- Email: [email protected]
Optical Recording of Neural Signals Evoked by Greater Superficial Petrosal Nerve Stimulation in Rat
Comp. Biochem. Physiol. Vol. 117A, No. 2, pp. 183–190, 1997 Copyright 1997 Elsevier Science Inc.
ISSN 0300-9629/97/$17.00 PII S0300-9629(96)00238-1
Optical Recording of Neural Signals Evoked by Greater Superficial Petrosal Nerve Stimulation in Rat Mamiko Yanaura, Satoshi Yamada, Satoru Shiono, and Michio Nakashima Advanced Technology R & D Center, Mitsubishi Electric Corporation, Amagasaki, Hyogo 661, Japan ABSTRACT. Electrical responses to greater superficial petrosal (GSP) nerve stimulation in a rat geniculate ganglion (GG) preparation were assessed by simultaneous multi-site optical recording. The GG/GSP nerve preparations were dissected out and were stained with a voltage-sensitive dye (RH155). Application of depolarizing square pulses to the GSP nerve fibers using a suction electrode evoked optical (absorbance) signals that were recorded simultaneously from many contiguous regions using a 24 3 24 photodiode matrix array with 448 active elements. Those optical signals were observed along the left half area of the GSP nerve. As the distance from the site of stimulation increased, the optical signals appeared to conduct with increasing time-delay. From the relationship between the peak latency and distance, the conduction velocity was estimated to be about 0.4 m/s. Tetraethylammonium affected the duration of the optical signals, and the signals disappeared in solutions containing tetrodotoxin (TTX) or in Na1-deficient solutions. The optical signals evoked by the GSP nerve stimulation are considered to be due to the action potentials propagating along the GSP of unmyelinated axons. comp biochem physiol 117A;2:183–190, 1997. 1997 Elsevier Science Inc. KEY WORDS. Geniculate ganglion, greater superficial petrosal nerve, gustatory system, optical recording, rat, voltage sensitive dye, electrical stimulation, conduction velocity, in vitro preparation
INTRODUCTION The geniculate ganglion (GG) is a small cranial sensory ganglion, locating dorsally to the inner ear at the junction of the facial nerve and greater superficial petrosal (GSP) nerve. The ganglion contains taste sensory neurons. The peripheral processes of these neurons innervate the palate via the GSP nerve, and the anterior dorsal tongue via the chorda tympany (CT) nerve (2,4,17). Recent behavioral and electrophysiological experiments have shown that the GSP nerve conveys important taste information (9,10) as well as the CT nerve (1,6,11,20,25). However, so far there have been a limited number of physiological studies on the GSP nerve (19), presumably because of the difficulty to detect its neural activities. On the other hand, multi-site optical recording with a voltage-sensitive dye has been shown to be a powerful tool with a number of advantages over electrophysiological recordings, allowing simultaneous and two-dimensional detection of neural activity (7,14,16,18,22,26). Optical recording is based on the finding that the absorption or fluorescence intensity of certain membrane-bound dyes is dependent on the transmembrane potential (5,8,21,23). London reported its first application to taste perception reAddress reprint request to: M. Yanaura, Mitsubishi Electric Corporation, Advanced Technology R & D Center, 8-1-1 Tsukaguchi-Honmachi, Amagasaki, Hyogo 661, Japan. Tel. 81-6-497-7066; Fax 81-6-497-7294. Received 13 December 1995; accepted 27 June 1996.
lated neural events, measuring optically spatio-temporal activity from the hamster gustatory cortex (15). This report describes the first application of the multisite optical recording to the peripheral gustatory nerve, attempting to detect neural signals from a rat in vitro GG preparation with the GSP nerve. By applying electrical stimulation to the GSP nerve, we could successfully detect action potential signals propagating along the GSP nerve of the GG/GSP nerve preparation. MATERIALS AND METHODS Preparations The experimental animals were adult female Sprague-Dawley rats weighing about 200 g. Animals were anesthetized by intraperitoneal injection of urethane (1.8 g/Kg). Following a tracheotomy, a GG/GSP nerve preparation (see Fig. 6B) was carefully dissected out by the technique reported by Harada and Smith (9). Briefly, an incision was made ventrally along the angle of the right mandible. The tympanic bulla was revealed, and a part of its ventral wall was removed to expose the inside of the bulla. The bone overlying the GG and the GSP nerves was removed carefully. Finally, the GG was then dissected out from the VIIth nerve with as long a piece of the GSP nerve as possible. The isolated GG/GSP nerve preparation of about 3 mm in length with a thickness of less than 1 mm was attached to the Sylgard (Dow Corning) bed of an experimental
184
chamber with tungsten pins. Its exposed surface facing to the tongue by drilling away the overlaying bone was fixed with an upside-down position. The preparation was constantly superfused by a bathing solution containing (in mM) 124 NaCl, 5 KCl, 1.3 MgSO4, 1.24 KH2PO4, 26 NaHCO3, 2.6 CaCl2, and 10.5 glucose, except during optical recording trials. The solution was equilibrated with oxygen at 22°C. The isolated preparation was stained for 20 min with 0.2 mg/ml of a voltage-sensitive pyrazolone-oxonol dye, RH155 (Nippon Kankoh Shikiso, also known as NK3041) dissolved in physiological solution. The preparation was then washed out with several changes of the bathing solution. About 10 min before optical recording trials, the bathing solution was changed to a solution containing 1 mM tetraethylammonium (TEA), in order to prolong the duration of action potentials unless otherwise stated. The perfusion was stopped during a recording trial. A cover glass was placed over the GG/GSP nerve preparation during a recording trial. Since the membrane-bound dye molecules gradually diffused out of the bathing solution, the preparation was restained for 10 min with the above dye solution after a set of optical recording trials. Electrical stimulations to the GSP nerve were made with a suction electrode by applying 5-200 µA depolarizing square-current pulses with a duration of 0.3 ms at 4 Hz, since the isolated size of the preparation was small (about 3 mm in length). One optical recording trial typically lasted for 10 sec. Therefore, 40 stimulating pulses were given in total during a recording trial. All experiments were carried out at room temperature (12).
Optical Recording In our optical recording, we detected the absorbance changes in light intensity transmitted through the GG/GSP nerve preparation. The optical recording system used in this experiment was described in our earlier report (18). Briefly, a Zeiss microscope equipped with a lamp housing (26 V 3 300 W tungsten-halogen lamp, Kondo-Silvania) and a bright-field condenser (0.3 n.a.) was used. The light was passed through a heat reflecting filter and a 720 6 25 nm interference filter before illuminating a stained preparation. Objective magnification of 203 (Epiplan-Neofluar, 0.3 n.a., Zeiss) was usually chosen to monitor neural signals. The transmitted light, passing through the preparation, was detected by a 24 3 24 photodiode matrix array with 448 active elements (Hamamatsu Photonix). One element covered an objective plane area of 46.5 3 46.5 µm2 at 203 magnification. The amplified output of each element was digitized at 1 ms intervals and stored on a magnetic disk of a HewlettPackard HP1000 micro 29 minicomputer (Yokogawa-Hewlett-Packard). Data acquisition, analysis, and display were done with programs written in Fortran.
M. Yanaura et al.
RESULTS Optical Response in the Rat GG/GSP Nerve As an example of GG/GSP nerve optical recording trials, Fig. 1 shows optically detected responses elicited by electrical stimulation to the rat GSP nerve bundles by applying 50 µA–0.3 ms–4 Hz negative square current pulses with a suction electrode (an electrical intensity of 10V). Figure 1 is an array representation, allocating 448 optical traces to corresponding photodiode positions (see Fig. 5 for an extended optical trace). Each of the 448 traces shows the optical absorbance change obtained by averaging optical responses to 40 electrical stimuli in a single recording trial. Optical signals were observed in the several vertical columns of photodiode elements covering a part of the GSP nerve. Those optical signals collected in each photodiode were estimated as total activities of nerve fibers or cell bodies within an area covered at its photodiode. On the other hand, on both sides adjacent to this area, there are traces with large amplitude noise. These traces correspond to the perimeter of the GSP nerve; photodiode elements seeing the perimeter gave much bigger optical changes without neural activities [called as vibration noise, (12,18)]. It was easy to evoke the action potentials when the cut end of the GSP nerve was stimulated beyond a threshold (see Figs. 2 and 3). Optical recording trials were carried out, varying the strength of the stimulating current pulse but fixing its duration at 0.3 ms (data not shown). The amplitude of optical signals and the size of response area were found to be dependent on the current strength. When the strength was smaller than 10 µA (1.5–2.0 V), no optical signal appeared. Very small signals were detected in much narrower area by 10 µA current. On the other hand, current strength of 50 or 100 µA induced maximum signal amplitude, although the current strength of more than 150 µA showed a tendency of decreasing the optical responses. To see optical responses from the whole area of the preparation, we carried out measurements on three contiguous areas of the preparation by sliding the 448-element photodiode array (see Fig. 6B, for a sketch of the preparation superimposed on the three positions of the photodiode array). As shown in Fig. 2, optically detected signals are observed mainly along the left half area of the GSP nerve from peripheral (top in this figure) to its central end (lower left) which is ascending to the nucleus of solitary tract (NST). The peak amplitude of the optical signals was found to be different from each other, depending on the observed positions along the GSP nerve fibers. In contrast there were no detectable signals in the GG, although cell bodies of the GG were clearly stained with the dye (see Discussion). Effect of Tetrodotoxin and Na1 -deficient Solutions If the optical signals reflect voltage-activated Na1 -dependent action potentials, they should be blocked by the Na1 channel blocker tetrodotoxin, TTX. We therefore exam-
Optical Recording of GSP Nerve Signals
185
FIG. 1. Traces of a 448-detector optical recording from a rat GG/GSP nerve preparation. Each of 448 traces shows the averaged
optical change recorded by a corresponding photodiode element. Forty trials were averaged. The GSP nerve was stimulated by applying 50 mA–0.3 ms–4 Hz depolarizing square current pulses with a suction electrode. In the lower right corner, the length of the vertical line represents a fractional change in intensity, DI, divided by the resting intensity, I.
ined the effect of TTX on the evoked optical signals, as shown in Fig. 3. Twenty µM TTX completely abolishes optical signals responding to the electrical stimulation (24). The optical signals also disappeared in a Na 1-deficient bathing solution. Figure 4A summarizes the results of experiments to see the effect of the Na 1-deficiency. We performed three optical recording trials and used 26 mM Na 1 concentration in the second trial instead of the normal Na 1 concentration. Figure 4A shows the change in signal amplitude thus obtained. Error bars represent standard deviations (SD) of four preparations. The optical signals were completely blocked in the Na 1-deficient solution and recovered fairly well to the initial amplitude in the normal bathing solution (the third trial). Significance of the analyzed data
was tested by Student’s t-distribution. No significance was observed between the first and third trials (p , 0.2), although the third trial in the experimental case (Fig. 4A) yielded smaller signals than the first trial. Separate sets of three optical recording trials were carried out in the normal bathing solution as control experiments, because the signal amplitude always gave non-negligible trial-to-trial variation, presumably due to the dye bleaching and restaining (see Methods). Figure 4B shows the results of these experiments (control). The responses in the second and third trials were likely to be the biggest and smallest, respectively (see Discussion). However, no significance was obtained between them (p , 0.1 for the first and second trials, p , 0.2 for the first and third trials).
186
M. Yanaura et al.
FIG. 2. Optical recordings from the whole area of the GG/GSP nerve preparation. Three measurements were done by sliding the 448-element photodiode array and were linked together (see Fig. 6B). Each trace shows the averaged optical change. The GSP nerve was stimulated by applying 100 mA–0.3 ms–4 Hz depolarizing square current pulses. Optical signals corresponding to action potentials are shown along the GSP nerves (see text). Top right shows scale bars; a timing of stimulation onset, time scale of 250 ms (horizontal) and fractional changes in intensity (DI/I), and the recording area covered by one photodiode element (46.5 3 46.5 mm2).
Effect of TEA We investigated the effect of TEA. In the optical recordings described so far, 1 mM TEA was added to the bathing solution to prolong the duration of action potentials. Figure 5 illustrates the comparison of the averaged optical signals in the presence (solid line) and in the absence of TEA (dotted line) (we tried to compare non-averaged signals, but the
signal in the absence of TEA was too small for comparison). Figure 5 indicates that the averaged signal in the presence of TEA was almost two times wider than that in the absence of TEA. Less importantly, the averaged signal in the presence of TEA is bigger than that in the absence of TEA. The apparently smaller signal in the absence of TEA simply reflects the fact that the sampling time of the optical recording system (1 ms) was not sufficiently short to pick up
Optical Recording of GSP Nerve Signals
FIG. 3. Effect of tetrodotoxin (TTX) on the optical signals. TTX at 20 mM was added to the bathing solution. Optical signals were completely blocked by TTX. Other experimental conditions and the preparation were the same as in Fig. 1.
187
M. Yanaura et al.
188
FIG. 4. Effect of Na1 -deficient solution. Choline chloride re-
placed NaCl in the normal bathing solution. We compared signals which were obtained from same spots on the preparations, and we calculated relative responses. A: Relative responses in the normal bathing solution and in the Na1-deficient solution. Some signals (n . 4) were compared between three topical recording trials in which signals were obtained from same spots on the preparation, and signals were normalized to the first trial. Experimental conditions were the same between the three trials except for the Na1 concentration. Four preparations were examined and error bars show SDs. The optical signals were completely blocked in the Na1 -deficient solution and recovered in the normal bathing solutions. B: Control experiments. Relative responses were calculated in the same way. Four preparations were examined.
the signal shape of action potentials in the absence of TEA. Conduction Velocity Optical signals responding to the electrical stimulation appeared with increasing time-delay along the GSP nerve,
FIG. 5. Effect of tetraethylammonium (TEA) on the optical
signals. Traces show the averaged optical signals in the presence of TEA (solid line) and in the absence of TEA (dotted line). The signals in the presence of TEA are wider and bigger than in the absence of TEA. Intensity of the stimulation was 100 mA. Other experimental conditions were the same as in Fig. 1.
presumably reflecting the propagation of action potentials. The peak time of a signal became later as the distance of the photodiode element detecting the signal increased from the site of stimulation. We, however, could not discriminate the difference of peak time between detectable signals in one horizontal row of the photodiodes, because of the time resolution of 1 ms. In Fig. 6A, the peak latency of optical signals are plotted against the distance. The peak latency was calculated as the average peak time of all detectable signal in one horizontal row of the photodiode array, while the abscissa represents the objective-plane distance between this horizontal row and the top row. Figure 6A shows two experimental results using a single preparation: one was performed using 203 objective magnification and moving the preparation into three positions (Fig. 6B), and the other using 103 objective magnification (not shown here). The open and closed symbols represent the results of the former and latter experiments, respectively. These two sets of data points fell fairly well on a single straight line (correlation, r 5 0.94). This linear relationship indicates that the impulse elicited by the electrical stimulus conducts at an uniform rate longitudinally along the GSP nerve bundle. From the reciprocal of the slope of the straight line, the conduction velocity was estimated to be about 0.4 m/s. We also measured conduction velocities with other preparations, and we obtained the conduction velocity of 0.4 6 0.11 m/s (average 6 SD, n 5 7). DISCUSSION In the present multi-site optical recording experiments, we detected optical signals responding to the GSP nerve stimulation (Fig. 1). The optical signals were completely blocked by the application of 20 µM TTX (24) and disappeared in the Na 1-deficient bathing solution (Fig. 3), and their peaks were widened by the application of 1 mM TEA. These observations indicate that the optical signals represent action potentials. The optical signals in the absence of TEA (Fig. 5) apparently look wider than usual action potentials obtained by electrophysiological studies (3). In addition to the averaging effect, the wider optical signals may be accounted for by the fact that each photodiode element received light concomitantly from many nerve fibers and thus detected their summed response. As for the amplitude of optical signals, we often observed clear optical signals without signal averaging in spite of a limited number of optical traces. The maximum signal to noise ratio was about 3 for a single trial (data not shown). The size of optical signals appeared to be affected by the amount of membrane-bound dye and the activity of a preparation. The membrane-bound dye was observed to decrease by perfusion of the bathing solution and to increase by restaining. On the other hand, the response of the preparation
Optical Recording of GSP Nerve Signals
189
FIG. 6. The peak times of optical signals as a function of distance. A: Plot of the averaged peak times from the stimulation time against the distance. A peak time was calculated as the average peak times of all detectable signal in one horizontal row of the photodiode array. SD were omitted, but the majority of them were less than 61 ms. Data were obtained from optical recordings using objective magnifications of 103 (closed circle), and 203 (open signs; circle: data from upper area in Fig. 6B, square: middle area, and triangle: lower area, respectively). B: Illustrations of the areas where we carried the measurements out on the preparation. Three measurements were done using objective magnification of 203. Distance is shown at right (mm). The preparation was the same as in Fig. 2.
fell off gradually after the dissection and possibly by photodynamic damage (8) during optical recording (Fig. 4). The response increase in the second recording trials was mainly due to the effect of the restaining, while the response decrease in the third trial was mainly due to the photodynamic damage and activity decrease. In Fig. 2 the optical signals were observed along the GSP nerve fibers, suggesting that these optical signals represent action potentials propagating from the peripheral palate to the NST along them. The optical signals in this figure were obtained from three successive measurements in a single series by sliding the 448-element photodiode array. We could not find the activity decrease due to the photodynamic damage within the single series. The optical signals of bigger spike amplitudes were found at the middle part along the GSP nerves, and small peaks were widely distributed along the left half of the GSP nerves until the cell body region located at the GG. This result suggests the possibility that a diameter of each fiber might become smaller near the cell body region, or fibers might spread widely in their horizontal row. Another explanation may be possible that small peaks might reflect a decrease in absorbance intensity through the GG/GSP nerve preparation depending on the supporting structure overlaying the GG/GSP nerves. The optical recording allowed us to detect the firing times of propagating action potentials at a large number of adjacent points on the GSP nerve, and to measure its conduction velocity. The conduction velocity was estimated to be
about 0.4 m/s, which we believe is a low estimate because the GSP nerve was likely to be not straight but bending in own preparation. Further, the optical measurements were performed at room temperature. It is easily expected that the conduction velocity at room temperature would be slower than that of living animals at about 37°C. Taking into account the estimated conduction velocity and the reported fact that about 70% of the GSP nerve fibers are unmyelinated (17), the optical signals seem to be due to action potentials in the unmyelinated fibers like C fibers (3). The estimated conduction velocity is similar to the optically measured velocity of the unmyelinated fibers in the skate cerebellum (13) and in the embryonic chick vagus nerve (12). We could not get optical signals of unmyelinated part of myelinated fibers (the remaining 30% of the GSP nerve fibers). For the conduction velocity of myelinated fibers (more than 3 m/s), almost the same peak latency is expected against a distance within 3 mm, because of the time resolution (1 ms). This relationship was, however, not found in our optical experiments (see Fig. 6). Although it is important to measure neural activity from a single taste sensory neuron in GG for understanding taste perception, we could not detect any evoked activity from the cell body area of GG. We discuss here two possibilities. First, the bundle structure of nerve fibers and their simultaneous firing give higher signal to noise ratio for the response in the fiber area. On the other hand, in the cell body area the number of overlapping membrane is considered to be
190
much smaller, and simultaneous firing among the cell bodies covered by a photodiode element is less likely to occur, making it the difficult to obtain optical signals. Using other dyes, especially fluorescence dyes, might improve the signal to noise ratio and obtain optical signals from cell bodies (27). Second, the cell soma of the GG might fail to fire action potentials because of lacking a spike initiation zone. Only the nerve fibers might convey taste signals into central nervous system (CNS), without passing through the cell bodies. In this preliminary report, we could show the possibility of detecting neural signals in the GG preparation by multisite optical recording. In order to clarify the functional roles of the CT and GSP nerves, optical recording of GG in vivo should be conducted in the next step. We are grateful to Dr. S. Harada for his technical suggestions and fruitful discussions.
References 1. Beidler, L. Properties of chemoreceptors of tongue of rat. J. Neurophysiol. 16:595–607;1953. 2. Boudreau, J.C.; Bradley, B.E.; Bierer, P.R.; Kruger, S.; Tsuchitani, C. Single unit recordings of the geniculate ganglion of the facial nerve of the cat. Exp. Brain Res. 13:461–488;1971. 3. Brinley, F.J., Jr. Excitation and conduction in nerve fibers. In: Mountcastle, V.B. (ed). Medical Physiology, Vol. 1. St Louis, MO: C.V. Mosby Company; 1980:46–81. 4. Cleaton-Jones, P. A denervation study of the taste buds in the soft palate of the albino rat. Arch. Oral Biol. 21:79–82;1976. 5. Cohen, L.B.; Cohen, A.I.; Hopp, H.-I.; London, J.A.; Schiminovich, D.; Wu, J.-Y.; Xiao, C.; Zecevic, D. Multisite optical measurement of membrane potential: Methodological considerations presented at Eleventh Ann. Meet. European Neurosci. Assoc. Tech. Workshop I:1–21;1988. 6. Fishman, I. Single fiber gustatory impulses in rat and hamster. J. Cell Comp. Physiol. 49:319–334;1957. 7. Grinvald, A.; Cohen, L.B.; Lesher, S.; Boyle, M.B. Simultaneous optical monitoring of activity of many neurons in invertebrate ganglion using a 124-element photodiode array. J. Neurophysiol. 45:829–840;1981. 8. Grinvald, A.; Frostig, R.D.; Lieke, E.; Hildersheim, R. Optical imaging of neural activity. Physiol. Rev. 68:1285 –1366;1988. 9. Harada, S.; Smith, D.V. Gustatory sensitivity of the hamster’s soft palate. Chem. Senses 17:37–51;1992. 10. Harada, S.; Enomoto, S.; Kasahara, Y. Gustatory response of the greater superficial petrosal nerve to l and d-amino acids on the soft palate in the rat. In: Kurihara, K.; Suzuki, N.; Ogawa, H. (eds). Olfaction and Taste XI. Tokyo: SpringerVerlag; 1994:90–91. 11. Jakinovich, W. Methyl 4,6-dichloro-4,6-dideoxy-a-d-agalactopyranoside: An inhibitor of sweet taste responses in gerbils. Science 219:408–410;1983.
M. Yanaura et al.
12. Kamino, K.; Katoh, Y.; Komura, H.; Satoh, K. Multiplesite optical monitoring of neural activity evoked by vagus nerve stimulation in the embryonic chick brain stem. J. Physiol. 409: 263–283;1989. 13. Konnerth, A.; Obaid, A.L.; Salzberg, B.M. Optical recording of electrical activity from parallel fibers and other cell types in skate cerebellar slices, in vitro. J. Physiol. 393:681–702; 1987. 14. London, J.A.; Cohen, L.B.; Wu, J.Y. Optical recordings of the cortical response to whisker stimulation before and after the addition of an epiletogenic agent. J. Neurosci. 9:2182–2190; 1989. 15. London, J.A. Optical recording of activity in the hamster gustatory cortex elicited by electrical stimulation of the tongue. Chem. Senses 15:137–143;1990. 16. Malonek, D.; Tootell, R.B.H.; Grinvald, A. Optical imaging reveals the functional architecture of neurons processing shape and motion in owl monkey area MT Proc. R. Soc. Lond. B 258:109–119;1994. 17. Miller, I.J.; Gomez, M.M.; Lubarsky, H.H. Distribution of the facial nerve to taste receptors in the rat. Chem. Senses Flav. 3:397–411;1978. 18. Nakashima, M.; Yamada, S.; Shiono, S.; Maeda, M.; Satoh, F. 448-Detector optical recording system: Development and application to Aplysia gill-withdrawal reflex. IEEE Trans. Biomed. Eng. 39:26–36;1992. 19. Nejad, M. The neural activities of the greater superficial petrosal nerve of the rat in response to chemical stimulation of the palate. Chem. Senses 11:283–293;1986. 20. Pfaffman, C. Gustatory afferent impulses. J. Cell Comp. Physiol. 17:243–258;1941. 21. Ross, W.N.; Salzberg, B.M.; Cohen, L.B.; Grinvald, A.; Davilla, H.V.; Waggoner, A.S.; Wang, C.H. Changes in absorption, fluorescence, dichronism, and birefringence in stained giant axon: Optical measurement of membrane potential. J. Membrane Biol. 33:141–183;1977. 22. Salzberg, B.M.; Grinvald, A.; Cohen, L.B.; Davila, H.V.; Ross, W.N. Optical recording of neural activity in an invertebrate central nervous system: Simultaneous monitoring of several neurons. J. Neurophysiol: 40:1281–1291;1977. 23. Salzberg, B.M. Optical recording of electrical activity in neurons using molecular probes. In: Barker, J.L.; McKelvy, J.F. (eds). Current Methods in Cellular Neurobiology, Electrophysiological Techniques, Vol. 3. New York: John Wiley & Sons Inc.; 1983:139–187. 24. Sakai, T.; Komuro, H.; Katoh, Y.; Sasaki, H.; Sato, Y.; Kamino, K. Optical determination of impulse conduction velocity during development of embryonic chick cervical vagus nerve bundles. J. Physiol. 439:361–381;1991. 25. Tamar, H. Taste reception in the opposum and bat. Physiol. Zool. 34:86–91;1961. 26. Zecevic, D.; Wu, J.-Y.; Cohen, L.B.; London, J.A.; Hopp, H.-P.; Falk, C.X. Hundreds of neurons in the Aplysia abdominal ganglion are active during the gill-withdrawal reflex. J. Neurosci. 9:3681–3689;1989. 27. Zecevic, D. Multiple spike-initiation zones in single neurons revealed by voltage-sensitive dyes. Nature 381:322–325; 1996.
ISSN 0300-9629/97/$17.00 PII S0300-9629(96)00238-1
Optical Recording of Neural Signals Evoked by Greater Superficial Petrosal Nerve Stimulation in Rat Mamiko Yanaura, Satoshi Yamada, Satoru Shiono, and Michio Nakashima Advanced Technology R & D Center, Mitsubishi Electric Corporation, Amagasaki, Hyogo 661, Japan ABSTRACT. Electrical responses to greater superficial petrosal (GSP) nerve stimulation in a rat geniculate ganglion (GG) preparation were assessed by simultaneous multi-site optical recording. The GG/GSP nerve preparations were dissected out and were stained with a voltage-sensitive dye (RH155). Application of depolarizing square pulses to the GSP nerve fibers using a suction electrode evoked optical (absorbance) signals that were recorded simultaneously from many contiguous regions using a 24 3 24 photodiode matrix array with 448 active elements. Those optical signals were observed along the left half area of the GSP nerve. As the distance from the site of stimulation increased, the optical signals appeared to conduct with increasing time-delay. From the relationship between the peak latency and distance, the conduction velocity was estimated to be about 0.4 m/s. Tetraethylammonium affected the duration of the optical signals, and the signals disappeared in solutions containing tetrodotoxin (TTX) or in Na1-deficient solutions. The optical signals evoked by the GSP nerve stimulation are considered to be due to the action potentials propagating along the GSP of unmyelinated axons. comp biochem physiol 117A;2:183–190, 1997. 1997 Elsevier Science Inc. KEY WORDS. Geniculate ganglion, greater superficial petrosal nerve, gustatory system, optical recording, rat, voltage sensitive dye, electrical stimulation, conduction velocity, in vitro preparation
INTRODUCTION The geniculate ganglion (GG) is a small cranial sensory ganglion, locating dorsally to the inner ear at the junction of the facial nerve and greater superficial petrosal (GSP) nerve. The ganglion contains taste sensory neurons. The peripheral processes of these neurons innervate the palate via the GSP nerve, and the anterior dorsal tongue via the chorda tympany (CT) nerve (2,4,17). Recent behavioral and electrophysiological experiments have shown that the GSP nerve conveys important taste information (9,10) as well as the CT nerve (1,6,11,20,25). However, so far there have been a limited number of physiological studies on the GSP nerve (19), presumably because of the difficulty to detect its neural activities. On the other hand, multi-site optical recording with a voltage-sensitive dye has been shown to be a powerful tool with a number of advantages over electrophysiological recordings, allowing simultaneous and two-dimensional detection of neural activity (7,14,16,18,22,26). Optical recording is based on the finding that the absorption or fluorescence intensity of certain membrane-bound dyes is dependent on the transmembrane potential (5,8,21,23). London reported its first application to taste perception reAddress reprint request to: M. Yanaura, Mitsubishi Electric Corporation, Advanced Technology R & D Center, 8-1-1 Tsukaguchi-Honmachi, Amagasaki, Hyogo 661, Japan. Tel. 81-6-497-7066; Fax 81-6-497-7294. Received 13 December 1995; accepted 27 June 1996.
lated neural events, measuring optically spatio-temporal activity from the hamster gustatory cortex (15). This report describes the first application of the multisite optical recording to the peripheral gustatory nerve, attempting to detect neural signals from a rat in vitro GG preparation with the GSP nerve. By applying electrical stimulation to the GSP nerve, we could successfully detect action potential signals propagating along the GSP nerve of the GG/GSP nerve preparation. MATERIALS AND METHODS Preparations The experimental animals were adult female Sprague-Dawley rats weighing about 200 g. Animals were anesthetized by intraperitoneal injection of urethane (1.8 g/Kg). Following a tracheotomy, a GG/GSP nerve preparation (see Fig. 6B) was carefully dissected out by the technique reported by Harada and Smith (9). Briefly, an incision was made ventrally along the angle of the right mandible. The tympanic bulla was revealed, and a part of its ventral wall was removed to expose the inside of the bulla. The bone overlying the GG and the GSP nerves was removed carefully. Finally, the GG was then dissected out from the VIIth nerve with as long a piece of the GSP nerve as possible. The isolated GG/GSP nerve preparation of about 3 mm in length with a thickness of less than 1 mm was attached to the Sylgard (Dow Corning) bed of an experimental
184
chamber with tungsten pins. Its exposed surface facing to the tongue by drilling away the overlaying bone was fixed with an upside-down position. The preparation was constantly superfused by a bathing solution containing (in mM) 124 NaCl, 5 KCl, 1.3 MgSO4, 1.24 KH2PO4, 26 NaHCO3, 2.6 CaCl2, and 10.5 glucose, except during optical recording trials. The solution was equilibrated with oxygen at 22°C. The isolated preparation was stained for 20 min with 0.2 mg/ml of a voltage-sensitive pyrazolone-oxonol dye, RH155 (Nippon Kankoh Shikiso, also known as NK3041) dissolved in physiological solution. The preparation was then washed out with several changes of the bathing solution. About 10 min before optical recording trials, the bathing solution was changed to a solution containing 1 mM tetraethylammonium (TEA), in order to prolong the duration of action potentials unless otherwise stated. The perfusion was stopped during a recording trial. A cover glass was placed over the GG/GSP nerve preparation during a recording trial. Since the membrane-bound dye molecules gradually diffused out of the bathing solution, the preparation was restained for 10 min with the above dye solution after a set of optical recording trials. Electrical stimulations to the GSP nerve were made with a suction electrode by applying 5-200 µA depolarizing square-current pulses with a duration of 0.3 ms at 4 Hz, since the isolated size of the preparation was small (about 3 mm in length). One optical recording trial typically lasted for 10 sec. Therefore, 40 stimulating pulses were given in total during a recording trial. All experiments were carried out at room temperature (12).
Optical Recording In our optical recording, we detected the absorbance changes in light intensity transmitted through the GG/GSP nerve preparation. The optical recording system used in this experiment was described in our earlier report (18). Briefly, a Zeiss microscope equipped with a lamp housing (26 V 3 300 W tungsten-halogen lamp, Kondo-Silvania) and a bright-field condenser (0.3 n.a.) was used. The light was passed through a heat reflecting filter and a 720 6 25 nm interference filter before illuminating a stained preparation. Objective magnification of 203 (Epiplan-Neofluar, 0.3 n.a., Zeiss) was usually chosen to monitor neural signals. The transmitted light, passing through the preparation, was detected by a 24 3 24 photodiode matrix array with 448 active elements (Hamamatsu Photonix). One element covered an objective plane area of 46.5 3 46.5 µm2 at 203 magnification. The amplified output of each element was digitized at 1 ms intervals and stored on a magnetic disk of a HewlettPackard HP1000 micro 29 minicomputer (Yokogawa-Hewlett-Packard). Data acquisition, analysis, and display were done with programs written in Fortran.
M. Yanaura et al.
RESULTS Optical Response in the Rat GG/GSP Nerve As an example of GG/GSP nerve optical recording trials, Fig. 1 shows optically detected responses elicited by electrical stimulation to the rat GSP nerve bundles by applying 50 µA–0.3 ms–4 Hz negative square current pulses with a suction electrode (an electrical intensity of 10V). Figure 1 is an array representation, allocating 448 optical traces to corresponding photodiode positions (see Fig. 5 for an extended optical trace). Each of the 448 traces shows the optical absorbance change obtained by averaging optical responses to 40 electrical stimuli in a single recording trial. Optical signals were observed in the several vertical columns of photodiode elements covering a part of the GSP nerve. Those optical signals collected in each photodiode were estimated as total activities of nerve fibers or cell bodies within an area covered at its photodiode. On the other hand, on both sides adjacent to this area, there are traces with large amplitude noise. These traces correspond to the perimeter of the GSP nerve; photodiode elements seeing the perimeter gave much bigger optical changes without neural activities [called as vibration noise, (12,18)]. It was easy to evoke the action potentials when the cut end of the GSP nerve was stimulated beyond a threshold (see Figs. 2 and 3). Optical recording trials were carried out, varying the strength of the stimulating current pulse but fixing its duration at 0.3 ms (data not shown). The amplitude of optical signals and the size of response area were found to be dependent on the current strength. When the strength was smaller than 10 µA (1.5–2.0 V), no optical signal appeared. Very small signals were detected in much narrower area by 10 µA current. On the other hand, current strength of 50 or 100 µA induced maximum signal amplitude, although the current strength of more than 150 µA showed a tendency of decreasing the optical responses. To see optical responses from the whole area of the preparation, we carried out measurements on three contiguous areas of the preparation by sliding the 448-element photodiode array (see Fig. 6B, for a sketch of the preparation superimposed on the three positions of the photodiode array). As shown in Fig. 2, optically detected signals are observed mainly along the left half area of the GSP nerve from peripheral (top in this figure) to its central end (lower left) which is ascending to the nucleus of solitary tract (NST). The peak amplitude of the optical signals was found to be different from each other, depending on the observed positions along the GSP nerve fibers. In contrast there were no detectable signals in the GG, although cell bodies of the GG were clearly stained with the dye (see Discussion). Effect of Tetrodotoxin and Na1 -deficient Solutions If the optical signals reflect voltage-activated Na1 -dependent action potentials, they should be blocked by the Na1 channel blocker tetrodotoxin, TTX. We therefore exam-
Optical Recording of GSP Nerve Signals
185
FIG. 1. Traces of a 448-detector optical recording from a rat GG/GSP nerve preparation. Each of 448 traces shows the averaged
optical change recorded by a corresponding photodiode element. Forty trials were averaged. The GSP nerve was stimulated by applying 50 mA–0.3 ms–4 Hz depolarizing square current pulses with a suction electrode. In the lower right corner, the length of the vertical line represents a fractional change in intensity, DI, divided by the resting intensity, I.
ined the effect of TTX on the evoked optical signals, as shown in Fig. 3. Twenty µM TTX completely abolishes optical signals responding to the electrical stimulation (24). The optical signals also disappeared in a Na 1-deficient bathing solution. Figure 4A summarizes the results of experiments to see the effect of the Na 1-deficiency. We performed three optical recording trials and used 26 mM Na 1 concentration in the second trial instead of the normal Na 1 concentration. Figure 4A shows the change in signal amplitude thus obtained. Error bars represent standard deviations (SD) of four preparations. The optical signals were completely blocked in the Na 1-deficient solution and recovered fairly well to the initial amplitude in the normal bathing solution (the third trial). Significance of the analyzed data
was tested by Student’s t-distribution. No significance was observed between the first and third trials (p , 0.2), although the third trial in the experimental case (Fig. 4A) yielded smaller signals than the first trial. Separate sets of three optical recording trials were carried out in the normal bathing solution as control experiments, because the signal amplitude always gave non-negligible trial-to-trial variation, presumably due to the dye bleaching and restaining (see Methods). Figure 4B shows the results of these experiments (control). The responses in the second and third trials were likely to be the biggest and smallest, respectively (see Discussion). However, no significance was obtained between them (p , 0.1 for the first and second trials, p , 0.2 for the first and third trials).
186
M. Yanaura et al.
FIG. 2. Optical recordings from the whole area of the GG/GSP nerve preparation. Three measurements were done by sliding the 448-element photodiode array and were linked together (see Fig. 6B). Each trace shows the averaged optical change. The GSP nerve was stimulated by applying 100 mA–0.3 ms–4 Hz depolarizing square current pulses. Optical signals corresponding to action potentials are shown along the GSP nerves (see text). Top right shows scale bars; a timing of stimulation onset, time scale of 250 ms (horizontal) and fractional changes in intensity (DI/I), and the recording area covered by one photodiode element (46.5 3 46.5 mm2).
Effect of TEA We investigated the effect of TEA. In the optical recordings described so far, 1 mM TEA was added to the bathing solution to prolong the duration of action potentials. Figure 5 illustrates the comparison of the averaged optical signals in the presence (solid line) and in the absence of TEA (dotted line) (we tried to compare non-averaged signals, but the
signal in the absence of TEA was too small for comparison). Figure 5 indicates that the averaged signal in the presence of TEA was almost two times wider than that in the absence of TEA. Less importantly, the averaged signal in the presence of TEA is bigger than that in the absence of TEA. The apparently smaller signal in the absence of TEA simply reflects the fact that the sampling time of the optical recording system (1 ms) was not sufficiently short to pick up
Optical Recording of GSP Nerve Signals
FIG. 3. Effect of tetrodotoxin (TTX) on the optical signals. TTX at 20 mM was added to the bathing solution. Optical signals were completely blocked by TTX. Other experimental conditions and the preparation were the same as in Fig. 1.
187
M. Yanaura et al.
188
FIG. 4. Effect of Na1 -deficient solution. Choline chloride re-
placed NaCl in the normal bathing solution. We compared signals which were obtained from same spots on the preparations, and we calculated relative responses. A: Relative responses in the normal bathing solution and in the Na1-deficient solution. Some signals (n . 4) were compared between three topical recording trials in which signals were obtained from same spots on the preparation, and signals were normalized to the first trial. Experimental conditions were the same between the three trials except for the Na1 concentration. Four preparations were examined and error bars show SDs. The optical signals were completely blocked in the Na1 -deficient solution and recovered in the normal bathing solutions. B: Control experiments. Relative responses were calculated in the same way. Four preparations were examined.
the signal shape of action potentials in the absence of TEA. Conduction Velocity Optical signals responding to the electrical stimulation appeared with increasing time-delay along the GSP nerve,
FIG. 5. Effect of tetraethylammonium (TEA) on the optical
signals. Traces show the averaged optical signals in the presence of TEA (solid line) and in the absence of TEA (dotted line). The signals in the presence of TEA are wider and bigger than in the absence of TEA. Intensity of the stimulation was 100 mA. Other experimental conditions were the same as in Fig. 1.
presumably reflecting the propagation of action potentials. The peak time of a signal became later as the distance of the photodiode element detecting the signal increased from the site of stimulation. We, however, could not discriminate the difference of peak time between detectable signals in one horizontal row of the photodiodes, because of the time resolution of 1 ms. In Fig. 6A, the peak latency of optical signals are plotted against the distance. The peak latency was calculated as the average peak time of all detectable signal in one horizontal row of the photodiode array, while the abscissa represents the objective-plane distance between this horizontal row and the top row. Figure 6A shows two experimental results using a single preparation: one was performed using 203 objective magnification and moving the preparation into three positions (Fig. 6B), and the other using 103 objective magnification (not shown here). The open and closed symbols represent the results of the former and latter experiments, respectively. These two sets of data points fell fairly well on a single straight line (correlation, r 5 0.94). This linear relationship indicates that the impulse elicited by the electrical stimulus conducts at an uniform rate longitudinally along the GSP nerve bundle. From the reciprocal of the slope of the straight line, the conduction velocity was estimated to be about 0.4 m/s. We also measured conduction velocities with other preparations, and we obtained the conduction velocity of 0.4 6 0.11 m/s (average 6 SD, n 5 7). DISCUSSION In the present multi-site optical recording experiments, we detected optical signals responding to the GSP nerve stimulation (Fig. 1). The optical signals were completely blocked by the application of 20 µM TTX (24) and disappeared in the Na 1-deficient bathing solution (Fig. 3), and their peaks were widened by the application of 1 mM TEA. These observations indicate that the optical signals represent action potentials. The optical signals in the absence of TEA (Fig. 5) apparently look wider than usual action potentials obtained by electrophysiological studies (3). In addition to the averaging effect, the wider optical signals may be accounted for by the fact that each photodiode element received light concomitantly from many nerve fibers and thus detected their summed response. As for the amplitude of optical signals, we often observed clear optical signals without signal averaging in spite of a limited number of optical traces. The maximum signal to noise ratio was about 3 for a single trial (data not shown). The size of optical signals appeared to be affected by the amount of membrane-bound dye and the activity of a preparation. The membrane-bound dye was observed to decrease by perfusion of the bathing solution and to increase by restaining. On the other hand, the response of the preparation
Optical Recording of GSP Nerve Signals
189
FIG. 6. The peak times of optical signals as a function of distance. A: Plot of the averaged peak times from the stimulation time against the distance. A peak time was calculated as the average peak times of all detectable signal in one horizontal row of the photodiode array. SD were omitted, but the majority of them were less than 61 ms. Data were obtained from optical recordings using objective magnifications of 103 (closed circle), and 203 (open signs; circle: data from upper area in Fig. 6B, square: middle area, and triangle: lower area, respectively). B: Illustrations of the areas where we carried the measurements out on the preparation. Three measurements were done using objective magnification of 203. Distance is shown at right (mm). The preparation was the same as in Fig. 2.
fell off gradually after the dissection and possibly by photodynamic damage (8) during optical recording (Fig. 4). The response increase in the second recording trials was mainly due to the effect of the restaining, while the response decrease in the third trial was mainly due to the photodynamic damage and activity decrease. In Fig. 2 the optical signals were observed along the GSP nerve fibers, suggesting that these optical signals represent action potentials propagating from the peripheral palate to the NST along them. The optical signals in this figure were obtained from three successive measurements in a single series by sliding the 448-element photodiode array. We could not find the activity decrease due to the photodynamic damage within the single series. The optical signals of bigger spike amplitudes were found at the middle part along the GSP nerves, and small peaks were widely distributed along the left half of the GSP nerves until the cell body region located at the GG. This result suggests the possibility that a diameter of each fiber might become smaller near the cell body region, or fibers might spread widely in their horizontal row. Another explanation may be possible that small peaks might reflect a decrease in absorbance intensity through the GG/GSP nerve preparation depending on the supporting structure overlaying the GG/GSP nerves. The optical recording allowed us to detect the firing times of propagating action potentials at a large number of adjacent points on the GSP nerve, and to measure its conduction velocity. The conduction velocity was estimated to be
about 0.4 m/s, which we believe is a low estimate because the GSP nerve was likely to be not straight but bending in own preparation. Further, the optical measurements were performed at room temperature. It is easily expected that the conduction velocity at room temperature would be slower than that of living animals at about 37°C. Taking into account the estimated conduction velocity and the reported fact that about 70% of the GSP nerve fibers are unmyelinated (17), the optical signals seem to be due to action potentials in the unmyelinated fibers like C fibers (3). The estimated conduction velocity is similar to the optically measured velocity of the unmyelinated fibers in the skate cerebellum (13) and in the embryonic chick vagus nerve (12). We could not get optical signals of unmyelinated part of myelinated fibers (the remaining 30% of the GSP nerve fibers). For the conduction velocity of myelinated fibers (more than 3 m/s), almost the same peak latency is expected against a distance within 3 mm, because of the time resolution (1 ms). This relationship was, however, not found in our optical experiments (see Fig. 6). Although it is important to measure neural activity from a single taste sensory neuron in GG for understanding taste perception, we could not detect any evoked activity from the cell body area of GG. We discuss here two possibilities. First, the bundle structure of nerve fibers and their simultaneous firing give higher signal to noise ratio for the response in the fiber area. On the other hand, in the cell body area the number of overlapping membrane is considered to be
190
much smaller, and simultaneous firing among the cell bodies covered by a photodiode element is less likely to occur, making it the difficult to obtain optical signals. Using other dyes, especially fluorescence dyes, might improve the signal to noise ratio and obtain optical signals from cell bodies (27). Second, the cell soma of the GG might fail to fire action potentials because of lacking a spike initiation zone. Only the nerve fibers might convey taste signals into central nervous system (CNS), without passing through the cell bodies. In this preliminary report, we could show the possibility of detecting neural signals in the GG preparation by multisite optical recording. In order to clarify the functional roles of the CT and GSP nerves, optical recording of GG in vivo should be conducted in the next step. We are grateful to Dr. S. Harada for his technical suggestions and fruitful discussions.
References 1. Beidler, L. Properties of chemoreceptors of tongue of rat. J. Neurophysiol. 16:595–607;1953. 2. Boudreau, J.C.; Bradley, B.E.; Bierer, P.R.; Kruger, S.; Tsuchitani, C. Single unit recordings of the geniculate ganglion of the facial nerve of the cat. Exp. Brain Res. 13:461–488;1971. 3. Brinley, F.J., Jr. Excitation and conduction in nerve fibers. In: Mountcastle, V.B. (ed). Medical Physiology, Vol. 1. St Louis, MO: C.V. Mosby Company; 1980:46–81. 4. Cleaton-Jones, P. A denervation study of the taste buds in the soft palate of the albino rat. Arch. Oral Biol. 21:79–82;1976. 5. Cohen, L.B.; Cohen, A.I.; Hopp, H.-I.; London, J.A.; Schiminovich, D.; Wu, J.-Y.; Xiao, C.; Zecevic, D. Multisite optical measurement of membrane potential: Methodological considerations presented at Eleventh Ann. Meet. European Neurosci. Assoc. Tech. Workshop I:1–21;1988. 6. Fishman, I. Single fiber gustatory impulses in rat and hamster. J. Cell Comp. Physiol. 49:319–334;1957. 7. Grinvald, A.; Cohen, L.B.; Lesher, S.; Boyle, M.B. Simultaneous optical monitoring of activity of many neurons in invertebrate ganglion using a 124-element photodiode array. J. Neurophysiol. 45:829–840;1981. 8. Grinvald, A.; Frostig, R.D.; Lieke, E.; Hildersheim, R. Optical imaging of neural activity. Physiol. Rev. 68:1285 –1366;1988. 9. Harada, S.; Smith, D.V. Gustatory sensitivity of the hamster’s soft palate. Chem. Senses 17:37–51;1992. 10. Harada, S.; Enomoto, S.; Kasahara, Y. Gustatory response of the greater superficial petrosal nerve to l and d-amino acids on the soft palate in the rat. In: Kurihara, K.; Suzuki, N.; Ogawa, H. (eds). Olfaction and Taste XI. Tokyo: SpringerVerlag; 1994:90–91. 11. Jakinovich, W. Methyl 4,6-dichloro-4,6-dideoxy-a-d-agalactopyranoside: An inhibitor of sweet taste responses in gerbils. Science 219:408–410;1983.
M. Yanaura et al.
12. Kamino, K.; Katoh, Y.; Komura, H.; Satoh, K. Multiplesite optical monitoring of neural activity evoked by vagus nerve stimulation in the embryonic chick brain stem. J. Physiol. 409: 263–283;1989. 13. Konnerth, A.; Obaid, A.L.; Salzberg, B.M. Optical recording of electrical activity from parallel fibers and other cell types in skate cerebellar slices, in vitro. J. Physiol. 393:681–702; 1987. 14. London, J.A.; Cohen, L.B.; Wu, J.Y. Optical recordings of the cortical response to whisker stimulation before and after the addition of an epiletogenic agent. J. Neurosci. 9:2182–2190; 1989. 15. London, J.A. Optical recording of activity in the hamster gustatory cortex elicited by electrical stimulation of the tongue. Chem. Senses 15:137–143;1990. 16. Malonek, D.; Tootell, R.B.H.; Grinvald, A. Optical imaging reveals the functional architecture of neurons processing shape and motion in owl monkey area MT Proc. R. Soc. Lond. B 258:109–119;1994. 17. Miller, I.J.; Gomez, M.M.; Lubarsky, H.H. Distribution of the facial nerve to taste receptors in the rat. Chem. Senses Flav. 3:397–411;1978. 18. Nakashima, M.; Yamada, S.; Shiono, S.; Maeda, M.; Satoh, F. 448-Detector optical recording system: Development and application to Aplysia gill-withdrawal reflex. IEEE Trans. Biomed. Eng. 39:26–36;1992. 19. Nejad, M. The neural activities of the greater superficial petrosal nerve of the rat in response to chemical stimulation of the palate. Chem. Senses 11:283–293;1986. 20. Pfaffman, C. Gustatory afferent impulses. J. Cell Comp. Physiol. 17:243–258;1941. 21. Ross, W.N.; Salzberg, B.M.; Cohen, L.B.; Grinvald, A.; Davilla, H.V.; Waggoner, A.S.; Wang, C.H. Changes in absorption, fluorescence, dichronism, and birefringence in stained giant axon: Optical measurement of membrane potential. J. Membrane Biol. 33:141–183;1977. 22. Salzberg, B.M.; Grinvald, A.; Cohen, L.B.; Davila, H.V.; Ross, W.N. Optical recording of neural activity in an invertebrate central nervous system: Simultaneous monitoring of several neurons. J. Neurophysiol: 40:1281–1291;1977. 23. Salzberg, B.M. Optical recording of electrical activity in neurons using molecular probes. In: Barker, J.L.; McKelvy, J.F. (eds). Current Methods in Cellular Neurobiology, Electrophysiological Techniques, Vol. 3. New York: John Wiley & Sons Inc.; 1983:139–187. 24. Sakai, T.; Komuro, H.; Katoh, Y.; Sasaki, H.; Sato, Y.; Kamino, K. Optical determination of impulse conduction velocity during development of embryonic chick cervical vagus nerve bundles. J. Physiol. 439:361–381;1991. 25. Tamar, H. Taste reception in the opposum and bat. Physiol. Zool. 34:86–91;1961. 26. Zecevic, D.; Wu, J.-Y.; Cohen, L.B.; London, J.A.; Hopp, H.-P.; Falk, C.X. Hundreds of neurons in the Aplysia abdominal ganglion are active during the gill-withdrawal reflex. J. Neurosci. 9:3681–3689;1989. 27. Zecevic, D. Multiple spike-initiation zones in single neurons revealed by voltage-sensitive dyes. Nature 381:322–325; 1996.