Hypoxia causes an increased production of erythropoietin (EPO), but the time course of the EPO response in humans has not been well characterized. This study examines the relationship between the duration of normobaric hypoxic exposure and plasma EPO levels in healthy human subjects. Six volunteers breathed a gas mixture of 10.5% O2-89.5% N2 continuously for 5, 60, and 120 or intermittently for 240 min. O2 saturations were maintained between 75 and 85% during the exposure. Arterial pH was 7.467 +/- 0.019, PO2 37.05 +/- 2.43 Torr, and PCO2 36.69 +/- 2.05 Torr. O2 half-saturation pressures of hemoglobin were normal for all subjects. Plasma EPO was measured every 30 min for 360 min by radioimmunoassay. No increase in EPO was seen after the 5- and 60-min exposures. However, a 50% increase was seen 240 min after the initiation of the 120-min hypoxic exposure (P less than 0.01). Intermittent exposure resulted in an increase of EPO by 52% 360 min after the onset of exposure (P less than 0.05). Therefore, exposing humans continuously to an inspiratory O2 fraction of 0.105 for 120 min or intermittently for 240 min provides a sufficient stimulus to increase production of EPO.

Establishment of a stable, transformed human renal carcinoma cell line that produces erythropoietin in vitro and has maintained this function continuously since 1981 and for >150 passages in monolayer culture was accomplished by transplantation of human renal clear cell carcinoma tissue from a patient with erythrocytosis into an immunosuppressed athymic mouse. In addition to its immunocrossreactivity with native human urinary erythropoietin, the tumor erythropoietin demonstrates biological activity in the in vitro mouse erythroid colony-forming unit assay and in tumor-bearing nude mice. The cloned renal carcinoma cell line has an abnormal human karyotype and has ultrastructural features characteristic of human renal clear cell carcinoma. This cell line provides a reproducible model system for the production of an erythropoietin-like material and for the study of its synthesis and secretion.Several investigators have attempted to develop in vitro model systems that produce erythropoietin on a continuous and reproducible basis. These studies utilized either normal kidney (1) or renal carcinoma cells (2, 3), testicular germ cells (4), or mouse erythroleukemia cells (5). However, establishment of a stable transformed human cell line capable of continuous production of erythropoietin for several years has not yet been achieved.We report the establishment of a human renal cell carcinoma cell line in culture that was generated by using an immunosuppressed nude mouse as a temporary host. These cells have been shown to produce erythropoietin on a continuous basis and have been maintained for >4 years and >150 passages in culture.

Significant amounts of active erythropoietin were extracted from the kidneys of normal rats, cattle, dogs, and rabbits by homogenization of the organs in 0.1 M phosphate buffer. The mean erythropoietin activities of the extracts, as determined by the starved-rat assay, were 0.26 U/g beef kidney, 0.41 U/g dog kidney, and 0.11 U/g rat kidney. The dog kidney extracts had a mean activity of 0.35 U/g, as measured by stimulation of hemoglobin synthesis in cultured bone marrow cells (in vitro assay) and produced a dose-dependent stimulation of 59Fe incorporation into circulating red cells when assayed in polycythemic mice. Extracts of rabbit kidney cortices had a mean activity of 2.12 U/g, as measured by stimulation of hemoglobin synthesis in cultured bone marrow cells. When the dog kidney homogenate was fractionated on DEAE-cellulose, all of the erythropoietin activity was adsorbed to the exchanger in the presence of 0.01 M acetate buffer, pH 4.5, and was completely eluted by 0.1 M Na2HPO4-0.5 M NaCl, pH 8. An antibody made against human urinary erythropoietin completely inactivated the erythropoietic factor in the dog kidney extract. Serum from a donor dog had no erythropoietin activity when assayed in the starved rat, suggesting that the factor in the extracts is intracellular erythropoietin rather than that contained in plasma trapped in the renal vasculature. The complete inactivation of the erythropoietic factor in these kidney homogenates by antierythropoietin and its behavior on DEAE-cellulose indicate that this factor is structurally similar to native plasma erythropoietin. The extracts are completely active without being incubated in the presence of serum.

We describe the development of a radioimmunoassay (RIA) for erythropoietin. Antisera were raised in rabbits with an impure human urinary erythropoietin preparation used as immunogen, but with pure human erytropoietin serving as the labeled antigen in the RIA and as a primary standard. The immunoreactivity of erythropoietin is not altered significantly by the mode of labeling with radioiodine, even though the biologic activity is lost. With this method, it is possible to detect 2- -3 mU of erythropoietin in a volume of 0.1--0.3 ml. Therefore, the method can be used for detection of normal and subnormal serum titers as well as elevated titers. RIA for erythropoietin does not distinguish between the native (active in vivo) and the asialo form (inactive in vivo) and cannot yet be used for routine monitoring of crude fractions during purification.

We measured nocturnal plasma erythropoietin concentration (EPO) in eight subjects with obstructive sleep apnea syndrome (OSAS) and eight healthy overweight subjects while they were undergoing nocturnal polysomnography. We also measured EPO (radioimmunoassay) after 120 min of exposure to 10.5% O2. The subjects with OSAS had a respiratory disturbance index (RDI) of 50.8 +/- 41.9 and a maximal O2 desaturation of 65 +/- 13.3%, and they spent 104.5 +/- 89.3 min out of a total sleep time of 356 +/- 54 min below 90% oxygen saturation (T90). Nocturnal EPO concentrations were normal and did not differ between the two groups. We found no correlation between the T90, T80, or RDI and either the mean or maximal EPO concentrations. After the exposure to 10.5% oxygen, during which oxygen saturations were between 80 and 85%, the healthy subjects and those with OSAS developed increases in EPO of 34 and 49%, respectively, 240 min after the initiation of exposure (p < 0.05). We conclude that the nocturnal hypoxemia occurring in these subjects with OSAS may not have been sufficiently severe to stimulate an increased EPO production. The lack of EPO response was not due to down-regulation of EPO production.