- Email: [email protected]
Location and taste responses of parabrachio-thalamic relay neurons in rats
EXPERIMENTAL
NEUROLOGY
Location
83, 507-5 17 ( 1984)
and Taste Responses of Parabrachio-Thalamic Relay Neurons in Rats
HISASHI OGAWA, TOMIO
HAYAMA,
AND SHIN’ICHI
ITO’
Department of Physiology, Kumamoto University Medical School, Honjo 2-2-1, Kumamoto 860, Japan Received April 25, 1983; revision received October I 7, 1983 Thii of the 55 taste units in the parabrachial nucleus were activated antidromically by stimulation of either or both of the ipsi- or contralateral thalamic taste areas Such parabrachio-thalamic taste relay neurons produced bilateral thalamic a&rent fibers (B type, N = 14), exclusively ipsilateral thalamic adherent fibers (I type, N = 12), or exclusively contralateral thalamic a&ent fibers (C type, N = 4). Most of the B-type neurons were excited best by NaCl among the four basic taste stimti, approximately one-half the I-type neurons by HCl. Most of the NaCl-best neurons were located in the medial part of the parabrachisl nucleus but most of the HCl-best neurons were in the lateral part. In addition, NaCl-best neurons had shorter ipsilateral latencies (modal value = 1.0 to 3.0 ms) from the ipsilateral thalamic taste area, whereas HClbest neurons had longer latencies (modal value = 4.0 to 6.0 ms). INTRODUCTION
The parabrachial nucleus (PB) receives a&rent axons from the nucleus of the solitary tract and projects to the thalamic taste area (TTA) and other nuclei in rats (5-7, 9, 11, 13, 14, 18, 20). A recent anatomical study (20) suggested that three kinds of thalamic tierent 6bers originate from the PB neurons; i.e., bilateral, ipsilateral, and contralateral. Although several investigations (10, 12, 17, 19) have been made on taste responsiveness of the PB neurons, it has not been clarified how the cells of origin of thalamic afferent fibers in the PB respond to the four basic taste stimuli or how they are distributed in the nucleus. Abbreviations: PB-parabrachial nucleus, ‘lTA-thalamic taste area, P-T-parabrachio-thalamic, BC-bmchium conjunctivum. ’ The authors express their gratitude to Dr. R. Norgren for his useful discumion of a revision of the manuscript. This research was supported by a grant from the Ministry of Education in Japan (58106008). 507 0014-4886/84 $3.00 copyli&t Q 1984 by Aadernic AUrightsofnpmductioninmyformrrnred
Pteu, Inc.
508
OGAWA,
HAYAMA,
AND IT0
We examined taste responses and loci of PB neurons sending their axons to either or both of the ipsi- and contralateral TTA (parabrachio-thalamic or P-T relay neurons). We found that P-T cells in the lateral and medial subnucleus were differentiated in terms of thalamic projection patterns and taste response profiles. MATERIALS
AND
METHODS
Female Sprague-Dawley rats (220 to 250 g) were anesthetized with amobarbital sodium (80 mg/kg, i.p.). After cannulation of the trachea and femoral vein, the rats were mounted on a stereotaxic instrument, according to the F&ova and Marsala method (2) and cerebellectomized by gentle aspiration to expose the dorsal surface of the pons. Subsequently, the animals were immobilized with curare and artificially ventilated. Rectal temperature was maintained at about 37°C by an electric blanket, and the ECG was monitored. Endtidal PCOz was maintained at 3.5 to 4.5%. Extracellular activity from PB neurons was isolated using the criteria of Bishop et al. (1) and glass microelectrodes filled with 2% pontamine sky blue in 0.5 M sodium acetate. Action potentials were recorded through conventional physiologic equipment consisting of a preamplifier, cathode-ray oscilloscope, and kymographic camera. The taste solutions, 0.1 M NaCl, 0.5 M sucrose, 0.01 N HCl, and 0.02 A4 quinine-HCl, were delivered by gravity flow to the whole oral cavity including the tongue and palate, from a system of overhead funnels at a rate of 3 ml/s. Tilting the animal’s head at 45” with ipsilateral side up and dissecting the ipsilateral cheek from the mouth to an anterior border of the mandibular bone facilitated flow over the root of the posterior tongue and soft palate. After rinsing the oral cavity with distilled water for 10 s, each taste solution was administered into the oral cavity for 10 s followed by another rinse with distilled water for at least 10 s. Taste stimuli were applied at intervals of at least 1 min. Both the taste stimuli and the rinsing water were administered at room temperature (25 to 28°C). Responsiveness of the neurons to mechanical stimulation was also examined by stroking, brushing, or pinching the tongue and palate with a glass rod, or a nonserrated forceps. Taste and mechanical responses were differentiated by the following criteria (12)-taste responses outlasted sapid stimulus flow, but were not evoked during water application and mechanical responses were elicited by the onset of water rinse, taste stimulation, and mechanical stimulation. Temperature sensitivity was not usually studied in these experiments, except when the discharge of a neuron was transiently facilitated or inhibited by taste solutions or water rinse, but not by mechanical stimulation. For cooling the oral cavity, tap water at 20°C was applied to the oral cavity, after preadapting it to 40°C water, and vice versa for warm stimulation (15).
PARABRACHIO-THALAMIC
TASTE
RELAY NEURONS
509
For each neuron, a given taste stimulus was tested several times. When the neurons produced consistent responses to taste stimulation, but not to mechanical or thermal stimulation, the response was recorded on an FM tape for further analysis. Because sapid stimulation of either the front (3, 16) or back (2 1) of the tongue can result in long-latency (e.g., 5 s) responses, we developed the following criterion for identifying gustatory neurons: during the first 10 s after stimulus application, any changes in discharge rate at least 1.0 s long and 2 SD above or below the prestimulus average. This criterion might lead to a few false-positive responses due to momentary large variation in spontaneous activity, but repeated trials reduced this possibility. Magnitude of taste responses was defined as the number of impulses in the first 5 s after the onset of stimulation after subtracting the number of impulses of the spontaneous discharge. Bilateral monopolar tungsten electrodes, insulated to the tip (0.5 to 1 MCI), were positioned 3.0 to 3.4 mm posterior to bregma., 1.5 to 1.7 mm off the midline, and 6.0 mm ventral to the cerebral surface to stimulate the TTA on each side. Electrical pulses (duration, 0.2 ms; maximum intensity, 300 FA) were applied to these thalamic electrodes to ant&on&ally activate the pontine taste neurons. With this stimulus intensity, physical spread of current did not exceed 1.0 mm. In one experiment, we measured the threshold current required to antidromically activate a PB neuron at various depths of a stimulating electrode from the surface of the cerebral cortex. The lowest threshold was about 5 PA in the middle of the ipsilateral TTA, whereas a current of as much as 300 PA was required to activate the neuron when the stimulation electrode was moved in either way as much as 1.0 mm away from the most sensitive depth. Antidromic invasion of a pontine neuron was inferred from the following criteria: invariant latency, colbsion of antidromic spikes with orthodromic ones evoked by the taste stimulation, and the ability to follow repetitive stimulation at more than 200 Hz (13). All recording sites were marked by the electrophoretic deposition of the dye from the recording electrodes ( 12). The stimulation sites in the thalamus were marked by passing a cathodal current through the stimulating electrodes. After the experiment, the loci of the dye spots and electrolytic lesions were established histologically. RESULTS Thalamic Projection of Parabrachial Neurons. In 34 of the 148 rats used in our experiment, thalamic electrodes were accurately positioned in both the ipsi and contralateral TTAs. Loci of some thalamic stimulation sites are shown in Fig. 1. of the 55 PB taste neurons isolated in these 34 animals, 30 units were activated from either or both of the thalamic stimulating electrodes. Fourteen
510
OGAWA,
HAYAMA,
AND IT0
AP 2.9 m AP 3.1 coo
e
.I!-
, AP
3.3
caudal FIG. 1. Stimulation sites in the thalamic taste area (TTA) in five experiments. Different marks indicate stimulation sites in different experiments. Each section separated by 200 pm. F-fomix; FMT-fasciculus mammillothalamicus; FB-fasciculus retroflexus; LM-lemniscus medialis; PF-nucleus parafascicularis; VB-nucleus ventralis posterior, pars arcuatus et pars extema.
units (B type) were activated antidromically from both sides of the thalamus (Fig. 2). Another 12 (I type) were excited antidromically only from the ip silateral side and the remaining 4 units (C type), from the contralateral side.
IOSZ-9 ipsi
sucllUIUIYlllYllOlllllUYgllllrll
;,i-
HCI I
I
lll’u~
3,
q-
contra
III,,1
quinin;l
39 FIG. 2. Responses of a taste neuron projecting bilaterally to the thalamic taste area (1052-3). A-unit in response to four basic taste stimuli. An arrow indicates onset of taste stimulation. B-antidromic response to thalamic stimulations. ipsi, ipsilateral thalamic stimulation; contra, contralateral stimulation. a, antidromic responses without any preceding spikes; band c, antidromic responses with preceding spikes. In b, antidromic spikes were abolished in collision with the preceding orthodromic spikes. Dots and arrowheads mean spikes and onset of thalamic stimulation, respectively. Each record consists of five superimposed traces.
PARABRACHIO-THALAM
IC TASTE
RELAY NEURONS
511
Antidromic latencies of the I-type units ranged from 1.5 to 7.6 ms (mean + SD = 4.2 -t 2.0 ms, N = 12). B-type unit late&es ranged from 1.2 to 5.1 ms (2.5 f 1.2 ms, N = 14) to ipsilateral stimulation, but 1.7 to 7.2 ms (3.4 + 2.0 ms, N = 14) for the contralateral side. In 11 of the 14 B-type units, the antidromic latency was shorter in response to the ipsilateral rather than contralateral thalamic stimulation. In response to the &lateral thalamic stimulation, the B-type neurons produced a spike with a signikantly shorter latency than the I-type neurons (P < 0.05, Mann-Whitney U test). The antidromic latencies of the C-type neurons ranged from 1.9 to 6.2 ms (4.13 f 1.9 ms, N = 4), not signifkantly different from the contralateral latencies of the B-type neurons (P > 0.05). Location of the Parabrachio-Thalamic Relay Neurons. Thirty P-T relay neurons were located in the medial and lateral part of the caudal PB, as well as inside the brachium conjunctivum (BC). The distribution was similar to that of gustatory neurons that were not antidromically invaded from the thalarnic electrodes (7). At more rostral levels, taste neurons were lkquently found dorsolaterally; caudally, they were found more medioventrally (Pii. 3). Nine of the 14 B-type and three of the four C-type P-T cells were found in the medial part or below the BC, whereas 6 of the 12 I-type neurons were
FIG. 3. Location of the P-T neurons in the PB. Recording siteswere plotted in five mntative coronal sections of the PB. Each section was separated by 300 pm. X, 0, and A represent neuron4 with bilateral, ipklateral, and contralateral thalamic a&rents, respectively. BC-brachium conjunctivum; IC-colliculus inferior; LC-locus ceruleus; MesV-nucleus mesencephalici n. trigemini; MV-nucleus motorius n. tripemini; PV-nucleus principalis n. Wigemik; SOL-nuckus olivaris superior, SV-nucleus supratrigeminalis.
512
OGAWA,
HAYAMA,
AND IT0
in the lateral part, or above the BC. These differential distributions of B-type and I-type neurons were statistically significant (P < 0.05, Fisher’s exact probability test). In addition there was a tendency for C-type neurons to be distributed in the rostral and ventral portion of the nucleus, I-type neurons in the rostral and dorsal portion, and B-type neurons were in the caudal and ventral portion. Taste Responsiveness of Parabrachio-Thalamic Relay Neurons to Four Basic Stimuli. Taste responsiveness of P-T taste relay neurons was examined by applying each of the four basic taste stimuli to the whole oral cavity (Fig. 4). Some of the B-type neurons produced a large magnitude of responses to NaCl and some of the I-type neurons responded best to HCl or quinine. One of the C-type neurons was predominantly sensitive to sucrose. To characterize the taste responsiveness of the P-T relay and nonrelay taste neurons, taste neurons were classified according to the stimulus which excited the neurons most effectively (4, 12) (Fig. 3). Both P-T relay and nonrelay taste neurons were primarily in the NaCl-, sucrose-, or HCI-best groups. Different projection types of P-T relay neurons showed different best stimulus categories. Eight (57%) of the 14 B-type neurons were NaCl-best, 5 (42%) of the 12 I-type neurons were HCI-best, and three of the four Cl-type
B-type
200 1
C-type
m. L
N&l
HCI
span.
Parabrachial
dis
units
4. Taste response profiles of P-T neurons. Taste stimuli: 0.1 M NaCl, 0.5 M sucrw, 0.01 NHCl, and 0.02 Mquinine-HCl. Spon. dis.-the average number of spontaneous discharges before the taste stimulations. O-inhibitory responses. FIG.
PARABRACHIO-THALAM
IC TASTE
513
RELAY NEURONS
type neurons were sucrose-best. The response magnitude of B-type neurons to NaCl(168.4 f 202.2 impuIses/5 s, N = 14) was significantly larger than that of I-type (43.3 + 72.2 impulses/5 s, N = 12), or C-type neurons (17.8 + 19.7 impulses/5 s, N = 4). (P < 0.01, Mann-Whitney U test). There was no significant difference in the response to other taste stimuli among three types of P-T cells. Quinine-best neurons (N = 2) were found only in the Itype neuron group. D@kentiation of Neurons in D@wnt Categories of Taste profiles. Antidromic latencies also varied among various best stimuhrs categories of PT neurons (Fig. 5). NaCl-best neurons showed short antidromic latencies in response to ipsilateral TTA stimulation, ranging from 1.2 to 4.2 ms with a mode at 1.0 to 3.0 ms, whereas the antidromic latencies of the HCl-best neurons ranged from 1.9 to 7.6 ms with two modes at 2.0 to 3.0 ms and 4.0 to 6.0 ms, signilicantly longer than those of the NaCl-best neurons (P < 0.05, Mann-Whitney U test). Four of the seven sucrose-best P-T neurons were excited antidromically from the ipsilateral TTA, three at a very short latency (1.3 to 2.4 ms), and the remaining one had a longer latency (7.6 ms). Neurons in different best-stimulus categories were local&d in different parts of the PB (Table 1). Nine of the 11 NaCl-best neurons were in the medial part of the nucleus, and six of the nine HCl-best neurons were in the lateral part. NaCl-best and HCl-best neurons were reciprocally distributed in the medial and lateral part of the PB (P < 0.01, Fisher’s exact probability test). In addition, the response magnitude to NaCl of the P-T cells in the E-type
I-type
WYW
quinine-beat ipsi Am-
c 0
I. 5 Antidromic
latency
L 0 (me)
a ’ 5
c
5. Latency distribution of various groups of P-T taste relay neurons in response to antidromic stimulation of the thalamic taste area. P-T taste relay neurons were m by the best of the four basic taste stimuli. ipsi-Responses to antidrmnic stimulation of the ipdated thalamic taste afea; contra-responses to the contralateral thalamic taste area FIG.
514
OGAWA,
HAYAMA, TABLE
AND IT0 1
Location of Various Types of Parabrachio-Thalamic (P-T) Relay Taste Neurons in the Parabrachial Nucleus (PB) and inside the Brachium Conjunctivum (BC) Type of P-T relay taste neurons” Location Lateral part of the PB Inside the BC Medial part of the PB Total
NaCl-best
sucrose-best
HCI-best
Quinine-best
Total
1 1 9
2 2 3
6 3 0
1 1 0
10 I 12
11
I
9
2
29
“P-T relay taste neurons producing excitatory responses to at least one of the four basic stimuli are listed.
medial part (136.0 f 132.8 impulses/5 s, N = 14) was significantly larger than those in the lateral part (15.7 + 14.1 impulses/5 s, N = 10) (P < 0.05, Mann-Whitney U test) but not different from those inside BC (146.6 + 248.5 impulses/5 s, N = 6). There was no difference in the response magnitude to HCl of the P-T cells in the medial and lateral part of the nucleus and the BC (33.2 f 32.4 impulses/5 s, N = 14; 22.7 + 22.5 impulses/5 s, N = 10; 99.9 f 8 1.O, N = 6) (P > 0.05, Mann-Whitney U test). Seven sucrose-best neurons were invariantly found in all three regions. Response magnitudes to sucrose of the P-T cells within the BC (27.1 f 16.4 impulses/5 s, N = 6) were significantly larger than those in the lateral part of the PB ( 18.1 + 40.1 impulses/5 s, N = 10) (P < 0.05, Mann-Whitney U test) but not different from those in the medial part (49.1 + 96.7 impulses/5 s, N = 14). Chemotopic localization was not observed with the non-P-T neuron group. DISCUSSION We found that parabrachial taste neurons were activated antidromically from ipsilateral, contralateral, or bilateral thalamic taste areas. When we identified taste responses, every precaution was used to differentiate them from mechanical or thermal responses. Some taste responses, however, occurred only at a long latency after the onset of stimulation, particularly when the circumvallate papilla was stimulated (2 1). For this reason, we adopted a less rigid criterion for defining gustatory responses (see Methods), which may have raised the risk of false positives. Three Groups of Thalamic Aferent Fibers from the Parabrachial Nucleus. Anatomical studies have shown that the PB projects to the TTA bilaterally
PAlMBR4CHIO-
THALAhJIC
TASTE
RELAY
NEURONS
515
(9, 11, 20), and that three kinds of thalamic atferent fibers originate from the PB: ipsiIateral, contraIateral, and bilateral (20). Our study revealed that most axons arising from taste neurons in the PB terminated either ipsiIateraUy or bilaterally in the thalamus. Double labeling of PB neurons with dyes injected into the TTA revealed that about 60% of the ipsilaterally projecting neurons were double-labeled, whereas about 40% of the contralaterally projecting neurons were double-labeled (20). In our experiment, 14 of the 26 (54%) P-T neurons activated antidromically from the ipsilateral TTA also were excited from the contralateral ‘ITA, and 14 of the 18 (78%) activated antidromically from the contralateral TTA were driven from the ipsilateral side as well. A somewhat smaller ratio of the C-type neurons was found in our study than expected from the anatomic data. Most axons of PB neurons projecting to the contralateral TTA pass close to the ipsilateral TTA (9). Therefore, it is possible that some of C-type neurons were identified as B type as a result of activation of fibers of passage near the ipsilateral TTA. Since the ipsi- and contralateral TTA are separated by at least 2 mm, it seems unlikely that an electrode in one TTA would stimulate axons in the opposite TTA even with the maximal strength current (300 PA) used in this study (see Methods). Fibers projecting to the forebrain or amygdala pass posterior and ventral to the ipsilateral TTA. Electrophysiological experiments (7-9) indicate that more than one-half the P-T taste neurons send collaterals to the amygdala or forebrain. A recent anatomical study (20), however, could not identify PB neurons that send axon collaterals to both the TTA and amygdala. In our study, it is possible that some of the I-type neurons sent axons to the amygdala, but were misidentified as I-type by activation of fibers of passage. This question remains to be answered. In our study, the three different types of relay neurons could be differentiated in other aspects. Many B-type neurons were located in the caudal and ventral PB, many I-type neurons in the rostral and dorsal PB, and C-type neurons in the rostral and ventral PB. The differential distribution of these P-T neurons was similar to the distribution identified anatomically (20). In addition, in response to ipsilateral thalamic stimulation, antidromic late&es of B-type neurons were sign&antly shorter than those of I-type neurons. In the cat (6), PB neurons projecting to the thalamus or amygdala also are topographically distributed. Parabrachial neurons projecting to the amygdala are larger than those to the thalamus. D@rential Taste SpeciJicity of Three DlQ$erent ReIay Neurons. The B-type neurons produced significantly larger responses to NaCl, and many were NaCl-best. On the other hand, some of the I-type neurons responded well to HCl or quinine, and many of these were HCl-best. The NaCl-best and HCl-best P-T neurons were more differentiable based on their locus in the
516
OGAWA,
HAYAMA,
AND IT0
PB and antidromic latencies than were the same neurons classified as B-type and I-type (Table 1). Neural responses to NaCl were significantly larger in the medial part of the PB than in the lateral part. This suggests that taste quality is represented topographically. The present study clarified that the P-T relay cells were somewhat chemotopically localized in the PB, and that the major groups of P-T cells, those with bilateral or ipsilateral projections, may be of different celI sizes. Neurons in the lateral part of the PB are excited by stimulation of the posterior oral cavity, those in the medial part by the anterior tongue (10). Given the present data, the medial and lateral part of the PB may receive differential inputs and have differential thalamic projections. Topographic localization of each category of the pontine taste neurons, classified according to the best stimulus, was also noticed in the hamster (19). In our study in the rat, topographic arrangement was evident only for the P-T relay neurons. REFERENCES 1. BISHOP, P. O., W. BURKE, AND R. DAVIS. 1962. The identification of single units in central visual pathways. J. Physiol. (London) 162: 409-431. 2. FIFKOVA, E., AND J. MARSALA. 1967. Stereotaxic atlases for the cat, rabbit and rat. Pages 653-695 in J. BURES, M. PETRAN, AND J. ZADER, Eds., Electrophysio~ogicai Methods in Biological Research. Academic Press, New York. 3. PISHMAN, I. Y. 1957. Single fiber gustatory impulses in rat and hamster. J. Cell Comp. Physiol. 49: 3 19-334. 4. PRANK, M. 1973. An analysis of hamster atferent taste nerve response functions. J. Gen. Physiol.
61: 588-618.
5. LASITER, P. S., AND D. L. GLANZMAN, 1983. Axon collaterals of pontine taste area neurons project to the posterior ventromedial thalamic nucleus and to the gustatory neocortex. Brain
Res. 258: 299-304.
6. NOMURA, S., N. MIZUNO, K. ITOH, T. SUGIMOTO, AND Y. NAKAMURA. 1979. Localization of parabrachial neurons projecting to the thalamus and the amygdala in the cat using horseradish peroxidase. Exp. Neural. 64: 375-385. 7. NORGREN, R. 1974. Gustatory alferents to ventral forebrain. Brain Res. 81: 285-295. 8. NORGREN, R. 1976. Taste pathways to hypothalamus and amygdala. J. Comp. Neural. 166: 3 l-48. 9.
NORGREN, R., AND C. M. LEONARD. 1974. Ascending central gustatory pathways. J. Cornp. Neurol.
150: 217-238.
10. NORGREN, R., AND C. PFAFFMANN. 1975. The pontine taste area in the rat. Brain Res. 91: 99-l 17. 11. OGAWA, H., AND T. AKAGI. 1978. Afferent connections to the posteromedial ventral nucleus from the pons and the rostra1 medulla in the rat. Kumamoto Med. J. 31: 54-62. 12. 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. 13. OCAWA, H., T. IMOTO, AND T. HAYAMA. 1980. Taste relay neurons in the sohtaty tract nucleus of rats. Neurosci. Lett. 18: 295-299. 14. OGAWA, H., AND J. KAISAKU. 1982. Physiological characteristics of the solitario-parabrachial relay neurons with tongue afferent inputs in rats. Exp. Brain Res. 48: 362-368.
PARABRACHIO-THALAMIC
TASTE
RELAY
NEURONS
517
15. OOAWA, H., M. SATO, AND S. YAMANITA. 1968. Multiple sensitivity of chorda tympani fibres of the rat and hamster to gustatory and thermal stimuli. J. Physiol. (London) 199: 233-240. 16. @JAWA, H., M. SATO, AND S. Y AMAsHIT& 1973. variability in impulse discharges in rat chorda tympani fibers in response to repeated gustatory stimulations. Physic. Behav. 11: 469479. 17. P~~~crrro, R. S., AND T. R. !kwrr. 1976. Gustatoti neural coding in the pans. Brain Rex 110: 283-300.
18. SAPER, C. B., AND A. D. LCWEY. 1982. Werent connections of the pambrachial nucleus in the rat. Brain Res. 197: 291-317. 19. VAN BUSKIRK, R. L., AND D. V. SMITH. 1981. Taste sensitivity of hamster parabrachial pontine neurons. J. Neurophysid 45: 144-17 1. 20. WISHART, K., AND D. VAN DER Kooy. 1981. The organization of the e&rent projections of the parabrachial nucleus to the forebrain in the rat: a retrograde fluorescent doublelabelling study. Brain Res. 212: 27 l-286. 2 1. YAMADA, K. 1966. Gustatory and thermal responses in the glossopharyn8eal nerve of the rat. Jap. J. Physiol. 16: 599-611.
NEUROLOGY
Location
83, 507-5 17 ( 1984)
and Taste Responses of Parabrachio-Thalamic Relay Neurons in Rats
HISASHI OGAWA, TOMIO
HAYAMA,
AND SHIN’ICHI
ITO’
Department of Physiology, Kumamoto University Medical School, Honjo 2-2-1, Kumamoto 860, Japan Received April 25, 1983; revision received October I 7, 1983 Thii of the 55 taste units in the parabrachial nucleus were activated antidromically by stimulation of either or both of the ipsi- or contralateral thalamic taste areas Such parabrachio-thalamic taste relay neurons produced bilateral thalamic a&rent fibers (B type, N = 14), exclusively ipsilateral thalamic adherent fibers (I type, N = 12), or exclusively contralateral thalamic a&ent fibers (C type, N = 4). Most of the B-type neurons were excited best by NaCl among the four basic taste stimti, approximately one-half the I-type neurons by HCl. Most of the NaCl-best neurons were located in the medial part of the parabrachisl nucleus but most of the HCl-best neurons were in the lateral part. In addition, NaCl-best neurons had shorter ipsilateral latencies (modal value = 1.0 to 3.0 ms) from the ipsilateral thalamic taste area, whereas HClbest neurons had longer latencies (modal value = 4.0 to 6.0 ms). INTRODUCTION
The parabrachial nucleus (PB) receives a&rent axons from the nucleus of the solitary tract and projects to the thalamic taste area (TTA) and other nuclei in rats (5-7, 9, 11, 13, 14, 18, 20). A recent anatomical study (20) suggested that three kinds of thalamic tierent 6bers originate from the PB neurons; i.e., bilateral, ipsilateral, and contralateral. Although several investigations (10, 12, 17, 19) have been made on taste responsiveness of the PB neurons, it has not been clarified how the cells of origin of thalamic afferent fibers in the PB respond to the four basic taste stimuli or how they are distributed in the nucleus. Abbreviations: PB-parabrachial nucleus, ‘lTA-thalamic taste area, P-T-parabrachio-thalamic, BC-bmchium conjunctivum. ’ The authors express their gratitude to Dr. R. Norgren for his useful discumion of a revision of the manuscript. This research was supported by a grant from the Ministry of Education in Japan (58106008). 507 0014-4886/84 $3.00 copyli&t Q 1984 by Aadernic AUrightsofnpmductioninmyformrrnred
Pteu, Inc.
508
OGAWA,
HAYAMA,
AND IT0
We examined taste responses and loci of PB neurons sending their axons to either or both of the ipsi- and contralateral TTA (parabrachio-thalamic or P-T relay neurons). We found that P-T cells in the lateral and medial subnucleus were differentiated in terms of thalamic projection patterns and taste response profiles. MATERIALS
AND
METHODS
Female Sprague-Dawley rats (220 to 250 g) were anesthetized with amobarbital sodium (80 mg/kg, i.p.). After cannulation of the trachea and femoral vein, the rats were mounted on a stereotaxic instrument, according to the F&ova and Marsala method (2) and cerebellectomized by gentle aspiration to expose the dorsal surface of the pons. Subsequently, the animals were immobilized with curare and artificially ventilated. Rectal temperature was maintained at about 37°C by an electric blanket, and the ECG was monitored. Endtidal PCOz was maintained at 3.5 to 4.5%. Extracellular activity from PB neurons was isolated using the criteria of Bishop et al. (1) and glass microelectrodes filled with 2% pontamine sky blue in 0.5 M sodium acetate. Action potentials were recorded through conventional physiologic equipment consisting of a preamplifier, cathode-ray oscilloscope, and kymographic camera. The taste solutions, 0.1 M NaCl, 0.5 M sucrose, 0.01 N HCl, and 0.02 A4 quinine-HCl, were delivered by gravity flow to the whole oral cavity including the tongue and palate, from a system of overhead funnels at a rate of 3 ml/s. Tilting the animal’s head at 45” with ipsilateral side up and dissecting the ipsilateral cheek from the mouth to an anterior border of the mandibular bone facilitated flow over the root of the posterior tongue and soft palate. After rinsing the oral cavity with distilled water for 10 s, each taste solution was administered into the oral cavity for 10 s followed by another rinse with distilled water for at least 10 s. Taste stimuli were applied at intervals of at least 1 min. Both the taste stimuli and the rinsing water were administered at room temperature (25 to 28°C). Responsiveness of the neurons to mechanical stimulation was also examined by stroking, brushing, or pinching the tongue and palate with a glass rod, or a nonserrated forceps. Taste and mechanical responses were differentiated by the following criteria (12)-taste responses outlasted sapid stimulus flow, but were not evoked during water application and mechanical responses were elicited by the onset of water rinse, taste stimulation, and mechanical stimulation. Temperature sensitivity was not usually studied in these experiments, except when the discharge of a neuron was transiently facilitated or inhibited by taste solutions or water rinse, but not by mechanical stimulation. For cooling the oral cavity, tap water at 20°C was applied to the oral cavity, after preadapting it to 40°C water, and vice versa for warm stimulation (15).
PARABRACHIO-THALAMIC
TASTE
RELAY NEURONS
509
For each neuron, a given taste stimulus was tested several times. When the neurons produced consistent responses to taste stimulation, but not to mechanical or thermal stimulation, the response was recorded on an FM tape for further analysis. Because sapid stimulation of either the front (3, 16) or back (2 1) of the tongue can result in long-latency (e.g., 5 s) responses, we developed the following criterion for identifying gustatory neurons: during the first 10 s after stimulus application, any changes in discharge rate at least 1.0 s long and 2 SD above or below the prestimulus average. This criterion might lead to a few false-positive responses due to momentary large variation in spontaneous activity, but repeated trials reduced this possibility. Magnitude of taste responses was defined as the number of impulses in the first 5 s after the onset of stimulation after subtracting the number of impulses of the spontaneous discharge. Bilateral monopolar tungsten electrodes, insulated to the tip (0.5 to 1 MCI), were positioned 3.0 to 3.4 mm posterior to bregma., 1.5 to 1.7 mm off the midline, and 6.0 mm ventral to the cerebral surface to stimulate the TTA on each side. Electrical pulses (duration, 0.2 ms; maximum intensity, 300 FA) were applied to these thalamic electrodes to ant&on&ally activate the pontine taste neurons. With this stimulus intensity, physical spread of current did not exceed 1.0 mm. In one experiment, we measured the threshold current required to antidromically activate a PB neuron at various depths of a stimulating electrode from the surface of the cerebral cortex. The lowest threshold was about 5 PA in the middle of the ipsilateral TTA, whereas a current of as much as 300 PA was required to activate the neuron when the stimulation electrode was moved in either way as much as 1.0 mm away from the most sensitive depth. Antidromic invasion of a pontine neuron was inferred from the following criteria: invariant latency, colbsion of antidromic spikes with orthodromic ones evoked by the taste stimulation, and the ability to follow repetitive stimulation at more than 200 Hz (13). All recording sites were marked by the electrophoretic deposition of the dye from the recording electrodes ( 12). The stimulation sites in the thalamus were marked by passing a cathodal current through the stimulating electrodes. After the experiment, the loci of the dye spots and electrolytic lesions were established histologically. RESULTS Thalamic Projection of Parabrachial Neurons. In 34 of the 148 rats used in our experiment, thalamic electrodes were accurately positioned in both the ipsi and contralateral TTAs. Loci of some thalamic stimulation sites are shown in Fig. 1. of the 55 PB taste neurons isolated in these 34 animals, 30 units were activated from either or both of the thalamic stimulating electrodes. Fourteen
510
OGAWA,
HAYAMA,
AND IT0
AP 2.9 m AP 3.1 coo
e
.I!-
, AP
3.3
caudal FIG. 1. Stimulation sites in the thalamic taste area (TTA) in five experiments. Different marks indicate stimulation sites in different experiments. Each section separated by 200 pm. F-fomix; FMT-fasciculus mammillothalamicus; FB-fasciculus retroflexus; LM-lemniscus medialis; PF-nucleus parafascicularis; VB-nucleus ventralis posterior, pars arcuatus et pars extema.
units (B type) were activated antidromically from both sides of the thalamus (Fig. 2). Another 12 (I type) were excited antidromically only from the ip silateral side and the remaining 4 units (C type), from the contralateral side.
IOSZ-9 ipsi
sucllUIUIYlllYllOlllllUYgllllrll
;,i-
HCI I
I
lll’u~
3,
q-
contra
III,,1
quinin;l
39 FIG. 2. Responses of a taste neuron projecting bilaterally to the thalamic taste area (1052-3). A-unit in response to four basic taste stimuli. An arrow indicates onset of taste stimulation. B-antidromic response to thalamic stimulations. ipsi, ipsilateral thalamic stimulation; contra, contralateral stimulation. a, antidromic responses without any preceding spikes; band c, antidromic responses with preceding spikes. In b, antidromic spikes were abolished in collision with the preceding orthodromic spikes. Dots and arrowheads mean spikes and onset of thalamic stimulation, respectively. Each record consists of five superimposed traces.
PARABRACHIO-THALAM
IC TASTE
RELAY NEURONS
511
Antidromic latencies of the I-type units ranged from 1.5 to 7.6 ms (mean + SD = 4.2 -t 2.0 ms, N = 12). B-type unit late&es ranged from 1.2 to 5.1 ms (2.5 f 1.2 ms, N = 14) to ipsilateral stimulation, but 1.7 to 7.2 ms (3.4 + 2.0 ms, N = 14) for the contralateral side. In 11 of the 14 B-type units, the antidromic latency was shorter in response to the ipsilateral rather than contralateral thalamic stimulation. In response to the &lateral thalamic stimulation, the B-type neurons produced a spike with a signikantly shorter latency than the I-type neurons (P < 0.05, Mann-Whitney U test). The antidromic latencies of the C-type neurons ranged from 1.9 to 6.2 ms (4.13 f 1.9 ms, N = 4), not signifkantly different from the contralateral latencies of the B-type neurons (P > 0.05). Location of the Parabrachio-Thalamic Relay Neurons. Thirty P-T relay neurons were located in the medial and lateral part of the caudal PB, as well as inside the brachium conjunctivum (BC). The distribution was similar to that of gustatory neurons that were not antidromically invaded from the thalarnic electrodes (7). At more rostral levels, taste neurons were lkquently found dorsolaterally; caudally, they were found more medioventrally (Pii. 3). Nine of the 14 B-type and three of the four C-type P-T cells were found in the medial part or below the BC, whereas 6 of the 12 I-type neurons were
FIG. 3. Location of the P-T neurons in the PB. Recording siteswere plotted in five mntative coronal sections of the PB. Each section was separated by 300 pm. X, 0, and A represent neuron4 with bilateral, ipklateral, and contralateral thalamic a&rents, respectively. BC-brachium conjunctivum; IC-colliculus inferior; LC-locus ceruleus; MesV-nucleus mesencephalici n. trigemini; MV-nucleus motorius n. tripemini; PV-nucleus principalis n. Wigemik; SOL-nuckus olivaris superior, SV-nucleus supratrigeminalis.
512
OGAWA,
HAYAMA,
AND IT0
in the lateral part, or above the BC. These differential distributions of B-type and I-type neurons were statistically significant (P < 0.05, Fisher’s exact probability test). In addition there was a tendency for C-type neurons to be distributed in the rostral and ventral portion of the nucleus, I-type neurons in the rostral and dorsal portion, and B-type neurons were in the caudal and ventral portion. Taste Responsiveness of Parabrachio-Thalamic Relay Neurons to Four Basic Stimuli. Taste responsiveness of P-T taste relay neurons was examined by applying each of the four basic taste stimuli to the whole oral cavity (Fig. 4). Some of the B-type neurons produced a large magnitude of responses to NaCl and some of the I-type neurons responded best to HCl or quinine. One of the C-type neurons was predominantly sensitive to sucrose. To characterize the taste responsiveness of the P-T relay and nonrelay taste neurons, taste neurons were classified according to the stimulus which excited the neurons most effectively (4, 12) (Fig. 3). Both P-T relay and nonrelay taste neurons were primarily in the NaCl-, sucrose-, or HCI-best groups. Different projection types of P-T relay neurons showed different best stimulus categories. Eight (57%) of the 14 B-type neurons were NaCl-best, 5 (42%) of the 12 I-type neurons were HCI-best, and three of the four Cl-type
B-type
200 1
C-type
m. L
N&l
HCI
span.
Parabrachial
dis
units
4. Taste response profiles of P-T neurons. Taste stimuli: 0.1 M NaCl, 0.5 M sucrw, 0.01 NHCl, and 0.02 Mquinine-HCl. Spon. dis.-the average number of spontaneous discharges before the taste stimulations. O-inhibitory responses. FIG.
PARABRACHIO-THALAM
IC TASTE
513
RELAY NEURONS
type neurons were sucrose-best. The response magnitude of B-type neurons to NaCl(168.4 f 202.2 impuIses/5 s, N = 14) was significantly larger than that of I-type (43.3 + 72.2 impulses/5 s, N = 12), or C-type neurons (17.8 + 19.7 impulses/5 s, N = 4). (P < 0.01, Mann-Whitney U test). There was no significant difference in the response to other taste stimuli among three types of P-T cells. Quinine-best neurons (N = 2) were found only in the Itype neuron group. D@kentiation of Neurons in D@wnt Categories of Taste profiles. Antidromic latencies also varied among various best stimuhrs categories of PT neurons (Fig. 5). NaCl-best neurons showed short antidromic latencies in response to ipsilateral TTA stimulation, ranging from 1.2 to 4.2 ms with a mode at 1.0 to 3.0 ms, whereas the antidromic latencies of the HCl-best neurons ranged from 1.9 to 7.6 ms with two modes at 2.0 to 3.0 ms and 4.0 to 6.0 ms, signilicantly longer than those of the NaCl-best neurons (P < 0.05, Mann-Whitney U test). Four of the seven sucrose-best P-T neurons were excited antidromically from the ipsilateral TTA, three at a very short latency (1.3 to 2.4 ms), and the remaining one had a longer latency (7.6 ms). Neurons in different best-stimulus categories were local&d in different parts of the PB (Table 1). Nine of the 11 NaCl-best neurons were in the medial part of the nucleus, and six of the nine HCl-best neurons were in the lateral part. NaCl-best and HCl-best neurons were reciprocally distributed in the medial and lateral part of the PB (P < 0.01, Fisher’s exact probability test). In addition, the response magnitude to NaCl of the P-T cells in the E-type
I-type
WYW
quinine-beat ipsi Am-
c 0
I. 5 Antidromic
latency
L 0 (me)
a ’ 5
c
5. Latency distribution of various groups of P-T taste relay neurons in response to antidromic stimulation of the thalamic taste area. P-T taste relay neurons were m by the best of the four basic taste stimuli. ipsi-Responses to antidrmnic stimulation of the ipdated thalamic taste afea; contra-responses to the contralateral thalamic taste area FIG.
514
OGAWA,
HAYAMA, TABLE
AND IT0 1
Location of Various Types of Parabrachio-Thalamic (P-T) Relay Taste Neurons in the Parabrachial Nucleus (PB) and inside the Brachium Conjunctivum (BC) Type of P-T relay taste neurons” Location Lateral part of the PB Inside the BC Medial part of the PB Total
NaCl-best
sucrose-best
HCI-best
Quinine-best
Total
1 1 9
2 2 3
6 3 0
1 1 0
10 I 12
11
I
9
2
29
“P-T relay taste neurons producing excitatory responses to at least one of the four basic stimuli are listed.
medial part (136.0 f 132.8 impulses/5 s, N = 14) was significantly larger than those in the lateral part (15.7 + 14.1 impulses/5 s, N = 10) (P < 0.05, Mann-Whitney U test) but not different from those inside BC (146.6 + 248.5 impulses/5 s, N = 6). There was no difference in the response magnitude to HCl of the P-T cells in the medial and lateral part of the nucleus and the BC (33.2 f 32.4 impulses/5 s, N = 14; 22.7 + 22.5 impulses/5 s, N = 10; 99.9 f 8 1.O, N = 6) (P > 0.05, Mann-Whitney U test). Seven sucrose-best neurons were invariantly found in all three regions. Response magnitudes to sucrose of the P-T cells within the BC (27.1 f 16.4 impulses/5 s, N = 6) were significantly larger than those in the lateral part of the PB ( 18.1 + 40.1 impulses/5 s, N = 10) (P < 0.05, Mann-Whitney U test) but not different from those in the medial part (49.1 + 96.7 impulses/5 s, N = 14). Chemotopic localization was not observed with the non-P-T neuron group. DISCUSSION We found that parabrachial taste neurons were activated antidromically from ipsilateral, contralateral, or bilateral thalamic taste areas. When we identified taste responses, every precaution was used to differentiate them from mechanical or thermal responses. Some taste responses, however, occurred only at a long latency after the onset of stimulation, particularly when the circumvallate papilla was stimulated (2 1). For this reason, we adopted a less rigid criterion for defining gustatory responses (see Methods), which may have raised the risk of false positives. Three Groups of Thalamic Aferent Fibers from the Parabrachial Nucleus. Anatomical studies have shown that the PB projects to the TTA bilaterally
PAlMBR4CHIO-
THALAhJIC
TASTE
RELAY
NEURONS
515
(9, 11, 20), and that three kinds of thalamic atferent fibers originate from the PB: ipsiIateral, contraIateral, and bilateral (20). Our study revealed that most axons arising from taste neurons in the PB terminated either ipsiIateraUy or bilaterally in the thalamus. Double labeling of PB neurons with dyes injected into the TTA revealed that about 60% of the ipsilaterally projecting neurons were double-labeled, whereas about 40% of the contralaterally projecting neurons were double-labeled (20). In our experiment, 14 of the 26 (54%) P-T neurons activated antidromically from the ipsilateral TTA also were excited from the contralateral ‘ITA, and 14 of the 18 (78%) activated antidromically from the contralateral TTA were driven from the ipsilateral side as well. A somewhat smaller ratio of the C-type neurons was found in our study than expected from the anatomic data. Most axons of PB neurons projecting to the contralateral TTA pass close to the ipsilateral TTA (9). Therefore, it is possible that some of C-type neurons were identified as B type as a result of activation of fibers of passage near the ipsilateral TTA. Since the ipsi- and contralateral TTA are separated by at least 2 mm, it seems unlikely that an electrode in one TTA would stimulate axons in the opposite TTA even with the maximal strength current (300 PA) used in this study (see Methods). Fibers projecting to the forebrain or amygdala pass posterior and ventral to the ipsilateral TTA. Electrophysiological experiments (7-9) indicate that more than one-half the P-T taste neurons send collaterals to the amygdala or forebrain. A recent anatomical study (20), however, could not identify PB neurons that send axon collaterals to both the TTA and amygdala. In our study, it is possible that some of the I-type neurons sent axons to the amygdala, but were misidentified as I-type by activation of fibers of passage. This question remains to be answered. In our study, the three different types of relay neurons could be differentiated in other aspects. Many B-type neurons were located in the caudal and ventral PB, many I-type neurons in the rostral and dorsal PB, and C-type neurons in the rostral and ventral PB. The differential distribution of these P-T neurons was similar to the distribution identified anatomically (20). In addition, in response to ipsilateral thalamic stimulation, antidromic late&es of B-type neurons were sign&antly shorter than those of I-type neurons. In the cat (6), PB neurons projecting to the thalamus or amygdala also are topographically distributed. Parabrachial neurons projecting to the amygdala are larger than those to the thalamus. D@rential Taste SpeciJicity of Three DlQ$erent ReIay Neurons. The B-type neurons produced significantly larger responses to NaCl, and many were NaCl-best. On the other hand, some of the I-type neurons responded well to HCl or quinine, and many of these were HCl-best. The NaCl-best and HCl-best P-T neurons were more differentiable based on their locus in the
516
OGAWA,
HAYAMA,
AND IT0
PB and antidromic latencies than were the same neurons classified as B-type and I-type (Table 1). Neural responses to NaCl were significantly larger in the medial part of the PB than in the lateral part. This suggests that taste quality is represented topographically. The present study clarified that the P-T relay cells were somewhat chemotopically localized in the PB, and that the major groups of P-T cells, those with bilateral or ipsilateral projections, may be of different celI sizes. Neurons in the lateral part of the PB are excited by stimulation of the posterior oral cavity, those in the medial part by the anterior tongue (10). Given the present data, the medial and lateral part of the PB may receive differential inputs and have differential thalamic projections. Topographic localization of each category of the pontine taste neurons, classified according to the best stimulus, was also noticed in the hamster (19). In our study in the rat, topographic arrangement was evident only for the P-T relay neurons. REFERENCES 1. BISHOP, P. O., W. BURKE, AND R. DAVIS. 1962. The identification of single units in central visual pathways. J. Physiol. (London) 162: 409-431. 2. FIFKOVA, E., AND J. MARSALA. 1967. Stereotaxic atlases for the cat, rabbit and rat. Pages 653-695 in J. BURES, M. PETRAN, AND J. ZADER, Eds., Electrophysio~ogicai Methods in Biological Research. Academic Press, New York. 3. PISHMAN, I. Y. 1957. Single fiber gustatory impulses in rat and hamster. J. Cell Comp. Physiol. 49: 3 19-334. 4. PRANK, M. 1973. An analysis of hamster atferent taste nerve response functions. J. Gen. Physiol.
61: 588-618.
5. LASITER, P. S., AND D. L. GLANZMAN, 1983. Axon collaterals of pontine taste area neurons project to the posterior ventromedial thalamic nucleus and to the gustatory neocortex. Brain
Res. 258: 299-304.
6. NOMURA, S., N. MIZUNO, K. ITOH, T. SUGIMOTO, AND Y. NAKAMURA. 1979. Localization of parabrachial neurons projecting to the thalamus and the amygdala in the cat using horseradish peroxidase. Exp. Neural. 64: 375-385. 7. NORGREN, R. 1974. Gustatory alferents to ventral forebrain. Brain Res. 81: 285-295. 8. NORGREN, R. 1976. Taste pathways to hypothalamus and amygdala. J. Comp. Neural. 166: 3 l-48. 9.
NORGREN, R., AND C. M. LEONARD. 1974. Ascending central gustatory pathways. J. Cornp. Neurol.
150: 217-238.
10. NORGREN, R., AND C. PFAFFMANN. 1975. The pontine taste area in the rat. Brain Res. 91: 99-l 17. 11. OGAWA, H., AND T. AKAGI. 1978. Afferent connections to the posteromedial ventral nucleus from the pons and the rostra1 medulla in the rat. Kumamoto Med. J. 31: 54-62. 12. 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. 13. OCAWA, H., T. IMOTO, AND T. HAYAMA. 1980. Taste relay neurons in the sohtaty tract nucleus of rats. Neurosci. Lett. 18: 295-299. 14. OGAWA, H., AND J. KAISAKU. 1982. Physiological characteristics of the solitario-parabrachial relay neurons with tongue afferent inputs in rats. Exp. Brain Res. 48: 362-368.
PARABRACHIO-THALAMIC
TASTE
RELAY
NEURONS
517
15. OOAWA, H., M. SATO, AND S. YAMANITA. 1968. Multiple sensitivity of chorda tympani fibres of the rat and hamster to gustatory and thermal stimuli. J. Physiol. (London) 199: 233-240. 16. @JAWA, H., M. SATO, AND S. Y AMAsHIT& 1973. variability in impulse discharges in rat chorda tympani fibers in response to repeated gustatory stimulations. Physic. Behav. 11: 469479. 17. P~~~crrro, R. S., AND T. R. !kwrr. 1976. Gustatoti neural coding in the pans. Brain Rex 110: 283-300.
18. SAPER, C. B., AND A. D. LCWEY. 1982. Werent connections of the pambrachial nucleus in the rat. Brain Res. 197: 291-317. 19. VAN BUSKIRK, R. L., AND D. V. SMITH. 1981. Taste sensitivity of hamster parabrachial pontine neurons. J. Neurophysid 45: 144-17 1. 20. WISHART, K., AND D. VAN DER Kooy. 1981. The organization of the e&rent projections of the parabrachial nucleus to the forebrain in the rat: a retrograde fluorescent doublelabelling study. Brain Res. 212: 27 l-286. 2 1. YAMADA, K. 1966. Gustatory and thermal responses in the glossopharyn8eal nerve of the rat. Jap. J. Physiol. 16: 599-611.