Cardiovascular Effects of Erythropoietin: An Update

Abstract

Erythropoietin (EPO) is a therapeutic product of recombinant DNA technology and it has been in clinical use as stimulator of erythropoiesis over the last two decades. Identification of EPO and its receptor (EPOR) in the cardiovascular system expanded understanding of physiological and pathophysiological role of EPO. In experimental models of cardiovascular and cerebrovascular disorders, EPO exerts protection by either preventing apoptosis of cardiac myocytes, smooth muscle cells, endothelial cells, or by increasing endothelial production of nitric oxide. In addition, EPO stimulates mobilization of progenitor cells from bone marrow thereby accelerating repair of injured endothelium and neovascularization. A novel signal transduction pathway involving EPOR - β-common heteroreceptor is postulated to enhance EPO-mediated tissue protection. A better understanding of the role of β-common receptor signaling as well as development of novel analogues of EPO with enhanced non-hematopoietic protective effects may expand clinical application of EPO in prevention and treatment of cardiovascular and cerebrovascular disorders.

I. Introduction

The humoral regulation of erythropoiesis was identified more than 50 years ago (Bonsdorff and Jalavisto, 1948). Studies on erythropoietin (EPO) have been severely restricted, due to its circulating levels being only picomolar, until EPO was purified for the first time in 1977 from human urine (Miyake et al., 1977) and the amino acid sequence of human urinary EPO was characterized (Lai et al., 1986). EPO is a member of the class I family of cytokines and is mainly synthesized and secreted by the kidney (Jelkmann, 1992). The ability of EPO to stimulate erythropoiesis expanded clinical application of recombinant EPO to patients suffering from anemia and chronic kidney disease. To date, recombinant EPO is the most successful biotechnology drug and indeed is the world’s largest selling biopharmaceutical. EPO mediates erythropoiesis by binding to its specific receptor, EPO-receptor (EPOR), expressed on the surface of immature erythroblasts (Noguchi et al., 1991). Studies over the last decade have demonstrated that EPO and EPOR are expressed in a number of cell types including those within the cardiovascular and nervous system, suggesting that the effects of EPO extend beyond regulation of erythropoiesis.

In the cardiovascular system, EPO exerts its effects on cardiac as well as the vascular tissues. EPO and EPOR are expressed in cardiomyocytes, vascular endothelial cells and smooth muscle cells (Anagnostou et al., 1994; Ammarguellat et al., 1996; Brines and Cerami, 2005). It is important to note that high doses are required to observe EPO-induced tissue protection, in comparison to doses of EPO required for hematopoietic effects (Anagnostou et al., 1994; Ammarguellat et al., 1996; Brines and Cerami, 2005). Based on the existing studies with these high doses, a tangential paradigm evolved. On one hand, protective effects of EPO and the underlying mechanisms were elucidated in cardiovascular and neurological system, prompting initiation of clinical trials in patients with myocardial infarction, aneurysmal subarachnoid hemorrhage and acute stroke. In contrast, adverse events associated with EPO therapy were identified, due to its pleiotropic effects mainly in the cardiovascular system including hypertension, thrombosis and augmented tumour angiogenesis. In this review, the present understanding of the beneficial and detrimental cardiovascular effects of EPO will be discussed.

II. Pharmacokinetics of Erythropoietin

A. Recombinant Human Erythropoietin

EPO was produced in Chinese hamster ovary cells by recombinant DNA technology (Sasaki et al., 1987; Sasaki et al., 1988). Following isolation of the human erythropoietin gene (Jacobs et al., 1985; Lin et al., 1985), it was inserted into and expressed by cultured mammalian cells, which are capable of synthesizing unlimited quantities of the hormone. Epoetin alfa, epoetin beta, and epoetin gamma are analogues of recombinant human EPO (rhEPO) derived from a cloned human erythropoietin gene, licensed by international regulatory bodies to stimulate erythropoiesis. All have the same 165 amino acid sequence with a molecular weight of 30,400 daltons and have the same pharmacological actions as native EPO (Jelkmann, 1992). After intravenous administration, epoetin is distributed in a volume comparable to the plasma volume. Epoetin plasma concentrations decay at a much slower rate after subcutaneous than intravenous administration (Table 1). Between epoetin alfa and epoetin beta exhibit some differences in their pharmacokinetic profiles, such as elimination half-life, due to differences in glycosylation pattern (Table 1) (Storring et al., 1998). Biological activity of EPO in vivo is abolished when EPO is deglycosylated, suggesting that the carbohydrate moiety is essential to prevent degradation and to delay clearance of EPO from the circulation (Takeuchi and Kobata, 1991).

Table 1.

Molecular characterization and elimination half-live of EPO and its analogues.

Drug	Molecule structure			Route of administration		References
	Molecular weight	Glycosylation type	Number of sialic acid	Intravenous	Subcutaneous
Epoetin alpha	30,400 daltons	3 N-linked oligosaccharide chains	10-14	4-8 hours	24 hours	(McMahon et al., 1990; Halstenson et al., 1991; Macdougall et al., 1991)
Epoetin beta	30,400 daltons	3 N-linked oligosaccharide chains	10-14	4-10 hours	13-28 hours	(McMahon et al., 1990; Halstenson et al., 1991; Macdougall et al., 1991)
Darbepoetin alpha (NESP)	37,100 daltons	5 N-linked oligosaccharide chains	22	~25 hours	~50 hours	(Macdougall et al., 1999; Egrie and Browne, 2001; Egrie et al., 2003; Macdougall et al., 2007)
CERA	~ 60,000 daltons	30 kDa methoxy-polyethylene glycol polymer chain	10-14	~134 hours	~139 hours	(Macdougall and Eckardt, 2006; Locatelli et al., 2007; Provenzano et al., 2007)

Normal serum concentrations of EPO for individuals with normal hematocrit range from 4-27 mU/mL but EPO levels can be increased 100-1000 times the normal serum EPO concentration in response to hypoxia and anemia (Jelkmann, 1992). Subcutaneous administration of a single 600 U/kg dose of epoetin alfa to healthy volunteers produced a peak serum concentration of over 1000 mU/mL after 24 hours (Ramakrishnan et al., 2004). Furthermore, it is important to consider differences in pharmacokinetic profile of rhEPO between human, dog, rat, and mouse for achieving the desired therapeutic effects with EPO. Indeed, the total body clearance of EPO increases in the order human = dog < rat ≪ mouse (Egrie et al., 1986; Bleuel et al., 1996).

B. New Generations of Erythropoietin-Related Drugs

1. NESP

The first-generation erythropoiesis stimulating agents were succeeded by the development and production of longer-acting EPO analogues. EPO is known to be desialylated in vivo, cleared from plasma, and is bound to galactose receptors in the liver (Egrie et al., 1993). Thus, there is a direct relationship between the amount of sialic acid-containing carbohydrates, plasma half-life, and in-vivo biological activity, and an inverse relationship with receptor affinity (Jelkmann, 1992; Egrie et al., 1993; Egrie and Browne, 2001). A second-generation NESP (darbepoetin) was created to test the hypothesis that an analogue with more sialic acid–containing oligosaccharides than EPO would have an extended circulating half-life and thereby an increased in vivo biological activity (Macdougall et al., 1999; Egrie et al., 2003). When compared with rhEPO, darbepoetin is a hyperglycosylated rhEPO analog with two extra carbohydrate chains with more sialic acid (Table 1) and has a molecular weight of 37,100 daltons (Macdougall, 2001). The amino acid sequence differs from that of native human EPO at five positions (Macdougall, 2001). Darbepoetin has a threefold longer circulating half-life than rhEPO (Table 1), and due to the pharmacokinetic differences, the relative potency of the two molecules varies as a function of the dosing frequency. Darbepoetin alfa is 3.6-fold more potent than rhEPO in increasing the hematocrit of normal mice when each is administered thrice weekly, but when the administration frequency is reduced to once weekly, darbepoetin alfa is approximately 13-fold higher in vivo potency than rhEPO (Macdougall, 2001; Egrie et al., 2003).

Epoetin and darbepoetin bind to the EPO receptor to induce signal transduction by the same intracellular molecules as native EPO. However, differences in the glycosylation pattern lead to variations in the pharmacodynamic profiles. Darbepoetin has approximately four-fold lower EPO receptor binding activity than rhEPO despite higher potency. The apparent paradox is explained by the counteracting effects of sialic acid containing carbohydrate on clearance (Egrie et al., 2003; Elliott et al., 2004).

2. CERA

More recently, a third-generation EPO-related molecule has been manufactured called continuous EPO receptor activator (CERA; methoxy polyethylene glycol-epoetin beta), which was created by inserting a single 30 kDa polymer chain into the EPO molecule (Macdougall and Eckardt, 2006). The elimination half-life of CERA in humans is considerably increased to about 130 hours and is comparable after intravenous or subcutaneous administrations (Table 1) (Macdougall et al., 2006). Furthermore, CERA at up to once monthly intervals sustains and maintains sustained and stable control of hemoglobin levels (Locatelli et al., 2007; Provenzano et al., 2007).

III. Signal Transduction by Erythropoietin

Survival, proliferation and differentiation of red blood cells by EPO are mediated by activation of the homodimeric EPO receptor, EPOR, which belongs to the cytokine receptor superfamily (Richmond et al., 2005). Binding of EPO to the erythrocytic EPOR induces a conformational change of the homodimeric-EPOR and triggers Janus protein tyrosine kinase 2 (Jak2) phosphorylation and activation (Witthuhn et al., 1993; Miura et al., 1994b). Phosphorylation of Jak2, in turn, phosphorylates several tyrosine residues on EPOR providing docking sites for binding of several intracellular proteins and activation of multiple signaling cascades.

The tissue protective effects of EPO, beyond the hematopoietic system, is mediated but not restricted to activation of homodimeric EPOR, but is also believed to be mediated by its actions on the heterotrimeric complex consisting of EPOR and the β-common receptor (βCR, also known as CD131). The β-common receptor is a shared receptor subunit of interleukin 3, interleukin 5 and granulocyte macrophage-colony stimulating factor (Brines et al., 2004; Brines and Cerami, 2005). This heterotrimeric complex is postulated to mediate EPO-induced cytoprotection in the nervous system, while the involvement of EPOR-βCR complex in EPO-mediated protection in the cardiovascular system has not yet been demonstrated. Nevertheless, following binding of EPO to either homodimeric or heterotrimeric receptor complex activates Jak2. Signal transduction pathways mediating tissue protection in the cardiovascular system activated by EPO-induced phosphorylation by Jak2 are shown in Figure 1.

Figure 1. Signaling pathways mediating EPO-induced protection in the cardiovascular system.

Binding of EPO to either the homodimeric EPO-receptor (EPOR) complex or the heterotrimeric EPOR-βCR complex first activates Janus tyrosine kinase 2 (Jak2). Phosphorylation and activation of Jak2 recruits secondary messengers and activates secondary signaling pathways including phosphatidylinositol 3-kinase (PI3K), mitogen-activated protein kinase (MAPK), and nuclear factor-κB. Some of the pathways mediating EPO-mediated protection are similar to those mediating erythropoiesis or neuroprotection. In particular, EPO-mediated inhibition of apoptosis involves activation of protein kinase B/Akt to inhibit caspase-9 or activate nuclear factor kappa B (NFκB) or inhibit Bad, or stimulate signal transducer and activator of transcription (STAT), or MAPK pathway to increase anti-apoptotic messengers in the mitochondria including Bcl-2 and Bcl-xL. The most common pathway for EPO-mediated cardioprotection as well as vascular protection is EPO-induced augmentation of nitric oxide (NO) production by PI3K/Akt phosphorylation and activation of endothelial nitric oxide synthase (eNOS). EPO is also shown to regulate intracellular Ca²⁺ by activating phospholipase C (PLC).

Stimulation with EPO, on one hand, activates the phosphatidylinositol 3-kinase (PI3K) signaling by recruiting its p85 regulatory subunit to EPOR (Miura et al., 1994a). Phosphorylation and activation of PI3K is coupled to activation of protein kinase B/Akt, a pathway known to stimulate anti-apoptotic signals that facilitate the inhibition of mitochondrial cytochrome c release, in part by translocating pro-apoptotic Bad to mitochondria, and help maintain mitochondrial membrane potential (Figure 1). EPO-induced Akt activation prevents nuclear translocation of the pro-apoptotic forkhead transcription factor FOXO3a by facilitating its binding with the protein 14-3-3 (Chong and Maiese, 2007). In addition, EPO stimulation in the myocardium may lead to phosphorylation and inhibition of glycogen synthase kinase-3β (GSK-3β) in either an Akt-dependent or an independent mechanism (Nishihara et al., 2006; Miki et al., 2009). In peripheral and cerebral vasculature, activation of Akt by EPO phosphorylates endothelial nitric oxide synthase (eNOS) and increases nitric oxide (NO) production (Santhanam et al., 2005; Santhanam et al., 2006; d’Uscio et al., 2007). Akt activation may also prevent activation of caspase-9 to prevent apoptosis in the cerebral vasculature (Chong et al., 2003).

Phosphorylation of Jak2 may directly activate inhibitor of nuclear factor kappa B kinase (IκB kinase or IKK), or indirectly activate Akt to stimulate nuclear translocation of nuclear factor kappa B (NFκB) for nuclear transcription (Figure 1) and exerts cardioprotection (Xu et al., 2005). In isolated perfused rabbit hearts, EPO has been shown to protect the heart against ischemia by activating signal transducer and activator of transcription (STAT-5) and/or mitogen activated protein kinase (MAPK) pathway (Rafiee et al., 2005).

IV. Mechanisms of Vascular Protective Effects of Erythropoietin

A. Erythropoietin and Endothelial Nitric Oxide Synthase

NO is a potent vasodilator and plays a key role in control of the cardiovascular system (Lüscher and Vanhoutte, 1990). NO is formed in endothelial cells from L-arginine by oxidation of its terminal guanidino-nitrogen, requiring the cofactors (6R)-5,6,7,8-tetrahydrobiopterin (BH₄), nicotinamide adenine dinucleotide phosphate-oxidase, flavin adenine dinucleotide, heme, and Zn²⁺ (Palmer et al., 1987; Palmer et al., 1988; Ignarro, 1990; Raman et al., 1998). The formation of NO occurs by activation of eNOS, which is expressed constitutively (Förstermann et al., 1991; Pollock et al., 1991). Relaxations in response to the abluminal release of endothelium-derived NO are associated with stimulation of soluble guanylyl cyclase and in turn formation of cyclic guanosine 3’,5’-monophosphate (cGMP) in vascular smooth muscle cells (Rapoport et al., 1983).

Accumulating experimental evidence suggests that EPO can exert non-erythropoietic effects in vascular endothelium and is increasingly regarded as a potent tissue protective cytokine. Indeed, EPO decreases tissue damage by inhibition of apoptosis and reduction of inflammatory cytokines (Iversen et al., 1999; Rui et al., 2005; Burger et al., 2006; Li et al., 2006). Indeed, in-vitro treatment with low dose of rhEPO increased eNOS protein expression and NO₂^-/NO₃^- levels in cultured endothelial cells (Table 2) (Wu et al., 1999; Banerjee et al., 2000; Beleslin-Cokic et al., 2004). However, incubation of human coronary artery endothelial cells with high dose of rhEPO for 24 hours inhibited eNOS expression and NO production (Wang and Vaziri, 1999). Moreover, asymmetric dimethylarginine concentrations were increased after high dose of rhEPO or NESP over longer period leading to a significant reduction of NO synthesis in cultured endothelial cells (Table 2) suggesting that high dose of EPO may have detrimental effects on endothelial function (Scalera et al., 2005).

Table 2.

In vitro effects of recombinant human EPO on eNOS expression in cultured endothelial cells.

Cell type	Dose of rhEPO	Treatment duration	Reguation	Method of detection	References
Human umbilical vein endothelial cells	0.1, 1, 20, and 40 U/mL	8 hours	Up-regulation	NO2-/NO3- bioassay	(Wu et al., 1999)
Human umbilical vein, dermis, and pulmonary artery endothelial cells	4 U/mL	1-6 days	Up-regulation	RT-PCR, L-arginine-to-L-citrulline conversion assay	(Banerjee et al., 2000)
Human bone marrow microvascular endothelial cells	5 U/mL	1 hour	Up-regulation	Western blot, NO3- bioassay	(Beleslin-Cokic et al., 2004)
Human umbilical vein endothelial cells	5 U/mL	0.5 hour	Up-regulation	NO3- bioassay	(Beleslin-Cokic et al., 2004)
Bovine aortic endothelial cells	0.1 - 10 U/mL	1-24 hours	No change	Western blot	(Lopez Ongil et al., 1996)
Human artery endothelial cells	5 U/mL	1 hour	No change	Western blot, NO3- bioassay	(Beleslin-Cokic et al., 2004)
Human coronary artery endothelial cells	5 and 20 U/mL	24 hours	Down-regulation	Western blot, NO2-/NO3- bioassay	(Wang and Vaziri, 1999)
Human umbilical vein endothelial cells	10, 50, 100, and 200 U/mL	24 hours	Down-regulation	NO2-/NO3- bioassay	(Scalera et al., 2005)

rhEPO, recombinant human erythropoietin; eNOS, endothelial nitric oxide synthase; NO₂^-, nitrite; NO₃^-, nitrate.

As illustrated in Table 3, in-vivo treatment with rhEPO has been shown to increase urinary NO₂^-/NO₃^- levels in normotensive rats (del Castillo et al., 1995; Tsukahara et al., 1997), increase vascular eNOS phosphorylation and eNOS protein expressions (Ruschitzka et al., 2000; Kanagy et al., 2003; d’Uscio et al., 2007), improve endothelium-dependent relaxations in isolated aortas of rats and mice (Tsukahara et al., 1997; Iversen et al., 1999; Ruschitzka et al., 2000; Kanagy et al., 2003), and improve endothelial function in predialysis patients (Kuriyama et al., 1996). Furthermore, in transgenic mice overexpression of human EPO markedly increased aortic eNOS protein expression, NO-mediated endothelium dependent relaxation, and circulating and vascular tissue NO levels (Table 3). These mice do not develop hypertension, stroke, myocardial infarction, or thromboembolic complications despite excessive erythrocytosis exhibiting very high hematocrit levels of 80% (Ruschitzka et al., 2000). The increased NO production in these animals appears to counteract increased expression of potent vasoconstrictor endothelin-1 (ET-1) (Quaschning et al., 2003). Indeed, EPO transgenic mice treated with the NO synthase inhibitor exhibit high systolic blood pressure and show increased mortality whereas wild-type siblings develop only hypertension (Ruschitzka et al., 2000). Despite concomitant activation of the ET-1 system observed in transgenic mice, elevated NO levels led to a pronounced vasodilation, thereby protecting the transgenic animals from cardiovascular complications. Consistent with this concept, studies on cultured endothelial cells demonstrated that inactivation of NO synthesis caused increased production of ET-1 (Boulanger and Lüscher, 1990). In addition, EPO increased systolic blood pressure, accelerated thrombus formation, and exacerbated medial thickening of injured carotid arteries in eNOS-deficient mice (d’Uscio et al., 2007; Lindenblatt et al., 2007). These observations underscore the importance of eNOS during vascular adaptation to increased circulating levels of EPO.

Table 3.

In vivo effects of recombinant human EPO on vascular eNOS in animal studies.

Animal species	Dose of rhEPO	Treatment duration	Cell tissue	Result	Method of detection	References
Nephrectomized Sprague-Dawley rat	150 U/kg, intraperitoneal, twice a week	6 weeks	Thoracic aorta	No change	Western blot, NOx bioassay	(Ni et al., 1998)
Rabbit	400 U/kg, intravenous, each other day	1 week	Carotid artery	No change	Endothelium-dependent vasodilation studies in isolated artery	(Noguchi et al., 2001)
Sprague-Dawley rat	100 U/kg or 300 U/kg, subcutaneous, each other day	2 weeks	Urine	Up-regulation	NOx bioassay	(Tsukahara et al., 1997)
Sprague-Dawley rat	150 U/kg, subcutaneous, three times per week	3 weeks	Urine	Up-regulation	NOx bioassay	(del Castillo et al., 1995)
Sprague-Dawley rat	~ 160 U/kg/day, subcutaneous	2 weeks	Thoracic aorta	Up-regulation	Western blot, NOx bioassay, endothelium-dependent vasodilation studies in isolated aorta	(Kanagy et al., 2003)
EPO-transgenic mice	n/a	n/a	Aorta	Up-regulation	Western blot, NOx bioassay, endothelium-dependent vasodilation studies in isolated aorta	(Ruschitzka et al., 2000)
C57BL/6 mice	1000 U/kg, subcutaneous, twice a week	2 weeks	Aorta	Up-regulation	Western blot	(d’Uscio et al., 2007)
C57BL/6 mice	1000 U/kg, intraperitoneal, for the initial 3 days	2 weeks	Endothelial progenitor cells	Up-regulation	Immunohistochemistry, NOx bioassay	(Urao et al., 2006)

Because of an increased number of circulating red blood cells by EPO, subsequent increase in shear stress is a powerful stimulus for up-regulation of eNOS in endothelial cells (Berk et al., 1995; Ruschitzka et al., 2000; Davis et al., 2001; Boo et al., 2002; Lam et al., 2006). However, administration of EPO for 3 days, although not affecting the number of circulating red blood cells, stimulated phosphorylation of eNOS to a similar degree as did treatment with EPO for 14 days. These findings suggest that EPO has a direct stimulatory effect on phosphorylation of eNOS in vascular endothelium and that this effect is independent of hematopoietic effects of EPO (d’Uscio et al., 2007). Most recently, it was demonstrated that activation of eNOS by hypoxia was abolished in EPO-receptor-deficient mice, the latter mutants also exhibiting an exacerbation of pulmonary hypertension and vascular injury. These findings suggest that the vascular protective effects of EPO are dependent on NO production via activation of the EPO receptor in endothelial cells (Beleslin-Cokic et al., 2004; Satoh et al., 2006).

B. Erythropoietin and protein kinase B/Akt

As discussed earlier, EPO elicits phosphorylation of Akt and this is dependent upon the activation of Jak2 and PI3K. Numerous studies demonstrated tissue protective effects of both rhEPO and darbepoetin alpha mediated by phosphorylation of Akt and a subsequent increase in the production of NO (Chong et al., 2002; Bahlmann et al., 2004b; Santhanam et al., 2006; Chong and Maiese, 2007; Lindenblatt et al., 2007; d’Uscio and Katusic, 2008). To date, the cardiovascular effects of third generation of EPO analogues are largely unknown.

C. Erythropoietin and Tetrahydrobiopterin

Tetrahydrobiopterin (BH₄) is an essential cofactor required for enzymatic activity of eNOS (Raman et al., 1998). The biosynthesis of BH₄ is dependent on activity of the rate-limiting enzyme GTP-cyclohydrolase I (GTPCH I) (Nichol et al., 1985). Recent study showed that EPO causes an increase in intracellular levels of BH₄ via activation of GTPCH I (d’Uscio and Katusic, 2008). Pharmacological inhibition of Jak2 with AG490 abolished EPO-induced BH₄ biosynthesis, suggesting that increased phosphorylation and activation of Jak2 are the molecular mechanisms underlying the observed effect of EPO on BH₄ synthesis. Further analysis revealed that PI3K activity is an upstream activator of Akt1, because pharmacological and genetic inactivation of PI3K/Akt1 abolishes the stimulatory effects of EPO on GTPCH I activity and biosynthesis of BH₄ in mouse aorta (d’Uscio and Katusic, 2008). The ability of EPO to upregulate GTPCH I activity and eNOS phosphorylation in a coordinated fashion is most likely designed to optimize the production of NO in vascular endothelium.

D. Erythropoietin and Antioxidant Enzymes

The local concentrations of NO in arterial wall are not only dependent on enzymatic activity of eNOS but are also determined by concentrations of superoxide anions (Harrison, 1994). Indeed, recent study showed that in wild-type mice, treatment with EPO increases vascular CuZn superoxide dismutase (SOD1) expression and effectively prevents vascular remodeling after carotid artery injury (d’Uscio et al., 2010). In contrast, genetic inactivation of SOD1 abolished ability of EPO to reduce concentrations of superoxide anions thereby suggesting that EPO exerts antioxidant effect in blood vessel wall by regulating expression and activity of SOD1 protein.

E. Adverse Effects of Erythropoietin

Normally, endothelial cells contribute to the regulation of blood pressure and blood flow by releasing vasodilators such as NO and prostacyclin, as well as vasoconstrictors including ET-1 and prostanoids (Lüscher and Vanhoutte, 1990). Long-term administration of rhEPO has been associated with hypertension (Krapf and Hulter, 2009). The mechanisms of hypertension induced by long-term administration of EPO include increased cytoplasmic Ca²⁺ concentrations and increased ET-1 production leading to a blunted response to the vasodilator NO in rabbits and rats (Bode-Boger et al., 1992; Vaziri et al., 1995). Chronic use of EPO raises cytosolic Ca²⁺ concentration in smooth muscle cells, endothelial cells or platelets (Van Geet et al., 1990; Neusser et al., 1993; Vaziri et al., 1996) and in turn, may contribute to impaired responses to NO donors. It is also suggested that EPO, at doses exceeding 200 U/mL, activates tyrosine phosphorylation of phospholipase Cγ, which hydrolyzes phosphatidyinositol bisphosphate to inositol 1,4,5-triphosphate and diacylglycerol. These second messengers may, in turn, elevate Ca²⁺ concentration as well as activate protein kinase C in vascular smooth muscle cells (Akimoto et al., 1999; Akimoto et al., 2001). Similar doses of EPO (200 U/mL) released ET-1 and increased constrictor prostanoids over dilator prostanoids in vitro in endothelial cells (Bode-Boger et al., 1996; Krapf and Hulter, 2009) and these observations may help to explain EPO-induced hypertension observed in animals and in humans.One implication of these results is that endothelial dysfunction predisposes to EPO dependent hypertension. In a recent study, EPO treatment caused hypertension in rats treated in combination with a NOS inhibitor but not in rats treated with EPO alone (Moreau et al., 2000). In addition, a study using polycythemic mice overexpressing rhEPO observed that NOS blockade caused the normotensive polycythemic mice to develop hypertension (Ruschitzka et al., 2000). Therefore, erythropoiesis with a raised hematocrit is not likely associated with an increased risk for hypertension and thrombosis as long as endothelial NO production serves as compensatory mechanism (Moreau et al., 2000; d’Uscio et al., 2007; Lindenblatt et al., 2007; d’Uscio and Katusic, 2008). Impaired endothelial-dependent dilation in hypertensives (Spieker et al., 2000) as well as in hemodialyzed patients (Joannides et al., 1997) and the impaired ability to synthesize endothelial NO may increase susceptibility to EPO-induced hypertension.

V. Erythropoietin and Cardioprotection

Endogenous EPO system facilitates cardiomyocyte survival after ischemia-reperfusion injury and accelerates left ventricular remodeling (Tada et al., 2006). Deficiency of endogenous EPO-EPOR system resulted in acceleration of pressure overload-induced cardiac dysfunction by accelerating left ventricular hypertrophy, dysfunction and reduced survival (Asaumi et al., 2007). Therapeutic potential of EPO to exert cardioprotection was demonstrated in models of myocardial ischemia wherein EPO inhibited apoptosis and augmented survival of cardiac myocytes (Calvillo et al., 2003; Parsa et al., 2003), mediated in part, by activating EPOR expressed on the cardiac myocytes (Wright et al., 2004). Favorable results from numerous studies offered promise for administering EPO to treat myocardial infarction (Cai et al., 2003; Calvillo et al., 2003; Moon et al., 2003; Parsa et al., 2003; Wright et al., 2004).

However, subsequent studies in-vitro identified activation of multiple signaling pathways in EPO-induced cardioprotection (Shi et al., 2004; Rafiee et al., 2005). In particular, EPO-mediated cardioprotection involves activation of one or more of the following pathways: Jak/STAT pathway, Jak2/PI3K/Akt/GSK3β pathway, Jak2/MAPK pathway as well as activation of protein kinase Cε. Cardioprotection by EPO also seems to be mediated by activation of potassium channels, in particular, K_ATP and mitochondrial calcium-activated potassium channels (Shi et al., 2004). Administration of EPO before ischemia, at the onset of ischemia or after reperfusion triggered response mimicking preconditioning in the ischemic myocardium (Parsa et al., 2003; Baker, 2005; Gross and Gross, 2006). However, the role of EPOR in this supposed pre-conditioning response remains to be clarified. In addition, the ability of EPO to stimulate eNOS activation may also facilitate cardiomyocyte survival, as observed during hypoxia-induced apoptosis of cardiac myocytes (Burger et al., 2006). As mentioned earlier, the doses of EPO for cardioprotection far-exceeded the doses of EPO used in treatment of anemia and clinical translations of these results may be hampered by the onset of widely-reported adverse events. In this regard, understanding of the role of EPOR-βCR receptor complex may help expand clinical applications of non-hematopoietic analogues of EPO to cardioprotection. Recent success with carbamylated EPO, which selectively binds to EPOR-βCR heteromer, in a mouse model of ischemia-reperfusion injury (Xu et al., 2009) provides a promising outlook toward expanding tissue protection of EPO in the clinic to cardioprotection.

Occurrence of anemia as a risk factor for morbidity and mortality in patients with chronic heart failure necessitated evaluation of EPO in clinical trials. Randomized, single-center studies successfully demonstrated treatment of anemia in patients with mild, moderate or severe heart failure with associated improvement in exercise capacity, cardiac and renal function and reduced use of diuretics (Silverberg et al., 2001; Mancini et al., 2003; Namiuchi et al., 2005; Palazzuoli et al., 2006). Recently, three trials evaluated the safety and efficacy of darbepoetin alfa (0.75 μg/kg once every 2 weeks) in symptomatic heart failure patients (Ponikowski et al., 2007; van Veldhuisen et al., 2007; Ghali et al., 2008). Incidence of adverse events in these trials was similar between placebo- and darbepoetin alfa-treated patients (Klapholz et al., 2009), while an increase in hemoglobin with darbepoetin alfa treatment tended to correlate with improved health-related quality of life. Results obtained in these clinical trials extend support as a proof of concept and encourage the need for larger outcomes trials for treatment of anemia with erythropoiesis-stimulating agents.

In addition to treatment of anemia, results from pre-clinical studies attribute multiple mechanisms of protection by EPO against myocardial disorders. Treatment with EPO may decrease apoptosis of myocytes, induce neovascularization by promoting myocardial angiogenesis, reduce collagen deposition in ischemic myocardium as well as improve left ventricular function. Employing either non-hematopoietic analogues of EPO or a novel EPO delivery system (Kobayashi et al., 2008) may expand therapeutic boundary for EPO-mediated cardioprotection. Success from these studies will lay the frame work for future clinical evaluation in the treatment of myocardial infarction.

VI. Erythropoietin and Cerebrovascular Disorders

Identification of EPO and EPOR in the brain (Tan et al., 1992; Masuda et al., 1994) expanded investigations into tissue protective effects of EPO, beyond the hematopoietic system. Initial research ascribed EPO-mediated protection to the cytokine’s ability to inhibit apoptosis in tissues adjacent to a pathological insult in the brain. The current understanding of the tissue protective effects of EPO in the nervous system involves interaction of the non-hematopoietic βCR with the classical EPOR and subsequent activation of multiple signaling cascades, as reviewed by Brines and Cerami (Brines and Cerami, 2005). In the nervous system, EPO mediates neuroprotection following ischemic, hypoxic, metabolic, neurotoxic and excitotoxic stress (Genc et al., 2004). EPO-mediated protective effects in the brain may involve one or more of the following mechanism: a) prevention of excitatory aminoacid release, b) inhibition of apoptosis, c) anti-oxidant effects, d) anti-inflammatory effects and e) stimulation of neurogenesis and angiogenesis.

EPO exhibits higher affinity for endothelium of cerebral arteries as compared to neuronal cells (Brines and Cerami, 2006). Prior studies have demonstrated that EPO activates cerebrovascular protective mechanisms (Grasso et al., 2002; Grasso, 2004; Santhanam et al., 2005; Santhanam et al., 2006). In cerebral arteries exposed to recombinant EPO, the expressions of eNOS and its phosphorylated (S1177) form were increased. Basal levels of cGMP were also significantly elevated consistent with increased NO production. Over-expression of EPO in cerebral arteries reversed vasospasm in rabbits induced by injection of autologous blood into cisterna magna. Arteries-transduced with recombinant EPO demonstrated significant augmentation of the endothelium-dependent relaxations to acetylcholine. Over-expression of EPO further increased the expression of phosphorylated Akt and eNOS and elevated basal levels of cGMP in the spastic arteries (Santhanam et al., 2005). Cerebrovascular protective effects of EPO against cerebral vasospasm following experimental subarachnoid hemorrhage may appear to be mediated in part by phosphorylation and activation of endothelial Akt/eNOS pathway.

Administration of EPO into the brain reduced neurologic dysfunction in rodent models of stroke (Sadamoto et al., 1998; Sakanaka et al., 1998; Bernaudin et al., 1999; Brines et al., 2000; Siren et al., 2001). Following the success of EPO in pre-clinical models of stroke and cerebral ischemia, a successful proof of concept clinical trial demonstrated that intravenously injected EPO was well tolerated in patients with acute ischemic stroke (Ehrenreich et al., 2002). In addition, outcomes from a recent Phase II trial reports beneficial effect of EPO, as observed from a reduction in delayed ischemic deficits after aneurysmal subarachnoid hemorrhage (Tseng et al., 2009).

However, results from the German Multicenter EPO Stroke Phase II/III trial warrants caution in administering EPO with thrombolytics in patients with acute stroke (Ehrenreich et al., 2009). In this trial, patients receiving EPO alone demonstrated a reduction in National Institute of Health Stroke Scale (NIHSS) score from day 1 to day 90, an index of improved neurological outcome. On the contrary, when combined with thrombolytic recombinant tissue plasminogen activator, patients receiving EPO demonstrated increased risk of complications including death, intracerebral hemorrhage, brain edema, and thromboembolic events compared to patients receiving placebo. Further understanding of the complex interactions between different components of the cardiovascular system initiated by EPO is necessary for designing better strategies to maximize therapeutic potential of this cytokine.

With new evidences pointing toward the role of non-hematopoietic βCR in EPO-mediated tissue protective effects (Grasso et al., 2004), use of non-erythrocytic analogues of EPO currently under development offer promise of therapy as adjunct for patients with stroke and other cerebral vascular disorders with enhanced safety and reduced adverse effects. Another novel strategy currently investigated for acute stroke has employed the neurotrophic ability of EPO to differentiate neural progenitor cells into neurons. The Phase IIb prospective randomized, double-blind study of NTx™-265, comprising human chorionic gonadotrophin (for proliferation of endogenous neural stem cells) and EPO (to differentiate these neural stem cells) aims to improve neurological outcome in acute stroke patients, and was based on the success of pre-clinical studies adopting similar strategy in rats (Belayev et al., 2009).

VII. Erythropoietin and Progenitor Cells

Seminal studies by Heeschen et al. (Heeschen et al., 2003) demonstrated that circulating levels of EPO in humans significantly correlated with the number of stem and progenitor cells in the bone marrow as well as to the number and function of circulating progenitor cells. In addition, treatment of mice with EPO increased the number of stem and progenitor cells in the bone marrow as well as an increase in the number of peripheral blood endothelial progenitor cells (EPCs), suggestive of stimulation of mobilization by EPO. Indeed, mobilization of progenitor cells by EPO stimulated postnatal neovascularization in a model of hind limb ischemia (Heeschen et al., 2003; Li et al., 2009; Kato et al., 2010). Numerous studies, subsequently, have established a crucial pro-angiogenic role of EPO, as indicated by enhanced mobilization of progenitor cells, to elicit vascular repair. As illustrated in Table 4, treatment with EPO significantly augments mobilization of diverse population of progenitor cells from bone marrow distinguished by their distinct phenotype and, in turn, the mobilized progenitor cells contributed to either neovascularization or repair of denuded endothelium.

Table 4.

Contribution to vascular repair by progenitor cells mobilized by EPO.

Species	Dose of EPO	Functional Outcome	Phenotype of mobilized progenitor cells	References
Mice, Humans	1000 U/kg	Increased angiogenesis in models of disc neovascularization and hind limb ischemia	Lin-1+/Sca-1+ stem cells	(Heeschen et al., 2003)
			Sca-1+/Flk-1+ or CD34+/Flk-1+ cells
Humans	5000 ± 674 U	Increased angiogenesis in EPCs of patients exposed to EPO	CD34+/CD45+ cells	(Bahlmann et al., 2004a)
			UEA-1+/acLDL-DiI+ cells
Mice	1000 U/kg	Inhibition of neointimal hyperplasia after wire injury of femoral artery	CD45dim/Flk-1+ cells	(Urao et al., 2006)
			Sca-1+/Flk-1+ cells
Mice	100 U	Accelerated revascularization in a model of hind limb ischemia	CXCR4+/VEGFR1+-hemangiocytes	(Jin et al., 2006)
Dogs	1000 U/kg	Augmented neovascularization in a model of myocardial infarction	CD34+-mononuclear cells	(Hirata et al., 2006)
			Di-acLDL+/UEA-I cells
Rats	40 μg/kg	Increased functional neovascularization following acute myocardial infarction	Ac-LDL+/Lectin+ cells per high powered field	(Westenbrink et al., 2007)
Mice	1000 U/kg	Epo-induced mobilization impaired in eNOS-/- mice	Lin-1-/Sca-1+/c-Kit+ cells	(Santhanam et al., 2008)
			CD34+/Flk-1+ cells
Mice	500 U/Kkg	Acceleration of smooth muscle lesion formation by EPO in a model of carotid artery ligation	Ter-119-/CD45lo/c-Kit+/Sca-1+ cells	(Janmaat et al.)

This table summarizes the results from studies wherein EPO was administered to stimulate mobilization of progenitor cells in a model of vascular injury. Studies relating increased endogenous levels of EPO with enhanced mobilization of progenitor cells and their contribution to vascular repair have not been discussed.

Multiple mechanisms are likely to mediate mobilization of progenitor cells by EPO (Heeschen et al., 2003; Aicher et al., 2005; Urao et al., 2006). Prior studies from our laboratory have demonstrated that activation of endothelial nitric oxide synthase is critical for EPO-induced mobilization of hematopoietic stem and progenitor cells and CD34⁺/Flk-1⁺ EPCs (Santhanam et al., 2008). The ability of EPO to up-regulate the antioxidant capacity of EPCs, in particular, activation of SOD1 may also help to explain the pronounced vascular protection observed with this pleiotropic cytokine (He et al., 2005).

Endogenous EPO/EPOR system in the vasculature also dictates mobilization of progenitor cells and facilitates vascular repair. Satoh et al. (Satoh et al., 2006) successfully demonstrated this crucial role of endogenous EPO/EPOR using EPOR null mutant mice that expresses EPO-R exclusively in the erythroid lineage (EPOR^-/- rescued mice). Lack of EPOR signaling in the hematopoietic system resulted in attenuated mobilization of Flk-1⁺/CD133⁺ EPCs (hematopoietic progenitor cell population) (Satoh et al., 2006), as well as alteration in VEGF/VEGFR signaling and impaired recovery after hind limb ischemia in mice (Nakano et al., 2007). During hypoxia-induced pulmonary hypertension, in addition to impaired endogenous mobilization, recruitment of administered progenitor cells was impaired in EPOR^-/- rescued mice (Satoh et al., 2006). It is likely that endogenous EPO-mediated effects are not restricted to its effects on mature vascular cells, but also contributes to mobilization, recruitment and activation of progenitor cells as demonstrated in either hypoxia-induced pulmonary hypertension (Satoh et al., 2006) or in a model of myocardial ischemia and reperfusion (Tada et al., 2006). Elevated plasma levels of EPO in patients with acute myocardial infarction correlated with mobilization of CD34⁺/CD133⁺/VEGFR2⁺ EPCs (Ferrario et al., 2007). It is therefore likely that the extent of alteration of endogenous EPO levels following a cardiovascular insult may dictate the proportion of endogenous progenitor cells mobilized in the mononuclear cell population, which in turn, may contribute to extent of damage or repair.

VIII. Conclusions

Extensive research has demonstrated that EPO does not only affect the hematopoietic system, but also plays important role in control of cardiovascular system. EPO has been shown to be vascular protective by exerting its effects on the endothelial cells and vascular smooth muscle cells. Tissue protective effects of EPO on the vasculature are mediated, in part, by prevention of apoptosis or stimulation of endothelial nitric oxide production and augmented vasodilatation. In addition, EPO regulates mobilization of pro-angiogenic cells, including EPCs, from the bone marrow and stimulates neovascularization. EPO may exert its beneficial effects on the vasculature by acting through its homodimeric EPOR or through EPOR-βCR and activation of Jak2 to stimulate PI3K/Akt, NFκB, MAPK signaling pathways. Further studies are however needed to clarify the receptor(s) involved in protective effects of EPO in different vascular cell types and injury models. In addition, exacerbation of injury in EPO-treated mice deficient in either eNOS or CuZn superoxide dismutase highlights the critical role of endogenous NO or antioxidant defense mechanisms in triggering adaptation of vascular wall to elevated doses of EPO, commonly used in preclinical and clinical settings. These studies may help to explain the adverse events precluding the clinical use of EPO in treatment of cardiovascular and neurological disorders.

Results obtained from clinical trials with EPO on cerebrovascular disorders, conducted over the last decade, suggests EPO as a promising lead towards designing and developing novel strategies for these disorders. Future research on better understanding of protective effects of the non-hematopoietic analogues of EPO as well as harnessing the EPO-induced ability to stimulate progenitor cell function may favor therapeutic benefits with minimal side effects associated with this pleiotropic cytokine.

Figure 2. Morphological studies of carotid arteries of eNOS-deficient mice undertaken fourteen days after injury.

Carotid arteries were stained with standard Verhoeff van-Giessen. Representative photomicrographs of uninjured carotid arteries (A), carotid arteries after injury (B), and injured carotid arteries of eNOS-deficient mice treated with EPO for 14 days (C). The media is demarcated by internal elastic lamina (open arrow) and external elastic lamina (black arrow). Original magnification x200. Size bar = 50 μm. D: Quantitative histomorphometric analyses of medial CSA in carotid arteries of eNOS-deficient mice without (-) and with (+) EPO treatment. Data are shown as means ± SEM (n=4-6). * P<0.05 vs. control uninjured; † P<0.05 vs. uninjured + EPO (ANOVA with Bonferroni’s).

Abbreviations

EPO

erythropoietin

EPOR

erythropoietin-receptor

βCR

β-common receptor

rhEPO

recombinant human erythropoietin

NESP

novel erythropoiesis stimulating protein

CERA

continuous erythropoietin receptor activator

Jak2

Janus kinase 2

NFκB

nuclear factor kappa B

STAT-5

signal transducer and activator of transcription 5

MAPK

mitogen activated protein kinase

PI3K

phosphatidylinositol 3-kinase

NO

nitric oxide

eNOS

endothelial nitric oxide synthase

cGMP

cyclic guanosine 3’,5’-monophosphate

ET-1

endothelin-1