Download hippocampo–cerebellar theta band phase synchrony in rabbits

Neuroscience 165 (2010) 1538 –1545
HIPPOCAMPO– CEREBELLAR THETA BAND PHASE SYNCHRONY
IN RABBITS
J. WIKGREN,*1 M. S. NOKIA1 AND M. PENTTONEN
stored within the cerebellar deep nuclei and/or the cerebellar cortical area HVI (Thompson and Steinmetz, 2009;
De Zeeuw and Yeo, 2005), whereas various cerebral areas can modulate the learning process. For instance,
when the cognitive demands of the task increase, hippocampal contribution becomes critical (Beylin et al., 2001;
Moyer et al., 1990; Woodruff-Pak and Disterhoft, 2008).
Hippocampal function, especially in the form of thetaband oscillation (⬃6 Hz; for a review see, e.g., Bland, 1986
or Buzsáki, 2002), has been linked to a host of cognitive
processes, most notably to learning and memory (Buzsáki,
2005; Hasselmo, 2005; Kahana, 2006). The relative power
of the hippocampal theta activity (theta ratio) recorded
before learning correlates strongly with the learning rate
during subsequent eyeblink conditioning in rabbits, as
shown both using the simple delay paradigm and using the
trace paradigm, during which learning is hippocampally
mediated (Berry and Thompson, 1978; Berry and Seager,
2001; Griffin et al., 2004; Nokia et al., 2008, 2009; Nokia
and Wikgren, 2009; Seager et al., 2002). Moreover, blocking hippocampal theta oscillation with, for example, scopolamine injections before eyeblink conditioning retards
behavioral learning and virtually abolishes any learningrelated unit responses in the hippocampus (Salvatierra
and Berry, 1989).
Previous research begs the questions, how do the
hippocampus and the cerebellum interact, and what is the
special role of the theta oscillation? In order for the hippocampus to modulate memory trace formation in the
cerebellum we have to assume interaction between these
areas. One indication of interaction between two brain
areas is synchronized oscillation (e.g. Fries, 2005; Singer,
1999). By synchronizing and desynchronizing their activity,
brain structures can select which inputs arrive and are sent
during a time window of maximal effect, i.e. cause a
change in the desired direction in the excitability of the
receiving group of neurons (Fries, 2005). We recorded
LFPs simultaneously from the rabbit hippocampus, cerebellar cortex, and medial prefrontal cortex (mPFC) during
trace eyeblink conditioning and assessed the degree of
phase synchrony (PS) (Palva et al., 2005; Lachaux et al.,
1998) between these structures. Since synchronous oscillatory activity supposedly indicates co-operation, and because hippocampal theta oscillation has been associated
with the learning rate during cerebellum-dependent eyeblink conditioning, we expected to observe (1) theta oscillation also in the cerebellum and (2) PS between the
hippocampus and the cerebellum at the theta band. In
addition, we aimed to determine whether the degree of PS
changes as a function of training and/or learning and
Department of Psychology, University of Jyväskylä, P.O. Box 35, FIN
40014, Jyväskylä, Finland
Abstract—Hippocampal functioning, in the form of theta
band oscillation, has been shown to modulate and predict
cerebellar learning of which rabbit eyeblink conditioning is
perhaps the most well-known example. The contribution of
hippocampal neural activity to cerebellar learning is only
possible if there is a functional connection between the two
structures. Here, in the context of trace eyeblink conditioning, we show (1) that, in addition to the hippocampus, prominent theta oscillation also occurs in the cerebellum, and (2)
that cerebellar theta oscillation is synchronized with that in
the hippocampus. Further, the degree of phase synchrony
(PS) increased both as a response to the conditioning stimuli
and as a function of the relative power of hippocampal theta
oscillation. However, the degree of PS did not change as a
function of either training or learning nor did it predict learning rate as the hippocampal theta ratio did. Nevertheless,
theta band synchronization might reflect the formation of
transient neural assemblies between the hippocampus and
the cerebellum. These findings help us understand how hippocampal function can affect eyeblink conditioning, during
which the critical plasticity occurs in the cerebellum. Future
studies should examine cerebellar unit activity in relation to
hippocampal theta oscillations in order to discover the detailed mechanisms of theta-paced neural activity. © 2010
IBRO. Published by Elsevier Ltd. All rights reserved.
Keywords: hippocampus, cerebellum, theta, oscillation, eyeblink conditioning, phase synchrony.
Common to all biologically meaningful learning is the engagement of multiple distinct phases and sub-processes
that are governed by different, sometimes widely distributed, brain structures. Rabbit eyeblink conditioning (Gormezano et al., 1962), where a neutral conditioned stimulus
(CS, e.g. a tone) is repeatedly paired with a reflex-eliciting
unconditioned stimulus (US, e.g. a corneal airpuff) provides an example of a model system of learning which
could be modified to tap into different aspects of the learning process. In this paradigm, the primary memory trace of
the motor conditioned response (CR) is thought to be
1
J. Wikgren and M. Nokia both contributed equally.
*Corresponding author. Tel: ⫹358-40-511-9562; fax: ⫹358-14-2602841.
E-mail address: [email protected] (J. Wikgren).
Abbreviations: ANOVA, analysis of variance; CS, conditioned stimulus; CR, conditioned response; EMG, electromyogram; FFT, fast fourier transform; HPC, hippocampus; HVI, cerebellar cortical area HVI;
LFP, local-field potential; LTP/LTD, long-term potentiation/depression;
mPFC, medial prefrontal cortex; PS, phase synchrony; US, unconditioned stimulus.
0306-4522/10 $ - see front matter © 2010 IBRO. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.neuroscience.2009.11.044
1538
J. Wikgren et al. / Neuroscience 165 (2010) 1538 –1545
whether, like the hippocampal theta ratio does, it predicts
learning rate.
EXPERIMENTAL PROCEDURES
Subjects and surgery
The subjects were 27 adult male New Zealand white rabbits
(Harlan, Netherlands, BV, Horst, Netherlands) aged ⬃4 months
and weighing ⬃2.7 kg at the time of surgery. The rabbits were
housed in individual metal cages on the premises of the animal
research unit of the University of Jyväskylä. Food and water were
freely available, and room temperature and humidity were controlled. All procedures were conducted during the light portion of
the 12/12 hour light/dark cycle and implemented in accordance
with the European Communities Council Directive (86/609/EEC,
http://eur-lex.europa.eu/LexUriServ/LexUriServ.do? uri⫽CELEX:
31986L0609:EN:HTML) on the care and use of animals for research purposes. Every effort to minimize the number of animals
and suffering caused to them was undertaken.
For a detailed description of the surgery see Nokia et al.
(2009). Two or three monopolar recording electrodes were chronically implanted into the right hippocampus 5 mm posterior and
4 – 6 mm lateral to the bregma and one electrode into the right
cerebellar cortex lobule HVI 1 mm anterior and 5 mm lateral to the
lambda. In addition, two electrodes were implanted into the left
mPFC 4 –5 mm anterior and 0.8 mm lateral to the bregma. The
electrodes were attached to a pin connector and the whole construction cemented in place with dental acrylic. At least 1 week
was allowed for postsurgical recovery.
Stimuli and procedure
Prior to the experiments, the rabbits were placed (for approximately 20 min) in a Plexiglas restraining box located in a ventilated, electrically insulated, and sound-attenuated conditioning
chamber to familiarize them with the experimental situation, and to
ensure the functioning of the implanted electrodes. Thereafter,
sessions were conducted once per day on consecutive days.
The CS was a 2 kHz, 85 dB, 200 ms tone and the US was a
100 ms corneal airpuff (0.35 bar source pressure, sound pressure
level 64 dB) delivered through a nozzle (inner diameter 2 mm)
placed approximately 1 cm away from the eye. A fan located
inside the conditioning chamber behind the rabbit created a
steady background noise of approximately 65 dB. E-Prime software (Psychology Software Tools Inc., Pittsburgh, PA, USA) was
used to control the presentation of stimuli.
The rabbits were randomly assigned to unpaired (n⫽10) and
paired (n⫽17) groups. Both received 10 daily sessions of either
trace eyeblink conditioning (paired) or explicitly unpaired training
(unpaired). The unpaired sessions consisted of 70 CS-alone and
70 US-alone trials presented in a random order with an intertrial
interval averaging out at 20 s (range, 15–25 s). The conditioning
sessions consisted of 80 trials: 60 conditioning, 10 CS-alone, and
10 US-alone trials were presented in a pseudorandom order with
an average intertrial interval of 40 s (range, 30 –50 s). During the
conditioning trials, CS onset preceded US onset by 700 ms, thus
creating a 500 ms trace period.
Recordings
Eyeblinks were measured using stainless steel wire hooks placed
around the upper and lower eyelids for the duration of the training
session. To acquire neural measures, a low-noise pre-amplifier
(MPA8I, Multi Channel Systems, Reutlingen, Germany) was directly attached to the electrode coupler anchored with dental
acrylic to the rabbit’s head. A flexible, insulated cable was used to
connect the animal to the amplifier (Axon Cyberamp 380, Molecular Devices Corporation, Union City, CA, USA). Both the neural
1539
data and the bipolar electromyogram (EMG) were recorded with
AxoScope (Molecular Devices Corporation) software and digitized
(Digidata 1322A, Molecular Devices Corporation) using a 10.26
kHz sampling rate. Before digitization, the LFPs were band-passfiltered between 0.1 and 4000 Hz, and the EMG was filtered
between 30 and 300 Hz.
Data analyses
Clampfit (Molecular Devices Corporation), MATLAB (The
MathWorks Inc., Natick, MA, USA) and SPSS (SPSS Inc., Chicago, IL, USA) were used for the data analyses.
Eyeblinks. The EMG signal was further high-pass filtered
over 100 Hz and Hilbert-transformed. Following this, an envelope curve following the peaks of the signal was calculated using
the real and imaginary parts of the Hilbert transformation. Baseline
EMG activity was calculated for each animal and session as the
mean of the maximum EMG amplitude during a 500 ms prestimulus period (MAXpre). In addition, the mean of the standard
deviation of the EMG activity during the 500 ms pre-stimulus
period (SDpre) was determined. Eyeblinks were defined as EMG
activity exceeding a threshold of [MAXpre⫹4⫻SDpre]. Blinks occurring during the 500-ms period following the offset of the tone
(trace period) were counted as CRs. The learning rate was quantified as the number of trials needed to reach the 5th CR and the
learning criterion.
Phase synchrony calculations. To obtain comparable data
from the control and experimental groups, only CS-alone and
US-alone trials were included in the analyses. Phase synchrony
(PS) between two signals was calculated as described by Palva et
al. (2005). First, the LFP signals were band-pass filtered (delta ⬃2
Hz, theta ⬃6 Hz, alpha ⬃12 Hz, beta ⬃20 Hz, and gamma ⬃60
Hz). Next, the filtered signals were transformed into a complex
form using the Hilbert transform. Following this, the amplitudes of
the signals were normalized to 1 (one) by dividing each data point
by its absolute value. Then, the phase difference of the two signals
in comparison was calculated by multiplying the first signal with
the complex conjugate of the second signal (each data point of
each trial). Finally, the PS was derived by averaging the phase
difference matrix over trials, taking the absolute value.
In order to standardize the PS values, 100 surrogate datasets
per each real PS were created by shuffling the real trials. The PS
was then determined for each surrogate dataset and the mean
and the 95th percentile of the surrogate PSs derived. The mean of
the surrogate PS values was subtracted from the real PS and the
outcome divided point-by-point by the difference between the 95th
percentile and the mean of the surrogate PSs ([realPS–surrPS]/
[surr95th percentile–surrPS]) yielding a standardized PS. Thus
standardized PS values exceeding 1 (one) represent statistically
significant synchrony at the level of P⬍.05. The group level statistical significance for the PS values was calculated using binomial statistics.
Theta ratio. The FFT was calculated from the 1-s pre-stimulus period data, with a resolution of 0.5 Hz. The theta ratio was
then calculated using (theta/[delta⫹theta]). The delta and theta
frequencies were used as the sole reference for theta, firstly,
because in absolute power, delta and theta frequencies are fairly
comparable and, secondly, because the absolute power of the
higher frequencies (8 Hz⬍) is considerably lower than that of
theta.
To compare PS during periods of a high versus low hippocampal theta ratio, the tone-alone (10) and airpuff-alone (10)
trials from each session (10) and each animal were sorted according to the hippocampal theta ratio calculated from the pre-stimulus
period. Then, the trials with the lowest and highest theta ratio were
selected to form sets of 10 trials per animal, pre-stimulus theta
ratio level and stimulus type.
1540
J. Wikgren et al. / Neuroscience 165 (2010) 1538 –1545
Fig. 1. Recording electrodes were placed in the dorsal hippocampus, the cerebellar cortex lobule HVI, and in the medial prefrontal cortex (mPFC).
All of the rabbits had at least one electrode in the hippocampal fissure or in the dentate gyrus/hilus and in the cerebellar cortex (A). In addition, 23
subjects had a recording electrode also in the mPFC (B). The open triangles refer to the subjects in the unpaired group and the filled circles to the
subjects in the paired group.
Statistical analyses. A paired samples t-test was used to
compare pre- and post-stimulus PS values. Analysis of variance
(ANOVA) for repeated measures was used for the assessment of
training and group effects. Pearson correlation coefficient was
used when examining the connection between PS and learning
rate.
was low (Fig. 3B, D), the cerebellar theta oscillation was
also weak.
Histology
To assess the degree of PS between the hippocampus
and the cerebellum and between the hippocampus and the
mPFC without contamination from the possible effects of
learning, the combined paired and unpaired group data
from only the first session was analyzed. Binomial test
(P⬍.01) showed significant PS only in the theta band (⬃6
Hz) oscillation occurring simultaneously in the hippocampus and in the cerebellar cortex (Fig. 4A, C). In response to
external stimulation, hippocampo– cerebellar theta band
PS was further increased (tone: t(26)⫽5.67, P⬍.001; airpuff: t(26)⫽5.47, P⬍.001; binomial test P⬍.001) (Fig. 4A).
The explicit lack of hippocampo– cerebellar PS was evident in all the other frequency bands (delta, alpha, beta,
and gamma) and during all three conditions (pre-stimulus,
After the experiments, the rabbits were anesthetized with an i.m.
injection of ketamine–xylazine cocktail and then overdosed with
an i.v. injection of pentobarbital (Mebunat vet, Orion-Yhtymä Oyj,
Espoo, Finland). A detailed description of the perfusion and histology procedure can be found in Nokia et al. (2009). The electrode-tip locations were determined from the stained slides with
the help of a microscope and stereotaxic atlases (Bures et al.,
1967; Lavond and Steinmetz, 2003).
RESULTS
Histology
All 27 subjects had recording electrodes both in the hippocampus and in the cerebellar cortex lobule HVI (Fig. 1),
and 23 also had an electrode in the mPFC (Fig. 1B).
Phase synchrony is specific to
hippocampo– cerebellar theta band oscillation
Behavioral results
The percentage of CRs (Fig. 2) increased across paired
treatment but remained stable across unpaired treatment
(block: F [4,100]⫽17.71, P⬍.001; group: F [1,25]⫽11.25,
P⬍.01; block⫻group: F [4,100]⫽4.91, P⬍.01; paired group,
block: F [4,64]⫽22.45, P⬍.001; unpaired group, block: F
[4,36]⫽3.27, ns).
Theta (⬃6 Hz) oscillation in the rabbit cerebellum
A clear theta band oscillation visible to the naked eye and
verified by FFT was found in the cerebellar cortex (lobule
HVI, Fig. 3). FFT showed that when the hippocampal theta
ratio was high (Fig. 3A, C), the cerebellar theta oscillation
was also strong, and when the hippocampal theta ratio
Fig. 2. Conditioned responding increased in the paired but not in the
unpaired group.
J. Wikgren et al. / Neuroscience 165 (2010) 1538 –1545
1541
Fig. 3. The power of the theta-band activity was high/low in the cerebellum (HVI) when it was high/low in the hippocampus (HPC). The results of fast
fourier transform (FFT, upper panels) run on hippocampal and cerebellar LFPs (lower panels, representative examples) during periods of high relative
power of hippocampal theta activity (theta/[delta⫹theta]) are shown in A (C) and during low relative power of hippocampal theta activity in B (D).
tone, and airpuff), as shown by binomial test (P⬍.001)
(Fig. 4A).
No statistically significant PS was found between the
hippocampus and the mPFC (Fig. 4B). In fact, the binomial
test showed a significant lack of PS in all bands (P⬍.05/
.001).
Hippocampo– cerebellar theta band PS remains
statistically significant across training, but
decreases slightly
Significant main effects of training block (5) and condition
(3) were found when the hippocampo– cerebellar PS during the pre-stimulus period and in response to the tone and
airpuff across training was examined in both the unpaired
group (block: F [4,108]⫽7.85, P⬍.001; condition: F [1,27]⫽
4.36, P⬍.05) and paired group (block: F [4,192]⫽7.51,
P⬍.001; condition: F [1,48]⫽16.24, P⬍.001) (Fig. 5). In
both groups, hippocampo– cerebellar PS was higher in
response to stimuli compared to the pre-stimulus period,
but decreased linearly across the training. The effect of
training type (unpaired vs. paired) and learning (fast vs.
slow learners) was also tested, but no connections were
found. In addition, no significant correlations were found
between hippocampo– cerebellar PS and learning rate (tri-
als to 5th CR and trials to criterion) in the paired group
(n⫽17), although a clear predictive connection was found
between the hippocampal theta ratio recorded during the
first session and the number of trials needed to reach the
5th CR (pre-CS theta ratio⫻5th CR: r⫽⫺.48, P⫽.05 and
post-CS theta ratio⫻5th CR: r⫽⫺.63, P⬍.01).
Based on the lack of significant group differences, data
from the paired (n⫽17) and unpaired (n⫽10) groups were
again combined (n⫽27) for further analyses.
Phase synchrony is stronger during episodes of a
high hippocampal theta ratio but increases in
response to external stimuli irrespective of the
on-going theta state
When the trials were divided into two classes based on the
degree of the hippocampal theta ratio (high vs. low) during
the absence of stimulation (pre-stimulus period), the PS
between the hippocampus and cerebellar cortex was statistically significant only when the hippocampal theta ratio
was high (binomial test P⬍.001) (Fig. 4D, E). However,
regardless of whether the hippocampal theta ratio was
high or low (i.e. the PS during the pre-stimulus period was
significant or not), the presentation of the tone or the airpuff
always elicited significant theta band PS between the hip-
1542
J. Wikgren et al. / Neuroscience 165 (2010) 1538 –1545
Fig. 4. Statistically significant phase synchrony (PS) was found only between the hippocampus and the cerebellar cortex and only in the theta-band
(⬃6 Hz). Binomial statistics showed significant PS only in the theta oscillation occurring simultaneously in the hippocampus and in the cerebellar cortex
(A). In response to external stimulation, this hippocampo– cerebellar theta-band PS was increased. No statistically significant PS was found between
the hippocampus and the mPFC (B). The hippocampo– cerebellar theta-band PS increased in response to both stimuli (tone and airpuff, A and C),
irrespective of the relative power of the hippocampal theta oscillation (low vs. high theta) immediately preceding stimulus onset (D and E). The PS
calculated from the real data are shown in the uppermost panel of each subplot, the PS calculated from the surrogate data obtained by shuffling real
data are shown in the middle panel, and the standardized PS is shown in the lowest panel. The vertical lines indicate the onset of the stimulus. The
asterisk reflects statistical significance of paired samples t-tests: *** P⬍.001. The horizontal dotted line shows the threshold value of 1 (one)
corresponding to statistical significance at the level of P⬍.05.
pocampus and the cerebellar cortex (binomial test P⬍
.001) (Fig. 4D, E).
DISCUSSION
The two main findings of this study were (1) a prominent
theta band oscillation in the cerebellar cortex of the rabbit
and (2) an increased and high degree of PS between the
theta band oscillation in the cerebellum and that in the
hippocampus during the presentation of external stimuli.
Hippocampo– cerebellar PS was especially strong when
the hippocampal theta ratio was high, but presenting the
animal with external stimuli increased the theta band PS
regardless of the ongoing theta ratio immediately preceding the stimulus presentation. However, we did not observe any changes in the hippocampo– cerebellar theta
band PS specific to paired training or learning, nor did we
detect any correlation between the level of spontaneous
hippocampo– cerebellar theta-band PS and learning rate.
To our knowledge, this is the first report on theta oscillation (⬃6 Hz) in the rabbit cerebellum. In rats, D’Angelo
et al. (2001) have shown that granule cells in the cerebellar
cortex exhibit bursting and resonance at the theta frequency range (3–12 Hz). Such rhythmicity might play an
important role in cerebellum-mediated learning. Indeed,
they have also shown that long-term potentiation/depression (LTP/LTD) is present in the mossy fiber-granule cell
synapses in the cerebellum when high-frequency stimulation of the former is paired with membrane depolarization
in the latter (D’Angelo et al., 1999; Armano et al., 2000;
Mapelli and D’Angelo, 2007). Whereas the relative magni-
J. Wikgren et al. / Neuroscience 165 (2010) 1538 –1545
1543
Fig. 5. The overall level of hippocampo– cerebellar PS decreased across training, but remained higher in response to external stimuli compared to
the pre-stimulus period at all times in both the unpaired (A) and paired (B) group.
tude of the hippocampal theta band oscillation affects the
behavioral learning rate and learning-related unit activity
during eyeblink conditioning (see, for example, Nokia et
al., 2008), the phase of hippocampal theta oscillation affects synaptic plasticity when LTP procedures are used
(Huerta and Lisman, 1995; Hölscher et al., 1997). Stimulating the hippocampus via the Schaffer collaterals during
the positive phase of hippocampal CA1 theta oscillation
produced LTP, whereas stimulation given during the negative phase reversed the previously established LTP
(Huerta and Lisman, 1995; Hölscher et al., 1997). In sum,
the responsiveness and plasticity of hippocampal output,
as well as that of the cerebellar input neurons, is synchronized by theta oscillation, now also shown to exist in the
cerebellar LFPs.
Not only did our present study show a theta oscillation
in the cerebellar LFPs but it also demonstrated that the
theta oscillation in the cerebellar cortex is in concert with
that in the hippocampus, as indicated by the significant PS.
Synchronous activity is thought to reflect the formation of
transient functional assemblies by anatomically distant
brain structures (Fries, 2005). Related to the role of the
hippocampal theta oscillation in learning and memory, in
healthy humans undergoing tasks requiring working memory, coherent activity at the theta-band between different
cortical sites has been observed by recording electroencephalogram (EEG; for a review, see Sauseng and Klimesch, 2008). In addition, in an early experiment by Livanov
(1977), visual stimulation caused limb movements at a
probability increasing along with the degree of coherence
between EEG signals measured above the visual and the
motor cortex in rabbits. Similarly, Schoffelen et al. (2005)
showed that subjects’ readiness to respond in a simple
reaction-time task closely correlated with the degree of
gamma-band coherence between neurons in the motor
cortex and spinal cord. Given that neural plasticity in both
the output area of the hippocampus (Huerta and Lisman,
1995; Hölscher et al., 1997) and in the input area of the
cerebellar cortex (D’Angelo et al., 1999; Armano et al.,
2000; Mapelli and D’Angelo, 2007) are related to the phase
of the theta cycle, and that theta oscillations in both areas
occur in a fixed temporal relation to each other, it is likely
that hippocampo– cerebellar theta band PS is relevant to
neural communication between these structures. It is also
to be noted that presenting a stimulus to the animal actually decreased the amount of theta band synchrony for a
brief time period. This observation conforms to the theory
of large-scale integration of neural activity (Rodriguez et
al., 1999) which states that transiently formed cognitive
states are separated by brief moments of desynchronization, i.e. active uncoupling of neural activity.
The hippocampo– cerebellar theta band PS observed
in our study suggests the formation of transient neural
assemblies between the structures, but, in the absence of
accompanying single-unit analyses, unfortunately is not
able to demonstrate it. However, evidence from other lines
of research indicates that hippocampal theta oscillation
might be the pacemaker for neuronal activity in learningrelevant regions. For instance, neuronal activity in prefrontal regions becomes synchronized with hippocampal theta
in the early phase of appetitively motivated trace conditioning (Paz et al., 2008). Similarly, Berke et al. (2004) found
coherent theta oscillations in the hippocampal CA1 and the
ventral/medial striatum, and showed that the striatal neurons tended to fire more frequently during the descending
phase of hippocampal theta oscillation, thus manifesting a
link between hippocampal theta and motor areas relevant
in consummatory behavior. Furthermore, Sirota et al.
(2008; see also Sirota and Buzsáki (2005) recently suggested that oscillatory synchrony might be the mechanism
through which the information flow between the hippocampus and the neocortex is controlled. There is also evidence
in anesthetized animals that cells in the neocortex are
synchronized with cells in the cerebellum, albeit not at
the theta band (Ros et al., 2009). Thus, further studies are
needed to investigate the possible connections between
the hippocampo– cerebellar theta oscillations and unit-activity in the cerebellar cortex/deep nuclei.
Much to our surprise, we found no connection between
hippocampo-cerebellar theta band PS and learning. This is
at odds with previous studies, which have shown a clear
predictive connection between the relative magnitude of
hippocampal theta oscillation and learning rate, as well as
changes across training/learning during trace eyeblink
1544
J. Wikgren et al. / Neuroscience 165 (2010) 1538 –1545
conditioning (Griffin et al., 2004; Nokia et al., 2009). This
difference in results could be explained by the difference in
the nature of the analyses used: in the present study the
phase of theta oscillation was examined whereas in previous studies analyses were targeted at the relative amplitude of theta oscillation. It is possible that hippocampo–
cerebellar theta band PS is a fundamental and rather
stable mediator of communication between the two structures, and thus does not show changes across training/
learning. In other words, the pattern of activation leading to
synchronized activity and mediating efficient communication remains stable, whereas learning-related changes occur, for instance in the number of cerebellar neurons firing
in relation to the theta band oscillation. It could well be that
cerebellar unit activity either increases during periods of
strong hippocampal theta oscillation, or becomes more
sharply synchronized with it. Further investigation of this
matter is needed.
Another explanation for the lack of learning-related
changes in PS between the cerebellar cortex and hippocampus in the present study could be that the cerebellar
cortex actually does not critically contribute to this form of
learning. For instance, it has been shown that impaired
cerebellar cortical function has a detrimental effect specifically on delay but not trace eyeblink conditioning in mice
(Kishimoto et al., 2001; Woodruff-Pak et al., 2006). As
hippocampal function modulates learning also during delay
conditioning, replicating the present study but using the
delay conditioning paradigm instead might, therefore, result in changes in PS as a function of training or learning.
In relation to more technical issues, it might be argued
that the hippocampo– cerebellar theta band PS observed
in this study resulted from volume conduction. However,
PS was seen only in the theta band, and only between the
hippocampus and the cerebellum, and not between the
hippocampus and the mPFC (which resides spatially
closer to the hippocampus), making it unlikely that volume
conduction could explain the results. It is more probable,
since the rhythm of hippocampal theta oscillation is in part
generated in the medial septum-diagonal band of Broca
(see Buzsáki, 2002), that it also plays a part in pacing
cerebellar theta oscillation. In the knowledge that both the
hippocampus and the cerebellum are connected with the
neocortex, it might be suggested that it provides a link
between them. Whether this is the case and, if so, through
which pathways, remains, however, yet to be determined.
CONCLUSION
In conclusion, the theta band PS found here between the
hippocampus and the cerebellum might reflect a mechanism by which these two structures can interact efficiently.
Via theta band PS the hippocampus might transiently form
neural assemblies with its target area in the cerebellum,
possibly enhancing synaptic modification and thus learning. In future studies it would be of importance to investigate whether firing in the cerebellar neurons is also
synchronized with the phase of hippocampal theta oscillation. Showing synchrony between cerebellar plastic-
ity and hippocampal theta oscillation would provide
strong evidence for the functional significance of cerebellar theta oscillation.
Acknowledgments—Supported by grants from the Academy of
Finland to JW (114258 and 130013) and from the Emil Aaltonen
foundation to MSN (Young researcher’s grant). The authors wish
to thank Matias Palva for advice on phase synchrony calculation
and Michael Dutton and Michael Freeman for language checking.
REFERENCES
Armano S, Rossi P, Taglietti V, D’Angelo E (2000) Long-term potentiation of intrinsic excitability at the mossy fiber— granule cell synapse of rat cerebellum. J Neurosci 20:5208 –5218.
Berke JD, Okatan M, Skurski J, Eichenbaum HB (2004) Oscillatory
entrainment of striatal neurons in freely moving rats. Neuron 43:
883– 896.
Berry SD, Seager MA (2001) Hippocampal theta oscillations and
classical conditioning. Neurobiol Learn Mem 76:298 –313.
Berry SD, Thompson RF (1978) Prediction of learning rate from the
hippocampal electroencephalogram. Science 200:1298 –1300.
Beylin AV, Gandhi CC, Wood GE, Talk AC, Matzel LD, Shors TJ
(2001) The role of the hippocampus in trace conditioning: temporal
discontinuity or task difficulty? Neurobiol Learn Mem 76:447– 461.
Bland BH (1986) The physiology and pharmacology of hippocampal
formation theta rhythms. Prog Neurobiol 26:1–54.
Bures J, Petran M, Zachar J (1967) Electrophysiological methods in
biological research. New York: Academic Press.
Buzsáki G (2002) Theta oscillations in the hippocampus. Neuron 33:
325–340.
Buzsáki G (2005) Theta rhythm of navigation: link between path integration and landmark navigation, episodic and semantic memory.
Hippocampus 15:827– 840.
D’Angelo E, Rossi P, Armano S, Taglietti V (1999) Evidence for NMDA
and mGlu receptor-dependent long-term potentiation of mossy
fiber— granule cell transmission in rat cerebellum. J Neurophysiol
81:277–287.
D’Angelo E, Nieus T, Maffei A, Armano S, Rossi P, Taglietti V, Fontana
A, Naldi G (2001) Theta-frequency bursting and resonance in
cerebellar granule cells: experimental evidence and modeling of a
slow K⫹-dependent mechanism. J Neurosci 21:759 –770.
De Zeeuw CI, Yeo CH (2005) Time and tide in cerebellar memory trace
formation. Curr Opin Neurobiol 15:667– 674.
Fries P (2005) A mechanism for cognitive dynamics: neuronal communication through neuronal coherence. Trends Cogn Sci 9:474 –
480.
Gormezano I, Schneiderman N, Deaux E, Fuentes I (1962) Nictitating
membrane: classical conditioning and extinction in the albino rabbit. Science 138:33–34.
Griffin AL, Asaka Y, Darling RD, Berry SD (2004) Theta-contingent trial
presentation accelerates learning rate and enhances hippocampal
plasticity during trace eyeblink conditioning. Behav Neurosci 118:
403– 411.
Hasselmo ME (2005) What is the function of hippocampal theta
rhythm?—linking behavioral data to phasic properties of field potential and unit recording data. Hippocampus 15:936 –949.
Hölscher C, Anwyl R, Rowan MJ (1997) Stimulation on the positive
phase of the hippocampal theta rhythm induces long-term potentiation that can be depotentiated by stimulation on the negative
phase in area CA1 in vivo. J Neurosci 17:6470 – 6477.
Huerta PT, Lisman JE (1995) Bidirectional synaptic plasticity induced
by a single burst during cholinergic theta oscillation in CA1 in vitro.
Neuron 15:1053–1063.
Kahana MJ (2006) The cognitive correlates of human brain oscillations. J Neurosci 26:1669 –1672.
Kishimoto Y, Hirono M, Sugiyama T, Kawahara S, Nakao K, Kishio M,
Katsuki M, Yoshioka T, Kirino Y (2001) Impaired delay but normal
J. Wikgren et al. / Neuroscience 165 (2010) 1538 –1545
trace eyeblink conditioning in PLCbeta4 mutant mice. Neuroreport
12:2919 –2922.
Lachaux JP, Rodriguez E, Martinerie J, Varela FJ (1998) Measuring
phase synchrony in brain signals. Hum Brain Mapp 8:194 –208.
Lavond DG, Steinmetz JE (2003) Handbook of classical conditioning.
London: Springer.
Livanov MN (1977) Spatial organization of cerebral processes. New
York: John Wiley & Sons.
Mapelli J, D’Angelo E (2007) The spatial organization of long-term
synaptic plasticity at the input stage of the cerebellum. J Neurosci
27:1285–1296.
Moyer JR Jr, Deyo RA, Disterhoft JF (1990) Hippocampectomy disrupts trace eye-blink conditioning in rabbits. Behav Neurosci 104:
243–252.
Nokia MS, Penttonen M, Korhonen T, Wikgren J (2008) Hippocampal
theta (3– 8 Hz) activity during classical eyeblink conditioning in
rabbits. Neurobiol Learn Mem 90:62–70.
Nokia MS, Penttonen M, Korhonen T, Wikgren J (2009) Hippocampal
theta band activity and trace eyeblink conditioning in rabbits. Behav Neurosci 123:631– 640.
Nokia MS, Wikgren J (in press) Hippocampal theta activity is selectively associated with contingency detection but not discrimination in rabbit discrimination-reversal eyeblink conditioning.
Hippocampus.
Palva JM, Palva S, Kaila K (2005) Phase synchrony among neuronal
oscillations in the human cortex. J Neurosci 25:3962–3972.
Paz R, Bauer EP, Paré D (2008) Theta synchronizes the activity of
medial prefrontal neurons during learning. Learn Mem 15:524 –
531.
Rodriguez E, George N, Lachaux JP, Martinerie J, Renault B, Varela
FJ (1999) Perception’s shadow: long-distance synchronization of
human brain activity. Nature 397:430 – 433.
1545
Ros H, Sachdev RN, Yu Y, Sestan N, MacCormick DA (2009) Neocortical networks entrain neuronal circuits in cerebellar cortex.
J Neurosci 29:10309 –10320.
Salvatierra AT, Berry SD (1989) Scopolamine disruption of septohippocampal activity and classical conditioning. Behav Neurosci
103:715–721.
Sauseng P, Klimesch W (2008) What does phase information of
oscillatory brain activity tell us about cognitive processes? Neurosci Biobehav Rev 32:1001–1013.
Schoffelen JM, Oostenveld R, Fries P (2005) Neuronal coherence as
a mechanism of effective corticospinal interaction. Science 308:
111–113.
Seager MA, Johnson LD, Chabot ES, Asaka Y, Berry SD (2002)
Oscillatory brain states and learning: impact of hippocampal thetacontingent training. Proc Natl Acad Sci U S A 99:1616 –1620.
Singer W (1999) Neuronal synchrony: a versatile code for the definition of relations? Neuron 24:49 – 65.
Sirota A, Buzsáki G (2005) Interaction between neocortical and hippocampal networks via slow oscillations. Thalamus Relat Syst 3:
245–259.
Sirota A, Montgomery S, Fujisawa S, Isomura Y, Zugaro M, Buzsáki G
(2008) Entrainment of neocortical neurons and gamma oscillations
by the hippocampal theta rhythm. Neuron 60:683– 697.
Thompson RF, Steinmetz JE (2009) The role of the cerebellum in
classical conditioning of discrete behavioral responses. Neuroscience 162:732–755.
Woodruff-Pak DS, Disterhoft JF (2008) Where is the trace in trace
conditioning? Trends Neurosci 31:105–112.
Woodruff-Pak DS, Green JT, Levin ST, Meisler MH (2006) Inactivation
of sodium channel Scn8A (Na-sub(v)1.6) in Purkinje neurons impairs learning in Morris water maze and delay but not trace eyeblink classical conditioning. Behav Neurosci 120:229 –240.
(Accepted 18 November 2009)
(Available online 27 November 2009)