Erythropoietin: An Historical Overview of Physiology, Molecular Biology, and Gene Regulation

1

Introduction

Although generally ascribed to the 19th-century physicians, Richard Bright, Robert Christison, and Pierre Rayer, the link between kidney disease and anemia was first described 18 centuries earlier by the Greek physician Aretaeus, the Cappadocian, who noted: “In all the species there are present paleness, difficulty of breathing, occasional cough; they are torpid, with much languor.” In recent years, studies into the regulation of red blood cell production by the renal hormone erythropoietin have not only confirmed this link but have also provided effective therapeutic strategies for renal anemia and fundamental molecular insights into mechanisms of oxygen sensing and signaling that underlie oxygen homeostasis throughout the animal kingdom (see Table 11.1 for summary).

However, following Aretaeus’s description, many years were to pass before key discoveries in the 17th and 18th centuries defined the importance of oxygen and oxygen transport systems to living organisms and provided the fundamental platform necessary for modern understanding. Indeed, the description of the blood circulation by William Harvey (1578–1657) in De Motu Cordis et Sanguinis in Animalibus in 1628 left open the question of its purpose. Richard Lower (1631–91), working in Oxford with Robert Hooke (1635–1702), noted that whereas the blood leaving the heart for the lungs was blue, that returning from the lungs to the heart was red. Lower mixed blood with air in a glass vessel and noted the same color change, concluding that: “Nitrous spirit of the air, vital to life is mixed with the blood during transit through the lungs.” Furthermore, by means of the vacuum pump specially contrived by himself and Robert Hooke, Robert Boyle (1627–91) was able to obtain “air” from blood in 1670.

For centuries, it had been recognized that there was some active part in the air. The Chinese had called it “yin.” The Italian polymath, Leonardo da Vinci (1452–1519), had stated that the air was not completely consumed in respiration or combustion and had claimed that there were two gases in the air. Robert Boyle had shown that a component of air was depleted by living animals. However, the nature of this “spirit of the air” was to remain elusive for another 100 years, in part delayed by the erroneous phlogiston theory of combustion. By heating mercuric oxide to release a gas that supported combustion and respiration, Priestley and Scheele identified the essential “dephlogistated air” or “fire air,” although by publishing first, in 1774, Priestley is commonly credited with the discovery. However, it was Lavoisier who overturned the phlogiston theory and, in 1777, coined the term oxygen, correctly describing the chemistry of combustion and concluding that biological energy metabolism was essentially the same process. Perhaps because of Lavoisier’s untimely end, under the guillotine during the French Revolution, the term did not come into general use until it was popularized in the book “The Botanic Garden” by Erasmus Darwin, grandfather of Charles Darwin. “The enamour’d oxygene. The common air of the atmosphere appears by the analysis of Dr. Priestley and other philosophers to consist of … about one-fourth of pure vital air fit for the support of animal life and of combustion called oxygene.”

The first consistent measurements of oxygen in blood were performed by Gustav Magnus in 1837. In showing that there was more oxygen in arterial than venous blood, he confirmed the role of the blood circulation in delivering oxygen to the tissues. He also showed that blood contained more oxygen than could be accounted for by simple solubility and that the uptake of oxygen by the blood could be blocked by carbon monoxide suggesting a specific carrier mechanism. It was Hoppe-Seyler, in 1862, and Stokes, in 1864, who demonstrated the reversible binding of oxygen to the pigmented hemoglobin in the red blood cells that accounted for the color change and facilitated the transport of oxygen by the blood. By independently demonstrating the production of red blood cells in the bone marrow, in a process initially termed hematopoiesis, or latterly and more specifically erythropoiesis, Ernst Neumann and Giulio Bizzozero, in the 1870s, set the scene for subsequent studies into the regulation of this process.

2

Hormonal Regulation of Erythropoiesis

It was only a few years after that the concept of hormones was first suggested by Henri Brown-Sequard in 1889, and that the idea of hormonal regulation of erythropoiesis was first formulated by Carnot and Deflandre in 1906. Their experiments involved injecting serum from rabbits, rendered anemic by venesection, into normal rabbits leading to an increased concentration of red blood cells in the recipients within 1–2 days. They concluded that the transferred serum contained a hematopoietic factor that they termed “hematopoïetine.” Subsequently, hematopoïetine would be substituted by the more specific term “erythropoietin.” However, initially, the existence of hematopoïetine was doubted for many years because, with a few exceptions, most investigators failed to reproduce the results of Carnot and Deflandre. Interest in the possibility of a humoral factor controlling erythropoiesis was rekindled following observations in parabiotic animal pairs in 1950. Kurt Reissmann and Gerhard Ruhenstroth-Bauer observed that induction of anemia or hypoxemia in one of the parabiotic animals would induce erythrocytic hyperplasia and reticulocytosis in the partner. Allan Erslev (1919–2003) is generally credited with providing definitive proof of the existence of erythropoietin in 1953, by transfusing large quantities of plasma from anemic rabbits. Plasma from anemic animals (but not control animals) generated a significant reticulocytosis in the recipient animal which, after repeated dosing, resulted in a rise in the hematocrit. With remarkable foresight, he also postulated: “Conceivably isolation and purification of this factor would provide an agent useful in the treatment of conditions associated with erythropoietic depression, such as chronic infection and chronic renal disease.”

3

Identification of the Site of Erythropoietin Production

The clinical observation that patients suffering from hypoxemia to the lower portion of the body due to a patent ductus arteriosus showed generalized erythroid hyperplasia suggested a link between the lower part of the body and the stimulation of erythropoiesis. This was consistent with the observation that patients suffering from significant renal impairment were frequently anemic. The important role of the kidney in erythropoietin production became apparent when Leon Jacobson (1911–92) and Eugene Goldwasser showed that nephrectomized rats failed to respond to venesection or cobalt chloride with the normal increase in erythropoietin activity, whereas the response was intact in rats subjected to hypophysectomy, thyroidectomy, splenectomy, adrenalectomy, and gonadectomy. Nevertheless, the failure of attempts to extract erythropoietin from the kidney led to doubt that the kidney was the direct source of erythropoietin. Instead, an alternative hypothesis was advanced in which the kidney secreted an enzyme (erythrogenin) that cleaved erythropoietin from a plasma protein. However, erythropoietin could not be reliably generated by the addition of kidney extract to normal plasma. The erythrogenin concept was finally disproved by Erslev, as late as 1974, by the demonstration of erythropoietin activity in isolated serum-free perfused kidneys from hypoxic rabbits. Further confirmation that the kidney produces erythropoietin directly came with the isolation of erythropoietin mRNA from the kidneys of hypoxic rodents. While mRNA studies confirmed the results of organ ablation studies in showing that the main sites of erythropoietin synthesis were the kidney in the adult and the liver in fetal and neonatal life, the spleen, lung, bone marrow, brain and testes were all shown to express small amounts of erythropoietin mRNA. The translational efficiency and potential function of erythropoietin produced in these sites are not known. For instance, it is unlikely that erythropoietin produced in the brain enters the systemic circulation because of the blood–brain barrier. The expression of erythropoietin receptor (EpoR) in the brain and the ability of erythropoietin to protect the brain from ischemic insult have led to the assumption that erythropoietin may act as a paracrine neuroprotective factor, although the physiological function of such an action is unclear.

Within the kidneys of hypoxic rats, erythropoietin activity was mainly found in the cortex and not in the medulla. Again, this result was borne out by later mRNA studies, which demonstrated erythropoietin expression in the interstitium of the renal cortex and showed colocalization with fibroblast markers implicating this cell lineage in renal erythropoietin production.

4

Assays of Erythropoietin

Early erythropoietin research was hampered by the low concentration of the hormone in the fluids and tissues to be studied, particularly in the basal state, which made its detection and quantitation unreliable. The first assays of erythropoietin activity utilized the rate of incorporation of radioactive iron-59 into hemoglobin as a measure of erythropoiesis in starved rats that had been injected with the material under test. These assays were rendered more sensitive by using “ex-hypoxic” polycythemic mice to reduce the rate of background erythropoiesis. Initial standardization employed the “cobalt unit” in which one unit produced the same erythrogenic response in the test animals as 5 μmol cobalt chloride. Later reference standards included preparations of sheep plasma, human urinary erythropoietin, and in 1992, a fully glycosylated purified recombinant human erythropoietin.

In vivo bioassays were costly, time-consuming, and lacked precision and sensitivity. Several more sensitive bioassay methods using cell culture were described. These methods were generally applicable to purified erythropoietin samples but were often affected by nonspecific inhibitors present in crude samples. Such methods were eventually replaced by radioimmunoassay in the late 1970s and early 1980s. Today, there are many commercially available enzyme-linked immunoassay kits available for the determination of erythropoietin levels.

5

Isolation and Characterization of Erythropoietin

Armed with the early bioassays of erythropoietic activity, researchers next turned their attention to the biochemical purification of erythropoietin. Early attempts at partial purification of erythropoietin from anemic rabbit serum proved remarkably informative. The erythropoietic activity was found to have an electrophoretic mobility similar to alpha-2 globulin, to be heat stable and to stain for carbohydrate. Erythropoietin was therefore deduced to be a glycoprotein. These studies also showed that erythropoietin contained hexose, hexosamine, and sialic acid and that erythropoietic activity was lost upon removal of neuraminic acid.

In a mammoth effort , first ovine erythropoietin was purified over a millionfold from anemic-sheep plasma, and then human erythropoietin was purified from 2550 L of urine from patients with aplastic anemia. The purified human erythropoietin was subjected to tryptic digestion and the resulting fragments were separated and sequenced. The partial amino acid sequences obtained enabled DNA probes to be made, which were then used to probe both genomic and cDNA libraries to identify and subsequently clone the erythropoietin gene. Finally, expression of the erythropoietin cDNA in Chinese hamster ovary cells resulted in the production of biologically active erythropoietin.

The human erythropoietin gene is a single copy gene containing five exons and is located on the long arm of chromosome 7 (7q11–q22). It encodes a 193-amino-acid prohormone, from which a 27-residue leader sequence as well as the carboxy-terminal arginine are cleaved prior to secretion. The resulting 165 amino acid, mature human erythropoietin, is an acidic glycoprotein with a molecular mass of 30.4 kDa that contains two bisulfide bridges. Circulating erythropoietin has several glycosylation isoforms with 40% of the molecule consisting of carbohydrate comprising three tetra-antennary N-linked (Asn24, Asn38, and Asn83) and one small O-linked (Ser126) glycans. The N-linked glycans are essential for the biological activity of erythropoietin and contain terminal sialic acid residues that protect the whole molecule from removal by galactose receptors expressed on hepatocytes. The addition of further N-linked glycans to recombinant erythropoietin by site-directed mutagenesis has been used to prolong in vivo activity of the molecule.

The cloning of the human erythropoietin gene and the production of recombinant human erythropoietin led very quickly to its clinical application in the treatment of the anemia of chronic kidney disease, thereby fulfilling Erslev’s earlier prediction.

6

Erythropoietin Effector Mechanisms

It is now nearly 50 years since self-replicating hematopoietic stem cells were first demonstrated in the bone marrow. Derived from these pluripotent stem cells, the erythroid lineage comprises “the burst-forming unit erythroid” (BFU-E) and the more differentiated “colony-forming unit erythroid” (CFU-E). Each BFU-E can generate 50–200 erythroblasts when exposed to high concentrations of erythropoietin. The CFU-Es are more sensitive to the effects of erythropoietin with the number of colonies of erythroblasts derived from each increasing in response to more modest erythropoietin concentrations. In 1992, Koury and Bondurant demonstrated that erythropoietin promotes red blood cell formation by preventing apoptosis in these cell lineages.

Although evidence for a cell-surface EpoR was first provided in 1974, it was not until 15 years later that the murine EpoR was first cloned and characterized as belonging to the cytokine class I receptor family. The receptor exists as a homodimer and crystallization studies reveal a conformational change on binding of erythropoietin that leads to activation of Janus kinase 2 (JAK2), which interacts with the cytoplasmic region of the receptor. The affinity of erythropoietin analogs for the receptor decreases with increasing glycosylation (see Chapter 4 ).