Mice

All animal procedures were approved by the Institutional Animal Care and Use Committee of Korea University (KOREA-2022-0142). Mice were housed in a temperature-controlled environment (22 °C) under a 12-h light-dark cycle with ad libitum access to water and a standard chow diet (SAFE R A03, France). T1r2-Cre knock-in mice were generated in our previous study.20ROSA26-lsl-tdTomato mice (JAX007908) were obtained from the Jackson Laboratory, and Pirt-Cre mice were provided by Xinzhong Dong at Johns Hopkins University. T1r2-tdTomato and Pirt-tdTomato mice were established by crossing ROSA26-lsl-tdTomato with T1r2-Cre knock-in mice or Pirt-Cre mice, respectively. Genotyping was confirmed by PCR amplification of genomic DNA extracted from tails using the MyTaq Extract-PCR kit (BIO-21127, Bioline, UK). The primer sequences were as follows: For T1r2-Cre, 5′-CAATGAGGCTGGGCATCGTCTAAG-3′ (forward) and 5′-CACCACTTGCAACTTGACTTTGAACTC-3′ (reverse); For Rosa26-lsl-tdTomato, 5′-CAACATGGCCGTCATCAAAGA-3′ (forward) and 5′-CTTGTACAGCTCGTCCATGCC-3′ (reverse); For Pirt-Cre, 5′-ATCCGTAACCTGGATAGTGAA-3′ (forward, specific to wild-type allele), 5′-CAACTTTGTGGTACCCGAAG-3′ (forward, specific to mutant allele), and 5′-TCCCTGGGACTCATGATGCT-3′ (reverse, for both alleles).

Surgical procedures

Adult mice (>8 weeks of age) were anesthetized with a mixture of 2,2,2-tribromoethanol and 2-methyl-2-butanol (intraperitoneal injection, 1:2 ratio, 400–500 μL per mouse). The GLx and CTx surgeries were performed as previously described.10 Briefly, a midline incision was made along the ventral neck for both surgeries. For GLx, the glossopharyngeal nerve was exposed by retracting the digastric muscles overlying the carotid sheath, isolated between the carotid arteries, and finally severed with fine scissors. For CTx, the digastric muscles were teased up, so that the submaxillary gland could be retracted. Then, the tympanic bulla was exposed by carefully removing connective tissue while avoiding blood vessels. After opening the bony shell of the tympanic bulla, the chorda tympani nerve was removed by gently pulling the malleus with fine forceps. Bilateral transections were performed to prevent potential contralateral nerve compensation. Sham surgeries consisting of nerve exposure and no actual cutting were conducted on control mice.

In vivo imatinib administration

For in vivo experiments, adult mice (>8 weeks of age) were intraperitoneally injected with imatinib mesylate (GC11759, GLPbio, Montclair, CA, USA). Imatinib was typically administered at 100 mg/kg, but in CTx-operated mice, the dose was halved due to their increased mortality at the standard dose.

In vivo experimental design in detail

For simple GLx experiments (Figs. 1 and 2), B6N or T1r2-tdTomato mice underwent the GLx operation as described above and were sacrificed at post-operation week 1, 2, or 4.

For imatinib treatment experiments (Fig. 3a–e), T1r2-tdTomato or B6N mice received daily injections of vehicle or imatinib daily for 12 consecutive days and were sacrificed on day 13.

For imatinib treatment experiments on GLxed mice for 2 weeks (Fig. 3f–i), GLx-operated T1r2-tdTomato mice were injected daily with vehicle or imatinib for 12 consecutive days, starting the day after GLx, and were sacrificed on day 13.

For early phase imatinib treatment experiments (Fig. 5a–e), T1r2-tdTomato mice received daily injection of vehicle or imatinib for 12 consecutive days, starting the day after GLx, and were sacrificed on the following day after a 15-day recovery period.

For late phase imatinib treatment experiments (Fig. 5f–j), T1r2-tdTomato mice were allowed to recover for 15 days following GLx, after which they received daily injection of vehicle or imatinib daily for 12 consecutive days and were sacrificed the following day.

For simple CTx experiments (Fig. 6a–d), B6N mice underwent the CTx operation as described above, and were sacrificed at post-operation week 1, 2, or 4.

For imatinib treatment experiments on CTxed mice for 2 weeks (Fig. 6e, f), T1r2-tdTomato mice were injected daily with vehicle or imatinib for 12 consecutive days, beginning the day after CTx, and sacrificed on day 13.

For simple GLx experiments (Fig. 7), T1r2-tdTomato, Pirt-tdTomato or B6N mice underwent the GLx operation as described above and were sacrificed at post-operation week 2, 3, or 4.

Tissue immunohistochemistry

Mice were euthanized and perfused with 0.1 mol/L phosphate-buffered saline (PBS) followed by 4% paraformaldehyde (PFA) in PBS. The entire tongue was dissected, post-fixed in 4% PFA at 4 °C overnight, and then transferred to 30% sucrose in PBS at 4 °C for several days until the tissue sank. Samples were embedded in Tissue-Tek OCT (4583, Sakura Finetek, Torrance, CA, USA). Using a cryostat microtome (CM3050S, Leica, Deer Park, IL, USA), anterior tongue sections were cut coronally at 45-μm thickness and collected free-floating in PBS. Posterior tongue sections were cut at 12-μm thickness and mounted directly onto slide glass (HMA-S9911, MATSUNAMI, Japan). The sections were then blocked in 5% goat or donkey serum in 0.2% Triton X-100 PBS (PBS-T) for 1 h at room temperature. Primary antibodies diluted in 0.2% PBS-T were applied to the sections and incubated overnight at 4 °C. After three 10-min washes with 0.2% PBS-T, the sections were incubated with secondary antibodies diluted in 0.2% PBS-T for 2 h at room temperature. After antibody staining, anterior tongue sections were mounted onto slide glass. DAPI (1:1 000; D9542, MilliporeSigma, Burlington, MA, US) was used for nuclear staining, and cover glasses were mounted using anti-fade fluorescence mounting medium (ab1041135, Abcam, Cambridge, UK). Images were acquired using either an LSM900 (Zeiss, Oberkochen, Germany) or K1-Fluo (Nanoscopesystems, Daejeon, Korea) confocal microscope Optical thickness of each focal plane was set at 1 μm. The displayed fluorescent images are z-stacks, representing the summation of signals from all optical sections.

Organoid culture

From the harvested tongues of adult mice (8–12 weeks), the lingual epithelial layers were peeled off following enzymatic digestion of the subepithelium via injection with Dispase II (10269638001, Roche, Basel, Swiss). CVP were dissected and embedded in Matrigel (356231, Corning Inc., Corning, NY, US). Ex vivo CVP tissues were cultured in standard organoid culture medium for 10 days at 37 °C and 5% CO2, before then being dissociated into single cells with 0.25% Trypsin (25200072, ThermoFisher Scientific, Waltham, MA, US) and filtered through a 70-µm strainer to remove cellular aggregates. The resulting cellular suspensions were then centrifuged at 450 × g for 5 min at 4 °C, resuspended in Matrigel, and cultured through another passage into medium. The complete culture medium contained 70% DMEM/F12 (11320033, Life Technologies, Carlsbad, CA, US), 20% R-spondin 1-conditioned medium (generated from an Rspo1-expressing cell line, SCC111, Merck, Rahway, NJ, US), and 10% Noggin-conditioned medium (generated from a Noggin-expressing cell line received as a gift from Dr. Peihua Jiang and then selected via 100 mg/mL Zeocin). The medium was also supplemented with 1% N2 (17502-048, Life Technologies), 2% B27 (17504044, Life Technologies), 1× penicillin-streptomycin (15140122, ThermoFisher Scientific), 10 μmol/L Y27632 (HY-10584, Med Chem Express, Monmouth Junction, NJ, US), 1 mmol/L N-acetylcysteine (A9165, MilliporeSigma), and 50 ng/mL FGF-b (AF-100-18B, ThermoFisher Scientific). The medium was replaced every 3–4 days depending on the density of the organoid culture over the ensuing 10 days. In Fig. 4, 2 μmol/L imatinib mesylate (GC11759, GLPbio, Montclair, CA, US) was administered for the +Ima condition, and R-spondin 1-conditioned medium was omitted for the −R condition. The +Ima-R condition included both modifications on day 5.

Organoid immunostaining

Whole-mount staining was performed on taste bud organoids. Taste bud organoids treated under each condition were harvested with cold PBS. They were fixed in 4% PFA in PBS for 15 min and then permeabilized in 0.2% PBS-T for 15 min. The organoids were then incubated with primary antibody diluted in PBS-T overnight at 4 °C. After three 10-min washes with PBS-T, the organoids were incubated with secondary antibody diluted in PBS-T for 1 h at room temperature. After DAPI staining, the organoids were mounted with coverslips using Vectashield (Vector Laboratories, Newark, CA, USA). At least 90 organoids were examined per group in triplicate experimental sets.

Antibodies

The primary antibodies used for tissue and organoid immunostaining included the following: rabbit anti-Krt14 (1:500, ab181595, Abcam), rat anti-Krt8 (1:1 000, Troma-1, DSHB, Iowa City, IA, US), rabbit anti-Krt8 (1:1 000, ab53280, Abcam), guinea pig anti-Krt13 (1:500, BP5076, OriGene, Rockville, MD, US), rabbit anti-NTPdase2 (1:1 000, mN2-36L I6, CHUQ, Québec, Canada), guinea pig anti-Trpm5 (1:500, generated in our previous paper54), goat anti-Car4 (1:500, AF2414, R&D Systems, Minneapolis, MN, US), goat anti-c-Kit (1:400, AF1356, R&D Systems), rabbit anti-GNAT3 (1:1 000, sc-395, Santa Cruz, Dallas, TX, US), goat anti-GNAT3 (1:1 000, OAEB00418, Aviva system, San Diego, CA, US), and rabbit anti-PLCβ2 (1:500, generated in this study), rabbit anti-PGP9.5 (1:400, GTX109637, Gene Tex, Irvine, CA, US), and goat anti-AIF-1/Iba1 (1:400, NB100-1028, Novus biological, Centennial, CO, US). The final polyclonal antibody against mouse PLCβ2 was generated in a rabbit immunized with synthetic peptides (1145-1158: EPLVSKADTQESRL).

The secondary antibodies used included the following: donkey anti-rabbit Alexa488 (1:1 000, A32790, Invitrogen, Waltham, MA, USA), donkey anti-rabbit Alexa555 (1:1 000, A31572, Invitrogen), donkey anti-goat Alexa647 (1:1 000, A32849, Invitrogen), donkey anti-goat Alexa555 (1:1 000, A32816, Invitrogen), donkey anti-guinea pig Alexa488 (1:1 000, 706-545-148, Jackson ImmunoResearch, West Grove, PA, USA), goat anti-rat Alexa647 (1:1 000, ab150159, Invitrogen), goat anti-guinea pig Alexa488 (1:1 000, A11073, Invitrogen), goat anti-rabbit Alexa488 (1:1 000, A32731, Invitrogen), and goat anti-rabbit Alexa555 (1:1 000, A32732, Invitrogen).

Quantitative measurements from confocal images

To assess the morphometric features of taste buds, including size, width, and height, we carefully delineated the virtual borders of each taste bud using the anti-Krt8 immunofluorescence signal, which is a reliable taste bud marker. Taste bud borders were outlined manually on high-resolution confocal images using the Zeiss LSM Image Browser and K1-Viewer to ensure accurate representations. For cell type identification, only taste cells that exhibited immunoreactive signals and contained a DAPI-stained nucleus were included in the count to avoid false positives from overlapping or non-cellular signals. Immunoreactive taste cells were identified using specific markers corresponding to the different cell types. The number of taste cells per taste bud was determined by counting all DAPI-stained nuclei within the outlined border. The percentage of immunoreactive taste cells for each marker was calculated by dividing the number of immunoreactive cells by the total number of taste cells within the taste bud. Statistical analyses were applied to compare these morphometric parameters between experimental groups.

To quantify taste bud organoid images, we calculated PAI and PEI as follows. For PAI, the area of each organoid and the area occupied by the immunoreactive signals were measured using Image J. PAI was then calculated for each organoid by dividing the immunoreactive area by the total organoid area. Since PAI is derived from individual organoids, the average and SEM values are indicated in the graph. For clear comparison across experimental conditions, the average PAI values of each condition were divided by the average PAI value of FRN control condition to generate normalize PAI. PEI was determined as the proportion of organoids containing at least one cells expressing immunosignals. Normalized PEIs are generated by dividing the PEI of each condition by the PEI of the FRN control condition.

Statistics

All graphical data are expressed as means ± standard error of the mean (SEM). GraphPad Prism 9 (GraphPad Software, San Diego, CA, USA) was used to analyze the data. Normality and homoscedasticity were tested using the Kolmogorov–Smirnov and Levene tests, respectively. Comparisons between two groups were assessed using unpaired two-tailed t-tests or Mann–Whitney U tests. Comparisons among four or more groups were performed using ANOVA with post-hoc Bonferroni corrections. For these cases, adjusted P-values are reported. Chi-square tests were used for comparing proportional parameters.