The essential role of glycoprotein hormone erythropoietin (Epo) and its receptor, EpoR, in erythroid development is well established: both the EpoR−/− and Epo−/− mouse embryos die on embryonic day 13 (E13) due to failure of definitive erythropoiesis in fetal liver (Wu et al. 1995). It has been suggested that Epo’s principal role during erythropoiesis is to protect erythroid progenitors from apoptosis (Koury and Bondurant, Science 1990). Bcl-xL, an anti-apoptotic member of the bcl-2 family, is induced by EpoR signaling in erythroid cells via the Jak2/Stat5 pathway (Silva et al., Blood 1996; Socolovsky et al., Cell, 1999). Bcl-xL is essential for erythroid maturation: bcl-xL−/− embryos die in utero at the same stage as as EpoR−/− mice, lacking definitive erythropoiesis (Motoyama et al., Science 1995; J Exp Med, 1999). Recenlty, it has been shown that over-expression of bcl-xL in primary wild-type erythroblasts confers Epo independence on these cells in vitro and allows them to complete their differentiaion into red blood cells (Dolznig et al., Curr Biol, 2002). Here we reasoned that if the principal function of EpoR signaling is suppression of apoptosis via bcl-xL, it should be possible to rescue all aspects of erythroid differentiation in EpoR−/− fetal liver progenitors by retrovirally-transducing these cells with bcl-xL.

We infected EpoR−/− fetal liver progenitors with bicistronic retroviral vectors expressing either bcl-xL or EpoR, each linked via an IRES sequence to a GFP reporter. Control EpoR−/− cells were infected with ‘empty’ bicistronic vector. Infection rates were in excess of 30% for all constructs, and transduced cells were identified for further analysis using GFP fluorescence. We examined terminal differentiation of the transduced EpoR−/− cells over the ensuing 48 hours, using several distinct assays, including their expression of the cell-surface differentiation markers CD71 and Ter119 by FACS, their ability to give rise to CFU-e colonies in semi-solid medium, their cell-cycle status using DNA content analysis and BrdU incorporation, and their maturation and hemoglobinization by diaminobenzidine staining and light microscopy.

We found that EpoR−/− progenitors transduced with bcl-xL were protected from apoptosis, and underwent morphological changes characteristic of erythroid maturation, including decreasing cell size, nuclear condensation and expulsion, and accumulation of hemoglobin. These cells also upregulated the erythroid-specific cell surface marker Ter119. However, unlike EpoR−/− cells transduced with EpoR, bcl-xL -transduced cells did not express high levels of CD71, and failed to give rise to CFU-e colonies in semi-solid medium. Instead, they gave rise to small colonies of 6 cells or less. Cell cycle analysis showed that, throughout the 48 hours of erythroid terminal differentiation, the population of bcl-xL-transduced EpoR−/− cells had a lower fraction of cells in S-phase than control, EpoR-transduced EpoR−/− cells. The cell-cycle status of control, terminally-differentiating wild-type erythroid fetal liver progenitors was not altered by transduction with bcl-xL, excluding the possibility that it directly inhibits S-phase.

Taken together our results indicate that bcl-xL does not rescue all aspects of erythroid differentiation in EpoR−/− erythroid progenitors. Specifically, the proliferative program during erythroid terminal differentiation is directly dependent on EpoR signaling, and is not simply a default pathway secondary to EpoR’s anti-apoptotic effect.

Disclosure: No relevant conflicts of interest to declare.

Author notes


Corresponding author