Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Immediate Early Gene Expression in Experimental Epilepsy

Immediate Early Gene Expression in Experimental Epilepsy Institute of Neuropathology, University of Heidelberg, D-69120 Heidelberg, Germany Neuronal excitation b y experimentally induced seizures elicits the rapid induction of a set of genes called immediate early genes (IEGs). The gene products of fos, jun and Krox, multimember gene families that belong t o the class of IEGs, participate in a fundamental biological control mechanism, the regulation of gene transcription. IEG encoded proteins act as third messengers in an intracellular signal transduction cascade between neural cell surface receptors, cytoplasmic second messenger systems and specific target genes i n the nucleus, a process for which the term 'stimulus transcription coupling' has been given. Almost all types of seizures cause dynamic alterations of IEG expression i n neurons of the limbic system, but also in non-limbic areas, such as the cortex, striatum and thalamus. IEG encoded transcription factors are thought t o up- or down-regulate effector genes with preferential expression in the central nervous system, including genes for neurotransmitt ers, growth factors, receptors, synaptic and axonal proteins. If the concept holds true that IEGs act as molecular switches converting epileptic short-term excitation of neurons into alterations of the molecular phenotype, future research may help t o explain hitherto unexplained phenomena in epileptogenesis including changes of synaptic efficacy, kindling and sprouting. Introduction Excitation of neurons in vitro and in vivo principally results in two types of adaptive responses (7,41). First, a rapid, short-lasting biochemical modification of pre-existing target molecules such as receptors, Corresponding author: Or. M . Kiessling, Institute of Neuropathology, University of Heidelberg, 0-691 20 Heidelberg, Germany Tel. +49 (6221) 562 603; Fax +49 (6221) 563 466 Ca 2+ translocation systems and neurotransmitter trafficking. The second type of response comprises a complex series of delayed but long-lasting cellular events, i.e., changes of the neuronal genetic program that ultimately requires de novo protein synthesis. According to current concepts, this genomic response is a prerequisite for long-term changes of neuronal plasticity under physiological and pathological conditions. The first set of genes activated by neuronal stimulation have been termed 'immediate early' or 'early response genes' (8,50,102). Characteristically, (i) these genes can be expressed under conditions of protein 'synthesis inhibition, ruling out that other newly synthesized gene products are responsible for their induction; (ii) IEG encoded mRNA and protein rapidly accumulate after stimulation, but also have a short half-life due to a limited stability of their mRNAs and a rapid turn-over of their proteins. Consequently, IEGs are thought to act as third messengers in an intracellular signal transduction cascade between cell surface receptors (first messengers), cytoplasmic second messenger systems and the nucleus, a process for which the term 'stimulus transcription coupling' has been chosen (73,74). Many immediate early genes (IEGs) participate in a fundamental biological control mechanism, the regulation of gene expression. *The genes fos, jun and KTOX belong to families of IEGs that code for such transcription factors (8,SO). Fos and jun were originally discovered as proto-oncogenes, i.e., normal cellular genes that share large areas of sequence identity with retroviral transforming oncogenes. They are thought to have important functions in cell cycle control as well as growth and differentiation. However, investigations of central nervous system (CNS) neurons have unequivocally demonstrated that IEGs also play a role in postmitotic cells with a defined molecular phenotype. In the CNS, distinct and specific patterns of IEG induction are observed under physiological conditions such as sensory (53), visual (90,119) and audi~~ *Common notation for IEGs exemplified for c-fos: c-fos refers to the gene; c-fos to the c-fos encoded mRNA; c-FOS to the c-fos encoded protein. M. Kiessling and P. Gass: IEG expression in epilepsy Table 1 Regional analysis of immediate early gene (IEG) induction after seizures Seizure Model PTZ in situ Hybridization Immunacytochemistry References Morgan et al. (72) Dragunow & Robert (23) Saffen et al. (95) c-fos C-FOS C-FOS PTZ PTZ c-fos. Krox-24, c-jun, jun B c-fos, c-jun c-fos c-fos c-fos, Krox-24, c-jun, jun B PTZ C-FOS, FRA1, FRA2 C-FOS, FRAl, FRA2 Sonnenberg et al. (107) Sonnenberget al. (108) Daval et al. (18) Cole et al. (16 ) PTZ, Kainate, Electroshock Electroshock Electroshock Electroshock C-FOS, FOS B, KROX-24 C-JUN,JUN B, JUN D C-FOS, FOS 6 , KROX-20, KROX-24 C-JUN,JUN B, JUN D C-FOS C-FOS C-FOS, FOS 8, KROX-20, KROX-24 C-JUN, JUN B, JUN D C-FOS. c-fos c-fos, c-jun, Krox-24 c-fos Dragunow et al. (27) Bicuculline Gass et al. (36) Kainic acid Kainic acid Kainic acid Le Gal La Salle 165) Popovici et al. (84) Gass et al. (38) Kindling Kindling Kindling Kindling . Dragunow & Robert (24) Shin et al. (103) Simonato et al. (104) Clark et al. (13) tory stimulation (29,32), nociception (48) or circadian light shifts (60,61,91). An even more pronounced induction of IEGs is found after pathophysiological stimuli such as epiIepsy (Table l), ischemia (1,39,58,82,112,113,116,117) and brain trauma (22,25). FOS, JUN and the Leucine Zipper Motif The genes fos and jun belong to multigene families and several members have been identified for each, c-fos (11 ) the fos-related antigens pa-1 (14), pa-2 4, (81) and fos B (121) as well as c-jun (94), jun B (93), and jun D (52,92). Characteristically, FOS and JUN proteins contain an evolutionary highly conserved structural domain called the 'leucine zipper' (64). This domain consists of four or five leucines orderly arranged at seven-residue intervals (so-called heptad repeat) and lining up in a column of a regulatory protein segment most likely structured as an alpha helix. These recurring leucine residues 'zip' two FOS/JUN or JUN/JUN proteins together forming a dimeric complex called activator protein 1 (AP-1) (Fig. 1) (62). In close vicinity to the leucine zipper domain is a basic (arginine and lysine rich) domain that is responsible for DNA binding to the AP-1 consensus sequence; a common transcription regulatory element in the promotor region of target or 'late response genes' regulated by IEGs (Fig. 1) (40,62,85). Dimerization via the leucine zipper is mandatory for correct positioning of the two basic DNA binding domains which are both necessary for binding to the AP-1 sequence (40,62,85). Whereas FOS proteins can only form heterodimers with members of the JUN family (15,79), all JUN proteins are able to form homodimeric complexes as well as heterodimeric complexes with FOS proteins or members of other leucine zipper families, such as CAMP/ Can+-response element binding protein 1 (CRE-BPI) (67) and MyoD (6). Dimer formation increases the variety of protein combinations as well as the number of usable DNA motifs. This combinatorial multiplicity is of functional relevance, since different AP-1 complexes have different molecular properties including DNA binding affinity and half-life, as well as up- or down-regulatory effects on target genes. For example, while c-FOS/c-JUN dimers are efficient activators of promotors containing AP-1 sites, c-FOS/JUN B complexes seem to M . Kiessling and F? Gass: IEG expression in epilepsy repress them under certain circumstances (11,98). Similar features have' been discovered by investigating autoregulatory effects of FOS / JUN expression: transcription of the c-fos gene is repressed by its own gene product if heterodimeric binding of c-FOS/c-JUN takes place at the c-fos promotor (63,96). In contrast, c-jun transcription is directly stimulated by its own gene product via a positive autoregulatory loop at the c-jun promotor (2). KROX and the Zinc Finger Motif The combinatory principle can also operate within a single transcription factor by variation of the choice, order and number of independent DNA binding modules within a protein. This applies to another family of transcription factors called 'zinc finger proteins'. This term was given for proteins with so-called 'zinc finger sequences', a motif that is currently considered to be the most frequent DNA binding domain (59). It consists of protrusions formed by a pair of cysteines and a pair of histidines grouped around a central zinc ion and separated by a (finger-like) loop of amino acids, folding into a specific DNA-binding minidomain (Fig. 2) (30). The zinc finger proteins that have been identified so far carry from as few as 2 to as many as 37 zinc fingers, which are each able to connect to specific binding sites located in a groove of the DNA helix. It has been estimated that as much as 1%of the DNA in human cells codes for proteins with zinc finger motifs. In the CNS, Krox-20 (10) [also termed EGR-2 ( 5 6 ) ] and Krox-24 (66) [also termed Zif1268 (12), EGR-1 (110), or NGFI-A (71)], are well-characterized members of the zinc finger family. Both molecules contain three zinc fingers that curl around almost one turn of the DNA helix. The three zinc fingers make contacts with three successive 3-base pair sites on the DNA, called early growth response (EGR) consensus sequence. Apart from immediate early genes another class of zinc finger proteins exists in the CNS: zinc finger sequences are also found as structural units of nuclear hormone receptors and constitute, among others, the DNA binding domains of the glucocorticoid and estrogen receptor (99). Here, however, the zinc binding amino acids are exclusively cysteines instead of cysteines and histidines. Basal Versus Constitutive Immediate Early Gene (IEG) Expression Although fos, jun and Klox were originally defined Figure 1 The leucine zipper motif. The gene products of fos (c-FOS, FOS B, FRA-1 or FRA-2) and jun (c-JUN, JUN B or JlJN D) associate with each other via the leucine zipper domain to form a dimeric protein complex. 'Zippering' occurs by an interaction of five leucine amino acids (L) located in a protein alpha helix at seven amino acid intervals. This 'heptad repeat' causes the leucines to be aligned in a plane along the length of the helix. JUN proteins can form homo- and heterodimers, but FOS homodimers do not occur due to structural constraints. Dimerization is a prerequisite for DNA-binding of basic (arginine/lysine rich) domains to their target sequence, the so-called activator protein 1 (AP-1) binding site (TGACTCA), a common transcription regulatory element in the promotor region of target genes (double helix). as members of rapidly and transiently inducible gene families, they show to a varying degree a basal expression in the CNS of normal animals, suggesting that they also play a role in physiological transcriptional control in the brain (21,26,70). In accordance with their function as transcription factors, immunoreactivity is almost exclusively found in the nucleus upon investigation by immunocytochemistry in sitri (21,26). Immunoelectron-microscopic studies have revealed that c-FOS immunoreactivity is associated with euchromatin supporting the putative action of c-FOS as regulatory element at the transcriptional level (77). In general, basal expression of FOS proteins is restricted to a few scattered neurons, preferentially located in the hippocampus, thalamus and the outer layers of the cerebral cortex (21,26). This expression is possibly due to excitation of specific neuronal circuits processing physiological input. Similar to FOS proteins, c-JUN and JUN B are almost completely absent under unstimulated conditions in the large majority of neurons. However, selective neuronal subpopulations express remarkably high levels of FOS and JUN suggesting a function in the processing of sensory input. Such neuroanatomical areas include the inferior colliculus in acoustic perception (29,32) or the visual cortex (90,119) following photic stimulation. In analogy, disruption of the normal light dark cycle evokes a prominent c-FOS and JUN B expression in the suprachiasmatic nucleus, a central pacemaker which generates an endogenous almost 24-hour M. Kiessling and P. Gass: IEG expression in epilepsy petitive NMDA receptor antagonist MK-801 (37). Despite considerable dispute in the scientific literature about unspecific IEG induction due to handling and stressful manipulations of untreated animals these findings rather suggest a specific role of at least certain IEG encoded proteins in the normal transcriptional control in the CNS. The cerebral pattern of IEG expression, however, changes dramatically after pathological excitation such as seizure activity. IEG Induction After Generalized Tonic-Clonic Seizures Figure 2 The zinc finger motif. Three tandem zinc fingers of KROX-24 protein (also termed Zif-268. EGR-1, NGFI-A) are complexed to their specific DNA-binding sites (EGR-consensus sequence) in the promotor region of a target gene (double helix). The protein as a whole spirals around in the major groove of the DNA helix, each finger making approximately equivalent contacts. The first half of each zinc-finger minidomain forms a p-sheet (bold line), the second half twists into an alpha-helix (tube). Binding of zinc by cysteines in the p-sheet and histidines in the alpha-helical regions draws the halves together near the base of the finger (see insert). The three zinc finger units are linked by short amino acid segments without characteristic three dimensional structure (thin lines). Insert: schematic view of a single zinc finger module, formed by a pair of cysteines and a pair of histidines grouped around a central zinc ion and separated by a (finger-like) loop of 12 amino acids. Modified according to reference 83. periodicity of metabolic and behavioural rhythms (60,61,91). In contrast to those IEGs with a very restricted basal expression, other members of the leucine zipper and zinc finger families demonstrate high expression levels even in the absence of intentional stimulation. Thus, JUN D exhibits high constitutive expression throughout the brain, most likely due to an endogenous permanent activation of the jzin D promotor (19). This high constitutive expression has prompted some authors to postulate that jun D acts as a 'housekeeping gene' controlling the expression of target genes involved in the maintenance of basic cellular metabolic processes (8,9). Similar to JUN D, KROX-24 shows distinct basal expression in the neocortex, preferentially in layers 11 to IV and VI, in the striatum as well as in the limbic system and limbic system associated structures (49,68). In contrast to JUN D, high KROX-24 expression is thought to be maintained by N-methyl-D-aspartate (Nh4DA) receptor mediated physiologic activity. It can be completely blocked by application of the non-com- Epileptic tonic-clonic seizure activity induced by pentylenetetrazole (PTZ) was the first experimental model demonstrating the inducibility of c-fus in the mammalian brain in vivo (72). Subsequently, a large number of experiments have further characterized the temporal and spatial expression pattern of c-fos but also of other IEGs in various seizure models (Table 1). Systemic administration of PTZ or bicuculline to rodents results in generalized tonic-clonic seizures comparable to human 'grand mall epilepsy (23,36,38,72,95,107,108). IEG expression has been investigated both on mRNA level by in sifti hybridization and Northern blot analysis, and on protein level by immunocytochemistry and Western blot analysis (Table 1). These studies have demonstrated a rapid and transient neuronal IEG induction with peak levels in the hippocampus and other parts of the limbic system, but expression also occurred in the somatosensory cortex, striatum and thalamus. Due to, the short half-life of IEG encoded mRNAs, immunocytochemical investigations have revealed different spatiotemporal expression profiles for individual IEGs more precisely than in situ hybridization studies (27,36,107,108). According to these studies, three major categories of IEG encoded transcription factors can best be defined by their staggered induction in the hippocampus: (i) c-FOS and KROX-24 showing a concurrent rapid rise with peak levels at two hours of postictal recovery and a decline to control levels within eight hours after seizure termination; (ii) c JUN and JUN B demonstrating a more gradual increase with peak intensities at four hours and return to baseline levels after 24 hours of postictal recovery; (iii) FOS B, JUN D and KROX-20 exhibiting a delayed induction and prolonged persistance up to 24 hours after seizure termination. All IEG encoded proteins with the exception of KROX-20 demonstrate the same sequence of induction in different hippocampal subfields, following the order dentate gyrus, CA1 and CA3, respectively. Interestingly, this order inversely correlates with the degree of neuronal vulnerability among hippocampal subpopulations after human generalized seizure disorders. However, it can not be ruled out that the sequence of IEG induction within the hippocampus M. Kiessling and P. Gass: IEG expression in epilepsy and among other neuronal populations is related t o the propagation of epileptic discharges or is due t o different molecular genetic properties of hippocampal neurons. KROX-20 is the only transcription factor investigated not induced in the hippocampus by generalized seizures (47). Although all FOS and JUN encoded proteins investigated are induced in the limbic system and somatosensory cortex by generalized seizures, c-JUN is n o t expressed in striatal neurons (36). Recent data have associated c-JUN induction with the capacity of regeneration after transection experiments (45,46). Therefore the expression of c-JUN may identify neuroanatomical areas with the potential for postictal regenerative synaptogenesis and sprouting. IEG Induction After Limbic Seizures but also some scattered pyramidal cells in the hippocampus. At no timepoint is IEG induction observed in the nuclei of degenerating neurons, corroborating findings in a model of global ischemia that expression of IEG encoded proteins identifies neurons at risk but which are destined t o survive (58). Kindling as a Model for Epileptogenesis Partial epilepsy of temporolimbic origin is the most common and often intractable type of human seizure disorders. Administration of the glutamate agonist kainic acid (KA) t o rats has been established as a standardized model for human temporal lobe epilepsy, mimicking clinical, electrophysiological and neuropathological features with high accuracy (5). KA-induced limbic seizures cause a highly specific regional pattern of IEG expression that is strikingly similar to that after generalized seizures of the PTZ/ bicuculline type. However, markedly prolonged timecourses of IEG expression are observed following KA-induced seizures, most likely reflecting prolonged seizure activity (Fig. 3) (38,65,84). Again, FOS B, JUN D and KROX-20 exhibit delayed induction and prolonged expression compared to c-FOS, c-JUN and JUN B (38). Despite this asynchronous activation of individual IEGs, the sequence of induction in the hippocampus (dentate gyrus > CA1> CA3) is identical to that after generalized seizures (Fig. 3). The further sequence of gene activation, progressing from hippocampus to limbic system associated areas and subsequently t o non-limbic structures, correlates well with the generation and propagation of paroxysmal activity produced by KA (38,65,84). Also, in analogy t o generalized seizures of the PTZ/ bicuculline type, the expression profile of c-FOS and KROX-24 is very similar within limbic system structures (36,38). In non-limbic areas such as somatosensory cortex and striatum, however, KROX-24 levels increase after the onset of seizure activity, but 24 hours later they drop markedly below baseline levels. Since high basal levels of KROX-24 are thought to be regulated by NMDA receptor mediated physiological neuronal activity (37,119), the marked reduction of KROX-24 i n the cerebral cortex at later time intervals probably reflects postictal depression of electrophysiological activity ( 5 ) . In contrast t o bicuculline- and PTZinduced experimental seizures, systemic KA-evoked limbic seizure activity can cause neuronal degeneration, affecting many neurons in the piriform cortex Kindling is an animal model of epileptogenesis, i n particular for human complex partial epilepsy. After an appropriate time interval of days or weeks, repeated administration of an initially subconvulsive electrical or pharmacological stimulus finally results in intense limbic and generalized motor seizures. Once established, the kindling effect is permanent. The molecular basis of kindling is unknown. Kindling depends on de novo protein synthesis and it is tempting to speculate that IEG induction could represent the link between repeated short-term stimuli, gene expression and alterations in the neuronal phenotype associated with the development of the kindled state (101). Periodic induction of afterdischarges (ADS), i.e., local electrical seizures of a few seconds without overt behavioural effects, is a prerequisite for kindling. A single AD can evoke increased expression of c-fos, c-jun and Krox-24 mRNA in the hippocampus (104). Similar t o generalized seizure activity different timecourses are found for individual transcription factors, the sequence of induction being c o x - 2 4 > c-fos > c-jun (104). There is an all or nothing relationship between induction of IEGs and the duration of ADS with a time threshold of about 30 seconds (103). The above threshold ADS induce c-fos in both naive and kindled animals to the same extent and with identical temporal profiles and the kindled state is not accompanied by long-term changes of basal IEG expression (103,104). Thus, IEG induction seems t o be involved in producing the kindled state, but once established, n o permanent up-regulation of IEG levels occurs . IEG Induction and Sprouting Synaptic reorganization of limbic system and temporal lobe pathways is a prominent feature in experimental epilepsy and in human seizure disorders (111). Thus, a long-term consequence of repetitive seizure activity is the sprouting of axons, in particular of hippocampal mossy fibers. Sprouting has been observed after experimental KA-induced seizures (33,78), kindling (86,111) and in human childhood epilepsy (87). This has prompted some authors to put forward the hypothesis that IEGs play a crucial role in the development of sprouting by initiating a chain of molecular events culminating i n morphological alterations (75). Several correlations exist between seizures, IEG induction and sprouting: (i) Induction of c-fos and sprouting evoked by repe- M. Kiessling and P. Gass: IEG expression in epilepsy titive PTZ seizures require similar dose thresholds (42,72); (ii) The most pronounced IEG induction and sprouting occur in the same CNS regions, such as dentate gyrus (Fig. 4); (iii) Sprouting requires only a few days to become apparent; and (iv) Anti- epileptic drugs that are known to block the induction of c-FOS and kindling also impair or block sprouting. Clinical implications of these data are at present speculative. If IEGs indeed play a role in the development of sprouting, then blockade of M. Kiessling and P. Gass: IEG expression in epilepsy Figure 3 Differential postictal kinetics of c-FOS and FOS B induction in the rat hippocampus after kainic acid induced limbic seizures. c-FOS and FOS 6 are virtually absent in untreated control animals KO). After kainic acid elicited limbic seizures, c-FOS and FOS 6 are markedly induced in neuronal nuclei of all hippocampal subpopulations, demonstrating a specific sequence of induction, i.e., dentate gyrus > CA1 > CA3. FOS 6 expression is clearly delayed and prolonged compared to c-FOS. Thus, early on heterodimeric AP-1 complexes are likely to be predominantely composed of c-FOS and one of the JUN proteins, whereas FOS B is probably the partner to form complexes with JUN during late postictal recovery periods. Vibratome sections: 50 pm. x 40. their expression might have a therapeutic impact on epileptogenesis. Mechanisms of IEG induction Since almost all types of seizures induce IEGs in identical brain regions and neuronal circuits one could speculate whether a single mechanism such as stimulation of a particular neurotransmitter system can account for all modes of postictal IEG induction. However, pharmacological studies with specific receptor antagonists suggest considerable diversity. In generalized seizures of the PTZI bicuculline type, glutamate receptors of the NMDA subtype apparently mediate the major component of IEG induction, since transcription factor synthesis is largely attenuated by NMDA receptor antagonists such as MK-801 (72,108). These data may indicate that by reducing the activity of GABA, receptors, KTZ and bicuculline cause a disequilibrium between inhibitory and excitatory neurotransmission, thereby increasing or potentiating NMDA receptor stimulation. In contrast, IEG induction after electrically evoked seizure discharges of the type utilized in clinical electroconvulsive treatment, can be markedly attenuated by agents directly or indirectly blocking voltage sen- sitive calcium channels (VSCC), but not by MK-801, suggesting that IEG induction in this model is mediated via VSCC (16). I n KA-evoked temporal lobe seizures, pretreatment with MK-801 similarly does not affect IEG induction in the limbic system. This finding is in line with the notion that IEG induction in these regions is mediated by high affinity glutamate receptors of the kainate I AMPA subtype rather than NMDA receptors (Fig. 6). MK-801 though abolishes IEG induction in sopatosensory cortex and striatum, suggesting that paroxysmal activity in nonlimbic neurons occurs by a transynaptic activation o€NMDA receptors, which in addition to the hippocampus are particularly enriched in these structures (Fig. 6). Alternatively, the blockade of IEG expression by MK-801 in cortex and striatum but not in the limbic system may reflect a different regional susceptibility of different brain regions to KA-induced seizures. How is receptor stimulation at the cell surface COUpled to transcriptional activation of IEGs in the nucleus. For c-fos, two important regulatory elements located in the 5' untranslated region of its promotor have been identified: (i) the serum response element (SRE) located at basepair -310, and (ii) the CAMP/ M. Kiessling and P Gass: IEG expression in epilepsy . KAINATE KAINATE +MK-801 CRE site in the c-fos promotor. Differential intracelMar Caz+ entry may thus provide a mechanism for the control of diverse genomic responses (3). Molecular Consequences of Differential IEG Induction Figure 4 Pharmacdogical modification of immediate early gene (IEG) expression after limbic seizures. At 3 hours after K provoked seizures, JUN B and c-FOS are markedly induced A in the limbic system (hippocampus, piriform cortex, amygdala), and also in non-limbic areas (neocortex, striatum and thalamus); A,C Pretreatment of animals with the non-competitive NMDA receptor antagonist MK-801 strongly attenuates IEG induction in non-limbic regions, suggesting that paroxysmal activity in these structures involves transynaptic activation of NMDA receptors; B,D Despite pretreatment with MK-801, seizure elicited IEG expression is unaltered in neurons of the limbic system indicating that IEG induction in these regions is rather mediated by high affinity glutamate receptors of the kainate subtype. (50 p m vibratome sections of the rat forebrain at the level of the anterior commissure (A,B) and a t the level of the dorsal hippocampus (C,D), x 10). The existence of related members such as the leucine zipper or the zinc finger proteins with a common biological function as transcription factors but different specificities could equip the cell with a mechanism for fine regulation of the genetic program. To elucidate the effects of the differential induction of IEGs, it is important to study the expression of fos and jzin on the protein level, because the combinatorial changes of leucine zipper proteins greatly expand their regulatory potential. Since the composition of Fos/Jun dimers is determined by the availability of the different components, the apparently stereotypic rapid induction and disappearance of c-FOS in all seizure models indicates, that FOS/ JUN heterodimers in the early postictal recovery period mainly consist of c-FOS and one of the rapidly synthesized JUN proteins (c-JUN, JUN B). In contrast, it can be assumed that due to their prolonged persistence AP-1 complexes during the late postictal recovery period are predominantly assembled of FOS B and JUN D. In vitro studies have demonstrated that AP-1 complex composed of FOS B and JUN D is the most stable dimer, thus extending the action of these transcription factors. The staggered induction of individual FOS /JUN proteins and the subsequent postictal. combinatorial changes allow a sequential and differential regulation of target gene expression. Similarly to AP-1 complexes, Krox-24 and Krox-20 demonstrate a staggered but overlapping induction and compete for the same DNA sequences, the EGR elements. Additional complexity arises from the fact, that expression of target genes may not be controlled by a single regulatory element, but that several elements (such as AP-1 or EGR) can coexist and cooperate.in a promotor region, and may allow for crosstalk between different transcription factors. IEG Regulated Target Genes Ca2+-response element (CRE) at basepair - 60 (102). The SRE can be activated by Caz+ influx through NMDA receptor gated ion channels and by growth factors like NGF, which cause a Ca2+ independent activation of c-fos. The intracytoplasmic effectors that convey these signals to the SRE are currently unknown. In contrast to Ca2+ fluxes via NMDA receptors, Ca*+ entry by VSCC induces c-fos via the CRE (3,102). VSCC mediated Ca2+ influx causes an activation of intracytoplasmic Ca2+/calmodulin protein kinase with subsequent phosphorylation of the so-called CRE binding protein, finally activating the The complex pattern of postictal cerebral IEG induction and/or suppression raises a critical question: Is IEG expression part of a cellular homeostatic response, endowing the CNS with the capacity to cope with the potentially harmful effects of excessively enhanced neuronal activity (e.g., by replenishment of neurotrophic factors, an up-regulation of inhibitory and/or a down-regulation of excitatory neurotransmitter systems) or does it alternatively contribute to the development of the epileptic state? To define the nature of cellular adaptive responses elicited by epilepsy and potentially mediated by IEGs requires the identification of specific target genes that are controlled by leucine zipper and zinc M. Kiessling and P. Gass: IEG expression in epilepsy finger proteins. Although it is generally believed that the products of 'primary response genes', like IEGencoded proteins switch on or turn off 'late response genes', the identification of target genes has proved technically difficult. In vitro studies have demonstrated that FOS/JUN complexes are capable of regulating genes with preferential expression in the nervous system, such as nerve growth factor (NGF), prodynorphin and proencephalin (44,80,109). The promoter regions of these genes have AP-1 sites, which can be activated by FOS/JUN complexes in transactivation assays in vitro (44,80,109). However, it has yet to be established whether IEG encoded proteins are the transcription factors which actually regulate these genes in vivo, as most interpretations of in vivo studies rely upon temporal or spatial correlations only. After experimental seizures, FOS / JUN induction precedes the up-regulation of genes coding for NGF (34), brain-derived neurotrophic growth factor (28, 54,120), proenkephalin (20), neuropeptide Y (4), glutamate decarboxylase (31), and the down-regulation of prodynorphin (20,57,76) and a putative kainate receptor (35). Altered hippocampal gene expression in human temporal lobe epilepsy also comprises an up-regulation of genes encoding acetylcholinesterase (43), somatostatin (89), as well as an up- and down-regulation of various excitatory and inhibitory amino acid receptors (69). These changes of gene expression, leading to altered neuromodulator, neurotransmitter and neurotrophin levels, altered receptor sensitivity as well as presumably altered regulatory and structural proteins of the axon and synapse, are likely to be key events in the molecular program underlying synaptic plasticity after repetitive seizures, but could either counteract or contribute to the epileptic state. Compared to leucine zipper molecules, even less is known about the regulation of target genes by zinc finger proteins. A recent study provided evidence, that KROX-20 regulates the expression of HoxB2, a gene that plays a crucial role in hindbrain segmentation during ontogenetic development (loo), but no relevant data are presently available on physiological or pathological neuronal excitation. IEG Induction in Transgenic Animals As an alternative approach to mapping IEG expression by in situ hybridization and immunocytochemistry, a fos-ZacZ fusion gene (that confers inducible P-galactosidase expression via a fused c-fos promotor) has been introduced into the germ line of mice by microinjection (105). In these mice strains, enzymatic detection of P-galactosidase activity identifies cell populations with basal or induced c-fos expression. One out of three fos-ZacZ strains originally obtained largely recapitulated the kinetics of c-fos expression in unstimulated animals and the induction profile during early postictal time inter- vals, although the longer half-life of the Fos-lacZ fusion protein compared to c-FOS slightly extends the period of expression (97). At late time intervals (four to ten days) after seizure induction, however, focal cytoplasmic p-galactosidase staining was noted in the pyramidal cell Iayer of the hippocampus and the dorsomedial nucleus of the amygdala and interpreted as a reinduction of fos-ZacZ (106). Marked differences were also observed in the percentage of Fos-lacZ positive nuclei compared to c-FOS in certain brain regions after PTZ and kainic acid induced seizures (105). These discrepancies with previous studies were explained as a possible consequence of cross-reactivity of many c-FOS antibodies with FOSrelated antigens. The different extent of fos-ZacZ expression first noted in transgenic mice was interpreted to reflect a different regulation of IEGs in these two seizure models mediated by distinct transcriptional control mechanisms. It has to be kept in mind, however, that a major caveat of the fos-ZacZ approach is a possible mutation in the transgene or an unpredictable effect of the transgene integration on the expression of the fos-ZacZ fusion construct. This is also reflected by the heterogeneity of strains, two of which completely failed to show increase of p-galactosidase activity in any brain region after PTZ induced seizures (105). A new approach to determine specific roles of individual IEGs in vivo is to produce transgenic' animals lacking particular IEGs by homologous recombination. Recently, two laboratories independently produced mutant mice strains lacking the c-fos gene (55,115). Surprisingly, these c-fos knock out mice that indeed do not express detectable levels of c-FOS, are viable. The mutants showed reduced viability immediately after birth, but a normal life expectancy of survivors with growth retardation, osteopetrosis, defects of female transmission and lyrnphopenia with extramedullary hematopoiesis. Subtle behavioral deficits including the lack of response to external stimuli, hyperactivity and disturbances of the adjustment to the light/dark cycle may indicate some alterations in neural function, but so far no morphological abnormalities have been detected in the CNS. IEG expression (apart from c-FOS) has not been investigated in c-fos knock out mice. In contrast to transgenic mice without the c-fos gene, mutant mice strains lacking the c-jtm gene are not viable but die in utero around embryonic day 10 (51). Therefore, in contrast to c-fos, a lack of c-jun can not be compensated during development and early postnatal life. According to our knowledge, no attempts have been made to produce KROX24 knock out mice. However, these mice would be of particular interest, since KROX-24 represents the only IEG thought to be involved in the induction of long-term potentiation (LTP), currently the most compelling model for learning and memory (17,88,118). M. Kiessling and P. Gass: IEG expression in epilepsy 7. Black IB. Adler JE, Dreyfus CF, Friedman WF, LaGamma EF, Roach AH (1987) Biochemistry of information storage in the nervous system. Science 236: 1263-1268 8. Bravo R (19901 Growth factor responsive genes in fibroblasts. Cell Growth Diff 1: 305-309 9. Bravo R (1990) Growth factor inducible genes in fibroblasts. In: Growth Factors, Differentiation Factors and Cyiokines, Herschmann A (ed.), pp. 324-343, Springer: Berlin , 10.Chavrier P Zerial M, Lemaire P Almendral J. Bravo R, , Charnay P (1988) A gene encoding a protein with zinc fingers is activated during GO/G1 transition in cultured cells. EMBOJ 7: 29-35 11. Chiu R, Angel P Karin M (1989) JUN B differs in its bio, logical properties from, and is a negative regulator of, c-JUN. Cell 59: 979-986 12. Christy BA, Lau LF, Nathans D (1988) A gene activated in mouse 3T3 cells by serum growth factors encodes a protein with zinc finger sequences. froc Natl Acad Sci USA 85: 7857-7861 13.Clark M, Post RM, Weiss SRB, Cain CJ, Nakajima T (1991 ) Regional expression of c-fos rnRNA in rat brain during the evolution of arnygdala kindled seizures. Mol Brain Res 11: 5 5 6 4 14. Cohen DR, Curran T (1988) Fra-1 : A serum-inducible, cellular immediate-early gene that encodes for a fos-related antigen. M o l Cell Biol 8: 2063-2069 15.Cohen DR, Ferreira PCP, Gentz R, Franza Jr BR, Curran T (1989) The product of a fos-related gene, fra-I, binds cooperatively to the AP-1 site with jun: Transcription factor AP-1 is comprised of multiple protein complexes. Gene Develop 3: 173-184 16.Cole AJ, Abu-Shakra S, Saffen DW, Baraban JM. Worley PF (1990) Rapid rise in transcription factor mRNAs in rat brain after electroshock-induced seizures. J Neurochem 55: 1920-1927 Conclusions The picture that emerges from the numerous studies of the IEG response to seizure activity is increasingly complex. The large number of seizure-induced transcription factors, their diverse sequence-specific interactions with other transcription factors and with DNA recognition sites provide combinatorial codes for highly intricate and organized patterns of gene expression. Subtle changes in the combination of transcription factors may have profound consequences for neuronal function and phenotype in response to seizure activity. Although many questions remain unanswered, there are intriguing findings suggesting that the induction of c-fos and other immediate early genes may be associated with epileptogenesis. Further attempts must be made to identify target genes for IEG encoded homo- and heterodimeric protein complexes, in order to elucidate the causal chain of the molecular events underlying morphological and functional consequences of seizure activity. Gain of function and/or loss of function studies in transgenic animals or using antisense technology may be the most valuable tool currently available to elucidate the link between IEG mediated regulation of target genes and adaptive neuronal responses in epilepsy. Acknowledgements We apologize to those individuals whose important original contributions are not cited because of space limitations. Co-workers we would like to thank are Drs. Rodrigo Bravo and Thomas Herdegen. Special thanks are extended to Dr. Frauke Bentzien for the graphics of Figures 1 and 2. Part of this work was supported by a grant from the Deutsche Forschungsgemeinschaft (SFB 31 7). http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Brain Pathology Wiley

Immediate Early Gene Expression in Experimental Epilepsy

Brain Pathology , Volume 3 (4) – Oct 1, 1993

Loading next page...
 
/lp/wiley/immediate-early-gene-expression-in-experimental-epilepsy-WZ6ruSjHWa

References (135)

Publisher
Wiley
Copyright
Copyright © 1993 Wiley Subscription Services, Inc., A Wiley Company
ISSN
1015-6305
eISSN
1750-3639
DOI
10.1111/j.1750-3639.1993.tb00766.x
Publisher site
See Article on Publisher Site

Abstract

Institute of Neuropathology, University of Heidelberg, D-69120 Heidelberg, Germany Neuronal excitation b y experimentally induced seizures elicits the rapid induction of a set of genes called immediate early genes (IEGs). The gene products of fos, jun and Krox, multimember gene families that belong t o the class of IEGs, participate in a fundamental biological control mechanism, the regulation of gene transcription. IEG encoded proteins act as third messengers in an intracellular signal transduction cascade between neural cell surface receptors, cytoplasmic second messenger systems and specific target genes i n the nucleus, a process for which the term 'stimulus transcription coupling' has been given. Almost all types of seizures cause dynamic alterations of IEG expression i n neurons of the limbic system, but also in non-limbic areas, such as the cortex, striatum and thalamus. IEG encoded transcription factors are thought t o up- or down-regulate effector genes with preferential expression in the central nervous system, including genes for neurotransmitt ers, growth factors, receptors, synaptic and axonal proteins. If the concept holds true that IEGs act as molecular switches converting epileptic short-term excitation of neurons into alterations of the molecular phenotype, future research may help t o explain hitherto unexplained phenomena in epileptogenesis including changes of synaptic efficacy, kindling and sprouting. Introduction Excitation of neurons in vitro and in vivo principally results in two types of adaptive responses (7,41). First, a rapid, short-lasting biochemical modification of pre-existing target molecules such as receptors, Corresponding author: Or. M . Kiessling, Institute of Neuropathology, University of Heidelberg, 0-691 20 Heidelberg, Germany Tel. +49 (6221) 562 603; Fax +49 (6221) 563 466 Ca 2+ translocation systems and neurotransmitter trafficking. The second type of response comprises a complex series of delayed but long-lasting cellular events, i.e., changes of the neuronal genetic program that ultimately requires de novo protein synthesis. According to current concepts, this genomic response is a prerequisite for long-term changes of neuronal plasticity under physiological and pathological conditions. The first set of genes activated by neuronal stimulation have been termed 'immediate early' or 'early response genes' (8,50,102). Characteristically, (i) these genes can be expressed under conditions of protein 'synthesis inhibition, ruling out that other newly synthesized gene products are responsible for their induction; (ii) IEG encoded mRNA and protein rapidly accumulate after stimulation, but also have a short half-life due to a limited stability of their mRNAs and a rapid turn-over of their proteins. Consequently, IEGs are thought to act as third messengers in an intracellular signal transduction cascade between cell surface receptors (first messengers), cytoplasmic second messenger systems and the nucleus, a process for which the term 'stimulus transcription coupling' has been chosen (73,74). Many immediate early genes (IEGs) participate in a fundamental biological control mechanism, the regulation of gene expression. *The genes fos, jun and KTOX belong to families of IEGs that code for such transcription factors (8,SO). Fos and jun were originally discovered as proto-oncogenes, i.e., normal cellular genes that share large areas of sequence identity with retroviral transforming oncogenes. They are thought to have important functions in cell cycle control as well as growth and differentiation. However, investigations of central nervous system (CNS) neurons have unequivocally demonstrated that IEGs also play a role in postmitotic cells with a defined molecular phenotype. In the CNS, distinct and specific patterns of IEG induction are observed under physiological conditions such as sensory (53), visual (90,119) and audi~~ *Common notation for IEGs exemplified for c-fos: c-fos refers to the gene; c-fos to the c-fos encoded mRNA; c-FOS to the c-fos encoded protein. M. Kiessling and P. Gass: IEG expression in epilepsy Table 1 Regional analysis of immediate early gene (IEG) induction after seizures Seizure Model PTZ in situ Hybridization Immunacytochemistry References Morgan et al. (72) Dragunow & Robert (23) Saffen et al. (95) c-fos C-FOS C-FOS PTZ PTZ c-fos. Krox-24, c-jun, jun B c-fos, c-jun c-fos c-fos c-fos, Krox-24, c-jun, jun B PTZ C-FOS, FRA1, FRA2 C-FOS, FRAl, FRA2 Sonnenberg et al. (107) Sonnenberget al. (108) Daval et al. (18) Cole et al. (16 ) PTZ, Kainate, Electroshock Electroshock Electroshock Electroshock C-FOS, FOS B, KROX-24 C-JUN,JUN B, JUN D C-FOS, FOS 6 , KROX-20, KROX-24 C-JUN,JUN B, JUN D C-FOS C-FOS C-FOS, FOS 8, KROX-20, KROX-24 C-JUN, JUN B, JUN D C-FOS. c-fos c-fos, c-jun, Krox-24 c-fos Dragunow et al. (27) Bicuculline Gass et al. (36) Kainic acid Kainic acid Kainic acid Le Gal La Salle 165) Popovici et al. (84) Gass et al. (38) Kindling Kindling Kindling Kindling . Dragunow & Robert (24) Shin et al. (103) Simonato et al. (104) Clark et al. (13) tory stimulation (29,32), nociception (48) or circadian light shifts (60,61,91). An even more pronounced induction of IEGs is found after pathophysiological stimuli such as epiIepsy (Table l), ischemia (1,39,58,82,112,113,116,117) and brain trauma (22,25). FOS, JUN and the Leucine Zipper Motif The genes fos and jun belong to multigene families and several members have been identified for each, c-fos (11 ) the fos-related antigens pa-1 (14), pa-2 4, (81) and fos B (121) as well as c-jun (94), jun B (93), and jun D (52,92). Characteristically, FOS and JUN proteins contain an evolutionary highly conserved structural domain called the 'leucine zipper' (64). This domain consists of four or five leucines orderly arranged at seven-residue intervals (so-called heptad repeat) and lining up in a column of a regulatory protein segment most likely structured as an alpha helix. These recurring leucine residues 'zip' two FOS/JUN or JUN/JUN proteins together forming a dimeric complex called activator protein 1 (AP-1) (Fig. 1) (62). In close vicinity to the leucine zipper domain is a basic (arginine and lysine rich) domain that is responsible for DNA binding to the AP-1 consensus sequence; a common transcription regulatory element in the promotor region of target or 'late response genes' regulated by IEGs (Fig. 1) (40,62,85). Dimerization via the leucine zipper is mandatory for correct positioning of the two basic DNA binding domains which are both necessary for binding to the AP-1 sequence (40,62,85). Whereas FOS proteins can only form heterodimers with members of the JUN family (15,79), all JUN proteins are able to form homodimeric complexes as well as heterodimeric complexes with FOS proteins or members of other leucine zipper families, such as CAMP/ Can+-response element binding protein 1 (CRE-BPI) (67) and MyoD (6). Dimer formation increases the variety of protein combinations as well as the number of usable DNA motifs. This combinatorial multiplicity is of functional relevance, since different AP-1 complexes have different molecular properties including DNA binding affinity and half-life, as well as up- or down-regulatory effects on target genes. For example, while c-FOS/c-JUN dimers are efficient activators of promotors containing AP-1 sites, c-FOS/JUN B complexes seem to M . Kiessling and F? Gass: IEG expression in epilepsy repress them under certain circumstances (11,98). Similar features have' been discovered by investigating autoregulatory effects of FOS / JUN expression: transcription of the c-fos gene is repressed by its own gene product if heterodimeric binding of c-FOS/c-JUN takes place at the c-fos promotor (63,96). In contrast, c-jun transcription is directly stimulated by its own gene product via a positive autoregulatory loop at the c-jun promotor (2). KROX and the Zinc Finger Motif The combinatory principle can also operate within a single transcription factor by variation of the choice, order and number of independent DNA binding modules within a protein. This applies to another family of transcription factors called 'zinc finger proteins'. This term was given for proteins with so-called 'zinc finger sequences', a motif that is currently considered to be the most frequent DNA binding domain (59). It consists of protrusions formed by a pair of cysteines and a pair of histidines grouped around a central zinc ion and separated by a (finger-like) loop of amino acids, folding into a specific DNA-binding minidomain (Fig. 2) (30). The zinc finger proteins that have been identified so far carry from as few as 2 to as many as 37 zinc fingers, which are each able to connect to specific binding sites located in a groove of the DNA helix. It has been estimated that as much as 1%of the DNA in human cells codes for proteins with zinc finger motifs. In the CNS, Krox-20 (10) [also termed EGR-2 ( 5 6 ) ] and Krox-24 (66) [also termed Zif1268 (12), EGR-1 (110), or NGFI-A (71)], are well-characterized members of the zinc finger family. Both molecules contain three zinc fingers that curl around almost one turn of the DNA helix. The three zinc fingers make contacts with three successive 3-base pair sites on the DNA, called early growth response (EGR) consensus sequence. Apart from immediate early genes another class of zinc finger proteins exists in the CNS: zinc finger sequences are also found as structural units of nuclear hormone receptors and constitute, among others, the DNA binding domains of the glucocorticoid and estrogen receptor (99). Here, however, the zinc binding amino acids are exclusively cysteines instead of cysteines and histidines. Basal Versus Constitutive Immediate Early Gene (IEG) Expression Although fos, jun and Klox were originally defined Figure 1 The leucine zipper motif. The gene products of fos (c-FOS, FOS B, FRA-1 or FRA-2) and jun (c-JUN, JUN B or JlJN D) associate with each other via the leucine zipper domain to form a dimeric protein complex. 'Zippering' occurs by an interaction of five leucine amino acids (L) located in a protein alpha helix at seven amino acid intervals. This 'heptad repeat' causes the leucines to be aligned in a plane along the length of the helix. JUN proteins can form homo- and heterodimers, but FOS homodimers do not occur due to structural constraints. Dimerization is a prerequisite for DNA-binding of basic (arginine/lysine rich) domains to their target sequence, the so-called activator protein 1 (AP-1) binding site (TGACTCA), a common transcription regulatory element in the promotor region of target genes (double helix). as members of rapidly and transiently inducible gene families, they show to a varying degree a basal expression in the CNS of normal animals, suggesting that they also play a role in physiological transcriptional control in the brain (21,26,70). In accordance with their function as transcription factors, immunoreactivity is almost exclusively found in the nucleus upon investigation by immunocytochemistry in sitri (21,26). Immunoelectron-microscopic studies have revealed that c-FOS immunoreactivity is associated with euchromatin supporting the putative action of c-FOS as regulatory element at the transcriptional level (77). In general, basal expression of FOS proteins is restricted to a few scattered neurons, preferentially located in the hippocampus, thalamus and the outer layers of the cerebral cortex (21,26). This expression is possibly due to excitation of specific neuronal circuits processing physiological input. Similar to FOS proteins, c-JUN and JUN B are almost completely absent under unstimulated conditions in the large majority of neurons. However, selective neuronal subpopulations express remarkably high levels of FOS and JUN suggesting a function in the processing of sensory input. Such neuroanatomical areas include the inferior colliculus in acoustic perception (29,32) or the visual cortex (90,119) following photic stimulation. In analogy, disruption of the normal light dark cycle evokes a prominent c-FOS and JUN B expression in the suprachiasmatic nucleus, a central pacemaker which generates an endogenous almost 24-hour M. Kiessling and P. Gass: IEG expression in epilepsy petitive NMDA receptor antagonist MK-801 (37). Despite considerable dispute in the scientific literature about unspecific IEG induction due to handling and stressful manipulations of untreated animals these findings rather suggest a specific role of at least certain IEG encoded proteins in the normal transcriptional control in the CNS. The cerebral pattern of IEG expression, however, changes dramatically after pathological excitation such as seizure activity. IEG Induction After Generalized Tonic-Clonic Seizures Figure 2 The zinc finger motif. Three tandem zinc fingers of KROX-24 protein (also termed Zif-268. EGR-1, NGFI-A) are complexed to their specific DNA-binding sites (EGR-consensus sequence) in the promotor region of a target gene (double helix). The protein as a whole spirals around in the major groove of the DNA helix, each finger making approximately equivalent contacts. The first half of each zinc-finger minidomain forms a p-sheet (bold line), the second half twists into an alpha-helix (tube). Binding of zinc by cysteines in the p-sheet and histidines in the alpha-helical regions draws the halves together near the base of the finger (see insert). The three zinc finger units are linked by short amino acid segments without characteristic three dimensional structure (thin lines). Insert: schematic view of a single zinc finger module, formed by a pair of cysteines and a pair of histidines grouped around a central zinc ion and separated by a (finger-like) loop of 12 amino acids. Modified according to reference 83. periodicity of metabolic and behavioural rhythms (60,61,91). In contrast to those IEGs with a very restricted basal expression, other members of the leucine zipper and zinc finger families demonstrate high expression levels even in the absence of intentional stimulation. Thus, JUN D exhibits high constitutive expression throughout the brain, most likely due to an endogenous permanent activation of the jzin D promotor (19). This high constitutive expression has prompted some authors to postulate that jun D acts as a 'housekeeping gene' controlling the expression of target genes involved in the maintenance of basic cellular metabolic processes (8,9). Similar to JUN D, KROX-24 shows distinct basal expression in the neocortex, preferentially in layers 11 to IV and VI, in the striatum as well as in the limbic system and limbic system associated structures (49,68). In contrast to JUN D, high KROX-24 expression is thought to be maintained by N-methyl-D-aspartate (Nh4DA) receptor mediated physiologic activity. It can be completely blocked by application of the non-com- Epileptic tonic-clonic seizure activity induced by pentylenetetrazole (PTZ) was the first experimental model demonstrating the inducibility of c-fus in the mammalian brain in vivo (72). Subsequently, a large number of experiments have further characterized the temporal and spatial expression pattern of c-fos but also of other IEGs in various seizure models (Table 1). Systemic administration of PTZ or bicuculline to rodents results in generalized tonic-clonic seizures comparable to human 'grand mall epilepsy (23,36,38,72,95,107,108). IEG expression has been investigated both on mRNA level by in sifti hybridization and Northern blot analysis, and on protein level by immunocytochemistry and Western blot analysis (Table 1). These studies have demonstrated a rapid and transient neuronal IEG induction with peak levels in the hippocampus and other parts of the limbic system, but expression also occurred in the somatosensory cortex, striatum and thalamus. Due to, the short half-life of IEG encoded mRNAs, immunocytochemical investigations have revealed different spatiotemporal expression profiles for individual IEGs more precisely than in situ hybridization studies (27,36,107,108). According to these studies, three major categories of IEG encoded transcription factors can best be defined by their staggered induction in the hippocampus: (i) c-FOS and KROX-24 showing a concurrent rapid rise with peak levels at two hours of postictal recovery and a decline to control levels within eight hours after seizure termination; (ii) c JUN and JUN B demonstrating a more gradual increase with peak intensities at four hours and return to baseline levels after 24 hours of postictal recovery; (iii) FOS B, JUN D and KROX-20 exhibiting a delayed induction and prolonged persistance up to 24 hours after seizure termination. All IEG encoded proteins with the exception of KROX-20 demonstrate the same sequence of induction in different hippocampal subfields, following the order dentate gyrus, CA1 and CA3, respectively. Interestingly, this order inversely correlates with the degree of neuronal vulnerability among hippocampal subpopulations after human generalized seizure disorders. However, it can not be ruled out that the sequence of IEG induction within the hippocampus M. Kiessling and P. Gass: IEG expression in epilepsy and among other neuronal populations is related t o the propagation of epileptic discharges or is due t o different molecular genetic properties of hippocampal neurons. KROX-20 is the only transcription factor investigated not induced in the hippocampus by generalized seizures (47). Although all FOS and JUN encoded proteins investigated are induced in the limbic system and somatosensory cortex by generalized seizures, c-JUN is n o t expressed in striatal neurons (36). Recent data have associated c-JUN induction with the capacity of regeneration after transection experiments (45,46). Therefore the expression of c-JUN may identify neuroanatomical areas with the potential for postictal regenerative synaptogenesis and sprouting. IEG Induction After Limbic Seizures but also some scattered pyramidal cells in the hippocampus. At no timepoint is IEG induction observed in the nuclei of degenerating neurons, corroborating findings in a model of global ischemia that expression of IEG encoded proteins identifies neurons at risk but which are destined t o survive (58). Kindling as a Model for Epileptogenesis Partial epilepsy of temporolimbic origin is the most common and often intractable type of human seizure disorders. Administration of the glutamate agonist kainic acid (KA) t o rats has been established as a standardized model for human temporal lobe epilepsy, mimicking clinical, electrophysiological and neuropathological features with high accuracy (5). KA-induced limbic seizures cause a highly specific regional pattern of IEG expression that is strikingly similar to that after generalized seizures of the PTZ/ bicuculline type. However, markedly prolonged timecourses of IEG expression are observed following KA-induced seizures, most likely reflecting prolonged seizure activity (Fig. 3) (38,65,84). Again, FOS B, JUN D and KROX-20 exhibit delayed induction and prolonged expression compared to c-FOS, c-JUN and JUN B (38). Despite this asynchronous activation of individual IEGs, the sequence of induction in the hippocampus (dentate gyrus > CA1> CA3) is identical to that after generalized seizures (Fig. 3). The further sequence of gene activation, progressing from hippocampus to limbic system associated areas and subsequently t o non-limbic structures, correlates well with the generation and propagation of paroxysmal activity produced by KA (38,65,84). Also, in analogy t o generalized seizures of the PTZ/ bicuculline type, the expression profile of c-FOS and KROX-24 is very similar within limbic system structures (36,38). In non-limbic areas such as somatosensory cortex and striatum, however, KROX-24 levels increase after the onset of seizure activity, but 24 hours later they drop markedly below baseline levels. Since high basal levels of KROX-24 are thought to be regulated by NMDA receptor mediated physiological neuronal activity (37,119), the marked reduction of KROX-24 i n the cerebral cortex at later time intervals probably reflects postictal depression of electrophysiological activity ( 5 ) . In contrast t o bicuculline- and PTZinduced experimental seizures, systemic KA-evoked limbic seizure activity can cause neuronal degeneration, affecting many neurons in the piriform cortex Kindling is an animal model of epileptogenesis, i n particular for human complex partial epilepsy. After an appropriate time interval of days or weeks, repeated administration of an initially subconvulsive electrical or pharmacological stimulus finally results in intense limbic and generalized motor seizures. Once established, the kindling effect is permanent. The molecular basis of kindling is unknown. Kindling depends on de novo protein synthesis and it is tempting to speculate that IEG induction could represent the link between repeated short-term stimuli, gene expression and alterations in the neuronal phenotype associated with the development of the kindled state (101). Periodic induction of afterdischarges (ADS), i.e., local electrical seizures of a few seconds without overt behavioural effects, is a prerequisite for kindling. A single AD can evoke increased expression of c-fos, c-jun and Krox-24 mRNA in the hippocampus (104). Similar t o generalized seizure activity different timecourses are found for individual transcription factors, the sequence of induction being c o x - 2 4 > c-fos > c-jun (104). There is an all or nothing relationship between induction of IEGs and the duration of ADS with a time threshold of about 30 seconds (103). The above threshold ADS induce c-fos in both naive and kindled animals to the same extent and with identical temporal profiles and the kindled state is not accompanied by long-term changes of basal IEG expression (103,104). Thus, IEG induction seems t o be involved in producing the kindled state, but once established, n o permanent up-regulation of IEG levels occurs . IEG Induction and Sprouting Synaptic reorganization of limbic system and temporal lobe pathways is a prominent feature in experimental epilepsy and in human seizure disorders (111). Thus, a long-term consequence of repetitive seizure activity is the sprouting of axons, in particular of hippocampal mossy fibers. Sprouting has been observed after experimental KA-induced seizures (33,78), kindling (86,111) and in human childhood epilepsy (87). This has prompted some authors to put forward the hypothesis that IEGs play a crucial role in the development of sprouting by initiating a chain of molecular events culminating i n morphological alterations (75). Several correlations exist between seizures, IEG induction and sprouting: (i) Induction of c-fos and sprouting evoked by repe- M. Kiessling and P. Gass: IEG expression in epilepsy titive PTZ seizures require similar dose thresholds (42,72); (ii) The most pronounced IEG induction and sprouting occur in the same CNS regions, such as dentate gyrus (Fig. 4); (iii) Sprouting requires only a few days to become apparent; and (iv) Anti- epileptic drugs that are known to block the induction of c-FOS and kindling also impair or block sprouting. Clinical implications of these data are at present speculative. If IEGs indeed play a role in the development of sprouting, then blockade of M. Kiessling and P. Gass: IEG expression in epilepsy Figure 3 Differential postictal kinetics of c-FOS and FOS B induction in the rat hippocampus after kainic acid induced limbic seizures. c-FOS and FOS 6 are virtually absent in untreated control animals KO). After kainic acid elicited limbic seizures, c-FOS and FOS 6 are markedly induced in neuronal nuclei of all hippocampal subpopulations, demonstrating a specific sequence of induction, i.e., dentate gyrus > CA1 > CA3. FOS 6 expression is clearly delayed and prolonged compared to c-FOS. Thus, early on heterodimeric AP-1 complexes are likely to be predominantely composed of c-FOS and one of the JUN proteins, whereas FOS B is probably the partner to form complexes with JUN during late postictal recovery periods. Vibratome sections: 50 pm. x 40. their expression might have a therapeutic impact on epileptogenesis. Mechanisms of IEG induction Since almost all types of seizures induce IEGs in identical brain regions and neuronal circuits one could speculate whether a single mechanism such as stimulation of a particular neurotransmitter system can account for all modes of postictal IEG induction. However, pharmacological studies with specific receptor antagonists suggest considerable diversity. In generalized seizures of the PTZI bicuculline type, glutamate receptors of the NMDA subtype apparently mediate the major component of IEG induction, since transcription factor synthesis is largely attenuated by NMDA receptor antagonists such as MK-801 (72,108). These data may indicate that by reducing the activity of GABA, receptors, KTZ and bicuculline cause a disequilibrium between inhibitory and excitatory neurotransmission, thereby increasing or potentiating NMDA receptor stimulation. In contrast, IEG induction after electrically evoked seizure discharges of the type utilized in clinical electroconvulsive treatment, can be markedly attenuated by agents directly or indirectly blocking voltage sen- sitive calcium channels (VSCC), but not by MK-801, suggesting that IEG induction in this model is mediated via VSCC (16). I n KA-evoked temporal lobe seizures, pretreatment with MK-801 similarly does not affect IEG induction in the limbic system. This finding is in line with the notion that IEG induction in these regions is mediated by high affinity glutamate receptors of the kainate I AMPA subtype rather than NMDA receptors (Fig. 6). MK-801 though abolishes IEG induction in sopatosensory cortex and striatum, suggesting that paroxysmal activity in nonlimbic neurons occurs by a transynaptic activation o€NMDA receptors, which in addition to the hippocampus are particularly enriched in these structures (Fig. 6). Alternatively, the blockade of IEG expression by MK-801 in cortex and striatum but not in the limbic system may reflect a different regional susceptibility of different brain regions to KA-induced seizures. How is receptor stimulation at the cell surface COUpled to transcriptional activation of IEGs in the nucleus. For c-fos, two important regulatory elements located in the 5' untranslated region of its promotor have been identified: (i) the serum response element (SRE) located at basepair -310, and (ii) the CAMP/ M. Kiessling and P Gass: IEG expression in epilepsy . KAINATE KAINATE +MK-801 CRE site in the c-fos promotor. Differential intracelMar Caz+ entry may thus provide a mechanism for the control of diverse genomic responses (3). Molecular Consequences of Differential IEG Induction Figure 4 Pharmacdogical modification of immediate early gene (IEG) expression after limbic seizures. At 3 hours after K provoked seizures, JUN B and c-FOS are markedly induced A in the limbic system (hippocampus, piriform cortex, amygdala), and also in non-limbic areas (neocortex, striatum and thalamus); A,C Pretreatment of animals with the non-competitive NMDA receptor antagonist MK-801 strongly attenuates IEG induction in non-limbic regions, suggesting that paroxysmal activity in these structures involves transynaptic activation of NMDA receptors; B,D Despite pretreatment with MK-801, seizure elicited IEG expression is unaltered in neurons of the limbic system indicating that IEG induction in these regions is rather mediated by high affinity glutamate receptors of the kainate subtype. (50 p m vibratome sections of the rat forebrain at the level of the anterior commissure (A,B) and a t the level of the dorsal hippocampus (C,D), x 10). The existence of related members such as the leucine zipper or the zinc finger proteins with a common biological function as transcription factors but different specificities could equip the cell with a mechanism for fine regulation of the genetic program. To elucidate the effects of the differential induction of IEGs, it is important to study the expression of fos and jzin on the protein level, because the combinatorial changes of leucine zipper proteins greatly expand their regulatory potential. Since the composition of Fos/Jun dimers is determined by the availability of the different components, the apparently stereotypic rapid induction and disappearance of c-FOS in all seizure models indicates, that FOS/ JUN heterodimers in the early postictal recovery period mainly consist of c-FOS and one of the rapidly synthesized JUN proteins (c-JUN, JUN B). In contrast, it can be assumed that due to their prolonged persistence AP-1 complexes during the late postictal recovery period are predominantly assembled of FOS B and JUN D. In vitro studies have demonstrated that AP-1 complex composed of FOS B and JUN D is the most stable dimer, thus extending the action of these transcription factors. The staggered induction of individual FOS /JUN proteins and the subsequent postictal. combinatorial changes allow a sequential and differential regulation of target gene expression. Similarly to AP-1 complexes, Krox-24 and Krox-20 demonstrate a staggered but overlapping induction and compete for the same DNA sequences, the EGR elements. Additional complexity arises from the fact, that expression of target genes may not be controlled by a single regulatory element, but that several elements (such as AP-1 or EGR) can coexist and cooperate.in a promotor region, and may allow for crosstalk between different transcription factors. IEG Regulated Target Genes Ca2+-response element (CRE) at basepair - 60 (102). The SRE can be activated by Caz+ influx through NMDA receptor gated ion channels and by growth factors like NGF, which cause a Ca2+ independent activation of c-fos. The intracytoplasmic effectors that convey these signals to the SRE are currently unknown. In contrast to Ca2+ fluxes via NMDA receptors, Ca*+ entry by VSCC induces c-fos via the CRE (3,102). VSCC mediated Ca2+ influx causes an activation of intracytoplasmic Ca2+/calmodulin protein kinase with subsequent phosphorylation of the so-called CRE binding protein, finally activating the The complex pattern of postictal cerebral IEG induction and/or suppression raises a critical question: Is IEG expression part of a cellular homeostatic response, endowing the CNS with the capacity to cope with the potentially harmful effects of excessively enhanced neuronal activity (e.g., by replenishment of neurotrophic factors, an up-regulation of inhibitory and/or a down-regulation of excitatory neurotransmitter systems) or does it alternatively contribute to the development of the epileptic state? To define the nature of cellular adaptive responses elicited by epilepsy and potentially mediated by IEGs requires the identification of specific target genes that are controlled by leucine zipper and zinc M. Kiessling and P. Gass: IEG expression in epilepsy finger proteins. Although it is generally believed that the products of 'primary response genes', like IEGencoded proteins switch on or turn off 'late response genes', the identification of target genes has proved technically difficult. In vitro studies have demonstrated that FOS/JUN complexes are capable of regulating genes with preferential expression in the nervous system, such as nerve growth factor (NGF), prodynorphin and proencephalin (44,80,109). The promoter regions of these genes have AP-1 sites, which can be activated by FOS/JUN complexes in transactivation assays in vitro (44,80,109). However, it has yet to be established whether IEG encoded proteins are the transcription factors which actually regulate these genes in vivo, as most interpretations of in vivo studies rely upon temporal or spatial correlations only. After experimental seizures, FOS / JUN induction precedes the up-regulation of genes coding for NGF (34), brain-derived neurotrophic growth factor (28, 54,120), proenkephalin (20), neuropeptide Y (4), glutamate decarboxylase (31), and the down-regulation of prodynorphin (20,57,76) and a putative kainate receptor (35). Altered hippocampal gene expression in human temporal lobe epilepsy also comprises an up-regulation of genes encoding acetylcholinesterase (43), somatostatin (89), as well as an up- and down-regulation of various excitatory and inhibitory amino acid receptors (69). These changes of gene expression, leading to altered neuromodulator, neurotransmitter and neurotrophin levels, altered receptor sensitivity as well as presumably altered regulatory and structural proteins of the axon and synapse, are likely to be key events in the molecular program underlying synaptic plasticity after repetitive seizures, but could either counteract or contribute to the epileptic state. Compared to leucine zipper molecules, even less is known about the regulation of target genes by zinc finger proteins. A recent study provided evidence, that KROX-20 regulates the expression of HoxB2, a gene that plays a crucial role in hindbrain segmentation during ontogenetic development (loo), but no relevant data are presently available on physiological or pathological neuronal excitation. IEG Induction in Transgenic Animals As an alternative approach to mapping IEG expression by in situ hybridization and immunocytochemistry, a fos-ZacZ fusion gene (that confers inducible P-galactosidase expression via a fused c-fos promotor) has been introduced into the germ line of mice by microinjection (105). In these mice strains, enzymatic detection of P-galactosidase activity identifies cell populations with basal or induced c-fos expression. One out of three fos-ZacZ strains originally obtained largely recapitulated the kinetics of c-fos expression in unstimulated animals and the induction profile during early postictal time inter- vals, although the longer half-life of the Fos-lacZ fusion protein compared to c-FOS slightly extends the period of expression (97). At late time intervals (four to ten days) after seizure induction, however, focal cytoplasmic p-galactosidase staining was noted in the pyramidal cell Iayer of the hippocampus and the dorsomedial nucleus of the amygdala and interpreted as a reinduction of fos-ZacZ (106). Marked differences were also observed in the percentage of Fos-lacZ positive nuclei compared to c-FOS in certain brain regions after PTZ and kainic acid induced seizures (105). These discrepancies with previous studies were explained as a possible consequence of cross-reactivity of many c-FOS antibodies with FOSrelated antigens. The different extent of fos-ZacZ expression first noted in transgenic mice was interpreted to reflect a different regulation of IEGs in these two seizure models mediated by distinct transcriptional control mechanisms. It has to be kept in mind, however, that a major caveat of the fos-ZacZ approach is a possible mutation in the transgene or an unpredictable effect of the transgene integration on the expression of the fos-ZacZ fusion construct. This is also reflected by the heterogeneity of strains, two of which completely failed to show increase of p-galactosidase activity in any brain region after PTZ induced seizures (105). A new approach to determine specific roles of individual IEGs in vivo is to produce transgenic' animals lacking particular IEGs by homologous recombination. Recently, two laboratories independently produced mutant mice strains lacking the c-fos gene (55,115). Surprisingly, these c-fos knock out mice that indeed do not express detectable levels of c-FOS, are viable. The mutants showed reduced viability immediately after birth, but a normal life expectancy of survivors with growth retardation, osteopetrosis, defects of female transmission and lyrnphopenia with extramedullary hematopoiesis. Subtle behavioral deficits including the lack of response to external stimuli, hyperactivity and disturbances of the adjustment to the light/dark cycle may indicate some alterations in neural function, but so far no morphological abnormalities have been detected in the CNS. IEG expression (apart from c-FOS) has not been investigated in c-fos knock out mice. In contrast to transgenic mice without the c-fos gene, mutant mice strains lacking the c-jtm gene are not viable but die in utero around embryonic day 10 (51). Therefore, in contrast to c-fos, a lack of c-jun can not be compensated during development and early postnatal life. According to our knowledge, no attempts have been made to produce KROX24 knock out mice. However, these mice would be of particular interest, since KROX-24 represents the only IEG thought to be involved in the induction of long-term potentiation (LTP), currently the most compelling model for learning and memory (17,88,118). M. Kiessling and P. Gass: IEG expression in epilepsy 7. Black IB. Adler JE, Dreyfus CF, Friedman WF, LaGamma EF, Roach AH (1987) Biochemistry of information storage in the nervous system. Science 236: 1263-1268 8. Bravo R (19901 Growth factor responsive genes in fibroblasts. Cell Growth Diff 1: 305-309 9. Bravo R (1990) Growth factor inducible genes in fibroblasts. In: Growth Factors, Differentiation Factors and Cyiokines, Herschmann A (ed.), pp. 324-343, Springer: Berlin , 10.Chavrier P Zerial M, Lemaire P Almendral J. Bravo R, , Charnay P (1988) A gene encoding a protein with zinc fingers is activated during GO/G1 transition in cultured cells. EMBOJ 7: 29-35 11. Chiu R, Angel P Karin M (1989) JUN B differs in its bio, logical properties from, and is a negative regulator of, c-JUN. Cell 59: 979-986 12. Christy BA, Lau LF, Nathans D (1988) A gene activated in mouse 3T3 cells by serum growth factors encodes a protein with zinc finger sequences. froc Natl Acad Sci USA 85: 7857-7861 13.Clark M, Post RM, Weiss SRB, Cain CJ, Nakajima T (1991 ) Regional expression of c-fos rnRNA in rat brain during the evolution of arnygdala kindled seizures. Mol Brain Res 11: 5 5 6 4 14. Cohen DR, Curran T (1988) Fra-1 : A serum-inducible, cellular immediate-early gene that encodes for a fos-related antigen. M o l Cell Biol 8: 2063-2069 15.Cohen DR, Ferreira PCP, Gentz R, Franza Jr BR, Curran T (1989) The product of a fos-related gene, fra-I, binds cooperatively to the AP-1 site with jun: Transcription factor AP-1 is comprised of multiple protein complexes. Gene Develop 3: 173-184 16.Cole AJ, Abu-Shakra S, Saffen DW, Baraban JM. Worley PF (1990) Rapid rise in transcription factor mRNAs in rat brain after electroshock-induced seizures. J Neurochem 55: 1920-1927 Conclusions The picture that emerges from the numerous studies of the IEG response to seizure activity is increasingly complex. The large number of seizure-induced transcription factors, their diverse sequence-specific interactions with other transcription factors and with DNA recognition sites provide combinatorial codes for highly intricate and organized patterns of gene expression. Subtle changes in the combination of transcription factors may have profound consequences for neuronal function and phenotype in response to seizure activity. Although many questions remain unanswered, there are intriguing findings suggesting that the induction of c-fos and other immediate early genes may be associated with epileptogenesis. Further attempts must be made to identify target genes for IEG encoded homo- and heterodimeric protein complexes, in order to elucidate the causal chain of the molecular events underlying morphological and functional consequences of seizure activity. Gain of function and/or loss of function studies in transgenic animals or using antisense technology may be the most valuable tool currently available to elucidate the link between IEG mediated regulation of target genes and adaptive neuronal responses in epilepsy. Acknowledgements We apologize to those individuals whose important original contributions are not cited because of space limitations. Co-workers we would like to thank are Drs. Rodrigo Bravo and Thomas Herdegen. Special thanks are extended to Dr. Frauke Bentzien for the graphics of Figures 1 and 2. Part of this work was supported by a grant from the Deutsche Forschungsgemeinschaft (SFB 31 7).

Journal

Brain PathologyWiley

Published: Oct 1, 1993

There are no references for this article.