THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 280, NO. 37, pp. 32405–32412, September 16, 2005 © 2005 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

Investigating the Receptor-independent Neuroprotective Mechanisms of Nicotine in Mitochondria* Received for publication, April 28, 2005, and in revised form, June 27, 2005 Published, JBC Papers in Press, June 27, 2005, DOI 10.1074/jbc.M504664200

Yu-Xiang Xie‡§, Erwan Bezard¶, and Bao-Lu Zhao‡1 From the ‡State Key Laboratory of Brain and Cognitive Science, Institute of Biophysics, Academia Sinica, Beijing 100101, China, ¶ CNRS UMR5543, University Victor Segalen-Bordeaux 2, 146 Rue Leo Saignat, 33076 Bordeaux, France, and §Graduate School of Chinese Academy of Sciences, Beijing, 100049, China Although nicotine has been associated with a decreased risk of developing Parkinson disease, the underlying mechanisms are still unclear. By using isolated brain mitochondria, we found that nicotine inhibited N-methyl-4-phenylpyridine (MPPⴙ) and calcium-induced mitochondria high amplitude swelling and cytochrome c release from intact mitochondria. Intra-mitochondria redox state was also maintained by nicotine, which could be attributed to an attenuation of mitochondria permeability transition. Further investigation revealed that nicotine did not prevent MPPⴙ- or calciuminduced mitochondria membrane potential loss, but instead decreased the electron leak at the site of respiratory chain complex I. In the presence of mecamylamine hydrochloride, a nonselective nicotinic acetylcholine receptor inhibitor, nicotine significantly postponed mitochondria swelling and cytochrome c release induced by a mixture of neurotoxins (MPPⴙ and 6-hydroxydopamine) in SH-SY5Y cells, suggesting that there is a receptor-independent nicotine-mediated neuroprotective effect of nicotine. These results show that interaction of nicotine with mitochondria respiratory chain together with its antioxidant effects should be considered in the neuroprotective effects of nicotine.

Nicotine intake, mostly through cigarette smoking, has been associated with a decreased risk of developing Parkinson disease (PD)2 (1, 2), a common neurodegenerative disorder caused by the death of the dopamine neurons in the substantia nigra pars compacta that project to the striatum (3). Although the mechanisms of neuroprotection by nicotine are not fully understood, and there are still some conflicting reports (4, 5), it is accepted that nicotine usually acts as a nicotinic acetylcholine receptor (nAChRs) agonist. Although nigrostriatal damage reduces expression of specific nAChRs subtypes (6), stimulation of these receptors has been proved to be neuroprotective (7). Besides its agonistic properties, nicotine has several receptor independent effects whereby it could protect neurons against toxins (8). Linert et al. (9) found that nicotine could strongly affect the course of the Fenton reac-

* This work was supported by a grant from the National Natural Science Foundation of China. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 To whom correspondence should be addressed: State Key Laboratory of Brain and Cognitive Science, Institute of Biophysics, Academia Sinica, Beijing 100101, People’s Republic of China. Tel.: 86-1-64888569; Fax: 86-1-64871293; E-mail: zhaobl@ sun5.ibp.ac.cn. 2 The abbreviations used are: PD, Parkinson disease; 6-OHDA, 6-hydroxydopamine; CsA, cyclosporin A; GFP, green fluorescence protein; lucigenin, bis-N-methylacridinium nitrate; luminol, 3-aminophthalhydrazide; MPP⫹, N-methyl-4-phenylpyridine; MPTP, N-methyl-1,2,3,6-tetrahydropyridine; mPT, mitochondria permeability transition; mPTP, mitochondria permeability transition pore; PTP, permeability transition pore; nAChRs, nicotinic acetylcholine receptors; NQR, NADH-ubiquinone reductase; PN, pyridine nucleotides; Rh123, rhodamine 123; ROS, reactive oxygen species; SOD, superoxide dismutase; SMF, submitochondria fragment.

SEPTEMBER 16, 2005 • VOLUME 280 • NUMBER 37

tion by inhibiting the oxidation of 6-OHDA in the presence of Fe(II, III) ions. Our previous work also indicated that nicotine acts as an effective reactive oxygen species (ROS) scavenger in the gas phase of cigarette smoking (10). Recently, Cormier et al. (11) showed that nicotine was an effective NADH competitor that inhibited the mitochondria NADHubiquinone reductase (NQR) activity and significantly decreased brain mitochondrial respiratory control ratio. They also found that nicotine significantly decreased superoxide anion generation in brain mitochondria, suggesting a possible direct role of nicotine on mitochondria respiratory chain independent of its receptor. Complex I inhibition or deficiency is thought to play a major role in PD etiology (12–14), and nicotine was reported to protect neurons against pro-parkinsonian neurotoxins such as N-methyl-4-phenylpyridinium ion (MPP⫹), rotenone, and 6-hydroxydopamine (6-OHDA) (15, 16), which are all complex I inhibitors. In addition, MPP⫹ loading triggers the release of cytochrome c from mitochondria by inducing an apoptosis cascade that ultimately leads to programmed cell death, a phenomenon that involves the mitochondria permeability transition (mPT) (17). Collective evidence suggests an intrinsic relationship between mitochondria respiratory chain and mitochondria permeability transition pore (mPTP) (18 –21), in particular at the site of mitochondria complex I (22–27). We hypothesize that nicotine could be neuroprotective without the need of nAChRs by directly interacting with mitochondria. Here we examined the effects of nicotine and MPP⫹ on mPT and cytochrome c release to illustrate a possible involvement of mitochondria in mediating the neuroprotective effect of nicotine. Mitochondria substrates for NQR were used to analyze the specific involvement of complex I. The Ca2⫹-induced mPT was also examined to display a universal effect of nicotine on mitochondria permeability transition. To illustrate if nicotine can penetrate the cell membrane and take effect independent of its receptor, mecamylamine, a nonselective nicotinic acetylcholine receptor inhibitor, was applied in the SY5Y cell line. A plasmid containing cytochrome c tagged with green fluorescence protein (c-GFP) was transferred to the cell to trace cytochrome c under the confocal microscope, and the photogrammetry on mitochondria swelling was also taken in the presence of the fluorescent mitochondria red dye called Mitotracker Red CMXRos.

MATERIALS AND METHODS Chemicals—(⫺)-Nicotine was obtained from Zhengzhou Tobacco Academy of China (Zhengzhou, China). Cytochrome c release apoptosis assay kit with mouse monoclonal cytochrome c antibody (reacts with denatured human, mouse, and rat cytochrome c) was purchased from Oncogene (San Diego); ␤-actin antibody was from Santa Cruz Biotechnology. Cytochrome c-GFP plasmid was kindly provided by Prof. D. C. Chang (Hong Kong University of Science and Technology). Dulbecco’s modified Eagle’s medium, fetal calf serum, and Lipofectamine 2000

JOURNAL OF BIOLOGICAL CHEMISTRY

32405

Nicotine Protects Mitochondria from Neurotoxic Insult were purchased from Invitrogen; luminol (3-aminophthalhydrazide) was from Arcos Organics (Geel, Belgium); lucigenin (bis-N-methylacridinium nitrate), horseradish peroxidase, catalase, and superoxide dismutase (SOD) were from bovine erythrocytes; mecamylamine hydrochloride, Mitotracker Red CMXRos, Ellman’s reagent, bovine serum albumin, MPP⫹, CsA (cyclosporin A), glutamate, succinate, mannitol, and EDTA were from Sigma; malate and NADH were from Amresco (Solon, OH); and Percoll was from Amersham Biosciences. Cell Treatments and Confocal Microscope Photogrammetry—The photogrammetry measurement were taken according to Gao et al. (28) with some modifications. Briefly, human SH-SY5Y cells were maintained in Dulbecco’s modified Eagle’s medium supplemented with 15% fetal calf serum plus 100 mg/ml penicillin and 100 units/ml streptomycin. Cells were plated onto glass bottom dishes and grown to about 75% consistency at 37 °C and 5% CO2. To image cytochrome c, cells were transfected with the cytochrome c-GFP plasmid using Lipofectamine 2000 and were allowed to express the fusion gene for 48 –72 h. To stain mitochondria, Mitotracker Red CMXRos was added to a final concentration of 20 nM. For living cell measurements, we used a laser-scanning confocal microscope (Olympus FV500, Tokyo, Japan) to image cytochrome c-GFP distribution and mitochondria morphology simultaneously. Cytochrome c-GFP and Mitotracker were observed at 488 and 568 nm, respectively. After preincubation with Mitotracker Red CMXRos and mecamymine for 15 min, the cells were washed with Dulbecco’s modified Eagle’s medium and placed with 50 ␮l of fresh culture medium containing mecamylamine. The dish was fixed on the microscope platform, and a single cell was selected (t0), and then 950 ␮l of culture medium containing the given chemicals was gently added. The photogrammetry started immediately (t1), and pictures were taken in 5-min intervals. Mitochondria Preparation—Mitochondria were extracted as described by Cormier et al. (11) with some modifications. Briefly, male Sprague-Dawley rats (weighing 200 –250 g) fasting overnight were killed by decapitation; the forebrains were removed and homogenized (6 ml/g of tissue) in ice-cold isolation buffer (20 mM Tris-HCl, 250 mM sucrose, 40 mM KCl, 2 mM EGTA, and 1 mg/ml bovine serum albumin, pH 7.2, at 4 °C) using a Potter-Elvehjem homogenizer. Mitochondria isolation was immediately performed at 4 °C by using differential centrifugation. The homogenate was centrifuged at 2,000 ⫻ g for 8 min to remove cell debris and nuclei. The pellet was discarded, and the supernatant was further centrifuged at 12,000 ⫻ g for 11 min. The pellet was washed and resuspended in 320 mM sucrose, 10 mM Tris base, pH 7.4. Mitochondria kept high respiratory control ratio for the further experiments performed in the next 4 h. To obtain intact mitochondria, Percoll gradient centrifugation was adopted as described by others (29). Rough mitochondria pellets were obtained as described above, and then the pellets were resuspended in 25 ml of 15% Percoll, and 3-ml fractions of this suspension were laid on two preformed layers consisting of 3.5 ml of 23% Percoll and 3.5 ml of 40% Percoll. The gradient was centrifuged for 5 min at 30,700 ⫻ g. The fraction accumulated at the interface of the two lower layers was collected and slowly diluted 1:4 with isolation buffer. The mixture was centrifuged twice at 12,000 ⫻ g for 11 min, producing a pellet that was resuspended in 300 ml of respiratory buffer at 4 °C. To get the submitochondria fragment (SMF), crude mitochondria were freeze-thawed three times, washed with 10 volumes of isolation medium, and centrifuged at 12,000 ⫻ g for another 10 min to collect the SMF pellets (30). Protein concentration was determined by the method of Lowry, with bovine serum albumin in storage buffer as control. Western Blotting Analysis of Cytochrome c Release from Intact Mitochondria and in Cell Culture—Intact mitochondria (1 mg/ml) from

32406 JOURNAL OF BIOLOGICAL CHEMISTRY

Percoll gradient centrifugation were immediately incubated in respiratory buffer (300 mM mannitol, 10 mM KH2PO4, 10 mM KCl, 5 mM MgCl2, pH 7.2) with and without the tested molecule at a given concentration (see Fig. 2) for 30 min at 25 °C, for a final volume of 1 ml. In these reactions, 5 mM glutamate and 2.5 mM malate were used as respiratory substrates for mitochondria complex I, whereas 5 mM succinate (plus 1 ␮M rotenone) was used as the respiratory substrate for mitochondria complex II. After centrifugation at 10,000 ⫻ g for 5 min, the clear supernatants were collected. 15 ␮g of protein equivalent for each sample was loaded in 12% SDS-polyacrylamide gel and then standard Western blot proceeded as instructed by manufacturer of the cytochrome c release apoptosis assay kit. In cell experiment, apoptosis was induced in cell cultures (about 5 ⫻ 107) by 2 mM MPP⫹ administration for 24 h. 1–10 ␮M nicotine or 10 ␮M CsA was used in the presence of 20 ␮M mecamylamine. To detect cytosol cytochrome c, cells were washed and collected by centrifugation at 600 ⫻ g for 5 min at 4 °C. The cytochrome c was separated by stroke and centrifugation according to the manufacturer’s instructions. Mitochondrial Swelling and Membrane Potential Assay—Mitochondria swelling was monitored via the decrease in absorbance at 540 nm in a Hitachi U-2010 ultraviolet spectrophotometer (Hitachi, Tokyo, Japan) equipped with thermostatic control. The system is the same as the standard incubation scheme described above, and chemical concentrations are indicated in the figure legends. In most cases, mitochondria were preincubated with nicotine (or CsA) alone or in the presence of low dose Ca2⫹ (25 ␮M) at 25 °C for 5 min. Measurement started immediately when MPP⫹ was added at t1, and 500 ␮M phosphate was added 2 min after that (t2). In those conditions, i.e. when the effects of high dose of calcium plus phosphate were tested, we only show the effect of nicotine at 100 ␮M. The change in mitochondria membrane potential was assayed by measuring the retention of rhodamine 123 (Rh123, 200 nM) under the same experimental conditions as described in swelling assay. Rh123specific fluorescence intensity was monitored at an excitation wavelength of 500 nm and an emission wavelength of 530 nm using a Hitachi F-4500 fluorescence spectrophotometer (Hitachi, Tokyo, Japan). The base lines were measured for 2 min, and chemicals were then added as described in the figure legends. Assay for Free Radical Generation and Elimination from Complex I—Superoxide anion and hydrogen peroxide generation from complex I were detected as described previously (30) by using lucigenin (50 ␮M) alone or luminol (50 ␮M) plus horseradish peroxidase (50 units)-derived chemiluminescence, respectively, in the ultra-weak luminescence analysis system (BPCL Inc., Beijing, China). Briefly, the reaction mixture contained 200 ␮g/ml SMF, 200 ␮M decylubiquinone(acoenzymeQ2 analog),50mM sodiumphosphatebuffer,pH 7.4, 200 ␮M NADH, and the tested molecules in given concentrations for a total volume of 1 ml. After preincubation at 25 °C for 3 min, SMF was added to launch the reaction, and chemiluminescence was monitored simultaneously. In some cases, 500 units/ml SOD or 400 units/ml catalase were added as positive controls. The integral of the signal peak reflects the formation of total superoxide anion and hydrogen peroxide. The relationship between chemical concentration and the superoxide anion and hydrogen peroxide formation is plotted as the integral area of the peak on the ordinate with the chemical concentration on the abscissa. Monitoring of Intra-mitochondria Oxidation and Reduction Status— The pyridine nucleotide oxidation-reduction status was evaluated based on endogenous NAD(P)H fluorescence (excitation-emission, 340 – 460 nm). The system was the same as described above, where 5 mM glutamate and 2.5 mM malate were added as substrates. To detect intra-mitochondria reduced thiol group concentrations, intact mitochondria pellets were collected as described for cytochrome c assay.

VOLUME 280 • NUMBER 37 • SEPTEMBER 16, 2005

Nicotine Protects Mitochondria from Neurotoxic Insult Then 100 ␮l of mitochondria solution (100 ␮g of protein) were added to 700 ␮l of methanol and 20 ␮l of Tris-EDTA (250/20 mM, pH 8.2, at 25 °C). 20 ␮l of Ellman’s reagent (10 mM) was then added, and the reaction was incubated for 15 min at room temperature. Absorbance of the medium was read at 410 nm against Ellman’s reagent. Statistical Analyses—Data were presented as mean ⫾ S.E. All statistical analyses were completed using the Origin program (Origin 7.5, OriginLab Corp.). Difference between groups was established by oneway or two-way analysis of variance. A probability level of 5% (p ⬍ 0.05) was considered significant.

RESULTS Nicotine Inhibited Cytochrome c Release and Postponed Mitochondria Swelling in SY5Y Cells Independently of nAChRs—After 24 h of incubation, 2 mM MPP⫹ triggered the release of cytochrome c from mitochondria into cytosol markedly, whereas 10 ␮M CsA and 1–10 ␮M nicotine significantly reduced/prevented concentration of cytochrome c in cytosol, whereas nontreated cells did not show cytosolic cytochrome c (Fig. 1A, Western blot). Under a fluorescence microscope, cytochrome c-GFP (Fig. 1B, grid 1) co-distributed with mitochondria (Fig. 1B, grid 2), and their overlap also normally appeared as thin filaments (Fig. 1B, grid 3), in keeping with a previous report (28). In the condition of swelling, mitochondria changed from filamentous to a spherical shape, and the diameter increased (compare Fig. 1B, grids 5 and 6). This morphological change is because of the fact that the volume-to-surface ratio is higher in a spherical shape than in a rod shape, which adapts to the material influx from cytoplasm during swelling (28). Because the sensitivity of the technique did not allow us to detect morphological swelling induced by MPP⫹ alone (as high as 10 mM) or together with Ca2⫹ within 30 min (data not shown), sensitive and differentiable effects were obtained by using a neurotoxin mixture containing 200 ␮M MPP⫹ and 5 ␮M 6-OHDA. It instantly resulted in mitochondria high amplitude swelling within 1 min (Fig. 1C, lane 1). At the same time, the co-distribution of green and red fluorescence was disturbed, where the red fluorescence (mitochondria) becomes convergent and spherical, whereas the green fluorescence (cytochrome c-GFP protein) quickly dispersed and separated from the mitochondria. Control experiments ensured that mecamylamine alone (1–50 ␮M) did not affect the toxic effects of the mixture. Nicotine alone (1–20 ␮M) or in combination with mecamylamine (in the absence of neurotoxin mixture) had no effect on the fluorescence and morphology (data not shown). CsA (10 ␮M), however, significantly inhibited mitochondria swelling and cytochrome c release. Indeed, green fluorescence co-distributed with the red one all through the experiment (Fig. 1C, lane 2). When cells were preincubated with 20 ␮M mecamylamine for 15 min followed by 1 ␮M nicotine for another 15 min (total volume 50 ␮l), the toxicity of the mixture was postponed because co-localization of cytochrome c-GFP and mitochondria was preserved until 25 min (Fig. 1C, lane 3). However, the high amplitude swelling was not prevented. Preincubation with 20 ␮M mecamylamine and 10 ␮M nicotine prevented both cytochrome c release (c-GFP remained inside mitochondria) and high amplitude swelling (Fig. 1C, lane 4).

FIGURE 1. Cytochrome c release and mitochondria swelling in SY5Y cells. A, SY5Y cells were treated with 2 mM MPP⫹ and 20 ␮M mecamylamine, and 10 ␮M CsA or 1–10 ␮M nicotine was added, respectively (lanes 1– 4). Nontreated cells did not show cytochrome c (cyt. c) release (lane 5). Cytoplasmic cytochrome c and actin were detected by Western blot. B, representative photographs showing SY5Y cell expressing cytochrome c-GFP

SEPTEMBER 16, 2005 • VOLUME 280 • NUMBER 37

plasmid (grid 1, green), Mitotracker Red CMXRos staining (grid 2, red), the overlap of GFP and Mitotracker Red CMXRos fluorescence (grid 3, yellow), and transmission imaging of the same cell (grid 4). Mitochondria morphology under normal (grid 5) and swelling (grid 6) conditions is also shown. C, by using these markers, cells were preincubated with 20 ␮M mecamylamine for 15 min and then treated with (lane 1) mixture only, (lane 2) mixture ⫹ 10 ␮M CsA, (lane 3) mixture ⫹ 1 ␮M nicotine, and (lane 4) mixture ⫹ 10 ␮M nicotine. Each experiment is a representative example of three to four similar experiments performed in separate dishes. White scale bar, 10 ␮m.

JOURNAL OF BIOLOGICAL CHEMISTRY

32407

Nicotine Protects Mitochondria from Neurotoxic Insult

FIGURE 2. Western blot detection of cytochrome c release with different electron donors at complex I (A and B) or II (C). A, Percoll gradient brain mitochondria (1 mg/ml) were treated in respiratory buffer for 30 min at 25 °C. Nicotine at various concentrations was added in the absence (lanes 1–5) or presence (lanes 6 –10) of 500 ␮M MPP⫹. B, the mitochondria were treated in respiratory buffer for 30 min at 25 °C (lanes 1– 8) in the presence of 50 ␮M Ca2⫹ plus 500 ␮M phosphate. Nicotine (lanes 1– 6), CsA (lane 7), or H2O (control, lane 8) were added, respectively. C, reaction system was the same as mentioned in A and B but 5 mM succinate plus 1 ␮M rotenone instead of glutamate/malate were used as electron donors for complex II. Chemical concentrations are as described in the figure. Densitometric measurement of control group was set as 100% (lane 5 in A, lane 8 in B, and lane 1 in C). Two-way analysis of variance was applied with MPP⫹ and nicotine (Nic) (or CsA) as independent variables. *, p ⱕ 0.05 compared with control group, and #, p ⱕ 0.05 compared with MPP⫹ (or Ca2⫹)-treated group.

Nicotine Inhibited Cytochrome c Release from Intact Mitochondria— The above experiments were performed on cells, whereas the following experiments used mitochondria preparations to investigate further the mechanisms of nicotine-mediated neuroprotection. When a substrate of mitochondria complex I was supplied, 500 ␮M MPP⫹ significantly increased cytochrome c release from intact mitochondria (Fig. 2A, 1ane 6) in comparison with control conditions (Fig. 2A, 1ane 5). When the experiment was changed to a mild condition (25 °C plus 15 min), no detectable cytochrome c was found in control group. However, 500 ␮M MPP⫹ still induced remarkable cytochrome c efflux (data not shown). Nicotine dose-dependently inhibited cytochrome c release from mitochondria, with or without MPP⫹ (Fig. 2A, lanes 1–10). Nicotine at 100 ␮M virtually abolished the cytochrome c release in both control and MPP⫹-treated groups (Fig. 2A). Nicotine also strongly inhibited Ca2⫹ plus phosphate-induced cytochrome c release (Fig. 2B). Nicotine, from 1 ␮M and beyond, inhibited cytochrome c efflux (Fig. 2B, lane 4) in comparison with the Ca2⫹-treated group (Fig. 2B, lane 1), even more potently than CsA (Fig. 2B, lane 8) under the same conditions. In the presence of succinate (plus rotenone), 500 ␮M MPP⫹ or a high dose of Ca2⫹ plus phosphate induced a 3– 4-fold significant increase in cytochrome c release (Fig. 2C, lanes 2 and 5) in comparison with the control condition (Fig. 2C, lane 1). Such dramatic increase is almost completely inhibited in the presence of 1 ␮M CsA (Fig. 2C, lanes 4 and 7). As 100 ␮M nicotine had no significant effect under both conditions (Fig. 2C, lanes 3 and 6), it may suggest a robust effect at the complex I level only.

32408 JOURNAL OF BIOLOGICAL CHEMISTRY

Nicotine Inhibited MPP⫹- and Ca2-induced Mitochondrial High Amplitude Swelling but Did Not Inhibit the Membrane Potential Loss l— Mitochondria preincubated with a low dose of Ca2⫹ (25 ␮M) had a negligible effect on the absorption at 540 nm, even when 500 ␮M phosphate was added at t2 (curve 1 in Fig. 3, A and B). Different concentrations of nicotine also had no significant effect on this condition (data not shown). On the contrary, MPP⫹ immediately and dose-dependently induced an absorption decline at 540 nm (Fig. 3A, curves 3–5), suggesting high amplitude swelling. The swelling induced by MPP⫹ at 500 ␮M was significantly inhibited by 1 ␮M CsA (curve 2 in Fig. 3B). Increased nicotine concentration from 0.1 to 1000 ␮M increasingly delayed the 500 ␮M MPP⫹-induced swelling (Fig. 3B, curves 3–7). Nicotine at 100 ␮M also partly inhibited the 50 ␮M Ca2⫹ plus 500 ␮M phosphate-induced mitochondria high amplitude swelling, whereas1 ␮M CsA almost completely prevented the absorption decline (Fig. 3C). Treatment of mitochondria with MPP⫹ resulted in an immediate increase in Rh123-specific fluorescence, indicating a decline in the mitochondria membrane potential (Fig. 4A, curves 1– 4). Increased MPP⫹ concentrations showed enhanced depletion effects, in keeping with previous reports (31). In Fig. 4B, curve 1 is a negative control where no MPP⫹ was added, and thus, no significant decrease in membrane potential was observed. The 500 ␮M MPP⫹-induced membrane potential loss (Fig. 4B, curve 3) is inhibited by 1 ␮M CsA (curve 2). Nicotine concentration below 100 ␮M had no effect on the Rh123 fluorescence

VOLUME 280 • NUMBER 37 • SEPTEMBER 16, 2005

Nicotine Protects Mitochondria from Neurotoxic Insult

FIGURE 3. Mitochondria swelling in response to MPPⴙ alone (A), 500 ␮M MPPⴙ plus CsA or nicotine at various concentrations (B), and high dose of Ca2ⴙ-phosphate (Pi) plus 100 ␮M nicotine or 1 ␮M CsA. (C) 5 mM glutamate and 2.5 mM malate were used as substrates. A and B, curve 1 is the nontreated group, and curve 2 shows the effects of 1 ␮M CsA. Curves 3–5 in A show the effects of MPP⫹ concentration at 100, 200, and 500 ␮M, respectively. B, curve 2 shows the effects of 500 ␮M MPP⫹ alone, whereas curves 3– 8 show the effects of a combination of 500 ␮M MPP⫹ and nicotine at 1000, 100, 10, 1, 0.1, and 0 ␮M, respectively. C, curves 1–3 represent the swelling of mitochondria treated with CsA ⫹ Ca2⫹, nicotine ⫹ Ca2⫹, or Ca2⫹ alone, respectively. Each figure is a representative example of three to five similar experiments performed in separate mitochondrial preparations.

FIGURE 4. Mitochondria membrane potential in response to MPPⴙ alone (A), 500 ␮M MPPⴙ plus CsA or nicotine at various concentrations (B), and high dose of Ca2ⴙ-phosphate (Pi) plus 100 ␮M nicotine or 1 ␮M CsA (C). 5 mM glutamate and 2.5 mM malate were used as substrates. Mitochondria were treated as described in the swelling assay but for the addition of 200 nM rhodamine 123. A, curves 1– 4 show the effects of 0, 100, 200, and 500 ␮M MPP⫹, respectively. B, curve 1 represents the control situation without MPP⫹. Curve 2 shows the effects of adding 1 ␮M CsA and 500 ␮M MPP⫹ on mitochondria membrane potential. Curves 3– 6 show the effects of 500 ␮M MPP⫹ after having preincubated the mitochondria with 0, 10, 100, and 1000 ␮M nicotine, respectively. C, curve 2 shows the mitochondria membrane potential loss induced by 50 ␮M Ca2⫹ to 500 ␮M Pi. Such loss is inhibited by 1 ␮M CsA (added at t0) as shown on curve 1 but is enhanced when mitochondria were preincubated with 100 ␮M nicotine as shown on curve 3. Each figure is a representative example of three to five similar experiments performed in separate mitochondrial preparations.

(data not shown). At 100 and 1000 ␮M, however, nicotine slightly enhanced the membrane potential decrease (Fig. 4B, curves 5 and 6). Fig. 4C represents a classic example of 50 ␮M Ca2⫹ plus 500 ␮M phosphate-induced membrane potential loss. CsA (1 ␮M) effectively inhibited the decline in membrane potential (Fig. 4, curve 1). However, nicotine at concentrations below 100 ␮M induced a partial decrease in membrane potential (Fig. 4, curve 3), representing a membrane potential-independent mechanism for modulating PTP. Nicotine Inhibited Mitochondria Complex I Electron Efflux and ROS Generation—By adding 200 ␮M NADH to SMF as substrate induced an acute increase in both superoxide anion and hydrogen peroxide chemiluminescence signal. SOD and catalase administration eliminated superoxide anion and hydrogen dioxide chemiluminescence signal for 85.5 and 81.7%, respectively. Preincubation with nicotine inhibited both superoxide anion and hydrogen peroxide generation (Fig. 5A), suggesting that nicotine decreased the electron flow at the complex I level. It is worth noticing that nicotine is more effective in scavenging hydrogen peroxide than superoxide anion

SEPTEMBER 16, 2005 • VOLUME 280 • NUMBER 37

under these conditions, with significant effects at 10⫺11 and 10⫺6 M, respectively. MPP⫹ in the experimental conditions had negligible effects on hydrogen peroxide generation in mitochondria, whereas it induced a slight but significant decrease in O2. generation (Fig. 5B). Nicotine Protected Intra-mitochondria Redox State—The negligible effect of nicotine on reduced pyridine nucleotides (PN) was found at all concentrations except at 1 mM, which increased reduced PN concentration for about 7.7%. Standard incubation of intact mitochondria (Fig. 6A, control column) decreased the relative content in reduced thiol groups in comparison with the nonincubated mitochondria (Fig. 6A, Blank column). Administration of nicotine dose-dependently preserved reduced thiol groups in mitochondria (Fig. 6A). On the contrary, MPP⫹ treatment dose-dependently decreased detectable reduced thiol groups (Fig. 6B). As expected from the above, pretreatment with various concentrations of nicotine significantly helped mitochondria to prevent the decrease in relative content of reduced thiol groups induced by 200 ␮M MPP⫹. Statistical significance (p ⬍ 0.05) was reached at a concentration above 10⫺6 M (Fig. 6C).

JOURNAL OF BIOLOGICAL CHEMISTRY

32409

Nicotine Protects Mitochondria from Neurotoxic Insult

FIGURE 5. Effect of nicotine (A) and MPPⴙ (B) on ROS accumulation in mitochondria complex I. In the presence of lucigenin (50 ␮M) alone or luminol (50 ␮M) plus horseradish peroxidase (50 units), mitochondria were preincubated with chemicals in a given concentration at 25 °C for 3 min. Chemiluminescence was then initiated by adding 200 ␮g/ml SMF and was continuously monitored at 25 °C. Values are expressed as percent of control and are mean ⫾ S.E. (n ⱖ 6). *, p ⱕ 0.05 compared with control group.

FIGURE 6. Detection of reduced thiol groups in intact mitochondria treated with nicotine (A), MPPⴙ (B), and a combination of 200 ␮M MPPⴙ plus nicotine at various concentrations (C). After 100 ␮g of protein were collected from the pellet, reduced thiol groups were detected with Ellman’s reagent. The Control group indicates mitochondria treated with only 5 mM glutamate, 2.5 mM malate as substrates, and the Blank group indicates Percoll gradient mitochondria without any treatment. *, p ⱕ 0.05 compared with control group; #, p ⱕ 0.05 compared with MPP⫹-treated group (applies to C only).

DISCUSSION The key finding of this study is to demonstrate that nicotine neuroprotective effects can be mediated directly through interaction with mitochondria independently from the nAChRs. We posit that nicotine, besides its previously known inhibition of complex I and decreased ROS generation mode (19), inhibits neurotoxin-induced CsA-sensitive mitochondria swelling and cytochrome c release through inhibition of the mPTP. Although in our experiments both nicotine and CsA inhibited cytochrome c efflux and mitochondria swelling, their inhibitory effects remain different in nature. Indeed, (i) nicotine did not inhibit cytochrome c release with succinate as mitochondria electron donors, whereas CsA did. (ii) Nicotine was not as efficacious as CsA in inhibiting the high amplitude swelling in intact mitochondria but has similar inhibitory effects on cytochrome c release. (iii) CsA protected mitochondria membrane potential, whereas nicotine decreased it. (iv) Although the unique nicotine-binding site in mitochondria is complex I (11), CsA specifically interacts with cyclophilin D in the PTP (32). The evidence suggests that nicotine and CsA affect differentially mPTP. Mitochondria respiratory chain, particularly complex I, is concerned in nicotine mode of action. We demonstrate that nicotine can affect mitochondria physiology in a receptor-independent manner. We applied mecamylamine, a nonselective nAChR antagonist in the SY5Y cell line. Confocal photogrammetry on mitochondria in these conditions demonstrated that 1 ␮M nicotine postponed cytochrome c release induced by a mixture containing MPP⫹ and 6-OHDA, two toxins that are both proved to inhibit

32410 JOURNAL OF BIOLOGICAL CHEMISTRY

mitochondria respiratory chain and induce cytochrome c release (33, 34). Nicotine at 10 ␮M, which also significantly decreased cytochrome c release from intact mitochondria, almost completely inhibited the toxicity of the neurotoxin mixture. Because mecamylamine blocked nAChR activity at the selected concentration, it is likely that nicotine is membrane-permeable and takes effect inside the cell. In addition, mitochondria swelling took place immediately after the toxin mixture was added, suggesting that mitochondria is not only the target but also an initiative factor in this process. The interactions between mitochondria respiratory chain and PTP have been investigated as the respiratory chain determined almost all aspects of the mitochondria characters (18, 35–39). Although the relationship between complex I and mPTP is shown through the correlation with free radical generation (40) and the redox state of ubiquinone (41, 42), Fontaine et al. (26) found that the mPTP opening was dramatically affected by the substrates used for energization. Indeed, lower Ca2⫹ load was required when electrons were provided to complex I rather than to complex II. This increased sensitivity of PTP opening does not depend on differences in membrane potential, matrix pH, Ca2⫹ uptake, oxidation-reduction status of pyridine nucleotides, or production of H2O2 but is directly related to the rate of electron flow through complex I. Rotenone, a specific complex I inhibitor, is reportedly more potent than CsA at inhibiting Ca2⫹-induced PTP opening in digitonin-permeabilized cells (43). As nicotine also binds mitochondria complex I and competes with NADH (11), nicotine would also decrease NADH consumption and inhibit electron flow through complex I. Our experi-

VOLUME 280 • NUMBER 37 • SEPTEMBER 16, 2005

Nicotine Protects Mitochondria from Neurotoxic Insult ments also demonstrate that nicotine dose-dependently inhibits NQR activity (data not shown) in agreement with previous reports (11). They further emphasize the complicated relationship between nicotine and complex I. Because electron flow is not easily detectable directly, we adopted the chemiluminescence method that detects ROS generation in complex I, partly reflecting the electron flow/leakage (30, 42). Decreased electron flow (and thus leakage) is further highlighted here by the nicotine-induced inhibition of both superoxide anion and hydrogen peroxide generation, which is associated with inhibited mPTP, decreased cytochrome c release, and reduced thiol group depletion. That nicotine did not prevent MPP⫹ and calcium-induced cytochrome c release in the presence of rotenone, a complex I inhibitor, and succinate, a substrate of complex II–IV, suggests that complexes II–IV of the mitochondrial respiratory chain are not directly involved in that process. In fact, nicotine was also found to have negligible effect on either mitochondria swelling or membrane potential in the same conditions (rotenone and succinate), further supporting the lack of interaction between nicotine and complex II–IV (44). mPTP is thought to be directly modulated by both electron flux and proton electrochemical difference (26). mPTP opening would result in a decreased membrane potential, whereas the reverse, i.e. collapse in membrane potential, does not necessary provoke mPTP opening (43, 45, 46). Here, however, nicotine induced a decrease in membrane potential, likely because of direct competition with NADH. As a result, this competition would inhibit the proton-motive force and decrease the proton electrochemical gradient, thus closing the mPTP. The relationship between membrane potential and mPTP is certainly more complex than thought earlier. The redox state of mitochondria is another key factor involved in modulation of opening-closing of mPTP (23, 47, 48). The redox state variations can be assessed through changes in both reduced PN and reduced thiol group contents. We considered in our experiments that nicotine competition with NADH should result in an increase in reduced PN concentration, a factor known to affect mPTP (49, 50). However, among all concentration considered, nicotine only slightly increased reduced PN content at 1 mM. Because nicotine significantly inhibited cytochrome c release at much lower concentrations (10 ␮M in MPP⫹ condition and 1 ␮M in high dose Ca2⫹ condition), accumulation of reduced PN is thus unlikely to mediate nicotine effects onto mPTP in our experimental conditions. Acute intracellular thiol depletion causes mitochondria permeability transition as shown after ebselen administration (51), a reduced thiol scavenger. Most interestingly, SOD pretreatment significantly decreased the level of superoxide radicals in ebselen-treated cells but had no effect on mitochondria membrane potential loss and subsequent apoptosis. This is suggestive of a free radical-independent apoptosis pathway (51). MPP⫹ administration, in our experiments, also induced a loss of reduced thiol group independent of ROS generation. Several reports argue against free radical involvement in MPP⫹ toxicity. For instance, MPP⫹ fails in promoting free radical formation in the following: (i) in rat brain synaptosomes and mitochondria (33) and (ii) in cell cultures treated with MPP⫹ (52–55). So the decrease in mitochondria-reduced thiol group content should be attributed to the opening of mPTP, which can release small molecules such as GSH from mitochondria. Again, MPP⫹-induced decrease in reduced thiol group was dose-dependently prevented by nicotine preincubation, comparably to what happened for cytochrome c. This nicotine effect on reduced thiol group, based on above data, may result from the ability of nicotine to inhibit mitochondria permeability transition. Because the retention of reduced thiol group may also prevent permeability transition (49, 56, 57), the two factors can combine to further inhibit the mPTP opening.

SEPTEMBER 16, 2005 • VOLUME 280 • NUMBER 37

Besides the inhibition of electron flee in complex I, the anti-oxidant properties of nicotine must be considered. Indeed, nicotine strongly inhibits the Fenton reaction (9), thereby providing a likely mechanism for preventing at least in part the ROS generation. Our experiments support the following hypotheses: (i) spontaneous ROS generation in mitochondria was inhibited by nicotine (Fig. 5A); (ii) reduced thiol group depletion caused by ROS was prevented by nicotine (Fig. 6A); and (iii) despite inhibition of the mitochondria complex I, the autoxidation and ROS formation that contribute to 6-OHDA-induced mitochondria swelling and cytochrome c release in SY5Y cells were also inhibited by nicotine (Fig. 1C). The nicotine concentration in the blood after smoking ranges from 20 nM to 1 ␮M (58, 59) but reaches higher concentrations when nicotine is taken via high dose patch/gum therapy (60). In our experiments, nicotine concentrations within that range did not fully inhibit both MPP⫹- and calcium-induced mitochondrial high amplitude swelling. However, at the same physiologically relevant doses of nicotine, the nicotine nonreceptor-mediated effects are most likely achieved through inhibition of cytochrome c release, a decrease in ROS generation, and maintenance of intra-mitochondria redox states. Altogether the present data suggest a receptor-independent neuroprotective mode of action of nicotine through modulation of complex I activity, its main binding site in mitochondria, and regulation of the mPTP resulting from subsequent decreased electron flow and release in complex I. Reduction in the mitochondria free radical generation and maintenance of intra-mitochondria redox state may also take part in the modulation process. Such nonexclusive and possibly additive mechanisms may explain, at least in part, the neuroprotective effect of nicotine. If these events also take place in in vivo models of PD, they deserve further investigation. Acknowledgments—We are grateful to Prof. Jian-Xin Xu for the instrumental help with various aspects of the experiments and to Xiang Shi and Yan Teng for their patient help in animal treatment and confocal microscopy. The cytochrome c-GFP plasmid was kindly provided by Prof. D.C. Chang (Hong Kong University of Science and Technology).

REFERENCES 1. Ross, G. W., and Petrovitch, H. (2001) Drugs Aging 18, 797– 806 2. Allam, M. F., Campbell, M. J., Hofman, A., del Castillo, A. S., and Navajas, R. F. (2002) Rev. Neurol. (Paris) 34, 686 – 689 3. Ehringer, H., and Hornykiewicz, O. (1960) Klin. Wochenschr. 38, 1236 –1239 4. Behmand, R. A., and Harik, S. I. (1992) J. Neurochem. 58, 776 –779 5. Alves, G., Kurz, M., Lie, S. A., and Larsen, J. P. (2004) Movement Disorders 19, 1087–1092 6. O’Neill, M. J., Murray, T. K., Lakics, V., Visanji, N. P., and Duty, S. (2002) Curr. Drug Targets CNS Neurol. Disord. 1, 399 – 411 7. Newhouse, P., Singh, A., and Potter, A. (2004) Curr. Top. Med. Chem. 4, 267–282 8. Quik, M., and Kulak, J. M. (2002) Neurotoxicology 23, 581–594 9. Linert, W., Bridge, M. H., Huber, M., Bjugstad, K. B., Grossman, S., and Arendash, G. W. (1999) Biochim. Biophys. Acta 1454, 143–152 10. Liu, Q., Tao, Y., and Zhao, B. L. (2003) Appl. Magn. Reson. 24, 105–112 11. Cormier, A., Morin, C., Zini, R., Tillement, J. P., and Lagrue, G. (2001) Brain Res. 900, 72–79 12. Mizuno, Y., Ohta, S., Tanaka, M., Takamiya, S., Suzuki, K., Sato, T., Oya, H., Ozawa, T., and Kagawa, Y. (1989) Biochem. Biophys. Res. Commun. 163, 1450 –1455 13. Schapira, A. H., Cooper, J. M., Dexter, D., Jenner, P., Clark, J. B., and Marsden, C. D. (1989) Lancet 1, 1269 14. Dawson, T. M., and Dawson, V. L. (2003) Science 302, 819 – 822 15. Sershen, H., Mason, M. F., Reith, M. E., Hashim, A., and Lajtha, A. (1986) Neuropharmacology 25, 1231–1234 16. Soto-Otero, R., Mendez-Alvarez, E., Hermida-Ameijeiras, A., Lopez-Real, A. M., and Labandeira-Garcia, J. L. (2002) Biochem. Pharmacol. 64, 125–135 17. Przedborski, S., Tieu, K., Perier, C., and Vila, M. (2004) J. Bioenerg. Biomembr. 36, 375–379 18. Chavez, E., Melendez, E., Zazueta, C., Reyes-Vivas, H., and Perales, S. G. (1997) Biochem. Mol. Biol. Int. 41, 961–968

JOURNAL OF BIOLOGICAL CHEMISTRY

32411

Nicotine Protects Mitochondria from Neurotoxic Insult 19. Cormier, A., Morin, C., Zini, R., Tillement, J. P., and Lagrue, G. (2003) Neuropharmacology 44, 642– 652 20. Stoltz, M., Rassow, J., Buckmann, A. F., and Brandsch, R. (1996) J. Biol. Chem. 271, 25208 –25212 21. Huang, X., Zhai, D., and Huang, Y. (2001) Mol. Cell. Biochem. 224, 1–7 22. Salvi, M., Brunati, A. M., Clari, G., and Toninello, A. (2002) Biochim. Biophys. Acta 1556, 187–196 23. Broekemeier, K. M., Klocek, C. K., and Pfeiffer, D. R. (1998) Biochemistry 37, 13059 –13065 24. Batandier, C., Leverve, X., and Fontaine, E. (2004) J. Biol. Chem. 279, 17197–17204 25. Fontaine, E., and Bernardi, P. (1999) J. Bioenerg. Biomembr. 31, 335–345 26. Fontaine, E., Eriksson, O., Ichas, F., and Bernardi, P. (1998) J. Biol. Chem. 273, 12662–12668 27. Leverve, X. M., and Fontaine, E. (2001) IUBMB Life 52, 221–229 28. Gao, W., Pu, Y., Luo, K. Q., and Chang, D. C. (2001) J. Cell Sci. 114, 2855–2862 29. Sims, N. R. (1990) J. Neurochem. 55, 698 –707 30. Zhao, Y., Wang, Z. B., and Xu, J. X. (2003) J. Biol. Chem. 278, 2356 –2360 31. Cassarino, D. S., Parks, J. K., Parker, W. D., Jr., and Bennett, J. P., Jr. (1999) Biochim. Biophys. Acta 1453, 49 – 62 32. Clarke, S. J., McStay, G. P., and Halestrap, A. P. (2002) J. Biol. Chem. 277, 34793–34799 33. Fonck, C., and Baudry, M. (2003) Brain Res. 975, 214 –221 34. Mazzio, E. A., Reams, R. R., and Soliman, K. F. (2004) Brain Res. 1004, 29 – 44 35. Reed, D. J., and Savage, M. K. (1995) Biochim. Biophys. Acta 1271, 43–50 36. Hirsch, T., Marzo, I., and Kroemer, G. (1997) Biosci. Rep. 17, 67–76 37. Greenamyre, J. T., Sherer, T. B., Betarbet, R., and Panov, A. V. (2001) IUBMB Life 52, 135–141 38. Maciel, E. N., Kowaltowski, A. J., Schwalm, F. D., Rodrigues, J. M., Souza, D. O., Vercesi, A. E., Wajner, M., and Castilho, R. F. (2004) J. Neurochem. 90, 1025–1035 39. Armstrong, J. S., Whiteman, M., Rose, P., and Jones, D. P. (2003) J. Biol. Chem. 278, 49079 – 49084 40. Brand, M. D., Affourtit, C., Esteves, T. C., Green, K., Lambert, A. J., Miwa, S., Pakay,

32412 JOURNAL OF BIOLOGICAL CHEMISTRY

J. L., and Parker, N. (2004) Free Radic. Biol. Med. 37, 755–767 41. Chan, T. S., Teng, S., Wilson, J. X., Galati, G., Khan, S., and O’Brien, P. J. (2002) Free Radic. Res. 36, 421– 427 42. Lambert, A. J., and Brand, M. D. (2004) J. Biol. Chem. 279, 39414 –39420 43. Chauvin, C., De Oliveira, F., Ronot, X., Mousseau, M., Leverve, X., and Fontaine, E. (2001) J. Biol. Chem. 276, 41394 – 41398 44. Pryor, W. A., Arbour, N. C., Upham, B., and Church, D. F. (1992) Free Radic. Biol. Med. 12, 365–372 45. Aronis, A., Komarnitsky, R., Shilo, S., and Tirosh, O. (2002) Antioxid. Redox Signal. 4, 647– 654 46. Brunner, M., Moeslinger, T., and Spieckermann, P. G. (2001) Comp. Biochem. Physiol. B 128, 31– 41 47. Lemasters, J. J., and Nieminen, A. L. (1997) Biosci. Rep. 17, 281–291 48. Chernyak, B. V. (1997) Biosci. Rep. 17, 293–302 49. Chernyak, B. V., and Bernardi, P. (1996) Eur. J. Biochem. 238, 623– 630 50. Savage, M. K., and Reed, D. J. (1994) Biochem. Biophys. Res. Commun. 200, 1615–1620 51. Yang, C. F., Shen, H. M., and Ong, C. N. (2000) Arch. Biochem. Biophys. 380, 319 –330 52. Choi, W. S., Yoon, S. Y., Oh, T. H., Choi, E. J., O’Malley, K. L., and Oh, Y. J. (1999) J. Neurosci. Res. 57, 86 –94 53. Lotharius, J., Dugan, L. L., and O’Malley, K. L. (1999) J. Neurosci. 19, 1284 –1293 54. Seyfried, J., Soldner, F., Kunz, W. S., Schulz, J. B., Klockgether, T., Kovar, K. A., and Wullner, U. (2000) Neurochem. Int. 36, 489 – 497 55. Fonck, C., and Baudry, M. (2001) Brain Res. 905, 199 –206 56. Armstrong, J. S., and Jones, D. P. (2002) FASEB J. 16, 1263–1265 57. Scarlett, J. L., Packer, M. A., Porteous, C. M., and Murphy, M. P. (1996) Biochem. Pharmacol. 52, 1047–1055 58. Benowitz, N. L., Kuyt, F., and Jacob, P., III (1982) Clin. Pharmacol. Ther. 32, 758 –764 59. Jarvik, M. E., Madsen, D. C., Olmstead, R. E., Iwamoto-Schaap, P. N., Elins, J. L., and Benowitz, N. L. (2000) Pharmacol. Biochem. Behav. 66, 553–558 60. Dale, L. C., Hurt, R. D., Offord, K. P., Lawson, G. M., Croghan, I. T., and Schroeder, D. R. (1995) J. Am. Med. Assoc. 274, 1353–1358

VOLUME 280 • NUMBER 37 • SEPTEMBER 16, 2005