Erythropoietin receptor contributes to melanoma cell survival in vivo

Abstract

Erythropoietin (Epo) is widely used clinically to treat anemia associated with various clinical conditions including cancer. Data from several clinical trials suggest significant adverse effect of Epo treatment on cancer patient survival. However, controversy exists whether Epo receptor (EpoR) is functional in cancer cells. In this study, we demonstrated that EpoR mRNA expression was detectable in 90.1% of 65 melanoma cell lines, and increased copy number of the Epo and EpoR loci occurred in 30 and 24.6% of 130 primary melanomas, respectively. EpoR knockdown in melanoma cells resulted in diminished ERK phosphorylation in response to Epo stimulation, decreased cell proliferation and increased response to the inhibitory effect of hypoxia and cisplatin in vitro. EpoR knockdown significantly decreased melanoma xenograft size and tumor invasion in vivo. On the contrary, constitutive activation of EpoR activated cell proliferation pathways in melanoma cells and resulted in increased cell proliferation and resistance to hypoxia and cisplatin treatment in vitro. EpoR activation resulted in significantly larger xenografts with increased tumor invasion of surrounding tissue in vivo. Daily administration of recombinant Epo fails to stimulate melanoma growth in vivo, but the treatment increased vascular size in the xenografts. Increased local recurrence after excision of the primary tumors was observed after Epo treatment. Epo induced angiogenesis in Matrigel plug assays, and neutralization of Epo secreted by melanoma cells results in decreased angiogenesis. These data support that EpoR is functional in melanoma and EpoR activation may promote melanoma progression, and suggest that Epo may stimulate angiogenesis and increase survival of melanoma cells under hypoxic condition in vivo.

Figure 1. Effects of EpoR knockdown in vitro

A. EpoR mRNA expression. Quantitative RT-PCR of EpoR in melanoma cells with control scrambled or EpoR shRNA. * indicates p<0.05 comparing to control shRNA. Bars show mean ± SEM from three separate experiments. B. EpoR protein expression. Western blot analysis of WM35, WM793 and 1205Lu cells transfected with control or EpoR shRNA. Representative image from three independent experiments. C. ERK phosphorylation in response to Epo. Melanoma cells with control or EpoR shRNA were stimulated with Epo (10U/ml). D. ERK phosphorylation in response to Epo. Melanoma cells with control or EpoR siRNA were stimulated with Epo (10U/ml). E. EpoR knockdown and cell migration. WM35, WM793 and 1205Lu melanoma cells were placed in growth factor reduced Matrigel Boyden chambers. Cells that migrated through the chamber were quantified. Bars show mean ± SEM from three separate experiments. * indicates p<0.05 comparing to 1205Lu control cells. F. Cell proliferation under nomoxia in vitro. WM35, WM793 and 1205Lu cells with control or EpoR shRNA were cultured under nomoxia and quantified. * indicates p<0.05 comparing to corresponding controls. G. Effect of Epo and cisplatin on cell proliferation. WM35 melanoma cells with control or EpoR shRNA were treated with cisplatin (100 µM), Epo (5U/ml) or both for 24 hours. Live cells were counted using trypan blue dye exclusion assay. Bars show mean ± SEM from three separate experiments. * indicates p<0.05 comparing to corresponding controls. NS indicates not statistically significant. H. Cell proliferation under hypoxia in vitro. 1205Lu cells with control or EpoR shRNA were cultured under 1% O2 for 48 hours, and cell proliferation was quantified. Bars show mean ± SEM from three separate experiments. * indicates p<0.05 compared with 1205Lu cells with control shRNA.

Figure 2. Effects of EpoR knockdown in vivo

WM35^EpoRkd, WM793^EpoRkd, 1205Lu^EpoRkd or respective control cells were injected into the flanks of nude mice. All mice were sacrificed after 5 weeks. A. Photographs of xenografts. Arrows point to the tumors and arrow head points to bulging abdomen with ascites in the mouse injected with 1205Lu cells. Bar indicates 4 cm. B. Tumor growth rate. After subcutaneous injection, tumor size was measured weekly. * indicates p<0.05 comparing to respective control cells. Bars show mean ± SEM from 4 xenografts. C. Histology of xenografts formed by WM793 or WM793^EpoRkd cells. Arrows points to the area with tumor necrosis. Bar indicates 3 mm. D. Mitotic rate in the xenografts. Mitotic rate was quantified per mm². * indicates p<0.05 compared with the control.

Figure 3. Effects of constitutive EpoR activation in vitro

We transfected the 451Lu melanoma cells with EpoR-R129C or control empty vectors. A. Activation of EpoR signaling pathways. Western blot analysis phosphorylated and total JAK2, ERK and AKT proteins in melanoma cells with control or EpoR-129C vectors. B. Cell proliferation under nomoxia. 451Lu, WM793, and 1361C cells with control or EpoR-R129C plasmids were cultured under normoxic condition and cell proliferation was quantified. * indicates p<0.05 comparing to respective controls. The data represent mean ± SEM from three separate experiments. C. Cell proliferation under hypoxia. 451Lu, WM793, and 1361C cells with control or EpoR-R129C plasmids were cultured under hypoxic condition and cell proliferation was quantified. * indicates p<0.05 comparing to respective controls. The data represent mean ± SEM from three separate experiments. D. Cell proliferation after Epo and cisplatin treatment. 451Lu melanoma cells with control or EpoR-R129C plasmids were treated with Epo (5U/ml), cisplatin (100µM), or Epo/cisplatin for 24 hours, and surviving cells were quantified. Bars show mean ± SEM from three separate experiments. * indicates p<0.05 compared to cells with control shRNA.

Figure 4. Effects of constitutive EpoR activation in vivo

A. Tumor growth rate in vivo. Melanoma cells with control or EpoR-R129C were injected into the flank area of NOD/SCID mice and tumor size was measured weekly. * indicates p<0.05 comparing to control. Bars show mean ± SEM from 5 xenografts. B. Morphology of xenografts. Bar indicates 1 cm. C. Tumor histology. Arrows points to tumor necrosis in xenografts formed by melanoma cells with control shRNA (left panel). EpoR-R129C expressing tumor cells invaded into surrounding femoral bone. Arrow points to bone (right panel). Bar indicates 200 µm. D. CAIX expression. Tumor sections were stained with an antibody to CAIX. Bar indicates 200 µm. E. Quantification of tumor necrosis. The area with tumor necrosis was measured using a micrometer. * indicates p<0.05 comparing to control.

Figure 5. Effects of recombinant Epo treatment in vivo

GFP-1232Lu melanoma cells were injected subcutaneously into athymic nude mice. Epo group received daily subcutaneous Epo injection (2000 U/kg) in 100ul saline and the control group receiving daily injection of saline (100 ul) for 8 weeks. A. Hematocrit. Hematocrit was measure after 8 weeks of Epo treatment. * indicates p<0.05 comparing to control. Bars show mean ± SEM from 4 mice. B. Tumor growth rate in vivo. Tumor size was measured twice weekly. C. Histology and CD34 stain of primary xenografts. The tumor vasculature in the Epo-treated group appeared more dilated than the control group. The sections were stained with anti-CD34 antibody. D. Open vessel density. Vessels with lumen width more than 0.03 mm were quantified per mm². * indicates p<0.05 comparing to control. E. The primary tumors were removed, and the wounds were closed. These mice were observed for another 6 weeks. The Epo treated group continued to receive twice a week injection of Epo (2000U/kg) during that period of time. In vivo imaging showed that the recurrent tumors were larger in Epo treated group. F. Tumor volume was measured at the end of the sixth week. The secondary tumors at the site of local recurrence were significantly larger in the Epo treated group. * indicates p<0.05 comparing to control. Bars show mean ± SEM from 4 mice.

Figure 6. Epo and angiogenesis in vivo

A and B. Matrigel plug assay (A: gross; B: Histology). Athymic nude mice received subcutaneous injection of growth factor reduced Matrigel plugs supplemented with saline, Epo or VEGF. Angiogenesis was observed in plugs supplemented with VEGF or Epo. Arrows point to the blood vessels. C. Modified Matrigel plug assay. Athymic nude mice received subcutaneous injection of growth factor reduced Matrigel plugs supplemented with WM35 melanoma cells or WM35 cells plus anti-Epo antibody. Arrows points to blood filled vessels (upper panels). The sections were stained with an antibody to CD34. Arrows points to the stained open blood vessels (lower panel). D. Quantitative analysis of vessel density. The vessel density in the plugs was counted per mm². * indicates p<0.05 comparing to control. Bars show mean ± SEM from 6 Matrigel plugs.