Retroviral Transduction and Engraftment Ability of Primate ... - Cell Press

ARTICLE

doi:10.1006/mthe.2002.0544, available online at http://www.idealibrary.com on IDEAL

Retroviral Transduction and Engraftment Ability of Primate Hematopoietic Progenitor and Stem Cells Transduced Under Serum-Free versus Serum-Containing Conditions Kimberley A. Kluge,* Aylin C. Bonifacino,* Stephanie Sellers, Brian A. Agricola, Robert E. Donahue, and Cynthia E. Dunbar† Hematology Branch, NHLBI, NIH, Bethesda, Maryland 20892, USA *These authors contributed equally to this work. †

To whom correspondence and reprint requests should be addressed. Fax: (301) 496-8396. E-mail: [email protected].

The ability to efficiently transduce hematopoietic stem and progenitor cells under serum-free conditions would be desirable for safety and standardization of clinical gene therapy protocols. Using rhesus macaques, we studied the transduction efficiency and engraftment ability of CD34enriched SCF/G-CSF mobilized progenitor cells (PBSC) transduced with standard amphotropic marking vectors under serum-free and serum-containing conditions. Supernatants were collected from producer cells 16 hours after serum-free medium or medium containing 10% fetal calf serum was added. Vector titers were approximately two- to threefold higher when producer cells were cultured in serum-containing medium. However, retroviral transduction of rhesus CFU-GM was improved using serum-free vector-containing medium. For analysis of engraftment with transduced cells, three macaques had CD34+ peripheral blood stem cells split into two fractions for transduction. One fraction was transduced using serum-free vector-containing medium, and the other fraction was transduced using standard serum-containing medium. The two fractions were re-infused simultaneously following total body irradiation. In all three animals, there was equivalent marking from both vectors for 7–9 months post-transplantation. These data are encouraging regarding the removal of serum-containing medium from clinical hematopoietic cell transduction protocols, given the lack of a detrimental effect on transduction and engraftment with transduced cells.

INTRODUCTION Hematopoietic stem cells have been a primary focus for gene therapy applications since the initial development of helper-free retroviral vector producer cell lines in the 1980s. Initial progress was slow despite encouraging murine studies, with clinical trials demonstrating levels of gene transfer far below clinically relevant thresholds [1]. However, over the past 3 years the field has been revitalized after studies using non-human primate and human/murine xenograft models began to report significant improvements in transduction of primitive repopulating cells with standard retroviral vectors. The addition of the cytokine FLT3 ligand (FLT3L) during transduction, along with the use of either a fibronectin fragment (retronectin) or autologous stromal cells to support primitive hematopoietic elements ex vivo, resulted in long-term in vivo marking levels of 5–30% in hematopoietic progeny

316

cells [2–5]. This progress has recently culminated in the first convincing evidence for efficacious clinical gene therapy in children with X-linked severe combined immunodeficiency [6]. These results were encouraging and have led investigators to once again consider clinical trials of gene therapy directed at hematopoietic stem cells (HSC). Several important questions remain to be answered, however, before more widespread clinical applications are developed. Initial preclinical studies of HSC gene therapy used vector collected from producer cell lines grown in standard fetal calf serum-containing medium. Beginning in the mid 1990s, two developments fueled interest in performing transductions under serum-free conditions. First, reports of bovine spongiform encephalitis virus transmission from cattle to humans resulted in serious concerns regarding the exposure of humans to any unnecessary or unregulated

MOLECULAR THERAPY Vol. 5, No. 3, March 2002 Copyright © The American Society of Gene Therapy 1525-0016/02 $35.00


ARTICLE

FIG. 1. RNA slot blot on retroviral vector producer cell line supernatants. Identical flasks of G1Na and LNL6 amphotropic vector producer clones were grown for 16 hours under standard serum-containing conditions (+serum) versus serum-free conditions (–serum). For each vector, supernatants were collected from three independent flasks for each condition. A neo probe was used for hybridization. The P values comparing +serum versus –serum viral RNA production was 0.0005 for LNL6 and 0.002 for G1Na.

bovine products, including fetal calf serum [7]. Second, several investigators reported that ex vivo expansion of primitive human hematopoietic cells assayed in vitro was superior under serum-free conditions [8,9]. This stimulated the commercial development of specific serum-free medium for the maintenance of primitive hematopoietic cells. Few studies have directly compared either the transduction efficiency or the engraftment potential of HSC cultured and transduced under serum-free versus serumcontaining conditions. Using an early serum-free medium formulation, we reported significantly impaired engraftment of murine HSC cultured for 4 days under serum-free conditions when directly compared with HSC cultured in the presence of serum, despite equivalent maintenance of less primitive CFU or CFU-S [10]. More recently, investigators have reported retention or even expansion of the number of primitive human cord blood cells able to engraft immunodeficient mice, when cultured in serumfree medium [11]. Transduction of committed progenitors as well as more primitive NOD/SCID repopulating cells under serum-free conditions has been reported [12,13]. However, the encouraging non-human primate and clinical trials noted above used serum-containing medium for

culture and transduction of CD34+ cells [2,6,14,15]. An early clinical trial used serum-free conditions, but was performed before the advances in transduction conditions such as the addition of FLT3 ligand and retronectin support [16]. To carry out serum-free retroviral transductions, vectorcontaining medium must be collected from producer cells grown in serum-free medium. Producer cells are commonly murine fibroblasts that are serum-dependent for growth and proliferation. Success therefore requires adequate production of vector particles by producer cells grown at least transiently without serum, efficient transduction of target cells under serum-free conditions, and preservation of stem cell characteristics including multipotentiality and homing ability after ex vivo culture in the absence of serum. To completely prevent transmission of bovine infectious agents, producer lines would need to be adapted to long-term serum-free culture, or particle purification procedures devised, but this study represents a first step to explore whether serum-free transduction of longterm repopulating cells is feasible and of equivalent efficiency to standard transduction procedures. In this study we directly compare transduction and engraftment of non-human primate repopulating cells cultured and exposed to vector under serum-free versus serum-containing conditions. The rhesus macaque and other non-human primate autologous transplantation models are the best preclinical assays available to predict human clinical trial results, given the similarities in stem cell behavior, vector receptor levels, and cytokine responsiveness of primitive hematopoietic cells [17,18]. The competitive repopulation design allows meaningful data to be

FIG. 2. Transduction of committed progenitor cells. At the end of a 96-hour culture and transduction period, cells were plated at low density in a standard CFU assay. On days 10–12, at least 20 individual well-separated CFU were plucked for PCR analysis of transduction efficiency from each transduction. Each individual CFU was digested, and analyzed by PCR simultaneously for neo (vector) and control actin sequences. Colonies were scored as positive for vector only if both the neo and actin bands were positive, and concurrent reagent and negative control CFU reactions were without neo bands. The percentage of neo+ colonies is shown for each transduction.

MOLECULAR THERAPY Vol. 5, No. 3, March 2002 Copyright © The American Society of Gene Therapy

317

ARTICLE


TABLE 1: In vitro culture and transduction results Pre-Transduction Monkey

% CD34+

RQ2317

93%

RQ2315 RQ2230d

RQ2228

71%

58%

43.6%

Transduction

Post-Transduction

Transduction Efficiency

Conditionsa: vector usedb

% CD34+

Cell number

No. neo + CFU/ no. analyzed (%)c

2.9 ⫻ 107

Serum (-): G1Na

78%

4.4 ⫻ 107

71/81 (87.6%)

2.9 ⫻ 107

Serum (+): LNL6

76%

4.4 ⫻ 107

33/58 (57.0%)

1.25 ⫻ 107

Serum (-): LNL6

85%

1.8 ⫻ 107

37/41 (90.2%)

7

Cell number

7

1.25 ⫻ 10

Serum (+): G1Na

84%

2.0 ⫻ 10

68/80 (85.0%)

1.13 ⫻ 107

Serum (-): G1Na

70%

2.0 ⫻ 107

18/20 (90.0%)

1.13 ⫻ 107

Serum (+): LNL6

83%

1.9 ⫻ 107

15/20 (75.0%)

2.9 ⫻ 107

Serum (-): G1Na

57%

5.5 ⫻ 107

39/40 (97.5%)

2.9 ⫻ 107

Serum (+): LNL6

57%

2.2 ⫻ 107

39/40 (97.5%)

a

Transduction conditions: total of 96 hours, with daily exchange of vector supernatant and cytokines SCF 100 ng/ml, FLT3 ligand 100 ng/ml, MDGF 100 ng/ml (Amgen), on retronectin. Serum (-) conditions: X-vivo 10 serum-free media (BioWhittaker) used to produce vector supernatant. Serum (+) conditions: DMEM + 10% FCS used to produce vector supernatant. b LNL6 and G1Na are equivalent titer (7 ⫻ 106 inf/particles/ml) vectors carrying the neomycin resistance gene. c Individual CFU plucked from methylcellulose culture and PCR for vector (neo) and control actin performed. Number of neo + colonies/number of actin + colonies shown. d This animal died in an anesthesia mishap prior to reinfusion, thus only in vitro data are available.

obtained using a relatively small number of animals, as individual variability in PBSC mobilization or transduction efficiency is controlled for by directly comparing results of serum-free versus serum-containing conditions in each animal.

RESULTS The retroviral vector producer cell lines used for these studies had not previously been grown under serum-free conditions. Thus we first analyzed whether adequate vector particles could be obtained when producer cell lines were changed to culture in serum-free medium optimized for hematopoietic progenitor growth (Ex Vivo 10) 16 hours before collection of supernatant [19]. The amphotropic producer clones for the neo vectors G1Na and LNL6 were grown to subconfluence in standard serum-containing medium, then the flasks had medium changed to either fresh DMEM with 10% FCS or Ex Vivo 10. After 16 hours, the vector-containing medium was collected. Viral RNA was isolated and RNA slot blots were performed. Both LNL6 and G1Na producer clones produced consistently approximately 3-fold less viral RNA in serum-free medium, compared with standard serum-containing medium (Fig. 1; P = .0005 for LNL6 and P = .002 for G1Na). We also compared biologic titers by making serial dilutions of either the serum-free or serum-containing vector into standard D10 medium, and assaying for transfer of G418 resistance to HeLa cells. G1Na had a titer of 4 ⫻ 105 under standard serum-containing conditions, and it dropped to 2 ⫻ 105 under serum-free conditions. LNL6 likewise had a 50% drop in biologic titer on HeLa cells, from 6 ⫻ 105/ml to 3 ⫻ 105/ml. The relatively small

318

decrease in both particle number and biologic titer after collecting vector from producer cells grown in serum-free conditions was encouraging. Although gene transfer to committed progenitor cells such as CFU-GM is not always predictive of gene transfer to long-term in vivo repopulating cells, we next compared the ability of serum-containing versus serum-free vectorcontaining medium to transduce these hematopoietic target cells. SCF/G-CSF mobilized rhesus macaque peripheral blood CD34+ cells were transduced for 96 hours in the presence of serum-free or serum-containing vector, supported by SCF, MGDF, FLT3L, and retronectin. Transduction of committed hematopoietic progenitors was assessed by performing PCR for vector sequences on DNA extracted from individual CFU-GM plated at the end of the transduction period. The overall transduction efficiency of CFU-GM with either LNL6 or G1Na under serum-free or serum-containing conditions was very high (Table 1 and Fig. 2). The efficiency of transduction was actually significantly higher under serum-free conditions when the data were analyzed in aggregate using Fischer’s exact test [20] (Table 1 and Fig. 2). We then tested the transduction efficiency and engrafting ability of rhesus macaque repopulating cells in three animals, using a competitive repopulation approach that allows significant insights into stem cell behavior using practical numbers of these large animals [21,22]. Each animal had SCF/G-CSF mobilized PBSC collected and CD34enriched. These target cells were split into two equal fractions, and each fraction was transduced with serum-free or serum-containing LNL6 or G1Na vector supernatant. After transplantation, PCR analysis of peripheral blood granulocytes and mononuclear cells allowed quantitation of the



ARTICLE

FIG. 3. Representative PCR analysis for vector sequences in peripheral blood cells post-transplantation. We carried out semiquantitative PCR for the neo gene in post-transplantation blood samples from two animals receiving G1Na and LNL6 transduced CD34-enriched SCF/G-CSF mobilized progenitor cells (PBSC). In each animal, one vector transduction was done under serum-free conditions and the other under serum-containing conditions. PCR products were separated on a gel to distinguish amplified G1Na versus LNL6 proviral sequences, due to a 16-bp polylinker insertion within the amplified region of G1Na. PCR for actin was performed to control for amplifiable DNA content. Standards consist of indicated percentages of dilutions of single copy vector DNA into control rhesus DNA. M, PB mononuclear cells; G, granulocytes; (-) control, normal rhesus PB; No DNA, reagent control; Day, day post-transplantation sample was obtained.

relative contribution of primitive engrafting cells transduced under serum-free versus serum-containing conditions. Representative PCR analyses are shown (Fig. 3), as well as the levels in granulocytes over time in the three animals at engraftment and at the most recent follow-up (Fig. 4). In all three animals, there were no significant differences between in vivo marking derived from cells transduced under serum-free versus serum-containing transduction conditions. Similar results were obtained in mononuclear cells (data not shown).

DISCUSSION The use of animal serum in cell culture, especially for clinical applications, has potential disadvantages. There is variability between lots of animal sera in their ability to support hematopoietic progenitors, requiring careful screening of new lots of sera. Although this may also be the case in components such as human albumin used to manufacture serum-free medium, generally this variability is less than between lots of animal serum containing many positive and negative predominantly undefined substances. Thus most hematopoietic cell research is now carried out using serum-free medium optimized to support proliferation or maintenance of hematopoietic progenitor cells. Exposure of clinical cell products to fetal bovine serum has concerned investigators and regulatory agencies for many years, with heightened sensitivity regarding pathogens such as the bovine spongiform encephalitis agent potentially transmitted by exposure to cattle tissues


accelerating the pressure to remove fetal calf serum from all clinical-grade hematopoietic cultures [7]. Methodology for culturing primitive hematopoietic cells ex vivo, either in suspension culture, co-culture on a stromal support layer, or in semi-solid medium, was originally developed relying on the inclusion of at least 10% animal sera, generally fetal calf serum alone or in combination with horse serum [23]. The components in serum necessary for successful maintenance of hematopoietic cells in culture were initially unknown. Beginning in the 1970s, individual substances required began to be identified, and the first serum-free CFU-GM cultures were reported [24]. More recently, many investigators have reported that primitive human hematopoietic cells can be maintained or expanded in optimized serum-free medium [23,25–28]. However, it is clear that CFU assays or enumeration of total CD34+ cells do not adequately measure or reflect the impact of culture on more primitive repopulating stem cells, and few comparative studies exist directly comparing in vivo engraftment ability of cells cultured short- or long-term under serum-free versus serumcontaining conditions. Our group previously used the murine competitive repopulation model to quantitate long-term repopulating stem cell activity, and found that murine marrow cells cultured for 4 days under serum-free conditions in the presence of IL3, IL6, and SCF had significantly less functional stem cell activity than cells cultured with the same cytokines in the presence of serum, but the serum-free medium used was not optimized for culture of murine hematopoietic cells, and therefore it is

319

ARTICLE


FIG. 4. Summary of vector copy numbers in peripheral blood cells from three animals receiving cells transduced under serum-containing versus serum-free conditions. The percentage marking in granulocytes is shown, assuming one vector copy per cell. All time points analyzed following transplantation are shown, with the percentage marking derived from CD34+ cells transduced under serumcontaining versus serum-free conditions delineated. Marking in peripheral blood mononuclear cells was very similar (data not shown).

unclear how these results translate to a large animal or human setting [10]. The avoidance of bovine serum has recently been addressed for clinical gene therapy applications. This modification is more complex than simply optimizing culture of the hematopoietic target cells in serum-free medium, because viral vectors used for transduction are produced by murine, canine, or human fibroblast cell lines that are serum-dependent for proliferation. Investigators have attempted to adapt these producer cell lines to sustained growth in low-serum or serum-free conditions, but this has proven difficult. As a first step it has been possible to adapt retroviral vector producer clones to short-term culture in serum-free medium, allowing collection of virtually serum-free vector-containing medium after vigorous washing of subconfluent producer cells and addition of fresh serum-free medium that can be harvested 12–24 hours later [19]. As we confirmed in our study, retroviral vector titer has been reported to be similar under serumfree versus serum-containing conditions [29,30]. A less than one log change in vector titer has not been reported to have an impact on gene transfer efficiency to large animal long-term repopulating stem cells [31]. Another approach has been to collect standard serum-containing vector, and then use concentrated and purified vector preparations that are 5–6 logs depleted of protein components to add back to serum-free cultures of hematopoietic cells [32]. Either approach has allowed equivalent transduction of human CFU or LTCIC as directly compared with standard serum-containing vector preparations [12,29]. All but one clinical trial published so far describing hematopoietic stem cell gene transfer used serum-containing vector culture medium. Malech and coworkers transferred producer cells to serum-free conditions 8 hours before harvesting vector. No long-term in vivo correction

320

was obtained, but this may have resulted from the lack of myeloablation and the use of now-obsolete transduction conditions, not the use of serum-free vector-containing medium [16]. Our current results in the non-human primate autologous transplantation model are encouraging. The overall in vivo marking levels were equivalent for cells transduced under serum-free versus serum-containing conditions. We conclude that gene transfer to primitive repopulating cells can occur under these conditions, and that transduced cells maintained repopulating ability. It is theoretically possible that transduction was less efficient in serum-free medium, but that maintenance of functional stem cells was greater, resulting in the same in vivo marking levels, vecause the competitive repopulation model cannot distinguish between transduction efficiency of stem cells and the engraftment efficiency of the transduced cells. The non-human primate model has been accurate for prediction of results in human hematopoietic stem cell gene transfer protocols, and thus our findings suggest that serum-free transductions of target human CD34+ PBSC is feasible and desirable.

MATERIALS

AND

METHODS

Rhesus autologous PBSC collection, transduction, and transplantation. Young rhesus macaques (Macaca mulatta) were housed and handled in accordance with the guidelines set by the Committee on Care and Use of Laboratory Animals. The animals were mobilized with recombinant human (rhu) SCF (200 ␮g/kg; Amgen, Thousand Oaks, CA) and rhu G-CSF (50 ␮g/kg; Amgen) given as daily subcutaneous injections for 5 days. The mobilized PB cells were collected by apheresis on day 5 [33]. The mononuclear cells were isolated by density gradient centrifugation over Ficoll-Paque Plus (Amersham Pharmacia Biotech, Uppsala, Sweden). CD34 enrichment was performed using the 12.8 anti-CD34 antibody and streptavidin-coated MACS beads as directed by the manufacturer (Miltenyi, Auburn, CA) [34]. The retroviral vectors G1Na and LNL6 contain an identical bacterial neomycin phosphotransferase (neo) gene [35]. There is a 16-bp difference



5⬘ of the neo gene that allows for simple discrimination of the two vectors by PCR or Southern blot [36]. To prepare fresh vector-containing medium for transductions, the producer cell lines were grown to 70% confluency at 37⬚C in 5% CO2, and 16–18 hours before supernatant collection, standard serum-containing medium was replaced with either fresh DMEM (Biofluids, Rockville, MD) containing 10% fetal calf serum (Hyclone, Logen, UT) or serum-free XVivo 10 (BioWhittaker, Walkersville, MD), both supplemented with 200 mM L-glutamine and 50 ␮g/ml gentamycin. Fresh vector-containing medium was collected 16 hours later and passed through a 0.22-␮m filter (Millipore, Bedford, MA) to remove cellular debris. For each animal used for the in vivo engraftment studies, CD34enriched PBSCs were divided into two equal fractions and each fraction was transduced with one of the two vectors. The vectors were reversed in each subsequent animal to eliminate possible differences in transduction efficiency due to inherent properties of the two vectors or their producer cell lines. All transductions were performed on Retronectin-coated plates (Takara, Otsu, Japan), as directed by the manufacturer. CD34+ cells were transduced for a total of 96 hours at a starting concentration of 1 to 3 ⫻ 105 cells/ml in undiluted vector-containing medium supplemented with stem cell factor (SCF; Amgen, Thousand Oaks, CA), FLT3 ligand (FLT3L; Immunex, Seattle, WA), and megakaryocyte growth and development factor (MDGF; Amgen), all at concentrations of 100 ng/ml, with daily replacement of vector-containing medium and cytokines. At the end of transduction, cells were collected via vigorous washing and a brief exposure to 0.25% trypsin, counted, and samples were aliquoted for CFU analysis. Each animal received total body irradiation (TBI) with two fractions of 500 rads. The two transduced fractions from each animal were pooled and infused fresh via a central venous catheter 1 day following TBI. Twentyfour hours later, the animals were started on G-CSF at 5 ␮g/kg/day intravenously daily until the total leukocyte count reached 6000/␮l. Standard supportive care was administered until engraftment [37]. Analysis of viral titer. We precipitated 1 ml of each supernatant with 30% polyethylene glycol 8000 (Fisher Scientific, Fair Lawn, NJ) and the pellets were resuspended in 200 ␮l Tris-EDTA buffer (pH 7.4) containing 10% (wt/vol) vanadyl ribonuclease complex (Gibco BRL, Gaithersburg, MD) and 100 ␮g/ml yeast tRNA (Gibco BRL), then lysed by adding 200 ␮l of 2⫻ lysis buffer (1% SDS, 0.6 M NaCl, 20 mM EDTA, and 20 mM Tris-HCL, pH 7.4). Retroviral RNA was extracted from the lysed solution by phenol/chloroform extraction. The RNA was reconstituted with 500 ␮l of a 7.5% formaldehyde solution containing 1.5 M NaCl and 150 mM sodium citrate, pH 7.0. We loaded 400 ␮l and 100 ␮l of each sample onto a Minifold II (Schleicher & Schuell, Keene, NH) apparatus with vacuum applied. After drying, the transfer membrane (Hybond N+, Amersham, Cleveland, OH) was hybridized with a neo probe created by PCR using the following primers: 5⬘ATGATTGAACAAGATGGATTGCA-3⬘ and 5⬘-AGGCATCGCCATGGGTCACGACGAGAT-3⬘. Biologic titering was performed using HeLa cells. Serum-free or serumcontaining vector culture medium was harvested from G1Na or LNL6 producer cells as described above, and serially diluted into fresh D10. Subconfluent HeLa cells were cultured overnight in the medium in the presence of 4 mg/ml protamine, and then split into fresh medium 24 hours later, in the presence of 0.5 mg/ml active G418 (Gibco). Medium was replaced every 72–96 hours, and 12–14 days later, the cells were fixed in 30% formalin and stained with crystal violet for macroscopic colony enumeration. Analysis of gene transfer efficiency. Peripheral blood (PB) samples were obtained immediately following recovery of the neutrophil count to greater than 1000/␮l, and then monthly. PB mononuclear cell and granulocyte fractions were obtained by density gradient centrifugation over Ficoll-Paque Plus (Amersham Pharmacia Biotech), resulting in greater than 95% purity of mononuclear cells and granulocytes. DNA was purified using the Puregene Genomic DNA Isolation kit as per the manufacturer’s instructions (Gentra Systems, Minneapolis, MN). DNA from individual CFU was prepared for PCR by digestion at 55⬚C for at least 1 hour after addition of 1 ␮l of proteinase K (20 mg/ml) in a total volume of 50 ␮l of distilled H2O, followed by inactivation of the proteinase K by heating to 99⬚C for 10 minutes.


ARTICLE

For PCR analysis, 100 ng purified DNA from blood samples or 10 ml digested CFU DNA was added to a master mix containing Buffer II, 2.5 mM MgCl2, dNTPs, and Amplitaq Gold polymerase (Applied Biosystems, Foster City, CA) and then divided equally between two tubes, one containing the primer pair for the neo gene, and one containing the primer pair for the amplification of a control ␤-actin gene, and these reactions were carried out as described [36]. The ␤-actin and neo PCR amplification reactions were performed in the presence of 32P-labeled dCTP (Amersham, Arlington Heights, IL). All samples were run with concurrently extracted normal rhesus peripheral blood DNA and reagents alone as negative controls. Positive controls consisted of log dilutions of DNA extracted from cell lines containing known copy numbers of vector sequences diluted into normal rhesus genomic DNA for both the neo and ␤-actin PCRs. The final PCR products were separated on 8% polyacrylamide gels. The band sizes were 483 bp for G1Na neo and 467 bp for LNL6 neo, and 232 bp for ␤-actin. All reactions were optimized to yield linear amplifications in the range of the intensity of the positive samples, and neo marking of the samples was quantitated by plotting the ratio of neo signal to ␤-actin signal intensity derived from PhosphorImager analysis (Molecular Dynamics, Sunnyvale, CA) on a standard curve derived from known copy number controls amplified concurrently. All quantitations on experimental samples were performed in the linear range of the standard curve. CFU assays. Total cell number and the frequency of colony-forming units (CFU-C) were determined for the mononuclear cells in the apheresis product, for the CD34-enriched cells, and at the end of the 96-hour transduction period for both fractions. CFU assays were performed using methylcellulose medium (StemCell Technologies, Vancouver, British Columbia, Canada) supplemented with 5 U/ml recombinant human erythropoietin (Amgen), 10 ng/mL IL-3 (Sandoz, East Hanover, NJ), 10 ng/ml granulocytemacrophage colony-stimulating factor (GM-CSF; Sandoz), and 100 ng/ml SCF (Amgen). At day 10, colonies were counted, and 20 well-separated individual colonies from each fraction were plucked into 50 ␮l of distilled water, digested with 20 mg/ml proteinase K at 55⬚C for 1 hour followed by 99⬚C for 10 minutes, and assessed for vector neo sequences by PCR, as described above. Simultaneous PCR for ␤-actin sequences was performed on each colony, and the percentage of transduction was calculated by dividing the number of CFU positive for the neo gene by the number of CFU positive for ␤-actin. Statistical analysis. The Student’s t-test was used to analyze the amount of viral RNA, and a paired Student’s t-test was used to compare in vivo marking levels of cells transduced under serum-containing versus serumfree conditions. Fischer’s exact test was used to compare the CFU transduction efficiency as described [20].

ACKNOWLEDGMENTS We thank Mark Metzger, Earl West, Barrington Thompson, and the LAMS and ROW staff for their fine care of the animals over the past decade. RECEIVED FOR PUBLICATION AUGUST 30, 2001; ACCEPTED JANUARY 8, 2002.

REFERENCES 1. Dunbar, C. E. (1996). Gene transfer to hematopoietic stem cells: implications for gene therapy of human disease. Annu. Rev. Med. 47: 11–20. 2. Kiem, H. P., et al. (1998). Improved gene transfer into baboon marrow repopulating cells using recombinant human fibronectin fragment CH-296 in combination with interleukin-6, stem cell factor, FLT-3 ligand, and megakaryocyte growth and development factor. Blood 92: 1878–1886. 3. Tisdale, J. F., et al. (1998). Ex vivo expansion of genetically marked rhesus peripheral blood progenitor cells results in diminished long-term repopulating ability. Blood 92:1131–1141. 4. Dao, M. A., Taylor, N., and Nolta, J. A. (1998). Reduction in levels of the cyclin-dependent kinase inhibitor p27kip-1 coupled with transforming growth factor b neutralization induces cell-cycle entry and increases retroviral transduction of primitive human hematopoietic cells. Proc. Natl. Acad. Sci. USA 95: 13006–13011 5. Conneally, E., Eaves, C. J., and Humphries, R. K. (1998). Efficient retroviral-mediated gene transfer to human cord blood stem cells with in vivo repopulating potential. Blood 91: 3487–3493. 6. Cavazzana-Calvo, M., et al. (2000). Gene therapy of human severe combined immun-

321

ARTICLE


odeficiency (SCID)-X1 disease. Science 288: 669–672. 7. Asher, D. M. (1999). The transmissible spongiform encephalopathy agents: concerns and responses of United States regulatory agencies in maintaining the safety of biologics. Dev. Biol. Stand. 100: 103–118. 8. Lebkowski, J., et al. (1993). Serum free culture of purified CD34+ cells yields large expansions of both myeloid and erythroid progenitors. Exp. Hematol. 21: 1021. 9. Petzer, A. L., Hogge, D. E., Lansdorp, P. M., Reid, D. S., and Eaves, C. J. (1996). Selfrenewal of primitive human hematopoietic cells (long-term-culture-initiating cells) in vitro and their expansion in defined medium. Proc. Natl. Acad. Sci. USA 93: 1470–1474. 10. Sekhar, M., Soma, T., and Dunbar, C. E. (1997). Murine long-term repopulating ability is compromised by ex vivo culture in serum-free media despite the preservation of committed progenitors. J. Hematother. 6: 543–549. 11. Conneally, E., Cashman, J., Petzer, A., and Eaves, C. (1997). Expansion in vitro of transplantable human cord blood stem cells demonstrated using a quantitative assay of their lympho-myeloid repopulating activity in nonobese diabetic-scid/scid mice. Proc. Natl. Acad. Sci. USA 94: 9836–9841. 12. Sekhar, M., et al. (1996). Retroviral transduction of CD34-enriched hematopoietic progenitor cells under serum-free conditions. Hum. Gene Ther. 7: 33–38. 13. Schiltz, A. J., et al. (1998). High efficiency gene transfer to human hematopoietic SCIDrepopulating cells under serum free conditions. Blood 92: 3163–3171. 14. Abonour, R., et al. (2000). Efficient retrovirus-mediated transfer of the multidrug resistance 1 gene into autologous human long-term repopulating hematopoietic stem cells. Nat. Med. 6: 652–658. 15. Wu, T., et al. (2000). Prolonged high-level detection of retrovirally-marked hematopoietic cells in non-human primates after transduction of CD34+ progenitors using clinically feasible methods. Mol. Ther. 1: 285–293. 16. Malech, H. L., et al. (1997). Prolonged production of NADPH oxidase-corrected granulocytes after gene therapy of chronic granulomatous disease. Proc. Natl. Acad. Sci. USA 94: 12133–12138. 17. Van Beusechem, V. W., and Valerio, D. (1996). Gene transfer into hematopoietic stem cells of nonhuman primates. Hum. Gene Ther. 7: 1649–1668. 18. Donahue, R. E., and Dunbar, C. E. (2001). An update on the use of non-human primate models for preclinical testing of gene therapy approaches targeting hematopoietic cells. Hum. Gene Ther. 12: 607–617. 19. Seppen, J., Kimmel, R. J., and Osborne, W. R. (1997). Serum-free production, concentration and purification of recombinant retroviruses. Biotechniques 23: 788–790. 20. Hedges, L. V., and Olkin, I. (1985). Statistical Methods for Meta-Analysis, Academic Press, San Diego. 21. Dunbar, C. E., et al. (1996). Improved retroviral gene transfer into murine and rhesus peripheral blood or bone marrow repopulating cells primed in vivo with stem cell factor and granulocyte colony-stimulating factor. Proc. Natl. Acad. Sci. USA 93: 11871–11876.

322

22. Kiem, H. P., et al. (1997). Gene transfer into marrow repopulating cells: comparison between amphotropic and gibbon ape leukemia virus pseudotyped retroviral vectors in a competitive repopulation assay in baboons. Blood 90: 4638–4645. 23. Lebkowski, J. S., et al. (1994). Rapid isolation and serum-free expansion of human CD34+ cells. Blood Cells 20: 404–410. 24. Guilbert, L. J., and Iscove, N. N. (1976). Partial replacement of serum by selenite, transferrin, albumin and lecithin in haemopoietic cell cultures. Nature 263: 594–595. 25. Paquette, R. L., et al. (1998). Ex vivo expansion and differentiation of unselected peripheral blood progenitor cells in serum-free media. J. Hematother. 7: 481–491. 26. Poloni, A., et al. (1997). The ex vivo expansion capacity of normal human bone marrow cells is dependent on experimental conditions: role of the cell concentration, serum and CD34+ cell selection in stroma-free cultures. Hematol. Cell Ther. 39: 49–58. 27. Mayani, H., Dragowska, W., and Lansdorp, P. M. (1993). Characterization of functionally distinct subpopulations of CD34+ cord blood cells in serum-free long-term cultures supplemented with hematopoietic cytokines. Blood 82: 2664–2672. 28. Migliaccio, G., et al. (1991). Effects of recombinant human stem cell factor (SCF) on the growth of human progenitor cells in vitro. J. Cell Physiol. 148: 503–509. 29. Glimm, H., et al. (1998). Efficient serum-free retroviral gene transfer into primitive human hematopoietic progenitor cells by a defined, high-titer, nonconcentrated vector-containing medium. Hum. Gene Ther. 9: 771–778. 30. Schilz, A. J., et al. (2000). MDR1 gene expression in NOD/SCID repopulating cells after retroviral gene transfer under clinically relevant conditions. Mol. Ther. 2: 609–618. 31. Van Beusechem, V. W., et al. (1993). Retrovirus-mediated gene transfer into rhesus monkey hematopoietic stem cells: the effect of viral titers on transduction efficiency. Hum. Gene Ther. 4: 239–247. 32. Kotani, H., et al. (1994). Improved methods of retroviral vector transduction and production for gene therapy. Hum. Gene Ther. 5: 19–28. 33. Donahue, R. E., et al. (1996). Peripheral blood CD34+ cells differ from bone marrow CD34+ cells in Thy-1 expression and cell cycle status in nonhuman primates mobilized or non mobilized with granulocyte colony-stimulating factor and/or stem cell factor. Blood 87: 1644–1653. 34. Berenson, R. J., et al. (1991). Engraftment after infusion of CD34+ marrow cells in patients with breast cancer or neuroblastoma. Blood 77: 1717–1722. 35. Cassel, A., Cottler-Fox, M., Doren, S., and Dunbar, C. E. (1993). Retroviral-mediated gene transfer into CD34-enriched human peripheral blood stem cells. Exp. Hematol. 21: 585–591. 36. Dunbar, C. E., et al. (1995). Retrovirally-marked CD34-enriched peripheral blood and bone marrow and cells contribute to long-term engraftment after autologous transplantation. Blood 85: 3048–3057. 38. Nienhuis, A. W., et al. (1987). Recombinant human granulocyte-macrophage colonystimulating factor (GM-CSF) shortens the period of neutropenia after autologous bone marrow transplantation in a primate model. J. Clin. Invest. 80: 573–577.
