Protective effects of voltage-gated calcium channel antagonists against zinc toxicity in SN56 neuroblastoma cholinergic cells

One of the pathological site effects in excitotoxic activation is Zn2+ overload to postsynaptic neurons. Such an effect is considered to be equivalent to the glutamate component of excitotoxicity. Excessive uptake of Zn2+ by active voltage-dependent transport systems in these neurons may lead to significant neurotoxicity. The aim of this study was to investigate whether and which antagonists of the voltage gated calcium channels (VGCC) might modify this Zn2+-induced neurotoxicity in neuronal cells. Our data demonstrates that depolarized SN56 neuronal cells may take up large amounts of Zn2+ and store these in cytoplasmic and mitochondrial sub-fractions. The mitochondrial Zn2+ excess suppressed pyruvate uptake and oxidation. Such suppression was caused by inhibition of pyruvate dehydrogenase complex, aconitase and NADP-isocitrate dehydrogenase activities, resulting in the yielding of acetyl-CoA and ATP shortages. Moreover, incoming Zn2+ increased both oxidized glutathione and malondialdehyde levels, known parameters of oxidative stress. In depolarized SN56 cells, nifedipine treatment (L-type VGCC antagonist) reduced Zn2+ uptake and oxidative stress. The treatment applied prevented the activities of PDHC, aconitase and NADP-IDH enzymes, and also yielded the maintenance of acetyl-CoA and ATP levels. Apart from suppression of oxidative stress, N- and P/Q-type VGCCs presented a similar, but weaker protective influence. In conclusion, our data shows that in the course of excitotoxity, impairment to calcium homeostasis is tightly linked with an excessive neuronal Zn2+ uptake. Hence, the VGCCs types L, N and P/Q share responsibility for neuronal Zn2+ overload followed by significant energy-dependent neurotoxicity. Moreover, Zn2+ affects the target tricarboxylic acid cycle enzymes, yields acetyl-CoA and energy deficits as well.

Therefore, the overall goal of this work was to investigate whether and which antagonists of the VGCCs might modify Zn 2+ -induced neurotoxicity in the neuronal cells. Thereby, such information would reveal the contribution of individual VGCC to the early stages of Zn 2+ -evoked excitotoxic neuronal injury.

Materials
Specified reagents and cell culture growth media were obtained from Sigma Aldrich (Poznan, Poland). Antibodies were derived from ABCAM (Straszyn, Poland). TSQ and Western Blot components were derived from Thermofisher (Warszawa, Poland). Cell culture disposables were purchased from Sarstedt (Stare Babice, Poland).

Cell cultures
SN56 cells (neuroblastoma cell line SN56.B5.G4) were seeded with a density 40 000 cells/cm 2 and cultured for 48 h in Dulbecco's modified Eagle medium containing 2 mM L-glutamine, 0.05 mg of streptomycin and 50 U of penicillin per 1 mL and 10% fetal bovine serum (FBS). For chronic Zn 2+ toxicity and VGCCs density studies (Fig 1), SN56 cells were cultured for an additional 24h in fresh FBS-DMEM media with or without 0.15 mmol/L Zn 2+ . Such Zn 2+ concentration was consistent with cytotoxic Zn 2+ concentrations used in our previous projects [7,[12][13].
A growing body of evidence emphasizes that neurons might actively remove Zn 2+ ions from their cytoplasm [26][27]. In vitro studies showed that such a process requires the physiological concertation of sodium ion concentration and seems to be the most effective after 30 min of in vitro incubation [27]. Since one the main goal of acute studies was to determinate the highest capacity of SN56 cells for uptake Zn 2+ , thus cells were placed for 30 min in depolarizing media containing 30 mM K + .
For Zn 2+ , Ca 2+ , pyruvate, lactate and acetyl-CoA assays, the medium was centrifuged and cell pellets were deproteinized with 4% HClO 4 and neutralized with 7.5 N K 2 CO 3 . For enzyme activity studies, cell pellets were re-suspended in 320 mmol/L sucrose and lysed with 0.2%.

Cell viability assay
Cells viability was assessed with trypan blue exclusion assay as described elsewhere [12][13]. In brief, 2.5 mln cells/1mL were dyed by 0.2% Trypan Blue solution (10 min, room temperature). Next, 20 μL of cell suspension was analyzed by counting chamber under light microscope (Axiovert 25, Zeiss). Non-viable cell was identified as unable to remove applied dye.

Pyruvate and lactate assays
Rates of pyruvate utilization and lactate accumulation were calculated from differences in metabolite contents in the whole medium at the times zero and after 30 min of incubation [35]. Lactate and pyruvate levels were determined using NADH/NAD conversion technique, at 340 nm and 37˚C. Both assays were initiated by the 10 μL addition of 4 U lactate dehydrogenase (EC 1.1.1.27).
1 mL of reaction buffer for lactate assay contained 70 mM glycine-NaOH buffer (pH = 10), 5 mM NAD and 100 μg of cell homogenate protein in a final volume of 0.7 mL.

TBARS assay
The lipid peroxidation products were estimated in the whole medium as thiobarbituric acid reactive substances (TBARS) at 535 nm in room temperature [38]. Briefly, 0.5 mg of cell homogenate protein was deproteinized by 10% trichloroacetic acid in a final volume of 0.6 mL (10 min, 4˚C, gentle shaking). Next, each sample was enriched by 0.2 mL of 2% thiobarbituric acid and heated for 20 min in 100˚C.

Western Blot analysis
SN56 cells were lysed with a RIPA buffer (50 mmol/L TRIS-HCl buffer (pH 7.4), 5 mmol/L EDTA, 100 mmol/L NaCl, 1% Triton-X100, 5% glycerol, 10 mmol/L KH 2 PO 4 ) supplemented with a protease inhibitor cocktail. Supernatants were boiled for 5 min with a 2.5% β-mercaptoethanol/Laemmli buffer and then loaded on 4-20% gradient SDS-PAGE gels. The obtained proteins were transferred to a PVDF membrane. Next, the membrane was incubated with primary antibodies in a WB buffer (5% bovine serum albumin/0.5% Tween20/ 25 mmol/L Trisbuffered saline, pH 7.4) at 4˚C, overnight. The following day, a after 3-time washing step, the PVDF membrane was incubated for 3h with an alkaline phosphatase-conjugated secondary antibody at room temperature (in a WB buffer). The PVDF membrane was developed using 5-bromo-4-chloro-3-indoylphosphate with Nitro blue Tetrazolium chloride. β-actin was used as a reference protein.

Protein assay
Protein levels were assayed using the Bradford method [39] with human immunoglobulin as a standard.

Statistics
Statistical differences between the two groups were assessed with the Mann-Whitney U test and for multiple comparisons the Kruskal-Wallis test with Dunn's post hoc test was used. Values of p<0.05 were considered statistically significant. The data presented are means ±SEM from 4 to 25 experiments.

Effect of Zn 2+ on voltage gated calcium channels in SN56 cells
Excessive release of Zn 2+ from pathologically depolarized glutaminergic terminals was identified as an important factor triggering early neurodegenerative events in postsynaptic neurons [11,23,40]. Our past studies on SN56 cells revealed that cholinotoxicity may be partially mediated by depolarization-dependent Zn 2+ uptake [7,[12][13]. However, no specific channel type has been identified yet. Here, we demonstrate, using specific antibodies, that the plasma membranes of SN56 cells contain the L, P/Q and N types of VGCCs (Fig 1A-1C). Our double staining with DAPI and specific antibody against L-VGCC, P/Q-VGCC as well as N-VGCC revealed that all viable SN56 cells were positive for studied channels (Fig 1A-1C).
One-day treatment of SN56 cells with 0.15 mmol/L Zn 2+ , in a non-depolarizing culture medium, reduced both cell viability and total protein by about 50% (Fig 1J and 1K). However, in cells which were subjected to this treatment no changes in the levels of any of the three VGCC proteins took place (Fig 1D-1I).

Intraneuronal Zn 2+ accumulation
The control level of intracellular Zn 2+ in SN56 cells incubated for 30 min in an FBS free and Zn 2+ -free depolarizing medium was equal to 0.6-0.8 nmol/mg protein (Fig 2A, Table 1). In these conditions, the addition of 0.10 and 0.15 mmol/L Zn 2+ increased the intracellular content of Zn 2+ to about 8 and 34 nmol/mg protein, respectively (Fig 2A). Furthermore, the acute accumulation of Zn 2+ in SN56 cells was 2.6 times higher than that found in those cells treated with Zn 2+ chronically in a non-depolarizing medium containing 10% FBS (Table 1). However, irrespective of the treatment conditions, the general patterns of Zn 2+ subcellular distribution were similar. Particularly, in SN56 cells, we observed minor mitochondrial and larger cytoplasmic pools, containing 12-20 and 88-80% of the whole cell content, respectively ( Table 1). The short-term treatment of depolarized SN56 cells with 0.15 mmol/L extracellular [Zn 2+ ], was accompanied by a relatively smaller, 30% elevation of intracellular Ca 2+ levels ( Table 1).

Effects of Zn 2+ uptake on energy metabolism in SN56 cells
The concentration-dependent Zn 2+ accumulation in SN56 cells brought about, progressing up to three-fold, an increase of TBARS synthesis, a 4.5-fold rise of GSSG level, and a 70% decrease of the GSH/GSSG ratio (Fig 2A and 2B) These changes were accompanied by dose-dependent inhibition of pyruvate utilization by SN56 cells, from 6.4 in controls to 3.4 nmol/min/mg protein at 0.15 mmol/L extracellular Zn 2+ (Figs 2C, 2E and 3C). Meanwhile, lactate accumulation decreased from 2.1 to 1.2 nmol/min/mg protein (Fig 2C). The rate of pyruvate oxidation, calculated as the difference between pyruvate uptake and lactate formation, decreased from 4.1 to 0.4 (Fig 2C). Zn 2+ [IC 50 ] for pyruvate utilization was equal to 0.082 mmol/L, which was nearly two times higher than [IC 50 ] for pyruvate oxidation (Fig 3D). On the other hand, PDHC activity was inhibited only by 25% and revealed no dependence on metal concentration (Fig 2D  and 2E).
The activities of aconitase and IDH-NADP were inhibited by accumulating Zn 2+ in a concentration-dependent manner (Fig 2D and 2F, Table 2). Zn 2+ [IC 50 ] for IDH-NADP was equal to 0.015 mmol/L and appeared to be 9 times lower than that of aconitase (Fig 2D and 2F). In the same conditions, a 55% reduction of the ATP level at a 50% elevation of the ADP level caused a respective decline of the ATP/ADP ratio (Fig 3A, 3C and 3D, Table 2). These shortterm alterations did not change the total cell number, but decreased the viable trypan blue excreting cell fraction by 40% (Fig 3B, Table 2).  Significant inverse relationships were found between an extracellular Zn 2+ concentration and the rates of: pyruvate uptake and oxidation, aconitase and IDH-NADP activities, as well as ATP levels in SN56 cells (Fig 3C). On the other hand, the Zn 2+ -evoked decrease of viable cell fraction displayed positive correlations with the declining values of metabolic parameters quoted above (Fig 3D).

Voltage-gated calcium channel blockers and Zn 2+ cytotoxicity
The saturating concentrations of voltage-gated calcium channel (VGCC) antagonists were determined experimentally (S1 Fig). In a Zn 2+ -free medium none of the antagonists (0.01 mmol/L nifedipine, 0.0005 mmol/L ω-conotoxin GVIA and 0.0002 mmol/L ω-conotoxin MVIIC) altered the low endogenous level of intracellular Zn 2+ (Table 2). In the absence of Zn 2+ , these Ca 2+ -antagonists altered neither lipid peroxidation, nor the fraction of non-viable SN56 cells, respectively (Table 2). Furthermore, no significant effects of the antagonists on acetyl-CoA and ATP levels or aconitase and IDH-NADP activities were detected ( Table 2). Nifedipine and GVIA decreased cellular Ca 2+ levels by about 33 and 18%, respectively. The MVIIC exerted no significant effect on intracellular Ca 2+ content ( Table 2). In a medium containing 0.15 mmol/L Zn 2+ , each of those VGCC antagonists reduced its influx into SN56 cells by 45-60% ( Table 2). All VGCCs also prevented Zn 2+ -induced excessive accumulation of Ca 2+ and partially prevented aconitase inhibition ( Table 2).
Despite these similarities, nifedipine exerted the strongest protective effects against Zn 2+ -toxicity, reducing the non-viable fraction from 40 to 17% (Fig 4E and 4F, Table 2). In the presence of nifedipine, Zn 2+ treatment did not promote TBARS synthesis at fully preserved acetyl-CoA content and partially protected the ATP level as well (Fig 4A and  4B, Table 2). Nifedipine also alleviated the inhibitory effects of Zn 2+ treatment on both aconitase and IDH-NADP activities (Fig 4C and 4D, Table 2). Our studies showed that both ω-conotoxins (GVIA and MVIIC) appeared to be weaker neuronal protectants than nifedipine. They reduced the fraction of trypan blue positive SN56 cells to 31 and 34%, respectively (Fig 4E and 4F, Table 2). However, these antagonists did not reduce the excessive synthesis of TBARS and did not augment the inhibition of IDH-NADP  Table 2. https://doi.org/10.1371/journal.pone.0209363.g004 Role of calcium channel antagonists in Zn 2+ neurotoxicity activity. GVIA and MVIIC were not fully effective in the maintenance of acetyl-CoA level, increasing it from 45 to 69% of control levels (Fig 4A and 4B, Table 2). These VGCC-induced alterations in reducing inhibitory Zn 2+ effects on enzyme activities and metabolite levels in SN56 cells correlated significantly with their capacities for reducing intracellular Zn 2+ overload or/and TBARS accumulation (Fig 4A-4F).

Discussion
It has been demonstrated that Zn 2+ accumulation in depolarized postsynaptic neurons via diverse transporting systems exerts a negative effect on a several intracellular targets (Fig 5). Its excessive uptake by neurons yields functional and structural impairment, which eventually leads to their death [4-5, 11, 24]. In our cholinergic cellular model of Zn 2+ neurotoxicity we used pathologically relevant concentrations of Zn 2+ (0.05-0.20 mmol/L), which correspond well with assessed cation levels in the synaptic cleft and whole brain tissue as well (Figs 2 and  3) [41][42]. The basal Zn 2+ content in SN56 cell was equal 0.6 nmol/mg protein (0.08 mmol/L of cell water), which is comparable with 0.1-0.2 mmol/L cation levels reported for the whole brain or primary neurons (Tables 1 and 2) [43][44][45][46].
It is known that postsynaptic cells take up zinc ions via different Zn 2+ -and Ca 2+ -specific channels, although their participation in Zn 2+ transport is mainly related to neurons [47]. Our immunostainings confirmed the existence of the main neuron-specific calcium channel classes on the surface of clonal cholinergic SN56 cells (Fig 1A-1I). These findings are compatible with patch camp studies revealing the functional existence of these VGCCs in SN56 cells [25]. Thereby, mechanisms of Zn 2+ overload presented by using the clonal SN56 neuronal cells may be compatible with those reported for in vivo and in vitro models of brain excitotoxicityrelated pathologies [48][49]. Therefore, we conclude that neither short-nor long-term increases of Zn 2+ at gluzinergic synapses trigger mechanisms of VGCCs down regulation in postsynaptic neurons (Fig 1A-1I). These observations point out that cholinergic neurons may not suppress VGCC expression as a protective mechanism against prolonged increases of Zn 2+ in the synaptic cleft (Fig 1D-1I). Furthermore, accumulation of Zn 2+ by chronically exposed SN56 cells in a non-depolarizing medium was about 3 times lower than in those exposed to Zn 2+ for 30 min in a depolarizing medium (Table 1). These differences in Zn 2+ levels may result from a depolarization-dependent increase of VGCC permeability for Zn 2+ [47,[49][50][51]. Relatively low levels of Zn 2+ accumulation in mitochondria may indicate that these organelles possesses an effective system of Zn 2+ excretion to the cytoplasm or/and relevant mechanisms preventing the fast inward transport of Zn 2+ from the cytoplasm (Table 1) [49,[52][53]. These findings remain in accord with primary neuron studies showing slow mitochondrial uptake of Zn 2+ via Ca 2+ -uniporters or other mitochondrial zinc-specific transporters [49,[52][53]. Nevertheless, this relatively small mitochondrial Zn 2+ accumulation significantly affected SN56 cell oxidative metabolism (Figs 2C-2F and 3) [51,[54][55][56]. On the other hand, the activities of cytoplasmic enzymes, such as ATP-citrate lyase or choline acetyltransferase, lactate dehydrogenase, autotaxin and caspase-3, appeared to be resistant to acute influences of high Zn 2+ concentrations [12][13][57][58]. Therefore, changes triggered by Zn 2+ excess in cytoplasmic acetyl-CoA levels should be considered secondary to alterations in its mitochondrial availability (Fig 4A  and 4B) [12][13].
The presented data demonstrate that Zn 2+ might trigger Ca 2+ overload in SN56 cells (Table 2), which may result from ATP deficits yielding a substrate concentration-dependent decrease of Ca 2+ -ATP-ase activity and augmentation of free radical-evoked impairment of cell viability (Table 2) [10,49,59]. Also, the prolonged cell depolarization associated with such conditions may contribute to Ca 2+ overload due to increased VGCC permeability (Table 2) [16,19]. In fact, Zn 2+ caused upregulation of T-type VGCC in hippocampal neurons in vivo [50]. Zn 2+ might also facilitate Ca 2+ entry through activation of the transient receptor potential channel A1 3+ [60].
Our past data revealed that Zn 2+ reversibly inhibits PDHC activity in cell homogenates due to the removal of lipoamide from E2 subunit binding sites [12][13]. Here, we investigated the pyruvate uptake/utilization in whole, not lysed, SN56 cells. Relatively strong inhibition of pyruvate oxidation in situ, triggered by Zn 2+ overload, indicates that its "free"-ion levels reached sufficient concentrations to inhibit PDHC in a similar manner as in homogenates (Fig  2A, 2C and 2D) [13]. Such conditions did not inhibit PDHC, as demonstrated here by unaltered complex activity measured in cell homogenates with saturating concentrations of its Role of calcium channel antagonists in Zn 2+ neurotoxicity substrates and cofactors (Fig 2D and 2E) [12][13]. This indicates that the short-term toxic effects of Zn 2+ on mitochondrial metabolism of pyruvate in the brain may be partially or fully reversible (Fig 2C and 2E) [6,12,20,61]. Accordingly, an excess of intramitochondrial Zn 2+ in vivo may be the single factor causing transient pyruvate derived acetyl-CoA deficits in excitotoxicity-affected neurons (Figs 2C-2E and 4,Tables 1-2).
It has been shown that the inhibitory effects of Zn 2+ on aconitase and IDH-NADP activities in SN56 homogenates were irreversible [12][13]. Aconitase inhibition was caused by the removal of non-hem iron ions from Fe-S clusters in the enzyme active center [12][13]62]. The existence of such mechanism in cells in situ is confirmed here, by finding that each of the VGCC antagonist, partially alleviated inhibition of aconitase activity by limiting Zn 2+ uptake ( Fig 4C, Table 2). On the other hand, Zn 2+ -evoked inhibition of IDH-NADP activity was suggested as being caused by its competitive interaction with magnesium cations in the enzyme active center [63]. However, our results suggest that in situ inhibition of IDH-NADP by accumulated Zn 2+ was overcome only by nifedipine. Nifedipine is the strongest L-type VGCC antagonist with pleiotropic, antioxidative properties preventing TBARS overproduction ( Fig  4B and 4D-4F, Table 2) [64][65]. On the other hand, ω-conotoxins, which were devoid of such bio-activity, failed to alleviate Zn 2+ evoked IDH-NADP inhibition (Fig 4A-4F, Table 2). ( Fig  4A-4F, Table 2). This indicates that Zn 2+ -induced inhibition of IDH-NADP activity was not triggered directly by Zn 2+ itself, but by its promotion of oxidative stress ( Table 2, Fig 4D) [9,[66][67]. These suggestions are confirmed here by (i) the existence of an inverse correlation between IDH-NADP activity and alterations in TBARS accumulation, as well as by a lack of correlation with intracellular Zn 2+ load (Fig 4D), (ii) Zn 2+ [IC 50 ] against IDH-NADP, being 9 times higher than against aconitase (Fig 2D and 2F), (iii) the existence of an inverse correlation between aconitase activity and Zn 2+ accumulation in SN56 cells and the absence thereof with TBARS accumulation, respectively (Fig 4B and 4C).
Thereby, the potencies of specific VGCC inhibitors in alleviating the cytotoxic effects of Zn 2+ may reflect the potential, fractional contribution of each channel to Zn 2+ neuropathology ( Table 2) [21]. These protective effects were demonstrated by the capacity of particular VGCCs to alleviate Zn 2+ -evoked inhibition of PDHC, aconitase and IDH-NADP activities ( Table 2, Fig 4A-4F). The respective significant correlation plots support the assumption that the neuroprotective effects of VGCCs antagonists were predominately linked with an alleviation of intracellular Zn 2+ overload and/or TBARS overproduction ( Table 2, Fig 4A-4F).
The presented data demonstrate that VGGCs facilitate Zn 2+ and Ca 2+ entry to chronically depolarized neurons. An excess of Zn 2+ in their intracellular compartments: (i) inhibits PDHC limiting provision of acetyl-CoA for the TCA cycle in their mitochondria (ii) obstructs its metabolic flow through the TCA cycle inhibiting aconitase and IDH-NADP (iii) yields suppression of ATP production activating free radical synthesis and GSH oxidation (iv) that results in loss of neuronal viability (Table 2, Figs 3A, 3C and 4A-4F) [68]. These data indicate that Zn 2+ overload through VGGs may be the earliest neurotoxic signal affecting multiple metabolic steps crucial for neuronal functions and survival. Rapid reversal of these effects through Zn 2+ removal may be important for cell recuperation during recovery after excitotoxic-events.