• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Mol Cell Neurosci. Author manuscript; available in PMC May 18, 2009.
Published in final edited form as:
PMCID: PMC2683345
NIHMSID: NIHMS23136

Egr3, a synaptic activity regulated transcription factor that is essential for learning and memory

Abstract

Learning and memory depend upon poorly defined synaptic and intracellular modifications that occur in activated neurons. Mitogen activated protein kinase-extracellular regulated kinase (MAPK-ERK) signaling and de novo protein synthesis are essential aspects of enduring memory formation, but the precise effector molecules of MAPK-ERK signaling in neurons are not well defined. Early growth response (Egr) transcriptional regulators are examples of MAPK-ERK regulated genes and Egr1 (zif268) has been widely recognized as essential for some aspects of learning and memory. Here we show that Egr3, a transcriptional regulator closely related to Egr1, is essential for normal hippocampal long-term potentiation (LTP) and for hippocampal and amygdala dependent learning and memory. In the absence of Egr3, the defects in learning and memory appear to be independent of Egr1 since Egr1 protein levels are not altered in amygdala, hippocampus or cortex. Moreover, unlike Egr1-deficient mice which have impairments in late phase hippocampal LTP and consolidation of some forms of long-term hippocampus- and amygdala-dependent memory, Egr3-deficient mice have profound defects in early- and late-phase hippocampal LTP, as well as short-term and long-term hippocampus- and amygdala-dependent learning and memory. Thus, Egr3 has an essential role in learning and memory processing that appears to be partly distinct from the role of Egr1.

Keywords: learning, memory, transcriptional regulation, plasticity, MAPK, Erk, mouse

Introduction

Neurons process and retain information by forming synaptic connections that are modified by the intensity and frequency of their historical activity. This remarkable capacity to regulate the efficacy of synaptic transmission is essential for continual remodeling of neural networks required for cognitive processes such as learning and memory. Long lasting synaptic enhancement of excitatory neurotransmission, known as long-term potentiation (LTP), is one of the best studied cellular forms of synaptic “plasticity.” Many types of synapses throughout the brain undergo LTP after appropriate patterned stimulation but the phenomenon has been best studied in the hippocampus, a brain structure that is essential for memory acquisition and consolidation (Milner et al., 1998). Enduring memory-related synaptic changes depend upon new gene transcription and translation (protein synthesis), but the specific molecular pathways involved have only been partially characterized. For example, activity dependent calcium flux into neurons through excitatory N-methyl D-aspartate (NMDA) glutamate receptors and/or voltage-gated calcium channels engages the mitogen-activated protein kinase (MAPK)-extracellular regulated kinase (ERK) signaling cascade which is essential for NMDA-dependent induction of LTP in area CA1 (English and Sweatt, 1996, 1997) and the dentate gyrus of the hippocampus in vitro (Coogan et al., 1999), and for regulation of LTP in vivo (Davis et al., 2000; Rosenblum et al., 2000). Moreover, pharmacologic inhibitors of ERK signaling impair contextual and spatial memory (Atkins et al., 1998; Blum et al., 1999). A major deficit in our understanding of MAPK-ERK signaling related to the regulation of LTP and neuronal information storage at the synaptic (LTP) and network (behavioral) level remains because the genes regulated by MAPK-ERK signaling are still poorly characterized (Thomas and Huganir, 2004).

Early growth response (Egr) transcription factors are among a relatively small number of regulatory immediate early genes (IEGs) that are expressed in response to neuronal activity and coupled to MAPK-ERK signaling. There are four structurally related Egr genes (Egr1–4) that encode proteins with highly conserved zinc finger DNA-binding domains. As MAPK-ERK effector molecules, they may regulate target gene expression required for long-term structural and/or physiologic synaptic changes associated with learning and memory. Egr1 (also known as zif268, NGFI-A, TIS8, Krox24 or ZENK) and Egr3 are the most abundant Egr proteins upregulated by synaptic activity in the brain (Li et al., 2005; O’Donovan and Baraban, 1999). Early studies correlated patterned synaptic activity required to induce LTP with de novo induction of Egr1 and Egr3 expression (Cole et al., 1989; Yamagata et al., 1994). Recently, Egr1 has been shown to be required for the maintenance of late phase (but not early phase) LTP, long-term memory consolidation and reconsolidation of previously formed memories (Bozon et al., 2003; Jones et al., 2001; Lee et al., 2004).

Although a role of Egr1 in learning and memory is now well established, it is not clear what role, if any, Egr3 may have in learning and memory processing. In recent studies, we showed that Egr1 and Egr3 can regulate target genes such as the plasticity associated Arc gene (Activity Regulated Cytoskeletal associated gene; also known as Arg3.1 (Link et al., 1995; Lyford et al., 1995)), suggesting that Egr1 and Egr3 may have some overlapping roles in regulating gene expression in the brain (Li et al., 2005). Similarly, other plasticity associated genes encoding the GABA receptor subunit 4 gene, GABRA4 (Roberts et al., 2005) and synaptic vesicle associated proteins, Syn1 (Thiel et al., 1994) and Syn2 (Petersohn et al., 1995) may be directly regulated by either Egr1 or Egr3. Thus, if Egr3 modulates the expression of important plasticity-associated genes in a physiologically relevant manner, it may be essential for some aspects of learning and/or memory. Here, we report that despite the fact that Egr3-deficient (Egr3−/−) mice appear to have normally developed brains and normal basal synaptic transmission in CA3-CA1 hippocampal neurons where Egr1 and Egr3 are highly expressed, they have abnormal LTP in CA1 neurons, they have profound impairments in context and cued-associative learning/memory, and they have profoundly impaired short-term and long-term object recognition memory. Thus, Egr3 has an essential role in learning and memory presumably by regulating effector target genes required for memory acquisition, consolidation and/or retrieval.

Results

Normal brain development and basal synaptic transmission in Egr3−/− mice

Egr3−/− mice were previously generated but brain development was not systematically studied (Tourtellotte et al., 2001; Tourtellotte and Milbrandt, 1998). Egr3−/− brains were compared to littermate WT to determine whether there were any detectable morphologic, pharmacologic and/or physiologic differences. Detailed examination of littermate WT and Egr3−/− embryonic, perinatal and adult brains showed no gross or microscopic differences (data not shown). The hippocampus in particular, a brain region essential for normal learning and memory and where Egr1 and Egr3 are strongly induced by synaptic activity, appeared structurally normal (Fig. 1A, A). In addition, there were no detectable differences between WT and Egr3−/− brains in the distribution and density of neurons (Fig. 1B, B), synaptic terminals (Fig. 1C, C), or parvalbumin expressing interneurons (Fig. 1D, D) in the hippocampus. Similarly, Golgi-Cox staining showed no obvious alterations in neuron structure, orientation or migration in any brain regions examined (data not shown).

Figure 1
Hippocampal development is normal in Egr3-deficient mice. Normal neuron distribution, synaptic terminal labeling and interneurons are identified in the hippocampus of (A, B, C, D) WT and (A, B, C, D′) Egr3−/− ...

Dendritic synaptic spines on CA1 hippocampal neurons, which are enriched for excitatory glutamatergic receptors and which may be correlated with structural changes related to long-term memory (reviewed in (Segal, 2005)), were examined using Golgi-Cox staining. Spine structure appeared normal and spine density was similar between WT (1.07 ± 0.03/μm) and Egr3−/− (1.01 ± 0.04/μm) CA1 neuron secondary dendrites (p > 0.3) (Fig. 2A). In addition, there was no significant difference in neuron density in CA1 (WT = 6.55 ± 2.07 × 105/mm3, Egr3−/−= 6.02 ± 2.03 × 105/mm3, p > 0.8) and CA3 (WT = 3.63 ± 1.15 × 105/mm3, Egr3−/−= 2.36 ± 1.05 × 105/mm3, p > 0.5) pyramidal neurons, or in the dorsal (d) (WT = 10.84 ± 3.00 × 105/mm3, Egr3−/− = 6.08 ± 2.03 × 105/mm3, p > 0.3) and ventral (v) (WT = 10.22 ± 2.86 × 105/mm3, Egr3−/−= 6.83 ± 1.85 × 105/mm3, p > 0.4) blades of the hippocampal dentate gyrus (dg) (Fig. 2B).

Figure 2
Normal CA1 hippocampal neuron spine density and hippocampal neuron numbers in Egr3−/− mice. (A) Golgi-cox staining shows normal CA1 pyramidal neuron morphology with typical primary apical dendrites (1°) and spine laden secondary ...

Since glutamatergic signaling is essential for normal learning and memory processing, the protein levels of several important glutamate receptor subunits were examined in WT and Egr3−/− amygdala, hippocampus and somatosensory cortex. There were no significant changes in the NMDA (NR1, NR2A and NR2B), AMPA (GluR1, GluR2) or metabaphoric (mGluR5) glutamate receptor subunits examined (Fig. 3A). Similarly, protein levels of Erk1/2 and activated (phosphorylated) Erk1/2 (p-Erk1/2), essential mediators of short-term memory and early-phase LTP (English and Sweatt, 1997), were similar between Egr3−/− and WT brains (Fig. 3A). Egr1 protein, which was previously shown to be essential for late-phase LTP and long-term memory (Bozon et al., 2003; Jones et al., 2001), was not significantly different between WT and Egr3−/− amygdala, hippocampus or somatosensory cortex (Fig. 3B).

Figure 3
Biochemical analysis of glutamatergic signaling molecules in WT and Egr3−/− hippocampus, amygdala and somatosensory cortex. (A) Glutamate receptor subunits, Erk1/2, and activated Erk1/2 (p-Erk1/2) protein levels were analyzed by Western ...

Basal synaptic transmission was examined in CA3-CA1 synapses since both pre-synaptic and post-synaptic neurons express high levels of Egr3 protein induced by strong CA3-CA1 neuronal activity (Li et al., 2005). Basal synaptic transmission, measured in acutely prepared hippocampal slices by varying the current pulse amplitude applied to CA3 Schaffer collaterals (20–250 μA) and by determining the amplitude of evoked excitatory post-synaptic potentials (EPSPs) in CA1 stratum radiatum, was not significantly different between WT and Egr3−/− mice (p > 0.5, Fig. 4A). Similarly, there was no significant difference between WT and Egr3−/− mice in paired pulse facilitation (PPF, a measure of presynaptic plasticity) of CA3-CA1 synapses when test pulses were administered to Schaffer collaterals at varying interpulse intervals (p > 0.8, Fig. 4B). Finally, the baseline Schaffer collateral synaptic EPSP initial slope was not significantly different between WT (0.73 ± 0.05 mV/ms, N=9) and Egr3−/− (0.65 ± 0.05 mV/ms, N=9) mice (p > 0.2).

Figure 4
Normal basal synaptic transmission in Egr3−/− hippocampal Schafer collateral synapses. (A) CA1 excitatory post synaptic potentials (EPSPs) elicited by stimulating WT and Egr3−/− Schaffer collateral synapses with varying ...

Hippocampal-dependent fear conditioning and novel object discrimination are impaired in Egr3−/− mice

To examine whether Egr3 is essential for normal learning and memory, WT and Egr3−/− mice were tested for classical contextual and cued associative fear conditioning, and novel object discrimination. However, a series of control experiments was first performed to determine whether abnormal behaviors elicited by the tests were due to sensory and/or motor abnormalities, or whether they were attributable to memory defects. Naïve WT and Egr3−/− mice showed similar activity when they were exposed to the experimental fear conditioning context environment (AC0, p = 0.46) and when they were exposed to the experimental context environment and presented with an audible tone (AT0, p = 0.36). Moreover, WT and Egr3−/− mice responded to foot shock with a significant increase in locomotor activity (AS, p <0.0001), and there was no difference in the induced activity response to shock between WT and Egr3−/− mice (AS, p = 0.6; Fig. 5A). WT and Egr3−/− mice also had similar behavioral responses to the foot shock. At low shock intensity, there was no significant difference between WT and Egr3−/− mice in their flinching reflex (p = 0.23) and at higher shock intensity there was no significant difference in the shock intensity required to elicit a vocalization response (p = 0.7; Fig. 5B). In addition, there was no detectable difference between WT and Egr3−/− mice in response to thermal nociceptive stimuli (p = 0.62; Fig. 5C). Thus, Egr3−/− mice do not have any obvious defects in their ability to perceive and respond to nociceptive stimuli from shock or heat.

Figure 5
Locomotor activity, shock threshold response, thermal nociception, non-associative learning and overtraining fear conditioning behavior compared between WT and Egr3−/− mice. (A) Naïve WT and Egr3−/− mice showed ...

To examine whether Egr3−/− mice can distinguish between different contexts and whether they have impairments in non-associative learning and/or memory, WT and Egr3−/− mice were allowed to freely explore an open field context. Their exploratory activity was measured during a 3 minute training period and training was repeated in the same context each day for 5 consecutive days. There was no significant effect of genotype on the level of exploratory activity (F1,6 = 0.2; p = 0.66), but there was a significant reduction in their activity over time (F4,35 = 4.4; p < 0.01; Fig. 5D), suggesting they have similar non-associative learning during habituation to the context. After weak contextual fear conditioning both WT and Egr3−/− mice exhibited significant decreased activity in response to shock-context pairing (p < 0.001) and the locomotor activity was significantly increased when the mice were subsequently tested in a novel context (p < 0.001). Thus, Egr3−/− mice are capable of at least some forms of non-associative learning and they have at least some capacity to discriminate between two different contexts.

To examine whether Egr3−/− mice have any capacity to learn from fear conditioning and to elicit the freezing behavior, WT and Egr3−/− mice were trained using a high reinforcement (overtraining) paradigm. On the first day of training, the mice received 7 consecutive (reinforcement) repetitions of context/tone-shock pairing, between which freezing was assessed (see methods). Although Egr3−/− mice showed a clear increase in freezing as the reinforcement training progressed during the first day of training, there was a significant effect of genotype (F6,70 =13.9; p < 0.001; Fig. 5F) on freezing with Egr3 mice consistently showing less freezing than WT during the reinforcement training. On the second day, after 7 rounds of reinforcement training, there was no significant difference in freezing behavior between WT and Egr3−/− mice (p = 0.25; Fig. 5F). Thus, Egr3−/− mice have some capacity for associative learning and they are capable of eliciting similar freezing behaviors as WT mice, if enough reinforcement training is provided.

If Egr3 is involved in regulating synaptic activity dependent genes in the hippocampus that are required for normal learning and memory, Egr3−/− mice may have learning and memory impairments in tasks that depend upon normal hippocampal function. We examined the performance of WT and Egr3−/− mice in contextual fear conditioning and novel object discrimination, both of which depend upon intact hippocampal function (Clark et al., 2000; Kim and Fanselow, 1992; Kim et al., 1993; Phillips and LeDoux, 1992; Zola et al., 2000). Naïve WT and Egr3−/− mice received either single context-shock pairing (weak training) or triple context-shock pairing (strong training). Memory was assessed by measuring the conditioned response (freezing) to the same environmental context 24 hours after weak training, and 0.5 hours and 24 hours after strong training. There was a significant effect of genotype on weak (F1,25 = 29.61, p < 0.0001) and strong (F1,25 = 48.73, p < 0.0001) training (Fig. 6A). Although naïve WT and Egr3−/− mice had a similar low frequency of freezing to context (p = 0.86), WT mice showed significantly increased freezing to context 24 hours after weak training relative to their naïve state (p < 0.0001), whereas Egr3−/− mice showed no significant increase in freezing to context relative to their naïve state (p = 0.16). Thus, Egr3−/− mice had no significant memory of the conditioned stimulus (context) 24 hours after weak training, which was manifest by significantly decreased freezing to context relative to conditioned WT mice (p < 0.0001; Fig. 6A). Similarly, 0.5 hours after strong training, WT mice showed markedly increased freezing to context compared to naïve WT mice and freezing remained elevated and unchanged for at least 24 hours after strong training (p > 0.6). By contrast, Egr3−/− mice showed a significant freezing to context 0.5 and 24 hours after strong training relative to naïve Egr3−/− mice (p < 0.0001 and p < 0.01, respectively), but freezing was significantly decreased both 0.5 and 24 hours after strong training compared to WT mice (p < 0.0001; Fig. 6A). Whereas freezing to context remained high and showed no significant change between 0.5 and 24 hours after strong training in WT mice (p= 0.2), the frequency of freezing to context significantly decreased between 0.5 and 24 hours after strong training in Egr3−/− mice (p < 0.05) (Fig. 6A). Thus, Egr3−/− mice have highly impaired short term (0.5 hour) and long term (24 hour) contextual fear memory, and by contrast to WT mice which retain a memory of the aversive context for at least 24 hours after strong training, in Egr3−/− mice the memory is significantly dissipated between 0.5 and 24 hours after strong training.

Figure 6
Examination of hippocampus and amygdala dependent memory in Egr3−/− mice. (A) Naïve WT and Egr3−/− mice showed very little freezing to context prior to training. However, 24 hours after weak training freezing to ...

WT and Egr3−/− mice were next tested for their ability to discriminate between novel (N) and familiar (F) objects and a significant effect of genotype was observed (F1,20 = 6.146, p < 0.05). WT and Egr3−/− mice were habituated over several trials to two objects within a context and they were subjected to training which showed that WT and Egr3−/− mice spent equal time exploring both F objects (p = 0.4 and p = 0.76, respectively; Fig. 6B (Train)). After habituation training, the mice were evaluated 10 minutes and 24 hours in the same context with a single familiar object replaced by a novel object. Rodents tend to explore novel objects more frequently in this task, reflecting a memory of the familiar object which is required to discriminate the familiar object from the novel object (Clark et al., 2000). Thus, the amount of time exploring the novel object relative to the familiar object is an indirect measure of memory for the familiar object (see methods). WT mice showed a significant propensity to explore the novel object both 10 minutes (p < 0.001) and 24 hours (p < 0.001) after training. By contrast, Egr3−/− mice showed no significant propensity to explore the novel object 10 minutes (p = 0.17) or 24 hours (p = 0.15) after training, suggesting that they had no memory of the “familiar” object encountered during prior habituation and training sessions.

Egr3 is up-regulated by synaptic activity in the amygdala and cortex, suggesting that it may have a more generalized function in regulating activity dependent gene expression outside of the hippocampus (Honkaniemi and Sharp, 1999; Yamagata et al., 1994). Thus, we examined the performance of WT and Egr3−/− mice in cued associative fear conditioning, a task which depends upon normal amygdala function (Kim and Fanselow, 1992; Kim et al., 1993; Phillips and LeDoux, 1992). WT and Egr3−/− mice received weak (single tone-shock pairing) and strong (triple tone-shock pairing) training and then were evaluated for their freezing response to tone in a novel context (see methods). A significant effect of genotype on weak (F1,25 = 4.90, p < 0.05) and strong cued associated fear conditioning (F1,25 = 7.583, p < 0.01) was observed. Both WT and Egr3−/− mice showed significant freezing to auditory cue relative to naïve mice (p < 0.0001 and p < 0.01, respectively) but the cued freezing in Egr3−/− mice was significantly decreased relative to WT mice 24 hours after weak training (p < 0.01). Similarly, cued freezing was significantly increased in both WT and Egr3−/− mice relative to naïve mice 0.5 (p < 0.0001 and p < 0.001, respectively) and 24 hours (p < 0.0001 and p < 0.001, respectively) after strong training, but Egr3−/− mice showed significantly decreased cued freezing relative to WT mice 0.5 (p <0.01) and 24 hours (p < 0.01) after strong training (Fig. 6C). There was no significant difference between the frequency of freezing either 0.5 or 24 hours after strong training in Egr3−/− mice. Thus, unlike contextual memory in Egr3−/− mice, which was completely dissipated 24 hours after weak training and nearly so 24 hours after strong training, cued memory was significantly impaired but not completely abolished 24 hours after either weak or strong training.

LTP is impaired in Egr3−/− hippocampal CA3-CA1 synapses

Since Egr3 expression is regulated by synaptic activity in CA1 neurons (Li et al., 2005; Yamagata et al., 1994) and because Egr3−/− mice showed clear hippocampus-dependent learning and memory defects (Fig. 6A, B), we examined whether Egr3−/− mice also had disrupted hippocampal LTP, a form of synaptic plasticity thought to be critical for normal learning and memory processing. LTP was reliably induced in WT and Egr3−/− CA3-CA1 synapses by either Schaffer collateral HFS or TBS in hippocampal slices (Fig. 7A, B). However, LTP was significantly decreased relative to WT 40 minutes after Schaffer collateral TBS (WT = 179.8 ± 4.7%, Egr3−/− = 153.3 ± 8.2%, p < 0.02, Fig. 7A) and similarly, LTP was significantly decreased relative to WT 30 minutes (WT = 196.6 ± 8.4%, Egr3−/− = 132.1 ± 9.4%, p < 0.01), 90 minutes (WT = 184.2 ± 7.4%, Egr3−/− = 129.6 ± 10.0%, p < 0.01) and 210 minutes (WT = 176.7 ± 22.8%, Egr3−/− = 122.2 ± 10.5%, p < 0.05) after Schaffer collateral HFS (Fig. 7B–D). Thus, LTP induction in Schaffer collateral synapses is similar but rapidly decreases in Egr3−/− mice relative to WT using two different LTP induction paradigms.

Figure 7
Defective LTP in Schaffer collateral synapses in Egr3−/− mice. LTP was reliably induced in Schaffer collateral synapses using either (A) TBS or (B) HFS in WT and Egr3−/− hippocampal slices. However, LTP decreased in Egr3 ...

Discussion

Egr3 is a poorly studied member of the Egr family of transcriptional regulators that is structurally related to Egr1 (zif268). Like Egr1, Egr3 is regulated by synaptic activity in vivo and in excitatory cortical and hippocampal neurons by an NMDA receptor/MAPK-ERK signaling dependent mechanism (Li et al., 2005; Li et al., 2004; O’Donovan and Baraban, 1999; Yamagata et al., 1994). Several studies have indicated that Egr1 is essential for maintenance of late phase LTP, long-term memory in hippocampal dependent tasks and reconsolidation of previously memorized tasks (Bozon et al., 2003; Jones et al., 2001; Lee et al., 2004), but whether Egr3 has any role in learning and/or memory has not been previously examined. In this study, we showed that germline loss of Egr3 function has no obvious effect on the structural development of the brain, neuron architecture, the expression of several glutamatergic receptor subtypes, Erk phosphorylation, or Egr1 expression. Moreover, at least in the restricted context of Schaffer collateral (CA3-CA1) synaptic function, basal synaptic transmission and presynaptic plasticity are normal in Egr3−/− mice. However, there were defects in the magnitude of LTP induction in Schaffer collateral synapses and profound impairments in contextual fear conditioning and novel object discrimination, two memory tasks that depend upon normal hippocampal function. Since only germline Egr3−/− mice have been generated to date, it was not possible to exclude the possibility that some of the behavioral and physiologic abnormalities observed in Egr3 mutant mice may have been due to unrecognized developmental defects. Conditional mutagenesis of Egr3 will make it possible to selectively inactivate Egr3 in postnatal brain, after development is largely complete, and to bypass defects in muscle spindle morphogenesis that lead to abnormal proprioception in germline Egr3−/− mice (Tourtellotte and Milbrandt, 1998). In fact, the proprioception defects present in germline Egr3−/− mice precluded testing their behavior in other sensitive hippocampus dependent tasks that require intact motor skill, such as the Morris water maze. Nevertheless, the hippocampus and amygdala dependent behavioral tasks that could be performed show a clear role for Egr3 in at least some forms of learning and memory, despite the fact that the precise target genes regulated by Egr3 in neurons are still not known.

Egr1 and Egr3 are non-redundant synaptic activity dependent transcriptional regulators

Egr1 and Egr3 have highly homologous c-terminal zinc finger DNA binding domains that recognize similar sequences in target gene promoters, raising the possibility that Egr proteins may have overlapping functions under some conditions (Swirnoff and Milbrandt, 1995; Tourtellotte et al., 2000). Clearly though, Egr1 and Egr3 do not have completely redundant function since they have differing expression patterns and the targeted mutant mice have differing phenotypes. For example, Egr1-deficient mice have fertility defects due to abnormal regulation of luteinizing hormone β peptide not shared by Egr3-deficient mice (Lee et al., 1996; Topilko et al., 1998) and Egr3-deficient mice have abnormalities in proprioception due to defective muscle stretch receptor development not shared by Egr1-deficient mice (Tourtellotte and Milbrandt, 1998). However, since Egr1 and Egr3 are often co-expressed in similar neurons after intense synchronous synaptic activity (i.e., seizure), drug treatment or natural behavior, it has been suggested that they may have some functional overlap in the central nervous system (CNS) (Li et al., 2005; Yamagata et al., 1994). The extent to which Egr1 and Egr3 may have overlapping function by regulating similar activity dependent target genes in the CNS is not known, but there do appear to be some differences between these two molecules. For example, Egr3 appears to have higher basal levels of expression in brains of untreated mice, Egr3 protein is more stable than Egr1 and hence remains in neurons for longer periods after activity mediated induction, and Egr1 and Egr3 have divergent activation domains which could recruit different cofactors to activate or repress the promoters of a diverse repertoire of target genes (Li et al., 2005; O’Donovan et al., 1999). The results of this study clearly highlight some functional differences between Egr1 and Egr3 in the CNS since Egr3−/− and Egr1−/− mice have some clear differences in learning and memory related behavior and LTP.

A role for Egr3 in Schaffer collateral-CA1 LTP

Whereas both Egr1−/− and Egr3−/− mice have normal basal synaptic transmission and short term plasticity (paired pulse facilitation), LTP is differentially affected. In particular, Jones et al. reported that Egr1-deficient mice have intact early LTP (up to 60 minutes following tetanic stimulation) but absent late LTP (24–48 hours) in the dentate gyrus in vivo. By contrast, in Egr3−/− mice early LTP in CA1 is impaired relative to WT mice, a difference that occurs too rapidly to be explained by deficient LTP-induced Egr3 protein synthesis and impaired target gene regulation. Thus, Egr3 appears to be required prior to conditioning stimulation to regulate proteins that support WT levels of the early phase of LTP, whereas Egr1 is required to regulate gene expression required for late-phase but not early-phase LTP (Jones et al., 2001).

A role for Egr3 in hippocampal and amygdala dependent learning and memory

Egr1 and Egr3 also appear to have some differing roles in hippocampus and amygdala dependent memory. With contextual fear conditioning, Egr3−/− mice had no memory of the aversive context 24 hours after weak training and they had highly impaired memory after strong training compared to WT mice. Interestingly however, Egr3 −/− mice were able to elicit the freezing behavior and they could be conditioned to freeze to context similar to WT, mice but only after a high degree of reinforcement training. Thus, Egr3-mediated gene expression is not absolutely required for context associated fear conditioning but rather, it seems to be required for the process to occur efficiently. By contrast, previous studies demonstrated that Egr1−/− mice had no abnormalities in contextual fear conditioning after either weak or strong training in similar behavioral paradigms (Ko et al., 2005). Using the germline Egr3−/− mice it was not possible to determine how Egr3 protein that was present within neurons from neuronal activity prior to the learning task and Egr3 protein that was induced during the learning task contributed to task-dependent learning acquisition and/or memory retrieval. Additional molecular-genetic studies designed to manipulate Egr3 protein levels prior to acquisition, during consolidation and during retrieval will be required to more precisely define when Egr3 is required for certain stages of learning and/or memory processing. Nevertheless, since contextual fear memory was significantly dissipated in Egr3−/− relative to WT mice between 0.5 and 24 hours after strong training, Egr3 protein may be required to stabilize the memory over time. Similarly, Egr3−/− mice had no short-term (10 minute) or long-term (24 hour) object recognition memory, whereas Egr1−/− mice were previously reported to have normal short-term but impaired long-term recognition memory in a similar behavioral paradigm (Jones et al., 2001). Thus, Egr3 is required for some forms of both short-term and long-term hippocampus-dependent memory, whereas Egr1 appears to be primarily required for long-term hippocampus-dependent memory. In amygdala-dependent memory, the differences between Egr1 and Egr3 were again apparent. Whereas, Egr3−/− mice had impairments in cued fear conditioning after weak and strong training that became apparent 30 minutes after strong training, Egr1−/− mice were previously shown to have comparatively mild defects in cued fear conditioning that only became apparent 3 days after conditioning (Ko et al., 2005).

Future work focused on defining the target gene networks that are differentially regulated by Egr1 and Egr3, and a controlled side-by-side comparison of Egr1- and Egr3-deficient mice, will be important for understanding how this family of MAPK-ERK coupled transcriptional regulators function in learning and memory processing. These current results identify Egr3 as a learning and memory related transcriptional regulator of considerable significance, particularly since Egr1−/− mice have relatively mild impairments in learning and memory compared to Egr3−/− mice.

Experimental Methods

Animals

Egr3 heterozygous mice from a BL/6-129Sv/J hybrid strain were backcrossed 4 generations to BL/6 isogenic mice and maintained as an inbred strain. The mice were generated and progeny were genotyped as previously described (Tourtellotte and Milbrandt, 1998; Whitehead et al., 2005). All experimental procedures complied with protocols approved by the Northwestern University and University of Nebraska Institutional Animal Care and Use Committees.

Histology and immunohistochemistry

Deeply anesthetized adult WT and Egr3−/− mice were perfused through the heart with freshly prepared 4% paraformaldehyde/0.1 M phosphate buffer (pH=7.4). Immunohistochemistry was performed on free floating sections prepared from cryoprotected brain that had been equilibrated in 30% sucrose/PBS for 24 hours prior to freezing as previously described (Li et al., 2005). Primary antibodies that cross-reacted with NeuN (MAB377, Chemicon) to detect neurons, Synaptophysin (SC-9116, Santa Cruz) to detect synaptic terminals and Parvalbumin (MAB#235, Swant) to detect interneurons were used. After primary antibody incubation, the sections were treated with a biotinylated secondary antibody generated in the appropriate species and the antibody complexes were detected using avidin-biotin histochemistry with DAB as a chromagen, according to the manufacturer’s specifications (Vector Labs).

In vitro slice preparation and field EPSP recordings

Hippocampal slices (350 μm) from 3–4 month old male mice (TBS experiments) or male and female mice (HFS experiments) were prepared as previously described (Dong and Xiong, 2006; Yun et al., 2005). The animals were decapitated after deep anesthesia and their brains were placed in ice-cold artificial cerebrospinal fluid (ACSF; 124 mM NaCl, 3 mM KCl, 2.4 mM CaCl2, 1.3 mM MgSO4, 1.25 mM NaH2PO4, 26 mM NaHCO3, 10 mM glucose and saturated with 95% O2/5% CO2 to a final pH=7.4). Vibratome slices were equilibrated for ≥2 hr. at room temperature in ACSF and were transferred as needed to a submersion chamber maintained at 32°C. Field excitatory postsynaptic potentials (fEPSPs) were recorded in the stratum radiatum of area CA1 in the hippocampus in response to electrical stimulation of the Schaffer collateral/commissural axons. Each fEPSP was an average of 5 (TBS protocol) or 3 (HFS protocol) consecutive sweeps generated in response to 0.05 Hz test pulses. Electrophysiological measures of basal synaptic transmission included input/output (I/O) relationships and short-term presynaptic plasticity (paired pulse facilitation, PPF). I/O relationships were measured by comparing EPSP amplitudes as a function of stimulus intensity using a pulse width of 0.1 msec and a current range of 20–250 μA. The current that produced approximately 35% of maximal responses (usually 20–40 μA) was used throughout all subsequent experiments for monitoring test pulses. The same current was used to evoke PPF which was calculated as a percentage of EPSP slopes (pulse 2/pulse 1) in response to pairs of stimuli delivered with interpulse intervals of 20 msec, 50 msec and 100 msec. The data were sampled at 10 kHz and low-pass filtered at 3 kHz. The baseline EPSP amplitude was recorded in response to 0.05 Hz test pulses for 10 minutes after which LTP was induced by applying either TBS which consisted of 10 bursts with an inter-burst interval of 200 ms (5 Hz) or HFS which consisted of 2 tetani of 1 second each at 100 Hz and with a 20 second inter-tetanus interval. During TBS and HFS, the stimulus intensity was not varied. The magnitude of LTP was expressed as the percentage of baseline and compared between groups of WT and Egr3−/− mice at 40 minutes after TBS and 30 minutes, 90 minutes and 210 minutes after HFS. All responses were normalized to the averaged 10 minute (TBS) or 30 minute (HFS) baseline measurements to allow comparison among slices. pClamp v. 8.0 (Axon Instruments, U.S.A.) was used for data collection and analysis.

Neuron counts

Neurons from defined anatomical regions of the: dorsal (d) and ventral (v) blade of the dentate gyrus as well as area CA1 and CA3 pyramidal neurons of the hippocampus were counted in 3 WT and 3 Egr3−/− adult brains (8–12 week old). The mice were perfused through the heart with 0.1 M PB, pH=7.4 followed by 4% paraformaldehyde/0.1 M PB, pH=7.4. The brains were embedded in paraffin and serial 6 μm coronal sections were generated and counterstained with hematoxylin and eosin. For each brain, neurons were counted from 4 consecutive sections each separated by 40 μm. The sections spanned a total rostral-caudal distance of 160 μm including and caudal to coordinates: interaural, 2.34 mm; bregma -1.46 mm (Paxinos and Franklin, 2001). The number of neurons within each analyzed region of the hippocampus was estimated using a statistically unbiased optical dissector protocol (StereoInvestigator, Microbrightfield). The neuron density (number of neurons per unit volume) was compared between WT and Egr3−/− brains for each of the areas of the hippocampus analyzed.

Golgi silver impregnation

Golgi-Cox silver impregnation of neurons was performed using commercially available reagents (FD Rapid Golgi Stain Kit, NeuroTechnologies). Twelve-week-old freshly dissected mouse brains were immersed in silver impregnation solution for 2 weeks, transferred to Solution C for 2 days, and cut into 100 μm sections on a cryostat. The sections were mounted on 3% gelatin-coated glass slides, air-dried for 1 week and then stained with Solution D and E followed by alcohol dehydration and cover-slip mounting with Permount.

Spine counts

Dendritic spines were counted in Golgi impregnated CA1 neurons using oil-immersion and 1000× total magnification. Spines were counted along a 40 μm distance of CA1 neuron secondary dendrites starting 10–30 μm from their point of origin from the primary dendrite. The number of individual spines in 40 μm segments of secondary dendrites were counted in 40 CA1 neurons from each of 3 WT and 3 Egr3−/− brains. The results were expressed as the number of spines per μm length of secondary dendrite.

Behavioral Studies

Conventional fear conditioning and novel object discrimination were performed as previously described with some modification (Jones et al., 2001; Radulovic et al., 1998). WT and Egr3−/− littermate male and female mice ranging in age from 8–12 weeks were tested.

Fear conditioning

Fear conditioning was performed using a computer-controlled activity monitor and shock/tone generator (TSE, Bad Homburg, Germany). The mice were exposed to two contexts: Context 1 consisted of a Plexiglas chamber (35 × 20 × 20 cm) inside of which was placed a removable shock grid made of stainless steel rods (4 mm diameter with 0.9 cm spacing). The shock grid was connected to a constant current power supply set to deliver a 2-second 0.7 mA shock. The shock grid and chamber were cleaned with 70% aqueous ethanol prior to testing each animal. Context 2 consisted of a Plexiglas chamber (35 × 20 × 20 cm) that was exposed to a constant 20-watt light source and no shock grid. The chamber was washed with 1% aqueous acetic acid before each mouse was placed into the chamber. A signal generator connected to a loudspeaker (Conrad, KT-25-DT) produced a 10 kHz, 75 dB tone and the locomotor activity (cm/s) of each mouse was automatically recorded by an infrared beam detector array in each of the experimental context chambers.

Training

Mice were subjected to two training conditions: Weak training (single conditioning trial) which consisted of a 3 minute exposure to context 1, followed by a tone and a single foot shock; Strong training (multiple conditioning trial) consisted of a 30 second exposure to context 1, followed by a tone and a foot shock with context and tone/shock pairing repeated 3 times.

Contextual fear conditioning

Contextual memory was tested 24 hours after weak training, and 0.5 and 24 hours after strong training. Testing was performed by re-exposing each mouse for 3 minutes to context 1 without the administration of either shock or tone.

Cued associative fear conditioning

Each mouse was exposed to context 2 and freezing response was measured for 3 minutes after presentation of a tone. For both contextual and cued associative fear conditioning, freezing was defined as a lack of movement associated with a crouching posture. Freezing behavior was assessed at the end of every 10 second interval by two trained observers during the 3 minute trial period (a total of 18 assessments were made during the 3 minute trial). The data were recorded as a fraction (percent) of the total measurements scored as freezing for each trial.

Novel object discrimination

The natural behavior of WT and Egr3−/− mice was observed in an open 25 × 22 × 13 cm field (context) which contained two objects. Each mouse was habituated to the context and objects for 10 minutes on day 1 by allowing them to freely explore the area and the objects. Further habituation was performed on days 2 and 3, which consisted of two 10 minute trials of context and object exploration interrupted by a 10 minute period in the home cage. On day 4 each mouse was subjected to a single 10 minute training session followed by testing. Testing was performed 10 minutes and 24 hours after the training session by replacing one of the familiar (F) objects with a novel (N) object (a different novel object was used for the 10 minute test and the 24 hour test). Measurements were made by recording the amount of time each mouse spent with both forelimbs oriented toward and within 4 cm of either object. The data were reported as the mean total time (sec.) spent exploring the novel and familiar object during the 3 minute trial period for WT and Egr3−/− mice.

The results obtained from the performance of male and female mice on a particular task were pooled since there was no gender affect for strong contextual fear conditioning (F1, 72 = 0.454, p = 0.50), strong cued associative fear conditioning (F1,72 = 2.95, p =0.1), weak contextual fear conditioning (F1,48 = 2.39, p = 0.14), weak cued associative fear conditioning (F1,48 = 0.20, p = 0.65) or novel object discrimination (F1,60 = 0.23, p = 0.64).

Shock threshold testing

Adult WT (N=6) and Egr3−/− (N=5) mice were placed inside the fear conditioning chamber with a removable shock grid floor. Individual mice were subjected to 2-second shocks with intensities ranging from 0.1 mA to 0.8 mA in 0.1 mA increments, similar to that described previously (Bourtchuladze et al., 1994). At each shock intensity, the animals were scored for their involuntary flinching or overt vocalization responses to shock. The lowest current intensity to elicit the particular behavior was recorded and the mean response thresholds were compared between WT and Egr3−/− mice.

Nociception testing

Nociception was tested by placing adult WT (N=5) and Egr3−/− (N=5) mice on a 55 °C hotplate and the latency to lick their hindpaw was measured.

Non-associative learning (habituation) testing

Adult WT (N=4) and Egr3−/− (N=3) mice were placed in a 35 × 20 × 20 cm context chamber and their activity during a 3 minute training period was recorded by infrared beam crossing instrumentation (TSE, Bad Homburg, Germany). The training was repeated for each mouse for 5 consecutive days and the mean activity was recorded and compared over time between WT and Egr3−/− mice.

Context discrimination

Adult WT (N=13) and Egr3−/− (N=11) mice were tested for their ability to discriminate between two different contexts. Mice received weak contextual fear conditioning and were tested 24 hours after training. The testing was performed by re-exposing the mice to the original context (context 1) that was used during the context-shock fear conditioning and their freezing behavior was recorded during a 3 minute trial. Subsequently, the mice were exposed to a novel context (context 2) for 3 minutes during which time freezing was measured.

Freezing test

To examine whether Egr3−/− mice can produce the expected freezing behavior after shock and to evaluate their capacity for associative learning, adult WT (N=5) and Egr3−/− (N=5) mice were exposed to a highly reinforced (overtaining) fear conditioning protocol similar to that previously reported for testing PKC-β-deficient mice (Weeber et al., 2000). The mice were placed in the fear conditioning chamber containing the shock grid for 1 minute, followed by a 30 second acoustic tone (10 kHz, 75 dB) and a 2-second 0.7 mA shock at the end of the tone. Freezing behavior was measured during this 90 second context-shock interval and the process was repeated 7 times during Day 1. The entire procedure was repeated again on Day 2.

Quantitative Western blotting

Microdissected hippocampus, amygdala and somatosensory cortex were collected from WT and Egr3−/− mice and lysed in RIPA buffer to collect total cellular protein as previously described (Li et al., 2005). Protein samples (100 μg/lane) were electrophoresed on 10% SDS poly-acrylamide gels and the protein was immobilized onto PVDF membranes (Immobilon-P, Millipore). The membranes were incubated at room temperature 1 hour with specific antibodies to: Egr3 (sc-191), NR1 (sc-9058), Erk-1/2 (sc-94), p-Erk1/2 (sc-16982) all from Santa Cruz Biotechnology, Santa Cruz, CA, NR2A (JH1817, R. Huganir, Johns Hopkins University, Baltimore, MD), NR2B (JH1741, R. Huganir, Johns Hopkins University, Baltimore, MD), GluR1 (AB1540, Chemicon, Temecula, CA), GluR2 (JH1707, R. Huganir, Johns Hopkins University, Baltimore, MD), mGluR5 (AB5675, Chemicon, Temecula, CA), and β-actin (A2228, Sigma-Aldrich, St. Louis, MO), followed by incubation with alkaline phosphatase-conjugated Goat anti-rabbit antibody (Jackson Immunoresearch) and visualized using enhanced chemiluminescence (ECL) (CDP-star, Tropix) on autoradiography film (Amersham hyperfilm ECL). To ensure that the ECL readout was within the linear detection range of the film, standard curves were generated by systematically varying the protein concentration of samples and subjecting them to ECL and densitometry. The blots probed for the proteins of interest were then processed for differing lengths of time in ECL and densitometry measurements were performed using the computer software Wincam 2.2 (Cybertech, Berlin, Germany) and ensuring that the densitometry readings fell within the linear range of the standard curves. Background signals were subtracted from a film region close to the bands of interest. The densitometric values were normalized to β-actin to control for variation in protein loading.

Statistical Analyses

All data were analyzed by two-way ANOVA using Genotype (WT and Egr3−/−) and Test (e.g., baseline, short-term memory and long-term memory) as grouping factors. Post-hoc analyses were performed using Scheffe’s test for multiple comparisons. Since all WT animals learned the particular task, comparisons grouped by Test were all significant: object discrimination (F2,60 = 3.704, p < 0.05), strong contextual fear conditioning (F2,75 = 43.968, p < 0.001), strong cued associative fear conditioning (F2,75 = 31.270, p < 0.0001), weak contextual fear conditioning (F2,75 = 36.925, p < 0.0001) and weak cued associative fear conditioning (F2,75 = 37.828, p < 0.0001). All values were expressed as mean ± SEM and when individual groups were compared Student’s unpaired t-test was used to compare the means. The I/O correlations and PPF were analyzed using repeated measures ANOVA with Genotype as the grouping variable. In all instances p < 0.05 was considered statistically significant.

Acknowledgments

We thank P. Penzes and R. Huganir for antibody reagents. We thank J. Disterhoft and R. Miller for comments on the manuscript, and members of the Tourtellotte lab for helpful discussion and advice. Technical assistance was provided by J. Whitehead. This study was supported by The National Institutes of Health (NS046468 and NS040748) and a Howard Hughes Faculty Scholar Award to W.G.T.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • Atkins CM, Selcher JC, Petraitis JJ, Trzaskos JM, Sweatt JD. The MAPK cascade is required for mammalian associative learning. Nat Neurosci. 1998;1:602–609. [PubMed]
  • Blum S, Moore AN, Adams F, Dash PK. A mitogen-activated protein kinase cascade in the CA1/CA2 subfield of the dorsal hippocampus is essential for long-term spatial memory. J Neurosci. 1999;19:3535–3544. [PubMed]
  • Bourtchuladze R, Frenguelli B, Blendy J, Cioffi D, Schutz G, Silva AJ. Deficient long-term memory in mice with a targeted mutation of the cAMP-responsive element-binding protein. Cell. 1994;79:59–68. [PubMed]
  • Bozon B, Davis S, Laroche S. A requirement for the immediate early gene zif268 in reconsolidation of recognition memory after retrieval. Neuron. 2003;40:695–701. [PubMed]
  • Clark RE, Zola SM, Squire LR. Impaired recognition memory in rats after damage to the hippocampus. J Neurosci. 2000;20:8853–8860. [PubMed]
  • Cole AJ, Saffen DW, Baraban JM, Worley PF. Rapid increase of an immediate early gene messenger RNA in hippocampal neurons by synaptic NMDA receptor activation. Nature. 1989;340:474–476. [PubMed]
  • Coogan AN, O’Leary DM, O’Connor JJ. P42/44 MAP kinase inhibitor PD98059 attenuates multiple forms of synaptic plasticity in rat dentate gyrus in vitro. J Neurophysiol. 1999;81:103–110. [PubMed]
  • Davis S, Vanhoutte P, Pages C, Caboche J, Laroche S. The MAPK/ERK cascade targets both Elk-1 and cAMP response element-binding protein to control long-term potentiation-dependent gene expression in the dentate gyrus in vivo. J Neurosci. 2000;20:4563–4572. [PubMed]
  • Dong J, Xiong H. Human immunodeficiency virus type 1 gp120 inhibits long-term potentiation via chemokine receptor CXCR4 in rat hippocampal slices. J Neurosci Res. 2006;83:489–496. [PubMed]
  • English JD, Sweatt JD. Activation of p42 mitogen-activated protein kinase in hippocampal long term potentiation. J Biol Chem. 1996;271:24329–24332. [PubMed]
  • English JD, Sweatt JD. A requirement for the mitogen-activated protein kinase cascade in hippocampal long term potentiation. J Biol Chem. 1997;272:19103–19106. [PubMed]
  • Honkaniemi J, Sharp FR. Prolonged expression of zinc finger immediate-early gene mRNAs and decreased protein synthesis following kainic acid induced seizures. Eur J Neurosci. 1999;11:10–17. [PubMed]
  • Jones MW, Errington ML, French PJ, Fine A, Bliss TV, Garel S, Charnay P, Bozon B, Laroche S, Davis S. A requirement for the immediate early gene Zif268 in the expression of late LTP and long-term memories. Nat Neurosci. 2001;4:289–296. [PubMed]
  • Kim JJ, Fanselow MS. Modality-specific retrograde amnesia of fear. Science. 1992;256:675–677. [PubMed]
  • Kim JJ, Rison RA, Fanselow MS. Effects of amygdala, hippocampus, and periaqueductal gray lesions on short- and long-term contextual fear. Behav Neurosci. 1993;107:1093–1098. [PubMed]
  • Ko SW, Ao HS, Mendel AG, Qiu CS, Wei F, Milbrandt J, Zhuo M. Transcription factor Egr-1 is required for long-term fear memory and anxiety. Sheng Li Xue Bao. 2005;57:421–432. [PubMed]
  • Lee JL, Everitt BJ, Thomas KL. Independent cellular processes for hippocampal memory consolidation and reconsolidation. Science. 2004;304:839–843. [PubMed]
  • Lee SL, Sadovsky Y, Swirnoff AH, Polish JA, Goda P, Gavrilina G, Milbrandt J. Luteinizing hormone deficiency and female infertility in mice lacking the transcription factor NGFI-A (Egr-1) Science. 1996;273:1219–1221. [PubMed]
  • Li L, Carter J, Gao X, Whitehead J, Tourtellotte WG. The Neuroplasticity-Associated Arc Gene Is a Direct Transcriptional Target of Early Growth Response (Egr) Transcription Factors. Mol Cell Biol. 2005;25:10286–10300. [PMC free article] [PubMed]
  • Li L, Carter J, Tourtellotte WG. Transcriptional regulation of the plasticity associated ARC gene by early growth reponse (Egr) transcription factors. Soc Neurosci Astr 636. 2004:16.
  • Link W, Konietzko U, Kauselmann G, Krug M, Schwanke B, Frey U, Kuhl D. Somatodendritic expression of an immediate early gene is regulated by synaptic activity. Proc Natl Acad Sci U S A. 1995;92:5734–5738. [PMC free article] [PubMed]
  • Lyford GL, Yamagata K, Kaufmann WE, Barnes CA, Sanders LK, Copeland NG, Gilbert DJ, Jenkins NA, Lanahan AA, Worley PF. Arc, a growth factor and activity-regulated gene, encodes a novel cytoskeleton-associated protein that is enriched in neuronal dendrites. Neuron. 1995;14:433–445. [PubMed]
  • Milner B, Squire LR, Kandel ER. Cognitive neuroscience and the study of memory. Neuron. 1998;20:445–468. [PubMed]
  • O’Donovan KJ, Baraban JM. Major Egr3 isoforms are generated via alternate translation start sites and differ in their abilities to activate transcription. Mol Cell Biol. 1999;19:4711–4718. [PMC free article] [PubMed]
  • O’Donovan KJ, Tourtellotte WG, Millbrandt J, Baraban JM. The EGR family of transcription-regulatory factors: progress at the interface of molecular and systems neuroscience. Trends Neurosci. 1999;22:167–173. [PubMed]
  • Paxinos G, Franklin KBJ. The mouse brain in stereotaxic coordinates. Academic Press; San Diego: 2001.
  • Petersohn D, Schoch S, Brinkmann DR, Thiel G. The human synapsin II gene promoter. Possible role for the transcription factor zif268/egr-1, polyoma enhancer activator 3, and AP2. J Biol Chem. 1995;270:24361–24369. [PubMed]
  • Phillips RG, LeDoux JE. Differential contribution of amygdala and hippocampus to cued and contextual fear conditioning. Behav Neurosci. 1992;106:274–285. [PubMed]
  • Radulovic J, Kammermeier J, Spiess J. Generalization of fear responses in C57BL/6N mice subjected to one-trial foreground contextual fear conditioning. Behav Brain Res. 1998;95:179–189. [PubMed]
  • Roberts DS, Raol YH, Bandyopadhyay S, Lund IV, Budreck EC, Passini MA, Wolfe JH, Brooks-Kayal AR, Russek SJ. Egr3 stimulation of GABRA4 promoter activity as a mechanism for seizure-induced up-regulation of GABA(A) receptor alpha4 subunit expression. Proc Natl Acad Sci U S A. 2005;102:11894–11899. [PMC free article] [PubMed]
  • Rosenblum K, Futter M, Jones M, Hulme EC, Bliss TV. ERKI/II regulation by the muscarinic acetylcholine receptors in neurons. J Neurosci. 2000;20:977–985. [PubMed]
  • Segal M. Dendritic spines and long-term plasticity. Nat Rev Neurosci. 2005;6:277–284. [PubMed]
  • Swirnoff AH, Milbrandt J. DNA-binding specificity of NGFI-A and related zinc finger transcription factors. Mol Cell Biol. 1995;15:2275–2287. [PMC free article] [PubMed]
  • Thiel G, Schoch S, Petersohn D. Regulation of synapsin I gene expression by the zinc finger transcription factor zif268/egr-1. J Biol Chem. 1994;269:15294–15301. [PubMed]
  • Thomas GM, Huganir RL. MAPK cascade signalling and synaptic plasticity. Nat Rev Neurosci. 2004;5:173–183. [PubMed]
  • Topilko P, Schneider-Maunoury S, Levi G, Trembleau A, Gourdji D, Driancourt MA, Rao CV, Charnay P. Multiple pituitary and ovarian defects in Krox-24 (NGFI-A, Egr-1)-targeted mice. Mol Endocrinol. 1998;12:107–122. [PubMed]
  • Tourtellotte WG, Keller-Peck C, Milbrandt J, Kucera J. The transcription factor Egr3 modulates sensory axon-myotube interactions during muscle spindle morphogenesis. Dev Biol. 2001;232:388–399. [PubMed]
  • Tourtellotte WG, Milbrandt J. Sensory ataxia and muscle spindle agenesis in mice lacking the transcription factor Egr3. Nat Genet. 1998;20:87–91. [PubMed]
  • Tourtellotte WG, Nagarajan R, Bartke A, Milbrandt J. Functional compensation by Egr4 in Egr1-dependent luteinizing hormone regulation and Leydig cell steroidogenesis. Mol Cell Biol. 2000;20:5261–5268. [PMC free article] [PubMed]
  • Weeber EJ, Atkins CM, Selcher JC, Varga AW, Mirnikjoo B, Paylor R, Leitges M, Sweatt JD. A role for the beta isoform of protein kinase C in fear conditioning. J Neurosci. 2000;20:5906–5914. [PubMed]
  • Whitehead J, Keller-Peck C, Kucera J, Tourtellotte WG. Glial cell-line derived neurotrophic factor-dependent fusimotor neuron survival during development. Mech Devel. 2005;122:27–41. [PubMed]
  • Yamagata K, Kaufmann WE, Lanahan A, Papapavlou M, Barnes CA, Andreasson KI, Worley PF. Egr3/Pilot, a zinc finger transcription factor, is rapidly regulated by activity in brain neurons and colocalizes with Egr1/zif268. Learn Mem. 1994;1:140–152. [PubMed]
  • Yun SH, Park KA, Sullivan P, Pasternak JF, Ladu MJ, Trommer BL. Blockade of nicotinic acetylcholine receptors suppresses hippocampal long-term potentiation in wild-type but not ApoE4 targeted replacement mice. J Neurosci Res. 2005;82:771–777. [PubMed]
  • Zola SM, Squire LR, Teng E, Stefanacci L, Buffalo EA, Clark RE. Impaired recognition memory in monkeys after damage limited to the hippocampal region. J Neurosci. 2000;20:451–463. [PubMed]
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...