21.
A.
Fire,
U.
Samuels,
P.
A.
Sharp,
J.
Biol.
Chem.
259,
2509
(1984);
D.
K.
Hawley
and
R.
G.
Roeder,
ibid.
262,
3452
(1987).
22.
B.
Corthesy
et
al.,
Science
239,
1137
(1988).
23.
S.
K.
Yoshinaga
and
L.
P.
Freedman,
unpublished.
24.
K.
A.
Jones
et
al.,
Cell
42,
559
(1985).
25.
C.
Scheidereit
and
M.
Beato,
Proc.
Natl.
Acad.
Sci.
U.S.A.
81,
3029
(1984).
26.
H.
M.
Jantzen
et
al.,
Cell
49,
29
(1987).
27.
We
thank
W.
Soeller
for
embryo
extracts
and
discus-
sions;
M.
Biggin,
B.
England,
D.
Granner,
J.
La-
Baer,
and
D.
Picard
for
DNA
reagents;
S.
Johnson
and
R.
Myers
for
critical
readings
of
the
manuscript;
and
K.
Mulherin
and
B.
Maler
for
preparation
of
the
text
and
figures,
respectively.
Supported
by
grants
from
the
National
Institutes
of
Health
and
the
National
Science
Foundation;
postdoctoral
support
was
from
the
American
Cancer
Society,
California
Division
(S.K.Y.),
and
from
the
Bank
of
America-
Giannini
Foundation
(L.P.F.).
18
April
1989;
accepted
2
June
1989
Monoclonal
Antibody-Mediated
Tumor
Regression
by
Induction
of
Apoptosis
BERNHARD
C.
TRAUTH,
CHRISTIANE
KiAS,
ANKE
M.
J.
PETERS,
SIEGFRIED
MATZKU,
PETER
MOLLER,
WERNER
FALK,
KLAUS-MICHAEL
DEBATIN,
PETER
H.
KRAMMER*
To
characterize
cell
surface
molecules
involved
in
control
of
growth
of
malignant
lymphocytes,
monoclonal
antibodies
were
raised
against
the
human
B
lymphoblast
cell
line
SKW6.4.
One
monoclonal
antibody,
anti-APO-1,
reacted
with
a
52-kilodalton
antigen
(APO-1)
on
a
set
of
activated
human
lymphocytes,
on
malignant
human
lymphocyte
lines,
and
on
some
patient-derived
leukemic
cells.
Nanogram
quantities
of
anti-APO-l
completely
blocked
proliferation
of
cells
bearing
APO-1
in
vitro
in
a
manner
characteristic
of
a
process
called
programmed
cell
death
or
apoptosis.
Cell
death
was
preceded
by
changes
in
cell
morphology
and
fragmentation
of
DNA.
This
process
was
distinct
from
antibody-
and
complement-dependent
cell
lysis
and
was
mediated
by
the
antibody
alone.
A
single
intravenous
injection
of
anti-APO-1
into
nul
nu
mice
carrying
a
xenotransplant
of
a
human
B
cell
tumor
induced
regression
of
this
tumor
within
a
few
days.
Histological
thin
sections
of
the
regressing
tumor
showed
that
anti-APO-1
was
able
to
induce
apoptosis
in
vivo.
Thus,
induction
of
apoptosis
as
a
consequence
of
a
signal
mediated
through
cell
surface
molecules
like
APO-1
may
be
a
useful
therapeutic
approach
in
treatment
of
malignancy.
CEELL
SURFACE
MOLECULES
ARE
CRU-
cial
in
lymphocyte
growth
control.
Such
molecules
may
function
as
re-
ceptors
for
growth-stimulating
cytokines
or
be
associated
with
receptors
and
transmit
signals
essential
for
growth
regulation.
Re-
ceptor
blockade
or
removal
of
the
stimulat-
ing
cytokines
can
lead
to
decreased
lympho-
cyte
growth.
Withdrawal of
interleukins
slow
human
lymphocyte
growth
and
finally
leads
to
a
characteristic
form
of
cell
death
called
"programmed
cell
death"
or
apoptosis
(1).
Apoptosis
is
the
most
common
form
of
eukaryotic
cell
death
and
occurs
in
embryo-
genesis,
metamorphosis,
tissue
atrophy,
and
tumor
regression
(2).
It
is
also
induced
by
cytotoxic
T
lymphocytes
and
natural
killer
and
killer
cells;
by
cytokines
like
tumor
necrosis
factor
(TNF)
and
lymphotoxin
(LT);
and
by
glucocorticoids
(1,
2).
The
most
characteristic
signs
of
apoptosis
are
segmentation
of
the
nucleus,
condensation
of
the
cytoplasm,
membrane
blebbing,
and
DNA
fragmentation
into
multimers
of
about
180
base
pairs
(called
a
"DNA
lad-
der")
(1,
2).
To
analyze
mechanisms
of
lymphocyte
growth
control
and
to
interfere
with
the
replication
of
lymphoid
tumor
cells
we
raised
monoclonal
antibodies
(MAbs)
against
cell
surface
molecules
involved
in
these
processes.
We
found
one
MAb
(anti-APO-1)
that
blocks
growth
and
induces
apoptosis
of
SKW6.4
cells
(3).
Anti-APO-1
(IgG3,
K,
KD
=
1.9
x
10-10)
bound
to
approximately
4
x
104
sites
on
the
surface
of
SKW6.4
cells
(4).
It
specifically
immunoprecipitated
an
endogenously
synthesized
protein
antigen
(APO-1)
from
SKW6.4
cells
which,
under
reducing
conditions,
was
observed
on
SDS-
Fig.
1.
Molecular
weight
of
the
cell
surface
antigen
97
-
APO-1:
immunoprecip-
itation
of
biosynthetically
la-
68-
beled
APO-1
from
the
sur-
face
of
SKW6.4
cells
with
f
either
isotype-matched
con-
43
-
_
trol
MAb
(left
lane)
or
anti-
APO-1
(right
lane).
The
numbers
on
the
left
margin
indicate
the
positions
of
the
25.7
size
markers.
Cells
(3
x
106)
were
labeled
with
60
,uCi
of
75Se-labeled
methionine
18.4
(Amersham,
Braunschweig,
FRG)
in
6
ml
of
methionine-free
culture
medium
(Biochrom,
Berlin)
for
48
hours.
After
washing,
the
cells
were
incubated
in
either
control
MAb
or
anti-APO-1
(1
,ug/ml)
at
4°C
for
45
min.
The
cells
were
washed
and
resuspended
in
lysis
buffer
(tris-buffered
saline,
pH
7.3,
1%
Nonidet
P-40,
1
mM
phenylmethylsulfonyl
fluoride,
0.1%
apro-
tinin)
at
room
temperature
for
30
min.
The
lysates
were
centrifuged
and
supernatants
were
incubated
with
protein
A-Sepharose
beads
(Phar-
macia,
Uppsala,
Sweden)
at
4°C
for
hour.
The
immune
complexes
were
washed
four
times
with
buffer
(tris-buffered
saline,
pH
7.3,
0.25%
Noni-
det
P-40)
and
resuspended
in
SDS-PAGE
sample
buffer
containing
5%
SDS
and
5%
2-mercapto-
ethanol.
The
samples
were
heated
to
95°C,
centri-
fuged,
and
counts
per
minute
of
the
supernatants
were
determined
in
a
-y-counter.
A
total
of
15,000
cpm
were
loaded
in
each
lane
and
analyzed
by
a
10%
SDS-PAGE
(18).
The
gel
was
dried
and
subjected
to
autoradiography.
polyacrylamide
gel
electrophoresis
(SDS-
PAGE)
as
a
main
band
of
52
kD
(Fig.
1).
Apart
from
actin
(43
kD),
which
was
non-
specifically
precipitated
with
IgG3,
anti-
APO-1
specifically
immunoprecipitated
a
minor
band
of
25
kD.
This
25-kD
protein
might
either
represent
a
degradation
prod-
uct
or
be
noncovalently
associated
with
the
52-kD
protein.
There
are
two
major
modes
of
death
in
nucleated
eukaryotic
cells.
Necrosis
as
a
result,
for
example,
of
complement
attack
is
characterized
by
swelling
of
the
cells
and
rupture
of
the
plasma
membrane
caused
by
an
increase
in
permeability.
Cells
that
under-
go
apoptosis,
however,
show
a
different
biochemical
and
morphological
pattern
(2).
This
pattern
corresponds
to
the
one
induced
by
anti-APO-1:
condensation
of
the
cyto-
plasm,
membrane
blebbing
(Fig.
2a),
and
endonuclease-induced
DNA
fragmentation
(5)
into
multimers
of
approximately
180
bp
B.
C.
Trauth,
C.
Klas,
A.
M.
J.
Peters,
W.
Falk,
P.
H.
Krammer,
Institute
for
Immunology
and
Genetics,
Ger-
man
Cancer
Research
Center,
Heidelberg,
Federal
Re-
public
of
Germany.
S.
Matzku,
Institute
for
Radiology
and
Pathophysiology,
German
Cancer
Research
Center,
Heidelberg,
Federal
Republic
of
Germany.
P.
Molier,
Institute
for
Pathology
of
the
University
of
Heidelberg,
Federal
Republic
of
Germany.
K.-M.
Debatin,
Oncology/Immunology
Section,
Univer-
sity
Children's
Hospital,
Heidelberg,
Federal
Republic
of
Germany.
*To
whom
correspondence
should
be
addressed:
REPORTS
301
ZnCI2).
Typical
yields
obtained
were
1
mg
of
>90%
pure
T7EX556
per
liter
of
starting
culture.
13.
F.
Payvar
et
al.,
Cell
35,
381
(1983).
14.
W.
C.
Soeller
et
al.,
Genes
Dev.
2,
68
(1988).
15.
U.
Heberlein
and
R.
Tjian,
Nature
331,410
(1988).
16.
J.
LaBaer,
thesis,
University
of
California,
San
Fran-
cisco
(1989).
17.
R.
Schule
et
al.,
Science
242,
1418
(1988).
18.
DNA
fragments
without
GREs
were
inserted
into
the
Xba
I
site
(see
Fig.
2)
of
-33
GRE6,
displacing
the
GRE
cassette
to
-262,
-366,
and
-1069
bp
from
the
start
of
transcription.
Weak
receptor-medi-
ated
activation
was
detected
when
GREs
were
situ-
ated
at
-262,
but
no
effect
was
observed
from
the
more
distal
sites.
19.
A.
Sergeant,
D.
Bohman,
H.
Zentgraf,
H.
Weihler,
W.
Keller,
J.
Mol.
Biol.
180,
577
(1984);
P.
Sassone-Corsi,
A.
Wilderman,
P.
Chambon,
Nature
313,
458
(1985);
H.
R.
Scholer
and
P.
Gruss,
EMBO
J.
4,
3005
(1985);
Y.-S.
Lin,
M.
F.
Carey,
M.
Ptashne,
M.
R.
Green,
Cell
54,
659
(1988).
20.
L.
P.
Freedman,
unpublished.
2I
JULY
I989
on August 4, 2009 www.sciencemag.orgDownloaded from
(Fig.
2b).
Affinity-purified
anti-APO-1
in-
duced
growth
retardation
and
cell
death
(Fig.
2c),
which
was
not
observed
with
either
an
isotype-matched,
control
MAb
(FII20)
[anti-MHC
(major
histocompatabi-
lity
complex)
class
I
antigens]
or
the
non-
binding
MAb
FI123.
Abrogation
of
[3H]thymidine
incorporation
along
with
in-
creased
trypan
blue
uptake
into
dead
cells
were
observed,
and
growth
of
104
SKW6.4
cells
in
200-,ul
cultures
was
blocked
by
more
than
95%
by
an
anti-APO-1
concentration
of
only
10
ng/ml
(Fig.
2c).
The
specificity
of
cell
death
induced
by
anti-APO-1
becomes
evident
from
the
fact
that
the
following
additional
control
MAbs
were
inactive
for
induction
of
apoptosis:
18
nonbinding
and
9
binding
MAbs
of
the
IgG3
isotype
(tested
by
immunofluorescence
on
SKW6.4
cells)
and
a
panel
of
MAbs
directed
against
known
antigens
on
the
cell
surface
of
SKW6.4
cells
including
CD
19,
CD20,
CD22,
MHC
class
II,
IgM
(immunoglobulin
M),
and
the
SKW6.4
idiotype
(6).
Cell
death
induced
by
anti-APO-1
was
complement-independent
and
occurred
un-
der
serum-free
culture
conditions
or
in
cul-
ture
medium
plus
serum
inactivated
at
56°C
for
30
min.
It
differed
from
death
mediated
by
complement-dependent
lysis
by:
(i)
mor-
phology
and
formation
of
a
DNA
ladder
(Fig.
2,
a
and
b),
(ii)
exogenous
Ca2+
independence
(7),
and
(iii)
delayed
trypan
blue
uptake
and
delayed
5'Cr
release
from
radiolabeled
target
cells
(8).
These
experi-
ments
indicate
that
cell
death
induced
by
anti-APO-
1
is
fumndamentally
different
from
antibody-
and
complement-dependent
cell
lysis.
To
assess
the
specificity
of
anti-APO-1,
a
restricted
panel
of
tumor
cell
lines
was
screened
for
expression
of
APO-1
and
sus-
ceptibility
to
growth
inhibition
and
apopto-
sis.
APO-
1
was
expressed
on
various
human
B
M
0
20
40
60
90
120
pK
Fig.
2.
Induction
of
growth
inhibition
and
apoptosis
by
anti-
APO-1.
(A)
The
T
cell
line
CCRF-CEM.S2
(19)
was
cultured
in
the
presence
of
purified
MAb
(1
,ug/ml)
in
a
microtiter
plate
for
2
hours
before
photography
(left
panel
control
MAb
13B1;
right
panel
anti-APO-1).
(B)
CCRF-CEM.S2
cells
(106
per
milliliter)
were
incubated
with
MAb
(1
p.g/mi)
in
culture
medium
at
37°C.
At
various
times,
aliquots
of
106
cells
were
100-
C
removed
and
DNA
was
prepared.
M,
marker;
I,
control
MAb
I3B1
for
2
hours;
lanes
3
to
7,
anti-APO-1
for
the
times
.^
0
0
indicated.
(C)
SKW6.4
cells
were
either
incubated
with
the
E
isotype-matched
control
MAb
FII20
(0),
FII23
(nonbinding
'
MAb)
(0),
or
anti-APO-1
(0)
in
microcultures
for
24
hours
°
before
labeling
with
[3H]thymidine
for
a
further
4
hours.
The
10i
data
represent
the
mean
of
duplicate
cultures
with
a
variation
of
a)
1
less
than
5%.
The
cells
were
cultured
in
RPMI
1640
medium
X
1
(Gibco,
Grand
Island,
New
York),
supplemented
with
2
mM
L-
a
1
glutamine,
streptomycin
(100
,ug/ml),
penicillin
(100
U/ml),
20
\
mM
Hepes
bufferpH
7.3,
and
10%
heat-inactivated
fetal
bovine
,
1
serum
(Conco
Lab-Division,
Wiesbaden,
FRG).
For
microcul-
M
tures,
1
x
104
cells
per
well
were
cultured
in
duplicates
in
flat-
.
\
bottom
96-well
microtiter
plates
(Tecnomara,
Fermwald,
FRG)
'
1
(200
,ul
final
volume
per
well).
After
24
hours,
the
cells
were
-.
labeled
with
0.5
,uCi
of
[3H]thymidine
(Amersham,
Braun-
0o11
schweig,
FRG)
for
4
hours.
Before
harvesting,
the
microcul-
1
01
102
103
104
tures
were
examined
bv
microscopic
inspection.
DNA
fragmen-
tation
1
x
106
cells
were
washed
with
cold
phosphate-buffered
MAb
(ng/ml)
saline
and
disrupted
with
NTE
buffer,
pH
8
(100
mM
NaC1,
10
mM
tris,
1
mM
EDTA)
containing
1%
SDS
and
proteinase
K
(0.2
mg/mi).
After
incubation
for
24
hours
at
37°C,
samples
were
extracted
twice
with
phenol
plus
chloroform
(1:
1,
v/v)
and
precipitated
by
ethanol.
The
DNA
was
dissolved
in
38
,ul
of
NTE
buffer
and
digested
with
ribonuclease
(1
mg/mi)
for
30
min
at
37C.
To
each
sample
10
RI1
of
loading
buffer
containing
15%
Ficoll
400
(Pharmacia,
Uppsala,
Sweden),
0.5%
SDS,
50
mM
EDTA,
0.05%
bromophenol
blue,
0.05%
xylene
cyanol
in
TBE
buffer
(2
mM
EDTA,
89
mM
boric
acid,
89
mM
tris,
pH
8.4)
were
added.
The
mixture
was
loaded
onto
a
1%
agarose
gel
and
stained
after
electrophoresis
with
ethidium
bromide
(0.5
,ug/ml).
The
size
marker
was
Hind
III
+
Eco
RI-digested
X
DNA.
2.5
E
0
2
(a
o
E
:
1.5
0
a)
@
1
-
E
a)
0.5
0
S
0
*
0
*
0
*
S.0
.
0
0
0
0
0
0
0
0 0
A%
0
oo
A
0
14
0
14
Time
after
injection
of
antibodies
(days)
Fig.
3.
Anti-APO-1-induced
regression
of
the
EBV-negative
Burkitt-like
lymphoma
BJAB
in
nu/nu
mice.
BJAB
cells
(4
x
107)
were
injected
subcutaneously
into
the
left
flank
of
nu/nu
mice.
After
5
weeks
(day
0)
the
mice
were
injected
with
500
,ug
of
MAb
into
the
tail
vein.
Control
MAb
FII20
(]);
FII23
(0);
I3B1
(A);
and
anti-APO-
1
(0).
Fourteen
days
later
the
size
of
the
tumors
was
measured
at
the
base
of
the
tumor;
the
tumors
from
individual
mice
are
represented
by
dots.
lymphoid
B
and
T
cell
lines
and
was
not
found
on
a
gibbon
or
mouse
T
cell
line
or
a
human
monocytic
cell
line
(Table
1).
Anti-
APO-1
blocked
proliferation
of
the
APO-1-
positive
cell
lines
listed
in
Table
1
via
induc-
tion
of
apoptosis,
and
formation
of
a
DNA
ladder
was
observed
in
each
case
(9).
Expres-
sion
of
APO-
1
was
not
restricted
to
cell
lines
in
vitro
but
could
be
found
on
leukemic
cells
freshly
isolated
from
patients
(Table
1).
Since
APO-
1
was
not
found
on
all
leukemic
cells
it
may
be
possible
that
anti-APO-l
defines
a
subpopulation
of
leukemias.
We
also
screened
human
B
and
T
cells
for
expression
of
APO-
1.
We
did
not
detect
APO-1
on
resting
B
cells.
However,
APO-1
was
expressed
on
activated
B
cells
(Table
1)
and
IgM
secretion
was
reduced
approxi-
mately
fourfold
by
3
days
of
treatment
with
anti-APO-1
(10).
Peripheral
resting
T
cells
did
not
express
APO-
1.
Activated
T
cells,
however,
expressed
APO-1
and
anti-APO-
1-induced
apoptosis
and
growth
inhibition
of
these
cells
(Table
1).
Thus,
our
data
suggest
that
APO-
1
is
a
species-specific
anti-
gen
expressed
on
activated
or
malignant
lymphocytes.
The
striking
effect
of
anti-APO-
1
in
vitro
prompted
us
to
test
its
effect
on
tumor
growth
in
vivo.
Although
the
Epstein-Barr
virus
(EBV)-negative,
Burkitt-like
lympho-
ma
BJAB
was
the
least
sensitive
to
anti-
APO-1
of
the
B
cell
panel
in
Table
1
and
expressed
only
approximately
1.5
x
104
APO-1
epitopes
per
cell
(4),
we
selected
BJAB
for
our
in
vivo
experiments.
The
reason
for
this
choice
was
that
only
BJAB
grew
to
large
tumor
masses
in
unirradiated
nu/nu
mice.
Five
weeks
after
injection
of
SCIENCE,
VOL.
24S
302
on August 4, 2009 www.sciencemag.orgDownloaded from
Fig.
4.
Localization
of
anti-
A
APO-1
in
xenografts
of
BJAB
1
2
3
in
nu/nu
mice
and
induction
of
apoptosis
of
the
tumor.
(A)
Up-
-
per
row:
uptake
of
'251-labeled
MAb
anti-APO-1
(50
p.g,
50
-
-
,uCi
per
mouse)
in
the
tumor
at
12
hours
(1),
48
hours
(2),
and
96
hours
(3)
after
intravenous
injection
of
the
MAb.
Lower
row:
uptake
of
I5-labeled
MAb
anti-APO-1
(500
,ug,
as
used
for
therapy;
50
,uCi
per
mouse)
(4)
and
FII20
(control
MAb,
binding
to
BJAB;
50
,tgr,
4
5
6
50
,uCi
per
mouse)
(5)
and
125-
labeled
FII23
(control
MAb,
nonbinding
to
BJAB;
50
,ug,
50
p.Ci
per
mouse)
(6)
in
48
hours,
respectively.
(B)
Ten
days
after
intravenous
injection
of
500
,ug
of
MAb
per
mouse
the
remaining
tumor
tissue
was
removed
and
fixed
with
formalin.
Paraffin
sections
of
the
tumor
were
stained
with
haematoxylin/eosin.
Left
panel,
tumor
after
treatmnent
with
control
MAb
FI120;
right
panel,
tumor
after
treatment
with
anti-APO-1.
Arrows
BJAB
cells
the
nu/nu
mice
carried
tumors
with
a
diameter
of
approximately
1.0
to
2.5
cm
(Fig.
3).
These
mice
were
injected
intra-
venously
with
purified
anti-APO-l
(500
p.g
per
mouse)
or
the
same
quantities
of
various
isotype-matched
control
antibodies
(FII20,
anti-MHC
class
I
antigens,
recognizing
5.8
x
105
sites
per
cell;
or
one
of
the
two
nonbinding
MAbs
F1123
and
13B1).
As
a
control
we
also
injected
anti-APO-1
(500
,ug
per
mouse)
into
three
nu/nu
mice
carrying
the
APO-1-negative
B
cell
tumor
OCILY1
with
tumor
diameters
of
1.5,
1.8,
and
3.4
cm,
respectively
(11)
(see
also
Table
1).
Two
days
after
anti-APO-1
injection,
a
whitish
discoloration
of
the
BJAB
tumors
was
observed
that
was
followed
by
rapid
tumor
regression.
Macroscopic
tumor
re-
gression
was
seen
in
10
of
11
treated
mice
within
less
than
14
days.
The
control
anti-
bodies
had
no
effect
(Fig.
3).
In
addition,
no
tumor
regression
was
observed
in
the
mice
carrying
OCI.LY1,
as
expected.
To
demonstrate
proper
localization
and
enrichment
of
the
injected
antibodies,
la-
beled
MAbs
were
visualized
by
autoradiog-
raphy
of
sections
of
the
BJAB
tumor
tissue
(Fig.
4a).
These
autoradiographs
showed
a
pronounced
binding
of
anti-APO-1
in
the
periphery
but
only
sparse
accumulation
in
the
center
of
the
tumor.
The
binding
control
MAb
FII20
showed
a
qualitatively
similar
binding
pattern.
There
was
no
localization
of
the
nonbinding
control
MAb
FI123
above
background.
Furthermore,
paired
la-
bel
experiments
(12)
with
labeled
anti-
APO-1
and
FII23
revealed
that
the
specific
enrichment
of
anti-APO-1
over
FI123
in
the
tumor
was
four-
and
sixfold
after
48
and
96
hours,
respectively.
The
main
purpose
of
our
experiments
was
to
assess
whether
anti-APO-
1
can
also
act
in
vivo.
Therefore,
the
tumor-bearing
mice
only
received
one
intravenous
injection
of
anti-APO-
1
at
a
dose
in
the
range
used
in
21
JULY
I989
indicate
host
vessels.
Final
magnification,
x
92.
MAbs
were
radioiodinated
according
to
the
IODO-Gen
method
(4).
Labeled
MAbs
were
injected
into
the
tail
vein
and
animals
were
killed
by
ether
anesthesia
at
the
predetermined
time
points.
The
tumors
were
excised
and
embedded
in
methylcellulose
and
20-,um
cryotome
sections
were
prepared.
Lyophilized
sections
were
placed
on
a
Kodak
X-omat
AR
film
for
autoradiography.
Table
1.
Reactivity
of
anti-APO-1
with
different
cells.
Cells
Relative
Effects
of
MAbs
on
Cells*
positive
fluorescence
[3H]thymidine
uptake
for
intensity
(103
cpm)t
APO-1
(anti-APO-1/
Type
Designation
(%)t
control)
Control
Anti-APO-1
Malignant
cell
lines
Hu
B
cells
SKW6.4
98
11.1
30.0
0.02
CESS
95
12.0
23.0
0.1
BJAB
80
2.1
70.0
7.0
OCI.LY1
0
1
14.5
15.8
Hu
T
cells
Jurkat
83
2.3
20.3
8.1
Molt
91
2.4
35.7
0.6
CCRF-CEM
64
1.9
16.2
0.5
Hu
myeloid
cells
U937
5
0.97
62.2
60.5
Gibbon
T
cells
MLA
144
0
0.96
34.3
35.0
Mouse
T
cells
EL4
0
1
44.8
45.3
Leukemic
cells
from
patients§
PreT-ALL
B.M.
54
4.4
T-ALL
D.A.
53
3.2
Common
ALL
W.N.
72
5.0
Normal
human
lymphocytes
T
cellsll
Resting
3
1.36
Activated
89
7.4
22.4
0.23
B
cells¶
Resting
0
0.9
Activated
91
1.1
*Hu,
human;
ALL,
acute
lymphocytic
leukemia.
tAliquots
of
106
cells
were
incubated
at
4°C
in
100
,ul
of
medium
with
control
MAb
(RF23
or
I3B1)
or
anti-APO-
1
for
30
min.
Then
the
cells
were
washed
and
stained
with
fluorescein
isothiocyanate-coupled
goat
anti-mouse
Ig
F(ab')2(70
iLg/ml)
and
analyzed
by
a
cytofluorograph
(Ortho
Diagnostic
Systems,
Westwood,
Massachusetts).
tCells
(104
per
well)
were
cultured
in
the
presence
of
MAb
(500
ng/ml)
for
24
hours
and
labeled
with
[3H]thymidine
for
2
hours
before
harvest;
the
data
represent
the
mean
of
duplicate
cultures
with
a
variation
of
less
than
5%.
§Bone
marrow
cells
isolated
from
the
patients
were
mOrphologicay
>95%
blasts
and
showed
the
foHowing
phenotype:
pre
T-ALL,
cytoplasmic
CD3+, CD5+,
CD7+,
CD34,
Tdt+,
CD2-,
surface
CD3-,
CD4-,
and
CD8-;
T-ALL
CD2',
cytoplasmic
CD3',
CD5',
CD7'
and
Tdt+,
surface
CD3-,
CD4-,
CD8-,
and
CD34-;
common-ALL
CD10+,
CD19+,
CD22+,
CD24+,
CD20-.
The
effect
of
anti-APO-1
on
these
leukemic
cells
was
not
tested,
because
they
died
under
normal
culture
conditions.
IIPeripheral
blood
mononuclear
cells
(PBMC)
from
healthy
volunteers
were
isolated
by
Ficoll
Paque
(Pharmacia
Inc.,
Uppsala,
Sweden)
density
centrifugation.
Adherent
cells
were
removed
by
adherence
to
plastic
culture
vessels
overnight.
T
cells
were
isolated
from
PBMC
by
rosetting
with
2
amino-ethylisothyouronium-bromide
(AET)-treated
she
red
blood
cells
as
described
(20).
Freshly
prepared
resting
T
cells
(2
x
10'
per
milliliter;
96%
OKT11+,
1%
Tac)
were
activated
with
phytohemagglutinin-M
(50
kg/mn)
and
PMA
(10
ng/nil)
(Sigma
Chemical
Co.,
Munich,
FRG).
Two,
7,
and
12
days
later
the
T
cells
were
fed
with
20
to
30
U/mi
of
recombinant
human
interleukin-2
(20
to
30
U/mi).
T
cells
(5
x
105
per
milliliter)
activated
for
12
days
(90%
OKT11+;
60%
Tac+)
were
cultured
in
the
presence
of
FI123
or
anti-APO-1
(1
pLg/ml)
in
triplicates
for
24
hours
and
then
labeled
with
[3H]thymidine
for
a
further
17
hours
(see
legend
to
Fig.
2).
¶Resting
B
cells
(35.8%
CD19')
were
isolated
by
two
rounds
of
rosetting
as
above,
followed
b:
separation
via
a
Sephadex
G-10
column
as
described
(21).
For
activated
B
cells,
PBMC
were
adjusted
to
2
x
10
cells
per
milliliter
and
cultured
in
the
presence
of
pokeweed
mnitogen
at
10
j±g/ml
(Serva,
Heidelberg,
FRG)
for
6
days.
Dead
cells
and
T
cells
were
then
eliminated
by
rosetting
wit-h
AET-treated
sheep
red
blood
cells
and
subsequent
centrifugation
over
Ficoll
Paque.
The
interphase
cells
were
used
as
activated
B
cells
(84%
sIgM+).
MAb
therapies.
In
other
therapy
schedules,
tumor
was
observed
in
three
of
the
ten
mice
however,
MAbs
are
injected
repeatedly
(13).
in
which
tumor
regression
had
been
ob-
In
our
experiments
regrowth
of
the
BJAB
served
(Fig.
3).
Regrowth
was
observed
at
REPORTS
303
on August 4, 2009 www.sciencemag.orgDownloaded from
the
margin
of
the
original
tumor
approxi-
mately
3
months
after
the
initial
macroscop-
ic
tumor
regression.
One
of
these
tumors
was
removed
and
found
to
express
APO-1
by
immunofluorescence
and
to
be
sensitive
to
anti-APO-l
in
vitro
at
a
MAb
concentra-
tion
similar
to
the
original
in
vitro
BJAB
tumor
cell
line
(Table
1).
To
determine
the
histology
of
the
regress-
ing
BJAB
tumors
we
prepared
thin
sections
of
tumors
from
MAb-treated
nu/nu
mice.
Ten
days
after
intravenous
injection
of
F1120,
BJAB
appeared
as
a
solid
tumor
composed
of
densely
packed
large
blasts
with
numerous
mitoses,
some
tumor
giant
cells,
and
rare
apoptotic
figures
(Fig.
4b,
left
panel).
The
tumor
was
penetrated
by
host
vessels.
In
contrast,
almost
all
remaining
BJAB
cells
of
mice
treated
with
anti-APO-1
(Fig.
4b,
right
panel)
showed
severe
cyto-
pathic
changes
including
nudear
pycnosis
and
cellular
edema
most
pronounced
in
perivascular
microareas.
These
morphologi-
cal
changes
are
characteristic
of
apoptosis.
Taken
together,
these
data
strongly
sug-
gest
that
apoptosis
is
induced
by
anti-APO-
1
and
is
the
mechanism
of
death
and
regres-
sion
of
BJAB
tumor
cells
in
vivo.
The
fact
that
FII20,
which
strongly
binds
to
the
cell
surface
of
BJAB
tumor
cells,
did
not
cause
regression
of
BJAB
also
precludes
the
possi-
bility
that
killer
cells
or
complement
that
might
have
bound
to
anti-APO-1
may
have
been
involved
in
the
growth
inhibition
and
tumor
regression.
We
showed
that
anti-APO-1
specifically
blocked
growth
and
triggered
programmed
cell
death
(apoptosis)
of
a
set
of
activated
normal
lymphocytes
and
cells
from
malig-
nant
lymphocyte
lines
after
binding
to
the
cell
surface
protein
antigen
APO-
1.
Recent-
ly,
it
has
been
shown
that
anti-CD3
induces
apoptosis
of
immature
thymocytes
in
vitro
(14).
Therefore,
it
has
been
suggested
that
CD3-triggered
apoptosis
might
be
responsi-
ble
for
negative
selection
of
T
cells
in
the
thymus.
Since
APO-1
is
expressed
on
ma-
ture
activated
lymphocytes,
additional
ex-
periments
will
be
needed
to
determine
whether
the
antigen
might
play
a
similar
role
in
the
downregulation
of
the
immune
response
and
be
involved
in
selection
and
elimination
of
lymphocytes.
It
has
previous-
ly
been
shown
that
LT,
TNF,
and
killer
cells
with
their
effector
molecules
induce
apopto-
tic
cell
death
(15).
As
anti-APO-1
also
induces
apoptosis
a
number
of
possibilities
might
be
considered
for
the
physiological
role
of
the
APO-1
antigen.
APO-1
might
be
a
receptor
for
cytotoxic
molecules
or
for
autocrine
growth
factors.
Alternatively,
it
could
be
a
molecule
essential
for
vertical
or
lateral
growth
signal
transduction.
Thus,
anti-APO-1
might
trigger
receptors
for
lytic
molecules
or
block
receptors
for
growth
signals.
Apoptosis
is
found
in
all
tissues
and
also
in
cells
from
lower
organisms
(16).
It
is
conceivable,
therefore,
that
several
distinct
cell
surface
antigens
with
a
different
tissue
distribution
are
involved
in
the
induction
of
apoptosis.
Elucidation
of
the
structure
of
APO-1,
its
possible
connection
to
the
cyto-
skeleton
and
the
molecular
events
following
anti-APO-1
binding
might
resolve
some
of
these
issues.
Our
data
might
also
have
clinical
rele-
vance.
APO-
1
was
found
on
some
lymphoid
tumor
cells
freshly
isolated
from
patients.
Thus,
anti-APO-1
might
be
usefufl
as
a
diagnostic
tool
to
define
subsets
of
normal
and
malignant
lymphocytes.
In
addition,
induction
of
apoptosis
may
have
implica-
tions
for
anti-tumor
therapy.
Antibodies
have
frequently
been
used
as
heteroconju-
gates
with
toxins
or
drugs
to
destroy
tumor
cells
(17).
Our
data,
however,
show
that
MAb
alone
can
be
lethal
to
target
cells.
Anti-APO-1
and
related
MAbs
might,
therefore,
be
considered
for
ex
vivo
or
in
vivo
therapy,
under
conditions
where
reac-
tivity
with
vital
normal
cells
can
be
excluded
or
tolerated.
Finally,
the
molecular
investi-
gation
of
cell
death
induced
by
anti-APO-1
might
lead
to
a
general
understanding
of
apoptosis.
In
this
case,
the
use
of
modified
or
normal
physiological
ligands
to
the
cell
surface
antigen
initiating
apoptosis
or
of
chemicals
interfering
with
the
apoptotic
sig-
nal
might
be
envisaged.
REFERENCES
AND
NOTES
1.
E.
Duvall
and
A.
H.
Wylie,
Immunol.
Today
7,
115
(1986).
2.
A.
H.
Wylie,
J.
F.
R.
Kerr,
A.
R.
Currie,
Int.
Rev.
Cytol.
68,
251
(1980).
3.
BALB/c
mice
were
immunized
once
per
week
over
a
4-week
period
by
intraperitoneal
injection
of
1
x
107
SKW6.4
cells.
Four
days
after
the
last
injection,
spleen
cells
from
immunized
animals
were
fused
with
the
P3.X63.Ag8.653
myeloma
[G.
Koh-
ler
and
C.
Milstein,
Nature
256,
495
(1975)].
Twelve
days
after
fusion
culture
supematants
from
wells
positive
for
growth
were
tested
for
their
ability
to
inhibit
growth
of
SKW6.4
cells.
Hybridomas
that
produced
blocking
MAbs
were
cloned
three
times
by
limiting
dilution
at
a
concentration
of
0.5
cells
per
well.
MAbs
were
purified
from
serum-free
cul-
ture
supernatant
by
means
of
a
protein
A-Diasorb
column
(Diagen,
Dusseldorf,
FRG).
Bound
MAbs
were
eluted
with
0.1M
NaCl
and
0.1M
glycine,
pH
2.8,
dialyzed
against
phosphate-buffered
saline
and
sterilized.
The
isotype
of
the
MAbs
was
determined
by
enzyme-linked
immunosorbent
assay
[S.
Kiesel,
et
al.,
Leuk.
Res.
11,
1119,
1987)
with
isotype-
specific
goat
anti-mouse
Ig
that
had
been
conjugated
with
horseradish
peroxidase
(Dunn,
Asbach,
FRG).
4.
Affinity
and
number
of
anti-APO-1
binding
sites
per
cell
were
determined
by
Scatchard
analysis
as
described
[I.
von
Hocgen,
W.
Falk,
G.
Kojouharoff,
P.
H.
Krammer,
Eur.
J.
Immunol.
19,
329
(1989)].
Briefly,
MAbs
were
iodinated
by
the
IODO-Gen
method
[P.
J.
Fraken
and
J.
C.
Speck,
Biochem.
Biophys.
Res.
Commun.
80,
849
(1980)].
Aliquots
of
5
x
10'
cells
were
resuspended
in
200
,ul
of
culture
medium
containing
0.1%
NaN3
and
different
con-
centrations
of
"LI-labeled
MAbs.
After
incubation
at
4°C
for
4
hours,
two
95-pAl
portions
were
re-
moved
and
centrifuged
as
described
above
by
von
Hoegen
et
al.
5.
A.
H.
Wyllie,
Nature
284,
555
(1980).
6.
Monoclonal
anti-CD19
(HD37)
and
anti-CD22
(HD39)
were
kindly
provided
by
B.
Dorken
(Poli-
clinic
of
the
University,
Heidelberg,
FRG)
and
monoclonal
anti-CD20
by
G.
Moldenhauer
(IV
Leukocyte
typing
workshop
and
conference,
Vien-
na,
Austria,
1989),
respectively.
The
18
nonbinding
and
9
binding
MAbs
of
the
IgG
3
isotype
(tested
by
inmmunofluorescence
on
SKW6.4
cells)
and
the
MAbs
directed
against
MHC
class
II,
IgM,
and
SKW6.4
Ig
idiotypes
were
raised
in
our
own
labora-
tory.
7.
The
kinetics
of
membrane
blebbing
induced
by
anti-APO-1
(within
30
min;
Fig.
2a)
was
not
influenced
by
the
presence
of
10
mM
EDTA
or
EGTA.
In
addition,
endonudease-mediated
DNA
fragmentation
induced
by
anti-APO-1
was
not
in-
hibited
by
the
Ca2"
channel
blockers
Furamicin
(50
FM)
or
Nifedipin
(50
;.M).
8.
When
51Cr-labeled
SKW6.4
cells
were
incubated
with
anti-APO-1
(1
pg/ml)
for
2,
4,
8,
and
24
hours,
the
specific
5"Cr
release
[R.
C.
Duke,
R.
Chervenak,
J.
J.
Cohen,
Proc.
Nati.
Acad.
Sci.
U.S.
A.
80,
6361
(1983)]
was
found
to
be
2.9%,
7.6%,
21.3%,
and
32.5%,
respectively.
Trypan
blue
uptake
was
measured
at
the
same
time
points:
2.5%,
4.7%,
10.6%,
and
73.6%,
respectively,
of
the
cells
were
trypan
blue-positive.
In
contrast,
2
hours
after
the
addition
of
MAbs,
plus
complement
the
specific
51Cr
release
was
108.7%
and
92.7%
of
the
cells
stained
with
trypan
blue.
9.
Two
hours
after
addition
of
MAbs
(1
Rg/ml)
the
genomic
DNA
of
each
tumor
line
was
isolated
and
analyzed
on
agarose
gels
as
described
(Fig.
2).
Inhibition
of
[3H]thymidine
uptake
by
anti-APO-1
was
paralleled
by
fragmentation
of
the
genomic
DNA.
This
was
not
observed
after
treatment
with
control
MAb
(13B1).
10.
Activated
B
cells
(106
per
milliliter)
were
incubated
in
the
presence
of
MAb
FI123
or
anti-APO-1
at
1
ug/mi.
After
3
days
the
culture
supematants
were
collected
and
the
IgM
concentration
measured
with
a
human
IgM-specific
ELISA
containing
HRPO-
conjugated
goat
anti-human
IgM
(Medac,
Ham-
burg,
FRG).
IgM
secretion
after
treatment
with
FII23
or
anti-APO-1
was
2100
and
550
ng/ml,
respectively.
11.
OCI-LY1
was
obtained
from
H.
Messner,
Ontario
Cancer
Institute,
Toronto,
Canada.
12.
D.
Pressman
et
al,
Cancer
Res.
17,
845
(1957).
13.
S.
L.
Brown
et
al.,
Blood
73,
651
(1989).
14.
C.
A.
Smith
et
al.,
Nature
337,
181
(1989).
15.
D.
S.
Schmid,
J.
P.
Tite,
N.
H.
Ruddle,
Proc.
Natl.
Acad.
Sci.
U.S.A.
83,
1881
(1986);
G.
B.
Dealtry,
M.
S.
Naylor,
W.
Fiers,
F.
R.
Balkwill,
Eur.
J.
Immunol.
17,
689
(1987);
M.
M.
Don
et
al.,
Aust.
J.
Exp.
Biol.
Med.
Sci.
55,
407
(1977);
C.
J.
Sanderson,
Biol.
Rev.
56,
153
(1981);
J.
H.
Russell
and
C.
B.
Dobos,
J.
Immunol.
125,
1256
(1980);
D.
M.
Howell
and
E.
J.
Martz,
Immunology
140,
689
(1988);
J.
C.
Hiserodt,
L.
J.
Britvan,
S.
R.
Tag
Targan,
J.
Immunol.
129,
1782
(1982);
J.
D.-E.
Young
and
C.-C.
Liu,
Immunol.
Today
9,
140
(1988).
16.
F.
Giorgi
and
P.
J.
Deri,
Embryol.
Exp.
Morphol.
35,
521
(1976).
17.
E.
S.
Vitetta
et
al.,
Science
219,
644
(1983).
18.
V.
K.
Laemmli,
Nature
277,
680
(1970).
19.
The
CCRF-CEM.S2
subclone
was
obtained
by
doning
cells
under
limiting
dilution
conditions
from
the
CCRF-CEM
T
cell
line
at
one
cell
per
well
in
96-
well
microtiter
plates.
CCRF-CEM.S2
was
selected
because
of
its
high
sensitivity
to
programmed
cell
death
induced
by
anti-APO-1
(500
ng/ml)
as
mea-
sured
by
microscopic
inspection
in
a
4-hour
culture.
20.
M.
Madsen
et
al.,
J.
Immunol.
Methods
33,
323
(1980);
M.
A.
Pellegrino
et
al.,
Clin.
Immunol.
Immunopathol.
3,
324
(1975).
21.
T.
R.
Jerrells,
J.
H.
Dean,
G.
L.
Richardson,
D.
B.
Hcrberman,
J.
Immunol.
Methods
32,
11
(1980).
22.
We
thank
K.
Hexel,
J.
K6llner,
R.
Kuihnl,
C.
Mandl,
and
W.
Muller
for
excellent
technical
assistance;
H.
Sauter
for
excellent
secretarial
assistance;
G.
Ham-
merling
and
G.
Moldenhauer
for
their
criticism;
B.
SCIENCE,
VOL.
245
304-
on August 4, 2009 www.sciencemag.orgDownloaded from
gie,
Bonn,
the
Ministerium
fuir
Wissenschaft
und
Kunst,
Stuttgart,
and
the
Tumor
Center
Heidel-
berg/Mannheim,
FRG.
A.M.J.P.
was
supported
by
the
Boehringer Ingelheim
Fonds,
Stuttgart,
FRG.
3
February
1989;
accepted
19
May
1989
The
Reservoir
for
HIV-1
in
Human
Peripheral
Blood
Is
a
T
Cell
That
Maintains
Expression
of
CD4
STEVEN
M.
SCHNITrMAN,*
MILTIADES
C.
PSALLIDOPOULOS,
H.
CLIFFORD
LANE,
Louis
THOMPSON,
MICHAEL
BASELER,
FERDINAND
MASSARI,
CECIL
H.
Fox,
NoRMAN
P.
SALZMAN,
ANTHONY
S.
FAUCI
Human
immunodeficiency
virus
type
1
(HIV-1)
selectively
infects
cells
expressing
the
CD4
molecule,
resulting
in
substantial
quantitative
and
qualitative
defects
in
CD4+
T
lymphocyte
function
in
patients
with
acquired
immunodeficiency
syndrome
(AIDS).
However,
only
a
very
small
number
of
cells
in
the
peripheral
blood
of
HIV-
1-infected
individuals
are
expressing
virus
at
any
given
time.
Previous
studies
have
demonstrated
that
in
vitro
infection
of
CD4+
T
cells
with
HIV-1
results
in
downregulation
of
CD4
expression
such
that
CD4
protein
is
no
longer
detectable
on
the
surface
of
the
infected
cells.
In
the
present
study,
highly
purified
subpopulations
of
peripheral
blood
mononuclear
cells
(PBMCs)
from
AIDS
patients
were
obtained
and
purified
by
fluorescence-automated
cell
sorting.
They
were
examined
with
the
methodologies
of
virus
isolation
by
limiting
dilution
analysis,
in
situ
hybridization,
immunofluorescence,
and
gene
amplification.
Within
PBMCs,
HIV-1
was
expressed
in
vivo
predominantly
in
the
T
cell
subpopulation
which,
in
contrast
to
the
in
vitro
observations,
continued
to
express
CD4.
The
precursor
frequency
of
these
HIV-l1-expressing
cells
was
about
1/1000
CD4+
T
cells.
The
CD4+
T
cell
population
contained
HIV-1
DNA
in
all
HIV-
1-infected
individuals
studied
and
the
frequency
in
AIDS
patients
was
at
least
1/100
cells.
This
high
level
of
infection
may
be
the
primary
cause
for
the
progressive
decline
in
number
and
finction
of
CD4+
T
cells
in
patients
with
AIDS.
T
HHE
HUMAN
IMMUNODEFICIENCY
virus
type
1
(HIV-1),
the
etiologic
agent
of
the
acquired
immunodefi-
ciency
syndrome
(AIDS),
selectively
infects
cells
expressing
the
CD4
molecule,
includ-
ing
T
lymphocytes
and
cells
of
the
mono-
cyte/macrophage
lineage
(1).
In
vitro
infec-
tion
of
cells
with
HIV-1
results
in
a
de-
creased
expression
of
the
CD4
molecule
on
the
surface
of
the
infected
cells
(2).
Patients
with
AIDS
have
severe
depres-
sion
of
the
normal
cell-mediated
immune
mechanisms
that
is
partially
attributed
to
the
considerable
depletion
of
CD4
lymphocytes
(3).
Despite
this,
examination
of
cells
from
lymph
nodes
and
peripheral
blood
from
patients
with
AIDS
and
AIDS-related
com-
plex
(ARC)
has
revealed
a
very
low
frequen-
cy
of
viral
RNA
synthesis,
generally
occur-
ring
in
1/100,000
to
1/10,000
of
total
mononuclear
cells
(4).
However,
it
is
possi-
ble
that
a
larger
proportion
of
cells
may
be
latently
infected
(containing
proviral
DNA
but
not
expressing
viral
mRNA
or
protein).
Until
the
development
of
gene
amplification
[polymerase
chain
reaction
(PCR)]
method-
ology
(5,
6),
HIV-1-infected
cells
not
ex-
pressing
virus
were
not
readily
detectable
by
available
techniques.
In
the
present
study,
blood
was
obtained
from
HIV-1
culture-positive
patients
with
AIDS
either
directly
in
heparinized
syringes
or
via
apheresis
and
subjected
to
Ficoll-
Hypaque
separation
(7).
First,
peripheral
blood
mononuclear
cells
(PBMCs)
from
pa-
tients
were
stained
with
fluorescein
isothio-
cyanate
(FITC)-conjugated
antibody
to
CD3
and
sorted
by
a
fluorescence-activated
cell
sorter
(FACS)
into
CD3+
and
CD3-
populations.
Sorted
cells
were
cocultivated
with
an
excess
of
normal
phytohemaggluti-
nin
(PHA)-stimulated
blast
cells
and
we
determined
the
time
to
peak
viral
expres-
sion,
a
highly
consistent
and
reproducible
parameter
of
viral
expression.
A
predom-
inance
of
HIV-1
expression
in
the
>98%
enriched
CD3+
population,
as
determined
by
the
time
to
peak
syncytia
formation
(Fig.
1A)
and
reverse
transcriptase
(RT)
activity
(Fig.
1B),
was
seen.
Similar
results
were
obtained
in
seven
additional
AIDS
patients.
Delayed
expression
of
HIV-1
in
cells
that
were
initially
99%
CD3-
cells
(Fig.
1B)
was
due
to
outgrowth
of
the
few
contaminating
CD3+
cells.
Phenotypic
analysis
of
nonco-
cultivated
enriched
CD3-
cells
grown
wuder
the
same
conditions
revealed
that
35
to
65%
of
the
cells
were
CD3+
by
day
10
in
culture.
In
the
second
series
of
experiments,
PBMCs
from
AIDS
patients
were
double-
stained
with
FITC-conjugated
anti-CD3
and
anti-CD4
and
sorted
by
FACS
into
CD3+/CD4+
and
CD3+/CD4-
popula-
tions.
These
sorted
cells
were
cocultivated
with an
excess
of
normal
PHA-stimulated
blast
cells
and
showed
a
predominance
of
HIV-
1
expression
in
the
highly
enriched
(98
to
99%)
CD4+
T
cell
population
as
deter-
mined
by
the
time
to
peak
syncytia
forma-
tion
(Fig.
IC)
and
RT
activity
(Fig.
ID).
Similar
results
were
obtained
in
seven
addi-
tional
AIDS
patients.
The
phenotypic
analy-
sis
of
freshly
sorted
CD3+/CD4+
cells
re-
vealed
a
greater
than
98
to
99%
CD4+
purity
in
most
experiments
when
stained
with
the
monoclonal
antibody
to
Leu
3a.
Again,
the
delayed
expression
of
HIV-1
in
cells
that
were
initially
99%
CD4-
(Fig.
ID)
was
most
likely
due
to
outgrowth
of
a
few
contaminating
CD4+
T
cells.
Phenotyp-
ic
analysis
of
non-cocultured
enriched
CD4-
T
cells
grown
under
the
same
condi-
tions
revealed
that
30
to
55%
of
the
cells
were
CD4+
by
day
10
in
culture.
In
situ
hybridization
for
HIV-1
viral
RNA
was
then
performed
at
time
zero
on
the
highly
enriched
CD3+/CD4+-
and
CD3+/CD4--sorted
PBMCs.
There
was
a
predominance
of
viral
expression
in
the
CD4+
T
cell
population
at
a
frequency
of
about
1/1000
cells
in
four
AIDS
patients
(X
±
SEM
per
1000
cells
was
0.95
±
0.21)
(Fig.
2A).
This
is
in
comparison
to
a
level
of
viral
expression
in
the
CD4-
T
cell
popula-
tion
of
<1/100,000
cells
(Fig.
2B),
which
is
equivalent
to
background
signal
in
controls.
The
frequency
of
in
situ-positive
CD4+
T
cells
remained
unchanged
in
three
of
the
patients
reexamined
at
6
to
12
months
after
the
initial
studies.
Indirect
immunofluorescence
studies
for
HIV-
1
viral
antigens
was
also
performed
at
time
zero
on
highly
enriched
CD3+/CD4+-
and
CD3+/CD4--sorted
PBMCs.
These
demonstrate
a
predominance
of
viral
expres-
sion
in
the
CD4+
T
cell
population
at
a
frequency
of
about
1/1000
cells
in
four
AIDS
patients
(X
±
SEM
per
1000
cells
was
1.10
±
0.35
(Fig.
2C).
This
is
in
com-
parison
to
a
level
of
viral
expression
in
the
CD4-
T
cell
population
of
<
1/10,000
cells
S.
M.
Schnittman,
H.
C.
Lane,
F.
Massari,
C.
H.
Fox,
A.
S.
Fauci,
Laboratory
of
Immunoregulation,
National
Institute
of
Allergy
and
Infectious
Diseases,
National
Institutes
of
Health,
Bethesda,
MD
20892.
M.
C.
Psallidopoulos,
L.
Thompson,
N.
P.
Salzman,
Division
of
Molecular
Virology
and
Immunology,
Georgetown
University
School
of
Medicine,
Washing-
ton,_DC
20007.
M.
Baseler,
Program
Resources,
Incorporated,
Freder-
ick,
MD
21701.
*To
whom
correspondence
should
be
addressed.
REPORTS
30S
Dorken
and
G.
Moldenhauer
for
providing
MAbs,
and
Dr.
H.
Messner
for
providing
tumor
cells.
We
also
thank
D.
Scheppelmann
and
H.-P.
Meinzer
for
the
computerized
image
analysis
of
normal
and
apoptotic
cells.
Supported
by
grants
from
the
Bundesministerium
fir
Forschung
und
Technolo-
21
JULY
I989
on August 4, 2009 www.sciencemag.orgDownloaded from