- Email: [email protected]
Latency of gustatory neural impulses initiated in frog tongue
Brain Research,424 (1987) 333-342
333
Elsevier BRE 13000
Latency of gustatory neural impulses initiated in frog tongue Toshihide Sato, Takenori Miyamoto and Yukio Okada Department of Physiology, Nagasaki UniversitySchool of Dentistry, Nagasaki (Japan) (Accepted 7 April 1987)
Key words: Latency; Frog tongue; Taste stimulus; Fungiform papillae; Gustatory neural impulse; Receptor potential
Latencies of gustatory neural impulses evoked by stimulation of the bullfrog tongue with the 4 basic taste substances (NaCI, acetic acid, quinine-HC1 (Q-HC1), sucrose), CaCI: and water were studied by recording antidromic impulses conducted to the fungiform papillae. Mean latencies of the impulses ranged from 58 to 107 ms when very strong stimuli, such as 2 M NaCI, 0.1 M acetic acid, 0.1 M QHCI, 2 M sucrose and 1 M CaCI:, were applied. Mean latency in response to water was 2.41 s. The time required for arrival of an applied taste stimulus on the taste receptor membrane was a mean of 20.1 ms. The time required for antidromic conduction from the impulse initiation site to the recording site was a mean of 2.4 ms. Electrical stimulation of the fungiform papilla with a strong intensity produced the impulse with a long and fluctuating latency. The mean minimum latency of the fluctuating impulse, from which the conduction time was subtracted, was 5.3 ms. Mechanical destruction of the taste disk situated at the top of the fungiform papilla resulted in a disappearance of the fluctuating impulse, suggesting that this was initiated synaptically via a depolarization of taste cells by electrical current. The minimum5.3-ms latency was likely to be the time required from the onset of taste cell depolarization to the initiationof an impulse at the first node of Ranvier of myelinated gustatory fiber. These results indicate that the latencies of 58-107 ms by strong taste stimulation were composed of the 30- to 79-ms latency of taste cell receptor potential and the remaining 28 ms latency, which was the sum of the time of stimulant diffusion, the time from taste cell depolarization to the first impulse and the time of impulse conduction.
INTRODUCTION Following application of a NaCI stimulus on the m a m m a l i a n tongue the latent period of impulses appearing in a gustatory neural fiber is reported to be 30-50 ms ~-4. However, the relations of the latent periods of gustatory neural responses to various kinds of taste stimuli and their concentrations are obscure. No systematic studies have b e e n carried out to understand the mechanism underlying the latency of gustatory neural impulses. In the present experiments, we studied the mechanisms determining the latency of gustatory neural impulses by examining (1) the relation between kinds of taste stimuli and latencies, (2) the properties of the impulses evoked by electrical stimulation of the fungiform papilla, and (3) the arrival time of applied taste stimuli on the taste receptor m e m b r a n e . Short c o m m u n i c a t i o n of some points of this experi-
ment has appeared elsewhere 14. MATERIALS AND METHODS
Preparation Forty-two adult bullfrogs (Rana catesbeiana) weighing 250-420 g were used in the experiments. The animal was anesthetized with an intraperitoneal injection of 50% (w/v) u r e t h a n - R i n g e r solution (2 g/kg body wt.). The animal was put in the supine position, and almost the entire tongue was pulled out from the mouth and fixed on the cork plate of an experimental chamber with steel pins. To remove the spontaneous contractions of the tongue both the hypoglossal nerves and the hypoglossal muscles were cut bilaterally. To record gustatory impulses the glossopharyngeal nerve on each side was dissected free from the surrounding connective tissues, cut centrally and immersed in mineral oil.
Correspondence: T. Sato, Department of Physiology, Nagasaki University School of Dentistry, 7-1, Sakamoto-machi, Nagasaki 852, Japan. 0006-8993/87/$03.50 (~) 1987 Elsevier Science Publishers B.V. (Biomedical Division)
334 Electrical recordings Single afferent nerve fibers innervating the frog 9ngue give rise to several branches under the dorsal pithelium and each branch supplies the fungiform ~apilla ~l. Therefore, electrical and chemical stimulaion of a fungiform papilla containing the taste disk •esulted in an appearance of orthodromic neural im9ulses in the central direction and antidromic neural mpulses in the peripheral direction (cf. Figs. 1A and 2A). According to the method of Rupuzzi and Casella 11, electrical activities of single gustatory nerve fibers were recorded antidromicaily from the fungiform papilla sucked with a suction electrode during electrical or chemical stimulation of the surrounding fungiform papillae. Diameters of the fungiform papillae in the bullfrogs ranged from 270 to 370/~m 18. To tightly suck the papilla the tips of suction electrodes, which were made with glass capillary tubes of 3 mm in diameter, had the inner diameter of 227 + 8 /~m (mean + S.E.M., n = 21) with a range of 166-282 /~m. An indifferent electrode was composed of a silver wire of 200/~m in diameter, which was coated with enamel excepting its tip. This was glued to the glass wall of the suction electrode and the tip was located at the base of the fungiform papilla sucked. Neural impulses were amplified with an R-C coupled amplifier, displayed on an oscilloscope, and photographed with a kymograph camera. By a band-pass filter of the amplifier the impulse components of higher frequencies than 1 kHz were cut off. Electrical and taste stimulation In some experiments electrical stimulation was given to a fungiform papilla which was deeply sucked by a stimulating suction electrode filled with a normal saline solution (cf. Figs. 1A and 7A). The other stimulating electrode of an enamel-coated silver wire was glued to the body of the stimulating suction electrode and its tip was located at the base of the fungiform papilla. Taste stimuli used as the 4 basic taste stimuli were 0.5 M NaCI, 3 mM acetic acid, 1 mM quinine-HCl (Q-HC1) and 0.5 M sucrose. Besides these, 1 mM CaCI 2 and deionized water (Milli Q reagent-grade water, Millipore, MA) were employed frequently. Much stronger concentrations of the 4 basic taste substances were also used to determine the shortest latencies of gustatory impulses. When examining the
dose-latency relationships, various concentrations of the 4 basic taste substances and CaC12 were used. These substances were all dissolved in deionized water. The taste stimuli were delivered to the tongue surface with a semi-automatically controlled gustatory stimulator described previously 13. A flow rate from the output injection needle of gauge no. 21 was adjusted as 0.123 ml/s because the more rapid flow rate produced a distortion of oscilloscope beam due to the mechanical movement of the recording suction electrode. The stimulator's nozzle was put near the recording suction electrode drawing the fungiform papilla and arranged just over a thin Ringer layer covering the tongue surface (cf. Fig. 2A). As soon as the first drop of a stimulus solution reached the tongue surface, a surge of stimulus arrival mark due to the physicochemical reaction appeared on the oscilloscope. Electrical stimulation of the nerve fibers innervating a fungiform papilla evoked the impulses propagating antidromically to a large number of neighboring papillae 1~. The area of the cross-connected papillae was measured by electrically stimulating one fungiform papilla and by recording the antidromic impulses from the surrounding neurally connected papillae. The area was usually a circle of 2.5-3.5 mm in diameter. The time which was taken for a stimulus solution to flow the surface area of the whole neurally connected papillae was checked as follows: Both the recording microelectrode filled with 0.3 M KCI and the gustatory stimulator nozzle filled with 0.5 M NaC1 stimulus were placed in the field of the crossconnected papillae as separate as possible, and the time interval between the onset of taste stimulation and the onset of physicochemical junction potential occurring at the microelectrode tip was measured. When the diameter of the cross-connected papillae was a maximum of 3.5 mm, the maximal time interval was as small as 2 ms. This indicates that a taste stimulus flowed without time delay on the whole area of cross-connected papillae under investigation. The tongue surface was always adapted to the flowing Ringer solution whose composition (in mM) was: NaC1 115, KCI 2.5, CaCI 2 1.8, Na2HPO 4 2.15, NaHzPO4 0.85. The pH was adjusted to 7.2 by adding bicarbonate or NaOH. Two min before application of a taste stimulus, the adapting Ringer flow was stopped, and then the tongue surface was covered
335 SSE
~
1
RSE
N
~="AP2 1 HZ
R1 >::a
",'l,,p, ,,~P~ GN ~
N
R2
1~
2 2 ms
1 Hz Ii/~:AP2 q
100 Hz R1
~lr ~ ~w"I AP2
R1 1:12
Fig. 1. Action potentials recorded simultaneously from the papillary nerve fibers and the whole glossopharyngeal nerve in response to anodal electrical stimulation of a fungiform papilla. A: a schematic arrangement of recording and stimulation. SSE, stimulating suction electrode filled with normal Ringer. RSE, recording suction electrode filled with normal Ringer. P, fungiform papilla. N, gustatory nerve fibers. GN, whole glossopharyngeal nerve. R1 and R2, amplifiers for antidromic impulses in the papillary nerve fibers and orthodromic impulses in the glossopharyngeal nerve, respectively. B: action potential from papillary nerve (R1) and glossopharyngeal nerve (R2) evoked by near-threshold stimulus intensity of 0.03 ms pulse at 0.5 Hz. AP1 and AP2, spike potential of low threshold and that of slightly higher threshold, respectively. C: the same as in B except for slight increase in stimulus intensity. D: the same as in C except for stimulus frequency at 100 Hz.
statically with a thin layer of Ringer solution. After taste stimulation the tongue surface was rinsed with Ringer. The onset of the oscilloscope sweep was triggered by current pulses controlling electromagnetic valves by which the nozzles of the gustatory stimulator were opened or closed. After chemical stimulation was applied to the tongue, the latent period of gustatory neural impulses was measured as the time interval between the stimulus touching artifact and the first impulse evoked in a gustatory nerve fiber. All the experiments were carried out at a room temperature of 22-26 °C. RESULTS
Characteristics of orthodromic and antidromic impulses in gustatory nerve fibers Following an electrical stimulation of a fungiform papilla, we could record both antidromic neural impulses from another neighboring fungiform papilla
and orthodromic impulses from a glossopharyngeal nerve. A n example is shown in Fig. 1. In the upper trace (R1) of Fig. 1B, which was recorded antidromically from the papillary nerve fibers (Fig. 1A), a stimulus intensity of near-threshold produced two biphasic spike potentials from two different fibers. The spike marked AP1 was produced by slightly lower threshold than that for the other spike marked AP2. In a lucky case as in this preparation, recording (R2) from the whole glossopharyngeal nerve also showed two biphasic orthodromic spikes, which corresponded to the two antidromic spikes in the papillary nerve fibers (R1). These nerve fibers fired antidromically or orthodromically when the surrounding fungiform papillae were stimulated with 1 m M CaCI 2 and deionized water, so that the fibers were presumed to be gustatory ones. In Fig. 1C, the papilla was electrically stimulated with a slightly increased intensity at 1 Hz. The compound action potentials composed of two kinds of spikes (AP1 and AP2) were seen in both the papillary nerve and the glossopharyngeal nerve. Even when the stimulus frequency was increased to 100 Hz, neither failure of the biphasic action potentials occurred in both nerves (Fig. 1D), nor the shape, height and latency of the action potentials changed. However, at the frequency of~ several hundred Hz the action potentials sometimes failed to appear and the biphasic shape sometimes changed into the monophasic. Electrical threshold for eliciting a neural impulse was lower when the top of the papilla was anodally stimulated than when the top was cathodally stimulated. The threshold of cathodal and anodal currents depended largely on the spike amplitudes: the smaller the amplitude, the higher the threshold. The difference in the threshold between cathodal and anodal currents was 0.61 _+ 0.08 p A (21 units) when a pulse duration was 0.03 ms. The ratio of cathodal to anodal current threshold was 1.32 _+ 0.05 (21 units). Since about 10 afferent fibers of varying diameters are present in the fungiform papilla 17'1s, the various amplitudes of antidromic impulses (peak to peak, 15-180 pV) were elicited in the neurally connected papilla by electrical and chemical stimulation of the neighboring papillae. In each gustatory unit, the maximal voltage deviation from the mean magnitude of the impulses was a range of +3.6/2V, which corre-
336 Suction electrode
R
A
inger layet ~
i
u
I
~
!
s
Nozzle of gustatory stimulator
~ drOO Of taste stimulant
t
i t ' l l f ~ilviiiilll
0.5 M sucrose o~,
0.5 M NaCI ee
E
I o"=
I° •
3 mM acetic acid
~ 9 ms
100 ms
F
1 mM oQ'HCIo
i
>=
>=
100 ms
Deionized water
400 ms
G
IS 100 ms
reached the Ringer solution layer covering the tongue surface, an electrical surge as physicochemical p h e n o m e n o n a p p e a r e d ( m a r k S in Fig. 2 B - G ) . A train of impulses in single gustatory neural units evoked by application of 0.5 M NaC1, 3 m M acetic acid, 1 mM Q-HCI, 0.5 M sucrose, 1 m M CaCI 2 and deionized water were shown in Fig. 2 B - G . In each record, the filled circles indicate the first impulses elicited in different gustatory units. Fig. 3 illustrates latencies of gustatory impulses of 50-171 units in response to the 4 basic stimuli, 1 m M CaCI 2 and deionized water. The mean latency was 158 ms for 0.5 M NaCt, 218 ms for 3 mM acetic acid, 207 ms for 1 m M Q-HC1 and 557 ms for 0.5 M sucrose. The mean latencies for 1 m M CaCI 2 and deionized water were 441 ms and 2.41 s, respectively. The latencies for NaCI, acetic acid and Q-HCI were relatively shorter, but those for the sucrose, CaC12 and water were longer and distributed broadly. Application of much stronger stimuli of 2 M NaCI, 0.1 M acetic acid, 0.1 M Q-HC1 and 2 M sucrose and
2s
Fig. 2. Antidromic impulses in single gustatory neural fibers evoked by chemical stimulation of the tongue. A: schematic drawing of electrical recording and chemical stimulation. B-G: gustatory impulses of papillary nerve fibers evoked by 4 basic taste stimuli, CaC12 and deionized water. S, electrical surge showing just a contact of stimulant on the tongue surface. Closed circles denote the first impulses in each unit.
30 - A
05 M NaCI n= 96
t l
0003~ acetic acid =
20 -
I
sponded to the maximal level of the white noises (+3.7 ¢tV) in the recording systems used. In other words, the percentage of the maximal deviation from the mean impulse magnitude decreased exponentially with increasing impulse height or fiber diameter. Very strong chemical stimulation of the tongue with 0.1 M Q-HC1, 2 M sucrose, 0.1 M acetic acid, 2 M NaCI and 1 M CaC12 also p r o d u c e d antidromic gustatory impulses of + 3 . 6 / ~ V deviation in every unit of the papilla. These observations indicate that the amplitude of impulses originating from single gustatory units of any diameters is physiologically constant.
Latency of gustatory neural impulses in response to 4 basic st°muff, CaCI2 and water In Fig. 2 A is shown a schematic arrangement for recording antidromic impulses and for applying taste stimulation. W h e n the first drop of a taste stimulant
[-h °.°°, n=125
[
I
I
I
I
200 400 600 800 1000 1200 ms
°-.c.
0.5 M sucrose =
lO
I 30
-E
o.oo1 M CaCI 2 n= 171
10
0
i
0
i
I
I
i
I
I
i
I
I
)
O 400 800 1200 1600 2000 2400ms F
Deionized water n=50
t
200 400 800 800 ms 0 Latency
2
4
6
I 8
t 10
I 12 s
Fig. 3. Histograms of latencies of gustatory neural impulses elicited by 4 basic taste stimuli, CaCI 2 and deionized water. The mean and S.E.M. was 158 + 11 ms for 0.5 M NaCI (A), 218 + 10 ms for 0.003 M acetic acid (B), 207 + 8 ms for 0.001 M QHCI (C), 557 + 41 ms for 0.5 M sucrose (D), 441 + 18 ms for 0.00l CaCI2 (E) and 2.41 _+0.25 s for deionized water (F).
337 1 M CaCI 2 caused gustatory impulses with shorter latencies of 75 + 7 ms (mean + S.E.M., n = 32), 58 + 13 ms (n = 10), 87 + 7 ms (n = 29), 107 _+ 15 ms (n = 12) and 82 + 13 ms (n = 20), respectively. The shortest latency in the gustatory unit was 8.6 ms for 2 M NaCl, 7.2 ms for 0.1 M acetic acid, 21.6 ms for 0.1 M Q-HCI and 45.5 ms for 1 M sucrose.
oo
oo[ ~
7.0: 5.0 3.0 1.0 0.6 0.4
= Acetic acid
0.4 ! ~ ~ ~ ' x 0.3
n=23 " ! %
0.2 0.1
0.2
0 (3o 0.6 0.4 ~,, 0.2
O.t
~
-
0.3
0
1.0
(~),~','~ Q-HCI
.~
~n=29
!
" \
,
i
30
\\
1'0-2
o.s ,
10-5 10~4 10-3 10-2 / ooj ~(~)~\ 5.0!
1'0-3
10.c~
:~'~~ =
1()-4
0
Relation between stimulus concentrations and latencies When the tongue surface was adapted to a normal Ringer solution, in 54% of 13 units examined no impulses were evoked by 0.1 M NaCI, whereas in the remaining 46% the latencies ranged from 0.5 to 5 s (Fig. 4A). Latencies of all 13 units decreased with increasing concentration of NaCI. In case of acetic acid stimulation, 35% of 23 units sampled had an infinite latency to 10-4 M, and 22% had an infinite latency to 10-3 M (Fig. 4B). The threshold for acetic acid in 65% of the units was lower than 10 -4 M. All the units showed the decrease in latency with increasing concentration. In case of Q-HCI stimulation, 65.5% of 29 units had no responses to 10-5 M. The remaining 34.5% units had a latency of 0.2-0.5 s at 10-5 M and showed a decrease in that with increasing concentration (Fig. 4C). Stimulation with sucrose did not produce any impulse in 80% of 25 units at 0.1 M and 52% at 0.3 M. Twenty percent of the units responded to 0.1 M sucrose with a long latency of 2 - 9 s (Fig. 4D). The latencies for all the sucrose-sensitive units also gradually decreased with an increase in concentration. With CaCI2, 32% and 11% of 19 units examined had an infinite latency at 10-5 and 10-4 M, respectively. The other units of 68% responded to 10-5 M CaC12 with latencies of 0.2-1.3 s (Fig. 4E).
Relation of diameter of neural fibers to latency
011
0.3
1.0
CaCI2
n=t,
1.0" ~ 0.8 0.6 O.4 0.2 0 lO-S 1'o-4 l'O-a 1'0-2 Concentration
Simultaneous recording from several papillary fibers within a single fungiform papilla after chemical stimulation of the neighboring fungiform papillae frequently showed that there was a linear relation between the amplitudes of gustatory neural impulses and the latencies (Fig. 5). The smaller the spike amplitude, the shorter the latency of the spike evoked by a taste stimulation. When stimulated with 6 taste stimuli, typical examples of these relations are illustrated in Fig. 5. In gustatory neural fibers there is a linear relation between the spike amplitude and the fiber diameter 15, so it is said that the smaller the fiber diameter, the shorter the spike latency.
(M)
Fig. 4. Relation between concentrations of 4 basic taste stimuli
Conduction time of impulses
and CaCI2 and lateneies of impulses in gustatory nerve fibers. Experimental method is shown in Fig. 2A. When no impulses were initiated during 10 s after application of lower concentrations of taste stimuli, the unit under investigation was regarded os havingan infinite latency.
The conduction times of impulses traveling along gustatory nerve fibers connecting between the fungiform papillae were calculated (inset in Fig. 6C,D). In an example shown in Fig. 6A and B, 3 connected fi-
338 500
A
NaCI
D
Acetic acid
E
sucrose
bers were found by electrical stimulation of various intensities. Supramaximal stimulation p r o d u c e d 3 spikes m a r k e d as O , • and II, whose latencies were shorter to anodal stimulation (Fig. 6A) than to cathodal stimulation (Fig. 6B). In a total of 40 connected fibers examined, the m e a n latency of their impulses was 2.1 ms to anodal stimulation (Fig. 6C), but 2.4 ms to cathodal stimulation (Fig. 6D). M e a n difference between the two latencies was 0.3 ms.
250
i
0 500
/?
B
E 250
e/
CaCI2
Y
Arrival time of taste stimulus on taste receptor membrane I
0 500
C
Q-HCI
oo
F
/B
H20
, o £ soY
,;o
,oo
40
Amplitude of gustatory neural impulse (~uV)
Fig. 5. Relation between latencies of gustatory neural impulses and their magnitudes. The taste stimuli used were 0.5 M NaCI, 0.003 M acetic acid, 0.001 M Q-HC1, 0.5 M sucrose, 0.001 M CaCIz and deionized water. The data of the same symbols on each straight line were obtained from gustatory nerve fibers of single fungiform papillae.
Fluctuating latency of neural impulses
riSE SSE
A
15
+
;~,
N
'
lo
~
NN
C
~
>
O,
2~
RSE SSE
z
O;
0
1
2
3
4
5
Before application of a taste stimulus, the tongue surface was covered stationarily with a thin layer of Ringer solution. The thickness of the Ringer layer measured from the top surface of the fungiform papillae was 56 _+ 8/~m (12 papillae). A f t e r a taste stimulus touched the surface of Ringer solution as in Fig. 2A, the time when the stimulus reached the surface of the taste disk situated at the top of the fungiform papillae was estimated by measuring the physicochemical junction potential between an applied 2-M NaCI solution and a microelectrode tip put on the papilla surface covered with much mucus. The mean arrival time so m e a s u r e d was 20 + 3 ms (48 papillae) with a range of 4 - 5 8 ms.
6
Latency (ms)
Fig. 6. Latencies of neural impulses evoked by single electrical stimulation of a fungiform papilla. Schematic arrangement of stimulating and recording suction electrodes is shown in the insets of C and D. A and B: examples of 3 action potentials (O, &, II) in different papillary nerve fibers to electrical stimulation of a fungiform papilla with 0.03 ms pulse. The top of the fungiform papilla was stimulated anodally in A or cathodally in B. C and D: histograms of latencies of the neural impulses to anodal (C) and cathodal (D) stimulations.
No antidromic impulses, e v o k e d by weak electrical stimulation of a fungiform papilla, had a fluctuation of latency and amplitude. H e r e we t e r m e d these impulses as the first group of impulses (letter I in Fig. 7 B - D ) . On the other hand, when the stimulus intensity was raised to more than 5.2 times (20 units), there a p p e a r e d the second group of impulses (SGI) (letter II in Fig. 7D) having a considerably fluctuating latency. In case of Fig. 7, SGI a p p e a r e d intermittently in spite of anodal stimulation at 1 Hz. S G I was p r o d u c e d by either anodal or cathodai stimulation of the fungiform papilla, but the threshold was lower when anodally stimulated than when cathodaily stimulated. Even at 0.5 Hz of stimulus frequency, the latency of SGI fluctuated. The mean m i n i m u m and maximum latencies of SGI in 24 neural units, from which the conduction time of the impulses was subtracted, were 5.3 + 0.4 ms and 10.0 --- 0.7 ms, respectively. A f t e r the taste disk situated at the top of the fungiform papillae was mechanically injured by a
339
SSE
RSE
I
f
I
I 2
I
I
3
4
A
N
E
4
-5 O.
B-¢xc-----
E
2
Stimulus
intensity
(pA)
0 I C 0
I
6
C
I
I
I
I
I
f
I
I
B •A
•
JI
~ e
te-e ~
4
..1
I
I
I
l
I
I
I
I
Stimulus d u r a t i o n
Fig. 7. Appearance of the second group of impulses in the papillary neural fibers following anodal strong stimulation of a fungiform papilla. A: arrangement of a stimulating suction electrode (SSE) and a recording one (RSE). P, fungiform papilla. N, gustatory nerve fibers. B - D : recordings of antidromic neural impulses to electrical stimulation at 0.5 Hz with 0.03-ms pulses of different strengthes. In B and C, only the first group of impulses designated I was evoked. Two neurally connected fibers existed between the two papillae. In D, the second group of impulses designated II following the first group was evoked by stimulation of stronger intensity. One spike in the first group was evoked in the record B, but two kinds of spikes in the first group making a compound action potential were evoked in the record C and D. In record D, 20 sweeps were superimposed, but 10 impulses in the second group were evoked intermittently.
pair of fine forceps, SGI disappeared, but the first group of impulses remained unchanged. The latency of SGI changed depending on stimulus intensity: the stronger the stimulus intensity, the shorter the latency (Fig. 8A). Lengthening the duration of an electrical pulse, keeping the intensity constant, did not change the latency of second group (Fig. 8B). SGI appeared intermittently depending on the frequency of repetitive stimulation of the fungiform papilla. In Fig. 9, this intermittent response pattern is shown in 5 units. One unit showed a mostly complete
I
1 1.0
0.5 (ms)
Fig. 8. A: relation between stimulus intensity and latency of the second group of impulses in 3 units. Stimulus duration was 0.1 ms. B: relation between stimulus duration and latency of the second group of impulses in 3 units. Stimulus intensity was 2 /tA. In A and B, the top of the fungiform papillae was anodally stimulated.
failure of second group even at 1 Hz and another unit that at 10 Hz. When the duration of a single electrical pulse was less than 1 ms, one impulse in the second group was evoked in most cases, but two or more impulses were
==
100 •
o
\\
"o o o
5Q.
50
•E_ 0
o
o~
1'0
0 Stimulus frequency (Hz)
Fig. 9. Relation between electrical stimulus frequency and occurrence of the second group of impulses. Data from 5 units. The papillae were anodally stimulated.
340
4 3 "S
2
=E 1
z
0
i
0
10
i
i
i
20
30
40
5O
Stimulus duration (ms)
Fig. 10. Relation between duration of electrical stimulus pulse and number of impulses in a single neural unit. Method of recording and stimulation was the same as in Fig. 7.
produced in some cases. As the stimulus duration was increased to more than 1 ms, the number of the impulses increased gradually (Fig. 10).
DISCUSSION The present experiment, in which the latency from the touch of a taste stimulus on the tongue surface to the first gustatory neural impulse was measured, showed that 2 M NaCI, 0.1 M acetic acid, 0.1 M Q-HCI, 2 M sucrose and 1 M CaCI 2 of very strong stimuli produced a mean latency of 58-107 ms.
Tas?cell
_ ~ Taste Stimulus
"f[~
B
~or./~/l['![
C --
~' potential /
D --
~ ] } /
E-
Suction electrode --
t
II fl II II
/
II : g'form papilla
~L'~'~//Z/Gu~stat~rynerve fiber dl Initiated spike /
Je I I .
/
Conducted spike ]Fig. l]. A possible explanation of mechanism of latency of gustatory neural impulseselicitedby chemicalstimulation.A: ap-
plicationof taste stimulus. B: generation of receptor potential
with latency b. C: generation of postsynaptic potential with latency c after B. D: impulses with latency d initiated at the first node of Ranvier of myelinated gustatory fiber after C. E: conducted spikes with latency e after D.
In Fig. 11, all processes of appearance of gustatory neural impulses by chemical stimulation of the fungiform papillae are illustrated schematically. Gustatory impulses to the chemical stimulation have been found to generate at the first node of Ranvier of myelinated nerve fibers running inside the fungiform papillae 8. Generation of an impulse at the first node of Ranvier may be as follows: arrival of a taste stimulus on the taste receptor membrane --~ depolarization of a taste cell --~ release of a transmitter --~ generation of a postsynaptic potential ~ initiation of a gustatory impulse at the first node of Ranvier. The real time course of these whole processes has not yet been clarified. It has been reported that latency of depolarizing receptor potential in a frog taste cell evoked by application of 0.5 M NaC1 and 15.6 mM acetic acid is 100-300 m s 12'16. Electrical stimulation of the fungiform papilla in a frog produced two kinds of neural impulses termed the first and second groups. The latency of the first group of impulses having a low threshold was shorter and very stable. On the other hand, the second group of impulses had the following characteristics: (1) a high threshold, (2) a long and fluctuating latency, (3) a decrease of latency with increasing stimulus intensity, (4) a failure of impulses at a low stimulus frequency, and (5) a repetitive firing in response to a single pulse of long duration. The two kinds of impulses in the frog fungiform papillae have already been reported by Nomura and Katsuhata 1° and Sato TM.Fluctuation of latency similar to that of the second group of impulses in the present study has been found in the endplate potential of the neuromuscular junction 6'7. Fig. 12 illustrates an assumption of stimulus current flows through a taste cell and a gustatory nerve fiber, when the top of the fungiform papilla was stimulated anodally or cathodally. Judging from the short and stable latency of the low-threshold impulses in the first group, a weak stimulus current probably did not flow the taste cell but mostly flowed the gustatory axon (Ia in Fig. 12). In case of the anodal stimulation the stable latency of an impulse was, on the average, 0.3 ms shorter than that in case of the cathodal stimulation. It is, therefore, likely that the impulse induced by anodal stimulation of the fungiform papilla might be initiated at the third node of Ranvier because this was situated just outside of the fungiform papilla sucked by the stimulating suction electrode (Figs. 11
341
I
ste cell
t~ D o. cO
E
It.
Fig. 12. Stimulus current flows through a gustatory axon (Ia), a taste cell (It) and an extracellular space (Ie) in the fungiform papilla.
and 12) 8, whereas the impulse by cathodal stimulation might be initiated at unmyelinated portion near the first myelin sheath. The fact that the threshold of cathodal current was 1.32 times higher than that of anodal current is taken to indicate that the membrane of the third node of Ranvier possessed higher excitability. Stronger stimulus current might flow through taste cells (It in Fig. 12) and produce a critical depolarization, which in turn might initiate a release of transmitter, an excitatory postsynaptic potential at the axon terminal and an impulse in the second group at the first node of Ranvier. It has been reported that, after chemical blockage of sensory synapses between taste cells and gustatory afferent fibers in the frog, the neural response composed of initial phasic and tonic components to anodal electrical stimulation of the tongue was mostly suppressed 5. This suggests that the response of gustatory nerve fibers by anodal stimulation of the tongue was induced synaptically via an excitation of taste cell 5'19.
In the present experiment, mechanical destruction of the taste disk situated at the summit of the fungiform papillae did not elicit the second group of impulses by anodal stimulation of the papilla, but still initiated the first group of impulses. This also suggests that the second group of impulses was synaptically elicited via a depolarization of taste cells by currents. Since it is suggested that a single papillary nerve fiber in the frog tongue branches many times and innervates a great number of taste cells 18, it is probable that the distances from the afferent terminals to the first node of Ranvier of a myelinated nerve fiber are diverse. Therefore, synaptic potentials elicited in the postsynaptic membranes following a depolarization of taste cells might spread, with various time courses, to the first node of Ranvier. It is assumed that the decayed synaptic potential of slow rise time slowly induces a critical depolarization at the impulse firing zone, so that the latency of the first gustatory impulse induced by anodal stimulation might take as long as 5 ms. From the data of Nomura and Sakada 9 who first recorded compound synaptic potental generated in the gustatory afferent terminals of the frog tongue, we can calculate the peak time of 10-25 ms in the gustatory synaptic potentials. By strong gustatory stimulation with very high concentrations, we found a latency of 58-107 ms in the gustatory neural impulses. These values include (i) the mean arrival time of a stimulus to receptor membrane (20.1 ms), (ii) the mean minimal delay (5.3 ms) from a depolarization of taste cell to an impulse initiation, and (iii) the mean conduction time of the impulse (2.4 ms). The percentage of a total of 27.8 ms in chemically evoked impulse latency of 58-107 ms is 26-48%. Therefore, the 52-74% (30-79 ms) of the latency elicited by strong chemical stimuli is due to the true latency between the arrival of a chemical stimulus to the receptor membrane and the onset of the receptor potential.
ACKNOWLEDGEMENTS This study was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan. We thank Miss Y. Kitada for her excellent assistance in preparing this paper.
342 REFERENCES 1 Beidler, L.M., Properties of chemoreceptors of tongue of rat, J. Neurophysiol., 16 (1953) 595-607. 2 Beidler, L.M., Physiological properties of mammalian taste receptors. In G.E.W. Wolstenholme and J. Knight (Eds.), Ciba Foundation Symposium on Taste and Smell in Vertebrates, Churchill, London, 1970, pp. 51-70. 3 Faull, J.R. and Halpern, B.P., Taste stimuli: time course of peripheral nerve response and theoretical models, Science, 178 (1972) 73-75. 4 Halpern, B.P. and Marowitz, L.A., Taste responses to lickduration stimuli, Brain Research, 57 (1973) 473-478. 5 Kashiwayanagi, M., Yoshii, K., Kobatake, Y. and Kurihara, K., Taste transduction mechanisum: similar effects of various modifications of gustatory receptors on neural responses to chemical and electrical stimulation in the frog, J, Gen. Physiol., 78 (1981) 259-275. 6 Katz, B. and Miledi, R., The effect of temperature on the synaptic delay at the neuromuscular junction, J. Physiol. (London), 181 (1965) 656-670. 7 Katz, B. and Miledi, R., The release of acetylcholine from nerve endings by graded electric pulses, Proc. R. Soc. London B, 167 (1967) 23-38. 8 Miyamoto, T., Mineda, T., Okada, Y. and Sato, T., Initiation site of gustatory neural impulses in frog tongue, Jpn. J. Oral Biol., 27 (1985) 389-391. 9 Nomura, H. and Sakada, S., Local potential changes at sensory nerve fiber terminals of the frog tongue. In C. Pfaffmann (Ed.), Olfaction and Taste III, Rockefeller University Press, New York, 1969, pp. 345-351.
10 Nomura, H. and Katsuhata, T., Responses of a chemoreceptor to mechanical and electrical stimulation in the frog tongue, Bull. Tokyo Dent. Coll., 14 (1973) 169-193. 11 Rapuzzi, G. and Casella, C., Innervation of the fungiform papillae in the frog tongue, J. Neurophysiol., 28 (1965) 154-165. 12 Sato, T., Multiple sensitivity of single taste cells of the frog tongue to four basic taste stimuli, J. Cell Physiol., 80 (1972) 207-218. 13 Sato, T., Adaptation of primary gustatory nerve responses in the frog, Comp. Biochem. Physiol., 43A (1972) 1-12. 14 Sato, T., Latency of taste nerve signals in frog (Rana catesbeiana), Experientia, 32 (1976) 877-879. 15 Sato, T. and Sugimoto, K., The adaptation of the frog tongue to bitter solutions: enhancing effect on gustatory neural response to salt stimuli, Comp. Biochem. Physiol., 62A (1979) 965-981. 16 Sato, T., Recent advances in the physiology of taste cells, Prog. Neurobiol., 14 (1980) 25-67. 17 Sato, T., Ohkusa, M. and Sugimoto, K., Incremental conduction velocity of single afferent fibers innervating frog taste organ, Jpn. J. Physiol., 30 (1980) 655-658, 18 Sato, T., Ohkusa, M., Okada, Y. and Sasaki, M., Topographical difference in taste organ density and its sensitivity of frog tongue, Comp. Biochem. Physiol., 76A (1983) 233-239. 19 Yamamoto, T., Yuyama, N. and Kawamura, Y., Responses of cortical taste cells and chorda tympani fibers to anodal DC stimulation of the tongue in rats, Exp. Brain Res., 40 (1980) 63-70.
333
Elsevier BRE 13000
Latency of gustatory neural impulses initiated in frog tongue Toshihide Sato, Takenori Miyamoto and Yukio Okada Department of Physiology, Nagasaki UniversitySchool of Dentistry, Nagasaki (Japan) (Accepted 7 April 1987)
Key words: Latency; Frog tongue; Taste stimulus; Fungiform papillae; Gustatory neural impulse; Receptor potential
Latencies of gustatory neural impulses evoked by stimulation of the bullfrog tongue with the 4 basic taste substances (NaCI, acetic acid, quinine-HC1 (Q-HC1), sucrose), CaCI: and water were studied by recording antidromic impulses conducted to the fungiform papillae. Mean latencies of the impulses ranged from 58 to 107 ms when very strong stimuli, such as 2 M NaCI, 0.1 M acetic acid, 0.1 M QHCI, 2 M sucrose and 1 M CaCI:, were applied. Mean latency in response to water was 2.41 s. The time required for arrival of an applied taste stimulus on the taste receptor membrane was a mean of 20.1 ms. The time required for antidromic conduction from the impulse initiation site to the recording site was a mean of 2.4 ms. Electrical stimulation of the fungiform papilla with a strong intensity produced the impulse with a long and fluctuating latency. The mean minimum latency of the fluctuating impulse, from which the conduction time was subtracted, was 5.3 ms. Mechanical destruction of the taste disk situated at the top of the fungiform papilla resulted in a disappearance of the fluctuating impulse, suggesting that this was initiated synaptically via a depolarization of taste cells by electrical current. The minimum5.3-ms latency was likely to be the time required from the onset of taste cell depolarization to the initiationof an impulse at the first node of Ranvier of myelinated gustatory fiber. These results indicate that the latencies of 58-107 ms by strong taste stimulation were composed of the 30- to 79-ms latency of taste cell receptor potential and the remaining 28 ms latency, which was the sum of the time of stimulant diffusion, the time from taste cell depolarization to the first impulse and the time of impulse conduction.
INTRODUCTION Following application of a NaCI stimulus on the m a m m a l i a n tongue the latent period of impulses appearing in a gustatory neural fiber is reported to be 30-50 ms ~-4. However, the relations of the latent periods of gustatory neural responses to various kinds of taste stimuli and their concentrations are obscure. No systematic studies have b e e n carried out to understand the mechanism underlying the latency of gustatory neural impulses. In the present experiments, we studied the mechanisms determining the latency of gustatory neural impulses by examining (1) the relation between kinds of taste stimuli and latencies, (2) the properties of the impulses evoked by electrical stimulation of the fungiform papilla, and (3) the arrival time of applied taste stimuli on the taste receptor m e m b r a n e . Short c o m m u n i c a t i o n of some points of this experi-
ment has appeared elsewhere 14. MATERIALS AND METHODS
Preparation Forty-two adult bullfrogs (Rana catesbeiana) weighing 250-420 g were used in the experiments. The animal was anesthetized with an intraperitoneal injection of 50% (w/v) u r e t h a n - R i n g e r solution (2 g/kg body wt.). The animal was put in the supine position, and almost the entire tongue was pulled out from the mouth and fixed on the cork plate of an experimental chamber with steel pins. To remove the spontaneous contractions of the tongue both the hypoglossal nerves and the hypoglossal muscles were cut bilaterally. To record gustatory impulses the glossopharyngeal nerve on each side was dissected free from the surrounding connective tissues, cut centrally and immersed in mineral oil.
Correspondence: T. Sato, Department of Physiology, Nagasaki University School of Dentistry, 7-1, Sakamoto-machi, Nagasaki 852, Japan. 0006-8993/87/$03.50 (~) 1987 Elsevier Science Publishers B.V. (Biomedical Division)
334 Electrical recordings Single afferent nerve fibers innervating the frog 9ngue give rise to several branches under the dorsal pithelium and each branch supplies the fungiform ~apilla ~l. Therefore, electrical and chemical stimulaion of a fungiform papilla containing the taste disk •esulted in an appearance of orthodromic neural im9ulses in the central direction and antidromic neural mpulses in the peripheral direction (cf. Figs. 1A and 2A). According to the method of Rupuzzi and Casella 11, electrical activities of single gustatory nerve fibers were recorded antidromicaily from the fungiform papilla sucked with a suction electrode during electrical or chemical stimulation of the surrounding fungiform papillae. Diameters of the fungiform papillae in the bullfrogs ranged from 270 to 370/~m 18. To tightly suck the papilla the tips of suction electrodes, which were made with glass capillary tubes of 3 mm in diameter, had the inner diameter of 227 + 8 /~m (mean + S.E.M., n = 21) with a range of 166-282 /~m. An indifferent electrode was composed of a silver wire of 200/~m in diameter, which was coated with enamel excepting its tip. This was glued to the glass wall of the suction electrode and the tip was located at the base of the fungiform papilla sucked. Neural impulses were amplified with an R-C coupled amplifier, displayed on an oscilloscope, and photographed with a kymograph camera. By a band-pass filter of the amplifier the impulse components of higher frequencies than 1 kHz were cut off. Electrical and taste stimulation In some experiments electrical stimulation was given to a fungiform papilla which was deeply sucked by a stimulating suction electrode filled with a normal saline solution (cf. Figs. 1A and 7A). The other stimulating electrode of an enamel-coated silver wire was glued to the body of the stimulating suction electrode and its tip was located at the base of the fungiform papilla. Taste stimuli used as the 4 basic taste stimuli were 0.5 M NaCI, 3 mM acetic acid, 1 mM quinine-HCl (Q-HC1) and 0.5 M sucrose. Besides these, 1 mM CaCI 2 and deionized water (Milli Q reagent-grade water, Millipore, MA) were employed frequently. Much stronger concentrations of the 4 basic taste substances were also used to determine the shortest latencies of gustatory impulses. When examining the
dose-latency relationships, various concentrations of the 4 basic taste substances and CaC12 were used. These substances were all dissolved in deionized water. The taste stimuli were delivered to the tongue surface with a semi-automatically controlled gustatory stimulator described previously 13. A flow rate from the output injection needle of gauge no. 21 was adjusted as 0.123 ml/s because the more rapid flow rate produced a distortion of oscilloscope beam due to the mechanical movement of the recording suction electrode. The stimulator's nozzle was put near the recording suction electrode drawing the fungiform papilla and arranged just over a thin Ringer layer covering the tongue surface (cf. Fig. 2A). As soon as the first drop of a stimulus solution reached the tongue surface, a surge of stimulus arrival mark due to the physicochemical reaction appeared on the oscilloscope. Electrical stimulation of the nerve fibers innervating a fungiform papilla evoked the impulses propagating antidromically to a large number of neighboring papillae 1~. The area of the cross-connected papillae was measured by electrically stimulating one fungiform papilla and by recording the antidromic impulses from the surrounding neurally connected papillae. The area was usually a circle of 2.5-3.5 mm in diameter. The time which was taken for a stimulus solution to flow the surface area of the whole neurally connected papillae was checked as follows: Both the recording microelectrode filled with 0.3 M KCI and the gustatory stimulator nozzle filled with 0.5 M NaC1 stimulus were placed in the field of the crossconnected papillae as separate as possible, and the time interval between the onset of taste stimulation and the onset of physicochemical junction potential occurring at the microelectrode tip was measured. When the diameter of the cross-connected papillae was a maximum of 3.5 mm, the maximal time interval was as small as 2 ms. This indicates that a taste stimulus flowed without time delay on the whole area of cross-connected papillae under investigation. The tongue surface was always adapted to the flowing Ringer solution whose composition (in mM) was: NaC1 115, KCI 2.5, CaCI 2 1.8, Na2HPO 4 2.15, NaHzPO4 0.85. The pH was adjusted to 7.2 by adding bicarbonate or NaOH. Two min before application of a taste stimulus, the adapting Ringer flow was stopped, and then the tongue surface was covered
335 SSE
~
1
RSE
N
~="AP2 1 HZ
R1 >::a
",'l,,p, ,,~P~ GN ~
N
R2
1~
2 2 ms
1 Hz Ii/~:AP2 q
100 Hz R1
~lr ~ ~w"I AP2
R1 1:12
Fig. 1. Action potentials recorded simultaneously from the papillary nerve fibers and the whole glossopharyngeal nerve in response to anodal electrical stimulation of a fungiform papilla. A: a schematic arrangement of recording and stimulation. SSE, stimulating suction electrode filled with normal Ringer. RSE, recording suction electrode filled with normal Ringer. P, fungiform papilla. N, gustatory nerve fibers. GN, whole glossopharyngeal nerve. R1 and R2, amplifiers for antidromic impulses in the papillary nerve fibers and orthodromic impulses in the glossopharyngeal nerve, respectively. B: action potential from papillary nerve (R1) and glossopharyngeal nerve (R2) evoked by near-threshold stimulus intensity of 0.03 ms pulse at 0.5 Hz. AP1 and AP2, spike potential of low threshold and that of slightly higher threshold, respectively. C: the same as in B except for slight increase in stimulus intensity. D: the same as in C except for stimulus frequency at 100 Hz.
statically with a thin layer of Ringer solution. After taste stimulation the tongue surface was rinsed with Ringer. The onset of the oscilloscope sweep was triggered by current pulses controlling electromagnetic valves by which the nozzles of the gustatory stimulator were opened or closed. After chemical stimulation was applied to the tongue, the latent period of gustatory neural impulses was measured as the time interval between the stimulus touching artifact and the first impulse evoked in a gustatory nerve fiber. All the experiments were carried out at a room temperature of 22-26 °C. RESULTS
Characteristics of orthodromic and antidromic impulses in gustatory nerve fibers Following an electrical stimulation of a fungiform papilla, we could record both antidromic neural impulses from another neighboring fungiform papilla
and orthodromic impulses from a glossopharyngeal nerve. A n example is shown in Fig. 1. In the upper trace (R1) of Fig. 1B, which was recorded antidromically from the papillary nerve fibers (Fig. 1A), a stimulus intensity of near-threshold produced two biphasic spike potentials from two different fibers. The spike marked AP1 was produced by slightly lower threshold than that for the other spike marked AP2. In a lucky case as in this preparation, recording (R2) from the whole glossopharyngeal nerve also showed two biphasic orthodromic spikes, which corresponded to the two antidromic spikes in the papillary nerve fibers (R1). These nerve fibers fired antidromically or orthodromically when the surrounding fungiform papillae were stimulated with 1 m M CaCI 2 and deionized water, so that the fibers were presumed to be gustatory ones. In Fig. 1C, the papilla was electrically stimulated with a slightly increased intensity at 1 Hz. The compound action potentials composed of two kinds of spikes (AP1 and AP2) were seen in both the papillary nerve and the glossopharyngeal nerve. Even when the stimulus frequency was increased to 100 Hz, neither failure of the biphasic action potentials occurred in both nerves (Fig. 1D), nor the shape, height and latency of the action potentials changed. However, at the frequency of~ several hundred Hz the action potentials sometimes failed to appear and the biphasic shape sometimes changed into the monophasic. Electrical threshold for eliciting a neural impulse was lower when the top of the papilla was anodally stimulated than when the top was cathodally stimulated. The threshold of cathodal and anodal currents depended largely on the spike amplitudes: the smaller the amplitude, the higher the threshold. The difference in the threshold between cathodal and anodal currents was 0.61 _+ 0.08 p A (21 units) when a pulse duration was 0.03 ms. The ratio of cathodal to anodal current threshold was 1.32 _+ 0.05 (21 units). Since about 10 afferent fibers of varying diameters are present in the fungiform papilla 17'1s, the various amplitudes of antidromic impulses (peak to peak, 15-180 pV) were elicited in the neurally connected papilla by electrical and chemical stimulation of the neighboring papillae. In each gustatory unit, the maximal voltage deviation from the mean magnitude of the impulses was a range of +3.6/2V, which corre-
336 Suction electrode
R
A
inger layet ~
i
u
I
~
!
s
Nozzle of gustatory stimulator
~ drOO Of taste stimulant
t
i t ' l l f ~ilviiiilll
0.5 M sucrose o~,
0.5 M NaCI ee
E
I o"=
I° •
3 mM acetic acid
~ 9 ms
100 ms
F
1 mM oQ'HCIo
i
>=
>=
100 ms
Deionized water
400 ms
G
IS 100 ms
reached the Ringer solution layer covering the tongue surface, an electrical surge as physicochemical p h e n o m e n o n a p p e a r e d ( m a r k S in Fig. 2 B - G ) . A train of impulses in single gustatory neural units evoked by application of 0.5 M NaC1, 3 m M acetic acid, 1 mM Q-HCI, 0.5 M sucrose, 1 m M CaCI 2 and deionized water were shown in Fig. 2 B - G . In each record, the filled circles indicate the first impulses elicited in different gustatory units. Fig. 3 illustrates latencies of gustatory impulses of 50-171 units in response to the 4 basic stimuli, 1 m M CaCI 2 and deionized water. The mean latency was 158 ms for 0.5 M NaCt, 218 ms for 3 mM acetic acid, 207 ms for 1 m M Q-HC1 and 557 ms for 0.5 M sucrose. The mean latencies for 1 m M CaCI 2 and deionized water were 441 ms and 2.41 s, respectively. The latencies for NaCI, acetic acid and Q-HCI were relatively shorter, but those for the sucrose, CaC12 and water were longer and distributed broadly. Application of much stronger stimuli of 2 M NaCI, 0.1 M acetic acid, 0.1 M Q-HC1 and 2 M sucrose and
2s
Fig. 2. Antidromic impulses in single gustatory neural fibers evoked by chemical stimulation of the tongue. A: schematic drawing of electrical recording and chemical stimulation. B-G: gustatory impulses of papillary nerve fibers evoked by 4 basic taste stimuli, CaC12 and deionized water. S, electrical surge showing just a contact of stimulant on the tongue surface. Closed circles denote the first impulses in each unit.
30 - A
05 M NaCI n= 96
t l
0003~ acetic acid =
20 -
I
sponded to the maximal level of the white noises (+3.7 ¢tV) in the recording systems used. In other words, the percentage of the maximal deviation from the mean impulse magnitude decreased exponentially with increasing impulse height or fiber diameter. Very strong chemical stimulation of the tongue with 0.1 M Q-HC1, 2 M sucrose, 0.1 M acetic acid, 2 M NaCI and 1 M CaC12 also p r o d u c e d antidromic gustatory impulses of + 3 . 6 / ~ V deviation in every unit of the papilla. These observations indicate that the amplitude of impulses originating from single gustatory units of any diameters is physiologically constant.
Latency of gustatory neural impulses in response to 4 basic st°muff, CaCI2 and water In Fig. 2 A is shown a schematic arrangement for recording antidromic impulses and for applying taste stimulation. W h e n the first drop of a taste stimulant
[-h °.°°, n=125
[
I
I
I
I
200 400 600 800 1000 1200 ms
°-.c.
0.5 M sucrose =
lO
I 30
-E
o.oo1 M CaCI 2 n= 171
10
0
i
0
i
I
I
i
I
I
i
I
I
)
O 400 800 1200 1600 2000 2400ms F
Deionized water n=50
t
200 400 800 800 ms 0 Latency
2
4
6
I 8
t 10
I 12 s
Fig. 3. Histograms of latencies of gustatory neural impulses elicited by 4 basic taste stimuli, CaCI 2 and deionized water. The mean and S.E.M. was 158 + 11 ms for 0.5 M NaCI (A), 218 + 10 ms for 0.003 M acetic acid (B), 207 + 8 ms for 0.001 M QHCI (C), 557 + 41 ms for 0.5 M sucrose (D), 441 + 18 ms for 0.00l CaCI2 (E) and 2.41 _+0.25 s for deionized water (F).
337 1 M CaCI 2 caused gustatory impulses with shorter latencies of 75 + 7 ms (mean + S.E.M., n = 32), 58 + 13 ms (n = 10), 87 + 7 ms (n = 29), 107 _+ 15 ms (n = 12) and 82 + 13 ms (n = 20), respectively. The shortest latency in the gustatory unit was 8.6 ms for 2 M NaCl, 7.2 ms for 0.1 M acetic acid, 21.6 ms for 0.1 M Q-HCI and 45.5 ms for 1 M sucrose.
oo
oo[ ~
7.0: 5.0 3.0 1.0 0.6 0.4
= Acetic acid
0.4 ! ~ ~ ~ ' x 0.3
n=23 " ! %
0.2 0.1
0.2
0 (3o 0.6 0.4 ~,, 0.2
O.t
~
-
0.3
0
1.0
(~),~','~ Q-HCI
.~
~n=29
!
" \
,
i
30
\\
1'0-2
o.s ,
10-5 10~4 10-3 10-2 / ooj ~(~)~\ 5.0!
1'0-3
10.c~
:~'~~ =
1()-4
0
Relation between stimulus concentrations and latencies When the tongue surface was adapted to a normal Ringer solution, in 54% of 13 units examined no impulses were evoked by 0.1 M NaCI, whereas in the remaining 46% the latencies ranged from 0.5 to 5 s (Fig. 4A). Latencies of all 13 units decreased with increasing concentration of NaCI. In case of acetic acid stimulation, 35% of 23 units sampled had an infinite latency to 10-4 M, and 22% had an infinite latency to 10-3 M (Fig. 4B). The threshold for acetic acid in 65% of the units was lower than 10 -4 M. All the units showed the decrease in latency with increasing concentration. In case of Q-HCI stimulation, 65.5% of 29 units had no responses to 10-5 M. The remaining 34.5% units had a latency of 0.2-0.5 s at 10-5 M and showed a decrease in that with increasing concentration (Fig. 4C). Stimulation with sucrose did not produce any impulse in 80% of 25 units at 0.1 M and 52% at 0.3 M. Twenty percent of the units responded to 0.1 M sucrose with a long latency of 2 - 9 s (Fig. 4D). The latencies for all the sucrose-sensitive units also gradually decreased with an increase in concentration. With CaCI2, 32% and 11% of 19 units examined had an infinite latency at 10-5 and 10-4 M, respectively. The other units of 68% responded to 10-5 M CaC12 with latencies of 0.2-1.3 s (Fig. 4E).
Relation of diameter of neural fibers to latency
011
0.3
1.0
CaCI2
n=t,
1.0" ~ 0.8 0.6 O.4 0.2 0 lO-S 1'o-4 l'O-a 1'0-2 Concentration
Simultaneous recording from several papillary fibers within a single fungiform papilla after chemical stimulation of the neighboring fungiform papillae frequently showed that there was a linear relation between the amplitudes of gustatory neural impulses and the latencies (Fig. 5). The smaller the spike amplitude, the shorter the latency of the spike evoked by a taste stimulation. When stimulated with 6 taste stimuli, typical examples of these relations are illustrated in Fig. 5. In gustatory neural fibers there is a linear relation between the spike amplitude and the fiber diameter 15, so it is said that the smaller the fiber diameter, the shorter the spike latency.
(M)
Fig. 4. Relation between concentrations of 4 basic taste stimuli
Conduction time of impulses
and CaCI2 and lateneies of impulses in gustatory nerve fibers. Experimental method is shown in Fig. 2A. When no impulses were initiated during 10 s after application of lower concentrations of taste stimuli, the unit under investigation was regarded os havingan infinite latency.
The conduction times of impulses traveling along gustatory nerve fibers connecting between the fungiform papillae were calculated (inset in Fig. 6C,D). In an example shown in Fig. 6A and B, 3 connected fi-
338 500
A
NaCI
D
Acetic acid
E
sucrose
bers were found by electrical stimulation of various intensities. Supramaximal stimulation p r o d u c e d 3 spikes m a r k e d as O , • and II, whose latencies were shorter to anodal stimulation (Fig. 6A) than to cathodal stimulation (Fig. 6B). In a total of 40 connected fibers examined, the m e a n latency of their impulses was 2.1 ms to anodal stimulation (Fig. 6C), but 2.4 ms to cathodal stimulation (Fig. 6D). M e a n difference between the two latencies was 0.3 ms.
250
i
0 500
/?
B
E 250
e/
CaCI2
Y
Arrival time of taste stimulus on taste receptor membrane I
0 500
C
Q-HCI
oo
F
/B
H20
, o £ soY
,;o
,oo
40
Amplitude of gustatory neural impulse (~uV)
Fig. 5. Relation between latencies of gustatory neural impulses and their magnitudes. The taste stimuli used were 0.5 M NaCI, 0.003 M acetic acid, 0.001 M Q-HC1, 0.5 M sucrose, 0.001 M CaCIz and deionized water. The data of the same symbols on each straight line were obtained from gustatory nerve fibers of single fungiform papillae.
Fluctuating latency of neural impulses
riSE SSE
A
15
+
;~,
N
'
lo
~
NN
C
~
>
O,
2~
RSE SSE
z
O;
0
1
2
3
4
5
Before application of a taste stimulus, the tongue surface was covered stationarily with a thin layer of Ringer solution. The thickness of the Ringer layer measured from the top surface of the fungiform papillae was 56 _+ 8/~m (12 papillae). A f t e r a taste stimulus touched the surface of Ringer solution as in Fig. 2A, the time when the stimulus reached the surface of the taste disk situated at the top of the fungiform papillae was estimated by measuring the physicochemical junction potential between an applied 2-M NaCI solution and a microelectrode tip put on the papilla surface covered with much mucus. The mean arrival time so m e a s u r e d was 20 + 3 ms (48 papillae) with a range of 4 - 5 8 ms.
6
Latency (ms)
Fig. 6. Latencies of neural impulses evoked by single electrical stimulation of a fungiform papilla. Schematic arrangement of stimulating and recording suction electrodes is shown in the insets of C and D. A and B: examples of 3 action potentials (O, &, II) in different papillary nerve fibers to electrical stimulation of a fungiform papilla with 0.03 ms pulse. The top of the fungiform papilla was stimulated anodally in A or cathodally in B. C and D: histograms of latencies of the neural impulses to anodal (C) and cathodal (D) stimulations.
No antidromic impulses, e v o k e d by weak electrical stimulation of a fungiform papilla, had a fluctuation of latency and amplitude. H e r e we t e r m e d these impulses as the first group of impulses (letter I in Fig. 7 B - D ) . On the other hand, when the stimulus intensity was raised to more than 5.2 times (20 units), there a p p e a r e d the second group of impulses (SGI) (letter II in Fig. 7D) having a considerably fluctuating latency. In case of Fig. 7, SGI a p p e a r e d intermittently in spite of anodal stimulation at 1 Hz. S G I was p r o d u c e d by either anodal or cathodai stimulation of the fungiform papilla, but the threshold was lower when anodally stimulated than when cathodaily stimulated. Even at 0.5 Hz of stimulus frequency, the latency of SGI fluctuated. The mean m i n i m u m and maximum latencies of SGI in 24 neural units, from which the conduction time of the impulses was subtracted, were 5.3 + 0.4 ms and 10.0 --- 0.7 ms, respectively. A f t e r the taste disk situated at the top of the fungiform papillae was mechanically injured by a
339
SSE
RSE
I
f
I
I 2
I
I
3
4
A
N
E
4
-5 O.
B-¢xc-----
E
2
Stimulus
intensity
(pA)
0 I C 0
I
6
C
I
I
I
I
I
f
I
I
B •A
•
JI
~ e
te-e ~
4
..1
I
I
I
l
I
I
I
I
Stimulus d u r a t i o n
Fig. 7. Appearance of the second group of impulses in the papillary neural fibers following anodal strong stimulation of a fungiform papilla. A: arrangement of a stimulating suction electrode (SSE) and a recording one (RSE). P, fungiform papilla. N, gustatory nerve fibers. B - D : recordings of antidromic neural impulses to electrical stimulation at 0.5 Hz with 0.03-ms pulses of different strengthes. In B and C, only the first group of impulses designated I was evoked. Two neurally connected fibers existed between the two papillae. In D, the second group of impulses designated II following the first group was evoked by stimulation of stronger intensity. One spike in the first group was evoked in the record B, but two kinds of spikes in the first group making a compound action potential were evoked in the record C and D. In record D, 20 sweeps were superimposed, but 10 impulses in the second group were evoked intermittently.
pair of fine forceps, SGI disappeared, but the first group of impulses remained unchanged. The latency of SGI changed depending on stimulus intensity: the stronger the stimulus intensity, the shorter the latency (Fig. 8A). Lengthening the duration of an electrical pulse, keeping the intensity constant, did not change the latency of second group (Fig. 8B). SGI appeared intermittently depending on the frequency of repetitive stimulation of the fungiform papilla. In Fig. 9, this intermittent response pattern is shown in 5 units. One unit showed a mostly complete
I
1 1.0
0.5 (ms)
Fig. 8. A: relation between stimulus intensity and latency of the second group of impulses in 3 units. Stimulus duration was 0.1 ms. B: relation between stimulus duration and latency of the second group of impulses in 3 units. Stimulus intensity was 2 /tA. In A and B, the top of the fungiform papillae was anodally stimulated.
failure of second group even at 1 Hz and another unit that at 10 Hz. When the duration of a single electrical pulse was less than 1 ms, one impulse in the second group was evoked in most cases, but two or more impulses were
==
100 •
o
\\
"o o o
5Q.
50
•E_ 0
o
o~
1'0
0 Stimulus frequency (Hz)
Fig. 9. Relation between electrical stimulus frequency and occurrence of the second group of impulses. Data from 5 units. The papillae were anodally stimulated.
340
4 3 "S
2
=E 1
z
0
i
0
10
i
i
i
20
30
40
5O
Stimulus duration (ms)
Fig. 10. Relation between duration of electrical stimulus pulse and number of impulses in a single neural unit. Method of recording and stimulation was the same as in Fig. 7.
produced in some cases. As the stimulus duration was increased to more than 1 ms, the number of the impulses increased gradually (Fig. 10).
DISCUSSION The present experiment, in which the latency from the touch of a taste stimulus on the tongue surface to the first gustatory neural impulse was measured, showed that 2 M NaCI, 0.1 M acetic acid, 0.1 M Q-HCI, 2 M sucrose and 1 M CaCI 2 of very strong stimuli produced a mean latency of 58-107 ms.
Tas?cell
_ ~ Taste Stimulus
"f[~
B
~or./~/l['![
C --
~' potential /
D --
~ ] } /
E-
Suction electrode --
t
II fl II II
/
II : g'form papilla
~L'~'~//Z/Gu~stat~rynerve fiber dl Initiated spike /
Je I I .
/
Conducted spike ]Fig. l]. A possible explanation of mechanism of latency of gustatory neural impulseselicitedby chemicalstimulation.A: ap-
plicationof taste stimulus. B: generation of receptor potential
with latency b. C: generation of postsynaptic potential with latency c after B. D: impulses with latency d initiated at the first node of Ranvier of myelinated gustatory fiber after C. E: conducted spikes with latency e after D.
In Fig. 11, all processes of appearance of gustatory neural impulses by chemical stimulation of the fungiform papillae are illustrated schematically. Gustatory impulses to the chemical stimulation have been found to generate at the first node of Ranvier of myelinated nerve fibers running inside the fungiform papillae 8. Generation of an impulse at the first node of Ranvier may be as follows: arrival of a taste stimulus on the taste receptor membrane --~ depolarization of a taste cell --~ release of a transmitter --~ generation of a postsynaptic potential ~ initiation of a gustatory impulse at the first node of Ranvier. The real time course of these whole processes has not yet been clarified. It has been reported that latency of depolarizing receptor potential in a frog taste cell evoked by application of 0.5 M NaC1 and 15.6 mM acetic acid is 100-300 m s 12'16. Electrical stimulation of the fungiform papilla in a frog produced two kinds of neural impulses termed the first and second groups. The latency of the first group of impulses having a low threshold was shorter and very stable. On the other hand, the second group of impulses had the following characteristics: (1) a high threshold, (2) a long and fluctuating latency, (3) a decrease of latency with increasing stimulus intensity, (4) a failure of impulses at a low stimulus frequency, and (5) a repetitive firing in response to a single pulse of long duration. The two kinds of impulses in the frog fungiform papillae have already been reported by Nomura and Katsuhata 1° and Sato TM.Fluctuation of latency similar to that of the second group of impulses in the present study has been found in the endplate potential of the neuromuscular junction 6'7. Fig. 12 illustrates an assumption of stimulus current flows through a taste cell and a gustatory nerve fiber, when the top of the fungiform papilla was stimulated anodally or cathodally. Judging from the short and stable latency of the low-threshold impulses in the first group, a weak stimulus current probably did not flow the taste cell but mostly flowed the gustatory axon (Ia in Fig. 12). In case of the anodal stimulation the stable latency of an impulse was, on the average, 0.3 ms shorter than that in case of the cathodal stimulation. It is, therefore, likely that the impulse induced by anodal stimulation of the fungiform papilla might be initiated at the third node of Ranvier because this was situated just outside of the fungiform papilla sucked by the stimulating suction electrode (Figs. 11
341
I
ste cell
t~ D o. cO
E
It.
Fig. 12. Stimulus current flows through a gustatory axon (Ia), a taste cell (It) and an extracellular space (Ie) in the fungiform papilla.
and 12) 8, whereas the impulse by cathodal stimulation might be initiated at unmyelinated portion near the first myelin sheath. The fact that the threshold of cathodal current was 1.32 times higher than that of anodal current is taken to indicate that the membrane of the third node of Ranvier possessed higher excitability. Stronger stimulus current might flow through taste cells (It in Fig. 12) and produce a critical depolarization, which in turn might initiate a release of transmitter, an excitatory postsynaptic potential at the axon terminal and an impulse in the second group at the first node of Ranvier. It has been reported that, after chemical blockage of sensory synapses between taste cells and gustatory afferent fibers in the frog, the neural response composed of initial phasic and tonic components to anodal electrical stimulation of the tongue was mostly suppressed 5. This suggests that the response of gustatory nerve fibers by anodal stimulation of the tongue was induced synaptically via an excitation of taste cell 5'19.
In the present experiment, mechanical destruction of the taste disk situated at the summit of the fungiform papillae did not elicit the second group of impulses by anodal stimulation of the papilla, but still initiated the first group of impulses. This also suggests that the second group of impulses was synaptically elicited via a depolarization of taste cells by currents. Since it is suggested that a single papillary nerve fiber in the frog tongue branches many times and innervates a great number of taste cells 18, it is probable that the distances from the afferent terminals to the first node of Ranvier of a myelinated nerve fiber are diverse. Therefore, synaptic potentials elicited in the postsynaptic membranes following a depolarization of taste cells might spread, with various time courses, to the first node of Ranvier. It is assumed that the decayed synaptic potential of slow rise time slowly induces a critical depolarization at the impulse firing zone, so that the latency of the first gustatory impulse induced by anodal stimulation might take as long as 5 ms. From the data of Nomura and Sakada 9 who first recorded compound synaptic potental generated in the gustatory afferent terminals of the frog tongue, we can calculate the peak time of 10-25 ms in the gustatory synaptic potentials. By strong gustatory stimulation with very high concentrations, we found a latency of 58-107 ms in the gustatory neural impulses. These values include (i) the mean arrival time of a stimulus to receptor membrane (20.1 ms), (ii) the mean minimal delay (5.3 ms) from a depolarization of taste cell to an impulse initiation, and (iii) the mean conduction time of the impulse (2.4 ms). The percentage of a total of 27.8 ms in chemically evoked impulse latency of 58-107 ms is 26-48%. Therefore, the 52-74% (30-79 ms) of the latency elicited by strong chemical stimuli is due to the true latency between the arrival of a chemical stimulus to the receptor membrane and the onset of the receptor potential.
ACKNOWLEDGEMENTS This study was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan. We thank Miss Y. Kitada for her excellent assistance in preparing this paper.
342 REFERENCES 1 Beidler, L.M., Properties of chemoreceptors of tongue of rat, J. Neurophysiol., 16 (1953) 595-607. 2 Beidler, L.M., Physiological properties of mammalian taste receptors. In G.E.W. Wolstenholme and J. Knight (Eds.), Ciba Foundation Symposium on Taste and Smell in Vertebrates, Churchill, London, 1970, pp. 51-70. 3 Faull, J.R. and Halpern, B.P., Taste stimuli: time course of peripheral nerve response and theoretical models, Science, 178 (1972) 73-75. 4 Halpern, B.P. and Marowitz, L.A., Taste responses to lickduration stimuli, Brain Research, 57 (1973) 473-478. 5 Kashiwayanagi, M., Yoshii, K., Kobatake, Y. and Kurihara, K., Taste transduction mechanisum: similar effects of various modifications of gustatory receptors on neural responses to chemical and electrical stimulation in the frog, J, Gen. Physiol., 78 (1981) 259-275. 6 Katz, B. and Miledi, R., The effect of temperature on the synaptic delay at the neuromuscular junction, J. Physiol. (London), 181 (1965) 656-670. 7 Katz, B. and Miledi, R., The release of acetylcholine from nerve endings by graded electric pulses, Proc. R. Soc. London B, 167 (1967) 23-38. 8 Miyamoto, T., Mineda, T., Okada, Y. and Sato, T., Initiation site of gustatory neural impulses in frog tongue, Jpn. J. Oral Biol., 27 (1985) 389-391. 9 Nomura, H. and Sakada, S., Local potential changes at sensory nerve fiber terminals of the frog tongue. In C. Pfaffmann (Ed.), Olfaction and Taste III, Rockefeller University Press, New York, 1969, pp. 345-351.
10 Nomura, H. and Katsuhata, T., Responses of a chemoreceptor to mechanical and electrical stimulation in the frog tongue, Bull. Tokyo Dent. Coll., 14 (1973) 169-193. 11 Rapuzzi, G. and Casella, C., Innervation of the fungiform papillae in the frog tongue, J. Neurophysiol., 28 (1965) 154-165. 12 Sato, T., Multiple sensitivity of single taste cells of the frog tongue to four basic taste stimuli, J. Cell Physiol., 80 (1972) 207-218. 13 Sato, T., Adaptation of primary gustatory nerve responses in the frog, Comp. Biochem. Physiol., 43A (1972) 1-12. 14 Sato, T., Latency of taste nerve signals in frog (Rana catesbeiana), Experientia, 32 (1976) 877-879. 15 Sato, T. and Sugimoto, K., The adaptation of the frog tongue to bitter solutions: enhancing effect on gustatory neural response to salt stimuli, Comp. Biochem. Physiol., 62A (1979) 965-981. 16 Sato, T., Recent advances in the physiology of taste cells, Prog. Neurobiol., 14 (1980) 25-67. 17 Sato, T., Ohkusa, M. and Sugimoto, K., Incremental conduction velocity of single afferent fibers innervating frog taste organ, Jpn. J. Physiol., 30 (1980) 655-658, 18 Sato, T., Ohkusa, M., Okada, Y. and Sasaki, M., Topographical difference in taste organ density and its sensitivity of frog tongue, Comp. Biochem. Physiol., 76A (1983) 233-239. 19 Yamamoto, T., Yuyama, N. and Kawamura, Y., Responses of cortical taste cells and chorda tympani fibers to anodal DC stimulation of the tongue in rats, Exp. Brain Res., 40 (1980) 63-70.