Nicotine stimulation of cortical neurons obtained from gestation day 19 rats provoked a dose-dependent release of aspartate, glutamate, glycine and GABA, indicating a functional role for the nicotinic receptor in this model. This release was exclusively Ca2+-dependent (vesicular release) in the case of aspartate and dual (Ca2+-dependent and Ca2+-independent) for glutamate, glycine and GABA. Nicotine also raised the membrane potential and the intracellular calcium concentration. These effects were specific, since they were reversed by hexamethonium, an antagonist of the nicotinic receptor. It was shown that L, N, and P/Q type Ca2+ channels are involved in nicotine-mediated Ca2+ entry into cortical neurons. Evaluation of the effects of nicotine on Ca2+ entry in isolated cells showed that 100% of the cells responded to nicotine, although the intensity of the response was variable: 63% of the neurons showed an increase in intracellular Ca2+ of 152 ± 5 grey levels, 25% of 88 ± 12 grey levels and 12% of 48 ± 1 grey levels. Tetrodotoxin, which blocks voltage-dependent Na+ channels, completely reversed nicotine-induced Ca2+ entry into single cells. This suggests that the Ca2+ increment is mediated by opening of Ca2+ channels and not by the nicotinic receptor.
Although its physiological and psychological effects have long suggested that nicotine exerts specific actions in the brain, the identification of neuronal nicotinic receptors has only been possible in the last few years due to the development of molecular biology techniques [reviewed by McGehee and Rolen (McGehee and Rolen, 1995)]. A variety of nicotinic acetylcholine receptor (nAchR) complexes in the central nervous system (CNS) are formed by a diverse array of subunits which confer different pharmacological and physiological properties to the receptors (Sargent, 1993; Rust et al., 1994). Although there is little experimental evidence (Anano et al., 1991; Cooper et al., 1991), the neuronal nAchR channel is assumed to be a pentamer formed by α and β subunits, although many alternative α/β stoichiometries may exist, and this could contribute to the particular functionality of the nAchR channel. The expression of specific nAchR subunits is important during brain development since it influences neuronal excitability (Margiotta and Gurantz, 1989). Nicotinic AchRs are present on autonomic neurons and adrenal chromaffin cells of the peripheral nervous system and on many neurons in the CNS. Many of the properties of the nAchR channels, such as ion selectivity and gating properties, resemble those of muscle AchR. However, neuronal AchRs are clearly distinct from muscle and are themselves diverse. The most important role performed by cholinergic receptors in the CNS is their participation in neuronal excitability modulation (Aquilonius and Gillberg, 1990; Vernino et al., 1992) and in neurotoxicity events (Slotkin et al., 1997). Cumulative evidence from animal and human studies has indicated that nicotinic systems play a major role in higher cognitive functions and dysfunctions. Nakayama et al. were able to show the presence of nicotinic receptors in cerebral cortex through immunocyto-chemical techniques (Nakayama et al., 1995). The number of these receptors seems to be diminished in neurological disorders such as Parkinson's and Alzheimer's diseases (Araujo et al., 1988; Schroder et al., 1995; Chesselli, 1997; Zamanim et al., 1997) and schizophrenia (Miller et al., 1996). Pharmacological and electrophysiological studies appear to suggest the existence of different isoforms of the nicotinic receptor in the neocortex and hippocampus (Wada et al., 1989, 1990; Seguela et al., 1993; Lobron et al., 1995), although the exact location of these isoforms is not known. This diversity of nicotinic receptors in the CNS makes functional and pharmacological variations possible (Papke et al., 1989; Luetje and Patrick, 1991). In dopaminergic nigrostriatal axons nicotine induces dopamine release after previous stimulation of glutamatergic neurons, which release glutamate (García Muñoz et al., 1996). In hippocampal synaptosomes release of noradrenaline may be modulated by nicotinic receptors containing α3 and β4 subunits (Clarke and Reuben, 1996). Furthermore, nicotine may induce increases in glutamate, dopamine, serotonin and acetylcholine release in the hippocampus, cerebellum, nucleous accumbens and striatum (Thoth et al., 1992; Marshall et al., 1996, 1997; Wilkie et al., 1996; Ferger and Kuschinsky, 1997). Conversely, Izenwasser et al. found that in the striatum nicotine inhibits [3H]dopamine uptake and induces its release (Izenwasser et al., 1991). Moreover, Waniewski and Martin reported that the stimulation of sympathetic ganglia with nicotinic or muscarinic agonists provoked the release of [3H]taurine (Waniewski and Martin, 1994). Although there is much data on the properties and functional characteristics of the different cholinergic receptors (AchR) in the CNS, most of these may be considered presynaptic receptors since the action of nicotine has generally been evaluated in terminal neurons (synaptosomes). However, nicotinic receptors are distributed throughout the neurons. Few investigations have centred on the action of AchRs on neuro-transmitter release in neuronal preparations, although the effects of nicotine in in vivo systems have also been studied (Thoth et al., 1993). The aim of the present study was to investigate the nicotinic receptor-mediated modulation of synaptic transmission in cortical neurons. Our results suggest that in cortical neurons in culture nicotine induces aspartate release by an exocytotic pathway and glutamate, glycine and GABA release by a dual mechanism. It was also established that nicotine increased the intracellular calcium concentration in all of the neurons examined.
Materials and Methods
Fura 2AM and bis-[1,3-diethylthiobarbiturate]trimethine oxonol (bis-oxonol) were obtained from Molecular Probes (Eugene, OR). Eagle's minimum essential medium (EMEM) was supplied by Bio-Whittaker and foetal calf serum (FCS) and horse serum (HS) by Sera-Lab (Sussex, UK). Nicotine, hexamethonium and muscimol came from Sigma (St Louis, MA). The remaining chemicals were reactive grade products from Merck (Darmstadt, Germany).
Cell Isolation and Culture
Brain neurons were obtained from foetal rat brains at gestation day 19 following the procedure described by Segal (Segal, 1983) with minor modifications. Isolated neurons were suspended in EMEM containing 0.3 g/l glutamine, 0.6% glucose, 10% FCS, 10% HS, 100 U/ml penicillin, 100 μg/ml streptomycin, 40 μg/ml gentamycin and 5 μg/ml imipenem. Cells, at a density of 2 × 106 cells/well, were plated onto plastic Petri dishes treated with 10 μg/ml poly-l-lysine to aid attachment. The plates were incubated in a humidified incubator in an atmosphere of 5% CO2/95% air at 37°C. After 72 h the incubation medium was replaced with fresh medium to which 10 μM cytosine arabinoside was added to prevent overgrowth of contaminating glial cells. Cells were used after 10–15 days culture. Cell viability was checked by the trypan blue exclusion method. Viability was routinely >95%. Cell purity was checked by both cell staining with cresyl violet to identify neurons and with a specific anti-GFAP antibody to identify glial cells.
After 10–15 days culture the cortical neurons were detached from the culture plates with trypsin, as indicated below. Cells were the fixed (for 30 min) in 2% paraformaldehyde and washed in phosphate-buffered saline (PBS) followed by treatment (1 h) with anti-rabbit GFAP antibody (diluted 1/500). Cells were once again washed in PBS and treated with anti-rabbit FITC-conjugated IgG at a dilution of 1/100 for 30 min and identified by flow cytometry. Under these conditions the glial cells in the cultures were estimated at 9 ± 3% of the total cell population (neural + glial cells).
Changes in intracellular calcium concentration, [Ca2+]i, were monitored by Fura 2AM fluorescence. Six day cultured cells were detached from the plates using trypsin (0.25% trypsin and 0.02% EDTA in Dulbecco's phosphate-buffered saline without calcium or magnesium) and washed twice using 1 ml of a Krebs HEPES solution (Locke medium) containing 140 mM NaCl, 4.4 mM KCl, 2.5 mM CaCl2, 1.2 mM Mg(SO4)2, 1.2 mM KH2PO4, 4.0 mM NaHCO3, 5.5 mM glucose, 0.58 mM ascorbic acid and 10 mM HEPES, adjusted to pH 7.5 and incubated with 5 μM Fura 2AM for 45 min at 37°C. Excessive dye was removed by washing the cells twice with fresh Locke medium followed by suspension in this medium at 1 × 106 cells/ml. After Fura 2AM treatment, cell viability was checked as indicated previously. Fluorescence (excitation wavelength 340/380 nm, emission 510 nm) was monitored at 37°C in a well stirred cuvette containing 1 ml of this suspension using a Perkin Elmer LS-50 spectro-fluorimeter (slits 5 nm excitation, 10 nm emission). At the end of each experiment 1% Triton X-100 was added to make the cells permeable and permit the dye to gain access to the extracellular Ca2+ (2.5 mM). This Ca2+ concentration saturated the dye and provided a measure of the maximum fluorescence signal (Fmax). To determine the minimum fluorescence signal (Fmin), 20 mM Tris base was added to raise the pH above 8.2, followed by 5 mM EGTA which reduced the Ca2+ to <1 nM. [Ca2+]i was calculated with the Grynkiewicz equation (Grynkiewicz et al., 1985):
where F is the ratio between fluorescence values at 340 and 380 nm, SF2 is the maximum fluorescence at 380 nm and SB2 is the minimum fluorescence at 380 nm. The equilibrium dissociation constant (Kd) for the complex [Ca2+]–Fura 2AM was of 224 nM.
Single Cell [Ca2+]i
Cells cultured on glass coverslips were loaded for 45 min with Fura 2AM (5 μM) dissolved in Locke medium. After incubation at 37°C in the dark, cells were washed with Locke medium and the coverlips placed under a Nikon inverted stage microscope under continuous perfusion with Locke medium. Light from a xenon lamp was filtered through two different band-pass filters (340 or 380 nm) in the excitation path and the specimen illuminated on the microscope stage by a dichroic mirror. Excitation wavelengths of 340 and 380 nm were alternately applied to the cells. The fluorescence emitted by the cells was passed through a band-pass filter (510 nm) and video images were obtained using a cold intensified camera. Ratios corresponding to 340/380 nm are given as grey levels. Output from the camera was digitalized and stored in a computerized imaging system (MiraCal). The reagents dissolved in Locke medium were applied in the perfusion medium. When tetrodotoxin (TTX) was used the order of nicotine and TTX applications was as follow. First neurons were stimulated with 10 μM nicotine and images were taken over 80 s. Then, cells were perfused for 10 min to permit recovery of the nicotinic receptor. Subsequently, neurons were stimulated with 10 μM nicotine plus 50 nM TTX and images taken for 80 s as before.
Changes in the membrane potential of neurons were monitored using the fluorescent dye bisoxonol. This is a lipophilic anion whose distribution across the membrane is dependent upon the membrane potential. Thus, an increase in bisoxonol fluorescence indicates that the membrane has been depolarized, allowing more of this negatively charged dye to enter the cells (Waggoner, 1979). Washed cells, as used for [Ca2+]i determination, were suspended in Locke medium at a density of 5 × 105 cells. Neurons in suspension were incubated with 0.2 μM bisoxonol for 10–20 min and placed in a fluorimeter. Fluorescence was measured at an excitation wavelength of 540 nm and emission wavelength of 565 nm and monitored at 37°C in a well stirred cuvette using a Perkin Elmer spectrofluorimeter. Drugs were added at the indicated concentrations. Controls were performed using Locke medium in place of the drug. Fluorescence intensity is reported in arbitrary units.
Amino Acid Secretion
High performance liquid chromatography (HPLC) of amino acids was performed according to the methods described by Márquez et al. (Márquez et al., 1986). Cells were washed twice at 10 min intervals with 1 ml of Locke medium. After removal of the medium cells were stimulated for 15 min periods at 37°C with 0.5 ml of fresh Locke medium containing the different secretagogues. The stimulating medium was then withdrawn and cells ruptured by the addition of 0.5 ml of distilled water. The concentration of amino acids was determined by reversed phase HPLC using pre-column derivation with dansyl chloride and UV detection at 254 nm. Integration of peaks was achieved using a Sprectraphysis integrator. Peaks were quantified by comparison with those obtained using simultaneously prepared amino acid standards. Separation of dansyl derivatives was performed using a 5 μM Spherisorb-ODS-2 column (15 × 0.46 cm).
Proteins were identified according to Bradford (Bradford, 1976). Results were expressed as nmol neurotransmitter/mg protein/well or as the percentage of amino acids released into the incubation medium with respect to the total amino acid content (incubation medium + cells).
Data are presented as the means of three or four separate experiments performed on different cell cultures. Each experiment was performed in duplicate using different batches of cells. Student's t-test was used to statistically compare data.
A dose-dependent release of the amino acids aspartate, glut-amate, glycine and GABA was shown when cortical neurons were stimulated with nicotine. The effect of nicotine was first observed at a nicotine dose of 50 μM. Glycine was released in greatest quantity (Fig. 1). Measurement of secretion mediated by nicotine in a medium with and without external calcium revealed that aspartate release was exclusively Ca2+-dependent (vesicular release), while the release of the remaining amino acids occurred through Ca2+-dependent and Ca2+-independent processes. The order of Ca2+-dependent release of these amino acids at a nicotine concentration of 200 μM was: aspartate ≃ glycine > glutamate > GABA. The Ca2+-dependent release of all of these amino acids was higher than that observed in calcium-free medium, except in the case of GABA, which was preferentially released by a Ca2+-independent process (Fig. 2).
The nicotinic acetylcholine receptor antagonist hexamethonium completely blocked the release of these amino acid neuro-transmitters evoked by nicotine, which implies that the effect may be attributed to activation of the nicotinic receptor (Fig. 3).
Nicotine also increased both the membrane potential, measured as arbitrary fluorescence units, and intracellular calcium levels in a dose-dependent manner. Both effects were inhibited by hexamethonium (Fig. 4A,B).
Release of the four amino acid neurotransmitters evoked by 200 μM nicotine was inhibited by verapamil, an antagonist of L type Ca2+ channels; ω-conotoxin GVIA, an antagonist of N type Ca2+ channels, inhibited the release of all the amino acid neuro-transmitters with the exception of aspartate. Further, ω-agatoxin GIVA, which blocks P/Q type Ca2+ channels, inhibited the release of aspartate, glycine and GABA, but not of glutamate (Fig. 5).
When the nicotine-mediated increments in intracellular calcium levels were measured in single cells it was observed that all of the neurons responded to nicotine, although these responses varied depending on the cell. Increases in intracellular Ca2+ levels of 152 ± 5 grey levels were recorded in 63% of the cortical neurons, of 90 ± 10 in 25% and of 48 ± 1 in 12% (Fig. 6).
Despite abundant data on AchR diversity in the CNS, there is still little evidence for classical nicotinic synaptic transmission. The results presented here clearly demonstrate that when cortical neurons are stimulated with concentrations of nicotine from 50 to 200 μM there is a dose-dependent release of aspartate, glutamate, glycine and GABA, which is exclusively Ca2+-dependent (exocytotic release) or dual (Ca2+-dependent and Ca2+-independent), depending on the amino acid. The nicotine concentration needed to induce amino acid release appears high compared with the findings of electrophysiological studies (McGehee et al., 1995). However, it is much lower than the concentrations used (1 mM or higher) in studies in which amino acid release was measured directly (Thoth et al., 1993). Besides its effects on amino acid neurotransmitter release, nicotine also induced membrane depolarization and intracellular Ca2+ increases. All these effects were blocked by hexamethonium, an antagonist of the nicotinic receptor, indicating their specificity. Intracellular calcium increases mediated by this cholinergic receptor were also shown by Mulle et al. in rat CNS neurons (Mulle et al., 1992). Further, Letz et al. observed that the stimulation of cholinergic receptors on pinealocytes may cause membrane depolarization and activation of L type Ca2+ channels (Letz et al., 1997).
The data presented not only demonstrate the presence of nicotinic receptors in cultured cortical neurons, but also indicate their specificity and functionality. These findings are in accordance with those obtained by Vidal and Changeus, who reported an increase in the negative wave of field potentials reflecting increased excitability of cortical neurons when neocortical slices were treated with acetylcholine or dimethylphenyl piperazinium (Vidal and Changeus, 1989).
In the present study nicotine was also able to induce the release of amino acid neurotransmitters in a calcium-free medium (Ca2+-independent release). This Ca2+-independent amino acid release may be attributable to the reverse action of amino acid transporters, given that the nicotinic receptor is a Na+-permeable channel and, when open, the resulting increased intracellular Na+ level is a condition required to activate amino acid transporters.
Our data seem to indicate that different Ca2+ channels might be involved in release of the different amino acid neurotrans-mitters evoked by nicotine, since blockers of L, N and P/Q type Ca2+ channels inhibited release of the inhibitory amino acid neurotransmitters (glycine and GABA). However, N type Ca2+ channel blockers did not affect aspartate release and P/Q type Ca2+ channel blockers did not affect release of glutamate. These data would appear to suggest that opening of one Ca2+ channel or another is important in secretion of the different amino acid neurotransmitters mediated by nicotine.
The fact that nicotine-stimulated cortical neurons release excitatory (aspartate and glutamate) and inhibitory (glycine and GABA) amino acids would seem to indicate that nicotinic receptors are able to modulate excitatory and inhibitory synapses in cortical neurons.
An observation that warrants particular attention was the high release of aspartate produced when cortical neurons were stimulated with high concentrations of nicotine. This might indicate that in this part of the brain the toxic effect attributed to overexcitability produced by nicotine may be due not only to glutamate release but also to aspartate release that, like glutamate, is able to bind to NMDA receptors. The high release of glycine mediated by nicotine is also of note. Glycine release was some four times higher than GABA release. As both these amino acids are inhibitory, it may be considered that in nicotine-stimulated cortical neurons glycine may mediate inhibition or modulate the response of the NMDA receptor. This matter requires further study.
It may be inferred from the study of Ca2+ entry mediated by nicotine in single neurons that most cortical neurons have nicotinic receptors, although the possibility that Ca2+ entry could be mediated by excitatory amino acids released by nicotine should not be discarded. The increment in intracellular Ca2+ mediated by nicotine could be due to Ca2+ entry through: (i) voltage-dependent Ca2+ channels, since nicotine induced membrane depolarization, and/or (ii) Ca2+ entry through the nicotinic receptors, since, according to some authors, neuronal AchRs show significant permeability to Ca2+ (Adams and Nulter, 1992; Vernino et al., 1992, 1994). This Ca2+ permeation is sufficient to activate Ca2+-dependent cellular processes. However, in the presence of TTX, which completely blocked the Ca2+ entry mediated by 10 μM nicotine, the intracellular Ca2+ increment evoked by nicotine appeared to be mediated by the opening of Ca2+ channels and not by the nicotinic receptor. In this case the nicotinic receptors in cortical neurons seems to be impermeable to Ca2+ ions. These results agree with those of Lena and Changeux, which showed that the depolarizing effect of nicotine in thalamic neurons appears to be mediated through entry of calcium through voltage-sensitive calcium channels (Lena and Changeux, 1997). This is not surprising given the two major functional properties of neuronal nAChRs established by Vernino et al. (Vernino et al., 1992). These authors described nAChRs which show substantial permeability to Ca2+ and a further population of receptors which do not appear to permit Ca2+ movement. These differences may be attributed to the high degree of heterogeneity of neuronal nAChRs brought about by the combination of different subunits to give rise to many structural and functional variants of this neuronal receptor (Boulter et al., 1987; Duvoisin et al., 1989; Couturier et al., 1990).
Herrero et al. recently demonstrated that the Ca2+ channel blockers used in the present study also block α7 and α3β4 AchRs in chromaffin cells (Herrero et al., 1999), giving rise to the possibility that the present effect could be due to blockade of the nAchR. However, our results indicate that the effect of the toxins is mediated by Ca2+ channel blockade given that (i) the toxins show different sensitivity in inhibiting release of the different amino acid neurotransmitters and (ii) TTX, a voltage dependent Na+ channel blocker, completely abolished the Ca2+ entry mediated by 10 μM nicotine.
It may be concluded that: (i) cortical neurons contain functional nicotinic receptors since when stimulated with nicotine these neurons release aspartate, glutamate, glycine and GABA; (ii) the mechanism by which nicotine induces amino acid release is exocytotic or dual, depending on the amino acid; (iii) the effect of nicotine is specific since it is blocked by hexamethonium; (iv) L, N and P/Q type Ca2+ channels are involved in the nicotine effect.
This work was supported by CAICYT grants PM98-0121 and CAM 08.8/0012/1998. E.L. is the recipient of fellowships from the Ministerio de Educación y Ciencia. S.V. is the recipient of a fellowship from UC. We thank M. García Mauriño for helping us with culture preparation.
Address correspondence to M.P. González, Instituto de Bioquímica (Centro Mixto CSIC-UCM), Facultad de Farmacia, Ciudad Universitaria, 28040 Madrid, Spain. Email: email@example.com