| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
RESEARCH COMMUNICATION |
a Institute of Physiology, University of Zürich-Irchel, 8057 Zürich, Switzerland;
b Max-Planck-Institute for Molecular Physiology, 44026 Dortmund, Germany; and
c Institute of Physiology, Medical University of Lübeck, 23538 Lübeck, Germany
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
isoforms was detected in all EBs
tested at different differentiation stages, we propose that HIF-1
modulates ß-globin expression during development.—Bichet, S.,
Wenger, R. H., Camenisch, G., Rolfs, A., Ehleben, W., Porwol, T.,
Acker, H., Fandrey, J., Bauer, C., Gassmann, M. Oxygen tension
modulates ß-globin switching in embryoid bodies.
Key Words: embryonic stem cells • hemoglobin • hypoxia-inducible factor • erythropoietin
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The mouse ß-like globin gene cluster resides on chromosome 7 and
consists of five actively transcribed genes (
y, ßH0,
ßH1, ßmaj, ßmin), which are expressed in
a stage-specific manner (3
, 4
). During
embryonic and fetal development, erythropoiesis is subjected to rapid
changes. At day 7.5 of gestation, embryonic y- and
ßH1-globin genes are expressed in nucleated cells of blood islands
that begin to form in the central area of the egg cylinder
(5)
. By day 9, a primitive capillary plexus is formed,
allowing these cells to enter the circulation (6)
. The egg
cylinder/yolk sac remains the primary site of blood formation until
approximately midgestation. After this stage, yolk sac erythropoiesis
declines; at day 10.5 of gestation, the fetal liver becomes the major
organ for erythropoiesis until late gestation (reviewed in ref 7
).
Non-nucleated erythrocytes of hepatic origin, entering circulation at
about day 12 postcoitum, contain both forms of fetal/adult globin
(ßmaj and ßmin) (8)
. Finally,
starting at day 16–17 of development, the site of erythropoiesis is
again switched to bone marrow and spleen.
The establishment of an in vitro model of mouse embryogenesis based on differentiating embryonic stem (ES)2cells has facilitated the investigation of several aspects of mouse development (reviewed in 9, 10 ). ES cells are derived from the inner cell mass of blastocysts and can be cultivated in vitro without losing their pluripotency by addition of the leukemia inhibitory factor (LIF) (11) . In the absence of LIF, ES cells differentiate spontaneously, thereby forming 3-dimensional structures termed embryoid bodies (EBs) (12) . These embryo-like structures are known to contain derivatives from the three primitive germ layers. Pluripotent ES cells undergo differentiation resulting in various committed cell types, including hematopoietic precursors (9 , 10 ). Keller and co-workers (13) recently reported that primitive erythrocytes and other hematopoietic lineages arise from a common ES cell-derived precursor within EBs.
We sought to analyze whether changes in oxygen supply affect globin switching during mammalian development. The rationale is that during early embryogenesis, the peri-implantation embryo resides in the uterus with reduced oxygen tension (14) whereas upon vascularization of the developing fetus, oxygen supply to the conceptus increases. Previous reports showing that EBs express the full complement of mouse embryonic globin genes and that a switch occurs to fetal/adult globin genes (15 , 16 ) made this system seem attractive for our purposes. Indeed, we recently demonstrated that EBs are able to sense changes in oxygen concentrations and specifically respond to reduced oxygenation by increased expression of the oxygen-regulated genes vascular endothelial growth factor and aldolase A (17) . In this work, we provide evidence that ß-globin switching is modulated in an oxygen-dependent manner.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
promoter/enhancer (20)
, SV40 splice and polyadenylation
sites, and a neomycin-resistance gene driven by RSV-LTR sequences
(R. H. Wenger, unpublished results). After linearization with
XmnI, 50 µg plasmid DNA was transfected into 1 x
107 CCE ES cells by electroporation at 960 µF and 250 V
in a 0.4 cm cuvette (GenePulser, BioRad, Richmond, Calif.). Stable
transfectants were selected with 0.5 mg/ml G418 (Gibco-BRL,
Gaithersburg, Md.); resistant clones were analyzed for EPO expression
by Northern blotting (see below) and radioimmunoassy (21)
.
mRNA analysis
EBs were collected at the indicated time points, diluted with 2
volumes of Iscove's modified Dulbecco's medium (IMDM), and
centrifuged at 150 x g for 10–20 min at 4°C.
Subsequently, RNA was isolated from the pellet (22)
and
Northern blotting was performed, as described previously
(23)
, using 10 µg of total RNA. Both the ßH1-globin
probe obtained by reverse transcriptase-polymerase chain reaction
(RT-PCR) from 13-day mouse yolk sac RNA and the
ßmaj-globin probe generated from the genomic clone GSE428
(16)
were kindly provided by M. V. Wiles. The EPO
probe was derived from the plasmid pe49f. The probe for the ribosomal
protein L28 (24)
was used for control hybridizations.
The two mRNA isoforms encoding mouse hypoxia-inducible factor-1
(HIF-1) were detected by RT-PCR, as described previously
(25)
. Two different forward primers, specific for either
exon I.1 or exon I.2, and a common reverse primer specific for exon III
were used. EPO mRNA in EBs was quantitated by competitive RT-PCR, using
a synthetic EPO cRNA template as competitor (21)
. Human
and mouse EPO mRNA within the transfected EBs was analyzed by RT-PCR,
each using one EPO primer set specific for human or mouse EPO mRNA. The
amplification profile was 94°C for 60 s, 55°C for 60 s,
and 72°C for 100 s. RT-PCR products were 395 bp and 280 bp for
mouse (35 cycles) and human (30 cycles) EPO, respectively.
Protein analysis
Expression of the two HIF-1 protein subunits, HIF-1, and aryl
hydrocarbon receptor nuclear translocator (ARNT; HIF-1ß) was examined
by Western blot analysis. Nuclear extracts (25 µg) derived from ES
cells were separated by 7.5% sodium dodecyl sulfate-polyacrylamide gel
electrophoresis (SDS-PAGE) and electrotransferred onto nitrocellulose.
The HIF-1 subunits were detected using a chicken polyclonal antibody
raised against human HIF-1 (26)
and a polyclonal rabbit
antibody raised against ARNT (kindly provided by Y. Fujii-Kuriyama).
The blots were subsequently incubated with a secondary rabbit
anti-chicken (G1351, Promega, Madison, Wis.) or a goat anti-rabbit
antibody coupled to horseradish peroxidase (Pierce, Rockford, Ill.) and
developed with SuperSignal CL-HRP chemiluminescence substrate (Pierce).
DNA binding activity of HIF-1 was determined by electrophoretic
mobility shift assay (EMSA) as described previously (27)
.
Briefly, nuclear extracts were prepared from normoxic and hypoxic ES
cell cultures and incubated with a radioactively labeled
oligonucleotide derived from the oxygen-responsive EPO 3' enhancer.
Where indicated, the monoclonal anti-HIF-1 antibody mgc3
(26)
or the monoclonal anti-ARNT antibody 2B10 (kindly
provided by G.H. Perdew) were subsequently added to the binding
reaction. The reaction mixtures were separated by nondenaturing PAGE
and the dried gels were evaluated by phosphorimaging (Molecular
Dynamics, Sunnyvale, Calif.).
Hemoglobin quantitation
At appropriate time points, 100–3000 EBs were collected,
counted, diluted with 2 volumes IMDM, and centrifuged at 150 x
g for 10–20 min. The pellet was resuspended in 2 ml dispase
II (2.4 U/ml) (Boehringer-Mannheim), supplemented with 0.1 U/ml of
collagenase, and incubated at 37°C for 10–20 min. After cell
counting using the trypan blue exclusion method (0.4% (w/v) in
phosphate-buffered saline) to assess cell viability, the dissociated
cells were lysed in 550 to 600 µl 0.01% (v/v) NP-40 in water (Fluka,
St. Louis, Mo.). One aliquot of the cell lysate was used for protein
quantitation (BioRad). The other aliquot was used to determine the
hemoglobin concentration in the dissociated cells. To this end, we
modified a 2,7-diaminofluorene (DAF) colorimetric assay, which is based
on the oxidative transition of DAF to fluorene blue by the
pseudoperoxidase activity of hemoglobin (28)
. A mixture of
15 µl of 1% DAF (w/v) (Sigma, St. Louis, Mo.) in 90% glacial acetic
acid with 1.5 ml of assay buffer (100 mM Tris phosphate buffer, pH 7.0,
6 M urea, 0.18% w/v H2O2) was added to 500
µl of the supernatant of the lysed cells. After incubation for 12 min
at 25°C, the absorbance was measured at 610 nm. Fresh mouse blood was
used as hemoglobin standard.
Spectrophotometry
Whole mount light absorption photometry was performed as
described previously (29)
. In brief, EBs at
differentiation days 7–15 were placed in a superfusion chamber
(30)
on a small bench containing small holes of about the
same diameter as the EBs. Isotonic salt solutions containing 5 mM
glucose were equilibrated with different
O2/CO2/N2 mixtures in order to
adjust varying oxygen tensions at pH 7.4 (10 ml/min at 36°C). EBs
were supplied symmetrically with nutrients by this procedure. The
superfusion chamber was mounted on the stage of a light microscope for
light absorption measurements. Light from a halogen bulb (12 V, 100 W)
transilluminated the EBs only, since the bench was made opaque by
spattering with gold to avoid uncharacteristic light scattering.
Following this procedure, only the light coming through the EB's
tissue into the objective (40x) was recorded by a photodiode-array
spectrophotometer (MCS 210, Zeiss) connected to the third ocular of the
microscope trinocular head via a light guide. Difference spectra
(anaerobic vs. aerobic steady-state) were recorded by defining 20%
O2, 3% CO2 and 77% N2
equilibrating conditions in the superfusion medium as reference
(aerobic steady-state), which was automatically subtracted from the
spectra recorded under reducing conditions by equilibrating the
superfusion medium with 3% CO2 and 97% N2.
The recorded spectra were evaluated as described previously
(29)
.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
The expression of fetal/adult (ßmaj) globin mRNA was first detected after 5–6 days of differentiation, peaked at day 10, and decreased continuously (but less substantially than the embryonic globin mRNA level) until it could no longer be detected after day 16 (Fig. 1A ). The expression pattern of ßmaj-globin also agrees with previous reports in which low mRNA signals were detected in EBs as early as day 4 (32) , increasing over the ensuing days and dropping to undetectable levels after day 15 (15) . Thus, the ß-globin expression pattern is highly reproducible in EBs.
ß-Globin switching is dependent on the oxygen concentration
Because the conceptus experiences changes in oxygen supply during
development, we sought to analyze whether increasing oxygen
concentrations influence ß-globin switching in EBs. To recapitulate
the in utero oxygenation conditions to a given extent, EBs were
cultured at varying oxygen concentrations, as described
(17)
. In brief, EBs were exposed to 1% O2 for
72 h, switched to 4% O2 for an additional 96 h,
and finally switched to 20% O2 until differentiation day
18. Compared to the normoxic cultivation of EBs, these changes in
oxygen concentration did not influence the average number of viable
cells forming an EB (data not shown). Total RNA extracted from the EBs
cultured at increasing oxygen concentrations was analyzed for embryonic
and fetal/adult ß-globin mRNA expression by Northern blotting.
Similar to the EBs shown in Fig. 1A
, specific ßH1-globin
mRNA was detected as early as at day 5 of differentiation, showed a
peak at day 8, and decreased after day 10 (Fig. 1B
). In
contrast to normoxically cultured EBs, the appearance of
ßmaj-globin mRNA was delayed by about 2 days, showed weak
signals until day 8 of differentiation, and reached a plateau at day 12
that, unlike normoxically grown EBs, persisted until day 18 when the
experiment was terminated (Fig. 1B
).
Kinetic analysis of hemoglobin protein formation in differentiating
EBs
Quantitation of hemoglobin concentration in EBs grown at
increasing concentrations of oxygen was performed by means of a
DAF-based colorimetric assay (28
, 33
) adapted
for hemoglobin detection in EBs (19)
. DAF colorimetry is
less hazardous and more sensitive than the benzidine-based method.
Figure 2
shows hemoglobin concentrations in EBs from four independent
experiments. No hemoglobin was detected within the first 4 days of
differentiation. Between days 5 and 9 we observed a first wave of
hemoglobin production, peaking at days 5–7 with a hemoglobin
concentration of about 1.1–2.6 pg/cell. A second increase in
hemoglobin was observed from day 12 until day 18, when the experiment
was terminated.
|
Since the DAF-based (as well as the benzidine-based) hemoglobin colorimetric assay relies on the pseudoperoxidase activity of hemoglobin, we sought to verify these results by spectrophotometric analysis of hemoglobin within single EBs. Figure 3A shows an anaerobic (N2-saturated perfusion medium) vs. aerobic steady-state spectrum of EBs at differentiation days 15 and 11. The difference spectrum at day 15 was dominated by hemoglobin with typical troughs at 538 nm and 577 nm and peaked at 554 nm and 590 nm, as indicated by the superimposed difference spectrum of isolated mouse hemoglobin (dotted line). This corresponds well with anaerobic vs. aerobic steady-state spectra of single EBs at the differentiating ages of 7 or 8 and, although less pronounced, 13 days, as shown in Fig. 3B . The different variations indicate different signal-to-noise ratios due to varying hemoglobin contents that peaked at days 8 and 15. The single peaks of the anaerobic vs. aerobic steady-state difference spectrum of the EB at day 11 (Fig. 3A ) coincide with the characteristic peaks of reduced cytochrome c (552 nm), cytochrome b (563 nm), and cytochrome aa3 (605 nm) of the respiratory chain as indicated by the different spectra of isolated cytochrome c, b, aa3, appearing as shadowed curves in Fig. 3A (34) . This observation indicates that the hemoglobin content at day 11 is very low. In summary, the pattern of appearance, disappearance, and reappearance of hemoglobin, as exemplified in Fig. 3B , very closely resembles the one obtained by DAF colorimetry. EBs apparently generate two waves of hemoglobin, an early and a late one.
|
The hypoxia-inducible factor-1 (HIF-1) is expressed in ES cells and
EBs
From fertilization until delivery, cells contributing to the
embryo face continuous changes in oxygen supply (14)
.
Reduced oxygenation is known to activate HIF-1, a heterodimeric
transcription factor consisting of the specific HIF-1 subunit and
the common heterodimerization partner ARNT (HIF-1ß) (reviewed in 35, 36
). Activation of HIF-1 in turn results in induced expression of a
variety of genes including erythropoietin (EPO), vascular endothelial
growth factor, transferrin, and glycolytic enzymes (35)
.
Thus, one might expect that reduced oxygenation influences expression
of hypoxically regulated genes in the growing conceptus, thereby
perhaps modulating developmental mechanisms such as ß-globin
switching.
As a first step, we tested normoxic (20% O2) and hypoxic
(1% O2 for 4 h) ES cells for expression of HIF-1
and ARNT (HIF-1ß) by Western blot analysis. In view of their
characteristic feature to grow in clumps, ES cells were generally grown
to 30–50% confluency only, thereby avoiding the formation of a
hypoxic microenvironment in normoxically cultured cells. Although ARNT
(but no HIF-1) protein could be detected in nuclear extracts derived
from normoxic cells, both HIF-1 subunits appeared upon hypoxic exposure
of the cells (Fig. 4A
).To test the ability of ES cells to form a functional HIF-1 complex, we
performed EMSAs using an HIF-1 binding oligonucleotide derived from the
3' region of the EPO gene (27)
. As shown in Fig. 4B
and as we reported previously (37
,
38
), a constitutive and a nonspecific factor present in
normoxic cell extracts bound the HIF-1 probe. An additional DNA binding
activity resulting in retarded electrophoretic mobility appeared
exclusively in hypoxic ES cell extracts. Supershift experiments using
monoclonal antibodies against HIF-1 and ARNT confirmed the identity
of HIF-1. Very recently, the ES cell line J1 has been reported to
express HIF-1 in nonhypoxic culture conditions (39)
.
However, further analysis of three additional unrelated ES cell lines
exposed to hypoxia or normoxia confirmed our results shown in Fig. 4A, B
(I. Kvietikova, D. Chilov, and M.Gassmann, unpublished
results). Thus, the constitutive expression of HIF-1 appears to be a
specific property of J1 cells that is not shared by other ES cells
analyzed so far.
|
Next, we tested our developing EBs for the presence of HIF-1 mRNA.
As reported previously, mice express two different HIF-1 isoforms
that contain alternative first exons (25
,
40
). These two mRNAs give rise to two predicted
translation products differing in the translational start site by 12
amino acids. Therefore, we used an RT-PCR approach capable of
distinguishing between the two HIF-1 mRNA isoforms
(25)
. As shown in Fig. 4C
, both mRNAs were
expressed at approximately constant ratios in all differentiating EBs
tested. Together, ES cells produce HIF-1 in an oxygen-dependent manner
and HIF-1 isoforms are expressed throughout differentiation to EBs,
suggesting that HIF-1 might regulate expression of a variety of genes
during EB development.
EPO is up-regulated in hypoxic EBs but does not influence
ß-globin switching
HIF-1 mediates hypoxic up-regulation of the hematopoietic growth
factor EPO, which is known to regulate survival, proliferation, and
differentiation of erythroid progenitor cells upon binding to the EPO
receptor (EPO-R) (reviewed in 41, 42
). Null mutant mice lacking
either EPO or EPO-R have been reported to exhibit altered yolk sac
erythropoiesis to some extent, but a drastic reduction of definitive
erythropoiesis in the fetal liver leading to embryonic death at day
13–13.5 postcoitum (43
, 44
). Thus, the
effects of EPO on primitive and definitive erythropoiesis are
substantially different. Since EPO is regulated in an oxygen-dependent
manner, one might envision that upon hypoxic stimulation of EBs, EPO
might differentially regulate embryonic and fetal/adult globin
expression.
To test this notion, we first determined by means of RT-PCR whether EPO
and EPO-R mRNA were both expressed in our developing EBs. In agreement
with previous reports (19
, 31
,
32
, 45
), specific EPO and EPO-R mRNA were
detected at comparable levels in EBs during all stages of
differentiation examined (Fig. 5A
).Hypoxic inducibility of EPO was analyzed by measuring EPO mRNA levels in
9-day-old EBs that were preincubated at 20% or 1% oxygen for 24 h prior to exposure to 1% or 20% oxygen, respectively, for 4 h.
Using a synthetic EPO cRNA as competitor template (21)
, we
showed for the first time by competitive RT-PCR analysis that EPO mRNA
was reversibly increased by a factor of 6 in EBs when exposed to
hypoxia (Fig. 5B
). To detect a putative function of EPO on
ß-globin switching directly and to avoid exposure of EBs to hypoxia,
which in turn induces expression of a variety of other genes, we
established an EPO-overexpressing ES cell subclone. Since human EPO
(hEPO) is able to activate mouse erythropoiesis (41)
, we
linked the hEPO cDNA to the SR promoter/enhancer sequence that we
had previously reported to efficiently overexpress a reporter gene in
the 1C11 cell subline, a derivative of the mouse embryonic carcinoma
(EC) cell line F9 (20)
. Of note, EC cells closely resemble
ES cells. Northern blot analysis revealed the presence of hEPO mRNA in
three stably transfected ES cell subclones (data not shown). Whereas
the parental ES cell line did not produce detectable levels of
immunoreactive EPO, the supernatant of the chosen subclone contained 26
±0.1 mU EPO/ml (n=4). Subsequently, the hEPO-overexpressing
ES cell subclone was allowed to differentiate to EBs at 20% oxygen,
and total RNA was extracted from growing EBs every other day starting
at differentiation day 4. The presence of both endogenous and exogenous
EPO mRNA was analyzed by RT-PCR, using a specific primer set for the
mouse and human EPO gene. Figure 5C
shows two bands of the
expected size in all mRNA samples tested, implying that both EPO genes
were expressed during EB differentiation.
|
Total RNA from the same experiment was used to study ß-globin expression (Fig. 5D ). Northern blot analysis revealed a similar pattern of embryonic and fetal/adult ß-globin expression as was observed in normoxically grown EBs derived from the untransfected parental control ES cell line (compare Fig. 1A and Fig. 5D ): the ßH-globin mRNA levels were highest on days 6–8 of differentiation and strongly declined after day 10. The ßmaj-globin mRNA was first detected at day 6, peaked at day 10, and decreased continuously until day 18. Thus, overexpression of hEPO does not influence the temporal expression pattern of embryonic and fetal/adult ß-globin during EB differentiation.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Analysis of hemoglobin in EBs grown at varying oxygen concentrations was performed by two independent methods. DAF colorimetry and spectrophotometry both revealed the presence of early and late waves of hemoglobin. Based on the ßH1- and ßmaj-globin mRNA expression pattern, it is tempting to speculate that the early wave represents yolk sac erythropoiesis whereas the second wave mirrors the switch to fetal liver erythropoiesis. Similarly, Wiles and Keller (16) discussed the appearance of a late wave of globinization in normoxic EBs at day 18 of differentiation.
Favorable effects of hypoxia on EB development
EBs exposed to varying oxygen concentrations grew equally well
compared with the control EBs cultured at normoxia, as judged by the
average numbers of cells within EBs. This agrees with two previous
reports in which we and others (17
, 49
)
observed that hypoxic exposure at the beginning of differentiation did
not impair the ES cells' viability and had a favorable effect on
plating efficiency and hematopoietic differentiation. These
observations might be explained by hypoxically induced expression of
several growth factors that interact in concert with each other during
EB development. In fact, EBs are known to express a variety of
oxygen-regulated growth factors (17
, 31
,
45
, 50
, 51
); two of them, EPO
(this work) and vascular endothelial growth factor (17)
,
were found to have increased steady-state mRNA levels after hypoxic
exposure of EBs. Apart from differentiating ES cells, enhanced
erythroid colony formation has been observed when human marrow was
cultured at 5% oxygen (52)
.
By analogy to the observations in cell culture and animal models,
hypoxic induction of specific gene transcription in EBs probably occurs
via activation of the ubiquitously expressed transcription factor
HIF-1, which in turn binds to the HIF-1 binding site present in the
flanking regions of oxygen-regulated genes (36)
. Indeed,
expression of HIF-1 and ARNT was observed in all ES cells tested.
Moreover, we found that both HIF-1 isoforms that emerge from two
alternative first exons (25)
were expressed in developing
EBs. Whether these two isoforms, one being expressed from a
tissue-specific and the other from a housekeeping-type promoter
(40)
, have distinct physiological functions during
differentiation remains to be elucidated. Together, we propose that,
upon reduced oxygen supply and HIF-1 activation, EBs are capable of
inducing required (growth) factors, which in turn support their own
differentiation.
EPO gene expression in EBs
Based on an early report suggesting that EPO might be involved in
the fetal to adult hemoglobin switch (53)
, we generated
EBs derived from an EPO-overexpressing ES cell subline. However, the
kinetics of ßH1 and ßmaj-globin mRNA expression was not
altered in EPO-overexpressing EBs (26 mU EPO/ml). This agrees with a
previous report showing that addition of recombinant EPO (2000 mU/ml)
did not influence the kinetics of ß-globin mRNA expression
(16)
. The abundance of both ß-globins, however, was
reported to be about threefold stronger in the presence of exogenous
EPO. In contrast, we did not observe a similar effect by overexpressing
EPO in our EBs. This discrepancy might be explained by the lower and
perhaps more physiological concentrations of EPO expressed from
transfected cells within the EBs and/or by different posttranslational
modifications.
Expression of EPO and EPO-R during the earliest stages of EB development prior to the onset of erythropoiesis raised the question of whether EPO plays a yet unknown developmental function apart from erythropoiesis (reviewed in ref 54 ). Knowing, for instance, that EPO has a mitogenic and chemotactic effect on endothelial cells that express EPO-R (55 , 56 ), one might postulate that EPO is involved in angiogenesis and/or vascular development. On the other hand, expression of EPO and EPO-R in the mammalian (including human) brain (21 , 57-61 ) allows to speculate that EPO is involved in brain development. In analogy to its function during erythropoiesis, where EPO prevents apoptosis of erythroid progenitor cells (42) , EPO might be involved in the regulation of apoptosis in the developing brain.
How does hypoxia influence ß-globin switching in EBs?
As concluded from the above observations, EPO alone does not
influence ß-globin switching in developing EBs. How, then, is this
switch regulated? One possibility is that the transcription factor
HIF-1 activates a yet unknown gene, which in turn modulates ß-globin
expression. On the other hand, one might postulate that HIF-1 binds
directly to a putative HIF-1 binding site present in the locus control
region or promoters of the ß-globin cluster, thereby repressing
embryonic globin expression and/or activating fetal/adult globin
expression. In a computer-assisted search using
A/(G)CGTG as query (35)
, putative
HIF-1 binding sites were detected in the ß-globin locus control
region. Knowing that binding to this consensus sequence might be
influenced by the neighboring nucleotides as well as by adjacent sites
for other transcription factors (36)
, these putative HIF
binding sites have to be analyzed for functionality by EMSA and
transient transfection experiments. Understanding the molecular
mechanism(s) of ß-globin gene regulation is a prerequisite for
therapeutic treatment of patients suffering from ß thalassaemia and
sickle cell anemia. Once these mechanims are fully understood, one
might envision the possibility of altering the switch from
embryonic/fetal to adult hemoglobin production in patients.
Evidence for the need of a hypoxic environment for normal
development
Hypoxic modulation of ß-globin switching supports our hypothesis
that a low oxygen environment found in utero is mandatory for the early
embryo to express oxygen-regulated genes in the correct temporal order,
thereby ultimately driving their development (17)
. Around
midgestation, the capillary network is present, allowing efficient
oxygen supply to the tissue and thereby decreasing hypoxia-induced gene
expression to constitutive levels. Targeted disruption of HIF-1, the
master regulator of oxygen homeostasis, resulted in embryonic lethality
at around midgestation due to cardiovascular malformations and
neuronal defects (39
, 62
). Further evidence
in support of our hypothesis came from recent studies showing that null
mutant mice lacking ARNT, the heterodimerization partner of HIF-1,
also died at midgestation. The ARNT-deficient fetuses were reported to
suffer from defective angiogenesis of the yolk sac and solid tissues
(63)
or from a failure to develop the embryonic component
of the placenta to vascularize and form the labyrinthine
spongiotrophoblast (64)
. In keeping with this, it has been
reported that human cytotrophoblasts continued proliferating and
differentiated poorly when cultured at 2% oxygen, a concentration
mimicking the uterine conditions at wk 10 of gestation. In contrast,
when grown at 20% oxygen, the cells stopped proliferating and
differentiated normally (65)
. Thus, the uterine oxygen
tension regulates placental growth. Additional evidence in support of
our hypothesis awaits studies of the developing embryo in which the
oxygen concentration has been altered in vivo.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Abbreviations: ARNT, aryl hydrocarbon receptor nuclear translocator; DAF, 2,7-diaminofluorene; EBs, embryoid bodies; EMSA, electrophoretic mobility shift assay; EPO, erythropoietin; EPO-R, EPO receptor; ES, embryonic stem; hEPO, human EPO; HIF, hypoxia-inducible factor; IMDM, Iscove's modified Dulbecco's medium; LIF, leukemia inhibitory factor; PAGE, polyacrylamide gel electrophoresis; RT-PCR, reverse transcriptase-polymerase chain reaction.
Received for publication July 20, 1998.
Revision received October 5, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Eur. J. Biochem. 1997;246:155-165.[Medline]
. FASEB J 1998;13:81-88.. Genes & Dev 1998;12:149-162. is encoded by two different mRNA isoformsexpression from a tissue-specific and a housekeeping-type promoter. Blood 1998;91:3471-3480. is required for solid tumor formation and embryonic vascularization. EMBO J 1998;17:3005-3015.[Medline]
This article has been cited by other articles:
|
|
 |
|
  H. Komita, X. Zhao, J. L. Taylor, L. J. Sparvero, A. A. Amoscato, S. Alber, S. C. Watkins, A. D. Pardee, A. K. Wesa, and W. J. Storkus CD8+ T-Cell Responses against Hemoglobin-{beta} Prevent Solid Tumor Growth Cancer Res., October 1, 2008; 68(19): 8076 - 8084. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Yoon, Y. D. Pastore, V. Divoky, E. Liu, A. E. Mlodnicka, K. Rainey, P. Ponka, G. L. Semenza, A. Schumacher, and J. T. Prchal Hypoxia-inducible Factor-1 Deficiency Results in Dysregulated Erythropoiesis Signaling and Iron Homeostasis in Mouse Development J. Biol. Chem., September 1, 2006; 281(35): 25703 - 25711. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Ateghang, M. Wartenberg, M. Gassmann, and H. Sauer Regulation of cardiotrophin-1 expression in mouse embryonic stem cells by HIF-1{alpha} and intracellular reactive oxygen species J. Cell Sci., March 15, 2006; 119(6): 1043 - 1052. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Hofer, R. H. Wenger, M. F. Kramer, G. C. Ferreira, and M. Gassmann Hypoxic up-regulation of erythroid 5-aminolevulinate synthase Blood, January 1, 2003; 101(1): 348 - 350. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Sun, K. Jin, X. O. Mao, Y. Zhu, and D. A. Greenberg Neuroglobin is up-regulated by and protects neurons from hypoxic-ischemic injury PNAS, December 6, 2001; (2001) 251466698. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. L. Ramirez-Bergeron and M. C. Simon Hypoxia-Inducible Factor and the Development of Stem Cells of the Cardiovascular System Stem Cells, July 1, 2001; 19(4): 279 - 286. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. T. Ruschitzka, R. H. Wenger, T. Stallmach, T. Quaschning, C. de Wit, K. Wagner, R. Labugger, M. Kelm, G. Noll, T. Rulicke, et al. Nitric oxide prevents cardiovascular disease and determines survival in polyglobulic mice overexpressing erythropoietin PNAS, October 10, 2000; 97(21): 11609 - 11613. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Sun, K. Jin, X. O. Mao, Y. Zhu, and D. A. Greenberg Neuroglobin is up-regulated by and protects neurons from hypoxic-ischemic injury PNAS, December 18, 2001; 98(26): 15306 - 15311. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||
