Serotonergic dorsal raphe nucleus projections to the cholinergic and noncholinergic neurons of the pedunculopontine tegmental region a light and electron microscopic anterograde tracing and immunohistochemical studyкод для вставкиСкачать
THE JOURNAL OF COMPARATIVE NEUROLOGY 382:302–322 (1997) Serotonergic Dorsal Raphe Nucleus Projections to the Cholinergic and Noncholinergic Neurons of the Pedunculopontine Tegmental Region: A Light and Electron Microscopic Anterograde Tracing and Immunohistochemical Study TERESA L. STEININGER,1 BRUCE H. WAINER,4,5 RANDY D. BLAKELY,2 AND DAVID B. RYE3,4,6* 1Committee on Neurobiology, The University of Chicago, Chicago, Illinois 60637 2Department of Pharmacology, Vanderbilt University, Nashville, Tennessee 37232 3Neuroscience Program, Emory University, Atlanta, Georgia 30322 4Department of Neurology, Emory University, Atlanta, Georgia 30322 5Department of Pathology, Emory University, Atlanta, Georgia 30322 ABSTRACT The serotonergic dorsal raphe nucleus is considered an important modulator of statedependent neural activity via projections to cholinergic neurons of the pedunculopontine tegmental nucleus (PPT). Light and electron microscopic analysis of anterogradely transported biotinylated dextran, combined with choline acetyltransferase (ChAT) immunohistochemistry, were employed to describe the synaptic organization of mesopontine projections from the dorsal raphe to the PPT. In a separate set of experiments, we utilized immunohistochemistry for the serotonin transporter (SERT), combined with ChAT immunohistochemistry at the light and electron microscopic levels, to determine whether PPT neurons receive serotonergic innervation. The results of these studies indicate that: (1) anterogradely labeled and SERT-immunoreactive axons and presumptive boutons invest the PPT at the light microscopic level; (2) at the ultrastructural level, dorsal raphe terminals in the PPT pars compacta synapse mainly with dendrites and axosomatic contacts were not observed; (3) approximately 12% of dorsal raphe terminals synapse with ChAT-immunoreactive dendrites; and (4) at least 2–4% of the total synaptic input to ChAT-immunoreactive dendrites is of dorsal raphe and/or serotonergic origin. This serotonergic dorsal raphe innervation may modulate cholinergic PPT neurons during alterations in behavioral state. The role of these projections in the initiation of rapid eye movement (REM) sleep and the ponto-geniculo-occipital waves that precede and accompany REM sleep is discussed. J. Comp. Neurol. 382:302–322, 1997. r 1997 Wiley-Liss, Inc. Indexing terms: serotonin transporter; brainstem; biotinylated dextran; double immunohistochemistry; REM sleep The pedunculopontine tegmental nucleus (PPT), situated in the dorsolateral mesopontine tegmentum, together with the caudally adjacent laterodorsal tegmental nucleus (LDT), contain a prominent group of cholinergic neurons which project widely throughout the brainstem and forebrain (De Lima and Singer, 1987; Hallanger et al., 1987; Hallanger and Wainer, 1988; Rye et al., 1988; Woolf and r 1997 WILEY-LISS, INC. Grant sponsor: National Institutes of Health; Grant number: NS-1766114; Grant number: DA-07390. *Correspondence to: Dr. David B. Rye, Department of Neurology, Emory University, Suite 6000 WMRB, Atlanta, GA 30322. Email: firstname.lastname@example.org Received 23 July 1996; Revised 16 January 1997; Accepted 22 January 1997 DORSAL RAPHE AFFERENTS TO THE PPT Butcher, 1989; Woolf et al., 1990). Innervation of thalamicspecific relay, nonspecific and reticular nuclei by these cholinergic cells (De Lima and Singer, 1987; Hallanger et al., 1987; Hallanger and Wainer, 1988), are viewed as important substrates in the induction and maintenance of prolonged states of electroencephalograph (EEG) desynchronization, such as those accompanying wakefulness and rapid eye movement (REM) sleep (McCormick, 1989; Steriade et al., 1990a, 1993; Steriade, 1992). Converging lines of evidence also implicate the PPT in the transfer of phasic ponto-geniculo-occipital (PGO) waves in the thalamus that appear prior to and during REM sleep (McCarley et al., 1978; Hu et al., 1989; Steriade and McCarley, 1990; Steriade et al., 1990b; Steriade, 1992). Elucidation of the neural and pharmacological substrates influencing cholinergic PPT neurons has therefore garnered particular attention given their potential importance in governing states of EEG desynchronization and the evocation of PGO spikes. In addition to representing a sine qua non for the occurrence of REM sleep, PGO waves have been proposed as the neural correlate underlying dream-like imagery (Steriade and McCarley, 1990; Steriade et al., 1990b; Steriade, 1992), hallucinations (Kitsikis and Steriade, 1981; Steriade and McCarley, 1990) and orienting responses (Bowker and Morrison, 1976). Hyperpolarization of tonically active PPT/LDT neurons, and thereby inhibition of cortical ‘‘arousal,’’ is felt to be mediated by serotonin, gammaaminobutyric acid (GABA), acetylcholine (Muhlethaler et al., 1990; Steriade and McCarley, 1990; Jones, 1991; Steriade, 1992; Leonard and Llinas, 1994), adenosine (Rainnie et al., 1994), and norepinephrine (Williams and Reiner, 1993), while hyperpolarization of a separate population of neurons displaying low-threshold calcium spikes contributes to a burst firing mode and PGO waves. Histamine (Muhlethaler et al., 1990) and glutamate (Sanchez et al., 1991) are proposed excitatory transmitters in the PPT/LDT. Very little is known, however, concerning the origin of these transmitters and the specific cell types of the PPT region with which they are in synaptic contact Abbreviations AQ, Aq cll Cnf cp CTF DR DTg IP LC LDT ll MEA mes5 ml mlf PAG PB PL PPT PTF RRF scp SN xscp 3n 4n 4N 4V cerebral aqueduct commissure of the lateral lemniscus cuneiform nulceus cerebral peduncle central tegmental field dorsal raphe dorsal tegmental nucleus interpeduncular nucleus locus coeruleus laterodorsal tegmental nucleus lateral lemniscus midbrain extrapyramidal area mesencephalic tract of the trigeminal nerve medial lemniscus medial longitudinal fasciculus periaqueductal gray matter parabrachial nucleus paralemniscal nucleus pedunculopontine tegmental nucleus pontine tegmental field retrorubral field superior cerebellar peduncle substantia nigra decussation of the superior cerebellar peduncle oculomotor nerve trochlear nerve trochlear nucleus fourth ventricle 303 (e.g., cholinergic versus noncholinergic/tonically active versus bursters). The spectrum of putative afferents to the mesopontine tegmentum region, containing PPT cholinergic neurons, is diverse. Comingling of cholinergic cells and smaller noncholinergic cells, and their close apposition to several other functionally distinct nuclei and fiber tracts has precluded a precise determination of which afferents actually terminate in the region, and of these, which terminate on cholinergic elements. Putative serotonergic afferents from the dorsal raphe nucleus (DR) to cholinergic PPT/LDT neurons have been proposed as one of the important neural substrates governing both the normal and pathological expression of REM sleep and PGO waves (Steriade and McCarley, 1990). The synaptic organization of dorsal raphe and/or serotonergic pathways to the PPT/LDT is nevertheless poorly described. Anatomical tracer injections into the mesopontine tegmentum, containing the PPT, retrogradely label DR neurons (Semba and Fibiger, 1992; Steininger et al., 1992), but fail to discriminate as to which specific cells receive DR afferents. Injections of tritiated amino acids (Bobillier et al., 1975; Pierce et al., 1976; Moore et al., 1978; Conrad et al., 1991) and Phaseolus vulgaris leucoagglutinin (PHA-L; Vertes, 1991; Vertes and Kocsis, 1994) into the DR delineate efferents which putatively innervate the PPT. Whether these efferents are actually serotonergic is unclear because only 40–60% of the total population of DR neurons employ serotonin as a neurotransmitter (Moore, 1981; Descarries et al., 1982). Moreover, descriptions of mesopontine tegmental projections from the DR generally lack detail because the cholinergic identity of PPT neurons was not examined. Synapses between serotonergic terminals and cholinergic elements have been described in the rat PPT and LDT, but the majority contact noncholinergic dendrites (Honda and Semba, 1994). Because these findings were not quantified, it is unclear as to: (1) the relative distribution of serotonergic efferents between various cholinergic and noncholinergic elements; (2) what extent serotonergic terminals contribute to the total synaptic innervation of cholinergic PPT neurons; and (3) whether these serotonergic efferents arise from the DR. In the present study, the anterograde transport of biotinylated dextran from the DR was therefore combined with immunocytochemistry for choline acetyltransferase (ChAT), the synthetic enzyme for acetylcholine, to provide a detailed description of projections to the cholinergic PPT and LDT nuclei at the light microscopic level. At the electron microscopic level, the synaptic connectivity of anterogradely labeled terminals in the pars compacta of the PPT at the electron microscpic level was established and quantified in detail. To investigate the putative serotonergic nature of these projections, immunohistochemistry for the serotonin transporter protein (SERT; Qian et al., 1995) was utilized as a marker for serotonergic axons and terminals, in combination with ChAT immunohistochemistry, and investigated in a like manner at both light and electron microscopic levels in a separate set of experiments. MATERIALS AND METHODS Injection of anterograde tracer Male Sprague-Dawley rats (250–350 g, Harlan) were anesthetized with chloral hydrate (400 mg/kg, i.p.) and placed in a stereotaxic apparatus. Stereotaxic injections of 10% biotinylated dextran amine (10,000 MW, lysine fix- 304 T.L. STEININGER ET AL. able; Molecular Probes, Eugene, OR) were made into the dorsal raphe nucleus using a micropipette at a lateral approach of 20°. The tracer was ejected iontophoretically with a positive current of 5 µA, at a 7 seconds on/7 seconds off cycle, for 30 minutes. After a survival period of 7–10 days, the animals were perfused and the brains sectioned as described below. All animal care and procedures employed as part of these studies were approved by the institutional animal care and use committee (IACUC). Perfusion/fixation Animals were deeply anesthetized with chloral hydrate and given an injection of sodium heparin (600 U i.p.) 15 minutes prior to the procedure. The animals were perfused transcaridally with fixative consisting of 300 ml of 3% paraformaldehyde and 0.1–0.2% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4 (PB), over a period of 15 minutes. The brains were postfixed in situ for 1 hour before removal from the skull. Brains were then dissected into blocks containing the injection site and the PPT, and sectioned at 40–50 µm on a vibrating microtome in PB. Tissue sections were collected into staining dishes and rinsed several times in 0.05 M Tris-buffered saline pH 7.6 (TBS). containing 0.025% sodium nitroprusside and 0.005% hydrogen peroxide. The reaction was allowed to proceed for 15–30 minutes and was stopped by several rinses in 0.01 M PB pH 6.6. Following the biotinylated dextran (BD)/ ChAT visualization procedures, sections were processed for light and electron microscopy as described below. Control sections, in which the primary antiserum was omitted, were processed in the same manner. The utility of these chromogen combinations at the light and electron microscopic level has been previously demonstrated by ourselves (Bolam et al., 1986; Levey et al., 1986; Rye et al., 1987) and others (Zaborszky and Heimer, 1989; Bolam et al., 1991; Lapper and Bolam, 1992; Corvaja et al., 1993; Zaborszky et al., 1993). SERT and serotonin (5-HT) immunohistochemistry The injected and transported biotinylated dextran was visualized by incubation of the sections in avidin-biotinperoxidase complex (ABC, Vectastain Elite kit, Vector Labs, Burlingame, CA) in TBS for 24–48 hours and rinsing in TBS (4 3 15 minutes). Two visualization procedures were followed: (1) visualization with 3,38 diaminobenzidine tetrahydrochloride (DAB, Sigma Chemicals, St. Louis, MO) or (2) visualization with nickel-enhanced DAB (NiDAB; Hsu and Soban, 1982; Hancock, 1986). For DAB visualization, sections were rinsed in 0.05 M Tris buffer pH 7.6, and incubated in 0.05% DAB and 0.01% hydrogen peroxide in 0.05 M Tris buffer for 5–10 minutes. The reaction was halted by extensive rinses in TBS. For nickel-enhanced DAB visualization, sections were rinsed in buffer containing imidazole (0.01 M) and sodium acetate (0.12 M; IAB; pH 7.2–7.4), and incubated in a solution of 0.05% DAB, 2.5% nickel ammonium sulfate and 0.005% hydrogen peroxide in IAB for 5–10 minutes, then extensively rinsed in IAB, followed by rinses in TBS. Rats were perfused as above with 3% paraformaldehyde and 0.05–0.1% glutaraldehyde in PB. Immunohistochemistry was performed as follows. Sections were incubated in 4% normal goat serum in TBS (diluent) for 2 hours to block nonspecific staining, then incubated in primary antiserum (affinity-purified, rabbit polyclonal anti-SERT antibody CT-2; 0.5–1.0 µg/ml) that was directed at the carboxy terminus of the SERT protein (Qian et al., 1995) or rabbit-anti-5-HT (Incstar, Stillwater, MN) for 48 hours. Following four 15-minute rinses in TBS, sections were then incubated overnight in biotinylated goat-anti-rabbit (Vectastain Elite kit, Vector Labs), and then rinsed in TBS and incubated in avidin-biotin-peroxidase complex (Vectastain Elite kit, Vector Labs) for 4 hours. Following rinses in TBS and 0.05 M Tris, the bound peroxidase was visualized with either NiDAB or DAB as described above. Control sections, in which the primary antiserum was omitted, were processed in the same manner. ChAT immunohistochemistry was then performed with the following modifications to the above procedure: (1) the diluent consisted of 4% normal goat serum in TBS; and (2) the secondary antibody was goat-anti-rat. ChAT immunoreactivity was visualized with either DAB (following visualization of SERT with NiDAB) or BDHC (following visualization of SERT with DAB) as the chromogen. Following the double immunohistochemical procedures, sections were processed for light and electron microscopy as described below. ChAT immunohistochemistry Preparation for light microscopy Anterograde tracing was combined with immunohistochemistry for ChAT. Following the above procedure, sections were incubated in 4% normal rabbit serum in TBS (diluent) for 2 hours to block nonspecific staining. Sections were then incubated for 24–48 hours at 4°C in primary antibody (monoclonal rat-anti-ChAT [AB8] 1:250; Levey et al., 1983) in diluent. Following the incubation and four 15-minute rinses in TBS, sections were incubated overnight in secondary antibody (rabbit-anti-rat, Incstar) 1:200 in diluent; rinsed in TBS (4 3 15 minutes) and incubated for 4 hours in rat peroxidase-anti-peroxidase (PAP, Sternberger, Baltimore, MD) 1:200 in diluent. Following four 15 minute rinses in TBS, the bound peroxidase was visualized with either DAB (after tracer visualization with NiDAB) or benzidine dihydrochloride (BDHC; after tracer visualization with DAB; Levey et al., 1986). The DAB reaction was performed as above. For the BDHC visualization, sections were rinsed in 0.01 M PB pH 6.6 and incubated in 0.01% BDHC (Sigma), in 0.01 M PB pH 6.6, Following immunohistochemical procedures, the sections were mounted onto gelatin-coated slides, air-dried, dehydrated in ethanol, cleared in xylene and coverslipped with DPX mountant (BDH Laboratory Supplies, Poole, England). Tissue sections were examined in the light microscope and drawings were made with the aid of a camera lucida drawing tube. Nuclear boundaries were determined with the aid of the rat brain atlas of Paxinos and Watson (1986), with mesopontine tegmental nuclear divisions following the convention adopted by Rye et al. (1987). The subnuclear organization of the dorsal raphe follows the terminology of Steinbusch et al. (1981) and Steinbusch and Nieuwenhuys (1983). Visualization of tracer Preparation for electron microscopy Following the staining procedures, sections were postfixed in 2% glutaraldehyde in PB for 2–12 hours, and osmicated in 1% osmium tetroxide in either 0.1 M cacodylate buffer, pH 7.5 (for NiDAB/DAB staining) or PB pH 6.6 DORSAL RAPHE AFFERENTS TO THE PPT (for DAB/BDHC staining). Sections were then rinsed, dehydrated in ascending concentrations of ethanol and then propylene oxide, infiltrated with resin (Durcupan ACM, Fluka, Ronkonkoma, NY), mounted onto microscope slides, and coverslipped with a plastic slide, which was removed following curing of the resin at 60°C for 48 hours. Blocks containing the region of the dorsal PPT pars compacta (PPT-pc; Fig. 2C) were cut out and glued to blank cylinders of resin for sectioning. Ultrathin sections (70–80 nm) were cut on an ultramicrotome (Sorvall MT2B), collected onto Formvar-coated single slot grids, stained with uranyl acetate and lead citrate, and examined in an electron microscope (Zeiss EM10 or Jeol JEM 100C). Electron microscopic analysis The region of the dorsal PPT-pc, illustrated in Figure 2C, was analyzed at the electron microscopic level. Given the variability in penetration of immunoreagents, only the most superficial 5–10 µm of each tissue block was examined. Three or more blocks of the PPT-pc were examined from each animal in order to determine: (1) the proportion of ChAT-immunoreactive (IR) versus non-ChAT-IR spines, dendrites, terminals or somata contacted by a labeled afferent; (2) the density of afferent labeling in the neuropil; and (3) the density of afferent labeling contacting ChAT-IR elements. Two analyses were performed separately for both anterogradely labeled and SERT-immunoreactive (SERT-IR) terminals. In the first analysis, a sampling region from the dorsal PPT-pc was examined in the electron microscope and all labeled terminals encountered in the ultrathin section were photographed at approximately 310,000 magnification. The numbers of BDlabeled or SERT-IR synaptic profiles in contact with ChAT-IR and non-ChAT-IR spines, dendrites, terminals and somata were counted. Dendrites postsynaptic to labeled elements were further classified as small (,1 µm diameter), medium (.1 µm but ,2 µm) or large (.2 µm). These absolute numbers were then converted into percentages to express the proportion of labeled terminals in contact with ChAT-IR elements (number of synaptic profiles with ChAT-IR elements/total number of synaptic profiles; see Table 1). In the second analysis, photomicrograph montages were made of randomly selected regions of sampled ultrathin sections, at approximately 37,000 magnification (an average of 3,000 µm2 was analyzed for each tissue block), to determine the density of afferent labeling in the neuropil and those contacting ChAT-IR elements. If more than 33% of the sampling region was taken up by blood vessels or myelinated axon bundles (e.g., commissure of the lateral lemniscus or superior cerebellar peduncle), new coordinates were determined. Electron micrographs were then made of BD-labeled or SERT-IR terminals, and analysis derived from 8–10 overlapping photographic prints. The following were counted in each montage: (1) labeled axon terminals; (2) labeled axon terminals forming synapses with ChAT-IR spines, dendrites or somata; (3) labeled axon terminals forming synapses with non-ChAT-IR spines, dendrites or somata; (4) unlabeled axon terminals; (5) unlabeled axon terminals forming synapses with ChAT-IR spines, dendrites or somata; (6) -unlabeled axon terminals forming synapses with non-ChAT-IR spines, dendrites or somata; and (7) labeled and unlabeled dendrites. Quantitation was also performed both for labeled synaptic profiles (S) and for total labeled terminals (synaptic profiles plus 305 TABLE 1. Synaptic Targets of Anterogradely Labeled and Serotonin Transporter-Immunoreactive Terminals in the PPT-pc1,2 Anterogradely labeled terminals Unlabeled dendrites small (,1 µm) medium (.1 µm; ,2 µm) large (.2 µm) total ChAT-IR3 Dendrites small medium large total Unlabeled terminals Total SERT-IR4 terminals n % of total n % of total 125 50 4 179 59.5 23.8 1.9 85.2 35 8 1 44 72.9 16.7 2.1 91.7 17 5 3 25 6 210 8.1 2.4 1.4 11.9 2.9 3 0 0 3 1 48 6.2 0 0 6.2 2.1 1Data are shown for those terminals having distinct synaptic specializations. pedunculopantine tegmental nucleus pars compacta. choline acetyltransferase immunoreactivity. 4SERT-IR, Serotonin transporter immunoreactivity. 2PPT-pc, 3ChAT-IR, terminal appositions, S 1 N; see Table 2). The absolute numbers were then converted into percentages to express the following characteristics of each labeled afferent: (1) the density of terminals in the neuropil (labeled terminals/ labeled 1 unlabeled terminals) and (2) the density of terminals contacting ChAT-IR elements (labeled terminals contacting ChAT-IR elements/total terminals contacting ChAT-IR elements; see Table 2). The density of afferent labeling on ChAT-IR elements was also expressed as a terminal-covering ratio for ChAT-IR dendrites, which was calculated in a manner similar to that of a somatic terminal-covering ratio described in the accompanying paper (Steininger et al., 1997). Measurements of terminals and postsynaptic elements were made from photographic prints. The length of dendritic membrane and the length of terminals (both synaptic profiles and terminal appositions) contacting the dendrite were measured by using an image analysis program (NIH Image). The terminal-covering ratio was determined by dividing the total length of labeled terminal by the total length of dendritic membrane. RESULTS Anterograde tracing from the dorsal raphe nucleus Biotinylated dextran injection sites in the dorsal raphe. Iontophoretic injections of the anterograde tracer BD were made into the DR. Four distinct subdivisions within the DR are recognized on the basis of cytoarchitecture and cell packing density (Steinbusch et al., 1981; Steinbusch and Nieuwenhuys, 1983; Fig. 1A): a dorsomedian division, the ventromedial division, paired lateral divisions and a pars caudalis. Injections of biotinylated dextran amine were directed at the dorsomedian subdivision, because our previous retrograde tracing studies indicated that the majority of dorsal raphe and/or serotonergic inputs to the PPT originate from cells in this division (Steininger et al., 1992). The locations of small, iontophoretic applications of BD were confirmed on both cyto- and chemoarchitecture grounds, as well as by determining that resultant patterns of labeling in brain regions outside the mesopontine tegmentum recapitulated those previously described for DR efferents (Vertes, 1991; Vertes and Kocsis, 1994). In all cases, moderate to dense anterogradely labeled fibers were observed in the bed nucleus of the stria 306 T.L. STEININGER ET AL. TABLE 2. Frequency of Anterogradely Labeled and SERT-Immunoreactive Terminals Synapsing or Contacting CHAT-Immunoreactive Dendrites in the PPT-pc1 Terminals synapsing with CHAT-IR dendrites Case Anterogradely labeled terminals B14 B23 B29 total SERT-immunoreactive terminals TR7 TR20 TR21 total Terminals apposed to CHAT-IR dendrites (S 1 N) Synaptic terminals in the neuropil ChAT-IR dendrites (n) Unlabeled terminals (n) Labeled terminals (n) Frequency of labeling (%) Unlabeled terminals (n) Labeled terminals (n) Frequency of labeling (%) Unlabeled terminals (n) Labeled terminals (n) Frequency of labeling (%) 114 93 64 271 93 47 46 186 4 2 1 7 4.1 4.1 2.1 3.6 152 87 76 315 6 4 2 12 3.8 4.4 2.6 3.7 1,651 1,420 1,751 4,822 62 53 31 146 3.6 3.6 1.7 2.9 49 38 71 158 46 24 45 115 1 1 2 4 2.1 4.0 4.3 3.4 75 48 72 195 1 3 3 7 1.3 5.9 4.0 3.5 643 1,110 1,158 2,911 22 24 19 65 3.3 2.1 1.6 2.2 1Data shown are for terminals that synapse with ChAT-immunoreactive dendrites (left set of columns) and for all terminals, both nonsynapsing and synapsing (N 1 S), apposed to ChAT-immunoreactive dendrites (center columns). Data in the rightmost column are for terminals making synaptic contact with both ChAT-immunoreactive and nonimmunoreactive elements in the neuropil (i.e., inclusive of all synaptic terminals in the neuropil). terminalis, amygdala, ventral striatum, lateral hypothalamic area, intralaminar and posterior thalamic nuclei, the posterior hypothalamus, ventral tegmental area, locus coeruleus, the parabrachial nucleus (both lateral and medial subnuclei), and pontine reticular formation (particularly medial regions). Moderate anterogradely labeled fibers were observed in the striatum, piriform cortex, nucleus accumbens, medial septum/diagonal band, substantia innominata, dorsomedial and ventromedial hypothalamic nuclei, lateral geniculate, zona incerta, entopeduncular nucleus, substantia nigra and medullary gigantocellular field (particularly the ventromedial zone immediately dorsal to the inferior olive). A few anterogradely labeled fibers were observed in the olfactory bulb, frontal cortex, suprachiasmatic nucleus, the thalamic reticular nucleus and other thalamic nuclei, hippocampus and the lateral habenula. Taken together, these observations, made us feel entirely confident in attributing the present findings to specific uptake and transport of BD by DR neurons. In order to determine whether the anterogradely labeled terminals make direct synaptic contact with the cholinergic PPT neurons, anterograde tracing was combined with ChAT immunohistochemistry. Two double-staining visualization techniques were followed. In one procedure, the anterograde tracer was visualized with DAB, which yielded a smooth, brown reaction product, and ChAT immunoreactivity was visualized with BDHC, which yielded a granular blue reaction product. Alternatively, the anterograde tracer was visualized with NiDAB, which yielded a dense, amorphous, black reaction product, and ChAT immunoreactivity was visualized with DAB (brown homogeneous reaction product). With both methodologies, the anterogradely labeled fibers and varicosities were easily distinguishable from the ChAT-immunoreactive cells and dendrites on the basis of the color and texture of the reaction products. The possible retrograde labeling of neurons, which has been observed with the biotinylated dextran tracer, was minimized in these studies by using iontophoretic injections methods and micropipetters with small tip diameters. The location of retrogradely labeled neurons in each case is described for each case below. Visualization of the anterograde tracer yielded a Golgilike dark brown (DAB) or black (NiDAB) staining of cell bodies and processes of the neurons in the DR that had taken up and transported the BD (Fig. 2A). Anterogradely labeled fibers and varicosities were dense in regions immediately surrounding the injection site. Labeled varicose fibers were also observed in the ependymal layer surrounding the cerebral aqueduct. The cases having injection sites largely restricted to the DR were utilized for analysis and these are depicted in Figure 1. An example of a BD injection site (Case B26) is illustrated in the photomicrograph in Figure 2A. Case B14. The injection site in case B14 (Fig. 1B) was the largest of the analyzed cases. The injection was centered at a midcaudal level of the DR and involved nearly the entire DR. The injection extended from the level of the trochlear nucleus, rostrally, to the level of the transition from the cerebral aqueduct to the fourth ventricle (at the rostral level of the LDT), caudally. A few neurons that had taken up the tracer were located in the ventrolateral periaqueductal gray (PAG), lateral to the lateral subdivision of the DR. Anterograde labeling was most extensive in this case as was the amount of retrograde labeling. A distinct cluster of retrogradely labeled neurons were observed bilaterally, rostral to the injection site in the ventrolateral central gray, immediately dorsal to the oculomotor nucleus. Scattered retrogradely labeled neurons (about 5–10 neurons per 40-µm section) were also observed in the pontine reticular formation, central tegmental field and dorsal periaqueductal gray. Retrograde labeling in other parts of the brain was minimal (,20 neurons total), and limited primarily to the lateral hypothalamic and preoptic areas. Case B23. The injection site in case B23 (Fig. 1B) was the second largest and well localized to the DR, and encompassed nearly the entire nucleus, including the ventral aspect which was not included in the injection site of case B14. This injection was centered at a mesencephalic level immediately caudal to the trochlear nucleus, extending from the level of the rostral trochlear nucleus to the level of the fourth ventricle in the rostral pons. The handful of retrogradely labeled neurons observed in this case were situated in the ventrolateral PAG, immediately dorsal to the oculomotor nucleus, as in case B14. Case B26. The injection site in case B26 (Figs. 1C, 2A) was smaller than those in cases B14 or B23, and was centered in the lateral portion of the dorsomedial subdivision of the DR (Steinbusch et al., 1981; Steinbusch and DORSAL RAPHE AFFERENTS TO THE PPT 307 Fig. 1. Biotinylated dextran injection sites in the dorsal raphe nucleus (DR). Camera lucida drawings of coronal sections at four rostrocaudal levels of the DR illustrate the location and extent of the tracer injections. The boundaries of the DR nucleus are illustrated in column A. The injection sites of case B14 (dotted area) and case B23 (striped area) are illustrated in column B. In column C, smaller injection sites of case B26 (dotted area) and B29 (striped area) are illustrated. Scale bar 5 1 mm. Nieuwenhuys, 1983), at a mesencephalic level caudal to the trochlear nucleus. The injection site extended from the level of the caudal region of the oculomotor nucleus to the level of the rostral LDT (caudal midbrain). Few (,10) retrogradely labeled neurons were observed in the mesopontine tegmentum. The ultrastructural preservation was not optimal for electron microscopic analysis; therefore, this case was utilized to illustrate anterograde labeling at the light microscopic level only. Case B29. The injection site in case B29 (Fig. 1C) was the smallest, and centered in the dorsomedial and ventromedial subdivisions of the DR (Steinbusch et al., 1981; Steinbusch and Nieuwenhuys, 1983). The anteroposterior extent of the injection was limited to sections, including the bulk of the trochlear nucleus through a mid-PPT level. Retrograde labeling outside the injection site was limited to a few neurons in the PAG adjacent to the DR. Anterograde labeling in the mesopontine tegmentum. Coronal sections of the mesopontine tegmentum processed for the visualization of BD and ChAT were examined in the light microscope. A moderate-to-dense plexus of anterogradely labeled varicose axonal processes were observed in the region of the cholinergic PPT neurons. The majority of projection axons were thin, varicose and infrequently studded with boutons. Thick tortuous axons that exhibited clusters of larger boutons were occasionally observed. A representative case (B26) was used to illustrate the pattern of anterograde labeling in the mesopontine tegmentum following DR injections. Camera lucida drawings were made to illustrate the relationship of anterogradely labeled varicosities to ChAT-immunoreactive (ChAT-IR) neurons at four rostrocaudal levels of the mesopontine tegmentum (Fig. 3A–D). Rostral (Fig. 3A). The major target of DR fibers at this level was the midbrain central gray, particularly the rostral DR and ventrolateral periaqueductal gray. Few anterogradely labeled fibers were observed in the lateral and dorsal PAG at this level. Moderate-to-dense anterograde labeling was also observed in tegmental regions rich in dopaminergic neurons, such as the caudal aspects of the ventral tegmental area and substantia nigra pars compacta, as well as the retrorubral field (RRF). Anterogradely 308 T.L. STEININGER ET AL. Fig. 2. Combined light microscopic visualization of anterograde labeling and choline acetyltransferase (ChAT) immunostaining (Case B26). A: Coronal section of the DR, depicting the appearance of a typical biotinylated dextran injection site. The neurons that have taken up and transported the tracer are filled with 3,38 diaminobenzidine (DAB) reaction product. B: A coronal section through the dorsal part of the pedunculopontine tegmental nucleus (PPT) pars compacta. The fibers of the commissure of the lateral lemniscus (arrows) are prominent landmarks of the pars compacta. The ChAT-immunoreactivity (IR) neurons are visualized with benzidine dihydrochloride (BDHC; granular reaction product). ChAT immunoreactivity is also visible in the trochlear nerve in the upper left corner of the photomicrograph. C: Higher magnification photomicrograph montage of the PPT region depicted in B. Numerous anterogradely labeled fibers and boutons are seen in the vicinity of the ChAT-IR neurons. Arrows mark the same fascicles of the commissure of the lateral lemniscus as in B, and the asterisks mark the location of the identical blood vessel in B and C for orientation purposes. Scale bars 5 500 µm in A; 250 µm in B; 100 µm in C. labeled fibers were a bit less dense dorsal to the RRF in the central tegmental field (CTF). Few axons were observed in the vicinity of the most anterior PPT neurons located ventrolaterally in the tegmentum between the RRF dor- sally, and substantia nigra (SN), ventrally. Anterograde labeling in the contralateral tegmentum was less dense, but described a similar pattern. The amount of contralateral anterograde labeling appeared to vary according to DORSAL RAPHE AFFERENTS TO THE PPT 309 Fig. 3. Anterograde labeling in the mesopontine tegmentum in case B26. A–D: Camera lucida drawings of coronal sections at four rostrocaudal levels in (BD)/ChAT double-stained sections, illustrating the pattern of DR innervation in this brain region. The shape and location of cholinergic neurons and anterogradely labeled varicosities are illustrated. At rostral levels of the PPT (A,B), very little of theanterograde labeling is directed to the region of the cholinergic neurons. The situation is much different at caudal levels (C,D), where much of the labeling coincides with cholinergic PPT neurons (and the adjacent MEA and LDT), and less anterograde labeling is observed elsewhere in the section, e.g., cuneiform nucleus (Cnf) and pontine tegmental field (PTF). Scale bar 5 500 µm. the extent of involvement of the contralateral DR at the injection site. Mid-PPT (Fig. 3B). The injection site was centered at this level, immediately caudal to the trochlear nucleus. Dense patches of anterogradely labeled varicosities were observed in a region that extended ventrolaterally from the central gray to the central tegmental field immediately dorsal to the PPT, and also in the tegmentum between the midline and the PPT neurons. While cholinergic neurons were more numerous than in the rostral level, they remained rather loosely arranged and had not yet reached the density observed in the PPT-pc. A few anterogradely labeled fibers and boutons were seen adjacent to cholinergic neurons. Midcaudal PPT (Fig. 3C). At this level of the mesencephalon, the majority of anterogradely labeled varicose fibers were concentrated in the region of the PPT, as illustrated in the photomicrographs in Figure 2B and 2C. At this level, cholinergic neurons were concentrated in the PPT-pc, situated within the commissure of the lateral 310 lemniscus (cll), lateral to the ascending limb of the superior cerebellar peduncle (scp). Many labeled varicosities were observed in putative contact with ChAT-IR cell bodies and dendrites. Labeled fibers were quite dense in the dorsal part of the pars compacta and the pars dissipatus ventromedial to the superior cerebellar peduncle, as compared to ventral aspects of the PPT-pc. Moderate amounts of anterogradely labeled fibers were observed in the midbrain extrapyramidal area (MEA), just medial to the cholinergic neurons in the PPT-pc. Moderately dense anterograde labeling was observed in the tegmentum just lateral to the median raphe and in the DR caudal to the injection site. Few anterogradely labeled fibers were seen ventral to the pontine tegmental field (PTF). Caudal PPT (Fig. 3D). Anterogradely labeled fibers were concentrated within and dorsal to cholinergic somata in the LDT. Labeled fibers were less dense in the lateral parabrachial subnuclei (PB) immediately caudal to the PPT-pc. A few anterogradely labeled varicose axons were observed in the caudal DR. A moderately dense plexus of labeled axons were present in the neuropil containing the remaining cholinergic PPT neurons, located ventromedial to the superior cerebellar peduncle and ventral to the LDT (i.e.; the ‘‘subcoeruleal region’’). The labeling observed in the medial tegmentum more rostrally was not observed at this level. Moderate anterogradely labeling was also observed in the pontine tegmental field ventral to the PPT. No significant anterograde labeling was observed in the pontine nuclei. Anterograde labeling: Electron microscopy Ultrastructural appearance. Anterograde tracing visualized with DAB (smooth brown reaction product) was paired with ChAT immunoreactivity visualized with BDHC (blue granular reaction product). Alternatively, anterograde tracing visualized with NiDAB (dense black reaction product) was paired with ChAT immunoreactivity visualized with DAB. At the electron microscopic level, the DAB reaction product was amorphous and associated with organelle membranes and inner plasmalemma, whereas the BDHC reaction product appeared as large crystalline granules dispersed in the cytoplasm, such that the combination of DAB with BDHC resulted in reliably distinct reaction products. In sections developed using the combination of NiDAB and DAB, the NiDAB and DAB reaction products were similarly amorphous. The NiDAB reaction product in the anterogradely labeled terminals, however, was more abundant and electron-dense than the DAB reaction product in the ChAT-immunoreactive elements. The distinction between NiDAB and DAB reaction products has been previously described by other investigators (Lapper and Bolam, 1992; Zaborszky et al., 1993). The NiDAB reaction product was observed in a few neurons in the PPT region that were retrogradely labeled from the injection (in case B14 only). The appearance of the NiDAB reaction product in these retrogradely labeled elements was reliably distinct from the DAB reaction product seen in ChAT-IR elements, i.e., the reaction product was more dense and completely filled the somatodendritic compartment, as opposed to the patchy, flocculent electron-dense reaction product observed in ChAT-IR somata and dendrites (Fig. 5J, K). A few DAB- or NiDABlabeled myelinated axons were also observed (Fig. 5I), although the distinction between ChAT-IR and anterogradely labeled myelinated axons was less clear. T.L. STEININGER ET AL. Quantitative analysis. Two analyses were performed: (1) the postsynaptic targets of anterogradely labeled terminals were determined and quantitated; and (2) the proportion of anterogradely labeled versus unlabeled terminals synapsing with ChAT-IR neurons was determined and compared to the pattern observed for noncholinergic elements in the surrounding neuropil. For the first analysis, photomicrographs were made of each anterogradely labeled terminal encountered in these blocks. A total of 353 anterogradely labeled terminals were photographed. Only those exhibiting clear synaptic specializations were selected for further analysis, for a total of 210 anterogradely labeled terminals. Anterogradely labeled terminals were typically bulb- or club-shaped, and ranged in size from 0.23 to 3.74-µm-long (parallel to synaptic membrane; mean 5 0.94 µm) by 0.15 to 1.58-µmwide (perpendicular to synaptic membrane; mean 5 0.61 µm). Of the 210 anterogradely labeled terminals analyzed, 179 (85.2%) made synaptic contact with unlabeled dendrites, 25 (11.9%) made synaptic contact with ChAT-IR dendrites, six (2.8%) made synaptic contact with unlabeled terminals (Table 1), and no synapses were observed between anterogradely labeled terminals and neuronal somata. The unlabeled dendrites contacted by anterogradely labeled terminals (Fig. 4) ranged from 0.15 to 2.72 µm in diameter. In a few instances, two anterogradely labeled terminals contacted the same dendritic target (Fig. 4F). Although synaptic vesicle profiles in the anterogradely labeled terminals were partially obscured by the DAB or NiDAB reaction product, the shape of these vesicles appeared to be round in most of the labeled terminals. No clear distinctions in vesicle type could be made; however, it appeared that some terminals contained smaller, more densely packed vesicles (Fig. 4G). Of the 179 anterogradely labeled terminals contacting unlabeled dendrites, 125 (69.8%) synapsed with small (,1 µm diameter) dendrites (Fig. 4B–D,E,G–I), 50 (23.8%) synapsed with medium dendrites (between 1 and 2 µm; Fig. 4A,F) and four (2.8%) synapsed with large dendrites (.2 µm). Synaptic specializations were both symmetric (Fig. 4A–C) and asymmetric (Fig. 4D–I). Asymmetric synapses were preferentially located on small dendrites, while symmetric synapses were more prevalent on medium and large dendrites. Of the synapses with small dendrites, 68 were asymmetric and 57 were symmetric. Of the synapses with medium dendrites, 19 were asymmetric, 30 were symmetric and one was unclear. Of the synapses with large dendrites, one was asymmetric and three were symmetric. Approximately 16% of the asymmetric synapses were associated with subjunctional dense bodies. Of the 210 anterogradely labeled terminals analyzed, 25 (11.9%) were found to make synaptic contact with ChAT-IR elements, and in all cases, these were dendrites of ChAT-IR neurons. Of these 25 synapses, 17 (68%) were with small ChAT-IR dendrites (Fig. 5B–F), five (20.0%) were with medium ChAT-IR dendrites (Fig. 5A) and three (12%) were with large ChAT-IR dendrites. Synapses between anterogradely labeled terminals and ChAT-IR dendrites were both asymmetric (Fig. 5C,D) and symmetric (Fig. 5A,B,E,F). Of the synapses made with small ChAT-IR dendrites, 12 were asymmetric and five were symmetric. Of the synapses with medium ChAT-IR dendrites, one was asym- DORSAL RAPHE AFFERENTS TO THE PPT Fig. 4. Electron photomicrographs depicting synaptic contacts, between anterogradely labeled axon terminals (t) and unlabeled dendrites (d), that illustrate the morphology of anterogradely labeled terminals and the caliber of unlabeled dendrites that are contacted. Synaptic specializations are symmetric (A–C) and asymmetric (D–I). 311 Note that the reaction product densely fills these terminals. In F, two anterogradely labeled terminals appose a single unlabeled dendrite; however, synaptic specializations are not apparent for the apposition on the right. Scale bar 5 0.5 µm. 312 metric, three were symmetric and one was unclear. Of the synapses with large ChAT-IR dendrites, one was asymmetric and two were symmetric. Of the 210 anterogradely labeled terminals analyzed, six (2.9%) made synapses with unlabeled terminals (axoaxonic synapses). The synapses were either symmetric (n 5 5) or asymmetric (n 5 1). The postsynaptic terminals contained either pleomorphic vesicles (Fig. 5G) or small round clear vesicles (Fig. 5H). None of the postsynaptic terminals were seen to be in synaptic contact with other elements in the plane of section. In the second analysis, 271 ChAT-IR dendritic profiles were examined from the same tissue sampled in the first analysis (Table 2). A total of 327 terminals were apposed to these ChAT-IR dendrites, and of those terminals, only 193 exhibited clear synaptic specializations. Of these 193 synaptic profiles, seven were anterogradely labeled, yielding a frequency of 3.6% of anterogradely labeled input to the ChAT-IR dendrites (which varied among cases from 2.1% to 4.1%). The proportion was found to be similar (3.7%) when labeled terminal appositions (i.e., those exhibiting no apparent synaptic specialization) were included. In the neuropil surrounding these ChAT-IR dendrites, 4,968 synaptic profiles were identified. Of these, 146 were anterogradely labeled, giving a frequency of labeling of 2.9% (range of 1.7–3.6%) in the neuropil. Serotonin transporter immunoreactivity It was of interest to determine whether inputs from the dorsal raphe to the PPT might be serotonergic, since serotonin has been implicated in behavioral state control, and since only a portion (40–60%) of neurons in the cytoarchitectonically defined DR are serotonergic (Moore, 1981; Descarries et al., 1982; Léger and Wiklund, 1982). We therefore employed SERT immunohistochemistry as a marker for the serotonergic terminals. As illustrated in Figure 6C,D, a complete absence of immunostaining was observed in the PPT, following preabsorption of the antiSERT antibody with SERT fusion protein (used as antigen at 100 µg/ml). Based on the high quality of immunostaining achieved with this antibody, it was our expectation that SERT immunohistochemistry would yield superior ultrastructure and staining when compared to immunohistochemistry using commercially avaiable antibodies to serotonin. Light microscopy. Serotonin transporter immunoreactivity (SERT-IR) visualized with DAB (smooth, brown) was paired with ChAT immunoreactivity visualized with BDHC (granular, blue), and SERT immunoreactivity visualized with NiDAB (dense, black) was paired with ChAT immunoreactivity visualized with DAB (smooth, brown). SERT-immunoreactive neurons were observed in the brainstem raphe nuclei and SERT-IR varicose axons were observed in many brain regions. The appearance and distribution of SERT-IR neurons in the dorsal raphe nucleus (Fig. 6A) were identical to that of 5-HT-immunoreactive neurons (not illustrated); i.e., they are present in four clusters (dorsomedian, ventromedian, lateral and caudal), are medium to large in size, and fusiform or ovoid in shape. SERT-IR fibers were dense in the region of the PPT, particularly in the dorsal region of the pars compacta (Fig. 6B). At least two classes of SERT-IR fibers were observed. The majority of axons were characterized by a uniformly thin caliber and were lightly studded with small boutons. A T.L. STEININGER ET AL. second type of labeled axon displayed a varying, but typically greater caliber along its course and was interrupted less regularly by fairly large varicosities. Putative contacts were observed between SERT-IR varicosities and ChAT-IR elements. The pattern of SERT immunoreactivity in the PPT region was different than the pattern of anterograde labeling from the dorsal raphe, in that more SERT-IR fibers were seen in the ventrolateral tegmentum encompassing the rostral third of the PPT. Electron microscopy. SERT-IR elements were identified by the presence of amorphous electron-dense DAB reaction product, which was present throughout the cytoplasm, and accumulated at the mitochondrial, vesicular and internal surface of neuronal membranes, and ChAT-IR elements were identified by the presence of BDHC crystals (Fig. 7A). SERT-immunoreactive fibers were less dense in sections prepared for electron microscopy, likely secondary to the limited penetration of the antibody with the omission of Triton-X from the diluent. The majority of the SERT-IR elements in the dorsal PPT-pc consisted of thin unmyelinated axons and preterminal axons. No SERT-IR dendrites or somata were observed in the PPT-pc. SERT-IR terminals (Fig. 7A–G) varied in size from 0.38 to 4.05 µm in length (parallel to synaptic membrane; mean 5 1.0 6 0.89 µm) and 0.15 to 0.80 µm in width (perpendicular to synaptic membrane; mean 5 0.45 6 0.15 µm). The analysis performed in SERT/ChAT double-stained tissue was the same as that performed for the anterograde tracing experiments described above: (1) the postsynaptic targets of SERT-IR terminals were determined and quantitated (see Table 1), and (2) the proportion of SERT-IR versus unlabeled terminals synapsing with ChAT-IR neurons was determined and compared to the proportion of SERT-IR terminals synapsing in the surrounding neuropil (see Table 2). In the first analysis, greater than 170 SERT-IR vesicle-containing axons were photographed and 48 exhibiting synaptic specializations were selected for analysis. Of these 48 SERT-IR synaptic profiles, 44 (91%) were found to synapse with unlabeled dendrites (Fig. 7B–F), three (6%) were found to synapse with ChAT-IR elements (Fig. 7A), and one (2%) synapsed with an unlabeled terminal (Fig. 7G). Of the non-ChAT-IR dendrites contacted by SERT-IR synaptic profiles, 35 (80%) were small (,1 µm diameter), with 16 describing symmetric and 16 asymmetric specializations with the morphological subtype of three profiles undeterminable. Medium-sized (.1 µm but ,2 µm diameter) non-ChAT-IR dendrites contacted by SERT-IR synaptic profiles totaled eight (18%) of total dendrites contacted), with specializations evenly distributed between symmetric and asymmetric types. One (2% of total dendrites contacted) non-ChAT-IR dendrite contacted by a SERT-IR synaptic profile was large (.2 µm diameter). All three ChAT-IR dendrites contacted by SERT-IR synaptic profiles were small with one symmetric and two asymmetric specializations. Numerous ChAT-IR myelinated axons were observed (Fig. 7H), but SERT-IR myelinated axons were only rarely observed. SERT-IR terminals were frequently seen in apposition to unlabeled terminals in the neuropil. No contacts were observed between SERT-IR terminals and neuronal somata. In a separate analysis, 158 ChAT-IR dendrites were examined in SERT/ChAT double-stained material (Table 2). A total of 202 terminals were apposed to these ChAT-IR dendrites, and of those terminals, only 119 exhibited clear synaptic specializations. Of these 119 synaptic profiles, DORSAL RAPHE AFFERENTS TO THE PPT Fig. 5. A–F: Electron photomicrographs illustrating synaptic contacts between anterogradely labeled axon terminals (t) and ChAT-IR dendrites (d). G–H: Anterogradely labeled terminals in synaptic contact with both an unlabeled dendrite (d) and an unlabeled terminal 313 (t). I: Labeled myelinated axons (a). J–K: Retrogradely labeled dendrites (d). Note the contrast between the appearance of the reaction product in these non-ChAT-IR dendrites and in the ChAT-IR dendrites illustrated in A–F. Scale bar 5 0.5 µm. 314 T.L. STEININGER ET AL. Fig. 6. A and B: Serotonin transporter (SERT)/ChAT doubleimmunostaining at the light microscopic level. A: A coronal section of the DR showing the location and appearance of SERT-IR neurons. (Subdivisions of the dorsal raphe are recognized as follows: dm, dorsomedian; vm, ventromedian; lat, lateral.) B: SERT-IR fibers and ChAT-IR neurons in the dorsal part of the PPT pars compacta. Numerous SERT-IR fibers are in putative contact with ChAT-IR somata. C: Coronal tissue section stained solely for SERT-IR demon- strates dense SERT-immunoreactive fibers in the PPT-pars compacta (marked by the fibers of the commissure of the lateral lemniscus, arrows). D: In an adjacent section, the same region shows no immunoreactivity following preabsorption of the primary antiserum (rabbitanti-SERT) with excess SERT fusion protein. Arrows in D indicate all fascicles in a similar location to that seen in C. Scale bars 5 500 µm for A; 100 µm for B–D. four were SERT-IR, yielding a frequency of 3.4% of synapses on ChAT-IR dendrites that were SERT-IR. This frequency was found to vary among cases from 2.1% to 4.3%. The frequency was similar when terminal appositions (i.e., those not having clear synaptic specializations) were included in the analysis. In the neuropil surrounding these ChAT-IR dendrites, 2,976 synaptic terminals were identified. Of those terminals, 65 were SERT-IR, giving a frequency of 2.2% (ranged from 1.6% to 3.3%) of all synapses on neuropil elements that were SERT-IR. lish the presence of SERT containing terminals in the PPT-pc; and (4) demonstrate an ultrastructural similarity in the PPT-pc between terminals anterogradely labeled from the dorsal raphe and those containing SERT. Approximately 12% of anterogradely labeled dorsal raphe terminals in the PPT-pc synapsed with ChAT-IR dendrites and 85% synapsed with unlabeled dendrites, while 3% synapsed with unlabeled terminals. Axosomatic synapses were not observed. Anterogradely labeled terminals constituted 2–4% of the total synapses on cholinergic dendrites, as well as in the surrounding neuropil, indicating that raphe inputs to the cholinergic PPT neurons were no more frequent than raphe input to noncholinergic elements in the PPT region. Parallel light and electron microscopic visualizations of SERT immunoreactivity as a specific marker for serotonergic terminals confirmed the presumptive serotonergic nature of labeled DR efferents. Dual labeling of either the anterograde tracer or SERT-IR with DISCUSSION The results of the present study: (1) confirm the existence of a projection from the dorsal raphe to the PPT-pc; (2) provide the first quantitative assessment of an identified afferent to neurochemically defined elements (i.e., cholinergic versus noncholinergic) in the PPT-pc; (3) estab- DORSAL RAPHE AFFERENTS TO THE PPT Fig. 7. SERT/ChAT double-immunostaining at the ultrastructural level. A: An apposition between a small SERT-IR terminal (t) and a ChAT-IR dendrite (d), which is identified as such by the presence of large (BDHC) crystals (arrows). B–F: SERT-IR terminals (t) in apposition to unlabeled dendrites (d). In B (the synapse on the left) and F, synaptic specializations are clearly asymmetric, while other examples 315 are symmetric. G: A SERT-IR terminal in apposition to an unlabeled terminal. H: A ChAT-IR myelinated axon (a), which contains a large BDHC crystal. I: A ChAT-IR dendrite containing crystalline BDHC receives an asymmetric synaptic contact from an unlabeled terminal containing round clear vesicles. Scale bars 5 1 µm for A; 0.5 µm for B–I. 316 T.L. STEININGER ET AL. ChAT reveals novel information concerning the synaptic organization of the PPT and extends previous findings, which noted only that the majority of 5-HT-IR terminals in the rat PPT are in synaptic contact with non-ChAT-IR dendrites (Honda and Semba, 1994). Technical considerations Biotinylated dextran as an anterograde tracer. Biotinylated dextran (BD; also, biotinylated dextran-amine, BDA) has recently been introduced as a neuronal tracer (Veeman et al., 1992; Rajakumar et al., 1993; Wouterlood and Jorritsma-Byham, 1993; Naito and Kita, 1994). Anterograde tracing with BD compares favorably with the widely used anterograde tracer, PHA-L in light microscopic (Veenman et al., 1992) and electron microscopic studies (Wouterlood and Jorritsma-Byham, 1993). The advantages of BD over PHA-L are: (1) enhanced anterograde labeling; (2) ease of detection (e.g., a single incubation in avidin-biotinperoxidase complex [ABC] versus a 2–3-step immunohistochemical procedure); (3) improved ultrastructure reflecting shorter incubation times; and (4) better tissue penetration of ABC compared to immunohistochemical reagents for PHA-L visualization (Wouterlood and Jorritsma-Byham, 1993). The single disadvantage of BD is the greater tendency for this tracer to be transported retrogradely, as compared to PHA-L. In theory, axon collaterals of retrogradely labeled neurons might also be labeled, thereby yielding false-positive results. Iontophoretic application of BD, which we employed here, as opposed to pressure injection, minimizes this potential problem (Veenman et al., 1992). Iontophoretic injection of BD produced excellent anterograde labeling and a minor amount of retrograde labeling, which increased proportionately with increasing injection micropipette tip diameters, as has been suggested by others (Veenman et al., 1992). The extent of retrograde labeling probably had little effect on the results because the pattern of DR efferents visualized with BD: (1) essentially mirrored that which we have observed using PHA-L, where retrograde labeling is not observed (Steininger and Wainer, 1991); and (2) varied little across the cases analyzed, independent of the degree of retrograde labeling. That the anterogradely labeled terminals had a similar morphology and pattern of postsynaptic targets in all three cases analyzed, and that this pattern was recapitulated by SERT-IR terminals, further substantiates the results. SERT immunoreactivity as a specific marker for serotonin (5-hydroxytryptamine; 5-HT) SERT belongs to a family of Na1/Cl2 dependent amine transporters, that also includes transporters for noradrenaline and dopamine. SERT exists on the plasma membrane of serotonergic presynaptic terminals and plays a prominent role in terminating the action of synaptically released 5-HT (Amara and Kuhar, 1993; Rudnick and Clark, 1993; Barker and Blakely, 1994). The recent cloning of the SERT gene (Blakely et al., 1991) led to the development of antibodies directed at an epitope highly specific to the SERT peptide and fusion protein (Qian et al., 1995). These antibodies demonstrated SERT-IR cells and axons throughout the rat, monkey and human brain in a pattern entirely consistent with the known localization of 5-HT uptake, SERT antagonist binding and 5-HT immunoreactivity (Qian et al., 1995). We therefore felt entirely confident employing these antibodies as specific markers for seroton- ergic terminals in the present study. In all experimental material generated as a portion of this analysis, we always confirmed that the pattern of SERT-IR cells and axons throughout the brain exhibited a strict correspondence with 5-HT visualized in adjacent sections. SERT expression has previously been attributed to both neuronal and glial elements (Anderson et al., 1992). SERT mRNA, for example, has been localized to neurons of the midbrain raphe (Blakely et al., 1991; Fujita et al., 1993; Lesch et al., 1993), as well as the frontal cortex which is devoid of serotonergic cell bodies, but rich in serotonergic innervation (Lesch et al., 1993). The latter finding suggests either transport of mRNA from dorsal raphe neurons for local translation, or a postsynaptic and possibly glial localization, given the well-known uptake of 5-HT by glia. In the present analysis, however, SERT-IR elements in the dorsal PPT-pc consisted only of thin unmyleinated axons and preterminal axons with no SERT-IR dendrites, neuronal somata or glia observed. The absence of SERT-IR in glia in the rat PPT-pc might reflect a unique feature of the rat PPT-pc or a true false-negative secondary to methodological considerations. Glia in the PPT-pc, for example, might express a unique SERT, generated by alternative splicing, which lacks the epitope recognized by our antibodies. Serotonin uptake in cultured glia may alternatively reflect a feature unique to cultured, reactive astrocytes and less so of glia in vivo. Single and dual labeling immuno-electron microscopy. The first methodological consideration relates to the potential sampling bias against distal dendrites of cholinergic neurons, which is a well-recognized limitation of immuno-electron microscopy. Although the majority of ChAT-IR dendrites visualized here were small (,1 µm, mean average diameter, see Table 1), the proportion and density of a labeled afferent (e.g., anterogradely labeled or SERT-IR) contacting ChAT-IR dendrites should be considered minimal estimates. Underestimation of distal dendrites may be a particular concern in dual immunolabeling, with BDHC as the second chromogen secondary to the limited penetration of the large, granular BDHC precipitate within small caliber dendrites, its relative solubility, and inconsistency in the BDHC visualization reaction, as has been previously noted (Lapper and Bolam, 1992). In the present studies, however, ChAT-IR dendrites visualized with BDHC were as abundant as those visualized with DAB (e.g., BDHC labeled dendrites were 7.7% of dendrites in the neuropil, compared to 6–14% seen in DAB material (see Steininger et al., 1997). A second methodological consideration relates to falsepositives that might occur secondary to difficulties in discriminating between two chromogens employed in dual immunolabeling procedures. Sequential immunoperoxidase staining with DAB and BDHC avoids this problem, since the two electron-dense markers contrast sharply, and are therefore reliably distinct at both the light and electron microscopic levels (Bolam et al., 1986, 1991; Levey et al., 1986; Zaborszky and Heimer, 1989). Combinations employing NiDAB and DAB as the two chromogens have also been extensively utilized in light and electron microscopic dual antigen localization studies (Zaborszky and Heimer, 1989; Lapper and Bolam, 1992; Zaborszky et al., 1993). At the electron microscopic level, the NiDAB and DAB reaction products were both amorphous, but were judged as distinct on the basis of the greater abundance and electrondense nature of the NiDAB reaction product, as has been DORSAL RAPHE AFFERENTS TO THE PPT noted by other investigators (Lapper and Bolam, 1992; Zaborszky et al., 1993). Misidentification of NiDAB (i.e., anterogradely labeled) as DAB (i.e., ChAT-IR) in terminals, and vice versa, would not be expected to significantly alter the results of our analysis, because in material processed only for ChAT-IR: (1) the frequency of ChAT-IR terminals in the PPT-pc (0.4–0.7%; Steininger et al., 1997) is much less than that of DR terminals (2–4%); and (2) innervation of ChAT-IR dendrites or somata by ChAT-IR terminals was not observed (Steininger et al., 1997; see, however, Honda and Semba, 1995). The labeled synaptic profiles contacting ChAT-IR dendrites analyzed in the present studies are therefore likely to represent anterograde labeling or SERT-IR, and not ChAT-IR. Several additional sources for false-positives in dual immunolabeling techniques exist and have been previously discussed in detail (Bolam et al., 1986; Levey et al., 1986). Tissue processed to control for interactions between the individual immunolabeling procedures indicated that potential false-positive staining was not encountered in the present studies. The reliability of the results is further supported by the observation that the two methods of visualizing DR efferents to the PPT-pc (e.g., anterograde labeling and SERT immunolabeling) produced very much the same results independent of the dual chromogen combination (e.g., NiDAB/DAB or DAB/BDHC) employed. Anterograde labeling of the dorsal raphe-PPT-pc projection Light microscopy. Injections of BD in the dorsal raphe anterogradely labeled terminals were concentrated in caudal and dorsal aspects of the PPT (i.e., the PPT-pc) and the LDT. Anterogradely labeled terminals were less evenly distributed throughout the tegmentum encompassing the full extent of the PPT, when compared to 5-HT-IR or SERT-IR axons and varicosities. This observation reflects either suboptimal labeling of DR efferents with the anterograde tracer or the presence of additional innervation to the PPT from other serotonergic nuclei such as the median raphe, supralemniscal nucleus (i.e., the B9 cell group) or raphe magnus (e.g., see Semba and Fibiger, 1992; Honda and Semba, 1994). Previous investigations have noted that anatomical tracer injections involving, but not restricted to, the PPT retrogradely label DR neurons (Semba and Fibiger, 1992; Steininger et al., 1992), while anterogradely labeled DR axons invest the dorsolateral mesopontine tegmentum containing the PPT (and other cell populations; Bobillier et al., 1975; Pierce et al., 1976; Moore et al., 1978; Conrad et al., 1991; Vertes, 1991; Vertes and Kocsis, 1994). These descriptions of the DR, as a source of PPT afferents, lack specificity because the methodologies, applied by themselves, cannot differentiate between the several distinct ‘‘nuclei’’ within the PPT region where an additional partial comingling of cholinergic and noncholinergic exists (Rye et al., 1987). Retrograde tracer injections, for example, cannot be restricted to a single cell population in the PPT region, so that this methodology identifies only putative afferents to cholinergic PPT neurons. Moreover, as has been previously demonstrated that lack of an appreciation for the detailed cyto- and chemoarchitecture of the dorsolateral mesopontine tegmentum can lead to misidentification of cholinergic PPT neurons as the primary recipients of various labeled afferents (see, e.g., Rye et al., 1987, 1996). The combination of anterograde labeling and immunocytochemistry, for ChAT employed here, 317 avoids these methodological concerns in revealing that cholinergic PPT neurons were invested by a moderately dense plexus of anterogradely labeled DR fibers, in accord with our previous experience with PHA-L (Steininger and Wainer, 1991). Numerous labeled varicosities were observed in putative contact with ChAT-IR somata and proximal dendrites (Fig. 2B,C). Electron microscopy. The postsynaptic targets of raphe-tegmental projections were determined by ultrastructural analysis of anterograde tracing combined with ChAT immunohistochemistry. Quantitative analysis demonstrated that 85% of terminals synapsed with unlabeled dendrites, 12% with ChAT-IR dendrites and 3% with unlabeled terminals. These results might suggest that anterogradely labeled synapses with ChAT-IR elements were relatively infrequent. Because ChAT-IR dendrites comprise only 6–14% of dendrites in the neuropil (Steininger et al., 1997), only 12% of terminals synapsed with ChAT-IR dendrites indicates a relatively even distribution of the projection between cholinergic and noncholinergic dendrites. The proportion of anterogradely labeled terminals synapsing with ChAT-IR dendrites (2–4%), for example, was determined to be similar to that of anterogradely labeled terminals synapsing in the neuropil. Analysis of ChAT-IR dendrites demonstrated that anterogradely labeled terminals accounted for 2–4% of their total synaptic contacts. This proportion should be considered a minimum for dorsal raphe inputs to the PPT, due to experimental limitations such as the inability to involve the entire DR in the injection site, penetration of visualization reagents, and difficulties in the preservation of membrane and synaptic structures. Because this is the first quantitative study of afferent innervation of the cholinergic PPT neurons, there is no basis for comparison of the proportion of inputs from the dorsal raphe with other sources of synaptic input. The proportion of input to PPT neurons from the DR (2–4%), which was likely to be an underestimate of the true level of innervation, however, represents an appreciable input. This raphe-PPT input, for example, approximates that described for thalamocortical synapses on pyramidal cell dendrites (6–20%; White and Hersch, 1981), 1.5–6.8% for corticocortical neurons (White and Hersch, 1982), and only 0.3–0.9% for corticostriatal synapses on medium spiny cells (Hersch and White, 1982). SERT-immunoreactive terminals in the PPT-pc In the present light microscopic experiments, a rather dense plexus of SERT-IR axonal processes was observed in the region of cholinergic PPT neurons. Labeled axons and varicosities were more evenly distributed along the full rostral-to-caudal extent of ChAT-IR PPT somata when compared to the distribution of anterogradely labeled DR efferents, as has been discussed above. This observation reflects either less than maximal labeling of DR efferents with the anterograde tracer or the presence of additional serotonergic innervation to the PPT from other nuclei, such as the median raphe, supralemniscal nucleus (i.e., the B9 cell group) or raphe magnus (e.g., see Semba and Fibiger, 1992; Honda and Semba, 1994). At the ultrastructural level, SERT immunoreactivity was localized primarily to presynaptic axonal elements and much less so to axonal processes. No dendritic or somatic structures were found to be SERT-IR. SERT-IR terminals were indistin- 318 guishable from anterogradely labeled terminals on the basis of their morphology, size, vesicle morphology and presence of mitochondria. Furthermore, the type of synaptic specialization and the distribution of postsynaptic targets were similar for both classes of terminals. Both SERT-IR and anterogradely labeled DR terminals, for example, primarily contacted small caliber, noncholinergic and cholinergic dendrites in equal proportions, and were equally as likely to describe symmetrical as well as asymmetrical membrane specializations. The two classes of labeled terminals were also found to constitute 2–4% of the synaptic input to cholinergic dendrites. These observations are consistent with the hypothesis that the majority, if not all, DR efferents directed at the PPT-pc are serotonergic. Of the SERT-IR terminals analyzed, however, fewer were found to synapse with ChAT-IR dendrites when compared to anterogradely labeled DR efferents (i.e., 6% vs. 12%). The most parsimonius explanation for this discrepancy is that SERT-IR may have been suboptimally labeled secondary to methodological considerations. Immuno-electron microscopy, for example, is well recognized to be plagued by perfusion/fixation conditions that may adversely affect antigenicity and/or penetration of immunoreagents, especially when compared to the less stringent conditions required for visualization of BD. Alternatively, decreased labeling for SERT-IR, when compared to DRlabeled efferents, may reflect the presence of other neurochemical inputs (i.e., nonserotonergic projections) from the DR to ChAT-IR dendrites in anterogradely labeled material. Although it is known that 40–60% of subprimate DR neurons do not employ 5-HT as a neurotransmitter (Moore, 1981; Descarries et al., 1982; Léger and Wiklund, 1982), quantitative data on the neurotransmitter content of the nonserotonergic DR neurons and/or potential coexpression of neurotransmitters are either conflicting or sparse. As many as 10% of neurons in the DR appear to be GABAergic (Stamp and Semba, 1995), with dopaminergic neurons described as sparse and diffusely distributed throughout the DR (personal observations; Ochi and Shimizu, 1978; Hökfelt et al., 1984; Geffard et al., 1987). Substance P colocalizes with 5-HT in a substantial population of DR neurons; however, the degree to which substance P localizes by itself to a unique population of DR neurons is unknown (Chan-Palay et al., 1978; Hökfelt et al., 1984; Baker et al., 1991). The projection patterns of these neurons and the remainder of non-5-HT DR somata are unknown, and whether they might project to cholinergic PPT neurons will need to be addressed in future studies. The ultrastructural characteristics of SERT-IR and anterogradely labeled DR terminals in the PPT-pc were similar to those described previously for 5-HT terminals in the rat LDT (Honda and Semba, 1994). The morphology and micro-environment of SERT-IR and/or DR-labeled terminals analyzed here also demonstrated features in common with 5-HT terminals described in several other brain regions. In several nuclei of the basal ganglia (Pasik et al., 1984; Corvaja et al., 1993), for example, a high proportion (up to 40%) of 5-HT terminals demonstrated subjunctional dense bodies, compared to the 16% of asymmetrical terminals containing subjunctional dense bodies recognized here. Although no attempts were made to quantify and distinguish features of the micro-environment of SERT-IR and/or DR-labeled terminals from nonlabeled terminals, it was our impression that they were more frequently apposed to unlabeled axon terminals and fewer T.L. STEININGER ET AL. displayed classic synaptic junctions when compared to unlabeled terminals. These features have previously been noted in serial thin sections of 5-HT terminals in the rat striatum (Soghomonian et al., 1989) and hippocampus (Oleskevich et al., 1991). The findings that 5-HT innervation is largely nonjunctional and possibly presynaptic, support arguments that not only in these brain regions, but also in the PPT, 5-HT may act by volume transmission at pre- as well as postsynaptic sites (Fuxe and Agnati, 1991). Immuno-electron microscopy for simultaneous visualization of pre- and postsynaptic markers of serotonergic innervation will be a necessary anatomical approach to further validate hypotheses regarding volume transmission. Finally, in noting that ,50% of SERT-IR and/or DR-labeled synaptic contacts make asymmetrical membrane specializations in the PPT-pc, our results vary from the rat LDT where membrane specializations described by 5-HT-IR were mostly (,70%) asymmetrical (Honda and Semba, 1994). Asymmetrical synaptic specializations also appear to be characteristic of 5-HT-containing terminals in other brain regions, such as the striatum (Pasik et al., 1984; Corvaja et al., 1993), globus pallidus (Pasik et al., 1984), substantia nigra (Corvaja et al., 1993), and hippocampus (Oleskevich et al., 1991). Our tendency to find a lesser frequency of asymmetric membrane specializations reflects either a unique feature of the rat PPT-pc, age or strain differences in the rats investigated, or possibly an inability to accurately distinguish synaptic features secondary to inadequate tissue preservation of ‘‘masking’’ by peroxidase-benzidine precipitates. Physiological and behavior significance of 5-HT in the PPT-pc The effect of serotonin on guinea pig PPT neurons (Muhlethaler et al., 1990; Leonard and Llinas, 1994) and rat LDT neurons (Luebke et al., 1992) has been studied with intracellular electrophysiological recording in the in vitro slice preparation. These studies demonstrate that serotonin potently hyperpolarizes up to 90% of physiologically (Muhlethaler et al., 1990) or histologically identified (Leonard and Llinas, 1994) cholinergic PPT neurons via activation of an outward K1 conductance. This effect appears to be mediated by postsynaptic 5-HT1 receptors, because hyperpolarization was unaffected by tetrodotoxin, blocked by the 5-HT antagonistic properties of spiperone (Leonard and Llinas, 1994), and mimicked by the agonists (1)-8-OH-DPAT (Leonard and Llinas, 1994) or 5-carboxamidotryptamine (Luebke et al., 1992). These findings are difficult to reconcile with autoradiographic ligand binding studies which have failed to localize 5-HT1 receptors in the PPT or LDT (Pazos and Palacios, 1985). Moreover, while intense 5-HT2 receptor immunoreactivity has been localized to cholinergic PPT and LDT neurons, but not noncholinergic neurons, in the rat (Morilak et al., 1993; Morilak and Ciaranello, 1993), this receptor subtype appears to mediate depolarizing responses (Bobker and Williams, 1991). The serotonin receptor subtype-mediating hyperpolarization of cholinergic PPT neurons, therefore, remains unclear and is certain to attract renewed attention as more specific and sensitive techniques for investigation of multiple 5-HT receptor subtypes become available. The intrinsic electrophysiological properties of cholinergic LDT/PPT neurons recognized by different laboratories are in disagreement, so it remains unclear how the excitability and discharge patterns of cholinergic cells are DORSAL RAPHE AFFERENTS TO THE PPT affected by the presence or absence of serotonin. Many cholinergic LDT cells from neonatal rat, for example, display low-threshold bursting secondary to activation of a voltage-sensitive calcium current in vitro (Luebke et al., 1992). It has therefore been hypothesized that during waking, when serotonergic dorsal raphe neuronal activity is maximal (McGinty and Harper, 1976; Trulson et al., 1981), bursting cholinergic neurons are relatively quiescent unless subject to sufficient depolarizing influences (Steriade et al., 1990b; Steriade, 1992). Conversely, during REM sleep when serotonergic dorsal raphe neurons cease firing (McGinty and Harper, 1976; Trulson et al., 1981), inhibition of cholinergic bursting neurons is removed. On the basis of these data, generation of a burst firing mode that favors PGO wave production during REM sleep requires deinactivation of the low-threshold calcium current by transient hyperpolarizing synaptic events. Both local cholinergic (Luebke et al., 1992; Leonard and Llinas, 1994) and GABAergic basal ganglia afferents (Datta et al., 1991; Rye et al., 1996; Rye and Bliwise, 1997) have been proposed as sources for these hyperpolarizing influences. Further evidence in support of a serotonergic dorsal raphe modulation of the discharge pattern of bursting cholinergic LDT/PPT neurons, and thereby PGO wave production, derives from several observations: (1) DR neurons cease firing immediately prior to the onset of each PGO spike (McGinty and Harper, 1976); (2) PGO waves are ‘‘released’’ in wake, and non-REM (NREM) sleep follows chemical depletion of serotonin (Dement et al., 1972; Jouvet, 1972), reversible cooling or lesions of the raphe (Jouvet, 1972), or parasaggital cuts that isolate the raphe from the PPT (Simon et al., 1973); (3) PGO burst neurons are antidromically activated from the lateral geniculate nucleus (McCarley et al., 1978; Sakai and Jouvet, 1980; Nelson et al., 1983; Sakai, 1985), whose exclusive source of tegmental afferents derives from cholinergic neurons in the PPT, and to a lesser extent, LDT (Sakai, 1980; Hallanger et al., 1987; Bickford et al., 1995; Wilson et al., 1995; J.R. Wilson and D.B. Rye, personal observations); and (4) PGO waves can be completely suppressed by injections of cholinergic (nicotinic) antagonists into the lateral geniculate nucleus (RuchMonachon et al., 1976; Hu et al., 1988), and lesions (Sakai, 1980; Webster and Jones, 1988) or reversible cooling (Laurent and Alayaguerrero, 1975) of the PPT region. The modest serotonergic dorsal raphe innervation to cholinergic PPT neurons demonstrated in the present study provides the first anatomical evidence in support of the hypothesis that 5-HT derived from the DR modulates phasic REM sleep phenomenology via modulation of the discharge of bursting cholinergic neurons. Inconsistent with this hypothesis are independent in vitro analyses of PPT neurons in the adult rat (Kang and Kitai, 1990) and guinea pig (Leonard and Llinas, 1994), and in vivo recordings in the adult cat (Steriade et al., 1990b), that suggest that the vast majority of low-threshold burst neurons are noncholinergic while nonbursting neurons are cholinergic. The validity of this hypothesis is also questioned by observations that the release of PGO waves requires lesions of several serotonergic raphe nuclei rather than lesions involving only the dorsal raphe (Simon et al., 1973). Our findings of a more robust and uniform innervation of the PPT region, particularly its rostal portion, by SERT-IR when compared to labeled DR efferents suggest the existence of an additional serotonergic innervation to the PPT. In fact, other investigators have noted potential serotoner- 319 gic afferents to the PPT from the median raphe, supralemniscal nucleus (i.e., the B9 cell group) and/or raphe magnus (e.g., see Semba and Fibiger, 1992; Honda and Semba, 1994). It remains to be determined if these putative afferents to the PPT are indeed serotonergic, and what their role is, if any, in modulating REM sleep. Models of serotonergic dorsal raphe/PPT interactions and REM sleep control that focus solely on PGO burst/ cholinergic neurons largely ignore the status of the substantial populations of nonbursting and/or noncholinergic (Rye et al., 1987) neurons in the LDT/PPT region. Based on in vivo studies, for example, PGO burst neurons represent only 9% of all neurons in the LDT/PPT region, with non-PGO-related neurons and PGO-related, but nonbursting, neurons constituting 76% and 15% of the total recorded cell population, respectively (Steriade et al., 1990a,b). Tonically active, non-PGO-related neurons with efferent projections outside of the lateral geniculate nucleus may modulate tonic aspects of REM sleep such as muscle atonia and/or EEG desynchronization (Steriade and McCarley, 1990). The majority of these cells, particularly those which are cholinergic, are also hyperpolarized by 5-HT (Luebke et al., 1992; Leonard and Llinas, 1994). The effects of 5-HT on noncholinergic neurons in the LDT/PPT region are less well described, possibly because their small size precludes ready intracellular sampling and voltage clamping. In the neonatal rat LDT, 6/7 bursting noncholinergic neurons were hyperpolarized by 5-HT, whereas only 2/8 nonbursting, noncholinergic neurons were similarly hyperpolarized (Luebke et al., 1992). In the adult guinea pig PPT, the hyperpolarizing effects of 5-HT on noncholinergic neurons are less ubiquitous, and when observed are generally weak (Leonard and Llinas, 1994; C. Leonard, personal communication). One might have predicted a more widespread response, given the present observations that dorsal raphe/serotonergic afferents to noncholinergic elements in the PPT region are as prevalent as those on cholinergic dendrites. These noncholinergic neurons are neurochemically and connectionally heterogeneous, and it has also been speculated that at least some are GABAergic interneurons (Scarnati et al., 1988; Jones, 1991). Putative dorsal raphe/serotonergic projections to noncholinergic interneurons may indirectly affect the functioning of cholinergic PPT neurons. Alternatively, dorsal raphe/serotonergic afferents of the PPT region may contact noncholinergic, presumptively glutamatergic, neurons which project to the basal ganglia (Rye et al., 1987; Lee et al., 1988; Lavoie and Parent, 1994a,b), midline thalamic nuclei (Hallanger et al., 1987), and pontine (Mitani et al., 1988; Yasui et al., 1990; Lai et al., 1993) and medullary (Rye et al., 1988; Shiromani et al., 1990; Yasui et al., 1990) reticular fields. It will be critical to determine if neurons classified as to their cholinergic/noncholinergic status, as well as their bursting status, display unique patterns of efferents, and if they do, whether they are differentially affected by serotonin. Such information will come when antidromic identification of a neuron’s projection pattern is combined with a determination of its neurochemical identity in the in vitro slice preparation. In summary, serotonergic/dorsal raphe projections to the PPT region are particularly dense in the dorsal, compact portion of the nucleus where they primarily contact small caliber, noncholinergic and cholinergic dendrites in equal proportions. This afferent source of the PPT region is likely to play a key role in neuromodulation of behavioral state 320 T.L. STEININGER ET AL. given the DR’s well-described state-related alterations in activity; i.e., serotonergic/dorsal raphe ‘‘tone’’ is a potent determinant of the ultimate discharge pattern of a majority of cholinergic and noncholinergic neurons in the PPT region. Specific REM sleep-related physiological activity is then dependent not only on the intrinsic electrophysiological properties of individual cholinergic and noncholinergic neurons, but also on their unique afferent connectivity (i.e., excitatory or inhibitory synaptic inputs). Putative sources of these afferent connections are numerous and originate from widespread brain regions (Steininger et al., 1992). A more comprehensive picture of the determinants of PPT activity will emerge when detailed quantitative analyses of the synaptology of identified afferents, as presented here, is applied to some of these putative afferent sources. Determination of the uniqueness of the efferent connections of neurons contacted by these afferents may ultimately identify specific neuronal populations and neural circuits that subserve individual components of REM sleep (e.g., see Discussion in Rye et al., 1996). Elucidation of the detailed synaptic circuitry of the PPT region is likely to impact directly upon our understanding of diseases characterized by abnormalities of REM sleep such as depression (McCarley, 1982; Steriade and McCarley, 1992; Gillin et al., 1993), schizophrenia (Karson et al., 1991; Yeomans, 1995), narcolepsy (Aldrich, 1991), and REM sleep behavior disorder (Rye and Bliwise, 1997). ACKNOWLEDGMENTS The authors express their thanks to Drs. Steven Hersch and Allan Levey for their critical insights into the experimental design and analysis as well as reading of the manuscript. The secretarial assistance of Ms. Katy Hair and the technical assistance of S. Edmunds and H. Rees are also gratefully acknowledged. This work was supported by the National Institutes of Health, NS-17661-14 (T.L.S.,B.H.W.), DA-07390 (R.D.B.), and a Cotzias Fellowship Award from the American Parkinson’s Disease Association (D.B.R.). LITERATURE CITED Aldrich, M. (1991) The neurobiology of narcolepsy. Trends Neurosci. 14:235–239. Amara, S., and M. Kuhar (1993) Neurotransmitter transporters: Recent progress. Ann. Rev. Neurosci. 16:73–93. Anderson, E., D. McFarland, and H. Kimelberg (1992) Serotonin uptake by astrocytes in situ. Glia 6:154–158. Baker, K.G., G.M. Halliday, J.-P. Hornung, L.B. Geffen, R.G.H. Cotton, and I. Törk (1991) Distribution, morphology and number of monoaminesynthesizing and substance P-containing neurons in the human dorsal raphe nucleus. Neuroscience 42:757–775. Barker, E., and R. Blakely (1994) Norepinephrine and serotonin transporters: Molecular targets of antidepressant drugs. In F. Bloom and D. Kupfer (eds): Psychopharmacology: The Fourth Generation of Progress. New York: Raven Press. pp. 321–333. Bickford, M., A. Günlük, D. Godwin, J. Gnadt, and S. Sherman (1995) Subcortical extraretinal projections to the monkey LGN. Neurosci. Abstr. 21:658. Blakely, R.D., H.E. Berson, R.T. Fremeau, Jr., M.G. Caron, M.M. Peek, H.K. Prince, and C.C. Bradley (1991) Cloning and expression of a functional serotonin transporter from rat brain. Nature 354:66–70. Bobillier, P., D. Salvert, M. Ligier, and S. Seguin (1975) Differential projections of the nucleus raphe dorsalis and nucleus raphe centralis as revealed by autoradiography. Brain Res. 85:205–210. Bobker, D.H., and J.T. Williams (1991) Ion conductances affected by 5-HT receptor subtypes in mammalian neurons. Trends Neurosci. 13:169–173. Bolam, J.P., C.A. Ingham, P.N. Izzo, A.I. Levey, D.B. Rye, A.D. Smith, and B.H. Wainer (1986) Substance P-containing terminals in synaptic contact with cholinergic neurons in the neostriatum and basal forebrain: A double immunocytochemical study in the rat. Brain Res. 397: 279–289. Bolam, J.P., C.M. Francis, and Z. Henderson (1991) Cholinergic input to dopaminergic neurons in the substantia nigra: A double immunocytochemical study. Neuroscience 41:483–494. Bowker, R., and A. Morrison (1976) The startle reflex and PGO spikes. Brain Res. 102:185–190. Chan-Palay, V., G. Jonsson, and S. Palay (1978) Serotonin and substance P coexist in neurons of the rat’s central nervous system. PNAS (USA) 75:1582–1586. Conrad, L.C.A., C.M. Leonard, and D.W. Pfaff (1991) Connections of the median and dorsal raphe nuclei in the rat: An autoradiographic and degeneration study. J. Comp. Neurol. 156:179–206. Corvaja, N., G. Doucet, and J. Bolam (1993) Ultrastructure and synaptic targets of the raphe-nigral projection in the rat. Neuroscience 55:417–427. Datta, S., R.C. Dossi, D. Pare, G. Oakson, and M. Steriade (1991) Substantia nigra reticulata neurons during sleep–waking states: Relation with ponto-geniculo-occipital waves. Brain Res. 566:344–347. De Lima, A.D., and W. Singer (1987) The brainstem projection to the lateral geniculate nucleus in the cat: Identification of cholinergic and monoaminergic elements. J. Comp. Neurol. 259:92–121. Dement, W.C., M.M. Mittler, and S.J. Henriksen (1972) Sleep changes during chronic administration of parachlorophenylalanine. Rev. Canad. Biol. 31:239–246. Descarries, L., K.C. Watkins, S. Garcia, and A. Beaudet (1982) The serotonin neurons in nucleus raphe dorsalis of adult rat: A light and electron microscope radioautographic study. J. Comp. Neurol. 207:239– 254. Fujita, M., S. Shimada, H. Maeno, T. Nishimura, and M. Tohyama (1993) Cellular localization of serotonin transporter mRNA in the rat brain. Neurosci. Lett. 162:59–62. Fuxe, K., and L. Agnati (1991) Volume Transmission in the Brain: Novel Mechanisms for Neural Transmission. New York: Raven Press. Geffard, M., S. Tuffet, N. Mons, and J.-L. Chagnaud (1987) Simultaneous detection of indoleamines and dopamine in rat dorsal raphe nuclei using specific antibodies. Histochem. 88:61–64. Gillin, J., R. Salin-Pascual, J. Velazquez-Moctezuma, P. Shiromani, and R. Zoltoski (1993) Cholinergic receptor subtypes and REM sleep in animals and normal controls. Prog. Brain Res. 98:379–387. Hallanger, A.E., and B.H. Wainer (1988) Ascending projections from the pedunculopontine tegmental nucleus and the adjacent mesopontine tegmentum in the rat. J. Comp. Neurol. 274:483–515. Hallanger, A.E., A.I., Levey, H.J. Lee, D.B. Rye, and B.H. Wainer (1987) The origins of cholinergic and other subcortical afferents to the thalamus in the rat. J. Comp. Neurol. 262:105–124. Hancock, M. (1986) Two-color immunoperoxidase staining: Visualization of anatomic relationships between immunoreactive neural elements. Am. J. Anat. 175:343–352. Hersch, S.M., and E.L. White (1982) A quantitative study of the thalamocortical and other synapses in layer IV of pyramidal cells projecting from mouse SmI cortex to the caudate-putamen nucleus. J. Comp. Neurol. 211:217–225. Hökfelt, T., R. Mårtensson, A. Björklund, S. Kleinau, and M. Goldstein (1984) Distribution maps of tyrosine hydroxylase immunoreactive neurons in the rat brain. In: T. Hökfelt and A. Björklund (eds): Handbook of Chemical Neuroanatomy, Classical Transmitters in the CNS, Vol. 2. Amsterdam: Elsevier, pp. 277–379. Honda, T., and K. Semba (1994) Serotonergic synaptic input to cholinergic neurons in the rat mesopontine tegmentum. Brain Res. 647:299–306. Honda, T., and K. Semba (1995) An ultrastructural study of cholinergic and non-cholinergic neurons in the laterodorsal and pedunculopontine tegmental nuclei in the rat. Neuroscience 68:837–853. Hsu, S.-M., and E. Soban (1982) Color modification of diaminobenzidine (DAB) precipitation by metallic ions and its application for double immunohistochemistry. J. Histochem. Cytochem. 30:1079–1082. Hu, B., D. Bouhassira, M. Steriade, and M. Deschenes (1988) The blockage of ponto-geniculo-occipital waves in the cat lateral geniculate nucleus by nicotinic antagonists. Brain Res. 437:394–397. Hu, B., M. Steriade, and M. Deschênes (1989) The cellular mechanism of thalamic ponto-geniculo-occipital waves. Neuroscience 31:25–35. Jones, B. (1991) Paradoxical sleep and its chemical and structural substrates in the brain. Neuroscience 40:637–656. DORSAL RAPHE AFFERENTS TO THE PPT Jouvet, M. (1972) The role of monoamine and acetylcholine-containing neurons in the regulation of the sleep-wake cycle. Ergbn. Physiol. 64:166–307. Kang, Y., and S. Kitai (1990) Electrophysiological properties of pedunculopontine neurons and their postsynaptic responses following stimulation of substantia nigra reticulata. Brain Res. 535:79–95. Karson, C., E. Garcia-Rill, J. Biedermann, R. Mrak, M. Husain, and R. Skinner (1991) The brain stem reticular formation in schizophrenia. Psych. Res.: Neuroimaging 40:31–48. Kitsikis, A., and M. Steriade (1981) Immediate behavioral effects of kainic acid injections into the midbrain reticular core. Behav. Brain Res. 3:361–380. Lai, Y., J. Clements, and J. Siegel (1993) Glutamatergic and cholinergic projections to the pontine inhibitory area identified with horseradish peroxidase retrograde transport and immunohistochemistry. J. Comp. Neurol. 336:321–330. Lapper, S.R., and J.P. Bolam (1992) Input from the frontal cortex and the parafascicular nucleus to cholinergic interneurons in the dorsal striatum of the rat. Neuroscience 51:533–545. Laurent, J.P. and F. Alayaguerrero (1975) Reversible suppression of ponto-geniculo-occipital waves by localized cooling during paradoxical sleep in cat. Exp. Neurol. 49:356–369. Lavoie, B., and A. Parent (1994a) Pedunculopontinenucleus in the squirrel monkey: Cholinergic and glutamaterigc projections to the substantia nigra. J. Comp. Neurol. 344:232–241. Lavoie, B., and A. Parent (1994b) Pedunculopontine nucleus in the squirrel monkey: Projections to the basal ganglia as revealed by anterograde tract-tracing methods. J. Comp. Neurol. 344:210–231. Lee, H.J., D.B. Rye, A.E. Hallanger, A.I. Levey, and B.H. Wainer (1988) Cholinergic vs. noncholinergic efferents from the mesopontine tegmentum to the extrapyramidal motor system nuclei. J. Comp. Neurol. 275:469–492. Léger, L., and L. Wiklund (1982) Distribution and numbers of indolamine cell bodies in the cat brainstem determined with Falck-Hillarp fluorescence histochemistry. Brain Res. Bull. 9:245–251. Leonard, C., and R. Llinas (1994) Serotonergic and cholinergic inhibition of mesopontine cholinergic neurons controlling REM sleep: An in vitro electrophysiological study. Neurosci. 59:309–330. Lesch, K.P., C.S. Aulakh, B.L. Wolozin, T.J. Tolliver, J.L. Hill, and D.L. Murphy (1993) Regional brain expression of serotonin transporter mRNA and its regulation by reuptake inhibiting antidepressants. Molec. Brain Res. 17:31–35. Levey, A.I., D.M. Armstrong, S.F. Atweh, R.D. Terry, and B.H. Wainer (1983) Monoclonal antibodies to choline acetyltransferase: Production, specificity, and immunohistochemistry. J. Neurosci. 3:1–9. Levey, A.I., J.P. Bolam, D.B. Rye, A.E., Hallanger, R.M. Demuth, M.-M. Mesulam, and B.H. Wainer (1986) A light and electron microscopic procedure for sequential double antigen localization using diaminobenzidine and benzidine dihydrochloride. J. Histochem. Cytochem. 34:1449– 1457. Luebke, J., R. Greene, K. Semba, A. Kamondi, R. McCarley, and P. Reiner (1992) Serotonin hyperpolarizes cholinergic low-threshold burst neurons in the rat laterodorsal tegmental nucleus in vitro. PNAS (USA) 89:743–747. McCarley, R. (1982) REM sleep and depression: Common neurobiological control mechanisms. Am. J. Psychiatr. 139:565–570. McCarley, R.W., J.P. Nelson, and J.A. Hobson (1978) Ponto-geniculooccipital (PGO) burst neurons: Correlative evidence for neuronal generators of PGO waves. Science 201:269–272. McCormick, D. (1989) Cholinergic and noradrenergic modulation of thalamocortical processing. Trends Neurosci. 12:215–221. McGinty, D., and R. Harper (1976) Dorsal raphe neurons: Depression of firing during sleep in cats. Brain Res. 101:569–575. Mitani, A., K. Ito, A. Hallanger, B. Wainer, K. Kataoka, and R. McCarley (1988) Cholinergic projections from the laterodorsal and pedunculopontine tegmental nuclei to the pontine gigantocellular tegmental field in the cat. Brain Res. 451:397–402. Moore, R.Y. (1981) The anatomy of central serotonin neuron systems in the rat brain. In B.L. Jacobs and A. Gelperin (eds): Serotonin Neurotransmission and Behavior. Cambridge, Mass.: MIT Press, pp. 35–71. Moore, R.Y., A.E. Halaris, and B.E. Jones (1978) Serotonin neurons of the midbrain raphe: Ascending projections. J. Comp. Neurol. 180:417–438. Morilak, D., and R. Ciaranello (1993) 5-HT2 receptor immunoreactivity on cholinergic neurons of the pontomesencephalic tegmentum shown by double immunoflourescence. Brain Res. 627:49–54. 321 Morilak, D.A., S.J. Garlow, and R.D. Ciaranello (1993) Immunocytochemical localization and description of neurons expressing serotonin2 receptors in the rat brain. Neuroscience 54:701–717. Muhlethaler, M., A. Khateb, and M. Serafin (1990) Effects of monoamines and opiates on pedunculopontine neurones. In: M. Mancia and G. Marini (eds): The Diencephalon and Sleep. New York: Raven Press, pp. 31–48. Naito, A., and H. Kita (1994) The cortico-nigral projection in the rat: An anterograde tracing study with biotinylated dextran amine. Brain Res. 637:317–322. Nelson, J., R. McCarley, and J. Hobson (1983) REM sleep burst neurons, PGO waves, and eye movement information. J. Neurophys. 50:784–796. Ochi, J., and K. Shimizu (1978) Occurrence of dopamine-containing neurons in the midbrain raphe nuclei of the rat. Neurosci. Lett. 8:317–320. Oleskevich, S., L. Descarries, K.C. Watkins, P. Séguéla, and A. Daszuta (1991) Ultrastructural features of the serotonin innervation in adult rat hippocampus: An immunocytochemical description in single and serial thin sections. Neuroscience 42:777–791. Pasik, P., T. Pasik, G. Holstein, and J.S. Pecci (1984) Serotonergic innervation of the monkey basal ganglia: An immunocytochemical light and electron microscopy study. In: J. McKenzie, R. Kemm and L. Wilcock (eds): The Basal Ganglia. New York: Plenum Press, pp. 115–129. Paxinos, G., and C. Watson (1986) The Rat Brain in Stereotaxic Coordinates. Orlando: Academic Press. Pazos, A., and J.M. Palacios (1985) Quantitative autoradiographic mapping of serotonin receptors in the rat brain. I. Serotonin-1 receptors. Brain Res. 346:205–230. Pierce, E.T., W.E. Foote, and J.A. Hobson (1976) The efferent connection of the nucleus raphe dorsalis. Brain Res. 107:137–144. Qian, Y., H.E. Melikian, D.B. Rye, A.I. Levey, and R.D. Blakely (1995) Identification and characterization of antidepressant-sensitive serotonin transporter proteins. J. Neurosci 15:1261–1274. Rainnie, D., H. Grunze, R. McCarley, and R. Greene (1994) Adenosine inhibition of mesopontine cholinergic neurons: Implications for EEG arousal. Science 263:689–692. Rajakumar, N., K. Eliscvich, and B. Flumerfelt (1993) Biotinylated dextran: A versatile anterograde and retrograde neuronal tracer. Brain Res. 607:47–53. Ruch-Monachon, M.A., M. Jalfre, and W. Haefely (1976) Drugs and PGO waves in the lateral genicualte body of the curarized cat. IV. The effects of acetylcholine, GABA and benzodiazepines on PGO wave activity. Arch. Int. Pharmacodyn. 219:308–325. Rudnick, G., and J. Clark (1993) From synapse to vesicle: The reuptake and storage of biogenic amine neurotransmitters. Biochim. Biophys. Acta. 1144:249–263. Rye, D., and D. Bliwise (1997) Movement Disorders Specific To Sleep And The Nocturnal Manifestations Of Waking Movement Disorders. In: R. Watts and W. Koller (eds): Movement Disorders: Neurologic Principles and Practice. New York: McGraw-Hill, Inc., pp. 687–713. Rye, D., C. Saper, H. Lee, and B. Wainer (1987) Pedunculopontine tegmental nucleus of the rat: Cytoarchitecture, cytochemistry, and some extrapyramidal connections of the mesopontine tegmentum. J. Comp. Neurol. 259:483–528. Rye, D., H. Lee, C. Saper, and B. Wainer (1988) Medullary and spinal efferents of the pedunculopontine tegmental nucleus and adjacent mesopontine tegmentum in the rat. J. Comp. Neurol. 269:315–341. Rye, D., R. Turner, J. Vitek, R. Bakay, M. Crutcher, and M. DeLong (1996) Anatomical Investigations Of The Pallidotegmental Pathway In Monkey And Man. In: H. Ohye, M. Kimura and J. McKenzie (eds): Basal Ganglia V (Proceedings of the Vth Meeting of the International Basal Ganglia Society). New York: Plenum, pp. 59–75. Sakai, K. (1980) Some anatomical and physiological properties of pontomesencephalic tegmental neurons with special reference to the PGO waves and postural atonia during paradoxical sleep in the cat. In J.A. Hobson and M.A.B. Brazier (eds): The Reticular Formation Revisited. New York: Raven Press, pp. 427–447. Sakai, K. (1985) Anatomical and physiological basis of paradoxical sleep. In D.J. McGinty, A. Morrison, R. Drucker-Colin and P.L. Parmeggiani (eds): Brain Mechanisms of Sleep. New York: Raven Press, pp. 111–137. Sakai, K., and M. Jouvet (1980) Brain stem PGO-on cells projecting directly to the cat dorsal lateral geniculate nucleus. Brain Res. 194:500–505. Sanchez, R., A. Kahteb, M. Måhlether, and C. Leonard (1991) Glutamate and NMDA actions on mesopontine cholinergic neurons in vitro. Neurosci. Abstr. 17:256. 322 Scarnati, E., F. Hajdu, C. Pacitti, and T. Tombol (1988) An EM and Golgi study on the connection between the nucleus tegmenti pedunculopontinus and the pars compacta of the substantia nigra in the rat. J. Hirnforsch. 29:95–105. Semba, K., and H. Fibiger (1992) Afferent connections of the laterodorsal and the pedunculopontine tegmental nuclei in the rat: A retro- and antero-grade transport and immunohistochemical study. J. Comp. Neurol. 323:387–410. Shiromani, P., Y. Lai, and J. Siegel (1990) Descending projections from the dorsolateral pontine tegmentum to the paramedian reticular nucleus of the caudal medulla in the cat. Brain Res. 517:224–228. Simon, R.P., M.P. Gershon, and D.C. Brooks (1973) The role of the raphe nuclei in the regulation of ponto-geniculo-occipital wave activity. Brain Res. 58:313–330. Soghomonian, J.-J., L. Descarries and K.C. Watkins (1989) Serotonin innervation in adult rat neostriatum. II. Ultrastructural features: A radioautographic and immunocytochemical study. Brain Res. 481:67– 87. Stamp, J., and K. Semba (1995) Extent of colocalization of serotonin and GABA in the neurons of the rat raphe nuclei. Brain Res. 677:39–49. Steinbusch, H.W.M., and R. Nieuwenhuys (1983) The raphe nuclei of the rat brainstem: A cytoarchitectonic and immunohistochemical study. In P.C. Emson (ed): Chemical Neuroanatomy. New York: Raven Press, pp. 131–207. Steinbusch, H.W.M., R. Nieuwenhuys, A.A.J. Verhofstad, and D. Van der Kooy (1981) The nucleus raphe dorsalis of the rat and its projections upon the caudatoputament. A combined cytoarchitectonic, immunohistochemical and retrograde transport study. J. Physiol. (Paris) 77:154– 174. Steininger, T.L., and B.H. Wainer (1991) Projections from the dorsal raphe nucleus to the pedunculopontine and laterodorsal tegmental nuclei. Neurosci. Abstr. 17:1041. Steininger, T.L., D.B. Rye, and B.H. Wainer (1992) Afferent projections to the cholinergic pedunculopontine tegmental nucleus and adjacent midbrain extrapyramidal area in the albino rat. J. Comp. Neurol. 321:515–543. Steininger, T., D. Rye, and B. Wainer (1997) Ultrastructural study of cholinergic and noncholinergic neurons in the pars compacta of the rat pedunculopontine tegmental nucleus. J. Comp. Neurol. 382:285–301. Steriade, M. (1992) Basic mechanisms of sleep generation. Neurol. 42(Suppl. 6):9–18. Steriade, M., and R. McCarley (1990) Brainstem Control of Wakefulness and Sleep. New York: Plenum Press. Steriade, M., S. Datta, D. Pare, G. Oakson, and R.C. Dossi (1990a) Neuronal activities in brain-stem cholinergic nuclei related to tonic activation processes in thalamocortical systems. J. Neurosci. 10:2541– 2559. Steriade, M., D. McCormick and T. Sejnowski (1993) Thalamocortical oscillations in the sleeping and aroused brain. Science 262:679–684. T.L. STEININGER ET AL. Trulson, M.E., B.L. Jacobs, and A.R. Morrison (1981) Raphe unit activity across the sleep-waking cycle in normal cats and in pontine lesioned cats displaying REM sleep without atonia. Brain Res. 226:75–91. Veenman, C., A. Reiner, and M. Honig (1992) Biotinylated dextran amine as an anterograde tracer for single- and double-labeling studies. J. Neurosci. Meth. 41:239–254. Vertes, R.P. (1991) A PHA-L analysis of ascending projections of the dorsal raphe nucleus in the rat. J. Comp. Neurol. 313:643–668. Vertes, R., and B. Kocsis (1994) Projections of the dorsal raphe nucleus to the brainstem: PHA-L analysis in the rat. J. Comp. Neurol. 340:11–26. Webster, H., and B. Jones (1988) Neurotoxic lesions of the dorsolateral pontomesencephalic tegmentum-cholinergic cell area in the cat. II. Effects upon sleep-waking states. Brain Res. 458:285–302. White, E.L., and S.M. Hersch (1981) Thalamocortical synapses of pyramidal cells which project from Sm1 to Ms1 cortex in the mouse. J. Comp. Neurol. 198:167–181. White, E.L., and S.M. Hersch (1982) A quantitative study of thalamocortical and other synapses involving the apical dendrites of corticothalamic projection cells in mouse SmI cortex. J. Neurocytol. 11:137–157. Williams, J., and P. Reiner (1993) Noradrenaline hyperpolarizes identified rat mesopontine cholinergic neurons in vitro. J. Neurosci. 13:3878– 3883. Wilson, J., A. Hendrickson, H. Sherk, and J. Tigges (1995) Sources of subcortical afferents to the macaque’s dorsolateral geniculate nucleus. Anat. Rec. 242:566–574. Woolf, N.J., and L.L. Butcher (1989) Cholinergic systems in the rat brain: IV. Descending projections of the pontomesencephalic tegmentum. Brain Res. Bull. 23:519–540. Woolf, N.J., J.B. Harrison, and J.S. Buchwald (1990) Cholinergic neurons of the feline pontomesencephalon. II. Ascending anatomical projections. Brain Res. 520:55–72. Wouterlood, F., and B. Jorritsma-Byham (1993) The anterograde neuroanatomical tracer biotinylated dextran-amine: Comparison with the tracer Phaseolus vulgaris-leucoagglutinin in preparations for electron microscopy. J. Neurosci. Meth. 48:75–87. Yasui, Y., D. Cechetto, and C. Saper (1990) Evidence for a cholinergic projection from the pedunculopontine tegmental nucleus to the rostral ventrolateral medulla in the rat. Brain Res. 517:19–24. Yeomans, J. (1995) Role of tegmental cholinergic neurons in dopaminergic activation, antimuscarinic psychosis and schizophrenia. Neuropsychopharmacology 12:3–16. Zaborszky, L., and L. Heimer (1989) Combinations of tracer techniques, especially HRP and PHA-L, with transmitter identification for correlated light and electron microscopic studies. In L. Heimer and L. Zaborszky (eds): Neuroanatomical Tract-Tracing Methods 2, Recent Progress. New York: Plenum Press, pp. 49–96. Zaborszky, L., W.E. Cullinan, and V.N. Luine (1993) Catecholaminergiccholinergic interaction in the basal forebrain. Prog. Brain Res. 98:31– 49.