Introduction

Copper oxide nanoparticles (CuO-NPs) have interesting physical and catalytic properties and are therefore frequently used as additives for industrial products such as paints, lubricants, heat-transfer fluids, conductive films, plastics and solid propellants [1–4]. In addition, due to their potent antimicrobial and biocidal properties CuO-NPs are used in medical products such as bandages and water filters [4–6]. The most important routes of exposure to CuO-NPs are inhalation, ingestion or skin contact [7]. The risk of occupational exposure to substantial amounts of copper-containing NPs is high for welders [8] and for workers operating universal electric motors [9].

CuO-NPs are considered as very toxic compared to other metal oxide nanoparticles [10]. Exposure of cells to CuO-NPs has been reported to cause inhibition of enzymes, expression of proinflamatory molecules and alterations in gene expression [1, 11, 12] as well as the activation of autophagic and apoptotic processes [12–15]. The generation of reactive oxygen species (ROS) and oxidative stress-mediated damage are considered to be involved in the observed cell toxicity of CuO-NPs [16, 17].

Wistar rats injected with CuO-NPs suffered from severe cognitive impairments including decreased learning and memory abilities [18]. To reach the brain, CuO-NPs have to cross the blood–brain barrier (BBB) which has been demonstrated for several types of inorganic nanoparticles including CuO-NPs [19–23]. In addition, CuO-NPs and other inorganic nanoparticles can reach the brain via nerve axons from the olfactory bulb [24–27].

The first cells in brain that encounter nanoparticles which have crossed the BBB from blood are astrocytes which are located between the endothelial cells of the vascular system and neurons [28]. Astrocytes have several essential homeostatic functions in the brain. These cells supply metabolites and gliotransmitters for neuronal function, protect the brain environment from toxins and oxidative stress and regulate the metal homeostasis of the brain [29–32]. Astrocytes have been suggested to play a key role in the copper metabolism of the brain [31, 33]. At least in culture, astrocytes efficiently accumulate, store and export ionic copper [34, 35]. However, an exposure to copper does also alter metabolic properties of astrocytes as copper accumulation in cultured astrocytes is accompanied by accelerated glycolytic flux and stimulated synthesis of glutathione (GSH) [36, 37].

Several studies have shown that cultured astrocytes efficiently accumulate silver and iron oxide nanoparticles and that these types of NPs have little toxic potential to astrocytes (for overview see: [32, 38]). In contrast, exposure of astrocytes to CuO-NPs has recently been show to acutely increase cellular ROS production and to severely damage cultured astrocytes most likely by copper-mediated oxidative stress [17]. As it is currently not known whether CuO-NPs in sub-toxic concentrations may affect metabolic properties of astrocytes, we have exposed cultured primary astrocytes to CuO-NPs and investigated the consequences of such an exposure on cell viability, glycolytic flux as well as the cellular contents of GSH and metallothioneins (MTs).

Materials and Methods

Materials

Fetal calf serum (FCS) and penicillin/streptomycin solution were obtained from Biochrom (Berlin, Germany) and Dulbecco’s modified Eagle’s medium (DMEM) from Gibco (Karlsruhe, Germany). Bovine serum albumin (BSA), sulfosalicylic acid, dithiothreitol, magnesium chloride hexahydrate, NADPH, NADP+, NAD+ and NADH were from Applichem (Darmstadt, Germany). Glutathione reductase, ATP, hexokinase, glucose-6-phosphate dehydrogenase, lactate dehydrogenase and glutamate pyruvate transaminase were purchased from Roche (Mannheim, Germany). The monoclonal mouse anti-MT antibody (clone E9; detects MTs 1 and 2) was purchased from Dako (Hamburg, Germany). The peroxidase-conjugated donkey anti-mouse antibody and the monoclonal mouse anti-actin antibody (clone Ab-5) were purchased from Dianova (Hamburg, Germany). Page ruler prestained protein ladder 10–170 kDa and hydrochloric acid were from Thermo Fisher Scientific (Fair Lawn, USA), Tween®20 from Serva (Heidelberg, Germany) and trichloroacetic acid from Jannsen (Geel, Belgium). All other chemicals were purchased at analytical grade from Merck (Darmstadt, Germany), Fluka (Buchs, Switzerland), Roth (Karlsruhe, Germany) or Sigma-Aldrich (Steinheim, Germany). Sterile 24-well cell culture plates and unsterile 96-well microtiter plates were from Sarstedt (Nümbrecht, Germany).

Primary Astrocyte Cultures

Astrocyte-rich primary cultures were prepared from brains of newborn Wistar rats as previously described [39, 40] by seeding 300,000 viable cells in wells of a 24-well culture plate in 1 mL culture medium (90 % DMEM containing 25 mM glucose, 10 % FCS, 18 U/mL penicillin G, 18 µg/mL streptomycin sulphate). The cultures were maintained in presence of 10 % CO2 in the humidified atmosphere of a Sanyo incubator (Osaka, Japan). Immunocytochemical staining for the astrocyte marker glial fibrillary acidic protein confirmed that these cultures contain predominately astrocytes [40, 41]. The culture medium was renewed every 7 days. Confluent cultures of an age between 14 and 28 days were used for the experiments described here.

Synthesis and Characterization of Copper Oxide Nanoparticles

CuO-NPs were synthesized by alkaline precipitation of copper nitrate and coated with 2,3-dimercaptosuccinic acid as described recently [17]. These particles are agglomerates of crystalline primary particles which have an average size of approximately 5 nm. Dispersed in culture medium CuO-NPs have a mean hydrodynamic diameter of 178 ± 12 nm and a ζ-potential of −10 mV [17]. A dispersion of CuO-NPs (30 mM copper) in water was diluted in culture medium to the final particle concentrations. The concentrations of CuO-NPs given here represent the concentration of total copper in a given incubation medium and not the concentration of particles.

Experimental Incubations

The cultures were washed twice with 1 mL prewarmed (37 °C) phosphate-buffered saline (PBS; 10 mM potassium phosphate buffer, pH 7.4, containing 150 mM sodium chloride) and subsequently incubated for up to 24 h at 37 °C in an incubator with 10 % CO2 in 1 mL culture medium containing CuO-NPs and/or other compounds in the concentrations given in the legends of the figures and tables. Ammonium tetrathiomolybdate (TTM) was prepared as 10 mM stock solution in water and diluted in culture medium to a final concentration of 100 µM.

To investigate the persistence of CuO-NP-induced metabolic alterations, astrocytes were preincubated for 24 h in 1 mL culture medium in the absence or presence of 100 µM CuO-NPs, washed twice with prewarmed (37 °C) PBS and subsequently incubated at 37 °C (in the humidified atmosphere of an incubator without CO2 supply) in 1 mL incubation buffer (IB; 20 mM HEPES, 5 mM glucose, 1.8 mM CaCl2, 1 mM MgCl2, 5.4 mM KCl, 145 mM NaCl, adjusted to pH 7.4) containing 100 µM of the membrane-impermeable copper chelator bathocuproine disulfonate (BCS). After a given incubation period, the incubation media were harvested and the cells were washed twice with 1 mL ice cold PBS. Dry cells were stored at −20 °C for quantification of copper and protein contents.

Copper and Protein Quantification

The copper content of CuO-NP-treated cells was determined by graphite furnace atomic absorption spectroscopy using a Varian (Darmstadt, Germany) AA-240Z spectrophotometer and a Varian GTA-120 graphite tube atomizer as recently described [17, 34]. The protein content of the cells per well was determined by the Lowry method [42] using bovine serum albumin as standard protein.

Viability and Vitality of Cultured Astrocytes

Cell vitality and functionality was assessed by quantification of the capacity of the cells to reduce 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), by determination of the cellular activity of the enzyme lactate dehydrogenase (LDH) and by propidium iodide (PI) staining. The MTT reduction capacity was quantified as described recently [17]. Briefly, after the experimental incubation the cells were washed with prewarmed PBS and incubated for additional 90 min at 37 °C with culture medium containing 0.5 mg/mL MTT. Subsequently, the cells were lysed in DMSO and the absorbance of the generated formazan was determined at 540 nm in wells of a microtiter plate by a Sunrise microtiter plate photometer (Tecan, Crailsheim, Germany). The activity of the enzyme LDH in cell lysates was measured by following the NADH consuming reduction of pyruvate to lactate as described previously [43]. The integrity of the cell membrane was determined by application of the fluorescent and membrane-impermeable dye PI [40]. The staining procedure was performed as recently described [17] with Hoechst 33342 (H33342) for nuclear staining of all cells present and with silver nitrate as positive control for cell toxicity.

Determination of Lactate, Glucose and Glutathione

The concentrations of extracellular lactate and glucose were determined by coupled enzymatic assays as described previously [40, 44]. The amount of glucose consumed was calculated as difference between the glucose content of the incubation medium harvested after a given incubation and the medium applied to the cells. The contents of total glutathione [GSx = amount of GSH plus two times the amount of glutathione disulfide (GSSG)] and GSSG in cell lysates and in incubation media were quantified by the Tietze method using a microtiter plate assay [40, 45]. For all experiments described here, the cellular GSSG levels were in the range of the detection limit of the assay used [40] and did not significantly contribute to the GSx values determined.

Western Blot Analysis

Western blots to test for the presence of MTs in astrocytes were performed as previously described [46, 47]. Briefly, after a given treatment, the cells in a well of a 24-well plate were directly lysed in 80 μL loading buffer (62.4 mM Tris/HCl, pH 6.8, 2 % (w/v) sodium dodecylsulfate, 10 % (w/v) glycerol, 0.01 % (w/v) bromphenole blue, 4 mM dithiotreitol). Lysates from three wells were pooled and stored at −20 °C. For Western blotting, the lysates were boiled for 10 min and subsequently incubated with 12.5 mM iodoacetamide for 15 min at 50 °C to mask sulfhydryl groups. Per lane of a polyacrylamide gel (4 % stacking gel, 15 % separation gel) 12.5 µg protein were loaded and separated for 1 h at 160 V. Subsequently the gel was electroblotted onto an Amersham Hybond™-C Extra nitrocellulose membrane (GE Healthcare, Munich, Germany) for 1 h at 360 mA. The membranes were washed with TBS-T (10 mM Tris/HCl, 150 mM NaCl, 0.1 % (v/v) Tween®20, pH 7.3) and subsequently unspecific binding sites were blocked for 1 h with 5 % (w/v) milk powder in TBS-T (TBS-T-MP). The membranes were incubated with the monoclonal anti-MT antibody (1:500 in TBS-T-MP) or the monoclonal anti-actin antibody (1:2,000 in TBS-T-MP) overnight at 4 °C on a Heidolph (Schwabach, Germany) Duomax 1030 shaker. After the incubation, the membranes were washed thrice for 15 min with TBS-T and subsequently incubated for 1 h with a peroxidase-conjugated anti-mouse antibody (1:20,000 in TBS-T-MP) at room temperature. After three additional washing steps, the blots were developed with the ECL Western blot detection reagents (GE Healthcare) on a Amersham Hyperfilm ECL X-ray film (GE Healthcare). The signals for iodoacetamide-conjugated MTs (12 kDa) and actin (42 kDa) were detected at the expected molecular masses.

Presentation of Data

The data shown in figures and tables represent means ± standard deviations of values derived from n experiments that were performed on independently prepared cultures. Figures showing Western blots and microscopic images are from representative experiments that were reproduced at least twice on independently prepared cultures. Statistical analysis between multiple groups of data was performed by ANOVA followed by the Bonferroni’s post hoc test, analysis of the significance of differences between two sets of data by the paired Student’s t test. p > 0.05 was considered as not significant.

Results

Accumulation and Toxicity of CuO-NPs in Cultured Astrocytes

To investigate cellular uptake and toxicity of CuO-NPs, cultured astrocytes were incubated with up to 300 µM (18.9 µg/mL) of copper as CuO-NPs for up to 24 h and the MTT reduction capacity as well as the specific cellular copper contents were determined (Fig. 1). CuO-NP-treated astrocytes showed a time- and concentration-dependent increase in their specific copper contents (Fig. 1a). Within 24 h of incubation with 100, 200 and 300 µM CuO-NPs, the cells increased their specific cellular copper content from an initial value of 0.9 ± 0.4 to 19.7 ± 2.9, 61.3 ± 8.2 and 62.2 ± 7.1 nmol/mg, respectively. However, the strong increase in the copper contents of CuO-NP-treated cells was accompanied by a time- and concentration-dependent decrease in the MTT reduction capacity of the cells (Fig. 1b), which is considered as indicator for compromised cell vitality of copper-treated cells [17, 34]. While the MTT reduction capacity of astrocytes that had been incubated with 100 µM CuO-NPs did not differ significantly to that of control cells (incubation without CuO-NPs), the MTT reduction capacity of cells exposed to 200 or 300 µM CuO-NPs was found significantly impaired already after 12 and 6 h of incubation, respectively. For these conditions, the MTT reduction capacity declined further with longer incubation time and reached after 24 h of incubation values that corresponded to around 40 (200 µM) and 30 % (300 µM) of the initial values (Fig. 1b). In contrast, control astrocytes and cultures exposed to 100 µM CuO-NPs still possessed 85 % of the initial MTT reduction capacity after 24 h of incubation (Fig. 1b). Similar experiments performed with CuCl2 in micromolar concentrations revealed that exposure of astrocytes for 24 h to CuCl2 lowered the cellular MTT reduction capacity to 72 ± 3 (100 µM), 61.5 ± 4.6 (200 µM) and 49.7 ± 4.5 % (300 µM) of the control values (n = 3) and significantly increased the specific cellular copper contents to 13.7 ± 1.1 (100 µM), 17.6 ± 2.2 (200 µM) and 20.2 ± 2.1 nmol/mg (300 µM).

Fig. 1
figure 1

Effects of a treatment of astrocytes with CuO-NPs on cellular copper content, cell viability and lactate generation. The cells were incubated without or with CuO-NPs in the indicated concentrations for up to 24 h and the specific cellular copper content (a), the cellular MTT reduction capacity (b) as well as the extracellular lactate concentration (c) were determined. The data shown represent means ± standard deviations of values derived from three experiments that were performed on independently prepared cultures. Asterisks in a and b indicate the significance of differences (ANOVA) to the respective values determined for control cells (incubation in the absence of CuO-NPs) (*p < 0.05; **p < 0.01; ***p < 0.001). Hashes in c indicate the significance of differences (t test) in the extracellular lactate concentrations of astrocytes that were incubated in the absence and the presence of 100 µM CuO-NPs (# p < 0.05; ## p < 0.01; ### p < 0.001)

Full size image

The concentration-dependent toxicity of CuO-NPs was confirmed by PI staining. While cultures incubated for 24 h without (Fig. 2a, b) or with 100 µM CuO-NPs (Fig. 2c, d) did hardly contain any PI-positive cells (Fig. 2a, c), a substantial number of PI-positive cells was detected after incubation of astrocytes with 200 (Fig. 2e) or 300 µM (Fig. 2g) CuO-NPs as well as after exposure to AgNO3 (Fig. 2i), which served as positive control for toxicity [48].

Fig. 2
figure 2

Cytochemical staining of CuO-NP-treated astrocytes for membrane integrity. Primary astrocyte cultures were incubated for 24 h in the absence (a, b) or the presence of CuO-NPs (c–h) in the indicated concentrations and subsequently stained for loss of membrane integrity with propidium iodide. Presence of all cell nuclei was visualized by staining with the membrane-permeable dye H33342. An incubation with silver nitrate (100 µM) for 24 h was used as positive control for impaired membrane integrity (i). The images are from a representative experiment that was reproduced twice on independently prepared cultures with similar outcome. The scale bar in a applies to all panels

Full size image

Cultured astrocytes contained after 24 h of incubation with 100 µM CuO-NPs a specific cellular copper content of 19.7 ± 2.9 nmol/mg, while cells incubated in the absence of nanoparticles contained only around 1 nmol copper/mg (Table 1). Despite of this severe increase in the cellular copper content, astrocyte cultures that had been incubated for 24 h with 100 µM CuO-NPs did not significantly differ from control cells (incubated in the absence of CuO-NPs) in their cellular MTT reduction capacity (Fig. 1a; Table 1), PI-staining (Fig. 2c), in their protein content (Table 1) or their specific cellular LDH activity (Table 1). This demonstrates that cultured astrocytes are not damaged by a 24 h incubation with 100 µM CuO-NPs. Thus, this incubation paradigm was considered as suitable to investigate the consequences of an exposure to sub-toxic concentration of CuO-NPs on metabolic properties of cultured astrocytes.

Table 1 Effects of CuO-NPs on cellular parameters of cultured astrocytes
Full size table

CuO-NP Exposure Affects Glycolytic Flux and the Cellular Levels of GSH and Metallothioneins

To test for a potential effect of CuO-NPs on the glucose metabolism, astrocytes were incubated without or with 100 µM CuO-NPs. After 24 h of incubation, glucose consumption and lactate release by CuO-NP-treated astrocytes were found significantly increased by 23 and 24 %, respectively, compared to control cells that were incubated without CuO-NPs (Table 1). During the first hours of incubation, almost identical extracellular lactate concentrations were determined for astrocytes treated without or with CuO-NPs. However, already after 12 h of incubation the extracellular lactate concentration was significantly increased for CuO-NP-treated astrocytes compared to control cells (Fig. 1c). In contrast, the extracellular lactate concentration of astrocytes that had been incubated with toxic concentrations of 200 or 300 µM CuO-NPs for 24 h were not increased (200 µM CuO-NPs: 6.6 ± 1.2 mM) or even lowered (300 µM CuO-NPs: 4.9 ± 1.0 mM) compared to control cells (6.7 ± 1.2 mM).

Exposure of astrocytes for 24 h with 100 µM CuO-NPs significantly increased the specific cellular GSx contents by about 36 % compared to the values determined for control cells (Table 1), while the specific cellular contents of GSSG remained very low (below 2 % of the GSx values) in both control and CuO-NP-treated astrocytes (Table 1). Finally, also the cellular level of MTs was found significantly elevated compared to control cells after 24 h of incubation with 100 µM CuO-NPs (Fig. 3).

Fig. 3
figure 3

Content of metallothioneins (MTs) in astrocytes. Cultured astrocytes were incubated for 24 h without or with 100 µM CuO-NPs in the presence or the absence of 100 µM tetrathiomolybdate (TTM) and their lysates were analyzed for MTs and actin by Western blotting. a shows a representative blot, b the results of the analysis of the signal intensities for blots performed on lysates derived from experiments on four independently prepared cultures. Asterisks indicate the significance of differences (ANOVA) compared to the data of control cells (treated without CuO-NPs and TTM) (***p < 0.001)

Full size image

Effects of the Copper Chelator TTM on the CuO-NP-Induced Alterations in Astrocytic Properties

To test for a potential involvement of copper ions that had been liberated from accumulated CuO-NPs in the observed effects of an CuO-NP-treatment on lactate production, glucose consumption and the cellular contents of GSx and MTs, astrocytes were incubated for 24 h with 100 µM CuO-NPs in the absence or presence of the membrane-permeable copper ion chelator TTM. Presence of TTM lowered the CuO-NP-dependent increase in specific cellular copper content by 47 % (Table 2), while the CuO-NP-induced increases in glucose consumption, lactate production and levels of MTs were completely prevented in the presence of TTM (Table 2; Fig. 3). In contrast, TTM did not affect the CuO-NP-induced increase in the specific cellular GSx contents but increased already in the absence of CuO-NPs the specific cellular GSx content (Table 2).

Table 2 Effects of the copper chelator TTM and/or CuO-NPs on astrocytes
Full size table

Test for Persistence of the CuO-NP-Induced Stimulation of Glycolytic Flux

To investigate whether the observed CuO-NP-induced alterations in astrocytic glucose metabolism depends on the presence of extracellular CuO-NPs, the cultures were pre-incubated for 24 h without or with 100 µM CuO-NPs, washed and subsequently incubated at 37 °C for 6 h in an copper-free incubation buffer containing 100 µM of the extracellular copper chelator BCS. During the 6 h main incubation the copper content of the CuO-NP-pre-incubated cells was lowered by 45 % from 19.7 ± 2.9 (Table 1) to 10.9 ± 1.1 nmol/mg (Table 3). This loss of cellular copper was accompanied by a matching increase in the extracellular copper content, as the medium harvested after the 6 h incubation contained 1.7 ± 0.3 nmol of copper which corresponds to 11.2 ± 1.1 nmol copper/mg protein. An analogous 6 h incubation of CuO-NP-pre-incubated astrocytes at 4 °C partially prevented (p < 0.05; n = 3) the loss of cellular copper (cellular content: 16.3 ± 2.1 nmol/mg) and the accumulation of extracellular copper content (5.2 ± 1.0 nmol/mg), suggesting that the reported temperature-sensitive astrocytic export of copper ions [35] is involved in the observed loss of cellular copper. However, as this copper loss was only partially prevented by lowering the incubation temperature to 4 °C, also detachment of CuO-NPs which had been extracellularly adsorbed to the cell membrane [17] is likely to contribute to this process.

Table 3 Persistence of alterations in cellular parameters after removal of CuO-NPs
Full size table

The additional 6 h incubation after a 24 h pre-incubation lowered the MTT reduction capacity of CuO-NP-treated astrocytes to some extent while the cellular protein content per well was not altered compared to control cells (Table 3). In addition, PI staining of the cells after the 6 h main incubation confirmed that the membrane integrity of the CuO-NP-pre-incubated cells was not impaired (Fig. 4).

Fig. 4
figure 4

Cytochemical staining of astrocytes for membrane integrity after removal of CuO-NPs. Primary astrocyte cultures were pre-incubated without (a, b) or with 100 µM of CuO-NPs (c, d) for 24 h, washed and subsequently incubated for additional 6 h in incubation buffer containing 100 µM BCS. Subsequently the cells were stained with propidium iodide and H33342. A pre-incubation with silver nitrate (100 µM) for 24 h followed by a 6 h main incubation without silver nitrate was used as positive control for impaired membrane integrity (e, f). The images are from a representative experiment that was reproduced twice on independently prepared cultures with similar outcome. The scale bar in a applies to all panels

Full size image

Astrocytes that had been pre-incubated for 24 h with 100 µM CuO-NPs maintained during a following 6 h main incubation an increased glyolytic flux. Compared to control cells that had been pre-incubated without nanoparticles, the amounts of lactate released and glucose consumed within 6 h by the cells was significantly increased by 24 and 70 %, respectively, in CuO-NP-treated cultures (Table 3) and the specific lactate release rate of these cells, as calculated from the almost linear increase in the extracellular lactate accumulation (Fig. 5a), was increased by 40 % compared to controls (Table 3).

Fig. 5
figure 5

Persistence of the stimulated lactate release and accumulation of extracellular glutathione in CuO-NP-treated astrocytes. Cultured astrocytes were pre-incubated for 24 h in the absence (Control) or the presence of 100 µM CuO-NPs. Subsequently the cells were washed and incubated for up to 6 h in copper-free incubation buffer containing 100 µM BCS. During the 6 h main incubation the extracellular contents of lactate (a) and GSx (b) were determined. The data shown represent means ± standard deviations of values derived from three experiments that were performed on independently prepared cultures. Asterisks indicate the significance of differences (t test) of data obtained for cells that had been pre-incubated without and with CuO-NPs (*p < 0.05; **p < 0.01)

Full size image

Effects of a CuO-NP-Pre-incubation on GSH Export

After a 24 h pre-incubation of astrocytes with 100 µM CuO-NPs, the specific cellular GSx content was significantly increased compared to control cells (Table 1). After a subsequent 6 h main incubation without CuO-NPs the specific GSx content of cells that had been exposed to CuO-NPs remained higher than that of the respective control cells, although this difference did not reach the level of significance (Table 3). In addition, the cellular GSSG levels were very low after the 6 h main incubation for cells that had been pre-incubated without or with CuO-NPs (Table 3).

Quantification of the extracellular GSx contents revealed that GSx accumulated quicker in the incubation medium of astrocytes that had been pre-incubated with CuO-NPs compared to control cells (Fig. 5b). After 6 h of incubation the extracellular GSx content of CuO-NP-pre-incubated cultures was significantly increased by 70 % and the GSx accumulation rate of such cells was doubled compared to the respective data of control cultures (Table 3). The extracellular GSx values accounted almost exclusively for GSH, as only minute amounts of extracellular GSSG (<0.8 nmol/mg) were detectable for astrocytes that had been pre-incubated without or with CuO-NPs (data not shown).

The sum of extracellular plus cellular GSx contents determined after the 6 h main incubation without (43.6 ± 14.1 nmol/mg; n = 3) and with CuO-NPs (58.2 ± 11.9 nmol/mg; n = 3) did not differ significantly (p > 0.05) from the values determined after the respective pre-incubation without or with CuO-NPs (Table 1).

Discussion

To test for the potential of CuO-NPs to compromise the viability of astrocytes and to affect metabolic properties of these cells, we exposed primary cultured astrocytes in serum-containing culture medium to CuO-NPs for 24 h and found that astrocytes accumulated such particles in a time- and concentration-dependent manner. Astrocytes tolerated concentrations of 100 µM CuO-NPs very well, but were severely damaged after exposure to higher particle concentrations. The observed accumulation of CuO-NPs by astrocytes and the toxicity induced by elevated concentrations of CuO-NPs are in line with results obtained recently for astrocytes that had been acutely exposed for up to 6 h to CuO-NPs [17]. Compromised viability of cultured astrocytes was previously observed for conditions which elevated the specific cellular copper content within a few hours after exposure to copper salts or CuO-NPs to values of above 10 nmol/mg, while lower specific copper contents were not accompanied by a compromised viability [17, 34, 35]. This suggests that cultured astrocytes can deal well with an elevated cellular copper content as long as it does not exceed a given threshold value.

Studies on astrocytes [17] and other cell types [12, 13, 49–52] revealed that CuO-NP-treated cells suffer from an increased formation of cellular ROS. For astrocytes, elevated ROS levels and toxicity were prevented by co-application of the membrane-permeable copper chelator TTM, suggesting that copper ions which have been liberated from the accumulated CuO-NPs are responsible for the ROS-induced toxicity [17]. Indeed, TTM prevented also the impairment of cellular MTT reduction capacity and membrane integrity during a 24 h exposure of cultured astrocytes to 200 µM CuO-NPs (data not shown).

The uptake of CuO-NPs into astrocytes in serum-containing medium is likely to be mediated by endocytosis as recently described for silver and iron oxide NPs [48, 53, 54]. At least for other cell types CuO-NPs were found to be present in cellular vesicles [50, 55, 56]. Such vesicles may rapidly enter the lysosomal pathway and the CuO-NPs will then encounter an environment that facilitates the release of copper ions from the particles [52, 56]. As alternative scenario for the observed copper accumulation by CuO-NP-treated astrocytes it can currently not be exclude that a part of the cellular copper is derived from uptake of copper ions which have been liberated from the applied CuO-NPs. However, astrocytes exposed to CuCl2 contained less copper and remained a higher MTT reduction capacity compared to cells exposed to the respective concentration of copper as CuO-NPs. In addition, the size distribution curves of the CuO-NPs in culture medium were not obviously shifted to smaller sizes within a 24 h incubation (data not shown), demonstrating that CuO-NPs are not rapidly disintegrate to copper ions under the conditions used. These results support the view that the effects observed for CuO-NP-treated astrocytes are predominantly caused by copper liberated in the cells after endocytotic uptake of CuO-NPs.

A treatment of astrocytes with 100 µM CuO-NPs stimulated glycolytic lactate production, elevated the specific cellular GSx content and induced the synthesis of MTs. Similar alterations in astrocytic metabolism were observed after exposure of viable astrocytes to CuCl2 [31, 36, 37]. These results suggest that not the CuO-NPs themselves but rather copper ions which are liberated from the accumulated CuO-NPs are responsible for the observed alterations in astrocytic metabolism. This hypothesis is strongly supported by the ability of the membrane-permeable copper chelator TTM to prevent the observed stimulation of lactate production and MT synthesis in CuO-NP-treated astrocytes. However, a potential TTM-induced disintegration of extracellular CuO-NPs and extracellular chelation of the liberated copper ions would also explain the abolishing effects of TTM on the CuO-NP-induced stimulation in lactate and MT synthesis. Indeed, although the size distribution curves of the dispersed CuO-NPs were not obviously shifted to smaller sizes during a 24 h incubation with an equimolar concentration of TTM (data not shown), the strong absorbance of 100 µM copper-free TTM at 320 and 470 nm [57] was lowered by 30–40 % after incubation with 100 µM copper as CuO-NPs, but completely abolished by application of 100 µM copper ions (data not shown). These data demonstrate that in presence of TTM at least some copper ions were extracellularly chelated by TTM, while the majority of the copper applied remained during incubation in nanoparticle form. This, extracellular chelation may partly contribute to the observed lowering effect by TTM on the cellular copper content of CuO-NP-treated astrocytes.

The ability of TTM to prevent the CuO-NP-induced stimulation of lactate and MT synthesis makes also a potential contribution of the coating material on these effects unlikely, consistent with results from previous studies which revealed that even large amounts of accumulated copper-free DMSA-coated iron oxide nanoparticles (400 nmol iron/mg) did not affect cell viability, GSH and glucose metabolism of cultured astrocytes [58].

The mechanisms involved in the stimulation of glycolytic lactate production and synthesis of GSH and MTs by copper ions are currently unclear. The copper-induced stimulation of glycolytic flux was prevented by an inhibitor of protein synthesis [37]. This would explain the delayed onset of the stimulation of lactate production by CuO-NPs, which was only observed after more than 8 h of incubation, as well as the persistence of this stimulation even after removal of extracellular CuO-NPs. Also the increased astrocytic levels of MTs are likely to be a consequence of elevated concentrations of copper ions as intracellular metal ions can stimulate MT synthesis in astrocytes [31, 47] and as the chelation of such ions is known to prevent MT synthesis [59–61].

The increased GSx levels in viable astrocytes that had been exposed to CuO-NPs are in line with literature data for CuCl2-treated astrocytes [36] and are similar to elevated astrocytic GSx values found after treatment of cultured astrocytes with ammonia, formaldehyde, arsenite or arsenate [36, 62–64]. Such elevated cellular GSx contents have been discussed as consequence of a stressor-induced stimulation of de novo GSH synthesis or of an accelerated uptake of GSH precursors. Also the accelerated GSH export from CuO-NP-treated astrocytes resembles that reported for CuCl2-treated astrocytes [36] and is likely to be a direct consequence of the copper-induced increase in cellular GSH concentration. GSH export from astrocytes is predominantly mediated by the multidrug-resistance protein 1 and has a KM-value of around 100 nmol/mg [62, 64]. Thus, the elevated cellular GSH contents of CuO-NP-treated cultures will directly cause an accelerated GSH export, as previously also reported for astrocytes exposed to copper chloride or arsenicals [36, 62, 64]. A contribution of GSSG export [65, 66] in the increased extracellular GSx accumulation of CuO-NP-treated cells can be excluded, as GSSG hardly contributed to the GSx values observed for cells and media.

Presence of TTM prevented the stimulation of lactate production and MT synthesis in CuO-NP-treated astrocytes, suggesting that intracellular copper ions which are liberated from internalized CuO-NPs mediate these processes. In contrast, presence of TTM was not able to prevent the increase in cellular GSx contents in CuO-NP-treated astrocytes. However, this does not challenge the concept that copper ions liberated from internalized CuO-NPs stimulate GSH synthesis, as TTM itself even in the absence of exogenous copper caused an increase in astrocytic GSx contents. This was not unexpected as TTM has been reported to increase GSH contents in rat hepatocytes [67] and as molybdate application to rats increased GSH levels in liver and kidney [68].

In conclusion, a treatment of cultured astrocytes with sub-toxic amounts of copper in form of CuO-NPs stimulated their glycolytic lactate production and elevated the cellular levels of GSH and MTs. Such alterations in astrocytic metabolism were also reported for astrocytes that had been treated with copper ions [31, 36, 37], suggesting that in astrocytes copper ions are rapidly liberated from internalized CuO-NPs which subsequently induce the observed metabolic alterations. As cellular GSH and MTs are involved in cellular copper storage [33] the increase in the cellular levels of these compounds is likely to contribute to the relative resistance of astrocytes against copper-induced ROS formation and oxidative damage. On the other hand, the copper-induced accelerated release of GSH and lactate may help to supply additional metabolites to neighboring neurons which could serve as precursors for neuronal GSH synthesis [69] and as energy substrate [70, 71], respectively. The observation that astrocytes in culture deal well with moderate concentrations of copper in form of CuO-NPs, suggests that also astrocytes in brain have the potential to accumulate CuO-NPs and to store copper liberated from such NPs, which may help to protect neighboring cells against a toxic potential of CuO-NPs.