G-PROTEIN ALPHA-SUBUNIT GI-ALPHA-2 MEDIATES ERYTHROPOIETIN SIGNAL-TRANSDUCTION IN HUMAN ERYTHROID PRECURSORS

Erythropoietin induces a dose-dependent increase in cytosolic calcium in human erythroblasts that is mediated by a voltage-independent Ca2channel. Inhibition of this response to erythropoietin by pertussis to xin suggests involvement of guanine nucleotide-binding regulatory prot eins (G-proteins). The role of G-proteins in regulation of the erythro poietin-modulated Ca2+ channel was delineated here by microinjection o f G-protein modulators or subunits into human erythroid precursors. Th is is the first report on the use of microinjection to study erythropo ietin signal transduction in normal precursor cells. Fura-2 loaded day -10 burst-forming units-erythroid-derived erythroblasts were used for microinjection and free intracellular calcium concentration ([Ca-i]) w as measured with digital video imaging, BCECF (1,2',7'-bis(2-carboxyet hyl)-5-(and -6-)-carboxyfluorescein) was included in microinjectate, a nd an increase in BCECF fluorescence was evidence of successful microi njection. Cells were microinjected with nonhydrolyzable analogues of G TP, GTP gamma S or GDP beta S, which maintain the a subunit in an acti vated or inactivated state, respectively. [Ca-i] increased significant ly in a dose-dependent manner after microinjection of GTP gamma S. How ever, injection of GDP beta S blocked the erythropoietin-induced calci um increase, providing direct evidence that activation of a G-protein is required. To delineate which G-protein subunits are involved, alpha or beta gamma transducin subunits were purified and microinjected as a sink for beta gamma or alpha subunits in the erythroblast, respectiv ely. Transducin beta gamma, but not alpha, subunits eliminated the cal cium response to erythropoietin, demonstrating the primary role of the alpha subunit. Microinjected antibodies to Gi alpha 2, but not Gi alp ha 1 or Gi alpha 3, blocked the erythropoietin-stimulated [Ca-i] rise, identifying Gi alpha 2 as the subunit involved. This was confirmed by the ability of microinjected recombinant myristoylated Gi alpha 2, bu t not Gi alpha 1 or Gi alpha 3 subunits, to reconstitute the response of pertussis toxin-treated erythroblasts to erythropoietin. These data directly demonstrate a physiologic function of G-proteins in hematopo ietic cells and show that Gi alpha 2 is required in erythropoietin mod ulation of [Ca-i] via influx through calcium channels.