Sleep EEG in mice that are deficient in the potassium channel subunit K.v.3.2

Brain Research 947 (2002) 204–211 www.elsevier.com / locate / bres

Research report

Sleep EEG in mice that are deficient in the potassium channel subunit K.v.3.2 Vladyslav V. Vyazovskiy a , Tom Deboer a , Bernardo Rudy b , David Lau b , ´ a , Irene Tobler a , * Alexander A. Borbely a

b

¨ ¨ , Winterthurerstrasse 190, CH-8057 Zurich , Switzerland Institute of Pharmacology and Toxicology, University of Zurich Departments of Physiology and Neuroscience, and Biochemistry, New York University School of Medicine, New York, NY 10016, USA Accepted 15 April 2002

Abstract Voltage-gated potassium channels containing the K.v.3.2 subunit are expressed in specific neuronal populations such as thalamocortical neurons and fast spiking GABAergic interneurons of the neocortex and hippocampus. These K 1 -channels play a major role in the regulation of firing properties in these neurons. We investigated whether the K.v.3.2 subunit contributes to the generation of the sleep electroencephalogram (EEG). The EEG of a frontal and occipital derivation of K.v.3.2-deficient mice and littermate controls was recorded during a 24-h baseline, 6-h sleep deprivation (SD) and subsequent 18-h recovery to assess also the effects of the K.v.3.2 subunit deficiency under physiological sleep pressure. The K.v.3.2-deficient mice had lower EEG power density in the frequencies between 3.25 and 6 Hz in nonREM (NREM) sleep and 3.25–5 Hz in REM sleep. These differences were more prominent in the frontal derivation than in the occipital derivation. The waking EEG spectrum was not affected by the deletion. In both genotypes SD induced a prominent increase in slow-wave activity in NREM sleep (mean EEG power density between 0.75 and 4.0 Hz), and a concomitant decrease in sleep fragmentation. The effects of SD did not differ significantly between the genotypes. The results indicate that K.v.3.2 channels may be involved in the generation of EEG oscillations in the high delta and low theta range in sleep. They support the notion that GABA-mediated synchronization of cortical activity contributes to the electroencephalogram.  2002 Elsevier Science B.V. All rights reserved. Theme: Neural basis of behavior Topic: Biological rhythms and sleep Keywords: Knock-out mouse; Potassium channel; K.v.3.2 subunit; Sleep EEG; Spectral analysis; GABAergic interneuron

1. Introduction Extensive studies have disclosed the mechanisms underlying the generation of basic electroencephalogram (EEG) rhythms at the molecular, cellular and network levels [4,20,29,30]. Extracellular and intracellular recordings have shown that single neurons may have intrinsic properties which allow them to generate rhythmic activity in

*Corresponding author. Tel.: 141-1-635-5957; fax: 141-1-635-5707. E-mail address: [email protected] (I. Tobler).

several typical EEG frequency bands, i.e. in delta [1], theta [7,22] and spindle frequencies [20]. The intrinsic properties of these neurons and synaptic interactions within large neuronal networks contribute to the genesis of EEG waves [29,30,34]. Activity patterns of thalamic and neocortical neurons are key elements in the generation of spindles and slow waves in the sleep EEG [1,30]. Moreover, neurons in the medial septal area fire in the theta frequency range and drive both the phase and frequency of hippocampal theta oscillations during active wakefulness and rapid eye movement (REM) sleep [4,32]. The discharge pattern of individual neurons is a result of

0006-8993 / 02 / $ – see front matter  2002 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 02 )02925-6

V.V. Vyazovskiy et al. / Brain Research 947 (2002) 204 – 211

numerous molecular events occurring at the cellular level. Thus, distinct features of the spiking activity of a neuron are brought about by specific properties of the cell, such as potassium channel-mediated currents ([12,21] reviewed in [26]). Inhibitory neurons play a role in coordinating and synchronizing cortical activity [14,16]. Functional networks of interconnected fast spiking GABAergic interneurons synchronize neuronal spiking activity in the neocortex and in the hippocampus [3]. Interneurons synchronize the firing of hippocampal pyramidal cells at frequencies between 4 and 7 Hz (low-theta range) via GABAA -receptor mediated mechanisms [9]. Recordings from interneurons in hippocampal slices of the rat showed that the cells display intrinsic oscillations of membrane potentials in the theta-frequency range that result from an interplay between voltage-dependent Na 1 and K 1 conductances [6]. The amplitude and frequency of these thetafrequency oscillations were strongly reduced by K 1 -channel blockers [6]. The distinctive firing phenotype of the ‘fast spiking’ GABAergic interneurons is determined in part by their K 1 channels belonging to the Kv3 subfamily ([8] reviewed in [26]). The K.v.3.2 K 1 -channel subunit belongs to the Kv3 subfamily of proteins that form voltage-gated potassium channels. K.v.3.2 mRNA transcripts are expressed most abundantly in the relay nuclei of the thalamus and in the GABAergic interneurons of the neocortex and hippocampus. Moderate amounts of these transcripts are found in the medial septum, locus coeruleus and basal nuclei [17,23– 25,33]. In mice in which the deletion of the K.v.3.2 gene led to a deficiency of this channel subunit (K.v.3.2 2 / 2), whole-cell electrophysiological recordings showed an impaired ability to fire high frequency spikes in fast spiking interneurons of cortical layers 5–6, in which K.v.3.2 subunits are normally abundant [21]. These mice also exhibited increased cortical excitability related to the impaired fast spiking in GABA-interneurons [21]. The behavior of mice deficient in the K.v.3.2 channel subunit was normal, with the exception of an open field test where these mice showed a small but significantly lower centerto-total distance ratio, which is an indicator of anxiety in the open field [21]. Given the ubiquitous distribution of the fast spiking interneurons containing K.v.3.2 channel subunits in the neocortex [8] and hippocampus [23,25], as well as the particular importance of the inhibitory network for synchronizing the activity of pyramidal cells [5], we hypothesized that the activity of K 1 channels containing K.v.3.2 protein may modulate the cortical EEG. To test this hypothesis we recorded EEG from an occipital and frontal derivation in K.v.3.2-deficient mice and their littermate controls during a 24-h baseline. A 6-h sleep deprivation (SD) was performed to investigate the role of the K.v.3.2 channel subunits under physiological sleep pressure during subsequent recovery.

205

2. Materials and methods

2.1. Animals The experiments were approved by the local governmental commission for animal research. Adult male K.v.3.2 2 / 2 mice (6th generation backcrossed with C57BL6 / Taconic; [21]) and their wild-type littermates (K.v.3.2 1 / 1) were used. The age at recording onset was 23.7960.04 (S.E.M.) and 24.0560.2 weeks for K.v.3.2 2 / 2 (n58) and K.v.3.2 1 / 1 (n56), respectively (n.s., two-tailed t-test). The animals were kept individually in Macrolon cages (36320335 cm) with food and water available ad libitum, and maintained in a 12-h light:12-h dark cycle (light from 10.00–22.00 h; 7 Watt Osram Dulux EL energy saving lamp, approximately 30 lux). Mean ambient temperature was 23.960.1 8C. The cages were equipped with PVC running wheels (15.0 cm diameter, 170 g).

2.2. Surgery The weight of the mice at surgery was 24.760.5 g and 26.360.7 g (K.v.3.2 2 / 2 and K.v.3.2 1 / 1 mice respectively; difference n.s.). Under deep pentobarbital anesthesia (Nembutal sodium, 80 mg / kg i.p., volume approximately 0.5 ml) the mice were implanted with gold-plated miniature screws (0.9-mm-diameter) which served as EEG electrodes. Screws were placed epidurally over the right occipital cortex (2–3 mm lateral to the midline and 2 mm posterior to the bregma), right frontal cortex (1 mm lateral to the midline and 1 mm anterior to the bregma). A reference electrode was placed over the cerebellum (1 mm posterior to the lambda, on midline). Two gold wires (diameter 0.2 mm) inserted into the neck muscles served to record the electromyogram (EMG). The electrodes were connected to stainless steel wires that were fixed to the skull with dental cement. At least 3 weeks were allowed for recovery.

2.3. Experimental protocol and data acquisition The two EEGs and the EMG were recorded continuously for 48 h. First, a 24-h baseline day was recorded beginning at light onset, followed by 6-h sleep deprivation (SD) starting at light onset and 18-h recovery. SD was performed by introducing objects (e.g., nesting material) into the cage, and later by tapping on the cage whenever the animal appeared drowsy and / or the EEG exhibited slow-waves [31]. Halfway through the SD the mice were provided with new cages, which induced additional stimulation and elicited exploratory behavior. To minimize stress, some material (soiled wood chippings) from the old cage was transferred to the new one. The mice were never disturbed during feeding and drinking. The bipolar EEG

206

V.V. Vyazovskiy et al. / Brain Research 947 (2002) 204 – 211

(occipital–cerebellum and frontal–cerebellum) and EMG signals were amplified (amplification factor approx. 2000), conditioned by analog filters (high-pass filter: 23 dB at 0.016 Hz; low-pass filter: 23 dB at 40 Hz, less than 235 dB at 128 Hz) sampled with 512 Hz, digitally filtered (EEG: low-pass FIR filter 25 Hz; EMG: band-pass FIR filter 20–50 Hz) and stored with a resolution of 128 Hz. EEG power spectra were computed for 4-s epochs by a fast fourier transform (FFT) routine. Adjacent 0.25-Hz bins were averaged into 0.5-Hz (0.25–5.0 Hz) and 1.0-Hz (5.25–40.0 Hz) bins, and those above 40 Hz were omitted. The EMG was full-wave rectified and integrated over 4-s epochs, and ambient temperature inside the cage was sampled at 4-s intervals. Running wheel activity was monitored as previously, based on counting wheel revolutions for 1-min intervals ([11]; Chronobiology Kit, Stanford Software Systems, Stanford, CA).

2.4. Vigilance states and analyses The three vigilance states non rapid eye movement (NREM) sleep, REM sleep and waking were scored for 4-s epochs as in a previous study [31]. Vigilance states were determined off-line by visual inspection of the occipital EEG and EMG records and the values of EEG power in the slow-wave range (0.75–4.0 Hz). Epochs containing EEG artifacts were excluded from spectral analyses of both EEG derivations (K.v.3.2 1 / 1, 13.8862.98 and K.v.3.2 2 / 2, 14.4462.88% of recording time). Most artifacts occurred during wakefulness. Vigilance states could be always determined. The number of brief awakenings (BA, brief episodes of waking ,16 s) per hour of sleep was computed to assess sleep continuity. The frequency of vigilance state episodes was computed as previously [10]. Frequency bands that differed significantly between

Fig. 1. EEG power density spectra in non rapid eye movement sleep (NREMS) and REMS for the 24-h baseline. Mean values (K.v.3.2 2 / 2 mice, n58; K.v.3.2 1 / 1 mice, n56) in mV 2 / Hz are plotted at the upper limit of each bin on a logarithmic scale, separately for the frontal (top panels) and occipital (bottom panels) derivations. Horizontal lines below the curves indicate frequency bins in which an unpaired t-test on log-transformed values was significant: frontal derivation: NREMS: 3.25–6.0 Hz; REMS: 3.25–5.0 Hz; occipital derivation: n.s.

V.V. Vyazovskiy et al. / Brain Research 947 (2002) 204 – 211

genotypes (3.25–6 Hz in NREM sleep and 3.25–5 Hz in REM sleep, Fig. 1) were selected for further analyses. To assess whether the changes in the spectrum were sleep specific the time-course of EEG power in the 3.25–5 and 3.25–6 Hz bands during waking was also computed. Differences in EEG power between genotypes were analyzed with one-, two-, or three-way ANOVA, factors ‘genotype’ and ‘2-h interval’. Contrasts were tested by post hoc two-tailed t-tests (for equal variances) if the ANOVA reached significance.

3. Results

3.1. Motor activity and vigilance states Running wheel activity (24-h 10-day mean, arbitrary units6S.E.M.: K.v.3.2 1 / 1, 567.4622.7; K.v.3.2 2 / 2, 536.9630.1; t-test, n.s.) and its 24-h time-course over a 10-day interval was similar for K.v.3.2 2 / 2 and K.v.3.2 1 / 1. The amount of sleep (Table 1), as well as its distribution did not differ between the genotypes during baseline. Nevertheless, sleep architecture was affected by the absence of the K.v.3.2 protein. The frequency of both NREM sleep and REM sleep episodes was lower in K.v.3.2-deficient mice than in wild-type mice (mean number of episodes per hour 6S.E.M., NREM sleep: K.v.3.2 2 / 2, 2.560.1; K.v.3.2 1 / 1, 3.160.1; P,0.01; REM sleep: K.v.3.2 2 / 2, 2.560.1; K.v.3.2 1 / 1, 3.160.2; P,0.05). Neither episode duration (not shown) nor the number of brief awakenings per hour of sleep (BA,

207

Table 1) differed between the genotypes. SD had no effect on the vigilance states in the K.v.3.2 2 / 2 mice and only minor but significant effects in the K.v.3.2 1 / 1 mice. Thus, in the latter NREM sleep and REM sleep were enhanced in the first 6-h interval after SD while the amount of waking was below baseline (Table 1). The comparison of the effects of SD between the genotypes was not significant. Similarly, the computation of vigilance states after SD for 2-h intervals showed no differences between the genotypes in the time-course or the amount of the vigilance states (three-way ANOVA, factors ‘genotype’, ‘day’ [BL vs. Rec] and ‘2-h interval’; n.s. for genotype and for interaction ‘genotype3day’). Sleep continuity was enhanced by SD in both genotypes. This was reflected by the significant decrease in the number of BA during the first 6 h of recovery (Table 1). Sleep deprivation was successful. Only a small amount of NREM sleep occurred during the 6-h SD, amounting to less than 9 min, and it did not differ between the genotypes (Table 1). No behavioral or EEG abnormalities were noticed in the K.v.3.2 2 / 2 mice prior to the experiment or during the SD procedure.

3.2. EEG power density The comparison of the EEG spectra revealed in NREM sleep a significantly lower EEG power density in the 3.25–6 Hz band of the frontal derivation in the K.v.3.2 2 / 2 mice compared to K.v.3.2 1 / 1 mice (Fig. 1). The difference in the occipital derivation was similar, but did not reach significance. In REM sleep, power density in the

Table 1 Vigilance states and brief awakenings in K.v.3.2-deficient mice (K.v.3.2 2 / 2) and wild-type mice (K.v.3.2 1 / 1) K.v.3.2 2 / 2

K.v.3.2 1 / 1

Waking

NREMS

REMS

BA

Waking

NREMS

REMS

BA

Baseline Light 0–6 h 7–12 h 12 h (0–12 h)

42.1 (3.1) 43.3 (4.6) 42.7 (2.2)

49.3 (2.8) 47.4 (4.0) 48.3 (1.9)

8.7 (0.4) 9.3 (0.7) 9.0 (0.3)

33.5 (3.4) 40.0 (5.0) 36.5 (4.0)

39.5 (4.0) 38.0 (2.0) 38.8 (2.0)

51.5 (3.2) 51.8 (1.7) 51.7 (1.8)

8.9 (0.8) 10.3 (0.6) 9.6 (0.4)

27.5 (3.2) 32.3 (3.1) 29.9 (3.0)

Dark 13–18 h 19–24 h 12 h (13–24 h)

97.0 (2.0) 80.7 (4.0) 88.8 (2.5)

2.6 (1.7) 17.2 (3.7) 10.0 (2.2)

0.4 (0.3) 2.0 (0.5) 1.2 (0.3)

19.9 (14.0) 36.1 (5.8) 36.8 (5.6)

93.3 (2.2) 70.8 (5.0) 82.1 (2.1)

6.0 (2.0) 25.9 (4.4) 16.0 (2.0)

0.6 (0.3) 3.3 (0.6) 1.9 (0.2)

26.6 (7.7) 31.2 (6.2) 31.0 (6.2)

Sleep deprivation 0–6 h

97.5 (0.7)

2.5 (0.7)

98.2 (1.0)

1.8 (1.0)

Recovery 7–12 h 13–18 h 19–24 h

40.0 (3.3) 97.6 (1.9) 76.6 (4.0)

50.1 (2.9) 2.1 (1.6) 20.4 (4.0)

0.0

9.8 (0.7) 0.3 (0.3) 3.0 (0.5)

0.0

26.9 (4.0)* 26.6 (13.0) 37.1 (6.6)

31.5 (2.3)** 90.0 (4.0) 68.9 (4.6)

57.3 (1.8)** 8.8 (3.4) 27.3 (4.0)

0.0

11.2 (0.6)** 1.2 (0.6) 3.8 (0.7)

0.0

24.8 (3.0)** 32.3 (10.9) 37.7 (5.1)

Mean values (S.E.M. in parentheses) of waking, NREM sleep (NREMS) and REM sleep (REMS) expressed as percentage of recording time for K.v.3.2 1 / 1 (n56) and K.v.3.2 2 / 2 (n58) for 6-h intervals, and the 12-h light and dark period during baseline, 6-h sleep deprivation (SD) and recovery. SD began at light onset. Brief awakenings (BA) are expressed as number of waking episodes ,16 s per hour of total sleep time. Effect of SD within a genotype: **P,0.001; *P,0.05, two-tailed paired t-test. No differences were found between genotypes (t-test, n.s.).

208

V.V. Vyazovskiy et al. / Brain Research 947 (2002) 204 – 211

Fig. 2. EEG power in the frontal derivation in the 3.25–6 Hz band for non rapid eye movement sleep (NREMS), and 3.25–5 Hz band for REMS and waking during the 24-h baseline, 6-h sleep deprivation (SD) and 18-h recovery for K.v.3.2 1 / 1 (n56) and K.v.3.2 2 / 2 (n58) mice. Curves connect 2 h mean values 6S.E.M. Only those intervals for which data was available for at least two individuals are included. Note the different scales for waking, NREMS and REMS. The effect of ‘genotype’ was significant for NREMS and REMS (two-way ANOVA, factor ‘genotype’, P,0.0001; interaction, n.s.) and the effect of ‘2 h interval’ was significant for NREMS and waking (factor ‘2 h interval’, P,0.0001; interaction, n.s.). Triangles indicate 2 h intervals that differed between genotypes (triangles: black, P,0.05; white, P,0.1; two-tailed unpaired t-test after significance in the ANOVA). There were no significant differences between the genotypes in the effect of SD.

3.25 to 5 Hz band in the frontal derivation was lower in K.v.3.2-deficient mice compared to K.v.3.2 1 / 1 mice (Fig. 1, right panel). Also this difference did not reach significance in the occipital derivation. However, the interaction ‘genotype’ and ‘derivation’ was significant in NREM sleep between 0.75 and 5 Hz and in REM sleep between 1.75 and 4.5 and 6.25–11 Hz (two-way ANOVA, P,0.05). The waking EEG did not differ significantly between the genotypes in either derivation (not shown). Higher frequencies, including all frequencies up to 40 Hz, were not affected by the channel subunit depletion in any vigilance state (not shown). Fig. 2 illustrates for the frontal derivation the timecourse of those two frequency bands which reached significance in Fig. 1. Both in NREM sleep and in REM sleep most 2-h intervals of the baseline light period were significantly lower in K.v.3.2 2 / 2 mice than K.v.3.2 1 / 1 mice. In waking, neither the 3.25–5 Hz band (Fig. 2) nor the 3.25–6 Hz band (not shown) differed between the genotypes (two-way ANOVA, factor ‘genotype’, n.s.). No genotype differences were found in the 2-h intervals of the occipital derivation (data not shown). After SD the variability of the mean values in the 3.25–5 Hz and 3.25–6 Hz bands was enhanced, and no significant differences between the genotypes were observed. In addition, the genotypes did not differ in the effect of SD on these bands (interaction ‘day [baseline, recovery]’ and ‘genotype’ was n.s.). To investigate whether the K.v.3.2 channel subunit may play a role at the transitions between vigilance states, the relative EEG power density in the 3.25–6.0 Hz band and in the spindle-activity band (10.25–15.0 Hz) was computed for 2 min before and after the transition from waking to NREM sleep. No differences were apparent between the genotypes in the 3.25–6.0 Hz band, whereas spindlefrequency activity was higher in K.v.3.2 2 / 2 mice during several 4-s epochs after the transition from waking to NREM sleep in the occipital derivation (Fig. 3). The computation of the delta band (0.75–4.0 Hz; SWA) in NREM sleep showed in both genotypes after SD the well-known marked increase in both derivations. The increase was most prominent in the initial 2-h recovery interval (% of corresponding 2-h interval in baseline: occipital: K.v.3.2 1 / 1, 134.362.6; K.v.3.2 2 / 2, 130.868.1; n.s.; frontal: K.v.3.2 1 / 1, 160.268.4; K.v.3.2 2 / 2, 178.6620.8; n.s.). In both genotypes the increase of SWA was greater in the frontal derivation than in the occipital derivation (first and second 2-h intervals of recovery, P,0.05). No significant differences were found in the effect of SD between K.v.3.2 1 / 1 and K.v.3.2 2 / 2 mice.

4. Discussion The results indicate that voltage-gated potassium chan-

V.V. Vyazovskiy et al. / Brain Research 947 (2002) 204 – 211

209

Fig. 3. Time-course of EEG power density in the 3.25–6.0 Hz band (upper panels) and the 10.25–15.0 Hz band (lower panels) for 2 min before and 2 min after transitions from waking to NREM sleep (NREMS) during baseline. The values are plotted relative to baseline, such that 100% represents total EEG power density in the corresponding frequency band during the 24-h baseline. Vertical lines indicate the transition from waking to NREM sleep. Triangles indicate 4-s epochs that differed between K.v.3.2 2 / 2 (open circles) and K.v.3.2 1 / 1 (black circles) mice (P,0.05; two-tailed unpaired t-test).

nels composed of the Kv3 subfamily may be involved in the generation of EEG oscillations in the high delta and low theta range during sleep. The EEG frequencies in sleep which were affected by the absence of the K.v.3.2 subunit (Fig. 1) appear to be similar to those involved in the GABAergic rhythmic synchronization of pyramidal cell activity, i.e. low theta frequencies (4–7 Hz) [9] and ‘slow theta frequency range’ (4–6 Hz) [27]. An intrinsic intracellular theta rhythm (3–6 Hz) could be recorded from CA1 cells of rat hippocampus under ketamine–xylazine anesthesia [22]. Deletion of the K.v.3.2 subunit may decrease the efficacy of GABAergic interneurons that express this subunit [8,23] resulting in reduced synchronization of neuronal activity during sleep as indicated by reduced power in the frequency range 3.25–5 or 3.25–6 Hz. This idea is consistent with a previous finding where increased seizure susceptibility in K.v.3.2 2 / 2 mice indicated a decreased performance of intracortical inhibitory circuitry [21]. Taken together, these results imply that mechanisms of EEG synchronization are impaired following disruption of the K.v.3.2 gene.

The GABAergic interneurons represent a large population of neocortical and hippocampal neurons [13,15]. It has been proposed that the ‘supernetwork’ of GABAergic interneurons may entrain large populations of pyramidal cells throughout the forebrain [5]. The GABA-mediated phasing of neuronal oscillations may represent a general mechanism for synchronization of cortical activity [9]. Synchronous oscillations imposed by the GABAergic network across the forebrain structures are thought to bring together anatomically distributed ensembles of neurons in order to establish functional relationships between them [5]. The coactivation of large neuronal populations leads to the synchronization of the timing of spikes and results in the genesis of macro-oscillations in the form of various EEG rhythms [29]. Thus, EEG power could be related to the level of synchronization in the neuronal activity in the region of recording [19]. The deletion of K.v.3.2 subunit did not affect higher EEG frequencies (up to 40 Hz), which is in contrast to a deletion of the K.v.3.1 subunit which resulted in an increase in gamma frequencies, i.e. 20–60 Hz [18].

210

V.V. Vyazovskiy et al. / Brain Research 947 (2002) 204 – 211

The decreased frequency of sleep episodes in the K.v.3.2 2 / 2 mice compared to K.v.3.2 1 / 1 mice, and the better consolidation of sleep may be a compensation for impaired EEG synchronization mechanisms in the null mice. The differences between the K.v.3.2 2 / 2 and K.v.3.2 1 / 1 mice did not include the waking spectrum. This might be related to the specific pattern of expression of the K.v.3.2 subunit in the brain. Thus, mRNA transcripts of the K.v.3.2 were found mainly in neuronal populations within the thalamus, basal forebrain and locus coeruleus [33], which are known to be involved in the regulation of vigilance states, as well as in the cortex [28]. The activity patterns of neurons in these regions differ dramatically between vigilance states and the normal function of K.v.3.2 channels may depend on the physiological state of these neurons. Only small genotype differences were observed at the transition from waking to sleep in the spindle-frequency band. This may be a consequence of the negligible expression of the K.v.3.2 subunit in the reticular thalamic nucleus [33], which is known to be critical in generating spindle-frequency activity [20]. Several other brain regions are important for spindle generation or propagation [30]. However, due to unusual properties of the Kv3 channels (high activation threshold, above 210 mV [21]) they may not contribute to those activities, which are known to occur at more hyperpolarized levels of membrane potential [30]. The more prominent differences in the spectra in the frontal EEG compared to the occipital EEG between the K.v.3.2 2 / 2 and K.v.3.2 1 / 1 mice may be related to regional specificity in the expression of the K.v.3.2 subunit [33]. However, the regional distribution of the K.v.3.2 subunit across neocortical areas is still unknown. A significantly larger number of GABAergic neurons per mm 3 of tissue was estimated in the adult rat in the occipital cortex compared to the frontal cortex [2]. Lower density of GABAergic neurons in the frontal area may result in decreased efficiency of inhibitory circuits and increased susceptibility to the deletion of the K.v.3.2 subunit. Sleep deprivation did not change the genotype differences observed during baseline. This implies that the activity of the channels plays a rather basic role in the genesis of EEG oscillations which is not exacerbated by increasing sleep pressure. The electrophysiological phenotype of the K.v.3.2 2 / 2 mice provides no evidence for a functional compensation during development as may occur when a gene is deleted [21]. In contrast, the behavior of these mice appeared normal except for minor differences [21]. The present analysis shows that the amount of sleep and sleep homeostatic mechanisms are not different, but that minor differences are present in the sleep EEG. Our results raise the question of the functionality of the redundancy in the potassium channel heterogeneity. In conclusion, our data suggest that the Kv3 potassium channels containing the K.v.3.2 subunit play a role in the

mechanisms generating EEG activity within the low theta frequency range.

Acknowledgements We thank Drs P. Achermann and J. Gottselig for comments on the manuscript. The study was supported by the Swiss National Science Foundation grant 3100053005.97, Human Frontier Science Program RG 81 / 96 and NIH grant NS30989.

References [1] F. Amzica, M. Steriade, Electrophysiological correlates of sleep delta waves, Electroencephalogr. Clin. Neurophysiol. 107 (1998) 69–83. [2] C. Beaulieu, Numerical data on neocortical neurons in adult rat, with special reference to the GABA population, Brain Res. 609 (1993) 284–292. [3] M. Beierlein, J.R. Gibson, B.W. Connors, A network of electrically coupled interneurons drives synchronized inhibition in neocortex, Nat. Neurosci. 3 (2000) 904–910. [4] B.H. Bland, S.D. Oddie, L.V. Colom, Mechanisms of neural synchrony in the septohippocampal pathways underlying hippocampal theta generation, J. Neurosci. 19 (1999) 3223–3237. [5] G. Buzsaki, J.J. Chrobak, Temporal structure in spatially organized neuronal ensembles: a role for interneuronal networks, Curr. Opin. Neurobiol. 5 (1995) 504–510. [6] C.A. Chapman, J.C. Lacaille, Intrinsic theta-frequency membrane potential oscillations in hippocampal CA1 interneurons of stratum lacunosum-moleculare, J. Neurophysiol. 81 (1999) 1296–1307. [7] C.A. Chapman, J.C. Lacaille, Cholinergic induction of theta-frequency oscillations in hippocampal inhibitory interneurons and pacing of pyramidal cell firing, J. Neurosci. 19 (1999) 8637–8645. [8] A. Chow, A. Erisir, C. Farb, M.S. Nadal, A. Ozaita, D. Lau, E. Welker, B. Rudy, K 1 channel expression distinguishes subpopulations of parvalbumin- and somatostatin-containing neocortical interneurons, J. Neurosci. 19 (1999) 9332–9345. [9] S.R. Cobb, E.H. Buhl, K. Halasy, O. Paulsen, P. Somogyi, Synchronization of neuronal activity in hippocampus by individual GABAergic interneurons, Nature 378 (1995) 75–78. [10] T. Deboer, P. Franken, I. Tobler, Sleep and cortical temperature in the Djungarian hamster under baseline conditions and after sleep deprivation, J. Comp. Physiol. 174 (1994) 145–155. [11] T. Deboer, I. Tobler, Running wheel size influences circadian rhythm period and its phase shift in mice, J. Comp. Physiol. 186 (2000) 969–973. [12] A. Erisir, D. Lau, B. Rudy, C.S. Leonard, Function of specific K 1 channels in sustained high-frequency firing of fast-spiking neocortical interneurons, J. Neurophysiol. 82 (1999) 2476–2489. [13] T.F. Freund, G. Buzsaki, Interneurons of the hippocampus, Hippocampus 6 (1996) 347–470. [14] M. Galarreta, S. Hestrin, A network of fast-spiking cells in the neocortex connected by electrical synapses, Nature 402 (1999) 72–75. [15] A. Gupta, Y. Wang, H. Markram, Organizing principles for a diversity of GABAergic interneurons and synapses in the neocortex, Science 287 (2000) 273–278. [16] J.R. Gibson, M. Beierlein, B.W. Connors, Two networks of electrically coupled inhibitory neurons in neocortex, Nature 402 (1999) 75–79. [17] R. Hernandez-Pineda, A. Chow, Y. Amarillo, H. Moreno, M.

V.V. Vyazovskiy et al. / Brain Research 947 (2002) 204 – 211

[18]

[19] [20]

[21]

[22] [23]

[24]

[25]

Saganich, E.V. de Miera, A. Hernandez-Cruz, B. Rudy, Kv3.1– Kv3.2 channels underlie a high-voltage-activating component of the delayed rectifier K 1 current in projecting neurons from the globus pallidus, J. Neurophysiol. 82 (1999) 1512–1528. R.H. Joho, C.S. Ho, G.A. Marks, Increased gamma- and decreased delta-oscillations in a mouse deficient for a potassium channel expressed in fast-spiking interneurons, J. Neurophysiol. 82 (1999) 1855–1864. W. Klimesch, Memory processes, brain oscillations and EEG synchronization, Int. J. Psychophysiol. 24 (1996) 61–100. M. von Krosigk, T. Bal, D.A. McCormick, Cellular mechanisms of a synchronized oscillation in the thalamus, Science 261 (1993) 361– 364. D. Lau, E. Vega-Saenz de Miera, D. Contreras, A. Ozaita, M. Harvey, A. Chow, J.L. Noebels, R. Paylor, J.I. Morgan, C.S. Leonard, B. Rudy, Impaired fast-spiking, suppressed cortical inhibition, and increased susceptibility to seizures in mice lacking Kv3.2 K 1 channel protein, J. Neurosci. 20 (2000) 9071–9085. L.S. Leung, C.Y. Yim, Intracellular records of theta rhythm in hippocampal CA1 cells of the rat, Brain Res. 367 (1986) 323–327. M. Martina, J.H. Schultz, H. Ehmke, H. Monyer, P. Jonas, Functional and molecular differences between voltage-gated K 1 channels of fast-spiking interneurons and pyramidal neurons of rat, J. Neurosci. 18 (1998) 8111–8125. H. Moreno, C. Kentros, E. Bueno, M. Weiser, A. Hernandez, E. Vega-Saenz de Miera, A. Ponce, W. Thornhill, B. Rudy, Thalamocortical projections have a K 1 channel that is phosphorylated and modulated by cAMP-dependent protein kinase, J. Neurosci. 15 (1995) 5486–5501. B. Rudy, C. Kentros, M. Weiser, D. Fruhling, P. Serodio, E. Vega-Saenz de Miera, M.H. Ellisman, J.A. Pollock, Region-specific expression of a K 1 channel gene in brain, Proc. Natl. Acad. Sci. USA 89 (1992) 4603–4607.

211

[26] B. Rudy, A. Chow, D. Lau, Y. Amarillo, A. Ozaita, M. Saganich, H. Moreno, M.S. Nadal, R. Hernandez-Pineda, A. Hernandez-Cruz, A. Erisir, C. Leonard, E. Vega-Saenz de Miera, Contributions of Kv3 channels to neuronal excitability, Ann. N.Y. Acad. Sci. 868 (1999) 304–343. [27] I. Soltesz, M. Deschenes, Low- and high-frequency membrane potential oscillations during theta activity in CA1 and CA3 pyramidal neurons of the rat hippocampus under ketamine–xylazine anesthesia, J. Neurophysiol. 70 (1993) 97–116. [28] M. Steriade, Basic mechanisms of sleep generation, Neurology 42 (1992) 9–17. [29] M. Steriade, P. Gloor, R.R. Llinas, F.H. Lopes de Silva, M.M. Mesulam, Basic mechanisms of cerebral rhythmic activities, Electroencephalogr. Clin. Neurophysiol. 76 (1990) 481–508. [30] M. Steriade, D.A. McCormick, T. Sejnowski, Thalamocortical oscillations in the sleeping and aroused brain, Science 262 (1993) 679–685. [31] I. Tobler, T. Deboer, M. Fischer, Sleep and sleep regulation in normal and prion protein-deficient mice, J. Neurosci. 17 (1997) 1869–1879. [32] C.H. Vanderwolf, The electrocorticogram in relation to physiology and behavior: a new analysis, Electroencephalogr. Clin. Neurophysiol. 82 (1992) 165–175. [33] M. Weiser, E. Vega-Saenz de Miera, C. Kentros, H. Moreno, L. Franzen, D. Hillman, H. Baker, B. Rudy, Differential expression of 1 Shaw-related K channels in the rat central nervous system, J. Neurosci. 14 (1994) 949–972. [34] A. Ylinen, I. Soltesz, A. Bragin, M. Penttonen, A. Sik, G. Buzsaki, Intracellular correlates of hippocampal theta rhythm in identified pyramidal cells, granule cells, and basket cells, Hippocampus 5 (1995) 78–90.