- Email: [email protected]
Recovery of Chorda Tympani Nerve Function Following Injury
EXPERIMENTAL NEUROLOGY ARTICLE NO.
141, 337–346 (1996)
0169
Recovery of Chorda Tympani Nerve Function Following Injury PETER CAIN, MARION E. FRANK,
AND
MICHAEL A. BARRY
Department of BioStructure and Function, MC 3705, University of Connecticut Health Center, Farmington, Connecticut 06030
The chorda tympani (CT) nerve carries taste information from the anterior tongue to the brain stem. Injury to the chorda tympani may result in loss or distortion of taste information. This study examined changes occurring in the hamster peripheral taste system during recovery from injury. The hamster chorda tympani nerve was crushed in the middle ear and the animals were allowed to survive from 2 to 16 weeks. At 2 weeks, CT fibers had degenerated distal to the crush site. Up to 16 weeks after crush, there were 67% fewer myelinated fibers in regenerated nerves than in controls. The mean area of the Ca21-ATPase-stained core of the fungiform taste buds was significantly smaller than in controls 2 weeks after injury, but recovered to control values by 4 weeks. Electrophysiological responses to taste stimuli were recorded from the chorda tympani distal to the injury. No responses were seen after 2 weeks; weak and unstable responses were seen after 3 weeks. By 4–8 weeks, relative responses to taste stimuli were similar to control responses, but the variability of the responses to sucrose was significantly greater than that in controls. The frequency of responses to the water rinse following taste stimuli, particularly sucrose, was also greater in the regenerated nerves. The abnormal electrophysiological responses to sucrose may be the result of the differential rate of return of fiber types and/or the transduction mechanisms. In some ways, recovery of the peripheral gustatory system after damage to the chorda tympani nerve recapitulates the later stages of taste bud development. r 1996 Academic Press, Inc.
INTRODUCTION
Gustatory information from fungiform papillae on the anterior tongue is carried by afferent fibers of the chorda tympani branch of the facial nerve to the brain stem. In addition, the chorda tympani (CT) contains preganglionic parasympathetic secretomotor fibers to the submandibular and sublingual salivary glands and efferent vasodilator fibers to the tongue (52). In humans, the chorda tympani may be damaged during middle ear surgery (12) and oral surgery (11). Damage to the CT may result in a loss of taste, dysgeusias (taste distortions), or taste phantoms (11,
26, 33, 45). Recovery of the gustatory system after damage has been examined in humans (58) and experimentally in other species, including cats, gerbils, hamsters, and rats. In hamsters and rats, taste buds exhibited ultrastructural changes and atrophy after chorda tympani or combined chorda tympani–lingual nerve denervation (5, 46, 48, 49, 56), but were shown to persist for long periods in this state (48, 56), retaining some normal characteristics (6, 49). In contrast, the majority of gerbil taste buds disappeared after denervation by chorda tympani and lingual nerves (48). Gerbil taste buds reappear approximately 2 days after reinnervation by the chorda tympani (15). In cats, 12 weeks after nerve crush and regeneration, there is a loss in the number of fungiform papillae but no difference in the number of taste buds per papilla (53). Composite response profiles recorded from regenerating gerbil CT nerves were similar to those of normal nerves (14). However, there were broad differences between nerves in their responses to sucrose and NaCl. In cats (52), the response profiles from regenerated whole nerves 12 weeks after crushing were similar to normal, while single-fiber recordings revealed less vigorous responses, slower conduction velocities, and responses to a narrower range of stimuli. Recovery of behavior mediated by the CT nerve was seen following bilateral CT nerve crush and regeneration in hamsters (7). We were interested in examining the physiological responses of the CT underlying the recovery of behavior. Observations in gerbils (14), in cats (52), and in our preliminary studies on hamsters (13) suggested that the recovery of the taste system after CT damage was neither as rapid nor as complete as previously proposed. We examined the physiological and anatomical processes underlying the pattern of chorda tympani recovery in the golden hamster, a species for which there is considerable normative data on the gustatory system. The chorda tympani has been used extensively in electrophysiological studies of responses to taste stimuli in hamsters (21, 23, 31). We hypothesized that after regeneration, there may be a differential recovery of gustatory sensitivities to various taste stimuli, but expected a similar number of myelinated fibers in the normal and recovered adult chorda tympani and no difference in the size and number of the fungiform taste buds on the tongue. We found many similarities be-
337
0014-4886/96 $18.00 Copyright r 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
338
CAIN, FRANK, AND BARRY
tween the recovering and normal chorda tympani, but there were differences indicating that the events witnessed in the recovering nerve may be similar to those observed during development. MATERIALS AND METHODS
Adult male Syrian hamsters (Mesocricetus auratus), weighing 95–175 g, obtained from a commercial supplier (Charles River Laboratories), were housed individually in plastic cages (24 3 45 3 15 cm) and maintained on a 12h:12h light:dark cycle. Animals were fed rodent diet (Agway Pro-Lab 3000) and water ad libitum. After arrival, the animals were allowed a minimum of 1 week of acclimatization to their new surroundings. Procedures Hamsters (n 5 27) were divided into two groups for the electrophysiological study: intact control (n 5 7) and experimental nerve crush (n 5 20). The experimental nerve crush animals were grouped according to length of recovery time. The nerve crush animals were anesthetized with sodium pentobarbital (90 mg/kg ip; Nembutal, Abbott Laboratories). Body temperature was maintained between 36 and 39°C. The right CT nerve was exposed by opening a small hole on the dorsal margin of the tympanic membrane. The nerve was seen hanging free as it passed through the middle ear, or could be brought into view by gently pulling it from behind the bony shelf that traverses the dorsal margin of the middle ear immediately behind the tympanic membrane. The nerve was crushed 153 at a single point with No. 5 Inox forceps (15, 52) as it passed through the middle ear. Control animals received no manipulations. Subjects recovered for a minimum of 12 days before they were reanesthetized with sodium pentobarbital (90 mg/kg ip), positioned in a nontraumatic head holder (19), and maintained at a surgical level of anesthesia with supplemental doses (pentobarbital, 45 mg/kg). Subjects were tracheotomized and the hypoglossal nerves were cut bilaterally to prevent tongue movement. The right CT was exposed in the infratemporal fossa, cut just distal to its exit from the tympanic bulla, dissected free of the underlying tissue, desheathed, and placed on a nichrome wire electrode (36 gauge, 127 µm diameter). The recording site was approximately 1 cm distal to the crush site. The lingual nerve was not disturbed. An indifferent electrode was placed in neighboring tissue. Physiological Measurements The multifiber responses of the CT were recorded while stimulus solutions were flowed into a glass flow-chamber encasing the anterior portion of the tongue
(28). Solutions were made from reagent grade chemicals and deionized water. Each stimulus presentation lasted between 10 and 20 s and was followed by a 45-s deionized water rinse. The interval between stimuli was at least 60 s. They were presented in the following order: 0.1 M sucrose, 0.3 M sucrose, 0.5 M sucrose, 0.03 M NaCl, 0.1 M NaCl, 0.3 M NaCl, 0.1 M sodium acetate, 0.3 M sodium acetate, 0.1 M KCl, 0.3 M KCl, amiloride series, 0.1 M NaCl, 0.003 M HCl. In selected animals, the stimulus series was presented more than once. The amiloride series was used to test the amiloridesensitive component of the NaCl response (30). Ten seconds of 0.1 M NaCl was followed by 10 s of 0.1 M NaCl in 10 µM amiloride, which was followed by 10 s of 0.1 M NaCl, with no water rinse between stimuli (30). The nerve responses were amplified 10003 with a battery-powered A.C. amplifier and then further amplified (1–1003) to obtain signals on the order of 1 V. These signals were monitored with an oscilloscope and loudspeaker and recorded on an analog tape recorder (TEAC). The responses were squared and summed with a time constant of 200 ms (PAVC 1A, Duck Engineering Design) and displayed on a chart recorder (AstroMed Dash IV) for analysis. As absolute response magnitudes of whole nerve recordings are very variable in control animals (9), the relative response, R, to each stimulus was calculated from these chart records as R 5 RX 4 RS 3 100, where RX is the average magnitude of pen deflection from baseline during Seconds 3–8 after onset of stimulus X, and RS is the same measurement for stimulus 0.1 M NaCl. The effect of amiloride on the neural response is expressed as percentage suppression and was calculated as % Suppression 5 100 2 (RS1a 4 RS 3100), where RS1a was the average magnitude of pen deflection from baseline during Seconds 3–8 of stimulation with 0.1 M NaCl plus 10 µM amiloride and RS equals the average response to 0.1 M NaCl for 5 s immediately preceding the presentation of NaCl plus amiloride. In those cases where the stimulus series was repeated, the mean relative response to each stimulus was calculated from the individual relative responses. The mean relative response was used to calculate group responses. Fiber Morphology In selected cases, the previously crushed CTs were removed and, along with previously crushed and normal CTs obtained from aldehyde-perfused animals not used for electrophysiology, were characterized using transmission electron microscopy. Survival times ranged from 2 to 16 weeks. Anesthetized animals were per-
339
CHORDA TYMPANI RECOVERY
fused intracardially with 0.1 M phosphate buffer (pH 7.4) containing heparin and lidocaine, followed by 2% paraformaldehyde and 1% glutaraldehyde in phosphate buffer. Both freshly removed and perfusion-fixed nerves were immersed in 2.5% glutaraldehyde and 1% paraformaldehyde in 0.1 M cacodylate buffer for at least 24 h. They were postfixed with 0.5% osmium tetroxide for 2 h and embedded in Spurr resin. Thin cross sections were collected on formvar-coated slot grids and stained with aqueous 2–5% uranyl acetate and Sato’s triple lead. Sections were taken from the same part of the nerve as the nerve recordings, about 3 mm proximal to the junction of the CT and lingual nerves. Measurements from all myelinated fibers were made from one thin section in each case. A nearby section was utilized to verify the number of profiles. Counts and measurements of fibers were made from photograph montages at a final magnification of 59403. A computer determined the diameter of each myelinated fiber (not including myelin) from a digitized outline of its cross section. The analyzer was blind to the side of the animal that had been crushed.
rho correlation coefficients, comparing the response patterns of each experimental group with that of the control. We compared the mean group neural responses to each stimulus with separate Kruskal–Wallis 1-way analyses of variance. In order to determine if the variability of neural response was related to the length of recovery, we compared the variance of the responses of each experimental group to the variance of the control group response with a two-tailed F test. The mean diameter of the myelinated fibers in the crushed nerves was compared with that of the intact nerves with an independent t test and the distribution of axon diameters determined. Post hoc comparisons were made using independent t tests corrected for multiple comparisons. The mean core areas of fungiform taste buds on the crush side were determined and compared, using paired t tests corrected for multiple comparisons, to the areas of fungiform buds on the contralateral (intact) side at 2 through 20 weeks survival time. The effect of recovery time on taste bud area was examined using Kruskal–Wallis 1-way analysis of variance. Criterion for significance was P 5 0.05 for all comparisons.
Taste Bud Morphology We also determined the mean core area of fungiform taste buds in the posterior half of the anterior tongue from selected animals. Animals were perfused intracardially with 0.1 M phosphate buffer (pH 7.4) containing heparin and lidocaine, followed by 2% paraformaldehyde and 1% glutaraldehyde in phosphate buffer. The tongues were postfixed in the same fixative for 2–12 h and soaked overnight in 30% sucrose in phosphate buffer. Serial 25-µm transverse frozen sections were cut and processed to reveal Ca21-ATPase after the method of Ando et al. (3; see also 4). Fungiform papillae were identified based on their large size, the presence of a long connective tissue core, and staining of the connective tissue core and nerve bundles with ecto-Ca21ATPase (4). From the section showing its largest diameter, the outline of the Ca21-ATPase-stained portion of the taste bud, which excludes peripheral supporting cells, was traced with a drawing tube. This outline was entered into a computer and the area within the outline determined. Analysis The independent variables were the length of recovery time and the taste stimuli. The dependent variables were: (i) the relative neural response to each stimulus, (ii) frequency of neural response to the water rinse, (iii) number and diameter of regenerated nerve fibers, and (iv) the mean taste bud area. The dependent variables were analyzed separately. Animals were grouped according to recovery time and a mean group response to each stimulus was calculated. We calculated the Spearman
RESULTS
Chorda Tympani Electrophysiology Taste stimuli. The mean responses (6SE) of each of the experimental groups to the different stimuli are shown in Fig. 1. We did not discern responses to taste stimuli above spontaneous discharges at 2 weeks after nerve crush. We first detected responses to taste stimuli 3 weeks after crush, but these responses were not stable. Response magnitude diminished during the period of time needed to apply the stimulus series. Stable responses were detected and recorded 4 weeks after nerve crush. We found strong positive correlations between the response patterns of the 4- to 5-week and the control groups (rho 5 0.76, P , 0.01), the 6- to 7-week and the control groups (rho 5 0.89, P , 0.01), and the 8-week and the control groups (rho 5 0.82, P , 0.01). We found no significant differences among the groups in mean relative responses to any stimulus. Thus neither the pattern of response across stimuli nor relative response magnitudes to the stimuli differed in control and recovered nerves. Experimental groups and the control group differed in the amount of variance in some responses (Fig. 1, asterisks). At 4 to 5 weeks of recovery, responses to 0.1 M (F8,7 5 92.7), 0.3 M (F8,7 5 47.4), and 0.5 M (F6,3 5 577) sucrose were more variable than control responses (P , 0.005). The responses at 4 to 5 weeks to 0.03 M NaCl (F8,7 5 7.38, P , 0.01) and 0.1 M KCl (F8,6 5 47.13, P , 0.005) also varied more than control responses. At longer survival times, only responses to sucrose were significantly more variable than control
340
CAIN, FRANK, AND BARRY
responses: for 6- to 7-weeks recovery, 0.1 M (F5,7 5 15.21), 0.3 M (F5,7 5 11.77), and 0.5 M (F5,3 5 44), and for 8-weeks recovery, 0.1 M (F7,7 5 44.26), 0.3 M (F7,7 5 40.37), and 0.5 M (F7,3 5 232.67) (P , 0.005). The variance of the sucrose responses was due in part to a subset of nerves that responded to sucrose with a decrease below baseline. At 4 to 5 weeks of survival, 2 of 8 nerves; at 6 to 7 weeks, 1 of 5 nerves; and at 8 weeks, 1 of 8 nerves responded to sucrose with a decrement. We did not see this decrement in response to sucrose in the seven nerves of the control group. Amiloride response. For all groups, the CT response to 0.1 M NaCl was strongly suppressed by the sodium channel inhibitor, amiloride. The recovering nerve was at least as sensitive to amiloride inhibition at 4 to 5 and
FIG. 2. The integrated multifiber responses of the CT to 0.1 M NaCl followed by 10 µM amiloride in 0.1 M NaCl solution are shown for a normal animal (a) and a 5-week recovery case (b). Upward pointing arrows indicate the onset of 0.1 M NaCl; downward pointing arrows indicate the onset of 10 µM amiloride 1 0.1 M NaCl. The difference in the signal-to-noise ratios between the 5-week subjects and the normal subjects is apparent. For this presentation, the scale of the chart recorder was adjusted as necessary to match the peak response of the two records visually.
FIG. 1. The mean relative responses (6SE) of the chorda tympani to different stimuli are shown by experimental group. (a) Responses from intact, normal nerves (n 5 7). (b) Responses from nerves crushed 8 weeks before test (n 5 7). (c) Responses from nerves crushed 6 to 7 weeks before test (n 5 4). (d) Responses from nerves crushed 4 to 5 weeks before test (n 5 8). Asterisks indicate that the variability of response differed significantly from controls at the 0.05 level.
6 to 7 weeks (83 and 86% suppression, respectively) as at 8 weeks (77%) and in controls (75%). Figure 2 shows the largest integrated multifiber responses of the CT to 0.1 M NaCl followed by 0.1 M NaCl in 10 µM amiloride for a 5-week subject and for a control. The near-total suppression of the NaCl response by amiloride at 5 weeks is evident. The differences in the signal-to-noise ratio between the 5-week subject and the control subject are notable in this comparison. Water response. We found that 45% (9/20) of the injured nerves and 43% (3/7) of control nerves responded to the deionized water rinse following at least one taste stimulus with a sudden increase in activity above the prestimulus baseline (Fig. 3). Responses were not the result of thermal or mechanical stimula-
341
CHORDA TYMPANI RECOVERY
Chorda Tympani Morphology
FIG. 3. The proportion of CT nerves responding to the deionized water rinse following the application of different stimuli to the tongue is shown for (a) normal animals, (b) animals 6–8 weeks after crush, and (c) animals 4–5 weeks after crush.
Figure 4 shows electron photomicrographs of chorda tympani fibers at various recovery times. Immediately following nerve crush, only a thin piece of the neuronal sheath was left. In one case, at 12 days after CT crush, there were no intact fibers. In another nerve, there were 26 myelinated and some intact unmyelinated fibers 14 days after crush. However, most fibers were missing or undergoing degeneration (Fig. 4). At 4 weeks, there were still signs of degeneration; the debris from degenerated fibers was sometimes visible in lysosomes within Schwann cell tubes. However, by 5 weeks no degenerative profiles were present and all fibers exhibited normal morphology. The distribution of myelinated axon diameters is shown in Fig. 5. There was a significant difference (t10 5 6.44, P , 0.001) between the average number of myelinated fibers in nerves examined 5 to 16 weeks after crush (151.3 6 20.4) (n 5 7) and that in the intact nerves (454.6 6 48.5) (n 5 5). We found no correlation between number of myelinated fibers and survival time 5–16 weeks after crush. The mean diameter of the myelinated axons (not including myelin) in the crushed nerves (1.35 6 0.04 µm) and intact nerves (1.33 6 0.08 µm) did not differ. The inset shows the distribution of fiber diameters 2 to 4 weeks after injury. The fiber diameters appear more heterogeneous at 2 to 4 weeks than at longer survival times. At 5–16 weeks the percentage of large fibers remained greater than in intact nerves; the percentage of axons larger than 2.4 µm was 5.07 6 0.47% at 5–16 weeks and 1.62 6 0.55% in intact nerves (t10 5 4.71, P , 0.001). Fungiform Taste Buds
tion. Stimuli were all at room temperature at the time of presentation, and mechanical stimulation would appear equally across all stimuli. Both the recovery time (F2,12 5 5.22, P 5 0.015) and the stimulus type (F4,12 5 14.91, P , 0.0005) affected the proportion of nerves responding. There were more responses to water following a greater variety of stimuli by the 4- to 5-week group. Five of the experimental nerves responding to water were in the 4- to 5-week recovery group (n 5 8), compared to two in the 6- to 7-week group (n 5 5), and two in the 8-week group (n 5 7). When the experimental animals were considered as a group, the greatest number of water responses, 38%, followed sucrose stimulation. Only 12% of the total NaCl presentations were followed by water responses, 5% of the sodium acetate presentations, 8% of the KCl presentations, and 7% of HCl presentations. When the experimental animals were divided according to recovery time, only the 4- to 5-week group responded significantly more often following sucrose stimulation than did the intact controls (x2 5 4.83, P 5 0.028) (Fig. 3).
Figure 6 shows the mean taste bud area on the crush or experimental side compared with the intact side for different survival times. Two to 3 weeks after nerve crush, the mean area of the Ca21-ATPase-stained core of fungiform taste buds ipsilateral to the crush was reduced to about 52% (731.9 6 98.6 µm2) of that on the intact, contralateral side (1391 6 97.4 µm2) (t4 5 31.55, P , 0.001). By 4 weeks, the taste bud area on the crush side had increased such that there were no longer significant differences between the crush and intact sides. There were no significant differences in taste bud number between the crush and intact sides at any survival time (Table 1). DISCUSSION
We found that the chorda tympani nerve reinnervates the tongue after crush injury with significantly fewer fibers. These fibers are sufficient to support normal taste bud size and number. During recovery, the overall responses of the injured nerve to salts and HCl
342
CAIN, FRANK, AND BARRY
FIG. 4. Electron micrographs of cross sections through the chorda tympani nerve: (a) Two weeks after nerve crush, a Schwann cell tube may contain a macrophage with numerous lysosomes filled with lipid (asterisk) or myelin whorls (arrowhead). (b) Ten weeks after CT nerve crush, myelinated and unmyelinated (arrowheads) fibers with normal morphology are seen. However, there are many fewer myelinated fibers (c) than in the intact CT (d) from the other side of the same animal.
are remarkably similar to those of controls, but responses to sucrose and water are notably dissimilar to those of controls. These dissimilarities may result from a differential rate of return of fibers and/or transduction mechanisms. We propose that the recovery of the peripheral gustatory nervous system after damage to the chorda tympani nerve may follow a sequence similar to that observed during development. Early Characteristics Our morphological studies of the chorda tympani nerve revealed substantial myelin degeneration 2 weeks after the crush. Jang and Davis (34) report a similar
time course; they demonstrated degenerating myelin from CT parasympathetic fibers 3 weeks after section. In addition, we found the mean area of fungiform taste buds on the experimental side 2–3 weeks after injury to be comparable to the 54% reduction seen in fungiform taste bud size following CT section (6). We found no decrease in taste bud numbers with chorda tympani nerve crush, a finding consistent with reports on taste bud denervation following chorda tympani or chorda tympani–lingual section (6, 49). However, Oakley et al. (48) found that 26% of hamster fungiform taste buds had disappeared 3 weeks after chorda tympani–lingual denervation.
343
CHORDA TYMPANI RECOVERY
TABLE 1 Mean Number (6SEM) of Taste Buds on the Posterior Half of the Fungiform Taste Bud Field Following Unilateral CT Nerve Crush Survival time
CT crush Intact
2–3 Weeksa
4–5 Weeksb
6–8 Weeks c
14.2 6 1.1 16.6 6 1.2
15.2 6 1.8 15.4 6 1.0
15.0 6 1.1 14.8 6 1.2
a
n 5 5. n 5 5. c n 5 4. b
FIG. 5. The average numbers of myelinated fibers in the intact and CT crush nerves, 5–16 weeks after injury, are shown arrayed by diameter. The inset shows the average number of fibers 2–4 weeks after injury.
We found that taste solutions that normally elicit near-saturating CT responses did not elicit discernible responses from the CT fibers that may have regenerated by 2 weeks. Brief multifiber responses were seen 3 weeks after injury, but neither the baseline (spontaneous) activity nor the responses were stable and both decreased with time. In the gerbil chorda tympani, multifiber responses recorded in the middle ear were demonstrated at 2 weeks after chorda tympani crush just proximal to its union with the lingual nerve (14).
FIG. 6. The mean taste bud area (µm2) on the control versus CT crush sides of the animals: 2–3, 4–5, and 6–8 weeks after unilateral injury. The single asterisk indicates a significant difference (P , 0.05) between the mean taste bud area of the control and CT crush sides at the same survival time.
The differences in the recovery times in the gerbil and hamster data may be due in part to the different distances to be traversed by the recovering axons. Functional recovery in the somatosensory system following peripheral nerve crush varies from slight to moderate at early stages after injury (24, 39). For example, Fugleholm et al. (24) were able to elicit action potentials with electrical nerve stimulation 16 days after crushing the afferent fibers of the cat tibial nerve; however, the current needed to elicit a discernible response was two orders of magnitude greater than that necessary to elicit a response in a normal nerve. Recovery Four months after injury, we found fewer myelinated fibers than normal in the CT. Thirty-seven percent of normal hamster CT fibers are myelinated and the parasympathetic component of the CT is primarily unmyelinated (34). Ninety-two percent of myelinated CT fibers are sensory; thus, most of the lost myelinated fibers in our study were sensory. In comparison, the number of myelinated fibers in the recovered mouse peroneal nerve was unaltered after a single crush (25), and the number of myelinated fibers in tributaries of the sciatic nerve increased after crush injury or transection and reapproximating the nerve (35, 36). The increase may be the result of sprouting by the recovering nerves. We observed neither morphological signs of sprouting nor an increase in the number of myelinated fibers, although sprouting of unmyelinated fibers was apparent at early stages of regeneration. The reduction in the number of CT myelinated fibers may be the result of ganglion cell death, or perhaps the failure of some myelinated axons to regenerate completely, although we did not examine the nerve at the crush site. Following CT section in the hamster, Whitehead et al. (57) found an 18% decrease (not significant) in the number of geniculate ganglion cells with CT processes. We noticed that axonal retrograde transport did not match control levels after 5 months recovery in that study
344
CAIN, FRANK, AND BARRY
(57). In other sensory nerves, the percentage of ganglion cell loss resulting from nerve section is typically 15–30% (for review see Ref. 2). Usually the effects of nerve crush are less severe, and thus the loss of 67% of myelinated fibers in the regenerated chorda tympani is unusual compared to other systems. Although the mean patterns of responses in recovering and normal nerves were similar, the variability of responses to sucrose in the regenerated nerve was significantly greater than that of normal animals. Normal patterns of response from regenerated fibers in gerbils (14) and cats (52) were reported. Cheal et al. (14) suggested that in gerbils, all receptor types and transduction mechanisms were functional. However, ‘‘fewfiber’’ responses to sucrose varied widely early in recovery and some regenerated fibers had low responses to sucrose (14). In contrast to responses to sucrose, we found that responses to NaCl by regenerated nerves showed minor differences in variance from the normal nerves and showed amiloride sensitivity at all survival times. Two possible explanations for the differences in recovery of sucrose and NaCl responses are: (i) sucrose-sensitive fibers may make more variable functional connections in the first 8 weeks after nerve injury, or (ii) reconstitution of transduction pathways for sucrose may vary during this period. The magnitude of the whole nerve response is directly related to the number of fibers present in the nerve at the time of recording (10, 22, 32). Based on response profiles, the afferent population of the normal chorda tympani nerve is composed of subpopulations of fibers (23). The magnitude of the whole nerve response to each stimulus type should primarily reflect the contribution of the appropriate subpopulations. Data from other species indicate that after regeneration, there is recovery of fiber specificity (14, 52). Thus, greater relative magnitude and low variability in response to one type of taste stimulus as compared to another during recovery may be due to differential recovery of functional subpopulations. In the normal hamster CT, fibers responding best to Na1/Li1 stimulation (N fibers) outnumber fibers responding best to sweet stimuli about 2 to 1 (23). In addition, the responses of some N and HCl-best (H) fibers to sweet stimuli are depressed below baseline levels in normal animals. The sucrose-induced depression of the whole nerve response below baseline levels that we observed may be due to the depression in activity of N or H fibers. This would indicate that the fibers present during the early stages of recovery were predominantly Na1/Li1 and/or acid-best and may dominate the whole nerve response. Sucrose-best fibers, as a class, may recover more slowly than N and H fibers. Smaller diameter, slower conducting, myelinated sensory fibers in the sciatic nerve recover significantly more slowly than the larger
diameter, faster conducting, myelinated sensory fibers (18). However, in the mouse peroneal nerve there is no evidence that one class of fibers recovers more slowly than another (25). Our data indicate that the diameter of many of the early returning myelinated fibers is larger than those appearing later; however, the diameters of the three functional classes of afferent fibers in the hamster CT are unknown. Alternatively, the recovery time necessary to produce an effective and stable pathway for sucrose transduction may be longer than the recovery time to produce an efficacious Na1 transduction pathway. The transduction of sweet stimuli has been shown to be effected via a second-messenger pathway (1, 16, 17, 37, 38, 41, 54), whereas the specific transduction of sodium salts is mediated by an apical (amiloride-sensitive) cation channel: sodium influx directly depolarizes the receptor cell (29, 30, for review see Ref. 38). Like sucrose, the effects of serotonin are also mediated via a second-messenger pathway. Following irreversible serotonin receptor inactivation, new receptor density was significantly less than in controls after 14 days (50). While most fungiform taste buds in the hamster survive denervation by the chorda tympani and lingual nerves, they exhibit morphological changes and atrophy (5, 6, 48, 56), including loss of visible taste pores (46, 48). It is likely that in these denervated taste bud cells, receptors, protein synthesis, and second-messenger systems are also reduced relative to normal. At 8 days after denervation, taste bud cells retain their elongated shape; however, no intragemmal nerves are visible, and the taste bud cells are intermediate in electron density, lack well-developed smooth and rough endoplasmic reticula, and possess more rounded nuclei with evenly dispersed chromatin (56). After 21 days, denervated taste buds show light and dark cells, but are half the size of controls (6, 56). Following CT nerve regeneration, functional taste receptor cells could arise from these atrophied taste bud cells and/or new cells may be produced by stem cells in the taste bud. In either case, the synthesis and assembly of the complex G-protein and second-messenger systems required for sucrose transduction may take more time than the production of apical sodium channels. The CT water response has been shown to be a function of adaptation to preceding stimuli, including the ionic constituents of saliva, in rats and cats (8, 42, 43). Earlier reports included the hamster among species that did not show an excitatory response to water (10). Our results show that a small proportion of control hamster CTs responded to the water rinse. However, we found that a greater proportion of nerves responded to water, and responded following a wider range of stimuli, during the early stages of recovery. In young rats, the responses to the water rinse following sugar stimulation (‘‘off responses’’) were larger and more
345
CHORDA TYMPANI RECOVERY
frequent than in adults (27). This suggests that the recovery of sugar-processing mechanisms after CT damage may follow a pattern similar to that which occurs in development.
system may follow a progression of events leading to a mature response. The variability of response to sucrose during recovery after CT injury may be due in part to similar events.
Recovery May Recapitulate Ontogeny In contrast to the response to sucrose, Hill (31) found the week-old hamster to be highly sensitive to Na1 relative to NH4Cl and found that the Na1 response was strongly inhibited by amiloride. As the hamster matures, Na1 sensitivity (relative to NH41) decreases (31). The suppression of the Na1 response by amiloride early in hamster CT recovery and during development demonstrates that NaCl transduction via the amiloridesensitive cation channel (29, 30, 38) is present and functioning early in both processes. Recovery of taste cells following denervation might recapitulate the later stages of taste bud development (47). Our data offer evidence on a number of levels that recovery of the peripheral taste system following chorda tympani nerve injury may follow a developmental sequence. Denervated taste buds (48, 56) are similar to the developing rat fungiform taste bud at Farbman’s (20) stages 3 and 5. Elongate epithelial cells and no discernible type I and type II (dark and light, respectively) cells are characteristics of the taste bud 8 days after denervation (48, 56) and at stage 3 of development (20). Type I and II cells are seen at stage 5 (20) and at 21 days after denervation (6, 56) although both the papilla and the taste bud are smaller than in the control (6, 48, 56) or adult (20). The time-dependent increase in consistently positive responses to sucrose in the regenerated hamster CT is similar to that seen during development (31). In the regenerated CT, the frequent responses to water rinses, particularly following sucrose, are similar to those observed in young rats (27). The need to express a variety of receptors in order to effectively transduce sugar stimuli (17, 21, 31, 51, 55) may contribute to the variabiilty of responses seen early in development and regeneration. Similarly in another system, serotonin receptor subtypes require variable times to recover following experimental inactivation (50). Other G-protein-coupled systems show increases in receptor numbers during postnatal development (e.g., cannabinoid (44); opioid (for review see Ref. 40)). In contrast, as early as Postnatal Day 3 in rats, mRNA expression for the cannabinoid receptor was at adult levels (44). Thus, the expression of receptor mRNA may precede receptor proliferation. While these data do not address the number of receptors that are necessary and sufficient for transduction, establishment of a mature number of receptors may take time. In addition, functional delays may also be the result of changes in effector coupling during development (40). These reports indicate that development of a second-messenger
ACKNOWLEDGMENTS We thank Jonathon Clive for assistance with statistical analysis. We thank David Larson and Connie Gillies for technical assistance. This work was supported in part by Grants T32-DC00025-06 (P.C.), R01-DC00058 (M.E.F.), and P50-DC00168 (M.A.B.) from NIDCD.
REFERENCES 1.
2.
3.
4.
5. 6.
7.
8.
9. 10. 11.
12. 13.
14.
15.
16.
17.
AKABAS, M. H. 1991. Mechanisms of chemosensory transduction in taste cells. In International Review of Neurobiology (J. R. Smythies and R. J. Bradley, Eds.) Vol. 32, pp. 242–279. Academic Press, San Diego. ALDSKOGIUS, H., J. ARDVIDSSON, AND G. GRANT. 1985. The reaction of primary sensory neurons to peripheral nerve injury with particular emphasis on transganglionic changes. Brain Res. Rev. 10: 27–46. ANDO, T., H. FUJIMOTO, H. MAYAHARA, H. MIYAJIMA, AND K. OGAWA. 1981. A new one-step method for histochemistry and cytochemistry of Ca21-ATPase activity. Acta Histochem. Cytochem. 14: 705–726. BARRY, M. A. 1992. Ecto-calcium dependent ATPase activity of mammalian taste bud cells. J. Histochem. Cytochem. 40: 1919– 1928. BARRY, M. A., AND M. E. FRANK. 1992. Response of the gustatory system to peripheral nerve injury. Exp. Neurol. 115: 60–64. BARRY, M. A., AND L. D. SAVOY. 1993. Persistence and calciumdependent ATPase staining of denervated fungiform taste buds in the hamster. Arch. Oral Biol. 38(1): 5–15. BARRY, M. A., D. C. LARSON, AND M. E. FRANK. 1993. Loss and recovery of sodium-salt taste following bilateral chorda tympani nerve crush. Physiol. Behav. 53: 75–80. BARTOSHUK, L. M. 1965. Effects of Adaptation on Responses to Water in Cat and Rat. Ph.D. Dissertation. Brown University, Providence, RI. BEIDLER, L. M. 1953. Properties of chemoreceptors of tongue of rat. J. Neurophysiol. 16: 595–607. BEIDLER, L. M., I. Y. FISHMAN, AND C. W. HARDIMAN. 1955. Species differences in taste responses. Am. J. Physiol. 181: 235–239. BLACKBURN, C. W., AND P. A. BRAMLEY. 1989. Lingual nerve damage associated with the removal of lower third molars. Br. Dent. J. 167: 103–107. BULL, T. R. 1965. Taste and the chorda tympani. J. Laryngol. Otol. 79: 479–493. CAIN, P., AND M. A. BARRY. 1993. Regeneration and functional response in the chorda tympani nerve in the golden Syrian hamster (Mesocricetus auratus). Chem. Senses 18: 263. CHEAL, M., W. P. DICKEY, L. B. JONES, AND B. OAKLEY. 1977. Taste fiber responses during reinnervation of fungiform papillae. J. Comp. Neurol. 172(4): 627–646. CHEAL, M., AND B. OAKLEY. 1977. Regeneration of fungiform taste buds: Temporal and spatial characteristics. J. Comp. Neur. 172(4): 609–626. CUMMINGS, T. A., AND S. C. KINNAMON. 1992. In situ recording from hamster taste cells: Responses to sweet stimuli and cAMP. Chem. Senses 17: 607–608. CUMMINGS, T. A., J. POWELL, AND S. C. KINNAMON. 1993. Sweet
346
18.
19. 20. 21.
22. 23.
24.
25.
26.
27.
28. 29.
30. 31. 32.
33. 34.
35.
36. 37. 38.
39.
CAIN, FRANK, AND BARRY taste transduction in hamster cells: Evidence for the role of cyclic nucleotides. J. Neurophysiol. 70(6): 2326–2336. DEVOR, M., AND R. GOVRIN-LIPPMANN. 1979. Selective regeneration of sensory fibers following nerve crush injury. Exp. Neurol. 65: 243–254. ERICKSON, R. P. 1966. Nontraumatic headholders for mammals. Physiol. Behav. 1: 97–98. FARBMAN, A. I. 1965. Electron microscope study of the developing taste bud in rat fungiform papilla. Dev. Biol. 11: 110–135. FAURION, A., AND C. VAYSETTES-COURCHAY. 1990. Taste as a highly discriminative system: A hamster intrapapillar single unit study with 18 compounds. Brain Res. 512: 317–332. FRANK, M. E. 1973. An analysis of hamster afferent taste nerve response functions. J. Gen. Physiol. 61: 588–618. FRANK, M. E., S. L. BIEBER, AND D. V. SMITH. 1988. The organization of taste sensibilities in hamster chorda tympani nerve fibers. J. Gen. Physiol. 91: 861–896. FUGLEHOLM, K., H. SCHMALBRUCH, AND C. KRARUP. 1994. Early peripheral nerve regeneration after crushing, sectioning, and freeze studied by implanted electrodes in the cat. J. Neurosci. 14(5): 2659–2673. GIANNINI, C., A. C. LAIS, AND P. J. DYCK. 1989. Number, size and class of peripheral nerve fibers regenerating after crush, multiple crush, and graft. Brain Res. 500: 131–138. GRANT, R., S. MILLER, D. SIMPSON, P. J. LAMEY, AND I. BONE. 1989. The effect of CT section on ipsilateral and contralateral salivary secretion and taste in man. J. Neurol. Neurosurg. Psychiatry 52: 1058–1062. HARADA, S., S. MAEDA, AND Y. KASAHARA. 1989. Sugar response characteristics in the rat chorda tympani nerve during development. Chem. Senses 14: 707–708. HARPER, H. W. 1979. Improved flow-chamber for taste stimulation. Physiol. Behav. 22: 1033–1035. HECK, G. L., S. MIERSON, AND J. A. DESIMONE. 1984. Salt taste transduction occurs through an amiloride-sensitive sodium transport pathway. Science 223: 403–405. HETTINGER, T. P., AND M. E. FRANK. 1990. Specificity of amiloride inhibition of hamster taste responses. Brain Res. 513: 24–34. HILL, D. L. 1988. Development of chorda tympani nerve taste responses in the hamster. J. Comp. Neurol. 268: 346–356. HILL, D. L., C. M. MISTRETTA, AND R. M. BRADLEY. 1982. Developmental changes in taste response characteristics of rat single chorda tympani fibers. J. Neurosci. 2: 782–790. HOUSE, H. P. 1963. Early and late complications of stapes surgery. Arch. Otolaryngol. 78: 606–613. JANG, T., AND B. J. DAVIS. 1987. The chorda tympani and glossopharyngeal nerves in the adult hamster. Chem. Senses 12: 381–395. JENQ, C., AND R. E. COGGESHALL. 1984. Effects of sciatic nerve regeneration on axonal populations in tributary nerves. Brain Res. 295: 91–100. JENG, C., AND R. E. COGGESHALL. 1984. Regeneration of axons in tributary nerves. Brain Res. 310: 107–121. KINNAMON, S. C. 1988. Taste transduction: A diversity of mechanisms. Trends Neurosci. 11: 491–496. KINNAMON, S. C., AND T. A. CUMMINGS. 1992. Chemosensory transduction mechanisms is taste. Annu. Rev. Physiol. 54: 715–731. KINNMAN, E., AND Z. WISENFELD-HALIN. 1993. Time course and characteristics of the capacity of sensory nerves to reinnervate skin territories outside their normal innervation zones. Somatosensory Motor Res. 10(4): 445–454.
40.
41. 42.
43.
44.
45. 46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
LESLIE, F. M., AND S. E. LOUGHLIN. 1992. Development of multiple opioid receptors. In Development of the Central Nervous System: Effects of Alcohol and Opiates (M. W. Miller, Ed.), pp. 255–283. Wiley–Liss, New York. MARGOLSKEE, R. 1993. The biochemistry and molecular biology of taste transduction. Curr. Opin. Neurobiol. 3: 526–531. MATSUO, R., AND T. YAMAMOTO. 1990. Taste nerve responses during licking behavior in rats: Importance of saliva in response to sweeteners. Neurosci. Lett. 108: 121–126. MATSUO, R., AND T. YAMAMOTO. 1992. Effects of inorganic constituents of saliva on taste responses of the rat chorda tympani nerve. Brain Res. 583: 71–80. MCLAUGHLIN, C. R., B. R. MARTIN, D. R. COMPTON, AND M. E. ABOOD. 1994. Cannabinoid receptors in developing rats: Detection of mRNA and receptor binding. Drug Alcohol Depend. 36: 27–31. MOON, C. N., AND E. W. PULLEN. 1963. Effects of chorda tympani section during middle ear surgery. Laryngoscope 73: 392–405. MYERS, W. E., T. P. HETTINGER, J. A. D’AMBROSIO, S. L. WENDT, C. B. PEARSON, M. A. BARRY, AND M. E. FRANK. 1994. Visualizing taste papillae in vivo with scanning electron microscopy of a high resolution cast. Chem. Senses 20(1): 1–8. OAKLEY, B. 1991. Neural–epithelial interactions in taste bud regeneration. In Regeneration of Vertebrate Sensory Receptor Cells (J. Whelan, Ed.), pp. 277–287. Ciba Foundation Symposium, London. OAKLEY, B., A. LAWTON, D. R. RIDDLE, AND L.-H. WU. 1993. Morphometric and immunocytochemical assessment of fungiform taste buds after interruption of the chorda-lingual nerve. Microsc. Res. Tech. 26: 187–195. OLIVER, S., AND M. C. WHITEHEAD. 1992. Morphometry and cellular dynamics of denervated fungiform taste buds in the hamster. Chem. Senses 17(5): 529–542. PINTO, W., AND G. BATTAGLIA. 1994. Comparative recovery kinetics of 5-hydroxytryptamine 1A, 1B, and 2A receptor subtypes in rat cortex after receptor inactivation: Evidence for differences in receptor production and degradation. Mol. Pharmacol. 46: 1111–1119. REHNBERG, B. G., T. P. HETTINGER, AND M. E. FRANK. 1990. The role of sucrose-sensitive neurons in ingestion of sweet stimuli by hamsters. Physiol. Behav. 48: 459–466. ROBINSON, P. P. 1989. The reinnervation of the tongue and salivary glands after lingual nerve injuries in cats. Brain Res. 483: 259–271. ROBINSON, P. P., AND P. A. WINKLES. 1991. The number and distribution of fungiform papillae and taste buds after lingual nerve injuries in cats. Archs. Oral Biol. 36(12): 885–891. STRIEM, B. J., U. PACE, U. ZEHAVI, M. NAIM, AND D. LANCET. 1989. Sweet tastants stimulate adenylate cyclase coupled to GTPbinding protein in rat tongue membranes. Biochem. J. 260: 121–126. TINTI, J. M., AND C. NOFRE. 1991. Why does a sweetener taste sweet? A new model. In Sweeteners: Discovery, Molecular Design, and Chemoreception (D. E. Walters, F. T. Orthoefer, G. E. Dubois, Eds.), pp. 206–213. Am. Chem. Soc., Washington, DC. WHITEHEAD, M. C., M. E. FRANK, T. P. HETTINGER, L.-T. HOU, AND H.-D. NAH. 1987. Persistence of taste buds in denervated fungiform papillae. Brain Res. 405: 192–195. WHITEHEAD, M. C., S. T. MCGLATHERY, AND B. G. MANION. 1995. Transganglionic degeneration in the gustatory system consequent to chorda tympani damage. Exp. Neurol. 132: 239–250. ZUNIGA, J. R., N. CHEN, AND I. J. MILLER. 1994. Effects of chorda-lingual nerve surgery and repair on human taste. Chem. Senses 19(6): 657–665.
141, 337–346 (1996)
0169
Recovery of Chorda Tympani Nerve Function Following Injury PETER CAIN, MARION E. FRANK,
AND
MICHAEL A. BARRY
Department of BioStructure and Function, MC 3705, University of Connecticut Health Center, Farmington, Connecticut 06030
The chorda tympani (CT) nerve carries taste information from the anterior tongue to the brain stem. Injury to the chorda tympani may result in loss or distortion of taste information. This study examined changes occurring in the hamster peripheral taste system during recovery from injury. The hamster chorda tympani nerve was crushed in the middle ear and the animals were allowed to survive from 2 to 16 weeks. At 2 weeks, CT fibers had degenerated distal to the crush site. Up to 16 weeks after crush, there were 67% fewer myelinated fibers in regenerated nerves than in controls. The mean area of the Ca21-ATPase-stained core of the fungiform taste buds was significantly smaller than in controls 2 weeks after injury, but recovered to control values by 4 weeks. Electrophysiological responses to taste stimuli were recorded from the chorda tympani distal to the injury. No responses were seen after 2 weeks; weak and unstable responses were seen after 3 weeks. By 4–8 weeks, relative responses to taste stimuli were similar to control responses, but the variability of the responses to sucrose was significantly greater than that in controls. The frequency of responses to the water rinse following taste stimuli, particularly sucrose, was also greater in the regenerated nerves. The abnormal electrophysiological responses to sucrose may be the result of the differential rate of return of fiber types and/or the transduction mechanisms. In some ways, recovery of the peripheral gustatory system after damage to the chorda tympani nerve recapitulates the later stages of taste bud development. r 1996 Academic Press, Inc.
INTRODUCTION
Gustatory information from fungiform papillae on the anterior tongue is carried by afferent fibers of the chorda tympani branch of the facial nerve to the brain stem. In addition, the chorda tympani (CT) contains preganglionic parasympathetic secretomotor fibers to the submandibular and sublingual salivary glands and efferent vasodilator fibers to the tongue (52). In humans, the chorda tympani may be damaged during middle ear surgery (12) and oral surgery (11). Damage to the CT may result in a loss of taste, dysgeusias (taste distortions), or taste phantoms (11,
26, 33, 45). Recovery of the gustatory system after damage has been examined in humans (58) and experimentally in other species, including cats, gerbils, hamsters, and rats. In hamsters and rats, taste buds exhibited ultrastructural changes and atrophy after chorda tympani or combined chorda tympani–lingual nerve denervation (5, 46, 48, 49, 56), but were shown to persist for long periods in this state (48, 56), retaining some normal characteristics (6, 49). In contrast, the majority of gerbil taste buds disappeared after denervation by chorda tympani and lingual nerves (48). Gerbil taste buds reappear approximately 2 days after reinnervation by the chorda tympani (15). In cats, 12 weeks after nerve crush and regeneration, there is a loss in the number of fungiform papillae but no difference in the number of taste buds per papilla (53). Composite response profiles recorded from regenerating gerbil CT nerves were similar to those of normal nerves (14). However, there were broad differences between nerves in their responses to sucrose and NaCl. In cats (52), the response profiles from regenerated whole nerves 12 weeks after crushing were similar to normal, while single-fiber recordings revealed less vigorous responses, slower conduction velocities, and responses to a narrower range of stimuli. Recovery of behavior mediated by the CT nerve was seen following bilateral CT nerve crush and regeneration in hamsters (7). We were interested in examining the physiological responses of the CT underlying the recovery of behavior. Observations in gerbils (14), in cats (52), and in our preliminary studies on hamsters (13) suggested that the recovery of the taste system after CT damage was neither as rapid nor as complete as previously proposed. We examined the physiological and anatomical processes underlying the pattern of chorda tympani recovery in the golden hamster, a species for which there is considerable normative data on the gustatory system. The chorda tympani has been used extensively in electrophysiological studies of responses to taste stimuli in hamsters (21, 23, 31). We hypothesized that after regeneration, there may be a differential recovery of gustatory sensitivities to various taste stimuli, but expected a similar number of myelinated fibers in the normal and recovered adult chorda tympani and no difference in the size and number of the fungiform taste buds on the tongue. We found many similarities be-
337
0014-4886/96 $18.00 Copyright r 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
338
CAIN, FRANK, AND BARRY
tween the recovering and normal chorda tympani, but there were differences indicating that the events witnessed in the recovering nerve may be similar to those observed during development. MATERIALS AND METHODS
Adult male Syrian hamsters (Mesocricetus auratus), weighing 95–175 g, obtained from a commercial supplier (Charles River Laboratories), were housed individually in plastic cages (24 3 45 3 15 cm) and maintained on a 12h:12h light:dark cycle. Animals were fed rodent diet (Agway Pro-Lab 3000) and water ad libitum. After arrival, the animals were allowed a minimum of 1 week of acclimatization to their new surroundings. Procedures Hamsters (n 5 27) were divided into two groups for the electrophysiological study: intact control (n 5 7) and experimental nerve crush (n 5 20). The experimental nerve crush animals were grouped according to length of recovery time. The nerve crush animals were anesthetized with sodium pentobarbital (90 mg/kg ip; Nembutal, Abbott Laboratories). Body temperature was maintained between 36 and 39°C. The right CT nerve was exposed by opening a small hole on the dorsal margin of the tympanic membrane. The nerve was seen hanging free as it passed through the middle ear, or could be brought into view by gently pulling it from behind the bony shelf that traverses the dorsal margin of the middle ear immediately behind the tympanic membrane. The nerve was crushed 153 at a single point with No. 5 Inox forceps (15, 52) as it passed through the middle ear. Control animals received no manipulations. Subjects recovered for a minimum of 12 days before they were reanesthetized with sodium pentobarbital (90 mg/kg ip), positioned in a nontraumatic head holder (19), and maintained at a surgical level of anesthesia with supplemental doses (pentobarbital, 45 mg/kg). Subjects were tracheotomized and the hypoglossal nerves were cut bilaterally to prevent tongue movement. The right CT was exposed in the infratemporal fossa, cut just distal to its exit from the tympanic bulla, dissected free of the underlying tissue, desheathed, and placed on a nichrome wire electrode (36 gauge, 127 µm diameter). The recording site was approximately 1 cm distal to the crush site. The lingual nerve was not disturbed. An indifferent electrode was placed in neighboring tissue. Physiological Measurements The multifiber responses of the CT were recorded while stimulus solutions were flowed into a glass flow-chamber encasing the anterior portion of the tongue
(28). Solutions were made from reagent grade chemicals and deionized water. Each stimulus presentation lasted between 10 and 20 s and was followed by a 45-s deionized water rinse. The interval between stimuli was at least 60 s. They were presented in the following order: 0.1 M sucrose, 0.3 M sucrose, 0.5 M sucrose, 0.03 M NaCl, 0.1 M NaCl, 0.3 M NaCl, 0.1 M sodium acetate, 0.3 M sodium acetate, 0.1 M KCl, 0.3 M KCl, amiloride series, 0.1 M NaCl, 0.003 M HCl. In selected animals, the stimulus series was presented more than once. The amiloride series was used to test the amiloridesensitive component of the NaCl response (30). Ten seconds of 0.1 M NaCl was followed by 10 s of 0.1 M NaCl in 10 µM amiloride, which was followed by 10 s of 0.1 M NaCl, with no water rinse between stimuli (30). The nerve responses were amplified 10003 with a battery-powered A.C. amplifier and then further amplified (1–1003) to obtain signals on the order of 1 V. These signals were monitored with an oscilloscope and loudspeaker and recorded on an analog tape recorder (TEAC). The responses were squared and summed with a time constant of 200 ms (PAVC 1A, Duck Engineering Design) and displayed on a chart recorder (AstroMed Dash IV) for analysis. As absolute response magnitudes of whole nerve recordings are very variable in control animals (9), the relative response, R, to each stimulus was calculated from these chart records as R 5 RX 4 RS 3 100, where RX is the average magnitude of pen deflection from baseline during Seconds 3–8 after onset of stimulus X, and RS is the same measurement for stimulus 0.1 M NaCl. The effect of amiloride on the neural response is expressed as percentage suppression and was calculated as % Suppression 5 100 2 (RS1a 4 RS 3100), where RS1a was the average magnitude of pen deflection from baseline during Seconds 3–8 of stimulation with 0.1 M NaCl plus 10 µM amiloride and RS equals the average response to 0.1 M NaCl for 5 s immediately preceding the presentation of NaCl plus amiloride. In those cases where the stimulus series was repeated, the mean relative response to each stimulus was calculated from the individual relative responses. The mean relative response was used to calculate group responses. Fiber Morphology In selected cases, the previously crushed CTs were removed and, along with previously crushed and normal CTs obtained from aldehyde-perfused animals not used for electrophysiology, were characterized using transmission electron microscopy. Survival times ranged from 2 to 16 weeks. Anesthetized animals were per-
339
CHORDA TYMPANI RECOVERY
fused intracardially with 0.1 M phosphate buffer (pH 7.4) containing heparin and lidocaine, followed by 2% paraformaldehyde and 1% glutaraldehyde in phosphate buffer. Both freshly removed and perfusion-fixed nerves were immersed in 2.5% glutaraldehyde and 1% paraformaldehyde in 0.1 M cacodylate buffer for at least 24 h. They were postfixed with 0.5% osmium tetroxide for 2 h and embedded in Spurr resin. Thin cross sections were collected on formvar-coated slot grids and stained with aqueous 2–5% uranyl acetate and Sato’s triple lead. Sections were taken from the same part of the nerve as the nerve recordings, about 3 mm proximal to the junction of the CT and lingual nerves. Measurements from all myelinated fibers were made from one thin section in each case. A nearby section was utilized to verify the number of profiles. Counts and measurements of fibers were made from photograph montages at a final magnification of 59403. A computer determined the diameter of each myelinated fiber (not including myelin) from a digitized outline of its cross section. The analyzer was blind to the side of the animal that had been crushed.
rho correlation coefficients, comparing the response patterns of each experimental group with that of the control. We compared the mean group neural responses to each stimulus with separate Kruskal–Wallis 1-way analyses of variance. In order to determine if the variability of neural response was related to the length of recovery, we compared the variance of the responses of each experimental group to the variance of the control group response with a two-tailed F test. The mean diameter of the myelinated fibers in the crushed nerves was compared with that of the intact nerves with an independent t test and the distribution of axon diameters determined. Post hoc comparisons were made using independent t tests corrected for multiple comparisons. The mean core areas of fungiform taste buds on the crush side were determined and compared, using paired t tests corrected for multiple comparisons, to the areas of fungiform buds on the contralateral (intact) side at 2 through 20 weeks survival time. The effect of recovery time on taste bud area was examined using Kruskal–Wallis 1-way analysis of variance. Criterion for significance was P 5 0.05 for all comparisons.
Taste Bud Morphology We also determined the mean core area of fungiform taste buds in the posterior half of the anterior tongue from selected animals. Animals were perfused intracardially with 0.1 M phosphate buffer (pH 7.4) containing heparin and lidocaine, followed by 2% paraformaldehyde and 1% glutaraldehyde in phosphate buffer. The tongues were postfixed in the same fixative for 2–12 h and soaked overnight in 30% sucrose in phosphate buffer. Serial 25-µm transverse frozen sections were cut and processed to reveal Ca21-ATPase after the method of Ando et al. (3; see also 4). Fungiform papillae were identified based on their large size, the presence of a long connective tissue core, and staining of the connective tissue core and nerve bundles with ecto-Ca21ATPase (4). From the section showing its largest diameter, the outline of the Ca21-ATPase-stained portion of the taste bud, which excludes peripheral supporting cells, was traced with a drawing tube. This outline was entered into a computer and the area within the outline determined. Analysis The independent variables were the length of recovery time and the taste stimuli. The dependent variables were: (i) the relative neural response to each stimulus, (ii) frequency of neural response to the water rinse, (iii) number and diameter of regenerated nerve fibers, and (iv) the mean taste bud area. The dependent variables were analyzed separately. Animals were grouped according to recovery time and a mean group response to each stimulus was calculated. We calculated the Spearman
RESULTS
Chorda Tympani Electrophysiology Taste stimuli. The mean responses (6SE) of each of the experimental groups to the different stimuli are shown in Fig. 1. We did not discern responses to taste stimuli above spontaneous discharges at 2 weeks after nerve crush. We first detected responses to taste stimuli 3 weeks after crush, but these responses were not stable. Response magnitude diminished during the period of time needed to apply the stimulus series. Stable responses were detected and recorded 4 weeks after nerve crush. We found strong positive correlations between the response patterns of the 4- to 5-week and the control groups (rho 5 0.76, P , 0.01), the 6- to 7-week and the control groups (rho 5 0.89, P , 0.01), and the 8-week and the control groups (rho 5 0.82, P , 0.01). We found no significant differences among the groups in mean relative responses to any stimulus. Thus neither the pattern of response across stimuli nor relative response magnitudes to the stimuli differed in control and recovered nerves. Experimental groups and the control group differed in the amount of variance in some responses (Fig. 1, asterisks). At 4 to 5 weeks of recovery, responses to 0.1 M (F8,7 5 92.7), 0.3 M (F8,7 5 47.4), and 0.5 M (F6,3 5 577) sucrose were more variable than control responses (P , 0.005). The responses at 4 to 5 weeks to 0.03 M NaCl (F8,7 5 7.38, P , 0.01) and 0.1 M KCl (F8,6 5 47.13, P , 0.005) also varied more than control responses. At longer survival times, only responses to sucrose were significantly more variable than control
340
CAIN, FRANK, AND BARRY
responses: for 6- to 7-weeks recovery, 0.1 M (F5,7 5 15.21), 0.3 M (F5,7 5 11.77), and 0.5 M (F5,3 5 44), and for 8-weeks recovery, 0.1 M (F7,7 5 44.26), 0.3 M (F7,7 5 40.37), and 0.5 M (F7,3 5 232.67) (P , 0.005). The variance of the sucrose responses was due in part to a subset of nerves that responded to sucrose with a decrease below baseline. At 4 to 5 weeks of survival, 2 of 8 nerves; at 6 to 7 weeks, 1 of 5 nerves; and at 8 weeks, 1 of 8 nerves responded to sucrose with a decrement. We did not see this decrement in response to sucrose in the seven nerves of the control group. Amiloride response. For all groups, the CT response to 0.1 M NaCl was strongly suppressed by the sodium channel inhibitor, amiloride. The recovering nerve was at least as sensitive to amiloride inhibition at 4 to 5 and
FIG. 2. The integrated multifiber responses of the CT to 0.1 M NaCl followed by 10 µM amiloride in 0.1 M NaCl solution are shown for a normal animal (a) and a 5-week recovery case (b). Upward pointing arrows indicate the onset of 0.1 M NaCl; downward pointing arrows indicate the onset of 10 µM amiloride 1 0.1 M NaCl. The difference in the signal-to-noise ratios between the 5-week subjects and the normal subjects is apparent. For this presentation, the scale of the chart recorder was adjusted as necessary to match the peak response of the two records visually.
FIG. 1. The mean relative responses (6SE) of the chorda tympani to different stimuli are shown by experimental group. (a) Responses from intact, normal nerves (n 5 7). (b) Responses from nerves crushed 8 weeks before test (n 5 7). (c) Responses from nerves crushed 6 to 7 weeks before test (n 5 4). (d) Responses from nerves crushed 4 to 5 weeks before test (n 5 8). Asterisks indicate that the variability of response differed significantly from controls at the 0.05 level.
6 to 7 weeks (83 and 86% suppression, respectively) as at 8 weeks (77%) and in controls (75%). Figure 2 shows the largest integrated multifiber responses of the CT to 0.1 M NaCl followed by 0.1 M NaCl in 10 µM amiloride for a 5-week subject and for a control. The near-total suppression of the NaCl response by amiloride at 5 weeks is evident. The differences in the signal-to-noise ratio between the 5-week subject and the control subject are notable in this comparison. Water response. We found that 45% (9/20) of the injured nerves and 43% (3/7) of control nerves responded to the deionized water rinse following at least one taste stimulus with a sudden increase in activity above the prestimulus baseline (Fig. 3). Responses were not the result of thermal or mechanical stimula-
341
CHORDA TYMPANI RECOVERY
Chorda Tympani Morphology
FIG. 3. The proportion of CT nerves responding to the deionized water rinse following the application of different stimuli to the tongue is shown for (a) normal animals, (b) animals 6–8 weeks after crush, and (c) animals 4–5 weeks after crush.
Figure 4 shows electron photomicrographs of chorda tympani fibers at various recovery times. Immediately following nerve crush, only a thin piece of the neuronal sheath was left. In one case, at 12 days after CT crush, there were no intact fibers. In another nerve, there were 26 myelinated and some intact unmyelinated fibers 14 days after crush. However, most fibers were missing or undergoing degeneration (Fig. 4). At 4 weeks, there were still signs of degeneration; the debris from degenerated fibers was sometimes visible in lysosomes within Schwann cell tubes. However, by 5 weeks no degenerative profiles were present and all fibers exhibited normal morphology. The distribution of myelinated axon diameters is shown in Fig. 5. There was a significant difference (t10 5 6.44, P , 0.001) between the average number of myelinated fibers in nerves examined 5 to 16 weeks after crush (151.3 6 20.4) (n 5 7) and that in the intact nerves (454.6 6 48.5) (n 5 5). We found no correlation between number of myelinated fibers and survival time 5–16 weeks after crush. The mean diameter of the myelinated axons (not including myelin) in the crushed nerves (1.35 6 0.04 µm) and intact nerves (1.33 6 0.08 µm) did not differ. The inset shows the distribution of fiber diameters 2 to 4 weeks after injury. The fiber diameters appear more heterogeneous at 2 to 4 weeks than at longer survival times. At 5–16 weeks the percentage of large fibers remained greater than in intact nerves; the percentage of axons larger than 2.4 µm was 5.07 6 0.47% at 5–16 weeks and 1.62 6 0.55% in intact nerves (t10 5 4.71, P , 0.001). Fungiform Taste Buds
tion. Stimuli were all at room temperature at the time of presentation, and mechanical stimulation would appear equally across all stimuli. Both the recovery time (F2,12 5 5.22, P 5 0.015) and the stimulus type (F4,12 5 14.91, P , 0.0005) affected the proportion of nerves responding. There were more responses to water following a greater variety of stimuli by the 4- to 5-week group. Five of the experimental nerves responding to water were in the 4- to 5-week recovery group (n 5 8), compared to two in the 6- to 7-week group (n 5 5), and two in the 8-week group (n 5 7). When the experimental animals were considered as a group, the greatest number of water responses, 38%, followed sucrose stimulation. Only 12% of the total NaCl presentations were followed by water responses, 5% of the sodium acetate presentations, 8% of the KCl presentations, and 7% of HCl presentations. When the experimental animals were divided according to recovery time, only the 4- to 5-week group responded significantly more often following sucrose stimulation than did the intact controls (x2 5 4.83, P 5 0.028) (Fig. 3).
Figure 6 shows the mean taste bud area on the crush or experimental side compared with the intact side for different survival times. Two to 3 weeks after nerve crush, the mean area of the Ca21-ATPase-stained core of fungiform taste buds ipsilateral to the crush was reduced to about 52% (731.9 6 98.6 µm2) of that on the intact, contralateral side (1391 6 97.4 µm2) (t4 5 31.55, P , 0.001). By 4 weeks, the taste bud area on the crush side had increased such that there were no longer significant differences between the crush and intact sides. There were no significant differences in taste bud number between the crush and intact sides at any survival time (Table 1). DISCUSSION
We found that the chorda tympani nerve reinnervates the tongue after crush injury with significantly fewer fibers. These fibers are sufficient to support normal taste bud size and number. During recovery, the overall responses of the injured nerve to salts and HCl
342
CAIN, FRANK, AND BARRY
FIG. 4. Electron micrographs of cross sections through the chorda tympani nerve: (a) Two weeks after nerve crush, a Schwann cell tube may contain a macrophage with numerous lysosomes filled with lipid (asterisk) or myelin whorls (arrowhead). (b) Ten weeks after CT nerve crush, myelinated and unmyelinated (arrowheads) fibers with normal morphology are seen. However, there are many fewer myelinated fibers (c) than in the intact CT (d) from the other side of the same animal.
are remarkably similar to those of controls, but responses to sucrose and water are notably dissimilar to those of controls. These dissimilarities may result from a differential rate of return of fibers and/or transduction mechanisms. We propose that the recovery of the peripheral gustatory nervous system after damage to the chorda tympani nerve may follow a sequence similar to that observed during development. Early Characteristics Our morphological studies of the chorda tympani nerve revealed substantial myelin degeneration 2 weeks after the crush. Jang and Davis (34) report a similar
time course; they demonstrated degenerating myelin from CT parasympathetic fibers 3 weeks after section. In addition, we found the mean area of fungiform taste buds on the experimental side 2–3 weeks after injury to be comparable to the 54% reduction seen in fungiform taste bud size following CT section (6). We found no decrease in taste bud numbers with chorda tympani nerve crush, a finding consistent with reports on taste bud denervation following chorda tympani or chorda tympani–lingual section (6, 49). However, Oakley et al. (48) found that 26% of hamster fungiform taste buds had disappeared 3 weeks after chorda tympani–lingual denervation.
343
CHORDA TYMPANI RECOVERY
TABLE 1 Mean Number (6SEM) of Taste Buds on the Posterior Half of the Fungiform Taste Bud Field Following Unilateral CT Nerve Crush Survival time
CT crush Intact
2–3 Weeksa
4–5 Weeksb
6–8 Weeks c
14.2 6 1.1 16.6 6 1.2
15.2 6 1.8 15.4 6 1.0
15.0 6 1.1 14.8 6 1.2
a
n 5 5. n 5 5. c n 5 4. b
FIG. 5. The average numbers of myelinated fibers in the intact and CT crush nerves, 5–16 weeks after injury, are shown arrayed by diameter. The inset shows the average number of fibers 2–4 weeks after injury.
We found that taste solutions that normally elicit near-saturating CT responses did not elicit discernible responses from the CT fibers that may have regenerated by 2 weeks. Brief multifiber responses were seen 3 weeks after injury, but neither the baseline (spontaneous) activity nor the responses were stable and both decreased with time. In the gerbil chorda tympani, multifiber responses recorded in the middle ear were demonstrated at 2 weeks after chorda tympani crush just proximal to its union with the lingual nerve (14).
FIG. 6. The mean taste bud area (µm2) on the control versus CT crush sides of the animals: 2–3, 4–5, and 6–8 weeks after unilateral injury. The single asterisk indicates a significant difference (P , 0.05) between the mean taste bud area of the control and CT crush sides at the same survival time.
The differences in the recovery times in the gerbil and hamster data may be due in part to the different distances to be traversed by the recovering axons. Functional recovery in the somatosensory system following peripheral nerve crush varies from slight to moderate at early stages after injury (24, 39). For example, Fugleholm et al. (24) were able to elicit action potentials with electrical nerve stimulation 16 days after crushing the afferent fibers of the cat tibial nerve; however, the current needed to elicit a discernible response was two orders of magnitude greater than that necessary to elicit a response in a normal nerve. Recovery Four months after injury, we found fewer myelinated fibers than normal in the CT. Thirty-seven percent of normal hamster CT fibers are myelinated and the parasympathetic component of the CT is primarily unmyelinated (34). Ninety-two percent of myelinated CT fibers are sensory; thus, most of the lost myelinated fibers in our study were sensory. In comparison, the number of myelinated fibers in the recovered mouse peroneal nerve was unaltered after a single crush (25), and the number of myelinated fibers in tributaries of the sciatic nerve increased after crush injury or transection and reapproximating the nerve (35, 36). The increase may be the result of sprouting by the recovering nerves. We observed neither morphological signs of sprouting nor an increase in the number of myelinated fibers, although sprouting of unmyelinated fibers was apparent at early stages of regeneration. The reduction in the number of CT myelinated fibers may be the result of ganglion cell death, or perhaps the failure of some myelinated axons to regenerate completely, although we did not examine the nerve at the crush site. Following CT section in the hamster, Whitehead et al. (57) found an 18% decrease (not significant) in the number of geniculate ganglion cells with CT processes. We noticed that axonal retrograde transport did not match control levels after 5 months recovery in that study
344
CAIN, FRANK, AND BARRY
(57). In other sensory nerves, the percentage of ganglion cell loss resulting from nerve section is typically 15–30% (for review see Ref. 2). Usually the effects of nerve crush are less severe, and thus the loss of 67% of myelinated fibers in the regenerated chorda tympani is unusual compared to other systems. Although the mean patterns of responses in recovering and normal nerves were similar, the variability of responses to sucrose in the regenerated nerve was significantly greater than that of normal animals. Normal patterns of response from regenerated fibers in gerbils (14) and cats (52) were reported. Cheal et al. (14) suggested that in gerbils, all receptor types and transduction mechanisms were functional. However, ‘‘fewfiber’’ responses to sucrose varied widely early in recovery and some regenerated fibers had low responses to sucrose (14). In contrast to responses to sucrose, we found that responses to NaCl by regenerated nerves showed minor differences in variance from the normal nerves and showed amiloride sensitivity at all survival times. Two possible explanations for the differences in recovery of sucrose and NaCl responses are: (i) sucrose-sensitive fibers may make more variable functional connections in the first 8 weeks after nerve injury, or (ii) reconstitution of transduction pathways for sucrose may vary during this period. The magnitude of the whole nerve response is directly related to the number of fibers present in the nerve at the time of recording (10, 22, 32). Based on response profiles, the afferent population of the normal chorda tympani nerve is composed of subpopulations of fibers (23). The magnitude of the whole nerve response to each stimulus type should primarily reflect the contribution of the appropriate subpopulations. Data from other species indicate that after regeneration, there is recovery of fiber specificity (14, 52). Thus, greater relative magnitude and low variability in response to one type of taste stimulus as compared to another during recovery may be due to differential recovery of functional subpopulations. In the normal hamster CT, fibers responding best to Na1/Li1 stimulation (N fibers) outnumber fibers responding best to sweet stimuli about 2 to 1 (23). In addition, the responses of some N and HCl-best (H) fibers to sweet stimuli are depressed below baseline levels in normal animals. The sucrose-induced depression of the whole nerve response below baseline levels that we observed may be due to the depression in activity of N or H fibers. This would indicate that the fibers present during the early stages of recovery were predominantly Na1/Li1 and/or acid-best and may dominate the whole nerve response. Sucrose-best fibers, as a class, may recover more slowly than N and H fibers. Smaller diameter, slower conducting, myelinated sensory fibers in the sciatic nerve recover significantly more slowly than the larger
diameter, faster conducting, myelinated sensory fibers (18). However, in the mouse peroneal nerve there is no evidence that one class of fibers recovers more slowly than another (25). Our data indicate that the diameter of many of the early returning myelinated fibers is larger than those appearing later; however, the diameters of the three functional classes of afferent fibers in the hamster CT are unknown. Alternatively, the recovery time necessary to produce an effective and stable pathway for sucrose transduction may be longer than the recovery time to produce an efficacious Na1 transduction pathway. The transduction of sweet stimuli has been shown to be effected via a second-messenger pathway (1, 16, 17, 37, 38, 41, 54), whereas the specific transduction of sodium salts is mediated by an apical (amiloride-sensitive) cation channel: sodium influx directly depolarizes the receptor cell (29, 30, for review see Ref. 38). Like sucrose, the effects of serotonin are also mediated via a second-messenger pathway. Following irreversible serotonin receptor inactivation, new receptor density was significantly less than in controls after 14 days (50). While most fungiform taste buds in the hamster survive denervation by the chorda tympani and lingual nerves, they exhibit morphological changes and atrophy (5, 6, 48, 56), including loss of visible taste pores (46, 48). It is likely that in these denervated taste bud cells, receptors, protein synthesis, and second-messenger systems are also reduced relative to normal. At 8 days after denervation, taste bud cells retain their elongated shape; however, no intragemmal nerves are visible, and the taste bud cells are intermediate in electron density, lack well-developed smooth and rough endoplasmic reticula, and possess more rounded nuclei with evenly dispersed chromatin (56). After 21 days, denervated taste buds show light and dark cells, but are half the size of controls (6, 56). Following CT nerve regeneration, functional taste receptor cells could arise from these atrophied taste bud cells and/or new cells may be produced by stem cells in the taste bud. In either case, the synthesis and assembly of the complex G-protein and second-messenger systems required for sucrose transduction may take more time than the production of apical sodium channels. The CT water response has been shown to be a function of adaptation to preceding stimuli, including the ionic constituents of saliva, in rats and cats (8, 42, 43). Earlier reports included the hamster among species that did not show an excitatory response to water (10). Our results show that a small proportion of control hamster CTs responded to the water rinse. However, we found that a greater proportion of nerves responded to water, and responded following a wider range of stimuli, during the early stages of recovery. In young rats, the responses to the water rinse following sugar stimulation (‘‘off responses’’) were larger and more
345
CHORDA TYMPANI RECOVERY
frequent than in adults (27). This suggests that the recovery of sugar-processing mechanisms after CT damage may follow a pattern similar to that which occurs in development.
system may follow a progression of events leading to a mature response. The variability of response to sucrose during recovery after CT injury may be due in part to similar events.
Recovery May Recapitulate Ontogeny In contrast to the response to sucrose, Hill (31) found the week-old hamster to be highly sensitive to Na1 relative to NH4Cl and found that the Na1 response was strongly inhibited by amiloride. As the hamster matures, Na1 sensitivity (relative to NH41) decreases (31). The suppression of the Na1 response by amiloride early in hamster CT recovery and during development demonstrates that NaCl transduction via the amiloridesensitive cation channel (29, 30, 38) is present and functioning early in both processes. Recovery of taste cells following denervation might recapitulate the later stages of taste bud development (47). Our data offer evidence on a number of levels that recovery of the peripheral taste system following chorda tympani nerve injury may follow a developmental sequence. Denervated taste buds (48, 56) are similar to the developing rat fungiform taste bud at Farbman’s (20) stages 3 and 5. Elongate epithelial cells and no discernible type I and type II (dark and light, respectively) cells are characteristics of the taste bud 8 days after denervation (48, 56) and at stage 3 of development (20). Type I and II cells are seen at stage 5 (20) and at 21 days after denervation (6, 56) although both the papilla and the taste bud are smaller than in the control (6, 48, 56) or adult (20). The time-dependent increase in consistently positive responses to sucrose in the regenerated hamster CT is similar to that seen during development (31). In the regenerated CT, the frequent responses to water rinses, particularly following sucrose, are similar to those observed in young rats (27). The need to express a variety of receptors in order to effectively transduce sugar stimuli (17, 21, 31, 51, 55) may contribute to the variabiilty of responses seen early in development and regeneration. Similarly in another system, serotonin receptor subtypes require variable times to recover following experimental inactivation (50). Other G-protein-coupled systems show increases in receptor numbers during postnatal development (e.g., cannabinoid (44); opioid (for review see Ref. 40)). In contrast, as early as Postnatal Day 3 in rats, mRNA expression for the cannabinoid receptor was at adult levels (44). Thus, the expression of receptor mRNA may precede receptor proliferation. While these data do not address the number of receptors that are necessary and sufficient for transduction, establishment of a mature number of receptors may take time. In addition, functional delays may also be the result of changes in effector coupling during development (40). These reports indicate that development of a second-messenger
ACKNOWLEDGMENTS We thank Jonathon Clive for assistance with statistical analysis. We thank David Larson and Connie Gillies for technical assistance. This work was supported in part by Grants T32-DC00025-06 (P.C.), R01-DC00058 (M.E.F.), and P50-DC00168 (M.A.B.) from NIDCD.
REFERENCES 1.
2.
3.
4.
5. 6.
7.
8.
9. 10. 11.
12. 13.
14.
15.
16.
17.
AKABAS, M. H. 1991. Mechanisms of chemosensory transduction in taste cells. In International Review of Neurobiology (J. R. Smythies and R. J. Bradley, Eds.) Vol. 32, pp. 242–279. Academic Press, San Diego. ALDSKOGIUS, H., J. ARDVIDSSON, AND G. GRANT. 1985. The reaction of primary sensory neurons to peripheral nerve injury with particular emphasis on transganglionic changes. Brain Res. Rev. 10: 27–46. ANDO, T., H. FUJIMOTO, H. MAYAHARA, H. MIYAJIMA, AND K. OGAWA. 1981. A new one-step method for histochemistry and cytochemistry of Ca21-ATPase activity. Acta Histochem. Cytochem. 14: 705–726. BARRY, M. A. 1992. Ecto-calcium dependent ATPase activity of mammalian taste bud cells. J. Histochem. Cytochem. 40: 1919– 1928. BARRY, M. A., AND M. E. FRANK. 1992. Response of the gustatory system to peripheral nerve injury. Exp. Neurol. 115: 60–64. BARRY, M. A., AND L. D. SAVOY. 1993. Persistence and calciumdependent ATPase staining of denervated fungiform taste buds in the hamster. Arch. Oral Biol. 38(1): 5–15. BARRY, M. A., D. C. LARSON, AND M. E. FRANK. 1993. Loss and recovery of sodium-salt taste following bilateral chorda tympani nerve crush. Physiol. Behav. 53: 75–80. BARTOSHUK, L. M. 1965. Effects of Adaptation on Responses to Water in Cat and Rat. Ph.D. Dissertation. Brown University, Providence, RI. BEIDLER, L. M. 1953. Properties of chemoreceptors of tongue of rat. J. Neurophysiol. 16: 595–607. BEIDLER, L. M., I. Y. FISHMAN, AND C. W. HARDIMAN. 1955. Species differences in taste responses. Am. J. Physiol. 181: 235–239. BLACKBURN, C. W., AND P. A. BRAMLEY. 1989. Lingual nerve damage associated with the removal of lower third molars. Br. Dent. J. 167: 103–107. BULL, T. R. 1965. Taste and the chorda tympani. J. Laryngol. Otol. 79: 479–493. CAIN, P., AND M. A. BARRY. 1993. Regeneration and functional response in the chorda tympani nerve in the golden Syrian hamster (Mesocricetus auratus). Chem. Senses 18: 263. CHEAL, M., W. P. DICKEY, L. B. JONES, AND B. OAKLEY. 1977. Taste fiber responses during reinnervation of fungiform papillae. J. Comp. Neurol. 172(4): 627–646. CHEAL, M., AND B. OAKLEY. 1977. Regeneration of fungiform taste buds: Temporal and spatial characteristics. J. Comp. Neur. 172(4): 609–626. CUMMINGS, T. A., AND S. C. KINNAMON. 1992. In situ recording from hamster taste cells: Responses to sweet stimuli and cAMP. Chem. Senses 17: 607–608. CUMMINGS, T. A., J. POWELL, AND S. C. KINNAMON. 1993. Sweet
346
18.
19. 20. 21.
22. 23.
24.
25.
26.
27.
28. 29.
30. 31. 32.
33. 34.
35.
36. 37. 38.
39.
CAIN, FRANK, AND BARRY taste transduction in hamster cells: Evidence for the role of cyclic nucleotides. J. Neurophysiol. 70(6): 2326–2336. DEVOR, M., AND R. GOVRIN-LIPPMANN. 1979. Selective regeneration of sensory fibers following nerve crush injury. Exp. Neurol. 65: 243–254. ERICKSON, R. P. 1966. Nontraumatic headholders for mammals. Physiol. Behav. 1: 97–98. FARBMAN, A. I. 1965. Electron microscope study of the developing taste bud in rat fungiform papilla. Dev. Biol. 11: 110–135. FAURION, A., AND C. VAYSETTES-COURCHAY. 1990. Taste as a highly discriminative system: A hamster intrapapillar single unit study with 18 compounds. Brain Res. 512: 317–332. FRANK, M. E. 1973. An analysis of hamster afferent taste nerve response functions. J. Gen. Physiol. 61: 588–618. FRANK, M. E., S. L. BIEBER, AND D. V. SMITH. 1988. The organization of taste sensibilities in hamster chorda tympani nerve fibers. J. Gen. Physiol. 91: 861–896. FUGLEHOLM, K., H. SCHMALBRUCH, AND C. KRARUP. 1994. Early peripheral nerve regeneration after crushing, sectioning, and freeze studied by implanted electrodes in the cat. J. Neurosci. 14(5): 2659–2673. GIANNINI, C., A. C. LAIS, AND P. J. DYCK. 1989. Number, size and class of peripheral nerve fibers regenerating after crush, multiple crush, and graft. Brain Res. 500: 131–138. GRANT, R., S. MILLER, D. SIMPSON, P. J. LAMEY, AND I. BONE. 1989. The effect of CT section on ipsilateral and contralateral salivary secretion and taste in man. J. Neurol. Neurosurg. Psychiatry 52: 1058–1062. HARADA, S., S. MAEDA, AND Y. KASAHARA. 1989. Sugar response characteristics in the rat chorda tympani nerve during development. Chem. Senses 14: 707–708. HARPER, H. W. 1979. Improved flow-chamber for taste stimulation. Physiol. Behav. 22: 1033–1035. HECK, G. L., S. MIERSON, AND J. A. DESIMONE. 1984. Salt taste transduction occurs through an amiloride-sensitive sodium transport pathway. Science 223: 403–405. HETTINGER, T. P., AND M. E. FRANK. 1990. Specificity of amiloride inhibition of hamster taste responses. Brain Res. 513: 24–34. HILL, D. L. 1988. Development of chorda tympani nerve taste responses in the hamster. J. Comp. Neurol. 268: 346–356. HILL, D. L., C. M. MISTRETTA, AND R. M. BRADLEY. 1982. Developmental changes in taste response characteristics of rat single chorda tympani fibers. J. Neurosci. 2: 782–790. HOUSE, H. P. 1963. Early and late complications of stapes surgery. Arch. Otolaryngol. 78: 606–613. JANG, T., AND B. J. DAVIS. 1987. The chorda tympani and glossopharyngeal nerves in the adult hamster. Chem. Senses 12: 381–395. JENQ, C., AND R. E. COGGESHALL. 1984. Effects of sciatic nerve regeneration on axonal populations in tributary nerves. Brain Res. 295: 91–100. JENG, C., AND R. E. COGGESHALL. 1984. Regeneration of axons in tributary nerves. Brain Res. 310: 107–121. KINNAMON, S. C. 1988. Taste transduction: A diversity of mechanisms. Trends Neurosci. 11: 491–496. KINNAMON, S. C., AND T. A. CUMMINGS. 1992. Chemosensory transduction mechanisms is taste. Annu. Rev. Physiol. 54: 715–731. KINNMAN, E., AND Z. WISENFELD-HALIN. 1993. Time course and characteristics of the capacity of sensory nerves to reinnervate skin territories outside their normal innervation zones. Somatosensory Motor Res. 10(4): 445–454.
40.
41. 42.
43.
44.
45. 46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
LESLIE, F. M., AND S. E. LOUGHLIN. 1992. Development of multiple opioid receptors. In Development of the Central Nervous System: Effects of Alcohol and Opiates (M. W. Miller, Ed.), pp. 255–283. Wiley–Liss, New York. MARGOLSKEE, R. 1993. The biochemistry and molecular biology of taste transduction. Curr. Opin. Neurobiol. 3: 526–531. MATSUO, R., AND T. YAMAMOTO. 1990. Taste nerve responses during licking behavior in rats: Importance of saliva in response to sweeteners. Neurosci. Lett. 108: 121–126. MATSUO, R., AND T. YAMAMOTO. 1992. Effects of inorganic constituents of saliva on taste responses of the rat chorda tympani nerve. Brain Res. 583: 71–80. MCLAUGHLIN, C. R., B. R. MARTIN, D. R. COMPTON, AND M. E. ABOOD. 1994. Cannabinoid receptors in developing rats: Detection of mRNA and receptor binding. Drug Alcohol Depend. 36: 27–31. MOON, C. N., AND E. W. PULLEN. 1963. Effects of chorda tympani section during middle ear surgery. Laryngoscope 73: 392–405. MYERS, W. E., T. P. HETTINGER, J. A. D’AMBROSIO, S. L. WENDT, C. B. PEARSON, M. A. BARRY, AND M. E. FRANK. 1994. Visualizing taste papillae in vivo with scanning electron microscopy of a high resolution cast. Chem. Senses 20(1): 1–8. OAKLEY, B. 1991. Neural–epithelial interactions in taste bud regeneration. In Regeneration of Vertebrate Sensory Receptor Cells (J. Whelan, Ed.), pp. 277–287. Ciba Foundation Symposium, London. OAKLEY, B., A. LAWTON, D. R. RIDDLE, AND L.-H. WU. 1993. Morphometric and immunocytochemical assessment of fungiform taste buds after interruption of the chorda-lingual nerve. Microsc. Res. Tech. 26: 187–195. OLIVER, S., AND M. C. WHITEHEAD. 1992. Morphometry and cellular dynamics of denervated fungiform taste buds in the hamster. Chem. Senses 17(5): 529–542. PINTO, W., AND G. BATTAGLIA. 1994. Comparative recovery kinetics of 5-hydroxytryptamine 1A, 1B, and 2A receptor subtypes in rat cortex after receptor inactivation: Evidence for differences in receptor production and degradation. Mol. Pharmacol. 46: 1111–1119. REHNBERG, B. G., T. P. HETTINGER, AND M. E. FRANK. 1990. The role of sucrose-sensitive neurons in ingestion of sweet stimuli by hamsters. Physiol. Behav. 48: 459–466. ROBINSON, P. P. 1989. The reinnervation of the tongue and salivary glands after lingual nerve injuries in cats. Brain Res. 483: 259–271. ROBINSON, P. P., AND P. A. WINKLES. 1991. The number and distribution of fungiform papillae and taste buds after lingual nerve injuries in cats. Archs. Oral Biol. 36(12): 885–891. STRIEM, B. J., U. PACE, U. ZEHAVI, M. NAIM, AND D. LANCET. 1989. Sweet tastants stimulate adenylate cyclase coupled to GTPbinding protein in rat tongue membranes. Biochem. J. 260: 121–126. TINTI, J. M., AND C. NOFRE. 1991. Why does a sweetener taste sweet? A new model. In Sweeteners: Discovery, Molecular Design, and Chemoreception (D. E. Walters, F. T. Orthoefer, G. E. Dubois, Eds.), pp. 206–213. Am. Chem. Soc., Washington, DC. WHITEHEAD, M. C., M. E. FRANK, T. P. HETTINGER, L.-T. HOU, AND H.-D. NAH. 1987. Persistence of taste buds in denervated fungiform papillae. Brain Res. 405: 192–195. WHITEHEAD, M. C., S. T. MCGLATHERY, AND B. G. MANION. 1995. Transganglionic degeneration in the gustatory system consequent to chorda tympani damage. Exp. Neurol. 132: 239–250. ZUNIGA, J. R., N. CHEN, AND I. J. MILLER. 1994. Effects of chorda-lingual nerve surgery and repair on human taste. Chem. Senses 19(6): 657–665.