Alterations in Seizure Susceptibility and in Seizure-induced Plasticity after Pharmacologic and Genetic Manipulation of the Fibroblast Growth Factor-2 System


  • Silvia Zucchini,

    1. Department of Clinical and Experimental Medicine, Section of Pharmacology, and Neuroscience Center, University of Ferrara, Ferrara, Italy
    Search for more papers by this author
  • Mario Barbieri,

    1. Department of Clinical and Experimental Medicine, Section of Pharmacology, and Neuroscience Center, University of Ferrara, Ferrara, Italy
    Search for more papers by this author
  • Michele Simonato

    1. Department of Clinical and Experimental Medicine, Section of Pharmacology, and Neuroscience Center, University of Ferrara, Ferrara, Italy
    Search for more papers by this author

Address correspondence and reprint requests to Dr. S. Zucchini at Department of Experimental and Clinical Medicine, Section of Pharmacology, University of Ferrara, via Fossato di Mortara 17–19, 44100 Ferrara, Italy. E-mail:


Summary: Purpose: The adult brain undergoes activity-dependent plastic modifications during pathologic processes that are reminiscent of those observed during development. For example, seizures induce neuronal loss, neurogenesis, axonal and dendritic sprouting, gliosis, and circuit remodeling. Neurotrophic factors and fibroblast growth factor-2 (FGF-2), in particular, are well-known mediators in each of these cellular events. The aim of this minireview is to summarize and discuss the data supporting the idea that FGF-2 may be involved in seizure generation and in their sequelae.

Methods: We used epilepsy models of kainate and kindling, with FGF-2 knockout mice and FGF-2 overexpressing mice.

Results: Seizures increase FGF-2 mRNA and protein levels in specific brain areas and upregulate the expression of its receptor FGFR-1. Short-term intrahippocampal injection of FGF-2 cause seizures, whereas long-term i.c.v. infusion of low-dose FGF-2 does not affect kainate seizures but promotes behavioral recovery and reduces hippocampal damage. Kainate seizure severity is not altered in FGF-2 knockout mice, but is increased in FGF-2 overexpressing mice.

Conclusions: FGF-2 is implicated in seizure susceptibility and in seizure-induced plasticity.

The fibroblast growth factor-2 (FGF-2, previously named basic fibroblast growth factor, bFGF) is a neurotrophic factor belonging to a family of structurally related proteins, the FGFs, that encompasses 23 members (1,2). Although the FGFs play key roles in a variety of biologic events in development and adult life, the specific function of each FGF in vivo is not yet fully established. FGF-2 is a single-chain polypeptide composed of 146 amino acids, originally purified from the bovine pituitary by high-affinity binding to heparin and named after its property of promoting the growth of fibroblasts (3). It is found in at least four distinct isoforms: the high molecular weight (HMW) forms (22, 22.5, and 24 kDa) are predominantly localized in the nucleus, whereas the 18-kDa form resides in the cytoplasm. The HMW and the 18-kDa forms differ in the amino-terminal extension and in the nuclear targeting signal (4,5). It is plausible that different molecular weight proteins with different subcellular localization also have different biologic effects. FGF-2 lacks a signal peptide for secretion via the classic endoplasmic reticulum/Golgi-dependent mechanism (6). However, it is thought that FGF-2 can be released by cells via alternative mechanisms, because it is detected in the extracellular environment in many tissues. Furthermore, passive processes such as cell death, wounding, or chemical injury can release FGF-2 from cells (7).

FGF-2 exerts its biologic activities by interacting with four high-affinity and one low-affinity cell-surface receptors, termed FGFRs (8). The low-affinity receptor is a heparan sulfate proteoglycan (9). High-affinity FGFRs (FGFR-1 through FGFR-4) show substantial sequence homology and appear to be similar in their general structure: they all are tyrosine kinase receptors with one membrane-spanning domain and two or three extracellular immunoglobulin-like domains. Alternative RNA splicing produces several different splice variants of the four high-affinity FGF receptors (10), and the isoforms differing in their extracellular domain are associated with altered ligand-binding properties (11). Binding of FGFs leads to dimerization of receptor monomers, to activation of intracellular kinase domains, and to receptor trans-phosphorylation at multiple tyrosine sites. A number of SH2 domain–containing signal-transduction molecules can bind to these phosphotyrosine sites, potentially activating several signaling pathways, including the Ras/mitogen-activated protein kinase (MAPK) and the PLCγ (8). Depending on the activated receptor subtype and signaling pathway, the FGFs can produce different cellular responses (8,11,12).

Among high-affinity FGFRs, FGFR-1 binds FGF-2 with the highest affinity (13). The low-affinity receptor seems to be critical for FGF signaling, because enzymatic removal of heparin/heparan sulfate from the cell surface or inhibition of sulfation abolishes the biologic activity of FGFs (14–16). It has been proposed that heparan sulfate may facilitate dimerization and activation of receptors (17).

In situ hybridization studies, in good agreement with immunohistochemistry, provide evidence that FGF-2 mRNA and protein are present in the astrocytes of all brain regions and in selected neuronal populations [for example, in the pyramidal neurons of the CA2 hippocampal region and in neurons in the retrosplenial cortex (18,19)].


FGF-2 is a pleiotropic factor that exerts its effects in different cells and organs as a neurotrophic and as a chemotactic factor (20). As a neurotrophic factor, as detailed later, FGF-2 can regulate proliferation, survival, and differentiation of nerve cells (20,21). It is a potent mitogen, capable of inducing proliferation of a wide variety of cell types derived from the mesoderm (22) and from the neuroectoderm (23). Furthermore, it enhances axonal branching (24–27) and synaptogenesis (28) in neurons. FGF-2 also is a potent chemotactic factor for fibroblast and endothelial cells and a potent angiogenic molecule: therefore it can stimulate wound healing and tissue repair (20).

Extensive evidence supports the notion that FGF-2 acts as a neurotrophic factor in the developing and in the adult nervous system, where it is involved in the regulation of plasticity and in the maintenance of structural integrity. FGF-2 expression is developmentally regulated in the brain: in the rat brain, FGF-2 mRNA and protein are detectable by embryonic day 16 (29). In the human fetal brain, it is highly expressed in neurons and, later in development, in glia cells (30). A differential expression of all FGF receptors has been observed in neuronal cultures and in pure astrocytic and microglial cultures during development (31). The FGF-FGFR system is thought to play a critical role in cell–cell signaling between neurons, astrocytes, and microglia during development. FGF-2 mRNA levels increase progressively from postnatal day 1 to postnatal day 21, remaining high in the adult and aged rat brain (32). In adult rats, FGF-2 mRNA is distributed throughout the brain, the highest levels being observed in the cerebral cortex, hippocampus, and spinal cord (32).


In vitro, exogenously administered FGF-2 exerts mitogenic and differentiative effects on E16 calbindin-expressing cultured hippocampal neurons (33), stimulates proliferation of cortical precursor cells (23), and acts sequentially with other growth factors in regulating the generation of neurons and astrocytes from committed neuronal progenitor cells (34). Different concentrations of FGF-2 may influence the fate of early cortical stem cells: low FGF-2 levels lead predominantly to development into neurons, but high levels generate glia as well as neurons (35). FGF-2 also increases survival and differentiation of cultured neonatal dentate granule cells (36).

FGF-2 appears to produce similar neurotrophic effects in vivo. Neurogenesis is inhibited by injection of an antibody against FGF-2 in the P1 rat brain (37). In line with these data, a single subcutaneous FGF-2 injection in P1 rats elicits a significant increase in hippocampal cell number (38). The effects of exogenous FGF-2 persist in the adult brain in the regions (subgranular zone of the dentate gyrus and subventricular zone) where neurogenesis persists during adulthood, eliciting an increase in mitotic nuclei (39).

Taken together, these observations support the notion that FGF-2 mitogenic and differentiating effects vary depending on the specific developmental stage, on the cell type, and on the in vivo environment. It seems likely that different FGF-2 isoforms and receptor subtypes, as well as site- and time-specific synergies with other neurotrophic factors, account for these complex effects.


FGF-2 exerts neuroprotective effects against a wide variety of insults, reducing brain cellular damage and improving functional recovery in experimental models of stroke, epilepsy, and traumatic brain and spinal cord injury. In rats subjected to focal ischemia, intravenous treatment with FGF-2 produces a significant reduction in cortical infarct volume and a marked improvement in rotarod performance (40). Prolonged i.c.v. administration of FGF-2 protects against seizure-induced cell loss and reduces long-term behavioral deficits (41). Moreover, FGF-2 increases long-term survival of spinal motor neurons and improves respiratory function after experimental spinal cord injury (42).


FGF-2 accelerates bifurcation and growth of axonal branches without affecting the elongation rate of primary axons in cultured rat hippocampal neurons (24). Other studies have documented FGF-2–induced enhancement of neurite growth in hippocampal and cortical neurons (25,26,36). A greater number of axon branches is expected to produce a greater number of synaptic contacts, and indeed, local application of FGF-2 has been found to concentration-dependently increase the number of excitatory synapses between hippocampal neurons; these synapses are morphologically mature and functionally active (28).


The adult brain undergoes activity-dependent plastic modifications during pathologic processes, reminiscent of those observed during development. For example, recurring seizures may cause molecular, cellular, or network changes leading to long-term alterations in neural circuits, which, in turn, may contribute to progressive exacerbation of the disease and to behavioral and cognitive decline (43). The acute and delayed cellular alterations induced by seizures include neuronal loss, neurogenesis, axonal and dentritic sprouting, gliosis, and circuit remodeling (44). Interestingly, neurotrophic factors (and FGF-2 in particular) can be hypothesized to be involved in these phenomena, because they are well-known physiologic mediators in each of these cellular events. Consistent with this idea, the expression of many neurotrophic factors has been reported to be altered in diverse, broadly distributed neuronal populations during epileptogenesis and after seizures.

Together with other neurotrophic factors, FGF-2 seems a good candidate for implication in the plastic changes associated with epilepsy because, as described earlier, it regulates neurogenesis, cell survival, axonal sprouting, and synaptogenesis. Furthermore, as detailed later, (a) seizures increase FGF-2 mRNA and protein levels in specific brain areas and upregulate the expression of FGFR-1 receptors; (b) acute intrahippocampal injection of FGF-2 causes seizures, whereas long-term i.c.v. infusion of low-dose FGF-2 does not affect kainate seizures but promotes behavioral recovery and reduces hippocampal damage; and (c) kainate seizure severity is not altered in FGF-2 knockout mice, but is increased in FGF-2–overexpressing mice.


FGF-2 gene expression is induced with similar patterns in different acute seizure models: increased FGF-2 mRNA levels have been observed after application of bicuculline (45), lesion of the dentate gyrus hilus (46), kainate (47–49), and electroconvulsive shock (50,51). This phenomenon is fast, marked, and transient, peaking at 6–24 h in different hippocampal subfields and in the cortex. Increased levels are observed in different regions (including neocortex, amygdale, and septum), but most greatly in the hippocampus. In the hippocampus of naive rats, FGF-2 is expressed diffusely in astrocytes and in CA2 pyramidal neurons (52–54). After kainate or hilar lesion seizures, increased FGF-2 mRNA levels in these cell populations, as well as new expression in CA1 pyramidal neurons and in dentate gyrus granule cells, have been observed (48,55).

In the kindling model, induction of FGF-2 mRNA expression is observed in the hippocampus, neocortex, and hypothalamus, in a more pronounced manner after a single afterdischarge, not accompanied by behavioral seizures, than after a fully kindled, generalized tonic–clonic seizure lasting more than a minute (56,57). Such apparently paradoxical findings may result from adaptive mechanisms taking place during the kindling process; for example, it is possible that different neuronal circuitries with opposite effects on FGF-2 expression might be activated with kindling development. A higher responsiveness to low-effective compared with maximally effective stimulations also was observed in the hippocampus after minimal as compared with maximal electroshock (50). These observations suggest that the duration and intensity of seizures within a specific area does not necessarily correlate with the magnitude of FGF-2 mRNA level increase, and that FGF-2 may be more directly implicated with epileptogenesis than with generalized seizure expression.

The time course of induction of FGF-2 protein correlates with the one of the mRNA. In the kainate model, FGF-2–like immunoreactivity is detectable 6 h after injection of the convulsant agent and peaks at 24 h, being localized mainly in the nuclei of astrocytes (49,58). After electroconvulsive shock, FGF-2 protein expression increases in the frontal and rhinal cortex, peaking at 20 h, and being still elevated at 72 h (59). These events are observed predominantly for the high-molecular-weight isoforms of FGF-2. These isoforms contain nuclear targeting sequences, may access the nucleus through nuclear pores, and can influence gene regulation, activating programs for cellular plasticity or proliferation (59). In another study, FGF-2 protein levels were found to remain elevated in astrocytes up to 30 days after kainate-induced seizures (60). Therefore it has been suggested that the transient pattern of FGF-2 mRNA elevation has prolonged translational effect that may be capable of influencing long-term plasticity changes.

Seizure-induced changes in the expression of only two (FGFR-1 and FGFR-3) of the four high-affinity tyrosine kinase FGF receptors have been studied in the kainate model. Seizure activity modulates FGFR-1 mRNA expression: in situ hybridization reveals increased FGFR-1 mRNA levels in the cell body of granular and pyramidal neurons, ≤24 h after kainate injection (47,49). FGFR-1 protein immunoreactivity is normally localized in astrocytes (61) and neurons (49). After kainate administration, FGFR-1–positive cells appear first (at 3 h) in the molecular layer of the dentate gyrus, then in other hippocampal subfields, and become very diffuse throughout the hippocampus and the cerebral cortex by 24 h (61). Thirty days after kainate injection, a strong FGFR-3 staining has been found in OX-42–positive reactive microglia in several brain areas, including hippocampus, amygdala, and piriform and entorhinal cortex (60).

In spite of some minor discrepancies between different models of epilepsy, these observations lead to the notion that epileptogenic seizures coordinately increase the expression of FGF-2 and of its receptor(s). These events are likely to occur as part of an adaptive response to excessive activation of specific neuronal circuitries and, as stated earlier on the basis on the well-known biologic effects of FGF-2, may take part in the plastic changes associated with epilepsy. This hypothesis has been pharmacologically and genetically investigated.


Unilateral injection of FGF-2 into the dentate region of the ventral hippocampus causes an immediate excitatory effect culminating in EEG and behavioral seizures (62). This acute effect of FGF-2 seems to be dose dependent: at high doses (50 ng, as compared with 25 ng); both the percentage of seizing animals and the duration of seizures are increased. These acute electrical alterations of hippocampal activity may result from a fast signal-transduction mechanism leading to changes in the synaptic network (62).

In contrast with the effects of acute administration, the prolonged infusion of low FGF-2 doses (2.5 ng/h) into the cerebral ventricle does not modify latency and duration of kainate seizures; however, it effectively prevents seizure-induced hippocampal cell loss and improves long-term behavioral recovery (41,63). The neuroprotective action of FGF-2 may depend on inactivation of N-methyl-d-aspartate (NMDA) receptors. Long-term (hours), but not short-term, treatment with FGF-2 has been reported to potentiate Ca2+-dependent inactivation of NMDA currents in hippocampal neurons (64). Another possible mechanism of this effect may be the induction of activin A (ActA), a cytokine belonging to the transforming growth factor-β superfamily. When coinjected with kainate in the hippocampus, FGF-2 prevents the loss of CA3 neurons in mice. In mice treated with kainate and FGF-2, but not in those treated with kainate alone, ActA-immunoreactivity was high in CA3, as well as in CA1 and CA2, neurons (65). Moreover, FGF-2 failed to protect CA3 neurons against kainate-induced death in presence of the ActA-neutralizing protein follistatin (65).


The generation of mice with homozygous deletion of the FGF-2 gene (FGF-2 knockout mice) and with a species-specific transgene inserted in the genome (human FGF-2 transgenic mice) have provided new tools for study of the molecular, physiologic, and pathologic roles of FGF-2 in vivo. The FGF-2 knockout mice (66), lacking all three murine FGF-2 isoforms, are viable, fertile, and without any obvious phenotypic difference from their wild-type littermates. However, these mice exhibit a significant reduction in the number of neurons in the neocortex, in particular in the motor area. Furthermore, they display a delayed wound healing. The susceptibility to seizures of FGF-2 knockout mice has been studied in the kainate model (67). Reportedly, the severity of kainate seizures did not differ between knockout and wild-type mice. Furthermore, the neurogenesis in the dentate gyrus was not significantly different between FGF-2 knockout and wild-type mice under resting conditions. In response to kainate seizures, however, whereas an increase in FGF-2 protein levels and neuroproliferation was observed in the hippocampus of wild-type mice, very low levels of neurogenesis were observed in the knockouts. By using a viral gene delivery system, high FGF-2 levels can be restored in the mutant mice, also restoring levels of neurogenesis comparable to the one of wild-type littermates. These data suggest that seizure-induced FGF-2 overexpression is necessary and sufficient to prime proliferation and differentiation of neuroprogenitor cells in the adult hippocampus (67).

In an attempt to elucidate further the effect of FGF-2 in epilepsy, we studied transgenic mice expressing the human FGF-2 (68,69). These mice have been previously characterized from a molecular and from other phenotypic points of view (68,69). Western blot analysis has shown that the murine and human FGF-2 isoforms are tissue-specifically expressed and translationally regulated. The human FGF-2 transgene is highly expressed in the brain. By gross examination, these mice are affected by skeletal malformations, such as shortening and flattening of long bones and moderate macrocephaly (68). In addition, without having any spontaneous or inherent vascular defect, they exhibit a predisposition to angiogenic reactions with subsequent amplified angiogenesis (69). Mice overexpressing FGF-2 have a latent epileptic phenotype, in that they are more susceptible to kainate seizures than are their wild-type littermates (Zucchini et al., unpublished observations). At a cellular level, the overexpression of FGF-2 does not lead to increased neurogenesis: under basal conditions, BrdU immunoreactivity, an index of neuroregeneration, does not differ between transgenic mice and their wild-type littermates, and both groups display equally increased cellular proliferation in response to kainate administration. Neuroprotection was explored by using the Fluoro-Jade B labelling, a marker of cell death (70): after kainate seizures of similar severity, we observed greater cell damage in wild-type than in FGF-2 transgenic mice. In wild-type animals, Fluoro-Jade B labelling displayed the typical pattern of neurodegeneration associated with kainate seizures, involving different hippocampal subfields (CA1, CA3, CA4) and extrahippocampal regions like thalamus and cortex. In contrast, in transgenic mice that underwent severe kainate seizures, Fluoro-Jade B–positive neurons were observed only in the hilar dentate gyrus regions (Fig. 1). These data are in good agreement with the neuroprotective action of FGF-2 in injury models. It can be hypothesized (see earlier) that seizures prompt the overproduction of FGF-2 in astrocytes and neurons, and, in turn, this newly produced FGF-2 enhances the production of ActA in selected neurons. In the hippocampus, ActA may reach high levels in CA1 and CA3 neurons, protecting these cell types from injury (71). Accordingly, in FGF-2 transgenic mice, we observed the preservation from degeneration of CA1 and CA3 neurons, but not of those in the hilus of the dentate gyrus. Further experiments are in progress to challenge this hypothesis.

Figure 1.

After kainate seizures of identical severity [scored as previously described (73)], cell damage is greater in NTg than in Tg mice. Shown are representative dorsal hippocampal, Fluoro-Jade B–stained sections prepared from a wild-type (nontransgenic, NTg), and from an FGF-2 transgenic (Tg) mouse. Numerous degenerating, Fluoro-Jade B–positive cells can be observed in the hilus, CA3 and CA1 pyramidal layer of the NTg mouse, and only in hilus on the Tg mouse.

It is well known that epileptic seizures can cause long-term synaptic rearrangements, like the sprouting of mossy fiber axons. Because these morphologic alterations are likely dependent on neurotrophic factors, and on FGF-2 in particular (72), we investigated the morphoregulatory effects of FGF-2 in transgenic mice, under basal conditions and after kainate seizures. Under control condition, neither control nor FGF-2–overexpressing mice display sprouting of the mossy fibers, as detected by using Timm staining. Thirty days after kainate treatment, sprouting was observed in similar grade both in transgenic and in control mice. Thus the overexpression of human FGF-2 does not seem to be sufficient to influence seizure-induced morphologic reorganization.

Therefore, the latent epileptic phenotype of FGF-2–overexpressing mice cannot be explained on the basis of effects on neurogenesis (not altered in the mutant mice), cell death (not altered under basal conditions, and actually reduced after kainate seizures), or sprouting of the mossy fibers (not altered). Another hypothesis could be that this phenotype depends on the FGF-2 ability to increase the number of excitatory synapses (28). To explore this possibility, we investigated the effect of FGF-2 overexpression on synaptic transmission, comparing the basal properties of the Schaffer collateral–CA1 synapse in hippocampal slices prepared from transgenic and wild-type mice. Extracellular field potential recordings in the dendritic region of area CA1 have been performed to generate input/output (I/O) curves for each mouse strain: these curves display the field excitatory postsynaptic potential slope as a function of the stimulation intensity. We found that the I/O curve was parallel shifted leftward in FGF-2 transgenic as compared with wild-type mice, indicating increased excitability in the former. Furthermore, we consistently recorded afterdepolarizations at the highest stimulation intensities in the transgenic, but not in the wild-type, group. Together, these data suggest that an anomalously increased excitatory synaptic input may cause the increased propensity to kainate seizures in FGF-2–overexpressing mice.


Taken together, the data presented suggest the following: (a) seizures (most notably epileptogenic seizures) induce the synthesis of FGF-2 in astrocytes and neurons constitutively expressing its gene and, ectopically, in neurons that do not normally produce this neurotrophic factor; and (b) pharmacologic and genetic evidence suggests that increased availability of FGF-2 increases seizure susceptibility (i.e., favors epileptogenesis), but reduces seizure-induced cell death. Dissecting out the mechanisms of these apparently contrasting effects of FGF-2 will be critical in pursuing the goal of controlling seizures and their consequences through modulation of the FGF-2 system.


Acknowledgment:  We thank Drs. Douglas Coffin, Donata Rodi, Anna Binaschi, Beatrice Paradiso, Andrea Buzzi, and Katja Perrone for their contribution to the experiments performed in their laboratory.