HUMAN GENE THERAPY 11:2515–2528 (December 10, 2000)Mary Ann Liebert, Inc.

Preincubation with Endothelial Cell Monolayers IncreasesGene Transfer Efficiency into Human Bone Marrow

CD341CD382 Progenitor Cells

JOHN P. CHUTE,1,2 ABHA SAINI,1 MARK WELLS,1 WILLIAM CLARK,1 ANDREA WU,1

DANIEL ST. LOUIS,1 PATRICK BLAIR,1 DAVID HARLAN,1,2 and SUMESH KAUSHAL1

ABSTRACT

Retroviral gene transfer studies targeting bone marrow CD34 1 CD38 2 stem cells have been disappointing be-cause of the rarity of these cells, their G0 cell cycle status, and their low or absent expression of surface retro-viral receptors. In this study, we examined whether preincubation of bone marrow CD34 1 CD382 stem cellswith a hematopoietically supportive porcine microvascular endothelial cell line (PMVECs) could impact thecell cycle status and expression of retroviral receptors in pluripotent CD34 1 CD38 2 cells and the efficiency ofgene transfer into these primitive target cells. PMVEC coculture supplemented with GM-CSF 1 IL-3 1 IL-6 1 SCF 1 Flt-3 ligand induced . 93% of the CD34 1 CD38 2 population to enter the G1 or G2/S/M phase whileincreasing this population from 1.4% on day 0 to 6.5% of the total population by day 5. Liquid cultures sup-plemented with the identical cytokines induced 73% of the CD34 1 CD38 2 population into cell cycle but didnot maintain cells with the CD34 1 CD382 phenotype over time. We found no significant increase in the lev-els of AmphoR or GaLVR mRNA in PMVEC-expanded CD34 1 CD38 2 cells after coculture. Despite this, theefficiency of gene transfer using either amphotropic vector (PA317) or GaLV vector (PG13) was significantlygreater in PMVEC-expanded CD34 1 CD382 cells (11.4 6 5.6 and 10.9 6 5.2%, respectively) than in eithersteady state bone marrow CD34 1 CD38 2 cells (0.6 6 1.7 and 0.2 6 0.6%, respectively; p , 0.01 and p , 0.01)or liquid culture-expanded CD34 1 CD38 2 cells (1.4 6 3.5 and 0.0%, respectively; p , 0.01 and p , 0.01). SincePMVEC coculture induces a high level of cell cycling in human bone marrow CD34 1 CD382 cells and expandshematopoietic cells capable of in vivo repopulation, this system offers potential advantages for application inclinical gene therapy protocols.

2515

OVERVIEW SUMMARY

Two requirements for successful retroviral gene therapyinto human hematopoietic stem cells are cell cycling in tar-get cells and maintenance of repopulating capacity. We havedeveloped an ex vivo culture system utilizing a porcine mi-crovascular endothelial cell line (PMVECs) that promotesthe expansion of primitive bone marrow CD34 1 CD382

progenitor cells while inducing the large majority of thesecells into cell division. Preincubation of bone marrowhematopoietic (CD34 1 CD382 ) progenitor cells for 5 dayson PMVEC monolayers followed by 3 days of transductionwith amphotropic vector or GaLV was associated with sig-nificantly increased efficiency of gene transfer into primi-

tive (CD34 1 CD38 2 ) target cells compared with nonex-panded (steady state) CD34 1 CD382 cells andCD34 1 CD38 2 cells expanded in liquid culture. Preincuba-tion with PMVEC monolayers therefore may offer advan-tages in clinical gene therapy models by supporting the gen-eration of genetically transduced CD34 1 CD38 2 cellscapable of in vivo repopulation.

INTRODUCTION

BONE MARROW CD341CD382 cells are highly enriched forpluripotent progenitor cells that account for long-term re-

population in vivo (Terstappen et al., 1991; Huang and Ter-

1NIDDK-Navy Transplantation and Autoimmunity Branch, Stem Cell Biology Laboratory, Naval Medical Research Institute, Bethesda, MD20889.

2Department of Medicine, Uniformed Services University for the Health Sciences, Bethesda, MD 20889.

stappen, 1994; Bhatia et al., 1997). The ex vivo expansion ofCD341CD382 stem cells has been extensively studied for ap-plication in gene therapy protocols (Von Kalle et al., 1994;Brugger et al., 1995; Reems and Torok-Storb, 1995; Shah etal., 1996; Zandstra et al., 1997; Tisdale et al., 1998). However,previous methods for expanding bone marrow hematopoieticprogenitor cells have been associated with the induction of anengraftment defect in the expanded cells as well as the differen-tiation of the stem cell population (Peters et al., 1995, 1996; Noltaet al., 1995; Ramshaw et al., 1995; Traycoff et al., 1996; Yone-mura et al., 1996; Veiby et al., 1997; Dannaeus et al., 1998;Varas et al., 1998). The loss of repopulating capacity inhematopoietic stem cells after ex vivo culture has been ascribedto the induction of cell cycling and alterations in adhesion mol-ecules or homing receptors on the expanded hematopoietic cells(Ramshaw et al., 1995; van der Loo and Ploemacher, 1995;Agrawal et al., 1996; Peters et al., 1996). In addition, studieshave shown that the most primitive CD341CD382 cells remainin the G0 phase after short-term (72–96 hr) in vitro exposure tovarious cytokine combinations (Agrawal et al., 1996; Jordan etal., 1996) and that these cells also have low or absent expressionof cell surface receptors necessary for retroviral transduction (Or-lic et al., 1996; Horwitz et al., 1999). The lack of cell cyclingand the absence of retroviral receptors on human CD341CD382

cells are two major obstacles to successful retrovirus-based genetherapy using hematopoietic stem cells (Miller et al., 1990; Roeet al., 1993; Knaan-Shanzer et al., 1995; Emmons et al., 1997).

We have previously described a porcine microvascular en-dothelial cell (PMVEC) coculture system that supports a 10- to15-fold expansion of the CD341 population and a unique 150-fold increase in the more primitive CD341CD382 subset inshort-term (7-day) cultures (Davis et al., 1995; Chute et al.,1999). We subsequently have demonstrated that human CD341

cells expanded in the PMVEC coculture system are capable ofproviding complete hematopoietic reconstitution in a SCID–Hu(severe combined immunodeficiency–human) model (Brandt etal., 1998) and that autologous bone marrow CD341 cells ex-panded on PMVEC monolayers are capable of engrafting andhematologically rescuing lethally irradiated baboons (Brandt etal., 1999).

Since gene therapy is a major clinical end point of stem cellex vivo expansion and since autologous bone marrow would bea preferred source of CD341CD382 hematopoietic stem cells forgene therapy protocols, we examined the effects of PMVEC co-culture on the cell cycle status and the expression of retroviralreceptors AmphoR and GaLVR in bone marrow CD341CD382

cells. In addition, we compared the efficiency of gene transferinto PMVEC-expanded CD341CD382 cells versus steady state(day 0) CD341CD382 cells and CD341CD382 cells expandedin liquid culture plus cytokines. Our results suggest that cocul-ture of human CD341CD382 cells on PMVEC monolayers pro-motes significantly higher efficiency of gene transfer into bonemarrow CD341CD382 cells compared with either day 0 or liq-uid culture-expanded cells. This higher level of gene transferinto PMVEC-expanded cells was most closely associated witha high level of cell cycling in the CD341CD382 population,whereas we found little change in the expression of either theamphotropic receptor (AmphoR) or the gibbon ape leukemiavirus receptor (GaLVR) under any condition. Since we havepreviously demonstrated the engraftment capacity of PMVEC-

expanded cells in both a SCID–Hu and nonhuman primatemodel, preincubation of CD341 cells with PMVEC monolay-ers prior to retroviral transduction may be an advantageousstrategy for clinical gene therapy protocols.

MATERIALS AND METHODS

Purification of CD341 cells

Human vertebral body bone marrow was procured from or-gan donors and CD341 cells were isolated as previously de-scribed (Davis et al., 1995) under a protocol that has been ap-proved by the institutional review board at the Naval MedicalResearch Center (Bethesda, MD). CD341 bone marrow pro-genitor cells were purified by positive immunomagnetic se-lection by rosetting low-density bone marrow mononuclearcells to magnetic beads (Dynal, Great Neck, NY) coated withbiotinylated anti-CD34 monoclonal antibody (MAb K6.1). Af-ter three or four cycles of magnetic attraction, the beads weredissociated from the cells with an excess of biotin (GIBCO,Grand Island, NY) and separated from the cells magneticallyand cryo-preserved. Cells by this procedure showed .97%positive reactivity with a second non-cross-blocking CD34-specific mono-clonal antibody (HPCA-2 conjugated with flu-orescein isothiocyanate [HPCA-2–FITC]; Becton DickinsonImmunocytometry Systems, San Jose, CA) by flow cytomet-ric analysis.

Cryopreserved CD341 cells were thawed rapidly at 37°C,diluted in a 103 volume of prewarmed (37°C) complete cul-ture medium consisting of Iscove’s modified Dulbecco’smedium (IMDM; GIBCO) supplemented with 10% heat-inac-tivated fetal calf serum (FCS; HyClone, Logan, UT), L-gluta-mine (100 mg/ml; GIBCO), and penicillin–streptomycin (100U/ml; GIBCO). Unless otherwise noted this culture mediumwill be referred to as complete culture medium. The thawedCD341 bone marrow cells were washed twice in complete cul-ture medium, and resuspended at 1 3 106 cells/ml. Cell viabil-ity was .95% as determined by trypan blue dye exclusion.

CD341 cells plus PMVEC coculture

PMVECs were cultured as previously described (Davis etal., 1995). Briefly, PMVECs were plated at concentrations of1 3 105 cells/well in gelatin-coated six-well tissue cultureplates (Costar, Cambridge, MA) containing 5 ml of M199 sup-plemented with 10% heat-inactivated FCS (HyClone), L-gluta-mine (100 mg/ml), heparin (50 mg/ml), endothelial cell growthfactor supplement (30 mg/ml; Sigma, St. Louis, MO), and peni-cillin–streptomycin solution (100 mg/ml). After 48–72 hr, theadherent PMVEC monolayers (70–80% confluent) werewashed twice with complete culture medium to remove anynonadherent PMVECs and the culture medium was replacedwith 7 ml of complete cell culture medium.

Purified CD341 cells (2 3 105) were added to each well. Cul-tures were treated with granulocyte-macrophage colony-stimu-lating factor (GM-CSF, 2 ng/ml), interleukin 3 (IL-3, 5 ng/ml),IL-6 (5 ng/ml), stem cell factor (SCF, 120 ng/ml), and Flt-3 lig-and (50 ng/ml; R&D Systems, Minneapolis, MN) and incubatedat 37°C in a humidified 5% CO2-in-air atmosphere. After 7 daysof culture, the monolayers were gently washed twice with com-

CHUTE ET AL.2516

plete culture medium to remove both the adherent and nonad-herent hematopoietic cells. Manual hemacytometer cell countswere performed with trypan blue exclusion dye.

Stroma-free liquid suspension cultures

Cell cultures were performed in medium as indicated above,but without PMVEC monolayers.

Cell cycle analysis

The surface, intracellular, DNA (SID) analysis was per-formed as described by Jordan et al. (1996). Briefly, on day 0and after 5 days of either PMVEC coculture or liquid cultureplus cytokines, CD341 cells were harvested, washed once, andresuspended in phosphate-buffered saline (PBS) plus 1.0%FCS. Cell surface staining was performed with anti-CD34 con-jugated with allophycocyanin (CD34–APC; Becton Dickinson)and anti-CD38 conjugated with phycoerythrin (CD38–PE; Bec-ton Dickinson). After surface staining was complete, cells wereresuspended in PBS and 0.4% formaldehyde (electron mi-croscopy [EM] grade; Electron Microscopy Sciences, Ft. Wash-ington, PA). Cells were then incubated on ice for 30 min, anequal volume of PBS plus 0.2% Triton X-100 (Sigma) wasadded, and the samples were left at 4°C overnight. The fixedand permeabilized cells were washed and resuspended in PBSplus 1% FCS and stained with Ki-67–FITC (MIB-1 clone; Im-munotech, Westbrook, ME). The cells were then washed andresuspended in PBS plus 1% FCS containing 7-aminoactino-mycin D (7-AAD, 0.5 mg/ml; Sigma). Samples were incubatedin 7-AAD overnight prior to analysis.

PCR-based liquid hybridization analysis of retrovirusreceptor mRNA expression in human bone marrowCD341 subsets

Total cellular RNA was isolated from adult populations of hu-man Lin2CD341 CD381 or Lin2CD341CD382 cells, usingRNA-STAT 60 (Tel-Test, Friendswood, TX) according to themanufacturer instructions. First-strand cDNA was synthesizedfrom equalized amounts of mRNA after two reactions withMoloney murine leukemia virus reverse transcriptase (Mo-MuLV RT) to ensure 100% incorporation (Stratagene, La Jolla,CA). cDNA was amplified with primers for the retroviral recep-tors for 35 cycles of 94°C for 1 min, 60°C for 1 min, and 72°Cfor 2 min in a programmable thermal controller (MJ Research,Watertown, MA), using the following primers: human AmphoR(sense primer, 59-CGG AAC ATC TTC TGT GCC TG?39; anti-sense primer, 59-GCT GGT CAT GAG AGA GCC GTG-39; frag-ment size, 220 bp), human GaLVR (sense primer, 59-GTA GTCCTT CTG AAA GCC CC-39; antisense primer, 59-CAC TGGAGT TTA TTT GGT TGC-39; fragment size, 330 bp). Glycer-aldehyde-3-phosphate dehydrogenase (GAPDH) polymerasechain reaction (PCR) was used to estimate experimental cDNAloading as previously described (Blair et al., 1998). Liquid hy-bridization of PCR products was conducted as described previ-ously (Blair et al., 1998), using the following probes: AmphoR(59-ATG GCT CTT CTC ATG TAT GGG-39), GaLVR (59-CCTGCC ACT GTG CCC CTC C-39), GAPDH (59-TCG CTC CTGGAA GAT GGT GAT GGG ATT-39). After hybridization, sam-ples were loaded onto 10% acrylamide gels, run at 140 V for 3hr, and and then exposed to X-ray film. Semiquantitative analy-sis of the bands was accomplished with a Molecular Dynamics

GENE TRANSFER INTO BONE MARROW STEM CELLS 2517

FIG. 1. Cell cycle status of human CD341CD382 cells on day 0, after 5 days of PMVEC coculture supplemented with GM-CSF 1 IL-3 1 IL-6 1 SCF 1 Flt-3 ligand, and after 5 days of liquid culture supplemented with the identical cytokines. Eachcolumn represents the mean percentage of cells in the G0 or G1/G2/S/M phase from three separate experiments. *The percentageof CD341CD382 cells that had entered into the cell cycle (G1/G2/S/M phase) was significantly greater (p , 0.001) followingPMVEC coculture than in the liquid culture group.

CHUTE ET AL.2518

FIG. 2. Representative FACS histograms demonstrating the cell cycle status of human CD341CD382 and CD341CD381 cellsubsets on day 0 (A), on day 5 of PMVEC coculture (B), and on day 5 of liquid culture (C). Numbers shown represent the per-centage of cells in each quadrant. The lower left quadrant represents cells in G0; upper left, G1; upper right, G2/S/M phase.

GENE TRANSFER INTO BONE MARROW STEM CELLS 2519

FIG. 3. Bar graphs showing the percentages of (A) CD341 cells and (B) CD341CD382 cells present in culture on day 5 ofPMVEC coculture versus liquid suspension culture (n 5 3). *p value for the difference between PMVEC culture and liquid cul-ture is 0.03; **p value for the difference between PMVEC culture and liquid culture is 0.01.

(Sunnyvale, CA) densitometer system. Levels of AmphoR andGaLVR mRNA in CD341 subsets were normalized to AmphoRand GaLVR levels in HeLa cells, using a modification of the for-mula described by Orlic et al. (1998):

1 23 1 2

5 relative level of receptor mRNA in human CD341 cells

Generation of Moloney murine leukemia virusretroviral vectors

Both amphotropic (Ampho/Mo-MuLV) and GaLV pseudo-type LAPSN (GaLV/Mo-MuLV) viral vectors were generated

GAPDH mRNA in HeLa cells}}}}receptor mRNA in HeLA cells

Receptor mRNA in CD341 cells}}}}GAPDH mRNA in CD341 cells

bearing the human placental alkaline phosphatase (HPAP) re-porter gene, using PA317 and PG13 packaging cell lines ac-cording to methods described previously (Miller et al., 1994;Kiem et al., 1998). Briefly, packaging cell lines were trans-fected by the calcium phosphate method using plasmidpLAPSN (Miller et al., 1994). Transfected packaging cell lineswere selected for G418 (neomycin) resistance, using a 400- to800-mg/ml concentration of Geneticin (GIBCO-BRL, Gaithers-burg, MD) in Dulbecco’s modified Eagle’s medium (DMEM;GIBCO) plus 10% FCS. Producer cell lines were expanded in225-cm2 tissue culture flasks (Falcon, Franklin Lakes, NJ) at37°C/5% CO2 using DMEM (GIBCO-BRL) with 10% FCS.Collection of virus-containing medium (VCM) commencedwhen packaging cells were about 80% confluent. At this stagethe medium was replaced with 25 ml of DMEM containing 10%FCS and cells were incubated at 32°C. Twenty-four hours laterthe medium was collected, filtered (0.45-mm pore size), and

TABLE 1. IN VITRO EXPANSION OF CD341 CELL SUBSETS DURING PMVEC COCULTURE VERSUS STROMA-FREE LIQUID CULTUREa

No. of cells procured (n 5 3)

Culture conditions Cell yield CD341 CD341CD382

Input (day 0) 2.0 3 105 1.9 (6 0.4) 3 105 2.7 (6 0.2) 3 103

PMVEC Coculture 13.0 (6 0.4) 3 105 4.9 (6 1.3) 3 105 8.1 (6 0.1) 3 104

(day 5) (6.5-fold) (2.6-fold) (30-fold)Stroma-free 11.3 (6 0.4) 3 105 2.7 (6 1.8) 3 105 8.8 (6 0.7) 3 103

liquid culture (5.7-fold) (1.4-fold) (3.3-fold)(day 5)

aPurified CD341 bone marrow cells (2 3 105) were placed in culture on day 0 (input) in each experiment (n 5 3). Fold expansion of cells after 5 days of culture is shown in parentheses for each subset.

CHUTE ET AL.2520

FIG. 4. Phenotype of bone marrow CD341 cells on day 0 (A) versus day 5 PMVEC culture (B) and day 5 liquid culture (C).CD341CD382 cells were defined as cells expressing CD34 and lacking CD38 expression compared with the isotype controls.

GENE TRANSFER INTO BONE MARROW STEM CELLS 2521

frozen at 270°C. VCM was titrated using serial 10-fold dilu-tions with Polybrene (8 mg/ml) and exposure to HT1080 cells(70–80% confluence) for 24 hr. After infection, cells were se-lected for G418 resistance, and at 7–10 days colonies were ex-amined for the presence of HPAP activity using nitroblue tetra-zolium (NBT) and 5-bromo-4-chloro-3-indolyl phosphate(BCIP) (BioRad, Hercules, CA). Packaging cell lines capableof producing higher numbers of HPAP-positive colonies inHT1080 cells were selected for this study. The titers of am-photropic and pseudotyped GaLV/Mo-MuLV ranged from 1 3

105 to 3 3 105 CFU/ml for all transfection experiments.

Transduction of human hematopoietic CD341 cellsubsets using retroviral supernatants

Moloney murine leukemia virus vectors encoding HPAPwere prepared with packaging cell lines PA317 and PG13 as

described above (Miller et al., 1994; Kiem et al., 1998). Freshlyprepared retroviral vector supernatants were cryopreserved at270°C in IMDM 1 10% FCS 1 1% penicillin–streptomycin.Bone marrow hematopoietic cells from day 0, after 5 days ofliquid culture, and after 5 days of PMVEC coculture were har-vested, washed, and resuspended in warm PG13 or PA317VCM-supplemented IL-3 (5 ng/ml) 1 SCF (120 ng/ml) 1

Polybrene (8 mg/ml) (transduction medium) at a concentrationof 2 3 105 cells/ml. A minimum of 2 3 106 CD341 cells wasutilized at the initiation of each experiment to ensure adequatenumbers of cells for analysis posttransfection. Two millilitersof each cell suspension was aliquoted into 14-ml round-bottompolypropylene tubes (Falcon) and centrifuged at 2000 rpm for2 hr at 37°C. After centrifugation, the pellets were gently re-suspended and the tubes were placed in a 37°C/5% CO2 hu-midified incubator overnight. The tubes were then centrifugedfor 10 min at 1200 rpm. The overlying transduction mediumwas removed and the cells were resuspended in fresh trans-duction medium. This was repeated every 24 hr for 3 days. Af-ter 3 days, the hematopoietic cells were washed twice with com-plete culture medium and cell counts and viability testing wereperformed. These cells were then stained with CD34–FITC andCD38–PE and the CD341CD382 and CD341CD381 popula-tions were sorted via fluorescence-activated cell sorting(FACS). Cytospin slides were then prepared from each popu-lation and transduction efficiency was measured by staining thesorted populations for HPAP activity.

Retroviral–human placental alkaline phosphatase staining

CD341CD382 and CD341CD381 cells were sorted byFACS and collected for HPAP staining and methylcellulosecolony-forming assays. Cytospins were prepared from 2 3 104

to 5 3 104 CD341CD382 cells and from 2 3 104 to 5 3 105

CD341CD381 cells and HPAP activity was tested as previ-

FIG. 5. Semiquantitative comparison of RT-PCR product from different human CD341 subsets on day 0, day 5 PMVEC co-culture, and day 5 liquid culture. The top row reflects the level of GaLVR mRNA expression, the middle row reflects AmphoRmRNA expression, and the bottom row shows GAPDH mRNA expression in each sample. At the right is the expression of mRNAfor each transcript in HeLa cells.

TABLE 2. RT-PCR ANALYSIS OF RETROVIRUS RECEPTOR

mRNA IN HUMAN BONE MARROW CD341 SUBSETSa

Condition/subset GaLVR AmphoR

Day 0/381 (2) 0.07 6 0.06 0.22 6 0.26Day 0/382 (2) 0.11 6 0.07 0.05 6 0.04

PMVEC Day 5/381 (2) 0.31 6 0.42 3.08 6 1.81PMVEC Day 5/382 (2) 0.08 6 0.04 0.16 6 0.21

LC Day 5/381 (2) 0.05 6 0.03 0.05 6 0.02LC Day 5/382 (2) 0.03 6 0.01 0.01 6 0.01

HeLa cells (2) 1.0 1.0

aThe number of samples analyzed is shown in parentheses.Receptor mRNA levels were normalized to the level of GAPDHmRNA in the same sample and mean levels of mRNA werecalculated relative to HeLa cells 6 SD.

CHUTE ET AL.2522

TABLE 3. EFFICIENCY OF RETROVIRAL GENE TRANSFER INTO CD341 CELL SUBSETS AT STEADY STATE

VERSUS AFTER EXPANSION IN LIQUID SUSPENSION CULTURE OR AFTER PMVEC COCULTUREa

PA317 PG13

Condition % HPAP positive p Value % HPAP positive p Value

CD341 CD38 2 Cells

Day 0 0.6 6 1.7 0.2 6 0.6

Liquid 1.4 6 3.5 0.29 0.0 0.17

PMVEC 11.4 6 5.6 ,0.01 10.9 6 5.2 ,0.01

CD341 CD38 1 Cells

Day 0 12.2 6 3.4 16.8 6 9.7

Liquid 11.2 6 3.3 0.24 11.5 6 9.2 0.12

PMVEC 10.9 6 7.0 0.44 15.3 6 8.6 0.16

aData shown represent the mean percentage of cells staining positively for human placental alkaline phosphatase from two separate experiments that tested both vectors (PA317 and PG13; titers, 1–3 3 105 CFU/ml). For each experiment, five randomlyselected high-power fields (340) were examined and the mean percentages were calculated on the basis of the number of cellscounted per high-power field. Cytospins were prepared from FACS-sorted CD341CD382 and CD341CD381 cells.

ously described (Kiem et al., 1998). Briefly, the cytospins wererinsed in PBS, fixed in 2% formaldehyde, rinsed twice in deio-dinized H2O (diH2O), and incubated at 65°C for 30 min in or-der to inactivate endogenous alkaline phosphatase activity. Theslides were then rinsed in diH2O and incubated in 100 mMTris-HCl (pH 10.0), X-phosphate (1 mg/ml), and NBT (1mg/ml) for 3 hr at 37°C. The incubation was ended by rinsingin diH2O. HPAP is heat stable relative to other endogenous al-kaline phosphatases (Kiem et al., 1998) and survives this heatinactivation step. All reagents other than NBT (BioRad) wereobtained from Sigma.

HPAP analysis of clonogenic progenitor cells

After transduction, 5 3 102 to 1 3 103 sorted CD341CD382

and CD341CD381 populations were placed in methylcellulosesemisolid medium supplemented with erythropoietin (EPO, 5U/ml), GM-CSF (2 ng/ml), IL-3 (5 ng/ml), IL-6 (5 ng/ml), andSCF (120 ng/ml) (R&D Systems) for 14 days. Colony-formingcells were tested for HPAP activity in the same manner as de-scribed above. Colony-forming assay dishes on day 14 wereheated to 65°C in a water bath and then allowed to cool at 4°Cfor 1 hr. The methylcellulose layers were then overlaid with theHPAP staining solution for 3 hr at 37°C. After 3 hr, the stain-ing solution was aspirated off and the colonies were counted.The transduced colonies generated from CD341CD382 cellswere compared with colonies generated from nontransducedCD341CD382 cells. Triplicate assays were set up for each datapoint per experiment.

Statistical analysis

Statistical comparisons between different populations ofCD341 cells under different treatment conditions were madeby using the Student t test.

RESULTS

Cell cycle (SID) analysis and phenotype of CD341

cells after PMVEC coculture versus liquid suspension cultures

We have previously shown that PMVEC coculture supple-mented with GM-CSF 1 IL-3 1 IL-6 1 SCF 1 Flt-3 ligandsupports a potent expansion of CD341 cells and of the prim-itive CD341CD382 subpopulation (Davis et al., 1995). In thisstudy, we first examined the cell cycle status of CD341 sub-sets during PMVEC coculture supplemented with GM-CSF 1IL-3 1 IL-6 1 SCF 1 Flt-3 ligand versus liquid culture plusthe identical cytokines. On day 0 (steady state), we found thatthe large majority (97.5 6 1.8%) of bone marrowCD341CD382 cells resided in G0, 2.1 6 1.5% were in G1, and0.3 6 0.3% were in G2/S/M phase (n 5 3; Fig. 1). After 5 daysof PMVEC coculture, 93.5 6 0.7% of the CD341CD382 pop-ulation had entered the G1 or G2/S/M phase and only 6.4 6

0.7% remained in G0. In comparison, on day 5 of liquid sus-pension culture, 73.0 6 2.2% of the CD341CD382 cells hadentered either the G1 or G2/S/M phase, but 26.7 6 2.2% re-mained in G0. The percentage of CD341CD382 cells that hadentered the cell cycle was significantly greater in the PMVECculture group than in the liquid culture group (p , 0.001). Arepresentative FACS histogram showing the percentages ofCD341CD382 cells in G0, G1, and G2/S/M phase on day 0,after 5 days of PMVEC coculture, and after 5 days of liquidculture is shown in Fig. 2A–C.

In concert with the high level of cell division that occurredin the CD341CD382 population during PMVEC coculture, weobserved a significant increase in the percentage ofCD341CD382 cells during the 5-day culture period. In con-trast, the CD341CD382 population was not maintained in liq-

GENE TRANSFER INTO BONE MARROW STEM CELLS 2523

FIG. 6. (A) Wrights–Giemsa staining of human CD341CD382 cells from two experiments after 5 days of PMVEC cocultureand 3 days of transduction with PG13, (B) high magnification view (3400) of HPAP–positive CD341CD382 cells from two ex-periments after expansion in PMVEC coculture and transduction with PG13, and (C) a CFU-GM generated from PMVEC-ex-panded CD341CD382 cells, which shows HPAP activity after transduction with PG13.

A

B

C

CHUTE ET AL.2524

uid suspension culture supplemented with the same cytokines.Figure 3 shows the results from three separate experiments inwhich the mean starting populations of bone marrow cells were97 6 2.6% CD341 and 1.4 6 1.0% CD341CD382. After 5days of PMVEC coculture, 37.7 6 1.7% of the hematopoieticcells remained CD341 (Fig. 3A) and the CD341CD382 popu-lation had increased to 6.5 6 1.3% of the total population (Fig.3B). In contrast, after liquid culture, 22.5 6 8.8% of the finalpopulation remained CD341 but the CD341CD382 populationdecreased to only 0.7 6 0.4%. Similarly, on day 5 of PMVECcoculture, we observed a 6.5-fold increase in total cells, a 2.6-fold increase in CD341 cells, and a 30-fold increase in theCD341CD382 subset (Table 1). On day 5 of liquid culture, weobserved a 5.7-fold increase in total cells, a 1.4-fold increasein CD341 cells, but only a 3.3-fold increase in cells bearing theCD341CD382 phenotype. The percentage of cells expressingCD34 and the CD341CD382 phenotype was significantlygreater after PMVEC coculture than after liquid culture CD341,p 5 0.03; CD341CD382, p 5 0.01). Figure 4 shows the phe-notype of bone marrow CD341 cells from a representative ex-periment on day 0 (Fig. 4A) and after 5 days of PMVEC co-culture (Fig. 4B) and liquid suspension culture (Fig. 4C).

Expression of AmphoR and GaLVR in bone marrowCD341 subpopulations on day 0 and after PMVECcoculture versus liquid culture

To assess how PMVEC coculture affected the expression ofAmphoR and GaLVR on sorted populations CD341CD382 andCD341CD381 cells, we performed semiquantitative RT-PCRon mRNA collected from CD341 subset samples from day 0and after PMVEC coculture and liquid suspension culture, andwe compared these mRNA levels with those of HeLa cells aspreviously described (Orlic et al., 1998). As shown in Fig. 5,we found that the expression of the amphotropic receptor mRNAand GaLVR in day 0 bone marrow CD341CD382 subsets waslow relative to expression in HeLa cells. After PMVEC cocul-ture, the expression of AmphoR mRNA and GaLVR in theCD341CD382 subset did not significantly change relative toHeLa mRNA levels. Similarly, after 5 days in liquid suspensionculture, the expression of AmphoR and GaLVR mRNA in theCD341CD382 subset remained low. We also found consistentlylow expression of GaLVR in all samples compared with the ex-pression of AmphoR in the same samples, suggesting a differ-ential rate of transcription for these two retroviral receptorswithin bone marrow CD341 subsets. Table 2 shows the calcu-lated levels of retroviral receptor mRNA expression in bone mar-row CD341 subsets relative to expression in HeLa cells. Themean levels of AmphoR and GaLVR mRNA in day 0CD341CD382 cells were 5 and 11% of the mean level in HeLacells, whereas PMVEC-expanded CD341CD382 cells had meanlevels that were 16 and 8% of the mean level in HeLa cells. Themean levels of AmphoR and GaLVR mRNA in CD341CD382

cells after liquid culture were also low at 1 and 3%.

Retroviral gene transfer into CD341CD382 cells onday 0 and after expansion in PMVEC coculture versus liquid culture

Purified bone marrow CD341 cells were expanded inPMVEC coculture supplemented with GM-CSF 1 IL-3 1 IL-6 1 SCF 1 Flt-3 ligand for 5 days and the expanded popula-

tion was then transduced for 3 days with Mo-MuLV vectors(1–3 3 105 CFU/ml) bearing the HPAP reporter gene. For com-parison, CD341 cells on day 0 and CD341 cells that weretreated in liquid culture plus the identical cytokines were trans-duced in an identical fashion with PA317 and PG13 vector su-pernatants. Using vector from PG13 (rvPG13), 10.9 6 5.2% ofthe CD341CD382 cells transduced after PMVEC coculturedemonstrated HPAP activity (Table 3). Using vector fromPA317 (rvPA317), 11.4 6 5.6% of the PMVEC-expandedCD341CD382 cells were HPAP positive. Wrights–Giemsastained PMVEC-expanded CD341CD382 cells from two sep-arate experiments after 3 days of transduction showed typicalblast cell morphology, with a high nuclear cytoplasmic ratioand a fine chromatin pattern (Fig. 6A). Figure 6B shows HPAPactivity in human CD341CD382 cells collected from two sep-arate experiments after PMVEC expansion and transductionwith rvPG13.

As shown in Table 3, the efficiencies of transduction into day0 CD341CD382 cells and CD341CD382 cells collected afterliquid culture were significantly lower compared with the effi-ciency of transduction into PMVEC-expanded CD341CD382

cells. Efficiencies of transduction into CD341CD381 cells in theday 0, liquid culture, and PMVEC coculture groups are alsoshown in Table 3. We observed no significant difference betweenthe efficiency of transduction into CD341CD381 cells on day 0or after 5 days of liquid culture or PMVEC coculture (Table 3).

Retroviral gene transfer into colony-forming cells

CD341CD382 cells that were collected after 5 days ofPMVEC coculture and 3 days of exposure to rvPG13 orrvPA317 were then placed in methylcellulose to assess HPAPexpression in 14-day colony-forming cells (CFCs). On day 14,35.1 6 15.5% of the CFCs derived from CD341CD382 cellstransduced with rvPA317 were HPAP positive. Similarly,38.1 6 17.5% of the CFCs derived from CD341CD382 cellsexposed to rvPG13 demonstrated HPAP positivity (Table 4). Arepresentative CFU-GM showing HPAP expression is shownin Fig. 6C. As controls, CFCs that were derived from non-transduced CD341CD382 cells were tested for HPAP activityafter heat inactivation and these colonies were HPAP negative(data not shown). For comparison, CD341CD382 cells on day0 and after liquid culture were also transduced with rvPG13 andrvPA317 for 3 days and then placed in methylcellulose colony-forming assays. The mean number of transduced colonies inboth the day 0 and liquid culture groups were significantly lowerthan in the PMVEC coculture group (Table 4). The expressionof HPAP in CFCs generated from CD341CD381 cells usingeither vector ranged from 8.3 to 21.4% in the day 0 group, from45.4 to 58.1% in the liquid culture group, and from 62.5 to64.2% in the PMVEC coculture group. The efficiency of trans-duction into CFCs generated from CD341CD381 cells afterliquid culture or PMVEC coculture was significantly higherthan the day 0 group, using either rvPA317 or rvPG13 (p ,

0.01 and p , 0.01; Table 4).

DISCUSSION

The CD341CD382 population within human bone marrowis enriched for cells with in vivo long-term repopulating ca-

GENE TRANSFER INTO BONE MARROW STEM CELLS 2525

pacity (Berardi et al., 1995; Hao et al., 1995; Bhatia et al., 1997)and, therefore, these cells are ideal targets for therapeutic genetransfer. However, the efficiency of retroviral gene transfer intothese cells has been shown to be consistently low (Landsdorpand Dragowska, 1993; Dao et al., 1998; McCowage et al.,1998). These disappointing results are attributable to the lackof ex vivo conditions that promote cell cycling in these cellswhile maintaining their primitive CD341CD382 phenotype. Inaddition, ex vivo conditions that upregulate the numbers ofretroviral receptors on the surface of CD341CD382 cells wouldbe considered advantageous for gene therapy protocols (Crooksand Kohn, 1993; Gentry and Smith, 1999). Methods to expandand induce cell cycling in hematopoietic stem cells in short-term ex vivo liquid cultures have been associated with the ac-quisition of an engraftment defect in the expanded graft (Peterset al., 1995; Traycoff et al., 1996; Yonemura et al., 1996) alongwith the differentiation of hematopoietic stem cells into com-mitted progenitors (Veiby et al., 1997; Varas et al., 1998). Wehave reported that a porcine microvascular endothelial cell(PMVEC) monolayer supplemented with cytokines supports theex vivo expansion of hematopoietic progenitor cells capable ofrepopulating SCID–Hu bone (Brandt et al., 1998) and rescuinglethally irradiated baboons in vivo (Brandt et al., 1999). In thisstudy, we have found that 5-day coculture of human bone mar-row CD341 cells with PMVEC monolayers followed by 3 daysof transduction with retroviral vector supernatant significantlyincreased the efficiency of retroviral gene transfer into theCD341CD382 subpopulation. In contrast, the efficiency ofretroviral gene trnsfer into nonexpanded (day 0) bone marrowCD341CD382 cells and CD341CD382 cells collected after liq-uid suspension culture was low.

There are two effects of PMVEC coculture that contribute tothe increased gene transfer efficiency we have observed withthis system: First, PMVEC coculture causes a high percentage

(.93%) of the CD341CD382 population to enter cell divisionafter only a short (5-day) period of coculture. In contrast, .95%of the day 0 CD341CD382 cells harvested from the bone mar-row reside in G0 and these cells could not be efficiently trans-duced by the same transduction methods. The level of cell cy-cling in hematopoietic stem cells has been consistentlyassociated with a higher efficiency of retroviral gene transfer inprevious studies (Miller et al., 1990; Knaan-Shanzer et al.,1995). Second, short-term (5- to 10-day) coculture of humanCD341 cells on PMVEC monolayers plus cytokines promotesan expansion of cells bearing the CD341CD382 phenotypewhile cell division is ongoing. In contrast, liquid suspension cul-tures supplemented with the identical cytokines induce a sig-nificant level of cell cycling by CD341CD382 cells, but cellsbearing the CD341CD382 phenotype decline to negligible lev-els by day 5. The significant increase in cells bearing theCD341CD382 phenotype after PMVEC coculture comparedwith liquid suspension cultures suggests that contact with en-dothelial cells promotes the maintenance of more primitive re-populating cells in culture. Glimm and Eaves (1999) have shownthat optimal cytokine combinations of IL-3 1 IL-6 1 GCSF 1

SCF 1 Flt-3 ligand cause measurable cell divisions amongCD341CD382 cells, but our results indicate that this combina-tion of cytokines in the absence of PMVEC monolayers rapidlyleads to differentiation of CD341CD382 cells. Other investiga-tors have demonstrated that a murine stromal cell line, AFT024(Thiemann et al., 1998), supplemented with IL-3 1 IL-6 1

SCF, can maintain low levels of CD341CD382 cells (range,0.11–2.30%) in 3- to 10-day cultures. These authors hypothe-sized that AFT024 might maintain primitive CD341CD382 cellsthrough an inhibition of apoptosis or differentiation in the prog-enitor cell pool. PMVEC monolayers supplemented with IL-3 1

IL-6 1 GM-CSF 1 SCF 1 Flt-3 ligand offer a specific advan-tage for gene transfer applications by promoting a high level of

TABLE 4. EFFICIENCY OF RETROVIRAL GENE TRANSFER INTO CFCS GENERATED FROM CD341 SUBSETS

ON DAY 0 VERSUS AFTER LIQUID SUSPENSION CULTURE AFTER PMVEC COCULTUREa

PA317 PG13

Condition % HPAP-positive CFC p Value % HPAP-positive CFC p Value

CD341 CD38 2 Cells

Day 0 0.0 0.0

Liquid 2.5 6 2.80 0.09 0.5 6 0.8 0.18

PMVEC 35.1 6 15.5 ,0.01 38.1 6 17.5 ,0.01

CD341 CD38 1 Cells

Day 0 8.3 6 7.2 21.4 6 12.2

Liquid 58.1 6 10.9 ,0.01 45.4 6 10.7 ,0.01

PMVEC 64.2 6 14.3 0.23 62.5 6 4.7 ,0.01

aData shown represent the mean percentage of colonies staining positively for human placental alkaline phosphatase from two separate experiments, which tested both vectors (PA317 and PG13; titers, 1–3 3 105 CFU/ml). For each experiment, triplicatemethylcellulose cultures were examined and mean percentages were calculated on the basis of the number of colonies countedper dish. Methylcellulose CFC cultures were prepared from FACS-sorted CD341CD382 or CD341CD381 cells.

cell cycling in CD341CD382 cells in short-term culture whilemaintaining a population of cells capable of in vivo engraftment.Therefore, expansion of human CD341CD382 cells on PMVECmonolayers followed by retroviral transduction should increasethe likelihood that gene-marked cells will give rise to long-termrepopulation in vivo.

Our studies indicate that the levels of both AmphoR andGaLVR mRNA in steady state human bone marrowCD341CD382 populations are low, and that these levels are notsignificantly increased after either PMVEC coculture or liquidsuspension culture. These results are consistent with observa-tions that bone marrow CD341CD382 cells express low levelsof retroviral mRNA compared with cryopreserved human cordblood CD341CD382 cells (22-fold higher levels) (Orlic et al.,1998) and G-CSF-mobilized peripheral blood (4-fold higher)(Horwitz et al., 1999). However, our data further suggest thatlow levels of expression of AmphoR and GaLVR mRNA in thetarget cells may not be an absolute limiting factor toward im-proving the efficiency of gene transfer into bone marrowCD341CD382 cells. Our results contrast with a recent analysisthat suggested that the efficiency of transduction of both celllines and primary peripheral blood CD341 cells with am-photropic retrovirus was dependent on the expression of the am-photropic receptor (PiT-2) on the cell surface (Macdonald et al.,2000). In our studies, PMVEC coculture caused modest or noincrease in the level of AmphoR mRNA and no significantchange in the levels of GaLVR mRNA expressed in bone mar-row CD341CD382 cells compared with day 0 bone marrowCD341CD382 cells. Therefore, in our studies, the significantlyhigher efficiency of gene transfer into PMVEC-expandedCD341CD382 cells appears to have been caused almost exclu-sively by the increased cell cycling in the target population. Al-ternatively, retroviruses are capable of infecting target cellsthrough receptors other than the identified envelope receptors(e.g., AmphoR or GaLVR) (Miller, 1996), and thus unknownreceptor changes that occur during PMVEC coculture also mayhave accounted for the increased transduction efficiency we haveobserved. Of note, the titers of retrovirus used in our studies(1–3 3 105 CFU/ml) were less than “high-titer” concentrations(.1 3 106 CFU/ml) (van Hennick et al., 1998) that have beenconsidered critical for effective transduction efficiency (Mulli-gan et al., 1993). This further indicates that the preincubation oftarget hematopoietic cells with endothelial monolayers augmentsthe efficiency of transduction into these primitive cells.

Although these studies demonstrate that bone marrowCD341CD382 cells can be successfully transduced afterPMVEC coculture, the efficiency of gene transfer might be fur-ther increased by utilizing umbilical cord blood CD341 cellsas opposed to adult bone marrow CD341 cells. The efficiencyof retroviral gene transfer into umbilical cord CD341 cells hasbeen reported to be as high as 47–54% (van Hennik et al., 1998;Hennemann et al., 1999). These higher efficiencies using cordblood CD341 cells may be due, in part, to significant ontogenicdifferences between bone marrow CD341 cells and umbilicalcord blood CD341 cells with regard to both the frequency ofprimitive cells and their proliferative potential as well as theirexpression of retroviral receptors (Orlic et al., 1998; Weekx etal., 1998; Zandstra et al., 1998). Nonetheless, autologous bonemarrow CD341 cells will continue to be a primary and pre-ferred source of stem cells for human gene therapy applications.

Since we have previously shown that both human and primatebone marrow progenitor cells expanded ex vivo on PMVECmonolayers are capable of long-term repopulation in trans-planted recipients, we are optimistic that transduction protocolsthat include preincubation of human bone marrow progenitorcells with endothelial monolayers will improve the likelihoodof long-term engraftment of gene-marked cells.

ACKNOWLEDGMENT

The views presented herein are those of the authors; no en-dorsement by the Department of the Navy has been given orshould be inferred. This work has been supported in part by theNaval Medical Research and Development Command, Re-search Task No. 63706.M00095.003.1458

REFERENCES

AGRAWAL, Y.P., AGRAWAL, R.S., SINCLAIR, A.M., YOUNG, D.,MARUYAMA, M., LEVINE, F., and HO, A.D. (1996). Cell-cyclekinetics and VSV-G pseudotyped retrovirus-mediated gene transferin blood-derived CD341 cells. Exp. Hematol. 24, 738–747.

BERARDI, A.C., WANG, A., LEVINE, J., LOPEZ, P., and SCAD-DEN, D.T. (1995). Functional isolation and characterization of hu-man hematopoietic stem cells. Science 267, 104–108.

BHATIA, M., WANG, J., KAPP, U., BONNET, D., and DICK, J.E.(1997). Purification of primitive human hematopoietic cells capableof repopulating immune-deficient mice. Proc. Natl. Acad. Sci. U.S.A.94, 5320–5325.

BLAIR, P.J., RILEY, J.L., LEVINE, B.L., LEE, K.P., CRAIGHEAD,N., FRANCOMANO, T., PERFETTO, S.J., GRAY, G.S., CAR-RENO, B.M., and JUNE, C.H. (1998). Cutting edge: CTLA-4 liga-tion delivers a unique signal to resting CD4 T cells that inhibits in-terleukin-2 secretion but allows Bcl-xL induction. J. Immunol. 160,12–16.

BRANDT, J.E., GALY, A., LUENS, K., TRAVIS, M., YOUNG, J.,TONG, J., CHEN, S., DAVIS, T.A., LEE, K., CHEN, B., TUSHIN-SKI, R., and HOFFMAN, R. (1998). Bone marrow repopulation byhuman marrow stem cells after long term expansion culture on aporcine endothelial cell line. Exp. Hematol. 26, 950–961.

BRANDT, J.E., BARTHOLEMEW, A., FORTMAN, J., NELSON, M.,BRUNO, E., CHEN, L., TURIAN, J., DAVIS, T., CHUTE, J.P., andHOFFMAN, R. (1999). Ex vivo expansion of autologous bone mar-row stem cells with porcine microvascular endothelial cells resultsin a graft capable of rescuing lethally irradiated baboons. Blood 94,106–113.

BRUGGER, W., HEIMFELD, S., BRENSON, R.J., MERTELS-MANN, R., and KANZ, L. (1995). Reconstitution of hematopoiesisafter high-dose chemotherapy by autologous progenitor cells gener-ated ex vivo. N Engl. J. Med. 333, 283–287.

CHUTE, J.P., SAINI, A., KAMPEN, R., WELLS, M., and DAVIS,T.A. (1999). A comparative study of the cell cycle status and prim-itive cell adhesion molecule profile of human CD341 cells culturedin stroma free versus porcine microvascular endothelial cell cultures.Exp. Hematol. 27, 370–379.

CROOKS, G.M., and KOHN, D.B. (1993). Growth factors increase am-photropic retrovirus binding to human CD341 bone marrow progen-itor cells. Blood 82, 3290–3297.

DANNAEUS, K., NILSSON, K., and JONSSON, J.I. (1998). Distinctactions of Flt-3 ligand and stem cell factor on myeloid lineage se-lection and maturation of granulocytes versus macrophages. Exp.Hematol. 26, 345–352.

CHUTE ET AL.2526

DAO, M.A., SHAH, A.J., CROOKS, G.M., and NOLTA, J.A. (1998).Engraftment and retroviral marking of CD341 and CD341CD382

human hematopoietic progenitors assessed in immune-deficientmice. Blood 91, 1243–1255.

DAVIS, T., ROBINSON, D., LEE, K., and KESSLER, S. (1995).Porcine brain microvascular endothelial cells support the in vitro ex-pansion of human primitive hematopoietic bone marrow progenitorcells with a high replating potential: Requirement for cell-to-cell in-teractions and colony-stimulating factors. Blood 85, 1751–1761.

EMMONS, R.V., DOREN, S., ZUJEWSKI, J., COTTLER-FOX, M.,CARTER, C.S., HINES, K., O’SHAUGHNESSY, J.A., LEITMAN,S.F., GREENBLATT, J.J., and DUNBAR, C.E. (1997). Retroviralgene transduction of adult peripheral blood or marrow derivedCD341 cells for six hours without growth factors or on autologousstroma does not improve marking efficiency assessed in vivo. Blood89, 4040–4046.

GENTRY, T., and SMITH, C. (1999). Retroviral vector mediated genetransfer into umbilical cord blood CD34brCD382CD332 cells. Exp.Hematol. 27, 1244–1254.

GLIMM, H., and EAVES, C.J. (1999). Direct evidence for multiple selfrenewal divisions of human in vivo repopulating cells in short termculture. Blood 94, 2161–2168.

HAO, Q.L., SHAH, A.J., THIEMANN, F.T., SMOGORZEWSKA,E.M., and CROOKS, G.M. (1995). A functional comparison ofCD341CD382 cells in cord blood and bone marrow. Blood 86,3745–3753.

HENNEMANN, B., CONNEALLY, E., PAWLIUK, R., LEBOULCH,P., ROSE-JOHN, S., REID, D., CHUO, J., HUMPHRIES, R., andEAVES, C.J. (1999). Optimization of retroviral-mediated gene trans-fer to human NOD/SCID mouse repopulating cord blood cellsthrough a systematic analysis of protocol variables. Exp. Hematol.27, 817–825.

HORWITZ, M., MALECH, H., ANDERSON, S., GIRARD, L., BO-DINE, D., and ORLIC, D. (1999). Granulocyte colony stimulatingfactor mobilized peripheral blood stem cells enter into G1 of the cellcycle and express higher levels of amphotropic retrovirus receptormRNA. Exp. Hematol. 27, 1160–1167.

HUANG, S., and TERSTAPPEN, W.M. (1994). Lymphoid and myeloiddifferentiation of single human CD341, HLA-DR1, CD382

hematopoietic stem cells. Blood 83, 1515–1526.JORDAN, C.T., YAMASAKI, G., and MINAMOTO, D. (1996). High

resolution cell cycle analysis of defined phenotypic subsets withinprimitive human hematopoietic cell populations. Exp. Hematol. 24,1347–1355.

KIEM, H.P., ANDREWS, R.G., MORRIS, J., PETERSON, L., HEY-WARD, S., ALLEN, J., RASKO, J., POTTER, J., and MILLER,A.D. (1998). Improved gene transfer into baboon marrow repopu-lating cells using recombinant human fibronectin fragment CH-296in combination with interleukin-6, stem cell factor, Flt-3 ligand, andmegakaryocyte growth and development factor. Blood 92, 1878–1886.

KNAAN-SHANZER, S., VALERIO, D., and VAN BEUSECHEM,V.W. (1995). Cell cycle state, response to hematopoietic growth fac-tors and retroviral vector-mediated transduction of human hemato-poietic stem cells. Gene Ther. 3, 323–333.

LANSDORP, P.M., and DRAGOWSKA, W. (1993). Maintenance ofhematopoiesis in serum free bone marrow cultures involves sequen-tial recruitment of quiescent progenitors. Exp. Hematol. 21, 1321–1327.

MACDONALD, C., WALKER, S., WATTS, M., INGS, I., LINCH,D., and DEVEREUX, S. (2000). Effect of changes in expression ofthe amphotropic retroviral receptor PiT-2 on transduction efficiencyand viral titer: Implications for gene therapy. Hum. Gene Ther. 11,587–595.

MCCOWAGE, G.B., PHILLIPS, K.L., GENTRY, T.L., HULL, S.,KURTZBERG, J., GILBOA, E., and SMITH, C. (1998). Multipa-

rameter-fluorescence activated cell sorting analysis of retroviral vec-tor gene transfer into primitive umbilical cord blood cells. Exp.Hematol. 26, 288–298.

MILLER, A.D. (1996). Cell-surface receptors for retroviruses and im-plications for gene transfer. Proc. Natl. Acad. Sci. U.S.A. 93, 11407–11413.

MILLER, D.G., ADAM, M.A., and MILLER, A.D. (1990). Gene trans-fer by retrovirus vectors occurs only in cells that are actively repli-cating at the time of infection. Mol. Cell. Biol. 10, 4239–4242.

MILLER, D.G., EDWARDS, R.H., and MILLER, A.D. (1994). Cloningof the cellular receptor for amphotropic murine retroviruses revealshomoloogy to that for gibbon ape leukemia virus. Proc. Natl. Acad.Sci. U.S.A. 91, 78–82.

MULLIGAN, R.C. (1993). The basic science of gene therapy. Science260, 926–932.

NOLTA, J.A., SMOGORZEWSKA, E., and KOHN, D. (1995). Analy-sis of optimal conditions for retroviral mediated transduction of prim-itive human hematopoietic cells. Blood 86, 101–110.

ORLIC, D., GIRARD, L., JORDAN, C., ANDERSON, S., CLINE, A.,and BODINE, A. (1996). The level of mRNA encoding the am-photropic retrovirus receptor in mouse and human hematopoieticstem cells is low and correlates with the efficiency of retrovirus trans-duction. Proc. Natl. Acad. Sci. U.S.A. 93, 11097–11102.

ORLIC, D., GIRARD, L.J., ANDERSON, S.M., PYLE, L.C., YODER,M.C., BROXMEYER, H.E., and BODINE, D.M. (1998). Identifica-tion of human and mouse hematopoietic stem cell populations ex-pressing high levels of mRNA encoding retrovirus receptors. Blood91, 3247–3254.

PETERS, S., KITTLER, E., RAMSHAW, H., and QUESENBERRY,P. (1995). Murine marrow cells expanded in culture with IL-3, IL-6, IL-11, and SCF acquire an engraftment defect in normal hosts.Exp. Hematol. 23, 461–469.

PETERS, S.O., KITTLER, E.L., RAMSHAW, H.S., and QUESEN-BERRY, P.J. (1996). Ex-vivo expansion of murine marrow cells withinterleukin-3 (IL-3), IL-6, IL-11, and stem cell factor leads to im-paired engraftment in irradiated hosts. Blood 87, 30–37.

RAMSHAW, H.S., RAO, S., CRITTENDEN, R., PETERS, S., WEIER,H., and QUESENBERRY, P. (1995). Engraftment of bone marrowcells into normal unprepared hosts: effects of 5-fluorouracil and cellcycle status. Blood 86, 924–929.

REEMS, J.A., and TOROK-STORB, B. (1995). Cell cycle and func-tional differences between CD341CD38hi and CD341CD38lo hu-man marrow cells after in vitro cytokine exposure. Blood 85, 1480–1487.

ROE, T., REYNOLDS, T.C., YU, G., and BROWN, P.O. (1993). In-tegration of murine leukemia virus DNA depends on mitosis. EMBOJ. 12, 2099–2108.

SHAH, A.J., SMOGORZEWSKA, E.M., HANNUM, C., andCROOKS, G.M. (1996). Flt-3 ligand induces proliferation of quies-cent human bone marrow CD341CD382 cells and maintains pro-genitor cells in vitro. Blood 87, 3563–3570.

TERSTAPPEN, L.W., HUANG, S., SAFFORD, M., LANSDORP, P.,and LOKEN, M. (1991). Sequential generations of hematopoieticcolonies derived from single nonlineage committed CD341CD382

progenitor cells. Blood 77, 1218–1227.THIEMANN, F.T., MOORE, K.A., SMOGORZEWSKA, E.M., LEM-

ISHKA, I.R., and CROOKS, G.M. (1998). The murine stromal cellline AFT024 acts specifically on human CD341CD382 progenitorsto maintain primitive function and immunophenotype in vitro. Exp.Hematol. 26, 612–619.

TISDALE, J.F., HANAZONO, Y., SELLERS, S., AGRICOLA, B.,METZGER, M., DONAHUE, R., and DUNBAR, C.E. (1998). Exvivo expansion of genetically marked rhesus peripheral blood pro-genitor cells results in diminished long-term repopulating ability.Blood 92, 1131–1141.

TRAYCOFF, C.M., CORNETTA, K., YODER, M.C., DAVIDSON,

GENE TRANSFER INTO BONE MARROW STEM CELLS 2527

A., and SROUR, E.F. (1996). Ex vivo expansion of murinehematopoietic progenitor cells generates classes of expanded cellspossessing different levels of bone marrow repopulating potential.Exp. Hematol. 24, 299–306.

VAN DER LOO, J.C., and PLOEMACHER, R.E. (1995). Marrow- andspleen-seeding efficiencies of all murine hematopoietic stem cell sub-sets are decreased by preincubation with hematopoietic growth fac-tors. Blood 9, 2598–2606.

VAN HENNIK, P., VERSTEGEN, M., BIERHUIZEN, M., LIMON,A., WOGNUM, A., CANCELAS, J., BARQUINERO, J.,PLOEMAC-HER, R., and WAGEMAKER, G. (1998). Highly effi-cient transduction of the green fluorescent protein gene in humanumbilical cord blood stem cells capable of cobblestone formation inlong term cultures and multilineage engraftment of immunodeficientmice. Blood 92, 4013–4022.

VARAS, F., BERNARD, A., and BUEREN, J.A. (1998). Restrictionsin the stem cell function of murine bone marrow grafts after ex vivoexpansion of short term repopulating progenitors. Exp. Hematol. 26,100–109.

VEIBY, O.P., MIKHAIL, A.A., and SNODGRASS, H.R. (1997).Growth factors and hematopoietic stem cells. Hematol. Oncol. Clin.North Am. 11, 1173–1184.

VON KALLE, C., KIEM, H., GOEHLE, S., DAROVSKY, B., HEIN-FELD, S., TOROK-STORB, B., STORB, R., and SCHEUMING, G.(1994). Increased gene transfer into human hematopoietic progeni-tor cells by extended in vitro exposure to a pseudotyped retroviralvector. Blood 84, 2890–2897.

WEEKX, S.F., VON BOCKSTAELE, D., PLUM, J., MOULIJN, A.,RODRIGUS, I., LARDON, F., DE SMEDT, M., NIHS, G.,LENJOU, M., LOQUET, P., BERNEMAN, Z., and SNOECK, H.

(1998). CD341CD382 and CD342CD382 human haematopoieticprogenitors from fetal liver, cord blood, and adult bone marrow re-spond differently to hematopoietic cytokines depending on the on-togenic source. Exp. Hematol. 26, 1034–1042.

YONEMURA, Y., KU, H., HIRAYAM, F., SOUZA, L., and OGAWA,M. (1996). Interleukin 3 or interleukin 1 abrogates the reconstitutingability of hematopoietic stem cells. Proc. Natl. Acad. Sci. U.S.A. 93,4040–4044.

ZANDSTRA, P.W., CONNEALLY, E., PETZER, A.L., PIRET, J.M.,and EAVES, C.J. (1997). Cytokine manipulation of primitive humanhematopoietic cell self-renewal. Proc. Natl. Acad. Sci. U.S.A. 94,4698–4703.

ZANDSTRA, P.W., CONNEALLY, E., PIRET, J.M., and EAVES, C.J.(1998). Ontogeny associated changes in the cytokine responses ofprimitive human haemopoietic cells. Br. J. Haematol. 101, 770–778.

Address reprint requests to:Dr. John P. Chute

Stem Cell Biology LaboratoryNIDDK-Navy Transplantation and Autoimmunity Branch

Building 46, Room 2417A8901 Wisconsin Avenue

Bethesda, MD 20889

E-mail: chutej@nmripo.nmri.nnmc.navy.mil

Received for publication June 27, 2000; accepted after revisionSeptember 29, 2000.

CHUTE ET AL.2528