Improved assessment of bioactive erythropoietin in human sera using a modified tritiated-thymidine-incorporation assay.

The reproducibility and utility of an in vitro bioassay for erythropoietin were evaluated, and modifications, including the use of frozen target cells, were made. The assay, based on 3H-thymidine incorporation into phenylhydrazine-treated mouse spleen cells, showed a dose-response curve from 0.2 to 20 mU of sheep plasma or human urinary erythropoietin, with a 50-100-fold increase in 3H-thymidine incorporation. The dose-response curve, using purified, recombinant erythropoietin was identical to that from material obtained from other sources. The use of frozen cells simplified the methodology, reduced the time for assay preparation, and lowered costs. The freezing process resulted in 15% attrition of the cells, but after one year of storage, this level was not increased nor was the response to erythropoietin diminished. The nonspecificity of the assay was reduced by supplementing the cells with critical nutrients, depleting fetal bovine serum of endogenous erythropoietin, and reducing solute variability by ultrafiltration and reconstitution of samples with culture medium. Heat inactivation diminished inhibitors in sera samples, but affinity chromatography using wheat-germ lectin-Sepharose was most effective for extracting erythropoietin and reducing nonspecific effects. The assay is sensitive, can be used to measure multiple samples rapidly, and should prove useful for complementing immunoassay techniques.