Effect of Valproate,Lamotrigine and Levetiracetam on Excitability and Firing Properties of CA1 Neurons in Rat Brain Slices

The purpose of this study was to analyze the rapid effects of the antiepileptic drugs valproate, lamotrigine, and levetiracetam on excitability and firing properties of hippocampal neurons. The drug effects on resting potential, action potential, and repetitive firing properties were studied in whole-cell current-clamp recordings of CA1 neurons in rat brain slices. Lamotrigine changed action potential rising slope by −24 ± 38 V/s (mean ± SD), peak amplitude by −6.8 ± 5.0 mV, and maximum firing frequency by −60 ± 13%. Lamotrigine thereto increased the voltage threshold by 4.3 ± 4.2 mV and augmented the action potential attenuation during repetitive firing. All effects were significant (P < 0.01 to P < 0.0002) compared to control cells. Valproate and levetiracetam showed no significant effects on these parameters. None of the tested drugs had a significant effect on the resting potential. The lamotrigine effects are consistent with sodium channel blocking which may explain or contribute to the antiepileptic mode of action. Valproate and levetiracetam did not show these effects and the mechanism of their antiepileptic action need to be different. These findings (valproate) differ in some respects from findings reported in cultured or dissociated neurons. In a slice where the neurons have largely preserved connections, drug effects are likely to be more similar to the therapeutic action in the brain.     

Macrophage colony stimulating factor prevents NMDA-induced neuronal death in hippocampal organotypic cultures

Macrophage colony stimulating factor (M-CSF) and its receptor are up-regulated in the brain in Alzheimer's disease (AD), in transgenic mouse models for AD, and experimental models for traumatic and ischemic brain injury. M-CSF induces activation and proliferation of microglial cells and expression of proinflammatory cytokines. We examined the role of M-CSF in excitotoxic neuronal cell death in organotypic hippocampal cultures. NMDA treatment induced neuronal apoptosis and caspase-3 activation in organotypic hippocampal cultures, whereas treatment with M-CSF protected hippocampal neurons from NMDA-induced apoptosis. Caspase-3 activation was inhibited by M-CSF treatment to the same degree as with the caspase inhibitor Z-VAD-FMK. These results suggest that M-CSF has neuroprotective properties through inhibition of caspase-3 that could promote neuronal survival after excitotoxic insult. The role of M-CSF in neurological disease should be reevaluated as a microglial activator with potentially neuroprotective effects.     

Effect of thrombin on survival of hippocampal neurons

The effect of thrombin on the rat hippocampal neurons death in model of neurotoxicity induced by hemoglobin or glutamate, was studied. Thrombin (10 nM) was shown to inhibit 100-mkM glutamate--or 10-mkM hemoglobin-induced apoptosis of the rat hippocampal neurons. With the aid of PAR1 (protease-activated receptor1) agonist peptide and PAR1 antagonist, the PAR1 was found to be necessary for protective action of thrombin in hippocampal neurons in models of neurotoxicity induced by hemoglobin or glutamate. Because the prolonged elevation [Ca2+] ib neurons is a critical part of neurodestructive processes in CNS, the effect of thrombin on Ca2+-homeostatis of neurons after its injury by the inducer of neuronal apoptosis: a synthetic agonist of the NMDA receptors N-methyl-D-aspartate (NMDA), was studied. We hypothesized that thrombin via receptors PAR may prove to be neuroprotective for the hippocampus. Thrombin was shown to stimulate via PAR1 a transient increase in [Ca2+] in neurons in a concentration-dependent manner. Thrombin (1 nM) decreased the [Ca2+] signal induced by activation of the NMDA-subtype of glutamate receptors. This thrombin effect may be one of the reasons of the protective action of thrombin in hippocampal neurons.     

N-Methyl-d-Aspartate- or Glutamate-Mediated Toxicity in Cultured Rat Cortical Neurons Is Antagonized by FPL 15896AR

Abstract: The neuroprotective action of ( S )-α-phenyl-2-pyridineethanamine dihydrochloride (FPL 15896AR), a novel noncompetitive N -methyl- d -aspartate (NMDA) receptor antagonist, was examined in primary rat cortical neuronal cultures. Exposure of cortical cultures to NMDA (50 µ M ) or glutamate (50 µ M ) for 15 min resulted in the death of 85–95% of the neurons during the next 24 h. This neurotoxicity was completely eliminated by adding FPL 15896AR (50 µ M ) to the cultures during the time of NMDA or glutamate exposure. Neuroprotective concentrations of FPL 15896AR also inhibited other acute effects of NMDA. FPL 15896AR (50 µ M ) prevented the loss of membrane-associated protein kinase C activity that developed by 4 h after transient exposure to 50 µ M NMDA or 50 µ M glutamate. FPL 15896AR also reduced by ∼35% the magnitude of NMDA-triggered increases in intracellular free Ca²⁺ concentration in the cortical cultures. These data indicate that NMDA-mediated toxicity in cultured cortical neurons can be blocked by the NMDA antagonist FPL 15896AR.     

Inhibiting PTEN Protects Hippocampal Neurons against Stretch Injury by Decreasing Membrane Translocation of AMPA Receptor GluR2 Subunit

The AMPA type of glutamate receptors (AMPARs)-mediated excitotoxicity is involved in the secondary neuronal death following traumatic brain injury (TBI). But the underlying cellular and molecular mechanisms remain unclear. In this study, the role of phosphatase and tensin homolog deleted on chromosome 10 (PTEN) in GluR2-lacking AMPARs mediated neuronal death was investigated through an in vitro stretch injury model of neurons. It was indicated that both the mRNA and protein levels of PTEN were increased in cultured hippocampal neurons after stretch injury, which was associated with the decreasing expression of GluR2 subunits on the surface of neuronal membrane. Inhibition of PTEN activity by its inhibitor can promote the survival of neurons through preventing reduction of GluR2 on membrane. Moreover, the effect of inhibiting GluR2-lacking AMPARs was similar to PTEN suppression-mediated neuroprotective effect in stretch injury-induced neuronal death. Further evidence identified that the total GluR2 protein of neurons was not changed in all groups. So inhibition of PTEN or blockage of GluR2-lacking AMPARs may attenuate the death of hippocampal neurons post injury through decreasing the translocation of GluR2 subunit on the membrane effectively.     

Novel role of KCNQ2/3 channels in regulating neuronal cell viability

Overactivation of certain K(+) channels can mediate excessive K(+) efflux and intracellular K(+) depletion, which are early ionic events in apoptotic cascade. The present investigation examined a possible role of the KCNQ2/3 channel or M-channel (also named Kv7.2/7.3 channels) in the pro-apoptotic process. Whole-cell recordings detected much larger M-currents (212 ± 31 pA or 10.5 ± 1.5 pA/pF) in cultured hippocampal neurons than that in cultured cortical neurons (47 ± 21 pA or 2.4 ± 0.8 pA/pF). KCNQ2/3 channel openers N-ethylmaleimide (NEM) and flupirtine caused dose-dependent K(+) efflux, intracellular K(+) depletion, and cell death in hippocampal cultures, whereas little cell death was induced by NEM in cortical cultures. The NEM-induced cell death was antagonized by co-applied KCNQ channel inhibitor XE991 (10 μM), or by elevated extracellular K(+) concentration. Supporting a mediating role of KCNQ2/3 channels in apoptosis, expression of KCNQ2 or KCNQ2/3 channels in Chinese hamster ovary (CHO) cells initiated caspase-3 activation. Consistently, application of NEM (20 μM, 8 h) in hippocampal cultures similarly caused caspase-3 activation assessed by immunocytochemical staining and western blotting. NEM increased the expression of extracellular signal-regulated protein kinases 1 and 2 (ERK1/2), induced mitochondria membrane depolarization, cytochrome c release, formation of apoptosome complex, and apoptosis-inducing factor (AIF) translocation into nuclear. All these events were attenuated by blocking KCNQ2/3 channels. These findings provide novel evidence that KCNQ2/3 channels could be an important regulator in neuronal apoptosis.     

Effects of cannabinoids Δ(9)-tetrahydrocannabinol, Δ(9)-tetrahydrocannabinolic acid and cannabidiol in MPP+ affected murine mesencephalic cultures

Cannabinoids derived from Cannabis sativa demonstrate neuroprotective properties in various cellular and animal models. Mitochondrial impairment and consecutive oxidative stress appear to be major molecular mechanisms of neurodegeneration. Therefore we studied some major cannabinoids, i.e. delta-9-tetrahydrocannabinolic acid (THCA), delta-9-tetrahydrocannabinol (THC) and cannabidiol (CBD) in mice mesencephalic cultures for their protective capacities against 1-methyl-4-phenyl pyridinium (MPP(+)) toxicity. MPP(+) is an established model compound in the research of parkinsonism that acts as a complex I inhibitor of the mitochondrial respiratory chain, resulting in excessive radical formation and cell degeneration. MPP(+) (10 μM) was administered for 48 h at the 9th DIV with or without concomitant cannabinoid treatment at concentrations ranging from 0.01 to 10 μM. All cannabinoids exhibited in vitro antioxidative action ranging from 669 ± 11.1 (THC), 16 ± 3.2 (THCA) to 356 ± 29.5 (CBD) μg Trolox (a vitamin E derivative)/mg substance in the 1,1-diphenyl-2-picrylhydrazyl radical (DPPH) assay. Cannabinoids were without effect on the morphology of dopaminergic cells stained by tyrosine hydroxylase (TH) immunoreaction. THC caused a dose-dependent increase of cell count up to 17.3% at 10 μM, whereas CBD only had an effect at highest concentrations (decrease of cell count by 10.1-20% at concentrations of 0.01-10 μM). It influenced the viability of the TH immunoreactive neurons significantly, whereas THCA exerts no influence on dopaminergic cell count. Exposure of cultures to 10 μM of MPP(+) for 48 h significantly decreased the number of TH immunoreactive neurons by 44.7%, and shrunken cell bodies and reduced neurite lengths could be observed. Concomitant treatment of cultures with cannabinoids rescued dopaminergic cells. Compared to MPP(+) treated cultures, THC counteracted toxic effects in a dose-dependent manner. THCA and CBD treatment at a concentration of 10 μM lead to significantly increased cell counts to 123% and 117%, respectively. Even though no significant preservation or recovery of neurite outgrowth to control values could be observed, our data show that cannabinoids THC and THCA protect dopaminergic neurons against MPP(+) induced cell death.     

Protective effect of 20-HETE inhibition in a model of oxygen-glucose deprivation in hippocampal slice cultures

Recent studies have indicated that inhibitors of the synthesis of 20-hydroxyeicosatetraenoic acid (20-HETE) may have direct neuroprotective actions since they reduce infarct volume after ischemia reperfusion in the brain without altering blood flow. To explore this possibility, the present study used organotypic hippocampal slice cultures subjected to oxygen-glucose deprivation (OGD) and reoxygenation to examine whether 20-HETE is released by organotypic hippocampal slices after OGD and whether it contributes to neuronal death through the generation of ROS and activation of caspase-3. The production of 20-HETE increased twofold after OGD and reoxygenation. Blockade of the synthesis of 20-HETE with N-hydroxy-N'-(4-butyl-2-methylphenol)formamidine (HET0016) or its actions with a 20-HETE antagonist, 20-hydroxyeicosa-6(Z),15(Z)-dienoic acid, reduced cell death, as measured by the release of lactate dehydrogenase and propidium iodide uptake. Administration of a 20-HETE mimetic, 20-hydroxyeicosa-5(Z),14(Z)-dienoic acid (5,14-20-HEDE), had the opposite effect and increased injury after OGD. The death of neurons after OGD was associated with an increase in the production of ROS and activation of caspase-3. These effects were attenuated by HET0016 and potentiated after the administration of 5,14-20-HEDE. These findings indicate that the production of 20-HETE by hippocampal slices is increased after OGD and that inhibitors of the synthesis or actions of 20-HETE protect neurons from ischemic cell death. The protective effect of 20-HETE inhibitors is associated with a decrease in superoxide production and activation of caspase-3.     

Growth factors prevent changes in Bcl-2 and Bax expression and neuronal apoptosis induced by nitric oxide

Recent studies have shown that nitric oxide (NO) donors can trigger apoptosis of neurons, and growth factors such as insulin-like growth factor-1 (IGF-1) and basic fibroblast growth factor (bFGF) can protect against NO-induced neuronal cell death. The purpose of this study was to elucidate the possible mechanisms of NO-mediated neuronal apoptosis and the neuroprotective action of these growth factors. Both IGF-1 and bFGF prevented apoptosis induced by NO donors, sodium nitroprusside (SNP) or 3-morpholinosydnonimin (SIN-1) in hippocampal neuronal cultures. Incubation of neurons with SNP induced caspase-3-like activation following downregulation of Bcl-2 and upregulation of Bax protein levels in cultured neurons. Treatment of neurons with a bax antisense oligonucleotide inhibited the caspase-3-like activation and neuronal death induced by SNP. In addition, treatment of neurons with an inhibitor of caspase-3, Ac-DEVD-CHO, together with SNP did not affect the changes in the protein levels, although it inhibited NO-induced cell death. Pretreatment of cultures with either IGF-1 or bFGF prior to NO exposure inhibited caspase-3-like activation together with the changes in Bcl-2 and Bax protein levels. These results suggest that the changes in Bcl-2 and Bax protein levels followed by caspase-3-like activation are a component in the cascade of NO-induced neuronal apoptosis, and that the neuroprotective actions of IGF-1 and bFGF might be due to inhibition of the changes in the protein levels of the Bcl-2 family.     

Effect of enteropeptidase on survival of cultured hippocampal neurons under conditions of glutamate toxicity

The effects of full-size bovine enteropeptidase (BEK) and of human recombinant light chain enteropeptidase (L-HEP) on survival of cultured hippocampal neurons were studied under conditions of glutamate excitotoxicity. Low concentrations of L-HEP or BEK (0.1–1 and 0.1–0.5 nM, respectively) protected hippocampal neurons against the death caused by 100 μM glutamate. Using the PAR1 (proteinase-activated receptor) antagonist SCH 79797, we revealed a PAR1-dependent mechanism of neuroprotective action of low concentrations of enteropeptidase. The protective effect of full-size enteropeptidase was not observed at the concentrations of 1 and 10 nM; moreover, 10 nM of BEK caused death of 88.9% of the neurons, which significantly exceeded the cell death caused by glutamate (31.9%). Under conditions of glutamate cytotoxicity the survival of neurons was 26.8% higher even in the presence of 10 nM of L-HEP than in the presence of 10 nM BEK. Pretreatment of cells with 10 nM of either form of enteropeptidase abolished the protective effect of 10 nM thrombin under glutamate cytotoxicity. High concentrations of BEK and L-HEP caused the death of neurons mainly through necrosis.     

Activated protein C via PAR1 receptor regulates survival of neurons under conditions of glutamate excitotoxicity

The effect of an anticoagulant and cytoprotector blood serine proteinase--activated protein C (APC)--on survival of cultured hippocampal and cortical neurons under conditions of glutamate-induced excitotoxicity has been studied. Low concentrations of APC (0.01-10 nM) did not cause neuron death, but in the narrow range of low concentrations APC twofold and stronger decreased cell death caused by glutamate toxicity. High concentrations of APC (>50 nM) induced the death of hippocampal neurons similarly to the toxic action of glutamate. The neuroprotective effect of APC on the neurons was mediated by type 1 proteinase-activated receptor (PAR1), because the inactivation of the enzyme with phenylmethylsulfonyl fluoride or PAR1 blockade by a PAR1 peptide antagonist ((Tyr1)-TRAP-7) prevented the protective effect of APC. Moreover, APC inhibited the proapoptotic effect of 10 nM thrombin on the neurons. Geldanamycin, a specific inhibitor of heat shock protein Hsp90, completely abolished the antiapoptotic effect of 0.1 nM APC on glutamate-induced cytotoxicity in the hippocampal neurons. Thus, APC at low concentrations, activating PAR1, prevents the death of hippocampal and cortical neurons under conditions of glutamate excitotoxicity.     

Stretch-induced injury in organotypic hippocampal slice cultures reproduces in vivo post-traumatic neurodegeneration: role of glutamate receptors and voltage-dependent calcium channels

The relationship between an initial mechanical event causing brain tissue deformation and delayed neurodegeneration in vivo is complex because of the multiplicity of factors involved. We have used a simplified brain surrogate based on rat hippocampal slices grown on deformable silicone membranes to study stretch-induced traumatic brain injury. Traumatic injury was induced by stretching the culture substrate, and the biological response characterized after 4 days. Morphological abnormalities consistent with traumatic injury in humans were widely observed in injured cultures. Synaptic function was significantly reduced after a severe injury. The N-methyl-D-aspartate (NMDA) receptor antagonist MK-801 attenuated neuronal damage, prevented loss of microtubule-associated protein 2 immunoreactivity and attenuated reduction of synaptic function. In contrast, the NMDA receptor antagonists 3-[(R)-2-carboxypiperazin-4-yl]-propyl-1-phosphonic acid (CPP) and GYKI53655, were neuroprotective in a moderate but not a severe injury paradigm. Nifedipine, an L-type voltage-dependent calcium channel antagonist was protective only after a moderate injury, whereas omega-conotoxin attenuated damage following severe injury. These results indicate that the mechanism of damage following stretch injury is complex and varies depending on the severity of the insult. In conclusion, the pharmacological, morphological and electrophysiological responses of organotypic hippocampal slice cultures to stretch injury were similar to those observed in vivo. Our model provides an alternative to animal testing for understanding the mechanisms of post-traumatic delayed cell death and could be used as a high-content screen to discover neuroprotective compounds before advancing to in vivo models.     

Involvement of the synapse-specific zinc transporter ZnT3 in cadmium-induced hippocampal neurotoxicity

The present study examined the involvement of zinc (Zn)-transporters (ZnT3) in cadmium (Cd)-induced alterations of Zn homeostasis in rat hippocampal neurons. We treated primary rat hippocampal neurons for 24 or 48 hr with various concentrations of CdCl₂ (0, 0.5, 5, 10, 25, or 50 μM) and/or ZnCl ₂ (0, 10, 30, 50, 70, or 90 μM), using normal neuronal medium as control. By The CellTiter 96 ^® Aqueous One Solution Cell Proliferation Assay (MTS; Promega, Madison, WI) assay and immunohistochemistry for cell death markers, 10 and 25 μM of Cd were found to be noncytotoxic doses, and both 30 and 90 μM of Zn as the best concentrations for cell proliferation. We tested these selected doses. Cd, at concentrations of 10 or 25 μM (and depending on the absence or presence of Zn), decreased the percentage of surviving cells. Cd-induced neuronal death was either apoptotic or necrotic depending on dose, as indicated by 7-AAD and/or annexin V labeling. At the molecular level, Cd exposure induced a decrease in hippocampal brain-derived neurotrophic factor-tropomyosin receptor kinase B (BDNF-TrkB) and Erk1/2 signaling, a significant downregulation of the expression of learning- and memory-related receptors and synaptic proteins such as the NMDAR NR2A subunit and PSD-95, as well as the expression of the synapse-specific vesicular Zn transporter ZnT3 in cultured hippocampal neurons. Zn supplementation, especially at the 30 μM concentration, led to partial or total protection against Cd neurotoxicity both with respect to the number of apoptotic cells and the expression of several genes. Interestingly, after knockdown of ZnT3 by small interfering RNA transfection, we did not find the restoration of the expression of this gene following Zn supplementation at 30 μM concentration. These data indicate the involvement of ZnT3 in the mechanism of Cd-induced hippocampal neurotoxicity.     

A novel peptide prevents death in enriched neuronal cultures

We have recently cloned a novel protein (activity-dependent neuroprotective protein, ADNP) containing an 8-amino-acid, femtomolar-acting peptide, NAPVSIPQ (NAP). Here we show, for the first time, that NAP exerted a protective effect on glia-depleted neurons in culture. The number of surviving neurons was assessed in cerebral cortical cultures derived from newborn rats. In these cultures, a 24-h treatment with the beta-amyloid peptide (the Alzheimer's disease associated toxin) induced a 30-40% reduction in neuronal survival that was prevented by NAP (10(-13)-10(-11) M). Maximal survival was achieved at NAP concentrations of 10(-12) M. In a second set of experiments, a 5-day incubation period, with NAP added once (at the beginning of the incubation period) exhibited maximal protection at 10(-10) M NAP. In a third set of experiments, a 10-min period of glucose deprivation resulted in a 30-40% neuronal death that was prevented by a 24-h incubation with NAP. Glucose deprivation coupled with beta-amyloid treatment did not increase neuronal death, suggesting a common pathway. We thus conclude, that NAP can prevent neurotoxicity associated with direct action of the beta-amyloid peptide on neurons, perhaps through protection against impaired glucose metabolism.     

Effects of Vinpocetine on mitochondrial function and neuroprotection in primary cortical neurons

Vinpocetine (ethyl apovincaminate), a synthetic derivative of the Vinca minor alkaloid vincamine, is widely used for the treatment of cerebrovascular-related diseases. One of the proposed mechanisms underlying its action is to protect against the cytotoxic effects of glutamate overexposure. Glutamate excitotoxicity leads to the disregulation of mitochondrial function and neuronal metabolism. As Vinpocetine has a binding affinity to the peripheral-type benzodiazepine receptor (PBR) involved in the mitochondrial transition pore complex, we investigated whether neuroprotection can be at least partially due to Vinpocetine’s effects on PBRs.Neuroprotective effects of PK11195 and Ro5-4864, two drugs with selective and high affinity to PBR, were compared to Vinpocetine in glutamate excitotoxicity assays on primary cortical neuronal cultures. Vinpocetine exerted a neuroprotective action in a 1–50 μM concentration range while PK11195 and Ro5-4864 were only slightly neuroprotective, especially in high (>25 μM) concentrations. Combined pretreatment of neuronal cultures with Vinpocetine and PK11195 or Ro5-4864 showed increased neuroprotection in a dose-dependent manner, indicating that the different drugs may have different targets. To test this hypothesis, mitochondrial membrane potential (MMP) of cultured neurons was measured by flow cytometry. 25 μM Vinpocetine reduced the decrease of mitochondrial inner membrane potential induced by glutamate exposure, but Ro5-4864 in itself was found to be more potent to block glutamate-evoked changes in MMP. Combination of Ro5-4864 and Vinpocetine treatment was found to be even more effective.In summary, the present results indicate that the neuroprotective action of vinpocetine in culture can not be explained by its effect on neuronal PBRs alone and that additional drug targets are involved.     

Leptin counteracts the hypoxia-induced inhibition of spontaneously firing hippocampal neurons: a microelectrode array study

Besides regulating energy balance and reducing body-weight, the adipokine leptin has been recently shown to be neuroprotective and antiapoptotic by promoting neuronal survival after excitotoxic and oxidative insults. Here, we investigated the firing properties of mouse hippocampal neurons and the effects of leptin pretreatment on hypoxic damage (2 hours, 3% O(2)). Experiments were carried out by means of the microelectrode array (MEA) technology, monitoring hippocampal neurons activity from 11 to 18 days in vitro (DIV). Under normoxic conditions, hippocampal neurons were spontaneously firing, either with prevailing isolated and randomly distributed spikes (11 DIV), or with patterns characterized by synchronized bursts (18 DIV). Exposure to hypoxia severely impaired the spontaneous activity of hippocampal neurons, reducing their firing frequency by 54% and 69%, at 11 and 18 DIV respectively, and synchronized their firing activity. Pretreatment with 50 nM leptin reduced the firing frequency of normoxic neurons and contrasted the hypoxia-induced depressive action, either by limiting the firing frequency reduction (at both ages) or by increasing it to 126% (in younger neurons). In order to find out whether leptin exerts its effect by activating large conductance Ca(2+)-activated K(+) channels (BK), as shown on rat hippocampal neurons, we applied the BK channel blocker paxilline (1 μM). Our data show that paxilline reversed the effects of leptin, both on normoxic and hypoxic neurons, suggesting that the adipokine counteracts hypoxia through BK channels activation in mouse hippocampal neurons.     

Induction of activin A is essential for the neuroprotective action of basic fibroblast growth factor in vivo

Exogenous application of neurotrophic growth factors has emerged as a new and particularly promising approach not only to promote functional recovery after acute brain injury but also to protect neurons against the immediate effect of the injury. Among the various growth factors and cytokines studied so far, the neuroprotective and neurotrophic profile of basic fibroblast growth factor (bFGF) is the best documented. Using an animal model of acute excitotoxic brain injury, we report here that the neuroprotective action of bFGF, which is now being tested in stroke patients, depends on the induction of activin A, a member of the transforming growth factor-beta superfamily. Our evidence for this previously unknown mechanism of action of bFGF is that bFGF strongly enhanced lesion-associated induction of activin A; in the presence of the activin-neutralizing protein follistatin, bFGF was no longer capable of rescuing neurons from excitotoxic death; and recombinant activin A exerted a neuroprotective effect by itself. Our data indicate that the development of substances influencing activin expression or receptor binding should offer new ways to fight neuronal loss in ischemic and traumatic brain injury.