Melatonin-induced methylation of the ABCG2/BCRP promoter as a novel mechanism to overcome multidrug resistance in brain tumour stem cells.

BACKGROUND: Current evidence indicates that a stem cell-like sub-population within malignant glioblastomas, that overexpress members of the adenosine triphosphate-binding cassette (ABC) family transporters, is responsible for multidrug resistance and tumour relapse. Eradication of the brain tumour stem cell (BTSC) compartment is therefore essential to achieve a stable and long-lasting remission. METHODS: Melatonin actions were analysed by viability cell assays, flow cytometry, quantitative PCR for mRNA expression, western blot for protein expression and quantitative and qualitative promoter methylation methods. RESULTS: Combinations of melatonin and chemotherapeutic drugs (including temozolomide, current treatment for malignant gliomas) have a synergistic toxic effect on BTSCs and A172 malignant glioma cells. This effect is correlated with a downregulation of the expression and function of the ABC transporter ABCG2/BCRP. Melatonin increased the methylation levels of the ABCG2/BCRP promoter and the effects on ABCG2/BCRP expression and function were prevented by preincubation with a DNA methyltransferase inhibitor. CONCLUSION: Our results point out a possible relationship between the downregulation of ABCG2/BCRP function and the synergistic toxic effect of melatonin and chemotherapeutic drugs. Melatonin could be a promising candidate to overcome multidrug resistance in the treatment of glioblastomas, and thus improve the efficiency of current therapies.

Fas/Fas ligand regulation mediates cell death in human Ewing's sarcoma cells treated with melatonin.

BACKGROUND: Despite recent advances in cancer therapy, the 5-year survival rate for Ewing's sarcoma is still very low, and new therapeutic approaches are necessary. It was found previously that melatonin induces cell death in the Ewing's sarcoma cell line, SK-N-MC, by activating the extrinsic apoptotic pathway. METHODS: Melatonin actions were analysed by metabolic viability/survival cell assays, flow cytometry, quantitative PCR for mRNA expression, western blot for protein activation/expression and electrophoretic mobility shift assay for transcription factor activation. RESULTS: Melatonin increases the expression of Fas and its ligand Fas L, this increase being responsible for cell death induced by the indolamine. Melatonin also produces a transient increase in intracellular oxidants and activation of the redox-regulated transcription factor Nuclear factor-kappaB. Inhibition of such activation prevents cell death and Fas/Fas L upregulation. Cytotoxic effect and Fas/Fas L regulation occur in all Ewing's cell lines studied, and do not occur in the other tumour cell lines studied where melatonin does not induce cell death. CONCLUSION: Our data offers new insights in the study of alternative therapeutic strategies in the treatment of Ewing's sarcoma. Further attention deserves to be given to the differences in the cellular biology of sensitive tumours that could explain the cytotoxic effect of melatonin and the increase in the level of free radicals caused by this molecule, in particular cancer types.

Antioxidants do not prevent acrylonitrile-induced toxicity.

Several reports have recently described that acrylonitrile (ACN) toxicity resides in its capacity for inducing oxidative stress. ACN can be conjugated with glutathione (GSH), diminishing its cellular content, or being metabolized to cyanide. In the present report, we determine the effect of ACN on the viability of primary-cultured astrocytes as well as the oxidative damage generated by ACN by measuring GSH levels in primary cultured astrocytes. We also analyzed whether the ACN (2.5mM) toxicity could be avoided by using antioxidants such as taurine (5mM), N-acetylcysteine (20 mM), trolox (100 microM), estradiol (10 microM) and melatonin (100 nM-1mM). In this cell culture model, antioxidants were not able to prevent ACN-induced cell damage, with the exception of NAC, confirming that only GSH seems to play a key role in ACN-derived toxicity. Additionally, we measured different parameters of oxidative stress such as catalase activity, lipid peroxidation and GSH concentration, as indicators of the potential oxidative stress mediated by the toxicity of ACN, after exposure of Wistar rats to a concentration of 200 ppm ACN for 14 days. At the concentration assayed, we did not find any evidence of oxidative damage in the brain of ACN-treated rats.

Molecular characterization of ciprofloxacin resistance of gonococcal strains in Spain.

BACKGROUND: Over the past several years, the emergence of gonococcal isolates with intermediate or full resistance to fluoroquinolones has become a significant concern in several countries, including Spain. GOAL: The goal was to determine the occurrence of ciprofloxacin resistance among Neisseria gonorrhoeae strains in Spain during 2000 to 2001 and determine the frequency and patterns of mutations at gyrA, gyrB, and parC genes in these isolates. STUDY DESIGN: Eleven ciprofloxacin-resistant strains (with MICs ranging from 1 to 64 micrograms/mL) and two intermediate isolates (with MICs of 0.12 and 0.5 microgram/mL) were found. Mutations were identified by polymerase chain reaction and direct sequencing of the amplified products. RESULTS AND CONCLUSIONS: Alterations at Ser-91 and Asp-95 in GyrA were detected in all strains except one, an isolate for which the MIC was 0.12 microgram/mL. Alterations in ParC were more variable, and there was no clear correlation between the number of parC mutations and the level of resistance. No alterations at gyrB gene associated with ciprofloxacin resistance were found. The resistance was distributed among different types of strains, suggesting that the increase in the incidence of ciprofloxacin-resistant strains in Spain was not exclusively due to the appearance of a single-strain outbreak.

Glutamate induces oxidative stress not mediated by glutamate receptors or cystine transporters: protective effect of melatonin and other antioxidants.

Glutamate is responsible for most of the excitatory synaptic activity and oxidative stress induction in the mammalian brain. This amino acid is increased in the substantia nigra in parkinsonism due to the lack of dopamine restraint to the subthalamic nucleus. Parkinson's disease also shows an increase of iron levels in the substantia nigra and a decrease of glutathione, the antioxidant responsible for the ascorbate radical recycling. Considered together, these facts could make the antioxidant ascorbate behave as a pro-oxidant in parkinsonism. Since both glutamate and ascorbate are present in the synaptosomes and neurons of substantia nigra, we tested 1) if glutamate is able to induce oxidative stress independently of its excitatory activity, and 2) if ascorbate may have synergistic effects with glutamate when these two molecules co-exist. Brains were homogenized in order to disrupt membranes and render membrane receptors and intracellular signaling pathways non-functional. In these homogenates glutamate induced lipid peroxidation, indicating that this amino acid also may cause oxidative stress not mediated by its binding to glutamate receptors or cystine transporters. Ascorbate also induced lipid peroxidation thus behaving as a pro-oxidant. Both substances together produced an additive effect but they did not synergize. Given that melatonin is a potent physiological antioxidant with protective effects in models of neurotoxicity, we tested the role of this secretory product on the pro-oxidant effect of both compounds given separately or in combination. We also checked the protective ability of several other antioxidants. Pharmacological doses of melatonin (millimolar), estrogens, pinoline and trolox (micromolar) prevented the oxidant effect of glutamate, ascorbate, and the combination of both substances. Potential therapeutic application of these results is discussed.

Melatonin regulates glucocorticoid receptor: an answer to its antiapoptotic action in thymus.

We have previouslyreported that low doses of melatonin inhibit apoptosis in both dexamethasone-treated cultured thymocytes (standard model for the study of apoptosis) and the intact thymus. Here we elucidate the mechanism by which this agent protects thymocytes from cell death induced by glucocorticoids. Our results demonstrate an effect of melatonin on the mRNA for antioxidant enzymes in thymocytes, also showing an unexpected regulation by dexamethasone of these mRNA. Both an effect of melatonin on the general machinery of apoptosis and a possible regulation of the expression of the cell death related genes bcl-2 and p53 are shown not to be involved. We found melatonin to down-regulate the mRNA for the glucocorticoid receptor in thymocytes (glucocorticoids up-regulate their own receptor). The decrease by melatonin of mRNA levels for this receptor in IM-9 cells (where glucocorticoids down-regulate it) demonstrates that melatonin actually down-regulates glucocorticoid receptor. These findings allow us to propose the effects of melatonin on this receptor as the likely mediator of its thymocyte protection against dexamethasone-induced cell death. This effect of melatonin, given the oxidant properties of glucocorticoids, adds another mechanism to explain its antioxidant effects.

Ultrastructural confirmation of neuronal protection by melatonin against the neurotoxin 6-hydroxydopamine cell damage.

6-Hydroxydopamine (6-OHDA) is a neurotoxin used in the induction of experimental Parkinson's disease in both animals and cultured neuronal cells. Biochemical and molecular approaches showed previously that low doses of 6-OHDA induced apoptosis in PC12 cells, while high doses of this neurotoxin induced necrosis. Melatonin has been shown to protect against the neuronal programmed cell death induced by 6-OHDA, although it was not able to prevent the massive necrotic cellular death occurring after the addition of high doses of the neurotoxin. In the present work, we demonstrate by ultrastructural analysis that although low doses of 6-OHDA induced apoptosis in PC12 cells, it also damaged the non-apoptotic cells, morphologically corresponding this damage to incipient and reversible necrotic lesions. When the doses of the neurotoxin increase, there are still apoptotic cells, although most of the cells show necrotic irreversible lesions. We also found that melatonin partially prevents the incipient necrotic lesions caused by low doses of 6-OHDA. The fact that melatonin was shown in previous work to prevent apoptosis caused by low doses of 6-OHDA, but not necrosis induced by high doses of the neurotoxin, seemed to indicate that this agent is only able to protect against apoptosis. However, our present results, melatonin preventing also the incipient necrotic neuronal lesions, suggest that this hormone may provide a general protection against cell death, suggesting that higher doses should be tried in order to prevent the necrotic cell death induced by high doses of the neurotoxin.

Melatonin decreases mRNA for histone H4 in thymus of young rats.

The antiproliferative properties of melatonin have been previously demonstrated for several normal and tumoral tissues. In a recent report we have shown that melatonin is able to inhibit programmed cell death in thymus both, in vivo and in vitro. Given that other authors have related programmed cell death and cell proliferation and that no previous reports on melatonin and cell division exist on thymus, we decide to study the possible antiproliferative effect of melatonin in this organ measured as the levels of mRNA for the histone H4. We found that melatonin inhibits cell division on thymus when administered chronically both, at high (500 microg/body weight) and low (50 microg/body weight) dose. We also found a circadian rhythm of the mRNA for histone H4, opposed to the one previously described for melatonin, supporting the negative regulation by this hormone of cell division on thymus. A single dose of melatonin (50 microg/body weight) was not able to decrease the levels of mRNA for H4 in the time-points studied but after two hours of its administration. Finally, we report the inhibitory effect of melatonin in the cell proliferation of Harderian gland, brain, lung and kidney.

Inhibition of cell proliferation: a mechanism likely to mediate the prevention of neuronal cell death by melatonin.

In a previous work we demonstrated that melatonin is able to prevent apoptosis induced by low doses of 6-hydroxydopamine (6-OHDA) in undifferentiated and neuronal PC12 cells. We also reported how this neurohormone was able to prevent the decrease in the mRNA for antioxidant enzymes caused by 6-OHDA. Although the antioxidant capability of melatonin seems to be clearly implicated in its antiapoptotic activity, literature suggests that its antiproliferative property could also be involved in its prevention of apoptosis. In the present work we demonstrated that melatonin is able to inhibit cell proliferation in undifferentiated PC12 cells, decreasing cell number and the total amount of DNA, and the mRNA for the histone H4, which are known to increase during DNA synthesis. Melatonin does not decrease the number of cells in nonproliferating PC12 cells, indicating that it does not cause cell death. Additionally, we demonstrate that other inhibitors of cell proliferation, as well as other antioxidants, are able to mimic the antiapoptotic effect of melatonin. This is interpreted to mean that melatonin acts by both mechanisms to inhibit apoptosis caused by 6-OHDA and the findings support the hypothesis of a relationship between oxidative stress and regulation of the cell cycle.

Melatonin prevents apoptosis induced by 6-hydroxydopamine in neuronal cells: implications for Parkinson's disease.

It was recently reported that low doses of 6-hydroxydopamine (6-OHDA) induce apoptosis of naive (undifferentiated) and neuronal (differentiated) PC12 cells, and this system has been proposed as an adequate experimental model for the study of Parkinson's disease. The mechanism by which this neurotoxin damages cells is via the production of free radicals. Given that the neurohormone melatonin has been reported 1) to be a highly effective endogenous free radical scavenger, 2) to increase the mRNA levels and the activity of several antioxidant enzymes, and 3) to inhibit apoptosis in other tissues, we have studied the ability of melatonin to prevent the programmed cell death induced by 6-OHDA in PC12 cells. We found that melatonin prevents the apoptosis caused by 6-OHDA in naive and neuronal PC12 cells as estimated by 1) cell viability assays, 2) counting of the number of apoptotic cells, and 3) analysis and quantification of DNA fragmentation. Exploration of the mechanisms used by melatonin to reduce programmed cell death revealed that this chemical mediator prevents the 6-OHDA induced reduction of mRNAs for several antioxidant enzymes. The possibility that melatonin utilized additional mechanisms to prevent apoptosis of these cells is also discussed. Since this endogenous agent has no known side effects and readily crosses the blood-brain-barrier, we consider melatonin to have a high clinical potential in the treatment of Parkinson's disease and possibly other neurodegenerative diseases, although more research on the mechanisms is yet to be done.

Melatonin increases gene expression for antioxidant enzymes in rat brain cortex.

During the last years several reports have demonstrated that melatonin is a efficient free radical scavenger and general antioxidant. In addition, it has been shown that this neurohormone is able to increase the activity of glutathione peroxidase in rat brain cortex as well as the gene expression for some antioxidant enzymes in the Harderian gland of female Syrian hamster. Also, it is well known that brain cells are particularly exposed to free radicals, with antioxidant enzymes as the major defense mechanism that the brain uses to neutralize reactive oxygen species. The aim of the present study was to examine the influence of melatonin on gene expression for antioxidant enzymes in rat brain cortex. Our results clearly demonstrate that exogenously administered melatonin increases the levels of mRNA for glutathione peroxidase, copper-zinc superoxide dismutase, and manganese superoxide dismutase in this tissue. These stimulatory effects are observed after both acute and chronic treatment with this hormone, producing in the latter case the more marked increase. We therefore conclude that melatonin exerts an important role in providing indirect protection against free radical injury by stimulating gene expression for antioxidant enzymes. Consequently, melatonin could be considered as a potential therapeutic agent in some age-related neurodegenerative diseases where excessive free radical production has been implicated.

Androgen-dependent mast cell degranulation in the Harderian gland of female Syrian hamsters: in vivo and organ culture evidence.

In previous articles we have reported the "disappearance" of Harderian gland mast cells (HGMC) after treatment with testosterone. In the present work we study: (a) if the apparent decrease in the number of mast cells caused by this androgen is real or is due to the fact that testosterone induces mast cell degranulation that avoids its recognition by toluidine blue staining; (b) if testosterone acts through its receptor directly on the Harderian gland (HG). In order to give an answer to the first question, we observed HG of female Syrian hamsters treated with testosterone under the electron microscope to find the possible degranulated mast cells not recognizable with the aid of the toluidine blue staining. We also studied in vivo and in vitro the effects of the beta-agonists isoproterenol and salbutamol, given that they increase cAMP and can therefore prevent degranulation of mast cells. Finally we have used cytocalasin B, which inhibits degranulation by blocking actin depolimerization. Both the beta-agonists and cytochalasin B were able to prevent the decrease of mast cells, as recognized by staining with toluidine blue after treatment with testosterone. Indeed, when observed under the electron microscope, abundant degranulated mast cells were found after treatment with testosterone. For solving the second issue we analyzed the effect of the antiandrogen cyproterone acetate in vivo and in vitro. Our results demonstrate that testosterone is able to induce degranulation of HGMC in the Syrian hamster Mesocricetus auratus and that this effect is achieved directly through its receptor on the Harderian gland.

Antioxidative protection in a high-melatonin organism: the dinoflagellate Gonyaulax polyedra is rescued from lethal oxidative stress by strongly elevated, but physiologically possible concentrations of melatonin.

Lethal oxidative stress was investigated in the dinoflagellate Gonyaulax polyedra by measuring the dying-peak of bioluminescence during circadian phases of low physiological light emission, low bioluminescence capacity, and low sensitivity to stimulatory agents. Measurements were carried out in constant darkness after transfer of cells from light at CT 6 (circadian time, 0600 hr). H2O2 (0.08 mM), when administered 1 hr after transfer of cells, led to a multifold, long-lasting enhancement of light emission, which is typical for lethal cell damage. At the circadian phases of investigation, melatonin did not substantially stimulate bioluminescence up to concentrations of 0.5 mM. At this concentration, addition of melatonin prevented the dying-peak and reduced bioluminescence to almost basal values. The high concentration of melatonin applied is not unphysiological in Gonyaulax, because the indoleamine can increase to levels of several millimolar, e.g., in response to temperature signals. These protective effects of melatonin seem to be caused mainly by the direct action of melatonin as an antioxidant, because the major enzymes of antioxidative protection were not stimulated by melatonin, although some of them responded to H2O2. The activities of neither superoxide dismutase, hemoperoxidase/catalase, glutathione peroxidase, nor haloperoxidase were enhanced under the influence of melatonin; glutathione S-transferase activity increased only slightly.

Neurohormone melatonin prevents cell damage: effect on gene expression for antioxidant enzymes.

It is well known that porphyrins cause a toxic light-mediated effect due to their capability to generate free radicals. Several reports have proved that melatonin is a potent free radical scavenger. The aim of this work has been to study the ability of melatonin to prevent the cell damage caused by porphyrins in the Harderian gland of female Syrian hamsters. Cell injury was evaluated estimating the percentage of damaged cells found in the gland and analyzing the degree of this damage at ultrastructural level. To explain the mechanism by which this hormone could prevent the cell damage caused by porphyrins, its capability to both decrease porphyrin synthesis and increase the mRNA levels for antioxidant enzymes was evaluated. Our results demonstrate that melatonin administration decreases the percentage of damaged cells, porphyrin synthesis, and aminolevulinate synthase (ALA-S) mRNA levels and increases the mRNA levels for manganese superoxide-dismutase and copper-zinc superoxide dismutase. When observed under an electron microscope, the lesions in the clear cells of the treated females were much less severe than in the corresponding cells of the control animals. Melatonin exerts a cytoprotective effect by inhibiting the ALA-S gene expression (and so porphyrin synthesis) and by raising the mRNA levels for several antioxidant enzymes.

Isolation and identification of sex-specific cDNA clones from the Syrian hamster harderian gland.

Syrian hamster Harderian glands show a typical sexual dimorphism, with males having two secretory cell types and females having one cell type and intraluminal porphyrin accretions, among other differences. Since these differences may be due to the expression of specific genes, our interest is to identify those genes and their role on the development and control of the sexual dimorphism. The experimental approach was to construct cDNA libraries for male and female Syrian hamster Harderian glands and then subtracted libraries for male vs. female and for female vs. male. By this method, cDNA libraries enriched either in male-specific or in female-specific clones were obtained. Clones from those libraries were checked for differential expression by using double colony hybridization with [32P]-cDNA from male and female glands. Then, the selected clones were checked again for expression in Harderian glands by Northern hybridization, using poly(A+) RNA from males, castrated males, and females. Finally, the clones were sequenced and compared to search for significant homologies. One of the male-specific clones showed strong homology with rat cytochrome p450b/e. Among the female-specific clones, homologies were found to the complement C3 fragment from several species, to sequences from the mouse mammary tumor virus, and to the subunits C1 and C2 of the rat prostatic steroid binding protein. Several other clones showed no significant homologies and need further characterization.

Castration increases cell damage induced by porphyrins in the Harderian gland of male Syrian hamster. Necrosis and not apoptosis mediates the subsequent cell death.

It is known that the Harderian gland of male Syrian hamster synthesizes a much smaller amount of porphyrins than the gland of the female and that castration greatly increases this synthesis. We have studied in this experimental model the behavior of the different classes of secretory cells and their role in the synthesis of porphyrins, attempting to clarify the participation of these compounds in the cell damage leading to the formation of clear cells previously described in the gland of females. We have also investigated the mechanism underlying the death of these secretory cells after porphyrin accumulation (necrosis vs apoptosis). To achieve this, we have utilized the following techniques: (a) morphometrical; (b) ultrastructural; (c) biochemical (fluorescence spectrophotometry); and (d) molecular (DNA nick-end labeling in methacrylate sections and dot blot analysis). The glands from male hamsters (serving as control) present a very low rate of damaged cells that progressively rises after castration. This rise runs parallel to that of porphyrin synthesis, porphyrin deposits, and the decrease of Type II secretory cells. The damage and subsequent death of the secretory cells in the gland is produced by the deposit of porphyrins in the mitochondrial membrane. This porphyrin accumulation leads to a complete mitochondrial destruction that finally results in cell death and its secretion into the lumen. We finally conclude that this event is not a physiological cell death (apoptosis) but the consequence of the toxic accumulation of porphyrins (necrosis).

Invasive processes in the normal Harderian gland of Syrian hamster.

In this contribution we will pay special attention to several morphological findings that we can observe, under some circumstances, in the normal Harderian gland of the Syrian hamster. The accumulation of porphyrins in this gland results in mitochondrial damage and extensive cell death. Many damaged cells are secreted into the lumen of the tubule-alveoli, but most of them seem to produce an invasive process that even affects the vascular components of the gland. In this way, many blood vessels are invaded and appear partially filled with the invasive mass, which sometimes totally occludes the lumen of the vessels. We have also observed other surprising features related to a special kind of activity in certain secretory cells. Such activity results in a peculiar "segregation" of a cytoplasmic fragment, containing the nucleus. The affected cells seem to gather up their cytoplasm and nucleus towards the basal zone, while the rest of the cell, including practically the whole amount of lipid droplets, is relegated to the vicinity of the lumen. All these phenomena seem finally to result in the detachment of some clusters, composed of a limited number of cells, which display a basophilic cytoplasm practically free of lipid droplets.

Regulation of the aminolevulinate synthase gene in the Syrian hamster Harderian gland: changes during development and circadian rhythm and role of some hormones.

The Syrian hamster Harderian gland has been advocated as a model to study the porphyrin biosynthetic pathway, since it shows by far the highest porphyrin concentration known to date. Another particular characteristic is the sexual dimorphism at both the morphological and the biochemical levels. We found a variation in the ALV-S (aminolevulinate synthase) gene expression according to sex, with females exhibiting much higher mRNA levels than do males. After castration, ALV-S mRNA rose considerably in males, this increase being inhibited by darkness or treatment with melatonin. Treatment with hCG or progesterone did not vary the ALV-S mRNA levels in females. Castrated males, however, showed a much larger increase when they were treated with hCG. No variations have been found in the expression of the ALV-S gene in female HG throughout the estrous cycle. During development, males and females showed similar ALV-S mRNA levels until they were 20 days old. Afterwards, they started showing gender-associated differences. In females, ALV-S mRNA levels rose during the first 3 months of life, and thereafter they decreased progressively with aging. A circadian rhythm has been found in the gene expression of ALV-S mRNA in females, showing very low levels in the morning and reaching a peak during the first hours of darkness. It was an endogenous rhythm, probably regulated at the transcriptional level. It is proposed that the light-dark period duration modulates this rhythm through the suprachiasmatic nucleus which in turn acts on the pineal secretion of melatonin that regulates ALV-S gene expression.

The pineal neurohormone melatonin prevents in vivo and in vitro apoptosis in thymocytes.

Recently, melatonin was found to be the most potent physiological free radical scavenger known to date. In this work, we attempted to define the role this neurohormone plays in the regulation of apoptosis, since the effect of bcl-2, the main gene implicated in its inhibition, acts via an antioxidant mechanism. We investigated the role of melatonin in cell death of thymus, a well known model for the study of apoptosis. Two sets of experiments were carried out: in vivo experiments, performed with Wistar rats, and in vitro experiments, performed with primary cultures of young Wistar rat thymocytes treated with glucocorticoids in order to induce apoptosis. Morphometrical studies in semithin sections of thymus and analysis of DNA fragmentation by gel electrophoresis show that physiological apoptosis occurring in thymus of 65 days old rats, is prevented by the daily administration of melatonin beginning when the rats were 25 days old. Also, we found that at a concentration of 10(-7) M, melatonin decreases by 35% the percentage of apoptotic cells induced by glucocorticoids in cultured thymocytes of 25 day old rats. 10(-9) M melatonin decreases cell death by 20%. Finally, melatonin at 10(-11) M did not have any effect. Several hypothesis are discussed to explain this effect: direct interaction of melatonin with glucocorticoid receptors in the thymus; induction of interleukin-4 release; direct genomic action modulating the expression of apoptosis-inhibiting genes; an effect on nitric oxide synthase; and finally, the antioxidant action of melatonin. Since apoptosis is a possible mechanism involved in neuronal death shown in several neurodegenerative diseases such as Parkinson or Alzheimer's diseases, investigative efforts should be directed to the possible role of melatonin in inhibiting cell death in tissues other that the thymus. Melatonin might be a potent therapeutic agent in some of these conditions.