Molecular and functional characterization of EPOR

Developmental and Comparative Immunology 45 (2014) 191–198
Contents lists available at ScienceDirect
Developmental and Comparative Immunology
journal homepage: www.elsevier.com/locate/dci
Molecular and functional characterization of erythropoietin receptor of
the goldfish (Carassius auratus L.)
Fumihiko Katakura a, Barbara A. Katzenback a, Miodrag Belosevic a,b,⇑
a
b
Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada
Department of Medical Microbiology & Immunology, University of Alberta, Edmonton, Alberta, Canada
a r t i c l e
i n f o
Article history:
Received 31 January 2014
Revised 25 February 2014
Accepted 26 February 2014
Available online 19 March 2014
Keywords:
Erythropoietin receptor
EPOR
Teleost
Goldfish
Erythropoiesis
a b s t r a c t
Erythropoietin receptor (EPOR) is a member of the class I cytokine receptor superfamily and signaling
through this receptor is important for the proliferation, differentiation and survival of erythrocyte progenitor cells. This study reports on the molecular and functional characterization of goldfish EPOR. The
identified goldfish EPOR sequence possesses the conserved EPOR ligand binding domain, the fibronectin
domain, the class I cytokine receptor superfamily motif (WSXWS) as well as several intracellular signaling motifs characteristic of other vertebrate EPORs. The expression of epor mRNA in goldfish tissues, cell
populations and cells treated with recombinant goldfish EPO (rgEPO) were evaluated by quantitative PCR
revealing that goldfish epor mRNA is transcribed in both erythropoietic tissues (blood, kidney and spleen)
and non-hematopoietic tissues (brain, heart and gill), as well as in immature erythrocytes. Recombinant
goldfish EPOR (rgEPOR), consisting of its extracellular domain, dose-dependently inhibited proliferation
of progenitor cells induced by rgEPO. In vitro binding studies indicated that rgEPO exists as monomer,
dimer and/or trimmer and that rgEPOR exists as monomer and/or homodimer, and when incubated
together, formed a ligand–receptor complex. Our results demonstrate that goldfish EPO/EPOR signaling
has been highly conserved throughout vertebrate evolution as a required mechanism for erythrocyte
development.
Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction
Erythrocytes are unique cells that carry oxygen, a molecule
absolutely required for energy production in all vertebrates. It follows that the maintenance of circulating erythrocyte volume is
critical for sustaining the life of animals. The life span of mature
human erythrocytes is estimated to be about 120 days (Shemin
and Rittenberg, 1946); therefore, the continuous production of
erythrocytes from hematopoietic stem cells must be tightly regulated. Erythropoiesis, the development of erythrocytes from their
progenitor cells, is primarily controlled by erythrocyte growth factor, erythropoietin (EPO) binding to its receptor (EPOR) which
causes proliferation, differentiation and survival of erythrocyte
progenitor cells, including burst-forming units-erythroid (BFU-E)
and colony-forming units-erythroid (CFU-E) (reviewed by
Constantinescu et al., 1999; Fisher, 2003).
⇑ Corresponding author at: Department of Biological Sciences, University of
Alberta, CW-405 Biological Sciences Building, Edmonton, Alberta T6G 2E9, Canada.
Tel.: +1 780 492 1266; fax: +1 780 492 2216.
E-mail address: mike.belosevic@ualberta.ca (M. Belosevic).
http://dx.doi.org/10.1016/j.dci.2014.02.017
0145-305X/Ó 2014 Elsevier Ltd. All rights reserved.
EPOR is a member of class I cytokine receptor superfamily that
includes the interleukin receptors, growth hormone receptors, and
colony-stimulating factor receptors (Huising et al., 2006). The
EPOR is composed of a signal peptide, a EPOR-ligand binding domain, a fibronectin type 3 domain, a class I cytokine receptor
superfamily motif (WSXWS), a trans-membrane domain, and an
intracellular cytoplasmic signaling domain containing two motifs
termed Box 1 and Box 2, important for signal transduction
(Constantinescu et al., 1999). EPO binding induces dimerization
and reorientation of the cell surface EPORs, triggering activation
of the Janus family protein tyrosine kinase 2 (JAK2) which transphosphorylates several intracellular EPOR tyrosine residues
(Livnah et al., 1999; Remy et al., 1999), and leads to the activation
of many signaling proteins, including signal transducer and activator of transcription factor 5 (STAT5), the PI-3 kinase and the protein
tyrosine phosphatases SHP1 and SHP2 (Sasaki et al., 2000;
Socolovsky et al., 1999; Uddin et al., 2000; Witthuhn et al., 1993;
Richmond et al., 2005). These signaling pathways eventually lead
to the proliferation, differentiation, survival, and maturation of
erythrocyte progenitor cells. The expression of EPOR and signaling
after EPO/EPOR interaction has been reported in non-erythroid tissues such as brain (Liu et al., 1997), retina (Grimm et al., 2002),
192
F. Katakura et al. / Developmental and Comparative Immunology 45 (2014) 191–198
heart (Wu et al., 1999), myoblasts (Ogilvie et al., 2000), and vascular endothelium (Anagnostou et al., 1994).
Recently, we identified and characterized goldfish EPO and
found it to be functionally similar to its mammalian counterpart
(Katakura et al., 2013). Recombinant goldfish EPO (rgEPO) promoted the proliferation, differentiation and survival of erythrocyte
progenitor cells and up-regulated epor mRNA levels in progenitor
cells. Studies using zebrafish also have reported the conservation
of EPO/EPOR and STAT5 signal transduction in erythropoiesis (Paffett-Lugassy et al., 2007). Moreover, EPO and EPOR of Xenopus laevis
were also shown to be involved in erythropoiesis (Aizawa et al.,
2005; Nogawa-Kosaka et al., 2010). Interestingly, the EPO/EPOR
system appears to be well conserved throughout the evolution of
vertebrates, despite the fact that lower vertebrates such as fish,
amphibians, reptiles and birds, have nucleated erythrocytes while
mammals have enucleated erythrocytes.
In this study, we cloned and functionally characterized goldfish
EPOR. Goldfish epor mRNA expression was measured in goldfish
tissues, different cell subpopulations and in FACS-sorted kidney
progenitor cells. We also measured the expression of epor in cells
from the primary kidney macrophage (PKM) cultures treated with
rgEPO, to determine whether rgEPO signaling maintained erythroid lineage cells in vitro. Further, we examined whether the
addition of soluble recombinant goldfish EPOR (the extracellular
domain of EPOR) abrogated the rgEPO-induced proliferation of
the primary kidney progentior cells.
2. Materials and methods
2.1. Fish
Goldfish (Carassius auratus L.) were obtained from Aquatic Imports (Calgary, AB). Fish were maintained in tanks with a continuous flow water system at 20 °C and with a 14 h light/10 h dark
period in the aquatic facilities of Biological Sciences building at
the University of Alberta. Fish were fed daily and were acclimated
for at least 3 weeks prior to use in the experiments. Prior to handling, fish were sedated using a tricaine methane sulfonate (TMS,
syn MS-222) solution of 40–50 mg/L in water. The animals in the
Aquatic Facility were maintained according to the guidelines of
the Canadian Council of Animal Care (CCAC).
2.2. Isolation and establishment of goldfish primary kidney
macrophage (PKM) cultures
Goldfish (10–15 cm) were anesthetized with TMS, and killed.
The isolation and cultivation of goldfish kidney leukocytes in complete NMGFL-15 medium containing 5% carp serum and 10% newborn calf serum was performed as previously described (Neumann
et al., 1998, 2000). Cells were cultured at 20 °C. The primary kidney
macrophage (PKM) cultures consisted of heterogeneous populations of cells including early progenitors (R1 gated cells), monocytes (R3 gated cells) and mature macrophages (R2 gated cells)
as determined by flow cytometry, morphology, cytochemistry
and function (Neumann et al., 1998, 2000).
2.3. DNA sequencing and in silico analyses of goldfish epor
The initial partial sequence for goldfish epor was identified
using homology-based primers (Integrated DNA Technologies,
IDT) designed against corresponding carp sequences in the NCBI
database (AB671215). RACE PCR (BD Sciences, Clonetech) was performed to obtain a full open reading frame for epor. All amplicons
were gel purified using the QIA Gel Extraction kit (Qiagen) and
cloned into the TOPO TA pCR2.1 vector (Invitrogen). Colony PCR
was used to identify positive colonies using the vector specific
M13 forward and reverse primers, plasmids isolated using the QIAspin Miniprep kit (Qiagen) and sequenced using a BigDye terminator v3.1 cycle sequencing dye and a PE Applied Biosystems 377
automated sequencer. Single pass sequences were analyzed using
4peaks software (http://mekentosj.com/4peaks/) and sequences
aligned and analyzed using BLAST programs (http://blast.ncbi.nlm.nih.gov/Blast.cgi). The complete list of primers used for
homology-based PCR, RACE PCR, Q-PCR, sequencing and recombinant protein expression is shown in Supplemental Table 1.
EPOR protein sequences from fish, amphibian and mammals
were aligned using Clustal Omega software (http://www.ebi.ac.uk/Tools/msa/clustalo/). Signal peptide regions of respective
EPOR proteins were identified using the SignalP 4.0 Server
(http://www.cbs.dtu.dk/services/SignalP/) and conserved motifs
were predicted using the SMART server (http://smart.embl-heidelberg.de/). Phylogenetic analysis was conducted using Clustal X and
NJ-plot software using the neighbor joining method and bootstrapped 10,000 times, with values expressed as percentages Supplemental Fig. 1). The full-length sequence of goldfish EPOR has
been submitted to GenBank (KC595243).
2.4. Isolation of goldfish splenocytes, peripheral blood leukocytes and
kidney neutrophils
Splenocytes from individual goldfish (n = 4) were obtained by
gently homogenizing the spleen through a wire mesh screen using
NMGFL-15 medium containing heparin and penicillin/streptomycin (Pen/Strep). Cell suspensions were layered over 51% percoll
and centrifuged at 430g for 25 min. Cells at the interface were
collected and any residual red blood cells were lysed using red
blood cell lysis buffer (144 mM NH4Cl, 17 mM Tris, pH 7.2).
Remaining cells were washed twice with incomplete NMGFL-15
medium.
To isolate peripheral blood leukocytes, four goldfish were bled
from the caudal vein using a 25G needle and heparinized syringe
to prevent clotting. The blood was transferred into a capillary glass
tubes. After sealing one side with clay, the tubes were centrifuged
at 1500g for 5 min. Leukocytes were collected by cutting the
tubes 2 mm below the leukocyte layer and suspended in NMGFL15 medium. After centrifugation at 250g for 10 min to pellet
the leukocytes, the supernatant was discarded and residual red
blood cells were lysed using the red blood cell lysis buffer. Cells
were washed twice with incomplete NMGFL-15 to remove residual
red blood cell lysis buffer.
Kidney neutrophils were isolated and cultured overnight in
complete NMGFL-15 medium containing 5% carp serum and 10%
newborn calf serum at 20 °C, as described (Katzenback and Belosevic, 2009). Following overnight incubation to remove residual
contaminating adherent monocytes/macrophages, suspension of
cells containing neutrophils was collected and washed twice with
NMGFL-15.
2.5. Sorting of goldfish R1 progenitor cells
Freshly isolated leukocytes from goldfish kidney were re-suspended to a concentration of 5–10 106 cells/mL in complete
NMGFL-15 for cell sorting on a FACS Aria flow cytometer (Becton
Dickinson) in the Department of Medical Microbiology and Immunology, University of Alberta, Flow Cytometry Facility. R1 gated
cells, consisting of mainly heterogeneous early progenitors, were
sorted based on their small size and low internal complexity into
15 mL tubes containing 7 mL of complete NMGFL-15 (supplemented with serum) medium containing 100 U/mL of penicillin/
100 lg/mL of streptomycin and 100 lg/mL of gentamicin and then
washed twice with incomplete NMGFL-15.
F. Katakura et al. / Developmental and Comparative Immunology 45 (2014) 191–198
2.6. Quantitative PCR analysis of goldfish epor expression in goldfish
tissues
Brain, heart, gill, kidney, spleen, intestine, liver, muscle and
blood were harvested from six fish and RNA isolated using Trizol
(Invitrogen) according to the manufacturer’s specifications. Five
microgram of total RNA was reverse transcribed into cDNA using
the Superscript II cDNA synthesis kit (Invitrogen) according to
the manufacturer’s instructions. Quantitative expression analysis
of goldfish epor was performed using an Applied Biosciences
7500 Fast real time machine and elongation factor-1 alpha
(ef-1a) was employed as an endogenous control. Primers for all
target genes were designed with the Primer Express software
(Applied Biosystems) (Supplemental Table 1). Thermocycling conditions were 95 °C for 10 min followed by 40 cycles of 95 °C for 15 s
and 60 °C for 1 min. Data were analyzed using the 7500 fast software (Applied Biosciences) using the delta delta Ct method and
are shown as the mean ± SEM of six fish (n = 6).
2.7. Quantitative PCR analysis of goldfish epor mRNA expression in
different cell populations
Freshly isolated kidney leukocytes, sorted R1 progenitor cells,
splenocytes, peripheral blood leukocytes and kidney neutrophils
were obtained as described in Sections 2.2, 2.4, and 2.5. Each cell
population was pooled from 4–6 fish and RNA was isolated using
Trizol (Invitrogen). One microgram of RNA was reverse transcribed
into cDNA using the Superscript II cDNA synthesis kit (Invitrogen)
according to the manufacturer’s description. Quantitative PCR
expression analysis of goldfish epor was performed three different
times (n = 3) on an Applied Biosystems 7500 fast real time PCR system using SYBR green reagents and data analyzed using the delta
delta Ct method. Each bar is representative of the mean ± SEM from
three different experiments.
2.8. Quantitative PCR analysis of goldfish epor expression in PKM
cultures
Cells from four individual fish (n = 4) were seeded into each
well of 12 well plates at a concentration of 1 106 cells/mL in
complete NMGFL-15 medium in the presence or absence of
100 ng/mL of rgEPO. Cells were harvested at 0, 2, 4 and 8 days post
cultivation. Cell suspensions were centrifuged at 400g for 5 min
to pellet cells and total RNA was isolated using Trizol (Invitrogen).
Total RNA (1 lg) was reverse transcribed into cDNA using the
Superscript II cDNA synthesis kit (Invitrogen) according to the
manufacturer’s specifications. Quantitative expression of epor
was performed using the 7500 Fast Real Time PCR machine (Applied Biosciences) using the same primers and thermocycling conditions as described in Section 2.6.
2.9. Prokaryotic expression of recombinant goldfish EPOR
The extracellular domain (ECD) portion of the goldfish EPOR sequence was amplified from goldfish kidney cDNA by PCR using
primers designed to meet the requirements of the pET SUMO
expression vector (Invitrogen). The resulting PCR product was gel
purified (QIAquick gel extraction kit), ligated into the pET SUMO
vector, transformed into competent Escherichia coli (One Shot
Mach1-T1), plated onto LB-Kanamycin plates and incubated overnight at 37 °C. Positive clones were identified by colony PCR, grown
in 1 mL of LB-Kanamycin and the plasmids containing the inserts
were isolated from the bacteria using a QIAquick Spin miniprep
kit (Qiagen). To determine that inserts were in the correct orientation and in frame the purified plasmids were sequenced using the
pET SUMO vector specific primers, SUMO forward and T7 reverse
193
(Supplemental Table 1). A plasmid containing the in frame insert
of goldfish EPO was transformed into BL21 (DE3) One Shot E. coli
(Invitrogen), induced with 0.3 mM Isopropyl b-D-1-thiogalactopyranoside (IPTG) and the optimal induction time deduced in pilot
runs. The bacteria were scaled-up accordingly, pelleted and frozen
at 20 °C until purification of recombinant proteins.
2.10. Purification of recombinant goldfish EPOR and EPO
IPTG-induced, pelleted E. coli cultures expressing rgEPOR were
lysed in 50 mL of 1 FastBreak Cell Lysis Reagent (Promega), and
incubated with MagneHis Ni-particles (Promega). A PolyATrack
System 1000 magnet (Promega) was used to retain the Ni-particles
bound to the recombinant EPOR, the supernatants discarded and
the beads washed extensively with wash buffer (100 mM HEPES,
20 mM imidazole, pH 7.5). The recombinant protein was eluted
from the beads using elution buffer (100 mM Hepes, 500 mM imidazole, pH 7.5). The recombinant EPOR was passed through an
EndoTrap Red endotoxin removal column (Cambrex) to remove potential traces of endotoxin, then dialyzed against 1 PBS, filter
sterilized (0.22 lm) and stored at 4 °C until use. A Micro BCA assay
(Pierce) was performed according to manufactures directions to
determine protein concentration. The identity of the protein was
confirmed by mass spectrometry.
The production and purification of recombinant goldfish EPO
has been previously described (Katakura et al., 2013).
2.11. Purification of polyclonal rabbit anti-recombinant goldfish EPO
IgG
Anti-rgEPO IgG was produced by immunizing a rabbit with recombinant goldfish EPO. After testing for antibody titres, rabbits
were exsanguinated and the serum collected. The rabbit IgG was
purified from rabbit serum using ammonium sulfate precipitation
followed by protein A affinity chromatography as described previously (Katzenback and Belosevic, 2012). Affinity purified antibodies were filter sterilized (0.22 lm), and stored at 20 °C until use.
2.12. In vitro cross-linking studies
For all binding studies, 2.5 lg of each recombinant protein was
incubated in conjugation buffer (20 mM Hepes) for 2 h. The rgEPO
and rgEPOR were incubated individually or in ligand–receptor
combinations. Following the initial conjugation period, the ligands,
receptors and ligand–receptor combinations were cross-linked
using 5 mM disuccinimidyl suberate (DSS, Thermo Scientific) for
30 min before terminating the reaction for 15 min by the addition
of 50 mM Tris (final concentration). The reaction was then resolved
using reducing SDS–PAGE and Western blotting with the antipolyHistidine antibody (Sigma–Aldrich, 1:5000 dilution) as a primary antibody followed by the anti-mouse IgG-HRP conjugate
(Bio-Rad, 1:10,000 dilution) as a secondary antibody or anti-rgEPO
(1:1000 dilution) as a primary antibody followed by the anti-rabbit
IgG-HRP conjugate (Bio-Rad, 1:10,000 dilution) as a secondary
antibody. Blots were developed using ECL substrates (Pierce) and
exposure to X-ray film (Eastman Kodak Co.).
2.13. Measurement of rgEPO-induced cell proliferation abrogated by
rgEPOR
R1 gated cells, consisting of mainly heterogeneous early progenitor cells, were sorted based on size and internal complexity
using a FACS Aria flow cytometer (Becton Dickinson) from freshly
isolated PKM cultures (n = 8). Cells were adjusted to a concentration of 2 105 cells/mL in NMGFL-15 medium containing 10% calf
serum, 1% carp serum, 100 U/mL of penicillin/100 lg/mL of
194
F. Katakura et al. / Developmental and Comparative Immunology 45 (2014) 191–198
streptomycin, and 100 lg/mL of gentamicin. Fifty microliter of this
cell suspension was added to each well of a 96 well plate and
50 lL/well of treatment agent suspended in NMGFL-15. Cells were
treated with either NMGFL-15 (negative control), cell conditioned
medium (CCM, positive control 1), recombinant EPO alone at a final
concentration of 50 ng/mL (positive control 2), recombinant EPOR
alone at a final concentration of 100 ng/mL, and rgEPO pre-incubated at 4 °C for 12 h with 0.01, 0.1, 1, 10, and 100 ng/mL of rgEPOR. Cell conditioned medium (CCM) was cell culture
supernatants from previous PKM cultures. The CCM is a complex
mixture containing hematopoietic and myelopoietic growth factors including kit ligand (= stem cell factor), erythropoietin (EPO),
granulin and leukemia inhibitory factor, and macrophage colony
stimulating factor 1. For each fish, the measurements of proliferation were done in triplicate and the results presented as mean ± SEM of the replicates per fish for three individual fish. Cell
proliferation was determined using the BrdU colorimetric assay
(Roche). Briefly, 15 lM of BrdU labeling reagent was added to each
well 24 h prior to development and plates were kept at 20 °C. Cell
proliferation was determined on day 8, and plates developed
according to the manufacturer’s protocols. The plate was read at
an absorbance of 450 nm using a microplate reader (VersaMax).
Values obtained for the cells alone group in NMGFL-15 medium
were subtracted from values obtained for experimental groups to
control for spontaneous cell proliferation due to production of
endogenous growth factors.
Phylogenetic analysis of the predicted goldfish EPOR protein sequence placed it with other teleost EPOR sequences. Goldfish EPOR
was most similar to carp (Cyprinus carpio) EPOR (BAL42514) with
87% identity, followed by zebrafish (Danio rerio) EPOR
(NP_001036799) with 73% identity, and only 32% amino acid identity with human EPOR (NP_000112) (Supplemental Fig. 1).
3.2. Expression analysis of epor in goldfish tissues and different cell
populations
Assessment of epor expression in the tissues of normal goldfish
revealed the highest epor mRNA transcript levels in goldfish blood,
followed by spleen, heart, kidney, gill and brain (Fig. 2). Lower epor
mRNA levels were observed in the liver, intestine and muscle
(Fig. 2).
To identify the cell populations expressing epor mRNA, kidney
leukocytes, R1-gated progenitor cells, splenocytes, peripheral
blood leukocytes and kidney neutrophils were isolated and epor
expression measured using quantitative PCR. The highest epor
mRNA transcript levels were observed in PBLs, followed by sorted
progenitor cells and kidney leukocytes (Fig. 3). The epor expression
levels in sorted R1-gated progenitor cells were approximately 2–3
folds higher than those of kidney leukocytes (Fig. 3). Lower epor
mRNA levels were observed in splenocytes and neutrophils (Fig. 3).
2.14. Statistical analysis
3.3. Expression analysis of epor in the PKM cultures
Statistical analysis was performed using a one-way analysis of
variance (ANOVA) with a Dunnett’s post hoc test for analysis of
epor expression in cell populations (Fig. 3), PKM cultures over time
(Fig. 4) and for analysis of proliferation of progenitor cells treated
with both rgEPO and rgEPOR compared to those treated with rgEPO alone (Fig. 6). An unpaired t-test was performed for analysis of
epor expression in PKM cultures treated with rgEPO compared to
time-matched controls (Fig. 4). A probability of P < 0.05 was considered significant.
To examine the epor gene expression in the PKM cultures treated or non-treated with rgEPO, the mRNA levels of epor were measured in cells from PKM cultures on days 0, 2, 4, and 8 of
cultivation. The epor mRNA transcription levels in the cells from
PKM cultures without rgEPO decreased significantly after 4 days
of cultivation (Fig. 4). On the other hand, epor mRNA transcript levels in the cells from PKM cultures stimulated with rgEPO remained
high throughout the observation period (Fig. 4).
3. Results
3.4. In vitro analysis of rgEPO binding to rgEPOR
3.1. In silico analysis of goldfish erythropoietin receptor
In order to confirm the direct interaction between the identified
goldfish EPOR protein and its cognate ligand, EPO, we produced
rgEPO consisting of only an active mature form and rgEPOR containing a N-terminal 6His tag and the mature, extracellular domain of goldfish EPOR. The rgEPO and rgEPOR were then used to
perform in vitro cross-linking studies using the chemical cross-linker, DSS. As predicted, rgEPOR and rgEPO in the absence of the
cross-linker were observed as monomers at approximately
38 kDa and 17 kDa, respectively (Fig. 5A, lane 2 and B, lane 1). In
the cross-linked sample of rgEPOR, two distinct bands were observed representing both monomeric rgEPOR as well as the
homodimeric form at 76 kDa (Fig. 5A, lane 4). Cross-linking of
rgEPO resulted in the appearance of three different bands
(Fig. 5B, lane 3), presumably due to dimerization and trimerization
of rgEPO in part. In the sample containing cross-linked rgEPO and
rgEPOR, multiple bands were observed; non-bound rgEPO and rgEPOR were observed at the predicted sizes, prominent bands of
approximately 93 kDa and 55 kDa were also observed (Fig. 5A
and B, lane 5). Presumably the higher band consisted of a monomeric rgEPO and two rgEPOR molecules and the lower band likely
representing a monomeric rgEPO bound by a single rgEPOR. The
cross-linking of co-incubated rgEPOR and bovine serum albumin
did to show an interaction (results not shown).
The complete open reading frame (1566 bp) and the un-translated regions (UTR) (136 bp of 50 UTR and 345 bp of 30 UTR) of
the goldfish epor cDNA transcript were obtained. In silico analysis
of the predicted goldfish EPOR protein sequence identified functional domains characteristic of the type I cytokine receptor superfamily. The goldfish EPOR protein sequence included a signal
peptide sequence with a predicted cleavage site between amino
acid 24 and 25, as well as extracellular, transmembrane and cytoplasmic domains (Fig. 1). In the extracellular domain, there are
conserved 4 cysteine residues required for disulfide bonding
(Fig. 1). Present in the extracellular region was an EPOR ligand
binding domain, a fibronectin type 3 domain and a WSXWS motif
(Fig. 1). In the cytoplasmic region, the Box 1 motif, necessary for
binding and activation of JAK2, as well as the Box 2 motif are conserved (Fig. 1). The conserved 6 tyrosine residues at a distal portion
of the cytoplasmic domain are phosphorylated following EPO
stimulation and act as binding sites for downstream signaling molecules (Fig. 1) (Richmond et al., 2005). Furthermore, there are 6
amino acids (DTYVTL), which appear to be similar to mammalian
sequences (DTYLVL) that are believed to be a consensus site for
STAT5 binding (Fig. 1).
F. Katakura et al. / Developmental and Comparative Immunology 45 (2014) 191–198
195
Fig. 1. Alignment of vertebrate EPOR sequences. Goldfish (Carassius auratus L., KC595243), zebrafish (Danio rerio, NP_001036799), African clawed frog (Xenopus laevis,
NP_001089173), human (Homo sapiens, NP_000112) and mouse (Mus musculus, NP_034279) EPOR protein sequences were aligned using ClustalW. The signal peptide
cleavage site for goldfish EPOR is denoted with an arrowhead, conserved cysteine residues are highlighted in black, the WSXWS, Box1 and Box2 motifs are boxed, and
conserved tyrosine residues are highlighted in grey. The tyrosine residues that are believed to be a consensus site for STAT5 binding in mouse EPOR are boxed by broken line;
putative STAT5 binding sites in other species are likewise boxed. The predicted EPOR ligand binding domain is indicated by an overhead box with oblique lines, the predicted
fibronectin type 3 (FN3) domain is indicated by an overhead grey line, and the trans-membrane (TM) domain is marked by an overhead black line. Amino acids that are
conserved in all sequences are denoted with an asterisk (⁄), with high identity (:), and with weak identity (.).
3.5. Recombinant goldfish EPOR abrogated rgEPO ability to induce
proliferation of progenitor cells
In order to determine whether a soluble form of the goldfish
EPOR protein inhibited progenitor cell proliferation induced by
goldfish EPO, we produced both the goldfish EPO and the mature,
extra-cellular portion of the goldfish EPOR as recombinant proteins
(rgEPO and rgEPOR, respectively) in E. coli. Treatment of sorted
progenitor (R1-gated) cells with CCM or 50 ng/mL of rgEPO induced a high proliferative response (Fig. 6). In contrast, treatment
of sorted progenitor cells with rgEPOR alone did not enhance progenitor cell proliferation (Fig. 6). Treatment of progenitor cells with
a pre-incubated protein mixture of rgEPO and rgEPOR induced reduced proliferation of progenitor cells compared to non-treated
progenitor cells in a dose-dependent manner. Concentrations of
10 ng/mL or greater of rgEPOR were sufficient to significantly inhibit the proliferation of progenitors isolated from all the fish examined (Fig. 6).
4. Discussion
Erythrocytes are key players in oxygen transport to our whole
body, and the regulation of their development through EPO/EPOR
signaling is important in maintaining circulating erythrocyte numbers during homeostatic and hypoxic conditions. Despite the fact
that erythropoiesis is a fundamental system in vertebrates, the
molecules that regulate erythrocyte development in lower vertebrates have remained largely unexplored. Here, we report on the
identification and molecular characterization of the erythropoietin
receptor of goldfish.
The identified goldfish EPOR is 522 amino acids and is predicted
to have the EPOR_Ligand binding domain containing four conserved cysteine residues, the fibronectin type 3 domain, and the
class I cytokine receptor superfamily motif (WSXWS). Goldfish
EPOR also has the motifs required for intracellular signaling such
as the Box 1 and 2 motifs, and a number of tyrosine residues that
are important for recruitment and activation of JAK/STAT, PI3K/
AKT, and RAS/MAPK signaling pathways (Constantinescu et al.,
1999; Richmond et al., 2005) that mediate erythroid progenitor cell
survival, proliferation and differentiation. Indeed, studies in zebrafish have demonstrated the necessity and importance of EPOR in
erythropoiesis. Zebrafish embryos injected with epor morpholinos
displayed a partial block in primitive erythropoiesis and a complete block in definitive erythropoiesis leading to an absence of
erythrocytes by 4 days post fertilization (Paffett-Lugassy et al.,
2007). Furthermore, morpholino-mediated knockdown of stat5 in
zebrafish caused inhibition of erythropoietic expansion despite
overexpression of epo mRNA, indicating that STAT5 is required
for EPO signaling in teleosts (Paffett-Lugassy et al., 2007). Alignment of EPORs shows that a putative STAT5 binding site is conserved between goldfish and zebrafish and suggests that EPO
signaling through EPOR in goldfish likely involves JAK/STAT activation. Collectively, these in silico results suggest that EPO/EPOR are
critical for erythropoiesis in teleosts and the signaling pathways
of EPORs are evolutionally conserved in vertebrates, despite the
low sequence identity shared between EPOR proteins across
vertebrates.
196
F. Katakura et al. / Developmental and Comparative Immunology 45 (2014) 191–198
Fig. 2. Quantitative expression analysis of goldfish epor in tissues of normal
goldfish. The tissues examined were: brain, heart, gill, kidney, spleen, intestine,
liver, muscle and whole blood. The relative mRNA levels of goldfish epor were
determined using elongation factor-1 alpha (ef-1a) as the endogenous control. Data
were normalized against brain epor mRNA levels. The relative mRNA levels of epor
are presented as mean ± SEM (n = 6).
Fig. 4. Quantitative expression analysis of goldfish epor in goldfish primary kidney
macrophages treated with or without 100 ng/mL of rgEPO for 0, 2, 4, and 8 days. The
reported relative epor mRNA levels were calculated using ef-1a as an endogenous
control. The data were normalized against the mRNA levels at 0 day. Mean ± SEM
(n = 4). Significance (P < 0.05) compared to the day 0 group is denoted by (a) and
significance (P < 0.05) between the presence and absence of rgEPO groups is
denoted by (⁄).
Fig. 5. Western blot analysis of rgEPO binding to rgEPOR. All binding studies were
performed by incubating 2.5 lg of rgEPO, rgEPOR, or combinations of rgEPO and
rgEPOR in conjugation buffer (20 mM Hepes), cross-linking with 5 mM disuccinimidyl suberate (DSS, final concentration) and terminated by addition of 50 mM
Tris (final concentration). Lane: 1. rgEPO; 2. rgEPOR; 3. rgEPO + DSS; 4. rgEPOR + DSS; 5. rgEPO and rgEPOR + DSS. The reactions were resolved by SDS–PAGE
followed by Western blotting using an anti-poly-histidine antibody for the
detection of rgEPOR (A) and the anti-rgEPO antibody (B).
Fig. 3. Quantitative expression analyses of goldfish epor in different cell populations. The cell populations examined included: kidney leukocytes; R1-gated kidney
progenitor cell; splenocytes; peripheral blood leukocytes (PBL); and kidney-derived
neutrophils. Analysis of the relative cell population expression data were for pooled
cells from 4–6 fish, performed three different times (n = 3). All results were
normalized against the epor mRNA levels measured in splenocytes with ef-1a as the
endogenous control. The mean ± SEM is shown and distinct letters above each bar
denote statistically different groups (P < 0.05).
Not surprisingly, the highest expression of the goldfish epor
mRNA was observed in the whole blood, where numerous erythrocytes exist, followed by spleen and kidney tissues, the site of erythrocyte turnover and the major hematopoietic tissues of teleosts,
respectively. Moderate expression of goldfish epor was also observed in non-hematopoietic tissues such as the heart, gill, and
brain, which is similar to the expression profile of epor in zebrafish
(Paffett-Lugassy et al., 2007). The epor expression in other goldfish
tissues is consitant with reports that mammalian EPO/EPOR signaling plays roles not only erythropoiesis but also several non-hematopoietic functions including neurogenesis, neuroprotection,
wound healing, and cardiovascular protection (Arcasoy, 2008;
Marzo et al., 2008). It would be hard to deny a possibility that there
would be residual erythrocytes in these tissues and these cells
would contribute to the epor mRNA transcript levels observed.
However, one report shows that EPO is involved in fin regeneration
of sea bass (Dicentrarchus labrax) (Buemi et al., 2009), demonstrating a role for EPO/EPOR signaling in a non-hematopoietic context
F. Katakura et al. / Developmental and Comparative Immunology 45 (2014) 191–198
197
Fig. 6. Recombinant goldfish EPOR abrogated rgEPO-induced proliferation of progenitor cells. The proliferative response of sorted R1 progenitor cell cultures established from
eight individual fish were treated for 8 days with either cell conditioned medium (CCM), 50 ng/mL of rgEPO, 50 ng/mL of rgEPO pre-incubated with each concentration of
rgEPOR, or 100 ng/mL of rgEPOR alone. The inhibition of proliferation was observed in 6 out of 8 cultures examined. The representative proliferative response of three fish is
shown. The values for non-treated cells were subtracted from experimental values. Each bar represents the mean ± SEM of triplicate cultures. Significant differences
compared to the cells treated with rgEPO alone are denoted by (⁄).
in teleosts. Furthermore, although only one epor gene has been
identified in zebrafish, multiple epo genes/transcripts have been
identified suggesting that the various forms of zebrafish EPO may
also have functions outside of the hematopoietic system (Chu
et al., 2008). In light of these observations, future studies should
examine the novel non-hematopoietic functions of teleost EPO/
EPOR.
The mRNA and protein levels of the mammalian EPOR are maximized at early developmental stages, from colony-forming uniterythroid (CFU-E) to pronormoblasts, and gradually decrease in
maturing erythrocytes, thus diminishing the cell’s response to
EPO (Fisher, 2003; Krantz, 1991). The quantitative PCR analysis of
various cell populations revealed that epor mRNA levels in sorted
R1 progenitor cells were significantly higher than those in kidney
leukocytes, indicating that the cells expressing high levels of epor
are appear to be in the immature and or differentiating cell populations. Our current results are in accordance with our previous report that the R1-gated population of goldfish kidney leukocytes
contains cells having potential to form erythroid colonies following
rgEPO stimulation (approximately 1.2%) (Katakura et al., 2013).
The epor mRNA levels in splenocytes were significantly lower than
that of kidney leukocytes, which may indicate that goldfish spleen
is not the primary site of erythropoiesis. On the other hand, the
epor mRNA levels in PBLs were significantly higher compared to
those in the R1-sorted progenitor cells. However, upon performing
colony-forming assays using rgEPO, we could not detect CFU-E in
goldfish PBLs (unpublished data). Collectively, these results suggest that goldfish epor is expressed in erythroid precursor cells at
late developmental stage just before they differentiate into mature
erythrocytes. In Xenopus spp. epor mRNA was detected in peripheral blood cells while EPOR protein was absent in circulating mature
erythrocytes (Aizawa et al., 2005). Further studies are necessary to
understand molecular mechanisms of the terminal differentiation
of erythrocytes in lower vertebrates.
We have previously shown that the proliferation, differentiation, and survival of erythroid progenitor cells were promoted by
rgEPO (Katakura et al., 2013). Presumably, the action of rgEPO
was mediated via the EPOR, promoting the survival of erythroid
lineage cells in the PKM cultures. To examine this in more detail,
we measured the epor mRNA levels in the PKM cultures treated
with rgEPO. The epor mRNA levels in the PKM cultures without
rgEPO decreased with cultivation time and were significantly lower on day 4 of cultivation compared to the epor expression in day 0
PKM cultures. The decline in epor expression with cultivation time
was consistent with our previous findings showing down-regula-
tion in the expression of the key transcription factors, gata1 and
lmo2, involved in erythroid progenitor cell development (Katzenback et al., 2011). Together, the decline in epor, gata1 and lmo2 in
PKM cultures in the absence of rgEPO suggests the loss of erythroid
lineage cells. In contrast, the epor mRNA levels in the PKM cultures
treated with rgEPO remained elevated in PKM cultures during the
8 days of cultivation, indicating that epor mRNA expression may be
a good marker for the presence of erythroid lineage cells in fish cell
cultures. Based on these observations, we are currently in the process of generating antibody that recognizes EPOR, and predict that
that antibody would be an invaluable marker of developing erythroid progenitor cells.
Various soluble cytokine receptors have been shown to arise by
proteolytic digestion or alternative splicing and have neutralizing
effects toward their ligand’s functions (Aizawa et al., 2005; Heaney
and Golde, 1996; Nakamura et al., 1992; Novick et al., 1989). In the
murine models, alternative splicing of the epor transcript has
shown to encode for a soluble EPOR (sEPOR) (Barron et al., 1994;
Kuramochi et al., 1990; Nagao et al., 1992; Todokoro et al.,
1991). Our previous studies (Barreda et al., 2005) have demonstrated the presence of regulatory soluble receptors in teleosts;
for example a novel soluble form of the CSF-1R was present in
the goldfish and recombinant goldfish soluble CSF-1R, in a dosedependent manner inhibited cell proliferation induced by rgCSF1 (Hanington et al., 2007). Although we were unable to identify
epor transcripts that would suggest alternative splicing to generate
a sEPOR in goldfish, we wanted to assess whether a recombinant
goldfish EPOR (extracellular domain) could have neutralizing effects on its ligand, EPO. Indeed, the rgEPOR inhibited the rgEPO-induced proliferation of progenitor cells isolated from goldfish
kidney. These data demonstrate that goldfish EPO binds to the
extracellular domain of goldfish EPOR and that this region of the
receptor is required for the induction of central biological function
of EPO.
The in vitro binding studies further demonstrated the biological
interaction between goldfish EPO and goldfish EPOR. In mammals,
EPORs are thought to form dimers in their unbound state in a conformation that prevents activation of JAK2 and then undergo a ligand-induced conformational change that allows for activation of
intracellular signaling (Livnah et al., 1999; Remy et al., 1999).
Our findings for goldfish support this in that unbound goldfish
EPORs formed homodimers and that goldfish EPO bound to the
receptor. In addition, cross-linking of rgEPO alone resulted in the
appearance of three different bands, suggesting that rgEPO formed
monomer, dimer, and a trimer. There are reports that human EPO
198
F. Katakura et al. / Developmental and Comparative Immunology 45 (2014) 191–198
monomers, dimers and trimers can be produced artificially, and
that EPO dimers have an increased plasma half-life in vivo and
are biologically more active in vitro and in vivo compared to the
EPO monomers (Dalle et al., 2001; Sytkowski et al., 1998). Our results indicate that rgEPO bound to rgEPOR as a monomer, suggesting that monomeric EPO is a primary activator of the downstream
signaling via EPOR in goldfish.
In summary, we identified and characterized EPO receptor in
goldfish, demonstrating parallel expression, structure and function
to its mammalian homolog. It is interesting that vertebrates have
great diversity of respiration systems, oxygen transporters, and
erythrocytes (nucleated and non-nucleated), yet they use the
EPO/EPOR signaling as a common mechanism for erythropoiesis.
Acknowledgements
This research was supported by a Grant from the Natural Sciences and Engineering Research Council of Canada (NSERC) to
M.B., a Japan Society for Promotion of Science (JSPS) to F.K. and a
NSERC-CGS doctoral scholarship to B.A.K. We thank Dorothy Kratochwil-Otto from the Department of Medical Microbiology and
Immunology Flow Cytometry Facility at the University of Alberta
for technical assistance in cell sorting. We also thank Jing Zheng
from the Department of Chemistry Mass Spectrometry Facility at
University of Alberta, Canada for technical assistance in mass spectrometry analysis.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.dci.2014.02.017.
References
Aizawa, Y., Nogawa, N., Kosaka, N., Maeda, Y., Watanabe, T., Miyazaki, H., Kato, T.,
2005. Expression of erythropoietin receptor-like molecule in Xenopus laevis and
erythrocytopenia upon administration of its recombinant soluble form. J.
Biochem. 138, 167–175.
Anagnostou, A., Liu, Z., Steiner, M., Chin, K., Lee, E.S., Kessimian, N., Noguchi, C.T.,
1994. Erythropoietin receptor mRNA expression in human endothelial cells.
Proc. Natl. Acad. Sci. 91, 3974–3978.
Arcasoy, M.O., 2008. The non-haematopoietic biological effects of erythropoietin.
Br. J. Haematol. 141, 14–31.
Barreda, D.R., Hanington, P.C., Stafford, J.L., Belosevic, M., 2005. A novel soluble form
of the CSF-1 receptor inhibits proliferation of self-renewing macrophages of
goldfish (Carassius auratus L.). Dev. Comp. Immunol. 29, 879–894.
Barron, C., Migliaccio, A.R., Migliaccio, G., Jiang, Y., Adamson, J.W., Ottolenghi, S.,
1994. Alternatively spliced mRNAs encoding soluble isoforms of the
erythropoietin receptor in murine cell lines and bone marrow. Gene 147,
263–268.
Buemi, M., Lacquaniti, A., Bolignano, D., Maricchiolo, G., Favaloro, A., Buemi, A.,
Grasso, G., Donato, V., Giorgianni, G., Genovese, L., 2009. The erythropoietin and
regenerative medicine: a lesson from fish. Eur. J. Clin. Invest. 39, 993–999.
Chu, C.-Y., Cheng, C.-H., Yang, C.-H., Huang, C.-J., 2008. Erythropoietins from
teleosts. Cell. Mol. Life Sci. 65, 3545–3552.
Constantinescu, S.N., Ghaffari, S., Lodish, H.F., 1999. The erythropoietin receptor:
structure, activation and intracellular signal transduction. Trends Endocrinol.
Metab. 10, 18–23.
Dalle, B., Henri, A., Rouyer-Fessard, P., Bettan, M., Scherman, D., Beuzard, Y., Payen,
E., 2001. Dimeric erythropoietin fusion protein with enhanced erythropoietic
activity in vitro and in vivo. Blood 97, 3776–3782.
Fisher, J.W., 2003. Erythropoietin: physiology and pharmacology update. Exp. Biol.
Med. 228, 1–14.
Grimm, C., Wenzel, A., Groszer, M., Mayser, H., Seeliger, M., Samardzija, M., Bauer,
C., Gassmann, M., Remé, C.E., 2002. HIF-1-induced erythropoietin in the hypoxic
retina protects against light-induced retinal degeneration. Nat. Med. 8, 718–
724.
Hanington, P.C., Wang, T., Secombes, C.J., Belosevic, M., 2007. Growth factors of
lower vertebrates: characterization of goldfish (Carassius auratus L.)
macrophage colony-stimulating factor-1. J. Biol. Chem. 282, 31865–31872.
Heaney, M.L., Golde, D.W., 1996. Soluble cytokine receptors. Blood 87, 847–857.
Huising, M.O., Kruiswijk, C.P., Flik, G., 2006. Phylogeny and evolution of class-I
helical cytokines. J. Endocrinol. 189, 1–25.
Katakura, F., Katzenback, B.A., Belosevic, M., 2013. Molecular and functional
characterization of erythropoietin of the goldfish (Carassius auratus L.). Dev.
Comp. Immunol..
Katzenback, B.A., Belosevic, M., 2009. Isolation and functional characterization of
neutrophil-like cells, from goldfish (Carassius auratus L.) kidney. Dev. Comp.
Immunol. 33, 601–611.
Katzenback, B.A., Belosevic, M., 2012. Colony-stimulating factor-1 receptor protein
expression is a specific marker for goldfish (Carassius auratus L.) macrophage
progenitors and their differentiated cell types. Fish Shellfish Immunol. 32, 434–
445.
Katzenback, B.A., Karpman, M., Belosevic, M., 2011. Distribution and expression
analysis of transcription factors in tissues and progenitor cell populations of the
goldfish (Carassius auratus L.) in response to growth factors and pathogens. Mol.
Immunol. 48, 1224–1235.
Krantz, S.B., 1991. Erythropoietin. Blood 77, 419–434.
Kuramochi, S., Ikawa, Y., Todokorof, K., 1990. Characterization of murine
erythropoietin receptor genes. J. Mol. Biol. 216, 567–575.
Liu, C., Shen, K., Liu, Z., Noguchi, C.T., 1997. Regulated human erythropoietin
receptor expression in mouse brain. J. Biol. Chem. 272, 32395–32400.
Livnah, O., Stura, E.A., Middleton, S.A., Johnson, D.L., Jolliffe, L.K., Wilson, I.A., 1999.
Crystallographic evidence for preformed dimers of erythropoietin receptor
before ligand activation. Science 283, 987–990.
Marzo, F., Lavorgna, A., Coluzzi, G., Santucci, E., Tarantino, F., Rio, T., Conti, E., Autore,
C., Agati, L., Andreotti, F., 2008. Erythropoietin in heart and vessels: focus on
transcription and signalling pathways. J. Thromb. Thrombolysis 26, 183–187.
Nagao, M., Masuda, S., Abe, S., Ueda, M., Sasaki, R., 1992. Production and ligandbinding characteristics of the soluble form of murine erythropoietin receptor.
Biochem. Biophys. Res. Commun. 188, 888–897.
Nakamura, Y., Komatsu, N., Nakauchi, H., 1992. A truncated erythropoietin receptor
that fails to prevent programmed cell death of erythroid cells. Science 257,
1138–1141.
Neumann, N.F., Barreda, D., Belosevic, M., 1998. Production of a macrophage growth
factor (s) by a goldfish macrophage cell line and macrophages derived from
goldfish kidney leukocytes. Dev. Comp. Immunol. 22, 417–432.
Neumann, N.F., Barreda, D.R., Belosevic, M., 2000. Generation and functional
analysis of distinct macrophage sub-populations from goldfish (Carassius
auratus L.) kidney leukocyte cultures. Fish Shellfish Immunol. 10, 1–20.
Nogawa-Kosaka, N., Hirose, T., Kosaka, N., Aizawa, Y., Nagasawa, K., Uehara, N.,
Miyazaki, H., Komatsu, N., Kato, T., 2010. Structural and biological properties of
erythropoietin in Xenopus laevis. Exp. Hematol. 38, 363–372.
Novick, D., Engelmann, H., Wallach, D., Rubinstein, M., 1989. Soluble cytokine
receptors are present in normal human urine. J. Exp. Med. 170, 1409–1414.
Ogilvie, M., Yu, X., Nicolas-Metral, V., Pulido, S.M., Liu, C., Ruegg, U.T., Noguchi, C.T.,
2000. Erythropoietin stimulates proliferation and interferes with differentiation
of myoblasts. J. Biol. Chem. 275, 39754–39761.
Paffett-Lugassy, N., Hsia, N., Fraenkel, P.G., Paw, B., Leshinsky, I., Barut, B., Bahary, N.,
Caro, J., Handin, R., Zon, L.I., 2007. Functional conservation of erythropoietin
signaling in zebrafish. Blood 110, 2718–2726.
Remy, I., Wilson, I.A., Michnick, S.W., 1999. Erythropoietin receptor activation by a
ligand-induced conformation change. Science 283, 990–993.
Richmond, T.D., Chohan, M., Barber, D.L., 2005. Turning cells red: signal
transduction mediated by erythropoietin. Trends Cell Biol. 15, 146–155.
Sasaki, A., Yasukawa, H., Shouda, T., Kitamura, T., Dikic, I., Yoshimura, A., 2000. CIS3/
SOCS-3 suppresses erythropoietin (EPO) signaling by binding the EPO receptor
and JAK2. J. Biol. Chem. 275, 29338–29347.
Shemin, D., Rittenberg, D., 1946. The life span of the human red blood cell. J. Biol.
Chem. 166, 627–636.
Socolovsky, M., Fallon, A.E., Wang, S., Brugnara, C., Lodish, H.F., 1999. Fetal anemia
and apoptosis of red cell progenitors in Stat5a/5b/ mice: A direct role for
Stat5 in Bcl-XL induction. Cell 98, 181–191.
Sytkowski, A.J., Lunn, E.D., Davis, K.L., Feldman, L., Siekman, S., 1998. Human
erythropoietin dimers with markedly enhanced in vivo activity. Proc. Natl. Acad.
Sci. 95, 1184–1188.
Todokoro, K., Kuramochi, S., Nagasawa, T., Abe, T., Ikawa, Y., 1991. Isolation of a
cDNA encoding a potential soluble receptor for human erythropoietin. Gene
106, 283–284.
Uddin, S., Kottegoda, S., Stigger, D., Platanias, L.C., Wickrema, A., 2000. Activation of
the Akt/FKHRL1 pathway mediates the antiapoptotic effects of erythropoietin in
primary human erythroid progenitors. Biochem. Biophys. Res. Commun. 275,
16–19.
Witthuhn, B.A., Quelle, F.W., Silvennoinen, O., Yi, T., Tang, B., Miura, O., Ihle, J.N.,
1993. JAK2 associates with the erythropoietin receptor and is tyrosine
phosphorylated and activated following stimulation with erythropoietin. Cell
74, 227–236.
Wu, H., Lee, S.H., Gao, J., Liu, X., Iruela-Arispe, M.L., 1999. Inactivation of
erythropoietin leads to defects in cardiac morphogenesis. Development 126,
3597–3605.