Evidence for erythropoietin signaling has been shown in several nonhematopoietic tissues, including many tumor types. Clinically, recombinant erythropoietin treatment of malignancy-related anemia has yet to be definitively associated with any modulation of chemotherapy or radiotherapy efficacy. Preclinically, recombinant erythropoietin has been shown to increase tumor oxygenation, but the direct effects of recombinant erythropoietin on tumor cells that express erythropoietin receptor are not yet fully characterized. This study examined the effects of exogenous recombinant erythropoietin on rodent mammary adenocarcinoma cells (R3230) in vitro and in vivo, and determined the effects of systemic recombinant erythropoietin on tumor growth delay in Taxol treatment. We showed that systemic recombinant erythropoietin treatment of rats bearing R3230 mammary carcinomas induced an increase in phospho-Akt levels within tumor cells. This was associated with a decrease in the frequency of apoptotic cells in tumors from recombinant erythropoietin–treated animals, but did not noticeably affect tumor growth rate. In vitro studies revealed that not only does recombinant erythropoietin induce Akt phosphorylation, but it also stimulates phosphorylation of p44/42 mitogen-activated protein kinases, Erk1 and Erk2. Activation of erythropoietin-mediated signaling in R3230 cells was associated with dose-dependent inhibition of apoptosis in response to Taxol treatment and serum starvation, an effect that was blocked by the addition of a phosphatidylinositol-3-kinase inhibitor. Despite its cytoprotective effects in vitro, recombinant erythropoietin did not significantly affect tumor growth delay in Taxol treatment. This study shows direct recombinant erythropoietin–mediated activation of specific intracellular signaling pathways in mammary adenocarcinoma cells in vivo and in vitro. Modulation of tumor apoptosis pathways by recombinant erythropoietin may have negative consequences by decreasing the chemosensitivity and radiosensitivity of erythropoietin receptor–positive breast tumors, although it did not have any obvious effects on growth with or without chemotherapy in this model. [Mol Cancer Ther 2006;5(2):356–61]
Erythropoietin is a glycoprotein hormone that regulates RBC production by binding to its specific cell surface receptor (EpoR) expressed in erythroid progenitor cells. There have been an increasing number of reports indicating that erythropoietin exhibits significant clinical effects in nonhematopoietic tissues. These diverse effects of erythropoietin are associated with the expression of EpoR in nonhematopoietic tissues, including cardiac muscle and nervous tissue, where erythropoietin acts as a tissue-protective cytokine (1). EpoR expression has also been reported in various tumor types, including cancer of the breast (2, 3), lung (2), colon (4), female reproductive tract (5), central nervous system (2), and various pediatric tumors (4). In hematopoietic cells, EpoR signaling is known to activate several intracellular kinase pathways, including the Jak/Stat, phosphatidylinositol-3-kinase (PI3K)/Akt, and p44/42 mitogen-activated protein kinase (MAPK) pathways (6). However, relatively little is known about EpoR signaling in tumor cells. Since the discovery of EpoR expression in cancer cells, there have been in vitro reports linking erythropoietin-EpoR signaling to the modulation of tumor cell proliferation (2, 7), chemosensitivity (8), and radiosensitivity (9). Recent reports suggest that signaling through EpoR may be important in breast tumorigenesis (3) and may decrease hypoxia-induced apoptosis in vitro (10). Recent clinical trials have raised concerns regarding the value of recombinant erythropoietin treatment during chemoradiation therapy of cancer patients (11, 12), although a recent metaanalysis showed that not only did recombinant erythropoietin treatment improve hemoglobin levels and decrease the need for transfusions in anemic cancer patients, but also provided suggestive but inconclusive evidence that recombinant erythropoietin may improve overall survival in these patients (13). This study is the first to characterize the direct effects of recombinant erythropoietin on mammary adenocarcinoma cells both in vitro and in vivo, focusing on the ability of recombinant erythropoietin to modulate tumor growth, apoptosis, and the activation of the PI3K/Akt and p44/42 MAPK pathways (14, 15). We also determined the effects of recombinant erythropoietin on tumor growth delay induced by Taxol treatment.
Materials and Methods
Female Fischer 344 rats (Charles River Laboratories, Raleigh, NC) were used for flank tumor studies. Tumor cells (5 million cells/200 μL) were injected into the subcutis of the right quadriceps muscle for flank tumors. Following injection of tumor cells, the animals were allowed free access to food and water ad libitum. Tumors were measured with calipers thrice a week, and tumor volumes were determined using the following formula (all measurements in millimeters):
volume = greatest transverse dimension × (orthogonal transverse dimension)2 × π/6.
All procedures and experiments were approved by the Duke Institutional Animal Care and Use Committee.
Recombinant Erythropoietin Tumor Growth Study
Eighteen days after tumor implantation, animals were randomized to treatment with s.c. saline (control) or recombinant erythropoietin (Procrit, Ortho-Biotech, Raritan, NJ) via an injection into the subcutis of the nape of the neck at a dose of 2,000 units/kg thrice a week. Animals received a total of six doses of saline or erythropoietin, and tumor excision was done 24 hours after the last injection.
Taxol Tumor Growth Delay Study
Seven to 10 days following tumor implantation, animals were randomized into six groups: (a) Taxol once weekly (10 mg/kg i.p.) + saline thrice weekly (50 μL s.c.), (b) Taxol once weekly (10 mg/kg i.p.) + recombinant erythropoietin thrice weekly (2,000 units/kg s.c.), (c) Taxol five times weekly (2 mg/kg i.p.) + saline thrice weekly (50 μL s.c.), (d) Taxol five times weekly (2 mg/kg i.p.) + recombinant erythropoietin thrice weekly (2,000 units/kg s.c.), (e) saline five times weekly (50 μL i.p.) + recombinant erythropoietin thrice weekly (2,000 units/kg s.c.), (f) saline five times weekly (50 μL i.p.) + saline thrice weekly (50 μL i.p. s.c.). Taxol was obtained as a sterile MDV manufactured by Bristol-Myers-Squibb (Princeton, NJ). Tumor measurements were done thrice per week, and the volume was determined using the formula above until a time of 40 days or five times its initial volume was reached. Once animals completed the study, they were euthanized with Euthasol.
Cell Cultures and Treatments
Rat mammary adenocarcinoma cells (R3230), previously reported to express EpoR (3), were used in all experiments. The cells were maintained in DMEM medium (Sigma, St. Louis, MO) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. R3230 cells were washed free of serum and incubated in serum-free medium containing recombinant erythropoietin (5, 10, or 100 units/mL; 24 hours) in the presence or absence of PI3K inhibitor wortmannin (Sigma) at a final concentration of 100 nmol/L (24 hours). In a separate experiment, cells were serum-starved and incubated with recombinant erythropoietin (10 or 100 units/mL) in the presence or absence of Taxol (0.5 μmol/L) for 24 hours.
R3230 flank tumors from the recombinant erythropoietin study were sectioned and stained for phospho-Akt, phospho-p44/42 MAPK (Cell Signaling Technology, Beverly, MA), Ki-67 (Biogenex, San Ramon, CA), and terminal nucleotidyl transferase–mediated nick end labeling (TUNEL; In situ Cell Death Detection Kit; Roche, Indianapolis, IN) using standard immunohistochemical techniques. Briefly, immunohistochemistry was done using 8 to 10 μmol/L serial sections of frozen tissue. Sections were fixed in 4°C cold acetone for 10 minutes. After endogenous peroxidase activity was quenched with 3% hydrogen peroxide for 15 minutes, tumor sections were blocked with 10% donkey serum for 15 minutes. Sections were incubated in primary rabbit polyclonal antibodies (dilution, 1:200) overnight at 4°C. After rinsing thrice with TBS, biotinylated donkey anti-rabbit antibody (Jackson ImmunoResearch, West Grove, PA) was applied for 30 minutes at room temperature. The slides were washed with TBS, followed by the application of an avidin-biotin complex (Vectastain ABC kit, Vector Lab, Inc., Burlingame, CA). The location of the reaction was visualized with 3,3′-diaminobenzidine tetrahydrochloride (DAB chromogen, Vector Lab). Slides were counterstained with hematoxylin. Omission of the primary antibody served as a negative control. For pAkt and pMAPK immunohistochemistry, staining was evaluated at a lower magnification (×200) by acquiring digital images from four to five randomly chosen fields which were then averaged. The extent of staining was quantified as the percentage of positively stained tumor cells. For Ki67 and TUNEL, slides were scanned at a low magnification (×100), and then cells with positively stained nuclei were counted in four to five random fields at a high magnification (×400). Consecutive sections stained with H&E were analyzed similarly to ensure a similar number of total cells per field. Mean proliferation and apoptotic indices were calculated manually by counting the ratio of positive nuclei to the total number of nuclei in each field.
Frozen R3230 flank tumors or monolayer cultures of R3230 cells were lysed and soluble proteins were analyzed by immunoblotting for phosphorylation of Akt and p44/42 MAPK in response to erythropoietin. Equal amounts of protein from each sample were subjected to SDS-PAGE, transferred to nitrocellulose membranes, and incubated with primary antibodies against phospho-Akt (Ser473, 1:200), phospho-Erk (Tyr202/204, 1:1,000), total Akt (1:1,000), or total Erk (1:1,000) from Cell Signaling Technologies. Following incubation with horseradish peroxidase–conjugated secondary antibodies, proteins were visualized using chemiluminescence (SuperSignal WestPico Chemiluminescence kit; Pierce, Rockford, IL) and autoradiography.
Cells undergoing recombinant erythropoietin and/or PI3K inhibitor wortmannin treatment, or recombinant erythropoietin and/or Taxol treatment were trypsinized and fixed with ice-cold ethanol, and then stained with propidium iodide. Samples were subjected to fluorescence-activated cell sorting analysis for cell cycle analysis. Apoptotic cells were identified as the subset with subdiploid nuclear content.
Statistical analyses were done using the unpaired Student's t test for comparison of means between two groups. For multigroup analyses, a one-way ANOVA and post hoc Student-Newman-Keuls method were used. In all analyses, P < 0.05 was considered significant.
Recombinant Erythropoietin Stimulates Akt Phosphorylation
We investigated the ability of recombinant erythropoietin to activate intracellular signal transduction in mammary adenocarcinoma cells. Recombinant erythropoietin–treatment of monolayer cultures of R3230 cells in vitro resulted in increased phosphorylation of Akt that was detected by immunoblotting (Fig. 1A). Furthermore, systemic recombinant erythropoietin given to rats with R3230 flank tumors was also associated with increased levels of phosphorylated Akt in the tumors (Fig. 1B). Immunostaining of the excised flank tumors for phospho-Akt revealed significantly increased Akt phosphorylation in tumor cells from recombinant erythropoietin–treated animals compared with negative controls (Fig. 2). There were no detectable differences in phospho-p44/42 MAPK signaling by immunohistochemical analysis of flank tumors, although recombinant erythropoietin treatment of R3230 cells in vitro induced phosphorylation of Erk1 and Erk2 in a biphasic fashion (Fig. 1C). Recombinant erythropoietin–mediated phosphorylation of Akt and Erk1 and Erk2 was not associated with an increase in tumor cell proliferation in vitro (data not shown).
Recombinant Erythropoietin Inhibits Apoptosis
R3230 tumor cells were subjected to serum starvation and treated with recombinant erythropoietin in the presence or absence of PI3K inhibitor wortmannin for 24 hours. Cell cycle analysis revealed a decrease in the number of cells undergoing apoptosis with increasing concentrations of recombinant erythropoietin (Fig. 3A). The addition of a PI3K inhibitor (wortmannin) blocked the ability of recombinant erythropoietin to inhibit apoptosis. In a separate experiment, R3230 tumor cells were serum-starved and treated with recombinant erythropoietin and/or Taxol. Recombinant erythropoietin induced a dose-dependent decrease in the number of cells undergoing Taxol-induced apoptosis (Fig. 3B).
Sections of tumors from the Taxol growth delay study were also analyzed for apoptotic index by TUNEL staining. Microscopic analysis revealed a significant decrease in the number of apoptotic cells in recombinant erythropoietin–treated tumors compared with placebo-treated tumors (Fig. 4A). Tumors treated with a conventional, high-dose Taxol dosing regimen showed a significantly higher apoptotic index than those treated with a metronomic Taxol dosing regimen. The addition of recombinant erythropoietin did not affect the apoptotic response to either dosing regimen.
These sections were also stained for Ki67, an endogenous marker of proliferation. Recombinant erythropoietin treatment alone had no effect on the proliferation of R3230 tumors (Fig. 4B). Both dosing regimens significantly inhibited proliferation in flank tumors, although there was no difference between the two dosing regimens and the addition of recombinant erythropoietin to either did not further affect proliferation index.
Recombinant Erythropoietin Does Not Affect Tumor Growth or Tumor Growth Delay
To investigate whether recombinant erythropoietin affected tumor growth rate, R3230 flank tumors were randomized to systemic recombinant erythropoietin treatment or saline control. There were no differences in tumor growth rates (tumor volumes) or proliferation index as determined by Ki-67 staining (data not shown). Recombinant erythropoietin increased the hematocrit of treated animals from a mean baseline of 45.9% to 63.4%.
The effect of recombinant erythropoietin on R3230 tumor growth delay in response to two different dosing regimens of Taxol was also determined. The two regimens consisted of either once weekly dosing (10 mg/kg) or metronomic dosing (five times a week at a daily dose of 2 mg/kg). Recombinant erythropoietin did not affect tumor growth delay with either treatment regimen (Fig. 5). The high-dose regimen was significantly more effective than the metronomic dose of Taxol in this model.
Erythropoietin signaling has recently been shown to be biologically important in nonhematopoietic tissues, inducing cardioprotective and neuroprotective phenotypes against various types of tissue injury in multiple experimental models. Activation of Akt has been shown to be important in mediating the cytoprotective effects of EpoR signaling in several tissue types. In human erythroid precursors, Akt activation is essential for the inhibition of apoptosis in response to erythropoietin (16). Furthermore, Chong et al. have shown that Akt activation plays an important role in the antiapoptotic effects of erythropoietin in vascular endothelial cells (17) and neural tissue (18).
There have been reports of functional EpoR expression in many malignant tumor types, although several studies have shown conflicting results regarding how erythropoietin-EpoR signaling could affect tumor biology and response to chemoradiotherapy (8, 9). These prior studies were conducted in vitro and may not reflect the phenotype of how erythropoietin will affect tumor cell function in vivo.
In the current study, we show that recombinant erythropoietin treatment of rodent mammary adenocarcinoma cells induces the phosphorylation of Akt and Erk1-2 in vitro and in vivo. Systemic recombinant erythropoietin given to rats harboring R3230 flank tumors resulted in a significant increase in Akt phosphorylation in tumor cells detected 24 hours after the drug was given. The ability of recombinant erythropoietin to modulate in vivo Akt phosphorylation was associated with decreased apoptotic index in tumors not treated with chemotherapy. The in vitro data also shows a dose-dependent effect for recombinant erythropoietin to decrease the apoptosis of R3230 cells during serum starvation (an effect that was blocked by PI3K/Akt inhibitor), and to decrease R3230 tumor cell apoptosis in response to Taxol treatment.
However, the activation of intracellular signaling by recombinant erythropoietin was not associated with an increase in tumor cell proliferation in vitro or tumor growth rate in vivo. Furthermore, the addition of recombinant erythropoietin to both Taxol dosing regimens did not significantly affect the proliferative or apoptotic response of tumors in vivo. Tumor growth rate is directly proportional to the growth fraction (actively dividing cells) and is inversely proportional to cell cycle time and the rate of cell loss (apoptosis and necrosis). If recombinant erythropoietin were to decrease the rate of tumor cell apoptosis, one would predict that the growth rate would increase. However, cells that are “rescued” from apoptosis by recombinant erythropoietin do not necessarily re-enter the growth fraction of the tumor. These cells could easily enter the G0 portion of the cell cycle as “resting” cells, and, as such, would not contribute to the active growth fraction of the tumor. This effect may apply in the presence or absence of antitumor therapy. The addition of systemic recombinant erythropoietin treatment to two dosing regimens of Taxol (a high dose once per week or low doses five times per week) did not significantly affect tumor growth delay.
This study shows a correlation between EpoR intracellular signaling and cytoprotection in breast cancer cells in vitro. The observation that systemic treatment with recombinant erythropoietin could activate EpoR's on tumor cells and inhibit apoptosis has potential implications for both the chemosensitivity and radiosensitivity of breast tumors that express EpoR, even though recombinant erythropoietin did not adversely affect Taxol treatment in this model. It is likely that the overall clinical effect of systemic recombinant erythropoietin treatment is dependent on a balance of several distinct actions on tumor cells and the tumor microenvironment. Any negative cytoprotective effects recombinant erythropoietin might have on tumor cells may be counterbalanced by improvements in tumor oxygenation (19) and hemoglobin levels also seen with systemic recombinant erythropoietin treatment, which would be expected to improve chemosensitivity.
No studies have thus far shown a direct recombinant erythropoietin effect on tumor growth in vivo. Systemic recombinant erythropoietin therapy was shown to improve tumor oxygenation in a hemoglobin-independent fashion in preclinical models (19), an effect which may be associated with the enhanced efficacy of chemoradiotherapy in vivo (20). It is important to establish whether the potential negative consequences (shown by the cytoprotective in vitro results) of recombinant erythropoietin therapy in patients with EpoR-positive breast tumors may outweigh the benefits of treating anemia with systemic recombinant erythropoietin. The results of this study indicate that the positive benefits of Taxol are unaffected by recombinant erythropoietin and that systemic recombinant erythropoietin for its indicated use in the treatment of malignancy-related anemia remains a safe therapeutic option. Additional studies in other tumor models are warranted, and future studies will further explore the clinical implications of the direct effects of recombinant erythropoietin on tumor cells to determine if its use could be fully optimized when combined with either chemotherapy and/or radiotherapy.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.