Neuroscience Vol. 38, No. 3, pp. 643-654, 1990 0306-4522/90 $3.00 + 0.00 Printed in Great Britain Pergamon Press plc 0 1990IBRO

SINGLE CHOLINERGIC MESOPONTINE TEGMENTAL PROJECT TO BOTH THE PONTINE AND THE THALAMUS IN THE RAT

K. SEMBA,*~~ P. B. I&I=* and H. C. FIBIGER* *Division of Neurological Sciences, Department of Psychiatry, University of British Columbia, Vancouver, BC, Canada V6T IWS tDepartment of Anatomy, Dalhousie University, Halifax, NS, Canada B3H 4H7

Abstract-Microinjections of the cholinergic agonist carbachol into a caudal part of the pontine reticular formation of the rat induce a rapid eye movement sleep-like state. This carbachol-sensitive region of the pontine reticular formation is innervated by choline@ neurons in the pedunculopontine and laterodorsol @mental nuclei. The same population of cholinergic neurons also project heavily to the thalamus, where there is good evidence that acetylcholine facilitates sensory transmission and blocks rhythmic thalamo- cortical activity. The present study was undertaken to examine the degree to which single choline@ neurons in the mesopontine tegmentum project to both the carbachol-sensitive region of the pontine reticular formation and the thalamus, by combining double fluorescent retrograde tracing and immuno- fluorescence with a monoclonal antibody to choline acetyltransferase in the rat. The results indicated that a subpopulation (5-21% ipsilaterally) of cholinergic neurons in the mesopontine tegmentum projects to both the thalamus and the carbachol-sensitive site of the pontine reticular formation, and these neurons represented the majority (45--@3%) of choline& neurons projecting to the pontine reticular formation site. The percentage of cholinergic neurons with dual projections was higher in the pedunculopontine @mental nucleus (627%) than in the laterodorsal tegmental nucleus (4-I 1%). In addition, mixed with cholinergic neurons in the mesopontine tegmentum, there was a small population of dually projecting neurons that did not appear to be cholinergic. Mesopontine cholinergic neurons with dual projections may simultaneously modulate neuronal activity in the pontine reticular formation and the thalamus, and thereby have the potential of concurrently regulating different aspects of rapid eye movement sleep.

The presence of an ascending cholinergic pathway the intrinsic membrane properties of thalamic arising from the and was first neurons.27-30 suggested by Shute and Lewis? in 1967 on the basis Cholinergic projections arising from the mesopon- of acetylcholinesterase histochemistry combined with tine tegmentum are by no means limited to ascending lesions. These findings have recently been confirmed pathways.45 Descending choline@ pathways also and extended by using immunohistochemistry with exist,4’a and there is good evidence that one such antibodies to choline acetyltransferase (CUT), a pathway is involved in the initiation of rapid eye specific marker for cholinergic neurons, combined movement (REM) sleep. Pharmacological data with retrograde tracing. 13~14,43*45The ascending cholin- suggest that choline@ activation of a certain region ergic projection to the thalamus has received much of the pontine reticular formation (PRF) induces a attention because of evidence indicating that acetyl- REM sleep-like state accompanying muscle atonia, choline has roles in the blockade of rhythmic cortical EEG desynchronization and ponto-geniculo- thalamocortical activity and facilitation of sensory occipital (PGG) spikes. ‘2*“*32*49This region lies medial transmission.27~“~5s Recent studies using in vitro to the motor trigeminal nucleus, and includes a slice preparations have begun to unravel ionic medial part of the nucleus subceruleus (this area is mechanisms underlying these effects in relation to often referred to as the lateral tegmental field,*O although the nucleus subceruleus is not always re- garded as part of it,‘& as well as a lateral part of the $To whom correspondence should be addressed at: Depart- ment of Anatomy, Dalhousie University, Halifax, NS, gigantocellular tegmental field).“J4aJ”~3’~37This region Canada B3H 4H7. of the PRF contains moderate concentrations Abbreviations: ChAT, choline acetyltransferase; FG, of muscarinic, mostly M2, receptors.4”s3 Kainate fluorogold; LDT, laterodorsal @mental nucleus; PGO, lesions in the cholinergic mesopontine tegmentum, ponto-geniculo-occipitah PI, propidium iodide; PPT, which reduce the density of ChAT-immunoreactive pedunculopontine tcgmental nucleus; PRF, pontine reticular formation; REM, rapid eye movement; fibres in both PRF and thalamus,2’ also markedly WGA-HRP, wheat germ agglutinin conjugated with reduce the proportion of time spent in REM sleep in horseradish peroxide. the cat63 (see also Refs 20, 33,42 and 62 for reviews).

643 644 K. SEMEA et al.

The crucial demonstration of cholinergic projections Perfusion to the carbachol-sensitive PRF has recently been Following survival times of 7-12 days, and up to 16 days reported: this PRF region receives a cholinergic in the three rats injected with FG alone, animals were deeply projection from the pedunculopontine (PPT) and anaesthetized with sodium pentobarbital, and perfused transcardially, using 50 ml of 0.01 M phosphate-buffered laterodorsal (LDT) tegmental nuc1ei,39 and this de- saline (pH 7.4) at room temperature, followed by scending pathway derives from IO-15% of cholin- 3Ot-400ml of 4% paraformaldehyde in 0.1 M phosphate ergic neurons in these nuclei in the cat.3’,48 Taken buffer (pH 7.4) at 4°C. Following perfusion, the brain together, these findings are consistent with the notion was removed from the skull and postfixed in the same that the activation of cholinergic input from the fixative for l-2 h at 4°C and then placed in a cryoprotectant solution containing 25% sucrose, 15% glycerol and 0.02% mesopontine tegmentum to the PRF is a key step in sodium azide in 0.05 M phosphate buffer for 2-4 days the initiation of REM sleep. at 4°C. Brains were cut at 30pm on a freezing microtome The above studies suggest that cholinergic neurons into five sets of serial sections in 0.05 M Tris-buffered saline. in the PPT and LDT are involved, on one hand, in One set was used to confirm injection sites in the thalamus and the medial pontine reticular formation, and a second REM sleep initiation through a descending projec- set for additional immunohistochemistry as described tion, and on the other, through an ascending path- below. way, in the modulation of thalamocortical function and the blockade of rhythmic oscillatory activity Immunqjhorescence during certain behavioural states. Do individual sections encompassing the PPT and LDT were cholinergic neurons in the mesopontine tegmentum immunohistochemically processed to visualize tegmental innervate both the carbachol-sensitive region of the cholinergic neurons, using a rat monoclonal antibody PRF, and the thalamus? In the present study this against ChAT, the specificity of which has been previously question was investigated using a combination of documented in detail6 In animals with both FG and PI double fluorescent retrograde tracing and ChAT- injections, a fluorescein isothiocyanate-conjugated anti-rat IgG raised in goat (Jackson Immunoresearch) was used as immunohistochemistry in the rat. Tracer injections in a secondary antibody; in three rats without PI injections, a the PRF were aimed at the carbachol-sensitive site secondary antibody conjugated with Texas Red was used. previously reported in rat” and cat.3’,49 The two secondary antibodies produced the same results. The immunohistochemical procedure has been described in EXPERIMENTAL PROCEDURES detail elsewhere.46 Animals Following immunohistochemical processing, sections Thirteen male Wistar rats (Charles River) weighing were mounted from 15% sucrose in 0.1 M phosphate buffer, 240-310 g were used. air dried, coverslipped with low-fluorescence, microscopic immersion oil, and examined and photographed on a Leitz Tracer injections microscope. Distributions of single-, double- and triple- Animals were anaesthetized with sodium pentobarbital labelled neurons were plotted and counted using five sec- (SOmg/kg, i.p.) for microinjections of two fluorescent trac- tions matched to five standardized sections, at 1.7, I .2, 0.7, ers. Fluorogold (FG, 4% in 0.9% sodium chloride, 0.05 ~1) 0.2 and -0.3 mm from the interaural line, redrawn from was injected into the carbachol-sensitive sites in the Paxinos and Watson.Nx37 These five sections covered the PRF:“.“.49 AP = 9.4-9.6 mm from bregma, ML = 0.8-l .05, entire extent of the cholinergic neurons in the mesopontine D = 9.1-9.3 from the surface of the skull. In 10 of the 13 tegmentum. Cholinergic neurons in the mesopontine teg- rats, an additional two injections were made into the mentum are distributed in a continuum and the boundaries thalamus with propidium iodide (PI, 10% in distilled water, between the PPT and the LDT are not always well defined.45 0.2 ~1 each): (1) AP = 1.8, ML = 0.85-0.9, D = 5.5; and (2) For the purpose of data analyses, the LDT and PPT AP = 3.1, ML = 0.9, D = 5.9, according to the brain atlas of were defined as the parts of the mesopontine tegmentum Paxinos and Watson.36 All tracer injections were conducted populated with cholinergic neurons within and outside. ipsilaterally. respectively, of the central gray. ~~____-.-- -.. ~- Abbreviations used in figures AM anteromedial thalamic nucleus PO postterior thalamic nucleus APT anterior PC pontine reticular formation pars caudalis AV anteroventral thalamic nucleus PnO pontine reticular formation pars oralis CM central medial thalamic nucleus Pr5 principal sensory trigeminal nucleus DR PY pyramidal tract DTg dorsal tegmental nucleus of Gudden RMg raphe magnus f fornix rs Hb habenula Rt reticular thalamic nucleus IC SC ic internal capsule see superior cerebellar peduncle LC locus ceruleus sm stria medullaris LD laterodorsal thalamic nucleus so superior olive LG lateral geniculate nucleus S5 sensory trigeminal nerve LP lateral posterior thalamic nucleus VL ventrolateral thalamic nucleus MD mediodorsal thalamic nucleus VLL ventral nucleus of the ml VM ventromedial thalamic nucleus mlf medial longitudinal fasciculus VP ventroposterior thalamic nucleus Mo5 motor trigeminal nucleus xscp decussation of the superior cerebellar peduncle ZI zona incerta Z parafascicular thalamic nucleus 7n facial nerve Dual projections of cholinergic tegmental neurons 645

RESULTS of the dorsomedial tegmental area, the caudal pon- The data described here are based on eight of the tine reticular nucleus and the dorsal part of the 13 rats, in which tracer injections were localized in nucleus subceruleus (terminology according to and around the carbachol-sensitive site of the PRF Paxinos and Watson37). These injection sites corre- and the thalamus, and in which ChAT immunohisto- sponded well to the region of the PRF where micro- chemistry was acceptable. Data analyses involved injections of carbachol are known to induce a multiple counting and plotting of cells for three REM-like state.“x3’” different fluorescent markers, singly as well as in Minimal deposits of FG were seen in the immediate various combinations. Although attempts were made vicinity of the needle track in three of the eight rats. to reduce fading of fluorescence to a minimum by In rat No. 2, a faint deposit was seen in the most frequent use of the shutter during examination as we11 caudal part of the inferior colliculus, with minimum as by improving coverslipping procedures (see Exper- invasion of the cerebellum, and part of the locus imental Procedures), some fading was inevitable. This ceruleus (Fig. 3). Similarly, in rat No. 3, deposits were resulted in partial data in some rats, because only found through a caudal part of the inferior colliculus ctearly stained material was included in the analyses. and the cerebelIum. In rat No. 8, tracer diffusion was One rat (No. 8) received no thalamic PI injections, seen along the upper part of the needle track in the and only labelling with FG and ChAT immuno- inferior colliculus and the dorsolateral part of the fluorescence was examined. central gray. All these deposits along the needle track were fairly faint in intensity, similar to that seen in “halo” regions around the bright “core” of an injec- FG injections in the PRF in eight rats are shown tion site. Only in rat No. 2, a few ChAT-immuno- in Fig. 1. The injection sites variably involved parts reactive cells were seen close to the tracer deposit

Fig. 1. The locations of FG injections in the pontine recticular formation in eight rats. In (B)--(E) vertical hatching indicates the intensely bright “core” regions of individual injection sites in odd-nabered rats and horizontal hatching indicates those in even-numbered rats. (B) Rat Nos 1 and 2. (C) Rat Nos 3 and 4. (D) Rat Nos 5 and 6. (E) Rat Nos 7 and 8. The numbers in (A) indicate distances in mm from the interaural line. 646 K. SEMBA er ai. along needle track at the lateral aspect of the rats was diffusion from the deposit sufficiently close caudalmost LDT. Because it was felt impossible to or extensive to encroach ChAT-immunoreactive determine if FG labefling of these particular cells was neurons in the mesopontine tegmentum. due to damage, for example, of distal dendrites, or As iflustrated in Fig. 2, two thalamic PI injections retrograde label!ing from the injection site, these cells in each of eight rats consistently included the were excluded from the present analyses. In no other mediodorsai nucleus and the habenula, as well as

A I3 D E

Fig. 2. The locations of PI injections in the thalamus in Seven rats. In each rat, two injections were made. Only intensely labelled “core” regions of individual in&xXionsites are indicated by hatching. In (B)-(E), vertical hatching represents the injection sites in odd-numbered rats, and horizontal hatching represents those in even-numbered rats. (B) Rat Nos 1 and 2. (C) Rat Nos 3 and 4. (D) Rat Nos 5 and 6. (E) Rat No. 7. The injection sites of rat No. 8 were confhmed but the sections were subsequently lost. The numbers in (A) indicate the distance in mm from the interaural line. Dual pro~tions of cholinergic tegmental neurons 647 parts of the intralaminar nuclei. In addition, the (n =: 3). Examples of triple labelled neurons are injections included parts of the anterior, ventral shown in Fig. 4. and/or lateral cell groups of the thalamus. In most of Cholinergic neurons in the mesopontine tegmen- the eight rats, injection sites also included a small part turn are distributed more or less in a continuum, of the fimbria and/or hipp~ampus, dorsally, and constituting the PPT rostrally and the LDT caudaily, part of the pretectal region, caudally. with the two nuclei co-existent at intermediate to caudal levels. However, the boundaries between the Choline acetyltransferase-immunoreactive neurons PPT and the LDT are not always obvious.4s To test projecting to zhe pontine reticular formation the possibility that the aforementioned differences At the level of the pons, neurons retrogradely between the PPT and LDT in the percentages of labelled with FG from the PRF were distributed in double and triple labelled neurons are related to the both the rostra1 and the caudal parts of the reticular Ievels along the Ion~tudinal axis, the two nuclei were formation, the central gray, the superior colliculus, combined, and percentages of triple labetled neurons the cuneiform nucleus, the nucleus of the lateral were compared as a function of rostrocaudal levels. lemniscus, and the (Fig. 3). In the Five equally spaced sections were examined which PPT/LDT ipsilateral to the injection, 1 l-22% (n = 4) covered the entire extent of cholinergic neurons in of CbAT-immunoreactive neurons were retrogradely the tegmentum (see Experimental Procedures). The labelled with FG from the PRF (Table 1). The results indicated that while triple 1a~IIed neurons PPT contained about twice as great a percentage of tended to be lower in percentage at the most caudal FG-labelled, ChAT-immunoreactive neurons as that level in two rats (Nos 1 and 2), no obvious topogra- in the LDT in three rats (Nos 1, 2 and 8). In a fourth phy along the rostrocaudal axis was seen in the other rat (No. 3), almost all double labelled neurons were rats (ranges for the five rostral to caudal levels in five found in the PPT, with only a few in the LDT. The rats; ipsilateral: O-29%, O-33%, 9-41%, 5-22%, FG injection in this rat was placed slightly more O-20%; contralatera1: O-15%, O-20%, 2-20%, caudal compared with the other three cases. l-15%, O-8%). Contralaterally, ChAT-immunoreactive neurons Neurons retrogradely labelled from both the thala- labelled with FG from the PRF represented 6-16% mus and PRF were rarely seen outside the PPT/LDT of all ChAT-immunoreactive neurons, a smaller per- (Fig. 3). Occasionally, however, neurons that con- centage compared to the ipsilateral side (Table 1). tained both of the retrograde markers, but lacked The ratios between the PPT and LDT were similar to ChAT immuno~activity, were seen mixed with those found on the ipsilateral side, with higher per- ChAT-immunoreactive neurons in the PPT] centages in the PPT (8-18%) than the LDT (l-12%). LDT. To examine the percentage of these dually projecting, but ChAT-negative PPT/LDT neurons, Choline acetyltransferase-immunoreactive neurons the PPT/LDT was first defined and demarcated in projecting to the thalamus section drawings by the presence of ChAT-immuno- Mesopontine tegmental neurons retrogradely reactive neurons. Then, neurons labelled with FG labelled with PI from the thalamus accounted for from the PRF within the demarcated PPT/LDT were 27-33% of the ChAT-immunoreactive neurons ipsi- examined for the presence of PI from the thalamus laterally, and S-15% contralaterally (n = 3, Table 1). and ChAT immunoreactivity. The percentages of these double labelled neurons in Of all PRF-projecting neurons in the PPT/LDT, the PPT (30-46%) were greater in the PPT than the 4-9% were retrogradely labelled from the ipsilateral LDT (19-26%) ipsilaterally; contralaterally the per- thalamus with PI (n = 5, Table 1). Triple labelled centages were similar (6-14% and 13-16% for the neurons accounted for 2-6% of the total PRF- PPT and the LDT, respectively). projecting cell population in the PPT/LDT. Thus, it could be derived that 50-100% of the dually project- Choline acetyltransferase-immunoreactive neurons ing PPT/LDT neurons were ChAT-immunoreactive, projecting to both pontine reticular formation and and that O-50% were ChATnegative. ChAT- thalamus immunoreactive neurons accounted for 3-l 1% of ChAT-immunoreactive PPT/LDT neurons that PRF-projecting neurons in the same regions (n = 6). were retrogradely labelled from both PRF and Generally lower percentages were obtained contra- thalamus represented 5-21% of the total ChAT- laterally for comparable cell populations. immunoreactive cell populations in the ipsilateral mesopontine tegmentum, whereas they represented DISCUSSION l-14% on the contralateral side (n = 5, Table 1). Again, higher percentages of these triple labelled The choline@ neurons in the PPT and LDT have neurons were seen in the PFT (6-27%) than the LDT previously been shown to give rise to a major ascend- (3-11%) ipsilatecally. This was also the case con- ing projection to the thalamus, and an apparently tralaterally (2-24% vs O-4%). The triple labelled smaller, descending projection to the carbachol- neurons accounted for 4588% of ChAT-immune- sensitive PRF region (see the introduction for refer- reactive neurons retrogradely labelled from the PRF ences). In the present study, it was shown that: (1) a 648 K. SEMBA et al.

-0.8

- PRF A ChAT l PF+F+ChAT . PRF+Thalamue 0 ChAT+Thaiamus * ~+~~+C~T

Fig. 3. An example (rat No. 2) of distributions of single, double and triple labelled neurons with retrograde labelhng with FG from the PRF and/or with PI from the thalamus, and ChAT immunoreac- tivity. The numbers indicate distances in mm from the interaural line. Each symbol represents one cell, except for FG-labelled and ChATpositive neurons for which one symbol represents five cells. Stippling indicates the injection site. Dual projections of cholinergic tegmental neurons 649

Table I. Percentages of double and triple labelled neurons in the pedunculopontine tegmental nucleus/laterodorsal tegmental nucleus following fluorogold injections into the carbachol-sensitive region of the pontine reticular formation and propidium iodide injections into the thalamus, and immunohistochemistry for choline acetyltransferase in eight rats %CbAT+ neurons %PRF-proj~ting neurons Projecting to Projecting to Projecting to Projecting to PRF and Projecting to thalamus and Rat PRF thalamus thalamus thalamus ChAT+ ChAT+ Ipsilateral 1 11(13/7) 27 (32/20) 5 (6/4) 7 11 5 2 22 (27113) 29 (30/26) 14 (19/4) 6 9 6 3 17(31/l) 33 (46/19) 15 (25/3) 9 7 6 4 - - 21(27/l 1) - - - 5 - 13 (18/Q - - - 6 - - 6 5 4 - 4 3 2 : &26,!14) - - 10 - Contralateral 15 (14/E) 1 fwv 4 8 2 : 1~~~~) 8 (6/13) 5 W4) 5 11 5 3 7 (9/5) 14(14/14) 5 (7/3) 6 4 3 4 - - 14 (2412) - - - 7 (12/3) - - - - 3 3 2 - 2 2 1 - - 16 - Tbe numbers in parentheses indicate respective numbers for the PPT and LDT for ChATpositive neurons. PRF-projecting neurons were not further divided because of relatively small percentages. subpopulation of cholinergic tegmental neurons pro- mean = 24%) than in the LDT (l-14%, mean = 9%, ject to both the carbachol-~nsitive PRF site and the Table 1). The difference appears to derive primarily thalamus; (2) a majority of cholinergic tegmental from the higher percentages in the PPT in the present, neurons projecting to the PRF site also innervate the as compared to the previous studies; in the LDT the thalamus; (3) the majority of dually projecting neur- percentages are rather similar between the present ons in the PPT/LDT regions are cholinergic; and (4) and the previous studies. One possibility is that this the number of cholinergic tegmental neurons project- is due to a species difference in the organization of ing to the carbachol-sensitive PRF region appears to mesopontine cholinergic projections. Thalamic pro- be smaller than that of those projecting to the jections from the meso~n~ne cholinergic neurons thalamus. appear to be generally similar between rat and cat The percentage of cholinergic PPT/LDT neurons (see review by Semba and Fibiger).45 projecting to the carbachol-sensitive PRF region A second possibility, which is not mutually exclu- ranged from 11 to 22%. The variability most likely sive with the first, is that the difference is due to reflects differences in the size of injections as well as different tracers used, WGA-HRP vs FG, and related the location of the injections in both the thalamus problems of uptake by ftbres of passage. This is and the PRF; different degrees of fading of fluor- potentially relevant because cholinergic neurons in escent markers might also have contributed to it. The the mesopontine tegmentum, predominantly those in present range of PRF-projecting cholinergic neurons the PPT (reportedly about 18% of the cholinergic cell is slightly higher than those previously reported, population therein), have been shown to project 10-1S??48 and S-IO%,” which were both based on caudally to innervate the medullary reticular for- wheat germ a~lutinin conjugated with horseradish mation in the rat,4ie*65and one major pathway for this peroxidase (WGA-HRP) retrograde transport com- projection is identified as the ventromedial branch of bined with ChAT immunohistochemistry in the cat. Probst’s tract,4’a which apparently courses near the The major difference, however, lies in the topography carbachol-sensitive PRF region. Although collateral- of PRF-projecting cholinergic neurons. In both previ- ization of cholinergic axons in the medulla has been ous studies using the cat, lower percentages of PRF- suggested to be considerable,4’” that between the projecting choline@ neurons were found in the PPT medullary and the pontine reticular formation re- compared with the LDT (10% vs 15% by Shiromani mains to be investigated. Depending on the degree of et a1.,4a and means of 5.2% vs 10.2% by Mitani et this particular pattern of collateralization, the uptake a1.).3’ In contradistinction, in the present study using by fibres of passage, if there were any, should or the rat, cholinergic neurons innervating the PRF should not affect the number of cells which would appeared to be more frequent in the PPT (13-31%, be labelled by tracer injections into the carbachol- 650 I(. SEMBA et d.

Fig. 4. Examples of neurons (arrows) in the PPT which are retrogradely Iabelled from the PRF with FG (A), ChAT-immunoreactive (B) and retrogradely labelled from the thalamus with PI (C). An example of a which is retrogradely labelled from the thalamus and immunoreactive for ChAT, but not labelled from the PRF, is indicated with arrowheads. In this particular set of micrographs, almost all PI-labelled neurons appear to be ChAT-immunoreactive. Scale bar = JO pm. Dual projections of choline@ tegmental neurons 651

sensitive PRF site. It remains to be clarified whether ing cholinergic fibres. 4ia Collateral projections to the the difference in the topography of PRF-projecting posterior thalamus and the raphe magnus have re- neurons between the present and the previous studies cently been reported.6s Taken together, these findings is due to different species or the type of tracers used. indicate that the axons of cholinergic neurons in the Twenty-seven to thirty-three per cent of all ChAT- m~opontine te~ent~ do collateralize to innervate immunoreactive neurons in the PPT/LDT were different target structures, although apparently not found to project to the thalamus ipsilaterally, and as extensively as reported with monoaminergic 8-15% contralaterally. These ranges are most likely axons.’ underestimated because tracer injections did not in- One important implication of the present findings volve all the thalamic nuclei. The present percentages is that some cholinergic neurons may simultaneously are lower than those previously reported following intluence postsynaptic neurons in the carbachol- large HRP injections into the thalamus (about sensitive PRF region and the thalamus. Thus, these 60%),” but comparable to the ranges of l-26% for neurons may be involved in both the initiation of the PPT and 3-47% for the LDT following REM sleep, through descending projections, and WGA-HRP injections into selected thalamic nuclei blockade of rhythmic thalamocortical activity and in the rat (quoted by Mitani et a1.,31as obtained from modulation of sensory transmission, via ascending one of the authors from the data of Hallanger et al.13). projections. Thalamocortical oscillation is blocked in Similar results may be seen in the ill~tra~ons of both waking (except during ~-spindles3,47) and REM comparable data obtained from rat” and cat.“s1+s6 sleep. One possibility, therefore, is that these dual The variation seen in the present and previous projection cholinergic neurons are involved in the studies most likely reflects differences in the size of blockade of rhythmic thalamocortical activity during tracer injections, as well as the tracer used. An REM sleep but not during waking. Using chronic additional contributing factor would be the regional single unit recordings from neurons in the LDT in the differences in the density of choline& fibres within cat, El Mansari et ai.’ described a sub~pulation the thalamus.26 (12%) of LDT neurons which were selectively active Because in the present study the tracer injections in during REM sleep but not during waking. The same the thalamus always involved several thalamic nuclei, authors also reported on additional subpopulations a question remains as to whether the axons of these of LDT neurons which became active during both doubly projecting neurons terminate selectively waking and REM sleep or during waking alone. within the thalamus. This would be an interesting Thus, it is possible that the dual projection neurons question in relation to the role of this projection in in the present study correspond to those LDT neur- the state-dependent regulation of thalamocortical ons that fire only during REM sleep. activity (see below). The key observation relevant to An alternative hypothesis is that mesopontine teg- this question is the previous report that cholinergic mental cholinergic neurons projecting to both thala- neurons in the PPT project to all thalamic nuclei, mus and PRF are active during waking and REM whereas those in the LDT innervate predominantly sleep. According to this scenario, additional neuronal limbic parts of the thalamus.‘3*U~4s*S6Because ChAT- concomitants during REM sleep (but not during positive neurons with dual projections were more waking) permit the cholinergic input to evoIve the set frequent in the PPT than in the LDT in the present of behaviours seen with pharmacological application study, it is likely that doubly projecting choline@ of cholinergic agonists to the PRF. In particular, it is neurons, as a cell population, innervate all thalamic well documented that noradrenergic, serotonergic nuclei. This hypothesis, however, remains to be tested and histaminergic neurons are all active during wak- directly. ing but cease firing during REM ~leep,‘~~““*and these The present study had added to the evidence of three transmitters have been shown to affect neurons axonal collateralization of choline@ mesopontine in the carbachol-sensitive region of the PRF,9J0,s8 tegmental neurons. In our previous study using the To the cholinergic agonist carbachol, PRF neurons rat, about 10% of the neurons in the mesopontine respond with depolarization with a decrease in con- tegmentum that were retrogradely labelled from the ductance,i2 which parallels the membrane potential reticular thalamic nucleus were also labelled from the changes seen in natural REM sleep.‘* Thus, it would cortex, these being located mostly in the PPT; many be considerable interest to examine the response of of these dual projection neurons were also ChAT- PRF neurons to acetylcholine in the presence and immunoreactive.22 There is also some evidence to absence of one or more of these amine transmitters. suggest that the collateralization of cholinergic fibres The significance of the mesopontine cholinergic occurs within the thalamu? and between other fore- input in thalamocortical functions has begun to brain regions.@ Axonal ~llaterali~tion of tegmentai be understood at the cellular level using both cholinergic neurons involving the innervation of the in uim ‘s~‘7~23~24*5sand in vitru preparations.27-M These has also been reported;44@ however, the studies support the notion that with differential cholinergic projections to the extrapyramidal system effects on principal relay neurons and inter- have been controversial. 4s Collateralization within neurons, acetylcholine has an overall facilitatory medullary regions have been suggested for descend- effect on sensory transmission along with blockade of 652 K. SEMBAet al. rhythmic oscillation. Additionally, the thalamic retic- innervation, the thalamus receives aminergic inputs ular neurons, through their widespread yet topo- as well.27,55Thus, again, it would be important to graphically organized GABAergic projections to investigate the effects of co-activation of multiple other thalamic nuclei, have an important role in inputs in thalamic neurons. The functional signifi- regulating rhythmic thalamocortical activity.54,S5 The cance of the present findings of axonal collateraliz- effects of acetylcholine on thalamic reticular neurons, ation of some tegmental cholinergic neurons would however, are at present controversial.‘5.23,28 Further- be best understood in view of the data regarding all more. the activity of thalamic reticular neurons, in these cellular events which, in concert, regulate state- turn, appears to be regulated by an additional cholin- dependent thalamocortical rhythmicity and sensory ergic13.22.25.35.5land a GABAergic?’ input from the transmission. basal forebrain, which has been suggested to play a role in cortical EEG desynchronization.3 In the basal Acknowledgements-We thank Felix Eckenstein for a gener- ous gift of monoclonal antibody to ChAT, and excellent forebrain of the cat, both waking/REM sleep-active technical assistance by Chui-Se Tham. This work was and slow wave sleepactive neurons have been supported by the Medical Research Council of Canada. recorded.5.5y~6” Finally, in addition to the cholinergic KS. and P.B.R. are MRC Scholars.

REFERENCES

I. Baghdoyan H. A., Rodrigo-Angulo M. L., McCarley R. W. and Hobson J. A. (1984) Site-specific enhancement and suppression of desynchronized sleep signs following cholinergic stimulation of three brainstem regions. Brain Res. 306, 39-52. 2. Baxter B. L. (1969) Induction of both emotional behavior and a novel form of REM sleep by chemical stimulation applied to cat mesencephalon. Expl Neurol. 23, 220-229. 3. Buzsaki G., Bickford R. G., Ponomareff G., Thal L. J., Mandel R. and Gage F. H. (1988) Nucleus basahs and thalamic control of neocortical activity in the freely moving rat. J. Neurosci. 8, 4007-4026. 4. Cortes R. and Palacios J. M. (1986) Muscarinic cholinergic receptor subtypes in the rat brain. I. Quantitative autoradiographic studies. Brain Res. 362, 227-238. - . 5. Detari L.. Juhasz G. and Kukorelli T. (1987) Neuronal firing in the nahidal region: tirine natterns durine sleep-wakefulness cycle in cats. Electroenceph. chn. Neurophysiol. 167, 1599186. - _ 1 6. Eckenstein F. and Thoenen H. (1982) Production of specific antisera and monoclonal antibodies to choline acetyltransferase: characterization and use for identification of cholinergic neurons. Eur. molec. Biol. Org. J. 1, 363-368. 7. El Mansari M., Sakai K. and Jouvet M. (1989) Unitary characteristics of presumptive cholinergic tegmental neurons during the sleepwaking cycle in freely moving cats. Expl Brain Res. 76, 519-529. 8. Foote S. L. and Morrison J. H. (1987) Extrathalamic modulation of cortical function. A. Rev. Neurosci. 10, 67-95. 9. Gerber U., Greene R. W., Haas H. L. and McCarley R. W. (1989) Receptor characterization of adrenergic effects in the medial pontine reticular formation in vitro. Sot. Neurosci. Abstr. 15, 194. 10. Gerber U., Greene R. W., Haas H. L. and McCarley R. W. (1990) Histamine enhances excitability of medial pontine recticular formation neurons in rat. In Sleep Research (eds Chase M. H., Lydic R. and O’Connor C.), Vol. 19, p. 14. Brain Information Service/Brain Research Institute, University of California Los Angeles, Los Angeles. 11. Gnadt J. W. and Pegram G. V. (1986) Cholinergic brainstem mechanisms of REM sleep in the rat. Brain Res. 384, 29-41. 12. Greene R. W., Gerber U. and McCarley R. W. (1989) Cholinergic activation of medial pontine recticular formation neurons in vitro. Brain Res. 476, 1544159. 13. Hallanger A. E., Levey A. I., Lee H. J., Rye D. B. and Wainer B. H. (1987) The origin of cholinergic and other subcortical afferents to the thalamus in the rat. J. camp. Neural. X2, 105-124. 14. Hallanger A. E. and Wainer B. H. (1988) Ascending projections from the pedunculopontine tegmental nucleus and the adjacent mesopontine tegmentum in the rat. J. camp. Neurol. 274, 483-515. 14a. Holstege G. and Kuypers H. G. J. M. (1977) Propriobulbar fibre connections to the trigeminal, facial and hypoglossal motor nuclei. Brain 100, 239-264. 15. Hu B., Steriade M. and Deschenes M. (1989) The effects of brainstem peribrachial stimulation on perigeniculate neurons: the blockade of spindle waves. Neuroscience 31, 1-12. 16. Hu B., Steriade M. and Deschenes M. (1989) The effects of brainstem peribrachial stimulation on neurons of the lateral geniculate nucleus. Neuroscience 31, 13-24. 17. Hu B., Steriade M. and Deschenes M. (1989) The cellular mechanism of thalamic ponto-geniculo-occipital waves. Neuroscience 31, 25-35. 18. Ito K. and McCarley R. W. (1984) Alterations in membrane potential and excitability of cat medial pontine reticular formation neurons during changes in naturally occurring sleepwake states. Brain Res. 292, 1699175. 19. Jacobs B. L. (1987) Brain monoaminergic unit activity in behaving animals. Prog. Psychobiol. Physiol. Psychol. 12, 171-206. 20. Jones B. E. (1985) Neuroanatomical and neurochemical substrates of mechanisms underlying paradoxical sleep. In Brain Mechanisms ofSleep (eds McGinty D. J. er a/.), pp. 139-156. Raven Press, New York. 21. Jones B. E. and Webster H. H. (1988) Neurotoxic lesions of the dorsolateral pontomesencephalic tegmen- tum-cholinergic cell area in the cat. I. Effects upon the cholinergic innervation of the brain. Brain Res. 451, 13-32. 22. Jourdain A., Semba K. and Fibiger H. C. (1989) Basal forebrain and mesopontine tegmental projections to the reticular thalamic nucleus: an axonal collateralization and immunohistochemical study in the rat. Brain Res. 505, 55-65. 23. Kayama Y., Sumitomo I. and Ogawa T. (1986) Does the ascending cholinergic projection inhibit or excite neurons in the rat thalamic reticular nucleus? J. Neurophysiol. 56, 1310-1320. Dual projections of cholinergic tegmental neurons 653

24. Kayama Y., Takagi M. and Ogawa T. (1986) Cholinergic influence of the laterodorsal tegmental nucleus on neUrOna activity in the rat lateral geniculate nucleus. Neuroscience 56, 1297-1309. 25. Levey A. I., Hallanger A. E. and Wainer B. H. (1987) Cholinergic neurons may influence the cortex via the thalamus. Neurosci. Left. 74, 7~13. 26. Levey A. I., Hallanger A. E. and Wainer B. H. (1987) Choline acetyltransferase immunoreactivity in the rat thalamus. J. camp. Neurol. 257, 317-332. 27. McCormick D. A. (1989) Cholinergic and noradrenergic modulation of thalamocortical processing. Trends Neurosci. 12, 215-221. 28. McCormick D. A. and Prince D. A. (1986) Acetylcholine induces burst firing in thalamic reticular neurones by activating a potassium conductance. Nature 319, 402-405. 29. McCormick D. A. and Prince D. A. (1987) Actions of acetylcholine in the guinea-pig and cat medial and lateral geniculate nuclei, in vitro. J. Physiol., Land. 392, 147-165. 30. McCormick D. A. and Pape H.-C, (1988) Acetylcholine inhibits identified interneurons in the cat lateral geniculate nucleus. Nature 334, 246248. 31. Mitani A., Ito K., Hallanger A. E., Wainer B. H., Kataoka K. and McCarley R. W. (1988) Choline@ projections from the laterodorsal and pedunculopontine tegmental nuclei to the pontine gigantocellular tegmental field in the cat. Brain Res. 451, 397-402. 32. Mitler M. M. and Dement W. C. (1974) Cataplectic-like behavior in cats after microinjections of carbachol in pontine reticular formation. Brain Res. 68, 335-343. 33. Morrison A. R. and Reiner P. B. (1985) A dissection of paradoxical sleep. In Brain Mechanisms of Sleep (eds McGinty D. J. et al.), pp. 97-109. Raven Press, New York. 34. Pare D., Smith Y., Parent A. and Steriade M. (1988) Projections of brainstem core cholinergic and non-cholinergic neurons of cat to intralaminar and reticular thalamic nuclei. Neuroscience 25, 69986. 35. Parent A., Pare D., Smith Y. and Steriade M. (1988) Basal forebrain cholinergic and noncholinergic projections to the thalamus and brainstem in cats and monkeys. J. camp. Neural. 277, 281-301. 36. Paxinos G. and Watson C. (1982) The Rat Brain in Stereotoxic Coordinates. Academic Press, Sydney. 37. Paxinos G. and Watson C. (1986) The Rat Brain in Stereotaxic Coordinates, 2nd edn. Academic Press, Sydney. 38. Porter L. L. and Asanuma C. (1989) Ultrastructural and immunohistochemical observations on a projection from the magnocellular basal nucleus in rats. Sot. Neurosci. Abstr. 15, 289. 39. Quattrochi J. J., Mamelak A. N., Madison R. D., Mcklis J. D. and Hobson A. (1989) Mapping neuronal inputs to REM sleep induction sites with carbachol-fluorescent microspheres. Science 245, 9844986. 40. Quirion R., Araujo D., Regenold W. and Boksa P. (1989) Characterization and quantitative autoradiographic distribution of [3H]acetylcholine muscarinic receptors in mammalian brain. Apparent labelling of an MZlike receptor sub-type. Neuroscience 29, 271-289. 41. Reiner P. B. (1986) Correlational analysis of central noradrenergic neronal activity and sympathetic tone in behaving cats. Brain Res. 378, 86-96. 41a. Rve D. B.. Lee H. J.. Saoer C. B. and Wainer B. H. (1988) Medullarv and smnal etferents of the oedunculoDontme tegmental nucleus and adiacent mesopontine tegmentum in’the rat. J.-camp. -NemoI. 269, 3155341: _ 42. Sakai K. (1987) Anatomical and physiological basis of paradoxical sleep. In Brain Mechanisms of Sleep (eds McGinty D. J. et al.), pp. 11l-137. Raven Press, New York. 43. Satoh K. and Fibiger H. C. (1986) Cholinergic neurons of the laterodorsal tegmental nucleus: efferent and afferent connections. J. camp. Neurol. 253, 277-302. 44. Scarnati E., Gasbarri A., Campana E. and Pacitti C. (1987) The organization of nucleus tegmenti pedunculopontinus neurons projecting to basal ganglia and thalamus: a retrograde fluorescent double labeling study in the rat. Neurosci. Lett. 79, 11-16. 45. Semba K. and Fibiger H. C. (1989) Organization of central cholinergic systems. Prog. Brain Res. 79, 37-63. 46. Semba K., Reiner P. B., McGeer E. G. and Fibiger H. C. (1988) Brainstem afferents to the magnocellular basal fore- brain studied by axonal transport, immunohistochemistry, and electrophysiology in the rat. J. camp. Neural. 267, 4333453. 47. Semba K., Szechtman H. and Komisaruk B. R. (1980) Synchrony among rhythmical facial tremor, neocortical ‘alpha’ waves, and thalamic non-sensory neuronal bursts in intact awake rats. Brain Res. 195, 281-298. 48. Shiromani P. J., Armstrong D. M. and Gillin J. C. (1988) Cholinergic neurons from the dorsolateral pons project to the medial pons: a WGA-HRP and choline acetyltransferase immunohistochemical study. Neurosci. Left. 95, 19-23. 49. Shiromani P. J. and McGinty D. J. (1986) Pontine neuronal response to local cholinergic infusion: relation to REM sleep. Brain Res. 386, 20-3 1. 50. Shute C. C. D. and Lewis P. R. (1967) The ascending cholinergic reticular system: neocortical, olfactory and subcortical projections. Brain 90, 497-520. 51. Smith T., Pare D., Deschenes M., Parent A. and Steriade M. (1988) Cholinergic and non-choline@ projections from the upper brainstem core to the visual thalamus in the cat. Expl Brain Res. 70, 166-180. 52. Sofroniew M. V., Priestley J. V., Consolazione A., Eckenstein F. and Cue110A. C. (1985) Cholinergic projections from the midbrain and pons to the thalamus in the rat, identified by combined retrograde tracing and choline acetyltrans- ferase immunohistochemistry. Brain Res. 329, 213-223. 53. Spencer D. G. Jr, Horvath E. and Traber J. (1986) Direct autoradiographic determination of Ml and M2 muscarinic acetylcholine receptor distribution in the rat brain: relation to cholinergic nuclei and projections. Brain Res. 380, 59-68. 54. Steriade M. and Deschenes M. (1984) The thalamus as a neuronal oscillator. Brain Res. Rev. 8, l-63. 55. Steriade M. and Llinas R. R. (1988) The functional states of the thalamus and the associated neuronal interplay. Physiol. Rev. 68, 649-742. 56. Steriade M., Pare D., Parent A. and Smith Y. (1988) Projections of cholinergic and non-cholinergic neurons of the brainstem core to relay and associational thalamic nuclei in the cat and macaque monkey. Neuroscience 25, 47-67. 57. Steriade M., Parent A., Pare D. and Smith Y. (1987) Cholinergic and non-choline&c neurons of cat basal forebrain project to reticular and mediodorsal thalamic nuclei. Brain Res. 40& 372-376. 58. Stevens D. R., Greene R. W. and McCarley R. W. (1989) Excitatory and inhibitory actions of sertonin on neurons of the rat medial pontine reticular formation in vifro. Sot. Neurosci. Absrr. 15, 5. 654 K. SEMBAet al

59. Szymusiak R. and McGinty D. (1986) Sleep-related neuronal discharge in the basal forebrain of cats. Brain Res. 370, 82-92. 60. Szymusiak R. and McGinty D. (1989) Sleepwaking discharge of basal forebrain projection neurons in cats. Brain Res. Bull. 22, 423-430. 61. Vanni-Mercier G., Sakai K. and Jouvet M. (1984) ‘Waking-state specific’ neurons in the caudal hypothalamus of the cat. C. r. hebd. Skanc. Acad. Sci., Paris 298, 195-220. 62. Vertes T. P. (1984) Brainstem control of the events of REM sleep. Prog. Neurobiol. 22, 241-288. 63. Webster H. H. and Jones B. E. (1988) Neurotoxic lesions of the dorsolateral pontomesencephalic tegmentum- cholinergic cell area in the cat. 11. Effects upon sleepwaking states. Brain Res. 452I, 285-302. 64. Woolf N. J. and Butcher L. L. (1986) Cholinergic systems in the rat brain. III. Projections from the pontomesencephalic tegmentum to the thalamus, tectum, basal ganglia, and basal forebrain. Brain Res. Btdi. 16, 603-637. 65. Woolf N. J. and Butcher L. L. (1989) Cholinergic systems in the rat brain. IV. Descending projections of the pontomesencephalic tegmentum. Brain Res. Bull. 23, 519-540.

(Accepted 10 May 1990)