(Received for publication, May 16, 1997)
From the Departments of Structural and Cellular
Biology and § Pediatrics, University of South Alabama,
Mobile, Alabama 36688
There are numerous similarities between the
erythroid and megakaryocytic lineages which suggest that commitment to
either lineage occurs relatively late in hematopoiesis. Commitment
toward megakaryocyte development requires obligatory silencing of
erythroid-specific genes. Therefore, we investigated the effects of
interleukin-6, a known inducer of thrombocyte production, on globin
gene expression during erythroid differentiation. Studies in K562 cells
demonstrated inhibition of globin gene mRNA production and
chain biosynthesis in the presence of exogenous interleukin-6 which was
abrogated by anti-interleukin-6 monoclonal antibody. Similar studies in primary erythroid progenitors showed inhibition of burst-forming unit-erythroid colony formation when interleukin-6 was added late in
cultures with decreased
and
globin gene mRNA production. Protein binding studies demonstrated an increase in activator protein-1
binding to its consensus sequence by 24 h of interleukin-6 treatment. Inhibition of activator protein-1 gene activity had no
effect on
gene silencing by interleukin-6. A potential
interleukin-6 response element was identified in the
globin gene.
Interleukin-6 treatment led to a rapid increase in protein binding
to the target DNA sequence. These results suggest that interleukin-6
may play an important role in globin gene silencing during
megakaryocytic lineage commitment.
The hematopoietic pathways for megakaryocytic
(MK)1 and erythroid (E)
differentiation are closely related. These lineages share common
hematopoietic-specific trans-acting factors including GATA-1 and nuclear factor-erythroid 2 (NF-E2) (1). GATA-1 is absolutely required for differentiation of the E lineage (2); and although it is
not required for MK differentiation, cell lines with the highest
concentrations of GATA-1 differentiate into megakaryocytes (3). NF-E2
is required for both high level enhancer activity of the globin
locus control region (LCR) (4) and production of platelets (3). The
regulation of E- and MK-specific genes shares many features (5, 6),
suggesting that restriction toward either pathway occurs relatively
late in hematopoietic hierarchy. Evidence for a common E/MK progenitor
is demonstrated by expression of both E and MK markers in
erythroleukemic cell lines (1, 7). Recently, a bipotent burst forming
unit (BFU)-E/MK progenitor has been identified in human bone marrow
(7). The factors that determine whether commitment to the E or MK
lineage will occur are largely unknown. Optimal development of BFU-E/MK requires a combination of several cytokines including erythropoietin, stem cell factor, interleukin (IL)-3, and megakaryocyte growth and
development factor, which are proposed to act sequentially on
progenitors at different stages of maturation (7). Therefore, at some
stage during maturation of the putative bipotential BFU-E/MK progenitor
cell an additional level of regulation responsible for the proper
silencing of globin genes (which are not normally expressed in
megakaryocytes) must occur.
In vivo studies in mice and in vitro studies in
low density bone marrow cells demonstrated stimulation of
megakaryocytopoiesis and platelet production by IL-6 (8-12). IL-6
works synergistically with IL-3 to stimulate early megakaryocyte
development and proliferation of hematopoietic progenitors (8-10, 13).
In this study we investigated the possible role of IL-6 in globin gene
silencing during megakaryocytic lineage commitment. Studies were
conducted using K562 cells, a human erythroleukemic cell line that
expresses both erythroid and megakaryocytic-specific genes in the
uninduced state and has the capacity to differentiate along the E or MK
lineage (14, 15). Levels of globin mRNA and protein
biosynthesis in K562 cells were analyzed after treatment with IL-6. We
observed an IL-6 concentration-dependent decrease in both
globin mRNA production and protein biosynthesis. This
inhibition was abrogated by the addition of anti-IL-6 monoclonal
antibody. K562 cells treated with IL-6 responded with increased
glycoprotein IIb (GpIIb) mRNA levels. Similar studies were
performed using mononuclear cells isolated from human peripheral blood.
IL-6 inhibited BFU-E colony formation when added late in cultures and
decreased
and
globin mRNA production. Gel mobility shift
assays (GMSAs) showed increased protein binding to a potential IL-6
response element (IL-6RE) after IL-6 treatment. These studies
collectively suggest that IL-6 serves as a negative regulator of globin
gene expression during megakaryocyte lineage commitment.
K562 cells were maintained in suspension cultures in Iscove's modified Dulbecco's medium (IMDM) containing 10% fetal bovine serum, penicillin (100 units/ml), and streptomycin (0.1 mg/ml). Cells were grown in IMDM without fetal bovine serum for 24 h prior to beginning analyses. Experiments were conducted with 6.6 × 105 cells/ml in IMDM containing hemin (100 µM), sodium butyrate (2 mM), or exogenous IL-6 (10, 25, 50, 100 ng/ml) and collected at 0-, 24-, 48-, and 72-h increments for analysis by RT-PCR and RNase protection, described below. Antibody studies were performed by incubating anti-IL-6 monoclonal antibody (5 or 10 µg/ml) with IL-6 protein (50 ng/ml) for 1 h at 37 °C prior to addition to K562 cells for 72 h. Curcumin experiments were performed by incubating K562 cells with 20 µM curcumin for 12 h prior to the addition of IL-6.
Linear Range Determination for RT-PCRK562 cells were
collected by centrifugation, and the RNA was isolated using RNA Stat-60
(TEL-TEST "B", Inc., Friendswood, TX) according to the
manufacturer's instructions. To determine the relative level of
mRNA production in K562 cells, RT-PCR was performed with
experimental primers specific for the A, IL-6, IL-6 receptor
(IL-6R), GATA-1, and GpIIb genes (see Table I). The constitutively
expressed glyceraldehyde-3-phosphate dehydrogenase (GAPD) gene was
utilized as an internal control. Reverse transcription was performed
with different concentrations of total cellular RNA ranging from 0.1 to
500 ng for each primer pair analyzed. The cDNA product was
amplified by PCR using primers for GAPD, A
, IL-6R, IL-6, GATA-1, and
GpIIb. Optimal annealing temperatures for experimental primer sets were
determined to be 60 °C for A
, 62 °C for IL-6R and GpIIb, and
65 °C for IL-6 and GATA-1. The linear ranges of starting RNA
concentrations which resulted in a directly proportional increase in
PCR product were different for each set of primers, therefore
additional experiments were performed in which the number of PCR cycles
was varied to ensure linearity with the starting RNA concentration of
50 ng/ml. The number of PCR cycles required for each primer set to
produce product in the linear range was determined to be 23 cycles for
GAPD; 35 cycles for IL-6, GATA-1 and GpIIb; and 38 cycles for A
and
IL-6R. Subsequent reverse transcription reactions were performed using 50 ng of total cytoplasmic RNA for first strand synthesis.
|
cDNA strands were synthesized using 50 ng of
total cellular RNA, 1 × PCR buffer (10 mM Tris-HCl,
pH 8.3, 50 mM KCl), MgCl2 (3.0 mM),
RNase inhibitor (33 units/µl, Promega, Madison, WI), random hexamer
primers (2.5 µM, Perkin-Elmer), and Moloney murine leukemia virus-reverse transcriptase (50 units, Promega) in a 20-µl
total reaction volume. Each cDNA product was split into two tubes
(25 ng); PCR reagents and primers specific for GAPD were added to the
first tube; and PCR reagents and primers specific for A, IL-6,
IL-6R, GATA-1, or GpIIb were added to the second tube at a total volume
of 50 µl for each reaction. The respective cDNAs were amplified
at 95 °C for 30 s, 60 °C (A
), 62 °C (IL-6R, GpIIb), or
65 °C (IL-6, GATA-1) for 1 min, and 72 °C for 2 min at 23 cycles
for GAPD, 35 cycles for IL-6 and GpIIb, and 38 cycles for A
, GATA-1,
and IL-6R. 10 µl of each PCR product was run on a 1% agarose gel,
stained with 1 µg/ml ethidium bromide, and photographed. The
fluorescence intensity was quantitated using scanning laser densitometry and analyzed using AAB Biomed Instrument Densitometric Analysis Software. The relative change in mRNA product for the genes analyzed was calculated as the ratio of fluorescence intensity of
the PCR product obtained with the experimental primers to that of
GAPD.
K562 cells were grown in IMDM and 1 × 106 cells were treated with IL-6 at a concentration of 100 ng/ml for 72 h. At the end of the treatment period cells were harvested, washed in phosphate-buffered saline, and pelleted. Cellular lysates were prepared by lysing the K562 cells in 200 µl of water, and protein concentrations were quantitated using the Bradford assay (16). Total cellular protein (100 µg) was lyophilized and then resuspended in 500 µl of 0.1% trifluoroacetic acid. Hemoglobin chains were separated by a modification of the previously described protocol of Shelton et al. (17). Briefly, sample was loaded onto a C4 column (Vydac Separations Group, Hesperia, CA) and eluted with a Beckman System Gold apparatus. The gradient was composed of 36% acetonitrile in 0.1% trifluoroacetic acid for buffer A, and 46% acetonitrile in 0.1% trifluoroacetic acid for buffer B. Initial conditions were 23% A and 77% B. The gradient was set to 100% B in 50 min at a flow rate of 1 ml/min and ultraviolet absorbance at 220 nM.
RNase Protection AnalysisRNA was isolated from either K562
cells or BFU-E colonies on day 14 as described previously above. Globin
mRNA was analyzed by RNase protection with the following probes:
pT7A (170) linearized with BstEII
to give a 170-bp protected fragment and pT7H
linearized
with HindIII to give a 205-bp protected fragment. The GAPD
probe (Ambion, Austin, TX) yields a 316-bp fragment and was used as an
internal control. RNA (1 µg) was hybridized overnight at 45 °C
with 106 cpm of each radiolabeled probe. After digestion
with RNase A, the protected fragments were separated on 6%
polyacrylamide, 8 M urea gels and autoradiographed without
intensifying screens. Human A
,
, and GAPD mRNAs were
quantitated using a GS-250 Molecular Imager (Bio-Rad). Human
and
globin gene expression was calculated as the percent
or
mRNA as a ratio to GAPD mRNA production.
Mononuclear cells were isolated from
the peripheral blood of each patient studied by density gradient
centrifugation using Histopaque-1077 (Sigma). The cells were washed
with IMDM containing 30% fetal bovine serum, 2 mM
glutamine, and 104 M
-mercaptoethanol,
then cells were mixed at a concentration of 3 × 105
cells/ml in MethoCult GF H4434 medium (StemCell Technologies Inc.,
Vancouver, B. C.). The mononuclear cells were incubated in multiwell
tissue culture plates in a humidified atmosphere containing 5%
CO2, at 37 °C. The following experimental conditions were analyzed in triplicate wells in the culture plates: an untreated control, IL-6 added on day 0, and IL-6 added on day 7 of the study period (IL-6 concentrations were 5, 25, or 50 ng/ml). The concentration with the greatest inhibitory effect on colony growth was determined to
be 50 ng/ml IL-6, therefore subsequent experiments were performed at
this concentration. BFU-E colonies were counted on day 14 using an
inverted microscope and harvested for RNA analysis by an RNase protection assay as described above.
Nuclear protein extraction was performed as described
previously by Andrews and Faller (18). In brief, approximately 5 × 106 K562 cells were harvested and resuspended in buffer
A (10 mM HEPES-KOH, pH 7.9, 1.5 mM
MgCl2, 10 mM KCl, 0.5 mM
dithiothreitol, 0.2 mM phenylmethylsulfonyl fluoride). The
suspension was placed on ice for 10 min, vortexed, pelleted, and
resuspended in buffer C (20 mM HEPES-KOH, pH 7.9, 25%
glycerol, 420 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 0.5 mM
dithiothreitol, 0.2 mM phenylmethylsulfonyl fluoride). The
suspension was placed on ice for 20 min, and the supernatant was stored
at 70 °C until use. GMSA was performed as follows. Single-stranded
sense and antisense oligonucleotides were annealed by heating in an
80 °C water bath for 5 min, then cooled to room temperature over 30 min. The double-stranded oligonucleotide probes (see Table I) used for
GMSA were end-labeled with [
-32P]ATP in a
T4 kinase reaction. Each protein binding experiment contained 4 µg of nuclear protein incubated in binding buffer (20%
glycerol, 5 mM MgCl2, 2.5 mM EDTA,
250 mM NaCl, 50 mM Tris-HCl, pH 7.5, 0.25 mg/ml
poly(dI-dC)·poly(dI-dC)) for 10 min at room temperature, then
radiolabeled probe was added and incubated for an additional 20 min at
room temperature. For competition experiments, 100-fold excess of each
competitor was preincubated with the nuclear extracts at room
temperature before the probe was added. The anti-Ap-1 antibody (Santa
Cruz Biotechnology, Inc. Santa Cruz, CA) studies were performed by
pretreating the nuclear extract with antibody (2 µg/reaction) for
4 h at 4 °C. The remainder of the procedure was as described
above. Protein-DNA complexes were resolved on a 4% nondenaturing
polyacrylamide gel that was dried and autoradiographed at
70 °C
with intensifying screens overnight.
Phorbol esters have been shown to induce MK
differentiation while simultaneously inhibiting E differentiation in
erythroleukemia cells (19, 20-22). The majority of work in this area
has been accomplished through the use of tetradecanoylphorbol acetate
and phorbol myristate acetate in K562 cells. During induction of MK development by phorbol esters there is decreased globin gene expression which occurs through negative regulation of both the rate of
production and the stability of
globin mRNA (21). In addition,
erythroleukemia cells induced with tetradecanoylphorbol acetate become
committed to the MK lineage (23). Similar results have been observed
with phorbol myristate acetate, which also induces the expression of
IL-6 and IL-6R genes in K562 cells (24). These studies suggest that the
changes associated with phorbol ester treatment may be mediated by an
IL-6 autocrine loop mechanism. Initial experiments were performed to
determine whether the K562 cell line variant being maintained in our
laboratory expressed the gene for IL-6 and its receptor endogenously.
Primer pairs for IL-6 and IL-6R were used to screen for mRNA
production by RT-PCR (Table I). The
ratios of mRNA production of IL-6 and IL-6R to that of GAPD were
found to be 0.49 and 1.0 respectively (data not shown). Subsequently,
we began investigating the role of IL-6 in erythroid versus
megakaryocytic lineage commitment by analyzing its effects on
globin gene expression. K562 cells were grown in increasing
concentrations of IL-6 (0, 10, 25, 50, and 100 ng/ml) in culture for
72 h followed by
globin mRNA quantitation by RT-PCR. The
level of detectable
globin gene expression decreased in a
concentration-dependent manner with a maximal (70%)
decrease at 100 ng/ml IL-6 (Fig. 1). This
response was determined to be highly significant using a multiple
regression model with a p value of 0.002 and
r2 of 0.86. This inhibition was significant at
25 ng/ml and maximal at 100 ng/ml, therefore subsequent experiments
were performed at the 50 ng/ml concentration of IL-6. The same RNA
samples obtained from K562 cells treated with increasing concentrations
of IL-6 were analyzed for expression of GATA-1, an E-specific nuclear trans-acting factor present in erythrocytes and
megakaryocytes (2). The expression of GATA-1 was unaffected by
increasing concentrations of IL-6 (data not shown) which suggests that
the down-regulation of
globin gene expression by exogenous IL-6 is
independent of significant changes in GATA-1 gene activity levels.
Similar experiments were performed to determine the effects of
exogenous IL-6 on globin chain biosynthesis. K562 cells were treated with IL-6 (100 ng/ml) for 72 h, and the total cellular protein was isolated for HPLC analysis.
Globin chain biosynthesis (expressed as the percent
globin chains of the total protein produced) was 3.2% in uninduced K562 cells (Fig.
2A). In contrast, the level of
globin protein produced after treatment with IL-6 was below the
level of detection (Fig. 2B). The protein analysis indicates
that IL-6 inhibits the production of
globin chains in K562 cells
which correlates with the results obtained for
globin mRNA
production.
To gain additional evidence that IL-6 was directly responsible for the
observed down-regulation of globin gene expression in our
experimental system, anti-IL-6 monoclonal antibody (5 and 10 µg/ml)
was incubated with IL-6 protein (50 ng/ml) for 1 h at 37 °C
prior to addition to K562 cells. Anti-IL-6 antibody treatment restored
globin mRNA levels to 80% (5 µg/ml) and 57% (10 µg/ml) of
untreated levels, demonstrating an abrogation of the inhibitory effects
of IL-6 (Fig. 1B). Thus, exogenous IL-6 treatment in K562 cells resulted in decreased
globin gene expression, which was abrogated by monoclonal antibodies to IL-6, and decreased
globin chain biosynthesis.
A second method, RNase protection assay, was used to analyze the
inhibitory effects of IL-6 on gene expression in K562 cells.
mRNA was quantitated as a ratio to GAPD (an internal control). Similar to the findings obtained by RT-PCR,
mRNA production was
decreased by 68% in the presence of 50 ng/ml IL-6 (Fig.
3). Likewise, pretreatment with anti-IL-6
antibodies resulted in
mRNA at 63% of base-line levels. These
findings are comparable to those obtained by RT-PCR analysis.
To determine whether the down-regulation of globin gene expression
by IL-6 was accompanied by an increase in MK-specific markers, studies
were performed to analyze the level of GpIIb gene expression, an early
marker for commitment to the MK lineage (25). GpIIb mRNA levels
increased 17% (p value = 0.01) by RT-PCR analysis
after 6 days of IL-6 (50 ng/ml) treatment (data not shown). The
sequential down-regulation of
globin mRNA production and the
up-regulation of GpIIb mRNA are similar to the results obtained in
K562 cells treated with tetradecanoylphorbol acetate (26). Together
these results suggest that IL-6 may be involved in E versus
MK lineage commitment.
Previous studies have demonstrated the
induction of globin gene expression in K562 cells treated with
hemin and sodium butyrate (27). Based on these data, we hypothesize
that IL-6 gene activity may be decreased when
globin gene
expression is induced by either of these two agents. Levels of IL-6
mRNA were analyzed at 0, 24, 48, and 72 h after the addition
of hemin (100 µM) or sodium butyrate (2 mM)
in cultures of K562 cells. A 99% and 50% decrease in IL-6 mRNA
was observed in sodium butyrate- and hemin-treated cells, respectively
(Fig. 4). These results suggest that
induction of
gene activity is associated with suppression of IL-6
gene expression and indirectly support a role for IL-6 in
globin
gene silencing during MK lineage commitment.
The results summarized for our studies in K562 cells include the
following: 1) endogenous expression of both IL-6 and IL-6R genes; 2)
simultaneous decrease in globin mRNA production and protein
biosynthesis in the presence of exogenous IL-6; 3) abrogation of the
inhibitory effects of IL-6 by anti-IL-6 monoclonal antibody; and 4)
stimulation of GpIIb mRNA production.
Although
erythroleukemia cell lines can serve as a paradigm for investigating
the process of hematopoietic differentiation, it is desirable to verify
findings in primary E cultures. Therefore, we performed parallel
experiments in primary cultures to corroborate the results obtained in
K562 cells. Peripheral blood mononuclear cells were isolated and grown
in 0.9% methylcellulose in the presence of stem cell factor,
granulocyte-macrophage colony-stimulating factor, IL-3, and
erythropoietin. BFU-E colonies from untreated controls and from
IL-6-treated mononuclear cells (IL-6 added on day 0 or day 7 of the
growth period) were counted on day 14. IL-6 added on day 0 resulted in
a 19% stimulatory effect on BFU-E colony growth compared with BFU-E
colonies grown in the absence of IL-6 (p < 0.05). IL-6
is known to enhance the activity of IL-3 which stimulates stem cells
and early progenitor cell growth in cultures (28). However, IL-6 added
on day 7 had a 92% inhibitory effect on BFU-E colony growth
(p < 0.01). These results are summarized in Fig.
5.
The total cytoplasmic RNA was isolated from the BFU-E colonies and the
levels of and
globin mRNA production were determined by an
RNase protection assay. Although IL-6 added on day 0 had a stimulatory
effect on total BFU-E colony number, mRNA data demonstrated an
inhibitory effect on
and
globin gene activity for both day 0 and day 7 conditions (Fig. 6). There was
a 63% and 80% decrease in
globin mRNA levels in BFU-E
colonies treated with IL-6 on day 0 and day 7, respectively
(p < 0.05). Likewise, a 79% (p < 0.05) decrease in
globin mRNA was also observed when IL-6 was added on day 7 (Table II). These
experiments give additional support for a role of IL-6 in globin gene
silencing and expand its inhibitory effects to both
and
globin
gene expression.
|
The 5 hypersensitive sites (5
HS) of the
globin LCR are required for high level expression of the globin
genes (29-32). Studies in K562 cells have demonstrated regulation of
the enhancer activity of the LCR, by proteins that bind to the
NF-E2/Ap-1 tandem repeat within 5
HS2 (33, 34). Therefore, we examined
the effects of IL-6 on protein binding to this tandem repeat. Initially
GMSA was performed with two oligonucleotide probes, one specific for Ap-1 and a second containing the NF-E2/Ap-1 tandem repeat of the
globin 5
HS2 (HS2 NF-E2). To determine specific binding patterns for
the Ap-1 and HS2 NF-E2 oligonucleotides, GMSA was performed with
extracts from untreated K562 cells. Equal amounts of nuclear proteins
(4 µg) were added to all GMSA reactions. The protein-DNA complexes
observed without competitor are shown in Fig.
7A, lanes 1 and
5. The experiments with the Ap-1 probe (Fig. 7A,
lanes 1-4) resulted in the formation of a single specific
protein-DNA complex. Specificity of the bands produced with each probe
was confirmed by cold competition reactions with the Sp-1 (Fig.
7A, lanes 4 and 8) and Oct-1 (data not
shown) oligonucleotides. In addition, studies performed with a
polyclonal anti-Ap-1-specific antibody resulted in a complete loss of
the single protein-DNA complex (data not shown), thus confirming the
protein in this complex to be Ap-1. Four specific protein-DNA complexes
were observed with the HS2 NF-E2 probe demonstrated by the cold
competition reaction shown (Fig. 7A, lane 6). The
B2 complex was specifically competed by the Ap-1 oligonucleotide (Fig.
7A, lane 7). Andrews et al. (4)
previously identified a single protein-DNA complex as NF-E2 with
nuclear extracts from mouse erythroleukemia cells and the NF-E2/Ap-1
tandem repeat, using p45 NF-E2 antiserum. In our study, a single
complex formed in experiments performed with the HS2 NF-E2 probe and
mouse erythroleukemia cell extract which migrated at the same rate as
the B4 complex produced in K562 cells (data not shown).
Experiments were completed to determine whether IL-6 had an immediate effect on the proteins bound to the Ap-1 or HS2 NF-E2 oligonucleotide probes. K562 cells were treated with IL-6 (50 ng/ml) for 0, 5, 15, 30, and 60 min and 4, 24, 48, and 72 h. There was an increase in the Ap-1 protein-DNA complex observed at 24 h with both the Ap-1 and HS2 NF-E2 oligonucleotides compared with uninduced extract (Fig. 7B, lane 4, and data not shown). The increased binding observed with the Ap-1 oligonucleotide was sustained in K562 cell nuclear extracts isolated after 48 and 72 h of IL-6 treatment (data not shown). These results suggest that Ap-1 may be a nuclear trans-acting factor that mediates the inhibitory effects of IL-6 on globin gene expression.
Additional experiments were performed with curcumin, a known inhibitor
of Ap-1 gene expression, to confirm the role of Ap-1 as a possible
mediator of the silencing effects of IL-6 on globin gene expression
(35). Various concentrations of curcumin were analyzed (5-50
µM) with maximum inhibition of Ap-1 and least toxicity to
the cells at a concentration of 20 µM (data not shown).
Subsequently, K562 cells were incubated in the presence of curcumin (20 µM) for 12 h prior to the addition of IL-6. RNase
protection assays showed no significant change in mRNA
production either in the absence or presence of curcumin. We conclude
from these results that IL-6 increases protein binding to the Ap-1
consensus sequence; however, the observed increase in Ap-1 is not
responsible for the inhibitory effects of IL-6 on
globin mRNA
production.
Cytokines are known to activate gene expression through the Janus
kinase/signal transducers and activators of transcription (STAT)
pathway via the STAT3 and STAT1 nuclear trans-factors (36). The sequences of the natural palindromic STAT-responsive elements vary
considerably, but they conform to the general structure
TT(N)5AA, where N is any DNA base (37). The motif with a
spacing of 4 bp selectively binds to complexes containing STAT3.
Although the majority of experimental data support an activator role
for the STAT proteins, there has been reported an indirect inhibitory effect for STAT1 (38). IL-6 activates gene expression via the STAT3 and
STAT1 nuclear trans-factors, which bind to specific DNA
sequences defined as IL-6REs. Therefore, a computer homology search was
performed comparing the palindromic STAT3 consensus sequence
TT(N)4AA with the A globin gene and promoter. At
nucleotide +11 relative to the cap site a potential IL-6RE TTCTGGAA
(A
IL-6RE) was identified which, by the palindromic spacing, would
be predicted to bind STAT3. Protein binding to this region in response
to IL-6 treatment in K562 cells was analyzed by GMSA (Table I). In
untreated K562 cells there was a single specific protein-DNA complex
(Fig. 8, lanes 1 and
3) which was not disrupted by cold competition reactions
with Sp-1 (Fig. 8, lane 2), Oct-1, transcription factor IID,
glucocorticoid response element, GATA-1, Ap-1, NF-E2, HS2 NF-E2, cyclic
AMP response element-binding protein, CAAT-binding transcription
factor/nuclear factor-1, and nuclear factor
B (data not shown). In
the presence of IL-6 (50 ng/ml) there was, by PhosphorImager quantitation of the acrylamide gels, a 30% increased protein binding at 30 min and 4 h of treatment with a further increase to 45% by
24 h. Further studies are being conducted to define the role of
nuclear trans-factors that bind to the A
IL-6RE in
response to IL-6, in regulating globin gene expression.
The molecular mechanisms that govern E lineage commitment are
largely undetermined. Utilization of human cell lines has served to
increase our understanding of globin gene regulation during differentiation and the cellular signals necessary for silencing E-specific genes during MK lineage commitment. Mixed E/MK colonies have
been described in both humans (3) and mice (39, 40). In humans, almost
all leukemic cell lines described as erythroleukemic or
megakaryoblastic express E- and MK-specific genes, and this dual
expression is found in the same cell (41, 42). Both K562 and human
erythroleukemia cells retain the ability to differentiate down the MK
or E lineage when exposed to specific inducers (26, 43). Recently,
Debili et al. (7) characterized a normal bipotential BFU-E/MK progenitor. This cell is found in the
CD34+/CD38 cell fraction isolated from bone
marrow samples by flow cytometry methods. Further evidence to support a
common bipotential E/MK progenitor is the shared expression of the
genes for the transcription factors GATA-1 (44), GATA-2 (5), stem cell
leukemia (45), and NF-E2 (22, 27). These genes are expressed in
multipotential progenitor, erythroblasts and megakaryocytes. GATA-1 is
a major activator of hematopoietic gene expression and is absolutely
required for erythropoiesis (2). The coordinate expression of GATA-1 and stem cell leukemia may be associated with the presence of a
conserved GATA motif in the stem cell leukemia promoter that binds
GATA-1 and mediates trans-activation of the stem cell
leukemia promoter (46). It is likely that transcription factor
hierarchies and networks of this type are important for hematopoietic
differentiation. It is clearly of interest to determine at which stage
of lineage commitment and differentiation the genes for particular
transcription factors are themselves active.
As E maturation progresses from embryonic to fetal to adult stage
erythroblast, there is a switch in globin chain production from to
to
globin, respectively (47). The expression of the individual
globin genes reflects the interactions of proximal and distal
cis-active regulatory elements with trans-acting
factors (47). An important distal cis-active regulatory
element is the
globin locus control region (30, 32) located
approximately 20 kilobases upstream of the
globin gene. This region
is composed of five DNase I-hypersensitive sites (5
HS1-5), four of
which are E-specific. A looping model has been proposed to explain how the LCR influences gene expression over many kilobases of DNA. Each
individual HS encompasses a 250-500-bp core DNA sequence and contains
binding motifs for E-specific and ubiquitous transcription activators
(29). The control mechanisms for globin gene silencing and loss of LCR
function during MK commitment are not known. Hematopoietic cell
differentiation signals are under the direct control of growth factors
that modulate the expression of nuclear trans-acting
factors. Therefore, it is reasonable to speculate that cytokines may
play a central role in globin gene silencing either through the loss of
LCR function or possible direct silencing of E-specific genes, during
MK lineage commitment.
The process of normal megakaryocyte maturation is stimulated by IL-6,
which synergizes with IL-3 to produce MK colonies from progenitor
cells, but IL-6 has no colony stimulating activity of its own. IL-6 is
a pluripotent cytokine with proinflammatory (48), hematopoietic (28),
and immunomodulatory (8) effects. Different human hematopoietic cell
lines produce IL-6 constitutively; these include K562, HEL, KU812,
Meg01, and Dami (49). Experimental data support the ordered silencing
of E-specific genes expressed in early progenitors when MK commitment
occurs. A role for IL-6 in the alternate lineage commitment process is
supported by the following experimental data. First, IL-6 has been
shown to inhibit the production of erythropoietin in perfused rat
kidneys (50). Second, induction of K562 and HEL cells with
tetradecanoylphorbol acetate results in silencing of E-specific genes
( globin) (21) and stimulation of MK-specific genes (GpIIb/IIIa)
(51). Finally, tetradecanoylphorbol acetate has been shown to induce
IL-6 and IL-6R gene expression in erythroleukemia cell lines (52).
Based on these experimental data we performed studies to analyze the direct effects of IL-6 on
globin gene expression in K562 and primary human progenitor cells. We demonstrated a direct
concentration-dependent suppression of
globin gene
expression in K562 cells with a concomitant increase in GpIIb mRNA
production, suggesting that IL-6 has the ability to divert K562 cell
differentiation toward the MK pathway. Conversely, when K562 cells were
grown in the two E differentiating agents, sodium butyrate or hemin,
endogenous IL-6 mRNA production was decreased to barely detectable
levels. We next extended our studies to primary cultures of human
progenitor cells and demonstrated the ability of IL-6 to inhibit BFU-E
colony growth significantly when added on day 7 in cultures. One might
speculate that on day 7 the multipotential progenitors present have
acquired the signals necessary for inhibition of E maturation by IL-6.
Maturation along the MK lineage did not occur in our culture system
because of a lack of the growth factors, thrombopoietin or IL-11,
required to support terminal MK maturation. Recent evidence suggests
that there may be factors in fetal bovine serum which inhibit MK
development in vitro (7). Navarro et al. (49)
demonstrated the expression of both the IL-6 and IL-6R genes during
normal megakaryocytopoiesis, suggesting that megakaryocyte maturation
may be regulated in part by an IL-6 autocrine loop. Clinical trials to
examine the efficacy of recombinant IL-6 have shown both increased
platelet counts and production of a rapid onset anemia (53, 54).
Therefore, collectively the in vitro and in vivo
data suggest an inhibitory role for IL-6 toward erythroid cell
commitment and maturation. The observation made in our primary culture
system has important implications for E maturation in vitro.
Whereas IL-6 has known proliferative effects toward early progenitor
cells, it may ultimately inhibit late E maturation.
The level of globin gene expression in E progenitor cells during
lineage commitment is modulated directly by cell-specific transcription
factors that bind to specific motifs in the locus control region and
proximal globin gene promoters. The LCR becomes active in
multipotential hematopoietic cells (55) and is required to establish an
active chromatin domain and expression of genes located in the globin locus (30). An important motif is the NF-E2/Ap-1 tandem repeat
that binds the nuclear trans-acting factor NF-E2, a
heterodimer of p45NF-E2, and the small Maf proteins p18 (MafF, MafG,
and MafK) which lack canonical trans-activation domains but
have dimerization and DNA binding domains (56). NF-E2 binds to the
NF-E2/Ap-1 tandem repeat and activates globin gene activity. Expression
of p45 NF-E2 is restricted to erythroid and mast cells, megakaryocytes
(2), and multipotential hematopoietic progenitors. A third ubiquitous
transcription factor Ap-1, a heterodimer of the two proto-oncogenes
c-fos and c-jun, binds to the Ap-1 site in the
NF-E2/Ap-1 tandem repeat, resulting in either transcription activation
or repression (57). Taken together these data suggest that a p45
NF-E2/Maf/c-jun/c-fos
concentration-dependent switch may regulate globin gene
expression through the NF-E2/Ap-1 binding site.
Experiments were completed to investigate possible molecular mechanisms
for inhibition of globin gene expression by IL-6. Protein binding to
the NF-E2/Ap-1 tandem repeat was performed using GMSAs. Nuclear
extracts from IL-6-treated K562 cells produced increased binding to the
Ap-1 consensus sequence by 24 h in culture (Fig. 7B).
In other systems IL-6 has been shown to induce the expression of Ap-1
via an IL-6RE (58). In view of the experimental data to support a
possible inhibitory role for homodimerization of Ap-1 at the tandem
NF-E2/Ap-1 site in the LCR, this observation suggested a possible
inhibitory role for Ap-1 in this setting. Additional studies with
curcumin, which specifically blocks Ap-1 gene expression, produced no
effect on the down-regulation of globin gene mRNA production in
the presence of IL-6. Therefore, the increase in Ap-1 binding may not
be involved in the mechanism for globin gene silencing by IL-6.
Gene activation by IL-6 occurs through the Janus kinase/STAT signal
transduction pathway via the STAT3 or STAT1 nuclear
trans-factors binding to specific DNA sequences defined as
IL-6REs, which activate gene expression. These target genes can then go
on either to activate or to suppress secondary genes. Such indirect
mechanisms for inhibiting gene activity have been defined for the
cyclin-dependent kinase inhibitor,
p21WAF1/CIP1 gene via STAT1 activation (38). A
potential IL-6RE TTCTGGAA was identified between nucleotides +11 and
+16 relative to the A gene cap site. This sequence is identical to
an IL-6RE identified in the promoter of the IL-6-responsive gene SAA3
(59). Protein binding experiments revealed a single specific
protein-DNA complex that was not competed in a specific manner by
several different oligonucleotides. Nuclear extract from K562 cells
treated with IL-6 showed 30% increased binding by 30 min, with a
sustained 45% increase in protein binding to the A
IL-6RE by
24 h (Fig. 8). Experiments to identify the protein(s) that binds
to the A
IL-6RE are in progress. Collectively the data presented in
our study support a possible role for IL-6 in E versus MK
lineage commitment and E-specific gene silencing. The molecular
mechanism for the inhibitory effects of IL-6 on globin gene expression
may involve the Janus kinase/STAT signal transduction pathway.