Sex as a biological variable. Male mice were specifically used when studying response to HDM as previous studies in our lab established that there was variability in brain activity and respiratory response based on flexiVent measurements in female mice (4). In establishing the Ascl1-mutant model and collecting epithelial cells for scRNA-Seq, equal numbers of male and female mice were studied. Similar findings were found for both sexes.

Animals. Mouse lines used were tamoxifen-inducible, knockin Cre recombinase drivers Ascl1creERT2 (39), Scgb1a1creERT2 (40), and Tp63creERT2 and constitutively active reporter Rosa26ZsGreen (41) and Rosa26tdTomato (41). Mice were purchased from Jackson Laboratory, with the exception of Ascl1fl (42) mice that were gifts from Francois Guillemot’s lab (Francis Crick Institute, London, United Kingdom) and Shhcre mice from Clifford J. Tabin’s lab (Harvard Medical School, Boston, Masschusetts, USA). Genotyping was performed on ear clips utilizing oligonucleotide primers previously reported for each strain. Early postnatal stage and adult mice aged 2–3 months were used in all experiments, with similar numbers of males and females. Mice were maintained in 12-hour light/12-hour dark cycle with food and water provided ad libitum. All mice were housed in American Association for Accreditation of Laboratory Animal Care–accredited facilities and labs at UCSD. All animal husbandry, maintenance, and experiments were performed in accordance with UCSD’s IACUC-approved protocols.

Tamoxifen induction of Cre recombination. Tamoxifen stock solution (10 mg/mL, MilliporeSigma) was prepared by sonication in corn oil and stored at –20°C. Intraperitoneal injections at the dose of 25 mg/kg were administered every other day for the periods indicated in the figure scheme for the Ascl1 mutant (Figure 1A). In remaining reporter mice, adult mice (at 8 weeks old) were injected with tamoxifen twice within 48 hours to induce cre for the activation of the reporter.

Allergen challenge with HDM and Notch agonist. HDM (Dermatophagoides pteronyssinus, Greer Labs) stock solution (5 mg/mL) was prepared in saline solution and stored at –20°C. Mice were anesthetized with isoflurane prior to nasal administration of 50 μg (dissolved in 20 μL of saline) of HDM extract. HDM in adult mice (at 8 weeks old) was administered on days 0, 7, 14, and 21 (Figure 2A). For controls, 20 μL of saline was administered intranasally to adult mice in the same regimen. All mice were euthanized under IACUC guidelines 24 hours following the last HDM challenge for analysis.

For Notch agonist, Yhhu 3792 (Tocris 6599) powder was ground with 0.5% DMSO and 1% Tween and diluted into suspension with saline to approximately 50 μg. Yhhu 3792 was administered intranasally following HDM; 0.5% DMSO and 1% Tween was administered as the vehicle group without Yhhu 3972.

Immunohistochemistry. Mice were euthanized by CO2 asphyxiation. For cryosections, lungs were immediately dissected following inflation at 35 cm H2O airway pressure with 4% (v/v) paraformaldehyde (PFA; Electron Microscopy Sciences), prepared fresh by dilution of a 32% stock solution in phosphate-buffered saline (PBS). Inflated lungs were fixed overnight at 4°C and dehydrated in 30% (w/v) sucrose in PBS overnight before embedding in optimum cutting temperature (O.C.T.; Sakura) compound and storage at –80°C. For flash-frozen sections, lungs were inflated with 20% O.C.T. compound and immediately embedded in O.C.T. compound before flash-freezing in liquid nitrogen. Tracheal samples were dissected and mounted similarly, without inflation.

Coronal sections (10 μm and 30 μm) were prepared using a cryostat, and cryosections were mounted to Superfrost Plus slides (Fisherbrand, Thermo Fisher Scientific). Sections were dried at room temperature for 15–60 minutes before washing twice for 5 minutes with PBS, postfixing with 4% PFA, and washing twice for 5 minutes with 0.1% (v/v) Tween 20 in PBS. Sections were incubated for 1.5 hours at room temperature in 5% (v/v) normal goat serum in 0.1% Tween 20 with PBS (blocking solution), then incubated overnight at 4°C with primary antibodies in blocking solution in a humid chamber. The following day, sections were washed 3 times for 10 minutes each in 0.1% Tween 20 PBS, then incubated for 1 hour at room temperature in blocking solution containing Alexa Fluor–conjugated (AF-conjugated) secondary antibodies and DAPI. Stained sections were finally washed 3 times for 5 minutes each in 0.1% Tween 20 PBS and mounted on coverslips with Fluoromount-G (SouthernBiotech) as the mounting medium. Mounted specimens were stored at 4°C until microscopy (ZEISS Axio Imager 2).

The following primary antibodies were used: mouse anti-MUC5AC monoclonal antibody (5 mg/mL, MilliporeSigma MAB2011), mouse anti-CGRP (human) monoclonal antibody (2 mg/mL, Lifespan Biosciences ab81887), mouse anti-TUJ1 monoclonal antibody (1 mg/mL, MilliporeSigma BAM1195), mouse anti-FOXJ1 monoclonal antibody (0.5 mg/mL, Invitrogen 5013931), rabbit anti-CGRP polyclonal antibody (2 mg/mL, MilliporeSigma C8198), rabbit anti-SYP polyclonal antibody (5 mg/mL, Thermo Fisher Scientific RB-1461-P1), rabbit anti-SCGB1A1 polyclonal antibody (5 mg/mL, Seven Hills Bioreagents WRAB-3950), rabbit anti-MUC5B polyclonal antibody (0.23 mg/mL, Cloud Clone Corp PAA684Mu01), rabbit anti-AGR2 monoclonal antibody (1 mg/mL, Cell Signaling Technology 130625), rabbit anti-LAMP3 monoclonal antibody (1 mg/mL, Synaptic Systems 391 005), rabbit anti-NRN1 polyclonal antibody (1 mg/mL, Invitrogen PA5-20368), and guinea pig anti-SYT1 monoclonal antibody (1 mg/mL, Synaptic Systems 105 318). The following secondary antibodies were used: Cy3-conjugated goat anti-mouse IgG (2 mg/mL, Jackson ImmunoResearch Labs 2338680), Cy5-conjugated goat anti-mouse AffiniPure IgG (2 mg/mL, Jackson ImmunoResearch Labs 2338714), FITC-conjugated goat anti-rabbit (2 mg/mL, Jackson ImmunoResearch Labs 14165), Cy3-conjugated goat anti-rabbit IgG (2 mg/mL, Jackson ImmunoResearch Labs 2338000), and AF488-conjugated goat anti–guinea pig (2 mg/mL, Invitrogen 2337415).

For NICD immunostaining of cryosections, tyramide signal amplification was performed using the detection kit (Invitrogen). Sections were stained with rabbit anti-NICD primary antibody (2 mg/mL, Cell Signaling Technology 4147S), washed, and incubated with goat anti-rabbit biotin secondary (Invitrogen 65-6140) followed by incubation with horseradish peroxidase streptavidin. Following washes, the sections underwent incubation with AF594-conjugated tyramide (prepared according to manufacturer’s instructions) (Invitrogen B40925) for 5 minutes at room temperature. Slides were then washed, mounted, and imaged.

RNAscope. Adult fresh-frozen sections from control and HDM-challenged mice were stained with probes from mouse Robo1 (ACD 475951-C1) and Il-33 (ACD 400591-C2) using the RNAscope Multiplex Fluoromount Reagent Kit v2 (ACD 323100). Staining was performed using manufacturer’s instructions for fresh-frozen sections. Sections were counterstained with CGRP antibody to detect the presence of PNECs.

Microscopy and imaging. Cryosections (10–30 μm) were imaged using a ZEISS Axio Imager 2/LSM 880 at 5×, 10×, 20×, or 40× objective. Remaining images were taken with ZEISS laser scanning confocal microscope with an inverted 40× or 63× oil immersion objective. Sections were selected to best match the representative field and to cover different depths, if necessary.

Transcriptional analysis (qRT-PCR). Dissected lungs were stored in TRIzol (Invitrogen) at –80°C until further processing. Total RNA was extracted using instructions from the RNeasy Mini RNA extraction kit (QIAGEN). cDNA was generated using the iScript Select cDNA Synthesis Kit (Bio-Rad). qRT-PCRs were performed using gene-specific primers and iTaq SYBR Green (Bio-Rad) by CFX Connect system (Bio-Rad). At least 3 technical and 3 biological replicates were performed for each gene, if not otherwise noted. Standard quantified values were normalized to the β-actin housekeeping genes, and relative changes in gene expression were calculated using the ΔΔCT method. All primer sequences used for qRT-PCR analysis are listed: Calca forward: 5′ CCTTTCCTGGTTGTCAGCATCTTG 3′, Calca reverse: 5′ CTGGGCTGCTTTCCAAGATTGAC 3′, Ascl1 forward: 5′ CTCCCCATTTGACGTCGTTG 3′, Ascl1 reverse: 5′ CTACGACCCTCTTAGCCCAG 3′, Tac1 forward: 5′ GGTCCGACAGTGACCAGATCAAG 3′, Tac1 reverse: 5′ AAAGAACTGCTGAGGCTTGGGTC 3′, Muc5ac forward: 5′ TGACTCATCTGCGTGCCTT 3′, Muc5ac reverse: 5′ AGGCCTTCTTTTGGCAAGGTT 3′, Scgb1a1 forward: 5′ ATGAAGATCGCCATCACAATCAC 3′, Scgb1a1 reverse: 5′ GGATGCCACATAACCAGACTCT 3′, Foxj1 forward: 5′ CTCCTATGCCACTCTCATCTGC 3′, Foxj1 reverse: 5′ GACAGGTTGTGGCGGATGGAAT 3′, β-actin forward: 5′ CGGCCAGGTCATCACTATTGGCAAC 3′, and β-actin reverse: 5′ GCCACAGGATTCCATACCCAAGAAG 3′.

Tissue processing and flow cytometry. Mice were anesthetized and administered an intravenous dose of AF700-conjugated CD45 (BioLegend, no. 103128; 10 μg per mouse) to discriminate between circulating and lung-resident immune cells. Five minutes later, the mice were euthanized, and their lungs were dissected. Each lung was mechanically dissociated in GentleMACS C tubes (Miltenyi Biotec) containing 5 mL RPMI 1640 (Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (FBS), 1 mM HEPES (Life Technologies), 1 mM MgCl2 (Life Technologies), 1 mM CaCl2 (MilliporeSigma), 0.525 mg/mL collagenase/dispase (Roche), and 0.25 mg DNase I (Roche). The GentleMACS “mouse lung 1-2” program (Miltenyi Biotec) was used for initial dissociation. Samples were then incubated at 37°C with shaking (~150 rpm) for 30 minutes. Following incubation, lungs were further dissociated using the GentleMACS “mouse lung 2-1” program and strained through a 70 μm filter (Miltenyi Biotec). Red blood cells (RBC) were lysed by adding 1 mL of RBC lysis buffer (BioLegend) for 1 minute at room temperature. The remaining cells were pelleted (1,500 rpm, 4°C, 5 minutes), counted with a hemocytometer (Thermo Fisher Scientific), and diluted to approximately 1 × 106 cells/mL. Cells were treated with Fc-block (5 mg/mL, BD Biosciences) before staining with a cocktail of surface marker antibodies.

For lung myeloid cells, 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 cells, 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), 1:200 V450-conjugated lineage mix (anti-CD19 [TONBO, no. 50-201-4944], anti-CD11c [TONBO, no. 50-201-4937], anti-F4/80 [TONBO, no. 50-201-4978], anti-NK1.1 [BD Biosciences, no. 560524], anti-TER119 [BD Biosciences, no. 560504], anti-TCR γδ [Invitrogen, no. 48-5711-82]), 1:100 BV510-conjugated anti-ST2 (BD Biosciences, no. 745080), 1:200 PE-Cy7-conjugated anti–TCR-β (BioLegend, no. 109222), 1:100 BV604-conjugated anti-CD4 (BioLegend, no. 100548), anti–cleaved caspase-3 (Cell Signaling Technology, Asp175, no. 9661), anti-Ki67 (Cell Signaling Technology D3B5, no. 9129), and 1:2,000 PerCP-Cy5.5-conjugated anti-CD90.2 (BioLegend, no. 105338). Cells were labeled with live/dead dye (Ghost Dye Red 780, TONBO, no. 13-0865-T100 or Ghost Dye Violet 450, TONBO, no. 13-0863-T100), then fixed using BD Biosciences Stabilizing Fixative. Samples were analyzed using a BD Biosciences FACSCanto RUO – ORANGE flow cytometer (405, 488, and 640 nm lasers) at the VA San Diego Health Care System Flow Cytometry Core. Data were further analyzed and visualized with FlowJo (Tree Star). Eosinophils, ILC2s, and Th2 cells were gated on live, resident CD45+ singlets.

Airway hyperreactivity assayed by flexiVent. For AHR measurements, mice were anesthetized and given acepromazine (10 mg/kg, intraperitoneal). They were then tracheotomized with a 20 G sterile catheter, connected to a flexiVent pulmonary mechanics apparatus (SCIREQ), and ventilated at a tidal volume of 9 mL/kg with a frequency of 150 breaths/min. The weight of each mouse was entered into flexiVent at the beginning of each measurement series. Positive end-expiratory pressure was maintained at 300 mm H2O. A 10-second nebulization was performed for each methacholine (Mch) dose (0, 6, 12, or 24 mg/mL in 0.9% NaCl). The resistance (Rrs) and elastance (Ers) of the respiratory system were determined in response to each Mch challenge, and the mean maximal Rrs and Ers across 12 measurements per dose were calculated. Statistical comparisons at each Mch concentration were carried out separately.

PNEC and epithelial cell enrichment by FACS. To enrich live PNECs for scRNA-Seq profiling, adult Ascl1creERT2 Rosa26tdTomato/+ mice (roughly 2- to 3-month old male and female littermates) were administered 2 doses of i.p. injection of tamoxifen at P12 and P14, then underwent HDM regimen over the course of 4 weeks from 8 weeks old. Mice were euthanized by CO2 asphyxiation 24 hours following the last HDM challenge and perfused with approximately 10 mL of PBS into the right cardiac ventricle.

Lungs were dissected and pooled (6 lungs per collection round) to ensure sufficient numbers of labeled PNECs for sorting. Whole lungs, excluding the trachea, were dissociated by cutting them into small pieces with surgical scissors, then enzymatically digested in RPMI containing 1 mg/mL collagenase type 4 (Roche), 5 U/mL dispase (Thermo Fisher Scientific), and 20 μg/mL DNase I (Roche) for 30 minutes at 37°C in a bacterial shaker. Following digestion, lung tissue was pressed through a 70 mm cell strainer, and enzymes were quenched with an excess volume of RPMI containing 10% (v/v) FBS. Samples were spun down at 1,500 rpm for 5 minutes at 4°C and incubated in RBC lysis buffer (BioLegend) for 3 minutes on ice. Five-minute spins continued to be performed between each step, and cell-free supernatants were removed through vacuum aspiration. RBC-lysed cell suspensions were depleted for endothelial cells and leukocytes using anti-CD31– (0.5 mg/mL, BioLegend, 102407) and anti-CD45– (0.5 mg/mL, BioLegend, 103102) coated magnetic microbeads and MACS separation columns, according to the manufacturer’s instructions (EasySep, STEMCELL Technologies). PBS with 2% FBS (FACS buffer) was used as the suspension solution for magnetic labeling and for cell application onto columns. Endothelial/leukocyte-depleted cell suspensions were stained with allophycocyanin-conjugated (APC-conjugated) anti-CD326 (EpCAM G8.8, 0.5 mg/mL, Invitrogen) at 1:200 dilution in FACS buffer. Stained cell suspensions were washed twice for 5 minutes each in FACS buffer to remove residual unbound antibodies and resuspended in FACS buffer containing 1 μg/mL DAPI as a live-dead cell indicator before proceeding to FACS.

Cell sorting was performed in the Sanford Consortium FACS facility using a BD Biosciences Influx high-speed sorter flow cytometer with a 100 μm nozzle. Unstained lung cell suspensions were used as negative gating controls for each fluorophore. Approximately 300 DAPI–APC+tdTomato+ live PNECs were collected into FACS buffer, and 25,000 DAPI–APC+tdTomato– live non-neuroendocrine epithelial cells were also collected. Data were analyzed and plotted using FlowJo v10.

scRNA-Seq. Following FACS, sorted cells were counted and processed using Chromium Single Cell 3′ v3 kit (10x Genomics). cDNA quality was checked (high quality) prior to submission. Sequencing was carried out on the NovaSeq 6000 (Illumina) platform at the Institute for Genomic Medicine, UCSD. Cell Ranger package v3.0.2 (43) (10x Genomics) was used to align the raw reads onto the mouse reference genome (GRCm38) and generate the feature-barcode matrix.

R package Seurat v4.1.3 was used to perform data quality control, normalization, principal components analysis, and UMAP. Cells with fewer than 200 or more than 2,500 unique feature counts, or more than 7% mitochondrial counts, were considered “low-quality” cells and thus removed from further analysis. Doublets were removed using “doubletFinder” (44, 45). Global-scaling method “LogNormalize” was used to normalize the feature expression. A total of 2,000 top variable features were identified by the function FindVariableFeatures and selected for subsequent principal components analysis. The top 20–50 significant genes were chosen to conduct cell clustering by using the UMAP algorithm with default settings. The expression level and feature of selected genes were profiled and visualized by R package ggplot2 v4.1.3. Populations of identified immune cells, endothelial cells, erythroid cells, and remnant cells that had no distinctive cluster properties were removed from analysis to focus solely on airway epithelial cells.

Statistics. Cell numbers were manually counted from images taken from each section. An average of 15–25 sections from the same lung were quantified for each individual sample. As indicated in the figure legends, 1-tailed Student’s t test, 1-way and 2-way ANOVA, and Tukey’s multiple comparisons test were used on the data sets. Results were presented as mean ± SD. All statistical analyses were performed using Prism version 6 software (GraphPad). A P value of less than 0.05 was considered statistically significant. Fluorescence intensity was quantified using mean intensity density over the total area (μm2) on ImageJ (NIH) for Notch1 intensity levels.

Study approval. The animal care and experiments were performed in compliance with institutional guidelines that were reviewed and approved by the UCSD Animal Care and Use Committee and conformed to the NIH Guide for the Care and Use of Laboratory Animals (National Academies Press, 2011). All animals were handled according to approved Institutional Animal Care and Use Committee protocols (S16187) at UCSD.

Data availability. Values for all data points for each graph are included in the Supporting Data Values file. The scRNA-Seq data (raw data and annotated data) are archived in National Center for Biotechnology Information Gene Expression Omnibus (GSE273687). All data, reagents, and models are available upon request made to the corresponding author.