Stimulation of Proliferation, Differentiation, and Function of Human Cells by Primate Interleukin 3

Abstract

Cloned gibbon interleukin 3 (gIL-3) was found to stimulate the proliferation and differentiation of human bone marrow cells to produce day-14 granulocyte, macrophage, granulocyte-macrophage, and eosinophil colonies in semisolid agar. In the presence of normal human plasma, gIL-3 stimulated megakaryocytes. In methylcellulose cultures, it stimulated erythroid colonies in the presence, but not in the absence, of erythropoietin. When mature human leukocytes were used, gIL-3 stimulated the function of purified mature eosinophils as measured by the capacity to kill antibody-coated target cells, to produce superoxide anions, and to phagocytize opsonized yeast particles in a manner similar to recombinant human granulocyte-macrophage colony-stimulating factor. In contrast, gIL-3 did not significantly stimulate any of the neutrophil functions tested, whereas human recombinant granulocyte-macrophage colony-stimulating factor was active in these assays. Among cytokines that are active on human hematopoietic cells, gIL-3 thus has a distinct set of functions and may predict the range of actions of the human molecule.

Full text

Proc.
Natl.
Acad.
Sci.
USA
Vol.
84,
pp.
2761-2765,
May
1987
Cell
Biology
Stimulation
of
proliferation,
differentiation,
and
function
of
human
cells
by
primate
interleukin
3
(colony-stimulating
factor/eosinophil)
A.
F.
LOPEZ*,
L.
B.
To*,
YU-CHUNG
YANGt,
J.
R.
GAMBLE*,
M.
F.
SHANNON*,
G.
F.
BURNS*,
P.
G.
DYSON*,
C.
A.
JUTTNER*,
S.
CLARKt,
AND
M.
A.
VADAS*
*Divisions
of
Human
Immunology
and
Haematology,
Institute
of
Medical
and
Veterinary
Science,
Box
14,
Rundle
Mall
Post
Office,
Adelaide
5000,
South
Australia,
Australia;
and
tGenetics
Institute
Inc.,
Cambridge,
MA
02140
Communicated
by
J.
F.
A.
P.
Miller,
December
18,
1986
ABSTRACT
Cloned
gibbon
interleukin
3
(gIL-3)
was
found
to
stimulate
the
proliferation
and
differentiation
of
human
bone
marrow
cells
to
produce
day-14
granulocyte,
macrophage,
granulocyte-macrophage,
and
eosinophil
colo-
nies
in
semisolid
agar.
In
the
presence
of
normal
human
plasma,
gIL-3
stimulated
megakaryocytes.
In
methylcellulose
cultures,
it
stimulated
erythroid
colonies
in
the
presence,
but
not
in
the
absence,
of
erythropoietin.
When
mature
human
leukocytes
were
used,
gIL-3
stimulated
the
function
of
purified
mature
eosinophils
as
measured
by
the
capacity
to
kill
anti-
body-coated
target
cells,
to
produce
superoxide
anions,
and
to
phagocytize
opsonized
yeast
particles
in
a
manner
similar
to
recombinant
human
granulocyte-macrophage
colony-stimu-
lating
factor.
In
contrast,
gIL-3
did
not
significantly
stimulate
any
of
the
neutrophil functions
tested,
whereas
human
recom-
binant
granulocyte-macrophage
colony-stimulating
factor
was
active
in
these
assays.
Among
cytokines
that
are
active
on
human
hematopoietic
cells,
gIL-3
thus
has
a
distinct
set
of
functions
and
may
predict
the
range
of
actions
of
the
human
molecule.
The
murine
(m)
cytokine
interleukin-3
(IL-3)
(1,
2),
also
known
as
multipotential
colony-stimulating
factor
(3)
and
persistent
cell-stimulating
factor
(4),
is
distinct
among
the
cloned
hematopoietic-stimulating
factors
in
having
the
ca-
pacity
to
stimulate
progenitor
cell
renewal.
This,
as
well
as
its
relative
lack
of
lineage
restriction,
suggested
that
mIL-3
is
active
at
a
more
primitive
level
than
the
colony-stimulating
factors
(CSF)
granulocyte-macrophage
(GM)-CSF,
macro-
phage
(M)-CSF,
granulocyte
(G)-CSF,
or
eosinophil
differ-
entiation
factor
(EDF)
and
that
it
may
have
a
special
role
in
leukemogenesis
(5).
We
have
cloned
the
gene
encoding
a
molecule
active
in
hematopoiesis
from
a
gibbon
(g)
cell
line
that
is
29%o
homol-
ogous
at
the
amino
acid
level
and
has
a
genomic
organization
similar
to
that
of
mIL-3
(6).
These
properties
led
to
the
designation
of
this
molecule
as
gIL-3.
In
this
communication
we
describe
some
of
the
biological
effects
of
gIL-3
and
show
that,
to
our
knowlege,
its
spectrum
of
activities
on
human
cells
differs
from
all
described
hema-
topoietic
factors.
METHODS
Cytokines.
The
gIL-3
used
was
in
a
COS
cell-conditioned
medium
obtained
by
transfecting
COS-1
cells
with
5
,ug
of
plasmid
DNA
containing
the
gIL-3
cDNA
(pMLA-CSF)
and
harvesting
the
gIL-3-containing
medium
72
hr
after
transfec-
tion
(6).
Recombinant
human
(rh)
GM-CSF,
with
endotoxin
at
<0.2
ng/ml,
was
purified
from
the
conditioned
medium
of
COS
cells
that
had been
transfected
with
cloned
human
GM-CSF
cDNA
in
the
p91203(B)
vector
as
described
(7).
Silver-
staining
of
the
NaDodSO4/polyacrylamide
gels
of
the
puri-
fied
GM-CSF
revealed
a
major
band
of
19
kDa.
rh
tumor
necrosis
factor
type
a
was
a
gift
from
Genentech,
(South
San
Francisco,
CA)
and
contained
cytotoxic
activity
(3.6
x
107
units/mg)
on
actinomycin
D-treated
L929
mouse
fibroblast
cells
and
endotoxin
at
0.8
ng/ml.
rh
tumor
necrosis
factor
type
a
was
produced
in
Escherichia
coli
(8)
and
purified
to
99.8%
purity.
Bone
Marrow
Cultures.
Erythroid
colony-forming
unit
assay.
Light-density
nonadherent
bone
marrow
cells
were
obtained
by
separation
on
a
Ficoll/Paque
(Pharmacia,
Swe-
den)
density
gradient
followed
by
a
60-min
incubation
with
carbonyl
iron
[200
mg
of
carbonyl
iron
per
15
x
106
cells
in
10
ml
of
RPMI
with
15%
(vol/vol)
fetal
calf
serum
(FCS)]
and
removal
of
monocytes
(containing
attached
or
phagocytized
iron
particles)
with
a
magnet.
Cells
were
cultured
in
0.9%
methylcellulose
(Fluka,
Sweden)
with
Iscove's
modified
Dulbecco's
medium
(Commonwealth
Serum
Laboratories,
Australia),
30%
(vol/vol)
FCS
(GIBCO),
0.66%
bovine
serum
albumin
(fraction
V,
Sigma),
and
20
,uM
2-mercaptoethanol
at
a
concentration
of
5
x
104
cells
per
ml
of
culture
medium.
The
cultures
were
stimulated
with
1
unit
of
high-purity
human
urinary
erythropoietin
(EPO)
per
plate
and
phytohemag-
glutinin-stimulated
leukocyte-conditioned
medium
(PHA-
LCM)
[5%
(vol/vol)],
rhGM-CSF
(100
ng/ml),
or
gIL-3
(1:200
dilution).
Control
cultures
with
no
added
stimulus
or
with
EPO
alone
were
also
prepared.
The
cultures
were
incubated
in
an
atmosphere
of
5%
C02/95%
air.
Hemoglobin-
containing
colonies
present
after
14
days
containing
>100
cells
were
scored
as
large
erythroid
colonies,
and
those
containing
40-100
cells
were
scored
as
small
erythroid
colonies.
Myeloid
colony-forming
unit
assay.
The
same
culture
system
was
used
except
that
0.3%
agar
(Difco)
replaced
0.9%
methylcellulose
and
that
EPO
was
omitted.
Aggregates
of
>40
cells
were
scored
as
colonies
after
14
days
incubation.
The
agar
discs
were
then
fixed
with
3%
(vol/vol)
glutaralde-
hyde
and
transferred
onto
individual
5
x
8
cm
glass
slides.
The
discs
were
dried
at
room
temperature
and
stained
with
luxol
fast
blue
and
a
combined
specific
and
nonspecific
esterase
stain.
Megakaryocyte
colony-forming
unit
assay.
The
same
cul-
ture
system
was
used
except
that
EPO
was
omitted
and
that
Abbreviations:
CSF,
colony-stimulating
factor;
EDF,
eosinophil
differentiation
factor;
EPO,
erythropoietin;
FCS,
fetal
calf
serum;
g,
gibbon;
G,
granulocyte;
h,
human;
HUVE,
human
umbilical
vein
endothelium;
IL-3,
interleukin
3;
m,
murine;
M,
macrophage;
PHA-
LCM,
phytohemagglutinin-stimulated
leukocyte
conditioned
medi-
um;
r,
recombinant.
2761
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
84
(1987)
Table
1.
Morphological
types
of
colonies
in
human
marrow
cultures
after
14
days
of
stimulation
in
agar
Total
Morphological
type,
Marrow
number
of
%
of
colonies
sample
Stimulus
colonies
G
GM
M
Eo
1
Medium
0
PHA-LCM
303
44
9 32
16
gIL-3
181
33
4
43
20
2
Medium
0
PHA-LCM
96
55
10
18
17
rhGM-CSF
103
59
7
28
5
gIL-3
91
66
1
12
11
Eo,
eosinophil.
15%
(vol/vol)
human
plasma
and
15%
(vol/vol)
FCS
were
substituted
for
30%
(vol/vol)
FCS.
After
a
14-day
incubation,
each
1-ml
culture
was
resuspended
in
Dulbecco's
phosphate-
buffered
saline
(PBS),
and
cytosmears
were
prepared
on
a
cytocentrifuge.
The
cytosmears
were
examined
for
the
pres-
enceofmegakaryocytesusingthealkalinephosphatase-mono-
clonal
anti-alkaline
phosphatase
technique
(9).
The
mono-
clonal
antibody
25E11,
which
recognizes
the
platelet
glyco-
protein
Ilb/IIIa
complex
(10)
served
as
the
primary
antibody.
Purification
of
Human
Neutrophils
and
Eosinophils.
Periph-
eral
blood
of
healthy
volunteers
was
centrifuged
on
a
hypertonic
gradient
of
metrizamide
(Nyegaard,
Oslo)
as
described
(11)
after
dextran
sedimentation.
The
purity
was
>95%
for
neutrophils
and
>92%
for
eosinophils.
The
cells
were
resuspended
in
Eagle's
minimal
essential
medium
supplemented
with
10%
(vol/vol)
FCS,
20
mM
Hepes
buffer
(pH
7.4),
and
antibiotics.
In
experiments
involving
adher-
ence,
blood
was
collected
using
0.2%
EDTA
as
anticoagu-
lant.
The
neutrophils
were
isolated
by
density-gradient
cen-
trifugation
in
Ficoll/Paque,
followed
by
sedimentation
in
dextran
and
hypotonic
lysis
of
erythrocytes.
The
preparation
contained
>97%
neutrophils
and
was
suspended
in
0.9%
NaCl
at
5
x
107
neutrophils
per
ml.
Antibody-Dependent
Cell-Mediated
Cytotoxicity
Assay.
51Cr-labeled,
trinitrophenyl-coupled
P815
cells
(4
x
103
cells
in
40
Al)
were
incubated
with
24
Al
of
rabbit
anti-
trinitrophenyl
(Miles-Yeda,
Rehovot,
Israel),
80
Al
of
puri-
EOSINOPHILS
801
-
40-
0
0
C.)
X.O
ni
i
GM-CSF
IL-3
_i
I
100
10
1
0
rh
GM-CSF(ng/ml)
10-2
10-3
lo0
0
IL-3
dilution
Table
2.
Enhancement
of
erythroid
colonies
by
gIL-3
Mean
number
of
erythroid
colonies
(large/small)
Stimulus
1
2
3
EPO
0/1
0/0
4/30
EPO/PHA-LCM
6/18
9/12
85/61
EPO/rhGM-CSF
2/14
0/5
4/53
EPO/gIL-3
0/4
1/7
20/65
The
mean
number
of
erythroid
colonies
from
three
experiments
is
reported.
A
small
colony
had
40-100
cells;
a
large
colony
had
>100
cells.
fled
human
neutrophils
or
eosinophils
(1.3
x
105
cells)
as
effector
cells,
and
16
,ul
of
rhGM-CSF,
gIL-3,
or
medium
for
2.5
hr
at
370C
in
V-bottom
microtiter
plates.
The
percent
cytotoxicity
was
calculated
as
described
(12).
Phagocytosis
Assay.
Eosinophils
were
suspended
at
2
x
107
cells
per
ml
in
PBS.
Dried
bakers'
yeast
(Tandaco,
Mel-
bourne,
Australia)
was
suspended
in
PBS
to
an
OD540
of
1.6.
A
250-1.l
aliquot
was
mixed
with
50
/.l
of
stimulator
(gIL-3,
rhGM-CSF,
or
medium
control),
with
100
Al
of
fresh
autol-
ogous
serum,
and
with
100
/A
of
cells
(13).
Tubes
were
incubated
for
1
hr
at
370C,
placed
on
ice,
and
then
centrifuged
at
4°C.
Cell
pellets
were
resuspended
in
50
Al
of
cold
PBS.
Smears
were
fixed
in
methanol
and
stained
with
Giemsa.
The
number
of
phagocytized
yeast
cells
were
counted
in
at
least
100
eosinophils
per
preparation.
Neutrophil
Adherence
Assay.
First-
or
second-passage
human
umbilical
vein
endothelial
(HUVE)
cells
were
isolated
and
maintained
as
described
(14).
HUVE
cells
were
plated
in
RPMI
1640
medium
with
20%
(vol/vol)
FCS
into
641-mm
diameter
wells
(Costar,
Cambridge,
MA)
at
2
x
104
cells
per
well
and
grown
to
confluence.
Prior
to
assay,
the
medium
was
removed,
and
the
HUVE
monolayer
was
washed
once
in
RPMI
1640
containing
10%
(vol/vol)
FCS.
To
each
well
was
added
a
total
volume
of
200
/ul
containing
neutrophils
and
the
activating
substance.
The
monolayers
were
incubated
at
37°C
for
30
min
in
an
atmosphere
of
5%
C02/95%
air,
and
then
the
nonadherent
cells
were
removed
by
aspiration,
and
each
well
was
washed
once.
All
medium
was
then
removed,
and
100
,ul
of
a
0.25%
solution
of
rose
bengal
in
PBS,
pH
7.3,
was
added
NEUTROPHILS
OJ
I
I
30
10
3
1
rh
GM-CSF(ng/mi)
10-2
3
x
10-2
10-3
10-4
glL-3
dilution
0
0
FIG.
1.
Antibody-dependent
cytotoxicity
of
trinitrophenyl-coupled
P815
cells
by
human
eosinophils
and
neutrophils
in
the
presence
of
various
concentrations
of
gIL-3
(o)
and
rhGM-CSF
(e).
The
anti-trinitrophenyl
IgG
dilution
was
1:300
for
eosinophils
and
1:3000
for
neutrophils.
No
cytotoxicity
was
observed
in
the
absence
of
antibody
with
or
without
gIL-3
or
rhGM-CSF.
2762
Cell
Biology:
Lopez
et
al.
U-

Proc.
Natl.
Acad.
Sci.
USA
84
(1987)
2763
Table
3.
gIL-3
stimulates
human
eosinophils
to
phagocytize
serum-opsonized
bakers'
yeast
Fresh
autologous
%
of
eosinophils
containing
phagocytized
yeast
cells
human
serum*
Stimulus
0
1
2
5%
(vol/vol)
None
94.7
±
0.9
4.3
±
0.7
1
±
0.6
gIL-3t
84.5
±
4.5t
10.3
±
1.2t
5.3
±
0.7t
rhGM-CSF
84
±
1.1t
12
±
0t
4
±
1t
2.5%
(vol/vol)
None
97
±
0.7
2
±
0.6
0.7
±
0.3
gIL-3
84
±
3.5t
12.7
±
2.3t
3.3
±
1.2t
rhGM-CSF
90
±
1.5t
7.7
±
1.9t
2.3
±
0.3§
The
percent
of
eosinophils
containing
zero,
one,
or
two
yeast
cells
is
shown
as
the
arithmetic
mean
±
SEM
of
triplicate
determinations.
*In
the
absence
of
serum
no
phagocytosis
was
observed
with
any
stimulus.
tgIL-3
was
used
at
a
final
dilution
of
1:300,
while
rhGM-CSF
was
used
at
30
ng/ml.
tDiffers
from
no
stimulus
samples
by
P
s
0.05.
§In
this
group
alone
there
were
several
eosinophils
with
three
particles.
to
each
well
for
10
min
at
room
temperature
(15).
After
aspiration
of
the
stain,
each
well
was
washed
twice
in
assay
medium
and
200
/,u
of
a
1:1
(vol/vol)
solution
of
ethanol/PBS
was
added.
When
a
uniform
release
of
stain
had
occurred,
30
min
later,
the
OD570
of
each
well
was
determined
using
an
ELISA
reader.
Adherence
is
proportional
to
the
difference
between
the
OD
of
each
well
minus
the
OD
of
wells
containing
HUVE
cells
only.
Superoxide
Production.
Purified
neutrophils
or
eosinophils
were
incubated
with
gIL-3,
rhGM-CSF,
or
medium
for
various
times
at
37°C.
Cells
(150
,ul)
were
then
added
to
a
mixture
of
100
,u1
of
freshly
prepared
cytochrome
c
(Sigma,
type
VI;
12.4
mg/ml),
100
,ul
of
fMet-Leu-Phe
(Sigma),
and
medium
to
1
ml.
The
mixtures
were
incubated
at
37°C
for
5
min,
after
which
the
cells
were
rapidly
cooled
and
pelleted
at
4°C.
The
supernatants
were
transferred
to
plastic,
disposable
cuvettes.
Superoxide
production
was
measured
in
duplicate
by
the
reduction
of
cytochrome
c
as
described
(16).
In
each
experiment
superoxide
dismutase
(Sigma)
inhibited
all
su-
peroxide
generation.
Chemotaxis.
The
chemotactic
response
of
neutrophils
in
a
gradient
of
fMet-Leu-Phe
was
tested
under
agarose
(17).
Agarose
(5
ml
of
0.5%
agarose;
type
II,
Sigma)
in
RPMI
1640
with
2%
(vol/vol)
FCS
and
20
mM
Hepes
was
poured
into
a
plastic
Petri
dish
(50
mm,
Kayline),
and
wells
2.4
mm
in
diameter
were
formed
2.4
mm
apart
in
a
horizontal
line
from
the
center
to
the
edge
of
the
plate.
Neutrophils
at
2.5
x
107
cells
per
ml
in
RPMI
and
2%
(vol/vol)
FCS
were
preincu-
bated
for
45-60
min
at
37°C
in
the
presence
or
absence
of
rhGM-CSF
at
100
ng/ml
or
gIL-3.
Ten
microliters
of
these
cells
was
added
to
the
center
well;
10
,ud
of
medium
was
added
to
the
inner
well;
and
10
,ul
of
fMet-Leu-Phe
was
added
to
the
outer
well.
The
Petri
dish
was
then
incubated
for
2
hr
at
370C
in
5%
C02/95%
air.
Then,
the
cells
were
fixed
in
methanol
at
40C
overnight,
followed
by
47%
(vol/vol)
formalin
for
30
min
at
25°C.
Migration
was
measured
under
x40
magnification
with
a
calibrated
graticule
(1
division
=
0.04
mm).
Chemo-
taxis
was
the
difference
between
directed
and
random
move-
ment.
RESULTS
Stimulation
of
Bone
Marrow
Cells.
gIL-3
stimulated
the
formation
of
myeloid
colonies
in
agar
after
14
days
of
incubation
(Table
1)
and
of
very
few
clusters
of
cells
after
7
days.
Morphological
examination
of
the
stained
agar
cultures
showed
that
gIL-3,
like
rhGM-CSF,
stimulated
granulocyte,
macrophage,
granulocyte-macrophage,
and
eosinophil
colo-
nies
(Table
1).
In
two
experiments
megakaryocytes
were
identified
in
cytosmear
preparations
of
gIL-3-
but
not
of
GM-CSF-stimulated
colonies.
gIL-3
also
stimulated
the
for-
mation
of
erythroid
colonies
in
the
presence
but
not
in
the
absence
of
pure
EPO
(Table
2).
Stimulation
of
Human
Granulocyte
Function.
Antibody-
dependent
cell-mediated
cytotoxicity
assay.
gIL-3
stimulated
eosinophils
to
kill
antibody-coated
tumor
target
cells
in
a
dose-dependent
manner
(Fig.
1).
The
degree
of
stimulation
was
similar
to
that
of
rhGM-CSF,
and
comparable
levels
of
killing
were
obtained
with
a
1:1000
dilution
of
gIL-3
and
rhGM-CSF
at
1
ng/ml.
In
contrast,
neutrophil-mediated
killing
was
not
significantly
enhanced
by
gIL-3
over
the
same
dose
range,
whereas
neutrophils
did
respond
to
stimulation
by
rhGM-CSF.
Phagocytosis.
gIL-3
stimulated
eosinophils
to
phagocytize
serum-opsonized
bakers'
yeast
(Table
3).
The
degree
of
40
n
20
._
Q
(0)
0
a)
Q
20.
E
c
ai
x
0
a
0.
0
rhGM-CSF
glL-3
10
3
1
0
rhGM-CSF,
ng/ml
,6L2
10-3
10-4
d
glL-3
dilution
FIG.
2.
rhGM-CSF
but
not
gIL-3
enhanced
fMet-Leu-Phe-stim-
ulated
superoxide
production
by
human
neutrophils.
Points
are
arithmetic
means
of
duplicate
determinations,
and
SDs
were
always
<15%
of
means.
Cell
Biology:
Lopez
et
al.

Proc.
Natl.
Acad.
Sci.
USA
84
(1987)
Table
5.
Effect
of
various
CSFs
on
human
myeloid
cells
Bone
marrow
Neutrophil
Eosinophil
CSF
colony
type(s)
function
function
hGM-CSF
G,
M,
Eo
+
+
m-
orhG-CSF
G
+
mEDF/Eo-CSF
Eo
-
+
gIL-3
G,
M,
Eo,
-
+
megakaryocytic
Eo,
eosinophil.
+,
Function
present;
-,
function
absent.
neutrophil
adherence,
while
phorbol
myristate
acetate
was
the
only
stimulus
tested
that
was
effective
on
eosinophils.
Chemotaxis.
The
ability
of
gIL-3
to
influence
neutrophil
movement
was
tested
in
a
migration
assay.
gIL-3
did
not
stimulate
or
inhibit
random
migration
of
neutrophils
or
their
unidirectional
movement
to
a
chemotactic
gradient
of
fMet-
Leu-Phe
(Table
4),
These
functions,
however,
could
be
inhibited
by
rhGM-CSF.
T
DISCUSSION
10-2
10i3
g
IL-3
none
dilution
FIG.
3.
Stimulation
of
neutrophil
(e)
and
eosinophil
(x)
adher-
ence
to
HUVE
cells
by
rhGM-CSF,
recombinant
human
tumor
necrosis
factor
type
a
(rhTNF-a),
and
gIL-3.
Each
point
is
the
arithmetic
mean
of
three
replicates
±
SEM.
Adherence
is
given
as
the
change
in
OD570
of
each
well.
phagocytosis
was
similar
to
that
obtained
with
rhGM-CSF,
at
the
two
concentrations
of
serum
used.
Superoxide
anion
production.
gIL-3
directly
stimulated
superoxide
production
by
eosinophils
(10.7
nmol
per
106
cells
per
15
min),
which
was
comparable
to
the
effect
of
rhGM-
CSF
(11
nmol
per
106
cells
per
15
min).
By
contrast
neither
cytokine
directly
stimulated
superoxide
production
from
neutrophils.
Preincubation
of
neutrophils
with
rhGM-CSF,
but
not
gIL-3,
strongly
enhanced
their
capacity
to
respond
to
fMet-Leu-Phe
(Fig.
2).
Adherence.
Because
some
cytokines
have
been
shown
to
influence
the
adherence
of
neutrophils
to
endothelial
cells,
gIL-3
was
tested
in
the
HUVE
cell
adherence
assay.
gIL-3
did
not
stimulate
either
eosinophils
or
neutrophils
to
adhere
to
HUVE
cells
(Fig.
3).
The
control
cytokines
tumor
necrosis
factor
type
a
and,
to
a
lesser
degree,
rhGM-CSF
stimulated
Table
4.
Effect
of
rhGM-CSF
and
gIL-3
on
neutrophil
migration
Migration,
mm
Stimulus
1
aM
0,1
1AM
0.01
ILM
None
Total
None
6.44
±
0.06
5.4
±
0.0
4.2
±
0.1
4.00
±
0
rhGM-CSF
5.86
±
0.08*
4.54
±
0.1*
4.14
±
0.1
gIL-3
6.5
±
0.1
5.2
±
0.1 4.1
±
0.1
Chemotactic
None
2.68
±
0.06
1.38
±
0.08
0.48
±
0
0.32
±
0
rhGM-CSF
2.18
±
0.03*
0.84
±
0.1*
0.18
±
0.06
gIL-3
2.5
±
0.17
1.52
±
0.06
0.24
±
0.0
Neutrophils
were
incubated
with
no
stimulus,
rhGM-CSF,
or
gIL-3,
and
fMet-Leu-Phe
at
1
iM,
0.1
,uM,
or
0.01
uM
was
added
as
chemoattractant.
As
a
control,
medium
without
fMet-Leu-Phe
was
also
used.
Total
migration
and
chemotactic
migration
[total
migration
-
(random
migration
plus
well
diameter)]
was
measured.
*Differs
from
control
sample
by
P
5
0.05.
Our
results
indicate
that
gIL-3
is
a
multipotential
proliferative
stimulus
for
human
cells
and
strongly
stimulates
some
func-
tions
of
mature
eosinophils
but
not
of
neutrophils.
gIL-3
stimulated
normal
human
nonadherent
bone
marrow
cells
to
produce
predominantly
day-14
colonies
similar
in
morphology
to
those
produced
by
GM-CSF
(Table
1),
thus
suggesting
that
gIL-3
acts
on
a
relatively
primitive
type
of
progenitor
cell.
In
the
presence
of
human
plasma,
gIL-3
also
stimulated
the
formation
of
megakaryocytes,
an
effect
com-
parable
to
that
of
mIL-3
on
mouse
bone
marrow.
On
the
other
hand,
its
capacity
to
stimulate
mast
cell
colonies
remains
to
be
elucidated.
In
the
presence
of
EPO,
gIL-3
stimulated
erythroid
colonies.
The
spectrum
of
activities
of
gIL-3,
therefore,
appears
to
be
similar
to
GM-CSF
except
for
the
megakaryocyte-stimulating
property.
However,
because
on-
ly
a
single
dose
of
each
cytokine
was
used
in
these
experi-
ments
and
because
gIL-3
was
not
purified,
no
conclusions
can
be
drawn
about
their
relative
potency.
gIL-3,
like
rhGM-CSF,
was
a
powerful
stimulator
of
mature
human
eosinophil
function
as
judged
by
enhancement
of
antibody-dependent
cell-mediated
cytotoxicity,
superox-
ide
production,
and
phagocytosis
(Fig.
1
and
Table
3).
In
contrast
to
the
rhGM-CSF,
however,
gIL-3
had
no
detectable
effect
on
any
neutrophil
function
we
studied.
In
this
regard
it
resembled
mEDF,
although
EDF
is
known
to
stimulate
only
eosinophil
colonies
from
human
bone
marrow
cells
(18)
whereas
gIL-3
clearly
stimulated
neutrophil
and
macrophage
maturation
as
well.
The
failure
of
gIL-3
to
stimulate
mature
neutrophil
function
was
surprising
in
view
of
its
ability
to
produce
day-14
granulocyte
colonies.
However,
since
no
day-7
granulocyte
colonies
were
produced
either,
it
is
possible
that
human
neutrophils lose
their
receptors
or
their
responsiveness
to
IL-3
as
they
mature.
In
the
mouse,
autoradiographic
studies
with
radiolabeled
mIL-3
revealed
a
decreasing
number
of
receptors
on
neutrophils
and
eosinophils
during
maturation;
however,
mature
neutrophils
and
eosinophils
were
clearly
labeled-eosinophils
bound
twice
as
many
mIL-3
molecules
as
did
neutrophils
(19).
An
alternative,
though
unlikely,
explanation
for
the
lack
of
human
neutrophil
stimulation
by
gIL-3
is
the
fact
that
gIL-3
differs
from
hIL-3
by
11
amino
acids
(6),
which
may
affect
the
interactions
of
the
gibbon
molecule
with
the
human
receptor.
It
is
clear
from
our
experiments
that
IL-3,
like
other
CSFs,
has
the
property
of
stimulating
granulocyte
function.
mIL-3
can
stimulate
murine
macrophages
in
vitro
(20)
and
in
vivo
(21),
although
no
effect
was
observed
on
mature
neutrophil
OA
-
0.3
0
U')
0.2.-
a
0
LUJ
u
z
LUJ
a0.1.
z
a
4c
0
H-
f
""v
100
10
rh
GM-CSF
(ng/ml)
103
102
rh
TNFa
(U/ml)
2764
Cell
Biology:
Lopez
et
al.

Proc.
Natl.
Acad.
Sci.
USA
84
(1987)
2765
Table
6.
Regions
of
homology
between
human
IL-3
and
mouse
T-cell
replacing
factor/B-cell
growth
factor
2
61
78
hIL-3
G E D
---Q
D
I
L
M
EN
N
L
R
R
P
N
L E
*
*
*
*
*
*
*
mTRF
G
E
I
F
G G
L
D
I
L
K
N
Q
T
V
R
GGTVE
63
83
125
138
hIL-3
E
F
R
R
K
L
T
F Y
L
K
T
L
-
E
*
*
*
*
*
*
*
mTRF
E
R
R
R
T
R
Q
F
-
L
D
Y
L
Q
E
106
119
Identical
amino
acids
are
indicated
by
asterisks
and
conserved
changes
with
dots.
mTRF,
T-cell
replacing
factor.
and
eosinophil
function
(21).
Thus
it
remains
to
be
seen
whether
there
are
fundamental
differences
between
mouse
and
primate
IL-3
in
their
capacity
to
bind
and
functionally
activate
mature
neutrophils
and
eosinophils.
The
biological
activities
of
gIL-3
are
compared
to
those
of
other
CSFs
known
to
stimulate
human
cells
in
Table
5.
Clearly
this
molecule
has
a
characteristic
range
of
actions.
It
has
broad
specificity
on
relatively
undifferentiated
bone
marrow
cells,
whereas
it
is
active
only
on
mature
eosinophils
and
not
mature
neutrophils.
It
is
also
apparent
that
three
cytokines
have
now
been
described
with
the
capacity
to
stimulate
mature
human
eosinophil
function:
rhGM-CSF
(13,
16),
mEDF
(18),
and
gIL-3.
We
found
some
homologies
between
g-
or
hIL-3
and
mouse
T-cell
replacing
factor/B-cell
growth
factor
2,
a
molecule
that
is
probably
identical
to
mEDF
(22).
A
computer
comparison
of
hIL-3
(or
gIL-3)
and
mouse
T-cell
replacing
factor
amino
acid
sequences
revealed
two
regions
that
were
47%
and
46%
homologous
(Table
6).
The
percentage
homology
was
calculated
over
regions
of
>15
amino
acids
and
includes
conserved
amino
acid
changes.
It
would
be
of
interest
to
determine
whether
these
regions
of
homology
are
related
to
eosinophil
stimulation.
An
eosinophil
stimulatory
molecule
is
produced
by
human
mononuclear
cells,
probably
monocytes,
after
culture
(23,
24).
It
has
been
proposed
that
this
molecule
is
similar
but
not
identical
to
GM-CSF,
or
CSF-a
(25).
Our
present
results
support
the
hypothesis
that
IL-3
is
at
least
one
of
the
factors
produced
under
these
culture
conditions
and
is
responsible
for
controlling
some
aspects
of
eosinophil
stimulation.
It
is
also
possible
that
GM-CSF
and
IL-3
may
regulate
different
aspects
of
eosinophil
function;
if
this
is
the
case,
eosinophilia
in
allergy
and
in
parasitism
may
be
stimulated
by
different
molecules
and
result
in
different
clinical
conditions
reflecting
the
different
roles
of
the
eosinophil.
We
thank
Maria
Ianella,
Carolyn
Lucas,
Judith
Russell,
Ann
Branford,
David
Haylock,
and
Rota
Koy
for
excellent
technical
assistance;
Drs.
G.
R.
Johnson
and
C.
G.
Begley
for
gift
of
reagents
and
useful
discussions;
and
Mari
Walker
for
typing
the
manuscript.
This
work
was
supported
by
grants
from
the
National
Health
and
Medical
Research
Council,
Anti-Cancer
Council
of
South
Australia.
1.
IhWe,
J.
N.,
Keller,
J.,
Oroszalan,
S.,
Henderson,
L.
E.,
Cope-
land,
T.
D.,
Fitch,
F.,
Prystowsky,
M.
B.,
Goldwasser,
E.,
Schrader,
J.
W.,
Palaszynski,
E.,
Dy,
M.
&
Lebel,
B.
(1983)
J.
Immunol.
131,
282-287.
2.
Fung,
M.
C.,
Hapel,
A.
J.,
Ymer,
S.,
Cohen,
D.
R.,
Johnson,
R.
M.,
Campbell,
H.
D.
&
Young,
I.
G.
(1984)
Nature
(Lon-
don)
307,
233-237.
3.
Cutler,
R.
L.,
Metcalf,
D.,
Nicola,
N.
A.
&
Johnson,
G.
R.
(1985)
J.
Biol.
Chem.
260,
6579-6587.
4.
Clark-Lewis,
I.,
Kent,
S.
B.
H.
&
Schrader,
J.
W.
(1984)
J.
Biol.
Chem.
259,
7488-7494.
5.
Schrader,
J.
W.
(1985)
Lancet
ii,
133-135.
6.
Yang,
Y.-C.,
Ciarletta,
A.
B.,
Temple,
P.
A.,
Chung,
M.
P.,
Kovacic,
S.,
Witek-Giannotti,
J.
S.,
Leary,
A.
C.,
Kriz,
R.,
Donahue,
R.
E.,
Wong,
G.
G.
&
Clark,
S.
C.
(1986)
Cell
47,
3-10.
7.
Wong,
G.
G.,
Witek,
J.
S.,
Temple,
P.
A.,
Wilkens,
K.
M.,
Leary,
A.
C.,
Luxenberg,
D.
P.,
Jones,
S.
S.,
Brown,
E.
L.,
Kay,
R.
M.,
Orr,
E.
C.,
Shoemaker,
C.,
Golde,
D.
W.,
Kauf-
man,
R.
J.,
Hewick,
R.
M.,
Wang,
E.
A.
&
Clark,
S.
C.
(1985)
Science
228,
810.
8.
Pennica,
D.,
Nedwin,
G.
E.,
Hayflick,
J.
S.,
Seeburg,
P.
H.,
Derynck,
R.,
Palladino,
M.
A.,
Kohr,
W.
J.,
Aggarwal,
B.
B.
&
Eoddel,
D.
V.
(1984)
Nature
(London)
312,
724-729.
9.
Cordell,
J.
L.,
Falini,
B.,
Erber,
W.
N.,
Ghosh,
A.
K.,
Abdulaziz,
Z.,
Macdonald,
S.,
Pulford,
K.
A.
F.,
Stein,
H.
&
Mason,
D.
Y.
(1984)
J.
Histochem.
Cytochem.
32,
219-229.
10.
Bums,
G.
F.,
Cosgrove,
L.,
Triglia,
T.,
Beall,
J.
A.,
Lopez,
A.
F.,
Werkmeister,
J.
A.,
Begley,
C.
G.,
Haddad,
A.
P.,
d'Apice,
A.
J.
F.,
Vadas,
M.
A.
&
Cawley,
J.
C.
(1986)
Cell
45,
269-280.
11.
Vadas,
M.
A.,
David,
J.
R.,
Butterworth,
A.,
Pisani,
N.
T.
&
Siongok,
T.
A.
(1979)
J.
Immunol.
122,
1228-1236.
12.
Vadas,
M.
A.,
Nicola,
N.
A.
&
Metcalf,
D.
(1983)
J.
Immunol.
130,
795-799.
13.
Metcalf,
D.,
Begley,
C.
G.,
Johnson,
G.
R.,
Nicola,
N.
A.,
Vadas,
M.
A.,
Lopez,
A.
F.,
Williamson,
D.
J.,
Wong,
G.
G.,
Clark,
S.
C.
&
Wang,
E.
A.
(1986)
Blood
67,
37-45.
14.
Gamble,
J.
R.,
Harlan,
J.
M.,
Klebanoff,
S.
J.
&
Vadas,
M.
A.
(1985)
Proc.
Natl.
Acad.
Sci.
USA
82,
8667-8671.
15.
Chong,
A.
S.
&
Parish,
C.
R.
(1985)
Cell.
Immunol.
92,
277-289.
16.
Lopez,
A.
F.,
Williamson,
D.
J.,
Gamble,
J.
R.,
Begley,
C.
G.,
Harlan,
J.
M.,
Klebanoff,
S.,
Waltersdorph,
A.,
Wong,
G.,
Clark,
S.
&
Vadas,
M.
A.
(1986)
J. Clin.
Invest.
78,
1220-1228.
17.
Nelson,
R.
D.,
Quil,
P.
G.
&
Simmons,
R.
L.
(1975)
J.
Immu-
nol.
115,
1650-1656.
18.
Lopez,
A.
F.,
Begley,
C.
G.,
Williamson,
D.
J.,
Warren,
D.
J.,
Vadas,
M.
A.
&
Sanderson,
C.
J.
(1986)
J.
Exp.
Med.
163,
1085-1099.
19.
Nicola,
N.
&
Metcalf,
D.
(1986)
J.
Cell
Physiol.
128,
180-188.
20.
Crapper,
R.
M.,
Vairo,
G.,
Hamilton,
J.
A.,
Clark-Lewis,
I.
&
Schrader,
J.
W.
(1985)
Blood
66,
859-865.
21.
Metcalf,
D.,
Begley,
C.
G.,
Johnson,
G.
R.,
Nicola,
N.
A.,
Lopez,
A.
F.
&
Williamson,
D.
J.
(1986)
Blood
68,
46-57.
22.
Sanderson,
C.
J.,
O'Garra,
A.,
Warren,
D.
J.
&
Klauss,
G.
G.
B.
(1986)
Proc.
Natl.
Acad.
Sci.
USA
83,
437-440.
23.
Veith,
M.
C.
&
Butterworth,
A.
E.
(1983)
J.
Exp.
Med.
157,
1828-1843.
24.
Vadas,
M.
A., Nicola,
N.,
Lopez,
A.
F.,
Metcalf,
D.,
John-
son,
G.
&
Pereira,
A.
(1984)
J.
Immunol.
133,
202-207.
25.
Thorne,
K.
J.
I.,
Richardson,
B.
A.,
Taverne,
J.,
Williamson,
D.
J.,
Vadas,
M.
A.
&
Butterworth,
A.
W.
(1986)
Eur.
J.
Immunol.
9,
1143-1150.
Cell
Biology:
Lopez
et
al.