The Journal of Neurophysiology Vol. 78 No. 2 August 1997, pp. 1030-1044
Copyright ©1997 by the American Physiological Society

Sequence of Single Neuron Changes in CA1 Hippocampus of Rabbits During Acquisition of Trace Eyeblink Conditioned Responses

Matthew D. McEchron and John F. Disterhoft

Department of Cell and Molecular Biology and Institute for Neuroscience, Northwestern University Medical School, Chicago, Illinois 60611

	ABSTRACT

Abstract Introduction Methods Results Discussion References

McEchron, Matthew D. and John F. Disterhoft. Sequence of single neuron changes in CA1 hippocampus of rabbits during acquisition of trace eyeblink conditioned responses. J. Neurophysiol. 78: 1030-1044, 1997. The sequence of changes in single neuron activity in the CA1 area of the rabbit hippocampus was examined during daily sessions (80 trials/session) of hippocampally dependent nonspatial trace eyeblink (i.e., nictitating membrane response) conditioning. Each trial for trace conditioned animals (n = 7) consisted of a tone conditioned stimulus (CS; 6 kHz; 90 dB, 100 ms) followed by a 500-ms silent trace period, then a corneal airpuff unconditioned stimulus (US; 3.0 psi; 150 ms). Control animals(n = 5) received unpaired CSs and USs. Most pyramidal (n = 309) and theta (n = 21) cells were recorded for a single day of training. The activity of cells for each animal were grouped according to: the day of training that CRs began to increase and the day of training that CR performance became asymptotic. Pyramidal cells from trace conditioned animals demonstrated several stages of learning-related activity: large increases in activity after both the CS and US early in conditioning on the day of training when CRs began to increase, smaller moderate increases in activity on the following days of training, and decreases in activity after the US during asymptotic CRs. Pyramidal cell-increases declined significantly across the trials of each daily session. Theta cells showed an activity pattern opposite to the pyramidal cells, consistent with the notion that theta cells have an inhibitory influence on pyramidal cells. Single pyramidal cells also were categorized into response profiles. Most pyramidal response profiles showed increases in activity specific to the day of initial CRs. Two of the pyramidal response profiles may be involved in assessing the temporal properties of the CS-US trace conditioning trial.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Experimental and clinical investigations have implicated the hippocampus and related structures as having important functions in learning and memory (Rempel-Clower et al. 1996; Squire 1992). A number of hippocampally dependent behavioral paradigms have been developed to examine the neural mechanisms of learning. The rabbit eyeblink (i.e., nictitating membrane response) classical conditioning paradigm developed by Gormezano and associates (1962) is one of these paradigms that has proven to be effective in this area of research (Disterhoft et al. 1994; Thompson et al. 1976). The methods and control procedures for eyeblink conditioning have been tested thoroughly, allowing the neural substrates of learning to be analyzed with respect to well-defined and precisely controlled nonspatial stimuli (for review, see Gormezano et al. 1983). In this paradigm a conditioned stimulus (CS; usually a tone) is followed by an unconditioned stimulus (US; usually a corneal airpuff). After repeated pairings, rabbits learn to associate these stimuli and begin to exhibit conditioned eyeblink responses (CRs) to the CS, which precede the US. Lesion studies have shown that the hippocampus is essential for learning this task when a silent 500-ms trace interval separates the CS from the US (trace conditioning) but not when the CS coterminates or overlaps with the US (delay conditioning) (Moyer et al. 1990; Schmaltz and Theios 1972; Solomon et al. 1986). A recent lesion study by Kim and colleagues (1995) found that the hippocampus was also essential for the consolidation but not the retention of trace eyeblink CRs. In their study, lesions placed in the hippocampus 1 day after asymptotic levels of CRs were reached disrupted the consolidation and performance of CRs, whereas lesions placed 1 mo after conditioning had no effect on CR performance. Their study, along with a number of other lesion studies (Akase et al. 1989; Kim and Faneslow 1992; Zola-Morgan and Squire 1990), suggests that the hippocampus may have a time-limited role in trace eyeblink conditioning and that important changes in neural activity may occur early during the periods of acquisition and consolidation of CRs.

A number of in vivo electrophysiological studies have examined the neuronal activity of the hippocampus during eyeblink conditioning, beginning with a study by Berger, Alger, and Thompson (1976). This group conducted a series of studies examining the multiple- and single-unit activity of the pyramidal cell layer of the hippocampus during delay eyeblink conditioning (Berger and Thompson 1978b; Berger et al. 1983). Although the delay task was not hippocampally dependent, they found that pyramidal cell activity increased during the initial trials of training. These increases in activity paralleled the amplitude-time course of the behavioral response and preceded it temporally. Our laboratory recently examined CA1 single neuron activity in animals overtrained in trace eyeblink conditioning (Weiss et al. 1996); however, the activity pattern of CA1 single neurons during the early periods of hippocampally dependent trace eyeblink conditioning is still unknown. Therefore, the present study sought to examine the time course of CA1 single neuron changes during the acquisition of trace eyeblink CRs by focusing on several specific time periods of learning: when CRs initially increased early in training and when CR performance became asymptotic later in training. This information may be particularly important for understanding the dynamics of information flow in the hippocampal circuit during associative learning.

	METHODS

Abstract Introduction Methods Results Discussion References

Subjects and surgery

Subjects were 12 young adult (3- to 4-mo old) New Zealand albino female rabbits. Surgery was carried out under National Institutes of Health and Northwestern University Animal Care and Use Committee approved procedures. Animals were anesthetized with ketamine (60 mg/kg im) and xylazine (10 mg/kg im), and the eyes were kept moist with a thin coat of antibacterial ophthalmic ointment. The skull was positioned in a stereotaxic frame with lambda 1.5 mm below bregma. The skull then was exposed and a 3-mm diam hole drilled above the left CA1 area of the hippocampus (~5.5 mm posterior and 5.5 mm lateral to bregma). Five self-tapping screws (2 × 1/4 in) were inserted ~2 mm into the skull to anchor the final dental cement-head assembly. Electrodes then were lowered stereotaxically into the left CA1 area of the hippocampus (~3 mm ventral to dura) until action potentials with pyramidal cell firing characteristics were recorded (Ranck 1973). Dental cement then was used to secure the electrodes to the skull and close the remaining wound area. A light-weight head restraining device containing four nylon bolts (6-32 × 3/4 in) for use during behavioral training also was cemented to the top of the skull anterior to the electrode assembly. Rabbits were given Buprenex (0.3 mg/kg sc) to minimize discomfort after recovery from anesthesia.

Behavioral training

Rabbits were allowed 5-7 days of recovery before any handling or testing. Animals then received two consecutive daily sessions of acclimation to the testing chamber followed by consecutive daily sessions of either trace eyeblink conditioning or pseudoconditioning. For acclimation and training sessions, animals were placed in a cloth restraining jacket and Plexiglas restraining box and then placed in a sound-attenuating chamber. The right eye was held open in a comfortable position with eyelid hooks attached to a velcro strap, and the head of the animal was secured in a comfortable position using the headbolt. The ends of rubber tubes (1 cm diam) were placed comfortably in each ear and served to deliver the auditory tone-CS (100 ms; 90 dB; 6 kHz; 5 ms rise/fall time) from headphones. An airpuff tube was placed 1 cm from the animal's right eyeball and served to deliver the airpuff US (150 ms; 3 psi). The airpuff was supplied by compressed air and controlled by a regulator and solenoid valve. The airpuff intensity was adequate to reliably evoke an eyeblink, which consisted of an extension of the nictitating membrane across the globe of the eyeball. An infrared sensor attached to the airpuff tube transduced extensions of the nictitating membrane (Thompson et al. 1994). Changes in voltage from the infrared sensor were collected by a computer, which also controlled the delivery of CSs and USs (Akase et al. 1994).

Each acclimation session lasted 30 min during which no stimuli were presented. After acclimation, trace-conditioned rabbits (n = 7) received daily sessions consisting of 80 trials presented at an intertrial interval of 30-60 s. Each trial consisted of the CS, followed by a 500-ms stimulus-free trace period, then the US. Pseudoconditioned control rabbits (n = 5) received daily sessions consisting of 80 CS alone trials and 80 US alone trials presented at an intertrial interval of 15-30 s. Trace- and pseudoconditioning sessions lasted ~1 h and consisted of the same number of CSs and USs. For trace- and pseudoconditioned animals, CRs were defined as eyeblinks that occurred in the 600-ms period after CS onset and produced changes in voltage >10 ms in duration and >4 SD of the mean baseline voltage.

Single neuron recording

Single neurons were recorded using single Teflon coated 50-µm-diam stainless steel recording wires arranged in a 2 × 4 array (1-mm spacing; NB Labs, Denison, TX). Each wire allowed approximately two to four neurons to be recorded and separated with software. Two of the animals that received trace conditioning were implanted with a single moveable tetrode (Gray et al. 1995), which allowed approximately five to seven single units to be recorded and separated. Tetrodes were fashioned by twisting and heating four teflon-coated nicrome microwires (14 µm OD each) according to the techniques described by Gray et al. (1995). The tetrodes were moved ventrally (100-200 µm) 10-20 min before several of the training sessions to obtain a greater number of single units.

Single neuron analogue signals were amplified (×10,000), filtered (10 kHz and 300 Hz), and collected with a 486 33-MHz IBM-compatible computer, which sampled each channel at 30-32 kHz. Single neuron data were collected 1 s before and 2 s after CS onset using software from DataWave Technologies (Longmont, CO). The software allowed voltage thresholds to be set for each recording channel during training to minimize recorded noise and to ensure that the sampling capacity of the system was not exceeded. A 1.5-ms epoch of data was recorded to disk when the threshold of a channel was exceeded. Single neurons were separated using measurements of action potential (spike) height and width. Variations of these measurements were used to form two-dimension "cluster" plots, which allowed neurons to be tracked across a single recording session and across recording sessions in rare instances. Single neurons with similar waveforms were further separated by waveform matching the unique segments of the individual single neuron action potential. The oscilloscope was monitored carefully during training, and the waveforms were analyzed after training to ensure that the same neurons were tracked across the entire training session. It is important to note that the cluster plots of spike waveform measurements used to separate single neurons changed from day to day in most cases. A conservative approach was used, and neurons were treated as the same neuron on multiple days only if a similar base rate, spike width, and cluster pattern was resolved on successive days. In almost all cases, unique spike and cluster characteristics were recorded on each of the days of training, and only seven neurons were tracked across multiple days of training. This does not rule out the possibility that small drifts in the electrode between recording sessions may have allowed new patterns of clusters to form that included one or two of the neurons from the previous recording day. However, the chances of this happening were minimized by carefully tracking the base rates and spike widths for each new cluster pattern formed. Technical problems prevented three of the recorded cells from being included in this study.

After spike separation, an average waveform was computed for each single neuron to aid in the classification of cell type. Action potential widths were calculated from each average waveform as the peak time minus the valley time. Putative pyramidal cells were separated from putative theta cells using measurements of action potential width and background firing rate. Using criteria similar to those described by Fox and Ranck (1981), cells with a spike duration 0.3 ms and background firing rate <8 Hz were classified as putative pyramidal cells, and cells with a spike duration <0.3 ms and a background firing rate 8 Hz were classified as putative theta cells. A small number of cells that did not meet the criteria of both firing rate and action potential width were classified only according to action potential width.

Grouping of neural and behavioral data

To examine changes in CA1 single neuron activity during multiple stages of conditioning, behavioral and neural data from each trace conditioned animal were analyzed in relation to the day of training when the initial increase in CRs occurred and the day of training when CRs became asymptotic. These days of training were determined for each individual trace conditioned animal using a variation of previous strategies (Disterhoft et al. 1977; Theios and Brelsford 1966) designed to reduce the error variance associated with the heterogeneity in the rate and pattern of eyeblink CRs between animals (Thompson et al. 1996a). The change in daily percent CRs was calculated for each animal [(percentage of CRs on day X)  (percentage of CRs on day X  1)] across the days of training to determine the SD of change in daily percent CRs. These SDs ranged from 9.58 to 40.7. The day of the initial increase in CRs was defined for each animal as the first day of training when the change in daily percent CRs was >1 SD of the mean. The daily neural and behavioral data for each trace conditioned animal were grouped into the following days of early conditioning: the day before an animal showed an initial increase in CRs (day before); the day of training when an animal showed the initial increase in CRs (day of initial CR increase); and the 2 days after an animal's CRs increased (day after CR increase and 2nd day after CR increase). For several of the animals neural data were obtained from only the day before and three subsequent days of training, therefore, group analyses were limited to these days.

The neural and behavioral data for each trace conditioned animal were also analyzed in relation to the days of training when CRs were asymptotic. Thus the daily neural and behavioral data for each trace conditioned animal were grouped into the day before an animal's highest percentage of CRs was exhibited (day before asymptotic CRs), the day an animal's highest percentage of CRs was exhibited (day of asymptotic CRs), and the subsequent day of training (day after asymptotic CRs). No neural activity was recorded from one of the trace conditioned animals on the day of asymptotic CRs. Separate analyses were performed on the data collected from the days of initial CRs and the days of asymptotic conditioning to isolate changes in hippocampal activity in relation to unique phases of conditioning. For several animals, the days of training immediately after the day of initial CR increase were also days of asymptotic CRs, however, only 33 trace conditioned cells overlapped between the days of initial and asymptotic CRs.

Each pseudoconditioned control animal (n = 5) was matched randomly to one of the trace conditioned animals. The ordinal days of training that corresponded to the day of initial CR increase and the day of asymptotic CRs from the trace conditioned animal (e.g., 2 and 5) were the same days of training analyzed from the matched pseudoconditioned animal (e.g., 2 and 5).

Analyses

Changes in single neuron activity were measured using standard t-test scores. For each neuron, standard test-scores were computed for each of three consecutive 200-ms periods following both the CS and the US to capture discrete short latency increases and decreases in activity. The standard test scores were computed by subtracting the number of action potentials in the 200-ms period preceding CS onset from the number of action potentials in the 200-ms periods after CS and US onset. The differences calculated for each period were then averaged across blocks of 5 or 20 trials and divided by the sample standard deviation of the difference. The test-score measures then could be averaged across neurons to compare the blocks of trials or days of training. The test scores also were used to determine if the increases or decreases in activity for each individual neuron were significant. To test significance, each 200-ms period was examined separately for each neuron, and significance (P < 0.05, two tailed) was required for at least one of the 20-trial blocks of a session.

Analyses of variance (ANOVAs) were performed using a general linear model with an unbalanced design. Significant interactions were subjected to follow-up one-way ANOVAs using the MSerror term from the main general linear model analyses. Significant main effects were followed up with Newman-Keuls post hoc tests. An alpha of 0.05 was required for all significant analyses. The relationship of behavior and single cell activity was analyzed with Pearson correlations. The daily sum of activity for each single neuron in 10-ms bins was correlated with the daily average voltage for the eyeblink response in 10-ms intervals.

Histology

Marking lesions were placed at the tips of all electrodes by passing DC current (25 µA) for 20 s. Animals were overdosed with sodium pentobarbital and perfused transcardially with saline (0.9% NaCl) followed by 10% formaldehyde. Brains implanted with stainless steel electrode-arrays were reacted with 10% potassium ferrocyanide to produce a blue electrode mark. Brains were then frozen, sectioned coronally (50-µm thick), mounted on albumin/gelatin coated slides, and stained with neutral red. A light microscope then was used to locate electrode tips.

	RESULTS

Abstract Introduction Methods Results Discussion References

Histology

Electrode tips were placed in the oriens, pyramidal, and stratum radiatum layer of the CA1 area of the dorsal hippocampus, ~3.0-3.5 mm caudal to bregma. A total of seven trace-conditioned rabbits and five pseudoconditioned rabbits had one or more electrodes placed in CA1.

Behavioral CRs

DAYS OF INITIAL CRS. For the trace conditioned animals, the day of initial CR increase was the second day of training for three animals, the third day for three animals, and the sixth day for one animal. The percentage of CRs for trace conditioned animals showed a sharp increase across blocks on the day of initial CR increase and near asymptotic levels on the following day of training. This increase in CRs can be seen in Fig. 1A. An ANOVA was conducted on these data with the factors group (trace- and pseudoconditioning), days (day before to 2nd day after CR increase), and blocks (four 20-trial blocks within each day). This analysis revealed a group × days effect [F(3,140) = 10.38, P = 0.0001, MSerror = 0.0523]. The percentage of CRs for trace-conditioned animals on the day of initial CR increase and the day after CR increase were both greater than the preceding day of training [F(3,140) = 32.6, P < 0.001], and group differences occurred on the day of initial CR increase and the two subsequent days of training [F(1,140) = 23.6, P < 0.001, F(1,140) = 63.2, P < 0.001, F(1,140) = 32.5, P < 0.001, respectively]. Figure 2 shows daily averages of CR amplitude for the same groups depicted in Fig. 1.

View larger version (44K): [in this window] [in a new window]   FIG. 1. Trace conditioned pyramidal cell activity after the tone-conditioned stimulus (CS) and airpuff unconditioned stimulus (US) showed large increases on the day when animals initially exhibited increases in conditioned responses (CRs). A: mean percentage of CRs across 4 20-trial blocks of each day of early conditioning (day before, day of initial CR increase, day after CR increase, 2nd day after CR increase). B: mean percentage of CRs across 20-trial blocks of asymptotic conditioning (day before asymptotic CRs, day of asymptotic CRs, day after asymptotic CRs). C and D: mean standard test-score calculated from the activity in the 200-ms period after CS onset for pyramidal cells recorded during the days of early and asymptotic conditioning, respectively. Increases in pyramidal activity from trace-conditioned animals were reduced significantly after the day of initial CR increase. E and F: mean standard test score calculated from the activity in the period 201-400 ms after US onset for all pyramidal cells recorded during the days of early and asymptotic conditioning, respectively. Pyramidal cells showed decreases in activity during asymptotic CRs. Bars indicate SE.

View larger version (41K): [in this window] [in a new window]   FIG. 2. Voltage measure (above) for eyeblink response (closure upward) averaged across animals for a single day of training, and perievent histograms (10-ms bins) averaged across pyramidal cells recorded for a single day of training. Action potentials (spikes) from each cell were summed across a single day of training then averaged across cells. A: activity and behavior on the day before initial CR increase for trace conditioning trials (top), CS trials during pseudoconditioning (middle), and US trials during pseudoconditioning (bottom). B: activity and behavior on the day of initial CR increase. Arrows 1 and 2 highlight increased trace conditioned activity after the CS and US, respectively. C: activity and behavior on the day of asymptotic CRs. Increased activity due to trace conditioning has diminished during asymptotic conditioning (Arrow 3). Bar = 1.5 V eyeblink.

DAYS OF ASYMPTOTIC CRS. The day of asymptotic CRs was the fifth day of training for three animals, the ninth day for two animals, and the third day for one animal. The percentage of CRs for trace-conditioned animals remained greater than controls during asymptotic conditioning, as shown in Fig. 1B. An ANOVA performed on the data in Fig. 1B with the factors group, days, and blocks revealed a group effect [F(1,102)=138.98, P = 0.0001, MSerror = 0.06].

Classification of single neurons

A total of 560 CA1 neurons were recorded and separated from the trace- and pseudoconditioned animals, however, 429 neurons were recorded on the days of early and asymptotic conditioning reported in this study. The cluster plots of spike waveform measurements used to separate single neurons changed each day in most cases, and only 7 of the 429 neurons were tracked across multiple days of training. The 429 neurons were separated into pyramidal cells and theta cells as described in the METHODS. Using these criteria, 21 neurons were classified as having background firing rates (14.3 ± 6.3 Hz) and spike widths (0.237 ± 0.047 ms) resembling theta cells. The remaining 408 nontheta cells were classified as having background firing rates (1.73 ± 2.38 Hz) and spike widths (0.372 ± 0.09 ms) within the range of pyramidal cells.

Thompson and Best (1989) have described "silent cells" in the hippocampus with very low background firing rates. A total of 99 out of the 408 pyramidal cells encountered had very low background firing rates (0.16 ± 0.14 Hz) and showed no changes in activity related to the CS or US. The remaining 309 nonsilent pyramidal cells (180 trace conditioned cells; 129 pseudoconditioned cells) were analyzed for early and/or asymptotic learning-related changes in unit activity. An average of 32 and 30 cells were analyzed for each day in each group during the days of early and asymptotic conditioning, respectively.

CA1 pyramidal cell activity during initial CRs

SINGLE NEURON ACTIVITY AFTER THE CS. One of the major aims of this study was to examine CA1 single neuron activity during the period of conditioning when trace CRs first show increases. Therefore the activity of the nonsilent pyramidalcells was examined relative to the day of initial CR increase. The perievent histograms in Fig. 2B show that a large increase in activity occurred during the first 200-ms period after the tone-CS onset (100-ms tone CS + 100 ms of the trace period) for trace-conditioned animals on the day of initial CR increase. Figure 1C shows that trace-conditionedactivity was learning related, increasing sharply on the day of initial CR increase and becoming smaller on the subsequent days of training. An ANOVA was conducted on the data in Fig. 1C and revealed a significant interaction of group × days [F(3,936) = 4.49, P = 0.004, MSerror = 1.346]. Group differences occurred on the day of initial CR increase and the following day of training [F(1,936) = 31.28, P < 0.001 and F(1,936) = 9.28, P < 0.01, respectively]. The activity of trace-conditioned animals on the day of initial CR increase was greater than all other days of early conditioning [F(3,936) = 6.24, P < 0.01].

Figure 1C also shows that the CS-evoked activity for the trace-conditioned group decreased across the blocks within the day of initial CR increase and the following days of training. The main ANOVA revealed a group × blocks interaction [F(1,936) = 2.36, P = 0.049], and group differences in the first, second, and third 20-trial block [F(1,936) = 21.93, P < 0.001 and F(1,936) = 9.75, P < 0.01, F(1,936) = 4.96, P < 0.01, respectively]. Tests also revealed that the activity from trace conditioned animals was greater in the first 20-trial block than the 4th 20-trial block [F(3,936) = 3.24, P < 0.05]. This may suggest that CA1 neurons play a more prominent role in learning at the beginning of the daily conditioning sessions.

Significant learning-related changes were not revealed in the analyses of the activity in the periods 201-400 ms and 401-600 ms after CS onset from the days of early conditioning.

SINGLE NEURON ACTIVITY AFTER THE US. Figure 2, A and B, shows that there was a large increase in pyramidal cell activity in the 400-ms period after the airpuff-US on the day of initial CR increase. An analysis of the single neuron activity in the period 200 ms after US onset revealed no significant effects. However, an analysis of the pyramidal cellactivity in the period 201-400 ms after US onset, shownin Fig. 1E, revealed a difference in activity between thetrace- and pseudoconditioned group on the day beforethat became larger on the day of initial CR increase. AnANOVA conducted on the data in Fig. 1E revealed a significant interaction of group × days [F(3,940) = 6.58, P = 0.0001, MSerror = 2.627]. The activity after the US on the day of initial CR increase was greater than the preceding and subsequent days of training for trace-conditioned animals [F(3,940) = 5.4, P < 0.05]. Group differences were present on the day before and the following day [F(1,940) = 9.12, P < 0.01, F(1,940) = 12.18, P < 0.001, respectively].

An ANOVA conducted on the single neuron activity in the period 401-600 ms after the onset of the US also revealed a significant group × days interaction, [F(3,940) = 7.01, P = 0.0001, MSerror = 2.411]. These data showed a pattern similar to the data in Fig. 1E.

SINGLE-UNIT CHANGES PRECEDE CRS. To better understand the time course of behavioral and single neuron changes, the data from the day before and the day of initial CR increase in Fig. 1, A, C, and E, were analyzed using five-trial blocks. These data are depicted in Fig. 3 and show that the learning-related changes in single neuron activity after the CS and US preceded the increase in CRs. The percentage of trace-conditioned CRs was at the level of pseudoconditioning during the first 5 trials on the day of initial CR increase, and group differences did not emerge until the 10th or 15th trial. However, group differences in CS-evoked activity occurred on the first five-trial block on the day of initial CR increase, and group differences in US-evoked activity occurred within the first five trials on the day before. ANOVAs conducted on the data in Fig. 3 revealed significant interactions of group × days for the changes in CRs and the changes in single neuron activity after the CS, [F(1,302) = 42.96,P = 0.0001, MSerror = 0.051 and F(1,832) = 14.18, P = 0.0001, MSerror = 1.161, respectively], and a significant group effect for the changes in activity after the US, [F(1,1851) = 25.82, P = 0.0001].

View larger version (28K): [in this window] [in a new window]   FIG. 3. Behavioral CRs and pyramidal-cell activity after the CS and US grouped into 5-trial blocks on the day before and the day of initial CR increase suggests that learning-related increases in neuronal activity preceded CRs. A-C: data from Fig. 1, A, C, and E, respectively, grouped into 16 5-trial blocks for each day.

CA1 pyramidal cell activity during asymptotic CRs

SINGLE NEURON ACTIVITY AFTER THE CS. The other major aim of this study was to examine the activity of CA1 pyramidal cells when conditioning was well established or asymptotic. The CS-evoked increase in activity persisted during asymptotic conditioning as seen in Figs. 1D and 2C. An ANOVA conducted on the standard test scores in Fig. 1D revealed a significant group effect [F(1,653) = 14.21, P = 0.0001, MSerror = 1.107]. An ANOVA was also conducted using the test scores calculated from the periods 201-400 ms and 401-600 ms after tone onset during asymptotic conditioning. The analysis of the 201- to 400-ms period revealed a group effect, F(1,653) = 6.90, P = 0.009 (MSerror = 1.291) and a pattern of activity similar to the data in Fig. 1D.

SINGLE NEURON ACTIVITY AFTER THE US. Unlike the changes in activity after the US early in conditioning, the pyramidal cells from trace-conditioned animals showed decreases in activity, or "inhibitory" responses, to the US during asymptotic conditioning. These decreases in activity occurred during the period 201-400 ms after US onset and can be seen in the average histograms in Fig. 2C. The mean standard test scores for this period are plotted in Fig. 1F, and the decreases in activity are indicated by the negative means. An ANOVA conducted on the data in Fig. 1F revealed significant effects of group [F(1,653) = 5.49, P = 0.019], and days [F(2,653) = 4.04, P = 0.018, MSerror = 2.458]. ANOVAs also were conducted on the standard test scores calculated from the activity in the periods 1-200 and 401-600 ms after US onset during the days of asymptotic conditioning, and neither of these analyses revealed significant effects.

COMPARISON OF THE DAYS OF INITIAL AND ASYMPTOTIC CRs. From Figs. 1 and 2 it is apparent that the trace-conditioned pyramidal cell activity after the CS and US was much greater on the day of initial CR increase than during asymptotic conditioning. One-way ANOVAs and post hoc tests applied to the trace-conditioned single neuron data in Fig. 1 confirmed that the activity after the CS and US on the dayof initial CR increase was greater than the activity on the day of asymptotic CRs, and day after asymptotic CRs,[F(2,301) = 5.88, P = 0.003, MSerror = 1.54 and F(2,301) = 16.08, P = 0.0001, MSerror = 2.76, respectively].

CA1 theta cell activity

The majority of the cells in this study were classified as pyramidal cells; however, there were 21 cells recorded with characteristics resembling theta cells. Unlike pyramidal cells, trace conditioned theta cell activity decreased after CS and US onset on the day of initial CR increase and on the two following days of training as seen in Fig. 4, A and B. This pattern of theta cell activity was opposite to the pattern exhibited by the pyramidal cells in Figs. 1 and 2. Figure 7F shows a decrease in activity for a single theta cell. AnANOVA was performed on the theta cell activity from the 200-ms period after CS onset shown in Fig. 4A and revealed a significant group × days interaction [F(3,118) = 3.13, P = 0.028, MSerror = 2.187]. Group differences occurred on the 2nd day after CR increase [F(1,118) = 12.4, P < 0.001], and trace-conditioned activity on the day before was greater than the 2nd day after CR increase [F(3,118) = 4.05, P < 0.01]. ANOVAs conducted on the theta cell activity recorded in the periods 201-600 ms after CS onset revealed no effects. The ANOVA performed on the theta cell activity occurring during the 200-ms period after US onset shown in Fig. 4B revealed a significant group × days interaction [F(3,118) = 5.31, P = 0.002, MSerror = 6.298]. Group differences occurred on the day of initial CR increase and the two following days of training [F(1,118) = 9.74, P < 0.01; F(1,118) = 7.81, P < 0.01; F(1,118) = 8.10, P < 0.01, respectively]. The ANOVA conducted on the theta cell activity in the period 401-600 ms after US onset (data not shown) also revealed a group effect [F(1,118) = 6.39, P = 0.013, MSerror = 6.283]. It is important to note that these theta cell analyses consist of a relatively small number of cells (average of 5 theta cells for each day in each group). Furthermore, not enough theta cells were obtained from the days of asymptotic conditioning to assess the late pattern of theta cell activity.

View larger version (33K): [in this window] [in a new window]   FIG. 4. Theta cells exhibited decreases in activity early in conditioning. A: mean standard test score calculated from activity in the period 200 ms after CS onset for all theta cells recorded during the days of early conditioning. B: mean standard test score calculated from activity in the period 200 ms after US onset for all theta cells recorded during the days of early conditioning. Pattern of theta cell activity was opposite to pyramidal cell activity.

View larger version (38K): [in this window] [in a new window]   FIG. 7. Perievent histograms of individual exemplar pyramidal cells (A-E) from each of the excitatory and inhibitory response profiles. Histograms show number of spikes from 1 cell in 10-ms bins across a single day of training. A: tone excitatory cell. B: trace excitatory cell. C: airpuff excitatory cell. D: trace inhibitory cell. E: airpuff inhibitory cell. F: an individual theta cell that showed an exemplar inhibitory response to the CS and US on the day of initial CR increase. Voltage measures above are average eyeblink responses (closure upward) of the animals from which these cells were recorded (Bar = 2 V eyeblink).

Classification of individual pyramidal cell responses

To understand the individual neuronal responses that contributed to the overall pattern of learning-related changes in CS-evoked neural activity, each of the 200-ms periods from each of the trace-conditioned nonsilent pyramidal cells was analyzed for significant increases and decreases in activity. A total of 109/180 (60.6%) trace-conditioned cells showed significant increases and/or decreases in activity for at least one of the 200-ms periods during at least one of the 20-trial blocks of training. Cells from trace-conditioned animals with significant changes in activity were divided into categories with the most consistent response patterns: cells with increases during the first 200 ms after CS onset (tone excitatory); cells with increases during periods 201-600 ms after CS-onset but no change during the first 200 ms after CS onset (trace excitatory); cells with increases during the 600 ms after US onset but no change during the 600 ms after CS onset (airpuff excitatory); cells with decreases during the 600 ms after CS onset but not during the 600 ms period after US onset (trace inhibitory); and cells inhibitory during the 600 ms after US onset (airpuff inhibitory). Note that airpuff inhibitory cells were often inhibitory during the 600 ms after CS onset as well.

Table 1 shows the proportions of these single cell response profiles recorded during the 4 days of initial CRs and the 3 days of asymptotic CRs. This table shows that the percentage of airpuff inhibitory cells more than triples from the days of initial to asymptotic CRs. Average perievent histograms of the cell types are shown in Figs. 5 and 6 and histograms of exemplar single neurons are shown in Fig. 7. During the days of initial CRs, tone excitatory cells responded primarily to the onset of both the CS and US. Moreover, 17/23 of the tone excitatory cells responded significantly during the 200-ms periods after both the CS and US onset, whereas only 7/23 of these cells responded significantly during the trace periods 201-600 ms after the CS. In contrast, only 4/19 of the trace excitatory cells responded during the first 200 ms after the US.

View larger version (37K): [in this window] [in a new window]   FIG. 5. Average perievent histograms (10-ms bins) for excitatory pyramidal cell response profiles recorded from trace-conditioned animals during the 4 days of initial CRs (left) and 3 days of asymptotic CRs (right). For each histogram, action potentials (spikes) from each cell were summed across a single session, then averaged across cells. A and B: cells with increases during the 1st 200 ms after CS onset (tone excitatory). C and D: cells with increases during the periods 201-600 ms after CS-onset, but no change during the 1st 200 ms after CS onset (trace excitatory). E and F: cells with increases during the 600 ms after US onset, but no change during the 600 ms after CS onset (airpuff excitatory). Voltage measures above are the eyeblink responses (closure upward) averaged across the animals from which these cells were recorded (Bar = 2 V eyeblink).

View larger version (38K): [in this window] [in a new window]   FIG. 6. Average perievent histograms (10-ms bins) for inhibitory pyramidal cell response profiles recorded from trace-conditioned animals during the 4 days of initial CRs (left) and 3 days of asymptotic CRs (right). For each histogram, spikes from each cell were summed across a single session, then averaged across cells. A and B: cells with decreases during the 600 ms after CS onset but not during the 600-ms period after US onset (trace inhibitory). C and D: cells inhibitory during the 600 ms after US onset (airpuff inhibitory). Airpuff inhibitory cells were often inhibitory during the 600 ms after CS onset as well. Voltage measures above are eyeblink responses (closure upward) averaged across the animals from which these cells were recorded (Bar = 2 V eyeblink).

Berger and Thompson (1978b) examined CA1 activity during delay eyeblink conditioning and found a large number of unit responses that paralleled the amplitude-time course of the behavioral CR. Table 1 shows that none of the average single cell response profiles from trace-conditioned animals showed a strong correlation with the behavioral response. Pyramidal cells that exhibited r  0.50 were neurons with cellular activity that accounted for 25% of the variance in the eyeblink response. Only 5% of trace conditioned pyramidal cells (9 cells) exhibited a r  0.50.

To understand how each of the pyramidal response profiles contributed to the early increases in single neuron activity, the mean standard test scores from each of the 200-ms periods were analyzed for each of the cell types during the days of initial CRs. These data are depicted in Figs. 8 and 9, which show that the single neuron activity on the Day of Initial CR Increase was greater than the other days for each of the cell types except the airpuff inhibitory cells. ANOVAs revealed significant effects of days for the tone and trace excitatory cells and trace and airpuff inhibitory cells and an interaction ofdays × periods for the airpuff excitatory cells [F(3,252) = 10.84, P = 0.0001, MSerror = 3.434; F(3,187) = 7.57, P = 0.0001, MSerror = 1.674; F(3,142) = 7.216, P = 0.019, MSerror = 2.118; F(3,169) = 5.38, P = 0.001, MSerror = 1.019; F(15,356) = 1.91, P = 0.022, MSerror = 1.398; respectively]. Each of the ANOVAs also revealed significant effects for the 200-ms periods. The cell types from the days of asymptotic CRs also were analyzed but revealed no consistent results.

View larger version (69K): [in this window] [in a new window]   FIG. 8. Mean standard test scores calculated from activity in each of the 200-ms periods after CS and US onset for each of the excitatory pyramidal cell response types. CS was presented during the 1st 100 ms of the 1st period, and the US during 1st 150 ms of the 4th period. Excitatory responses were greatest at the beginning of each daily session, therefore, standard test scores from each of the cell types were averaged across the 1st 40 trials of each day of early conditioning. Each of the excitatory cell types showed greater increases in activity on the day of initial CR increase compared with other days of training.

View larger version (46K): [in this window] [in a new window]   FIG. 9. Mean standard test scores calculated from activity in each of the 200-ms periods after CS and US onset for each of the inhibitory pyramidal cell response types. CS was presented during the 1st 100 ms of the first period, and the US during the 1st 150 ms of the fourth period. Inhibitory responses were greatest during the 2nd half of each daily session, therefore, standard test-scores from each of the cell types were averaged across the last 40 trials of each day of early conditioning. Trace inhibitory cells showed greater increases in activity during the airpuff period on the day of initial CR increase compared with other days of training.

Additional analyses compared the activity on CR trials to non-CR trials for each of the 200-ms periods after the CS and US. No significant effects or consistent trends were revealed by the overall analyses or the analyses of the individual cell types. This null result may be partially due to the finding that the largest and most prominent changes in activity occurred very early on the day of initial CR increase when CR levels were still relatively low.

An ANOVA compared the background firing rates of each cell type recorded during both early and asymptotic conditioning. This analysis revealed no significant effects. However, the background rate of the airpuff excitatory cells in Fig. 5 does appear to drop slightly from the days of initial to asymptotic CRs.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

Sequence of changes in CA1 during trace conditioning

Pyramidal cells in CA1 demonstrated several stages of learning-related activity in the present study: large increases in activity after both the CS and US on the day of initial CR increase, smaller moderate increases in activity on the following days of training, and decreases in activity after the US during asymptotic CRs. This sequence of pyramidal cell activity is consistent with the notion that the large increase in activity on the day of initial CR increase represents a stage of activity that is critical for the initial events of learning. On this day the learning-related increase in activity after the CS preceded learning by at least 5-10 trials. Furthermore, the activity after the US was greater in the trace-conditioned group compared with the pseudoconditioned group on the day before the initial CR increase. It is possible that this early increase in trace-conditioned single neuron activity after the US was the result of an additive neural response of the combined CS- and US-evoked neural activity similar to reflex facilitation or modification (Gormezano et al. 1983; Thompson et al. 1976; Weisz and McInerney 1990). The additive neural response to the paired CS-US presentation may have been a precursor for the critical neural change exhibited on subsequent days of training. Berger and Thompson's (1978b) work also showed an early change in hippocampal multiple-unit activity after the US, which developed before the changes in CS-evoked activity and preceded the emergence of delay-eyeblink CRs.

Later during asymptotic CRs, pyramidal cells from trace-conditioned animals exhibited a decrease in activity after the US. In addition, the percentage of airpuff inhibitory cells more than tripled from the days of initial to asymptotic CRs. Several other studies have reported decreases in hippocampal neuronal activity late in training (Freeman et al. 1996; Rolls et al. 1989; Sears and Steinmetz 1990; Weiss et al. 1996). Sears and Steinmetz (1990) hypothesized that reductions in hippocampal activity during asymptotic conditioning may represent a "filtering" of irrelevant or well-learned stimuli after the memory consolidation process has occurred. A reduction in stimulus-evoked activity also occurred across trials during the days of initial CRs, falling back to a zero-level of change at the end of each day of training. This suggests that the largest learning-related neural changes in CA1 were transient, occurring primarily in the earliest days of conditioning and in the early trials of the conditioning sessions. The sharp drop off in activity across days and across the trials within each day may reflect periods when the hippocampus was playing a limited role and other brain regions were playing a larger role in the learning process.

A number of hippocampal lesion studies support the notion that the decreases in activity obtained during asymptotic conditioning may represent periods when the hippocampus was playing a limited role in conditioning (Akase et al. 1989; Kim and Faneslow 1992; Kim et al. 1995; Winocur 1990; Zola-Morgan and Squire 1990). In one study by Kim and colleagues (1995), hippocampal lesions 1 day, but not 30 days, after rabbits reached a criterion of 8 CRs out of 10 trials disrupted trace eyeblink CR performance. Their behavioral criterion was similar to the mean percentage of CRs displayed on the day of asymptotic CRs in the present study. As others have proposed (e.g., McClelland et al. 1995; Zola-Morgan and Squire 1990), our data suggest that the hippocampus is similar to a temporary buffer where critical learning-related changes occur during the earlier periods of conditioning, whereas later long-term storage for associative conditioning occurs in other neural regions independent of the hippocampus. In vitro studies from our laboratory (Moyer et al. 1996; Thompson et al. 1996b) also support the notion of a temporary buffer by showing that pyramidal neurons exhibit increased excitability when examined in brain slices 1-72 h after rabbits reached a criterion of 80% trace conditioned CRs, but not 7 days after animals reached behavioral criterion.

The number of theta cells was small compared with the number of pyramidal cells, as anticipated (Fox and Ranck 1981). The theta cells recorded in this study appeared to show a pattern of activity complementary to that of pyramidal cells during the initial stages of conditioning. Although the theta and pyramidal cell activity patterns were not exact opposites, our data are consistent with the notion that theta cells have inhibitory inputs to pyramidal neurons (Andersen 1975; Foster et al. 1987; Fox and Ranck 1981; Pang and Rose 1989; Ranck 1973; Stewart 1993; Thompson et al. 1990).

Technical limitations permitted only a few cells to be recorded on multiple days of training in the present study. Nevertheless, a large and comparable number of single neurons were recorded and isolated from each day of training for both trace- and pseudoconditioned animals, allowing comparisons of neural activity between different days of training. The level of neural activity from control animals remained stable across the days of training, suggesting that the changes in single neuron activity from trace-conditioned animals across days were associative and not due to physiological changes related to surgery, the animals' condition, daily neuronal sampling changes, or sensitization from the tone-CS and airpuff-US.

Individual pyramidal cell response profiles

Single pyramidal cell response profiles were isolated that showed increases in activity specific to the day of initial CR increase. Analyses revealed that the overall and individual response profiles do not consist solely of trial-by-trial activity that predicts the occurrence of a CR. Moreover, the activity of these response profiles was not consistent with that of a premotor structure that drives the CR but rather the activity of these profiles appeared to reflect the coding of the CS and US and their temporal relations. Specifically, the tone excitatory cells responded primarily to the onset of the CS and US, whereas the trace excitatory cells exhibited an increased level of activity mainly during the trace interval between the two stimuli. These cells types show similarities to hippocampal neurons with onset and delay properties encountered in other studies (Eifuku et al. 1995; Hampson et al. 1993; Watanabe and Niki 1985). It is possible that together these cells may mark the onset of the CS and US and the time interval between these stimuli. Solomon (1980) previously has suggested that hippocampal neurons may somehow be involved in assessing the temporal contiguity of the CS and US and the appropriate timing of the behavioral response. This suggestion is supported by the finding that hippocampal lesions produce inappropriately timed short-latency "nonadaptive" eyeblink responses during hippocampally dependent trace eyeblink conditioning (Moyer et al. 1990; Solomon et al. 1986). These nonadaptive responses are similar to the short-latency eyeblink responses observed in animals with anterior cerebellar cortex lesions (Perrett et al. 1993). On the basis of these observations, Mauk proposed that the cerebellar cortex is involved in the timing of CRs (Raymond et al. 1996). Together these lesion and recording data suggest that a hippocampal-cerebellar circuit may be operating during trace eyeblink conditioning to assess the CS-US temporal relationship and determine the appropriate timing of behavioral responses (Berger and Bassett 1992; Weiss and Disterhoft 1996).

It appears from the present study and other work from our laboratory (Weiss et al. 1996) that the activity of hippocampal neurons during trace eyeblink conditioning does not consist of a unitary modeling response but rather is made up of several different single neuron response profiles. Heterogeneous hippocampal neuron response profiles also have been reported in other learning paradigms (e.g., Eichenbaum et al. 1989; Hampson et al. 1993). Recently, a number of studies have used ensemble analyses to explain the relationship of these heterogeneous hippocampal unit response profiles (Deadwyler and Hampson 1995; Deadwyler et al. 1996; Eichenbaum et al. 1989; Sakurai 1996; Wilson and McNaughton 1993, 1994). All of these studies suggest that the ensemble activity of groups of hippocampal cells encodes more information than single or average cell activity. Future studies employing ensemble analyses of hippocampal unit activity may provide a more detailed understanding of the heterogeneous response patterns that we and others have observed in CA1 single pyramidal neurons.

Functional properties of hippocampal neurons during learning

Berger and Thompson's (1978b) early electrophysiological work was the first to describe associative changes in hippocampal neuronal activity during eyeblink conditioning. Their group showed that learning-related changes in CA1 activity occur in a simple delay eyeblink conditioning task (Berger and Thompson 1978a; Berger et al. 1983). It is not entirely clear why hippocampal changes occur during delay eyeblink conditioning, as several studies have shown that this task can be acquired normally without an intact hippocampus (Akase et al. 1989; Schmaltz and Theios 1972; Solomon and Moore 1975). It has been suggested that changes in activity occur in the nonhippocampally dependent delay paradigm because the hippocampus is included in one of several parallel circuits that participate in a variety of associative-learning paradigms (Disterhoft et al. 1986; Thompson et al. 1980) and that the hippocampus may be essential only during certain learning situations that require association of more complex, configural representations of stimuli (Eichenbaum and Cohen 1988; Solomon 1980).

It is well known that hippocampal neurons are involved in situations that require animals to process spatial information (Fox and Ranck 1975; O'Keefe and Dostrovsky 1971; O'Keefe and Nadel 1978). The majority of the recording studies that have examined the learning-related correlates of hippocampal activity have used learning tasks that require spatial processing of information (e.g., Deadwyler et al. 1996; Eichenbaum et al. 1989; Eifuku et al. 1995; Gothard et al. 1996; O'Keefe and Speakman 1987; Quirk et al. 1990; Rolls et al. 1993). These and other recording studies have shown that hippocampal neurons encode information about an animal's place (Fox and Ranck 1975; O'Keefe and Nadel 1978; Ranck 1973), direction (McNaughton et al. 1983; Muller et al. 1994), and spatial orientation in the environment (O'Keefe and Conway 1978). However, in the hippocampally dependent trace eyeblink conditioning paradigm of the present study, animals were required to associate nonspatial stimuli in a temporal or configural representation (i.e., the CS-trace-US association). A number of other studies have shown that learning-related changes in hippocampal neuronal activity occur during hippocampally dependent nonspatial tasks (Eichenbaum et al. 1987; Hampson et al. 1993; Sakurai 1996; Wible et al. 1986; Young et al. 1994). The evidence from spatial and nonspatial learning studies has led several investigators to suggest that the role of hippocampal neurons is to encode relationships or associations between multiple stimuli or contingencies, including the configural properties of stimuli simultaneously present in the environment (Eichenbaum and Cohen 1988; McNaughton 1991). This definition of hippocampal function is consistent with the anatomic organization and cellular mechanisms that have been shown to support associative neuronal changes in the hippocampus (McNaughton 1991) and is consistent with the changes in hippocampal activity reported in this trace eyeblink conditioning study.

	ACKNOWLEDGEMENTS

  The authors thank Dr. Maria Carillo, M. Kronforst-Collins, B. Mossadeghi, M. Oh, J. Power, Dr. L. T. Thompson, and Dr. Craig Weiss at Northwestern University Medical School and Dr. Ed Green at the University of Miami for technical contributions and advice. The authors also thank P. Garcia at DataWave Technologies for technical assistance.

  This research was supported by National Institute of Health Grants RO1 MH-47340, RO1 AG-08796, and F32 AG-05711.

	FOOTNOTES

  Address for reprint requests: J. F. Disterhoft, Dept. of Cell and Molecular Biology, Northwestern University Medical School, 303 East Chicago Ave., Chicago, IL 60611. E-mail: jdisterhoft{at}nwu.edu

  Received 8 November 1996; accepted in final form 1 May 1997.

	REFERENCES

Abstract Introduction Methods Results Discussion References

• AKASE, E., ALKON, D. L., DISTERHOFT, J. F. Hippocampal lesions impair memory of short-delay conditioned eye blink in rabbits. Behav. Neurosci. 103: 935-943, 1989.[Medline]
• AKASE, E., THOMPSON, L. T., DISTERHOFT, J. F. A system for quantitative analysis of associative learning. II. Real-time software for MS-DOS microcomputers. J. Neurosci. Methods 54: 119-130, 1994.[Medline]
• ANDERSEN, P. Organization of hippocampal neurons and their interconnections. In: The Hippocampus, edited by R. L. Isaacson, and K. H. Pribram . New York: Plenum Press, 1975, vol. 1, p. 155-175
• BERGER, T. W., ALGER, B. E., THOMPSON, R. F. Neuronal substrates of classical conditioning in the hippocampus. Science 192: 483-485, 1976.[Abstract/Free Full Text]
• BERGER, T. W., BASSETT, J. L. System properties of the hippocampus. In: Learning and Memory: The Behavioral and Biological Substrates, edited by I. Gormezano, and E. A. Wasserman . Hillsdale, NJ: Lawrence Erlbaum Associates, 1992, p. 275-320
• BERGER, T. W., RINALDI, P. C., WEISZ, D. J., THOMPSON, R. F. Single-unit analysis of different hippocampal cell types during classical conditioning of rabbit nictitating membrane response. J. Neurophysiol. 50: 1197-1219, 1983.[Abstract/Free Full Text]
• BERGER, T. W., THOMPSON, R. F. Identification of pyramidal cells as the critical elements in hippocampal neuronal plasticity during learning. Proc. Natl. Acad. Sci. USA 75: 1572-1576, 1978a.[Abstract/Free Full Text]
• BERGER, T. W., THOMPSON, R. F. Neuronal plasticity in the limbic system during classical conditioning of the rabbit nictitating membrane response. I. The Hippocampus. Brain Res. 145: 323-346, 1978b.[Medline]
• DEADWYLER, S. A., BUNN, T., HAMPSON, R. E. Hippocampal ensemble activity during spatial delayed-nonmatch-to-sample performance in rats. J. Neurosci. 16: 354-372, 1996.[Abstract/Free Full Text]
• DEADWYLER, S. A., HAMPSON, R. E. Ensemble activity and behavior: what's the code? Science 270: 1316-1318, 1995.[Abstract/Free Full Text]
• DISTERHOFT, J. F., COULTER, D. A., ALKON, D. L. Conditioning-specific membrane changes of rabbit hippocampal neurons measured in vitro. Proc. Natl. Acad Sci. USA 83: 2733-2737, 1986.[Abstract/Free Full Text]
• DISTERHOFT, J. F., KWAN, H. H., LO, W. D. Nictitating membrane conditioning to tone in the immobilized albino rabbit. Brain Res. 137: 127-143, 1977.[Medline]
• DISTERHOFT, J. F., THOMPSON, L. T., MOYER, J. R. Cellular mechanisms of associative learning in the hippocampus. In: The Memory System of the Brain, edited by and J. Delacour . Singapore: World Scientific, 1994, p. 431-492
• EICHENBAUM, H., COHEN, N. J. Representation in the hippocampus: what do hippocampal neurons encode? Trends Neurosci. 11: 244-248, 1988.[Medline]
• EICHENBAUM, H., KUPERSTEIN, M., FAGAN, A., NAGODE, J. Cue-sampling and goal-approach correlates of hippocampal unit activity in rats performing an odor-discrimination task. J. Neurosci. 7: 716-732, 1987.[Abstract]
• EICHENBAUM, H., WIENER, S. I., SHAPIRO, M. L., COHEN, N. J. The organization of spatial coding in the hippocampus: a study of neural ensemble activity. J. Neurosci. 9: 2764-2775, 1989.[Abstract]
• EIFUKU, S., NICHIJO, H., KITA, T., ONO, T. Neuronal activity in the primate hippocampal formation during a conditional association task based on the subject's location. J. Neurosci. 15: 4952-4969, 1995.[Abstract]
• FOSTER, T. C., CHRISTIAN, E. P., HAMPSON, R. E., CAMPBELL, K. A., DEADWYLER, S. A. Sequential dependencies regulate sensory evoked responses of single units in the rat hippocampus. Brain Res. 408: 86-96, 1987.[Medline]
• FOX, S. E., RANCK, J. B. Localization and anatomical identification of theta and complex spike cells in dorsal hippocampus in rats. Exp. Neurol. 49: 299-313, 1975.[Medline]
• FOX, S. E., RANCK, J. B. Electrophysiological characteristics of hippocampal complex-spike cells and theta cells. Exp. Brain. Res. 41: 399-410, 1981.[Medline]
• FREEMAN, J. H., CUPPERNELL, C., FLANNERY, K., GABRIEL, M. Context-specific multi-site cingulate cortical, limbic thalamic, and hippocampal neuronal activity during concurrent discriminative approach and avoidance training in rabbits. J. Neurosci. 16: 1538-1549, 1996.[Abstract/Free Full Text]
• GOTHARD, K. M., SKAGGS, W. E., MOORE, K. M, MCNAUGHTON, B. L. Binding of hippocampal CA1 neural activity to multiple reference frames in a landmark-based navigation task. J. Neurosci. 15: 823-835, 1996.
• GORMEZANO, I., KEHOE, E. J., MARSHALL-GOODELL, B. Twenty years of classical conditioning research in the rabbit. Prog. Psychobiol. Physiol. Psychol. 10: 197-275, 1983.
• GORMEZANO, I., SCHNEIDERMAN, N., DEAUX, E., FUENTES, I. Nictitating membrane: classical conditioning and extension in the albino rabbit. Science 138: 33-34, 1962.[Abstract/Free Full Text]
• GRAY, C. M., MALDONADO, P. E., WILSON, M., MCNAUGHTON, B. Tetrodes markedly improve the reliability and yield of multiple single-unit isolation from multi-unit recordings in cat striate cortex. J. Neurosci. Methods 63: 43-54, 1995.[Medline]
• HAMPSON, R. E., HEYSER, C. J., DEADWYLER, S. A. Hippocampal cell firing correlates of delayed-match-to-sample performance in the rat. Behav. Neurosci. 107: 715-739, 1993.[Medline]
• KIM, J. J., CLARK, R. E., THOMPSON, R. F. Hippocampectomy impairs the memory of recently, but not remotely, acquired trace eyeblink conditioned responses. Behav. Neurosci. 109: 195-203, 1995.[Medline]
• KIM, J. J., FANESLOW, M. S. Modality-specific retrograde amnesia of fear. Science 256: 675-677, 1992.[Abstract/Free Full Text]
• MCCLELLAND, J. L., MCNAUGHTON, B. L., O'REILLY, R. C. Why there are complementary learning systems in the hippocampus and neocortex: Insights from the successes and failures of connectionist models of learning and memory. Psychol. Rev. 102: 419-457.
• MCNAUGHTON, B. L. Associative pattern completion in hippocampal circuits: new evidence and new questions. Brain Res. 16: 202-204, 1991.
• MCNAUGHTON, B. L., BARNES, C. A., O'KEEFE, J. The contributions of position, direction, and velocity to single unit activity in the hippocampus of freely-moving rats. Exp. Brain Res. 53: 41-49, 1983.
• MOYER, J. R., DEYO, R. A., DISTERHOFT, J. F. Hippocampectomy disrupts trace eye-blink conditioning in rabbits. Behav. Neurosci. 104: 243-252, 1990.[Medline]
• MOYER, J. R., THOMPSON, L. T., DISTERHOFT, J. F. Trace eyeblink conditioning increases CA1 excitability in a transient and learning-specific manner. J. Neurosci. 16: 5536-5546, 1996.[Abstract/Free Full Text]
• MULLER, R. U., BOSTOCK, E., TAUBE, J. S., KUBIE, J. L. On the directional firing properties of hippocampal place cells. J. Neurosci. 14: 7235-7251, 1994.[Abstract]
• O'KEEFE, J., CONWAY, D. H. Place units in the hippocampus of the freely moving rat. Exp. Neurol. 51: 78-109, 1978.
• O'KEEFE, J., DOSTROVSKY, J. The hippocampus as a cognitive map. Preliminary evidence from unit activity in freely moving rats. Brain Res. 34: 171-175, 1971.[Medline]
• O'KEEFE, J., NADEL, L. In: The Hippocampus as a Cognitive Map., . London: Oxford University Press, 1978
• O'KEEFE, J., SPEAKMAN, A. Single unit activity in the rat hippocampus during a spatial memory task. Exp. Brain Res. 68: 1-27, 1987.[Medline]
• PANG, K., ROSE, G. M. Differential effects of methionine⁵-enkephalin on hippocampal pyramidal cells and interneurons. Neuropharmacology 28: 1175-1181, 1989.[Medline]
• PERRETT, S. P., RUIZ, B. P., MAUK, M. D. Cerebellar cortex lesions disrupt learning-dependent timing of conditioned eyelid responses. J. Neurosci. 13: 1708-1718, 1993.[Abstract]
• QUIRK, G. S., MULLER, R. U., KUBIE, J. L. The firing of hippocampal place cells in the dark depends on the rat's recent experience. J. Neurosci. 10: 2008-2017, 1990.[Abstract]
• RANCK, J. B. Studies on single neurons in dorsal hippocampal formation and septum in unrestrained rats. I. Behavioral correlates and firing repertoires. Exp. Neurol. 41: 461-555, 1973.[Medline]
• RAYMOND, J. L., LISBERGER, S. G., MAUK, M. D. The cerebellum: a neuronal learning machine? Science 272: 1126-1131, 1996.[Abstract]
• REMPEL-CLOWER, N. L., ZOLA, S. M., SQUIRE, L. R., AMARAL, D. G. Three cases of enduring memory impairment after bilateral damage limited to the hippocampal formation. J. Neurosci. 16: 5233-5255, 1996.[Abstract/Free Full Text]
• ROLLS, E. T., CAHUSAC, P.M.B., FEIGENBAUM, J. D., MIYASHITA, Y. Responses of single neurons in the hippocampus of the macaque related to recognition memory. Exp. Brain. Res. 93: 299-306, 1993.[Medline]
• ROLLS, E. T., MIYASHITA, Y., CAHUSAC, P.M.B., KESNER, R. P., NIKI, H., FEINENBAUM, J. D., BACH, L. Hippocampal neurons in the monkey with activity related to the place in which a stimulus is shown. J. Neurosci. 9: 1835-1845, 1989.[Abstract]
• SAKURAI, Y. Hippocampal and neocortical cell assemblies encode memory processes for different types of stimuli in the rat. J. Neurosci. 16: 2809-2819, 1996.[Abstract/Free Full Text]
• SCHMALTZ, L. W. AND THEIOS, J. Acquisition and extinction of a classically conditioned response in hippocampectomized rabbits (Oryctolagus cuniculus). J. Comp. Physiol. Psychol. 89: 328-333.
• SEARS, L. L., STEINMETZ, J. E. Acquisition of classically conditioned-related activity in the hippocampus is affected by lesions of cerebellar interpositus nucleus. Behav. Neurosci. 104: 681-692, 1990.[Medline]
• SOLOMON, P. R. A time and a place for everything? Temporal processing views of hippocampal function with special reference to attention. Physiol. Psychol. 8: 254-261, 1980.
• SOLOMON, P. R., MOORE, J. W. Latent inhibition and stimulus generalization of the classically conditioned nictiating membrane response in rabbits (Oryctolagus cuniculus) following hippocampal ablation. J. Comp. Physiol. Psychol. 89: 1192-1203, 1975.[Medline]
• SOLOMON, P. R., VANDER SCHAAF, E. R., THOMPSON, R. F., WEISZ, D. J. Hippocampus and trace conditioning of the rabbit's classically conditioned nictitating membrane response. Behav. Neurosci. 100: 729-744, 1986.[Medline]
• SQUIRE, L. R. Memory and the hippocampus: a synthesis from findings with rats, monkeys, and humans. Psychol. Rev. 99: 195-231, 1992.[Medline]
• STEWART, M. Disinhibition of hippocampal pyramidal cells during the transition into theta rhythm. Exp. Brain Res. 93: 1-5, 1993.[Medline]
• THEIOS, J., BRELSFORD, J. W. A Markov model for classical conditioning: application to eye-blink conditioning in rabbits. Psychol. Rev. 73: 393-405, 1966.[Medline]
• THOMPSON, L. T., BEST, P. J. Place cells and silent cells in the hippocampus of freely-behaving rats. J. Neurosci. 9: 2382-2390, 1989.[Abstract]
• THOMPSON, L. T., DEYO, R. A., DISTERHOFT, J. F. Nimodipine enhances spontaneous activity of hippocampal pyramidal neurons in aging rabbits at a dose that facilitates associative learning. Brain Res. 535: 119-130, 1990.[Medline]
• THOMPSON, L. T., MOYER, J. R., AKASE, E., DISTERHOFT, J. F. A system for quantitative analysis of associative learning. I. Hardware interfaces with cross-species applications. J. Neurosci. Methods 54: 109-117, 1994.[Medline]
• THOMPSON, L. T., MOYER, J. R., DISTERHOFT, J. F. Trace eyeblink conditioning in rabbits demonstrates heterogeneity of learning ability both between and within age groups. Neurobiol. Aging 17: 619-629, 1996a.[Medline]
• THOMPSON, L. T., MOYER, J. R., DISTERHOFT, J. F. Transient changes in excitability of rabbit CA3 neurons with a time-course appropriate to support memory consolidation. J. Neurophysiol. 76: 1836-1849, 1996b.[Abstract/Free Full Text]
• THOMPSON, R. F., BERGER, T. W., BERRY, S. D., HOEHLER, F. K., KETTNER, R. E., WEISZ, D. J. Hippocampal substrate of classical conditioning. Physiol. Psychol. 8: 262-279, 1980.
• THOMPSON, R. F., BERGER, T., CEGAVSKE, C., PATTERSON, M. M., ROEMER, R., TEYLER, T., YOUNG, R. The search for the engram. Am. Psychol. 31: 209-227, 1976.[Medline]
• WATANABE, T., NIKI, H. Hippocampal unit activity and delayed response in the monkey. Brain Res. 325: 241-254, 1985.[Medline]
• WEISS, C., DISTERHOFT, J. F. Eyeblink conditioning, motor control, and analysis of limbic- cerebellar interactions. Behav. Brain Sci. 19: 479-481, 1996.
• WEISS, C., KRONFORST-COLLINS, M. A., DISTERHOFT, J. F. Activity of hippocampal pyramidal neurons during trace eyeblink conditioning. Hippocampus 6: 192-209, 1996.[Medline]
• WEISZ, D. J., MCINERNEY, J. An associative process maintains reflex facilitation of the unconditioned nictitating membrane response during the early stages of training. Behav. Neurosci. 104: 21-27, 1990.[Medline]
• WIBLE, C. G., FINDLING, R. L., SHAPIRO, M., LANG, E. J., CRANE, S., OLTON, D. S. Mnemonic correlates of unit activity in the hippocampus. Brain Res. 399: 97-110, 1986.[Medline]
• WILSON, M. A., MCNAUGHTON, B. L. Dynamics of the hippocampal ensemble code for space. Science 261: 1055-1058, 1993.[Abstract/Free Full Text]
• WILSON, M. A., MCNAUGHTON, B. L. Reactivation of hippocampal ensemble memories during sleep. Science 265: 676-679, 1994.[Abstract/Free Full Text]
• WINOCUR, G. Anterograde and retrograde amnesia in rats with dorsal hippocampal or dorsomedial thalamic lesions. Behav. Brain Res. 38: 145-154, 1990.[Medline]
• YOUNG, B. J., FOX, G. D., EICHENBAUM, H. Correlates of hippocampal complex-spike cell activity in rats performing a nonspatial radial maze task. J. Neurosci. 14: 6553-6563, 1994.[Abstract]
• ZOLA-MORGAN, S. M., SQUIRE, L. R. The primate hippocampal formation: evidence for a time-limited role in memory storage. Science 250: 288-290, 1990.[Abstract/Free Full Text]

0022-3077/97 $5.00 Copyright ©1997 The American Physiological Society