Erythropoietin production: evidence for multiple oxygen sensing pathways.

The production of erythropoietin (Epo), the glycoprotein hormone which controls red blood cell formation, is regulated by feedback mechanisms sensing tissue oxygenation. The mechanism of the putative oxygen sensor has yet to be elucidated. There is evidence that at least two pathways participate in hypoxia signal transduction. One appears to involve a specific haem protein, and a second implicates reactive oxygen species (ROS). Iron catalyses the generation of intracellular ROS and therefore alters the cellular redox state. We have investigated the effect of modulating intracellular iron content on Epo production in Hep 3B cells. Iron chelation stimulates Epo production at 20% O2 and enhances Epo production at 1% O2, but it has no additive effect on cobalt-induced Epo production. Excess molar iron inhibited Epo production in response to hypoxia, desferrioxamine (DFO) and cobalt chloride and inhibited the DFO-enhancing effect of hypoxia-induced Epo production. We found that sulphydryl oxidising agents exert a differential inhibitory effect on hypoxia-induced versus DFO-induced Epo production, providing further evidence that multiple pathways of oxygen sensing exist.

Hemoglobin Old Dominion/Burton-upon-Trent, beta 143 (H21) His-->Tyr, codon 143 CAC-->TAC--a variant with altered oxygen affinity that compromises measurement of glycated hemoglobin in diabetes mellitus: structure, function, and DNA sequence.

OBJECTIVE: To determine the nature and characteristics of a unique hemoglobin variant that causes a spurious increase in glycated hemoglobin (HbA1c). MATERIAL AND METHODS: Blood specimens from four unrelated persons with this hemoglobin variant were examined by conventional laboratory methods, including electrophoresis, high-performance ion-exchange chromatography, and isoelectric focusing; by amino acid sequence analysis, polymerase chain reaction-based DNA sequence analysis, and electrospray ionization mass spectrometry, to establish the molecular structure; and by studies of oxygen affinity under varied conditions, to define the functional characteristics of the hemoglobin variant. RESULTS: The unique hemoglobin variant observed in these four cases is due to the mutation CAC-->TAC, at beta-globin gene codon 143, corresponding to beta 143 (H21) His-->Tyr. This amino acid substitution affects an important 2,3-diphosphoglycerate binding site and slightly increases the oxygen affinity of the hemoglobin variant. CONCLUSION: A hitherto unrecognized hemoglobin variant, encountered in four unrelated persons of Irish or Scots-Irish ancestry, hemoglobin Old Dominion/Burton-upon-Trent, beta 143 (H21) His-->Tyr, has now been characterized at the molecular, structural, and functional levels. Although it is associated with a slight increase in oxygen affinity, it is without hematologic effect, and its only clinical significance is that it coelutes with HbA1c on ion-exchange chromatography and thereby causes a spurious increase in HbA1c and compromises the use of this analyte to monitor the treatment of diabetes mellitus.

Serum erythropoietin levels in paroxysmal nocturnal haemoglobinuria: implications for therapy.

In order to assess the rationale and possible indications for the use of recombinant erythropoietin in paroxysmal nocturnal haemoglobinuria (PNH), we have measured endogenous erythropoietin (Epo) levels in 18 patients with PNH and in 44 patients with iron deficiency anaemia (IDA). In both groups of patients we found a significant inverse correlation between Epo and haemoglobin (Hb). However, the mean Epo level was significantly higher in the PNH group (385 mU/ml) than in the IDA group (136 mU/ml). The range of Epo levels at any given Hb was greater in the PNH group than in the IDA group. There was a significant positive correlation between Epo and absolute reticulocyte count. Since Epo administration is unlikely to benefit patients with high levels of endogenous Epo, we conclude that in the majority of patients with PNH there is no indication for treatment with Epo.

Expression and mutagenesis of recombinant human and murine erythropoietins in Escherichia coli.

Expression of the polypeptide hormone erythropoietin (EPO) in Escherichia coli by four bacterial expression vectors was examined. Complementary DNAs encoding human and murine EPO were amplified by polymerase chain reaction (PCR) and cloned into the glutathione-S-transferase (GST) fusion vector, pGEX-2T. Human EPO DNA was also cloned into the vectors, pET14b, pIN III-Omp A2 and pT7/7. Expression of human and murine EPO was obtained using constructs based on pGEX-2T. For constructs based on the other vectors, expression of EPO was absent or occurred at low levels, despite attempts to optimise conditions. Human and murine EPO, expressed as fusion proteins with GST, were partially soluble and displayed EPO bioactivity. Soluble GST-EPO fusion proteins were affinity purified on immobilised glutathione. Insoluble protein could also be purified by elution from gel slices following SDS-PAGE to yield either fusion protein or, after treatment with thrombin, unmodified EPO which was both soluble and bioactive. The pGEX expression system was evaluated as a means of analysing the structure-function relationships of EPO by in vitro mutagenesis. Three human and three murine EPO mutants were constructed and expressed as GST fusion proteins. Following purification, biological activity was evaluated using assays for bioactivity, immunoactivity and GST activity. The pGEX expression system complements eukaryotic systems described previously for expression of EPO and should provide much useful information about the structure-function relationships of the hormone.

Structure-function relationships of the erythropoietin molecule.

The tertiary structure of erythropoietin (EPO) remains to be elucidated by X-ray crystallography. Although the amino acid sequence of EPO is known, the specific features that confer its biological activity are not well understood. In order to study the structure-function relationships of EPO by in vitro mutagenesis, we have used the vector pGEX-2T to express human and murine EPO fused to the carboxyl terminus of glutathione S-transferase (GST) in E. coli. The fusion proteins were the predicted size (46 kDa) by SDS-PAGE. GST-huEPO eluted from glutathione-agarose using reduced glutathione (GSH) was tested by radioimmunoassay and in a mouse spleen cell assay (MSCA). Dose-response curves parallel to recombinant human EPO (rHuEPO) were obtained in both assays. The ratio of immuno- to bioactivity was 4.7:1. Thus the presence of the 26 kDa GST protein at the end terminus of EPO does not abrogate biological activity. GST-mEPO also gave dose-response curves parallel to rHuEPO in the MSCA but not in the RIA. The wild-type murine and three mutant GST-EPO fusion proteins (166 Des-Arg, Glu 159-->Val, and Arg 163-->Glu) were tested in the MSCA and assayed for GST activity. The ratio of bioactivity to enzyme activity for the Arg 163-->Glu mutant was approximately one third of the value obtained for each of the other fusion proteins, indicating that arginine at 163 is functionally important for EPO activity. The availability of these human and murine gene constructs in pGEX should facilitate site-directed mutagenesis and permit detailed studies of the structure-function relationships for the two erythropoietins.

Effect of recombinant human erythropoietin on anemia caused by a murine mammary carcinoma.

CA NT is a transplantable murine mammary carcinoma that causes progressive anemia accompanied by granulocytosis and splenomegaly. Serum erythropoietin (Epo) levels, as measured by RIA, did not become elevated in anemic tumor-bearing mice; there was no correlation between hematocrit and serum Epo levels. Treatment with recombinant human (rHu) Epo prevented anemia in tumor-bearing mice when given in large doses, commencing on days 3-5 of tumor growth. Recombinant human Epo-treated mice had smaller spleens than controls. When treatment commenced on day 7, the development of anemia was retarded but not completely prevented. Treatment commenced on day 14 was less effective. This study demonstrates that treatment with rHu Epo can markedly influence the course of tumor-induced anemia.

The metabolism of erythropoietin in the normal and uraemic rabbit.

The metabolism of recombinant human erythropoietin (rHuEpo) labelled with 125I has been investigated in five normal and nine 5/6-nephrectomised rabbits. The plasma erythropoietin half-life was significantly prolonged at 5.1 +/- 1.2 h (mean +/- SD) in the 5/6-nephrectomised rabbits, compared to 3.0 +/- 0.4 h in sham-operated controls (P less than 0.001). The disappearance of 125I-labelled rHuEpo is biphasic. Examination of serum by fast protein liquid chromatography (FPLC) following administration of 125I-labelled rHuEpo by FPLC showed a single peak of radioactivity in all rabbits except two of the nephrectomised group. In serum from both of these animals a second labelled peak was found, corresponding to material of MW 200,000-250,000 D. We conclude that the 5/6-nephrectomised rabbit provides a stable model for the study of hormonal metabolism in chronic renal failure.

Serum erythropoietic activity in acute anemia--an animal model.

Serum erythropoietic activity and reticulocyte response to anemia were investigated using a rabbit model. In hemolytic anemia, induced by injections of phenylhydrazine on Day 0 the hemoglobin reached a nadir (mean, 6.23 g/dl) on Day 4 when SEA was maximal (mean, 765 mU/ml). In animals venesected on Day 0 and Day 1 to produce anemia of equal severity, the SEA was maximal (mean 235 mU/ml) on Day 2. In both groups the reticulocyte response peaked on Day 7--at 34% for the hemolytic group and 21% for the venesected group. The 2,3-diphosphoglycerate, measured on Day 4, was significantly reduced in the PHZ-treated group. In the venesected group the 2,3-DPG increased between Day 0 and Day 4. There were no concurrent changes in acid-base balance. These results imply that the degree of anemia is only one of the factors which influence the level of circulating SEA.

The initiation of erythropoiesis following renal transplantation.

Erythropoiesis has been examined in relation to kidney function in 38 patients during the 3-month period following successful renal transplantation, using serial determinations of erythropoietin, haemoglobin, and creatinine. Two peaks of serum erythropoietin were observed: an early peak that occurred within 2 days of transplantation and was observed in ten patients, and a late one between 8 and 30 days, observed in 28 patients. The early peak did not produce an increase in haemoglobin and occurred only in the presence of delayed onset of graft excretory function when serum creatinine was greater than 1000 mumols/l. The ineffectiveness of the early peak may be due to the uraemic environment, which is probably a sequel of the tubular damage associated with postoperative acute tubular necrosis. The late peak followed a decrease in serum creatinine to less than 200 mumols/l and was associated with an increase in haemoglobin of 3-4 g/dl during the next 2-6 weeks.

Comparison of the mouse spleen cell assay and a radioimmunoassay for the measurement of serum erythropoietin.

The mouse spleen cell assay (MSCA) has been compared with a radioimmunoassay for the measurement of serum erythropoietin (Ep). In 20 normal subjects the serum values ranged from 15 to 73 mU/ml for the MSCA compared with 5-30 mU/ml for the RIA. For normal sera there was no correlation between the results of the two assays. In 37 patients with anaemias of differing aetiologies and at various stages of treatment values ranged from 10 to 3645 mU/ml for the MSCA and 13-10,000 mU/ml for the RIA. Although patient values from the two assays were highly correlated (r = 0.98, P less than 0.001), the MSCA results were generally lower. These discrepancies can be largely accounted for by two factors. Firstly the MSCA is sensitive to non-specific matrix effects. Secondly, heat inactivation of serum, a prerequisite for the MSCA, but not for the RIA, destroys a variable and unpredictable proportion of the Ep in the test sera leading to an underestimation of Ep in the MSCA. We conclude that the RIA is more reliable than the MSCA which, in its present form, cannot be recommended for the accurate measurement of serum erythropoietin.

Erythropoietic response to hypobaric hypoxia in rabbits.

Animals subjected to hypoxia become hypocapnic and after some hours show an increase in circulating erythropoietin. The steps involved in the increased production of erythropoietin in response to hypoxia are not fully understood, although it has been postulated that changes in coincident variables such as acid-base balance may contribute to the mechanism of increased erythropoietin production. A rabbit model has been used to determine the physiological changes which occur in short-term hypobaric hypoxia. After 1 h, no changes were found in pCO2, pH, P50, base excess, standard bicarbonate or serum erythropoietic activity (SEA). After 3 h the pCO2, pH, base excess and standard bicarbonate had decreased while the P50 and SEA had increased. After 6 h, although the pCO2 was still significantly reduced, the pH, base excess and standard bicarbonate had returned to the initial levels and maximal SEA values. 20-fold greater than the pre-hypoxia values were found. Overall the data are consistent with the view that the magnitude of the erythropoietic response to hypoxia is modified by changes in acid-base balance.

Red cell pyrimidine 5' nucleotidase deficiency: determination of nucleotidase activity and nucleotide content using HPLC.

The hereditary deficiency of erythrocyte pyrimidine 5' nucleotidase has been investigated using an HPLC anion-exchange procedure designed to measure both red cell nucleotide content and enzymic activity. Red cell nucleotide profiles were determined using a low pH phosphate buffer salt-gradient system in 20 min, whereas a low pH buffer alone permitted the determination of enzymic activity with each of six different nucleotide substrates in less than 4 min. Both the red cell nucleotide profiles and the enzymic activity of haemolysates from two affected brothers and their children agreed well with previously published values. This unified approach should prove useful for detailed studies of deficiencies involving isoenzymes of pyrimidine nucleotidase.

In vitro bioassay of erythropoietic activity in serum using mouse spleen cells. The effect of heat inactivation on serum erythropoietin.

Untreated human serum is known to be toxic to in vitro assays for erythropoietin, including the mouse spleen cell assay system (MSCA). This phenomenon had previously been shown to be mediated by complement-dependent IgM heteroantibodies and can be overcome by heating the serum at 56 degrees C for 30 minutes. Using the MSCA, we have found that the toxic effect of serum could also be removed by treatment with a precipitating antibody against the C3c component of complement. The effects of the two methods of complement inactivation on the measurement of stimulatory activity in serum have been compared. For normal serum, the results after heat inactivation and antibody treatment were similar. In contrast, serum from a patient with aplastic anemia gave a result equivalent to 327 mU erythropoietin/ml after heat treatment, but after antibody treatment equivalent to 1,520 mU erythropoietin/ml. Gel permeation chromatography of unheated, heated, and antibody-treated sera showed that heating markedly reduced the activity of the erythropoietin peak. Seventy percent of the activity of partially purified urinary erythropoietin was lost during heating in the presence of normal serum. In addition, heating caused the appearance of high molecular weight compounds that are stimulatory in the MSCA. The level of this activity appeared to be directly related to the stimulatory activity of the unheated serum.

Increased plasma glycosidase and protease activity in uraemia: possible role in the aetiology of the anaemia of chronic renal failure.

We have measured plasma N-acetyl-beta-D-glucosaminidase (EC 3.2.1.30) and neuraminidase (EC 3.2.1.18) activities as markers of glycosidase activity and immunoreactive trypsin (EC 3.4.21.4) levels as a marker of proteolytic potential in the plasma of normal and uraemic subjects. The levels of all of these enzymes are significantly elevated in the plasma of uraemic subjects when compared to normal. We have postulated that the combined attack of glycosidases and proteases on erythropoietin will lead to fragmentation of this glycoprotein hormone with loss of activity. This may be a major contributory cause to the anaemia of chronic renal failure.

Separation of erythropoietic stimulators in severe anaemia using high performance gel permeation chromatography.

Patients with various types of anaemia, but with comparable haemoglobin levels, show a wide range of serum erythropoietic activity. We have developed a method for the fractionation of serum by HPLC followed by bioassay of the individual fractions, using the mouse spleen cell microassay. Up to three distinct peaks of erythropoietic activity corresponding to molecular weights (MW) greater than 300,000, 250,000-300,000 and 40,000 have been found in serum from both normal and anaemic subjects. The erythropoietic profiles of the sera examined differ markedly in anaemias of different aetiology. The chemical nature and the physiological significance of the stimulators remain to be investigated.