The erythropoietin receptor (EPOR) and its cognate ligand, erythropoietin (EPO), are required for maintaining adequate levels of circulating erythrocytes during embryogenesis and adulthood by promoting erythroid mitogenesis, survival, and differentiation. Before the cloning of the EPOR cDNA, radiolabeled EPO was used to demonstrate specific binding to normal erythroid progenitors, murine and human erythroleukemia cells, and cells from human fetal liver. Then in 1989, the EPOR cDNA has been cloned by transfection of recombinant plasmid pools of murine cDNA from murine erythroleukemia (MEL) cell line into COS cells. A single cDNA was isolated that confers to COS cells the ability to bind EPO (D’Andrea et al. 1989). This murine EPOR cDNA, expressed in COS cells, generates high-affinity (30 pM) and low-affinity (210 pM) receptors. In 1990, the human homologue of the murine EPOR has been isolated from an erythroleukemia line and from fetal liver (Jones et al. 1990). Both cDNA and protein sequences of the human receptor are 82% similar to the sequences of the murine receptor. Scatchard analysis of the EPO-EPOR binding revealed the presence of two receptor species having apparent dissociation constants of 30 pM and 210 pM. Also, it has been shown that approximately 200 EPORs are present on the surface of normal erythroid progenitor cells (CFUe) in the bone marrow and that this number increases to about 1000 per cell on certain cell lines. Since then, the EPOR gene has been cloned from 126 organisms.
Gene and Protein Structures
In Xenopus laevis, the identities of the entire deduced amino acid sequences with human and murine EPOR are 33.3% and 34.2%, respectively. The sequence reveals the cysteine residues, the WSXWS motif, and two additional cysteine residues. In the cytoplasmic domain, Box-1/2 motifs are conserved. No tyrosine kinase catalytic domain exists whereas seven tyrosine residues are present. Interestingly, after the Box-2 motif, there is an insertion of 55 residues including one tyrosine that have no similar sequence related to mammalian EPOR.
Sequence identity between zebrafish EPOR and other vertebrates is relatively low: Fugu rubripes, 44%; Tetraodon nigroviridis, 41%; Xenopus laevis, 22%; Xenopus tropicalis, 22%; Mus musculus, 26%; and Homo sapiens, 27%. Despite this, the zebrafish EPOR retains key structural features. In the extracellular domain, it possesses the four highly conserved cysteine residues and the “WSxWS” domain. In the cytoplasmic domain, there are the conserved Box-1 and Box-2 domains, as well as five of eight tyrosine residues (Paffett-Lugassy et al. 2007).
Distribution of the Erythropoietin Receptor
In mouse embryos, EPOR expression begins at day 7.5 in both endothelial and primitive blood cells. An increased expression is found in the yolk sac vasculature and vitelline vessels. In Xenopus laevis, in situ hybridization studies revealed that EPOR is expressed in the ventral blood island of the developing tadpole, the site of primitive hematopoiesis, the equivalent to the mammalian yolk sac. Later, EPOR transcripts are detected in circulating blood (Yergeau et al. 2006). During zebrafish development, EPOR transcripts are detected in presumptive hematopoietic precursors, then in circulating erythrocytes, and finally only in the brain.
Erythropoietin Receptor and Signaling Pathways
The interaction between EPOR and JAK2 is initiated in the endoplasmic reticulum and promotes maturation of the receptor as well as its addressing to cell surface. Homodimerization of EPOR in response to EPO induces conformational changes in the extracellular domain of the receptor that subsequently allows the activation of two JAK2 molecules associated with the Box-1/2 motifs. JAK2 molecules are activated by autophosphorylation and in turn phosphorylate eight tyrosine residues in the cytoplasmic domain (Kubatzky et al. 2005). They serve as docking sites for several proteins, which subsequently activate the above signaling pathways (Fig. 4). The signal transducer and activator of transcription 5 (STAT5) is recruited to phosphorylated tyrosine residues at Y343 and Y401 through its Src homology 2 (SH2) domain. The STAT3 is recruited through the phosphorylated Y431. STAT3 and STAT5 are both phosphorylated, and then they homodimerize and translocate to the nucleus where they act as transcription factors. The G-protein RAS, bounded to phosphorylated tyrosine by several adaptor proteins, acts as a switch, activating both the rat sarcoma/mitogen-activated protein kinase/extracellular signal-regulated kinase (RAS/MEK/ERK) and the phosphatidylinositol 3-kinase/serine/threonine kinase (PI3K/Akt) pathways. The RAF protein kinase activates sequential phosphorylation of MEK/ERK whose ERK, the last element of the cascade, translocates into the nucleus. It activates cofactors such as the ETS transcription factor ELK-1, to transcribe target genes. The RAF/MEK/ERK pathway also is activated by protein kinase C (PKC) via phosphorylation of RAF or MEK. EPOR cytoplasmic domain also associates with the growth factor receptor protein 2 (GRB2) and the p85 α subunit of phosphatidylinositol 3 kinase (PI3K) via phosphorylated tyrosine residues at Y464 and Y429/431 or Y479, respectively. GRB2 induces the activation of ERK through the MEK pathway. The p85 regulatory subunit of PI3K induces the activation of Akt through the phosphorylation of phosphatidylinositol resulting in the activation of several transcription, translation, and regulation factors. Downstream genes, mainly BCL-xL, BCL-3, cyclin D1, c-MYC, and c-FOS, whose transcription and translation are stimulated by these pathways, impact biological functions of EPOR such as proliferation, differentiation, regulation of apoptosis, and drug resistance. In addition, Y464 and/or Y479 recruits the SH2 domain of the Src tyrosine kinase Lyn. In turn, Lyn phosphorylates the CRKL adapter protein, recruited on Y460, resulting in the augmentation of MAPK activation.
In contrast to phosphorylation that activates EPOR signaling, dephosphorylation tightly downregulates EPOR activities. The SH2 domain-containing tyrosine phosphatase 1 (SHP1) binds phosphorylated Y429 and Y431 and subsequently dephosphorylates JAK2, thereby inactivating the kinase and downregulating the signaling cascade. The cytokine-inducible SH2-domain-containing protein (CIS) interacts with phosphorylated Y401. Suppressor of cytokine signaling 3 (SOCS3) interacts with phosphorylated Y401 or Y429 and Y431. CIS and SOCS3 both inhibit the activation of pathways mediated by STAT5 and ERK. Interestingly, the transcription of SOCS3 and CIS is induced by EPO, thus allowing a negative feedback loop to regulate EPOR biological functions. Also, it has been suggested that CIS may promote EPOR degradation leading to EPOR downregulation. Two other pathways of EPO-induced EPOR downregulation have been proposed. Ubiquitination of EPOR at Lys256 provokes efficient EPO-induced receptor internalization, and ubiquitination at Lys428 promotes trafficking of EPOR to the lysosomes for degradation. Therefore, ubiquitination controls EPOR downregulation, downstream signaling, and its biological functions.
EPOR is mainly involved in proliferation, differentiation, and regulation of apoptosis during erythropoiesis. Erythropoiesis is a central feature of vertebrate development. Erythroid cells in vertebrates come from primitive erythropoiesis in the embryonic yolk sac or its equivalent and definitive erythropoiesis in the yolk sac or the aorta–gonad–mesonephros region, shift to the kidney, spleen, liver, or bone marrow, and last for the life span of the organism. The role of EPOR was first demonstrated in mice that do not express the EPOR. These mice have fewer primitive erythrocytes in the yolk sac blood islands and die between day 13 and 15 of gestation, owing to failure of definitive fetal liver erythropoiesis. By culturing fetal livers from embryos, the authors show that BFU-E and CFU-E progenitors were present suggesting that EPORs are not required for erythroid lineage commitment or for the proliferation and differentiation of BFU-E to CFU-E progenitors. The results reveal an essential role for EPOR in regulating definitive erythropoiesis by controlling processes such as proliferation, survival, and irreversible terminal differentiation of the late CFU-E progenitors, resulting in the formation of mature circulating red blood cells. Detailed examination of EPOR-null mouse embryos also show cardiac ventricular hypoplasia and increased apoptosis in the myocardium and brain. It has been reported that EPOR signaling regulates neural progenitor cell (NPC) differentiation and that EPOR expression levels is downregulated as NPCs terminally differentiated into mature neurons (Chen et al. 2007). Interestingly, conditional EPOR deletion in the brain leads to reduced cell proliferation in the subventricular zone where in vivo neurogenesis takes place in adult mice (Tsai et al. 2006).
In zebrafish, knockdown of the EPOR causes a decrease in primitive erythropoiesis and a complete block of definitive erythropoiesis (Paffett-Lugassy et al. 2007).
In adults, the main function of the EPO is the regulation of erythropoiesis by binding and activating EPOR on the surface of erythroid progenitor and precursor cells in the bone marrow. The responses of hematopoietic cells to EPO depend, at least partly, on the extent of EPOR expression. EPOR expresses at low levels on early erythroid progenitor cells (i.e., BFU-E); its expression then increases during erythroid differentiation by the CFU-E stage and is downregulated during late erythropoiesis, so that reticulocytes and mature red blood cells do not express EPOR.
The discovery of EPOR expression in a number of non-hematopoietic tissues suggests that the effects of EPOR activation extend beyond regulation of erythropoiesis and provides support for the various biological functions of EPO-EPOR. EPOR signaling promotes endothelial cell proliferation, tissue protection, and especially amelioration of neuronal recovery from injury (Brines and Cerami 2006). It also promotes endothelial cell and smooth muscle cell protection and cardiomyocyte survival after ischemia–reperfusion injury. These cytoprotective effects of EPOR signaling appear to require nanomolar concentrations of EPO that are not normally reached in plasma, whereas only low picomolar concentrations are required for erythropoiesis. Such differential binding affinities allow specific activation of erythroid and non-erythroid EPOR, preventing cross talk between the erythropoietic and cytoprotective effects of EPO.
EPOR signaling has protective effects on cardiac as well as vascular tissues, by either preventing apoptosis of cardiac myocytes, smooth muscle cells, and endothelial cells or by increasing endothelial production of nitric oxide. Accumulating experimental evidence suggests that EPOR signaling contributes to cardiomyocyte survival after ischemia-reperfusion injury. The EPO-induced cardioprotection was at least partly related to apoptosis inhibition and augmented cardiomyocyte survival, mediated by activating EPOR and/or EPOR/βCR complex (Xu et al. 2009). Moreover, numerous studies provided the evidence of EPOR localization in endothelial cells. It was therefore suggested that EPOR signaling was involved in increased angiogenesis in vascular system and survival of endothelial cells (Teng et al. 2011). Overall current evidence suggests that the vascular protective effects of EPOR signaling are dependent on nitric oxide (NO) production through increased of endothelial nitric oxide synthase (eNOS) phosphorylation via activation of the EPOR in endothelial cells, mediated by JAK2 activation and then PI3K and Akt phosphorylation (Satoh et al. 2006). This EPOR signaling increase in NO production improves the endothelium-dependent relaxation of arteries and then contributes to prevent cardiovascular disorders.
Because EPOR expression persists in the human brain after birth and throughout adulthood, EPOR signaling has been suggested to play a specific role in the brain. Using different animal models, it has been shown that exogenous EPO administration attenuated brain damage related to several tissue injuries during hypoxia/ischemia, after brain trauma, and neurotoxic and excitotoxic insults (Ogunshola and Bogdanova 2013). After binding with EPOR and/or EPOR/βCR, the cerebrovascular protective effects of EPOR activation involve several signaling pathways that contribute to explain the decreased excitatory amino acid release, inhibition of apoptosis, and stimulation of neurogenesis and angiogenesis. These cerebrovascular protective effects of EPOR signaling may appear to be mediated by NO production via activation of the Akt/eNOS signaling pathway and by inhibition of inducible nitric oxide synthase (iNOS) activity, preventing the formation of NO excess. Moreover, previous studies demonstrated the crucial role of endogenous EPO/EPOR on the mobilization, recruitment, and activation of endothelial progenitor cells, which in turn may contribute to vascular repair and neovascularization (Heeschen et al. 2003).
Erythropoietin Receptor and Pathologies
Mutations in the EPOR result in primary familial and congenital polycythemia (PFCP). PFCP, also known as familial erythrocytosis, is an inherited hematological disorder affecting bone marrow progenitor cells of the erythroid lineage. It is characterized by elevated hematocrit and hemoglobin levels or an increased red cell mass. It is also characterized by an hypersensitivity of erythroid progenitors to EPO and low serum levels of EPO. It is caused by uncontrolled red blood cell production. The phenotype is associated with normal leukocyte and thrombocyte counts, absence of splenomegaly, and lack of progression to clonal bone marrow disorders and leukemia (Bento et al. 2016). Actually, more than 22 EPOR mutations associated with PFCP phenotype have been described within exon 8 which encodes the C-terminal domain of the protein. Eighteen are frameshift mutations leading to a truncation of 59–84 amino acids in the C-terminal region (Bento et al. 2016). Also, in a kindred with autosomal dominantly inherited familial erythrocytosis, a nonsense mutation has been reported. It affects codon 399 in exon 8 leading to an EPOR peptide that is truncated by 110 amino acids at its C-terminal region (Arcasoy et al. 2003). Other EPOR mutations in PFCP have been found to be generated by insertion, deletion, or duplication (see Table 2 in Huang et al. 2010).
Several studies have reported EPOR expression in primary cancers as well as in tumor cell lines such as non-small cell lung cancer, head and neck squamous cell carcinoma, renal cell carcinoma, breast, prostate and gastric carcinomas, solid tumors of nervous system, tumors of female reproductive organs, and malignant melanoma (see Table 1 in Hardee et al. 2006). EPOR signaling may contribute to the growth and survival of these cells and not only enhance but also attenuate the resistance of cancer cells to different therapies. This is why today the role of EPOR signaling in tumor cells remains controversial (Szenajch et al. 2010).
In acute lymphoblastic leukemia, where chromosomal rearrangements are a hallmark, four different rearrangements of the EPOR have been described. They result in its cytoplasmic-tail truncation at residues similar to those mutated in PFCP. These rearrangements provoke the stabilization of EPOR and its hypersensitivity to EPO and heighten the activation of STAT5/3 signaling (Iacobucci et al. 2016).
Animal Models of Human Erythropoietin Receptor Pathologies
A mouse model of PFCP has been generated by replacing the murine EPOR with a truncated human EPOR that has been identified in PFCP patients. Within 3–6 weeks of age, the mice develop polycythemia. They show increased hematocrit and elevated hemoglobin concentrations, thus mimicking the human phenotype. This animal model can be used to investigate EPOR function in erythroid tissue as well as in non-erythropoietic tissues that express EPOR.
The EPOR is a member of the cytokine receptor family. The main role of EPOR and its ligand EPO is to maintain adequate levels of circulating erythrocytes during embryogenesis and adulthood inducing erythroid proliferation, differentiation, and survival. The EPOR gene has been cloned and mapped in humans on chromosome 19p13.2. EPOR has a cytoplasmic domain, a single transmembrane segment, and an extracellular region. As in cytokine receptors, the intracellular domain possesses the proline rich Box-1 motif, which together with the less-conserved distal Box-2 containing aromatic and acidic residues serves to bind the JAK2 proteins close to the cell membrane. The transmembrane part of the protein is responsible for receptor homodimerization in response to EPO. The homodimer allows the activation by autophosphorylation of JAK2 and by consequence the phosphorylation of eight tyrosines in the cytoplasmic domain. They constitute docking sites for several cytoplasmic proteins and activate signaling pathways. EPOR is present in hematopoietic tissues of embryos and adults where it promotes proliferation, differentiation, and regulation of apoptosis during erythropoiesis. In adults, EPOR is also expressed on numerous non-erythropoietic cells, associated with βCR or epinephrine B4 receptors to form heterodimer/trimers named NEPOR. The signaling pathways they generate, though still not well understood, lead to protective effects for tissues which express them. Mutations in human EPOR and chromosomal rearrangements are responsible of pathologies such as erythrocytosis and lymphoblastic leukemia. EPOR gene expression is widespread in both human cancer cell lines and in primary tumors. Although evidence from many experiments supports an effect of EPOR signaling on cell growth enhancement in tumor cell lines, further investigation is necessary to clarify the relationship among functional EPOR expression and EPOR signaling pathways in cell growth and survival in human tumors.
- Bento C, McMullin MF, Percy M, Cario H. Primary familial and congenital polycythemia. In: Pagon RA, Adam MP, Ardinger HH, Wallace SE, Amemiya A, LJH B, et al., editors. GeneReviews. Seattle: University of Washington; 2016.Google Scholar
- Kubatzky KF, Moucadel V, Constentinescu S. Signal transduction of the erythropoietin receptor: role of the transmembrane region. Hématologie. 2005;11:19–28.Google Scholar