Enhancement of long-term memory retention and short-term synaptic plasticity in cbl-b null mice

Abstract

The cbl-b gene is a member of the cbl protooncogene family. It encodes a protein with multiple domains, which can interact with other proteins in a variety of signaling pathways. The functions of cbl family genes in the brain are unknown. In this report, we used genetic, immunohistochemical, behavioral, and electrophysiological approaches to study the role of cbl-b in learning and memory. Cbl-b null mice developed normally and had no abnormalities in their locomotor performance. In spatial learning and memory studies, cbl-b null and WT mice performed similarly during training. To test memory retention, two probe trials were used. cbl-b null mice performed slightly better 1 day after training. However, in the probe trial 45 days after training, the cbl-b null group showed significantly higher memory retention than WT mice, suggesting an enhancement of long-term memory. Using electrophysiological approaches, we found there was enhanced paired-pulse facilitation in the Schaffer Collateral-CA1 glutamatergic synapses of the cbl-b null mice. On the other hand, there was no difference in long-term potentiation between the two groups of mice. In summary, we provide evidence that (i) cbl-b protein is concentrated in the synaptic regions of CA1, CA3, and the dentate gyrus of the hippocampus; (ii) cbl-b null mice have enhanced long-term memory; and (iii) cbl-b null mice show an enhancement in short-term plasticity. These results indicate that cbl-b is a negative regulator of long-term memory, and its neuronal mechanism regulates synaptic transmission in the hippocampus.

With the human and mouse genomes fully sequenced, many candidate genes that may be related to the process of learning and memory are available for gene targeting and other mutagenesis studies. Analysis of the behavioral, physiological, and biochemical consequences of these mutations will help implicate molecular pathways important for learning and memory, as well as their underlying synaptic mechanisms.

Cbl-b belongs to the cbl-protooncogene family. It contains multiple functional domains, including a tyrosine kinase-binding (TKB) domain, a RING-finger domain, and a proline-rich region, as well as an ubiquitin-associated domain. The TKB domain has been shown to recognize specific phosphotyrosine residues on activated tyrosine kinases (1, 2). The RING-finger domain is the site whereby cbl family proteins recruit ubiquitin-conjugating enzymes, which add ubiquitin to targeted proteins. These domains are required for cbl family proteins to regulate signaling transduction and protein degradation. C-cbl and cbl-b genes are strongly expressed in thymus, testes, and other tissues (3). Gene targeting in mice has shown that cbl-b is involved in pivotal events of lymphocyte activation (4, 5).

Although earlier Northern hybridization demonstrated that cbl-b mRNAs are expressed in many tissues, including the brain (3), the distribution of the cbl-b protein and its function in the brain are still unknown.

To understand the role of the cbl-b protein during mouse learning and memory, we hypothesize that, like other signaling proteins encoded by protooncogenes, it is expressed in the brain and participates in modulating neuronal activities, including learning and memory. Here we show that the cbl-b protein is located in the hippocampal region. Cbl-b null mice have normal motor coordination and learning in the Morris water-maze task but enhanced retention of long-term memory. These results suggested that cbl-b is a negative regulator in long-term memory. To further investigate synaptic mechanisms, we studied paired-pulse facilitation (PPF) and long-term potentiation (LTP). We found that PPF of glutamatergic synaptic transmission was significantly enhanced in CA1 neurons in the cbl-b null mice, whereas LTP had no change. These findings further suggest that glutamatergic transmission is facilitated and may contribute to the enhanced long-term memory retention found in cbl-b null mice.

Results

Distribution of the cbl-b Protein in the Mouse Brain.

Using a cbl-b-specific antibody (c-20) against the cbl-b C terminus, not crossreacted to cbl, we located the cbl-b protein in several regions of the mouse brain, where the cbl-b protein was located (bright light-stained structures) in the hippocampus, the cerebellum (data not shown), and some nuclei in the midbrain (data not shown). In the hippocampus, cbl-b was detected in CA1, CA3 (Fig. 1A), and the dentate gyrus (DG) (Fig. 1D). Remarkably, there was a high level of cbl-b in the dendrites or the regions surrounding the dendrites of the neurons of CA1, CA3, and DG (Fig. 1 B–D). On the other hand, there was a relatively low level of cbl-b (dark areas) in the somatic compartments of the neurons in the CA1, CA3, and DG regions (Fig. 1 B–D). This specific cellular distribution suggests a role for cbl-b in synaptic transmission received by the dendrites.

Fig. 1.

Distribution of cbl-b immunoreactivity in mouse hippocampus. (A) Localization of cbl-b immunoreactivity, stained as the bright fluorescence, in the hippocampus. The CA1 and CA3 regions are labeled. The highest concentration is in the dendritic compartment of pyramidal neurons. (B) Enlarged image of the CA1 region of the hippocampus. (C) Enlarged CA3 region. Cbl-b immunoreactivity is concentrated in the dendritic compartment of the pyramidal cells, sparing the cell bodies (darker area). (D) Cbl-b immunoreactivity in the dentate gyrus of the hippocampus. (Scale bars, 100 μm.)

Overall Condition and Locomotor Activities of cbl-b Null Mice.

The cbl-b null mice (Fig. 2), which were used in the experiment, are five to six generations into the C57BL/6 and contain 96.88–98.44% genetic background from the C57BL/6 strain. This backcrossing will reduce the behavioral variation caused by the different backgrounds of different strains of mice. Overall, cbl-b null mice developed normally without any visible defect. They appeared normal when they walked and swam. To test muscle strength, motor coordination, and balance, the hanging wire, vertical pole, and rotarod tests were used. In the hanging wire test, cbl-b null mice gripped the wires tightly and stayed without falling down, as did WT mice (58.5 ± 0.97 vs. cbl-b−/− 58.1 ± 1.22, P > 0.05, t test; Fig. 3A). In the vertical pole test, both WT and cbl-b null mice held the pole firmly while the pole reached a 90° angle. Both were able to maintain their position on the rod (58.13 ± 1.07 vs. cbl-b−/− 57.56 ± 1.43, P > 0.05; Fig. 3B). In the rotarod test, both WT and cbl-b null mice were able to keep from falling by walking forward on the rotating cylinder. There was no difference between the two groups (156.86 ± 8.47 vs. cbl-b−/− 170.17 ± 6.62, P > 0.05; Fig. 3C). To test swimming ability, both cbl-b null and WT mice were placed in the water tank for 2 min. The swimming patterns and speeds were recorded and analyzed. Both mice showed similar swimming patterns, and their swimming speeds showed no significant difference (21.14 ± 0.73 vs. cbl-b−/− 20.13 ± 0.72, P > 0.05; Fig. 3D). These results demonstrated that cbl-b null mice had adequate muscle strength, motor coordination, and swimming ability.

Fig. 2.

Genotyping of cbl-b mice. The upper arrow represents the mutant band, and the lower arrow represents WT band. 1 kb, 1-kb DNA ladder.

Fig. 3.

The locomotor activities of WT and cbl-b null mutant mice. The comparison of locomotor activities using hanging wire test, n = 10/each group (A); vertical pole test, n = 16/each group (B); rotarod test, n = 18 per group (C); and swimming speed, n = 15 per group (D). The white columns represent WT mice, and the black columns represent cbl-b null mutants. Vertical bars on the top of the columns indicate SEM.

Cbl-b Null Mice Had Normal Learning Ability but an Enhancement of Long-Term Memory.

In the spatial learning experiment, the Morris water maze was used as a model. The performance of each mouse in the swimming tank was monitored and recorded by a water 2020 HVS Image (Middlesex, U.K.) system and analyzed by the software. Both cbl-b null and WT B6 mice were trained for a series of 7 days to locate the hidden platform under the water. Cbl-b null and WT mice performed similarly, because the learning curves of both groups showed no significant difference (Fig. 4A). To test memory retention, each trained mouse was tested in probe trials 1 and 45 days after training. Both WT and cbl-b null mice performed well in the probe trial 1 day after training, although cbl-b null mice did slightly better than WT. This difference did not reach statistical significance (Fig. 4B). To assess longer-term memory, the probe trial was administered 45 days after training. The cbl-b null group still showed a clear preference for the target quadrant (40.4%) [F (3, 44) = 13.68, P < 0.0001, ANOVA], whereas the WT group did not show this preference for the target quadrant (29.5%) [F (3, 64) = 2.354, P = 0.079, ANOVA]. It was clear that cbl-b null mice had significantly greater memory retention than WT mice (Fig. 4C). These data demonstrated that there was an enhancement of long-term memory in the cbl-b null mice.

Fig. 4.

The effect of cbl-b on learning and memory. (A) Morris water-maze performance of WT (n = 21) and cbl-b (n = 14) null mice. The line with green circles represents WT, and the line with red squares represents cbl-b null mutant mice. (B) The probe trial tested 1 day after training. The green columns represent WT (n = 21), and the red, cbl-b null mice (n = 13). (C) The probe trial performed 45 days after training, WT (n = 17) and cbl-b null mice (n = 12). In both probe trials, Quadrant 2 is the target quadrant. ∗∗, P = 0.02.

Cbl-b Null Mice Showed Enhanced Paired-Pulse Facilitation.

To investigate synaptic mechanisms of long-term memory that involve the cbl-b protein, we studied short-term plasticity of glutamatergic and GABAergic synapses in the CA1 region of the hippocampal slices of WT, cbl-b+/−, and cbl-b null mice. In this study, we used two age groups, 9- and 2-month-old mice. In 9-month-old mice, when the paired-pulse interval was 30 ms, the excitatory postsynaptic current (EPSC)2/EPSC1 ratio in WT mice was 168%, whereas the EPSC2/EPSC1 ratio in cbl-b null mice was 277% (t = 3.091, P < 0.01). When the intervals increased to 50 and 100 ms, the differences of EPSC2/EPSC1 ratio of both mice were still significant (t = 3.24, P < 0.01 and t = 2.18, P < 0.05, respectively) (Fig. 5A and B). However, the baseline EPSC1s were not significantly different between groups (Fig. 5C). Therefore, the increased ratio of EPSC2/EPSC1 in cbl-b null mice was caused by an increment of EPSC2 rather than a reduction of EPSC1. The inhibitory postsynaptic current (IPSC)2/IPSC1 ratios in both groups of 9-month-old mice were statistically the same at 30-, 50-, and 100-ms intervals (Fig. 5D). In 2-month-old mice, when the paired-pulse interval was 30 ms, the EPSC2/EPSC1 ratio in cbl-b+/− was 213%, whereas the ratio in cbl-b null mice was 384%. When the interval was 50 ms, the ratio in cbl-b+/− was 211%, whereas in cbl-b null mice, it was 348%. The differences within both interval groups were statistically significant (P < 0.01 and P < 0.05, respectively) (Fig. 5E). However, the IPSC2/IPSC1 ratios in both 2-month-old mice were not statistically different (data not shown). These data showed clearly that cbl-b null mice have enhanced paired-pulse facilitation, and this facilitation is specific to the glutamatergic synapses.

Fig. 5.

The enhancement of paired-pulse facilitation in the glutamatergic synapses of the CA1 neurons in the cbl-b null mutant mice. (A) The representative diagrams of the paired-pulse facilitation of WT (Left) and cbl-b null mice (Right) at interpulse interval of 30 ms. The paired stimulus pulses occurred as indicated by the vertical transients. (B) The paired-pulse facilitation in 9-mo-old WT and cbl-b null mice. The green columns represent WT (n = 16), and the red columns represent cbl-b null mice (n = 18). (C) The comparison of EPSC1, the EPSC evoked by the first pulse, of glutamatergic synapses of the CA1 neurons of the WT and cbl-b null mice. (D) No difference in the transmission of GABAergic synapses between the WT and cbl-b null mutant mice. (E) Comparison of paired-pulse facilitation between the neurons of 2-mo-old cbl-b+/− and cbl-b null mice. The yellow columns represent cbl-b+/− mice (n = 8), and the red columns, cbl-b null mice (n = 7). ∗, P < 0.05; ∗∗, P < 0.01.

LTP.

LTP was induced in 2-month-old WT and cbl-b null mice. Excitatory postsynaptic potentials (EPSPs) of both types of mice had equivalent slopes in the control period (−175.7 ± 10.6 mV per sec, mean ± SEM, n = 5; null −178.4 ± 14.90 mV per sec, n = 5; P > 0.88, t test) (Fig. 6B). The synaptic response was similarly potentiated by tetanic stimulations to an equivalent degree (posttetanic potentiation at its peak: WT 1.974 ± 0.162, normalized to control values; Null 1.763 ± 0.130; P > 0.34) (Fig. 6C). In the time segment between 45 and 60 min, EPSPs showed sustained potentiation to an equivalent degree (LTP: WT 1.241 ± 0.027; Null 1.243 ± 0.050; P > 0.97) (Fig. 6C). Although there was more variability in the posttetanic phase as compared with the later sustained phase for both genotypes, there were no differences in the size and kinetics of LTP induced in null and WT mice (Fig. 6C).

Fig. 6.

Comparison of LTP induced in CA1 striatum radiatum of hippocampal slice preparation from WT and cbl-b null (null) mice. (A) Electrode configuration for the LTP study. Stimulation (Stim) and recording (Rec) electrodes are inserted to the depth of 200 μm from the surface of 400-μm-thick slice preparation. (B) Example EPSPs by test-pulse stimulations before and after the tetanic protocol. EPSP intensity was quantified as the onset slope. Four traces each before (open triangles) and after (filled triangles) the start of tetanic stimulations (downward arrows, Inset) were superimposed (low-pass-filtered at 500 Hz post hoc for presentation purposes), mean trace for each four was calculated (not shown), and corresponding onset slopes (control and posttetanic, respectively) were extrapolated from quasilinear portions (width, 1 ms). (C) Mean EPSP time courses in WT (filled circles, n = 5) and null (open circles, n = 5). Time 0 (0 min) corresponds the last control test pulse before tetanic stimulations. Vertical bars indicate SEM.

Discussion

In previous studies, the use of genetically altered mice to study learning and memory has enhanced our knowledge of how genes regulate these processes.

Here, we have shown with immunohistochemical labels that the cbl-b protein is located preferentially in specific regions of the mouse brain. High expression regions include the hippocampus, cerebellum, and some nuclei of the midbrain regions. Within the hippocampal region, cbl-b is located in CA1, CA3, and dentrate gyrus. In the CA1, CA3 regions, the cbl-b protein has a unique cellular distribution. The immunostaining of cbl-b was very low in the somatic region but substantially higher in the distal part of the dendritic area of the pyramidal cells. This unique cellular localization strongly suggests a role of cbl-b in synaptic transmission. Although in this study we have located cbl-b in the synaptic regions of the CA3 axonal–CA1 dendritic region, it is not clear whether the cbl-b protein is located in pre- and/or postsynaptic compartments. Our next step will be a more detailed localization, which can be determined by using techniques such as electron microscopy, as well as coimmunolocalization of cbl-b protein with pre- or postsynaptic marker proteins.

Measurement of learning and memory in animal models depends on the performance of animals in well defined tasks. In spatial learning tasks, such as the Morris water-maze task, mice must use their motor ability to perform learning tasks. Therefore, the integrity of motor function is critical for their learning performance. In this report, we have shown that cbl-b null mice showed normal motor coordination and locomotive activity in the hanging wire, vertical pole, rotarod, and swimming tests. Therefore, the observed enhancement in the probe test after 45 days can be attributed to a specific effect of cbl-b on long-term memory.

Although the synaptic mechanisms of learning and memory have not been fully revealed, it is generally accepted that short- and long-term synaptic modifications must be involved. A well characterized form of short-term plasticity is paired-pulse facilitation frequently analyzed at the Schaffer collateral–CA1 pyramid cell synapses (6, 7). Our electrophysiological data showed enhanced paired-pulse facilitation in Schaffer collateral–CA1 synaptic transmission in the brain slice of cbl-b−/− mice. This facilitation occurred for the glutamatergic, but not GABAergic, synapses. Because paired-pulse facilitation is mainly involved in presynaptic function and related to increasing neurotransmitter release, the result suggests there is a facilitation of release of glutamate from the glutamatergic synapses onto the postsynaptic neurons of cbl-b null mice. Knocking out other mouse oncogenes resulted in a variety of phenotypes, mainly causing deficits in learning and memory. In this report, we showed the cbl-b null mutant had an enhancement in long-term memory and paired-pulse facilitation. It is clear that in WT mice, cbl-b protein plays a role as a negative regulator, as shown in the immune system (8, 9). Presynaptic release of neurotransmitters significantly depends on presynaptic Ca²⁺ concentrations (7, 10). Protein sequence analysis reveals that cbl-b has an EF-hand domain within its conserved N terminus. The EF-hand domain is shown as a Ca²⁺-binding domain in many proteins (11). Therefore, it is possible that one role of the cbl-b protein is to regulate the concentration of Ca²⁺ in the presynaptic regions of the Schaffer collateral–CA1 glutamatergic synapses. Another possible mechanism may involve regulation of presynaptic membrane channels (e.g., K⁺ channels) that in turn control transmitter release.

Our results showed no difference in the size and kinetics of LTP induced in cbl-b null and WT mice. Therefore, LTP does not appear to be the main mechanism responsible for the enhancement of long-term memory shown in cbl-b null mice.

Cbl family proteins are substrates of many receptor tyrosine kinases (RTKs) and can be activated by phosphorylation (1, 2). Recent reports show that c-cbl protein acts as a ubiquitin ligase (E3 enzyme) that binds and forms a complex with the RTKs and then facilitates the transfer of ubiquitin from ubiquitin-conjugating enzymes to the RTKs (12, 13). This E3 enzyme activity results in the monoubiquitylation or multiubiquitylation and further degradation of RTKs in the lysosome or proteasome. This controlled termination of RTK signaling has emerged as a key control for fine-tuning RTK signaling. C-cbl and cbl-b proteins have a high degree of sequence homology throughout whole molecules. They share several functional domains, which include the tyrosine kinase-binding (TKB), RING-finger, and praline-rich domains, as well as a ubiquitin-associated/LZ domain (1, 2). This similarity at the molecular level implicates a functional similarity between both proteins. Indeed, recently, cbl-b has been identified as a RING-type E3 ubiquitin ligase to target signaling proteins in T cells (14). To date, the molecular mechanisms of the cbl-b protein in the brain are still unknown, but our data suggest that cbl-b is a negative regulator, and it may work in a similar way to contribute to the degradation of RTKs in neurons. Additional experiments are needed to determine which RTK will be targeted by cbl-b. One candidate is the receptor for brain-derived neurotrophic factor, which is a RTK in the brain and has been shown to rapidly potentiate synaptic transmission at glutamatergic synapses by enhancing transmitter release (15).

Our study revealed that cbl-b null mutants had only enhanced long-term memory and an unchanged learning ability compared with WT mice. This suggests cbl-b may play an important role in the process of long-term memory retention after, but not during, the learning process. Although learning and memory are closely related processes, there should be some unique molecules or pathways used specifically for each process. How long-term memory is recorded, restored, and recalled at the molecular level is not clear. Changes in synapses, such as strengthening preexisting synapses and forming new synapses, are certainly required. That cbl-b null mice showed significantly more improved memory than WT mice after a longer delay (45 days), and that cbl family proteins are associated with RTKs of many growth factors suggest a hypothesis that the cbl-b protein may participate in the process that leads to morphological changes in the synapses of neurons during memory processes. Additional studies of cbl-b null mice after learning are needed to reveal how this unique regulatory mechanism is involved in the formation, storage, and retention of memory.

Materials and Methods

Cbl-b Null Mice Maintenance and Genotyping.

All mice were housed in regular cages on a 12-h light/12-h dark schedule with access to food and water ad libitum. Cbl-b null mice were created by using homologue recombination (4). To reduce the variation caused by different genetic backgrounds, cbl-b mice were backcrossed to C57BL/6 mice for five to six generations before use in this study. Cbl-b+/− and cbl-b null mice were produced by crossing cbl-b+/− and/or cbl-b null mice.

For genotyping, the genomic DNAs of each mouse were extracted from the tip of the tail. DNA amplification was performed with a modified protocol (16) by using puReTaq Ready-To-Go PCR Beads (Amersham Pharmacia Biosciences). Three oligonucleotides were used as two sets of PCR primers (only + strand is shown) to identify the mutant and WT amplified fragments (5′-GAATGGCAGAAATCCTGGTG-3′, 5′-TTCCTCGTGCTTTACGGTAT-3′, and 5′-CCCAGCAAAAGTAGCCAATG-3′). The PCR products were electrophoretically separated in agarose gel and stained with ethidium bromide, imaged, and recorded with FUJIFILM LAS-1000 Gel Documentation system (Fuji, Tokyo). Cbl-b−/− and WT have a fragment of 700 and 432 bp, respectively; cbl-b+/− have both fragments (Fig. 2).

Immunohistochemical Staining.

Mouse brains were dissected and fast-frozen on dry ice. Using a cryostat, a series of brain slices were sectioned at 10 μm and stored at −80°C before use. For immunohistochemical staining, the slices were thawed and fixed with 4% formaldehyde in PBS (pH 7.4) for 5 min. After being washed three times using PBS, the slices were permeabilized with 1% Triton X-100 in PBS (pH 7.4). Heat-inactivated 10% normal horse serum in PBS (pH 7.4) was used to block the nonspecific binding sites. For cbl-b protein staining, an anti-cbl-b antibody, C-20: sc-1435 (Santa Cruz Biotechnology) with 1:200 dilution was incubated with slices at room temperature for 8 h. After the slices were washed with PBS three times, a secondary anti-goat IgG conjugated with fluorescein (Vector Laboratories) was added and incubated for 1 h in the dark at room temperature. Slices were washed three times before being sealed with VECTASHIELD (Vector Laboratories). Slices were observed under a Nikon fluorescence microscope, and images were recorded by using the manufacturer’s software.

Testing Motor Activity.

To measure the motor strength, balance, and coordination of mice, a series of tests were performed by using a standardized protocol (ref. 17, pp. 52–55). For the hanging wire test, a mouse was placed onto a standard wire cage lid. After the mouse gripped the lid, the wire cage lid was slowly turned upside down, and the length of time the mouse held onto the wire was recorded.

To test the motor balance ability of the mice, a vertical pole test was used. A mouse was placed onto a horizontally positioned wood pole (2 cm in diameter and 50 cm long). After mouse placement, the pole was raised gently and slowly to a 90° position. The length of time for the mouse to fall off the pole was recorded. A cutoff time of 1 min was used for the hanging wire and vertical pole tests.

To test motor coordination, a mouse was placed on a rotating cylinder (3 cm diameter) and given two practice trials. The rod was rolling at a standard speed of five revolutions per minute. To maintain balance, the mouse needed to adjust its position by constantly walking forward. If the mouse fell, an automatic timer in the equipment recorded the staying time on the rod. For this test, a 3-min cutoff time was used.

To gauge swimming ability, the swimming speed of each mouse was recorded automatically by a camera during the first day of water-maze training and analyzed by using water 2020 software.

The Morris Water-Maze Task.

The swimming pool was a circular metal tank 127 cm in diameter, painted white. The hidden platform, 10 cm in diameter, was clear Plexiglas and submerged 1 cm below the surface of the water. To make the platform invisible, the water was colored white by using powdered milk. Extramaze visible cues of various geometric shapes were posted on the walls of the room. The water-maze procedures were slightly modified versions of those described (ref. 17, pp. 87–95). Adult male mice, 3- 6 months old, were trained in the water-maze task for 7 consecutive days, and each training day included six trials per mouse, separated by 10-min intervals. The starting points for mouse placement in each training episode were varied. Each mouse was allowed to search for the underwater platform for 120 sec. After landing on the hidden platform, each mouse was allowed to remain for 30 sec before being returned to its cage. Mice that failed to land on the platform by themselves within the time limit were manually guided to it.

To determine mouse memory retention of the hidden platform position during the 7 days of training, two probe trials, 1 and 45 days after the training, were carried out. In each trial, mice were placed in the water and allowed to search for 1 min. During the training sessions and probe trials, the performance of each mouse was tracked by an overhead video camera and observed by the experimenter through a monitor in a corner of the room, where swimming mice could not see the experimenter. The data were analyzed by using the water 2020 software (HVS Image).

Paired-Pulse Facilitation.

Slice preparation.

Transverse hippocampal slices (300 μm thick) were prepared from WT, cbl-b+/−, and cbl-b−/− mice, which were provided to the experimenter randomly and blindly. After sedation with isoflurane, the mouse was decapitated and its brain bisected sagittally and removed into ice-cold artificial cerebrospinal fluid (aCSF) containing: 124 mM NaCl, 3 mM KCl, 1.25 mM NaH₂PO₄, 1.3 mM MgSO₄, 26 mM NaHCO₃, 2.4 mM CaCl₂, and 10 mM glucose, saturated with 95% O₂/5% CO₂, at pH 7.4. Transverse hippocampal slices were cut in ice-cold modified aCSF, containing decreased Ca²⁺ (0.5 mM) and increased Mg²⁺ (4 mM), with a VT 1000S microtome (Leica, Deerfield, IL). Slices were then placed on a nylon mesh, submerged in normal aCSF bubbled with 95% O₂/5% CO₂ continuously, and allowed to recover from slicing trauma for at least 1 h at 32°C before being stored at room temperature for later recordings.

Electrophysiological recordings.

One slice was transferred to the perfusion chamber on a Burleigh Gibraltar fixed stage system (Burleigh Instruments, Fisher, NY) and perfused with aCSF at a rate of 2–3 ml per min at room temperature. CA1 pyramidal cells were identified visually by using an Axioskop 2FS microscope (Zeiss) equipped with a ×40 water-immersion objective coupled with an infrared differential interference contrast camera system. Whole-cell patch-clamp recordings were made with a dual-headstage MultiClamp 700A amplifier (Axon Instruments, Union City, CA). Membrane current and potential signals were digitized and analyzed with Digidata 1322A and pClamp 8.0 systems (Axon Instruments). Patch pipettes of ≈5 MΩ were pulled with a Narishige PP-830 puller and fire-polished with a Narishige MF-83 microforge (Narishige, Greenvale, NY). The pipette solution had the following composition unless otherwise stated: 90 mM K⁺ gluconate, 45 mM KCl, 1.7 mM NaCl, 0.1 mM CaCl₂, 2.7 mM MgCl₂, 10 mM Hepes, 1.1 mM EGTA, 5 mM phosphocreatine-Na⁺, 3.5 mM ATP-K⁺, 0.3 mM GTP-Na⁺, pH 7.2, and 290 mOsm. Series resistance and membrane capacitance were compensated ≈70% and monitored during experiments. Precise positioning of recording and stimulation electrodes was achieved with PCS-5400 piezoelectric micromanipulators (Burleigh Instruments). To activate GABAergic interneuron-to-pyramidal cell synapses, a concentric bipolar stimulation electrode was placed at the stratum pyramidal layer approximately <100 μm away from the pyramidal neurons being recorded. The stimulation strength was set at 5 μA above the current necessary to evoke just observable synaptic responses, and the duration was fixed at 200 μs. 6-Cyano-7-nitroquinoxaline-2,3-dione and 2-amino-5-phosphonovaleric acid (10 μM each) were included in the perfusion solution to block glutamatergic contamination. Glutamatergic synaptic responses were evoked by stimulating the Schaffer Collateral pathway in the presence of 10 μM bicuculline. Paired-pulse protocols were composed of two identical stimulation pulses separated by various delays, as indicated in the result and Fig. 5.

LTP.

Hippocampal slices were prepared from either WT or Cbl-b null (Null) mice, which were provided to the experimenter randomly and blindly. Mice were sedated with isoflurane and decapitated. The hippocampus was dissected out and placed in refrigerated aCSF, which consists of: 119 mM NaCl, 1.3 mM MgSO₄, 2.5 mM CaCl₂, 1 mM NH₂PO₄, 26.2 mM NaHCO₃, and 11 mM d-glucose, equilibrated with 5% CO₂ and 95% O₂. Transverse slices (400 μm thick) were prepared with a Vibratome, transferred to a small recording chamber (200 μl; modified from RC-26LP, Warner Instruments, Hamden, CT) on a fixed-stage microscope (Zeiss Axioskop 2FS Plus) and superfused at the rate of 4 ml per min with aCSF at 23°C. This temperature was chosen for slower synaptic kinetics, which could be faster at higher temperatures (e.g., 27°C or 32°C) and overlapped with stimulation artifacts in preliminary experiments, adapted from a room-temperature protocol (18). Preincubation for at least 1 h was allowed before recording. Recording and stimulation with glass-pipette electrodes (≈7 MΩ, filled with aCSF) were inserted into the slice to the depth of 200 μm from the slice surface. Their tips were positioned at a distance of ≈150 μm from each other within the striatum radiatum of the dorsal CA1 region (Fig. 6A). The electrode closer to CA2/CA3 was assigned for stimulation, the other for recording. Field potentials were recorded with the amplifier EPC-9 (Version D, HEKA Electronics, Lambrecht/Pfalz, Germany) in the fast current-clamp recording mode with zero holding current. Potential was sampled at 10 kHz (pulse software, HEKA Electronics) and digitally filtered at 3 kHz for real-time display. The stimulator (Master-8, AMPI) was driven by the EPC-9 by using pulse software (HEKA Electronics), and 0.4-ms pulses were delivered via a constant-current isolator (PS/U6, Grass Instruments, Quincy, MA). Current delivered through the electrode was monitored as a voltage drop across a 10-kΩ resistor positioned serially in the stimulation circuit. Stimulation intensity was adjusted so that field EPSP showed an onset slope of approximately −175 mV per sec measured at the quasilinear portion after onset (Fig. 6B), which occurred at a latency of ≈5 ms after the onset of stimulation pulse. Single-pulse stimulation (test pulse) was delivered at an interstimulus interval of 30 sec. After stable EPSPs were recorded for 15 min, LTP was induced by a tetanic stimulation protocol: two 1-sec trains of pulses at 100 Hz delivered at a 30-sec interval, with intermittent intervening test pulses (Fig. 6B Inset, arrows). EPSPs were normalized to mean control values and recorded for 60 min after the tetanic stimulation (Fig. 6C). A data point was omitted from the mean calculations for the time point if the control potential (before test stimulus) in the sweep was not stable (one WT point at 51.5 min in Fig. 6C).

Statistic Analyses.

The data in this study were presented as mean ± SEM and were subjected to either ANOVA or a t test. P < 0.05 was considered statistically significant.

Abbreviations

LTP

long-term potentiation

RTK

receptor tyrosine kinase

aCSF

artificial cerebrospinal fluid

EPSC

excitatory postsynaptic current

IPSC

inhibitory postsynaptic current

EPSP

excitatory postsynaptic potential.