Hypoxia and estrogen are functionally equivalent in breast cancer-endothelial cell interdependence

First to determine whether TG1-1 cells are indeed responsive to hypoxia, we cultured cells under hypoxic conditions, specifically 1% O₂, in a sealed hypoxic chamber for the indicated number of hours. We observed an increase in HIF-1α in nuclear lysates and used TATA binding protein (TBP) as a nuclear loading control (Figure1A). Cells were also treated with cobalt chloride (CoCl₂), a HIF prolyl hydroxylase antagonist, used as a positive control for HIF-1α induction (Figure1B). HIF-1α accumulation peaked rapidly between 3-6 hours for both treatments and then returned to basal levels. To further demonstrate HIF-1α localization to the nucleus, TG1-1 cells were either untreated (left) or treated with CoCl₂ (right) for 24 hours and stained for VEGF (green) and HIF-1α (red). The panel on the right demonstrates an increase in HIF-1α staining intensity as well as co-localization with the nuclear DAPI stain compared to the left panel with low level diffuse HIF-1α cellular staining. Together these suggest that HIF-1α is an acceptable readout of hypoxia in TG1-1 cells.

Recent work has focused on the oxygen independent activation of HIF-1α in hormone responsive tissues by estrogen. To verify whether estrogen was able to induce HIF-1α in breast cancer cells in vitro, we treated the estrogen receptor positive TG1-1 cells with E₂ and observed an induction of HIF-1α in nuclear lysates at approximately 24 hours (Figure2A). Further, treatment of cells for 24 hours with E₂ and the pure anti-estrogen Fulvestrant abrogated E₂ induced accumulation of HIF-1α comparable to cells treated with the HIF-1α inhibitor YC1 (Figure2B) validating the E₂ stimulation of HIF-1α.

Stimulation of HIF-1α leads to dimerization with HIF-1β and nuclear translocation, where the heterodimer acts as a transcription factor leading to production of pro-angiogenic proteins. To test whether HIF-1α induction was functional, cytoplasmic cell lysates were isolated from TG1-1 cells treated with CoCl₂ and E₂ and western blots performed and probed for vascular endothelial growth factor (VEGF). Consistent with previous literature, hypoxia signaling leads to the expression of the pro-angiogenic protein VEGF in breast cancer cells in vitro (Figure3A). We also observed an increase in VEGF in cells treated with E₂ for 24 hours, an effect abrogated by Fulvestrant and more profoundly byYC1 (Figure3B). To measure functional secretion of VEGF, we performed and ELISA and observed a marked increase in VEGF secretion when cells were treated with E₂ or grown under hypoxic conditions (1% O₂). Similar to western blot observation, treatment of cells with YC-1 abrogated VEGF secretion, thus demonstrating the importance on HIF-1 in estrogen induced VEGF secretion (Figure3C). Thus, the pro-angiogenic effect of E₂ on breast cancer cells is not solely dependent on the nuclear translocation of estrogen receptor (ER) but rather on HIF-1α translocation as well.

We present evidence that E₂ stimulation of HIF-1α and VEGF is PI3K dependent. TG1-1 cells treated with E₂ for 24 hours show an increase in PI3K levels, an effect abrogated by Fulvestrant, further indicating functional ER signaling (Figure4A). Treatment of TG1-1 cells with E₂ for 24 hours in conjunction with the PI3K inhibitor Wortmannin prevented E₂ up regulation of HIF-1α (Figure4B). We observed the inhibition of PI3K also diminished E₂ stimulation of VEGF in cells treated for 24 hours (Figure4C). Thus, E₂ signals via the PI3K pathway to stimulate HIF-1α, and inhibition of this prosurvival pathway abrogates E₂ induction of angiogenic proteins.

Lastly, we sought to determine the cellular mechanism of estrogen induced neovasculogenesis in breast cancer progression. The process of neovasculogenesis is indispensible for tumor proliferation and metastasis and occurs largely in hypoxic tissues in which rapid tumor development quickly outgrows existing vasculature. Previously data from our laboratory demonstrated that E₂ enhanced breast tumor neovasculogenesis in vivo, however the mechanism remains unclear. To address this question, culture media from TG1-1 cells treated with E₂ with and without the anti-estrogen Fulvestrant or the HIF-1α inhibitor YC1 was used in an in vitro migration assay of human umbilical vein endothelial cells (HUVECs), culture media from untreated TG1-1 cells served as a control. We observed a significant increase in HUVEC migration toward the media of E₂ treated cells, which was abrogated by Fulvestrant as well as YC1 (Figure5A). Thus, E₂ stimulation of HIF-1α and consequent up regulation of VEGF leads to endothelial cell migration toward breast tumor cell secreted proteins. HUVEC migration experiments in which TG1-1 cells cultured under hypoxic conditions (1% O²) with and without E₂ and Fulvestrant also demonstrated only a modest synergistic effect of hypoxia and estrogen on endothelial cell migration which was not abrogated by the anti-estrogen (Figure5B). To further characterize the role of estrogen we focused on another phenotypic characteristic of vasculogenesis in vitro, namely the formation of tube shaped structures by endothelial cells utilizing the tubulogenesis assay. Briefly, HUVEC cells were plated over a bed of matrigel to simulate extracellular matrix and were exposed to either control basal media or conditioned media harvested from TG1-1 cultured cells as previously mentioned. We observed an increase in tube number and length when endothelial cells were treated with tumor cell condition media from E₂ stimulated cells (Figure6B). Estrogen induced tubulogenesis was abrogated in the presence of the anti-estrogen Fulvestrant (ICI), the HIF-1 inhibitor YC1, and cycloheximide. Analysis of lysates from tumor cells revealed that cycloheximide treatment prevents estrogen upregulation of HIF-1α, highlighting the importance of de novo protein synthesis in estrogen induced vasculogenesis in vitro (Figure6A). Media from TG1-1 cells grown under hypoxic conditions and concurrently treated with estrogen and inhibitors was also investigated for the impact on endothelial cell tube formation. When comparing conditioned media from tumor cells simultaneously grown under hypoxic conditions (1% O₂) with those also treated with E₂, we observed an increase in tubule formation which was abrogated by the HIF-1 inhibitor (Figure6C). This further highlights the importance of both hypoxia and estrogen as determinants of tumor induced neovasculogenesis.

The carcinoma cell line used for this study was TG1-1, a mouse mammary epithelial cell line, and the primary human endothelial cells HUVECs (human umbilical vein endothelial cells). TG1-1 was cultures in DMEM (Mediatech, Herndon, VA) supplemented with 10% fetal bovine serum (FBS) (Atlanta Biologicals, Atlanta, GA), penicillin 10,000 IU/mL, streptomycin 10,000 μg/mL (Mediatech) and 2mM L-glutamine (Mediatech). HUVEC (human umbilical vein endothelial cell) cells were obtained from American Type Culture Collection (ATCC) (Manassas, VA) and grown in F12K (Mediatech) supplemented with 10% FBS, 0.1mg/mL Heparin, 0.03mg/mL endothelial cell growth supplement (Sigma Aldrich). Cells were grown at 37°C in a humidified atmosphere with 5% CO₂ unless otherwise noted. For the cellular factors studied in this manuscript, the estrogen dependent, hypoxia induced change TG1-1 cells respond to, specifically producing VEGF, were similar to human breast cancer cells and hence the interdependent interaction of TG1-1 and HUVEC was rationalized.

For experiments, TG1-1 cells were grown to 75-80% confluence and serum-starved overnight in phenol red-free DMEM (Mediatech) supplemented with penicillin and streptomycin. For hypoxia experiments cells were grown in 1% O₂ in a modular incubator chamber (Billups-Rothenberg Inc.). Addition of E₂ to breast cancer cells was always on serum starved cells so as to define the estrogenic mediated increase in protein expression.

Cells were harvested using 0.25% trypsin (Mediatech), washed with PBS, and lysed (1X10⁶/100μL of lysis buffer) using the radioimmunoprecipitation assay (RIPA) buffer [50mM Tris–HCl (pH 7.4), 150 mM NaCl, 0.2% sodium deoxycholate, 0.1% SDS, 0.5% NP40, 1mM Pefabloc] and incubated on ice for 30 minutes with vortexing every 5 minutes. Samples were centrifuged at 14,000 rpm for 30 minutes at 4°C then supernatants collected for whole cell lysates. For nuclear/cytoplasmic isolation we used the NE-PER Nuclear and Cytoplasmic Extraction Kit from Thermo Scientific and followed manufacturer’s directions. Cell lysates (10-20μg) were subjected to 10% SDS-PAGE under reducing conditions (presence of β-mercaptoethanol) as previously described. Proteins were transferred to Immobilon-P membranes at 220 mA for 2 h and membranes were blocked in 5% dried milk in TBST [200mM Tris–HCl, pH 7.4, 150mM NaCl, and 0.01% Tween-20 added fresh/liter of 1X TBS (TBST)] for 2 h at room temperature on a shaker. After blocking, membranes were incubated overnight at 4°C with either HIF-1α (Abcam, Cambridge, MA), TBP (Abcam), VEGF (Santa Cruz Biotechnology, Santa Cruz, CA), Actin (Santa Cruz), or PI3K (Cell Signaling Technology) antibody in TBST. Membranes were subsequently washed three times in TBST and incubated with the respective horseradish peroxidase (HRP) conjugated secondary antibody (Pierce, Rockford, IL), for 2 h at room temperature in TBST containing 2% dried milk. Membranes were then washed three times with TBST and developed using ECL substrate (Pierce) and detected on Denville autoradiography film.

The RayBio Mouse VEGF Quantikine ELISA Kit (RayBiotech, Inc.,) was used to quantitate the levels of VEGF in conditioned media obtained from TG1-1 cells according to manufacturer’s instructions. ELISA was performed on each sample in duplicate. Protein content of cell pellets was determined in duplicate by using the Bradford protein assay (Bio-Rad).

BD Biocoat Control inserts (BD Biosciences, Bedford, MA) with 8-μm pore membrane filters were used for migration experiments as previously described. Briefly, TG1-1 cells were serum starved overnight for 16 h; the media was then replaced with serum free, phenol red-free DMEM (Mediatech) supplemented with ± 10^-8 M Estradiol (Sigma Aldrich), ± 10^-6 M Fulvestrant (Sigma), ± 10^-5 M YC-1 and culture media was subsequently harvested 24 h later. Conversely, TG1-1 cells were incubated under hypoxic conditions and media collected after 24 h. HUVEC cells were then harvested by trypsinization and 2.5 x 10⁴ cells were seeded in the upper chamber in 500 μl of serum-free, phenol red-free DMEM. The lower chamber contained 750 μl of the harvested TG1-1 media. After 18 h of incubation non-adherent cells were removed from the upper chamber using a cotton swab. Migrant cells on the lower surface were then fixed with 100% methanol and stained using 1% toluidine blue, 1% borax stain and then washed twice in distilled water. Inserts were allowed to dry and then visualized under 10x magnification. Experiments were performed in duplicate and data represents number of migrating cells per 10x field and normalized to cell counts of control treatment groups.

96 well plates were coated with 100 μl of growth factor reduced, phenol red-free Matrigel (BD Biosciences). HUVEC cells were harvested by trypsinization then added at a concentration of 10,000 cells/well in serum-free, phenol red-free DMEM or TG1-1 cell conditioned culture media as previously described. Plates were then incubated at 37°C for 4-6 h and visualized using 5x magnification, images were obtained using Axiovert 4.0.

TG1-1 cells were harvested as described and seeded at a density of 1 x 10⁴ into 8 well chamber slides (Becton Dickson) in complete DMEM and were allowed to adhere 24 h. Media was then removed and replaced with serum-free, phenol red-free DMEM and cells were serum starved overnight. Starvation media was removed and replaced with DMEM supplemented with ± 10^-8 M E2 ± 10^-6 Fulvestrant ± 10^-5 YC1 or grown under hypoxic conditions. Media was then removed and cells washed three times with phosphate buffered saline (PBS). Cells were then fixed with 4% para-formaldehyde at room temperature for 15 minutes then washed again three times with PBS. Cells were then permeabilized with 0.5% Triton-X for 5 min at room temperature and again washed three times with PBS. Cells were then blocked in 0.2% Triton-X, 10% goat serum (Sigma) and 3% bovine serum albumin (BSA) for 30 min at room temperature. Cells were then incubated overnight at 4°C with either HIF-1α or VEGF antibody in 1% BSA. Wells were then washed three times with PBS and incubated with the respective secondary antibody conjugated to either Alexa-fluor 488 or 595. Images were gathered using the Axiovision Rel 4.8 program under 40x magnification on the Axiovert 200M microscope (Carl Zeiss MicroImaging Inc., Thornwood, NY).