From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
RAPID COMMUKICATIOK
Dynamics of GATA Transcription Factor Expression During Erythroid
Differentiation
By Mark Leonard, Martha Brice, James Douglas Engel, and Thalia Papayannopoulou
Although the formation of terminally differentiated erythroid cells has been shown to require the presence of a
functional GATA-1 gene in vivo, the role of this transcription factor and other members of the GATA family at earlier stages of erythroid differentiation is unclear. In this report, the expression of GATA-1, GATA-2, and GATA-3 has
been examined in enriched peripheral blood progenitors
before and after culture in a well-characterized liquid culture system. In addition primary leukemic cells as well as
several erythroleukemic and nonerythroid cell lines were
analyzed for GATA factor expression. The results show
that the profile of GATA factor expression in erythroid cells
is distinct from that of myeloid or lymphoid lineages. Erythroleukemic cell lines express little or no GATA-3, but
high levels of GATA-1 and GATA-2. When they are induced t o display the terminal erythroid phenotype, little
change in the level of GATA-1 is detected but a significant
decline in the levels of GATA-2 is observed commensurate
with the degree of maturation achieved by the cells. Enrichment of erythroid progenitors from peripheral blood leads
to selection of cells that express both GATA-1 and GATA2. As the enriched populations are cultured in suspension
in the presence of multiple cytokines, the levels of both
GATA-1 and GATA-2 initially increase. However, in cultures containing only erythropoietin, which show exclusive
erythroid differentiation, the levels of GATA-1 continue t o
increase, whereas GATA-2 expression declines as erythroid maturation progresses. In contrast, cultures lacking
Epo (ie, with interleukin-3 and kit ligand) display limited
progression towards both the myeloid and erythroid pathways, and high levels of expression of both GATA-1 and
GATA-2 are maintained. Despite the initial upregulation of
GATA-1 expression in the latter cultures, terminal erythroid differentiation does not occur inthe absence of erythropoietin. These results indicate that GATA-1 upregulation
is associated with both the initiation and the maintenance
of the erythroid program, but that these t w o processes appear t o be under separate regulatory control. Thus, the dynamic changes in the levels of different GATA factors that
occur during primary erythroid differentiation suggest that
the levels of these factors may influence the progression t o
specific hematopoietic pathways.
0 1993 by The American Society of Hematology.
T
and after induction of erythroid differentiation. We find
that both GATA- 1 and GATA-2 are expressed in enriched
progenitor populations and both are upregulated further in
the progeny of CD34( +) cells under the influence of hematopoietic cytokines, especially interleukin-3 (IL-3) and kit ligand (KL). However, in contrast to GATA-1, which is further induced in Epo-containing cultures in parallel with
terminal erythroid differentiation, a significant decline in
GATA-2 is observed in the same cultures. A similar phenomenon is also observed upon induction of erythroleukemia cell lines. It appears that the initial upregulation of
GATA-1 expression occurs in the absence of Epo and is
accompanied by the initiation of globin synthesis. However,
this increased level of GATA- 1 is not by itself sufficient for
the maintenance or stabilization of the erythroid phenotype
and for the survival of these cells. Thus, although it is possi-
RANSCRIPTION FACTORS recognizing the GATA
DNA consensus motif have now been described in a
large number of vertebrate species and appear to exhibit a
conserved pattern of expression during development and, in
GATA- 1, the first characterparticular, in hematopoie~is."~
ized factor in this multigene family, is abundantly expressed
in erythroid-lineage cells and has proven to be a pivotal
regulator of terminal erythroid differentiation in vivo."
Furthermore, its presence has been recently documented in
proliferating, factor-dependent cell lines" and in primary
hematopoietic cells with progenitor cell characteristics.11J2
However, GATA-1 is also expressed in nonerythroid lineages (megakaryocytic and mast cells and in testes)l3-l5and
additional members of the GATA factor family are expressed in erythroid cells.* Thus, the precise role of GATA- 1
in defining erythroid specificity and the stage of erythroid
differentiation at which GATA- 1 exerts its specific effects
are unclear. Furthermore, parameters that govern the initiation and the maintenance of the erythroid phenotype have
not been clearly defined. The contribution of other transcription factors expressed in erythroid cells (ie, SCL, NFE2, Myb, Rb, or other GATA members) or the effects of
growth factor receptor- (ie, erythropoietin receptor [EpoR]) mediated signals potentially modulating the stabilization and/or induction of GATA-1 during terminal erythroid differentiation have not been explored.
We have investigated the patterns of expression of
GATA- 1, GATA-2, and GATA-3 in enriched populations
of progenitors from normal peripheral blood and have observed their changes during differentiation under the influence of hematopoietic cytokines. In addition to primary
cells, we have also examined the relative changes in abundance of GATA factors in erythroleukemia cell lines before
Blood, VOI 82, NO4 (August 15). 1993: pp 1071-1079
From the Division of Hematology, University of Washington,
Seattle, WA 98195; andthe Department OfBiochemistry,Northwestern University, Evanston, IL.
Submitted March 31, 1993; accepted April 30, 1993.
Supported by National Institutes of Health Grants No. HL46557,
GM28846, and DK30852. M.L. is a Leukemia Society of America
Fellow.
Address reprint requests to Thalia Papayannopoulou, MD, Division ofHematology, University of Washington, RM-IO, Seattle, WA
98195.
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.section I734 solely to
indicate thisfact.
0 1993 by The American Society of Hematology.
0006-49 71/93/8204-0039$3.00/0
1071
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
LEONARD ET AL
IO72
ble that increased GATA-1 expression is required for the
initiation of erythroid phenotype and globin expression, additional interactive regulatory events are needed for its
maintenance. GATA-3 does not seem to have a clear role in
human erythromyeloid differentiation.
MATERIALS AND METHODS
Cells
Peripheral blood cells from normal individuals were used as a
source of normal cells and blood from leukemic (acute myeloid
leukemia [AML]or chronic myeloid leukemia [CML]) patients as a
source of primary leukemic blasts. Fetal liver samples were obtained, with informed maternal consent, from spontaneousor therapeutic abortions through the Central Laboratory for Human Embryology, University of Washington, Seattle. All procedures were
performed according to institutionally approved protocols. To
enrich for hematopoietic progenitors, peripheral blood mononuclear cells (PBMC), recovered after centrifugation of platelet apheresis collections, were subjected to a direct, positive immunoadherence to Petri dishes precoated with anti-CD34 or anti-c-kit
monoclonal antibody (SR- 1)16 according to previously described
procedures.17 Adherent cells were recovered from the plates after
flushing with warm Iscove’s modified Dulbecco’s medium/fetal calf
serum (IMDM/FCS) and designated as PBMC-P. If enough cells
were availableafter the first direct panning, a second round ofdirect
adherencewas performed to further increasethe specificity of adherence or the progenitor enrichment.”
Cell Lines
Eight human erythroleukemia lines were examined: KU8 12,”
JK-1,” MB02;’ HEL,” OClMl and OClM2,2* CMK?’ and
F-36.24In addition, several nonerythroid lines (HL60 and EM-2),
as well as Epstein-Barr virus (EBVttransformed lymphocytes and
other cell lines (IM9, U266, CEM, HSB2, HuT78, MOLT-3, and
HELA cells)25were studied.
Cell Cultures
Liquid suspension cultures. Enriched populations of progenitors were cultured usually (1 X lo5cells) in 1- or 2-mL aliquots for 6
to 14 days in stationary suspension cultures. Eighty percent to 90%
of the medium and additives were replaced every 48 hours. The
basic suspension medium was serum-containing (IMDM + 20%
FCS + IO% human AB serum) with 1% bovine serum albumin
(BSA),
mol/L 2-mercaptoethanol.To these media, the various
recombinant hematopoietic growth factors were added in microliter quantities: IL-3 (GeneticsInstitute, Cambridge, MA) was used
at 50 U/mL; KL (stem cell factor [SCF]; Amgen, Thousand Oaks,
CA) at 50 ng/mL; Epo (Genetics Institute) at 2 lU/mL; and PIXY
32 I (a granulocyte-macrophage colony-stimulating factor [GMCSF]/IL-3 fusion protein; Immunex, Seattle, WA) at 10 pg/mL.
Clonal cultures. Peripheral blood cells before and after their
purification were cultured in clonal methylcellulosecultures to assess the frequency of clonogenic hematopoietic progenitors and the
degree of enrichment. Methylcelluloseculture media were as previously described12and colonies of different lineages (burst-forming
unit-erythroid [BFUe], colony-forming unit granulocyte-macrophage [CFU-GM], CFU-Mix)were evaluated in situ on the basis of
specific morphologic criteria.”
Erythroid induction of erythroleukemia lines. MB02 and F-36
cells were induced with erythropoietin (6 lU/mL) for about 2
weeks, whereas JK-I was induced by hemin (50 pmol/L) and
OClM2 by &ALA (500 pmol/L) for about 1 week.
Globin Analysis
Presence of globin was assessed by immunofluorescenceon fixed
centrifuge-prepared smears. Antibodies specific for @- or y-globin,
prepared in our
were used at optimal concentrations
in phosphate-bufferedsaline (PBS)/BSA. In addition to immunofluorescence,globin mRNA was evaluated by a slot blot hybridization
assay as previously described?’ Probes used and the procedure were
described previously in detai1F7
Reverse Transcription-PolymeraseChain Reaction (RTPCR)
Quantitative RT-PCR analysis was performed essentially as deTotal RNA was isolated from the cells indicated using
standard protocols.30 Approximately 1 pg of total RNA was denatured (65°C for 5 minutes) before use as a template in a 20 pL
cDNA synthesis reaction. To reduce variations in cDNA synthesis,
all RNA samples were made into cDNA at the same time using a
single master-mix containing (per sample) 1X RT-PCR buffer (50
mmol/L KCl, 20 mmol/L Tris-HC1 [pH 8.4 at 23”C], 2.5 mmol/L
MgC12, 100 pdmL BSA, 2.5 mmol/L dithiothreitol [DTL], 1
mmol/L each dNTP), 17.5 U RNasin ribonuclease inhibitor (Promega, Madison, WI), 100 pmol random d(N), primers (Pharmacia,
Piscataway, NJ), and 8 U AMV reverse transcriptase (Promega).
The reaction was incubated for 10 minutes at 22°C followed by 90
minutes at 42°C. Aliquots of the cDNA reactions were analyzed for
GATA factor expression in 100 pL PCR reactions. Again, to reduce
sample to sample variability all cDNAs to be analyzed on a given
gel were amplified at the Same time using a master-mix containing
(per sample) I X Taq DNA polymerase buffer (Promega), 0.2
mmol/L each dNTP, 25 pmol of each primer (see below), 0.2 pL
[aS2P]-dCTP(3,000 Ci/mmol; ICN, Irvine, CA), and 2.5 U Taq
DNA polymerase (Promega). PCR conditions were 94°C for 2.5
minutes, followed by 21 to 26 cycles (depending on the samples
under analysis)of94”C for 1 minute, 64°C for 1 minute, and 72°C
for I minute. Control reactions were performed to ensure that the
conditions used were within the linear range of PCR amplification
(control lanes of figures, and data not shown). All reactions contained, as an internal control, primers to coamplify the S14 rihosomal protein gene product present in similar abundance to the
GATA factors (and thus having a similar amplification profile).
Samples were analyzed on 6% denaturing polyacrylamide gels
(loaded for equivalent internal control) and exposed to autoradiography before quantitation on a Molecular Dynamics PhosphorImager (Molecular Dynamics, Sunnyvale, CA). The results shown
throughout represent typical data from analysis on at least two separate occasions using at least two independent cDNA preparations.
It should be noted that this methodology permits quantitative comparison of the level of a given GATA factor mRNA in different
samples. However, due to differences in primer annealindextension characteristicsand amplicon length/sequence,etc, it is not possible to compare the absolute abundance of different mRNAs
(GATAs or S14) in these experiments.
The following primers were used for PCR human small ribosomal protein 14 (S 14) sense (5’) GGCAGACCGAGATGAATCCTCA (3’) and antisense (5’) CAGGTCCAGGGGTCTTGGTCC
(Y),nucleotides 187 to 208 and 331 to 3 11, respectively,ofthe S14
sequence’”; hGATA- I sense (5’) GATCCTGCTCTGGTGTCCTCC (3’) and antisense (5’) ACAGTTGAGCAATGGGTACACC
(Y),nucleotides I16 to 136 and 298 to 276, respectively, of the
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
1073
GATA FACTORS IN ERYTHROPOIESIS
Myeloid
hGATA- 1 sequence"; hGATA-2 sense (5') CCCTAAGCAGCGCAGCAAGAC (3') and antisense (5') GATGAGTGGTCGGTTCTGGCC (Y),nucleotides 10 12 to I032 and I I74 to 1 154, respectively, of the hGATA-2 sequence'% hGATA-3 sense (53
GTACAGCTCCGGACTCTTCCC (3') and antisense (5') CTGCTCTCCTGGCTGCAGACA (33, nucleotides 887 to 907 and
1 146 to I 126, respectively, of the hGATA-3 sequence.33
EL Lines
I
,
-
T-cell Lines
RESULTS
Expression of GATA Transcription Factors in Human Cell
Lines and Primary Leukemic Cells
A number of studies have used Northern blot analysis
to determine the tissue distribution of GATA factor
mRNAs.- Recently, such studies have been confirmed and
extended using more sensitive RT-PCR technique^.^^"*^^ To
confirm the validity of this methodology as used here, total
RNA was isolated from a variety of hematopoietic cell lines
and a limited number of primary leukemic cells. hGATA-1,
hGATA-2, and hGATA-3 expression was determined by
RT-PCR. The results (Figs 1 through 3) show clearly distinct patterns of expression for hGATA- I and hGATA-2 on
the one hand and hGATA-3 on the other. Concordant with
previous Northern blot analysis, hGATA- I expression was
detected in all eight of the erythroid cell lines examined and
(at greatly reduced levels) in certain myeloid samples, but
no expression was observed in either B- or T-lymphoid cells.
Like hGATA-I, hGATA-2 is expressed at variable levels in
all of the erythroid lines analyzed, as well as certain myeloid
samples (Fig 2), but is not detected in T or B lymphocytes.
In contrast, hGATA-3 is abundantly expressed only in Tlymphoid cells, although this mRNA is also present at very
much reduced levels in EBV-transformed B cells. (The
hGATA-3 expression detected in the peripheral blood AML
sample [Fig 21 likely reflects the presence of 20% mature
lymphoid cells, as assessed by immunophenotyping.)
GATA mRNA abundance was measured in four of the
erythroleukemia lines before and after induction of differentiation to determine whether the absolute or relative levels of hGATA- 1 and hGATA-2 expression are altered during erythroid differentiation in vitro. All of the lines tested
became hemoglobinized, as judged by visual inspection for
the acquisition of red pigmentation after induction, and the
proportion of cells positive for human globin gene expression by immunofluorescence (Table I). Although many
cells showed morphologic signs of maturation such as decreasing cell size, accumulation of hemoglobin in the cytoplasm, and pycnotic nucleus, significant numbers of undifferentiated blast-like cells remained at the end of induction.
When these cells were analyzed for GATA expression, little
change in the level of hGATA-I mRNA was detected (Fig
3). However, a significant reduction in the level of hGATA2 was observed, particularly in the lines that showed better
hemoglobinization and maturation, such as MB02 (Fig 3).
Similarly, in the two independent primary erythroid (red)
colonies assayed, higher levels of GATA-1 expression coincided with relatively lower levels of GATA-2 (Fig 3). These
observations suggest that terminal erythroid differentiation
is associated with a reduction in the levels of the transcrip
CATA-1
Myeloid
-I)
EL Lines
-
T-cell Lines
Fig 1. Expression of GATA factors in two myeloid (HL60 and
EM-2). seven erythroleukemia lines (KU812, JK-1 , MB02, HEL,
OClMl, OCIM2, and CMK), and several other lines ofnone*roid
origin including B cells (EBV-transformed lymphocytes, WM9, and
U266). T cells (CEM. HSB2, HUT78, and MOLT-3). and HELAcells.
Note that all erythroleukemia lines express GATA-1 and GATA 2
(with one exception), but show virtually no GATA-3 expression. In
contrast, T-cell lines abundantly express GATA-3 and no other
GATA factors.
tion factor hGATA-2. Although the various leukemic myeloid cells tested express little or no hGATA-I, this mRNA
is, surprisingly, essentially as abundant in primary GM + M
myeloid colonies as in primary BFUederived colonies. This
may reflect an active expression of hGATA-1 in proliferating myeloid precursors, before its inactivation as maturation progresses, as previously reported," or may be due to
development of cells of eosinophilic lineage in these colo-
GATA Factor Expression in Enriched Primary Human
Progenitor Cells
The observation that more than one hGATA factor is
expressed in a number of transformed erythroid cell lines
led us to ask whether these GATA factors are similarly expressed in primary erythroid cells and, if so, whether there
are hierarchical differences in the expression of these factors
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
LEONARD ET AL
1074
A
hCATA-34
hCATA-3
hCATA-I +
hCATA-1
hS14
hS14-D
B
hCATA-2
B
hGATA-24
Fig 3. GATA factor expressionin colonies of BFUe and CFU-GM
origin (“plucked“ from methylcellulose clonal cultures) and in four
uninduced (unind) and induced (ind) erythroleukemic cell lines.
GATA-1 and GATA-2 are detected in all colonies except the most
mature erythroid ones, which show mainly GATA-1. Also note that
induction of erythroleukemic cells is accompanied by downregulation of GATA-2, especially in MB02 cells.
S I 4 4
Fig 2. GATA factor expression in primary leukemic cells. (Two
AML samples, one from peripheral blood, the other from mamw,
and two CML samples, one in accelerated phase and the other in
chronic phase, both from peripheral blood.) GATA-1 and GATA-2
are detected in CML samples; GATA-2 but not GATA-1 is detected
in AML samples and in HL60 cells.
during erythroid lineage differentiation. PBMCs are composed mostly of T- and B-lymphoid cells with a variable
proportion of monocytes (and often a small proportion of
contaminating granulocytes). RNA from isolated total
PBMCs was analyzed for hGATA factor expression. As expected, abundant hGATA-3 expression was observed, but
little or no hGATA-I or hGATA-2 could be detected (Fig
4). The PBMC samples were enriched for progenitors by
immunoadherence to anti-CD34 or to anti-c-kit antibody
(SR-l)-coated plates.” Through this process, greater than
100-fold enrichment of hematopoietic progenitors was
achieved (up to 30% of cells in the enriched fraction are
progenitor cells, compared with less than 0. I% in the starting population). Progenitor-enriched cells consistently
showed significantly reduced hGATA-3 mRNA levels and
increased hGATA- 1 and hGATA-2 (Fig 4). Moreover,
hGATA- 1 and hGATA-2 appear to be enriched with similar
profiles during the purification process (Fig 4).
Table 1. Proportion of Globin (7.8. {, or 6) and Benzidine-Positive
Cells in Four Erythroleukemic Lines Before ( ) and
After Their Induction ( )
-
+
Benr (+)
Globin (+) Cells (%)
Cells
r
c
(%I
----+
+
+ - + - +
B
I
EL
CellLine
OCIM2’
JK-1’
MB02’
F-36’
15
20
9
18
55
30
100
72
<1
5
3
<1
10
17
36
1
16
25
0
3
64
45
0
5
Pellets from all induced cells had red color.
0
5
0
0
8
9
10
0
<5
<1
49
97
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
1075
GATA FACTORS IN ERYTHROPOIESIS
C
?
oo 0
zzi =
m
a a .a. .
nn
Cultureday
GATA-3
GATA-1
*-
-
0 0 12
00
0 0
8
I
m
a
a b c
0 0 0 6 6 6
-
*-
data from erythroleukemic cell lines, samples of enriched
progenitor cells were cultured in suspension either in the
presence of early acting cytokines IL-3, KL, or PIXY-321,
or in the presence of Epo. Under the influence of IL-3 and
KL there is significant cell proliferation and progenitor amplification during the first week in culture.” Concurrently,
cells with lineage-specific differentiative features begin to
emerge. An interesting and unexpected feature was the fact
that cells with both myeloid (ie, granule formation) and erythroid features were detected in these cultures. The presence
of cells at later stages of erythroid differentiation (CFUe/
proerythroblasts) was suspected on the basis of morphologic
-+
*
--7
*
J
Y
Y
I
s’4
ma
@-om
Dm
+
-I
+
is
7
X
H
a
Culture day
GATA-2
0
0
-
0 0 047 0 4 7
0 4 7
0 4 7
0
GATA-3
S14Fig 4. GATA factor expression in PBMCs before (PBMC-Un)
and after (PBMC-P)their enrichment for progenitor cells. The three
sets of samples on the right are from three separate experiments.
Note that both GATA-1 and GATA-2 become readily detectable
after purification, whereas expression of GATA-3 is greatly diminished. In addition to the three independent samples of PBMC-P at
day 0, one sample was also tested after 6 days (far right, a, b, and
c), and another after 1 2 days in culture (left set of samples, PBMCP). Samples day 6a and PBMC-P day 12 contained Epo in culture,
whereas sample day 6b contained 11-3
KL in serum and sample
day 6c contained the same cytokines in serum-free medium. Note
that the Epo-containingcultures (PBMC-Pday 12, and day 6a) had
mainly GATA-1 expression with very little GATA-2 or GATA-3.
Likewise, preparationsfrom fetal liver (FL, -80% erythroid cells) or
purified erythroblasts (FL-E)from fetal liver (by panning to antibody
SFL23.6-coated plates) expressed primarily GATA-1.
GATA-1
+
GATA-2
S14
In addition to the peripheral blood samples, progenitors
were also enriched from fetal liver (Fig 5). Progenitorenriched samples (FL-PI and FL-P2, Fig 5) were free of
contaminating erythroblasts and the predominant progenitor cell was the BFUe.” Interestingly, in these samples,
hGATA- I mRNA was abundant, whereas very low levels of
hGATA-3 and GATA-2 expression were detected (Fig 5).
hGATA-I and hGATA-2 Are Abundantly Expressed in
ProliJerating Progenitors, But Their Expression Diverges
in Terminally Differentiated Erythroid Cells
To further analyze the pattern of expression of hGATA
factors during erythroid maturation, and to test whether
terminal erythroid differentiation of normal cells is associated with reduced GATA-2 expression, as suggested by
Fig 5. GATA factor expressionin progenitor-enrichedcells from
two fetal livers (FL-P1 and FL-P2) and from one peripheral blood
sample at days 0,4, and 7 (four sets of samples, right) after culture
in the presence of growth factors, either in serum-containing (S )
or serum-free medium (S ). Purified fetal liver samples (FL-P1 and
FL-PP)had a blast-like morphologywith no differentiated cells present. The proportion of BFUe in day 0 samples of peripheral blood
was about 25% with an absence of globin-positive cells. Note the
upregulation of GATA-1 and GATA-2 expressionin samples during
the first week in culture. In Epo-containing cultures from the PB
sample and in enriched populations of fetal liver that contain mainly
erythroid progenitor cells, there is primarily GATA-1 expression,
whereas without Epo both GATA-1 and GATA-2 are expressed. At
day 7, Epo-containingcultures had 54%6-globin-positive cells and
12%erythroid progenitors, IL-3 KL (S ) cultures had 14%6-globin-positive cells and 13%progenitors, and PIXY-321 cultures had
4%B-globin-positive cells and 12%progenitors. Furthermore, cultures without Epo had, in addition, differentiating myeloid cells.
+
-
+
+
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
LEONARD ET AL
1076
in IL-3 and KL cultures without any evidence of further
maturation, whereas globin-positive cells in Epo-containing
cultures increase to more than 90% by day 12 or 13.
The levels of hGATA- 1 and GATA-2 were similar at day
0 in enriched populations of progenitors and they were
found to increase further in culture (Figs 4 and 5). In experiments in which cells were studied under the stimulation of
only Epo, GATA-1 expression remained high during the
observation period ( 1 to 12 days), whereas GATA-2 expression declined (see Fig 5 , Epo, days 4 and 7; and Fig 4,
PBMC-P, day 12). Thus, abundant hGATA-2 expression
seems to be associated with cell and progenitor proliferation
(with IL-3 and KL), but terminal erythroid differentiation
(in the presence of Epo) is associated with a drastic reduction in the levels of this transcription factor.
DISCUSSION
GATA-I
Fig 6. Morphologic appearance of erythroid cells presentin Epoonly cultures at 7 days. Note the characteristic proerythroblast-like
or CFUe-like appearance of these blasts.
criteria (Fig 6) and confirmed by slot-blot hybridization assays for globin RNA (Fig 7) and by antiglobin immunofluorescence (data not shown). These results were obtained not
only in serumcontaining cultures but in serum-free cultures and under conditions in which anti-Epo antibody was
present." Furthermore, no endogenous Epo mRNA was
detected in these cultures by RT-PCR (data not shown).
However, in the absence of Epo, erythroid cells are significantly diminished at later times in culture (12 to 14 days)
and the population is virtually all granulomonocytic in nature. On the other hand, if Epo is present as a single factor in
these cultures, a healthy basophilic blast population (resembling proerythroblasts, Fig 6) is present in significant numbers by the first week in culture, and 50% to 70% of these
contain &globin. At later times (days 12 to 13), there is
increased b-globin mRNA expression (virtually all the cells
present are P-globin-positive) and accumulation of smaller,
more mature, strongly benzidine-staining cells. Thus, during the second week in culture there is a major difference
between the Epocontaining cultures and those with only
IL-3 and KL, the frequency of globin-positive cells declines
Enrichment of hematopoietic progenitors from peripheral blood selects for cells expressing GATA-1. Committed
progenitors for all lineages are concurrently purified in these
samples, raising the possibility that GATA-I is expressed in
multipotential hematopoietic lineage progenitors. Consistent with this, GATA-I expression is detected in a multipotential bone marrow-derived cell line." However, whether
heterogeneity exists in expression of GATA- 1 among different types of progenitorscannot be determined from the present data.
When progenitors are cultured in suspension in the presence of various cytokines there is a significant upregulation
of GATA-1 expression, in agreement with a previous report." In addition to cell proliferation and progenitor
(BFUe and CFU-GM) amplification, there is a progressive
decrease in CD34+ cells and a concomitant significant in-
P
d.2
1 m
1
IL-3+KL
1
1
d.6
-
d.13
d.6 0
d.6
d.13
d.13
d.13
d.2
Y
.-
Y
d.2
Fig 7. Globin mRNA levels wand y) in one PBMC-P sample cultured with either Epo or with IL-3
KL. Note progressive globin
mRNA accumulation in Epo-containing cultures (days 1 through
13) whereas in 11-3
KL cultures, globin mRNA levels peak at day
6 to 7.
+
+
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
GATA FACTORS IN ERYTHROPOIESIS
crease in cells that have acquired lineage-specificdifferentiative features. For example, in cultures without Epo, apart
from granulomonocytic differentiation, we have noted the
generation of late erythroid progenitors (CFUe) and of globin-positive cells (in greater numbers than CFUe or BFUe)
within the first week in cu1t~re.I~
Thus, in contrast to previous r e p ~ r t s , ~ 'there
, ' ~ is not exclusively myeloid differentiation in the absence of Epo in these cultures. Our findings
raise the possibility that the upregulation of GATA-1 may
not be due just to the proliferation of CD34+progeny," but
rather associated mainly with cell progeny displaying globin
transcription. Thus, low levels of GATA-1 may be found in
all types of progenitors, especially proliferating ones, but
upregulation occurs only in erythroid and not in nonerythroid progenitors. Although further studies with separated progenitors are needed to confirm these points, indirect lines of evidence support the view that myeloid
leukemia cell lines do not express significant levels of
GATA- 1;this factor has been found only in a small number
of AML, especially in those of the undifferentiated type36or
in those from CML patients (Fig 2), in categories in which a
mixed-lineage leukemia including erythroid cells is very
likely.
Despite the initial upregulation of GATA-1 and the expression of globin, globin-positive cells generated in the absence of Epo are unable to mature further without this cytokine and die premat~re1y.I~These cells are greatly
diminished after 2 weeks in culture and GATA- 1 expression
is also greatly reduced14 (data not shown). In contrast, in
cultures containing Epo (Fig 1, PBMC-P, day 12), high levels of GATA-1 are maintained during the second week. One
interpretation of these results is to implicate GATA-1 in
both the initiation and maintenance of the erythroid program. However, these two events appear to be under separate regulatory control. Initiation is an Epo-independent
event, whereas, for maintenance, additional events, triggered by Epo-Rfligand-dependent interactions, are needed
and these are critical for the survival and further maturation
of these erythroid cells. It is possible that, although GATA-1
may upregulate Epo-R, only subthreshold levels of Epo-R
may be reached in the absence of Epo-Rlligand interact i o n ~ It. ~is~also possible that, in addition to Epo, other
downstream-activated genes (such as SCL) are needed to
sustain high levels of GATA- 1 and maintain the erythroid
program.
GATA-2
Detailed studies of GATA-2 expression during development in frogs and chicken^^^^^^^ have shown that GATA-1
and GATA-2 are expressed concurrently in development
from the very onset of hematopoiesis. Similar studies of
GATA-2 expression in developing human hematopoietic
cells are not available. In the present studies, we show that
progenitor enrichment leads to selection of cells expressing
GATA-2 in addition to GATA- 1. Likewise, culture of progenitors leads to upregulation of GATA-2 in addition to
GATA-1, indicating that GATA- 1 and GATA-2 are modulated in tandem at early stages of hematopoietic differentia-
1077
tion. It remains to be seen whether GATA-2 precedes
GATA-1 expression in the hematopoietic hierarchy (ie, is
present in preprogenitors with the CD34+/CD38- phenotype), as recently suggested," or whether there is heterogeneity in the levels of GATA-2 expression in committed progenitors. The profile of GATA-1 and GATA-2 expression
during progenitor enrichment suggests that it is very likely
that the same target cells express both of these factors. The
presence of both GATA-1 and GATA-2 in all erythroleukemia lines tested would tend to support this conclusion (although erythroleukemiacells have more than one potentiality). Preparations of fetal liver that are highly enriched for
erythroid progenitors express GATA- 1 at much higher levels than GATA-2 (Fig 5). Furthermore, upregulation of
GATA-2 is reduced or minimal in Epo-only containing cultures (in which exclusively erythroid differentiation is observed, Figs 4 and 5) compared with non-Epo-containing
cultures in which a more mixed population is present. Thus,
it would seem that, at a certain stage in erythroid differentiation, the expression of GATA- l and GATA-2 diverge; earlier stages display GATA-2 expression,but with more differentiation and maturation the GATA-2 expression is
dramatically reduced. The fact that GATA-2 is diminished
as erythroid cells differentiate is supported by several additional pieces of evidence; small bursts (from methylcellulose
cultures) with mostly differentiated mature erythroid precursors display much less GATA-2 expression than GATA1 (Fig 3). Furthermore, upon differentiation, erythroleukemia lines display a downregulation of GATA-2 (eg, MB02,
Fig 3) proportional to the degree of terminal maturation
achieved; the more advanced the differentiation, the more
conspicuous the GATA-2 downregulation. Furthermore,
modest overexpression of GATA-2 in committed chick erythroid progenitors blocks terminal differentiati~n.~'
GATA-3
GATA-3 expression in human hematopoietic cells does
not appear to follow the expression pattern seen in chickens
and frogs.2s3From our studies in primary hematopoietic
cells and cell lines, it is clear that the GATA-3 expression is
associated only with the T-cell lineage. In contrast to findings in
there is absence of GATA-3 expression
in the terminal phase of human erythroid differentiation.
In summary, our data suggest that differentiallyregulated
expression of the GATA family of transcription factors is
associated with lineage selection during hematopoietic differentiation. We propose that initiation of the terminal erythroid program of gene expression and globin transcription
is accompanied by an upregulation of GATA-1 and GATA2, and does not require the presence of Epo. However, for
maintenance of the erythroid program, additional regulatory events are required. For example, the second phase of
GATA- 1 upregulation may be dependent on Epo-Rfligand
interactions or other downstream erythroid-specificregulatory genes; the downregulation of other factors such as
GATA-2, or c-myb, may possibly also be critical. GATA-3
appears to be confined to lymphoid populations and not to
be involved in human erythromyeloid cytodifferentiation.
Nevertheless, a host of basic questions relevant to regulation
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
LEONARD ET AL
1078
of early stages of hematopoietic differentiation by GATA
transcription factors remain currently unanswered. At what
stage of differentiation is the expression of GATA-2 abrogated in erythroid versus myeloid cells? Are there quantitative differencesin the expression of GATA- 1 between megakaryocytic and erythroid cells at equivalent stages of
differentiation? What are the necessary interactions, positive or negative, for sustaining and further upregulating
GATA-1 in erythroid cells? As more sophisticated approaches become available for separation of progenitor subsets and testing them for transcription factor expression,
several of these questions may be resolved.
ACKNOWLEDGMENT
We are grateful to Dr T.H. Price and the Puget Sound Blood
Center for providing peripheral blood mononuclear cells, and to Dr
A.G. Fantel (Central Laboratory of Human Embryology, University of Washington) for fetal liver samples. We are also indebted to
the following for their generous gifts of the monoclonal antibodies
and cytokines used in this work V.C. Broudy for the monoclonal
antibody to c-kit (SR-1) and I.D. Bemstein for the anti-CD34 antibody 12.8; Genetics Institute for IL-3 and Epo; Amgen for KL
(SCF); and Immunex for PIXY-32 1. Our special thanks to B. Nakamoto for preparation of the many cell lines used here, to D. Farrer
for technical assistance, and to B. Lenk for expert secretarial assistance.
REFERENCES
1. Orkin SH: GATA-binding transcription factors in hematopoi-
etic cells. Blood 80575, 1992
2. Engel JD, George KM, KO U,Kornhauser JM, Leonard
MW, Ting P, Yamamoto M: Transcription factor regulation of hematopoietic lineage cells. Semin Hematol28:158, 1991
3. Zon LI, Tsai SF, Burgess S, Matsudaira P, Bruns GAP, Orkin
SH: Expression ofGATA-binding proteins during embryonic development in Xenopus laevis. Proc Natl Acad Sci USA 88:10642,
1991
4. Tsai SF, Martin DI, Zon LI, D’Andrea AD, Wong GG, Orkin
SH: Cloning of cDNA for the major DNA-binding protein of the
erythroid lineage through expression in mammalian cells. Nature
339:446, 1989
5. Evans T, Felsenfeld G: The erythroid-specific transcription
factor Eryfl: A new finger protein. Cell 58:877, 1989
6. Trainor CD, Evans T, Felsenfeld G, Boguski MS: Structure
and evolution of a human erythroid transcription factor. Nature
343:92, 1990
7. Pevny L, Simon MC, Robertson E, Lekin WH, Tsai SF,
DAgati V, Orkin SH, Costantini F Erythroid differentiation in
chimaeric mice blocked by a targeted mutation in the gene for transcription factor GATA-I. Nature 349:256, 1991
8. Yamamoto M, KO U,Leonard MW, Baug H, Orkin SH,
Engel JD: Activity and tissue-specific expression of the transcription factor NF-E1 multigene family. Genes Dev 4:1650, 1990
9. Simon MC, Pevny L, Wiles MV, Keller G, Costantini F, Orkin SH: Rescue of erythroid development in gene targeted GATAI - mouse embryonic stem cells. Nature Genet 1:92, 1992
IO. Crotta S, Nicolis S, Ronchi A, Ottolenghi S, Ruzzi L, Shimada Y, Migliaccio AR, Migliaccio G: Progressive inactivation of
the expression of an erythroid transcriptional factor in GM- and
G-CSF-dependent myeloid cell lines. Nucleic Acids Res 18:6864,
I990
1 1. Sposi NM, Zon L1, Care A, Valtieri M, Testa U, Gabbianelli
M, Mariani G, Bottero L, Mather C, Orkin SH, and Peschle C: Cell
cycle-dependent initiation and lineage-dependent abrogation of
GATA- 1 expression in pure differentiating hematopoietic progenitors. Proc Natl Acad Sci USA 89:6353, 1992
12. Mouthon M-A, Bernard 0, Mitajavila M-T, Romeo P-H,
Vainchenker W, Mathieu-Mahul D: Expression of tat-1 and
GATA-binding proteins during human hematopoiesis. Blood
81:647, 1993
13. Martin DI, Zon LI, Mutter G, Orkin SH: Expression of an
erythroid transcription factor in megakaryocytic and mast cell lineages. Nature 344:444, 1990
14. Romeo P-H, Prandini MH, Joulin V, Mibnotte V, Pemant
M, Vainchenker W, Marguerie G, Uzan G: Megakaryocytic and
erythrocytic lineages share specific transcription factors. Nature
344:447, 1990
15. Ito E, Toki T, Ishihara H, Ohtani H, Gu L, Yokoyama M,
Engel JD, Yamamoto M: Erythroid transcription factor GATA-1 is
abundantly transcribed in mouse testis. Nature 362:466, 1993
16. Broudy VC, Lin NL, Zsebo KM, Papayannopoulou Th: Isolation and characterization of a monoclonal antibody (SR-I) that
recognizes the stem cell factor receptor (c-kit). Clin Res 39:210A,
1991
17. Papayannopoulou Th, Brice M, Blau T Kit ligand in synergy
with IL-3 amplifies the Epo independent, globin synthesizing progeny of normal human BFUe in suspension cultures: Physiologic
implications. Blood 8 1 :299, 1993
18. Fukuda T, Kishi K, Ohnishi Y, Shibata A: Bipotential cell
differentiation of KU-8 12: Evidence of a hybrid cell line that differentiates into basophils and macrophage-like cells. Blood 70:6 12,
1987
19. Hitomi K, Fujita K, Sasaki R, Chiba H, Okuno Y, Ichiba S,
Takahashi T, Imura H: Erythropoietin receptor of a human leukemic cell line with erythroid characteristics. Biochem Biophys Res
Commun 143:902, 1988
20. Morgan DA, Gumucio DL, Brodsky I: Granulocyte-macrophage colony-stimulating factor-dependent growth and erythropoietin-induced differentiation of a human cell line MB-02. Blood
78:2860, 1991
21. Martin P, Papayannopoulou Th: HEL cells: A new human
erythroleukemia cell line with spontaneous and induced globin expression. Science 216:1233, 1982
22. Papayannopoulou Th, Nakamoto B, Kurachi S, Tweeddale
M, Messner H: Surface antigenic profile and globin phenotype of
two new human erythroleukemia lines: Characterization and interpretations. Blood 72: 1029, 1988
23. Sat0 T, Fuse A, Eguchi M, Hayashi Y, Sugita K, Nakazawa
S, Minato K, Shima Y, Komori I, Sunami S, Okimoto Y, Nakajima
H: Establishment and characterization of a megakaryoblastic cell
line (CMK) from a Down’s syndrome patient with acute megakaryoblastic leukemia. Exp Hematol 15:495, 1987
24. Chiba S, Takaku F, Tange T, Shibya K, Misawa C, Saski K,
Miyagawa K, Yazaki Y, Hirai H: Establishment and erythroid differentiation of a cytokine-dependent human leukemic cell line
F-36: A parental line requiring granulocyte-macrophage colonystimulating factor or interleukin-3, and a subline requiring erythropoietin. Blood 78:2261, 1991
25. Leonard MW, Lim K-C, Engel JD: Expression of GATA
transcription factor family during early development and differentiation. Development (submitted)
26. Stamatoyannopoulos G, Farquhar M, Lindsley D, Brice M,
Papayannopoulou Th, Nute PE: Monoclonal antibodies specific
for globin chains. Blood 61:530, 1983
27. Constantoulakis P, Nakamoto B, Papayannopoulou Th, Stamatoyannopoulos G Fetal calf serum contains activities which in-
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
GATA FACTORS IN ERYTHROPOIESIS
duce fetal hemoglobin in adult erythroid cell cultures. Blood
75:1862, 1990
28. Camp TA, Rahal JO, Mayo K E Cellular localization and
hormonal regulation of follicle-stimulatinghormone receptor messenger RNAs in the rat ovary. Mol Endocrinol5:1405, 1991
29. Foley KP, Engel JD. Individual stage selector element mutations lead to reciprocal changes in 8- vs €-globingene transcription:
Genetic confirmation of promoter competition during globin gene
switching. Genes Dev 6:730, 1992
30. Chromczynski P, Sacchi N Single-stepmethod of RNA isolation by acid guanidinium thiocyanate-phenolchloroform extraction. Anal Biochem 162:156, 1987
3 1. Zon LI, Tsai S-F, Burgess S, Matsudaira P, Bruns GAP, Orkin SH: The major human erythroid DNA-binding protein (GF-1):
Primary sequence and localization of the gene to the X chromosome. Proc Natl Acad Sci USA 87:668, 1990
32. Dorfman DM, Wilson DB, Bruns GA, Orkin SH: Human
transcription factor GATA-2. Evidence for regulation of preproendothelin-1 gene expression in endothelial cells. J Biol Chem
267:1279, 1992
33. KO U,Yamamoto M, Leonard MW, George KM, Ting P,
Engel JD: Murine and human T-lymphocyteGATA-3factors medi-
1079
ate transcription through a cis-regulatory element within the human T-cell receptor delta gene enhancer. Mol Cell Biol 11:2778,
1991
34. Hromas R, Zon L, Friedman AD: Hematopoietic transcription regulators and the origins of leukemia. Clin Rev Oncol/Hemato1 12:167, 1992
35. Yamaguchi Y, Albee EA, Zon LI, Tenen DG, Ackerman SJ:
GATA-1 transcription factor mRNA is expressed in association
with granule protein genes during IL-5 induced eosinophilopoiesis.
Blood 80:248a, 1992 (abstr, suppl 1)
36. It0 E, Toki T, Arai K, Kawauchi K, Tsuda H, Yokoyama M:
Expression of a lineage specific transcriptional factor GATA-1 in
leukaemic blasts from patients with infantile leukaemia. Br J Hematol 80561, 1992
37. Chiba T, Ikawa Y, Todokoro K GATA-1 transactivates
erythropoietin receptor gene, and erythropoietin receptor-mediated signals enhance GATA-1 gene expression. Nucleic Acids Res
19:3843, 1991
38. Breigel K, Liu K-C, Planck C, Beug H, Engel JD, Zenke M:
Ectopic expression of a conditional GATA-Z/estrogen receptor chimera arrests erythroblast differentiation in a hormone-dependent
manner. Genes Dev 7:1097, 1993
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
1993 82: 1071-1079
Dynamics of GATA transcription factor expression during erythroid
differentiation
M Leonard, M Brice, JD Engel and T Papayannopoulou
Updated information and services can be found at:
http://www.bloodjournal.org/content/82/4/1071.full.html
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:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
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.