8/7/2019 Paper PU.1

http://slidepdf.com/reader/full/paper-pu1 1/10

Available online at www.sciencedirect.com

Leukemia Research 32 (2008) 121–130

PU.1 is dispensable to block erythroid differentiationin Friend erythroleukemia cells

Marıa Jose Fernandez-Nestosa, Pablo Hernandez,Jorge B. Schvartzman, Dora B. Krimer ∗

Department of Cell and Developmental Biology, Centro de Investigaciones Biol´ ogicas,

Consejo Superior de Investigaciones Cientıficas (CSIC), 28040-Madrid, Spain

Received 29 January 2007; received in revised form 29 January 2007; accepted 7 May 2007

Available online 21 June 2007

Abstract

Friend murine erythroleukemia cell lines derive from erythroblasts transformed with the Friend complex where the spleen-focus form-

ing virus integrated in the vicinity of the Sfpi-1 locus. Erythroleukemia cells do not differentiate and grow indefinitely in the absence of 

erythropoietin. Activation of the transcription factor PU.1, encoded by the Sfpi-1 gene, is thought to be responsible for the transformed

phenotype. These cells can overcome the blockage and reinitiate their differentiation program when exposed to some chemical inducers such

as hexamethylene bisacetamide. In this study, we established cell cultures that were capable to proliferate unconstrained in the presence of the

inducer. Resistant cell lines restart erythroid differentiation, though, if forced to exit the cell cycle or by overexpressing the transcription factor

GATA-1. Unexpectedly, expression of PU.1 was suppressed in the resistant clones albeit the spleen-focus forming virus was still integrated

in the proximity of the Sfpi-1 locus. Exposure to 5-Aza-2-deoxycytidine activates PU.1 expression suggesting that the PU.1 coding gene is

highly methylated in the resistant cells. Altogether these results suggest that PU.1 is dispensable to block erythroid differentiation.

© 2007 Elsevier Ltd. All rights reserved.

Keywords: Erythroleukemia; Friend cells; PU.1; GATA-1; SFFV; ETS transcription factor

1. Introduction

Friend disease is an acute erythroleukemia that occurs

in susceptible strains of mice following infection with

a replication-defective spleen focus-forming virus (SFFV)

and its natural helper Friend murine leukemia virus (F-

MuLV) [1–3]. Mice infected with the Friend complex

(SFFV + MuLV) develop a multistep tumorigenesis with two

clearly defined stages. During the initial stage, a 55 kDa

glycoprotein (gp55) encoded by the env gene of SFFV acti-

vates the erythropoietin receptor (Epo-R) causing a massive

production of Epo-independent erythroblasts [4]. Gp55 also

activates sf-STK, a truncated form of the receptor tyrosine

∗ Corresponding author at: Department of Cell and Developmental Biol-

ogy, Centro de Investigaciones Biologicas (CSIC), Ramiro de Maeztu 9,

28040 Madrid, Spain. Tel.: +34 918373112x4238.

E-mail address: dbkrimer@cib.csic.es (D.B. Krimer).

kinase STK, promoting an abnormal erythroid proliferation

[5,6]. This stage is designated preleukemic as cells retain the

capacity to terminally differentiate and are not transplantable

when injected into anemic or X-irradiated mice [7]. A second

tumorigenic stage arises when SFFV integrates upstream of 

the promoter of the Sfpi-1 gene [8–10]. This leads to the acti-

vation of the gene product PU.1, an Ets transcription factor

that normally remains silent in the erythroid lineage [11]. At

this stage cells are blocked in their differentiation program

and are capable to grow indefinitely in semisolid or liquid

cultures in vitro. PU.1 inappropriate expression is thought

to be the main cause of the transformed phenotype. More-

over, substantial evidence supports the oncogenic properties

of PU.1 in transgenic mice and long-term bone marrow cell

cultures [3,12,13].

Murine erythroleukemia (MEL) cell lines derive from

Friend complex transformed erythroblasts isolated from the

second stage of the disease [14]. As in untransformed

0145-2126/$ – see front matter © 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.leukres.2007.05.008

8/7/2019 Paper PU.1

http://slidepdf.com/reader/full/paper-pu1 2/10

122 M.J. Fern´ andez-Nestosa et al. / Leukemia Research 32 (2008) 121–130

erythroblasts, MEL cells do not differentiate and grow indef-

initely in the absence of erythropoietin. MEL cells can

overcome the blockage, however, and reinitiate differenti-

ation after exposure to different chemical inducers. This

feature makes MEL cells an invaluable model to study repro-

gramming of tumor cells to a non-malignant phenotype and

to analyze the mode of action of different chemotherapeuticcompounds. MEL differentiation in vitro involves two par-

tially overlapping events. Cell cycle withdrawal occurs in

the initial phases and is essential to proceed to the differ-

entiation stage. Some observations indicate that phenotypic

differentiation and terminal cell division, however, are not

necessarily coupled [15]. Down regulation of genes charac-

teristic of a proliferative status, including several oncogenes

such as myc, myb and PU.1, together with terminal cell

divisions and G1 accumulation, go along with cell cycle

arrest [16–20]. Concomitantly, the expression of a number

of differentiation markers and chromatin remodeling lead to

reactivation of the erythroid differentiation program [21–23].

In this complicated scenario, crosstalk between PU.1 and theerythroid transcription factor GATA-1 appears to be criti-

cal. PU.1 can bind and repress GATA-1 acting as a negative

regulator of erythroid development [24–27]. Differentiation

blockage, however, can be overcome by providing the cells

with additional GATA-1 [28].

In an attempt to identify targets for the chemical inducers

of differentiation in MEL cells we established resistant MEL

cultures capable to grow indefinitely in the presence of hex-

amethylene bisacetamide (HMBA). These variant cell lines

can resume differentiation when forced to withdraw from the

cell cycle or, alternatively, when overexpress GATA-1. Unex-

pectedly, we found that expression of PU.1 was suppressed inthe resistant clones even though SFFV remained integrated

at a similar location than in the MEL parental lines. Treat-

ment with 5-Aza-2-deoxycytidine (5-azaC) restored PU.1

expression suggesting that the gene coding for PU.1 is highly

methylated in HMBA-resistant cells.

2. Materials and methods

2.1. Cell culture

MEL-DS 19 cells were maintained in Dulbecco’s modi-

fied Eagle’s (DME) medium supplemented with 10% fetal

bovine serum (FBS) and 100 units/ml of penicillium and

streptomycin (Gibco). DP18 cells were maintained in alpha

minimum essential medium (-MEM) supplemented with

10% fetal bovine serum (FBS) and 100 units/ml penicillium

and streptomycin (Gibco). Differentiation was induced by

exposing logarithmically growing cultures to 5 mM HMBA

or 1% DMSO. MEL-resistant cells (MEL-R) were estab-

lished by successive passes in the presence of 5 mM HMBA

and were routinely cultured in DME with the inducer.

Hemoglobinized cells were monitored by determining the

proportionof benzidine-stainingpositivecells (B+)in the cul-

ture. To analyze epigenetic changes of the PU.1 locus MEL

and MEL-R cells were grown in the absence or presence

of either 0,4M 5-Aza-2-deoxycytidine (5-azaC, Sigma) or

50 nM Trichostatin A (TSA, Sigma). At specific time points,

cells were harvested and PU.1 expression was analyzed by

RT-PCR.

2.2. Flow cytometry

Flow cytometry was performed using an EPICS XL flow

cytometer (Beckman-Coulter). For DNA content analysis,

1× 106 MEL and MEL-R cells were centrifuged at 1000 ×g

for 5 min, washed in PBS and resuspended in PBS containing

0.05% NP-40, 60g/ml RNAse A and 25g/ml propidium

iodide (PI). Samples were incubated for 20 min at 37 ◦C

and analyzed using FL3 (620/25) filters. Green fluorescent

protein (GFP) positive cells were sorted with a fluorescence-

activated cell sorter (FACS Vantage, Becton-Dickinson).

2.3. RT-PCR analysis

Total RNA was isolated from 5 × 106 cells using Trizol

reagent as described by the supplier (Gibco BRL). For semi-

quantitative RT-PCR reactions 5g of total RNA extracted

from MEL and MEL-R cells was reverse transcribed with

M-MLV reverse transcriptase (USB) as previously described

[21]. PCR amplification was performed using 200M each

nucleotide, 0.5M sense and antisense primers and 5 units

of Recombi-Taq (LINUS). The conditions for the amplifi-

cation cycles were: denaturing at 95 ◦C for 2 min, followed

by 30 cycles with 94 ◦C for 1 min, elongation at 72 ◦C for

1 min and a final elongation at 72◦

C for 5 min. The follow-ing primers wereused: Myb fw5 -GAT GAT GGG CGC CCC

ACT CAA-3, rv 5-CTC ACA TGA CCA GAG TTC GAG

C-3, Myc fw 5-ACG ACG ATG CCC CTC AAC-3, rv 5-

GTT TAT GCA CCA GAG TTT CG-3, Mad1 fw 5-GAA

GAT CTG CAG GAT GGC GAC AGC CG-3, rv 5-GAA

GAT CTG AGG TGG AGA GGA CTC TC-3, p27 fw 5-

GAG GAG GAA GAT GTC AAA CGT-3, rv 5-CAC CGG

AGC TGT TTA CGT CT-3, GATA-1 fw 5-CCC CAG TGT

TCC CAT GGA TTT TCC TG-3, 5-AAG GTC AAG GCT

ATT CTG TGT ACC T-3,EKLFfw5-TAG CCT CAT AGC

CCA TGA GGC AGA AGA-3, rv 5-CAT CCC CAG TCC

TTG TGC AGG ATC AC-3, PU.1 fw 5-GGA TCT GAC

CAA CCT GGA GC-3, rv 5-GCT CAG GAG CCT GGC

GGTCTC T-3,Fli-1fw5-GAA CTC TGG CCT CAA CAA

AAG-3 and rv 5-TGT CTC TGT TGG ATG TGG CTG-

3, GAPDH fw 5-ACA ACA GTC CAT GCC ATC AC-3

and rv5-TCC ACC ACC CTG TTG TA-3. PCR products

were resolved on agarose gels and visualized after ethidium

bromide staining.

2.4. Plasmid construction and DNA transfections

The expression vector producing the GFP-Mad fusion pro-

tein was constructed by subcloning a 1577 bp BamHI DNA

8/7/2019 Paper PU.1

http://slidepdf.com/reader/full/paper-pu1 3/10

M.J. Fern´ andez-Nestosa et al. / Leukemia Research 32 (2008) 121–130 123

fragment encoding GFP-Mad into the pMTHbetaglobin.neo

vector (kindly provided by T. Bender) digested with the same

enzyme. For RNAi experiments a 64 bp BglII–BamHI frag-

ment spanning position 987–1005 on c-myc exon II was

cloned into the pSUPER neo + gfp vector (Oligo Engine)

digested with BglII. The target sequence for myc spe-

cific knockdown was 5

GAACATCATCATCCAGGAC3

.The inducible vector pEBBpuro GATA-1-ER (a gener-

ous gift from A. Skoultchi) was used to produce a

conditionally active form of GATA-1 [25]. Stable trans-

fectants of MEL and MEL-R were prepared as described

previously [20]. Briefly, recombinant DNA was intro-

duced into exponentially growing MEL cells by lipofectine

(Gibco) and after 6 h incubation cells were distributed into

96-well plates. The transfectants were selected and main-

tained in growth medium containing 5g/ml puromycin

or 600g/ml G418. Cell differentiation was tested by cul-

turing in the absence or presence of 5 mM HMBA with

either 100M ZnCl2 (for GFP-Mad) or 10−7 M -estradiol

(for GATA-1-ER).

2.5. Northern and Southern-blot analysis

Total RNA was isolated using the Trizol Reagent (Gibco

BRL) in accordance with the manufacturer’s protocol. Pre-

hybridization and hybridization were carried out as described

previously [21]. Probes for Northern experiments were

obtained by radioactive labeling a 600 bp EcoRI frag-

ment for α-globin and a 700 bp PstI fragment for c-myc

[20].

For Southern analysis high molecular weight DNA

was isolated from mouse tissues and MEL cells anddigested with BamH1 and EcoRV restriction enzymes.

After separation on 0.7% agarose gels the DNA was trans-

ferred to Zeta-Probe blotting membranes (BioRad) and

hybridized to a PU.1 specific PstI fragment localized 500 bp

upstream from the 5 SFFV LTR, kindly provided by F.

Moreau-Gachelin (probe A in [8]). This probe was labeled

by a non-radioactive procedure and used as described

[29].

2.6. Antibodies and immunoblot analysis

Transfected MEL and MEL-R cells (1 × 107) were pel-

leted, washed twice in cold PBS and lysed in Laemmli

buffer (65 mM Tris–HCl pH 6.8, 10% glycerol, 5% 2-

mercaptoethanol, 1% SDS) containing protease inhibitors

(SIGMA). Proteins (20–50g) were resolved in a 10% SDS-

polyacrylamide gel electrophoresis and transferred to PVDF

membranes (Millipore). Primary antibodies included rabbit

polyclonal anti-Mad C-19 antibody (1:1000; Santa Cruz), rat

monoclonal GATA-1 N6 antibody (1:500; Santa Cruz), rabbit

polyclonal anti-PU1 antibody (1:1000; Santa Cruz), mouse

monoclonal c-myc Ab-5 antibody (1:300; NeoMakers) and

goat polyclonal anti-gp70 (1:10,000, a generous gift from S.

Ruscetti).

3. Results

3.1. Establishment of HMBA-resistant cell lines

The exposure of MEL cells to HMBA leads to

commitment to terminal erythroid differentiation [14]. Spon-

taneous differentiation, measured by benzidine-reactivehemoglobinized cells (B+ cells), was less than 1% in

untreated cultures and increased gradually after 48–72 h with

the inducer. By 5 days about 90% of the cells became B+

(Fig. 1a). We generated HMBA-resistant cell lines (MEL-R)

by growing MEL cells continuously in the presence of 5 mM

HMBA. Cells that initially did not differentiate grew steadily

and after 45–70 days the percentage of B+ cells dropped

to near the parental uninduced levels (Fig. 1a). To exclude

that the observed resistance was restricted to this particular

chemical inducer, we extended our studies to DMSO. Sim-

ilarly, the percentage of differentiation was negligible after

10 days (Fig. 1b). We next tested if resistance required the

continuous presence of the inducer or, alternatively, if a sub-population with a particular deficit has been selected. HMBA

was removed from the cultures and restored after 10, 20 or

30 days. As in cells grown in the continuous presence of 

HMBA, the percentage of B+ cells observed in MEL-R cul-

tures remained very low (Fig. 1c). All these results indicated

that MEL-R cloneshave acquired (or lost) characteristics that

render them incapable to activate the mechanism triggered by

the inducer.

MEL-R cells showed also some phenotypic differences

compared to the parental line. An increment in cell size

(Fig. 1d) and a doubling time prolonged to 16–18 h instead

of 10–12 h seen in MEL cells (data not shown). We examinedalso the distribution of cells in each phase of the cell cycle by

propidium iodide staining and flow cytometry (Fig. 1e). As

previously mentioned, an accumulation in G1 was observed

in MEL cells treated with HMBA already at 48 h that was

prolonged all the way to the final stages (96 h). The profile

of MEL-R cells growing continuously in HMBA, however,

coincided with that one corresponding to uninduced MEL

cells (0 h) and showed a distribution characteristic of a pro-

liferative state. Expression markers of proliferation such as

c-myc or of erythroid differentiation such as α-globin con-

firmed that MEL-R cells remained in a stable proliferative

state having lost the ability to react to the chemical inducer

(Fig. 1f).

3.2. PU.1 expression is suppressed in MEL-R cell lines

The expression pattern of genes characteristic of pro-

liferative or differentiated states was further analyzed by

semi-quantitative RT-PCR in the parental and MEL-R clones.

To this end we included protoncogenes such as c-myc and

c-myb as well as antagonists of the proliferative state such

as mad1, p27  and p21 (Fig. 2a). In a different group we

investigated genes related to the erythroid phenotype either

as promoters of differentiation such as GATA-1 and EKLF 

8/7/2019 Paper PU.1

http://slidepdf.com/reader/full/paper-pu1 4/10

124 M.J. Fern´ andez-Nestosa et al. / Leukemia Research 32 (2008) 121–130

Fig. 1. Establishment and characteristics of MEL-R cells in the continuous presence of HMBA. (a) Percentage of benzidine positive cells (B+) of MEL

cells growing in the absence (0) or in the presence of 5 mM HMBA. Undifferentiated cells after 5–6 days of culture in HMBA were considered MEL-R. (b)

Percentage of B+ cells in MEL and MEL-R cultures after exposure to 1% DMSO. (c) MEL-R cells remain resistant even when culturing the cells without

HMBA. Percentage of B+ cells after 10, 20 and 30 days in the absence of the inducer (lower section of the y axis) followed by 5 days with 5 mM HMBA (upper

section of the y axis). (d) Increase cell size in MEL-R cells. MEL-R and MEL cells untreated (0 h) or treated with HMBA for 5 days (120 h) were analyzed

by flow cytometry. On the right, aliquots of the same samples were tested for B+ cells by the benzidine assay. (e) Cell cycle parameters of MEL-R cells were

similar to those observed for non-treated MEL cells. Flow cytometry analysis of MEL cells uninduced (0 h) and treated with 5 mM HMBA for 48 and 96 h,

and MEL-R cells growing in the presence of HMBA. (f) Expression markers of proliferation ( c-myc) or erythroid differentiation (α-globin) showed the same

profile in uninduced MEL and MEL-R cultures. Northern blots of MEL cells uninduced (0) or treated with HMBA for 5 days (96) and MEL-R cells hybridized

to c-myc and α-globin specific probes. Ethidium bromide staining was used as a control of integrity and equal loading of the samples.

or blockers of the erythroid differentiation program such as

PU.1 and Fli-1 (Fig. 2b). In all cases, except for PU.1, the

expression pattern obtained for MEL-R clones was similar to

that one observed for the undifferentiated parental MEL cells

(lanes0andRin Fig. 2a andb). As expected, PU.1 expression

was very high in undifferentiated MEL cells. As cells com-

mitted to differentiate, however, there was a temporal decline

observed at the 48-h sample followed by a minor recover at

later stages. This bimodal pattern was previously observed

andit wasspeculatedto be the consequence of a general down

regulation during early cell commitment [20]. PU.1 expres-

sion, however, wasabolished in the MEL-R pools established

from three independent experiments (Fig. 2c, MEL-R, MEL-

R2 anddata notshown) as well as in single-cell derived clones

(Fig. 2c, R7A and R11A). RT-PCR results were confirmed by

Western blot analysis of PU.1 from different resistant lines

(Fig. 2c).

3.3. The SFFV glycoprotein gp55 is expressed in MEL-R

cell lines and SFFV remains integrated at the Sfpi-1

locus as in the parental lines

Theabsenceof PU.1 expression in MEL-Rcellswas unex-

pected. In 70–95% of Friend virus-induced erythroleukemias

SFFV integrates at the Sfpi-1 locus and causes the activation

of the gene that encodes for the PU.1 protein [8–10]. PU.1,

on the other hand, requires the constitutive activation of gp55

to block differentiation [30]. We wanted to check if the lack 

of PU.1 expression was associated with changes in the activ-

ity of gp55 or was the result of rearrangements at the SFFV

integration site. First, we analyzed the expression of gp55

in uninduced and differentiated MEL cells and in MEL-R

cells by Western blotting (Fig. 3a). As a control we included

extracts of the cell line DP18 derived from mice infected

only with SFFV (a generous gift of Dr. Y. Ben-David). The

results showed that gp55 was highly expressed in both MEL

and MEL-R cells. A lower activity of gp55 was detected in

HMBA-treated MEL cells (96 h) when more than 90% of 

the cells were already differentiated. An additional band was

detected that correspondedto thegp70 glycoproteinof MuLV

that cross-reacted with the probe used.

Next, we investigated whether or not the absence of PU.1

expression in MEL-R cells was the result of SFFV rear-

rangements at the Sfpi-1 locus. Genomic DNA isolated from

Friend cell lines and Balb/c mice tissues was digested with

the restriction enzymes BamHI and EcoRV and analyzed by

Southern blotting. After hybridization with a specific SFFV

probe (probe A in [8]) a 10.6 kb band for the BamH1 and a

8/7/2019 Paper PU.1

http://slidepdf.com/reader/full/paper-pu1 5/10

M.J. Fern´ andez-Nestosa et al. / Leukemia Research 32 (2008) 121–130 125

Fig. 2. PU.1 expression was lost in MEL-R cells. (a) RT-PCR analysis for

genes linked to cell proliferation in the parental MEL cells untreated (0) ortreated with HMBA (48, 96) and in MEL-R cells. PCR products were nor-

malized to GAPDH, electrophoresed on a 1% agarose gel and stained with

ethidium bromide. (b) RT-PCR analysis for genes involved in the regula-

tion of erythroid differentiation. PCR products were normalized and stained

as in (a). (c) Western blot analysis of MEL and MEL-R total lysates for

PU.1 expression. MEL corresponds to the parental cell line, MEL-R and

MEL-R2 correspond to HMBA-resistant cell lines established in two inde-

pendent experiments, R7A and R11A correspond to MEL-R single-cell

derived clones isolated by cell sorting, R-HMBA correspond to MEL-R

cultures in theabsence of HMBAfor 30 days andrestoredto a culture media

with 5 mM HMBA. -tubulin was used as a protein loading control.

22.5 kb for the EcoRV treatments corresponding to the nor-

mal allele was evident in all samples (Fig. 3b). An additional

band of about 6 or 9 kb consistent with the size expected for

the rearranged allele was observed in both Friend cell lines.

Altogether, these results demonstrated that in MEL-R cells

SFFV was still integrated upstream of the Sfpi-1 locus at a

location similar to that one observed for the parental line.

Moreover, gp55 remained active in MEL-R cells.

3.4. Ectopic expression of Mad1 or siRNA knockdown of 

Myc induces MEL-R cell differentiation in the presence

of HMBA

MEL terminal differentiation is preceded by an arrest of 

cell proliferation and accumulation of cells in the G1 phase

[31]. Overexpression of the oncogene c-myc prevents exit

from the cell cycle and inhibits cell differentiation [32–34].

On the contrary, the ectopic expression of mad1, an antagonist

of c-myc, results in typical MEL cell differentiation [35]. We

wanted to test if MEL-R cell lines react in a similar way when

cells were forced to withdraw from the cell cycle. To this end

we transfected MEL-R cells with a Zn-inducible expression

vector that expresses a chimeric GFP-Mad protein. Pools of 

transfectants were further selected by cell sorting based on

GFP expression and physically separated in 96-well plates

to obtained single-cell derived clones. Western blots con-

Fig. 3. The SFFV envelope protein gp55 is expressed in MEL-R cells. (a)

Western blot analysis for the expression of gp55 in MEL cells untreated (0)

or treated with HMBA for 5 days (96), MEL-R cells (R) and the DP18 cell

line.The antibodyused to detect gp55cross-reacted withthe MuLV gp70. (b)

SFFV integration in MEL-R occurred at the same location as in the parental

MEL cell line. MEL and MEL-R genomic DNA was isolated, treated with

BamHI and EcoRV restriction enzymes and analyzed by Southern blotting.

A non-radioactive labeled fragment corresponding to a PU.1 specific probe

(probe A in [8]) was used. Balb/c genomic DNA was included as a control.

firmed that GFP-Mad was expressed in MEL-R clones at

similar levels than endogenous Mad1 in differentiated MEL

cells (Fig. 4a and b). Cell differentiation was monitored mea-suring the percentage of B+ cells in cultures grown in the

absence or presence of Zinc with or without HMBA (Fig. 4c

and d). The results showed that spontaneous differentiation

was very low in all transfectant cultures in the absence of 

HMBA. The highest value observed was 25.9% for clone

19 that also expressed high amounts of ectopic GFP-Mad

(Fig. 4b). Differences between cultures with or without Zn

were minimal probably due to leakiness of the metalloth-

ionein promoter. When HMBA was added to the cultures,

however, the percentage of B+ cells increased significantly

for all clones reaching very high values in clones 8, 19

and 22 (Fig. 4c and d). These results indicated that ectopic

expression of mad1 makes MEL-R cells sensitive to HMBA

and capable to trigger differentiation toward the erythroid

lineage.

The ability of Mad1 to induce differentiation has been

correlated with its capacity to antagonize with Myc causing

cell cycle arrest (reviewed in [36]). The fact that Myc and

Mad1 exert opposite effects led us speculate that inhibition

of Myc could interfere with cell cycle progression allow-

ing differentiation of MEL-R cells. To test this hypothesis

we stably transfected MEL-R cells with an expression vec-

tor coding for a 20 bp siRNA targeted to exon 2 of  c-myc

and containing a puromycin-GFP selectable marker. Selec-

8/7/2019 Paper PU.1

http://slidepdf.com/reader/full/paper-pu1 6/10

126 M.J. Fern´ andez-Nestosa et al. / Leukemia Research 32 (2008) 121–130

Fig. 4. Overexpression of Mad1 induced MEL-R cells differentiation in the

presence of HMBA. (a) Western blot analysis of the levels of Mad1 in MEL

cells untreated (0 h) or treated with 5 mM HMBA during 5 days (96h) and

MEL-Rcells. (b) GFP-Mad stable transfectants isolatedby cellsortingbased

on GFP expression were analyzed by Western blotting with an anti-Mad

antiserum. Numbers above the panel represent individual transfectants, M:

mock transfectant. (c) Percentage of differentiation of GFP-Mad transfec-tants in (b)was determined by thebenzidine assay. Cells were cultured in the

absence or presence of Zn,with or without 5 mMHMBA. Each barindicates

the mean and standard deviation of three independent experiments. (d) Ben-

zidine staining of GFP-Mad transfectants cultured with or without HMBA.

Representative fields of samples obtainedfrom experiments describedin (c).

tion by cell sorting and culture of single-cell derived clones

were performed as for mad1 and the decrease in the lev-

els of Myc was confirmed by Western blotting ( Fig. 5a).

Most of the transfected clones showed a significant reduction

in growth rate and their viability was sometimes seriously

compromised (data not shown). The analysis of B+ cells

showed that the siRNA-mediated depletion of Myc did not

allow spontaneous differentiation of MEL-R transfectants

(Fig. 5b). However, when HMBA was added to the culture,

a significant percentage of cells differentiated (Fig. 5b and

c). These results are consistent with the notion that prolif-

eration should diminish to allow differentiation to proceed.

siRNA-mediated depletion of Myc appears to disrupt normal

cell cycle progression. Nevertheless, as it also occurs when

overexpressing Mad1, this is not sufficient to eliminate the

differentiation block and additional mechanisms, triggered

by HMBA, are required for the activation of terminal cell

differentiation.

3.5. Stable MEL-R transfectants that express a

conditional estrogen-activated form of GATA-1

differentiate even without the inducer 

The transcription factor GATA-1 is a key regulator of ery-

thropoiesis that coordinates proliferation arrest with cellular

maturation acting both as an activator and a repressor of different target genes [37]. Previous studies in MEL cells

showed that the blockage to differentiation could be over-

come by providing the cells with additional GATA-1 [28].

Using a similar approach we checked if the ectopic expres-

sionof GATA-1 renders analogous results in HMBA-resistant

cell lines. A conditional GATA-1 fused to the ligand-binding

domain of the estrogen receptor (GATA1-ER) was inserted

in the pEBBpuro expression vector and transfected into MEL

and MEL-R cell lines. Puromycin-resistant clones were iso-

lated by limited dilution and analyzed for the expression of 

the chimeric GATA1-ER protein (Fig. 6a). We next examined

the effect of forced GATA-1 expression on the production

of B+ cells when the clones were grown in the presenceof -estradiol with or without HMBA (Fig. 6b). The results

showed that the ectopic expression of GATA-1 stimulated the

production of hemoglobinized cells. Clones that efficiently

expressed GATA-1-ER reached values close to 50% when

exposed to -estradiol. Addition of HMBA to the media

enhanced this effect somewhat. These results demonstrated

that overexpression of GATA-1 was sufficient to trigger dif-

ferentiation in MEL-R cells. The differentiation attained by

the HMBA-resistant clones, however, was less than that

acquired by the parental MEL cell line ([28] and the results

shown in Fig. 6c). In MEL cells GATA-1 antagonizes with

PU.1 and differentiation is associated with the ability to over-express GATA-1 to overcome a putative PU.1 inhibition. As

PU.1 expression was missing in MEL-R cells, a different

mechanism or an additional partner must affect differentia-

tion activation in a less efficient way.

3.6. The DNA methylation inhibitor 

5-Aza-2-deoxycytidine restores PU.1 expression in

MEL-R cell lines

Epigenetic changes such as methylation of DNA or histone

deacetylation have been proved to be major players in tran-

scriptional gene silencing. Reactivation of silenced genes can

be obtained by treatments with 5-Aza-2-deoxycytidine (5-

azaC) or trichostatin A (TSA), which are known as inhibitors

of methyltransferases or histone deacetylases, respectively

[38,39]. One of the major differences between MEL-R cells

and their parental MEL cells was the unexpected inactivation

of PU.1. To further explore the nature of this inactivation we

asked whether or not 5-azaC or TSA could restore the expres-

sion of PU.1. Exposure of MEL-R cells to 5-azaC resulted in

a significant increase in PU.1 expression compared to that one

observed for uninduced MEL cells (Fig. 7a). On the contrary,

TSA was unable to reverse PU.1 silencing in MEL-R cells

(Fig. 7b). These results suggested that the PU.1 locus is pref-

8/7/2019 Paper PU.1

http://slidepdf.com/reader/full/paper-pu1 7/10

M.J. Fern´ andez-Nestosa et al. / Leukemia Research 32 (2008) 121–130 127

Fig. 5. Specific knockdown of Myc expression led to cell differentiation in response to HMBA in MEL-R cells. (a) MEL and MEL-R cells were stably

transfected with a vector expressing Myc specific siRNA as described in Section 2. Knockdown efficiency was assessed by Western blot analysis in controls(−) and transfected cells (+) with anti-Myc antisera. -Tubulin was used as protein loading control. (b) Percentage of B+ cells in knockdown Myc transfectants

cultured in the absence (−) or presence (+) of 5 mM HMBA. (c) Benzidine staining of mock and siRNA-Myc transfectants from the experiments described in

(b).

Fig. 6. Activation of conditional GATA-1-ER led MEL-R transfectants to

differentiate without the inducer. (a) Western blot analysis of lysates derived

from MEL and MEL-RGATA-1-ER transfectants stimulated by-estradiol.

Equal amounts of protein (50g) were fractionated by SDS-polyacrylamide

gelelectrophoresisand analyzedby immunoblottingwith an anti-GATA anti-

body. M: mock.Numbersabove thepanel correspond to differentclones. (b)

Percentage of benzidine positive cells (B+) in the GATA-1 stable transfec-

tants activated by -estradiol and cultured in the absence (−) or presence

(+) of 5 mM HMBA. Transfectants 9R, 5R and 1R correspond to 9, 5 and

1 clones in (a). (c) Northern blot analysis for -globin mRNA in MEL and

MEL-R GATA-1 stable transfectants stimulated by -estradiol. Ethidium

bromide staining was used as a control of integrity and equal loading of the

samples.

Fig. 7. Reactivation of silenced PU.1 by 5-azaC treatment in MEL-R cells.

(a) RT-PCR analysis for the expression of PU.1 in MEL and MEL-R cells

afterexposure to 0.4M 5-azaC. PCRproducts werenormalized to GAPDH,

electrophoresed on a 1% agarose gel and stained with ethidium bromide. (b)

RT-PCR analysis for the expression of PU.1 in MEL-R cells after 5 or 24-h

exposure to 50 nM TSA.

erentially methylated in MEL-Rcells causing silencing of the

gene and possibly their inability to respond to differentiation

inducers.

4. Discussion

PU.1 and GATA-1 are specific transcription factors that

regulate myeloid and erythroid lineage development, respec-

tively. Several lines of evidence suggest that PU.1 and

GATA-1 antagonizeeach otherfunctionand thatlineage com-

mitment depends on the relative ratio of these two proteins

[40]. The reciprocal negative regulation of PU.1 and GATA-

1 resides in a physical interaction that takes place between

the C-terminal zinc finger domain of GATA-1 and the Ets

8/7/2019 Paper PU.1

http://slidepdf.com/reader/full/paper-pu1 8/10

128 M.J. Fern´ andez-Nestosa et al. / Leukemia Research 32 (2008) 121–130

domain of PU.1 [24–26]. Sometimes, additional co-factors

collaborate to form repressive or active protein complexes.

Rekhtman and co-workers showed that PU.1 must cooperate

with the retinoblastoma protein (pRB) to repress GATA-1

and block erythroid differentiation [41]. Later on, the same

group demonstrated that PU.1 binds GATA-1 on its target

genes and recruits a repression complex that includes pRB,HP1 and the histone methyltransferase Suv39h, creating

an inactive chromatin structure [23]. In the myeloid lineage

GATA-1 inhibits PU.1 function by disrupting PU.1’s interac-

tion with its essential co-factor c-Jun [26]. Recent molecular

analysis revealed that the interaction between GATA-1 and

PU.1 is very weak  [42]. These results led the authors to

suggest a possible mechanism that allow rapid changes in

regulation.

In the present study we analyzed the differentiation block-

age of established murine erythroleukemia cell lines that

are resistant to the action of chemical inducers. MEL-R

cells do not react to HMBA or DMSO and cells remain in

an undifferentiated state even in the continuous presenceof the inducers. We showed that MEL-R cells could dif-

ferentiate, however, if cell cycle progression is inhibited.

Differentiation of the parental MEL cell lines is also accom-

panied by withdrawal from the cell cycle. Previous reports

indicated that the constitutive expression of several genes

associated with cell proliferation such as the oncogenes c-

myc and c-myb blockcell differentiation[32–34,43]. Besides,

overexpression of Mad1, a natural antagonist of c-Myc, inter-

rupts the differentiation process [35]. Here we demonstrated

that overexpression of Mad1, led to a moderate percent-

age of hemoglobinized cells in MEL-R cells. Interestingly,

HMBA potentiates this effect. Mad1 transfectants that effi-ciently expressed the exogenous protein, as clones 8, 9 and

22 (Fig. 4b and c) reached very high differentiation values.

A clear effect on differentiation was also observed when

siRNA-Myc transfectants were cultured with the inducer

(Fig. 5b and c). In those cases, however, a lower percent-

age of B+ cells was observed most likely due to the fact that

knockdown of Myc was only partial. Cells transfected with

forms producing more efficient Myc interference would not

probably be viableas suggestedby several clonesthat showed

extreme growth retardation (data not shown).

In addition to cell cycle arrest, MEL cell differentiation

requires the activity of specific master regulators such as

GATA-1. Previous results showed that stable transfectants

that express a conditional estrogen-activated form of GATA-

1 differentiate without the inducer [28]. Consistent with

those results and using similar constructs, here we demon-

strated that ectopic expression of GATA-1 was sufficient to

overcome differentiation blockage in MEL-R cells. Taken

together, these observations suggest that MEL-R cells resis-

tance to chemical inducers is associated with a failure to

inactivate proliferation signals. Enforcing cell cycle arrest

by deregulating c-Myc or Mad1, on the other hand, might

restore the requirements for cell differentiation and render

the cells responsive to the inducers. It is tempting to spec-

ulate that the increase in cell size we observed for MEL-R

cells is connectedto this lack for properinactivation.Previous

data demonstrated a direct association between deregulation

of c-Myc and cell growth [44,45]. We detected no major

differences in Myc content or any mutation that increases

Myc stability (data not shown), although we do not dis-

card other possible modifications. It was recently shown thataberrant stabilization of the c-Myc protein occur in some

lymphoblastic leukemias due to abnormal phosphorylation

at two conserved sites [46].

Contrary to the results obtained with the Myc and Mad

transfectants, which relies on the presence of HMBA, the

ectopic expression of GATA-1 by itself is sufficient to induce

differentiation in MEL-R cells. As previously reported for

the parental MEL cell line [28], the level of expression

of GATA-ER is lower than that of endogenous GATA-1

(Fig. 6a), yet there is an increase in differentiation. This

might reflect a sensitive balance requirement for GATA-1 or,

alternatively, an endogenous GATA-1 impeded to react. The

effect of ectopic GATA-1 expression on MEL-R is probablyconnected to both differentiation requirements: proliferation

arrest and phenotypic differentiation. GATA-1 forms distinct

activating and repressing protein complexes linking differ-

ent partners [37,47,48]. Rylski and co-workers showed that

GATA-1 inhibits c-myc expression and causes cellcycle arrest

[37]. On the other hand, GATA-1 is able to establish distinct

activating complexes that are essential for erythroid differen-

tiation [47,48].

The most striking result obtained in the present study was

the inactivation of PU.1 observed for MEL-R cells. Activa-

tion of PU.1 due to proviral insertion upstream of the Sfpi-1

gene is a primary transforming event associated with SFFV-induced erythroleukemia [3]. The constitutive up-regulation

of this gene is thought to be the main cause for a blockage

in the differentiation process of the affected erythroblasts.

Additional data confirms that in MEL cells overexpression

of PU.1 impedes chemically induced differentiation [18,49]

and that PU.1 silencing obtained via siRNA leads to termi-

nal differentiation [50]. There is also substantial evidence to

suggest that PU.1 blocks erythroid differentiation by binding

to and repressing GATA-1 [25,27]. In MEL-R cell lines PU.1

remains silent and confined to a hypermethylated region, as

suggested by the azacytidine results. The SFFV glycopro-

tein gp55 is expressed in both the MEL parental cell line and

MEL-R cells and the proviral DNA is integrated upstream

of the PU.1 promoter in both cases. These results ruled out

the possibility that a distinct retrovirus (i.e. MuLV) could be

responsible for the MEL-R cells transformation or that SFFV

integrated at a different genomic location. It is not clear, how-

ever, why MEL-R cells are blocked and unable to complete

the differentiation program and in addition do not respond

to chemical inducers. It is possible that a different partner

other than PU.1 could inhibit GATA-1 expression in MEL-R

cells and also restrain their response to the inducer. Several

GATA-1 partners have been described as possible transcrip-

tional repressors [51] and current experiments are in progress

8/7/2019 Paper PU.1

http://slidepdf.com/reader/full/paper-pu1 9/10

M.J. Fern´ andez-Nestosa et al. / Leukemia Research 32 (2008) 121–130 129

in our laboratory to characterize those putatively involved in

MEL-R cells.

Acknowledgments

We thankDrs. T. Bender for the pMTHbetaglobin.neovec-tor, A.I. Skoultchi for pEBBpuro GATA-1-ER plasmid, F.

Moreau-Gachelin for the PU.1 probe, Y. Ben-David for the

DP-18 cell line and S. Ruscetti for the gift of anti-gp70 anti-

body. We thank P. Santos for help with the 5-AzaC and TSA

experiments. We are grateful to M.L. Martınez for technical

assistance. This work was supported in part by the Spanish

Ministerio de Educacion y Ciencia Grants BIO2005-02224

and BFU2004-00125. MJFN is supported by the Consejerıa

de Educacion de la Comunidad de Madrid, Fondo Social

Europeo and a fellowship from the Residencia de Estudiantes

(CSIC)-Ayuntamiento de Madrid.

References

[1] Ben-David Y, Bernstein A. Friend virus-induced erythroleukemia and

the multistage nature of cancer. Cell 1991;66:831–4.

[2] FriendC. Cell-freetransmissionin adult Swiss mice ofa disease having

the character of a leukemia. J Exp Med 1957;105:307–18.

[3] Ruscetti SK. Deregulation of erythropoiesisby the Friend spleen focus-

forming virus. Int J Biochem Cell Biol 1999;31:1089–109.

[4] Li JP, D’Andrea AD, Lodish HF, BaltimoreD. Activationof cellgrowth

by binding of Friend spleen focus-forming virus gp55 glycoprotein to

the erythropoietin receptor. Nature 1990;343:762–4.

[5] Nishigaki K, Thompson D, Hanson C, Yugawa T, Ruscetti S. The

envelope glycoprotein of friend spleen focus-forming virus covalentlyinteracts withand constitutively activates a truncatedform of the recep-

tor tyrosine kinase Stk. J Virol 2001;75:7893–903.

[6] Persons DA, Paulson RF, Loyd MR, HerleyMT, BodnerSM, Bernstein

A, et al. Fv2 encodes a truncated form of the Stk receptor tyrosine

kinase. Nat Genet 1999;23:159–65.

[7] Lee CR, Cervi D, Truong AH, Li YJ, Sarkar A, Ben-David Y. Friend

virus-induced erythroleukemias: a unique and well-defined mouse

model for the development of leukemia. Anticancer Res 2003;23:

2159–66.

[8] Moreau-Gachelin F, Tavitian A, Tambourin P. Spi-1 is a putative onco-

gene in virally induced erythroleukemias. Nature 1988;331:277–80.

[9] PaulR, Schuetze S, KozakSL, KabatD. A common site forimmortaliz-

ing proviral integrations in Friend erythroleukemia: molecular cloning

and characterization. J Virol 1989;63:4958–61.

[10] Paul R, Schuetze S, Kozak SL, Kozak CA, Kabat D. The Sfpi-1 provi-ral integration site of Friend erythroleukemia encodes the ets-related

transcription factor Pu.1. J Virol 1991;65:464–7.

[11] SchuetzeS, PaulR, Gliniak BC, KabatD. Role ofthe PU.1 transcription

factor in controlling differentiation of Friend erythroleukemia cells.

Mol Cell Biol 1992;12:2967–75.

[12] Kabat D. Molecular biology of Friend viral erythroleukemia. Curr Top

Microbiol Immunol 1989;148:1–42.

[13] Ney PA, DAndrea AD. Friend erythroleukemia revisited. Blood

2000;96:3675–80.

[14] Marks P, Rifkind R. Erythroleukemic Differentiation. AnnRevBiochem

1978;47:419–48.

[15] Matushansky I, Radparvar F, Skoultchi AI. Manipulating the onset

of cell cycle withdrawal in differentiated erythroid cells with cyclin-

dependent kinases and inhibitors. Blood 2000;96:2755–64.

[16] Kirsch IR, Bertness V, Silver J, Hollis GF. Regulated expression of the

c-myb and c-myc oncogenes during erythroid differentiation. J Cell

Biochem 1986;32:11–21.

[17] Lachman HM, Skoultchi AI. Expression of c-myc changes during

differentiation of mouse erythroleukaemia cells. Nature 1984;310:

592–4.

[18] Rao G, Rekhtman N, Cheng G, Krasikov T, Skoultchi AI. Dereg-

ulated expression of the PU.1 transcription factor blocks murine

erythroleukemia cell terminal differentiation. Oncogene 1997;14:

123–31.

[19] Vanegas N, Castaneda V, Santamarıa D, Hernandez P, Schvartzman

J, Krimer D. Cloning, sequencing and expression in MEL cells of 

cDNA encoding the mouse ribosomal protein S5. Biochem Biophys

Acta 1997;1357:1–4.

[20] Vanegas N, Garcıa-Sacristan A, Lopez-Fernandez L, Parraga M, Del

Mazo J, Hernandez P, et al. Differential expression of Ran GTPase

during HMBA-induced differentiationin murine erythroleukemia cells.

Leukemia Res 2003;27:607–15.

[21] Garcıa-SacristanA,Fernandez-NestosaMJ, Hernandez P, Schvartzman

JB, Krimer DB. Protein kinase clk/STY is differentially regu-

lated during erythroleukemia cell differentiation: a bias toward the

skipped splice variant characterizes postcommitment stages. Cell Res

2005;15:495–503.

[22] Krimer DB, Cheng G, Skoultchi AI. Induction of H3.3 replace-

ment histone mRNAs during the precommitment period of murine

erythroleukemia cell differentiation. Nucleic Acids Res 1993;21:

2873–9.

[23] Stopka T, Amanatullah DF, Papetti M, Skoultchi AI. PU.1 inhibits

the erythroid program by binding to GATA-1 on DNA and creating

a repressive chromatin structure. EMBO J 2005;24:3712–23.

[24] Nerlov C, Querfurth E, Kulessa H, Graf T. GATA-1 interacts with

the myeloid PU.1 transcription factor and represses PU. 1-dependent

transcription. Blood 2000;95:2543–51.

[25] Rekhtman N, Radparvar F, Evans T, Skoultchi AI. Direct interaction

of hematopoietic transcription factors PU.1 and GATA-1: functional

antagonism in erythroid cells. Genes Dev 1999;13:1398–411.

[26] Zhang P, Behre G, Pan J, Iwama A, Wara-Aswapati N, Radom-

ska HS, et al. Negative cross-talk between hematopoietic regulators:GATA proteins repress PU.1. Proc Natl Acad Sci USA 1999;96:

8705–10.

[27] Zhang P, Zhang X, Iwama A, Yu C, Smith KA, Mueller BU, et al. PU.1

inhibits GATA-1 function and erythroid differentiation by blocking

GATA-1 DNA binding. Blood 2000;96:2641–8.

[28] Choe KS, Radparvar F, Matushansky I, Rekhtman N, Han X, Skoultchi

AI. Reversal of tumorigenicity and the block to differentiation in ery-

throleukemia cells by GATA-1. Cancer Res 2003;63:6363–9.

[29] Olavarrieta L, Martinez-Robles ML, Hernandez P, Krimer DB,

Schvartzman JB. Knotting dynamics during DNA replication. Mol

Microbiol 2002;46:699–707.

[30] Quang CT, Wessely O, Pironin M, Beug H, Ghysdael J. Cooperation

of Spi-1/PU.1 with an activated erythropoietin receptor inhibits apop-

tosis and Epo-dependent differentiation in primary erythroblasts and

induces their Kit ligand-dependent proliferation. EMBO J 1997;16:5639–53.

[31] Marks PA, Richon VM, Rifkind RA. Cell cycle regulatory proteins

are targets for induced differentiation of transformed cells: molecular

and clinical studies employing hybrid polar compounds. Int J Hematol

1996;63:1–17.

[32] Coppola JA, ColeMD. Constitutive c-myc oncogeneexpression blocks

mouse erythroleukaemia cell differentiation but not commitment.

Nature 1986;320:760–3.

[33] Dmitrovsky E, Kuehl WM, Hollis GF, Kirsch IR, Bender TP, Segal S.

Expression of a transfected human c-myc oncogene inhibits differenti-

ation of a mouse erythroleukaemia cell line. Nature 1986;322:748–50.

[34] Prochownik EV, Kukowska J. Deregulated expression of c-myc

by murine erythroleukaemia cells prevents differentiation. Nature

1986;322:848–50.

8/7/2019 Paper PU.1

http://slidepdf.com/reader/full/paper-pu1 10/10

130 M.J. Fern´ andez-Nestosa et al. / Leukemia Research 32 (2008) 121–130

[35] Cultraro CM, Bino T, Segal S. Function of the c-Myc antagonist Mad1

during a molecular switch from proliferation to differentiation. Mol

Cell Biol 1997;17:2353–9.

[36] Grandori C, Cowley SM, JamesLP, Eisenman RN. The Myc/Max/Mad

network and the transcriptional control of cell behavior. Annu Rev Cell

Dev Biol 2000;16:653–99.

[37] Rylski M, Welch JJ, Chen YY, Letting DL, Diehl JA, Chodosh LA, et

al. GATA-1-mediated proliferation arrest during erythroid maturation.

Mol Cell Biol 2003;23:5031–42.

[38] Issa JP, Kantarjian H, Azacitidine. Nat Rev Drug Discov 2005;

(Suppl):S6–7.

[39] Yoshida M, Kijima M, Akita M, Beppu T. Potent and specific inhibi-

tion of mammalian histone deacetylase both in vivo and in vitro by

trichostatin A. J Biol Chem 1990;265:17174–9.

[40] Cantor AB, Orkin SH. Hematopoietic development: a balancing act.

Curr Opin Genet Dev 2001;11:513–9.

[41] Rekhtman N, Choe KS, Matushansky I, Murray S, Stopka T, Skoultchi

AI. PU.1 and pRB interact and cooperate to repress GATA-1 and block 

erythroid differentiation. Mol Cell Biol 2003;23:7460–74.

[42] Liew CW, Rand KD, Simpson RJ, Yung WW, Mansfield RE, Crossley

M, et al. Molecular analysis of the interaction between the hematopoi-

etic master transcription factors GATA-1 and PU.1. J Biol Chem

2006;281:28296–306.

[43] Clarke M, Kukowska-Latallo J, Westin E, Smith M, Prochownik E.

Constitutive expression of a c-myb cDNA blocks erythroleukemia cell

differentiation. MolCellBiol 1988;8:884–92.

[44] Iritani BM, Eisenman RN. c-Myc enhances protein synthesis and cell

size during B lymphocyte development. Proc Natl Acad Sci USA

1999;96:13180–5.

[45] Johnston LA, Prober DA, Edgar BA, Eisenman RN, Gallant P.

Drosophila myc regulates cellular growth during development. Cell

1999;98:779–90.

[46] Malempati S, Tibbitts D, Cunningham M, Akkari Y, Olson S, Fan G,

et al. Aberrant stabilization of c-Myc protein in some lymphoblastic

leukemias. Leukemia 2006;20:1572–81.

[47] Rodriguez P, Bonte E, Krijgsveld J, Kolodziej KE, Guyot B, Heck 

AJR, et al. GATA-1 forms distinct activating and repressive complexes

in erythroid cells. EMBO J 2005;24:2354–66.

[48] Welch JJ, Watts JA, Vakoc CR, Yao Y, Wang H, Hardison RC, et al.

Global regulation of erythroid gene expression by transcription factor

GATA-1. Blood 2004;104:3136–47.

[49] Yamada T,Kondoh N, Matsumoto M, Yoshida M, Maekawa A, Oikawa

T. Overexpression of PU.1 induces growth and differentiation inhibi-

tion and apoptotic cell death in murine erythroleukemia cells. Blood

1997;89:1383–93.

[50] Atar O, Levi BZ. PU.1 silencing leads to terminal differenti-

ation of erythroleukemia cells. Biochem Biophys Res Commun

2005;329:1288–92.

[51] Ferreira R, Ohneda K, Yamamoto M, Philipsen S. GATA1 function,

a paradigm for transcription factors in hematopoiesis. Mol Cell Biol

2005;25:1215–27.