Basic fibroblast growth factor modulates sensitivity of cultured hippocampal pyramidal neurons to glutamate cytotoxicity: interaction with ganglioside GM1

Transient upregulation of the glial glutamate transporter GLAST in response to fibroblast growth factor, insulin-like growth factor and epidermal growth factor in cultured astrocytes

Although expression of the glial glutamate transporter GLAST is tightly regulated during development and under pathophysiological conditions, little is known about endogenous modulators of GLAST expression. Because growth factors are generally believed to regulate glial functions, we addressed their possible contribution to GLAST regulation in cultured rat astrocytes. Of the six growth factors tested (basic fibroblast growth factor (bFGF), insulin-like growth factor-1 (IGF-1), epidermal growth factor (EGF), insulin, platelet-derived growth factor, and hepatocyte growth factor), bFGF, IGF-1 and EGF enhanced [3H]glutamate transport activity in a concentration-dependent manner. These effects were accompanied by an increase in the Vmax value for transport activity and in GLAST protein and mRNA levels, which suggests that GLAST expression is transcriptionally regulated by the growth factors. Interestingly, the effects reached a peak after 36 hours of exposure to growth factors, and rapidly returned to baseline by 48 hours. A combination of IGF-1 with either bFGF or EGF showed an additive effect on the glutamate uptake activity, but a combination of bFGF and EGF did not. Pharmacological blockade of protein kinase C inhibited the effects of IGF-1 and EGF, but not bFGF. By contrast, genistein, an inhibitor of tyrosine kinases, blocked the effects of bFGF and EGF without affecting the effect of IGF-1. These results suggest that the growth factors activate different signaling pathways for GLAST upregulation. The present study may indicate a novel regulatory system of glial glutamate transporters.

Locally reduced levels of acidic FGF lead to decreased expression of 28-kda calbindin and contribute to the selective vulnerability of the neurons in the entorhinal cortex in Alzheimer's disease

Recent studies demonstrate that a disturbed calcium-homeostasis leading to increased susceptibility to excitotoxic triggers plays a major role in the neurodegenerative process initiating in layer 2 of the entorhinal cortex (EC2) during Alzheimer's disease (AD). Thus, proteins binding free Ca++ (i.e. calbindin) and factors regulating these proteins are of great importance for the neuroprotective-neurotoxic balance in the affected brain regions. In the present combined human and in vitro study evidence is provided that altered levels of the acidic fibroblast growth factor (aFGF) and calbindin expression are concomitantly present in EC2 neurons and have interactive effects. A dramatic loss of aFGF- and calbindin-labeled EC2 neurons was found. Further analysis of the surviving EC2 neurons revealed a strong immunoreactivity to calbindin and aFGF. In vitro experiments show that aFGF regulates calbindin expression, because treatment of differentiating neurons with recombinant aFGF increases calbindin expression in a time-dependent fashion. The data imply that a reduced expression of aFGF in EC2 neurons of AD brains leads to lower levels of calbindin resulting in decreased neuroprotective capacity.

Culture and regeneration of human neurons after brain surgery

Cortical human brain tissue was obtained from 11 craniotomies for intractable epilepsy or tumor resection. Neuregen transport medium preserved viability at 4 degrees C during transfer to the culture laboratory. Cells were isolated and cultured by methods previously developed for adult rat neurons (Brewer GJ. Isolation and culture of adult rat hippocampal neurons. J. Neurosci. Meth. 1997:71:143-55). In about 40% of the cases, cultures regenerated with a majority of neuron-like cells that stained for neurofilament and not GFAP. After 3 weeks of culture from a 70 year old meningioma case, synapse-like structures were revealed by electron microscopy. Trophic support from basic human recombinant fibroblast growth factor was synergistically improved with the steroid hormone dehydroepiandrosterone 3-sulfate. Another 40% of the cases resulted in cultures that were predominantly GFAP positive astroglia. The remaining 20% of the cases did not regenerate cells with neuron-like or glial processes. Three postmortem cases did not regenerate neurites. These methods may aid development of human culture models of epilepsy as well as human pharmacology, toxicology and development of improved methods for brain grafts.

Systemic treatment with GM1 ganglioside improves survival and function of cryopreserved embryonic midbrain grafted to the 6-hydroxydopamine-lesioned rat striatum

Cryopreservation may allow long-term storage of embryonic ventral mesencephalon (VM) for neural transplantation. We investigated whether the ganglioside GM1 or the lazaroid tirilazad mesylate (U-74006F) could improve survival of grafts derived from cryopreserved VM in a rat model of Parkinson's disease. VM was dissected from rat embryos (E14-E15), frozen and stored in liquid nitrogen under controlled conditions, thawed, dissociated, and then grafted into the 6-hydroxydopamine-lesioned rat striatum. In Experiment I, VM fragments were exposed in vitro either to GM1 (100 microM) or to lazaroid (0.3 microM) during all preparative steps. In Experiment II, rats receiving GM1-pretreated VM were, in addition, treated systematically with GM1 (30 mg/kg) daily for 3.5 weeks. Rats grafted with untreated cryopreserved or fresh VM were used as controls, respectively. Rats receiving fresh VM control grafts showed complete recovery from lesion-induced rotations after 6 weeks whereas rats grafted with cryopreserved VM (untreated or pretreated) did not recover. Cryografts contained significantly less (18%, control; 23%, GM1; and 12%, lazaroid) tyrosine hydroxylase-positive cells compared to fresh grafts (1415 +/- 153; mean +/- SEM). Graft volume was also significantly smaller after cryopreservation. In contrast, with additional systemic GM1 treatment cryografts contained almost the same number of tyrosine hydroxylase-positive cells (376 +/- 85) as fresh grafts (404 +/- 56), which was significantly more than that of untreated cryografts (147 +/- 20), showed a significantly larger volume (0.15 mm(3)) compared to that of untreated grafts (0.08 mm(3)) (fresh controls, 0.19 mm(3)), and induced significant and complete functional recovery in the rotation test. In conclusion, systemic treatment of rats with GM1 improved the low survival and functional inefficacy of grafts derived from cryopreserved VM whereas tissue pretreatment alone with either GM1 or lazaroid was not effective.

GM2 promotes ciliary neurotrophic factor-dependent rescue of immortalized motor neuron-like cell (NSC-34)

We have examined whether ciliary neurotrophic factor (CNTF) can alter serum-free cell survival of immortalized motor neuron-like cells, which were established by fusing mouse neuroblastoma N18TG2 with mouse motor neurons. One of the cell lines, NSC-34 exhibited cell survival in the presence of CNTF. NSC-34 preserves the most characteristics of motor neurons, such as the formation of neuromuscular junctions on co-cultured myotube. GM2 ganglioside is characteristic of motor neurons, and expressed highly in NSC-34. When NSC-34 was cultured with exogenous GM2 ganglioside and CNTF, GM2 facilitated the cell survival effect of CNTF. In the addition, beta 1,4 N-acetylgalactosaminyltransferase (GM2 synthase) activity was enhanced up to 3.9-fold by culture in the presence of CNTF. GM2 might be a functional modulator of CNTF in motor neurons. It might be presented to cell surface by its enzyme activation, and become a signal of early stage, when CNTF rescues motor neurons.

Promotion of neuronal survival by GM1 ganglioside. Phenomenology and mechanism of action

The purpose of this article is to review recent findings regarding the mechanisms by which GM1 may mimic or potentiate certain actions of neurotrophic factors, including promotion of neuronal survival. It is proposed that the neuroprotective activity of GM1 is due, at least in part, to its ability to favor the dimerization of neurotrophic factor tyrosine kinases and thereby mimicking the action of their corresponding ligands. This may manifest both in the absence of ligand (thereby triggering a subset of neurotrophic-factor responses such as prevention of apoptosis) and in the presence of ligand (thereby potentiating responses to neurotrophic factors).

Neurite outgrowth is enhanced by anti-idiotypic monoclonal antibodies to the ganglioside GM1

Exogenously added gangliosides enhance sprouting, neurite outgrowth, and other neuronal activities; this effect may be initiated when a ganglioside binds to a membrane protein or when a ganglioside intercalates into the plasma membrane. To test whether binding to membrane proteins is sufficient for ganglioside-mediated activity, anti-idiotypic antibodies were generated that mimic the functional binding sites of the ganglioside GM1 as described by M. J. Riggott and W. D. Matthew (1996, Glycobiology, 6, 581-589). These anti-idiotypic antibodies are proteinaceous probes that model the biochemical and biological effects of gangliosides. Those anti-idiotypic ganglioside (AIG) monoclonal antibodies (mAb's) were selected based on their ability to bind a known GM1 binding protein, the beta-subunit of cholera toxin. These studies described neuronal cell surface proteins that were identified by immunocytochemistry and Western blotting using these AIG mAb's. Here we show that AIG mAb's mimic the functional properties of GM1 in that they facilitate neurite outgrowth from central and peripheral nervous system neurons in in vitro bioassays. In addition, AIG mAb binding modulates second messenger activity, suggesting that membrane protein binding alone is sufficient to invoke intracellular activation. The similarity in the pattern of protein tyrosine phosphorylation evoked by GM1 and the anti-idiotypic ganglioside antibodies suggests that the AIG mAb's modulate neurite outgrowth in a manner similar to that of GM1. Because antibodies cannot intercalate into the plasma membrane, these results suggest that the ganglioside GM1 can mediate neuronal cellular activity by binding to cell surface proteins.

Basic fibroblast growth factor is highly neuroprotective against seizure-induced long-term behavioural deficits

Basic fibroblast growth factor has been reported to protect neurons of various structures from excitotoxic damage. To study the effects of basic fibroblast growth factor on seizure-induced brain damage we infused the growth factor into the lateral ventricles of 35-day-old rats receiving convulsant dosages of kainic acid. Artificial cerebrospinal fluid or basic fibroblast growth factor at dosages of 0.5 ng/h or 2.5 ng/h was infused into the lateral ventricle continuously for seven days starting two days before and continuing for five days after the animals had kainic acid-induced status epilepticus. At age 80 days the animals underwent behavioural testing using the water maze, open field, and handling tests and at age 95 days were tested for seizure threshold using flurothyl inhalation. Neither artificial cerebrospinal fluid or basic fibroblast growth factor modified the latency or duration of the acute seizures following kainic acid. However, rats infused with 2.5 ng/h, but not 0.5 ng/h of basic fibroblast growth factor, had fewer spontaneous recurrent seizures, a higher seizure threshold, better performance in the handling, open field and water maze test, and less cell loss in the hippocampus when compared to rats receiving artificial cerebrospinal fluid or 0.5 ng/h of basic fibroblast growth factor. These results show that basic fibroblast growth factor has a dose-related neuroprotective effect against seizure-induced long-term behavioural deficits when administered by osmotic pump prior to seizure onset. This neuroprotective effect is not related to an anticonvulsant effect.

Transient cerebral ischaemia in Mongolian gerbils pre-exposed to hypoxia

The objective of this study was to clarify whether pre-exposure to hypoxia influences neuronal death following transient cerebral ischaemia. Twenty gerbils were exposed to 10% oxygen in a chamber for 3 weeks. The other control gerbils (n = 20) were fed in normoxia for 3 weeks. Both carotid arteries in the neck were occluded with aneurysm clips for 5 minutes under halothane anaesthesia in 30 gerbils, recirculated and then fed in normoxia. Five animals in both groups were sacrificed before, and 2, 4, and 7 days after surgery. The animals were fixed with 4% paraformaldehyde and histological study was performed. Immunohistochemical study was also done with antibodies against basic fibroblast growth factor (bFGF) and vascular endothelial growth factor (VEGF). The neuronal death in the hippocampus was more severe in the hypoxic group. Expression of both bFGF and VEGF was obvious in the cingulate cortex, corpus callosum and internal capsule before clipping in the hypoxic group, but not observed in the normoxic group before clipping. We observed the expression of both bFGF and VEGF widely in the brain at 2 and 4 days after recirculation in both groups. The expression in the hypoxic group was much more prominent than that in the normoxic group. These expressions were not observed at 7 days in both groups. Pre-exposure to hypoxia followed by transient cerebral ischaemia accelerated neuronal death in the hippocampus, and induced the more obvious expression of both VEGF and bFGF compared with those in the normoxic group.

Elevated basic fibroblast growth factor levels in stroke-prone spontaneously hypertensive rats

Basic fibroblast growth factor is a biologically active polypeptide with mitogenic, angiogenic and neurotrophic properties. In the present study, the temporal and spatial expressions of basic fibroblast growth factor in stroke-prone spontaneously hypertensive rats were compared to two related strains of rat: spontaneously hypertensive rats and normotensive Wistar Kyoto rats. Higher levels of total RNA concentration were found in cerebral cortex of four-week-old stroke-prone rats compared to spontaneously hypertensive rats and Wistar Kyoto rats. Northern blot analysis showed no changes in levels of basic fibroblast growth factor messenger RNA with increasing age in cerebral cortex of Wistar Kyoto and spontaneously hypertensive rats. However, significant increases were found in 26- and 38-week-old stroke-prone rats compared to four-week-old stroke-prone rats. Although messenger RNA increases were also found in subcortical and cerebellar regions, a significant difference in levels of basic fibroblast growth factor messenger RNA was observed only in cerebral cortices among these three strains. This age-related increase in basic fibroblast growth factor messenger RNA correlated with the increase incidence of stroke in stroke-prone rats. Immunohistochemical study further revealed a dramatic increase in levels of basic fibroblast growth factor immunoreactivity in cerebral cortex of 30-week-old stroke-prone rats as compared to young stroke-prone rats, as well as age-matched Wistar Kyoto and spontaneously hypertensive rats. This increase in basic fibroblast growth factor immunoreactivity therefore appears very specific to aged stroke-prone rats. However, immunoreactivity decreased once severe tissue damages were observed in the cerebral cortex. Basic fibroblast growth factor-positive cells were diffusely expressed in cerebral cortex; double staining with glial fibrillary acidic protein showed the majority of these basic fibroblast growth factor-positive cells to be astrocytes. In summary, although young stroke-prone spontaneously hypertensive rats showed significantly higher RNA concentration, significant increases in levels of basic fibroblast growth factor, including both messenger RNA and protein expression, were observed in aged stroke-prone rats with a high incidence of stroke. These findings suggest the possibility that basic fibroblast growth factor may play a role in the developmental sequelae of cerebral lesions in stroke-prone spontaneously hypertensive rats.

Cultured skin fibroblasts as a cell model for investigating schizophrenia

Cultured skin fibroblasts, among other non-neuronal cells (e.g. platelets, lymphocytes, red blood cells), provide an advantageous system for investigating dynamic molecular regulatory processes underlying abnormal cell growth, metabolism, and receptor-mediated signal transduction, without the confounding effects of disease state and its treatment in a variety of brain disorders, including schizophrenia, and are useful for studies of systemic biochemical defects with predominant consequences for brain function. These cells are also useful for studying aspects of neurotransmitter functions because the cells express enzymes involved in their metabolism, as well as their receptors with complete machinery for signal transduction. These processes also function predictably with receptors that are transfected in fibroblasts. This review will focus on the use of cultured skin of which have also been studied in post-mortem brains. These mechanisms might involve DNA processing and mitogenesis, cell-cell adhesion molecules, actions of growth factors, oxidative damage, and membrane phospholipid derived second messengers. This review will further discuss the implications of these processes to clinical and structural brain abnormalities. An understanding of these biochemical processes might help establish therapeutic implications and identify the risk for illness through experimental strategies such as epidemiology, family pedigree and high risk populations. Finally, despite some methodological limitations, skin fibroblasts are relatively easy to grow and maintain as primary cultures or as immortalized cell lines for long periods of time for use in investigating newly identified biochemical abnormalities.

Effects of basic fibroblast growth factor and ganglioside GM1 on neuronal survival in primary cultures and on eight-arm radial maze task in adult rats following partial fimbria transections

The effects of basic fibroblast growth factor (bFGF) and ganglioside GM1 (GM1) were evaluated alone and simultaneously in two types of experiments. First, the neuronal survival of primary culture neurons from fetal rat brain was measured. Then, performance on radial maze task in adult male rats following bilateral partial Fimbria-Fornix transections (F-F lesion) was tested. In primary culture neurons, bFGF (1-10 ng/ml) supported the neuronal survival from three regions (hippocampus, cortex and septum) of embryonic rat brain. However, GM1 (0.1-10 micrograms/ml) did not support the neuronal survival from any regions. Survival of cultured neurons was not supported by addition of 0.1 ng/ml bFGF, but when bFGF (0.1 ng/ml) and GM1 (0.1, 1 microgram/ml) were given to the cultured neurons simultaneously, the number of surviving neurons increased significantly. In the eight-arm radial maze task, where only the same four arms were baited, F-F lesion produced substantial memory impairment. In this task, administration of bFGF (10 micrograms/ml) or GM1 (1 mg/ml) alone did not produce any effects. However, when they were given simultaneously, the number of working memory errors decreased significantly, in spite of no amelioration for hippocampal choline acetyl transferase (ChAT) depletion. These findings indicate that actions of bFGF may be potentiated by the addition of GM1 in both primary neuronal cultures and radial maze task performance. These results suggest that the combination of bFGF and GM1 may synergistically improve spatial memory deficits.

A comprehensive analysis of the distribution of FGF-2 and FGFR1 in the rat brain

We have examined the cellular distribution of both FGF-2 and FGFR1 immunoreactivity and their mRNAs throughout the normal adult rat brain in order to reconcile numerous disparate findings in the published literature. The results confirm a widespread distribution of FGF-2 and FGFR1 in the rat brain, and different regions express distinct patterns of FGF-2 and FGFR1 mRNA and protein: neuronal and non-neuronal cells show different subcellular distributions that vary according to the area where they are located. The intensity of the staining and hybridization also varies according to the loci examined and the cell type involved. Astrocytes contain the highest levels of FGF-2 and FGFR1 mRNAs, and characteristically, possess high levels of immunoreactive FGF-2 within the nucleus. Amongst non-neuronal cells, oligodendrocytes do not synthesize or contain significant levels of FGF-2 immunoreactivity however, they do express FGFR1 mRNA. In these cells, immunoreactive FGFR1 is mainly associated with the myelin sheaths of neuronal fibers. In ventricular systems, ependymal cells synthesize and contain immunoreactive FGFR1. In contrast, only cells lining the lateral wall of the IIIrd ventricle express FGF-2 mRNA. Subependymal cells contain high levels of both FGF-2 and FGFR1 immunoreactivity. Neurons express low levels of FGF-2 mRNA and immunoreactive FGF-2 is localized predominantly to the perikaryon. However, selected populations of neurons, such as CA2 field of the hippocampus, show high levels of FGF-2 mRNA, in which the nucleus is strongly immunopositive. Similarly, high levels of FGFR1 mRNA are localized to select populations of neurons (e.g. amygdala). FGFR1 immunoreactivity is mainly associated with myelinated fiber tracts (e.g. striatum), and some neurons show immunoreactivity in the perikaryon (e.g. hippocampus), the nucleus (e.g. mesencephalic trigeminal nucleus), or in axonal projections (e.g. hypothalamus). Remarkably, in many of the areas studied, FGF-2 and FGFR1 mRNA and/or their translated protein do not co-localize in neurons (e.g. neo-cortices) or even in the same regions of the brain (e.g. substantia nigra). In other instances, mRNAs for both FGF-2 and FGFR1 colocalize (e.g. supraoptic nucleus). The brain, in contrast to peripheral tissues, contains high levels of FGF-2 and actively expresses its gene under normal physiological conditions. The highly specific anatomical distribution of immunoreactive FGF-2 in neuronal and non-neuronal brain cells, supports the notion that it plays a multifunctional role in the CNS under normal physiology. By correlating the localization and the synthesis of FGF-2 and one of its high affinity receptors, FGFR1, in the CNS, it should be possible to obtain a better understanding of the roles of FGF-2 in normal and pathological conditions.

Auditory brainstem responses of CBA/J mice with neonatal conductive hearing losses and treatment with GM1 ganglioside

Exogenous administration of GM1 ganglioside to CBA/J mice with a neonatal conductive hearing loss ameliorates the atrophy of spiral ganglion neurons, ventral cochlear nucleus neurons, and ventral cochlear nucleus volume. The present investigation demonstrates the extent of a conductive loss caused by atresia and tests the hypothesis that GM1 ganglioside treatment will ameliorate the conductive hearing loss. Auditory brainstem responses were recorded from four groups of seven mice each: two groups received daily subcutaneous injections of saline (one group had normal hearing; the other had a conductive hearing loss); the other two groups received daily subcutaneous injections of GM1 ganglioside (one group had normal hearing; the other had a conductive hearing loss). In mice with a conductive loss, decreases in hearing sensitivity were greatest at high frequencies. The decreases were determined by comparing mean ABR thresholds of the conductive loss mice with those of normal hearing mice. The conductive hearing loss induced in the mice in this study was similar to that seen in humans with congenital aural atresias. GM1 ganglioside treatment had no significant effect on ABR wave I thresholds or latencies in either group.

Basic fibroblast growth factor protects against excitotoxicity and chemical hypoxia in both neonatal and adult rats

Basic fibroblast growth factor (bFGF) is a polypeptide growth factor that promotes neuronal survival. We recently found that systemic administration of bFGF protects against both excitotoxicity and hypoxia-ischemia in neonatal animals. In the present study, we examined whether systemically administered bFGF could prevent neuronal death induced by intrastriatal injection of N-methyl-D-aspartate (NMDA) or chemical hypoxia induced by intrastriatal injection of malonate in adult rats and 1-methyl-4-phenylpyridinium (MPP+) in neonatal rats. Systemic administration of bFGF (100 micrograms/kg) for three doses both before and after intrastriatal injection of either NMDA or malonate in adult rats produced a significant neuroprotective effect. In neonatal rats, bFGF produced dose-dependent significant neuroprotective effects against MPP+ neurotoxicity, with a maximal protection of approximately 50% seen with either a single dose of bFGF of 300 micrograms/kg or three doses of 100 micrograms/kg. These results show that systemic administration of bFGF is effective in preventing neuronal injury under circumstances in which the blood-brain barrier may be compromised, raising the possibility that this strategy could be effective in stroke.

Enhanced restoration of striatal dopamine concentrations by combined GM1 ganglioside and neurotrophic factor treatments

Intraperitoneal injection of GM1 ganglioside or intracerebroventricular infusion of basic fibroblast growth factor (FGF-2) or epidermal growth factor (EGF) partially restored dopamine and 3,4-dihydroxyphenylacetic acid (DOPAC) levels in the striatum of young MPTP-treated mice. Combined treatments of GM1 ganglioside with FGF-2 or EGF produced a greater restoration of striatal dopamine levels than treatments with GM1 or either of the neurotrophic factors alone. GM1 treatment, but not trophic factor treatments caused significant sparing of substantia nigra pars compacta (SNc) tyrosine hydroxylase (TH)-positive neurons. These results confirm previous findings that GM1 provides trophic support for damaged dopamine neurons and suggests that GM1, FGF-2, and EGF may also enhance dopaminergic function in residual neurons. The results also suggest that a potentially fruitful approach to treating degenerative disorders of the dopamine system may be the use of combined trophic factor therapies.

The effect of cycloheximide and ganglioside GM1 on the viability of retinotectally projecting ganglion cells following ablation of the superior colliculus in neonatal rats

The time-course and extent of death of retinal ganglion cells (RGCs) following ablation of the superior colliculus (SC) in neonatal Wistar rats has recently been described [Harvey, A. R. and Robertson, D. (1992) J. Comp. Neurol., 325, 83-94]. Normal and pyknotic nuclei of retinotectally projecting ganglion cells were visualized using the fluorescent retrograde tracer diamidino yellow (DY), which had been injected into the SC at P2 (day of birth = P0), 2 days prior to tectal removal. The present report sets out to determine whether cycloheximide, an inhibitor of protein synthesis, or ganglioside GM1 reduced this lesion-induced RGC death. All surgery was carried out under ether anaesthesia; DY was injected into the left SC at P2 and the injected area was removed at P4. Cycloheximide (20-500 ng) was injected into the vitreous chamber of the right eye immediately after the lesion and again 11-12 h later. In some rats, cycloheximide administration was delayed until 12 h after the SC ablation. Control rats received SC lesions alone or lesions plus sham eye injections of saline. Different doses of GM1 were applied i.p. or intraocularly. Rats were perfused 24 h after the SC lesion, at the time of peak RGC death. Retinae of lesion only or sham eye injected rats contained approximately 11% pyknotic RGCs and the density of normal RGCs was approximately 3400/mm2. The rate of pyknosis in cycloheximide treated retinae was reduced to approximately 3%. Normal RGC density in these retinae was approximately 5500/mm2, similar to that found in retinae of unlesioned animals.(ABSTRACT TRUNCATED AT 250 WORDS)

Fibroblast growth factors: activities and significance of non-neurotrophin neurotrophic growth factors

Although first characterized by virtue of their ability to stimulate endothelial cell proliferation in vitro and angiogenesis in vivo, the fibroblast growth factors are now also well recognized for their neurotrophic activities. Understanding the physiological significance of these multifunctional, virtually ubiquitous and pluripotential molecules, however, remains enigmatic. Recent advances describing their molecular, biochemical and biological characteristics has led to a better understanding of their role in the central nervous system.

Fast and widespread increase of basic fibroblast growth factor messenger RNA and protein in the forebrain after kainate-induced seizures

Basic fibroblast growth factor promotes the survival and outgrowth of neurons and protects neurons from glutamate mediated excitotoxicity. The present study investigates the effects of kainate-induced epileptic seizures on the cellular expression of basic fibroblast growth factor messenger RNA and protein. Seizures were induced by injection of 12 mg/kg kainic acid. Rats were killed 3 h, 6 h, and 24 h after injection of the drug and analysed by radioactive and non-radioactive in situ hybridization as well as immunohistochemistry for glial fibrillary acidic protein and basic fibroblast growth factor. Radioactive in situ hybridization revealed a fast (6 h), strong (300-400% of control) and widespread increase of basic fibroblast growth factor messenger RNA after kainate-induced seizures. Non-radioactive in situ hybridization using digoxigenin-labeled riboprobes combined with glial fibrillary acidic protein immunohistochemistry showed that basic fibroblast growth factor messenger RNA was markedly increased in astroglial cells throughout the brain. Immunohistochemistry for basic fibroblast growth factor revealed labeling of nuclei in astrocytes in many forebrain areas and in neurons in area CA2 and fasciola cinereum. Kainate markedly increased basic fibroblast growth factor-like immunoreactivity in nuclei of astrocytes in several forebrain areas. This effect peaked 24 h after injection. It is concluded that basic fibroblast growth factor may play a neuroprotective role in kainate mediated excitotoxicity as seen from a massive and widespread astroglial increase in basic fibroblast growth factor messenger RNA and -like immunoreactivity. These effects may, to a large degree, be mediated through the excessive release of endogenous glutamate, induced by the epileptic seizures, leading to activation of glutamate receptors on astroglial cells through volume transmission, i.e. via diffusion of electrochemical signals in the extracellular fluid pathways.

Basic fibroblast growth factor protects cerebellar neurons in primary culture from NMDA and non-NMDA receptor mediated neurotoxicity

We have investigated the ability of bFGF to protect cerebellar neurons from neurotoxicity by excitatory amino acids. We have found that preincubation with 1-2.5 nM bFGF for 1-6 days significantly protected neurons from excitotoxic damage via NMDA receptors as well as ionotropic non-NMDA receptors. bFGF neuroprotection appeared not to be dependent upon neuronal differentiation and was not mimicked by other neurotrophins including BDNF, NT-3 and NGF. A greater rise in extracellular calcium-dependent cGMP formation, following either depolarization or excitatory amino acid receptor activation was observed in bFGF-pretreated neurons. We suggest that neuroprotection from excitotoxicity following bFGF treatment may be associated to the modulation of neurochemical pathways dependent upon extracellular calcium influx.