Erythroid colony formation in cultures of mouse and human bone marrow: Analysis of the requirement for erythropoietin by gel filtration and affinity chromatography on agarose‐concanavalin A

Abstract Assay of red cell progenitors by colony formation in culture is expected to allow study of early events in erythroid differentiation in animal models and in man. In this communication, a simplification of the culture method for mouse bone marrow cells originally reported by Stephenson et al. ('71) is described. In the modified procedure, plasma clot is replaced by methyl cellulose, and scoring of erythroid colonies is done directly in the plates without staining. With additional modification the system proved applicable to the culture of erythroid colonies from human bone marrow as well. The development of both mouse and human erythroid colonies was dependent on “erythroid colony stimulating activity” (E‐CSA) supplied by extracts of anemic sheep plasma, or by extracts of urine from anemic patients. Over a certain range, the number of colonies was a function of dose of E‐CSA. In addition to E‐CSA, these extracts also possessed erythropoietin activity as measured in plethoric mice. The possible chemical equivalence of urinary E‐CSA and erythropoietin was suggested by their similar behavior on both gel filtration and affinity chromatography on agarose‐concanavalin A. The finding further suggests that the culture method might prove useful for the bioassay of erythropoietin. Granulocyte colony stimulating activity (G‐CSA) was also present in the urine extracts, as detected in cultures of mouse bone marrow. Virtually all of this activity was bound on agarose‐con A, whereas only a small fraction of E‐CSA was retained on this material. Agarose‐con A may thus be useful for the purification of erythropoietin.