In the present study, we use Egr-1 knockout mice to investigate the contribution of Egr-1 to different types of fear memory. First, we studied two forms of associative fear memory: contextual and auditory fear conditioning. Contextual fear conditioning is thought to be hippocampus-dependent, while auditory fear conditioning is amygdala-dependent^[48]; but see reference^[49]. We then investigated the role of Egr-1 in trace memory^[50] as well as in the extinction of fear memory. Although there were no differences in response to contextual and trace memory or in the extinction of fear memory, there was a significant decrease in fear responses during late auditory fear memory in Egr-1 knockout versus wild-type mice. Our results support a selective role for Egr-1 in the processing of long-term auditory fear memory.

1 MATERIALS AND METHODS

1.1 Animals and treatment

Adult male mice (wild-type and mutant Egr-1 mice generated by Dr. J. Milbrandt) were used. Wild-type and homozygous mutant Egr-1 mice were obtained by crossing heterozygous mutant mice bearing a targeted mutation of the Egr-1 gene. Genotypes were determined by PCR analysis^[51] of genomic DNA extracted from mouse ear tissue. Mice were maintained in a C57BL/6 strain background and were age-matched in each experiment. Wild-type and mutant mice were well groomed and showed no signs of abnormality or any obvious motor defects. No indication of tremor, seizure or ataxia was observed. As it was impos sible to visually distinguish mutant mice from wild-type mice, experimenters were blind to the genotype. The Animal Care and Use Committees at Washington University and the University of Toronto approved the experimental protocols.

1.2 Fear conditioning

The following experiments were performed in a conditioning shock chamber (30.5 cm×24.1 cm×21.0 cm) (Med Associates, Georgia, Vermont). Mice were allowed to habituate to the chamber for 2 min before fear conditioning. The conditioned stimulus (CS) used was an 85 dB tone at 2800 Hz for 30 s, and the unconditioned stimulus (US) was a continuous scrambled foot shock at 0.75mA for 2 s. During training, mice were presented with a 30 s tone (CS) and a shock (US) starting 28 s after the onset of the CS. Three CS/US pairings were delivered during multi-shock conditioning, while only one was used for single shock experiments. After the CS/US pairing, mice were allowed to stay in the chamber for an additional 30s for the measurement of immediate freezing. Freezing was scored manually every 10 s. For contextual memory, each mouse was placed back into the shock chamber and the freezing response was recorded for 3min. For auditory fear memory, the mice were put into a novel chamber (different floor, ceiling and walls) and monitored for 3min before the onset of a tone identical to the CS, which was delivered for 3min, and freezing responses were recorded. Contextual and auditory fear memory was measured 1 h, 1, 3, 7, and 14 d after training for all animals. To measure trace memory, we used a trace fear conditioning paradigm as described^[50]. For this paradigm, the US was delivered 30 s after the end of the CS (trace) and mice were subjected to three training trails. To test the extinction of fear memory, mice were trained with three shock-tone pairings and fear responses were measured during five trails at 1 h intervals in both the context where they had received the shock-tone pairing and in a novel chamber before and after the onset of the tone (CS). The percentage change in fear memory was normalized to control responses.

1.3 Elevated plus maze

The elevated plus maze (Med Associates, Georgia, Vermont) consists of two open arms and two closed arms situated opposite each other and separated by a 6cm square center platform. Each runway is 6 cm wide and 35 cm long. The open arms have lips that are 0.5 cm high and the closed arms are surrounded on three sides by 20 cm walls. The floors and walls are black polypropylene and the floors are 75 cm from the ground. For each test, the animal is placed in the center square and allowed to move freely for 5 min. The number of entries and time spent in each arm is recorded.

1.4 Open field activity

To record horizontal locomotor activity we used the Activity Monitor system from Med Associates (43.2 cm×43.2 cm×30.5 cm) (Med Associates, St. Albans, VT). Briefly, this system uses paired sets of photo beams to detect movement in the open field and movement is recorded as beam breaks. The open field is placed inside an isolation chamber with dim illumination and a fan. Each subject was placed in the center of the open field and activity was measured for 30 min.

1.5 Slice electrophysiology

Mice were anesthetized with halothane and transverse slices of amygdala and cortex were rapidly prepared and maintained in an interface chamber at 30℃, where they were subfused with artificial cerebrospinal fluid (ACSF) consisting of (mmol/L) NaCl 124, CaCl₂ 4.4, MgSO₄ 2.0, NaHCO₃ 25, Na₂HPO₄ 1.0, glucose 10, and bubbled with 95% O₂ and 5% CO₂. In all experiments, slices recovered in the chamber for at least 2 h before recording. In amygdala slices, a bipolar tungsten stimulating electrode was placed in the ventral striatum, and an extracellular recording electrode (3~12 MW filled with ACSF) was placed in the lateral amygdala. In cortical slices, a bipolar tungsten stimulating electrode was placed in layer Ⅴ, and extracellular field potentials were recorded using a glass microelectrode placed in layer Ⅱ/Ⅲ. Synaptic responses were elicited at 0.02 Hz. For inducing LTP, we used five trains of theta-burst stimulation (TBS) at the same intensity of testing stimulation (each train contains four pulses at 100 Hz; delivered at 200 ms interval). We found that this protocol induced reliable LTP in the auditory cortex and amygdala (see Results).

1.6 Immunocytochemistry

Mice were deeply anesthetized with sodium pentobarbital (50 mg/kg) and transcardially perfused with heparinized saline (100 000 IU/L heparin, 0.1 mol/L PBS; 0.9% NaCl) followed by 4% paraformaldehyde in 0.1 mol/L PBS, pH 7.4. Brains were removed and stored in the same fixative overnight at 4℃, then cryopreserved in 30% sucrose in PBS buffer. Slices (14 μm) from frozen sections of the entire brain were cut. The primary antibodies used in this study were directed against the following antigens (using the stated dilutions): astrocytes [glial fibrillary acidic protein (GFAP) rabbit polyclonal, 1:4; Incstar, Stillwater, Minn.]; neurons [neuronal nuclear antigen (NeuN) mIgG1, 1:500]; Species-specific secondary antibodies (1:200 dilution) were conjugated to Cy3, fluorescein isothiocyanate (Jackson Immunoresearch, West Grove, Penn.) or Alexa 488 (1:200 dilution; Molecular Probes). The samples receiving Hematoxylin-eosin were embedded in paraffin, cut in sections 4 μm thick and stained.

1.7 Data analysis

Results were expressed as means ± SEM. Statistical comparisons were made with one- or two-way analysis of variance (ANOVA) with the Student-Newmann-Keuls test used for post hoc comparisons. In all cases, P<0.05 was considered statistically significant.

2 RESULTS

2.1 Anatomy of memory-related central regions in Egr-1 knockout mice

In general, Egr-1 knockout mice are visually indistinguishable from wild-type littermates. To determine whether Egr-1 knockout mice have neuroanatomic abnormalities in central regions related to sensory transmission and fear memory, we carried out histochemical experiments in several brain areas. Analysis of serial coronal sections, examined by light microscopy, showed no detectable morphological differences in the auditory cortex, amygdala and hippocampus. Higher magnification of the stained sections further demonstrated no apparent differences in the number and distribution of cells in these areas ( Fig.1). A recent study in the hippocampus showed that both neuronal and glial cell populations were not affected in Egr-1 knockout mice^[42]. We wanted to confirm this finding in other central areas such as the amygdala. As shown in Fig.2, we found no difference in neuronal population and distribution between Egr-1 knockout and wild-type mice. We also used GFAP as a marker of glial cells. GFAP staining demonstrated that glial cells were similar between Egr-1 knockout and wild-type mice.

2.2 Contextual and auditory fear memory

Previous studies show that fear conditioning activates Egr-1 in the amygdala, a structure critical for fear memory. However, no study has reported a change in short or long term fear memory in mice lacking Egr-1. We assessed two forms of associative emotional memory in wild-type and Egr-1 knockout mice: contextual and auditory fear conditioning. Preliminary studies in wild-type mice show that three shock-tone pairings produced long-term fear memory that lasted for at least 2 weeks after conditioning (Fig.3, n=6). Pairing the tone with a single shock resulted in fear memory that lasted for about 1~3 d after training (Fig.4, n=6). We next measured fear responses from Egr-1 knockout mice during contextual and auditory conditioning 1 h, 1 d, 1 and 2 weeks after receiving multiple shock-tone pairings (Fig.3). As shown in Fig.3A, there was a significant reduction in the freezing response during auditory conditioning in Egr-1 knockout mice as compared to wild-type mice [F(1, 50)=19.5, P<0.001] (n=6 for each group). Post hoc analysis revealed that the reduction was selective for late responses (P<0.01 for day 3, P< 0.005 for 1 week and P<0.01 for 2 weeks), while early responses (P=0.6 for 1 h and P=0.4 for 1 d) were not significantly different between wild-type and mutant mice. Unlike auditory fear memory, contextual memory was not significantly different between genotypes [F(1,50)=3.5, P=0.07](Fig.3B). This finding indicates that Egr-1 preferentially contributes to late auditory fear memory.