Mice
All mice were housed, and all experimental procedures were carried out in American Association for Accreditation of Laboratory, Animal Care-certified laboratory animal facilities at the University of California, San Diego (UCSD). All animal procedures were approved by the Institutional Animal Care and Use Committee at UCSD. Animals were maintained under constant environmental conditions (temperature in rooms is 68–72 °F and humidity 30–70%), with food and water provided ad libitum in a 12/12 h light/dark cycle. Adult mice from strains C57BL/6 J (no. JAX 000664), c-Kitw-sh/w-sh (no. JAX 030764), Fos2A-iCreER (TRAP2, no. JAX 030323), Rosa-lxl-tdTomato (Ai14, no. JAX 007914), Rosa-lxl-DTR (no. JAX 016603), Th-cre (no. JAX 008601), Chat-cre (no. JAX 031661), CAG-Sun1/sfGFP (no. JAX 030952) and Rosa-Zsgreen (no. JAX 007906) were purchased from the Jackson laboratory. Strain Dbh-cre (MMRRC 036778) was purchased from the Mutant Mouse Resource and Research Center. All cre lines were maintained in a B6 background and were viable and fertile with no detectable abnormal phenotypes. Both male and female mice were used for experiments. Mice were at least 6 weeks old when subjected to HDM challenge, stereotaxic injection or surgery.
HDM challenge
Mice were anaesthetized using isoflurane. 50 μg 20 μl−1 HDM extract (Dermatophagoides pteronyssinus, GREER Labs) was introduced intranasally for 4 consecutive weeks on days 0, 7, 14 and 21. For controls, 20 μl of saline was used instead of HDM on the same regime. Mice were euthanized 1.5 h following the last challenge for Fos RNAscope or nTS snRNA-seq; 2 h after the last challenge for FOS antibody immunostaining; 24 h after the last challenge for flexiVent assay, periodic acid–schiff (PAS) staining and quantitative PCR (qPCR); and 3 days after the last challenge for flow cytometry.
Lung tissue PAS staining
Mice were euthanized by CO2 inhalation. The lungs were inflated with 4% paraformaldehyde (PFA) at 35 cm H2O airway pressure; they were postfixed in 4% PFA overnight and then prepared for paraffin sections at width 6 μm. Goblet cells were stained using a PAS staining kit (Sigma).
Tissue collection and immunofluorescence staining
Mice were euthanized by CO2 inhalation followed by transcardial perfusion with PBS and 4% PFA. Subsequent to postfixing overnight in 4% PFA, tissues were washed in PBS followed by overnight sucrose dehydration. Brainstem blocks were sectioned at 25, 40 or 99 μm thickness in a rostral to caudal sequence, and lung blocks were sectioned at 99 or 300 μm thickness for parasympathetic neuron/airway staining. All sections were processed for immunostaining following the standard protocol. Primary antibodies used include rabbit anti-c-FOS (SYSY, no. 226 008, 1:300), rabbit anti-DBH (Sigma, no. AB1585, 1:300), rabbit anti-Dsred (Takara, no. 632496, 1:300), rabbit anti-VAChT (SYSY, no. 139 103, 1:300), mouse anti-alpha Smooth Muscle Actin-FITC (Sigma, no. F3777, 1:300), rabbit anti-TRPV1 (Alomone labs, no. ACC-030, 1:300) and chicken anti-GFP (abcam, no. ab13970, 1:300). Secondary antibodies used include goat anti-rabbit FITC, goat anti-rabbit Cy3 and goat anti-rabbit Cy5 (all from Jackson Immuno Research Labs, all 1:300). Slides were mounted with Vectashield (Vector Labs) and imaged using an Olympus VS200 Slide Scanner or Leica SP8 confocal microscope. For quantification of the total number of FOS+ neurons by antibody staining and the overlap between Fos and the top marker gene of each individual nTS cluster by RNAscope following HDM, we counted total cell numbers in 20 serial sections (at 25 µm thickness each, with 75 µm interval between sections, together representing ¼ of the whole nTS regions (2 mm)) and recorded these as one data point for one animal. For quantification of FOS+ neurons without the total label on the y axis, we counted three sections of nTS bregma regions with the most concentrated signals following serial section of the whole nTS (three sections were chosen between bregma −7.20 and −8.08 mm at the same stereotaxic coordinates between groups), and the average number was used as one data point for one animal. We used Qupath for overlap quantification in the nTS.
RNAscope
Vagal ganglia or brainstems were sectioned at 25 μm thickness. All staining procedures were performed using the RNAscope Fluorescent Multiplex Kit (Advanced Cell Diagnostics, no. 320850) following the manufacturer’s instructions. The following probes from Advanced Cell Diagnostics were used: Mm-Fos (no. 316921), Mm-Dbh (no. 407851), Mm-Phox2b (no. 407861), Mm-Otp (no. 516391), Mm-Mecom (no. 432231), Mm-Kcns3 (no. 467371), Mm-Mafa (no. 556931), Mm-Egr3 (no. 431101), Mm-Crym (no. 466131), Mm-Gbx2 (no. 314351), Mm-Col23a1 (no. 432681), Mm-Npy (no. 313321), Mm-Cdh3 (no. 514591), Mm-Nr4a2 (no. 423351), Mm-Lhx9 (no. 495431), Mm-Pou3f1 (no. 436421), Mm-Tac2 (no. 446391), Mm-St18 (no. 443271), Mm-Igf1 (no. 443901), Mm-Lama2 (no. 424661), Mm-Th (no. 317621), Mm-Tacr1 (no. 428781), Mm-Trpv1 (no. 313331), Mm-Gfp (no. 400281), Mm-tdTomato (no. 317041), Mm-Chat (no. 408731), Mm-Adra1a (no. 408611) and Mm-Adra1b (no. 413561).
Vagotomy
Mice were anaesthetized with a mixture of ketamine (100 mg kg−1) and xylazine (10 mg kg−1) by intraperitoneal injection (the same approach was used for all anaesthesia in our study unless otherwise notated). Fully anaesthetized mice were placed ventral side up on a stereotaxic frame, then 70% ethanol was sprayed on the throat to wet the fur. The skin was lifted to make a vertical cut (1 cm) on the throat, then one side of the vagal nerve was isolated and teased away from the carotid artery using small, curved forceps. Unilateral vagotomy was conducted by lifting the vagal nerve and cutting with straight scissors. Control sham operations were performed by lifting the vagal nerve and releasing it intact. Following vagotomy, the wounds were sutured and the area was disinfected with povidone-iodine. For all surgeries in our study, unless otherwise notated, mice were positioned on a heating pad to maintain body temperature and ophthalmic ointment was applied to maintain eye lubrication during surgery. Postoperative analgesia was provided with Buprenorphine SR (0.1 mg kg−1, subcutaneous injection). Mice were allowed to recover for 1 week before being subjected to allergen challenge.
4-OHT injection and TRAP2 labelling
4-Hydroxytamoxifen (4-OHT, Sigma, no. H6278) was dissolved at 20 mg ml−1 in ethanol by shaking at 37 °C for 15 min, followed by aliquoting and storage at −20 °C for up to several weeks. Before use, 4-OHT was redissolved in a 1:4 mixture of castor oil/sunflower seed oil (Sigma, nos. 259853 and S5007). Ethanol was fully evaporated by vacuum under centrifugation. To determine brainstem neurons activated by allergen challenge but not by food consumption30, mice were fasted for 12 h before allergen challenge and 4-OHT injection; they were then placed back on a regular diet following cre activation. Mice were maintained in their home cages for 1 week to allow tdTomato expression before treatment.
CLARITY-based brain clearing
A hydrogel based on 1% acrylamide (1% acrylamide, 0.125% Bis, 4% PFA, 0.025% (w/v) VA-044 initiator, in 1× PBS) was used for all CLARITY preparations. Following transcardial perfusion with 4% PFA and postfixation, brainstems were transferred to 1% hydrogel for 48 h to allow monomer diffusion. Samples were degassed and polymerized for 4–5 h at 37 °C. Samples were washed with 200 mM NaOH-boric buffer (pH 8.5) containing 8% SDS for 6–12 h and then transferred to a flow-assisted clearing device using a temperature-control circulator. Next, 100 mM Tris-boric buffer (pH 8.5) containing 8% SDS was used to accelerate clearing, after which samples were washed in PBS + 0.1% Triton X for at least 24 h at 37 °C. Samples were incubated in a refractive index-matching solution (refractive index = 1.45) for 8 h at 37 °C and then 6–8 h at room temperature before confocal imaging.
Brainstem stereotaxic injection
Fully anaesthetized mice were placed in a stereotaxic frame with the head angled at 45°. A midline incision was made though the animal’s skin, posterior neck muscles and dura mater were pulled to expose the medulla between the occipital bone and C1 vertebra. Based on the stereotaxic coordinates of mouse brain31 and using obex as a reference point, injections were made into either bilateral nTS (0.1 mm rostral to obex, 0.2 mm lateral to midline, 0.25 mm under the medullary surface), bilateral NA (0.65 mm caudal to obex, 1.25 mm lateral to midline, 0.45 mm under the medullary surface) or bilateral DMV (0.05 mm caudal to obex, 0.1 mm lateral to midline, 0.1 mm under the medullary surface) using a calibrated glass micropipette attached to a Nanoject II injector (Drummond) and microprocessor pump (Pneumatic PicoPump, WPI). Each injection lasted no less than 10 min. Following injection, the glass micropipette was left in place for an additional 10 min before slow withdrawal. DTX (2 ng 200 nl−1), anti-DBH–SAP (42 ng 200 nl−1, advanced targeting system, no. IT-03), SSP–SAP (3.25 ng 200 nl−1, advanced targeting system, no. IT-11), blank–SAP (advanced targeting system, no. IT-21) or virus (AAV2/9-flex-tdTomato, 1.3 × 1013 genome copies (gc) ml−1, AAV2/8-flex-hM4D(Gi)-mCherry, 2.07 × 1013 gc ml−1, AAV2/8-DIO-hM3D(Gq)-mCherry, 8 × 1012 gc ml−1, AAV2/8-CMV-flex-TVA-mCherry-2A-oG, 1.39 × 1013 gc ml−1, EnvA G-Deleted Rabies-EGFP, 5.0 × 107 gc ml−1; Boston Children’s Hospital Vector Core and Salk GT3 core) were used. Blank–SAP cannot enter cells and is thus non-toxic to neurons, serving as the appropriate control for DBH–SAP or SSP–SAP administration. DBH–SAP was previously validated as specifically ablating DBH+ neurons with no effect on neighbouring neurons18,32,33. Because a period of 2 weeks is necessary to eliminate DBH+ neurons using DBH–SAP32, mice were injected with SAP 1 week before the first sensitization and 2 weeks before the second challenge.
Chemogenetic manipulation
Either CNO (Sigma, no. C0832, 1 mg kg−1, dissolved in 0.9% NaCl) or vehicle (0.9% NaCl) was injected intraperitoneally following expression of hM3D(Gq), hM4D(Gi) or mCherry in bilateral nTS or NA. The CNO concentration of 1 mg kg−1 used was effective34,35,36 and without apparent non-specific effects37,38. For hM4D(Gi) inhibition, CNO was injected 1 h before every of the second to fourth challenge; for hM3D(Gq) activation, CNO was injected in place of the fourth HDM challenge. Mice were euthanized 24 h later for flexiVent assay.
Airway hyperreactivity assayed by flexiVent
Anaesthetized mice were paralysed with acepromazine (10 mg kg−1, intraperitoneal injection). Mice were tracheotomized with a 20 G sterile catheter and attached to a flexiVent pulmonary mechanics apparatus (SCIREQ). Mice were ventilated at 9 ml kg−1 tidal volume and a frequency of 150 beats min−1. The weight of each animal was entered into flexiVent at the start of each round of assay. Pre-scans were carried out as part of the flexiVent programme, allowing for the calculation of lung size. Positive end-expiratory pressure was set at 300 mm H2O. The nebulizer was activated for 10 s to deliver each dose of methacholine (0, 6, 12 or 24 mg ml−1, dissolved in 0.9% NaCl). The Rrs and Ers of the respiratory system were determined in response to aerosolized methacholine challenges, and the mean maximal elastance and resistance of 12 measurements by dose were then calculated. Statistical analysis at each methacholine concentration was performed separately.
Tissue processing and flow cytometry
Anaesthetized mice were injected with AF700-counjugated CD45 (BioLegend, no. 103128, 10 µg per mouse) by intravenous injection to distinguish circulating immune cells and resident immune cells within the lung. Mice were euthanized 5 min later for lung harvest. Whole lungs were mechanically dissociated in GentleMACS C tubes (Miltenyi Biotec) containing 5 ml of RPMI 1640 (Thermo Scientific) with 10% fetal bovine serum, 1 mM HEPES (Life Technology), 1 mM MgCl2 (Life Technology), 1 mM CaCl2 (Sigma), 0.525 mg ml−1 collagenase/dispase (Roche) and 0.25 mg DNase I (Roche) by running the mouse lung 1-2 program on GentleMACS (Miltenyi Biotec). Lung pieces were then digested by shaking at around 150 rpm for 30 min at 37 °C. Following incubation, lung pieces were mechanically dissociated further using the mouse lung 2-1 program on GentleMACS, followed by straining through a 70 μm filter. Red blood cells were removed by the addition of 1 ml of RBC lysis buffer (BioLegend) to each tube and incubation at room temperature for 1 min. Single-cell suspensions were pelleted (1,500 rpm, 4 °C, 5 min), counted with a haemocytometer and diluted to around 1 × 106 cells ml−1. Diluted cells were stained with Fc blocking antibody (5 mg ml−1, BD) before incubation with a surface marker antibody cocktail. For lung myeloid tissue, the following antibodies were used: 1:100 BV605-conjugated anti-F4/80 (BioLegend, no. 123133), 1:500 BV510-conjugated anti-CD45 (BioLegend, no. 110741), 1:1,000 APC-conjugated anti-CD11c (BioLegend, no. 117310), 1:1,000 PE-Cy7-conjugated anti-Ly6G (BioLegend, no. 560601) and 1:2,000 PE-CF594-conjugated anti-CD11b (BioLegend, no. 101256). For lung lymphoid tissue, the following antibodies were used: 1:200 FITC-conjugated anti-CD45 (BioLegend, no. 103108), 1:100 APC-Cy7-conjugated anti-IL-7Ra (BioLegend, no. 135040), V450-conjugated Lineage mix (1:200 anti-CD19, TONBO, no. 50-201-4944), 1:500 anti-CD11c (TONBO, no. 50-201-4937), 1:500 anti-F4/80 (TONBO, no. 50-201-4978), 1:100 anti-NK1.1 (BD, no. 560524), 1:100 anti-TER119 (BD, no. 560504), 1:100 anti-TCR gamma delta (Invitrogen, no. 48-5711-82), 1:100 BV510-conjugated anti-ST2 (BD, no. 745080), 1:200 PE-Cy7-conjugated anti-TCR-beta (BioLegend, no. 109222), 1:100 BV604-conjugated anti-CD4 (BioLegend, no. 100548) and 1:2,000 PerCP-Cy5.5-conjugated anti-CD90.2 (BioLegend, no. 105338). Cells were then stained using live/dead dye (1:1,000, Ghost Dye Red 780 (TONBO, no. 13-0865-T100 for myeloid tissue, 1:500 and Ghost Dye Violet 450, TONBO, no. 13-0863-T100)) before fixing using BD Stabilizing Fixative and transfer to fluorescent activated cell sorting tubes. Flow cytometry was analysed on a BD FACS Canto RUO – ORANGE analyser with three lasers (405, 488 and 640 nm) at the Flow Cytometry Core at VA San Diego Health Care System and San Diego Veterans Medical Research Foundation. All data were further analysed and plotted with FlowJo software (Tree Star). Eosinophils, group 2 innate lymphoid cells and T-helper 2 cells were gated on live, resident CD45+ singlets.
Isolation of nTS and NA nuclei
Mice were euthanized using CO2 inhalation. For nTS, brainstems were acutely harvested from either (1) 4 groups of adult wild-type naive mice (n = 4 biological repeats, two males in each group); (2) mice at 1.5 h following the fourth saline treatment (2 males in the group); or (3) mice at 1.5 h following the fourth HDM challenge (2 males in the group). The nTS was visualized by microscopy and harvested based on anatomical landmarks. The reason for using males was based on our observation that males show less Fos background in saline control groups compared with females, providing a more consistent baseline for our study. Similarly, male mice were used in multiple, recently published single-cell RNA-seq datasets14,15,27,39. For NA, brainstems were acutely harvested from adult Chat–cre; CAG-Sun1/sfGFP mice (n = 7, 4 males and 3 females). NA was identified by nucleus-localized GFP fluorescence signals based on anatomical landmarks while avoiding DMV and 12N regions that also express Chat. NA samples from 7 mice were pooled for the snRNA-seq experiment. One pooled sample was assayed similarly in the other two studies with single-cell/-nucleus RNA-seq of the NA12,13. Dissected tissues were placed in liquid N2 immediately and either stored at −80 °C or sent directly for nuclei isolation.
On the day of nuclei dissociation, dissected tissues were transferred into 1 ml of douncing buffer (0.25 M sucrose, 25 mM KCl, 5 mM MgCl2, 10 mM Tris-HCl pH 7.5, 1 mM DTT (no. D9779, Sigma) and 1× cOmplete EDTA-free protease inhibitor (no. 05056489001, Roche, DB-DP, 0.1% Triton-X)). Pestled samples were filtered with 30 μm CellTrics and transferred to prechilled low-bind Eppendorf tubes. Samples were spun and sequentially resuspended in douncing buffer, permeabilization buffer (5% IGEPAL-CA630, no. I8896, Sigma, 0.2% DTT, 1 mM cOmplete EDTA-free protease inhibitor and 1× PBS) and tagmentation buffer (66 mM Tris-acetate pH 7.8 (no. BP-152, Thermo Fisher Scientific)), 132 mM K-acetate (no. P5708, Sigma), 20 mM Mg-acetate (no. M2545, Sigma) and 32 mM DMF (no. DX1730, Millipore) and counted using a haemocytometer.
snRNA-seq and data analysis
Single-nucleus RNA sequencing experiments were carried out by the Center for Epigenomics, UCSD. Nuclei were processed into complementary DNA libraries using the Chromium Single Cell 3’ v3 kit (10X Genomics) and sequencing was carried out on the NovaSeq (Illumina) platform. The CellRanger software package from 10X Genomics (v3.0.2) was used to align raw reads onto the mouse reference genome (GRCm38) and generate the feature-barcode matrix. CellBender (v0.3.0)40 was then used to remove technical artefacts and ambient RNA to produce improved estimates of gene expression. The R package Seurat (v4.0)29 was then used to perform data quality control, normalization, principal components analysis, UMAP generation and differential gene expression testing. Nuclei with above 5% mitochondrial reads and greater than 2,000 unique genes were considered high-quality cells and were filtered for further analyses, following the filtering criteria commonly used in neuronal snRNA-seq studies including those on brainstem neurons12,13,14. In addition, DoubletFinder (v2.0)41 was used to remove doublets and SCTransform was used to normalize feature expression. Harmony (v1.2.0)42 was used to integrate individual datasets across three conditions (naive, saline and HDM). In total 42,157 nuclei were recovered, including 39,626 neurons. We then extracted nTS data from baseline naive condition (n = 4 biological repeats, n = 2 mice in each group) to profile marker gene expression across all nTS clusters. To determine dimensions for optimized clustering, we tested up to 50 principal components and evaluated the optimal cutoff using an elbow plot. In our study, we settled on using the first 20 principal components for clustering and projection with both UMAP and t-distributed stochastic neighbour embedding. We also tested a range of clustering resolutions (from 0.1 to 2.0) that were evaluated with clustree (v0.5.1; https://github.com/lazappi/clustree)43. In this dataset the resolution was set to 1.0, resulting in 25 interim clusters. We plotted a density UMAP using geom_density_2d and stat_density_2d (https://ggplot2.tidyverse.org/reference/geom_density_2d.html) from ggplot2 (v3.3.2)44 for visual identification of high-density regions that represent potential unique cell populations. Using these two methods, coupled with manual inspection of top markers, we combined several clusters with shared markers to ensure that annotated clusters would show unique transcriptional profiles, resulting in 18 distinct clusters (further details in Supplementary Note 1, Supplementary Fig. 2a–i and Supplementary Tables 1 and 2). Following combination, we reordered cluster ID based on the number of cells in each cluster and renumbered the largest cluster as cluster 1. For rigorous definition of marker genes for each cluster, we screened each cluster’s top marker genes (Seurat, FindAllMarkers) using ViolinPlot, FeaturePlot and DotPlot. Using the Seurat default dotplot setting (R package Seurat (v4.0)29, percentage expressed was plotted based on the actual percentage of cells expressing selected marker genes in a given cluster. Average expression was plotted based on the average normalized single-cell expression value, using the Seurat default dotplot setting (R package Seurat v4.0)29, with maximum average expression threshold set at 2.5 and everything higher set to this; and with minimum average expression threshold set at −2.5 and everything lower set to this. The colour bar shown on the right of the plot represents the range of scaled normalized expression values for the genes shown in that plot. We provide a list of the top 100 marker genes of nTS clusters in Supplementary Table 3.
For NA, 10,072 nuclei were recovered, including 7,664 neurons. Using Chat as a positive control gene to identify NA neurons, we removed clusters not showing Chat expression from ViolinPlot, FeaturePlot and DotPlot. We then followed the same pipeline given above to process our data and used Harmony (v1.2.0)42 to integrate our data with two published NA single-cell datasets12,13. We provide a list of the top 100 marker genes in Supplementary Table 4.
qPCR
Total RNA was extracted from lungs using Trizol (Invitrogen) and the RNeasy Mini RNA extraction kit (Qiagen). qPCR with reverse transcription was then performed to obtain corresponding cDNA using THE iScript Select cDNA Synthesis Kit (Bio-Rad). qPCR was performed with the CFX ConnectTM system (Bio-Rad) using SYBR Green (Bio-Rad). At least three technical and three biological replicates were performed for each gene. Primer sequence: 5′-CGGCCAGGTCATCACTATTGGCAAC-3′, 5′-GCCACAGGATTCCATACCCAAGAAG-3′ for Actb (β-actin); 5′-TGACTCAATCTGCGTGCCTT-3′, 5′-AGGCCTTCTTTTGGCAGGTT-3′ for Muc5ac; 5′-GGTCTCAACCCCCAGCTAGT-3′, 5′-GCCGATGATCTCTCTCAAGTGAT-3′ for Il4; 5′-CCTCTTCGTTGCATCAGGGT-3′, 5′-GATCCTCCTGCGTCCATCTG-3′ for Il5; 5′-AAAGCAACTGTTTCGCCACG-3′, 5′-CCTCTCCCCAGCAAAGTCTG-3′ for Il13.
Plethysmography
Ventilatory parameters during normoxia was measured in unrestrained male mice using a whole-body barometric plethysmograph modified for continuous flow45,46,47. Flow was maintained constant through the chamber while a pressure transducer (45 mMP with 2 cm H2O diaphragm, Validyne) recorded changes attributable to the warming and expansion of inhaled gases. Mice were weighed and sealed into an individual plethysmograph chamber along with a temperature and humidity probe (Thermalert TH5, Physitemp). A constant gas flow (335 ml min−1) was delivered using a rotameter (no. 603, Matheson) and measured with a flow meter (Sables System International, Inc.) upstream of the chamber. Gases exited the chamber through a valve and into a vacuum pump (Model 25, Precision Scientific Co.) to isolate pressure changes from respiration in the chamber during constant flow with high input and output impedances. All ventilatory parameters were recorded on an analogue-digital acquisition system (PowerLab 8SP, AD Instruments) and analysed with LabChart 8-Pro Software, sampling at a rate of 1 kHz. Mice were allowed to acclimatize to the chamber and constant air flow for 45 min (normoxia, 21% of O2) and were then exposed to a 5 min challenge of hypoxic gas (10% O2) to test responsiveness; they were then exposed to normoxic gas for 15 min. A minimum of 30 s between 10 and 15 min of normoxia exposure was analysed. Respiratory frequency (breaths min−1) was measured from cyclic peaks in the plethysmograph pressure pulses, and tidal volume (ml) was measured from calibration pulses using equations from ref. 48. Minute ventilation (\({\dot{{\rm{V}}}}_{{\rm{i}}}\), ml min−1 kg) was calculated from respiratory frequency and tidal volume and normalized to body mass ventilation. Oxygen consumption \(({\dot{{\rm{V}}}}_{{{\rm{O}}}_{2}})\) and carbon dioxide production \(({\dot{{\rm{V}}}}_{{{\rm{CO}}}_{2}})\) were calculated by recording inspired and expired oxygen and carbon dioxide fractions using an O2/CO2 analyser. The ratio of \({\dot{{\rm{V}}}}_{{\rm{i}}}\,/{\dot{{\rm{V}}}}_{{{\rm{O}}}_{2}}\) to \({\dot{{\rm{V}}}}_{{\rm{i}}}\,/{\dot{{\rm{V}}}}_{{{\rm{CO}}}_{2}}\) thus provides a more precise estimation of mouse ventilation without the confounding factors produced by changes in metabolic rate.
CUBIC tissue clearing and LSFM
Clear, unobstructed brain/body imaging cocktails and computational analysis protocol (CUBIC) buffers were prepared accordingly49. Following sufficient tissue clearing in R1 buffer, tissues were embedded in 2% low-melting agarose and then incubated in R2 solution before imaging. Cleared samples were imaged using a Zeiss Z.1 light sheet fluorescence microscope (LSFM). Vagal ganglia were imaged using a ×5 objective (LSFM ×5/0.1 numerical aperture (NA)) and a 1.45 ×5 CLARITY specific chamber. Lung samples were imaged using a ×2.5 objective (LSFM ×2.5/0.1 NA) and a 1.45 ×2.5 CLARITY specific chamber (Translucence Biosystems).
CTB488 injection
Fully anaesthetized mice were placed ventral side up on a stereotaxic frame. The extrathoracic trachea was exposed via a neck incision. The trachea was carefully lifted without bleeding. Using a 10 μl Hamilton glass micropipette fitted with a 32 G needle, 10 μl of CTB488 (1 mg ml−1, Thermo Scientific, no. C22841) was injected into the dorsal aspect of the trachea from both sides, about three to five cartilage rings caudal to the larynx. Sutured mice were allowed to recover on a thermal pad before being returned to housing. Mice were euthanized 1 week following CTB488 injection for brainstem collection.
Cannula implantation at bilateral NA
Fully anaesthetized mice were mounted on a stereotaxic frame. A midline incision was made to show bregma, the skull was cleaned with hydrogen peroxide and small holes were drilled through the skull at the designated stereotaxic coordinates of NA (6.7 mm caudal to bregma, 1.25 mm lateral to midline, 4.25 mm under the dura)31. A bilateral guide cannula affixed with two units of 26 G stainless-steel tubing (P1 Technologies, Inc.) was stereotaxically implanted 0.5 mm above the NA region. Guide cannulae were secured to the skull using superglue and dental cement. A matching dummy cannula were inserted into the guide cannula and secured with a dust cap to ensure guide cannula patency. Mice were allowed to recover in their home cages for 1 week before challenge and drug delivery. On the day of flexiVent assay, starting from baseline measurement, 1 μl of prazosin (0.4 mg ml−1 in ultrapure water, no. P7791, Sigma), terazosin (0.4 mg ml−1 in ultrapure water, no. T4680, Sigma) or ultrapure water control was consecutively microinjected into bilateral NA through the guided cannula using a two-channel syringe pump (no. R462, RWD).
AAV infection of vagal ganglia
The vagal ganglia of anesthetized mice were surgically exposed by making an incision along the neck. A micropipette containing 200 nl of AAV2/9-syn-flex-GFP (2 × 1013 gc ml−1; Boston Children’s Hospital Vector Core) was inserted into bilateral vagal ganglia. Mice were euthanized 3 weeks later for harvesting of vagal ganglia and brainstem.
Statistics and reproducibility
Statistical analyses were calculated with Microsoft Excel and performed using Prism (GraphPad), with statistical tests and sample sizes reported in figure legends. Data in graphs are presented as mean ± s.e.m. and statistical tests are two-sided, unless otherwise indicated. All replicates were biological, unless otherwise indicated. All representative images are from at least three independent experiments, and details are described in figure legends. Sample sizes were determined based on previous expertise and publications in the field. Exact sample sizes are described in each figure legend. Investigators were blinded to group allocations for FOS antibody staining and flexiVent experiments associated with Figs. 1 and 3–6 and Extended Data Figs. 1–3 and 6–10; group allocation was not blinded in other experiments. Significance is defined as P < 0.05, with significance annotations of *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001. Absence of significant differences (P > 0.05) is indicated by NS (not significant).
Materials availability
All reagents and materials used in this study are commercially available.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.