Recovery of two independent sweet taste systems during regeneration of the mouse chorda tympani nerve after nerve crush

Experimental manipulation

All experimental procedures were approved by the committee for Laboratory Animal Care and Use at Kyushu University (Fukuoka, Japan) and were in accordance with the Animal Care Guidelines of Kyushu University. Subjects were adult male and female C57BL/6NCrj mice (Charles River Japan, Tokyo, Japan) that were 8–20 weeks of age, ranging in weight from 20 to 32 g. At 8–10 weeks of age, mice were divided into five groups, one intact control group and four nerve-crush groups, which were later examined at 2, 3, 4 and ≥ 5 weeks after their bilateral CT nerves were crushed. Animals of the four nerve-crush groups were used to obtain nerve recordings at the time-points of 14–16, 21–23, 28–30 and ≥ 35 days (< 50 days) after the CT nerve crush and regeneration. The procedure used to crush the CT nerve crush was described previously (Yasumatsu et al., 2003). Animals were anesthetized with pentobarbital sodium (50–60 mg/kg, i.p.; Somnopentyl, Schering-Plough Co., Kenilworth, NJ, USA). The bilateral CT nerves were exposed at ∼5 mm rostrally apart from their entry to the bulla and repeatedly crushed at a single point with number 5 forceps until only a thin strand of nerve sheath remained (5–10 crushes).

Recordings of responses from intact and regenerated chorda tympani nerves

At selected times after the nerve crush, the experimental groups were reoperated under pentobarbital anesthesia (again 50–60 mg/kg) to expose the regenerated nerve and to dissect a single fiber or a few fibers from the nerve for electrophysiological recording. The procedures for dissection and recording of responses from the regenerated nerve and fibers were the same as those used previously (Ninomiya, 1998; Yasumatsu et al., 2003). Under pentobarbital anesthesia, the trachea of each mouse was cannulated and the mouse was then fixed in the supine position with a head holder to allow dissection of the CT nerve. The right CT nerve was dissected free from surrounding tissues after removal of the pterygoid muscle and cut at the point of its entry to the bulla. For whole-nerve recording, the entire nerve was placed onto an Ag/AgCl electrode. For single-fiber recording, a single fiber or a few fibers of the nerve were teased apart with a pair of needles and lifted onto the electrode. An indifferent electrode was placed in nearby tissue. Neural activities resulting from chemical and electrical stimulations (ESs) of the tongue were fed into an amplifier (K-1; Iyodenshikogaku, Nagoya, Japan), and monitored on an oscilloscope and audiomonitor. Whole-nerve responses were integrated with a time constant of 1.0 s and recorded on a computer for later analysis using a PowerLab system (PowerLab/sp4; AD Instruments, Australia).

Chemical and electrical stimulations of the tongue

The anterior half of the tongue was enclosed in a flow chamber made of silicone rubber (Ninomiya & Funakoshi, 1981). Solutions were delivered into the chamber by gravity flow and flowed over the tongue for a controlled period. Solutions used as chemical stimuli were: 0.1 m NH₄Cl, 0.1 m NaCl, 0.01 m HCl, 0.02 m quinine HCl, 0.01–1 m sucrose (Suc), 0.1–8 mm SC45647 (SC), 0.01–1 m glucose (Glc), 0.5 m maltose (Mal), 0.5 m fructose (Fru), 1 m sorbitol, 20 mm saccharin Na (Sac), 0.3 m l-alanine and 0.5 m l-proline (Wako Pure Chemicals Industries, Osaka, Japan). These chemicals were dissolved in distilled water and used at ∼24 °C. For whole-nerve recording in the second week, the response of the CT to ES applied to the tongue was tested because the 2-week group responded to chemical stimuli only slightly or not at all. For ES, an Ag/AgCl electrode was placed on the inside wall of the flow chamber or was placed directly on the tongue when required. The Ag/AgCl indifferent electrode was positioned in nearby tissue. Anodal current was passed through the tongue from a ramp current generator (Densi-Sekkei, Nagoya, Japan). The bathing medium used during the current stimulation was 1 mm NaCl (Ninomiya & Funakoshi, 1981). As found in our previous study (Yasumatsu et al., 2003), anodal current with an intensity of 20 µA (rate of raise, 100 µA/s; duration, ∼20 s) provoked robust responses in the CT at 2 weeks after nerve crush. Therefore, we used the response to anodal current as the standard for calculating relative magnitudes of response to chemical stimuli at each experimental time-point. During chemical stimulation of the tongue, the test solution flowed for ∼30 s at the same flow rate as the distilled water used for rinsing the tongue (∼0.1 mL/s). The tongue was rinsed with distilled water for an interval of ∼1 min between successive stimulations. The order of chemical stimulation for whole-nerve recording was: 0.1 m NH₄Cl, 0.01–1 m Suc, 0.1–8 mm SC, 0.1 m NH₄Cl, 0.01–1 m Glc, 0.5 m Mal, 0.5 m Fru, 1 m sorbitol, 20 mm Sac, 0.3 m l-alanine and 0.5 m l-proline. To examine Gur inhibition of the CT responses, the tongue was treated with 30 µg/mL (∼7.13 µm) Gur dissolved in 5 mm phosphate buffer (pH 6.8; made with Na₂HPO₄.12H₂O and NaH₂PO₄.2H₂O) for 5 min as described in our previous studies (Ninomiya & Imoto, 1995; Ninomiya et al., 1999). After Gur treatment, responses to chemical and electrical stimulation were recorded. The stability of each preparation was monitored by periodic application of ES or 0.1 m NH₄Cl. A recording was considered to be stable when ES or NH₄Cl response magnitudes at the beginning and end of each stimulation series deviated by no more than 15%. Only responses from stable recordings were used in the data analysis. The order for single-fiber recording was 0.5 m Suc, 0.1 m NaCl, 0.01 m HCl and 0.02 m quinine HCl. When a single fiber responding to 0.5 m Suc was found, responses of the fiber to Suc before and after the Gur treatment were examined. In both whole-nerve and single-fiber recording experiments, Gur was applied to the tongue only once and therefore data from only one preparation were obtained from each animal.

Data analysis

In the analysis of whole-nerve responses, the magnitude of the integrated responses from 5 to 25 s (every 5 s) after stimulus onset was measured and averaged. The averaged response magnitude for each test stimulus was normalized to the response magnitude for ES (20 µA anodal current), which was taken as unity (1.0), and this relative response was used for statistical analysis. Repeated anova and Student's t-test were performed to evaluate the effect of Gur on taste responses statistically. In the analysis of single-fiber responses, single fibers were identified by their uniform spike height, singular wave form and intervals between contiguous spikes (Ninomiya, 1998; Kawai et al., 2000). Frequency/time histograms of impulse discharges before, during and after chemical stimulation of the tongue were calculated by means of a spike-analysis system (SAS-1; Iyodensikogaku). For data analysis, we used the net average frequency for the first 10 s after the stimulus onset, which was obtained by subtracting the spontaneous frequency for the 10-s period before stimulation. We calculated the percentage of Gur inhibition of the response to 0.5 m Suc for each fiber and used the 60% control level to classify the fibers as either GS-type (< 60%) or GI-type (≥ 60%). Our previous study showed that the 60% level was the most appropriate boundary for classifying the two groups of fibers with different Gur sensitivity in mice (Ninomiya et al., 1997).

CRNA probes for in-situ hybridization

Each mouse was anesthetized with pentobarbital sodium (50–60 mg/kg, i.p.; Somnopentyl, Schering-Plough Co.), killed by cervical dislocation and the tongue rapidly removed. After washing the tongues with a normal extracellular solution (140 mm NaCl, 5 mm KCl, 1 mm MgCl₂, 1 mm CaCl₂, 5 mm Glc, 1 mm sodium pyruvate, 10 mm HEPES-NaOH, pH 7.4), 0.2 mL of normal extracellular solution containing 1.0 mg/mL elastase (Roche, Indianapolis, IN, USA) was injected between the epithelium and muscle layers of the tongue. After incubation in normal extracellular solution continuously bubbled with 95% O₂ and 5% CO₂ for 15–20 min at 26 °C, the epithelial sheet containing fungiform papillae (FP) was peeled off from the underlying muscle. The epithelium was washed with normal extracellular solution, pinned serosal side up in a silicone dish and incubated in divalent cation-free extracellular solution containing 2 mm EDTA for 20 min at room temperature (26°C). Taste buds were individually removed from fungiform papillae (FP) by sucking with a 100 µm transfer pipette. One hundred taste buds from the taste papillae were transferred to an Eppendorf tube containing 100 µL (1 volume) of lysis buffer (4 m guanidine thiocyanate, 25 mm trisodium citrate, pH 7.0, 0.5% sodium N-lauroyl-sarcosine, 0.1 m 2-mercaptoethanol) and homogenized. Yeast transfer RNA (1.0 µg) was used as a carrier. Sequentially, 0.1 volume of 2 m Na-acetate (pH 4.0), 1 volume of acidic phenol and 0.2 volume of chloroform : isoamyl alcohol (49 : 1) were added, the mixture being vortexed after each addition. After incubation at 4 °C for 15 min and centrifugation at high speed, the aqueous phase was recovered and RNA was precipitated with 1 volume of isopropanol. After centrifugation, the pellet was resuspended in 1 volume of lysis buffer and RNA was reprecipitated with isopropanol. After two washings with 75% ethanol, the pellet was dried, dissolved in water. A cDNA was generated by reverse transcription [oligo (dT)₁₂₋₁₈ primer] with the superscript preamplification system (Gibco/BRL, Gaithersburg, MD, USA). Polymerase chain reactions (PCRs) were carried out with an equivalent of 10 taste buds per reaction. Genomic DNA did not contribute to the signal as suggested by two protocols. In the first, RNA was treated in parallel in the presence and absence of reverse transcriptase, and the material was then used for PCR. In the absence of reverse transcriptase, there was no amplification of fragments of the expected size. Primers were chosen to span one or more introns in order to exclude confusion with amplified fragments from genomic DNA. PCR reactions led in this case to the amplification of two bands, either specific for the genomic DNA (characterized by a longer size owing to the presence of at least one intron) or for the reverse-transcribed complementary DNA. The primers used for DNA amplification were as follows: T1r3, 5′-TGCTGCTATGACTGCGTGGAC-3′ and 5′- AAGAAGCACATAGCACTTGGG-3′ (NM_031872, 910 bp); and α-gustducin, 5′-AGATGGGAAGTGGAATTAGTTCAGA-3′ and 5′-GCTCAGAAGAGCCCACAGTCTTTGA-3′ (X65747, 1069 bp). PCR was performed on PE9700 with the following conditions: 95 °C for 5 min (one cycle); 94 °C for 30 s, 58 °C for 30 s, 72 °C for 120 s (40 cycles); and 72 °C for 5 min (one cycle). The PCR solution contained 10 mm Tris/HCl, 50 mm KCl, 1.5 mm MgCl₂, 0.5 µm of each primer, 200 µm deoxyribonucleoside 5′-triphosphate and 0.05 unit/mL of Ex Taq polymerase (Takara Bio, Otsu, Japan). These DNA fragments were purified and cloned into the pGEM T-Easy vector (Promega, Madison, WI, USA), confirmed by direct sequencing and digested with appropriate restriction enzymes. Biotin-labeled antisense RNA probes were generated by in-vitro transcription using a digoxigenin-RNA labeling mix and SP6 or T7 RNA polymerase (Roche).

in-situ hybridization

Frozen blocks of the dissected anterior and posterior parts of the tongue embedded in the OCT compound (Sakura Finetechnical, Tokyo, Japan) were sectioned into 5–7 µm thick slices, which were mounted on silane-coated glass slides. The cryosections were fixed in 4% paraformaldehyde in phosphate-buffered saline for 10 min at room temperature, treated twice with 0.1% diethyl pyrocarbonate in phosphate-buffered saline for 15 min, washed with 5 × SSC (standard saline citrate) for 15 min at room temperature and then prehybridized in hybridization buffer consisting of 50% formamide, 5 × SSC, 5 × Denhardt's solution, 500 µg/mL denatured salmon testis DNA, 250 µg/mL denatured baker's yeast tRNA and 1 mm dithiothreitol for 1 h at room temperature. Hybridization was carried out in a hybridization buffer to which was added 200 ng/mL antisense riboprobe for 18 h at 58 °C. After hybridization, sections were washed twice in 5 × SSC for 5 min each and twice in 0.2 × SSC for 30 min each at 65 °C. Subsequently, the sections were immersed in Tris-buffered saline consisting of 50 mm Tris/HCl (pH 7.5) and 150 mm NaCl for 5 min at room temperature, put in the blocking solution containing 0.5% blocking reagent (Roche) in Tris-buffered saline for 30 min and incubated with anti-digoxigenin Fab fragments conjugated with alkaline phosphatase (1 : 400 dilution in the blocking solution) for 60 min at room temperature. After three washes of 5 min each in Tris buffer, consisting of 50 mm Tris/HCl (pH 7.5), 150 mm NaCl and 0.05% Tween 20, sections were immersed in alkaline phosphatase buffer, consisting of 100 mm Tris/HCl (pH 9.5), 100 mm NaCl and 50 mm MgCl₂, for 5 min. The signals were developed using nitroblue-tetrazolium chloride and 5-bromo-4-chloro-3-indolylphosphate as chromogenic substrates. The reaction was then stopped by rinsing the slides in Tris-EDTA (TE) buffer, after which they were mounted. The signal specificities of mRNA for each gene in the taste tissues were tested by using a sense probe as a negative control.

Reappearance of gurmarin sensitivity during chorda tympani nerve regeneration

At 2 weeks after the CT nerve crush, the regenerated nerve responded to anodal current but showed no significant response to 0.5 m Suc (Fig. 1). After 3 weeks of regeneration, the response to Suc reappeared but was not significantly inhibited by lingual treatment with Gur. At 4 weeks, the response to Suc increased and was significantly inhibited by Gur. Concentration-dependent responses for Suc before and after Gur are shown in Fig. 2. Statistically significant inhibition of the Suc response by Gur was not observed at 3 weeks (repeated anova; F_1,50 = 0.5; P > 0.05) but was clearly evident 4 weeks after the nerve crush (F_1,60 = 24.0; P < 0.001; t-test; P < 0.05 for 30 mm−1 m Suc). At ≥ 5 weeks, the concentration-dependent responses for Suc before and after Gur became similar to those in the intact control group (before Gur, F_1,60 = 1.23, P > 0.05; after Gur, F_1,60 = 0.23, P > 0.05) and the inhibition of responses to Suc by Gur was significant (F_1,60 = 14.5; P < 0.01 for the ≥ 5 week group; F_1,60 = 21.5; P < 0.001 for the control group).

Recovery of gurmarin-sensitive responses to various sweet compounds

Figure 3 shows the concentration-dependent responses for SC and Glc. As compared with the responses to Suc, the responses to SC were more strongly inhibited by Gur (by ∼80% of control), whereas the responses to Glc were more weakly inhibited (by ∼30% of control). Thus, the magnitude of Gur inhibition of CT responses varied among sweet compounds. In contrast, the recovery of Gur sensitivity in the responses to these compounds after the nerve crush appeared very similar to the recovery of Gur sensitivity in the response to Suc. Gur inhibition of the SC and Glc responses was not observed at 3 weeks (F_1,32 = 2.27, P > 0.05 for SC; F_1,24 = 0.79, P > 0.05 for Glc) but it became statistically significant at 4 weeks after the nerve crush (F_1,32 = 9.35, P < 0.05 for SC; F_1,24 = 5.60, P < 0.05 for Glc). At ≥ 5 weeks, the concentration-dependent responses for SC and Glc before and after Gur became similar to those in the intact control group (before Gur, F_1,32 = 0.12, P > 0.05 for SC; F_1,24 = 1.48, P > 0.05 for Glc; after Gur, F_1,32 = 0.55, P > 0.05 for SC; F_1,24 = 0.55, P > 0.05 for Glc), with significant inhibition of the responses to SC and Glc by Gur (≥ 5 week group, F_1,32 = 17.74, P < 0.01 for SC; F_1,24 = 10.72, P < 0.05 for Glc; control group, F_1,48 = 15.74, P < 0.01 for SC; F_1,36 = 27.48, P < 0.001 for Glc).

Figure 4 shows the responses to six sweet compounds before and after Gur treatment in intact and experimental mice at 3, 4 and ≥ 5 weeks after the nerve crush. As observed for Suc, SC and Glc, the responses to all six compounds were not significantly inhibited by Gur at 3 weeks (t-test, Ps > 0.05). At 3 weeks the magnitude of this GI taste response had already reached more than 50% of the GI response in intact mice. At 4 weeks, significant inhibition by Gur was observed in the responses to Mal, Sac, l-alanine and l-proline (Ps < 0.05), whereas inhibition for Fru and sorbitol was not yet significant (Ps > 0.05). At ≥ 5 weeks, the responses to all six compounds became indistinguishable from those in the control group (Ps > 0.05) and all showed clear inhibition by Gur (Ps < 0.05–0.01).

Kinetic analyses of concentration-dependent responses for the gurmarin-sensitive and gurmarin-insensitive components of the sucrose response during recovery after the nerve crush

We calculated the dissociation constant and maximum response (Lineweaver & Burk, 1934) on the basis of the concentration-dependent responses for the GS and GI components of the Suc response at four different concentrations (e.g. 0.1, 0.3, 0.5 and 1.0 m) from 3 to ≥ 5 weeks after the nerve crush. As shown in Table 1, for both the GS and GI components, the maximum response values gradually recovered to the control levels during this period. In contrast, during regeneration, the dissociation constant values of the GI component ranged from 0.84 to 0.89 and those of the GS component ranged from 0.42 to 0.51. In both the GI and GS components, the dissociation constant values did not substantially change (t-test, P > 0.05).

Single-fiber response analyses on gurmarin sensitivity at 4 weeks after the nerve crush

We examined the single-fiber responses of the regenerated CT to various taste stimuli at 4 weeks after the nerve crush. In total, we successfully recorded spike activities from 80 regenerated single fibers. Sixty-five of the 80 responded to some of the taste stimuli (0.1 m NaCl, 0.1 m KCl, 0.01 m HCl, 0.02 m quinine HCl and 0.5 m Suc), whereas the remaining 15 fibers did not respond to any of these taste stimuli. Of 65 taste-responsive fibers, 23 responded predominantly to Suc. Of these 23 fibers, we found that six fibers (GS-type) showed clear inhibition of the Suc response by Gur (to 0–32% of control), whereas the remaining 17 fibers (GI-type) did not (ranging from 69 to 114% of control) (Fig. 5). Therefore, at 4 weeks, the GS and GI response components were clearly segregated based on both whole-nerve integrated response and single-fiber response analyses.

Expression of T1R3 and gustducin during regeneration

Antisense riboprobes specific for T1R3 and gustducin mRNAs were applied to horizontal sections of the mouse fungiform papillae after the CT nerve crush. T1R3 and α-gustducin mRNAs were detected in some cells in taste buds of the fungiform papillae throughout the period tested (Fig. 6). T1R3 and α-gustducin mRNAs were clearly detected in circumvallate papillae. As reported in our previous studies (Shigemura et al., 2005; Yasumatsu et al., 2005b), taste buds in fungiform papillae were reduced in size and number at 2 weeks after the CT nerve crush. Some shrunken fungiform papillae or dermal indentations into the epithelium without taste buds were also detected. Although the size of taste buds was reduced, a few positive taste cells for T1R3 and α-gustducin were detected at this period. The intensity of the signals for T1R3 and α-gustducin was weaker than that of intact control mice. At 3 weeks, the size of taste buds increased (Shigemura et al., 2005) and hybridization signal for T1R3 and α-gustducin were more clearly detected. At 4 and 5 weeks, the expression patterns for T1R3 and α-gustducin were almost the same as those at 3 weeks. Staining using a sense probe was negative in control fungiform and circumvallate papillae, and regenerating fungiform papillae.