Inducible cAMP early repressor acts as a negative regulator for kindling epileptogenesis and long-term fear memory

Abstract

Long-lasting neuronal plasticity as well as long-term memory (LTM) requires de novo synthesis of proteins through dynamic regulation of gene expression. cAMP-responsive element (CRE)-mediated gene transcription occurs in an activity-dependent manner and plays a pivotal role in neuronal plasticity and LTM in a variety of species. To study the physiological role of inducible cAMP early repressor (ICER), a CRE-mediated gene transcription repressor, in neuronal plasticity and LTM, we generated two types of ICER mutant mice: ICER-overexpressing (OE) mice and ICER-specific knock-out (KO) mice. Both ICER-OE and ICER-KO mice show no apparent abnormalities in their development and reproduction. A comprehensive battery of behavioral tests revealed no robust changes in locomotor activity, sensory and motor functions, and emotional responses in the mutant mice. However, long-term conditioned fear memory was attenuated in ICER-OE mice and enhanced in ICER-KO mice without concurrent changes in short-term fear memory. Furthermore, ICER-OE mice exhibited retardation of kindling development, whereas ICER-KO mice exhibited acceleration of kindling. These results strongly suggest that ICER negatively regulates the neuronal processes required for long-term fear memory and neuronal plasticity underlying kindling epileptogenesis, possibly through suppression of CRE-mediated gene transcription.

Figure 1.

Characterization of ICER-OE mice. A, Transgene constructs for ICER transgenic mice generation were composed of a promoter region from the Ca²⁺/calmodulin-dependent protein kinase IIα gene (pCaMKIIα), 5′-untranslated region (5′-UTR) derived from pNN265, 3′-untranslated region containing a poly(A) signal sequence (SV40pA1 or SV40pA2), and either ICER-I or ICER-II cDNA. The probes for detecting transgenic ICER mRNAs are indicated by thick lines (Probe 1 and Probe 2). B, Expression of transgenic ICER mRNA in mouse brain. Sagittal sections from an ICER-I-OE mouse (line I-15) and a non-TG littermate were hybridized with digoxigenin-labeled cRNA probe 1. The transgene-derived ICER-I mRNA was preferentially expressed in forebrain structures. Scale bars, 500 μm. A coronal section from ICER-II-OE mice (line II) was hybridized with digoxigenin-labeled cRNA probe 2. ICER mRNA was highly expressed in layers II–III and V–VI of neocortex, hippocampus, piriform cortex, and amygdala. Scale bars, 500 μm. C, Northern blot analysis of total RNA prepared from the forebrain of non-TG and ICER-OE mice (lines I-15 and II). ³²P-labeled probe 1 hybridized to transgenic ICER-I mRNA (left, I-15) and ³²P-labeled probe 2 hybridized to both ICER-I and ICER-II mRNAs (right, I-15 and II). Endogenous ICER mRNA in non-TG mice was below the detection limit of probe 2 under the condition used (non-TG). D, Quantitative RT-PCR detection of ICER mRNA in the hippocampus of each transgenic line. Each value was calculated from four to six independent samples and presented as means ± SEM of mRNA levels normalized to mRNA levels of non-TG mice. E, Western blot analysis of protein lysates prepared from the hippocampi of non-TG and ICER-OE mice (lines I-15, I-23, and II); blots were probed with anti-CREM antibody. The 19 kDa band corresponding to ICER protein is indicated by an arrow.

Figure 2.

Generation of ICER-KO mice. A, For construction of the targeting vector, two loxP sequences (arrowheads) were inserted into the 5′- and 3′-flanking region of the P2 exon of the CREM gene. For negative and positive selection of recombinant ES clones, a PGK–neo cassette (neo) flanked by frt sequences and a diphtheria toxin A gene (DT-A), respectively, were inserted into the targeting vector. The P2 exon flanked by two loxP sequences was deleted by microinjecting a Cre recombinase-expressing vector into the fertilized eggs carrying the targeted genomic allele. The location of the probe used for Southern blot analysis is indicated by a thick line. E, EcoRI; S, SacI. B, Southern blot analysis of genomic DNA prepared from WT and recombinant (Targeted) ES cells. A ³²P-labeled DNA probe detected a single 15.0 kb EcoRI fragment in WT cells and an additional 8.1 kb fragment in targeted cells. C, PCR-based genotyping of ICER-KO mice; the primer pair (arrows) specific for the P2 exon distinguished WT, heterozygous (+/−), and homozygous (KO) mice. D, ICER-specific primer pair (described in the supplemental Materials and Methods, available at www.jneurosci.org as supplemental material) was used to amplify the PCR products from brain cDNA prepared from the RNA of WT mice but not from brain cDNA prepared from the RNA of KO mice. E, Western blot analysis of lysates prepared from hippocampi of WT, ICER-KO, and ICER-OE (line II) mice. Blots were probed with anti-CREM or anti-CREB antibody. We detected no apparent difference in CREM isoform and CREB expression levels in all groups of mice. The additional band corresponding to ICER protein in ICER-OE mice is indicated by an arrow.

Figure 3.

Fear conditioning in ICER-OE mice. A, B, Conditioned freezing to tone and context was measured in lines I-15 (A) and II (B) and was compared with conditioned freezing displayed by non-TG littermates. No difference was observed between non-TG mice and both ICER-OE lines of mice at 1 h after conditioning (1-h tone). However, 24 h (24-h tone) and 48 h (48-h context) after conditioning, tone-dependent and context-dependent freezing, respectively, were significantly attenuated in ICER-OE mice. Data are means ± SEM; number of animals tested are in parentheses. **p < 0.01, ***p < 0.001 compared with non-TG littermates.

Figure 4.

Fear conditioning in ICER-KO mice. A, Conditioned freezing to tone and context after subjecting mice to a strong training protocol. KO mice and WT littermates exhibited similar freezing levels during the tone-dependent test performed 1 h (1-h tone) and 24 h (24-h tone) after conditioning and during the context-dependent test performed 48 h after conditioning (48-h context). B, Conditioned freezing to tone and context after subjecting mice to a weak training protocol. We observed no significant differences in conditioned freezing between KO and WT mice 1 h after conditioning (1-h tone). However, KO mice showed increased freezing during both pre-tone and tone presentation at the tone-dependent test performed 24 h after conditioning (24-h tone). Context-dependent freezing in KO mice was also enhanced when tested 48 h after conditioning (48-h context). Data are means ± SEM; number of animals are in parentheses. *p < 0.05, **p < 0.01 compared with WT littermates.

Figure 5.

Amygdala kindling in ICER-OE and ICER-KO mice. Days required to reach each convulsive stage. A–C, Mice were defined as fully kindled (“full”) when they had stage 5 or severer convulsions on 3 consecutive days: ICER-OE mice I-15 (A), II (B), and ICER-KO mice (C). Development of kindling was retarded in ICER-OE mice (lines I-15 and II). In contrast, ICER-KO mice showed accelerated kindling. Data are means ± SEM. Numbers in parentheses (n = x), Total mice number per group; (x), number of mice that reached a particular convulsive stage in case the number is fewer than the total mice number per group. D, Percentage induction of fully generalized kindling in ICER-OE and ICER-KO mice. Only some ICER-OE mice reached a fully kindled state, in contrast to the majority of nonmutant controls and ICER-KO mice. *p < 0.05, **p < 0.01, ***p < 0.001 compared with nonmutant littermates.

Figure 6.

Simplified schematic diagram of the hypothetical involvement of ICER in the transcription of CRE-containing genes. Other CREMs are omitted for simplification. In WT mice, CREB is phosphorylated by protein kinases activated through neuronal excitation, and then phosphorylated CREB activates CRE-containing gene transcription. Dephosphorylation of CREB by protein phosphatase inactivates CREB activity and stops gene transcription. Suppression of activated gene transcription is also achieved by ICER. Phosphorylated CREB binds to the P2 promoter of the CREM/ICER gene and induces ICER expression. ICER is a transcription repressor that suppresses CRE-dependent activation of gene expression, as well as ICER expression itself, by forming an ICER homodimer and/or an ICER-CREB heterodimer. In ICER-OE mice, suppression of CRE-mediated gene transcription (as well as ICER gene expression) is enhanced by an excess amount of ICER. In ICER-KO mice, lack of ICER causes prolonged activation of CRE-mediated gene transcription that results from the loss of an ICER-mediated suppression mechanism. Accumulated gene products resulted from the transcription–translation of the gene are indicated as filled squares. P, Phosphate group; P2, intronic promoter of the CREM/ICER gene.