THE JOURNAL OF COMPARATIVE NEUROLOGY 306:708-722 (1991)

Mapping Study of the Parabrachial Taste-Responsive Area for the Anterior Tongue in the Golden Hamster CHRISTOPHER B. HALSELL AND MARION E. FRANK Department of BioStructure and Function and Center for Neurological Sciences, The University of Connecticut Health Center, Farmington, Connecticut 06030

ABSTRACT The locations of taste-responsive areas within the brainstem parabrachial nucleus (PBN), an obligatory taste relay in the golden hamster (Mesocricetus auratus),were mapped in relation to cytoarchitectural boundaries. The PBN was systematically searched for multiunit neural activity in response to a taste mixture composed of 0.1 M sucrose, 0.03 M NaC1, and 0.1 M KC1 applied to the anterior tongue. Taste responses were located exclusively in one of three subdivisions of the medial PBN, which is thought to be specialized for gustatory processing, and in one of six subdivisions of the lateral PBN, which is thought to be specialized for general visceral processing. Based on Nissl-stained material, both the medial and lateral PBN subdivisions in the hamster were similar to those reported for the rat PBN. The largest group of taste-responsive cells encompassed two-thirds of the central medial subdivision, while a smaller group of taste cells was exclusively located within the ventral lateral subdivision. The two taste-responsive subdivisions are separated by the superior cerebellar peduncle and contain diverse cell types. The finding that anterior tongue taste may be exclusively represented in circumscribed cytoarchitecturally defined parts of two PBN divisions suggests that taste information from the anterior tongue is required for both specific gustatory and general visceral functions. Key words: gustatory,pons, brainstem, cytoarchitectonics,subnuclei

The parabrachial nucleus (PBN) is located in the dorsal pontine tegmentum and surrounds the superior cerebellar peduncle (SCP) or brachium conjuctivum. In hamsters, rats, and rabbits, some parabrachial neurons respond to gustatory stimuli (Norgren and Pfaffmann, ’75; Yamamoto et al., ’80; Van Buskirk and Smith, ’81). The portion of the PBN that is taste-responsive has been called the “pontine taste area” (Norgren and Leonard, ’71, ’73; Norgren and Pfaffmann, ’75). This area is located primarily medial to the SCP, with some overlap lateral to the peduncle. However, the location of the pontine taste area has not been related to the anatomical organization of the PBN. The PBN has been considered to be “viscerotopically” organized with the gustatory system, sometimes considered a special visceral sense, represented medially and the general visceral system represented laterally (Norgren, ’84; Cechetto, ’87). There is also evidence of convergent information from these two sensory systems onto PBN neurons (Hermann and Rogers, ’85). Individual neurons are coactivated with an additive response following simultaneous stimulation of the hepatic-vagal afferents and tongue taste buds. The possibility of part of the pontine taste area being O

1991 WILEY-LJSS, INC.

a substrate for convergence of information from the ascending gustatory and general visceral pathways is intriguing. In hamsters and rats, gustatory information synapses within the PBN before ascending to higher brain regions, making this an obligatory third-order gustatory relay. Studies using lesions or anterograde transport of tracers injected into the rostra1 taste area of the nucleus of the solitary tract resulted in degenerated fibers or label in the PBN but not in the thalamus (Norgren and Leonard, ’73; Norgren, ’78; Travers, ’88). In contrast, similar studies in primates resulted in label in the ventral posteromedial thalamic nucleus, but not in the PBN (Beckstead et al., ’80). It appears that whatever role the PBN has in the ascending central gustatory system, it is not phylogenetically preserved. The rodent ascending gustatory system does not follow the classical lemniscal sensory organization wherein brainstem nuclei project to thalamic sensory centers en route to ventral forebrain regions and cortex. The inclusion of a pontine relay in the rodent provides a substrate for divergent projections directly to multiple regions with different Accepted January 16,1991

709

PONTINE TASTE AREA IN HAMSTER generalized functions. This is borne out by the findings that PBN efferents terminate bilaterally in the thalamus, hypothalamus, amygdala, bed nucleus of stria terminalis, and insular cortex (Norgren and Leonard, '73; Saper and Loewy, '80; Lasiter et al., '82; Fulwiler and Saper, '84; Yamamoto and Kitamura, '90). The areas of efferent termination suggest that gustatory and general visceral information is directly relayed to cognitive, memorial, and homeostatic regulatory centers. The origin of the efferents within the PBN seems organized with respect to anatomically defined subdivisions (Fulwiler and Saper, '84). However, the relationship of location of taste-responsive neurons and location of efferent neurons is uncertain. One reason for this is that locations of functionally defined neurons have not been precisely related to the subnuclear organization within the PBN. This anatomical organization has recently been characterized in the rat (Fulwiler and Saper, '84; Kolesarova and Petrovicky, '87). Nissl-stained sections were used to parcel both the medial and lateral PBN into multiple, anatomically distinct subdivisions. The role of the PBN in the central gustatory system cannot be well understood until the receptive properties of taste-responsive neurons are related to boundaries defined on cytoarchitectural grounds. The purpose of the present study, therefore, is to provide a precise anatomical localization of taste-responsive neurons in the PBN. Locations of multiunit neuronal activity in response to sapid stimulation of the anterior tongue were mapped in the PBN of the golden hamster. The stimulus was limited to the anterior tongue to selectively activate taste mediated by the chorda tympani branch of the facial cranial nerve, thereby eliminating influences of nerves serving taste bud populations in other oral taste fields. The general anatomical organization of the hamster PBN is described for the first time so the physiologically defined map of taste-responsive sites can be placed in perspective. Cytoarchitecture of the physiologically defined pontine taste area is described in detail. The current results indicate that taste-responsive areas are confined to two anatomically distinct subdivisions of the

~~

Abbreviations Parabrachial cm dm vm cl dl el il sl vl 4v DTg

LC LL Me5ILC Mo5 Pr5 PY

s5

SCP SOC SubC SuVe vsc

nucleus subdivisions central medial dorsal medial ventral medial central lateral dorsal lateral external lateral internal lateral superior lateral ventral lateral fourth ventricle dorsal tegrnental nucleus locus coeruleus lateral leminiscus, tract and nuclei mesencephalic trigeminal and locus coeruleus nuclei motor trigeminal nucleus principal sensory trigeminal nucleus pyramids sensory root trigeminal nerve superior cerebellar peduncle superior olivary complex subcoeruleus nucleus superior vestibular nucleus ventral spinocerebellar tract

PBN, one in each of the gustatory and general visceral regions.

MATERIALS AND METHODS Animal preparation Twenty-four adult male golden Syrian hamsters (Mesocricetus auratus) obtained from Charles River Supply, between 50 and 300 days old and weighing between 100 and 160 g, were used in mapping experiments. Animals were anesthetized with sodium pentobarbital (Nembutal,Abbott Labs., 100 mg/kg, i.p.). Subsequent doses (50 m&g) were administered to maintain areflexia of hindlimb withdrawal during the course of the experiment. Body temperature was maintained between 36" and 38°C with a Deltaphase Isothermal Pad (Braintree Scientific). The animal was positioned in a non-traumatic headholder with the head angled down 27" to facilitate breathing. The skin and muscle overlying the occipital plate were reflected and the occipital plate was removed. The dorsal part of the cerebellum was aspirated, reducing the amount of tissue the electrode passed through and helping to alleviate clogging of the electrode tip.

Stimulus application For mapping studies, the search stimulus consisted of a mixture of 0.03 M NaC1,O.l M sucrose, and 0.1 M KCI. This mixture was chosen because response properties of chorda tympani whole nerve and single fibers to these stimuli are well characterized in the hamster (Frank, '73; Frank et al., '88). Mixtures of taste stimuli have been shown to stimulate neurons in a manner similar to the individual components of those mixtures in the hamster PBN (Travers and Smith, '84). Therefore, the mixture probably evokes responses from all neurons that are responsive to any of the individual components. A system employing bottles of solutions pressurized to 0.5 atm above ambient was used for stimulus application (McPheeters et al., '90). When a solenoid was opened the appropriate solution would flow from the bottles through the solenoid and tubing to the tongue. When searching for taste-responsive sites, the solenoid was opened manually to stimulate the tongue followed by at least 30 seconds of water rinse. When a taste-responsive site was encountered, a computer (IBM-PCAT using Keithley System 570 software) controlled the stimulus application. The protocol consisted of 10 seconds of no stimulus, 12 seconds of taste stimulus, 24 seconds of a water rinse, and then 14 seconds of no stimulus. This sequence was repeated twice at each taste-responsive site. A flow chamber was placed on the tongue to limit the stimulus solutions to the anterior tongue taste buds, which are innervated by the chorda tympani branch of the facial nerve. The stimulus was limited to the anterior tongue to eliminate influences of other taste bud populations within the oral cavity, which have not been characterized as well (Frank et al., '88; Hanamori et al., '88). Also, all oral taste bud populations cannot be easily and adequately stimulated simultaneously (Travers et al., '86; Frank, '91).

Recording and marking The location of taste-responsive neurons was systematically searched for throughout the dorsal pontine tegmentum. In these experiments only multiunit activity (typically four to ten neurons per recording site) was mapped. To avoid the lateral venous sinuses, electrode penetrations

710

C.B. HALSELL AND M.E. FRANK responsive activity was variable because of the aspiration of variable amounts of the dorsal cerebellum. The electrodes were glass micropipettes (1.2 mm o.d./ 0.68 mm i.d. filament borosilicate glass capillaries, World Precision Instruments) pulled with a Narishige PP-83 micropipette puller. These were filled with 4% horseradish peroxidase (HRP, Boehringer Mannheim grade 1) in a 0.5 M KW0.05 M Tris buffer, pH 7.6. The impedance of these electrodes ranged from 1.4-1.6 MQ and tip diameters ranged from 1.2 to 2.4 pm. Electrode penetrations were made in a grid pattern with either 100 or 200 pm steps in the rostrocaudal and mediolateral dimensions. Multiunit taste-evoked neural activity was amplified with a differential amplifier (DAM 60, World Precision Instruments) and monitored with a storage oscilloscope and audio monitor. Electrode penetrations were made at least 100 or 200 km beyond (medial, lateral, rostral, or caudal) penetrations in which taste-evoked activity was seen. At each penetration site, the electrode was advanced ventrally through the tissue in 50 pm steps by means of a micromanipulator (Narishige hydraulic microdrive) until taste-responsive activity was detected. Thereafter, the electrode was advanced in 25 km steps. This allowed the dorsal and ventral borders of the tasteresponsive cell group to be defined more precisely. Two types of taste activity were noted when defining a taste-responsive area. A weak response to stimulation was characterized by an apparent increase in baseline amplitude due to increased neuronal spiking with a low signal-tonoise ratio. A typical weak response is shown in Figure 1A,B. Although this increase in baseline height appeared very small as visualized on the oscilloscope, the response could always be heard on the audio monitor. Typically, as the electrode was advanced ventrally through the PBN at a taste-responsive site, weak taste responses were detected for about 50 pm. The size of the weak zone allows estimation of the distance between active neurons and the electrode tip. As the electrode advanced further, a region of strong taste responses was encountered, followed by another zone of weak taste responses. Not all electrode penetrations yielded a weak taste zone before or after a strong taste response was encountered. A strong response to stimulation was characterized by an increase in the

Fig. 1. Photographs of oscilloscope traces of multiunit activity in response to taste stimulation of anterior tongue. A Example of a typical weak taste response. This type of response is characterized by an increase in the baseline height. The arrowheads mark 12 seconds of stimulus application. Water rinse starts at the same time the stimulus is stopped. Scale bar is 10 seconds. B: The portion of the record in A marked by the dashed lines. The time scale in this record has been expanded to show the response more clearly. Scale bar in A represents 2 seconds for record in B. The arrowhead marks stimulus onset. C: Example of strong taste recorded at the site indicated by the star in Figure 5C. The arrowheads mark the 12 seconds of stimulus application, followed by water rinse. Scale bar in A represents 10 seconds for record C.

were angled 34” off vertical and advanced through the cerebellum. The zero reference point for the rostral-caudal and medial-lateral electrode coordinates was the calamus scriptorius-that is, the convergence of the gracile funiculi at the caudal end of the fourth ventricle. Taste-responsive activity was generally found about 4.0 mm rostral and 1.6 mm lateral to this convergence. The zero reference point for the dorsal-ventral coordinates was the remaining surface of the cerebellum after aspiration. The depth of taste-

Fig. 2 (see pages 711-7131, A-F: Low-power photomicrographs of 20 pm-thick cresyl-violet-stainedtransverse sections through the PBN. Section A is the most caudal; section F is the most rostral. A’-F‘: Line drawings corresponding to photomicrographs. Section numbers indicate rostrocaudal levels through PBN, relative to the caudal (section 1) and rostral (section 45) poles of the PBN. A, B: Caudal PBN with only the medial division present. The lateral division is not present and the superior cerebellar peduncle (SCP) has not separated from the cerebellum Only the central medial (cm) subdivision is present in A. All three medial subdivisions-cm, dorsal medial (dm), and ventral medial (vm)-are present in B. C, D: Middle PBN where both medial and lateral division cells are well represented. All three medial subdivisions are still present, but reduced in size. The SCP has separated from the cerebellum and the ventral spinocerebellar tract (vsc) is present. The lateral subdivisions present are central lateral (cl), dorsal lateral (dl), ventral lateral (vl), and external lateral (el). The “waist” area as defined in rat (Fulwiler and Saper, ’84)is outlined by the dashed line in D’. E, F: I n rostral PBN, only the lateral division is present. By this level, dl and vl have ended and the superior lateral (sl) subdivision replaces el in section F. The internal lateral subdivision (il) is present. The sl, il, and cl subdivisions are gradually replaced by the lateral lemniscus, about 200 +m rostral to section F. For all the sections, dorsal is toward the top of the page, medial is to the right, and the scale bar is 300 +m.

PONTINE TASTE AREA IN HAMSTER

711

712

C.B. HALSELL AND M.E. FRANK

PONTINE TASTE AREA IN HAMSTER

713

714

frequency of spikes well above baseline spontaneous activity. A typical strong response is shown in Figure 1C. The locations of both the weak and strong taste responses were used to define a taste-responsive area. This definition does tend to overestimate the taste area by somewhat less than the width of the weak taste-responsive zone. Iontophoretic deposits of HRP were used to mark the electrode tip location, by anodal current passing through the recording electrode (10 seconds, 2 pA, 10 ms on/off). This technique allows a very small ( < 100 pm diameter) spot of HRP to be deposited (McPheeters et al., '90). The small HRP deposit can be placed very close to an area of interest, which reduces error when locations of electrode tracks are reconstructed after the tissue has been histologically processed. In the majority of the experiments, the deposit was placed one electrode penetration (100 pm) away from a penetration where taste-responsive activity was detected. Two deposits were made: one medial and one lateral to the group of penetrations with taste activity. Often, deposits were made at either or both the dorsal and ventral extremes of taste-responsive activity within a single electrode penetration. No more than eight deposits were made in any one experiment.

C.B. HALSELL AND M.E. FRANK tected (30% sucrose), sectioned at 20 pm, and stained with cresyl violet as above. Subdivisions were defined on the basis of somal shape, size, and density by using low-power objectives. Figure 2 is a series of low-power photomicrographs of six levels through the PBN. In this particular animal, the PBN extended through 45 sections (20 pm thick). To calculate neuronal size and shape, soma containing nuclei with a visible nucleolus were drawn by using a drawing tube attachment on a Leitz Orthoplan 2 microscope and a high-power objective (X63 oil immersion). The somal long axis and perpendicular short axis were measured. From these measurements, the average diameter and a relative measure of shape were calculated. The shape factor is derived from the ratio of the width to the length, where a perfect circle has a ratio of one.

RESULTS Anatomical topography

The parabrachial nucleus is composed of two cell groups, the medial and lateral divisions, separated by the superior cerebellar peduncle (SCP) or brachium conjuctivum. In the Histology hamster, the average rostrocaudal length of the complete At the conclusion of mapping sessions, each animal was PBN (both medial and lateral divisions) was about 900 pm. given a lethal dose of pentobarbital (200 mgkg animal The medial division extends throughout the caudal half of weight i.p.) and perfused transcardially with 20 ml of 0.1 M the PBN, with an average rostrocaudal length of 550 pm. phosphate buffer (pH 7.3), containing 0.002% CaCl, and The lateral division extends throughout the rostra1 two0.05% lidocaine, followed by 100 ml fixative (1%paraformal- thirds of the PBN, with an average rostrocaudal length of dehyde and 2.5% glutaraldehyde in phosphate buffer). 650 pm. These two divisions overlap in the middle third of Following perfusion, the head was placed back on the the PBN by about 250 pm. Within each of these two headholder and insect pins were driven into brain areas divisions, multiple subdivisions can be distinguished. Medial division. The bulk of the medial division is distant from the pons by means of the same electrode holder employed during the experiment. The insect pins situated between the SCP and the intermingled cells of the were used as guides, and the brain was blocked at the same mesencephalic trigeminal and locus coeruleus nuclei (Me5/ angle as the recording electrode angle. The tissue was LC). The ventral border curves along the ventromedial edge placed in 30% sucroseiphosphate buffer and subsequently of the SCP extending ventrally close to the principal frozen sectioned in the transverse plane at either 20 or 40 sensory trigeminal nucleus and Kolliker-Fuse nucleus. The pm on a sliding microtome. The sections were processed dorsal border abuts the fourth ventricle. The medial diviaccording to the Hanker-Yates method (Hanker et al., '77) sion is composed of three subdivisions: the dorsal medial, and counterstained for Nissl substance with cresyl violet central medial, and ventral medial. The central medial subdivision (cm) is the largest of the (Chroma-Gesellschaft). Drawings were made of sections containing the visible medial division subnuclei, measuring 500-600 pm rostroHRP reaction product. The locations of electrode tracks caudally (Fig. 2A-D). At its widest point it measures were then plotted on these drawings on the basis of the 600-700 pm mediolaterally and 200-250 pm dorsovendistances of tracks from the deposits as determined by trally. Caudally, cm cells gradually replace the superior micromanipulator readings. In many instances the elec- vestibular nucleus. This transition is clearly evident due to trode tracks were visible in the tissue. The sites of taste- the difference in size between the large vestibular neurons responsive activity were then related t o PBN subdivisions. and the smaller PBN cells (Fig. 2A). Rostrally, cm extends There was only a small amount of shrinkage (about 10%) the furthest of the medial division subnuclei, ending where associated with the tissue processing, as attested by known the SCP contacts Me5/LC as the peduncle moves medially distances between adjacent visible tracks or HRP deposits. toward its decussation in the midbrain. Compare Figure 2, The locations of taste activity were not corrected for sections D and E, which are 160 pm apart. shrinkage, which introduced an error of up to 10% in the In Nissl-stained transverse sections, cm contains three locations of our recording sites. Including this error, the general cell types. The most distinct cell type is that of the anterior tongue taste responses are confined to two distinct large, dark-staining elongate cells which have an average subdivisions. diameter of 13.54 2 0.34 (S.E.M.) and a shape factor of The anatomical topography of the hamster PBN was 0.51 i 0.02 (S.E.M.).These cells have an oblique orientaexamined in Nissl-stained material. Three normal animals tion when viewed in transverse sections (Fig. 3B). The large were anesthetized and perfused as above, except the perfu- round-oval cells have an average diameter of 13.37 c 0.26 sate was 100 ml of a 4% paraformaldehyde/O.l M phosphate and a shape factor of 0.74 2 0.02 (Fig. 3B). These palebuffer solution. The head was left undisturbed for at least 2 staining cells are also generally oriented obliquely. The hours following perfusion; the brain was removed and medium round cells have an average diameter of 9.02 r blocked in the transverse plane. The block was cryopro- 0.15 and a shape factor of 0.85 +- 0.09 (Fig. 3B), tending to

PONTINE TASTE AREA IN HAMSTER

715

Fig. 3. High-power photomicrographs of the cell types in the taste-responsive regions of PBN. A: Cell types in the taste-responsive ventral lateral (vl) subdivision and the neighboring dorsal lateral (dl) and external lateral (el) subdivisions. The elongate cells of vl are oriented obliquely; the curved arrow points to one example. The more round cells of the dorsal lateral (dl) subdivision sit dorsal to vl. B: There

are three cell types in the central medial subdivision: large elongate (curved arrows), large oval (large arrows), and small round (open arrows). Notice oblique orientation of the elongate and oval cells. For both photomicrographs dorsal is toward the top of the page and lateral is left. The scale bar is 40 p,m.

stain more darkly than the large round-ovals, but they do not have any particular orientation. The dorsal medial subdivision (dm) is a cell-sparseregion sitting dorsal to cm and abutting the fourth ventricle (Fig. 2B-D). The cells in dm are mostly light-staining medium round-ovals with an average diameter of 8.79 t 0.3 and a shape factor of 0.73 0.03. A few cells similar to the large round-ovals in cm are also present. The ventral medial subdivision (vm) borders cm ventrally and extends ventrally and laterally to include cells intercalated within ventrally streaming fibers of the SCP (Fig. 2B-D). Cells ofvm are similar to those in cm, but there tend to be proportionally more elongate cells. These elongates are oriented more horizontally than those of cm. Lateral division. The lateral division occupies the rostral two-thirds of the PBN (Fig. 2C-F) and is bordered dorsally and laterally by the ventral spinocerebellar tract, except in the rostralmost PBN, where the lateral border is the lateral lemniscus. Rostrally, the lateral division is bordered by the pedunculopontine tegmental nucleus. The SCP forms the medial border and the ventral border extends to the principal sensory trigeminal nucleus and the dopaminergic A-5 cell group. The lateral division is composed of six subdivisions: the central lateral, dorsal lateral, external lateral, internal lateral, superior lateral, and ventral lateral.

The central lateral subdivision (cl) is the largest of the lateral subdivisions and is present throughout the complete rostrocaudal extent of the lateral division (Fig. 2C-F). Cells of cl first appear caudally where the SCP exits from the cerebellum and extends the furthest rostral of any PBN subdivision. The cells of cl are mostly small round cells (average diameter 6.6 0.21, shape factor 0.85 ? 0.02) and medium oval cells (average diameter 7.51 -C 0.81, shape factor 0.6 5 0.02). The nuclei of the round cells are more centered within the soma, whereas those of the oval cells are more eccentric. Most of cl appears as a loose collection of intermingled round and oval cells. The internal lateral subdivision (ill is a fairly homogeneous cluster of the round cells, located in the rostral- and dorsalmost part of cl (Fig. 2E,F). These cells appear to be the same type as the cl round cells. The dorsal lateral (dl) and ventral lateral (vl) subdivisions are two thin strips of cells located in the middle third of the PBN, situated just ventral to cl (Fig. 2C,D). These two subdivisionsare squeezed between the ventral spinocerebellar tract and the SCP. The vl subdivision is very small, measuring only 100-200 pm rostrocaudally, 400 pm mediolaterally, and 50 pm dorsoventrally. The cells of vl are medium and large elongate cells (average diameter 9.75 ? 0.27 and 11.22 & 0.23, respectively, shape factor 0.41 0.1, Fig. 3A) and are oriented obliquely, curving

*

*

*

C.B. HALSELL AND M.E. FRANK

716

ical border is reflected by the anatomically defined ventral border of cm, where there is very little overlap between the cells of cm and Me5/LC. In contrast, the dorsal border of cm is composed of “fingers” of cells extending dorsally between the fascicles of SCP fibers (Figs. 2C,D, 5). Taste responses in lateral division. Taste responses in the lateral division were restricted to the dorsal edge of the SCP, corresponding to vl. Responses were detected at only one rostrocaudal level in each of seven animals. The very small size of vl, as described above, may account for the limited success in mapping taste responses in this area. It is possible that this subdivision was missed in these other experiments since the rostrocaudal electrode tracks were 100 pm apart. However, at the rostrocaudal level where taste was found, in five of the seven experiments, taste responses were observed in successive mediolateral penetrations through this region (Figs. 4B, 5B), demonstrating a taste region of about 200-400 pm. In the other two experiments, only a single penetration was positive for taste responses in vl, although other electrode tracks penetrated this subdivision. This made it impossible to determine the mediolateral extent of taste within vl in these cases. The Functional topography mediolateral borders of the taste-responsive region are not The pons was mapped for the location of taste-responsive definitively defined due to the distances between electrode neurons in 24 animals. In all 24 animals taste activity was penetration, and because vl cells are intermingled with cells found in the medial division. Taste-responsive activity was medially and laterally. In all cases when taste-evoked also detected in the lateral division, but only in seven activity was detected in vl, it was always found throughout animals. In three animals the complete PBN was examined the complete dorsoventral extent of vl. The ventral border in a grid pattern consisting of 100-200 pm steps in the of taste responses corresponded very well to the border of vl horizontal plane. For each of these three animals, 34-48 and the peduncle. Because the extent of taste-responsive penetrations were made through the pons, with 12-16 neurons within vl was not defined completely, it is possible penetrations yielding taste-responsive activity. The details that not all of vl is specific for anterior tongue taste of this report are based on these three cases. Incomplete stimulation. maps for the other 21 animals were consistent with these Although the extent of taste responses in vl was not three cases. completely mapped, it is clear that taste responses in the Taste responses in medial division. Neurons responsive lateral division are limited to vl. Taste responses were never to anterior tongue stimulation were generally in the middle detected in the thin band of small round cells of dl which and lateral parts of the medial division, within and re- borders vl dorsolaterally. The lateral border of taste activity stricted to cm (Fig. 2). No taste-responsive activity was was always at the border of vl and el. No taste activity was found in either the dm or vm subdivisions bordering cm detected rostral and medial to vl in cl. (Fig. 2). Taste-responsive activity was never detected in any other The location of taste-responsive areas within cm are PBN subdivision besides cm and vl. Cases in which taste illustrated in Figures 4 and 5, each showing a separate activity was detected were used to establish this exclusivemapping experiment. Both experiments demonstrate the ness. In the initial search for the PBN in hamster, no taste tightly organized taste-responsive region within the medial activity at all was observed in some experiments. The division. As can be seen in Figure 4(C-E) and Figure 5(C), electrode tracks in these experiments were localized to successive penetrations revealed that taste-responsive neu- subdivisions other than cm or vl. Occasionally taste was rons form a continuous, uninterrupted zone within cm. detected from cells between the fascicles of the SCP. The This taste-responsive cell group generally measured 300- cells within the SCP fascicles did not have an appearance 400 pm long rostrocaudally, extending from the caudal similar to any other PBN cell type. They twisted around border of cm to almost the rostral extent of cm, where SCP fiber bundles, which gave most a curved, fusiform appearcontacts Me5/LC. Taste responses were generally not found ance. at this rostralmost extent of cm, although in one experiment weak taste activity was detected. At its widest point, the taste-responsive group measured 3 0 0 4 0 0 pm mediolatDISCUSSION erally, but at the rostral and caudal extremes it generally Anatomical topography narrowed to only 100 pm (see Fig. 4C-E). Within a single The anatomical organization of the PBN in hamster has electrode track, taste responses typically extended from the ventral edge of the SCP to about 100 pm dorsal to the been briefly described previously (Halsell and Frank, ’89; Me5ILC complex. The total size of this taste-responsive cell Whitehead, ’90). The anatomy of the PBN had also been described in rat, rabbit, cat, monkey, and human, but group is about two-thirds the total size of cm. The ventral border of the taste-responsive group is the without precise definitions of the subdivisions (see Kolesamost discrete. As the electrode moved ventrally through a rova and Petrovicky, ‘87, for review). Only more recently taste-responsive area, strong taste responses changed to has the cytoarchitectonic organization of the PBN been well weak taste responses and weak taste responses changed to described in rat (Fulwiler and Saper, ’84; Kolesarova and no taste responses in less than 50 pm. This sharp physiolog- Petrovicky, ’87).

along the dorsolateral edge of the SCP (Figs. 2D, 3B). Bordering vl dorsolaterally, dl is composed of small roundoval dark-staining cells with an average diameter of 6.48 0.18 and a shape factor of 0.64 0.02 (Fig. 3B). The external lateral subdivision (el) extends throughout the rostral PBN. In the middle PBN, it is situated ventral to dl and vl (Fig. 2C-E). This subdivision is ovoid shaped in transverse sections and is very distinct since it is the most cell dense of any PBN subdivision. Most of the cells are darkly staining oval-multipolar cells (average diameter 8.93 0.18, shape factor 0.66 2 0.02). Often the long axes of the cells are oriented vertically or obliquely. These cells are distinctly characterized by having numerous proximal dendrites visible with Nissl stains. The superior lateral subdivision replaces el rostrally (sl, Fig. 2F). The transition between el and sl is gradual, though evident due to the lower cell density of sl. The cells of sl are very similar to el but have no particular orientation and are lighter staining. Rostrally, sl is replaced by the lateral lemniscus.

*

*

*

PONTINE TASTE AREA IN HAMSTER

717

C A D

E

Fig. 4. Example of experiment from one animal where the complete PBN was mapped. The line drawings of transverse sections show reconstructed electrode penetrations through the PBN. Section A is from the rostral third PBN; section F is just caudal to the PBN. Vertical lines indicate the location of individual electrode tracks. The shaded bars on some of the vertical lines indicate where tasteresponsive activity occurred along those electrode tracks. Tasteresponsive activity was detected only in the central medial (cm) and ventral lateral (vl) subdivisions. Both weak and strong taste responses are included in the shaded bars. The irregularly shaped dark spots are

the locations of horseradish peroxidase deposits (B, C, E). The filled circles and asterisks above the electrode tracks relate to the inset in the upper left. Lateral is to the left and dorsal is toward the top of the page. The sections are 100 pm apart except for B and C, which are 200 pm apart. Inset: A horizontal view of the electrode track penetrations. The six levels marked A-F correspond to the six drawn sections. The filled circles indicate electrode tracks where no taste was detected. The asterisks indicate electrode tracks where taste-responsive activity was detected. Lateral is left and rostral is toward the top of the page. Scale bar is 1mm for all drawings.

C.B. HALSELL AND M.E. FRANK

718

TABLE 1. Comparison of PBN Subnuclei Between Hamster and Rat

Rat Hamster Halsell and Frank (present study)

L

cm central medial dm dorsal medial vm ventral medial

m m m exm

cl

central lateral

C

medial medial medial external medial central lateral

dl el

dorsal lateral external lateral

d el

dorsal lateral external lateral

il sl vl

internal lateral superior lateral ventral lateral

I

internal lateral superior lateral ventral lateral

-

Fig. 5. Example of an experiment in one animal showing line drawings of three transverse sections through the caudal, middle, and rostra1 PBN. Taste-responsive activity was detected in both the central medial (section C)and ventral lateral (section B) subdivisions. No taste responses were detected in the rostralmost PBN (section A), which includes the "waist" area. Taste responses were not observed in the dorsal medial (dm) or ventral medial (vm) subdivisions. Orientation and data presentation as in Figure 4, but the scale bar is 0.5 mm.

Fulwiler and Saper ('84)

S V

-

Kolesarova and Petrovicky ('87) M M Ve

medialis medialis ventralis

DM dorsomedialis, DL dorsolateralis

-

L DL D

lateralis. dorsolateralis dorsalis

DL

dorsolateralis

-

The general anatomical organization of the PBN in the hamster is similar to the previously described PBN in the rat (Fulwiler and Saper, '84; Kolesarova and Petrovicky, '87). The present study in hamster and the two previous studies in rat have based the parceling of the PBN on cytoarchitectonic criteria using Nissl stains. The only real differences between these studies are variations in the number, and subtle localizations, of subnuclei. The more recent PBN parceling in the rat based the names of the subdivisions on general direction (Kolesarova and Petrovicky, '87), whereas the earlier study in rat named subdivisions as parts of the traditional medial and lateral divisions (Fulwiler and Saper, '84). The designation of medial and lateral divisions is based on the vertical orientation of the SCP in the transverse plane in the human (Olszewski and Baxter, '82). This designation is somewhat confusing in hamster and rat since the SCP is oriented obliquely causing the lateral division to actually be situated more dorsolateral and the medial division to be situated ventromedial. However, for consistency in the literature and to aid in comparisons between species, the general terminology of the subdivisions as introduced by Fulwiler and Saper ('84) for the rat has been kept. See Table 1 for a comparison of the PBN subdivisions in the rat and hamster and the study by Petrovicky and Kolesarova ('89) comparing the general organization of the PBN in birds and selected mammals, including human. In hamster, the medial division has been parceled into three subdivisions, whereas in rat, the medial division was only divided into two subnuclei (Fulwiler and Saper, '84; Kolesarova and Petrovicky, '87). In hamster, three cell types are evident in cresyl-violet-stained material: large elongate, large round-oval, and medium round. Three similar cell types have been reported in rat by using thionin staining: large fusiform or multipolar, medium polygonal, and small round cells (Fulwiler and Saper, '84). All three cell types in rat have larger average diameters than their corresponding types in hamster. The differences between average diameters of the largest and smallest cells in rat are the same as in hamster. The absolute difference in size may be due to histological techniques rather than a difference between species. Also in rat, but using cresyl-violet, this same area was reported to contain only two cell types: medium oval and small round cells (Kolesarova and Petrovicky, '87). No soma1 sizes were reported for these cells. The cell-sparse dorsal medial subdivision, as defined herein, has not been described previously. This region was included in the medial subdivision, or not mentioned, in previous anatomical characterizations of the rat PBN (Ful-

PONTINE TASTE AREA IN HAMSTER wiler and Saper, '84; Kolesarova and Petrovicky, '87). In hamster, vm is considered a separate subdivision from cm because no taste activity was detected in this area and it contains a higher proportion of elongate cells, which are oriented in a somewhat more horizontal fashion than those in cm. This area is similar to that described as the subnucleus ventralis in the rat based on cresyl-violet staining (Kolesarova and Petrovicky, '87). However, this area was not separated into a distinct subdivision based on thionin staining in the rat, but considered a concentration of fusiform or multipolar cells along, but within, the ventrolatera1 border of the medial subdivision (Fulwiler and Saper, '84).The most laterally located cells of vm (intercalated within the ventrally streaming fibers of SCP) in hamster were considered in rat to be part of subnucleus ventralis (Kolesarovaand Petrovicky, '87) or a separate subdivision, the external medial (Fulwiler and Saper, '84). In hamster, the lateral division has been divided into six subnuclei, whereas in rat, this division was divided into either four or eight subnuclei (Fulwiler and Saper, '84; Kolesarova and Petrovicky, '87). The description of lateral subdivisions in rat (Fulwiler and Saper, '84) and our lateral subdivisions in hamster are in close correspondence. One major difference of interest is that in rats the ventral lateral subdivision is situated further medial, closer to and including the "waist" area (Fulwiler and Saper, '84).The waist area is located in the rostral two-thirds of the PBN, where the medial portion of the SCP constricts, forming a bridge of cells connecting the medial and lateral divisions of the PBN. Some investigators consider the waist area as a separate subdivision, the subnucleus interstitialis (Kolesarova and Petrovicky, '87). In hamster, vl is situated more laterally and does not include the waist area (Fig. 2D', outlined by the dashed line). This is because the cell types in the waist area appear similar to those of cl. In rat thionin-stained material, the ventral lateral subdivision is reported to contain cell types like the central lateral, but more densely packed (Fulwiler and Saper, '84). In rat cresyl-violet-stained material, this same area is defined as the subnucleus dorsolateralis, which contains only oval-shaped cells (Kolesarova and Petrovicky, '87). The cells in the hamster vl seem to be more elongate than those in the corresponding area in rat. In the course of this study, tissue stained histochemically for acetylcholinesterase, cytochrome oxidase, NADHdehydrogenase, and NADPH-diaphorase was screened to determine the usefulness of enzymatic markers in subdividing the PBN. These enzymatic markers were not as useful in subdividing the PBN as Nissl stains (unpublished observations). The staining pattern for these enzymes is fairly uniform throughout the PBN, though the el subdivision does consistently stain lighter than the rest of the subnuclei with actelycholinesterase and cytochrome oxidase. Also, with each of these markers, numerous fibers which course through vm are visible in unstained relief, which helps to distinguish between vm and cm. The lack of correlation between Nissl-definedsubdivisionsand histochemical staining patterns in the PBN is in marked contrast to the nucleus of the solitary tract. In this nucleus, cytochrome oxidase and NADH-dehydrogenase have clear differential staining patterns which are consistent with cytoarchitecturally defined subdivisions and primary afferent terminal fields (Whitehead, '88; Halsell et al., '89; Lasiter and Kachele, '89; Herbert et al., '90).

719

Functional topography Taste responses in medial division. Taste-responsive areas in the medial division described in this report appear similar to those previously reported for anterior tongue stimulation in hamsters, rats, and rabbits (Norgren and Pfaffmann, '75; Van Buskirk and Smith, '81; DiLorenzo and Schwartzbaum, '82; Travers and Smith, '84). It is difficult to determine exactly the taste-responsive locations in these reports since only the borders of SCP and MeS/LC and the dorsal surface of the pons, and not the boundaries of the PBN, are drawn. In the hamster studies, most single units are reported to be located in the caudal third of the PBN and in the middle of the medial division along the ventral edge of SCP (Van Buskirk and Smith, '81; Travers and Smith, '84). Generally these units seem to be located within our multiunit-mapped taste-responsive area in cm, which is described for the first time in this study. The only mammalian PBN systematically mapped for taste responses is the rat (Norgren and Pfaffmann, '75). This map for anterior tongue stimulation based on multiunit recordings is similar to our results in the hamster. The majority of responses are located in the caudal and lateral half of the medial division. However, the medial taste area in the rat is larger in that it encompasses more of the medial division than in the hamster. A possible explanation for this difference is that in the rat study the locations of evoked activity from multiple animals were fit onto one prototypical brain. Small errors associated with reconstructing the recording sites can be compounded because of variable alterations of tissue size during histological processing and differences among brains of the same species. These sources of error can be greatly minimized or eliminated by completely mapping an area for multiunitevoked activity in a single animal as was done in the present study. Compare the sections in Figures 4D and 5C, which are similar rostrocaudal locations in two different animals, but the location of the taste-responsive groups is slightly different within cm. Thus, reconstructing sites to an individual, not a prototypical, brain may account for the finding that the pontine taste area reported here is smaller than those previously described. There does not seem to be any difference in the relative sizes of the anatomically defined subnuclei between the rat and hamster as discussed above. Anatomically, the rostral and caudal borders of cm are not very clear. There is mixing of the cells in cm and the superior vestibular nucleus caudally and cm and the pedunculopontine tegmentum nucleus rostrally. This could account for the finding that the taste-responsive area in cm is narrower at the rostral and caudal borders than in the middle (Fig. 4C-E), which was also seen in the rat multiunit study (Norgren and Pfaffmann, '75). The borders of the taste map suggest that the tasteresponsive portion of cm is a separate subdivision. However, anatomically cm appears fairly uniform with respect to the distribution of cell types in Nissl-stained material (Fig. 2). There is a slight difference in cell density within cm, where the more centrally located cells are more densely packed. Since the taste-responsive area overlaps this area, this slight difference in cell density does not seem to be functionally significant in terms of the organization of anterior-tongue taste. It is possible that the nontasteresponsive part of cm represents the other oral cavity taste fields. To establish this, complete multiunit maps need to be made for each taste field. However, the oral fields do not

720 appear to have distinct spatial representations within cm, since in rats, single units responsive to anterior and posterior tongue stimulation were intermingled (Norgren and Pfaffmann, '751, and some single units respond to both tongue and palate stimulation (Ogawa et al., '82). In rat, taste responses were not seen in the area corresponding to the hamster ventral medial subdivision (Norgren and Pfaffmann, '75; Van Buskirk and Smith, '81; Travers and Smith, '84).In hamster, taste activity was also never detected in the most medial portion of the medial division, the dm subdivision. Previous reports in rat agree with this finding. Taste responses in lateral division. The complete borders of the taste area in the lateral division have not been as well defined as the borders of the medial taste area. Our mapping of the vl cell group was incomplete and tasteresponsive areas in this very small subdivision could have been skipped over as the electrode was moved in 100 pm steps. This is one reason why the physiologically defined boundaries of vl taste are not clear rostrocaudally and mediolaterally. Also, vl cells are somewhat intermingled with surrounding cells of the cl and el subdivisions. In this study, the taste-responsive sites were restricted to the dorsal edge of the SCP, within the vl subdivision. In previous studies in the hamster and rat, the majority of the lateral taste-responsive sites are also along the dorsal edge of the SCP (Norgren and Pfaffmann, '75; Van Buskirk and Smith, '81;Ogawa et al., '84,'87; Travers and Smith, '84; Hill, '87). In rats, the reported taste-responsive sites along the dorsal edge of SCP appear to be located within the ventral lateral subdivisiondefined anatomically by Fulwiler and Saper ('84;Norgren and Pfaffmann, '75; Ogawa et al., '84,'87; Hill, '87). However, there is a difference between the lateral taste area in hamsters and rats. As discussed above, the ventral lateral subdivision is situated further medial, closer to, and including the "waist" area in rats (Fulwiler and Saper, '84).In hamsters, vl is situated more laterally and does not include the waist area. Although the majority of reported lateral taste sites appear to correspond to vl, some units have been located in other parts of the lateral division. In one study in hamster, a group of taste-responsive neurons appear to be located very dorsally in the rostral lateral division (Van Buskirk and Smith, '81),which may constitute a third group of taste-responsive neurons. In the present study, taste responses were never encountered this far rostral and dorsal nor were they in other studies (Norgren and Pfaffmann, '75; Ogawa et al., '84,'87; Travers and Smith, '84;Hill, '87). Marking lesions placed 0.5-1.0 mm above or below recording sites (Van Buskirk and Smith, '81)may introduce an error of +- 150-200 pm (DiLorenzoand Schwartzbaum, '82). Even these seemingly small errors could make it appear as if the lateral taste group is situated in a more dorsal position. In the present study, we have tried to minimize this problem by using multiple small deposits of horseradish peroxidase (Fig. 6). Because the deposits have less adverse effects on the tissue integrity they can be placed closer to the area of interest, thereby reducing errors introduced by reconstructing sites via measurements. In many cases deposits were made right at the dorsal andlor ventral border between taste-responsive and nontasteresponsive activity. In the present study, taste-responsive

C.B. HALSELL AND M.E. FRANK

Fig. 6. A: Photomicrograph of section used to draw Figure 3B, showing the relationship of the horseradish peroxidase (HRP) deposits and the ventral lateral subdivision (vl).An HRP deposit is visible on the left side of the photo, which marks the lateral border of vl. The large arrow marks the medial border of vl. An HRP deposit with a center located about 80 pm rostral to this section was just visible at this point. B: Photomicrograph of section used to draw Figure 4C, showing the relationship of the HRP deposits and the central medial subdivision (cm). In this section two HRP deposits were made in the same electrode track marking the dorsal and ventral borders of taste activity. Having both on the same section helps verify that the sections are indeed parallel to the tracks. The arrow marks the location of a third HRP deposit, which is faint since the center of this deposit is located about 60 pm rostral to this section. Scale bar for both photomicrographs Is 300 pm.

activity was tested every 25 pm, to minimize this type of measurement error. In rats, some taste-responsive sites were also reported in a very lateral area of the parabrachial region (Norgren and Pfaffmann, '75). This area appears in rat to be the external medial subdivision situated at the ventralmost end of SCP (Fulwiler and Saper, '84).In hamster this area corresponds to the lateralmost part of the ventral medial subdivision. We never found taste in this area, which is consistent with other reports on hamsters (Van Buskirk and Smith, '81; Travers and Smith, '84). Taste responses in superior cerebellar peduncle. In the present study, multiunit taste-responsive sites were detected within the SCP but not as frequently as reported previously for rat (Norgren and Pfaffmann, '75; Ogawa et al., '84,'87; Hermann and Rogers, '85; Hill, '87) or hamster (Van Buskirk and Smith, '81;Travers and Smith, '84). Many cells are seen between the fascicles of the SCP in the caudal part of the PBN in rats (Fulwiler and Saper, '84).In

PONTINE TASTE AREA IN HAMSTER the hamster, on the other hand, there seem to be fewer large groups of cells and more single cells spread out between the fiber bundles of the SCP (Figs. 2,6). Although not included in any PBN subdivision, the taste-responsive neurons within the SCP can be considered part of the pontine taste area. Significance of two taste areas. The finding of two separate taste areas may be related to the general organization of sensory modalities within the PBN. In rodents, the medial division is considered to have a gustatory function (special visceral) since this area contains neurons responsive to gustatory stimuli, as discussed above, and receives direct ascending input from the gustatory rostra1 nucleus of the solitary tract (Norgren and Leonard, '73; Norgren, '78, '84; Travers, '88). The lateral PBN division is considered to have a general visceral autonomic function and receives ascending input from the visceral caudal nucleus of the solitary tract (Loewy and Burton, '78; Norgren, '78; Kalia and Sullivan, '82; Cechetto and Calaresu, '83; Cechetto, '87). Lateral division neurons respond to multiple general visceral sensory modalities such as respiratory cycles, tooth pulp noxious stimuli, cardiovascular states, such as hemorrhage (rapid changes in blood pressure), and fluid regulation (Hugelin and Vibert, '74; Mraovitch et al., '82; Hermann and Rogers, '85; Ohman and Johnson, '86; Rehnig et al., '86; Ward, '89). Spinothalamic and spinoamygdaloid pathways carrying nociceptive information also relay through the lateral PBN (Cechetto et al., '85; Bernard et al., '89; Hylden et al., '89). Since the gustatory nucleus of the solitary tract in rat and hamster projects heavily to both cm and vl, as well as lightly to other subdivisions (Norgren, '78; Travers, '88; unpublished observations), it is possible that the two parts of the pontine taste area receive the same gustatory information. The role of the two taste areas would then be to differently process and route ascending gustatory information. The medial taste area may be in a pathway involved in discrimination of taste qualities for cognition via efferent projections to the thalamus and cortex (Norgren and Leonard, '73; Lasiter et al., '82; Fulwiler and Saper, '84; Yamamoto and Kitamura, 'go), as well as providing memorial and nonvoluntary aspects of taste processing via efferent projections to the amygdala (Fulwiler and Saper, '84). In contrast, the lateral taste area could be involved in a pathway of converging gustatory and visceral information via efferent projections to the amygdala, hypothalamus, and bed nucleus of stria terminalis (Saper and Loewy, '80; Fulwiler and Saper, '84). Convergence of gustatory and general visceral information does occur in the general area of the ventral lateral subdivision, where neurons respond to co-activation of vagal and gustatory afferents (Hermann and Rogers, '85). The PBN may be the first central synapse for this convergence since no co-activationis detected in the nucleus of the solitary tract, the first central synapse for the gustatory system (Hermann et al., '83). Convergence of gustatory and visceral information is necessary for autonomic regulation and certain behaviors, such as conditioned taste aversions. A discrete group of taste neurons in the visceral domain of the PBN, as found in the vl subdivision, would be an appropriate anatomical substrate for this processing. It should be noted, however, that there does not appear to be a complete separation of PBN efferents. For example, both medial and ventral lateral cells project to the thalamic taste area in rat (Fulwiler and Saper, '84). It is

721 possible that the efferents of visceral, gustatory, and multimodal PBN neurons are intermingled. In summary, within the hamster parabrachial nucleus there are two separate and anatomically distinct subdivisions that contain neurons responsive to taste stimulation of the anterior tongue. The central medial subdivision is large, comprising most of the medial division. The ventral lateral subdivision is a small group of cells along the dorsal edge of the superior cerebellar peduncle. It is speculated that these two groups of cells are involved in different gustatory functions.

ACKNOWLEDGMENTS The authors thank Dr. Michael F. Huerta for lending us his invaluable scientific expertise and Drs. D. Kent Morest and Ken Hutson for advice on the normal anatomy and for a critical reading of a previous version of this manuscript. This work was supported by NIH grants DC 00853, DC 00168. and DE 07131.

LITERATURE CITED Beckstead, R.M., J.R. Morse, and R. Norgren (1980) The nucleus of the solitary tract in the monkey: Projections to the thalamus and brain stem nuclei. J. Comp. Neurol. 190:259-282. Bernard, J.F., M. Peschanski, and J.M. Besson (1989) A possible spino (trigeminol-ponto-amygdaloid pathway for pain. Neurosci. Lett. 100~8388. Cechetto, D.F. (1987) Central representation of visceral function. Fed. Proc. 46(1): 17-23. Cechetto, D.F., and F.R. Calaresu (1983) Parabrachial units responding to stimulation of buffer nerves and forebrain in the cat. Am. J. Physiol. 245:R811-R819. Cechetto, D.F., D.G. Standaert, and C.B. Saper (1985) Spinal and trigeminal dorsal horn projections to the parabrachial nucleus in the rat. J. Comp. Neurol. 240:153-160. DiLorenzo, P.M., and J.S. Schwartzbaum (1982) Coding of gustatory information in the pontine parabrachial nuclei of the rabbit Magnitude of neural response. Brain Res. 251t229-244. Frank, M. (1973) An analysis of hamster afferent taste nerve response functions. J. Gen. Physiol. 62~588-618. Frank, M.E. (1991) Taste responsive neurons of the glossopharyngeal nerve of the rat. J. Neurophysiol. (in press). Frank, M.E., S.L. Bieber, and D.V. Smith (1988) The organization of taste sensibilities in hamster chorda tympani nerve fibers. J. Gen. Physiol. 91:861-896. Fulwiler, C.E., and C.B. Saper (1984) Subnuclear organization of the efferent connections of the parabrachial nucleus in the rat. Brain Res. Rev. 7:229-259. Halsell, C.B., and M. Frank (1989) Cytoarchitecture of the pontine taste area in the golden hamster. Chem. Senses 24:707. Halsell, C., L. Savoy, M. Huerta, and M. Frank (1989) Cytochrome oxidase activity in the gustatory central nervous system. Chem. Senses 24:209. Hanamori, T., I.J. Miller, and D.V. Smith (1988) Gustatory responsiveness of fibers in the glossopharyngeal nerve. J. Neurophysiol. 60:47&498. Hanker, J.S., P.E. Yates, C.B. Metz, and A. Rustioni (1977) A new specific, sensitive and non-carcinogenic reagent for the demonstration of horseradish peroxidase. J. Histochem. 9: 789-792. Herbert, H., M.M. Moga, and C.B. Saper (1990) Connections of the parabrachial nucleus with the nucleus of the solitary tract and the medullary reticular formation in the rat. J. Cornp. Neurol. 293~540-580. Hermann, G.E., and R.C. Rogers (1985) Convergence of vagal and gustatory afferent input within the parabrachial nucleus of the rat. J. Auton. Nerv. Syst. 13:1-17. Hermann, G.E., N.J. Kohlerman, and R.C. Rogers (1983) Hepatic-vagal and gustatory afferent interactions in the brainstem of the rat. J. Auton. Nerv. Syst. 9:477495. Hill, D.L. (1987) Development of taste responses in the rat parabrachial nucleus. J. Neurophysiol. 57:481495.

722 Hugelin, B.A., and J.F. Vibert (1974) A stereotaxic model of pneumotaxic oscillator based on spatial and temporal distributions of neuronal bursts. J. Neurophysiol. 37: 9 1- 107. Hylden, J.L., F. Anton, and R.L. Nahin (1989) Spinal lamina I projection neurons in the rat: Collateral innervation of parabrachial area and thalamus. Neuroscience 285'7-37. Kalia, M., and J.M. Sullivan (1982) Brainstem projections of sensory and motor components of the vagus nerve in the rat. J. Comp. Neurol. 211.248-264. Kolesarova, D., and F'. Petrovicky (1987) Parabrachial nuclear complex in the rat (Nuclei Parabrachiales and Nucleus Koelliker-Fuse). J. Hirnforsch. 5517-527. Lasiter, P.S., and D.L. Kachele (1989) Elevated NADH-dehydrogenase activity characterizes the rostral gustatory zone of the solitary nucleus in rat. Brain Res. Bull. 22.777-781. Lasiter, P.S., D.L. Glanzman, and P.A. Mensah (1982) Direct connectivity between pontine taste areas and gustatory neocortex in rat. Brain Res. 234:lll-121. Loewy, A.D., and H. Burton (1978) Nuclei of the solitary tract: Efferent projections to the lower brain stem and spinal cord of the cat. J. Comp. Neurol. 181;421-450. McPheeters, M., T.P. Hettinger, S.C. Nuding, L.D. Savoy, M.C. Whitehead, and M.E. Frank (1990) Taste-responsive neurons and their locations in the solitary nucleus of the hamster. Neuroscience 34:745-758. Mraovitch, S., M. Kumada, and D.J. Reis (1982) Role of the nucleus parabrachialis in cardiovascular regulation in cat. Brain Res. 23257-75. Norgren, R. (1978) Projections from the nucleus of the solitary tract in the rat. Neuroscience 3.207-218. Norgren, R. (1984) Central neural mechanisms of taste. In J.M. Brookhart, V.B. Mountcastle, and I. Darian-Smith (eds): Handbook of Physiology, Section 1:The Nervous System, Vol. 111. Bethesda, Maryland: American Physiological Society, pp. 1087-1128. Norgren, R., and C.M. Leonard (1971) Taste pathways in rat brainstem. Science 173,1136-1139. Norgren, R., and C.M. Leonard (1973) Ascending central gustatory pathways. J. Comp. Neurol. 150:217-238. Norgren, R., and C. Pfaffmann (1975) The pontine taste area in the rat. Brain Res. 91:99-117. Ogawa, H., T. Hayama, and S. Ito (1982) Convergence of input from tongue and palate to the parabrachial nucleus neurons of rats. Neurosci. Lett. 28:9-14.

C.B. HALSELL AND M.E. FRANK Ogawa, H., T. Hayama, and S. Ito (1984) Location and taste responses of parabrachio-thalamic relay neurons in rats. Exp. Neurol. 83t507-517. Ogawa, H., T. Hayama, and S. Ito (1987) Response properties of the parabrachio-thalamic taste and mechanoreceptive neurons in rats. Exp. Brain Res. 68t449457. Ohman, L.E., and A.K. Johnson (1986) Lesions in lateral pardbrachial nucleus enhance drinking to angiotensin I1 and isoproteronal. Am. J. Physiol. 251:R504-R509. Olszewski, J., and D. Baxter (1982) Cytoarchitecture of the Human Brain Stem (2nd ed.). Basel: Karger. Petrovicky, P., and Kolesarova (1989) Parabrachial nuclear complex. A comparative study of its cytoarchitectonics in birds and some mammals including man. J. Hirnforsch. 5:539-550. Rehnig, H.P., J. Brankack, and F. Klingberg (1986) Tooth pulp evoked potentials in trigeminal nuclei and parabrachial region of the freely moving rat. Biomed. Biochim. Acta 4.485489. Saper, C.B., and A.D. Loewy (1980) Efferent connections of the parabrachial nucleus in the rat. Brain Res. 197.291-317. Travers, J.B. (1988) Efferent projections from the anterior nucleus of the solitary tract of the hamster. Brain Res. 457:l-11. Travers, S.P., and D.V. Smith (1984) Responsiveness of neurons in the hamster parabrachial nuclei to taste mixtures. J. Gen. Physiol. 84:221250. Travers, S.P., C. Pfaffmann, and R. Norgren (1986) Convergence of lingual and palatal gustatory neural activity in the nucleus of the solitary tract. Brain Res. 365:305-320. Van Buskirk, R.L., and D.V. Smith (1981) Taste sensitivity of hamster parabrachial pontine neurons. J. Neurophysiol. 45:144-171. Ward, D.G. (1989) Neurons in the parabrachial nuclei respond to hemorrhage. Brain Res. 491:80-92. Whitehead, M.C. (1988) Neuronal architecture of the nucleus of the solitary tract in the hamster. J. Comp. Neurol. 276t547-572. Whitehead, M.C. (1990) Subdivisions and neuron types of the nucleus of the solitary tract that project to the parabrachial nucleus in the hamster. J. Comp. Neurol. 3021-21. Yamamoto, T., and R. Kitamura (1990) A search for the cortical gustatory area in the hamster. Brain Res. 51Ot309-320. Yamamoto, T., R. Matsuo, and Y. Kawamura (1980) The pontine taste area in the rabbit. Neurosci. Lett. 16:5-9.

Mapping study of the parabrachial taste-responsive area for the anterior tongue in the golden hamster.

The locations of taste-responsive areas within the brainstem parabrachial nucleus (PBN), an obligatory taste relay in the golden hamster (Mesocricetus...
2MB Sizes 0 Downloads 0 Views