Multimodal responses of taste neurons in the frog nucleus tractus solitarius

0361-9230/87 $3.00 + .OO

Brain Research Bulletin, Vol. 18, PP. 87-97, 1987.B Pergamon Journals Ltd., 1987.Printed in the U.S.A.

Multimodal Responses of Taste Neurons in the Frog Nucleus Tractus Solitarius TAKAMITSU Department

of Physiology,

HANAMORI,’

NOBUSADA

Miyazaki Medical College,

ISHIKO

Miyazaki, 889-16, Japan

AND DAVID

V. SMITH

Department of Otolaryngology and Maxillofacial Surgery University of Cincinnati College of Medicine, Cincinnati, OH 45267 Received

27 August 1986

HANAMORI, T., N. ISHIKO AND D. V. SMITH. Multimodal responses of taste neurons in the frog nucleus tractus BRAIN RES BULL 18(l) 87-97, 1987.-The responses of 216 neurons in the nucleus tractus solitarius (NTS) of the American bullfrog were recorded following taste, temperature, and tactile stimulation. Cells were classified on the basis of their responses to 5 taste stimuli: 0.5 M NaCI, 0.0005 M quinine-HC1 (QHCI), 0.01 M acetic acid, 0.5 M sucrose, and deionized water (water). Neurons showing excitatory responses to 1, 2, 3, or 4 of the 5 kinds of taste stimuli were named Type I, II, III, or IV, respectively. Cells whose spontaneous rate was inhibited by taste and/or tactile stimulation of the tongue were termed Type V. Type VI neurons were excited by tactile stimulation alone. Of the 216 cells, 115 were excited or inhibited by taste stimuli (Types I-V), with 35 being Type I, 34 Type II, 40 Type III, 2 Type IV and 4 Type V. The remaining 101 cells were responsive only to tactile stimulation (Type VI). Of those 111 cells excited by taste stimulation (Types I-IV), 106 (95%) responded to NaCl, 66 (5%) to acetic acid, 44 (4%) to QHCI, 10 (9%) to water, and 9 (8%) to warming. No cells responded to sucrose. Of the 111 cells of Types I-IV, 76 (6%) were also sensitive to mechanical stimulation of the tongue. There was some differential distribution of these neuron types within the NTS, with more narrowly tuned cells (Type I) being located more dorsally in the nucleus than the more broadly tuned (Type III) neurons. Cells responding exclusively to touch (Type VI) were also more dorsally situated than those responding to two or more taste stimuli (Types II and III).

solitarius.

Glossopharyngeal

nerve

Nucleus tractus solitarius

Taste

IN mammals, gustatory information from the tongue is conveyed by the chorda tympani (CT) branch of the facial (VIIth) nerve and the glossopharyngeal (IXth) nerve, whereas somatosensory information is conveyed by the lingual branch of the trigeminal (Vth) nerve and IXth nerve. The CT nerve innervates the taste buds of the fungiform papillae on the anterior two-thirds of the tongue and the IXth nerve innervates those of the vallate and foliate papillae on its posterior third. Afferent fibers of these two nerves terminate in the rostra1 part of the nucleus tractus solitarius (NTS) in an overlapping rostral to caudal projection [2, 7, 8, 34, 371. In frogs, on the other hand, only the IXth nerve carries afferent information from the tongue, both gustatory and somatosensory [9, 20, 301. The frog’s IXth nerve divides into medial and lateral branches shortly before it reaches the tongue, with the medial branch supplying the caudal twothirds and the lateral branch the rostra1 third of the tongue

Multimodal responses

Bullfrog

Gustation

[13]. Afferent fibers of the frog’s IXth nerve project into the NTS, as in mammals Ill, 12, 351. Thus, even though frogs are the lowest vertebrates to have a true tongue, the neural organization of taste input is somewhat different from that seen in mammals. A number of investigators have examined the response of the frog’s IXth nerve to gustatory stimulation of the tongue. Integrated responses of the IXth nerve show that frogs are sensitive to a wide range of stimuli [13, 24, 25, 30, 39, 401. The IXth nerve of the bullfrog (Ram caresbeiuna) responds well to salts, acids and quinine and is somewhat sensitive to sugars [13,25], although the sugar sensitivity is suppressed by the addition of salts to the solution [24]. In addition to its responsiveness to gustatory stimuli, the frog’s IXth nerve also responds well to mechanical [9, 20, 301 and thermal [IS] stimuli. The responses of single fibers from the IXth nerve of the

‘Requests for reprints should be addressed to Dr. Takamitsu Hanamori, Department of Otolaryngology and Maxillofacial Surgery, University of Cincinnati, College of Medicine, 231 Bethesda Avenue, ML 528, Cincinnati, OH 45267.

87

88

HANAMORI

Japanese common frog (Rana nigromaculata) were studied by Kusano [19] and by Kimura [14]. Most fibers in the IXth nerve were responsive to a broad range of chemicals [ 14,191. Of the 105 fibers recorded by Kusano [19], 94 responded to more than one kind of taste stimulus. Thus, peripheral taste fibers in frogs are typically broadly responsive across a range of gustatory stimuli. On the other hand, the responsiveness of peripheral taste fibers to tactile stimulation is not as well documented. In an investigation of mechanosensitive fibers in the frog’s IXth nerve, Yamane [38] demonstrated that 19 of 20 touch fibers also responded to at least one of several gustatory stimuli. In a study of water-sensitive fibers of the frog’s IXth nerve, Andersson and Zotterman [l] found no mechanical sensitivity. Thus, even though many touch fibers may respond to taste [38], the degree of mechanical sensitivity in individual fibers is not clear. The responsiveness of the NTS to gustatory stimulation has been studied in mammals by many investigators [4,6,21, 26, 29, 361. In hamsters, cells in the medulla are somewhat more broadly responsive to gustatory stimuli than CT fibers [33,36]. Gustatory-responsive neurons in the rat NTS have been shown to respond also to tactile and thermal stimuli [21,29]. In contrast to the relatively extensive work on mammals, there have been few investigations of the medullary gustatory relay in frogs. Kumai [18] studied tasteevoked multiunit activity of the NTS of Japanese common frogs (Rana nigromaculata) and found that quinine produced the greatest response, followed by NaCl, HCl and CaCl,. Hanamori and Ishiko [lo] have recorded negative field potentials in the medulla of the American bullfrog (Runa catcsbeiana) evoked by electrical stimulation of the IXth nerve. However, the response characteristics of single NTS neurons in frogs to gustatory or mechanical stimulation of the tongue have never been investigated. In the present study, we recorded the responses of individual NTS cells in bullfrogs to gustatory, thermal and mechanical stimulation of the tongue.

METHOD

Sixty American bullfrogs (Rana catcgsbeiana), weighing between 250 and 500 g, were used. The frogs were immobilized with intraperitoneal injection of suxamethonium chloride (50 mg/kg). A craniotomy exposed the dorsal surface of the brain from the obex to the optic lobes. The posterior choroid plexus and the dura overlying the exposed area were carefully removed. In some experiments, a portion of the IXth nerve near the tympanic membrane was exposed for recording. Throughout the experiment the frog was held rigidly by clamping the upper jaw and was placed in a shallow pool of water to prevent drying of the skin. To allow stimulation of the tongue, the frog’s mouth was opened and a small acrylic platform was placed inside the mouth between the upper and lower jaws and behind the tongue, which was fixed to the platform with pins. Extracellular unit responses to gustatory, thermal and mechanical stimulation of the tongue were recorded from the NTS with glass micropipettes filled with either 2 M NaCl or 2 M NaCl containing Fast Green FCF. Electrode impedances ranged from 1 to 10 Ma. A silver wire reference electrode was placed into the muscle near the dorsal surface of the cervical vertebrae. Neural activity was amplified (WPI Model M-701), displayed on an oscilloscope and audio monitor, and stored

ET AL.

on magnetic tape. The taped data were later displayed on an oscilloscope and photographed with a kymograph camera. The number of spikes elicited by the various stimuli was determined by counting the impulses on these photographic records. The activity of the IXth nerve was recorded by a pair of silver wire electrodes, amplified by conventional means, integrated electronically, and displayed on an inkwriter. Impulse discharges from NTS cells were also integrated and monitored on an inkwriter during the course of the experiment. The taste stimuli were: 0.5 M NaCl, 0.0005 M quinineHCl (QHCI), 0.01 M acetic acid, 0.5 M sucrose and deionized water (water). All of the stimuli except water and 0.5 M NaCl were dissolved in 0.01 M NaCl, which was also used as a rinse between stimuli, since the frog’s tongue is known to respond to water 11, 19, 401 and a rinse of 0.01 M NaCl produces a minimal response to distilled water [39]. All stimuli were delivered at room temperature (about 20°C). The concentrations of these stimuli are those that produce midrange integrated responses in the frog’s IXth nerve [25.32]. For thermal stimulation, 0.01 M NaCl at 30°C was applied after adaptation of the tongue to 0.01 M NaCl at 20°C. Stimuli were dehvered to the tongue from an overhead funnel via gravity flow at a rate of 6 ml/set and were allowed to flow for 5 sec. Between taste stimuli, 0.01 M NaCl flowed for 1 min, the tongue sat in 0.01 M NaCl for 3 min, and the flow resumed for 1 min just prior to the next stimulus, producing an interstimulus interval of 5 min. The taste and thermal stimuli were delivered following a flowing rinse in order to minimize the tactile response to solution flow. The tongue was mechanically stimulated with a small brush and the receptive field of the neurons responding to mechanical stimulation was determined. Since some neurons had a wide tactile receptive tield, mechanical stimulation was delivered not only to the tongue but also to other areas of the frog’s body. The following procedure was used to quantify the locations of taste-sensitive neurons in the NTS. For 17 cells responding to gustatory stimulation, the recording site was histologically determined by the iontophoretic injection of Fast Green FCF. Frogs were perfused with 10% formalin and the brains removed, cut in 50 pm sections and stained with cresyl violet. The histologically determined recording sites were represented in a 3-dimensional coordinate system, in which X indicates the lateral distance from the midline, Y the rostra1 distance from the obex, and Z the depth from the dorsal surface of the medulla (see Fig. 9). The mean values (tS.E.) of X, Y and Z (n=17) were 0.89t0.05, 1.53t0.07 and 0.78a0.07 mm, respectively. For the same 17 cells. the recording sites were also measured from the scale on the micromanipulator. The mean values (%S.E.) obtained by this method were 0.81 kO.03, 1.68?0.08 and 0.82tO.OS mm for X, Y and Z, respectively (n=17). Since the distances obtained by these two mehtods were not significantly different (two-tailed r-test, p>O.O5), the locations of the remaining neurons were determined using the micromanipulator scale. Of all the neurons isolated in this experiment, 70% had no spontaneous discharge. In addition, among the 30% showing a spontaneous discharge, about half of them produced a bursting discharge at irregular intervals of more than 30 set with no impulses between the bursts. Thus, for a large majority of the cells the occurrence of a response to stimulation was easy to determine. For the remaining neurons (about 15% of the cells), which had spontaneous discharge rates of 0.4-6.0 Hz, a response criterion was developed. A

RESPONSES

89

OF FROG NTS NEURONS TABLE

1

RESPONSE TYPES OF NTS CELLS Stimuli Type

NQAWST

I

+ + 0 0 0 0

0 0 + +

+ + + + + +

+ +

II

0

0 0

0 + 0 + + +

0 0 0 0 0 +

35

0 0 + +

0 0 0 0 + +

0 +

0 0 0 0 0 0 0

34

0 0 0 0 0 0

0 +

0 0

0 0 0 0 + +

0

+

+

+

0

+

+ _ I 0

0 I 0 0

0

0

IV

+

V

I _

0 -

+

0

0

0

VI

0 0 0 0 0 0

0

+ + + +

Total

0 0 0 0 0 0

+ + + + + +

0 0 0 0 +

No. of Cells

0 0 0 0 + +

0 0 0 0 0 0 0

+ + + + + +

III

Warm

0 0 +

Total

0 + 0 + +

0 0 + +

40

0 +

1 29 2 5 2 1

+

0

2

2

0 0 I 0

0 _

1 1 1 1

4

0 0

0 I I -

0

+

0

101

101

0 + +

216

+, - and 0: neuronal activity was excited, inhibited and unchanged, respectively. I: stimulus was not applied. N: 0.5 M NaCl, Q: 0.5 mM quinine hydrochloride, A: 0.01 M acetic acid, W: water, S: 0.5 M sucrose, T: touch, Warm: 0.01 M NaCl at 30°C.

was termed excitatory or inhibitory if the rate of discharge over the 2.5set poststimulus interval increased or decreased by more than 30% of the immediately preceding 2.5 set of spontaneous rate. If impulse frequency did not vary by more than 30% during stimulation, it was not considered to be a response. Since the waveforms of the action potentials recorded from all cells were diphasic or triphasic and many displayed the typical a-b inflection, they were considered to arise from cell bodies rather than from entering first-order fibers. response

RESULTS

Response

Characteristics

of NTS Neurons

Responses to the entire array of stimuli were recorded from 216 neurons. These cells were then classified into 6 types on the basis of their response patterns, as shown in Table 1. Neurons responding to 1, 2,3 and 4 of the 5 kinds of taste stimuli (NaCI, QHCl, acetic acid, and water; none responded to sucrose) were termed Type I, II, III and IV,

respectively. In this classification scheme the particular gustatory sensitivity of a cell was not important (e.g., whether it was NaCl-best or QHCl-best), but how broadly responsive the cell was to the taste stimuli was the critical variable [23]. If the cells were not excited by any of these stimuli but had their ongoing spontaneous discharge inhibited by taste and/or tactile stimulation, they were termed Type V. Neurons responsive only to tactile stimulation of the tongue were named Type VI. The distributions of the sensitivities among these 6 neuron types are shown in Table 1. There were 35 neurons of Type I, most of which responded exclusively to NaCl or to NaCl and tactile stimulation (see Table 1). The responses of a Type I cell are shown in Fig. 1. This cell responded vigorously to 0.5 M NaCl and considerably less to tactile stimulation of the rostra1 portion of the tongue. Of 35 Type I neurons, 31 (8%) responded to NaCl, 2 (6%) to QHCI, and 2 (6%) to acetic acid. Sensitivity to tactile stimulation was found in 17 (4%) of the cells and was associated with sensitivity to any one of the effective taste stimuli. One (4%) of the Type I neurons also responded to warming the tongue (30°C 0.01 M NaCl).

90

HANAMORI

05M NoCl

L

OSmM cwwlc

0 OIM

ACdlC

octd

Water

I

I

Worming

PSCC

FIG. 1. Responses of a Type I cell to chemical, tactile and thermal stimulation. Stimulus onset and duration are indicated by the horizontal line below each oscillographic record.

Of the 34 Type II cells, 21 (62%) responded to NaCl and acetic acid, 7 (2%) responded to NaCl and QHCl, 5 (14%) responded to NaCl and water, and 1(3%) responded to QHCl and acetic acid (see Table 1). As in the Type I cells, most Type II neurons responded to NaCl (33/34=97%). A response occurred to acetic acid in 22 (65%), to QHCl in 8 (24%), and to water in 5 (15%) of these cells. Tactile responses occurred in 21 (62%) of the Type II cells and were associated with responses to any pair of taste stimuli. None of the Type II neurons responded to warming the tongue. Forty neurons were classified as Type III, responding to three of the taste stimuli. Of these cells, 37 (93%) were sensitive to NaCl, QHCl and acetic acid and 3 (8%) to NaCl, acetic acid and water (see Table 1). All of the Type III cells (100%) responded to NaCl, 37 (93%) of them responded to QHCl, and all of them (100%) responded to acetic acid. Three (8%) of these neurons responded to water, always in conjunction with responses to NaCl and acetic acid. Eight (20%) of the Type III cells responded to warming and 36 (90%) to touching the tongue. Sensitivity to warming and to tactile stimulation occurred in conjunction with that to any combination of gustatory sensitivities. The responses of a Type III neuron to NaCl, QHCl, acetic acid and tactile stimulation of the tongue and ipsilateral forelimb are shown in Fig. 2. Of the 36 Type III neurons that responded to touch, 10 responded to tactile stimulation of the skin on other parts of the body, such as the forelimb or hindlimb, as well as to bilateral stimulation of the tongue. The tactile receptive fields of Type I and Type II cells, on the other hand, were always limited to the ipsilateral tongue. Two of the cells in the NTS responded to four taste stimuli (NaCl, QHCl, acetic acid and water) and were classified as Type IV (see Table 1). In addition to their broad gustatory sensitivities, these cells also responded to tactile stimulation of

E7‘ AL.

the ipsilateral tongue. Type V cells were those that were inhibited by taste, tactile or thermal stimulation of the tongue. Of the four cells in this category, one was inhibited by NaCl alone, one by acetic acid and touch, one by NaCl and QHCl and one by warming the tongue. Unfortunately, two of these cells were not tested by the entire array of stimuli. The responses of one of the cells that was completely tested are shown in Fig. 3, where it may be seen that NaCl completely suppressed the ongoing spontaneous activity and QHCl and acetic acid produced increased discharge rates. The responses during water and thermal stimulation were complex and were complicated by the bursting nature of the spontaneous discharge. The responses of the 111 neurons showing excitatory responses to one or more gustatory stimuli (Types I-IV) are shown in Fig. 4. In this figure, the cells are arranged along the abscissa in decreasing order of their responses to NaCl. The broad sensitivity of these cells across taste quality and sensory modality is evident from the figure. Of these II I neurons, 106 (95%) responded to NaCl, 49 (44%) to QHCl, 66 (5%) to acetic acid, 10 (%) to water, 0 to sucrose (not shown), 76 (68%) to touch, and 9 (8%) to warming. Most of the responses to quinine and to warming occurred in Type III cells, whereas the responses to touch were distributed across all four cell types. The distributions of responses to the various stimuli across the four cell types are shown as average profiles in Fig. 5. The responses to NaCl and touch are relatively evenly distributed across the four cell types. Type III cells. on the other hand, are much more responsive to acetic acid, quinine and warming than the other cell types. Water is a relatively more effective stimulus for Type 1V cells than for any of the others. Thus, the sensitivity of these cells to tactile stimulation is unrelated to their breadth of responsiveness to chemical stimuli. Pearson correlation coefficients (r) were computed between all possible pairs of stimuli across the 111 cells ofTypes I-IV. These correlations are depicted as profiles for each stimulus in Fig. 6. In this figure, each stimulus is assumed to correlate perfectly (r= + 1.O) with itself. Statistically significant correlations are indicated by asterisks (see caption). The responses to NaCl correlated significantly with those to acetic acid (+0.35), water (+0.25) and touch (+0.25). Quinine responses correlated with those to acetic acid (+0.36) and warming (+0.57). The responses to acetic acid also correlated with those to touch (+0.27) and warming (+0.40). With n=lll. r>+0.35,p<0.001 and r>+0.24,p
Of the five chemical stimuli employed to test the responof these cells, four were effective stimuli for at least some of the neurons. Sucrose was completely ineffective as a stimulus for these medullary cells. The cells were divided into types on the basis of their responses to NaCl, acetic siveness

acid,

QHCl

and water.

The breadth

of each cell’s

respon-

siveness to these four stimuli was calculated using the measure (H) developed by Smith and Travers [331. Breadth of responsiveness (H) is given by the equation H(P,,

.I-‘,,)

=

-K

;

<.

/IJ,IO~,,,, _.

:_I

where H=breadth of responsiveness, K is a scaling constant, and p, is the proportional response to each of n stimuli. The pi for each cell are derived by converting the neural responses to the four stimuli (NaCl, acetic acid, QHCl and

RESPONSES

91

OF FROG NTS NEURONS 0.5M NaCl

0.5mM Quinine

0.5M Sucrose

-

Touch

O.OlM Acetic acid

Water

Touch &,rei~mb)

FIG. 2. Responses of a Type III cell to chemical, tactile and thermal stimulation. Stimulus onset and duration are indicated by the horizontal line below each oscillographic record.

0 5mM

QUlWlC

OOlM Acetic ood

Water

O.SM

sucrose

Touch

Warmmg

-

zscc FIG. 3. Responses of a Type V cell to chemical, tactile and thermal stimulation. Stimulus onset and duration are indicated by the horizontal line below each oscillographic record.

water)

to proportional responses, the response to each stimulus being expressed as a proportion of the total response to all four. If the logarithms ofpi are taken to the base

of 10 and K=1.661, then H=l.O when the responses to all four stimuli are equal @i= l/n =0.25), and H=O.O when there is a response to only one of the stimuli @i= I .O, 0, 0, and 0). Thus, this measure reflects the breadth of tuning of a cell across the stimuli employed (see [33] for more detail). For the 111 cells responding to gustatory stimulation, the breadth of responsiveness (H) was calculated and the distributions of this measure within each of the cell types (I-IV) are shown in Fig. 7. Breadth of responsiveness ranged from 0.0 (Type I cell) to 0.961 (Type IV cell). All cells of Type I, by definition, were completely specific to one of these stimuli (H=O). Cells

of each type were successively

more broadly tuned, with the mean value of H being 0.418 for Type II cells, 0.709 for Type III cells and 0.919 for Type IV neurons. There was little overlap in the distributions of this measure between the four cell types that were defined on the basis of the number of stimuli to which they responded. Gustato~ cells in the frog NTS ranged from being completely specific to one of these stimuli (Type I) to being extremely broadly responsive across stimuli (Types III and IV). The mean breadth of responsiveness for all cells (Types I-IV) was 0.400. Spontaneous

Discharge

Of the 111 neurons responding to gustatory stimulation (Types I-IV), 61 (55%) were spontaneousty active. The re-

92

HANAMORI 150

E7 AL

N&l

‘00 1I i

NaCl 1

100

j

Touch

IOO-

Touch

1

Warming

WarmIng

I

1

5

15 Type

2s

I ceils

35

5

15 Type

Warmmg

25 II cells

FIG. 4. Responses (impulses/t.5 set) of 111 NTS cells (Types I-IV) to gustatory, tactile and thermal stimulation of the tongue. Cells are divided into types and ranked along the abscissa within each type according to the magnitude of their response to NaCI. No cells responded to sucrose, which is not shown.

maining JO neurons (45%) showed negligible spontaneous activity (less than 0.4 Hz, i.e., less than 1 impulse/2.5 set). Of the 61 spontaneously active cells, 32 showed a bursting discharge at irregular intervals, with few spikes occurring between bursts. The remaining 29 cells showed a relatively constant rate of spontaneous discharge, ranging from 0.4 to 6 Hz (mean=2.72 Hz). The four Type V cells had a spontaneous discharge rate between 3 and 11 Hz (mean=5.4 Hz). None of the Type VI neurons, which responded only to mechanical stimulation, were spontaneously active. Concentration-Response

Function

The relationship between the rate of impulse discharge of six cells in the NTS and the concentration (M) of NaCl is shown in Fig. 8A (filled circles). The responses of one cell to five concentrations of NaCl are shown in the oscillographic records in Fig. 8B. The concentration-response function of NTS cells (filled circles) is almost the same as that of the integrated IXth nerve response (open circles). Each function is expressed relative to the response to 1.0 M NaCl. The values in the figure are based on data from six frogs, one IXth nerve from each of six animals and one NTS cell from each of six additional frogs. The function for NTS neurons

was obtained by averaging the responses of three different cell types (one Type I, one Type II, and four Type III neurons). Localizution

of NTS Cells

The positions of the 216 cells responding to natural stimulation of the tongue were calculated relative to obex, midline, and the dorsal surface of the medulla (see the Method section). The positions of Type I (open circles) and Type 111 (tilled circles) cells in the X-Y (A) and X-Z (B) coordinates are shown in Fig. 9. In this coordinate scheme, X and Y are medial-lateral and rostral-caudal dimensions, respectively, and Z is the dorsal-ventral dimension (see diagram, Fig. 9). As may be seen in Fig. 9, cells of Types I and III were distributed across the medulla between 0.4 and 1.2 mm lateral to the midline(X), between 0.5 and 2.6 mm rostral to the obex (Y), and between 0.3 and 1.3 mm ventral to the dorsal surface of the medulla (Z). Type III cells (filled circles, Fig. 9B) were distributed more ventrally within the medulla than those of Type I (open circles) (two-tailed t-test, p
RF$PONSES

93

OF FROG NTS NEURONS

30

20

g s

1

Type

Type 1

L

10

Q

0 !I-_L

? ? 2 Gj

1

Type III (n=40)

30

20

0 IILL NAOWS

FIG. 5. of

10

\

-05 L

F

0

.-01 ”

S=sucrose,

. . f. . L NoCl

z

8-o4

s .-

T Warm

Mean responses seven stimuli.

W=water,

10

(n=2)

!LliLL

10

each

Type IV

-L

E IX

II

(ne34)

(n&3

N

A Q W

S

T Warm

(impulsed2.5

set) of cell Types I-IV to N=NaCI, A=acetic acid, Q=QHCI, T=touch, Warm=warming.

Quinine

9

,.+

4..

Warming

Water

w f

z

a.*

F 35 b v 0

.

i-l t/

-04

N

0

A

W

T

Warm

L..‘.” N

0

A

W

T

Warm

N

Q

A

.

W

T

Warm

Stimulus FIG. 6. Correlation profiles among six stimuli. Each profile shows the magnitude of the Pearson correlation coefftcient (r) between the responses to one stimulus and each of the others across the 1I 1cells of Types I-IV. Each stimulus is assumed to correlate perfectly with itself (r= + 1.O). ***~
94

HANAMORI ET At,.

Breadth of chemical responsiveness

0-i)

FIG. 7. Distributions of the breadth of responsiveness (H) of cells in each of 4 types (Types I-IV). H was calculated from the responses to four chemical stimuli (NaCl, acetic acid, QHCI and water).

0 0 0625M

0 125M

1 OM NoCl ,

01

10

Concentration ( M) FIG. 8. Concentration-response functions of 6 NTS neurons (A, filled circles) and 6 integrated IXth nerve preparations (A, open circles) to NaCl stimulation. Responses are expressed in proportion to the response to 1.0 M NaCI. The responses of one NTS cell to five concentrations of NaCl are shown in B.

RESPONSES

OF FROG NTS NEURONS

95

A 2.5

X(mm)

-0

0.5

05

1.0

1.0

X(mm)

FIG. 9. Distribution of NTS cells (Types I and III) in the brainstem shown in three-dimensional coordinates. X and Y are medial-lateral and rostra!-caudal dimensions and Z is the do~al-vent~l dimension (see di~m}. Measurements are in mm from the midline (X), obex (Y), and dorsal surface of the brainstem (Z). Type I ceils are depicted by open circles and Type III by tilled circles.

TABLE 2 LOCATIONS

OF EACH CELL TYPE WITHIN THE BRAINSTEM

Recording Site (mean + SE., Type

*o.ol~p
X

t0.001
mm)

Y

No. of Cells

Z

Q?
the mean value of the dorsal-vent~l position between Type I and Type III cells (p
DISCUSSION Breadth

of Gustatory

Responsiveness

Most (761111) of the gustatory-responsive cells recorded from the medulla of the frog were responsive to more than one of the chemicals applied to the tongue (Table 1, Figs. 2 and 4). This broad tuning is consistent with the sensitivities of medullary cetls in mammals [4, 6, 21, 26, 29, 33, 361. However, in the frog, 351111 (32%) of the cells were responsive to only one of the gustatory stimuli, usually NaCl (31/35). Such narrow specificity has not been seen in mammalian NTS cells. For example, there were no cells in the

96

HANAMORI

hamster NTS that responded to only one of four prototypic stimuli [36]. A comparison of the mean breadth of responsiveness of NTS cells in frog (H=0.400) and hamster (H=0.698) [36] demonstrates that medullary cells in the frog are, on the average, more narrowly tuned to an array of gustatory stimuli that differ widely in taste quality. Cells in the frog NTS, however, appear to be more broadly responsive than those in fish. Of 245 cells recorded from the facial lobe of the carp, 110 (45%) responded to only one of four stimuli, usually to NaCl (102/110) [23]. The remaining 13.5 cells were responsive to more than one of the stimuli. Thus, the data that are available suggest that the relative specificity of medullary cells in the frog is intermediate between that of fish and mammals. In a systematic study of single fibers in the glossopharyngeal nerve of the frog, Kusano 1191 reported that only a few (191105, 18%) fibers responded exclusively to one class of gustatory stimuli, either single stimuli or monovalent salts. Most fibers in the frog glossopharyngeal nerve were responsive to more than one type of chemical stimulus [ 191. Similarly, in the glossopharyngeal nerve of the carp, only 17 of 114 fibers (15%) responded to a single gustatory stimulus 1171. In mammals, a similar proportion of peripheral fibers are specifically tuned. In the rat, 9 of 48 chorda tympani fibers (1%) showed sensitivity to only one kind of gustatory stimulus [27], whereas in the hamster 5 of 28 fibers (18%) [27] or 23 of 79 fibers (2%) [5] responded to only one of the prototypical taste stimuli. Most peripheral gustatory fibers are broadly tuned across the stimulus array, regardless of whether the fibers are fish, amphibian, or mammalian. Thus, in fish and amphibians, there appears to be a sharpened tuning in medullary cells, where there are greater proportions of narrowly tuned cells than in the periphery. This sharpening does not occur in mammals, where medullary cells are more broadly tuned than peripheral fibers (see

WI). Multimodal

Sensitivity

Multiunit responses to mechanical stimulation of the tongue have been recorded from the same areas of the frog medulla that respond to gustatory stimulation [18]. The results of the present investigation demonstrate that many individual cells in the frog NTS show multimodal sensitivity. Sixty-eight percent (76/l 11) of the gustatory-responsive cells (Types I-IV) were also responsive to mechanical stimulation of the tongue. A similar proportion of multimodal cells has been reported to exist in the rat NTS, where 27 of 38 cells (71%) were responsive to both taste and touch in one study [29], 13 of 21 (6%) in another [21], and 50 of 78 (64%) in another [26]. In the facial lobe of the carp, 90 percent of the cells responsive to gustatory stimulation were also responsive to tactile stimulation of their receptive fields [23]. Thus, the multimodal sensitivity of medullary cells to both taste and tactile inputs appears to be a general rule in a number of divergent species. It would appear that medullary cells are multimodal in their responsiveness because of converging input from peripheral fibers, but the degree to which peripheral fibers in the frog are sensitive to both taste and touch is not clear. Many investigations of single fibers in the frog glossopharyngeal nerve have not tested for multimodal sensitivity [14, 15, 191. From the results of other studies, it is clear that some taste fibers do not respond to mechanical stimulation [1,30]. In an investigation of tactile fibers in the frog’s IXth nerve, however, Yamane [38] found that 19 of 20

E7- A i..

touch-sensitive fibers also responded to chemical stimulation of the tongue. A study of the conduction velocities of IXth nerve fibers suggests that rapidly adapting touch fibers are of a larger diameter than those responding to gustatory stimuli [9], although slowly adapting touch (i.e., pressure sensitive) fibers fell within the same range. Thus, data on the multimodal sensitivity of frog glossopharyngeal fibers are not at all conclusive. Studies on other species do not help to clarify this situation. For example, in a study of glossopharyngeal fibers in the mudpuppy, all taste fibers were shown to be responsive to tactile stimulation [31]. On the other hand, IXth nerve fibers in the carp [ 171and palatine nerve fibers in the puffer [16] are sensitive to either taste or touch, but never to both. More parametric data from fibers in the frog’s 1Xth nerve are necessary to establish the degree to which the multimodal sensitivity of medullary cells is due to convergence. In contrast to the correspondence between frog and mammalian medullary cells in their joint sensitivities to taste and touch, there is a large difference in their mutual sensitivities to taste and temperature. Of the 111 taste-sensitive cells in the frog NTS, only 9 (8%) responded to warming of the tongue (Table 1, Fig. 4). In the rat, on the other hand, 68 percent of medullary taste cells responded also to warming and 93 percent responded to cooling [29]. This pattern of sensitivity of the frog’s medullary taste cells to thermal stimulation is consistent with what is known about the temperature sensitivity of its glossopharyngeal nerve. In a multiunit study of the frog IXth nerve, Dodt [3] demonstrated a vigorous response to warming but no response to cooling of the tongue. There are no glossopharyngeal fibers in the frog that respond to cooling, but some smaller diameter fibers respond to warming [15]. Fibers sensitive to chemicals were shown to respond to temperatures higher than 30°C [IS], which was the temperature employed in the present study. In general, the temperature sensitivity of amphibian skin is predominantly a response to warming. Unmyelinated fibers in cutaneous nerves of the toad respond to heat and cold in addition to nociceptive stimuli, but much more to warming than to cooling [22]. Thus, the relative lack of thermal sensitivity in the frog medulla in comparison to that of mammals appears to reflect differences already apparent in the peripheral nervous system. Anatomical

Organization

The taste-responsive cells recorded in the present study were distributed throughout the ipsilateral medulla from 0.5 to 2.6 mm rostra1 to the obex (mean= 1.5 mm) and 0.3 to 1.3 mm lateral to the midline (mean=0.7 mm). This distribution corresponds almost exactly with the location of the maximal amplitude of the negative field potential evoked by electrical stimulation of the glossopharyngeal nerve in this species (I .5 mm rostral, 0.5 mm lateral) [lo]. This area also overlaps with the major region of termination of IXth nerve afferent fibers, as shown by application of HRP to the frog glossopharyngeal nerve [11,35]. Although the nucleus of the solitary tract is not anatomically well-defined in frogs [28], the 17 cells that were histologically verified in the present study were all found in close proximity to the solitary tract. An examination of the distributions of the various cell types (I-VI) within the medulla revealed some significant differences in their locations. Cells that responded to only a single gustatory stimulus (Type I) were located dorsal to those responding to three taste stimuli (Type III), as may be seen in Fig. 9B and in Table 2. Neurons exclusively respon-

RESPONSES

OF FROG NTS NEURONS

97

sive to touch were also found dorsal to the more broadly tuned (Types II and III) taste neurons (Table 2). Thus, cells that were narrowly tuned, either to taste or to touch, were found primarily dorsal to those with broader sensitivity. There were no differences among the various cell types in

their rostral-caudal or medial-lateral distributions. The more dorsal location of tactile cells in the frog is in marked contrast to the situation in rats, where evoked multiunit activity from the medulla demonstrates that tactile sensitivity is ventral to taste and temperature sensitivity [211.

REFERENCES 1. Andersson, B. and Y. Zotterman. The water taste in the frog. Acta Physiol Stand 20: 95-100, 1950. 2. Blomquist, A. J. and A. Antem. Localization of the terminals of the tongue afferents in the nucleus of the solitary tract. J Camp Neural 124: 127-130, 1965. 3. Dodt, E. Schmerzimpulse bei Temperaturreizen. Acta Physiol &and

31: 83-%,

1954.

4. Doetsch, G. S. and R. P. Erickson. Synaptic processing of taste-quality information in the nucleus tractus solitarius of the rat. J Neurophysiol 33: 490-507, 1970. 5. Frank, M. An analysis of hamster afferent taste nerve response functions. J Gen Physiol 61: 588-618, 1973. 6. Ganchrow, J. R. and R. P. Erickson. Neural correlates of gustatory intensity and quality. J Neurophysiol 33: 768-783, 1970. 7. Halpem, B. P. and L. M. Nelson. Bulbar gustatory responses to anterior and to posterior tongue stimulation in the rat. Am J Physic>/ 209: 105-110, 1965. 8. Hamilton, R. B. and R. Norgren. Central projections of gUStatory nerves in the rat. J Comp Neurol 222: 560-577, 1984. 9, Hanamori, T. and N. Ishiko. Conduction velocity of the 1Xth nerve fibers innervating taste organs in the rostra1 and caudal tongue region in bullfrog. Chem Senses 6: 175-187, 1981. 10. Hanamori, T. and N. Ishiko. Surface and intramedullary potentials evoked by stimulation of the glossopharyngeal nerve in frogs. Brain Res 260: 51-60, 1983. 11. Hanamori, T. and N. Ishiko. Intraganglionic distribution of the primary afferent neurons in the frog glossopharyngeal nerve and its transganglionic projection to the rhombencephalon studied by HRP method. Brain Res 260: 191-199, 1983. 12. Herrick, C. J. The fasciculus solitarius and its connections in amphibians and fishes. J Comp Neural 81: 307-331, 1944. 13. Ishiko, N., T. Hanamori and N. Murayama. Frog’s tongue receptive areas: Neural organization and gustatory function. Expc&lntia

35: 773-774, 1979.

14. Kimura, K. Adaptation of chemoreceptors of frog as analyzed by activity in single gustatory fibers. Kumamoto Mud J 14: 149-156, 1961. 15. Kimura, K. Effects of temperature on the response of chemoreceptors in frog tongue. Kumamoto Med J 15: 73-82, 1962. 16. Kiyohara, S., 1. Hidaka, J. Kitoh and S. Yamashita. Mechanical sensitivity of the facial nerve fibers innervating the anterior palate of the puffer, Fu#u paradalis. and their central projection to the primary taste center. J Camp Physiol [A] 157: 705-716, 1985. 17. Konishi, J. and Y. Zotterman. Taste function in the carp: An electrophysiological study on gustatory fibers. Acta Physid Stand 52: 150-161,

1961.

18. Kumai, T. Gustatory responses in the medulla of the frog. Brain Res 196: 508-512, 1980.

19. Kusano, K. Analysis of the single unit activity of gustatory receptors in the frog tongue. Jpn J Physic)/ 10: 620-633, 1960. 20. Kusano, K. and M. Sato. Properties of fungiform papillae in frog’s tongue. Jpn J Physiol 7: 324-338, 1957. 21. Makous, W., S. Nord, B. Oakley and C. Pfaffmann. The gustatory relay in the medulla. In: Olfaction and Taste I. edited by Y. Zotterman. Oxford: Pergamon Press, 1963, pp. 381-393. 22. Maruhashi, J., K. Mizuguchi and I. Tasaki. Action currents in single afferent nerve tibres elicited by stimulation of the skin of the toad and the cat. J Physic>/ fLond) 117: 129-151, 1952.

23. Marui, T. Taste responses in the facial lobe of the carp, Cyprinus carpio L. Brain Res 130: 287-298, 1977. 24. Miyake, M., N. Kamo, K. Kurihara and Y. Kobatake. Physico-chemical studies of taste reception. V. Suppressive effect of salts on sugar response of the frog. B&him Biophys Acta 436: 856-862, 1976. 25. Morimoto, K. and M. Sato. Is serotonin a chemical transmitter in the frog taste organ? Life Sci 21: 1685-16%, 1977. 26. Ogawa, H., T. Imoto and T. Hayama. Responsiveness of solitario-parabrachial relay neurons to taste and mechanical stimulation applied to the oral cavity in rats. Exp Brain Res 54: 349-358, 1984. 27. Ogawa, H., M. Sato and S. Yamashita. Multiple sensitivity of chorda tympani fibres of the rat and hamster to gustatory and thermal stimuli. J Physiol (Land) 199: 223-240, 1968. 28. Opdam, P., M. Kemali and R. Nieuwenhuys. Topological analysis of the brain stem of the frogs Rana esklenta and i?ana catcsbeiana.

J Como Neural

165: 307-332. 1976.

29. PfatTmann, C., R. P. Erickson, G. P. Proher and B. P. Halpem. Gustatory discharges in the rat medulla and thalamus. In: Sensory Communication, edited by W. A. Rosenblith. New York: John Wilev. 1961. DD. 455-473. 30. Pumphrey, R. J. Nerve ~&&es from receptors in the mouth of the frog. J Cell Camp Physiol 6: 457-467, 1935. 3 1. Samanen, D. W. and R. A. Bernard. Response properties of the glossopharyngeal taste system of the mudpuppy (Necturus maculosus). II. Responses of individual first-order neurons. J Camp Physic)/ 143: 151-158, 1981.

32. Sato, T. Multiple sensitivity of single taste cells of the frog tongue to four basic taste stimuli. J CeN Phvsiol So: 207-218. 1972. 33. Smith, D. V. and J. B. Travers. A metric for the breadth of tuning of gustatory neurons. Chem Senses Flavor 4: 215-229, 1979. 34. Stanford, T. R. and M. Whitehead. Central organization of afferent fibers of the posterior tongue in the hamster. Sot Neurosci Abstr 10: 858, 1984. 35. Stuesse, S. L., W. L. R. Cruce and K. S. Powell. Organization

within the cranial IX-X complex in ranid frogs: A horseradish peroxidase transport study. j Camp Neurol222: 358-365, 1984. 36. Travers, J. B. and D. V. Smith. Gustatorv sensitivities in neurons of the hamster nucleus tractus solit&ius. Sens Prw<~sscs3: l-26, 1979. 37. Whitehead, M. C. and M. E. Frank. Anatomy of the gustatory system in the hamster: Central projections of the chorda tympani and the lingua1 nerve. J Camp Neural 220: 378-395, 1983. 38. Yamane, S. Influence of ions and chemical substances on the response of the frog’s tongue to mechanical stimulation. Camp B&hem Physiol [A] 61: 451-459, 1978. 39. Yamashita, S. Stimulating effectiveness of cations and anions on chemoreceptors in the frog tongue. Jpn J Physiol 13: 54-63, 1963. 40. Zotterman, Y. The response of the frog’s taste fibers to the application of pure water. Acta Physiol Stand 18: 181-189,

1949.