Gene Expression in Mice After High Efficiency 
Retroviral-Mediated Gene Transfer 
Abstract. A retroviral expression vector IN2) containing the selectable gene, neo". 
has keen used to determine the optimal conditions for infecting marine hematopoiet- 
ic progenitor cells at high efficiency. After infected hone marrow cells were 
introduced into lethally irradiated mice, the presence, stability, and expression of the 
vector l)NA sequences were analyzed either in individual spleen foci 10 days later or 
in the MimhI. bone marrow, and spleens of mice 4 months later. When hone marrow 
cells were collared in mediam containing virus with tilers of more than /(/* colony- 
forming anils per milliliter in the presence of purified marine interleukin-}. more than 
85 percent of the resalting foci contained vector DNA. This proviral vector DNA was 
intact. Efficient expression of the neo * gene was demonstrated in most of the DNA- 
positive fttci examined. The spleens of reconstituted animals lover a long term) 
contained intact "vector DNA" and the blood and hone marrow expressed the neo* 
gene in some animals. Thus, a retroviral vector can he used to introduce intact 
exogenous DNA sequences into hematopoietic stem cells with high efficiency and 
with substantial expression. 
Martin A. Eg litis 
Philip Kantoff 
Laboratory of Molecular Hematology . 
National Heart. Lang, and Blood 
Institute. National Institutes of 
Health. Bethesda. Maryland 20205 
Eli Gilboa 
Department of Molecular Biology. 
Princeton University. 
Princeton. New Jersey 08544 
W. French Anderson 
luihoralory of Molecular Hematology. 
National Heart. Ijmg. and Blood 
Institute 
The life cycle of retroviruses makes 
them attractive candidates for use as 
agents for gene transfer. Several features 
are particularly relevant for their poten- 
tial use for gene therapy II). Recombi- 
nant retroviral "vectors" can be used to 
introduce new genetic material into the 
progenitor cells of the hematopoietic 
system of the mouse 12-4). Joyner et al. 
12) detected expression of a transferred 
neo K gene in individual CFU-GM colo- 
nies in vitro at an efficiency of 0.3 per- 
cent. Subsequently, Miller et al. U) 
showed transfer of a functional human 
HPRT (hypoxanihine phosphoribosyl- 
transferase) gene into hematopoietic 
stem cells that subsequently colonized 
the hematopoietic system of a whole 
mouse. Williams et al. (4). using the 
helper-free system that we utilize below, 
showed that a retroviral vector could be 
used to introduce DNA sequences con- 
taining a neo R gene into about 15 percent 
of the CFU-S. We have characterized in 
vivo an efficient new retroviral vector 
derived from Moloney murine leukemia 
virus (Mo-MLV) and have determined 
the conditions for bone marrow gene 
transfer so that more than 85 percent of 
CFU-S are infected and the transferred 
gene is expressed efficiently. 
M DECEMBER IW5 
The proviral form of the recombinant 
retrovirus N2 < Fig. I) can produce high 
tiler virus. Infection with this virus re- 
sults in turn in a stable provirus capable 
of expressing genes. Vector develop- 
ment has been described (5). A large 
portion of the Mo-MLV codirtg sequence 
has been deleted in N2 and replaced with 
the bacterial neomycin-resistance gene 
(neo R ) which confers on eukaryotic cells 
resistance to the neomycin analogue 
G4I8. After transfection into the helper- 
free cell line di2 (6) and subsequent selec- 
tion in G4I8. individual clones were 
isolated that produced N2 virus at liters 
ranging from less than 10' cfu/ml (colo- 
ny-forming units) to more than 10* 
cfu/ml. More than 50 percent of isolated 
colonies generated virus at a titer more 
than 10* cfu/ml. When NIH-3T3 cells 
infected with N2 virus produced by the 
high-titer clone F-5B were examined by 
restriction enzyme analysis and South- 
ern blotting, no evidence for deletions or 
rearrangements in the vector DNA was 
found. Furthermore, there was signifi- 
cant expression of neo R -coded phospho- 
transferase activity in these cells. 
To determine optimal time for co-culti- 
vation of bone marrow cells with the F- 
5B cells, we first established the time 
course for viral particle production. 
Soon after the F-5B cells had reached 
confluence, the medium was changed, 
and the titer was measured for the next 
72 hours. The effective liter increased 
rapidly for the first 24 hours and then 
continued to increase slowly over the 
next 48 hours. Hence, for bone marrow 
infections, fresh marrow cells were plat- 
ed onto the F-5B cells 24 hours after a 
medium change. 
The optimal period of time for co- 
cultivation of bone marrow cells with F- 
5B cells (Table I) was 24 hours. Bone 
marrow cells were also co-culturcd with 
F-5B cells in the presence or absence of 
purified growth factor interleukin-3 (IL- 
3) (7). but there was only a small benefi- 
cial effect of IL-3 (Table I). 
Since these infection efficiencies (86 
percent) are greater than those reported 
earlier (15 percent) (4). the potential 
mouse strain specificity of these results 
was investigated (Table I). A small in- 
crease in the efficiency of CFU-S infec- 
tion was observed when mice of the 
DBA/2J strain were utilized (86 percent), 
in comparison to another mouse line 
used in retroviral studies. NFS/N mice 
(74 percent). 
These results indicate that vector ti- 
lers of more than 10* cfu/ml are very 
efficient at introducing exogenous genes 
into the murine hematopoietic stem cell 
(CFU-S) population. To determine the 
efficiency of bone marrow infection with 
lower titer virus, bone marrow cells were 
co-cultured with sub-confluent plates of 
F-5B (Table I). In addition, bone mar- 
row cells were co-cultivated with indi- 
vidual ii/2 cell clones (obtained at the 
same lime as the higher titer clone F-5B) 
having lower viral tilers (Table I). No 
evidence of successful stem cell infec- 
tion was found until the titer of the virus 
in the medium was 6 x I0 4 cfu/ml or 
greater (Fig. 2). regardless of whether 
the liter was the result of diluted, high 
titer cells or the total productivity of a 
given clone. The proportion of infected 
stem cells increased as the viral titer 
increased, with efficiencies more than 80 
percent being obtained when titers were 
s2 x I0 5 cfu/ml. The effect of IL-3 was 
to slightly increase only both the overall 
proportion of CFU-S infected (Table 1 
and Fig. 2) and the average number of 
splenic foci found after infection (9.6 
versus 8.4) (Table 1). 
LTR 5' V 
LTR 
Fig. I. Diagram of the inte- 
grated vector (proviral) 
N2. 0 to t.5 and 3.0 to .3.8 
kb: Moloney murine leuke- 
mia virus sequences; 1.5 to 
1 3.0 kb box: Tn5 sequence 
containing the neo* gene 
Scale tn kitobases (Bgl 1-Bam HI fragment 
from Tn5) (ML the hatched area is the coding sequence. LTR. long terminal repeal: 5'. the 
donor splice site at the 5' end; tli. packaging sequence: restriction enzvme sites: Sa c U . Pst 
i: E. Eco Rl: X. Xho 1: C. Cla I. 
JWS 
.1 L. 
20 2 5 
[236] 
Recombinant DNA Research, Volume 12 
