Absence of Functional Epo Receptors In Normal and Cancerous Human Tissues and Cell Lines,

Abstract 3169

Erythropoietin (Epo) is a late acting growth factor, named because early studies suggested it had a singular effect on stimulation of red blood cell formation (erythropoiesis). Epo functions by activation of Epo receptors (EpoR) and while early studies suggested EpoR transcripts were expressed exclusively in the erythroid compartment, transcripts were later detected in other tissues and cell types including tumor cells with more sensitive RT-PCR strategies. This suggested that activation of EpoR by ESAs on tumor cells may explain results from some clinical studies where cancer patients in the ESAs treatment arm had decreased locoregional control and/or decreased survival. While EpoR transcripts could be detected, it was not clear that they were translated into EpoR protein or that the EpoR protein was functional. Attempts to examine EpoR protein expression in tumors were inhibited by nonspecificity and low sensitivity of available anti-EpoR antibodies.

Recently specific and sensitive monoclonal antibodies were identified to human EpoR enabling experiments to explore EpoR protein expression in cells and tumors by western: CD34⁺ cells differentiated into erythroid progenitors and Epo-dependent UT-7/Epo cells served as positive controls and 769-P cells as a negative control. The latter expressed minimal EpoR mRNA and undetectable EpoR protein according to [¹²⁵I]rHuEpo binding and western analysis. The limit of detection of EpoR by the most sensitive antibody, A82 in westerns was estimated to be less than 100 total (intracellular plus extracellular) EpoR/cell. By comparison, positive control cell types expressed over 10,000 EpoR/cell. In positive controls a 59 kDa protein, shown to be full-length EpoR, was readily detected with A82 with smaller EpoR proteins also present, but no proteins were detected in 769-P cells. The smaller proteins detected by A82 in the positive controls were shown by Mass-Spec analysis to be EpoR fragments. In contrast to the positive controls, EpoR proteins were not detected in either normal human breast or cancerous breast tissue (both Her2⁺ and Her2⁻) and there was little/no EpoR detected in a breast cancer lines (N=17) including MCF-7, though a band similar to EpoR in size was detected with the Santa Cruz anti-EpoR polyclonal, anti-peptide antibody, M-20. We therefore attempted to characterize this band. The band was not detected in MCF-7 cells with another specific anti-EpoR antibody (ab10653) nor was it observed in any immunoprecipitation:immunoblotting (IP:IB) experiments with combinations of 5 different anti-EpoR antibodies, though EpoR was readily detected in positive control. In contrast, the band was detected in IP:IB experiments with M-20 suggesting the protein is likely a non-EpoR protein that cross-reacted solely with that polyclonal antibody. Similar to breast tissues, EpoR was not detected with A82 in normal human primary keritinocytes, melanocytes and normal skin tissue. It was also not detected in squamous skin carcinoma and melanoma tissue or in cell lines derived from skin tumors (N=5). Matched normal and cancerous tissues from lung, colon, larynx, tongue and ovary were also negative for EpoR with A82. In contrast to tissues, low-level EpoR protein was detected in several of 21 cell lines derived from lung tumors. The EpoR protein in the one with the highest level (H838) was estimated to be ∼1,000–2,000 EpoR/cell, 10-fold lower than in erythroid progenitor cells and 40-fold lower than in UT-7/Epo cells. We therefore examined EpoR signaling in these cells. No phosphorylation of AKT or STAT5 (pAKT, pSTAT5) was detected in H838 cells with rHuEpo treatment. In control experiments increased pAKT was observed in H838 with a growth factor cocktail (EGF/ HGF/ IGF-1) and in UT-7/Epo cells treated with rHuEpo. These results suggested the EpoR in H838 cells was not functional. A factor dependent (IL-3) cell line (32D) was also engineered to express similar levels of EpoR as H838: these cells grew with IL-3 (10 ng/ml) but not with medium lacking growth factors, nor in medium containing 10 U/ml rHuEpo. In contrast, 32D cells with 20-fold higher levels of EpoR grew immediately with rHuEpo and at a rate similar to that with IL-3 but did not grow in absence of stimulation. These results suggest that EpoR expression level is important for Epo dependent growth of 32D cells, and that insufficient levels of EpoR in H838 cells may be one explanation for non-response to rHuEpo.