Neuroglobin in Breast Cancer Cells: Effect of Hypoxia and Oxidative Stress on Protein Level, Localization, and Anti-Apoptotic Function

The over-expression of human neuroglobin (NGB), a heme-protein preferentially expressed in the brain, displays anti-apoptotic effects against hypoxic/ischemic and oxidative stresses enhancing neuron survival. As hypoxic and oxidative stress injury frequently occurs in fast proliferating neoplastic tissues, here, the effect of these stressors on the level, localization, and anti-apoptotic function of NGB in wild type and NGB-stable-silenced MCF-7 breast cancer cells has been assessed. The well-known endogenous NGB inducer 17β-estradiol (E2) has been used as positive control. The median pO2 present in tumor microenvironment of breast cancer patients (i.e., 2% O2) does not affect the NGB level in breast cancer cells, whereas hydrogen peroxide and lead(IV) acetate, which increase intracellular reactive oxygen species (ROS) level, enhance the NGB levels outside the mitochondria and still activate apoptosis. However, E2-induced NGB up-regulation in mitochondria completely reverse lead(IV) acetate-induced PARP cleavage. These results indicate that the NGB level could represent a marker of oxidative-stress in MCF-7 breast cancer cells; however, the NGB ability to respond to injuring stimuli by preventing apoptosis requires its re-allocation into the mitochondria. As a whole, present data might lead to a new direction in understanding NGB function in cancer opening new avenues for the therapeutic intervention.


Introduction
Neuroglobin (NGB) is a relatively recent discovered monomeric heme-protein so named because of its preferential expression in the nervous system [1]. NGB over-expression, driven by transiently transfected pcDNA vector, protects cultured neurons against hypoxia [2], enhances neuron survival under anoxia or oxygen/glucose deprivation [3], and displays neuro-anti-NGB, anti-Bcl2 antibodies and Annexin V-FITC Apoptosis Detection Kit were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). The chemiluminescence reagent for Western blot super power ECL was obtained from Bio-Rad (Milan, Italy). All the other products were from Sigma-Aldrich. Analytical or reagent grade products were used without further purification.
Preparation and purification of human recombinant NGB NGB cDNA was cloned into the pET3a vector (Novagen EMD Biosciences, Inc., Madison, WI, USA). The overexpression of NGB was induced in the Escherichia coli strain BL21(DE3)pLysS (Invitrogen, Carlsbad, California, USA) by treatment with 0.4 mM of isopropyl-D-thiogalactopyranoside (IPTG) in the presence of the heme-precursor aminolevulinic acid (1 mM). Soluble cell extract was loaded onto a DEAE-Sepharose Fast Flow (GE Healthcare Biosciences, Amersham Biosciences Ltd, UK) anion-exchange column equilibrated with 5 mM Tris-HCl, pH 8.5 and fractions were eluted with a NaCl gradient (from 0 to 300 mM). Eluted NGB was further purified by passage through a Sephacryl S-100 (GE Healthcare Biosciences, Amersham Biosciences Ltd, UK) gel filtration column. The protein obtained was > 98% pure on SDS-PAGE. The NGB concentration was determined spectrophotometrically, acquiring UV-visible spectra on a Cary 300 spectrophotometer (Varian, Palo Alto, CA). Five ng of recombinant NGB (final dilution: 1μg/1μl) were loaded in Western blot and the intensity of the bands was compared by densitometric analyses (see below). Note that, due to recombinant NGB purification, its migration on SDS PAGE resulted faster than that of NGB present in whole cell lysates.
Hypoxic treatment MCF-7 cell lines were grown to 70% confluence in 6-well plates and stimulated with either vehicle or E2 (10 nM). After 2h of stimulation, cells were cultured in normoxia using an incubator (KW Apparecchi Scientifici, Siena, Italy) set at 5% CO 2 , 21% O 2 (atmospheric oxygeñ 140 mmHg), and 37.0°C in a humidified environment. For the experiments under hypoxia, a water-jacketed incubator (Forma Scientific, Marietta, OH, USA) has been used to provide a customized and stable humidified environment through electronic control of CO 2 (5%), O 2 , and temperature (37.0°C). The O 2 tension was set and maintained constantly at 2% (~14mmHg) by injecting N 2 automatically in the chamber.

Western blot assay
Protein extraction and Western blot assay were performed as reported elsewhere [9]. Briefly, after treatment, cells were lysed and solubilized in the sample buffer containing 0.125 M Tris-HCl, pH 6.8, and 10% (w/v) SDS [9]. Total proteins were quantified using the Bradford Protein Assay. Solubilized proteins (20 μg) were resolved by 7% or 15% SDS-PAGE at 100 V for 1 h at 24.0°C and then transferred to nitrocellulose with the Trans-Blot Turbo Transfer System (Bio-Rad, Hercules, CA) for 7 min. The nitrocellulose was treated with 3% (w/v) BSA in 138.0 mM NaCl, 25.0 mM Tris, pH 8.0, at 24.0°C for 1 h and then probed overnight at 4.0°C with either anti-NGB (final dilution 1:1000), anti-PARP-1 (final dilution 1:1000), and anti-Hypoxiainducible factor-1α (HIF1α) (final dilution 1:1000) antibodies. The nitrocellulose was stripped by the Restore Western Blot Stripping Buffer (Pierce Chemical, Rockford, IL, USA), for 10 min at room temperature, and then probed with anti-β-tubulin antibody (final dilution 1:1000) to normalize protein loaded. The antibody reaction was visualized with the chemiluminescence Western blotting detection reagent (Amersham Biosciences, Little Chalfont, UK). The densitometric analyses were performed by ImageJ software for Microsoft Windows (National Institutes of Health, Bethesda, MD, USA).

Intracellular ROS measurement
Cells were seeded in clear bottom 96-well microplate with 2.5×10 4 cells per well. After allowing cells to adhere overnight, the medium was removed and cells washed once with serumfree medium. Then, cells were incubated with DCFH-DA (20 μM) at 37°C, 30 min in the dark. After this time, the DCFH-DA solution was removed and cells washed once with serum-free medium and treated with selected compounds; background wells (untreated stained cells) as well as blank wells (medium only) were included. The microplates were read in the presence of compounds and media on a multi-label plate reader (VICTOR™ X3 Multilabel Plate Reader, PerkinElmer, Waltham, MA, USA) with excitation wavelength at 485 nm and emission wavelength at 535 nm to measure fluorescence intensity for each time interval (from 0 to 6 h). The fluorescence was registered as arbitrary units, the ratio between the single treatment induced fluorescence, and the vehicle fluorescence was plotted for each time considered.

Stress and Apoptosis Signaling Measurement
For the simultaneous detection of 19 signaling molecules that are involved in the regulation of the stress response and apoptosis, the PathScan1 Stress and Apoptosis Signaling Antibody Array Kit has been used according to the manufacturer's instructions (Cell Signaling Technology, Danvers, MA, USA). Briefly, MCF-7 cells were grown until 80% confluence, treated with the selected compounds, and lysed in 1X Cell Lysis Buffer to collect cell lysates. The arrayblocking buffer was added to each well for 15 min at room temperature. Then, 30 μg of solubilized proteins were added to wells and incubated for 2 h at room temperature. Subsequently, the Detection Antibody Cocktail supplied with the kit was added and maintained for 1 h at room temperature. The slide was then incubated for 30 min with horseradish peroxidaselinked streptavidin solution at room temperature. Finally, the slide was covered with Lumi-GLO/Peroxide reagent (supplied with the kit) and exposed to chemiluminescence film (Amersham Biosciences, Little Chalfont, UK) for 2 to 60 sec. The images were then acquired and the signal intensity was measured using the ImageJ software for Microsoft Windows (National Institutes of Health, Bethesda, MD, USA).

Quantitative Real-Time Polymerase Chain Reaction
The sequences for gene-specific forward and reverse primers were designed using the Oligo-Perfect Designer software program (Invitrogen). The following primers were used: for human NGB 5'-GTCTCTCCTCGCCTGAGTTC-3'(forward) and 5'-GACTCACCCACTGTCGAG AA -3' (reverse) and for human GAPDH, 5'-CGAGATCCCTCCAAAATCAA-3' (forward) and 5'-TGTGGTCATGAGTCCTTCCA-3' (reverse). Total RNA was extracted from cells using TRIzol Reagent (Invitrogen) according to the manufacturer's instructions. To determine NGB gene expression levels, cDNA synthesis and qPCR were performed using the GoTaq two-step RT-qPCR system (Promega) in an ABI Prism 7900HT Sequence Detection System (Applied Biosystems, Foster City, CA) according to the manufacturer's instructions. Each sample was tested in triplicate and the experiment repeated twice. All primers used were optimized for real-time amplification in a standard curve amplification (>98% for each pair of primers) and verifying the production of a single amplicon in a melting curve assay. Results were normalized to the expression of GAPDH mRNA. The relative level for NGB gene, reported in arbitrary units, was calculated using the 2-ΔΔCt method.

Apoptosis measurement
Phosphatidylserine externalization was quantified by flow cytometry by using the Annexin V-FITC Apoptosis Detection Kit including propidium iodide (PI) according to the manufacturer's guideline (Santa Cruz, CA, USA). Briefly, both attached and floating cells were collected after treatment(s), washed twice with cold PBS and re-suspended in the annexin-binding buffer at a concentration of~1×10 6 cells/ml; 100 μl of the cell suspension (~1×10 5 cells) were transferred to a culture tube and 2.5 μl of annexin V-FITC and 10 μl of PI were added. After incubation in the dark (15 min at room temperature), 400 μl of the binding buffer were added and cells were analyzed immediately by flow cytometry with the DAKO Galaxy flow-cytometer equipped with HBO mercury lamp. Analysis by flow cytometry used the FL1 (FITC) and FL3 (PI) laser lines; each sample was assessed using a collection of 10,000 events. Each experiment was carried out in triplicate and the fluorescence was calculated using a FloMax © Software.

Mitochondria isolation
Cell fractionation was performed using ApoAlert™ Cell Fractionation kit (Clontech Laboratories Inc. Mountain View, CA, USA) according to manufacturer's instructions. After stimulation, cells were harvested with trypsin (1%, v/v), suspended with complete medium, and centrifuged at 600g for 5 min. Pellet was suspended in Fractionation Buffer Mix containing DTT 1 mM and homogenized in a Dounce tissue grinder. Homogenate was centrifuged at 700g for 10 min. Pellet was suspended in Fractionation Buffer Mix to obtain mitochondrial fraction. The mitochondrial TNF-receptor associated protein 1 (TRAP-1) and cytosolic Protein Phosphatase 2A (PP2A) markers were used as mitochondrial fraction purity indicators. Protein concentration of each fraction was determined using Bradford protein assay. Lysate of each fraction was then processed for Western Blot or used for immunoprecipitation assay.

Statistical analysis
The statistical analysis was performed by Student's t-test with the INSTAT software system for Windows. In all cases, only probability (p) values below 0.05 were considered significant.

Effect of hypoxia on NGB levels
Neither 24 h (Fig 1) nor 48 h (data not shown) of physiological hypoxia (2% O 2 ) increases NGB protein levels in comparison to normoxia (21% O 2 ) in breast cancer cells. However, E2 treatment (10 nM, 24 h) still induces the up-regulation of NGB and the hypoxia sensor HIF1α (Fig 1).

Effect of ROS-inducing compounds on NGB levels
To determine the role of NGB during oxidative stress injury, MCF-7 cells have been treated with H 2 O 2 and Pb(IV), (a pollutant which induces oxidative stress and mitochondrial-dependent apoptosis) [25]. E2 treatment has been used as the positive control. Neither vehicle nor E2 (10 nM) enhance ROS level in MCF7 cells, whereas cell treatment with either H 2 O 2 (400 μM) or Pb(IV) (200 μM) increased the ROS production reaching maximum levels after 30 min of treatment (14.68 ± 0.04 and 11.00 ± 1.49 fold over the control, respectively) (Fig 2A). Moreover, the E2-, H 2 O 2 -, and Pb(IV)-activated signaling pathways involved in cell response to stress and apoptosis has been evaluated by the PathScan array kit (S1 Fig). As shown in Fig 2B, E2 activates the phosphorylation of AKT, SMAD2, and SAPK/JNK in line with its well-known function as activator of MCF-7 survival, proliferation, and migration [26]. On the other hand, H 2 O 2 enhances PARP-1 cleavage and AKT phosphorylation, impairing SAPK/JNK phosphorylation and decreasing survivin levels; while, Pb(IV) increases the phosphorylation of SMAD2 and SAPK/JNK as well as the PARP-1 cleavage and survivin level (Fig 2B). Finally, the capability of E2, H 2 O 2 , and Pb(IV) to modify the level of NGB mRNA ( Fig 2C) and protein (Fig 2D,  2F and 2F') has been evaluated. E2 induces the increase of NGB mRNA 4 h after treatment ( Fig  2C), whereas neither H 2 O 2 nor Pb(IV) modulate NGB mRNA levels ( Fig 2C). Conversely, like E2, both H 2 O 2 and Pb(IV) increase NGB protein levels ( Fig 2D). To quantify the results obtained by Western blot, the intensity of the NGB bands was compared with that obtained loading 5 ng of recombinant NGB. MCF-7 cells contain a very low basal level of NGB (30 ± 3.3 ng/mg protein lysate) which significantly doubles 24 h after E2 (60 ± 3.2 ng/mg protein lysate), H 2 O 2 (48 ± 2.2 ng/mg protein lysate), and Pb(IV) (46 ± 2.1 ng/mg protein lysate) treatment ( Fig 2E). In order to obtain clear evidence how NGB level could be regulated by H 2 O 2 and Pb (IV), MCF-7 cells were treated with either the proteasomal inhibitor, MG-132 (1 μM for 30 min), the lysosomal inhibitor, Chloroquine (Chloroq, 10 μM for 30 min), and the translational inhibitor Cicloheximide (Ciclohex, 10 μM for 30 min) before the treatment with the ROSinducers. Fig 2F and 2F' show that NGB level is reduced by ciclohex and increased by lysosomal degradation. H 2 O 2 and Pb(IV) treatments do not modify this trend suggesting that ROS-inducing compounds could increase NGB levels by inhibiting lysosomal degradation and increasing NGB translation. Intriguingly, MG-132 does not modulate NGB level, but completely impairs H 2 O 2 and Pb(IV) effect in enhancing NGB level (Fig 2F and 2F'). As a whole, these data indicate that ROS-inducing compounds increase NGB protein levels in MCF-7 cancer cells activating specific and diverse pathways.

Effect of H 2 O 2 and Pb(IV) on mitochondrial NGB localization
The NGB localization into the mitochondrial compartment is necessary to act as an anti-apoptotic protein in several cell lines [9,23,24]. This prompted us to verify if the selected compounds could modify NGB mitochondrial localization. Confocal microscopy analyses show the co-localization of NGB with the mitochondrial marker COX-4 (cytochrome c oxidase-4) ( Fig  3A). Although the confocal microscopy allows a purely qualitative analysis, the analysis with the 8.2 IMARIS software demonstrated that only E2 treatment (24h) significantly raises the NGB-COX-4 merged signals (Fig 3A). The increase of the NGB localization at mitochondrial level has been confirmed in isolated mitochondria by using cell fractionation kit. Fig 3B confirms the purity of mitochondrial fraction, in fact PP2A, cytosolic marker is absent; while TRAP-1, mitochondrial marker, is evident. As expected [23], E2 (10 nM; 24 h) increases the mitochondrial NGB content; conversely, H 2 O 2 treatment (400 μM; 24h) does not modify the NGB mitochondrial localization (Fig 3C). Intriguingly, this more sensitive method demonstrates that Pb(IV) (200 μM, 24 h) treatment determines a significant decrease of the NGB amount in the mitochondrial fraction (Fig 3C).

Effect of H 2 O 2 and Pb(IV) on the NGB anti-apoptotic function
Recently, it has been demonstrated that NGB is an E2 compensatory protein, which up-regulation counteracts apoptosis induced by oxidative stress in several cancer cell lines [23,24]. This  Fig 4B). Surprisingly, although both H 2 O 2 and Pb(IV) increase NGB level, 24 h after treatment they activate the PARP-1 cleavage, another pro-apoptotic marker, both in the presence and in the absence of NGB (Fig 4C and 4D). This unexpected result lead us to evaluate if E2-induced NGB over-expression in mitochondria could protect MCF-7 cells from Pb(IV)-induced apoptosis. To reach this aim, MCF-7 cells were stimulated with E2 for 24 h before the Pb(IV) treatment. Fig 4E clearly indicate that E2 pre-treatment reduces Pb(IV)-induced PARP cleavage in ScNGB MCF-7, while this effect is completely impaired in ShNGB MCF-7 (Fig 4F).

Discussion
Here, we investigated the putative role of endogenous level of NGB in breast cancer cells as a stress sensor and as a compensatory protein, which responds to the injuring stimuli inhibiting the trigger of mitochondria-dependent apoptosis. For the first time, we showed that endogenous NGB is a ROS-inducible protein in MCF-7 cells.
Hypoxia is a common feature of solid tumors and the involvement of NGB in the shortterm adaptation of cancer cells has been hypothesized [21]. In fact, NGB co-localizes with the hypoxia-inducible metallo-enzyme carbonic anhydrase IX in different human primary tumor specimens [21]. On the other hand, the hypoxia-dependent up-regulation of NGB mRNA has been assessed in lung cancer cells even if no information is available on the protein level [22]. Although NGB is not transcriptionally regulated by HIF1α the major intracellular oxygen sensor [27], this globin has been proposed to be a member of the hypoxia-inducible protein family [4]. Despite this evidence, our results indicate that 2% O 2 , which resembles the median pO 2 present in breast cancer microenvironment [28], does not up-regulate NGB levels in MCF-7 cells suggesting that NGB is not required for MCF-7 cell adaptation to hypoxic conditions. Of note, myoglobin, which is strongly up-regulated by hypoxia in MCF-7 cells [29], may attend to this function thus suggesting a cell context dependent modulation of NGB from hypoxia As well hypoxia, oxidative stress is characteristic of both tumor development and cancer cells resistance to antitumor drugs. High ROS levels, as occur in fast proliferating tumor tissues [19], could lead to severe cellular damage and, consequently, to cell death. However, cancer cells established several mechanisms to counteract the oxidative stress-induced apoptosis and, generally, display an antioxidant capacity higher than that of normal cells [19]. Different intracellular pathways could converge to alter the cellular metabolism and to adapt cancer cells to both intrinsic and extrinsic oxidative stress conditions [18]. However, it remains unsolved the question about the possible role of endogenous NGB in non-nervous cancer cells as an oxidative stress sensor; this function requires NGB activation or induction by stressing conditions. Although the PathScan assay and NGB mRNA did not define a unique common pathway linking E2 and stressor-inducing NGB up-regulation, the results reported here clearly demonstrate that in MCF-7 cancer cells the apoptotic inducers modulate the level of NGB. Indeed, cell treatment with H 2 O 2 and Pb(IV) leads to a rapid increase of intracellular ROS production, and upregulate NGB protein levels. H 2 O 2 and Pb(IV) effect on NGB level seems to be mediated by the inhibition of NGB lysosomal degradation and by the activation of translation as demonstrated by cell pre-treatment with Chloroquine and Cicloheximide. Contrarily, MG-132 does not modulate NGB level, but completely impairs H 2 O 2 and Pb(IV) effect in enhancing NGB level. Recently, a role for MG-132 and Chloroquine in the activation and inhibition, respectively, of autophagy in breast cancer cells has been reported [30,31] rendering particularly intriguing these results. Indeed, our results seem suggest that the autophagic process is involved in H 2 O 2and Pb(IV)-NGB accumulation breast cancer cells. Finally, the involvement of AKT, SMAD2, SAPK/JNK, survivin, and protein stability, evidenced in this study, could converge in ROSinducing pathways (e.g., Nuclear factor erythroid-derived 2, NRF2) to increase NGB levels and, ultimately, to cancer cell survival. Although studies more detailed are requested to define the ROS-induced pathways and their functional outcomes, the data reported in this study enlarge the physiological role of NGB in breast cancer cells pointing to its up-regulation as possible ROS sensor as reported in brain-derived cells [2-4, 6, 11, 17]. In particular, in human and rat pheochromocytoma PC12 cell lines, the H 2 O 2 treatment increases the level of transiently transfected NGB and induces a conformational change of NGB allowing the globin recruitment at the plasma membrane lipid rafts where it acts as a guanine-dinucleotide dissociation inhibitor suppressing the Gαs activity and leading to neurons protection against apoptosis [17]. However, data reported here unexpectedly indicate that the increased level of NGB induced by H 2 O 2 and Pb(IV) is not sufficient to counteract the ability of these substances to induce the apoptotic death in MCF-7. This result has been further confirmed by NGB silencing experiments in which both H 2 O 2 and Pb(IV) still activate the PARP-1 cleavage, an apoptotic marker in MCF-7 cells. However, increasing NGB levels by 24 h E2 treatment reduces Pb(IV) activation of PARP-1 cleavage. This result is in line with the E2 protective effect against H 2 O 2induced apoptosis previously reported [23][24] strongly confirming the anti-apoptotic role of NGB.
Remarkably, E2-induced NGB up-regulation exerts anti-apoptotic function directly at the mitochondrial compartment by interacting with cytochrome c and impairing its release to cytosol and the consequent activation of the intrinsic apoptotic pathway upon oxidative stress injury [7,23]. Present data indicate that in MCF-7 cells, only the E2 treatment almost doubles the NGB amount in the mitochondrial fraction, whereas H 2 O 2 and Pb(IV), which increase NGB level in the whole cell, do not affect the mitochondrial protein amount. All together, these data indicate that the increase of intracellular NGB levels induced by H 2 O 2 and Pb(IV) is not sufficient to reset the intrinsic apoptotic pathway, which requires the re-allocation of NGB into mitochondria. In line with this idea, Yu and coworkers [6] demonstrated that in primary mouse cortical neurons NGB mitochondrial localization increased after pathological oxygenglucose deprivation conditions conferring neuroprotection. Although it could be possible that only the high NGB levels obtained by transfection with NGB-encoding plasmid allow the NGB recruitment on lipid rafts reported in neurons after oxidative stress [17], the possibility that E2, oxygen-glucose deprivation, and oxidative stress signaling could induce different NGB conformation that affect the protein translocation into mitochondria should be taken into account.
An opposite function has been attributed to NGB in hepatocarcinoma cells. In this cell line, NGB acts as a mediator between oxygen/ROS signals and the cytosolic signaling cascade that regulates the cell proliferation [20]. Unfortunately, also this evidence has been obtained in cells transiently transfected with the NGB-encoding plasmid which induces high non-physiological protein levels that could be necessary for NGB involvement in intracellular signaling cascade [20]. However, the involvement of NGB in cancer cell proliferation has not been confirmed by MCF-7 cell growth curves, which result similar to the control even when NGB was stably silenced [23]. These results indicate that the over-expression of NGB does not fully represent the physiological behavior of NGB.
In conclusion, NGB level could be considered as a sensor of ROS being up-regulated by ROS (i.e., H 2 O 2 ) and by ROS-inducing substances (i.e., Pb(IV)), whereas its function as an anti-apoptotic protein is strictly linked to its level and intracellular localization. However, oxidative conditions increase the heme-Fe-based NGB reactivity by formation of the labile Cys46-Cys55 disulfide bond and of the Tyr44/His64/heme propionate interaction [32,33]. Consequently, ROS and ROS-generating compounds induce high level of oxidized NGB that increase, for example, NGB activity as free radical scavenger [34] supporting the role of NGB as a compensatory protein in breast cancer cells. Although further studies in breast cancer cells are needed to identify up-stream and down-stream NGB regulating pathways and the intracellular NGB trafficking, present data might lead to a new direction in understanding NGB function cancer and neuroprotection opening new avenues for the therapeutic intervention based on the development of inhibitors impairing NGB-dependent cell protection.