Neurochemical Research, Vol. 29, No. 1, January 2004 (© 2004), pp. 189–197

Taurine as a Modulator of Excitatory and Inhibitory Neurotransmission* Abdeslem El Idrissi1 and Ekkehart Trenkner1,2 (Accepted August 7, 2003)

We present data that summarize our findings on the role of taurine in the central nervous system and in particular taurine’s interaction with the inhibitory and excitatory systems. In taurine-fed mice, the expression level of glutamic acid decarboxylase (GAD), the enzyme responsible for GABA synthesis, is elevated. Increased expression of GAD was accompanied by increased levels of GABA. We also found in vitro, that taurine regulates neuronal calcium homeostasis and calciumdependent processes, such as protein kinase C (PKC) activity. This calcium-dependent kinase was regulated by taurine, whereas the activity of protein kinase A (PKA), a cAMP-dependent, calciumindependent kinase, was not affected. Furthermore, as a consequence of calcium regulation, taurine counteracted glutamate-induced mitochondrial damage and cell death.

abundant in cerebral cortex, diencephalon, and cerebellum, glycine is an important inhibitory transmitter in the brain stem and spinal cord (10–12). Functional GABAergic neurons and GABAA receptors are expressed at very early stages of brain development. However, because of the chloride gradient, the activation of GABAA receptors produces a robust depolarization. Furthermore, in the neonatal brain, the main ionotropic receptors (GABAA, NMDA, and AMPA/KA) display a sequential developmental pattern of participation in neuronal excitation (13). Although GABA is the main inhibitory transmitter in the adult, it provides the main excitatory drive to neurons at early stages of development. This early depolarizing action of GABA is due to a chloride gradient that leads to depolarization of young neurons rather than the hyperpolarization observed in adults. As development proceeds, neuronal intracellular chloride decreases and GABA becomes inhibitory (13). During this developmental stage, when GABA is excitatory, taurine might play a critical role as a modulator of neuronal excitability through calcium regulation and thus compensate for the lack of receptormediated neuronal inhibition. The neonatal brain contains high levels of taurine (14–16). As the brain matures, its taurine content declines

INTRODUCTION Glutamate is the major excitatory neurotransmitter in the brain (1). Its functions are mediated via different receptor subtypes. Activation of glutamate receptors causes extracellular calcium influx and mobilization of additional calcium from intracellular stores (2). Calcium serves physiologically important functions as second messenger (3). However, excessive elevation of intracellular calcium levels results in structural damage to neurons (4–7). Thus the control of intracellular calcium concentrations is fundamental for neuronal survival and function. On the other hand, inhibitory neurotransmission is mediated by the neuroactive amino acids GABA and glycine, which are structurally similar to taurine. These are the major inhibitory neurotransmitters in the mammalian CNS (8–10). Although GABAergic synapses are * Special issue dedicated to Dr. Herminia Pasantes-Morales. 1 New York State Institute for Basic Research in Developmental Disabilities and The Center for Developmental Neuroscience at The City University of New York, Staten Island, New York 10314. 2 Address reprint requests to: Institute for Basic Research in Developmental Disabilities, 1050 Forest Hill Road, Staten Island, New York 10314. Tel: 718-494-5249; Fax: 718-494-0622; E-mail. trenkner@ hotmail.com

189 0364-3190/04/0100–0189/0 © 2004 Plenum Publishing Corporation

190 and reaches stable adult concentrations that are second to those of glutamate, the principal excitatory neurotransmitter in the brain. Taurine levels in the brain significantly increase under stressful conditions (17), suggesting that taurine may play a vital role in neuroprotection. We have suggested that a possible mechanism of taurine’s neuroprotection lies in its calcium modulatory effects and agonistic role on glycine and GABAA receptors. The considerable pool of taurine in the brain must have a functional significance. We have previously shown both in vitro and in vivo that taurine modulates both excitatory and inhibitory signals. Taurine works concomitantly with GABA to activate GABAA receptors, thus enhancing neuronal inhibition. Taurine also acts downstream of glutamate receptor activation through the regulation of cytoplasmic and intramitochondrial calcium homeostasis (18–22), thereby preventing neuronal damage associated with excitotoxicity. These two mechanisms indicate that taurine achieves the same purpose of modulating neuronal excitability through at least two independent mechanisms: direct enhancement of GABAergic function and indirect depression of glutamatergic neurotransmission.

EXPERIMENTAL PROCEDURE Cerebellar Granule Cell Culture. Cerebellar granule cells were prepared from 7-day-old mice as previously described (20,22,23). Briefly, the entire cerebellum was removed and single cell suspensions prepared by trypsinization and trituration in 1% trypsin in Ca2⫹/Mg2⫹ -free isotonic phosphate buffer (CMF-PBS). Cells were washed in CMF-PBS and resuspended in culture medium (MEM), supplemented with 0.25% glucose, 2 mM glutamine, 10% HS, 5% FCS, and 25 U/ml of both penicillin and streptomycin. Cells were seeded into poly-Dlysine (PDL) coated dishes and incubated at 37°C in a moist chamber under 5% CO2. After 24 h in vitro the medium was replaced with serum-free medium containing 15% N-2 supplement. The mitotic inhibitor Ara C was added during medium exchange (2 ␮M); this curtailed the number of astrocytes that develop in cultures. The cultures were maintained in a humidified CO2/air (5%/95%) atmosphere at 37°C and monitored daily by phase contrast microscopy. 36 ⫺ Cl Influx Measurements. Cerebellar granule cells were cultured for 5 days in vitro. Growth medium was removed, and cultures were washed three times with Earl’s balanced salt solution (EBSS). After equilibration in 1 ml of EBSS for 20 min at 37°C, 36Cl⫺ influx measurements were initiated by replacing this solution with an identical one containing 5 ␮Ci/ml 36Cl⫺, 50 ␮M GABA or taurine. Cells were incubated in this mixture for 10 s. Free radioligand was removed by three rapid (5 s) rinses with ice-cold of EBSS. The radioactivity was extracted using 2 ⫻ 0.5 ml distilled water that lysed the cells, followed by 2 ⫻ 0.5 ml methanol extraction. The lysate was counted by standard liquid scintillation spectrometry. Western Blotting for GAD Level. Brains were dissected within 3 min from the sacrifice of the animal and immediately frozen on dry ice. Total soluble and membrane bound proteins were extracted (24).

El Idrissi and Trenkner Ten micrograms of proteins were loaded onto a 12.5% polyacrylamideSDS gel and transferred onto PVDF membranes (Millipore). For GAD detection, the membrane was simultaneously probed with two antibodies: AB1511, a rabbit polyclonal antibody (1:5000 dilution, Chemicon) that recognizes both GAD65 and GAD67 and with anti–␤-actin (1:200000 dilution) as a control for equal protein loading. After incubation with the primary antibodies overnight at 4°C, alkaline phosphatase–linked goat anti-mouse or anti-rabbit secondary antibodies (1:5000 dilution, Sigma) were incubated with primary antibodies followed by the reaction with CDP-Star (NEB) reagent according to the supplier’s instructions. Membranes were exposed to x-ray film for 2 to 20 min. Scans of autoradiograms were analyzed with AIDA software (Raytest). Serial dilutions of a brain protein extract were run in parallel with each set of Western blots. Analyses of these samples were used to identify the linear response range for each antibody. 45 Ca2⫹ Influx Measurements. Cells were washed twice with Mg2⫹-free Locke’s solution (154 mM NaCl, 5.6 KCl, 3.6 mM NaHCO3, 1.3 mM CaCl2, 5.6 mM glucose, and 5 mM HEPES [pH 7.4]). Additions were made to a final volume of 0.25 ml including 2 ⫻ 105 cpm of 45CaCl2, which was added 10 s before the addition of the agonist. After 5 min at room temperature, the cells were rapidly washed three times in 0.5 ml of Locke’s solution containing 10 ␮M MK-801, 1 mM MgCl, and 1 mM EGTA. High concentration of MK-801 and Mg2⫹ were used in the 45Ca2⫹ uptake-assay to ensure fast channel block. EGTA was used to chelate nonspecific membranebound 45Ca2⫹. Finally the total amount of 45Ca2⫹ was determined in the lysate after the cells were dissolved in 0.5 ml of 0.1 M NaOH. In all cases 45Ca2⫹ uptake was measured in triplicate. Protein Kinase C Assay. PKC activity was measured by incorporation of 32P from [␥-32P]ATP into a synthetic peptide from myelin basic protein (MBP) that contains the consensus site for PKC phosphorylation as a specific substrate. Cells were rinsed with PBS, and collected in an extraction buffer (20 mM Tris [pH 7.5], 0.5 mM EDTA, 0.5 mM EGTA, 0.5% Triton X-100, and 2.5 ␮g/ml of the protease inhibitors aprotinin and leupeptin). Cells were then homogenized on ice on a precooled dounce homogenizer and centrifuged for 2 min in a microcentrifuge. To activate PKC, 25 ␮l of the supernatant was added to 5 ␮l of lipid preparation (100 ␮M phorbol 12-myristate 13-acetate [PMA], 2.8 mg/ml phosphatidyl serine, Triton X-100, and 10 ␮l H2O and incubated at room temperature for 20 min. After the 20 min 10 ␮l of [32P]ATP/substrate (250 ␮M Ac-MBP 4-14, 100 ␮M ATP, 5 mM CaCl2, 100 mM MgCl2, 20 mM Tris, pH 7.5) were added to each sample, immediately placed at 30°C, and incubated for 5 min. The reaction was stopped by removing and spotting 25 ␮l from each tube onto a phosphocellulose disc to immobilize 32P-labeled peptide. Unincorporated [32P]ATP was removed by washing with 1% phosphoric acid and water. PKC-specific activity was expressed as pmol/min and was normalized to total protein. cAMP-Dependent Protein Kinase (PKA). Assays were performed according to the supplier’s description (GIBCO BRL, Grand Island, NY, USA) by adding 10 ␮l of cell homogenate to 20 ␮l of assay mixture containing 50 mM Tris (pH 7.5), 10 mM MgCl2, 100 ␮M [32P]ATP, 0.25 mg/ml BSA, and 50 ␮M kemptide (Leu-Arg-Arg-AlaSer-Leu-Gly). The synthetic kemptide contains the consensus site for PKA phosphorylation and was used as a specific substrate. PKA is a cAMP-dependent kinase. We determined the activity of PKA induced by endogenous cAMP “active PKA” then we added exogenous cAMP to activate all the available PKA “total PKA.” Cellular ATP Levels. Cellular ATP levels were measured using a highly sensitive luciferase-based ATP assay, Bioluminescence Assay kit (Promega, Madison, WI, USA) and a luminometer (MLX Microtiter Plate Luminometer, Dynex Technologies, Chantilly, VA,

Taurine as a Modulator of Excitatory and Inhibitory Neurotransmission USA). Because the buffer used for luciferase assays was incompatible with any of the available colorimetric protein assays, we were unable to normalize ATP values to concurrent measures of the total cellular protein. Instead, the total ATP signal from cultured neurons, exposed to various treatments, was expressed in nanomoles. Values were obtained from standard curve dilutions performed for each experiment. A standard curve was created by measuring solutions of known ATP concentrations. Assessment of Cell Survival. Neuronal viability was determined biochemically by double staining the cultured neurons for 15 min with fluorescein diacetate (FDA, 15 ␮g/ml) and propidium iodide (PI) (4.5 ␮g/ml). FDA is converted to an intensely green fluorescent lipophobic product in metabolically active cells through the activation of intracellular esterases, while PI is a DNA-binding dye that enters dead cells through damaged membranes and yields a bright red fluorescent complex. Cellular fluorescence was determined using an ELISA plate reader (CytoFluor, Multiwell plate reader, Series 4000, Millipore, Bedford, MA, USA). For the FDA signal, excitation wavelength was 485 nm and emission wavelength was 530 nm; for PI the excitation wavelength was 530 nm and emission wavelength was 645 nm. Cell survival was determined using the following formula: % Live cells ⫽ {(EF ⫺ BF)/[(CF ⫺ BF) ⫹ CP]} ⫻ 100 where EF is the FDA fluorescence value from live cells in experimental wells; BF represents the FDA fluorescent value obtained from wells where all neurons were killed with 1 M glutamate, representing fluorescense values from the remaining non-neuronal cells (possibly a few fibroblast and glia cells); CF represents the FDA fluorescent value obtained from untreated controls, and CP is the PI fluorescent value obtained from untreated controls. The denominator of the equation represents the total number of neurons. Some cultures were fixed in a mixture of 2% paraformaldehyde and 0.01% glutaraldehyde at 37°C for 15 min. We found that this fixation method best preserves the cell morphology. After fixation, cell viability was determined by phase contrast microscopy. Statistical Analysis. Multifactorial ANOVAs and covariance were used to identify overall condition effects. Significant changes were determined by post hoc comparisons of means using the Tukey HSD test. Significance was set at a confidence level of 95%. Data are presented as mean ⫾ SEM.

191

with lower affinity than their respective agonists, GABA and glycine (26,27). Taurine Activates Chloride Currents Through GABAA Receptors There is increasing evidence supporting the presence of functional interactions between GABA and taurine (27–29). Taurine has been shown to increase plasma membrane chloride conductances by affecting bicucullinesensitive chloride channels (27). Taurine has also been shown to act as a partial agonist of GABAA receptors in synaptic membranes (27). Taurine has been suggested to be the endogenous ligand for glycine receptors during neocortical development (26), and, recently, taurine was shown to activate the corticostriatal pathway through activation of glycine receptors (30). We show here that taurine activates Cl⫺ influx through GABAA receptors in cerebellar granule cells in vitro (Fig. 1) through a direct interaction with GABAA receptors. Consistent with these results, we found that taurine has anticonvulsive effects in the pharmacologically induced seizure model. Taurine (43 mg/kg) significantly reduced the seizure severity when injected 10 min before seizure induction with kainic acid (31). We found that the anticonvulsive effect of taurine was mediated by interaction with the GABAA receptors in vivo and by activation of chloride conductance (31). Taurine Increases GABA Levels in the Brain The close structural and functional similarities of taurine and GABA are not limited to binding to the recognition site of GABA on the GABAA receptor and

RESULTS Interaction of Taurine with the Inhibitory System Exogenous application of taurine inhibits neuronal depolarization; thus taurine was viewed to act as an inhibitory neurotransmitter. However, the calciumindependent release of taurine under potassiumdepolarizing conditions, together with the lack of specific taurine receptors, excluded taurine from a formal classification as neurotransmitter. Nevertheless, the inhibitory actions of taurine are well documented in several brain regions (8,25,26), and have been shown to be mediated through activation of GABAA and glycine receptors, both of which are responsible for fast synaptic inhibition. Taurine has been found to interact with these receptors

Fig. 1. Taurine induces chloride uptake into cerebellar granule cells. Cells were treated with GABA (50 ␮M) or taurine (50 ␮M) and chloride uptake was initiated for 10 s. Both GABA and taurine induced a significant increase in 36Cl⫺ influx. The fact that bicuculline (10 ␮M), a GABAA receptor antagonist, blocked taurineinduced chloride influx indicates that taurine activates chloride currents through GABAA receptor.

192 to the activation of chloride current by taurine, but they extend to alterations of the biosynthetic pathway of GABA. We have shown previously that chronic treatment with taurine affects GABA synthesis (31). Mice fed with taurine in drinking water (0.05% for 4 weeks) show significantly elevated levels of GABA in the brain (31). Consistent with the effect of taurine on GABA synthesis, we found that in taurine-fed mice approximately a twofold increase in the expression of both isoforms of glutamic acid decarboxylase (GAD) (GAD 65 and 67) in the brain occurred as determined by Western blotting (Fig. 2). Interaction of Taurine with Excitatory System Unlike GABAergic and glycinergic systems, the interaction of taurine with the glutamatergic system occurs downstream of glutamate receptor activation. We have previously shown that taurine does not directly interact with any of the glutamate receptors, including the strychnineinsensitive glycine recognition site of the NMDA receptors (20). However, taurine regulates calcium homeostasis in a variety of cell types (14,15,20,32). As a consequence of

Fig. 2. Representative Western blot chemiluminescence of GAD and ␤ actin from brains of control and a taurine-fed mouse. The Western blot was probed simultaneously with a rabbit polyclonal antibody that recognizes both isoforms of GAD (65 and 67 kDa) and a monoclonal antibody that recognizes ␤ actin. Comparison of GAD to actin ratios showed a significant increase in GAD expression in taurine-fed mice (mean ⫾ SEM, 216 ⫾ 14.6; P ⬍ .01)

El Idrissi and Trenkner calcium regulation, taurine regulates several calciumdependent reactions. Here we present data that pertain to neuronal calcium processes and their regulation by taurine. Taurine Downregulates Glutamate-Induced Calcium Uptake Cerebellar granule cells express both glutamate and GABA receptors. Activation of GABAA receptors leads to increased chloride influx (Fig. 1) and hyperpolarization, whereas activation of glutamate receptors leads to sodium and calcium influx, resulting in depolarization. Time course studies of glutamate-induced 45Ca2⫹ uptake showed that depolarization with 1 mM glutamate caused a significant increase in 45Ca2⫹ uptake over time. The linear increase of intracellular 45Ca2⫹ uptake continued over time up to 30 min (Fig. 3). However, in cultures pretreated with taurine (1 mM) for 24 h before the addition of glutamate, 45Ca2⫹ uptake was significantly lower than in control cultures (Fig. 3). Consistent with radioligand uptake studies, using free-cytoplasmic calcium indicators, we showed that in cells pretreated with taurine (1 mM, 24 h) glutamate elicited a rapid and significant increase in free-cytoplasmic calcium concentrations ([Ca2⫹]i) similar to non–taurine treated control cultures. But this increase was not sustained and [Ca2⫹]i recovered slowly toward baseline after 10 min following bath application of glutamate (20,21). Interestingly, however, the calcium regulatory effects of

Fig. 3. Time course of calcium accumulation under glutamatedepolarizing conditions Calcium uptake was determined over time in enriched cerebellar cultures grown in serum-free medium for 5 days in vitro. Cultures were pretreated with taurine (1 mM) for 24 h before exposure to glutamate. 45Ca2⫹ accumulation was determined after depolarization with 1 mM glutamate for the indicated time. Each data point represents the mean ⫾ SEM of at least three sets of separate experiments. Glutamate caused a significant increase in 45Ca2⫹ uptake at 5 min (P ⬍ .05), and thereafter (P ⬍ .0001). However, when cultures were pretreated with taurine, a significant reduction of calcium accumulation was observed 15 min after depolarization (P ⬍ .05) (from Ref. 20).

Taurine as a Modulator of Excitatory and Inhibitory Neurotransmission taurine could not be elicited if taurine was applied simultaneously with glutamate (20), indicating that taurine does not interfere with the mechanisms of calcium influx (such as interaction with glutamate receptors), but rather taurine affects downstream events involved in maintaining calcium homeostasis. Furthermore, we showed that taurine regulates mitochondrial calcium homeostasis (20,21), because these organelles play a crucial role in calcium sequestration in CGCs. Therefore it was not surprising to observe that calcium regulation in the mitochondria by taurine preserved mitochondrial function (20,21).

Taurine Regulates the Activity of CalciumDependent Kinases PKC has been shown as one of the calciumactivated enzymes linking glutamate receptor activation to excitotoxicity (33–35). Exposure to glutamate resulted in a significant increase in PKC activity, suggesting that prolonged translocation of PKC may be related to glutamate neurotoxicity. Consistent with this, downregulation of PKC achieved almost complete neuroprotection from glutamate toxicity (34,36). We investigated the effects of taurine on the activity of this enzyme in the presence of glutamate. As has been reported earlier (34,35), glutamate caused a significant increase in PKC activity (Fig. 4). Pretreatment of the cultures

Fig. 4. Regulation of PKC activity by taurine. Cerebellar granule cells were cultured in serum-free medium for 5 days in vitro, and treated with taurine (1 mM) 24 h before the assay. Cells were depolarized with glutamate in the presence or absence of taurine for 30 min. Culture medium was removed and cells lysed in ice-cold homogenization buffer as described. Data represent the mean ⫾ SEM of three separate experiments. Post hoc tests indicated that glutamate caused a significant increase in PKC activity as compared to controls (P ⬍ .005). Pretreatment with taurine inhibited the glutamate-induced PKC activity, and the resulting activity levels were not statistically different from controls. Addition of MK801 (10 ␮M), 2 min before glutamate depolarization, resulted in activity levels significantly lower than controls (P ⬍ 0.05).

193

with taurine (1 mM, 24 h), however, prevented the glutamate-induced increase in PKC activity. In the presence of taurine alone, PKC activity was lower than untreated controls (Fig. 4). The effects of taurine on kinase activities was limited to calcium-dependent kinases, because pretreatment with taurine did not affect protein kinase A (PKA) activity, a cAMP-dependent kinase (Fig. 5). This activity level was based on endogenous cAMP-induced activity, which is the amount of cytoplasmic cAMP found after treatment with taurine and glutamate. Furthermore, exogenous application of cAMP to the assay mixture, which activated all of the existing cAMP-dependent PKA, showed that the total amount of PAK was not affected by taurine, suggesting that taurine did not affect the expression levels of this enzyme. In summary, taurine actions are limited to the calcium-dependent PKC but not to the cAMP-dependent PKA. Prevention of excessive activation of PKC and subsequent substrate hyperphosphorylation is another mechanism for the neuroprotective effects of taurine against glutamate excitotoxicity.

Taurine Preserves Mitochondrial Function One of the earliest events of excitotoxicity is the impairment of energy metabolism. As previously shown, taurine controls glutamate excitotoxicity through modulation of mitochondrial function (20,21). This was demonstrated by monitoring mitochondrial activity using rhodamine 123 uptake into the mitochondria as an

Fig. 5. PKA expression or activity were not affected by taurine. Cerebellar granule cells were cultured in serum-free medium for 5 days in vitro, and treated with taurine (1 mM) 24 h before the assay. Cells were depolarized with glutamate (1 mM) in the presence or absence of taurine for 30 min. Culture medium was removed and cells lysed in ice-cold homogenization buffer. PKA activity was determined in the absence of exogenously added cAMP “active PKA” or in the presence of added cAMP “total PKA” (see Experimental Procedure). Data represent the mean ⫾ SEM of three separate experiments.

194 indicator for the mitochondrial electrochemical gradient (20). Here we measured cellular ATP levels as another biochemical marker for mitochondrial function. Addition of glutamate resulted in a dose-dependent and rapid depletion of cellular ATP levels (Fig. 6). High levels of glutamate (1 and 0.1 mM) initially induced similar reductions in ATP levels, whereas low concentrations (10 and 1 ␮M) had no effects on ATP levels when cultures were treated with glutamate for 15 min. However, after glutamate was removed and cells were allowed to incubate for an additional 4 h, cultures treated with 0.1 mM glutamate showed a significant recovery of ATP levels, whereas ATP levels in cultures treated with 1 mM glutamate were found to have significantly less ATP than the initial 15 min treatment with glutamate, indicating the irreversibility of the mitochondrial damage. The further decrease in ATP at later time point (4 h) may be attributable in part to cell death that took place during this period (Fig. 7). The presence of taurine (1 mM) did not affect the levels of ATP in the presence of glutamate during the initial phase of treatment. However, ATP levels were significantly higher in taurine-treated

El Idrissi and Trenkner cultures when they were exposed to glutamate for 15 min and allowed to recover for 4 h. Neuroprotective Role of Taurine One functional consequence of regulating intracellular calcium is preserving the mitochondria from the damaging effects of calcium overloads and therefore subsequent neuronal death (37–40). Treatment with glutamate, NMDA, or kainate (100 ␮M) resulted in significant cell death of cerebellar granule cells (Fig. 7). The excitotoxicity was glutamate-specific, because neurons were protected by the NMDA or kainate receptor antagonists, MK801 and DNQX, respectively (Fig. 7). When cells were pretreated with taurine (1 mM, 24 h), excitotoxicity induced by excessive activation of glutamate receptors was significantly reduced. We have, as well as others, previously shown a correlation between elevations of cytoplasmic calcium and subsequent cell death (20,21,40,41). Taurine was more efficient in preventing excitotoxicity when cytoplasmic calcium loads were modest, such as in the presence of kainate as compared to glutamate.

Fig. 6. Taurine induces recovery of cellular ATP levels after glutamate treatment. Cerebellar granule cells were pretreated with taurine (1 mM) for 24 h before stimulation with glutamate. In one experimental group, cultures were treated with glutamate for 15 min and ATP levels were measured (acute). In the second group, cultures were treated with glutamate for 15 min, culture medium was removed, and cultures were allowed to recover for 4 h; then ATP levels were determined (recovery). Each data point represents mean ⫾ SEM of ATP concentrations obtained from three separate experiments. A two-way ANOVA showed significant main effect of the treatment time (acute and recovery) [F(1,32) ⫽ 20.13, P ⬍ .0001] and concentrations of glutamate [F(3,32) ⫽ 769.27, P ⬍ .0001]. The interaction between concentrations of glutamate and treatment time was also significant [F(3,32) ⫽ 82.36, P ⬍ .0001]. Post hoc tests indicated that glutamate at 0.1 and 1 mM induced a statistically significant (P ⬍ .001) decrease in cellular ATP levels in all groups examined, whereas 1 and 10 ␮M had no significant effect. Taurine did not improve ATP levels in cultures treated with glutamate for 15 min, whereas in cultures that were allowed to recover, taurine significantly increased ATP levels.

Taurine as a Modulator of Excitatory and Inhibitory Neurotransmission

Fig. 7. Taurine protects against glutamate excitotoxicity. Cerebellar granule cells were grown in serum-free medium for 4 days in vitro. Cultures were then treated with taurine (1 mM) and incubated for an additional 24 h. Cells were treated with 0.1 mM glutamate, NMDA or kainate for 15 min in the presence or absence of MK801 or DNQX as indicated (10 ␮M, NMDA and kainate receptor antagonists, respectively). After 15 min, drugs were removed by rinsing the cells with growth medium and cultures were returned to the incubator for an additional 4 h. Cells were labeled with fluorescein diacetate (FDA, 15 ␮g/ml) and propidium iodide (PI) (4.5 ␮g/ml) and cellular fluorescence was determined. An ANOVA test showed that treatment of cultures with glutamate, NMDA, or kainate induced a significant decrease in cell viability [F(3,96) ⫽ 569.7, P ⬍ .0001], whereas taurine significantly improved cell viability [F(2,96) ⫽ 625.01, P ⬍ .001]. A two-way ANOVA revealed that the interaction between taurine and glutamate receptor agonists was significant [F(6,96) ⫽ 120.87, P ⬍ .0001].

DISCUSSION Taurine is structurally related to GABA and acts itself as an inhibitory amino acid during development. Here, we review data that pertain to the interaction of taurine with the inhibitory and excitatory systems, particularly the GABAergic and the glutamatergic neurotransmission. The abundance, distribution, and physiological functions of taurine in the organism are ubiquitous. The contrast between its ubiquity and the paucity of clear physiological actions perhaps has attracted many researchers to the field of taurine. Based on the pioneering work and engagement of Dr. Pasantes-Morales and others, the taurine field has been lifted out of obscurity. The biological importance of taurine is becoming established. Levels of taurine are relatively high in neuronal cells at early stages of brain development (14–16). As brain development proceeds, taurine levels decline and reach concentrations that are second to those of glutamate, the principal excitatory neurotransmitter in the brain. The considerable pool of taurine in the developing brain must have a functional application. We have, as well as others, previously shown both in vitro and in vivo that taurine modulates both excitatory and inhibitory signals (16,20, 25,28,42). Taurine works concomitantly with GABA to activate GABAA receptors, thus enhancing neuronal inhi-

195

bition. Taurine also acts downstream of glutamate receptor activation through the regulation of cytoplasmic and intramitochondrial calcium homeostasis (20,21), therefore preventing neuronal damage associated with excitotoxicity. These two mechanisms indicate that taurine achieves the same goal: modulation of neuronal excitability through at least two independent means, direct enhancement of GABAergic function and indirect depression of glutamatergic neurotransmission. GABA, the main inhibitory transmitter in the adult, provides the main excitatory drive to neurons at early stages of development because of a chloride gradient that leads to depolarization of young neurons rather than the hyperpolarization observed in adults. During this developmental stage, when GABA is excitatory, taurine may play a critical role as a modulator of neuronal excitation by compensating for the lack of receptormediated neuronal inhibition. By contrast to GABA and glycine, taurine regulates calcium homeostasis and thus neuronal excitability. The early “critical periods” of brain development correspond to the time of maximal synapse elaboration, complex pathway formation, and high seizure sensitivity (13,43). This increased excitability in the developing brain appears to be secondary to a developmental imbalance between maturation of excitatory and inhibitory circuits. We suggest that taurine plays a critical role in brain development during this transient and critical stage through its unique interaction with both the excitatory and inhibitory system. Several pharmacological and electrophysiological studies indicate that taurine induces an increase in chloride conductance that is sensitive to GABAA and glycine receptor antagonists (25–27,44). However, the interaction site of taurine with these receptors has yet to be established. Consistent with these studies, we report here that taurine induces chloride influx into CGCs in vitro similar to that induced by GABA. Taurine-induced chloride influx was inhibited by the GABAA receptor antagonist, bicuculline (Fig. 1), indicating that taurine directly interacts with GABAA receptors in vitro. We also found that subcutaneous injection of taurine (43 mg/kg) significantly reduced the severity of kainic acid–induced seizures and subsequent neuronal death (31). Furthermore, the interaction of taurine with the GABAergic system in vivo extends beyond the activation of GABAA receptors. We found that chronic treatment (4 weeks) of mice with taurine in drinking water (0.05%) resulted in a significant increase in brain GABA levels (31). Concomitant with this increase of GABA levels, the expression of glutamic acid decarboxylase (GAD) was also elevated (Fig. 2). However, it remains to be determined if the activity of GAD is also increased. It is not clear, at present, if the

196 increased GAD levels are directly related to taurine treatment or just a secondary adaptive mechanism to changes in the function of the inhibitory system in response to chronic treatment with taurine. Chronic exposure of GABAA receptors to taurine, in addition to the endogenous agonist GABA, may affect the normal function of these receptors, and as compensatory mechanisms to these changes in the inhibitory system there is an increase in GABA synthesis. The sequence of events leading to increased GABA and GAD expression levels remain to be elucidated, but the functional significance of increased GABA and GAD levels would be an enhanced inhibition. We used mouse cerebellar granule cells (CGCs) in vitro as a model to investigate the interaction of taurine with the glutamatergic system. CGCs respond to depolarization with a rise in intracellular calcium that is initiated by calcium entry through NMDA receptor channels (5,6,44,45). Pretreatment of CGCs with taurine resulted in a significant reduction in glutamate-induced calcium accumulation (Fig. 3). Although the initial increase in intracellular calcium concentrations was not affected by the presence of taurine after glutamate depolarization, cytoplasmic calcium concentrations returned to baseline after approximately 10 min (20,21). In the CNS, calcium is thought to play a key role in mediating glutamate excitotoxicity (for review see 4–6,45). Glutamate-induced neuronal necrosis is preceded by a rapid increase in cytoplasmic free calcium. There is also increasing evidence that mitochondrial dysfunction plays a primary role in the initiation of both necrotic and apoptotic neuronal cell death (46). Furthermore, after exposure to glutamate, mitochondrial depolarization has been reported as an early event associated with neuronal calcium loading (37–40). We have further examined the effects of taurine on intracellular calcium regulation and its consequences on neuronal function. We found that taurine downregulated glutamate-induced protein kinase C (PKC) activity, but did not affect PKA activity. PKC has been shown as one of the calcium-activated enzymes linking glutamate receptor activation to excitotoxicity (34–36), therefore it serves as an excellent example for taurine’s role as a calcium modulator. Because the neurotoxicity is probably caused by the destabilized intracellular Ca2⫹ homeostasis, it has been suggested that the proteins involved might be the ones regulating the rate of Ca2⫹ influx or efflux (36). It was shown that PKC reduces the Mg2⫹ block of the Ca2⫹-permeable NMDA receptor channel (47), suggesting that the phosphorylation of these receptors, involved in glutamate toxicity, may enhance the NMDA-mediated responses.

El Idrissi and Trenkner In addition to glutamate receptors, cysteine sulfinic acid decarboxylase (CSAD), the rate-limiting enzyme in taurine biosynthesis, is another substrate for PKC phosphorylation. It has been shown that phosphorylation of CSAD by PKC leads to its activation (17), resulting in increased taurine biosynthesis. Consistent with this, taurine levels in the brain have been shown to significantly increase under stressful conditions (17), suggesting that taurine may play a vital role in neuroprotection. The functional consequences of calcium regulation are clearly related to protection from glutamate excitotoxicity (Fig. 7), because several intermediary steps link the initial calcium influx to the resulting neuronal death. We have examined several crucial steps in this pathway and found that taurine regulates intramitochondrial calcium homeostasis (18–21), which preserved mitochondrial function. As shown in Fig. 6, pretreatment of CGCs with taurine resulted in an increase in cellular ATP levels. Because mitochondria are integral organelles for neuronal function, preserving their function proved to be neuroprotective (Fig. 7)

CONCLUSION We suggest a dual cellular function for taurine: modulation of excitatory and inhibitory neurotransmission. Although these two systems are differentially affected by taurine, the net result is enhanced inhibition and reduced excitation. The culmination of these two functions, and possibly others, might be the mechanisms through which taurine regulates neuronal development and survival and protects against glutamate excitotoxicity.

ACKNOWLEDGMENTS This work was supported in part by funds from New York State Office of Mental Retardation and Developmental Disabilities and the Center of Developmental Neuroscience and Developmental Disabilities.

REFERENCES 1. Watkins, J. C. and Evans, R. H. 1981. Excitatory amino acid transmitter. Ann. Rev. Pharmacol. Toxicol. 21:165–204. 2. Jaffe, D. B. and Brown, T. H. 1994. Metabotropic glutamate receptor activation induces calcium waves within hippocampal dendrites. J. Neurophysiol. 72:471–474. 3. Kater, S. B., Mattson, M. P., Cohan, C., and Connor, J. 1988. Calcium regulation of the neuronal growth cone. TINS 11:315–321. 4. Mattson, M. P. 1994. Calcium and neuronal injury in Alzheimer’s disease. Contributions of beta-amyloid precursor protein mismetabolism, free radicals, and metabolic compromise. Ann. NY Acad. Sci. 15:50–76.

Taurine as a Modulator of Excitatory and Inhibitory Neurotransmission 5. Hartley, D. M., Kurth, M. C., Bjerkness, L., Weiss, J. H., and Choi, D. W. 1993. Glutamate receptor-induced 45Ca2⫹ accumulation in cortical cell cultures corelates with subsequent neuronal degeneration. J. Neurosci. 13:1993–2000. 6. Choi, D. W. 1992. Excitotoxic cell death. J. Neurobiol. 23: 1261–1276. 7. Mattson, M. P. 1992. Calcium as sculptor and destroyer of neural circuitry. Exp. Gerontol. 27:29–49. 8. Curtis, D. R. and Johnston, G. A. R. 1974. Amino acid transmitter in the mammalian central nervous system. Ergbn. Physiol. 69:97–188. 9. Kelly J. S. and Krnjevic, K. (1969) The action of glycine on cortical neurons. Exp Brain Res. 9:155–163. 10. Aprison, M. H., Shank, R. P., Davidoff, R. A., and Werman, R. 1968. The distribution of glycine, a neurotransmitter suspect in the central nervous system of several vertebrate species. Life Sci. 7:583–590. 11. Betz, H. 1987. Biology and structure of the mammalian glycine receptor. Trends Neurosci. 10:113–117. 12. Langosch, D., Becker, C.-M., and Betz, H. 1990. The inhibitory glycine receptor: A ligand-gated chloride channel of the central nervous system. Eur. J. Biochem. 194:1–8. 13. Ben-Ari, Y. 2002. Excitatory actions of GABA during development: The nature of the nurture. Nature Neurosci. 3:728–739. 14. Huxtable, R. J. 1992. The physiological actions of taurine. Physiol. Rev. 72:101–163. 15. Sturman, J. A. 1993. Taurine in development. Physiol. Rev. 73: 119–147. 16. Huxtable, R. J. 1989. Taurine in the central nervous system and the mammalian actions of taurine. Prog. Neurobiol. 32:471–533. 17. Wu, J. Y., Tang, X. W., Schloss, J. V., and Faiman, M. D. 1998. Regulation of taurine biosynthesis and its physiological significance in the brain. Adv. Exp. Med. Biol. 442:339–345. 18. Trenkner, E., El Idrissi, A., and Harris, C. 1996. Balanced interaction of growth factors and taurine regulate energy metabolism, neuronal survival, and function of cultured mouse cerebellar cells under depolarizing conditions. Adv. Exp. Med. Biol. 403:507–517. 19. El Idrissi, A., Harris, C., and Trenkner, E. 1998. Taurine modulates glutamate- and growth factors-mediated signaling mechanisms. Adv. Exp. Med. Biol. 442:385–396. 20. El Idrissi, A., and Trenkner, E. 1999. Growth factors and taurine protect against excitotoxicity by stabilizing calcium homeostasis and energy metabolism. J. Neurosci. 19:9459–9468. 21. El Idrissi, A., and Trenkner, E. (2003) Taurine regulates mitochondrial calcium homeostasis. Pages 527–536, in Lombardini, J. B., Schaffer, S. W. and Azuma, J. (eds.), Taurine 5 beginning the 21st century, Adv. Exp. Med. Biol., Vol. 526, New York: Kluwer. 22. Trenkner, E., and Sidman, R. L. 1977. Histogenesis of mouse cerebellum in microwell cultures: Cell reaggregation and migration, fiber and synapse formation. J. Cell Biol. 75:915–940. 23. Trenkner, E. 1991. Culturing nerve cells. Pages 283–307, in Banker, G. and Goslin, K. (eds.), Cerebellar cells in culture, MIT Press. 24. Benke, D., Mertens, S., Trzeciak, A., Gillessen, D., and Mohler, H. 1991. GABAA receptors display association of gamma 2-subunit with alpha 1- and beta 2/3-subunits. J. Biol. Chem. 266: 4478–4483. 25. Ye, G. L., Tse, A. C. O., and Yung, W. H. 1997. Taurine inhibits rat substantia nigra pars reticulata neurons by activation of GABA- and glycine-linked chloride conductance. Brain Res. 749:175–179. 26. Flint, A. C., Liu, X., and Kriegstein, A. R. 1998. Nonsynaptic GlyR activation during early neocortical development. Neuron 20:43–53. 27. Quinn, M. R. and Harris, C. L. 1995. taurine allosterically inhibits binding of [35S]-t-butylbicyclophosphorothionate (TBPS) to rat brain synaptic membranes. Neuropharmacology 34:1607–1613.

197

28. Kuriyama, K. and Hashimoto, T. 1998. Interrelationship between taurine and GABA. Adv. Exp. Med. Biol. 442:329–337. 29. Frosini, M., Sesti, C., Dragoni, S., Valoti, M., Palmi, M., Dixon, H. B., Machetti, F., and Sgaragli, G. 2003. Interactions of taurine and structurally related analogues with the GABAergic system and taurine binding sites of rabbit brain. Br. J. Pharmacol. 138:1163–1171. 30. Chepkova, A. N., Doreulee, N., Yanovsky, Y., Mukhopadhyah, D., Haas, H. L., and Sergeeva, O. A. 2002. Long-lasting enhancement of corticostriatal neurotransmission by taurine. Eur. J. Neurosci. 16:1523–1530. 31. El Idrissi, A., Messing, J., Scalia, J., and Trenkner, E. 2003. Prevention of epileptic seizures through taurine. Pages 515–525, in: Lombardini, J. B., Schaffer, S. W. and Azuma, J. (eds.), taurine 5 beginning the 21st century, Adv. Exp. Med. Biol., Vol. 526, New York: Kluwer. 32. Militante, J. D. and Lombardini, J. B. 1998. Pharmacological characterization of the effects of taurine on calcium uptake in the rat retina. Amino Acids 15:99–108. 33. Choi, D. W. 1990. The role of glutamate neurotoxicity in hypoxicischemic neuronal death. Ann. Rev. Neurosci. 13:171–182. 34. Favaron, M., Manev, H., Siman, R., Bertolino, M., Szekely, A. M., DeErausquin, G., Guidotti, A., and Costa, E. 1990. Downregulation of protein kinase C protects cerebellar granule neurons in primary culture from glutamate-induced neuronal death. Proc. Natl. Acad. Sci. USA 87:1983–1987. 35. Eboli, M. L., Mercanti, D., Cioti, M. T., Squinto, A., and Castellani, L. 1994. Glutamate-induced phosphorylation in cerebellar granule cells: Role of protein kinase C. Neurochem. Res. 19:1257–1264. 36. Manev, H., Costa, E., Wroblewski, J. T., and Guidotti, A, 1990. Abusive stimulation of excitatory amino acid receptors: A strategy to limit neurotoxicity. FASEB. J. 4:2789–2797. 37. Isaev, N. K., Zorov, D. B., Stelmashook, E. V., Uzbekov, R. E., Kozhemyakin, M. B., and Victorov, I. V. 1996. Neurotoxic glutamate treatment of cultured cerebellar granule cells induces Ca2⫹-dependent collapse of mitochondrial membrane potential and ultrastructural alterations of mitochondria. FEBS Lett. 392:143–147. 38. Khodorov, B., Pinelis, V., Vergun, O., Storozhevykh, T., and Vinskaya, N. 1996. Mitochondrial deenergization underlies neuronal calcium overload following a prolonged glutamate challenge. FEBS Lett. 397:230–234. 39. White, R. J. and Reynolds, I. J. 1996. Mitochondrial depolarization in glutamate stimulated neurons: An early signal specific to excitotoxin exposure. J. Neurosci. 16:5688–5697. 40. Keelan, J., Vergun, O., and Duchen, M. R. 1999. Excitotoxic mitochondrial depolarization requires both calcium and nitric oxide in rat hippocampal neurons. J. Physiol. 520:797–813. 41. Eimerl, S. and Schramm, M. 1994. The quantity of calcium that appears to induce neuronal death. J. Neurochem. 62:1223–1226. 42. Pasantes-Morales, H., Franco, R., Ochoa, L., and Ordaz, B. 2002. Signaling events during swelling and regulatory volume decrease. Neurochem. Res. 25:1301–1314. 43. Holmes, G. L. and Ben-Ari, Y. 1998. Seizures in the developing brain: Perhaps not so benign after all. Neuron 21:1231–1234. 44. Choi, D. W., Koh, J. Y., and Peters, S. 1988. Pharmacology of glutamate neurotoxicity in cortical cell culture: attenuation by NMDA antagonists. J. Neurosci. 8:185–196. 45. Siesjo, B. K. 1988. Historical overview: Calcium, ischemia, and death of brain cells. Ann. NY Acad. Sci. 522:638–631. 46. Beal, M. F., Hyman, B. T., and Koroshetz, W. 1993. Do defects in mitochondrial energy metabolism underlie the pathology of neurodegenerative diseases? TINS 6:125–131. 47. Chen, L. and Huang, L.-Y. M. 1992. Protein kinase C reduces Mg2⫹ block of NMDA-receptor channels as a mechanism of modulation. Nature 356:521–523.