Advances in adenoviral vectors: from genetic engineering to their ...

mustardnimbleΒιοτεχνολογία

11 Δεκ 2012 (πριν από 4 χρόνια και 11 μήνες)

203 εμφανίσεις

0892-6638/97/0011-061
5/$01
.50
©
FASEB
615
Advances
in
adenoviral
vectors:
from
genetic
engineering
to
their
biology
PATRICE
YEW
AND
MICHEL
PERRICAUDET
CNRS
URA
1301/Rh#{244}ne-Poulenc
Rorer
Gencell,
Laboratoire
de
Genetique
des
Virus
Oncogenes,
Institut
Gustave
Roussy,
94805
Villejuif
Cedex,
France
ABSTRACT
Ad2
and
Ad5
belong
to
a
group
of
hu-
man
cytolytic
viruses
that
target
the
respiratory
air-
ways
for
reproduction,
whereas
latent
infections
establish
within
other
tissues.
Signals
therefore
exist
that
control
this
dlichotomic
process
in
different
cell
types,
perhaps
including
cis
and/or
trans
elements
of
viral
origin.
Since
1993,
Ad2-
and
Ad5-based
adeno-
viruses
lacking
all
or
part
of
the
El
regulatory
region
have
been
undergoing
evaluation
in
phase
I
trials
that
target
cancer
and
cystic
fibrosis.
These
viruses
are
extremely
attenuated
and
actually
do
not
reproduce
in
most
human
cells.
However,
they
retain
most
of
the
virus
genetic
program
and
often
promote
a
sig-
nificant
cytotoxicity
after
infection,
emphasizing
the
need
to
further
cripple
the
virus
biology
to
extend
the
duration
of
transgene
expression,
if
required.
We
will
review
the
strategies
currently
followed
to
engi-
neer
a
professional
lytic
virus
for
epithelial
cells
into
an
innocuous
gene
delivery
vehicle.
Potential
effects
on
the
transducing
properties
of
the
vector
that
may
result
from
the
inactivation
of
viral
activities
that
nor-
mally
allow/regulate
extrachromosomal
gene
expres-
sion
during
wild-type
infection
are
discussed.-Yeh,
P.,
Pemcaudet,
M.
Advances
in
adenoviral
vectors:
from
genetic
engineering
to
their
biology.
FASEBJ.
11,
615-623
(1997)
Key
Words:
adenovirus
.
viral
capsid

DNA
replication
gutless
THE
VIRUS
LIFE
CYCLE:
REPRODUCTION
VS.
LATENCY
Infection
with
subgroup
C
adenoviruses
(Ad2
and
Ad5)
is
common
in
humans,
in
whom
they
induce
a
mild,
sometimes
subclinical,
acute
disease
within
the
respiratory
airways
of
immunocompetent
individuals.
Infected
individuals
develop
neutralizing
antibodies
to
the
virus
and
retain
a
long-lasting
humoral
and
cellular
immunity
(ref
1
and
references
therein),
most
likely
because
latent
infections
occur
within
ad-
enoid
and
lymphoid
tissues
(for
a
review,
see
ref
2).
Ad2
and
Ad5
induce
a
cytopathic
productive
cycle
in
a
wide
variety
of
human
adherent
cells
in
vitro.
The
isolation
of
persistently
infected
cell
cultures
that
have
become
resistant
to
virus
infection
in
vitro
has
also
been
described
(3).
The
most
accessible
structural
component
of
the
viral
capsid
is
the
penton
capsomer,
a
heteromeric
structure
composed
of
the
penton
base
subunit,
from
which
the
fiber
protein
projects
outward
(for
a
re-
view,
see
ref
4).
This
structure
is
essential
for
virus
entry.
In
the
currently
accepted
model,
the
fiber
pro-
tein
binds
with
high
affinity
to
a
yet
unknown
cellular
receptor.
An
interaction
between
the
penton
base
and
ctv
integrins
subsequently
leads
to
endocytosis
of
the
virus
by
clathrin-coated
vesicles.
Epithelial
cells
express
both
types
of
receptors
and
constitute
natural
target
cells
in
vivo.
Infection
by
a
fiber-independent
route
may
also
occur,
as
recently
suggested
in
vitro
for
hematopoietic
cells
(5).
Viral
entry
is
associated
with
sequential
uncoating
of
the
virion
within
the
early
endosome,
a
process
that
somehow
leads
to
the
release
within
the
cyto-
plasm
of
a
nucleoparticle
capable
of
migrating
and
delivering
the
viral
chromosome
to
the
nuclear
pore
complexes
(for
a
review,
see
ref
4).
A
55
kDa
protein
of
viral
origin
(terminal
protein
[TP],
see
below)2
remains
covalently
attached
to
the
viral
genome
dur-
ing
these
processes
and
apparently
drives
the
viral
chromosome
to
specific
sites
of
the
nuclear
matrix,
where
viral
transcription
is
initiated
(6,
7).
Expres-
sion
of
the
virus
genome
is
a
timely
(see
below)
and
spatially
organized
process:
snRNPs
and
nuclear
in-
clusion
bodies
are
reorganized,
the
host
chromatin
and
the
nucleolus
are
“pushed”
against
the
nuclear
border,
and
critical
host
proteins
are
concentrated
to
novel
(virus-induced)
nuclear
inclusion
bodies
(e.g.,
NF-1,
a
factor
required
for
the
initiation
of
viral
DNA
replication)
and
fibrous
structures
in
conjunction
‘Correspondence:
CNRS
URA
1301/Rh#{244}ne-Poulenc
Rorer
Gencell,
Laboratoire
de
Genetique
des
Virus
Oncogenes,
In-
stitut
Gustave
Roussy,
39
rue
Camille
Desmoulins,
94805
Vil-
lejuif
Cedex,
France.
2Abbreviations:
LTU,
late
transcription
unit;
ITR,
inverted
terminal
repeats;
RCA,
replication-competent
adenoviruses;
AdPol,
a
140
kDa
DNA
polymera.se;
gpl9K,
a
19
kDa
glyco-
protein;
TNF,
tumor
necrosis
factor;
LTR,
long-terminal
re-
peat;
RSV,
Rous
sarcoma
virus.
A
p1X
Va2
E1A
EIB
E2B
E2A
40
50
0
20
30
B
E3A
E3B
E4
60
70
60
90
100
LTU
late
primary
transcript
(28
kb)
123
Li
I
POST-TRANSCRIPTiONAL
ISATURATION
L2
L3
L4
L5
Late
mRNAs
:::
616
Vol.
11
July
1997
The
FASEB
Journal
YEH
AND
PERRICAUDET
with
virus-encoded
regulatory
proteins
(see
refs
8,
9
and
references
therein).
The
Ad5
chromosome
is
a
linear
double-stranded
DNA
molecule
of
approximately
36
kb.
It
is
con-
densed
inside
the
capsid
in
a
chromatin-like
structure
that
results
from
its
association
with
a
set
of
viral
polypeptides
that
are
synthesized
during
the
late
phase
of
infection.
A
higher
level
of
organiza-
tion
into
eight
supercoiled
domains
is
also
apparent
within
the
capsid,
perhaps
timing
expression
of
the
virus
genetic
program
after
nucleus
entry
(for
a
re-
view,
see
ref
4).
The
viral
chromosome
encodes
eight
p0111-dependent
transcription
units,
which
are
conventionally
referred
to
as
early,
intermediate,
and
late
depending
on
their
timely
expression
with
regard
to
the
onset
of
viral
DNA
replication.
There
are
five
early
coding
regions
(E1A,
E1B,
E2,
E3,
and
E4),
two
early
delayed
(intermediate)
transcription
units
(pIX
and
Wa2),
and
a
late
transcription
unit
(LTU)
that
mostly
encodes
structural
proteins
for
the
capsid
and
the
internal
core.
An
additional
re-
gion
is
transcribed
by
RNA
polymerase
III
and
en-
codes
two
small
virus-associated
RNAs
that
bind
to
and
inactivate
a
cellular
protein
kinase
(PKR),
which
disables
translation
initiation
factor
eIF2
after
infection
(for
a
review,
see
ref
4).
One
hundred
and
three
base
pair
inverted
terminal
repeats
(ITR)
at
the
ends
of
the
chromosome
function
as
replication
origins.
The
so-called
P
sequence
at
the
immediate
vicinity
of
the
left
ITR
is
another
nondispensable
cis-
acting
sequence,
as
it
is
required
during
packaging
of
the
progeny
chromosomes
during
the
terminal
phase
of
infection
(Fig.
1).
Viral
reproduction
is
a
highly
orchestrated
process
that
occurs
within
dedicated,
viral-induced
subnu-
clear
compartments.
Transcription
from
the
viral
chromosome
has
been
detected
within
specific
struc-
tures
that
are
also
the
sites
for
viral
DNA
replication
(see
ref
8
and
references
therein).
At
the
molecular
level,
completion
of
the
infectious
cycle
relies
on
the
timely
expression
of
a
set
of
regulatory
proteins
that
will
progressively
subvert
endogenous
pathways
criti-
cal
for
cell
viability,
viral
DNA
replication
and
ex-
pression,
and
productivity.
As
mentioned,
adenoviruses
target
quiescent
(epithelial)
cells
for
re-
production.
Accordingly,
E1A
encodes
oncogenes
whose
expression
will
drive
the
cell
into
the
S-phase
of
the
cell
cycle,
setting
the
stage
for
efficient
multi-
plication.
Unprogrammed
entry
into
the
S-phase
is
a
conflictual
signal
to
which
mammalian
cells
respond
by
undergoing
p53-dependent
and
p53-independent
apoptosis.
Regulatory
proteins
from
regions
E1B
and
E3
are
likely
to
counteract
premature
cell
death
in
vivo,
whereas
others
will
ensure
viral
DNA
replication
(E2),
preferential
synthesis
of
the
virion
structural
proteins
(E4),
packaging,
and
eventually
release
of
the
progeny
virions
(for
a
review,
see
ref
4;
see
also
ref
10).
Figure
1.
Schematic
organization
of
the
Ad5
chromosome.
A)
The
viral
chromosome
is
nearly
36
kb
long
and
conventionally
divided
into
100
map
units.
There
are
eight
polli-dependent
transcription
units:
except
pIX,
they
all
are
alternatively
spliced
(see
panel
Bfor
an
example).
Arrows
refer
to
the
pri-
mary
transcripts.
The
viral
transcription
units
tend
to
encode
functionally
related
proteins:
EIA
encodes
apoptotic
genes
and
E1B
expresses
anti-apoptotic
activities.
Region
E2
(E2A
and
E2B)
encodes
proteins
required
during
replication
of
the
viral
DNA,
E3
encodes
modulators
of
the
host
response,
and
E4
is
required
for
late
gene
expression
(see
text).
Most
struc-
tural
proteins
of
the
capsid
and
the
internal
core
are
encoded
within
the
late
transcription
unit
(LTU).
The
black
boxes
rep-
resent
the
viral
ITR
required
for
the
replication
process.
‘F
refers
to
the
packaging
sequence.
B)
The
28
kb
long
late
pri-
mary
transcript
is
processed
by
alternative
polyadenylation
and
splicing.
The
overall
process
is
dependent
on
regulatory
proteins
from
region
E4.
All
mRNAs
exhibit
the
same
5’
end
(tripartite
leader),
a
structure
that
facilitates
their
selective
CAP-independent
translation
in
the
late
phase
of
infection
(for
a
recent
review,
see
ref
4).
PRODUCTION
OF
CLINICAL-GRADE
RECOMBINANT
ADENOVIRUSES
Since
1993,
human
subgroup
C
adenoviruses
lacking
all
or
part
of
the
E1A
and
E1B
regions
have
been
evaluated
in
phase
I
trials
that
target
cancer
and
cystic
fibrosis.
Viruses
lacking
El
are
extremely
attenuated:
most
human
cells
are
not
permissive
anymore
unless
the
missing
genes
are
provided
in
trans,
or
if
infection
takes
place
within
appropriate
cellular
backgrounds
(including
p53-deficient;
see
below).
Efficient
ampli-
fication
of
El-deleted
adenoviruses
requires
that
the
E1A
and
E1B
genes
be
provided
in
trans
during
the
production
process.
Whereas
expression
of
these
genes
can
lead
to
cellular
transformation
in
rodent
cells,
their
successful
coexpression
in
a
human
cel-
lular
context
is
obviously
a
rare
event,
most
likely
be-
cause
a
subtle
balance
of
E1A
(apoptotic)
and
E1B
(anti-apoptotic)
gene
expression
is
required
to
allow
full
transformation
and/or
because
an
appropriate
status
of
the
cell
cycle
regulators
these
viral
onco-
genes
target
(e.g.,
RB,
p53)
may
be
mandatory.
Ac-
_._,.
E1A
E1B
3
HOMOLOGOUS
RECOMBINATJON
El
-deleted
virus
RCA
(Ei)
E3
,5E4
-
\/_
-
E1A
E1B
pIX
IVa2
B
:t.
--
C
Iva2
LTU
E2A+E26
E3
pIX
E4
LTU
ADENOVIRAL
VECTOR
ADVANCES
617
cordingly,
the
293-derived
cell
line
has
constituted
for
a
long
time
the
unique
example
of
human
pri-
mary
cells
successfully
transformed
by
the
El
genes
of
Ad5
(11).
This
cell
line
also
efficiently
transcom-
piements
the
E1A
and
E1B
genes
and
is
used
to
am-
plify
El-deleted
adenoviruses
to
titers
up
to
1011
infectious
particles/mi.
El-deleted
adenoviruses
are
mostly
constructed
by
homologous
recombination
between
overlapping
fragments
transfected
into
293
cells
(for
a
review,
see
ref
12).
This
nonclonal
process
is
time-consuming
as
it
requires
severai
rounds
of
plaque
purification
onto
293-derived
monolayers.
Alternative
strategies
have
thus
been
developed
to
recover
the
recombinant
vi-
rus
in
a
more
straightforward
and
clonal
fashion.
For
example,
the
Ad2
genome
was
first
introduced
in
a
yeast
artificial
chromosome
from
which
infectious
re-
combinant
genomes
were
generated
by
recombina-
tional
cloning
in
Saccharomyces
cerevisiae
(13).
Gene
repiacement
procedures
have
also
been
developed
to
construct
infectious
recombinant
backbones
by
ho-
mologous
recombination
in
Escherichia
coli
(14,
15).
The
recombinant
backbone
can
be
truely
cloned
in
these
microorganisms,
thereby
only
generating
the
corresponding
virus
upon
transfection
in
293
cells.
Multiple
modifications
all
along
the
viral
backbone
were
also
introduced
by
consecutive
rounds
of
re-
combinational
cloning
in
E.
coli,
with
no
need
to
con-
struct
and
isolate
intermediate
viruses
(15).
The
293
cell
line
exhibits
a
large
fragment
from
the
left
end
of
the
Ad5
chromosome
that
includes
the
left
ITR,
the
packaging
sequence,
the
totality
of
regions
E1A
and
E1B/pIX,
as
well
as
the
distal
part
of
the
Wa2
gene
(11,
16).
The
presence
of
identical
sequences
within
the
293
cell
line
and
the
viral
chro-
mosomes
allows
reintroduction
of
the
El
genes
back
into
the
viral
backbones
by
homologous
recombina-
tion
during
batch
preparations
(Fig.
2),
leading
to
the
emergence
of
replication-competent
adenovi-
ruses
(RCA)
that
will
progressively
outgrow
the
re-
combinant
virus
during
propagation.
Because
of
the
cytolytic
nature
of
adenoviruses,
an
important
issue
is
the
production
of
clinical
batches
that
are
not
con-
taminated
with
RCA.
Three
directions
are
being
fol-
lowed
to
solve
this
issue.
For
example,
A549-derived
cell
lines
have
been
engineered
that
express
minimal
El
sequences
of
AdS
from
a
heterologous
promoter
(17,
18).
RCA
emergence
by
homologous
recombi-
nation
has
been
rendered
impossible
in
this
setting
as
there
is
no
recombinogenic
sequences
on
the
left
side
of
the
El
deletion
and
because
the
ITR/P
se-
quences
have
not
been
integrated.
The
transcomple-
mentation
potential
of
A549-derived
packaging
cell
lines
may
be
lower
than
that
of
293,
as
suggested
in
one
study
by
a
nearly
10-fold
reduction
in
productiv-
ity
(17).
The
remodeling
of
the
vector
backbone
pro-
vides
another
way
of
reducing
the
emergence
of
RCA
while
maintaining
the
productivity
of
293
cells
(12,
transgene
pIX
lVa2
A
293
I--
Figure
2.
Backbone
engineering
to
lower
RCA
emergence
dur-
ing
viral
propagation
in
293.
A)
The
emergence
of
RCAs
arises
by
homologous
recombination
between
identical
sequences
(hatched
box)
in
the
293
chromosome
and
the
viral
back-
bones.
EIA
and
E1B
sequences
are
represented
by
a
black
box.
B)
The
relocation
of
the
E4
sequences
in
place
ofEl
generates
an
E1E4
(defective)
virus
after
homologous
recombination
between
the
cellular
and
viral
chromosomes.
E4
refers
to
a
deletion
encompassing
both
E4
ORF3
and
E4
ORF6
(see
text).
C)
Remodeling
the
backbone
of
El-deleted
adenovi-
ruses
by
relocating
the
pIX
sequences
between
the
L5
and
E4
regions
(16)
1)
reduces
RCA
frequency
by
spliting
the
recom-
binogenic
sequences
in
the
backbone
and
2)
generates
a
de-
fective
E1E4
adenovirus
upon
homologous
recombination
between
cellular
and
viral
pIX
sequences.
16).
For
example,
it
is
possible
to
reposition
the
non-
dispensable
E4
region,
which
normally
is
located
next
to
the
right
ITR,
at
the
immediate
vicinity
of
the
left/P
sequences
so
an
ElE4
defective
virus
emerges
after
homologous
recombination
between
the
cellular
and
the
viral
genomes
(Fig.
2).
The
re-
location
of
the
pIX
gene
between
the
fiber
(L5)
and
the
E4
gene
also
diminishes
RCA
emergence,
partic-
ularly
because
the
recombinogenic
sequences
have
been
split
within
the
backbone
(16).
The
design
of
E1/E4
doubly
defective
adenoviruses
provides
an
ad-
ditional
way
of
reducing
RCA
emergence
(see
below).
BIOLOGY
OF
El-DELETED
ADENOVIRUSES
El-deleted
adenoviruses
retain
most
(80%)
of
the
vi-
ral
genetic
information
(Fig.
1).
Significant
early
and
late
gene
expression
occurs
after
infection
of
estab-
lished
cell
lines
of
human
origin
(ref
19
and
refer-
ences
therein).
A
bona
fide
replication
of
the
vector
backbone
is
also
apparent:
for
example,
a
25-
(e.g.,
HeLa,
KB,
Hep3B)
to
60-fold
(HepG2)
replication
has
been
observed
as
early
as
3
days
postinfection
(p.i.)
in
conditions
that
transduced
70
to
90%
of
the
618
Vol.
11
July
1997
The
FASEB
Journal
YEH
AND
PERRICAUDET
cells.
In
these
conditions,
infection
of
most
cell
lines
of
human
origin
is
associated
with
phenotypic
abnor-
malities,
including
cell
detachment,
enlargement,
rounding
and
clumping,
and
eventually
cell
death
(19),
presumably
because
endogenous
activities
can
somehow
substitute
for
the
transactivating
function
of
E1A.
Infection
can
even
be
productive
in
certain
cell
lines
(e.g.,
HCT116,
a
human-derived
colon
car-
cinoma),
although
with
a
moderate
amplification
level
U.
F.
Dedieu,
personal
communication).
The
ability
of
El-deleted
adenoviruses
to
selectively
prop-
agate
within
certain
tumor
cell
lines
of
human
origin
may
open
up
new
strategies
to
fight
cancer.
Indeed,
a
phase
I
clinical
trial
in
patients
with
p53-deficient
head
and
neck
squamous
cell
carcinoma
has
been
initiated
to
assess
whether
an
E1B
55K-deficient
ad-
enovirus
could
be
used
for
selective
viral
amplifica-
tion
and
reiterative
tumor
cell
killing.
This
novel
approach
(“adenotherapy”)
is
based
on
the
obser-
vation
that
in
vitro
infection
of
human
cell
lines
lack-
ing
p53
(a
most
common
feature
in
human
carcinoma)
with
the
mutant
adenovirus
was
produc-
tive
(and
thus
cytopathic)
even
though
it
was
down
a
100-fold
in
cells
with
a
normal
p53
status
(20),
per-
haps
because
E1A-induced
p53-dependent
apoptosis
occurred
prematurely
and
hampered
viral
produc-
tion
in
the
absence
of
E1B
55K
(see
above).
These
data
demonstrate
that
the
E1B
55K
protein
indeed
is
not
mandatory
for
efficient
export
of
the
viral
late
mRNAs.
That
A549-derived
cell
lines
(pS3
wild-type)
expressing
E1A
genes
and
only
part
of
E1B
(i.e.,
en-
coding
a
truncated
55K-encoding
gene)
can
propagate
El-deleted
adenoviruses
(17)
suggests
that
propagation
of
an
E1B
55K-deficient
virus
may
not
be
restricted
to
p53-mutated
cells
but
could
involve
other
factors
along
the
pS3
pathway.
Infection
with
human
recombinant
adenoviruses
is
not
restricted
to
a
particular
cell
type
or
host.
For
example,
they
can
be
used
to
transduce
a
variety
of
dividing
(e.g.,
activated
smooth
muscle
cells,
tumor
cells)
and
quiescent
cells
(e.g.,
from
the
liver,
mus-
cles,
brain).
C57BL/6
mice
have
been
widely
used
in
preclinical
studies
because
this
inbred
strain
is
par-
ticularly
reactive
to
viral
infections,
including
Ad5,
against
which
they
mount
a
strong
lymphocytic
infil-
tration
within
the
infected
organ
(21).
C57BL/6
mice
are
not
permissive
for
AdS
and
they
also
strongly
react
to
El-deleted
adenoviruses.
For
exam-
ple,
systemic
injection
of
a
lacZ-expressing
recombi-
nant
virus
typically
triggers
a
dose-dependent
inflammation
and
necrosis
together
with
the
recruit-
ment
of
macrophages
and
lymphocytes
within
the
in-
fected
organ
(mostly
liver).
Most
important,
these
responses
invariably
translate
into
a
limited
survival
of
the
transduced
hepatocytes,
even
in
C57BL/6
lacZ-transgenic
mice
immunotolerant
for
this
E.
coli-
derived
protein
(ref
19
and
references
therein).
Because
transgene
persistance
can
be
extended
in
immunodeficient
mice,
it
is
likely
that
the
recogni-
tion
of
viral
antigen
by
immune
effectors
participates
in
this
destructive
process.
Indeed,
early
and
late
viral
gene
expression
has
been
documented
in
C57BL/6
mice,
together
with
some
replication
of
the
viral
backbone
(22).
Mice
hepatocytes
infected
with
an
El-
defective
adenovirus
have
also
been
shown
to
un-
dergo
apoptosis
(23),
again
illustrating
the
need
to
further
cripple
the
virus
genetic
program
to
engineer
an
innocuous
gene
delivery
vehicle,
if
required.
ADDITIONAL
INACTiVATION
OF
EARLY
TRANSCRIPTION
UNITS
Targeting
E2
Region
E2
encodes
proteins
required
for
the
repli-
cation
of
the
viral
chromosome.
It
is
divided
into
two
subregions
(E2A
and
E2B)
depending
on
the
utili-
zation
of
two
polyadenylation
sites
(Fig.
1).
Subre-
gion
E2A
encodes
a
72
kDa
protein
that
is
also
involved
in
transcriptional
control,
viral
assembly,
and
host
range
(for
a
review,
see
ref
4).
This
pleio-
tropic
protein
binds
heterogeneous
nuclear
RNA,
as
well
as
single-stranded
replication
intermediates
within
virus-induced
subnuclear
structures.
Subre-
gion
E2B
encodes
a
140
kDa
DNA
polymerase
(AdPol)
and
a
80
kDa
protein
(pTP).
AdPol
and
pTP
are
synthesized
in
catalytic
amounts
during
the
early
phase
of
infection
and
function
as
heterodimers
dur-
ing
initiation
of
DNA
replication
in
the
nucleus.
pTP
exhibits
a
nuclear
localization
signal
and
also
associ-
ates
with
AdPol
in
the
cytoplasm,
allowing
its
nuclear
import.
pTP
is
covalently
bound
to
the
ITRs
during
the
initiation
of
replication
and
is
later
cleaved
into
a
55K
product
(TP),
most
likely
within
the
virions.
Both
pTP
and
TP
have
been
shown
to
interact
with
the
nuclear
matrix
(6,
24).
In
particular,
covalently
bound
pTP
(i.e.,
postreplication)
may
promote
interaction
of
the
viral
chromosomes
with
a
limited
number
of
(spe-
cific)
sites
within
the
nuclear
matrix
(24).
H5ts125
contains
a
thermosensitive
mutation
lo-
cated
within
the
E2A
72K-encoding
sequence
that
has
been
first
included
in
the
backbone
of
an
El-
deleted
virus
to
further
cripple
the
vector’s
biology.
Propagation
of
the
doubly
defective
virus
was
conse-
quently
abrogated
when
infected
293
cells
were
grown
at
nonpermissive
temperature
(39#{176}C),
whereas
it
was
almost
normal
at
32#{176}C.
Mice
hepatocytes
in-
fected
with
the
doubly
defective
virus
were
shown
to
persist
for
a
longer
period
of
time
than
their
coun-
terparts
infected
with
an
El-defective
control
virus.
Cellular
infiltration
of
the
infected
organ
was
also
de-
layed/reduced
consecutive
to
the
inclusion
of
the
mutation
within
the
backbone
(25).
ts125,
however,
is
a
point
mutation
and
may
not
abrogate
all
the
reg-
ulatory
functions
of
the
72K
protein.
Several
groups
ADENOVIRAL
VECTOR
ADVANCES
619
have
thus
deleted
the
E2A
gene
from
the
viral
back-
bone.
For
example,
El/E2A
doubly
deleted
viruses
have
been
constructed
in
A549
cells
engineered
to
transcomplement
both
regions
(18)
or
in
293
cells
engineered
to
express
the
AdS
E2A
gene
from
its
own
promoters
(26).
Deleting
E2A
from
the
viral
back-
bone
is
indeed
much
more
efficient
than
the
ts125
mutation
in
crippling
viral
replication
and
gene
ex-
pression
in
a
short-term
in
vitro
assay
(18).
The
dou-
bly
deleted
adenovirus
was
also
capable
of
transducing
mice
hepatocytes
in
vivo
(26).
The
host
responses
to
the
virally
infected
cells
and
the
persist-
ance
of
transgene
expression
that
can
be
achieved
in
the
absence
of
viral
DNA
replication
have
not
been
studied
yet.
The
293
cells
engineered
to
express
the
Ad2
140
kDa
DNA
polymerase
from
the
Rous
sarcoma
virus
long-terminal
repeat
(RSV
LTR)
transcomplement
the
growth
defect
of
H5ts36
(27),
an
AdPol
mutant
virus
unable
to
initiate
DNA
replication
at
38.5#{176}C.
pTP
has
also
been
targeted
for
inactivation.
For
ex-
ample,
293
cells
engineered
to
express
the
AdS
pTP
gene
have
been
used
to
construct
a
virus
in
which
a
large
deletion
within
the
pTP-encoding
sequences
has
been
introduced
(28).
These
novel
transcomple-
menting
cell
lines
should
soon
allow
the
construction
of
the
corresponding
E1/E2B
doubly
defective
vi-
ruses
and
the
assessment
of
their
biology.
Targeting
E4
Another
way
to
cripple
viral
gene
expression
is
to
si-
lence
the
E4
locus,
a
complex
regulatory
unit
that
en-
codes
seven
different
polypeptides.
This
approach
has
been
followed
by
several
groups
(23,
29-32),
essen-
tially
because
E4
subvert
endogenous
gene
expression
at
different
(i.e.,
transcriptional
and
post-transcrip-
tional)
levels
(for
a
review,
see
ref
4).
For
example,
the
E4
ORF6/7
polypeptide
competes
with
RB
pro-
teins
for
binding
to
the
E2F
transcription
factor,
trans-
forming
it
into
a
virus-specific
transactivator.
Another
E4
gene
product,
E4
ORF4,
has
been
shown
to
bind
and
regulate
cellular
phosphatase
2A,
modulating
the
activity
of
viral
(e.g.,
E1A)
and
cellular
transcription
factors.
The
E4
ORF3
and
E4
ORF6
proteins
share
a
redundant
function
that
somehow
regulates
the
mat-
uration
of
the
28
kb
late
primary
transcript
and/or
the
nuclear
export
of
the
corresponding
mRNAs
(Fig.
1).
In
contrast
to
the
other
E4
proteins,
this
regulatory
function
is
indispensable
for
virus
growth.
Although
E4
ORF6
and
E4
ORF3
exhibit
a
redundant
activity,
they
localize
to
different
subnuclear
compartments
during
infection
(9).
E4
ORF6
is
indeed
more
potent
than
E4
ORF3
and
exhibits
additional
properties.
For
example,
E4
ORF6
interacts
with
the
E1B
55K
protein,
a
complex
that
facilitates
the
cytoplasmic
accumula-
tion
of
late
mRNAs
at
the
expense
of
most
endoge-
nous
mRNAs
(for
a
review,
see
ref
4).
The
E4
ORF6
protein
also
displays
oncogenic
properties,
presum-
ably
because
it
inhibits
p53-dependent
apoptosis
after
binding
to
this
tumor
suppressor
(ref
33
and
refer-
ences
therein).
El/E4
doubly
defective
adenoviruses
(e.g.,
lacking
both
E4
ORF3
and
E4
ORF6)
can
be
efficiently
grown
in
293
cells
expressing
only
the
E4
ORF6
gene
(23,
32).
Extensive
E4
deletions
can
interfere
with
ex-
pression
from
the
adjacent
unit
(LS),
thereby
reduc-
ing
fiber
accumulation
and
viral
productivity
(32).
The
biology
of
E1/E4
doubly
defective
adenoviruses
has
been
assessed
in
vitro
and
in
vivo.
As
mentioned
previously,
infection
of
tumor-derived
cell
lines
of
hu-
man
origin
with
an
El-deleted
adenovirus
is
not
an
innocuous
event.
We
thus
compared
the
behavior
of
otherwise
“isogenic”
El-
and
E1/E4-defective
ade-
noviruses
in
these
model
systems
and
concluded
that,
per
se,
the
additional
deletion
of
E4
translates
into:
1)
a
lower,
most
likely
only
delayed,
replication
of
the
viral
backbone,
2)
an
almost
complete
shutoff
of
late
gene
and
protein
expression,
and
3)
no
apparent
vi-
rus-induced
cytotoxicity
(19).
Also
(and
in
contrast
to
their
isogenic
El-deleted
control
virus),
El/E4
doubly
defective
adenoviruses
were
not
amplified
when
HCT116
cells
were
similarly
infected
(E.
Vigne
andJ.
F.
Dedieu,
personal
communication).
Finally,
and
in
contrast
with
batches
of
El-deleted
adenovi-
ruses,
RCA
contamination
of
batches
of
El/E4
dou-
bly
defective
adenoviruses
has
never
been
detected
(E.
Vigne,
personal
communication).
Whereas
the
additional
inactivation
of
E4
drasti-
cally
disabled
viral
gene
expression
in
vitro,
trans-
gene
expression
(i.e.,
lacZ
expressed
from
the
RSV
LTR)
was
quantitatively
unchanged
in
a
short-term
in
vitro
assay
(19).
However,
this
did
not
apply
in
vivo.
Indeed,
tail
vein
injection
of
identical
doses
of
iso-
genic
El-
and
E1/E4
doubly
defective
adenoviruses
to
C57BL/6
lacZ-transgenic
mice
was
associated
as
early
as
4
days
postinfection
with
a
four-
to
sevenfold
reduction
in
virally
encoded
3-galactosidase
activity
in
the
liver.
Despite
a
much
longer
persistence
of
the
viral
chromosome
in
a
linear,
nonmethylated
status,
transgene
expression
from
the
vector
backbone
was
also
more
transient
in
the
absence
of
E4
(19).
In
con-
trast,
expression
of
lacZ
from
a
composite
CMV
en-
hancer/-actin
promoter
has
been
shown
to
last
for
at
least
2
months
in
another
lacZ-transgenic
murine
model
(23).
Together,
these
studies
demonstrate
that
inactivation
of
E4
is
a
prerequesite
for
stable
gene
delivery,
but
that
transgene
expression
from
an
El/E4
doubly
deleted
backbone
is
critically
depen-
dent
on
the
identity
of
the
promoter
used
to
drive
its
expression
in
vivo.
Engineering
E3
Transcription
of
the
E3
locus
occurs
throughout
in-
fection,
either
from
its
own
NFIL6-and
NFtB-respon-
620
Vol.
11
July
1997
The
FASEB
Journal
YEH
AND
PERRICAUDET
sive
promoter
or
as
part
of
the
LTU
(see
ref
10
and
references
therein).
Another
level
of
complexicity
is
added
by
the
use
of
two
polyadenylation
sites
(defin-
ing
the
E3A
and
E3B
families)
and
by
alternative
splicing
(Fig.
1).
E3
encodes
at
least
seven
polypep-
tides,
some
of
unknown
function.
The
whole
locus
is
dispensable
for
viral
growth
in
vitro
and
is
often
re-
moved
from
the
vector
backbone
to
extend
its
clon-
ing
capacity.
Preclinical
studies
suggest
that
E3
is
required
in
vivo
to
control
the
host
inflammatory
and
immune
responses
during
a
productive
infection
and
perhaps
to
establish
latency
within
lymphoid
tissues,
as
suggested
(34).
For
instance,
E3A
encodes
a
19
kDa
glycoprotein
(gpl9K)
that
binds
and
retains
MHC
class
I
molecules
within
the
endoplasmic
retic-
ulum
in
appropriate
in
vitro
conditions
(35).
The
E3B-encoded
proteins
have
also
been
shown
to
protect
infected
cells
from
tumor
necrosis
factor
a
(TNFa)
-mediated
cytolysis,
and
they
hamper
ex-
pression
of
phospholipase
A2
and
biosynthesis
of
in-
flammatory
mediators
(see
ref
36
and
references
therein).
Recent
observations
add
to
the
complexic-
ity
of
the
E3
locus:
1)
ADP
(for
adenovirus
death
pro-
tein)
is
an
E3A-encoded
l1.6kDa
protein
that
has
been
shown
in
vitro
to
facilitate
cytolysis
and
release
of
the
viral
progeny
in
the
late
phase
of
infection
(10),
2)
a
10.4K/14.5K
complex
between
two
E3B
proteins
has
also
been
reported
to
down-regulate
vi-
ral
gene
expression
at
the
translational
level
(37),
and
3)
gpl9K
may
function
beyond
the
sole
aspect
of
reducing
the
export
of
MHC
class
I
molecules
to
the
cell
surface
as
its
accumulation
within
the
retic-
ulum
somehow
up-regulates
NF-xB
binding
and
tran-
scriptional
activity
(38),
perhaps
another
way
to
protect
the
infected
cells
from
TNFa-mediated
cytol-
ysis
as
recently
reported
(see
ref
39
and
references
therein).
The
in
vivo
behavior
of
E3
deletion
mutants
has
been
extensively
studied
in
mice
(e.g.,
see
ref
34
and
references
therein).
However,
mice
are
not
permis-
sive
for
subgroup
C
adenoviruses,
and
as
such
are
not
the
most
relevant
models
to
decipher
the
biological
consequences
associated
with
the
removal
of
regula-
tory
proteins
that
are
critical
to
adenovirus
infections
in
nature.
This
may
explain
why
the
immune
re-
sponses
against
the
virus
and
the
virally
infected
cells
are
quite
different
in
mice
and
humans.
Thus,
the
absence
of
viral
replication
would
explain
why
the
CTL
response
against
adenovirus-infected
cells
is
di-
rected
mostly
against
early
gene
products
in
mice,
whereas
major
T
epitopes
have
been
assigned
to
the
late
(capsid)
proteins
in
humans
(40).
It
is
also
pos-
sible
that
infected
cells
of
human,
but
not
murine
origin,
could
present
MHC-restricted
viral
epitopes
from
the
virions
in
the
absence
of
de
novo
gene
ex-
pression,
as
recently
suggested
(1).
Subgroup
C
ad-
enovirus
infections
are
also
extremely
common
in
humans,
where
they
prime
both
the
humoral
and
cel-
lular
arms
of
the
immune
system
(ref
1
and
refer-
ences
therein).
Priming
may
even
cross
the
adenovirus
subgroup
specificity
stricto
senso,
as
hu-
man
memory
CTL
directed
against
AdS-infected
cells
have
been
reported
to
lyse
cells
infected
with
Adl
1
(subgroup
B).
Finally,
the
extent
to
which
endoge-
nous
factors
will
substitute
for
the
regulatory
activities
deleted
from
the
backbone
(e.g.,
E1A)
is
likely
to
be
species
dependent.
Given
the
complexity
of
the
E3
locus
and
the
limited
understanding
of
its
biological
significance
during
a
productive
or
a
latent
infection,
it
seems
quite
difficult
at
the
present
time
to
ascribe
particular
deletions
to
vectors
aimed
at
being
used
with
a
destructive
(e.g.,
cancer)
or
a
conservative
(e.g.,
cystic
fibrosis)
clinical
approach.
GENETIC
ENGINEERING
OF
THE
CAPSU)
Ideally,
adenoviral
vectors
should
deliver
and
express
their
transgene
only
in
the
targeted
cells
in
vivo.
How-
ever,
the
widespread
distribution
of
the
cellular
re-
ceptor
for
the
fiber
protein
precludes
the
targeting
of
specific
cell
types.
Barriers
also
exist
that
limit
ad-
enovirus-mediated
gene
transfer
in
vivo.
For
exam-
ple,
intratumoral
administration
of
a
recombinant
adenovirus
to
cancer
lung
patients
is
associated
with
dissemination
(41).
Because
of
the
sequential
nature
of
the
virus-cell
interaction,
most
approaches
aimed
at
enhancing
or
directing
gene
delivery
to
specific
cell
types
target
the
fiber
protein.
Genetic
engineer-
ing
of
the
fiber
protein
is
obviously
tricky
because
deletions
as
small
as
two
residues
from
its
carboxyl
terminus
prevented
proper
trimerization
of
the
mol-
ecule
in
vitro,
whereas
addition
of
6
residues,
but
not
27,
was
permissive
(42).
In
another
study,
addition
of
a
polylysine
peptide
to
the
carboxyl-terminal
end
of
the
fiber
allowed
the
recovery
of
a
recombinant
ad-
enovirus
that
was
shown
to
increase
gene
transfer
from
10-
to
300-fold
in
vitro
(43).
An
epidemiologic
approach
is
also
useful
in
pro-
viding
clues
to
alter
the
virus
tropism
in
vivo.
For
ex-
ample,
Ad5
and
Ad3
(subgroup
B)
bind
to
different
receptors,
and
binding
specificity
could
be
ex-
changed
in
vitro
by
swapping
the
knob
domain
of
the
fiber
protein
(44).
Another
study
has
taken
advan-
tage
of
the
observation
that
the
amino-terminal
do-
main
of
the
fiber,
which
anchors
the
molecule
into
the
vertex
capsomer
of
the
capsid,
represents
its
most
conserved
domain
within
different
serotypes.
Accord-
ingly,
the
substitution
of
the
fiber
of
AdS
by
that
of
Ad7
(subgroup
B)
within
the
virions
has
been
suc-
cessful
(45).
Viral
specificity
has
also
been
altered
in
vitro
by
genetic
engineering
of
the
penton
base,
for
example,
by
exchanging
the
neighboring
sequence
of
its
binding
motif
(46)
or
after
deletion
of
this
adhesive
motif
(ref
3
and
references
therein).
CsCI
purification
helper
adenovirus
ADENOVIRAL
VECTOR
ADVANCES
621
“GUTLESS”
ADENOVIRAL
VECTORS?
Different
strategies
are
being
developed
to
remove
all
transcription
units
from
the
viral
backbone
(“gut-
less”
adenoviruses).
A
major
limitation
here
is
to
transcomplement
the
LTU
to
levels
high
enough
so
the
structural
proteins
are
synthesized
to
levels
com-
patible
with
high
titers.
In
all
cases,
this
has
been
achieved
by
using
a
replicative
transcomplementa-
tion
approach.
For
example,
the
LTU
can
be
pro-
vided
in
transwithin
a
helper
adenovirus
(12,
47,
48).
In
principle,
this
strategy
extends
the
cloning
capac-
ity
of
the
vector
to
its
maximal
limit
(e.g.,
that
of
the
adenovirus
capsid).
However,
it
invariably
leads
to
the
copropagation
of
the
helper
virus
with
the
mini-
virus
so
extensive
purification
is
required
to
reduce
the
contamination
of
the
production
batch
(Fig.
3).
Because
the
minivirus
is
amplified
and/or
packaged
less
efficiently
than
the
helper
virus,
it
is
necessary
to
provide
it
with
a
selective
advantage.
This
can
be
achieved
by
engineering
both
viruses
(e.g.,
the
mini-
virus
retains
E4
whereas
the
“helper”
virus
is
E4-de-
fective),
so
each
virus
requires
the
presence
of
the
other
for
its
own
propagation
in
293
cells
(12).
The
introduction
of
a
weaker
encapsidation
signal
on
the
helper
virus
(48)
or
its
removal
during
reiterative
am-
plification
steps
have
also
been
described
(49).
It
is
also
possible
to
transcomplement
the
LTU
in
cis
before
its
removal
from
the
backbone
during
viral
amplification.
By
definition,
the
cloning
capacity
of
recombinant
miniviruses
recovered
by
this
strategy
is
that
of
the
undeleted
adenovirus.
Specific
removal
of
adenoviral
helper
sequences
has
been
achieved
with
the
loxP-specific
cre
recombinase
from
bacterio-
phage
P1.
For
example,
direct
repeats
of
loxP
sites
“framing”
a
25
kb
fragment
of
viral
sequences
that
included
pIX,
E2,
and
most
of
the
LTU
(i.e.,
Ll,
L2,
L3,
and
L4
families)
have
been
included
in
the
back-
bone
of
an
El-deleted
recombinant
virus
(22).
Am-
plification
of
this
virus
in
293
cells
expressing
the
crc
recombinase
in
trans
resulted
in
the
precise
excision
of
the
25
kb
fragment
and
the
emergence
of
a
mini-
virus
that
retained
only
the
L5
(fiber
encoding)
gene,
the
whole
E4
locus,
and
the
cis
elements
required
for
replication
and
packaging
(Fig.
3).
In
vitro
and
in
vivo
assessment
of
the
biology
of
this
gutless
adeno-
virus
has
outlined
its
intrinsic
disabled
properties,
in-
cluding
the
instability
of
its
chromosome
after
infection.
Coinfection
with
an
El-deleted
adenovirus
increased
the
stability
of
the
gutless
backbone
in
vitro
and
reestablished
(RSV
LTR-driven)
transgene
ex-
pression
in
C57BL/6
hepatocytes
(22).
Trans-acting
factors
of
viral
origin
thus
apparently
promoted,
di-
rectly
or
indirectly,
the
stabilization
of
the
gutless
backbone
and/or
it
allowed
up-regulation
of
trans-
gene
expression
from
the
RSV
LTR.
As
stated
by
the
authors,
these
observations
may
be
crucial
to
the
de-
sign
of
adenovirus-based
vectors
for
gene
therapy,
as
they
indicate
that
overcrippling
the
virus
genetic
pro-
gram
may
have
pernicious
effects
on
their
transduc-
ing
properties.
Indeed,
nuclear
processes
such
as
transcription
and
RNA
processing
most
likely
occur
in
association
with
the
nuclear
matrix.
Specific
matrix-bound
com-
plexes
may
even
bridge
these
processes
(for
example,
see
ref
SO
and
references
therein).
Matrix
attach-
ment
regions
are
potential
cis-acting
players
that
could
modulate
expression
of
eukaryotic
genomes
in
such
a
scheme.
Does
the
routing
of
the
AdS
chro-
mosome
to
appropriate
subnuclear
compartments
also
apply
within
a
permissive,
a
non-,
or
a
“semiper-
missive”
context?
Indeed,
recent
data
suggest
that
cis-
A
TA
TAPE’
ransgene
mini-pIasmid”
helper
adenovirus
(e.g.,
5E4)
..Tralrstection
ection
B
__
#{149}
E4
Transfecon
-
(293cre)
oxp
LTU
h,xP
\../
Figure
3.
Minimal
vectors.
A)
The
helper
virus
approach.
A
plasmid
carrying
a
minimal
adenoviral
genome
generates
the
corresponding
“mini-adenovirus”
if
the
missing
functions
are
provided
in
trans
by
a
helper
adenovirus.
The
presence
of
E4
in
the
mini-adenovirus
genome
and
its
absence
in
the
helper
virus
ensure
its
propagation
(12).
Alternatively,
a
weaker
encapsidation
sequence
may
be
included
within
the
backbone
of
the
helper
adenovirus
(see
ref
48
for
details).
B)
Excision
of
intervening
sequences
by
the
loxP-specific
cre
recombinase
(22;
see
also
ref
49).
Because
of
its
smaller
density,
the
mini-adenovirus
can
be
separated
from
the
helper
virus
by
CsCl
gradient
purification.
622
Vol.
11
July
1997
The
FASEB
Journal
YEH
AND
PERRICAUDET
and/or
trans
elements
of
viral
origin
promote
(i.e.,
directly
or
indirectly)
a
strong
interaction
of
the
viral
chromosomes
with
the
nuclear
matrix,
in
close
asso-
ciation
with
spliceosome
components
(8).
In
this
re-
gard,
covalently
bound
TP
and
pTP
complexes
are
potential
candidates
involved
in
the
routing
of
the
viral
genome
to
appropriate
locations
after
nuclear
entry
and
viral
DNA
replication,
respectively
(6,
24).
How
this
would
translate
in
the
case
of
defective
back-
bones
for
which
long-term
expression
of
one
partic-
ular
gene
(i.e.,
the
transgene)
is
targeted
while
that
of
the
remaining
genes
should
ideally
be
shut
off
re-
mains
elusive.
These
issues
may
even
be
contradic-
tory,
especially
if
replication
of
the
backbone
is
abrogated.
That
a
gutless
adenovirus
unable
to
rep-
licate
was
associated
with
transient
transgene
expres-
sion
unless
putative
viral
factors
were
provided
in
trans
would
support
this
hypothesis,
as
suggested
(22).
The
study
of
wild-type
adenoviruses
has
proved
unvaluable
to
decipher
many
fundamental
biological
processes;
a
study
of
their
defective
derivatives
will
likely
contribute
to
our
understanding
of
eukaryotic
gene
regulation
in
the
near
future.
Eventually,
it
will
also
provide
a
clear
picture
of
the
extent
to
which
the
“guts”
of
the
virus
can
be
removed
without
disabling
expression
of
the
transgene
they
carry.
The
authors
wish
to
thank
the
members
of
the
laboratory,
in
particular
E.
Vigne
andJ.
F.
Dedieu,
for
helpful
discussions
and
critical
comments.
We
acknowledge
that
important
publications
in
the
field
could
not
be
referenced
due
to
space
limitations.
The
work
conducted
in
this
laboratory
would
not
have
been
made
possible
without
the
financial
contributions
from
Rhone-Poulenc
and
the
BioAvenir
Program
from
the
French
Ministry
of
Research
and
Industry.
REFERENCES
1.
Smith,
C.
A.,
Woodruff,
L.
S.,
Kitchingman,
G.
R.,
and
Rooney,
C.
M.
(1996)
Adenovirus-pulsed
dentntic
cells
stimulate
human
virus-specific
T-cells
responses
in
vitro.
J.
Virol.
70,
6733-6740
2.
Hierholzer,
J.
C.
(1992)
Adenovinises
in
the
immunocompro-
mised
host.
Clin.
Microbiol.
Rev.
5,
262-274
3.
Freimuth,
P.
(1996)
A
human
cell
line
selected
for
resistance
to
adenovirus
infection
has
reduced
levels
of
the
virus
receptor.
j
Virol.
70,
4081-4085
4.
Shenk,
T.
(1996)
Adenoviridae:
the
viruses
and
their
replication.
In
Virology
(Fields,
B.
N.,
Knipe,
D.
M.,
Howley,
P.
M.,
et
al.,
eds)
pp.
2111-2148,
Raven
Publishers,
Philadelphia
5.
Huang,
S.,
Kamata,
T.,
Takada,
V.,
Ruggeri,
Z.
M.,
and
Nemerow,
G.
R.
(1996)
Adenovirus
interaction
vith
distinct
integrins
me-
diates
separate
events
in
cell
entry
and
gene
delivery
to
hema-
topoieuc
cells.
j
Virol.
70,
4502-4508
6.
Schaack,J.,
Ho,
W.,
Freimuth,
P.,
and
Shenk,
T.
(1990)
Adeno-
virus
terminal
protein
mediates
both
nuclear
matrix
association
and
efficient
transcription
of
adenovirus
DNA.
Genes
&
Den.
4,
1197-1208
7.
Ishov,
A.
M.,
and
Maul,
G.
G.
(1996)
The
periphery
of
nuclear
domain
10
(ND
10)
as
site
of
DNA
virus
deposition.
j
Cell
Biol.
134,
815-826
8.
Besse,
S.,
and
Puvion-Dutilleul,
F.
(1995)
Anchorage
of
adeno-
viral
RNAs
to
clusters
of
interchromatin
granules.
Gene
Exp.
5,
79-92
9.
Carvalho,
T.,
Seeler,J.-S.,
Ohman,
K.,Jordan,
P.,
Pettersson,
U.,
Akusjarvi,
G.,
Carmo-Fonseca,
M.,
and
Dejean,
A.
(1995)
Tar-
geting
of
adenovirus
E1A
and
E4-Orf
proteins
to
nuclear
ma-
trix-associated
PML
bodies.
j
Cell
Biol.
131,
45-56
10.
Tollefson,
A.
E.,
Scaria.
A.,
Hermiston,
T.
W.,
Ryerse,J.
S.,
Wold,
L.J.,
and
Wold,
W.
S.
M.
(1996)
The
adenovirus
death
protein
(E3-1
1.6K)
is
required
at
very
late
stages
of
infection
for
effi-
cient
cell
lysis
and
release
of
adenovirus
from
infected
cells.
j
Virol.
70,
2296-2306
11.
Graham,
F.
L.,
Smiley,
J.,
Russel,
W.
C.,
and
Nairn,
R.
(1977)
Characteristics
of
a
human
cell
line
transformed
by
DNA
from
human
adenovirus
type
5.].
Gen.
Virol.
36,
59-72
12.
Vigne,
E.,
Dedieu,
J.-F.,
Orsini,
C.,
Latta,
M.,
Klonjkowski,
B.,
Prost,
E.,
Lakich,
M.
M.,
Kremer,
E.
J.,
Denfle,
P.,
Perricaudet,
M.,
and
Yeh,
P.
(1996)
Improvement
of
adenoviral
vectors
for
human
gene
therapy.
In
Genetic
Approaches
to
Noncommunicable
Diseases
(Berg,
K.,
Boulyjenkov,
V.,
and
Christen,
V.,
eds)
pp.
113-131,
Springer-Verlag
Berlin,
Heidelberg
13.
Ketner,
G.,
Spencer,
F.,
Tugendreich,
S.,
Connelly,
C.,
and
Hie-
ter,
P.
(1994)
Efficient
manipulation
of
the
human
adenovirus
genome
as
an
infectious
yeast
artificial
chromosome
clone.
Proc.
NatI.
Acad.
Sci.
USA
91,
6186-6190
14.
Chartier,
C.,
Degryse,
E.,
Gantzer,
M.,
Dieterl#{233},
A.,
Pavirani,
A.,
and
Mehtali,
M.
(1996)
Efficient
generation
of
recombinant
ad-
enovirus
vectors
by
homologous
recombination
in
Escherichia
coli.
J.
Virol.
70,
4805-48
10
15.
Crouzet,
J.,
Naudin,
L.,
Orsini,
C.,
Vigne,
E.,
Ferrero,
L.,
Le
Roux,
A.,
Benoit,
P.,
Ustta,
M.,
Torrent,
C.,
Den#{232}fle,
P.,
Mayaux,
J.-F.,
Perricaudet,
M.,
and
Yeh,
P.
(1997)
Recombinational
con-
struction
in
E.
coli
of
infectious
adenoviral
genomes.
Foc.
Nati.
Aca4.
Sd.
USA
In
press
16.
Hehir,
K.
M.,
Armentano,
D.,
Cardoza,
L.
M.,
Choquette,
T.
L.,
Berthelette,
P.
B.,
White,
G.
A.,
Couture,
L.
A.,
Everton,
M.
B.,
Keegan,J.,
Martin,J.
M.,
Pratt,
D.
A.,
Smith,
M.
P.,
Smith,
A.
E.,
and
Wadsworth,
S.
C.
(1996)
Molecular
characterization
of
rep-
lication-competent
variants
of
adenovirus
vectors
and
genome
modifications
to
prevent
their
occurrence.
J.
Virol.
70,
8459-
8467
17.
Imler,
J.
L.,
Chartier,
C.,
Dreyer,
D.,
Dieterl#{233},
A.,
Saintemarie,
M.,
Faure,
T.,
Pavirani,
A.,
and
Mehtali,
M.
(1996)
Novel
corn-
pleinentation
cell
lines
derived
from
human
lung
carcinoma
A549
cells
support
the
growth
of
El-deleted
adenovirus
vectors.
Gene
Titer.
3,
75-84
18.
Gorziglia,
M.
I.,
Kadan,
M.J.,
Yei,
S.,
Lim,J.,
Lee,
G.
M.,
Luthra,
R.,
and
Trapnell,
B.
C.
(1996)
Elimination
of
both
El
and
E2a
from
adenovirus
vectors
further
improves
prospects
for
in
vivo
human
gene
therapy.J.
Virol.
70,
4173-4178
19.
Dedieu,J.-F.,
Vigne,
E.,
Torrent,
C.,Jullien,
C.,
Mahfouz,
1.,
Cail-
laud,J.-M.,
Aubailly,
N.,
Orsini,
C.,
Guillaume,J.-M.,
Opolon,
P.,
Dela#{233}re,
P.,
Perricaudet,
M.,
and
Yeh,
P.
(1997)
Long-term
gene
delivery
into
the
liver
of
immunocompetent
mice
with
E1/E4-
defective
adenoviruses.j
Virol.
71,
4626-4637
20.
Bischoff,J.
K.,
Kim,
D.
H.,
Williams,
A.,
Heise,
C.,
Horn,
S.,
Muna,
M.,
Ng,
L.,
Nyc,
J.
A.,
Sampson-Johannes,
A.,
Fattaey,
A.,
and
Mc-
Cormick,
F.
(1996)
An
adenovirus
mutant
that
replicates
selectively
in
p53-deficient
human
tumor
cells.
Science
274,
373-376
21.
Ginsberg,
H.
S.,
Moldawer,
L.
L.,
Sehgal,
P.
B.,
Redington,
M.,
Kilian,
P.
L.,
Chanock,
R.
M.,
and
Prince,
G.
A.
(1991)
A
mouse
model
for
investigating
the
molecular
pathogenesis
of
adenovi-
rus
pneumoniae.
Proc.
NatI.
Acad.
Sci.
USA
88,
1651-1655
22.
Lieber,
A.,
He,
C-V.,
Kinllova,
I.,
and
Kay,
M.
A.
(1996)
Recombinant
adenoviruses
with
large
deletions
generated
by
crc-mediated
exci-
sion
exhibit
different
biological
properties
compared
with
first-gen-
eration
vectors
in
vitro
and
in
vivo.j
VitaL
70,
8944-8960
23.
Gao,
G.-P.,
Yang,
V.,
and
Wilson,J.
M.
(1996)
Biology
of
adeno-
virus
vectors
with
El
and
E4
deletions
for
liver-directed
gene
therapy.
J.
Virol.
70,
8934-8943
24.
Fredman,J.
N.,
and
Engler,J.
A.
(1993)
Adenovirus
precursor
to
terminal
protein
interacts
with
the
nuclear
matrix
in
vivo
and
in
vitro.].
ViroL
67,
3384-3395
25.
Engelhardt,
J.
F.,
Ye,
X.,
Doranz,
B.,
and
Wilson,
J.
M.
(1994)
Ablation
of
E2A
in
recombinant
adenoviruses
improves
trans-
gene
persistence
and
decreases
inflammatory
response
in
mouse
liver.
Proc.
NaIL
Acad.
Sci.
USA
91,
6
196-6200
26.
Zhou,
H.,
O’Neal,
W.,
Morral,
N.,
and
Beaudet,
A.
L.
(1996)
Development
of
a
complementing
cell
line
and
a
system
for
con-
ADENOVIRAL
VECTOR
ADVANCES
623
struction
of
adenovirus
vectors
with
El
and
E2a
deleted.
J.
ViroL
70,
7030-7038
27.
Amalfitano,
A.,
Begy,
C.
R.,
and
Chamberlain,J.
S.
(1996)
Im-
proved
adenovirus
packaging
cell
lines
to
support
the
growth
of
replication-defective
gene-delivery
vectors.
Proc.
Nati.
Acad.
So.
USA
93,
3352-3356
28.
Schaack,J.,
Guo,
X.,
and
Langer,
S.J.
(1996)
Characterization
of
a
replication-incompetent
adenovirus
type
5
mutant
deleted
for
the
preterminal
protein
gene.
Proc.
NaIL
Acad.
Sci.
USA
93,
14686-14691
29.
Krougliak,
V.,
and
Graham,
F.
L.
(1995)
Development
of
cell
lines
capable
of
complementing
El,
E4,
and
protein
IX
defective
adenovirus
type
5
mutants.
Human
Gene
Ther.
6,
1575-1586
30.
Wang,
Q.,
and
Finer,
M.
H.
(1995)
A
packaging
cell
line
for
propagation
of
recombinant
adenovirus
vector
containing
two
lethal
gene-region
deletions.
Gene
Titer.
2,
775-783
31.
Yeh,
P.,
Dedieu,J.-F.,
Orsini,
C.,
Vigne,
E.,
Den#{232}fie,
P.,
and
Per-
ricaudet,
M.
(1996)
Efficient
dual
transcomplementation
of
ad-
enovirus
El
and
E4
regions
from
a
293-derived
cell
line
expressing
a
minimal
E4
functional
unit.
j
Virol.
70,
559-565
32.
Brough,
D.
E.,
Lizonova,
A.,
Hsu,
C.,
Kulesa,
V.
A.,
and
Kovesdi,
I.
(1996)
A
gene
transfer
vector-cell
line
system
for
complete
functional
complementation
of
adenovirus
early
regions
El
and
E4.j
Viral.
70,
6497-6501
33.
Moore,
M.,
Horikoshi,
N.,
and
Shenk,
T.
(1996)
Oncogenic
po-
tential
of
the
adenovirus
E4orf6
protein.
Proc.
Nat!.
Acad.
Sci.
USA
93,
11295-11301
34.
Sparer,
T.
E.,
Tripp,
R.
A.,
Dillehay,
D.
L.,
Hermiston,
T.
W.,
Wold,
W.
S.
M.,
and
Gooding,
L.
R.
(1996)
The
role
of
human
adenovirus
early
region
3
proteins
(gpl9K,
10.4K,
14.5K,
and
14.7K)
in
a
murine
pneumonia
model.j
ViraL
70,
2431-2439
35.
Flomenberg,
P.,
Piaskowski,
V.,
Truitt,
R.
L.,
and
Casper,
J.
T.
(1996)
Human
adenovirus-specific
CD8
T-cell
responses
are
not
inhibited
by
E3-l9K
in
the
presence
of
gamma
interferon.
].
ViroL
70,
6314-6322
36.
Thorne,
T.
E.,
Voelkel-Johnson,
C.,
Casey,
W.
M.,
Parks,
L.
W.,
and
Laster,
S.
M.
(1996)
The
activity
of
cytosolic
phospholipase
A2
is
required
for
the
lysis
of
adenovirus-infected
cells
by
tumor
necrosis
factor.
j
ViroL
70,
8502-8507
37.
Zhang,
X.,
Bellett,
A.
J.
D.,
Hla,
K.
T.,
Voss,
T.,
Mullbacher,
A.,
and
Braithwaite,
A.
W.
(1994)
Down-regulation
of
human
ade-
novirus
Ela
by
E3
gene
products:
evidence
for
translational
con-
trol
of
Ela
by
E3
14.5K
and/or
10.4K
products.
J.
Gen.
ViroL
75,
1943-1951
38.
PahI,
H.
L.,
Sester,
M.,
Burgert,
H.-G.,
and
Baeuerle,
P.
A.
(1996)
Activation
of
transcription
factor
NF-B
by
the
adenovirus
E3/
19K
protein
requires
its
ER
retention.j
Cell.
Biol.
132,
511-522
39.
Bannaga,
M.
(1996)
Life-death
balance
within
the
cell.
Science
274,
724
40.
Flomenberg,
P.,
Piaskowski,
V.,
Truitt,
R.
L.,
and
Casper,
J.
T.
(1995)
Characterization
of
human
proliferative
T
cell
responses
to
adenovirus.].
Infect.
Die.
171,
1090-1096
41.
Tursz,
T.,
Le
Cesne,
A.,
Baldeyrou,
P.,
Gautier,
E.,
Opolon,
P.,
Schatz,
C.,
Pavirani,
A.,
Courtney,
M.,
Lamy,
D.,
Ragot,
T.,
Saul-
nier,
P.,
Andremont,
A.,
Monier,
R.,
Perricaudet,
M.,
and
Le
Chevalier,
T.
(1996)
Phase
I
study
of
a
recombinant
adenovirus-
mediated
gene
transfer
in
lung
cancer
patients.].
Nail.
Cancer
Inst.
88,
1857-1863
42.
Hong,
J.
S.,
and
Engler,
J.
A.
(1996)
Domains
required
for
as-
sembly
of
adenovirus
type
2
fiber
trimers.j
ViroL
70,
7071-7078
43.
Wickham,
T.
J.,
Roelvink,
P.
W.,
Brough,
D.
E.,
and
Kovesdi,
I.
(1996)
Adenovirus
targeted
to
heparan-containing
receptors
in-
creases
its
gene
delivery
efficiency
to
multiple
cell
types.
Nature
BiotechnoL
14,
1570-1573
44.
Stevenson,
S.
C.,
Rollence,
M.,
White,
B.,
Weaver,
L.,
and
Mc-
Clelland,
A.
(1995)
Human
adenovirus
serotypes
3
and
5
bind
to
two
different
cellular
receptors
via
the
fiber
head
domain.].
ViroL
69,
2850-2857
45.
Gall,
J.,
Kass-Eisler,
A.,
Leinwand,
L.,
and
Falck-Pedersen,
E.
(1996)
Adenovirus
type
5
and
7
capsid
chimera:
fiber
replace-
ment
alters
receptor
tropism
without
affecting
primary
immune
neutralization
epitopes.].
ViroL
70,
2116-2123
46.
Wickham,
T.J.,
Carrion,
M.
E.,
and
Kovesdi,
I.
(1995)
Targeting
of
adenovirus
penton
base
to
new
receptors
through
replace-
ment
of
its
RGD
motif
with
other
receptor-specific
peptide
mo-
tifs.
Gene
Ther.
2,
750-756
47.
Mitani,
K.,
Graham,
F.
L.,
Caskey,
C.
T.,
and
Kochanek,
S.
(1995)
Rescue,
propagation,
and
partial
purification
of
a
helper
virus-
dependent
adenovirus
vector.
Proc.
Nail.
Acad.
Sri.
USA
92,3854-
3858
48.
Kochanek,
S.,
Clemens,
P.,
Mitani,
K.,
Chen,
H-H.,
Chan,
S.,
and
Caskey,
C.
T.
(1996)
A
new
adenoviral
vector:
replacement
of
all
viral
coding
sequences
with
28
kb
of
DNA
independently
ex-
pressing
both
full-length
dystrophin
and
-galactosidase.
Proc.
NaIl.
Acad.
Sci.
USA
93,
573
1-5736
49.
Parks,
R.
J.,
Chen,
L.,
Anton,
M.,
Sankar,
U.,
Rudnicki,
M.
A.,
and
Graham,
F.
L.
(1996)
A
helper-dependent
adenovirus
vector
system:
removal
of
helper
virus
by
crc-mediated
excision
of
the
viral
packaging
signal.
Proc.
Nail.
Acad.
Sri.
USA
93,
13565-13570
50.
Mortillaro,
M.J.,
Blencowe,
B.J.,
Wei,
X.,
Nakayasu,
H.,
Du,
L.,
Warren,
S.
L.,
Sharp,
P.
A.,
and
Berezney,
R.
(1996)
A
hyper-
phosphozylated
form
of
the
large
subunit
of
RNA
polymerase
II
is
associated
with
splicing
complexes
and
the
nuclear
matrix.
Proc.
Nail.
Acad.
Sri.
USA
93,
8253-8257