCGCACCATCTTCTTCAAG/3BHQ_1/-3′), and 5% dimethyl sulfoxide (Sigma-Aldrich) in a 10-μl final reaction volume. The LightCycler 2.0 System (Roche Applied Science) was used for amplification of the target sequence with the following cycling parameters: 95°C for 10 min, 1 cycle; 95°C for 10 s, 60°C for 20 s, 45 cycles; and 40°C for 30 s, 1 cycle). Results were analyzed using LightCycler software, version 3.5, and copy number was determined by the second derivative maximal analysis method (i.e., the cycle number at which sample fluorescence is first detected above background fluorescence). A standard curve was generated by first isolating a HeLa cell clone in selective medium containing 0.2 μM MTX after exposure to DHFR lentivirus vector CCDG (see above) at low multiplicity as a source of DNA containing a single copy integrant. CCDG/HeLa DNA was mixed with unexposed HeLa cell DNA in 10-fold serial dilutions to generate a standard curve covering a range of 104 copies to 1 copy of the integrated vector sequence per 60 ng of template DNA. Negative controls for each reaction included no template control, HeLa cell DNA, and DNA extracted from C57BL/6 mouse bone marrow.
Statistical Analysis. Animal survival in response to MTX administration was evaluated by the Kaplan-Meier product limit method (Kaplan and Meier, 1958). The log rank statistic was determined using Prism 4 software (GraphPad Software, Inc., San Diego, CA) to test for differences among recipient cohorts (Peto and Peto, 1972). Data from other experiments were evaluated by unpaired t test analysis.
Generation and in Vitro Characterization of Lentivirus Vectors Transducing MTXr-DHFR and GFP. To achieve high-efficiency gene transfer in hematopoietic cell targets, we constructed lentivirus vector plasmids pEFDIG and pDL2G, containing a murine MTXr-Tyr22DHFR variant cDNA along with the eGFP marker gene under transcriptional control of the human elongation factor 1-α (EF-1α) promoter. The Tyr22DHFR and eGFP coding sequences were inserted as part of a bicistronic expression cassette (pEFDIG) or as a genetic fusion (pDL2G) (Fig. 1). The EFDIG and DL2G vectors were packaged into vesicular stomatitis virus G protein pseudotyped virions and concentrated 100-fold. Titer was assessed in comparison with the eGFP-encoding vector CSIIEG (Fig. 1). Vector titers ranged between 107 and 108 IU/ml (Table 1), with reasonably good correlation between titer assessed by flow cytometry for GFP expression and by MTXr colony formation for DHFR expression. Mean fluorescence intensity (MFI) of GFP-positive 3T3 TK– cells transduced at a multiplicity of infection of 0.2 or less was higher for the DL2G fusion in comparison with the EFDIG-infected cells, indicating that DL2G may be a more reliable vector design to achieve GFP marking of transduced cell populations.
To characterize Tyr22DHFR and eGFP protein expressed in transduced cells, clones resistant to a high level of MTX (100 μM, to select for clones expressing a high level of DHFR) were isolated for examination by fluorescence microscopy, and extracts were prepared for Western blot analysis (Fig. 2). Tyr22DHFR and eGFP proteins were clearly identifiable in Western blots (Fig. 2A). The presence of a 48-kDa band in MTXr-DL2G lysates immunoreactive with both anti-eGFP and anti-DHFR antibodies confirmed successful generation of the Tyr22DHFR-eGFP fusion protein, consistent with the molecular masses of eGFP (27 kDa), DHFR (21 kDa), and the glycine-serine linker (0.8 kDa). In contrast, a 21-kDa band was detected in EFDIG-transduced cell lysates probed with anti-DHFR antibody, consistent with the molecular weight of endogenous murine DHFR. A 27-kDa band was also detected in GFP-probed EFDIG lysates, consistent with the molecular mass of eGFP. In uninfected and DL2G-transduced cells, endogenous DHFR was detected at greater exposure times (data not shown). Fluorescence microscopy of MTXr-clones showed a higher level of fluorescence in DL2G-positive cells in comparison with EFDIG-positive cells (Fig. 2B).
To determine the effect of the DL2G fusion on DHFR enzyme activity, 3T3 TK– cells were transduced at a multiplicity of 1 (based on GFP titer), maintained in culture for 7 days, and then harvested for extract preparation and DHFR enzyme assay. The enzyme activity per milligram of total protein was approximately 4 times lower in cells transduced with DL2G compared with EFDIG in the absence of MTX, despite similar levels of GFP expression as determined by flow cytometry (Table 2). When drug was added to the extracts, enzyme activity in EFDIG- and DL2G-transduced cells decreased by 27 and 76%, respectively. When units per milligram was evaluated within the context of copy number, there was no significant difference between EFDIG- and DL2G-transduced cells, indicating the maintenance of DHFR enzyme activity of the fusion protein.
Transplantation with EFDIG-Transduced Marrow Protects Recipients from Lethal Doses of MTX. To determine the level of chemoprotection conferred by lentivirus-mediated Tyr22DHFR gene transfer, we conducted dose-escalation studies in animals that had received either CSIIEG- or EFDIG-transduced marrow. Following overnight transduction in the presence of IL-3, IL-6, and SCF, 3 × 106 transduced, unfractionated marrow cells were infused into sublethally irradiated (700 cGy) recipients. Animals were then administered either PBS or MTX at ascending doses, from 0.25 mg/kg/day on day 1 up to 2 mg/kg/day on day 35. Mice that were transplanted with GFP-transduced marrow and subsequently administered MTX did not survive dose escalation to 2 mg/kg/day (Fig. 3A). Two of the CSIIEG-marrow recipients administered MTX that survived until day 35 exhibited poor health, as indicated by 25% loss of body weight, lethargy, and hematocrits below 15 (Fig. 3B). In contrast, animals transplanted with EFDIG-transduced marrow exhibited a significant survival advantage (p = 0.0102), and most of these animals survived escalation of the MTX dose up to 2 mg/kg/day. EFDIG-marrow recipients recovered to healthy hematocrit levels by day 20, exhibiting a similar hematocrit recovery pattern as that observed for the PBS-administered control cohort. Histopathological analysis of marrow from MTX-treated control (GFP-transduced) recipients revealed severe hypoplasia, consistent with myelosuppression. In contrast, bone marrow from the EFDIG/MTX and CSIIEG/PBS cohorts showed marrow hyperplasia. Histopathological analysis of samples from the ileum revealed normal crypt/villus ratios and a lack of lesions (data not shown).
Donor cell engraftment levels were assessed in recipient animals by flow cytometric analysis for CD45.2+ leukocytes in peripheral blood at 4 and 8 weeks post-transplantation. For PBS-administered control animals, donor engraftment increased from 60% at 4 weeks to 82% at 8 weeks (Table 3). Lower levels of engraftment were observed in MTX-treated EFDIG-marrow recipients 4 weeks post transplantation (30%). By week 8, engraftment increased to 82%. GFP marking was higher in MTX-treated EFDIG marrow recipients (36%) compared with PBS-treated animals (15%) during the dose escalation, but marking decreased to 9 to 10% in both cohorts 1 month after MTX withdrawal. These data show selective expansion of Tyr22DHFR-expressing cells resulting from MTX administration.
MTX Dose-Escalation Studies in DL2G-Transduced BMT Recipients. To assess whether the Tyr22DHFR-GFP fusion protein protects recipients from MTX-toxicity, we conducted MTX dose-escalation studies in mice transplanted with 2 × 106 DL2G-transduced marrow cells. Donor CD45.2 marrow cells were harvested and transduced overnight with DL2G or CSIIEG in the presence of IL-3, IL-6, and SCF as described under Materials and Methods. In comparison with the EFDIG-transduced marrow transplant experiment described above, fewer cells were infused due to decreased viable cell yield after overnight transduction. After transplantation, recipient animals were administered either PBS or MTX at an initial dose of 0.25 mg/kg, subsequently increasing the dose to 0.5 mg/kg and then 1.0 mg/kg on day 9 post-transplant. Control CSIIEG-transduced marrow recipients did not survive MTX dose escalation to 1 mg/kg/day and succumbed by day 16 (p = 0.0398) (Fig. 3C). Declining health of all animals, including PBS-treated controls, necessitated a 5-day suspension of MTX administration, from day 13 to 17 post-transplant. Once the animals regained their health based on weight, appearance, and behavior, treatment was resumed and continued until the end of the 30-day dose escalation. We observed that DL2G marrow recipients survived MTX dose escalation to 1 mg/kg/day. In comparison with the MTX-treated CSIIEG-marrow recipients, MTX- and PBS-treated DL2G-marrow recipients recovered to normal hematocrit levels by day 21 (Fig. 3D). Histopathological analysis of marrow and small intestine (ileum) sections from MTX-treated CSIIEG-marrow recipients revealed severe marrow hypoplasia but no evidence of GI atrophy, respectively (data not shown). In contrast, DL2G-transduced marrow recipients treated with MTX or PBS exhibited hyperplastic marrow and normal GI histology.
MTX- and PBS-treated DL2G-marrow recipients exhibited 9 and 14% CD45.2 donor cell marking, respectively, in peripheral blood at the endpoint of drug administration (31 days). By 8 weeks after marrow transplantation, donor cell engraftment increased to 59 and 60% for these cohorts, respectively (Table 3). In PBS-administered control animals, GFP marking in peripheral blood was stable between 4 and 8 weeks (3% GFP-positive PBMCs). In contrast, GFP marking in MTX-treated DL2G-marrow recipients decreased by 8 weeks (from 13 to 2% GFP-positive PBMCs). GFP marking was observed in donor myeloid and lymphoid lineages at mean frequencies from 10 to 60% (Table 4). The greater level of GFP marking in MTX-treated animals implies a selective advantage for Tyr22DHFR-expressing donor hematopoietic cells over untransduced donor cells during MTX administration. Although antifolate administration alone is not sufficient for in vivo selection of hematopoietic stem cells (Allay et al., 1998), Tyr22DHFR-myeloid and lymphoid progeny seem to be less sensitive to MTX toxicity; therefore, they may have a survival advantage over their untransduced counterparts.
Proviral integrant frequencies (determined by real-time quantitative PCR) in the bone marrow of primary transplant recipients were similar among those animals within each experiment (Table 3). The higher proviral integrant frequency in experiment 1 compared with experiment 2, in combination with the reduced number of marrow cells transplanted in experiment 2, may account for the decreased recovery time and higher level of chemoprotection achieved in MTX-treated EFDIG-marrow recipients (2 versus 1 mg/kg/day).
Repopulation of Secondary Recipients with Lentiviral Vector-Transduced Hematopoietic Stem Cells. To confirm transduction of long-term repopulating stem cells, bone marrow was harvested from primary recipients, and 5 × 106 cells were transplanted into each of three lethally irradiated CD45.1 secondary recipients. Peripheral blood from the secondary recipients was analyzed for CD45.2 donor and GFP+ marking by flow cytometry at 8 and 16 weeks post-transplantation. By 16 weeks, the mean engraftment level of CD45.2+ cells in the peripheral blood of secondary recipients was not significantly different between cohorts (p < 0.05). GFP marking was stable over the 4-month period and similar to the GFP-marking level observed for EFDIG-transduced primary marrow transplant recipients (Fig. 4A). By comparison, engraftment of CD45.2+ cells in DL2G-marrow recipients was lower, probably due to the fact that 1 million fewer cells were transplanted into primary recipients. GFP marking in secondary DL2G-marrow recipients was substantially lower (1% GFP+), likely due to reduced stem cell marking (Fig. 4B). The level of GFP expression persisted in secondary recipients, as revealed by stable mean fluorescence intensity over time observed in secondary recipients of EFDIG-, DL2G-, and CSIIEG-transduced marrow. These results allow us to conclude that: 1) EFDIG and DL2G LV treatment of unfractionated marrow supported the transduction of hematopoietic stem cells; 2) primitive hematopoietic cells engrafted, repopulated, and gave rise to MTX-resistant progeny in primary recipients sufficient to provide significant chemoprotection and prolonged animal survival during MTX dose escalation; and 3) genetically marked hematopoietic stem cells reconstituted the bone marrow of lethally irradiated secondary recipients, giving rise to progeny that persisted in the peripheral blood.
Previous studies have shown that transplantation with drug-resistant DHFR transgenic marrow or with marrow transduced using a γ-retrovirus encoding human MTXr-DHFR conferred significant chemoprotection to transplant recipients (May et al., 1995, 1996; James et al., 1999; Budak-Alpdogan et al., 2004). The primary goal of the studies described in this article was to test the effectiveness of lentivirus vectors for Tyr22DHFR gene transfer into murine bone marrow and protection of transplant recipients from MTX toxicity. We constructed both a Tyr22DHFR-IRES-eGFP bicistronic vector and a Tyr22DHFR-eGFP genetic fusion for comparative in vivo chemoprotection studies, with the rationale that the eGFP coding sequence located 3′ to the IRES may not always be expressed to the same degree as the upstream Tyr22DHFR (Mizuguchi et al., 2000; Yu et al., 2003). In two separate in vivo chemoprotection studies, EFDIG- and DL2G-transduced marrow recipients recovered to healthy hematocrit levels, and they had a significant survival advantage over control GFP-transduced marrow recipients during MTX administration (p < 0.05). Animals that received GFP-transduced bone marrow exhibited severe myelosuppression (hematocrit < 20 and marrow hypoplasia upon histopathological analysis), and they did not survive MTX dose escalation to 1 mg/kg/day. We conclude that both EFDIG and DL2G lentiviral vectors confer a substantial chemoprotective effect in mouse marrow transplant recipients administered MTX.
To facilitate vector titering and the tracking of DHFR-transduced hematopoietic cells, we included a GFP coding sequence in both of the vectors tested in our study. For vector EFDIG, expression of the downstream eGFP sequence relies on internal translation from a picornavirus IRES sequence positioned between the DHFR and eGFP coding sequence. Although there is ample evidence in the literature for the effectiveness of this strategy for expression of two genes from the same vector, maintenance of expression can be problematic, particularly for the downstream coding sequence. Persons et al. (1997) solved this problem for DHFR-expressing Moloney murine leukemia virus vector by placing the GFP coding sequence in the upstream position of a GFP-IRES-DHFR transcript, thereby forcing expression of both genes due to the selective pressure exerted on DHFR in the downstream position. In this study, we generated a Tyr22DHFR-eGFP fusion protein so as to more intimately link GFP expression to drug-resistant DHFR function. Similar use of a human Phe22/Ser31 DHFR-GFP fusion protein was previously reported by Mayer-Kuckuk et al. (2002). We verified the predicted size of the DHFR-GFP fusion protein by Western blot using both anti-GFP and anti-DHFR antibodies, as well as maintained GFP fluorescence and DHFR enzyme activity in cell extracts. Direct comparison of cells transduced with EF-1α-regulated vectors (EFDIG versus CSIIEG; Table 1) revealed a 1.8-fold reduction in GFP mean fluorescence intensity resulting from the DHFR-GFP fusion (175 versus 322, respectively). However, GFP fluorescence of the IRES regulated construct EFDIG was reduced 3.3-fold (MFI = 98) compared with CSIIEG, thus exemplifying our motivation in seeking a more reliable strategy for GFP coexpression in these vectors. The effectiveness of the fusion design was also apparent in the maintenance of MFI in DL2G transduced peripheral blood cells of secondary transplant recipients (Fig. 4B), the low level of stem cell transduction in this experiment notwithstanding.
Therapeutic application of antifolates such as MTX is limited by systemic toxicity and tumor-acquired drug resistance (Bertino, 1993). Low- and intermediate-dose MTX therapy causes cytopenia among hematopoietic lineages, thereby compromising overall patient health. The primary acute toxicities associated with high-dose MTX treatment are myelosuppression and mucositis (Moe and Holen, 2000). This is especially limiting in applications where GI sensitivity to MTX precludes dose escalation to a level that would improve the likelihood for patient relapse-free survival. As an approach to protect against antifolate toxicities, ex vivo γ-retroviral transduction studies in mice have shown that virus-mediated MTXr-DHFR gene transfer supports higher dose antifolate administration while limiting myelotoxicity. Protection of animals from lethal doses of methotrexate by ex vivo retroviral vector-mediated transduction of hematopoietic stem cells was demonstrated by Williams (Corey et al., 1990) and also by Bertino (Capiaux et al., 2003). Our laboratory reported similar protection against lethal doses of MTX in animals transplanted with transgenic marrow expressing one of several MTX-resistant DHFR variants (May et al., 1995, 1996; Morris et al., 1996; James et al., 1997, 2000). Sorrentino's group demonstrated selective protection against the antifolate trimetrexate mediated by Tyr22DHFR transduction at the stem cell level (Allay et al., 1998). Here, we have extended these chemoprotection studies to include a lentiviral vector-mediated approach with the intention of providing a more potent means of achieving Tyr22DHFR gene transfer into hematopoietic targets.
There are several hematological observations that were made as a part of this study. The evidence from our study supports transduction of hematopoietic stem cells. GFP+ donor leukocytes of both myeloid and lymphoid lineages observed at 8 weeks post-transplant could have resulted from transduction of stem cells or of more committed myeloid and lymphoid progenitors. Drug resistance gene expression in committed progenitors is important, in fact, to confer chemoprotection in the period soon after bone marrow transplant (in mice, the first 2–3 weeks) (James et al., 1999). However, the data from our secondary transplant recipients indicate that gene transfer into primitive stem cells was achieved in the primary-transplanted animals, most probably giving rise to at least some transduced cells in the peripheral blood starting 3 to 4 weeks post-transplant. We also observed that transgene expression based on GFP marking in the peripheral blood was significantly increased during MTX treatment and then subsided 1 month after drug withdrawal (p < 0.05). In contrast, GFP marking did not decrease significantly in PBS-treated animals after cessation of injections (p > 0.05). This is consistent with a survival advantage imparted on more differentiated cells during MTX administration, but not on more primitive stem cells. This is also consistent with previous reports that selection of DHFR-transduced stem cells requires the use of a more specific DHFR inhibitor (i.e., trimetrexate) in combination with a nucleoside transport inhibitor (Allay et al., 1997). Temporary expansion of more differentiated cell types has also been reported for nonhuman primates transplanted with DHFR-transduced marrow upon administration of trimetrexate (Persons et al., 2004). Such temporary expansion of DHFR+ peripheral blood lymphocytes nonetheless implies a certain degree of chemoprotection achieved, with potential therapeutic benefit. Finally, gene transfer was achieved after a simple overnight exposure of marrow cells to DHFR lentivirus vector at an infection multiplicity of 10, resulting in a moderate overall proviral copy number in peripheral blood cells (0.5 copies/cell in expt 1 and 0.2–0.3 copies/cell in expt 2). Nonetheless, the low gene transfer frequency into stem cells observed in these experiments (11% in expt 1 and 1% in expt 2) was sufficient to protect transplant recipients from myelosuppression and to support MTX dose escalation to 2 and 1 mg/kg per day, respectively. Such chemoprotection was previously observed in mice engrafted at low levels (down to 1%) with transgenic marrow expressing drug-resistant DHFR (James et al., 2000).
Clinical testing of drug-resistance genes has included introduction of the MDR1 gene as a means of protecting against the toxicity of paclitaxel chemotherapy (Abonour et al., 2000, and references therein) and introduction of the methylguanine methyltransferase gene to confer resistance to alkylating agents (Cornetta et al., 2006). These studies have targeted human CD34+ hematopoietic stem cells using retroviral vectors for introduction of the MDR1 or O6-methylguanine-DNA methyltransferase genes. In particular, Abonour et al. (2000) reported successful MDR gene transfer into CD34+ cells engrafted in recipient patients, with a transient increase in MDR1-positive cells observed in the peripheral blood after a course of paclitaxel chemotherapy. These results imply protection against the toxicity of chemotherapy brought about by introduction of the MDR gene. We suggest that effective protection against antifolate toxicity may similarly be brought about by introduction of DHFR genes that encode antifolate-resistant enzyme activity, and at the higher frequency afforded by the use of lentivirus rather than murine retrovirus vectors. The studies described herein show that ex vivo lentiviral transduction of mouse bone marrow supports DHFR gene transfer into HSCs and more differentiated hematopoietic lineages, providing short- and long-term chemoprotection to marrow transplant recipients during MTX dose escalation. These studies also demonstrate the effectiveness of lentivirus vectors for this purpose, using ex vivo transduction conditions (i.e., short-term exposure) that are relevant to the clinical setting. Systemic chemoprotection coupled with antifolate chemotherapy may thus provide more effective treatment of antifolate-responsive tumors.
This work was supported by National Institutes of Health Grant CA060803 (to R.S.M.). This work was presented at the 9th Annual Meeting of the American Society of Gene Therapy; 2006 May 31–Jun 4 (Gori et al., 2006a); and the 35th Annual Meeting of the International Society for Experimental Hematology; 2006 Sept 27–30 (Gori et al., 2006b).
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
ABBREVIATIONS: MTX, methotrexate; DHFR, dihydrofolate reductase; MTXr, methotrexate-resistant; HSC, hematopoietic stem cell; IRES/ires, internal ribosomal entry site; GFP, green fluorescent protein; PCR, polymerase chain reaction; eGFP, enhanced green fluorescent protein; DMEM, Dulbecco's modified Eagle's medium; P/S, 100 U/ml penicillin, 100 μg/ml streptomycin; FBS, fetal bovine serum; TK–, lacking thymidine kinase; LV, lentivirus vector; Rx, treatment; BMT, bone marrow transplant; PBS, phosphate-buffered saline; IL, interleukin; SCF, stem cell factor; cGy, centiGray; MFI, mean fluorescence intensity; GI, gastrointestinal; PBMC, peripheral blood mononuclear cell; expt, experiment; MDR, multidrug resistance; CFU, colony-forming unit.
- Received June 7, 2007.
- Accepted June 21, 2007.
- The American Society for Pharmacology and Experimental Therapeutics