The main result reported here is that one week CEF treatment resulted in a delayed reduction in EEG theta power (7–9 Hz) in both frontal and parietal derivations. This decrease peaked at day 10, 2 days after the end of treatment, and disappeared by day 16. In addition, we found that the same CEF treatment increased motor activity, especially during those experimental days in which the EEG changes are more prominent.
Since EEG is sensitive to different environmental conditions and animal manipulations 
, we assessed the effects of CEF by using a within-subjects design, comparing data before and after CEF treatment. To rule out the possibility that EEG changes reflected the chronic effect of i.p. injections and/or treatment withdrawal rather than the actual effect of CEF, we continued to treat the animals with daily saline i.p. injections until the end of the experiment. In this way, the animals were exposed every day to the same manipulation for the entire length of the experiment, minimizing possible interferences of animal handling on the EEG analysis.
Although it has been well documented that CEF treatment can affect the expression of the major Glu transporter GLT-1 in rodents 
, we did not measure GLT-1 levels in treated animals because of our experimental design. Nonetheless, using the same treatment schedule, we previously showed that CEF-induced GLT-1 modifications persist for at least four days after the end of treatment, and return to baseline level eight days after the end of treatment 
. The delayed emergence of electrophysiological and behavioral changes relative to GLT-1 up-regulation, which can be documented at day 8 
, may indicate that all the synaptic modifications observed and their functional consequences 
require time to affect the dynamics of large neuronal populations and lead to behavioral changes.
In rodents, theta oscillations are easily observed in the hippocampus, but can also be detected in other cortical and subcortical brain structures 
. Historically, they have been associated to learning and memory and voluntary movements 
. Lesions at different level of the hippocampal formation correlate to sizeable reductions of the hippocampal theta rhythm and to impairments of memory tasks 
. Parallel augmentation of long-term synaptic potentiation and theta activity was described by Maren and collaborators in a contextual fear-conditioning task 
, and more recent electrophysiological studies showed an increase of theta rhythm associated with induction of long-term potentiation 
, thereby suggesting a close relationship between theta oscillations and long-term synaptic plasticity 
. Recently, Omrani and colleagues demonstrated that CEF-induced GLT-1 up-regulation at CA3 synapses impairs long-term plasticity, by preventing perisynaptic metabotropic receptors activation 
. It is worth noting that our results are in line with these findings, and it is therefore conceivable to hypothesize that changes in synaptic plasticity can interfere with the normal function of the hippocampus, which in turn can generate a lower degree of synchronization, visible as a reduction in EEG theta power. Omrani and colleagues also showed that baseline fEPSPs were not affected by CEF treatment, indicating that the glutamatergic function is altered by GLT-1 up-regulation only when synapses undergo long-term plasticity 
. It is therefore conceivable that AMPA receptors activation at the glutamatergic synapses, even in a condition of reduced synaptic plasticity, could ensure a normal EEG signal. This might explain why the observed decrease in theta activity was not associated with an overall decrease in EEG power over the entire range of frequencies.
It is worth noting that we also found modifications of NREM and REM sleep power spectrum at day 10. Specifically, the analysis of NREM power spectrum showed a significant reduction of a broad band ranging between 7 and 13 Hz, which includes the spindle frequency band, suggesting that other pattern generators beside the hippocampus might be affected by CEF treatment. Interestingly, in a previous study we found that the GLT-1 is up-regulated by CEF in several brain regions, including thalamus and cerebral cortex, brain regions crucial in spindles formation and propagation 
. REM sleep normally is characterized by a robust theta activity in rodents, and several lines of evidence indicate that modifications in waking theta activity can be reflected in REM theta activity 
. Along this line, we found a decrease in theta power during REM sleep resembling the one observed during waking, although less prominent. The combined theta reduction in waking and in REM sleep induced by CEF could be ascribable to an impairment of a common circuit promoting theta formation in the two behavioral states.
Rats treated with CEF showed an increase of motor activity occurring after the end of treatment, i.e., when EEG alterations were clearly evident. Reportedly, theta is the electrical sign of activity in a forebrain mechanism that is organizing higher voluntary motor acts 
. Evidence suggests that there are two distinct types of hippocampal theta rhythm in behaving animals: the first (4–7 Hz) appears when animals are immobile or during repetitive acts such as sniffing or whiskers movements 
; the second one is more directly linked to voluntary motor behaviors (i.e., walking, running, rearing etc) and it is characterized by higher frequency (7–10 Hz) 
. In physiological conditions, motion speed is associated to the frequency of theta rhythm, as indicated by studies reporting that the higher is the speed, the faster is frequency 
. Moreover, electrophysiological studies demonstrated that theta activity recorded while the animal was running is larger than during walking, thus suggesting that the amplitude of theta rhythm increases as the intensity of movement increases 
. On the contrary, our results show a reduction of EEG theta power when motor activity is increased. By assessing separately motor activity during active and quiet waking we found that only active waking (mostly characterized by exploratory behavior) was affected by CEF treatment, thereby revealing a dissociation between theta power and motor activity. Of note ketamine, another drug that affects glutamatergic transmission, when administered to mimic pathological states (i.e., schizophrenia) causes both an increase in motor activity and a decrease in EEG theta power 
. Thus, it appears that the reduction of theta power, at least in non-physiological, drug-induced conditions, can be associated with an increase in motor activity. In addition, drug-induced motor activity is often mediated by aberrant activation of basal ganglia and cerebral cortex 
. It is therefore possible that glutamatergic alterations also affecting those structures 
may contribute to the increase of motor activity we observed. In support to the possible co-occurrence of theta activity reduction and increased motor activity, we also asked whether the time courses of those alterations were correlated. We found indeed a significant correlation between the EEG changes and the motor modifications, supporting a link between them, although a causal relationship still remains to be established.
In addition, there is evidence that theta oscillations play a role in integrating sensorimotor information 
. It is well known that a simple measure to evaluate sensorimotor integration in animals and in humans is prepulse inhibition of the startle reflex 
, which reflects the ability of the brain to temporarily adapt to a strong sensory stimulus when a preceding weaker signal is given to warn the organism 
. PPI is a neurological phenomenon regulated by a complex interplay between several brain structures, including cortical and subcortical sites 
. Interestingly, we recently demonstrated that the PPI was impaired in rats treated with CEF and that this deficit was blocked by dihydrokainate, a selective GLT-1 inhibitor 
, thus suggesting that sensorimotor integration is affected by GLT-1 up-regulation.
Taken together, these data indicate that GLT-1 up-regulation, by modifying the efficacy of the glutamatergic transmission, impairs the activity of widespread neural circuits and that the reduction in EEG theta power could be its electrophysiological signature. In addition, the increased motor activity and PPI alterations previously described suggest that neural circuits involved in sensorimotor control are particularly sensitive to changes in the efficacy of glutamatergic transmission induced by GLT-1 up-regulation.