Interactions between the basal ganglia, the pontine parabrachial region, and the trigeminal system in cat

NeuroscienceVol. 19, No. 2, pp. 411425. 1986 Printed in Great Britain

0306-4522/86$3.00+ 0.00 Pergamon Journals Ltd 0 1986IBRO

INTERACTIONS BETWEEN THE BASAL GANGLIA, THE PONTINE PARABRACHIAL REGION, AND THE TRIGEMINAL SYSTEM IN CAT J. S. SCHNEIDER Dept of Neurology, UCLA School of Medicine, University of California, Los Angeles, CA 90024, U.S.A. Abstract-Anatomical studies utilizing wheat germ lectin-bound horseradish peroxidase demonstrated direct connections between the pontine parabrachial region and the substantia nigra pars reticulata and to a lesser extent, the entopeduncular nucleus as well as a number of other forebrain regions including the amygdala, hypothalamus, thalamus, bed nucleus stria terminalis and substantia innominata. The pontine parabrachial region was also shown to receive direct inputs from the spinal trigeminal system and to send axons to areas surrounding trigeminal and hypoglossal motor areas. Once the anatomical connections were determined, electrophysiological studies were undertaken to investigate some of the functional aspects of these connections between the pontine parabrachial, basal ganglia and trigeminal systems. Extracellular single unit recordings were obtained from 228 cells in the dorsal pontine parabrachial region of the cat. These cells were tested for responsiveness to trigeminal

sensory stimulation and activation of basal ganglia outputs (i.e. substantia nigra and entopeduncular nucleus). Twenty-two percent of pontine parabrachial cells responded to only trigeminal stimulation; 4% responded to entopeduncular nucleus only; 37% responded to substantia nigra only, and 28% responded to both substantia nigra and trigeminal stimulation. Furthermore, 43% of pontine parabrachial cells with both substantia nigra and sensory response had the sensory response altered by a preceding stimulus to the substantia nigra. Thus, the substantia nigra is shown to exert influences on both the spontaneous activities and afferent responses of pontine parabrachial neurons. The significance of these findings are discussed in relation to the importance of descending basal ganglia influences and ascending influences from the pontine parabrachial region on various sensorimotor activities.

The basal ganglia have often been associated with various aspects of trigeminally mediated behaviors. Basal ganglia lesions in rats,i6 cats?’ and monkeys*O result in oral motor dysfunction and feeding deficits. Single unit activity in various basal ganglia structures (striatum, globus pallidus/entopeduncular nucleus, and substantia nigra) in cats is influenced by trigeminal sensory stimulation3’.33*35while electrical stimulation of basal ganglia structures exerts strong modulatory influences on both sensory” and motor components of the trigeminal system.3 A high proportion of single cells in the substantia nigra pars reticulata of monkeys are related to jaw, lip and tongue movements.5 In humans, basal ganglia disease virtually always results in some degree of oral motor dysfunction. I8 Tardive dyskinesia, a drug induced buccolingual motor disorder is attributed to damage in the basal ganglia.’ In short, the literature strongly suggests a basal ganglia involvement in actions mediated by the trigeminal system. However, while the basal ganglia is undoubtedly capable of influencing the trigeminal system, the basis of this interaction is Abbreviations: CN, cuneiform nucleus; CSF, cerebrospinal

fluid; C-T, conditioning-test; E-I, excitation-inhibition; EPN, entopeduncular nucleus; FTL, rostra1 reticular formation; HRP, horseradish peroxidase; LC, locus coeruleus; MLR, mesencephalic locomotor region; NTPP, nucleus tegmenti pedunculopontinus; PB, pontine parabrachial region; SN, substantia nigra; WGA-HRP, wheat germ agglutinin-horseradish peroxidase. 411

unknown since direct connections between the two systems have never been observed. The present study has been undertaken to examine the possible involvement of one particular area, the dorsal pontine parabrachial region (PB), in mediating some basal ganglia influences on trigeminally mediated activities. The PB, which extends from the rostra1 pole of the trigeminal motor nucleus to the caudal level of the inferior colliculus, was chosen for study because of its close association with the trigeminal system. Ventrocaudally, the PB is contiguous with the supratrigeminal region, which contains commissural inhibitory interneurons for the trigeminal motor system.2’ The PB receives input from the spinal trigeminal nuclei as well as the nucleus tractus solitarius (which also receives some trigeminal as well as gustatory and visceral afferents).i4a2’ The PB projects to areas around the hypoglossal and trigeminal motor nuclei and trigeminal main sensory nucleus39 as well as to the spinal trigeminal nucleus interpolar&39 cerebellar vermis3’ and spinal cord.6*‘2*39 In the present experiment we tried to find anatomical connections between the basal ganglia, PB and trigeminal system. The activity of single PB neurons was then recorded and assessed for responsiveness to trigeminal sensory stimulation and to activation of basal ganglia output structures, i.e. entopeduncular nucleus (EPN) and substantia nigra (SN). PB cells were examined for degree of basal ganglia-trigeminal convergence and for basal ganglia modulatory effects on trigeminal sensory responses.

412

J. S. SCH~~DER EXPERIMENTAL

PROCEDURES

In a total of 30 cats, afferent and efferent projections of the PB area were determined by wheatgerm &tin-bound horseradish peroxidase (WGA-HRP) labeling techniques. In all cats, WGA-HRP, 2% w/v in 0.9% NaCl solvent was used for injection Both male and female cats weighing between 2.5 and 4.0 kg were used for this study. Volumes of 0.05-0.10 ~l/injection were pressure injected via a Hamilton syringe into the PB area on one side. In 7 cats, the injection cannula was dated into the PB area (A: -3.5, L: 3.5. D: - 2.5) by driving through the cerebellum at a 62” angle from posterior to anterior to avoid hitting the tentorium. One cat received similar marker injections into the cerebellar white matter as control for marker leakage along the cannula track. One other cat received HRP injections into the trigeminal sensory nucleus on one side and the vestibular nucleus on the other side as controls for spread of WGA-HRP to PB adjacent structures. Four cats received marker injections directly into the PB area under direct visual guidance by methods similar to those previously outlined. In 4 additional cats, a related procedure was performed in which the occipital cortex was removed, a small hole was drilled in the tentorium and the injection cannula was advanced straight through the cerebellum into the PB. For all surgical procedures ketamine (8 mg/kg) and acepromazine (0.1 cc) were used as anesthetic agents. LidoCaine was also used to infiltrate exposed muscles and wound margins. Marker injections were made slowly over the course of 5-10min. Once the desired volume was injected, the cannula was left in place for 30 min, after which it was slowly removed and the animal’s wounds were closed. In an additional IO cats, WGA-HRP injections were made into the SN (n = 5) or the EPN (n = 5). For SN injections, the cannula was advanced into the area of the SN at a 45” angle off the midline (A: 4.5, L: 4.0, D: -4.5). For EPN injections, the cannula was advanced into the area of the EPN at either a 30” angle off the midline (n = 3), or by a straight drop (n = 2) (A: 11.5, L: 5.5, D: - 3.0). All surgical procedures as well as ail injection parameters were the same as described above. Forty-eight hours post-injection, the animals were deeply anesthetized with sodium pentobarbital (Nembutal) and prepared for transcardial perfusion. Heparin (10,000 units/ml, 0.2 cc, to decrease clotting) and 0.2~~ of 1% sodium nitrite (to help keep vessels open) were injected into the left ventricle and aliowed to circulate for approximately 1 min before the actual perfusion began. The animals were ‘then perfused with 0.9% NaCl (11, ZS’C), 3.8% oaraformaldehvde (2 1. 4°C. in 0.1 M uhosuhate buffer, pH = 7.4) and-graded’conckntrations of sucrose (10 and 30%, 750 ml each solution, 4”C, in 0.1 M phosphate buffer, pH = 7.4). The brains were then removed and immersed in 30% sucrose for 24 h for cryoprotection. Frozen sections (to0 pm thickness) were cut in the coronal ptane and placed in cold 0.1 M phosphate buffer @H = 7.4). Peroxidase histochemistry was by a modified de Olmos procedure’ and included nickel ammonium sulfate intensification of a 3,3’,5,5’,-tetramethylbenzidine (T’MB) reaction product. Alternate sections from each brain were counterstained with Neutral Red to reveal the cytology of the labeled neurons and nuclear groups. The tissue was subsequently examined utilizing both transmitted bight-meld and dark-field microscopy (Leitz Dialux 20 system). Camera lucida drawings were made of cell bodies in the SN following PB WGA-HRP injections and of PB cells following SN and EPN injections. These cells were subsequently measured for total somatic area, major diameter (through long axis of soma), and minor diameter (taken at a right angle to a point midway on the long axis) with micr~omputer. Cell soma diameters of WGA-HRPlabeled neurons were measured by averaging the major

diameter with the minor diameter. Comparisons of the means of average soma diameters was made using standard statistical techniques. Electrophysiology

Twenty-three adult cats weighing between 2.5 and 4.0 kg were anesthetized iv. with sodium methohexital (Brevital), a short-acting barbiturate. The femoral vein was cannulated and the trachea intubated. The animal was placed in a standard stereotaxic frame, the scalp incised, and exposed muscles reflected. A portion of the calvarium overlying the parabrachial region was removed and holes were drilled in the bone overlying the EPN, the SN, and the trigeminal ganglion, for placement of stimulating electrodes. Posterior portions of the occipital cortex and underlying white matter were removed by suction and the bony tentorium extirpated. In 10 animals, portions of the cerebellum were also removed by suction to allow direct visuali~tion of the dorsal pans and placement of recording electrodes into the PB under visual guidance. After completion of surgery, the animal was administered gallamine triethiodide (8 mg/kg) i.v. and artificially respired. The cistema magna was cut to drain cerebrospinal fluid (CSF) and a bilateral pneumothorax was performed to help lessen pulsatile effects. During the course of the recording session, the animal was maintained on alpha chloralose anesthesia (50 mg/kg). A three-prong stimulating “comb” electrode constructed from “ooo” stainless steel insect pins, insulated except for 0.5 mm at the tip and with tip separations of OS-I.0 mm and with tips cut forming a 45” bevel, was placed in the SN [A: 4.5, L: 2.5-5.5, H: -4.0 (medial pin)]. A similar stimulation electrode (although not beveled) was placed in the EPN (A 11.5, L: 6.5, H: -2.5). A bipolar electrode, constructed from the same materials as the combs was placed in the trigeminal ganglion and cemented in place at a location identified both by stereotaxic coordinatesra and by low-threshold stimulation-induced jaw movement. Substantia nigra and EPN electrodes were not cemented in place so that various stimulation locations could be tested during a recording session. Needle stimulation electrodes were also placed in the face of the cat bilaterally. Basal ganglia stimulation electrodes were placed ipsilateral to the recording site, since pilot electrophysiological and anatomical studies suggested primarily a unilateral basal gangliapontine parabrachial connection.’ Extracellular single unit recordings were obtained from a glass micropipette electrode filled with 1.6M potassium citrate, introduced stereotactically into the PB region. Spontaneous activity and responses to stimulation were recorded on FM tape and analyzed with a microcomputer. Mean firing rates and standard deviations were calculated for putative responses. These were compared to mean firing rates and standard deviations of baseline (non-stimulation) periods. Differences of at least two standard deviations were considered to be responses. If a cell was found to have a response to trigeminal sensory stim~ation, the effects of a preceding stimulus pulse to the basal ganglia on the sensory response were assessed. In this conditioning-test paradigm (C-T), basal ganglia stimulation (single pulses, OSms duration, 0.5 Hz) served as the conditioning stimulus while trigeminal sensory stimulation (single pulses, 0.5 ms duration, 0.5Hz) was always the test stimulus. Means and standard deviations were calculated for sensory response with and without preceding basal ganglia stim~ation. Differences of at least two standard deviations were considered to be an effect. In an effort to control for possible artefactual effects of current spread from basal ganglia stimulation sites, several procedures were performed. In some animals, stimulating electrodes were placed in the internal capsule adjacent to the EPN. If a cell had a response to SN stimulation, the SN electrode was driven either ventral& into the cerebral peduncle or dorsally above the SN (into the medial lemnis-

Fig. 1. HRP injection into PB region. (A) Bar = 1 mm. (B) SN pars reticulata labeled cells. Bar = 25 pm. (C) SN pars lateralis labeled cells. Bar = 50 pm. (D) Terminal field (and cell) labeling in SN pars reticulata. Bar = 50pm. (E) SN pars lateralis labeled neuron. Bar = 25 pm. (F) Retrorubral labeled cells. Bar = 25 pm.

413

Fig. 3. Additional transport from PB HRP injection. (A) Terminal field labeling in entopeduncular nucleus. Bar = 25 pm. (B) Labeled entopeduncular nucleus cell. Bar = 50 pm. (C) Terminal field and cell labeling (arrow) in the central nucleus of the amygdala. Bar = 50 pm. (D) Thalamic terminal field labeling (ventro-posterior medial nucleus). Bar = 250 pm. (E) Darkfield photograph of spinal trigeminal (nucleus interpolaris) labeling. Bar = 50 pm. (F) Terminal field labeling (and some cell labeling) in the bed nucleus of the stria terminalis. Bar = 250 pm. 414

Fig. 5. Substantia nigra HRP injection (A). Bar = 1 mm. (B) Darkfield photograph showing labeled neurons in the dorsal PB region. Bar = 1mm. (C) Terminal field (and labeled cells) in dorsomedial PB region. Bar = 50 pm. (D) Fusiform PB cell. Bar = 2Qm. (E) Nucleus tegmenti pedunculopontinus (NTPP) labeled cells. Bar = 50pm.

415

Interactions between BG, PB and TS in cat

417

cus and other dorsally traveling fiber systems) and the effects of stimulation were again assessed. Stimulation was also varied between the different leads of the comb electrodes to assess effects of stimulating different regions of the SN. In all cases, stimulus current was monitored and never exceeded 250 PA. Strength duration curves were constructed and chronaxies measured for some SN stimulation responses and SN C-T effects on PB sensory responses. Thresholds were determined for either evoking a response in a PB unit or for inhibiting a PB sensory response by successively longer duration SN stimulation pulses (range: 0.01-3.0 ms). Obtained threshold values were then plotted and rheobase and chronaxie determined from these curves. These curves and chronaxie measurements were compared to data obtained from animals in which SN stimulation effects on superior colliculus neurons (via a known SN output pathway) were recorded. At the conclusion of the recording session, a metal lesioning electrode was passed through the glass micropipette and small electrolytic lesions were made at various anterior-posterior locations (including the last recording site) and at various depths. Small marking lesions were also made at all stimulation sites. The animal was then killed with barbiturates and the brain perfused with saline followed by 10% formalin. Frozen sections (80 pm) were cut, mounted, and stained with Cresyl Violet. Careful reconstructions were made of the PB region of the brainstem and of stimulation sites. Only those cells definitely located within the boundaries of the PB region were included for analysis. RESULTS Anatomy Terminology. The dorsal pontine tegmental region contains principally the parabrachial (PB) nucleus, the cuneiform nucleus (CN), the locus coeruleus (LC) and the trigeminal mesencephalic nucleus (Mes. 5). The PB is a grouping of cells surrounding and embedded within the ascending limb of the brachium conjunctivum corresponding to an area often referred to as the pontine taste area.26s29The PB region referred to in this paper extends from the region of the caudal inferior colliculus through the area of the trigeminal motor nucleus. This PB region can be distinguished from nearby cell groups such as the mesencephalic locomotor region (MLR),“v~~ which is more rostral, and dorsal, the cuneiform nucleus (CN), also more rostra1 and dorsal, and the nucleus tegmenti pedunculopontinus (NTPP), which also surrounds the ascending limb of the brachium conjunctivum but is situated far more rostra1 than the PB. Parabrachial horseradish peroxidase injections. De-

posits of WGA-HRP were placed within the PB region in 9 cats. Figures 1A and 2A show representative cases. Both the ipsilateral SN and the EPN contained labeled neurons and diffuse terminal field labeling (Figs lB, 2, 3A and B). Cell labeling in the SN was far more numerous than in the EPN. In the SN, most of the cells were within the middle to lateral portions of the pars reticulata or within pars lateralis. Sparse terminal field and cell labeling was also seen in the retrorubral area (substantia nigra pars dorsalis3*). Labeled SN cells were found mostly at the more caudal levels. Substantia nigra neurons projecting to

Fig. 2. Substantia nigra and trigeminal afferents to the PB. WGA-HRP was injected in the PB region (black area denotes effective injection site, lines denote extent of local transport). (A) HRP injection centered in PB region. (B) Dots denote locations of HRP-filled cells in SN pars reticulata, pars lateralis and throughout the rostrocaudal extent of the spinal trigeminal system (although densest near the interpolaris and paratrigeminal nuclei). Sst = spinal trigeminal tract, 7N = facial nerve, ICP = inferior cerebellar peduncle, RN = red nucleus, SN = substantia nigra, SpV = spinal trigeminal nucleus, BC = brachium conjunctivum, MV = trigeminal motor nucleus, SO = superior olivary nucleus.

the PB had a mean somatic area of 192.3 f 64pm* (based on 84 cells), were mostly triangular or fusiform in shape (Fig. 4), with mean soma diameter of 30.8 k 8.8 pm. Labeled cells bodies in the EPN were quite rare and situated at several rostrocaudal levels (as was diffuse terminal field labeling) and scattered through the medial-lateral extent of the nucleus. Many labeled fibers could be traced coursing above the optic tract and seemingly ascending into the EPN. Other fibers could be traced as following a course laterally and turning ventrally toward the amygdala. Most of the cell labeling in the trigeminal system was observed in the spinal trigeminal nuclei (spinal 5) (Figs 2 and 3E). All levels of the ipsilateral spinal 5 system (i.e. nucleus oralis, interpolaris, and caudalis) contained some HRP labeled neurons, with most labeling in the oralis and interpolaris zones. Labeled cells were also seen in the paratrigeminal nucleus in the dorsal part of the spinal trigeminal tract near the obex. Most of these labeled neurons were medium-

J. S.

418

SCHNEIDER

NT

Fig. 4. Camera lucida drawings of SN cells which project to the PB (top), PB cells which project to the SN (middle), and nucleus tegmenti pedunculopontinus (NTPP) cells which project to the SN (bottom). Note the larger size and polymorphic shapes of NTPP cells as compared to the primarily fusiform PB cells. Considering morphological differences in inputs and outputs of these regions, they most likely represent independent neural populations.

sized (range = 2&35pm) with either ovoid, triangular, or occasionally fusiform cell bodies. When WGA-HRP injections were contained within the PB region, labeled cells were only rarely

observed in the trigeminal main sensory nucleus or trigeminal motor nucleus. Labeling in other regions was consistent with previously published reports of PB afferents (Table I). In two brains, WGA-HRP injections into the dorsal pons rostra1 to the central PB region resulted in much fewer labeled SN pars reticulata and spinal 5 cells and a greater amount of retrorubral and rostra1 reticular formation (FTL) labeling. There was virtually no labeling of other structures known to project to or receive a projection from the PB. In two brains, WGA-HRP injections encompassed a considerable portion of the ventral PB region, the supratrigeminal area, and part of the trigeminal main sensory nucleus. Again, few HRP-positive SN cells were seen, with labeling more in rostra1 FTL areas. Virtually no labeling was seen in other structures connected with the PB with the exception of fairly heavy field labeling in the VPM nucleus of the thalamus and extensive cell and field labeling in the trigeminal main sensory and motor nuclei bilaterally. Two other cats received WGA-HRP injections which only minimally involved the ventrolateral parabrachial region. No SN labeling was observed. Sparse terminal field labeling could be seen in the amygdala and VPM nucleus of the thalamus. Labeling was seen throughout several trigeminal regions, as the injection site impinged upon the main sensory nucleus and lateral portions of the supratrigeminal region. Injections made directly into the cerebellar white matter (brachium conjunctivum), the vestibular nuclei, and the trigeminal main sensory nucleus (all areas which may have been affected, to some extent, by spread of WGA-HRP from PB injections sites) resulted in no basal ganglia labeling or pattern of labeling similiar to that seen with injections involving the PB region. Substantia nigra and entopeduncular nucleus horseradish peroxidase injections. Deposits of WGA-HRP

were placed in the SN in 5 cats and in the EPN in 5 cats. Figures 5-7 show representative cases. Injections in the SN resulted in cell and terminal field labeling at several PB levels, extending from the

Table 1. Cell and terminal field labeling observed following pontine parabrachial hdrseradish peroxidase injections Brain region N. solitary tract Lat. tegmental field (FTL) Gigantocellular tegmental field (FTG) Bed nucleus of stria terminalis Central nucleus of amygdala Substantia innominata VM and lateral hypothalamus VPM nucleus of thalamus Spinal trigeminal nuclei Substantia nigra Entopeduncular nucleus

Cell labeling

Terminal field labeling

X X

X X

X

X

X

X

X

X

X X

X

X

X

X

X

X

VM = ventromedial; VPM = ventroposterior-medial.

‘I

Interactions between BG, PB and TS in cat

419

Fig, 6. Parabrachial-substantia nigra projection. (A) Substantia nigra WGA-HRP injection sites. Blackened area denotes effective injection site; lines denote local transport. (B) Dots denote locations of HRP-tilled cells in the parabrachial region following the injection pictures in (A). The ~~b~chial-Nina projection is primarily ipsilateral. A considerable amount of terminal field and fiber labeling was also observed in the Darab~chial reaion. Each drawina is a comnosite of 2 100 urn thick sections. (Cf Labekxl cells and t&minal field w&e also found in The nucleis tegmenti ~~n~ulopontinus (NTPP).

caudal inferior colliculus to the level of the trigeminal

motor nucleus. Cells containing HRP were found in all PB subdivisions. A number of HRP-filled cells were also embedded within the fibers of the brachium conjunctivum. At more caudal levels (through the trigeminal motor nucleus) labeled PB cells were situated in more dorso- and ventrolateral PB regions. Diffuse terminal field labeling was observed at all PB levels extending to the region of the Kolliker-Fuse nucleus. Cell and terminal field labeling was also seen in the NTPP rostra1 to the PB.’ The PB neurons labeled from an SN WGA-HRP injection were mostly small and fusiform or spindleshaped, spherical, or more rarely, polygonal (Fig. 4). Mean somatic areas of PB labeled neurons (sampled from various PB regions) was 96.5 + 39.1 pm (n = 92 cells). Mean soma diameter of dorsal PB neurons was 11.OI_+2.7 pm (n = 60), while ventral PB cells had a slightly but not significantly (P 2 0.05) smaller mean diameter of 9.8 _t I.8 pm (n = 32). The more rostra1 NTPP region contained fewer small spindle-shard cells and more small polygonai and medium to large spindle and polygonal-shaped cells than the PB region (Fig. 4). Mean somatic area of NTPP ceils was 109.3 + 30.3pm, while mean soma diameter was 14.1 f 1.8 pm (n = 32). Both these measures are significantly different (P I 0.05) from those of PB neurons. Injections of WGA-HRP into the EPN (Fig. 7) resulted in PB cell labeling from just posterior to the inferior

colliculus

to the trigeminal

motor

nucleus.

Cells were distributed both in dorsal and ventral regions with a number of cells embedded within fibers of the brachium ~onjuncti~. While the jority of labeled neurons were ipsilateral to

PB the mathe

injection site, a small number of contralateral cells also contained HRP. Very little PB terminal field labeling was observed. Following EPN HRP injections, some labeled cells and terminal field were also seen in the NTPP region. Parabrachial neurons labeled from EPN injections were generally small fusiform or spherical cells, with a mean somatic area of 84.2 _I 25.90 pm2 (n = 128 cells). Mean somatic diameter of dorsal PB cells was 12.1 + 2.4 pm (n = 40 cells), ventral PB cells were slightly smaller with a mean diameter of 9.8 + 2. t pm (n =40 &Is). Labeled pedunculopontine cells had mean somatic areas of 112.3 f. 18.5 pm2 (n = 32 cells) and mean somatic diameters of 16.4 + 2.8 pm [significantly larger than PB cells (P I O.OS)]. Cases in which the HRP injection was too ventral and centered in the optic tract with minima1 EPN involvement resulted in no PB labeling. Injections which were too lateral (putamen-claustrum) or too dorsal (internal capsule) without appreciable EPN involvement also resulted in no PB labeling. One injection which only involved the rostra1 portions of the EPN but involved the areas ventral to the EPN resulted in slight PB cell labeling. Electrophysioiogy Characteristics of pontine parabrachial neurons. A total of 228 cells were recorded from the parabrachial region. These cells had a range of firing rates from very slow (I 1.0 spike/s) to fast (30 spikes/s) (mean = 12.5 spikes/s f 12.2). Cells were recorded from several PB rep;ions. Parabrachial divisions will be noted as dorsal, ventral, medial and lateral. Figure 8 shows the locations of cells responsive to basal ganglia stimulation, trigeminal sensory stimulation

J. S. SCHNEIDER

420

(12%) responded to trigeminal stimulation with excitation followed by inhibition (E-I) and two cells (4%) responded with only inhibition. PB cells which responded to trigeminal stimulation with excitation or excitation followed by inhibition had a mean excitatory response latency of 7.9 ms (SD = 5.9) and a mean inhibitory duration of 400 ms (SD = 405, range 150-1400 ms). Examples of PB cell responses to trigeminal stimulation are shown in Fig. 9. Cells responsive to stimulation were sensitive to weak electrical stimulation of the trigeminal ganglion (1.0 V, 10-40 PA) as well as to natural stimulation of the face (i.e light pressure on face, intra-oral structures, and/or tongue or brushing of hairs on face). Although peripheral sensory fields were not systematically mapped in this study, receptive fields generally ranged in size from small (i.e. small localized region on gum) to large (i.e. entire perioral region including vibrissae). Responses to basal ganglia stimulation. Responses of PB cells to electrical stimulation of the SN were tested in 175 cells; responses to EPN stimulation were tested in 124 cells. Only six cells (4%) responded to EPN stimulation alone and only four cells (3%)

Fig. 7. Parabrachial-entopeduncular projection. (A) WGAHRP injection sites in the entopeduncular nucleus. Blackened center denotes effective injection site; lines denote extent of local transport. (B) Dots denote locations of HRP-filled neurons in the parabrachial area following the injections pictured in (A). The parabrachial-entopeduncular projection is mostly ipsilateral with only a small contralateral component. Virtually no parabrachial terminal field was observed. Each drawing id a composite of two 100 urn thick sections. (C) ~ Labeled cells and terminal field were found in the nucleus tegmenti pedunculopontinus (NTPP). BC = brachium conjunctivum, EN = entopeduncular nucleus, MV = trigeminal motor nucleus, OT = optic tract, Inf. C = inferior colliculus. I

only, both basal ganglia and trigeminal sensory stimulation. and non-responsive cells. Cells responsive to all types of stimulation used were found throughout the PB and interspersed within the fibers of the brachium conjunctivum. The most medial and lateral portions of the PB were not sampled as extensively as the central and lateral regions of the nucleus and therefore cannot be dismissed as having little trigeminal or basal ganglia connections. Responses to trigeminal stimulation. Forty-nine cells (22%) of 225 PB cells had some type of response to trigeminal stimulation. Forty-one of these 49 (84%) cells responded to trigeminal sensory stimulation with excitation only. The majority of these excitatory responses were single spike responses, while bursts of spikes to trigeminal stimulation were only seen with high intensity stimulation. Six cells

Fig. 8. Locations of cells recorded in the parabrachial region. (A) Locations of cells responsive to basal ganglia stimulation [either SN (filled circles) or EPN (open circles)]. (B) Locations of cells responsive to trigeminal sensory stimulation only (open circles) and to both trigeminal and basal ganglia stimulation (filled circles) and to both trigeminal and basal ganglia stimulation (filled circles). (C) Locations of non-responsive cells (neither trigeminal nor . . .. nasal gangua responses).

Inte~ctions

between BG, PB and TS in cat

421

Fig. 9. Substantia nigra (SN) and trigeminal sensory effects on parabrachial (PB) neurons. (A) Top oscillograph tracing shows a PB excitatory response to SN stimulation; middle tracing shows a PB inhibitory response to SN stimulation; and lower tracing shows a PB excitatory-inhibitory response (with rebound excitation) to SN stimulation. Calibration: Top = 5 ms, middle and lower = 0.1 s. (B) Top, middle, and lower tracings show PB excitatory inhibitory and excitatory-inhibitory response, respectively, to trigeminal ganglion stimulation. Calibrations are the same as in (A). (C) Illustration of the differences in latency of inhibitory effects on PB spontaneous firing from (top) SN and (bottom) trigeminal ganglion on the same neuron. Calibration = 0.2 s. Arrows denote stimulus artifact.

responded to both EPN and ~geminal stimulation. Considering the paucity of EPN effects on PI3 cells, basal ganglia effects, in the remainder of the paper, will refer primarily to SN influences on the PB area. In contrast to the effects of EPN stimulation, 65 PB cells (37%) responded in some way to SN stimulation. Of these, 17 cells (26%) responded to SN stimulation alone. Seven of these responses were excitatory; IO were inhibitory. A total of 48 cells (29%) responded to both SN and trigeminal ganglion stimulation. Of these, half had excitatory responses to SN stimulation (mean latency = 7.7 f 4.6 ms), with others responding with E-I sequences and pure inhibitions (8 and 16 cells, respectively) (mean latency of inhibition = 5.2 + 2.6 ms) (Fig. 9A). PB cells responding to trigeminal sensory stimulation in a constant and reproducible manner were tested for modulatory effects from basal ganglia stimulation. The sensory test stimulus was just above threshold so that each stimulus evoked a neural response on each trial. A total of 86 cells were tested in this condition-test (C-T) paradigm. Nineteen of these cells were subjected to EPN stimulation preceding the sensory stimulation. Two of these

cells (10%) displayed an EPN modulatory effect (inhibition) on the sensory response. Thirty-one out of 72 PB cells (43%) tested in the C-T paradigm had their sensory responses altered by prior SN stimulation (Fig. 10). In all cases, SN stimulation preceding the sensory stimulus caused either a total inhibition of the sensory response or partial inhibition of the sensory response with a concomitant lengthening of response latency. Nineteen of the 31 cells (61%) tested had overt SN response. Cells, responding to SN stimulation with excitation, still had inhibitory effects upon evoked sensory responses. Since many cells had fairly low spontaneous firing rates in the absence of appropriate stimulation and inhibitory responses may not be readily detected extracellularly, ceils were tested in the C-T paradigm even if they showed no overt SN response. Twelve cells without an overt SN response (39%) showed an inhibition of the test sensory response in the C-T paradigm. Cells with overt SN responses had effective mean C-T intervals of 14.7 ms (SD = 11.3) to 285.1 ms (SD = 102.2), whereas cells without overt SN responses had mean C-T intervals 16.2 ms (SD = 14.6) to 150.0 ms (SD = 42.9).

J. S.

422

tkHNElDER

chronaxie was derived after calculation of chronaxie for each cell. These measurements are regarded as

useful in determining what elements are being activated during electrical stimulation of brain.22 The strength duration curve for SN stimulation effects on the PB is shown in Fig. 11. The mean chronaxie of the neural population which mediated inhibitory effects at PB neurons was calculated at 650 ps. For comparison, strength~uration curves were constructed for SN stimulation effects on superior colliculus neurons. The mean chronaxie of the neural population mediating inhibitory effects on coiliculus cells was 700 ,us. These values were not significantly different (P > O.OS), indicating that a similar neural population

was probably

SN ‘001

6.

C-T

responsible

INHIBITORY (

t

Chronwe

EFFECT

for both effects.

ON PB CElLS

= 590psecl

I

--lb-A

DURATIQN

Fig. 10. Substantia nigra m~u~at~o~ of PB sensory responses. Arrows denote stimulus artifact. (A) Upper trace shows a PB unit response to trigeminal ganglion stimulation, Middle trace shows the same unit’s response to ganglion stimulation preceded by IOms by a puke to the ipsilateral SN. Lower trace shows that the response to ganglion stimulation is inhibited by SN stimulation when the SN stimulation occurs at least 12ms prior to it. (B) Trace shows excitatory response to ipsilateral EPN stimulation.

A small number of PB neurons (n = 5) displayed antidromic response to EPN (n = 2) and SN (n = 3) stimulation. Antidromic~ty of these responses was confirmed by invariant, short response latencies, high frequency following (75-100 Hz), and spike collision.’ StimuIation controlprocedures. The use of electrical brain stimulation, particularly in basal ganglia structures, presents certain problems, most important of which is current spread to adjacent

structures.

Several

procedures were employed in the present study to determine to what extent current spread may have contributed to the effects of basal ganglia stimulation on PB neurons.

Strength-duration curves were constructed for the SN stimulation effect on PB neurons. Rheobase current

and chronaxie

were then determined.

Mean

SN

I mroc)

INHIBITORY

EFFECT

1001 . PB Cells

= 80 E 3 0 6o : z 4 40 it,,

(Chronaxle =650pssrj

0 SC C611% t Chronaxre = 7OOpascl

I

3 20I =I= OL

I

OS

I 40

I” 20

I $5

DURATION

I 30

i S’ISOE)

Fig. Il. Strength duration curves for SN stimulation effects on PB neurons. Top: Strength duration curve constructed from data on SN inhibitory effect on trigeminal sensory responses in PB neurons. Curve shows mean (and standard deviation) changes in threshold current as a function of varying stimulus pulse duration. Current values were converted to percentages of maximum threshold (current at shortest pulse duration to cause an effect). Bottom: Curves, constructed

similarly

as that above,

show

that there is no curve or effects on either PB or suoerior results suggest that a similar

significant difference in either the strength duration chronaxies

for SN inhibitorv

CO~I~~~IUS (SC) ceils. the&

neural population was mediating both effects.

Interactions between BG, PB and TS in cat

423

The mean chronaxies of the population mediating SN axons may retrogradely transport HRP.r9 Some conC-T inhibitory effects on PB sensory responses trol of this problem was provided by comparing (590 ps) and mediating SN excitatory effects (680 ps) transport resulting from injections made from did not differ (P < 0.05) from that yielding SN different approaches, i.e. angle drops and straight inhibitory effects on PB cells but did differ drops (ex. in the PB: 62” angle from posterior to significantly from those of the surrounding pyramidal anterior, straight drop into the PB through the tract (150 ps) or medial lemniscus (40-100 ~s),“*~* cerebellum, straight drop into PB, cerebellum resuggesting that these fiber systems may not have been moved). The possibility still exists that some transthe source of SN evoked responses. port resulted from broken axons within the injection During SN stimulation trials, stimulating current site itself. However, injections of HRP into the SN was monitored and observed never to exceed 500 y A, (retrogradely labeled by PB injections) produced with the bulk of stimulation trials ranging between signifi~nt anterograde labeling in the PB. This to100 and 250 PA. Previous studies indicate that curgether with the el~trophysiolo~~ findings suggest rents of this intensity should spread approximately that the connections described are not artifactual. In 3004OO~m from the stimulation site.28 This inforaddition, a descending nigral projection passing mation together with the histologically verified SN through the PB region en route to lower structures stimulation sites, suggests that stimulation was most has not been previously described, making direct probably contained within the boundaries of the SN. uptake into cut fibers from this system an unlikely During several SN stimulation trials, the stimulating alternative. electrode was raised or lowered in 0.5 mm steps and The SN-PB pathway appears to be more extensive PB cell responses were observed. In many cases than an EPN-PB projection. These differential SN movement of the electrode 0.5 mm from the effective and EPN projections to the PB are interesting since stimulation site resulted in loss of PB response. Often, the EPN and the SN pars reticulata have been the PB response could be brought back by increasing considered to be part of a single cell group separated the stimulus intensity. In the remaining cases, moveby internal capsule fibers.8*23The present finding ment of the electrode 1.0 mm from the effective suggests that functionally, the characteristics of these stimulation site eliminated the PB response. In some two basal ganglia output pathways may be somewhat instances, either the medial or lateral pin of the SN different. This is consistent with the findings of stimulating comb was situated outside the SN. In tieLong et aL5 which suggest that SN pars these cases, stimulation between these pins and an reticulata/pars lateralis neurons are functionally different (i.e. more concerned with orolingual behavadjacent one in the SN failed to elicit a response, iors) from EPN neurons (more concerned with limb while stimulation between the two pins located within the SN resulted in a PB response. motor behavior). This idea is further supported by Electrodes were placed above the EPN in the the findings that the SN pars reticulata and not the internal capsule for control stimulation. PB cells EPN projects to other output areas such as the which responded to EPN stimulation were un- medullary reticular formation” and the superior colresponsive to internal capsule stimulation. liculus? Although labeled fibers could be seen to descend from the PB to the region near trigeminal motor nucleus and further posteriorly near the hypoDISCUSSION glossal nucleus, no distinct terminal labeling could be The PB is considered to participate in various discerned within these structures. While labeled fibers functions including: transmission of gustatory inforpass through and around these motor regions, it is mation, control of respiration, locomotor activities, impossible at this level of study to determine with emotional behaviors, regulation of waking and sleep certainty whether any of these fibers terminate in states, and the integration of sensory information for these regions. the control of orolingual/orofacial motor activIn electrophysiological studies, SN and to a lesser ities.‘3*2s~34 The results of the present study suggest extent, EPN stimulation affected the spontaneous that some of the functions of the PB, particularly firing of PB neurons. Furthermore, a conditioning those which may involve the trigeminal system, may stimulus pulse applied to the SN (or EPN) was to some extent be modulated by a descending basal capable of inhibiting PB cells responses to trigeminal ganglia inguence. sensory stimulation. Inhibitory effects from the SN to the PB could be The pontine PB region appears to have anatomical connections with the SN pars reticulata as well as mediated via GABAergic SN reticulata output neuwith parts of the spinal trigeminal system. Results of rons. This has recently been suggested intracellularly studies using the HRP transport technique must, by Noda and Oka. 24 Excitatory effects from SN however, be evaluated cautiously. There is a fiber of stimulation pose more of an interpretation problem passage problem with this technique which must be and their true origin remains to be verified. One considered. The use of the WGA-HRP conjugate possibility is an excitatory output from the SN. This should reduce the probability of HRP uptake into has been suggested by York and Fabera and Deniau undamaged fibers of passage.” However, damaged et aL6in studies of the nigrotectal pathway. However,

424

J. S. SCHNEIDER

later work by Chevalier et ~1.~ suggested that excitatory SN effects on the superior colliculus were primarily due to effects on fibers passing through and around the SN. Presently it is possible that SN stimulation antidromically activated ascending PB fibers coursing over the SN. Such effects could be expressed synaptically if the ascending fibers also collateralize cells in the PB (or elsewhere that project back on PB cells). Altematively, SN stimulation may have activated pyramidal tract fibers or descending medial forebrain bundle fibers coursing to the PB region. However, stimulation applied to white matter tracts (medial lemniscus, pyramidal tract) was not as effective in influencing PB cell activity as stimulation applied through electrodes in the SN itself. Strengthduration and chronaxie data also suggested that fiber tract stimulation did not account for the observed effects. Following HRP injections into the SN or the EPN, labeled cells in the rostra1 peribrachial NTPP region were observed. Many large cells were seen interspersed with small fusifo~ and spheroid cells. The PB, on the other hand, contained primarily small fusiform or spheroid cells. On the basis of cell morphology and distribution patterns following HRP injections, it is suggested that these two peribrachial regions (PB and NTPP) may be functionally independent populations. The SN-EPN projections presently described are also not to be confused with those believed to terminate in the CN-MLR. The SN-EPN have been described as projecting to the CN-MLR at stereotaxic levels of P. 1.5-2.0 in the cat.‘“.” The SN projects to the more caudai PB at a level of approximately P. 4.5. The fact that SN stimulation can influence the responses of PB cells to sensory stimulation, and that SN stimulation alone can influence PB spontaneous activity suggests at least two ways in which the SN may influence this system. First, SN effects on PB cells may be exerted postsynaptically. Both SN and trigeminal sensory information are received indepen-

dently by some PB neurons whose excitability and response to sensory input are mediated by the SN at the PB neuron. This might be the sequence of events occurring at PB neurons with overt SN responses and with SN C-T inhibitory effects on trigeminal sensory inputs. Second, SN effects on PB cells may be exerted presynaptically. The SN would presynaptically affect trigeminal sensory inputs to PB cells, thus affecting the PB response to sensory stimulation and ultimately influencing PB output. These effects might be occurring at PB cells with no overt SN responses but with potent SN C-T inhibitory effects on trigeminal sensory inputs. These effects may be mediated di- or polysynaptically, possibly via the reticular formation. The wealth of forebrain connections, access to the spinal cord, and the cerebellum, as well as its connections with the basal ganglia make the PB region intriguing. The PB receives gustatory, visceral, and trigeminal somatosensory inputs and may indirectly affect ingestion-related activities.34 Through extensive reciprocal connections with several forebrain structures, the PB may influence ingestion by relaying relevant information to forebrain structures and integrating outputs from the forebrain to control oroingestive behaviors. It is interesting that cells in the SN that project to the PB arise in the same regions (i.e. lateral pars reticulata, pars lateralis) where the activity of SN cells are reported to be most influenced by jaw and tongue movements5 Thus, the PB region may represent an area through which a basal ganglia influence over oroingestive behavior may be exerted. Ack~ouiiedgeme~~s-The

author wishes to thank Dr N. A. Buchwald for allowing the use of his resources for this project, Dr R. S. Fisher for advice on the anatomicaf procedures, R. A. Koelling and Gloria Allen for technical assistance, Drs T. I. Lidsky and J. S. Schwartzbaum for helpful comments on the manuscript, and Setsuko Kashitani for preparation of the manuscript. This research was supported by USPHS Grants HD 04612, HD 05958 and DE 07282.

REFERENCES

basis of tardive dyskinesia. In Tardive Dyskinesiu (eds Fann W. E., Smith R. C., Davis J. M. and Domino E. F.). SP Medical, New York. Beckstead R. M., Edwards S. B. and Frankfurter A. (1981) A comparison of the intranigral distribution of nigrot~tal neurons labeled with horseradish peroxidase in the monkey, cat, and rat. J. Neurosci. I, 121-125. Buchwald N. A., Schneider J. S., Levine M. S. and Hull C. D. (1983) Basal ganglia influences on brainstem neurons. Sot. Neurosci. Absr. 9, 950. Chevelier G., Deniau J. M., Thierry A. M. and Feger .I. (1981) The nigrotectal pathway. An electrophysiological reinvestigation in the rat. Bruin Res. 213, 253-263. DeLong M. R.. Crutcher M. D. and Georgopoulos A. P. (1983) Relations between movement and single cell discharge in the substantia nigra of the behaving monkey. J. Neurosci. 3, 159991606. Deniau J. M., Chevalier G. and Feger J. (1978) Electrophysiological study of the nigro-tectal pathway in the rat. Neurosci. Lett. 10, 27-35. de Olmos J. S. (1977) An improved HRP method for the study of central connections. Expl Brain Res. 29, 541-551. Edley S. M. and Graybiel A. M. (1983) The afferent and efferent connections of the feline nucleus tegmenti pedunculopontinus, pars compacta. J. camp. Neural. 217, 1877215. Fuller J. H. and Schlag J. D. (1976) Determination of antidromic excitation by the collision test: Problems of interpretation. Bruin Ref. 112, 283-298. Garcia-Rill E., Skinner R. D. and Gilmore S. A. (1981) Paltidal projections to the mesencephalic locomotor region (MLR) in the cat. Amer. J. Anat. 161, 31 I-321.

1. L. Baldessarini R. J. and Tarsy D. (1980) The pathophysiologic

2. 3. 4. 5. 6. 7. 8. 9. IO.

Interactions between BG, PB and TS in cat

425

11. Garcia-Rill E., Skinner R. D., Jackson M. B. and Smith M. M. (1983) Connections of the mesencephahc locomotor region (MLR). I. Substantia nigra afferents. Brain Res. Bull. 10, 57-62. 12. Holstege G. and Kuypers H. G. J. M. (1982) The anatomy of brainstem pathways to the spinal cord in cat. A labeled amino acid tracing study. In Decending Pathways to rhe Spinal Cord (eds Kuypers H. G. J. M. and Martin G. F.), pp. 145-176. Elsevier, Amsterdam. 13. Katayama Y., Watkins L. R., Becker D. P. and Hayes R. L. (1984) Evidence for involvement of cholinoceptive cells of the parabrachial region in environmentally induced nociceptive suppression in the cat. Bruin Res. 299, 348-353. 14. King G. W. (1980) Topology of ascending brainstem projections to nucleus parabrachialis in the cat. J. comp. Neurol. 191, 615-638. 15. Labuszewski T. and Lidsky T. I. (1979) Basal ganglia influences on brainstem trigeminal neurons. Expl Neurol. 65, 471477. 16. Levine M. S., Ferguson N., Kreinick C. J., Gustafson J. W. and Schwartzbaum J. S. (1971) Sensorimotor dysfunctions and aphagia and adipsia following pallidal lesions in rats. J. camp. Physiol. Psychol. 77, 282-293. 17. Lidsky T. I., Schneider J. S. and Harper J. A. (1981) Jaw movements evoked by substantia nigra stimulation: Role of extranigral current spread. Brain Res. Bull. 5, 487489. 18. Martin J. P. (1967) The Basal Ganglia and Posfure. Philadelphia, Lippincott.

19. Mesulam M.-M. (1982) Principles of horseradish peroxidase neurohistochemistry and their applications for tracing neural pathways-axonal transport, enzyme histochemistry, and light microscopic analysis. In Tracing Neural Connections with Horseradish Peroxidase (ed. Mesulam M.-M.), pp. 1-151. John Wiley, New York. 20. Mettler F. A. (1945) Effects of bilateral stimultaneous subcortical lesions in the primate. J. Neuropath. exp. Neurol. 4, 99122. 21. Mizuno N., Nomura S., Itoh K., Nakamura

Y. and Konishi A. (1978) Commissural interneurons for masticatory motoneurons: A light and electron microscope study using the horseradish peroxidase tracer technique. Expl Neurol.

59, 254-262. 22. Morgan L. 0. (1927) The corpus striatum. A study of secondary degenerations following lesions in man and of symptoms and acute degenerations following experimental lesions in cats. Arch. Neurol. Psych&. 18, 495-549. 23. Nauta H. J. W. (1979) A proposed conceptual reorganization of the basal ganglia and telencephalon. Neuroscience 4,

1875-1881. 24. Noda T. and Oka H. (1984) Nigral inputs to the pedunculo-pontine

region: intracellular analysis. Brain Res. 322,

332-336. 25. Nomura S., Mizuno N., Itoh K., Matsuda K., Sugimoto T. and Nakamura

Y. (1979) Localization of parabrachial nucleus neurons projecting to the thalamus or the amygdala in the cat using horseradish peroxidase. Expl Neural. 64,

375-385. 26. Norgren R. and Leonard C. M. (1971) Taste pathways in the rat brainstem. Science 173, 11361139.

27. Panneton W. M. and Burton H. (1983) Trigeminal and spinal projections to the parabrachial area in the cat. Sot. Neurosci. Abst. 9, 246. 28. Ranck J. B. (1975) Which elements are excited in electrical stimulation of mammalian central nervous svstem: A review. I Brain Res. 98, 417440. 29. Saper C. B. and Loewy A. D. (1980) Efferent connections of the parabrachial nucleus in the rat. Bruin Res. 197,291-317. 30. Schneider J. S., Denaro F. J., Olazabal U. E. and Laird H. 0. (1981) Stereotaxic atlas of the triaeminal Y eanalion in ~ Y rat, cat, and monkey. Brain Res. Bull. 7, 93-95. 31. Schneider J. S. and Lidsky T. I. (1981) Processing of somatosensory information in striatum of behaving cats. J. Neurophysiol. 45, 841-85 1. I

32. Schneider J. S., Manetto C. and Lidsky T. I. (1985) Substantia nigra projection to medullary reticular formation: Relevance to oculomotor and related motor functions in cat. Neurosci. Let?. 62, 16. 33. Schneider J. S., Morse J. R. and Lidsky T. I. (1982) Somatosensory properties of globus pallidus neurons in awake cats. Expl Brain Res. 46, 311-314. 34. Schwartzbaum J. S. (1983) Electrophysiology of taste-mediated functions in parabrachial nuclei of behaving rabbit. Brain Res. Bull. 11, 61-90. 35. Schwarz M., Sontag K.-H. and Wand P. (1984) Sensory-motor processing in substantia nigra pars reticulata in conscious cats. J. Physiol. 347, 1299147. 36. Shik M. L., Severin F. V. and Orlovskii G. N. (1966) Control of walking and running by means of electrical stimulation of the midbrain. Biophysics 11, 756765. 37. Somana R. and Walberg F. (1979) The cerebellar projection from the parabrachial nucleus in the cat. Bruin Res. 172, 144149. 38. Szabo J. (1980) Distribution of striatal afferents from the mesencephalon in the cat. Brain Res. 188, 3-21. 39. Takeuchi Y., Uemura M., Matsuda K., Matsushima R. and Mizuno N. (1980) Parabrachial nucleus neurons projecting

to the lower brainstem and the spinal cord. A study in the cat by the Fink-Heimer and the horseradish peroxidase methods. Expl Neural. 70, 403-413. 40. York D. H. and Faber J. E. (1977) An electrophysiological study of nigrotectal relationships: a possible role in turning behavior. Brain Res. 130, 383-386. (Accepted 8 April 1986)