ISOLATED LARYNGEAL Pou2f3-EXPRESSING CELL AND METHOD FOR MONITORING COUGH RESPONSE IN MOUSE

Abstract

To provide an isolated cell capable of regulating cough.SOLUTION: The problem is solved by an isolated laryngeal PCC expressing Calhm3 protein.SELECTED DRAWING: None

Description

[Technical Field]

[0001] Disclosed herein are isolated laryngeal Pou2f3-expressing cells and methods for monitoring cough responses in mice. [Background technology]

[0002] Coughing and swallowing are responses to chemicals in the throat that protect the airway from aspiration and inhalation of noxious substances. Chronic cough is a condition in which the cough reflex is hypersensitive, but due to a lack of understanding of the mechanisms of coughing, it remains prevalent worldwide (Non-Patent Documents 1 and 2). [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] Chung, KF et al. Cough hypersensitivity and chronic cough. Nat Rev Dis Primers 8, 45 (2022) [Non-patent document 2] Chung, KF, McGarvey, L. & Mazzone, SB Chronic cough as a neuropathic disorder. Lancet Respir Med 1, 414-422 (2013). Summary of the Invention [Problem to be solved by the invention]

[0004] An objective of the present invention is to provide isolated cells that control coughing, and a method for monitoring the coughing response in mice, which has previously been difficult to monitor. [Means for solving the problem]

[0005] The present invention includes the following embodiments. Section 1. Isolated laryngeal Pou2f3-expressing cells expressing Calhm3 protein. Section 2. Item 1. A laryngeal Pou2f3-expressing cell according to Item 1, which further expresses Calhm1 protein. Section 3. A laryngeal Pou2f3-expressing cell according to item 2, which does not express Dclk1 protein. Section 4. 1. A method for monitoring a cough response in a mouse, comprising: administering a test substance into the subglottic cavity; Observe respiratory and glottic opening movements and a monitoring method comprising: Section 5. It also includes observing digastric muscle dynamics. The administration of the test substance does not change the dynamics of the digastric muscle. Item 4. The monitoring method described in item 4. Section 6. Item 5. The monitoring method according to Item 4, wherein the test substance is administered by aerosolizing the test substance. [Effects of the Invention]

[0006] Isolated cells that control coughing can be provided, and methods for monitoring the cough response in mice can be provided. [Brief explanation of the drawings]

[0007] [Figure 1]This figure shows that hypopharyngeal and laryngeal PCCs form channel synapses with vagal afferents. (a, e, h, i) Representative images of immunostaining for GCaMP3 and tdTomato, PLCB2 (e), P2X2 (h), or P2X3 (i) in the indicated tissues of Calhm1 / 3 dual reporter mice. All images were counterstained with DAPI. (a, b) Images are epifluorescence images, and the boxed regions in each image were reacquired by confocal imaging (a1-d1). (f) The percentage of CALHM1 / 3+ cells among PLCB2+ cells in the indicated tissues. (g) The percentage of PLCB2+ cells among CALHM1 / 3+ cells in the indicated tissues. jl, n, and o show representative images of immunostaining for P2X2, CALHM1, cytochrome c, and tdTomato in the hypopharyngeal taste buds (jl) and laryngeal epithelium (n, o) of Calhm3-tdTomato mice. j is an image acquired by confocal imaging. k is an image of the boxed area in j reacquired by super-resolution imaging. n shows an image acquired by super-resolution imaging. l and o are enlarged views of the boxed areas in k and n, respectively. m and p show the fluorescence intensity along the dashed lines in l and o, respectively. Scale bars indicate (ad): 1 mm, (a1-d1, e, hj): 10 μm, (k, n): 1 μm, (l, o): 200 nm. The numbers of animals used in the experiments were 4, 5, 5, 3 (ad); 2, 3, 5, 5, 4, 3, 2 (e.g., from left to right); and 6, 3 (h); 6, 3 (i); 8 (jm) and 3 (np), respectively. [Figure 2] The expression patterns of the reporter in 44 tissues are shown. [Figure 3] The results of RT-PCR evaluation of Calhm1, Calhm3, Pou2f3, and Actb (β-actin) mRNA expression in various tissues collected from wild-type mice are shown. [Figure 4]Confocal imaging of GCaMP3 (green), tdTomato (red), and TRPM5 (blue) in the tissues indicated in the figure from Calhm1 / 3 dual reporter mice. Nuclear staining is with DAPI (gray). [Figure 5] Confocal fluorescence imaging of EYFP (green) and PLCB2 (blue) in the indicated tissues of Vglut2-ChR2-EYFP mice. Scale bars are 20 μm. The number of animals used in the experiment was 2 (a); 2, 3, 5, 4, 4, 3, 2 (b, left to right); 2, 2, 3, 5, 5, 4, 4 (c, left to right); 2 (d); 2, 2, 3, 3, 2, 3, 2 (e, left to right); and 3 (f). [Figure 6] The probability of CALHM1 / 3 co-expression along the airways (among cells expressing at least one subunit) is shown. [Figure 7] Figure 1 shows the involvement of hypopharyngeal purinergic PCCs in the swallowing reflex. (a) Uniform Manifold Approximation and Projection (UMAP) plot showing 12 color-coded cell clusters identified by scRNA-seq of 432 cells isolated from the hypopharyngeal taste buds of 12 Pkd2l1-GCaMP3 mice. The plot also shows the approximate location of expression of standard taste cell markers. (b) Heatmap of log-normalized expression of the cell class-defining genes and standard taste bud-associated genes identified in (a). (c) Plot showing log-normalized expression of PCC-associated genes. [Figure 8]Figure 1 shows the involvement of hypopharyngeal purinergic PCCs in the swallowing reflex. a) Experimental setup for monitoring swallowing induced by chemical stimulation of the hypopharynx and larynx. b) and c) show representative recordings during stimulation with 20 mM denatonium benzoate (Den, between dashed lines) in wild-type (b) and Calhm3 KO mice (c). The inverted triangle indicates the timing of hyoid elevation. d) Swallowing frequency following various stimuli (saline, water, 750 mM NaCl (high salt), 25 mM citric acid (CA), 10 mM cycloheximide (CHX), Den, 30 mM acesulfame K (Ace K), 2 mM SC-45647 (SC), and 30 mM monosodium glutamate + 1 mM inosine monophosphate (MSG / IMP)). n = 6 (WT) and 9 (KO). Graphs show mean ± SEM. *** indicates P<0.001 (unpaired t-test). e shows the illumination scheme for optogenetic stimulation of the hypopharynx, subglottis, trachea, or superior laryngeal nerve (SLN, not shown) in Calhm1-ChR2 mice. Note that the hypopharynx and subglottis illumination overlap slightly. f shows the pharmacological sensitivity of swallowing responses induced by Den, water, and high salt to AF-353 (125 μM). n=6. Graphs show mean ± SEM. * indicates P<0.05, and *** indicates P<0.001 (post-hoc Dunnett's test). g shows the number of swallows during optogenetic stimulation of the indicated regions. n=9, 7, 9, 7, 8, 5, 6, 5 (left to right). Graphs show mean ± SEM. *** indicates P<0.001 (post-hoc Tukey-Kramer test). [Figure 9]Purinergic PCCs in the larynx and their characteristics. (a) Results of transcriptome analysis of isolated laryngeal epithelium (11 mice). UMAP plots of 56,013 cells with color-coded cell IDs are shown. Twelve transcriptome cell classes, including circled PCCs, are color-coded. (b) Results of transcriptome analysis of laryngeal PCCs. UMAP plots of 576 cells with color-coded cell IDs are shown for six transcriptionally distinct subclasses, PCC1–6. Classification by combined expression of Plcb2 and Dclk1 is also shown. (c) Violin plots showing the log-normalized expression of the subclass-defining genes classified in (b). (d) Violin plots showing the results of custom target enrichment using linear amplification and PCR. Target enrichment included enrichment of Calhm1, Calhm3, T2r105, and Scn2a transcripts in PCC5. Violin plots show the log-normalized expression of genes in PCC subclasses before (original) and after target enrichment. (e) Whole-mount staining of PLCB2, DCLK1, and tdTomato in the subglottic epithelium of Calhm3-tdTomato mice. Nuclear staining was performed with DAPI. Arrows indicate PLCB2+ / DCLK1- PCCs. Scale bars indicate 20 μm. Images are representative of two mice. (fi) Results of ex vivo single-cell intracellular calcium ion imaging (f) and extracellular ATP imaging (gi) in the subglottic epithelium of Calhm1-GCaMP3 (3 mice, 6 experiments) and Calhm1-GRABATP1.0 (2 mice, 4 experiments). In (f), each row shows the change in GCaMP3 fluorescence intensity (color scale) and time course (x-axis) for each cell in response to stimulation. The stimuli were 5 mM quinine HCl (QHCl), 10 mM denatonium benzoate (Den), and 0.3 mM 3-oxo-C12-homoserine lactone (HSL). The dashed line indicates the duration of the stimulation. (g) Spatially separated GRABATP1.0 responses to 10 mM Den and 10 μM ATP administered luminally.The top row shows representative images of GRABATP1.0 fluorescence (ΔF), and the bottom row shows traces (Z-scores). Images show peak responses. Scale bars indicate 10 μm. Traces represent the Den-responsive (blue) and ATP-responsive (pink) regions, as circled in the images. Black bars indicate the duration of stimulation (20 s). h and i show an overview of the experiment shown in g, showing peak responses in the Den-responsive (h) and ATP-responsive (i) regions. n = 5 cells. Graphs show mean ± SEM. * indicates P < 0.05, *** indicates P < 0.001 (post-hoc Dunnett's test). [Figure 10]This figure shows an overview of single-cell gene expression analysis of laryngeal epithelial cells using droplet-based 3' scRNA-seq, custom linear amplification, and PCR-based target enrichment techniques. (a) shows the experimental scheme for droplet-based 3' scRNA-seq of acutely isolated laryngeal epithelial cells from 11 WT mice (6 females, 5 males), combined with linear amplification and PCR-based target enrichment of four genes. (b) shows the scheme for linear amplification and PCR-based target enrichment. To improve specificity while maintaining the efficiency of conventional methods for detecting low-abundance genes, we developed a novel targeting method using three rounds of linear amplification and PCR in a nested configuration. Each cDNA amplicon consists of a 16-nucleotide (nt) 10x barcode from Illumina TruSeq Read 1, a 12-nt unique molecular identifier (UMI), a 30-nt poly(dT), a cDNA insert, and a 30-nt template switching oligo (TSO). Target enrichment involves 20 cycles of linear amplification using a single primer (gene-specific primer 1) that binds to a specific target sequence in the cDNA library, allowing for selective enrichment of the desired target. Double-stranded DNA is then amplified by three cycles of PCR using the universal Read 1 primer and gene-specific primer 1 (Round 1). For more specific and efficient enrichment, two additional rounds (Round 2 and Round 3) of linear amplification and PCR were performed using nested gene-specific primers (gene-specific primers 2 and 3). [Figure 11]Characteristics of PCC5 cells are shown. (a) Heatmap of log-normalized expression of PCC-associated receptor genes in cells classified into cell classes identified by 3' scRNA-seq (left) and log-normalized expression of PCC-associated receptor genes in cells classified into cell classes identified by full-length scRNA-seq (right). (b) Heatmap of log-normalized expression of voltage-gated Na+ channel genes in cells classified into cell classes identified by 3' scRNA-seq (left) and log-normalized expression of PCC-associated receptor genes in cells classified into cell classes identified by 3' scRNA-seq (right). Data are omitted in the left panel for the following genes, as no expression was detected: in a, T2r103, T2r114, T2r115, T2r119, T2r120, T2r121, T2r122, T2r123, T2r124, T2r125, T2r131, T2r134, T2r136, T2r139, T2r140, and T2r144; in b, Scn10a and Scn11a. [Figure 12] Design of the Cre-dependent TIGRE2.0-GRABATP1.0 reporter allele for generation of Calhm1-GRABATP1.0 mice. The conditional allele was inserted into the Igs7 locus using CRISPR / Cas9 genome editing technology. [Figure 13]Figure 1 shows the involvement of laryngeal purinergic PCCs in cough and cough hypersensitivity. (a) Schematic diagram of the experimental setup for simultaneous subglottic chemical stimulation and cough monitoring. A stimulating solution was sprayed into the subglottic cavity, and glottic opening area, respiratory dynamics, and digastric muscle activity were measured using an endoscope, a thermistor, and EMG, respectively. (b) Representative examples of responses during mechanical stimulation (left) and 25 mM citric acid (CA; right) in wild-type (WT) mice. The dashed line indicates the level of complete glottic closure. (c) An enlarged view of the box in (b). The inspiratory and expiratory phases are highlighted in blue and pink, respectively. (d) Physiological events related to breathing. Three indices, Index GA, Index Resp, and Index EMG, were calculated from the glottic opening area, respiratory dynamics, and EMG measurements, respectively. These indices allowed for clear classification of cough (blue circle), swallowing (red circle), and normal breathing (black circle) events. Figures e and f show representative recordings of 20 mM denatonium benzoate nebulization (Den; dashed line indicates the onset of stimulation) in WT mice (e) and Calhm3 KO mice (f). The location of coughing is indicated by an asterisk. Respiratory scalogram and phase plot analysis revealed Den-induced changes in WT mice. Figures g and h show graphs of cough counts (g) and delta respiratory power (h) from the experiments shown in Figures e and f. There were eight WT and eight KO mice, respectively; graphs show mean ± SEM; * indicates P < 0.05, *** indicates P < 0.001 (post-hoc Tukey-Kramer test). Figure i shows the number of coughs during optogenetic stimulation in the designated area (dashed line) in Figure 8e. In the graph, n = 9, 7, 9, 7, 8, 5, 6, 5 (from left to right); graphs show mean ± SEM; ** indicates P < 0.01 (post-hoc Tukey-Kramer test). j shows the pharmacological sensitivity of the cough response induced by optogenetic activation of subglottic PCCs to AF-353 (125 μM). n = 5; graphs show mean ± SEM; *** indicates P < 0.001 (post-hoc Dunnett test).(k) shows the number of coughs induced by various stimuli in WT and Calhm3 KO mice after subglottic administration of mock (-) or Alternaria extract (+). Stimuli were Dental acid and citric acid (CA). Each group (mock-WT, Alternaria-WT, mock-KO, Alternaria-KO) consisted of eight individuals; the graph shows the mean ± SEM; ** indicates P<0.01 (post-hoc Tukey-Kramer test). (l) shows a schematic diagram of epithelial chemosensation by laryngeal purinergic PCCs as the initiation site of coughing and swallowing via transmission to the vagus nerve, and oral purinergic PCCs as taste sensors. DETAILED DESCRIPTION OF THE INVENTION

[0008] 1. Isolating Laryngeal Pou2f3-expressing Cells Pou2f3-expressing cells are sometimes referred to herein as Pou2f3-expressing chemosensory cells (hereinafter "PCCs"). Laryngeal PCCs are purinergic PCCs derived from the larynx that express Calhm3 protein. Laryngeal PCCs preferably express Calhm1 protein. Laryngeal PCCs more preferably do not express Dclk1 protein. Laryngeal PCCs even more preferably express Plcb2 protein. Laryngeal PCCs even more preferably express T2r proteins such as T2r104 protein and T2r105 protein. Laryngeal PCCs are Pou2f3-positive chemosensory cells. Hereinafter, these proteins will be referred to as "marker proteins."

[0009] The human Calhm1 protein is registered with the National Center for Biotechnology Information (NCBI) under Accession No. NP_001001412.3. The mouse Calhm1 protein is registered with NCBI under Accession No. NP_001074740.1.

[0010] The human Calhm3 protein is registered under NCBI Accession No. NP_001123214.1, and the mouse Calhm3 protein is registered under NCBI Accession No. J3QMI4.1.

[0011] Human Dclk1 protein has been registered under NCBI Accession Nos. NP_001182344.1, NP_001182345.1, NP_001182359.1, NP_001317000.1, NP_001317001.1, and NP_004725.1. Mouse Dclk1 proteins have been registered under NCBI Accession Nos. NP_001104521.1, NP_001104522.1, NP_001104523.1, NP_001182467.1, NP_001182468.1, NP_001182469.1, NP_001344395.1, NP_001344397.1, NP_001344398.1, NP_001344404.1, NP_001344405.1, and NP_064362.1.

[0012] Human Plcb2 proteins are registered under NCBI Accession Nos. NP_001271226.1, NP_001271227.1, NP_001271228.1, and NP_004564.2. Mouse Plcb2 proteins are registered under NCBI Accession Nos. NP_001277719.1 and NP_808236.2.

[0013] The human T2r10 protein is registered under NCBI Accession No. NP_076410.1, and the mouse T2r104 protein is registered under NCBI Accession No. NP_996894.1.

[0014] The human T2r10 protein is registered under NCBI Accession No. NP_076410.1. The mouse T2r105 protein is registered under NCBI Accession No. NP_065247.1.

[0015] The human Pou2f3 protein is identified under NCBI Accession No. Registered under the following names: NP_055167.2, NP_001231611.1, XP_011541041.1, XP_011541042.1, XP_011541043.1, XP_011541044.1, XP_011541045.1, XP_016872976.1, XP_024304192.1, XP_054224314.1, XP_054224315.1, XP_054224316.1, XP_054224317.1, XP_054224318.1, XP_054224319.1, XP_054224320.1. The mouse Pou2f3 protein is registered under NCBI Accession Nos. NP_035269.2 and XP_011240722.1.

[0016] Here, whether a marker protein is "expressed" or "not expressed" can be determined by immunohistochemical staining, single-cell RNA sequencing (scRNA-seq, preferably droplet-based 3' scRNA-seq), RT-PCR, etc. When determining whether a marker protein is "expressed" or "not expressed" by immunohistochemical staining, a cell can be determined to be "expressed" if the fluorescent signal of the marker protein in the cell is higher than the detection limit of the measurement system. Furthermore, a cell can be determined to be "not expressed" if the fluorescent signal of the marker protein in the cell is lower than the detection limit of the measurement system. In the case of immunohistochemical staining, the fluorescent signal is confirmed visually, for example, using a fluorescent microscope. Cells with a stronger fluorescent signal than other cells can be determined to be "expressing" the marker protein. When determining whether a marker protein is "expressed" or "not expressed" by scRNA-seq, a cell can be determined to be "expressed" if the mRNA expression level of the marker protein is higher than the detection limit of the measurement system and is higher than "0." Furthermore, a cell can be determined to be "not expressed" if the mRNA expression level of the marker protein is lower than the detection limit of the measurement system and is lower than "0." More specifically, expression levels can be assessed by mapping scRNA-seq data (read sequences) obtained using a next-generation sequencer to a reference sequence and counting the number of reads. A read count of "0" indicates "not expressed," whereas a read count greater than "0" indicates "expressed." Regarding the determination of "expressed" or "not expressed" by RT-PCR, if an amplification product derived from the mRNA of each marker protein is detected by PCR, it can be determined to be "expressed." Furthermore, if no amplification product derived from the mRNA of each marker protein is detected by PCR, it can be determined to be "not expressed." The amplification products can be determined by gel electrophoresis or capillary electrophoresis.

[0017] By "isolated" it is intended that the laryngeal PCC is separated from living or fixed tissue (including dead cells).

[0018] Laryngeal PCCs can be isolated from tissue, for example, by the following method.

[0019] Mice are deeply anesthetized and transcardially perfused with phosphate-buffered saline. The larynx is incised to expose the inner surface of the larynx. Approximately 100 μL of the first enzyme mixture is injected under the laryngeal epithelium and left at room temperature (approximately 20-28°C) for 10 to 25 minutes. This reaction allows the epithelium to peel off from the dermis. For example, Tyrode's solution contains 140 mM sodium chloride, 5 mM potassium chloride, 2 mM calcium chloride, 1 mM magnesium chloride, 10 mM glucose, 5 mM sodium pyruvate, and 10 mM [4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid] [HEPES], adjusted to pH 7.4 with sodium hydroxide. The enzymes added to Tyrode's solution to prepare the first enzyme mixture can include, for example, approximately 0.5-3 mg / ml of DNase I and approximately 1-5 mg / ml of Dispase II (8 U / ml-40 U / ml). The detached epithelium is incubated in approximately 100 μl of a second enzyme mixture at 37°C for approximately 15 to 30 minutes. The second enzyme mixture can be prepared, for example, by dissolving approximately 15 U / ml to 60 U / ml of papain, approximately 0.5 mg / ml to 3 mg / ml of DNase I, approximately 1 mM to 4 mM of DL-dithiothreitol, and 5 μg / ml to 15 μg / ml of L-cysteine ​​in calcium ion-free Tyrode's solution. After digestion of the epithelium in the second enzyme mixture, approximately 50 μl of a stop solution is added. Examples of the stop solution include Tyrode's solution containing 1 to 4 μg / ml of leupeptin. A cell suspension can be obtained by pipetting the tissue containing the stop solution using a glass capillary pipette.

[0020] Laryngeal PCCs in a cell suspension can be isolated using a cell sorter (preferably a FACSseq™ cell sorter), the AS ONE Cell Picking System, or the like.

[0021] 2. Method for monitoring cough response in mice This embodiment involves administering a test substance to the subglottic space and observing respiratory and glottic opening dynamics.

[0022] Specifically, mice were anesthetized, for example, by intraperitoneal injection of urethane (1.3 g / kg body weight, Sigma-Aldrich, Catalog No. U2500). Respiratory secretions were suppressed by intraperitoneal administration of, for example, 0.1 mg / kg atropine. The trachea was then surgically exposed, and a small window (hereinafter referred to as the "tracheal window") was opened in the anterior wall below the cricoid cartilage to expose the subglottic space.

[0023] Respiratory dynamics are measured by placing a thermistor (e.g., Warner Instruments, Hamden, CT, USA, Catalog No. TC-324C) in the lower trachea. The thermistor signal is connected to a temperature monitor (Warner Instruments, Hamden, CT, USA, Catalog No. AN6N4-GC11KA143L / 37C) and recorded using a data acquisition system (e.g., PowerLab8 / 35, AD Instruments, Colorado Springs, CO, USA, Catalog PL3508). The collected data is quantitatively analyzed using, for example, MATLAB (registered trademark) (MathWorks, Neatec, MA, USA). For example, a scalogram (time-resolved power across frequencies) of respiratory dynamics is created using wavelet transform. Wavelet transform can be performed, for example, by convolving a band-pass filtered (0.5-30 Hz) signal with a Complex Morlet wavelet as the mother wavelet. The effect of stimulation was evaluated by substituting the scalogram values ​​into the following equation:

[0024]

number

[0025] where Pf pre and P f Post indicates the mean power at each frequency (2–10 Hz) before and after stimulation, respectively. Negative and positive values ​​of this index (ranging from -1 to 1) indicate decreases and increases in respiratory dynamics, reflecting respiratory rate and depth, respectively.

[0026] Digastric muscle contraction can be monitored by electromyography (EMG). A 0.30 mm diameter silver wire (e.g., Nilaco Corporation, Tokyo, Japan, catalog AG-401325) is inserted into the muscle. The signal from the muscle is bandpass filtered to obtain a signal in the 0.3–10 kHz range, amplified 1,000 times, and then recorded using a differential amplifier (e.g., World Precision Instruments, Sarasota, FL, USA, catalog number DAM-80) and a PowerLab 8 / 35.

[0027] The dynamics of the glottic opening (i.e., the glottis) were observed from below using a 1.7 mm diameter endoscope (Vison Well Corporation, Tokyo, Japan, catalog number 1.7-010-030-065) inserted into the trachea through the tracheal window. The endoscope can be connected to a high-speed monochrome camera (120 fps, Argo Corporation, Osaka, Japan, catalog number DMK33UP1300).

[0028] The glottic opening can be assessed using video captured at 120 fps using an endoscope. The region of interest (ROI) is determined by calculating the pixel-by-pixel absolute difference in the glottic opening between -10 seconds before the stimulus and 20 seconds after the stimulus, with the stimulus onset being set as 0. The ROI is set to include the glottic opening at maximum opening. The brightness intensity of each pixel is measured, and the ROI is binarized using thresholding. The glottic opening within the ROI is then defined as the zero-intensity region.

[0029] The cough response can be identified by an initial deep inspiration with glottal expansion (inspiratory effort) followed by a forceful, long expiration against a constricted glottis (expiratory effort) without digastric muscle activity.

[0030] Normal respiratory and glottic dynamics can be distinguished from those during cough response using cluster analysis, for example, a Gaussian mixture model (GMM). For each dynamic, three indices (IndexResp, IndexGA, and IndexEMG) are calculated and then classified into 1 to 10 clusters using a 3D Gaussian mixture model. Next, the Bayesian information criterion (BIC) for each clustering is calculated to determine the "true" number of clusters. The BIC is calculated according to the following formula, and the number of clusters with the smallest average BIC is determined.

[0031] BIC=-2ln(L)+kln(n) where L, k, and n denote the likelihood of each physiological event, the number of parameters estimated by the model, and the number of physiological events, respectively.

[0032] Next, GMM fitting using the expectation-maximization algorithm is performed for 1 to 10 clusters, for example, 1,000 times, and the BIC is calculated for each iteration. For further validation, an x-means clustering algorithm is used. These methods can be performed using custom scripts, for example, with Python ver. 3.9 (Python Software Foundation, Wilmington, Delaware, USA) and additional modules such as scikit-learn (GMM fitting and BIC calculation) and PyClustering (x-means clustering).

[0033] Cough-inducing chemicals, chemicals to be tested for cough induction, and negative control chemicals can be administered to mice in aerosol form. Chemicals can be sprayed as aerosols into the subglottic cavity through the tracheal window using a mist sprayer (Qomolangma Brave Knight, Seattle, WA, USA, catalog number B08MZ9JX3N) placed 15 cm away from the mouse. For example, spraying each chemical for 3 seconds allows the mouse to count a cough response within 10 seconds. Mechanical stimulation can also be achieved by delivering airflow (50 kPa) through a 0.8 mm diameter tube into the throat from a distance of 1–2 cm for 10 seconds.

[0034] Chemicals that induce a cough response include, for example, denatonium benzoate and citric acid. [Example]

[0035] The present invention will be described in more detail below using examples, but the present invention should not be construed as being limited to these examples.

[0036] I. Materials and Methods 1. Animals and Genetically Modified Animals (1)Animals All animal experiments were conducted in accordance with protocols approved by the Institutional Animal Care and Use Committees of Kyoto Prefectural University of Medicine and the RIKEN Center for Integrated Medical Sciences. All procedures related to animal care and treatment were conducted in accordance with the Japanese Act on Welfare and Management of Animals and the Guidelines for Proper Conduct of Animal Experiments issued by the Science Council of Japan. Every effort was made to minimize animal suffering during experiments and to reduce the number of animals used in this study. Animals were housed in a temperature- and humidity-controlled animal room under a 12:12-h light / dark cycle and provided with water and standard rodent chow ad libitum. Experiments were conducted using adult animals of both sexes (≥70 days old for mice) unless otherwise specified. C57BL / 6 mice were obtained from Shimizu Experimental Materials Co., Ltd. (Kyoto, Japan). Mouse genotypes were determined by PCR.

[0037] (2) Calhm1 / 3 dual reporter mice GCaMP3 and tdTomato were expressed under the control of the Calhm1 and Calhm3 promoters, respectively. Calhm1-Cre mice were transfected with the Cre-dependent GCaMP3 reporter Rosa26 LSL-GCaMP3 Mice (Jackson Laboratory, Ai38, strain number 014538, RRID:IMSR_JAX:014538) and Calhm3 tdTomato The mice were crossed with mice carrying the Calhm1 allele. Cre / + ::Rosa 26LSL-GCaMP3 / + ::Calhm3 tdTomato / + Mice were transfected with CALHM1 / 3 + Calhm1 / 3 dual reporter mice were used to identify the cells.

[0038] (3) Calhm3 KO mice and Calhm3-tdTomato mice Calhm3 in which the first 117 nucleotides starting from the initiation codon of exon 1 of Calhm3 were replaced with the tdTomato sequence containing a stop codon at the 3' end tdTomato Allele (Calhm3 tdTomato / tdTomato) homozygous mice (Calhm3 tdTomato / tdTomato ) were used as Calhm3 KO mice. Mice heterozygous for the allele (Calhm3 tdTomato / + ) are hereafter referred to as Calhm3-tdTomato mice. + was used to label the cells.

[0039] (4) Vglut2-ChR2-EYFP mice Slc17a6-Cre mice (Jackson Laboratory, Vglut2-ires-cre, line number 016963, RRID:IMSR_JAX:016963) were transfected with the Cre-dependent ChR2-EYFP reporter Rosa26 LSL-ChR2-EYFP The mice were crossed with mice (Jackson Laboratory, Ai32, line number 012569, RRID:IMSR_JAX:012569) and transfected with Slc17a6 Cre / + ::Rosa26 LSL-ChR2-EYFP / + These mice are hereinafter referred to as Vglut2-ChR2-EYFP mice. + It was used to fluorescently label afferent nerve fibers.

[0040] (5) Pkd2l1-GCaMP3 mice Pkd2l1-Cre mice were kindly provided by Dr. C.S. Zuker (Columbia University, New York, USA) and were used as Rosa26 mice. LSL-GCaMP3 The mice were crossed with Pkd2l1-Cre::Rosa26 LSL-GCaMP3 / + We generated mice, hereafter referred to as Pkd2l1-GCaMP3 mice, and used them to visualize taste bud cells in the hypopharynx.

[0041] (6) Calhm1-ChR2 mice Calhm1-Cre mice were cloned as Rosa26 LSL-ChR2-EYFP and crossed with mice to express Calhm1 Cre / + ::Rosa26 LSL-ChR2-EYFP / +We generated mice, hereafter referred to as Calhm1-ChR2 mice, and used them in PCC optogenetic experiments.

[0042] (7) Calhm1-GCaMP3 mice Calhm1-Cre mice were cloned as Rosa26 LSL-GCaMP3 and crossed with mice, and Calhm1 Cre / + ::Rosa26 LSL-GCaMP3 / + Hereafter, these mice are called Calhm1-GCaMP3 mice. 2+ was used for imaging.

[0043] (8) Calhm1-GRAB ATP1.0 mice A Cre-dependent fluorescent extracellular ATP indicator tool line was generated using a C57BL / 6 strain with a high-level expression reporter system called TIGRE2.0, which contains two insulator sequences (Ins), a TRE2 promoter, a loxP-flanked STOP cassette (LSL), and a GPCR activation-based ATP sensor, GRAB. ATP1.0 The Ins-TRE2-LSL-GRABATP1.0-Ins-CAG-LSL-tTA2 allele (TIGRE2.0-GRABATP1.0) was designed using the tTA2 sequence, the woodchuck hepatitis virus posttranscriptional regulatory element (WPRE), and the bovine growth hormone polyadenylation site (bGHpA). ATP1.0 The Igs7 gene was targeted and inserted into the Igs7 locus using CRISPR / Cas9 genome editing technology. The CRISPR / Cas9 genome editing was performed at the University of Tsukuba Laboratory Animal Resource Center. The resulting mice were viable and fertile, and were characterized using the TIGRE2.0-GRAB gene. ATP1.0 When mice carrying the allele were bred, the expected genotype was obtained at the expected frequency.

[0044] For single PCC ATP imaging, Calhm1-Cre mice were cultured in TIGRE2.0-GRAB ATP1.0 and crossed with mice, and Calhm1 Cre / + ::Igs7 TIGRE2.0-GRABATP1.0 / + Hereafter, this mouse will be referred to as Calhm1-GRAB ATP1.0 Called Mouse. Calhm1-GRAB ATP1.0 In mouse subglottic epithelium, PLCB2 + GRAB specifically entopic to the cell membrane of PCCs ATP1.0 Expression was confirmed.

[0045] 2. Single-cell RNA sequencing (scRNA-seq) Hypopharyngeal taste buds (defined below) were acutely isolated from 12 Pkd2l1-GCaMP3 mice (6 females and 6 males, 3 technical replicates). Mice were deeply anesthetized by intraperitoneal (ip) injection of sodium pentobarbital (>100 mg / kg body weight, Somnopentyl, Kyoritsu Pharmaceutical, Tokyo, Japan) and transcardially perfused with phosphate-buffered saline (PBS: 137 mM NaCl, 2.7 mM KCl, 10 mM NaHPO, 1.8 mM KHPO, pH 7.4). The hypopharynx was excised and its inner surface was exposed through a dorsal incision. A mixture of enzymes (DNase I, 1 mg / ml, Sigma-Aldrich, St. Louis, MO, USA, Cat. No. DN25; Dispase II, 2 mg / ml, Roche, Basel, Switzerland, Cat. No. 4942078001; Collagenase A, 1 mg / ml, Roche, Cat. No. 10103578001; Elastase, 0.2 mg / ml, Wako, Osaka, Japan, Cat. No. 058-05361) diluted with freshly prepared Tyrode's solution (140 mM NaCl, 5 mM KCl, 2 mM CaCl, 1 mM MgCl, 10 mM glucose, 5 mM Na-pyruvate, and 10 mM [4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid] [HEPES], pH adjusted to 7.4 with NaOH) was injected under the hypopharyngeal epithelium and incubated for 25 min at room temperature to remove the epithelium. The detached epithelium was incubated in the same enzyme mixture for 15 min, then Ca2+ The samples were then incubated for an additional 15 minutes in Tyrode's solution (140 mM NaCl, 5 mM KCl, 5 mM ethylene glycol-bis(2-aminoethylether)-N,N,N',N'-tetraacetic acid [EDTA], 10 mM glucose, 5 mM sodium pyruvate, and 10 mM HEPES, pH adjusted to 7.4 with NaOH). Hypopharyngeal taste buds identified under a fluorescent microscope using GCaMP3 were transferred with a glass capillary pipette to a 0.25% trypsin solution containing 1 mg / ml DNase I (Thermo Fisher Scientific, Waltham, MA, USA, catalog 15090046). After incubating the taste buds at 37°C for 20 minutes, the tissue was triturated with a glass capillary pipette to prepare a single-cell suspension. This cell suspension was used for scRNA-seq library preparation, as described below.

[0046] Laryngeal epithelium was acutely isolated from 11 wild-type C57BL / 6 mice (6 females and 5 males, 3 technical replicates) and 36 Calhm3-tdTomato mice (18 females and 18 males, 4 technical replicates). Animals were anesthetized with an overdose of sodium pentobarbital and transcardially perfused with PBS. The larynx was dissected and an incision was made dorsally to expose the inner surface. The laryngeal epithelium (mouse) from the glottic epithelium to the subglottis was enzymatically dissected.

[0047] Specifically, the laryngeal epithelium (mouse) from the supraglottic to subglottic region was enzymatically stripped by subepithelial injection of 100 μl of a freshly prepared enzyme mixture (DNase I: 1 mg / ml, Dispase II: 16 U / ml, diluted in Tyrode's solution) or 250 μl (guinea pig) by incubation at room temperature for 20 min. The stripped epithelium was then incubated in 100 μl of papain solution at 37°C for 20 min. The papain solution consisted of papain (26 U / ml: Worthington Biochemical, Lakewood, NJ, USA, catalog LS003126; DNase I, 1 mg / ml), DNase I (1 mg / ml), DL-dithiothreitol (2 mM), Sigma-Aldrich, cat. D9779), L-cysteine ​​(10 μg / ml, Sigma-Aldrich, cat. C7352), and Ca. 2+ The tissue was dissolved in Tyrode's solution containing no leupeptin. 50 μl of stop solution (Tyrode's solution containing 2 μg / ml leupeptin, Sigma-Aldrich, catalog number L0649) was then added. The tissue was then triturated with a glass capillary pipette to prepare a single-cell suspension. This cell suspension was used for scRNA-seq library preparation, as described below.

[0048] Single cells were isolated from the cell suspension by FACS onto 96-well plates, and scRNA-seq libraries were prepared using Smart-seq3 (M Hagemann-Jensen et al., Nat Biotechnol. 2020 Jun;38(6):708-714. doi: 10.1038 / s41587-020-0497-0.), a plate-based full-length scRNA-seq method. Alternatively, single cells were isolated from the cell suspension into droplets using the Chromium Single Cell 3' Reagent Kits v3.1 (10x Genomics), and droplet-based scRNA-seq libraries were prepared.

[0049] Next, the cDNA sequences, cellular barcodes, and molecular barcodes of the scRNA-seq library were sequenced using a DNA sequencer DNBSEQ G400RS (MGI Tech).

[0050] 3. Preparation of Target Enriched Libraries by Linear Amplification and PCR Amplification First, a full-length cDNA library containing the target (template) was prepared with the Chromium Single Cell 3' Reagent Kits v3.1 (10x Genomics), and a gene-specific forward primer (F1 primers listed below) was used to linearly amplify specific first-stranded cDNA by 20 cycles of denaturation-annealing-extension. Next, the Read1 primer (5'-CACTCTTTCCCTACACGACGCTCTTCCGAT*C*T-3': SEQ ID NO: 1, *phosphorothioate linkage) was added to the linear amplification reaction, and specific double-stranded cDNA was amplified by three cycles of PCR (Round 1). To further enrich for specificity and efficiency, two additional rounds (Rounds 2 and 3) of linear and PCR amplification were performed using nested gene-specific forward primers (F2 and F3 primers listed below). Mouse Calhm1-F1: 5'-ATGGATAAGTTTCGGATGATCTTCCAGTTC-3' (SEQ ID NO: 2) Mouse Calhm1-F2: 5'-GCAATCCAACCAAGAGTCCTTCATGAATGG-3' (SEQ ID NO: 3) Mouse Calhm1-F3: 5'-CATCTGTGGCATCATGGCGCTGGCCAGTGC-3' (SEQ ID NO: 4) Mouse Calhm3-F1: 5'-ATGGATAGGTTCCGGATGCTCTTCCAGC-3' (SEQ ID NO: 5) Mouse Calhm3-F2: 5'-ACCTCCAGTCCAGCTCGGAGTCGGTGATGA-3' (SEQ ID NO: 6) Mouse Calhm3-F3: 5'-GGCATTTGCCTCCTGCTGGCTGCTGTCAC-3' (SEQ ID NO: 7) Mouse Scn2a-F1: 5'-AGGAGGTGTCTGCTATTGTCATCCAGCGAG-3' (SEQ ID NO: 8) Mouse Scn2a-F2: 5'-CTTACAGACGCTATCTTCTGAAACAGAAAG-3' (SEQ ID NO: 9) Mouse Scn2a-F3: 5'-GACAAGGGTAAAGAAGATGAGGGAACGCCCATC-3' (SEQ ID NO: 10) mouse T2r105-F1: 5'-ATGCTGAGTGCGGCAGAAGGCATCCTCC-3' (SEQ ID NO: 11) Mouse T2r105-F2: 5'-TTCCATTGCAACTGTTGAAGCTGGGCTGGG-3' (SEQ ID NO: 12) mouse T2r105-F3: 5'-GGGAACACATTTATTGCACTGGTAAACTGCATGG-3' (SEQ ID NO: 13)

[0051] The linearly amplified and PCR-amplified target cDNA was fragmented using Chromium Single Cell 3' Reagent Kits v3.1 (10x Genomics), purified using 0.6 volumes of SPRIselect bead solution (Beckman Coulter Life Sciences, Brea, CA, 648 USA, cat. B23318), and finally converted into a 10x Genomics v3.1 library.

[0052] 4. Bioinformatics Analysis The raw sequencing data from plate-based scRNA-seq were processed using the zUMIs (S Parekh et al., Gigascience. 2018 Jun; 7(6): giy059. doi: 10.1093 / gigascience / giy059) pipeline, and the raw sequencing data from droplet-based scRNA-seq were processed using the CellRanger (10x Genomics) pipeline to obtain gene-cell expression matrices for each cell. Bioinformatics analysis was then performed using the single-cell analysis tool Seurat (https: / / satijalab.org / seurat / ).

[0053] 5. Immunohistochemical staining Mice subjected to immunohistochemistry were anesthetized and perfused with cold PBS followed by 4% paraformaldehyde in PBS (cold 4% PFA-PBS). Tissues were then harvested and fixed overnight in 4% PFA-PBS at 4°C. Fixed tissues were cryoprotected sequentially in 20% sucrose in PBS, followed by 30% sucrose in PBS at 4°C, embedded in OCT compound, and frozen sections were prepared.

[0054] For immunohistochemistry staining of P2X3, cytochrome c, CALHM1, and TRPM5, antigen retrieval was performed using antigen retrieval solution (for P2X3, cytochrome c, and CALHM1, DAKO, Carpinteria, CA, catalog S1700) or antigen retrieval solution pH 9 (for TRPM5; DAKO, catalog S2386) before blocking. Sections were incubated in antigen retrieval solution preheated to 80°C for 20 minutes and then washed with PBS.

[0055] The primary antibodies used for immunochemical staining were as follows: chicken anti-GFP(for staining GCaMP3, EYFP, and GRAB ATP1.0Aves Labs, Davis, CA, USA, cat. GFP-1020)、goat anti-tdTomato(SICGEN, Cantanhede, Portugal, cat. AB8181)、rabbit anti-RFP(for staining tdTomato; Rockland Immunochemicals, Limerick, PA, USA, cat. 600-401-379)、rabbit anti-TRPM5(1 / 500; Alomone). labs, Jerusalem, Israel, cat. ACC-045); rabbit anti-P2X2(1 / 500; Sigma-Aldrich, cat. P7982); Biotechnology, Dallas, TX, USA, cat. sc-206), guinea pig anti-PLCB2 (1 / 200 to 1 / 500; kindly gifted by Dr. Thomas E. Finger), guinea pig anti-CALHM166 (1 / 10,000), rat anti-EpCAM (1 / 1,000; Santa Cruz Biotechnology, cat. sc-53532); cat. ab31704).

[0056] Other anti-mouse IgG, Alexa Fluor The 405 has a 250-inch range.donkey anti-mouse IgG, Alexa Fluor 405(Abcam, cat. 175659)、 donkey anti-guinea pig IgG, biotinylated(Millipore, Burlington, MA, USA, cat. AP193B)、donkey anti-rat IgG, Alexa Fluor Plus 405(Thermo Fisher Scientific, cat. A48268)、donkey anti-rabbit IgG, Alexa Fluor Plus 405(Thermo Fisher Scientific, cat. A48258)、donkey anti-rat IgG, Alexa Fluor 488(Thermo Fisher Scientific, cat. A-21208)、donkey anti-mouse IgG, Alexa Fluor 488(Thermo Fisher Scientific, cat. A-21202)、donkey anti-chick IgY, Alexa Fluor 488(Jackson ImmunoResearch, West Grove, PA, USA, cat. 703-545-155)、donkey anti-goat IgG, Alexa Fluor 488(Thermo Fisher Scientific, cat. A-11055)、donkey anti-goat IgG, Alexa Fluor 546(Thermo Fisher Scientific, cat. A-11056)、donkey anti-rabbit IgG, Alexa Fluor 546(Thermo Fisher Scientific, cat. A-10040)、donkey anti-rabbit IgG, Alexa Fluor 647(Thermo Fisher Scientific, cat. A-31573)、donkey anti-goat IgG, Alexa Fluor 647(Thermo Fisher Scientific, cat.A-21447), donkey anti-mouse IgG, Alexa Fluor 647 (Thermo Fisher Scientific, cat. A-31571). .

[0057] 6. Physiological Measurement of Swallowing Mice were anesthetized by intraperitoneal injection of urethane (1.3 g / kg body weight, Sigma-Aldrich, catalog number U2500). The trachea was surgically exposed and transversely cut below the cricoid cartilage at the level of the 5-6 tracheal rings. An inflow cannula was inserted through the mouth into the hypopharynx and positioned above the laryngeal inlet, and an outflow cannula was inserted proximally into the cut trachea. The inflow and outflow cannulas were connected to a peristaltic pump (Minipuls Evolution, Gilson, WI, USA, catalog number F110701). To avoid solution leakage and mechanical reactions caused by cannula movement, the delivery tube was secured with silicone sealant (DentSilicon-V and Duplicon, Matsukaze, Kyoto, Japan). Respiratory dynamics were measured by placing a thermistor (e.g., Warner Instruments, Hamden, CT, USA, catalog number TC-324C) in the lower trachea. The thermistor signal was connected to a temperature monitor (Warner Instruments, Hamden, CT, USA, catalog number AN6N4-GC11KA143L / 37C) and recorded with a data acquisition system (e.g., PowerLab8 / 35, AD Instruments, Colorado Springs, CO, USA, catalog number PL3508). The collected data was used to monitor the contraction of the digastric muscle via electromyography (EMG). A 0.30 mm diameter silver wire (Nilaco Corporation, Tokyo, Japan, catalog number AG-401325) was inserted into the muscle. The muscle signal was bandpass filtered in the 0.3–10 kHz range, amplified 1,000 times, and acquired using a differential amplifier (World Precision Instruments, Sarasota, FL, USA, catalog number DAM-80) and the PowerLab8 / 35.

[0058] The EMG power was calculated using Matlab (registered trademark) (MathWorks, Neatec, MA, USA) after passing the signal through a 0.3–1 kHz bandpass filter. Hyoid movements were recorded with a video camera (60 fps, Shodensha, Osaka, Japan, catalog number BA200HD).

[0059] Water, saline, and other stimulus solutions were perfused through the inflow cannula at a constant rate (0.1 ml / min). Saline (containing 10 mM HEPES, 154 mM NaCl, pH 7.4) was used as the baseline solution.

[0060] Various chemicals were dissolved in saline to form stimulation solutions: concentrated salt (750 mM, NaCl), citric acid (25 mM, pH 2.8), cycloheximide (10 mM, Sigma-Aldrich, catalog no. C7698), denatonium benzoate (10 mM, Sigma-Aldrich, catalog no. D5765), acesulfame K (30 mM, Sigma-Aldrich, catalog no. 04054), SC-45647 (2 mM), monosodium glutamate and inosine 5'-monophosphate (30 and 1 mM, Sigma-Aldrich, catalog nos. G5889 and I4625).

[0061] To visualize the timing of stimulus administration, Fast Green FCF (0.1 μg / μl; Nacalai Tesque, Kyoto, Japan, catalog number 15939-54) was added to the stimulus solution. Chemical stimulation was performed by perfusing saline for at least 4 minutes, followed by 30 seconds of stimulation with one chemical. When consecutive stimuli were administered to the same animal, saline perfusion was performed for at least 4 minutes, followed by washing, followed by 30 seconds of stimulation.

[0062] During stimulation, observations were made for a swallowing response, which was identified by the simultaneous occurrence of apnea, contraction of the digastric muscle, and elevation of the hyoid bone.

[0063] The effect of AF-353 (Selleck Biotech, Tokyo, Japan, catalog S0405), an inhibitor of P2X3, was evaluated as follows.

[0064] First, the animals were perfused with saline containing 0.5% propylene glycol (0.5% PG saline, vehicle) for 15 minutes, after which the chemicals were administered and swallowing responses were counted. Next, the animals were perfused with 0.5% PG saline containing 125 μM AF-353 (AF-353) for 15 minutes, after which the same chemicals were administered and swallowing responses were counted. After this, the animals were perfused with 0.5% PG saline, and after AF-353 was washed out for at least 30 minutes (recovery), the same chemicals were administered and swallowing responses were counted. The chemicals were prepared in 0.5% PG saline or 0.5% PG saline containing AF-353 (125 μM).

[0065] 7. Optogenetic Analysis of PCCs CALHM1 +For optogenetic activation of PCCs, four throat regions of anesthetized Calhm1-ChR2 mice were illuminated with blue light at 10 mW, 10 Hz, and 50 ms pulses for 10 seconds using the tip of an optical fiber. A Civil Lasers (Hangzhou, China, catalog number NDB7875-E / TTL) blue laser source and a RIGOL (Beijing, China, catalog number DS1022Z) function generator were used. The optical fiber (400 μm core diameter, 0.39 NA, Thorlabs, Newton, NJ, USA, catalog number FT400EMT) was positioned in the target region using a micromanipulator (Narishige Co., Ltd., Tokyo, Japan, catalog number SM-25A). Light from the optical fiber was delivered to the following four locations: (1) When the optical fiber was positioned toward the dorsal surface of the subglottis, the entire hypopharyngeal region containing the taste buds and a portion of the subglottic epithelium were illuminated. (2) When the optical fiber was placed in the proximal trachea and oriented longitudinally toward the subglottic cavity, it primarily illuminated the entire subglottic epithelium and, presumably, the hypopharynx due to light penetration. (3) The light was irradiated perpendicularly to the lower trachea. (4) The light was irradiated above the superior laryngeal nerve (see Figure 8e below). When the light was irradiated at positions (3) and (4), the hypopharynx and larynx were not exposed to light. Swallowing and coughing events were counted during the 10-second light pulse stimulation, and the average count was calculated from multiple trials.

[0066] The dynamics of the glottic opening were obscured by direct light illumination of the subglottis and surrounding tissues (see Figure 8e). Therefore, optogenetic analysis was limited to observing respiration and digastric electromyography (EMG). To evaluate the ability to accurately classify respiration, coughing, and swallowing during optogenetic stimulation, we constructed an unsupervised classification model (k = 3) using the Scipy module in Python and trained it with the k-means algorithm (k = 3) using the IndexEMG and IndexResp datasets. The agreement between manual and k-means classification was 98.0%, and the results of the k-means algorithm mirrored the manual classification.

[0067] In the experiment shown in Figure 13j, the effect of AF-353 was assessed as follows: Optogenetically induced cough responses were measured after 15 min of subglottic application of PG saline (2 μL, 0.5%). Cough responses were then measured after 15 min of treatment with AF-353 (2 μL of 125 μM in 0.5% PG saline) and 20 min after repeated intensive irrigation of the subglottic surface with saline.

[0068] 8. Alternaria extract administration On days 0 and 3, mice were anesthetized with isoflurane (4.5% for induction and 2.0–2.5% for maintenance; Mylan, Tokyo, Japan) using an inhalation anesthesia device (Univentor 400 Anesthesia Unit, Univentor, Zeitung, Malta, Catalog No. U-410). Alternaria alternata extract (6.83 mg / mL, ITEA Inc., Tokyo, Japan, Catalog No. 10117) or vehicle (filter-sterilized PBS) solution was administered as follows: 30 μL of Alternaria alternata extract solution or vehicle solution was administered through the mouth using a 24 G blunt needle placed under otoscopic observation (Hallowell EMC, Pittsfield, MA, USA, Catalog No. 000A3754) to reach the subglottic space just below the vocal cords. On day 4, mice were subjected to a coughing experiment using Den (denatonium) or CA (citric acid).

[0069] 9. Ca in Single PCCs 2+ and ATP imaging Calhm1-GCaMP3 and Calhm1-GRAB were analyzed according to the method described above in Section 2. Single-cell RNA sequencing (scRNA-seq). ATP1.0 Subglottic epithelium was enzymatically stripped from mice. For fluorescence imaging, the isolated epithelium was flattened onto a thin silicone sheet, SYLGARD 184 (Dow, Midland, MI, USA, ~200 μm thick), mounted in a perfusion chamber (Warner Instruments, catalog no. RC-27L) and permeated with Tyrode's solution.

[0070] GRAB, a GFP-based fluorescent sensor ATP1.036 Imaging of GCaMP335 and GCaMP335 was performed using an epifluorescence microscope as previously described. Chemical stimuli, including Den (denatonium benzoate), QHCl (quinine hydrochloride dihydrate, 5 mM; Tokyo Chemical Industry Co., Ltd., Tokyo, Japan; catalog number Q0030), HSL (3-oxo-C12-homoserine lactone, 0.3 mM; Sigma-Aldrich; catalog number O9139), and ATP (adenosine 5'-triphosphate disodium salt hydrate, 10 μM; Sigma-Aldrich; catalog number A2383), were dissolved in Tyrode's solution and administered to tissues fixed on a silicone sheet using a fast-step perfusion system (VC-8 valve controller and SF-77B Fast-Step Perfusion System; Warner Instruments). As a control for mechanical stimulation, Tyrode's solution was administered in the same manner. All recordings were performed at room temperature (24 ± 1 °C). After imaging, Calhm1-GRAB ATP1.0 Mouse epithelial sheets were fixed in 4% PFA-PBS overnight at 4°C and subjected to immunofluorescence staining, except that primary antibody labeling was performed at 4°C.

[0071] Fluorescence images were preprocessed with MatLab® for motion correction (NoRMCorre73). 2+ For imaging, source extraction (constrained non-negative matrix factorization of microendoscopic data) and baseline subtraction were performed using the CaImAn package in MatLab®. The extracted data were normalized using R version 4.2.2 (https: / / www.R-project.org / ). For ATP imaging, circular ROIs (10-pixel diameter) were manually set in Fiji, and the extracted data were normalized using IgorPro (waveMetrics, Lake Oswego, Oregon, USA). Z-scores were used for this normalization. The Z-score was calculated as (F-(mean[F baseline ])) / SD[F baseline] where F baseline is the fluorescence signal for 20 seconds immediately before the onset of stimulation. A representative pseudocolor image of the peak fluorescence change was generated in Fiji after image denoising.

[0072] 10. Physiological Measurement of Cough in Mice Mice were anesthetized by intraperitoneal injection of urethane (1.3 g / kg body weight). Respiratory secretions were suppressed by intraperitoneal administration of atropine (0.1 mg / kg; Tanabe Mitsubishi Pharma, Osaka, Japan). The trachea was surgically exposed, and a small window was made in the anterior wall below the cricoid cartilage to expose the subglottic space. Respiratory dynamics and digastric muscle activity were measured as described in Section 6. Physiological Measurements of Cough. The glottic opening (i.e., the glottis) was observed from below using a 1.7 mm diameter endoscope (Vison Well Corporation, Tokyo, Japan, catalog number 1.7-010-030-065) inserted through the tracheal window and placed in the tracheal window. The endoscope was connected to a high-speed monochrome camera (120 fps, Argo Corporation, Osaka, Japan, catalog number DMK33UP1300). Chemicals (PBS, 20 mM denatonium benzoate in PBS, and 25 mM citric acid in water) were sprayed as aerosols into the subglottic cavity through the tracheal window using a mist sprayer (Qomolangma Brave Knight, Seattle, WA, USA, catalog number B08MZ9JX3N) positioned 15 cm from the animal. Each spray lasted 3 seconds, and cough responses were counted within 10 seconds after the spray began. For mechanical stimulation, airflow through a 0.8 mm diameter tube (50 kPa) was delivered into the throat from a distance of 1–2 cm for 10 seconds.

[0073] Physiological parameters (respiratory dynamics, EMG signals, and glottal opening area) were quantified using Matlab®. Respiratory dynamics scalograms (time-resolved power across frequencies) were generated by wavelet transformation. Bandpass-filtered signals (0.5–30 Hz) were convolved with Complex Morlet wavelets. The effect of stimulation on respiratory dynamics was analyzed by substituting the values ​​obtained from the scalogram into the following equation:

[0074]

number

[0075] where P f pre and P f post The indexes (x, y, y, y) indicate the mean power at each frequency (2-10 Hz) before and after stimulation, respectively. Negative and positive values ​​of this index (ranging from -1 to 1) indicate decreases and increases in respiratory rate and depth, respectively.

[0076] The glottic opening was observed using endoscopically recorded video at 120 fps. The region of interest (ROI) was calculated by calculating the pixel intensity difference between each frame from -10 s before stimulus onset to 20 s after stimulus onset, with stimulus onset being set as 0. This setting was set to include the glottic opening at maximum opening. The brightness intensity of each pixel was measured, and the ROI was binarized by thresholding. The glottic opening within the ROI was defined as the binarized zero-intensity region. A cough response was identified by an initial deep inspiration with glottal expansion (inspiratory effort), followed by a strong, long expiration against a constricted glottis without digastric muscle activity (expiratory effort).

[0077] 11. Gaussian Mixture Models with Bayesian Information Criterion Based on the patterns of the three physiological parameters measured in this study (respiratory dynamics, glottal opening dynamics, and digastric muscle activity), it was easy to manually classify events as normal breathing, coughing, and swallowing. To objectively validate the manual classification, we performed cluster analysis to assess how many distinct clusters the physiological events in the data could be classified into. Briefly, we calculated three indices (IndexResp, IndexGA, and IndexEMG) for each physiological event and classified them into 1–10 clusters using a three-dimensional Gaussian mixture model (GMM). Finally, to determine the "true" number of clusters, we calculated the Bayesian information criterion (BIC) for each clustering using the following formula: Generally, the number of clusters with the smallest average BIC value is estimated to be the optimal classification result.

[0078] BIC=-2ln(L)+kln(n) where L, k, and n denote the likelihood of each physiological event, the number of parameters estimated by the model, and the number of physiological events, respectively. GMM fitting using the expectation-maximization algorithm was repeated 1,000 times for 1–10 clusters, and the BIC was calculated for each iteration. For further validation, x-means clustering was used. These methods were implemented using custom scripts in Python ver. 3.9 (Python Software Foundation, Wilmington, Delaware, USA) with additional modules installed, including scikit-learn (GMM fitting and BIC calculation) and PyClustering (x-means clustering).

[0079] 12. General Statistical Analysis The sample size was determined to be adequate by post-hoc power analysis using G*Power 3.178 (power = 0.8, alpha = 0.05). Other common statistical analyses were performed using unpaired or paired t-tests, analysis of variance with post-hoc Dunnett's or Tukey-Kramer tests. All analyses were two-tailed. These statistical analyses were performed using the Statistics and Machine Learning Toolbox in Matlab®.

[0080] II. Results 1. Laryngeal PCC Channel Synapses Using the known channel synapse-specific proteins CALHM1 / 3 as a guide, we searched for extraoral channel synapses in the following five steps.

[0081] In the first step, we generated transgenic dual reporter mice and confirmed the distribution of cells expressing the reporter genes. These mice express the reporter genes GCaMP3 under the control of the Calhm1 promoter and tdTomato under the control of the Calhm3 promoter. We screened 44 organs for the expression of the two reporter genes in the transgenic dual reporter mice. Cells expressing both reporter genes (hereafter referred to as "CALHM1 / 3") were then screened. + These cells (referred to as "Cahlm1 / 3 cells") were present in the vomeronasal organ, throat, trachea, and thymus, but not in most other organs (Fig. 1a-d, Fig. 2). On the other hand, cells expressing both reporter genes in the vomeronasal organ, throat, trachea, and thymus were rare EpCAM-positive cells. Expression of Calhm1 and Calhm3 mRNA in each organ was confirmed by RT-PCR (Fig. 3). In the throat, CALHM1 / 3 + The cells were concentrated in taste bud-like structures located on the dorsal surface of the arytenoids of the hypopharynx (hereafter referred to as "hypopharyngeal taste buds"). CALHM1 / 3 cells were also expressed on the laryngeal surface, including the supraglottic and subglottic epithelia. + The cells were present in isolation.

[0082] In the second step, immunohistochemical analysis was performed to identify CALHM1 / 3 + The cell type of the cells was identified. + The cells were found to express the PCC marker proteins PLCB2 and TRPM5 (Fig. 1e, Fig. 4). (Hereafter, PLCB2-expressing cells are referred to as "PLCB2 + cells,” and TRPM5-expressing cells were “TRPM5 + (Hereinafter referred to as "PLCB2 cells"). + CALHM1 / 3 in cells + The proportion of cells varied between tissues and vice versa, suggesting intra-tissue heterogeneity of PCCs (Fig. 1f, g).

[0083] In the third step, we investigated the distribution of afferent nerve fiber markers VGLUT2 and CALHM1 / 3. + The distribution of PCC was observed by immunohistochemical analysis. CALHM1 / 3 was detected in the distal trachea and thymus. + In tissues containing PCCs, VGLUT2-positive (hereafter referred to as "VGLUT2 + ") nerve fibers and CALHM1 / 3 + Contact with PCC was observed, and CALHM1 / 3 + Contact between the cells and the afferent nerve was suggested (Fig. 5). + CALHM1 / 3 cells were detected most frequently in the throat along the respiratory tract (Figure 6). + This suggests that cells are physiologically important in this tissue region.

[0084] As a fourth step, we investigated the afferent vagus nerve markers P2X2 and P2X3, and CALHM1 / 3. + The distribution of PCC was observed by immunohistochemical analysis. CALHM1 / 3 in both the hypopharynx and larynx + PCCs express P2X2 (Fig. 1h) and P2X3 (Fig. 1i) (respectively, “P2X2 + nerves, P2X3 + The cerebellum was innervated by vagal afferents (referred to as "nerve").

[0085] As a fifth step, high-resolution immunofluorescence imaging of throat PCCs allowed us to identify synaptic mitochondria, CALHM1, and P2X2 + The distribution of the neurons was observed. As a result, the synaptic mitochondria, CALHM1, and P2X2, which are characteristic of channel synapses, were detected. + The close arrangement of nerves was identified (Fig. 1j-p).

[0086] All these data strongly suggest the existence of channel synapses between vagal neurons and a subset of PCCs in the hypopharyngeal and laryngeal epithelia.

[0087] 2. Purinergic PCCs in the hypopharynx trigger the swallowing reflex Because PCCs in hypopharyngeal taste buds express channel synapses, we investigated whether PCCs in hypopharyngeal taste buds express other proteins in common with PCCs in lingual taste buds.

[0088] First, we performed full-length single-cell RNA sequencing (scRNA-seq) of taste buds isolated from the hypopharyngeal epithelium of Pkd2l1-GCaMP3 mice. In these mice, taste buds fluoresce due to the fluorescent protein GCaMP3, which we used as a landmark for retrieval. We obtained a total of 432 single-cell transcriptomes, with an average of 8,594 genes detected per cell, and identified 12 cell classes by unsupervised clustering (Figure 7a, b).

[0089] The results showed that hypopharyngeal and lingual taste buds share some common cellular components, but have different cellular compositions. Specifically, both hypopharyngeal and lingual taste buds express Krt5. + Progenitor cells, Shh + Progenitor cells, Entpd2 + Supporting cells, Pkd2l1 + Acid-sensing cells, and Pou2f3 + They had PCC (also known as type II) cells.

[0090] On the other hand, hypopharyngeal PCC is T1r2 / T1r3+ (Cluster 9) and T2r + The subclasses (cluster 10 / 11) accounted for 6.0% and 10.6%, respectively, but no subclass expressing only T1r1 / T1r3 was observed in the lingual taste bud PCC.

[0091] T1R1 / T1R3 and T1R2 / T1R3 are heterodimeric GPCRs that respond to amino acids and sugars, respectively. T2Rs are a group of GPCRs that respond to a variety of substances, including toxic, irritant, metabolic, and infectious agents.

[0092] Cluster 10 / 11 expressed virtually all T2r. Calhm1 and Calhm3 expression was predominant in clusters 9 / 10 / 11 (Fig. 7b, c).

[0093] Cells in clusters 9 / 10 / 11 also expressed all proteins that link receptor (T1R and T2R) activation to CALHM1 / 3 activation in tongue PCCs, including Gnat3, Plcb2, Itpr3, Trpm5, and voltage-gated Na channels (Scn genes) (Fig. 7b, c).

[0094] These results suggest that the gene expression profiles analyzed by scRNA-seq suggest that hypopharyngeal PCCs sense T1R and T2R agonists and transmit information to the vagus nerve via purinergic channel synapses.

[0095] Cholinergic neurotransmission mediated by PCCs in the nasal and tracheal epithelium is involved in the inhibitory control of breathing, but the involvement of purinergic PCCs in airway neural reflexes is unknown.

[0096] Swallowing is an airway defense mechanism that safely transports ingested materials into the esophagus. Swallowing consists of three phases: oral, pharyngeal, and esophageal. The hypopharynx plays an active role in initiating the involuntary pharyngeal phase. Here, we administered various chemicals to the hypopharyngeal and laryngeal surfaces by perfusion and observed swallowing responses to these chemicals (Figure 8a). In anesthetized wild-type (WT) mice, repetitive swallowing was induced by typical swallowing-inducing stimuli (water, high salt concentration, and citric acid), but not by T1R agonists (acesulfame K, SC-45647, and glutamate) (Figure 8b, d). These findings are consistent with previously published studies. Meanwhile, this study demonstrates for the first time that T2R agonists (cycloheximide and denatonium) also induce swallowing (Figure 8d). Calhm3 knockout (KO) and the selective P2X3 receptor antagonist AF-353 inhibited T2R ligand-induced swallowing (Fig. 8f), indicating the possible involvement of channel synapses and purinergic extracellular signals, respectively.

[0097] Furthermore, in mice expressing channelrhodopsin 2 (ChR2) under the control of the Calhm1 promoter (Calhm1-ChR2), blue light (445 nm) irradiation of the hypopharyngeal epithelium induced more swallowing than did irradiation of the subglottic surface, trachea, and superior laryngeal nerve (Fig. 8e, g), whereas blue light did not induce swallowing in control mice that did not express ChR2. These results suggest that depolarization of hypopharyngeal PCCs is sufficient to trigger swallowing.

[0098] These data suggest that activation of T2R and subsequent channel synapses in hypopharyngeal PCCs (cluster 10 / 11) triggers swallowing.

[0099] 3. Purinergic PCCs in the larynx Next, CALHM1 / 3 + We investigated the role of channel synapses in the laryngeal epithelium, another area where PCCs exist.

[0100] First, we performed high-throughput droplet-based 3' scRNA-seq in the laryngeal epithelium from the supraglottis to the subglottis of wild-type mice. We first profiled the transcriptomes of 56,013 single cells, detecting an average of 3,679 genes per cell. Using unsupervised clustering, we identified 10 epithelial cell types, including PCCs, which accounted for 1.0% of the total population, and two immune cell classes (Fig. 9a).

[0101] Pou2f3 + Re-clustering only PCCs confirmed their subdivision into six subclasses, designated PCC1-6 (Figure 9b, c). Contrary to expectations generated by reporter screening (Figure 1), Calhm1 and Calhm3 were barely detected in the original dataset, suggesting low transcription levels (Figure 9d). Therefore, we developed a specific and efficient target enrichment method for the cDNA library, combining linear amplification with nested gene-specific primers and PCR (Figure 10a, b). This target enrichment analysis revealed that the rare PCC subclass, PCC5, which accounts for 8.7% of PCCs and 0.091% of the total epithelial population, abundantly expresses Calhm1, Calhm3, T2r105, and Scn2a (Figure 9d). This suggests that subclass PCC5 has the ability to convert ATP release in response to T2R ligands present in the luminal cavity.

[0102] Laryngeal PCCs can also be classified by the combined expression of the canonical PCC markers Plcb2 and Dclk1. Subclass PCC5 expresses Plcb2 + / Dclk1 -The phenotypes of T2r members were characterized by the following: (Fig. 9b, c, d). High-throughput 3' scRNA-seq analysis revealed that only two T2r members (referred to as "Tas2r" in Fig. 9 and Fig. 11), T2r104 and T2r105, were detected in the subclass PCC5 (Fig. 11a, b). To investigate gene expression in more detail, Calhm3 cells collected from Calhm3-tdTomato mice by fluorescence-activated cell sorting were analyzed. + We performed full-length scRNA sequencing on laryngeal epithelial cells. We profiled 661 cells (average 8,740 genes per cell) and identified Plcb2 + / Dclk1 - The pattern identified the subclass PCC5.

[0103] Highly sensitive full-length scRNA sequencing revealed that each subclass PCC5 expresses many T2R members and two Scn genes (Figure 11a, b). These results indicate a wide chemoreceptive range and electrical excitability. Transcripts for Calhm1, Calhm3, and Scn genes were not present in previously reported single-cell transcriptome atlases of tracheal PCCs. Furthermore, all previously reported solitary PCCs with neural functions are characterized by Chat expression and cholinergic neurotransmission. However, Chat expression was not prominent in the subclass PCC5 among laryngeal PCCs (Figure 9c). Therefore, subclass PCC5 appears to have distinctive gene expression among solitary PCCs. The gene expression of subclass PCC5 is consistent with that of T2Rs in the tongue and hypopharyngeal taste buds, as suggested by the expression of canonical T2R signaling-related molecules and gene ontology analysis. + Similar to PCC (Fig. 9d), the Calhm1 expression in the exfoliated subglottic epithelium was significantly increased when measured using GCaMP3 fluorescence as an index. + Cells release intracellular Ca2 + Increasing concentrations responded to different classes of T2R ligands, including denatonium, quinine, and 3-oxo-C12-homoserine lactone (Fig. 9f).

[0104] Finally, we tested whether laryngeal PCCs release ATP in response to T2R stimulation. We used the genetically encoded extracellular ATP sensor protein, GRAB, to detect ATP release from a small plasma membrane region of rare cells. ATP1.0 A Cre-dependent reporter allele of GRAB was engineered under the control of the Calhm1 promoter. ATP1.0 Mice expressing Calhm1-GRAB ATP1.0 We generated a new generation of high-expression reporter (TIGRE2.0) in laryngeal PCCs (Fig. 12a). ATP1.0 In this mouse model, GRAB expression was confirmed in response to denatonium. ATP1.0 Cells exhibiting a rapid, localized increase in cell surface fluorescence were observed (Fig. 12b, c). The localized ATP release evoked by T2Rs is consistent with the localization of CALHM1 / 3 channels at the synaptic site of ATP release (Fig. 1n-p).

[0105] 4. Cough induction by PCC Based on the anatomical location, gene expression, T2R responsiveness, and ATP release potential suggested by cellular functions, we hypothesized that subclass PCC5 mediates cough in response to noxious substances. Cough is defined by coordinated glottal and respiratory movements.

[0106] However, despite indirect evidence from acoustic and whole-body plethysmography, the ability of mice to cough remains controversial. To directly assess the cough reflex in anesthetized mice, we applied a spatiotemporally controlled chemical stimulus (in the form of an aerosol) to the subglottic surface while simultaneously observing glottal opening, respiration, and digastric muscle activity (Figure 13a). Citric acid, a typical cough stimulus, evoked a characteristic response in mice that differed from normal breathing and swallowing (Figure 13b–d). Specifically, this response consisted of forced inspiration through an open glottis followed by forced expiration through a narrowed glottis, without the digastric muscle activity characteristic of coughing. These results demonstrate that mice are capable of coughing, and that the symptom can be reliably identified by its characteristic breathing pattern and glottal movement. In this experimental model, nebulization of denatonium, a T2R ligand, was confirmed to induce coughing in half of the tested mice (Figure 13e, g). Even in mice that did not exhibit coughing, denatonium induced significant changes in respiration, observed as an increase in amplitude, expressed as a change in respiratory power (Fig. 13e, h). Calhm3 KO completely inhibited denatonium-induced coughing and respiration changes, strongly suggesting the involvement of channel synapses. On the other hand, acid-induced coughing and respiration changes were unaffected in Calhm3 KO. These results demonstrate that channel synapses mediate T2R ligand-induced coughing but are not involved in acid-induced coughing (Fig. 13f-h).

[0107] Because channel synapses are activated by depolarization, we tested whether optogenetic activation of subglottic PCCs in Calhm1-ChR2 mice could induce coughing. Optogenetic activation of PCCs by irradiating the subglottis with blue light potently induced coughing, with the effect decreasing with distance from the subglottis (Figure 8e, Figure 13i). This indicates that depolarization of subglottic epithelial PCCs induces coughing. Furthermore, AF-353 suppressed coughing induced by optogenetic activation of PCCs (Figure 13j). This result indicates the involvement of purinergic neurotransmission in the initiation of coughing.

[0108] Finally, because the most well-described function of PCCs is their role in initiating and amplifying mucosal type 2 immunity to helminths and allergens, and type 2 inflammation is a hallmark of allergic cough hypersensitivity in chronic airway diseases such as asthma, we observed a role for the subclass PCC5 in pathological cough. In mice subglottically administered with a mold aeroallergen (Alternaria extract), the cough reflex to denatonium was significantly enhanced compared to mock-treated mice. Furthermore, denatonium-induced coughing was absent in both mock-treated and Alternaria-treated Calhm3 KO mice (Figure 13k).

[0109] These data suggest that epithelial aeroallergen sensing enhances cough sensitivity to T2R stimulation and that the PCC5 subclass is involved in the induction of this allergic cough hypersensitivity.

Claims

1. Isolated laryngeal Pou2f3-expressing cells expressing Calhm3 protein.

2. The laryngeal Pou2f3-expressing cell of claim 1, which further expresses Calhm1 protein.

3. A laryngeal Pou2f3-expressing cell as described in claim 2, which does not express Dclk1 protein.

4. 1. A method for monitoring a cough response in a mouse, comprising: administering a test substance into the subglottic cavity; Observe respiratory and glottic opening movements and a monitoring method comprising:

5. It also includes observing digastric muscle dynamics. The administration of the test substance does not change the dynamics of the digastric muscle. The monitoring method according to claim 4.

6. 5. The monitoring method according to claim 4, wherein the test substance is administered by aerosolizing the test substance.

Images