Abstract
Analysis of Cell Surface Erythropoietin Receptor (EpoR) Expression and Function in Epithelial Human Tumor Tissues Reveals No Detectable Cell Surface Expression or Function
A number of studies have reported that EpoR mRNA is detectable at low levels in human epithelial tumor tissues and have raised the possibility that this surface EpoR protein expression could render human tumor cells responsive to Epo with inferred implications for the use of erythropoietic stimulating agents in the Oncology setting. To date the only data supporting this hypothesis has been generated using mRNA analysis or protein analysis using antibodies that have subsequently been shown to not be specific for EpoR.
To address these possible issues more directly we have analyzed EpoR expression and function in a large panel of human primary tumor tissues isolated by surgical resection from a variety of epithelial tumor cells (including colon, non small cell lung, breast and ovarian tumors). Solid human tumor tissue was disaggregated using a treatment shown to preserve cell surface EpoR density and function in positive controls (below). Flow cytometric analysis of primary tumor cells in the disaggregated tumor population was performed using multiple markers specific for epithelial tumor cells, as well as viability dyes and apoptotic markers to exclude non-viable and or apoptotic cells from the analysis. Using anti-EpoR monoclonal antibodies with high sensitivity and specificity for EpoR, no expression of cell surface EpoR was detected in primary tumor cells in any of the more than 60 human epithelial tumor specimens analyzed. In contrast, high levels of expression were observed in a positive control cell line (UT7/Epo) analysed in parallel, as well as a physiologically relevant primary tissue (differentiated erythroid progenitor cells). Notably, in a fraction of the ovarian and breast tumor tissues cell surface EpoR expression was identified in CD45+ tumor infiltrating leukocytes. No EpoR expression was detected in non tumor cells that were not CD45+ suggesting no contribution from other stromal elements in the tumor. These observations may explain the reported detection of EpoR mRNA in a subset of breast and other solid tumor patients, as those previous analyses involved bulk tumor tissue and did not allow analysis of tumor cell specific expression.
To test the possibility that low cell-surface EpoR protein density on tumor cells may be sufficient for function but undetectable by flow cytometry, we evaluated the ability of primary human tumor tissue to support an EpoR-driven signal pathway. EpoR function was analyzed in primary human tumor cells treated with a range of concentrations of recombinant human Epo (rHuEpO; 0.1–300U/mL) under conditions shown to result in productive EpoR dependent signaling in positive control cells/tissues. Analysis of possible EpoR-driven signaling was determined by intracellular flow cytometry using antibodies specific for the phosphorylated forms of STAT5, Akt, Erk1/2, p70S6, STAT3, STAT1, STAT6, JNK, and c-jun. Attribution of any detected signaling to viable tumor cells was performed via the use of a combination of tumor cell specific and viability/apoptosis markers. No evidence of downstream signaling was observed in primary tumor cells in epithelial tumors from over 60 patients at any concentration of rHuEpo, whereas UT7/Epo cells, treated in parallel, showed robust Epo concentration-dependent activation of signaling. Furthermore, activation of these signaling proteins was detected when the same primary human tumor cells were treated in parallel with a cocktail of known human tumor growth factors, confirming that these cells are capable of responding to exogenous stimuli using the same pathways as EpoR and that these signaling events can be readily detected using the platform. Taken together these data support the hypothesis that tumor cells in solid human tumors do not express functional cell surface EpoR and are not responsive to physiological or therapeutically relevant concentrations of Epo or indeed very high levels of Epo (300U/mL).
Disclosures: McCaffery:Amgen Inc.: Employment. Rossi:Amgen Inc.: Employment. Paweletz:Amgen Inc.: Employment. Tudor:Amgen Inc.: Employment. Elliott:Amgen Inc.: Employment. Busse:Amgen Inc.: Employment. Fitzpatrick:Amgen Inc.: Employment. Patterson:Amgen Inc.: Employment.
Author notes
Corresponding author