Copper Uptake in Mammary Epithelial Cells Activates Cyclins and Triggers Antioxidant Response

The toxicologic effects of copper (Cu) on tumor cells have been studied during the past decades, and it is suggested that Cu ion may trigger antiproliferative effects in vitro. However, in normal cells the toxicologic effects of high exposures of free Cu are not well understood. In this work, Cu uptake, the expression of genes associated with cell cycle regulation, and the levels of ROS production and related oxidative processes were evaluated in Cu-treated mammary epithelial MCF10A nontumoral cells. We have shown that the Cu additive is associated with the activation of cyclin D1 and cyclin B1, as well as cyclin-dependent kinase 2 (CDK2). These nontumor cells respond to Cu-induced changes in the oxidative balance by increase of the levels of reduced intracellular glutathione (GSH), decrease of reactive oxygen species (ROS) generation, and accumulation during progression of the cell cycle, thus preventing the cell abnormal proliferation or death. Taken together, our findings revealed an effect that contributes to prevent a possible damage of normal cells exposed to chemotherapeutic effects of drugs containing the Cu ion.


Introduction
Recent advances in biochemical tools have highlighted the extraordinary array of functions of Cu in living organisms [1]. Cu is required for binding to the dual specificity mitogen-activated protein kinase kinase 1 MEK1 and promotion of mitogen-activated protein kinase MAPK signaling and tumorigenesis by v-raf murine sarcoma viral oncogene homolog B BRAF in mammary tumors [2]. However, despite the enormous expansion in our knowledge of Cu biology that has occurred over the last decades, we are only just beginning to unravel the complexity of the role of this transition metal in the regulation of cellular processes and cell cycle.
Cu is an essential trace element and its intracellular concentrations are tightly controlled. Within the cell, Cu is distributed by metallochaperones and plays a fundamental role in regulating cell growth, altering gene expression (due to oxidation of guanosine and adenosine residues in nucleic acids or changes in transcription factor/growth factor activities) [3]. A critical factor in the development of cancer is angiogenesis, which endows continuous supplying of nutrients, growth factors, and signaling agents to malignant tissue [4][5][6][7]. This angiogenic response in tumor is stimulated by ceruloplasmin, the plasma Cu-carrier [6,8,9]. Although these studies with cancer cells and tumors strongly suggested that Cu plays an essential role in cell growth and proliferation, little is known about underlying molecular mechanisms. Also, Cu is involved in redox reactions that generate intracellular reactive oxygen species (ROS), mainly by Fenton reaction, and a number of reports point to a relationship between Cu, ROS production, and cancer development [10,11], and recently the role of Cu metabolism in resistance of cancer cells to cisplatin [12][13][14]. The redox status of cells is influenced by the homeostasis of reactive species, since ROS might act as secondary messengers in the regulation of pathways associated with cell proliferation, differentiation, and apoptosis [15,16]. Based on these findings, some studies suggested that elevated Cu levels and increased oxidative stress may be used in selective cancer therapy [17,18]; however, the effect of Cu-stimulation in cell proliferation 2 Oxidative Medicine and Cellular Longevity and its relationship with ROS needs to be well elucidated, especially in nontumoral cells.
The aim of the present study was to clarify the connection of the Cu with cell cycle activation in normal epithelial cells and to determine the mechanism by which this ion, supplied as CuSO 4 , stimulates the cell cycle of breast epithelial cells in vitro. For this purpose, the Cu uptake, the expression of genes associated with cell cycle regulation, and the levels of ROS production and related oxidative processes were evaluated in Cu-treated mammary epithelial MCF10A nontumoral cells.

Chemicals.
Unless otherwise stated, chemicals were obtained from Sigma-Aldrich (USA) and were of analytical grade: solutions were prepared using Milli-Q water (Millipore, Billerica, USA).

Cell Proliferation
Assays. MCF-10A cells were incubated in 24-well plates at a density of 4 × 10 4 cell/cm 2 for the period of 24 h under the conditions described above. Aliquots of freshly prepared solutions of CuSO 4 were added separately to the culture medium (less than 1% of total volume) in order to attain final concentrations in the range 25.0-1000 M, and the plates were incubated for further 24-48 h. On the basis of the results obtained subsequent experiments were conducted by incubating treated cells to CuSO 4 at final concentrations of 50 M ( = 5) and control cells on unsupplemented medium. Typically, cells were plated onto the medium at a density of 4 × 10 4 cells/cm 2 to give monolayers of approximately 50-60% cell confluence and incubated for 48 h. Following incubation, cells were trypsinized, washed with phosphate buffered saline (PBS: 137 mM NaCl and 2.7 mM KCl in 10 mM phosphate buffer at pH 7.4), stained with Trypan Blue (T8154, Sigma Aldrich, St. Louis, USA), and counted under an optical microscope using a Neubauer's chamber [19].

Analysis of PCR Data.
The relative levels of expression of target genes were evaluated using the comparative cycle threshold method as described by Medhurst et al. [20]. A value for was determined from the fluorescence detected within the geometric region of the semilog amplification plot and represented the PCR cycle number at which fluorescence was detectable above an arbitrary threshold established on the basis of the variability of baseline data during the first 15 cycles.

Solid Sampling in Graphite Furnace
Atomic Absorption Spectroscopy (GFAAS). Experimental parameters were obtained from Carvalho Do Lago et al. [21] and the new developed methodology with dried cells [20]. Briefly, a model ZEEnit 600 (Analytik Jena) atomic absorption spectrometer, equipped with a transversely heated graphite atomizer, an inverse and transversal 2-and 3-field mode Zeeman effect background corrector, manual sampling accessory, and hollow Cu cathode lamp, was employed to determine intracellular Cu concentrations. Pyrolytically coated heated graphite tubes and pyrolytically coated boat-type solid sampling platforms were employed throughout. Argon (99.998% v/v; Air Liquide, Mauá, Brasil) was used as protective and purge gas. All measurements were based on integrated absorbance values acquired with the aid of Windows NT software.

Determination of Cu Content of Cultured Cells.
MCF10A cells that had been plated and incubated in the presence or absence of CuSO 4 (50.0 M) as described above were trypsinized and adherent cells were combined, washed five times with phosphate buffered saline (PBS: 137 mM NaCl and 2.7 mM KCl in 10 mM phosphate buffer at pH 7.4) containing 1.0 mM EDTA in order to remove residual Cu(II), and dried for 1 week in a desiccator. Experiments were conducted in triplicate or quintuplicate using plates of surface area 75 or 150 cm 2 . For the GFAAS determination of Cu, procedures were followed as from Carvalho Do Lago et al. [21].

Extraction of Nuclei.
Nuclei from MCF10A cells that had been incubated in the presence or absence of CuSO 4 (50.0 M) as described above were isolated using published procedures [22,23]. Briefly, cells were inoculated at a density of 4 × 10 4 cell/cm 2 into culture bottles containing appropriate medium (150 cm 2 surface area of culture) and incubated for 48 hours at 37 ∘ C in an atmosphere of 5% CO 2 in air at a relative humidity of 80%. Experiments were conducted in quadruplicate. Following incubation, cells were trypsinized and adherent cells were combined, washed with PBS, and centrifuged (250-300 ×g, 10 minutes, 4 ∘ C) and the pellet was resuspended in 2 mL of lysis buffer (10 mM NaCl, 3 mM MgCl 2 , and 0.5% Tergitol NP-40 in 10 mM Tris buffer at pH 7.5) and left on ice for 5 minutes. Cells were subsequently centrifuged (500 ×g, 5 minutes, 4 ∘ C) and the pellet was resuspended in 2 mL of lysis buffer and recentrifuged. The pellet from the second centrifugation (containing extracted nuclei) was dried in an oven at 30 ∘ C and subsequently analyzed by GFAAS. The purities of the extracted nuclei were determined by Western blot analyses using rabbit antihistone H3 Nterminal and rabbit anti-human Cu/Zn SOD1 polyclonal antibodies.

Generation of Intracellular Reactive Oxygen Species.
MCF10A cells that had been plated and incubated in the presence or absence of CuSO 4 (50.0 M) were treated with trypsin/EDTA (1 mM, 25200-056, Gibco, Waltham, USA) solution, washed three times with PBS, and suspended in a 50.0 M solution of the oxidation-sensitive nonfluorescent probe 2 ,7 -dichlorodihydrofluorescein diacetate (DCFH, D6883, Sigma Aldrich, St. Louis, USA) [24,25]. Following incubation at 37 ∘ C for 45 minutes [19], the cells were washed three times with PBS and the levels of intracellular fluorescence were determined immediately by flow cytometry at 530 nm using a Cytomics FC 500 MPL (Beckman Coulter) instrument [26,27]. Assays were conducted at least in quintuplicate and >20,000 viable cells from each sample were analyzed per assay, in arbitrary units of fluorescence.

Determination of GSH/GSSG Ratio.
MCF10A cells that had been plated and incubated in the presence or absence of CuSO 4 (50.0 M) as described above were trypsinized and adherent cells were combined, washed three times with cold PBS, and centrifuged (1500 ×g, 3 minutes, 4 ∘ C). The cells were resuspended in 400 L of cold water and lysed by freezing in a mixture of dry ice and ethanol for 10 minutes. After thawing, proteins were precipitated with 100 L of 10% sulfosalicylic acid (390275, Sigma Aldrich, St. Louis, USA) and centrifuged (4000 ×g, 5 minutes, 4 ∘ C). The protein concentration in the pellet was determined, and the levels of GSH and GSSG in the supernatant were assessed using the protocol of Martín et al. [28]. For GSH quantification, the assay mixture contained 100 L of supernatant, 790 L of a 0.1 M sodium phosphate buffer containing 0.05% EDTA at pH 7.0, 100 L of 6 mM 5,5 -dithiobis (2-nitrobenzoic acid) (DTNB, D8130, Sigma Aldrich, St. Louis, USA) dissolved in glutathione assay buffer (GAB; 125 mM sodium phosphate containing 6.3 mM EDTA), and 10 L glutathione reductase (55 U/mL, G3664, Sigma Aldrich, St. Louis, USA). For GSSG quantification, the assay mixture comprised 100 L of supernatant, 190 L of 0.5 M phosphate buffer at pH 6.8, 700 L of 0.3 mM NADPH (N5130, Sigma Aldrich, St. Louis, USA), prepared in GAB, and 10 L of glutathione reductase (55 U/mL). The reaction rate was estimated from the change in absorbance at 412 nm after 3 minutes at 25 ∘ C (for GSH) or at 340 nm after 16 minutes at 30 ∘ C (for GSSG) [28][29][30]. The accuracy of the GSH reference standard was measured with DTNB (D8130, Sigma Aldrich, St. Louis, USA), using a molar extinction coefficient of 13,600 with an absorbance of 412 nm [31]. GSSG was standardized by measuring the decline of NADPH in the presence of glutathione reductase, taking into consideration that the molar extinction coefficient of NADPH to be 6270 at 340 nm and 1 mol of NADPH converts 1 mol of GSSG to 2 mol of GSH [32]. The specific protein complexes formed following treatment with specific secondary antibody (anti-mouse or anti-rabbit IgG-peroxidase conjugate, A4416 or A0545, Sigma Aldrich, St. Louis, USA) were detected using SuperSignal West Pico chemiluminescent substrate (34080, Thermo Scientific, Waltham, USA).

Statistical Analyses.
All experiments were repeated at least five times (except where stated otherwise) and the results are expressed as mean values ± standard deviations. Analysis of variance (ANOVA) with the Bonferroni correction was used to evaluate the differences between means with the level of significance set at < 0.05. For real-time PCR experiments, values obtained for each cell lineage were entered into a two-way ANOVA with factors time and treatment, and pairwise comparisons were performed using the Tukey HSD test with the level of significance set at < 0.05.

Viability Test and Measurement of Intracellular Copper
Levels. The effects of Cu on the viabilities of the nontumor line MCF10A were initially evaluated using the Trypan Blue exclusion test. On the basis of concentration-dependent studies (Figure 1), proliferation of MCF10A was not observed at concentrations above 75.0 M CuSO 4 after 24 hr of incubation (Figure 1(a)), but cell viability was significantly reduced after 48 hr (Figure 1(b)) when levels of Cu were equal to or greater than 200.0 M. We did not observe differences in cell proliferation or cell death when comparing untreated and treated cells after 24 hours, 10.83 ± 0.14 × 10 5 viable cells/mL versus 11.91 ± 0.76 ( = 7.32 × 10 −2 ) for 25 M Cu, or versus 11.50 ± 0.43 ( = 6.46 × 10 −2 ) for 50 M Cu, or  (Figure 1(b)). Due to the nonproliferative effect on MCF10A, the concentration of 50.0 M CuSO 4 was chosen in whole study.

Real-Time PCR Analyses of Cyclins D1 and B1 in Epithelial
Cell Line MCF10A. In order to assess the mechanism by which Cu(II) salt stimulates cell proliferation in human epithelial cells, we investigated the regulation of genes involved in cell cycle progression. Specific primers were designed and their reliability verified on the basis of amplification plots obtained with serially diluted cDNA (1, 1/3, 1/9, and 1/27), linear regression analyses, and dissociation melting curves. With these procedures, we validated specific primers designed for human cyclin D1 (Figures 3(a)-3(c)) and cyclin B1 (Figures 3(d)-3(f)) as well as for GAPDH (Figures 3(g) 48 hours. Western blots obtained using mouse anticyclin dependent kinase 2 (anti-CDK2) and mouse anti-tubulin (internal standard) as primary antibodies revealed that CDK2 expression levels were upregulated in MCF10A cells in the presence of Cu after 48 hr (Figure 4).

Quantification of the Intracellular Glutathione Levels.
In order to clarify the effects of Cu entrance in MCF10A cells, the GSH/GSSG ratios in cells that had been exposed to Cu(II) for 48 hours were determined as suggested by Estrela et al. [38]. We measured the total glutathione level in cell to ensure the level of this endogenous antioxidant was not changed during the Cu treatment. Control and Cutreated cells exhibited glutathione total level of 7.35 ± 0.76 and 8.19 ± 0.21 ( = 0.1375), respectively, indicating no significant changes in the total glutathione (GSH + GSSG, Figure 6(a)). Cells of MCF10A that had been incubated on control medium for 48 hours exhibited GSH/GSSG ratio of 3.20 ± 0.05 (Figure 6(b)). Inclusion of CuSO 4 in the medium increased to 5.72 ± 0.86 ( = 0.00896) the GSH/GSSG ratio.

Discussion
Cu complexes can induce apoptosis in cancer cells as a result of damage inflicted to the organelle [19,39,40], and this process could engender misinterpretation of the actual effect of the free metal excess on the cell cycle of normal cells. In the present study, the culture medium was supplemented with the free salt CuSO 4 in order to investigate the influence of free intracellular Cu on the proliferation of normal epithelial MCF10A cells in vitro. The specific choice of the free Cu ion and its concentration was based on results recently reported by Carvalho Do Lago et al. [21]. Also, malignant cells typically possess increased levels of oxidant species that contribute naturally to the enhancement of apoptosis [41], while the lower oxidant levels of the nonmalignant MCF10A should contribute to their resistance to Cu-stimulated cell death.
Progression through the different phases of the cell cycle is regulated by specific combinations of cyclins and CDKs. Cyclins and CDKs can be involved in processes other than that of cell proliferation, as demonstrated by the reported association between cyclin expression and cell cycle reentry leading to apoptosis in neurodegenerative processes triggered by oxidative stress [42].  Figure 5: Cu-treated MCF10A cells produced less reactive oxygen species (ROS) than their untreated counterparts. The generation of intracellular ROS, cells that had been exposed to Cu(II), was estimated using the membrane-permeable nonfluorescent cell probe DCFH.  In the present study, normal MCF10A cells showed a mild upregulation of cyclins D1 and B1 on exposure to Cu. Interestingly, we observed that cyclin B1 was upregulated when the internal concentration of Cu exceeded 100 g/g cells, while cyclin D1 was upregulated when Cu concentrations were within the range 150−275 g/g cells. MCF10A cells permitted the entry of Cu such that the internal concentration of the metal was sufficiently elevated to cause specific upregulation of cyclins mRNA levels. Indeed, Cu was detected in the nuclei of MCF10A cells, indicating that this ion may trigger changes in gene expression, including those related to cell cycle progression. Malignant cells typically possess increased levels of oxidant species that contribute naturally to the enhancement of proliferation [41], while the lower oxidant levels of the nonmalignant MCF10A should contribute to their resistance to Cu-stimulated cell proliferation as we observed.
The concept that redox cycling controls the mammalian cell cycle through the modulation of intracellular antioxidant/oxidant species, particularly thiol-containing molecules such as GSH, has received much consideration in the literature (for a review [15]). In cancer cells, the GSH/GSSG ratio has been shown to influence the regulation of the cell cycle, mutagenic mechanisms, DNA synthesis, growth, and multidrug and radiation resistance, and GSH levels are typically higher in tumor tissue in comparison with normal tissue [38,43]. In the present study, we observed that internalization of Cu, induced by treatment with CuSO 4 , decreased ROS levels and increased the GSH/GSSG ratio. Cu resistance has been also observed in platinum drug resistance on cancer [44]. The exposure of Cu(II) soluble in tumor mammalian cells MCF7 led to clear increase in the proliferation of the cells due to Cu uptake and disturbances of the redox status [45].
If we compare the degree of cell proliferation, the expression of genes associated with cell cycle regulation, and the levels of ROS production and related oxidative processes, in copper-treated mammary epithelial cell lines with equivalent metabolic rates, namely, MCF7 tumor cells [45] and the epithelial MCF10A nontumor cells, we can observe different behavior between cells. The observed copper-stimulated proliferation of tumor cells was not correlated with cyclin upregulation or increased cytosolic concentration of the metal, but rather with enhanced ROS generation and elevated levels of lipid peroxidation, which gave rise to alterations in the topography of the cell membranes [45]. In contrast, we observed here in this work that copper did not stimulate the proliferation of nontumorigenic epithelial MCF10A cells, although the enhanced intracellular uptake of the metal was accompanied by a moderate overexpression of cyclins D1 and B1, and an increase in the ratio of reduced to oxidized glutathione (GSH/GSSG).
Based on our findings, we suggest that normal mammary cells respond to increased levels of intracellular Cu by triggering cyclin mRNA expression and elevating the concentration of the endogenous antioxidant GSH. In turn, increased levels of antioxidant prevent abnormal ROS formation, which could cause oxidative stress and cell death. Finally, our findings point to proteins involved in cell cycle such as cyclins and CDK2 as a novel target for Cu interaction. More detailed investigation into its molecular mechanism will be important for our understanding of Cu metabolism in normal cells. We observed here that enhanced intracellular uptake and accumulation of Cu were followed by an overexpression of cyclin D1 and cyclin B1, with an increase in the ratio of reduced to oxidized glutathione (GSH/GSSG), but without ROS production. The results presented herein provide new insights into the molecular link between Cu excess and the control of cell cycle and, consequently, the mechanism by which changes in redox balance and ROS accumulation regulate cell proliferation.