Background: Endothelial cells have been reported to express surface erythropoietin receptors (EpoR) and erythropoietin (Epo) has been reported to stimulate endothelial cell proliferation, migration and tumor angiogenesis. Published studies have reported human endothelial cells express high levels of surface EpoR (10,000 and 27,000R/cell) with an unusual and low binding affinity (2,400 to15,000 pM Kd). These data contrast with erythroid progenitors that have low levels of surface EpoR (135–1,000) with high binding affinity (100–200 pM Kd). Of concern, studies reporting endothelial cell responses to Epo were often conflicted or inconclusive. Therefore, we have investigated the validity of these studies through analysis of EpoR transcription, EpoR protein and surface expression in primary human endothelial cells, and examined the effect of recombinant murine Epo (rMuEpo; Amgen) on angiogenesis in vivo.

Methods: Cell surface EpoRs were measured by [125I]rHuEpo competitive binding assays using cultured intact whole cells derived from primary human endothelial tissue from: umbilical vein (HUVEC); lung microvasculature (HMVEC-LBl); adult dermal microvasculature (HMVEC-dBLAd) and coronary artery (HCAEC). EpoR positive (UT-7/Epo:promegakaryocytic leukemia) and EpoR negative (769-P: renal cell adenocarcinoma) cell lines were included as controls. EpoR transcript and protein levels were determined by quantitative RT-PCR and Western blot analysis using the anti-EpoR peptide polyclonal antibody, M20 (Santa Cruz Inc) respectively. Rat corneal angiogenesis studies were performed with rMuEpo at three different concentrations (500, 150 or 50 ng/ml). Recombinant human VEGF (420 ng/ml) was included as a positive control. Nylon disks were soaked with growth factors or vehicle then implanted into the corneal stroma of female CD rats (n=8). Seven days after disk implantation, the angiogenic response was evaluated from digital images of each rat cornea. All images were evaluated blinded.

Results: EpoR transcripts were detected in endothelial cultures but levels were over 50 fold lower than bone marrow and EpoR positive UT-7/Epo cells. EpoR protein was not detected in endothelial cells analyzed by Western analysis. In concordance with these data, specific [125I]rHuEpo binding to intact endothelial cells was not detected, demonstrating lack of surface EpoR protein. In angiogenesis studies, rMuEpo (at all concentrations) did not induce a significant angiogenic response above that of the vehicle, even though rMuEpo was bioactive in proliferation assays using EpoR expressing hematopoietic cell lines (EC50 = 1.7 ng/ml). As expected, positive control rHuVEGF provided a strong angiogenic response.

Conclusions: Endothelial cells express very low levels of EpoR mRNA, and EpoR protein was undetectable within the whole cell or on the cell surface. The lack of an angiogenic effect of rMuEpo strongly suggests that the hypothesis that erythropoiesis-stimulating agents stimulate endothelial cells and promote vascular effects or tumor angiogenesis should be reconsidered.

Author notes

Disclosure:Employment: All authors are employees of Amgen, Inc. Ownership Interests:; All authors hold stock in Amgen, Inc.