A stable human-derived packaging cell line for production of high titer retrovirus/vesicular stomatitis virus G pseudotypes.

Abstract

We have generated a human 293-derived retroviral packaging cell line (293GPG) capable of producing high titers of recombinant Moloney murine leukemia virus particles that have incorporated the vesicular stomatitis virus G (VSV-G) protein. To achieve expression of the retroviral gag-pol polyprotein, the precise coding sequences for gag-pol were introduced into a vector which utilizes totally nonretroviral signals for gene expression. Because constitutive expression of the VSV-G protein is toxic in 293 cells, we used the tetR/VP 16 transactivator and teto minimal promoter system for inducible, tetracycline-regulatable expression of VSV-G. After stable transfection of the 293GPG packaging cell line with the MFG.SnlsLacZ retroviral vector construct, it was possible to readily isolate stable virus-producing cell lines with titers approaching 10(7) colony-forming units/ml. Transient transfection of 293GPG cells using a modified version of MFG.SnlsLacZ, in which the cytomegalovirus IE promoter was used to drive transcription of the proviral genome, led to titers of approximately 10(6) colony-forming units/ml. The retroviral/VSV-G pseudotypes generated using 293GPG cells were significantly more resistant to human complement than commonly used amphotropic vectors and could be highly concentrated (> 1000-fold). This new packaging cell line may prove to be particularly useful for assessing the potential use of retroviral vectors for direct in vivo gene transfer. The design of the cell line also provides at least theoretical advantages over existing cell lines with regard to the possible release of replication-competent virus.

Full text

Proc.
Natl.
Acad.
Sci.
USA
Vol.
93,
pp.
11400-11406,
October
1996
Colloquium
Paper
This
paper
was
presented
at
a
colloquium
entitled
"Genetic
Engineering
of
Viruses
and
of
Virus
Vectors,
"
organized
by
Bernard
Roizman
and
Peter
Palese
(Co-chairs),
held
June
9-11,
1996,
at
the
National
Academy
of
Sciences
in
Irvine,
CA.
A
stable
human-derived
packaging
cell
line
for
production
of
high
titer
retrovirus/vesicular
stomatitis
virus
G
pseudotypes
(retrovirus
vector/gene
therapy/293
cells/transient
transfection)
DANIEL
S.
ORY*,
BEVERLY
A.
NEUGEBORENt,
AND
RICHARD
C.
MULLIGANtt
Whitehead
Institute
for
Biomedical
Research,
and
Department
Biology,
Massachusetts
Institute
of
Technology,
Cambridge,
MA
02142
ABSTRACT
We
have
generated
a
human
293-derived
retroviral
packaging
cell
line
(293GPG)
capable
of
producing
high
titers
of
recombinant
Moloney
murine
leukemia
virus
particles
that
have
incorporated
the
vesicular
stomatitis
virus
G
(VSV-G)
protein.
To
achieve
expression
of
the
retroviral
gag-pol
polyprotein,
the
precise
coding
sequences
for
gag-pol
were
introduced
into
a
vector
which
utilizes
totally
nonretro-
viral
signals
for
gene
expression.
Because
constitutive
expres-
sion
of
the
VSV-G
protein
is
toxic
in
293
cells,
we
used
the
tetR/VP
16
transactivator
and
tet°
minimal
promoter
system
for
inducible,
tetracycline-regulatable
expression
of
VSV-G.
After
stable
transfection
of
the
293GPG
packaging
cell
line
with
the
MFG.SnlsLacZ
retroviral
vector
construct,
it
was
possible
to
readily
isolate
stable
virus-producing
cell
lines
with
titers
approaching
107
colony-forming
units/ml.
Tran-
sient
transfection
of
293GPG
cells
using
a
modified
version
of
MFG.SnlsLacZ,
in
which
the
cytomegalovirus
IE
promoter
was
used
to
drive
transcription
of
the
proviral
genome,
led
to
titers
of
_106
colony-forming
units/ml.
The
retroviral/VSV-G
pseudotypes
generated
using
293GPG
cells
were
significantly
more
resistant
to
human
complement
than
commonly
used
amphotropic
vectors
and
could
be
highly
concentrated
(>1000-fold).
This
new
packaging
cell
line
may
prove
to
be
particularly
useful
for
assessing
the
potential
use
of
retroviral
vectors
for
direct
in
vivo
gene
transfer.
The
design
of
the
cell
line
also
provides
at
least
theoretical
advantages
over
existing
cell
lines
with
regard
to
the
possible
release
of
replication-
competent
virus.
Currently,
retroviral-mediated
gene
transfer
is
widely
utilized
to
obtain
efficient
transduction
of
mammalian
cells
in
vitro,
and
to
date,
has
been
the
gene
transfer
method
of
choice
for
clinical
protocols
aimed
at
the
evaluation
of
ex
vivo
strategies
for
gene
therapy
(1).
While
standard
murine-based
retroviral
vectors
are
well
suited
for
use
in
such
ex
vivo
applications,
the
vectors
have
found
only
limited
use
in
strategies
involving
direct
in
vivo
gene
transfer
(1).
One
major
limitation
of
the
commonly
used
vector/packaging
cell
systems
is
the
inability
to
easily
purify
and
concentrate
the
large
amounts
of
virus
often
needed
for
direct
in
vivo
gene
transfer
applications.
A
second
limitation
relates
to
the
sensitivity
of
virus
with
am-
photropic
host
range
to
inactivation
by
human
serum
(2-4).
A
final
limitation
of
all
murine-based
vectors
is
their
inability
to
integrate
in
quiescent
cells
(5,
6).
One
recent
advance
that
may
prove
to
be
important
for
the
eventual
use
of
retrovirus
vectors
for
direct
in
vivo
gene
transfer
was
the
demonstration
that
it
is
possible
to
generate
retrovirus
vector
particles
which
have
incorporated
the
vesic-
ular
stomatitis
virus
G
(VSV-G)
protein
(7).
The
resulting
VSV-G/retroviral
pseudotypes
possessed
the
wide
host
range
of
VSV
and
could
be
highly
concentrated
without
loss
of
biological
activity
(8).
This
finding
follows
very
early
studies
which
had
demonstrated
the
capacity
of
retroviruses
and
VSV
to
form
viral
pseudotypes
upon
coinfection
of
cells
with
both
viruses
(9).
In
the
recent
work,
the
procedure
used
to
generate
virus
involved
the
use
of
transient
transfection
techniques
to
express
the
VSV-G
protein,
since
the
constitutive
expression
of
significant
levels
of
VSV-G
in
most
cells
is
toxic.
However,
this
method
of
virus
production
significantly
limits
the
evalu-
ation
of
the
potential
applications
of
the
viral
pseudotypes,
since
only
small
amounts
of
virus
can
be
easily
produced.
To
overcome
these
difficulties,
we
have
generated
a
stable
human-
derived
cell
line
which
constitutively
expresses
the
necessary
retroviral
proteins
for
packaging
and
provides
for
large
amounts
of
the
VSV-G
protein
by
inducible
expression.
We
describe
here
the
manner
in
which
the
cell
line
was
constructed
and
some
of
the
characteristics
of
the
virus
that
is
generated
from
the
cells.
MATERIALS
AND
METHODS
Cell
Lines
and
Drug
Selections.
Adenovirus
5-transformed
human
embryonic
kidney
293
cells
(10)
were
obtained
from
B.
Panning
(Whitehead
Institute).
The
293
cells
were
grown
in
293
growth
medium
containing
Dulbecco's
modified
eagle
medium
(DMEM)
(GIBCO/BRL),
10%
(vol/vol)
inactivated
fetal
bovine
serum
(IFS)
(Sigma),
2
mM
L-glutamine
(GIBCO/BRL),
and
50
units/ml
penicillin
and
streptomycin
(GIBCO/BRL).
Drug
selections
in
transfected
293
cells
were
performed
at
2
,ug/ml
puromycin
(Sigma),
0.3
mg/ml
G418
(GIBCO/BRL)
and
100
jig/ml
Zeocin
(Invitrogen).
All
growth
media,
except
where
noted,
was
supplemented
with
1
,ug/ml
tetracycline.
NIH
3T3
cells
(ATCC
CRL
1658)
were
grown
in
DMEM
containing
10%
(vol/vol)
calf
serum
(Sig-
ma),
and
50
units/ml
penicillin
and
streptomycin.
Mus
dunni
cells
were
grown
in
DMEM
containing
5%
(vol/vol)
calf
serum
(Sigma),
and
50
units/ml
penicillin
and
streptomycin.
Abbreviations:
VSV-G,
vesicular
stomatitis
virus
G;
CMV,
cytomeg-
alovirus;
HCMV,
human
CMV;
cfu,
colony-forming
unit;
RT,
reverse
transcriptase;
IFS,
inactivated
fetal
bovine
serum;
RCV,
replication-
competent
virus;
MuMLV,
Moloney
murine
leukemia
virus;
IE,
immediate
early.
*Present
address:
Cardiovascular
Division,
Washington
University
School
of
Medicine,
660
South
Euclid
Avenue,
St.
Louis,
MO
63110.
tPresent
address:
Howard
Hughes
Medical
Institute,
The
Children's
Hospital,
and
Department
of
Genetics,
Harvard
Medical
School,
Boston,
MA
02115.
tTo
whom
reprint
requests
should
be
sent
at
the
present
address:
Howard
Hughes
Medical
Institute,
The
Children's
Hospital,
Boston,
MA
02115.
11400
The
publication
costs
of
this
article
were
defrayed
in
part
by
page
charge
payment.
This
article
must
therefore
be
hereby
marked
"advertisement"
in
accordance
with
18
U.S.C.
§1734
solely
to
indicate
this
fact.

Proc.
Natl.
Acad.
Sci.
USA
93
(1996)
11401
Plasmid
Constructs.
The
plasmid
pBC.tTA
(see
Fig.
1)
was
constructed
from
pBC12/cytomegalovirus
(CMV)/interleu-
kin
2
(11)
by
replacement
of
the
interleukin
2
sequences
(bp
756-1439)
with
the
tet
transctivator
gene
from
pUHD10-1
(12).
To
construct
pMDtet.G
(Fig.
1),
the
1.6-kb
EcoRI
fragment
from
pSVGL
(13)
containing
the
VSV-G
gene
was
cloned
into
the
EcoRI
cloning
site
in
pMD.tet
which
is
within
exon
3
of
the
genomic
human
B3-globin
sequence.
pMDtet
was
constructed
with
a
0.47-kb
XhoI-BamHI
fragment
from
pUHC
13-3
(12),
which
contains
the
tet
operator
and
minimal
human
cytomeg-
alovirus
(HCMV)
enhancer-promoter
sequences,
a
1.34-kb
BamHI-XbaI
fragment
from
pUCMd,3s(R)S
(14)
that
in-
cludes
the
genomic
human
3-globin
sequences
from
the
BamHI
site
in
exon
2
through
690
bp
in
the
3'
untranslated
region,
and
a
3.06-kb
XbaI-XhoI
fragment
from
pSL301
(Invitrogen).
To
construct
pMD.gagpol
(see
Fig.
1),
PCR
was
performed
with
pCRIPenv-
(15)
using
the
following
pairs
of
primers:
5'-CGGAATTCATGGGCCAGACTGTTACC-3'
and
5'-
AGCAACTGGCGATAGTGG-3
',
5
'-CGGAATTCT-
TAGGGGGCCTCGCGG-3'
and
5'-ACTACATGCTGAAC-
CGGG-3'.
The
PCR
products
were
digested
with
EcoRI
and
XhoI
and
with
EcoRI
and
HindIll,
respectively,
to
generate
0.94-kb
EcoRI-XhoI
and
0.94-kb
HindIII-EcoRI
fragments.
These
fragments
were
ligated
with
the
3.3-kb
XhoI-HindIII
fragment
from
pCRIPenv-
and
with
pUC19,
which
had
been
linearized
with
EcoRI
and
calf
intestinal
phosphatase
treated,
to
produce
pUC19.gagpol.
The
5.2-kb
EcoRI
fragment
from
pUC19.gagpol
was
cloned
into
the
EcoRI
cloning
site
in
pMD
to
yield
pMD.gagpol.
pMD
was
constructed
with
the
3.1-kb
EcoRI-BamHI
fragment
from
pBC12/CMV/interleukin
2
that
includes
the
pXF3
backbone
and
HCMV
enhancer-
promoter
region
and
the
previously
described
1.34-kb
BamHI-
XbaI
fragment
derived
from
pUCMd,Bs(R)S.
The
3.1-kb
EcoRI-BamHI
and
1.34-kb
BamHI-XbaI
fragments
were
ligated
after
the
EcoRI
and
XbaI
overhangs
were
blunt-ended
by
Klenow
treatment.
The
plasmids
pJ6fQpuro
and
pJ6fQbleo
conferring
resistance
to
puromycin
and
bleomycin
(and
zeocin),
respectively,
were
kindly
provided
by
J.
Morgenstern
(16).
The
plasmid
pSV2neo
confers
resistance
to
G418
(17).
Retroviral
Vectors.
MFG.SnlsLacZ
(see
Fig.
1)
was
kindly
provided
by
0.
Danos
(18).
This
vector
is
a
derivative
of
MFG
(19)
in
which
mutations
have
been
introduced
at
nucleotides
412
(A
to
T),
429
(T
to
A),
and
631
(C
to
T)
[nucleotide
625
of
the
Moloney
murine
leukemia
virus
(MuMLV)
sequence].
These
substitutions
produce
the
sequence,
ATGGGC-
CCGGGGTAG,
thereby
preventing
expression
of
the
N-
terminal
portion
of
gag
that
would
otherwise
be
expressed
by
the
vector.
The
AU3nlslLacZ
retroviral
vector
was
constructed
by
precise
replacement
of
the
U3
region
in
the
5'
long-terminal
repeat
of
MFG.SnlsLacZ
with
the
HCMV
enhancer-promotor
(bp
-671
to
-2)
(20).
For
the
construction
of
AU3nlsLacZ,
a
701-bp
fragment
encoding
the
HCMV
promoter
was
gener-
ated
by
PCR
with
the
pMD
plasmid
as
the
template
with
the
pair
of
primers,
5'-GGGCCCAAGCTTCCCATTGCAT-
ACGTTGTATC-3'
and
5'-GGACTGGCGCCGGTTCAC-
TAAACGAGCTC-3',
creating
a
5'
HindIII
site
and
a
3'
KasI
site.
The
PCR
product
was
digested
with
HindIII
and
KasI
to
yield
a
677-bp
fragment.
The
91-bp
KasI-StyI
was
isolated
from
the
3'
long-terminal
repeat
of
MFG
(19).
The
253-bp
Styl-EagI
and
the
4994-bp
EagI-ScaI
fragments
were
isolated
from
MFG.SnlsLacZ,
and
the
backbone
for
AU3nlsLacZ
is
a
2.65-kb
HindIll-SmaI
fragment
from
pUC18.
DNA
Transfection.
Stable
transfection
of
293
cells
was
performed
by
the
calcium
phosphate
precipitation
method
(22)
with
5
,ug
pBC.tTA,
5
,ug
pMDtet.G,
and
1
,ug
pJ6flpuro.
For
all
stable
and
transient
transfections,
plasmid
DNA
was
prepared
by
double
banding
on
CsCl
density
gradients
(23).
Cells
(1.5
x
106)
were
plated
on
60-mm
dishes
in
4
ml
293
growth
media
the
night
before
transfection.
Chloroquine
(final
concentration,
25
AM)
and
tetracycline
(final
concentration,
1
,ug/ml)
were
added
to
the
media
5
min
before
transfection.
The
media
was
changed
7
h
posttransfection.
The
transfected
cells
were
plated
48
h
posttransfection
by
limiting
dilution
in
media
containing
puromycin
and
tetracycline
and
independent
clones
were
isolated.
293G
cells
were
always
grown
in
293
growth
medium
supplemented
with
tetracycline
and
puromycin
(293G
growth
medium).
Stable
transfection
of
the
293G
cells
was
performed
by
the
calcium
phosphate
precipitation
method
with
10
,tg
pMD.gagpol
linearized
with
ScaI
and
2
,tg
pSV2neo.
Cells
(2
x
106)
were
plated
on
60-mm
dishes
in
4
ml
293G
media
the
night
before
transfection.
Chloroquine
(final
concentration,
25
AM)
was
added
to
the
media
5
min
before
transfection.
The
media
was
changed
7
h
posttransfection.
The
transfected
293G
cells
were
plated
by
limiting
dilution
48
h
posttransfection
in
293G
growth
medium
supplemented
with
G418
and
indepen-
dent
clones
were
isolated.
293GPG
cells
were
grown
in
293G
growth
medium
supple-
mented
with
G418
(293GPG
medium).
Stable
transfection
of
the
293GPG
cells
with
MFG.SnlsLacZ
was
performed
by
the
calcium
phosphate
precipitation
method
with
12.5
,ug
MFG.-
SnlsLacZ
linearized
with
Asel
and
2.5
,ug
pJ6fQbleo
linearized
with
AflIII.
Cells
(4
x
106)
were
plated
on
60-mm
dishes
in
4
ml
293GPG
media
the
night
before
transfection.
The
media
was
changed
9
h
posttransfection.
The
transfected
293GPG
cells
were
plated
by
limiting
dilution
48
h
posttransfection
in
293GPG
media
supplemented
with
zeocin
and
independent
clones
were
isolated.
Transient
transfections
with
293GPG
cells
were
performed
on
60-mm
dishes
where
4-5
x
106
cells
were
plated
the
night
before
in
4
ml
293
growth
medium.
Four
micrograms
of
AU3nlsLacZ
was
diluted
into
300
,l
OptiMEM
(GIBCO/
BRL)
and
incubated
at
room
temperature
for
30
min
with
25
,ul
lipofectamine
(GIBCO/BRL)
diluted
into
300
Al
Opti-
MEM.
The
DNA-lipofectamine
mixture
was
diluted
into
2.4
ml
OptiMEM
and
layered
on-
top
of
the
293GPG
cells,
which
had
been
rinsed
30
min
before
transfection
and had
media
replaced
with
2
ml
OptiMEM.
Seven
to
8
h
posttransfection,
2
ml
293
media
was
added,
and
the
media
was
changed
at
24
h
with
2.5
ml
293
media.
The
supernatant
was
harvested
at
72
h
and
viral
titers
determined
as
described
below.
Analysis
of
VSV-G
Expression
in
Transfected
Cells.
The
pMDtet.G
and
pBC.tTA
cotransfected
clones
were
screened
for
inducible
VSV-G
expression
by
plating
each
clone
in
parallel
into
two
35-mm
tissue
culture
dishes
at
50%
conflu-
ence.
The
following
day
one
plate
was
rinsed
twice
with
2
ml
293G
media
without
tetracycline
and
maintained
in
this
media.
At
48
h
the
postnuclear
cellular
lysates
were
prepared
and
the
paired
samples
run
on
a
7.5%
SDS/PAGE
under
reducing
conditions.
The
gels
were
transferred
onto
nitrocellulose
(0.45
mm;
Schleicher
&
Schuell)
with
a
semidry
electroblotter
(Owl
Scientific,
Woburn,
MA).
Western
blot
analysis
was
performed
by
using
a
murine
monoclonal
anti-VSV-G
IgG
(Sigma)
at
a
dilution
of 1:800
and
a
peroxidase-conjugated
F(ab')2
frag-
ment
donkey
anti-mouse
IgG
(Jackson
ImmunoResearch
Lab-
oratories)
was
used
at
a
dilution
of
1:10,000.
Detection
by
chemiluminescence
was
performed
using
commercially
avail-
able
reagents
(Renaissance;
New
England
Nuclear).
Assays
For
Reverse
Transcriptase
(RT)
and
18-Galactosi-
dase
Activity.
293G
cells
transfected
with
pMD.gagpol
were
screened
for
RT
activity
in
the
culture
medium
of
subconflu-
ent
clones
growing
in
24-well
culture
dishes
as
described
(24).
Cells
were
stained
for
B3-galactosidase
activity
as
described
(25).
Viral
Titers,
Virus
Concentration,
and
Stability
of
Virus
to
Human
Serum.
To
determine
viral
titers,
NIH
3T3
cells
were
plated
at
1
x
105
cells
per
well
in
6-well
culture
dishes
16
h
before
infection
and
incubated
for
24
h
with
serial
dilutions
of
Colloquium
Paper:
Ory
et
aL

11402
Colloquium
Paper:
Ory
et
al.
viral
supernatants
containing
8
,ug/ml
polybrene
(Sigma).
Viral
titer
was
determined
as
the
average
number
of
cells
with
blue
nuclei
(13-galactosidase-producing
cells)
per
20
1-mm2
fields
(2-3
x
104
cells)
multiplied
by
a
factor
to
account
for
plate
size,
dilution
of
viral
stock,
and
division
of
target
cells
in
tissue
culture
wells.
Viral
supernatants
harvested
either
from
stable
virus-producing
cell
lines
or
transiently
transfected
cells
were
concentrated
by
ultracentrifugation
(8).
The
stability
of
the
GPGnlsLZ
pseudotyped
retrovirus
and
the
TCRIPLacZ
amphotropic
retrovirus
was
determined
in
normal
human
serum.
Twenty
microliters
virus
harvest,
which
was
diluted
1:5
in
20
mM
Hepes
buffer
(pH
7.0),
10
,ul
of
PBS,
or
10
,ul
of
Gal(al-3)galactose
(final
concentration,
10
mg/ml)
(Dextra
Laboratories,
Reading,
U.K.),
and
20
,ul
of
fresh
normal
human
serum,
heat-inactivated
human
serum,
or
IFS
were
mixed
on
ice
and
then
incubated
at
37°C
for
1
h
(26).
Heat
inactivation
of
the
human
serum
was
carried
out
at
56°C
for
1
h.
The
virus-serum
mixture
was
diluted
in
1.5
ml
DMEM
with
8
,ug/ml
polybrene.
Serial
dilutions
of
the
virus-serum
mixture
were
incubated
with
NIH
3T3
cells
in
6-well
dishes
as
described
above
to
determine
viral
titers.
Relative
titers
(per-
cent)
were
determined
for
fresh
and
heat-inactivated
human
serum
treatment
as
compared
with
IFS
treatment.
Helper
Virus
Assay.
Retroviral
stocks
were
assayed
for
replication-competent
virus
(RCV)
by
a
vector
rescue
assay
(27).
Mus
dunni
cells
with
a
stably
integrated
MFGLacZ
(MDZ),
were
plated
at
a
2
x
105
cells
per
10-cm
dish.
The
following
day
the
MDZ
cells
were
infected
with
viral
stocks
that
were
0.45
,mM
filtered
and
to
which
8
Ag/ml
polybrene
was
added.
The
media
was
changed
at
24
h
and
again
3
days
later.
The
24-h
MDZ
supernatant
was
harvested
and
passed
through
a
0.45
,uM
filter,
8
,ug/ml
polybrene
was
added,
and
the
supernatant
was
overlayed
on
naive
Mus
dunni
cells,
which
were
plated
the
day
before
at
4
x
105
cells
per
10-cm
dish.
Twenty-four
hours
later
the
supernatant
was
removed
and
the
cells
supplied
with
normal
media.
The
following
day
the
cells
were
stained
for
jB-galactosidase
activity.
The
entire
plate
was
scanned
with
a
light
microscope
(X4
phase
objective)
for
the
presence
or
absence
of
LacZ
expressing
cells
(blue
cells).
To
determine
the
sensitivity
of
the
assay,
graded
dilutions
of
a
titered
4070A
amphotropic
virus
stock
(Tektagen,
Malvern,
PA)
were
used
in
place
of
the
test
virus.
The
assay
was
shown
to
be
able
to
detect
one
particle
of
4070A
per
8
ml
supernatant.
RESULTS
General
Strategy
for
Construction
of
Packaging
Cell
Line.
Although
most
previously
described
retroviral
packaging
cell
lines
have
been
derived
from
murine
cell
lines
(15,
28-30),
we
chose
to
use
the
human-derived
cell
line
293
(10)
as
the
parental
cell
line
for
our
studies
for
several
reasons.
First,
in
contrast
to
murine
cells,
human
cells
do
not
harbor
a
large
number
of
endogenous
retroviral
genomes
nor
express
endog-
enous
viral-like
RNAs
that
may
contribute
to
the
generation
of
replication-competent
virus
through
recombination
events
involving
packaged
vector
sequences
(31-33).
In
addition,
retrovirus
vectors
produced
from
human
cells
have
been
shown
to
be
resistant
to
the
mechanisms
of
virus
inactivation
involving
natural
antibodies
and
complement
that
occur
when
virus
derived
from
murine
cells
is
exposed
to
human
serum
(2,
4).
Finally,
293
cells
can
be
transiently
transfected
at
high
efficiency,
a
property
potentially
useful
for
generating
small
quantities
of
high
titer
virus
in
a
rapid
fashion
(22).
Because
the
high-level
constitutive
expression
of
VSV-G
is
toxic
to
cells,
we
employed
the
tetracycline-regulatable
gene
expression
system
of
Gossen
and
Bujard
(12)
to
provide
for
the
inducible
expression
of
VSV-G.
The
expression
construct
used
to
express
the
VSV-G
protein,
pMDtet.G
(Fig.
1),
contains
a
minimal
CMV
immediate
early
(IE)
promoter
to
which
seven
tet
operator
sequences
(12)
are
linked
upstream,
and
an
cMv
promoter
pBC.tTA
pMDtet.G
pMD.gagpol
MFG.S.nlsLacZ
Preproinsulin
genomic
tTA
intron
poly
A
CMV
l-globin
V-globin
(tetO
)7
min
intron
VSV-G
poly
A
CMV
P-globin
P-globin
promoter
intron
gag-pol
polyA
MuLV
MuLV
LTR
SD
Tp
SA
nisLacZ
LTR
CmV
MuLV
promoter
LTR
SD
TP
SA
nlsLacZ
MuLV
LTR
AU3nIsLacZ
101
......I
FIG.
1.
Schematic
diagrams
of
plasmid
and
retroviral
constructs.
The
construction
of
pBC.tTA,
pMDtet.G,
and
pMD.gagpol
is
detailed
in
Materials
and
Methods.
The
pBC.tTA
construct
encodes
the
VP16-
tet
transactivator
fusion
protein.
In
pMDtet.G
expression
of
the
VSV-G
protein
is
under
the
control
of
the
inducible
tet°/CMV
minimal
promoter
sequences.
The
pMD.gagpol
construct
encodes
the
MuLV
gag-pol
sequences.
MFG.SnlsLacZ
is
a
replication-defective
retroviral
vector
with
splice
donor
(SD)
and
splice
acceptor
(SA)
sites
in
which
a
nuclear
localizing
LacZ
(nlsLacZ)
has
been
cloned
into
the
ATG
of
the
env
gene
(19).
AU3nlsLacZ
encodes
a
nuclear
localizing
f3-galactosidase
under
the
control
of
the
HCMV
enhancer-promoter.
intervening
sequence
and
poly(A)
site
from
the
human
f3-glo-
bin
gene.
For
expression
of
the
tet/VP16
transactivator
(12),
we
used
the
vector
pBC.tTA,
which
utilizes
the
full
CMV
IE
promoter
and
an
intervening
sequence
and
poly(A)
signal
from
the
rat
insulin
II
gene
(Fig.
1).
The
tet/VP16
transacti-
vator
binds
to
the
tet
operator
sequences
in
the
promoter
region
of
pMDtet.G
and
activates
transcription
of
VSV-G
from
the
minimal
CMV
promoter
(12).
Transcription
is
sup-
pressed
in
the
presence
of
tetracycline
and
is
activated
when
tetracycline
is
removed
from
the
media.
For
expression
of
the
MuMLV
gag-pol
sequences,
we
used
the
expression
vector
pMD,
which
employs
the
CMV
IE
promoter
and
an
intervening
sequence
and
poly(A)
site
from
the
human
13-globin
gene
for
expression
of
inserted
sequences
(Fig.
1).
Notably,
this
vector
contains
no
retroviral
sequences.
A
segment
of
the
MuMLV
genome
that
precisely
encodes
gag-pol
was
then
inserted
into
the
vector
(see
Fig.
1).
These
design
features
help
to
minimize
the
overlap
of
sequences
between
different
vectors
and
the
packaging
cell
sequences
that
can
often
occur,
and
which
have
been
previously
shown
to
contribute
to
the
formation
of
RCV
(34,
35)
In
comparison
to
pCRIPenv-,
the
construct
used
in
the
generation
of
TCRIP
cells
(15),
pMD.gagpol
generated
significantly
higher
levels
of
RT
activity
during
transient
transfection
of
293
cells
(data
not
shown).
The
overall
scheme
for
generating
the
293GPG
packaging
cells
is
shown
in
Fig.
2.
The
first
step
in
the
generation
of
the
cells
was
the
isolation
of
a
293-derived
cell
line
that
expressed
VSV-G.
Rather
than
sequentially
introduce
the
pMDtet.G
and
pBC.tTA
constructs
into
cells,
we
chose
to
simultaneously
introduce
both
constructs
via
a
tripartite
cotransfection
with
a
selectable
marker
in
the
hopes
of
providing
a
natural
selection
for
integrants
that
express
an
acceptable
level
of
VSV-G
in
the
repressed
state.
Having
identified
a
clone
of
cells
that
express
high
levels
of
VSV-G
in
the
absence
of
tetracycline
and
low
basal
levels
in
the
presence
of
tetracycline,
the
pMD.gagpol
vector
was
then
introduced
by
cotransfection
into
the
cells.
Candidate
packaging
cell
lines
were
then
screened
for
high
RT
levels
in
the
media
as
described
below.
Proc.
Natl.
Acad.
Sci.
USA
93
(1996)

Proc.
Natl.
Acad.
Sci.
USA
93
(1996)
11403
A
NGF
Tet
+
+
293
Co-transfect
pBC.tTA
pMDtet.G
1
2
3
4
5
6
7
8
Screen
inducible
VSV-G
expression
qo-4...
293GPG
FIG.
2.
Strategy
for
generation
of
pseudotyped
packaging
cell
line.
293
cells
were
cotransfected
with
pBC.tTA
and
pMDtet.G.
Clones
were
screened
by
Western
blot
analysis
for
induction
of
VSV-G
expression.
The
293G
clone
was
transfected
with
pMD.gagpol
and
clones
were
screened
by
RT
activity.
The
293GPG
clone
demonstrated
the
highest
level
of
RT
activity
and
was
used
to
generated
stable
producer
cell
lines.
Construction
of
a
293-Derived
Cell
Line
Which
Expresses
VSV-G
in
an
Inducible
Fashion.
To
generate
a
cell
line
capable
of
expressing
VSV-G,
293
cells
were
cotransfected
by
calcium
phosphate
precipitation
techniques
with
equimolar
amounts
of
pBC.tTA
and
pMDtet.G
(Fig.
1)
and
with
a
plasmid
encoding
resistance
to
puromycin
as
a
selectable
marker.
The
pBC.tTA-
and
pMDtet.G-transfected
cells
were
then
cultured
in
media
containing
puromycin
and
1
,g/ml
tetracyline
during
selection
(to
prevent
expression
of
the
VSV-G).
Seventy-two
indepen-
dent
drug-resistant
clones
were
subsequently
isolated
and
screened
by
removal
of
tetracycline
from
the
growth
medium.
Western
blot
analysis
identified
12
clones
that
exhibited
high
levels
of
VSV-G
expression
in
the
absence
of
tetracycline,
yet
low
or
no
detectable
VSV-G
expression
in
the
presence
of
tetracycline
(data
not
shown).
The
293
clone
chosen
for
further
study,
termed
293G,
demonstrated
particularly
high
levels
of
VSV-G
expression
per
mg
of
cellular
protein,
comparable
to
twice
the
amount
of
VSV-G
expressed
after
transient
trans-
fection
with
pMD.G
(Fig.
3A).
The
two
VSV-G
bands
detected
in
de-repressed
293G
cells
(Fig.
3A,
lane
3)
represent
the
completely
glycosylated
(upper
band)
and
an
incompletely
glycosylated
(lower
band)
form
of
VSV-G.
Treatment
of
the
postnuclear
cellular
lysate
from
the
293G
cells
with
N-
glycosidase
F
demonstrates
a
single
unglycosylated
VSV-G
band
(Fig.
3A,
lane
9).
The
observation
of
incomplete
glyco-
sylation
of
VSV-G
in
the
293G
cells
suggests
that
the
extremely
high
level
of
expression
of
VSV-G
may
overwhelm
the
capacity
of
the
cells
for
glycosylation
or
that
the
large
quantity
of
VSV-G
protein
may
stimulate
intracellular
recycling
with
deglycosylation
(36).
The
high-level
expression
of
VSV-G
observed
after
trans-
fection
with
the
VSV-G
expression
constructs
was
consistently
associated
with
significant
morphologic
changes
in
the
cells.
In
transiently
transfected
293
cells
(data
not
shown)
and
in
derepressed
293G
cells,
we
observed
formation
of
large
multi-
nucleated
syncytia
of
cells,
the
appearance
of
which
correlated
precisely
with
VSV-G
expression
(Fig.
3B).
The
VSV-G
pro-
tein
has
a
putative
fusagenic
domain
spanning
amino
acids
123
FIG.
3.
Induction
of
VSV-G
expression
in
293G
and
293GPG
cell
lines.
(A)
Western
blot
analysis
of
SDS/PAGE
(7.5%)
of
cell
lysates
(10
jig/lane)
from
293
cells
(lane
1),
293G
cells
(lanes
2,
3,
and
9),
and
293GPG
cells
(lanes
4,
5,
and
10).
Lanes
6,
7,
8,
and
11
show
5,ug,
10,g,
20,ug,
and
10
,ug
of
cell
lysate
from
293
cells
transiently
transfected
with
pMD.G,
respectively.
Tet
(+)
indicates
growth
in
the
presence
of
tetracycline
(1
jig/ml)
and
Tet
(-)
indicates
growth
for
48
hours
in
the
absence
of
tetracycline.
NGF
(+)
indicates
treatment
with
2
units
N-glycosidase
F
(Boehringer
Mannheim)
for
24
h
at
37°C
before
analysis.
(B)
Morphology
(x
100
field)
of
293G
(a
and
b)
and
293GPG
(c
and
d)
cells
grown
in
the
presence
of
tetracycline
(1
,Lg/ml)
(a
and
c)
and
in
the
absence
of
tetracycline
(b
and
d)
for
72
h.
Prominent
syncytia
formation
is
observed
in
293G
(b)
and
293GPG
(d)
cells
after
induction
of
VSV-G
expression.
to
137
(37),
which
facilitates
fusion
between
the
membrane
of
the
enveloped
virus
and
the
plasma
membrane
of
target
cell.
The
high-level
cell
surface
expression
of
VSV-G
in
our
tran-
sient
and
stable
cell
lines
may
promote
fusion
of
plasma
membranes
of
adjacent
cells
in
response
to
local
pH
changes
(38).
Construction
of
a
Packaging
Cell
Line
Which
Expresses
Both
VSV-G
and
MuMLVgag-pol.
To
generate
a
stable
cell
line
which
expresses
both
VSV-G
and
gag-pol,
pMD.gagpol
was
linearized
at
the
Scal
site
and
introduced
into
293G
cells
along
with
a
plasmid
encoding
resistance
to
neomycin,
using
calcium
phosphate
transfection
techniques.
Sixty-nine
G418-resistant
clones
(15)
were
isolated,
and
each
culture
supernatant
was
screened
for
the
level
of
RT
activity.
Twenty-four
positive
clones
with
RT
activity
equivalent
to
or
greater
than
that
of
TCRE
(15)
were
identified
on
an
initial
screen
(data
not
shown).
The
clone
selected
for
further
study,
293GPG,
re-
leased
-25-fold
more
RT
activity
than
that
released
by
TCRE
and
TCRIP
(15)
and
-
10-fold
more
activity
than
that
released
by
Anjou
65
cells
(22)
(Fig.
4).
Removal
of
tetracycline
from
the
growth
medium
of
293GPG
cells
demonstrated
the
con-
tinued
presence
of
inducible
VSV-G
expression
by
Western
97.4
kD
-
293G
9
10
11
69
-
Transfect
pMD.gagpol
Screen
RT
activity
46
-
B
Colloquium
Paper:
Ory
et
al.

11404
Colloquium
Paper:
Ory
et
al.
S
Blank
293
293GPG
*
293GPG
1:5
PCRE
I
1
293GPG
1:10
TCRIP
*
*
293GPG
1:25
ANJOU
65
*
293GPG
1:50
293GPG
1:100
FIG.
4.
RT
activity
of
supernatant
from
293GPG
clone.
293GPG
cells
were
grown
to
75%
confluence
in
60-mm
dishes,
the
media
was
changed,
and
supernatant
was
harvested
24
h
later.
RT
assay
was
performed
as
described.
Supematants
from
RT
reaction
cocktail
without
supernatant
(Blank)
and
293
cells
are
negative
controls.
Supernatants
from
TCRE,
PCRIP,
and
Anjou
65
cells
are
positive
controls.
Supernatant
from
the
293GPG
cells
was
diluted
1:5,
1:10,
1:25,
1:50,
and
1:100
as
indicated.
blot
analysis
(Fig.
3A)
and
by
the
formation
of
syncytia
(Fig.
3B).
Production
and
Characteristics
of
Recombinant
Virus
from
293GPG
Cells.
To
examine
the
capacity
of
the
293GPG
cells
to
produce
high
titers
of
recombinant
retrovirus
vectors,
the
cells
were
cotransfected
with
the
retroviral
vector
MFG.-
SnlsLacZ,
linearized
by
AseI,
and
a
plasmid
encoding
resis-
tance
to
zeocin
(16).
Sixteen
independent
drug-resistant
clones
were
isolated.
The
clones
were
then
cultured
in
tetracycline-
free
media
and
the
supernatants
were
harvested
at
96
h
and
were
used
to
infect
NIH
3T3
cells
for
determination
of
viral
titer
by
5-bromo-4-chloro-3-indolyl
13-D-galactoside
(X-Gal)
staining.
Three
of
the
clones,
termed
GPGnlsLZ2,
GPGnlsLZ3,
and
GPGnlsLZ4,
generated
virus
with
titers
of
_107
colony
forming
units
(cfu)/ml.
To
further
determine
the
optimal
time
for
virus
harvest
after
removal
of
tetracycline,
supernatants
from
the
three
previously
selected
clones
were
collected
serially
at
successive
time
points
(Fig.
5).
Maximal
virus
production
per
24-h
period
was
shown
to
occur
between
48
and
96
h
after
removal
of
tetracycline.
Because
293
cells
have
been
shown
to
yield
efficient
tran-
sient
gene
expression
after
transfection
(22),
we
next
examined
1X108--
F
1x107
lxi06
-c
1
xl05-
1X1041
1Xl310
the
ability
to
harvest
high
titer
virus
after
transient
transfection
of
the
cells
with
a
retroviral
vector.
Although,
as
shown
above,
cells
stably
transfected
with
the
MFG.S
vector
yielded
high
titers
of
virus,
preliminary
transient
transfection
studies
with
the
MFG.S
vector
yielded
titers
of only
_105
cfu/ml
(data
not
shown).
Based
on
the
studies
of
others
(39),
which
indicated
the
high
transcriptional
activity
of
the
CMV
IE
promoter
after
transient
transfection
of
293
cells,
we
generated
a
derivative
of
MFG.S,
termed
AU3nlslLacZ,
in
which
the
enhancer-
promoter
region
of
the
5'
long-terminal
repeat
was
replaced
with
the
complete
CMV
IE
promoter
in
such
a
way
that
transcription
would
initiate
at
the
proper
viral
start
site
(Fig.
1).
Using
this
construct
in
conjunction
with
lipofectamine
transfection,
we
were
able
to
obtain
an
average
efficiency
of
transfection
of
the
293GPG
cells
of
40%.
More
importantly,
when
virus
was
harvested
24-72
h
after
transfection,
titers
exceeding
106
cfu/ml
were
obtained
(Table
1).
To
determine
whether
virus
produced
from
293GPG
cells
could
be
efficiently
concentrated,
a
large
amount
of
culture
supernatant
was
generated
from
the
GPGnlsLZ2
and
GPGnlsLZ3
clones.
The
pseudotyped
virus,
initially
possess-
ing
titers
of
1.6
x
106
cfu/ml,
could
be
concentrated
by
ultracentrifugation
>
1000-fold
to
achieve
titers
of
>
109
cfu/ml
with
65-67%
recovery
of
the
infectious
viral
particles
(Table
1).
Virus
produced
from
the
GPGnlsLZ
cell
lines
were
exam-
ined
for
the
presence
of
RCV
using
a
sensitive
assay
involving
a
Mus
dunni
cell
line
which
harbors
a
highly
transmissible
retroviral
genome
encoding
LacZ
(21).
Helper
virus
assays
performed
on
unconcentrated
viral
supernatants
as
well
as
supernatants
from
transient
transfections
demonstrated
that
the
retroviral
stocks
generated
from
the
packaging
cells
were
free
of
RCV
(Table
1).
It
was
not
possible
to
perform
cocultivation-based
helper
virus
assays
on
the
293GPG-
derived
producer
cell
lines,
since
cocultivation
of
the
cells
with
the
Mus
dunni
indicator
cells
led
to
extensive
cell
fusion
(data
not
shown).
Finally,
because
retroviruses
produced
from
human
cells
are
known
to
be
resistant
to
inactivation
by
human
serum
(2,
4),
we
investigated
the
extent
to
which
the
viral
pseudotypes
produced
from
the
293GPG
packaging
cell
line
were
resistant
to
inactiva-
tion
by
serum.
Amphotropic
virus
from
a
TCRIPLZ
producer
clone
and
pseudotyped
virus
from
the
GPGnlsLZ4
clone
were
incubated
with
human
serum,
and
the
relative
titers
of
each
virus
stock
were
then
determined.
The
titer
of
the
amphotropic
virus
was
reduced
by
-250-fold
after
incubation
with
human
serum,
but
not
after
incubation
with
heat-inactivated
human
serum
(Fig.
6).
In
contrast,
incubation
of
the
VSV-G/retrovirus
pseudotypes
with
human
serum
resulted
in
only
a
5-fold
reduction
in
titer.
In
light
of
the
demonstrated
role
of
natural
human
antibodies
directed
against
proteins
carrying
Gal(al-3)galactose
terminal
carbohydrates
in
the
virus
inactivation
process
(3,
4),
we
next
E
GPGnIsLZ2
*
GPGnIsLZ3
3E
GPGnIsLZ4
0
0-24
24-48
48-72
72-96
96-120
120-144
144-168
168-192
Time
(hours)
FIG.
5.
Time
course
of
VSV-G
pseudotyped
virus
production
by
GPGnlsLZ
clones.
Clones
GPGnIsLZ2,
GPGnlsLZ3,
and
GPGnlsLZ4
were
grown
to
95%
confluence
in
100-mm
dishes
in
tetracycline
(1
,ug/ml).
Cells
were
rinsed
and
placed
in
5
ml
growth
media
without
tetracycline.
Supernatants
were
harvested
and
replaced
with
fresh
media
(no
tetracycline)
at
24
intervals.
Supernatants
were
titered
on
NIH
3T3
cells.
Titer
is
expressed
as
cfu/ml.
Times
indicate
the
period
after
tetracycline
removal
during
which
viral
production
was
assessed.
Proc.
Natl.
Acad.
Sci.
USA
93
(1996)

Proc.
Natl.
Acad.
Sci.
USA
93
(1996)
11405
Table
1.
Viral
titers
from
293GPG
and
GPGnlsLZ
producer
cells
Producer
cells
Retroviral
vector*
Viral
titer,
units/mlt
Helper
virus
assayt
GPGnlsLZ2§
1.2
x
107
GPGnlsLZ3
9.8
x
106
GPGnlsLZ4
8.4
x
106
Concentrated
GPGnlsLZ2
5.4
x
109
ND
Concentrated
GPGnlsLZ311
3.4
x
109
ND
293GPG
(18)**
AU3nlsLZ
2.8
x
106
293GPG
(18)
AU3nlsLZ
3.0
x
106
293GPG
(20)
AU3nlsLZ
1.1
x
106
ND
293GPG
(20)
AU3nlsLZ
2.5
x
106
ND
*nlsLacZ
virus
produced
by
transient
transfection
with
AU3nlsLZ.
tnlsLacZ
virus
titers
determined
on
NIH
3T3
cells.
tHelper
virus
assay
performed
using
Mus
dunni
LacZ
mobilization
assay.
Less
than
one
amphotropic
4070A
virus
particle
per
milliliter
was
able
to
be
detected
by
this
method.
ND,
not
determined.
§Twenty-four
hour
viral
supernatant
harvest
after
removal
of
tetracycline
from
growth
media
(72-
to
96-hr
collection
for
GPGnlsLZ2,
48-
to
72-hr
collection
for
GPGnlsLZ3,
and
72-
to
96-hr
collection
for
GPGnlsLZ4).
sUnconcentrated
titer
was
1.6
x
106,
virus
was
concentrated
>3300-fold
with
65%
total
virus
recovery.
'Unconcentrated
titer
was
1.6
x
106,
virus
was
concentrated
>3200-fold
with
67%
recovery
of
total
virus.
**Numbers
in
parentheses
indicate
cell
passage
at
time
of
transfection.
examined
the
inactivation
of
virus
exposed
to
human
serum
in
the
presence
of
an
excess
of
Gal(al-3)galactose.
In
the
case
of
amphotropic
virus,
incubation
with
10
mg/ml
Gal(al-3)galactose
completely
blocked
virus
inactivation,
while
in
the
case
of
the
VSV-G/retroviral
pseudotypes,
the
titer
was
unaffected.
(Fig.
6).
DISCUSSION
The
293GPG
packaging
cell
line
described
above
possesses
a
number
of
features
that
should
greatly
facilitate
the
further
evaluation
of
the
potential
applications
of
retroviral
vectors.
Most
importantly,
the
cell
line
makes
it
possible
both
to
generate
stable
virus-producing
cell
lines
which
produce
very
high
titers
of
VSV-G/retroviral
pseudotypes
and
to
generate
virus
rapidly
by
transient
transfection
techniques.
Because
of
the
ability
of
the
pseudotypes
to
be
efficiently
concentrated,
it
will
now
be
feasible
to
generate
the
large
amounts
of
extremely
high
titer
(>
109
cfu/ml)
virus
critical
for
examining
the
potential
utility
of
retroviral
vectors
for
in
vivo
infection
and
the
ability
to
transduce
cells
refractory
to
infection
by
standard
vectors.
While
the
virus
produced
from
293GPG
cells
is
not
fully
resistant
to
human
serum,
it
is
significantly
more
resistant
than
amphotropic
virus
generated
from
murine
cells.
Based
on
our
Gal(al-3)galactose
blocking
studies,
we
believe
that
the
100%
10%-
MO
a:
1%-
0%
E
HS
*
IHS
L
HS
+gal
GPGnIsLZ
FIG.
6.
Sensitivity
to
human
serum
of
viruses
produced
by
TCRIPLZ
and
GPGnlsLZ4
producer
cell
lines.
Viruses
produced
from
amphotropic
TCRIPLZ
producer
cells
and
VSV-G
pseudotyped
GPGnlsLZ4
producer
cells
were
incubated
for
1
h
at
37°C
with
fresh
human
serum
(HS)
and
heat-inactivated
human
serum
(IHS)
in
the
presence
and
absence
of
10
mg/ml
Gal(al-3)galactose.
Surviving
titers
of
LacZ
virus
were
determined
by
infection
of
NIH
3T3
cells.
The
serum
sensitivity
assay
was
performed
twice,
and
averaged
relative
titers
for
HS
and
IHS
treatment
versus
IFS
treatment
are
shown.
residual
sensitivity
of
the
virus
pseudotypes
to
human
serum
may
be
due
to
natural
IgM
antibodies
known
to
react
against
VSV
(26).
Cosset
et
al.
(40)
have
recently
described
another
human-derived
packaging
cell
line
capable
of
producing
re-
combinant
retrovirus
that
appears
to
be
fully
resistant
to
inactivation
by
human
serum.
In
those
studies,
however,
either
the
amphotropic
envelope
or
an
envelope
derived
from
the
virus
RD
114
(41)
was
used
rather
than
the
VSV-G
protein.
Another
potentially
useful
packaging
cell
line
for
the
produc-
tion
of
VSV
G/retroviral
pseudotypes
has
recently
been
described
by
Neinhuis
and
coworkers
(42).
That
cell
line
makes
use
of
murine
cells
and
a
strategy
for
the
regulated
expression
of
VSV-G
similar
to
the
one
we
have
employed.
As
a
consequence
of
the
use
of
the
VSV-G
protein
as
the
determinant
of
the
host
range
of
virus
produced
by
the
293GPG
cell
line,
the
packaging
cell
line
exhibits
several
somewhat
unusual
properties
relative
to
other
packaging
cell
systems.
First,
while
the
expression
of
VSV-G
in
293GPG
cells
is
tightly
regulated
by
tetracycline,
the
cells
nevertheless
express
sufficient
levels
of
VSV-G
to
yield
titers
of
103
cfu/ml
in
the
presence
of
tetracycline.
One
concern
that
has
not
yet
been
fully
evaluated
is
that
the
low
levels
of
VSV-G
expression
in
the
presence
of
tetracycline
may
lead
to
the
gradual
loss
of
packaging
function
with
the
prolonged
culture
of
the
cells
due
to
a
growth
selection
against
VSV-G-expressing
cells.
While
we
have
observed
that
viral
titers
exceeding
106
cfu/ml
can
still
be
obtained
by
transient
transfection
after
20
passages
of
the
cells,
the
prolonged
passage
of
any
working
stock
of
the
cells
should
probably
be
avoided.
A
second
novel
property
of
the
cells
is
that,
in
addition
to
spontaneously
fusing
upon
induction
of
VSV-G
expression,
the
cells
also
promote
the
fusion
of
target
cells
with
which
they
are
cocultivated.
While
this
property
of
the
packaging
cells
makes
the
use
of
cocultivation
techniques
for
achieving
the
highly
efficient
transduction
of
cells
impractical,
it
is
hopeful
that
the
ability
to
generate
highly
concentrated
virus
stocks
will
compensate
for
this
limitation.
Another
interesting
property
of
293GPG
cells
is
the
potential
susceptibility
of
stable
virus-producing
cell
lines
generated
from
the
cells
to
superinfection
by
the
virus
produced
from
the
same
cells.
Although
the
titers
of
virus
released
from
the
cells
in
the
presence
of
tetracycline
are
low,
we
do
not
yet
know
the
extent
to
which
reinfection
of
virus-producing
cells
may
occur.
Finally,
because
the
packaging
cells
are
routinely
maintained
in
media
containing
tetracycline,
puromycin,
and
G418,
other
selectable
markers
must
be
used
for
stably
introducing
vector
constructs
into
the
cells.
In
addition
to
zeocin,
we
have
also
Colloquium
Paper:
Ory
et
al.
L.

11406
Colloquium
Paper:
Ory
et
al.
successfully
used
hygromycin
and
histidinol
(43)
selections
with
the
293G
and
293GPG
cells.
Most
of
the
problems
related
to
the
generation
of
RCV
by
different
retroviral
packaging
cell
lines
can
be
practically
eliminated
through
the
use
of
specific
vectors
(34,
35)
and/or
the
practice
of
isolating
clonal
virus
producing
cell
lines
and
screening
the
cell
lines
for
helper.
Nevertheless,
several
of
the
design
features
of
the
293GPG
packaging
cell
line
may
also
provide
at
least
some
theoretical
advantages
over
existing
cell
lines
with
regard
to
the
possibility
of
release
of
RCV.
One
problem
with
all
existing
murine-based
packaging
cell
lines
is
the
presence
of
both
endogenous
retroviral
DNA
sequences
(31,
33)
and
retroviral-like
RNAs
(44)
that
are
efficiently
packaged
and
transmitted
to
cells,
and
may
contribute
to
the
generation
of
helper
virus
under
certain
conditions.
Another
issue
relates
to
the
potential
of
different
packaging
cell
lines
to
give
rise
to
helper
virus
due
to
the
overlap
of
sequences
between
the
particular
vector
used
and
the
precise
packaging
sequences
present
in
the
packaging
cell
line
(34,
35).
A
more
global
inherent
defect
in
the
design
of
all
murine
and
human-
derived
packaging
cell
lines
is
the
ability
of
retroviral
RNAs
which
lack
packaging
sequences
to
be
packaged,
albeit
at
low
efficiency,
and
transmitted
to
cells
(40).
Even
in
the
case
of
third
generation
packaging
cells
(15),
it
has
been
possible
to
observe
the
transmission
of
viral
packaging
functions
to
re-
cipient
cells.
Cosset
et
at.,
for
example,
have
recently
docu-
mented
and
quantitated
the
transfer
of
both
gag
and
env
encoding
genomes
derived
from
both
a
packaging
construct
used
to
generate
TCRIP
cells
(15)
and
a
construct
used
in
the
generation
of
the
human-derived
FLY
cell
line
(40),
another
third
generation
packaging
cell
line.
In
our
laboratory's
un-
published
studies
with
TCRE/TCRIP
cells
and
the
parental
MFG
vector,
which
contains
an
extended
gag
ORF
(19),
we
have
also
obtained data
consistent
with
the
transfer
of
pack-
aging
functions
and
the
possible
emergence
of
helper
virus
in
the
context
of
high
titer
cross-infections
employed
to
generate
complex
populations
of
virus-producing
cells.
The
design
of
the
293GPG
cells
may
be
relevant
to
each
of
the
above
issues.
In
the
construction
of
the
293GPG
cell
line,
we
have
used
only
the
precise
viral
sequences
necessary
to
encode
gag-pol
and
an
expression
vector
that
utilizes
totally
nonretroviral
sequences.
We
have
also
utilized
totally
nonret-
roviral
sequences
to
provide
for
the
host
range
of
the
virus
produced
from
the
cells
rather
than
use
conventional
retroviral
env
gene
expression
constructs.
Depending
on
the
vector
used
in
conjunction
with
293GPG
cells,
this
design
feature
of
the
cells
may
reduce
the
probability
of
undesirable
recombination
events.
More
significantly,
it
is
hopeful
that
the
removal
of
all
extraneous
viral
sequences
in
the
transcript
used
to
express
gag-pol
and
the
use
of
nonretroviral
transcripts
encoding
VSV-G
will
reduce
the
efficiency
with
which
those
transcripts
can
be
packaged
and
transmitted
to
cells,
relative
to
that
which
occurs
with
conventional
third
generation
packaging
cell
tran-
scripts.
This
hypothesis
will
need
to
be
tested
directly.
We
thank
Dr.
Jean
Schaffer
for
helpful
discussions
and
critical
review
of
this
manuscript.
D.S.O.
is
supported
by
a
National
Institutes
of
Health
Physician
Scientist
Award
(HL02910).
Support
for
this
work
was
also
provided
for
by
a
Program
of
Excellence
in
Molecular
Biology
Grant
from
the
National
Heart,
Lung,
and
Blood
Institute
(HL41484).
1.
Mulligan,
R.
C.
(1993)
Science
260,
926-932.
2.
Takeuchi,
Y.,
Cosset,
F.-L.
C.,
Lachmann,
P.
J.,
Okada,
H.,
Weiss,
R.
A.
&
Collins,
M.
K.
L.
(1994)
J.
Virol.
68,
8001-8007.
3.
Takeuchi,
Y.,
Porter,
C.
D.,
Strahan,
K.
M.,
Preece,
A.
F.,
Gustaf-
asson,
K.,
Cosset,
F.-L.,
Weiss,
R.
A.
&
Collins,
M.
K.
L.
(1996)
Nature
(London)
379,
85-88.
4.
Rother,
R.
P.,
Fodor,
W.
L.,
Springhorn,
J.
P.,
Birks,
C.
W.,
Setter,
E.,
Sandrin,
M.
S.,
Squinto,
S.
P.
&
Rollins,
S.
A.
(1995)
J.
Exp.
Med.
182,
1345-1355.
5.
Roe,
T.,
Reynolds,
T.
C.,
Yu,
G.
&
Brown,
P.
0.
(1993)
EMBO
J.
12,
2099-2108.
6.
Lewis,
P.
F.
&
Emerman,
M.
(1994)
J.
Virol.
68,
510-516.
7.
Emi,
N.,
Friedmann,
T.
&
Yee,
J.-K.
(1991)
J.
Virol.
65,
1202-
1207.
8.
Burns,
J.
C.,
Friedmann,
T.,
Driever,
W.,
Burrascano,
M.
&
Yee,
J.-K.
(1993)
Proc.
Natl.
Acad.
Sci.
USA
90,
8033-8037.
9.
Zavada,
J.
(1972)
J.
Gen.
Virol.
15,
183-191.
10.
Grahm,
F.,
Smiley,
R.
&
Nairu,
R.
(1977)
J.
Gen.
Virol.
36,
59.
11.
Cullen,
B.
R.
(1986)
Cell
46,
973-982.
12.
Gossen,
M.
&
Bujard,
H.
(1992)
Proc.
Natl.
Acad.
Sci.
USA
89,
5547-5551.
13.
Rose,
J.
K.
&
Gallione,
C.
(1981)
J.
Virol.
39,
519-528.
14.
Sadelain,
M.,
Wang,
C.
H.,
Antoniou,
M.,
Grosveld,
F.
&
Mul-
ligan,
R.
C.
(1995)
Proc.
Natl.
Acad.
Sci.
USA
92,
6728-6732.
15.
Danos,
0.
&
Mulligan,
R.
C.
(1988)
Proc.
Natl.
Acad.
Sci.
USA
85,
6460-6464.
16.
Morgenstern,
J.
P:
&
Land,
H.
(1990)
NucleicAcids
Res.
18,
1068.
17.
Southern,
P.
J.
&
Berg,
P.
(1982)
J.
Mol.
Appl.
Genet.
1,
327-341.
18.
Berns,
A.
J.,
Clift,
S.,
Cohen,
L.
K.,
Donehower,
R.
C.,
Dranoff,
G.,
Hauda,
K.
M.,
Jaffee,
E.
M.,
Lazenby,
A.
J.,
Levitsky,
H.
I.
&
Marshall,
F.
F.
(1995)
Hum.
Gene
Ther.
6,
347-368.
19.
Riviere,
I.,
Brose,
K.
&
Mulligan,
R.
C.
(1995)
Proc.
Natl.
Acad.
Sci.
USA
92,
6733-6737.
20.
Boshart,
M.,
Weber,
F.,
Jahn,
G.,
Dorsch-Hasler,
K.,
Flecken-
stein,
B.
&
Schaffner,
W.
(1985)
Cell
41,
521-530.
21.
Dranoff,
G.,
Jaffee,
E.,
Lazenby,
A.,
Golumbek,
P.,
Levitsky,
H.,
Brose,
K.,
Jackson,
V.,
Hamada,
H.,
Pardoll,
D.
&
Mulligan,
R.
C.
(1993)
Proc.
Natl.
Acad.
Sci.
USA
90,
3539-3543.
22.
Pear,
W.
S.,
Nolan,
G.
P.,
Scott,
M.
L.
&
Baltimore,
D.
(1993)
Proc.
Natl.
Acad.
Sci.
USA
90,
8392-8396.
23.
Sambrook,
J.,
Fritsch,
E.
F.
&
Maniatis,
T.
T.
(1989)
Molecular
Cloning:
A
Laboratory
Manual
(Cold
Spring
Harbor
Lab.
Press,
Plainview,
NY),
2nd
Ed.
24.
Goff,
S.,
Traktman,
P.
&
Baltimore,
D.
(1981)
J.
Virol.
38,
239-248.
25.
Lim,
K.
&
Chae,
C.-B.
(1989)
BioTechniques
7,
576-579.
26.
Beebe,
D.
P.
&
Cooper,
N.
R.
(1981)3.
Immunol.
126,1562-1568.
27.
Osbourne,
W.
R.
A.,
Hock,
R.
A.,
Kaleko,
M.
&
Miller,
A.
D.
(1990)
Hum.
Gene
Ther.
4,609-615.
28.
Miller,
A.
D.
&
Buttimore,
C.
(1986)
Mol.
Cell.
Biol.
6,
2895-
2902.
29.
Mann,
R.,
Mulligan,
R.
C.
&
Baltimore,
D.
(1983)
Cell
33,
153-159.
30.
Markowitz,
D.,
Goff,
S.
&
Bank,
A.
(1988)
J.
Virol.
62,
1120-
1124.
31.
Scadden,
D.
T.,
Fuller,
B.
&
Cunningham,
J.
M.
(1990)
J.
Virol.
64,
424-427.
32.
Cosset,
F.
L.,
Girod,
A.,
Flamant,
F.,
Dryand,
A.,
Ronfort,
C.,
Valsesia,
S.,
Molina,
R.
M.,
Faure,
C.,
Nigon,
V.
M.
&
Verdier,
G.
(1993)
Virology
193,
385-395.
33.
Purcell,
D.
F.
J.,
Broscius,
C.
M.,
Vanin,
E.
F.,
Buckler,
C.
F.,
Neinhuis,
A.
W.
&
Martin,
M.
A.
(1996)
J.
Virol.
70,
887-897.
34.
Miller,
A.
D.
&
Rosman,
G.
J.
(1989)
BioTechniques
7,
982-990.
35.
Miller,
A.
D.
(1990)
Hum.
Gene
Ther.
1,
5-14.
36.
Hebert,
D.
N.,
Foellmer,
B.
&
Helenius,
A.
(1995)
Cell
81,
425-433.
37.
Zhang,
L.
&
Ghosh,
H.
P.
(1994)
J.
Virol.
68,
2186-2193.
38.
Blumenthal,
R.
A.,
Bali-Puri,
A.,
Walter,
A.,
Corell,
D.
&
Eidel-
man,
0.
(1987)
J.
Biol.
Chem.
262,
13614-13619.
39.
Finer,
M.
H.,
Dull,
T.
J.,
Qin,
L.,
Farson,
D.
&
Roberts,
M.
R.
(1994)
Blood
83,
43-50.
40.
Cosset,
F.-L.,
Takeuchi,
Y.,
Battini,
J.-L.,
Weiss,
R.
A.
&
Collins,
M.
K.
L.
(1995)
J.
Virol.
69,
7430-7436.
41.
Reeves,
R.-
H.
&
O'Brien,
S.
J.
(1984)
J.
Virol.
52,
164-171.
42.
Yang,
Y.,
Vanin,
E.
F.,
Whitt,
M.
A.,
Fornerod,
M.,
Zwart,
R.,
Scheiderman,
R.
D.,
Grosveld,
G.
&
Neinhuis,
A.
W.
(1995)
Hum.
Gene
Ther.
6,
1203-1213.
43.
Hartman,
S.
C.
&
Mulligan,
R.
C.
(1988)
Proc.
Natl.
Acad.
Sci.
USA
85,
8047-8051.
44.
Torrent,
C.,
Bordet,
T.
&
Darlix,
J.
L.
(1994)
J.
Mol.
Biol.
240,
434-444.
Proc.
Natl.
Acad.
Sci.
USA
93
(1996)