Journal of Neurochemistry, 2007, 100, 736–746 doi:10.1111/j.1471-4159.2006.04228.x Melatonin prevents glutamate-induced oxytosis in the HT22 mouse hippocampal cell line through an antioxidant effect specifically targeting mitochondria Federico Herrera, Vanesa Martin, Guillermo Garcı́a-Santos*, Jezabel Rodriguez-Blanco, Isaac Antolı́n and Carmen Rodriguez* Departamento de Morfologı́a y Biologı́a Celular and *Instituto Universitario de Oncologı́a del Principado de Asturias. Facultad de Medicina. Oviedo, Spain Abstract The pineal hormone melatonin has neuroprotective effects in a large number of models of neurodegeneration. Melatonin crosses the blood–brain barrier, shows a decrease in its nocturnal peaks in blood with age that has been associated with the development of neurodegenerative disorders, and has been shown to be harmless at high concentrations. These properties make melatonin a potential therapeutic agent against neurodegenerative disorders but the pathways involved in such neuroprotective effects remain unknown. In the present report we study the intracellular pathways implicated in the complete neuroprotection provided by melatonin against glutamate-induced oxytosis in the HT22 mouse hippocampal cell line. Our results strongly suggest that melatonin prevents oxytosis through a direct antioxidant effect specifically tar- geted at the mitochondria. Firstly, none of the described transducers of melatonin signalling seems to be implicated in the neuroprotection provided by this indole. Secondly, melatonin does not prevent cytosolic GSH depletion-dependent increase in reactive oxygen species (ROS), but it totally prevents mitochondrial ROS production despite the fact that the latter is much higher than the former. And finally, there is a high correlation between the concentration at which melatonin and closely related indoles exert a direct antioxidant effect in vitro and a neuroprotective effect against glutamate-induced oxytosis. Keywords: antioxidant, glutamate, melatonin, mitochondria, oxidative stress. J. Neurochem. (2007) 100, 736–746. Melatonin is a hormone produced mainly by the pineal gland and which has a wide spectrum of biological effects. With respect to its therapeutic potential, melatonin has been shown to modulate immune response (Guerrero and Reiter 2002; Carrillo-Vico et al. 2003); to have antiproliferative effects on tumoral cells (Hill and Blask 1988; Cos et al. 1998; Martin et al. 2005); and to display cytoprotective properties (Sainz et al. 1995; Antolı́n et al. 2002). Melatonin prevents neuronal cell death in a large number of neurodegeneration models both in vivo and in cell cultures (Mayo et al. 1998; Antolı́n et al. 2002; Andrabi et al. 2004; Beni et al. 2004). This kind of cytoprotective effect is especially interesting for three major reasons. Firstly, melatonin easily crosses the blood–brain barrier (Costa et al. 1995) so that it can be administered either orally or by intravenous injection with no need to apply the treatment directly into the brain. Secondly, the fall in nocturnal melatonin peaks in blood with increasing age, parallels quite closely the appearance of chronic neurodegenerative disorders in populations, suggesting that the lack of melatonin in the elderly may contribute to the appearance of these kinds of diseases (Reiter et al. 1996). And, finally, patients treated with high doses of melatonin do not present any harmful 736 Received March 30, 2006; revised manuscript received July 5, 2006; accepted September 17, 2006. Address correspondence and reprint requests to Carmen Rodriguez. Departamento de Morfologı́a y Biologı́a Celular. Facultad de Medicina. C/Julian Claveria, 33006 Oviedo, Spain. E-mail: carro@uniovi.es Abbreviations used: BSO, L-buthionine sulfoximine; CaCaM, Ca2+calmodulin complex; COX, cyclooxygenase; DCFH-DA, 2¢,7¢-dichlorodihydrofluorescein; D-123, dihydrorhodamine 123; FACS, flow cytometry analysis; GSH, glutathione; LOX, lipoxygenase; LPO, lipid peroxidation; MTT, 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide; NFjB, nuclear factor kappaB; PBS, phosphate-buffered saline; ROS, reactive oxygen species; SOD, superoxide dismutase.  2006 The Authors Journal Compilation  2006 International Society for Neurochemistry, J. Neurochem. (2007) 100, 736–746 Melatonin blocks oxytosis targeting mitochondria 737 side-effects (Lissoni et al. 2001), quite the opposite to what is found when other compounds such as L-DOPA or NMDA receptor antagonists are used for improving neurodegeneration symptoms and/or decreasing neuronal death (Ziv et al. 1997; Lipton 2004). On the basis of these data, melatonin has been proposed as a possible preventive treatment against neurodegenerative disorders. However, the mechanisms involved in the neuroprotection provided by melatonin remain unknown. The wide range of potential transducers of its signal accounts for the wide spectrum of melatonin biological effects. Melatonin has at least two pertussis toxin-sensitive G-coupled membrane receptors (MT1 and MT2) and an ubiquinone reductase 2-like membrane receptor (MT3) (Witt-Enderby et al. 2003); it is an antagonist of the Ca2+- calmodulin complex (CaCaM) (Leon et al. 2000); it has direct and indirect antioxidant properties (Martin et al. 2002b; Tan et al. 2002; Rodrı́guez et al. 2004); it regulates the activation of several transcription factors (i.e. nuclear factor NFjB and AP-1) (Gilad et al. 1998; Urata et al. 1999); and it might activate ROR/RZR nuclear orphan receptor (Becker-André et al. 1994; Becker-André et al. 1997). Moreover, binding sites for melatonin have been found in mitochondria (Yuan and Pang 1991; Poon and Pang 1992). There, melatonin activates mitochondrial complexes I and IV and increases oxidative phosphorylation in isolated mitochondria through an unknown mechanism (Martin et al. 2002a). It also seems to bind to – and to inactivate – the mitochondrial permeability transition pore, thus preventing cytochrome c release and excitotoxic NMDA-induced cell death in cultures of striatal neurons (Andrabi et al. 2004). Glutamate is the main excitatory neurotransmitter in the central nervous system, reaching concentrations of 1– 10 mM in the synaptic cleft and intraneuronal compartments (Dzubay and Jahr 1999). Disturbance of glutamate levels is the primary cause of neuronal death in stroke, mechanical trauma and seizure, and it is considered to play a role in some chronic neurodegenerative disorders such as Parkinson’s or Alzheimer’s diseases (Lipton 2004). Glutamate can induce cell death by two different pathways: excitotoxicity and oxytosis. While excitotoxicity is triggered by the overactivation of glutamate ionotropic receptors and the consequent massive influx of extracellular Ca2+, glutamate-induced oxytosis is triggered by the blockade of the cystine/glutamate antiporter, which causes the progressive depletion of glutathione, the major intracellular antioxidant (Tan et al. 2001). In glutamate-induced oxytosis, glutathione depletion initially induces a slight, linear increase in oxidative stress (Tan et al. 1998). When glutathione levels drop under a certain threshold (after 6–8 h), 12-lipoxygenase (12-LOX) is activated and translocated to the membrane fraction (Li et al. 1997a). Then, 12-LOX induces both an exponential burst of reactive oxygen species (ROS) from mitochondria through an unknown mechanism and an exponential increase in intracellular Ca2+ levels through activation of soluble guanylate cyclase (sGC) (Li et al. 1997a, 1997b). Finally, the exponential increase of both oxidative stress and Ca2+ influx causes a massive, sudden cell death, which is morphologically distinguishable after 8–10 h of incubation with glutamate. Blockade of any of these steps is enough to prevent cell death (Tan et al. 2001). Glutamate induces HT22 mouse hippocampal cell death exclusively through the oxytotic pathway as these cells do not express functional ionotropic receptors (Maher and Davis 1996). Lezoualc’h et al. (1996) have previously described that melatonin prevents neuronal death in this experimental model, but they did not carry out a more profound investigation into the mechanisms involved in the melatonin neuroprotective effect. The aim of the present work is to study the intracellular pathways implicated in the neuroprotection provided by melatonin against glutamate-induced oxytosis in the HT22 mouse hippocampal cell line. Our results strongly suggest that melatonin prevents cell death through a direct antioxidant effect specifically targeted at mitochondria. Materials and methods Materials HT22 cells were kindly provided by Dr David Schubert (The Salk Institute for Biological Studies, La Jolla, CA, USA). Culture plates were acquired from Falcon (Becton Dickinson BioScience, Le Pont de Claix, France). Foetal bovine serum was obtained from Gibco (Invitrogen Life Technologies, Spain). Reagents were purchased from Sigma (Sigma Chemical Co., St. Louis, MO, USA) unless otherwise indicated. Luzindole and L-NAME were acquired from Tocris (Avonmouth, UK), 5-HETE and 12-HETE from Alexis Biochemicals (QBiogene-Alexis Ltd, Nottingham, UK), and CGP52608 was kindly provided by Novartis Pharma AG (Basel, Switzerland). Cell cultures HT22 cells were cultured and seeded as described previously (Herrera et al. 2004). Mitochondrial DNA-depleted HT22 cells (HT22 q-cells) were established by the method of King and Attardi (1989). HT22 wild-type cells were grown for 2 months in Dulbecco’s modified Eagle’s medium containing 4.5 g/L glucose, 110 mg/L pyruvate, 250 ng/mL ethidium bromide, 50 lg/mL uridine, 10% foetal bovine serum and 1% of antibiotic-antimycotic mixture containing 10 000 units penicillin, 10 mg streptomycin, and 25 lg amphotericin. Ethidium bromide inhibits mitochondrial DNA replication and the expression of mitochondrial DNA-encoding genes. Mitochondrial DNA encodes for most subunits of mitochondrial respiration complexes and for a pyrimidine-synthesising enzyme (King and Attardi 1989). Therefore, q-cells have no functional electron transport chain and need uridine to proliferate. Additionally, q-cells die when pyruvate is removed from the culture medium, although there is no explanation for this dependence (King  2006 The Authors Journal Compilation  2006 International Society for Neurochemistry, J. Neurochem. (2007) 100, 736–746 738 F. Herrera et al. and Attardi 1989), and lose their ability to reduce 3-(4,5-dimethyl-2thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) (Liu et al. 1997). As expected, HT22 q-cell growth and survival is dependent on uridine and pyruvate, respectively; they do not reduce MTT; and, moreover, they do not present basal fluorescence when incubated with D-123 (dihydrorhodamine), a stain which fluorescence depends on mitochondrial activity (data not shown), indicating that mitochondrial DNA depletion was successful. Viability of cell cultures after treatments was determined by either the MTT or the trypan blue exclusion assays as described previously (Martin et al. 2002b). Trypan blue exclusion assay was carried out to determine HT22 q-cell viability after glutamate treatment as these cells are unable to reduce MTT. In vitro lipid peroxidation assay In vitro lipid peroxidation assays were performed as previously described (Herrera et al. 2001; Garcı́a-Santos et al. 2004). Briefly, rat brain homogenates (10% w/v in Tris-HCl 20 mM) were incubated at 37C for 1 h with 100 lM FeSO4 with or without different concentrations (100 nM)1 mM) of melatonin, N-acetyltryptamine, 5-methoxytryptamine or tryptamine. Afterwards, lipid peroxidation in samples was determined by the LPO586 kit (OXIS International Inc., Portland, OR, USA) following the manufacturer’s instructions. Electron microscopy After treatment, cells were collected by trypsinisation and samples were processed for electron microscopy as follows. Cells were washed once in phosphate-buffered saline (PBS) and fixed for 30 min in 3% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4). After removal of glutaraldehyde, cells were embedded in 2% w/v agar in PBS and postfixed in OsO4 for 15 min at 4C. Samples were dehydrated in acetone and embedded in SPURR resin (EMS, Fort Washington, PA, USA). Ultrathin sections (70–90 mm) were cut with an Ultracut-E ultratome (Reichert-Jung, Vienna, Austria) and stained with 2% w/v uranyl acetate and Reynolds’ lead citrate (Reynolds 1963). Photographs were taken using a transmission electron microscope EM-109 (Zeiss, Oberkochen, Germany) working at 80 kV. Superoxide dismutase (SOD) activity and glutathione (GSH) determination GSH levels were measured as previously described (Martin et al. 2002b). Superoxide dismutase activity was determined by the method of Paoletti and Mocali (1990). Briefly, cells were resuspended in 100 lL of TDB buffer (triethanolamine : diethanolamine : HCl, pH 7.4) and sonicated. Afterwards samples were centrifuged at 10 000 · g 15 min at 4C, supernatants were collected, and protein content was determined by the method of Bradford (1976). Thirty micrograms of protein in a volume of 20 lL were added to 180 lL of the reaction mixture (360 lM NADPH, 3 mM EDTA, 1.5 mM MgCl2, in TDB buffer pH 7.4). Then, 20 lL of 10 mM mercaptoethanol were added and the drop in NADPH absorbance at 340 nm was kinetically measured (during 20 min, every 30 s) in an automatic microplate reader (lQuant, Bio-Tek Instruments, Inc.). One SOD enzymatic unit (U) is the amount of enzyme needed to prevent 50% of the absorbance drop induced by mercaptoethanol. Flow cytometry analysis (FACS) Total intracellular ROS or mitochondrial ROS were evaluated by fluorescent probes 2¢,7¢-dichlorodihydrofluorescein diacetate (DCFH-DA) and dihydrorhodamine 123 (D-123), respectively. Intracellular Ca2+ levels were determined by fluorescent probe Fluo-3 AM. After treatment, cells were collected by trypsinisation, washed once, resuspended in 500 lL of DCFH-DA (10 lM), D-123 (10 lM) or Fluo-3 AM (1 lg/mL) in serum free medium and incubated for 15 min at 37C in darkness. Cells were then centrifuged and resuspended in 500 lL of PBS. Five lL of 50 lg/mL propidium iodide were added to each tube and fluorescence of 10 000 living cells per group (cells without propidium iodide uptake) was measured in a Beckman Coulter FC500 flow cytometer. The excitation wavelength for every fluorescence probe was 488 nm and emission wavelengths were 525 nm for DCF, D123 and Fluo-3 AM, and 675 nm for propidium iodide. Molecular biology studies Northern blot analyses were carried out as previously described (Mayo et al. 2002). For western blots, SDS-PAGE electrophoresis of 50 lg of total protein and transference of proteins to Hybond-P PVDF membranes (Amersham Biosciences, Little Chalfont, UK) were carried out. PVDF membranes were blocked with 5% (w/v) non-fat dry milk (Santa Cruz Biotechnologies, Santa Cruz, CA, USA) in TBS for 1 h at room temperature. Afterwards they were incubated overnight at 4C with either anti-12-LOX (1 : 1000 in blocking solution, rabbit, polyclonal, QBiogene-Alexis Ltd, Nottingham, UK) or anti-GAPDH (1 : 500 in blocking solution, rabbit, polyclonal, Santa Cruz Biotechnologies) antibodies. Filters were then washed three times with TBS-Tween (0.05% v/v) and incubated for 1 h at room temperature with an anti-rabbit antibody linked to horseradish peroxidase (1 : 10 000 in blocking solution, Calbiochem, Darmstadt, Germany). Filters were washed again and detection was carried out by ECL kit (Amersham Life Sciences) following the manufacturer’s instructions. For electromobility shift assays (EMSA), nuclear extracts were prepared following the method described by Dignam et al. (1983). Oligonucleotide probes containing the consensus sequence for ROR/ RZR orphan nuclear receptor (Proligo, Boulder, CO, USA), AP-1 and NFjB (Santa Cruz Biotechnologies) were labelled with a-[32P]ATP (3000 Ci/mmol) using T4 polynucleotide kinase 5¢-end labelling kit and purified on Microspin G25 columns (all from Amersham Life Sciences, UK). Nuclear extracts (10 lg) were incubated for 30 min on ice with 0.4 ng of labelled oligonucleotide and 100 lg/mL poly(dI-dC) in a final volume of 20 lL. For competition studies, a 100-fold excess of unlabelled oligonucleotide were added to nuclear extracts immediately before the addition of the labelled probe. DNAprotein complexes were resolved on a 6% non-denaturing polyacrylamide gel at 250 V for 1.5 h in 1X TBE. Gels were then dried and exposed to Kodak Biomax X-ray film at ) 80C for 24 h. Statistical analysis Results shown are the average of at least three independent experiments. Data are represented as the mean ± SEM. Significance was tested by one-way ANOVA test followed by a Student-NewmanKeuls multiple range test. The correlation coefficient and its p-value were tested by the Pearson Product Moment Correlation.  2006 The Authors Journal Compilation  2006 International Society for Neurochemistry, J. Neurochem. (2007) 100, 736–746 Melatonin blocks oxytosis targeting mitochondria 739 Results Millimolar but not lower concentrations of melatonin totally prevent glutamate-induced cell death To determine the concentration of melatonin required to completely prevent glutamate-induced cell death, we incubated HT22 cells with or without glutamate 10 mM and with or without melatonin at concentrations ranging from 1 nM to 1 mM. Only melatonin concentrations higher than 0.25 mM prevent significantly glutamate-induced cell death and, only 1 mM melatonin was able to completely perform this prevention it (Fig. 1a), as previously described (Lezoualc’h et al. 1996). Incubation with 1 mM melatonin – either alone or in the presence of 10 mM glutamate- decreased cell (a) (b) Fig. 1 Melatonin prevents glutamate-induced cell death only at millimolar concentrations. (a) MTT reduction assay. Cells were incubated with or without several concentrations of melatonin (0.1–1 mM) and/or 10 mM glutamate, and viability was determined after 24 h of incubation. *Significant vs. group without additives; #Significant vs. 10 mM glutamate, p < 0.01. (b) Melatonin (1 mM) prevents morphological changes induced by 10 mM glutamate. Light microscopy pictures (left panels, · 400) and electron microscopy pictures (right panels, · 8200) clearly show that melatonin prevents cell death, chromatin (ch) condensation (arrowheads) and membrane blebbing (arrows) induced by glutamate. The morphology of mitochondria (mit) does not change after any treatment (inserts, · 16 500).  2006 The Authors Journal Compilation  2006 International Society for Neurochemistry, J. Neurochem. (2007) 100, 736–746 740 F. Herrera et al. number by inhibition of cell proliferation but not by induction of cell death, as confirmed by trypan blue exclusion assay and by optical microscopy (data not shown). Figure 1b (right panels) also shows the ultrastructure of HT22 cells treated with or without glutamate 10 mM and with glutamate 10 mM plus melatonin 1 mM. Ultrastructural features of glutamate-treated cells resemble apoptotic cell death as we observe chromatin condensation and blebbing of cell membrane, but cell and nuclear membranes remain unbroken and intracellular organelles are not swollen. Incubation with 1 mM melatonin completely prevents the ultrastructural changes induced by glutamate. Neuroprotective effect of melatonin is not mediated by classical transduction pathways To determine whether the previously described melatonin signal transduction pathways were implicated in the observed neuroprotective effect, a series of experiments with agonists or blockers of these pathways were carried out, the results of which are summarised in Table 1. Some of the effects of melatonin are mediated by binding and activation of its membrane receptors and a consequent decrease of cyclic AMP intracellular levels, especially those that can be observed at nanomolar concentrations of the indole (Vanecek and Klein 1992; Witt-Enderby et al. 2003). However, pre-incubation for 6 h either with luzindole (10– 50 lM) or with pertussis toxin (10 ng/mL) – two blockers of melatonin membrane receptor activation – was unable to prevent the neuroprotective effect of melatonin (Table 1). Moreover, the neuroprotective effect was not prevented by CPT-cAMP (10 nM)1 mM), a permeable analogue of cyclic AMP (Table 1), further indicating that membrane receptor activation is not implicated in the neuroprotection provided by melatonin. Melatonin also inhibits neuronal nitric oxide synthase (nNOS) activity through antagonism with CaCaM complex (Leon et al. 2000) as well as the expression of the inducible isoform of NOS through inhibition of NFjB activation (Gilad et al. 1998), both enzymes presenting pro-oxidant activity. However, NOS activity does not seem to be involved in glutamate-induced oxytosis since the NOS inhibitor L-NAME does not prevent cell death (Table 1), as previously described (Li et al. 1997b). Furthermore, the CaCaM antagonist W7 (10–100 lM) and the NFjB inhibitor parthenolide (1–50 lM) did not prevent glutamate-induced cell death (Table 1). Additionally, neither glutamate nor melatonin alters the activation state of NFjB in HT22 cells after 4 or 8 h of incubation (data not shown). Thus, our results strongly indicate that NOS, CaCaM complex and NFjB transcription factor are not implicated either in glutamate-induced oxytosis or in the neuroprotection provided by melatonin in this experimental model. The induction of a rise in GSH levels by melatonin has been described in a number of cell types (Urata et al. 1999; Martin et al. 2005). In endothelial cells this is achieved by activation of AP-1 transcription factor and the consequent increase in c-glutamyl cysteine synthetase (c-GCS) expression, the rate-limiting enzyme in GSH synthesis (Urata et al. 1999). Given that GSH depletion is involved in glutamateinduced oxytosis, melatonin could be regulating GSH synthesis and consequently preventing cell death. Nevertheless, L-buthionine sulfoximine (BSO) – a specific inhibitor of Table 1 Pharmacological approach to the pathways implicated in prevention of glutamate-induced oxytosis by melatonin. Cells were incubated with or without melatonin 1 mM, with or without glutamate 10 mM, and with or without the compounds shown in the Table for 24 h. All drugs were co-incubated, except the melatonin receptor inhibitors luzindole and pertussis toxin, which were pre-incubated for 6 h prior to the addition of melatonin and/or glutamate. After treatment, cell viability was determined by MTT reduction assay. When a drug significantly (p < 0.05) prevented glutamate-induced cell death or the neuroprotection provided by melatonin, the percentage of prevention is presented in parenthesis Compound Action Luzindole (10–50 lM) Pertussis toxin (10 ng/mL) CPT-cAMP (10 nM)1 mM) L-NAME (1 nM)1 mM) W7 (10–100 lM) Parthenolide (1–50 lM) BSO (1–10 lM) Acetylsalicylic acid (1 lM)1 mM) Piroxicam (5–100 lM) Diclofenac (50–100 lM) CGP52608 (1 lM) 5-HETE (100 nM-2 lM) 12-HETE (100 nM-2 lM) Mel membrane receptor antagonists cAMP permeable analogue NOS inhibitor CaCaM antagonist NFjB inhibitor ã-GCS inhibitor COX inhibitors ROR/RZR agonist Product of 5-LOX Product of 12-LOX Blocks mel protection Prevents cell death NO NO NO n.d. n.d. NO NO n.d. n.d. n.d. n.d. NO NO NO NO NO NO NO NO NO NO NO NO YES (40%) NO NO n.d., not determined.  2006 The Authors Journal Compilation  2006 International Society for Neurochemistry, J. Neurochem. (2007) 100, 736–746 Melatonin blocks oxytosis targeting mitochondria 741 c-GCS – did not prevent the melatonin neuroprotective effect (Table 1), and neither melatonin nor glutamate altered AP-1 transcription factor activation state in HT22 cell line after 4 or 8 h of incubation (data not shown). Inhibition by melatonin of the expression of the prooxidant enzyme cyclooxygenase (COX) has been also reported (Dong et al. 2003; Mayo et al. 2005). However, COX activity does not appear to be implicated in glutamateinduced oxytosis, as indomethacin does not prevent cell death (Li et al. 1997a). We observed that typical COX inhibitors such as acetylsalicylic acid, piroxicam and diclofenac did not prevent cell death (Table 1). Our results therefore indicate that COX expression inhibition by melatonin plays no part in the neuroprotective effect, and further confirm that COX activity does not play a role in glutamateinduced oxytosis. Melatonin has been shown to activate orphan nuclear receptor ROR/RZR, repressing 5-lipoxygenase (5-LOX) expression in lymphocytes (Steinhilber et al. 1995). The ROR/RZR agonist CGP52608 (1 lM) prevented 40% of the glutamate-induced cell death (Table 1). In spite of this, neither melatonin nor glutamate changed the activation state of ROR/RZR in HT22 cells after 4 or 8 h of incubation (data not shown). Additionally, cells were incubated with 5-HETE, the main metabolite of 5-LOX, at concentrations that prevent the biological effects caused by 5-LOX inhibition in other experimental models (Gosh and Myers 1998). The 5-LOX metabolite was unable to prevent the neuroprotective effect (Table 1), further confirming that ROR/RZR activation and 5-LOX down-regulation by melatonin are not involved in this effect of the indole. It is also known that melatonin inhibits 12-LOX activity and expression in rat pinealocytes (Zhang et al. 1999), at high and low concentrations, respectively, and this enzyme is involved in glutamate-induced cell death (Li et al. 1997a). However, neither melatonin nor glutamate modified the expression of 12-LOX after 4 or 8 h of incubation (data not shown), and 12-LOX metabolite 12-HETE was also unable to prevent the melatonin effect (Table 1) at concentrations previously described to prevent the biological effects caused by 12-LOX inhibition (Tang et al. 1996; Pidgeon et al. 2002). Finally, our group had previously shown that physiological concentrations of melatonin can regulate copper-zinc and manganese superoxide dismutase (CuZn & MnSODs) expression and reports on the regulation of their activity by the indole also exist (Rodrı́guez et al. 2004). We analysed both parameters by northern blot and the NADPH oxidation assay, respectively (see Materials and methods). We were unable to detect MnSOD by Northern blot, indicating that HT22 cells have low levels of this enzyme, and CuZnSOD expression remains unaltered after glutamate and/or melatonin treatment (data not shown). Total SOD activity assays showed that none of the treatments alter the activity of these enzymes, further confirming the results obtained by Northern blot (data not shown). In summary, these results clearly indicate that the previously described transducers of the melatonin biological effects are not involved in the observed neuroprotective effect in this experimental model. Melatonin specifically prevents the events that characterise the second phase of the cell death process The effect of the indole in the sequence of events of the glutamate-induced oxytosis process (i.e. GSH depletion, cytosolic and mitochondrial oxidative stress, and intracellular Ca2+ increase) was then studied. Although BSO did not prevent the neuroprotective action of melatonin, we first analysed the effect of melatonin on GSH depletion after glutamate incubation. As shown in Fig. 2a, melatonin cannot prevent glutamate-induced GSH depletion after 8 and 12 h of incubation. This result further confirms that the rise in c-GCS expression and GSH levels shown in human endothelial cells after melatonin treatment is not involved in melatonin prevention of glutamate-induced cell death in HT22 cells. The effect of melatonin on the increase in ROS production after 8 and 12 h of glutamate incubation was then investigated. Analysis of DCF fluorescence by flow cytometry shows that melatonin can not prevent the initial glutamateinduced increase of ROS (8 h) but it almost completely prevents the exponential increase of these reactive species (12 h) (Fig. 2b). A certain level of ROS production, not prevented by melatonin after 12 h, is equal to the ROS levels found after 8 h of glutamate incubation. As the slight increase of ROS at 8 h is GSH depletion-dependent and the exponential increase of ROS at 12 h is dependent on mitochondrial function (Tan et al. 1998), these results suggest that melatonin specifically prevents mitochondrial ROS production. Another significant observation was that cells incubated only with melatonin for 12 h display a basal ROS production lower than control cells. Besides the mitochondrial oxidative stress, an increase in intracellular Ca2+ levels, typically appearing between 30 min and 1 h after mitochondrial oxidative stress, also characterises the second phase of this cell death process (Tan et al. 1998). Thus, the effect of melatonin in glutamate-induced increase in intracellular Ca2+ levels after 13 h of incubation was also determined. As shown in Fig. 2c, melatonin completely prevents the increase in intracellular Ca2+ levels induced by glutamate. Melatonin thus prevents the particular events of the second phase of the cell death process, but not the events characterising the first phase. The prevention by melatonin of glutamate-induced Ca2+ increase may be mediated by the blockade of mitochondrial oxidative stress, as the two events are interdependent and mitochondrial oxidative stress appears earlier than the Ca2+ rise in the sequence of events leading to cell death (Tan et al. 1998). Moreover, it has been shown that melatonin is able to  2006 The Authors Journal Compilation  2006 International Society for Neurochemistry, J. Neurochem. (2007) 100, 736–746 742 F. Herrera et al. a wide variety of antioxidant effects (Tan et al. 2002; Rodrı́guez et al. 2004), which could in turn lead to the blockade of glutamate-induced Ca2+ increase. We consequently focused our analysis on the prevention of mitochondrial-derived oxidative stress by melatonin. For this purpose we carried out two complementary experiments. First, we analysed the effect of glutamate with or without melatonin in HT22 q-cells on GSH and ROS levels after 12 h of incubation. As these cells have no functional mitochondria, glutamate is unable to induce mitochondrial stress and cell death (data not shown). However, glutamate is able to induce GSH depletion and the consequent GSH depletion-dependent ROS increase in these cells (Fig. 3a and b, respectively). Melatonin is unable to prevent either the GSH depletion or the ROS increase induced by glutamate in HT22 q-cells, further suggesting that its antioxidant effect is not targeting the cytosol, at least in this experimental model. We next analysed mitochondrial ROS production in wildtype HT22 cells after 12 h of incubation with glutamate and/ or melatonin using the fluorescent ROS-sensitive probe D123. This probe is commonly used either as a marker of mitochondrial function or as a specific indicator of mitochondrial ROS production (Jou et al. 2004). Thus, any change in D-123 fluorescence should be caused specifically by changes in mitochondrial ROS production. Glutamate produces a striking increase in D-123 fluorescence after 12 h of incubation and melatonin completely prevents this increase (Fig. 3c). Moreover, melatonin also decreases the basal production of ROS measured by D-123 as previously observed by means of DCF fluorescence analysis. These results – together with those obtained with HT22 q-cells – strongly indicate that the antioxidant effect of melatonin in HT22 cells specifically targets the mitochondria. Fig. 2 Melatonin effects on the sequence of events occurring in glutamate-induced oxytosis. GSH (a), ROS (b) and Ca2+ (c) intracellular levels were determined after incubation with 10 mM glutamate and/or 1 mM melatonin. ROS levels were determined by DCFH-DA fluorescence probe, which detects both cytosolic and mitochondrial ROS. GSH and ROS levels were measured after 8 and 12 h of incubation, whereas Ca2+ levels were only measured after 13 h of incubation, as no changes can be found before that time. *Significant vs. group without additives; #Significant vs. 10 mM glutamate 12 h, p < 0.01. block L-type Ca2+ channels, although this blockade depends on membrane receptor activation (Vanecek and Klein 1992) and these channels are not involved in the glutamate-induced Ca2+ increase (Li et al. 1997b). However, melatonin displays In vitro antioxidant properties of melatonin and closely related indoles and their effectiveness in preventing glutamate-induced cell death are highly correlated To confirm that the observed neuroprotective effects of melatonin are related to its free radical scavenging properties, we carried out a series of experiments with melatonin and three closely related indoles. First of all we analysed the effect of concentrations ranging from 1 nM to 1 mM of melatonin (Nacetyl-5-methoxytryptamine), N-acetyltryptamine, 5-methoxytryptamine and tryptamine on glutamate-induced oxytosis. Secondly, we analysed the effect of these indoles at the same concentrations on in vitro lipid peroxidation (LPO) induced by 100 lM Fe2SO4 in rat brain homogenates. From these data, we calculated the EC50 values of the indoles in both effects and the correlation coefficient between the resulting effective concentrations as is shown in Table 2. N-acetyltryptamine does not prevent cell death nor LPO at the tested concentrations, whereas the calculated EC50 values for the other indoles in preventing both events are highly correlated (r2 ¼ 0.99973, p-value ¼ 0.0147).  2006 The Authors Journal Compilation  2006 International Society for Neurochemistry, J. Neurochem. (2007) 100, 736–746 Melatonin blocks oxytosis targeting mitochondria 743 Table 2 EC50 of antioxidant and neuroprotective effects of melatonin and closely related indoles. Direct antioxidant properties were evaluated by the in vitro lipid peroxidation assay, inducing lipid peroxidation in rat brain homogenates with FeSO4 100 lM and determining prevention of lipid peroxidation by concentrations of indoles ranging from 1 nM to 1 mM. The neuroprotective effect of the indoles against glutamate-induced cell death was evaluated by the MTT reduction assay. The EC50 values were calculated from regression curves of the obtained experimental data Fig. 3 Melatonin effect on ROS production is targeted on mitochondria. GSH (a) and ROS (b) levels in HT22 q-cells were determined after incubation with 10 mM glutamate and/or 1 mM melatonin. ROS levels were determined by fluorescent probe DCFH-DA, which detects both mitochondrial and cytosolic ROS production. Mitochondrial ROS production (c) was determined in HT22 wild-type cells after incubation with 1 mM melatonin and/or 10 mM glutamate by fluorescent probe D-123, which specifically detects mitochondrial ROS production. *Significant vs. group without additives; #Significant vs. 10 mM glutamate, p < 0.01. Discussion Melatonin completely prevents mitochondrial ROS production, Ca2+ increase and cell death induced by glutamate in HT22 cells. The results shown above strongly suggest that Indoleamine EC50 MTT EC50 LPO Melatonin Tryptamine 5-Methoxytryptamine N-Acetyltryptamine 790.11 ± 21.39 17.80 ± 1.29 41.18 ± 5.47 – 434.36 ± 14.30 29.52 ± 2.89 31.17 ± 2.21 – melatonin prevents glutamate-induced oxytosis through a direct antioxidant effect specifically targeting mitochondria. None of the described transducers of melatonin signalling seems to be implicated in the neuroprotection provided by the indole. Melatonin membrane receptors are not involved as well-known antagonists of these receptors were unable to prevent the neuroprotective effect. In fact, most biological effects of pharmacological concentrations of melatonin are found to be independent of membrane receptor activation (Pappolla et al. 2002; Martin et al. 2005), with some exceptions (Jiao et al. 2004). Although CGP52608 also prevented cell death, there is evidence that the orphan nuclear receptor ROR/RZR is not implicated in the melatonin effect. The activation state of the orphan receptor is not altered by melatonin and/or glutamate incubation at the time points studied. Furthermore, although melatonin can repress the expression of the 5-LOX gene by activating the ROR/RZR in human B lymphocytes (Steinhilber et al. 1995), the 5-LOX main metabolite 5-HETE could not block the neuroprotection provided by the indole. Additionally, the existing literature supports our results. Melatonin binding to and activation of ROR/RZR remains controversial (Becker-André et al. 1994, 1997), and most of the reports implicating ROR/RZR in a given melatonin effect are based only on the similarity between melatonin and CGP52608 effects. In fact, CGP52608 is a little studied compound that could in reality prevent cell death by any pathway aside from that of ROR/ RZR activation. For example, other thiazolidine derivatives have been shown to inhibit mitochondrial respiration (Hildyard et al. 2005) or be a source of cysteine (Wlodek et al. 1993), which might explain the neuroprotective effect of CGP52608. In other experimental models, melatonin up-regulates the activity or the expression of antioxidant enzymes, such as CuZn- and MnSODs and c-GCS (Antolı́n et al. 1996; Urata et al. 1999; Mayo et al. 2002), and down-regulates the activity or the expression of pro-oxidant enzymes such as LOXs, COXs or the neuronal and inducible isoforms of NOS  2006 The Authors Journal Compilation  2006 International Society for Neurochemistry, J. Neurochem. (2007) 100, 736–746 744 F. Herrera et al. (Steinhilber et al. 1995; Gilad et al. 1998; Zhang et al. 1999; Leon et al. 2000; Dong et al. 2003; Mayo et al. 2005). However, these indirect antioxidant effects do not seem to be involved in the melatonin neuroprotective effect on the HT22 system, according to the present results. Melatonin also completely prevents the glutamateinduced increase in intracellular Ca2+ levels. 12-LOX activation is responsible for both mitochondrial oxidative stress and Ca2+ increase (Li et al. 1997a, 1997b) and melatonin is able to inhibit the expression or the activity of 12-LOX in rat pinealocytes (Zhang et al. 1999). However, in our model, melatonin did not inhibit 12-LOX expression and the 12-LOX main metabolite 12-HETE did not prevent the melatonin effect, suggesting that inhibition of 12-LOX expression or activity by melatonin is not involved in melatonin neuroprotection against glutamate-induced oxytosis. The prevention of glutamate-induced Ca2+ increase may be mediated by the prevention of mitochondrial oxidative stress, as both events have been described to be interdependent and mitochondrial oxidative stress appeared before Ca2+ increase in the sequence of events leading to cell death (Tan et al. 1998). In support of this possibility, the Ca2+ channels described as being regulated by melatonin (i.e. L-type, voltage-dependent) are not implicated in glutamate-induced oxytosis. Moreover, these channels are regulated by melatonin membrane receptor activation (Vanecek and Klein 1992), a pathway that does not mediate the neuroprotective effect of melatonin against glutamateinduced oxytosis. Prevention of mitochondrial oxidative stress is sufficient to stop glutamate-induced oxytosis in HT22 cells (Tan et al. 1998). Melatonin prevents the mitochondrial ROS increase but not the cytosolic GSH depletion-dependent ROS increase, even though the former is much higher than the latter. The results obtained in HT22 q-cells, which lack mitochondrial function, demonstrate that melatonin does not prevent GSH depletion-dependent ROS increase in this experimental model. A complete and specific prevention of mitochondrial ROS increase by melatonin was further confirmed by using D-123, a ROS-sensitive fluorescent dye commonly used as a specific indicator of mitochondrial ROS production. Thus, mitochondria seem to be the specific target of the melatonin antioxidant and neuroprotective effects. Finally, there is a close correlation between the concentration of melatonin and that of closely related indoles in which their direct antioxidant properties in vitro and their neuroprotective effect against glutamate-induced oxytosis can be observed. Once we have observed a complete and specific antioxidant effect of melatonin against glutamateinduced mitochondrial oxidative stress, and discarded every previously described indirect antioxidant pathway, this correlation supports the implication of the direct antioxidant properties of this molecule in its neuroprotective effects against glutamate-induced oxytosis. On the basis of the present results and the previously published data, there are two possible pathways by which melatonin might prevent glutamate-induced mitochondrial oxidative stress and cell death in HT22 cells. Melatonin might increase the efficiency of the mitochondrial respiration chain through the activation of complexes I and IV (Martin et al. 2002a; Leon et al. 2005), thus preventing the increase in mitochondrial ROS. Should this be the case, this effect would very likely depend on direct antioxidant properties of melatonin, bearing in mind the high correlation between the in vitro antioxidant effect of melatonin and closely related indoles and their neuroprotective effect. Although melatonin has been shown to bind to and inhibit the mitochondrial permeability transition pore, thus preventing cytochrome c release from mitochondria and cell death in NMDA-induced cell death in striatal neurons (Andrabi et al. 2004), cytochrome c is not released from the mitochondria in glutamateinduced oxytosis (Tan et al. 2001). Therefore, it is very unlikely that this pathway is involved in melatonin effects on the HT22 system. Alternatively, melatonin might directly and specifically scavenge mitochondrial free radicals. Melatonin has been shown to have mitochondrial binding sites in pigeon and guinea pig brains (Yuan and Pang 1991; Poon and Pang 1992). Thus it is plausible that melatonin accumulates specifically in HT22 cell mitochondria, attaining a concentration high enough to cause a direct antioxidant effect only inside this organelle, but not in the cytosol. Nevertheless, other unknown effects of direct antioxidant properties of melatonin at the mitochondria should not be ruled out. In conclusion, we present strong evidence that antioxidant properties of melatonin are involved in its neuroprotective effect against glutamate toxicity in HT22 cells, and that the antioxidant effect specifically targets mitochondrial ROS production. Moreover, we have also discarded the participation of other previously described melatonin transducers as well as some of the main factors involved in glutamateinduced oxytosis, such as GSH or 12-LOX. The possible pathways to be studied in the future are strictly limited to those that could be regulated by the antioxidant properties of melatonin. Mitochondrial dysfunction and oxidative stress are common features in a huge number of chronic and acute neurodegenerative disorders (Beal 2005). Melatonin has been shown to protect against a wide variety of neuronal insults and to have effects at the mitochondrial level, but little is known of the mechanisms implied in these effects (Leon et al. 2005). The therapeutic use of melatonin in neurodegeneration depends very much on the complete understanding of the pathways implicated in the neuroprotection provided by this indole. Bearing in mind our results and those found in the literature, further studies on the mitochondrial effects of melatonin in experimental models of neurodegeneration should be carried out.  2006 The Authors Journal Compilation  2006 International Society for Neurochemistry, J. Neurochem. (2007) 100, 736–746 Melatonin blocks oxytosis targeting mitochondria 745 Acknowledgements Work supported by FICYT (PB02-085) and DGES (SAF200303303) grants to CR; ISCIII, Red de Centros de Cancer RTICCC (CO3/10) and FISS grant (PI031596) to IA. FH was supported by a fellowship from the Fondo de Investigaciones Sanitarias (BEFI 00/ 9412). VM was supported by FICYT and JR-B by MAPFRE and DANONE. GG-S acknowledges a fellowship from FICYT (BP03112). Authors also appreciate the technical assistance of Fernando Jañez and Carlos Villa from the Electron Microscopy Service of the University of Oviedo. References Andrabi S. A., Sayeed I., Siemen D., Wolf G. and Horn T. F. (2004) Direct inhibition of the mitochondrial permeability transition pore: a possible mechanism responsible for anti-apoptotic effects of melatonin. FASEB J. 18, 869–871. Antolı́n I., Rodrı́guez C., Sainz R. M. et al. (1996) Neurohormone melatonin prevents cell damage: effect on gene expression for antioxidant enzymes. FASEB J. 10, 882–890. Antolı́n I., Mayo J. C., Sainz R. M., del Brı́o M. A., Herrera F., Martı́n V. and Rodriguez C. (2002) Protective effect of melatonin in a chronic experimental model of Parkinson’s disease. Brain Res. 943, 163– 173. Beal M. F. (2005) Mitochondria take center stage in aging and neurodegeneration. Ann. Neurol. 58, 495–505. Becker-André M., Wiesemberg I., Schaeren-Wiemers N., André E., Missbach M., Saurat J. H. and Carlberg C. (1994) Pineal gland hormone melatonin binds and activates an orphan of the nuclear receptor subfamily. J. Biol. Chem. 269, 28 531–28 534. Becker-André M., Wiesemberg I., Schaeren-Wiemers N., André E., Missbach M., Saurat J. H. and Carlberg C. (1997) Erratum to J. Biol. Chem. (1994) 269: 28 531–28 534. J. Biol. Chem. 272, 16 707. Beni S. M., Kohen R., Reiter R. J., Tan D. X. and Shohami E. (2004) Melatonin-induced neuroprotection after closed head injury is associated with increased brain antioxidants and attenuated late-phase activation to NF-kappaB and AP-1. FASEB J. 18, 149– 151. Bradford M. M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of proteindye binding. Anal Biochem. 72, 248–254. Carrillo-Vico A., Garcı́a-Maurino S., Calvo J. R. and Guerrero J. M. (2003) Melatonin counteracts the inhibitory effect of PGE2 on IL-2 production in human lymphocytes via its mt1 membrane receptor. FASEB J. 17, 755–757. Cos S., Fernandez R., Guezmes A. and Sánchez-Barceló E. J. (1998) Influence of melatonin on invasive and metastatic properties of MCF-7 human breast cancer cells. Cancer Res. 58, 4383–4390. Costa E. J. X., López R. H. and Lamy-Freund M. T. (1995) Permeability of pure lipid bilayers to melatonin. J. Pineal Res. 19, 123–126. Dignam J. D., Lebovitz R. M. and Roeder R. G. (1983) Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucl Acid Res. 11, 1475– 1489. Dong W. G., Mei Q. YuJ. P., Xu J. M., Xiang L. and Xu Y. (2003) Effects of melatonin on the expression of iNOS and COX-2 in rat models of colitis. World. J. Gastroenterol. 9, 1307–1311. Dzubay J. A. and Jahr C. E. (1999) The concentration of synaptically released glutamate outside of the climbing fiber-Purkinje cell synaptic cleft. J. Neurosci. 19, 5265–5274. Garcı́a-Santos G., Herrera F., Martin V., Rodriguez-Blanco J., Antolı́n I., Fernández-Marı́ F. and Rodrı́guez C. (2004) Antioxidant activity and neuroprotective effects of zolpidem and several synthesis intermediates. Free Radic. Res. 38, 1289–1299. Gilad E., Wong H. R., Zingarelli B., Virag L., O’Connor M., Salzman A. L. and Szabo C. (1998) Melatonin inhibits expression of the inducible isoform of nitric oxide synthase in murine macrophages: role of inhibition of NFjB activation. FASEB J. 12, 685–693. Gosh J. and Myers C. E. (1998) Inhibition of arachidonate 5-lipoxygenase triggers massive apoptosis in human prostate cancer cells. Proc. Natl Acad. Sci. U S A 95, 13 182–13 187. Guerrero J. M. and Reiter R. J. (2002) Melatonin-immune system relationships. Curr. Top. Med. Chem. 2, 167–179. Herrera F., Sainz R. M., Mayo J. C., Martin V., Antolı́n I. and Rodrı́guez C. (2001) Glutamate induces oxidative stress not mediated by glutamate receptors or cystine transporters: protective effect of melatonin and other antioxidants. J. Pineal Res. 31, 356–362. Herrera F., Mayo J. C., Martin V., Sainz R. M., Antolı́n I. and Rodrı́guez C. (2004) Cytotoxicity and oncostatic activity of the thiazolidinedione derivative CGP52608 on central nervous system cancer cells. Cancer Lett. 211, 47–55. Hildyard J. C., Ammala C., Dukes I. D., Thomsom S. A. and Halestrap A. P. (2005) Identification and characterization of a new class of highly specific and potent inhibitors of the mitochondrial pyruvate carrier. Biochim. Biophys. Acta 1707, 221–230. Hill S. M. and Blask D. E. (1988) Effects of the pineal hormone melatonin on the proliferation and the morphological characteristics of human breast cancer cells (MCF-7) in culture. Cancer Res. 48, 6121–6126. Jiao S., Wu M. M., Hu C. L., Zhang Z. H. and Mei Y. A. (2004) Melatonin receptor agonist 2-iodomelatonin prevents apoptosis of cerebellar granule neurons via K+ current inhibition. J. Pineal Res. 36, 109–116. Jou M. J., Peng T. I., Reiter R. J., Jou S. B., Wu H. Y. and Wen S. T. (2004) Visualization of the antioxidative effects of melatonin at the mitochondrial level during oxidative stress-induced apoptosis of rat brain astrocytes. J. Pineal Res. 37, 55–70. King M. P. and Attardi G. (1989) Human cells lacking mtDNA: Repopulation with exogenous mitochondria by complementation. Science 246, 500–503. Leon J., Macı́as M., Escames G., Camacho E., Khaldy H., Martin M., Espinosa A., Gallo M. A. and Acuña-Castroviejo D. (2000) Structure-related inhibition of Calmodulin-dependent neuronal nitric-oxide synthase activity by melatonin and synthetic kynurenines. Mol. Pharmacol. 58, 967–975. Leon J., Acuña-Castroviejo D., Escames G., Tan D. X. and Reiter R. J. (2005) Melatonin mitigates mitochondrial malfunction. J. Pineal Res. 38, 1–9. Lezoualc’h F., Skutella T., Widmann M. and Behl C. (1996) Melatonin prevents oxidative stress-induced cell death in hippocampal cells. Neuroreport 7, 2071–2077. Li Y., Maher P. and Schubert D. (1997a) A role for 12-lipoxygenase in nerve cell death caused by glutathione depletion. Neuron 19, 453– 463. Li Y., Maher P. and Schubert D. (1997b) Requirement for cGMP in nerve cell death caused by glutathione depletion. J. Cell Biol. 139, 1317– 1324. Lipton S. A. (2004) Failures and successes of NMDA receptor antagonists: Molecular basis for the use of open-channel blockers like memantine in the treatment of acute and chronic neurologic insults. NeuroRx 1, 101–110. Lissoni P., Rovelli F., Malugani F., Bucovec R., Conti A. and Maestroni G. J. (2001) Anti-angiogenic activity of melatonin in advanced cancer patients. Neurol. Endocrinol. Lett. 22, 45–47.  2006 The Authors Journal Compilation  2006 International Society for Neurochemistry, J. Neurochem. (2007) 100, 736–746 746 F. Herrera et al. Liu Y., Peterson D. A., Kimura H. and Schubert D. (1997) Mechanism of cellular 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) reduction. J. Neurochem. 69, 581–593. Maher P. and Davis J. P. (1996) The role of monoamine metabolism in glutamate toxicity. J. Neurosci. 16, 6394–6401. Martin M., Macı́as M., Leon J., Escames G., Khaldy H. and AcuñaCastroviejo D. (2002a) Melatonin increases the activity of the oxidative phosphorylation enzymes and the production of ATP in rat brain and liver mitochondria. Int. J. Biochem. Cell Biol. 34, 348–357. Martin V., Sainz R. M., Antolı́n I., Mayo J. C., Herrera F. and Rodrı́guez C. (2002b) Several antioxidant pathways are involved in astrocyte protection by melatonin. J. Pineal Res. 33, 204–212. Martin V., Herrera F., Carrera M. P., Garcı́a-Santos G., Antolı́n I., Rodrı́guez-Blanco J. and Rodrı́guez C. (2005) Intracellular signalling pathways involved in the cell growth inhibition of glioma cells by melatonin. Cancer Res. 66, 1081–1088. Mayo J. C., Sainz R. M., Urı́a H., Antolı́n I., Esteban M. M. and Rodriguez C. (1998) Melatonin prevents apoptosis induced by 6-hydroxydopamine in neuronal cells: Implications for parkinson’s disease. J. Pineal Res. 24, 179–192. Mayo J. C., Sainz R. M., Antolı́n I., Herrera F., Martin V. and Rodrı́guez C. (2002) Melatonin regulation of antioxidant enzyme gene expression. Cell Mol. Life Sci. 59, 1706–1713. Mayo J. C., Sainz R. M., Tan D. X., Hardeland R., León J., Rodrı́guez C. and Reiter R. J. (2005) Anti-inflammatory actions of melatonin and its metabolites, N1-acetyl-N2-formyl-5-methoxykynuremine (AFMK) and N1-acetyl-5-methoxykynuremine (AMK), in macrophages (2005). J. Neuroimmunol. 165, 139–149. Paoletti F. and Mocali A. (1990) Determination of superoxide dismutase activity by purely chemical system based on NAD(P)H oxidation. Methods Enzymol. 186, 209–220. Pappolla M. A., Simovich M. J., Bryant-Thomas T. et al. (2002) The neuroprotective activities of melatonin against the Alzheimer protein are not mediated by melatonin membrane receptors. J. Pineal Res. 32, 135–142. Pidgeon G. P., Kandouz M., Meram A. and Honn K. V. (2002) Mechanisms controlling cell cycle arrest and induction of apoptosis after 12-lipoxygenase inhibition in prostate cancer cells. Cancer Res. 62, 2721–2727. Poon A. M. and Pang S. F. (1992) 2-[125I]iodomelatonin binding sites in spleens of guinea pigs. Life Sci. 50, 1719–1726. Reiter R. J., Pablos M. I., Agapito T. T. and Guerrero J. M. (1996) Melatonin in the context of the free radical theory of aging. Ann. N.Y. Acad. Sci. 786, 362–378. Reynolds E. S. (1963) The use of lead citrate at high pH as a electronopaque stain in electron microscopy. J. Cell Biol. 17, 208–212. Rodrı́guez C., Mayo J. C., Sainz R. M., Antolı́n I., Herrera F., Martin V. and Reiter R. J. (2004) Regulation of antioxidant enzymes: a significant role for melatonin. J. Pineal Res. 36, 1–9. Sainz R. M., Mayo J. C., Urı́a H., Kotler M., Antolı́n I., Rodrı́guez C. and Menendez-Pelaez A. (1995) The pineal hormone melatonin prevents in vivo and in vitro apoptosis of thymocytes. J. Pineal Res. 19, 178–188. Steinhilber D., Brungs M., Werz O., Wiesemberg I., Danielsson C., Kahlen J. P., Nayeri S., Schrader M. and Carlberg C. (1995) The nuclear receptor for melatonin represses 5-lipooxygenase gene expression in human B lymphocytes. J. Biol. Chem. 270, 7037– 7040. Tan S., Sagara Y., Li Y., Maher P. and Schubert D. (1998) The regulation of reactive oxygen species production during programmed cell death. J. Cell Biol. 141, 1423–1432. Tan S., Schubert D. and Maher P. (2001) Oxytosis: a Novel Form of Programmed Cell Death. Curr. Topics Med. Chem. 1, 497–506. Tan D. X., Reiter R. J., Manchester L. C. et al. (2002) Chemical and physical properties and potential mechanisms: melatonin as a broad spectrum antioxidant and free radical scavenger. Curr. Top. Med. Chem. 2, 181–197. Tang D. G., Chen Y. Q. and Honn K. V. (1996) Arachidonate lipoxygenases as essential regulators of cell survival and apoptosis. Proc. Natl Acad. Sci. USA 93, 5241–5246. Urata Y., Honma S., Goto S., Todoroki S., Iida T., Cko S., Honma K. and Kondo T. (1999) Melatonin induces c-glutamylcysteine synthetase mediated by activator protein-1 in human cells. Free Radic. Biol. Med. 27, 838–847. Vanecek J. and Klein D. C. (1992) Melatonin inhibits of gonadotropinreleasing hormone-induced elevation of intracellular Ca2+ in neonatal rat pituitary cells. Endocrinology 130, 701–707. Witt-Enderby P. A., Bennet J., Jarzynka M. J., Firestine S. and Melan M. A. (2003) Melatonin receptors and their regulation: biochemical and structural mechanisms. Life Sci. 72, 2183–2198. Wlodek L., Rommelspacher H., Susilo R., Radomski J. and Hofle G. (1993) Thiazolidine derivatives as source of free 1-cysteine in rat tissue. Biochem. Pharmacol. 46, 1917–1928. Yuan H. and Pang S. F. (1991) [125-I]Iodomelatonin-binding sites in the pigeon brain: binding characteristics, regional distribution and diurnal variation. J. Endocrinol. 128, 475–482. Zhang H., Akbar M. and Kim H. Y. (1999) Melatonin: an endogenous negative modulator of 12-lipoxygenation in rat pineal gland. Biochem. J. 344, 487–493. Ziv I., Zilkha-Falb R., Offen D., Shirvan A., Barzilai A. and Melamed E. (1997) Levodopa induces apoptosis in cultured neuronal cells – a possible accelerator of nigrostriatal degeneration in Parkinson’s disease? Mov. Disord. 12, 17–23.  2006 The Authors Journal Compilation  2006 International Society for Neurochemistry, J. Neurochem. (2007) 100, 736–746