Erythropoietic Agents as Neurotherapeutic Agents: What Barriers Exist?

September 3, 2002

Erythropoietin is the primary physiological regulator of erythropoiesis, and it exerts its effect by binding to cell surface receptors. It has recently been shown that both erythropoietin and its receptor are found in the human cerebral cortex, and that, in vitro, the cytokine is synthesized by astrocytes and neurons, has neuroprotective activity, and is up-regulated following hypoxic stimuli. In animal models, exogenous recombinant human erythropoietin has been reported to be beneficial in treating experimental global and focal cerebral ischemia and reducing nervous system inflammation

ABSTRACT: Erythropoietin is the primary physiological regulator of erythropoiesis, and it exerts its effect by binding to cell surface receptors. It has recently been shown that both erythropoietin and its receptor are found in the human cerebral cortex, and that, in vitro, the cytokine is synthesized by astrocytes and neurons, has neuroprotective activity, and is up-regulated following hypoxic stimuli. In animal models, exogenous recombinant human erythropoietin has been reported to be beneficial in treating experimental global and focal cerebral ischemia and reducing nervous system inflammation. These findings suggest that exogenous administration of erythropoietic agents (darbepoetin alfa [Aranesp], epoetin alfa [Epogen, Procrit], and epoetin beta [NeoRecormon]) may be a potential therapeutic tool for central nervous system injury. However, transport of protein therapeutics to the brain’s extracellular environment via systemic blood supply generally does not occur due to the negligible permeability of the brain capillary endothelial wall. Therefore, in order to pharmacologically exploit and fully realize the therapeutic benefits of exogenous erythropoietic agents in CNS dysfunction, mechanisms of action and the potential impact of biodistribution barriers need to be elucidated. [ONCOLOGY 16(Suppl 10):91-107, 2002]

This report is structuredto servetwo purposes. The first is tosummarize current understanding of the biology of erythropoietin in the centralnervous system (CNS) and to highlight missing pieces of the puzzle. Some of thebasic questions to be addressed include: What is the role of erythropoietin inthe CNS? Does it differ from its function in systemic erythropoiesis? Iserythropoietin an instructive or survival/trophic factor during CNS developmentand in the adult brain? Is there a link between CNS and peripheralerythropoietin regulation? Do exogenously administered erythropoietic agentscross the blood-brain barrier/blood-cerebrospinal fluid barrier? Doerythropoietic agents need to cross the blood-brain barrier in order to elicitcentral therapeutic effects?

The second purpose of this report is to discuss ongoingresearch addressing these questions. Suggestions of how this information may beintegrated to fully realize the potential benefits of intervention witherythropoietic agents to prevent or reduce pathological neuronal loss in CNSdisease or injury are proposed where appropriate.

As it is beyond the scope of this report to provide acomprehensive review of erythropoietin literature, the reader is referred toavailable reviews.[1-7]

Erythropoietin: Gene Structure and Activity

The human endogenous erythropoietin (EPO) gene, a single-copygene located on chromosome 7, consists of 5 exons and 4 introns. Erythropoietinis initially formed as a 197-amino-acid glycoprotein. The leader sequence andterminal amino acid are cleaved prior to secretion of the mature 165-amino-acidpeptide with a molecular weight of 34,000 Da.[8] Endogenous erythropoietin hasthree N-linked glycosylation sites and one O-linked glycosylation site.Glycosylation of these sites results in four complex carbohydrate chainscontaining sialic acid,[9-11] and the duration of erythropoietin circulation inblood appears to be governed by these moieties.

For biological activity, erythropoietin amino acids 150-160interact with the erythropoietin receptor.[12] Studies have shown that there isa direct relationship between sialic acid-containing carbohydrate content,serum half-life, and in vivo biological activity, and an inverse relationshipwith its receptor binding affinity.[13]

Erythropoietic stimulation was believed to be the solephysiological function of erythropoietin. However, evidence now exists thatsuggests that erythropoietin also acts as an instructive signal during fetaldevelopment and an antiapoptotic agent promoting survival and differentiation inthe brain, heart, and uterus. The main erythropoietin production site is thekidney in the adult and the liver in the fetus.

Erythropoietin Receptors

The action of erythropoietin is mediated by binding to thespecific erythropoietin receptor, which belongs to the family of cytokinereceptors that do not have a tyrosine kinase domain. This family includesreceptors for growth hormone, granulocyte-colony stimulating factor (G-CSF [Neupogen]), and thrombopoietin, among others.[14] The erythropoietinreceptor is a 507-amino-acid (55,000-68,000 Da) polypeptide with a singledomain,[15] which has an unusual tryptophan-serine-X (any amino acid)-tryptophan-serineextracellular motif.[16] The extracellular N-terminal region contains theerythropoietin-binding domain and the C-terminal intracellular regionparticipates in signal transduction.[17]

Once erythropoietin binds to the erythropoietin receptor (EPO-R),rapid tyrosine phosphorylation is induced by receptor dimerization. Thephosphorylated tyrosines act as docking sites for intracellular proteins, whichare subsequently phosphorylated, leading to activation and downstream signaltransduction. Signal termination occurs via dephosphorylation and inactivationof the kinase, which down-regulates the signaling cascade.[18]

In mammalian embryos, EPO-Rs are initially found in the yolksac blood islands, where ‘primitive’ erythropoiesis occurs and then latershifts to the fetal liver. At term, liver erythropoiesis is largely attenuated,supposedly by the increase of glucocorticoid levels.[19] The EPO-Rs areinitially manifested in burst-forming unit-erythroid (BFU-e) cells and reachmaximum expression in the colony-forming unit-erythroid (CFU-e) cells, followingwhich receptor numbers decline sharply (see Figure1). In small animals such asrats and mice, "definitive" erythropoiesis occurs in the bone marrowand spleen, where erythropoietin receptors are localized. In adult humans,erythropoiesis largely occurs in the bone marrow. The erythropoietin requirementis absolute for "definitive" erythropoiesis in the developing fetusand human adult erythropoiesis.[20-22]

Erythropoietic Agents

The erythropoietic agents (darbepoetin alfa [Aranesp],epoetin alfa [Epogen, Procrit], and epoetin beta [NeoRecormon]) have thebiologic activity of endogenous erythropoietin, which is mediated viaerythropoietin receptor. Recombinant human erythropoietin (rHuEPO) has an aminoacid sequence identical to that of human erythropoietin. Endogenouserythropoietin and rHuEPO have microheterogeneity in their structures and arecomprised of several isoforms, including some with charge differences.[23]Charge differentiation by isoelectric focusing shows that rHuEPO-alpha andrHuEPO-beta patterns are similar (isoelectric point [pI] 4.4-5.1), butdistinguishable from purified urinary erythropoietin, which is more acidic (pI3.92-4.42). Such differences permit the differentiation of exogenouslyadministered erythropoietin from endogenous protein.[23-25]

Darbepoetin alfa is a new erythropoietic agent generated bysite-directed mutagenesis of the erythropoietin gene, resulting in an increasednumber of glycosylation sites and greater carbohydrate content.[13]Consequently, darbepoetin alfa has an increased serum residence time and greaterpotency in vivo.[26] In clinical trials of patients with renal failure,darbepoetin alfa was shown to have a threefold longer terminal half-life thanepoetin alfa (25.3 h vs 8.5 h) and to be as efficacious as epoetin alfa despitereductions in dosing frequency to weekly and once-every-2-week regimens.[26,27]In cancer patients receiving chemotherapy, previous studies have shown thatdarbepoetin alfa effectively increases hemoglobin concentrations whenadministered once weekly, once every 2 weeks, or once every 3 weeks. [28-30]

The quest to identify small-molecule agonists of theerythropoietin receptor led to the discovery of both peptide[31-33] andnonpeptide[34] erythropoietin mimetics. These molecules mimic EPO-R binding andactivate the signal transduction pathways used by erythropoietin. While thesemolecules have yet to be exploited clinically, they have proven to be valuabletools for investigating the biological and functional structure of EPO-R, whichwill be useful for continued exploration of erythropoietin mimetic agents.

What Is the Role of Erythropoietin in the CNS? Is it Different From Systemic Erythropoiesis?

Human Brain Development

Erythropoietic stimulation was thought to be the solephysiological function of erythropoietin. Besides this classical function oferythropoietin, however, data indicate that erythropoietin activity and tissueerythropoietin receptor distribution is not restricted to the erythroid lineage.In all instances, erythropoietin receptor must be expressed at the site ofaction and erythropoietin must bind to its specific receptor in order to elicitbiological effects. Erythropoietin response has been identified in endothelial,renal, neuronal, and cardiac cells in vitro.[35,36] Moreover, in vivo studieshave demonstrated erythropoietin involvement in neovascularization,neuroprotection, and uterine angiogenesis.[37-39]

Expression of erythropoietin and erythropoietin receptor inhuman brain tissue ranging from 5 weeks postconception to adult has been studiedby immunohistochemistry in preserved sections following elective abortion,surgical removal, or autopsy. Results demonstrate that erythropoietin and EPO-Rexpression is evident from 5 weeks through to adulthood and has a definitivetemporal- and spatial-dependent distribution through development (see Figure2A). Mouse EPO-R-targeted deletion studies have shown that the absence oferythropoietin receptor significantly affects brain[40] (see Figure2B) andcardiac development[41] as early as embryonic day 10.5. These data do notclearly demonstrate whether erythropoietin acts only as a survival/trophicfactor, or also has instructive potential (see Figure3). Embryonic death isseen at embryonic day 13.5 in erythropoietin receptor null mice. Findings inrodent studies appear to correlate well with observations in mammaliandevelopment (see Figure 2B).[40]

EPO/EPO-R in CNS:In Vitro Evidence

EPO-R expression has been documented in the embryonic, fetal,and adult brain in mice, rat, monkeys, and human cell cultures and tissuesections.[42-45] Neuronal cell lines such as NT2 (human committed neuronalprecursor cell line, inducible to differentiate into postmitotic neurons),[46]PC12 (rat pheochromocytoma cell line, which can differentiate into a neuronalphenotype),[47] SN6 (a cholinergic hybridoma cell line with neuronalproperties),[48] and SK-N-MC cells (a human neuroblastoma cell line)[49] allexpress erythropoietin receptor. Interestingly, binding of erythropoietin toPC12 cells increases the intracellular concentration of calcium andmonoamines,[36] suggesting that erythropoietin might have a role in affectingneuronal excitability possibly by lowering levels of monoamines such asserotonin and noradrenaline in the CNS.

Studies have shown that purified erythropoietin from ratprimary brain cultures had a higher biological activity than erythropoietinisolated from rat serum. The erythropoietin receptor cloned from PC12 cells appears to be identical to that from raterythroid cells. However, there were significant differences noted in the ligandbinding properties between two cell lineages (PC12 (Kd = 16nM), and erythroid cells (Kd = 95pM for high affinity sites and 1.9 nM for low affinity sites).[36] Western blotanalysis of serum and neuronal erythropoietin showed that neuronalerythropoietin was smaller and had a different sialic acid profile, which couldaccount for higher receptor affinity and greater biological activity.[36]Similar to peripheral erythropoietin, recombinant human erythropoietin-bindingaffinity to the erythropoietin receptor on neuronal cells is higher than onerythroid cells.[36] The impact of this finding on dosing strategies to elicitrequired response is unknown.

Immunofluorescence studies also revealed that EPO-R islocalized on the membrane of neuronal cells, and that an additional soluble formof the receptor is localized in the cytoplasm.[50,51] Comparison betweenmembrane-associated EPO-R and soluble EPO-R (sEPO-R) revealed that the sEPO-Rwas smaller in size.[51] The physiologic function of sEPO-R has not beencharacterized. However, the possibility that sEPO-R levels are inversely linkedto treatment response appears to be a logical hypothesis.

EPO/EPO-R in CNS:In Vivo Evidence

In vivo and in vitro data from erythroid progenitors andneuronal cell lines show that erythropoietin receptor expression regulates thebiological effectiveness of erythropoietin. In contrast to erythroid progenitorcells, which express high levels of erythropoietin receptor and are directlyresponsive to erythro-poietin stimulation, neuronal cells EPO-R expression is up-regulated in response to hypoxia or anemic stressindicating an increase in neuronal cell sensitivity to erythropoietin underinjury.[52] Other data from animal disease models further indicatetissue-specificity and developmental regulation of erythropoietin/EPO receptor expression in the brain.[53,45,54]

Following focal permanent ischemia, erythropoietin anderythropoietin receptor expression varies with time. After occlusion,erythropoietin localizes in endothelial cells, microglia/macrophage-like cells,and reactive astrocytes.[55] In all instances, erythropoietin receptorexpression was shown to always precede that of erythropoietin for each celltype. These results suggest that erythropoietin and erythropoietin receptor areactively formed at the time a focal cerebral infarct develops, and that thereexists an erythropoietin/EPO-R response system to injury that is likely to havea neuroprotective role following ischemia.

The neuroprotective activity of erythropoietin has beeninvestigated in stroke-prone spontaneously hypertensive rats with permanentocclusion of the left middle cerebral artery,[56] as well as in gerbils withmild to lethal ischemic damage.[39] Place navigation testing of rats in theMorris water maze indicated that rHuEPO infusion into the cerebroventriclesalleviated the ischemia-induced navigation disability and also preventedneuronal death at the infarct foci.[56] Consistent with in vitro data, in situhybridization indicated that following occlusion, erythropoietin receptor mRNAwas up-regulated at the site of injury, again suggesting that there is anerythropoietin/EPO-R-mediated response to injury, which in turn might helpminimize tissue damage.

Similarly, infusion of recombinant erythropoietin into thelateral ventricles of gerbils rescued hippocampal CA1 neurons destined forapoptosis and prevented ischemia-induced learning disability compared tosham-operated, saline-treated control animals. In parallel experiments, infusionof sEPO-R resulted in neuronal degeneration and impaired learning abilitycompared to control animals. Infusion of denatured sEPO-R did not elicit the same neurodegeneration/learning deficits, suggestingthat complex formation between infused sEPO-R and endogenous brainerythropoietin inhibits the erythropoietin/EPO-R interaction that mediatesneuroprotection[39] (see Figure 4). This set of experiments provides the mostcompelling evidence of the possible role of unbound endogenous erythropoietin inbrain interstitial space in promoting neuronal survival. This leads to theconsideration of the mechanism of action of erythropoietin in the CNS.

What Is the Mechanism of Action Underlying Erythropoietin NeuroprotectiveActivity?

Is There a Link Between CNS and Peripheral EPO Regulation?

Erythropoietin production is hypoxia-inducible. Oxygensensors play a prominent role in maintaining an optimal level of oxygen partialpressure (pO2)in various organs by matching oxygen demand and oxygen supply (see Table1). Themolecular mechanism of oxygen-sensing has not been fully elucidated; however, ahemeprotein that does not participate in the mitochondrial energy production hasbeen implicated.[57] Studies with erythropoietin-producing HepG2 cells suggestthat reactive cytochrome beta-generated oxygen species, in direct correlationwith cellular pO2,serve as ancillary messengers of the oxygen sensing signal cascade determiningthe stability of transcription factors and/or the gating of ion channels.Consequently, hypoxia as well as other metabolic disturbances, includinghypoglycemia and strong neuronal depolarization, generate mitochondrial reactiveoxygen species that may increase CNS erythropoietin expression through hypoxia-inducible factor-1 (HIF-1).[58]

In contrast, neuronal injury is characterized byerythropoietin receptor expression preceding that of erythropoietin,[55] asdiscussed above. Erythropoietin may thus protect nervous tissue under anycondition characterized by a relative deficiency of ATP in the face of increasedmetabolic demands. The mechanism may ultimately involve maturation andup-regulation of the expression of erythropoietin receptor and the Bcl-2 familyof antiapoptotic proteins, which ensure optimal survival, proliferation, anddifferentiation of cells.[59,60] Bcl-2 family proteins are likely key effectorsof growth factor receptor-mediated survival signals. The balance of Bcl-2family antiapoptotic (Bcl-2, Bcl-xL, A1 and MCL1) and proapoptotic (bax, Bad,Bak, and Bcl-xS) proteins has been shown to be a critical determinant for cellsurvival, proliferation, and differentiation of studied cells.[61,62]

Investigations of whether endogenous erythropoietin andexogenously administered recombinant erythropoietin facilitation of neuronalsurvival is through up-regulation of Bcl-xL have been performed. Recombinanterythropoietin (1 mU/mL) was shown to up-regulate Bcl-xL mRNA and proteinexpression in cultured neurons. Infusion of recombinant erythropoietin (5 U/dfor 28 days) caused significantly more intense expressions of Bcl-xL mRNA andprotein in the hippocampal CA1 field of ischemic gerbils than did vehicleinfusion. These findings suggest that erythropoietin prevents delayed neuronaldeath in the hippocampal CA1 field, possibly through up-regulation of Bcl-xL,which is known to facilitate neuron survival. In vivo, infusion ofrecombinant erythropoietin (5 U/d for 28 days into the cerebroventricles) ingerbils following ischemia showed intense expression of Bcl-xL mRNA and proteinin the hippocampal CA1 field of animals treated with rHuEPO compared to animalsinfused with saline.[60] These findings suggest that recombinant erythropoietinneuroprotection was likely through the up-regulation of Bcl-xL.

The involvement of Bcl-xL in erythropoietin-mediatedantiapoptotic activity has been further implicated by studies that demonstratedthat binding of erythropoietin to its receptor activates signal transducers andactivators of transcription (Stat) proteins.[63] Stats are important mediatorsof cytokine and growth factor-induced signal transduction and have been shownto play a role in survival and proliferation of hematopoietic cells both invitro and in vivo and to contribute to the growth and viability of cells.

Fetal anemia and increased apoptosis of fetal liver erythroidprogenitors were found in Stat5a(-/-)5b(-/-) mice. Adult Stat5a(-/-)5b(-/-) micewere shown to be deficient in generating high erythropoietic rates in responseto stress even though they had near-normal hematocrit, and to have persistentanemia despite a marked compensatory expansion in their erythropoietictissue.[63] These findings are likely explained by noting that inStat5a(-/-)5b(-/-) mice, there was an initial increase in early erythroblastmass with the majority failing to progress to differentiation. Degree of anemia,increased apoptosis of early erythroblasts, and decreased expression of Bcl-xLwere all shown to be correlated.[64]

The interaction between stem cell factor and its specificreceptor c-Kit, erythropoietin receptor, Stat5, and Bcl-xL, is clarifiedsomewhatby studies using an erythroid progenitor cell line from mice deficient in thehemapoietic lineage-specific transcription factor GATA-1.[59] GATA-1 isexpressed in pluripotent progenitor cells prior to commitment to a singlelineage. GATA-1 deficiency is embryonic lethal at the yolk sac stage anddisruption of GATA-1 produces maturation arrest late in erythroiddevelopment.[64]

Stem cell factor, c-Kit, erythropoietin, erythropoietinreceptor, and Stat5 are involved in the survival and proliferation of erythroidprogenitors via the regulation of Bcl-2 expression. Stem cell factor stimulationof c-Kit was shown to be essential for erythropoietin receptor and Stat5 proteinexpression maintenance, which results in significantly enhanced Bcl-xL inductionand survival of erythroid progenitors in response to erythropoietin stimulation.Similar results were observed upon restoration of GATA-1 in the presence oferythropoietin alone, thus demonstrating that both erythropoietin and stem cellfactor regulate the erythropoietin receptor-Stat5-Bcl-xL pathway, where c-Kitand EPO-R have distinct temporal roles during erythroid maturation.

Separate studies have produced opposing data for Bcl-xLinvolvement in erythropoietin-mediated anti-apoptotic activity. These studiesdemonstrated that human erythroid colony-forming cells undergo rapid apoptosisin the absence of stem cell factor and erythropoietin. Erythropoietin and stemcell factor synergistically activate mitogen-activated protein kinase to promotegrowth and maintain survival of these cells.[65] Apoptosis was partiallyinhibited when either recombinant erythropoietin or stem cell factor was addedto cells, and was totally prevented in the presence of both factors.

Treatment of erythroid cells with blockers of specific kinasesignal transduction pathways demonstrated that both erythropoietin and stem cellfactor induced activation of phosphoinositide 3-kinase (PI3K) where stem cellfactor caused activation of protein kinase B (PKB), an anti-apoptosis signal,while erythropoietin led to activation of extracellular signal-regulated kinases(ERKs). Activation of ERKs was correlated with the expression of theantiapoptotic protein Bcl-XL, which is consistent with the preceding discussion.However, addition of a specific inhibitor of mitogen-activated protein kinase/ERKkinase (MEK), PD98059, inhibited cell growth but had no effect on theantiapoptotic activity of either stem cell factor or erythropoietin, suggestingthe following:

  • Stem cell factor and erythropoietin support survival andgrowth of human erythroid colony-forming cells through different signalingpathway;

  • Stem cell factor and erythropoietin transduce distinctlydifferent signals through activation of PI3K; and

  • The overexpression of Bcl-XL correlates witherythropoietin-mediated activation of ERKs is coincidental and is likely notlinked to prevention of apoptosis of human erythroid colony-forming cells.

The effects of erythropoietin on Ca2+ uptake, membrane potential, cell survival,release, and biosynthesis of dopamine and nitric oxide have also beeninvestigated in neuronal cell lines. Recombinant HuEPO(1) induced membrane depolarization, (2) increased survival of cells culturedwithout serum and nerve growth factor, (3) increased dopamine release, and (4)increased nitric oxide production and dose-dependent 45Ca2+uptake in differentiated PC12 cells. These effects of recombinant erythropoietinwere all inhibited by nicardipine (a Ca2+channel blocker) or antierythropoietin antibody.[66]

Separate studies investigating the effects of erythropoietinon neurosecretion in clonal rat PC12 cells showed similar results witherythropoietin and an erythropoietin mimetic peptide, EMP1. These studiesdemonstrated that erythropoietin suppresses neurotransmitter release throughactivation of erythropoietin receptor linked to JAK2.[67] Erythropoietin wasalso shown to protect neurons from glutamate toxicity.[44] Taken together, theseresults suggest that erythropoietin interacts with neuronal cells by affectingCa2+homeostasis and erythropoietin stimulates neuronal function and viability viaactivation of Ca2+ channels.

Data continue to be generated pertaining to the mechanism ofaction of erythropoietin and erythropoietic agents in the CNS and periphery (seeFigure 5). While in vitro studies are a useful tool, studies with cell lines arefraught with reprogramming risks where cells acquire, reacquire, or losemolecular characteristics that they do not normally possess in vivo, at the timeof isolation or following different culture manipulations. Thus efforts toconfirm in vitro findings in vivo to determine potential clinical significanceare imperative.

Do Exogenously Administered Erythropoietic Agents Cross the Blood-BrainBarrier/Blood-Cerebrospinal Fluid Barrier?

Endogenous erythropoietin protein is detectable in thecerebrospinal fluid (CSF) of human neonates and adults. Previous studies showthat the rank order of erythropoietin levels at basal state in the CSF inneonates with asphyxia > neonates with intraventricular hemorrhage >preterm and term neonates > infants > normal adults with depression >adults with traumatic brain injury. Treatment of neonates (1,200 U/kg/wksubcutaneous [SC] or 1,400 U/kg/wk intravenous [IV] or 6,000 U rHuEPO IV) didnot result in elevated erythropoietin concentrations in the cerebral spinalfluid.[68,69] However, simultaneous serum and CSF samples showed that, inpatients with traumatic injury, erythropoietin concentrations correlate with thedegree of blood-brain barrier dysfunction.[43] Sampling times, patientconditions, and analysis methods vary widely in the studies presented in Table 2and data are only presented for qualitative comparison[69a].

Local injection of recombinant erythropoietin into thelateral ventricles was shown to:

  • Promote the survival of septal cholinergic neurons inadult rats subjected to fimbria-fornix transection[70,49];

  • Attenuate place navigation disability and corticalinfarction induced by permanent occlusion of the middle cerebral artery[56]; and

  • Rescue neurons destined for apoptosis in a gerbil ischemiamodel.[39]

Interestingly, and contrary to human studies which suggestthat neither endogenous erythropoietin nor recombinant erythropoietin cross theintact blood-brain barrier, separate studies suggest the direct effect ofsystemically administered recombinant erythropoietin (intraperitoneal and subcutaneous) in stroke,blunt trauma, acute experimental autoimmune encephalitis (EAE)rat model,[3] and subarachnoid hemorrhage rat/rabbit models.[71] Available dataare summarized in Table 2.[71a-71e]

Doses, dosing routes, and visualization/drug levelmeasurement tools used in these studies confound critical evaluation of observedresults and have had limited successful replication by other laboratories.[6]Nonetheless, at the bare minimum, data from experimental animal models of CNSinjury/disease provide phenomenologic evidence supporting the activity oferythropoietic agents in reducing/treating CNS injury and improving cognitivefunction. Hence, further exploration of erythropoietic agents as potentialcandidates for treatment of CNS disease of various pathophysiological originsshould be encouraged, albeit using well-validated procedures (with appropriatecontrols) that enable study comparison.

What Are the Potential Strategies for Using Erythropoietic Agents asNeurotherapeutic Agents?

Strategies for using erythropoietic agents asneurotherapeutic drugs must include an understanding of the optimal time fortherapeutic intervention and target therapeutic drug concentrations and relateddose requirements for reliable response. The latter requirements are the focusof the remainder of this article, including important principles of thebiodistribution of drugs into the CNS following intrathecal/cerebroventricularand systemic drug administration and available strategies for developingeffective erythropoietic agent-based therapies using systemic administration.

Occasionally, the plasma concentration vs time profile may bemodified to optimize drug concentrations at the site of drug activity. However,drugs administered systemically often have poor access to the CNS because of theblood-brain and blood-CSF barriers. Thus, drug development and interventiontherapy development for the CNS must also deal with the challenge of achievingeffective concentrations at the site of action.

Overview of Blood-Brain and Blood-CSF Barrier Transport Biology

Parenteral administration routes would be preferable for theintervention of CNS injury with erythropoietic agents because of the potentialof general exposure of the drug to the entire CNS. Currently used parenteralroutes for erythropoietic agents are intravenous, subcutaneous, andintraperitoneal. The biodistribution of these drugs throughout the body dependson the rates of absorption at the site of administration, distribution to thesite of action, metabolism and excretion from the blood, as well as the extentof binding to plasma proteins.

The major advantages of intravenous drug administration arerapid and complete absorption. However, unlike other parts of the body,specialized barriers regulate drug distribution from the blood into protectedcompartments like the prostate, eye, and CNS. These are privileged sites inwhich the concentration vs time profile may differ substantially from thatobserved for the plasma.[72] Consequently, understanding drug distributionacross the blood-brain and blood-CSF barriers is of crucial importance for thedevelopment of erythropoietic drugs as neuroprotective agents.

The blood-brain barrier (BBB), blood-CSF barrier (B-CSFB),and blood-ocular barrier (BOB) share similar embryological origin, microanatomy,and many physiologic functions.[72] Structurally, the barriers consist of tightjunctions between endothelial cells. Functionally, endothelial cell barriersregulate transfer of sugars, amino acids, organic acids, and ions according tomolecular size, protein-binding affinity, lipophilicity, and degree ofionization at the relevant anatomical compartment pH.[73-76] Furthermore, activeand/or saturable transport systems and enzymatic degradation contribute to theselectivity of these barriers and regulate the effective penetration of avariety of small molecule and protein therapeutic agents.[77,78]

More recently, the BBB has been shown to be an importantcommunication interface between the CNS and peripheral tissues thorough itscontrol of the exchange of peptides and regulatory proteins.[79,80]

Studies have also shown that the penetration of severalantimicrobial agents is similar in both the eye and CSF following systemic drugadministration.[81] This observation may have practical as well as theoreticalimplications since, in the absence of data for site-specific pharmacokinetics, vitreous penetration and CSF penetration may serve as pharmacokineticsurrogates for each other for some molecules.

Drug translocation rates are dependent on the functionalanatomy and physiology of the CNS as well as pharmacokinetics in the blood. Drugtransfer into the CNS may occur by passive diffusion, active, and/or saturablesystem transport. Ultimately the concentrations of drug achieved in the CSF orparenchyma depend upon competing rate constants describing uptake and efflux.There is a sidedness to these transport mechanisms, which is demonstrated bydifferent influx and efflux rate constants.

The distribution of erythropoietic agents into the CSF is ofprimary importance for overall cerebral biodisposition of drug since the CSF isthe major pathway for drug distribution to different targets in the centralnervous system. Drug access to the CSF is regulated by the choroid plexuses.[82]The endothelial cells of the BBB and choroid plexus endothelial cells of theblood-cerebrospinal fluid barrier share structural and functional similarities.Specific morphological properties of the choroidal epithelium demonstrate itsrole in protecting the brain by enzymatic degradation of exogenous and endogen-
ous toxins and drugs, and by vectorial clearance of neurotoxins/metabolites intothe blood via transport proteins of the multidrug resistance family.

Immunohistochemical studies have demonstrated thaterythropoietin receptor is present around brain capillaries.[3,6] Transmissionelectron microscopy further revealed EPO-R positive expression within astrocyticend-feet surrounding capillaries and on capillary endothelial cells.[3] Thesefindings have led to the hypothesis that erythropoietic agents may bepreferentially transported across the blood-brain barrier via erythropoietinreceptor.[3] This hypothesis presumes that transport of these agents is byspecific receptor-mediated transcytosis, which requires the binding of a ligandto its specific membrane receptor and its transport across the blood-brainbarrier within vesicles, or by other saturable processes, which do not requirevesicles.[83-85]

Specific receptors have been identified for various proteinson brain capillaries, including insulin,[86] insulin-like growth factor-1(IGF-1),[87] transferrin,[88] interleukin-1 (IL-1),[89] leptin, [90] andlow-density lipoprotein.[91] Saturable transport across the blood-brain barrierhas been demonstrated for insulin,[92] transferrin,[88] low-densitylipoprotein,[91] IgG,[93] angiotensin II,[94] leptin, arginine vasopressin,neurotrophins, tumor necrosis factor, pancreatic polypeptide, interleukin-1, andmany other peptides and regulatory proteins. [89,95-99] Most noteworthy is thefinding that protein size does not appear to be a limiting factor inreceptor-mediated transport. Insulin is a relatively small protein at 6,000 Dacompared to transferrin, which has a molecular weight of 77,000 Da.

Blood-Brain Barrier/Blood CSF Barrier Permeability to Darbepoetin Alfa andEpoetin Alfa

There are common characteristics of proposed proteintransport systems across the blood-brain barrier/blood-CSF barrier.

  • Ligands are transported based on their affinity to and thecapacity of the particular receptor.

  • Transport is dictated by the ligand’s physicochemicalcharacteristics including size, mass, and charge.[85]

Alternative transport mechanisms for therapeutic proteinaccess into privileged compartments include absorptive-transcytosis, which maynot involve specific membrane receptors, cytoplasmic-shuttle systems like "argosomes,"which have been shown to transport small proteins across epithelia in Drosophila,[100]and non-specific leakage. In particular, glycoproteins, of which erythropoietinis one, are able to cross the blood-brain barrier by means of adsorptivetranscytosis.[101,102]

In light of this information, it has been postulated thatmass, size, charge, and apparent receptor affinity differences in thepeptide-carbohydrate forms between darbepoetin alfa and epoetin alfa might leadto differences in blood-CSF permeation.

The permeation of darbepoetin alfa and epoetin alfa into thecerebrospinal fluid was investigated in rats dosed intravenously with epoetinalfa (5,000 U/kg ~ 25 µg/kg,n = 12) or darbepoetin alfa (25 µg/kg,n = 12). Matched serum samples (collected by terminal cardiac puncture) and CSFsamples (from the posterior fossa) were collected predose and at 0.5, 2, 4, 6,and 8 hours postdose, n = 2/time-point. Darbepoetin alfa and rHuEPOconcentrations were determined by ELISA, and sample protein integrity incollected CSF and serum samples was confirmed by sodium dodecyl sulphatepolyacrylamide gel electrophoresis (SDS-PAGE).

Both darbepoetin alfa andrHuEPO were measurable in rat CSF up to 8 hours after IV administration of thesehigh doses (see Figure 6). The protein integrity of darbepoetin alfa and rHuEPOin collected serum and CSF samples was determined by SDS-PAGE—a single bandfor each sample was taken as an indication of an intact protein. The resultsshowed that the protein concentration determined by the immunoassay was fromintact darbepoetin alfa or rHuEPO. The calculated area under theconcentration-time curve (AUC(0-8)),by noncompartmental analysis, was (mean ± SE) 4,500 ± 310 ng h/mL in serum for darbepoetin alfa vs 370,000 ± 13,000 mU h/mL in serum for rHuEPO, and 3.6 ± 0.07ng h/mLin CSF for darbepoetin alfa vs 340 ± 40 mU ´ h/mL in CSF for epoetin alfa (see Table 3). The ratio of CSF AUC:serum AUCprovides an indication of CSF penetration. Thus, penetration into the CSF forsystemically administered drug (95% confidence interval) for darbepoetin alfawas mean 0.079% (0.081%-0.084%) and for rHuEPO was mean 0.089% (0.081%-0.099%).

Albumin (66,000 Da) is generally excluded from the CNS by theblood-CSF barrier and is used as a marker for barrier intactness. In our study,the penetration of both rHuEPO and darbepoetin alfa into the CSF was of asimilar magnitude to that seen with albumin[103] and, therefore, it is possiblethat a similar nonreceptor mediated mechanism is involved (see Table3). Ourdata are consistent with results from monkey data,[71e] which indicate similarfindings (penetration ranging from 0.03% to 0.22%). These data do not provideany evidence of a saturable transport mechanism of either darbepoetin alfa orrHuEPO into CSF. However, they show that mass, size, charge, and apparentreceptor affinity differences in the peptide-carbohydrate forms betweendarbepoetin alfa and rHuEPO do not appear to lead to differences in blood-CSFpermeation for these drugs. Further investigations to explore darbepoetin alfa/epoetinalfa blood-brain barrier translocation and to determine mechanisms of transportare in progress.

Do Erythropoietic Agents Need to Cross the Blood Brain Barrier in Order toElicit Central Therapeutic Effects?

The effect of hyperbaric oxygen in a rat middle cerebralartery occlusion/reperfusion (MCAO) model showed that such therapy appliedshortly (6 hours) after reperfusion significantly reduced infarct area comparedto untreated animals.[104] Separate studies showed that hyperbaric oxygenpreconditioning induced ischemic tolerance in transient but not permanent MCAOrats.[105] The role of improved tissue oxygenation in dampening/reversing CNSdamage has also been demonstrated in a limited number of patients as shown bysuccessful hyperbaric oxygen therapy in hemorrhagic stroke patients.[106]Further substantiation of the importance of tissue oxygen in CNS outcomefollowing traumatic injury has been demonstrated by findings withhemoglobin-based oxygen carriers that show improved outcome in animal models andpatients with impaired perfusion (eg, stroke or myocardial infarction).[107]Taken together, these results suggest that oxygenation and cell rescue fromapoptosis determine outcomes following CNS injury. Thus systemicallyadministered exogenous erythropoietic agents may act indirectly on the CNS byimproving general oxygenation, while local erythropoietin response to injuryrescues cells from apoptosis.

Erythropoietic Agent Strategies for CNS Intervention

Erythropoietic agents show promise for providing alternatetherapeutic approaches for CNS injury and cognitive dysfunction. Experimentalrodent disease model data demonstrate the neuroprotective/neurotrophic effectsof erythropoietic agents administered locally or systemically. Clinical datafrom controlled trials supporting the clinical application of erythropoieticagents as neuroprotective agents are currently not available. However, there arenumerous anecdotal reports of positive neurocognitive effects of these agentsfrom studies conducted to prevent or treat anemia.[108,109] Thus, the currentlevel of interest in determining possible clinical benefits of the interventionof erythropoietic agents in CNS injury is warranted.

Barriers for realizing the full benefits of erythropoieticagent application in CNS injury/disease likely exist, especially if this noveluse for these agents becomes just "another therapeutic application."Application of erythropoietic agents for neuroprotection/neutrophic effect hasmuch to gain from as well as much to contribute to our growing understanding ofthe mechanisms underlying erythropoietin and erythropoietic agent biologicactivity throughout the body.

It also remains unknown to what extent the blood-brainbarrier/blood-CSF barrier will restrict the clinical benefit of systemicallyadministered erythropoietic agents. Blood-brain barrier/blood-CSF barrierpermeability limitations become important only when data shows unequivocallythat direct interaction of exogenous drug and cells at the site of drug actionis necessary for intervention rather than indirect mechanisms of action (tissueoxygenation). Incidentally, disease or injury may increase the permeability ofthe blood-brain barrier/blood-CSF barrier to systemically administerederythropoietic agents, allowing therapeutic drug levels in the CNS to beachieved readily.

Finally, the blurring of clinical discipline boundaries, bythe widening potential therapeutic spectrum of erythropoietic agents, provides aunique opportunity for cross-discipline collaboration to expedite the process ofbringing the benefits of these agents to the clinic. In this way, the continuedexploration of erythropoietic agents will enrich and benefit frominterdisciplinary knowledge exchange, while simultaneously providing new avenuesfor the treatment of injury/disease and improvement of quality of life forpatients.


1. Sasaki R, Masuda S, Nagao M: Erythropoietin: Multiplephysiological functions and regulation of biosynthesis. Biosci BiotechnolBiochem 64(9):1775-1793, 2000.

2. Sasaki R, Masuda S, Nagao M: Pleiotropic functions andtissue-specific expression of erythropoietin. News Physiol Sci16:110-113, 2001.

3. Brines ML, Ghezzi P, Keenan S, et al: Erythropoietincrosses the blood-brain barrier to protect against experimental brain injury. ProcNatl Acad Sci 97(19):10526-10531, 2000.

4. Tong EM, Nissenson AR: Erythropoietin and anemia. SeminNephrol 21(2):190-203, 2001.

5. Mulcahy L: The erythropoietin receptor. Semin Oncol 2(suppl8):19-23, 2001.

6. Dame C, Juul SE, Christensen RD: The biology oferythropoietin in the central nervous system and its neurotrophic andneuroprotective potential. Biol Neonate 79:228-235, 2001.

7. Cerami A, Brines M, Ghezzi P, et al: Neuroprotectiveproperties of epoetin alfa. Nephrol Dial Transplant 17(suppl 1):8-12,2002.

8. Recny MA, Scoble HA, Kim Y: Structural characterization ofnatural human urinary and recombinant DNA-derived erythropoietin. Identificationof des-arginine 166 erythropoietin. J Biol Chem 262(35):17156-17163,1987.

9. Nissenson AR, Nimer SD, Wolcott DL: Recombinant humanerythropoietin and renal anemia: Molecular biology, clinical efficacy, andnervous system effects. Ann Intern Med 114(5):402-416, 1991.

10. Eschbach JW, Haley NR, Adamson JW: The anemia of chronicrenal failure: Pathophysiology and effects of recombinant erythropoietin. ContribNephrol 78:24-37, 1990.

11. Besarab A: Optimizing epoetin therapy in end-stage renaldisease: The case for subcutaneous administration. Am J Kidney Dis 2 (suppl1):13-22, 1993.

12. Koury ST, Koury MJL: Erythropoietin production by thekidney. Semin Nephrol 13(1):78-86, 1993.

13. Egrie JC, Browne JK: Development and characterization ofnovel erythropoiesis stimulating protein (NESP). Nephrol Dial Transplant16(suppl 3):3-13, 2001.

14. Boulay JL, Paul WE: The interleukin-4-related lymphokinesand their binding to hematopoietin receptors. J Biol Chem 267(29):20525-20528,1992.

15. D’Andrea AD, Fasman GD, Lodish HF: Erythropoietinreceptor and interleukin-2 receptor beta chain: A new receptor family. Cell58(6):1023-1024, 1989.

16. Constantinescu SN, Ghaffari S, Lodish HF: Theerythropoietin receptor: Structure,activation and intracellular signal transduction. Trends Endocrinol Metab 10(1):18-23,1999.

17. D’Andrea AD, Yoshimura A, Youssoufian H, et al: Thecytoplasmic region of the erythropoietin receptor contains nonoverlappingpositive and negative growth-regulatory domains. Mol Cell Biol 11(4):1980-1987,1991.

18. Lacombe C, Mayeux P: Biology of erythropoietin. Haematologica 83(8):724-732, 1998.

19. Billat CL, Felix JM, Jacquot RL: In vitro and in vivoregulation of hepatic erythropoiesis by erythropoietin and glucocorticoids inthe rat fetus. Exp Hematol 10(1):133-140, 1982.

20. Hisakawa H, Sugiyama D, Nishijima, et al: Humangranulocyte-macrophage colony-stimulating factor (hGM-CSF) stimulates primitiveand definitive erythropoiesis in mouse embryos expressing hGM-CSF receptors butnot erythropoietin receptors. Blood 98(13):3618-3622, 2001.

21. Detmer K, Walker AN: Bone morphogenetic proteins actsynergistically with haematopoietic cytokines in the differentiation ofhaematopoietic progenitors. Cytokine 17(1):36-42, 2002.

22. Jegalian AG, Acurio A, Dranoff G, et al: Erythropoietinreceptor haploinsufficiency and in vivo interplay with granulocyte-macrophagecolony-stimulating factor and interleukin 3. Blood 99(7):2603-2605,2002.

23. Lasne F, de Ceaurriz J: Recombinant erythropoietin inurine. Nature 405(6787):635, 2000.

24. Wide L, Bengtsson C, Berglund B, et al: Detection inblood and urine of recombinant erythropoietin administered to healthy men. MedSci Sports Exerc 27(11):1569-1576, 1995.

25. Gareau R, Brisson GR, Chenard C, et al: Total fibrin andfibrinogen degradation products in urine: A possible probe to detect illicitusers of the physical-performance enhancer erythropoietin? Horm Res44(4):189-192, 1995.

26. Macdougall IC, Gray SJ, Elston O, et al: Pharmacokineticsof novel erythropoiesis stimulating protein compared with epoetin alfa indialysis patients. J Am Soc Nephrol 10:2392-2395, 1999.

27. Heatherington AC, Cosenza ME, Watson A, et al:Establishment of a PK-PD relationship for novel erythropoiesis stimulatingprotein (NESP) in dogs. Pharmaceut Sci Suppl 1,1999.

28. Pirker R, Vansteenkiste J, Gateley J, et al: A phase III,double-blind, placebo-controlled randomized study of novel erythropoiesis-stimulatingprotein (NESP) in patients undergoing platinum treatment for lung cancer. EurJ Cancer 37(6):264, 2001.

29a. Glaspy J, Jadeja JS, Justice G, et al: A dose-findingand safety study of novel erythropoiesis stimulating protein (NESP) for thetreatment of anaemia in patients receiving multicycle chemotherapy. Br JCancer 84 (suppl 1):17-23, 2001.

29b. Glaspy JA, Jadeja JS, Justice G, et al: Darbepoetin alfagiven every 1 or 2 weeks alleviates anaemia associated with cancer chemotherapy.Br J Cancer 87(3):268-276, 2002.

30. Kotasek D, Albertsson M, Mackey J, et al: Randomized,double-blind, placebo-controlled, dose-finding study of darbepoetin alfaadministered once every 3 (Q3W) or 4 (Q4W) weeks in patients with solid tumors(abstract 1421). Proc Am Soc Clin Oncol 21:356a, 2002.

31. Wrighton NC, Farrell FX, Chang R, et al: Small peptidesas potent mimetics of the protein hormone erythropoietin. Science273(5274):458-464, 1996.

32. McConnell SJ, Dinh T, Le MH, et al: Isolation oferythropoietin receptor agonist peptides using evolved phage libraries. BiolChem 379(10):1279-1286, 1998.

33. Naranda T, Wong K, Kaufman RI, et al: Activation oferythropoietin receptor in the absence of hormone by a peptide that binds to adomain different from the hormone binding site. Proc Natl Acad Sci U S A96(13):7569-7574, 1999.

34. Qureshi SA, Kim RM, Konteatis Z, et al: Mimicry oferythropoietin by a nonpeptide molecule. Proc Natl Acad Sci U S A96(21):12156-12161, 1999.

35. Anagnostou A, Lee ES, Kessimian N, et al: Erythropoietinhas a mitogenic and positive chemotactic effect on endothelial cells. ProcNatl Acad Sci U S A 87(15):5978-5982, 1990.

36. Masuda S, Nagao M, Takahata K, et al: Functionalerythropoietin receptor of the cells with neural characteristics. Comparisonwith receptor properties of erythroid cells. J Biol Chem268(15):11208-11216, 1993.

37. Ribatti D, Presta M, Vacca A, et al: Human erythropoietininduces a pro-angiogenic phenotype in cultured endothelial cells and stimulatesneovascularization in vivo. Blood 93(8):2627-2636, 1999.

38. Yasuda Y, Masuda S, Chikuma M, et al: Estrogen-dependentproduction of erythropoietin in uterus and its implication in uterineangiogenesis. J Biol Chem 273(39):25381-25387, 1998.

39. Sakanaka M, Wen TC, Matsuda S, et al: In vivo evidencethat erythropoietin protects neurons from ischemic damage. Proc Natl Acad Sci U S A 95(8):4635-4640, 1998.

40. Wu H, Liu X, Jaenisch R, et al: Generation of committederythroid BFU-E and CFU-E progenitors does not require erythropoietin or theerythropoietin receptor. Cell 83(1):59-67, 1995.

41. Ogilvie M, Yu X, Nicolas-Metral V, et al: Erythropoietinstimulates proliferation and interferes with differentiation of myoblasts. JBiol Chem 275(50):39754-39761, 2000.

42. Digicaylioglu M, Bichet S, Marti HH, et al: Localizationof specific erythropoietin binding sites in defined areas of the mouse brain. ProcNatl Acad Sci U S A 92(9):3717-3720, 1995.

43. Marti HH, Gassmann M, Wenger RH, et al. Detection oferythropoietin in human liquor: Intrinsic erythropoietin production in thebrain. Kidney Int 51:416-418, 1997.

44. Morishita E, Masuda S, Nagao M, et al: Erythropoietinreceptor is expressed in rat hippocampal and cerebral cortical neurons, anderythropoietin prevents in vitro glutamate-induced neuronal death. Neuroscience 76(1):105-116, 1997.

45. Juul SE, Yachnis AT, Christensen RD: Tissue distributionof erythropoietin and erythropoietin receptor in the developing human fetus. EarlyHum Dev 52(3):235-249, 1998.

46. Pleasure SJ, Lee VM: NTera 2 cells: A human cell linewhich displays characteristics expected of a human committed neuronal progenitorcell. J Neurosci Res 35(6):585-602, 1993.

47. Greene LA, Tischler AS: Establishment of a noradrenergicclonal line of rat adrenal pheochromocytoma cells which respond to nerve growthfactor. Proc Natl Acad Sci U S A 73(7):2424-2428, 1976.

48. Hammond DN, Wainer BH, Tonsgard JH, et al: Neuronalproperties of clonal hybrid cell lines derived from central cholinergic neurons.Science 234(4781):1237-1240, 1986.

49. Tabira T, Konishi Y, Gallyas F Jr: Neurotrophic effect ofhematopoietic cytokines on cholinergic and other neurons in vitro. Int J DevNeurosci 13(3-4):241-252, 1995.

50. Assandri R, Egger M, Gassmann M, et al: Erythropoietinmodulates intracellular calcium in a human neuroblastoma cell line. J Physiol 516(Pt 2):343-352, 1999.

51. Nagao M, Masuda S, Abe S, et al: Production and ligand-bindingcharacteristics of the soluble form of murine erythropoietin receptor. BiochemBiophys Res Commun 188:888-897, 1992.

52. Sinor AD, Greenberg DA: Erythropoietin protects culturedcortical neurons, but not astroglia, from hypoxia and AMPA toxicity. NeurosciLett 290(3):213-215, 2000.

53. Liu C, Shen K, Liu Z, et al: Regulated humanerythropoietin receptor expression in mouse brain. J Biol Chem 272(51):32395-32400,1997.

54. Dame C, Bartmann P, Wolber E, et al: Erythropoietin geneexpression in different areas of the developing human central nervous system. BrainRes Dev Brain Res 125(1-2):69-74, 2000.

55. Bernaudin M, Marti HH, Roussel S, et al: A potential rolefor erythropoietin in focal permanent cerebral ischemia in mice. J Cereb BloodFlow Metab 19(6):643-651, 1999.

56. Sadamoto Y, Igase K, Sakanaka M, et al: Erythropoietinprevents place navigation disability and cortical infarction in rats withpermanent occlusion of the middle cerebral artery. Biochem Biophys Res Commun 253(1):26-32, 1998.

57. Porwol T, Ehleben W, Brand V, et al: Tissue oxygen sensorfunction of NADPH oxidase isoforms, an unusual cytochrome aa3 and reactiveoxygen species. Respir Physiol 128(3):331-348, 2001.

58. Chandel NS, Maltepe E, Goldwasser E, et al: Mitochondrialreactive oxygen species trigger hypoxia-induced transcription. Proc NatlAcad Sci U S A 95(20):11715-11720, 1998.

59. Kapur R, Zhang L: A novel mechanism of cooperationbetween c-Kit and erythropoietin receptor. Stem cell factor induces theexpression of Stat5 and erythropoietin receptor, resulting in efficientproliferation and survival by erythropoietin. J Biol Chem 276(2):1099-1106,2001.

60. Wen TC, Sadamoto Y, Tanaka J, et al: Erythropoietinprotects neurons against chemical hypoxia and cerebral ischemic injury byup-regulating Bcl-xL expression. J Neurosci Res 67(6):795-803, 2002.

61. Oltvai ZN, Korsmeyer SJ: Checkpoints of dueling dimersfoil death wishes. Cell 79(2):189-192, 1994.

62. Chao DT, Korsmeyer SJ: BCL-2 family: Regulators of celldeath. Annu Rev Immunol 16:395-419, 1998.

63. Socolovsky M, Fallon AE, Wang S, et al: Fetal anemia andapoptosis of red cell progenitors in Stat5a-/-5b-/- mice: A direct role forStat5 in Bcl-X(L) induction. Cell 98(2):181-189, 1999.

64. Fujiwara Y, Browne CP, Cunniff K, et al: Arresteddevelopment of embryonic red cell precursors in mouse embryos lackingtranscription factor GATA-1. Proc Natl Acad Sci U S A 93(22):12355-12358,1996.

65. Sui X, Krantz SB, Zhao ZJ: Stem cell factor anderythropoietin inhibit apoptosis of human erythroid progenitor cells throughdifferent signalling pathways. Br J Haematol 110(1):63-70, 2000.

66. Koshimura K, Murakami Y, Sohmiya M, et al: Effects oferythropoietin on neuronal activity. J Neurochem 72(6):2565-2572, 1999.

67. Kawakami M, Iwasaki S, Sato K, et al: Erythropoietininhibits calcium-induced neurotransmitter release from clonal neuronal cells. BiochemBiophys Res Commun 279(1):293, 2000.

68. Juul SE, Harcum J, Li Y, et al: Erythropoietin is presentin the cerebrospinal fluid of neonates. J Pediatr 130:428-430, 1997.

69. Juul SE, Stallings SA, Christensen RD: Erythropoietin inthe cerebrospinal fluid of neonates who sustained CNS injury. Pediatr Res46(5):543-547, 1999.

69a. Buemi M, Allegra A, Corica F, et al: Intravenousrecombinant erythropoietin does not lead to an increase in cerebrospinal fluiderythropoietin concentration. Nephrol Dial Transplant 15:422-423, 2000.

70. Konishi Y, Chui DH, Hirose H, et al: Trophic effect oferythropoietin and other hematopoietic factors on central cholinergic neurons invitro and in vivo. Brain Res 609(1-2):29-35, 1993.

71a. Grasso G, Buemi M, Alafaci C: Beneficial effects ofsystemic administration of recombinant human erythropoietin in rabbits subjectedto subarachnoid hemorrhage. Proc Natl Acad Sci USA 99(8):5627-5631, 2002.

71b. Grasso G, Passalacqua M, Sfacteria A, et al: Doesadministratin of recombinant human erythropoietin attenuate the increase ofS-100 protein observed in cerebrospinal fluid after experimental subarachnoidhemorrhage? J Neurosurg 96(3):565-570, 2002.

71c. Springborg JB, Ma X, Rochat P, et al: A singlesubcutaneous bolus of erythropoietin normalizes cerebral blood flowautoregulation after subarachnoid hemorrhage in rats. Br J Pharmacol 135(3):823-829,2002.

71d. Grasso G: Neuroprotective effect of recombinant humanerythropoietin in experimental subarachnoid hemorrhage. J Neurosurg Sci 45(1):7-14,2001.

71e. Farell F, et al: EPO crosses the BBB: An analysis in anonhuman primate. Blood 98(11):148b, 2000.

72. Barza M, Cuchural G: General principles of antibiotictissue penetration. J Antimicrob Chemother 15(suppl A):59-75,1985.

73. Rabkin MD, Bellhorn MB, Bellhorn RW: Selected molecularweight dextrans for in vivo permeability studies of rat retinal vasculardisease. Exp Eye Res 24(6):607-612, 1977.

74. Oldendorf WH: Stereospecificity of blood-brain barrierpermeability to amino acids. Am J Physiol 224:967-969, 1973.

75. Habgood MD, Begley DJ, Abbott NJ: Determinants of passivedrug entry into the central nervous system. Cell Mol Neurobiol20:231-253, 2000.

76. van de Waterbeemd H, Camenisch G, Folkers G, et al:Estimation of blood-brain barrier crossing of drugs using molecular size andshape, and H-bonding descriptors. J Drug Target 6(2):151-165, 1998;Oldendorf WH: Stereospecificity of blood-brain barrier permeability to aminoacids. Am J Physiol 224:967-969, 1973.

77. Sande MA, Sherertz RJ, Zak O, et al: Factors influencingthe penetration of antimicrobial agents into the cerebrospinal fluid ofexperimental animals. Scand J Infect Dis Suppl 14:160-163, 1978.

78. Liu W, Jumbe N, Kaw P, et al: Comparison of quinolonepharmacokinetics in vitreous humor and CSF following IV drug administration inalbino rabbits. IOVS 40(4):S88, 1999.

79. Banks WA, Kastin AJ: Editorial Review: Peptide transportsystems for opiates across the blood-brain barrier. Am J Physiol259:E1-E10, 1990.

80. Banks WA, Kastin AJ: Passage of peptides across theblood-brain barrier:Pathophysiological perspectives. Life Sci 59(23):1923-1943, 1996.

81. Kaw P, Jumbe N, Liu W, et al: The penetration oftrovafloxacin into the eye and CSF of rabbits (abstract 469-B429). IOVS40:S88, 1999.

82. Ghersi-Egea JF, Strazielle N: Brain drug delivery, drugmetabolism, and multidrug resistance at the choroid plexus. Microsc Res Tech52(1):83-88, 2001.

83. Pardridge WM: Drug delivery to the brain. J CerebBlood Flow Metab 17(7):713-731, 1997.

84. Kastin AJ, Pan W, Maness LM, et al: Peptides crossing theblood-brain barrier: Some unusual observations. Brain Res848(1-2):96-100, 1999.

85. Bickel U, Yoshikawa T, Pardridge WM: Delivery of peptidesand proteins through the blood-brain barrier. Adv Drug Deliv Rev46(1-3):247-279, 2001.

86. Pardridge WM, Eisenberg J, Yang J: Human blood-brainbarrier insulin receptor. J Neurochem 44(6):1771-1778, 1985.

87. Duffy KR, Pardridge WM, Rosenfeld RG: Human blood-brainbarrier insulin-like growth factor receptor. Metabolism 37(2):136-140,1988.

88. Fishman JB, Rubin JB, Handrahan JV, et al:Receptor-mediated transcytosis of transferrin across the blood-brain barrier. JNeurosci Res 18(2):299-304,1987.

89. Banks WA, Ortiz L, Plotkin SR, et al: Human interleukin(IL) 1 alpha, murine IL-1 alpha and murine IL-1 beta are transported from bloodto brain in the mouse by a shared saturable mechanism. J Pharmacol Exp Ther259(3):988-996, 1991.

90. Banks WA: Leptin transport across the blood-brainbarrier: Implications for the cause and treatment of obesity. Pharm Des7(2):125-133, 2001.

91. Dehouck B, Fenart L, Dehouck MP, et al: A new functionfor the LDL receptor: Transcytosis of LDL across the blood-brain barrier. JCell Biol 138(4):877-889, 1997.

92. Baura GD, Foster DM, Porte D Jr, et al: Saturabletransport of insulin from plasma into the central nervous system of dogs invivo. A mechanism for regulated insulin delivery to the brain. J Clin Invest92(4):1824-1830, 1993.

93. Zhang Y, Pardridge WM: Mediated efflux of IgG moleculesfrom brain to blood across the blood-brain barrier. J Neuroimmunol114(1-2):168-172, 2001.

94. Prat A, Biernacki K, Wosik K, et al: cell influence onthe human blood-brain barrier. Glia 36(2):145-155, 2001.

95. Zlokovic BV, Banks WA, El Kadi H, et al: Transport,uptake, and metabolism of blood-borne vasopressin by the blood-brain barrier.Brain Res 590:213-218, 1992.

96. Pan W, Banks WA, Kastin AJ: Permeability of theblood-brain barrier to neurotrophins. Brain Res 788:87-94, 1998.

97. Gutierrez EG, Banks WA, Kastin AJ: Murine tumor necrosisfactor alpha is transported from blood to brain in the mouse. JNeuroimmunol 47:169-177, 1993.

98. Banks WA, Kastin AJ, Huang W, et al: Leptin enters thebrain by a saturable system independent of insulin. Peptides 17:305-311,1996.

99. Banks WA, Kastin AJ, Jaspan JB: Regional variation intransport of pancreatic polypeptide across the blood-brain barrier of mice. PharmacolBiochem Behav 51:139-147, 1995.

100. Greco V, Hannus M, Eaton S: Argosomes: A potentialvehicle for the spread of morphogens through epithelia. Cell 106(5):633-645,2001.

101. Broadwell RD: Transcytosis of macromolecules through theblood-brain barrier: A cell biological perspective and critical appraisal. ActaNeuropathol (Berl) 79:117-128, 1989.

102. Raub TJ, Audus KL: Adsorptive endocytosis and membranerecycling by cultured primary bovine brain microvessel endothelial cellmonolayers. J Cell Sci 97:127-138, 1990.

103. Poduslo JF, Curran GL, Berg CT: Macromolecularpermeability across the blood-nerve and blood-brain barriers. Proc Natl AcadSci U S A 91(12):5705-5709, 1994.

104. Yin W, Badr AE, Mychaskiw G, et al: Down regulation ofCOX-2 is involved in hyperbaric oxygen treatment in a rat transient focalcerebral ischemia model. Brain Res 926(1-2):165-171, 2002.

105. Xiong L, Zhu Z, Dong H, et al: Hyperbaric oxygenpreconditioning induces neuroprotection against ischemia in transient notpermanent middle cerebral artery occlusion rat model. Chin Med J (inEnglish) 113(9):836-839, 2000.

106. Lim J, Lim WK, Yeo TT, et al: Management of haemorrhagicstroke with hyperbaric oxygen therapy—A case report. Singapore Med J42(5):220-223, 2001.

107. Standl T: Haemoglobin-based erythrocyte transfusionsubstitutes. Expert Opin Biol Ther 1(5):831-43, 2001.

108. O’Shaughnessy J, Vukelja S, Savin M, et al: Effects ofepoetin alfa (Procrit) on cognitive function, mood, asthenia, and quality oflife in women with breast cancer undergoing adjuvant or neoadjuvantchemotherapy: A double-blind, randomized, placebo-controlled trial (abstract1449). Proc Am Soc Clin Oncol 21:363a, 2002.

109. Stivelman J: Benefits of anaemia treatment on cognitivefunction. Nephrol Dial Transplant 15(suppl 3):29-35, 2000.

110. Masuda S, Kobayashi T, Chikuma M, et al: The oviductproduces erythropoietin in an estrogen- and oxygen-dependent manner. Am JPhysiol Endocrinol Metab 278(6):E1038-E1044, 2000.

111. Chikuma M, Masuda S, Kobayashi T, et al: Tissue-specificregulation of erythropoietin production in the murine kidney, brain, and uterus.Am J Physiol Endocrinol Metab 279(6):E1242-E1248, 2000.

112. Ehrenreich H, Hasselblatt M, Piotr L, et al:Erythropoietin treatment for acute stroke: A randomized double-blind proof-ofconcept trial in man. 27th International Stroke Conference. February 7-9, 2002.Abstract #500030.

113. McLay RN, Kimura M, Banks WA, et al:Granulocyte-macrophage colony-stimulating factor crosses the blood-brain andblood-spinal cord barriers. Brain 120(11):2083-2091, 1997.

114. Alfrey CP, Rice L, Udden MM, et al: Neocytolysis:Physiological down-regulator of red-cell mass. Lancet 349(9062):1389-1390,1997.

115. Rice L, Alfrey CP, Driscoll T, et al: Neocytolysiscontributes to the anemia of renal disease. Am J Kidney Dis 33(1):59-62, 1999.