This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Li, L. Articles by Tourtellotte, W. G. Search for Related Content PubMed PubMed Citation Articles by Li, L. Articles by Tourtellotte, W. G.

The Neuroplasticity-Associated Arc Gene Is a Direct Transcriptional Target of Early Growth Response (Egr) Transcription Factors

Lin Li,^1, John Carter,^1, Xiaoguang Gao,¹ Jennifer Whitehead,¹ and Warren G. Tourtellotte^1,2,3*

Departments of Pathology (Neuropathology),¹ Neurology,² The Institute for Neuroscience, Northwestern University, Chicago, Illinois 60611³

Received 3 August 2005/ Accepted 2 September 2005

	ABSTRACT

Top Abstract Introduction Materials and Methods Results Discussion References

 
Early growth response (Egr) transcription factors (Egr1 to Egr4) are synaptic activity-inducible immediate early genes (IEGs) that regulate some aspects of synaptic plasticity-related to learning and memory, yet the target genes regulated by them are unknown. In particular, Egr1 is essential for persistence of late-phase long-term potentiation (L-LTP), for hippocampus-dependent long-term memory formation, and for reconsolidation of previously established memories. Here, we show that Egr1 and Egr3 directly regulate the plasticity-associated activity-regulated cytoskeletal-related (Arc) gene, a synaptic activity-induced effector molecule which is also required for L-LTP and hippocampus-dependent learning and memory processing. Moreover, Egr1-deficient and Egr3-deficient mice lack Arc protein in a subpopulation of neurons, while mice lacking both Egr1 and Egr3 lack Arc in all neurons. Thus, Egr1 and Egr3 can indirectly modulate synaptic plasticity by directly regulating Arc and the plasticity mechanisms it mediates in recently activated synapses.

	INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

 
Neurons process and retain information by establishing networks of synaptic connections that are modified by repeated activation. These "plasticity" changes are mediated by synaptic activity-regulated intraneuronal signal transduction pathways, some of which induce new gene expression and protein synthesis. Immediate early genes (IEGs) are rapidly induced as the earliest genomic response to synaptic activity (6). IEGs encode both transcriptional regulators and direct effector molecules, such as structural proteins, signaling enzymes, and growth factors that participate in synaptic and structural alterations made by neurons in response to their activity-dependent history (8). Transcriptional regulatory IEGs further diversify neuron responses to activation by broadening the regulatory control over long-term plasticity changes involved in complex processes, such as learning and memory. However, the precise molecular mechanisms mediated by activity-induced transcriptional regulators are poorly defined because the target genes they regulate are unknown.

The early growth response (Egr) family of activity induced IEGs encodes four zinc-finger transcription factors that are designated Egr1 (also known as zif268, NGFI-A, Krox24, Tis8, and ZENK), Egr2 (also known as Krox20), Egr3 (also known as Pilot), and Egr4 (also known as NGFI-C and PAT133) (20). They have nearly identical zinc finger DNA binding domains near their C terminus, they are often coregulated by similar stimuli in many cell types (including neurons), and they can transactivate gene expression by binding a GC-rich motif (Egr response element [ERE]) located in the promoter regions of target genes (31). In neurons, Egr genes are robustly and transiently expressed by elevated cytosolic calcium produced by synaptic activity (7, 18), and their expression is coupled with N-methyl D-aspartate (NMDA) receptor activation and intracellular mitogen-activated protein kinase (MAPK) signaling in excitatory glutamatergic synapses, by either abnormal (seizure induced) or physiologic synaptic activity. A critical role for the archetypal family member Egr1 (zif268) in regulating some neuroplasticity responses has been clearly established. For example, early studies demonstrated a strong correlation between Egr1 gene expression and stabilization of long-term potentiation (LTP) (a particular form of synaptic plasticity) in response to appropriate-patterned synaptic activity (1, 40). Recently, studies using Egr1-deficient mice and antisense oligonucleotides have confirmed a critical role for Egr1 in LTP persistence (late-phase LTP [L-LTP]), long-term memory in hippocampus-dependent tasks (12) and in reconsolidation of previously memorized tasks (5, 13). Thus, despite the fact that Egr1 has a clear role in mediating gene expression required for some learning and memory processes, the specific molecular mechanisms involved are poorly defined because the target genes that Egr1 regulates in the brain have not yet been identified. Moreover, comparatively little is known about the potential role of other Egr transcription factors in plasticity-related mechanisms, although they are regulated by similar intracellular signal transduction pathways and they may also mediate some forms of neuroplasticity, such as stabilization of LTP (38, 41).

In a previous study, we identified the activity-regulated cytoskeletal associated (Arc) gene as a potential target gene regulated by Egr transcription factors (2). Arc (also known as Arg3.1) is a particularly interesting effector plasticity-associated molecule because, like Egr transcription factors, it is rapidly induced by synaptic activity (15, 17), its expression depends upon excitatory synaptic NMDA receptor activation (28) and intracellular MAPK signaling (36), and it has a critical role in maintaining LTP and long-term, but not short-term, memory formation (9). In activated neurons, Arc mRNA is rapidly distributed into dendrites, where it is targeted to recently activated synapses and locally translated into protein, suggesting that it has a direct role in the plasticity responses made within synapses after activation (26, 27). Moreover, similar to Egr transcription factors, the Arc gene is regulated as an immediate early gene that is rapidly upregulated without the need for new protein synthesis. However, the Arc gene is also regulated as a delayed target gene by a poorly understood protein synthesis-dependent mechanism that appears to modulate its expression beyond the brief transient immediate early phase (35).

Based upon evidence that the Arc gene may be regulated as a delayed target gene and that there are numerous similarities between Egr and Arc gene regulation and function related to memory processing, we examined whether Egr transcription factors are necessary for regulating Arc during its protein synthesis-dependent phase of expression. We found that enforced expression of Egr1 and Egr3 in myotubes and primary hippocampal and cortical neurons was sufficient to induce Arc expression by directly binding and transactivating the Arc promoter through a species-conserved high-affinity ERE in its proximal promoter. Accordingly, when Egr1 and Egr3 proteins were coinduced by seizure-related synaptic activity in vivo, the protein synthesis-dependent phase of Arc expression required Egr3. However, during physiologic synaptic activity induced by a novel environment, both Egr1 and Egr3 were required to induce normal levels of Arc expression in the brain and in mice lacking both Egr1 and Egr3 proteins, Arc protein was absent in activated neurons.

Egr1 and potentially Egr3 are important transcriptional regulators of some plasticity changes associated with learning and memory but to what extent they regulate genes capable of modifying synaptic function or influencing the structural arrangement of synapses is not known. These results identify the Arc gene as the first known direct Egr target gene that has a clear role in mediating some forms of synaptic plasticity during learning and memory processing. Hence, Egr1 and Egr3 can indirectly regulate the physiologic properties of recently activated synapses by directly modulating Arc gene expression.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

 
Animals and experimental treatments. (i) Animals. Egr1-deficient and Egr3-deficient mice were generated and genotyped as previously described (14, 32, 34, 37). Egr1-deficient mice were backcrossed 10 generations to C57BL/6J mice and Egr3-deficient mice were backcrossed four generations to C57BL/6J mice. Littermate wild-type mice were used as controls for all experiments in which comparisons were made between wild-type and Egr gene-deficient mice. All experimental procedures complied with protocols approved by the Northwestern University Institutional Animal Care and Use Committee.

(ii) Kainic acid-mediated seizure induction. Seizures were induced by kainic acid (KA) (20 mg/kg; Sigma, St. Louis, MO) intraperitoneal injection (i.p.) in adult mice (>15 weeks of age). Control mice received phosphate-buffered saline (PBS) injections, and some experimental mice received PBS and some received cycloheximide (CHX) (120 mg/kg; Sigma, St. Louis, MO) 15 min prior to KA administration to inhibit protein synthesis. Seizures were induced in all KA treated mice which were sacrificed 2 or 4 h after seizure or PBS treatment. No differences in the intensity or duration of the seizures were noted between wild-type and mutant mice.

(iii) Exposure to novel environment. To examine the role of Egr genes in regulating Arc expression during physiologic synaptic activity, 21-day-old (P21) wild-type, Egr1-deficient, Egr3-deficient, and Egr1/Egr3 double-deficient (Egr1/3 dKO) mice were subjected to a novel environment exploration paradigm to induce Egr and Arc gene expression by increasing physiologic synaptic activity. The exploration paradigm was similar to that previously reported to induce Arc gene expression and consisted of an open square box measuring 61 by 61 cm with 20-cm-high walls (23). The box was partitioned into a three by three grid with seven identical three-dimensional objects fixed in place. The box was located in a room with several landmarks on the walls and ceiling. Wild-type and knockout mice were removed from their littermates and allowed to explore the environment for 5 min. The mice were placed in a new grid every 15 s to maintain a novel environment and to facilitate exploration. Each grid was visited two or three times during the exploration session, and there was no obvious difference in the exploratory behaviors exhibited by wild-type or mutant mice. Following the exploration session each mouse was returned to a cage in isolation. Four hours after exposure to a novel environment, the mice were anesthetized and the brains were processed as indicated (below) for in situ hybridization analysis and immunohistochemistry.

Tissue preparation. Anesthetized (ketamine, 150 mg/kg; xylazine 10 mg/kg; i.p.) P21 and adult wild-type and Egr gene-deficient mice were examined. In some mice, brains were removed fresh, microdissected to obtain dorsal hippocampus (coordinates were as follows: interaural, 2.34 mm; bregma, 1.46 mm) (22) and trunk/barrel somatosensory cortex (coordinates were as follows: interaural, 2.34 mm, bregma, 1.46 mm) (22), snap frozen in liquid nitrogen, and stored at –80°C for subsequent nuclear protein, total protein, or RNA preparation. For in situ hybridization and immunohistochemical analysis, anesthetized mice received intracardiac phosphate-buffered saline (0.1 M, pH 7.2) followed by aldehyde perfusion (4% paraformaldehyde, 0.1 M phosphate buffer [PB], pH 7.4). The brains were removed, midsagitally hemisected, cryoprotected in 30% sucrose-PBS overnight, frozen in OCT embedding media (VWR, Westchester, PA), and stored at –80°C. The brains were cut on a freezing microtome into four parallel series of 8- to 15-µm-thick sections.

In situ hybridization. In situ hybridization was performed using digoxigenin-labeled riboprobes with slight modification to previously published protocols (2). The probes were generated by PCR, and the amplicons were polished (End-it; Epicenter, Madison, WI), subcloned into the EcoRV site of Bluescript (Stratagene, La Jolla, CA) and sequence verified. Sense and antisense riboprobes were synthesized using in vitro transcription and digoxigenin-labeled UTP. The sense probe was used on parallel tissue sections as a control for nonspecific labeling and in all cases the sense probes showed no hybridization signal. The probes used for in situ hybridization spanned coding sequences as follows: for the Egr3 gene (GenBank NM_018781), nucleotides [nt] 890 to 1389; Egr1 gene (GenBank NM_007913), nt 1589 to 2088; and Arc gene (GenBank NM_018790), nt 577 to 1082. In all cases in which gene expression comparisons were made between treatment conditions or genotypes, a complete set of tissue sections was processed together and the chromogen reactions were terminated simultaneously. All experiments were repeated three or four times to ensure consistency and accurate interpretation. Representative photographs were obtained with a Spot digital camera (Diagnostic Instruments, Sterling Heights, MI) attached to a Nikon E600 microscope and by using the Spot image acquisition software. The digital images were acquired using the internal light meter to determine camera shutter speed and no additional manipulations were applied to the acquired images.

Immunohistochemistry. Immunohistochemistry was performed on paraformaldehyde-fixed and frozen tissue sections. Tissue sections were incubated with blocking buffer (3% normal serum-0.1% Triton X-100 in PBS) for 1 h at room temperature. The sections were incubated in primary antibody diluted in blocking buffer overnight at 4°C. The sections were washed in PBS/0.1% Triton X-100 and incubated with appropriate fluorescent secondary antibodies (Cy3 or Cy5, Jackson Immunoresearch) for 1 h at room temperature. The following antibodies were used: rabbit anti-Egr3 antibody (sc-191 1:1000), rabbit anti-Egr1 antibody (sc-110, 1:1,000), rabbit anti-Arc (sc-15325, 1:1,000), mouse anti-Arc (sc-17839, 1:100) (all from Santa Cruz Biotechnology, Santa Cruz, CA), Cy3-conjugated goat anti-rabbit and Cy5-conjugated goat anti-rabbit antibody (both from Jackson Immunoresearch, West Grove, PA). Fluorescent images were captured with a Zeiss LS510 confocal microscope with identical aperture and photomultiplier tube voltage settings used for all of the acquired images. In some cases unmanipulated images were assembled into montages with Photoshop (Adobe) to use for neuron counting.

Western blotting. Microdissected hippocampi from adult wild type, Egr1-deficient and Egr3-deficient mice receiving either KA to induce seizures or PBS were homogenized in RIPA buffer (20 mM Tris, pH 7.7, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) supplemented with 10 nM NaF and Complete protease inhibitors (Roche, Alameda, CA). Eighty micrograms of total cellular protein was resolved by SDS-PAGE on a 10% acrylamide gel and transferred to a nitrocellulose membrane. Blots were probed with anti-Egr1 (sc-110), anti-Egr3 (sc-191), anti-Arc (sc-17839), or antiactin (sc-1616) antibodies, all from Santa Cruz Biotechnology, Santa Cruz, CA, and then with an horseradish peroxidase-conjugated goat anti-rabbit secondary antibody (Jackson ImmunoResearch, West Grove, PA) and visualized using SuperSignal West Pico chemiluminescent substrate (Pierce, Rockford, IL).

Nuclear extracts and electrophoretic mobility shift assays (EMSA). DNA binding proteins were extracted from nuclei by using a slight modification of previously described procedures (3). Briefly, mice were injected with KA to induce seizures (20 mg/kg, i.p) or with PBS, and hippocampi were dissected 2 h later. Hippocampi were dissociated by passage though a 26-gauge needle and resuspended in 400 µl ice-cold lysis buffer (10 mM HEPES, pH 7.9, 1.5 mM MgCl₂, 10 mM NaCl, 0.5 mM dithiothreitol [DTT], 1 mM Na₃VO₄, 10 mM NaF, 20 mM 2-phosphoglycerate, Complete protease inhibitors [Roche, Alameda, CA]). Dissociated cells were allowed to swell on ice for 15 min, after which 25 µl of 10% NP-40 was added, and the tubes were vortexed for 10 s. After centrifugation and supernatant removal, the nuclear pellet was resuspended in 100 µl ice-cold extraction buffer (20 mM HEPES, pH 7.9, 0.4 mM NaCl, 1 mM EDTA, 1 mM DTT, 1 mM Na₃VO₄, 10 mM NaF, 20 mM 2-phosphoglycerate, Complete protease inhibitors [Roche, Alameda, CA]) for 60 min. Binding reactions using the nuclear protein extracts were carried out at room temperature for 20 min with 15 µg of nuclear extract, 0.5 ng ³²P-labeled oligonucleotide probe, 1 mg poly(dI-dC) in a binding buffer consisting of 10 mM Tris (pH 7.5), 26 mM KCl, 1 mM MgCl₂, 0.5 mM ZnCl₂, 1 mM EDTA, 5 mM DTT, and 5% (vol/vol) glycerol. For supershift experiments, nuclear extracts were preincubated with 2 µg of anti-Egr1 (sc-189x), anti-Egr2 (sc-190x), anti-Egr3 (sc-191x), or anti-Egr4 (sc-19868) polyclonal antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) for 30 min on ice before initiating the binding reaction. The sequence of the sense strand of the double-stranded oligonucleotides used as probes, with mutations of the wild-type sequence in bold, are as follows: E1 (–86 to –66) 5'-CTGGGCGCCGCCAATGGGAG-3', E2 (–54 to –34) 5'-AGCTGCCGCCCACGGGCCCC-3', E2m1 5'-GCTCCACGAGCTGCCAAACACGGGCCCCGCGC-3', E2m2 5'-AGCTGCCGCCCACGGAAACCGCGCAGCATAAATAGC-3'. All probes were 5' end labeled with T4-polynucleotide kinase and [-³²P]ATP (Perkin Elmer, Boston, MA).

Luciferase assays and reporter constructs. Plasmids for luciferase experiments were constructed by PCR using mouse genomic DNA as the template. The amplified promoter fragments were cloned into pGL3 Basic luciferase reporter vector (Promega, Madison, WI) and sequence verified. Luciferase reporter plasmids (0.5 µg) were cotransfected into SH-SY5Y-trkA cells (a generous gift of Sven Påhlman, Lund University, Sweden) with pRL-TK (Promega, Madison, WI) and 1 µg of expression plasmid consisting of full-length rat Egr1, Egr3, or Nab2 cloned into pcDNA3 (Invitrogen, Carlsbad, CA) or pcDNA3 alone. Twenty-four hours after transfection, lysates were processed for luciferase activity using the Dual Luciferase Assay (Promega, Madison, WI) according to the manufacturer's specifications.

Adenovirus preparation. Recombinant adenoviruses were generated using homologous recombination in Escherichia coli as previously described (10). The generation, characterization, and amplification of the enhanced green fluorescent protein (EGFP) Egr3, and transcriptionally inactive Egr3 (Egr3_Tr) adenoviruses were previously described (2). The Egr1-expressing adenovirus was generated for this study by using a full-length N-terminal hemagglutinin-tagged rat Egr1 cDNA (GenBank NM_012551) which was cloned into pAdTrackCMV. The recombinant Egr1 pAdTrack shuttle vector was recombined by homologous recombination into the adenoviral genomic plasmid (pAdEasy) in E. coli BJ5183 to generate a recombinant replication-deficient adenoviral genome plasmid. The replication-deficient EGFP, Egr1, Egr3, and Egr3_Tr expression viruses were packaged and amplified in transfected HEK-293 cells (ATCC, Manassas, VA), purified/concentrated on cesium chloride gradients, and titered by using EGFP fluorescence and 50% culture infective doses.

Primary cell culture and adenoviral infection. (i) Myotubes. Primary myoblasts were isolated and differentiated into myotubes for 10 days in vitro (DIV) as previously described (2).

(ii) Neurons. Primary neuron cultures were prepared from hippocampi or cerebral cortices of mouse embryos as described previously (4). Briefly, dorsal hippocampi (coordinates were as follows: interaural, 2.34 mm; bregma, 1.46 mm) (22) and trunk/barrel somatosensory cortices (coordinates were as follows: interaural, 2.34 mm; bregma, 1.46 mm) (22) were dissected from embryonic day 16.5 (E16.5) C57BL/6J mice, treated with 0.25% trypsin for 15 min at 37°C, and dissociated by trituration with 0.25% DNase I (Sigma, St. Louis, MO). Dissociated cell suspensions were plated at 1.25 x 10⁶ cells per well on six-well tissue culture plates or at 3 x 10⁶ cells per dish on 60-mm tissue culture dishes coated with poly-L-lysine, in minimum essential medium (MEM, Mediatech, Inc. VA) supplemented with 10% horse serum (Harlan, Indianapolis, IN), 100 U/ml penicillin, and 0.1 mg/ml streptomycin. After 4 h, neurons were maintained in glia-conditioned MEM containing N2 supplements plus 0.1% ovalbumin (Sigma, St. Louis, MO) and 0.1 mM sodium pyruvate (Sigma, St. Louis, MO). Cytosine-ß-D-arabinofuranoside (2.0 µM; Sigma, St. Louis, MO) was added on the third day after plating (3 DIV) to inhibit cellular mitosis of contaminating glial cells.

(iii) Adenoviral infections. Myotubes were infected with Egr3 and Egr3_Tr adenovirus with a multiplicity of infection of 100 to obtain 100% infection efficiency without identifiable toxicity. Cultured mouse hippocampal and cortical neurons were infected with EGFP, Egr1, or Egr3 adenovirus at 6 DIV at a multiplicity of infection of 1,000. The neuron cultures were >95% free of glial contamination, and the infections were 60 to 70% efficient as determined by averaging several random fields of the number of green (infected) neurons relative to the number of neurons (DAPI [4',6'-diamidino-2-phenylindole] stained) present (data not shown). There was no morphological evidence of cytotoxicity and the neurons were harvested in Trizol (Invitrogen, Carlsbad, CA) 24 h following infection.

RNA preparation, qRT-PCR, and semiquantitative gene expression analysis. Total RNA was isolated from adenovirus-infected primary myotubes (7 DIV), differentiated hippocampal neurons (6 DIV), and cortical neurons (6 DIV) with Trizol (Invitrogen, Carlsbad, CA). Total RNA (0.5 to 1.0 µg) was reverse transcribed using random octomer priming and Powerscript reverse transcriptase according to the manufacturer's specifications (BD Biosciences Clontech, Palo Alto, CA). Quantitative reverse transcription-PCR (qRT-PCR) was performed on an SDS5700 sequence detector (Applied Biosystems, Foster City, CA) by using SYBR green (Molecular Probes, Eugene, OR) fluorescence chemistry. For expression analysis studies, non-intron-spanning primers that amplified coding sequence were designed for each target gene. The RNA was treated with RNase-free DNase (Promega, Madison, WI) and parallel reverse transcription reaction mixtures processed without reverse transcriptase were prepared from the treated RNA to control for possible genomic DNA contamination. For chromatin immunoprecipitation (ChIP) analysis, primers spanning appropriate regions of genomic DNA were generated. For all primer pairs, melt curves were generated to identify the optimal temperature to quantify the total fluorescence after each amplification cycle so that only the gene-specific amplicon fluorescence was measured during the PCR. Moreover, serial dilutions of mouse genomic DNA were used to generate standard curves for each primer pair to characterize their linear dynamic range and to provide relative DNA concentrations that correlated with gene expression (or relative DNA enrichment in the case of ChIP analysis) at particular threshold cycle numbers in the linear range of amplification. For each cDNA or ChIP sample, the relative level of target and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) sequence was interpolated from the gene-specific standard curves. The relative level of each target sequence was normalized to the level of GAPDH, and the values were again normalized to a reference condition (such as gene expression in EGFP virus infected cells or relative DNA enrichment in ChIP experiments processed with immunoglobulin G [IgG] instead of immunospecific antibody) to assess change in expression/DNA enrichment relative to that condition. The non-intron-spanning primer sequences used for qRT-PCR analysis are available upon request. They amplified the Egr1 (GenBank NM_012551, nt 790 to 1190), Egr3 (GenBank NM_018781.1, nt 163 to 563), GAPDH (GenBank NM_001001303, nt 210 to 635), and Arc (GenBank NM_018790, nt 1175 to 1483) genes.

Chromatin immunoprecipitation. Mice were injected with PBS or KA to induce seizures (20 mg/kg, i.p.), and hippocampi were dissected 2 h later. Hippocampi were dissociated by passage though a 26-gauge needle and DNA/protein complexes were cross-linked in PBS containing 1% formaldehyde for 10 min at room temperature. After being washed twice in PBS, the tissue fragments were lysed in 500 µl nuclear lysis buffer (50 mM Tris, pH 8.1, 10 mM EDTA, 1% SDS, Complete protease inhibitors [Roche, Alameda, CA]). The lysates were incubated on ice for 15 min, and the chromatin was sonicated 10 times for 15 s using a cone sonicator (Sonic Dismembrator, model 300; Fisher Scientific, Pittsburg, PA) at 30% power. Insoluble material was cleared by centrifugation, the sheared chromatin was diluted 10-fold in dilution buffer (16.7 mM Tris, pH 8.1, 0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 167 mM NaCl) and incubated overnight in the presence of 5 µg of rabbit anti-Egr1, rabbit anti-Egr3 antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) or normal rabbit IgG (Jackson Immunoresearch, West Grove, PA). Immunocomplexes were captured using 50 µl of protein A/G-agarose (Santa Cruz Biotechnology, Santa Cruz, CA) which had been previously blocked with sheared salmon sperm DNA. The beads were sequentially washed for 5 min in low-salt buffer (20 mM Tris-HCl, pH 8.1, 0.1% SDS, 1% Triton X-100, 2 mM EDTA, 150 mM NaCl), high-salt buffer (20 mM Tris-HCl, pH 8.1, 0.1% SDS, 1% Triton X-100, 2 mM EDTA, 500 mM NaCl), LiCl buffer (10 mM Tris, pH 8.1, 0.25 M LiCl, 1% sodium deoxycholic acid, 1 mM EDTA), and then twice in 10 mM Tris (pH 7.2)-1 mM EDTA. The immunocomplexes were eluted in 0.1 M NaHCO₃ (pH 9.0)-1% SDS while shaking for 30 min. Cross-links between protein and DNA were reversed at 65°C for 5 h in the presence of 10 µg RNase-0.3 M NaCl and the chromatin was ethanol precipitated. The pellets were resuspended in proteinase K buffer (10 mM Tris, pH 8.5, 1 mM EDTA, 0.25% SDS) and incubated for 3 h at 42°C with 15 µg/ml of proteinase K (Invitrogen, Carlsbad, CA) and the remaining DNA was purified using QIAquick spin columns (QIAGEN, Valencia, CA). Quantitative PCR (qPCR) was performed as described above using the following primers (numbering according to previously published mouse Arc promoter sequence [36]): to amplify the distal Arc promoter (nt –1842 to –1532), Pr1 (5'-GTGCAGACAGACCAGGGTCCAGT-3') and Pr2 (5'-ACACACCCAGGGCTTCCATCC-3'), and to amplify the proximal Arc promoter (nt –518 to –210), Pr3 (5'-ACACACCCAGGGCTTCCATCC-3') and Pr4 (5'-GACTAATGTGCTCTGCTGCGCG-3'). To control for nonspecific DNA that was captured in the ChIP library, a region of the GAPDH coding sequence was amplified using primers 5'-ACGGCAAATTCAACGGCACAGTCA-3' and 5'-GCTTTCCAGAGGGGCCATCCACAG-3', and relative enrichment of Arc sequences in the ChIP DNA were normalized to GAPDH. Alternately, the ChIP DNA was used as a template in a standard PCR and the products were electrophoresed on a 2% agarose gel.

Neuron counts. Neurons were counted in 21-day-old mice that had been exposed to a novel environment 4 h prior to sacrifice using images of random 0.1-mm² high power fields (HPFs) obtained by confocal microscopy. Immunofluorescence histochemistry was performed for Arc and Egr1, Arc and Egr3, or Arc counterstained with SYTOX green (Invitrogen, Carlsbad, CA) to label DNA in all neurons. For colocalization studies in somatosensory cortex, the percent of Arc-positive neurons that were either Egr1 or Egr3 positive were averaged from a total of six HPFs from two wild-type mice and three HPFs per mouse. For quantitation of Arc in wild-type and Egr gene-deficient neurons, Arc-immunopositive neurons were counted as a percent of the total SYTOX green staining neurons in each of six total HPF from three mice of each genotype and two closely related tissue sections per mouse. For all of the mice, the neuroanatomical regions analyzed corresponded to the dorsal hippocampus (coordinates were as follows: interaural, 2.34 mm; bregma, 1.46 mm [22]) and trunk/barrel somatosensory cortex (coordinates were as follows: interaural, 2.34 mm; bregma, 1.46 mm [22]).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

 
Egr1 and Egr3 induce Arc gene expression in multiple cell types. The Arc gene was identified as a potential Egr target gene from an Affymetrix microarray analysis intended to characterize Egr3 target genes that control intrafusal muscle fiber morphogenesis in muscle spindle stretch receptors (2). The microarray results were confirmed by comparing Arc expression in primary murine myotubes infected with an adenovirus that expressed EGFP and Egr3_Tr to that in myotubes infected with an adenovirus that expressed EGFP and transcriptionally active Egr3 (Egr3). Arc was upregulated 11.3-fold in primary myotubes expressing Egr3 relative to Egr3_Tr (Fig. 1A). In the hippocampus, Egr1 and Egr3 are the major Egr-related proteins upregulated after seizure-induced synaptic activity (21). Therefore, we examined whether Egr1 and/or Egr3 could regulate Arc gene expression in a neuronal context in differentiated murine primary hippocampal and cortical neurons using adenoviruses to enforce expression of EGFP, EGFP and Egr1, or EGFP and Egr3. Arc gene expression was upregulated by Egr1 and Egr3 250- and 168-fold, respectively, in hippocampal neurons (Fig. 1B) and by 8- and 15.5-fold, respectively, in cortical neurons (Fig. 1C). The difference in the relative magnitudes of induction of hippocampal and cortical neurons was apparently due to the fact that hippocampal neurons had very low basal levels of Arc expression, whereas cortical neurons had higher levels under these culture conditions (data not shown). Nevertheless, either Egr1 or Egr3 significantly induced Arc gene expression in primary myotubes and neurons in vitro.

View larger version (77K): [in this window] [in a new window]   FIG. 1. Arc expression is induced by Egr1 and Egr3 in vitro and is coexpressed with Egr1 and Egr3 in the hippocampal neurons (n) after seizure in vivo. (A) qRT-PCR analysis of Arc expression in primary (1°) myotubes differentiated 10 DIV and infected for 24 h with adenoviruses expressing EGFP and transcriptionally active Egr3 (Egr3) and adenoviruses expressing EGFP and Egr3Tr. Arc was induced 11.4 ± 1.4-fold in myotubes expressing Egr3 relative to myotubes expressing Egr3Tr. (B) In primary hippocampal neurons (n.) differentiated 3 DIV and infected for 24 h with EGFP and Egr1 (Egr1) or Egr3 and EGFP (Egr3), Arc expression was induced 250.1 ± 76.1-fold and 220.0 ± 23.7-fold, respectively, relative to neurons infected with a virus expressing only EGFP. (C) In primary cortical neurons differentiated 3 DIV and infected for 24 h with Egr1- or Egr3-expressing adenovirus, Arc expression was induced 8.0 ± 2.0-fold and 15.4 ± 2.6-fold, respectively, relative to neurons infected with a virus expressing only EGFP. (D) After KA-induced seizure, Egr1, Egr3, and Arc are coinduced, each with a stereotypic time course. The transient Arc protein levels are most closely correlated with the transient levels of Egr3 protein. By contrast, Egr1 protein has a shorter transient time course after seizure. Two hours after seizure, Egr1 (E) is robustly induced and coexpressed with Arc (E'), similar to Egr3 (F), which is also coexpressed with Arc (F') in hippocampal dentate gyrus granule neurons. Scale = 50 µm. *, P < 0.01, Student's t test relative to conditions without asterisks. Values represent means ± standard deviations of three or four experimental replicates for each condition.

The proximal promoter region of the Arc gene contains a high-affinity ERE that is necessary and sufficient for Egr-mediated Arc transcription. Since Egr proteins regulate gene transcription through an incompletely defined mechanism that involves obligate binding to a GC-rich sequence (ERE) in target gene promoters, we examined whether Egr1 and/or Egr3 could directly bind Arc promoter sequences in vivo and hence whether they might directly regulate Arc gene expression. Chromatin from the hippocampus of unseized control mice and mice subjected to KA induced seizures 2 h prior to hippocampal dissection was sheared into fragments averaging 0.5 to 1 kb in length. Genomic DNA bound to either Egr1 or Egr3 was isolated by ChIP, and PCR primers were used to amplify precipitated DNA representing either a distal segment of the Arc promoter (primers Pr1 and Pr2) (Fig. 2A) or a proximal segment near the transcription start site (primers Pr3 and Pr4) (Fig. 2A) (36). In ChIP samples from unseized and seized hippocampus, neither Egr1 nor Egr3 was bound to distal (Pr1 and Pr2) regions of the Arc promoter. However, when proximal primers were used for PCR (Pr3 and Pr4), Arc genomic DNA was amplified from seized hippocampus but not unseized hippocampus, indicating that both Egr1 and Egr3 were bound to a region of genomic DNA near the Arc transcription start site after they were induced by seizures.

View larger version (26K): [in this window] [in a new window]   FIG. 2. Egr1 and Egr3 bind the Arc promoter in vivo after seizure and transactivate the Arc promoter through a response element in the proximal promoter. (A) Egr1 and Egr3 bind to a region of the Arc promoter that is proximal to the transcription start site. The position of the distal (Pr1 and Pr2) and proximal (Pr3 and Pr4) primers used for PCR and qPCR are shown relative to the TATA box and transcription start site (+1) of the mouse Arc gene (numbering according to reference 36). For each primer pair, endpoint PCR and qPCR are shown with ChIP libraries generated from unseized and seized mouse hippocampus by using Egr1 (1)- and Egr3 (3)-specific antibodies or nonimmune IgG. Neither Egr1 nor Egr3 were bound to a distal region of the Arc promoter (Pr1-Pr2) before or after seizure. However, they were bound to a proximal region of the Arc promoter (Pr3-Pr4) in vivo after seizure and enriched 21.9- ± 5.5- and 41.7- ± 22.9-fold in ChIP libraries generated with Egr1- and Egr3-specific antibodies, respectively, relative to nonimmune IgG. (B) Luciferase reporter plasmids containing portions of the proximal Arc promoter –775 to –87 nucleotides upstream of the transcription start site were significantly activated by Egr1 and Egr3 relative to the CMV expression vector alone. All of the promoter fragments down to –87 nucleotides upstream of the transcription start site were activated by Egr1 and Egr3, suggesting that the ERE was located within this region of the proximal promoter. *, P < 0.01, Student's t test relative to conditions without asterisks; values represent means ± standard deviations of three or four experimental replicates.

To more precisely define the ERE within the Arc promoter, genomic sequence consisting of the –87 nucleotide regulatory fragment from mouse, rat, and human was aligned to the TATA box upstream of the transcription start site, and two GC-rich domains (E1 and E2) that could contain Egr protein binding sites were identified (Fig. 3A). To examine whether either of these regions is capable of binding to Egr1 and/or Egr3 proteins, nuclear extracts from seized and unseized microdissected hippocampus and radiolabeled oligonucleotide probes representing the E1 and E2 domains were used in EMSA. By contrast to the E1 probe, which formed no specific protein-DNA complexes with nuclear extracts from unseized or seized hippocampus, the E2 probe generated multiple complexes, indicating that specific proteins within the hippocampal extracts were capable of binding this small region of the Arc gene promoter in vitro (Fig. 3B). Interestingly, in unseized hippocampus, no complexes were formed, indicating that all of the E2 DNA binding proteins were induced by seizure. To examine whether any of these complexes contained Egr proteins, Egr protein-specific antibodies were incubated with the radiolabeled oligos and nuclear extracts. Whereas antibodies that cross-reacted with Egr2 and Egr4 had no effect on the formation of the protein-DNA complexes, the Egr1 antibody specifically supershifted the high-molecular-weight complex and the Egr3 antibody supershifted the lower-molecular-weight complexes (Fig. 3B, asterisk). The supershifting that occurred for multiple low-molecular-weight complexes by the Egr3 antibody is consistent with previous results demonstrating the presence of multiple Egr3 isoforms produced by alternative translation start sites (19) and that all of these Egr3 isoforms are recognized by the C-terminal-epitope-specific Egr3 antibody. Moreover, Egr1 and Egr3 are the only Egr-related proteins that bind to the E2 domain, consistent with previous results indicating that they are the major Egr proteins upregulated in hippocampus after seizure-induced synaptic activation (19). Finally, there is no evidence that other proteins bind to the E2 domain after seizure-induced synaptic activation, since all of the protein-DNA complexes formed were identified by supershifting with either Egr1- or Egr3-specific antibodies.

View larger version (39K): [in this window] [in a new window]   FIG. 3. Egr1 and Egr3 bind to a high-affinity response element that is necessary and sufficient to transactivate Arc gene expression. (A) The proximal Arc promoters from mice, rats, and humans were aligned to the TATA box to identify conserved regions that might represent Egr binding sites. Two GC rich domains were identified (E1 and E2). The E2 domain appeared to contain a highly conserved response element (ERE; gray). (B) Nuclear proteins from unseized and seized mouse hippocampi were able to bind to probes representing the E2 domain but not the E1 domain by EMSA. The nuclear proteins were induced by seizure and the protein DNA complexes specifically contained Egr1 and Egr3 but not other Egr proteins as demonstrated by supershifting of the Egr-containing complexes by Egr-specific antibodies (asterisk). Moreover, no additional binding proteins were observed, as only Egr1- and Egr3-containing complexes were bound to the E2 domain. (C) To determine whether Egr proteins were bound to the putative ERE in the E2 domain, EMSA was used with a probe that contained a mutation expected to disrupt Egr protein binding within the core binding domain (E2m1) and a probe that contained a mutation outside of the core binding domain not expected to disrupt Egr protein binding. (D) The mutation within E2m1 completely abrogated Egr protein binding with nuclear proteins obtained from seized mouse hippocampus. Egr1 and Egr3 proteins were capable of binding to the probe containing mutations outside of the core binding domain (E2m2) similar to the wild-type sequence. (E, left panel) When the E2m1 mutation was introduced into the –775Arc luciferase reporter construct, activation by either Egr1 or Egr3 was completely abrogated. (E, center panel) Egr3 appeared to consistently activate the Arc promoter better than Egr1, but this was also observed on the minimal 4x ERE reporter, which has been previously shown to result from increased levels of Egr3 expression/stability relative to Egr1 from the CMV expression plasmids. (E, right panel) Egr1- and Egr3-dependent activation of –775Arc promoter was subject to repression by the Egr corepressor Nab2 which completely abrogated Arc transactivation. *, P < 0.01, Student's t test relative to conditions without asterisks, values represent means ± standard deviations of three experimental replicates.

We next examined whether this newly identified ERE was necessary for Egr-mediated Arc gene expression. Indeed, when the E2m1 Egr protein-binding mutation was introduced into the –775Arc luciferase reporter construct, transactivation of the proximal Arc promoter by either Egr1 or Egr3 was completely abrogated in SH-SY5Y human neuroblastoma cells (Fig. 3E, left panel). Egr3 consistently activated the –775Arc promoter better than Egr1 when an equal mass of cytomegalovirus (CMV) expression vector was used. Similar results were also observed when a minimal 4x ERE reporter construct was used (Fig. 3E, center panel), consistent with previous results demonstrating increased expression/stability of Egr3 relative to Egr1 when expressed in vitro from these CMV expression constructs (24). Finally, in NIH 3T3 cells, both Egr1 and Egr3 significantly activated the –775Arc promoter, and this activation was completely abrogated by coexpression of the Egr coregulator Nab2 (29) (Fig. 3E, right panel).

Egr3 mediates the protein synthesis-dependent phase of Arc gene expression in the hippocampus after seizure. In the rat hippocampal dentate gyrus granule neurons, the Arc gene appears to be regulated as an immediate early gene that is rapidly induced after synaptic activation in the absence of new protein synthesis. However, it is also regulated by a protein synthesis-dependent mechanism that modulates its expression beyond the transient immediate early phase (35). To examine whether Egr1 and/or Egr3 may mediate protein synthesis-dependent Arc gene expression in mouse hippocampus, we first identified a time point after seizure when Arc expression is completely dependent upon new protein synthesis. Adult wild-type littermate mice were injected either with phosphate-buffered saline or with KA, and they were sacrificed 2 and 4 h after KA-induced seizure or PBS injection. In addition, the mice received either PBS or CHX 15 min prior to the PBS or KA treatment to inhibit protein synthesis. Compared to the low basal levels of Egr1 (not shown), Egr3 (Fig. 4A), and Arc (Fig. 4A') expression in caged adult mice treated with PBS, mice receiving KA-induced seizures 2 h prior to analysis, but without CHX pretreatment, showed marked induction of Egr1 (Fig. 5A ', inset), Egr3 (Fig. 4B), and Arc expression (Fig. 4B'). Unlike with Egr1 and Egr3, where the RNA was confined largely to the neuronal somata (Fig. 4B), Arc RNA was additionally distributed into the dendrites of dentate gyrus granule neurons (Fig. 4B', arrow), consistent with previous results in rats (17). Similarly, after KA-induced seizure and CHX pretreatment, Egr1 (data not shown), Egr3 (Fig. 4C), and Arc (Fig. 4C') genes were highly induced, consistent with their regulation as immediate early genes that do not require synaptic activation induced protein synthesis. By 4 h after KA-induced seizure in mice not receiving CHX pretreatment, Egr1 expression had returned to baseline (Fig. 5D', inset), and increased Egr3 (Fig. 4D) and Arc gene expression (Fig. 4D') was still present in the hippocampus, consistent with the results from the kinetic analysis of protein induction in the hippocampus using Western blotting (Fig. 1D). However, 4 h after KA-induced seizure in mice receiving CHX pretreatment, both Egr1 (not shown) and Egr3 (Fig. 4E) were superinduced, but Arc expression was completely abrogated (Fig. 4E'). These results identify two time points after KA-induced synaptic activation in mice when the protein synthesis-independent (2 h postseizure) and protein synthesis-dependent (4 h postseizure) Arc expression levels can be clearly distinguished in hippocampal dentate gyrus granule neurons (compare Fig. 4C' and E').

View larger version (84K): [in this window] [in a new window]   FIG. 4. Arc is transiently expressed as a protein synthesis-independent immediate early gene and its expression is modulated by a protein synthesis-dependent mechanism. In caged wild-type adult mice treated with PBS, Egr3 (A) and Arc (A') expression were very low in the hippocampus. Two hours after KA-induced seizure, Egr3 (B) and Arc (B') were markedly upregulated. Unlike Egr3 mRNA (B), which was localized within the neuron somata and nuclei, Arc mRNA (B', arrow) was also characteristically transported into the dendrites of hippocampal dentate gyrus neurons. When mice were treated with CHX 15 min prior to seizure to inhibit protein synthesis, there was no detectable difference in Egr3 (C) or Arc (C') expression compared to mice not receiving CHX. By 4 h after KA-induced seizure in the absence of CHX treatment, Egr3 (D) and Arc (D') expression were still elevated above basal (PBS treatment) levels. In CHX-pretreated mice 4 h after KA-induced seizure, Egr3 (E) expression was superinduced, consistent with its regulation as an immediate early gene; however Arc (E') expression was completely abrogated. Thus, Arc expression is independent of protein synthesis within at least 2 h after KA-induced seizure but by 4 h after seizure, Arc expression is completely regulated by a protein synthesis-dependent mechanism. Scale = 0.5 µm.

View larger version (141K): [in this window] [in a new window]   FIG. 5. The protein synthesis-dependent phase of Arc gene expression requires Egr3 after seizure-induced synaptic activation. Similar to Arc (A), Egr3 (A'), and Egr1 (A', inset), which were markedly upregulated in wild-type mice independent of new protein synthesis 2 h after seizure, Arc (B) and Egr3 (B') were appropriately induced in Egr1-deficient mice and Arc (C) and Egr1 (C') were appropriately induced in Egr3-deficient mice, indicating that similar seizure induced activity was generated the Egr gene-deficient mice to induce qualitatively equivalent levels of protein synthesis-independent Arc expression. However 4 h after seizure, when Arc expression was dependent upon new protein synthesis, Arc (D) and Egr3 (D') were still upregulated and (D', inset) Egr1 had returned to basal levels. In Egr1-deficient mice, Arc (E) and Egr3 (E') were also elevated similar to wild-type mice. However, in Egr3-deficient mice, Arc (F) expression was markedly diminished and Egr1 (F') was slightly elevated above basal levels. Thus, protein synthesis-dependent Arc expression requires Egr3 after seizures in the hippocampus. Scale = 0.5 µm.

To understand why Egr3, but not Egr1, was necessary for protein synthesis-dependent Arc gene expression after seizure, we examined the relative levels of Egr1, Egr3, and Arc proteins in the hippocampus from wild-type, Egr1-deficient, and Egr3-deficient mice after seizure when Arc expression was dependent upon new protein synthesis (4 h postseizure). Similar to the kinetics of Egr1, Egr3, and Arc protein induction that was observed in wild-type mice 4 h after seizure (Fig. 1D), we found that in Egr1-deficient mice, Egr3 protein was present at a high level similar to wild-type mice, but in Egr3-deficient mice, Egr1 protein was present at a very low level, similar to the situation with wild-type mice (Fig. 6). Moreover, Arc protein was markedly decreased in Egr3-deficient mice compared to either wild-type or Egr1-deficient mice, consistent with the in situ hybridization results (Fig. 5F). Thus, despite the fact that both Egr1 or Egr3 can bind the Arc ERE and transactivate Arc expression after seizure in vivo, because the kinetics of Egr1 expression overlapped with the immediate early protein synthesis-independent phase of Arc expression and Egr3 was simultaneously elevated with Egr1 during this period (Fig. 1D), it was not possible to distinguish a specific role for Egr1 in regulating Arc expression in this seizure paradigm. However, since Egr3 protein remained elevated beyond the transient immediate early phase of Arc expression, when Egr1 protein had returned to basal levels, it was possible to clearly identify an essential role for Egr3 in regulating the protein synthesis-dependent phase of Arc gene expression in this paradigm.

View larger version (45K): [in this window] [in a new window]   FIG. 6. Prolonged elevation of Egr3 protein in the hippocampus after seizure explains why Egr3, but not Egr1, regulates the protein synthesis-dependent phase of Arc expression. Hippocampal lysates obtained four hours after seizure, when Arc expression depends upon protein synthesis in wild-type hippocampus, contain very low levels of Egr1 and high levels of Egr3 protein because of the difference in kinetics between the two proteins after seizure induction. Similarly, Egr3 protein levels remain high in the hippocampus of Egr1-deficient mice, and consequently, Arc protein levels are comparable to wild-type levels. By contrast, in the hippocampus of Egr3-deficient mice, Egr1 levels are very low, as they are in wild-type mice 4 h after seizure, and without Egr1 or Egr3, Arc protein levels are very low compared to those in wild-type or Egr1-deficient mice. These results indicate that Egr1 is dispensable for regulating the protein synthesis-dependent phase of Arc expression in this seizure paradigm. In Egr1-deficient mice, there are high levels of Egr3 which are sufficient to regulate Arc expression whereas in Egr3-deficient mice there is an insufficient amount of Egr1 protein to compensate for Egr3 loss leading to correspondingly very low levels of Arc protein.

View larger version (167K): [in this window] [in a new window]   FIG. 7. Arc, Egr1, and Egr3 are coexpressed in hippocampal dentate gyrus granule neurons and cortical neurons during physiologic synaptic activity. Four hours after mice were subjected to a novel environment, Egr1 (A), Arc (B), and Egr3 (C) were coexpressed in a subpopulation of hippocampal dentate gyrus granule neurons. Some Arc-expressing neurons coexpressed Egr1 (white-filled arrowheads), some coexpressed Egr3 (open arrowheads), and some coexpressed both Egr1 and Egr3 (black-filled arrowheads). In the somatosensory cortex, in Arc protein (D)-expressing neurons, Egr1 protein (E) was present in 94.6 ± 4.1% of the neurons (F). Arc protein (G)-expressing neurons also expressed Egr3 (H) in 96.6 ± 2.4% of the neurons (I). (A through C and D through H) Scale = 50 µm. (F and I) Values represent means ± standard deviations of neuron counts from six 0.1-mm2 high power fields from two wild-type 21-day-old mice exposed to a novel environment.

View larger version (115K): [in this window] [in a new window]   FIG. 8. Arc expression requires Egr1 and Egr3 in the context of physiologic synaptic activity. (A) In 21-day-old mice, four hours after exposure to a novel environment, Arc was expressed at high levels by many cortical and hippocampal neurons. In the hippocampus, a subpopulation of neurons (arrow) expressed high levels of Arc (E), and in the frontal cortex (somatosensory cortex shown), pyramidal neurons in superficial and deep layers of cortex preferentially expressed Arc (I). In Egr1-deficient mice (B) and Egr3-deficient mice (C), Arc expression was decreased compared to that in wild-type mice, which was reflected by a decreased number of Arc-expressing hippocampal dentate gyrus granule neurons (F and G, arrows) and a decrease in the number and intensity of Arc-expressing neurons in cortex (J and K). (D) In Egr1/3 dKO mice, Arc expression was very low compared to that in wild-type mice. In the hippocampal dentate gyrus, no Arc-expressing neurons were identified (H) and similarly in cortex, Arc expression was barely detectable (L). Boxes in panels A through D orient the magnified regions shown in panels E through L. (A through D) Scale = 1 mm. (E through H) Scale = 100 µm. (I through L) Scale = 50 µm.

View larger version (68K): [in this window] [in a new window]   FIG. 9. Arc protein is decreased in Egr-deficient cortical and hippocampal neurons. Many cortical neurons (somatosensory cortical neurons shown here) contain high levels of Arc protein in wild-type (A) but not Egr1/3 dKO (B) mice after identical treatment to novel environmental stimuli. The cortical neurons were apparently activated similarly because the level of c-fos, an activity regulated immediate early protein, was similar in wild-type (C) relative to Egr1/3 dKO (D) cortical neurons (red labeling corresponds to Arc or c-fos protein and green corresponds to SYTOX green staining for DNA [scale bar = 100 µm]). (E) In wild-type somatosensory cortex, 17.5% of neurons were immunoreactive for Arc, whereas 2.2%, 1.3%, and 0% of neurons were immunopositive for Arc in Egr1-deficient, Egr3-deficient, and Egr1/3 dKO mice, respectively. (F) In wild-type mice, 1.6% of hippocampal dentate gyrus neurons were immunoreactive for Arc protein. By contrast, in Egr1-deficient and Egr3-deficient mice, 0.8% and 0.2% of hippocampal dentate gyrus neurons, respectively, were immunoreactive for Arc protein. In hippocampal dentate gyrus neurons from Egr1/3 dKO mice, no immunoreactive neurons (0%) were identified. (E and F) Values cited are shown as horizontal bars which represent the medians of six measurements from three animals and two similar tissue sections analyzed for each genotype. *, P < 0.05; **, P < 0.005, Student's t test relative to wild type.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In primary myotubes, and hippocampal and cortical neurons, enforced expression of either Egr1 or Egr3 was capable of significantly upregulating Arc gene expression. Moreover, after seizure-induced synaptic activation, Egr1 and Egr3 were bound in vivo to a species-conserved high-affinity response element in the proximal Arc promoter that was necessary and sufficient for Arc transactivation in vitro. Transcriptional activation of Arc by either Egr1 or Egr3 was completely repressed by coexpression of the Egr coregulatory molecule Nab2, suggesting that Egr-mediated Arc regulation may be subject to complex control in vivo (29). Considering that Egr1 and Egr3 were both bound to the Arc ERE after their induction, it was surprising to find that only Egr3 was required for the protein synthesis-dependent phase of Arc expression after synchronous activation by seizure in vivo. The results were not explained by differences in seizure response to kainic acid treatment between Egr1-deficient and Egr3-deficient mice, since the animals displayed highly reproducible seizures that were subjectively equivalent in intensity and duration, and the immediate early components of Arc expression within the first two hours after seizure were identical between wild-type and Egr gene-deficient mice. Rather, the protein synthesis-dependent phase of Arc expression depended upon Egr3 after synchronous synaptic activation by seizure most likely because it was expressed with prolonged kinetics relative to Egr1 (21). Thus, after a single-seizure epoch, Egr3 was essential for regulating the protein synthesis-dependent phase of Arc expression presumably because it was the only Egr protein available to regulate Arc expression 4 h after seizure (Fig. 1D).

During physiologic synaptic activity, when neurons were asynchronously activated by exposure to a novel environment, both Egr1 and Egr3 were required for normal levels of Arc expression in cortical and hippocampal neurons. Egr1 and Egr3 were differentially expressed in some neurons after physiologic activation since they expressed Egr1, Egr3, or both Egr1 and Egr3. After seizure, Egr1 and Egr3 are coactivated and transiently expressed with differing stereotypical kinetics (Fig. 1D). If the differential pattern of Egr1 and Egr3 expression after physiologic synaptic activity was a result of coactivation and differing kinetics of expression then all neurons that expressed Egr1 should have also expressed Egr3, which was not the case (compare Fig. 7A and C). Moreover, while most neurons that expressed Arc also expressed Egr1, Egr3, or both Egr1 and Egr3, some Egr-expressing neurons did not express Arc. This suggests that additional regulatory mechanisms might exist to prevent Arc expression in some circumstances despite the presence of Egr1 and/or Egr3. One possibility is that coregulatory factors, such as Nab2, which can inhibit Egr-mediated Arc gene transactivation (Fig. 1E, right panel), may be differentially regulated in these neurons to provide an additional level of control over Arc transactivation (29). Similarly, the majority of Arc-expressing cortical neurons also expressed either Egr1 or Egr3 but the extent to which these neuron populations overlapped was not determined because antibodies that permit simultaneous labeling of Egr1, Egr3, and Arc were not available. Nevertheless, the data indicate that Egr1 and Egr3 are differentially regulated by physiologic synaptic activation to express Egr1, Egr3, or both Egr1 and Egr3 proteins in particular neurons at a specific time. Consistent with this hypothesis, we found that the frequency of Arc-expressing cortical and hippocampal dentate gyrus neurons, after similar exposure to a novel environment, was decreased in Egr1- and Egr3-deficient mice relative to wild-type mice. In addition, Arc mRNA was barely detectable and Arc protein was completely absent in cortical and hippocampal dentate gyrus neurons from Egr1/3 dKO mice after exposure to a novel environment. While these results seem to confirm the fact that Arc expression depends upon Egr-mediated transcription, they could be explained by decreased neuronal activity in the mutant mouse brains. Similar levels of expression of the activity-regulated immediate early gene c-fos in cortical neurons between wild-type and Egr1/3 dKO mice strongly argues against this possibility, although c-fos expression was decreased in hippocampal dentate gyrus neurons. Thus, in the hippocampal dentate gyrus neurons from Egr 1/3 dKO mice, it is not clear whether the lack of Arc-containing neurons can be definitively attributable to loss of Egr-dependent regulation.

Arc is an unusual plasticity-associated molecule that is induced by NMDA receptor activation in glutamatergic neurons (28). Arc mRNA is transported to recently activated synapses where it is locally translated into protein and integrated into the NMDA multiprotein receptor complex (11), thereby potentially altering the response properties of recently activated synapse associated receptors. Indeed, Arc appears to have an important role in activity-induced synaptic plasticity related to LTP stabilization and consolidation of hippocampal-dependent memory (9), but a more thorough characterization of Arc function will require analysis of mice with neuron-specific loss of Arc, since germ line mutations are embryonic lethal (16). By directly regulating Arc in a subpopulation of neurons, Egr1 would be expected to indirectly influence the properties of recently activated synapses. Thus, we propose that the learning and memory defects that have been defined in Egr1-deficient mice are, at least in part, explained by deregulation of Arc expression in a subpopulation of neurons. Whether Egr3-deficient mice have any defects in LTP stabilization and learning/memory behavior or whether Egr1/3 dKO mice have exacerbated defects is currently under investigation. Ultimately, Arc regulation by Egr transcription factors may be complex, considering the fact that Egr1 and Egr3 appear to have overlapping regulatory control of Arc expression in some neurons and distinct regulatory control in subpopulations of neurons where only a single Egr protein is expressed at any given time.

	ACKNOWLEDGMENTS

 
We thank R. Miller for comments on the manuscript. The Egr1, Egr3, and Nab2 cDNAs were generously provided by J. Milbrandt (Washington University, St. Louis, MO) and J. Svaran (University of Wisconsin, Madison, WI). Technical assistance was provided by Y. Albert.

This study was supported by The National Institutes of Health (NS046468 and NS040748) and a Howard Hughes Faculty Scholar Award to W.G.T. J.C. was supported by a predoctoral fellowship from NIH (CA009560) and the NIH Medical Scientist Training Program (GM008152).

	FOOTNOTES

 
* Corresponding author. Mailing address: Ward 7-110, Department of Pathology, W127, Northwestern University, 393 E. Chicago Ave., Chicago, IL 60611. Phone: (312) 503-2415. Fax: (312) 503-2459. E-mail: warren{at}northwestern.edu.

These authors contributed equally to this work.

	REFERENCES

Top Abstract Introduction Materials and Methods Results Discussion References

 
1. Abraham, W. C., M. Dragunow, and W. P. Tate. 1991. The role of immediate early genes in the stabilization of long-term potentiation. Mol. Neurobiol. 5:297-314.[Medline]

2. Albert, Y., J. Whitehead, L. Eldredge, J. Carter, X. Gao, and W. G. Tourtellotte. 2005. Transcriptional regulation of myotube fate specification and intrafusal muscle fiber morphogenesis. J. Cell Biol. 169:257-268.[Abstract/Free Full Text]

3. Andrews, N. C., and D. V. Faller. 1991. A rapid micropreparation technique for extraction of DNA-binding proteins from limiting numbers of mammalian cells. Nucleic Acids Res. 19:2499.[Free Full Text]

4. Banker, G., and K. Goslin. 1998. Culturing nerve cells, 2nd ed. MIT Press, Cambridge, Mass.

5. Bozon, B., S. Davis, and S. Laroche. 2003. A requirement for the immediate early gene zif268 in reconsolidation of recognition memory after retrieval. Neuron 40:695-701.[CrossRef][Medline]

6. Clayton, D. F. 2000. The genomic action potential. Neurobiol. Learn. Mem. 74:185-216.[CrossRef][Medline]

7. Ghosh, A., D. D. Ginty, H. Bading, and M. E. Greenberg. 1994. Calcium regulation of gene expression in neuronal cells. J. Neurobiol. 25:294-303.[CrossRef][Medline]

8. Guzowski, J. F. 2002. Insights into immediate-early gene function in hippocampal memory consolidation using antisense oligonucleotide and fluorescent imaging approaches. Hippocampus 12:86-104.[CrossRef][Medline]

9. Guzowski, J. F., G. L. Lyford, G. D. Stevenson, F. P. Houston, J. L. McGaugh, P. F. Worley, and C. A. Barnes. 2000. Inhibition of activity-dependent arc protein expression in the rat hippocampus impairs the maintenance of long-term potentiation and the consolidation of long-term memory. J. Neurosci. 20:3993-4001.[Abstract/Free Full Text]

10. He, T. C., S. Zhou, L. T. da Costa, J. Yu, K. W. Kinzler, and B. Vogelstein. 1998. A simplified system for generating recombinant adenoviruses. Proc. Natl. Acad. Sci. USA 95:2509-2514.[Abstract/Free Full Text]

11. Husi, H., M. A. Ward, J. S. Choudhary, W. P. Blackstock, and S. G. Grant. 2000. Proteomic analysis of NMDA receptor-adhesion protein signaling complexes. Nat. Neurosci. 3:661-669.[CrossRef][Medline]

12. Jones, M. W., M. L. Errington, P. J. French, A. Fine, T. V. Bliss, S. Garel, P. Charnay, B. Bozon, S. Laroche, and S. Davis. 2001. A requirement for the immediate early gene Zif268 in the expression of late LTP and long-term memories. Nat. Neurosci. 4:289-296.[CrossRef][Medline]

13. Lee, J. L., B. J. Everitt, and K. L. Thomas. 2004. Independent cellular processes for hippocampal memory consolidation and reconsolidation. Science 304:839-843.[Abstract/Free Full Text]

14. Lee, S. L., Y. Sadovsky, A. H. Swirnoff, J. A. Polish, P. Goda, G. Gavrilina, and J. Milbrandt. 1996. Luteinizing hormone deficiency and female infertility in mice lacking the transcription factor NGFI-A (Egr-1). Science 273:1219-1221.[Abstract]

15. Link, W., U. Konietzko, G. Kauselmann, M. Krug, B. Schwanke, U. Frey, and D. Kuhl. 1995. Somatodendritic expression of an immediate early gene is regulated by synaptic activity. Proc. Natl. Acad. Sci. USA 92:5734-5738.[Abstract/Free Full Text]

16. Liu, D., D. Bei, H. Parmar, and A. Matus. 2000. Activity-regulated, cytoskeleton-associated protein (Arc) is essential for visceral endoderm organization during early embryogenesis. Mech. Dev. 92:207-215.[CrossRef][Medline]

17. Lyford, G. L., K. Yamagata, W. E. Kaufmann, C. A. Barnes, L. K. Sanders, N. G. Copeland, D. J. Gilbert, N. A. Jenkins, A. A. Lanahan, and P. F. Worley. 1995. Arc, a growth factor and activity-regulated gene, encodes a novel cytoskeleton-associated protein that is enriched in neuronal dendrites. Neuron 14:433-445.[CrossRef][Medline]

18. Murphy, T. H., P. F. Worley, and J. M. Baraban. 1991. L-type voltage-sensitive calcium channels mediate synaptic activation of immediate early genes. Neuron 7:625-635.[CrossRef][Medline]

19. O'Donovan, K. J., and J. M. Baraban. 1999. Major Egr3 isoforms are generated via alternate translation start sites and differ in their abilities to activate transcription. Mol. Cell. Biol. 19:4711-4718.[Abstract/Free Full Text]

20. O'Donovan, K. J., W. G. Tourtellotte, J. Millbrandt, and J. M. Baraban. 1999. The EGR family of transcription-regulatory factors: progress at the interface of molecular and systems neuroscience. Trends Neurosci. 22:167-173.[CrossRef][Medline]

21. O'Donovan, K. J., E. P. Wilkens, and J. M. Baraban. 1998. Sequential expression of Egr-1 and Egr-3 in hippocampal granule cells following electroconvulsive stimulation. J. Neurochem. 70:1241-1248.[Medline]

22. Paxinos, G., and K. B. J. Franklin. 2001. The mouse brain in stereotaxic coordinates. Academic Press, San Diego, Calif.

23. Ramirez-Amaya, V., A. Vazdarjanova, D. Mikhael, S. Rosi, P. F. Worley, and C. A. Barnes. 2005. Spatial exploration-induced arc mRNA and protein expression: evidence for selective, network-specific reactivation. J. Neurosci. 25:1761-1768.[Abstract/Free Full Text]

24. Sevetson, B. R., J. Svaren, and J. Milbrandt. 2000. A novel activation function for NAB proteins in EGR-dependent transcription of the luteinizing hormone beta gene. J. Biol. Chem. 275:9749-9757.[Abstract/Free Full Text]

25. Sham, M. H., C. Vesque, S. Nonchev, H. Marshall, M. Frain, R. D. Gupta, J. Whiting, D. Wilkinson, P. Charnay, and R. Krumlauf. 1993. The zinc finger gene Krox20 regulates HoxB2 (Hox2.8) during hindbrain segmentation. Cell 72:183-196.[CrossRef][Medline]

26. Steward, O., and P. Worley. 2002. Local synthesis of proteins at synaptic sites on dendrites: role in synaptic plasticity and memory consolidation? Neurobiol. Learn. Mem. 78:508-527.[CrossRef][Medline]

27. Steward, O., and P. F. Worley. 2001. A cellular mechanism for targeting newly synthesized mRNAs to synaptic sites on dendrites. Proc. Natl. Acad. Sci. USA 98:7062-7068.