**4. Hippocampal theta activity during negative patterning**

It has been known that electroencephalography (EEG) was useful for neural activity of hippocampus without extensive hippocampal lesions in rodents. When we implanted a recording polar into the rats' hippocampal CA1 area, we can observe rhythmic EEG patterns. The EEG activity was grouped: theta waves (6–12 Hz), beta waves (12–30 Hz), gamma waves (30–100 Hz), and ripple waves (100–200 Hz). Specifically, it was known that hippocampal theta wave is related to psychological state and behavior. Several studies have reported that hippocampal theta waves strongly are related to locomotor behavior such as running, jumping, ricking, and operant response [21–24]. In addition, it has reported that hippocampus theta activity is related to learning and memory [25–43]. Masuoka et al. [29] showed that rats' hippocampal theta activity increased during elevated radical eight mazes. Also, Olvera-Cortés et al. [30, 31] revealed that the hippocampal theta waves change during performance of spatial learning task. Thus, in addition, theta waves are thought to occur by the synchronization of neurons in the whole hippocampal formation [44], which would reflect hippocampal neural activity [45–48].

Recently, we have examined hippocampal theta activity in rats during the acquisition stages (early, middle, and late) of the negative patterning task (T+, L+, TL−) [38]. We observed a transient decrease in hippocampal theta power immediately after the presentation of a compound stimulus during the late stage of learning in the negative patterning task (**Figure 1**). In addition, the magnitude of the decrease in theta power strongly correlated with improved performance in the negative patterning task (**Figure 2**). Grastyán et al. [49] examined the relationship between hippocampal theta activity and the acquisition of an orientative conditioned response (CR) for a tone stimulus presentation in cats. Although the hippocampal theta activity increased with an association between stimulus and orientative CR, the hippocampal theta wave decreased after the formation of this association. Thus, the transient decrease in hippocampal theta activity during the late stage of learning in the negative patterning task observed in the current study may be related to mastery of the negative patterning task. However, our previous reports showed a greater decrease in hippocampal theta activity in the late learning stage of a negative patterning task compared to the simple discrimination task [32, 38]. Therefore, we suggest that the decrease in hippocampal theta power is induced by hippocampus-mediated information processing for compound stimuli in the negative patterning task. This is in agreement with the concepts of the configural association theory and conflict resolution model.

Further studies revealed characteristics of compound stimuli inducing a decrease in theta power by comparing simultaneous feature-negative (T+, TL−) and compound stimulus discrimination tasks (T1 +, T2 L−) [32]. In feature negative tasks, the compound stimulus had overlapping elements with single stimuli because these stimuli were composed of tone stimuli with the same frequency component. However, the compound stimulus in the compound stimulus discrimination task did not have overlapping elements with single stimuli because they were composed of tone stimuli with different frequency components (*T*<sup>1</sup> : 2000 Hz, *T*<sup>2</sup> : 4000 Hz). These studies reported a transient decrease in hippocampal theta activity following the presentation of a compound stimulus during the simultaneous feature-negative task compared to the simple discrimination task but not during the compound stimulus discrimination task. The compound stimulus of the simultaneous feature-negative task had an overlapping element shared with the single stimulus. This may justify the transient decrease in hippocampal theta activity during response inhibition for the compound stimulus of negative patterning and simultaneous feature-negative tasks. Therefore, we proposed that the decrease in hippocampal theta power is related to behavioral inhibition for conflict stimulus discrimination in which the single stimuli have overlapping elements.

theta waves strongly are related to locomotor behavior such as running, jumping, ricking, and operant response [21–24]. In addition, it has reported that hippocampus theta activity is related to learning and memory [25–43]. Masuoka et al. [29] showed that rats' hippocampal theta activity increased during elevated radical eight mazes. Also, Olvera-Cortés et al. [30, 31] revealed that the hippocampal theta waves change during performance of spatial learning task. Thus, in addition, theta waves are thought to occur by the synchronization of neurons in the whole hippocampal formation [44], which would reflect hippocampal neural activity

Recently, we have examined hippocampal theta activity in rats during the acquisition stages (early, middle, and late) of the negative patterning task (T+, L+, TL−) [38]. We observed a transient decrease in hippocampal theta power immediately after the presentation of a compound stimulus during the late stage of learning in the negative patterning task (**Figure 1**). In addition, the magnitude of the decrease in theta power strongly correlated with improved performance in the negative patterning task (**Figure 2**). Grastyán et al. [49] examined the relationship between hippocampal theta activity and the acquisition of an orientative conditioned response (CR) for a tone stimulus presentation in cats. Although the hippocampal theta activity increased with an association between stimulus and orientative CR, the hippocampal theta wave decreased after the formation of this association. Thus, the transient decrease in hippocampal theta activity during the late stage of learning in the negative patterning task observed in the current study may be related to mastery of the negative patterning task. However, our previous reports showed a greater decrease in hippocampal theta activity in the late learning stage of a negative patterning task compared to the simple discrimination task [32, 38]. Therefore, we suggest that the decrease in hippocampal theta power is induced by hippocampus-mediated information processing for compound stimuli in the negative patterning task. This is in agreement with the concepts of the configural association theory and conflict

Further studies revealed characteristics of compound stimuli inducing a decrease in theta power by comparing simultaneous feature-negative (T+, TL−) and compound stimulus

had overlapping elements with single stimuli because these stimuli were composed of tone stimuli with the same frequency component. However, the compound stimulus in the compound stimulus discrimination task did not have overlapping elements with single stimuli because they were composed of tone stimuli with different frequency components

theta activity following the presentation of a compound stimulus during the simultaneous feature-negative task compared to the simple discrimination task but not during the compound stimulus discrimination task. The compound stimulus of the simultaneous feature-negative task had an overlapping element shared with the single stimulus. This may justify the transient decrease in hippocampal theta activity during response inhibition for the compound stimulus of negative patterning and simultaneous feature-negative tasks. Therefore, we proposed that the decrease in hippocampal theta power is related to behavioral inhibition for conflict stimulus discrimination in which the single stimuli have

L−) [32]. In feature negative tasks, the compound stimulus

: 4000 Hz). These studies reported a transient decrease in hippocampal

[45–48].

14 Electroencephalography

resolution model.

(*T*<sup>1</sup>

discrimination tasks (T1

: 2000 Hz, *T*<sup>2</sup>

overlapping elements.

+, T2

**Figure 1.** The change in theta power during a presentation of compound stimuli of the negative patterning task by using wavelet analysis (A). Upper side shows the change in hippocampal theta activity along a time course during compound stimulus on the early stage, Middle side shows theta activity on the middle stage, and Lower side shows theta activity on late stage of negative patterning task. The *x*-axis is time (ms), and the *y*-axis is frequency (Hz). In each panel, the period is from 500 ms before stimulus onset to 4000 ms after stimulus onset. The mean hippocampal theta power during 500 ms before stimulus onset was counted as the −500-ms period (no stimuli were present and no rats pressed the lever during this period), and the relative theta power calculated for each period (per 250 ms) was normalized to that during the −500 ms period (relative theta activity of each period = theta power of each period/theta power at the −500-ms period). Panel B contains a comparison of the mean (± S.E.M.) relative hippocampal theta activity at 6–12 Hz among each learning stage (early, middle, and late) throughout the time course of the experiment during compound stimuli of the negative patterning task. Two-way within-subjects ANOVA suggests that there is a significant interaction of learning stages (early, middle, and late) × epochs (−500 to 4000 ms, with each 250 ms; *F*(36,180) = 2.37, *p* < 0.05) and a significant effect of epochs (*F*(18,90) = 4.80, *p* < 0.05), but no significant effect of stages (*F*(2,10) = 0.97, n.s.) on relative hippocampal theta power during compound stimulus of the negative patterning task. *Post-hoc* tests showed that there was a significant simple main effect in the 250 and 500-ms epochs during compound stimulus. Multiple comparisons revealed that hippocampal theta power decreased in the 250-ms epochs during nonRFTs in the late stage compared with the early stage (*p* < 0.05) and in the 500-ms epochs during nonRFTs in the middle and late stages compared with the early stage (\* *p* < 0.05). Panel C contains a comparison of the mean (± S.E.M.) relative hippocampal theta activity at 6–12 Hz among each learning stage (early, middle, and late) throughout the time course of the experiment during nonreinforced stimulus of the simple discrimination task. This figure was referred to Sakimoto et al. [38].

**Figure 2.** A comparison of the mean relative hippocampal theta activity between tasks. Panel A shows the relative hippocampal theta power during the 500-ms epochs between the negative patterning and simple discrimination task groups. A group (negative patterning task and simple discrimination task groups) × stage (early, middle, and late) ANOVA for hippocampal theta activity during a 500-ms epoch in the nonRFT showed a significant interaction (*F*(2,20) = 6.12, *p* < 0.05). Multiple comparisons revealed that hippocampal theta power decreased during the late stage in the negative patterning task compared to the simple discrimination task group (*p* < 0.05; \*: *p* < 0.05). Hippocampal theta power during the 500 ms nonRFT correlated with the discrimination rate in the negative patterning task (*r* = −0.70, *p* < 0.05; panel B) but not the simple discrimination task (*r* = −0.06, *p* = n.s; panel C). This figure was referred to Sakimoto et al. [38].

#### **5. Why does hippocampal theta amplitude decline?**

Hippocampal theta power is affected by the activity of cholinergic and γ-aminobutyric acid (GABA) neurons of the medial septal/diagonal band area [44]. Monmaur and Breton [50] demonstrated that theta activity increases when the cholinergic agonist, carbachol, is injected into the intra septum in freely moving rats. In addition, Sun et al. [51] reported that hippocampal theta activity is abolished by the GABA antagonist bicuculline. Thus, we propose that the transient decrease in hippocampal theta activity during compound stimulus learning in the negative patterning task is induced by the activity of septal cholinergic or GABAergic neurons, or their interaction. In future studies, the relationship between the negative patterning task and septal cholinergic and/or GABAergic activity should be examined. Because septo-hippocampal GABAergic input to CA1 is essential for the generation of theta waves [52], the transient decrease in presynaptic GABA release may cause a transient decrease in the hippocampal theta power immediately after stimulus presentation. We recently analyzed synaptic plasticity using the slice patch-clamp technique and measured a rapid decrease in presynaptic GABA release at hippocampal CA1 synapses immediately after the nonspatial contextual learning task, inhibitory avoidance (IA) [53]. Compared to untrained controls, the paired pulse ratio (PPR) of evoked inhibitory postsynaptic current (IPSC) increased immediately after IA training (at 0 min), suggesting an acute decrease in the probability of presynaptic GABA release. As the PPR returned to baseline 5 min after the training, the decrease in presynaptic GABA release seems to be transient. Moreover, we observed a sustained increase in the miniature excitatory postsynaptic current (mEPSC) and miniature inhibitory postsynaptic current (mIPSC) amplitudes 5–30 min after the IA task, suggesting long-term postsynaptic strengthening of α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) and GABAA receptor-mediated synapses. In addition, the long-term increase in mIPSC frequency is probably due to an increase in the number of GABAA receptor-mediated inhibitory synapses after the training [53].
