Copper activates HIF-1α/GPER/VEGF signalling in cancer cells.

Copper promotes tumor angiogenesis, nevertheless the mechanisms involved remain to be fully understood. We have recently demonstrated that the G-protein estrogen receptor (GPER) cooperates with hypoxia inducible factor-1α (HIF-1α) toward the regulation of the pro-angiogenic factor VEGF. Here, we show that copper sulfate (CuSO4) induces the expression of HIF-1α as well as GPER and VEGF in breast and hepatic cancer cells through the activation of the EGFR/ERK/c-fos transduction pathway. Worthy, the copper chelating agent TEPA and the ROS scavenger NAC prevented the aforementioned stimulatory effects. We also ascertained that HIF-1α and GPER are required for the transcriptional activation of VEGF induced by CuSO4. In addition, in human endothelial cells, the conditioned medium from breast cancer cells treated with CuSO4 promoted cell migration and tube formation through HIF-1α and GPER. The present results provide novel insights into the molecular mechanisms involved by copper in triggering angiogenesis and tumor progression. Our data broaden the therapeutic potential of copper chelating agents against tumor angiogenesis and progression.


INTRODUCTION
Copper, which is an essential trace element naturally occurring in soil, water and air, acts as a catalytic and/or structural cofactor in a wide array of important biological processes like embryogenesis, growth, homeostasis and angiogenesis [1,2]. An elevated exposure to copper may be mainly consequent to environmental pollution from the manufacture of wire, sheet metal, pipe and other metal products [2]. In addition, mining, waste dumps, combustion of fossil fuels, wood production and phosphate fertilizers release copper in the environment, thus contributing to the actual exposure in humans [2][3][4]. To date, mismanaged or high copper levels have been involved in the generation of oxidative stress [5] which plays an important role in cancer development [6]. In this regard, it should be mentioned that physiological concentrations of copper range approximately from 18 to 31 µM [7], while serum copper levels have been found in cancer patients from 50 µM to 205 µM or even at mM concentrations [8][9][10]. Of note, elevated copper concentrations were correlated with cancer stage and/or progression in diverse types of tumors, thus suggesting that copper may be a useful prognostic factor and a marker of responsiveness to therapy [reviewed in 8]. On the basis of these findings, a number of studies investigated the stimulatory action of copper on VEGF production and tumor angiogenesis [11][12][13] and the repressive effects exerted by copper-chelating on HIF-1α mediated expression of VEGF [14,15].
Numerous G-protein coupled receptors (GPCRs) contribute to the angiogenic switch through mechanisms that include their functional interaction with HIF-1α toward VEGF expression [16]. In this regard, our recent study has shown that the G protein estrogen receptor (GPER) cooperates with HIF-1α in order to modulate www.impactjournals.com/oncotarget VEGF in hypoxic breast tumor microenvironment [17].
In addition, we have demonstrated that estrogenic GPER signalling activates HIF-1α/VEGF transduction pathway leading to angiogenesis and tumor growth [18].
Here, we provide novel evidence on the mechanisms by which copper triggers the EGFR/ERK/c-fos signalling cascade along with GPER and HIF-1α toward VEGF expression and function in cancer cells. We also show that GPER may be considered as an additional target of copper chelating agents, hence broadening the therapeutic potential of these chemicals against tumor angiogenesis and progression.

CuSO 4 induces the expression of the proangiogenic factor VEGF
Considering that copper and its chelating agents have been involved in tumor angiogenesis [5], we asked whether copper sulfate (CuSO 4 ) may induce the expression of the pro-angiogenic factor VEGF and its transcriptional regulator HIF-1α in SkBr3 breast cancer cells and HepG2 hepatocellular carcinoma cells. Of note, CuSO 4 induced the mRNA expression of both HIF-1α ( Figure 1A) and VEGF ( Figure 1B) in a dose dependent manner, starting from 25 µM and reaching the strongest stimulation upon concentrations ranging from 100 to 200 µM. Taking into account these results and considering that in previous studies relevant biological responses to copper exposure were observed up to 500 µM [19][20][21], in the subsequent assays of the current study 200 µM CuSO 4 were used. First, we determined that CuSO 4 up-regulates in a time-dependent manner the mRNA expression of HIF-1α ( Figure 1C) and VEGF ( Figure 1D) in SkBr3 and HepG2 cells. Thereafter, we ascertained that the well-acknowledged copper chelating agent TEPA [14,22] as well as the extensively used ROS scavenger NAC [reviewed in 23] prevent the mRNA induction of HIF-1α ( Figure 1E) and VEGF ( Figure 1F) and the transactivation of a VEGF promoter construct ( Figure 1G) upon treatment with CuSO 4 . As copper has been previously involved in HIF-1α responses to low oxygen conditions [14,15], we then assessed the effect of TEPA on the action of the hypoxia-mimetic agent CoCl 2 . As expected, CoCl 2 induced the mRNA expression of HIF-1α ( Figure 2A) and VEGF ( Figure 2B) as well as the transactivation of a VEGF promoter construct ( Figure 2C) in SkBr3 and HepG2 cells. Interestingly, these effects were abolished in the presence of TEPA and rescued adding CuSO 4 to SkBr3 and HepG2 cells ( Figure 2). Results similar to those observed using CoCl 2 were obtained culturing SkBr3 and HepG2 cells in a low oxygen tension (2% O 2 ) (Supplementary Figure 1A-1C).
Collectively, these findings suggest that CuSO 4 may be involved in the activation of HIF-1α/VEGF transduction signalling in cancer cells. On the basis of our recent findings suggesting that a functional crosstalk between HIF-1α and GPER may occur toward the VEGF expression in hypoxic conditions [24,[17][18], we next determined that the up-regulation of GPER mRNA expression induced by CuSO 4 in SkBr3 and HepG2 cells ( Figure 3A, 3B) is abolished in the presence of both TEPA and NAC ( Figure 3C). Moreover, the transactivation of a GPER promoter construct triggered by CuSO 4 was prevented using TEPA and NAC ( Figure 3D). Notably, the GPER mRNA induction and the GPER promoter transactivation induced by CoCl 2 were prevented in the presence of TEPA and rescued adding CuSO 4 ( Figure  3E, 3F). Results comparable to those observed upon CoCl 2 treatment were obtained culturing cells in a low oxygen tension (2 % O 2 ) (Supplementary Figure 1D-1E). Cumulatively, these data recall previous studies showing that the inhibitory effects of TEPA on hypoxia-induced responses are rescued by CuSO 4 in a dose-dependent manner [14].
Altogether, these data indicate that GPER may be included among the transduction mediators triggered by copper, in particular in stressful conditions characterized by a low oxygen tension in cancer cells.

Molecular mechanisms involved in the stimulatory actions elicited by CuSO 4
As c-fos expression is a molecular sensor of both GPER and HIF-1α signalling [18,[25][26], we also demonstrated that c-fos mRNA increase upon CuSO 4 stimulation ( Figure 3G, 3H) is abrogated in the presence of TEPA and NAC ( Figure 3I). Nicely fitting with these results, the transactivation of a c-fos luciferase construct and AP1-luc promoter sequence induced by CuSO 4 was repressed in the presence of TEPA and NAC ( Figure 3J). Recapitulating the aforementioned findings, the protein induction of c-fos, HIF-1α and GPER observed upon CuSO 4 treatment was abrogated in the presence of TEPA and NAC in SkBr3 and HepG2 cells ( Figure 4A-4D). Given that the activation of EGFR/ERK signalling triggers transduction mechanisms leading to gene expression changes as mentioned above [17-18, 25, 27-28], we ascertained that the EGFR and ERK1/2 phosphorylation induced by CuSO 4 in both SkBr3 and HepG2 cells ( Figure  5A, 5B) is blocked in the presence of the EGFR tyrosine kinase inhibitor AG1478 (AG) and the MEK inhibitor PD98059 (PD) ( Figure 5C, 5D) as well as using TEPA and NAC ( Figure 5E, 5F). Further corroborating these data, the up-regulation of c-fos, HIF-1α, GPER and VEGF mRNA expression ( Figure 6A-6D) as well as the transactivation of fos-luc, AP1-luc, GPER-luc and VEGFluc reporter constructs ( Figure 6E  were abolished in the presence of AG and PD. Analogous findings were obtained evaluating the regulation of c-fos, HIF-1α and GPER protein expression in SkBr3 and HepG2 cells ( Figure 6G, 6H).
Immunofluorescence experiments performed in SkBr3 cells showed that TEPA, NAC, AG and PD prevent also the increase of VEGF protein expression upon CuSO 4 treatment (Figure 7). In addition, the HIF-1α protein increase triggered by CuSO 4 was no longer evident transfecting SkBr3 and HepG2 cells with a plasmid encoding a dominant/negative c-fos mutant (DN/c-fos) ( Figure 8A, 8B). In accordance with the aforementioned results, the up-regulation of GPER ( Figure 8A, 8B) and VEGF ( Figure 8C, 8D) protein levels upon CuSO 4 treatment was prevented by DN/c-fos, as evaluated by immunoblotting and immunofluorescence assays, respectively. As demonstrated in our previous investigations, in hypoxic tumor microenvironment HIF-1α mediates the expression of GPER that contributes to the regulation and function of VEGF [17,24]. Likewise, we found that the GPER protein up-regulation induced by CuSO 4 as well as the transactivation of the GPER promoter were abolished knocking down HIF-1α expression ( Figure 9A-9F). In addition, the silencing of HIF-1α prevented the CuSO 4 -induced activation of the VEGF promoter construct ( Figure 9G, 9H) as well as the up-regulation of VEGF protein expression ( Figure 9I-9K). Of note, GPER was required for VEGF protein induction

HIF-1α and GPER are required for VEGFinduced endothelial tube formation, cell migration and proliferation induced by CuSO 4
Having established that HIF-1α and GPER cooperate in triggering the up-regulation of VEGF by CuSO 4 , we assessed in HUVECs the involvement of HIF-1α and GPER in the formation of tubule-like structures that represent a useful experimental model of angiogenic process [29]. Interestingly, a ramified network of tubules was generated in HUVECs cultured in conditioned medium from CuSO 4 -treated SkBr3 cells ( Figure 11A). However this effect was prevented by knocking down the expression of HIF-1α or GPER ( Figure 11B-11H). The addition of VEGF to the medium collected from CuSO 4 -treated and GPER-silenced SkBr3 cells rescued the generation of tubule structures in HUVECs ( Figure  11C). Figure 11 (panels D-F) recapitulates these results, suggesting that VEGF may be considered as a target of copper-activated HIF-1α/GPER signalling toward new blood vessels formation. As in previous studies VEGF boosted endothelial cells migration [30][31] we then evaluated whether HIF-1α and GPER are involved in the migration of HUVECs. Conditioned medium from SkBr3 cells exposed to CuSO 4 induced the migration of HUVECs ( Figure 12A), however this response was abrogated silencing HIF-1α and GPER expression ( Figure 12B-12E).
Indeed, the addition of VEGF rescued cell migration culturing HUVECs in medium collected from SkBr3 cells which were GPER-silenced and treated with CuSO 4 ( Figure 12C). Next, we determined that HIF-1α and GPER are required for SkBr3 cell proliferation induced by CuSO 4 , as this response was prevented knocking-down their expression (Supplementary Figure S2A-C). Likewise, the growth effects elicited by CuSO 4 was abolished in the presence of TEPA (Supplementary Figure S2D). Altogether, these findings suggest that copper may trigger relevant biological actions through HIF-1α/GPER/VEGF transduction signalling in both cancer and endothelial cells toward angiogenesis and tumor progression.

DISCUSSION
The present study provides novel evidence regarding the molecular mechanisms by which copper may trigger the expression and function of VEGF toward angiogenesis and tumor progression. In particular, we have shown that copper activates the EGFR/ERK/c-fos transduction pathway leading to the expression of HIF-1α, GPER and VEGF in breast and hepatic cancer cells. In this regard,     we demonstrated that a functional cooperation between HIF-1α and GPER contributes to VEGF regulation in cancer cells exposed to copper. Recalling previous studies on the capability of copper chelating agents to elicit anti-tumor effects [5,32], we have also evidenced that these chemicals exert an inhibitory action on HIF-1α/GPER/ VEGF transduction pathway. Next, we have found that HIF-1α and GPER are required for endothelial tube formation and cell migration stimulated by VEGF as well as for copper-induced proliferation of breast cancer cells.
The role of copper in tumor initiation and progression has been extensively investigated both in vitro and in vivo [8,33]. In this context, high copper levels ranging from 50 to 200µM have been correlated with incidence and recurrence in cancer patients [8,9]. In accordance with these findings, we have ascertained that copper exerts stimulatory effects on gene expression starting from a concentration of 25 µM, even though the maximal responses were observed using a concentration of 200 µM. Hence, the last amount was used in all assays to better evaluate the potential of copper to activate the aforementioned biological activity. Previous studies have disclosed that certain effects elicited by copper in cancer cells rely on the generation of reactive oxygen species (ROS), which act as second messenger in triggering stimulatory signals [8]. In this regard, it has been shown the transduction mechanisms involved, that include the activation of the EGFR/ERK pathway and the expression of genes mediating growth responses like c-fos [reviewed in 5]. On the basis of these observations, it could be argued that copper may mimic some biological features which characterize the hypoxic tumor environment.
HIF-1 acts as a survival factor upon low oxygen conditions regulating the expression of genes involved in cell metabolism, migration, invasion and angiogenesis [34][35]. In this vein, it is worth mentioning that copper was shown to increase HIF-1α stabilization and accumulation [19]. Further extending these findings, our current results indicate that copper is also able to induce HIF-1α expression, thus providing a new mechanism through which this chemical may be involved in cancer progression. Previous studies have determined that GPER contributes together with HIF-1α to the adaptive responses to hypoxic tumor microenvironment [17,24]. Nicely fitting with these observations, the present data reveal that copper induces the expression of GPER through HIF-1α, leading to the regulation of VEGF in breast cancer cells and cancer associated fibroblasts (CAFs) [17]. The stimulatory role of copper in cancer development has been also proved by copper chelating agents as a reduction in tumor volume, vascular permeability, tumor's microvascular supply and micrometastasis generation has been reported lowering copper levels in diverse experimental models [5]. Extending the current knowledge on the action of anti-copper drugs like TEPA, our data indicate that these chemicals may also target HIF-1α/GPER signalling among the multifaceted responses triggered in cancer cells.
Here, we have provided novel evidence regarding the action elicited by copper toward tumor angiogenesis and progression. On the basis of the present findings GPER may be included together with HIF-1α and VEGF among the molecular targets of copper chelating agents in combination therapies. Nevertheless, further studies are needed to better define the role of copper on the functional interaction between GPER, HIF-1α and VEGF in malignant cells and tumor microenvironment.

Cell cultures
We used SkBr3 breast cancer cells and HepG2 hepatocarcinoma cells that represent a valuable tool for the evaluation of the transduction pathways activated by copper in cancer cells. As both cell lines express GPER, which has been involved with the angiogenic process within the tumor microenvironment [17][18], this model system is suitable to ascertain the contribution of GPER to copper action toward tumor angiogenesis.

Gene reporter assays
The 2.6 kb VEGF promoter-luciferase construct containing full-length VEGF promoter sequence (22,361 to +298 bp relative to the transcription start site) used in luciferase assays was a kind gift from dr. P. Soumitro (Harvard Medical School, Boston, Massachusetts). The GPER promoter-luciferase construct (pGPER 2.9 kb) was obtained as previously described [24].
The luciferase reporter plasmid for AP-1 responsive collagen promoter was a kind gift from H. Van Dam (Department of Molecular Cell Biology, Leiden University, Leiden, Netherlands). The luciferase reporter plasmid for c-fos, encoding a -2.2 kb 5' upstream fragment of human c-fos, kindly provided by K. Nose (Department of Microbiology, Showa University School of Pharmaceutical Sciences, Hatanodai, Shinagawa-ku, Tokyo, Japan). SkBr3 and HepG2 cells (1 x 10 5 ) were plated into 24-well dishes with 500µL/well culture medium containing 10% FBS. Transfections were performed using X-treme GENE 9 DNA transfection reagent as recommended by the manufacturer (Roche Diagnostics, Milan, Italy), with a mixture containing 0.5µg of reporter plasmid and 10 ng of pRL-TK. After 24 h, cells were treated with CuSO 4 , alone and in combination with TEPA, NAC, AG1478 and PD98059, as indicated. For co-transfection experiments, cells were previously transfected with control shRNA, shHIF-1α or shGPER using X-treme GENE 9 DNA transfection reagent (Roche Diagnostics, Milan, Italy). A mixture containing 0.5 µg of reporter plasmid and 10 ng of pRL-TK was then transfected by using X-treme GENE 9 DNA Transfection. After 8 hours, cells were treated for 18 hours with CuSO 4 in serum free medium. Luciferase activity was measured with the Dual Luciferase Kit (Promega, Milan, Italy) normalized to the internal transfection control provided by Renilla luciferase activity. The normalized relative light unit values obtained from cells treated with vehicle were set as 1-fold induction, upon which the activity induced by treatments was calculated.

Gene expression studies
Total RNA was extracted from cell cultures using the TRIzol commercial kit (Life Technologies, Milan, Italy) according to the manufacturer's protocol. RNA was quantified spectrophotometrically and quality was checked by electrophoresis through agarose gels stained with ethidium bromide. Only samples that were not degraded and showed clear 18 and 28 S bands under UV light were used for RT-PCR. Total cDNA was synthesized from the RNA by reverse transcription as previously described [17]. The expression of selected genes was quantified by real-time PCR using Step One (TM) sequence detection system (Applied Biosystems Inc, Milan, Italy), following the manufacturer's instructions. Gene-specific primers were designed using Primer Express version 2.0 software (Applied Biosystems. Inc., Milan, Italy) and are as follows: : HIF-1α Fwd:

Gene silencing experiments
Cells were plated onto 10-cm dishes and prior to treatments cells were transfected for 24 hours using X-treme GENE 9 DNA Transfection Reagent (Roche Diagnostics, Milan, Italy) with a control shRNA, shHIF-1α, shGPER, a control vector and the plasmid DN/c-fos, encoding a c-fos mutant that heterodimerizes with c-fos dimerization partners but not allowing DNA binding (kindly obtained from Dr. C. Vinson, NIH, Bethesda, MD, USA). The HIF-1α shRNA and the respective control plasmid were purchased from SABioscience Corporation (Frederick, MD, USA). The silencing of GPER expression was obtained by the construct which we have previously described and used [68].

Immunofluorescence assay
Fifty percent confluent cultured SkBr3 cells grown on coverslips were serum deprived and then treated for 12 hours with CuSO 4 alone and in combination with TEPA, NAC, AG1478 and PD98059, as indicated. Where required, cells previously transfected for 24 h with shHIF-1α or shGPER and respective negative control plasmids (as described above) and then treated for 18 hours with CuSO 4, Then cells were fixed in 4% paraformaldehyde, permeabilized with 0.2% Triton X-100, washed three times with PBS and incubated overnight with a mouse primary antibody against VEGF (C-1) (Santa Cruz Biotechnology, DBA, Milan, Italy). After incubation, the slides were extensively washed with PBS and incubated with 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI), (1:1000), (Sigma-Aldrich, Milan, Italy) and donkey anti-mouse IgG-FITC (1:300; purchased from Alexa Fluor, Life Technologies, Milan, Italy). The slides were imaged on the Cytation 3 Cell Imaging Multimode reader (BioTek, Winooski, VT) and analysed using the software Gen5 (BioTek, Winooski, VT).

Conditioned medium
SkBr3 cells were cultured in regular growth medium, then cells were washed twice with PBS and transfected for 24 hours in serum-free RPMI-1640 with shHIF-1α, shGPER or control shRNA using X-treme GENE 9 DNA Transfection Reagent (Roche Diagnostics, Milan, Italy). Cells were treated for 18 hours with CuSO 4 , culture medium was then replaced for additional 18 hours with medium without serum. Thereafter, the supernatants were collected, centrifuged at 3,500 rpm for 5 minutes to remove cell debris and used as conditioned medium in HUVECs. www.impactjournals.com/oncotarget

Tube formation assay
The day before the experiment, confluent HUVECs were starved overnight at 37 °C in serum free medium (EBM, Lonza, Milan, Italy). Growth factor-reduced Matrigel® (Cultrex, Trevigen Inc, USA) was thawed overnight at 4 °C on ice, plated on the bottom of prechilled 96well-plates and left at 37°C for 1 h for gelification. Starved HUVECs were collected by enzymatic detachment (0.25% trypsin-EDTA solution, Life Technologies, Milan, Italy), counted and resuspended in conditioned medium from CAFs. Then, 10,000 cells/well were seeded on Matrigel and incubated at 37 °C. Tube formation was observed starting from 2 h after cell seeding and quantified by using the software NIH ImageJ (National Institutes of Health (NIH), Rockville Pike, Bethesda, Maryland, USA) .

Migration assay
Twelve-well plates were coated with 500 μL fibronectin for 2 hours at 37°C (Sigma Aldrich, Milan, Italy). HUVECs were allowed to grow in regular growth medium until they reached a 70% to 80% confluence. Next, to create a scratch of the cell monolayer, a p200 pipette tip was used. Cells were washed twice with PBS and then incubated in medium collected from SkBr3 cells as previously described. The migration assay was evaluated after 24 hours of treatment.

MTT growth assay
For quantitative proliferation assay, cells (1 x 10 5 ) were seeded in 24-well plates in regular growth medium. Cells were washed once they had attached and then incubated in medium containing 2.5% charcoalstripped FBS with the indicated treatments; medium was renewed every day (with treatments) before dimethylthiazoldiphenyltetrazoliumbromide (MTT, Sigma-Aldrich, Milan, Italy) assay which was performed according to the manufacturer's protocol. A concentration of 250ng/L of the control shRNA, shHIF-1α or shGPER plasmids was transfected using X-treme GENE 9 DNA Transfection Reagent the day before treatments. The absorbance was measured using a FLX-800 microplate fluorimeter (Bio-Tek Instruments, Inc., Winooski, VT, USA) at a test wavelength of 570 nm. Each experiment was performed at in triplicate.

Statistical analysis
Statistical analysis was performed using ANOVA followed by Newman-Keuls' testing to determine differences in means. p < 0.05 was considered statistically significant.