H U M A N G E N E T H E R A P Y 8:1815-1824 (October 10, 1997) Mary Ann Liebert, Inc.

Selective I m m u n o a f f i n i t y - B a s e d E n r i c h m e n t o f C D 3 4 + Cells

T r a n s d u c e d w i t h Retroviral V e c t o r s C o n t a i n i n g a n

Intracytoplasmatically T r u n c a t e d V e r s i o n o f the H u m a n

L o w - A f f i n i t y N e r v e G r o w t h F a c t o r R e c e p t o r ( A L N G F R ) G e n e

BORIS FEHSE,' ALMUT UHDE,' NATALIA FEHSE,' HANS-GEORG ECKERT,^ JOHANNES CLAUSEN,' RUDIGER RUGER,3 STEFAN KOCH,^ WOLFRAM 0STERTAG,2 AXEL R. ZANDER,' and

MARCUS STOCKSCHLADER'

ABSTRACT

Human hematopoietic stem cells remain one of the most promising target cells for gene therapeutic approaches

to treat malignant and nonmalignant diseases. To rapidly characterize transduced cells and to isolate these

from residual nontransduced, but biologically equivalent, cells, we have used a Moloney murine leukemia

virus (Mo-MuLV)-based retroviral vector containing the intracytoplasmatically truncated human low-affin­

ity nerve growth factor receptor ( A L N G F R ) c D N A as a marker gene. Supernatant transduction of CD34"'"

cells (mean purity 9 7 % ) in fibronectin-coated tissue culture flasks resulted in 5.5-45% (mean 2 6 % ) trans­

duced cells expressing A L N G F R (LNGFR"*" cells). After transduction, more than 6 5 % ofthe transduced cells

remained CD34"'". Compared with control (mock- and nontransduced) CD34"'" cells, transduction did not de­

crease the cloning efficiency of CD34"'" cells. Immunomagnetic selection of the transduced cells with a mono­

clonal anti-LNGFR antibody resulted in > 9 0 % LNGFR"*" cells. Further phenotypic characterization of these

highly enriched LNGFR"*" cells indicated that the majority co-expressed the C D 3 4 and C D 3 8 antigens. These

results show that transduced cells expressing an ectopic cell-surface protein can be rapidly and conveniently

quantitated and characterized by fluorescence-activated cell sorting (FACS) analysis and fast and efficiently

enriched by immunoadhesion using magnetic beads. The use of cell-surface reporters should facilitate opti­

mization of methods of gene transfer into more primitive hematopoietic progenitors.

OVERVIEW SUMMARY isting retroviral vectors because gene transfer efficiency can be measured almost immediately after transduction.

In the present study, primary human CD34+ cells as well as TF-1 cells have been efficiently transduced using a Moloney murine leukemia virus (Mo-MLV) vector con- I N T R O D U C T I O N taining a truncated version of the human low-affinity nerve growth factor receptor gene (ALNGFR). Fast and conve- " I ^ etroviral vector-mediated gene transfer is currently nient quantitation of the transduction efficiency by fluores- JVused in many gene therapy protocols for advanced cancer, cence-activated cell sorting (FACS) analysis as well as en- human immunodeficiency vims (HIV) infection, and the treat-richment of transduced cells by immunoadhesion were ment of inherited and acquired blood diseases as well as in gene possible through the ectopically expressed transgene. Ex- marking studies (Gilboa et al, 1986; Karlsson, 1991; Anderson pression of A L N G F R in hematopoietic progenitor cells will et al, 1992; Miller, 1992). Hematopoietic stem and progenitor be a useful tool for the selection of transgene-expressing cells cells remain one of the most promising target cells because trans-and for the design of novel and further development of ex- fer and expression of therapeutic genes in these cells may result

'Bone Marrow Transplantation, University Hospital Eppendorf, Hamburg, Germany. Heinrich-Pette-Institut, Hamburg, Germany. 'Boehringer Mannheim, Penzberg, Germany.

1815

1816 FEHSE ET AL.

in transient or long-lasting clinical improvement of acquired or inherited disease (Miller, 1990; Y u et al, 1994).

For many years, the standard approach of measuring gene transfer efficiency in hematopoietic cells has been the determi­nation of the number of cells that had been transduced with vec­tor-encoded genes conferring cellular resistance to dmgs such as neomycin, hygromycin, or methotrexate (Bayever, 1990). However, this method is limited by prolonged in vitro selec­tion and nonspecific toxicity of the selection dmgs to primary cells. Other methods to determine the number of transduced cells include the polymerase chain reaction (PCR) or in vitro, ex vivo, or in situ staining with substrates for the bacterial ji-galactosidase (/3-Gal), the human placental alkaline phos­phatase (AP), or green fluorescent protein (GFP). A n altema­tive strategy has relied on the expression of transgene-encoded cell-surface antigens such as human C D 2 4 (Pawliuk et al, 1994), its murine homolog heat-stable antigen (HSA; Conneally et al, 1996; Medin et at., 1996), murine C D 2 (Champseix et al, 1996), human C D 4 ^ (Tran et al, 1995), the multidmg re­sistance gene mdr-1 (Choi et at., 1991; Ward et at., 1994; Richardson and Bank, 1995), and the full-length or intracyto­plasmatically truncated human low-affinity nerve growth fac­tor receptor ( L N G F R and A L N G F R ) (Mavilio et at., 1994; Rudoll etal, 1996).

In this study, we demonstrate efficient transduction of CD34'^ cells with a retroviral vector containing the A L N G F K c D N A . Transduced C D 3 4 ^ cells could be rapidly and accu­rately identified through fluorescence-activated cell sorting (FACS) analysis and were efficiently enriched for expression of the transgene through magnetic cell sorting. Transduction with and enrichment for A L N G F R did not lead to apparent bi­ological modifications of the target cells.

MATERIALS AND METHODS

Primaiy cells and cell lines

After informed consent, CD34+-selected human hematopoi­etic progenitor cells were obtained from healthy bone marrow donors. Mononuclear cells were separated by centrifugation on a Ficoll gradient (Biochrom, Berlin, Gennany) according to the manufacturer's instmctions and enriched for C D 3 4 ^ cells to near homogeneity as shown by F A C S analysis using the M A C S prog­enitor cell isolation kit (Miltenyi Biotec, Bergisch-Gladbach, Germany).

Amphotropic packaging cell lines were based on G P -I-envMsA\2 cells (Markowitz et al, 1988a,b) and established by Boehringer Mannheim (Boehringer Mannheim, Penzberg, Ger­many) releasing high-titer (1 X 10* cfu/ml) replication-deficient retroviral vector (LXSN-based constmct; Miller and Rosman, 1989). The retrovims contained the c D N A of A L N G F R (obtained from C. Bordignon, Milano, Italy) as well as the neomycin phos­photransferase gene (neo). In another constmct, the neo gene was removed from the vector in an effort to improve vector titer and expression of A L N G F R (Fig. 1). This second constmct will be used in a gene-marking study. Packaging cell lines producing replication-deficient retroviral vectors were kept in Dulbecco's modified Eagle's medium ( D M E M + Glutamax; GIBCO-BRL, Paisley, Scotland) supplemented with 1 0 % heat-inactivated fetal

calf semm (FCS; PAA, Linz, Austria) and sodium pymvate (f.c.

1 m M ; GIBCO-BRL). •Viral titers were estimated by the infection of Rat-2 cells with

serial 10-fold dilutions of vims-containing supematant and sub­sequent G418 selection for «ec»-containing vectors or F A C S analysis for vectors containing only A L N G F R . Rat-2 cells were kept in D M E M -I- Glutamax supplemented with 1 0 % heat-inac­tivated FCS and sodium pymvate. TF-1 cells, a C D 3 4 + cell line established by Kitamura et at. (1989) from a patient with ery­throleukemia, were cultured in R P M I (Sigma, Irvine, U K ) sup­plemented with 1 0 % heat-inactivated FCS, 1 m M sodium pym­vate, 2 m M glutamine (GIBCO-BRL), and 500 U/ml rhGM-CSF (Leucomax; Sandoz, Numberg, Germany). All cells were kept at 37°C in a humidified atmosphere of 5 % C O 2 in air.

Transduction protocol

Immediately after isolation, 5 X 10^-1 X 10* C D 3 4 + cells were transduced with retrovims-containing supematant in un-coated (Greiner, Solinger, Germany) or fibronectin-coated tis­sue culture (Falcon, Heidelberg, Germany) flasks. Supematant was fresh in most experiments and cryopreserved in some (Table 1). During transducdon, cells were kept in a - M E M (GIBCO-BRL) supplemented with 1 0 % horse s e m m (GIBCO-B R L ) , and 1 0 % F C S (PAA) containing 1 U/ml of rhuSCF (10 ng/ml), 150 U/ml of interieukin-1/3 (IL-1/3) (3 ng/ml), 20-200 U/ml of IL-3 (20-200 ng/ml) ± 100 U/ml of IL-6 (5 ng/ml) ± Epo (1 U/ml) (Recormon), and 4 /J-g/ml of protamine sulfate (Merck, Darmstadt, Germany). All growth factors were ob­tained from Boehringer Mannheim. Supematant was added at the beginning of the cultures (single transduction), again (dou­ble transduction), and again (triple transduction) every 12-24 hr before the cells were harvested for assaying colony forma­tion (CFU-GM, BFU-E) and A L N G F R expression by F A C S analysis. The viability of the cells after retroviral transduction was always higher than 9 5 % as determined by trypan blue (GIBCO-BRL) exclusion and/or propidium iodide (PI) exclu­sion (2 /xg/ml; Sigma, Deisenhofen, Germany).

Long-term bone marrow cultures and long-term culture initiating cells

After transduction, I X 10' cells were cultured on irradiated (15 Gy) bone marrow-derived stromal layers in T25 flasks in

LIR

ALNGFR

LTR

B LlR

A I NGFR

S\'

\F0

LTR

1 kb

FIG. 1. A L N G F R retroviral vectors based on L X S N (Miller and Rosman, 1989). A. Retroviral vector containing A L N G F R c D N A (obtained from C. Bordignon, Milano, Italy). B. Retro­viral vector containing the A L N G F R c D N A and the neomycin phosphotransferase c D N A (NEO) from transposon Tn5; S V in­dicates the SV40 eariy promoter.

TRANSDUCTION OF CD34- CELLS WITH RETROVIRAL VECTORS CONTAINING ALNGFR 1817

Table 1. Transduction Efficiency of CD34+ Cells with Retroviral Vectors Containing the ALNGFR cDNA ± the neo

Cell source

LP BM BM BM LP BM BM BM BM BM BM BM BM BM

CD34

purity (%)

96 N.D. 97 96 96 95 95 97 97

N.D. 97

>99 >99 N.D.

Incubation

period (hr)

40 40 40 40 40 40 40 66 66 40 64 64 64 64

Fresh supernatant

(cycles)

3 3 3 2 3 3 3

factors

S136E S136E S136E S136E S136E S136E S136E S136 S136 S13 S13 S13 S13(lo) S13(lo)

Transduction

Without F N

<l 1

<1 1 6.7 6.4 5 2 3.5 0.3 5.5

N.D. N.D. 12.7

efficiency

With F N

N.D. N.D. N.D. N.D. N.D. N.D. 15.5 26 5.5 30 29 32 44.5

Comments

neo neo neo neo

cryo cryo

cryo

»LNGFR+-cells in %. LP, Leukapheresis (GCSF-mobilized peripheral blood progenitor cells); BM, bone marrow; CD34 purity as determined by

FACS analysis; S, human recombinant stem cell factor (10 ng/ml); 1, human recombinant IL-1 (3 ng/ml); 3 human recombinant IL-3 (200 ng/ml and [lo] 20 ng/ml); 6, human recombinant IL-6 (5 ng/ml); E, human recombinant erythropoietin; transduction efficiency, the percentage of cells expressing ALNGFR (FACS analysis); FN, fibronectin; neo, neomycin phosphotransferase containing vectors; cryo, cryopreserved supematant; N.D., not done.

10 ml of Myelocult (Stem Cell Technologies, Remagen, Ger­many) supplemented with 10~^ M hydrocortisone (Sigma). Cul­tures were maintained at 33°C with weekly half-medium changes. At the time of medium change, harvested cells were stained and FACS-analyzed to determine the percentage of

ALNGFR-expressing calls. After 6 weeks, the nonadherent and adherent cells ofthe long-term bone marrow cultures (LTBMC) were harvested and plated in methylcellulose for measuring long-term culture initiating cells (LTC-IC) (Conneally et al, 1996).

NGF-07100a B NGF-07000a

Sh

I - I LU a.

•../•-<»;v vc:-: v;'-' •".• •• :.. • sy.

<'y:-•. ••- .*•'-• .'{•'•'>,'. ••'•• ':•'.:. - '/•'•• '''•'•

.:-.-.':.X.~\lfJSt- itit *,'''-'7-.'l- ••:• • -••••

i' 'I ''i Y I' I *i—I—f-i—I—I I I—r—1—I—r-800 400 600 800

SSC-HvSSC-Height >

UJ a. U.'

/ X cu -J li-O <»•

W W ' - ' - .

••-"v:>'\;^{;<:'.,"••}:'.:.-: ''

'i?. -'.. :"••-.

1000 ^ I I I I I I I I [ I I I I [ I I I I I I I I I [

'0 £00 400 600 SSC-H^SSC-Height >

800 1000

FIG. 2. Transduction efficiency is increased in the presence of FN. CD34+ cells were infected under the same conditions in regular! uncoated (A) and in FN-coated tissue culture flasks (B). The transduction efficiency was significantly increased when infection occurred in FN-coated flasks: 3 1 % versus 6% ALNGFR+ cells, respectively. Thirteen out of 35 randomly picked colonies from 14- to 17-day-old methylcellulose cultures (35%) contained ALNGFR D N A as determined by retrovims-specific nested

PCR after infection on fibronectin.

1818 FEHSE ET AL.

HQF-981002

ll>.

Cf> y. " 1 1 1 UJ <M

y-t'

(3 / /Si

cu • _l -U - o

-'.'

/\

' --^T>". V J-.J-—

J H ^ I -tSHHj^^lc, ''

-^l^.f/..

-^^^fi,K Mui^Hffi^^^-''' --^SHOHhEK^ >j .*•!.: •. • JB vZ cfiux' iu

« H P ^ ^ ' ' ' ^ B " " .>7^v;. :•

' ' 1 1 ' 1 1 ' 1

.-, • 4 6 5 :

• ,. . . • t' •'

B

'e 200 400 600 800 1000 SSC-H\SSC-Height >

NGF-081003

m

1 1 UJ 04

Qi

1 CU

B X

»'--'-'V • • < 1 • I 1 ' ^0 £00 400 600 SSC-H\SSC-Height -

NGF-081004

800 1000 ->

2

21

O

; : | p l

;^94^

'' ' • I ' '' ' 11 ' ' ' 1 ' ' '' 1 -0 £00 400 600 800 1000

SSC-H\S3C-Height >

FIG. 3. Transfer of the ALNGFR gene into CD34+ cells al­lows efficient enrichment of genetically transduced cells using immunomagnetic beads (MACS). A. Hematopoietic progenitor cells after transduction with the ALNGFR vector (46% ALNGFR + ). B. LNGFR-negative (ALNGFR") fraction after enrichment (8% of the cells remained ALNGFR-positive). C. LNGFR-positive (ALNGFR+ ) fraction after enrichment (94% of the cells were ALNGFR+ ).

F A C S analysis

F A C S analysis was performed on a F A C S c a n (Beckton Dick­inson, San Jose, C A ) . Data were analyzed using Lysis II soft­ware (Beckton Dickinson). CD34+-enriched cells were stained as previously described (Eckert et al, 1996) with a phyco­erythrin (PE)-coupled antibody (HPCA-2; Becton Dickinson, Heidelberg, Germany) not interfering with the antibody ( Q B E N D / 1 0 ) used for C D 3 4 enrichment.

Aliquots of mock-infected and ALNGFR-infected cells were incubated with an unconjugated anti-LNGFR-antibody (Boehringer Mannheim). Cells were washed twice in phos­phate-buffered saline (PBS) ( G I B C O ) and stained with a PE-coupled anti-IgGi antibody (Dako, Glostmp, Denmark). In some experiments, cells were then stained with a fluorescein isothiocyanate (FITC)-coupled anti-CD34 antibody (HPCA-2; Becton Dickinson).

After enrichment of transduced cells using the anti-LNGFR-antibody and magnetic beads coupled to an anti-lgGi antibody ( M A C S ) , cells were stained with a PE-coupled anti-lgGj anti­body (Dako) for detection of A L N G F R expression. Subse­quently, A L N G F R + and A L N G F R " fractions were double-stained with anfi-CD34-PE (Becton Dickinson) and anti-CD38-FITC (Dako).

Methylcellulose assays

One thousand and/or 3,000 hematopoietic progenitor cells were plated per dish in 1.2 ml of Iscove's modified M D M (GIBCO-BRL) containing 0.80% Methocult H4100, 5 % phythemagglutinin-leukocyte-conditioned medium (PHA-LCM; Stem Cell Technologies, Remagen, Gennany), 3 0 % FCS (PAA), 1% bovine semm albumin (BSA) (Boehringer Mannheim), 5 U/ml Epo (Boehringer Mannheim), 10""* M mer-captoethanol (Merck), and 4 m M glutamine (GIBCO-BRL), es­sentially as described (Eckert et al, 1996). CFU-GM and BFU-E were counted after 14-17 days and clones were picked for PCR analysis.

Polymerase chain reaction

Colonies randomly picked from 14- to 17-day-old methyl­cellulose cultures were washed and lysed as previously de­scribed (Eckert et at., 1996). One-tenth of the lysate was sub­jected to a nested polymerase chain reaction (PCR) in a U N O thermocycler (Biometra, Gottingen, Germany) for detecdon of retroviral LNGFR sequences using Goldstar Taq polymerase (Eurogentec, Seraing, Belgium). The primers used in the first round of amplification (annealing t°, 58°C) were BF07 (5'-GTA-GGA-GAC-GAG-AAC-CTA-AAA-CAG-3') and BF08 (5'-CCA-GAA-GCA-GCA-ACA-GCA-GCA-GGC-3'). For the second round of PCR (annealing t°, 58°C), the following primers were used: BF09 (5'-GCC-AGA-CTG-TTA-CCA-CTC-CC-3') and BFIO (5'-CGC-TCG-ACT-TCC-AGC-TCG-GT-3'). Primers BF07 and BF09 are located within the retro­viral vector sequences, whereas BF08 and BFIO represent .sequences of the ALNGFR gene. The resulting amplification product was 529 bp long. Negative colonies were analyzed for the presence of amplifiable genomic D N A using primers spe­cific for the human hematopoietic cell kinase (HCK) gene as described (Eckert et at., 1996).

TRANSDUCTION OF CD34+ CELLS WITH RETROVIRAL VECTORS CONTAINING ALNGFR 1819

RESULTS

Fibronectin increases gene transfer into CD34'^ progenitor cells

CD34"^ hematopoietic progenitor cells were isolated from human bone marrow mononuclear cells (MNC) using M A C S technology. CD34 purity after enrichment ranged from 9 3 % to 9 9 % (mean 97%) as shown by F A C S analysis (Table 1).

After incubation periods between 40 and 68 hr in the pres­ence of a minimal growth factor cocktail stem cell factor (SCF) (10 ng/ml), IL-1^ (3 ng/ml), and IL-3 20 (ng/ml), the total num­ber of hematopoietic cells increased at least 1.5-fold. More than 6 5 % of these cells remained CD34"'" after infection as shown by FACS analysis (not shown). Despite the limited expansion of cell numbers, the absolute number of CD34"'" cells after in­cubation for 40-68 hr remained approximately constant in com­parison to preincubation CD34 values.

Using the above-mentioned conditions, transduction rates as detennined by FACS analysis ranged from 0.3 to 12.7% (Table 1). Transduction efficiencies were at least 4—7 times higher (5.5^6%) in the presence of fibronectin (FN)-coated flasks (Table 1). A representative result of such an experiment is shown in Fig. 2.

Cloning efficiency of expanded cells and gene transfer efficiency into C F U s

One thousand and/or 3,000 ALNGFR-, mock-, and nonin­fected CD34"'' cells were seeded into methylcellulose-contain-ing medium to estimate the number of precursor cells (CFU). The cloning efficiency was estimated as the ratio of cells giv­ing rise to viable colonies and the number of seeded cells. Cloning efficiencies ranged from 7 % to 12%, depending on the plated cell number (3,000 and 1,000 cells, respectively). No dif­ferences in cloning efficiencies between the ALNGFR-infected cells and the two control groups (mock- and noninfected) were observed. However, the cloning efficiency after cell expansion appeared to be slightly lower than before expansion using freshly isolated CD34+ cells.

To determine the percentage of C F U containing the retro­vims-transduced A L N G F R gene, colonies were picked after 14-17 days and analyzed by retrovims-specific nested PCR. The percentage of ALNGFR-positive colonies was in agree­ment with the percentage of ALNGFR-expressing cells in the

FACS analysis (Fig. 2).

Efficient enrichment of transduced cells by ectopic cell-surface expression of A L N G F R

Hematopoietic progenitor cells transduced with A L N G F R were labeled with an anti-LNGFR antibody and incubated with a magnetic-beads-coupled anti-IgGi antibody. ALNGFR-posi­tive cells were enriched using two passages over mini-MACS columns. Both cell populations (ALNGFR+ and A L N G F R " cells) were collected and analyzed for LNGFR, CD34, and CD38 expression. Efficient enrichment of ALNGFR-express­ing cells (Fig. 3c) could be shown in four independent experi­ments (85%, 91%, 91%, 94%) whereas the percentage of ALNGFR-expressing cells in the negative fraction was repro­ducibly lower than 1 0 % (Fig. 3b). The recovery rate of AL­

NGFR-expressing cells in the A L N G F R + fraction ranged from 50 to 70%.

FACS analysis of CD34 and CD38 antigen expression showed no differences in the number of CD34'^ cells in the two cell fractions (—65%). There was, however, a striking differ­ence in the number of CD34+CD38" between the A L N G F R + and A L N G F R " cell populations. As shown in Fig. 4, nearly all of the CD34+CD38" cells were found in the A L N G F R " cell fraction.

Cells from the two fractions (LNGFR+ cells and L N G F R " cells) were plated in methylcellulose to analyze the influence of the enrichment procedure on the viability of the cells as well as on the transfer efficiency into C F U at the D N A level by PCR. No loss of colony-forming capability in comparison to un­transduced cells could be detected (see above). In particular, in the experiment shown in Fig. 3, the mean cloning efficiency (four assays) was 7.3% (60 C F U - G M and 13 BFU-E) for A L N G F R - and 7.5% (64 C F U - G M and 11 BFU-E) for the en­riched L N G F R + cells, respectively (1,000 cells/dish; Table 2). Ninety percent (16 out of 18) of the L N G F R """-derived colonies were shown to contain an integrated A L N G F R gene, whereas only one out of 17 analyzed colonies from the L N G F R " frac­tion (6%) was PCR positive (Fig. 5). PCR results are in very good agreement with the results of A L N G F R expression (FACS analysis. Fig. 3).

Long-term expression of ALNGFR in vitro

A L N G F R gene expression was monitored in five long-term bone marrow cultures of transduced primary CD34"'" cells (two B M donors). The initial gene transfer rates were 2 4 % and 37%, respectively. Within the first week, the fraction of ALNGFR-expressing cells decreased to approximately one-third of the ini­tial value (mean, 7 % and 12%, respectively). Thereafter, the relative number of ALNGFR-expressing cells remained almost constant (Fig. 6b,c). At day 28, the percentages of ALNGFR-expressing cells had decreased to 15-20% of the initial values (absolute numbers, 3.5% and 8%, respectively). PCR analysis of 100 LTC-IC-derived clones showed a gene transfer rates of about 10% at the genomic level. These data are in good agree­ment with the expression data. In a third long-term experiment with a similar initial gene transfer rate of about 35%, the per­centage of ALNGFR"^ LTC-IC was also 1 0 % (11 of 109). Long-term expression of the A L N G F R gene was also analyzed in

Table 2. Cells Enriched for the Expression of A L N G F R (see Fig. 3c) Show N o Signihcant Difference to

A L N G F R - Cells (Fig. 3b) in Their Capability to Give Rise to Hematopoietic Colonies (1,000 Cells/Dish)

Assay 1 Assay 2 Assay 3 Assay 4 Mean

AINGFR+

CFU-GM

89 78 45 44 64

cells

BFU-E

13 19 8 5 11

ALNGFR-

CFU-GM

63 73 56 49 60

cells

BFU-E

11 14 16 11 13

1820

A

FEHSE ET AL.

£:NGF-075001

16" 161 102 162 FLl-HMGG-1 Control FITC >

B £:NGF-07£004

_ TTT' 1— I '—r-

FL1-H\CD 38 FITC DAKO —

2:NGF-075004 D £:NGF-075005

t d " l6» 162 1§3 FL1-H\CD 38 FITC DAKO >

16" -ti)° l6» i'i)2 FL1-H\CD 38 FITC DAKO ->

FIG. 4. Ennchment of ALNGFR-expressing cells allows the phenotypic characterization of genetically transduced versus non­transduced cells. Different cell fractions were analyzed for the expression of lineage-specific markers of early hematopoietic nroe-emtor cells (CD34 and CD38 antigen). A. Isotype PE and FITC control antibodies. B. Neariy all of the cells used for transduc tion were CD34+, the percentage of CD34"^CD38" cells was 5 % (nonnal range). The percentage of CD34+ cells does not differ significantly between the LNGFR-negative (C) and the LNGFR-positive fraction (D). However more primitive CD34-^CD38-

rllir^Tr.lT'^ u n ' ™ ^ ' " " ^ " ' ' ^ " ^'''"°" * ' " ^ - ^' LNGFR-positive fraction contains only a very low number of CD34+CD38 cells (1%)(D).

TRANSDUCTION OF CD34- CELLS W I T H RETROVIRAL VECTORS CONTAINING A L N G F R 1821

19 38

FIG. 5. Efficient enrichment of transduced cells could also be confirmed at the D N A level. Only a small percentage (I out of 17 clones) of the CFUs from the LNGFR-negative fraction (Fig. 3B) contained the transgene (lanes 2-18), whereas nearly all (16 out of 18 clones) C F U s from the LNGFR-positive frac­tion (Fig. 3C) gave rise to a signal (expected size, 529 bp) in P C R analysis (lanes 21-38). Lanes 1 and 20, P C R controls; lane 19, a 100-bp ladder (MBI Fermentas, Vilnius, Litva).

vitro using the human C D 3 4 + cell line TF-1. After retroviral transduction, TF-1 cells were immunomagnetically enriched to a purity of 9 7 % , as shown by F A C S analysis. Expression of the transgene without selection has been stable for more than 29 weeks (69% A L N G F R + cells) (Fig. 6a).

DISCUSSION

In this study, we have tested the utility of a cell-surface anti­gen, the truncated human low-affinity nerve growth factor re­ceptor A L N G m , for the immediate postinfection determina­tion of the transduction efficiency and for the selective enrichment of retrovirally transduced primary C D 3 4 + cells as well as hematopoietic TF-1 cells.

Gene transfer into hematopoietic cells holds much promise, but is cunentiy limited by a number of issues, including the low gene transfer efficiency (Miller and Rosman, 1989; Cor­netta et at. 1991; Brenner et at., 1993; Knaan-Shanzer et al, 1996). Therefore, transduction protocols must be improved by designing optimized vectors (Baum et al, 1995) and transduc-fion conditions (Moritz etal, 1994; Bordignon etal, 1995; Ha­nenberg et al, 1996). Although recombinant retrovimses pro­vide an attractive vehicle for gene transfer, retroviral transduction of hematopoietic stem cells remains a challenge because the majority of the most primitive hematopoietic cells is in Go of the cell cycle (Springett et at., 1989; Miller et al, 1990). Strategies to increase the transduction rate of primitive hematopoietic cells include the use of a combination of growth factors (Nolta and Kohn, 1990; Heimfeld etal, 1991; Tsuji et al, 1992; Brugger et al, 1995; Nolta et al, 1995) and of com­ponents of the bone m a n o w extracellular matrix ( E C M ) such as FN, as well as chymotryptic or recombinant fragments of this molecule (Moritz et al, 1994, 1996; Verfaille et al, 1994; Hanenberg et al, 1996). W e found that short-term incubation

(40-66 hr) of hematopoietic progenitor cells in fibronectin-coated flasks with medium containing a minimal growth factor combination of human SCF, IL-1/3, and IL-3 resulted in effi­cient transduction of C D 3 4 + cells (5.5^4.5% A L N G F R ^ cells). Immunophenotypic characterization of transduced, AL­NGFR-expressing cells after enrichment through A L N G F R -based immunoadhesion revealed that most of the A L N G F R ^ cells were CD34"^CD38 + . C D 3 4 + C D 3 8 - cells, a population presumed to contain the most primitive pluripotent hematopoi­etic stem cells, were predominantly present in the L N G F R " fraction, confirming the difficulty of infecting early hematopoi­etic progenitor cells. When transduced CD34"^ cells (24% and 3 7 % L N G F R + cells, respectively) were cultured in L T B M C , the proportion of LNGFR"^ cells decreased over time (28 days). However, a significant number of cells still expressed A L N G F R after 4 weeks of culture (3.5% and 8%, respectively). Further analysis of LTC-IC by P C R showed that approximately 1 0 % of LTC-IC contained the proviral D N A . Taken together, these data suggest that the described culture conditions (growth fac­tors, fibronectin) can lead to retroviral transduction of com­mitted and progenitor cells as early as LTC-IC.

Due to the well-known shortcomings of conventional meth­ods such as d m g selection (e.g., G418), D N A blotting, or semi­quantitative P C R to determine the transduction efficiency, sev­eral groups have used retroviral vectors encoding cell-surface proteins to quantify gene transfer into hematopoietic cells. These marker proteins include the human low-affinity nerve growth factor receptor, human CD4^, the multidmg resistance gene mdr-f human CD24, the murine homolog heat-stable anti­gen, mutated murine prion protein (mPrP; Tumas et at., 1996), and murine CD2. The more recently described intracytoplas­matically tmncated version of the L N G F R ( A L N G F R ) offers several advantages. Because of its human origin and the intra-cytoplasmatic tmncation, it should neither be immunogenic nor mediate signal transduction. Our results show that quantitative determination ofthe transduction efficiency of C D 3 4 + cells by F A C S is fast (within 48 hr) and conelates well with traditional determinadon methods such as P C R analysis of single colonies (CFU-GM, BFU-E) of methylcellulose cultures, which usually takes 2-3 weeks. The rapid phenotypic characterization of transduced cells will have important implications for the design and development of novel vector constmcts and should facili­tate optimization of methods of gene transfer into more primi­tive hematopoietic progenitors. Furthermore, because of the ease and sensitivity of methods for monitoring expression of the transgene in blood and bone m a n o w cells, vector constmcts encoding proteins such as A L N G F R should also be well suited for further studies of vector modifications that may abrogate promoter shutdown in primitive hematopoietic cells and their long-term progeny (Pawliuk et al, 1994; Baum et at., 1995; Larochelle et at., 1996).

A L N G F R expression can also be used to phenotype target and progeny cells. Double-staining of peripheral blood cells has provided evidence that hematopoietic lineages were able to ex­press the transfened gene (Pawliuk et al, 1994). In the case of hematopoietic stem cells, multiparameter F A C S analysis per­mits the quantification of the gene transfer efficiency in defined subsets (e.g., C D 3 4 + C D 3 8 + and CD34"^CD38", respectively) whose phenotype is not altered by retroviral transduction by it­self (Ming et al, 1994). W e were able to confirm through this

1822 FEHSE ET AL.

. <Sr, I -I I UJ CL I r -^^ _i / X I cu

17:NGF1£9004 B NGF109006

69'>^

'I I—r • t—[—I—r '1 • -T *-\—I—I—r—I—)—r-£00 400 600 SSC-H\SSC-Height >

/\ I ' * I I UJ Q. I ^ , z: "* _) / X I cu

800 1000

•r'^^Mi^h^^'':--.. •i7•.•:![ / 5'?;••'•• ;.../•

B X

?fexl^"

' 1 ' £00 400 600

SSC-H\SSC-Height >

' I ' I ' I I 800 1000

10 15 20

days in LTBMC

25 30

F I G . 6. Long-term in vitro expression of A L N G F R . A. A L N G F R expression of T F l cells after 29 weeks in culture ( 6 9 % A L ­NGFR"*^). TFl cells were transduced, enriched (97% ALNGFR+), and kept in culture without selection. B. ALNGFR expression of transduced CD34^ cells after 3 weeks in long-term bone manow culture (LTBMC). CD34+ cells were transduced (initial transduction rate, 24%) and cultured on a preformed, inadiated (15 Gy) bone manow-derived stroma. C. Time course (28 days) of ALNGFR expression of transduced CD34"^ cells in LTBMC. Transduced CD34+ cells from two independent bone marrow donors were monitored for ALNGFR expression in five LTBMC (3 and 2 cultures, respectively).

analysis that CD34+CD38"'" cells are more effecdvely trans­duced than CD34+CD38" cells. Moreover, the ectopic expres­sion of cell-surface markers allows the selection of transduced cells (77-100%) using flow cytometry or magnetic beads

(Pawliuk el al. 1994; Conneally et al, 1996). Evidence that se­lection of transduced cells leads to an almost pure population of cells expressing the transgene (human CD24) after trans­plantation has been published (Pawliuk et al, 1994) in the

TRANSDUCTION OF CD34^ CELLS WITH RETROVIRAL VECTORS CONTAINING ALNGFR 1823

mouse model. In this study, 100% of retrovirally infected day-12 CFU-S in the sorted CD24"'" fraction gave rise to spleen colonies expressing CD24. In contrast, despite the detection of intact provims in approximately half of the CFU-S in the un­sorted or C D 2 4 " fraction, only a minority of the colonies ex­pressed detectable levels of CD24. Using immunoadhesion (mag­netic beads), we were able to enrich ALNGFR-expressing cells to a purity of greater than 90%. By using clinically approved de­vices such as the Ceprate avidin colunm (CellPro), large-scale enrichment of transduced cells will also be possible.

Although the A L N G F R marker gene and other surface mark­ers appear extremely useful for optimizing ex vivo gene trans­fer procedures in human T cells (Mavilio et al, 1994; Bonini et al, 1997) and hematopoietic stem cells (Hanenberg et al, 1996; Phillips et al, 1996), several caveats remain. First, it should be demonstrated that gene expression is stable over rel­atively long periods of time in culture. In an ongoing study with TF-1 cells, we have been able to show that the expression of the A L N G F R transgene is nearly constant (>90%) after a time period of more than 8 weeks and thereafter decreases only very slowly (69% after 29 weeks). Second, the utility of A L N G F R for following genetically modified cells in vivo in hu­man clinical trials is just now beginning to be explored. One potential concem is the possibility that A L N G F R may elicit an immune reaction that could promote the elimination of gene-modified cells. So far, there is no evidence that this will be a problem (Bonini et al, 1996, 1997). A second concem is that A L N G F R could have unintended biologic effects, such as aber­rant homing of transduced cells. Preclinical (own observation) and toxicological studies in a murine model (Boehringer Mannheim), however, have not yet shown any adverse effects. Further studies will determine whether any of these concems are warranted.

Our results show that the use of the A L N G F R cell-surface antigen as a retrovirally encoded marker allows the rapid, effi­cient, and nontoxic in vitro selection of infected primary cells, facilitates tracking and phenotyping of their progeny, and will provide a unique tool to identify elements that regulate the ex­pression of transduced genes in the most primitive hematopoi­

etic cells.

ACKNOWLEDGMENT

This work was supported by the German Bundesministerium fur Bildung und Forschung ( B M B F 01 K V 9531).

REFERENCES

ANDERSON, W.F. (1992). Human gene therapy. Science 256, 808. BAUM, C, HEGEWISCH-BECKER, S., ECKERT, H., STOCKING, C, and OSTERTAG, W. (1995). Novel retroviral vectors for effi­cient expression of the multidrug resistance (mdr 1) gene in early hematopoietic cells. J. Virol. 69, 7541-7547.

BA"yE"VER, E. (1990). Gene transfer into hematopoietic cells. Blood 75, 1587.

BONINI, C, 'VERZELETTI, S., TRAVERSARI, C, ZAPPONE, E., SERVIDA, P., ROSSINI, S., MAVILIO, P., and BORDIGNON, C.

(1996). A pilot study of HSV-Tk gene transfer into donor peripheral blood lymphocytes for controlled GvL after allo-BMT. Fourth Meet­ing of the EWGT, Leiden.

BONINI, C, FERRARI, G., VERZELETTI, S., SERVIDA, P., ZAP­PONE, E., RUGGIERI, L., PONZONI, M., ROSSINI, S., MAVILIO, F., TRAVERSARI, C, and BORDIGNON, C. (1997). HSV-TK gene transfer into donor lymphocytes for control of allogeneic graft-ver­sus-leukemia. Science 276, 1719-1724.

BORDIGNON, C, NOTARANGELO, L.D., NOBILI, N., FERRARI, G., CASORATI, G., PANINE, P., MAZZOLARI, E., MAGGIONI, D., ROSSI, C, SERVIDA, P., UGAZIO, A.G., and MAVILIO, F. (1995). Gene therapy in peripheral blood lymphocytes and bone mar­row for ADA immunodeficient patients. Science 270, 470-475.

BRENNER, M.C, RILL, D.R., HOLLADAY, M.S., HESLOP, H.E., MOEN, R.C, BUSCHLE, M., KRANCE, R.A., SANTANA, V.M., ANDERSON, W.F., and IHLE, J. (1993). Gene marking de­termines whether autologous marrow infusion restores long-term haemopoiesis in cancer patients. Lancet 342, 1134—1137.

BRUGGER, W., HEIMFELD, S., BERENSON, R., and MERTELS­MANN, R. (1995). Reconstitution of hematopoiesis after high-dose chemotherapy by autologous progenitor cells generated ex vivo. N. Engl. J. Med. 333, 283-287.

CHAMPSEIX, C, MARECHAL, V., KHAZAAL, I., SCHWARTZ, O., FOURNIER, S., SCHLEGEL, N., DRANOFF, G., DANOS, O., BLOT, P., VILMER, E., HEARD, J.M., PEAULT, B., and LEHN, P. (1996). A cell surface marker gene transferred with a retroviral vector into CD34-F cord blood cells is expressed by their T-cell prog­eny in the SCID-hu thymus. Blood 88, 107-113.

CHOI, K., FROMMEL, T., STERN, R., PEREZ, C, KRIEGLER, M., TSUSUO, T., and RONINSON, I. (1991). Multidrug resistance af­ter retroviral transfer of the human MDRl gene correlates with P-glycoprotein density in the plasma membrane and is not affected by cytotoxic selecdon. Proc. Natl. Acad. Sci. USA 88, 7386-7390.

CONNEALLY, E., BARDY, P., EAVES, C, THOMAS, T., CHAP-PEL, S., SPALL, E., and HUMPHRIES, K. (1996). Rapid and effi­cient selection of human hematopoietic cells expressing murine heat-stable antigen as an indicator of retroviral-mediated gene transfer. Blood 87, 456-464.

CORNETTA, K., MORGAN, R., and ANDERSON, F. (1991). Safety issues related to retroviral-mediated gene transfer in humans. Hum. Gene Ther. 2, 5-15.

ECKERT, H.-G., STOCKSCHLADER, M., JUST, U., HEGEWISCH-BECKER, S., GREZ, M., UHDE, A., ZANDER, A., OSTERTAG, W., and BAUM, C. (1996). High-dose multidrug resistance in pri­mary human hematopoietic progenitor cells transduced with opti­mized retroviral vectors. Blood 88, 3407-3415.

GILBOA, E., EGLITIS, M., KANTOFF, P., and ANDERSON, W.F. (1986). Transfer and expression of cloned genes using retroviral vec­tors. BioTechniques 4, 504-514.

HANENBERG, H., XIAO, L.X., DILLOO, D., HASHINO, K., KATO, I., and WILLLVMS, D.A. (1996). Colocalization of retrovirus and target cells on specific fibronectin fragments increases genetic trans­duction of mammalian cells. Nature Med. 2, 876-882.

HEIMFELD, S., HUDAK, S., WEISSMAN, I., and RENNICK, D. (1991). The in vitro response of phenotypically defined mouse stem cells and myeloerythroid progenitors to single or multiple growth factors. Proc. Natl. Acad. Sci. USA 88, 9902-9906.

KARLSSON, S. (1991). Treatment of genetic defects in hematopoietic cell function by gene transfer. Blood 78, 2481-2492.

KITAMURA, T., TANGE, T., TERASAWA, T., CHIBA, S., KUWAKI, T., MIYAGAWA, K., PIAO, Y.-F., MIYAZONE, K., URABE, A., and TAKAKU, F. (1989). Establishment and charac­terization of a unique human cell line that proliferates dependentiy on GM-CSF, 11-3, or erythropoietin (1989). J. Cell. Physiol. 140, 323-329.

1824 FEHSE ET AL.

KNAAN-SHANZER, S., VALERIO, D., and VAN BEUSECHEM, V. (1996). Cell cycle state, response to hematopoietic factors and retro­viral vector-mediated transduction of human hematopoietic stem cells. Gene Ther. 3, 323-333.

LAROCHELLE, A., VORMOOR, J., HANENBERG, H., W A N G , L, BHATIA, M., LAPIDOT, T., MORITZ, T., MURDOCH, B., XIAO, X., KATO, I., "WILLIAMS, D., and DICK, J. (1996). Identification of primitive human hematopoietic cells capable of repopulating NOD/SCID mouse bone marrow: Implications for gene therapy. Na­ture Med. 2, 1329-1337.

MARKOWITZ, D., GOFF, S., and BANK, A. (1988a). Construction and use of a safe and efficient amphotropic packaging cell line. Vi­rology 167, 400-406.

MARKO'WITZ, D., GOFF, S., and BANK, A. (1988b). A safe pack­aging cell line for gene transfer: Separating viral genes on two dif­ferent plasmids. J. Virol. 62, 1120-1124.

MA'VILIO, F., FERRARI, G., ROSSINL S., NOBILI, N., BONINI, C, CASORATI, G., TRAVERSARI, C, and BORDIGNON, C. (1994). Peripheral blood lymphocytes as target cells of retroviral vector me­diated gene transfer. Blood 83, 1988-1997.

MEDIN, J., MIGFTA, M., PAWLIUK, R., JACOBSEN, S., AMIRI, M., KLUEPFEL-STAHL, S., BRADY, R., HUMPHRIES, K., and KARLSSON, S. (1996). A bicistronic therapeutic retroviral vector enables sorting of transduced CD34"'" cells and corrects the enzyme deficiency in cells from Gaucher patients. Blood 87, 1754—1762.

MILLER, A.D. (1990). Progress toward human gene therapy. Blood 76, 271-278.

MILLER, A.D. (1992). Human gene therapy comes of age. Nature 357, 455^60.

MILLER, A.D., and R O S M A N , G. (1989). Improved retroviral vectors for gene transfer and expression. BioTechniques 7, 980-990.

MILLER, D., A D A M , M., and MILLER, A.D. (1990). Gene transfer by retrovirus vectors occurs only in the cells that are actively repli­cating at the time of infection. Mol. Cell. Biol. 10, 4239^242.

MING, L., M A R U Y A M A , M., ZHANG, N., LEVINE, F., FRIED­M A N N , T., and HO, A. (1994). High efficiency retroviral-mediated gene transduction into CD34+ cells purified from peripheral blood of breast cancer patients primed with chemotherapy and granulocyte-macrophage colony-stimulating factor. Hum. Gene Ther. 5,203-208.

MORITZ, T., PATEL, V., and VvTLLIAMS, D. (1994). Bone marrow extracellular matrix molecules improve gene transfer into human hematopoietic cells via retroviral vectors. J. Clin. Invest. 93, 1451-1457.

MORITZ, T., DLTIT, P., XIAO, X., CARSTANJEN, D., VIK, T., HA­NENBERG, H., and WILLIAMS, D.A. (1996). Fibronectin improves transduction of reconstituting hematopoietic stem cells by retroviral vectors: evidence of direct viral binding to chymotryptic carboxy-terminal fragments. Blood 88, 855-862.

NOLTA, J., and KOHN, D. (1990). Comparison ofthe effects of growth factors on retroviral vector-mediated gene transfer and the prolifer­ative status of human hematopoietic progenitor cells. Hum. Gene Ther. 1, 257-268.

NOLTA, J., S M O G O R Z E W S K A , E., and KOHN, D. (1995). Analysis

of optimal conditions for retroviral-mediated transduction of primi­tive human hematopoietic cells. Blood 86, 101-110.

PA-WLIUK, R., KAY, R., LANSDORP, P., and HUMPHRIES, R.K. (1994). Selection of retrovirally transduced hematopoietic cells us­ing CD24 as a marker of gene transfer. Blood 84, 2868-2877.

PHILLIPS, K., GENTRY, K., M c C O W A G E , G., GE^BOA, E., and SMITH, C. (1996). Cell-surface markers for assessing gene transfer into human hematopoietic cells. Namre Med. 2, 1154-1156.

RICHARDSON, C, and BANK, A. (1995). Preselection of transduced murine hematopoietic stem cell populations leads to increased long-term stability and expression of the human multiple drug resistance gene. Blood 86, 2579-2589.

RUDOLL, T., PHILLIPS, K., LEE, S.W., HULL, S., GASPAR, O., SUCGANG, N., GILBOA, E., and SMITH, C. (1996). High-effi­ciency retroviral vector mediated gene transfer into human periph­eral blood CD4-I- T lymphocytes. Gene Ther. 3, 695-705.

SPRINGETT, G., M O E N , R., ANDERSON, S., BLAESE, R., and AN­DERSON, F. (1989). Infection efficiency of T Iymphcx;ytes with am­photropic retroviral vectors is cell cycle dependent. J. Virol. 63,3865.

TRAN, A., ZHANG, D., BYRN, R., and ROBERTS, M. (1995). Chimeric f-receptors directed human natural killer effector function to permit killing of NK-resistant tumor cells and HIV-infected T lym­phocytes. J. Immun. 159, 1000-1009.

TSUJI, K., L Y M A N , S., SUDO, T., CLARK, S., and O G A W A , M. (1992). Enhancement of murine hematopoiesis by synergistic inter­actions between steel factor (ligand for c-kit), interleukin-11, and other early acting factors in culture. Blood 79, 2855-2860.

T U M A S , D., SPRANGRUDE, G., BROOKS, D., WILLIAMS, D., and CHESEBRO, V. (1996). High-frequency cell surface expression of ^ foreign protein in murine hematopoietic stem cells using a new retroviral vector. Blood 87, 509-517.

VERFAILLE, C, BENIS, A., McGLAVE, P., and McCARTHY, J. (1994). Adhesion of committed human progenitors to synthetic pep­tides from the C-terminal heparin-binding domain of fibronectin: Co­operation between the integrin alpha4betal and the CD44 adhesion receptor. Blood 84, 1802-1811.

W A R D , M., RICHARDSON, C, PIOLI, P., SMITH, L., PODDA, S., GOFF, S., HESDORFFER, and C, BANK, A. (1994). Transfer and expression of the human multiple drug resistance gene in human CD34+ cells. Blood 84, 1408-1414.

YU, M., POESCHLE, E., and WONG-STAAL, F. (1994). Progress to­wards gene therapy for HIV infection. Gene Ther 1, 13-26.

Address reprint requests to: Dr. Marcus Stockschlader

Innere Medizin I Hdmatotogie/Onkologie

Klinikum der Atbert-Ludwigs-Universitdt Freiburg Hugstetter Str. 55

79106 Freiburg

Received for publication January 28, 1997; accepted after re­vision August 6, 1997.