Biphasic effects of copper on neurotransmission in rat hippocampal neurons

Authors


Address correspondence and reprint requests to Dr Carlos Opazo, Departamento de Fisiología, Facultad de Ciencias Biológicas, Universidad de Concepción, P.O. Box 160-C, Concepción Chile. E-mail: carlosopazo@udec.cl

Abstract

J. Neurochem. (2011) 119, 78–88.

Abstract

The importance of copper in the CNS is well documented, but the mechanisms related to its brain functions are poorly understood. Copper is released at the synaptic cleft, where it may modulate neurotransmission. To understand the functional impact of copper on the neuronal network, we have analyzed the synaptic activity of primary rat hippocampal neurons by using different approaches including whole cell patch clamp, recording of calcium transients, immunofluorescence and western blot. Here, we show that copper produces biphasic changes in neurotransmission. When copper is acutely applied to the plate it blocks neurotransmission. Interestingly, when it is applied for 3 h to hippocampal neurons it mainly increases the frequency and amplitude of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)ergic currents (control: 0.21 ± 0.05 Hz/22.9 ± 1.3 pA; copper: 0.68 ± 0.16 Hz/30.5 ± 2.5 pA), intracellular calcium transients (control: 0.05 ± 0.013 Hz; copper: 0.11 ± 0.02 Hz) and evoked AMPA currents (control: EC50 8.3 ± 0.5 μM; copper: EC50 2.9 ± 0.2 μM). Moreover, our results suggest that copper increases GluA1 subunit levels of the AMPA receptor through the anchorage of AMPA receptors to the plasma membrane as a result of PSD-95 accumulation. We also found that copper-treated neurons displayed an undistinguishable neurotransmission to control neurons after 24 h of treatment, indicating that changes in neurotransmission induced by copper at 3 h of incubation are homeostatically regulated after long-term exposure to the metal. Together, our data reveal an unexpected biphasic effect of copper on neurotransmission, which may be relevant to understand the effects of this ion in brain diseases that display copper dyshomeostasis such as that observed in Alzheimer’s disease (AD).

Abbreviations used
AD

Alzheimer’s Disease

AMPA

α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid

APV, DL-2-amino-5-phosphonovaleric acid; CNQX

6-cyano-7-nitroquinoxaline-2,3-dione

GABA

γ-aminobutyric acid

Copper is an essential transition metal that participates in the regulation of brain physiology, being a key structural component of various proteins and a co-factor for enzymes that are critical for brain function, including enzymes involved in antioxidant defense and cellular respiration (Mathie et al. 2006). More recently, some reports have described the effect of copper at the synaptic level, where it modulates complex parameters such as LTP (Goldschmith et al. 2005; Leiva et al. 2009) and receptor pharmacology (Weiser and Wienrich 1996; Vlachováet al. 1996; Sharonova et al. 1998). The average copper concentration in the CSF has been estimated to be approximately 70 μM, whereas the normal extracellular copper concentration in the brain is 0.2–1.7 μM (Stuerenburg 2000; Tarohda et al. 2004). Nevertheless, these levels may be exceeded in the synaptic cleft under either normal or pathological conditions because copper can be released to the extracellular space from synaptosomes (Kardos et al. 1989). In fact, copper is released in a calcium dependent manner under high potassium (Kardos et al. 1989) and chemical stimulation driven by NMDA activation (Schlief et al. 2005). Moreover, micromolar concentration of copper (400 μM) is present in senile plaques in AD brains (Lovell et al. 1998), which could be a source of copper for the neurons surrounding these pathological structures. In addition, there are several reports that support a clear link between brain copper dyshomeostasis and AD (Bush 2003).

Under both scenarios, physiological or pathological, post-synaptic receptors may be targets for copper. Indeed, it has been observed that acute applications of copper rapidly blocks GABA and NMDA neurotransmission on rat olfactory bulb neurons (Trombley and Shepherd 1996), and also inhibits AMPAergic neurotransmission on rat cortical neurons (Weiser and Wienrich 1996). Indeed, Cu2+ inhibits AMPA/kainate receptors expressed by rat cortical neurons in culture, with an IC50 of 4.3 ± 0.6 μM (with 100 μM kainate, holding potential −60 mV) (Weiser and Wienrich 1996); thus, suggesting that copper has an acute modulatory role on the AMPA receptor in neurotransmission. To date, the effects that copper may exerts on neurotransmission for prolonged periods of incubation remain unknown and therefore here we have studied the neurotransmission of primary cultures of hippocampal neurons exposed to low micromolar concentrations of copper at different times (0, 3 and 24 h). Remarkably our results suggest that copper after 3 h of incubation, induces an increase in synaptic activity through AMPA receptor at the post-synaptic density via a mechanism that involves the accumulation of PSD95 protein. Thus, our findings represent a novel mechanism for the action of copper, which may have implication for neurophysiology and neuropathology of CNS.

Materials and methods

Primary cultures of rat hippocampal neurons

Hippocampal neurons were obtained from 18-day pregnant Sprague-Dawley rats and maintained for 10–14 days in vitro (DIV) as previously described (Aguayo and Pancetti 1994). Animals were obtained from the animal house of Catholic University of Chile (Santiago, Chile). All animals were handled in strict accordance with NIH recommendations and approved by the appropriate committee at the University of Concepción (Concepción, Chile).

Electrophysiology

Experiments were performed in the ‘whole-cell’ configuration. Recording pipettes were pulled from borosilicate glass (WPI, Sarasota, FL, USA) in a horizontal puller (Sutter Instruments, Novato, CA, USA). Membrane currents were measured using an Axopatch-200B amplifier (Molecular Devices, LLC, Sunnyvale, CA, USA) and an inverted microscope (Eclipse TE200-U; Nikon Instruments Inc., Tokyo, Japan). Data were collected, stored and analyzed using a data acquisition system card (Molecular Devices, LLC) and the pClamp9 software (Molecular Devices, LLC). For synaptic activity records, data were analyzed using the Minianalysis software, obtaining the frequency, amplitude and decay time of the records. All experiments were performed at room temperature (20–25°C) using a membrane potential of −60 mV. Data are given as means ± SEM and are obtained from more than five experiments.

Solutions and drugs

The intracellular medium contained (in mM): 120 KCl, 2 MgCl2, 2 ATP-Na2, 10 BAPTA, 0.5 GTP and 10 HEPES (pH 7.4). The extracellular medium contained (in mM): 150 NaCl, 5.4 KCl, 2 CaCl2, 1 MgCl2, 10 glucose and 10 HEPES (pH 7.4). Records were performed in the presence of 100 nM TTX to inhibit action potentials as described (Sepúlveda et al. 2009). AMPA currents were isolated with bicuculline (5 μM), DL-2-amino-5-phosphonovaleric acid (APV) (5 μM) and strychnine (1 μM). For GABA currents we used 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) (4 μM), APV (5 μM) and strychnine (1 μM).

Immunofluorescence

Hippocampal neurons plated in 35 mm dish were washed in phosphate-buffered saline (PBS) (pH 7.4) and fixed with cold methanol (−20°C) for 10 min. Then, the dish was washed again in PBS and neurons were permeabilized and blocked for 30 min with PBST (PBS + triton 0.1%) and bovine serum albumin 10%. Then, cells were incubated with the following primary antibodies for 16 h: anti-MAP2, 1 : 600 (Santa Cruz Biotechnology, CA, USA); anti-GluA1, 1 : 200 (Santa Cruz Biotechnology); anti-GABAAα-1 subunit, 1 : 200 (Neuromab, CA, USA); anti-PSD95, 1 : 200 (NeuroMab); and anti-GluA2, 1 : 200 (Chemicon, Temecula, CA, USA). Secondary antibodies conjugated with FITC, Cy3 and Cy5 were used for fluorescent staining (Jackson ImmunoResearch Laboratories, PA, USA). All of them were used at 1 : 200 for 2 h. Finally, samples were mounted in fluorescent mounting medium (Dako, CA, USA) and images were obtained under a Nikon Eclipse confocal microscope (Nikon Instruments Inc.) The immunoreactivity of the different synaptic proteins was quantified at the primary processes with the aid of the ImageJ software (NIH). Fluorescent signal was quantified as relative units using a region of interest of 10 μm of length. To study the immunoreactivity of GluA1 subunits in non-permeabilized neurons, the neuronal primary cultures were fixed with paraformaldehyde (4%) for 10 min, as previously reported (Noel et al. 1999). We confirm the non-permeabilized state of the neurons by negative immunostaining with MAP2 antibody.

Western blots

Equal amounts of proteins were separated by 10–12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis as previously described (Avila et al. 2010). Protein bands were transferred onto nitrocellulose membranes, blocked with 5% milk and incubated with the following primary antibodies: anti-α-tubulin, 1 : 5000 (Sigma, St Louis, MO, USA); anti-GluA1, 1 : 500 (extracellular N-terminal epitope; Santa Cruz Biotechnology); anti-GABAAα-1 subunit, 1 : 500 (Neuromab); anti-PSD95 1 : 1000 (Neuromab); β-actin, 1 : 2000 (Santa Cruz Biotechnology); and anti-Ctr1, 1 : 500 (kindly provided by Professor Dennis Thiele and Dr. Yasuhiro Nose). Immunoreactive bands were detected with secondary antibodies conjugated with HRP (Santa Cruz Biotechnology) and were visualized with ECL plus western blotting detection system (PerkinElmer, MA, USA).

Analysis of intracellular calcium transients

Neurons were loaded with Fluo-4 AM (1 μM in pluronic acid/dimethylsulfoxide; Molecular Probes, Eugene, OR, USA) for 30 min at 37°C and then washed twice with external solution as described above and incubated for 30 min at 37°C. Neurons were mounted in a perfusion chamber that was placed on the stage of an inverted fluorescent microscope (Eclipse TE), equipped with a xenon lamp and a 40× objective (22–24°C). Cells were subsequently illuminated for 200 ms using a computer-controlled Lambda 10-2 filter wheel (Sutter Instruments) and regions of interest were simultaneously selected on neuronal somata containing Fluo-4 fluorescence (Ex:Em;480 : 510 nm) in a field having usually more than 10 cells. Images were collected at 2–5 s intervals during a continuous 5-min period of recording with a 12-bit cooled SensiCam camera (PCO, Kelheim, Germany). Finally, calcium transients, as defined by their TTX sensitivity (Gu et al. 1994), were acquired and analyzed off line with Axon Instruments Workbench 2.2 software.

Results

Copper increases synaptic activity in primary cultures of rat hippocampal neurons

The effect of copper on synaptic activity has not been described in detail to date. So far, it is known that copper inhibits AMPAergic activity in cortical neurons when it is acutely applied to neurons by direct perfusion (Weiser and Wienrich 1996). In fact, AMPA evoked currents in hippocampal neurons in the presence of copper (acute application) were decreased compared with the control neurons treated with AMPA alone (Fig. 1a). On these primary neurons copper inhibited AMPA currents with an IC50 22.9 ± 3.8 μM (Fig. 1b). Moreover, the synaptic activity decreases in the presence of copper (Fig. 1c–e). These results are in agreement with previous reports that described the inhibitory effect of copper on neurotransmission when it is applied acutely to the cortical neurons (Weiser and Wienrich 1996). To evaluate if these changes were maintained after hours of exposure, rat primary hippocampal cultures (10–14 days in vitro) were treated with copper (CuCl2; up to 10 μM) for a short period of time (3 h). Using the whole cell mode in the patch clamp we found that copper induced a significant increase either in the frequency (control: 0.36 ± 0.17 Hz; copper: 0.74 ± 0.36 Hz; ***< 0.001), amplitude (control: 22.92 ± 1.34 pA; copper: 30.50 ± 2.52 pA; **< 0.01) and the time constants (control: 30.78 ± 1.58 ms; copper: 37.06 ± 1.68 ms; *< 0.05) of synaptic events in cells pre-treated with copper (Figs 2 and 3). Under our experimental conditions, 10 μM of copper was the minimal concentration able to increase synaptic activity (Fig. 2). Moreover, these results were specific for copper because under similar experimental conditions other transition metals such as iron and zinc did not increased synaptic activity (Fig. 3). Interestingly, longer time copper applications (24 h) did not change the synaptic activity (Fig. 4), indicating that the effect observed at 3 h was regulated by neurons after 24 h of treatment, as it is expected for a homeostatic neurophysiological mechanism (Watt et al. 2000). However, zinc and iron, were effective in blocking the amplitude of the synaptic events after 24 h (Fig. 4). These changes in the electrophysiological recordings were not explained by changes in neuronal viability because neither iron (10 μM) nor zinc (10 μM) affected the viability of these primary cultures after 24 h of treatment (data not shown). We also did not observe a decrease in the cell viability of primary hippocampal cultures treated with copper (10 μM) for 3 to 24 h (data not shown).

Figure 1.

 Acute application of copper inhibits AMPA evoked currents and decreases synaptic activity in rat hippocampal neurons. (a) Representative traces of AMPA evoked currents in the absence or presence of copper (up to 500 μM). (b) The graph summarizes the inhibitory effect of copper on AMPA evoked currents (IC50 = 22 ± 3.8 μM). (c) Typical miniature synaptic currents of rat hippocampal neurons treated acutely (perfusion) with copper 10 μM. Graphs summarize data obtained in these experiments with CuCl2 (10 μM): (d) frequency (Hz); (e) amplitude (pA). The bars are mean ± SEM from at least six neurons (*< 0.05, **< 0.01).

Figure 2.

 Minimal concentration of copper to increase synaptic activity in rat hippocampal neurons. (a) Typical miniature synaptic currents of rat hippocampal neurons treated in the absence or presence of copper (1, 5 and 10 μM) for 3 h. (b) The graph summarizes the frequency (Hz) obtained in these experiments. The bars are mean ± SEM from at least six neurons (**< 0.01).

Figure 3.

 Copper increases synaptic activity in rat hippocampal neurons. (a) Typical miniature synaptic currents of rat hippocampal neurons treated for 3 h in the absence or presence of the transition metals (10 μM) indicated in the figure. Graphs summarize data obtained in these experiments: (b) frequency (Hz); (c) amplitude (pA) and (d) decay (ms). The bars are mean ± SEM from at least six neurons (*< 0.05, **< 0.01, ***< 0.001).

Figure 4.

 Copper application for 24 h does not change the synaptic activity in rat hippocampal neurons. (a) Typical miniature synaptic currents of rat hippocampal neurons treated for 24 h in the absence (control) or presence of copper (10 μM). A zoom of representative miniature synaptic events of rat hippocampal neurons treated with copper for 24 h is shown. Graphs summarize data obtained in these experiments with copper (10 μM): (b) frequency (Hz); (c) amplitude (pA). Typical miniature synaptic currents of rat hippocampal neurons treated for 24 h in the absence (control) or presence of iron (10 μM) or zinc (10 μM) are summarized on the graphs: (d) frequency (Hz) and (e) amplitude (pA). The bars are mean ± SEM from at least six neurons (*< 0.05, ***< 0.001).

It has been previously demonstrated that calcium transients reflect changes in neurotransmission (Carrasco et al. 2007; Sepúlveda et al. 2009; Avila et al. 2010). Therefore, to further examine the effect of copper on the function of the neuronal network at 3 h of treatment, changes in calcium transients were subsequently recorded (Fig. 5e and f). We observed a significant increase in the frequency of calcium transients in copper-treated neurons (control: 0.05 ± 0.01 Hz; copper: 0.11 ± 0.02 Hz, respectively; *< 0.05) (Fig. 5f), which correlated with the increase in the frequency of synaptic events, supporting the effect of copper on neurotransmission.

Figure 5.

 Synaptic activity induced by copper is mediated by an increase in AMPA currents and calcium transients. (a) Representative traces of AMPA minis isolated by pharmacology maneuvers [APV (5 μM); strychnine (1 μM); bicuculline (5 μM)] in control neurons and copper-treated neurons for 3 h. Graphs summarize data obtained in these experiments: (b) frequency (Hz); (c) amplitude (pA) and (d) decay (ms). The bars are mean ± SEM from at least six neurons (**< 0.01, ***< 0.001). (e) Calcium transient records of control and copper-treated (10 μM) neurons for 3 h. (f) Frequency of calcium transient showing an increase after copper treatment (CuCl2, 10 μM) for 3 h (*< 0.05).

Copper enhances AMPAergic neurotransmission

To identify which neurotransmission network was involved in the synaptic changes induced by copper at this time (3 h), we pharmacologically isolated the AMPAergic [bicuculline (5 μM); APV (5 μM) and strychnine (1 μM)] and GABAergic [CNQX (4 μM); APV (5 μM) and strychnine (1 μM)] currents. These experiments showed that AMPAergic neurotransmission was the most affected mini current in copper-treated neurons (Fig. 5a). All parameters including frequency, amplitude and time constant of AMPA receptors were modified significantly (Fig. 5b–d). Indeed, although both the frequency and the amplitude of synaptic currents were enhanced (frequency, control: 0.21 ± 0.05 Hz, copper: 0.68 ± 0.16 Hz, **< 0.01; amplitude, control: 11.90 ± 0.45 pA, copper: 15.63 ± 0.39 pA; ***< 0.001), the time constant of AMPA events was decreased in copper-treated neurons (control: 8.99 ± 0.52 ms; copper: 5.75 ± 0.11 ms; ***< 0.001) (Fig. 5b–d). Interestingly, GABAergic neurotransmission was also affected, and changes observed were mainly restricted to changes in amplitude and time constant parameters, but not in the frequency of these synaptic events (control: 0.108 ± 0.009 Hz, copper: 0.119 ± 0.006 Hz; NS = 0.18) (Fig. 6). In this case, both amplitude (control: 13.48 ± 0.97 pA; copper: 17.65 ± 0.71 pA; **< 0.01) and time constant of GABA events (control: 21.52 ± 1.76 ms; copper: 29.94 ± 2.16 ms) were increased in copper-treated cells (Fig. 6a–c). The increase in the amplitude of GABAergic currents correlated with an increase in GABAA receptors after 3 h treatment with copper (Fig. 6d). These results indicate that changes in total synaptic activity induced by copper are because of changes in AMPAergic and GABAergic neurotransmission.

Figure 6.

 Copper does not affect the frequency of GABA currents, but increases the amplitude and decay time. GABA mini currents were isolated by pharmacology maneuvers [AP-V (5 μM); strychnine (1 μM); CNQX (4 μM)] in control neurons and copper-treated (10 μM) neurons by 3 h. Graphs summarize data obtained in these experiments for: (a) frequency (Hz); (b) amplitude (pA) and (c) decay (ms). The bars are mean ± SEM from at least six neurons (**< 0.01). (d) Confocal microphotographs of untreated (control) or copper-treated (10 μM) hippocampal neurons for 3 h, showing immunoreactivity for GluA2 (red) and GABAAα-1 (green) subunits. Right panels show the merge (yellow) of GluA2 (red) and GABAAα-1 (green) subunits immunoreactivity. Scale bar represents 25 μm.

Copper increases functional AMPA receptors

The fact that copper-treated neurons displayed an increase in amplitude of synaptic events may be as a result of the modifications at the post-synaptic level. In fact, the levels of GABAA receptors are increased in copper-treated neurons after 3 h of treatment (Fig. 6d). In the case of AMPAergic currents, amplitude was also increased after 3 h treatment with copper (Fig. 5), suggesting that a major number of receptors should be at the plasma membrane. We analyzed the localization of AMPA receptors using an antibody that detects the GluA1 subunit of these excitatory receptors. We found that GluA1 staining was significantly increased (***< 0.001) at MAP2-positive dendritic zones of copper-treated neurons (Fig. 7a and b), which seemed to be localized at the plasma membrane. Moreover, we found that another subunit of the AMPA receptor, GluA2, was also increased after 3 h of treatment with copper (Fig. 6d). To confirm that AMPA receptors were functionally active and effectively localized at the post-synaptic membrane, we performed studies of evoked currents using the AMPA ligand (Fig. 7c). These studies revealed that copper-treated neurons were more sensitive to AMPA compared with control neurons, showing a significant shift to the left in the dose–response curve. In fact, the EC50 for AMPA observed in control neurons (EC50 8.3 ± 0.5 μM) was decreased to an EC50 of 2.9 ± 0.2 μM in copper-treated (10 μM; 3 h) neurons (Fig. 7d). Furthermore, the desensitization of AMPA receptors was slower in copper-treated neurons as shown by the values for peak/plateau of the currents evoked by AMPA (control: 1.84 ± 0.58 pA; copper-treated: 1.28 ± 0.09 pA; *< 0.05), which favor a longer opening time for AMPA receptors. However, when copper was co-applied with AMPA (‘acute application of copper’), AMPA evoked currents were effectively inhibited by the presence of copper (IC50 22.9 ± 3.8 μM) (Fig. 1a and b), in agreement with previous results recorded in cortical neurons (Weiser and Wienrich 1996). All together, our data indicate that neurons behave differently to copper under acute versus prolonged incubation time, through mechanisms that may or may not be connected. For example, 3 h of copper treatment may arise from a permanent block of AMPA receptors through homeostatic or anti-homeostatic mechanisms (Carrasco et al. 2007), which then may increase these receptors at the synaptic membrane. Thus, to explore this idea using a different pharmacological approach, we incubated hippocampal cultures for 3 h with 4 μM of the AMPA blocker CNQX (See Figure S1), at a concentration that efficiently inhibits AMPA receptors (Carrasco et al. 2007). After 3 h of incubation, we did not observe any change in total synaptic activity (See Figure S1), indicating that blockade of AMPA receptors at 3 h did not induce a compensatory response at the synapses, suggesting that the effect of copper on neurotransmission after 3 h of incubation is probably not related to the acute inhibitory effect of copper on AMPA currents (Fig. 1) (Weiser and Wienrich 1996).

Figure 7.

 Membrane anchorage of functional AMPA receptors is induced by copper in hippocampal neurons. (a) Confocal microphotographs of untreated (control) or copper-treated (10 μM) hippocampal neurons for 3 h, showing immunoreactivity for GluA1 subunit (red) and MAP2 (green). Scale bar represents 20 μm. (b) The graph shows a quantitative analysis of GluA1 subunit immunoreactivity shown in panel A. The bars are mean ± SEM from at least six neurons (***< 0.001). (c) Representative traces of AMPA evoked currents on untreated neurons and copper-treated (10 μM) neurons for 3 h. (d) AMPA dose–response curves for untreated (EC50 = 8.26 ± 0.5 μM) and copper-treated (10 μM) neurons for 3 h (EC50 = 2.89 ± 0.2 μM) (< 0.001).

To evaluate if the changes observed for AMPA and GABAA receptors by immunofluorescence were because of changes in the expression of these receptors, we performed western blot analysis using anti-GluA1 and anti-GABAAα-1 subunit antibodies. Interestingly, this study showed no differences in the levels of both proteins between control and copper-treated (3 h) neurons (Fig. 8a–c), indicating that the mechanism behind the effect of copper on neurotransmission should be explained by changes in the localization of receptors. In fact, immunoflurescence studies performed in non-permeabilized neurons show the presence of large clusters of GluA1 subunits at the plasma membrane, which were undetectable in control neurons (Fig. 8d and e). The result supports the idea that copper increases AMPAergic neurotransmission by the clustering of AMPA receptors at the plasma membrane. Interestingly, and in agreement with the literature (Nose et al. 2010), we observed that the total levels of the main copper transporter located at the plasma membrane, Ctr1, were decreased (Fig. 8f and g) after 3 h of treatment with copper, indicating that this brain primary culture possesses the machinery to regulate the copper uptake at the plasma membrane. These results also indicate that total levels of AMPA and GABAA receptors are not affected by copper in the same fashion than Ctr1.

Figure 8.

 Copper increases the clustering of AMPA receptors at the plasma membrane. (a) Western blot of control and copper-treated (10 μM) neurons for 3 h, showing the levels of GluA1 and GABAAα-1 subunits. The graph shows a quantitative analysis of the immunoreactivity for (b) GluA1 and (c) GABAAα-1 subunits. (d) Confocal microphotographs of untreated (control) or copper-treated (10 μM) hippocampal neurons for 3 h, showing immunoreactivity for GluA1 subunits in non-permeabilized neurons. The arrow heads indicate small clusters, and the arrows indicate large clusters. Scale bar represents 15 μm. (e) The graph shows a quantitative analysis of the size of GluA1 clusters (μm2; *< 0.05). (f) Western blot of control (left line) and copper-treated (10 μM; right line) neurons for 3 h, showing the immunoreactivity for the copper transporter Ctr1. β-actin was used as internal control. (g) The graph shows a quantitative analysis of the immunoreactivity for Ctr1 (*< 0.05).

To analyze whether changes in AMPA receptor were accompanied with an increase in scaffolding proteins necessary for its accumulation at the synaptic membrane, we further evaluated the levels and localization of PSD95, a critical protein for AMPA receptor anchorage at the synapses (Colledge et al. 2003). Remarkably, these studies indicated that protein levels of PSD95 are indeed increased in copper-treated neurons (Fig. 9a–c), thus offering a potential explanation for the clustering of AMPA receptors in the post-synaptic membranes. Interestingly, when hippocampal neurons were treated with a copper chelator, TTM (25 μM), the levels of PSD95 were decreased (Fig. 9d), indicating the importance of copper on the regulation of PSD95 and the function of AMPA receptors.

Figure 9.

 PSD95 levels are regulated by copper. (a) Confocal microphotographs of untreated (control) or copper-treated (10 μM) neurons for 3 h, showing immunoreactivity for PSD95. Scale bar of 30 μm. (b) The graph shows a quantitative analysis of PSD95 immunoreactivity shown in panel A. The bars are mean ± SEM from at least six neurons (***< 0.001). (c) Western blot showing the increase of PSD95 levels from neurons treated with copper (up to 10 μM). α-tubulin was used as internal control. (d) Western blot showing the decrease of PSD95 levels from neurons treated with the copper chelator, TTM (25 μM).

Discussion

The importance of copper in the CNS is well documented, but the mechanisms behind its brain functions are unknown (Linder and Hazegh-Azam 1996). Brain copper deficiency is observed in Menke’s disease, which affects brain physiology, as patients display gray matter degeneration, Purkinje cell abnormalities and hippocampal neuronal loss (Okeda et al. 1991). AD is another pathology characterized by neurodegeneration that produces a broad spectrum of symptoms that have been linked to copper brain depletion as cupro-proproteins such as ceruloplasmin are decreased (Connor et al. 1993) or less active as observed by superoxide dismutase (Omar et al. 1999; Maynard et al. 2005). Currently, the effects of copper in CNS diseases are thus poorly understood.

It has been suggested that copper is released at the synaptic cleft (Hartter and Barnea 1988; Hopt et al. 2003), where it may modulate neurotransmission (Weiser and Wienrich 1996; Schlief et al. 2005, 2006). In this regard, this is the first time that copper is shown to enhance synaptic activity after 3 h of incubation (Figs 2 and 3). This effect is specific because it is not produced by 3 h applications of other transition metals, such as zinc or iron (Fig. 3). Both frequency and amplitude of total synaptic events were increased in copper-treated neurons, indicating that pre- and post-synaptic components are involved (Fig. 3b and c). For instance, a pre-synaptic mechanism is supported by the observation that the frequency of calcium transients was also augmented in copper-treated neurons (Fig. 5e and f). On the other hand, a post-synaptic mechanism is suggested by the increase in amplitude displayed in copper-treated neurons (Fig. 3c), which also correlated with an increase of GluA1 synaptic levels (Fig. 7). Interestingly, we have found that this increase was related to an increase in PSD95 (Fig. 9), indicating that AMPA receptor anchorage at post-synaptic membranes may be occurring under these experimental conditions. We also found that TTM, a copper II chelator, mirrored the effect of copper on PSD95, indicating that somehow PSD95 expression is under the regulation of copper (Fig. 9), through a direct or indirect mechanism. For example, copper could directly interact with PSD95 increasing its stability or decreasing its degradation by the proteasome (Colledge et al. 2003). In fact, PSD95 has in its structure several metal-coordinating amino acids, such as histidine and tyrosine that may be targeted by copper. Conserved histidine residues (H130 and H225) are located in the PSD95 PDZ domains, which are important for the interaction of PDS95 with stargazin (Schnell et al. 2002). Therefore, these histidines could be targeted by copper, increasing AMPA clustering driven by PSD95-stargazin interaction. However, other alternative mechanisms may explain the effect of copper on neurotransmission. For example, copper could increase phosphorylation of serine-295 of PSD95 (Kim et al. 2007), which is important for AMPAergic neurotransmission (Kim et al. 2007). Moreover, copper could regulate PSD-95 palmitoylation, which occurs on cysteine residues (Craven et al. 1999) that could also be a possible target for copper. All these possible mechanisms deserve further examination.

All together, these results suggest that neurons exposed to a copper-enriched environment display a more active synaptic transmission, which might be important in neurodegenerative diseases characterized by an imbalance in copper homeostasis (Bush 2003).

AMPA-evoked currents were significantly increased in copper-treated neurons, showing a ligand shift to the left reflected by a significant change in EC50 (control: 8.3 ± 0.5 μM; copper-treated: 2.9 ± 0.2 μM, < 0.001), indicating that these neurons have more AMPA receptors at the plasma membranes as it is suggested by the immunofluorescence analysis and AMPA-evoked currents (Fig. 7). Moreover, desensitization was also decreased in these neurons keeping AMPA channels open for longer times (Fig. 7c). However, we cannot discard that other auxiliary proteins, such as TARP, may be involved in this effect (Farrant and Cull-Candy 2010).

Finally, copper also increased the amplitude of the GABAergic transmission but without affecting frequency of these synaptic events (Fig. 6), indicating only a post-synaptic mechanism probably involving changes in receptor pharmacology, as time constant of GABA events was increased in copper-treated neurons by a mechanism that was not explored in the present work, which may involve homeostatic or anti-homeostatic mechanisms (Carrasco et al. 2007).

The molecular mechanism of copper (3 h) in neurotransmission seemed to be unrelated to a compensatory action resulting from the blockade of AMPA receptor, which was able to inhibit AMPA evoked currents in hippocampal neurons when copper was acutely applied to neurons (Fig. 1a and b), as it has been previously described in cultured rat cortical neurons (Weiser and Wienrich 1996). In this regard, we examined the total synaptic events in neurons after blockade of AMPA receptors for 3 h with CNQX, which is a specific and potent antagonist of AMPA currents. This pharmacological maneuver did not affect any parameter of the total mini currents, indicating that at this time frame a compensatory mechanism is not responsible for receptor blockade (See Figure S1). Therefore, we speculate that the mechanism behind the effect of copper on this neuronal network involves intracellular changes independent of AMPA receptor blockade. Interestingly, after 24 h exposure of copper, the synaptic activity returned to the control levels, indicating that changes in neurotransmission observed at 3 h are under homeostatic regulation.

In summary, in this work, we have shown that copper, in a time and concentration dependent manner, may induce different effects on neurotransmission, indicating that a fine-tuning of this essential metal is needed by neuronal cells to maintain adequate synaptic function. In fact, an increase in the brain concentration of copper, as well as a decrease in the levels of this metal, can lead to serious illness (Bush 2003). Therefore, future studies are needed to better understand the mechanisms behind copper effects on living neurons, which may lead to prevent or to treat copper-related neurodegenerative diseases and neurological conditions, which remain untreatable to date.

Acknowledgements

We would like to thank the members of the laboratory, especially Claudia Lopez and Laurie Aguayo for their technical assistance. This work was supported by the Chilean Government (Anillo-PBCT ACT-04 project to LGA, GVD and CO), FONDECYT 1100502 (LGA), FONDECYT 1100942 (GDV) and DIUC grant Nº 205.033.101-1.0 (CO). CP and FJS are recipients of CONICYT fellowships. We thank Professor Dennis Thiele and Dr Yasuhiro Nose for providing us anti-Ctr1 antibody. We are grateful to the “Centro de Microscopía Avanzada (CMA) Bío Bío for technical assistance in confocal microscopy. The authors declare no conflict of interest.

Ancillary