Behavioural Processes 101 (2014) 97–102
Contents lists available at ScienceDirect
Behavioural Processes journal homepage: www.elsevier.com/locate/behavproc
Brain electrophysiological activity correlates with temporal processing in rats Minoru Hattori a , Shogo Sakata b,∗ a b
Institute of Biomedical & Health Sciences, Hiroshima University, Hiroshima 734-8553, Japan Graduate School of Integrated Arts and Sciences, Hiroshima University, Higashi-Hiroshima 739-8521, Japan
a r t i c l e
i n f o
Article history: Received 15 April 2013 Received in revised form 12 September 2013 Accepted 17 September 2013 Keywords: Hippocampus Striatum Timing EEG power
a b s t r a c t In this study, we present electroencephalographic (EEG) recording data obtained in correlation with timing behavior in rats trained in a 30-s peak interval (PI) procedure. The distribution of lever press responses was found to be Gaussian, peaking at approximately 30 s: lever pressing behavior increased for 30 s then decreased after the reinforcement time. We recorded EEG activity in the hippocampus (hippocampal theta wave) and striatum during the task, and evaluated whether the EEG power correlated with the behavior pattern. We found that the striatum EEG, but not the hippocampal theta wave, showed a good correlation with the response pattern in the 30-s PI. This result suggests that striatum neurons fired more synchronously at the time of reinforcement, thus supporting a critical role for synchronization of firing of striatal neurons in regulating timing mechanisms. This article is part of a Special Issue entitled: Associative and Temporal Learning. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Timing is everything in living animals. It is very important to study timing in order to understand animal behavior. Interval timing is the ability of animals to perceive and process durations in the seconds-to-minutes range (Buhusi and Meck, 2005). How do animals estimate the passage of time? Numerous human and animal studies have demonstrated that the brain is equipped with multiple time-measurement systems associated with interval-timing behavior. The role of the hippocampus in temporal discrimination learning was initially explored by Meck et al. (1984). More recently, the hippocampus has become a primary target region in the study of the cognitive functions associated with temporal discrimination learning (Howard and Eichenbaum, 2013; Sakata, 2006; Yin and Troger, 2011). For example, in studies using the peak-interval (PI) procedure, rats with hippocampal lesions show a small leftward shift in their temporal discrimination function (Meck et al., 1984). Moreover, a frequency analysis of the hippocampal theta wave revealed an increase in power during a duration discrimination task compared with that observed in a simple reaction-time task (Sakata and Onoda, 2003). The hippocampal EEG theta wave is a large amplitude sinusoidal wave reflecting brain oscillation activity in the hippocampus (see Fujisawa and Buzsaki, 2011;
∗ Corresponding author. Tel.: +81 82 424 6581; fax: +81 82 424 0759. E-mail address: [email protected] (S. Sakata). 0376-6357/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.beproc.2013.09.011
Fig. 1 for typical raw wave data). These results suggest that the hippocampus contributes to timing and the perception of time. A variety of neuropsychological, functional neuroimaging and pharmacological studies suggest that the striatum, the major input structure of the basal ganglia complex, is an essential component of the neural networks involved in time perception. Neurons in the striatum are able to read time codes emitted by oscillator cells in the cortex (Matell et al., 2003). Impaired time perception has been found in patients with Parkinson’s disease (PD), attributed to the presence of a dysfunctional dopaminergic striatal pacemaker in such patients (Dusek et al., 2012; Malapani et al., 1998). The hippocampal formation sends projections to several structures, including the prefrontal cortex and ventral striatum. Both structures exhibit coherent theta rhythmicity. Previous studies have demonstrated that coherence between the striatum and hippocampus at the theta frequency is strong in the ventral/medial striatum and weak in the dorsal/lateral striatum (Berke et al., 2004). Pathways interconnecting the hippocampus and striatum are thought to use theta rhythms to transfer and coordinate neural representations in hippocampus–striatum circuits in relation to procedural maze tasks (DeCoteau et al., 2007). Should the hippocampus and striatum contribute to time perception, then the theta power of the hippocampus and striatum should increase during interval timing tasks. The purpose of this study was to examine the variation in the hippocampus and striatum theta power related to temporal discrimination tasks, in particular during the peak-interval (PI) procedure. Among temporal discrimination tasks, the PI procedure is a
98
M. Hattori, S. Sakata / Behavioural Processes 101 (2014) 97–102
Fig. 1. The experimental procedures of the PI training and PI test sessions. The horizontal lines represent time. The square peaks on the “tone” axis indicate the presentation of pure tones. The vertical lines on the “response” axis indicate the lever response. The oblique lines indicate the supply of food pellets. The PI task consisted of two types of trials: FI trials (50%) and peak trials (50%). The rats were trained using a 30-s FI 30 schedule (FI), during which the first lever press that occurred 30 s after the beginning of the signal triggered the delivery of a food reward. PI sessions included probe trials in which the lever was extended for 90 s and the lever presses were not reinforced. In the test session, the EEG recording window started 4 s before stimulus onset and ended 80 s after stimulus onset in the peak trial.
powerful tool for studying timing in rats (Meck et al., 2008). In the PI procedure, the animals are trained to respond in the presence of a stimulus after a given amount of time has elapsed. Therefore, we hypothesized that the pattern of response in a peak procedure should correlate with theta waves in the hippocampus and striatum. 2. Methods 2.1. Animals The subjects were six male Wistar rats (CLEA Japan, Inc., Osaka, Japan). Rats were 12 weeks old when the experiment began. Rats were housed in individual cages and kept on a 12:12-h light–dark cycle (lights on at 8:00 a.m.). Throughout the experiment, all rats were maintained at 85% of their ad libitum weight. Water was freely available throughout the experiment. The experiment was conducted in accordance with the guidelines published by the Japan Society for Animal Psychology and approved by the Committee of Experimental Animals at Hiroshima University (D08-6). 2.2. Apparatus Behavioral training and EEG recording sessions took place in a standard operant chamber (ENV-007 CT; MED Associates, Inc., Georgia, VT, USA). According to the procedure, 45-mg food pellets (F0165; Bio Serv, Frenchtown, NJ, USA) were delivered in a cup was placed at the center of the front wall at the floor level. A food dispenser delivered the pellets to the cup with a click noise. A lever was positioned on the left side of the front aluminum wall. The chambers were equipped with a house light and a back wall-mounted speaker to provide tone stimuli. The experimental control and lever pressing data recording was performed using a PC running an inhouse software program developed in Delphi (Borland Software Corporation). The chambers were housed in a soundproof, ventilated and electrically shielded room and monitored with a video camera (Mother tool CO. Model MTB-2081P). 2.3. Procedures 2.3.1. Initial training In the initial preparation training session, the rats received 3 days of training for lever pressing for up to 60 pellets or 30 min. Next, the rats progressed through the following training sequence: (a) a continuous reinforcement (CRF) schedule for 60 pellets; (b) a fixed ratio 3 (FR3) reinforcement schedule; (c) an FR6 reinforcement schedule; (d) an FR10 reinforcement schedule for 60 trials per session. One training and testing session were conducted per day. Afterwards, rats were trained in the timing procedures described below.
2.3.2. Fixed-interval (FI) training All rats received 15 sessions of a 30-s FI schedule in which a tone signal (2000 Hz, 80 dB) signaled the to-be-timed fixed interval. The sessions consisted of 60 trials, during which the first lever press that occurred 30-s after the beginning of the signal triggered the delivery of a food reward and terminated the tone signal for the duration of the variable intertrial interval (ITI) (average ITI 30 s, range 20–40 s). 2.3.3. Peak-interval (PI) task training After the FI training, the rats received 50 sessions of PI task training (Fig. 1). The PI session consisted of 60 trials, of which 50% were FI trials and 50% were PI trials. During the PI trials, the lever was extended for 90 s and the lever presses were not reinforced. The peak trials and FI trials were randomly intermixed with the restriction that no more than four peak trials occurred consecutively. As in the FI training, the trials were separated by variable ITIs (average 30-s, range 20–40 s). Afterwards, the rats underwent surgery for electrode implantation. 2.4. Surgery The electrode implantation procedure was similar to that used in a previous study (Sakimoto et al., 2013a, 2013b). The electrodes were implanted stereotaxically into the hippocampal region (3.5 mm posterior from bregma, 1.5 mm lateral from the midline and 2.5 mm beneath the skull surface) and the striatum region (0.7 mm anterior from bregma, 2.8 mm lateral from the midline and 4.6 mm beneath the skull surface). Other electrodes were implanted into the frontal cortex, parietal cortex, primary auditory cortex and cerebellar cortex. The EEG data for the frontal cortex, parietal cortex and primary auditory and cerebellar cortex are not shown due to the presence of artifacts. The data obtained from the hemisphere with fewer artifacts and less noise in each rat were used for data analysis. The electrodes were connected to a nine-pin connector (Amphenol, Wallingford, CT, USA) and fixed with dental cement to the screws and skull. After surgery, each rat was given a recovery period of at least 1 week. After the recovery period, the rats were re-trained in the PI task. The rats’ EEG data were recorded throughout the PI task following the extinction schedule. 2.5. Test session and extinction session After the recovery period, the rats were re-trained in the PI task. The EEG data were then recorded during the peak trials for three PI test sessions. Finally, rats were also tested in three extinction sessions similar to the 30-s PI sessions, except no reinforcement was delivered. All parameters were the same as those used during training.
M. Hattori, S. Sakata / Behavioural Processes 101 (2014) 97–102
99
2.6. Behavioral response and recording and EEG analysis 2.6.1. Behavioral response analysis The time of each lever press during each session was recorded in 3-s intervals in both the PI and extinction tasks. The response rate was normalized in amplitude relative to the maximum response rate in the PI task. The response functions were separately fitted to Gaussian functions using a curve-fitting software program (MATLAB R2007b, MathWorks, Inc., Natick, MA, USA). The mean of the fitted function was used as a measure of the peak time. The ratio of the R2 fit values obtained from the Gaussian function was used as a measure of the temporal control exhibited by each rat. 2.6.2. Recording and analysis of EEG The EEG waveforms were referenced to the nasal skull and recorded from each channel. The waveforms were amplified (System 360; NEC Sanei, Tokyo, Japan) and digitized at a sampling rate of 1 kHz, with a time constant of 3.0 s. All EEG analyses were performed using the MATLAB software version R2007b. The EEG analysis focused on the three PI task sessions and the extinction task sessions. The recording and analysis window lasted from 4 s before stimulus onset to 80 s after stimulus onset. The EEG data were computed using an FFT analysis with a 4 s point size. The analysis window was divided into 20 4-s periods, and the epoch average EEG power was computed for each period. The mean EEG power at 4 s before stimulus onset was set as the baseline (no stimuli were present during this period), and the relative EEG power calculated for each subperiod was normalized relative to the 4-s period before stimulus onset (relative theta power of each period = theta power of each period/theta power during the 4-s period before stimulus onset). We analyzed the 6- to 12-Hz frequency band of theta waves in the hippocampus and striatum. Trials with artifacts were eliminated from the FFT analysis. The analysis of theta power was conducted in the probe trials only, that is, the non-RFT trials. The EEG data recording and analysis included three sessions with the PI task and extinction task. 2.7. Histology
Fig. 2. Diagram of the locations of electrode placement in the rat brains. This figure was modified from Paxinos and Watson (1986). The filled circles indicate the placement of the electrode tips (N = 3). Top: striatum, bottom: hippocampal CA1.
3. Results 3.1. Recording sites for EEG The electrode tips were placed in the hippocampal CA1 and striatum areas. Fig. 2 shows the locations of electrode placement for the EEG recording in the rats in which EEG data was free of artifacts (N = 3). 3.2. PI task and extinction task After 50 sessions of PI 30-s training, the mean estimated peak time during the last three sessions was 36.8 ± 5.63 s, with a mean fitting R2 of 0.86. Following electrode implantation, EEG artifacts were observed in half of the rats. Therefore, data analysis was performed on data obtained from the remaining three rats. In these three rats, the mean peak time was 34.9 ± 6.42 s. The fitting R2 in each rat was very high: 0.98, 0.93 and 0.89, respectively. In the
At the end of the experiment, all rats were deeply anesthetized with an overdose of intraperitoneal thiamylal sodium (50 mg/kg) and perfused with saline followed by 10% buffered formalin solution. The brains were removed, fixed for 24 h in 10% buffered formalin and then soaked in 30% sucrose in phosphate buffer, after which they were frozen and sectioned into 50-m slices. The locations of the electrode tips were verified under a microscope. 2.8. Statistical analysis One-tailed t-tests were used to compare the mean relative response rate and the mean relative EEG power in the hippocampal CA1 and striatum in the PI task and the extinction task. Since the intervals of behavioral response (3-s intervals) and EEG theta power (4-s intervals) differed, the one-tailed Pearson correlations between the behavioral responses and EEG theta power were calculated using 12-s, 24-s, 36-s, 48-s, 60-s and 72-s data as the common divisors. Following electrode implantation, EEG artifacts were observed in half of the rats. Therefore, data analysis was performed on data obtained from the remaining three rats. Statistical analyses were performed using the IBM SPSS Statistics 21.0 software package (IBM SPSS, Chicago, IL, USA), with a level of statistical significance set at an alpha = 0.05. All results are expressed as the mean ± SEM.
Fig. 3. Response functions during the probe trials of the PI task and extinction task (N = 3). The time of each lever press in each of the three sessions was recorded in 3-s intervals. The behavioral response was transferred into the relative response rate, meaning that the maximum response rate was 1.0 in the PI task. The filled circles indicate PI tasks. The open circles indicate extinction tasks. The vertical error bars indicate the SEM.
100
M. Hattori, S. Sakata / Behavioural Processes 101 (2014) 97–102
Fig. 4. Relative EEG power values (theta band; 6–12 Hz) observed during the probe trials of the three PI task sessions and three extinction task sessions in the hippocampus (left) and striatum (right) (N = 3). The analysis window was divided into 20 4-s periods. The relative EEG power calculated for each period was normalized relative to the 4-s period before stimulus onset. The filled circles indicate PI tasks. The open circles indicate extinction tasks. The vertical error bars indicate the SEM.
extinction task, the rats’ responses showed no peak pattern with fewer lever press responses compared with that observed in the PI task (t(29) = 14.245, p < 0.001). The response patterns observed in the PI and extinction tasks are shown in Fig. 3.
3.3. EEG power Fig. 4 shows the relative change in power in the hippocampal CA1 and striatum during the task. The striatum EEG power increased with lever pressing, exhibiting a larger change than the hippocampal CA1 EEG power. After the PI task, in the extinction
sessions, both the hippocampal CA1 and striatum EEG power values showed no changes compared to the base power values. The mean EEG power values of the hippocampal CA1 and striatum were significantly increased in the PI task compared to those observed in the extinction task (hippocampal CA1: t(19) = 22.447, p < 0.001, striatum: t(19) = 9.316, p < 0.001).
3.4. Lever response and EEG power A strong correlation was observed between the relative change in EEG power and the increase in lever responses. Therefore,
Fig. 5. Scatterplot with the regression lines of the correlation between the relative theta EEG power and the lever response values in both tasks (N = 3). Since the intervals of the behavioral responses (3-s intervals) and EEG theta power (4-s intervals) differed, the correlation between the behavioral response and EEG theta power was calculated using the 12-s, 24-s, 36-s, 48-s, 60-s and 72-s data as the common divisors. Left: hippocampal CA1, right: striatum. The filled circles indicate PI tasks. The open circles indicate extinction tasks.
M. Hattori, S. Sakata / Behavioural Processes 101 (2014) 97–102
Pearson’s correlation coefficients were calculated to determine the degree of correlation between the lever response values and EEG power in the hippocampal CA1 and striatum. In the PI task, a statistical significance correlation was observed between the lever response values and the striatum EEG power (r = 0.401, p = 0.045), but not the hippocampal CA1 EEG power (r = −0.172, p = 0.247). In the extinction task, no reliable correlations were observed between the lever response values and the hippocampal CA1 (r = 0.144, p = 0.285) or striatum EEG power (r = 0.041, p = 0.435) (Fig. 5).
101
Therefore, the enhanced power of the theta rhythm may not reflect lever pressing, but rather temporal processing. A limitation of this study is the small number of subjects in which EEG data was recorded free of artifacts. Although compatible with other data in the literature, this study should be followed by further studies to clarify the brain circuits involved in interval-timing behavior. More studies correlating physiological and psychological data during interval-timing experiments must be conducted before a clear timing circuit theory would be developed. Acknowledgments
4. Discussion The purpose of this study was to examine variations in the hippocampus and striatum EEG theta power in the peak interval (PI) procedure. We found that the relative EEG theta power reliably increased in the PI task compared to that observed in the extinction task in both regions. Moreover, in the PI task, a statistical significant correlation was observed between the lever response values and the striatum EEG power, but not between the response values and the hippocampal CA1 EEG power. Because an increased EEG power is indicative of firing synchronization of single units, this result indicates that striatum neurons fired more synchronously at the time of reinforcement, thus supporting a critical role for synchronization of firing of striatal neurons in interval timing. The EEG power of the striatum region exhibited a good correlation with the behavior patterns, suggesting that the striatum is important for regulating timing mechanisms in the brain. Indeed, Chiba et al. (2008) analyzed single-unit activity in the striatum in a monkey performing a duration discrimination task, and reported that striatum neurons were able to encode the duration of timing cues. These results provide support for the striatal beat frequency model (Matell and Meck, 2004), which assumes that at the onset of a timed signal, the populations of cortical and thalamic neurons phase reset and synchronize then begin oscillating at their endogenous periodicity (Allman and Meck, 2012). Although we found EEG power differences in the hippocampus between the PI task and the extinction task, the pattern did not correlate with any obvious timing responses/lever pressing. These results raise the possibility that the hippocampus is also important for timing, especially at the starting point. For example, Onoda et al. (2003) examined event-related potentials (ERPs) in rats performing a timing task and reported that the P3-like potential in the hippocampus was larger than that observed in the cortex and thalamus. They suggested that the hippocampal theta activity is reset by stimulus presentation during tasks using working memory. Buhusi and Meck (2006) evaluated the effects of the PI task with gaps and distracters and discussed four hypotheses of timing, including memory decay. One hippocampal lesion study found that the timing behavior was clearly affected during the PI task with gaps (Meck et al., 1984). The authors discussed the role of the hippocampal function with working memory and response inhibition. These results suggest that the hippocampus may play a role in the formation of memory, but it may not have such strong effects on timing. Finally, the hippocampus and striatum theta rhythm is known to be increased during active locomotion (Ledberg and Robbe, 2011; Yamin et al., 2013). It is possible that other variables, such as lever press behaviors, may account for some of the EEG theta power observed in the PI task. If the enhanced power of the hippocampal theta rhythm merely reflects lever pressing, a strong correlation between the lever response values and EEG theta power should have been observed in the extinction task. However, in our study, no correlations were observed between the lever response values and hippocampal CA1 or striatum EEG power in the extinction task.
This work was supported by Grants-in-Aid for Scientific Research (16530471, 16203038, 21300125, 25280051) from the JSPS and a Hiroshima University Research Grant. We would like to thank Erika Takahashi for her excellent assistance with the experiment. References Allman, M.J., Meck, W.H., 2012. Pathophysiological distortions in time perception and timed performance. Brain: A Journal of Neurology 135 (Pt 3), 656–677, http://dx.doi.org/10.1093/brain/awr210. Berke, J.D., Okatan, M., Skurski, J., Eichenbaum, H.B., 2004. Oscillatory entrainment of striatal neurons in freely moving rats. Neuron 43 (6), 883–896, http://dx.doi.org/10.1016/j.neuron.2004.08.035. Buhusi, C.V., Meck, W.H., 2005. What makes us tick? Functional and neural mechanisms of interval timing. Nature Reviews Neuroscience 6 (10), 755–765, http://dx.doi.org/10.1038/Nrn1764. Buhusi, C.V., Meck, W.H., 2006. Time sharing in rats: a peak-interval procedure with gaps and distracters. Behavioural Processes 71 (2–3), 107–115, http://dx.doi.org/10.1016/j.beproc.2005.11.017. Chiba, A., Oshio, K., Inase, M., 2008. Striatal neurons encoded temporal information in duration discrimination task. Experimental Brain Research 186 (4), 671–676, http://dx.doi.org/10.1007/s00221-008-1347-3. DeCoteau, W.E., Thorn, C., Gibson, D.J., Courtemanche, R., Mitra, P., Kubota, Y., Graybiel, A.M., 2007. Learning-related coordination of striatal and hippocampal theta rhythms during acquisition of a procedural maze task. Proceedings of the National Academy of Sciences of United States of America 104 (13), 5644–5649, http://dx.doi.org/10.1073/pnas.0700818104. Dusek, P., Jech, R., Sieger, T., Vymazal, J., Ruzicka, E., Wackermann, J., Mueller, K., 2012. Abnormal activity in the precuneus during time perception in Parkinson’s disease: an fMRI study. PLoS One 7 (1), e29635, http://dx.doi.org/10.1371/journal.pone.0029635. Fujisawa, S., Buzsaki, G., 2011. A 4 Hz oscillation adaptively synchronizes prefrontal, VTA, and hippocampal activities. Neuron 72 (1), 153–165, http://dx.doi.org/10.1016/j.neuron.2011.08.018. Howard, M.W., Eichenbaum, H., 2013. The hippocampus, time, and memory across scales. Journal of Experimental Psychology. General, http://dx.doi.org/10.1037/a0033621. Ledberg, A., Robbe, D., 2011. Locomotion-related oscillatory body movements at 6–12 Hz modulate the hippocampal theta rhythm. PLoS One 6 (11), e27575, http://dx.doi.org/10.1371/journal.pone.0027575. Malapani, C., Rakitin, B., Levy, R., Meck, W.H., Deweer, B., Dubois, B., Gibbon, J., 1998. Coupled temporal memories in Parkinson’s disease: a dopamine-related dysfunction. Journal of Cognitive Neuroscience 10 (3), 316–331. Matell, M.S., Meck, W.H., 2004. Cortico-striatal circuits and interval timing: coincidence detection of oscillatory processes. Brain Research. Cognitive Brain Research 21 (2), 139–170, http://dx.doi.org/10.1016/j.cogbrainres.2004.06.012. Matell, M.S., Meck, W.H., Nicolelis, M.A., 2003. Interval timing and the encoding of signal duration by ensembles of cortical and striatal neurons. Behavioral Neuroscience 117 (4), 760–773. Meck, W.H., Church, R.M., Olton, D.S., 1984. Hippocampus, time, and memory. Behavioral Neuroscience 98 (1), 3–22. Meck, W.H., Penney, T.B., Pouthas, V., 2008. Cortico-striatal representation of time in animals and humans. Current Opinion in Neurobiology 18 (2), 145–152, http://dx.doi.org/10.1016/j.conb.2008.08.002. Onoda, K., Takahashi, E., Sakata, S., 2003. Event-related potentials in the frontal cortex, hippocampus, and cerebellum during a temporal discrimination task in rats. Brain Research. Cognitive Brain Research 17 (2), 380–387. Paxinos, G., Watson, C., 1986. The Rat Brain in Stereotaxic Coordinates, 2nd ed. Academic Press, San Diego. Sakata, S., 2006. Timing and hippocampal theta in animals. Reviews in the Neurosciences 17 (1–2), 157–162. Sakata, S., Onoda, K., 2003. Electrophysiological correlates of interval timing. In: Meck, W.H. (Ed.), Functional and Neural Mechanisms of Interval Timing. CRC Press, Boca Raton, pp. 339–349. Sakimoto, Y., Hattori, M., Takeda, K., Okada, K., Sakata, S., 2013a. Hippocampal theta wave activity during configural and non-configural
102
M. Hattori, S. Sakata / Behavioural Processes 101 (2014) 97–102
tasks in rats. Experimental Brain Research 225 (2), 177–185, http://dx.doi.org/10.1007/s00221-012-3359-2. Sakimoto, Y., Takeda, K., Okada, K., Hattori, M., Sakata, S., 2013b. Transient decline in rats’ hippocampal theta power relates to inhibitory stimulus-reward association. Behavioural Brain Research 246, 132–138, http://dx.doi.org/10.1016/j.bbr.2013.02.012.
Yamin, H.G., Stern, E.A., Cohen, D., 2013. Parallel processing of environmental recognition and locomotion in the mouse striatum. Journal of Neuroscience 33 (2), 473–484, http://dx.doi.org/10.1523/JNEUROSCI.4474-12.2013. Yin, B., Troger, A.B., 2011. Exploring the 4th dimension: hippocampus, time, and memory revisited. Frontiers in Integrative Neuroscience 5, 36, http://dx.doi.org/10.3389/fnint.2011.00036.
Contents lists available at ScienceDirect
Behavioural Processes journal homepage: www.elsevier.com/locate/behavproc
Brain electrophysiological activity correlates with temporal processing in rats Minoru Hattori a , Shogo Sakata b,∗ a b
Institute of Biomedical & Health Sciences, Hiroshima University, Hiroshima 734-8553, Japan Graduate School of Integrated Arts and Sciences, Hiroshima University, Higashi-Hiroshima 739-8521, Japan
a r t i c l e
i n f o
Article history: Received 15 April 2013 Received in revised form 12 September 2013 Accepted 17 September 2013 Keywords: Hippocampus Striatum Timing EEG power
a b s t r a c t In this study, we present electroencephalographic (EEG) recording data obtained in correlation with timing behavior in rats trained in a 30-s peak interval (PI) procedure. The distribution of lever press responses was found to be Gaussian, peaking at approximately 30 s: lever pressing behavior increased for 30 s then decreased after the reinforcement time. We recorded EEG activity in the hippocampus (hippocampal theta wave) and striatum during the task, and evaluated whether the EEG power correlated with the behavior pattern. We found that the striatum EEG, but not the hippocampal theta wave, showed a good correlation with the response pattern in the 30-s PI. This result suggests that striatum neurons fired more synchronously at the time of reinforcement, thus supporting a critical role for synchronization of firing of striatal neurons in regulating timing mechanisms. This article is part of a Special Issue entitled: Associative and Temporal Learning. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Timing is everything in living animals. It is very important to study timing in order to understand animal behavior. Interval timing is the ability of animals to perceive and process durations in the seconds-to-minutes range (Buhusi and Meck, 2005). How do animals estimate the passage of time? Numerous human and animal studies have demonstrated that the brain is equipped with multiple time-measurement systems associated with interval-timing behavior. The role of the hippocampus in temporal discrimination learning was initially explored by Meck et al. (1984). More recently, the hippocampus has become a primary target region in the study of the cognitive functions associated with temporal discrimination learning (Howard and Eichenbaum, 2013; Sakata, 2006; Yin and Troger, 2011). For example, in studies using the peak-interval (PI) procedure, rats with hippocampal lesions show a small leftward shift in their temporal discrimination function (Meck et al., 1984). Moreover, a frequency analysis of the hippocampal theta wave revealed an increase in power during a duration discrimination task compared with that observed in a simple reaction-time task (Sakata and Onoda, 2003). The hippocampal EEG theta wave is a large amplitude sinusoidal wave reflecting brain oscillation activity in the hippocampus (see Fujisawa and Buzsaki, 2011;
∗ Corresponding author. Tel.: +81 82 424 6581; fax: +81 82 424 0759. E-mail address: [email protected] (S. Sakata). 0376-6357/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.beproc.2013.09.011
Fig. 1 for typical raw wave data). These results suggest that the hippocampus contributes to timing and the perception of time. A variety of neuropsychological, functional neuroimaging and pharmacological studies suggest that the striatum, the major input structure of the basal ganglia complex, is an essential component of the neural networks involved in time perception. Neurons in the striatum are able to read time codes emitted by oscillator cells in the cortex (Matell et al., 2003). Impaired time perception has been found in patients with Parkinson’s disease (PD), attributed to the presence of a dysfunctional dopaminergic striatal pacemaker in such patients (Dusek et al., 2012; Malapani et al., 1998). The hippocampal formation sends projections to several structures, including the prefrontal cortex and ventral striatum. Both structures exhibit coherent theta rhythmicity. Previous studies have demonstrated that coherence between the striatum and hippocampus at the theta frequency is strong in the ventral/medial striatum and weak in the dorsal/lateral striatum (Berke et al., 2004). Pathways interconnecting the hippocampus and striatum are thought to use theta rhythms to transfer and coordinate neural representations in hippocampus–striatum circuits in relation to procedural maze tasks (DeCoteau et al., 2007). Should the hippocampus and striatum contribute to time perception, then the theta power of the hippocampus and striatum should increase during interval timing tasks. The purpose of this study was to examine the variation in the hippocampus and striatum theta power related to temporal discrimination tasks, in particular during the peak-interval (PI) procedure. Among temporal discrimination tasks, the PI procedure is a
98
M. Hattori, S. Sakata / Behavioural Processes 101 (2014) 97–102
Fig. 1. The experimental procedures of the PI training and PI test sessions. The horizontal lines represent time. The square peaks on the “tone” axis indicate the presentation of pure tones. The vertical lines on the “response” axis indicate the lever response. The oblique lines indicate the supply of food pellets. The PI task consisted of two types of trials: FI trials (50%) and peak trials (50%). The rats were trained using a 30-s FI 30 schedule (FI), during which the first lever press that occurred 30 s after the beginning of the signal triggered the delivery of a food reward. PI sessions included probe trials in which the lever was extended for 90 s and the lever presses were not reinforced. In the test session, the EEG recording window started 4 s before stimulus onset and ended 80 s after stimulus onset in the peak trial.
powerful tool for studying timing in rats (Meck et al., 2008). In the PI procedure, the animals are trained to respond in the presence of a stimulus after a given amount of time has elapsed. Therefore, we hypothesized that the pattern of response in a peak procedure should correlate with theta waves in the hippocampus and striatum. 2. Methods 2.1. Animals The subjects were six male Wistar rats (CLEA Japan, Inc., Osaka, Japan). Rats were 12 weeks old when the experiment began. Rats were housed in individual cages and kept on a 12:12-h light–dark cycle (lights on at 8:00 a.m.). Throughout the experiment, all rats were maintained at 85% of their ad libitum weight. Water was freely available throughout the experiment. The experiment was conducted in accordance with the guidelines published by the Japan Society for Animal Psychology and approved by the Committee of Experimental Animals at Hiroshima University (D08-6). 2.2. Apparatus Behavioral training and EEG recording sessions took place in a standard operant chamber (ENV-007 CT; MED Associates, Inc., Georgia, VT, USA). According to the procedure, 45-mg food pellets (F0165; Bio Serv, Frenchtown, NJ, USA) were delivered in a cup was placed at the center of the front wall at the floor level. A food dispenser delivered the pellets to the cup with a click noise. A lever was positioned on the left side of the front aluminum wall. The chambers were equipped with a house light and a back wall-mounted speaker to provide tone stimuli. The experimental control and lever pressing data recording was performed using a PC running an inhouse software program developed in Delphi (Borland Software Corporation). The chambers were housed in a soundproof, ventilated and electrically shielded room and monitored with a video camera (Mother tool CO. Model MTB-2081P). 2.3. Procedures 2.3.1. Initial training In the initial preparation training session, the rats received 3 days of training for lever pressing for up to 60 pellets or 30 min. Next, the rats progressed through the following training sequence: (a) a continuous reinforcement (CRF) schedule for 60 pellets; (b) a fixed ratio 3 (FR3) reinforcement schedule; (c) an FR6 reinforcement schedule; (d) an FR10 reinforcement schedule for 60 trials per session. One training and testing session were conducted per day. Afterwards, rats were trained in the timing procedures described below.
2.3.2. Fixed-interval (FI) training All rats received 15 sessions of a 30-s FI schedule in which a tone signal (2000 Hz, 80 dB) signaled the to-be-timed fixed interval. The sessions consisted of 60 trials, during which the first lever press that occurred 30-s after the beginning of the signal triggered the delivery of a food reward and terminated the tone signal for the duration of the variable intertrial interval (ITI) (average ITI 30 s, range 20–40 s). 2.3.3. Peak-interval (PI) task training After the FI training, the rats received 50 sessions of PI task training (Fig. 1). The PI session consisted of 60 trials, of which 50% were FI trials and 50% were PI trials. During the PI trials, the lever was extended for 90 s and the lever presses were not reinforced. The peak trials and FI trials were randomly intermixed with the restriction that no more than four peak trials occurred consecutively. As in the FI training, the trials were separated by variable ITIs (average 30-s, range 20–40 s). Afterwards, the rats underwent surgery for electrode implantation. 2.4. Surgery The electrode implantation procedure was similar to that used in a previous study (Sakimoto et al., 2013a, 2013b). The electrodes were implanted stereotaxically into the hippocampal region (3.5 mm posterior from bregma, 1.5 mm lateral from the midline and 2.5 mm beneath the skull surface) and the striatum region (0.7 mm anterior from bregma, 2.8 mm lateral from the midline and 4.6 mm beneath the skull surface). Other electrodes were implanted into the frontal cortex, parietal cortex, primary auditory cortex and cerebellar cortex. The EEG data for the frontal cortex, parietal cortex and primary auditory and cerebellar cortex are not shown due to the presence of artifacts. The data obtained from the hemisphere with fewer artifacts and less noise in each rat were used for data analysis. The electrodes were connected to a nine-pin connector (Amphenol, Wallingford, CT, USA) and fixed with dental cement to the screws and skull. After surgery, each rat was given a recovery period of at least 1 week. After the recovery period, the rats were re-trained in the PI task. The rats’ EEG data were recorded throughout the PI task following the extinction schedule. 2.5. Test session and extinction session After the recovery period, the rats were re-trained in the PI task. The EEG data were then recorded during the peak trials for three PI test sessions. Finally, rats were also tested in three extinction sessions similar to the 30-s PI sessions, except no reinforcement was delivered. All parameters were the same as those used during training.
M. Hattori, S. Sakata / Behavioural Processes 101 (2014) 97–102
99
2.6. Behavioral response and recording and EEG analysis 2.6.1. Behavioral response analysis The time of each lever press during each session was recorded in 3-s intervals in both the PI and extinction tasks. The response rate was normalized in amplitude relative to the maximum response rate in the PI task. The response functions were separately fitted to Gaussian functions using a curve-fitting software program (MATLAB R2007b, MathWorks, Inc., Natick, MA, USA). The mean of the fitted function was used as a measure of the peak time. The ratio of the R2 fit values obtained from the Gaussian function was used as a measure of the temporal control exhibited by each rat. 2.6.2. Recording and analysis of EEG The EEG waveforms were referenced to the nasal skull and recorded from each channel. The waveforms were amplified (System 360; NEC Sanei, Tokyo, Japan) and digitized at a sampling rate of 1 kHz, with a time constant of 3.0 s. All EEG analyses were performed using the MATLAB software version R2007b. The EEG analysis focused on the three PI task sessions and the extinction task sessions. The recording and analysis window lasted from 4 s before stimulus onset to 80 s after stimulus onset. The EEG data were computed using an FFT analysis with a 4 s point size. The analysis window was divided into 20 4-s periods, and the epoch average EEG power was computed for each period. The mean EEG power at 4 s before stimulus onset was set as the baseline (no stimuli were present during this period), and the relative EEG power calculated for each subperiod was normalized relative to the 4-s period before stimulus onset (relative theta power of each period = theta power of each period/theta power during the 4-s period before stimulus onset). We analyzed the 6- to 12-Hz frequency band of theta waves in the hippocampus and striatum. Trials with artifacts were eliminated from the FFT analysis. The analysis of theta power was conducted in the probe trials only, that is, the non-RFT trials. The EEG data recording and analysis included three sessions with the PI task and extinction task. 2.7. Histology
Fig. 2. Diagram of the locations of electrode placement in the rat brains. This figure was modified from Paxinos and Watson (1986). The filled circles indicate the placement of the electrode tips (N = 3). Top: striatum, bottom: hippocampal CA1.
3. Results 3.1. Recording sites for EEG The electrode tips were placed in the hippocampal CA1 and striatum areas. Fig. 2 shows the locations of electrode placement for the EEG recording in the rats in which EEG data was free of artifacts (N = 3). 3.2. PI task and extinction task After 50 sessions of PI 30-s training, the mean estimated peak time during the last three sessions was 36.8 ± 5.63 s, with a mean fitting R2 of 0.86. Following electrode implantation, EEG artifacts were observed in half of the rats. Therefore, data analysis was performed on data obtained from the remaining three rats. In these three rats, the mean peak time was 34.9 ± 6.42 s. The fitting R2 in each rat was very high: 0.98, 0.93 and 0.89, respectively. In the
At the end of the experiment, all rats were deeply anesthetized with an overdose of intraperitoneal thiamylal sodium (50 mg/kg) and perfused with saline followed by 10% buffered formalin solution. The brains were removed, fixed for 24 h in 10% buffered formalin and then soaked in 30% sucrose in phosphate buffer, after which they were frozen and sectioned into 50-m slices. The locations of the electrode tips were verified under a microscope. 2.8. Statistical analysis One-tailed t-tests were used to compare the mean relative response rate and the mean relative EEG power in the hippocampal CA1 and striatum in the PI task and the extinction task. Since the intervals of behavioral response (3-s intervals) and EEG theta power (4-s intervals) differed, the one-tailed Pearson correlations between the behavioral responses and EEG theta power were calculated using 12-s, 24-s, 36-s, 48-s, 60-s and 72-s data as the common divisors. Following electrode implantation, EEG artifacts were observed in half of the rats. Therefore, data analysis was performed on data obtained from the remaining three rats. Statistical analyses were performed using the IBM SPSS Statistics 21.0 software package (IBM SPSS, Chicago, IL, USA), with a level of statistical significance set at an alpha = 0.05. All results are expressed as the mean ± SEM.
Fig. 3. Response functions during the probe trials of the PI task and extinction task (N = 3). The time of each lever press in each of the three sessions was recorded in 3-s intervals. The behavioral response was transferred into the relative response rate, meaning that the maximum response rate was 1.0 in the PI task. The filled circles indicate PI tasks. The open circles indicate extinction tasks. The vertical error bars indicate the SEM.
100
M. Hattori, S. Sakata / Behavioural Processes 101 (2014) 97–102
Fig. 4. Relative EEG power values (theta band; 6–12 Hz) observed during the probe trials of the three PI task sessions and three extinction task sessions in the hippocampus (left) and striatum (right) (N = 3). The analysis window was divided into 20 4-s periods. The relative EEG power calculated for each period was normalized relative to the 4-s period before stimulus onset. The filled circles indicate PI tasks. The open circles indicate extinction tasks. The vertical error bars indicate the SEM.
extinction task, the rats’ responses showed no peak pattern with fewer lever press responses compared with that observed in the PI task (t(29) = 14.245, p < 0.001). The response patterns observed in the PI and extinction tasks are shown in Fig. 3.
3.3. EEG power Fig. 4 shows the relative change in power in the hippocampal CA1 and striatum during the task. The striatum EEG power increased with lever pressing, exhibiting a larger change than the hippocampal CA1 EEG power. After the PI task, in the extinction
sessions, both the hippocampal CA1 and striatum EEG power values showed no changes compared to the base power values. The mean EEG power values of the hippocampal CA1 and striatum were significantly increased in the PI task compared to those observed in the extinction task (hippocampal CA1: t(19) = 22.447, p < 0.001, striatum: t(19) = 9.316, p < 0.001).
3.4. Lever response and EEG power A strong correlation was observed between the relative change in EEG power and the increase in lever responses. Therefore,
Fig. 5. Scatterplot with the regression lines of the correlation between the relative theta EEG power and the lever response values in both tasks (N = 3). Since the intervals of the behavioral responses (3-s intervals) and EEG theta power (4-s intervals) differed, the correlation between the behavioral response and EEG theta power was calculated using the 12-s, 24-s, 36-s, 48-s, 60-s and 72-s data as the common divisors. Left: hippocampal CA1, right: striatum. The filled circles indicate PI tasks. The open circles indicate extinction tasks.
M. Hattori, S. Sakata / Behavioural Processes 101 (2014) 97–102
Pearson’s correlation coefficients were calculated to determine the degree of correlation between the lever response values and EEG power in the hippocampal CA1 and striatum. In the PI task, a statistical significance correlation was observed between the lever response values and the striatum EEG power (r = 0.401, p = 0.045), but not the hippocampal CA1 EEG power (r = −0.172, p = 0.247). In the extinction task, no reliable correlations were observed between the lever response values and the hippocampal CA1 (r = 0.144, p = 0.285) or striatum EEG power (r = 0.041, p = 0.435) (Fig. 5).
101
Therefore, the enhanced power of the theta rhythm may not reflect lever pressing, but rather temporal processing. A limitation of this study is the small number of subjects in which EEG data was recorded free of artifacts. Although compatible with other data in the literature, this study should be followed by further studies to clarify the brain circuits involved in interval-timing behavior. More studies correlating physiological and psychological data during interval-timing experiments must be conducted before a clear timing circuit theory would be developed. Acknowledgments
4. Discussion The purpose of this study was to examine variations in the hippocampus and striatum EEG theta power in the peak interval (PI) procedure. We found that the relative EEG theta power reliably increased in the PI task compared to that observed in the extinction task in both regions. Moreover, in the PI task, a statistical significant correlation was observed between the lever response values and the striatum EEG power, but not between the response values and the hippocampal CA1 EEG power. Because an increased EEG power is indicative of firing synchronization of single units, this result indicates that striatum neurons fired more synchronously at the time of reinforcement, thus supporting a critical role for synchronization of firing of striatal neurons in interval timing. The EEG power of the striatum region exhibited a good correlation with the behavior patterns, suggesting that the striatum is important for regulating timing mechanisms in the brain. Indeed, Chiba et al. (2008) analyzed single-unit activity in the striatum in a monkey performing a duration discrimination task, and reported that striatum neurons were able to encode the duration of timing cues. These results provide support for the striatal beat frequency model (Matell and Meck, 2004), which assumes that at the onset of a timed signal, the populations of cortical and thalamic neurons phase reset and synchronize then begin oscillating at their endogenous periodicity (Allman and Meck, 2012). Although we found EEG power differences in the hippocampus between the PI task and the extinction task, the pattern did not correlate with any obvious timing responses/lever pressing. These results raise the possibility that the hippocampus is also important for timing, especially at the starting point. For example, Onoda et al. (2003) examined event-related potentials (ERPs) in rats performing a timing task and reported that the P3-like potential in the hippocampus was larger than that observed in the cortex and thalamus. They suggested that the hippocampal theta activity is reset by stimulus presentation during tasks using working memory. Buhusi and Meck (2006) evaluated the effects of the PI task with gaps and distracters and discussed four hypotheses of timing, including memory decay. One hippocampal lesion study found that the timing behavior was clearly affected during the PI task with gaps (Meck et al., 1984). The authors discussed the role of the hippocampal function with working memory and response inhibition. These results suggest that the hippocampus may play a role in the formation of memory, but it may not have such strong effects on timing. Finally, the hippocampus and striatum theta rhythm is known to be increased during active locomotion (Ledberg and Robbe, 2011; Yamin et al., 2013). It is possible that other variables, such as lever press behaviors, may account for some of the EEG theta power observed in the PI task. If the enhanced power of the hippocampal theta rhythm merely reflects lever pressing, a strong correlation between the lever response values and EEG theta power should have been observed in the extinction task. However, in our study, no correlations were observed between the lever response values and hippocampal CA1 or striatum EEG power in the extinction task.
This work was supported by Grants-in-Aid for Scientific Research (16530471, 16203038, 21300125, 25280051) from the JSPS and a Hiroshima University Research Grant. We would like to thank Erika Takahashi for her excellent assistance with the experiment. References Allman, M.J., Meck, W.H., 2012. Pathophysiological distortions in time perception and timed performance. Brain: A Journal of Neurology 135 (Pt 3), 656–677, http://dx.doi.org/10.1093/brain/awr210. Berke, J.D., Okatan, M., Skurski, J., Eichenbaum, H.B., 2004. Oscillatory entrainment of striatal neurons in freely moving rats. Neuron 43 (6), 883–896, http://dx.doi.org/10.1016/j.neuron.2004.08.035. Buhusi, C.V., Meck, W.H., 2005. What makes us tick? Functional and neural mechanisms of interval timing. Nature Reviews Neuroscience 6 (10), 755–765, http://dx.doi.org/10.1038/Nrn1764. Buhusi, C.V., Meck, W.H., 2006. Time sharing in rats: a peak-interval procedure with gaps and distracters. Behavioural Processes 71 (2–3), 107–115, http://dx.doi.org/10.1016/j.beproc.2005.11.017. Chiba, A., Oshio, K., Inase, M., 2008. Striatal neurons encoded temporal information in duration discrimination task. Experimental Brain Research 186 (4), 671–676, http://dx.doi.org/10.1007/s00221-008-1347-3. DeCoteau, W.E., Thorn, C., Gibson, D.J., Courtemanche, R., Mitra, P., Kubota, Y., Graybiel, A.M., 2007. Learning-related coordination of striatal and hippocampal theta rhythms during acquisition of a procedural maze task. Proceedings of the National Academy of Sciences of United States of America 104 (13), 5644–5649, http://dx.doi.org/10.1073/pnas.0700818104. Dusek, P., Jech, R., Sieger, T., Vymazal, J., Ruzicka, E., Wackermann, J., Mueller, K., 2012. Abnormal activity in the precuneus during time perception in Parkinson’s disease: an fMRI study. PLoS One 7 (1), e29635, http://dx.doi.org/10.1371/journal.pone.0029635. Fujisawa, S., Buzsaki, G., 2011. A 4 Hz oscillation adaptively synchronizes prefrontal, VTA, and hippocampal activities. Neuron 72 (1), 153–165, http://dx.doi.org/10.1016/j.neuron.2011.08.018. Howard, M.W., Eichenbaum, H., 2013. The hippocampus, time, and memory across scales. Journal of Experimental Psychology. General, http://dx.doi.org/10.1037/a0033621. Ledberg, A., Robbe, D., 2011. Locomotion-related oscillatory body movements at 6–12 Hz modulate the hippocampal theta rhythm. PLoS One 6 (11), e27575, http://dx.doi.org/10.1371/journal.pone.0027575. Malapani, C., Rakitin, B., Levy, R., Meck, W.H., Deweer, B., Dubois, B., Gibbon, J., 1998. Coupled temporal memories in Parkinson’s disease: a dopamine-related dysfunction. Journal of Cognitive Neuroscience 10 (3), 316–331. Matell, M.S., Meck, W.H., 2004. Cortico-striatal circuits and interval timing: coincidence detection of oscillatory processes. Brain Research. Cognitive Brain Research 21 (2), 139–170, http://dx.doi.org/10.1016/j.cogbrainres.2004.06.012. Matell, M.S., Meck, W.H., Nicolelis, M.A., 2003. Interval timing and the encoding of signal duration by ensembles of cortical and striatal neurons. Behavioral Neuroscience 117 (4), 760–773. Meck, W.H., Church, R.M., Olton, D.S., 1984. Hippocampus, time, and memory. Behavioral Neuroscience 98 (1), 3–22. Meck, W.H., Penney, T.B., Pouthas, V., 2008. Cortico-striatal representation of time in animals and humans. Current Opinion in Neurobiology 18 (2), 145–152, http://dx.doi.org/10.1016/j.conb.2008.08.002. Onoda, K., Takahashi, E., Sakata, S., 2003. Event-related potentials in the frontal cortex, hippocampus, and cerebellum during a temporal discrimination task in rats. Brain Research. Cognitive Brain Research 17 (2), 380–387. Paxinos, G., Watson, C., 1986. The Rat Brain in Stereotaxic Coordinates, 2nd ed. Academic Press, San Diego. Sakata, S., 2006. Timing and hippocampal theta in animals. Reviews in the Neurosciences 17 (1–2), 157–162. Sakata, S., Onoda, K., 2003. Electrophysiological correlates of interval timing. In: Meck, W.H. (Ed.), Functional and Neural Mechanisms of Interval Timing. CRC Press, Boca Raton, pp. 339–349. Sakimoto, Y., Hattori, M., Takeda, K., Okada, K., Sakata, S., 2013a. Hippocampal theta wave activity during configural and non-configural
102
M. Hattori, S. Sakata / Behavioural Processes 101 (2014) 97–102
tasks in rats. Experimental Brain Research 225 (2), 177–185, http://dx.doi.org/10.1007/s00221-012-3359-2. Sakimoto, Y., Takeda, K., Okada, K., Hattori, M., Sakata, S., 2013b. Transient decline in rats’ hippocampal theta power relates to inhibitory stimulus-reward association. Behavioural Brain Research 246, 132–138, http://dx.doi.org/10.1016/j.bbr.2013.02.012.
Yamin, H.G., Stern, E.A., Cohen, D., 2013. Parallel processing of environmental recognition and locomotion in the mouse striatum. Journal of Neuroscience 33 (2), 473–484, http://dx.doi.org/10.1523/JNEUROSCI.4474-12.2013. Yin, B., Troger, A.B., 2011. Exploring the 4th dimension: hippocampus, time, and memory revisited. Frontiers in Integrative Neuroscience 5, 36, http://dx.doi.org/10.3389/fnint.2011.00036.
