- Email: [email protected]
Increase in antidromic excitability in presumed serotonergic dorsal raphe neurons during paradoxical sleep in the cat
Brain Research 898 (2001) 332–341 www.elsevier.com / locate / bres
Research report
Increase in antidromic excitability in presumed serotonergic dorsal raphe neurons during paradoxical sleep in the cat Kazuya Sakai*, Sylvain Crochet INSERM U480, Universite´ Claude Bernard, Lyon 1, Department of Experimental Medicine, 8 Avenue Rockefeller, 69373 Lyon, Cedex 08, France Accepted 30 January 2001
Abstract Putative serotonergic dorsal raphe (DRN) neurons display a dramatic state-related change in behaviour, discharging regularly at a high rate during waking and at progressively slower rates during slow-wave sleep (SWS) and ceasing firing during paradoxical sleep (PS). Using the antidromic latency technique and extracellular recording, we have examined the change in neuronal excitability of presumed serotonergic DRN neurons during the wake–sleep cycle in freely moving cats. We found that, under normal conditions, suprathreshold stimulation of the main ascending serotonergic pathway resulted in a marked decrease in both the magnitude and variability of antidromic latency during PS, while subthreshold stimulation led to a marked increase in antidromic responsiveness during PS compared with during other behavioural states. The antidromic latency shift resulted from a change in the delay between the initial segment (IS) and soma-dendritic (SD) spikes, the antidromic latency being inversely related to the interval between the stimulus and the preceding spontaneous action potential. A marked decrease in the magnitude and variability of antidromic latency was also seen following suppression of the spontaneous discharge of DRN neurons by application of 5-HT autoreceptor agonists or muscimol, a potent GABA agonist. A marked IS–SD delay or blockage of SD spikes was, however, seen in association with the PS occurring during recovery from 5-HT autoreceptor agonist or during muscimol application. The present findings are discussed in the light of previous in vitro intracellular recording data and our recent findings of the disfacilitation mechanisms responsible for the cessation of discharge of DRN neurons during PS. 2001 Elsevier Science B.V. All rights reserved. Theme: Neural basis of behaviour Topic: Biological rhythms and sleep Keywords: Antichromic excitability; Serotonergic dorsal raphe neurons; Paradoxical sleep
1. Introduction It is well known that putative 5-hydroxytryptamine (5HT, serotonin)-containing neurons in the dorsal raphe nucleus (DRN) display a dramatic state-related change in neuronal activity during the wake–sleep cycle, discharging slowly and regularly at a high rate during waking (W) and at progressively slower rates during slow-wave sleep (SWS) and exhibiting virtually complete cessation of discharge during paradoxical sleep (PS), also known as rapid eye movement (REM) sleep [9,23,35,43]. In this respect, DRN neurons are very similar to norepinephrine (NE)-containing neurons in the locus coeruleus (LC) and *Corresponding author. Tel.: 133-4-7877-7122; fax: 133-4-78777172. E-mail address: [email protected] (K. Sakai).
histamine (HA)-containing neurons in the tuberomamillary nucleus of the posterior hypothalamus. These neurons are, therefore, collectively referred to as PS-off or REM-off neurons [40]. In vitro extracellular and intracellular recordings from DRN neurons in brain slices have demonstrated that serotonergic DRN neurons show a long-duration action potential, a large after-hyperpolarisation (AHP) after each action potential and a consistent depressant response to lysergic acid diethylamide (LSD) and 5-HT agonists [2,11,26,36,44,45,50]. In vitro slice experiments have also shown that the post-spike AHP is mediated by a Ca 21 dependent K 1 conductance [4,11], while that induced by LSD or 5-HT is caused by an increase in a resting potassium conductance [3,48,50]. A marked early transient outward potassium current (IA ) and a low-threshold Ca 21 current (LTS) have also been demonstrated in serotonergic DRN neurons [1,3,8,34]. Although much is known about
0006-8993 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 01 )02210-7
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341
the membrane properties of serotonergic DRN neurons, no information is available regarding changes in their membrane potential and neuronal excitability during the sleep– waking cycle, largely due to technical difficulties. In a previous paper, we reported the existence of presumed non-monoaminergic PS-off neurons in the medulla and showed that the antidromic responses of these neurons were completely or partially blocked during PS with a prolongation of the antidromic latency, suggesting that inhibitory synaptic volleys impinge on these non-monoaminergic PS-off neurons during PS [33]. However, no information is available about changes in the antidromic responses of putative central monoaminergic neurons during different behavioural states. Recently, we found that the cessation of discharge of putative serotonergic DRN neurons during PS is not caused by g-aminobutyric acid (GABA)-mediated inhibition, but by disfacilitation, resulting from the cessation of discharge of NE- or HAcontaining neurons during PS [30]. Using extracellular recording and the antidromic latency technique, a useful tool for measuring changes in membrane potential [22], we have therefore studied changes in the neuronal excitability of presumed serotonergic DRN neurons during the wake– sleep cycle in the freely-moving cat.
2. Materials and methods A total of nine adult cats of both sexes were anaesthetised with pentobarbital (Nembutal, 25 mg / kg, i.v.) for implantation of the polygraphic recording electrodes using standard techniques, as previously described [10]. In addition, bipolar stimulating electrodes, consisting of two stainless steel wires (200 mm diameter, 1.0 mm apart, bared 0.5 mm at the tip), were placed in a major ascending 5-HT pathway at the level of the ventral tegmental area (VTA: A4.0, L2.0, HC-5.5) and lateral posterior hypothalamus (A9.0, L3.0, HC-4.0) [20] in order to identify antidromic responses of serotonergic DRN neurons [30]. For extracellular single unit recordings, two bundles of six flexible, Formvar-coated, stainless steel wires (32 mm in diameter) were implanted in the DRN through the guide cannulae (24 gauge) of a mechanical microdrive assembly, fixed to the skull at an angle of 508 to the horizontal. For microdialysis application of drugs, one guide cannula (23 gauge) was incorporated into the microdrive assembly between the two microelectrode bundles (1 mm separation between the centres of the guide cannulae), as previously described [31]. After recovery from surgery, the animals were housed individually in sound-attenuated, dimly illuminated electrically-shielded recording chambers maintained at 24– 268C. Food and water were available ad libitum. The cats were habituated to the experimental conditions until they displayed normal sleep–waking cycles. All experimental procedures followed EEC Guidelines (86 / 609 / EEC) and
333
every effort was made to minimise the number of animals used and any pain or discomfort. The polygraphic measurements included neocortical and dorsal hippocampal electroencephalograms (EEG), an electrooculogram (EOG), a neck muscle electromyogram (EMG) and ponto–geniculo–occipital (PGO) waves recorded from the dorsal lateral geniculate nucleus (LGN). The polygraphic data were digitised using a CED 1401 data processor (Cambridge Electronic Design, Cambridge, UK) at a sampling rate of 250 Hz, and stored initially on a personal computer, then on digital tape. Unit recordings were made differentially between active and reference microelectrodes within the same bundle in order to avoid artefacts due to movements of the animal. The unit activities were amplified using a conventional amplifier (Model P15, Grass Instrument, Quincy, USA) with low and high cut-off filters of 100 Hz and 10 kHz, respectively, and were monitored on a storage-oscilloscope (5103N, Tectronix, Les Ulis, France) and on a digital memory oscilloscope equipped with a processor for spike waveform averaging (DRO 1604, Gould Electronique, Longjumeau, France). Single unit activity was separated from background noise to obtain on-line output of the unit discharge on an oscilloscope and on a computer connected to the CED 1401 intelligent interface. The unit activities were digitised using a CED 1401 data processor at a sampling rate of either 20 or 25 kHz and analysed using Spike2 software (CED). Stimulation was performed with square pulses (0.2–0.5 ms), usually delivered at 0.6–0.8 Hz, using a WPI 302 stimulator (World Precision Instruments, Florida, USA) and a stimulus isolation unit (Model 305-2, WPI). The stimulation current was adjusted to a setting that antidromically activated the neuron in 100% of the non-collision trials (suprathreshold current) or was below the threshold for antidromically activating the neuron during W (subthreshold current). For neurons with multiple antidromic latencies, the higher stimulation current eliciting the shortest latency response was used for the threshold current. In the present study on presumed serotonergic DRN neurons showing a low safety factor for antidromic invasion of the soma [5,46,47], we used the collision of spontaneous and stimulation-evoked action potentials (collision test) as the criterion for differentiating between antidromic invasion and synaptic activation (Fig. 1) [15,22]. The antidromic latency was measured from the onset of the stimulus (S) to the middle of the linear portion of the somato-dendritic (SD) component of the spike. Mean latencies were calculated from continuous recordings lasting longer than 1 min using 0.02- or 0.04-ms bins for steady states of W, SWS and PS. The conduction velocity was estimated from the straight line distance between recording and stimulating sites and the shortest antidromic latency. Individual DRN neurons were identified as serotonergic according to previously established criteria [14,20] as
334
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341
propylamino)tetralin hydrobromide (8-OH-DPAT), muscimol and bicuculline methiodide (all from Sigma–Aldrich, Saint Quentin, France). All drugs were dissolved in Ringer’s solution just before use. At the end of the experiments, all animals were deeply anaesthetised, then several unit recording sites were coagulated by passing a 20 mA cathodal current for 10 s and the locations of the recording microelectrode and stimulating electrodes were determined histologically on 25 mm sagittal sections, as previously described [31].
3. Results
Fig. 1. Antidromic responses of a presumed serotonergic DRN neuron following stimulation of the main ascending 5-HT pathway at the level of the ventral tegmental area (VTA). Antidromic action potentials are shown in A–E and a collision with a spontaneous action potential in F (shown by black arrow). Note the marked variability in latency and the lengthening of the antidromic latency when the stimulus was delivered shortly after spontaneous spike at an interval longer than the collision interval. A, five superimposed sweeps. The onset of the stimulus (0.2 ms) is indicated by the black triangle. The vertical dashed line indicates the onset of the shortest antidromic latency.
follows: (1) long-duration (.2 ms) bi- or triphasic action potential, (2) slow (,6 Hz) discharge activity during waking, (3) reduction in spontaneous discharge rate during PS, and (4) histological localisation of recording sites to the DRN. In five cats, a microdialysis probe (Eicom A-L-50-01, Kyoto, Japan) was inserted through the guide cannula into the middle portion of the DRN. The probe membrane was 1 mm in length and had an external diameter of 0.23 mm. Unit recordings and microdialysis infusion of drugs were then carried out for 5–6 consecutive days. During experiments, the probe was continuously perfused at a flow rate of 5 ml / min with either Ringer’s solution or the test drug dissolved in Ringer’s solution, as previously described [31]. The main drugs used were 5-methoxy-N,N-dimethyltryptamine (5-MeODMT), 8-hydroxy-2-(n-di-
A total of 75 presumed serotonergic DRN neurons were activated antidromically by unilateral stimulation of the main serotonergic ascending pathway at the level of the VTA and / or lateral posterior hypothalamus. Of these 75 neurons, 49 were subjected to continuous antidromic stimulation at a slow stimulation rate (,1 / s) during a sleep–waking cycle including at least one PS episode. The results presented were mainly obtained from 32 typical presumed serotonergic DRN neurons which displayed slow regular firing during quiet waking (QW) (mean6S.D.5 2.660.9 spikes / s) and virtually complete suppression of discharge during PS (0.0160.05 spikes / s). The estimated mean conduction velocity (6S.D.) of these neurons was 0.9260.49 m / s (range 0.24–2.50), this slow conduction velocity being indicative of non-myelinated or finely myelinated axons, a characteristic of serotonergic DRN neurons [6]. When stimulated during W, all neurons displayed post-stimulus suppression of spontaneous activity for a period of 0.15–0.20 s. In those neurons tested, complete suppression of spontaneous discharge was seen following systemic (i.m.) administration of either 50–200 mg / kg of 5-MeODMT (n511) or 10–25 mg / kg of 8-OHDPAT (n56). As described for other central monoaminergic neurons [18,25,27,41,42], every presumed serotonergic DRN neuron displayed a large variability in latency (.0.1 ms) when suprathreshold stimulation was applied during W at the slow stimulation rate (Fig. 1). When the neurons were stimulated with twin pulses separated by less than 10 ms, a complete blockage of the SD (or B) spike (Fig. 2A and B) or a marked IS–SD delay (or A–B break) (Fig. 2B) was observed in the second response, which had a longer latency than the first (Fig. 2), a phenomenon commonly seen in slowly conducting, small myelinated and unmyelinated fibre systems, such as presumed catecholaminergic neurons in the locus coeruleus [18] and nucleus commissuralis [25] and presumed cholinergic neurons in the mesopontine tegmentum [32]. In contrast, other central neurons having a high safety factor for antidromic invasion, such as pyramidal tract neurons [38] and presumed non-monoaminergic PS-off neurons in the medulla [33], have been reported to display fixed latency (,0.1 ms) and
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341
335
Fig. 2. Antidromic responses of a presumed serotonergic DRN neuron to double-stimulation of the VTA. Note the complete blockage of the SD spike (A) or the IS–SD break (B) when double stimuli were applied at an inter-pulse interval of less than 10 ms (shown by black arrows). Note also the lengthening of the second response latency. The onset of the stimuli (0.2 ms) are indicated by the black triangles.
faithful response to high frequency (.200 Hz) stimulation (for review, see Ref. [22]). Fig. 3 shows dot displays of antidromic latencies obtained using either suprathreshold (A) or subthreshold (B) stimulation of the main ascending serotonergic pathway during the sleep–waking cycle. Using suprathreshold stimulation, every typical presumed serotonergic DRN neuron (n532) displayed a reduction in the magnitude and variability of the antidromic latency during PS compared with during other behavioural states (Fig. 3A), the latency shift being related to the spontaneous firing rate. Marked shortening of antidromic latency was also seen during SWS in a group of neurons that ceased firing during SWS accompanying PGO waves (Fig. 3A-2). Overall, the shortest and longest latencies were seen during PS and W, respectively, and a gradual shortening of the antidromic latency was seen during PS (Fig. 3). Using subthreshold stimulation, all seven neurons tested showed an increased antidromic responsiveness during PS compared with during other states (Fig. 3B-2), the responsiveness being inversely related to the spontaneous firing rate (Fig. 3B1,2). As previously reported by Gustafsson and Lipski [17], the latency shift (S–SD delay) was mainly due to a change in the delay between the IS and SD spikes (Fig. 4), and, as shown in Fig. 5, the latency was inversely related to the interval between the stimulus and the preceding spontaneous action potential, especially in the case of intervals of less than 1 s. There was a marked decrease in antidromic latency and latency variability when the stimulus was applied 1–2 s after the spontaneous action potential during the period corresponding to deep SWS accompanying PGO waves. Fig. 6 shows the relationship between the magnitude of the latency variation (mean latency during W-mean latency during PS) and the initial
Fig. 3. Antidromic responses of four different presumed serotonergic DRN neurons during wake–sleep states following stimulation of the main ascending 5-HT pathway at the level of the VTA. A and B: suprathreshold (A) and subthreshold (B) stimulation. The upper traces in each record show dot displays of the antidromic latency, while the lower traces show the spontaneous discharge rate (spikes / s) after removal of the antidromic spike responses and stimulus artefacts. Note the decrease in the magnitude and variability of the antidromic latency (A) and the increase in the antidromic responsiveness (B) during PS and the SWS episodes with reduced spontaneous discharge activity. Also note the gradual shortening of the antidromic latency seen during PS. EEG, neocortical electroencephalogram; PGO, dorsal lateral geniculate EEG, showing ponto– geniculo–occipital (PGO) waves.
latency seen during W. No significant correlation was seen with stimulation of the VTA (r50.005) or the lateral posterior hypothalamus (HYP: r50.45). The change in antidromic latency was then studied following systemic (i.m.) or microdialysis application of the 5-HT autoreceptor agonists, 5-MeODMT or 8-OHDPAT, or the GABA agonist, muscimol, all of which induce suppression of spontaneous discharge in serotonergic DRN neurons [12,30,36,43,44]. As shown in the example in Fig. 7B, systemic administration of 50 (n53), 100 (n53) or 200 mg / kg (n53) of 5-MeODMT or 10 (n52) or 25 mg / kg (n51) of 8-OHDPAT led to complete suppression of spontaneous discharge in all typical presumed DRN neurons studied.
336
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341
Fig. 4. Antidromic latency variation during wake–sleep states. A, antidromic responses during W, SWS and PS obtained by signal averaging. The number of samples (n) is indicated in parenthesis. B, antidromic action potentials seen during W (1), SWS (2) and PS (3) (three superimposed sweeps). Note that the change in antidromic latency results from a change in the IS–SD delay. The vertical dotted line represents the onset of IS spike.
Following administration of the 5-HT autoreceptor agonists, the antidromic latency became progressively shortened in parallel with the gradual decrease in spontaneous discharge rate, reaching a plateau at the same time as discharge activity completely ceased. This latency shift was caused by shortening of the IS–SD delay (Fig. 7b), the magnitude of the shift being similar to, or greater than, that seen during a normal PS episode (Fig. 7a). Both the magnitude and variability of the antidromic latency then increased during recovery, paralleling the increase in spontaneous discharge rate (Fig. 7B). When PS occurred during the recovery period, a marked IS–SD delay or complete blockage of the SD component of the antidromic spike was noted in all neurons tested (Fig. 8), a result exactly the opposite of that seen during PS under normal conditions. Three other neurons were tested using microdialysis application of 100 mM muscimol. As illustrated in Fig. 9, muscimol completely suppressed the spontaneous firing of DRN neurons 20 to 30 min after the start of drug application. As with the 5-HT autoreceptor agonists, a decrease in both the magnitude and variability of antidromic latency was seen which paralleled the drug-induced decrease in spontaneous discharge activity. However, using
Fig. 5. Two different examples of the relationship between the antidromic latency and the interval between the stimulus and the preceding spontaneous action potential. Suprathreshold stimulation of the VTA at 0.6 Hz was used. Note that the antidromic latency is inversely related to the interval, especially when the interval is less than 1 s.
muscimol, a marked increase in antidromic latency and latency variability was seen during both PS and the SWS episode immediately preceding it, a result contrasting sharply with that seen under normal conditions.
4. Discussion The antidromic latency technique was introduced by Merrill [24] and has been used in many extracellular studies, including some on sleep [33,37–39]. It is based on the principle that the delay between the stimulus (S) and the SD component of the antidromic spike invariably depends on the soma membrane polarisation, with soma membrane hyperpolarisation increasing the delay and depolarisation shortening it. The main advantage of this technique is that intracellular recordings, which are difficult to realise in behaving animals, can be avoided. Using both intracellular and extracellular recordings, Gustafsson and Lipski [17] showed that changes in the latency to the SD spike (S–SD delay) are mainly caused by changes in the IS–SD delay, which varies from 10 to 100 ms per mV change in membrane potential. The shortening of the antidromic latency, as observed extracellularly, is associ-
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341
Fig. 6. Relationship between the initial antidromic latency seen during W and the maximal latency variation seen during W and PS. No significant correlation was seen using suprathreshold stimulation of either the VTA (upper plot; F51.13, P50.30) or the lateral posterior hypothalamus (HYP) (lower plot; F52.30, P50.16).
ated with excitation or disinhibition of the cell, while its prolongation is related to inhibition or disfacilitation (see Ref. [22] for review). In the present study, we found that, under normal conditions, presumed serotonergic DRN neurons display both a reduction in magnitude and variability of antidromic latency and an increase in antidromic responsiveness during PS compared with during other behavioural states, suggesting an increase in neuronal excitability of presumed serotonergic DRN neurons during PS, even though spontaneous discharge activity is almost completely suppressed during this state. This increase in neuronal excitability seems to be due to the mechanism of disinhibition. Previous intracellular studies in vivo [4] and in vitro [8,11,44,49,50] have shown that typical serotonergic DRN neurons exhibit a resting membrane potential of approximately 260 mV and a large (10–20 mV) and long-lasting ($150–300 ms) AHP after each action potential. The AHP, mediated by a Ca 21 -dependent K 1 conductance [4,11], is followed by gradual inter-spike depolarisation, a plateau period, and then a depolarising pre-potential (LTS) which triggers the succeeding spike. The activation threshold for the LTS is approximately 260 mV [8]. It therefore appears that spontaneously discharging serotonergic DRN
337
neurons are in a hyperpolarised state during most of the inter-spike interval. In the present study, the antidromic latencies of presumed serotonergic DRN neurons were inversely related to the interval between the stimulus and the preceding spontaneous action potential, particularly when the interval was less than 1 s (Fig. 5). When the interval was greater than 1 s, as seen during deep SWS with reduced spontaneous discharge activity, there was a marked reduction in both antidromic latency and latency variability. It therefore seems that the marked prolongation of antidromic latency and variability seen during W and SWS can be ascribed to the prominent post-spike AHP, an intrinsic property of serotonergic DRN neurons. It should be mentioned that gradual shortening of the IS–SD delay was also seen during PS (Fig. 3A), suggesting gradual depolarisation of the soma membrane during this sleep state. It is open to question whether this depolarisation, associated with cessation of spontaneous discharge, represents a restorative process of serotonergic DRN neurons. The present findings appear to be inconsistent with the assumption that the cessation of discharge of serotonergic DRN neurons seen during PS is due to strong inhibition, since no prolongation of antidromic latency or blockage of the SD spike was seen during PS under normal conditions. As briefly described in the Introduction, putative serotonergic DRN neurons selectively cease firing during PS, and it has been postulated that this may result from GABAmediated inhibition and promote PS generation [16,28,29]. However, in agreement with a previous report that the suppression of DRN neuronal activity during PS is unaffected by iontophoretic application of bicuculline, a potent GABA antagonist [21], we recently found that the cessation of discharge of DRN neurons is not attenuated by microdialysis application of bicuculline at concentrations that are able to completely antagonise the potent depressant effect of muscimol, but is completely blocked by either histamine or phenylephrine, a selective a 1 -adrenoceptor agonist. Application of mepyramine, a specific H 1 histamine receptor antagonist, or prazosin, a specific a 1 adrenoceptor antagonist, suppresses the spontaneous discharge of DRN neurons during QW and SWS [30]. In addition, systemic [13] or microdialysis application of WAY100635, a specific 5-HT 1A autoreceptor antagonist, to the DRN [30] does not reverse the cessation of firing of DRN neurons during PS. These findings strongly suggest that this cessation of DRN unit activity during PS is not caused by GABA- or 5-HT-mediated inhibition, but by the mechanism of disfacilitation resulting from the cessation of discharge of NE- or HA-containing neurons during PS. It has previously been shown that spontaneous activity of serotonergic DRN neurons results from noradrenergic drive in the anaesthetised rat [7] and that NE depolarises serotonergic DRN neurons via a 1 -adrenoceptors in vitro by decreasing a resting K 1 conductance [1,49]. In addition, a 1 -adrenoceptor stimulation is reported to suppress the IA , which slows the rate of depolarisation of serotoner-
338
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341
Fig. 7. Effects of administration of 8-OH-DPAT on the antidromic latency of a presumed serotonergic DRN neuron. Left: latency and spontaneous discharge rate in normal control (A) and before, and after, administration of 8-OH-DPAT (10 mg / kg, i.m.) (B). Upper trace, each latency is shown by a dot; lower trace, spontaneous discharge rate (spikes / s) after removal of the antidromic responses and stimulus artefacts. Right: a, superimposed antidromic action potentials (n510) during SWS and PS under normal conditions seen at the times marked 1 and 2 in A. b, Superimposed antidromic action potentials (n510) seen at the times marked 1–4 in B, i.e. (1) shortly before and (2) shortly after 8-OH-DPAT administration, (3) at the time when spontaneous discharge activity was suppressed and (4) during recovery. Note the shortening of the IS–SD delay which was associated with suppression of spontaneous unit discharge seen during PS (a-2) or following 8-OH-DPAT administration (b-3). A, B, a and b, suprathreshold stimulation of the VTA at 0.6 Hz.
gic DRN neurons [1], while the LTS, which accelerates the rate of depolarisation, is not affected by noradrenergic input [8]. It should be noted that the ability of a phasic auditory stimulus to elicit a DRN unit response persists during PS [19], consistent with the idea that the excitability of serotonergic DRN neurons is not reduced during PS. The effects of HA on the membrane potential of serotonergic DRN neurons are not yet known. As during PS, shortening of the antidromic latency was also seen following application of 5-HT autoreceptor agonists or muscimol, a GABA agonist, all of which cause suppression of spontaneous discharge in presumed serotonergic DRN neurons. Previous intracellular recording studies have shown that both 5-HT autoreceptor agonists and GABA induce a dose-dependent hyperpolarisation (,20 mV) in serotonergic DRN neurons [3,8,36,48,50]. 5-HT appears to have no direct effect on the A current or the LTS [8]. 5-HT and LSD appear to block the gradual depolarisation or decay in the post-spike AHP by opening resting K 1 channels with the result that the neurons remain hyperpolarised and lose their rhythmic firing pattern [4,8]. It is possible that the hyperpolarisation induced by these drugs is smaller than that caused by the
post-spike AHP, at least at the concentrations used in the present study and in conscious waking cats, in which serotonergic DRN neurons receive tonic excitatory inputs from NE- or HA-containing, and other, possibly glutamatergic, neurons [30]. It remains to be determined whether larger doses of 5-HT agonists or higher concentrations of muscimol lead to prolongation or blockage of the SD spike in serotonergic DRN neurons during W. A previous intracellular study in anaesthetised rats [4] has shown that administration of LSD produces hyperpolarisation (4–8 mV) in serotonergic DRN neurons, but does not raise the threshold for spiking induced by intracellular depolarising pulses, consistent with the assumption that DRN neurons remain responsive to phasic excitatory volleys when the 5-HT autoreceptor is stimulated. When PS occurred during recovery from 5-HT autoreceptor agonist administration or during muscimol application, a marked IS–SD delay or complete blockage of the SD component of the antidromic spike was seen in every presumed serotonergic DRN neuron tested, suggesting the occurrence of a greater degree of hyperpolarization during the PS episodes. These effects may be attributed to the combination of the inhibition induced by
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341
339
Fig. 9. Antidromic latency change during PS episodes during microdialysis application of muscimol. The upper trace is a dot display of the antidromic latency of a presumed serotonergic DRN neuron following suprathreshold stimulation of the VTA, while the lower trace shows the spontaneous discharge rate (spikes / s), the antidromic spikes and stimulus artefacts being removed from the data. The data were obtained 60 min after the start of application of 100 mM muscimol to the DRN near the unit recording site. Note the virtually complete suppression of spontaneous discharge activity during muscimol application and the marked increase in antidromic latency and latency variability seen during PS. Also note the recovery of spontaneous discharge activity seen during W after microdialysis application of 100 mM bicuculline. The bars under the drug names indicate the drug application period. EEG, neocortical electroencephalogram; EMG, electromyogram; EOG, electrooculogram; PGO, dorsal lateral geniculate EEG, showing ponto–geniculo–occipital (PGO) waves.
Fig. 8. Antidromic latency change during PS following 8-OH-DPAT administration. The upper trace in A is a dot display of the antidromic latency of a presumed serotonergic DRN neuron following suprathreshold stimulation of the VTA, while the lower trace shows the spontaneous discharge rate (spikes / s), the antidromic spike responses and stimulus artefacts being removed from the data. The data were obtained 50–90 min after 8-OH-DPAT (10 mg / kg, i.m.) administration. The mean discharge rate of this neuron during QW was 3.0 spikes / s in control periods. B, superimposed antidromic action potentials at the times indicated as 1–4 in A, i.e. (1) during SWS before PS, (2) during the early phase of PS, (3) during the late phase of PS and (4) during SWS after the PS episode. Note the marked IS–SD break (B-2) or complete blockage of the SD spike (B-3) seen during PS following 8-OH-DPAT administration. The black arrows in B indicate the IS–SD break (B-2) or the blockage of the SD spike (B-3).
activation of 5-HT autoreceptors or GABA receptors and the disfacilitation due to the cessation of discharge of NE or HA neurons during PS. Future intracellular recordings in conscious cats are required to elucidate the exact membrane mechanisms underlying the change in antidromic latency of putative serotonergic DRN neurons during the wake–sleep states. In conclusion, the present findings suggest that, under normal conditions, the neuronal excitability of presumed serotonergic DRN neurons increases during PS mainly due to the mechanism of disinhibition as a result of the
removal of the marked AHP that follows each action potential.
Acknowledgements This work was supported by INSERM U480.
References [1] G.K. Aghajanian, Modulation of a transient outward current in serotonergic neurons by a 1 -adrenoceptors, Nature 315 (1985) 501– 503. [2] G.K. Aghajanian, H.J. Haigler, F.E. Bloom, Lysergic acid diethylamide and serotonin: direct actions on serotonin-containing neurons in rat brain, Life Sci. 11 (1972) 615–622. [3] G.K. Aghajanian, J.M. Lakoski, Hyperpolarization of serotonin neurons by serotonin and LSD: studies in brain slices showing increased K 1 conductance, Brain Res. 305 (1984) 181–185. [4] G.K. Aghajanian, C.P. Vandermaelen, Intracellular recordings from serotonergic dorsal raphe neurons: pacemaker potentials and the effect of LSD, Brain Res. 238 (1982) 463–469. [5] G. Aston-Jones, M. Segal, F. Bloom, Brain aminergic axons exhibit marked variability in conduction velocity, Brain Res. 195 (1980) 215–222. [6] E. Azmitia, P. Gannon, The ultrastructural localization of serotonin immunoreactivity in myelinated and unmyelinated axons within the
340
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20] [21]
[22]
[23] [24]
[25]
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341 medial forebrain bundle of rat and monkey, J. Neurosci. 3 (1983) 2083–2090. J.M. Baraban, G.K. Aghajanian, Suppression of firing activity of 5-HT neurons in the dorsal raphe by alpha-adrenoceptor antagonists, Neuropharmacology 19 (1980) 355–363. T.M. Burlhis, G.K. Aghajanian, Pacemaker potentials of serotonergic dorsal raphe neurons: contribution of a low-threshold Ca 21 conductance, Synapse 1 (1987) 582–588. R. Cespuglio, H. Faradji, M.E. Gomez, M. Jouvet, Single unit recording in the nuclei raphe dorsalis and magnus during the sleep–waking cycle of semi-chronic prepared cats, Neurosci. Lett. 24 (1981) 133–138. S. Crochet, K. Sakai, Effects of microdialysis application of monoamines on the EEG and behavioral states in the cat mesopontine tegmentum, Eur. J. Neurosci. 11 (1999) 3738–3752. V. Crunelli, S. Forda, P.A. Brooks, K.C.P. Wilson, J.C.M. Wise, J.S. Kelly, Passive membrane properties of neurons in the dorsal raphe and periaqueductal grey recorded in vitro, Neurosci. Lett. 40 (1983) 263–268. C.A. Fornal, W.J. Litto, C.W. Metzler, F. Marrosu, K. Tada, B.L. Jacobs, Single-unit responses of serotonergic dorsal raphe neurons to 5-HT 1A agonist and antagonist drug administration in behaving cats, J. Pharmacol. Exp. Ther. 270 (1994) 1345–1358. C.A. Fornal, C.W. Metzler, R.A. Gallegos, S.C. Veasey, A.C. McCreary, B.L. Jacobs, WAY-100635, a potent and selective 5hydroxytryptamine 1A antagonist, increases serotonergic neuronal activity in behaving cats: comparison with (S)-WAY-100135, J. Pharmacol. Exp. Ther. 278 (1996) 752–762. C.A. Fornal, C.W. Metzler, F. Marrosu, L.E. Ribiero-do-Valle, B.L. Jacobs, A subgroup of dorsal raphe serotonergic neurons in the cat is strongly activated during oral–buccal movements, Brain Res. 716 (1996) 123–133. J.H. Fuller, J.D. Schlag, Determination of antidromic excitation by the collision test: problems of interpretation, Brain Res. 112 (1976) 283–298. D. Gervasoni, C. Peyron, C. Rampon, B. Barbagli, G. Chouvet, N. Urbain, P. Fort, P.H. Luppi, Role and origin of the GABAergic innervation of dorsal raphe serotonergic neurons, J. Neurosci. 20 (2000) 4217–4225. B. Gustafsson, J. Lipski, Effect of membrane polarization and synaptic activity on the timing of antidromic invasion, Brain Res. 181 (1980) 61–74. P.G. Guyenet, The coeruleospinal noradrenergic neurons: anatomical and electrophysiological studies in the rat, Brain Res. 189 (1980) 121–133. J. Heym, M.E. Trulson, B.L. Jacobs, Raphe unit activity in freely moving cats: effects of phasic auditory and visual stimuli, Brain Res. 232 (1982) 29–39. B.L. Jacobs, E.C. Azmitia, The structure and function of the brain serotonin system, Physiol. Rev. 72 (1992) 165–229. E.S. Levine, B.L. Jacobs, Neurochemical afferents controlling the activity of serotonergic neurons in the dorsal raphe nucleus: microiontophoretic studies in the awake cat, J. Neurosci. 12 (1992) 4037–4044. J. Lipski, Antidromic activation of neurons as an analytic tool in the study of the central nervous system, J. Neurosci. Meth. 4 (1981) 1–32. D.J. McGinty, R.M. Harper, Dorsal raphe neurons: depression of firing during sleep in cats, Brain Res. 101 (1976) 569–575. E.G. Merrill, Temporal patterns of antidromic invasion latencies for the respiratory neurons of the nucleus retroambigualis in cats, J. Physiol. (Lond.) 223 (1972) 18–20P. S.D. Moore, P.G. Guyenet, An electrophysiological study of the forebrain projection of nucleus commissuralis: preliminary identification of presumed A2 catecholaminergic neurons, Brain Res. 263 (1983) 211–222.
[26] S.B. Mosko, B.L. Jacobs, Recording of dorsal raphe unit activity in vitro, Neurosci. Lett. 2 (1976) 195–200. [27] S. Nakamura, Some electrophysiological properties of neurons of rat locus coeruleus, J. Physiol. 267 (1977) 641–658. [28] D. Nitz, J. Siegel, GABA release in the dorsal raphe nucleus: role in the control of REM sleep, Am. J. Physiol. 273 (1997) R451–R455. [29] C.M. Portas, M. Thakkar, D. Rainnie, R.W. McCarley, Microdialysis perfusion of 8-hydroxy-2-(di-n-propylamino)tetralin (8-OH-DPAT) in the dorsal raphe nucleus decreases serotonin release and increases rapid eye movement sleep in the freely moving cat, J. Neurosci. 16 (1996) 2820–2828. [30] K. Sakai, S. Crochet, Serotonergic dorsal raphe neurons cease firing by disfacilitation during paradoxical sleep, NeuroReport 11 (2000) 3237–3241. [31] K. Sakai, S. Crochet, Role of dorsal raphe neurons in paradoxical sleep generation in the cat: no evidence for a serotonergic mechanism, Eur. J. Neurosci. 13 (2001) 103–112. [32] K. Sakai, M. El Mansari, M. Jouvet, Inhibition by carbachol microinjections of presumptive cholinergic PGO-on neurons in freely moving cats, Brain Res. 527 (1990) 213–223. [33] K. Sakai, N. Kanamori, Are there non-monoaminergic paradoxical sleep-off neurons in the brainstem?, Sleep Res. Online 2 (1999) 57–63. [34] M. Segal, A potent transient outward current regulates excitability of dorsal raphe neurons, Brain Res. 359 (1985) 347–350. [35] K. Shima, H. Nakahama, M. Yamamoto, Firing properties of two types of nucleus raphe dorsalis neurons during the sleep–waking cycle and their responses to sensory stimuli, Brain Res. 399 (1986) 317–326. [36] J.C. Sprouse, G.K. Aghajanian, Electrophysiological responses of serotoninergic dorsal raphe neurons to 5-HT 1A and 5-HT 1B agonists, Synapse 1 (1987) 3–9. [37] M. Steriade, Mechanisms underlying cortical activation: neuronal organization and properties of the midbrain reticular core and intralaminar thalamic nuclei, in: O. Pompeiano, C. Ajmone Marsan (Eds.), Brain Mechanisms and Perceptual Awareness, Raven Press, New York, 1981, pp. 327–377. ˆ [38] M. Steriade, M. Deschenes, Inhibitory processes and interneuronal apparatus in motor cortex during sleep and waking. II. Recurrent and afferent inhibition of pyramidal tract neurons, J. Neurophysiol. 37 (1974) 1093–1113. [39] M. Steriade, J.A. Hobson, Neuronal activity during the sleep– waking cycle, Prog. Neurobiol. 6 (1976) 155–376. [40] M. Steriade, R.W. McCarley, Brainstem Control of Wakefulness and Sleep, Plenum Press, New York, 1990. [41] M. Takigawa, G.J. Mogenson, A study of inputs to antidromically identified neurons of the locus coeruleus, Brain Res. 135 (1977) 217–230. [42] J.M. Tepper, S. Nakamura, S.J. Young, P.M. Groves, Autoreceptormediated changes in dopaminergic terminal excitability: effects of striatal drug infusions, Brain Res. 309 (1984) 317–333. [43] M.E. Trulson, B.L. Jacobs, Raphe unit activity in freely moving cats: correlation with level of behavioral arousal, Brain Res. 163 (1979) 135–150. [44] C.P. Vandermaelen, G.K. Aghajanian, Electrophysiological and pharmacological characterization of serotonergic dorsal raphe neurons recorded extracellularly and intracellularly in rat brain slices, Brain Res. 289 (1983) 109–119. [45] C.P. Vandermaelen, G.K. Matheson, R.C. Wilderman, L.A. Petterson, Inhibition of serotonergic dorsal raphe neurons by systemic and iontophoretic administration of buspiron, a non-benzodiazepine anxiolytic drug, Eur. J. Pharmacol. 1986 (1986) 123–130. [46] R.Y. Wang, G.K. Aghajanian, Collateral inhibition of serotonergic neurons in the rat dorsal raphe nucleus: pharmacological evidence, Neuropharmacology 17 (1978) 819–825. [47] K. Watabe, T. Shinba, T. Satoh, The characteristics of the antidromic
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341 discharge of cat dorsal raphe neurons during repetitive activation, Brain Res. 327 (1985) 231–239. [48] J.T. Williams, W.F. Colmers, Z.Z. Pan, Ligand- and voltage-activated inwardly rectifying currents in dorsal raphe neurons in vitro, J. Neurosci. 8 (1988) 3499–3506. [49] M. Yoshimura, H. Higashi, Noradrenaline mediated slow excitatory
341
synaptic potentials in rat dorsal raphe neurons in vitro, Neurosci. Lett. 61 (1985) 305–310. [50] M. Yoshimura, S. Higashi, 5-Hydroxytryptamine mediates inhibitory postsynaptic potentials in rat dorsal raphe neurons, Neurosci. Lett. 53 (1985) 69–74.
Research report
Increase in antidromic excitability in presumed serotonergic dorsal raphe neurons during paradoxical sleep in the cat Kazuya Sakai*, Sylvain Crochet INSERM U480, Universite´ Claude Bernard, Lyon 1, Department of Experimental Medicine, 8 Avenue Rockefeller, 69373 Lyon, Cedex 08, France Accepted 30 January 2001
Abstract Putative serotonergic dorsal raphe (DRN) neurons display a dramatic state-related change in behaviour, discharging regularly at a high rate during waking and at progressively slower rates during slow-wave sleep (SWS) and ceasing firing during paradoxical sleep (PS). Using the antidromic latency technique and extracellular recording, we have examined the change in neuronal excitability of presumed serotonergic DRN neurons during the wake–sleep cycle in freely moving cats. We found that, under normal conditions, suprathreshold stimulation of the main ascending serotonergic pathway resulted in a marked decrease in both the magnitude and variability of antidromic latency during PS, while subthreshold stimulation led to a marked increase in antidromic responsiveness during PS compared with during other behavioural states. The antidromic latency shift resulted from a change in the delay between the initial segment (IS) and soma-dendritic (SD) spikes, the antidromic latency being inversely related to the interval between the stimulus and the preceding spontaneous action potential. A marked decrease in the magnitude and variability of antidromic latency was also seen following suppression of the spontaneous discharge of DRN neurons by application of 5-HT autoreceptor agonists or muscimol, a potent GABA agonist. A marked IS–SD delay or blockage of SD spikes was, however, seen in association with the PS occurring during recovery from 5-HT autoreceptor agonist or during muscimol application. The present findings are discussed in the light of previous in vitro intracellular recording data and our recent findings of the disfacilitation mechanisms responsible for the cessation of discharge of DRN neurons during PS. 2001 Elsevier Science B.V. All rights reserved. Theme: Neural basis of behaviour Topic: Biological rhythms and sleep Keywords: Antichromic excitability; Serotonergic dorsal raphe neurons; Paradoxical sleep
1. Introduction It is well known that putative 5-hydroxytryptamine (5HT, serotonin)-containing neurons in the dorsal raphe nucleus (DRN) display a dramatic state-related change in neuronal activity during the wake–sleep cycle, discharging slowly and regularly at a high rate during waking (W) and at progressively slower rates during slow-wave sleep (SWS) and exhibiting virtually complete cessation of discharge during paradoxical sleep (PS), also known as rapid eye movement (REM) sleep [9,23,35,43]. In this respect, DRN neurons are very similar to norepinephrine (NE)-containing neurons in the locus coeruleus (LC) and *Corresponding author. Tel.: 133-4-7877-7122; fax: 133-4-78777172. E-mail address: [email protected] (K. Sakai).
histamine (HA)-containing neurons in the tuberomamillary nucleus of the posterior hypothalamus. These neurons are, therefore, collectively referred to as PS-off or REM-off neurons [40]. In vitro extracellular and intracellular recordings from DRN neurons in brain slices have demonstrated that serotonergic DRN neurons show a long-duration action potential, a large after-hyperpolarisation (AHP) after each action potential and a consistent depressant response to lysergic acid diethylamide (LSD) and 5-HT agonists [2,11,26,36,44,45,50]. In vitro slice experiments have also shown that the post-spike AHP is mediated by a Ca 21 dependent K 1 conductance [4,11], while that induced by LSD or 5-HT is caused by an increase in a resting potassium conductance [3,48,50]. A marked early transient outward potassium current (IA ) and a low-threshold Ca 21 current (LTS) have also been demonstrated in serotonergic DRN neurons [1,3,8,34]. Although much is known about
0006-8993 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 01 )02210-7
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341
the membrane properties of serotonergic DRN neurons, no information is available regarding changes in their membrane potential and neuronal excitability during the sleep– waking cycle, largely due to technical difficulties. In a previous paper, we reported the existence of presumed non-monoaminergic PS-off neurons in the medulla and showed that the antidromic responses of these neurons were completely or partially blocked during PS with a prolongation of the antidromic latency, suggesting that inhibitory synaptic volleys impinge on these non-monoaminergic PS-off neurons during PS [33]. However, no information is available about changes in the antidromic responses of putative central monoaminergic neurons during different behavioural states. Recently, we found that the cessation of discharge of putative serotonergic DRN neurons during PS is not caused by g-aminobutyric acid (GABA)-mediated inhibition, but by disfacilitation, resulting from the cessation of discharge of NE- or HAcontaining neurons during PS [30]. Using extracellular recording and the antidromic latency technique, a useful tool for measuring changes in membrane potential [22], we have therefore studied changes in the neuronal excitability of presumed serotonergic DRN neurons during the wake– sleep cycle in the freely-moving cat.
2. Materials and methods A total of nine adult cats of both sexes were anaesthetised with pentobarbital (Nembutal, 25 mg / kg, i.v.) for implantation of the polygraphic recording electrodes using standard techniques, as previously described [10]. In addition, bipolar stimulating electrodes, consisting of two stainless steel wires (200 mm diameter, 1.0 mm apart, bared 0.5 mm at the tip), were placed in a major ascending 5-HT pathway at the level of the ventral tegmental area (VTA: A4.0, L2.0, HC-5.5) and lateral posterior hypothalamus (A9.0, L3.0, HC-4.0) [20] in order to identify antidromic responses of serotonergic DRN neurons [30]. For extracellular single unit recordings, two bundles of six flexible, Formvar-coated, stainless steel wires (32 mm in diameter) were implanted in the DRN through the guide cannulae (24 gauge) of a mechanical microdrive assembly, fixed to the skull at an angle of 508 to the horizontal. For microdialysis application of drugs, one guide cannula (23 gauge) was incorporated into the microdrive assembly between the two microelectrode bundles (1 mm separation between the centres of the guide cannulae), as previously described [31]. After recovery from surgery, the animals were housed individually in sound-attenuated, dimly illuminated electrically-shielded recording chambers maintained at 24– 268C. Food and water were available ad libitum. The cats were habituated to the experimental conditions until they displayed normal sleep–waking cycles. All experimental procedures followed EEC Guidelines (86 / 609 / EEC) and
333
every effort was made to minimise the number of animals used and any pain or discomfort. The polygraphic measurements included neocortical and dorsal hippocampal electroencephalograms (EEG), an electrooculogram (EOG), a neck muscle electromyogram (EMG) and ponto–geniculo–occipital (PGO) waves recorded from the dorsal lateral geniculate nucleus (LGN). The polygraphic data were digitised using a CED 1401 data processor (Cambridge Electronic Design, Cambridge, UK) at a sampling rate of 250 Hz, and stored initially on a personal computer, then on digital tape. Unit recordings were made differentially between active and reference microelectrodes within the same bundle in order to avoid artefacts due to movements of the animal. The unit activities were amplified using a conventional amplifier (Model P15, Grass Instrument, Quincy, USA) with low and high cut-off filters of 100 Hz and 10 kHz, respectively, and were monitored on a storage-oscilloscope (5103N, Tectronix, Les Ulis, France) and on a digital memory oscilloscope equipped with a processor for spike waveform averaging (DRO 1604, Gould Electronique, Longjumeau, France). Single unit activity was separated from background noise to obtain on-line output of the unit discharge on an oscilloscope and on a computer connected to the CED 1401 intelligent interface. The unit activities were digitised using a CED 1401 data processor at a sampling rate of either 20 or 25 kHz and analysed using Spike2 software (CED). Stimulation was performed with square pulses (0.2–0.5 ms), usually delivered at 0.6–0.8 Hz, using a WPI 302 stimulator (World Precision Instruments, Florida, USA) and a stimulus isolation unit (Model 305-2, WPI). The stimulation current was adjusted to a setting that antidromically activated the neuron in 100% of the non-collision trials (suprathreshold current) or was below the threshold for antidromically activating the neuron during W (subthreshold current). For neurons with multiple antidromic latencies, the higher stimulation current eliciting the shortest latency response was used for the threshold current. In the present study on presumed serotonergic DRN neurons showing a low safety factor for antidromic invasion of the soma [5,46,47], we used the collision of spontaneous and stimulation-evoked action potentials (collision test) as the criterion for differentiating between antidromic invasion and synaptic activation (Fig. 1) [15,22]. The antidromic latency was measured from the onset of the stimulus (S) to the middle of the linear portion of the somato-dendritic (SD) component of the spike. Mean latencies were calculated from continuous recordings lasting longer than 1 min using 0.02- or 0.04-ms bins for steady states of W, SWS and PS. The conduction velocity was estimated from the straight line distance between recording and stimulating sites and the shortest antidromic latency. Individual DRN neurons were identified as serotonergic according to previously established criteria [14,20] as
334
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341
propylamino)tetralin hydrobromide (8-OH-DPAT), muscimol and bicuculline methiodide (all from Sigma–Aldrich, Saint Quentin, France). All drugs were dissolved in Ringer’s solution just before use. At the end of the experiments, all animals were deeply anaesthetised, then several unit recording sites were coagulated by passing a 20 mA cathodal current for 10 s and the locations of the recording microelectrode and stimulating electrodes were determined histologically on 25 mm sagittal sections, as previously described [31].
3. Results
Fig. 1. Antidromic responses of a presumed serotonergic DRN neuron following stimulation of the main ascending 5-HT pathway at the level of the ventral tegmental area (VTA). Antidromic action potentials are shown in A–E and a collision with a spontaneous action potential in F (shown by black arrow). Note the marked variability in latency and the lengthening of the antidromic latency when the stimulus was delivered shortly after spontaneous spike at an interval longer than the collision interval. A, five superimposed sweeps. The onset of the stimulus (0.2 ms) is indicated by the black triangle. The vertical dashed line indicates the onset of the shortest antidromic latency.
follows: (1) long-duration (.2 ms) bi- or triphasic action potential, (2) slow (,6 Hz) discharge activity during waking, (3) reduction in spontaneous discharge rate during PS, and (4) histological localisation of recording sites to the DRN. In five cats, a microdialysis probe (Eicom A-L-50-01, Kyoto, Japan) was inserted through the guide cannula into the middle portion of the DRN. The probe membrane was 1 mm in length and had an external diameter of 0.23 mm. Unit recordings and microdialysis infusion of drugs were then carried out for 5–6 consecutive days. During experiments, the probe was continuously perfused at a flow rate of 5 ml / min with either Ringer’s solution or the test drug dissolved in Ringer’s solution, as previously described [31]. The main drugs used were 5-methoxy-N,N-dimethyltryptamine (5-MeODMT), 8-hydroxy-2-(n-di-
A total of 75 presumed serotonergic DRN neurons were activated antidromically by unilateral stimulation of the main serotonergic ascending pathway at the level of the VTA and / or lateral posterior hypothalamus. Of these 75 neurons, 49 were subjected to continuous antidromic stimulation at a slow stimulation rate (,1 / s) during a sleep–waking cycle including at least one PS episode. The results presented were mainly obtained from 32 typical presumed serotonergic DRN neurons which displayed slow regular firing during quiet waking (QW) (mean6S.D.5 2.660.9 spikes / s) and virtually complete suppression of discharge during PS (0.0160.05 spikes / s). The estimated mean conduction velocity (6S.D.) of these neurons was 0.9260.49 m / s (range 0.24–2.50), this slow conduction velocity being indicative of non-myelinated or finely myelinated axons, a characteristic of serotonergic DRN neurons [6]. When stimulated during W, all neurons displayed post-stimulus suppression of spontaneous activity for a period of 0.15–0.20 s. In those neurons tested, complete suppression of spontaneous discharge was seen following systemic (i.m.) administration of either 50–200 mg / kg of 5-MeODMT (n511) or 10–25 mg / kg of 8-OHDPAT (n56). As described for other central monoaminergic neurons [18,25,27,41,42], every presumed serotonergic DRN neuron displayed a large variability in latency (.0.1 ms) when suprathreshold stimulation was applied during W at the slow stimulation rate (Fig. 1). When the neurons were stimulated with twin pulses separated by less than 10 ms, a complete blockage of the SD (or B) spike (Fig. 2A and B) or a marked IS–SD delay (or A–B break) (Fig. 2B) was observed in the second response, which had a longer latency than the first (Fig. 2), a phenomenon commonly seen in slowly conducting, small myelinated and unmyelinated fibre systems, such as presumed catecholaminergic neurons in the locus coeruleus [18] and nucleus commissuralis [25] and presumed cholinergic neurons in the mesopontine tegmentum [32]. In contrast, other central neurons having a high safety factor for antidromic invasion, such as pyramidal tract neurons [38] and presumed non-monoaminergic PS-off neurons in the medulla [33], have been reported to display fixed latency (,0.1 ms) and
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341
335
Fig. 2. Antidromic responses of a presumed serotonergic DRN neuron to double-stimulation of the VTA. Note the complete blockage of the SD spike (A) or the IS–SD break (B) when double stimuli were applied at an inter-pulse interval of less than 10 ms (shown by black arrows). Note also the lengthening of the second response latency. The onset of the stimuli (0.2 ms) are indicated by the black triangles.
faithful response to high frequency (.200 Hz) stimulation (for review, see Ref. [22]). Fig. 3 shows dot displays of antidromic latencies obtained using either suprathreshold (A) or subthreshold (B) stimulation of the main ascending serotonergic pathway during the sleep–waking cycle. Using suprathreshold stimulation, every typical presumed serotonergic DRN neuron (n532) displayed a reduction in the magnitude and variability of the antidromic latency during PS compared with during other behavioural states (Fig. 3A), the latency shift being related to the spontaneous firing rate. Marked shortening of antidromic latency was also seen during SWS in a group of neurons that ceased firing during SWS accompanying PGO waves (Fig. 3A-2). Overall, the shortest and longest latencies were seen during PS and W, respectively, and a gradual shortening of the antidromic latency was seen during PS (Fig. 3). Using subthreshold stimulation, all seven neurons tested showed an increased antidromic responsiveness during PS compared with during other states (Fig. 3B-2), the responsiveness being inversely related to the spontaneous firing rate (Fig. 3B1,2). As previously reported by Gustafsson and Lipski [17], the latency shift (S–SD delay) was mainly due to a change in the delay between the IS and SD spikes (Fig. 4), and, as shown in Fig. 5, the latency was inversely related to the interval between the stimulus and the preceding spontaneous action potential, especially in the case of intervals of less than 1 s. There was a marked decrease in antidromic latency and latency variability when the stimulus was applied 1–2 s after the spontaneous action potential during the period corresponding to deep SWS accompanying PGO waves. Fig. 6 shows the relationship between the magnitude of the latency variation (mean latency during W-mean latency during PS) and the initial
Fig. 3. Antidromic responses of four different presumed serotonergic DRN neurons during wake–sleep states following stimulation of the main ascending 5-HT pathway at the level of the VTA. A and B: suprathreshold (A) and subthreshold (B) stimulation. The upper traces in each record show dot displays of the antidromic latency, while the lower traces show the spontaneous discharge rate (spikes / s) after removal of the antidromic spike responses and stimulus artefacts. Note the decrease in the magnitude and variability of the antidromic latency (A) and the increase in the antidromic responsiveness (B) during PS and the SWS episodes with reduced spontaneous discharge activity. Also note the gradual shortening of the antidromic latency seen during PS. EEG, neocortical electroencephalogram; PGO, dorsal lateral geniculate EEG, showing ponto– geniculo–occipital (PGO) waves.
latency seen during W. No significant correlation was seen with stimulation of the VTA (r50.005) or the lateral posterior hypothalamus (HYP: r50.45). The change in antidromic latency was then studied following systemic (i.m.) or microdialysis application of the 5-HT autoreceptor agonists, 5-MeODMT or 8-OHDPAT, or the GABA agonist, muscimol, all of which induce suppression of spontaneous discharge in serotonergic DRN neurons [12,30,36,43,44]. As shown in the example in Fig. 7B, systemic administration of 50 (n53), 100 (n53) or 200 mg / kg (n53) of 5-MeODMT or 10 (n52) or 25 mg / kg (n51) of 8-OHDPAT led to complete suppression of spontaneous discharge in all typical presumed DRN neurons studied.
336
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341
Fig. 4. Antidromic latency variation during wake–sleep states. A, antidromic responses during W, SWS and PS obtained by signal averaging. The number of samples (n) is indicated in parenthesis. B, antidromic action potentials seen during W (1), SWS (2) and PS (3) (three superimposed sweeps). Note that the change in antidromic latency results from a change in the IS–SD delay. The vertical dotted line represents the onset of IS spike.
Following administration of the 5-HT autoreceptor agonists, the antidromic latency became progressively shortened in parallel with the gradual decrease in spontaneous discharge rate, reaching a plateau at the same time as discharge activity completely ceased. This latency shift was caused by shortening of the IS–SD delay (Fig. 7b), the magnitude of the shift being similar to, or greater than, that seen during a normal PS episode (Fig. 7a). Both the magnitude and variability of the antidromic latency then increased during recovery, paralleling the increase in spontaneous discharge rate (Fig. 7B). When PS occurred during the recovery period, a marked IS–SD delay or complete blockage of the SD component of the antidromic spike was noted in all neurons tested (Fig. 8), a result exactly the opposite of that seen during PS under normal conditions. Three other neurons were tested using microdialysis application of 100 mM muscimol. As illustrated in Fig. 9, muscimol completely suppressed the spontaneous firing of DRN neurons 20 to 30 min after the start of drug application. As with the 5-HT autoreceptor agonists, a decrease in both the magnitude and variability of antidromic latency was seen which paralleled the drug-induced decrease in spontaneous discharge activity. However, using
Fig. 5. Two different examples of the relationship between the antidromic latency and the interval between the stimulus and the preceding spontaneous action potential. Suprathreshold stimulation of the VTA at 0.6 Hz was used. Note that the antidromic latency is inversely related to the interval, especially when the interval is less than 1 s.
muscimol, a marked increase in antidromic latency and latency variability was seen during both PS and the SWS episode immediately preceding it, a result contrasting sharply with that seen under normal conditions.
4. Discussion The antidromic latency technique was introduced by Merrill [24] and has been used in many extracellular studies, including some on sleep [33,37–39]. It is based on the principle that the delay between the stimulus (S) and the SD component of the antidromic spike invariably depends on the soma membrane polarisation, with soma membrane hyperpolarisation increasing the delay and depolarisation shortening it. The main advantage of this technique is that intracellular recordings, which are difficult to realise in behaving animals, can be avoided. Using both intracellular and extracellular recordings, Gustafsson and Lipski [17] showed that changes in the latency to the SD spike (S–SD delay) are mainly caused by changes in the IS–SD delay, which varies from 10 to 100 ms per mV change in membrane potential. The shortening of the antidromic latency, as observed extracellularly, is associ-
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341
Fig. 6. Relationship between the initial antidromic latency seen during W and the maximal latency variation seen during W and PS. No significant correlation was seen using suprathreshold stimulation of either the VTA (upper plot; F51.13, P50.30) or the lateral posterior hypothalamus (HYP) (lower plot; F52.30, P50.16).
ated with excitation or disinhibition of the cell, while its prolongation is related to inhibition or disfacilitation (see Ref. [22] for review). In the present study, we found that, under normal conditions, presumed serotonergic DRN neurons display both a reduction in magnitude and variability of antidromic latency and an increase in antidromic responsiveness during PS compared with during other behavioural states, suggesting an increase in neuronal excitability of presumed serotonergic DRN neurons during PS, even though spontaneous discharge activity is almost completely suppressed during this state. This increase in neuronal excitability seems to be due to the mechanism of disinhibition. Previous intracellular studies in vivo [4] and in vitro [8,11,44,49,50] have shown that typical serotonergic DRN neurons exhibit a resting membrane potential of approximately 260 mV and a large (10–20 mV) and long-lasting ($150–300 ms) AHP after each action potential. The AHP, mediated by a Ca 21 -dependent K 1 conductance [4,11], is followed by gradual inter-spike depolarisation, a plateau period, and then a depolarising pre-potential (LTS) which triggers the succeeding spike. The activation threshold for the LTS is approximately 260 mV [8]. It therefore appears that spontaneously discharging serotonergic DRN
337
neurons are in a hyperpolarised state during most of the inter-spike interval. In the present study, the antidromic latencies of presumed serotonergic DRN neurons were inversely related to the interval between the stimulus and the preceding spontaneous action potential, particularly when the interval was less than 1 s (Fig. 5). When the interval was greater than 1 s, as seen during deep SWS with reduced spontaneous discharge activity, there was a marked reduction in both antidromic latency and latency variability. It therefore seems that the marked prolongation of antidromic latency and variability seen during W and SWS can be ascribed to the prominent post-spike AHP, an intrinsic property of serotonergic DRN neurons. It should be mentioned that gradual shortening of the IS–SD delay was also seen during PS (Fig. 3A), suggesting gradual depolarisation of the soma membrane during this sleep state. It is open to question whether this depolarisation, associated with cessation of spontaneous discharge, represents a restorative process of serotonergic DRN neurons. The present findings appear to be inconsistent with the assumption that the cessation of discharge of serotonergic DRN neurons seen during PS is due to strong inhibition, since no prolongation of antidromic latency or blockage of the SD spike was seen during PS under normal conditions. As briefly described in the Introduction, putative serotonergic DRN neurons selectively cease firing during PS, and it has been postulated that this may result from GABAmediated inhibition and promote PS generation [16,28,29]. However, in agreement with a previous report that the suppression of DRN neuronal activity during PS is unaffected by iontophoretic application of bicuculline, a potent GABA antagonist [21], we recently found that the cessation of discharge of DRN neurons is not attenuated by microdialysis application of bicuculline at concentrations that are able to completely antagonise the potent depressant effect of muscimol, but is completely blocked by either histamine or phenylephrine, a selective a 1 -adrenoceptor agonist. Application of mepyramine, a specific H 1 histamine receptor antagonist, or prazosin, a specific a 1 adrenoceptor antagonist, suppresses the spontaneous discharge of DRN neurons during QW and SWS [30]. In addition, systemic [13] or microdialysis application of WAY100635, a specific 5-HT 1A autoreceptor antagonist, to the DRN [30] does not reverse the cessation of firing of DRN neurons during PS. These findings strongly suggest that this cessation of DRN unit activity during PS is not caused by GABA- or 5-HT-mediated inhibition, but by the mechanism of disfacilitation resulting from the cessation of discharge of NE- or HA-containing neurons during PS. It has previously been shown that spontaneous activity of serotonergic DRN neurons results from noradrenergic drive in the anaesthetised rat [7] and that NE depolarises serotonergic DRN neurons via a 1 -adrenoceptors in vitro by decreasing a resting K 1 conductance [1,49]. In addition, a 1 -adrenoceptor stimulation is reported to suppress the IA , which slows the rate of depolarisation of serotoner-
338
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341
Fig. 7. Effects of administration of 8-OH-DPAT on the antidromic latency of a presumed serotonergic DRN neuron. Left: latency and spontaneous discharge rate in normal control (A) and before, and after, administration of 8-OH-DPAT (10 mg / kg, i.m.) (B). Upper trace, each latency is shown by a dot; lower trace, spontaneous discharge rate (spikes / s) after removal of the antidromic responses and stimulus artefacts. Right: a, superimposed antidromic action potentials (n510) during SWS and PS under normal conditions seen at the times marked 1 and 2 in A. b, Superimposed antidromic action potentials (n510) seen at the times marked 1–4 in B, i.e. (1) shortly before and (2) shortly after 8-OH-DPAT administration, (3) at the time when spontaneous discharge activity was suppressed and (4) during recovery. Note the shortening of the IS–SD delay which was associated with suppression of spontaneous unit discharge seen during PS (a-2) or following 8-OH-DPAT administration (b-3). A, B, a and b, suprathreshold stimulation of the VTA at 0.6 Hz.
gic DRN neurons [1], while the LTS, which accelerates the rate of depolarisation, is not affected by noradrenergic input [8]. It should be noted that the ability of a phasic auditory stimulus to elicit a DRN unit response persists during PS [19], consistent with the idea that the excitability of serotonergic DRN neurons is not reduced during PS. The effects of HA on the membrane potential of serotonergic DRN neurons are not yet known. As during PS, shortening of the antidromic latency was also seen following application of 5-HT autoreceptor agonists or muscimol, a GABA agonist, all of which cause suppression of spontaneous discharge in presumed serotonergic DRN neurons. Previous intracellular recording studies have shown that both 5-HT autoreceptor agonists and GABA induce a dose-dependent hyperpolarisation (,20 mV) in serotonergic DRN neurons [3,8,36,48,50]. 5-HT appears to have no direct effect on the A current or the LTS [8]. 5-HT and LSD appear to block the gradual depolarisation or decay in the post-spike AHP by opening resting K 1 channels with the result that the neurons remain hyperpolarised and lose their rhythmic firing pattern [4,8]. It is possible that the hyperpolarisation induced by these drugs is smaller than that caused by the
post-spike AHP, at least at the concentrations used in the present study and in conscious waking cats, in which serotonergic DRN neurons receive tonic excitatory inputs from NE- or HA-containing, and other, possibly glutamatergic, neurons [30]. It remains to be determined whether larger doses of 5-HT agonists or higher concentrations of muscimol lead to prolongation or blockage of the SD spike in serotonergic DRN neurons during W. A previous intracellular study in anaesthetised rats [4] has shown that administration of LSD produces hyperpolarisation (4–8 mV) in serotonergic DRN neurons, but does not raise the threshold for spiking induced by intracellular depolarising pulses, consistent with the assumption that DRN neurons remain responsive to phasic excitatory volleys when the 5-HT autoreceptor is stimulated. When PS occurred during recovery from 5-HT autoreceptor agonist administration or during muscimol application, a marked IS–SD delay or complete blockage of the SD component of the antidromic spike was seen in every presumed serotonergic DRN neuron tested, suggesting the occurrence of a greater degree of hyperpolarization during the PS episodes. These effects may be attributed to the combination of the inhibition induced by
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341
339
Fig. 9. Antidromic latency change during PS episodes during microdialysis application of muscimol. The upper trace is a dot display of the antidromic latency of a presumed serotonergic DRN neuron following suprathreshold stimulation of the VTA, while the lower trace shows the spontaneous discharge rate (spikes / s), the antidromic spikes and stimulus artefacts being removed from the data. The data were obtained 60 min after the start of application of 100 mM muscimol to the DRN near the unit recording site. Note the virtually complete suppression of spontaneous discharge activity during muscimol application and the marked increase in antidromic latency and latency variability seen during PS. Also note the recovery of spontaneous discharge activity seen during W after microdialysis application of 100 mM bicuculline. The bars under the drug names indicate the drug application period. EEG, neocortical electroencephalogram; EMG, electromyogram; EOG, electrooculogram; PGO, dorsal lateral geniculate EEG, showing ponto–geniculo–occipital (PGO) waves.
Fig. 8. Antidromic latency change during PS following 8-OH-DPAT administration. The upper trace in A is a dot display of the antidromic latency of a presumed serotonergic DRN neuron following suprathreshold stimulation of the VTA, while the lower trace shows the spontaneous discharge rate (spikes / s), the antidromic spike responses and stimulus artefacts being removed from the data. The data were obtained 50–90 min after 8-OH-DPAT (10 mg / kg, i.m.) administration. The mean discharge rate of this neuron during QW was 3.0 spikes / s in control periods. B, superimposed antidromic action potentials at the times indicated as 1–4 in A, i.e. (1) during SWS before PS, (2) during the early phase of PS, (3) during the late phase of PS and (4) during SWS after the PS episode. Note the marked IS–SD break (B-2) or complete blockage of the SD spike (B-3) seen during PS following 8-OH-DPAT administration. The black arrows in B indicate the IS–SD break (B-2) or the blockage of the SD spike (B-3).
activation of 5-HT autoreceptors or GABA receptors and the disfacilitation due to the cessation of discharge of NE or HA neurons during PS. Future intracellular recordings in conscious cats are required to elucidate the exact membrane mechanisms underlying the change in antidromic latency of putative serotonergic DRN neurons during the wake–sleep states. In conclusion, the present findings suggest that, under normal conditions, the neuronal excitability of presumed serotonergic DRN neurons increases during PS mainly due to the mechanism of disinhibition as a result of the
removal of the marked AHP that follows each action potential.
Acknowledgements This work was supported by INSERM U480.
References [1] G.K. Aghajanian, Modulation of a transient outward current in serotonergic neurons by a 1 -adrenoceptors, Nature 315 (1985) 501– 503. [2] G.K. Aghajanian, H.J. Haigler, F.E. Bloom, Lysergic acid diethylamide and serotonin: direct actions on serotonin-containing neurons in rat brain, Life Sci. 11 (1972) 615–622. [3] G.K. Aghajanian, J.M. Lakoski, Hyperpolarization of serotonin neurons by serotonin and LSD: studies in brain slices showing increased K 1 conductance, Brain Res. 305 (1984) 181–185. [4] G.K. Aghajanian, C.P. Vandermaelen, Intracellular recordings from serotonergic dorsal raphe neurons: pacemaker potentials and the effect of LSD, Brain Res. 238 (1982) 463–469. [5] G. Aston-Jones, M. Segal, F. Bloom, Brain aminergic axons exhibit marked variability in conduction velocity, Brain Res. 195 (1980) 215–222. [6] E. Azmitia, P. Gannon, The ultrastructural localization of serotonin immunoreactivity in myelinated and unmyelinated axons within the
340
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20] [21]
[22]
[23] [24]
[25]
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341 medial forebrain bundle of rat and monkey, J. Neurosci. 3 (1983) 2083–2090. J.M. Baraban, G.K. Aghajanian, Suppression of firing activity of 5-HT neurons in the dorsal raphe by alpha-adrenoceptor antagonists, Neuropharmacology 19 (1980) 355–363. T.M. Burlhis, G.K. Aghajanian, Pacemaker potentials of serotonergic dorsal raphe neurons: contribution of a low-threshold Ca 21 conductance, Synapse 1 (1987) 582–588. R. Cespuglio, H. Faradji, M.E. Gomez, M. Jouvet, Single unit recording in the nuclei raphe dorsalis and magnus during the sleep–waking cycle of semi-chronic prepared cats, Neurosci. Lett. 24 (1981) 133–138. S. Crochet, K. Sakai, Effects of microdialysis application of monoamines on the EEG and behavioral states in the cat mesopontine tegmentum, Eur. J. Neurosci. 11 (1999) 3738–3752. V. Crunelli, S. Forda, P.A. Brooks, K.C.P. Wilson, J.C.M. Wise, J.S. Kelly, Passive membrane properties of neurons in the dorsal raphe and periaqueductal grey recorded in vitro, Neurosci. Lett. 40 (1983) 263–268. C.A. Fornal, W.J. Litto, C.W. Metzler, F. Marrosu, K. Tada, B.L. Jacobs, Single-unit responses of serotonergic dorsal raphe neurons to 5-HT 1A agonist and antagonist drug administration in behaving cats, J. Pharmacol. Exp. Ther. 270 (1994) 1345–1358. C.A. Fornal, C.W. Metzler, R.A. Gallegos, S.C. Veasey, A.C. McCreary, B.L. Jacobs, WAY-100635, a potent and selective 5hydroxytryptamine 1A antagonist, increases serotonergic neuronal activity in behaving cats: comparison with (S)-WAY-100135, J. Pharmacol. Exp. Ther. 278 (1996) 752–762. C.A. Fornal, C.W. Metzler, F. Marrosu, L.E. Ribiero-do-Valle, B.L. Jacobs, A subgroup of dorsal raphe serotonergic neurons in the cat is strongly activated during oral–buccal movements, Brain Res. 716 (1996) 123–133. J.H. Fuller, J.D. Schlag, Determination of antidromic excitation by the collision test: problems of interpretation, Brain Res. 112 (1976) 283–298. D. Gervasoni, C. Peyron, C. Rampon, B. Barbagli, G. Chouvet, N. Urbain, P. Fort, P.H. Luppi, Role and origin of the GABAergic innervation of dorsal raphe serotonergic neurons, J. Neurosci. 20 (2000) 4217–4225. B. Gustafsson, J. Lipski, Effect of membrane polarization and synaptic activity on the timing of antidromic invasion, Brain Res. 181 (1980) 61–74. P.G. Guyenet, The coeruleospinal noradrenergic neurons: anatomical and electrophysiological studies in the rat, Brain Res. 189 (1980) 121–133. J. Heym, M.E. Trulson, B.L. Jacobs, Raphe unit activity in freely moving cats: effects of phasic auditory and visual stimuli, Brain Res. 232 (1982) 29–39. B.L. Jacobs, E.C. Azmitia, The structure and function of the brain serotonin system, Physiol. Rev. 72 (1992) 165–229. E.S. Levine, B.L. Jacobs, Neurochemical afferents controlling the activity of serotonergic neurons in the dorsal raphe nucleus: microiontophoretic studies in the awake cat, J. Neurosci. 12 (1992) 4037–4044. J. Lipski, Antidromic activation of neurons as an analytic tool in the study of the central nervous system, J. Neurosci. Meth. 4 (1981) 1–32. D.J. McGinty, R.M. Harper, Dorsal raphe neurons: depression of firing during sleep in cats, Brain Res. 101 (1976) 569–575. E.G. Merrill, Temporal patterns of antidromic invasion latencies for the respiratory neurons of the nucleus retroambigualis in cats, J. Physiol. (Lond.) 223 (1972) 18–20P. S.D. Moore, P.G. Guyenet, An electrophysiological study of the forebrain projection of nucleus commissuralis: preliminary identification of presumed A2 catecholaminergic neurons, Brain Res. 263 (1983) 211–222.
[26] S.B. Mosko, B.L. Jacobs, Recording of dorsal raphe unit activity in vitro, Neurosci. Lett. 2 (1976) 195–200. [27] S. Nakamura, Some electrophysiological properties of neurons of rat locus coeruleus, J. Physiol. 267 (1977) 641–658. [28] D. Nitz, J. Siegel, GABA release in the dorsal raphe nucleus: role in the control of REM sleep, Am. J. Physiol. 273 (1997) R451–R455. [29] C.M. Portas, M. Thakkar, D. Rainnie, R.W. McCarley, Microdialysis perfusion of 8-hydroxy-2-(di-n-propylamino)tetralin (8-OH-DPAT) in the dorsal raphe nucleus decreases serotonin release and increases rapid eye movement sleep in the freely moving cat, J. Neurosci. 16 (1996) 2820–2828. [30] K. Sakai, S. Crochet, Serotonergic dorsal raphe neurons cease firing by disfacilitation during paradoxical sleep, NeuroReport 11 (2000) 3237–3241. [31] K. Sakai, S. Crochet, Role of dorsal raphe neurons in paradoxical sleep generation in the cat: no evidence for a serotonergic mechanism, Eur. J. Neurosci. 13 (2001) 103–112. [32] K. Sakai, M. El Mansari, M. Jouvet, Inhibition by carbachol microinjections of presumptive cholinergic PGO-on neurons in freely moving cats, Brain Res. 527 (1990) 213–223. [33] K. Sakai, N. Kanamori, Are there non-monoaminergic paradoxical sleep-off neurons in the brainstem?, Sleep Res. Online 2 (1999) 57–63. [34] M. Segal, A potent transient outward current regulates excitability of dorsal raphe neurons, Brain Res. 359 (1985) 347–350. [35] K. Shima, H. Nakahama, M. Yamamoto, Firing properties of two types of nucleus raphe dorsalis neurons during the sleep–waking cycle and their responses to sensory stimuli, Brain Res. 399 (1986) 317–326. [36] J.C. Sprouse, G.K. Aghajanian, Electrophysiological responses of serotoninergic dorsal raphe neurons to 5-HT 1A and 5-HT 1B agonists, Synapse 1 (1987) 3–9. [37] M. Steriade, Mechanisms underlying cortical activation: neuronal organization and properties of the midbrain reticular core and intralaminar thalamic nuclei, in: O. Pompeiano, C. Ajmone Marsan (Eds.), Brain Mechanisms and Perceptual Awareness, Raven Press, New York, 1981, pp. 327–377. ˆ [38] M. Steriade, M. Deschenes, Inhibitory processes and interneuronal apparatus in motor cortex during sleep and waking. II. Recurrent and afferent inhibition of pyramidal tract neurons, J. Neurophysiol. 37 (1974) 1093–1113. [39] M. Steriade, J.A. Hobson, Neuronal activity during the sleep– waking cycle, Prog. Neurobiol. 6 (1976) 155–376. [40] M. Steriade, R.W. McCarley, Brainstem Control of Wakefulness and Sleep, Plenum Press, New York, 1990. [41] M. Takigawa, G.J. Mogenson, A study of inputs to antidromically identified neurons of the locus coeruleus, Brain Res. 135 (1977) 217–230. [42] J.M. Tepper, S. Nakamura, S.J. Young, P.M. Groves, Autoreceptormediated changes in dopaminergic terminal excitability: effects of striatal drug infusions, Brain Res. 309 (1984) 317–333. [43] M.E. Trulson, B.L. Jacobs, Raphe unit activity in freely moving cats: correlation with level of behavioral arousal, Brain Res. 163 (1979) 135–150. [44] C.P. Vandermaelen, G.K. Aghajanian, Electrophysiological and pharmacological characterization of serotonergic dorsal raphe neurons recorded extracellularly and intracellularly in rat brain slices, Brain Res. 289 (1983) 109–119. [45] C.P. Vandermaelen, G.K. Matheson, R.C. Wilderman, L.A. Petterson, Inhibition of serotonergic dorsal raphe neurons by systemic and iontophoretic administration of buspiron, a non-benzodiazepine anxiolytic drug, Eur. J. Pharmacol. 1986 (1986) 123–130. [46] R.Y. Wang, G.K. Aghajanian, Collateral inhibition of serotonergic neurons in the rat dorsal raphe nucleus: pharmacological evidence, Neuropharmacology 17 (1978) 819–825. [47] K. Watabe, T. Shinba, T. Satoh, The characteristics of the antidromic
K. Sakai, S. Crochet / Brain Research 898 (2001) 332 – 341 discharge of cat dorsal raphe neurons during repetitive activation, Brain Res. 327 (1985) 231–239. [48] J.T. Williams, W.F. Colmers, Z.Z. Pan, Ligand- and voltage-activated inwardly rectifying currents in dorsal raphe neurons in vitro, J. Neurosci. 8 (1988) 3499–3506. [49] M. Yoshimura, H. Higashi, Noradrenaline mediated slow excitatory
341
synaptic potentials in rat dorsal raphe neurons in vitro, Neurosci. Lett. 61 (1985) 305–310. [50] M. Yoshimura, S. Higashi, 5-Hydroxytryptamine mediates inhibitory postsynaptic potentials in rat dorsal raphe neurons, Neurosci. Lett. 53 (1985) 69–74.