Functional erythropoietin receptor is undetectable in endothelial, cardiac, neuronal, and renal cells.

Erythropoiesis stimulating agents (ESAs) have been reported to activate erythropoietin receptors (EpoR) on cell types, including endothelial, neuronal, renal tubule, and cardiac cells. ESAs have also been reported to promote angiogenesis. However, those findings are controversial and confounded by methodologic issues. We show that EpoR mRNA was detected in essentially all cell types examined, including primary human endothelial, renal, cardiac, and neuronal cells but 10- to 100-fold lower than Epo-responsive cells using quantitative reverse-transcribed polymerase chain reaction. Total endothelial EpoR protein examined using a new monoclonal antibody was low to undetectable. Surface EpoR on endothelial cells was not detected using [(125)I]-rHuEpo surface-binding studies. There was no evidence of ESA-induced intracellular signaling in endothelial cells. There was a similar lack of EpoR expression and signaling in other cell types examined. Experiments were performed examining ESA function on these cells. An in vivo rat corneal angiogenesis assay demonstrated neo-vessel formation in response to recombinant human vascular endothelial growth factor (rHuVEGF). However, recombinant mouse Epo did not induce vessel formation. Similarly, ESAs did not reproducibly provide cytoprotection to neuronal, renal, or cardiac cells. Taken together, our data challenge the notion of presence or function of EpoR on nonhematopoietic cells, and call into question the preclinical basis for clinical studies exploring direct, "pleiotropic" actions of ESAs.

Absence of functional EpoR expression in human tumor cell lines.

Certain oncology trials showed worse clinical outcomes in the erythropoiesis-stimulating agent (ESA) arm. A potential explanation was that ESA-activated erythropoietin (Epo) receptors (EpoRs) promoted tumor cell growth. Although there were supportive data from preclinical studies, those findings often used invalidated reagents and methodologies and were in conflict with other studies. Here, we further investigate the expression and function of EpoR in tumor cell lines. EpoR mRNA levels in 209 human cell lines representing 16 tumor types were low compared with ESA-responsive positive controls. EpoR protein production was evaluated in a subset of 66 cell lines using a novel anti-EpoR antibody. EpoR(+) control cells had an estimated 10 000 to 100 000 EpoR dimers/cell. In contrast, 54 of 61 lines had EpoR protein levels lower than 100 dimers/cell. Cell lines with the highest EpoR protein levels (400-3200 dimers/cell) were studied further, and, although one line, NCI-H661, bound detectable levels of [(125)I]-recombinant human Epo (rHuEpo), none showed evidence of ESA-induced EpoR activation. There was no increased phosphorylation of STAT5, AKT, ERK, or S6RP with rHuEpo. In addition, EpoR knockdown with siRNAs did not affect viability in 2 cell lines previously reported to express functional EpoR (A2780 and SK-OV-3). These results conflict with the hypothesis that EpoR is functionally expressed in tumors.

Lack of expression and function of erythropoietin receptors in the kidney.

BACKGROUND: Erythropoiesis-stimulating agents (ESAs) stimulate formation of red blood cells by binding to and activating Epo receptors (EpoR) on erythroid progenitor cells. Beyond successful treatment of anemia, ESAs have been reported to reduce damage following insult to organs, including the kidney, possibly via direct activation of EpoR. However, data on ESA effects outside hematopoietic functions are conflicting. Furthermore, limited use of appropriate EpoR-positive and EpoR-negative controls and lack of specific anti-EpoR antibodies make interpretation of data difficult. Recently positive and negative control cell types were validated and a sensitive and specific anti-EpoR antibody (A82) that detects low levels of EpoR protein was described. METHODS: A82 was used to measure EpoR protein levels in tissues, human renal cells and human cell lines by western blot analysis. Surface EpoR was examined on renal cells by measuring binding of [125I]-rHuEpo or antibodies. Renal cells and cell lines were treated with rHuEpo to see if phosphorylation of signaling proteins or proliferation/survival could be induced. Small inhibitory RNA (siRNA) were used to determine if EpoR knockdown affected cell viability. RESULTS: Total EpoR protein was low to undetectable in tissues and renal cells with no detectable EpoR on cell surfaces. EpoR knockdown had no effect on viability of renal cell lines. RHuEpo had no detectable effect on intracellular signaling on renal cell lines with no growth-promoting or survival effect on primary human renal cells. CONCLUSIONS: These results suggest that functional EpoR protein is absent on renal cells and that non-EpoR mechanisms should be explored to explain non-hematopoietic effects of ESAs.

Expanding the spectrum of genetic elements transferable by retroviral vectors.

Retroviral vectors are powerful tools to study gene function. However, conventional methods require a cellular transcription step to generate the genomic RNA for viral production. This limits the scope of genetic elements that may be transferred by these vectors, excluding many key gene regulatory signals, including RNA editing motifs, alternative splicing, and various promoter/enhancer constellations, as well as cytotoxic genes. To address this problem, we devised a simple approach where in vitro-synthesized vector genomic RNA is transfected into the cytoplasm of a packaging cell, allowing immediate viral particle assembly. We demonstrate that high-titer retroviruses that efficiently transduce mammalian cell lines and primary cells are readily generated. Importantly, we show that an intron-containing expression cassette can be transferred by this method, leading to increased expression levels in the target cell. Further, we demonstrate that the cap structure is not required for retroviral packaging, thus avoiding translation of vector-encoded genes in the packaging cell. This allows the retroviral transfer of cytotoxic genes or proteins that otherwise inhibit viral production.

Multiple roles for the receptor tyrosine kinase axl in tumor formation.

A focus of contemporary cancer therapeutic development is the targeting of both the transformed cell and the supporting cellular microenvironment. Cell migration is a fundamental cellular behavior required for the complex interplay between multiple cell types necessary for tumor development. We therefore developed a novel retroviral-based screening technology in primary human endothelial cells to discover genes that control cell migration. We identified the receptor tyrosine kinase Axl as a novel regulator of endothelial cell haptotactic migration towards the matrix factor vitronectin. Using small interfering RNA-mediated silencing and overexpression of wild-type or mutated receptor proteins, we show that Axl is a key regulator of multiple angiogenic behaviors including endothelial cell migration, proliferation, and tube formation in vitro. Moreover, using sustained, retrovirally delivered short hairpin RNA (shRNA) Axl knockdown, we show that Axl is necessary for in vivo angiogenesis in a mouse model. Furthermore, we show that Axl is also required for human breast carcinoma cells to form a tumor in vivo. These findings indicate that Axl regulates processes vital for both neovascularization and tumorigenesis. Disruption of Axl signaling using a small-molecule inhibitor will hence simultaneously affect both the tumor and stromal cell compartments and thus represents a unique approach for cancer therapeutic development.

Epo receptors are not detectable in primary human tumor tissue samples.

Erythropoietin (Epo) is a cytokine that binds and activates an Epo receptor (EpoR) expressed on the surface of erythroid progenitor cells to promote erythropoiesis. While early studies suggested EpoR transcripts were expressed exclusively in the erythroid compartment, low-level EpoR transcripts were detected in nonhematopoietic tissues and tumor cell lines using sensitive RT-PCR methods. However due to the widespread use of nonspecific anti-EpoR antibodies there are conflicting data on EpoR protein expression. In tumor cell lines and normal human tissues examined with a specific and sensitive monoclonal antibody to human EpoR (A82), little/no EpoR protein was detected and it was not functional. In contrast, EpoR protein was reportedly detectable in a breast tumor cell line (MCF-7) and breast cancer tissues with an anti-EpoR polyclonal antibody (M-20), and functional responses to rHuEpo were reported with MCF-7 cells. In another study, a functional response was reported with the lung tumor cell line (NCI-H838) at physiological levels of rHuEpo. However, the specificity of M-20 is in question and the absence of appropriate negative controls raise questions about possible false-positive effects. Here we show that with A82, no EpoR protein was detectable in normal human and matching cancer tissues from breast, lung, colon, ovary and skin with little/no EpoR in MCF-7 and most other breast and lung tumor cell lines. We show further that M-20 provides false positive staining with tissues and it binds to a non-EpoR protein that migrates at the same size as EpoR with MCF-7 lysates. EpoR protein was detectable with NCI-H838 cells, but no rHuEpo-induced phosphorylation of AKT, STAT3, pS6RP or STAT5 was observed suggesting the EpoR was not functional. Taken together these results raise questions about the hypothesis that most tumors express high levels of functional EpoR protein.

Identification of a sensitive anti-erythropoietin receptor monoclonal antibody allows detection of low levels of EpoR in cells.

Erythropoietin (Epo) binds and activates the Epo receptor (EpoR) on the surface of erythroid progenitor cells resulting in formation of erythrocytes. Recently, EpoR was reported to be expressed on non-erythroid cells suggesting a role for Epo outside of erythropoiesis. However those studies employed antibodies with questionable specificity and the significance of the observations are controversial. In order to accurately determine the expression of EpoR proteins in cells, we have generated a panel of novel anti-human EpoR monoclonal antibodies. One of these antibodies (A82) was particularly sensitive and it detected the EpoR protein on intact cells by flow cytometry and by western blot analysis with cell lysates. Both methods were optimized and using them, EpoR protein was detected by western immunoblotting with lysates from fewer than 200 EpoR positive control cells and the positive signals were proportional to EpoR protein expression level with a minimal signal in EpoR negative cells. The proteins detected by western blot analysis using A82 included full-length EpoR ( approximately 59kDa) as well as smaller EpoR fragments derived from the EPOR gene. These results indicate that A82 can be used to examine low level EpoR expression in cells by western and flow cytometry allowing an improved understanding of EpoR expression and metabolism.

European bat lyssavirus in Scottish bats.

We report the first seroprevalence study of the occurrence of specific antibodies to European bat lyssavirus type 2 (EBLV-2) in Daubenton's bats. Bats were captured from 19 sites across eastern and southern Scotland. Samples from 198 Daubenton's bats, 20 Natterer's bats, and 6 Pipistrelle's bats were tested for EBLV-2. Blood samples (N = 94) were subjected to a modified fluorescent antibody virus neutralization test to determine antibody titer. From 0.05% to 3.8% (95% confidence interval) of Daubenton's bats were seropositive. Antibodies to EBLV-2 were not detected in the 2 other species tested. Mouth swabs (N = 218) were obtained, and RNA was extracted for a reverse transcription-polymerase chain reaction (RT-PCR). The RT-PCR included pan lyssavirus-primers (N gene) and internal PCR control primers for ribosomal RNA. EBLV-2 RNA was not detected in any of the saliva samples tested, and live virus was not detected in virus isolation tests.