Phosphorylation is considered a main mechanism modulating nNOS (neuronal nitric oxide synthase) function to reduce NO production. In the present study, the effects of nNOS phosphorylation on redox signalling, including that of NO, ROS (reactive oxygen species), and 8-nitro-cGMP (8-nitroguanosine 3′,5′-cyclic monophosphate), a downstream messenger of redox signalling, were investigated. In vitro experiments revealed that a phosphorylation-mimic mutant of nNOS (Ser847 replaced with aspartic acid, 847D) increased uncoupling to produce a superoxide. In addition, nicotine, which triggers an influx of Ca2+, induced more ROS and 8-nitro-cGMP production in 847D-expressing PC12 cells than WT (wild-type)-expressing cells. Additionally, nicotine-induced phosphorylation of nNOS at Ser847 and increased ROS and 8-nitro-cGMP production in rat CGNs (cerebellar granule neurons). In CGNs, the NOS (nitric oxide synthase) inhibitor L-NAME (NG-nitro-L-arginine methyl ester) and superoxide dismutase completely inhibited ROS and 8-nitro-cGMP production, whereas the CaMK (Ca2+/calmodulin-dependent protein kinase) inhibitor KN93 mildly reduced this effect. Nicotine induced HO-1 (haem oxygenase 1) expression in CGNs and showed cytoprotective effects against apoptosis. Moreover, 8-nitro-cGMP treatment showed identical effects that were attenuated by KN93 pre-treatment. The present paper provides the first substantial corroboration for the biological effects of nNOS phosphorylation at Ser847 on redox signalling, including ROS and intracellular 8-nitro-cGMP generation in neurons, which possibly play roles in neuroprotection.
In the nervous system, NO is involved in multiple physiological and pathological processes, including neurotransmission, development, neuroprotection, apoptosis, vascular regulation and neuroinflammation . NO is produced by NOSs (nitric oxide synthases), which catalyse the conversion of L-arginine to L-citrulline and NO using NADPH as an electron donor . NOSs generate not only NO, but also ROS (reactive oxygen species) such as superoxide through uncoupling reactions . The isoform expressed in neuronal cells, designated as nNOS (neuronal NOS) [1,4,5], is activated by intracellular Ca2+ concentration via the reversible binding of Ca2+/CaM (calmodulin) . It was shown that nNOS knockdown induced cell death in rat CGNs (cerebellar granule neurons)  and that nNOS-deficient mice exhibited impaired spatial performance in the Morris water maze , had increased grooming and anxiety-related behaviours , and failed to consolidate olfactory-cued long-term recognition memory . On the basis of these findings, nNOS appears to participate in diverse neuronal functions. Phosphorylation of nNOS is considered an important mechanism for the regulation of nNOS function [10–17]. nNOS is phosphorylated at different residues by various protein kinases. Although the physiological roles of these phosphorylation events are still poorly characterized, phosphorylation at Ser847 has been well studied. It was reported that CaMK-Iα (Ca2+/CaM-dependent protein kinase Iα), CaMK-IIα, CaMK-IV and p90 ribosomal S6 kinase-1 phosphorylate nNOS at Ser847, resulting in a reduction in its NO synthesis activity [11–13,16,17]. Previous papers reported that phosphorylation of nNOS at Ser847 was detected under normal physiological conditions [11,15,16]. On the other hand, the phosphorylation was also reported under non-physiological conditions [13,18–22]. Thus this phosphorylation appears to be involved in the regulation of physiological and pathological functions, but the functions and the molecular mechanisms have been poorly understood. Furthermore, to date, no information is available concerning the effects of nNOS phosphorylation on ROS production.
The major NO signalling pathway is regulated by a cGMP-dependent mechanism [2,23]. We have previously reported that a novel nitrated cyclic nucleotide 8-nitro-cGMP (8-nitroguanosine 3′,5′-cyclic monophosphate) functions as not only a cGMP analogue, but also an electrophilic signal reacting with thiol groups to form a protein–S-cGMP adduct (S-guanylation) . Indeed, we previously showed that 8-nitro-cGMP plays crucial roles under various pathophysiological conditions via S-guanylation [25,26]. 8-Nitro-cGMP is produced endogenously in various cells in NO- and ROS-dependent manners [24,25,27,28], probably via the formation of reactive nitrogen oxide species such as peroxynitrite. It is now clear that ROS functions as not only a toxicant, but also a signalling molecule [24,29,30], therefore 8-nitro-cGMP is considered a downstream messenger of NO/ROS redox signalling.
The nAChRs (nicotinic acetylcholine receptors) are pentameric ligand-gated channels that generate a cation-selective pathway across the plasma membrane . Stimulation of nAChRs by nicotine leads to elevation of intracellular Ca2+ levels through various pathways, which then regulates diverse biological functions depending on the activation of specific signalling cascades . The increase in intracellular Ca2+ levels that arises from the activation of nAChRs can activate CaMKs and nNOS [31–34], therefore we predict that nicotine regulates nNOS activity via a complicated mechanism. In addition, previous papers reported that nAChR activation promotes neurogenesis and survival of CGNs [35,36], suggesting potential regulatory roles for nAChR during cerebellar development. However, the mechanism has not been fully elucidated.
To date, the role of phosphorylation in nNOS at Ser847, especially in terms of redox signalling, is still poorly characterized. In the present study, we investigated the role of phosphorylation at Ser847 in the regulation of redox signalling. We studied the effects of phosphorylation on nNOS activity in vitro and redox signalling, including 8-nitro-cGMP production in either nNOS-expressing PC12 cells or rat CGNs using nicotine as a stimulant. Moreover, we examined whether the regulation of NO/ROS redox signalling by nNOS phosphorylation is associated with neuroprotection induced by nicotine.
The HRP (horseradish peroxidase)-conjugated anti-nNOS monoclonal antibody , anti-(phospho-Ser847 nNOS) polyclonal antibody (NP847)  and anti-8-nitro-cGMP monoclonal antibody  were prepared as described previously. 8-Nitro-cGMP was synthesized as described previously . The anti-HO-1 (haem oxygenase 1) antibody was purchased from Stressgen Bioreagents. PC12 cells were a gift from Professor Masami Takahashi (Kitasato University School of Medicine, Kanagawa, Japan). Bovine brain CaM, bovine immunoglobulin, Cu-Zn SOD (superoxide dismutase), BH4 (tetrahydrobiopterin), RPMI 1640 medium, anti-actin polyclonal antibody and HRP-conjugated secondary antibodies were purchased from Sigma–Aldrich. 2′,5′-ADP–Sepharose was purchased from GE Healthcare. Catalase, haemoglobin, nicotine and MEM (minimum essential medium) were purchased from Nacalai Tesque. DMPO (5,5-dimethylpyrroline-N-oxide) was obtained from Radical Research. L-NAME (NG-nitro-L-arginine methyl ester) and DAN (2,3-diaminonaphthalene) were purchased from Dojindo Laboratories. DHE (dihydroethidium) was purchased from ABD Bioquest. HS (horse serum) was purchased from Invitrogen. FBS was obtained from MultiSer (Cytosystems). KN93 was purchased from Wako Pure Chemical Industries. All other chemicals and reagents were from common suppliers and were of the highest grade commercially available.
We generated a mutation that mimics phosphorylation of nNOS at Ser847 . A previously described cDNA encoding rat nNOS  was used as the template for constructing the aspartic acid replacement mutant. For PCR mutagenesis, the forward and reverse primers used were as follows: 5′-AAGGTCCGATTCAACGACGTCTCCTCCTATTCT-3′ and 5′-AGAATAGGAGGAGACGTCGTTGAATCGGACCTT-3′ respectively, in which the target sites of mutagenesis are underlined. After sequencing, the insert fragment released via BamHI/EcoRI digestion was used as the template for overlap extension PCR. After sequencing, the product encoding full-length nNOS was released via NdeI/XbaI digestion and cloned into the pCW vector, which was designated as 847D. The 847A mutant, described previously , was used as a mutation control for nNOS activities.
Expression and purification of nNOSs
nNOS activity measurement
NO generation and NADPH oxidation activity were measured [4,11]. The standard assay was performed at room temperature (23°C) in a final reaction mixture of 300 μl containing 50 mM sodium phosphate buffer (pH 7.4), 10 μM oxyhaemoglobin, 0.2 mM NADPH, 0.625 μM CaM, 1 mM CaCl2, 120 units of catalase, 60 units of SOD, 10 μM BH4, 4 μM FAD, 4 μM FMN and 0.0625 μM nNOS in the presence or absence of 1 mM L-arginine. Reactions were initiated by the addition of nNOS.
nNOS activity in PC12 cells was measured by the DAN assay as described previously  with minor modifications. Unless otherwise indicated, cells were washed three times with incubation buffer (154 mM NaCl, 5.6 mM KCl, 3.6 mM NaHCO3, 5.6 mM glucose, 2.3 mM CaCl2, 5 mM Hepes and 5 mM L-arginine, pH 7.4) at room temperature, followed by incubation in the same buffer in a humidified atmosphere at 37°C for 30 min. Next, the buffer was discarded, and the cells were treated with 50 μM nicotine in incubation buffer at 37°C for 60 min in a humidified atmosphere. The nitrite levels in culture supernatants were measured using DAN as follows. Samples (100 μl) in 0.6-ml microtubes were mixed with 10 μl of fresh DAN (0.5 mg/ml in 0.62 M HCl) for 10 min at room temperature in the dark. Reactions were terminated with 5 μl of 2.8 M NaOH. The formation of DAN end-products was measured by an FP-610 spectrofluorometer (JASCO) with excitation at 365 nm and emission at 450 nm using NaNO2 as the standard.
Uncoupling efficiency was stoichiometrically determined by the combination of spectrophotometry and HPLC. Briefly, reaction mixtures of 50 mM sodium phosphate buffer (pH 7.4) containing 1 mM L-arginine, 0.1 mM CaCl2, 0.625 μM CaM, 4 μM FAD, 4 μM FMN, 400 units/ml catalase, 200 units/ml SOD, 0.0625 μM nNOS, 10 μM BH4 and 40 μM NADPH were incubated at room temperature. The consumption of NADPH in the reaction mixtures was spectrophotometrically monitored as decreases in absorbance at 340 nm using a U-3210 spectrophotometer (Hitachi). After depletion of NADPH, 0.1 M perchloric acid was added to the reaction mixtures to precipitate proteins. After centrifugation at 15000 g for 15 min at 4°C, the L-citrulline content of the supernatants was determined using RP (reverse phase)-HPLC. RP-HPLC was performed on a Mightysil RP-18 column (250 mm×2.0 mm inner diameter; Kanto Chemical), using precolumn derivatization with o-phthalaldehyde and fluorescence detection. As 1.5 mole of NADPH is utilized to produce 1 mole of NO and L-citrulline when NOS is fully coupled, the uncoupling efficiency (%) was calculated using formula X=(1−1.5A/B)×100, in which A and B are the value of the L-citrulline production and the value of the NADPH consumption respectively.
EPR spectroscopy for superoxide generation
Superoxide generation induced by nNOS was quantified via EPR spectroscopy using DMPO as the spin trap [24,40]. The reaction catalysed by nNOS was initiated by addition of 0.2 mM NADPH to reaction mixtures containing 0.0625 μM nNOS, 1 mM CaCl2, 0.625 μM CaM, 0.1 mM DTPA (diethylenetriaminepenta-acetic acid) and 45 mM DMPO in 100 mM sodium phosphate buffer (pH 7.4) at room temperature. After the reaction mixtures were kept standing for 2 min, EPR spectroscopy was performed. To investigate differences between WT and 847A, EPR spectrometry was performed after a 4 min reaction. Spin adducts of superoxide were identified and quantified by using an X-band EPR spectrometer (JES-RE1X, JEPL). The DMPO adduct of superoxide (DMPO-OOH) was quantified by normalizing the intensity of EPR signals obtained for DMPO-OOH to that of the standard reference signal of manganese oxide. Spectra were recorded at room temperature under the following conditions: modulation frequency, 100 kHz; modulation amplitude, 0.079 mT; scanning field, 335±5 mT; receiver gain, 250; response time, 0.1 s; sweep time, 2 min; microwave power, 10 mW; and frequency, 9.421 GHz.
Production of stable transformants
The gene encoding the nNOS 847D mutant was stably transfected into PC12 cells using the ViraPower Lentiviral Expression System (Invitrogen). Next, the nNOS mutant gene was subcloned into the pENTR vectors and then transferred into the pLenti6/V5-DEST vector using the LR Clonase Enzyme Mix Kit (Invitrogen) according to the manufacturer's instructions. The resultant plasmids (pLenti6/V5-DEST/nNOS 847D mutant) were cotransfected into 293FT cells using the ViraPower Packaging Mix (Invitrogen). The medium was replaced 24 h after initial transfection, and on the third day, the medium containing recombinant lentiviruses was collected. PC12 cells were cultured in the presence of viral solutions containing 10 μg/ml polybrene (Sigma–Aldrich). Infected cells were selected by 2.5 μg/ml blasticidin (Invitrogen), and surviving cells that expressed 847D at appropriate levels were used for subsequent analyses. Expression of 847D was confirmed by Western blotting using monoclonal antibodies specific for nNOS. Control and WT nNOS-expressing PC12 cells were prepared as described previously .
Cell culture and stimulation
PC12 cells were maintained in RPMI 1640 medium supplemented with 5% (v/v) FBS, 5% (v/v) HS and 1% (v/v) penicillin/streptomycin in a humidified atmosphere at 37°C. Cells were plated at a density of 3.0×105 cells/well in 12-well plates (Iwaki) for the DAN assay and 1.0×106 cells per 35-mm-diameter culture dishes to prepare cell extracts for Western blot analyses. For fluorescence microscopic detection of intracellular ROS generation and immunocytochemistry, cells were seeded on circular coverslips at a density of 2.0×105 cells/well in 24-well plates (Iwaki).
Preparation of CGNs was conducted in compliance with the Guideline for Animal Experimentation at Osaka Prefecture University, with an effort to minimize the number of animals used and their suffering. CGNs were obtained from 7–10-day-old Wistar rats as described previously . Cells were plated at a density of 4.0×105 cells/well in 24-well culture plates for the MTS assay and 2.0×106 cells per 60-mm-diameter culture dish to prepare cell extracts for Western blot analyses. For fluorescence microscopic detection of intracellular ROS generation and immunocytochemistry, cells were seeded on circular coverslips at a density of 1.0×105 cells/well in 4-well culture plates (Nalge Nunc International). CGNs were maintained in MEM containing 25 mM K+ (HK-MEM) with 5% (v/v) FBS, 5% (v/v) HS and 1% (v/v) penicillin/streptomycin in a humidified atmosphere at 37°C. After 12 days in vitro, cells were used for each experiment.
To study the formation of ROS and 8-nitro-cGMP, cells were stimulated with nicotine (50 μM) for 5, 10, 30 and 60 min and 24 h in a humidified atmosphere at 37°C. To investigate the mechanism of ROS and 8-nitro-cGMP production and nicotine-induced cytoprotection against apoptosis, cells were stimulated in the presence of membrane-permeable ROS scavengers (200 units/ml PEG-SOD)  for 30 min, a CaMK inhibitor (0.1 or 10 μM KN93) for 1 h, a membrane-permeable HO-1 inhibitor [20 μM PEG-ZnPP (PEGylated zinc protoporphyrin)]  for 1 h, a NOS inhibitor (0.2 mM L-NAME) for 1 h or 10 μM 8-nitro-cGMP for 24 h, followed by analyses for 8-nitro-cGMP formation and ROS generation and the MTS assay.
To study nicotine-induced nNOS phosphorylation at Ser847, CGNs were pre-incubated for 60 min with HK-MEM either in the presence or in the absence of 10 μM KN93 and then stimulated with HK-MEM either alone or with 50 μM nicotine for 10 min in a humidified atmosphere at 37°C. To analyse the mechanism of nicotine-induced cytoprotection against apoptosis, CGNs pre-treated with HK-MEM either with or without 50 μM nicotine over various times were incubated with MEM containing 5 mM K+ (LK-MEM) in the presence or absence of 50 μM nicotine for 24 h in a humidified atmosphere at 37°C.
Western blot analysis
Cells treated as described above were washed with PBS (pH 7.4) three times and lysed with 0.5 ml of lysis buffer [1% Triton X-100, 1% proteinase inhibitor cocktail (Nacalai Tesque), and 1% phosphatase inhibitor cocktail (Nacalai Tesque) in PBS] followed by sonication. After centrifugation at 15000 g for 15 min at 4°C, supernatants were collected. Proteins in cell lysates were heat-denatured, separated via SDS/PAGE and transferred to nitrocellulose membranes (Hybond-C, GE Healthcare). After blocking with Blocking One (Nacalai Tesque), membranes were incubated with antibodies in TBST (TBS containing 0.05% Tween 20). Antibodies used for Western blot analysis were anti-nNOS and NP847 for nicotine-induced phosphorylation of nNOS, as well as anti-HO-1 and anti-β-actin for nicotine-induced protein expression. After three washes in TBST, membranes were incubated with an HRP-conjugated secondary antibody at room temperature for 1 h. After three washes in TBST, immunoreactive bands were detected using a chemiluminescence reagent (SuperSignal Reagent, Thermo Fisher Scientific) with a luminescent image analyser (LAS-1000 mini; Fujifilm).
To confirm the purity of recombinant nNOS proteins, we performed Western blot analysis using the anti-nNOS antibody.
Detection of intracellular ROS production
Intracellular ROS generation was quantified by fluorescence microscopy with the oxidant-sensitive dye DHE. For fluorescence microscopy, PC12 cells and CGNs were treated with nicotine for various times. After stimulation, cells were washed with PBS three times and then stained with 10 μM DHE in PBS for 30 min in a humidified atmosphere at 37°C in the dark. Cells were washed carefully with PBS containing 0.1 mg/ml ascorbic acid and fixed with 4% paraformaldehyde (pH 7.4) for 15 min in the dark. Cells were washed again with PBS containing 0.1 mg/ml ascorbic acid, mounted with mounting medium (KPL), covered with coverslips, and examined using an inverted fluorescence microscope (Ti-S/L100; Nikon). Images were captured and processed by Nikon Ti-S/L100 software. Further image processing and quantification were performed using Adobe Photoshop Elements v.7.0. Fluorescence intensity values from three different experiments were obtained.
The formation of 8-nitro-cGMP in PC12 cells and CGNs was analysed by immunocytochemistry using an anti-8-nitro-cGMP monoclonal antibody [24–28]. After nicotine stimulation, cells were washed with PBS three times and fixed with Zamboni fixative (4% paraformaldehyde and 10 mM picric acid in 0.1 M phosphate buffer, pH 7.4) at 4°C for 7 h in the dark. After three washes with PBS, cells were permeabilized with 0.5% Triton X-100 at room temperature for 15 min and washed again with PBS. To block non-specific antigenic sites, cells were incubated at 4°C overnight with BlockAce (Snow Brand Milk Products). Next, cells were incubated overnight with the anti-8-nitro-cGMP-specific monoclonal antibody (1 μg/ml) at 4°C in Immunoreaction Enhancer Solution 1 (Toyobo). After three rinses with TBST, cells were incubated with Cy3 (indocarbocyanine)-labelled goat anti-(mouse IgG) antibody for 1 h at room temperature in the dark in Immunoreaction Enhancer Solution 2 (Toyobo). Cells were washed with TBST three times, mounted with mounting solution, covered with coverslips, and examined using a Nikon Ti-S/L100 microscope. Images were captured and processed by Nikon Ti-S/L100 software. Further image processing and quantification were performed using Adobe Photoshop Elements v.7.0.
To analyse the mechanism of nicotine-induced cytoprotection against apoptosis, CGNs were pre-treated with HK-MEM either in the presence or absence of KN93 (0.1 μM), PEG-ZnPP (20 μM) or L-NAME (0.2 mM) for 1 h. Subsequently, cells were treated with 50 μM nicotine or 10 μM 8-nitro-cGMP for 24 h and then incubated in LK-MEM with or without 50 μM nicotine for 24 h in a humidified atmosphere at 37°C. After treatment, cell viability was determined using the MTS assay (Promega). The cell viability of all treatment groups was normalized to that of untreated cells.
All experiments were performed at least three times. The values for individual experiments are presented as the means±S.E.M. Statistical significance was determined by the one-way ANOVA, two-way ANOVA or Student's paired t test using GraphPad Prism software. P<0.05 was considered significant.
Activity of the phosphorylation-mimic mutant of nNOS
Mutated nNOS (847D), in which Ser847 was substituted with an aspartic acid, has been shown to exhibit NO synthesis activity similar to that of Ser847-phospho-nNOS [11,16]. In the present study, we examined the enzymatic activity (NO synthesis and NADPH oxidation) and uncoupling efficiency of the 847D (phosphomimic) and the 847A (mutation control) mutants. The nNOS mutants expressed and purified from E. coli cells by affinity chromatography on 2′,5′-ADP–Sepharose had the same molecular mass as WT (wild-type) nNOS (Figure 1A). To investigate the effects of nNOS phosphorylation at Ser847 on its enzymatic activity, we first determined the rate of NO generation and NADPH oxidation by WT, 847D and 847A. The initial rate at which NO was produced by WT, 847D or 847A in the presence of L-arginine was 183.8±6.2, 140.1±7.8 or 177.0±5.6 nmole·min−1·mg of protein−1 respectively (Figure 1B). The NO production activity of 847D was approximately 76% of that of WT. These results are consistent with previous results [11,16]. The initial rate of NADPH oxidation catalysed by WT, 847D or 847A was 532.3±46.1, 832.6±53.2 or 482.9±21.5 nmole·min−1·mg of protein−1 in the absence of L-arginine, and 303.1±14.6, 390.6±19.5 or 275.7±7.2 nmole·min−1·mg of protein−1 in the presence of L-arginine respectively (Figure 1C). These data are consistent with previous reports  in which Ser847 phosphorylation was found to increase nNOS reductase activity. The stoichiometric analysis further revealed that the uncoupling efficiency of WT, 847D or 847A was 48.3%, 61.5% or 50.1% respectively (Figure 1D); suggesting that nNOS phosphorylation at Ser847 regulates uncoupling efficiency. Overall, there were no significant differences in the enzyme activities, including uncoupling efficiency, between WT and the 847A mutant.
Effects of the nNOS 847D mutation on its enzymatic activity
ROS generation activity of phosphorylated nNOS
Several previous studies demonstrated that nNOSs catalyse superoxide formation; that is, NOSs utilize uncoupled electrons from NADPH to produce superoxide [3,44]. The production of superoxide was examined by EPR spin trapping using the superoxide-specific spin trap DMPO (Figure 2). The DMPO-OOH spectra showing superoxide trapping [24,40,45] were produced via the reaction of DMPO with WT, 847D or 847A nNOS in the absence of L-arginine (Figures 2A and 2B, upper traces). The signals were nullified by addition of SOD to the reaction mixture, indicating that superoxide was the primary radical responsible for the observed EPR signals. Addition of the NOS inhibitor L-NAME also inhibited oxygenase domain-dependent electron uncoupling , revealing the domain responsible for superoxide generation. The amount of superoxide generated by 847D was approximately 200% of that generated by WT (Figure 2C); however, the amount generated by 847A was identical to that generated by the WT. These results suggest that the differences between WT and 847D are not due to substitution of Ser847 by other amino acids, but are probably due to phosphomimic substitution of Ser847.
Generation of superoxide by WT or mutant nNOS
Nicotine-induced nNOS activation and 8-nitro-cGMP formation in nNOS-expressing PC12 cells
As the in vitro experiments revealed no significant differences in enzymatic activity between WT and 847A nNOS, we generated PC12 cells that stably express WT and 847D to investigate intracellular nNOS activity. The protein expression levels of WT and 847D in the respective transfected PC12 cells were identical, whereas the control PC12 cells showed no endogenous nNOS expression (results not shown). Nicotine, which activates intracellular nNOS, is known to facilitate Ca2+ influx into the cell via the nAChR [31–34]. nAChR activation promotes the neurogenesis and survival of CGNs [35,36], in which CaMK-IIα and nNOS are both expressed . Although it has been reported that nicotine exhibits cytotoxicity under certain conditions , nicotine does not exhibit cytotoxicity under the conditions employed in the present study [32,33]. Nicotine-induced NO production in nNOS-expressing PC12 cells was measured by fluorescence analysis using the nitrite-specific probe DAN. Nicotine-induced nitrite generation was detected in nNOS-expressing cells, but not control cells (Figure 3A). In addition, cells expressing WT nNOS generated more nicotine-induced nitrite than those expressing 847D nNOS. These results were consistent with a previous report . Additionally, previous reports showed that the enzymatic activity of 847A was identical to that of WT in cell lines [12,16]. These results suggest that nicotine treatment induced intracellular NO production, which was decreased by nNOS phosphorylation at Ser847 in cell cultures.
Characterization of nicotine-induced nNOS activity in PC12 cells
Using fluorescence microscopy with the ROS-specific probe DHE, we found that nicotine treatment increased the fluorescence intensity only in nNOS-expressing cells and that 847D-expressing cells showed higher intensity than WT-expressing cells (Figures 3B and 3C).
8-Nitro-cGMP formation proceeds through an NO-dependent manner in various cell types. Our previous studies also demonstrated that ROS play an important role in 8-nitro-cGMP production [24,25,27,28]. In the present study, we analysed 8-nitro-cGMP formation in PC12 cells expressing WT and 847D. The time course of nicotine-induced 8-nitro-cGMP formation was determined in nNOS-expressing PC12 cells by immunocytochemistry using an 8-nitro-cGMP-specific antibody. The fluorescence intensity increased only in nNOS-expressing cells and was higher in 847D-expressing cells than in WT-expressing cells at all time points (Figures 4A and 4B). In contrast, the control cells showed only marginal immunoreactivity. Next, nicotine-dependent 8-nitro-cGMP formation was examined in the presence of the NOS inhibitor L-NAME or the ROS scavenger PEG-SOD. As shown in Figures 4(C) and 4(D), the nicotine-induced increase of fluorescence intensity was diminished by pre-treating cells with L-NAME and PEG-SOD. Taken together, these results suggest that nicotine-induced 8-nitro-cGMP production in PC12 cells was increased by nNOS phosphorylation at Ser847 and was dependent on NO and ROS derived from nNOS.
Analysis of the mechanism of 8-nitro-cGMP production in nicotine-treated PC12 cells
Determination of the relationship between 8-nitro-cGMP formation and nNOS phosphorylation at Ser847 in CGNs
nNOS phosphorylation at Ser847 in nicotine-treated CGNs was detected by immunoblotting with a phosphopeptide antiserum, NP847, specific for phospho-Ser847 in nNOS as previously described [11–13,15]. As shown in Figures 5(A) and 5(B), phosphorylation at Ser847 was significantly increased after treatment of CGNs with 50 μM nicotine for 10 min (Figures 5A and 5B, middle lane). However, pretreatment of CGNs with 10 μM KN93, a selective inhibitor of CaMKs, almost completely blocked nicotine-induced nNOS phosphorylation at Ser847 (Figures 5A and 5B, right-hand lane). In addition, we detected nicotine-induced ROS generation in CGNs by fluorescence microscopy. A significant increase in fluorescence intensity was observed in CGNs treated with nicotine for 10 min, and this increase was completely blocked by pre-treatment with 150 μM L-NAME and 200 units/ml PEG-SOD (Figures 5C and 5D). Notably, the fluorescence intensity in CGNs treated with nicotine in the presence of 10 μM KN93 was significantly attenuated compared with that in nicotine-stimulated cells in the absence of KN93; however, the intensity was still higher than that in unstimulated cells (Figures 5C and 5D). As shown in Figure 6, when CGNs were treated with nicotine for 24 h, the anti-8-nitro-cGMP antibody showed strong staining in the cells, and the immunoreactivity was completely blocked by pre-treatment with either L-NAME or PEG-SOD, indicating that 8-nitro-cGMP formation stimulated by nicotine in CGNs depended on the presence of NOS activity and superoxide. The fluorescence intensity in KN93-pre-treated cells was significantly greater than that in unstimulated cells, but was significantly less than that in nicotine-stimulated cells (Figure 6), suggesting that nNOS phosphorylation at Ser847 by CaMKs may enhance nicotine-induced 8-nitro-cGMP formation in CGNs.
Detection of nNOS phosphorylation at Ser847 and its effects on ROS generation in nicotine-treated CGNs
Analysis of the relationship between 8-nitro-cGMP formation and nNOS phosphorylation at Ser847 in nicotine-treated CGNs
Nicotine protects CGNs against apoptotic cell death via induction of HO-1 expression
As nicotine has been shown to protect CGNs from apoptotic neuronal cell death induced by K+ deprivation (or low concentration of K+; LK) [35,36], we first examined nicotine-induced cytoprotective effects against apoptosis induced by K+ deprivation. As shown in Figure 7(A), K+ deprivation significantly induced cell death; in contrast, nicotine attenuated the cytotoxicity in a time-dependent manner, and 24 h treatment completely blocked the cytotoxicity. Nicotine alone did not exhibit cytotoxicity under the conditions of the present study (Figure 7A). We found that nicotine increased the protein level of HO-1 in CGNs (Figure 7B). We previously found that HO-1 acted downstream of 8-nitro-cGMP through the Keap1 (Kelch-like ECH-associated protein 1)–Nrf2 pathway and exerted cytoprotective effects against oxidative damage [25,47]. As shown in Figure 7(C), pre-treatment with either L-NAME or PEG-ZnPP completely blocked nicotine-induced cytoprotection, whereas KN93 mildly attenuated the cytoprotective effect. It is worth noting that the above-mentioned inhibitors did not affect cell viability (results not shown). 8-Nitro-cGMP treatment for 24 h induced HO-1 expression in CGNs (Figure 7D). Moreover, pre-treatment of 8-nitro-cGMP exerted neuroprotective effects against apoptosis (Figure 7E).
Analysis of the mechanism of nicotine-induced cytoprotection against apoptosis in CGNs
Phosphorylation of nNOS at Ser847 was detected under normal physiological conditions [11,15,16]. However, the phosphorylation has also been observed under non-physiological conditions [13,18–22]. Phosphorylated nNOS has been reported to attenuate NO production in vitro and in vivo [11–17]. In the present study, we investigated the effect of nNOS phosphorylation at Ser847 on not only NO production, but also ROS generation (Figures 1B and 2). Our in vitro analysis shows that recombinant WT nNOS exhibited a lower uncoupling efficiency than the phosphorylation-mimic mutant 847D (Figures 1D).
nNOS has previously been demonstrated to generate superoxide, in addition to NO, by utilizing uncoupling electrons [3,44]. Therefore we expected that the amount of superoxide produced by 847D would be higher than that produced by WT. This speculation was supported by our EPR spin trapping analysis, which revealed that purified 847D generated 200% more superoxide than WT (Figure 2). Our results thus indicate that nNOS phosphorylation at Ser847 not only down-regulates the NO production activity of nNOS [11–13,15–17], but also increases ROS generation.
8-Nitro-cGMP is a novel nitrated second messenger, and its formation depends on both NO and ROS production [24,25,27,28], probably via the generation of reactive nitrogen oxide species such as peroxynitrite. Indeed, 8-nitro-cGMP is a downstream messenger of NO/ROS redox signalling. When synthesized within the same cell, superoxide and NO combine spontaneously to form peroxynitrite. In the present study, we confirmed that 8-nitro-cGMP formation was induced by nicotine treatment in PC12 cells and CGNs (Figures 4 and 6) and that nicotine-induced 8-nitro-cGMP formation was regulated via phosphorylation at Ser847 by CaMKs in the cells (Figure 6). These results provide the first indication of 8-nitro-cGMP formation in neuronal cells under physiological conditions. Although the importance of peroxynitrite is poorly understood, our results suggest that peroxynitrite may participate in the regulation of redox signalling mechanisms via 8-nitro-cGMP formation. 8-Nitro-cGMP synthesized endogenously in cells and tissues exhibits cytoprotective functions [24,25,47]. In the present study we emphasize that one of the important biological properties of 8-nitro-cGMP is its moderate electrophilicity without any apparent cytotoxicity, even at high (up to micromolar) concentrations for any type of cultured cells . Therefore, although peroxynitrite may be mostly toxic to cells, once it induces the formation of 8-nitro-cGMP, this non-toxic secondary product may mediate the physiological cellular response to oxidative stress. For example, in lipopolysaccharide/cytokine-treated rat C6 glioma cells  and peritoneal macrophages obtained from salmonella-infected mice , 8-nitro-cGMP was shown to activate the Keap1–Nrf2 pathway via S-guanylation of the redox sensor protein Keap1 which results in Nrf2 activation and expression of cytoprotective genes, including the antioxidative enzyme HO-1 [25,47]. We have previously reported in vivo formation of 8-nitro-cGMP in the liver and heart, and induction of HO-1 mediated by 8-nitro-cGMP in the liver of Salmonella-infected mice . Although our preliminary MS and immunohistochemical analyses revealed in vivo formation of 8-nitro-cGMP in rat brains, we are still in the process of performing these studies to fully understand the mechanisms of 8-nitro-cGMP formation and the functions involving neuroprotective gene induction in vivo. These data will be the subject of a future publication.
In the present study, we analysed ROS production by using a ROS-specific fluorescent probe, DHE. Although DHE has been used as a superoxide-specific fluorescent probe, recent papers reported that both superoxide-specific oxidation and non-specific oxidation of DHE occur in cells under oxidative conditions [48–50]. Due to the spectral overlapping in the fluorescent response of superoxide-derived and non-specific oxidized fluorescent products formed from DHE, caution should be exercised when performing fluorescent microscopic analysis with the probe . Several methods have been proposed to solve this imaging problem, including the use of specific inhibitors [50,51]. Thus, in the present study, the nicotine-induced increase of fluorescence was completely blocked by pre-treatment with L-NAME or PEG-SOD. The results strongly suggested that the nicotine-induced fluorescence was derived from the marker product oxidized by nNOS-derived superoxide.
Nicotine has been reported to directly or indirectly induce an increase in intracellular Ca2+ levels by stimulation of nAChRs, and consequently activates CaMKs [31–34]. In total, 17 nAChR subunits have been identified to date and are expressed in various cells, including PC12 cells and rat cerebellum neurons [52–54]. We have previously reported that activated CaMK-IIα phosphorylates nNOS at Ser847 . Therefore we assume that stimulation of nAChR by nicotine results in phosphorylation of nNOS at Ser847. Indeed, the present study showed that nicotine enhanced the phosphorylation of nNOS at Ser847, which was significantly blocked by KN-93 (an inhibitor of CaMKs) in CGNs (Figures 5A and 5B). Nicotine exerts a variety of pharmacological actions in neuronal systems. Although previous reports suggested that NO generated by NOS is dependent on nicotine [55–57], it has not been determined whether nicotine exposure to cells induces ROS generation by nNOS. In the present study, we show for the first time that ROS were generated by nNOS in nicotine-treated nNOS-expressing PC12 cells and CGNs (Figures 3B, 3C, 5C and 5D). We also found in CGNs that nicotine induced nNOS phosphorylation at Ser847 and that nicotine-induced ROS generation was attenuated by the CaMK inhibitor KN93, with concomitant inhibition of phosphorylation at Ser847 (Figure 5). These results suggest that nicotine-induced ROS generation by nNOS is regulated by phosphorylation at Ser847 by CaMKs in neuronal cells. However, KN93 has been found to non-specifically block Ca2+ channels in PC12 cells . In addition, Kajiwara et al.  have reported significant inhibitory effects of KN93 on nicotine-induced Ca2+ influx in nNOS-expressing PC12 cells. As nAChRs are pentamers of subunits, composed of multiple subunits, there is enormous potential for variation of the above-mentioned subunits. Thus the effects of KN93 on nicotine-induced Ca2+ influx may differ among cell types and tissues. Therefore caution should be exercised when using KN93 as CaMK inhibitor.
To assess the biological relevance of nicotine-induced 8-nitro-cGMP production and its regulation by nNOS phosphorylation in CGNs, we investigated nicotine-induced neuroprotection against K+ deprivation-induced apoptosis in CGNs. Cultured CGNs, which have been widely used as a model of neuronal survival, can be maintained in culture medium by elevating extracellular potassium levels (HK; 25 mM) to induce depolarization [59–61]. The depolarization is presumed to mimic endogenous excitatory activity. In fact, when a culture medium containing HK is changed to that containing a low, but more physiological, K+ concentration (5 mM), CGNs undergo apoptotic cell death, mimicking the naturally occurring death of granule cells during cerebellar development in vivo [59,62]. More importantly, such K+ deprivation-induced cell death is known to be caused by ROS-mediated oxidative stress [63,64]. Fucile et al.  reported that nicotine had protective effects against K+ deprivation-induced apoptosis in CGNs; however, the molecular mechanism remains unclear. In the present paper, we demonstrated that nicotine enhanced phosphorylation of nNOS at Ser847 in CGNs, and phosphorylated nNOS at Ser847 increased its uncoupling reaction. The nNOS uncoupling reaction may be detrimental to neuronal cells, because uncoupled nNOS may produce a strong oxidant, peroxynitrite, which is in turn involved in neuronal injury . However, we previously reported that 8-nitro-cGMP, generated by peroxynitrite, could induce an antioxidant adaptive response in the cell via the S-guanylation of the redox sensor protein Keap1, which results in Nrf2 activation and expression of cytoprotective genes, including HO-1 . Therefore in the present study we focused on the physiological, rather than toxic, aspect of the uncoupling reaction in terms of 8-nitro-cGMP formation, as regulatory mechanism of NO/ROS redox signalling. The present study demonstrated that stimulation of nAChR with nicotine resulted in 8-nitro-cGMP formation, induction of HO-1 expression which caused cytoprotective effect, and consequently facilitated neuroprotection against apoptosis induced by LK treatment of CGNs. Since HO-1 is an antioxidative enzyme , HO-1 protects CGNs from oxidative stress induced by uncoupling reaction and K+ deprivation-induced cell death. It is also interesting that pre-treatment of CGNs with KN93 attenuated nicotine-induced cytoprotective functions (Figure 7C). This result suggests that nNOS phosphorylation at Ser847 may contribute to regulation of neuronal cell death.
CaMK-IIα-null mice had an infarct size almost twice that of WT mice after transient focal cerebral ischaemia , indicating that CaMK-IIα had a neuroprotective effect under non-physiological condition. However, the exact mechanism underlying the neuroprotection by CaMK-IIα has not been clarified. Since nNOS-knockout mice showed considerably decreased brain damage after cerebral ischaemia , excessive production of NO by nNOS was assumed to aggravate the neuronal injury. The phosphorylation of nNOS at Ser847 by CaMK-IIα in the hippocampus in cerebral ischaemia decreased NO production, thereby controlling the overproduction of harmful NO , which is thus suggested to be a possible neuroprotective mechanism after hippocampal ischaemia. Since there were no reports on the generation of ROS by the uncoupling reaction of phosphorylated nNOS at Ser847, the present study focuses on ROS production from nNOS with or without Ser847 phosphorylation. We demonstrated that phosphorylated nNOS at Ser847 by CaMK-IIα enhanced the uncoupling reaction to alter the ratio of ROS to NO, and thereby induced intracellular 8-nitro-cGMP formation, so that neuronal cell death was blocked by the induction of HO-1 protein via the Keap1–Nrf2 cascade. Therefore the phosphorylation of nNOS could be implicated, at least in part, in the neuroprotective mechanism of CaMK-IIα.
In conclusion, we demonstrated for the first time that phosphorylation of nNOS at Ser847 may increase its ROS generation and change the ratio of NO to ROS to regulate the intracellular formation of 8-nitro-cGMP in PC12 cells and CGNs. Moreover, our results suggested that nicotine may have protective effects against neuronal apoptosis via 8-nitro-cGMP formation and downstream HO-1 expression. In a predictive model shown in Figure 8, we propose that nNOS phosphorylation at Ser847 regulates redox signalling and 8-nitro-cGMP formation and is involved in neuronal cell death of CGNs. Overall, the present study identified, for the first time, a novel role of nNOS phosphorylation in the redox signalling mechanism.
A predictive model for the regulation of redox signalling by nNOS phosphorylation at Ser847 in CGNs
Ca2+/calmodulin-dependent protein kinase
cerebellar granule neuron
DMPO adduct of superoxide
minimum essential medium containing 25 mM K+
haem oxygenase 1
Kelch-like ECH-associated protein 1
minimum essential medium containing 5 mM K+
minimum essential medium
nicotinic acetylcholine receptor
NG-nitro-L-arginine methyl ester
8-nitroguanosine 3′,5′-cyclic monophosphate
neuronal nitric oxide synthase
nitric oxide synthase
PEGylated zinc protoporphyrin
reactive oxygen species
TBS containing 0.05% Tween 20
Shingo Kasamatsu and Hideshi Ihara designed the experiments, analysed and interpreted the results, and wrote the paper. Hideshi Ihara conducted the experiments. Yasuo Watanabe, Tomohiro Sawa and Takaaki Akaike contributed to data interpretation and writing and revision of the paper.
We thank Akira Wadano for his excellent editing of the paper prior to submission.
This work was supported, in part, by the Ministry of Education, Sciences, Sports and Technology (MEXT) Grants-in-Aid for Scientific Research [grant numbers 22500337 and 25430069] Grants-in-Aid for Scientific Research on Innovative Areas (Research in a Proposed Area) [grant number 20117001] and the Japan Science and Technology Agency (JST) Precursory Research for Embryonic Science and Technology (PRESTO) program [grant number 09801194] and MEXT-Supported Program for the Strategic Research Foundation at Private Universities (2013–2017) (to Y.W.) and a grant from Takeda Science Foundation (to Y.W.).