Inhibition of the Erythropoietin

From by guest on January 26, 2015. For personal use only.
Inhibition of the Erythropoietin-Induced Erythroid Differentiation by
Granulocyte-Macrophage Colony-Stimulating Factor in the Human UT-7
Cell Line Is Not Due To a Negative Regulation of the
Erythropoietin Receptor
By 0. Hermine, A. Dubart, F. Porteux, P. Mayeux, M. Titeux, D. Dumenil, and W. Vainchenker
The human pluripotent UT-7 cell line is growth factor-dependent for proliferation and differentiation. We have previously
shown that (11 granulocyte-macrophage colony-stimulating
factor (GM-CSF) and erythropoietin (Epo) induce a myeloid
and erythroid pattern of differentiation, respectively; (2) GMCSF acts predominantly over Epo for cell differentiation; (3)
GM-CSF induces a rapid downmodulation (4 hours) of Epo
receptors (Epo-R) at the mRNA and binding site levels; and
(4) in contrast, Epo has no effect on GM-CSF receptor (GMCSF-R) expression. These results suggested that UT-7 cell
commitment or differentiation may be directed by a hierarchical action of growth factors through an early and rapid
transmodulation of growth factor receptors. To test this hypothesis, we introduced and expressed the murine Epo-R
(muEpo-R) in UT-7 cells using a retroviral strategy. Two retroviral vectors were constructed: one carrying the neomycin
resistance gene, and another carrying a mouse Epo-R cDNA
devoid of its regulatory untranslated 3’sequence placed under the transcriptional control of the viral long terminal repeat element (LTR) and the neomycin resistance gene. Three
UT-7/Epo-R infected clones (12.6, I O ) and one UT-7lneomycin clone (Neo) were selected in medium containing G418.
After growth factor deprivation (18 hours), Epo-Rs were expressed at the same level (approximately 6,000 receptors
per cell) in all four clones 12, 6, IO, Neo, and in parental
UT-7 cells, and exhibited similar affinity (0.1 t o 0.2 nmol/L).
Cross-linking experiments showed that Epo is associated
with three proteins of about 66, 85, and 100 kD in cells of
parental UT-7, as well as in cells of clones 10 and 12. An
inhibitory antibody directed specifically against the human
Epo-R (huEpo-R Ab) abolished almost completely the crosslinking on parental UT-7 cells, but not on cells of clone 12,
demonstrating that more than 90% cell surface Epo-Rs were
of murine origin. The presence of GM-CSF significantly reduced the number of Epo-Rs expressed on parental UT-7
cells, but not on cells of clones 12, IO, and 6. HuEpo-R Ab
inhibited Epo-induced parental UT-7 cell growth, but not
that of cells of clone 12, suggesting that the muEpo-R is able
t o induce human UT-7 cell proliferation. When cells of clone
12 were switched from a medium containing GM-CSFt o one
with Epo, cell surface glycophorin A (GPA) was induced, as
in parental UT-7 cells without inhibition by the huEpo-R Ab,
demonstrating that the muEpo-R is also able t o transduce
a differentiation signal in human cells. However, in cells of
clones 12, 6, IO, and Neo, as well as in parental UT-7 cells,
the induction of GPA by Epo was inhibited by GM-CSF. This
finding demonstrates that, although GM-CSF does not downregulate muEpo-R binding sites on UT-7/muEpo-R infected
clones, it still inhibits the effects of Epo on cell differentiation. Therefore, hierarchical regulation induced by growth
factors for cell commitment or differentiation more likely
acts downstream of cell surface receptors at either the signal
transduction or transcriptional levels.
0 1996 by The American Society of Hematology.
differentiation of the target cells bearing the appropriate
growth factor receptor. In contrast, their role as inducers of
differentiation and commitment of a multipotent cell is still
a matter of debate. On one hand, a stochastic model hypothesizes that pluripotent stem cells are intrinsically committed
to lineage restricted progenitors without influence of external
stimuli, growth factors simply permitting their survival, amplification, and further de~elopment.~.’
On the other hand, in
some instructive models, it is postulated that growth factors
interact with their receptors at the stem cell or multipotent
cell levels to direct the commitment or restriction toward the
different hematopoietic lineages.6 It has been reported that
growth factors may act competitively by an internal autocrine mechanism'^' or hierarchically by triggering their own
receptors on cell s~rface.’,’~
In the latter hypothesis, the fate
of the pluripotent cell would depend on the concentration of
the encountered growth factors and on the number and affinities of their receptors expressed on the cell surface. We
have recently shown that the commitment and differentiation
along the erythroid or myeloid pathways of the multipotent
UT-7 cell line are also directed hierarchically by erythropoietin (Epo) or granulocyte-macrophage colony-stimulating factor (GM-CSF), respectively.” In this cell model, GM-CSF
acts dominantly over Epo and rapidly downregulates the
number of Epo receptors (Epo-R) on the cell surface. This
downregulation is mediated by decreased mRNA levels. This
model suggests that receptor transmodulation could be involved in the UT-7 cell commitment or differentiation along
EMATOPOIESIS IS A continuous process leading to
the production of mature blood cells of various lineages from a population of pluripotent stem cells that are
capable of both extensive self-renewal and maturation. The
mechanisms that regulate the commitment of stem cells or
the lineage restriction of multipotential progenitors are
poorly understood.’ However, several speculative models
have been proposed based on studies of progenitor cell colony growth assays, in vitro differentiation of cell lines, and
in vivo
Most studies have focused on the role of growth factors.
It is well established that hematopoietic growth factors are
able to sustain proliferation and survival at all stages of
From INSERM U 362, lnstitut Gustave Roussy, Villejuif; Hdpital
Necker, Service d’Hdmatologie Clinique, Paris; and INSERM U 363,
ICGM, Hopital Cochin, Paris, France.
Submitted May I , 1995; accepted October 6, 1995.
Supported by grants from the Association de la recherche sur le
Cancer (ARC 6688) (Villejuij France) and from la Ligue Nationale
contre le Cancer (Paris, France).
Address reprint requests to 0. Hermine, MO, INSERM U 362,
Institut Gustave Roussy, 94805 Villejuif Cedex, France.
The publication costs of this article were defrayed in part by puge
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 1996 by The American Society of Hematology.
Blood, Vol 87, No 5 (March 1). 1996:pp 1746-1753
From by guest on January 26, 2015. For personal use only.
one restricted lineage. It could not, however, be ruled out
that other mechanisms were involved in these processes.
To define the role of Epo-R transmodulation by GM-CSF,
we have retrovirally transduced the muEpo-R cDNA into
UT-7 cells. The muEpo-R cDNA was placed under the transcriptional control of the retroviral LTR and its posttranscriptional regulatory sequences were removed to avoid a downregulation by GM-CSF. We show that muEpo-R is functional
in transducing both cell proliferation and differentiation signals in UT-7 cells, and although Epo-Rs in transfected cells
were not transmodulated, GM-CSF still acted as a dominant
factor over Epo, strongly suggesting that cell commitment
or differentiation is not a consequence of a receptor transmodulation.
pBT Zen mR-Epo SVNeo
5.3 kb
4.9 kb
2 kb
Cell lines and culture. UT-7 cells, kindly provided by N. Komatsu (Tichi Medical School. Tochigi-Ken. Japan).” were grown at
37°C in Iscove‘s modified Dulbecco‘s medium (IMDM) (Jbio. Paris.
France) containing 10% fetal calf serum (FCS) (Jbio) and the appropriate growth factors in a humidified atmosphere with 5% CO,.
Human recombinant Epo (Cilag, France) and GM-CSF (Amgen.
Thousand Oaks, CA) were used at 2 UlmL and 2.5 nglmL, respectively. The cell concentration was maintained between 2 and 5 X
IO’ cellslmL by diluting the cells with fresh medium every 3 to 4
Inhibiton?anti-Epo ond CM-CSFreceptor monoclonal rintihodies.
Monoclonal antibodies (MoAb) directed against the human Epo-R
(16.5.1) (huEpo-R) and the huGM-CSF-R were obtained from Genetics Institute (Cambridge, MA) and lmmunex (Seattle, WA). respectively. Both MoAbs are IgGl and inhibit the binding of Epo
and GM-CSF to their respective receptors.
Cell pro1;feration and d;fferentiationa.rsa\:v. Cells ( I X 1 04/mL)
were cultured in a 96-well microculture plate (Falcon no. 3072:
Becton Dickinson, Mountain View, CA) with Epo (2 UlmL) or GMCSF (2.5 nglmL) in the presence or absence of inhibitory antibodies
directed against the EPO-R or GM-CSF-R. Cell numbers and viability were then assessed by the trypan blue dye exclusion test. Cell
differentiation was assessed by analysis of GPA expression by flow
Flow cvtometp analysis. Cell surface antigens were detected by
indirect immunofluorescence staining with an anti-GPA MoAb
(CLB-ery 1, CLB, Rotterdam, Netherland). Cells were incubated for
30 minutes at 4°C with the diluted MoAb. After washing twice in
RPMl (Jbio). the cells were reincubated with fluorescent sheep Fab‘
antimouse Ig (Silenus, Hawthorn, Australia) for 30 minutes at 4°C.
After two washes in RPMI, cells were analyzed according to their
membrane antigenic density using a flow cytometer (FacSort, Becton
Dickinson). The negative control was determined on cells indirectly
labeled with an irrelevant lgGl antibody.
Constructions and retroviral infections q f UT-7cells. The retroviral vector encoding Epo-R was constructed by inserting 1.500
bp of the muEpo-R cDNA (Sal I fragment) in the Xho I site of the
pBTZen-SVNeo vector,” in which the neomycin resistance gene is
expressed under transcriptional control of the SV40 promoter. The
muEpo-R cDNA fragment starts 24 bases upstream of the ATG
codon and ends at the stop codon, and is driven by the viral long
terminal repeat (LTR). This construct (Fig I), or the Neo construct
(not shown), were transfected into the ecotropic packaging cell line
#2.“ Individual neomycin-resistant clones were derived, and their
supernatants tested for viral production on NIH 3T3 cells. The EpoR virus-producing clone, which produces the largest number of viruses, was finally selected. and its viral integrated structure and
23.1 kb
9.4 kb
6.6 kb
4.3 kb
2.3 kb
- W r
0.5 kb
- ”
Fig 1. (A) Retroviral construct. The Epo-R cDNA coding sequence
was cloned at the Xho I site of the pBTZen-SVNeo vector. Splice
donor (D) and acceptor (A) sites and ATG and TGA codons are indicated. The size of the different expected mRNAs are given. DNAs and
RNAs were prepared from three clones of Epo-R virus infected UT-7
cells (clones 6,10, and 12). IBI Southern blot analysis. A 10-pg portion
of genomic DNA of the different samples was digested with bglll and
electrophoresed on a 0.8% agarose gel. (C) Northern blot analysis. A
total of 5 p g RNA of the different samples were electrophoresed on
a formaldehyde-agarosegel. After transfer both membranes were
hybridized with a Neo probe.
expression were tested. The titers of the clones ($2 Neo and $2
Epo-R) used for infection were 5 x IO‘ infectious viral particles
per milliliter of supernatant. Supernatants of ecotropic $2 clones
containing the vectors were used to infect the amphotropic helper
cell line JJCRIP in the presence of 8 mglmL of polybrene (Aldrich.
Steinheim, Germany); 48 hours after infection, the $CRIP cells
supernatant was collected. UT-7 cells (2 x 10’ cells; 4 X IO‘lmL
final concentration) were then infected by incubation for 48 hours
in a mixture of 3 mL filtered (0.45 pmollL sterile filter (Nalgene.
Rochester, NY)) viral supernatant, 2 mL of a cell culture medium
containing GM-CSF (2.5 nglmL final concentration), and polybrene
(4 mglmL final concentration). UT-7 cells (10, 50. or 100 X 10‘1
mL) were then plated in a semisolid medium containing 0.8% methylcellulose in IMDM supplemented with 10% FCS, 1% bovine serum
albumin (BSA), 2.5 nglmL GM-CSF, and 0.8 mg/mL (3418. Individual clones were selected after 10 days, and expanded in liquid medium in presence of G418 (0.8 mglmL) and GM-CSF (2.5 nglmL).
DNA and RNA analvsis. Total cellular RNA was isolated according to the method of Chomczynsky and Sacchi.” RNA ( 5 to
10 pg per lane) was size-fractionated by formaldehydelagarose gel
electrophoresis and then transferred to a nylon membrane (Hybond
N; Amersham, Les Ulis, France). The membrane was hybridized
with a labeled Neo cDNA probe (Amersham Multiprime DNAlabeling system and a-”P-dCTP).
From by guest on January 26, 2015. For personal use only.
Genomic DNA was extracted using the standard method," digested with BgnI restriction enzyme using conditions recommended
by the supplier. After separation on 0.8% agarose gel, DNA fragments were transferred onto a nylon filter and hybridized to a 32Plabeled Neo cDNA probe (Amersham Multiprime DNA-labeling
system and w3*P-dCTP).
Binding experiments. Epo receptor characteristics were studied
as reported earlier." Briefly, Epo was iodinated with a specific activity ranging from 500 to 2,000 Ci/mmol using lodogen and 0.5 to 2
X 10' cells (parental UT-7 cells, and cells of clones 10, 6 , and 12)
were incubated for two hours at 25°C with various concentrations
of '2'I-Ep~in 100 pL Iscove's modified Dulbecco medium containing 5% FCS and 0.1% sodium azide. After this time, the cells
were centrifuged, and aliquots of the supernatants were used to
determine the free hormone concentrations. The cells were then
washed three times with ice cold PBS and the radioactivity bound
to the cells was measured. Nonspecific binding was determined by
incubating the cells with 100-fold excess of unlabeled ligand and
the specific binding was analyzed using an unweighted least-square
regression fitting method (Enzfitter, Biosoft, Paris, France). In some
experiments, the number of Epo binding sites was estimated from
triplicate measurements realized using a saturating (2 nmol/L) '"7Epo concentration.
Cross-linking experiments. Cell labeling with "'I-Epo and
cross-linking with Disuccinimidyl suberate (DSS) (Pierce Chemical
CO,Rockford, IL) were done as previously described." Briefly, the
cells were labeled with '*'I-Epo for 30 minutes at 37°C in the presence or absence of inhibitory antibodies. Sodium azide (0.1%) was
added to prevent internalization of the Epo-Epo-R complexes. The
cells were washed twice with ice cold PBS and cross-linked for 30
minutes on ice with 0.2 mmol/L DSS. After two washes in PBS
containing 0.1 molL ethanolamine pH 8.00, the cells were solubilized in 25 mmom HEPES pH 7.4 containing 150 mmoVL NaCI,
5 mmol/L EDTA, 1% Triton XIOO, 1 mmol/L phenylmethylsufonyl
fluoride, and 1 pg/mL each of aprotinin, leupeptin, and pepstatin
(all from Sigma, St Louis, MO). After centrifugation (27,OOOg 20
minutes), cellular extracts were immunoprecipitated with anti-Epo
antibodies and protein A Sepharose and immunoprecipitated products were analyzed by sodium dodecyl sulfate-polyacrylamide gel
electrophoresis (SDS-PAGE) and autoradiography.
Transfected murine Epo-R is expressed on the cell surface
of human UT-7 cells, and is not downregulated by GM-CSF.
The human UT-7 cell line is strictly growth factor-dependent
for survival, proliferation, and differentiation. We have previously shown that UT-7 cells express GM-CSF and EpoR, and that Epo-Rs are rapidly downregulated by GM-CSF
during the first 4 hours of exposure. This effect occurred
mainly at the transcriptional, and to a lesser extent, at the
posttranscriptional levels (Herrnine et al, in preparation). To
overcome this effect of GM-CSF on Epo-R expression, we
introduced the murine cDNA Epo-R with a retroviral vector
into UT-7 cells. This construct was devoid of the untranslated 3' regulatory elements" and was transcriptionally controlled by the viral LTR (Fig 1A).
Three cell lines were derived from muEpo-R vector-infected UT-7 cells (clones 12, 6 , and 10) and expanded in
liquid culture. To establish integration of the retrovirus in
the individual clones, Southern blots were performed using
BglII (Fig 1B). This restriction enzyme cuts twice within the
proviral genome and is useful for comparisons of proviral
Table 1. Binding of '251-Epoto UT-7 Cells and Cells of Clones 12, 10,
and 6 in the Presence or Absence of GM-CSF
17,680 2 190
18,700 2 520
16,370 2 440
5,140 2 145
18,600 2 390
18,200 i- 150
Cells were cultured in the presence of GM-CSF (2.5 ng/mL) or deprivated for
18 hours, and then were incubated with 2 nmol/L of '251-Epofor 2 hours at 25°C
with 0.1% sodium azide. The data are presented in cpmlcell as the average of
triplicate measurements ? SD. The results represent the average of three experiments.
integration between different lines. Two bands were detected
for each clone, a common one (2.1 kb) corresponding to the
internal proviral fragment, and a second one corresponding
to the 3' proviral junction fragment, which differed in the
three different clones, demonstrating that they derived from
independent infection events. A Neo vector infected clone
(Neo) andor the parental cell line UT-7 were used as a
control throughout this study.
To check transcription of the proviral genome and expression of the muEpo-R, Northern blot analysis and Epo binding
assays were performed. Using a Neo probe, Northem analysis identified three transcripts of the expected size, ie, 3.6,
3.2, and 2 kb for Neo infected cells (data not shown) and
5.3, 4.9, and 2 kb for Epo-R infected cells (Fig IC). In
contrast, no signal was observed in the parental UT-7 cell
line. Using binding experiments, we showed that surprisingly, and despite high expression of muEpo-R mRNA (hybridized with a Neo probe on Fig IC), muEpo-R transfected
cells (clones 12, 6 , IO) had the same number of Epo-R on
their cell surfaces as the UT-7 parental cell line and the Neo
transfected cells (Neo) (Table 1 ). Furthermore, Scatchard
analysis performed on clones 12 and 10 confirmed that they
expressed a comparable number of Epo-R per cell with kD
values close to parental UT-7 cells (Fig 2). However, because
human recombinant '"I-Epo binds mu and hu Epo-R with
roughly the same affinity," we could not determine from
these results whether murine Epo-Rs were expressed on the
surface of the cells from clones 6 , 10, 12. To answer this
question, we performed cross-linking experiments with Epo
on parental UT-7 cells and cells from clones 12 and 10
in the presence or absence of an inhibitory anti-huEpo-R
antibody. This antibody specifically inhibits Epo binding to
the huEpo-R, but not to the muEpo-R (data not shown). As
reported earlier, '251-Ep0cross-linking showed the presence,
in UT-7 cells, of three proteins with molecular masses of 66
kD (corresponding to the Epo-R binding chain), 85 M), and
100 kD (corresponding to the putative Epo-R accessory proteins).'"'" In the parental UT-7 cells, addition of the inhibitory antihuman Epo-R MoAb inhibited the cross-linking of
In contrast, although clones 12 and 10 exhibited
the same pattern of Epo cross-linked proteins as parental
UT-7 cells, the inhibitory anti-huEpo-R MoAb had little or
no effect (Fig 3), demonstrating that virtually all (>90%)
cell surface Epo-Rs were of murine origin.
In agreement with this finding, GM-CSF (2.5 ng/mL) had
no effect on the number of Epo-R or on their affinity with
Epo as assessed by '251-Ep0binding assay (Table 1) and
From by guest on January 26, 2015. For personal use only.
0 -GM-CSF 5000 R
0 +GM-CSF 1990 R
CllO Cl12
-+ - +-
B: pM
Clonel 0
0 -GM-CSF 6900 R
~ d = 4 7 pM
0 +GM-CSF 6300 R
6: pM
Clonel 2
-GM-CSF 7200 R
+GM-CSF 6900 R
B: pM
Fig 3. Murine Epo receptors are expressed on cell surface of
clones 12 and 10. '251-Epoprotein was cross-linked t o parental UT-7
cells (PI, and cells of clones 10 (CI 10) and 12 (CI 12). Cells were labeled
with 500 pmol/L of lZ51-Epoin the absence (-1 or in the presence (+I
of antihuman Epo-R MoAb. Epo and Epo-R were cross-linked with
DSS and complexes were analyzed as described in Materials and
Methods. Arrows on right side indicate the molecular weights of
proteins cross-linked with Epo.
Scatchard analysis (Fig 2). In contrast, and as reported earlier," GM-CSF decreased significantly the Epo-R number
of the UT-7 parental cell line from 5,000 to 2.000 (receptors/
cell) with no affinity change (kD = 390 pmol/L v 310 pmol/
L) (Fig 2).
Taken together, these results demonstrate that the transduced cell lines express a high number of muEpo-R on their
cell surface. These receptors are associated with the same
putative accessory proteins as the huEpo-R but are not downregulated by GM-CSF.
Tronsdiicecl mirEpo-R is able to induce proliferotion of
humori UT-7 cells in the presence of Epo. We then examined whether the muEpo-R is capable of transducing a proliferative signal when triggered by Epo. The growth response
Fig 2. Number and affinity of Epo receptors on parental UT-7 cells
and cells of clones 10 and 12. Scatchard analysis of Epo binding sites
on UT-7 cells, clones 10 and 12 in the presence ( 0 )or absence (0)
GM-CSF (after 18 hours of growth factor deprivation). The number
of Epo receptors and their affinities are indicated for each condition.
Cells were grown in presence of 2.5 n g h L GM-CSF, and then washed
twice and resuspended in presence or not of GM-CSF (2.5 ng/mL)
for 18 hours. '251-Epo binding was then determined by Scatchard
analysis using 0.5 x 10'cells/100 FL per point as described in Materials and Methods. B (bound) or F (Free) '251-Epo.
From by guest on January 26, 2015. For personal use only.
Table 2. Growth of Parental UT-7 Cells and Transduced UT-7 Cells
in the Presence of Epo or GM-CSF
Table 3. Transduced muEpo-R is Able to Induce Proliferationin the
Human UT-7 Cells
6,300 t 200
6,100 t 200
6,600 f 400
4,200 f 200
4,700 f 400
80 f 20
3,200 2 80
2.600 2 80
3,400 2 100
2,900 2 80
2,580 2 60
2,600 _i 80
1,800 80
1.500 _t 60
3,100 t 120
2,700 _t 80
UT-7 cells and cells from clones 12, 10,6, and Neo were grown in the presence
of either Epo (2 U/mL) or GM-CSF (2.5 ng/mL) in 96-well microplates (lo3cells/
100 pLI. The number of viable cells per 100 pL was determined by a trypan
blue exclusion on day 4. The data represent the mean t SD of a representative
experiment made in quadruplicate.
was assessed by counting the viable cells after 96 hours of
culture (Table 2). As reported earlier, parental UT-7 cells
responded to GM-CSF (2.5 n g k ) and Epo ( 2 U k ) . Clones
12 and 10 grew in the presence of GM-CSF and Epo, as
well as the parental cell line, suggesting that muEpo-R is
able to transduce a growth signal in human cell lines (Table
2). Clone 6 grew slower than the others in the presence of
Epo. However, the same growth alteration was observed in
the presence of GM-CSF. Moreover, transfected clones
could be maintained in the presence of Epo or GM-CSF for
-............ ,,e
Antibody dilutions
12: iw
Antibody Dilutions
Fig 4. Inhibitory concentration of antihuman Epo-R and antihuman GM-CSF-R antibodies. Cells ([email protected]/lOOpLJ were cultured for 4
days in a 96-well mimcutture plate with Epo (2 UlmLJ or GM-CSF
(2.5 ng/mLJ in the presence of various dilutions of inhibitory antihuman Epo-R or antihuman GM-CSF-R MoAbs. Cell numbers per 100
p L and v i a b i l i were then assessed by the trypan blue dye exclusion
Epo-R moAb
Cells from clone 12 (103/100pLJwere cultured with Epo or GM-CSF
i n the presence or absence of inhibitory concentration of anti-huEpoR or anti-huGM-CSF-R MoAbs. The number of viable cells per 100
fiL was determined by a trypan blue exclusion on day 4. The results
represent the data of one experiment made in quadruplicate.
a long period of time, and deprivation of Epo from the culture
medium resulted in a rapid cell death in 2 or 3 days as for
the parental UT-7 (data not shown). Although the number
of residual human Epo-R on cell surface of clones 12, 6,
and 10 was extremely low, we could not rule out that they
were responsible for the entire growth signal. Indeed, I O 0
to 300 Epo-R per cell are sufficient to induce a biological
effect in normal erythroid progenitors or in some Epo-responsive cell lines.*'-'' Therefore, we repeated the same experiments in the presence of the inhibiting antibody directed
against the huEpo-R (16.1.5). We first determined the inhibitory concentration of antibodies directed against huEpo-R
and huGM-CSF-R used as a control, which inhibit parental
UT-7 cell growth in the presence of Epo ( 2 U/mL) or GMCSF (2.5 ng/mL), respectively (Fig 4). We next grew clone
12 in the presence of Epo (2 U/mL) or GM-CSF (2.5 ng/
mL) with an inhibitory concentration of antibody directed
against huEpo-R (dilution was 1/50 vol/vol) or huGM-CSFR (dilution was 1/2,000 vol/vol). The antibody directed
against the hu Epo-R did not inhibit the growth of clone 12
in presence of either Epo or GM-CSF, whereas the antibody
directed against the huGM-CSF-R did so in the presence of
GM-CSF, but not of Epo (Table 3). Similar results were
found in the two other clones. These results demonstrate that
muEpo-R can transduce a proliferative signal in the human
UT-7 cell line.
MuEpo-R is able to induce erythroid differentiation of
the human UT-7 cells in the presence of Epo. We have
previously shown that UT-7 is a multipotent cell line." Epo
or GM-CSF can induce erythroid or myeloid programs of
differentiation, which are well assessed by GPA or CD33
expression, respectively. In the presence of GM-CSF, the
level of GPA expression is low and increases on exposure
to Epo, increasing to a maximum in about 10 days of culture.
Therefore, to assess the ability of muEpo-R to induce cell
differentiation, cells from clone 12 grown in a GM-CSFcontaining medium were switched to an Epo-containing medium with the presence of an inhibitory concentration of
antibody directed against huEpo-R (16.1.5), or an antihuGM-CSF-R (as control) for 10 days. After this period,
GPA expression was determined by flow cytometry analysis.
As shown in Fig 5, Epo markedly increased GPA expression
in clone 12, while the inhibitory antibody directed against
huEpo-R did not prevent this induction, demonstrating that
the muEpo-R is also able to induce an erythroid program of
From by guest on January 26, 2015. For personal use only.
Epo+ anti-hGM-CSF-R
;I i
Clone 10
. .
i j;
GM-CSF + Epo
Fig 5. Differentiation of muEpo-R transduced cells in the presence
of inhibitory concentration of anti-huEpo-R and anti-huGM-CSF-R
antibodies. Cells from clone 12 were grown in presence of GM-CSF
12.5 ng/mL) for at least five passages, washed, and thereafter reincubated in the presence of either GM-CSF (2.5 nglmLl or Epo (2 UlmL)
and with inhibitory concentration of anti-huEpo-R or anti-huGMCSF-R for 10 days. Cells were then analyzed for GPA expression as
described in Materials and Methods. An lgGl irrelevant antibody was
used as the negative control (IgG1).
differentiation in a human cell line. Similar results were
found in the two other clones.
Although GM-CSF does not downregulate Epo-R number
on the cell sugace of transfected clones, it still acts dominantly over Epo to inhibit erythroid differentiation. We
have previously shown that the GM-CSF effect on UT-7
differentiation predominates over that of Epo. In the presence
of GM-CSF plus Epo, the expression of GPA was identical
to that obtained with GM-CSF alone." To assess whether
GM-CSF still inhibited Epo-induced cell differentiation in
muEpo-R expressing cells (clones 12, 6, lo), and in Neotransfected cells (Neo), we grew them in the presence of
Epo, washed and then switched them to an Epo, GM-CSF,
or GM-CSF plus Epo-containing medium for 10 days. GPA
expression was then assessed by flow cytometry analysis. In
clones 12, 10 (Fig 6), and 6 (data not shown), GPA expression was downmodulated by GM-CSF and by the combination of Epo plus GM-CSF with the same magnitude, as observed with the cells of Neo clone (Fig 6) and UT-7 parental
cells (data not shown). Using the same conditions of culture,
Clone Ne0
Fig 6. Expression of GPA on transduced UT-7 cells in the presence
of GM-CSF, Epo, and a combination of Epo plus GM-CSF. Cells from
clones 12,10, and Neo were grown in the presence of Epo (2 UlmL)
for at least five passages, washed, and thereafter reincubated in the
presence of Epo (2 UlmLl. GM-CSF (2.5 nglmL), or a combination of
Epo (2 UlmLl plus GM-CSF (2.5 nglmL1 for 10 days. Cells were then
analyzed for GPA expression as described in Materials and Methods.
lgGl irrelevant antibody was used as the negative control (IgG1).
. ....
From by guest on January 26, 2015. For personal use only.
Table 4. Binding of '251-Epoto UT-7 Cells and Cells of Clones 12 in
the Presence or Absence of GM-CSF at Day 10
+GM-CSF 2,270 ? 115 2,840 i- 270 4,975 t 540 7,960 t- 720 6,090
-GM-CSF 6,935 ? 30 6,030 i 120 5,170 ? 50 6,935 2 615 6,150
+ Epo
Cells were cultured in the presence of Epo (2 U/mL). washed, and then switched
to an Epo (2 UlmL), GM-CSF (2.5 ng/mLI, or GM-CSF (2.5 ng/mL) + Epo ( 2 UlmL)
medium for 10 days. For binding analysis cells were washed and resuspended
in the presence or not of GM-CSF for 18 hours and were then incubated with 2
nmollL of '''I-Epo for 2 hours at 25°C with 0.1% sodium azide. The data are
presented in number of receptorslcell as the average of triplicate measurements
t SD. The results represent the average of three experiments.
the number of Epo-R was also examined at day IO. As
shown in Table 4, on clone 12 in the presence of GM-CSF,
downregulation of Epo-R did not occur. In contrast, in UT7 parental cells the Epo-R number was still downmodulated
by GM-CSF with the same magnitude as at day 1 (Fig 2).
Taken together these findings demonstrate that although
GM-CSF does not downregulate Epo-R, it still acts dominantly over Epo to inhibit erythroid differentiation.
Understanding of the mechanisms involved in stem cell
commitment has been hampered by the difficulties of purifying, identifying, and maintaining pluripotent cells in sufficient number. An alternative is provided by the study of
permanent pluripotent hematopoietic cell lines dependent on
the presence of growth factors for survival, proliferation,
commitment, and differentiation. Studies using cell lines,
and to a lesser extent, normal bone marrow cells, have established a hierarchical modulation of growth factor effects for
cell differentiation. For example, interleukin (1L)-3 or GMCSF may inhibit macrophagic," erythr~id,'~
or granulo~ytic'~
differentiation induced by M-CSF, Epo, and G-CSF, respectively. On one hand, it has been suggested, but not demonstrated, that transmodulation of receptors may be one mechanism involved in this hierarchy.'."' On the other hand, in
some murine growth factor receptor transduced cell lines,
this phenomenon may occur even without negative regulation of receptor^.^^.^^
We have previously shown that, although of leukemic
origin, the human UT-7 cell line provides a good model to
study lineage restriction hierarchically induced by growth
factors." Indeed, UT-7 cells may display features of megakaryocytic, eosinophilic, basophilic, and erythroid lineages.
In this cell line, the presence of Epo favors differentiation
toward the erythroid lineage (demonstrated by the presence
of erythroblasts that are hemoglobinized and express high
levels of glycophorin A), whereas GM-CSF, as well as IL3, even in the presence of Epo, inhibits this Epo-induced
erythroid differentiation. This effect was associated with a
rapid downmodulation of Epo-R within the first 4 hours of
GM-CSF exposure. In this report, we wondered whether this
GM-CSF- induced Epo-R downmodulation was responsible
for the inhibition of erythroid differentiation. Thus, to address this question, we have transduced and highly expressed
a muEpo-R species that cannot be downregulated in the UT-
7 cells to study the effects of a combination of Epo plus
As expected from the high homology between human and
murine Epo-Rs," we first demonstrated that muEpo-R is able
to transduce both proliferative and differentiative signals in
a human cell line. Furthermore, in UT-7, the murine binding
Epo-R chain was able to associate with two human 85 and
105 kD proteins, thus suggesting a strong homology between
these putative accessory chains in different species. Although clones 6, 10, and 12 expressed variable amounts
of the endogenous huEpo-R mRNA, the number of Epo-R
expressed were constant and roughly identical to that o f
parental cells. Moreover, virtually all cell surface receptors
were of murine origin. This finding strongly suggests that
the Epo-R number expressed at the cell surface level may
be limited by the requirement of an accessory chain for its
membrane transportation. Further work is needed to understand the mechanisms operating at the molecular level to
translocate Epo-R to the cell surface in these infected cells.
Although muEpo-R was not downregulated by GM-CSF,
GM-CSF still acted dominantly over Epo to inhibit erythroid
differentiation. This result shows that the dominant effect
of GM-CSF over Epo is not a consequence of an Epo-R
downmodulation. Alternatively, because Epo-R expression
can also be considered as an early erythroid differentiation
marker, GM-CSF inhibition of Epo-R expression may reflect
a rapid evidence of the GM-CSF driven inhibition of the
erythroid differentiation program. Thus, Epo-R inhibition is
more likely a consequence and not the cause of the GMCSF-induced inhibition of the erythroid differentiation.
Taken together these findings strongly suggest that the
hierarchy between growth factors for cell commitment or
differentiation is probably not directed by transmodulation
of receptors. The levels at which GM-CSF can inhibit the
differentiation signal transduced by the Epo-R are numerous
and are under investigation. Activated Epo-R and GM-CSFR may compete for various intracellular proteins that transduce differentiation signals to the nucleu~.'~
However, both
receptors used the same known transduction signalization
such as the Ras and the JAK2/STATS pathways." Therefore,
inhibition of erythroid differentiation by GM-CSF may be
related to another mechanism. At the nucleus level, differential activation by Epo or GM-CSF of DNA binding proteins,
or proto-oncogenes expression such as csuch as GATA
myc, or c-myb)"," may also contribute to the inhibition of
erythroid commitment or differentiation. Alternatively. GMCSF may reduce the length of the G I phase of the cell cycle
and thus inhibits erythroid differentiation as suggested by
Krosl et
in Epo-R-transfected B d F 3 cell line treated
with IL-3.
In conclusion, our data suggest that the dominant effect
of GM-CSF over Epo is not a consequence of a receptor
transmodulation, but more likely occurs at the signal transduction or transcriptional level. Therefore, the modulation
of receptors would be considered more as an early marker
of differentiation.'2,33However, this conclusion is based on
the use of a leukemic cell line, and it would be important,
but rather difficult, to reproduce such results on normal
multipotent progenitors.
From by guest on January 26, 2015. For personal use only.
The authors are indebted to J.-L. Villeval and I. Dusanter for
helpful discussion and critical review of the manuscript, to Genetics
Institute (Cambridge, MA), to Immunex (Seattle, WA), and Amgen
(Thousand Oaks, CA) for providing the antihuman Epo-R MoAb,
the anti-GM-CSF-R MoAb, and the human rhGM-CSF, respectively.
I . Metcalf D: The molecular control of cell division, differentiation commitment and maturation in haemopoietic cells. Nature
339:27, 1989
2. Ogawa M, Porter PN, Nakahata T Renewal and commitment
to differentiation of hemopoietic stem cells (an interpretative review). Blood 61:823, 1983
3. Ogawa M: Differentiation and proliferation of hematopoietic
stem cells. Blood 8 1:2844, 1993
4. Nakahata T, Cross AJ, Ogawa M: A stochastic model of selfrenewal and commitment to differentiation of the primitive hemopoietic stem cells in culture. J Cell Physiol 113:455, 1982
5. Fairbaim LJ, Cowling GJ, Reipert BM, Dexter TM: Suppression of apoptosis allows differentiation and development of a
multipotent hemopoietic cell line in the absence of added growth
factors. Cell 74:823, 1993
6. Van Zant G, Goldwasser E: Competition between erythropoietin and colony-stimulating factor for target cells in mouse marrow.
Blood 53:946. 1979
7. Hermine 0, Beru N, Pech N, Goldwasser E: An autocrine role
of erythropoietin in mouse hematopoietic cell differentiation. Blood
78:2253, 1991
8. Pech N, Hermine 0, Goldwasser E: Further study of internal
autocrine regulation of multipotent hematopoietic cells. Blood
82: 1502, 1993
9. Gliniak BC, Rohrschneider LR: Expression of the M-CSF receptor is controlled posttranscriptionally by the dominant actions of
GM-CSF or Multi-CSF. Cell 63:1073, 1990
IO. Walker F, Nicola NA, Metcalf D, Burgess A: Hierarchical
down-modulation of hemopoietic growth factor receptors. Cell
43:269, 1985
1 I . Hermine 0, Mayeux P, Titeux M, Mitjavila MT, Casadevall
N, Guichard J, Komatsu N, Suda T, Miura Y, Vainchenker W,
Breton-Gorius J: Granulocyte-macrophage colony-stimulating factor
and erythropoietin act competitively to induce two different programs of differentiation in the human pluripotent cell line UT-7.
Blood 80:3060, 1992
12. Komatsu N, Nakauchi H, Miwa A, Ishihara T, Eguchi M,
Moroi M, Okada M, Sat0 Y, Wada H, Yamata Y, Suda T, Miura
Y: Establishment and characterization of a human leukemic cell line
with megakaryocytic features: Dependency on granulocyte-macrophage colony stimulating factor, interleukin 3, or erythropoietin for
growth and survival. Cancer Res 51:341, 1991
13. Johnson GR, Gonda TJ, Metcalf D, Hariharan IK, Cory S: A
lethal myeloproliferative syndrome in mice transplanted with bone
marrow cells infected with a retrovirus expressing granulocyte-macrophage colony stimulating factor. EMBO J 8:441, 1989
14. Mann R, Mulligan RC, Baltimore D: Construction of a retrovirus packaging mutant and its use to produce helper-free defective
retrovirus. Cell 33: 153, 1983
15. Chomczynski P, Sacchi N: Single-step method of RNA isolation by acid guanidium thiocyanate-phenol-chloroform extraction.
Anal Biochem 162:156, 1987
16. Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning. A
Laboratory Manual. Cold Spring Harbor, NY, Cold Spring Harbor
Laboratory, 1989
17. Dusanter-Fourt I, Casadevall N, Lacombe C, Muller 0, Billat
C, Fischer S, Mayeux P: Erythropoietin induces the tyrosine phosphorylation of its own receptor in human erythropoietin-responsive
cells. J Biol Chem 267:10670, 1992
18. Dubart A, Feger F, Lacout C, Goncalves F, Vainchenker W,
Dumenil D: Murine pluripotent hematopoietic progenitors constitutively expressing a normal erythropoietin receptor proliferate in response to erythropoietin without preferential erythroid cell differentiation. Mol Cell Biol 14:4834, 1994
19. Youssoufian H, Longmore G, Neumann D, Yoshimura A,
Lodish H F Structure, function, and activation of the erythropoietin
receptor. Blood 81:2223, 1993
20. Mayeux P, Lacombe C, Casadevall N, Chretien S, Dusanter
I, Gisselbrecht S: Structure of the murine erythropoietin receptor
complex. Characterization of the erythropoietin cross-linked proteins. J Biol Chem 266:23380, 1991
21. Mayeux P, Billat C, Jacquot R: The erythropoietin receptor
of rat erythroid progenitor cells. J Biol Chem 262:13985, 1987
22. Takahashi T, Chiba S, Hirano N, Yazaki Y, Hirai H: Characterization of three erythropoietin (Epo)-binding proteins in various
human Epo-responsive cell lines and in cell transfected with human
Epo-receptor cDNA. Blood 85:106, 1995
23. Broudy VC, Lin N, Brice M, Nakamoto B, Papayannopoulou
T: Erythropoietin receptor characteristics on primary human erythroid cells. Blood 77:2583, 1991
24. Krosl J, Damen JE, Krystal G, Humphries RK: Erythropoietin
and interleukin-3 induce distinct events in erythropoietin receptorexpressing BAF3 cells. Blood 8550, 1995
25. Fukunaga R, Ishisaka-Ikeda E, Nagata S: Growth and differentiation signals mediated by different regions in the cytoplasmic
domain of granulocyte colony-stimulating factor receptor. Cell
74:1079, 1993
26. Jones SS, D’Andrka AD, Haines LL, Wong GG: Human
erythropoietin receptor: Cloning, expression, and biologic characterization. Blood 76:31, 1990
27. Williamson EA, PhD., Boswell HS, MD.: Signal transduction
during myeloid cell differentiation. Curr Opin Hematol 2:29, 1995
28. Gouilleux F, Pallard C, Dusanter-Fourt I, Wakao H, Haldosen
LA, Norstedt G, Levy D, Groner B: Prolactin, growth hormone,
erythropoietin and granulocyte-macrophage colony stimulating factor induce MGF-Stat5 DNA binding activity. EMBO J 14:2005,
29. Komatsu N, Yamamoto M, Fujita H, Miwa A, Hatake K,
Endo T, Okano H, Katsube T, Fukumaki Y, Sassa S, Miura Y:
Establishment and characterization of an erythropoietin-dependent
subline, UT-7lEp0, derived from human leukemia cell line, UT-7.
Blood 82:456, 1993
30. Clarcke M, Kukowska-Latallo J, Westin E, Smith M, Prochownick E: Constitutive expression of a c-myb cDNA blocks Friend
murine erythroleukemia cell differentiation. Mol Cell Biol 8:884,
3 1. Arsura M, Luchetti MM, Erba E, Golay J, Rambaldi A, Introna M: Dissociation between p93B-myb and p75c-myb expression
during the proliferation and differentiation of human myeloid cell
lines. Blood 83:1778, 1994
32. Just U, Friel J, Heberlein C, Tamura T, Baccarini M, Tessmer
U, Klingler K, Ostertag W: Upregulation of lineage specific receptors
and ligands in multipotential progenitor cells is part of an endogenous program of differentiation. Growth Factors 9:291, 1993
33. McClanahan T, Dalrymple S, Barkett M, Lee F: Hematopoietic growth factor receptor genes as markers of lineage commitment
during in vitro development of hematopoietic cells. Blood 81:2903,
From by guest on January 26, 2015. For personal use only.
1996 87: 1746-1753
Inhibition of the erythropoietin-induced erythroid differentiation by
granulocyte-macrophage colony-stimulating factor in the human UT-7
cell line is not due to a negative regulation of the erythropoietin
O Hermine, A Dubart, F Porteux, P Mayeux, M Titeux, D Dumenil and W Vainchenker
Updated information and services can be found at:
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
Information about ordering reprints may be found online at:
Information about subscriptions and ASH membership may be found online at:
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.