Signaling pathways activated in erythroid cells have been recently characterized at the proteomic level by Reverse Phase Protein Array (RPPA) (Hricik T. et al., Am. J. Hematol. 2013). These experiments were performed on erythroblasts (Ery) obtained in HEMA (Human Erythroid Massive Amplification) cultures of healthy donors (adult blood, AB), CB and JAK2V617F positive PV patients. They showed that AB Ery express the highest levels of CD63, a member of the Transmembrane 4 Superfamily Proteins that quenches cKIT tyrosine kinase function and of cKIT phosphorylation at Y721 that mediates the interaction between cKIT and CD63 (Anzie et al, Blood 2002; 99: 4413). Since experiments in mice demonstrated that fetal cells are more responsive to cKIT stimulation by Stem Cell Factor (SCF) than the adult ones, we hypothesized that the stringency of the interaction between CD63 and cKIT may determine different levels of signaling activation and of proliferation in response to SCF in Ery from AB, CB and PV. To test this hypothesis, response to SCF of Ery obtained from an extended number of AB, CB and PV (>5 subjects each) was characterized through biological (proliferation assays), biochemical (receptor downmodulation by FACS and confocal analyses) and signaling (RPPA) studies.
CB Ery achieved maximal proliferation with 30 ng/mL of SCF while maximal expansion of both AB and PV Ery required as much as 100 ng/mL of SCF. Anti-cKIT antibodies neutralized 100% of the proliferation response induced by 30 ng/mL of SCF in Ery from CB and AB but only 50% of that elicited by the same SCF concentration in Ery from PV.
By FACS, SCF exposure downmodulated readily cKIT expression on Ery from CB (100% downmodulation with 10 ng/mL of SCF at 2 h), to a limited extend that on Ery from AB (100% downmodulation with 100 ng/mL at 2 h) and poorly that on Ery from PV (50% downmodulation with 100 ng/mL at 2 h). Confocal microscopy analysis showed that CD63 and cKIT were expressed in different proportions in Ery from AB, CB and PV with Ery from AB expressing the greatest CD63 levels and those from PV the greatest cKIT levels. Furthermore, upon SCF stimulation, CD63 and cKIT became strongly co-localized in Ery from AB but CD63/cKIT co-localization remained poor in Ery from CB and barely detectable in those from PV.
RPPA determinations indicated that SCF stimulation induced different numbers of events in Ery from AB (8 hits), from CB (24 hits) and PV (42 hits). In AB Ery, SCF induced phosphorylation of cKITY721, of STAT3/STAT5 and of factors involved in transcription/translation (eIF4G, FKHR and S6 Ribosomal protein). In CB Ery, the numerous responses elicited by SCF stimulation included activation of mTOR/AKT pathway components (mTOR, AKT and p70 S6K). In PV Ery, SCF activated multiple signaling pathways involved in cell survival and expansion such as the mTOR pathway (mTOR, p70, S6 Ribosomal protein), the AKT pathway (FKHR, Glycogen Synthase Kinase 3 beta) and the Ras/MAPK pathway (c-Raf, ERK 1/2, Ras-GRF1, RSK3). We are currently validating by western blot and loss-of-function studies which of these pathways plays a major role in eliciting the abnormal amplification observed with Ery from PV.
These observations suggest that CD63 association with cKIT mediates the quenching of cKIT response occurring during the ontogenesis of human erythroid cells (CB-derived vs AB-derived Ery) and that this regulatory mechanism is disrupted as part of the pathogenetic process of PV. We propose that the signaling events identified by this proteomic profiling may represent targets to improve ex-vivo expansion of hematopoietic stem cells/progenitor cells or erythroblasts for cell therapy purposes or druggable targets for treatment of PV.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.